(Routledge Studies in Metaphysics) William M.R. Simpson, Robert C. Koons, James Orr - Neo-Aristotelian Metaphysics and The Theology of Nature-Routledge (2021)

“This book ofers a fresh, bold, stimulating take on some of the hardest
questions philosophers and theologians alike are engaged with, arguing

that only a metaphysics developed on the basis of the Abrahamic-
Aristotelian tradition can enable us to understand today’s world of science,
the relation of God to it, and our place within it. A hugely enjoyable
read – even for those who will fnd themselves in disagreement.”
– Prof. Anna Marmodoro, Professor of
Metaphysics, Durham, UK
“Neo-Aristotelianism is one of the important movements in contemporary

metaphysics: this collection provides a wide-ranging sample of the state
of the art from its outstanding exponents and shows how it can also be
an important force in theology.”
– Prof. John Marenbon, Professor of Medieval
Philosophy, Cambridge, UK
“While scholasticism has been on the margins of Christian theology for

some decades, this volume suggests the possibility of a powerful and
fruitful intellectual revival. The talented authors of this conceptually well-
organized volume orchestrate the use of scholastic principles to analyze
some of the most important and difcult questions in contemporary
theology, with both clarity and depth. In doing so they delineate a
compelling vision of harmony between revealed theology, philosophy,
and modern science. A wonderful book!”
– Prof. Thomas Joseph White, OP, Rector Magnifcus of the
Pontifcal University of St Thomas, Angelicum, Rome, Italy
“Despite its long neglect in mainstream philosophy of science, neo-

Aristotelian metaphysics has enormously interesting and creative
applications for contemporary science, as this new collection of essays
amply demonstrates. But perhaps even more importantly, the question of
God can then be manoeuvred back into contestation in a way occluded
or outlawed by other metaphysical alternatives. This important new
book brings together senior and junior scholars to help carve a new
trajectory through long-disputed territory.”
– Prof. Sarah Coakley, FBA, Norris-Hulse
Professor of Divinity Emerita, University of
Cambridge, and Honorary Fellow,
Oriel College, Oxford, UK
“When the Scientifc Revolutionaries of the seventeenth century left

their Aristotelian homeland for a far country, little did they realise that
their distant descendants might need to return one day. I have heard
much whispered about the promise of neo-Aristotelianism for the
science-and-theology enterprise, but until seeing this volume had never
been able to grasp the vision. This book will surely be an indispensible
guide for years to come, providing lucid overviews of contemporary
historical, philosophical, scientifc and theological perspectives on neo-
Aristotelianism, as well as much original research looking forward. ”
– Prof. Mark Harris, Professor of Natural Science
and Theology, Edinburgh, UK
Neo-Aristotelian Metaphysics and
the Theology of Nature
This book explores the relationship between a scientifcally updated Aristotelian

philosophy of nature and a scientifcally engaged theology of nature. It features
original contributions by some of the best scholars engaging with Aristotelianism
in contemporary metaphysics, philosophy of science, and philosophical theology.
Despite the growing interest in Aristotelian approaches to contemporary
philosophy of science, few metaphysicians have engaged directly with the question
of how a neo-Aristotelian metaphysics of nature might change the landscape for
theological discussion concerning theology and naturalism, the place of human
beings within nature, or the problem of divine causality. The chapters in this
volume are collected into three thematic sections: Naturalism and Nature, Mind
and Nature, and God and Nature. By pushing the current boundaries of neo-
Aristotelian metaphysics to recover the traditional notion of substantial forms in
physics, reframe the principle of proportionality in biology, and restore the hierarchy
of being familiar to ancient philosophy, this book advances a metaphysically
unifed framework that accommodates both scientifc and theological knowledge,
enriching the interaction between science, philosophy, and theology.
Neo-Aristotelian Metaphysics and the Theology of Nature will be of interest to
scholars and advanced students working in metaphysics, philosophy of science,
natural theology, philosophical theology, and analytic theology.
William M. R. Simpson is Junior Research Fellow at Wolfson College, Cambridge,

and Honorary Research Fellow in the Philosophy of Nature at the University of
St Andrews. He is the author of Surprises in Theoretical Casimir Physics (2014)
and the co-editor, with Robert C. Koons and Nicholas J. Teh, of Neo-Aristotelian
Perspectives on Contemporary Science (Routledge, 2017).
Robert C. Koons is Professor of Philosophy at the University of Texas at Austin.

He is co-editor with George Bealer of The Waning of Materialism: New Essays in
the Philosophy of Mind (2010) and with William M. R. Simpson and Nicholas J.
Teh of Neo-Aristotelian Perspectives on Contemporary Science (Routledge, 2017).
He is also the co-author of Metaphysics: The Fundamentals (2015).
James Orr is University Lecturer in Philosophy of Religion at the University of

Cambridge, and formerly the McDonald Postdoctoral Fellow in Theology, Ethics,
and Public Life at Christ Church, Oxford. He is a trustee of St. Paul’s Theological
Centre, London, and a member of the Archbishop of Canterbury’s advisory group,
the Lambeth Partnership.
Routledge Studies in Metaphysics
Fundamental Causation
Physics, Metaphysics, and the Deep Structure of the World
Christopher Gregory Weaver
Eliminativism, Objects, and Persons

The Virtues of Non-Existence
Jiri Benovsky
Thisness Presentism
An Essay on Time, Truth, and Ontology
David Ingram
Metametaphysics and the Sciences

Historical and Philosophical Perspectives
Edited by Frode Kjosavik and Camilla Serck-Hanssen
Authority and the Metaphysics of Political Communities

Gabriele De Anna
Language and Word

A Defence of Linguistic Idealism
Richard Gaskin
Neo-Aristotelian Perspectives on Formal Causation

Edited by Ludger Jansen and Petter Sandstad
A Powerful Particulars View of Causation

R. D. Ingthorsson
Neo-Aristotelian Metaphysics and the Theology of Nature

Edited by William M. R. Simpson, Robert C. Koons, and James Orr
For more information about this series, please visit: www.routledge.com/

Routledge-Studies-in-Metaphysics/book-series/RSM
Neo-Aristotelian Metaphysics
and the Theology of Nature
Edited by William M. R. Simpson,

Robert C. Koons, and James Orr
First published 2022
by Routledge
605 Third Avenue, New York, NY 10158
and by Routledge
2 Park Square, Milton Park, Abingdon, Oxon, OX14 4RN
Routledge is an imprint of the Taylor & Francis Group, an informa business
© 2022 selection and editorial matter, William M. R. Simpson, Robert C.
Koons, and James Orr; individual chapters, the contributors
The right of William M. R. Simpson, Robert C. Koons, and James Orr to
be identifed as the authors of the editorial material, and of the authors for
their individual chapters, has been asserted in accordance with sections 77
and 78 of the Copyright, Designs and Patents Act 1988.
With the exception of Chapters 1 and 2, no part of this book may be
reprinted or reproduced or utilised in any form or by any electronic,
mechanical, or other means, now known or hereafter invented, including
photocopying and recording, or in any information storage or retrieval
system, without permission in writing from the publishers.
Chapters 1 and 2 of this book are available for free in PDF format as
Open Access from the individual product page at www.routledge.com.
They have been made available under a Creative Commons Attribution-
Non Commercial-No Derivatives 4.0 license.
Trademark notice: Product or corporate names may be trademarks or
registered trademarks, and are used only for identifcation and explanation
without intent to infringe.
Library of Congress Cataloging-in-Publication Data
A catalog record for this book has been requested
ISBN: 978-0-367-63714-9 (hbk)
ISBN: 978-0-367-64698-1 (pbk)
ISBN: 978-1-003-12586-0 (ebk)
DOI: 10.4324/9781003125860
Typeset in Sabon
by Apex CoVantage, LLC
Contents
Acknowledgements ix
Prologue xi
J O H N M A R E NB O N
Introduction: Refections on Science, Theology, and the New

Aristotelianism 1
W I L L I A M M . R. SIMP SO N , RO B E RT C. KO O N S, AND JAMES OR R
PART 1
Naturalism and Nature 19
1 From Quantum Physics to Classical Metaphysics 21

W I L L I A M M . R. SIMP SO N
2 Essential Thermochemical and Biological Powers 66

R O B E RT C . K O O N S
3 Restoring the Hierarchy of Being 94

D AV I D S . O DE RB E RG
4 Evolution and the Principle of Proportionality 125

S TE P H E N B O ULTE R
PART 2
Mind and Nature 149
5 Free Will in a Network of Interacting Causes 151

TI M O TH Y O ’CO N N O R
6 Animal Powers and Human Agency 169

J A N I C E TZ U LIN G CH IK
viii Contents
7 The Power to Perform Experiments 192
D A N I E L D . DE H A A N
8 Logical and Mathematical Powers 220

A N TO N I O R AMO S- DÍA Z
9 Persons, Souls, and Life After Death 245

C H R I S TO P H E R H AUSE R
10 Angels, Principalities, and Powers 267

TR AV I S D U MSDAY
PART 3
God and Nature 291
11 Grounding and Participation in God 293

R O S S D . I N MAN
12 God and Hylomorphism 322

A N N E S I E B E LS P E TE RSO N
13 Natural and Supernatural 343

E D WA R D F E SE R
14 God, Chance and Evolution: In Memory of

Benjamin Arbour 364
A L E X A N D E R R. P RUSS
15 Teleology, Providence, and Powers 383

S I M O N M A R IA KO P F
16 Divine Lawmaking: Powers and Kinds 408

JAMES ORR
Epilogue 432
A N N A M A R MO DO RO
About the Contributors 437

Index 441
Acknowledgements
The editors owe a debt of gratitude to many colleagues and professional

bodies who have assisted in bringing their plans for this volume to fruition.
William M. R. Simpson would like to acknowledge a number of insti-
tutions for their support, including Wolfson College (Cambridge), the
Philosophy and Divinity Faculties of the University of Cambridge, the
Von Hügel Institute for Critical Catholic Inquiry (Cambridge), the Logos
Institute (St Andrews), and the John Templeton Foundation (which sup-
ported him fnancially via the international project, “God and the Book of
Nature”, Grant Id: 61507). Robert C. Koons would like to acknowledge
with gratitude the support of the James Madison Program in American
Ideals and Institutions at Princeton University, of which he was a fellow
in 2014–15.
The editors would like to thank all of the philosophers who have
contributed chapters to this edited collection for their fne scholarship
and professionalism: Stephen Boulter, Janice Tzuling Chik, Daniel D.
De Haan, Travis Dumsday, Edward Feser, Christopher Hauser, Ross D.
Inman, Simon Maria Kopf, Timothy O’Connor, David S. Oderberg, Anne
Siebels Peterson, Alexander R. Pruss, and Antonio Ramos-Díaz. In par-
ticular, the editors would like to single out Daniel D. De Haan, David S.
Oderberg, and Travis Dumsday for special thanks regarding the critical
input and advice that they provided at various stages during the prepara-
tion of this volume, and to thank Marcus Ackermann for compiling the
book’s index.
The editors would also like to acknowledge a debt of gratitude to the
many scholars and experts who have assisted in the process of providing
critical feedback on the various contributions to this volume, or provided
other forms of advice and support for this intellectual project, includ-
ing Paul Audi, Nancy Cartwright, Hasok Chang, Randolph Clarke,
Andrew Davison, Thomas Davenport, Michael Dodds, Barbara Drossel,
Jason Eberl, Hans Halvorson, Mark Harris, Simon Horsley, Christian
Kietzmann, John Marenbon, Anna Marmodoro, John Pemberton, Javier
Sanchez-Canizares, and Robert Verrill.
x Acknowledgements
Finally, the editors would like to thank John Marenbon for writing the
prologue to this volume, which deftly situates this philosophical discus-
sion, and Anna Marmodoro for writing the epilogue, which helpfully
suggests a number of ways in which the discussion begun here might be
taken forward.
Prologue
Diferent as the essays are that make up this collection, they all have
one thing in common. They envisage themselves as being, in some sense,
Aristotelian, or rather, as the title insists, ‘neo-Aristotelian’. What does
this mean? How do the studies here relate to the Aristotelian tradition?
The tradition of Aristotle has a place within broadly Western philos-
ophy unlike that of any other philosopher. Plato, Descartes, Kant and
Hegel, for instance, all had followers not just in their own times but in
later ages. But Platonism, Cartesianism, Kantianism and Hegelianism are
each particular threads within the wider fabric of the history of philoso-
phy. By contrast, Aristotelianism was dominant in three of the four main
branches of Western philosophy, from c. 500 until the late seventeenth
century in the Latin tradition, from c. 900 for roughly a millennium in
the Arabic tradition and from the twelfth to the ffteenth centuries in the
Jewish tradition. It was not just a matter of accepting certain central con-
cepts, arguments and positions, which went back to Aristotle: concepts,
arguments and positions were, indeed, adopted, though rarely without
contention and dissenters, or without development or often profound
change. Aristotelianism also provided an overall intellectual framework.
It was, indeed, not the only intellectual framework: another, diferent in
each case, and interlocking only awkwardly with it, was provided by the
scripture and doctrines of Christianity, Islam and Judaism. Nonetheless,
Aristotelianism was almost inescapable.
In the ancient world, Aristotelianism had never held this dominant
position. Like Plato, Aristotle had his followers in the centuries after his
death, but his school was one among many, including Epicureans, Scep-
tics, Cynics and the Stoics, who were the most infuential philosophers in
the Roman Republic and in the frst two centuries of the Roman Empire.
Then, with Plotinus in the third century, it was Platonism which took
over the lead and brought about the disappearance of the other schools.
Platonism was the quasi-ofcial philosophy of the later Empire. Plotinus
does, indeed, take a good deal from Aristotle, but he presents him as an
antagonist. His pupil and editor, Porphyry, took a diferent approach,
however. Porphyry argued that Plato and Aristotle, despite appearances,
xii Prologue
did not disagree, because they were concerned with diferent felds: Plato
investigated super-sensible reality, Aristotle the world manifest to our
senses. He therefore introduced Aristotle into what became the curricu-
lum of the Platonic Schools of Athens and Alexandria. Before students
reached the Platonic dialogues at the end of their studies, they would work
through Aristotle’s corpus, beginning with his logic, to which Porphyry
himself wrote an Introduction, which became standard in antiquity and
beyond. As a result, there grew up a great bulk of Aristotelian commen-
tary, in which many of the late ancient philosophers did their fnest work.
By late antiquity, Aristotelianism was, then, a powerful force, but within
a scheme of learning where Plato was regarded as supreme.
Both Arabic and Medieval Greek philosophy directly, and Jewish and
Latin philosophy indirectly, took their origin from the Platonic Schools
– but without Plato. In the two or three centuries following the fall of
Alexandria to the Muslims in 641, many of the Greek philosophical and
scientifc texts found there were translated, often via Syriac, into Arabic.
The frst of the Arabic philosophers, al-Kindî (d. 870), was more inter-
ested in the versions he had of writings by the Plotinus and Proclus than in
Aristotle. But in the tenth century Muslim and Christian philosophers in
Baghdad set about close textual study of Aristotle and revived the ancient
tradition of commentary. By contrast, although the outstanding Baghdad
Peripatetic, al-Farabi, was clearly interested in Plato, he knew, it seems,
only a little of his work, through epitomes. The mould of Aristotelian-
ism in the Islamicate world was cast, however, by Avicenna (Ibn Sina; c.
970–1037). Avicenna considered that he was an expositor of Aristotelian
philosophy, but his main works consist, not of line-by-line commentaries
or paraphrases, but of philosophical encyclopaedias in which he reformu-
lates what he found in Aristotle’s texts in accord with his own understand-
ing. The result is a far more rigorous system than Aristotle himself ever
devised, based on Aristotelian concepts and positions, but as developed,
recombined and rethought by a philosopher perhaps no less bold and
original than Aristotle himself.
It was this brand of Aristotelianism which became the staple of phi-
losophy in the Islamicate world everywhere except in twelfth-century
Andalusia, where Averroes (Ibn Rushd) continued the Baghdad method of
direct commentary on Aristotle and consistently preferred Aristotle’s own
positions to Avicenna’s alterations of them. Otherwise the translated texts
of Aristotle dropped out of use, and even the name of Aristotle was rarely
mentioned. Rather, Avicenna’s writings themselves formed the object of
a long commentary tradition, an Aristotelianism without the fgure of
Aristotle, which fourished even as the distinction between philosophical
thinking and theological speculation became blurred.
In Greek, the tradition of Aristotelian commentary was continued, in
close connection with late ancient exegesis, by writers such as Psellos,
Italos and Eustratios in the eleventh and early twelfth centuries, and in
Prologue xiii
the thirteenth century by Blemmydes and Metochites. If the Aristotelian
tradition was not obviously dominant in Byzantium, as it was elsewhere,
this was because of the importance of the Christianized Platonism of
pseudo-Dionysius.
Jewish thinkers in eleventh- and twelfth-century Andalusia shared the
same Aristotelian philosophical culture as Averroes. Although he did
not share Averroes’s desire to comment on Aristotle’s texts, the greatest
of these thinkers, Maimonides, believed that Aristotle had reached the
extreme possible for human intellect. He accordingly treated Aristote-
lian science as a fxed body of knowledge, representing what humans
could know of the natural world. His task was not to add to it, but to
consider its limitations and its compatibility with Judaism. Most of the
Hebrew-writing Jewish philosophers in the following centuries took over
this attitude and looked especially to Averroes, rather than Aristotle’s own
texts, for their knowledge of the Aristotelian system, although some went
far further in explicitly following (an Averroistic) Aristotle, whilst a few
were critical of Aristotle.
Although it was not until the mid-thirteenth century that Latin scholars
had the full range of Aristotelian texts that had already been available
in Arabic for centuries, even from the ninth century and strikingly in
the twelfth, Aristotle’s and Aristotelian logic was central to the school
curriculum and shaped the philosophical culture. From about 1200 the
universities were the main institutions for higher education, and the larg-
est of their faculties, the Faculty of Arts, rapidly became faculties for
the study of Aristotle, their syllabus organized according to the diferent
texts by Aristotle, which were discussed with the help of Avicenna and,
especially, of Averroes’s commentaries. Philosophers in the universities,
therefore, unlike those in the Islamicate world, continued to engage in
close, textual exegesis of Aristotle. All who went to the ‘higher’ Faculty
of Theology needed to have this Aristotelian training in the Arts Faculty
(or its equivalent), and so their work as theologians was developed within
the Aristotelian framework. Only in the course of the seventeenth century
did the Aristotelian curriculum fully lose its grip on most, though not all,
universities.
Although almost all of Aristotle’s writings were studied in all the four
branches, Arabic, Greek, Jewish and Latin, two sides of his work were
particularly important: his logic and his natural science. From the time
of al-Ghazali (d. 1111), the study of logic, based on the version which
Avicenna had developed from Aristotelian syllogistic, in its Avicennian
form, became a compulsory part of madrasa education. In Europe, from
the turn of the ninth century, after students had learned Latin, the study
of logic provided the foundation for all further study, both in universities
(where many students left after having studied nothing but logic) and
sub-university institutions. Although new branches of logic were invented,
the Aristotelian organon, available in full by the early thirteenth century,
xiv Prologue
was always central. Aristotelian natural science (usually enriched, in its
astronomy, by not completely concordant Ptolemaic ideas) provided an
understanding of the physical world which was questioned much less
often than Aristotle’s contentions about logic, semantics, epistemology,
philosophy of mind and action and metaphysics, where opinions, though
usually taking the form in the Latin world of diferent supposed interpre-
tations of Aristotle, were always divided.
The fact that thinkers disagreed about most things in all these areas of
philosophy did not make them any the less part of the Aristotelian tradi-
tion, since this tradition was not a set of views. It was, indeed, possible
for thinkers to stay resolutely outside the tradition, even while informing
themselves about it well (as in the case of, for instance, Ibn Taymiyya)
or badly (as with Petrarch). But some philosophers who explicitly and
relentlessly attacked Aristotelianism – for instance, Suhrawardi, Nicho-
las of Autrecourt and Hasdai Crescas (the latter two among the few to
reject even Aristotle’s natural science) – were, through their education,
methods and the context within which they worked, Aristotelians despite
themselves.
I was asked to write this Prologue as an historian of (medieval) philoso-

phy. Some readers may think that I have taken the job description too
literally, by flling it up with a lump of history of philosophy. Others, I
hope, will already have realized that it is only through at least this much
history that I can provide the necessary backing for the point that will
strike any historian of medieval philosophy faced by a collection such as
this – a point that might be summarized by observing what wisdom and
tact the editor has shown, when choosing the book’s title, in referring
not to ‘Aristotelian metaphysics’ but rather ‘neo-Aristotelian metaphys-
ics’. Aristotelianism is not a type or school of philosophy, even less a
philosophical position. Unlike a philosophical school or position, no one
chooses Aristotelianism. In Europe before 1700, in Islamicate lands until
more recently, people were born and educated into it. Now its doors are
closed to us as philosophers, though not as historians, for whom it is
the indispensable label for a long-enduring, geographically widespread
cultural condition in pre-modern Europe, large parts of Asia and some
of Africa.
By contrast, contemporary philosophers may well choose, as the
contributors to this volume have done, to be neo-Aristotelians: to take
concepts and positions from Aristotle, just as the Platonists, Cartesians,
Kantians and Hegelians (who should probably more properly have their
descriptions prefxed by ‘neo-’) look back and borrow from their respec-
tive Great Philosophers. It is an easy error for historians of philosophy
to judge this sort of relation between present and past philosophy from
Prologue xv
the wrong perspective, criticizing the ways in which the ideas and argu-
ments of whatever long-dead Great Philosopher have been distorted or
misrepresented. In doing so, they are overlooking what, as historians, they
should know well. When historians present the philosophy of the past to
readers today, they engage in a difcult process of intellectual translation
from one conceptual scheme to another. Unless the conceptual schemes
happen to be very close, this translation will always require explanatory
notes. Rather than being able to say that the GP in question held p, they
need to say something like, ‘Given that GP accepted q, r and s, and many
other people at the time believed t, GP held p’, where q, r, s and t are all
positions that today seem untenable. Such explanations ft well into an
historical essay, but they make impossible the direct engagement with the
GP that a present-day philosopher needs. Philosophers now who want
to use philosophers of the past for their own thinking need to do so
straightforwardly, without a baggage of explanations, and so cannot but
distort them. But they should not be criticized for doing what is necessary
to their job of being philosophers. Rather, when a philosopher (not an
historian) writes Plato, Descartes, Kant, Hegel – or Aristotle – we should
hasten to put scare quotes round the name. The ‘philosophical conversa-
tion of mankind’, as it has grandiloquently been called, takes place as if
over fuzzy telephone wires or between interlocutors trying, and mostly
failing, to make themselves heard over the hubbub.
There is, however, a complication that sets the neo-Aristotelianism
exemplifed by this collection from the other GP-isms and which may
be a cause of confusion to readers, and perhaps even to some of the
writers themselves. Whereas the (neo)-Kantian or (neo)-Hegelian seeks
stimulus directly from a (necessarily inaccurate) understanding of Kant
or Hegel themselves, the neo-Aristotelians writing here look not just to
Aristotle, but also to other fgures in the Aristotelian tradition, above all
Aquinas, for ideas. Moreover, in some cases, at least, their conception of
Aristotelian theories is strongly infuenced by Aquinas’s interpretation of
them. But, given that their aim is to write frst-order philosophy, not to
give an accurate account of what Aristotle (or Aquinas) thought, there is
no reason to be disturbed by this phenomenon, the result in part of the
historical weight of the Aristotelian tradition discussed in the frst part of
this Prologue. Confusion arises only if, as a result, it is thought that this
tradition is still living, and the writers here belong to it. For that, as has
been explained, is now impossible.
John Marenbon
Trinity College, Cambridge
Introduction
Refections on Science, Theology,
and the New Aristotelianism
William M. R. Simpson, Robert C. Koons,
and James Orr
Decline and Fall of the Old Aristotelianism

The philosophical framework built by the later Plato (Theaetetus, Par-
menides, Sophist, Statesman, Laws, Timaeus), Aristotle, and his succes-
sors at the Lyceum, together with the great philosophers of late antiquity
(the Neo-Platonists) and the High Middle Ages (Christian, Jewish, and
Islamic), has endured for so long and through so many intellectual revolu-
tions as to earn the sobriquet of ‘the perennial philosophy.’ This philoso-
phy shaped and structured most of the formal theology of the Abrahamic
faiths. When Augustine praised the “Platonici” in The City of God, he
was referring to the Neo-Platonists, whose work built upon Aristotle as
well as Plato. Thomas Aquinas referred to Aristotle simply as “The Phi-
losopher” and wrote commentaries on many of his greatest works. Even
after the Renaissance and the early modern period, with their revival
of the ideas of Aristotle’s ancient competitors – the Democritean, the
Epicurean, the Pyrrhonian, and the Stoic – it is almost certainly true that
the majority of teachers of philosophy in the Western world continued
to work within what might broadly be called the Abrahamic-Aristotelian
tradition.
Nonetheless, we would not wish to deny that a formidable crew of bril-
liant minds thrust upon the Western world a great challenge to Aristotle in
the seventeenth century through to the nineteenth century, which included
such luminaries as Bruno, Descartes, Hobbes, Spinoza, Francis Bacon,
Galileo, Gassendi, Boyle, Newton, Hume, and Kant. And, no doubt, a
reaction of some sort was warranted. Europe had become too deferen-
tial to Aristotle’s authority, contrary to the free spirit of exploration and
hypothesis which inhabits Aristotle’s organon of the scientifc method.
We maintain, however, that what needed correcting lay more in the
periphery than the core of Aristotle’s metaphysical system. The reaction
was an overreaction. In fact, many of Aristotle’s scientifc inaccuracies
constituted anomalies and puzzles for his system rather than central
theorems. For example, the indestructibility of the heavenly spheres
(which Aristotle accepted on the basis of observation) sat awkwardly
DOI: 10.4324/9781003125860-1
2 Simpson, Koons, and Orr
with Aristotle’s positing of matter as the basis for the possibility of sub-
stantial change, since the heavens were clearly material in nature. A more
measured reaction would have cleansed Aristotle’s system of these errors
whilst leaving the foundations intact. Unfortunately, the revolution went
too far, as revolutions usually do, to the detriment of the ‘classical theism’
generated through the confuence of Greek philosophy with the Abraha-
mic faiths. Much of the progress made in philosophy and science in the
last hundred years can be construed as a recovery of Aristotelian insights
that were trampled upon in the enthusiastic confusion that resulted from
the heady discoveries of the Scientifc Revolution.
Rise of the New Aristotelianism

The resurgence of Aristotelian concepts, theses, and explanations which
has taken place in the late twentieth- and early twenty-frst-century phi-
losophy, and which may have far-reaching implications for philosophical
theology and the burgeoning discipline of analytic theology, has many
roots. Key elements of Aristotelianism (such as teleology) were reintro-
duced into philosophy by Leibniz. In the nineteenth century, Aristotelian
and scholastic themes were prominent in the work of Trendelenburg, Marx
(in the Economic and Philosophical Manuscripts of 1844), Nietzsche, and
Brentano, and this revival continued into the twentieth century, in the
work of Meinong, Husserl, and Heidegger. Brentano infuenced Brown
University’s Roderick Chisholm, who prepared a generation of American
metaphysicians, supervising over 50 dissertations. At Cambridge, Peter
Geach (a logician) and Elizabeth Anscombe (a philosopher of language
and mind, as well as Wittgenstein’s student and literary executor) com-
bined their great expertise in analytic philosophy with a deep interest in
scholastic philosophy, especially Thomas Aquinas’s Aristotelianism.
Another crucial step in the process of restoration within mainstream
analytic philosophy was the resuscitation of the notion of causation, which
Bertrand Russell had summarily dismissed in 1914 as a “relic of a bygone
age.” Anscombe’s inaugural lecture at Cambridge in 1971, “Causality and
Determination,” was a watershed event. Interest in causal powers and
dispositions, a central element in Aristotelian metaphysics, accelerated
at the end of the twentieth century and the beginning of the twenty-frst,
with pioneering work by C. B. Martin, Alexander Bird, Stephen Mum-
ford, Brian Ellis, and George Molnar (Powers: A Study in Metaphysics,
2003). Kit Fine liberated the theory of modality from its dependence on
Leibniz’s abstract possible worlds and brought Aristotelian essence back
to fore in an important article in 1994 (“Essence and Modality”), building
on crucial work by Saul Kripke (Naming and Necessity, 1980) and David
Wiggins (Sameness and Substance, 1980).
The restoration has not been confned to the rarefed world of meta-
physics. Nancy Cartwright’s groundbreaking work within the philosophy
Introduction 3
of science – beginning with her infuential How the Laws of Physics Lie
(1983) – built on the new foundations of causal realism, further liberating
philosophy from the obsession with passively observed regularities that
had resulted from David Hume’s anti-Aristotelian philosophy. In The
Dappled World (2008), Cartwright drew the attention of philosophers of
science back to scientifc practices, describing how experimental scientists
isolate and interact with the systems they are studying so as to manifest
the powers they possess, explicitly drawing upon Aristotle in her discus-
sion of causal powers. Cartwright’s work – in Hunting Causes and Using
Them (2007) – also demonstrated that the notion of causation, even if it
doesn’t appear in our most ‘fundamental’ theories of particles and grav-
ity, is indispensable in all of the so-called special sciences, like chemistry,
astronomy, and the social sciences.
At the same time, a revolution in the domain of ethics and value theory
was taking place. Here again, Anscombe was a pioneer, whose master-
ful and provocative essay, “Modern Moral Philosophy,” written in 1958,
issued a radical challenge to the moral philosophy of her day. The bombshell
which Anscombe dropped upon modern ethics created conceptual space for
the development of the virtue ethics tradition, consolidated by Alasdair
MacIntyre (After Virtue, 1981), Philippa Foot (Virtues and Vices, 1978),
Julia Annas, Linda Zagzebski, and Rosalind Hursthouse, among others, all
of whom built upon Aristotle’s virtue-based theory in The Nicomachean
Ethics. Most (although not all) of virtue ethics is explicitly Aristotelian in
inspiration, with Aristotle’s conception of happiness (eudaemonia) coming
into renewed attention, fnding its foundation in the teleology of human
nature – especially in work by Foot (Natural Goodness, 2001), MacIntyre
(Dependent Rational Animals, 1999), and Michael Thompson.
Of course, scholars of Aristotle and the Aristotelian tradition have also
played an essential role, especially those who were most familiar with
issues in contemporary philosophy, such as David Charles, Miles Burn-
yeat, Montgomery Furth, Theodore Scaltsas, Sarah Broadie, Frank A.
Lewis, and Jonathan Barnes. Likewise, scholars of Aristotelian scholasti-
cism have been critical in forwarding these developments, such as Peter
Geach, Norman Kretzmann, Eleonore Stump, John Haldane, Marilyn M.
Adams, Calvin Normore, Etienne Gilson, Jacques Maritain, and Gyula
Klima. Through their patient labours, these scholars revealed the great
debt that the modern sciences owe to Aristotle and medieval scholasticism
and debunked many of the myths about the supposedly dogmatic and
rigid tenor of Aristotle’s system.
Modern Theology and the New Aristotelianism

Whilst a number of the scholars who have been instrumental in bringing
about the ‘neo-Aristotelian’ counter-revolution taking place in philosophy
have either been theists or sympathetic towards theism, most theologians
were entirely unafected by the post-war turn to Aristotle in Anglophone
philosophy for many years. In part this was due to the fact that it was
perceived as being largely confned to normative ethics and value theory.
However, there were also reasons internal to the various traditions of
theological inquiry which have dominated the modern Divinity Faculties
of the United States and of Europe. For much of the twentieth century,
infuential Protestant and Catholic theologians readily accepted Martin
Heidegger’s generalised indictment of metaphysics as complicit in what
he termed the “onto-theo-logical error,” namely the mistake of reifying
merely ‘ontic’ notions (such as causation, reason, or contingent existence)
before projecting them as a metaphysically ultimate ‘ontological’ ground
of reality (such as the causa sui of Descartes, the ultima ratio of Leibniz,
and the ens realissimum of Kant). While its infuence on analytic phi-
losophy was negligible, Heidegger’s critique had a signifcant impact on
continental philosophy. Since many theologians found it easier to absorb
and recover from this kind of objection than the criticisms raised by the
early analytic tradition against the meaningfulness of religious claims,
their dialogue with continental philosophy was able to continue on the
understanding that they would avoid staking out substantive metaphysi-
cal positions, while productive exchanges with analytic philosophers
waned considerably.
Karl Barth’s suspicion of natural theology, explosively expressed in
his famous exchange with Emil Brunner in 1934 and articulated more
fully in his Giford Lectures (1937–8), injected into the bloodstream of
twentieth-century Protestant theology a spirit of scepticism towards meta-
physical speculation as such. This attitude dovetailed with a strong meth-
odological preference for positive revelation over natural theology that,
he insisted, stretched back to Luther and Calvin and distinguished the
Reformed position from scholastic philosophy. Notwithstanding his con-
tested historical account of Protestantism’s hostility towards metaphysics
and natural theology – one belied by the rich tradition of Reformed scho-
lasticism documented in recent decades by Richard Muller and others –
Barth’s enduring infuence on historical and systematic theologians in
Germany and throughout the English-speaking world tended to insulate
them from the fruitful philosophical attention that was beginning to be
brought to bear on many important questions concerning the coherence
of doctrine and the rationality of religious commitment in general. Many
theologians failed to recognise the potential of the slow disintegration
of positivism and the ‘return’ to metaphysics for shielding theological
claims from sceptical headwinds, a reluctance that was compounded by
a longstanding suspicion of rationalism that liberal Protestantism inher-
ited from Friedrich Schleiermacher in the nineteenth century and that
reached its apogee in fgures such as Rudolf Bultmann, Paul Tillich, and
John Robinson. Consequently, many infuential theological movements
dramatically downplayed the distinctiveness of those claims in favour
Introduction 5
of what amounted to a thoroughgoing agnosticism about the benefts
of philosophical inquiry. That agnosticism has continued to motivate
more recent developments such as the Kantian liberal theology associ-
ated with fgures such as John Hick, Gordon Kaufman, and Don Cupitt,
which characterised doctrinal claims as culturally contingent expressions
of religious experience, or the postliberal Yale School of Hans Frei and
George Lindbeck, which contextualised confict between religious claims
as merely surface tensions arising from diferences between scriptural nar-
ratives or grammatical schemes.
Among Catholic theologians, the First Vatican Council (1869–70) and
Leo XIII’s encyclical Aeterni Patris (1879), as well as Pius X’s condemna-
tion of certain trends in secular philosophy in Pascendi Gregis (1907),
catalysed an intellectually confdent retrieval of scholastic philosophy
and, in particular, of the thought of Aquinas. This was conceived as a
long overdue response to the philosophical assaults on the rationality of
religious commitment that had begun with the philosophical critiques of
Hume and Kant in the eighteenth century and continued with increasing
ferocity into the nineteenth and twentieth centuries with the anthropo-
logical, sociological, and psychological critiques of Feuerbach, Nietzsche,
Marx, Frazer, and Freud. The manualist tradition synthesised Aristotelian
and Thomistic ideas to produce inventive analyses of the nature of sub-
stance, universals, causation, modality, and mind, which it combined with
a programmatic emphasis on the distinctions between act and potency and
between essence and existence. But the ecclesial authority that sustained
this renaissance sowed the seeds of dissension that would arrest its devel-
opment for many years to come. The rapid institutionalisation of scholas-
tic ideas, especially in the catechetical manuals, which were composed in a
rather arid style and disseminated widely in Catholic seminaries, gave rise
to what has been termed – in retrospect, rather unfairly – the ‘sawdust’
Thomism associated with fgures such as Joseph Gredt, Cardinal Mercier,
and Réginald Garrigou-Lagrange.
Within Catholic circles there were diferent reactions to this neo-scho-
lastic, manualist approach to philosophy, including strongly negative
reactions among those who perceived it to be dry and insular, sparking
a variety of neo-Thomist developments in response to neo-Kantianism,
French positivism, phenomenology, and existentialism. This wide variety
of neo-Thomists included such Catholic thinkers as Joseph Maréchal,
Jacques Maritain, Etienne Gilson, Charles DeKoninck, Karl Rahner, and
Bernard Lonergan, among others. Meanwhile, the initially contentious
criticisms of manualist approaches to the relationship between nature
and grace, which had been raised by Henri de Lubac, came to be widely
accepted by leading Catholic thinkers in the wake of the Second Vati-
can Council (1962–5). This shift in intellectual and doctrinal sympathies
was followed by the displacement of Aquinas from Catholic seminaries
for several decades. The victory of the so-called nouvelle théologie was
generally perceived by theologians and leading clerics to have brought the
scholastic renaissance to an end, in part because it advanced a theology of
nature that drew more heavily on Augustine and Christian Platonism than
Aristotle and Aquinas, a preference that continues to underpin other theo-
logical movements, notably the constellation of theologians who gather
under the banner of Radical Orthodoxy. And while there was a revival
of interest in Aristotelian and Thomistic ideas in fgures such as Herbert
McCabe, Fergus Kerr, and David Burrell, their work tended to be fltered
through a Wittgensteinian philosophy of language that made them shy
away from making metaphysically substantive claims.
Despite these unpromising historical hinterlands, the trafc between
Athens and Jerusalem in more recent years has never been busier and only
promises to increase in the years to come. The revolution in the metaphys-
ics of modality that began with Kripke and Putnam was an event that
not only expedited a revival of confdence in neo-Aristotelian metaphys-
ics more generally but also paved the way for fresh and philosophically
sophisticated approaches to theistic arguments and traditional doctrinal
problems. For philosophical theologians seeking to escape the long shad-
ows of Kantian scepticism and Humean positivism, this development
came as a relief. It led to the emergence of the movement of ‘Analytical
Thomism’ in the wake of Anscombe and Geach (the term was coined
by Haldane), which is reviving Catholic metaphysical interest in Aristotle
and Aquinas, and, more recently, the movement of ‘Analytic Theology,’
which, though by no means exclusively committed to neo-Aristotelianism,
remains broadly sympathetic to many of its distinctive positions. There
is considerable scope for future interactions between these two outward-
looking movements, which both deploy the methods and ideas of analytic
philosophy and share a common commitment to engaging mainstream
philosophy as it is practiced within the university environment. Finally,
looser but increasingly extensive circles of metaphysicians have emerged
who take up Aristotelian and Thomistic insights to develop theoretical
frameworks that are not only highly congenial to theological assumptions,
but also capable of yielding more promising solutions to longstanding
metaphysical puzzles than approaches that reject those assumptions. There
is, in short, every reason for theologians to welcome the return of rigorous
neo-Aristotelian approaches to causation, teleology, essence, and powers
and to be gratifed by the demise of defationary stances towards these
philosophical ideas. We trust that the 16 essays presented in this volume
will serve in underscoring these claims.
Part 1
In what follows, we ofer brief summaries of each of the chapters included
in this edited collection. The frst part of this volume is concerned with
the philosophy of nature.
Introduction 7
In Chapter 1, William M. R. Simpson introduces Aristotle’s doctrine
of hylomorphism, which conceived the natural world as consisting of
‘substances’ which are metaphysically composed of ‘matter’ and ‘sub-
stantial form.’ Hylomorphism was repudiated by the mechanistic and
corpuscularian philosophers of the seventeenth century, who rejected tele-
ology in nature and denied any role for substantial form in bringing about
change, and was ultimately succeeded by the dogma of microphysicalism
in mainstream philosophy, which conceives the natural world as con-
sisting fundamentally of some set of microscopic constituents. Simpson
argues that the doctrine of hylomorphism – like the concept of causal
powers – is ripe for rehabilitation in the light of contemporary phys-
ics and ofers hylomorphic accounts of two diferent interpretations of
quantum mechanics: frst, a hylomorphic account of the de Broglie-Bohm
theory, in which he posits a single cosmic substance with a teleological
nature; second, a hylomorphic account of the ‘contextual wave function
collapse’ theory recently proposed by the physicist Barbara Drossel and
the cosmologist George Ellis, in which he posits an ontology of ‘thermal
substances.’ Simpson suggests that the inorganic world consists of thermal
substances embedded within a cosmic thermal substance (without sharing
parts), all generated from the primordial substance of the Big Bang.
Robert C. Koons extends this hylomorphic picture of nature in Chap-
ter 2, in which he elaborates a ‘theory of thermal substances’ that covers
the inorganic world and explains how biological entities like plants and
animals can be conceived as substances possessed of a ‘distinctively bio-
logical teleology.’ According to Koons, the ‘emergence of new powers at
the macroscopic, biological scale should be unsurprising, given the fact
that, according to our most recent quantum mechanical models, we see
strong (ontological) emergence at the mesoscopic scale in solid-state phys-
ics and chemistry.’ In seeking to explain how these two domains of fun-
damental substances are related to one another, he proposes that thermal
substances serve as ‘virtual parts’ and as ‘proximate matter’ for biological
substances. This theory can account for the compatibility between the
material requirements of biological essences and the attributes of inor-
ganic material substances, without violating the unity of biological sub-
stances and reducing them to the sum of their physical parts.
In Chapter 3, David S. Oderberg proposes that biological and inorganic
substances are situated within a metaphysical hierarchy of being, arguing
for the retrieval of a foundational doctrine of Western philosophy. After
surveying the fate of this doctrine in the nineteenth century and distancing
himself from a variety of sociopolitical misapplications of the idea (which,
he argues, involve a serious metaphysical mistake), he puts forward a
rigorous defnition of ‘metaphysical superiority’ in terms of ‘generic pow-
ers,’ derived from Aristotle and Aquinas, which he subsequently defends
against challenges and counterexamples. According to his power-centred
criteria, the bottommost rung of the hierarchy of being is occupied by
‘pure potency’ (prime matter), followed by the inorganic (bodies), the
vegetative (plants), the sentient (animals), the rational (man), pure mind
(angels), and fnally ‘pure act’ (God). Oderberg argues that the concept
of a hierarchy of being still has much to commend itself to philosophers
and theologians and deserves serious reconsideration.
Stephen Boulter considers in Chapter 4 a potential challenge to hier-
archical views of nature presented by the theory of evolution. According
to the principle of proportionality, an (efcient) cause cannot confer on
an efect that which it does not itself possess. According to the theory
of evolution, however, life on this planet has involved a series of transi-
tions in which entities with new capacities and operations arose from a
causal base lacking those capacities and operations. Whilst the principle
of proportionality does not rule out causes bring about efects with new
properties per se, it purports to rule out the possibility of what is ‘less
perfect’ bringing about what is ‘more perfect.’ If evolutionary transitions
cannot be reconciled with the principle of proportionality, however, then
one is forced either to deny that any fundamental transitions have ever
taken place in history (and hence to deny any hierarchy of being among
biological substances) or to assert that each fundamental transition must
have involved a divine intervention. Boulter argues that a naturalistic
interpretation of the evolutionary transitions can be shown to be consis-
tent with a naturalistic interpretation of the principle of proportionality
by distinguishing neo-Platonic from Aristotelian ‘perfections’ in both the
‘order of dependence’ and the ‘order of eminence,’ and by adopting a
further principle, viz., the principle of the conservation of perfection.
Part 2
The second part of this volume is focused on the place of minded beings
within the natural order, including animal (the sentient), human (the ratio-
nal), and angelic beings (pure mind).
In Chapter 5, Timothy O’Connor canvasses a range of attempts to
defend an account of causation within a broadly neo-Aristotelian ontol-
ogy of causal powers that preserves consistency with the natural sciences.
Paying special attention to problems arising from analyses of objects and
efcient causation, O’Connor advances his own preferred solution to these
difculties that makes room for a libertarian account of human agency
and deliberate choice. More precisely, he confnes himself to an analysis
of simple entities that broadly resembles Aristotelian hylomorphism by
construing natural properties, which he construes as both qualitative and
dispositional, as grounded by way of a theoretically primitive connection
to a bare particular to form a structured unity. Rejecting four-dimension-
alism and three-dimensional eternalism as incompatible with the Aris-
totelian intuition that causation is a fundamentally productive relation,
O’Connor proposes a choice between a version of Growing Block Theory
Introduction 9
and a version of presentism, each minimally modifed to accommodate
truths about past. By way of a preliminary conclusion, O’Connor claims
that the core elements of causation involve the manifestation by objects
of powerful qualities that produces a fundamental, external, and produc-
tive relation with events as their efects. With this framework in place,
he turns to consider how best to understand the exercise of causal power
involving the deliberate choice of an agent to control intentional actions.
He argues that it is the manifestation of a specifc kind of fundamental
open-ended power, namely the power to cause an intention to act, a view
he is at pains to distinguish from the so-called noncausalist versions of
libertarianism – in particular that advanced by E. J. Lowe – by noting
that his own account afrms that reasons for action and having desires
for action can be part of the antecedent conditions that make the choices
an agent contemplates more or less probable.
In Chapter 6, Janice Tzuling Chik develops a more self-consciously
Aristotelian paradigm for understanding human agency by parsing it
naturalistically as the exercise of animal powers. She begins by critically
contrasting her animalist conception of human agency with a rival Car-
tesian conception that includes beliefs and desires among the antecedent
conditions of an action. For Chik, the Cartesian view she opposes pres-
ents an unduly intellectualist analysis of human agency that implausibly
rules out of its conception of agency the actions of non-rational animals,
whereas Aristotelian animalism, she contends, successfully preserves both
the distinctiveness of human agency and a broadly naturalistic analysis of
it that avoids familiar objections to Cartesian theories of action, including
worries about the causal efcacy of the mental in a physical domain. She
goes on to argue that Aristotelian animalism yields an attractively inte-
grated picture of action and motion, which is compatible with a disjunc-
tivist view that treats action and motion as both categorically distinct and
lacking any shared element or causal factor. The disjunctivism of this view
consists in the distinction between construing action as a psychophysical
process and motion as a purely physical one. Chik concludes that this
analysis makes sense of the plausible intuition that non-human animals
possess agency, a form of activity that by unifying mental and physical
phenomena can be accurately characterised as a psychophysical process.
In Chapter 7, Daniel D. De Haan considers an unusual specifcation
of the familiar problem articulated by Wilfrid Sellars in relation to the
compatibility of the manifest and scientifc images of human beings: he
examines the power of a human being to conduct scientifc experiments.
De Haan contends that this is a power which must be explained in terms
that are intelligible within both the manifest and the scientifc images. He
begins by arguing that it is a necessary condition of conducting a scien-
tifc experiment that a human being exercises rational embodied control
in doing so before proposing that an ontology of causal powers ofers
the best explanatory account of how human beings have the ability to
perform scientifc experiments and to draw conclusions from them. De
Haan’s central contention is that any analysis of a person exercising the
ability to conduct an experiment must involve reference to features of the
relevant scenario that are intelligible in terms of both the scientifc image
and the manifest image at the same level of explanation. Moreover, hav-
ing a grip on the manifest image of rational embodied action involved in
the ability to undertake a scientifc experiment is a condition of the pos-
sibility of arriving at any scientifc conclusions at all. De Haan concludes
that analysing this ability as a causal power defects objections facing
standard causalist theories of human agency to the efect that action as
construed on such theories is both overdetermined and susceptible to devi-
ant causal chains where mismatches arise between an agent’s intention
and the resulting action.
In Chapter 8, Antonio Ramos-Díaz takes up the Kripke-Ross Argu-
ment, a straightforward but strikingly potent case against computational
theories of mind and in support of the immateriality of thought. The
argument is premised on the alleged incompatibility between the formally
determinate character of logical and mathematical thought and the for-
mally indeterminate character of physical processes and material phe-
nomena, whether paradigmatically computational objects or otherwise,
where ‘formally’ is intended to specify the invariable form that structures
our logical or mathematical understanding rather than its variable con-
tent. Ramos-Díaz argues that the objection that the argument derives
metaphysical conclusions from epistemological premises does not work,
since even if the argument only alleges an incompatibility between the
epistemic transparency of the formal determinacy of mathematical and
logical thought and the epistemic opacity of the formal determinacy of the
physical, that would still imply a diference between logical powers and
physical powers that undercuts the contention that the former are entirely
physical. He also notes that even if the physicalist bites the bullet with an
externalist theory of mental content that locates determination beyond
the brain, the formal indeterminacy of the physical remains, since there is
no more reason to suppose that physical conditions extrinsic to a person
are more formally determinate than his intrinsic physical features. Finally,
he defects the claim that physical computing mechanisms count as evi-
dence against the Kripke-Ross Argument by observing that the logical
and mathematical powers of human beings cannot be explained in terms
of a computer’s logical and mathematical powers, since a computer only
possesses its logical and mathematical powers extrinsically – that is, in
virtue of being programmed by human beings who possess those powers
intrinsically, which is what supporters of the argument claim cannot be
explained in purely physicalist terms. Ramos-Díaz suggests that while an
Aristotelian theory of mind better explains the immaterial powers for logi-
cal and mathematical reasoning than a Cartesian one, it is not immune to
objections; but what the Kripke-Ross Argument does successfully show,
Introduction 11
he concludes, is that no purely physicalist theory of mind could account
for those powers.
In Chapter 9, Christopher Hauser addresses a well-known problem
confronting Thomistic hylomorphist accounts of the human person.
Defenders of this version of hylomorphism hold that a human person is
a composite of matter and substantial form endowed with mental and
physical powers, that souls are not identical to persons, and that souls
continue to exist independently of their material substrate beyond the
biological death of a person. Hauser distinguishes corruptionism, accord-
ing to which souls but not persons continue to exist after biological death,
from survivalism, which claims that human persons can continue to exist
after biological death, albeit in a disembodied state. He argues that cor-
ruptionism turns out to be a self-defeating position for the Thomistic
hylomorphist. The reason for supposing that the soul exists beyond
death, according to Thomists, is that it is endowed with certain powers
of thought, as Ramos-Díaz argues in the previous chapter. As Hauser
observes, the claim that souls that do not inform bodies can think is one
that most corruptionists accept. But thinking is only possible if there is
an entity capable of engaging in thought. Given corruptionism, a human
person cannot exercise the power of thought, but Hauser insists that there
is no reason to suppose that a soul is capable of thinking whether or not
it informs a body. Human persons think, but all Thomist hylomorphists
agree that souls are not human persons. In particular, Hauser contends
that it is incoherent to claim that souls are capable of having self-refer-
ential thoughts, since souls are not identical to the self, and it would be
implausible to suppose that a human person contains two thinkers (a
thinking person and a thinking soul). He concludes that the objection
forces Thomist hylomorphists to endorse survivalism as the more plau-
sible account of the post-mortem condition of human persons.
In Chapter 10, Travis Dumsday argues that scholastic angelology pro-
vides signifcant but hitherto neglected resources for substance dualists
to articulate more cogent accounts of the interaction between immaterial
substances and material objects. He lays out a strategy for how con-
temporary substance dualists could draw on these resources, noting one
way in which angelology ofers a possible solution to a familiar problem
facing substance dualists, namely the apparently arbitrary limitation of
the causal powers of an embodied but naturally immaterial substance.
Drawing on a close analysis of the substance ontologies of Aquinas,
Bonaventure, and Alexander of Hales, Dumsday’s central contention is
that, given that angels are metaphysically possible, it must be the case
either that hylomorphism is mistaken to insist that prime matter plays
an essential role in the individuation of substances or that not all sub-
stances are hylomorphic composites. Christian angelology, he concludes,
supplies important theoretical tools for developing and defending a neo-
Aristotelian philosophy of nature.
Part 3
The third part of this volume is focused on God’s relation to nature.
In Chapter 11, Ross D. Inman aims at achieving greater precision in
the metaphysics of participation, a notion that plays an important role in
soteriology and the theology of creation and has attracted considerable
interest among philosophical theologians and historians of doctrine in the
last 20 years, although it is normally associated with the Platonic rather
than the Aristotelian tradition. Inman notes that, for Christian theology,
creaturely participation in God cannot be understood either as a creature’s
being a part of God or as God’s being a part of the creature. Instead,
participation must be understood in terms of dependency on God, both in
terms of existence and in terms of what the creature is. Inman documents
the importance of the tradition of divine exemplarism as an elaboration of
this second form of dependency. Inman argues that divine fundamentality
requires that every explanation of existence facts and of essential facts
must terminate in, at least in part, facts about God. Inman argues that
this explanatory dependence on God must also be immediate. Inman then
turns to contemporary metaphysical work on the notion of grounding,
work which was inspired by the hierarchical structure of Aristotle’s meta-
physics. Inman argues that only a thoroughly neo-Aristotelian version
of grounding theory will work as an account of participation, namely,
one that builds on Kit Fine’s model of essential explanation. More spe-
cifcally, Inman identifes the essential identity-grounding relation as the
best candidate for elucidating creaturely dependence on God. Inman’s
neo-Aristotelian synthesis attests to the capacity of classical theism to
unify two central strands in classical philosophy – namely, the Aristotelian
doctrine of essentialism and the Platonic doctrine of participation – that
have often been treated as being opposed to each other.
Anne Siebels Peterson, in Chapter 12, takes up Aristotle’s own theology
concerning the question of divine unity and simplicity. Peterson urges that
we should not overlook the important dimension of analogy in our predi-
cation of unity to God, as compared with the unity of created substances.
Peterson explores two approaches in recent literature to understanding
the unity of hylomorphic substances: the conceptual distinction model
of Scaltsas, Marmodoro, and Gill, and the ontological distinction model
of Frank Lewis, Haslanger, Koslicki, and (in a somewhat diferent vein)
Loux, locating Koons’s and Simpson’s conceptions of hylomorphism (in
Chapters 1 and 2) within the ontological distinction model. For Lewis,
Haslanger, and Koslicki, the unity of a hylomorphic substance depends
on the simplicity of one of its components (namely, the substantial form).
On the one hand, Peterson contends that the unity of God – as Aristotle
conceives it in Metaphysics Lambda – is not mediated by the simplicity
of one of His parts, and that Aristotle’s God also fails to fall within the
scope of Loux’s alternative model of unity. On the other hand, God does
Introduction 13
seem to meet the criterion for substantial unity ofered by the conceptual
distinction model due to his simplicity (partlessness). However, this model
misses the role that God’s unique kind of actuality plays in securing His
unity in Aristotle’s system. This unique kind of actuality consists in divine
thought, and this actuality has a unique unity precisely because God’s
thinking has God Himself as His unique proper object. Peterson concludes
by suggesting that the closest creaturely analogue to God’s unity would
be something with a particular kind of life; namely, the life of thought.
In Chapter 13, Edward Feser seeks to elucidate the traditional distinc-
tion in theology between the natural and supernatural, making use of the
resources of Aristotelian-Thomistic metaphysics. As Feser notes, this dis-
tinction is critical to the understanding of key concepts like grace, miracle,
faith, and creation. Feser dismisses several inadequate conceptions of the
distinction, including the popular option of defning the ‘natural’ in terms
of what is studied by natural science (especially physics). As Feser points
out, such a proposal would attempt to give an epistemological defnition
of an obviously metaphysical distinction. Feser ofers philosophers and
theologians a fresh start by turning to Aristotle’s notion of nature (physis)
as an internal principle of rest and change. This in turn leads us to the
characteristic notions of active and passive powers and of substantial
form. Feser then takes on Thomas Aquinas’s distinction between essence
and existence, which provides us with a powerful way of understanding
God’s role as frst cause. Feser explains why, on this view, divine conser-
vation of the creation and divine concurrence with all creaturely action
does not entail occasionalism but instead preserves genuine secondary
causation. In light of this, we can simply defne the supernatural as what
no creature could possibly bring about. Feser concludes by defending a
traditional Thomistic conception of the relation between our natural and
supernatural ends, by articulating an account of the possibility of miracles
and by defending the reasonableness of faith in divine mysteries.
In Chapter 14, Alexander R. Pruss turns the discussion back to the
theology of evolution and proposes a new way of reconciling Darwinian
evolution (in which chance or randomness play an essential role) and
a theology of providence (according to which, for example, the physi-
cal structure of human beings is what it is by virtue of a divine ‘blue-
print’). Pruss argues that fve pre-existing reconciliations are fawed or
in need of supplementation, including determinism, generalised Molin-
ism, Thomism, the multiverse, and an appeal to some combination of
divine skill and sheer luck. The problem resolves into a familiar one in
the Aristotelian-Thomist tradition: how to reconcile God’s primary causa-
tion (including meticulous control) with genuine secondary causation, in
this case of a probabilistic, indeterministic nature. When God wills that
some specifc event occurs, then its objective probability would seem to
be one, which is prima facie incompatible with that same event’s being
the result of a chancy process. Conversely, if we insist that the event really
is chancy, we seem to be forced to read that chanciness back into God’s
own deliberative or volitional processes. The solution to this problem
turns on discovering a viable interpretation of probability. The familiar
interpretations (indiference, frequency, epistemic, propensity) have valid
but limited spheres of application, but none of them are of help in this
case. David Lewis’s best-ft version of frequentism points Pruss toward
a fourth, explicitly theistic interpretation: the objective probability of an
event of type A (given background K) is an ideal relative of frequency of A
within K, as intended by God. That is, God intends that the actual relative
frequency shall approximate the frequency posited by some generalisa-
tion, thereby making the statistical generalisation into a probabilistic law
of nature. Such statistical lawfulness arguably has great value, both instru-
mental and intrinsic, so the account does not attribute intentions to God
in an ad hoc way. And, unlike Lewis’s best-ft frequentism, divine proba-
bilistic laws will count as real explanations of actual statistical regulari-
ties. When this is applied to evolution, we can make sense of a particular
mutation’s being both chancy (relative to the world’s probabilistic laws)
and specifcally willed by God (for the sake of some providential plan).
Simon Maria Kopf tackles the same problem from a diferent perspec-
tive in Chapter 15. Does Christian theology require a teleological concep-
tion of biological evolution, according to which the emergence of human
beings was virtually inevitable? And if so, is this compatible with our best
evolutionary theories? Would replaying the tape of life produce essentially
the same results? Kopf argues that theology does not require an overall
teleology of life but needs instead merely ‘local teleologies,’ grounded in
the causal powers of individual organisms. Kopf rejects the identifcation
of providence with predictable and directional progress in evolution. By
distinguishing between the necessity of consequence with the necessity of
the consequent, we can afrm the necessary emergence of mankind (given
God’s actual plan), while denying that it has the kind of absolute necessity
that would be incompatible with worldly chance in evolution. Building on
the philosophy of Thomas Aquinas, Kopf argues that we must not identify
God’s providential direction of nature as a whole for divine purposes with
any immanent teleology, opening up the possibility that a particular end
state, even if intended by God, could still be a matter of chance in the strict
Aristotelian sense; namely, in the sense that this end state is not fxed by
the immanent teleology of creatures.
In Chapter 16, James Orr explores the connections between laws of
nature, powers, and natures. Orr argues that neo-Aristotelian perspec-
tives on these issues would be enhanced by considering non-naturalistic,
and even theistic, hypotheses. In particular, theism ofers an attractive
explanation for the peculiar modal status of laws of nature, lying, as it
does, between accidental generalisations and metaphysical necessities. Orr
suggests that God’s foresight could explain the harmony among the actual
repertoire of powers (both active and passive) – a harmony that is needed
Introduction 15
for each power to fnd appropriate manifestation partners. In addition,
divine conceptualism or exemplarism ofer metaphysical advantages in
providing an ontological basis for uninstantiated properties, which seem
to be needed for laws involving mathematical functions defned over inf-
nite domains.
For Further Reading

For students or scholars who are interested in becoming better acquainted
with the neo-Aristotelian revival in philosophy, we ofer the following
reading suggestions:
Ethics
G. E. M. Anscombe, “Modern Moral Philosophy,” Philosophy 33 (1958): 1–19.
William Fitzpatrick, Teleology and the Norms of Nature (New York: Garland,
2000).
Philippa Foot, Virtues and Vices (Oxford: Blackwell, 1978).
Philippa Foot, Natural Goodness (Oxford: Clarendon Press, 2001).
Peter Geach, “Good and Evil,” Analysis 17 (1956): 33–42.
Peter Geach, The Virtues (Cambridge: Cambridge University Press, 1977).
Rosalind Hursthouse, On Virtue Ethics (Oxford: Oxford University Press, 1999).
Rosalind Hursthouse, Gavin Lawrence, and Warren Quinn (eds.), Virtues and
Reasons (Oxford: Oxford University Press, 1995).
Alasdair MacIntyre, Dependent Rational Animals (Chicago: Open Court, 1999).
John McDowell, Mind, Value, and Reality (Cambridge, MA: Harvard University
Press, 1998).
Martha C. Nussbaum, The Fragility of Goodness (Cambridge: Cambridge Uni-
versity Press, 1986).
Michael Thompson, Life and Action: Elementary Structures of Practice and Prac-
tical Thought (Cambridge, MA: Harvard University Press, 2008).
Metaphysics
Kit Fine, “Essence and Modality: The Second Philosophical Perspectives Lec-
ture,” Philosophical Perspectives 8 (1994): 1–16.
Kit Fine “Things and Their Parts,” Midwest Studies in Philosophy 23 (1999):
61–74.
Ruth Grof and John Greco (eds.), Powers and Capacities in Philosophy: The
New Aristotelianism (New York: Routledge, 2013).
John Hawthorne, Metaphysical Essays (Oxford: Clarendon Press, 2006).
Jonathan D. Jacobs (ed.), Causal Powers (Oxford: Oxford University Press,
2017).
Mark Johnston, “Hylomorphism,” The Journal of Philosophy 103 (2006):
652–698.
Kathrin Koslicki, The Structure of Objects (New York: Oxford University Press,
2008).
Michael Loux, “Aristotle’s Constituent Ontology,” in Dean Zimmerman (ed.),
Oxford Studies in Metaphysics, vol. 2 (Oxford: Clarendon Press, 2006),
pp. 207–250.
E. J. Lowe, The Four-Category Ontology: A Metaphysical Foundation for Natu-
ral Science (Oxford: Clarendon Press, 2006).
Anna Marmodoro, “Aristotle’s Hylomorphism, without Reconditioning,” Philo-
sophical Inquiry 36 (2013): 5–22.
C. B. Martin, “Substance Substantiated,” Australasian Journal of Philosophy 58
(1980): 3–10.
Trenton Merricks, Objects and Persons (Oxford: Oxford University Press, 2003).
George Molnar, Powers: A Study in Metaphysics (Oxford: Oxford University
Press, 2003).
Daniel D. Novotny and Lukas Novak (eds.), Neo-Aristotelian Perspectives in
Metaphysics (New York: Routledge, 2016).
David S. Oderberg, Real Essentialism (New York: Routledge, 2008).
Alexander R. Pruss, Actuality, Possibility, and Worlds (New York: Continuum,
2011).
Michael C. Rea, “Hylomorphism Reconditioned,” Philosophical Perspectives 25
(2011): 341–358.
Tuomas Tahko (ed.), Contemporary Aristotelian Metaphysics (Cambridge: Cam-
bridge University Press, 2013).
David Wiggins, Sameness and Substance (Cambridge, MA: Harvard University
Press, 1980).
Philosophy of Science
Alexander Bird, Nature’s Metaphysics: Laws and Properties (New York: Oxford
University Press, 2010).
Nancy Cartwright, How the Laws of Physics Lie (Oxford: Oxford University
Press, 1983).
Nancy Cartwright, Nature’s Capacities and Their Measurement (Oxford: Oxford
Nancy Cartwright, The Dappled World (Cambridge University Press, 2008).
Brian Ellis, Scientifc Essentialism (New York: Cambridge University Press, 2001).
Rom Harré and E. H. Madden, Causal Powers (Oxford: Blackwell, 1977).
Stephen Mumford and Rani Lill Anjum, Getting Causes From Powers (Oxford:
Oxford University Press, 2011).
Emergence
P. W. Anderson, “More is Diferent,” Science 177 (1972): 393–396.
Terence W. Deacon, “The Hierarchic Logic of Emergence: Untangling the Inter-
dependence of Evolution and Self-organization,” in B. H. Weber and D. J.
Depew (eds.), Evolution and Learning: The Baldwin Efect Reconsidered
(Cambridge MA: MIT Press, 2003), pp. 273–308.
Carl Gillett, “Samuel Alexander’s Emergentism: Or, Higher Causation for Physi-
calists,” Synthese 153 (2006): 261–296.
Carl Gillett, “On the Implications of Scientifc Composition and Completeness:
Or, the Troubles, and Troubles, of Non-Reductive Physicalism,” in Antonella
Introduction 17
Corradini and Timothy O’Connor (eds.), Emergence in Science and Philosophy
(New York: Routledge, 2010), pp. 25–45.
Carl Gillett, Reduction and Emergence in Science and Philosophy (Cambridge:
Cambridge University Press, 2016).
Robin F. Hendry, “Is there Downward Causation in Chemistry?,” in D. Baird, L.
McIntyre, and E. R. Scerri (eds.), Philosophy of Chemistry: Synthesis of a New
Discipline (Dordrecht: Springer, 2006), pp. 173–189.
Kalevi Kull, Terrence Deacon, Claus Emmeche, Jesper Hofmeyer, and Frederik
Stjernfelt, “Theses on Biosemiotics: Prolegomena to a Theoretical Biology,”
Biological Theory 4/2 (2009): 167–173.
Robert Laughlin, A Diferent Universe: Remaking Physics from the Bottom
Down (New York: Basic Books, 2005).
Michael Polanyi, “Life Transcending Physics and Chemistry,” Chemical Engi-
neering News 45 (1967): 54–66.
Michael Polanyi, “Life’s Irreducible Structure,” Science 160 (1968): 1308–1312.
Mariam Thalos, Without Hierarchy: The Scale Freedom of the Universe (Oxford:
Aristotle’s Science
David Charles, Aristotle on Meaning and Essence (Oxford: Clarendon Press,
2002).
David Charles (ed.), Defnition in Greek Philosophy (Oxford: Oxford University
Press, 2010).
Montgomery Furth, Substance, Form and Psyche: An Aristotelian Metaphysics
(Cambridge: Cambridge University Press, 1988).
Allan Gotthelf, “Aristotle’s Conception of Final Causality,” The Review of Meta-
physics 30/2 (1976): 226–254.
Allan Gotthelf and J. G. Lennox (eds.), Philosophical Issues in Aristotle’s Biology
(Cambridge: Cambridge University Press, 1987).
Monte Johnson, Aristotle on Teleology (Oxford: Oxford University Press, 2005).
J. G. Lennox, Aristotle’s Philosophy of Biology: Studies in the Origins of Life
Science (Cambridge: Cambridge University Press, 2001).
J. G. Lennox and Robert Bolton (eds.), Being, Nature, and Life in Aristotle (Cam-
bridge: Cambridge University Press, 2010).
G. E. R. Lloyd, Aristotelian Explorations (Cambridge: Cambridge University
Press, 1996).
Theodore Scaltsas, David Charles, and Mary Louise Gill (eds.), Unity, Identity, and
Explanation in Aristotle’s Metaphysics (Oxford: Oxford University Press, 1994).
Philosophical Theology
Edward Feser, Five Proofs of the Existence of God (San Francisco: Ignatius Press,
2017).
Brian Leftow, God and Necessity (Oxford: Oxford University Press, 2012).
Alvin Plantinga, Does God Have a Nature? (Milwaukee: Marquette University
Press, 1980).
Alexander Pruss, The Principle of Sufcient Reason: A Reassessment (Cambridge:
Cambridge University Press, 2006).
Alexander Pruss and Joshua Rasmussen, Necessary Existence (Oxford: Oxford
J. J. C. Smart and John Haldane, Atheism and Theism (Oxford:
Wiley-Blackwell).
Richard Swinburne, The Existence of God (Oxford: Oxford University Press,
2004).
Richard Swinburne, The Coherence of Theism (Oxford: Oxford University Press,
2016).
Thomas Joseph White, Wisdom in the Face of Modernity: A Study in Thomistic
Natural Theology (Ave Maria University, FL: Sapientia Press, 2009).
Part 1
Naturalism and Nature

1 From Quantum Physics to
Classical Metaphysics
William M. R. Simpson
1. Introduction
What is the world made of, and how do we ft into it? If it is made out
of something, then are there many things or is there really just one thing,
of which everything else is derivative? Philosophical questions about the
nature of matter did not begin with the Scientifc Revolution, nor with the
advent of modern philosophy. In fact, in seeking to make sense of modern
theories like quantum mechanics, both pioneering physicists, like Werner
Heisenberg, and leading philosophers of science, like Nancy Cartwright,
have found themselves reaching as far back as the metaphysics of Aristotle
for inspiration, whose philosophy was shaped in turn by Plato and the
pre-Socratics.
In this chapter, I shall argue that the existence of quantum entanglement
and the phenomenon of emergence in fnite temperature quantum systems
have called into question the microphysicalist conception of nature that
dominated the landscape of modern philosophy, in which the world of
ordinary experience was reducible to a spatiotemporal arrangement of
microscopic constituents. I shall further argue that they have opened a
path toward a non-reductive conception of nature familiar to medieval
theologians like Thomas Aquinas, in which many of the objects of ordi-
nary experience possess ‘forms’ that determine their natures.
Although the philosophy of the Middle Ages was far from monolithic,
medieval metaphysics from the thirteenth century can be broadly charac-
terised within the Latin tradition by its explicit commitment to Aristotle’s
hylomorphic analysis of substances in terms of ‘matter’ (hyle) and ‘form’
(morphe).1 Whilst strict demarcations between medieval and early mod-
ern philosophy are increasingly discouraged by historians of philosophy,
the widespread rejection of form as the determining principle of matter,2
led by the early modern philosophers René Descartes and John Locke,
is nonetheless a striking discontinuity between medieval and modern
notions of the natural order. The abandonment of Aristotle’s hylomorphic
doctrine of substances laid the foundation for a microphysicalist concep-
tion of nature in which the whole of reality is reducible to some set of
DOI: 10.4324/9781003125860-3
22 William M. R. Simpson
microscopic constituents whose physical properties are arranged accord-
ing to universal laws. Indeed, the rise of microphysicalism has sometimes
been presented as one of the lasting and inevitable triumphs of modern
science, in which an opaque medieval philosophy was forced into retreat
by a more perspicuous account of nature committed solely to the existence
of properties which can be measured.
I aim to put such hackneyed claims into question by raising doubts
about the compatibility of microphysicalism with contemporary physics
and by drawing upon the doctrine of hylomorphism to make sense of
physical phenomena at certain points where a dogmatic commitment to
microphysicalism may be inhibiting understanding.3 Just as Aristotle’s
conception of causal powers has recently been reclaimed in contemporary
philosophy, I shall argue, hylomorphism is ripe for rehabilitation. The
discussion is divided into the following sections.
In Section 2, I consider Aristotle’s doctrine of hylomorphism, as it was
interpreted by Aquinas, and suggest that the widespread rejection of sub-
stantial form was partly the consequence of a ‘physicalisation’ of hylo-
morphism in which matter came to be seen as having physical properties
independently of substantial form, combined with a faith in the explana-
tory power of microphysical reductionism which contemporary science no
longer supports. In Section 3, I argue that the phenomenon of quantum
entanglement gives us good reason to question whether microphysical sys-
tems have intrinsic physical properties, and I sketch my recent neo-Aris-
totelian account of the de Broglie-Bohm version of quantum mechanics in
which substantial form plays a fundamental role in grounding the physical
properties of particles. In this metaphysical model, the only fundamental
physical entity is the cosmos. In Section 4, I argue that the phenomenon
of emergence in fnite temperature quantum systems gives us good reason
to doubt whether the world is a single, closed quantum system whose
behaviour can be understood in terms of universal laws. I ofer a hylomor-
phic account of the ‘contextual wave function collapse’ theory of quantum
mechanics put forward by the physicists Barbara Drossel and George Ellis,
for whom the evolution of quantum systems is open to their ‘classical’
environments and subject to local, macroscopic boundary conditions. In
this alternative metaphysical model, there are a plurality of substantial
forms which determine the natures of diferent ‘thermal substances’.
2. Matter Physicalised
2.1. Modern Microphysicalism

The writer and lay-theologian, C. S. Lewis, who held the chair of Medi-
aeval and Renaissance Literature at Magdalene College, Cambridge,
described the highest achievements of the Middle Ages as the ‘medieval
synthesis itself, the whole organisation of their theology, science, and
From Quantum Physics to Metaphysics 23
history into a single, complex, harmonious mental model’ (Lewis, 1964,
p. 11). The medieval mind had a genius for developing systems in which
‘highly original and soaring philosophical speculation squeezes itself into
a rigid dialectical pattern copied from Aristotle’ (p. 10).
Modern philosophy, in contrast, has typically adopted a more piece-
meal approach to reality. Among analytic philosophers, the question of
how human beings ft into the world, for example, is sometimes framed
(ironically) in the following way: when God made the physical world,
did he have to add anything in order for there to be human agents, who
(apparently) have causal powers to act in pursuit of their various purposes
and goals? Under these rules, the aim of the game for philosophers of
mind, or theologians, or anybody whose subject matter is deemed less
respectable than physics, is to relate those things that they want to talk
about to the fundamental physical facts, without helping themselves,
ontologically speaking, to anything more than is strictly necessary.
According to David Lewis – a leading analytic philosopher of the last
century and a staunch microphysicalist – the answer to this ontological
question is a decided negative:
all there is to the world is a vast mosaic of local matters of particular

fact, just one little thing and then another . . . We have geometry:
a system of external relations of spatio-temporal distances between
points . . . And at those points we have local qualities: perfectly natu-
ral intrinsic properties which need nothing bigger than a point at
which to be instantiated. For short, we have an arrangement of quali-
ties. And that is all.
(Lewis, 1986, p. ix)
For the modern disciple of David Hume, who denied any necessary
connections between the properties of things and the causal powers we
happen to associate with them, the law-like way in which the world
appears to be organised refects nothing more than a human habitus for
systematisation. According to David Lewis, the sparse natural properties
picked out by our best physics are related within a ‘best system’, in which
a law of nature is simply a ‘contingent generalization that appears as a
theorem (or axiom) in each of the true deductive systems that achieve a
best combination of simplicity and strength’ (Lewis, 1973, p. 73). Since
dispositions are to be worked out in terms of counterfactuals that depend
upon the truth of physical laws, and laws are contingent regularities in the
spatiotemporal distribution of sparse natural properties, a neo-Humean
microphysicalist, such as Lewis, is committed to the following claim:
Proposition 2.1. The whole truth about nature supervenes upon

the intrinsic physical properties of (and spatiotemporal relations
between) some set of fundamental microphysical constituents.
Yet Humeanism is now on the defensive in academic philosophy: the Aris-
totelian notion that things have ‘powers’ to bring about necessary change
was reintroduced within mainstream analytic philosophy by Rom Harré
and E. H. Madden (Harré and Madden, 1973). The concept of causal
powers was further developed by George Molnar in the 1990s (Molnar,
2006), and has recently become the foundation of a non-Humean theory
of causation put forward by Stephen Mumford and Rani Lill Anjum
(Mumford and Anjum, 2011).4 It has been championed in the philosophy
of mind by John Heil and C. B. Martin (Heil, 2003; Heil, 2012; Martin,
2007) and advanced in the philosophy of science by Brian Ellis and Alex-
ander Bird (Ellis, 2001; Bird, 2007), among others.5
A world with powers is a world in which entities have fundamental
agency, since causal powers are features of reality that bring about change
by natural necessity, and they are irreducible to ‘categorical’ properties
which make no reference to change. The case for reconceiving natural
properties as having (or being) causal powers, however, has done little
to extend the fundamentally real domain beyond the microphysical to
include macroscopic agents such as human beings.
Analytic theologians have typically found microphysicalist accounts
of nature too sparse for articulating a theological anthropology, whether
or not they admit causal powers. Something more is often supposed
to be needed in order to get the facts right about human persons as
purposive and responsible agents, created in imago dei. Yet whatever
these extra-physical entities are taken to be – whether they’re Cartesian
souls which somehow act upon physical bodies or mental properties
which are somehow distinct from physical properties – our ‘best phys-
ics’ is still widely supposed to provide a unifed account of the material
world in terms of microphysical properties that can be measured and
manipulated. The quandary which confronts those who are tempted
to add something non-physical to this picture of nature, however, is
whether the microphysical description of reality that they are seeking
to augment should be regarded as ‘causally closed’. This is the kind
of tribute that modern philosophers are generally expected to pay for
the triumph of scientifc empiricism and reductive simplicity over the
kind of rationalism and ontological extravagance exhibited by ancient
philosophers like Plato, for whom the truth about nature was a matter
of armchair speculation.
2.2. The Doctrine of Hylomorphism

It would be caricature, however, to portray every ancient and medieval
philosopher as being insensitive to the role of empirical investigation in
fnding out about nature. Although Aristotle was famously the protégé
of Plato, and both philosophers aspired to universal truths about real-
ity, including the truth about human souls, it is widely acknowledged
by scholars of ancient philosophy that these great thinkers of antiquity
diverged in their methodology and their metaphysics.6
For Plato, reality lay beyond our experience of the world of concrete
physical things in a transcendent realm of universals that he called the
‘forms’, which somehow cause the plurality of particulars with which we
are acquainted – or which serve as the mental templates by which ‘the
Demiurge’ moulds the world’s matter – and in which these particulars
are said to ‘participate’. A form is the universal essence which bestows
qualitative similarity upon particular things – such as the form of the
triangle, or the form of man – and can only be known through a kind
of philosophical recollection (see e.g. [Plato, Meno, 71–81, 85–86]).7 In
his treatise On Generation and Corruption, however, Aristotle criticised
Plato’s treatment of the forms on the grounds that, since they are tran-
scendent entities, they cannot function as efcient causes in the physical
world, and it is implausible that embodied beings such as ourselves could
ever come to know them (Aristotle, Gen. & Corr., 335b, 18–24). If the
forms are to explain the characteristic activities of concrete particulars,
they must be immanent within the physical world, and embodied beings
must come to know them by causally interacting with the substances that
they are said to ‘in-form’ (Aristotle, Metaphysics III, 34). It is substances,
not forms, that act as efcient causes in nature. For Aristotle, ontological
commitment is guided by the Eleatic principle:8
Proposition 2.2. A thing exists just in case that thing has the causal
power to afect and/or to be afected, to bring about change and/
or to sufer change.
Whilst the ontological status of the forms in Aristotle’s metaphysics is

still vigorously debated, Aristotle maintained a robust commitment to the
concrete reality of in-formed substances, which are fundamental wholes
whose causal powers are irreducible to the powers of their material parts.
His broadly empirical stance is manifest in his systematic study of natural
phenomena and his indefatigable classifcation of substances into natural
kinds. For Aristotle, philosophical inquiry began with the study of nature
as it presents itself in ordinary experience, proceeding to more abstract
refections upon the nature of time and change (in the Physics), and thence
to fundamental questions about the nature of being (in the Metaphysics).
For these reasons, I suggest, it is not unreasonable to consider Aristotle
as a kind of proto-scientist who inaugurated the empirical tradition of
scientifc inquiry – a tradition characterised by a remarkable confdence in
the powers of reason and observation to uncover the truths about nature.
To understand Aristotle’s philosophical account of nature, we must
begin with the reality of change in the natural world. Aristotle distin-
guished two ways of being in his account of change: there is being-in-
potency (or potentiality) and there is being-in-act (or actuality) (see
[Aristotle, Physics, I.7]).9 According to Aristotle,10 the natural order is
fundamentally composed of substances, which are concrete particulars
which have the potential to change in various ways. For example, ani-
mals are organic substances which actualise their potential for gathering
fesh by exercising their powers of growth and nutrition.11 Yet organic
substances cannot persist through all kinds of change but are transitory
entities, which are subject to processes of death and decay. Aristotle intro-
duced the co-relative concepts of matter (hyle) and form (morphe) in
his account of how substances change (see e.g. [Aristotle, Physics, II.3]).
Matter is that which changes and gets determined (or actualised), and
form is that which determines (or actualises) matter. Both metaphysical
principles are required to explain the changes that we observe in ordinary
experience, along with the concept of privation, which is the lack of the
form that is required by whatever the goal of the change happens to be.
For instance, when an animal consumes a plant, by exercising its pow-
ers of growth and nutrition, it transforms the matter of the plant into
its own fesh. In so doing, the substance of the animal is a subject of
change: by gathering fesh where there was previously a privation, the
matter of this substance is determined by a diferent accidental form. In
being consumed by the animal, the matter of the plant is also a subject of
change: by being transformed into the fesh of the animal, the matter of
the plant is stripped of powers that are essential to the nature of the plant
and acquires powers that are essential to the nature of the animal. In this
case, it is the matter underlying the substances of the plant and the animal
which is said to be determined (or actualised) by the substantial form of
the animal, since the animal now exists where its form was previously in
privation.
Aristotelian substances have a per se unity which other kinds of entities
lack. In order to distinguish an Aristotelian substance, like an animal,
from an aggregate, like a pile of sand, we must distinguish between actual
and potential parts. An aggregate is composed of actual parts that have
their own determinate physical natures. These parts can exist indepen-
dently of the wholes of which they are parts, and they retain their natures
and identities even whilst they are composing them. They are separate
actualities which can be analysed in terms of matter and form. An aggre-
gate derives its nature and being from the sum of its actual parts.
An Aristotelian substance, by contrast, does not consist of actual physi-
cal parts into which it can be decomposed. Rather, all the physical parts
of a substance are dependent for their physical natures upon the substan-
tial whole of which they are part, whose physical existence depends on
the action of a substantial form. Whilst a living substance, such as an
animal, can decompose into a collection of non-living chemicals, which
do not depend upon the original substance for their existence or their
physical natures, these physical entities are not numerically identical to
any of the parts of the substance that existed prior to its decomposition.
The separate physical entities into which a substance may decompose are
said to exist only in potential, just so long as the substance itself exists.12
The metaphysical unity of the substance pertains to its having a single
nature, upon which the natures of all of the parts of the substance jointly
depend. It is the metaphysical unity of the substance that distinguishes it
from an aggregate.13
According to a traditional line of interpretation that is often attributed
to Aquinas, substantial form is both the principle of unity of a substance
and that by which substances are fundamentally and objectively what
they are (Aquinas, De Principiis Naturae, 5, 30). For Aquinas, whilst a
substantial form is not a physical part of a substance, a substantial form
may be said to unite itself to the matter of a substance in virtue of being
the formal cause of its nature (Aquinas, De Ente et Essentia, 6). Matter
and form, on this account, may be thought of as metaphysical parts of a
substance.14 The matter which is in-formed by the substantial form of the
substance is the prime matter underlying all substances, which is the very
potentiality for substantial being. These ‘metaphysical parts’ cannot be
physically isolated by manipulating the entity in a scientifc experiment
but can only be separated by metaphysical abstraction, which is a purely
intellectual process. Without form as well as matter, a physical substance
would be incomplete.
Aquinas considered the world to be carved into a plurality of sub-
stances with diferent natures: among the world of inorganic substances,
he distinguishes various minerals and metals; among the world of organic
substances, he admits a hierarchy of plants, animals, and people, accord-
ing to their various powers. Animals have powers of perception and
locomotion (as well as various passions) that plants do not, and humans
have rational or intellectual powers that other animals do not. Since all
of these substances are subject to generation and corruption, however,
and neither arise out of nor disappear into nothingness, we require an
explanation of substantial change. According to Aquinas’s interpretation
of Aristotle, the fact that there are substances in the world with diferent
powers, which are not simply aggregates of more fundamental physical
parts, is explained by the existence of diferent substantial forms, which
determine the prime matter underlying all substances in diferent ways
(Aquinas, De Principiis Naturae, 5). This interpretation of hylomorphism
is thus committed to the following proposition:
Proposition 2.3. The physical parts of a (macroscopic) substance do

not have microphysical properties independently of the substances
of which they are part.
This proposition contradicts the claim that the whole truth about nature –
including the truth about plants, animals, and people – supervenes upon
an arrangement of microscopic properties (2.1). According to Aristotle
and Aquinas, many of the objects of ordinary experience are substances
which are irreducible to their parts, including biological agents like human
beings. So why was this metaphysical account abandoned in favour of
microphysical reductionism?
2.3. Matter and Form in Late Scholasticism

Whilst the concepts of matter and form were widely deployed in scholastic
metaphysics in accounting for the nature and unity of substances, these
doctrines were developed by some scholastic philosophers in ways that
were incompatible with the Aristotelian-Thomistic conception of hylo-
morphism. No single systematic view characterises the scholastic period.
Nonetheless, it is possible to pick out some suggestive tendencies within
medieval metaphysics that served as a prelude to microphysicalism.15 I
shall use the term physicalise to refer to the tendency to treat the matter
of a substance as having the same ontological standing as the substance
by attributing to matter intrinsic properties or causal powers.16
The necessity of some material substrate underlying all forms of change
was widely accepted within the Latin tradition. As Franco Burgersdijk
observed, ‘all seem to have granted to Aristotle that the generation and
corruption of natural things requires a common subject’.17 In this way,
medieval scholastics sought to afrm the continuity of natural processes
and to avert the supposition that change must involve creation from noth-
ing. However, Aquinas’s characterisation of this substrate as a determin-
able potentiality was widely criticised by other scholastics for failing to
bottom out in anything concrete or determinate and was never widely
accepted, even in the thirteenth century. William of Ockham, writing in
the early fourteenth century, echoed Averroes in requiring that matter
should have extension.18 Duns Scotus insisted against Aquinas that this
material substrate should have actual parts.19 The metaphysical misgiv-
ings concerning a merely determinable substrate were starkly expressed
by the seventeenth-century Ockhamist, André Dabillon, who insisted
that either ‘the things that compose an actual being actually exist, or a
substantial whole would be composed of nothing’. Since this position is
untenable, he claimed that it must be the case that both ‘matter and form
are real substantial beings that exist actually in nature’.20
In addition to physicalising the material substrate of change, increasing
its independence from substantial form, some scholastics seem to attribute
a quasiphysical status to form, suggesting that it interacts with other enti-
ties like an efcient cause. A signifcant example of this tendency lies in
the widespread rejection of the so-called unitarian doctrine of substantial
form.21 In Aristotelian-Thomistic hylomorphism, a substance has a single
substantial form, which is the principle of unity of the substance and
determines its nature. For some scholastics, however, such as Scotus, a
plurality of substantial forms were said to exist within the same substance.
For example, the form of corporeity (by which an animal is embodied),
and the form of the soul (by which an animal is living), were held to be
present simultaneously within a human substance.
Yet if multiple substantial forms can exist within a substance that simul-
taneously determine its causal powers, wherein lies the unity of the sub-
stance, and how do they determine its properties? For certain scholastics,
it seems the temptation was to preserve their commitment to the unifying
character of forms by portraying them as elements within the compos-
ite with powers to organise their various parts into a functional whole.
Francisco Suárez, a philosopher of the late sixteenth and early seventeenth
centuries, seems to be considering such a position when he writes,
The aggregation of multiple faculties or accidental forms in a simple

substantial subject is not enough for the constitution of a natural
thing . . . A form is required that, as it were, rules over all those
faculties and accidents, and is the source of all actions and natural
motions of such a being.22
Yet once matter and form have been physicalised, formal and efcient
causation become difcult to distinguish from one another. For the Aris-
totelian-Thomistic hylomorphist, substances share a substrate of deter-
minable potentiality, which is in-formed one way and then another. For
certain scholastics, however, substantial forms seem to act directly as ef-
cient causes, since that is the only way in which they can make a diference
to things that have their own intrinsic and determinate natures.23 Whilst
the metaphysical roles of matter and form remained the same in inten-
tion, the ways in which their tasks were implemented rapidly shifted from
their Aristotelian-Thomistic moorings. Matter and form were physicalised
and formal causation confounded with efcient causation, a circumstance
that would place substantial forms in direct competition with physical
mechanisms.
2.4. The Rise of Microphysicalism

The mechanical philosophy of the seventeenth century, far from arising
in a philosophical vacuum, represents a development in these tendencies
within medieval philosophy, which culminated in the complete physi-
calisation of the material substrate of change, combined with an explicit
rejection of the notion of substantial forms. Abandoning the hylomor-
phism of late scholasticism, the corpuscularianists proposed an alternative
ontology consisting of corpuscles arranged within physical space which
have intrinsic and determinate properties, echoing the atomism of Leu-
cippus and Democritus that Aristotle had so vehemently opposed. In this
new vision of the world, nature wears all of her properties on her sleeves
where they are exposed to being measured and manipulated, unlike the
metaphysical principles underlying the world in medieval philosophy,
which are only uncovered through a process of metaphysical abstraction.
Not all of the early mechanists were committed to an ontology of dis-
crete corpuscles. In the mechanical philosophy of Descartes, reality was
divided between thinking things (res cogitantes) and extended things (res
extensa), where the extended things are wholly characterised by the geo-
metric properties of ‘shape, size [and] position’ and their motions are
governed by universal laws. A thoroughgoing reductionist in his approach
to the material world – which included fora and fauna and the animal
kingdom but excluded human persons who are immaterial souls – Des-
cartes believed ‘there is nothing in all of nature whose character (ratio)
cannot be deduced through these same principles’ and dismissed substan-
tial form as ‘a philosophical being unknown to me’.24, 25
It would be difcult to overstate the enduring infuence of Descartes’s
metaphysics, although Cartesian physics was a short-lived afair by con-
trast: Isaac Newton rejected Descartes’s identifcation of matter with
extension and proceeded to develop an alternative account of motion that
would rapidly secure Newtonian mechanics as the archetype of modern
physics.26 Nonetheless, the common commitment to a physicalised form
of matter continued to set the agenda for natural philosophers from the
seventeenth century. Buoyed by swift advances in the experimental sci-
ences, corpuscularianism swiftly supplanted scholasticism in many parts
of Europe, as scientists like Robert Boyle contrived plausible mechanical
explanations for natural phenomena, specifcally targeting cases in physics
where scholastics had attributed phenomena to the activities of forms.27
Henry Oldenburg, who served as the frst secretary for the Royal Society,
memorably complimented Boyle for having ‘driven out that drivel of sub-
stantial forms’ which ‘has stopped the progress of true philosophy, and
made the best of scholars not more knowing as to the nature of particular
bodies than the meanest ploughmen’.28
Whilst corpuscularianists maintained a commitment to the notion of
a material substrate underlying all change – in Boyle’s view, a ‘substance
extended, divisible, and impenetrable’29 – the doctrine of substantial forms
was swiftly abandoned during the course of the seventeenth century (albeit
with some notable dissenters, such as Leibniz (1976)). This extirpation
of form was accompanied by a diferent account of the generation and
corruption of the things we encounter in experience. As Silva observes,
without forms to determine the intrinsic powers of substances, natural
philosophers increasingly relied upon ‘laws of nature that extrinsically
guided the movements of corpuscles and atoms in a void’ to explain how
things (apparently) come into and out of being (p. 64), and thus ‘intrin-
sic natural formal causes were replaced by extrinsically imposed laws of
nature’ (Silva, 2019, p. 65).30 According to Boyle, the material world that
is laid bare by the physical sciences should be regarded as a ‘contrivance
of brute matter managed by certain laws of local motion’ (Boyle, 2000,
vol. 10, p. 447). The matter of which everything is made persists through
time and only changes with respect to accidents like position.
By the eighteenth century, David Hume had sought to ‘introduce the
experimental method into moral subjects’, in A Treatise on Human Nature
(1739–40), laying the foundations for a microphysicalist account of the
mind, and the French mechanist Pierre Simon de Laplace had given voice
to a mechanical stance toward the whole of nature that would dominate
the imagination of philosophers until the turn of the twentieth century.
According to the new philosophy of nature, the state of the whole cosmos
at any future time, including human beings, is entirely fxed by the pres-
ent locations and momenta of small particles and the laws of Newtonian
mechanics.31 The only changes in such a world are the accidental changes
in the arrangements of the particles. This is the world of ‘classical physics’
with which most philosophers are familiar, in which the mechanical laws
of physics determine all the physical possibilities of nature. It is a causally
closed world in which substantial forms have no role to play in bringing
about change.
3. Form in Standard Quantum Mechanics

The confdence of the corpuscularianists in their capacity to explain every-
thing in terms of the properties of material corpuscles, and the animus
displayed by early philosophers of science like Oldenburg against the
medieval scholastics, efectively banished the doctrine of hylomorphism
from mainstream philosophical discourse. Yet the microphysicalism that
came to dominate the Anglophone philosophical tradition, which eventu-
ally dropped any commitment to corpuscles, has run into difculties with
contemporary physics. It is widely accepted, for example, that the theory
of quantum mechanics, which superseded classical physics, is incompat-
ible with the doctrine of Humean supervenience, as it was formulated
by Lewis (2.1). As Maudlin has pointed out: the standard doctrine of
Humean supervenience is inconsistent with the existence of quantum-
entangled states, which are indeterminate states of a physical system that
cannot be characterised in terms of the intrinsic properties of spatially
separated microscopic constituents (Bell, 1964; Maudlin, 2007). In what
follows, I shall discuss this problem and consider how it can be addressed
by adopting a hylomorphic framework in which matter ontologically
depends upon form for its physical properties.32
3.1. Quantum Entanglement

The fundamental mathematical object within quantum theory is the
quantum state, which encodes the probability of an arbitrarily compli-
cated physical system having a particular confguration. In the famous
EPR experiment involving two microscopic particles originally proposed
by Einstein and his collaborators (Einstein et al., 1935), as it was sub-
sequently presented by David Bohm, one particle is constrained to be
‘spin-up’ |↑〉 when another is measured to be ‘spin-down’ |↓〉, and vice
versa, however far apart the two particles are spatially separated.33 The
physical system in this case is said to be in a ‘quantum superposition’ that
is described by the singlet state:
1
˜ 1, 2
=
2
(° 1
˛ 2−˛ 1
° 2
). (1)
According to this formalism, when a system comprised of two particles

(1 and 2) is in the singlet state ˜ 1, 2 : there is a probability of 1/2 that we
will observe particle 1 to be ‘spin up’ ° 1 and particle 2 to be ‘spin down’
˛ 2 ; and there is a probability of 1/2 that we will observe particle 1 to
be ‘spin down’ ˛ 1 and particle 2 to be ‘spin up’ ° 2. There are no other
physically possible outcomes.
3.1.1. Non-Local Correlations

The challenge that such systems pose to the doctrine of Humean super-
venience is the fact that we cannot explain their measurement statistics
in terms of the local properties of their constituent parts. Suppose two
quantum-entangled particles that are emitted from a common source fy
of in opposite directions, and two experimenters (traditionally, ‘Alice’
and ‘Bob’) measure the spins of each particle once they are sufciently
separated using diferent measuring devices. Let φA specify the confgura-
tion of Alice’s apparatus, and A the outcome of her experiment; likewise,
φB for Bob’s apparatus, and B for his outcome. Let λ denote whatever in
the past may have infuenced the behaviour of the system that is being
measured. In this example, λ includes the physical state of the two-particle
system, prior to measurement. The measuring apparatus in each case is
a Stern-Gerlach device, in which a pointer has the possibility of being
defected up or down, and the confguration parameters are the two angles
of polarisation of each device, φA and φB. These parameters can be set
at an appropriate angle for measuring vertical spin (that is, ‘spin-up’ or
‘spin-down’), but can also be adjusted separately to produce a range of
measurement outcomes. According to the physicist John Bell (Bell, 1964),
the principle of locality requires:
P°a ,°b ( A | B,˛ ) = P°a ( A | ˛ ) (2)
P°a ,°b (B | A,˛ ) = P°b (B | ˛ ) . (3)
The formalisation of the frst equation can be read as follows: if the prin-
ciple of locality is true, then the probability P for Alice obtaining outcome
A can be fxed by conditionalising on the confguration of her apparatus
φA and whatever in the past infuenced its behaviour λ, such as the local
properties of the particle she measures. Signifcantly, in a world in which
locality holds and the two measurements are conducted simultaneously,
conditionalising on the confguration of Bob’s apparatus φB and outcome
B does not change the probabilities for Alice’s outcome. Hence, the two
probabilities in the frst equation are stated to be equal. This is also the
case for Bob’s outcome with respect to Alice’s apparatus, hence the equal-
ity of the two probabilities in the second of the two equations.
However, Bell’s theorem demonstrates that the principle of locality
is violated by the phenomenon of quantum entanglement. According
to quantum mechanics, the probabilities for obtaining a measurement
outcome in one part of the experiment depend on the outcome obtained
in the other part of the experiment, in spite of the fact that the two mea-
surement events are represented as ‘space-like separated’ in the theory of
special relativity, because they are conducted simultaneously within the
frame of reference of the experiment.
To visualise the limits on signalling which are imposed by the theory of
relativity (in this case, between the two wings of the EPR experiment), the
mathematician Minkowski suggested imagining a fash of light, confned
to a two-dimensional plane, which spreads out in a circle from an event
E at some time t (Minkowski, 1908). If we graph the growing circle using
time as the vertical axis, we obtain a ‘light-cone’ for event E that extends
to include any past event t´ < t in which a signal could have been sent
which would have time to reach E and causally infuence this event. Any
event which falls outside of the light-cone of E is ‘space-like separated’
from E and cannot be causally related to E by a classical mechanism.34
Since the measurement events in the two wings of the EPR experiment
corresponding to A and B do not belong within each other’s past light-
cones, they are space-like separated. In other words, the assumption that
the behaviour of the particles can be explained by a (subluminal) physical
mechanism that governs their local properties implies one set of mea-
surement statistics, whereas quantum mechanics predicts another. Signif-
cantly, in the case of the EPR experiment, quantum mechanics predicts
that the measurement statistics will depend on the relative angle between
the two devices, φA – φB; a fact that neither particle, considered separately,
is in a position to ‘know’.
Whilst the EPR experiment was originally proposed as a thought
experiment by Albert Einstein and his associates, and was intended as a
reductio ad absurdum of quantum mechanics, subsequent experiments –
in particular, those of Alain Aspect in the 1980s (Aspect et al., 1982) –
are now widely regarded as having confrmed the statistics predicted by
quantum mechanics and established non-locality as an empirical fact.35
Contra Laplace, we cannot explain the physical behaviour of everything
in nature in terms of the mechanical forces between physical bodies and
the intrinsic properties of their microscopic constituents. The phenome-
non of quantum entanglement suggests that the mechanical stance toward
nature which was adopted by the corpuscularianists ofers an efective
description of physical phenomena which holds only at certain scales.
3.1.2. The Measurement Problem

The existence of quantum superpositions, such as the quantum states of
entangled particles, confronts us with the additional problem of reconcil-
ing the formalism of quantum mechanics with the reality of determinate
measurement outcomes. Prior to any measurement of a quantum system –
or any collapse-inducing event, the quantum state evolves according to
the time-dependent Schrödinger equation,
ˇ˜
Hˆ ˜ = i˜ , (4)
ˇt
where Ĥ is the Hamiltonian of the system which represents its energy,

and ħ is the reduced Plank’s constant. The formal solution of Equation
(4) is the quantum state (or, wave function) ˜ (t ) . The quantum state
can be expressed in terms of an operator Uˆ , such that the state of a sys-
tem at some arbitrary time t can be obtained from its state at some ear-
lier time t = 0 through the action of this operator: ˜ (t ) = Uˆ (t ) ˜ (0) .36
This formula tells us how to start from a given state of a system and evolve
the probability amplitudes for all the possible confgurations of the sys-
tem in time. Yet suppose we perform a ‘non-demolition’ measurement on
the system, which does not destroy the quantum system being measured
(Dong et al., 2008). After this measurement, we know more about the
state of the system than the information contained in the quantum state.
For example, the measurement outcome of the EPR experiment will
have ruled out one of the combined states of the two particles to which
˜ assigns a non-zero probability; perhaps the state in which particle 1
is spin-down and particle 2 is spin-up ˜ 1, 2 = ˛ 1 ° 2 . To obtain the cor-
rect results for future experiments, we have to update the wave function
of the system we are measuring with the empirical knowledge we have
gained from our experiment. Yet this updating is not performed by the
time evolution operator Û. For instance, suppose at time t we fnd that
particle 1 is spin-up and particle 2 is spin-down: ˜ 1, 2 = ° 1 ˛ 2 . The wave
function has to undergo the following discontinuous modifcation:
˜ (t −˝t ) = Uˆ (t −˝t ) ˜ (0)

1,2 1,2
˜ (t + ˝t ) =° 1
˛ 2, (5)
1, 2
where δt denotes an infnitesimal period of time. This discontinuous
change in a system’s quantum state is known as the ‘collapse of the wave
function’, and it is necessary to properly account for any non-demolition
experiment. There is no agreed understanding of this physical process
(Omnés, 1994). Even if the phenomenon of decoherence is taken into
account, in which the quantum nature of the system is said to ‘leak’ into
its environment, the time evolution operator must be supplemented with
a discontinuous change in the system’s state. This disconnect between the
quantum formalism describing the physical state of a system, which can
only be specifed empirically in terms of an indeterminate superposition
of mutually exclusive measurement outcomes, and the ‘classical’ world
of observation occupied by scientists, in which measurement always give
rise to a determinate physical outcome, naturally gives rise to the question
of why this apparent indefniteness in the microrealm is not transmitted
up to the macroworld (as in the notorious ‘Schrödinger’s cat’ thought
experiment). Is quantum mechanics about an objective world that exists
independently of scientists and their measurements, or does it merely keep
track of the state of our knowledge during the course of an experiment?
3.1.3. Matter Without Intrinsic Physical Properties

According to Bell, any realist approach to quantum mechanics that seeks
to explain the existence of determinate measurement outcomes must
come to terms with a dilemma: either the dynamics of standard quantum
mechanics is wrong, and the wave function evolves according to a non-
linear Schrödinger dynamics that permits the wave function to collapse
independently of the ‘observations’ of any scientist, or there are ‘hid-
den variables’ in addition to the wave function, which evolve according
to some non-linear dynamics of their own (Bell, 1987). In either case,
standard quantum mechanics must be regarded as physically incomplete.
Maudlin has argued that the choice comes down to two possibilities
(Maudlin, 1995): either we should adopt something like the modifed
Schrödinger dynamics proposed in Ghirardi et al. (1986) (GRW theory),
in which the wave function undergoes spontaneous collapse, or something
like the pilot wave theory of de Broglie and Bohm (Bohm, 1951, 1952;
de Broglie, 1928), which includes an equation of motion for a particle
confguration.37
The GRW theory seizes the frst horn of the dilemma by incorporating a
stochastic mechanism which produces random ‘hits’ on the wave function
that occur universally for microscopic particles and result in an objective
collapse of the wave function (Ghirardi et al., 1986). The efects of this
non-linear modifcation to the Schrödinger equation become signifcant
when a large number of quantum-entangled particles are involved, such
as the particles composing a measuring device. The theory of Bohmian
mechanics seizes the second horn of the dilemma by positing a global
confguration of particles whose trajectories are choreographed by the
wave function (Bohm, 1951, 1952; de Broglie, 1928). The guiding equa-
tion for the particles depends in a non-linear way upon the wave function,
which evolves according to the standard Schrödinger equation.
According to Allori et al. (2008), however, the diference between these
apparently opposite approaches to fxing standard quantum mechanics
is not as stark as it seems. GRW theory and Bohmian mechanics may be
interpreted as sharing a common structure: ‘they are ultimately not about
wave functions but about “matter” moving in space, represented by either
particle trajectories, felds on space-time, or a discrete set of space-time
points.’ This confguration of matter makes up the world of macroscopic
objects, including our measuring devices, and ‘the role of the wave func-
tion . . . is to govern the motion of the matter’ (p. 353).
In the Bohmian primitive ontology, this confguration of primitive mat-
ter consists of N discrete particles. The trajectory of every particle is gov-
erned by an equation of motion that is frst-order in time, which depends
upon both the universal wave function ψ and the positions {Q1, . . ., QN}
of all the other particles comprising the global confguration:
dQ i ˜ i ˜
= ˙i˜ (Q1 , …, Q N ) mi Im 2 (6)
dt ˜
where ˙i˜ is the velocity of particle i at time t, and mi is its gravitational

mass.38 In the GRWm primitive ontology for GRW theory, which was
frst suggested in Ghirardi et al. (1995), the matter is not discrete but
consists of infnitely divisible gunk. It is a matter-feld whose distribu-
tion of matter-density m(x, t) expands with the unitary evolution of the
wave function ψ and contracts in spontaneous localisation events, which
is governed by an equation of the form:
N
2
m ( x, t ) = mi dq1 ˛dq N ˝ (q i − x) ˜ (q1 ,˛, q N , t ) , (7)
R3 N
i=1
where the sum ranges over the N quantum ‘particles’ in the physical
system (from i = 1 to i = N), and the integral ranges over the whole of
the 3N-dimensional confguration space in which the wave function
ψ is defned. (In computing m(x, t) for a given position x and time t,
the integration of the wave function ψ combined with the Dirac delta
function δ(qi – x) gives us the marginal distribution of the ith degree of
freedom qi ∈ R3, by integrating out all other variables qj, then one takes
the mass-weighted sum of these contributions for all of the particles.)
In both cases, the microscopic description of standard quantum
mechanics is completed, and the existence of determinate measurement
outcomes is ontologically explained, by the introduction of a spatiotem-
poral distribution of primitive matter, which is choreographed by a wave
function that is defned as a holistic feature of the physical system. In both
cases, this distribution of matter lacks any intrinsic physical properties.
The particles (or parcels of gunk) are holistically individuated by the
distance relations in which they stand, deriving their physical natures
from the wholes of which they are parts. The wave function of quan-
tum mechanics, which is defned in a high-dimensional confguration
space, is not included as part of the fundamental ontology, to avoid
creating an interaction problem involving two separate domains, but
is assigned a nomological role in explaining the motion of the matter
in physical space.
It must be emphasised that the matter in these primitive ontologies can-
not be read of the formalism of standard quantum mechanics or deduced
from any other physical concept or from the vocabulary of any physical
theory. It is posited for the sake of empirical adequacy, since there are
facts about which way the pointers on our measuring devices are point-
ing. A primitive ontology approach aims to explain the measurement
outcomes of quantum experiments, like the EPR experiment, and more
generally to explain the behaviour of the macroscopic objects upon which
scientists depend to make their measurements, by ofering an account of
the empirical content of a physical theory that is exhausted by its state-
ments about this primitive distribution of matter (Maudlin, 2019).
3.2. Cosmic Hylomorphism

There is good reason to think, however, that a fundamental ontology con-
sisting solely of matter without intrinsic physical properties is too sparse
to account for the truth of the laws that determine the spatiotemporal
development of the matter. Something more is needed, in addition to a
distribution of particles or gunk, in order to explain the phenomena that
scientists observe in quantum experiments.
3.2.1. Bohmian Mechanics

I wish to focus in the remainder of this section on the Bohmian solution
to the measurement problem, which attributes to the cosmos a universal
wave function that does not collapse. In Section 4, I will consider a dif-
ferent approach to quantum mechanics, in which the wave function is
permitted to collapse.
Bohmian mechanics takes the non-locality of quantum mechanics seri-
ously, providing an account of quantum experiments in terms of both the
universal wave function ψ and a set of N particles. Although we cannot
measure the universal wave function, nor can we trace the trajectories of
the Bohmian particles, it is possible under certain conditions to provide
an empirically adequate account of a subsystem of physical particles in
terms of the efective wave function of the subsystem, which encodes the
statistical information about the positions of the particles that can be
recovered in an experiment. According to Bohmian mechanics, it is only
efective wave functions which are subject to collapse.
For Bohmians, the quantum spin of a particle is explained by the efec-
tive wave function of the particle and its initial position. Suppose we
direct a spin 1/2 particle at the magnets of a Stern-Gerlach device, such
that its wave function splits into ‘up’ and ‘down’ wavepackets. The way
in which the particle travels within the apparatus – its ‘choice’, let’s say,
to travel up rather than down – is determined by its initial position in
relation to the magnets and their felds. If this particle is entangled with
another particle in the singleton state (1), however, that choice changes
the wave function of the other particle, guaranteeing that it will go down
when it is measured. In efect, the wave function of the second particle
has been forced to evolve non-locally in three-dimensional space, with the
local behaviour of particle 1 nomically determining the wave function of
particle 2. Consequently, after the measurement of particle 1, the ‘spin’ of
particle 2 – and hence its trajectory – is no longer a function of its position
in relation to the magnets used to measure it.
If the orientation of the magnetic feld used to measure particle 1 were
reversed, however, and one measured the spin of particle 1, particle 1
would travel down rather than up, assuming the initial position of particle
1 to be the same with respect to the apparatus. In that case, one would
obtain a wave function for particle 2 in which it travelled up instead,
irrespective of its initial position. The Bohmian is thus committed to the
existence of action at a distance – an action that cannot be explained by
mechanical contact or the intermediation of physical felds – since the
scientist’s choice of the orientation of the magnetic feld which measures
the spin of particle 1 afects not only the motion of particle 1 but also of
particle 2, however far apart the two particles are from one another.
The Bohmian is also committed to the existence of a world in which
there are defnite physical objects and determinate measurement out-
comes: there is always a fact of the matter about where every particle
stands in relation to one another, and the macroscopic devices which
are used to make measurements are composed of particles. According to
the kinematic interpretation of Bohmian mechanics, the particles do not
have any intrinsic physical properties, but every particle in the cosmos is
assigned an instantaneous velocity which depends on the defnite positions
of all the other particles and the universal wave function. The Schrödinger
equation and the guiding equation comprise the non-classical dynamics
of a particle confguration that exists independently of our observations.
The Bohmian theory, for all practical purposes, is empirically equivalent
to standard (non-relativistic) quantum mechanics.39 Agreement with the
Born rule, which gives the probability that a measurement of a quantum
system will yield a given result, is secured via the quantum equilibrium
hypothesis: specifcally, if the initial confguration of the particles at t may
be supposed to be randomly distributed with a probability distribution
2
ˆ (t = 0) = ˜t=0 , then it follows as a consequence of the Schrödinger
equation and the Bohmian equation of motion that this relationship will
2
hold at some later time t > 0 for the distribution ˆ (t ) = ˜t .40 Although
the particles have determinate positions, we cannot know where all of
the particles are, and so we must resort to the probabilities of quantum
mechanics in order to make any predictions. Nature does not wear all
of her properties on her sleeves. Nonetheless, the physical state of the
world is completely specifed by the wave function and the positions of
the particles.
3.2.2. Rival Views of Laws

Yet why would particles without physical properties follow the trajec-
tories laid out for them in Bohmian mechanics, supposing a theory like
Bohmian mechanics were true? The Armstrong-Dretske-Tooley con-
ception of lawhood as a second-order relation between universals, for
example, is unsuitable, since the particles do not have any properties that
could instantiate necessitation relations and the wave function is not part
of the primitive ontology. According to the ‘primitive ontology approach’
to quantum mechanics espoused by Allori et al. (2008), the wave function
of quantum mechanics plays a nomological role in the temporal develop-
ment of the particles. However, there are two ways of spelling out this
role: namely, by appealing to a Humean account of laws, in which laws
are merely summaries of regularities for all space and time, or to some
form of dispositionalism, in which laws are grounded in powers (Esfeld
et al., 2017).
Michael Esfeld has proposed an ontology for Bohmian mechanics in
which the world is composed of ‘matter points’, which are nothing over
and above the distance relations in which they stand (Esfeld and Deckert,
2017). To accommodate the truth of the Bohmian law of motion, which
is specifed in terms of both the particles and the wave function, Esfeld
adopts the Mill-Ramsey-Lewis account of laws, in which a regularity
only qualifes as a law of nature just in case it is an axiom in the ‘best
system’ that balances strength and simplicity in deriving the facts about
the positions of the matter points.41 According to this Humean account
of the quantum state, the Bohmian law and the quantum state supervene
upon nothing less than the global confguration of particles for all time.
However, many philosophers have found Humean accounts of laws
to be deeply unsatisfactory. For instance, Humeanism fails to capture
the intuition that laws have metaphysical work to do in explaining what
happens. According to Humeans, the Bohmian law depends for its law-
fulness upon the global confguration of particles and is thus constituted
by that which it seeks to explain. Yet ‘a fact cannot be used to explain
itself’ (Armstrong, 1983, p. 40), as Armstrong complained. In making this
claim, Armstrong was echoing Plato’s insight that things are explained by
being referred to a ‘higher principle’. The Mill-Ramsey-Lewis account,
however, not only fails to maintain the necessary metaphysical distance
between explanans and explanandum, but also reverses the proper order
of explanation: laws are supposed to explain instances that fall under
them, yet the lawfulness of laws in this model is grounded in the instances
they are supposed to explain. Doubts have also been raised about the
logical coherence of the Super-Humean’s attempt to make the doctrine
of Humean supervenience compatible with quantum entanglement.42 For
example, in rejecting an ontology of sparse natural properties in favour
of a supervenience base consisting solely of positions, the Super-Humean
removes an objective constraint upon what could count as a ‘best system’
of laws, leading to the well-known problem of ‘immanent comparisons’
and to subjectivism about the laws of nature (Matarese, 2018).
In order to build a metaphysical model that incorporates a power to
choreograph the trajectories of the particles according to the Bohmian
law of motion, however, a dispositionalist must answer a number of
metaphysical questions. For instance: what sort of thing might be sup-
posed to possess a causal power which could ground the Bohmian law of
motion? Some dispositionalists have been tempted to attribute Bohmian
particles with their own intrinsic dispositions. For example, Suárez has
suggested that each Bohmian particle has powers to change its velocity
which depend upon the spatial confguration of all the other particles
for their stimulation (Suárez, 2015). It is doubtful, however, that this
conspiracy model ofers any advantage over Super-Humeanism. For one
thing, Bohmian dispositionalism fails to capture intuitively correct coun-
terfactuals about what would happen in Small Worlds which have only
a few particles, since the lawfulness of the Bohmian law of motion is not
due to any intrinsic features of the particles which could be duplicated in
other possible worlds, and so we should expect a Small World to have dif-
ferent laws. For another, it seems that Bohmian dispositionalism is subject
to a serious dilemma: either time must be regarded as discrete rather than
continuous, or the powers of the particles fail to determine the particles’
trajectories (Simpson and Pemberton, 2021).
A dispositionalist might propose that the plurality of particles has a
single, collective property – namely, the property of instantiating a power
to choreograph their trajectories according to the Bohmian law. This
property would be intrinsic to the plurality of particles, rather than to any
of the individual particles. Yet such an account faces two metaphysical
challenges (Simpson, 2021). In the frst place, the Bohmian law of motion
is specifed not merely in terms of the positions of all the particles but in
terms of a universal wave function, which does not supervene upon the
particle confguration but evolves according to the Schrödinger equation.
To explain the law-like behaviour of the particle confguration, this global
power would have to persist through time and ground the lawfulness of
the Schrödinger equation. In the second place, in order to capture intui-
tively correct counterfactuals about what would happen in Small Worlds
which have only a few particles (Carroll, 1994; Demarest, 2017), the
identity of this global power should not depend upon the number of par-
ticles in the global confguration. There would have to be possible worlds
in which the same global power is instantiated by a diferent number of
particles. What we require, then, is a metaphysics that can explain both
the diachronic and transworld sameness of this collective causal power.
Cosmic Hylomorphism provides such a model, in which the only funda-
mental physical entity is the cosmos.
3.2.3. Cosmic Form

According to the Cosmic Hylomorphist, the cosmos is a substance which
is metaphysically composed of both matter and a Cosmic Form (Simpson,
2021; cf. Schafer, 2010). The matter of this substance may be identifed
with the Bohmian particles, which are endowed with causal powers to
change their velocities. However, these particles are not substances which
possess their own essential and intrinsic powers. Rather, their causal pow-
ers are metaphysically grounded in the Cosmic Form, which manifests
a cosmic process.43 The material particles, thus empowered, are trans-
formed into the integral physical parts of a cosmic whole.
In this model, Cosmic Form is a simple and fundamental particular
with the power to manifest a cosmic process (cf. Koons, 2018). The Cos-
mic Form is not located in physical space, along with the particles it
transforms, since it does not stand in any distance relations, nor is it the
efcient cause of their motion, since they have their own causal powers
to bring about change. Rather, the Cosmic Form brings about change
indirectly, in the course of manifesting the cosmic process, by grounding
the powers of the particles at each moment of time. In so doing, the Cos-
mic Form unites itself to all of the particles to compose a single substance
with an intrinsic power to choreograph their trajectories according to the
Bohmian law.44 It is the Cosmic Form that explains the diachronic and
transworld sameness of this collective power.
Although this neo-Aristotelian account of Bohmian mechanics departs
in signifcant ways from Aristotle’s hylomorphic account of nature, as
it was understood by Aquinas, it nevertheless deploys Aristotle’s four-
fold conception of causation to provide a non-mechanical explanation
of quantum phenomena. The similarities with the ancient doctrine of
hylomorphism are evident in at least three respects.
In the frst place, the material particles that it posits are at least analo-
gous to Aristotle’s concept of matter: by supplying a persisting substrate
which has the potential to bear causal powers, they serve as the material
cause of the cosmic whole. The Cosmic Substance has no physical exis-
tence apart from its matter. Unlike Aquinas’s concept of prime matter, the
particles are metaphysically discrete and stand in distance relations. Prime
matter is more than a substratum of bare potentiality, on this account,
since it is spatially extended. The particles constitute a materia prima,
however, in the restricted sense that they lack any intrinsic physical fea-
tures. The phenomenon of quantum entanglement gives us good reason
to doubt that particles have intrinsic physical properties independently of
the wholes of which they are parts (Section 3.1).
In the second place, the Cosmic Form is analogous to Aristotle’s concept
of substantial form: by grounding the causal powers of the particles, the
Cosmic Form acts as the formal cause of the Cosmic Substance, causing it
to be the kind of unifed substance that it is. Unlike Aristotle’s account of
the forms, there is only one substance in nature, and none of the objects
of ordinary experience – including biological entities – are said to have
substantial forms. A ‘formal cause’ that operates instantaneously upon
matter can provide a non-mechanical explanation of the non-local cor-
relations associated with quantum phenomena, however, since it is only
efcient causes which are mediated in space and time. The Cosmic Form
can explain the motion of the Bohmian particle confguration without
violating the ‘superluminal ban’ in physics.
In the third place, this account reintroduces an element of teleology
within physics: the ultimate explanation for the motion of the Bohm-
ian particles is that the activity of the Cosmic Substance is a teleological
process. The ordering of this process cannot be explained mechanically
in terms of laws that connect one set of properties instantiated at time t
with another set of properties instantiated at time t > t, since the wave
function is not an element of the fundamental ontology and the particles
do not have intrinsic physical properties which stand in law-like relations.
Rather, the ordering of this process is only explained by taking a teleologi-
cal stance, in which the temporal development of the Cosmic Substance
has an intrinsic direction (that is, the substance admits a fnal cause).
Unlike Aristotle’s concept of fnal causes, the fnal cause of the Cosmic
Substance can be described mathematically in terms of the boundary con-
dition on a universal wave function defned in an abstract confguration
space, which contains information about possible trajectories of particles
in diferent possible worlds. However, the universal wave function of the
cosmos is not a property that we can measure.
Cosmic Hylomorphism is of course a far cry from that medieval vision
of nature in which many of the objects of ordinary experience, including
human beings, were substances with their own natures. Nonetheless, it
represents a decisive break with the microphysical reductionism that
dominated the philosophy of the last century and the corpuscularian
assumption that nature wears all her properties on her sleeves for sci-
entists to inspect. According to the Cosmic Hylomorphist, the cosmos
is unintelligible apart from its substantial form, which grounds the
properties of its matter, but the only way in which its matter and form
can be separated is by metaphysical abstraction rather than scientifc
manipulation.
Cosmic Hylomorphism ofers a way of understanding what the world
might be like, if the laws of Bohmian mechanics were true, in which the
universal wave function is not an element of the fundamental ontology, but
represents an intrinsic power of a Cosmic Substance that is grounded in
its substantial form. Yet there are reasons to doubt the claim that the cos-
mos can be represented as a single quantum system with a universal wave
function – even in principle – which I shall discuss in the following section.
4. Form in Quantum Statistical Mechanics

The phenomenon of quantum entanglement is not the only reason to recon-
sider the supposed redundancy of hylomorphism. The microphysicalism
that dominated analytic philosophy in the last century has also run into
difculties in accommodating novel and robust phenomena that emerge
at higher scales. According to the standard monistic approach to realism
adopted by many philosophers, a physical theory is supposed to provide
a universally true and exhaustive description of the world in all physical
respects and at all physical scales. However, the turn toward scientifc prac-
tices in the philosophy of science has given rise to a pluralistic understand-
ing of scientifc theories, which has called into question the philosopher’s
aspiration of advancing a fundamental physical ontology. In what follows, I
shall discuss the dilemma that an alternative pluralistic approach to realism
faces between pragmatism and reductionism, in the context of quantum
statistical mechanics, and will propose a way of splitting the horns of this
dilemma by adopting a hylomorphic approach to realism that admits a plu-
rality of fundamental substances (instead of a single Cosmic Substance).45
4.1. A Question of Context
4.1.1. Unitarily Inequivalent Representations

According to the standard approach to scientifc realism, a scientifc the-
ory’s explanatory virtues gives us good reason to attribute (some aspect
of) that theory with representational content. There is a striking disparity,
however, between the standard approach to realism and the way that
quantum theories are used to explain phenomena in practice. For the
purposes of this chapter, I shall focus on phenomena described by the
theory of quantum statistical mechanics, although the same difculties
arise in quantum feld theory.46
At the core of any quantum theory are the canonical commutation rela-
tions between conjugate quantities such as position and momentum – or
the anticommutation relations that hold between the Pauli spin matrices –
which encode Heisenberg’s uncertainty principle. Any quantum theory
which specifes a quantum state ˜ of a physical system defned in a Hil-
bert space H, and a set of bounded self-adjoint operators corresponding
to observables (H, O ˆ ) which act upon the quantum state O ˆ ˜ , must
i i
realise the Weyl algebra associated with these relations. When the opera-
tors on a Hilbert space conform to these commutation relations, they are
said to be a ‘representation’ of these relations.
Unitary equivalence is widely considered the standard of empirical
equivalence: if two representations are unitarily equivalent, there is some
unitary operator Uˆ that transforms one representation into the other
U: H → H, such that they both determine the same expectation values
for the various observables which they defne: U −1O ˆ U = Ô . However,
i i
the theory of quantum statistical mechanics generates a continuum of uni-
tarily inequivalent representations in the so-called thermodynamic limit
(Ruetsche, 2002, 2003, 2006, 2013), where it is necessary to adopt models
which have infnite degrees of freedom in order to describe many kinds of
physical phenomena. Whilst the Stone-von Neumann theorem establishes
that any pairs of distinct representations for a fnite system will be unitar-
ily equivalent to the irreducible Schrödinger representation, since there
is a unitary operator that transforms one into the other, infnite systems
admit infnitely many Hilbert-space representations which fall outside of
the scope of the Stone-von Neumann theorem (Ruetsche, 2011, chs. 2–3).
Let us consider the example of a ferromagnet. When a physical system
experiences a phase transition, certain properties of the system undergo
discontinuous change due to some macroscopic change in their immedi-
ate external conditions. An iron bar that is at thermal equilibrium, for
instance, exhibits a paramagnetic phase above a critical temperature
T ≥ Tc, in which it experiences no net magnetisation. Below this critical
temperature, however, it exhibits a ferromagnetic phase, in which it expe-
riences spontaneous magnetisation. In the presence of an external mag-
netic feld, the ferromagnet admits two possible metastable states which
are characterised by opposite magnetic polarisations. These two states, as
it turns out, must be defned using unitarily inequivalent representations.
Proof: To demonstrate this inequivalence for the more mathematically

minded (others may prefer to skip this argument), suppose we set up
a Hilbert space for a system whose ground state is characterised by
a sequence sk = +1 for k ∈ Z, then add all the sequences that can be
obtained by making fnitely many local modifcations to this sequence
which replace some of the entries with −1.47 Let us label this Hilbert space
H+. A set of operators ˇˆ zj may then be introduced such that sequences sk
whose jth entry is ±1 correspond to the eigenvector in the Hilbert space
associated with the eigenvalue ±1. A magnetic polarisation observable
m̂ can be defned for the composite system with the components:
k= N
1
m̂iN = ˇˆ ik , i {x, y, z} ,
2N + 1 k=−N
(8)
which has a limit N → ∞ in the weak topology of H+. Let [sk]j ∈
{−1, +1} denote the jth entry in the sequence sk. For every state, the
expectation value of m̂ will be oriented along the z axis and will take
the value of +1 in the thermodynamic limit, breaking the rotational
symmetry of the system’s dynamics, since only a fnite number of
spins take the value −1. But what about states that break this sym-
metry in the opposite direction? For these cases, we begin with a
ground state characterised by a sequence sk = −1 for k ∈ Z, adding
all the sequences that can be obtained by making fnitely many local
modifcations which replace some of the entries with +1. Let us label
this alternative Hilbert space H−.
It is obvious that our two representations of the ground state
are inequivalent. The proof may proceed by contradiction. Sup-
pose there is some unitary transformation, U: H+ → H−, such that
N−
Uˇˆ +nU −1 = ˇˆ n− for all n, which implies that m̂ = Um ˆ N +U −1 , and
+ −
suppose that ˜ and ˜ are unit vectors in the Hilbert spaces
H+ and H− respectively. Assuming that these two vectors are related
by the transformation |ψ+> = U−1 |ψ−>, it follows that
˜ + mzN + ˜ + = ˜− mzN− ˜− . (9)
However, this identity does not hold in the thermodynamic limit N → ∞,

since the right-hand side evaluates as +1 and the left-hand side as −1.
These two representations do not admit a unitary transformation,
and the Hamiltonians defned on these two infnite models describe
physically diferent situations.
Nonetheless, the use of infnite models defned in the thermodynamic

limit, which deploy unitarily inequivalent representations, turns out to
be necessary for empirical adequacy in describing quantum systems.48
The statistical physics of fnite systems identifes equilibrium states with
unique Gibbs states (Ruetsche, 2011, p. 3), implying that the phase avail-
able to a system at temperature T is unique for all T. Yet this is contrary to
what we observe in experiments. According to Ruetsche, it is ‘only in the
thermodynamic limit [that] one can introduce a notion of equilibrium that
allows what the Gibbs notion of equilibrium for fnite systems disallows:
the multiplicity of equilibrium states at a fnite temperature implicated in
phase structure’ (p. 3). It is a detail which the more pragmatically minded
physicist will doubtless be tempted to gloss over, but one which a scientifc
realist in search of an ontology will have to take into consideration.
4.1.2. Scientifc Pluralism

According to the standard approach to realism, a realist should be preoc-
cupied with the counterfactual question: what would the world be like if
this scientifc theory provided a universally true and exhaustive description
of the physical world? According to Ruetsche, however, it is incumbent
upon any would-be interpreter to answer the question: ‘does this interpre-
tation allow the theory to discharge all of its scientifc duties?’ The problem
with the standard approach to realism, from this practical standpoint, is
that ‘there often isn’t a single interpretation under which a theory enjoys the
full range of virtues realists are wont to cite as reasons for believing that
theory’ (Ruetsche, 2011, p. 5). In practice, physicists deploy a plurality of
inequivalent representations to capture the phenomena in which they are
interested, for which the correct choice is context-dependent.
In our example of the ferromagnet, it is evident that the quantum theory
that describes the behaviour of the physical system admits of (at least)
two representations, and that these representations are not empirically
equivalent. There is no obvious reason for preferring one to the other, and
there is good reason to doubt that they share a common structure which
completely determines the physical content of both representations. How
should we proceed?
On the one hand, to privilege the physical content of one particular
representation (‘Hilbert space conservatism’) would be to reduce the num-
ber of physically signifcant states to a subset of those that are generally
accepted within successful scientifc practices (Ruetsche, 2003). On the
other hand, to confne the physical content of a quantum theory to the
algebraic structure that is shared by diferent Hilbert space representa-
tions (‘algebraic imperialism’) would be to reduce the number of physi-
cally signifcant observables that are measured within successful scientifc
practices. Either of these moves fails to support the explanatory agenda of
our best quantum theories and is inconsistent with the practice of adopt-
ing a realist stance toward physical theories in virtue of their explanatory
successes. It seems there is no single universal representation in which all
the observables of a complex system can be defned and in which they
evolve in a continuous way.
Porter Williams argues that philosophers should content themselves
with adopting an efective realism (Williams, 2017). According to efective
realists, physical theories are not absolutely true in all physical respects and
at all physical scales, but that should not matter to philosophers of science,
as they are only ever applied within limited regimes in successful scientifc
practices. The task of the efective realist is to identify elements described
by an efective theory that are ‘stable and robust’, inasmuch as they are
detectable and measurable and ‘can be expected to survive future episodes
of theory change’ (p. 218). It is these elements to which we should be onto-
logically committed.49 Williams articulates his brand of efective realism
within the context of quantum feld theory, which is widely acknowledged
to be an efective feld theory that breaks down at small scales and that
depends upon the practice of renormalisation to secure limited regimes
within which empirically adequate models can be constructed.
There is much to recommend this pluralistic approach to realism to
metaphysicians seeking to be guided by scientifc practices. As Williams
observes, the attempt to extract a single fundamental ontology from a
physical theory, which characterises the standard approach to realism,
‘leaves one with an interpretation unequipped to support the theory in
the performance of its explanatory duties’ (p. 228), whereas a pluralistic
approach to realism which acknowledges the context-dependence of our
interpretations opens the prospect of uncovering ‘a rich, layered ontology’
that is otherwise hidden from the standard realist (p. 220). There is also
something salutatory about his exhortation to technically minded phi-
losophers of physics that the business of constructing ontologies ‘contains
considerable amounts of art as well as science’ (p. 233). It is questionable,
however, whether efective realism can steer a middle course between
the Scylla of pragmatic empiricism and the Charybdis of microphysical
reductionism.
On the one hand, it is unclear why philosophers should prefer an efec-
tive realism that embraces a complex ontology to a pragmatic empiri-
cism that shuns ontological commitment, on the basis of the reasons put
forward by Williams. As Ruetsche points out, both approaches oppose
physical fundamentalism by seeking to disentangle the question of explan-
atory power from the question of universal truth (Ruetsche, 2020). The
philosopher who is seeking to extract a fundamental ontology from some
physical theory T ‘has to endow a physical theory with explanatory self-
sufciency. Otherwise, something exogenous to T is required to explain
why T works in some regimes but not others, and why things aren’t quite
the way T says they should be’ (p. 308). The efective realist, however,
believes we can construct explanations of T’s success which make no
appeal to T’s truth, since we have good reasons to think that T is only an
efective theory, and hence to reject the fundamentalist’s assumption of T’s
explanatory self-sufciency. According to Ruetsche, ‘Fundamentalism’s
sin isn’t to interpret T. It’s to decide that, having interpreted T, no further
explanatory work remains – and thereby suppress the possibility that T
is merely efective’ (p. 309). Ruetsche argues that efective realism can be
exchanged for a humble empiricism which eschews ontological commit-
ment without sacrifcing explanation. ‘The humble empiricists answer is
that our theory succeeds as well as it does where it does, not because it’s
true, but because, whatever the true theory is, our theory approximates it
in its domain of application’ (p. 310).
On the other hand, it is unclear why philosophers who are strongly
committed to realism over pragmatic empiricism should prefer an efec-
tive realism that produces a fragmented picture of reality to some form of
weak emergence that maintains a commitment to the unity of nature. It
may be argued that the interpretation of quantum theories only appears
to be context-dependent because we are cognitively incapable of model-
ling complex systems without introducing approximations, and that the
existence of unitarily inequivalent representations of quantum theories is
merely an artefact of reifying the infnite degrees of freedom which are
introduced in the thermodynamic limit (or in quantum feld theories).
After all, it is doubtful that many physicists take any infnities literally,
regarding them rather as idealisations or abstractions (Ellis et al., 2018).
According to Wallace (2011), the problem of unitarily inequivalent rep-
resentations is circumvented in conventional quantum feld theory by the
introduction of cutofs which limit its application to systems with only
fnite degrees of freedom. Yet, as Fraser (2009) has argued, the necessity
of cutofs suggests conventional quantum feld theory is only an efective
feld theory and unsuitable for standard interpretation. In that case, it may
be urged, realists will have to wait for some better theory.
4.2. Hylomorphic Pluralism

For the contemporary philosopher of science, it may seem as though the
options lie between some kind of fragmented pluralism, which rules out a
single unifed account of how everything hangs together, or some sort of
eschatological monism, which allows the possibility of a theory of every-
thing but admits that such a theory may lie forever beyond our grasp. Yet
there may be a third way, which afrms the plurality of physical being
and the existence of entities at diferent scales, without abandoning the
commitment to a fundamental ontology.
4.2.1. Contextual Wave Function Collapse

I wish to consider the possible existence of macroscopic entities between
the microscopic and the cosmic scale, which are neither integral parts of
a single Cosmic Substance nor reducible to a single set of microscopic
constituents, but which may nonetheless be said to have microscopic
parts. For an Aristotelian who is committed to the Eleatic principle, a
substance which has fundamental physical being is marked by its posses-
sion of irreducible causal powers (Section 2). A macroscopic substance
would have to have causal powers over and above the powers of any of its
constituents to be included in the fundamental ontology. Yet is there room
in contemporary physics, which is widely supposed to concern the proper-
ties of point-sized particles and felds that make up everything else, for
the existence of substances which have irreducibly macroscopic powers?
In fact, there are many cases where microscopic physical theories
contain physical quantities which are not determined by the theory but
depend upon large-scale properties of the physical system. For example,
most physical theories are written in terms of diferential equations, yet
physical models can only produce testable predictions once boundary
conditions for these equations have been specifed. These physical bound-
ary conditions often have a macroscopic origin.
The standard approach to realism downplays the role of macroscopic
properties in specifying the boundary conditions of physical systems,
because it assumes that the total state of the physical world can be
uniquely represented by our best physical theory as a single closed sys-
tem which evolves according to the universal laws of our ‘best physics’.
For example, in the de Broglie-Bohm version of quantum mechanics, the
temporal development of the matter is determined for the whole confgu-
ration by a law of nature which is specifed solely in terms of microscopic
quantities (positions) and a global quantity (the wave function) defned
at the cosmic scale. The evolution of the total confguration of matter
does not depend upon any macroscopic (intermediate-scale) quantities.
Yet there are good reasons, as we have seen, to doubt whether there is an
interpretation of quantum mechanics that is true in all physical respects
and at all physical scales.
An alternative ‘contextual’ model of the quantum dynamics is avail-
able, however, proposed by Barbara Drossel and George Ellis, in which
the interaction of a quantum system with the intrinsic heat bath of a
macroscopic system, such as the measuring device that a scientist might
deploy in a laboratory, plays a key role in solving the measurement prob-
lem (Drossel and Ellis, 2018). In this interpretation, quantum systems
are individual systems, which are ‘open’ to the infuence of a ‘classical’
environment that cannot be modelled as a single quantum system, and
the macroscopic, thermal properties of certain features of the environment
have the power to collapse their wave function.
The CWC model (contextual wave function collapse) drops the assump-
tion of physical monism that underpins the choice between GRW theory
and Bohmian mechanics (Section 3.1.2) by denying that the theory of
quantum mechanics is endowed with explanatory self-sufciency and
allowing something exogenous to quantum mechanics to complete its
physical description in diferent contexts. According to Drossel, a ther-
malised system cannot be described by a many-particle wave function
(Drossel, 2017). There is more to the intrinsic heat bath of a macroscopic
system than the sum of its microscopic parts. Whilst this approach to
quantum mechanics has certain features in common with the GRW and
Bohmian theories, it also difers from them in signifcant ways.
As in the GRW modifcation of quantum mechanics, the CWC model
seizes the frst horn of Bell’s dilemma, allowing the wave function to
become localised with respect to position. It likewise distinguishes sci-
entifc measurements from localisation events in general, removing the
necessity of a scientifc ‘observer’ in order for there to be any facts about
the physical properties of quantum systems. Unlike GRW theory, how-
ever, the corrections that achieve these localisations of the wave func-
tion depend upon the macroscopic context of a physical system, which
includes systems which instantiate macroscopic thermal properties, rather
than the introduction of an additional collapse mechanism. In short, the
CWC model proposes incorporating a feedback loop – from a particle,
via the intrinsic heat bath of a macroscopic system, back to the particle –
which introduces non-linear terms into the Schrödinger equation:
ˇ˜
i˜ = H˜ + f (˜ ) ˜ (10)
ˇt
where H is the Hamiltonian of an idealised closed system and f is a non-
linear contribution due to the external environment. This extra term is
physically motivated: it can be accounted for in terms of thermodynamics
and solid-state physics (Drossel and Ellis, 2018, pp. 13–19).50
As in Bohmian mechanics, the CWC model relies upon the efects of the
environment upon the measuring process to explain why the outcomes of
quantum experiments, such as the EPR experiment (Section 3.1), conform
to standard quantum statistics and obey Born’s rule for connecting quan-
tum ‘observables’ with the wave function of a physical system. Unlike
Bohmian mechanics, the CWC model does not conceive the environment
that is relevant to the measuring process in terms of a many-particle sys-
tem with a wave function that is subject to the unitary and reversible time
evolution described by Schrödinger’s linear dynamics. In fact, the heat
bath of any fnite temperature system that is capable of collapsing the
wave function is characterised as having only a limited ‘memory’, since it
radiates irreversibly into the heat sink of its surroundings. Consequently,
the CWC model does not leave any physical system entangled with any
part of its environment beyond the usual time scale of decoherence, plac-
ing a limit on the extent of quantum entanglement in nature. According
to the CWC model, the heat bath of a macroscopic instrument of mea-
surement serves as a bridge between quantum systems and their classical
environment, since the heat bath of a macroscopic system can induce the
collapse of the wave function.
The CWC theorist agrees with the GRW and Bohmian theorists that
standard quantum mechanics is physically incomplete. Unlike GRW
theory and Bohmian mechanics, however, CWC theory does not ofer a
universal way of completing quantum mechanics that purports to describe
every physical aspect of the world at every physical scale, but restricts
itself to describing the quantal properties of open quantum systems whose
dynamics depend upon macroscopic features of their classical environ-
ments. From this standpoint, a purely quantal system (in which f → 0
in (10)) is an idealised isolated system that does not exist in the real
world. Drossel and Ellis believe that we should reject ‘the untestable and
implausible claim that the environmental heat bath can be described by
an infnite-precision wave function that is subject to unitary time evolu-
tion’ (p. 4).
In this contextual approach to quantum mechanics, classical properties
are supposed to be higher-level, strongly emergent properties of physical
systems, which have top-down causal powers to change the microscopic
properties of systems. They derive these causal powers from the funda-
mental role that they play in defning the Hilbert spaces and the time
scales within which the unitary time evolution of an open quantum system
takes place.
4.2.2. Rival Views of Macroscopic Entities

Yet how are macroscopic properties like temperature and chemical
entropy, which characterise complex physical systems like measuring
devices, supposed to ‘emerge’ from simpler microphysical systems, and
what is the nature of the microphysical substrate from which a macro-
scopic entity is supposed to emerge? Drossel and Ellis’s commitment to
strong emergence may seem to confront an ontological dilemma between
microphysical reductionism or substance dualism.
On the one hand, suppose the causal powers of an emergent macro-
scopic whole can be explained entirely in terms of the properties and
causal powers of its microscopic parts. These microphysical parts do not
depend upon the whole for their physical properties but have their own
intrinsic natures. In that case, the emergent macroscopic whole must be
ontologically reducible to or supervenient upon the aggregation of its
microscopic parts and should not be counted among the fundamental
entities of nature. Yet CWC theory does not appear to be compatible with
microphysical reductionism, since it provides the macroscopic properties
of one system (such as a measuring device) with an irreducible role to
play in collapsing the wave function of another system, thus endowing
macroscopic properties with top-down causal powers.
It may be argued that the macroscopic properties of an emergent
whole do not locally supervene upon its microscopic constituents, but
globally supervene upon nothing less than the entire cosmos: the cos-
mos as a whole is an emergent entity with its own intrinsic nature. In
that case, the macroscopic emergent whole should be regarded as an
integral part of the physical cosmos and should not be counted among
the fundamental entities of nature. Such entities are merely artefacts
of the boundary conditions that we impose upon physical systems. Yet
CWC theory does not seem to be compatible with cosmic holism, since
it rejects the claim that the cosmos as a whole has a universal wave func-
tion as an ‘untestable and implausible’ assumption (Drossel and Ellis,
2018, p. 4) and insists on adopting a contextual approach to the wave
function’s dynamics.
On the other hand, suppose that the activities of an emergent macro-
scopic whole cannot be explained in terms of the causal powers of its
microscopic constituents, but that the macroscopic whole has a novel
and irreducible causal power which directly acts upon the microscopic
constituents and causes them to change their collective behaviour. In that
case, there would be good reason to count this macroscopic entity as a
separate entity that interacts with these microscopic entities rather than
being composed of them (Gillett, 2016, p. 247). If the relation between
this macroscopic entity and these microscopic entities is merely a causal
relation between diferent entities which have their own intrinsic causal
powers, then the macroscopic entity and the microscopic entities are
ontologically independent substances, since they do not depend upon one
another for their physical natures. Yet CWC theory does not appear to
be compatible with a dualism of microscopic and macroscopic entities,
since it does not supply a complete quantum mechanical characterisation
of any microscopic entity but only characterises quantum systems within
diferent macroscopic contexts.
It may be argued that the emergent causal power which is exercised by
the macroscopic whole is not a fundamental power of the whole, which
has its own nature and primitive identity, but is a ‘structural power’ that
is instantiated by its microscopic constituents, which arises from a kind
of conspiracy that occurs between them whenever they are spatiotem-
porally confgured in a certain way. In that case, we would have good
reason to count this macroscopic entity as a functional whole, since its
behaviour would be a non-linear function of the intrinsic properties of its
microscopic constituents. Such macroscopic entities are not fundamental
substances, as they do not constitute any addition to being over and above
their constituents. Yet CWC theory leaves us unenlightened concerning
the microscopic domain of these non-linear functions, since it has nothing
to say about the properties of microscopic constituents independently of
their diferent macroscopic contexts. What we require, then, is a meta-
physics that can elucidate a non-causal relation of composition between
the emergent macroscopic whole and its microscopic parts. Hylomorphic
Pluralism provides such a model.
4.2.3. Substantial Forms

Hylomorphic Pluralism afrms the existence of a variety of substances
in nature which are metaphysically prior to their physical parts. These
substance are not built of entities which possess intrinsic physical natures
apart from the substances of which they are parts. Rather, their material,
spatially defned parts are ontologically dependent for their causal pow-
ers, and hence their physical natures, upon the wholes of which they are
parts. Hylomorphic Pluralism, I suggest, can make sense of the existence
of wholes with ‘emergent’ macroscopic causal powers.
In the frst place, Hylomorphic Pluralism ofers an alternative account
of composition to microphysical reductionism. In this model, a macro-
scopic entity which has irreducible causal powers is not an arrangement
of microphysical constituents which have their own intrinsic physical
natures. Rather, this entity is metaphysically composed of both mat-
ter (which has no intrinsic causal powers) and substantial form (which
determines the powers of the substance): the essential powers of the sub-
stance are metaphysically grounded in its substantial form.
Unlike the Cosmic Hylomorphist, the Hylomorphic Pluralist does not
posit the existence of a single Cosmic Substance, of which everything else
in the world is only a part, but admits a plurality of diferent substances,
which exist at diferent physical scales. Among the substances which make
up the physical world, it may be supposed, are some of the objects of
ordinary experience that were familiar to the investigations of Aristotle,
such as plants, animals, and human beings, although I shall leave discus-
sion of the biological world to the next chapter.
In the second place, Hylomorphic Pluralism is able to avoid dualism
by providing an account of the composition of a metaphysically unifed
whole in which the parts are not themselves substances. In this metaphysi-
cal model, the world is not divided into microscopic entities that have
quantal properties, on the one hand, and macroscopic entities that have
classical properties, on the other hand, which somehow interact with
one another despite being governed by incommensurate laws. Rather, we
can think of the (inorganic) world as being made of thermal substances
which have both quantal and classical properties, as Koons and I have
suggested elsewhere (Koons, 2019; Simpson, 2019), which are metaphysi-
cally composed of matter and substantial form. A thermal substance,
on this view, is not a unifed and fundamental physical whole because
it micromanages its microphysical parts by pushing them around, and
the relation between its substantial form and its material parts is not an
(efcient) causal relation. Rather, the substantial form is the formal cause
of the unity of the substance by determining the powers of its parts, and
the relation between the substantial form of the substance and its material
parts is one of metaphysical grounding.
Unlike the Cosmic Hylomorphist, the Hylomorphic Pluralist does not
conceive of physical substances as existing in a state of causal isolation,
but allows substances to interact with one another through the exercise
of their causal powers. A substance therefore has an accidental form, in
addition to its substantial form, which is subject to change as it interacts
with other substances.51 The causal powers through which one substance
interacts with another substance depend jointly upon both its substantial
form and its accidental forms.
I shall outline briefy how this hylomorphic metaphysics might be
applied to a contextual approach to quantum mechanics, whilst restrict-
ing the scope of this toy model to thermal phenomena. According to this
metaphysical model, the physical world is ‘tiled’ with thermal substances
which have intrinsic causal powers,52 such that any change in the physi-
cal world necessarily involves a change in one or more of these physical
substances, which is brought about through the exercise of their causal
powers. The matter of a substance Si is a parcel of gunk mi(x,t).53 The
matter-felds of two substances Si and Sj are distinct from one another in
virtue of the prior distinctness of their two parcels of gunk mi(x,t) and
mj(x,t); they do not share any metaphysical parts. The metaphysical sub-
strate from which physical substances are carved lacks any intrinsic causal
powers: it only has the potential to bear causal powers. A parcel of matter-
density must be combined with a substantial form to compose a physical
substance. A parcel of matter-density which – per impossibile – were not
combined with a substantial form would be a ‘compositional zombie’ that
lacked any causal powers. When substances interact with one another by
exercising their causal powers, they exchange matter-density with each
other’s matter-felds.
We may think of a substance as having an internal matter fow which is
choreographed by a wave function via the matter-feld equation (7). The
wave function evolves according to a non-linear Schrödinger equation
(10), where the non-linear terms depend on the macroscopic context in
which the substance is situated. There is no universal wave function in this
model: it is only possible to defne a wave function in those circumstances
in which the system has been isolated from the noise of its environment
and its boundary conditions are relatively stable. The laws in this model
are thus ‘dappled’ laws (to use Cartwright’s nomenclature), since they are
not universal regularities but are restricted to the ‘nomological machine’
constructed by the experimentalist (Cartwright, 1999). However, the
Hylomorphic Pluralist is committed to a realist approach to quantum
mechanics: the wave function governed by the non-linear Schrödinger
equation represents a real power of the substance to regulate its own
parcel of matter, which is grounded in its substantial form.54 We might
think of this power as a multi-track power, where each track corresponds
to a diferent macroscopic context.
In addition to having a power to regulate their own parcels of matter,
the thermal substances comprising the environment of any quantum
system have causal powers to bring about changes in the matter of other
substances. These macro-level powers of the substance are sustained
in existence by the micro-level parts of the substance. The matter of a
thermal substance, however, does not relate to space and time merely
as a set of discrete physical units, and the macroscopic properties of a
thermal substance at a particular time are not reducible to the spatial
arrangement of its matter at a particular time: the material parts of the
substance also ‘cooperate’ with one another over time as an undivided
continuum, generating dynamical situations which can only be modelled
using Hamiltonian functions defned on an infnite (continuum) model,
cf. Koons (2019) and Simpson (2019). The proper way of taking this
limit for a given substance (that is, the correct choice of physical repre-
sentation) is determined by the substantial form of the substance, which
confers a monadic nature upon the substance as a whole and grounds
the powers of its micro-level parts.
The Hylomorphic Pluralist is committed to a realist approach to the
macroscopic properties of physical substances, but one that resists micro-
physical reduction: each thermal substance has a Hilbert-space represen-
tation that defnes macroscopic observables for its emergent properties,
but diferent substances admit diferent representations. The macroscopic
properties of a thermal substance cannot be explained mechanically in
terms of laws that connect one set of microscopic properties instanti-
ated at time t with another set of microscopic properties at time t´ >
t, where these microscopic properties are defned in terms of a unique
representation which characterises the total state of the physical world.
Nonetheless, there is no need to suppose that the macroscopic powers
of thermal substances demand new fundamental forces (or non-physical
‘confguration forces’) in order to make a causal diference to where their
matter ends up, since the quantum dynamics is context-dependent and the
macroscopic properties of thermal substances play a role in defning the
Hilbert spaces and time scales within which the unitary time evolution of
quantum systems takes place.
The Hylomorphic Pluralist, unlike the efective realist, can thus steer
a course between pragmatic empiricism and microphysical reductionism.
On the one hand, the Hylomorphic Pluralist can agree with the prag-
matic empiricist about the disconnect between the explanatory power of
a theory and the universal truth of its description, whilst maintaining an
attitude of ontological commitment. Quantum theories are not successful
because they uncover a single set of microscopic constituents of which
everything is made, but because they isolate certain causal powers of
substances which manifest under certain conditions. These causal powers
are real properties of diferent thermal substances, but they are metaphysi-
cally grounded in their substantial forms.
On the other hand, the Hylomorphic Pluralist can agree with the micro-
physicalist that the thermodynamic limit is an idealisation, without aban-
doning an ontological commitment to the reality of macroscopic entities.
Quantum theories about fnite temperature systems are not successful
because they truthfully represent the degrees of freedom in a physical
system but because their diferent representations capture something of
the diferent ways in which their material parts cooperate in determining
the thermal properties of a system. These properties are real properties
of thermal substances, which are grounded in their substantial forms.
The necessity of adopting a change in the physical representation of a
system – when it undergoes a phase transition, for example – is due to
the occurrence of a substantial change taking place in the system,55 which
is a discontinuous change in which the microscopic powers of a parcel of
matter are redetermined by a diferent substantial form.
The Hylomorphic Pluralist, like the efective realist, afrms the existence
of entities which exist at diferent physical scales, but remains committed
to a fundamental ontology of substances which are metaphysically com-
posed of matter and form. Matter and form are afrmed for the sake of
both empirical and explanatory adequacy, rather than being read of the
formalism of quantum physics or isolated in a scientifc experiment. We
can only reach them by metaphysical abstraction in the context of an
empirical inquiry that afrms the reality of change (Section 2.2). Like the
people in Plato’s cave, we fnd ourselves unable to look directly at the
forms, which cast shadows on the wall of our cave. Unlike the situation
in Plato’s cave, however, these forms are immanent within substances
which we can manipulate within diferent scientifc practices, and they
cast diferent shadows depending on how they are manipulated. The role
of the metaphysician is like that of someone who infers the shape of an
object from the variety of shadows that it projects as it is held to the light
in diferent ways.
5. Concluding Remarks
At the beginning of this chapter, I observed that the Aristotelian concept
of causal powers, which was widely rejected during the Scientifc Revolu-
tion, is now at the forefront of contemporary philosophical discussions.
In the course of this chapter, I have argued that it also time to reclaim
the Aristotelian doctrine of hylomorphism. By restoring the concept of
substantial form, it is possible to provide a realist account of quantum
physics in terms of an ontology of powers.
The two hylomorphic models which I have discussed correspond to two
incompatible versions of quantum mechanics: Cosmic Hylomorphism
ofers an ontological account of Bohmian mechanics, which attributes
the cosmos a universal wave function and a global confguration of par-
ticles (Section 3); Hylomorphic Pluralism ofers an ontological account of
CWC theory, which divides the world into separate parcels of gunk and
requires the wave function of quantum mechanics to collapse (Section
4). Granted the unitarian doctrine of substantial form, which Aquinas
afrmed and Scotus rejected, the ontologies of these models would seem
to be mutually exclusive: if there is a Cosmic Substance, of which every
particle is a part, there can be no thermal substances, on pain of violating
the unity of the Cosmic Substance – unless the thermal substances could
be embedded in a Cosmic Substance without being mereological parts of
it (cf. Dumsday, 2016). We might think of the cosmos as beginning as a
single substance, from which macroscopic thermal substances are subse-
quently derived.56 There is scope here for developing a neo-Aristotelian
cosmology.
Both of my metaphysical models, I suggest, also hold interest for con-
temporary discussions of theology and science, although each of them
has its limitations. Cosmic Hylomorphism opens the prospect of reviv-
ing something like the neo-Platonic conception of a ‘world soul’, which
could play a role in a scientifcally informed account of special divine
action that is able to resist common objections (Dumsday, 2019a). The
global perspective of Cosmic Hylomorphism, however, encounters dif-
fculties in explaining the emergence of specifc physical entities, which
a localist perspective like Hylomorphic Pluralism can readily accom-
modate. Although the sketch of Hylomorphic Pluralism that I have
ofered in this chapter only includes thermal substances, the following
chapter shows how to extend this account to include biological as well
as inorganic entities, gaining the explanatory advantage of being able
to include human beings within nature as embodied and responsible
agents, created in imago dei.
In putting forward these two hylomorphic models, however, I do
not wish to suggest that the quantum theories for which they ofer
ontologies are without their difculties or theoretical costs: Bohmian
mechanics only secures agreement with the Born rule of standard quan-
tum mechanics via the quantum equilibrium hypothesis, which has been
criticised for being ‘artifcial’ (Valentini, 2019); CWC theory rejects
the theory of decoherence, a theory many physicists fnd appealing.
Both Bohmian mechanics and CWC theory come at the cost of denying
Lorentz invariance, sacrifcing an observational symmetry of special
relativity.
Whether these are perceived as theoretical costs, and if so, just how they
will be weighed against one another, will vary among both philosophers
and physicists. Some physicists – particularly mathematical theorists – are
strongly attracted to simplicity and symmetry, although such an attrac-
tion is not without its perils, since it can lead to physical models that are
oversimplifed or are endowed with symmetries that do not appear to
correspond to anything in reality. Other physicists – perhaps the more
practically minded experimentalists – may be sympathetic to some degree
of dappledness in nature’s laws.
In any case, I do not wish to suggest that Bohmian mechanics and CWC
theory are the only quantum mechanical theories which admit hylomor-
phic models. Adrian Kent has proposed a way of solving the measure-
ment problem that does not entail a rejection of decoherence theory and
which may admit a hylomorphic interpretation (Kent, 2015).57 Alexander
R. Pruss has ofered a hylomorphic account of quantum mechanics that
adapts the so-called travelling minds interpretation (Pruss, 2017). There
are doubtless other areas for exploration.
Before closing, I should emphasise that the two metaphysical mod-
els that I have outlined in this chapter are not simple restorations of
Aristotle’s original doctrine of hylomorphism: the matter they posit is
not merely a substrate of potentiality but consists of particles or par-
cels of gunk which play a role in individuating physical objects, and
the notion of formal causation is explicated in terms of a grounding
relation. These are features which may attract the censure of classical
Aristotelians or Thomists. I am ofering these ‘neo-Aristotelian’ models,
however, as philosophical sketches which are intended to inspire further
discussion.58
Notes
1. The term ‘hylomorphism’ is a portmanteau of these Greek words.
2. Widespread, but not universal: Leibniz was a notable exception to this trend
(Leibniz, 1976).
3. Physicists have often denied the intelligibility of quantum mechanics (Dürr
et al., 2012, ch. 4).
4. Anna Marmodoro has expounded its Aristotelian roots by linking the concept
to Aristotle’s discussion of potentiality and actuality (Marmodoro, 2014, ch. 1).
5. For a recent discussion of the return of powers to mainstream philosophy, see
Lagerlund et al. (2021).
6. Famously, in Raphael’s School of Athens, Plato is portrayed pointing a fnger
upward toward the sky, whilst one of Aristotle’s hand is stretched toward the
earth with his fngers splayed.
7. For a discussion of universals in Plato and Aristotle, see Scaltsas (1994).
8. See Plato’s Sophist in (Cooper, 1997, 247d-e, 269).
9. In my references to Aristotle, I am relying upon Barnes (1984).
10. That is, according to how Aristotle has generally been understood. For an
alternative reading, in which Aristotle is fundamentally committed to powers
(or power tropes), see Marmodoro (2014).
11. That is, in virtue of their having animal forms, or souls.
12. See e.g. [Arist., Gen. & Corr., I.10] in Barnes (1984); in particular, 327b23–
b33. Also, [Arist., Metaphys., b5–15 & 1040b5–15].
13. For further discussion of integral wholes and potential parts, see Simons
(2000, ch. 9).
14. According to Aquinas, in De Principiis Naturae (following the translation in
Skrzypek, 2019): ‘Matter and form are said to be intrinsic to a thing, in that
they are parts constituting that thing’ (ch. 3); ‘matter and form are said to be
related to one another . . . they are also said to be related to the composite as
parts to a whole and as that which is simple to that which is composite’ (ch.
4). For a contemporary account of forms as metaphysical parts of substances,
see Koslicki (2008, 2018).
15. For a more detailed narrative about the physicalisation of matter and form,
see Pasnau (2011). Sections 2.3 and 2.4 also draw upon parts of a previously
published paper of mine concerning the fall and rise of hylomorphism, and
some sentences are verbatim (Simpson, 2018).
16. Following Pasnau’s discussion in Pasnau (2011).
17. [Burgersdijk, F. 1650, Collegium Physicum, II.34, pp. 1314], as translated in
Pasnau (2011).
18. For a discussion of Ockham’s view, see Pasnau (2011), ‘Matter and extension’.
19. See [Scotus, Rep. II.12.2 n. 7 (XI:322b)] in Wolter and Bychkov (2004).
20. See [Dabillon 1643, Physique, I. 3.2, p. 103] as cited in Pasnau (2011).
21. William de la Mare targeted Aquinas afrmation of unicity in Correctorium
Fratris Thomae in 1279.
22. [Suárez, Disputationes metaphysicae, 15.1.7] as quoted in Pasnau (2011).
The text is somewhat ambiguous.
23. A Scotist might seek to develop a response to my physicalisation argument in
which some instances of causation could count as being efcient and formal,
rather than having to count as one or the other.
24. From C. Adam and P. Tannery eds., Oeuvres de Descartes, rev. ed., 12 vols.
(Paris: Vrin/CNRS, 1964–76), as quotes and translated in Pasnau (2011).
25. See René Descartes to Morin, 12 September 1638, in The Philosophical Writ-
ings of Descartes, vol. 3, ed. J. Cottingham et al. (Cambridge: Cambridge
University Press).
26. Signifcantly, Newton rejected the claim that a physical explanation had to
trace everything down to the motion of the smallest corpuscles. His theory
of universal gravitation aforded a physical explanation even though it did
not describe what push or pull between corpuscles caused gravitational
phenomena.
27. See R. Boyle, The Origin of Forms and Qualities, in The Works of Robert
Boyle, ed. M. Hunter and E. Davis (London: Pickering & Chatto, 1999–2000).
28. H. Oldenburg, Correspondence, ed. and trans. A.R. Hall and M.B. Hall
(Madison: University of Wisconsin Press, 1965), III:67.
29. See Works of Boyle, V:305.
30. I have not space here to address Newton’s changing views on powers, or
Thomas Reid’s defence of dispositionalism against Hume, etc. For a more
adequate historical account, see Lagerlund et al. (2021).
31. See P.-S. Marquie de Laplace, A Philosophical Essay on Probabilities, trans.
F. W. Truscott and F. L. Emory (New York: John Wiley & Sons, 1902).
32. Section 3 draws upon the hylomorphic interpretation of Bohmian mechanics
suggested in my doctoral thesis (Simpson, 2019), which was further devel-
oped in Simpson (2021).
33. In Einstein’s original version of the EPR experiment, the measurements are of
position and momentum.
34. For a more detailed but accessible discussion, see Maudlin (2011).
35. For three recent experimental tests of Bell’s inequalities, which close two
‘loopholes’ in Bell’s theorem and underscore the failure of locality, see Gius-
tina et al. (2015); Hensen et al. (2015); Shalm et al. (2015).
36. The operator Û is a unitary operator. The meaning of ‘unitarity’ is that the proba-
bilities computed from ˜ always sum to unity. The operator Uˆ redistributes
the probabilities between diferent possibilities as time goes on. It also induces
superpositions.
37. Maudlin admits that besides GRW theory and Bohmian mechanics, ‘there
are others that can survive the test’ (Maudlin, 1995, p. 14). I discuss a third
possibility called CWC theory in Section 4, which is like GRW theory except
that the wave function collapse dynamics is context-dependent.
38. The particles are each attributed gravitational mass, mi, but the COW
experiment suggests mass delocalises over the wave function and need not be
regarded as an intrinsic property (Brown, 1996). (COW refers to the scien-
tists: Colella, Overhauser and Werner.)
39. Concerning relativistic versions of Bohmian mechanics, see Dürr et al.
(2014).
2
40. Since the theory allows non-equilibrium solutions ˆ ˜ , this restriction
to quantum equilibrium has generally been taken as a necessary postulate to
introduce the element of randomness that is essential to quantum mechanics
(Dürr et al., 1992). Some see this move as ‘artifcial’, e.g. Valentini (2019).
41. For details of this account of laws, see Ramsey (1978), Lewis (1973,
pp. 73–75), Lewis (1987, postscript), and Mill (1875, Book III Chapter IV).
42. For further discussion, see Lazarovici (2018), Matarese (2018), Simpson
(2020), and Wilson (2018).
43. This model deploys the notion of grounding championed in Schafer (2009,
2010).
44. Concerning the nature of this substantial unity, see Simpson (2021, Section 3)
and Peterson (2018).
45. Section 4 draws upon my hylomorphic interpretation of contextual wave
function collapse theory in my doctoral work (Simpson, 2019).
46. This section discusses a problem I describe at greater length in (Simpson et
al. in progress).
47. This technique maintains the tradition of using separable Hilbert spaces in
quantum mechanics.
48. In statistical physics, classical models of coupled spins (like Glauber models)
are also used which show phase transitions for fnite systems. These mod-
els have fip rates that depend on the orientation of their neighbours. With
increasing system size, however, the time required for fipping the preferred
orientation becomes larger than the lifetime of the universe. So for all practi-
cal purposes, ergodicity is broken and diferent phases coexist. I would like
to thank Barbara Drossel for pointing this out.
49. Cf. Brian Ellis’s causal process realism (Ellis, 2001, pp. 157–160).
50. Of course, this is only a simplifed example of what this feedback might look
like. The fnal equation in an elaborated theory of feedback will look more
complicated and will involve stochasticity.
51. For a discussion of accidental forms as metaphysical parts of material sub-
stances, see Skrzypek (2019).
52. Regarding the ‘tiling constraint’, see Schafer (2010). See also its application
in Koons (2019).
53. In another version of this model, the gunk might be replaced with infnitely
divisible ‘extended simples’. For the sake of simplicity and continuity, however,
I have inherited the gunk from the GRWm ontology.For further discussion of
material composition and substance ontologies, see Dumsday (2019b, chs. 4–5).
54. On this view, the linear Schrödinger equation familiar to standard quantum
mechanics is an idealisation that only exists in the limit of perfect isolation f
→ 0, which cannot be attained in practice.
55. As opposed to an accidental change; the only kind of change admitted in the
mechanical stance.
56. The Cosmic Substance, in that case, would also be a thermal substance.
57. Correspondence with Robert Verrill.
58. The author would like to thank (in alphabetical order) Thomas Davenport,
Barbara Drossel, Travis Dumsday, George Ellis, Simon Horsley, Robert C.
Koons, Anna Marmodoro, John Pemberton, Javier Sanchez-Canizares, and
Robert Verrill, among other friends and colleagues too numerous to name, for
critical feedback and discussions.
References
Allori, V., Goldstein, S., Tumulka, R., and Zanghì, N. (2008). On the common
structure of Bohmian mechanics and the Ghirardi-Rimini-Weber Theory: Dedi-
cated to GianCarlo Ghirardi on the occasion of his 70th birthday. The British
Journal for the Philosophy of Science, 59(3):353–389.
Armstrong, D. M. (1983). What is a Law of Nature? By D. M. Armstrong. Cam-
bridge University Press, Cambridge.
Aspect, A., Dalibard, J., and Roger, G. (1982). Experimental test of Bell’s inequal-
ities using time-varying analyzers. Physical Review Letters, 49:1804–1807.
Barnes, J. (1984). Complete Works of Aristotle, Volume 1. The Revised Oxford
Translation. Princeton University Press, Princeton.
Bell, J. S. (1964). On the Einstein Podolsky Rosen paradox. Physics Physique
Fizika, 1(3):195–200.
Bell, J. S. (1987). Speakable and Unspeakable in Quantum Mechanics. Cam-
Bird, A. (2007). Nature’s Metaphysics: Laws and Properties. Oxford University
Press, Oxford.
Bohm, D. (1951). Quantum Theory. Prentice-Hall, Englewood Clifs.
Bohm, D. (1952). A suggested interpretation of the Quantum Theory in terms of
“Hidden” Variables. I. Physical Review, 85(2):166–179.
Boyle, R. (2000). The Works of Robert Boyle, Vol. 10: Notion of Nature and
other publications of 1684–6. Pickering & Chatto, London.
Brown, H. R. (1996). Bovine metaphysics: Remarks on the signifcance of the
gravitational phase efect in quantum mechanics. In Perspectives on Quantum
Reality, pp. 183–193. Springer, Dordrecht.
Carroll, J. W. (1994). Laws of Nature. Cambridge University Press, Cambridge,
1st edition.
Cartwright, N. (1999). The Dappled World by Nancy Cartwright: A Study of the
Boundaries of Science. Cambridge University Press, Cambridge.
Cooper, J. M., editor (1997). Plato: Complete Works. Hackett Publishing, India-
napolis; Cambridge.
De Broglie, L. (1928). La nouvelle dynamique des quanta [The new dynamics of
quanta]. In Electrons et photons. Rapports et discussions du cinqui’eme Conseil de
physique tenu ‘a Bruxelles du 24 au 29 octobre 1927 sous les auspices de l’Institut
international de physique Solvay. Paris: Gauthier-Villars, pp. 105–132. English
translation. In Bacciagaluppi, G. and Valentini, A., editors, Quantum Theory at the
Crossroads: Reconsidering the 1927 Solvay Conference, pp. 341–371. Cambridge
University Press, Cambridge.
Demarest, H. (2017). Powerful properties, powerless laws. In Jacobs, J. D., editor,
Causal Powers, pp. 38–53. Oxford University Press, Oxford.
Dong, C.-H., Yang, L., Imoto, N., Gaddam, V., Xiao, Y.-F., and Özdemir, Ş. K.
(2008). Quantum nondemolition measurement of photon number via optical
Kerr effect in an ultra-high-Q microtoroid cavity. Optics Express,
16(26):21462–21475.
Drossel, B. (2017). Ten reasons why a thermalized system cannot be described by a
many-particle wave function. Studies in History and Philosophy of Modern Phys-
ics, Part B: Studies in History and Philosophy of Modern Physics, 58:12–21.
Drossel, B. and Ellis, G. (2018). Contextual wavefunction collapse: An integrated
theory of quantum measurement. New Journal of Physics, 20:113025.
Dumsday, T. (2016). Non-mereological pluralistic supersubstantivalism: An alter-
native perspective on the matter/spacetime relationship. Canadian Journal of
Philosophy, 46(2):183–203.
Dumsday, T. (2019a). Breathing new life into the world-soul? Revisiting an old
doctrine through the lens of current debates on special divine action. Modern
Theology, 35(2):301–322.
Dumsday, T. (2019b). Dispositionalism and the Metaphysics of Science. Cam-
Dürr, D., Goldstein, S., Norsen, T., Struyve, W., and Zanghì, N. (2014). Can
Bohmian mechanics be made relativistic? Proceedings of the Royal Society A:
Mathematical, Physical and Engineering Sciences, 470(2162):20130699.
Dürr, D., Goldstein, S., and Zanghì, N. (1992). Quantum equilibrium and
the origin of absolute uncertainty. Journal of Statistical Physics,
67:843–907.
Dürr, D., Goldstein, S., and Zanghì, N. (2012). Quantum Physics Without Quan-
tum Philosophy. Springer-Verlag, Berlin, Heidelberg.
Einstein, A., Podolsky, B., and Rosen, N. (1935). Can quantum-mechanical
description of physical reality be considered complete? Physical Review,
47(10):777–780.
Ellis, B. (2001). Scientifc Essentialism. Cambridge University Press, Cambridge.
Ellis, G. F. R., Meissner, K. A., and Nicolai, H. (2018). The physics of infnity.
Nature Physics, 14(8):770–772.
Esfeld, M. and Deckert, D.-A. (2017). A Minimalist Ontology of the Natural
World. Routledge, London.
Esfeld, M., Lazarovici, D., Lam, V., and Hubert, M. (2017). The physics and
metaphysics of primitive stuf. The British Journal for the Philosophy of Sci-
ence, 68:133–161.
Fraser, D. (2009). Quantum feld theory: Underdetermination, inconsistency, and
idealization*. Philosophy of Science, 76(4):536–567.
Ghirardi, G. C., Grassi, R., and Benatti, F. (1995). Describing the macroscopic
world: Closing the circle within the dynamical reduction program. Foundations
of Physics, 25(1):5–38.
Ghirardi, G. C., Rimini, A., and Weber, T. (1986). Unifed dynamics for micro-
scopic and macroscopic systems. Physical Review D, 34(2):470–491.
Gillett, C. (2016). Reduction and Emergence in Science and Philosophy. Cam-
Giustina, M., Versteegh, M. A. M., Wengerowsky, S., Handsteiner, J., Hochrainer,
A., Phelan, K., Steinlechner, F., Kofer, J., Larsson, J.-Å., Abellán, C., Amaya,
W., Pruneri, V., Mitchell, M. W., Beyer, J., Gerrits, T., Lita, A. E., Shalm, L. K.,
Nam, S. W., Scheidl, T., Ursin, R., Wittmann, B., and Zeilinger, A. (2015).
Signifcant-Loophole-Free Test of Bell’s Theorem with entangled photons.
Physical Review Letters, 115(25):250401.
Harré, R. and Madden, E. H. (1973). Natural powers and powerful natures.
Philosophy, 48(185):209–230.
Heil, J. (2003). From an Ontological Point of View. Oxford University Press,
Oxford.
Heil, J. (2012). The Universe As We Find It. Oxford University Press, Oxford.
Hensen, B., Bernien, H., Dréau, A. E., Reiserer, A., Kalb, N., Blok, M. S., Ruiten-
berg, J., Vermeulen, R. F. L., Schouten, R. N., Abellán, C., Amaya, W., Pruneri,
V., Mitchell, M. W., Markham, M., Twitchen, D. J., Elkouss, D., Wehner, S.,
Taminiau, T. H., and Hanson, R. (2015). Loophole-free Bell inequality viola-
tion using electron spins separated by 1.3 kilometres. Nature,
526(7575):682–686.
Kent, A. (2015). Lorentzian quantum reality: Postulates and toy models. Philo-
sophical Transactions of the Royal Society A, Mathematical, Physical and Engi-
neering Sciences, 373(2047):1–9.
Koons, R. C. (2018). Forms as simple and individual grounds of things’ natures.
Metaphysics, 1(1):1–11.
Koons, R. C. (2019). Thermal substances: A Neo-Aristotelian ontology of the
quantum world. Synthese, 4(2):1–22.
Koslicki, K. (2008). The Structure of Objects. Oxford University Press, Oxford.
Koslicki, K. (2018). Form, Matter, Substance. Oxford University Press, Oxford.
Lagerlund, H., Hill, B., and Psillos, S., editors (2021). Reconsidering Causal
Powers: Historical and Conceptual Perspectives. Oxford University Press,
Oxford.
Lazarovici, D. (2018). Super-Humeanism: A starving ontology. Studies in History
and Philosophy of Science, Part B: Studies in History and Philosophy of Mod-
ern Physics, 64:79–86.
Leibniz, G. W. (1976). Discourse on Metaphysics. In Philosophical Papers and
Letters, pages 303–330. Springer, Netherlands, Dordrecht.
Lewis, C. S. (1964). The Discarded Image: An Introduction to Medieval and
Renaissance Literature. Cambridge University Press, Cambridge.
Lewis, D. (1973). Counterfactuals. Harvard University Press, Cambridge, MA.
Lewis, D. (1986). On the Plurality of Worlds. Blackwell, Oxford.
Lewis, D. (1987). A subjectivist’s guide to objective chance. In Philosophical
Papers Volume II, pp. 83–113. Oxford University Press, Oxford.
Marmodoro, A. (2014). Aristotle on Perceiving Objects. Oxford University Press,
Oxford.
Martin, C. B. (2007). The Mind in Nature. Oxford University Press, Oxford.
Matarese, V. (2018). A challenge for Super-Humeanism: The problem of imma-
nent comparisons. Synthese, 197, 4001–4020 (2020).
Maudlin, T. (1995). Three measurement problems. Topoi, 14(1):7–15.
Maudlin, T. (2007). The Metaphysics Within Physics. Oxford University Press,
Oxford.
Maudlin, T. (2011). Quantum Non-Locality and Relativity: Metaphysical Intima-
tions of Modern Physics. Wiley-Blackwell, Oxford.
Maudlin, T. (2019). The universal and the local in quantum theory. In Philoso-
phers Look at Quantum Mechanics, pp. 45–60. Springer, Cham.
Mill, J. S. (1875). A System of Logic, Ratiocinative and Inductive. Longmans,
London.
Minkowski, H. (1908). Raum und Zeit [Space and time]. Physikalische Zeitschrift,
10:75–88.
Molnar, G. (2006). Powers: A Study in Metaphysics. Oxford University Press,
Oxford.
Mumford, S. and Anjum, R. L. (2011). Getting Causes From Powers. Oxford
University Press, Oxford.
Omnés, R. (1994). The Interpretation of Quantum Mechanics. Princeton Univer-
sity Press, Princeton.
Pasnau, R. (2011). Metaphysical Themes 1274–1671. Oxford University Press,
Oxford.
Peterson, A. S. (2018). Unity, plurality, and hylomorphic composition in Aristo-
tle’s metaphysics. Australasian Journal of Philosophy, 96(1):1–13.
Pruss, A. R. (2017). A travelling forms interpretation of quantum mechanics. In
Simpson, W. M. R., Koons, R. C., and Teh, N. J., editors, Neo-Aristotelian
Perspectives on Contemporary Science, pp. 105–122. Routledge, London.
Ramsey, F. P. (1978). Universals of Law and Fact. [Manuscript]. http://www.
dspace.cam.ac.uk/handle/1810/194721
Ruetsche, L. (2002). Interpreting quantum feld theory*. Philosophy of Science,
69(2):348–378.
Ruetsche, L. (2003). A matter of degree: Putting unitary inequivalence to work.
Philosophy of Science, 70(5):1329–1342.
Ruetsche, L. (2006). Johnny’s so long at the ferromagnet. Philosophy of Science,
73(5):473–486.
Ruetsche, L. (2011). Interpreting Quantum Theories. Oxford University Press,
Oxford.
Ruetsche, L. (2013). Unitary equivalence and physical equivalence. In The Oxford
Handbook of Philosophy of Physics. Oxford University Press, Oxford.
Ruetsche, L. (2020). Perturbing realism. In French, S. and Saatsi, J., editors, Sci-
entifc Realism and the Quantum, pp. 293–314. Oxford University Press,
Oxford.
Scaltsas, T. (1994). Substances and Universals in Aristotle’s ‘Metaphysics’. Cor-
nell University Press, Ithaca; London.
Schafer, J. (2009). On what grounds what. In Manley, D., Chalmers, D. J., and
Wasserman, R., editors, Metametaphysics New Essays on the Foundations of
Ontology, pp. 347–383. Clarendon Press, Oxford.
Schafer, J. (2010). Monism: The priority of the whole. Philosophical Review,
119(1):31–76.
Shalm, L. K., Meyer-Scott, E., Christensen, B. G., Bierhorst, P., Wayne, M. A.,
Stevens, M. J., Gerrits, T., Glancy, S., Hamel, D. R., Allman, M. S., Coakley,
K. J., Dyer, S. D., Hodge, C., Lita, A. E., Verma, V. B., Lambrocco, C., Tor-
torici, E., Migdall, A. L., Zhang, Y., Kumor, D. R., Farr, W. H., Marsili, F.,
Shaw, M. D., Stern, J. A., Abellán, C., Amaya, W., Pruneri, V., Jennewein, T.,
Mitchell, M. W., Kwiat, P. G., Bienfang, J. C., Mirin, R. P., Knill, E., and Nam,
S. W. (2015). Strong loophole-free test of local realism. Physical Review Let-
ters, 115(25):250402.
Silva, I. (2019). From extrinsic design to intrinsic teleology. European Journal of
Science and Theology, 15(3):61–78.
Simons, P. (2000). Parts: A Study in Ontology. Oxford University Press, Oxford.
Simpson, W. M. R. (2018). Knowing nature: Beyond reduction and emergence. In
Torrance, A. B. and McCall, T. H., editors, Knowing Creation. Zondervan,
Grand Rapids.
Simpson, W. M. R. (2019). What’s the Matter? Toward a Neo-Aristotelian Ontol-
ogy of Nature. PhD thesis, University of Cambridge, Peterhouse.
Simpson, W. M. R. (2020). What’s the matter with Super-Humeanism? The Brit-
ish Journal for the Philosophy of Science, 00:1–21.
Simpson, W. M. R. (2021). Cosmic hylomorphism. European Journal for Philoso-
phy of Science, 11(28):1–25.
Simpson, W. M. R. and Pemberton, J. (2021). Cosmic hylomorphism vs Bohmian
dispositionalism: Implications of the ‘no-successor problem’. Manuscript in
submission.
Skrzypek, J. W. (2019). Accidental Forms as Metaphysical Parts of Material Sub-
stances in Aquinas’s Ontology. In Pasnau, R., editor, Oxford Studies in Medi-
eval Philosophy Volume 7, Oxford University Press, Oxford.
Suárez, M. (2015). Bohmian dispositions. Synthese, 192(10):3203–3228.
Valentini, A. (2019). Foundations of statistical mechanics and the status of the
Born rule in de Broglie-Bohm pilot-wave theory. In Allori, V., editor, Statistical
Mechanics and Scientific Explanation, pp. 423–477. World Scientific,
Singapore.
Wallace, D. (2011). Taking particle physics seriously: A critique of the algebraic
approach to quantum feld theory. Studies in History and Philosophy of Mod-
ern Physics, 42(2):116–125.
Williams, P. (2017). Scientifc realism made efective. The British Journal for the
Philosophy of Science, 70(1):209–237.
Wilson, A. (2018). Super-Humeanism: Insufciently naturalistic and insufciently
explanatory. Metascience, 27(3):427–431.
Wolter, A. B. and Bychkov, O. V. (2004). John Duns Scotus: The Examined
Report of the Paris Lecture: Reportatio 1-A (2 vols.). franciscanpublications.
com.
2 Essential Thermochemical
and Biological Powers
Robert C. Koons
1. Introduction
There are three great options in the philosophy of nature: materialism,
cosmic monism, and plural holism. These correspond to the metaphysi-
cal priority of, respectively, the very small, the very large, and the inter-
mediate. Human beings and other organisms fall into the intermediate
category. I will argue in Section 2 that a philosophy of nature that gives
pride of place to thought and responsible, intentional action, while avoid-
ing Cartesian dualism and idealism, must embrace the Aristotelian option
of plural holism (Inman 2018).
Before turning to the details of contemporary quantum science, I will
sketch the basic requirements of an Aristotelian pluralism in Section 3.
Aristotelian philosophy of nature maintains a unique balance between
top-down (formal) and bottom-up (material) modes of explanation. This
balance requires the use of a repertoire of basic Aristotelian concepts,
including proximate and prime matter, substantial form, quantitative
accidents, and integral and virtual parts.
The greatest challenge to the viability of Aristotelian natural philosophy
comes from the apparent atomism of modern science since the Scientifc
Revolution. I will argue in Section 4 that the discovery of the quantum
world in the early twentieth century efected a kind of Aristotelian coun-
ter-revolution, displacing the ostensible atomism of the Newton-Maxwell
model with an irreducible holism, a holism that is most apparent in quan-
tum chemistry and thermodynamics. Consequently, I will argue in Section
5 that the world according to a neo-Aristotelian framework should consist
of living organisms and what I shall call ‘thermal substances’ (along with
remnants of these). In that section I will develop the theory of thermal
(inorganic) substances in some detail, addressing the questions of their
origin and individuation.
With the theory of thermal substances in place, I will be able to turn
to an account of the world of organisms in Section 6. We will see there
that thermal substances serve as virtual parts and as proximate matter
for living organisms. This opens up the possibility of a formal mode of
DOI: 10.4324/9781003125860-4
Thermochemical and Biological Powers 67
causation by the souls or substantial forms of living organisms, avoiding
some of the difculties of Cartesian interactionism. In Section 7, I discuss
how an Aristotelian philosophy of nature will enable us to rehabilitate
what Wilfred Sellars called the manifest image of the world (Sellars 1962),
undergirding the veridicality and reliability of human observation and
experimentation.
2. The Three Options in the Philosophy of Nature

Aristotle’s metaphysics clearly assigns the status of fundamental to living
organisms, despite their intermediate size. Organisms are neither mere
heaps of atoms nor mere fragments of the whole cosmos. They are instead
among the primary beings of the world – the things that have unity and
exist in the strictest, most central sense.
Since the Scientifc Revolution of the seventeenth century, a kind of
philosophical atomism has tended to dominate our understanding of
nature. On this view, the power and nature of any composite material
entity depends on the powers, natures, and mutual arrangements and
motions of the smallest bits of matter. This would seem to leave no
room for genuine human agency, as Plato recognized in the Phaedo
(98c–99b). The agency of the particles leaves nothing for reason or
free will to do, except in a subordinate and derived way. The atom-
istic materialist cannot fnd a place in the world for genuine rational
powers, a kind of fundamental responsiveness of the human mind to
reasons and evidence.1
In response, many theists have embraced a kind interactionist dualism
or some form of idealism or cosmic holism (e.g., Rowan Williams – see
Pickstock 2015), trying to carve out real space for the domain of reason.
However, there are severe theological and apologetic costs to these dualist
and idealist stratagems. Both dualists and idealists must posit a prob-
lematic explanatory gap between natural phenomena and our internal
sensations or “phenomenal qualia” (Levine 2000). The prospects for any
simple, law-like relationship between microphysical properties and sen-
sory qualia are extremely dim, as noted by Robert Adams (1987). Instead,
we are left with massively gerrymandered and anomalous correlations
between physical conditions and experiential qualities, correlations that
can never be illuminated by causal mechanisms. Dualists and idealists
also face difcult questions about how spiritual realities can interact with
physical processes without violating physical symmetries and conserva-
tion laws. Dualists and idealists seem to be stuck with a fruitless quest for
some elusive vital force (élan vital) by which the mind can move funda-
mental particles (see Lowe 1992). Ethically, dualists and idealists run the
risk of downplaying the importance of bodily integrity, since they make
the human body wholly extrinsic to the human person as such (see, for
example, Lee and George 2009).
68 Robert C. Koons
If we set aside dualism and idealism as problematic and implausible,
then we are forced to choose between the three remaining options: atom-
ism, monism, and pluralism. Atomists hold that only extremely small
entities, like subatomic particles or point-intensities of felds, can be meta-
physically fundamental, while monists (like Jonathan Schafer) take that
there is only one fundamental entity, the entire cosmos (Schafer 2010).
Pluralists assume that we can fnd fundamental entities at many diferent
scales of size, including intermediate-sized entities like living organisms.
From a traditional theistic point of view, there are at least three reasons
for favoring pluralism: preserving human agency, securing our knowledge
of necessary and normative truths, and buttressing the teleological argu-
ment for God’s existence.
First, both physical monism and atomism threaten human agency. If
reason is to have any power, the human being must be capable (sans
reason) to arrive at more than one conclusion (whether theoretical or
practical). And which alternative conclusion we do reach must be explain-
able in terms of our reasons and acts of will – it cannot be exhaustively
explained at either the atomic or the cosmic level without introducing
an implausible coincidence of overdetermination, that is, an ad hoc, pre-
established harmony between the material and the rational (of the sort
proposed by Gottfried Leibniz).
Second, for the same reason, monism and atomism threaten human
epistemology. Our non-empirical knowledge of necessary facts requires
that those facts have some direct impact on our faculty of intuition. The
intuitions we form must not be explainable in terms of processes at the
atomic or cosmic level, processes with no real, constitutive connection to
the relevant necessary facts. This is especially problematic for our knowl-
edge of purely normative facts, since normative facts seem to play no role
in determining the nature or movements of either material atoms or the
physical cosmos as a whole. Any morality with an intellectual compo-
nent (anything, that is, beyond the most voluntarist of divine command
or social-convention theories) requires both real human agency and real
knowledge of normative reasons (see Koons 2017, 2019b).
Third, Aristotelian pluralism provides strong grounds for inferring the
existence of a transcendent Creator, as exemplifed by Thomas Aquinas’s
Fifth Way (see Feser 2008, 110–119). To be viable, pluralism must reject
Humean doubts about causation, embracing a robust realism about causal
powers. As pluralists, Aristotelians attribute such robust causal powers to
composite material substances, including living organisms. Causal powers
correspond to teleological structures, since each causal power is intrinsi-
cally ordered to a particular result. If there are fundamental causal powers
possessed by living organisms, then those powers defne a distinctively
biological teleology.
These causal powers must themselves be explained in terms of the
natures or essences of these composite substances. These natures must
somehow be imposed upon appropriate material parts or components.
Hence, there must be compatibility between the material requirements
of the biological essences and the attributes of inorganic material things.
Otherwise, it would be impossible for living things to grow, develop,
or reproduce. For example, there must be a compatibility between inor-
ganic substances (like water, O2, CO2, or various mineral salts) and the
substance of living organisms like plants, a compatibility that enables
the latter to absorb and integrate the former. Moreover, there must be an
adequate causal explanation of why these particular biological essences
are actually imposed upon parts of the natural world. Even if, as Aristo-
tle supposed, all of the species of organisms had been exemplifed from
eternity past, we would still need an explanation for the imposition of
these specifc essences on those particular streams of material substrate.
We would still need to know the ground for the existence of any living
things at all. Given, as we now know, that life is a relatively late arrival,
the need for an adequate cause of its emergence becomes especially acute.
Darwinian natural selection, even if correct as an explanation of the
generation of new species, falls short of providing a metaphysically
sufcient explanation of this fact (see Stephen Boulter’s chapter in this
volume). Evolution can describe how the life-engendering powers of the
natural world came to be exercised and in what order they were exercised,
but it cannot explain why those powers are there in the frst place. Given
the metaphysical fundamentality of organisms, we cannot hope to explain
the origin or development of life simply by describing possible trajectories
of the constituent particles. Even if matter fell by chance into optimal
spatial arrangements and mutual motions, it wouldn’t thereby constitute
an Aristotelian substance. And given the non-eternity of species, possible
future species must exist in the mind of something like a Craftsman, in
the way that the form of a house is present in the mind of an architect.
3. Building a Pluralist Philosophy of Nature

Building a pluralist philosophy of nature is not a trivial task. Pluralists
must make room in the world for both top-down explanation (explaining
parts in terms of wholes) and bottom-up explanation (explaining wholes
in terms of parts), while atomists and monists need only one or the other
mode. For pluralists, the world consists of substances, fragments of sub-
stances, and heaps of substances. For atomists, the world consists only
of substances and heaps, while for the monist it consists only of a single
substance and its fragments.
Moreover, it is clear that many substances (like organisms) interact
with their environment through their parts. Hence, the powers of the
whole substance must in some way be dependent on the disposition of
those parts. In addition, the very survival of a substance depends on the
appropriate cooperation of its parts. At the same time, there must be
70 Robert C. Koons
something that unifes those parts (and just those parts) into a single sub-
stance. For Aristotelians, this something is known as a substantial form.
Each substance has a single substantial form that makes it what it is and
that unifes its parts, both at a time and through time.2 The substantial
form of a substance does not simply consist in the nature of its parts and
their mutual arrangement in space – it is that which ultimately grounds
and explains those natures and that arrangement.
For composite, material substances, substantial form cannot be the
whole story. There must also be that on which the substantial form
operates. This is the substance’s matter. The primary metaphysical role
of matter is that of individuation. Chunks of matter individuate a sub-
stance and its parts from substances and parts of the same natural kind.
They ground the mutual distinctness of things that are specifcally the
same (the same in kind). This individuating role is what gives the Aris-
totelian a unique and attractive account of natural sameness (see Brower
2017; Koons 2018b). If we consider matter in its pure function as a
bare individuator, we arrive at the concept of prime matter. A chunk of
prime matter has no positive nature, quality, or quantity of its own. It
simply individuates its substance or part of a substance from others of
the same kind.3
However, prime matter is never wholly on its own and so never actually
bare. It is always of necessity informed by a substantial form, and this
informing results in various layers of what is called proximate matter.
The human being, for example, is obviously composed of various kinds
of tissue, such as bone, muscle, skin, and blood. Each chunk of tissue cor-
responds to a chunk of prime matter, but a chunk that has been natured
and qualifed by the human being’s one substantial form. Each piece of
proximate matter is composed of parts of more elementary, less proximate
kinds of matter. Such intermediate forms of matter in a living organism
include proteins, lipids, and carbohydrates, which in turn are composed
of still less proximate matter, like carbon, hydrogen, and oxygen, with the
most elemental forms of matter (protons, neutrons, electrons) constituting
a layer just “above” that of prime matter. It is the one substantial form
of the human being (the human soul) that is responsible for the character
of each of these layers of matter above the level of prime matter, but
each level has an important role to play in explaining the persistence and
varying powers of the whole substance. The explanatory role that these
layers play is called material causation. The substantial form plays the
complementary role of formal causation.
Given the critical role that substantial form must play in the Aristote-
lian framework, it is impossible for one substance to contain another or
even to overlap another substance. Substances in the Aristotelian scheme
satisfy what Jonathan Schafer calls the tiling constraint (Schafer 2010).
The material universe consists entirely of non-overlapping substances
(and remnants of these). Parts of substances, therefore, cannot be actual
substances themselves. Their natures or identities (or both) are inextrica-
bly tied to the whole substance of which they are a part.
Given the tiling constraint, all parts of substances must be either inte-
gral parts or virtual parts. An integral part of a substance is a part whose
whole nature and individual identity is tied to that substance. My hand,
for example, is essentially a hand, so essentially a part of a human being.
Integral substances can sometimes exist independently of their “host”
substance, but they persist only as remnants of that substance and never as
substances in their own right. Their natures are such that it is metaphysi-
cally impossible for an integral part to exist except as a part or a remnant
of a particular substance. Their individual identities are irrevocably tied
to the organism from which they originated.
In contrast, virtual parts of a substance have intrinsic natures that are
independent of the whole. This does not violate the unicity of substan-
tial form, since the virtual parts have only potential existence within the
whole substance. Moreover, the fact that this inorganic substance can
exist as a virtual part of the organism is grounded in the organism’s own
substantial form.
For any virtual part of a substance, there are many empirically indis-
tinguishable counterparts existing as actual substances in their own right
and not as mere virtual parts. The water in my veins, for example, cor-
responds chemically and thermodynamically to batches of water existing
in the inorganic world as actual, independent substances. In this case the
tiling constraint is satisfed by stipulating that the virtual part no longer
has actual existence while part of my body. It exists only potentially, con-
tributing to the persistence and powers of my body but not constituting
at the same time a distinct substance. For this reason, none of the water
in my veins can be numerically identical to any inorganic sample of water
that I ingested, since the inorganic water was a substance in its own right
and not merely a virtual part of another substance. This transmutation of
inorganic substances into virtual parts and vice versa is an unavoidable
theoretical cost of the hylomorphic picture, but it is a cost well worth
paying.
The structured parts of a substance (e.g., organs and tissues of the
living body) are its integral parts, while the kinds of stufs contained by
the substance (water, lipids, proteins, nucleic acids) are or correspond to
virtual parts.
Aristotelian natural philosophy includes the reifcation of certain fea-
tures of substances – the accidents. These accidents are abstract particu-
lars, corresponding to the modes or tropes of modern metaphysics. Each
particular accident is simultaneously both a real entity, tied essentially to
a single subject, and a feature of that subject. So, Socrates’s musicality is
a classic example of such an accident. It is that by which Socrates is musi-
cal. The accident is ontologically dependent on Socrates, and it exactly
resembles other instances of the same kind, such as Plato’s musicality.
72 Robert C. Koons
These individual accidents can themselves be generated and corrupted,
and they enter into other causal and explanatory relations.
The very fact that substances have parts of any kind is the responsibility
of the substantial form. The substantial form is responsible both for pro-
viding its integral parts with their natures and with transmuting external
inorganic substances into its virtual parts. Substantial form also imposes
forms of quantity on batches of prime matter – so-called quantitative acci-
dents of volume, mass, and relative position. These quantitative accidents
stand in part-whole relations to one another. For example, the location
of my left elbow is contained within the location of my left arm. These
quantitative accidents are responsible for the possibility of my body’s hav-
ing quantitative or material parts, both integral and virtual. The identities
of these quantitative accidents have a dual anchor: in the frst place, to
the particular substances to which they belong, in the second place, to
the particular packet of prime matter that they inform. It is because the
quantitative accidents are tied inextricably to the individual identity of
the substance to which they belong that all integral parts of the substance
are similarly so tied.
4. The Case for Thermal Substances

Integrating this Aristotelian natural philosophy with modern science faces
the problem of the inorganic world. Aristotelians will always count living
organisms among the world’s substances. Given the tiling constraint, the
sum total of substances and their remnants must together constitute the
whole of the material world. It is clear that organisms alone cannot ft
this bill. Much of the world is not and never has been part of any living
organism. Where can we fnd the fundamental Aristotelian substances in
that inorganic world, needed to complete the plural holist picture? I will
argue that we can take our cues from the holistic character of quantum
chemistry and thermodynamics. I have developed and defended a theory
of thermal substances in some recent work (Koons 2018a, 2019a, 2021),
building on the pioneering work on quantum chemistry by the theoreti-
cal chemist Hans Primas (1980, 1981), and William M. R. Simpson very
ably lays out the case for the existence of such substances in the preceding
chapter and in his doctoral thesis (Simpson 2019), building on the recent
proposal of contextual wave function collapse by the cosmologist George
Ellis and the physicist Barbara Drossel (Drossel 2015; Ellis and Drossel
2018).
On this picture, none of the entities described in so-called fundamental
physics are in fact fundamental; that is, none of them are Aristotelian
substances. Instead, quarks, electrons, photons, and the rest have merely
virtual presence in true material substances.
Quantum theory provides a fertile feld for the theory of Aristotelian
accidents. However, we cannot identify particles with accidents in a
one-to-one fashion. Rather, it is pluralities of indistinguishable particles
(which I will call congeries) that are to be identifed with Aristotelian
accidents. In standard interpretations of quantum theory (in fact, all the
interpretations except the Bohmian one),4 particles lack determinate and
enduring identities. Both Dirac-Fermi and Einstein-Bose statistics treat
elementary particles as mere fgures in an accounting book, like dollars
in the bank, rather than as enduring nodes of possibility, reidentifable as
particular individuals at diferent times and in varying possible scenarios
(see the chapters in Part I in Castellani 1998, and Redhead and Teller
1991). In nearly all situations, particles lack location and many intrinsic
properties (like momentum or spin). In fact, the location of a particle
(when not actually measured) cannot be considered to be determinately
restricted to any fnite region of space (Clifton and Halvorson 2001).
And the actual number of particles in a system is not an absolute fact but
depends on one’s frame of reference (Fraser 2008).
Each congeries of indistinguishable particles is, nonetheless, a real
entity, but an entity that is ontologically dependent on the thermal sub-
stance that contains it. Such congeries of particles correspond with the
thermal substance’s power of afecting and being afected by other thermal
substances at the quantum level. An individual particle is fully actualized
when such an interaction occurs. At that moment, the individual particle
exists as a distinct individual – an individual accident in the category of
action or passion.
A congeries of indistinguishable particles is a virtual quantitative part
of a thermal substance, corresponding to a quantitative accident. In tra-
ditional Aristotelianism, a solid bronze sphere has a top half that cor-
responds with a certain hemispheric accident of quantity. So, a congeries
of indistinguishable particles, as a quantitative accident of the thermal
substance, corresponds to a virtual quantitative part of that thermal sub-
stance. Such congeries are analogous to the quantities of elements (earth,
water, air, fre) that existed virtually in Aristotle’s theory of mixtures (in
On Generation and Corruption, Book 1). A congeries of particles is onto-
logically dependent on the thermal substance and not an actual substance
in its own right. It doesn’t exist in full actuality but only in virtue of the
power of the substance to act and be acted on in certain ways. In its virtual
presence in the substance, a congeries of quantum particles counts as a
virtual part of the proximate matter of that substance – again, analogous
to the way that elemental quantities counted in Aristotle’s chemistry.
The substantial form of the whole thermal substance is responsible
for the existence, nature, and mutual combination of the quantal enti-
ties that make up its proximate matter. At the same time, the result-
ing proximate quantum-level matter must satisfy certain conditions
for the whole substance to persist and to have the contingent powers
that it does have. Hence, there is a legitimate place for bottom-up
material explanation in the Aristotelian scheme. We can partly explain
74 Robert C. Koons
the persistence of a substance and its actions upon and reactions to its
environment in terms of the powers and potentialities corresponding
to its quantum constituents. There are two complementary modes of
explanation: material and formal, and both are prior to the usual expla-
nation in terms of efcient causation, which is invoked in explaining
the production of changes.
This theory of thermal substances and their quantum accidents agrees
with Niels Bohr’s Copenhagen interpretation of quantum mechanics on
two points: the incompleteness of the quantum-level description of the
world and the existence of defnite states within the complementary “clas-
sical” world. I difer with Bohr, however, in refusing to separate reality
into two disjoint domains (quantal and classical). Instead, the world con-
sists of thermal substances and their virtual quantal parts. Each thermal
substance has both classical and quantal properties. The classical proper-
ties form a non-trivial core or center of mutually commutable observables,
defning super-selection sectors (Primas 1980, 1981). These classical prop-
erties never enter into quantum superpositions. In contrast, the disjoint
class of quantal properties are typically found in superposed states, with
non-trivial probabilities assigned to the constituents of orthogonal bases
of observables. Two quantal properties do not typically commute, since
they belong to the same super-selection sector.
The theory of thermal substances denies the completeness of the micro-
physical. One important aspect of the scientifc revolution of the seven-
teenth and eighteenth centuries was its emphasis on what Aristotelians
would label material causation and explanation through hypothetical
necessity. If large systems are to behave as they are observed to do, they
must be composed of parts with intrinsic natures and mutual arrange-
ments in space that are capable of sustaining the observed collective
behavior. From an Aristotelian point of view, this analytic approach is
perfectly legitimate. It is illuminating to learn that water is composed
(virtually) of H2O molecules, and that cells contain (virtually) double
helices of DNA.
Where microphysicalism went wrong was in insisting that macroscopic
phenomena can be exhaustively explained in terms of microphysical facts.
Microphysicalists assume that all facts supervene on the microphysical
entities and their arrangements in space. This is true both of atomists
of various kinds and those who believe that matter is infnitely divisible,
like Empedocles or Descartes. The motions of large bodies depend on
their composition and the motions of those components, and not vice
versa. On this view, all true explanation on this view is bottom-up, from
the very small to the large, and never top-down. This means that for the
microphysicalist, there can be no room for substantial form – with one
possible exception. If there are fundamental, indivisible particles, they and
they alone could have substantial forms. Anti-atomists like Descartes or
Empedocles must reject substantial forms altogether, since on their view
there are no true unities in nature, merely various uniform and infnitely
divisible continua.
In any case, on this view, neither people nor organisms more generally
nor any of the many things that we can perceive have substantial forms
at all. Consequently, microphysicalists must deny the reality of all the
so-called secondary qualities, such as color, smell, or taste, since these, if
they existed, would be accidents of macroscopic substances.
Long before the quantum revolution, this anti-realism about the “mani-
fest image” of the world (to use Wilfred Sellars’s phrase) threatened to
undermine the scientifc enterprise itself, since all of our observations and
experimental results presuppose the real existence of, well, observations
and experiments, neither of which can easily be accommodated within a
microphysicalist image of the world. Experiments require experimental
conditions or setups, and these are defnable only in macroscopic terms.
If the macroscopic world is merely a world of misleading appearances,
how are experiments possible?
So, there were always grounds for suspicion about the microphysical-
ist picture, but the success for so long of what is now called “classical
physics,” the physics of Newton and Maxwell, and even of the theories
of relativity, suggested that such a picture must be true, whatever its
philosophical and epistemological conundrums, and that we just need
sufcient imagination to see how. This presumption of ultimate coher-
ence changes dramatically with the quantum revolution. The revolution
has not perhaps made microphysicalism completely untenable, but it has
clearly put it on the defensive and opened up the live possibility of resur-
recting substantial forms at macroscopic scales, including the scale of
human beings and other organisms.
Any quantum theory of chemistry and thermodynamics must come to
grips with the Stone-von Neumann theorem, proved by Marshall Stone
and John von Neumann in 1931 (von Neumann 1931). The theorem
states that any two “irreducible representations” of the canonical commu-
tation relations of quantum theory are unitarily equivalent, whenever the
system contains only fnitely many degrees of freedom. Any two n-param-
eter systems are, therefore, unitarily equivalent. Unitary equivalence is
taken, in quantum theory, to represent macroscopic indistinguishability.
When using fnite quantum representations, there could be no diference
at the macroscopic scale in chemical composition, or phase state of matter
(solid, liquid, gas), or entropy. This would seem to make chemistry and
thermodynamics impossible.
Practicing quantum theorists avoid this negative result by taking their
models to the thermodynamic limit (Liu 1999; Bangu 2009; Sewell 2002).
Such models represent the macrophysical state of a substance as con-
sisting of an infnite number of infnitesimal particles. In this way, they
can escape the constraints of the Stone-von-Neumann theorem, which
applies only to fnite systems. It is only at the thermodynamic limit, also
76 Robert C. Koons
known as the ‘continuum limit’ (Compagner 1989), that the models have
enough internal structure to distinguish diferent chemical compounds
and diferent phases of matter. And only at that limit can we give rigor-
ous defnitions of central notions in the physics of complex systems like
temperature and entropy.
There are three reasons for taking models to unrealistic limits. The frst
two are purely fctional: namely, using an unrealistic model because it
makes calculations simpler, or because doing so enables us to ignore rela-
tively insignifcant factors (like friction or gravity). The continuum limit
of quantum thermodynamics exemplifes a third reason: because only the
unrealistic model contains the sort of mathematical structure needed to
represent the phenomena faithfully.
When we use an unrealistic model for the third reason, the model cap-
tures some truth that a more “realistic” model of the same kind could
not in principle capture. This is exactly what we should expect in the
presence of substantial form. The substantial form imposes a structure
on the whole system that cannot in principle be explained by the system’s
parts and their arrangement, given a system with only a fnite number
of fundamental particles. The need for an infnite model refects a gap
in the bottom-up explanation of the phenomenon in question. Only the
presence of a top-down explanans (that is, the form) can account for the
inadequacy of the fnite models.
The use of infnite models in quantum statistical mechanics exactly
realizes a hylomorphic model of explanation, with its inclusion of two
complementary modes of explanation, namely, formal and material. The
fact that the models are derived by taking fnite models of the quantum
components of the system to the infnite limit represents the role of mate-
rial causation. In modeling the behavior of a body of water, we begin with
a fnite model of water molecules (with the appropriate numbers of pro-
tons, neutrons, and electrons). The fact that we must take these models to
an “unrealistic” limit (with an infnite number of infnitesimal molecules)
represents the role of formal causation. Material causation cannot fully
account for the system’s thermodynamic properties.
Substantial form is also responsible for the chemical composition of
substances. The geometrical structure of molecules cannot be captured
by quantum mechanics in the absence of the method of taking models to
the thermodynamic/continuum limit. It is true that, once we have the geo-
metrical structure (as revealed by experiment), we can use fnite quantum
models to explain the stability of that structure. This is just what hylomor-
phists would expect. Finite quantum models provide the material mode
of explanation – they explain how stable chemical forms are possible.
However, quantum theory in that simple form cannot give us the formal
explanation of the molecular structure. Finite quantum models, unlike
the corresponding models of classical mechanics, cannot explain spon-
taneous symmetry breaking. Such breaking of symmetries occurs only
in the “unrealistic” models at the thermodynamic limit (Woolley 1988;
Strocchi 1985, 117–118; Ruetsche 2006; Earman 2004). Therefore, all
asymmetric chemical structure, like the left-handedness of organic amino
acids, must be imposed from above, by substantial form (Woolley 1988;
Hendry 2010).
At the level of microphysics, there is no true irreversibility or direction
to time. Any purely quantum transition can be reversed in time, so long
as we reserve electric charge and parity (left- and right-handedness) at the
same time. There is, therefore, no microphysical process that is intrinsi-
cally directed toward the future. If, however, we accept thermodynam-
ics as a fundamental science applying to thermal substances (Prigogine
1997), then the Second Law of Thermodynamics (the monotonic increase
in entropy) gives us an objective, fundamental, and intrinsic direction in
time associated with every actual motion. And this gives us natural teleol-
ogy, as systems naturally seek their equilibrium state. The sum total of
quantum-level facts does not fx the direction of time, but the substantial
forms of thermal substances do.
Thus, quantum thermodynamics and chemistry provide us with
examples in which the macroscopic features of the system do not
supervene on and therefore cannot be exhaustively explained by the
microphysical facts. That is, we could have two situations that are
microphysically indistinguishable and yet chemically and thermody-
namically diferent. Consequently, these chemical diferences cannot
in principle be explained exhaustively at the microphysical level. An
adequate science of matter needs to combine bottom-up (material)
and top-down (formal) modes of explanation. The result is not a radi-
cal form of theoretical pluralism or the disunity of science, since, as
quantum thermodynamics demonstrates, we can combine both modes
in a single model.
In fact, even apart from these considerations about quantum chem-
istry and thermodynamics, pure quantum theory itself (in the form of
Schrödinger or matrix dynamics) indicates the incompleteness of the
quantum domain, as recognized by Niels Bohr. The predictions of quan-
tum dynamics all take the form of probabilities, but probabilities of what?
The standard answer (following Bohr) is: the probability of measure-
ment results. But what are measurements, and how do they have defnite
results? This leads to what is known as the measurement problem or
measurement paradox.
A quantum measurement consists in an interaction between a human
experimenter, various experimental materials (instruments, laboratory
setups, and the like), and a source of quantum particles. But macroscopic
entities (like experimenters and their instruments) are themselves entirely
composed of quantum particles, and so quantum dynamics should apply
to them as well. This leads to an infnite regress: probabilities of prob-
abilities of probabilities, ad infnitum.
78 Robert C. Koons
The famous thought experiment of Schrödinger’s cat illustrates the frst
step of the paradox. If the cat is poised to observe some quantum mea-
surement, and if we treat the cat itself as a quantum system, then the inter-
action between the cat and the quantum phenomenon (say, an electron
that could go either up or down) will yield no defnite result. Instead, the
electron will begin in a superposed Up/Down state, and the cat will come
to be in a superposed Observe-up/Observe-down state, until we open the
box and observe what the cat has actually observed. But the observer of
the cat could be treated as yet another quantum system, resulting in an
infnite regress.
Aristotelian pluralists deny that macroscopic entities like human experi-
menters and their instruments can be represented adequately by fnite
quantum models. Thermal substances have classical, mutually commuting
properties, like chemical composition, temperature, and phase of matter,
properties that never enter into quantum superpositions. When a quan-
tum power interacts with a thermal substance and produces a change in
classical properties, a “measurement” has occurred with a defnite result.
I can illustrate the hylomorphic solution to the measurement paradox
by introducing the thought experiment of Schrödinger’s ice cube. We put
an ice cube in a box, and attach it to a system that responds to some
quantum-level event, an event in a 50/50 superposed state. If the system
results in an Up event, the ice cube melted, and if it results in Down
event, the ice cube remains frozen. Now the ice cube is entirely com-
posed of protons, neutrons, and electrons, and so it is subject to quantum
modeling. However, the ice cube is a thermal substance, and so it has a
substantial form that imposes a phase of matter (namely, solidity) upon
those particles. The distinction between two phases of matter occurs only
at the level of form – it is not determined by the quantum state of the
constituent particles. Consequently, it is impossible for a cube to be in
a Frozen/Unfrozen superposed state. Such a state simply does not exist.
So, we can defne a measurement event to be an event involving non-
quantal properties (accidents) of substances. Whenever a quantum system
produces such an event, a “measurement” occurs, regardless of whether
the substance is an organism or merely a thermal substance (like an ice
cube). Consciousness need not be involved, and so we escape idealism.
This account, I suggest, is consistent with the contextual wave function
collapse model put forward by the physicists Barbara Drossel and George
Ellis (Drossel 2015; Ellis and Drossel 2018), and discussed by William M.
R. Simpson in Chapter 1.
For the hylomorphist, thermal substances and organisms have defnite
positions in space at each moment in time, even if none of their quan-
tum components do. Each quantum particle is, except for the moment in
which its position is measured, located vaguely everywhere, with a certain,
fnite probability (Clifton and Halvorson 2001). Thermal substances and
organisms, in contrast, have a defnite, actual location at each moment.
Individual quantum particles are really just momentary accidents of sub-
stances, and so the locations of the particles do not fx the location of
the substance. When not actualized by measurement, individual quantum
particles are merely powers of interaction, typically non-localized powers.
Congeries of such particles (in which the particles lack individual identi-
ties) are virtual parts of the substance.
Bohr was right in thinking that quantum mechanics indicates the
incompleteness of the quantum world. But he was wrong (at least, as he
is commonly interpreted) in thinking that the two domains could be kept
separate through a dualism of objects or entities. I speculate that this mis-
take results from unfamiliarity with the hylomorphic solution, in which
congeries of quantum particles act as the virtual proximate matter for
non-quantal forms. We now know that even macroscopic objects can have
quantal aspects – for example, superconductors and supercooled fuids,
which exemplify exotic behavior thanks to quantum coherence efects.
Hylomorphists are not committed to a dichotomous quantum/classical
dualism but to a system in which complementary entities (namely, sub-
stances and accidents, form and matter) coexist in mutual dependency.
Prompted by the urging of John Bell (1987), defenders of microphysi-
calism have sought an alternative strategy for resolving the measurement
paradox. These eforts have led to revived interest in Everett’s many-
worlds interpretation, David Bohm’s pilot-wave interpretation, and a
family of objective bottom-up collapse theories, including the Ghirardi-
Rimini-Weber (GRW) theory.5 These various microphysicalist reactions
to the paradox explain why we have a multiplicity of “interpretations”
of quantum theory. There was no such plurality of ontological interpreta-
tions of Newton and Maxwell – those theories seemed to point clearly to
the truth of microphysicalism. Quantum theory no longer does so. Saving
microphysicalism requires ad hoc supplementation.
Now, no one can be forced by quantum mechanics to embrace hylo-
morphism. Nonetheless, the hylomorphic rejection of microphysicalism
preserves the simplest and most natural interpretation of the quantum
formalism. It is well supported by the use of the thermodynamic limit in
chemistry and thermodynamics, as required by the Stone-von-Neumann
theorem. Unlike the other interpretations, hylomorphism does not require
any ad hoc modifcations or unverifable additions, and it accords best
with the actual practice of scientists. Practicing quantum scientists, like
everyone else, are implicit Aristotelians, as the philosopher of science
Nancy Cartwright (1994, 1999) has argued since the 1990s.
5. The World of Thermal Substances

A quantitative part of a substance is a part in a very ordinary familiar
way, as a fnger or a particular pint of blood are parts of an individual
organism. The substantial form and the so-called prime matter of the
80 Robert C. Koons
substance, as well as the accidents of a substance, in contrast, are not
quantitative parts. Individual quantum particles are, somewhat surpris-
ingly, also not quantitative parts of thermal substances. Instead, they are
merely potential accidents of quantitative parts. Unlike quantitative parts,
individual quantum particles have no persisting particular identities. A
congeries of indistinguishable particles belonging to a single thermal sub-
stance constitutes a virtual quantitative part of that substance and so part
of that substance’s proximate matter. Such virtual parts may be said to
have a real but derived identity as persisting, re-identifable parts of the
substance.
The tiling constraint dictates that no substance can have other actual
substances as quantitative parts. To do so would fatally compromise the
per se unity of the containing substance. The requirements of per se unity
of the composite substance are so great that we have to think of the
quantitative parts as also metaphysically dependent on the whole, and
not vice versa. As a result, Aristotelians are committed to what has been
called the Homonymy Principle: no quantitative part of a substance can
exist except as a part of that substance.
Can either accidents or quantitative parts of substances persist
beyond the demise of their substances? Accidents and quantitative
parts are dependent in some way on their host substances, but does
this dependence rule out the possibility of their extended survival?
Thomas Aquinas wrote (In Metaphysica, VII, L8, 1459) that substance
is the ‘active principle’ of accidents. It is impossible for accidents to
be prior to substances ‘in defnition (ratio), time, or generation’ (op.
cit. L13, 1579). Accidents do not have ‘perfect being (esse perfectum)’
unless they exist in a subject (L9, 1477). Priority in time means that
no accident or part can exist before its substance. Priority in defnition
means that the identity of the accident or part is derived from that
of its subject. This non-priority of accidents in both time and defni-
tion seems compatible, however, with some accidents continuing to
exist after their substances have been destroyed. In one case, as is well
known, Aquinas explicitly afrmed the possibility of the persistence
of accidents in the absence of their subjects: the accidents of the Host
in the Eucharist (Summa Theologica, Part III, q77 a1). If my theory
of thermal substances is correct, this kind of survival of accidents is
actually quite common even within the natural order and does not
demand any ad hoc metaphysics.
I have argued elsewhere (Koons 2020) that there are good grounds
within Aristotelian philosophy for positing that accidents in the category
of action can readily survive separation from and the demise of their
substantial host. For instance, Aristotle describes the archer’s accident of
action surviving in the arrow throughout its fight, and presumably this
would be true even if the archer died while the arrow was in fight. This
is extremely relevant, since quantum particles are also accidents in the
category of action. A quantum particle represents a way in which one
thermal substance can act upon another.
As we’ve seen, both accidents and quantitative parts have a dual depen-
dence on their substances. Every accident or part must receive its existence
at some point in time from a substance. Second, each accident or part is
individuated by its substance. No two congeries of particles of the same
kind are fundamentally or primitively distinct from each other. Instead, if
they have real distinctness, they derive that distinctness from the numeri-
cal distinctness of their substances (Brower 2017; Koons 2018b). This
means that the individual identity of each particle or collectivity of par-
ticles is essentially tied to that of a particular substance (thermal substance
or organism). Consequently, no particle or collectivity of particles can
be transferred from one substance to another. Once a quantum particle
has interacted with a thermal substance, its individual identity has come
to an end. However, as I’ve argued, that does not rule out the possibility
of an individual particle’s continuing to exist beyond the demise of its
substance.
The theory of thermal substances satisfes a version of the tiling
constraint. The material world consists entirely of disjoint actual sub-
stances and remnants of such substances, along with their accidents.
No two actual substances overlap, and so the plurality of actual sub-
stances (including remnants of extinct substances) constitutes a set of
mutually disjoint entities that collectively exhaust the whole material
universe.
Every actual entity in the world is either (i) a substance, (ii) an accident
of a substance, (iii) a fragment or remnant of a substance, or (iv) a group
or (v) a heap of substances and fragments.6 The fve categories can be
distinguished based on their degree of unity. A substance has absolute or
per se unity. An accident, fragment (a quantitative part), or remnant of a
substance has a similar degree of unity, but one that is dependent on the
unity of the whole substance. A group of substances has what Thomas
Aquinas called the unity of order, an imperfect form of unity. A heap has
the lowest degree of unity – the unity of contact or contiguity.
A group has greater unity than a mere heap, but less than that of a true
substance. When substances form a group, the members gain novel causal
powers, both active and passive, through membership. Like substances,
facts about groups are not reducible to intrinsic facts about their members
and the members’ spatiotemporal relations. In addition, like a substance,
a group can persist despite a changing roll of members. There are two
key diferences between groups and substances. First, a group does not so
absorb its members in such a way that the members cannot belong simul-
taneously to several groups, while no material entity can belong at the
same time to two distinct substances. Some but not all of the irreducible
powers of the members are grounded by the nature of the group. Second,
the team cannot undergo any intrinsic change except via change of some
82 Robert C. Koons
of its members, while substances can undergo change as a whole, with the
implications of the change percolating downward to the parts.
It is clear that organisms can form groups. Human beings form a bewil-
dering variety of social groups and institutions. Many other animals also
work together in analogous social structures. I would conjecture that
inorganic thermal substances can also cooperate in group-like structures.
Stable patterns seen in phenomena like Rayleigh-Bénard convection cells
(Drazin and Reid 2004), whirlpools (Steward 2012, 241), the water cycle
(Oderberg 2006), hurricanes, stars, and certain kinds of planets may indi-
cate such irreducible collective behavior, which nonetheless falls short of
the per se unity of substances (see Anderson 1972; Laughlin 2005).
Since developing my account of thermal substances several years ago,
I am often presented with two sorts of questions. These are questions of
origin and questions of individuation.
Let’s take the question of origin. How did the frst thermal substances
come into being in the early universe? What process combined the quarks,
electrons, and photons into stable, thermal substances? And, if the frst
thermal substances were formed by some congealing of fundamental par-
ticles, doesn’t that entail that there was either a time in which there were
no substances at all or a time in which particles were substances?
This question presupposes the very bottom-up imperialism that I am
taking pains to deny. The frst thermal substances did not form by some
conglomeration of fundamental particles. Rather, thermal substances
were there at the very beginning. If the universe is fnite in extent (a
so-called closed universe), then I would expect that a single thermal sub-
stance emerged from the Big Bang, a thermal substance at a cosmic scale.
As time passed, the original thermal substance decomposed into ever
smaller substances, frst at the scale of galaxy clusters, then at that of
galaxies, then solar systems, and fnally planetary components. Particles
have always had only virtual presence within substances.
I think that the same top-down perspective would restructure our
inquiry into the origin of life. Instead of thinking of the original organ-
isms as resulting from the clumping together of molecules in a pond, we
should look frst at the whole pond, or indeed at the whole solar system.
The original organisms must have resulted from the partial disintegra-
tion of some larger pre-organic system, a system that encompassed the
precursors of both the population of living things and their environment.
Similarly, we shouldn’t think of multicelled organisms as resulting from
the clumping together of one-celled organisms, but rather as the breaking
of of multicelled organisms from a pre-existing community of simpler,
less diferentiated life.
Let’s turn next to the questions of individuation.7 How many thermal
substances are there? How are they individuated? For instance, is the earth
a single substance? The lithosphere of the earth? The mantle, core, or
crust? Tectonic plates? Mountain ranges? Rocks or pebbles? Homogenous
crystals? Or are all of these mere groups or heaps of substances? How
many substances do the world’s oceans contain? Or the earth’s atmo-
sphere? How many substances occupy interplanetary or interstellar space?
In my view, these are open, empirical questions. We cannot settle them
from the armchair or by careful phenomenological examination of ordi-
nary experience. We need to develop full theories of collective phenom-
ena. The study of such phenomena (which physicists term “emergence”)
is still in its infancy. It is only in the last 40 years or so that sustained
investigation into these matters has been undertaken. Hylomorphism can
be helpful, by ruling out facile, microphysicalist answers, answers that
suggest that there is nothing fundamental or deep to study here, since
everything is supposedly reducible “in principle” to microphysics. Ernest
Rutherford is reported to have said that all of science is either ‘atomic
physics or stamp collecting’ Such microphysical imperialism relegates all
the “special” sciences to second-class status, simply arranging on the page
facts that are fully explained only by so-called fundamental physics.
If hylomorphism true, each of the special sciences is equally funda-
mental. The world cannot be captured in microphysical terms alone. The
natures and accidents of thermal substances and organisms do not even
supervene on the character and arrangement of micro-particles. (This is
consistent with admitting that no macroscopic change can occur without
concurrent microscopic change.) As we descend to the quantum scales,
things become less defnite and more dependent, and not the reverse. It is
actually fnite, “realistic” quantum physics that is non-fundamental, since
there are no “quantum substances” per se, but only quantal aspects (acci-
dents) of thermal substances and organisms. The role of particle physics is
simply to provide the base models from which the truly realistic, infnite
algebras of quantum chemistry and thermodynamics can be built, char-
acterizing the essences of thermal substances (both actual and virtual).
How big can thermal substances be? Thermal substances in principle
can exist at any scale, from single particles (or even fraction of a particle)
to the entire cosmos. I conjecture that very small substances are quite
short-lived – substances in the late stages of corruption or the early stages
of generation. Very small substances can perhaps be sustained in labora-
tory conditions. In the wild, they will, I think, generally be much larger.
In the absence of empirical inquiry, I can’t answer the questions about
the individuation of thermal substances with any confdence, but I can
suggest some criteria for individuation:
1. Sharp boundaries or discontinuities, in both space and time (i.e.,

sharp transitions) are a necessary condition for distinguishing two
thermal substances. But such boundaries may not be a sufcient
condition. Some thermal substances (like perhaps convection cells)
might include some internal boundaries. Nonetheless, where there
are no sharp boundaries, where there is perfect continuity in
84 Robert C. Koons
temperature, chemical composition, and density, we should count
the continuum as contained by a single, enduring substance.8
2. Strongly collective powers are a necessary condition for substantial
unity, both at a time and through time. A substance must have
causal powers, both active and passive, that do not supervene upon
and so are not determined by the powers and arrangements of its
parts. As we have seen, this condition is met by all bodies with
chemical and thermodynamic properties. Again, such strongly col-
lective powers may not be a sufcient condition for substantial unity
at a point in time: it could be that a group of substances possesses
some strongly collective powers, over and above the powers of the
individuals. This happens in the case of human societies, for
example.
3. The complete absorption of the causal powers of parts is a sufcient
condition for substantial unity at a point in time. Partial absorption
of powers results in groups, not in the per se unity of a substance.
In contrast, the parts of a substance are so completely absorbed
into the nature of the whole that they cannot simultaneously belong
in the same way to two distinct wholes.
6. Irreducible Powers of Organisms

How do organisms ft into the world of thermal substances? Thermal sub-
stances have a virtual presence within an organism. These virtually present
thermal substances correspond to actual quantitative parts of the organ-
ism, contributing to the explanation of the causal powers of those parts.
Consequently, there is a metaphysically fundamental diference between
inorganic water (for example) and water as it exists in organisms, although
clearly inorganic water can play a role of material causation or explana-
tion in relation to organic processes, even if organic and inorganic water
are empirically indistinguishable in their chemical actions. The chemical
properties of the water have diferent metaphysical explanations, depending
on the kind of substance (thermal or organic) to which it belongs.
The chemical and thermodynamic properties and the associated causal
powers that the quantitative parts of the living body possess are partly
determined by the soul, the substantial form of the body, along with the
holistic accidents of the organism, like perception or thought. The inter-
action is not one of action/passion, as between two substances, but top-
down formal determination. Nonetheless, changeable mental attributes
can make a real diference to the operation of bodily parts, and vice versa.
The soul can guide the breaking of microscopic symmetries, imposing
asymmetric accidents on the results, without requiring any novel force
(such as Bergson’s élan vital) or violation of conservation laws.
Is there immanent teleology in nature outside of human thought and
intention? Thomists and other Aristotelians argue that the answer is ‘Yes’.
In fact, whenever a thing acts according to its intrinsic power and potenti-
alities, immanent teleology exists (Rota 2011). Fundamental causal pow-
ers as Aristotle conceives of them are inherently teleological. To have the
power to produce E in circumstances C is to have the C-to-E transition
as one of one’s natural functions. Indeed, as George Molnar (2003) has
pointed out, the ontology of causal powers builds intentionality into the
very foundations of natural things. To have a power is to be in a kind
of intentional state, one that is in a real sense “about” the efects one is
predisposed to produce.
On the Aristotelian model, biological teleology requires just two things:
a robust causal powers metaphysics and real causal powers at the level of
biological organs and organisms. Such real powers require, in turn, sub-
stantial forms at the level of whole organisms. The substantial form of an
organism is called its soul (psyche). In nonhuman animals and in plants,
the soul is nonrational. Human souls possess additional rational powers,
powers of scientifc understanding and deliberation about the good.
This emergence of new powers at the macroscopic, biological scale
should be unsurprising, given the fact that, according to our most recent
quantum mechanical models, we see strong (ontological) emergence at
the mesoscopic scale in solid-state physics and chemistry. Mesoscopic
systems, like ferromagnets, superconductors, and Bose-Einstein conden-
sates, all exhibit dynamical behaviors that are irreducible to the micro-
states of the constituent particles, namely spontaneous symmetry breaking
and thermodynamic irreversibility. In a similar way, we should expect
the biological functioning of organisms to be irreducible to the chemical
and thermodynamic facts about the virtual thermal substances that cor-
respond to their actual quantitative parts.
Evolution itself presupposes teleology in the very idea of reproduction,
and so evolution requires irreducible causal powers at the organismic
level. No organism ever produces an exact physical duplicate of itself.
In the case of sexual reproduction, the children are often not even close
physical approximations to either parent at any stage. An organism suc-
cessfully reproduces itself when it successfully produces another instance
of its own biological kind. This presupposes a form of teleological realism,
since biological kinds are individuated teleologically, that is, in terms of
their fundamental causal powers.
Richard Dawkins has suggested that we think of organisms as mere
“robots” that our DNA molecules have “programmed” for reproducing
themselves (Dawkins 1976, xxi). In fact, DNA molecules never succeed
in producing perfect physical duplicates of themselves, and even if they
did, the mere physical duplication of the molecule would not constitute
reproducing oneself. Suppose, for example, that an eccentric billionaire
builds a chemical factory that does nothing but fll barrels with copies of
his own genome, launching them into deep space. No one would think
that such a man had succeeded in procreating trillions of descendants. A
86 Robert C. Koons
DNA molecule counts as a copy of one of one’s genes only when it is suc-
cessfully fulflling the function of a gene within a living organism, indeed,
within a living organism of the appropriate teleologically defned kind.
Alexander Pruss and I have argued that functionalist theories of mind
require an account of normativity (Koons and Pruss 2017). The argument
can be extended to functionalist accounts of biology. The functional dis-
positions that are supposed to be defnitive of mental or biological states
can only be defned relative to the normal state of the organism, where the
“normality” involved is a normative notion, not merely a matter of aver-
ages or actual frequencies. There are two prima facie plausible accounts
of the natural basis of normativity: Aristotelian powers and evolutionary
accounts.
An Aristotelian can give a straightforward account of such normativity:
A substance is supposed to produce E on occasions of C if and only

if its nature includes a C-to-E power.
The other potential source of normativity is evolutionary selection.

For example, Ruth Millikan attempted defne normality in terms of
adaptations:
If a system x belongs to a reproductive family F, then x is supposed

to produce E under circumstances C if and only if doing so is one of
F‘s adaptations.
This seems to be the most promising alternative to the Aristotelian

account. However, such evolutionary accounts are highly vulnerable to
hypothetical counterexamples. Pruss proposed (in Koons and Pruss 2017)
the thought experiment of the Great Grazing Ground: a hypothetical
world in which organisms in our history were maximally profcient in
reproduction, thanks to the intervention of benevolent aliens. In such a
world, even if the causal path leading to each of us were identical to the
actual historical path, none of us would be conscious, since no distinc-
tion between normal and abnormal states could exist. Without that dis-
tinction, the relevant functional states supposedly defning consciousness
could not be instantiated.
Pruss’s thought experiment brings out very vividly how Millikan’s
defnition of biological teleology fails to capture any form of immanent
teleology. The present function of an organ or organelle depends on her
defnition on remote facts in the past, and even on past facts that are
causally unrelated to the present. If human thought and intention depend
on the teleology of the human body, then thought and intention are also
extrinsic to our present constitution and operation, which is incredible.
As I mentioned in Section 2, an Aristotelian account of organisms pro-
vides clear advantages for epistemology. Take, for example, our perception
of secondary qualities. Unless we are perceiving real qualities in nature, and
qualities that really are as we perceive them to be, all of our empirical knowl-
edge is vulnerable to skeptical challenge. Microphysicalists cannot suppose
that secondary qualities (as we perceive them) are real, since scientifc theory
has no place for any counterpart to color, smell, and so on at the micro-
physical level. In contrast, hylomorphists can suppose thermal substances to
have fundamental powers of mutual interaction that correspond closely with
the appearance of color, sound, and odor. The quantum-level interactions
of particles are less, rather than more, fundamental than these chromatic,
acoustic, and olfactory interactions between thermal substances. Organisms
have evolved fundamental passive powers of responding reliably to these
active powers, resulting in veridical perception of the qualitative features
of inorganic substances. This provides a metaphysical foundation for J. J.
Gibson’s (1979) ecological theory of color, rehabilitating the “secondary”
qualities to frst-class status.
Anti-realists about secondary qualities could object that the hylomor-
phic view fails to secure the veridicality of color perception, since colors as
they appear to us are categorical (non-dispositional) and non-relational,
while the qualities of thermal substances are merely powers to afect
other substances. But in a causal-powers ontology, the activation of a
causal power is a categorical and intrinsic property of its bearer. When
we perceive the color of a substance, we are perceiving an accident of
action, the actualization of a corresponding power to act. An accident of
action can both be intrinsic to a substance and make essential reference
to something beyond it. The critic is right in thinking that colors as we
perceive them are not perceived as the actualization of something relative
to us, but as the actualization of a mind-independent feature. That aspect
of our phenomenology is preserved in the Aristotelian account, since the
qualities involve powers to act on both inorganic thermal substances and
our sensory organs.
But is the Aristotelian account of organisms compatible with the facts
of biological evolution? Wasn’t Aristotle committed to the eternal fxity
of the species? In fact, the fxity of species is no more central to Aristotle’s
system than is the constancy of the celestial spheres. Aristotle’s natu-
ral philosophy depends on the existence of individual substantial forms
and on the existence of relations of objective resemblance between those
forms, resulting in a nested, species-genus structure of taxonomy. All of
these features are fully compatible with (and even partly explained by)
the theory of evolution. And, as I argued earlier in this section, evolution
actually depends on the existence of organismic substantial forms.
In principle, there would be no problem for hylomorphism even if every
biological individual were ontologically sui generis. And the transition
from one species to another is not a problem. Aristotle recognized that
the environment plays a role, along with the parents, in every case of
reproduction: “man and the sun beget man” (Physics ii.2, 194b13).
88 Robert C. Koons
But Aristotle’s theory does require a principled distinction between sub-
stantial forms and mere accidents, and a neo-Darwinian theory would
seem to blur that distinction, supposing that the essence of a biological
individual consists of nothing but the accumulation of favored accidents.
Darwinian evolution depends on a pre-existing genetic variety within spe-
cies, which corresponds (in Aristotelian terms) to the members’ possession
of contrary contingent accidents. As a new species emerges, what had been
mere contingent accidents take on new functions, enabling the organisms
with those accidents to better exploit some ecological niche. Therefore, it
seems, at least prima facie, that all evolutionary change involves merely
changes in the distribution of accidents within related populations. Where
is the need for the substantial form and its essence?
For Aristotelians, the distinction between substantial form and accident
depends on a set of asymmetric explanatory relations. A substantial form
explains the organism’s potentiality for certain accidents, and not vice
versa. This is compatible with neo-Darwinian theory, which is silent on
the explanatory priority between biological forms.
For hylomorphists, there is a traditional distinction between “proper”
and contingent accidents. An accident is proper (a proprium) if it fows
from the specifc nature of the substantial form; otherwise, it is contin-
gent.9 Among contingent accidents, some are permanent (like sex or
handedness) and others are changeable. The permanent but contingent
accidents are explained by the combination of the nature of the substan-
tial form with certain contingent facts about the process by which the
organism was generated and originally developed. Changeable accidents
are explained by combining the nature of the substantial form with con-
tingent facts about the subsequent history of the organism.
Now, the crucial question is: can the Aristotelian explain how some
accident could be a contingent accident of an ancestral form of the organ-
ism but a proper accident of the organism itself (or vice versa)? Doesn’t
this require a sharp, un-Darwinian transition between one generation (for
which the accident is contingent) and the next one (for which the accident
is proper)?
Not necessarily. It could be that during evolution a population comes
to realize (unstably) substantial forms that belong simultaneously to two
distinct species. The substances in such a population would have two,
contrary propria, each one a diferentia of a diferent species. Each mem-
ber of the population would be abnormal relative to one or both of its
species, lacking one or both of the contrary propria. Relative to one spe-
cies, the accident is permanent but contingent, and relative to the other it
is proper. In such a population, there would be an unusual and unstable
overdetermination in the explanation of the possession of certain proper
accidents. The presence of a proper accident that is a diferentia of one of
the two species would be explained twice over: once by direct implication
from one of the specifc natures of the substantial form and in another
way by contingent factors in the substance’s generation and development
that result in the organism’s deformation relative to the other species. The
transitional organism’s substantial form would have a natural disposition
toward both contrary diferentiae. This would be an unusual situation but
not an impossible one.
Alternatively, even if it were true that each organism must belong to a
unique species, this would not create an insuperable problem for reconcil-
ing Aristotle and evolution. We could suppose that transitional popula-
tions consist of a mixture of two species, with members of each species
able to mate with the other, and with ofspring of both species possible
from a single pair of parents.10 The coexistence of distinct species would
refect the fact that the population stands on a borderline between two
competing explanatory schemes. Some organisms will fall on one side of
the boundary or the other, due to its particular confguration of accidents.
Eventually, selective pressures could eliminate the representation in the
population of the older species, completing the origin of a new species,
which would eventually become reproductively isolated. The Aristotelian
concept of species would not perfectly coincide with modern biological
uses in evolutionary settings, but the correspondence would in general be
quite close.
7. Conclusion
Aristotelian pluralism carves a middle path between atomism and cosmic
monism, securing a foundation for the manifest image of human life.
It acknowledges the homeliness of the world – a place in which human
freedom, agency, and knowledge can exist without threat of nihilism or
corrosive skepticism. Modern science seemed to threaten this world with
a universal acid of atomistic reductionism, but the implications of the
quantum revolution enable us to set the world right again.
Notes
1. See, for example, Lewis (1947, chapter 2), Plantinga (2010, chapter 10),
Koons (2017, 2018c), and Steward (2012).
2. I side here with Thomas Aquinas (and, I believe, with Aristotle himself)
in afrming the “unicity” of substantial form: that is, the thesis that every
substance has a single substantial form. Some contemporary theorists (e.g.,
Jaworksi 2016) side instead with Scotus and most later scholastic philoso-
phers in allowing for multiple substantial forms. They face what I take to be
a decisive objection: if the substantial form is that by which everything in the
substance (both material parts and accidents) receive their actual existence,
then multiple substantial forms would introduce an unacceptable form or
overdetermination or redundancy in nature.
3. My views about the role of matter have changed signifcantly from my 2014
paper (Koons 2014). I now think that matter’s role as an enduring substrate
through change (emphasized in Aristotle’s Physics) is secondary to its more
90 Robert C. Koons
central role as individuator. I now think of the persistence of matter through
substantial change as non-fundamental, as the persistence of a kind of ens suc-
cessivum (a series of fundamentally distinct, time-bound entities that are tied
together into a kind of causal sequence). Thus, I have moved somewhat in the
direction of Scaltsas (1994) and Marmodoro (2013). However, unlike Scaltsas
and Marmodoro, I still believe that there are real and not merely conceptual
distinctions among a substance, its form, and its matter. The three must be
really distinct in order to play three distinct explanatory roles in metaphysics.
4. In the GRWm ontology, parcels of material “gunk” do have defnite loca-
tions, but particles are not part of the fundamental ontology.
5. The Drossel-Ellis account, like GRW, involves objective quantum collapses,
but in their account the collapses are precipitated in a top-down fashion by
thermal properties of the system.
6. To be absolutely complete, I would have to add heaps of groups and groups
of groups to the list.
7. These were pressed at an American Catholic Philosophical Association meet-
ing in 2019 by physicist Stephen Barr.
8. Is the existence of sharp boundaries an ontological or merely an epistemologi-
cal condition? That is, could there be two thermal substances in contact with
each other, where the boundary in space (or time) is merely a metaphysical
boundary but not a physical or chemical discontinuity? If such were possible,
it would be difcult for us to detect the existence of two adjoining thermal
substances, as opposed to a single one. Perhaps we could use facts about the
past or future history of the substances to make the distinction, relying on
empirical laws of the generation and corruption of thermal substances.
9. Although propria are fully explained by a substance’s essence, it is possible
for a substance to lack one of its propria, through genetic defect or injury.
Being bipedal is a proprium of human nature, but not all human beings have
two legs in fact.
10. It is important to note that metaphysical species do not necessarily correspond
one to one with what biologists mean by “species”. Biologists use a number
of criteria to distinguish one species from another, and none of these criteria
will always carve nature at the joints, metaphysically speaking. Nonetheless,
just as new species in the various biological senses arise by descent with
modifcation, the same must be true of metaphysical species.
Bibliography
Adams, Robert M. (1987), “Flavors, colors, and god,” in The Virtue of Faith and
Other Essays in Philosophical Theology, New York: Oxford University Press,
pp. 243–262.
Anderson, P. W. (1972), “More is diferent,” Science, New Series 177(4047):
393–396.
Bangu, Sorin (2009), “Understanding thermodynamic singularities: Phase transi-
tions, data and phenomena,” Philosophy of Science 76:488–505.
Bell, J. S. (1987), Speakable and Unspeakable in Quantum Mechanics, Cam-
bridge: Cambridge University Press.
Brower, Jefery (2017), “Aquinas on the individuation of substances,” Oxford Studies
in Medieval Philosophy 5(1). https://doi.org/10.1093/oso/9780198806035.
003.0004.
Cartwright, Nancy (1994), Nature’s Capacities and Their Measurement, Oxford:
Clarendon Press.
——— (1999), The Dappled World: A Study of the Boundaries of Science, Cam-
bridge: Cambridge University Press.
Castellani, Elena, ed. (1998), Interpreting Bodies: Classical and Quantum Objects
in Modern Physics, Princeton: Princeton University Press.
Clifton, Robert and Hans Halvorson (2001), “Entanglement and open systems in
algebraic quantum feld theory,” Studies in History and Philosophy of Modern
Physics 32:1–31.
Compagner, Aaldert (1989), “Thermodynamics as the continuum limit of statisti-
cal mechanics,” American Journal of Physics 57(2):106–117.
Dawkins, Richard (1976), The Selfsh Gene, Oxford: Oxford University Press.
Drazin, P. G. and W. H. Reid (2004), Hydrodynamic Stability, second edition,
Cambridge: Cambridge University Press.
Drossel, Barbara (2015), “On the relation between the second law of thermody-
namics and classical and quantum mechanics,” in Why More Is Diferent:
Philosophical Issues in Condensed Matter Physics and Complex Systems, B.
Falkenburg and M. Morrison, eds., Berlin: Springer, pp. 41–54.
Earman, John (2004), “Curie’s principle and spontaneous symmetry breaking,”
International Studies in Philosophy of Science 18:173–198.
Ellis, George F. R. and Barbara Drossel (2018), “Contextual wavefunction col-
lapse: An integrated theory of quantum measurement,” New Journal of Phys-
ics, 20:113025.
Feser, Edward (2008), The Last Superstition: A Refutation of the New Atheism,
South Bend, IN: St. Augustine Press.
Fraser, Doreen (2008), “The fate of ‘particles’ in quantum feld theories with
interactions,” Studies in History and Philosophy of Modern Physics
39:841–859.
Gibson, J. J. (1979), The Ecological Approach to Visual Perception, Boston:
Houghton-Mifin.
Hendry, Robin Findlay (2010), “Emergence vs. reduction in chemistry,” in Emer-
gence in Mind, Cynthia MacDonald and Graham MacDonald, eds., Oxford:
Oxford University Press, pp. 205–221.
Inman, Ross D. (2018), Substance and the Familiarity of the Familiar: A Neo-
Aristotelian Mereology, New York: Routledge.
Jaworski, William (2016), Structure and the Metaphysics of Mind: How Hylo-
morphism Solves the Mind-Body Problem, Oxford: Oxford University Press.
Koons, Robert C. (2014), “Staunch vs. faint-hearted hylomorphism: Toward an
Aristotelian account of composition,” Res Philosophica 91 (2014):1–27.
——— (2017), “The epistemological and ontological superiority of hylomor-
phism,” Synthese https://doi.org/10.1007/s11229-016-1295-6 (2017).
——— (2018a), “Hylomorphic escalation: A hylomorphic interpretation of
quantum thermodynamics and chemistry,” American Catholic Philosophical
Quarterly 92:159–78.
——— (2018b), “Forms as simple and individual grounds of things’ natures,”
Metaphysics 1(1):1–11, https://doi.org/10.5334/met.4.
——— (2018c), “The argument from intuition,” in Two Dozen (or So) Argu-
ments for God: The Plantinga Project, Trent Dougherty and Jerry Walls, eds.,
Oxford: Oxford University Press, pp. 238–260.
——— (2019a), “Thermal substances: A Neo-Aristotelian ontology for the quantum
world,” Synthese 198:2751–2772, https://doi.org/10.1007/s11229-019-02318-2.
92 Robert C. Koons
——— (2019b), “Hylomorphism and our knowledge of value,” in Being, Good-
ness, and Truth: Proceedings of the Society for Medieval Logic and Metaphys-
ics, volume 16, Alex Hall, ed., Newcastle upon Tyne: Cambridge Scholars
Publishing, pp. 78–90.
______ (2020), “Remnants of substances: A neo-Aristotelian resolution of the
puzzles,” Quaestiones Disputatae 10(2):53–68.
——— (2021), “Powers ontology and the quantum revolution,” European Jour-
nal for Philosophy of Science 11(1):1–28.
Koons, Robert C. and Pruss, Alexander R. (2017), “Must a functionalist be an
Aristotelian?,” in Causal Powers, Jonathan D. Jacobs, ed., Oxford: Oxford
University Press, pp. 194–204.
Laughlin, Robert B. (2005), A Diferent Universe: Reinventing Physics from the
Bottom Down, Cambridge, MA: Basic Books.
Lee, Patrick and Robert P. George (2009), Body-Self Dualism in Contemporary
Ethics and Politics, Cambridge: Cambridge University Press.
Levine, J. (2000), Purple Haze: The Puzzle of Conscious Experience, Cambridge,
MA: MIT Press.
Lewis, C. S. (1947), Miracles: A Preliminary Study, London: Centenary Press.
Liu, Chuang (1999), “Explaining the emergence of cooperative phenomena,” Phi-
losophy of Science 66(Proceedings):S92–S106.
Lowe, E. J. (1992), “The problem of psychophysical causation,” Australasian
Journal of Philosophy 70(3):263–276, https://doi.org/10.1080/000484
09212345161.
Marmodoro, Anna (2013), “Aristotle’s hylomorphism without reconditioning,”
Philosophical Inquiry 36(1–2):5–22.
Molnar, George (2003), Powers: A Study in Metaphysics, Oxford: Clarendon
Press.
Oderberg, David (2006), “Teleology: Organic and inorganic,” in Contemporary
Perspectives on Natural Law: Natural Law as a Limiting Concept, Ana Marta
González, ed., Aldershot, UK: Ashgate, pp. 259–279.
Pickstock, C. J. C. (2015), “Matter and mattering: The metaphysics of Rowan
Williams,” Modern Theology 31(4):599–617.
Plantinga, Alvin (2010), Where the Confict Really Lies: Science, Religion, and
Naturalism, New York: Oxford University Press.
Prigogine, Ilya (1997), The End of Certainty: Time, Chaos, and the New Laws of
Nature, New York: Free Press.
Primas, Hans (1980), “Foundations of theoretical chemistry,” in Quantum
Dynamics for Molecules: The New Experimental Challenge to Theorists, R. G.
Woolley, ed., New York: Plenum Press, pp. 39–114.
——— (1981), Chemistry, Quantum Mechanics, and Reductionism: Perspectives
in Theoretical Chemistry, Berlin: Springer-Verlag.
Redhead, Michael and Paul Teller (1991), “Particles, particle labels, and quanta:
The toll of unacknowledged metaphysics,” Foundations of Physics,
21:43–62.
Rota, Michael (2011), “Causation,” in The Oxford Handbook of Aquinas, Brian
Davies and Eleonore Stump, eds., Oxford: Oxford University Press,
pp. 104–114.
Ruetsche, Laura (2006), “Johnny’s so long at the ferromagnet,” Philosophy of
Science 73:473–486.
Scaltsas, Theodore (1994), “Substantial holism,” in Unity, Identity, and Explana-
tion in Aristotle’s Metaphysics, Theodore Scaltsas, David Charles, and Mary
Louise Gill, eds., Oxford: Clarendon Press, pp. 107–128.
Schafer, Jonathan (2010), “Monism: The priority of the whole,” Philosophical
Review 119:31–76.
Sellars, Wilfred (1962), “Philosophy and the scientifc image of man,” in Frontiers
of Science and Philosophy, Robert Colodny, ed., Pittsburgh, PA: University of
Pittsburgh Press, pp. 35–78.
Sewell, Geofrey L. (2002), Quantum Mechanics and its Emergent Macrophysics,
Princeton, NJ: Princeton University Press.
Simpson, W. M. R. (2019), What’s the Matter? Toward a Neo-Aristotelian Ontol-
ogy of Nature, Doctoral thesis, University of Cambridge.
Steward, Helen (2012), A Metaphysics for Freedom, Oxford: Oxford University
Press.
Strocchi, Franco (1985), Elements of Quantum Mechanics of Infnite Systems,
Singapore: World Scientifc.
Von Neumann, John (1931), “Die Eindeutigkeit der Schrödingerschen Opera-
toren,” Mathematische Annalen 104: 570–588.
Woolley, R. G. (1988), “Quantum theory and the molecular hypothesis,” in Mol-
ecules in Physics, Chemistry, and Biology, Vol. I, Jean Maruani, ed., Dordrecht:
Kluwer Academic, pp. 45–89.
3 Restoring the Hierarchy
of Being
David S. Oderberg
1. Introduction
The idea of a ‘Great Chain of Being’ was, historically, one of the most
durable and persistent of highly general ideas embedded in Western phi-
losophy.1 From Plato through much of the eighteenth century, the vast
majority of philosophers and other thinkers believed in what I will call,
more prosaically, a hierarchy of being in the universe. Yet during the
nineteenth century the idea of a hierarchy of being died out, not to be
revived or reconsidered at any time since. It went the way of so many
other theories and ideas prominent in medieval philosophy and almost
invariably rooted in Greek thought, usually that of Aristotle but in this
case primarily the thought of Plato. The post-medieval period, as a whole,
is unifed most clearly by its taking a wide broom to what is now thought
of as the detritus of a ‘pre-Enlightenment’ worldview. The hierarchy of
being along with so much else was swept away, never, it is almost univer-
sally hoped, to return.
As a point of methodology, philosophers ought not to be too ready to
brush aside ideas and theories that have a literal two millennia of intel-
lectual support behind them. If their inheritance is a rejection of some
such idea or theory, it behoves them to be open-minded concerning its
re-evaluation. Can the hierarchy of being be restored, or at least recovered
from the landfll of intellectual history? Once we separate that which is
viable from that which is not in the thicket of ingredients that make up
the historical Great Chain, I think we can restore a theory that stands
up to a modern critical eye. The process of restoration does require the
reafrmation of assumptions and positions that will still not fnd favour
with many philosophers; one cannot sell the entire farm in order to give
the hierarchy of being a new lease of life. That said, I believe we can do
much to rehabilitate it in a form, and with all the appropriate caveats, that
gives it renewed interest to a relatively unprejudiced mind.
The following discussion aims to show how this can be done. It begins
with some necessary explanation of the hierarchy in its historical con-
text, but then moves towards a dehistoricisation of the idea, stripping
DOI: 10.4324/9781003125860-5
Restoring the Hierarchy of Being 95
it down to its basic elements – these being grounded in Aristotle and
Aquinas. I aim to show how the notion of metaphysical superiority
at the core of the hierarchy can be given a rigorous defnition able to
handle both old and new counterexamples that have or can be levelled
against it. I leave aside the implications of a plausible hierarchy, about
which it would be far too early to speculate. In this sense, the ambit of
the discussion is narrow – the metaphysical defence of a hierarchy of
being recognisable to its past adherents yet plausible to relatively recep-
tive modern mind.
2. Lovejoy and the Hierarchy

As is well known to historians of ideas, Arthur Lovejoy’s seminal book
The Great Chain of Being virtually founded the discipline of the history of
ideas with its groundbreaking study of the hierarchy of being.2 The book
had an enormous infuence, spawning critical analyses and retrospective
studies both of its correctness and of its impact.3 From Plato to Romanti-
cism, via Aristotle, the neo-Platonists, St Augustine, St Thomas Aquinas,
Bruno, Spinoza, Locke, Leibniz, Kant, along with lesser philosophers and
also historians, naturalists, poets, and novelists, Lovejoy managed to cap-
ture the idea of the hierarchy as a subtle complex of sub-ideas adhered
to in one form or other, though they agreed largely on essentials, by two
millennia of thinkers. Sometimes only a poet has the pithiness and turn
of phrase to encompass a deep philosophical idea:
Vast Chain of Being! Which from God began,

Natures ethereal, human, angel, man,
Beast, bird, fsh, insect, what no eye can see,
No glass can reach; from Infnite to thee,
From thee to Nothing.4
What place was occupied by the idea of a hierarchy of being, so admirably

and succinctly expressed by Pope? Lovejoy rightly saw it as ‘a part of the
history of Western man’s long efort to make the world he lives in appear to
his intellect a rational one.’5 It answered the question: why should ‘the world
as a whole be what it is?’6 If, asserted Lovejoy, the world’s ‘being’, ‘extent’,
and ‘range of diversity which its components exhibit’ have no ‘intelligible
reason’, if they ‘might equally well have been other than they are’, then ‘the
constitution of the world is but a whim or an accident’.7 He explicitly added,
‘conformity to the very curious set of primary laws which empirical science
discovers’ – at least part of which we now think of as the ‘fne-tuning’ of the
universe – was an element of the explanandum. Such conformity, accord-
ing to so many thinkers throughout much of intellectual history, could not
be a ‘brute fact’.8 Clearly, then, the Principle of Sufcient Reason under-
lay the idea of the hierarchy as an explanatory concept:9 in general terms,
96 David S. Oderberg
the contingent diversity and behaviour of the universe and everything in it
required a complete and adequate explanation.
One might argue about the explanatory reach or adequacy of the hier-
archy. Would it explain conformity to the peculiar laws of our cosmos?
One might question the extent to which the hierarchy consisted of a genu-
ine explanation of known phenomena or else a putative explanation in
search of phenomena to explain. Whatever one’s view, the picture we
ended up with, transmitted from generation to generation over millennia,
involved a
conception of the plan and structure of the world which, through the
Middle Ages and down to the late eighteenth century, many philoso-
phers, most men of science, and, indeed, most educated men, were to
accept without question – the conception of the universe as a ‘Great
Chain of Being’.10
It was
composed of an immense, or – by the strict but seldom rigorously

applied logic of the principle of continuity – of an infnite, number
of links ranging in hierarchical order from the meagerest kind of
existents, which barely escape nonexistence, through ‘every possible’
grade up to the ens perfectissimum – or, in a somewhat more ortho-
dox version, to the highest possible kind of creature, between which
and the Absolute Being the disparity was assumed to be infnite –
every one of them difering from that immediately above and that
immediately below it by the ‘least possible’ degree of diference.11
As Lovejoy explains, the hierarchy is a master idea constituted by three

main sub-ideas. According to the Principle of Plenitude, ‘the universe is a
plenum formarum [plenum of forms] in which the range of conceivable
diversity of kinds of living things is exhaustively exemplifed’ (emphasis in
original).12 Put in more general terms, ‘[A]ll conceptual possibilities must
be realised in actuality.’13 The Principle of Continuity states: ‘If there is
between two given natural species a theoretically possible intermediate
type, that type must be realized.’14 If Continuity were false then Plenitude
would also be false. But if Plenitude were false, there would be a lack or
defect in what neo-Platonists thought of as the Good or the Absolute,
what for Christian adherents of the hierarchy is God, what for any sup-
porter of the hierarchy was the eternal and infnitely fecund source of the
boundless diversity of the cosmos. The necessary outpouring of creative
diversity would have a limit, would contain gaps into which further cre-
ative acts could be poured but were not.
The third principle in Lovejoy’s account – the one to which he gives
least attention yet which, if there is to be a promising restoration of the
hierarchy, should be given most attention – is the Principle of Gradation:
all things may be ‘arranged in a single graded scala naturae [ladder of
nature] according to their degree of “perfection”’.15 In fact it is Gradation,
rather than Plenitude or Continuity, that historically represented the hier-
archy in visual terms. In books of philosophy, theology, alchemy, natural
history, and politics, generations of educated people were presented with
a ladder or set of steps from the least perfect to the most perfect of beings
in the universe, with God or the Creator above all. Angels were immedi-
ately below God and at the top of creation, followed by man, other living
creatures, non-living creatures, and (although not depicted, for obvious
reasons) prime matter or pure potentiality, the closest being to nothing-
ness and the most removed from the ‘pure actuality’ that is God.
3. The Breaking of the Chain

As we know, the Great Chain of Being eventually broke. So cleanly did it
break that there has not been a hint, in the last two hundred years, of its
making the slightest comeback. Its fracture was the result of a confuence
of numerous causes, some theological, others philosophical, yet others
political or scientifc. One might say that for every domain of human
knowledge in which the hierarchy played a foundational role, there
emerged from within that domain a current of thought – of thinkers –
railing against it or at least chipping away at the hierarchy, weakening
its links until, in our own age, nothing of it remains. As the hierarchy
was arguably the ideological centrepiece of the pre-Enlightenment, pre-
Scientifc Revolution intellectual framework, so its nullity is the hallmark
of the ‘modern’ or ‘scientifc’ worldview. This even though many Enlight-
enment thinkers, scientists included, persisted in adhering to some form
of the hierarchy.
For Lovejoy, the main causes of the breaking of the Chain were the
infuences of Evolutionism and of Romanticism.16 Evolutionism – as
opposed to the nineteenth-century theory of evolution – was a move-
ment of thought encompassing both a particular biological outlook and
a general adherence to principles of development and progress that ‘tem-
poralised’ the Chain of Being, to use Lovejoy’s term. Although Evolution-
ists did not abandon the idea of gradation in nature, the gradation was
no longer part of a static, immovable metaphysical structure but instead
an unfolding process that was essentially temporal and dynamic. The
specifc view of evolution that found its apotheosis in Charles Darwin’s
theory already had its seeds in Evolutionist philosophy both metaphysical
and natural. We fnd anything from strands of Evolutionism to concrete
empirical theories (minus the range and wealth of data crucial to Darwin)
in thinkers as diverse as Leibniz, Voltaire, Kant, Bufon, and Lamarck –
although both Kant and Leibniz also afrm the metaphysical hierarchy,
the latter more clearly and explicitly than the former.17
The other solvent of the Chain, as Lovejoy saw it, was the emergence of
Romanticism in the late eighteenth century and its development through
the nineteenth, as represented by poets, philosophers, and theologians
such as Herder, Schleiermacher, and Schelling.18 In Schelling,
[n]ot only had the originally complete and immutable Chain of Being
been converted into a Becoming, in which all genuine possibles are,
indeed, destined to realization grade after grade, yet only through a
vast, slow unfolding in time; but now God himself is placed in, or
identifed with, this Becoming.’19
In Romanticism, an ‘absolute and static world’ is transformed into ‘one of

diversity and becoming’.20 Movement, process, and self-development, the
unfolding of divinity in nature, self-realization through the striving of the
spirit: all of these Romantic themes contributed to the complete inversion
of the ‘entire Platonic and Aristotelian scheme of the Universe’.21
Lovejoy is surely right that these two currents of thought undermined
the hierarchy, though one might argue over how dominant either of
them was. Moreover, without going deeply into historical questions of
infuence and priority we should also, for background, note other impor-
tant movements and schools of thought that undermined the hierarchy.
Protestantism overall,22 with its emphasis on the democratisation and
privatisation of religion, was inherently antithetical to static, heavily
metaphysical hierarchies of being. The infuence here is twofold. On one
hand, political and ecclesiastical hierarchies were often bound up with
the primary metaphysical hierarchy of being, so Protestantism’s corrosion
of the primary hierarchy was indirectly due to its overall hostility to the
hierarchy’s political and ecclesiastical manifestations. On the other, given
general Protestant hostility to Aristotelian and Scholastic metaphysics,
it was unlikely that a hierarchy developed and refned through medieval
philosophy could survive the success of Protestantism in many parts of
formerly Catholic Europe.
The Scientifc Revolution of the seventeenth and eighteenth centuries
also played its role in the hierarchy’s demise.23 Through Galileo and New-
ton, the Copernican Principle became established in physics and cosmol-
ogy: man was not a ‘privileged observer’ of the universe.24 We are now
accustomed to the magisterial pronouncements of public science (whether
from scientists or, as we might call the other variety, scientistes) that
man is an insignifcant species on an equally insignifcant planet located
on a minor spiral arm of one among a hundred billion galaxies foat-
ing around an indiferent and possibly infnite universe.25 It is easy to
see that any conceptual scheme resembling the hierarchy of being could
not survive this fundamental decentralisation of mankind. If we add into
the mix the general, irresistible trend of anti-feudalism and burgeoning
political egalitarianism, along with the Enlightenment – at least in its more
revolutionary and anti-theistic moments, given that many Enlightenment
thinkers continued to believe in something like the hierarchy – we end
up with a mix of philosophy, theology, economics, politics, and social
attitudes that pulled the hierarchy to pieces, placing it ostensibly beyond
rehabilitation.
It would be a caricature, however, to suggest that the backlash against
the hierarchy was an assault by early modern sophistry against a pristine
medieval structure whose pure authenticity had survived the test of time.
The hierarchy of being, as passed down through the intellectual genera-
tions, had acquired an encrustation of philosophical limpets and concep-
tual barnacles that made it, quite rightly, appear increasingly less plausible
to the early modern mind and its heirs. These excesses did the hierarchy
no favours. At best, they appeared as bewildering, metaphysically weak
epicycles that served up glib answers to serious questions. At worst, they
acted as a kind of metaphysical Star Chamber, wielded to uphold an exist-
ing order of things because that order had to be that way lest the beauty,
symmetry, and completeness of the hierarchy be found wanting. Putting
it in fairly general terms, Lovejoy tells us:
Since every place in the scale must be flled, and since each is what it
is by virtue of the special limitations which diferentiate it from any
other, man’s duty was to keep his place, and not to seek to transcend
it . . . The good for a being of a given grade, it seemed evident, must
consist in conformity to its type.26
More specifcally and with greater portentousness, he goes on:
[N]o great improvement in men’s political behavior or in the organi-

zation of society could be hoped for . . . The principles of plenitude
and gradation could, in this way, among their many uses, be made to
serve the purposes of a species of pessimistic and backhanded apolo-
getic both for the political status quo and for the accepted religion.27
As Pope puts it with his usual succinct clarity:
Order is Heav’n’s frst law; and this confest,

Some are, and must be, greater than the rest,
More rich, more wise . . .28
From Samuel Richardson’s novel Pamela we have:
Wise Providence
Does various parts for various minds dispense:
The meanest slaves, or those who hedge and ditch,
Are useful, by their sweat, to feed the rich.
The rich, in due return, impart their store,
Which comfortably feeds the lab’ring poor.
Nor let the rich the lowest slave disdain:
He’s equally a link of Nature’s chain;
Labours to the same end, joins in one view;
And both alike the Will Divine pursue;
And, at the last, are levell’d, king and slave,
Without distinction, in the silent grave.29
The appeal to a hierarchy of being in order to reinforce social, political,

economic, and religious roles was one of the main excesses, as I would put
it, of hierarchical thinking. Yet we need to tread carefully when exposing
the mistake involved in this sociopolitical accretion to the ontological
structure. We can readily admit, as nearly all of us moderns would, that
an ossifed hierarchy of the type espoused even in the eighteenth century
was bound to perpetuate and reinforce many kinds of social injustice,
artifcially cramping the horizons not only of groups but also of individu-
als with their legitimate aspirations of improvement and advancement.
We dispose of this worldview too quickly, however, merely by noting the
injustice. Rather, there was a metaphysical mistake at its heart, namely
that of confusing the essence of the kind with the essence of the role. In
other words, there is no such thing as a monarchical essence, or a feudal
essence, a papal or episcopal essence, or a knightly essence in the sense of
belonging to the bearer of any such role. What has the essence is the role
itself: the monarchical role is essentially defned in terms of its essential
constituents – being a head of state, practising one-person rule, of a noble
lineage, ceremonially installed, and so on (the essence of the monarchical
role being an interesting metaphysico-political question in its own right).
The same for any other role, be it political, social, religious, or economic.
The monarch, however, does not have the monarchical essence. The mon-
arch’s essence is the same as yours, mine, and the serf’s – that of a human
being, a rational animal. The monarchical essence belongs to the role,
not the person occupying the role. We can reasonably suspect that the
sociopolitical excesses of the hierarchy, as seen in Pope and Richardson,
were encouraged by a confused and illegitimate metaphysical transfer
of essentialist thinking about an ontological hierarchy to the concept of
a sociopolitical one. This does not, though, rule out sociopolitical hier-
archies. All it means is that whether any such hierarchy is justifed, and
whether and to what extent it is fxed, will be a matter for debate outside
metaphysics – debate in ethics, political philosophy, theology, economics,
and the like.
A fnal observation to make before moving on to more positive ideas
concerns what might be considered an excess as well, but of a kind more
understandable and excusable. This is the extravagance involved in apply-
ing the metaphysical hierarchy to empirical science, particularly biology.
So, for example, we have Charles Bonnet, the eighteenth-century natu-
ralist, producing a hierarchy in which, whilst man is at the sublunary
summit (as per the traditional scale), sea anemones are at the bottom of
the ladder of the living, with tube worms30 above moths, eels above water
snakes, and gall insects31 above tapeworms. Descending the ladder, we
fnd Bonnet placing stones above salts, salts above metals, earth above
water, water above air, and air above fre (the ‘four elements’ still in the
picture).32 It is easy for us to scof at these pseudo-gradations, but given
the state of physical and biological knowledge at the time one could no
doubt have found a fair bit of good sense in Bonnet’s distinctions. To
that extent, what come across as yet more excrescences of the metaphysi-
cal hierarchy could more generously be considered legitimate scientifc
hypotheses, respectable for their time if now found lacking in empirical
support (for reasons all too familiar to Aristotle himself, as we will see).
4. Can We Do Better?
If we are to recover the hierarchy as something at least worth considering
by the contemporary metaphysical mind, we need to get back to basics
as well as enter some methodological caveats. By getting back to basics
I mean, frst, putting aside the moral and political excesses of the hierar-
chy as either wrong or at least not justifed by distinctively metaphysical
truths. They are, in my view, a distraction from the proper understanding
of the hierarchy as a metaphysical structure. Secondly, we should attenu-
ate the connections between the hierarchy and the metaphysical specu-
lations of neo-Platonism, so admirably detailed by Lovejoy.33 To think
of the hierarchy as an ‘emanation’ of the ‘superabundance’ or ‘infnite
fecundity’ of ‘The Absolute’ might well be translatable into terms more
easily grasped by those not initiated into the profundities of neo-Platonist
metaphysics, in which there may be important truths embedded. Never-
theless, to build that into the recovery plan for a slimmer – albeit to the
modern mind still overly elaborate – structure would not, I submit, serve
the cause of appreciating what is plausible about it. Thirdly, we should
not, in explicating the hierarchy, assume the existence of God or any other
immaterial agents. Nor, at the other extreme, should its explication be
treated simultaneously as a proof of their existence. Methodologically,
our approach should be to take on the content of the basic hierarchy as
handed down by its previous adherents and to show what the structure
would look like and how it would be justifed on metaphysical grounds,
were there to be, as elements in it, God and other immaterial agents. By
not assuming the existence of God (and the rest), then, I mean that God’s
existence should not be regarded as necessary for the hierarchy to be
coherent. God would have to exist in order for the hierarchy itself to exist,
at least in its traditional form, but that would be a diferent question.
Correspondingly, although it would be misplaced to treat an argument
for the coherence of the hierarchy – for its plausibility as a structure – as
an argument for the existence of God at least with deductive or even
strong inductive force, one might think of the hierarchy as suggestive of
the existence of God. I will say (perhaps disappointingly) little about this
at the end of the discussion.
In order to restore a hierarchy that is more plausible in its own terms,
we need to get back to the spirit of Aristotelian-Thomistic metaphysics.
Here is where the core of the hierarchy is to be found. This is not, I want
to stress, an exercise in exegesis. The basic materials of the hierarchy are
indeed to be found in Aristotle and Aquinas.34 To make progress, however,
we should be concerned far less with exegetical niceties and much more
with the substance of the central ideas that can give us a hierarchy of more
than historical curiosity. In this spirit, we should frst focus on essential
powers – powers that are of the essence of their bearers. These could be
powers that enter into the very defnition of a thing, such as rationality
(for humans) or sentience (for animals), or they could be powers that
‘fow’ from the essence whilst not being a constituent part of it, such as the
ability to laugh (humans) or the ability to bend in the wind (most plants).
A power (or other property in the strict Scholastic sense of ‘proprium’)
that fows35 from the essence is necessary to a kind of entity given the
kind to which it belongs. Nevertheless, damage to a member of the kind,
such as disease or the absence of parts, can result in loss of the power or
a frustration of its manifestation. Normal functioning for the individual
according to its kind, then, is required for all of its powers to be present
and exercisable.36 We should focus on essential powers because the hier-
archy is supposed to be a necessary one as opposed to a mere potentially
temporary arrangement of things. That is, the hierarchy is supposed to be
necessary given the existence of its members. Their own existence may be
contingent, of course, but their place in the hierarchy would be necessary
as long as all the members existed. It would not, by contrast, make for a
hierarchy of more than ephemeral interest if the very same members could
be arranged diferently, which would be possible were the structure based
on merely contingent powers.
Next, the hierarchy needs to be highly generic. What counts as highly
generic is, admittedly, somewhat vague and also interest-dependent, but
there is nothing subjective about the salient joints at which the hierarchy
is carved. We have already seen that any attempt to rank all species (at
whatever level) will lead to excesses and absurdities of the kind we (albeit
understandably) found in Bonnet. At the other extreme, it would be wilful
blindness to refuse to rank anything on the spurious Aristotelian-Thomis-
tic ground that the hierarchy has always been about ranking substances,
and since every member is a substance they are all on a metaphysical
par. This kind of monism is quite obviously too much of a fattening to
be faithful to the joints of reality. Perhaps we might make one cleavage
only – between the living and the non-living? Yet if we make that – as we
should – why not go further? Can we not fnd anything of fundamental
metaphysical interest and importance separating kinds of living things?
Might we not be able to separate levels of non-living things as well? As
we will see, although signifcant generic rankings among the living are
plausible, they are less so in the case of the non-living – with perhaps one
exception.
Recall Lovejoy’s dissection of the Chain into the principles of Pleni-
tude, Continuity, and Gradation. By far the most attention is given
by him to Plenitude,37 with roughly equal attention to Continuity and
Gradation after that. Whether or not this bias is historically accurate,
on my view we need a reorientation, making the Principle of Gradation
the centrepiece of the hierarchy. Here is a more recent statement of the
principle: ‘In material and living bodies we fnd an ascending order of
perfections in which the higher beings have their own perfections as well
as those of the lower level of being.’38 Needless to say, by ‘perfections’
the author does not mean that which makes a thing perfect, since no
material or living body is or can be perfect, at least on the Aristotelian-
Thomistic view of reality. He means, rather, essential elements as well
as the properties that fow from them: those features or characteristics
whereby a thing is of the essential kind to which it belongs. To put it
tautologically, if we want a restoration of the hierarchy we need to focus
on what is hierarchical about it, not on the other – albeit interesting –
ancillary features.
Does this mean we should cast aside Plenitude and Continuity? Perhaps
in the way they were so often understood, as represented by Lovejoy, we
need to retire them from the conceptual feld. Given Continuity’s depen-
dence on Plenitude, as shown earlier, it is really the latter that has argu-
ably caused enough neo-Platonist intellectual chaos over the centuries to
bring about the downfall of the hierarchy. Yet perhaps we can, one day,
perform an intellectual makeover on both principles so as to render them
ft for a restored hierarchy based on Gradation. Consider these Scholastic
formulations of Plenitude, whose author remarks explicitly on Lovejoy’s
‘amazing misunderstanding of the principle of plenitude’:39 (P1) ‘By the
free choice of the Creator the universe of being contains all essential levels
of perfections and of natures’; (P2) ‘The superior one is represented by
many inferior beings.’40 The interpretation of both (P1) and (P2) is open
for debate, but suppose we read them as variants of the following idea:
it is necessary, for the representation of the Creator’s creative act as both
free and good, that there be a sufcient diversity of created essences of
varying levels of perfection.
This version of the Principle of Plenitude, obviously weaker that the
neo-Platonist version, encompasses two key thoughts of Aquinas. First,
God’s free creative act communicates His goodness to creatures. Their
existence thereby represents His goodness, but this representation would
not be adequate if there were only one kind of creature on its own. Hence
the very nature of God’s creative act, as defned by its objective, logically
requires that many diverse kinds of creature be produced.41 Secondly,
although the divine creation participates in and represents the goodness
of the Creator, still God could have created things superior to those He
did create.42 In other words, adequate representation (and glorifcation)
of the Creator does not entail the existence of a best possible universe,
assuming even that the concept of a best possible universe were coherent.
And this is what we should expect given the unbridgeable gap between
the infnitude of God and the fnitude of any possible creation (another
key theme of the hierarchy through the ages).
What version of the Principle of Continuity might be worth consid-
eration? Note that because Continuity is entailed by Plenitude in their
strong formulations set out earlier, if we relinquish the latter then the
former loses its most important justifcation. If we moved, as Lovejoy
would have us, towards a dynamic understanding of Continuity, we
would likely fnd the beginning of justifcation in terms of the need for
transitional forms in evolution. Even here, however, the principle would
have to be weakened from one involving a maximally strong modal claim
about logically possible intermediaries to one proposing the existence of
those intermediaries needed biologically to explain actual evolutionary
transitions. Without Plenitude, however, the logical claim about a static,
gapless hierarchy looks unmotivated.
Consider, then, these alternative formulations of a weaker Principle of
Continuity: (C1) ‘The order of the universe displays a gradual scale of
perfections from end to end through all essentially diferent intermediate
steps’; (C2) ‘Every superior nature in its least perfection or operation
borders on the highest perfection or operation of the nature ranking next
below it in the scale of being.’43 We can plausibly read (C1) and (C2) as
variants of the following thesis: the universe displays a hierarchy of over-
lapping essences at some metaphysically signifcant generic level. Again,
this idea is embedded in Thomistic thinking, as developed from Aristotle’s
own thought about the powers of the soul.44
Both Plenitude and Continuity in their Scholastic forms play roles in the
formulation of the hierarchy I will give. The role of Plenitude is implicit
and will merely be gestured at towards the end. That of Continuity is
explicitly built into the hierarchy by virtue of its Aristotelian-Thomistic
form. Both are worthy of further exploration, but the emphasis must
be on Gradation as the keystone of the hierarchy. Consider even that in
the case of Continuity in particular, it is not mere overlap that matters
but graded overlap, as both (C1) and (C2) make clear. Maybe the three
principles handed down by Lovejoy’s reconstruction ought to be reduced
to two, with a single Principle of Continuous Gradation accompanied
by the Principle of Plenitude. That said, I think we should regard the
phenomenon of overlap as playing metaphysical second fddle to that
of superiority, since it is the latter that makes the hierarchy a hierarchy.
5. Defning Superiority
Before proceeding to a rigorous defnition of metaphysical superiority,
now is the time to present the hierarchy in its full Aristotelian-Thomistic
form:
The Hierarchy
Creator of Pure Act God Mind without body

[Active intellect without reason]
Pure Mind Angels [Passive intellect without reason]
Living Rational Man Body, life, mind, reason
Organic Sentient Animals Body, life, and mind without reason
Vegetative Plants Body and life without mind
Inorganic Bodies Body without life
Receiver of Pure Potency Prime matter [Matter without form]
Figure 3.1 Diagram of the hierarchy of being
Here are a few comments on the structure. (i) The generic beings and
kinds of being on display are standard for the Scholastic conception,
albeit put together in a complete way not usually seen. (ii) With regard
to the horizontal lines, prime matter is separated from all the rest due
to its not being a substance. It has no independent existence, requiring
complementation by substantial form in order for a material substance
to exist. God and angels are separated from the rest because they are
immaterial. God is separate from the rest for obvious reasons, including
that He is not a species, a genus, or a member of any species or genus.
(iii) With regard to the left-hand side, prime matter as pure potency
receives the material forms above it, as already indicated. In the opposite
direction, God as pure actuality creates all that is below. (iv) Moving
further to the right, everything from God to plants is enclosed in a curly
bracket because all of these are living beings. Within these, humans,
animals and plants are enclosed in a smaller bracket because they are
the organic living beings, that is, they rely on material constituents and
conditions for their lives. (v) Further along, as God is pure actuality
and prime matter is pure potency, so in between there is a denomina-
tion for each grade of being according to its specifc diference – what
separates it from everything else in the hierarchy. These are the usual
Scholastic denominations, but note that by calling angels pure minds I
do not mean that God is not a pure mind; rather, angels are the only cre-
ated pure minds. (vi) On the right-hand side, the specifc diferences are
explicitly laid out, thereby demonstrating the overlap between adjacent
members of the hierarchy – the kind of continuity that is central to the
Aristotelian-Thomistic conception.
To elaborate, we should note that prime matter overlaps bodies
because bodies really contain prime matter. Angels and humans over-
lap because they both have minds; the same for God and angels. That
said, overlap in the hierarchy is no more a transitive relation than any-
where else in metaphysics. God and angels overlap humans since all
have minds, but neither God nor angels overlap bodies or prime matter
even though humans do (on the standard Thomistic view).45 (vii) God
and angels are grouped together as minds without bodies although in
no way whatsoever belonging to any common species or genus, since
God is in neither. In a more tentative way, I have explained the difer-
ence between them, as far as having a mind goes, in terms of the kind
of mind. Although neither God nor angels have reason – they do not
ratiocinate – they both have intellects. God’s intellect is wholly active,
not in the same way as a human’s active intellect, which engages in
reasoning and abstraction, but in the way of an omniscient being. The
intellect of angels is, I venture to suggest, either partly or purely passive
and non-discursive. Angels know only what is imparted to them by God
and humans.46 I will say more about this later.
We need a concept of metaphysical superiority that will explain the gra-
dation at the centre of this classic Scholastic hierarchy of being. Talking
now about metaphysical species and their members, we want to rigorise
this informal idea: (members of)47 species S1 is (are) superior to (members
of) species S2 just in case S1 can do what S2 can do and more. (Recall
principle 210 earlier: “In material and living bodies we fnd an ascending
order of perfections in which the higher beings have their own perfections
as well as those of the lower level of being.”) Further, we need to expli-
cate ‘can do’ in terms of essential powers, as implied earlier, otherwise
we would not be talking about species qua species. (One species of bird
might be contingently superior to another species of bird when it comes
to nest building – imagine that the nest-building superiority is not due to
the birds’ intrinsic natures but rather to the vagaries of available environ-
mental nest material – but this is not the sort of superiority we should
have in our sights.)
Again, as suggested earlier, the powers we are concerned with must
be highly generic, not merely essential. There might be, say, two highly
similar species of fungus one of which is more toxic to grapes than the
other is. This does not make for the metaphysical superiority of one fun-
gus species over the other even if the diference is essential. Nor is there
such superiority from the mere fact of having an extra, highly specifc
power – say, one fungus being toxic to grapes and apples and the other
only to grapes. One problem is that such precise and minimal diferences
between species are rare – perhaps found only in cryptic species that are
easily confused as conspecifc – which means there will be no complete
ordering available among the species, as no doubt Bonnet will have con-
cluded at some point in his career.48 A related problem is that if S1 has a
power making it putatively superior to S2 and S2 has a diferent power
making it putatively superior to S1, neither will have absolute superior-
ity over the other. It is easy to suppose this to be the case throughout
nature if we focus on powers that are less than highly generic. How,
though, do we know whether a complete ordering in terms of absolute
superiority is even possible? That depends on whether there is a suit-
ably generic level at which absolute superiority can be identifed. The
Aristotelian-Thomistic hierarchy sketched previously shows just what
that level is. In other words, the right generic level and the possibility of
a ranking in terms of absolute superiority come as a package: if we can
identify one, we can identify the other. Our hierarchy, then, will be at
the level of the most generic powers of the essences of the most generic
metaphysical species of things consistent with a complete ranking by
absolute superiority.
Further, the defnition of superiority in terms of which the hierarchy
is explained will need to account for certain data, or better certain non-
negotiable ideas likely to be had by its adherents. (1) Plants are supe-
rior to mere bodies, that is to say inorganic bodies. Plants are not ‘just
another’ kind of body. To some reductionists, of course, plants and most
if not all other living things are just more kinds of body, with no features
incapable of being explained in wholly physico-chemical terms. I doubt
that any supporter of the traditional hierarchy is a reductionist of this
sort, nor is this chapter addressed to such reductionists. (2) A second
idea that should be explicable in terms of the defnition of superiority
is that angels – immaterial fnite minds – are superior to human beings.
They are not just another kind of human (which would be obvious to
any Scholastic philosopher), nor could humans ever become angels and
vice versa. Needless to say, it had better not turn out that humans are
superior to angels.49 (3) A datum you do not need to be a Scholastic
philosopher to accept is that bats are not superior to mammals. Bats are
just another kind of mammal, so to say that they are superior to mam-
mals cannot possibly be correct.50 (4) I do not think we should say that
rocks are superior to electrons or vice versa. Our defnition of superior-
ity should not entail otherwise. (5) Nor should it entail that, say, bats
are superior to dolphins or vice versa, these both being no more than
kinds of mammal, even though bats have specifc powers that dolphins
do not have, and conversely.
Here, then, is a rigorous formulation of the idea of metaphysical
superiority:
Species S1 is superior to species S2 =def:

(I)
(I.i) The set of S2’s generic powers is a proper subset P of
S1’s generic powers; and
(I.ii) S1’s generic powers minus P are not wholly explained
by any of the powers in P.
OR
(II)
(II.i) A proper subset of S2’s generic powers is a proper
subset P of S1’s generic powers; and
(II.ii) S1’s generic powers minus P are not wholly explained
by or identical to any of S2’s generic powers or any
powers of some species S3 to which S1 is superior accord-
ing to (1).
The main idea behind this defnition is to capture the earlier-mentioned

intuition that absolute metaphysical superiority involves being able to do
what something else can do and more. Correlative with this, however, is
the need not only to rule out intuitive cases of non-superiority but also
to account for the diferent kinds of relationship we see on the classical
hierarchy. In particular, we need to show how superiority holds between
the rational, sentient, vegetative, and inorganic on the one hand, as well
as between the incorporeal (God and angels) and all the rest on the other.
The best way of unpacking this is to go through some examples; the
reader can then work through others for himself.
Let us begin with clause (I): how does the hierarchy measure up? Three
examples show how it works. (1) The powers of a body are a proper
subset of the powers of a plant. A plant is not a mere body but it is a
species of body nonetheless, subject to the laws of physics and chemistry.
Since it is not a mere body, it has further powers (those proper to living
things) beyond those of a mere body – the vegetative powers51 – so the
set of generic powers belonging to inorganic bodies is a proper subset of
those belonging to plants (I.i). Next, on my non-reductionist assumption
about biology the vegetative powers are not wholly explained by any
of the inorganic bodily (corporeal) powers. That is precisely the non-
reductionist point: not that reductionists mistakenly think plants are
inorganic, of course, but that what is nominally organic about plants is
(for the reductionist) nothing but a repertoire of inorganic powers, wholly
explicable by the laws of physics and chemistry. This is what the non-
reductionist denies, and this denial is built into the classical hierarchy. To
be sure, vegetative powers are partly explained by bodily ones – physics
and chemistry (even inorganic chemistry) enter into vegetative processes.
But no non-reductionist would say otherwise. As far as the hierarchy
goes, however, clause (I.ii) is still satisfed: the vegetative powers are not
wholly explained by any of the mere corporeal powers; there is more to
the essence of the vegetative than the merely bodily.
By contrast, bodies are not superior to plants on the defnition. (I) is
not satisfed since the set of vegetative powers is not a proper subset of the
set of corporeal powers. By (II), we can fnd a proper subset P of plants’
powers that is a proper subset of the corporeal powers – being space-occu-
pying, for example; so (II.i) is satisfed. Nevertheless, the corporeal powers
minus P – being time-occupying, having shape, and so on – are identical
to generic vegetative powers, since plants too are time-occupying, have
shape, and so on. Hence (II.i) is not satisfed.
(2) Applying the same reasoning to animals and plants, we see that
the set of vegetative powers is a proper subset of the generic powers
of animals, since the latter also have sentience (consciousness of some
kind, no matter how rudimentary).52 However, those sentient powers are
not wholly explained by the vegetative powers of animals – even though
all animals have vegetative powers as part of their essence and those
vegetative powers enter essentially into the distinctively animal powers
(no perception without nutrition, for example). The non-reductionist
assumption, again, is that these higher sentient powers cannot be wholly
explained in terms of any of the powers possessed by mere vegetative
beings, that is, by plants. Again, (I) is satisfed.
By contrast, plants do not turn out to be superior to animals: (I) does
not apply since the set of generic animal powers is not a subset of the
generic powers of plants. Still, might plants be superior to animals accord-
ing to the second disjunct (II) of the defnition? Various proper subsets
of generic animal powers are indeed proper subsets of generic vegetative
powers: the nutritive powers, for instance, or the reproductive powers,
or the union of both sets; so (II.i) is satisfed. Whichever P we choose,
however, it turns out that the vegetative powers minus P are either wholly
explained by or identical to some of the generic animal powers. In this
case, they are all identical to generic animal powers, precisely because
animals also have the generic vegetative powers. So (II.ii) is not satisfed.
Hence plants are not superior to animals according to (II) either.
One might object that since photosynthesis, for example, is a vegetative
power but not an animal power, there must be some vegetative power not
belonging to P, for any P we choose, that is not explained by or identical to
any animal power. So why aren’t plants superior to animals after all? The
fact that some animals do photosynthesise (some kinds of bacteria, for
example) is irrelevant. What matters is that photosynthesis is not a generic
vegetative power. Rather, it is a kind of nutritive power, and it is the
generic nutritive power of plants that will either be part of P or identical to
an animal power that is not part of P (if we chose, say, only reproductive
powers, or reproductive and locomotive powers, as the members of P).
Photosynthesis does not, therefore, make plants metaphysically superior
to animals.
(3) The set of generic animal powers is a proper subset of human generic
powers: humans are a species of animal. So (I.i) is satisfed. However,
humans also have rational powers53 that cannot be wholly explained in
terms of the sentient powers. So (I.ii) is satisfed. Hence, by (I), humans are
superior to mere animals although they are a species of animal: the human
powers minus the animal powers are not wholly explained by any of the
animal powers. Once again, this is a non-reductionist presupposition for
the purpose of defning the metaphysical hierarchy.
Animals, by contrast – mere or non-rational animals – are not superior
to humans. (I) is not satisfed since the set of human powers is not a subset
of the set of animal powers. By (II.i), a proper subset of human powers is
a proper subset P of animal powers: one could, for example, choose the
reproductive powers, or sentient powers, or the union of both. Whichever
P one chooses, however, the remaining animal powers will be identical to
human powers – again, taken at the most generic level. So if sentience is
the only power excluded from P, then since it still exists in both animals
and humans, animals will not come out superior to humans according to
(II.ii). The same for any other animal power excluded from P.
What we should not want is a defnition yielding the result that bats are
superior to mammals. Intuitively, bats are just one among the mammals;
indeed every mammal, however fascinating in itself, is just one among the
mammals, except when considering humans and their special rational and
intellectual powers. We should not say that at the generic metaphysical
level, where the joints of nature are saliently carved, we will fnd bats
standing to all the other mammals as humans to all the other mammals (or
to all the other animals, for that matter). Again, whereas bats, intuitively,
are just another kind of mammal, we should not – as non-reductionists –
think of plants as just another kind of body.
Fortunately, (I) does preserve this distinction. In order to show this,
though, we do need to descend to more specifc powers, because now we
are considering bats as against other mammals; so we need to consider
the generic powers of mammals qua mammals, along with the powers had
by specifc kinds of mammal. We cannot avoid getting into the biologi-
cal weeds; nor should any Scholastic metaphysician be afraid to do so.
The generic bat powers minus the generic mammalian powers are, quite
simply, the power of fight and all the powers that depend on it (perhaps
certain powers of limb movement, for example). But the power of fight
in bats presupposes and is wholly explained by the generic powers of
mammals.54 Note that the full account of bat wing homology appeals to
more than just mammals – for instance tetrapod reptiles and the fore-
limb structure of birds – but this is due to the descent of mammals, on
the standard account, from non-mammalian ancestors. The complete
explanation of bat fight in terms of bats being mammals includes, then,
whatever it is about mammals that explains their generic limb structure.
The most specifc part of the explanation will be in terms of mammalian
limb structure and the adaptations undergone by bats so as to result in the
power of fight. The more general part of the explanation will be in terms
of whatever adaptations mammals themselves, as a metaphysical species,
have in common with certain non-mammals. The whole explanation of
bat fight, then, will still be in terms of the fact that bats are mammals. So
bats do not turn out, on (I), to be superior to mammals. What about (II)?
We can easily fnd a proper subset P of mammalian powers that is a proper
subset of bat powers – say the powers that depend on having hair, or the
power of lactation in the female, or the union of these – but it will remain
the case that the bat powers minus P are identical to or (in the case of
fight, as per earlier) wholly explained by the generic mammalian powers.
Note that my use of the phrase ‘wholly explain’ is not meant to exclude
the explanatory role of selection pressures in the emergence of a charac-
teristic, which applies across the board. Rather, I am focusing on spe-
cies relationships only: qua variation on a type, the bat wing structure is
wholly explained by its being a mammal, not by any species relationship
with, say, fying insects – where the similarity is a matter, it is thought, of
convergent evolution.
We certainly do not want mammals to turn out superior to bats (not
merely counterintuitive but absurd). On (I) this is impossible since the set
of generic bat powers (which, of course, includes fight) is not a subset of
the mammalian powers. On (II.i), a proper subset of bat powers – pick
for instance the set containing lactation and the powers that depend on
having hair – is a proper subset P of the mammalian powers. (II.ii) fails,
however, because the mammalian powers minus P are wholly explained
by or identical with the generic bat powers. For example, the power to
send oxygen-rich blood around the body is identical in all mammals.
Hence, and fortunately, the defnition does not entail that mammals are
superior to bats.
Now we want to capture the slightly more complicated relationship
between disembodied substances – God and the angels – as well as
between both of these and the embodied substances. For God is not a
species of angel and vice versa, and neither are species of anything below
them on the hierarchy, nor is anything below them a species of God or
angel; in particular, angels are not a species of human nor are humans
a species of angel. It might be replied that not everything is a species of
something below it in the hierarchy as diagrammed earlier. Humans are a
species of animal but animals are not a species of plant. Nor are humans a
species of plant even though plants are a species of body.55 And nothing is
a species of prime matter, even though prime matter is a component of all
material substances. So why should God and the angels pose a problem?
For one thing, God is not in a species or genus, is not even a singleton
species, and has no overlap with bodies in terms of powers. As utterly
distinct from His creation, yet with humans made in His image, how can
we specify His superiority in a way that comes out clear and plausible?
Angels, too, are not bodies and have no bodily powers, yet they share
important powers with humans whilst being quite distinct species from
humans (each angel being its own singleton species). Humans have pow-
ers distinct from angels and vice versa – yet the classical hierarchy has
angels above humans, not the reverse.
How, given all of this, are we to account for the superiority relations?
Again, the intuition we want to capture is that where two beings or kinds
of being share certain powers yet are in a superiority relation, we should
have (i) an explanation of why the unshared powers are unshared and,
more importantly, (ii) an explanation of the superiority generated by some
of those unshared powers such that asymmetry is preserved.
Before looking further at angels and humans, consider the more quotid-
ian example of bats and dolphins. Bats and dolphins are species of mam-
mal. We have no intuition that one is metaphysically superior to the other.
Still, each has powers the other doesn’t have, so maybe we can call that
a kind of limited or local superiority across a particular dimension (e.g.
locomotion, whether by air or sea), but we do not want to say that bats
are metaphysically superior to dolphins or vice versa. They are just two
kinds of mammal. Take bats to be our S1 and dolphins our S2. Clause (I)
of our defnition clearly does not apply since the generic powers of neither
S1 nor S2 are a proper subset of the other’s. We move to clause (II). By
(II.i), there is a proper subset of dolphin powers that is a proper subset P
of bat powers – these being the shared mammalian powers (lactation, res-
piration via lungs, etc.). The unshared bat powers are those of fight (and
the unshared dolphin powers are those of swimming). Looking at clause
(II.ii), the bat powers minus P, in other words the power of fight (and the
powers that depend on it), are explained by the generic mammalian pow-
ers of dolphins. Note: I am not asserting that the power of fight in bats is
explained by the fact that dolphins have certain powers. That would be
an absurd claim. Rather, I am asserting that it is explained by the powers
that dolphins possess – certain generic mammalian powers – that are also
possessed by bats, therefore not lifting bats as a species above dolphins in
the metaphysical hierarchy.
Since, as explained earlier, the whole explanation of bat fight is in terms
of the fact that bats are mammals, their lack of superiority to dolphins
consists in the fact that bat fight is wholly explained by mammalian pow-
ers possessed, inter alia, by dolphins. Again, to put it more casually, the
special bat powers do not place bats above dolphins in the metaphysical
hierarchy because those powers are adequately explained by something
that dolphins are, namely mammals. In this case, bats too are mammals,
but it is the comparison to dolphins – with the focus on what dolphins
are – that explains the lack of superiority to them in particular on the
part of bats.
It might be objected that I am cherry picking favourable examples, so
let us consider a more challenging case in order to show the robustness
of the defnition and to highlight the underlying metaphysical principles.
Consider bats and insects. Our defnition should not yield the result that
bats are metaphysically superior to insects or vice versa, no matter how
much such a thought would have appealed to many eighteenth- and even
nineteenth-century naturalists. Neither bats nor insects are a metaphysi-
cal species of the other, let alone biological. More generally, chordates
(to which metaphysical species/biological phylum bats belong) are not
a metaphysical species of arthropods (to which metaphysical species/
biological phylum insects belong) and conversely.
We should be able to see immediately that clause (I) is not satisfed.
What about clause (II)? By (II.i), a proper subset of insects’ generic powers
is a proper subset P of bats’ generic powers (I will run the argument in this
direction but I leave the other direction to the reader). These generic pow-
ers include the generic powers of all organisms – reproduction, homeosta-
sis, and so on – plus those of the animals – sentience and locomotion – and
also those powers which allow chordates and arthropods to be within the
same tree of life, part of a unifed taxonomy involving specifc morpholo-
gies in the broadest sense, such as body plans and organ types. In evolu-
tionary terms, this will be governed by the putative last common ancestor
of the two phyla – perhaps an annelid worm-like animal, as proposed in
1875 by Anton Dohrn.56
Speaking in more abstract metaphysical terms, we can focus for exam-
ple on segmentation as common to both phyla: arthropods and chordates
all have segmented body plans giving rise to certain bodily powers unique
to segmented organisms, such as certain kinds of independent movement,
organ regeneration, and bodily growth. But the phyla diverge signifcantly,
of course – as anyone who has seen a bat and a fy will testify. One key
divergent feature is dorso-ventral inversion – roughly speaking, opposite
orientations of the nerve cord, heart, and blood fow. The chordate ori-
entation – to which all the distinctively mammalian powers ultimately
owe their existence via diference in body plan between chordates and
arthropods – is not part of the P shared by chordates and arthropods. But
it is explained wholly by the powers due to the species relation chordates
bear to arthropods, namely being common species of segmented organ-
ism (whatever the specifc details of any last common ancestor or descent
with modifcation). In other words, the inverted orientation of chordates
is explained by the powers they owe to being a kind of segmented organ-
ism, which arthropods are as well. The inversion is possible because of
the segmentation, whatever the historical details. Hence the defnition is
satisfed and, since chordates are not metaphysically superior to arthro-
pods, neither are bats superior to insects – or vice versa.
It is, of course, impossible to examine every possible challenge to the
defnition derived from the incredible diversity found in the tree of life.
It is difcult to rule out all possible counterexamples. Perhaps we will
have to indulge in some chisholming57 of the defnition in order to arrive
at the bulletproof status, not (too) divergent from intuition, that meta-
physicians should be bold enough to seek. Or perhaps the chisholming
will reveal unfxable faws in the attempt to give metaphysical superiority
a precise criterion. I submit, however, that such cases as the ones I have
just discussed reveal more general principles that should help us when
confronting alleged counterexamples.
One is that biological non-reductionism should be handled with care.
It is one thing to assume, plausibly, that sentience cannot be wholly
explained in terms of non-sentient features of organisms: awareness, no
matter how primitive, looks on its face to be of a diferent metaphysi-
cal order from even the most complex entities and processes that are
wholly without awareness. The same goes for trying to explain rational-
ity in terms of awareness, no matter how complex the latter may be: the
abstractness of rational thought again seems to be of a diferent order from
consciousness on its own. By contrast, taking a non-reductionist approach
to, say, mammalian echolocation, or insect fight, seems unmotivated and
somewhat bizarre. We should therefore expect that when comparing even
quite disparate species, such as bats and insects, or whales and worms, or
horses and dogs, or even bacteria and fungi, we should be able to trace a
path – whether historical (phylogenetic) or conceptual (morphology and
engineering, including genetic processes) – whereby what is distinctive of
one species is ultimately explained in terms of what it has in common with
the comparator species.
Secondly, and related to this thought, if the tree of life is really a single
tree – more abstractly, if biological taxonomy is unifed – then we should
not expect that global unity to be swamped by local disunities. In other
words, if there is a metaphysical hierarchy, with the inevitable breaks or
leaps in the chain, this should be understood as limited in scope compared
to the unity and gradualness found in between the breaks. Otherwise
the very enterprise of biological explanation will look hopeless. So we
should look for, and expect to fnd, plausible explanations of species char-
acteristics in terms of commonalities with other species, just as biology
postulates – and did postulate both before and after Darwin. Hence the
appeal to homologies, common selection pressures, shared genes, com-
mon morphological constraints, and so on. Our defnition of metaphysical
superiority should, as Quine might have put it, minimally mutilate all of
this explanatory work.
We are fnally in a position to consider the much thornier relation of
angels to humans. It should be clear that (I) is not satisfed since the
powers of a human are not a proper subset of the powers of an angel:
humans have rationality and they have bodily powers, but these do not
belong to immaterial minds. (Recall: angels do not reason, they receive
knowledge.) We need, then, to look at part (II) of the defnition. On (II.i),
there is a proper subset of the powers of a human that is a proper subset
of the powers of an angel, namely intellect and will. Now consider clause
(II.ii). The angelic powers minus P are the incorporeality of angels and the
special, essentially non-discursive nature of their intellects.58 But these are
not explained by or identical to any generic human powers. So the frst
half of (II.ii) is satisfed. What about the second half? Again, the angelic
powers minus P – incorporeality and non-discursivity – are not explained
by or identical to any species below angels in the hierarchy according to
(I). We know this because we know there are no species below angels
whose powers are a subset of the angelic powers: that all such species are
bodies is enough to prove that. Further, suppose there were a hypothetical
non-angelic species above humans, such that angels failed to be superior
to it according to (I). Such a species would still, ex hypothesi, be superior
to humans, which does not entail that angels were thereby inferior to
humans. For the purposes of the second half of (II.ii), what is relevant
would be a case where a species inferior to humans according to (I)59
somehow wholly explained the angelic powers minus the powers angels
shared with humans. But there is no such species. Moreover, suppose
there were a species intermediate between angels and humans, such that:
(i) angels were superior to it according to (I); (ii) but the angelic pow-
ers minus the powers angels shared with humans were wholly explained
by the powers of that hypothetical species. Assuming one could even
make sense of such a species standing in those hierarchical relations, by
the transitivity of superiority, given that this intermediate species was ex
hypothesi superior to humans, angels would still be superior to humans
according to (II). I should add that there are independent reasons for
doubting the very possibility of a species intermediate between humans
and angels, making the previous speculations somewhat tangential.60
To elaborate a little, on the classic understanding of angels each angel
belongs to its own singleton species.61 No angelic species is superior to
itself, so that cannot explain the angelic powers they do not share with
humans. Even if they all belong to a genus – the angelic genus, of which
each individual angel is a singleton species – that will not explain the
powers they do not share with humans as opposed to merely restating the
fact of their being angels. The truth is that what explains those unshared
powers is the direct creation of angels in their entirety by God, who saw
ft to produce incorporeal intellectual substances.
To speak casually again, nothing about this special creation brings
angels down to the same metaphysical level as humans, in the way that
bats and dolphins are just kinds of mammals on the same level. This
is so even though humans themselves have specially created immaterial
souls. For it is having such souls that raises them above everything else in
creation apart from the incorporeal angelic substances.
Our defnition also clarifes the superiority of angels over the spe-
cies below humans. We have seen that angels are not superior to them
according to (I), so they must be according to (II). I submit we take a
very simple explanatory route here: angels share no generic powers with
animals, plants, or inorganic bodies.62 Consider any of them: since the
null set is standardly taken to be a proper subset of every non-empty set,
(II.i) is trivially satisfed. (II.ii) is not trivially satisfed, but it is evidently
satisfed nonetheless – as brief refection on what has already been said
should demonstrate.
A worry about asymmetry immediately arises, however. For why don’t
humans turn out to be superior to angels on this defnition, contrary to
the putative facts of the hierarchy? This result does not follow, due to the
diference between the intellects of humans and angels. A proper subset
P of angelic powers is a proper subset of human powers – once again,
intellect and will. Further, human generic powers minus P – the animal
powers along with the specifcally discursive feature of human rationality
– are not wholly explained by or identical to any of the angelic powers.
They are, however, wholly explained by or identical to the powers of some
species to which humans are superior according to (I), namely animals.
Tautologically, the human animal powers themselves are identical, generi-
cally, to the powers that all animals share as sentient (aware, not neces-
sarily feeling) organic bodies. Nevertheless, it might be objected, human
rationality is not identical to or wholly explained by any of these powers
(otherwise humans would not be superior to animals) since humans have
immaterial intellects (like angels) – so how can being an animal explain
this at all?
The answer is that we must be careful about what ‘this’ refers to. We
should distinguish between the intellect itself – the capacity for under-
standing by means of abstraction and the possession of concepts – and
its specifcally discursive nature in humans, which requires the combi-
nation and division of the concepts delivered by the intellect and their
formation into judgments and inferences. In a broad sense, the intellect
encompasses all of these activities – precisely because, when we think of
the intellect, we think of the human intellect (and possibly the intellect
in other animals, if they have it). More narrowly, however, the intellect
is a power of understanding not essentially connected to any discursive
activity. That is why we can truly say that both God and angels have
intellects. The reason that humans have discursive intellects – what we call
rationality, or the power to reason – is that, unlike God and the angels,
humans are conscious, living bodies. Having a body essentially limits
what a human can know at any one time, or over time, both qualitatively
and quantitatively. It entails having sensory organs with dedicated and
fnite channels of information of specifc kinds. There can be no instanta-
neous connections between concepts and judgments, as is possible for the
immaterial angelic mind. Humans have to interpret the information they
get, combine information from diferent channels or sources (including
any innate knowledge), and arrive at further judgments. It is corporeality
that makes all of this necessary, so although the human intellect as such is
not wholly explained by having a body,63 it is the discursive nature of that
intellect – the power of ratiocination possessed by humans – that is wholly
explained by humans being bodily creatures. And this is compatible with
humans – like angels – having immaterial intellects. For this reason, (II.
ii) of the defnition is not satisfed and humans are not superior to angels.
We can also explain, on my defnition, why God is superior to a rock.
We know that clause (I) does not apply because the powers of a rock are
not a subset of the divine powers. We turn, then, to (II): what powers of
a rock, one might ask in horror, are a proper subset of God’s powers?
The answer, of course, is that there are no such powers. But this is to
say that P is the null set, and the null set is a proper subset of every non-
empty set, including the set of divine powers. But the set of divine powers
minus P, which just is the set of divine powers, is in no way explained
by or identical to any of the generic powers of a rock or any powers of a
species to which God is superior by (I). So, on (II), God is quite evidently
superior to a rock. To show that a rock is not superior to God, we note
again that the null set is a proper subset of the divine powers that is also
a proper subset of the rock’s powers, satisfying (II.i). But by (II.ii), the
rock’s powers minus the null set – that is, the rock’s powers – are wholly
explained by the divine powers: after all, God created rocks so how could
it be otherwise? Note that all we are talking about are the generic bodily
powers, which on the classical conception are a direct part of the divine
creation. On this conception, rocks are no more superior to water or to
electrons than bats are to dolphins. Whatever the explanatory relations
between, say, bat fight and mammalian morphology, or between rocks
and electrons (rocks are made up of electrons among other particles), it
remains that qua bodies, or qua animals, the classical conception sees
these archetypes, so to speak, as objects of direct divine creation.64
What about God’s superiority to man and angels? As to the former,
again we must consider part (II). Here, we should say that P consists of the
intellect and will, since God also has an intellect and will.65 However, we
see when applying (II.ii) that none of the divine powers minus P – pick any
of the ‘omni’ powers you like66 – are wholly explained by or identical to
any of the human powers or those of any species to which God is superior
according to (I). Again, with God and angels we see that P consists of will
and intellect but also the incorporeal powers. Yet none of the divine pow-
ers minus P are explained by or identical to any of the angelic powers or
any powers of a species to which God is superior according to (I).
As for asymmetry, are humans superior to God? Again we refer to (II).
A proper subset P of God’s powers is a proper subset of human powers,
namely intellect and will (II.i).67 The human powers minus P – the rational
and animal powers and (roughly) the powers they entail68 – are not identi-
cal to any of the divine powers. Nevertheless, any human who is an entire
special object of divine creation will – like the angels – have all of his
powers wholly explained by the divine powers in a direct way.69 Those –
nearly all of us – who are not entire special objects of divine creation70
will have our powers minus P indirectly explained by the divine creative
powers or we can think of them as identical to or wholly explained by
the powers of animals, to which humans are superior according to (I) – as
explained earlier. However we interpret the explanation – and the versions
just given are not, I submit, competing explanations – humans do not turn
out superior to God.
Perhaps angels turn out to be superior to God on the defnition? Fortu-
nately, this is not so either. Again, on (II.i) the proper subset P consists of
intellect, will, and the incorporeal powers. On (II.ii), the angelic powers
minus P – we can include the receptivity of the angelic intellect to divine
communication71 – are wholly explained by the divine powers, since each
individual angel is directly created by God. Each angel is given all of its
powers immediately by God as a unique creation – a singleton species –
rather than as a member of a created species containing multiple instances.
Rocks, by contrast, have their generic bodily powers wholly explained
by divine creation of the bodily species containing multiple instances.
In between, humans have some of their powers wholly explained by the
species to which they belong (animal, body) and others wholly explained
by direct divine creation (intellect and will as parts of the created human
soul). It is this intermediate status that ensures that humans are superior
to animals and bodies (also plants, given the vegetative powers), are not
superior to angels, and are not superior to God.
For anyone suspicious that my defnition of metaphysical superiority
threatens to confate the creator and the creature, enough has been said,
I submit, to dispel that understandable concern.
6. Conclusion
Although there is still plenty that could and should be said about the
hierarchy of being in a modern context, I hope I have done enough to
show that the classic Aristotelian-Thomistic hierarchy can, with suitable
assumptions taken on and metaphysical prejudices put aside, be defended
as coherent in its own right and largely in accord with common sense,
given the many assumptions. It may well be that further sub-hierarchies
exist and can be embedded within the general hierarchy; many, however,
that were recognised until the nineteenth century are rightly rejected for
the various reasons given earlier.
Let me emphasise a few points before bringing this discussion to an end.
First, nothing presented here is meant to be incompatible with thinking
of the hierarchy more loosely, and with some neo-Platonic colouring, in
terms of concepts such as Godlikeness, an increase in power or nobil-
ity as one ascends the hierarchy, or the retreat from matter. Second, the
hierarchy I present does not exclude local cyclicality – the food chain, for
example.72 Third, the hierarchy does not exclude local reverse superiority:
the notorious idea that cockroaches would, unlike us, survive a nuclear
holocaust, or the thought that tortoises and redwoods put our lifespans
to shame, only highlight the fact that the metaphysical hierarchy is not
about one-dimensional phenomena such as survival or longevity. Con-
nected to this is the more general fourth point that the hierarchy is not
about a being’s mere fourishing in its niche, how it manages as against
its competitors for resources, and the like.
Finally, I repeat that nothing I have presented here constitutes an argu-
ment for the existence of God – or of angels or anything else, for that mat-
ter. The hierarchy can, nevertheless, be viewed as what might be called a
‘suggestive schema’. Consider the analogy of coming across a large jigsaw
puzzle consisting of thousands or tens of thousands of pieces – and yet
missing one single piece. The entire picture, in other words, is complete –
but for that single, missing piece. You know what the piece should look
like; you can even reconstruct it. You just do not know whether it exists
anywhere. Still, it would be reasonable to have confdence that it must be
somewhere – perhaps mislaid or forgotten, or even deemed unnecessary
given the virtual completeness of the whole. Yet the belief that it must
exist would hardly be the height of irrationality.73
Notes
1. From now on I will omit the qualifer ‘Western’, on the implicit understand-
ing that I am confning my discussion to this domain. The idea of a chain
or hierarchy of being is not wholly absent from non-Western traditions of
thought, though it is far less prominent; see Khomiakov (2005): 989–990 for
the briefest of summaries.
2. Lovejoy (1936).
3. See, for example, Bynum (1975); Mahoney (1987); Wilson (1980), (1987).
4. Lovejoy (1936): 60, quoting Alexander Pope, Essay on Man (1733): First
Epistle, lines 237–241, Pattison (ed.) (1881): 35.
5. Lovejoy (1936): 47.
6. Ibid.: 47, quoting T. H. Green (1836–1882) who himself, like probably the
majority of philosophers since his time, thought such a question ‘unanswer-
able’. Lovejoy notes the startling contrast between the attitude of Green and
those who have come after him with that of thinkers from Plato to the late
eighteenth century.
7. Ibid.: 47.
8. Ibid.
9. Ibid.: 73, 116, 177, inter alia.
10. Ibid.: 59.
11. Ibid.
12. Ibid.: 52.
13. Wilson (1980): 260.
14. Lovejoy (1936): 58.
15. Ibid.: 58.
16. Lovejoy (1936): chs. IX and X.
17. See Lovejoy (1936) for accounts of the views of these (minus Lamarck) and
other Evolutionists – in particular Leibniz (ch. V).
18. Lovejoy (1936): ch. X.
19. Ibid.: 325–326.
20. Wilson (1980): 256.
21. Lovejoy (1902): 462, quoted also in Wilson (1980): 257.
22. More as a general religious tendency rather than any particular denomina-
tion. For within the multitude of Protestant denominations, one could fnd
anything from the rigidly hierarchical (Lutheran, Anglican) to the virtually
communistic (Levellers, Anabaptists).
23. For some discussion, see Burtt (1932), especially pp. 84–85, 89–90, 175–176,
300–303.
24. See Peacock (1999): 66 for a modern statement of the principle.
25. Perhaps best encapsulated in the oft-quoted remark of Stephen Hawing from
a 1995 interview (whose original I have not been able to check for veracity):
‘the human race is just a chemical scum on a moderate size planet, orbiting
round a very average star in the outer suburb of one among a billion galaxies’
(quoted by, inter alia, Raymond Tallis in Philosophy Now 89 (2012), article
entitled ‘You Chemical Scum, You’).
26. Lovejoy (1936): 200.
27. Ibid.: 203–204.
28. Pope, Essay: Fourth Epistle, lines 49–51, Pattison (ed.) (1881): 60.
29. Richardson (1914/1740): 230.
30. Of many species within the phylum Annelida and in other taxa.
31. Species within the family Cynipidae.
32. Archibald (2014): 8, reproducing a fgure from Bonnet’s Traité d’Insectologie
(1745). See the page following the end (p. xxxii) of the Preface to Part 1
of the Traité at https://archive.org/details/traitdinsectolog00bonn/page/n41/
mode/2up [last accessed 23/04/20].
33. Lovejoy (1936) passim, especially chapter II.
34. See, for example, Aristotle, De Anima, 414a29–415a13, Ross (1931); Aqui-
nas, Commentary on the De Anima Book 2, Lecture 5, (1994): 90–96; Aqui-
nas, Summa Contra Gentiles III.22.7, (1956a): 86–87. For Aquinas’s account
of the hierarchy in the Sentences, the Summa Theologica and other writings,
see Mahoney (1982): 169–172.
35. A felicitous term of art going back to Locke.
36. For more on the diference between constituents of essence and powers that
fow from the essence, see Oderberg (2011).
37. As a glance at the Index demonstrates.
38. Wuellner (1956): 52, principle 210.
39. Ibid.: 54. Perhaps Fr Wuellner is being a little unfair, since it may be the neo-
Platonists and their disciples who misunderstood plenitude, whilst Lovejoy
understood them only too well.
40. Ibid.: 52, principles 216 and 217.
41. Summa Theologica I.47.1 resp.; Aquinas (1921): 257. ‘For He brought
things into being in order that His goodness might be communicated to
creatures, and be represented by them; and because His goodness could not
be adequately represented by one creature alone, He produced many and
diverse creatures, that what was wanting to one in the representation of the
divine goodness might be supplied by another. For goodness, which in God
is simple and uniform, in creatures is manifold and divided; and hence the
whole universe together participates [in] the divine goodness more perfectly,
and represents it better than any single creature whatever.’(Produxit enim res
in esse propter suam bonitatem communicandam creaturis, et per eas reprae-
sentandam. Et quia per unam creaturam sufcienter repraesentari non potest,
produxit multas creaturas et diversas, ut quod deest uni ad repraesentandam
divinam bonitatem, suppleatur ex alia, nam bonitas quae in Deo est simplic-
iter et uniformiter, in creaturis est multipliciter et divisim. Unde perfectius
participat divinam bonitatem, et repraesentat eam, totum universum, quam
alia quaecumque creatura.)
42. Summa Theologica I.25.6 resp.; Aquinas (1920): 355. ‘I answer that, The
goodness of anything is twofold; one, which is of the essence of it – thus,
for instance, to be rational pertains to the essence of man. As regards this
good, God cannot make a thing better than it is itself; although He can make
another thing better than it; even as He cannot make the number four greater
than it is; because if it were greater it would no longer be four, but another
number. . . . Another kind of goodness is that which is over and above the
essence; thus, the good of a man is to be virtuous or wise. As regards this
kind of goodness, God can make better the things He has made. Absolutely
speaking, however, God can make something else better than each thing made
by Him.’(Respondeo dicendum quod bonitas alicuius rei est duplex. Una
quidem, quae est de essentia rei; sicut esse rationale est de essentia hominis.
Et quantum ad hoc bonum, Deus non potest facere aliquam rem meliorem
quam ipsa sit, licet possit facere aliquam aliam ea meliorem. Sicut etiam
non potest facere quaternarium maiorem, quia, si esset maior, iam non esset
quaternarius, sed alius numerus. . . . Alia bonitas est, quae est extra essen-
tiam rei; sicut bonum hominis est esse virtuosum vel sapientem. Et secundum
tale bonum, potest Deus res a se factas facere meliores. Simpliciter autem
loquendo, qualibet re a se facta potest Deus facere aliam meliorem.)
43. Wuellner (1956): 52, principle 212 and 214.
44. Summa Theologica I.78.1, Aquinas (1922b): 75–79; Summa Contra Gentiles
II.95, Aquinas (1956b): 323–325; Summa Contra Gentiles II.68, Aquinas
(1956b): 203–207. (Note Aquinas indulging a little, in the third location,
in some Aristotelian-style biology, albeit this is hardly central to his overall
metaphysical view of the hierarchy.)
45. Summa Theologica I.51.1, Aquinas (1922a): 17–19.
46. Along, perhaps, with what they actively know ab initio by their very natures.
See further Aquinas, Summa Theologica I.58, Aquinas (1922a): 80–94.
47. I will omit further talk of species members, leaving it understood.
48. Not that I have checked.
49. Even though angels can assume human form, which is not the same as having
a human body. See further Summa Theologica I.51, Aquinas (1922a): 17–25.
50. In Oderberg (2007): 251 I said the following: ‘F-type capacities are superior
to G-type capacities just in case the former entail the latter but not vice
versa.’ This was a very loose way of putting it. Strictly, it cannot be cor-
rect since bat-type capacities entail mammalian capacities (because bats are
mammals) but not vice versa (not all mammals are bats); yet it is absurd to
say that bats are superior to mammals. One might think of defending the
one-way entailment view by replying that not all bat-type capacities entail
mammalian capacities (e.g. one does not have to be a mammal in order to
be able to bite). At the right level of specifcity, however, we can see that the
reply is specious: the way in which bats bite – with their specifc teeth and jaw
structure, etc. – entails having mammalian capacities. In any case, consider
that angelic capacities are superior to human capacities although the former
do not entail the latter. So the one-way entailment approach I put forward in
Real Essentialism was meant only as a loose approximation to a more rigor-
ous formulation such as that proposed here.
51. For more on the vegetative powers, see Oderberg (2020): 85–91.
52. On the sentient powers, see Oderberg (2020): 91–96.
53. On which see Oderberg (2020): 101–110.
54. For a standard explanation of bat wing homologies, see Ridley (2004): 55–57.
55. A look at the Porphyrian Tree will show why this is the case: see Oderberg
(2007): 93–105.
56. Dohrn (1875); for more recent speculation, see Gerhart (2000).
57. http://luigigobbi.com/philosophicallexicon.htm [last accessed 24/7/21].
58. Intellect itself might be common to both angels and humans, but the non-
discursive character of the angelic intellect is not; I interpret this as a sec-
ond-order power of angels. Similarly, the discursive character of the human
intellect – that is, rationality – is a second-order power of humans. It is a
feature of their specifc kind of intellect and is unshared with angels.
59. But not according to (II), which would make this half of the defnition of
superiority circular!
60. See Oderberg (2013).
61. Summa Theologica I.50.4, Aquinas (1922a): 13–14.
62. Doesn’t my diagram of the hierarchy group angels, humans, animals and
plants as living? Yes, but in the case of angels and the rest I take this to be an
analogical grouping, not a literal sharing of powers. Angels do not literally
grow, develop, self-maintain their internal processes, and so on.
63. Though it partly is, inasmuch as the human intellect is specially created to
inform matter, that is, to be the form of a body – unlike the specially created
angelic intellect.
64. Note in passing: there is no space to demonstrate that the superiority to prime
matter of all species and entities above it also conforms to the defnition. I
leave this as an exercise for the reader, with the hint that the powers of prime
matter are no more than passive receptivity to form and passive receptivity
to spatio-temporal dimensionality. (Feel free to email me your solution and I
will give my response.)
65. I do not see an incompatibility between man’s being made in the image of God
(implying a non-literal comparison between one and the other) and man’s
having some of the powers – intellectual and volitional – God has, albeit not
in the way God has them.
66. Including the second-order powers of the powers belonging to P – the wholly
unlimited character of the divine intellect and the totally unconstrained free-
dom of the divine will, for example.
67. We are, to reiterate, not made in the image of God if this much is not true.
68. More precisely: the powers they entail that do not also entail the powers in P.
69. The frst human, so most defenders of the hierarchy would believe.
70. Even though our souls are (including our intellects).
71. This is an aspect of the specifc non-discursivity of the angelic intellect, which
is passive in nature. God also has a non-discursive intellect but it is active in
nature. (I have shown this on the diagram.) So we can truly say that God and
the angels have intellects, but as between them, and between either of them
and humans, the specifc intellectual powers difer markedly.
72. I promised the rapper Zuby I would mention his remarkable observation
made on social media: ‘What if humans are actually farmed by plants? They
feed us oxygen until we get big, die and become fertiliser for them’ (tweet by
@ZubyMusic on 2 April 2020). There is more smartness in those words than
I have read in many philosophical texts. I have not verifed whether Zuby is
the primary source, but I doubt it was a professional philosopher.
73. An earlier version of this chapter was presented at the Humane Philosophy
seminar, Oxford, in February 2020. I am grateful to the participants for the
many comments received there. I would also like to thank two anonymous
readers for the editors of this volume, who made many suggestions that have
led to signifcant improvements.
Bibliography
Aquinas, St Thomas (1956a) Summa Contra Gentiles, Book 3 (Providence, Part 1)
of On the Truth of the Catholic Faith (Summa Contra Gentiles), trans. Vernon
J. Bourke (Garden City, NY: Image Books/Doubleday & Company, Inc.).
——— (1956b) Summa Contra Gentiles, Book 2 (Creation) of On the Truth of
the Catholic Faith (Summa Contra Gentiles), trans. J.F. Anderson (Garden
City, NY: Image Books/Doubleday & Company, Inc.).
——— (1920) Summa Theologica, Part 1, QQ.I-XXVI, vol. 1 of The ‘Summa
Theologica’ of St Thomas Aquinas, literally translated by the Fathers of the
English Dominican Province (London: Burns Oates & Washbourne Ltd.).
——— (1921) Summa Theologica, Part 1, QQ.XXVII-XLIX, vol. 2 of The
‘Summa Theologica’ of St Thomas Aquinas, literally translated by the Fathers
of the English Dominican Province (London: Burns Oates & Washbourne
Ltd.).
——— (1922a) Summa Theologica, Part 1, QQ.L-LXXIV, vol. 3 of The ‘Summa
English Dominican Province (London: Burns Oates and Washbourne Ltd.).
——— (1922b) Summa Theologica, Part 1, QQ.LXXV-CII, vol. 4 of The ‘Summa
English Dominican Province (London: Burns, Oates and Washbourne Ltd.).
——— (1994) Commentary on Aristotle’s De Anima, trans. K. Foster O.P., and
S. Humphries O.P. (Notre Dame, IN: Dumb Ox Books). (Orig. pub. 1951.)
Archibald, David J. (2014) Aristotle’s Ladder, Darwin’s Tree: The Evolution of
Visual Metaphors for Biological Order (New York: Columbia University
Press).
Burtt, E.A. (1932) The Metaphysical Foundations of Modern Physical Science
(London: Routledge).
Bynum, W.F. (1975) ‘The Great Chain of Being after Forty Years: An Appraisal’,
History of Science 13: 1–28.
Dohrn, A. (1875) Der Ursprung der Wirbelthiere und das Princip des Function-
swechsels (Leipzig: Verlag Von Wilhelm Engelmann).
Gerhart, J. (2000) ‘Inversion of the Chordate Body Axis: Are there Alternatives?’,
Proceedings of the National Academy of Sciences 97: 4445–4448.
Khomiakov, M. (2005) ‘Hierarchy and Order’, in M.C. Horowitz (ed.) New Dic-
tionary of the History of Ideas (Detroit: Thomson Gale).
Lovejoy, A.O. (1902) ‘Religion and the Time-Process’, American Journal of The-
ology 6: 439–472.
——— (1936) The Great Chain of Being: A Study of the History of an Idea
(Cambridge, MA: Harvard University Press).
Mahoney, E.P. (1982) ‘Metaphysical Foundations of the Hierarchy of Being
According to Some Late-Medieval and Renaissance Philosophers’, in P. More-
wedge (ed.), Philosophies of Existence, Ancient and Medieval (New York:
Fordham University Press): 165–257.
——— (1987) ‘Lovejoy and the Hierarchy of Being’, Journal of the History of
Ideas 48: 211–230.
Oderberg, D.S. (2007) Real Essentialism (London: Routledge).
——— (2011) ‘Essence and Properties’, Erkenntnis 75 (2011): 85–111.
——— (2013) ‘Could There Be a Superhuman Species?’, The Southern Journal of
Philosophy 52 (2014): 206–226.
——— (2020) The Metaphysics of Good and Evil (London: Routledge).
Pattison, M. (ed.) (1881) Pope: Essay on Man (Oxford: Clarendon Press).
Peacock, J.A. (1999) Cosmological Physics (Cambridge: Cambridge University
Press).
Richardson, S. (1914/1740) Pamela (London: J. M. Dent & Sons Ltd).
Ridley, M. (2004) Evolution, Third Edition (Oxford: Blackwell, 2004).
Ross, W.D. (1931) Aristotle: De Anima (Oxford: Clarendon Press). (Vol. III of
The Works of Aristotle.)
Wilson, D.J. (1980) ‘Arthur O. Lovejoy and the Moral of The Great Chain of
Being’, Journal of the History of Ideas 41: 249–265.
——— (1987) ‘Lovejoy’s The Great Chain of Being after Fifty Years’, Journal of
the History of Ideas 48: 187–206.
Wuellner, B.S.J. (1956) Summary of Scholastic Principles (Chicago: Loyola Uni-
versity Press).
4 Evolution and the Principle
of Proportionality
Stephen Boulter
Introduction
Life on this planet has involved a series of fundamental transitions in
which entities with new capacities and operations arose from a causal
base lacking those capacities and operations. There is the origin of life
itself, involving the transition from the abiotic to the biotic, perhaps the
most difcult transition to explain. Further signifcant transitions include
the origin of replicating molecules, followed by the emergence of chromo-
somes, followed in turn by the arrival of the eukaryotes, which then gave
rise to multicellular organisms, and eventually to organisms with varying
degrees of sentience and rationality. In each case something strikingly new
made its appearance on the world stage.
As Darwin said in the closing passages of On the Origin of Species,
there is “grandeur in this view of life”, in which “from so simple a begin-
ning endless forms most beautiful and most wonderful have been, and are
being evolved.”1 It is hard not to agree. Nonetheless, it is this very gran-
deur that gives some pause, for the transitions in the history of life raise a
metaphysical puzzle. It is a widely held intuition amongst metaphysicians –
ancient, medieval and modern – that causes must be commensurate with
their efects, which means, at least as a frst approximation, that a cause
must “have all that is required to produce” its efect.2 Much depends on
precisely what the cause must “have”, and in what sense the cause must
have what is needed in order to be commensurate with its efect. But at
least at frst blush we have a puzzle, for the fundamental transitions in the
history of life appear to involve causes conferring on efects signifcantly
new features that they themselves do not have in any straightforward or
obvious sense. How is this possible? Such events would seem to fy in the
face of the scholastic principle of proportionality, viz., the axiom that a
cause cannot confer on an efect that which it does not itself have, as this
would be tantamount to creation ex nihilo.
The sense that there is something metaphysically curious about the his-
tory of life is only heightened when scientists themselves confess that they
have yet to provide compelling accounts of these fundamental transitions.
DOI: 10.4324/9781003125860-6
126 Stephen Boulter
Indeed, it is routinely admitted that the circumstances necessary to gener-
ate a new kind of entity are invariably extremely improbable, particularly
so in the case of the origins of life.3 These admissions confrm the suspi-
cion in some quarters that the difculty in accounting for these transitions
naturalistically is due not to the fact that they involve highly improbable
confuences of circumstances but that they are metaphysically impossible.4
If so, then it is an understatement to call the history of life endlessly
“grand”, “beautiful” and “wonderful”. “Miraculous” would be more
appropriate.
Now if the transitions in the history of life were indeed metaphysically
impossible, then the following dilemma would present itself: One would
be forced either to reject the textbook account of the history of life as
provided by biology or to assert that each fundamental transition in that
history involved a “gift from above”, that is, an intervention in which the
First Cause raised a secondary cause above its natural capacity. The earth
did indeed bring forth grass, and the waters did indeed bring forth moving
creatures, but they did so only with divine assistance.5 This horn of the
dilemma leads to the surprising conclusion that acceptance of evolution-
ary theory, combined with acceptance of the principle of proportionality,
forces a commitment to a form of theism.6
The primary burden of this chapter is to resist both horns of this
dilemma. The frst horn is to be rejected on the methodological grounds
that one’s metaphysics ought to be cut to ft the best scientifc theories
available, with evolutionary theory being one of our fnest. If the principle
of proportionality really is incompatible with evolutionary theory, then
so much the worse for the principle of proportionality.7 Methodological
considerations also motivate the rejection of the second horn because it
is both theologically8 and metaphysically unsound to advert to the First
Cause when secondary causes will do.9 This chapter is thus an extended
refection on the prima facie incompatibility of evolutionary theory with
the principle of proportionality. The challenge is to explain how both can
be maintained within a naturalistic metaphysical framework.
But a crucial clarifcation is needed at the outset regarding what the
principle of proportionality cannot mean. The core intuition is that every
efect has an adequate (i.e. proportional or commensurate) efcient cause.
At issue then is what constitutes “adequacy”. A very rough guide to the
thought is that a cause cannot give what it does not have. But this will
not do as it stands. In particular, the principle of proportionate causation
cannot rule out the emergence of new powers and operations arising from
a causal base lacking those powers simply in virtue of their being in some
sense “new”. It is not as though the principle of proportionality maintains
that all the properties of an efect must also be possessed by the cause.10
For example, hydrogen and oxygen combine to produce water, which
has powers and operations that hydrogen and oxygen lack, but no one
suggests that this involves a violation of the principle of proportionate
Evolution and Proportionality 127
causation. And this is just as well, for otherwise the principle of propor-
tionality would render problematic physics and chemistry, as well as all
of biology, and this would surely be enough to deprive the principle of
any plausibility.
Now this means a further notion must be brought into our discussion of
“adequacy”, namely the notion(s) of “perfection”. For what the principle
of proportionality rules out are cases of the less “perfect” causing the
more “perfect”, or cases of an efect “exceeding” its cause, or of an efect
“being more powerful” than its cause.11 The important point for present
purposes, however, is that being “new”, “novel” or “emergent” is not
equivalent to being “more perfect”, and so the initial statement of our
puzzle needs to be refned. It can be set out an aporia with the following
three propositions:
1. Evolution on this planet has involved a series of fundamental transi-

tions in which entities with a set of powers and operations arise
from a causal base lacking those powers and operations.
2. The fundamental transitions in the history of life include the emer-
gence of life itself, but also sentience and rationality, powers and
operations deemed “perfections” within the neo-scholastic meta-
physical framework.
3. The principle of proportionality: Every efect must have an adequate
natural efcient cause, that is, a cause sufciently “perfect” to have
the power to produce the efect.
The challenge then is to unravel this refned version of our aporia.12 To

this end I frst set out a list of the various transitions in evolutionary his-
tory that are putative violations of the principle of proportionality. Sec-
ond, I provide a fuller statement of the principle of proportionality itself
as found in the writings of the scholastics. Third, I set out a taxonomy
of the various kinds of “perfection” based on the distinctions between
privative and negative perfections, on the one hand, and the orders of
ontological dependence and eminence on the other. With this taxonomy
in place, I then consider whether the transitions in the history of life do
indeed violate the principle of proportionality by involving naturalisti-
cally impossible increases in perfection in any of these categories. As will
become clear, dismissing any such violation concerning privative perfec-
tions and perfections in the order of ontological dependence is (relatively)
straightforward. And while more care is needed in handling negative per-
fections in the order of eminence, it can be shown that they do not involve
any violation of the principle of proportionality either. The upshot is that
naturalism and the principle of proportionality are compatible with the
textbook account of the fundamental transitions in the history of life.
These transitions are indeed “miraculous” in the sense that they remain
extremely improbable events (given what we know), but there is no reason
128 Stephen Boulter
to believe that they required special divine intervention in the natural
order of things. Our dilemma is thus avoided.
The Evolutionary Transitions

To fx ideas let us begin with the problematic transitions in the history of
life. In The Origins of Life John Maynard Smith and Eors Szathmary set
out the series of fundamental transitions that occupy their attention qua
biologists.13 These are as follows:
Abiotic → Biotic
Replicating molecules → Populations of molecules in protocells
Independent replicators → Chromosomes
RNA as gene and enzyme → DNA genes protein enzymes
Prokaryotes → Eukaryotes
Asexual clones → Sexual populations
Single-celled organisms → Animals, plants, fungi
Solitary individuals → Colonies with non-reproductive castes
Primate societies → Human societies with language
Each of these transitions marks the emergence of something funda-

mentally new, and it is these transitions that need to be squared with
the principle of proportionality if our dilemma is to be avoided. Now
Maynard Smith and Szathmary ofer plausible guesses as to how these
transitions occurred, guesses based on our best current evidence. They
also ofer general observations regarding certain patterns they discern
in the series as a whole. As we shall see, one of these patterns proves
particularly suggestive when dealing with the challenge of squaring
the transitions with the principle of proportionality. But for now it is
enough that these transitions be clearly before us and to note, at least
in passing, that sentience and rationality do not appear explicitly in
their number.
The Principle of Proportionality

The scholastic principle of proportionality is one of a set of axioms or
“intuitions” regarding causation in general. In particular the principle
serves as a general guide to thinking about the minimal requirements an
entity must meet in order to be a candidate cause for a given efect, viz.,
that any candidate cause must be “proportional” to the efect. The con-
temporary formulation of the proportionality condition is that a cause is
commensurate or proportional with an efect if it has “all that is required
to produce the efect, and as little as possible that is not”.14 Turning to
the scholastic principle in particular, I begin with Peter Cofey’s collec-
tion of scholia on causation from his Ontology or the Theory of Being.15
Although I will be relying primarily on Anselm, Aquinas, Scotus, Ockham
and Suárez as sources of scholastic thought on this matter, Cofey nicely
codifes the primary intuitions and conclusions regarding causation in
general and the principle of proportionality in particular that animate
scholastic thought generally. These are:
1. The Principle of Causality
This gathers a number of intuitions, including that whatever happens has

a cause; whatever begins to be has a cause; whatever is contingent has a
cause; nothing occurs without a cause.16
2. Operari sequitur esse
Any act of being necessarily involves operational powers because part of

what it means to be is to be at least a potential cause. Moreover, those
powers depend on what kind of entity it is.
3. Omnes agens agit inquantum est in actu
All agents act insofar as they are in act. The higher an agent is on the scale
of being, the “fuller” its act of being, the higher its operations and efects.
4. From a known efect, of whatsoever kind, we can argue with certainty

to the existence of an adequate efcient cause and to some knowledge
of the nature of such a cause.
This epistemological principle is grounded in
5. The Principle of Proportionality: Every efect must have an adequate

efcient cause
An adequate efcient cause is a cause sufciently perfect, sufciently high

on the scale of being, to have the active power to produce the efect in
question (otherwise the efect would be partially uncaused – which is
impossible).
6. An efect cannot as such be more perfect than its adequate (created)

cause
This is a further specifcation of “adequate”. The idea is that it is

inconceivable that a less perfect power could bring about the more
perfect because the efect is ontologically dependent upon the cause.
Adequacy requires that the cause be at least as perfect as the efect, if
not more so.
130 Stephen Boulter
7. Nihil agit ultra suam speciem; omne agens agit simile sibi
Nothing acts in a manner above its own kind; (as) everything produces
its like.
At this point an obvious problem arises: Bearing in mind that the prin-
ciple of proportionality does not rule out causes bringing about efects with
“new” properties per se, it does remain difcult to understand how causes
can generate efects higher up the scale of being. And the scholastics were
very aware of this difculty because they had their own extensive budget of
putative counterexamples. For example, we fnd the scholastics wondering:17
i. How living organisms can arise from putrefaction

ii. How celestial bodies can produce living things
iii. How heat produces fesh
iv. How the earth can bring forth vegetation
v. How the causes of the death of an organism can generate the form
of a cadaver
vi. How the heat of the sun can convert elements into minerals when
elements are found in the “bowels of the earth” and so shielded
from the heat of the sun
vii. How accidents can produce substances
viii. How semen can produce an organised foetus
ix. How God, whose simplicity and perfection is existence itself, can
bring forth the higher perfections of life, sentience and rationality18
That these causal relationships obtain was not in question for the scho-
lastics, merely how such relationships were possible. And the scholastics
had two ways of dealing with these prima facie counterexamples to the
principle of proportionality. The frst was the claim that
8. The actuality of the efect need not be in its adequate created cause
actually and formally, but merely potentially or virtually.
This allows for the evident fact that efects often have properties the causes
lack, whether those properties be higher on the scale of being or merely
new and emergent. As we shall see, this appeal to the virtual or potential
existence of properties in causes provides much room for manoeuvre. But
there was one further solution that was always available. (At least as a
last resort) one could claim that
9. Created causes have a passive obediential capacity (potentia obedi-

entialis) whereby their nature can be so elevated by the First Cause
that they can produce, with His special supernatural concursus,
efects of an entirely higher order than those within the ambit of
their natural powers.
Let this serve as a statement of the general principle of proportionality as
I will understand it for the remainder of this chapter. Matters will become
more precise once we have considered the notion(s) of perfection in detail
in the next section. But before doing so it is worth pausing to notice
something important about the problem cases that taxed the scholastics
themselves, namely, that cases i–viii involving causation in the natural
order no longer trouble us. At issue is why this is so. And it is clear that
these puzzles arose primarily because of their impoverished state of infor-
mation, not because there is something problematic about the principle of
proportionality per se. Moreover, their information was impoverished in
two distinct ways. First, they were puzzled primarily by putative cases of
causation in the natural order where in fact no causation was occurring
(cases i–vii). Second, they were puzzled by cases where causation was
indeed occurring, but they lacked the information necessary to make the
causal process intelligible – the role of semen in the organisation of the
foetus being a clear case in point. The frst point to recognise is that none
of these problematic cases now forces an appeal to the potentia obedi-
entialis of secondary causes and divine concursus. Second, we should be
mindful of the fact our current state of information is very likely to be
impoverished in the same two ways and so responsible for the generation
of our own problem cases.19
“Perfection” Disambiguated
If new or emergent properties are permitted by the principle of propor-
tionality – by invoking the potential or virtual existence of the efect in the
cause – what does the principle purport to rule out? We know that the less
perfect are not permitted to be the cause of the more perfect. But what,
exactly, is a “perfection”, and what is it about perfections that renders
causal relations between inferior and superior entities problematic? These
questions go to the heart of our aporia.
Unfortunately answering these questions is far from straightforward
because there are (at least) two distinct intellectual traditions that make
use of the notion of perfection – the Anselmian and the Aristotelian – and
these notions are only loosely related. Equivocation thus abounds, and
it is very easy to misread what is intended by an author’s use of the term
“perfection”.
We can trace the Anselmian notion of a perfection that is at work in
the Great Chain of Being and his version of the Ontological Argument to
Plato and Plotinus.20 The idea here is that all entities can be ordered hier-
archically starting from God, the highest link in the chain, down through
the various created substances – angels, humans, animals, plants, inani-
mate minerals – descending further into the various accidents and modes
of substances, and fnally terminating in absolute Non-Being. Using this
Great Chain of Being we can then indicate roughly that a perfection is any
132 Stephen Boulter
property at all the possession of which moves one up the Chain. Existence
is therefore a perfection, as it moves one from absolute Non-Being to
the lowest of real entities. Life is also a perfection as it marks the transi-
tion from the inanimate to the animate; sentience the transition from the
plants to the animals, rationality the transition from non-human animals
to humans, and so on.
Now if we look for a general account of what all these disparate prop-
erties have in common we fnd suggestions to the efect that perfections
are “anything that is better than that which is not-it”; or “that which is
absolutely and in an unqualifed sense better than everything incompatible
with it”; or even “a perfection is a notion that includes no imperfection”.21
Unfortunately these phrases do not add much to our understanding of
perfections, but they do at least chime with Aquinas’s assertion that the
“good, true and noble” are instances of perfections.22 Later writers are
more specifc. Wuellner provides the following list of criteria for judging
degrees of perfection. One entity might be “better” than another because
of their comparative
• degree of act vs potency

• degree of immateriality
• immanence of action
• range and depth of operation and causal infuence
• freedom from substantial change
• unity and simplicity of structure
• natural control of other things.23
Although this list provides examples of perfections, it still leaves us with

a disparate collection of properties with no obvious unifying factor.
Some order is brought to these matters by following Scotus’s lead. Sco-
tus distinguishes diferent kinds of perfections by distinguishing diferent
hierarchical orders. He ofers four divisions of prior and posterior, but
we need attend only to the frst division for present purposes. The frst
division is between the order of ontological dependence and the order of
eminence. The order of dependence orders entities according to “what
depends on what”, while the order of eminence orders entities accord-
ing to degrees of “nobility” and “dignity”.24 And this distinction seems
well drawn. For considerations of “nobility” and “dignity” seem to be of
an entirely diferent nature from those keyed to ontological dependency.
Indeed the order of eminence falls under suspicion of being an anthro-
pocentric projection entirely out of keeping with biologically informed
thinking, while the order of dependence is built into the very fabric of the
natural order as there are real ontological dependency relations found in
real things themselves.25 More on this shortly.
Now in addition to this Anselmian approach to perfection, there is also
the Aristotelian tradition which ties the notion of perfection tightly to
the notion of completion. Aristotle says that an entity is perfect when
“in respect of excellence and goodness [it] cannot be excelled in its kind”
(emphasis added).26 Here the idea is that a perfection is a property that
an entity requires in order to be whole and in the best possible state given
the kind of entity it is. A perfection, as the Latin suggests, is whatever
“completes” an entity, making it lacking in nothing which is suitable to
it. It is in this sense that the adult and the child of the same species are
related as perfect and imperfect.27 And it is this sense of “perfection”
that Aquinas is employing when discussing God’s perfection in the early
passages of the Summa Theologiae. He says, “That is perfect which lacks
nothing of the mode of its perfection.”28
Now, although both accounts of perfection contain the notion of
completion (if we take God’s esse to be the most complete as it lacks no
perfections), it remains the case that the Anselmian and the Aristotelian
notions are quite distinct. According to the Aristotelian account, a per-
fection is a property that makes an entity better than other entities of
the same kind, while on the Anselmian account a perfection is a prop-
erty that makes an entity “better” or more “noble” than another entity,
whether that entity is of the same kind or not. Anselmian perfections
are thus absolute in a way Aristotelian perfections are not. Ockham
recognises two distinct kinds of perfection on precisely these lines. 29
And Suárez will eventually label these two distinct notions “negative”
and “privative” perfections respectively.30 A privative perfection is
closely connected to Aristotle’s idea of completion according to which
an entity is perfect when it is whole and lacking nothing required by its
nature, a privation being the lack of something expected or called for
by a particular entity’s nature – sentience in an animal, for example.
By contrast, a negative perfection is not related to a standard set by a
particular nature but is in some sense absolute. In this case an entity
might lack a perfection that is not expected or required by its nature –
for example, life in a stone.
Further distinctions in kinds of perfection can be found, but these cap-
ture the fundamental ideas of both traditions.31 But the main issue we
face is that there is no univocal sense of “perfection”. Rather we have a
collection of notions that are at best analogously related to the notion of
completion. This means there is no simple answer as to whether the fun-
damental transitions in the history of life involve metaphysically problem-
atic increases in “perfection”, for the answer will turn on which notion
of perfection is at issue. It is surprising then to fnd authors neglecting to
specify which notion of perfection they are operating with. These lines
from Maurice de Wulf are a case in point, although he is far from being
the only ofender:
in general all living creatures are diferent from and superior to inor-
ganic bodies, such as a molecule of water or a loadstone, because they
134 Stephen Boulter
possess a form which is superior in perfection to any forms found in
the inorganic world.32
This is a perfectly unhelpful statement as de Wulf does not go on to specify

which notion of perfection he is employing, nor does he say why the forms
of living things are superior to those of inorganic bodies.
Nonetheless, progress can be made in these matters by the introduc-
tion of a taxonomy of the basic kinds of perfections to be found at least
implicitly in the scholastic tradition. Using the aforementioned two dis-
tinctions, we can say that any putative perfection must fall into one of
four possible kinds:
i. Privative perfections in the order of dependence

ii. Privative perfections in the order of eminence
iii. Negative perfections in the order of dependence
iv. Negative perfections in the order of eminence
With this taxonomy in place we can now proceed more securely to con-
sider, kind by kind, whether the fundamental transitions in the history
of life violate the principle of proportionality by involving metaphysi-
cally dubious increases in perfection. But we should note frst that these
perfections are not all equally metaphysically secure. Privative perfec-
tions in either order, and negative perfections in the order of ontological
dependence, present no particular metaphysical difculties. However, a
question mark hangs over negative perfections in the order of eminence.
Therefore perfections of kinds i–iii are the most pressing as far as reconcil-
ing evolutionary theory with the principle of proportionality is concerned,
as it is not clear what weight should be given to considerations regarding
perfections in kind iv.33
Privative Perfections (Cases I and II)

It is clear that the cases of privative perfections in both the order of depen-
dence and the order of eminence can be set aside. All privative perfections
are species or kind relative, and so these perfections allow one to rank
specimens of the same kind in hierarchical order but do not allow a rank-
ing of entities in diferent kinds. The incommensurability of the privative
perfections of diferent kinds of entity means there can be no question of
one kind of entity being more or less perfect than another kind of entity
in virtue of a privative perfection, and so there can be no question of a
less privatively perfect kind of entity bringing about a more privatively
perfect kind of entity, as there is no ground for the alleged comparison.
And since the relevant transitions in the history of life are between enti-
ties of fundamentally diferent kinds, we can conclude immediately that
the allegedly problematic transitions in the history of life do not involve
metaphysically dubious increases in privative perfections in either order.
And so in at least these instances we can say that evolutionary theory and
the principle of proportionality are not in confict. This is an important
result for neo-scholastics because of their deeper commitment to the Aris-
totelian rather than the Platonically inspired Anselmian tradition.
Negative Perfections in the Order of Dependence

(Case III)
The next order of business is to consider whether the transitions in the
history of life violate the principle of proportionality by involving meta-
physically problematic increases in negative perfections in the order of
ontological dependence. And the crucial point in these cases is the idea
that all causes in the order of ontological dependence can produce efects
in the same ontological category as itself or an efect in a category onto-
logically dependent upon the category of the cause. And so the principle
of proportionality rules out the following possibilities in this order:
• The non-existent cannot be the efcient cause of any being

whatsoever
• A being of reason cannot be the efcient cause of any real being
• A potential being cannot be the efcient cause of an actual being
• Compound substances cannot be the efcient cause of simple
substances
• Accidents cannot be the efcient cause of substances
• Contingent substances cannot be the efcient cause of necessary
substances
It is likely that it is with this order of dependence in mind that Aquinas

writes, “efects correspond proportionally to their causes, so that we attri-
bute actual efects to actual causes, potential efects to potential causes,
and particular efects to particular causes, universal efects to universal
causes . . . frst action belongs to the frst agent”, as each instance turns
on an ontologically signifcant category with no cause bringing about an
efect that the cause itself depends upon.34
But again it is plain that the fundamental transitions in the history of life
do not violate the principle of proportionality when we consider negative
perfections in the order of dependence, since they do not involve causes
generating efects on which they are themselves ontologically dependent.
This is clear enough from a cursory glance at the fundamental transitions
highlighted at the outset, as in no case is there any question of the causal
base being ontologically dependent on the newly generated kind of entity.
Quite the reverse, as typically the new kind of entity incorporates the
old kind of entity as a constituent part. And this general point remains
the case when attention shifts to the perfections of life, sentience and
136 Stephen Boulter
rationality. For the transitions from the abiotic to the biotic, asentience to
sentience, and arationality to rationality, while clearly highly signifcant,
are not transitions on a metaphysical par with transitions from non-being
to being; from existence qua mere being of reason to real being; from mere
potential being to actual being; or from compound to simple being. In
no case do we have accidents generating substances or contingent beings
generating necessary beings. So we can conclude that the transitions in
the history of life do not involve violations of principle of proportional-
ity where negative perfections in the order of ontological dependence
are concerned. And so the compatibility of evolutionary theory and the
principle of proportionality is secured in the third of our four possible
cases. Again this is a singularly important point, for perfections in the
order of ontological dependence are metaphysically very secure, and so
any violation of the principle of proportionality in this domain would
have been very damaging indeed.
One likely objection to this conclusion will be found amongst those who
maintain the “immateriality” and hence the separability of the human
intellect from the body. It will be claimed that a violation of the principle
of proportionality in the order of dependence has indeed occurred in the
history of life because fully material or corporeal causes have purportedly
given rise to immaterial efects when the transition from the arational
to the rational occurred, since the immaterial does not depend upon the
material.35
In my view the burden of proof lies with those who would uphold
this thesis. It is true that Aquinas thought the intellect immaterial,36 and
in this he was following Aristotle.37 But in both their cases postulating
an immaterial component to the psyche is in tension with their general
metaphysical framework.38 Moreover, Aristotle and Aquinas were forced
down this uncomfortable path by the difculty of explaining how abstract
thought is possible in a concrete material organ. Scotus realised that this
difculty was rooted in the unnoticed confation of three distinct senses
of “immaterial”: (a) non-organic; (b) non-extended; (c) abstract.39 Con-
sequently there is no forced move from the abstract content of a concept
to the immateriality or non-corporeality, and hence the separability of
the intellect. There is clearly a warrant for claiming that the intellect has
abstract concepts; but this provides no warrant for the claim that the
human intellect has knowledge with no link to an extended bodily organ,
and so must itself be immaterial.40 Moreover, the challenge of fnding an
activity of the intellect that has no dependence on a material organ has
yet to be discharged.41 And since an organic intellect actually sits better
within the frameworks of both Aristotle and Aquinas, to say nothing of
contemporary naturalistic metaphysics, it makes sense to welcome rather
than resist this conclusion. But the point for now is that the immateriality
of the intellect is not the basis of a threat to evolutionary theory and the
principle of proportionality because there is no need to maintain that the
wholly material processes of evolution have produced an ontologically
independent immaterial efect.
Negative Perfections in the Order of Eminence

Shifting to matters of “nobility” and “dignity”, might life, sentience and
rationality be absolute or negative perfections in the order of eminence? If
so, is there a violation of the principle of proportionality to be found here?
The quickest response is simply to deny that this category of perfection
has any members. This would be in keeping with the suspicion that nega-
tive perfections are anthropocentric projections or a dispensable meta-
physical hangover of Neoplatonism. For life, sentience and rationality
are the standout contenders for inclusion in this category, but it is clear
that they are in fact privative perfections. Life in a plant, sentience in a
mouse, and rationality in humans are indeed perfections, but life is not a
perfection in a stone, any more than sentience is in a plant, or rationality
in a bird. Similar remarks apply in the cases of a prokaryote having an
enclosed nucleus, a planarian or annelid worm discovering a sex drive,
or a lemur with a gift for the gab. And Scotus provides the explanation
of these claims:
Wisdom is not better in a dog, because nothing is goodness in that

with which it is incompatible.42
One merit of this response is that it provides a short and decisive pro-
nouncement on the fnal category of perfections that might have gener-
ated a threat to the principle of proportionality. For it has already been
established that privative perfections generate no threat to the principle
of proportionality in the context of the transitions in the history of life.
But perhaps we should not foreclose too quickly here. Perhaps it is
worth considering what might be said if one accepted, for the sake of
argument, that Anselmian perfections are metaphysically respectable
properties, and that life, sentience and rationality are indeed genuine
Anselmian perfections. How might the principle of proportionality fare
in that hypothetical context?
Two considerations suggest themselves. The frst is to employ the stan-
dard move in the scholastic playbook and advert to the notion of virtual or
potential existence. Since life, sentience and rationality are all accidents of
substances, their virtual pre-existence cannot be barred by any metaphysi-
cal considerations. Scotus is very clear about this in his commentary on
Book IX of Aristotle’s Metaphysics where, following Aristotle, he writes:
What is peculiar to substance is that there must always exist before-

hand in complete actuality another substance which produces it,
but it is not necessary that a quality or quantity should pre-exist
138 Stephen Boulter
otherwise than potentially. It is impossible for quality or quantity
to exist after being non-existent, unless an efcient cause brings this
about. And it is not necessary that the “quantum” or “quale” pre-
exist in actuality.43
Thus one might claim that in the right circumstances abiotic entities are
able to generate biotic entities because life pre-exists virtually in the abi-
otic, and similarly with the cases of sentience and rationality. Of course
it remains the case that these transitions still lack scientifcally respect-
able explanations. But we should recall the Aristotelian principle that
potentiality is an integral component of the natural order,44 and the fur-
ther principle of modal epistemology that actuality is the surest guide
to possibility in the natural order. While much that is possible remains
unactualised, we can safely assume that anything that has been actualised
is possible.45 Therefore Aristotelians should be prepared to say that the
actual history of life, as documented in textbook biology, has revealed the
potentialities of the substances involved. And just as at one point it was
unimaginable how semen could play a role in the organisation a foetus,
so too the emergence of life from the abiotic will eventually be provided
a naturalistic account, and similarly with sentience and rationality. This
would leave us with the following fve apparently compossible claims:
1. There can be no more negative perfection in an efect than is con-

tained in its efcient cause; but
2. Negative perfections, such as life, sentience and rationality, can exist
virtually or potentially in the cause and actually in the efect;
3. Efcient causality occurs when an agent actualises a potential in a
patient via an action;
4. The potentiality contained in the joint actualities of the agent-patient
pair is known only empirically by noting the actualisation of these
potentialities as they occur (that is, there is no a priori knowledge
of the potential perfections of agent and patients);
5. The fundamental transitions in the history of life have unveiled the
virtual potentialities of life, sentience and rationality in the relevant
substances involved.
I submit that these propositions contain no inconsistency. If this is right,

then again there is no contradiction between evolutionary theory and the
principle of proportionality even in the case of negative perfections in the
order of eminence.
The appeal to virtual pre-existence is the traditional method of han-
dling putative counterexamples to the principle of proportionality. But a
non-traditional response is also available. The key claim in this approach
is that the arrival of life, sentience and rationality, despite their status
as Anselmian perfections, do not mark an absolute increase in overall
negative perfection in or of the cosmos. Rather, throughout the fundamen-
tal transitions in the history of life the cosmos displays the same overall
level of negative perfection as life “advances” from one level to the next.
This view, which might be called the conservation of overall perfection,
is suggested by several considerations.
First, there is one negative perfection which even Aristotelians are likely
to accept with equanimity, namely existence. Indeed, it is the only Ansel-
mian perfection that can be readily admitted into a naturalistic ontology.
And as Aquinas says, “being itself is the most perfect of all things.”46
Therefore, everything that has an act of existence has a claim to this
negative perfection insofar as it exists. And certainly all substances have
an equal claim to this negative perfection insofar as their mode of exis-
tence qua substances is the same across the category. This opens up the
possibility of seeing each transition in the history of life, remarkable as
they are, as simply new ways of sharing in the primary perfection since
none of the transitions involves a move from non-substantial cause to a
substantial efect.
Second, if negative perfections turn on notions of “nobility” and “dig-
nity”, then one might refect on Aristotle’s point that “all things have
something godlike in their nature.”47 His emphasis on the continuity of
living things is a further welcome antidote to anthropocentric thinking.48
We should also note his observations in Parts of Animals. Speaking of the
less glamorous of living creatures he says:
if some have no graces to charm the sense, yet even these, by disclos-
ing to intellectual perception the artistic spirit that designed them,
give immense pleasure to all who can trace links of causation, and are
inclined to philosophy. . . . We therefore must not recoil in childish
aversion from the examination of the humbler animals. Every realm
of nature is marvellous.49
Everything insofar as it exists, particularly every living creature, is

“noble”, is “dignifed”, is “marvellous”, and so has a claim to equal
respect. As Aquinas never tires of saying, all beings qua being are good.
And I take this to be part of the motivation for these lines:
The judgment . . . of the goodness of anything [depends] upon what

it is in itself, and on its reference to the whole universe, wherein every
part has its own perfectly ordered place.50
What is more, the perfection of the cosmos did not need to wait for the
arrival of the “higher” animals, for “[t]he universe in its beginning was
perfect as regards the species of things.”51 This train of thought suggests
that it is unjustifed to claim that certain kinds of beings enjoy a higher
degree of “nobility” or “dignity” than other kinds simply in virtue of
140 Stephen Boulter
their kind. This is an unseemly anthropocentric projection which hampers
development of a truly global ethic that embraces all of humanity as well
as the environment and all other living creatures as rightful objects of
moral concern.
Third, regarding each transition there is a natural tendency to notice
frst the novel increases in operation or power enjoyed by the new kind of
entity. What is less immediately obvious, however, is that the new powers
and operations come with new difculties, problems and defects. This bal-
anced vision reveals that each transition in the history of life involves both
an increase and decrease in perfection indicative of an overall conserva-
tion of perfection in an evolving universe. This is suggested by Maynard
Smith and Szathmary’s study of the origins of life. They noted a general
pattern in each of the fundamental transitions, namely,
Entities that were capable of independent replication before the tran-

sition could afterwards replicate only as part of a larger whole.52
Loss of operational control would normally be seen as a decrease in

perfection by traditional standards. Prokaryotes provide one instance
of this pattern. Other instances include the cells of higher organisms
which depend on being part of a multicellular organism for their activi-
ties; individuals of sexually reproducing species who now only reproduce
as members of a sexual population; and members of colonies, ants, for
example, which can only reproduce as members of their colony, whereas
their ancestors were able to reproduce as a sexual pair.53
This general pattern of counterbalancing perfections is continued as we
mount the scale of being. For example, each of the transitions involves an
increase in the internal complexity of the relevant entities. Yet simplicity
is traditionally seen as a perfection. So whatever is gained in operational
powers after a transition is ofset by the increase in complexity.54
Moreover, the transition from abiotic to biotic involves the emergence
for the frst time of mistakes.55 No doubt this is in part due to the greater
internal complexity of the relevant entities. However that may be, life
cannot be seen as an unqualifed advance against the background of a
mistake-free inanimate order, given that these mistakes are often non-
trivial and introduce the very possibility of sufering for the frst time.
Again, living things, animals in particular, need to be equipped to deal
with their precarious form of existence in a way abiotic entities do not.
In particular there is a need for forms of perception to guide movement
to secure resource requirements and avoid predation. From this perspec-
tive sentience is seen to be both a perfection and a compensation for a
handicap due to a need not present amongst abiotic entities and plants.
Similar remarks apply to rationality in humans. Our intellectual capaci-
ties are both a perfection and a necessary compensation for a handicap not
present in the other higher animals, namely, the lack of natural weapons
or defences against both predators and the elements. The lesson here is
that nature provides what is needful; nobility is not a consideration.
Moreover, the “advance” from the sensory experiences of the sensitive
soul to the abstract concepts of the intellective soul is not a move up from
a material sensory organ to an immaterial intellect, but a sideways move
from informationally rich to informationally impoverished but useful rep-
resentations. This is because concept formation via abstraction involves
ignoring information available in perception. But if information – truth –
is a good, then this transition involves sacrifcing quantity of information
for quality.
Finally, voluntary action is one of the benefts associated with rationality
because it involves an increase in control over behaviour; but it also intro-
duces the possibility, not to say inevitability, of moral mistakes and sufer-
ing of a kind and intensity not known to animal, plant or inanimate entities.
The idea here is that when we view each transition in the history of life
as marking a step in the direction of higher dignity because it is a step
in our direction – as has traditionally been our wont – we are focusing
one-sidedly on the positive gains and neglecting to notice that these gains
are compensations for challenges not faced by entities allegedly lower in
dignity than ourselves. A more balanced assessment suggests that each
transition involves both increases and decreases in negative perfection,
leaving the quantity of overall perfection as it was. We are left then with
the image of a universe that was perfect to begin with, with the emergence
of life and its subsequent history displaying diferent but equally perfect
states of the universe. If this is right, then the transitions themselves do
not involve violations of the principle of proportionality with respect
to hypothetical negative perfections either, and so the compossibility of
evolutionary theory, the principle of proportionality and naturalistic
metaphysics is secured once again.56
Conclusion
The burden of this chapter has been to show that increases in perfection in
the living world are not inconsistent with the principle of proportionality
employed in a naturalistic context. The key to meeting this challenge was
recognising that there are at least four distinct kinds of perfection that need
to be clearly distinguished. With this more secure appreciation of the possi-
ble perfections involved, what emerges is that the transitions do not violate
the principle of proportionality. This was easily recognised in the cases of
privative perfections and perfections in the order of dependence. Negative
perfections in the order of eminence were less straightforward. But there
are at least three distinct ways of showing that the fundamental transitions
in the history of life do not violate the principle of proportionality with
respect to this category of negative perfection either. The result is that pro-
ponents of naturalistic metaphysics can quite happily accept the textbook
142 Stephen Boulter
account of the transitions in the history of life while continuing to respect
the scholastic principle of proportionality. This is to be welcomed, as only a
scientifcally informed metaphysics has a right to our allegiance. This result
is also consistent with various versions of theistic evolution, although it has
not been my concern to show that here. The main casualty of this exercise
has been an element of Neoplatonism which has been entangled but never
successfully blended with Aristotelianism, viz., the traditional account of
negative perfections in the order of eminence. Unreconstructed Neopla-
tonism is inconsistent with the history of life as we know it. But even here
not all is lost. If Neoplatonists accept the idea that there is a conservation
of perfection across transitions, then their intuitions regarding the dignity
and nobility of human beings need not be discarded, merely balanced by a
similar respect accorded to other creatures.
Notes
1. Charles Darwin (2006, p. 760).
2. This turn of phrase is due to Yablo. The scholastic principle of proportionality
has close afnities with the contemporary principle of causal commensuration
in that it codifes a set of intuitions regarding what we take to be a possible
cause of a given efect. The contemporary formulation of the shared core idea
is that a cause is commensurate with an efect if it has “all that is required
to produce the efect, and as little as possible that is not” (Yablo, 1992).
This is thought to be equivalent to the claim that a commensurate cause is
sufcient but also necessary for a given efect. For contemporary discussion
and applications of this principle see Robert Audi (2011), Graham Oddie
(2013 delivered at the APA Central Division Meeting) and Tyler Huismann
(2016). See also Descartes’s “Axioms or Common Notions” in his Second Set
of Replies for his application of the principle (2008, p. 166).
3. It is not unusual to fnd passages like the following: “Overall, what the feld
of protein evolution needs are some plausible, solid hypothesis to explain how
random sequences of amino acids turned into the sophisticated entities we
recognise today as proteins. Until that happens, the phenomenon of the rise
of proteins will remain ‘something close to a miracle’” (from interview with
Dan Tawfx, Weizmann Institute of Science).
4. Douglas Axe is a case in point (2000, 2004). He has calculated the odds of pro-
tein evolution occurring by chance to be one in a trillion, trillion, trillion, tril-
lion, trillion, trillion. Of course we need to distinguish two senses of “miracle”:
(a) An extremely improbable event (given what we know), and (b) an event
requiring divine intervention – a “supernatural” occurrence. But it is tempting
to infer a miracle in sense (b) from a miracle in sense (a), as does Axe.
5. Genesis 1, verses 11 and 20.
6. This would appear to be the position favoured by M. Dodds (2012, pp.
201–202).
7. This is not a case of naïve scientism but a judgment based on a diferential
certainty argument, which favours saving the basic facts of the history of
life rather than the more recherché claims of causal metaphysics if the two
really are incompatible. For a defence of this basic principle see Chapter 2 of
Boulter (2007). Of course accommodation is not always one-way. For studies
of cases where scholastic metaphysics in particular can help solve puzzles in
evolutionary theory see Part II of Boulter (2013).
8. In order to motivate a claim regarding the principles by which one created
substance produces another, Suárez says,
if this is not so, then it will be necessary in many cases to attribute
the entire efecting of a substance to universal causes or even to the
First Cause – something that ought to be avoided in philosophy as far
as possible. For the proper management of the universe demands that
whatever can be brought about appropriately and connaturally by sec-
ondary causes should be brought about by them.
(Metaphysical Disputations 18.2, p. 62)
Here Suárez is following Aquinas very closely indeed (Disputed Questions
on Truth, q. 11, a. 1). This is not the place to discuss the merits of theistic
versions of evolutionary theory. For present purposes the point is merely
that there is no reason in principle to think that the transitions in the history
of life cannot be given a naturalistic interpretation. For highly competent
accounts of theistic versions of evolutionary theory see Jitze M van der Meer
(2021), Mariusz Tabaczek (2019) and Michael Chaberek (2017).
9. Here I am generalising Morgan’s Canon, a version of the Scholastic principle
of economy. He writes, “In no case is an animal activity to be interpreted in
terms of higher psychological processes if it can be fairly interpreted in terms
of processes which stand lower in the scale of psychological evolution and
development” (1903, p. 59). The so-called zebra principle of medical diagno-
sis also applies here – never advert to the exotic when the mundane will do.
10. We do occasionally fnd less careful statements of the principle, as in the
case of Dante: “Whatever is changed from potentiality into act is changed
by something which actually exists in the form to which it is changed; if an
agent tried to act otherwise, he would act in vain” (1957, p. 18). But this is
a naïve and untenable version of the principle.
11. Passages containing such expressions are ubiquitous in the writings of the
scholastics. For illustrative examples from Aquinas Summa Theologiae I.95.1,
1–2.66.1, 1–2.63.2 obj 3, 2–2.24.6, SCG 1.67, Summa Contra Gentiles 1.41,
3.120.
12. As ever, the model here is Aristotle (2011): An aporia begins with statements
that are “true but unclarifed” and we hope “to achieve an outcome that is
both true and clear” (The Eudemian Ethics, II.I). Those long familiar with
this particular problem will know that George Klubertanz (1941) cast it in
the form of a dilemma as well.
13. (2000, p. 17).
14. Yablo (1992).
15. (2018, pp. 445–448).
16. For possible counterexamples from within the Aristotelian tradition see Sor-
abji’s (1980) discussion of coincidental occurrences in Chapter 1.
17. See Suárez (1994), Metaphysical Disputations, 17.2 and 18.2, for extended
discussion of some of these cases.
18. Aquinas writes,
A living thing is more perfect than that which merely exists; and an intel-
ligent thing than what merely lives. Therefore life is more perfect than
existence; and knowledge than life. But the essence of God is existence
itself. Therefore He has not the perfections of life, and knowledge, and
other similar perfections (despite being the First Cause).
(Summa Theologiae I. 4.2)
19. I will return to case ix in due course.
144 Stephen Boulter
20. See his Monologium, ch. 15 in Anselm (1988). For the infuence of Neoplatonism
in scholastic thought and Aquinas in particular see Wayne J. Hankey (2019).
21. Scotus, De Primo Principio, ch. 1, section 1, and ch. IV, section 3, B.
22. Aquinas, Summa Theologiae I, q. 2, a. 3.
23. Wuellner (1956, p. 226).
24. In the course of disambiguating the notions of prior and posterior, Scotus says,
essential order seems to be divided by a primary division . . . namely, into
the order of eminence and the order of dependence. In the frst mode the
prior is called eminent, and the posterior is called that which is exceeded.
To put it briefy, whatever is more perfect and noble according to essence
is thus prior . . . In the second mode the prior is called that upon which
something depends, and the posterior is called that which depends . . .
the prior according to nature and essence is that which is able to exist
without the posterior, but not conversely.
(De Primo Principio, p. 5)
25. The confation of the order of dependence and eminence is not always
avoided and can lead to muddled thinking. Even Scotus fails to keep them
apart, as when he says that act and potency are ordered as prior and poste-
rior in the order of eminence, while substance and accident are ordered as
prior and posterior in the order of dependence. It would be more natural to
say that in both cases the appropriate order is the order of dependence. But
the confation is often more egregious. Consider this passage from Augus-
tine’s The Teacher:
Adeodatus: I don’t know whether any objection can be raised against the
rule according to which everything that exists on account of another is
said to be inferior to that on account of which it exists.
Augustine: We’ll discuss this more opportunely and more carefully at
another time. Right now what you concede is sufcient for what I am try-
ing to establish. You grant that knowledge of things is more valuable than
the signs of things, and for this reason knowledge of the things signifed
should be preferable to knowledge of their signs. Doesn’t it seem so to you?
Adeodatus: Surely I haven’t conceded that knowledge of things is
superior to the knowledge of signs, and not just superior to the signs
themselves, have I? So I have misgivings about agreeing with you on
that score. The name “flth” is better than the thing it signifes. What if
knowledge of this name is then likewise to be preferred to knowledge of
the thing, although the name itself is inferior to that knowledge? There
are obviously four things here: (a) the name; (b) the thing; (c) knowledge
of the name; (d) knowledge of the thing. Just as (a) surpasses (b), then,
why shouldn’t (c) also surpass (d)? Yet even if (c) were not to surpass (d),
surely (c) isn’t then to be subordinated to (d), is it?
(1995, 9.27)
In this passage we have both orders invoked at least implicitly; but it is clear
that the order of eminence is about human attitudes, not mind-independent
features of things in the natural order.
26. Metaphysics, V, ch. 16. Emphasis added. Aristotle (1941).
27. Summa Theologiae I.II, q. 91, a. 5.
28. Summa Theologiae I, q. 4, a. 1.
29. Second Quodlibet, q. 13. Ockham (1991).
30. Metaphysical Disputations 30.1.1.
31. For example, it was common to distinguish between pure and impure perfec-
tions. Pure perfections are properties that do not imply an upper limit, while
impure perfections are those which do. This distinction is at work in this
passage from Leibniz:
We must know what a perfection is. One thing which can be afrmed
about it is that those forms or natures which are not susceptible of it to
the highest degree, say the nature of numbers or of fgures, do not permit
perfection. This is because the number which is the greatest of all (that is,
the sum of all the numbers), and likewise the greatest of all fgures, imply
contradictions. The greatest knowledge, however, and omnipotence con-
tain no impossibility. Consequently power and knowledge do admit of
perfection, and in so far as they pertain to God they have no limits.
(Discourse on Metaphysics, 2005, p. 1)
32. (2017, p. 84).
33. For Thomists this becomes most pressing when it is noticed that Aquinas uses
the Aristotelian notion of privative perfection in discussing God’s perfection,
while seemingly relying on the Anselmian notion in the course of the fourth
way in Summa Theologiae I, q. 2, a. 3. There is pressure to fnd houseroom
for both kinds of perfection if at all possible.
34. Summa Contra Gentiles II, ch. 21.
35. For extended discussion of two versions of Thomistic hylomorphism that
maintain the immateriality of the intellect in this sense, see Chapter 9 of this
volume by Christopher Hauser.
37. De Anima, III, ch. 5 in his (1941). Aristotle (1941).
38. The immateriality of the intellect sits badly with the thesis that the soul is the
form of the body and the thesis that a human person is a metaphysical unity.
39. Opus Oxoniense, IV, dist. XLIII, qu. 2 in his (1987). Scotus (1987).
40. It is interesting to note that Buridan follows Scotus’s lead (Questions on Aristo-
tle’s De Anima, 3.4–5, in Klima, Allhof and Vaidya, 2007). He maintains that
the intellective soul is as “organic” as the sensitive soul, and the worry expressed
by Aristotle and Aquinas regarding the nature of abstract concepts goes unmen-
tioned. After providing four separate reasons for thinking the intellective soul
cannot be immaterial, he concludes, “natural reason would dictate that the
human intellect is drawn forth out of the power belonging to matter, and that it
is generable and corruptible” (ibid., 3.5). And of course this result was restated
by Pomponazzi. For a good discussion of Pomponazzi see Spruit (2020).
41. See Ramos-Díaz’s contribution to this collection for an extended account
and defence of an argument to this efect developed by James Ross for the
immateriality of the intellect based on semantic considerations. The idea
is that formal thought has the property of being semantically determinate
while no physical system can be semantically determinate. Hence there is
an “in principle diference” between formal thought and any physical body.
Ramos-Díaz suggests that this diference is sufcient to warrant the claim
that formal thought could not be generated by the purely physical processes
of evolution. For my part, I hesitate to take semantic considerations as a
frm basis for metaphysical theses. In this I depart from the analytical tradi-
tion which takes the philosophy of language to be First Philosophy. But
setting that aside, even when one accepts that there is an “in principle dif-
ference” between formal thought and any bodily organ, the crucial question
remains, viz., whether this diference amounts to a real distinction allowing
for the two-way separability of the intellect from any bodily organ. While
the Ross argument does establish that there is some kind of distinction
between formal thought and any bodily organ, it does not establish that
the distinction is real (as opposed to formal, modal or merely mental).
146 Stephen Boulter
Moreover, nothing short of the actual separation of an act of formal thought
from any bodily organ can establish such a distinction. But if it has yet to be
established that the intellect can exist without the body, then there is as yet
no reason to think that evolutionary processes have violated the principle
of proportionality in the order of dependence. For further contemporary
debate see Robert Pasnau (2011).
42. De Primo Principio, p. 79.
43. 2000, p. 211. Aristotle (1941).
44. Mariusz Tabaczek (2020, p. 25) emphasises this point. There are
two potentialities (levels) inherent in the very fabric of the cosmos (1) the
pure potentiality of prime matter which can be actualised by all possible
types of substantial forms, proper for both inanimate and animate natural
kinds; (2) potentiality of primary matter underlying each and every instan-
tiation of secondary matter, which is specifed by the substantial form and
accidental forms characteristic of a particular natural kind it belongs to.
45. Suárez is my guide here. When discussing whether we can know that two
entities are really distinct and so could potentially be separated, when they
have never actually been separated, either in the course of nature of through
divine intervention, he writes:
a sure sign is hardly forthcoming when things such that naturally they
are always and necessarily found to be really united to one another, and
have not thus far been separated by divine intervention. I do not deny that
many things of this kind may be found to be really distinct . . . I am merely
pointing out that we lack a universal sign for discerning such a distinction.
(Metaphysical Disputation 7, 2007, section 2.21)
47. Eudemian Ethics, VI, 14, 1153b, 32–33.
48. History of Animals, Book VIII, ch. 1, in his (Aristotle 1941).
49. Book I, ch. 5, in his (Aristotle 1941). Emphasis added.
50. Aquinas, Summa Theologica, I. q. 49, a. 3.
51. De Potentia I, q. 3, a. 10, ad. 2; Aquinas (2012).
52. 1999, p. 19.
53. Ibid., p. 19.
54. While simplicity in God is a perfection, Aquinas does say that simplicity in
material things is not (Prologue to Summa Theologiae I, q. 3). It is not entirely
clear to me how this double standard is maintained if simplicity is a negative
perfection.
55. My thanks to David S. Oderberg who drew this point to my attention. He
uses precisely this point to diferentiate the abiotic from the biotic: “only
living things make mistakes – as manifested by disease, injury and various
kinds of maladaptation to the environment. Organisms get things wrong.
Even DNA itself is full of ‘copying errors’, leading to mutations and genetic
illnesses. But atoms and molecules do not make mistakes.”
56. This conservation of perfection thesis has a striking parallel with what might
be called a conservation of information thesis forwarded by Ashley. He writes,
[A] new species is not a “greater emerging from the less” because the
amount of information it contains in integrated form is no greater than
the amount of information present in the historical process. What is spread
out in history is condensed, as it were, in the emerging new species.
(1972, p. 215)
My thanks to Mariusz Tabaczek for bring this paper to my attention.
References
Anselm (1988) Basic Writings. La Salle, IL: Open Court Publishing.
Aquinas, Thomas. Disputed Questions on Truth. Available at: https://documen-
tacatholicaomnia.eu/03d/1225-1274,_Thomas_Aquinas,_The_29_questions_on_
Truth_(Mulligan_Translation),_EN.pdf
Aquinas, Thomas. Summa Contra Gentiles. Available at: https://isidore.co/aqui-
nas/english/ContraGentiles.htm
Aquinas, Thomas (1999) Summa Theologiae. Available at: www.documentacath-
olicaomnia.eu/03d/1225-1274,_Thomas_Aquinas,_Summa_Theologiae_%
5B1%5D,_EN.pdf
Aquinas, Thomas (2012) The Power of God. Translation by Richard Regan.
Oxford: Oxford University Press.
Aristotle (1941) The Basic Works of Aristotle. Edited by Richard McKeon. New
York: Random House.
Aristotle (2011) The Eudemian Ethics. Translation by A. Kenny. Oxford: Oxford
University Press.
Ashley, Benedict (1972) “Causality and evolution”, The Thomist, 36, 199–230.
Audi, Robert (2011) “Causation, coincidence and commensuration”, Philosophi-
cal Studies, 162, 447–464.
Augustine (1995) Against the Academicians and The Teacher. Translation by
Peter King. Indianapolis: Hackett Publishing.
Axe, Douglas (2000) “Extreme functional sensitivity to conservative amino acid
changes on Enzyme exteriors”, Journal of Molecular Biology, 301(3), 585–595.
Axe, Douglas (2004) “Estimating the prevalence of protein sequences adopting
functional enzyme folds”, JMB, 341(5), 1298–1315.
Boulter, Stephen (2007) The Rediscovery of Common Sense Philosophy. Hound-
mills: Palgrave Macmillan.
Boulter, Stephen (2013) Metaphysics from a Biological Point of View. Hound-
mills: Palgrave Macmillan.
Chaberek, Michael (2017) Aquinas on Evolution. Leicester: The Chartwell Press.
Dante, Alighieri (1957) On World-Government. Translation by Herbert Schnei-
der. Indianapolis: The Bobbs-Merrill Co.
Darwin, Charles (2006) The Four Great Books of Charles Darwin. Edited by
Edward O. Wilson. New York: W. W. Norton.
Descartes, René (2008) The Philosophical Writings of Descartes, Vol. II. Edited
by John Cottingham, Robert Stoothof and Dugald Murdoch. Cambridge:
Cambridge University Press.
de Wulf, Maurice (2017) Medieval Philosophy, Illustrated from the System of
Thomas Aquinas. Trieste: Trieste Publishing.
Dodds, M. (2012) Unlocking Divine Action. Washington, DC: Catholic Univer-
sity of America Press.
Hankey, Wayne J. (2019) Aquinas’ Neoplatonism in the Summa Theologica on
God: A Short Introduction. South Bend, IN: St Augustine Press.
Huismann, Tyler (2016) “Aristotle on accidental causation”, Journal of the
American Philosophical Association, 561–575.
Jitze, M. van der Meer (2021) “The concept of natural selection in Theodosius
Dobzhansky: Its development and interpretation”, in Evolutionary Biology:
New Perspectives on its Development, Vol. 3, Richard Delisle (ed.). Switzer-
land: Springer Verlag.
148 Stephen Boulter
Gyula Klima, Fritz Allhof and Anand Vaidya (eds.) (2007) Medieval Philosophy:
Essential Readings with Commentary. Oxford: Blackwell.
Klubertanz, George (1941) “Causality and evolution”, The Modern Schoolman,
19(1), 11–14.
Leibniz, G.W. (2005) Discourse on Metaphysics and The Monadology. Transla-
tion by George Montgomery. Mineola, NY: Dover Publications.
Maynard Smith, John and Szathmary, Eors (2000) The Origins of Life. Oxford:
Oxford University Press.
Morgan, C. Lloyd (1903) An Introduction to Comparative Psychology. London:
W. Scott.
Ockham, William (1991) Quodlibetal Questions. Translation by Alfred Freddoso
and Francis Kelley. New Haven: Yale University Press.
Oddie, Graham (2013) “An Argument against Commensurate Truthmakers”,
delivered at the APA Central Division Meeting.
Pasnau, Robert (2011) “Philosophy of mind and human nature: Biological or
theological,” in The Oxford Handbook of Aquinas. Oxford: Oxford University
Press.
Cofey, Peter (2018) Ontology or the Theory of Being. Miltons Keynes: Alpha
Editions.
Scotus, John Duns (1949) The De Primo Principio of John Duns Scotus. Transla-
tion by Evan Roche. Washington: The Franciscan Institute.
Scotus, John Duns (1987) Philosophical Writings. Edited by Allan Wolter. India-
napolis: Hackett Publishing.
Scotus, John Duns (2000) A Treatise on Potency and Act: Questions on the Meta-
physics of Aristotle, Book IX. Translation by Allan Wolter. New York: The
Franciscan Institute.
Sorabji, Richard (1980) Necessity, Cause and Blame: Perspectives on Aristotle’s
Theory. London: Duckworth.
Spruit, Leendert (2020) “The Pomponazzi afair: The controversy of the immor-
tality of the soul”, in Routledge Companion to Sixteenth-Century Philosophy.
Henrik Lagerlund and Benjamin Hill (eds.). New York: Routledge.
Suárez, Francisco. Metaphysical Disputation 30. Translation by Sydney Penner.
Available at www.sydneypenner.ca/su/DM_30_1.pdf.
Suárez, Francisco (1994) On Efcient Causality: Metaphysical Disputations 17,
18 & 19. Translation by Alfred Freddoso. New Haven: Yale University Press.
Suárez, Francisco (2007) On the Various Kinds of Distinction: Metaphysical Dis-
putation 7. Translation by Cyril Vollert. Milwaukee: Marquette University
Press.
Tabaczek, Mariusz (2019) “What do god and creatures really do in an evolution-
ary change? Divine concurrence and transformation from a Thomist perspec-
tive”, American Catholic Philosophical Quarterly, 93(3), 445–482.
Tabaczek, Mariusz (2020) “The metaphysics of evolution: From Aquinas’ inter-
pretation of Augustine’s concept of rationale seminales to the contemporary
Thomists accounts of species transformation”, Nova et Vetera, English edition,
18(3), 945–972.
Wuellner, Bernard (1956) Summary of Scholastic Principles. Chicago: Loyola
University Press.
Yablo, Stephen (1992) “Mental causation”, The Philosophical Review, 101,
145–180.
Part 2
Mind and Nature

5 Free Will in a Network of
Interacting Causes
Timothy O’Connor
A number of contemporary metaphysicians are engaged in the project of

working out a pared-down, broadly neo-Aristotelian account of the natu-
ral world that provides a plausible framework for modern science.1 Much
as one seems to fnd diferences in detail within a consistent broad vision
at diferent locations in Aristotle’s corpus, so contemporary theorists have
forged diferent footpaths through the neo-Aristotelian grove. Two impor-
tant, contested categories in the recent discussion are substance/object and
efcient causation. In what follows, I have two aims. The frst aim is to
indicate very general analyses of these categories that I favor and to situate
and motivate them within this recent discussion. These partial analyses
leave open many difcult questions of detail. My intention here is to say
just enough to enable me to fulfll my second aim, which is to propose and
defend a way of extending such an attractive, broad analysis to human
persons and the power of deliberate choice, in a manner that undergirds
an indeterministic account of free will.
I. Property and Object

In thinking about the metaphysical category of substance (or, as I shall
often say, object), it is helpful to start by thinking about simple (noncom-
posite) entities. We don’t know at present what the simple entities are that
constitute the building blocks of our universe or even whether they have
the character of particles or felds (or both). For concreteness in what
follows, let’s suppose they include the known specifc varieties of quark,
lepton, and boson particles. On physical theory, each such particle type is
defned by a privileged few invariant properties, which are characterized
in turn in terms of the way the particles interact. The neo-Aristotelian
takes the success of such theorizing as indication that the core defning
features of each particle type constitute a real essence, and the functional/
interactional means of identifying the sparse ‘natural’ properties should be
taken as essential characterizations: they are ‘causal powers’. To identify
such properties as powers (better: as powerful qualities)2 is to assert at a
frst pass that their essence in some manner involves potentiality. Objects
DOI: 10.4324/9781003125860-8
152 Timothy O’Connor
that have them are necessarily disposed to interact with other objects in
certain ways. Sometimes such interactions appear to be structured, such
that one object or set of objects causes (or contributes to causing) certain
efects in another, without reciprocation. But, as C. B. Martin (1993,
2008) and John Heil (2012) emphasize, often it will instead be a matter of
mutual manifestation, with the interacting partners co-producing a shared
state or process, as with the mutual repulsion of adjacent electrons in the
absence of countervailing forces.
I will say more later concerning the metaphysics of (efcient) causa-
tion that grounds the varieties of true causal explanations. First, let us
observe that for properties to play a fundamental role in our understand-
ing of how the world unfolds – for properties to be causal powers – they
need to be fundamental realities (contra nominalism) and immanent to
things (contra Platonism). Whatever other uses they may serve, spatio-
temporally transcendent Platonic universals are not suited to the needs
of sober empirical theorizing (as Armstrong 1978 convincingly argued).
How then should we understand the immanence of natural properties?
Not as separable parts, surely! The very idea of an isolated quantity of
mass that isn’t the mass of anything is absurd – a seeming impossibility.
E. J. Lowe (2012) and John Heil (2012) argue that from the premise that
immanent properties (which they dub ‘modes’) are inherently dependent
entities, not capable of independent existence, we should conclude that
they cannot be constituents: entities that enter into the makeup of a
thing, conferring a distinctive kind of structure to it, prompting the ques-
tion whether a thing’s modes collectively exhaust its being or instead
conjoin with some further, non-qualitative ingredient. However, the
inference from inherent dependence to non-constituency is unconvinc-
ing. The notion of constituency is not analytically connected to that of
real separability. Further, Lowe and Heil allow that modes are ‘aspects’
of objects, particular (while being such that distinct individuals can have
exactly resembling modes), located where the objects are, perceivable in
some cases, and partly responsible for objects having causal powers of
various kinds. How can they fll all these roles without being (proper)
constituents of the objects?
The case for constituency (in Scotistic terms, a real and not merely
formal distinction between property and the object it characterizes) is bol-
stered by the fact that there is some overlap of properties among distinct
basic kinds. For example, mass is a defning characteristic of many particle
types. This is a case of a shared determinable where the types may vary
in their specifc magnitude of it. But there are also cases of shared deter-
minates, as in the case of electrons and positrons, which possess identical
values of mass, spin, and magnetic moment, difering only in the direction
of their charge. That fundamental qualities of the same precise kind can
enter into qualitatively nonidentical individuals force us to acknowledge
that they are ingredients contributing to distinct kinds of structure.
Free Will in a Network of Causes 153
That said, Lowe and Heil are right to underscore that these features of
things are inherently dependent entities, not capable of existence apart
from the objects they characterize. That dependency seems, frstly, more
congruent with their being particular, rather than universals: ‘tropes’
that depend on the particular objects they help to constitute rather than
Armstrongian immanent universals that exist in a plurality of scattered
objects, more loosely dependent for their existence only on there being
some or other ‘host’ objects and not on any particular object.3 Secondly,
the ontological dependence of properties implies that the way they enter
into the makeup of an object importantly difers from the way basic
objects enter into the makeup of a composite object, enough so that we
should understand property constituency to be a fundamentally distinct
relation from the part-whole relation between objects.
We have two broad options for a theory of property constituency. The
frst option is to say that objects are (mere) bundles of tropes, unifed by
a primitive relation. Concerning this proposal I will merely echo the com-
plaint of others that it implausibly suggests that tropes are, in the words of
A. J. Ayer, ‘junior substances’. I prefer an alternative, substratum-attribute
account, on which individual basic objects consist in a substratum (a.k.a.
bare particular or thin particular) being joined in a theoretically primitive
way to a cluster of natural properties, the object’s powerful qualities.4
Some critics have charged that the substratum-attribute theory of
objects is incoherent. The theory requires that substrata ‘in themselves’
have no properties, and, say the critics, nothing can be like that. (And
isn’t the very assertion that a substratum lacks properties self-refuting?)
This objection fails to recognize the complementary roles that ontology
and ideology play in any theory. We of course can truly say some things
about substrata. (If we couldn’t, we wouldn’t have a theory!) We can say,
for example, that substrata are inherently incomplete, particular entities;
that they are necessarily joined in a primitive way with particular natural
properties/powers, resulting in individuals that interact with one another
over time; and that the two categories of substratum and attribute play
diferent roles in our thinking about objects (as refected in the fact that
powers, and not substrata, ground objective similarity and causal infu-
ence).5 We don’t say that these statements are true in virtue of substrata
being joined ‘in themselves’ to a further, special class of tropes or in virtue
of a fundamental join relation – moves destined to be futile, as one can-
not ‘ontologize’ all true predication.6 Instead, they constitute the theory’s
ideological machinery, a set of primitive truths that identify the theoretical
roles of the theory’s most basic categories.7
So, a basic object (one not composed of object-parts) is a structured
unity featuring a substratum and a small number of particular powerful
qualities (as the case may be, certain values of mass, charge, spin, and
magnetic moment). A fully worked-out theory would provide some basis
for saying that the substratum and each of the qualities of a given object
essentially reside within it and for distinguishing the way essential and
accidental properties enter into the makeup of an object. Doing so seems
to require an account on which structured individuals are no less ontologi-
cally fundamental than their constituents are. Although it is embedded in
a diferent framework, Peter Simons’s (1994) ‘nuclear’ bundle theory is
suggestive of a direction this might go.
II. Time
One other issue that needs to be addressed before turning to neo-
Aristotelian options for a metaphysical theory of causation is the nature
of time. Of course, this is itself a large and vexed matter, with four broad
views commanding some allegiance, each of which faces signifcant chal-
lenges. Here I will only gesture at reasons for neo-Aristotelians embracing
the perspective on free will that I seek to elaborate later to limit their
consideration to variants on two of these views, the choice between which
I leave open for present purposes.
Four-dimensionalism maintains that natural reality encompasses the
objects and events at all past, present, and future times; that time does
not ‘pass’; and that there is no ontological priority relation among objects
and events based on their temporal ordering, any more than there is such
a relation based on ordering along a spatial axis. This last thesis conficts
with the neo-Aristotelian’s understanding of efcient causation as a fun-
damental generative relation.8
Three-dimensional eternalism accepts the frst of the four-dimensional-
ist theses while afrming passage as a fundamental aspect of time. While
objects and events at all times exist, there is a further fundamental and
ever-changing fact about which such entities exist or occur now. This pos-
tulate at least opens up the possibility of robustly ontological past-future
causal dependence. As critics have observed, simply adding the passage of
time (a moving ‘now’, or present moment) to an ontology of objects and
events over all times renders it a ghostly phenomenon; the ever-changing
facts concerning what is happening in the metaphysical now have no
discernible impact on events.9 Time may pass, but past entities remain as
they were, such as Aristotle in all of his varying states of fowering youth
and aging decrepitude, and likewise for entities in the far-of future. Such
a view seems no more compatible with Aristotelian productive causa-
tion than is four-dimensionalism. But the strangeness of this picture and
puzzles to which it gives rise have led recent proponents to add more fun-
damental change to objects themselves: only present entities are, we might
say, fully concretely real, so the passage of time marks real and continuous
change.10 Arthur Prior (who died in 1969) may exist, as does his storied
painful experience at the dentist in 1954, but not as he and it did when
they were present. Prior changed from being alive, talking, ambulatory,
and so on, to being none of these things; his toothache exists, but it is not
the experience of pain within a fesh-and-blood human being. In general,
such now-past entities exist only ‘abstractly’, as stripped-down loci that
serve as referents of our true statements concerning the past. Puzzled
questions concerning details needed to work out views of this sort will
naturally occur to the reader. I here set these aside. I note that the broad
approach does give footing for a kind of irreducible productive causation.
We may say that such causes, when present, make concretely real what
had previously been merely abstract. (No sooner do events become real
than they retreat to abstraction once again. ‘From abstraction you came,
and to abstraction you shall return.’) We might well question whether
this provides the basis for a satisfactory account of productive causation.
But I want to highlight a diferent reason for deeming it inadequate for
the purpose I have set for myself in this chapter. The conception of free
will that I aim to elucidate is one on which our free choices help to settle
what the future will be (in the tiny corner of it that concerns ourselves and
that is in the sphere of our direct and indirect infuence). But this cannot
be so if all future events exist and are fully determinate. On this concep-
tion of temporal reality, my choices may give (or contribute to giving)
future events a concrete embodiment. It may also be that, were I not to
have so chosen, diferent events would have occurred instead (contrary
to one variety of fatalism). It may even be that some of my choices do not
have sufcient antecedent causes (contrary to causal determinism). Even
so, it remains that my choices do not settle what any subsequent aspect
of an eternally determinate reality shall be, as there is no time at which
the character of reality at subsequent times is not fully settled. (Even on
eternalism, every proposition concerning a specifc event is of course made
true at a particular time or temporal interval, not by a prior choice or
anything antecedent, but by the event itself. But whether there will be such
an event is not metaphysically unsettled or ‘open’, as reality on eternalism
is neither added to nor subtracted from.)
It seems to me in any case that modifed three-dimensional eternalism is
an unnatural stopping point on the road to a third theory of time. Given
that past and future events are conceived as not having the fully concrete
reality that they had or will have when they are present, the natural next
step is to embrace the more straightforwardly intelligible claim that causes
bring their efects into being tout court, rather than merely ‘making them
to be concrete’. It is through the passage of time that concrete realities
recede into a permanent, ghostly status. If the world is fully deterministic,
then sufcient grounds for future truth reside in the dispositional structure
of the present. If the world is not deterministic, then the claim that there
is nonetheless fully determinate future truth is unmotivated. The neo-
Aristotelian gives a non-defationary understanding to the truism that the
future will be what it will be because of the sum total of its antecedent
causes. From this she should naturally conclude that, if the world’s causes
are not fully determining, what the future will be is not fully settled. With
this last, we have arrived at a third theory of time, growing block. More
exactly, we have arrived at a modifed version of growing block theory,
on which the reality of the past is downgraded to a skeletal structure
providing minimal truthmakers.
A fourth theory, presentism, ostensibly eschews altogether the reality
of the past (and the future). However, under pressure to account for the
grounds of truths concerning the past, recent theorists tend to postulate
presently existing surrogates, in the form, for example, of primitive, past-
referring properties such as its having been the case that the dinosaurs once
roamed, and also including haecceities of individuals who have ceased to
be. It is in virtue of such properties that our world is distinguished from
a world whose frst moment is a state identical to the present state of
our world in all its non-tensed intrinsic properties. If the presentist also
accepts the existence of a geometrically defned manifold of spatiotem-
poral points (an essential part of the theory of general relativity),11 then
the gap between modifed growing block and modifed presentism is quite
narrow, with a similar stock of basic entities, diferently arranged. I take
both theories to be in need of further development that cannot be pursued
here. The later accounts of the metaphysics of causation and then of free
will are consistent with either of these two theories of time.
III. Causation
Contemporary neo-Aristotelians are united in thinking that efcient cau-
sation is a real and irreducible feature of the world. But they difer in their
theoretical characterization of it. One dispute concerns whether there is a
fundamental relation of causation. They also disagree concerning whether
causes are substances/objects, properties, events, or facts. In addressing
these questions, let us again focus for now on causal processes featur-
ing (presumed) basic objects/substances. We will take up the question of
composite powers when we consider our ultimate target of the powers of
human beings (and, in particular, the power of choice).
The Relation of Causation

John Heil (2016, 132f; 2017, §5; and following Martin 1993, 2008)
contends that causation never has the asymmetrical character of cause
and efect but instead consists in the mutual manifestation of certain dis-
positional properties (causal powers) of substances. As an example of
what he takes to be a completely general point, it is at best misleading to
say that a quantity of water exercised its power to dissolve some salt in
a given situation; a more metaphysically perspicuous description will say
instead that when the two are brought into contact, the distinctive powers
in each substance jointly manifest themselves in the form of the solute.
Heil further contends that since properties are inherently dispositional,
or tendencies towards such manifestations, manifestations are always
necessary in the appropriate context. In consequence, causation qualifes
as an internal relation, grounded in its relata, rather than being a further,
fundamental tie between them.
Heil notes that many will argue that this picture of the causal nexus
is threatened by the success of physical theories on which certain funda-
mental processes are taken to unfold nondeterministically, as with the
decay of radioactive particles. Some neo-Aristotelians see this as indicat-
ing irreducibly probabilistic causal tendencies. If there are such ‘chancy’
tendencies, their manifestation, when they occur, are not grounded solely
in facts about the arrangement of powers. However, Heil argues that
a more plausible (neo-Aristotelian) metaphysical interpretation of these
stochastic phenomena is instead that they are uncaused – spontaneous
occurrences that are ‘sprinkled in’ to an otherwise fully necessary, unfold-
ing dispositional matrix. Each spontaneous occurrence then reverberates
throughout the nexus in the fully necessitating fashion of mutual mani-
festations (2016, 134).
This is unconvincing. At the most general level, the neo-Aristotelian’s
rejection of the alternate neo-Humean vision of metaphysically indepen-
dent (‘loose and separate’) events stems from the latter’s countenancing of
entirely groundless contingency, made worse by the ‘cosmic coincidence’
of metaphysically independent events falling into deep regular patterns.
Heil’s proposal cedes ground to the neo-Humean: a steady stream of brute
fundamental occurrences that nonetheless conform to statistical patterns
(enshrined, e.g., in the half-life of radioactive substances). This conces-
sion is surely unwarranted, given the ready availability of an alternative
account, according to which such substances are manifesting a nondeter-
ministic, self-directed propensity, the character of which is well-defned
(the emission of a nuclear particle), but whose timing is undetermined on
each occasion and is (largely) impervious to external stimulus.
Heil proposed to reduce the ostensible relation of causation to the ‘join-
ing’ of the relata in a spatiotemporal relation that simply is their mutual
manifestation. Anna Marmodoro proposes to do so through an enriched
ontology on which causal powers undergo internal transitions of state
(from ‘potential’ to ‘activated’) in the presence of suitable partner-powers
(2017, 59). If this is accepted, we are naturally led to ask what is respon-
sible for such state transitions. Here the specifcation is vague: ‘It is the
“coming together” in appropriate conditions of the partner-powers that
accounts for their mutual activation’ (65). Accounts for in what sense and
in what manner? As with Heil and Martin, so here language is used in
places that suggests more than is ofered, for example:
Because of this mutual dependence, partner-powers manifest them-

selves in activities that are co-determined, co-varying, and co-
extensive in time . . . It follows that if causation is explained in terms
of causal powers, there is reciprocity in it in the sense that both causal
partner-powers get activated, each ‘facilitating’ the activation and the
activity of the other.
(66, n21, emphasis in original)
‘Co-determining’ and ‘facilitating’ the activity of the power-partner

sounds relational, but the reality of a relation is disavowed: ‘All there is to
their causal “interaction” is their mutual and simultaneous manifestation
. . . There is . . . no relation bridging the two’ (70, emphasis in original).
It seems that what we are left with, in this fnal analysis, is an account on
which powers necessarily activate in diverse particular ways in the pres-
ence of structures of other powers of diverse kinds. This is a real internal
change that happens of necessity in the right circumstance but without
there being a real relation between any aspect of the circumstance and
it. What a power does is manifest itself in one of the ways of which it
is capable; towards others, its presence is merely the occasion on which
the other power manifests itself in a certain characteristic way. When the
gloss of other-directed causal language is stripped away in favor of an
austere metaphysical description capturing how the account sees things,
all activity seems weirdly internal. It’s a coherent description, as best I
can judge, but one that is not highly motivated apart from a compelling
theoretical reason for eschewing real relations altogether.
Lowe (2016) ofers such a reason in the context of proposing yet
another non-relational account.12 He complains that real relations in gen-
eral are ‘ontologically weird’, since ‘[a] relational trope or mode would
be an “abstract” particular that was dependent for its existence and iden-
tity on two distinct and quite possibly ontologically independent objects’
(111). Applying it to the case of a causal relation between a power (or
its bearer) and its manifestation in some distinct object, the complaint is
that there would be a particular instance of causing that could not exist
absent the entities it relates, and yet without its existing ‘in’ either of
them in the way that a monadic power dependently exists in its bearer.
Monadic tropes by their nature are ‘adjectival’ aspects of things that are
grasped in thought by abstracting away from the thing’s other aspects.
That we cannot say something analogous for a relational trope vis-à-vis
its two or more supporting substances brings out the weirdness of the
very notion (112).
Here we see a further consequence of our earlier rejection of Lowe’s
non-constituent understanding of tropes/powers. If, contra Lowe, these
are not merely formal realities contemplatable through abstraction but
real entities in their own right, giving a distinctive kind of ontological
structure to objects, then there is no mystery to overcome in there being
real causal relations between objects as well, ontologically dependent on
each of them. They are not entities afxed between the objects from out-
side, as it were; they are the activity of the one towards the other (perhaps
matched by a corresponding relation fowing in the other direction, some-
times or always).13
The Kind of Entity Causes Are

According to some recent neo-Aristotelians, strictly speaking, neither
substances nor the events they undergo are the causes of manifestations
(efects), the constituent causal powers themselves are. For Marmodoro
(2017), this thesis is a natural consequence of her having adopted a bundle
theory on which objects are constituted by powers, a theory I rejected
earlier in the chapter. But Stephen Mumford and Rani Anjum (2011) also
accept the thesis while embracing an ontology of enduring substances.
They say that an account on which the substance that has the power is
the cause of its manifestations does not tell ‘a precise enough story. It is
a something about the substance that does its causal work’ (2, empha-
sis in original). Their point here is obscure. No one who embraces any
form of neo-Aristotelian ontology featuring causal powers will deny that
something about substances (i.e., their specifc causal powers) is central to
accounting for what they do: it is built into the very idea of substances as
possessing distinctive powers to act in some ways and not in others. But
in the particular framework of enduring objects/substances whose prop-
erties/powers are ontologically dependent on their bearers (in contrast
to both bundle-theoretic and four-dimensionalist frameworks), enduring
objects are the fundamental agents. In virtue of having the powers they
do, they are powerful in particular ways. And they exercise those powers
when in circumstances that stimulate, or remove obstacles, or enable,
their acting, or interacting. To suppose otherwise is to conceive objects
as mere skeletal arenas in and through which powers do their thing. We
are back to the ‘junior substance’ picture that A. J. Ayer rightly derided.
Summing the previous, I am led to the following understanding:
• objects (individually or collectively) cause the changing events that

unfold over time, manifesting their powerful qualities;
• the exercise of power is a real/ontologically fundamental, external,
and productive relation between an object(s) and an event(s); and
• there are, at least typically, necessary external and/or internal condi-
tions (events) for every manifestation.
These elements are, in abstract terms, the fundamental ontological reality

undergirding all true causal discourse.
The Ordinary Language of Causation

How one identifes what ‘causes’ what in everyday conversational con-
texts is actually another question, or set of questions, because ordinary
causal discourse – discourse that is not partly stipulative for specialized
theoretical purposes, as with fundamental metaphysics – is partly answer-
able to the purposes behind ordinary language usage. When philosophers
give metaphysical theories of ‘causation’, they are not theorizing about an
entity that is squarely in view in everyday discourse involving that term
(although it is taken to fgure centrally in the grounding of such discourse
where true). There is often confusion on this point in much discussion. We
need consider only the fact that it is perfectly unproblematic to say that
my failing to water my neighbor’s plant caused it to die, but no powers
theorists will allow that absences exercise causal power, and such every-
day statements are surely not evidence that most people are implicitly
committed to the denial of any metaphysics that embraces causal powers.
As with other matters, so here we may plausibly suppose that untutored
thinkers’ usage only imperfectly tracks the features of fundamental reality
that ground the truths of causal dependency that they express.14
Something similar applies even to the more rigorous usage of scientifc
discourse. Powers theorist Helen Steward (2012) helpfully suggests that,
insofar as we are trying to be faithful to the full range of theoretical con-
cerns driving scientifc investigation, we should recognize that there is no
single type of relation that is the causal relation (210) for such purposes. She
proposes a three-fold division of the kinds of causes of interest, as follows:
movers: substances, collections of substances, or other things that

act (e.g., felds)
makers-happen: events that trigger substances into action
(I take it that these are typically events of substances being ‘joined’
into confgurations, or coming to acquire new powers.)
matterers: facts that are causally relevant to other facts
(These capture general features that are invariant when certain
types of efects are produced, mapping type-level property
co-variations.)
All of these foci are of practical and scientifc importance, answering to

distinct concerns. And once one appreciates the ways that everyday causal
assertions, scientifc analysis, and metaphysical analysis both overlap and
diverge, the path is clear to embrace the propriety of event-, fact-, and
substance-causal statements, regardless of one’s metaphysical views con-
cerning the fundamental dynamical structure of reality. All such discourse
can be grounded in a causal powers ontology such as I have proposed.
IV. Free Will

‘Free will’ refers to an apparent power of human beings to control our
own intentional actions through deliberate choice, in such a way that
we may properly be held morally responsible for those actions and their
foreseeable consequences. Human beings are composed of several levels
of nested parts that are integrated in a structured hierarchy, an organic
(and thus necessarily dynamic) unity. So in asking how this power is best
characterized, we need to consider in general terms the nature of macro-
scopic powers – the powers of composed objects.
Many composite powers are mere ‘resultants’ of the powers of an
object’s parts, with mass being perhaps the most straightforward example.
The mass of a composite is the sum of the masses of each of its parts at any
level of decomposition. A composite’s exercising this power in a particular
way in a particular context and on a particular occasion ‘consists in’ the
individual exercises of power of each of its mass-bearing parts. There
is a subtle and vexed issue lurking in the phrase that I’ve put in quotes:
is this a plural identity, or an intimate-but-not-quite-identity relation of
‘constitution’, or some third such? Fortunately, we need not consider that
issue here. It’s plainly the case that if human choice stands to the activity
of neurons in executive control regions of the brain in the way that the
exercise of the macroscopic power of mass stands to the masses of the
composite’s parts, then human beings lack free will. The exercise of free
will, whatever else it might be, cannot be a merely additive resultant of
the exercise of a person’s unthinking and unwilling parts, as it would then
fail to be a distinctive kind of power.
What alternatives are there, for the case of conscious choice, to the aus-
terely reductionist model of composite power that mass exemplifes? For
more than a century, the preferred English term for what we are after has
been ‘emergence’. The term is meant to signify a macroscopic-microscopic
relation that involves both dependence and (distinguishing it from mere
resultancy) autonomy. The task of spelling out each of these constituent
features has proved challenging as well as controversial, and scientists and
philosophers have done so in a wide variety of ways.15 But in recent years,
a philosophical consensus of sorts is emerging that theories of emergence
may be usefully grouped into one or the other of two broad families, often
dubbed ‘weak’ and ‘strong’. Composite powers are weakly emergent just
in case (i) they exhibit some interesting kind of ‘autonomy’ (perhaps non-
aggregativity, or multiple realizability, or distinctive efcacy) and (ii) in
both existence and exercise, they are asymmetrically wholly determined
by the powers and arrangements of their fundamental parts, consistent
with the ‘causal closure’ of the fundamental physical realm.
Which specifc account of weak emergence best characterizes real-world
phenomena is controversial, but the existence of phenomena needing such
characterization is not.16 It enshrines the central physicalist commitment
of ontological reductionism, while eliding the latter’s ‘nothing-buttery’
excesses by afrming the reality of composite powers that harness micro-
scopic powers in ways that result in distinctive causal patterns. (It is dis-
puted whether there is a place for irreducible weakly emergent powers
within a sparse-property framework, such as the current one, but I will not
press that concern here.)17 Nonetheless, weak emergence seems insufcient
as the basis for free will. If the full panoply of microscopic powers nec-
essarily and asymmetrically determine all macroscopic powers and their
exercise, then what our future actions shall be is settled fundamentally by
nonrational and nonvolitional forces. (Determination ‘from below’ is as
much an obstacle to freedom as is determination ‘from before’, i.e., causal
determinism). What free will requires is of course as controversial as any
philosophical question. Views leaning towards the more defationary end of
the spectrum can accommodate its being a merely weakly emergent power.
My aim is not to defend the need for a more robust construal but rather
to elucidate my own commitments within the neo-Aristotelian framework
that is partially delineated earlier. As I maintain that free will does indeed
require that through conscious choice we sometimes are fundamental, as
opposed to constitutively derivative, diference-makers, I need recourse to
the general concept of strong emergence.18
A composite object’s power is strongly emergent only if it is funda-
mental, not being constituted or ‘realized’ by any structure consisting in
the powers of its parts and their arrangement. This is its understanding
of emergent autonomy. Given this, it is natural within our framework to
characterize the other, dependence condition in causal terms: the com-
posites parts, or some subset thereof, suitably arranged, jointly cause and
sustain the existence of the fundamental power of the whole, even as the
exercise of that power in turn has ‘downward’ efects on the composite’s
components.19
Along with some other theorists of action, I recognize a distinctive kind
of occurrent mental state that is termed an ‘intention’.20 It’s a state that
strongly disposes the agent to a course of action (where actively formed,
as in decision, its onset resolves prior uncertainty about what to do). It
represents an action-type and perhaps also the purpose behind it. And
ordinarily it is executive, leading directly to the attempt to act and guiding
its completion via perceptual feedback.
The power of choice, as I think of it, just is the power to form such an
intention. In Jacobs and O’Connor (2013), I endorsed a view on which
there was no general power to form choices, only a vast multiplicity of
specifc powers of choice, individuated by their intentional content (hav-
ing the form do X, do Y, etc. – or perhaps, in at least some cases, do X in
order to further goal G). The guiding thought was that, within a sparse
ontology, powers must always be fully determinate, never determinable.
Given the plausible assumption that one cannot choose at a given time to
perform an action for which one has no motivation whatsoever, or has
no capacity to represent (owing to limits of information or imagination),
it follows that one has a handful of specifc powers to choose at any time
from among the vast range of humanly possible such powers, and the
suite of powers one has is constantly changing over time with the change
of thoughts, desires, aims, and context. I now think that it was a mistake
to deny the ontological reality of an enduring, more general power of
choice in which are partly grounded the particular ‘all-in’ powers one
variably has on any given occasion to choose this or that. We cannot give
a satisfactory account of organisms (or any kind of complex system) with-
out recourse to such general capacities; any ‘sparse’ theory of immanent
properties must be structured to accommodate this explanatory need.
By its embrace of such a fundamental and open-ended power of choice,
my view can seem quite similar to those of a group of theorists of the will
that are commonly described as ‘noncausalists’, who also endorse a fun-
damental power of choice.21 We further agree that free will and determin-
ism are incompatible. Choosing freely entails that all the causal conditions
surrounding my choice were consistent with (at least) my intending to do
A and my not so intending, and frequently also with my intending to do
B or C, signifcantly distinct action types. This is of course consistent with
recognizing that there are enabling conditions that are necessary but not
sufcient for free choices (e.g., being aware of alternatives available to
me, desiring to resolve uncertainty, and being motivated towards certain
alternatives with variable degrees of strength).
But we characterize the power of choice diferently. As they see it,
the power of choice should be understood as a noncausal, ‘spontaneous’
power, whereas I understand it to be a specifc kind of causal power: a
power to cause an intention to act. (Furthermore, for some of them but
not for me, its exercise is by its very nature not causally determined by
antecedent factors.)
I have come to wonder just how substantive our disagreement is. I will
explore this question in relation to the account proposed by E. J. Lowe
(2008). (My earlier discussion of Lowe’s views reveals the key diferences
between his neo-Aristotelian framework and mine, which concern the
way powers immanently depend on substances and whether there is a
relation of causation. I believe that these diferences do not account for
the disagreement I am about to discuss.)
Here is the key passage explaining why Lowe believes the will (or
choice) is not a causal power:
A causal power is a power to cause some object to act in a certain

way. . . . The will, however, is not a causal power, inasmuch as its
exercise does not consist in the causing of some relevant kind of efect,
even though its exercise does normally have an efect of a predictable
kind, namely, an action-result appropriate to the kind of action that
the agent wills to perform. . . . [it] consists, in itself, merely in my
willing to do something.
(2008, 149–150, emphasis in original)
Lowe goes on to identify a distinct general category of power from causal

power, ‘spontaneous power’, which is neither a causal power nor a causal
liability and for which there are no deterministic or probabilistic causes
of its exercise. Alongside will itself, he includes the stochastic power of
radioactive atoms to decay.22
Lowe and I thus agree that when I freely form an intention, I exercise a
power such that the intention is essentially a manifestation of this power
and there is no prior set of conditions that trigger the manifestation to
occur just when it does. Now, contrary to what Lowe states, I would
have thought that its exercise does always result in a relevant kind of
efect, with the only salient diference being that the characteristic efect
(the onset of an intention) occurs within the acting substance rather
than an external object. (And we might say the same with regard to the
nonrational analogue of particle decay.) So why not just distinct species
of causal power? The distinctions between those powers that manifest
within and those which manifest without, and (what may be orthogonal)
between those that are necessarily ‘triggered’ into activity and those that
are not, are signifcant ones, but I don’t see in Lowe’s remarks a convinc-
ing basis for seeing either distinction as demarcating a line between causal
powers and another kind of power.
It may be that Lowe’s analysis is partly driven by one further diference
between our accounts of freely forming an intention (willing). I afrm,
while Lowe denies, that there are antecedent conditions or infuences that
make contemplated possible choices more or less probable until a choice
is made.23 As I see it, these prominently include motivational states, along-
side purely physiological internal states. Lowe, however, insists that we
cannot think of the infuence of reasons on practical deliberation in causal
terms. Given my own view, it is open to me to see the grasping of reasons
for action and the having of desires as acting in tandem with the general
capacity to form intentions, giving me ‘all-in’, more specifc powers to
intend to A, to intend to B, etc. Whereas on Lowe’s account, there is only
an open-ended power to form intentions. When he writes, ‘its exercise
does not consist in the causing of some relevant kind of efect,’ by ‘rel-
evant kind’ he might mean the very specifc efect of intending to A (and
not B). His view would then be that causal powers, indeterministic or not,
are directed at specifc efects, whereas the noncausal power that is will
is open-ended, potentially being manifested in any number of intentions
having very diferent contents.
The central drawback of the rejection of causal infuence on the will’s
operation by Lowe and other ‘noncausalists’ is intimated in my title: we,
with our various capacities that include will or choice, are embedded in
(and plausibly are composed of) a network of interacting causal agents.
And, taking a wider temporal view, the process by which there came to
be reasoning and willing agents such as ourselves was a very gradual
and incremental one. The strong emergence of a distinct category of
power whose operation is not fully integrated with those surrounding it
in its local region of the network is a priori implausible. Accordingly, the
burden is heavy for the philosopher who would argue that such a com-
mitment is necessary for a satisfactory account of our freedom.24
Notes
1. ‘Pared-down’ because they do not deploy Aristotle’s central doctrine of hylomor-
phism. There is a smaller but growing number of contemporary theorists who
argue that jettisoning hylomorphism was a mistake, including three contributors
to this volume: namely, William M. R. Simpson (Chapter 1), Robert C. Koons
(Chapter 2), and Christopher Hauser (Chapter 9). See note 4 for brief discussion
of hylomorphism’s relation to the alternative analysis of objects proposed here.
2. There has been signifcant debate concerning the adequacy and/or intelligibil-
ity of a view on which natural intrinsic properties are ‘pure powers’, such
that their dispositional tendencies exhaust their ‘natures’. I follow those who
maintain that such properties are at once qualities – characters that make for
objective, intrinsic similarity of objects – and dispositions towards efects.
(See, e.g., Martin 2008; Jacobs 2011; Heil 2012; Corry 2019; and Williams
2019, §5.3.)
3. I don’t want to be diverted into a full-scale defense of the preferability of
tropes, of which I have become convinced only recently. Armstrong is moti-
vated not only by theoretical economy (he suggests that embracing univer-
sals allows us to analyze both exact and partial qualitative resemblance) but
also by the needs of an austerely reductive naturalism that eschews abstracta
altogether. I take the latter project to be hopeless. If one accepts the need
for abstract universal concepts (or an equally non-immanent surrogate), the
small theoretical economy of immanent universals as against tropes appears
to me to be outweighed by its internal challenges that Armstrong admirably
acknowledges but fails to overcome. That said, if one brackets such chal-
lenges, much of what I will go on to propose could be paralleled within an
immanent universals theory.
4. This view bears a family resemblance to Aristotle’s hylomorphism, a structured
view in which objects are said to be matter-form compounds, and Aristotle’s
distinctive theoretical concept of matter plays a somewhat similar role to indi-
vidual substrata. Aristotle’s view is complex and notoriously difcult to inter-
pret fully, and I shall not attempt to make a detailed comparison here. For a
helpful recent analysis, see Koons (unpublished ms.). For recent arguments that
hylomorphism is made plausible (somewhat ironically) by the advent of quan-
tum theory in physics, see Koons (2014) and both William M. R. Simpson’s
(Chapter 1) and Robert C. Koons’s (Chapter 2) contributions to this volume.
Finally, for trenchant discussion of the diferences between a hylomorphic view
and the sort advanced later on the nature of human beings, see Koons (2018).
I recognize that hylomorphism has certain theoretical advantages, particularly
in handling the crucial concept of essence, which issue I gesture at in the fnal
paragraph of this section in the text. I hope to systematically compare substra-
tum-object and hylomorphic theory in future work.
5. Granted, the sortal concepts our theory permits us to apply to its posited
substrata are few and informationally ‘thin’. But the natural reply is that the
simplicity/thinness of the concept appropriately matches its job description
of mere individuation.
6. As Lewis (1983) observed in criticising the overweening ambition of Arm-
strong’s own substratum-attribute theory, which has perpetuated some of
the confusion. For just this reason, I reject David Armstrong’s (1997, 2004)
truthmaker-maximalist argument for states of afairs.
7. Cf. Sider (2006).
8. If one embraces a neo-Humean reduction of causation to facts concerning
recurring deep patterns of some kind in the spatiotemporal distribution of
qualities, then four-dimensionalism can allow for causal dependence of some
events on others. But since causation reduces to noncausal facts on this view,
it is not a fundamental feature of reality.
9. Sider (2011), 259f.
10. See Sullivan (2012) and Cameron (2015) for diferent ways of working out
such an account.
11. Dean Zimmerman (2011) proposes this as one of two options available to the
presentist for developing an account that is adequate to the empirical basis of
the theory as standardly interpreted.
12. ‘A manifestation [the efect] of this particular power could not exist in the
absence of this power, even though this particular power could exist in the
absence of this or any other manifestation of it. . . . We may say if we please
that the power and its manifestation stand in a ‘relation’ of asymmetrical exis-
tential dependency, but such ontological dependency facts do not involve real
relations, that is special relational universals or tropes’ (2016, 108, emphasis
in original).
13. Some philosophers seem to fnd implausible a productive, real-relation under-
standing of causation because they think it is wedded to a punctiliar picture
of causal interactions, whereas much of modern science describes continuous
processes. But (as many contemporary neo-Aristotelians emphasize) the basic
picture can easily be generalized, allowing for both episodic and continuous
manifestations of causal power.
14. Cf. Ted Sider (2011, 15–16 and 75–76). E. J. Lowe (2008, 142, 144) is getting
at a similar point, I think, when he suggests that event and substance causal
descriptions are both unproblematic, and we can even see how to translate
one way of speaking into another, while going on to contend that the most
perspicuous ontological analysis will identify substances as causes.
15. For a thorough summary and analysis of extant views, see O’Connor (2020).
16. The assumption that a single characterization is adequate to all weakly emer-
gent phenomena is common but nowhere adequately defended, and it seems
to me that it ought to be resisted.
17. See O’Connor (2000) §3.1 for an overview of arguments pro and con.
18. One can run an argument for the incompatibility of micro-determination and
freedom that parallels the much-discussed ‘consequence argument’ champi-
oned by van Inwagen (1983) and others for the incompatibility of causal
determinism and freedom. In my (2000), I defend a version of the latter
argument (Ch. 1) and the need for a strongly emergent understanding of the
power of free choice (Ch. 6).
19. The existence of fundamental powers implies a greater kind of substantial
unity to strongly emergent as against weakly emergent wholes. It plausibly
implies that such wholes have distinctive substrata. It is a challenging problem
to allow for such a fundamental substratum-of-the-whole melded to a suite
of emergent powers and a dynamically changing set of object-parts. There is
some theoretical pressure towards recognising an emergent individual entirely
distinct from, albeit causally dependent on, the underlying system of physi-
cal parts. O’Connor (2018) frames this problem without resolving it. The
signifcant and continuous causal dependencies on component-level processes
for emergent-level function provide, I believe, the basis for arguing for an
essential incompleteness of the emergent elements considered apart from the
body. But I haven’t the space to develop this suggestion here.
20. This notion and its relevance to free will has been extensively explored by Al
Mele; see, for example, Mele (2009).
21. Representatives include Ginet (2008), Goetz (1988), McCann (1998), and
Pink (2016). I discuss Lowe (2008) in the text since he develops his view
within a metaphysical framework similar to my own.
22. According to current theory, certain kinds of atoms have unstable nuclei
that ‘decay’ by emitting subatomic particles. The decay of individual atoms
is indeterministic and is largely impervious to its external conditions (while
there may be general necessary conditions at play, individual particle emis-
sions appear to lack stimulus conditions). For each type of radioactive sub-
stance, there is a general rate of decay for an aggregate sample of it that is
indicated by its half-life, which specifes a time interval by which half of the
sample will have decayed. But there is no recurring pattern in the timing of
individual emissions. (I note that there have been some recent claims of evi-
dence that decay rates can be accelerated through energy excitation; whether
this proves correct or not, the stochasticity of individual emissions remains
unchallenged.)
23. This makes particle decay more disanalogous to free choices on my view.
24. This publication was made possible through the support of a joint grant
from the John Templeton Foundation and the Fetzer Institute. The opinions
expressed in this publication are those of the author(s) and do not necessarily
refect the views of the John Templeton Foundation or the Fetzer Institute.
References
Armstrong, D. M. (1978). Universals and Scientifc Realism, Vol. I: Nominalism
and Realism, and Vol. II: A Theory of Universals. New York: Cambridge Uni-
versity Press.
Armstrong, D. M. (1997). A World of States of Afairs. New York: Cambridge
University Press.
Armstrong, D. M. (2004). Truth and Truthmakers. New York: Cambridge Uni-
versity Press.
Cameron, Ross (2015). The Moving Spotlight: An Essay on Time and Ontology.
Corry, Richard (2019). Power and Infuence. Oxford: Oxford University Press.
Ginet, Carl (2008). “In Defense of a Non-Causal Account of Reasons Explana-
tions.” The Journal of Ethics 12: 229–237.
Goetz, Stewart (1988). “A Noncausal Theory of Agency.” Philosophy and Phe-
nomenological Research 49: 303–316.
Heil, John (2012). The Universe as We Find It. New York: Oxford University Press.
Heil, John (2016). “Causal Relations.” In The Metaphysics of Relations, edited
by A. Marmodoro and D. Yates, 127–137. Oxford: Oxford University Press.
Heil, John (2017). “Real Modalities.” In Causal Powers, edited by Jonathan
Jacobs, 90–104. Oxford: Oxford University Press.
Jacobs, Jonathan (2011). “Powerful Qualities, Not Pure Powers.” Monist 94:
81–102.
Jacobs, Jonathan and Timothy O’Connor (2013). “Agent Causation in a Neo-
Aristotelian Metaphysics.” In Mental Causation and Ontology, edited by R. D.
Ingthorsson and Sophie Gibb, 173–192. Oxford: Oxford University Press.
Koons, Robert (2014). “Staunch vs. Faint-Hearted Hylomorphism: Toward an
Aristotelian Account of Composition.” Res Philosophica 91: 1–27.
Koons, Robert (2018). “Against Emergent Individualism.” In The Blackwell
Companion to Substance Dualism, edited by J. Loose, A. Menuge, and J. P.
Moreland, 377–393. Oxford: Wiley Blackwell.
Koons, Robert (unpublished). “Sorting Out Aristotle: A Comparison of Four
Modern Models of Aristotelian Form.”
Lewis, David (1983). “New Work for a Theory of Universals.” Australasian Jour-
nal of Philosophy 61: 343–377.
Lowe, E. J. (2008). Personal Agency: The Metaphysics of Mind and Action.
Lowe, E. J. (2012). “A Neo-Aristotelian Substance Ontology: Neither Relational
nor Constituent.” In Contemporary Aristotelian Metaphysics, edited by T.
Tahko, 229–248. Cambridge: Cambridge University Press.
Lowe, E. J. (2016). “There Are (Probably) No Relations.” In The Metaphysics of
Relations, edited by A. Marmodoro and D. Yates, 100–112. Oxford: Oxford
University Press.
Marmodoro, Anna (2017). “Aristotelian Powers at Work: Reciprocity Without
Symmetry in Causations.” In Causal Powers, edited by Jonathan Jacobs,
57–76. Oxford: Oxford University Press.
Martin, C. B. (1993). “Power for Realists.” In Ontology, Causality and Mind –
Essays in Honour of D. M. Armstrong, edited by John Bacon et al., 175–186.
New York: Cambridge University Press.
Martin, C. B. (2008). The Mind in Nature. New York: Oxford University Press.
McCann, Hugh (1998). The Works of Agency. Ithaca: Cornell University Press.
Mele, Al (2009). Efective Intentions. New York: Oxford University Press.
Mumford, Stephen and Rani Anjum (2011). Getting Causes From Powers.
O’Connor, Timothy (2000). Persons and Causes: The Metaphysics of Free Will.
New York: Oxford University Press.
O’Connor, Timothy (2018). “For Emergent Individualism.” In The Blackwell
Companion to Substance Dualism, edited by J. Loose, A. Menuge, and J. P.
Moreland, 369–376. Oxford: Wiley Blackwell.
O’Connor, Timothy (2020). “Emergent Properties.” In The Stanford Encyclope-
dia of Philosophy (Fall 2020 Edition), edited by Edward N. Zalta. URL =
<https://plato.stanford.edu/archives/fall2020/entries/properties-emergent/>.
Pink, Thomas (2016). Self-Determination: The Ethics of Action, Vol. 1. Oxford:
Sider, Theodore (2006). “Bare Particulars.” Philosophical Perspectives 20(1):
387–397.
Sider, Theodore (2011). Writing the Book of the World. New York: Oxford Uni-
versity Press.
Simons, Peter (1994). “Particulars in Particular Clothing Three Trope Theories of
Substance.” Philosophy and Phenomenological Research 54(3): 553–575.
Steward, Helen (2012). A Metaphysics for Freedom. Oxford: Oxford University
Press.
Sullivan, Meghan (2012). “The Minimal A: Theory.” Philosophical Studies 158:
149–174.
Williams, Neil (2019). The Powers Metaphysic. Oxford: Oxford University Press.
Zimmerman, Dean (2011). “Presentism and the Space-Time Manifold.” In The
Oxford Handbook of Philosophy of Time, edited by Craig Callender, 163–246.
6 Animal Powers and Human
Agency
Janice Tzuling Chik
All living things both move and are moved for the sake of something, so
that this is the limit of all their movement – that for the sake of which.
Now we see that the animal is moved by intellect, imagination, purpose,
wish, and appetite. And all these are reducible to thought and desire.
(Aristotle, De Motu Animalium, VII, 700b 15–18)
1. Introduction
Does the nature of an agent matter for a theory of action? For instance,
if animalism – the view that human beings are essentially animals (or,
more strongly, that we could not exist except as animals) – is true, does
its truth determine the acceptability of the standard account of action
(the causal theory of action, or causalism)? This chapter argues in the
afrmative. Instead of structuring its explanation of action around states
and events, an animalist-centred account of agents and agency insists on
the signifcance of the whole animal as the agent of what happens. This
chapter explains why the latter is incompatible with causalism.
Historically, Aristotle’s contribution to this conception of agents and
agency is unrivalled.1 For him, human action essentially refers to ani-
mal powers, for movement is linked to sense perception, and both are
related to animal desire, which ‘moves the animal’. This thesis in general
has resulted in contentious interpretations, however. In particular, the
Aristotelian view is cited in favour of the causal theory of action, on
which action is paradigmatically rational: by this causalists mean that
it is an event caused by certain necessary and sufcient mental anteced-
ents. Insofar as this chapter is concerned with an Aristotelian account of
action as performed by animal agents, its frst task will be to show that
such an account is incompatible with the standard account. Contra its
proponents, the causal theory cannot accommodate Aristotle’s animalist
paradigm of agency. This is because the causalist assumes event causation
as the paradigm for explaining how desire and belief motivate action. But
an Aristotelian account is neither event causationist nor intellectualist in
DOI: 10.4324/9781003125860-9
170 Janice Tzuling Chik
its conception of agency. The causal theory’s intellectualist approach, and
its particular conditions of rational mind, excludes non-rational animals
from an account of who counts as an agent, resulting in an anti-natural-
istic narrative.
Central to this problem is the post-Cartesian framework presumed by
many contemporary action theorists. If agents are analytically divided
as constituents of events or facts of a mental type and constituents of
events or facts of a physical type, then the problem of causal exclusion
persists, that is, of how the efcacy of mental events can be established
in physical domains. A metaphysical gap appears as well in the post-
Cartesian concept of action, between the latter’s causal relata, between
agential consciousness and its bodily efects, and no apparent way of
reconciling them beyond the causalist’s requirement that the latter be
‘caused in the right way’ by the former. It is furthermore impossible,
elaborating on this account, to preserve a concept of non-human animal
agents, if the causalist takes up Davidson’s argument that the capacity
for articulating reasons provides the only basis for identifying one’s
movements as rational and therefore intentional.2 Even intelligent
animals are therefore incapable of action, for this account requires
propositional speech as an indicator of mental content, separated from
the mere bodily movements that are undeniably shared by beings both
rational and non-rational.
In contrast, this chapter argues that action is a species of movement,
not strictly identical with a bodily event, that is distinctive of a living
animal. It advances a contemporary interpretation of Aristotle as ofer-
ing a disjunctivist account of action, based on the idea that action is a
psychophysical process: the actualization of a power that is essentially
both physical and psychological, simultaneously and irreducibly. That
intentional phenomena should exhibit such psychophysicality is an unsur-
prising and widespread feature of the natural world, given the diversity
of its denizens’ many agents, who are themselves, as animals, psycho-
physical creatures. In other words, consideration of the concept of animal
agents makes it entirely unproblematic that the concept of action should
be understood as a psychophysical phenomenon.
Although seemingly obvious, the thesis that action is a kind of move-
ment belonging to a living animal has been rejected in a surprising variety
of ways. Rejection of the thesis may be summarized as adopting one of
two general approaches. Reductionists claim that action consists in, no
more and no less, the movement of molecular particles and quantum
entities that materially constitute the animal agent. Others deny that
any animals other than human agents are able to perform actions, for
reasons having to do with their conception of practical rationality. The
latter group can be identifed as ‘intellectualists’ about action. Despite
their diferent approaches, both reductionists and intellectualists about
action fall into a causalist category of understanding agency: they both
Animal Powers and Human Agency 171
assume that there is a common physical element shared by both actions
and non-actions, that is, a common physical element that either makes
up the essential constituent of action (as is claimed by the reductionists
about action) or is rejected as the distinguishing feature of action over
mere movement (as is claimed by intellectualists about action). Between
the two general approaches, the notion of action as a causal relation is
cast to extremes, of a certain kind. On one extreme end, action under-
stood as a type of causation between molecules, events, and facts (and
perhaps also properties, depending on how we understand them) assumes
a notion of causation that is universally applied to all natural phenom-
ena, whether they are agential occurrences or not. At the other extreme,
causation of action is viewed as unique and radically diferent from all
other causal types, where all of these non-agential types are potentially
grouped together into a single overarching type: namely, that of event or
fact causation.
Some varieties of agent causation, such as accounts advanced by
O’Connor (2000) and Lowe (2010), defend some form of substance
dualism that treats the agent as an exceptional cause within the scheme
of natural events. This problematic view need not be associated with
an account of agent cause, however, as some have assumed (Bratman
2000). Merely positing an agent or even substance cause does not entail
a Cartesian structure of causation, according to which such a cause is a
purely mental res cogitans, with fundamentally physical efects manifested
only in the res extensa. The Aristotelian approach to action considered
in this chapter is compatible with certain varieties of agent or substance
causation, in which Cartesian schemas are not assumed. This is because
the agent is an animal: a living, sensing, and motile organism possessed of
capacities that are simultaneously psychological and physical.
2. The Causal Theory’s Aristotelian Heritage?
2.1. Contours of the Causal Theory

The causal theory defnes action as the event of a physical movement (i)
that is caused in the right way by another event or state, and (ii) whose
cause is, specifcally, something mental or psychological.3 Causalists
adopt a variety of combinations involving these two elements: some are
not event theorists, and thus embrace (ii) but not (i), while others accept
both.4 However broadly understood, causalism appears to have the fol-
lowing main components or conditions:
c1. A causal relation between an agent’s reasons or intentions5 and

his intentional movements;
c2. Analysis of reasons6 in terms of compound psychological states,
for example, beliefs and desires, and
c3. Action identifed as the physical movements caused by the psy-
chological states, which constitute the agent’s reasons for acting,
and without which such movements would not count as action.
We may also take Aguilar and Buckaref’s defnition of the causal theory
of action as standard:
(CTA) Any behavioral event A of an agent S is an action if and only

if S’s A-ing is caused in the right way and causally explained by some
appropriate nonactional mental item(s) that mediate or constitute S’s
reasons for A-ing.
(2010: 1)
Some causalists are also infuential interpreters of Aristotle, including

the earlier David Charles (1984), Jennifer Hornsby (2011), Alfred Mele,
and, earliest and arguably most infuentially among them, Donald David-
son. Some causal theorists explicitly argue that Aristotle defended a causal
theory of action, while others take the weaker position of suggesting that
Aristotle serves as one inspiration for modern causalism and thus can
be named as an ally of causalist approaches. Aguilar and Buckaref, for
example, describe ‘Aristotle’s account of the etiology of action [as] inher-
ited by later causal theories of action’ and claim that ‘to the extent that
Aristotle had a theory of action, his theory is clearly a progenitor of the
CTA’ (2010: 3–4).
Davidson often regards action causalism as standing substantially in
consonance with an Aristotelian approach to agency.7 Given that David-
son’s aim is to establish a causal relation between reasons – defned as the
mental states they comprise – and the physical movements they produce,
we may fnd surprising the association between Aristotle and his particu-
lar causal approach, for there is no evidence that Aristotle specifcally
conceived of reasons in terms of mental states or events.8 Nevertheless,
Davidson writes, ‘[T]he best argument for a scheme like Aristotle’s is
that it alone promises to give an account of the “mysterious connection”
between reasons and actions’ (1963: 693). On Davidson’s interpretation,
the causal relation promised by Aristotle’s account specifcally involves
the concept of a reason which, as the cause of action, can be broken down
into compound mental states, a belief and a desire (or, in lieu of ‘desire’,
what Davidson calls a ‘pro-attitude’).
2.2. Coope
What then is the evidence for thinking that Aristotle inspired or was even
himself committed to aspects of causalism, either in its totality or at least in
part? Is (CTA) in fact a neo-Aristotelian account? In what follows, I assess
an argument by Ursula Coope (2007) in support of the causal theory’s
core commitment that action is identifed with a bodily movement (c3) or,
according to (CTA), that it is a ‘behavioral event A of an agent S’.
Wittgenstein’s provocative query in the Philosophical Investigations is
now a standard starting point for contemporary philosophy of action:
‘What is left over if I subtract the fact that my arm goes up from the fact
that I raise my arm?’9 For the causalist, the question rightly suggests that
action requires the causal combination of certain elements: for any move-
ment to qualify as action, it must be caused by the appropriate mental
events or states. In other words, something more must be added to mere
movement (or a ‘bare movement’), for the latter to count as an action,
such as my raising my arm. Thus the causalist’s reply to Wittgenstein
would be: what is left over from the conceptual subtraction is none other
than the requisite mental states or events which stand in the proper causal
relation to one’s bodily movement.
Coope’s primary aim is to show how Aristotle’s view departs from the
standard causal view; she also argues that Aristotle and the causalist share
a commitment to identifying movements with actions:
The standard causal account, like Aristotle’s, identifes my action of

moving my body with my body’s movements. So according to both
accounts, when I raise my arm, my arm’s going up is the same as the
event of my raising my arm. Moreover, Aristotle, like the proponents
of the standard account, holds that human and animal actions must
be caused, in some sense, by a combination of the agent’s desire and
the agent’s being in some cognitive relation to the object of this desire.
(2007: 110)
On Coope’s interpretation, Aristotle broadly agrees with the causalist

as regards the claim in (c1): that there is some causal connection between
movement and psychological states. Aristotle also afrms (c2), construed
generally: that the psychological states in question are a combination of
the cognitive and conative elements in sensing and thinking creatures.
Insofar as such states can be implicated in explanations of the agent’s
movements, it is even possible to construe Aristotle as agreeing that such
states constitute reasons for acting. Finally, Aristotle’s view, according
to Coope, also partially afrms (c3), which claims an identity relation
between action and bodily movements.
Coope is not uncritical of the causal theory, and she ultimately endeav-
ours to ofer an alternative to it: like many of its critics (Velleman, Horn-
sby, Lavin), Coope accuses (CTA) of perpetuating a problem of ‘the
disappearing agent’. This problem arises from the causalist’s reductive
concept of agency, which posits,
when I raise my arm, all that happens is that my beliefs and desires
cause my arm to go up. But in that case, so the criticism goes, I don’t
really do anything: I am merely an arena in which some events (my
beliefs and desires) cause others (the movements of my body). What
the standard account fails to recognize (on this view) is that an action
is an agent’s bringing something about.
(2007: 110–111, emphasis in original)
Aristotle’s non-reductive view of action avoids this difculty, for ‘he

does not share the assumption that the only things that do any causal
work are events and states’ (2007: 111). Rather, an event is an action if it
is the exercise of some kind of causal power. Coope wishes to retain the
general intuition that action is ‘a causing of something’, while rejecting the
causal theorist’s stipulation that the causation must be a relation involving
a specifc conception of beliefs and desires, where the occurrence of such
mental states constitutes a necessary and sufcient condition10 by which
an event counts as an action (2007: 110). She proposes the following
Aristotelian account of agency:
P1: An agent’s action of moving her body is identical to her body’s

movement11 (e.g., her moving her arm is identical with her arm’s
rising).
P2: A change counts as an action if and only if it is the exercise of
a certain kind of causal power (e.g., an agent’s arm going up
counts as an action, viz., her raising her arm, only because it is
the exercise of one of her causal powers).
Coope argues that P1 and P2, taken together, avoid the problem of dis-
appearing agency. They also amount to a position sustained by Aristotle,
or one that he might have accepted. On Coope’s Aristotelian account, an
action is a change brought about only by an agent, and contra (CTA), is
not reducible to the causation of one event by another. Since an agent’s
raising her arm is not something that can be reduced to a set of mental or
psychological events that cause her arm to rise, the concept of action as
belonging to an agent is thereby preserved in the claim that to act just is
to move one’s body (P1). Such movement is the exercise of a causal power
of an agent, a phenomenon resistant to causal reduction. Granting P1 and
P2, Coope moves to defne action in the following terms:
P3: (Action as change) The exercise of a certain kind of causal

power of an agent, also understood as the partial fulflment of
that capacity (in Aristotelian terms, ‘the incomplete actuality of
the moveable’).
2.3. Challenges to the Interpretation

To what extent can Aristotelianism in fact be reconciled with the causal
theory of action? In this section, I articulate some concerns about this
prospect: while recognizing that Aristotelianism shows signifcant diver-
gence from (CTA), Coope’s account also allows for substantial overlap
with (CTA). Is this overlap justifed?
First, Coope acknowledges that Aristotle diverges from causalism in
aspects of (c3): according to (CTA), physical movements are identifed
as action in virtue of their being produced by beliefs and desires. The
primary support for this claim is De An. III.9–11, where Aristotle cites
both the causal contribution of desire and nous (which includes imagina-
tion) in the production of agential self-movement.12 As Coope argues,
‘though Aristotle thinks that human actions are explained by desires and
beliefs, it is not because they are explicable in this way that they count
as actions’ (110, emphasis in original). A correct reading of (P3) entails
not that movement is necessarily made into action by the psychological
states of the agent, but merely that action can be explained with reference
to such psychological facts; that action is thus explicable does not entail
that the occurrence of such psychological states makes it the case that a
movement is an action.13
Coope’s clarifcation amounts to a rejection of the causal theory alto-
gether, for it is (CTA)’s core thesis that mental states or events cause
intentional bodily movements, and as such are explained by them. Unsur-
prisingly, causalists have rallied behind Aristotle’s general statement that
‘every action must be due to a cause’ (Rhet. I).14 On Davidson’s view,
there can be no way to understand this thesis other than in the light of
NE VI.2: ‘An origin of action – its efcient, not its fnal cause – is choice,
and that of choice is desire and reasoning with a view to an end’ (1139a
31–32).15 Interpretation of the latter statement, in itself, is obviously not
constrained by any one of the propositions defning causalism, (c1)–(c3).
For causalists, claims concerning action’s ‘origin’ necessarily point to its
principal efcient cause, a choice made by the agent, understood in terms
of thought and desire; although, Davidson remarks, ‘it is generally oti-
ose to mention both’ (1963: 688). Coope’s clarifcation of (P3) therefore
amounts to a denial of the causalist’s core thesis, which requires explana-
tion of action in terms of beliefs and desires to provide a critical linkage
to reasons, qua the causal determinants of action itself.
Second, Coope’s insistence that action is identical with bodily movement
is puzzling. The causalist’s solution to Wittgenstein’s question at PI §621
is that the rising of one’s arm is identical to one’s having raised it. But by
the causalist’s own lights, this appears inconsistent. For there can be arm
risings that are not arm-raisings, and Wittgenstein’s provocative arithmetic
suggests that something extra must be added to mere movement to produce
action: according to (CTA), such is the case wherever the necessary mental
states or events are absent. So how can it be that action is identical with
bodily movement, if something else must be added to the bodily movement
in order to qualify it as action? There seem to be two possibilities here. First,
we may conclude (as I propose later) that it cannot be that a mere rising of
the arm is identical to the agent’s having raised it, for even on the causalist’s
account, the necessary and relevant mental states or events must be added
to the mere movements to produce actions, such as someone’s raising her
arm. Alternatively, identity between action and a bodily movement can
be retained only on the basis that the causal antecedents of action are not
counted as constitutive parts of it: for if they did, it could not be claimed
simpliciter that action is identical with a bodily movement.
The latter alternative preserves identity between action and bodily
movement, but at the cost of excluding anything psychological from
action itself: the psyche and its powers are merely external antecedents,
which may qualify a movement etiologically but do not in themselves
serve as the constituents of action. But this implication bucks our most
common-sense intuitions about intentional action being something intrin-
sically psychological. Particularly for critics of causalism worried about
the problem of disappearing agency, such an account poses a particularly
worrisome variant of disappearing agency, for here we are faced with
the prospect of a disappearing psyche. Without the psychological, what
remains that is still agential?
Finally, Coope’s causalist interpretation of Aristotle leaves unanswered
the crucial question: what explains the basic idea that desire and thought
may explicate action, let alone play a causal role? In the next part, I argue
that it is only the fact that the agent is an animal, possessor of powers
that are at once psychological and physical, and relating specifcally to
purposive locomotion. Beginning instead with this Aristotelian paradigm,
we can resolve the difculties raised in this part – or prevent them from
arising in the frst place.
3. Action as Psychophysical Process

I defend a diferent Aristotelian account of agency, on which mere bodily
movement is entirely unlike an action: contra the interpretation consid-
ered earlier in §2.2, no identity relation necessarily holds between an
action and a mere bodily movement. This account defends their distinct
causal characters and diferentiates their metaphysical constitutions,
implying a certain kind of disjunctivist argument: action and movement
are distinct phenomena, with no common element or causal factor shared
between them.16 This is because action is a psychophysical process, and
mere bodily movement is not. This account is able to explain why, on all
Aristotelian theories (including those appropriated by causalists), desire
and judgment are thought to explicate action but not bodily movement
simpliciter. The conventional explanans provided are not mental events
pertaining exclusively to rational creatures like ourselves. Rather, they
specify the essential character of an animal, whose sensitive soul defnes
it paradigmatically as an agent.
The causes of animal movement are many, according to Aristotle in De
An. III.10, and although those cited by the causalist may be among these,
neither desire nor deliberation are the frst cause, but rather ‘the object of
appetite (for this, though unmoved, causes movement by being thought of
or imagined)’ (433b 11 f). Indeed, relative to the truly primary cause that
is the agent’s desired object, desire and belief play more of an intermediate
causal role; Aristotle describes desire as ‘that which both moves and is
moved’, that is, a moved mover. The agent’s object of desire, in contrast, is
an uncaused cause, and therefore primary. Davidson’s appropriation of NE
VI.2, then, is a gross oversimplifcation of Aristotle’s thought.17 If ‘reasons
for acting’ are defned exclusively in terms of psychological states qua the
mental causes of action, we neglect the possibility that an agent may cite
a reason for doing P as ‘for the sake of’ achieving Q, or ‘in order to’ do
Q – and that such an expression in fact reveals the metaphysical structure of
many actions.18 In citing reasons that express purposiveness or fnal causal-
ity, agents themselves articulate a non-causalist structure of action wherein
one’s physical movements are not efciently produced by mental states.
Where, by nature, some agents cannot cite reasons, we have little reason to
doubt that their intentional movements may be any less forward-looking.
3.1. Charles
The evident neglect of De Anima 3.10 by action theorists is redressed by
David Charles, whose aim is to develop Aristotle’s statement that ‘desire
moves the animal’ (433b 13f). Charles’s principal claim is that desire,
along with other afections such as fear and anger, is ‘treated by Aristotle
as inseparable in defnition as well as in existence from states of the body’
(2011: 77). In the case of anger, there is a type of desire specifcally for
revenge, understood in accordance with the principal thesis: ‘[T]he type
of desire for revenge which defnes anger cannot be defned without refer-
ence to certain physiological processes’, which according to Aristotle, is
constituted by a ‘boiling of the blood’ (2011: 77). Charles explains, ‘The
[physiological processes] play a role in determining the type of desire that
is present in anger. Desire of this type is an inextricably and essentially
psychophysical process. It is a boiling-of-the-blood-type of desire for
revenge’ (2011: 77). He also adds a further clause to his principal claim:
‘The type of desire for revenge which the angry experience resists decom-
position into distinct physical and psychological components’ (2011: 77).
Charles’s argument, that desire is ‘an inextricably and essentially psy-
chophysical process’, is based on several key passages from De Anima.19
The frst includes Aristotle’s elaboration on his comment that ‘desire
moves the animal’:
De Anima 3.10: ‘The instrument by which desire moves the animal is

a bodily one: this is why it must be investigated among the functions
common to body and soul.’
(433b 19–20)
Charles also considers De An. 1.1:20
De An. 1.1(a): ‘It is clear that the emotions are enmattered formulae
and so their defnitions will be of the following form: to be angry is a
process of this type of body or part or capacity of such a body caused
in this way for the sake of such and such a goal.’
(403a 24–7)
De An. 1.1(b): ‘To be angry is a given type of process, the boiling of

the blood around the heart for the sake of revenge.’
(403a 31)
De An. 1.1(c): ‘No one considers the properties of matter which are
inseparable [from this type of body/matter] not that is as properties
separable [from this type of body/matter] but the physicist considers
all the deeds and properties of this type of body and matter of this
type.’
(403b 10–12)
For Aristotle, an afection ‘is essentially enmattered because its form is

an enmattered form: one which is to be understood as essentially enmat-
tered in this type of physical process’ (2011: 79). Furthermore, desire on
Aristotle’s view ‘resists decomposition into distinct physical and psycho-
logical components’:21
De An. 1.1(d): ‘We have said that the afections of the soul are insepa-
rable from the physical matter of living beings in the way in which
anger and fear are inseparable and not in the way in which line and
plane are.’
(403b 17–19)
Next, Charles adapts this hylomorphic analysis to the case of action

and activity, referring to Aristotle’s example of weaving in De An. I.4,
408b 11–13. Weaving is an instance in which an agent intentionally
moves her body ‘to achieve a given goal’ (2011: 81). It is a process that
is characterized by both psychological features and physical properties,
and ‘[t]he psychological features that are essential to weaving . . . are
inseparable in defnition from the processes with physical properties to
which they belong. They are inextricably psychophysical features’ (2011:
81). In addition, ‘[t]he psychophysical features (specifed in [the previous])
are essential to the identity of the processes to which they belong: the
processes to which the psychophysical features belong are essentially psy-
chophysical processes (i.e. weaving). . . . There is no other process (other
than the one specifed [as psychophysical]) which is essential to weaving’
(2011: 81). This defnition of action is supported by Aristotle’s remark
in De Generatione Animalium, that ‘the movements of the instruments
employed [by the craftsman] contain the formula [logos] of the craft’
(744b 32 f). Both the craftsman and the weaver are engaged in activity
whose movements contain the ‘formula of their craft’. Concerning the
craftsman in De Gen. An., Charles elaborates:
his movements contain the formula (logos) of the art in that one is not
able to defne them (say what these processes are) without reference to
the goals of the craftsman and his/her skill. While they are essentially
spatial, there is no defning what they are without reference to the
goals and know-how of the craftsman. The relevant movements are
the ones which are guided in the appropriate way by his/her skill. In
the earlier case, the weaver spins, cards the wool, and ties the knots in
the way required to make the coat she aims to produce. Her actions
are vivid examples of inextricably psychophysical processes.
(2011: 87)22
It may be thought that the weaving machine itself possesses a structure

that is psychophysical: it is conceptually defned as a physical instrument
for the sake of spinning thread. Although Charles extends the concept of
psychophysicality to objects such as Aristotle’s ‘snub’ (Physics II. 2, 194a
5–8) and a house (403b 4–6), it seems more appropriate to say that these
examples exhibit ‘form-in-matter’ or ‘hylomorphism’ as a more generic
category, since psychophysicality is a characteristic of living substances.
Aristotle defnes ‘soul’ as the frst actuality of a living organic body, where
the sense of ‘organon’ corresponds closely to ‘tool’ or ‘instrument’: one
used naturally for realizing the animal’s vital activities (De An. I).23 Like
a weaving machine, an animal’s body, qua corporeal instrument with
specifc purposes and functions, is governed by psychological powers and
processes; unlike a machine being manipulated by a weaver, an animal
body is governed not by an external psychological agent, but by the very
agent whose body it is and which essentially constitutes the agent herself.
The kind of desire that moves the animal is essentially enmattered in a
specifc type of bodily process, which partially defnes the desire to move
as the kind of desire it is. Contra Menn (2002), it is not the soul ‘as user
of the body . . . separate from the body, standing to it as user to used’ that
moves the animal (2011: 91). On this account of action as psychophysical
process, Charles argues, it is ‘the person (or the composite), not the soul,
that moves his hand, grieves, and so on’, for ‘the composite . . . not the
soul, is the appropriate subject of the essentially psychophysical processes
that lead to action’ (2011: 92).24
This account unequivocally rejects the causalist premise that the causes
of action are mental or psychological states or events in the mind of
the agent, for whatever antecedents to action may occur, this Aristote-
lian account emphasizes that they are themselves psychophysical. And
although at times we may describe these, as Charles does earlier, as ‘the
essentially psychophysical processes that lead to action’, it may be that
desire and thought also are partly constitutive of the action itself, not
merely contemporaneous with its performance. Where the causes of
action are thought to be purely mental and action itself wholly physi-
cal, we should not be surprised to fnd a vitiating gap, metaphysical and
temporal, arise between them. However, where the causes of action, even
should they be conceived as events or states, are classifed as the very same
kind of phenomenon that action itself is, we should not fnd it implausible
that they are indeed parts of one and the same process.25 Sensitive desire,
as a power specifc to an animal, is not an event or state that ends when
the lioness seizes its quarry: the perceived edibility of the antelope is a
judgment that can only become more profound with desire that waxes
(and eventually wanes) with the temporal progression of her act of eating.
Desire evidently informs animal action; indeed, it forms an essential part
of it, and is what constitutes it qua something intentional.
3.2. Aristotelian Disjunctivism?

On the Aristotelian account of action presented, the distinction between
an action and an involuntary movement is straightforward: one is a
psychophysical process, and the other is not. Certain bodily powers
belonging to an animal, such as digestive, circulatory, and immunologi-
cal functions, are represented by purely physiological movements: they
are defned as such by explicitly denying them reference to anything
psychological. This is not because the owner of such processes cannot
possibly be aware of them: we may indeed take notice of the efects of
digestion or an immunological response in our bodies, as the weaver may
note that the machine tends to slacken the thread at certain moments as
she pedals. Such movements are purely physiological processes because
they occur outside of our agential powers and, for the most part, outside
of our awareness.
In contrast, voluntary and intentional action has psychological and
physical phenomena constituents ‘essentially inextricable’ from one
another, both in existence and in defnition. On this account, certain
physical events or processes involved in action cannot be defned without
reference to the psychological events or processes that characterize the
moments throughout which an agent desires to move and carries out her
desired task. The events, parts, and processes that essentially constitute
an action cannot be either wholly physical or wholly psychological, for
they are ‘essentially inextricable’ from one another in just the way defned.
Mere movement – an involuntary response or non-voluntary process of
the body – lacks the ‘essentially inextricable’ feature of action. Therefore,
there is no overlap between ‘mere movement’ and ‘action’, and we cannot
treat ‘action’ as if it is ‘mere movement’ qualifed by the psychological. It
is possible to advance this Aristotelian account of agency as a variety of
disjunctivism about action.
An advantage of this Aristotelian disjunctivist account lies in its ability
to resolve the causalist’s problem of deviant causal chains. Davidson’s
unfortunate climber loosens his hold on the rope because his very desire
and belief to do so absolutely unnerves him: this is an efect of the neces-
sary and sufcient causes stipulated by (CTA), but the causal relation
itself is not sound (the action is not ‘caused in the right way’). By itself,
even Davidson admits, (CTA) fails to provide an explanation for this dif-
fculty. An Aristotelian disjunctivist approach to agency, in contrast, is not
committed to mental antecedents as the necessary and sufcient causes of
action; as a result, it is able to analyze the unnerved climber’s letting go of
the rope as mere movement rather than action. According to this Aristote-
lian approach, the fact that what precipitates the climber’s movements are
certain mental or psychological states – and the fact that they correspond
precisely to the causalist’s criteria for the causes of action – is irrelevant
to what happens. The climber’s desire to lessen the weight of his burden
may itself be a psychophysical process, but it is one that instead produces
an involuntary physiological efect. Nothing psychophysical follows, and
indeed, the causes of what happens need not have been psychophysical
in the frst place (we can imagine him letting go of the rope having been
knocked unconscious by falling debris).
Disjunctivism itself comes in diferent varieties. Originally a strategy
for resolving problems of perceptual experience, its proponents diferen-
tiate between epistemological and metaphysical forms: that is, between
experiences that do or do not provide reasons for knowing, and between
experiences of essentially disparate kinds. Alex Byrne and Heather Logue
argue that the metaphysical variant ‘leads naturally if not inexorably to
[the epistemological variant]’, but not conversely.26 This seems especially
true for a disjunctivism about action, which may likewise be diferenti-
ated as epistemological (experiences that bear reasons for acting or not)
or metaphysical (experiences that are essentially of one kind or another);
one’s disjunctivist theory can evidently be both.27
The Aristotelian account of action as psychophysical process may be
construed as a metaphysical variant of disjunctivism capable of explaining
why actions are intentional (and, in the rational case, done for reasons)
and mere bodily movements are not. A standard way of articulating meta-
physical disjunctivism in general is J. M. Hinton’s proposal, in Experiences
(1973), that no common constituent or element is shared by the disjuncts.
As Byrne and Logue comment, ‘there is no “kind of experience common
and peculiar” to the good case and the bad cases – they have no “common
element”’ (emphasis in original).28 Others ofer variations on this thesis:29
McDowell (1983) denies a ‘highest common factor’ across disjuncts,30
while Snowdon (1980) denies any ‘single sort of state of afairs’ obtain-
ing in good and bad cases,31 and Martin (2004) asserts no ‘distinctive
mental event or state common to these various disjoint situations’.32 My
proposed metaphysical disjunctivism about action, in turn, argues that
no common factor or constituent exists between action and bodily move-
ment. Action disjunctivism is therefore incompatible with (CTA), which
assumes that an action is a bodily movement, caused in the right way by
mental states or events. According to the latter, action and movement are
metaphysically overlapping concepts: one is a causal constituent of the
other, a common element shared by both.
Precisely why disjunctivism forecloses this causalist picture has seemed
obscure to some critics. Jonathan Dancy accuses disjunctivism of making
‘no move beyond’ asserting metaphysical disjuncts, or the claim ‘that there
are broadly two types of experience. . . . We are told virtually nothing
about the natures of these diferent types of experiences’.33 If we return to
Coope’s Aristotelian account of agency, it may seem that Dancy’s critique
is vindicated. For Coope wishes to retain the causalist’s identity relation
between action and bodily movement, while also arguing that her raising
an arm is ‘of quite a diferent kind’ from the rising of an arm caused by
alien limb syndrome. She writes:
A movement of this kind [i.e., an involuntary arm movement] is not

the fulflment of a potential anything has to be in some particular end
state. If it is right that such accidental movements do not qualify as
Aristotelian changes, then this has an interesting consequence for his
views about action. On Aristotle’s view, there is not one type of event,
my arm’s going up, which could occur because I sufer from anarchic
hand syndrome. The event which occurs when I raise my arm is of
quite a diferent kind from the event that would occur if my arm
moved about because I sufered from this condition. The one event
is a change and hence a progression towards some end; the other is
not. According to Aristotle, if my arm’s rising is a change, it must
have some agent (though the agent could be someone other than me,
and it could, even, be something inanimate). On the other hand, if
it is not a change, then it is not even a candidate for being identical
with an action of the sort I have discussed in this paper: an action of
changing something.
(2007: 135–136, emphasis in original)
Initially, it may seem that Coope’s account coheres with disjunctivism

and that my critique in §2.3 was unjustifed: raising an arm voluntarily
and seeing it rise involuntarily are two diferent kinds of events, for the
frst is a change and the other is not. My concern is with how this apparent
disjunctivism is compatible with a causalist commitment to a common
constituent – that is, a bodily movement – shared by both action and mere
movement. For she argues, ‘the standard causal account, like Aristotle’s,
identifes my action of moving my body with my body’s movement. So
according to both accounts, when I raise my arm, my arm’s going up
is same as the event of my raising my arm’ (2007: 110). Plausibly, we
could identify ‘arm-risingv’ (a voluntary movement) versus ‘arm-risingi’
(an involuntary or non-voluntary movement). In Coope’s terms, arm-
risingv is an incomplete fulflment of an agent’s potential to raise her arm:
it is an event exhibiting end-directedness. Arm-rising i denotes a mere
bodily movement, one lacking in end-directedness, for it does not fulfl
any capacity or potential of an agent to move her body ‘towards some
end’. The problem is that causalism does not specify whether the body
movement identifed with action in Coope’s account refers to arm-risingv
or arm-risingi. If we say that action is identical with arm-risingv but not
arm-risingi, we apparently forfeit the causalist claim that action is bodily
movement qualifed by certain mental states or events. That is, we must
choose either
P1: An agent’s [action of moving her body]v is identical to [her

body’s movement]i: for example, [her moving her arm]v is iden-
tical with [her arm’s rising]i
or
P1*: An agent’s [action of moving her body]v is identical to [her

body’s movement]v: for example, [her moving her arm]v is iden-
tical with [her arm’s rising]v.
According to (CTA) in §2.1, what grounds the fact that the [action of
moving her body]v is a voluntary movement is that it ‘is caused in the
right way and causally explained by some appropriate nonactional mental
item(s)’. This is the case for either P1 or P1*. But what grounds the fact
that in P1* this voluntary action is identical to [a body movement]v rather
than [a body movement]i? If we say that it is the same ‘appropriate non-
actional mental item(s)’ cited by (CTA), which causally explains an action
and its [body movement]v alike, then it is (1) trivial to assert P1*, and (2)
the causal theory loses its primary explanatory power. Alternatively, if we
retain P1, we also retain all of the challenges raised by Davidson’s deviant
causal chains, and we are left with an explanatory puzzle about how an
action of arm-risingv is distinct from arm-risingi and yet also sometimes
identical to it. In other words, Coope’s account is susceptible to Dancy’s
critique that ‘it makes no move beyond’ merely stating that action is a
change and mere movement is not (1995: 435).
My disjunctivist proposal assumes a metaphysical commitment to a
concept of the agent as an animal, innately a possessor of powers that
range from sensing, desiring, aiming, and doing.34 This proposal is not
uncontroversial; it overturns the standard assumption that the concept of
action may be investigated independently of what constitutes the agent
herself. As Coope argues, in the earlier passage, ‘if my arm’s rising is a
change, it must have some agent (though the agent could be someone
other than me, and it could, even, be something inanimate).’ If action
is not something necessarily done by an agent understood as an animal,
specifed by capacities that are at once psychological and physical, then
surely it could be movement caused by something other than an agent,
even something that is non-agential itself. But then what happens would
not be action. The metaphysical ground on which my disjunctivist pro-
posal stands must be a consideration of the animal agent herself, who
senses, desires, forms cognitive judgments (and, in the human case, delib-
erates and decides), and moves to act. Where she plays no such role and
exercises none of the agential powers uniquely possessed by an animal,
that which happens cannot be described as action.
I will consider another objection to my disjunctivist account. It is evi-
dent that the Aristotelian appeal to action as purposive or end-directed
does not in fact rule out the possibility that purposive motion might be
partly constituted or produced by the very capacity by which one also
moves simpliciter, whether purposively or not. Does the account of action
as psychophysical process fare better against this problem? Certain func-
tions of an animal body, such as breathing or blinking,35 present potential
difculties for our disjunctivist account of action. These are movements of
the body that occur without any agential control or awareness, but which
can become agential should we choose to fx our attention and control
over their occurrence. If the psychophysical account of action is assumed,
one may object, the purported disjuncts of action versus bodily movement
appear overly severe: surely it appears to be the case that my controlled
breathing is still a bodily movement, qualifed now by my agency. Does
action disjunctivism not produce an odd picture, on which the expansion
of the lungs at t1 is mere movement, while the expansion of the lungs at
t2 (when I have become aware of my breathing) has nothing to do with
mere movement, being instead action? How can what patently appears
to be the same sort of phenomenon be rejected as such?
Plausibly, the disjunctivist may insist that what happens at t1 has no
common physical constituent with what happens at t2, given the tendency
of a psychophysical process to assume an appropriate physiological sub-
ject that corresponds to its specifc form. For instance, ‘boiling of the
blood’ (or high blood pressure, in contemporary terms) presumably also
occurs outside of one’s being angry, but anger assimilates into psycho-
physical phenomena what otherwise would have been a purely physiologi-
cal movement. Likewise, in the case of action, the very same muscles that
I use to raise my arm are otherwise constantly engaged in kinematic and
fexion activities of which I am wholly unconscious. Breathing and blink-
ing and other animal processes may seem dissimilar to these examples, in
that my coming to be aware of them entails that I do the very same thing
that my body otherwise would have done involuntarily. However, the
possibility of agential control in these cases only highlights the distinc-
tion of the psychophysical as integrating animal mind and body: it is not
merely that the body is governed by mind, but that the phenomenon pro-
duced by such governance is itself substantially altered from what before
might have been merely physiological. Agential blinking and breathing
are fundamentally distinct from their involuntary counterparts, as rais-
ing the arm is fundamentally distinct from an arm merely rising; these
diferences, furthermore, would no doubt be validated from an empiri-
cal standpoint.36 This account, by no means comprehensive, nonetheless
suggests a plausible alternative to (CTA) in metaphysically probing the
truism that in raising an arm, one’s arm rises: Aristotelian disjunctivism,
as I have proposed it, allows us to see how action and physiological occur-
rence are conceptually and analytically distinct, even when unifed as a
psychophysical compound.
3.3 Action Beyond Cartesianism

On Haddock and Macpherson’s analysis, disjunctivism fundamentally
challenges Cartesian conceptions of mind and body. The latter conceives
the agent as a possessor of subjective experiences, diferentiated according
to distinct internal and external contents:
The Cartesian picture of the mind understands the inner world as

constitutively independent of anything outer: it is possible that the
outer world difers radically, and the inner world remains exactly as
it is. There are various ways of understanding the notion of an inner
world. We might understand it as the idea of (amongst other things)
the subject’s experiences. . . . We might understand it as (simply) the
idea of a subject’s experientially acquired reasons – those reasons that
a subject enjoys when it looks to him as if things are a certain way.
And we might also understand it as the idea of the phenomenology
of experience.37
This constitutive independence of the mental from the physical frames

our conception of (CTA), according to which an action’s mental causes
are distinguished from the ‘behavioural event A of an agent S’, in a
relationship of cause to efect. According to its proponents, the kind of
cause at work cannot be part of its efects: otherwise, the cause would
be, absurdly, a cause of itself. Sarah Broadie’s critique of (CTA) is
that it ‘locates the causation in the wrong place. . . . For it places [the
causal relation] between events or states of afairs or facts of which [the
agent] is a constituent’.38 Her critique refers to a Cartesian ontologi-
cal partitioning between types of events or states, mental and physical,
but also implies that such a metaphysics treats the agent as an after-
thought, a mere logical constituent of the causal relation who is herself
ontologically divided. In other words, (CTA) re-presents a variant of
the mind-body problem not only for the concept of action, but for our
understanding of an agent as well.
The mind-body problem specifc to agency seems particularly intracta-
ble given the apparent impossibility of causal closure: how do we explain
the efcacy of mental events in purely physical domains? Perhaps not
every physical efect has a complete physical cause, in which case we
lack causal closure. The difculty can be resolved by positing that bodily
movements are caused by the physical ‘realizers’ of mental events, such
as neurophysiological events or states. But this kind of solution carries
a notorious cost: it literally excludes mental events, such as intentions,
desires, or reasons, from the account of action, and renders them epiphe-
nomenal – causally inefcacious – henceforth to be dismissed as artefacts
of ‘folk psychology’. But this appears to be a cost that relatively few
causalists are willing to accept.
In contrast, the disjunctivist theory of action advanced in this chap-
ter proposes an anti-Cartesian metaphysics of agency that avoids these
notorious challenges. Haddock and Macpherson propose, ‘[T]he mark of
disjunctivism, in all of its varieties, is a conception of the inner and the
outer as sufused.’39 The metaphor of sufusion is suggestive, ofering a
relation of direct access between perceiver and object, and between agent
and action; it forecloses the possibility of alienation and impermeability,
simply by rejecting a Cartesian account of mind and body:
We might understand the idea of the outer as consisting at least of

the subject’s bodily movements. We might understand it as (simply)
the idea of those of a subject’s reasons that concern the behaviour
of others – reasons that specifcally concern how things are with the
bodies of others. Metaphysical disjunctivism about the surface sees
some bodily movements as partly constituted by inner states. And its
epistemological variant sees some of the reasons that concern behav-
iour as also concerning the state of the other’s inner world (as when
subjects see that the other’s behaviour is expressing the fact that they
are in pain). We might think of disjunctivism about bodily movements
as an inverted, but structurally analogous, version of disjunctivism
about experience, because it asserts, not the sufusion of the inner by
the outer, but that of the outer by the inner. Put together, both kinds
of disjunctivism yield a conception of the inner and the outer as each
sufusing the other.40
I have advanced the Aristotelian account of action in §3.1 as an instan-

tiation of this metaphysical disjunctivism. Whether or not the historical
interpretation is persuasive, the anti-Cartesian theory of action ofered
promotes a unifed ontological arena in which intentional phenomena,
at once mental and physical, are natural properties of the animal whose
powers specify it uniquely as an agent in the world. This chapter’s pro-
posal may seem radical, as occupying a distinct alternative to contempo-
rary accounts of agency; as Charles remarks, Aristotle’s approach ‘shows
that there are more ways of understanding the mind-body relation than
are dreamed of in most post-Cartesian philosophy’ (2011: 93). If such
an approach is at all plausible, however, I have argued that we must also
reject the possibility that Aristotelianism might be claimed as progenitor
to the causal theory of action.
Notes
1. This chapter focuses on Aristotle’s contributions on the subject in Categories,
De Anima, De Generatione Animalium, De Motu Animalium, and Physics.
Other texts, not addressed closely here but relevant to the present inquiry
also, are Eudemian Ethics, Metaphysics, Nicomachean Ethics, and Rhetoric.
2. See especially Davidson’s ‘Actions, Reasons and Causes’ (1963) and ‘Agency’
(1971).
3. There are variations on the temporal structure of such causes to their efects:
some causalists believe that the cause must be antecedent to the physical
movement, while others believe it could be simultaneous. See Arthur Danto
(1963, 1965), George Wilson (1989), David Velleman (1992), Berent Enç
(2006), and Michael Smith (2012). For the present discussion, I will assume
that a general characterization of causalism sufces.
4. It should be noted that the acceptance of (ii) but not (i) is compatible with the
idea of agent causation.
5. Davidson, for instance, makes equivalent an agent’s primary reason and the
intention with which something was done: ‘To know a primary reason why
someone acted as he did is to know an intention with which the action was
done’ (1963: 689).
6. In light of the defnition of reasons ofered here, it should be clear that the
discussion henceforth focuses on motivating rather than normative reasons.
7. This is not to claim, of course, that Davidson saw all of his views on action
as corresponding to Aristotle’s: he disagreed radically with Aristotle over the
latter’s conception of akrasia, for instance.
8. As will be argued, Aristotle does refer to thought and desire as being involved
in movement, but the connection between thoughts and desires qua mental
states as constituting reasons for acting are tenuous. A historical grounds for
skepticism about the latter as an Aristotelian thesis is that Aristotle himself
employed no term that translates as ‘mental’.
9. Wittgenstein, Philosophical Investigations, §621.
10. The other condition concerns the causal relation between the movement(s)
and the mental states: the body movements must be ‘caused in the right way’
in order to count as action.
11. I critique this defnition below in §3.2. It appears that, for Coope, such move-
ments are by defnition intransitive.
12. He says at De An. III.9–11, ‘these two, appetite and practical thought, seem
reasonably considered as the producers of movement; for the object of appe-
tite produces movement, and therefore thought produces movement, because
the object of appetite is its beginning’ (433a 19–20).
13. The problem of deviant causal chains is notable for emphasising this point.
14. Aristotle’s statement is consistent with an agent-causal view, that is, one deny-
ing that the causal contribution towards an action is necessarily an event.
15. Many translations render this passage erroneously: ‘The origin of action – its
efcient, not its fnal cause – is choice’ (emphasis added). Aristotle’s text merely
says that choice (prohairesis) is an origin of action, among others that may not
include the deliberation characteristic of prohairesis. At NE III, 112a 14 f he
writes, ‘Not everything voluntary is chosen.’ On this basis one can understand
his view that non-rational animals and children can also be (voluntary) agents.
See also De An. III.10 for Aristotle’s more extensive argument that both appe-
tite and mind are capable (and necessary) for causing movement.
16. I must emphasize that this is only a certain kind of disjunctivist argument: it
appropriates elements from metaphysical disjunctivists such as Martin, rather
than epistemological disjunctivists such as McDowell, as will be explained
later. As disjunctivists comprise a large and diverse society of thinkers, it
must be said at the outset that not all disjunctivists (e.g. of a McDowellian
following) will be inclined towards the arguments advanced.
17. For Aristotle, appetite (epithumia) is but one of three types of desire (orexis).
The causalist interpretation does not take sufcient account of cases in which
the agent’s reason for acting in fact specifcally excludes the relevant desire.
Aristotle observes in De Anima III.9, ‘it is not appetite which is responsible
for movement; for the self-controlled, though they may crave and desire, do
not do these things for which they have an appetite, but follow their reason’
(433a 9 f). By itself, this comment does not necessarily vitiate the causalist’s
commitment to a belief and desire pair as constituting the primary origin
and cause of action, since causalists could argue that acting against a specifc
desire does not exclude acting from a secondary desire, that is, the desire to
follow one’s reason rather than one’s impulse.
18. See Ginet (1990).
19. All cited translations of these passages are Charles’s.
20. Aristotle’s remarks on the afections could also be added here: ‘In most cases it
seems that none of the afections, whether active [πασχειν] or passive [ποιεîν],
can exist apart from the body. This applies to anger, courage, desire and sensa-
tion generally, though possibly thinking is an exception. But if this too is a kind
of imagination . . . even this cannot exist apart from the body’ (403a 6–7).
21. Charles draws especially from the latter two passages, De An. 1.1(c) and De
An. 1.1(d), to develop his understanding of desire as ‘inseparable in defnition
as well as in existence from states of the body’.
22. Elsewhere, Charles emphasizes that his understanding of action as an ‘inextri-
cably psychophysical process’ is based on the way in which explanation of the
elements of action refer to one another, rather than a mere a priori claim about
the craftsman’s skill being ‘necessarily enmattered’ in certain physical processes.
Furthermore, the term ‘inextricably psychophysical’ goes beyond merely com-
bining psychological and physical elements of action. Charles writes,
Activities (or processes) which are ‘common’ to the body and the soul, as
we have seen, cannot be defned as a combination of psychological and
physical components. One cannot say what they are in purely psycho-
logical terms as they are types of expanding and contracting, becoming
hard and soft, which generate pushing/pulling. Like anger, they are not
simply necessarily enmattered in some physical process. Rather, one can-
not defne what they are without reference to expanding or contracting,
pulling or pushing, and so forth. The latter features are of their very
essence. Nor can one defne the relevant type of physical process without
reference to their psychological aspects. This is what I intend by calling
these processes ‘inextricably psychophysical’.
(2011: 86–87)
23. For further commentary, see Christopher Shields (1988) and Sarah Waterlow
(1982).
24. ‘For the soul, conceived of as a purely psychological phenomenon (or set of
such phenomena), is not the type of thing which can be the subject of these
inextricably psychophysical processes. At this point, Aristotle is best seen as
challenging the moves that lead to dualism at the frst step: it is not the soul
(properly speaking) that moves the animal but the (psychophysical) person’
(2011: 92).
25. Here, I apparently depart from Charles’s treatment of desire, which is unfor-
tunately scant. Like the causalist, Charles seems to assume throughout his
argument that desire is a kind of antecedent cause to the movement of the
animal. Since desire itself is a psychophysical phenomenon, it may be objected
that it cannot serve as a constituent for action, since the form was earlier spec-
ifed as the craftsperson’s logos (in De Gen. An. 744b 32 f). As I understand
Charles, the spatial aspects of an action are inextricably defned with respect
to the ‘goals and know-how of the craftsman’, which include desire, but not
exhaustively (2011: 87). Because of defnitional inextricability, whatever is
included in the craftsperson’s logos will have spatial features as part of its
concept (thus desire in action is psychophysical, as is the intention or goals
and technical knowledge of a human craftsperson). Thanks to an anonymous
referee for raising this objection.
26. Byrne, Alex and Logue, Heather. ‘Either/Or’ (2008). In Disjunctivism: Per-
ception, Action, Knowledge, eds. Adrian Haddock and Fiona Macpherson.
Oxford: Oxford University Press. p. 67.
27. Since my proposed disjunctivism about action is intended to be compatible
with animalism and the paradigm of animal agents, any variety of episte-
mological disjunctivism implied by my metaphysical account must allow for
animal agency in its conception of ‘acting for reasons’. McDowell’s theory
(1986, 1996) is therefore incompatible with my disjunctivist approach, as
his epistemological account excludes nonrational animals. In any case, my
account concerns action, not knowledge per se, and follows Anscombe’s Aris-
totelian view in Intention: ‘Intention appears to be something that we can
express, but which brutes (which e.g. do not give orders) can have, though
lacking any distinct expression of intention’ (1957: 5).
28. Byrne, Alex and Logue, Heather. ‘Either/Or’ (2008). In Disjunctivism: Per-
ception, Action, Knowledge, eds. Adrian Haddock and Fiona Macpherson.
Oxford: Oxford University Press. p. 60.
29. All examples taken from Byrne and Logue’s ‘Either/Or’, in Haddock and
Macpherson, eds. (2008: 60).
30. McDowell, John. ‘Criteria, Defeasibility, and Knowledge’ (1983). Proceed-
ings of the British Academy, 68: 455–479.
31. Snowdon, Paul. ‘Perception, Vision, and Causation’ (1980). Proceedings of
the Aristotelian Society, 81: 175–192. p. 186.
32. Martin, M.G.F. ‘The Limits of Self-Awareness’ (2004). Philosophical Studies,
120: 37–89. p. 37. Also see Martin (2002, 2006).
33. Dancy, Jonathan. ‘Arguments From Illusion’ (1995). The Philosophical
Quarterly, 45, 181: 421–438. p. 435.
34. I have defended this account in detail elsewhere. See ‘Thomistic Animalism’,
New Blackfriars (2019), and ‘Action, Animacy, and Substance Causation’, in
Neo-Aristotelian Perspectives on Contemporary Science (2018), eds. William
Simpson, Robert Koons, and Nicholas Teh. London: Routledge.
35. These examples are specifc to land mammals like us, but one can speculate
about examples pertaining to other animal forms: perhaps a fsh’s gill move-
ments are similar, or the wing movements of a bird fying a great distance.
36. See, for instance, Itzhak Fried et al. ‘Volition and Action in the Human Brain:
Processes, Pathologies, and Reasons’ (2017). The Journal of Neuroscience,
37, 45: 10842–10847.
37. Haddock, Adrian and Macpherson, Fiona (eds.). Disjunctivism: Perception,
Action, Knowledge (2008). Oxford: Oxford University Press. pp. 21–22.
38. Broadie, Sarah. ‘Agency and Determinism in A Metaphysics for Freedom’
(2013). Inquiry: An Interdisciplinary Journal of Philosophy, 56, 6: 571–582.
p. 574.
39. Haddock, Adrian and Macpherson, Fiona (eds.). Disjunctivism: Perception,
Action, Knowledge (2008). Oxford: Oxford University Press. p. 22.
40. Ibid.
Works Cited
Aguilar, J. and Buckaref, A. Causing Human Actions: New Perspectives on the
Causal Theory of Action (2010). Cambridge, MA: MIT Press.
Anscombe, Elizabeth. Intention (1957). Cambridge, MA: Harvard University
Press.
Aristotle. Categories.
——— De Anima.
——— De Generatione Animalium.
——— De Motu Animalium.
——— Eudemian Ethics.
——— Metaphysics.
——— Nicomachean Ethics.
——— Physics.
——— Rhetoric.
Bratman, Michael. ‘Refection, Planning and Temporally Extended Agency’
(2000). Philosophical Review, 109: 35–61.
Broadie, Sarah. ‘Agency and Determinism in a Metaphysics for Freedom’ (2013).
Inquiry: An interdisciplinary Journal of Philosophy, 56, 6: 571–582.
Byrne, Alex and Logue, Heather. Disjunctivism: Contemporary Readings (2008).
Cambridge, MA: MIT Press.
Charles, David. Aristotle’s Philosophy of Action (1984). London: Duckworth.
——— ‘Aristotle’s Psychological Theory’ (2009). In Proceedings of the Boston
Area Colloquium in Ancient Philosophy, 2008, Volume 24, ed. J. J. Cleary and
Gary M. Gurtler.
——— ‘Desire in Action: Aristotle’s Move’ (2011). In Moral Psychology and
Human Action in Aristotle, eds. Michael Pakuluk and Giles Pearson. New
York: Oxford University Press.
Chik, Janice Tzuling ‘Action, Animacy, and Substance Causation’ (2018). In Neo-
Aristotelian Perspectives on Contemporary Science, eds. William Simpson,
Robert Koons, and Nicholas Teh. New York: Routledge.
——— ‘Thomistic Animalism’ (2019). New Blackfriars, 100, 1090: 645–662.
Coope, Ursula. ‘Aristotle on Action’ (2007). In Proceedings of the Aristotelian
Society Supplementary Volume LXXXI.
Dancy, Jonathan. ‘Arguments From Illusion’ (1995). The Philosophical Quar-
terly, 45, 181: 421–438.
Danto, Arthur. ‘Basic Actions’ (1965). American Philosophical Quarterly, 2, 2:
141–148.
——— with Sidney Morgenbesser. ‘What We Can Do’ (1963). The Journal of
Philosophy, 60, 15: 435–445.
Davidson, Donald. ‘Actions, Reasons and Causes’ (1963). Journal of Philosophy,
60: 685–700.
——— ‘Agency’ (1971). In Agent, Action and Reason, eds. Robert Binkley, Rich-
ard Bronaugh, and Ausonia Marras. Toronto: University of Toronto Press.
Enç, Berent. How We Act: Causes, Reasons, and Intentions (2006). New York:
Fried, Itzhak, Haggard, Patrick, He, Biyu and Schurger, Aaron. ‘Volition and
Action in the Human Brain: Processes, Pathologies, and Reasons’ (2017). The
Journal of Neuroscience, 37, 45: 10842–10847.
Ginet, Carl. On Action (1990). Cambridge, UK: Cambridge University Press.
Haddock, Adrian and Macpherson, Fiona, eds. Disjunctivism: Perception, Action,
Knowledge (2008). Oxford: Oxford University Press.
Hinton, J.M. Experiences: An Inquiry Into Some Ambiguities (1973). Oxford:
Hornsby, Jennifer. ‘Actions in Their Circumstances’ (2011). In Essays on Ans-
combe’s Intention, eds. Anton Ford, Jennifer Hornsby, and Frederick Stout-
land. Cambridge, MA: Harvard University Press.
Lowe, E.J. Personal Agency: The Metaphysics of Mind and Action (2010). New
York: Oxford University Press.
Martin, M.G.F. ‘The Transparency of Experience’ (2002). Mind and Language,
4, 4: 376–425.
——— ‘The Limits of Self-Awareness’ (2004). Philosophical Studies, 120: 37–89.
——— ‘On Being Alienated’ (2006). In Perceptual Experience, eds. T. Gendler
and J. Hawthorne. New York: Oxford University Press.
Menn, Stephen. ‘Aristotle’s Defnition of Soul and the Programme of the De
Anima’ (2002). Oxford Studies in Ancient Philosophy, 22: 83–139.
McDowell, John. ‘Criteria, Defeasibility, and Knowledge’ (1983). Proceedings of
the British Academy 68: 455–479.
——— ‘Singular Thought and the Extent of Inner Space’ (1986). In Subject,
Thought, and Context, eds. P. Petit and J. McDowell. Oxford: Clarendon Press.
——— Mind and World (1996). Cambridge, MA: Harvard University Press.
O’Connor, Timothy. Persons and Causes: The Metaphysics of Free Will (2000).
Shields, Christopher. ‘Soul as Subject in Aristotle’s De Anima’ (1988). The Clas-
sical Quarterly, 38, 1: 140–149.
Smith, Michael. ‘Four Objections to the Standard Story of Action (and Four
Replies)’ (2012). Philosophical Issues, 22: 387–401.
Snowdon, Paul. ‘Perception, Vision, and Causation’ (1980). Proceedings of the
Aristotelian Society, 81: 175–192.
Velleman, David. ‘What Happens When Someone Acts?’ (1992). Mind, 101, 403:
461–481.
Waterlow, Sarah. Nature, Change and Agency in Aristotle’s Physics (1982).
Oxford: Clarendon Press.
Wilson, George. The Intentionality of Human Action (1989). Redwood City:
Stanford University Press.
Wittgenstein, Ludwig. Philosophical Investigations (2001/1953), trans. G.E.M.
Anscombe. Oxford: Basil Blackwell.
7 The Power to Perform
Experiments
Daniel D. De Haan
Do humans have abilities to perform scientifc experiments? Do humans

possess real powers for performing scientifc experiments? I shall treat
these two questions in turn where the frst will bring us to the second. I
shall argue that the scientifc image of humans must cohere with the mani-
fest image of humans as having the ability to exercise rational embod-
ied control. This is because the scientifc image of humans depends on
the results of experiments performed by scientists, experiments that are
instances of the manifest image of humans with the power to exercise
rational embodied control over physical phenomena. Since the scientifc
image of humans cannot be incompatible with the existence of scientists
who perform scientifc experiments, it cannot be incompatible with the
manifest image of humans that exercise rational embodied control in
scientifc experiments. Let us start with why someone might argue the
scientifc image discredits the manifest image and thereby undermines the
theses that humans can perform scientifc experiments and are endowed
with the real powers to do so.
1. Manifest and Scientifc Images of Humans

It has become a cliché for philosophers and scientists to pronounce that
advances in scientifc enquiry have disclosed a scientifc image of humans
that has overturned our commonsense or manifest image of humans.
When Wilfred Sellars coined his post-Kantian distinction between scien-
tifc and manifest images of humans, he was identifying a troubling ten-
sion that he sought to alleviate by articulating how these two images could
be reimagined in deference to the scientifc image yet without jettisoning
the “irreducible core of the framework of persons.”1 Many since Sellars
have ignored his insightful and nuanced conciliatory handling of this ten-
sion by endorsing the very confrontational interpretations of this distinc-
tion that Sellars cogently argued against. Reductive physicalists argue that
the manifest image can be reductively explained by the scientifc image.
Eliminative physicalists go so far as to contend that the manifest image
will be eliminated and wholly replaced by the scientifc image because the
DOI: 10.4324/9781003125860-10
The Power to Perform Experiments 193
results of scientifc experiments have purportedly established a form of
scientifc naturalism that falsifes the manifest image of humans.2
These bold contentions overlook a number of important facts that
place unavoidable obstacles in the path of these eforts to bulldoze
over the manifest image of humans; obstacles these glib revolutionary
viewpoints must confront before anyone should lend any plausibility
to their ungrounded promissory explanations. Some of these obstacles
come from important work in philosophy of science that has turned
to focus on the history and practices of scientists. These investigations
have overturned many armchair physicalists’ assumptions about what it
is scientifc investigations really discover about nature by their experi-
ments and theorizing. They have also led many to endorse, or at least
sympathize with, Nancy Cartwright’s “empiricist dictum . . .: Construct
our scientifc image of the world from our scientifc practices that prove
successful in interacting with it.”3 This dictum has provided scientifc
support for an ontology of manifesting but also defeasible causal powers
over and against a conception of exceptionless universal laws of nature.4
This approach to articulating our scientifc image of the world must also
include an account of its dependence upon the scientists and their suc-
cessful scientifc practices.
This chapter focuses on one of the signifcant obstacles to the radical
revisionist viewpoint that comes from considering what scientifc practices
in themselves require in order to secure any scientifc image of the world
and of humans in it. The obstacle I focus on is the fact that humans
can perform scientifc experiments and that this fact is included among
the explananda of both manifest and scientifc images of humans. My
argument will be prosecuted as follows. First, I establish why rational
embodied control is indispensable to the ability to perform a scientifc
experiment. Second, I contend that an ontology of causal powers provides
the best explanation for this ability to make the causal diference in reality
required for scientists to efectively perform scientifc experiments and
arrive at scientifc truths. I then conclude by considering a major objection
against the alternative standard causal theory of action, which would also
undermine the thesis that scientists perform experiments. I show why neo-
Aristotelian causal power realism does not face this objection in its expla-
nation of how scientists exercise real rational embodied causal control
within their experiments, which supplies one more reason for endorsing
neo-Aristotelian causal power realism over rival causal theories.
To clarify the diferent stages of my argument I shall stipulate a termi-
nological distinction between abilities and powers. An “ability” denotes a
notion of discourse that is underdetermined with respect to its ontological
underpinnings. Talk of abilities might be indispensable to intelligible and
true discourse, but this discourse does not of itself disclose any ontological
commitments. I use the term “power” ontologically and will later argue
that true discourse about abilities that make a diference in the world
194 Daniel D. De Haan
is best explained by endorsing power realism, wherein powers are the
truthmakers that ground true discourse about abilities.
2. Rational Embodied Control

What is necessary for performing scientifc experiments? We can distill
the panoply of features required for performing scientifc experiments
down to a scientist’s abilities for rational embodied control over physi-
cal phenomena: in physics, scientists orchestrate the collision of proton
beams in the Large Hadron Collider; in chemistry, scientists can synthe-
size an extraordinary range of substances, like ammonia from nitrogen
and hydrogen; in biology, scientists can edit genomes using CRISPR. This
distillation down to three constitutive and interconnected abilities is not a
reductive explanation of scientifc experimentation, but rather what Peter
Strawson called a connective one. None of these three factors – rational,
embodied, control – is more primary or basic than the others, rather their
interconnection is one of mutual presupposition. Let us examine each of
these features in turn.
2.1. Rational
It should be uncontroversial that scientifc experiments are rational. Being
“rational” is a personal-level attribute in contrast to the subpersonal-level
attributes of a human person’s nervous and endocrine systems. An agent
is rational insofar as it is a person whose observing, thinking, and acting
can be intelligently constituted and guided by reasons, as opposed to
non-intelligently and mechanically processing information or responding
to stimuli. This notion of “rational” therefore excludes much of what
cognitive scientists and even functionalists in philosophy of mind call
reasons and rationality, which are often ascribed to overtly subpersonal-
level objects, states, networks, and processes.
Being rational comprises observing, thinking, speaking, and acting intel-
ligently and voluntarily according to reasons; this includes both practical
and theoretical reason. Understood reasons can intelligently govern and
guide what one is thinking and what one is doing, wherein thinking and
speaking are themselves typically forms of intentional acting, and embod-
ied intentional acting is a manifestation of thinking rationally. To exercise
practical reason is to perform actions intentionally – often accompanied by
or even constituted by deliberation and decision – and voluntarily or at will
with respect to the commencement, and/or continuation, and/or cessation
of some overtly observable or unobservable psychological behavior – like
calibrating a laser or performing calculations “in one’s head.”
Human actions are intentional because they actually realize and embody
one’s practical reasons or intentions for acting as one does. The embodied
psychological behaviors of scientists are not mere refexes or instinctual
responses, for then their behaviors would be nonrational. Similarly, the
embodied psychological behaviors of scientists engaged in scientifc prac-
tices are not instances of faking or incompetently calibrating a laser for
an experiment, for then their behaviors would be irrational. That is to
say, they would be acting against the norms of rationality for truthfulness,
integrity, and the deployment of scientifc competencies in the intelligent,
intentional, and voluntary calibration of a laser for an experiment.
Human actions are voluntary insofar as scientists perform their experi-
ments at will, that is, they have the two-way ability to perform experiment
φ or not to perform experiment φ. This, of course, drastically underde-
scribes how complex the performance of an experiment is, for it omits the
social collaboration involved in most scientifc experiments that rely on
the coordination of multiple scientists intentionally and voluntarily per-
forming a scientifc experiment by working together. It also obscures how,
within such complex ongoing activities as collaboratively performing an
experiment together, a team of scientists is comprised of individual sci-
entists who must intentionally and voluntarily deliberate together about
and enact what they will φ, whether they will φ, when they will φ, where
they will φ, how they will φ, and why they will φ in this way and at this
time rather than φ in that way and at that time.5
For example, for any neuroscientifc experiment on long-term memory
in rodents that employs optogenetic stimulation, neuroscientists need to
answer and then intentionally and voluntarily enact their answers to these
questions, the following questions, and many more. Given past experimen-
tal data and theorizing relevant to these questions, neuroscientists design-
ing, enacting, and interpreting their optogenetics experiment on rodent
long-term memory need to decide what kind of mice to use, how many mice
are sufcient for a control group, what kind of task(s) should be employed,
how long the mice should train on the task(s), what kind of diet the mice
should have, when trials should start following recovery from surgery,
and so on to conduct their experiments applying optogenetic stimulation.
Within the trials themselves experimenters need to work through a similar
battery of questions concerning when to apply optogenetic stimulation
vis-à-vis the task(s), for how long during each task, how many times, how
long should each trial run, and so on. In each case a variety of variables,
counterfactuals, and contravening factors need to be intelligently discerned
and experimentally controlled for, mitigated, or eliminated.
It would not be difcult to explicate – and in more detail and with
greater depth beyond the surface level surview I have presented here – the
myriad ways rationality is constitutive of performing scientifc experi-
ments, including the experimental intelligence required to identify and
overcome contravening and confounding factors. We can fnd a litany of
examples on display within the methods, critical discussions, and other
sections of scientifc studies. What these discussions exhibit beyond the
basic and straightforward application of scientifc rationality to the design,
implementation, and interpretations of results of scientifc experiments is
a rational ability for critical self-refection upon the scientifc reasoning
and intelligent adherence to the canons and norms of rationality that are
themselves embodied within the performance of a scientifc experiment.
And on some occasions, but especially within philosophical studies on
these scientifc practices of experimentation and theorization, we fnd dis-
closed the ability of reason to engage in critical refection upon the argu-
ments for and against rival understandings of the very canons of scientifc
enquiry and norms of rationality, along with the debates concerning the
criteria for how and when to apply them. Scientifc rationality therefore
involves and relies upon both reason’s ability to employ the norms and
standards of rationality within experimental investigations of domains of
empirical phenomena, and the rational ability to refect critically on the
norms and standards of rationality that govern the former experimental
investigations of domains of empirical phenomena.
Signifcantly, the norms of rationality that are learned and efcaciously
employed in all scientifc thinking and acting cannot be established or
rationally vindicated on the basis of any scientifc experimental results.
This is because norms of rationality are presupposed by all scientifc
experimentation and theorization and so also by any scientifc image of
the world or of humans that is reliant on scientifc reasoning. Scientifc
investigations do not supply or vindicate the norms of rationality; rather,
the manifest image of humans is the indispensable source for the norms
of rationality and supplies the very criteria and grounds by which we can
vindicate or revise the normative standards of rationality. Such vindica-
tions or revisions are independent from the scientifc investigations and
experiments that fundamentally rely on the normative space of reasons
and their embodied efcacy.
Against this cursory account of rationality one might object that many of
the forms of scientifc reasoning mentioned here are rarely if ever explicitly
thematized and actually articulated. But this is not an objection so much as
a reminder to make clear that this account of rationality can only be cor-
rectly understood and interpreted as one that is informed and qualifed by
the treatments of embodiment, expertise, tacit knowledge, know-how, and
intentional action of Merleau-Ponty, Polanyi, Ryle, and Anscombe; diverse
threads that are often woven together in explicitly complementary ways
by Charles Taylor, Alasdair MacIntyre, and many others, but especially
by the radical enactive cognition theorists. The inescapable dimension of
embodiment that keeps emerging in our treatment of scientifc rationality
brings us to the second feature of scientifc experiments.
2.2. Embodiment
In order to give some focus to the wide-ranging issue of embodiment, I
shall enter into its signifcance by feshing out some of its most salient
connections with rationality. We can start with the fact that human ratio-
nality is the embodied rationality of social animals. Humans are altricial
mammals who are unable to exercise their rational capacities for inten-
tional and voluntary action for many years after birth. Humans only learn
how to exercise these capacities in virtue of the social scafolding that
nurtures, cares for, and educates humans into becoming animals who can
exercise their rational capacities for linguistic, practical, and theoretical
interactions with other humans. But even enabling and later engaging
in embodied rational activities presupposes the rational transformation
of the many embodied capacities we share with other animals, such as
capacities for enactive perception, motivation, emotion, associative and
social mimetic learning, and the palpably embodied haptic abilities for
manipulating physical phenomena. For example, engaging in scientifc
observation and experimentation requires rationally transforming one’s
perceptual capacities to achieve sufcient observational acuity to select
and discriminate intelligently among what is relevant versus irrelevant
within observable data. Engaging in these scientifc activities also involves
the rational transformation of one’s embodied capacities for motivation
and emotion that enable the kinds of discipline, persistence, resilience, and
passion required for the arduous painstaking work that comprises scien-
tifc experimentation. Experimental work is, after all, full of long tedious
hours of trial and error accompanied by seemingly endless dead ends.
Furthermore, because humans are rational animals, they can neither
exercise their rational capacities ab initio nor immediately rationally
transform their embodied capacities for enactive perception, motivation,
emotion, and haptic manipulation that constitute most embodied rational
activities, like performing a scientifc experiment. There is a long drawn
out socially informed ontogenetic process of enablings and transforma-
tions of proto-rational and later rational capacities that is worked out
through many phases of human development. And because development
is essential to human embodiment, being socially dependent in myriad
ways throughout life is thereby also essential to being embodied rational
agents. To become sufciently educated to have the proximate potentiali-
ties for becoming a scientist requires not only that all of our capacities
for perception, motivation, emotion, language, intelligence, and will have
been enabled and cultivated by the right caregivers who are provided with
the ecological and social resources needed for engendering these basic
human ontogenetic transformations. It also requires a vast educational
infrastructure populated by competent educators that can prepare devel-
oping minds to become potential initiates into scientifc experimentation
and theorization. But even that is often not sufcient, for what is ordinar-
ily required today is a still more advanced educational formation that
can transform educated potential scientists into fully initiated and col-
laborating participants in normal science, and this depends on elaborate
social institutions with instructors, mentors, and colleagues all working
together. Finally, the growing number of scientifc experimental mega-
projects (e.g., NASA, HGP, CERN, LIGO-Virgo) also depend on extraor-
dinary international political and economic resources, cooperations, and
the institutions and populations that make them possible, in addition to
the publicly and privatively funded universities and other research institu-
tions we ordinarily identify with normal science and its research.
In short, the essential embodiment of human persons is manifested in
humans being developing and dependent rational animals. And scientifc
practices, including experimentation, are constituted by their own distinct
forms of embodied human development and social dependencies. Once
again, it is these personal-level developmental factors from the manifest
image of humans that are indispensable to scientifc practices and so to
any critical refection on the curricula and standards of scientifc educa-
tion and methods.
Embodiment also sets in relief the constitutive nexus wherein rationality
and control are intertwined, for we exercise rational control both over our
own enactive perceptions, motivations, emotions, desires, and beliefs and
“also” over our own bodies and other bodies. I say “also” because I do
not intend to suggest, as most philosophers of mind and other exponents
of the discourse of “folk psychology” do, that we can fully understand
what it is to exercise rationality – not to mention rational control over our
perceptions, motivations, emotions, desires, and beliefs – apart from being
embodied animals that exercise control over our own and other bodies in
our surrounding environment. No fully satisfactory explication of what
perceptions, motivations, emotions, desires, and beliefs are could be intel-
ligible apart from the embodied psychological behaviors that are partially
constitutive of human abilities for perception, emotion, desire, belief, and
so forth.6 For instance, while it is obvious that enactive perceptions are
constituted by perspectival observations due to orientations, comport-
ments, and other constraints of embodiment, even emotions, desires, and
beliefs, like perceptions, are identifed by their objects, and the objects of
these capacities largely consist in the embodied objects of our environ-
ments and our enactive engagements with them. It is only through devel-
opmentally later transformations, via social, linguistic, and conceptual
learning, that we eventually become enabled to have emotions, desires,
and beliefs concerning objects that are interiorized, transcend physical
objects, or, as Charles Taylor argues, constitute and create meaningful dis-
course beyond the ecological afordances of objects in our environment.7
2.3. Control
The myriad illustrations of humans exercising rational embodied control
over physical phenomena are endless. I shall limit the scope of this broad
topic by focusing on how rational embodiment pertains to control within
scientifc experiments. Roughly, rational embodied control consists in the
ability to make a diference with respect to either the inception, and/
or continuation, and/or cessation of the directed and purposeful move-
ments or changes of one’s own body and of other surrounding bodies
in the ways one intends to make a diference with respect to what these
objects’ capacities aford for movement and change. Adult humans can
ordinarily control whether and how they walk across a room insofar as
they can intentionally initiate, continue, and terminate taking steps across
the room. But if one’s leg has been anesthetized or one sufers from tin-
nitus, vertigo, or other disabilities, it can be difcult or impossible even
to commence walking across a room. Some events we either do not or
cannot initiate, but we can nevertheless control either an ongoing fow or
an outcome by modulating and infuencing various factors. For example,
when we try to inhibit a sneeze, reduce a fever by taking paracetamol, or
steer a kayak out of a ferce river current.
Human bodily movements are characteristically, though not always,
rationally guided, purposeful, and controlled bodily movements in con-
trast to mere mechanical and purposeless (which is not to say directionless)
bodily movements. Rationally directed and controlled hand movements,
among other haptic movements, are manifested in the complex ordered
and purposeful activities that comprise making a fre or a loaf of bread,
buttering toast with a knife, replacing a transmission, calibrating a laser,
surgically installing an optogenetic optical fber within the medial ento-
rhinal cortex of a rodent, manufacturing and installing the “quadruple-
pendulum system, supported by an active seismic isolation platform”8 of
the mirrors within the LIGO (Laser Interferometer Gravitational-Wave
Observatory) interferometers in Hanford, Washington, and Livingston,
Louisiana, USA. Clearly, performing a scientifc experiment falls on this
enormous list of embodied forms of rationally controlled engagements
with the world, or umwelt as Rom Harré characterizes it. And without the
discoveries made possible by these scientifc experiments, humans would
be oblivious to the hidden properties of nature.
The umwelt is that region of the world which is available to a species

by virtue of their capacities to register and explore it. The human
umwelt is bounded at any moment by the human capacities for per-
ception and the instruments currently available to extend them. The
ontology of the umwelt was greatly illuminated by Niels Bohr. He
realized that human beings could ascribe only dispositions to the
world behind appearances. He also realized that those dispositions
were defned by the available experimental equipment. They are, in
the last analysis, Gibsonian afordances, not occurrent properties. . . .
There is indeed a world beyond the observable – but it is an umwelt,
and its properties, for the human species, are the afordances it has, as
made humanly discernible through experimental apparatus (emphasis
in original).9
A scientifc experimenter attempts to exercise control over every known
and relevant factor that can be controlled, while also mitigating or intel-
ligently identifying and accounting for all of the factors that cannot be
controlled or eliminated. Insofar as the experimenter – or often enough
experimenters – can achieve the construction of a rationally controlled
experimental context, they can thereby either allow nature to do whatever
it may do, or they can direct and even force nature to disclose and mani-
fest its powers. If this kind of rational embodied control were delusional
or impossible, then we would have to discount countless scientifc discov-
eries and theories grounded in experimental results. Like, for instance, the
1910 experimental results of Robert Millikan and Harvey Fletcher that
purported to isolate and measure the charge of an electron by their inge-
nious oil-drop experiment. The performance of the Millikan experiment
can now – through the developmental social ratcheting-up efects of inno-
vations in scientifc understanding, education, and technology – be taught
to and reenacted by college physics students.10 Briefy and simplistically, in
the Millikan oil-drop experiment, aspirated and ionized oil droplets in an
upper chamber fall into a lower chamber that has two parallel horizontal
charged plates. Gravity pulls the oil droplets down, but the electric attrac-
tion and repulsion due to the oppositely charged plates can work together
to pull the charged oil droplets up. By turning of and on the potential dif-
ference between the charged plates Millikan in 1910 could – and college
students repeating the experiment today can – manipulate the movements
of the charged oil droplets either to fall slowly under gravity resisted by
drag or to move upwards by the additional contribution of electric force
from the charged plates. It is this experimental setup that allowed Mil-
likan to perform an experiment to measure the charge of electrons.
As Thomas Kuhn and many others have detailed, the experimental
work of normal science can be directed to multiple ends, but it is rarely,
if ever, conceived, designed, and implemented apart from a theoretical
paradigm that inspires and channels the peculiarities of the structure of
the experiment itself.
The existence of the paradigm sets the problem to be solved; often the
paradigm theory is implicated directly in the design of apparatus able
to solve the problem. Without the Principia, for example, measure-
ments made with the Atwood machine would have meant nothing
at all.11
So also the Millikan experiment presumes theoretical accounts of gravity,

electromagnetism, the drag force on the droplet due to the air, and so forth
to design, manufacture, and employ two charged plates to manipulate
the movements of ionized oil droplets by altering the voltage across the
plates as well as the ability to isolate the relevant experimental factors
from other diference makers. All of these critical considerations and
experimental operationalizations are at the heart of a rationally informed
and controlled scientifc experiment. Signifcantly, objections and con-
cerns about the reliability of Millikan’s results or the multiple sources of
potential experimental error within his experiment have not challenged
or undermined the ability of scientifc experimenters to achieve rational
embodied control over physical phenomena, like manipulating the move-
ments of charged oil droplets. Instead, they have inspired the application
of more critical reasoning and intelligence in the theoretical modeling,
design, and execution of experiments.12
2.4. The Indispensability of Rational Embodied Control

One could fll volumes detailing further the multitudinous ways the abili-
ties for rational embodied control are constitutive elements in all scientifc
experiments, but the digest of thick descriptions I have expounded should
be sufcient to illustrate why abilities for rational embodied control are
indispensable for performing scientifc experiments.
If scientifc experiments are not rational, then scientifc experiments
and their results (a) cannot intelligently operationalize and assay theoreti-
cal paradigms, concepts, models, hypotheses, predictions, and so on, (b)
cannot count as rational evidence that verifes, confrms, or falsifes pre-
dictions, hypotheses, models, theories, paradigms, and so on, and (c) so
cannot be the truth-seeking, truth-tracking, and critically self-correcting
activities we understand and hold scientifc practices to be. But scientifc
experiments do aford us with these consequents and many more, so the
antecedent is false.
Embodiment is the constitutive nexus uniting rationality and control
in scientifc experiments. This is why the thesis that scientifc experiments
need not be embodied is oxymoronic. Of course, this notion of scientifc
experiments presumes the truth of some form of critical and explanatory
empirical realism and the falsity of idealism, but that is a pair of presump-
tions I and most plain persons, scientists, philosophers, and theologians
willingly hold and have more than sufcient rational justifcation for
holding. Consider nevertheless the absurdities that would follow from
this oxymoronic thesis. If scientifc experiments were not embodied but
disembodied, then scientifc experiments and their results would totally
fail to be empirical and so would be indistinguishable from “thought
experiments” in philosophy and mathematics; they might probe intuitions
but they would not probe the recalcitrant mettle of nature. Furthermore,
by failing to be empirical, scientifc experiments would also essentially fail
as empirical tests and so could not assay embodied observations, measure-
ments, manipulations, or predictions concerning any mind-independent
empirical phenomena. A scientifc experiment that is neither empirical
nor tests any empirical phenomena, because it is not embodied, is clearly
absurd.
For the sake of completeness, consider further that if scientifc experi-
ments were not controlled, then scientifc experiments and their results
would fail to provide rational grounds for distinguishing them from lucky
guesses, superstitions, conspiracy theories, and a host of falsifed scientifc
theories. Indeed, if any scientifc theories or predictions could be “falsi-
fed” on this supposition, then it would not be in virtue of any controlled
scientifc experiments, for there are none on this supposition. If scientists
do not exercise causal control over physical phenomena, then why are
photons fying back and forth along the arms of the LIGO interferom-
eters? If scientists did not cause this, then scientists have not performed a
controlled scientifc experiment to detect gravitational waves. Similarly,
if scientists did not cause the proton beams to fy through CERN’s Large
Hadron Collider, if they could not control the trajectory of them into
ATLAS and CMS detectors to the extent that they understand their mas-
sive experimental apparatus to do just this, then scientists did not perform
a controlled scientifc experiment to identify and establish the existence of
a Higgs boson. Clearly, scientists do exercise causally efcacious rational
embodied control over physical phenomena, including the fundamental
particles of nature.
In short, if one could establish that the manifest image of humans as
agents with rational embodied control was false, then it would thereby
also establish that truth-establishing scientifc experiments do not exist,
and this would undermine all scientifc theorizing that is based upon
data assumed to have been accurately acquired by the rational embodied
control of scientifc experiments. If the scientifc image of humans truly
discredits the manifest image of humans, then it also discredits the corner-
stone of all scientifc enquiry along with any scientifc image of humans
or the world that is derived from scientifc enquiry.
Certain kinds of error theorists and skeptics might welcome this con-
clusion. They wish to afrm these antecedents and push back against my
modus tollens arguments in defense of the abilities for rational embod-
ied control required for scientifc experiments. How should we respond?
This scientifc, if not global, skepticism will rightly be met with a volley
of objections. The most fundamental objection we can raise is to ask
upon what evidential or other grounds can anyone cogently establish
that scientifc experiments cannot be constituted by rational embodied
control because humans do not have rational embodied control over any
phenomena?
Appeals to the results of scientifc experiments are not available either
to motivate or to justify these skeptical contentions, since these skepti-
cal conclusions presume scientifc experiments are discredited. Similarly,
skeptics’ own criticisms of faulty and defective applications of rationality
presuppose that rationality is the critical self-corrective ability detailed
previously; we can only rationally concede the rational grounds of their
critiques if we reject their skeptical rejections of rationality. Skeptical
responses to these basic challenges often try to turn the tables by arguing
that there is a desperate need to make extraordinary radical revisions to
all of our notions concerning the norms and standards of rationality. The
problem is, any such “proposal” for radical revision remains literally
unintelligible and inconceivable at this point, and this defect even infects
any “grounds” or “arguments” proposed for entertaining the “plausibil-
ity” of a radical total revision of our epistemological and other notions.
How can we even concede as “plausible” that which we have been given
no plausible account of, and so cannot give any arguments for, and for any
thesis which essentially contends that concepts like “rational,” “truth,”
“plausible,” and “argument” need to be eliminated or radically trans-
formed in their basic meanings?13
More can always be said about arguments against and in defense of
these radical skeptical and error-theoretic positions, but even wading this
far into the shallow end of the pool of these debates is enough to disclose
the gravamen underwriting the insurmountable difculties facing these
radical revisionist positions. For each positive argument they enunciate
in favor of why some skeptical thesis is true, namely, that some domain
of knowledge is fundamentally in error, they expose their own skepti-
cal thesis and arguments to the retort that this new contention entails a
self-defeating global skepticism that undermines the arguments for their
skeptical thesis.
What I have prosecuted thus far is a distinctive form of transcendental
argument against the claims that the scientifc image of humans has dis-
credited the manifest image of humans. I follow Charles Taylor’s account
of (a) what makes certain transcendental arguments valid, (b) why they
remain open to revisions and further objections, and signifcantly for
this chapter, (c) why they always leave us with unsettled metaphysical
business.14 The argument presented here establishes why the scientifc
image of humans cannot be at odds with the manifest image, for the only
way we can get at anything that might justifably be deemed a scientifc
image of humans or the world will be arrived at via performing scientifc
experiments. And scientifc experiments are specifc manifestations of the
completely quotidian abilities of the manifest image of humans to exercise
rational embodied control over physical phenomena. The scientifc image
of humans is one that needs to be of a piece with the existence of scientists
and their scientifc practices. I have argued that scientists and their scien-
tifc practices manifestly inhabit and depend upon the sociopolitical world
of the manifest image of humans as agents that exercise rational embodied
control. True, scientifc images of humans certainly do enrich, interro-
gate, and sometimes revise signifcant aspects of our manifest images of
humans. But true scientifc images do not achieve this by presumptively
dismissing manifest images in the problematic and self-defeating ways
I have argued against.15 True scientifc images achieve such ennobling
goals by explaining the subpersonal biological systems that undergird
our perceptions, emotions, and embodied movements, by challenging our
prior, limited, and erroneous manifest and scientist images about physi-
ological and psychological disorders, and by radically revising our views
about the biological origins of organisms like humans and other animals.
This Taylorian transcendental argument brings out how enormous the
descriptive and explanatory burden is for radical revisionists. It sets in
relief the Olympian proportions of the challenge they must overcome,
rather than the elementary school track meet they so often portray the
manifest image to be. It concludes that the performance of scientifc
experiments would be unintelligible and incapable of securing facts and
verifying scientifc theories without the basic constitutive features of ratio-
nal embodied control. But, as Taylor shows, transcendental arguments
do not settle metaphysical questions; rather, they are indispensable pro-
legomena to our metaphysical investigations. So even if scientifc experi-
ments are unintelligible apart from rational embodied control and other
thick descriptive notions drawn from true manifest and scientifc images
of humans, why should we countenance the existence of an ontological
image of humans with real powers to perform experiments?
3. From Abilities to Powers

A number of argumentative gears are available for shifting from the
conclusions of our Taylorian transcendental argument in defense of the
indispensability of abilities for rational embodied control for performing
scientifc experiments to the question of what this commits us to onto-
logically. I start with the existence question and then turn to why these
existing abilities should be identifed with neo-Aristotelian causal powers.
3.1. The Existence of Abilities

There are good arguments for why even scientifcally inclined ontological
naturalists should hold that these abilities exist. Ontological naturalism
is committed to the existence of whatever is required to achieve efec-
tive and true scientifc enquiries. Many ontological naturalists, follow-
ing Quine, maintain that our ontology must recognize the existence of
whatever is disclosed and required by our best scientifc theories. For
example, because scientifc theories depend on mathematics, the ontologi-
cal reality of mathematics must be acknowledged. Going beyond Quinean
ontology, metaphysics still has more work to do than fll out ontological
grocery lists, for acknowledging the indispensable existence of mathemati-
cal objects neither precludes nor settles debates over what mathematical
“objects” are and what their mode of being might be.16 In addition to
mathematics, our scientifc theories also depend on whatever is essential
to the scientifc practices that are indispensable to our scientifc theories,
and these practices include conducting rigorous scientifc experiments
without which the armchair speculations of scientifc theorizing would be
indistinguishable from philosophy and mathematics. William Jaworski,
drawing on Cartwright and others, argues:
Empirical methods and techniques appear to provide important

sources of ontological commitment as well. The idea is roughly that
in constructing and executing experiments we are often implicitly
committed to a range of assumptions that carry ontological commit-
ments of their own but that might nevertheless remain unstated in
the more canonical descriptions and explanations we give. If this is
the case, then we need to expand the basic Quinean thesis to accom-
modate these further commitments. On this expanded understanding,
ontological naturalism says that we are committed not only to the
existence of the entities needed to make our best empirical descrip-
tions and explanations true, but also to the existence of the entities
needed to make our best empirical methods and techniques efective.17
In order for our scientifc practices to be efective and arrive at true descrip-
tions and explanations of empirical phenomena, we must be committed
to the existence of the entities required for our scientifc practices to be
efective and arrive at truths. Our transcendental argument established
that scientifc experiments cannot be efective and arrive at truths without
abilities for rational embodied control. Any true ontological naturalist
image of humans is therefore committed to the existence of these abilities.
A second argument is based on the Eleatic principle that what exists
is what plays a causal role or makes a causal diference.18 The ability to
exercise rational embodied control over physical phenomena can only
make a causal diference in reality if this ability is real, like either an
event-cause or a causal power whose causal manifestations make a dif-
ference in reality. If a scientist performing the Millikan experiment is not
making a diference to the movements of the oil droplets by charging
or discharging the plates by turning on or of the voltage; if instead the
changing movements of the oil droplets are somehow inexplicably occur-
ring and have no causal connection to the scientist’s intentional action of
turning on and of the voltage by moving her hand, which is fipping the
switch, which is charging or discharging the plates, which is etc., then
the scientist cannot know why some of the charged droplets of oil started
to move upward when and while this process was occurring. In short, if
scientists cannot confrm that their rational embodied control is causally
efcacious in their experiments, then scientifc experiments do not secure
facts and cannot verify or falsify scientifc theories. But we are correct to
reckon scientifc experiments secure facts and can verify or falsify scien-
tifc theories, because scientists’ abilities for rational embodied control are
causally efcacious in their experiments. And, because they are causally
efcacious, these abilities exist.
3.2. Rival Conceptions of the Ontological Nature of These
Abilities
These two arguments establish the existence of the abilities defended by
our Taylor-style transcendental argument. What is the ontological nature
of these existing abilities? Neo-Aristotelians maintain that if these abilities
exist, they exist as real powers of humans for rational embodied control
over physical phenomena. Why hold that for these contentions to come
out true, these abilities need to be real powers?
First, our Taylorian transcendental argument has cleared away from
the metaphysical table the major ontological alternatives to causal powers
insofar as the plausibility of these rival positions on causation and agency
require the assumptions of physicalism, which – as we will see – these
transcendental arguments undermine. Second, given the neo-Aristotelian
setting of this chapter, and that space does not permit us to rehearse the
extensive debates and arguments for why neo-Aristotelians defend and
many others have converted to power realism in recent decades, a pre-
sumption in favor of an ontology of powers is already on the table. If any
ability is real, then it is a power. It will nevertheless be worth surveying
some of the major problems faced by those courageous souls who have
tried to carve out an ontological place for mental causation and agency
within the diminished ontological space and resources permitted by physi-
calism. Later we will address why neo-Aristotelian causal power realism
does not face similar difculties.
Unlike neo-Aristotelian causal power realism, most positions in the
metaphysics of causation aim to downgrade or reduce the apparent real-
ity of human agents with causally efcacious rational embodied control
over physical phenomena. But these diminished accounts of causal agency
are deeply problematic. Many face the problem of deviant causal chains,
which shows that humans cannot exercise the right kind of ongoing causal
control over physical phenomena required for intentional actions, like
performing an experiment. The disappearance of the agent problem dis-
closes that the standard causal story of action deprives the human agent
of any active control over what the agent does, since it is event causes
within the agent that determine what happens. The standard story “fails
to cast the agent in his proper role. In this story, reasons cause an inten-
tion and an intention causes bodily movement, but nobody – that is, no
person – does anything” (emphasis in original).19 In philosophy of mind,
all eforts to defend the reality of mental causation confront the problem
of causal overdetermination. Since physical causes independent of any
mental causes are sufcient causes of physical efects – like performing an
experiment – there is no causal exigency for postulating mental causes.
In addition to problems with the philosophical coherence of these etio-
lated metaphysical theories of causation and agency, there are also fun-
damental difculties with the assumptions of physicalism that underwrite
them. These assumptions lead many philosophers to agree with Derk
Pereboom that “given our scientifc understanding of the world, how
could there exist anything as fabulous as an agent-causal power? It would
appear that our natural scientifc theories could not yield an account of
a power of this sort.”20 Indeed, the major motivation for adopting onto-
logical downgrading strategies in the frst place is based upon a funda-
mental faith in an austere physicalism that purportedly some present or
future theoretical physics can establish. The purported scientifc support
for physicalism leaves no place for real and causally efcacious rational
embodied control, which is why defenders of the reality of mental causa-
tion within this paradigm face an uphill battle of trying to secure any
form of mental causal efcacy over and above real physical properties
and causes.
Against the assumptions of physicalism stands our Taylorian transcen-
dental argument, which contributes to the existing battery of arguments
that reveal that “physicalism” is ill-conceived.21 In particular, these argu-
ments demonstrate that the presumed scientifc point of departure of
physicalism fundamentally misrepresents and overlooks the indispens-
able features of the manifest image of humans that are required for the
scientifc practices of physics. Theoretical physics cannot put in question
the reality of causally efcacious embodied rational control over physical
phenomena without undermining the very scientifc experimental results
required to vindicate its scientifc theories. No one should commence
their enquiries concerning human agency with the self-defeating physi-
calist assumptions that theoretical physics has already established mental
causation (probably) does not exist. Contrary to physicalism, the mani-
fest image remains the point of departure for our philosophical enquiries
concerning the ontological commitments required for human abilities for
rational embodied control. These Taylorian transcendental arguments
overturn the mistaken assumption that scientifc theories rule out ratio-
nal causal agency and that we must settle for downgraded conceptions
of rational agency instead. They correct the physicalist’s image of reality,
which assumes a distorted vision of both the manifest and the scientifc
images of humans and the world.
3.3. Power Realism

We have just presented arguments for setting aside the physicalist wor-
ries that motivate inherently problematic downgraded theories of causal
agency. In this subsection, I sketch an account of causal powers that
illustrates how powers fulfll the ontological, causal, and epistemological
explanatory roles that abilities for rational embodied control play in our
manifest, scientifc, and ontological images of humans.
We cannot move into this metaphysical territory concerning the nature
of powers to perform experiments independent from a host of other
presumed metaphysical commitments about what there is, what grounds
what, what causes what, and so forth. But I also hope that my argu-
ments intimate why none of these metaphysical issues can be confronted
and settled independent from addressing critical questions concerning
the ontological image of humans as psychological subjects who engage
in ontological enquiries that draw on manifest and scientifc images of
humans and the world. Put otherwise, the truth of a neo-Aristotelian
powers ontology is not going to be established independent from what
we know about the roles and reality of powers as humans who exercise
our own powers in concert with the co-manifestations of a panoply of
powerful particulars in nature.22 This point is not unique to neo-Aristo-
telianism, for the austere ontology of a neo-Humean mosaic also cannot
be established independently from a neo-Humean epistemological error
theory that justifes discounting aspects of the manifest and scientifc
images of humans and the world.
Even though we cannot rehearse the debates over powers, we do need
to say something about the view of causal powers being presumed and
how it supplies answers to the aforementioned three major problems that
plague its rivals, problems that are apropos the question of whether sci-
entists can perform experiments. There are a number of competing neo-
Aristotelian accounts of causal powers, but only some of these debates
matter for our purposes.23
First, because neo-Aristotelians hold that causal powers are funda-
mental, irreducible, frst-order properties of reality, they reject “categori-
calism,” which is the view that fundamental properties are not powers.
Second, power realism rejects Humean and other theories of causation
that hold all causation consists in temporally successive independent
events, that causation requires a necessary connection among events or
the instantiation of an exceptionless generalization or universal law of
nature. Power realism therefore rejects the standard causal story of action
as well. Neo-Aristotelians defend a realist theory of causation consisting
in simultaneous and reciprocal causal manifestations among causal power
partners.24 “[W]hen two causal partner-powers are mutually activated,
their two manifestations occur simultaneously in one event, but they are
two diferent types of activity.”25 It is the co-manifestations or co-activa-
tions of causal powers that make a causal diference in reality, not laws
of nature; laws of nature are, at best, abstract descriptions of the regular,
but defeasible, co-manifestations of causal powers.
Third, the co-manifestation of causal power partners is at the heart
of the neo-Aristotelian conception of powers and causation, and it
brings out the way diferent causal interactions or activities – which
are often ongoing causal processes26 – are comprised of diferent active
and passive causal roles flled by diferent power partners. 27 Fourth,
following more neo-Aristotelian power realists, like Anna Marmodoro,
the manifestations of powers do not essentially consist in the giving or
production of another power; rather the manifestation of a power is an
internal transition of one and the same power shifting from being in a
state of potentiality to a state of activation or of acting, manifesting,
and causing. Power partners are mutually activated whenever they are
co-manifesting.28 Fifth, strictly speaking the causal agents and patients
of action and passion interactions are typically substances or powerful
particulars that realize or exercise their powers in concert with their
causal power partners.29
Sixth, one of the central explanatory strengths of power realism over
rival explanations of causation is its elegant account of counterfactuals
and the defeasibility of occurrences and its attendant arguments, which
many believe have refuted most, if not all, attempts to explanatorily reduce
powers or dispositions.30 The co-manifestations or causal occurrences of
reciprocal powers are always liable to being prevented, infuenced, modu-
lated, or otherwise altered by subtractive and additive interferences, fnks,
masks, antidotes, blockers, absences, and by any other counteracting con-
founds or more powerful powers.31 This means in any individual case we
need to identify and distinguish a power’s canonical way of acting and
so also its canonical power partners and their canonical ways of acting,
which give us their canonical ways of co-acting or co-manifesting, from
the outcome or what happens when these canonical co-manifestations of
power partners are modulated or interfered with by the manifestations of
other powerful powers.32
Seventh, given the varieties of causal activities and canonical active
and passive causal roles among any multitude of co-manifesting pow-
ers, along with the variety of powerful powers that can interfere with
or modulate the outcome of co-manifesting powers, neo-Aristotelian
power realism is thereby committed to causal pluralism. This is signifcant
since it also rejects most of the basic assumptions taken for granted by
proponents of causal monism and sufciency, which generate the afore-
mentioned problem of overdetermination for mental causation.33 For
example, the account of event causation presumed by Jaegwon Kim’s
principle of explanatory exclusion, that “No event can be given more
than one complete and independent explanation,” does not consider two
central features of neo-Aristotelian causal power realism. First, since any
power’s manifestations are defeasible, there can be a manifesting power
that would be, ceteris paribus, a sufcient cause for an efect, but if a
second power manifests it can prevent or modulate the manifestation-
outcome of the frst otherwise sufcient cause. Second, strictly speak-
ing these two manifestation-outcomes are not the same, as disjunctivists
rightly point out. And neo-Aristotelian causal power realists can explain
why the outcomes or efects are not the same since they are the results
of distinct nexus of co-manifesting powers. Neo-Aristotelian causal plu-
ralism therefore endorses a quite diferent understanding of the causal
exclusion principle and of the problem of causal overdetermination. An
efect, like a bodily movement, which results from the manifestation of
nonrational instinctual causal powers, is simply a diferent efect from
any bodily movements that results from rational causal powers whose
co-manifestations with powers for bodily movement can override or over-
power nonrational instinctual causal powers.
This means that rational mental causation is not essentially problematic
for neo-Aristotelians, whereas it presents a notoriously insurmountable
problem for anti-Aristotelians in the philosophy of mind who face the fol-
lowing standard dilemma. Either contend that the normativity of rational-
ity and its genuine mental causation can be epistemologically naturalized
and reduced to physical causes (where many philosophers, if not most,
agree all promissory attempts to naturalize rational normativity by reduc-
ing it to causal co-variation have failed). Or, contend that reasons are
genuine mental causes of physical efects and are distinct from the physical
events that are already independent sufcient causes of the same physical
efects. But this horn introduces the problem of overdetermination, which
renders all mental causes superfuous. Neither horn of the dilemma can
secure the rational embodied control required for scientifc experiments
detailed by our Taylorian transcendental arguments. That this dilemma
faces every major non-Aristotelian position in the philosophy of mind
provides one more example of the explanatory superiority of the neo-
Aristotelian position over its rivals.34
Eighth, following Helen Steward, I distinguish between the realization
of one-way powers from the exercise of two-way powers.35 A negatively
charged electron’s one-way power for electromagnetic attraction is real-
ized whenever its necessary conditions for manifestation obtain, but an
agent empowered with a two-way power exercises its two-way power by
φ-ing or refraining from φ-ing. When a scientist reenacting the Millikan
experiment today intentionally and voluntarily fips a voltage switch in
order to alter the movement of ionized oil droplets, she is exercising her
two-way power for rational embodied control.
Ninth, powers are identifed by their manifestations and so a two-way
power exercised by an intentional and voluntary action is identifed by
the details of an accurate philosophical analysis of human action. This is
another way of articulating Aristotle’s heuristic in De Anima II.4 that we
know natures by their powers, and powers by their activation or opera-
tions, and operations are known by their objects. The neo-Aristotelian
approach to this two-way power is therefore signifcantly diferent from
and arrives at a remarkably diferent account of human action as an
explanandum and of its two-way power explanans than so-called agent-
causation or source incompatibilism theories. The latter two, as their critics
point out, are eclectic retreats or revolts (however one sees it) that remain
captive to an essentially anti-Aristotelian ontology. When a human per-
forms an intentional and voluntary action, that is, to put it simply, when
a human acts she does not cause her action, and her action is not an efect
of her as an agent-cause. Rather, following Steward, Alvarez, Hyman, and
others, her acting is her causing.
To act is to cause some kind of change, the agent being the one that
causes the change and the patient being the one that undergoes it, the
kind of act depending on the kind of change. Agents and patients are
individual substances, such as a particular cat or man or stone, or else
particular quantities of a kind of material, such as a particular vial of
poison or a particular ounce of gold. But “agent” and “patient” do
not refer to diferent kinds of entities, but to complementary roles.36
On this neo-Aristotelian account of an agent exercising her two-way causal

power for acting or refraining from acting, the disappearance of the agent
problem does not even arise. For it is the agent herself that settles what
she is doing, not some concatenation of causes happening within her.37
Accordingly, whenever a scientist exercises her two-way causal power by
acting or refraining from acting, she is causing or refraining from causing.
And since her causation—like the other manifestations of causal pow-
ers—occurs in concert with the activation of reciprocal causal powers,
whenever she is acting she is thereby also co-manifesting her exercised
two-way causal power with other causal powers. By intentionally fipping
the voltage switch on a power supply, she is causing the co-manifestation
of coordinated passive causal roles of the movements of her fngers, hand,
wrist, arm, and so on, each of which also thereby plays an active causal
role in the process of her embodied rationally controlled fipping of the
voltage switch. With respect to the voltage switch itself, its passive causal
role of being on or of is being actualized by the active causal process of
the scientist’s embodied rationally controlled fipping of the voltage switch,
which enables the switch’s active causal role within the chain of active and
passive causal roles and powers involving capacitors, inductors, varistors,
transformers, and whatever else is employed for charging or discharging
the plates within the lower chamber of the Millikan oil-drop experiment.
In sum, the neo-Aristotelian aims to get the correct account of the
descriptive and normative features of human action without skipping out
on or evading a serious metaphysical explanation of how human actions,
as the exercise of causal powers, make a causal diference in reality. Many
more ontological details and arguments would be required to outft this
picture properly, but I will turn now to the problem of deviant causal
chains and conclude with an explanation for why this neo-Aristotelian
causal power theory of human action is not beleaguered by this problem.
3.4. The Problem of Deviant Causal Chains

Let us start with why the problem of deviant causal chains arises for
the standard causal theories of human action, often called “the standard
story,” a problem that many regard as a decisive objection against it.38
The standard story holds that intentional actions are events, paradigmati-
cally movements of the agent’s body, that are caused in the right way by
antecedent mental causes, such as by intentions or other pro-attitudes. We
only identify the event of some bodily movement as an “action” because
it was caused in the right way by its antecedent beliefs and desires. The
problem for the standard story is explaining how the antecedent mental
events can cause the bodily movements in the right way given its commit-
ments to two theses about causation and action. First, there is no logical
or categorial connection between the successive events of cause and efect.
Second, there is nothing intrinsic or essential to the bodily movements that
are actions that distinguish them from mere bodily movements. Given
these standard story commitments, all attempts to explain this in the right
way are open to counterexamples known as deviant causal chains.
Donald Davidson’s famous example is that of a climber who desires to
remove the dangerous weight of holding a fellow climber on a rope and
who also believes that he could loosen his grip on the rope and remove
this burdensome load; however, his belief and desire startle him so much
that they cause his grip on the rope to be loosened. Clearly this bodily
movement is not an intentional action, but since it was caused by the
climber’s belief and desire (or intentions on other versions), the standard
story has no way to distinguish this bodily movement from bodily move-
ments that are actions caused in the right way. Davidson admitted his own
version of the standard story did not have the resources to meet this devi-
ant causal chains objection, and while there is not a universal consensus,
there is widespread agreement that this objection is fatal to the standard
story’s neo-Humean causal theory of action.39
As with the problems of causal overdetermination and the disappear-
ance of the agent, this problem is relevant to the reality of the abilities
for performing scientifc experiments. Consider the difculties that arise
if one endorses the standard causal story as an explanation of such abili-
ties instead of neo-Aristotelian power realism. Given the abject failure
of the standard story to answer the deviant causal chains objection, this
entails that the standard story fundamentally fails to explain any actions,
including the intentional action of performing a scientifc experiment. We
have established, independently of rival ontological theories of causation,
that performing a scientifc experiment is an intentional action constituted
by rational embodied control. But the standard story cannot distinguish
the rational embodied control of actions from mere bodily movements,
that is, the standard story cannot distinguish the controlled action of
intentionally loosening one’s grip on a rope from the uncontrolled bodily
movement of one’s grip being loosened caused by being startled by one’s
beliefs and desires or intentions. Consequently, the standard story fails
to provide an ontology of causation that can explain how the ability to
exercise rational embodied control can really make a causal diference
in the right way that constitutes the intentional actions of performing
scientifc experiments.
The problem of deviant causal chains does not even arise for the neo-
Aristotelian causal powers explanation of human action. There are many
other important points of diference, but I shall stick to three. First, neo-
Aristotelianism starts by detailing a fundamentally diferent notion of
action. Actions are intrinsically and essentially distinct from mere bodily
movements. This is why, for neo-Aristotelianism, the climber’s bodily
movements that involved his grip on the rope being loosened by being
startled are intrinsically distinct from an intentional action of loosening
one’s grip on a rope.
Second, as noted earlier, intentional actions are not caused by the agent
or its powers, so actions are not efects of some extrinsic prior mental
event-cause or agent-cause. Rather, the agent’s exercising of his or her
two-way power is the agent’s manifesting its two-way power, and this
manifesting is the agent’s acting or causing. As John Hyman argues at
length, “To act is to intervene, to make a diference, to make something
happen, to cause some kind of change.”40 Consequently, there is no prob-
lem here of trying to explain how to get the action caused in the right way,
since “actions” are not caused. An intentional action itself is a causing; an
action consists in an agent’s intentionally making a diference or causing
some kind of change.
Third, neo-Aristotelians reject both of the aforementioned planks of
the standard story. When powers co-manifest, they manifest via diferent
reciprocal active and passive causal roles, and these co-manifestations
constitute either one simultaneous event or one unifed ongoing process or
action-passion interaction among agents and patients. So there is not only
a categorial connection between cause and action insofar as intentional
acting is itself an intentional causing, but there is also a categorial con-
nection between cause and being caused. And, on one neo-Aristotelian
view, an efect simply is the outcome of a total nexus of reciprocal powers
co-manifesting in their diferent causal roles as active and passive causes.
Someone might raise the objection that the problem of deviant causal
chains has a diferent form for neo-Aristotelians, insofar as the manifesta-
tion of the climber’s intention to loosen his grip could startle the climber
so much that the manifestation of the power to startle overpowers the
climber’s intention and thereby becomes the dominating active cause in
the loosening of the grip on the rope. This objection, however, misun-
derstands that actions are intrinsically actions, and insofar the agent’s
intentions are thwarted or prevented from manifesting, say, by a more
powerful startle, then no “human action” occurs or a failed action of
trying happens. Accordingly, a neo-Aristotelian powers ontology has a
causal explanation for why its account of intentional action is commit-
ted to a form of disjunctivism. For “human actions” – like intentionally
loosening one’s grip on a rope – are individuated by the active-passive
causal constitution of the co-manifestation of reciprocal powers. If the
active causal role is not fulflled by the manifestation of a power for
intentional action, then no human action has occurred. Consequently, the
“human action” of intentionally loosening one’s grip constituted by bodily
movements that co-manifest as a passion-action event with the power for
intentional action is fundamentally diferent from the mere “human behav-
ior” of a startled grip being loosened. Since this latter human behavior is
constituted by the co-manifestation passion-action event of the powers for
bodily movements and being startled by one’s thoughts.
A related problem for the standard story is that its exclusive focus on
antecedent mental causes as the defning feature of actions41 “makes it
impossible for them to give any account whatever of the most salient
diferentiating characteristic of action: during the time a person is per-
forming an action he is necessarily in touch with the movements of his
body in a certain way.”42 As we have seen, ongoing causal control over
phenomena is essential for performing scientifc experiments, and it is part
and parcel of the neo-Aristotelian explanation of activities and processes
that they consist of the ongoing co-manifestations of causal powers. But
the standard story fails to explain this quotidian feature of action and
so also fails as a causal explanation of the ability to perform scientifc
experiments.
I have argued, contrary to radical revisionist positions, that the scientifc
image of humans and the world cannot dispense with the manifest image
of humans because our scientifc image depends upon human scientists’
abilities to perform scientifc experiments. I started with a Taylorian tran-
scendental argument that established the abilities to perform scientifc
experiments are simply instances of the manifest image of the human
abilities to exercise rational embodied control over physical phenomena.
The manifest image of humans is therefore indispensable to all scientifc
images of the world and of humans, which depend upon the discoveries
and results of scientifc experiments. But this transcendental argument
does not give us an ontology; that takes another step. I then presented
two arguments for why these abilities need to exist in order to fulfll
their causal explanatory roles as indispensable to performing scientifc
experiments. Next I argued that these existing abilities are best explained
by a causal powers ontology and showed why power realism does not
face such insurmountable problems as overdetermination, disappearance
of agents, and deviant causal chains, which confute the standard causal
story of human action and elucidate why it cannot provide an ontological
explanation of the abilities for rational embodied control.
One fnal worry concerning fundamentality merits a few concluding
remarks. Even if the scientifc image must concede that these causal powers
from the manifest image are ontologically indispensable, this chapter’s
arguments do not preclude the possibility that the scientifc image might
discover that human agents cannot be fundamental entities within the
ontological image of the world. Human agency might be real, but it might
only be a derivative mode of being that is ontologically grounded in the
fundamental particles of physics.
While the neo-Aristotelian image of humans in the world does not
ignore the dependency of human agency and other powerful particulars
on basic ontological powers of the physical stuf of the cosmos, it also
rejects the ontological thesis that only completely ungrounded entities
can be real substances or fundamental entities. Following the Eleatic
principle, it regards causation as an ontological guide to the existence of
the basic substances and properties of reality, even if all properties are
grounded in substances. The arguments of this chapter have shown that
the question of what grounds what cannot be answered independently
from establishing the existences of the whats. For all of the indispensable
ways fundamental particles make an ontological diference in reality, we
could not even discover and establish the existence of these basic bits and
bobs as framing features of our scientifc image of the world if human
scientists did not have the causally efcacious scientifc know-how, say,
to collide beams of protons together to manifest these fundamental par-
ticles and their hidden powers. In short, establishing the existence of the
fundamental particles of physics presupposes the existence of the real
irreducible and ineliminable center of human agency that can exercise
efcacious causal powers over and above whatever happens to be the dis-
positions of the fundamental particles that scientists cause to co-manifest
in their experiments. Hence, even if it could be shown that fundamental
particles comprise an ungrounded ground, this ontological image of an
ungrounded ground could not exclude the basic irreducible existence and
causal efcacy of humans as agents for real change in nature. And this is
a scientifc and ontological image of humans in the world that is perfectly
compatible with the neo-Aristotelian manifest image of humans in the
world empowered with the power to perform experiments.
Notes
I would like to thank Rob C. Koons, Hasok Chang, William M. R. Simpson,
Brandon Dahm, and Fr. Thomas Davenport O.P. for helpful feedback, com-
ments, and corrections on earlier drafts of this chapter.
1. Wilfrid Sellars, “Philosophy and the Scientifc Image of Man,” in Science,
Perception, and Reality (Ridgeview Publishing Company, 1963), pp. 1–40,
p. 39.
2. Exponents of eliminative and reductive physicalism are the primary advocates
of the radical replacement of the manifest image by the scientifc image of
humans. But radical replacement is also entailed by all those versions of
nonreductive physicalism, like anomalous monism and realization functional-
ism, which Jaegwon Kim and others have demonstrated cannot consistently
maintain both physicalism and the real causal efcacy of reasons or rational
agency. See Jaegwon Kim, Physicalism, or Something Near Enough (Princeton
University Press, 2008). For eliminativist and reductionist views, see Daniel
Wegner, The Illusion of Conscious Will, 2nd ed. (MIT Press, 2018); John
Bickle, Philosophy and Neuroscience: A Ruthlessly Reductive Account (Klu-
wer Academic Publishers, 2003), ch. 3; P. S. Churchland, Brain-Wise (MIT
Press, 2002), ch. 5; Paul Churchland, Neurophilosophy at Work (Cambridge
University Press, 2007). For a helpful summation and critique of forms of
reductionism employed in cognitive neuroscience, see A. Chemero & C. Hey-
ser, “Methodology and Reduction in the Behavioral Neurosciences: Object
Exploration as a Case Study,” in The Oxford Handbook of Philosophy and
Neuroscience (Oxford University Press, 2009), pp. 68–90. For detailed cri-
tique of reductive accounts of agency, see Erasmus Mayr, Understanding
Human Agency (Oxford University Press, 2011).
3. Nancy Cartwright, Nature, the Artful Modeler: Lectures on Laws, Science,
How Nature Arranges the World and How We Can Arrange It Better (Open
Court, 2019), ch. 2.
4. Nancy Cartwright, The Dappled World (Cambridge University Press, 1999).
See also infra nn. 23–24.
5. Helen Steward, A Metaphysics for Freedom (Oxford University Press, 2012),
chs. 2–3.
6. Alasdair MacIntyre, “What Is a Human Body?,” in Tasks of Philosophy
(Cambridge University Press, 2006), pp. 86–103.
7. Charles Taylor, The Language Animal: The Full Shape of the Human Linguis-
tic Capacity (Harvard University Press, 2016).
8. B. P. Abbott et al. “Observation of Gravitational Waves From a Binary Black
Hole Merger,” Physical Review Letters, 116.061102 (11 February 2016): 4.
9. Rom Harré, “What Is Real in Psychology: A Plea for Persons,” Theory and
Psychology, 2.2 (1992): 153–158, p. 154. See also, Rom Harré, “New Tools
for Philosophy of Chemistry,” HYLE – International Journal for Philosophy
of Chemistry, 20 (2014): 77–91.
10. I must thank Fr. Thomas Davenport, O.P., for his detailed feedback on my
presentation of the Millikan experiment, and for drawing my attention to a
common discrepancy found in discussions of Millikan experiments. There are
scores of online video demonstrations of the “Millikan experiment” that use
power supplies with an adjustable voltage dial, but Millikan himself ended
up opting for an on-of voltage switch. I shall be describing the experimental
apparatus employed by Millikan in his experiment; however, nothing philo-
sophically signifcant hangs on this.
11. Thomas Kuhn, Structure of Scientifc Revolutions, 2nd ed. (University of
Chicago Press, 1970), p. 27.
12. For an excellent discussion of the Millikan experiment, see Cartwright,
Nature, The Artful Modeler, ch. 1.
13. See Hilary Putnam, Representation and Reality (MIT Press, 1988), p. 60.
reference and truth are the fundamental notions of the fundamental exact
science: the science of logic. Why don’t the eliminationists speak of “folk
logic” as well as of “folk psychology”? I once put just this question
to Paul Churchland, and he replied, “I don’t know what the successor
concept [to the notion of truth – H. P.] will be.” This is honest enough!
Churchland is aware that the notion of truth is in as “bad shape” as the
notion of belief from his point of view, and accepts the consequence: we
must replace the “folk” notion of truth by a more scientifc notion. But
the innocent reader of Churchland’s writings is hardly aware that he is
also being asked to reject the classical notion of truth!
Similar objections can be raised against non-eliminativist error-theoretic
forms of Humeanism. Stroud reads Hume as a kind of non-eliminativist
error theorist wherein all our basic beliefs about causation, for example, are
fundamentally mistaken (i.e., error theory), but we can do nothing to avoid
having these beliefs about causation in nature (i.e., non-eliminativism); see
Barry Stroud, Hume (Routledge, 1977), ch. 10 (esp. p. 248). It is worth
nothing that others reject this error-theoretic interpretation of Hume; see
Helen Beebee, “Hume and the Problem of Causation,” in P. Russell (ed.),
The Oxford Handbook of Hume (Oxford University Press, 2016), ch. 12.
For error theory eliminativism on rational control for agents, see Richard
Double, The Non-Reality of Free Will (Oxford University Press, 1991), ch. 8
and pp. 220–221; Gregg Caruso, “Free Will Eliminativism: Reference, Error,
and Phenomenology,” Philosophical Studies, 172.10 (2015): 2823–2833.
14. Charles Taylor, “The Validity of Transcendental Arguments,” Proceedings
of the Aristotelian Society, 79 (1979): 151–165, p. 165:
Transcendental arguments thus turn out to be quite paradoxical things.
I have been asking here what arguments of this kind prove, and how
they prove it. They appear to be rather strange in both these dimensions.
They prove something quite strong about the subject of experience and
his place in the world; and yet since they are grounded in the nature of
experience, there remains an ultimate, ontological question they cannot
foreclose – for Kant, that of the things in themselves; for the thesis of
embodied agency, the basic explanatory language of human behaviour.
When we ask how they prove what they prove, we see another paradox-
ical mixture. They articulate a grasp of the point of our activity which
we cannot but have, and their formulations aspire to self-evidence; and
yet they must articulate what is most difcult for us to articulate, and
hence are open to endless debate. A valid transcendental argument is
indubitable; but it is hard to know when you have one, at least one with
an interesting conclusion. But then that seems true of most arguments
in philosophy.
15. There are number of reasons for why the manifest image cannot be a lad-
der we kick away once we arrive at the scientifc image. First, because, like
the manifest image, the scientifc image is asymptotic ideal of fnite human
enquiries. Second, if the ladder metaphor is taken in an eliminativist way,
then there is the onus of establishing that there is an eliminativist scientifc
image at the other end of the ladder, for example, like some radically revised
and more explanatory notion of truth. So far, no plausible proposals even
exist. If the metaphor is taken in a reductionist way, then the ladder is not
really kicked away, unless some further argument can show why the mani-
fest image is wholly or partially falsifed and needs to be replaced. This is
because most forms of inter-theoretic reduction involve theoretical take over,
not the falsifcation of the less explanatory theory. But again, so far, even
these reductionist proposals are mere explanatory promissory notes, that is,
reductionist theorists’ pipe dreams based on a positivist naturalism’s faith in
some scientifc explanatory eschaton.
The more fundamental problem is the fact that the manifest image is itself
not a theory, even if theorists often confuse their own theories about com-
monsense, manifest images, or folk psychology, for what folk in fact think.
So the very suggestion of theoretical reduction is out of place without some
radically revised notion of reductionism that can overcome the cogent argu-
ments of Wittgensteinians, Heidegger, Michael Polanyi, Charles Taylor, and
Alasdair MacIntyre that no such theory, folk or otherwise, can be given that
exhaustively thematizes our embodied coping and everyday tacit practices.
In short, the kicked away ladder promissory note has enormous explanatory
burdens it has not faced. I thank Hasok Chang for raising this objection.
16. Jonathan Schafer, “On What Grounds What,” in David Manley, David J.
Chalmers & Ryan Wasserman (eds.), Metametaphysics: New Essays on the
Foundations of Ontology (Oxford University Press, 2009), pp. 347–383.
17. William Jaworski, Structure and the Metaphysics of Mind (Oxford University
Press, 2016), p. 19.
18. Jaworski, Structure and the Metaphysics of Mind, pp. 29–32.
19. David Velleman, “What Happens When Someone Acts?,” Mind 101 (1992):
461–481, p. 461. Velleman distinguishes the action theory problem of the dis-
appearing agent from the mental causation mind-body problem, p. 469. See
also Jennifer Hornsby, “Agency and Actions,” Royal Institute of Philosophy
Supplement, 55 (2004): 1–23; Steward, A Metaphysics for Freedom; Derk
Pereboom, Free Will, Agency, and Meaning in Life (Oxford University Press,
2014).
20. Pereboom, Free Will, Agency, and Meaning in Life, p. 69.
21. Tim Crane and D. H. Mellor, “There Is No Question of Physicalism,” Mind,
99.394 (1990): 185–206; Carl Hempel, “Reduction: Ontological and Lin-
guistic Facets,” in S. Morgenbesser et al. (eds.), Essays in Honor of Ernest
Nagel (St Martin’s Press, 1969), pp. 189–207; Robert C. Koons and George
Bealer (ed.), The Waning of Materialism (Oxford University Press, 2010).
22. For arguments that only an active being could acquire causal concepts, see
G. H. von Wright, Causality and Determinism (Columbia University Press,
1974), p. 52; P. M. S. Hacker, Human Nature: The Categorial Framework
(Blackwell, 2007), ch. 3; S. Mumford & R. L. Anjum, Getting Causes From
Powers (Oxford University Press, 2011), ch. 9.
23. For some of the relevant literature, see R. L. Anjum & S. Mumford, What
Tends to Be: The Philosophy of Dispositional Modality (Routledge, 2018);
Cartwright, The Dappled World; Travis Dumsday, Dispositionalism and the
Metaphysics of Science (Cambridge University Press, 2019); Brian Ellis, Sci-
entifc Essentialism (Cambridge University Press, 2001); S. French, A., Bird,
B. Ellis, S. Mumford, & S. Psillos, “Looking for Laws,” Metascience, 15
(2006): 437–469; R. Grof & J. Greco, Powers and Capacities in Philosophy:
The New Aristotelianism (Routledge, 2013); Jonathan Jacobs, Causal Powers
(Oxford University Press, 2017); Anna Marmodoro (ed.), The Metaphysics
of Powers: Their Grounding and Their Manifestations (Routledge, 2010).
24. Anna Marmodoro, “Aristotelian Powers at Work: Reciprocity Without Sym-
metry in Causation,” in Jacobs (ed.), Causal Powers; C. B. Martin, The Mind
in Nature (Oxford University Press, 2008); Mumford & Anjum, Getting
Causes From Powers; Jaworski, Structure and the Metaphysics of Mind.
25. Marmodoro, “Aristotelian Powers at Work,” p. 71.
26. Steward, Metaphysics for Freedom, ch. 3.
27. Marmodoro, “Aristotelian Powers at Work,” pp. 72–75.
28. Marmodoro, “Aristotelian Powers at Work.”
29. On the interrelatedness of both agents and events for causation, see Steward,
Metaphysics for Freedom, ch. 8; Hacker, Human Nature, ch. 3; Erasmus
Mayr, Understanding Human Agency (Oxford University Press, 2011), chs.
8–9; John Hyman, Action, Knowledge, and Will (Oxford University Press,
2015), chs. 2–3.
30. For an extensive summary of the argumentative case against such reduc-
tive simple, causal, and Lewis’s reformed conditional analyses of powers,
see Mayr, Understanding Human Agency, chs. 6–7, see also the references in
supra nn. 23–24.
31. See references in nn. 23–24 and Stephen Mumford, “Laws and Their Excep-
tions,” in Walter Ott & Lydia Patton (eds.), Laws of Nature (Oxford Univer-
sity Press, 2018), ch. 11.
32. See Nancy Cartwright & Pedro Merlussi, “Are Laws of Nature Consis-
tent With Contingency?,” in Ott & Patton (eds.), Laws of Nature, ch. 12,
pp. 239–243.
33. Jaegwon Kim, “Mechanism, Purpose, and Explanatory Exclusion,” Philo-
sophical Perspectives 3 (1989): 77–108, p. 79. I thank Rob C. Koons for
urging me to clarify my view on this problem.
34. On the hard problem of content against naturalized epistemology see Dan-
iel D. Hutto & Erik Myin, Radicalizing Enactivism: Basic Minds Without
Content (MIT Press, 2016). On the problem of overdetermination, see Kim,
“Mechanism, Purpose, and Explanatory Exclusion”; Jaworski, Structure and
the Metaphysics of Mind.
35. Steward, A Metaphysics for Freedom, pp. 155–156. On two-way powers,
see also Anthony Kenny, Will, Freedom and Power (Blackwell, 1975); Maria
Alvarez, “Agency and Two-Way Powers,” Proceedings of the Aristotelian
Society. New Series, 113 (2013): 101–121; Hacker, Human Nature, chs. 3–7;
Mayr, Understanding Human Agency.
36. Hyman, Action, Knowledge, Will, p. 42.
37. See Steward, A Metaphysics for Freedom, pp. 62–66.
38. See Mayr, Understanding Human Agency, ch. 5; Steward, A Metaphysics for
Freedom, pp. 55–69.
39. Donald Davidson, “Freedom to Act,” in Essays on Actions and Events,
2nd ed. (Oxford University Press, 2001), p. 79. For a helpful comparison of
Davidson’s and other versions of the standard story with Anscombe’s rival
account of action, see Frederick Stoutland, “Introduction: Anscombe’s Inten-
tion in Context,” in Anton Ford, Jennifer Hornsby, & Frederick Stoutland
(eds.), Essays on Anscombe’s Intention (Harvard University Press, 2011), pp.
1–22.
40. Hyman, Action, Knowledge, Will, p. 33.
41. Steward has demonstrated that this focus on the necessity of mental anteced-
ents to action is itself problematic; see Steward, A Metaphysics for Freedom,
pp. 66–69.
42. Harry G. Frankfurt, “The Problem of Action,” American Philosophical
Quarterly 15.2 (1978): 157–162, p. 158.
8 Logical and Mathematical
Powers
Antonio Ramos-Díaz
Many traditional arguments for the immateriality of the intellectual pow-

ers of human beings have tended to focus, primarily, on the content and
universality of conceptual understanding. Aristotelian immateriality argu-
ments in particular trade on our capacity to abstract the essence of things
and form universal concepts. Platonic arguments, in a similar manner,
tend to focus on the universality and other features of the content of
our thoughts derived from the Ideas or Forms, which cannot come from
anything material. We shall call such arguments universality and content-
based arguments for the immateriality of the intellect.1 Mathematics can
play an important role in these traditional arguments, insofar as math-
ematical phenomena (e.g., numbers, geometrical objects, sets, infnitary
functions, etc.) have features that impede not only their identifcation
with material phenomena but also their becoming the content of any
purely material power. But there is a diferent line of argument that does
not trade primarily on the conceptual content or universality of math-
ematical thought and objects but on the form or structure of mathematical
thought. We shall call such arguments formal-structural arguments for the
immateriality of the intellect.
The present chapter will argue that our mathematical and logical pow-
ers display features that cannot be explained in materialist terms.2 It will
develop not a universality or content-based argument but a formal-struc-
tural argument for the immateriality of our mathematical and logical
operations and powers. In particular, it argues that the actualization or
activity of those powers consists in the actualization of forms or structures
that cannot be actualized determinately by material powers alone. This
actualization occurs whenever we do mathematics or logic. Admittedly,
the expression “doing mathematics” can pick out an array of diferent
activities and tasks, since mathematical and logical practice is wide, com-
plex, and varied, with specialized felds and with ordinary objectives in
the context of everyday reasoning. But an interesting aspect of our math-
ematical and logical activity is the application of abstract rules and the
carrying out of mathematical or logical operations (including, but not
limited to, computations). It is this aspect of human mathematical and
DOI: 10.4324/9781003125860-11
Logical and Mathematical Powers 221
logical activity that I shall argue cannot be a wholly physical or material
process. The aim of this chapter is to show that these types of powers
cannot be explained either constitutively or genetically (cf. Nagel 2012)
in terms of material processes alone.
In the frst section of the chapter, I develop and defend an argument
I call the Kripke-Ross Argument (“KRA”),3 which establishes the non-
physicality of our logical and mathematical operations and powers. In the
second section, I consider some intuitive and pressing objections to the
argument. In the third section, I address some possible misunderstand-
ings that may arise in interpreting the argument, in particular its use of
“form”. Finally, I conclude with a brief consideration of KRA’s place
within an Aristotelian framework.
1. The Argument: KRA

I formulate the argument as follows:
KRA.
A. Logical-mathematical activity is formally determinate.4
B. No physical process or mechanism is formally determinate.
C. Ergo, logical-mathematical activity is not (entirely) a physical
process.
The second premise is taken from Kripke’s unpublished arguments against

computational and functionalist theories of the mind, though it is admit-
tedly stronger than anything Kripke explicitly said.5 The frst premise is
taken from James Ross’s argument against physicalist theories of human
formal understanding, modifed here in terms of acts or activities. In what
follows, “formal operations” is used as shorthand for mathematical or
logical operations. The notion of “form” at play here will be made clear
later on. The term “pure function” will also be used, in keeping with
Ross’s own use of the terms, to refer to infnitary mathematical (or logical)
functions.6 We shall use both expressions interchangeably. In developing
the argument, we have limited ourselves to basic mathematical and logi-
cal operations, since if the following considerations apply to the basic
building blocks of advanced and complex mathematical activity, they
will apply to such an advanced and complex mathematical activity too.
1.1. The Formal Indeterminacy of the Physical

To say that nothing physical is or can be formally determinate is to
say that the physical cannot be in any intrinsic way an instance of a
determinate formal operation. By “formally determinate” we mean the
requirement that the realizer(s) of a formal operation be, on account of
222 Antonio Ramos-Díaz
its own intrinsic powers, metaphysically sufcient for the determinate
realization of a unique and defnite formal operation, so as to exclude
incompossible ones from being equally well realized at the same time and
in the same respect. Incompossible functions are functions that might be
indistinguishable from each other for a subset of their argument-value
(or input-output) relations but incompatible for the totality of their argu-
ment-value relations. For instance, Kripke’s (Kripke 1982, 9 f.) well-
known incompossible addition and quaddition functions yield the very
same values (or output) for any arguments (or input) less than or equal
to 57, divergent values otherwise.7 Our claim is that physical processes
and mechanisms necessarily satisfy incompossible functions and therefore
are formally indeterminate. (As will become clear later in the chapter, and
contrary to Kripke’s skeptic, this argument cannot be coherently extended
to our intellectual powers.) This is no small thing: if physical proper-
ties and structural patterns do not sufce to metaphysically determine
defnite mathematical or logical operations, then which formal structures
a physical mechanism realizes cannot be something determined by the
physical structure of the mechanism alone. The argument for the formal
indeterminacy of the physical goes as follows:
1. Take any physical object or mechanism whatsoever, call it φ, and

constrain it however much you’d like so that it cannot count as com-
puting every function and so that we can label exactly its properties
and behavior. (No triviality problems and no labelling problems.)8
2. Suppose φ is a physical system whose purely physical components,
changes, and processes compute a pure function f (say, the addition
or the identity function).
3. Then there will be at least one other incompossible pure function
g which all the physical changes, processes, and facts of φ equally
well satisfy and realize.
4. Thus, φ cannot exclude incompossible functions from being equally
realized or satisfed.
5. But incompossible pure functions, such as f and g, by defnition
exclude one another from being equally satisfed or realized at the
same time and in the same respect.
1. For, in part, a genuine case of a pure function is determinate
with respect to which counterfactuals are true of it. But a case
that is both genuinely of f and genuinely of g at the same time
and in the same respect will equally satisfy incompossible, that
is, contradictory, counterfactuals – which is absurd.
6. Therefore, since nothing about φ can exclude incompossible pure
functions from being equally well realized or satisfed, it follows
that no such pure function is/can be realized or satisfed by φ.
7. Therefore, no φ can realize a pure function (i.e., φ is formally
indeterminate).
Consider the following simple electrical circuit, made to compute the
identity function (Stabler 1987):
[I] – – – – – – – – – – – – [O]9
This simple circuit takes sequences of voltage pulses at one end (input
source) and conducts them to the other end (output source). Using
unary notation, sequences of voltage pulses can be mapped onto the
domain and range of the identity function to the efect that the circuit
reliably computes the identity function, which is infnite. Any physical
implementation of this circuit, however, will be constrained by innu-
merable physical factors: physical laws and necessities of nature, tem-
poral constraints, fnitary processing capacity (e.g., quantity of voltage
pulses, current conduction speed, etc.), decay or wear, malfunction,
manufacturing, and so on. Yet, if built appropriately and if the physical
circuit sufers no malfunction or decay, absent external impediments it
will compute the identity function for as long as it can. Stabler believes
that this simple example, which can be extended to more complex
infnitary formal operations, shows how the structure of a physical
mechanism can instantiate determinately and objectively the structure
of a formal operation. The objective of Stabler’s analysis is to exclude
the sort of formal indeterminacy Kripke argued physical computing
mechanisms are inevitably subject to: for any pure function f that a
physical computing mechanism P is said to compute, there will be at
least some other incompossible quus-like pure function f*, incompos-
sible with f, that all the physical properties of P satisfy at the same time
and in the same respect.
Unfortunately for Stabler, any purported physical instantiation of
that simple circuit will satisfy some nonstandard formal operation.10
Here’s why. The conductor will be limited in its operative capacity,
being capable of conducting only fnitely many sequences of voltage
pulses. Likewise, the input and output sources will have limited, fnite
processing capacities. No matter what materials are used, what physical
properties the input-output ports and the conductor are made of, lim-
ited lifespan and limited operative capabilities guarantee that a physical
mechanism cannot compute every possible argument of an infnitary
mathematical function. Let P be a physical instantiation of the identity
computing circuit diagram. Now suppose P’s conductor will, for the
sake of argument, decay or wear out after 10 sequences of voltage
pulses (“10 sequences of voltage pulses” is just a case in point; for it is
a truism that any physical component of a computing mechanism will
at some point reach its limits or endpoint and that infnitary functions
like identity have arguments too great to ever be physically process-
able). The physical circuit will reach its computational capacity at 10
sequences of voltage pulses.11 But P seems to be just as much an identity
computing mechanism as a quidentity computing mechanism, where
quidentity is defned:
f(x) = y {if x ≤ 10, y = x; y = 0 otherwise}
So, if P is a quidentity computing mechanism, it continues to compute

the quidentity function for sequences of voltage pulses greater than 10.
But if P is an identity computing mechanism, P is either miscomputing
or not computing at all after sequences greater than 10. The question is
whether all the relevant physical facts and properties of P alone sufce
to determine which antecedent is true of P.12 Moreover, there are further
possible functions (e.g., “schmidentity”, a function that yields the same
outputs as identity and quidentity for any arguments less or equal to 10,
but then diverges from both functions for any argument greater than 10
but less than or equal to 100, and then for any argument greater than 100
but less than or equal to 1,000, and so on exponentially) that P satisfes
at the same time and in the same respect but that are incompossible with
one another – quidentity is but one.13 Since this is generalizable to any
physical mechanism because of the physical features and limitations that
all such mechanisms are necessarily subject to, it follows that any physical
computing mechanism satisfes incompossible mathematical (and logical)
functions.
1.2. Some Necessary Features of Physical Mechanisms

The argument trades on several necessary features of physical objects.
For one, a physical mechanism (artifcial or organic) is only capable of
fnitely many steps and processes, so that both its actual operations and
its potential operations can also only ever be fnite in number. So, for
instance, any physical instantiation of Stabler’s identity circuit diagram
will only ever conduct fnitely many voltage pulses, and it’s only capable
of conducting fnitely many. Given the fnitude of both its actual and
potential processes, a physical mechanism can be made out to accord
with incompossible infnitary functions.14 Another necessity of nature is
that physical mechanisms decay, break down, malfunction, and cease to
exist. Breakdowns and malfunctions seem to be normative: they indicate
at least that the mechanism is failing to follow some norm or rule that it
is supposed to be following.15 In general, whether a physical mechanism
malfunctions depends on which function it is supposed to be realizing. So
whether the decay of P’s conductor after 10 sequences of voltage pulses
constitutes a malfunction or proper functioning will depend, in part, on
which mathematical function P is supposed to be computing. But if this
function is indeterminate, then it is likewise indeterminate whether P mal-
functions or functions properly. This point is important and we will come
back to it later.
1.3. Causal-Counterfactual Accounts of Formal Determinacy
There is a natural and intuitive response to this argument: “just look at
the intrinsic physical structure of P, determine its conditions of normal
operation and establish what P would’ve done had its conductor not
decayed.” The strategy is to allow the physical structure of the mechanism
to determine the counterfactuals that are true of it. Stabler (1987, 11)
proposes the following defnition of physical implementation of a math-
ematical function in order to block indeterminacy concerns:
1. there is an “interpretation” or “realization” function In which maps

a set of fnite sequences of physical “input” states of the system
onto the domain of f, and an interpretation function Out which
maps a set of fnite sequences of physical “output” states onto the
range of f, such that
2. physical laws guarantee that (in certain circumstances C, and if the
system satisfed conditions of normal operation N for long enough),
if the system went successively through the states of an input sequence
i, it would go successively through the states of the corresponding
output sequence f where Out(f) = f(In(i)).
To put it simply, the idea is that if P were to continue operating under

normal conditions (i.e., without decay, breakdown, or malfunction),
physical laws guarantee that it would compute the identity function by
causally securing the continuation of physical states and state transitions
that can be mapped to the range and domain of the identity function.
One major problem with Stabler’s account has already been noted by Jef
Buechner (2011): it is patently circular. “Normal conditions of operation”
is a normative notion that works as a placeholder for a defnite function.
Hence, Stabler’s response is a petitio: if P were to operate or continue
operating under normal conditions (read: if P were to continue comput-
ing the identity function without impediments), then it would (continue
to) compute the identity function. This is trivially true and circular.16 The
problem is that the criteria for what counts as normal and abnormal with
regards to P’s computations is dependent on the formal operation the
mechanism is supposed to be carrying out. But which formal operation
that is, is precisely what is in question! What counts as “conditions of
normal operation” for P will vary with the intended formal operation and
design (e.g., for a quidentity designer, P could very well be an adequate
and properly functioning quidentity device, while for an identity designer,
a malfunctioning identity device).
There is, nevertheless, something to Stabler’s response: it is true that if
we look at the physical structure and architecture of the system, it simply
isn’t the case that it is indeterminate what the mechanism would have
done had the conductor not decayed. On the basis of P’s physical structure
and environment, we can know that certain physically determinate pro-
cesses and facts would obtain if the conductor were not to decay. But
the question is whether the physical laws that determine and guarantee a
specifc set of physical processes and operations also sufce to determine
and guarantee a unique or defnite set of formal operations. And it seems
that by themselves they do not.17 Any set of physical facts that physical
laws guarantee cannot exclude incompossible pure functions from being
equally well realized. Even an actual infnity of physical cases would not
sufce, since no physical mechanism can do both what it does and what
it would have done instead (see Ross 1992, 141, n. 10). This is what KRA
purports to establish. If one modifes P so as to make it a digital device,
capable of processing digits (in particular, numerals), its physical structure
guarantees certain systematic causal patterns and manipulations of those
digits that, absent impeding conditions, will continue to be manifested in
the right set of circumstances for a fnite amount of time. But all this says
is that P is physically determinate in its actual and potential causal powers
(something KRA does not deny). That is to say, physical laws guarantee
ceteris paribus that such-and-such physical events will happen but it can-
not guarantee that such-and-such events will be of a defnite mathematical
or logical form, because such-and-such events though physically determi-
nate will not exclude incompossible formal operations. (This is, in part,
why the epistemological objection does not work [see §2.1].) Conditions
N of normal operation are not merely physical conditions of operation:
they are the physically relevant conditions for satisfying some assigned
mathematical form. From what physical laws guarantee, no unique and
determinate formal law can be determined. Hence the asymmetry between
the determinate physical structure of P and the indeterminate formal
structure of P – and of any physical object, for that matter.
It will not do to respond by saying that a physical object like P realizes
both functions: that at the same time and in the same respect, P is both an
identity computing device and a quidentity computing device. For, noth-
ing can realize incompossible functions at the same time and in the same
respect, on pain of contradiction. Such a conjunctivist response would
make it true of P that if P were not to malfunction, P would produce
an identity output and P would not produce an identity output (for it
would produce, instead, a quidentity output, namely 0 voltage pulses).
But it cannot be the case that if P were computing identities it would also
be doing the very same thing that would be yielding non-identities.18 A
disjunctivist response, according to which either P realizes identity or P
realizes quidentity (but not both), will also fail because none of the physi-
cal properties of P can provide truthmakers for any of the disjuncts.19
1.4. No Formal Operations in Matter

Further exploration of the implications of this argument for realist
accounts of physical computation is beyond the scope of this chapter.20
The purpose of the previous discussion is to highlight that formal inde-
terminacy aficts physical computing mechanisms qua physical, not qua
computational. As such, the indeterminacy argument can be run on physi-
cal objects of all kinds, whether natural or artifcial, whether paradig-
matically computational (e.g., digital computers, physically implemented
neural networks) or non-computational (e.g., rivers, metabolic processes,
tropical storms).21 In fact, it can be run on human beings qua physical
organisms (e.g., physical behavior, dispositions, genetics, neurons, etc.).
For example: take any physical feature or property of human beings (e.g.,
physical behavior or neurological processes) that purportedly realizes def-
nite mathematical-logical operations, then specify the physical processes
and mechanisms that instantiate or generate such determinate pure opera-
tions and specify the determinate pure operations that those processes
instantiate in virtue of their physical properties alone; then run a quus-like
argument over those processes and cash out on the indeterminacy: there
will always be some other mathematical-logical function incompossible
with the one assigned to the physical processes that these will equally well
realize at the same time and in the same respect. If this is the case, then
physical objects and mechanisms are not, on account of their physical
properties alone, the sorts of things that can do mathematics and logic.
This is because “doing mathematics” consists, partly, in grasping and
carrying out determinate formal operations.
Some authors (see, e.g., Churchland 1981; Dennett 1987, 1995) would
perhaps welcome this result, extending it to human beings (understood as
wholly physical objects) and arguing that there is no such thing as intrinsic
or underived mathematical and logical powers. But this reasoning does
not pass the test of intelligibility. We will come back to this objection later.
For now, simply take the prima facie obvious proposition that human
beings do mathematics and logic. So human beings are the sorts of beings
that can do mathematics and logic, that really have mathematical and
logical powers and therefore can carry out determinate formal opera-
tions. The question is, on account of what? If KRA is right, it cannot
be on account of their physical properties alone. Hence, human beings’
mathematical and logical powers are not entirely physical powers.
2. Objections and Replies

In what follows I address some of the most obvious objections. Since
some possible objections have already been answered elsewhere by other
authors, I will not consider them here.22
2.1. Underdetermination or Indeterminacy?

One predictable objection to KRA is that it fallaciously arrives at a meta-
physical conclusion from epistemological premises. What KRA demon-
strates, at most, is that we cannot know (i.e., it cannot be epistemically
determined)23 whether a physical mechanism realizes defnite formal
operations, not that it realizes no such operations. Hence, what KRA
shows, if anything, is the formal underdetermination (epistemic) of the
physical, not its formal indeterminacy (metaphysical). It is important to
stress that KRA seeks to establish a metaphysical and categorical con-
clusion from metaphysical and categorical premises, not from epistemo-
logical ones. So, what can justify a reinterpretation of the premises in
epistemological terms? It is possible that the objection hinges on an anti-
metaphysical view of the import of philosophical arguments. If so, this
cannot be resolved here. But it is also possible that the objection hinges
on a misunderstanding of the indeterminacy argument, given Ross’s own
mention of well-known underdetermination arguments (Quine 1969,
2013; Goodman 1983). At any rate, the objection reinterprets KRA in
the following manner:
(EA) It cannot be known whether the physical is formally determinate.

(EB) It can be known that mathematical and logical activity are
formally determinate.
(EC) Hence, mathematical and logical activity are not (entirely)
physical.
The argument has the following form: “It cannot be known that x is F.
It can be known that y is F. Therefore, x ≠ y”. To see what’s problematic
about this argument, says the objector, consider that this argument does
not rule out that x is in fact F. So, though it cannot be known, x could still
be F. Suppose x is F. Then both x and y are Fs and having F would hardly
ground an ontological diference between x and y. But the objector must
account for what, other than a metaphysical or categorical asymmetry,
can explain such modally strong epistemic asymmetry. Formal determina-
tion, as opposed to physical determination, is something that cannot be
known and attributed determinately to physical mechanisms. Even an
omniscient being would not be able to know.24 Yet it can be known and
attributed determinately to our mathematical-logical powers and activ-
ity. Explanatorily speaking, this leaves physicalism in principle unable to
provide any de re claims about the mathematical and logical features of
physical mechanisms and processes – human psychology and rationality
included if wholly physical.
Even if, for the sake of argument, we read KRA in epistemological terms
such that it can only be said to establish that the physical is necessarily
subject to formal underdetermination, this would still deal a devastating
blow to the explanatory pretensions of materialist accounts of intellectual
powers. For one, any theory that implies that mathematical-logical activ-
ity is epistemically inaccessible makes for a poor psychological theory of
our mathematical and logical powers. It makes knowledge and under-
standing of anything mathematical and logical impossible and leaves us
unable to explain well-established mathematical and logical practice.
Moreover, if true, it would be impossible for us to know whether the
physical is formally underdetermined. This is because in order to under-
stand and formulate intelligibly a formal underdetermination argument,
not only does one need to reason in logically determinate ways (what else
is an argument if not of a given logical form?), but one has to grasp and
think in terms of the formally determinate operations that the physical
cannot rule out. But if our very thinking were physical, our very reasoning
would be formally underdetermined. Thus, it would be underdetermined
whether we were arguing about formal underdetermination, and we could
not know this. So, the epistemological version is no win for physicalism.25
Consequently, even though KRA is not an epistemological argument,
we can use this objection to generate a compelling epistemological argu-
ment against materialism. Under the epistemological reading, it turns out
that materialism, if true, implies that it cannot be known whether def-
nite formal operations are ever realized, since the only admissible realizer
candidates are physical mechanisms and processes. But then, if math-
ematical and logical powers were purely material powers, it should not
be knowable whether they are formally determinate and determinable. So
mathematical-logical activity should be epistemically inaccessible, to the
efect that one could never be in a position know whether one or anyone
else can carry out a defnite formal operation or reason in any formally
determinate way. But such activity is not epistemically inaccessible (see
our following arguments). We know (or can know) that mathematical and
logical activity is formally determinate. We know that we can carry out
formally determinate operations. Hence, our mathematical-logical powers
are not (entirely) physical.26 Thus, the epistemological argument would
look more like the following:
(EA) If x is wholly physical, then it cannot be known whether x is

formally determinate.
(EB) It can be known whether x is formally determinate.
(EC) Hence, x is not wholly physical.
So, although KRA is not an epistemological argument, even under an

epistemological reading we would also get an interesting and strong
argument against materialism. Premise (EA´) would be derived from the
Kripkean argument that physical properties necessarily satisfy incompos-
sible formal operations and, on the epistemological reading, necessarily
block any epistemic determination as to which formal properties are true
of it. But no such thing is true about intellectual powers, which neither
satisfy incompossibles nor block epistemic determination. Hence, intellec-
tual powers have properties that are not physical. So, the epistemological
version still seems to imply an in principle diference between physical
powers and intellectual formal powers.
2.2. Sui Generis Physical Facts?
One could also try to object to KRA by saying that formally determinate
physical powers do exist but that they are sui generis physical powers,
that is, they belong to some sui generis physical level27 – of which, presum-
ably, we (can) have no knowledge. This is problematic in several ways.
We concede that mathematical-logical powers are sui generis, but it is not
clear what it would mean to still call them physical. The idea seems to be
that sui generis physical facts are not subject to indeterminacy concerns.
But the culprit of the indeterminacy is not being non-sui generis; it’s being
physical. So, what exactly is the content of the term “physical” in “sui
generis physical facts”? If these sui generis facts are still intelligibly said to
be physical, then the argument will apply. But if the discontinuity is such
that the sui generis physical facts have no ontological traits in common
with all the non-sui generis physical facts, then it seems that the subject
has been changed and the term “physical” is being employed in utterly
equivocal terms. In such a case, it is no longer clear that this response
saves physicalism, since it is no longer clear in what sense such sui generis
physical facts would still be physical.
2.3. Intrinsic vs. Extrinsic Formal Powers

Some authors would most likely reject the previous attempts to salvage
the formal determination of the physical and simply argue that there is no
such thing as a formally determinate physical process. So, they would be
more than happy to concede premise (B) of KRA. They would, however,
adamantly reject premise (A) if by it we mean a kind of human activity.
For, if humans are nothing but physical mechanisms, then its powers and
activities are entirely physical and hence are plagued by the same formal
indeterminacy that aficts any other physical thing. This objection simply
denies that mental states have any sort of “formality”, just as it denies
the have any sort of “intentionality” and “conceptual content”. Ross
discusses several problems with this objection (2008, 121–122, 1992,
145–148), as does Feser (2013, 17–18). I’d like to emphasize the point
that it is a wholly unintelligible objection.
The failure of this objection will illustrate why the Kripkean argu-
ment cannot be extended to our mathematical-logical powers, on pain of
absurdity and unintelligibility. Ross (1992, 145; cf. 2008, 121) put the
objection as follows:
We do not really add, either; we just simulate addition. Pure addition

is just as much an idealization as E = mc2. Of course, we can defne
such pure functions but cannot realize them; that is just a case of
the many functions we can defne which cannot be computed by any
fnite automation, or any other computer either. In a word, the fact
that there is no pure addition and no pure conjunction or modus
ponens is no odder than the fact that there are no perfect triangles.
The idea would be that, just as material powers and processes satisfy
incompossible formal operations, so do human powers and acts, and
therefore indeterminacy aficts all aspects of reality – human and nonhu-
man. So, we do not really think in formally determinate ways and our
reasoning is never really of any defnite logical or mathematical form.
Nothing intrinsic to our rational powers determines any defnite formal
operation. We, if anything, only simulate determinate mathematical and
logical activity.
The denialist response is defective in at least two ways. First, simulation
is a parasitic concept. It is a concept that requires, both in its understand-
ing and in its competent application to particular cases, prior grasp of
non-simulation or of the simulated thing. What this means is that in order
to understand what a simulation is, we must understand the distinction
between the simulation and the simulated (the real McCoy). If you cannot
tell the diference between a simulation and a non-simulation, then either
you don’t understand the concept of simulation or you do not understand
the application of those concepts to the particular case being considered.
To say we only simulate addition, identity, or modus ponens is to claim
understanding of what it is to simulate those pure functions and therefore
what it is to actually carry them out. Otherwise, one has no criteria for
stating that a given doing is a simulation, as opposed to the real thing.
The objection defeats itself. Second, as is the case with the epistemological
objection, if we cannot think intrinsically in any determinate formal way,
we would not be able to comprehend and follow formal indeterminacy
arguments. First, because we would not be able to think in terms of any
defnite argument, since our reasoning can never be of defnite logical
form. There’d be no matter of fact as whether we were reasoning validly
or invalidly, since there’d be no matter of fact as to the logical form of
our thinking. Second, because it requires that we do the very things it
denies we can do. It is, as Ross says, a performative contradiction. If you
deny that we ever add, because a putative case of addition is compatible
with quaddition, you are displaying a determinate grasp or understanding
of formal operations. The defnition of quidentity provides a rule for a
mathematical operation, but if I cannot think in terms of any such rule, I
cannot understand the rule nor what it means for something to satisfy (or
fail to satisfy) it. We would not even be able to defne it. In fact, if humans
do not ever reason in accord with determinate logical and mathematical
forms, there is no fact of the matter as to what the denial of formal deter-
minacy consists in. Since the very claim that we cannot think in formally
determinate ways requires that we (be the sort of beings that can) think
in formally determinate ways, the denial of the formal determination of
intellectual activity is incoherent, as it implies its own unintelligibility.28
Note that one of the implications of KRA is that the formal determina-
tion of physical mechanisms and processes is extrinsic to them. Noth-
ing intrinsic to them determines which formal features are true of them.
According to some perspectivalist accounts (see, e.g., Rescorla 2013,
2014, 2015), the formal determination of the physical is supplied by its
integration in our interpretative practices, which settle defnitively what
the mechanism is mathematically or logically doing. According to Kripke
(Buechner 2011; cf. 2018), the formal determination of a physical mecha-
nism is settled by the intentions of its designer which, in certain cases,
have become part of the shared background conditions against which
we interpret the operations of, say, calculators and such. If, however,
human beings are entirely physical mechanisms, then it seems that in
order for their mathematical and logical activity to be formally determi-
nate it would have to be extrinsically so and therefore dependent on some
other being that could interpret their operations in formally determinate
ways. Perhaps an intelligent design proponent would happily embrace this
conclusion. But this has unpalatable implications. First, on this view it
remains that we do not really do mathematics and logic because we do not
really have mathematical and logical powers: intrinsically speaking, you
never reason in any defnite mathematical or logical form. Our thinking
is only extrinsically and relationally mathematical and logical (dependent
on the interpretation assigned to our acts by some nonhuman external
agent not susceptible to the same considerations).29 Second, it is not clear
that we have determinate access to that interpretation, so that it becomes
a monumental epistemic hurdle to discover and to know what we are
actually doing when we are doing mathematics and logic. But again, this
is confused. We know that we can add, deduce in accord with modus
ponens, conjoin, divide, and so on, and we know this without needing to
consult an external agent’s interpretation of our acts. We understand the
argument from formal indeterminacy, whether we agree with it or not,
without the need to know some external’s agent stipulation. Plus, as has
already been shown, it is incoherent and unintelligible to sustain that our
formal powers are not intrinsically formally determinate.
It seems Dennett (1987, 287–321, 1995) would take the sort of deter-
minacy in display in formal thinking to be extrinsic to it, determined
by and derived from Mother Nature (genes, natural selection, etc.).
But, of course, if evolution and genetics are purely physical (which they
surely are), they are likewise devoid of any intrinsic and original (i.e.,
non-derived) formal determinacy.30 The buck cannot stop with Mother
Nature, since Mother Nature is simply a placeholder for purely physical
processes and mechanisms. So, Mother Nature cannot confer any formal
determination, extrinsic or otherwise, as it itself is in need of extrinsic
formal determination in order to instantiate formal features. But we have
already shown that this cannot be the case with our mathematical and
logical powers, which must be intrinsically formally determinate. Hence,
such powers are not entirely physical and their coming about cannot
be entirely accounted for by explanations that appeal merely to Mother
Nature (i.e., evolution).31
3. What About Conceptual Content?

Last but not least we come to the issue of semantics and conceptual con-
tent. Thus far we’ve been talking about formal operations without much
mention about meaning. This is intentional. We want to highlight formal,
as opposed to content, determination – interconnections notwithstanding.
The notion of formal determination is taken from Ross (1992, 2008), but
it needs explaining. Unfortunately, Ross’s own use of the term is confusing
at times and has led several authors (Dillard 2011; Feser 2013; Dillard
2013; Reppert 2012) to read into Ross’s text a long-standing tradition of
arguing for the immateriality of the human mind on the basis of concep-
tual content.32
3.1. Form and Content

In order to explain the notion of “formality” at play here, consider that
logical-mathematical operations have both an invariable component and
a variable component.33 The invariable component is that which remains
fxed through suitable transformations of the rule’s, function’s, or opera-
tion’s admissible variable components. It accounts for the normativity of
the transformations. It is what “encompasses” or logically contains all
cases whatsoever of the function. So, for example, when one applies a
given logical rule (e.g., modus ponens, disjunctive syllogism, NAND, etc.),
one reasons in a way that is at least determinate as to form,34 that is, as to
the structural features of the operation in question and the proper trans-
formation of the variable component: in an inferential rule like modus
ponens, the variable component will be whichever propositions one
applies the rule to; the invariable component will be the specifc type of
inferential pattern or structure of reasoning that the rule normatively sub-
mits propositions to. For example, the argument “If both conjuncts are
true, then the conjunction is true. Both conjuncts are true. Therefore, the
conjunction is true.” has as its variable components propositions about
conjunction. But the argument itself is not a conjunction: the argument is
about conjunction and can, in fact, be about anything else, but its inferen-
tial structure is what makes it an application of the logical rule of modus
ponens. In syllogistics, BARBARA is a type of argument that can take an
indefnite number of (the right kind of) diferent categorical propositions
(the admissible variable component), without change or transformation
in the structure of the argument (the invariable component). It is that
structure which grounds the normative properties of the argument (i.e.,
its validity).
Something analogous obtains in the case of arithmetical operations
like addition, or algorithms for computing the addition function: the
operation remains the same through suitable transformations of its
admissible variable component (i.e., pairs of numbers). The invariable
component, together with the rule’s determination of the admissible
variable component, is what allows for the identity or sameness of the
operation through substitutions of what is variable (e.g., “2 + 2”, “67
+ 58”, and “347 + 550” all have diferent variable components yet
the same operative structure).35 The addition function “+(x, y) = z” –
where “x” and “y” are natural numbers and “z” is their sum – does
not lose its form no matter which numbers we “plug” in for x and y:
variation in the arguments of the operation results not in a change in
what is being done but in a change in what it is being done to. It is
the “what is being done”, rather than the “what it is being done to”,
that we are after here. We call the “what is being done” (invariable
component) of the operation its form, the “what it is being done to”
(variable component) its content.36 And it is precisely logical and math-
ematical form that KRA, if sound, shows cannot be either identifed
with a material form or realized determinately by physical processes.37
Of course, the operation as a whole can be represented schematically,
in which case in order to know what operation “+(x, y) = z” represents
one must know what those symbols mean. But when one considers
the operation itself as a doing (i.e., as act or activity), not merely as
represented schematically and in terms of our knowledge of the mean-
ing of such representation or notation, but in terms of what it is to
apply the rule or what it is to carry out the operation (as well as what
it is for a language to be sensitive to and preserve the structure of the
operation), it becomes readily apparent that in doing such-and-such
logical or mathematical operation one is reasoning in some determi-
nately structured way.38 It is the structure of the thinking, considered
as invariable component, that we are arguing is immaterial. Content-
based arguments are not incompatible with this and may actually be
a necessary correlative to formal-structural arguments, but their focus
on the variable component as conceptualized makes them a diferent
kind of immateriality argument.
Two further points. First, the diferential characteristic, from the point
of view of ontology, of mathematical and logical form is that it is not
subject to nor constituted by physical laws.39 The structures of mathemat-
ics as a whole are not constituted by nor subject to any physical laws
and necessities, though their physical implementation or realization is
(e.g., the categorial diference between logical irreversibility and physi-
cal irreversibility). The whole project of explaining physical computa-
tion, for examples, depends on the presumed conceptual intelligibility of
physical processes and mechanisms determinately and objectively real-
izing or implementing such categorially diferent (i.e., abstract) structures
and operations. Similarly, physicalist or materialist accounts of human
mathematical and logical activity depend on the physical (e.g., our neu-
rophysiology or neurobiology) being able to generate powers that can
actualize determinately and intrinsically formal (i.e., mathematical and
logical) features – which are not constituted by physical laws, necessi-
ties, and conditions. Second, it should be noted that the form of such
operations can be the content of some other intellectual act, as when one
ponders about a given logical rule or reasons and proves a theorem about
a given arithmetical operation. Clearly, thinking about a given operation
is not the same as actually carrying out the operation. Why? In part,
because when one thinks about addition, addition is the content, not the
form, of the thinking; hence, to think about addition is not to add and
to add is not (the same as) to think about addition but to carry it out. In
sum: there is a diference between thinking about a rule and following or
applying the rule.
3.2. No Content Fallacy Threat

Once one realizes that the determinacy in question concerns primarily
the form of the understanding or the intellectual act, not the content,
one begins to see why objections like Pasnau’s “content fallacy” (Pasnau
1998), which arises for some content-based arguments, do not arise for
KRA. The content fallacy occurs when one infers from the fact that the
objects (content) of thought have such-and-such properties that there-
fore thought itself must have such-and-such properties. So, for example,
to infer from the fact that thoughts can be about objects that are imma-
terial that therefore thought itself must be immaterial is to commit the
content fallacy – as fallacious as inferring from the fact that thoughts
can be about fying sorcerers that therefore thoughts themselves must
be fying, and so on. But to think this objection can arise for KRA is
wholly misplaced. For, KRA is not about the content of thoughts but
about the forms of thought. We do not argue from formal operations
as the content of some thinking (e.g., thinking about addition) to the
immateriality of the thinking, but rather from the formal operation as
the form of the thinking (e.g., adding) to the immateriality of that kind
of thinking. Once one shows that such forms cannot be physically real-
ized in any intrinsic and determinate way but are realized intrinsically
and determinately in intellectual acts (e.g., calculations, argumentations,
etc.), one shows that such kind of intellectual acts cannot be (wholly)
physical. The accusation that KRA goes from the properties of the con-
tent of thought to the properties of thought itself only arises if one
construes KRA, primarily, as a content-based argument.40 Since KRA is
not, primarily, a content-based argument, the content fallacy does not
arise.41 Conceptual content matters, but we do not think it should be
allowed to constitute the basis of KRA.42
4. What Happened to Physical Computation?
Lastly, it is important to say something about physical computation, on
pain of unwarranted conclusions being drawn from KRA. KRA is com-
patible with perspectivalist accounts of physical computation (Schweizer
2016; Rescorla 2013, 2015). It only undermines realist (e.g., structuralist,
mechanistic, etc.) accounts of physical computation. A full treatment of
what it means to physically compute mathematical and logical functions,
in light of KRA, will have to wait for another occasion. But I shall stress
three diferent points. First, perspectival accounts of physical computa-
tions do not compromise the objectivity of those computations, only their
mind-independence. It is an objective fact, for example, whether a given
computational system is a classical digital computer or a neural network,
just as it is an objective fact whether a computing mechanism computes
an XOR gate. But the “objectivity” of this fact is constituted by (1) the
physical mechanism having the appropriate structural features, (2) there
being a well-defned formal structure or operation, and (3) the integration
of those physical structural features to an interpretative formal practice
for which such structural patterns can reliably count as instantiating that
formal operation, thus excluding by stipulation incompossible operations.
Physical computing mechanisms can achieve enough structural reliabil-
ity so as to extrinsically count as realizing a defnite formal operation.
But absent extrinsic determination, it will be intrinsically indeterminate
amongst incompossible mathematical and logical operations.43 Yet for
practical purposes, this does not matter: what’s important is that the
structural properties of a physical mechanism be such that it can be said
to compute efectively the desired formal operations, even if not exhaus-
tively. All KRA implies is that the physical computation of mathemati-
cal-logical functions is necessarily extrinsic, whereas ours is necessarily
intrinsic. So, materialism gets it backwards: the mathematical-logical
powers of physical computing mechanisms are explained in terms of the
mathematical-logical powers of human beings, not the other way around.
5. Conclusion
We have argued that KRA poses a forceful challenge to materialism. If
sound, it shows that mathematical and logical powers of human beings are
not entirely physical powers. In this respect, materialism about our formal
powers stands refuted. But more needs to be said. One of the most press-
ing questions is, how do we make sense of mathematical and logical activ-
ity as a human activity? Certainly, human beings qua biological organisms
are subject to all of the physical necessities and limitations that give rise
to formal indeterminacy. Given KRA, it would have to be on account of
the nonphysical aspects of their intellectual powers that human beings
can possess formal understanding and apply formal rules determinately.
Limitations on these powers would be, presumably, attributable to the
corporeality of human beings.44 But we need to account for how this is
supposed to work and how it is possible. Mathematical-logical activity
requires what Lonergan calls the image (Lonergan 1992): it needs the
appropriate signs or symbols, the appropriate representational medium,
for it to take place and to progress. Arabic numerals made possible dis-
coveries and progress that seem unimaginable with Roman numerals. If
our mathematical powers were purely immaterial, it is not clear why this
should be so. So we need to explain how physical necessities impose limi-
tations on our formal powers, which are purportedly immaterial, and why
it is that such immaterial powers cannot be as they are without appropri-
ate material media. It must also be explained how mathematical prac-
tice and knowledge are possible (a problem all theories of mathematical
knowledge, except fctionalism, must account for). Mathematical practice
has shared internal standards of excellence,45 epistemic patrimony (e.g.,
a body of mathematical knowledge accumulated though millennia and
transmitted through generations), competitions, prizes, medals, and so
forth. But if all our interactions and exchanges (whether in person, writ-
ten, or verbal) are mediated by something physical or material and the
material is formally indeterminate, how can any mathematical and logical
truths ever be communicated?
Given the prima facie deep interconnection between our mathematical-
logical powers and our material reality (i.e., corporeality), it seems that
strong forms of Cartesian dualism might be ill-suited to the explanatory
task. Aristotelianism may be in a better position to explain actual math-
ematical activity, integrating the non-physicality of intellectual powers to
a metaphysical framework that afrms corporeality as an essential fea-
ture of human beings. Yet this needs explaining. There are also questions
about whether Aristotelianism can accommodate KRA completely.46 Be
that as it may, we believe KRA to be a strong and formidable argument.
KRA does not concern itself with the question of how, in fact, humans
reason on average or whether classical logic actually captures the every-
day reasoning of human beings but rather with the fact that we can think
in formally determinate ways, and in certain technical cases, without
semantic determinacy: we can create, generate, and understand context-
free grammars, we can invent formal languages with determinate syntactic
and derivational rules but without any plausible semantic candidate, and
we can separate the form from the content and compute accordingly, and
so on. If KRA is correct, any being that can do that does it on account of
powers that cannot be entirely material.
Notes
1. For some contemporary examples of Aristotelian universality and content-
based arguments, see Feser (2013), Haldane (2013), and Klima (2009).
2. I take logical and mathematical powers to encompass (1) the natural capac-
ity of human beings for mathematics and logic, (2) the acquired ability for
mathematics and logic, and (3) the actual exercise of that ability: the activity
(i.e., doing mathematics and logic).
3. KRA is the argument that results from combining Kripke’s unpublished anti-
functionalist argument with Ross’s anti-materialist argument. My objective
here is, in the words of Kripke, “to present the argument as it struck me, as
it presented a problem for me, rather than to concentrate on the exegesis of
specifc passages”(Kripke 1982, viii). For a full explanation and defense of
KRA, see Ramos Díaz (2019).
4. I will, at times, simply write “mathematical-logical” for “mathematical and
logical”.
5. We will therefore refer to the argument underpinning this premise as “Krip-
kean”, in order to avoid attributing to Kripke an argument he did not explic-
itly countenance. Kripke’s lectures on computational functionalism remain
largely unpublished. Discussions can be found in Stabler (1987), Buechner
(2011), Burgess (2016), and Buechner (2018). Kevin Lande’s (Lande 2011)
unpublished argument against functionalism, strangely similar to Kripke’s,
gives a clear presentation of a Kripkean style argument. It should be noted
that Kripke does not use the term “formal determination”. This terminology
is taken entirely from Ross.
6. “A pure function is one satisfed by an infnite number of functors, and deter-
minate for each n-tuple, like ‘+’, and complete in each and every instance (that
is, it does not consist in a relationship among its cases, but explains them)”
(Ross 2008, 197, n. 9; emphasis added).
7. Quaddition is an operation that gives the same values as addition for any
operand less than or equal to 57; 5 for any operands greater than 57. For the
plus/quus argument, see Kripke (1982).
8. Triviality arguments against functionalism and computationalism purport to
show that the notion of “realization of a function” or the notion of “com-
putation” is too trivial to have any theoretical merit since, under appropriate
descriptions, anything can be said to realize a function or to compute. For
a classic treatment of triviality arguments, see Godfrey-Smith (2009). Label-
ing problems (see Buechner 2011, 352) concern the nonarbitrary labeling of
certain physical states and processes, to the exclusion of others, as the ones
that satisfy the abstract states of a given functional or computational diagram.
9. Interestingly, Stabler’s example can also be said to implement the addition
function, with “0” as a fxed operand: [0, n] – – – – – – – – – – – – [0, n]. Of
course, 0 plus any number equals the number itself. Now, if P is an addition
circuit, it is the sort of circuit that would generate the sum of the two numbers
if a real number other than 0 was admitted. But if P is an identity computing
circuit, then it is the sort of circuit that would never generate the sum of any
argument. These counterfactuals contradict each other. More on this later.
10. Some formal languages (e.g., frst-order logics) satisfy nonstandard models.
See Ramos Díaz (2019, 167–169) for a rebuttal of the claim that therefore our
formal understanding is indeterminate amongst nonstandard, incompossible
models. On a slightly diferent note: what counts as “standard” and what
counts as “nonstandard” is admittedly relative to mathematical practice. Sup-
pose there is a twin earth, like ours in every respect except that TE humans
do not have mathematics but rather quathematics (quarithmetic, qualgebra,
etc.). What counts as standard in their mathematical practice will be, natu-
rally, diferent (quaddition, quidentity, etc., are all standard functions in TE).
If a TE philosopher were to generate an argument like KRA, he could very
well formulate the argument with their nonstandard functions of addition
and identity.
11. For simplicity’s sake, we let the frst sequence of voltage pulses be of one volt-
age pulse, the second sequence of two, the third of three, and so on until 10,
such that the number of voltage pulses stand in a one-to-one correspondence
with the number of sequences of voltage pulses, starting from 1.
12. Notice that we could have introduced the example using the quidentity func-
tion frst, and then introduce the identity function as the “nonstandard” func-
tion that generates the indeterminacy problem.
13. Relative to schmidentity, P can be said to malfunction or miscompute after
10. So it seems to have equal claim in explaining the malfunction or miscom-
putation of P.
14. This too is true with respect to fnitary functions whose domain and range
include arguments too great to be physically processable and computable in
a fnite amount of time.
15. This is not entirely precise, but it will do for our purposes.
16. For a similar point made against non-Aristotelian subjunctive and counter-
factual accounts of functions, see Koons and Pruss (2017).
17. What determines that the physical causal patterns and processes that would
obtain if the conductor were not to decay after 10 sequences of voltage pulses
aren’t abnormal conditions of operation? They would be if the function to
be computed were “quidentity”. To a quidentity designer, the decay of the
conductor at that stage could be part of the normal conditions of operation.
To an identity designer, the decay clearly signals a malfunction or a limitation
(say, the architectural inadequacy) of the device. But the physical facts of the
mechanism cannot settle who is right since these are compatible with both
designs. The confusion arises from the fact that the expression “conditions
N of normal operation” disguises the fact that there are physical conditions
of normal operation and formal conditions of normal operation, the former
making reference to physical laws and necessities and the latter to formal
(mathematical or logical) laws and necessities. KRA does not deny the former
kind of conditions; it denies that they are sufcient for generating or instan-
tiating determinately the latter kind.
18. Again, the qualifer “at the same time and in the same respect” is essential
here, otherwise one might mistakenly take this argument to be about parallel
processing or computation.
19. For a fuller exposition and refutation of conjunctivist and disjunctivist
responses, see Lande (2011, 5–7).
20. Realist accounts of physical computation hold that mathematical functions
are computed in a mind-independent way by physical computing mechanisms.
There are extreme forms of realism like ontic pancomputationalism (Piccinini
and Andersen 2018; Piccinini 2015, 51–73), according do which everything
physical computes. For non-realist accounts compatible with KRA, see for
example perspectivalist accounts (Rescorla 2013; cf. Coelho Mollo 2020).
21. Assuming that pancomputationalism is false. KRA does not imply that there
is no diference between physical computing mechanisms and physical non-
computing mechanisms, as if nothing makes the former a more appropriate
candidate for computations.
22. For a refutation of equivalence classes and causal-counterfactual responses,
see Buechner (2011, 355–356, 357–358).
23. The epistemological conclusion has to be this modally strong because the
upshot of the reasoning underpinning KRA is that any physical mechanism
whatsoever will necessarily satisfy incompossible formal operations.
24. “[The sceptic] claims that an omniscient being, with access to all available
facts, still would not fnd any fact that diferentiates between the plus and the
quus hypotheses. Such an omniscient being would have neither need nor use
for simplicity considerations” (Kripke 1982, 39).
25. Add to the fact that it makes identity theses and supervenience accounts
epistemically bankrupt, as neither could secure any de re statements in their
explanation of psychological phenomena with formal features (e.g., mental
calculations, valid reasoning).
26. This is not an argument from introspection, but an argument from operations
or acts. You can refute someone who insists that it cannot be known whether
humans can walk by simply walking. In this sense, you know you can add by
adding, similar to the way you knowing you can walk by walking.
27. This is, in fact, Dillard’s view (2011, 2013). Due to space limitations, we can-
not address Dillard’s objections. For a discussion and refutation of Dillard’s
epistemological objections, particularly his “mitosis/schmitosis” counterex-
ample, see Ramos Díaz (2019, 156–160, 182–189).
28. This, incidentally, is the very same reason that the skeptical paradox becomes
incoherent when extended to human understanding. It remains coherent
when applied to physical and phenomenal properties (e.g., behavior disposi-
tions, qualia, past event, sensations, etc.) but becomes absurd when extended
to human thought.
29. Similar to a point made in Koons and Pruss (2017, 199), an “extrinsic deter-
mination only” account leads to an infnite regress that would make any
extrinsic determination impossible, since there would be no source of original
or intrinsic determination that could create the conditions for the possibility
of extrinsic determination. Extrinsic determination is parasitic or consequent
on intrinsic determination.
30. For a classic treatment of diferences between original, derived, and “as-if”
intentionality, see Searle (1992, 78f).
31. I cannot here develop this argument with respect to causal theories (Adams
and Aizawa 2021) and teleological functionalist accounts (Neander and
Schulte 2021), but it should be clear that neither of them will succeed in
solving the genetic problem (the problem of explaining the coming about or
emergence of our formal powers). I will address this in a subsequent paper.
32. To be clear, Ross’s own description of the indeterminacy argument as one of
the “jewels of analytic philosophy” appeals to arguments by Quine, Good-
man, and Kripke, who were generally concerned with meaning. His formula-
tion of his own indeterminacy argument relies at times on seemingly semantic
language (e.g., “formal content”). So, it is not unreasonable to interpret Ross’s
argument as content-based. But I think it is mistaken. This is analogous to
reading Ross’s argument in epistemological terms, which is reasonable given
Ross’s own seemingly epistemological language at times and clear appeals
to the underdetermination arguments of Quine, Goodman, and Duhem – all
of whom frame their arguments to varying degrees in epistemological terms.
Nevertheless, the epistemological interpretation is mistaken, as Feser (2013,
22–23, n. 65) has pointed out.
33. This terminology – “invariable and variable components”, as well as its
explication, is taken from (Koslicki 2008). Koslicki uses it as a general meta-
physical characterization of the structure of objects; I’ll use it only as a char-
acterization of pure formal structures.
34. I say “at least” because in non-ordinary technical circumstances (e.g., Aris-
totelian or Boolean syllogistics, propositional and predicate logic, abstract
algebra) one’s reasoning and operations will be determinate as to form but not
necessarily as to content, since one abstracts away from it in such specialized
settings.
35. In what follows, a rule is a norm or standard for a certain action that deter-
mines what it is to act in accord with it and what it is to not. A concrete
operation is a particular application of a rule.
36. Though we identify the invariable component (structure) of the operation
with its form, with that which makes it be the kind of act or operation it is
and susceptible to mathematical or logical taxonomy, it is not necessary to
cash out the distinction between form and content in terms of “invariable vs.
variable components”, though it is extremely useful and illuminating to do
so. In the case of logical operations, conceptual content (e.g., propositions) is
more readily seen to be the variable component, whereas in a mathematical
function like “+(x, y) = z”, whose possible variable components are pairs of
natural numbers and not propositions, the notion of conceptual content may
seem inadequate. All I can add is that, whatever else might be said, when
one considers the function as an intellectual doing it becomes clear that the
variable component (i.e., specifc numbers) is the content of the reasoning
inasmuch as when one adds natural numbers the content of the thinking will
be whatever natural numbers one is adding.
37. The concepts of “being truth-preserving in virtue of form” or “validity-pre-
serving operations” pick out features not had by physical processes, since they
cannot be of any defnite logical-mathematical form.
38. If one is engaged in mental arithmetic, or mental proofs of frst-order logic,
one is not meaning anything; but one is doing much. This requires structur-
ing one’s meaningful thoughts and reasoning in determinate ways that are
irreducible to the content.
39. This is not the same as saying that it is abstract (in platonic terms), though it
is not incompatible with it either. Aristotelians and Wittgensteinians would
agree that mathematical and logical forms are not subject to physical laws
and necessities and yet resist the move to claim that such forms are mind-
independent substances existing in some third realm.
40. An example of a content-based version of KRA is Feser’s (2013). There
are important diferences between Feser’s argument and mine, but I cannot
address them here. I will do so in a separate paper.
41. We believe this to be hermeneutically correct as an interpretation of Ross’s
own argument. If one interprets Ross’s indeterminacy argument as primarily
a content-based argument, one cannot explain certain key parts of his texts.
To mention one: after having laid out the principal argument, Ross goes on
to propose a “parallel reasoning as to content”, asking, “Can the reasoning
above be applied to content as well as to form’?” (Ross 2008, 123). If the
entire argument had been about content, this would make no sense: there
would be no parallel argument as to content. Likewise, “asserting, in any one
of its senses, cannot be ‘halfway’ between opposed forms; it would not be
asserting then. And so on, for every form of thinking as process” (Ross 2008,
123; emphasis added). In our formulation of the argument, “operation” and
“activity” have taken the place of “process”, a term we’ve only used in refer-
ence to the physical.
42. This is, in part, why the indeterminacy problems posed by KRA are diferent
from Quine’s “gavagai” (Quine 1969, 2013) and Goodman’s “blue/grue”
examples (Goodman 1983). KRA concerns operations and acts such as fol-
lowing and applying, with understanding, defnite mathematical and logical
rules. “Grue” and “gavagai” signal concepts (literally, meanings) not acts
or operations of any sort: there is no such thing as gavagaing and grueing,
except equivocally. KRA is concerned primarily with operations and acts,
as opposed to concepts (although both are interconnected, as has already
been explained). It deals with things we can do, not merely think about, with
understanding.
43. Interestingly, Dewhurst (2020; cf. 2018, 2014) believes that there is no such
thing as miscomputation in physical computing mechanisms. In simple terms,
the idea is that the physical mechanisms that are said to miscompute are
really following the rule that is encapsulated in their physical structure, so
they are not really miscomputing. Physical mechanisms are acting in accord
with their own internal physical structure, and thus intrinsically speaking do
not miscompute. A computation only counts as deviant or as a miscomputa-
tion relative to an intended design, not relative to the mechanism’s physical
structure. Now, recall that, as KRA shows, the physical structure can be made
out to accord with incompossible formal rules even if the physical mechanism
does not last enough to manifest deviant behavior relative to the rules (e.g.,
let P reach its lifetime span at nine sequences of voltage pulses; all physical
facts of P are compatible with identity and quidentity). So whatever rules
are encapsulated in the physical structure of a computing mechanism, these
will not be mathematical or logical rules. At any rate, Dewhurst has given an
excellent argument for why physical computation is not an adequate notion
for explaining human computational powers and for why human beings are
not physical computing mechanisms, because human beings do miscompute.
44. This is, precisely, Ross’s view (see, e.g., Ross 1992, 143–144, n. 13). One
would likewise need an account of the physiological and psychological media
of such immaterial powers if one hopes to give an integrated and holistic
explanation of human formal activity, which is embodied.
45. For the notion of normative practices with internal standards of excellence
in the context of musical practice, see for example Ross (1993). The analysis
is surprisingly applicable, mutatis mutandis, to mathematical practice.
46. For example, Aristotelian approaches to mathematics tend to explain the
origin of mathematics in terms of the interaction between our capacity for
abstraction and the real magnitudes of material things. So, a mathematical
object emerges when we abstract away from all features of material things
except their quantitative dimensions, now universalized and dematerialized.
But if material things are formally indeterminate, it is not clear how one can
abstract from a material thing a defnite abstract object, rule, or operation. In
any case, I do not think mathematical operations are in any way abstracted
from matter, even if mathematical objects (e.g., natural numbers, geometrical
fgures) are.
Bibliography
Adams, Fred, and Ken Aizawa. 2021. “Causal Theories of Mental Content.” In
The Stanford Encyclopedia of Philosophy, edited by Edward N. Zalta, Spring
2021. Metaphysics Research Lab. Stanford University. https://plato.stanford.
edu/archives/spr2021/entries/content-causal/.
Buechner, Jef. 2011. “Not Even Computing Machines Can Follow Rules.” In
Saul Kripke, edited by Alan Berger, 343–368. Cambridge: Cambridge Univer-
sity Press. https://doi.org/10.1017/CBO9780511780622.015.
———. 2018. “Does Kripke’s Argument Against Functionalism Undermine the
Standard View of What Computers Are?” Minds and Machines 28 (3): 491–
513. https://doi.org/10.1007/s11023-018-9466-5.
Burgess, John P. 2016. “Kripke on Functionalism.” Crítica: Revista Hispano-
americana de Filosofía 48 (144): 3–18.
Churchland, Paul M. 1981. “Eliminative Materialism and the Propositional Atti-
tudes.” The Journal of Philosophy 78 (2): 67. https://doi.org/10.2307/2025900.
Coelho Mollo, Dimitri. 2020. “Against Computational Perspectivalism.” The
British Journal for the Philosophy of Science, December. https://doi.org/10.1093/
bjps/axz036.
Dennett, D. C. 1987. The Intentional Stance. Cambridge, MA: MIT Press.
———. 1995. Darwin’s Dangerous Idea: Evolution and the Meanings of Life.
New York: Simon & Schuster.
Dewhurst, Joe. 2014. “Mechanistic Miscomputation: A Reply to Fresco and
Primiero.” Philosophy & Technology 27 (3): 495–498. https://doi.org/10.1007/
s13347-013-0141-8.
———. 2018. “Computing Mechanisms Without Proper Functions.” Minds and
Machines 28 (3): 569–588. https://doi.org/10.1007/s11023-018-9474-5.
———. 2020. “There Is No Such Thing as Miscomputation.” http://philsci-
archive.pitt.edu/17000/.
Dillard, Peter. 2011. “Two Unsuccessful Arguments for Immaterialism.” Ameri-
can Catholic Philosophical Quarterly 85 (2): 269–286. https://doi.org/10.5840/
acpq201185215.
———. 2013. “Ross Revisited: Reply to Feser.” Edited by Robert E. Wood.
American Catholic Philosophical Quarterly: 139–147. https://doi.org/10.5840/
acpq201311216.
Feser, Edward. 2013. “Kripke, Ross, and the Immaterial Aspects of Thought.”
American Catholic Philosophical Quarterly 87 (1): 1–32. https://doi.
org/10.5840/acpq20138711.
Godfrey-Smith, Peter. 2009. “Triviality Arguments against Functionalism.” Phil-
osophical Studies 145 (2): 273–295. https://doi.org/10.1007/
s11098-008-9231-3.
Goodman, Nelson. 1983. Fact, Fiction, and Forecast. 4th ed. Cambridge, MA:
Harvard University Press.
Haldane, John. 2013. “Is the Soul the Form of the Body?” American Catholic
Philosophical Quarterly 87 (3): 481–493. https://doi.org/10.5840/
acpq201387334.
Klima, Gyula. 2009. “Aquinas on the Materiality of the Human Soul and the
Immateriality of the Human Intellect.” Philosophical Investigations 32 (2):
163–182. https://doi.org/10.1111/j.1467-9205.2008.01368.x.
Koons, Robert C., and Alexander Pruss. 2017. “Must Functionalists Be Aristo-
telians?” In Causal Powers, edited by Jonathan D. Jacobs. Vol. 1.
O x f o r d : Oxford University Press. https://doi.org/10.1093/oso/
9780198796572.003.0013.
Koslicki, Kathrin. 2008. The Structure of Objects. Oxford; New York: Oxford
University Press.
Kripke, Saul A. 1982. Wittgenstein on Rules and Private Language: An Elemen-
tary Exposition. Repr. Cambridge, MA: Harvard University Press.
Lande, Kevin. 2011. “Realizing Functions? An Argument Against Functional-
ism.” Unpublished, 1–8.
Lonergan, Bernard J. F. 1992. Insight: A Study of Human Understanding. Edited
by Frederick E. Crowe and Robert M. Doran. Vol. 3. Collected Works of
Bernard Lonergan. Toronto: Published by University of Toronto Press for
Lonergan Research Institute of Regis College.
Nagel, Thomas. 2012. Mind and Cosmos: Why the Materialist Neo-Darwinian
Conception of Nature Is Almost Certainly False. New York: Oxford University
Press.
Neander, Karen, and Peter Schulte. 2021. “Teleological Theories of Mental Con-
tent.” In The Stanford Encyclopedia of Philosophy, edited by Edward N. Zalta,
Spring 2021. Metaphysics Research Lab. Stanford University. https://plato.
stanford.edu/archives/spr2021/entries/content-teleological/.
Pasnau, Robert. 1998. “Aquinas and the Content Fallacy.” The Modern School-
man 75 (4): 293–314. https://doi.org/10.5840/schoolman199875431.
Piccinini, Gualtiero. 2015. Physical Computation: A Mechanistic Account.
Piccinini, Gualtiero, and Neal G. Andersen. 2018. “Ontic Pancomputational-
ism.” In Physical Perspectives on Computation, Computational Perspectives on
Physics, edited by Michael E. Cufaro and Samuel C. Fletcher. 1st ed., 23–38.
Cambridge: Cambridge University Press. https://doi.org/10.1017/
9781316759745.
Quine, W. V. 1969. Ontological Relativity and Other Essays. 10th–11th print ed.
New York: Columbia University Press.
———. 2013. Word and Object. New ed. Cambridge, MA: MIT Press.
Ramos Díaz, Antonio. 2019. “Mind and Formal Structures: On the Kripke-Ross
Argument Against Naturalizing Formal Understanding.” PhD Thesis. KU
Leuven.
Reppert, Victor. 2012. “The Argument from Reason.” In The Blackwell Compan-
ion to Natural Theology, edited by William Lane Craig and J. P. Moreland,
344–390. Oxford: Wiley-Blackwell.
Rescorla, Michael. 2013. “Against Structuralist Theories of Computational
Implementation.” The British Journal for the Philosophy of Science 64 (4):
681–707. https://doi.org/10.1093/bjps/axs017.
———. 2014. “A Theory of Computational Implementation.” Synthese 191 (6):
1277–1307. https://doi.org/10.1007/s11229-013-0324-y.
———. 2015. “The Representational Foundations of Computation.” Philosophia
Mathematica 23 (3): 338–366. https://doi.org/10.1093/philmat/nkv009.
Ross, James. 1992. “Immaterial Aspects of Thought.” The Journal of Philosophy
89 (3): 136–150. https://doi.org/10.2307/2026790.
———. 1993. “Musical Standards as a Function of Musical Accomplishment.”
In The Interpretation of Music: Philosophical Essays, edited by Michael Krausz,
89–102. Oxford; New York: Clarendon Press; Oxford University Press.
———. 2008. Thought and World: The Hidden Necessities. Notre Dame, IN:
University of Notre Dame Press.
Schweizer, Paul. 2016. “In What Sense Does the Brain Compute?” In Computing
and Philosophy, edited by Vincent C. Müller, 63–79. Cham: Springer Interna-
tional Publishing. https://doi.org/10.1007/978-3-319-23291-1_5.
Searle, John R. 1992. The Rediscovery of the Mind. Representation and Mind.
Cambridge, MA: MIT Press.
Stabler, Edward P. 1987. “Kripke on Functionalism and Automata.” Synthese 70
(1): 1–22.
9 Persons, Souls, and Life
After Death
Christopher Hauser
1. Two Rival Versions of Thomistic Hylomorphism:

Corruptionism and Survivalism
It is evident to each of us that he or she engages in a variety of mental
activities. It is also evident to each of us that he or she presently has a
body. But what is this entity that thinks, senses, remembers, and so on and
presently has a body? Put more simply, what are we? What are human
persons? Philosophers both past and present have proposed a variety of
answers to this question. Of the theories defended in our contemporary
context, several, including Thomistic Hylomorphism, Animalism, Con-
stitutionalism, and Emergent Individualism, can claim to be “neo-Aristo-
telian” in one respect or another.1 This chapter will focus on just one of
these contemporary neo-Aristotelian theories: Thomistic Hylomorphism,
a theory inspired by Thomas Aquinas’s hylomorphic account of what we
are, which in turn was inspired by Aristotle’s hylomorphic account of
what we are.
Like other hylomorphists, Thomistic Hylomorphists maintain that
human persons are enduring individuals “composed of” or “constituted
from” (in a to-be-specifed sense of the term) matter (hylē) and a cer-
tain kind of substantial form (morphē). Thomistic Hylomorphists add
that human persons difer from other material substances in that human
persons have substantial forms that can exist without informing a body.
Adopting the Aristotelian use of the term “soul,” Thomistic Hylomor-
phists call the substantial forms of living things “souls” and the substan-
tial forms of human persons “intellective (or rational or human) souls.”
For the purposes of this chapter, we can defne Thomistic Hylomorphism
as the conjunction of the following four theses:
A. Human persons are enduring individuals “composed of” or “con-

stituted from” (in a broad, to-be-specifed sense of the term) matter
and a certain kind of substantial form (viz., an intellective soul).
B. Human persons have certain powers (including all of their mental
powers, e.g., their powers to think, to sense, to imagine, etc.) at
DOI: 10.4324/9781003125860-12
246 Christopher Hauser
least partially in virtue of having the kind of souls that they do. (In
the traditional Aristotelian jargon, a person’s soul is a “principle”
[archē, principium] of his or her mental powers).
C. Human persons are not identical to their souls.
D. The souls of human persons can continue to exist separately, that
is, without informing a body, after the persons to which these souls
belong die.
These four claims leave unanswered many further questions, such as the
following: just what kind of entity is a substantial form?; how exactly are we
to characterize the body (or underlying matter) informed by such a form?;
what exactly is the “informing” relationship holding between a human per-
son’s form (or soul) and matter (or body)?; and in what sense exactly is a
human person “composed of” or “constituted from” this form (or soul) and
the matter informed by it? Though questions such as these must be answered
to arrive at a fully feshed-out version of Thomistic Hylomorphism, the
points made in this chapter do not require any particular answers to these
questions. For that reason, I make no attempt to answer them here.2
The focus of this chapter is the fourth of the aforementioned four theses,
viz., the Thomistic Hylomorphist’s claim that the souls of human persons
can continue to exist separately, that is, without informing a body, after
the persons to which these souls belong die. Much of the recent scholarly
discussion of Thomistic Hylomorphism has centered on this thesis and,
more specifcally, the question of whether human persons can continue
to exist along with their souls after they die or whether only their souls
can continue to exist in this separated, disembodied, post-mortem state.
Two rival groups of Thomistic Hylomorphists have formed, Survivalists
and Corruptionists, with the following opposed views:
Corruptionist Thomistic Hylomorphism: The souls of human persons

can continue to exist separately, that is, without informing a body,
after these persons die, but these persons cannot exist without their
souls informing a body and hence cannot continue to exist after their
souls separate from their bodies at their deaths.
Survivalist Thomistic Hylomorphism: Not only the souls of human

persons but also human persons themselves can continue to exist
after their souls separate from (i.e., cease to inform) their bodies at
their deaths. While in this separated state, these persons have souls
but no bodies.
In some cases, discussion of these two rival views has centered on the
correct interpretation of Thomas Aquinas, the fgure whose ideas are
the inspiration for all contemporary Thomistic Hylomorphists.3 In other
cases, however, the discussion has focused not on this historical and exe-
getical question but rather on the grounds that contemporary Thomistic
Persons, Souls, and Life After Death 247
Hylomorphists have to favor one position or the other, regardless of which
position Aquinas himself might have held. It is to this latter discussion and
not the former discussion that this chapter aims to contribute.
On the one side, contemporary Corruptionists have ofered several
philosophical objections to Survivalism. For example, some Corruption-
ists have argued that (a) Survivalism requires that a human person could
be composed of just one proper part, viz., a soul, and that (b) this is incon-
sistent with the allegedly true mereological principle that nothing can have
just one proper part.4 Other Corruptionists have claimed that Survivalism
requires an implausible and decidedly un-Aristotelian rejection of the idea
that having a body (or matter) is part of the essence of a human being.5
On the opposite side, Survivalists have argued that Corruptionism is
inconsistent with the religious beliefs often held by Thomistic Hylomor-
phists. One such objection claims that (a) Corruptionism is consistent
with the traditional Judeo-Christian belief in a non-immediate bodily
resurrection only if human persons can cease to exist for a time before
beginning to exist again at a later time and that (b) this is problematic
because having such “gappy existence” is impossible.6 Another religious
objection charges that Corruptionism is inconsistent with the traditional
Christian belief that human persons (and not just their souls) will experi-
ence, between their deaths and bodily resurrection, a disembodied afterlife
or interim state in which such persons, among other things, undergo pur-
gation in Purgatory, ofer intercessory prayers for the living, and, in the
case of the saints, enjoy a beatifc communion with God.7
This chapter makes a new contribution to this debate by developing a
heretofore undiscussed and, in my view, decisive objection against Cor-
ruptionist Thomistic Hylomorphism. In particular, I argue that if the cen-
tral claim of Corruptionism were true, that is, if it were true that human
persons cannot exist without their souls informing a body, this would
undermine the grounds on which all Thomistic Hylomorphists, including
Corruptionists, rely to justify their claim that our souls can continue to
exist after our deaths. If I am right, then Thomistic Hylomorphists have
grounds for thinking that our souls can continue to exist after our deaths
only if they allow, as Survivalists do and Corruptionists do not, that we
human persons can continue to exist after our deaths, with souls but no
bodies. It follows that Corruptionists face a choice: give up a central thesis
of Thomistic Hylomorphism, viz., the thesis that our souls can continue
to exist separately from our bodies after our deaths, or give up their Cor-
ruptionism and become Survivalists.
2. The Overall Argument

Corruptionists maintain that human persons cannot exist disembodied,
that is, without their souls informing a body. But I claim that if this were
true it would undermine the grounds that all Thomistic Hylomorphists,
including Corruptionists, have for believing that our souls can continue
to exist after our deaths. In this section, I present the overall argument for
this thesis. In §§3–4, I defend the argument’s key premise in further detail.
The argument has two parts. The frst part begins with an uncontro-
versial observation:
1. The Thomistic Hylomorphist’s claim that a human person’s soul

after the person’s death depends for its justifcation on the claim
that a human person’s soul is the principle of a power, viz., the
power to think, which can be exercised without the soul’s informing
a body.
The “power to think” in question is the power to formulate concepts of

universals or kinds (e.g., the concept of a dog, the concept of color, the
concept of a circle, etc.); to entertain, assent to the truth of, or deny the
truth of propositions employing these concepts (propositions like all dogs
are mammals, Fido is a dog, shapes and colors are properties, etc.); and
to reason using these concepts (e.g., Fido is mammal since Fido is a dog
and all dogs are mammals). For simplicity, I simply refer to this power
as the “power to think.” Thomistic Hylomorphists, following Aquinas,
hold that, unlike all of our sensory powers (e.g., our powers to see, hear,
imagine, etc.), whose exercise necessarily depends on one or more parts of
the body (e.g., our eyes, ears, etc.) undergoing certain changes, our power
to think can be exercised even in the absence of any such body-involving
changes. From this, Thomistic Hylomorphists infer that this power to
think is one which can be exercised without one’s soul informing a body
at all, and, from this in turn, Thomistic Hylomorphists infer that the soul,
as the principle or ground of such a power, likewise can exist without
informing a body. (To be clear, I’m not claiming here that this is a sound
argument but only that this accurately refects the Thomistic Hylomor-
phist’s grounds for claiming that human souls can exist separately, i.e.,
without informing a body).8
Next, consider that
2. Thinking can occur only if something can do the thinking in ques-

tion; thinking requires a thinker.
Like any other mental activity, the occurrence of thinking requires that
something performs the activity, that is, the thinking, in question. This is
self-evident and needs no further defense.
From 1 and 2, it straightforwardly follows that
3. The Thomistic Hylomorphist’s claim that a human person’s soul

after that person’s death depends for its justifcation on the claim
that there is something which can exercise the relevant power to
think while the soul is not informing a body. (From 1 and 2)
This is the conclusion of the frst part of my argument.

The second part of the argument consists in showing that, if Corrup-
tionism were true, Thomistic Hylomorphists would have no basis for
claiming that there is something that can exercise the relevant power to
think while the soul is not informing a body. First, consider that
4. Given Corruptionism, human persons cannot exercise the relevant

power to think while their souls are not informing a body.
The central claim of Corruptionism is that human persons cannot exist

while their souls are not informing a body. Since one cannot think while
one does not exist, it follows straightforwardly from Corruptionism that
human persons cannot exercise the relevant power to think while their
souls are not informing a body.
But if human persons cannot exercise the relevant power to think after
their deaths, what else can? The only other candidate for this post-mortem
thinker is the person’s separated soul:
5. If there is something which can exercise the relevant power to think

while the soul is not informing a body and this something is not
the human person to which the soul previously belonged, it is the
soul itself.
However, I claim that
6. There is no reason to believe that the souls of human persons can

think (i.e., exercise the relevant power for thinking), either while
informing a body or while not informing a body.
This is likely to be the most controversial premise of the argument. I

defend it in §§3–4 below. For now, I observe that, if true, 4, 5, and 6
imply that
7. Given Corruptionism, Thomistic Hylomorphists have no reason

to believe that there is something that can exercise the relevant
power to think while the soul is not informing a body. (From 4,
5, and 6)
This concludes the second part of the argument.

Combining the conclusions of the two parts of the argument (viz., 3
and 7), we reach the overall conclusion:
8. Given Corruptionism, Thomistic Hylomorphists have no reason to
believe that a human person’s soul can continue to exist separately,
that is, without informing a body, after that person’s death. (From
3 and 7)
In other words, if it were true, as Corruptionists claim, that human per-

sons cannot survive the separation of their souls from their bodies, this
would undermine the grounds that all Thomistic Hylomorphists, includ-
ing Corruptionists, invoke to justify their belief that our souls can exist
separately from our bodies after our deaths.
As I indicated earlier, the step in this argument most likely to be
challenged by Corruptionists is 6. All Thomistic Hylomorphists main-
tain that there is something that can think while the soul is separated.
Indeed, the frst part of the argument (1–3) makes clear that Thomistic
Hylomorphists must maintain this, for it is implied by the grounds they
ofer for believing that the souls of human persons can exist without
informing bodies. Survivalists insist that this disembodied thinking
thing is the human person: while the soul is informing a body, the thing
which can think is the human person; likewise, Survivalists claim, the
thing which can think while the soul is not informing a body is the
same human person.9 Corruptionists, however, cannot accept this: as
4 makes clear, the Corruptionist thesis that human persons cannot
exist without their souls informing a body entails that human persons
cannot think without their souls informing a body (since nothing can
think while it does not exist). So, as 5 indicates, if Corruptionists are
to sustain their claim that something can exercise the relevant power
to think after our deaths, they must claim that this something is the
separated soul itself.
As a matter of fact, if one reviews the literature, one will see that all
Corruptionists say exactly this: separated human souls can think, that
is, exercise the relevant power to think. But Corruptionists are divided
about how exactly to explain this claim. Some Corruptionists maintain
that human souls can think both after our deaths, once separated and no
longer informing a body, and before our deaths, while still informing a
body.10 Other Corruptionists, however, maintain that human souls can
think only after our deaths, once separated and no longer informing a
body.11
In what follows, I discuss each of these positions in turn. In §3, against
the frst Corruptionist position, I argue not only that there is no reason to
believe that we have souls that can think prior to our deaths but also that
there is good reason not to believe this, given its deeply counterintuitive
implications. In §4, I build on this by arguing, against the second Cor-
ruptionist position, that if there is no reason to believe that our souls can
think prior to our deaths, then there is also no reason to believe that they
can think after our deaths.
3. Both My Soul and I Think: A Baseless and
Counterintuitive Claim
It is evident that we human persons can engage in a variety of mental
activities. Each of us is aware that he himself thinks, senses, imagines,
remembers, etc. Thomistic Hylomorphists accept this datum of experience
and add that we have such powers, for example to think, sense, imagine,
remember, and so on, at least partly in virtue of having the souls that
we do (see claim (B) in §1). In the case of our “sensory” powers (i.e.,
our powers to see, hear, imagine, etc.), Thomistic Hylomorphists see no
reason to claim that our souls themselves have the powers in question.
In other words, our having the souls (or substantial forms) that we do at
least partially grounds our having the relevant powers in a way similar
to the way in which a material object’s having mass grounds its power
to exert an attractive force on other material objects. Just as a material
object’s mass doesn’t itself exert the attractive force but rather the mate-
rial object does, by virtue of having mass, likewise our having the souls
that we do at least partially grounds our having certain powers (to see,
hear, imagine, and so forth) without our souls themselves having these
powers. We can see, hear, imagine, and so on because we have the souls
that we do, but not because our souls are themselves things that can see,
hear, imagine, and so on.
But what of our intellects? That is, what of our “power to think” as
I have been using the term? Survivalists claim that, just like our sensory
powers, our power to think is a power possessed by us in virtue of our
having the souls that we do but exercised by us human persons alone, not
by us and our souls. We have this power to think because we have the
souls that we do but not because our souls themselves have this power to
think; rather, again, just as an object’s mass grounds its power (and ten-
dency) to exert a certain attractive force without the mass itself exerting
that attractive force, likewise, according to the Survivalist, one’s having
a human or intellective soul grounds one’s having the power to think
without one’s soul itself having that power. Now some Corruptionists
partly agree with this claim: they maintain that, while informing bodies
and hence parts of human persons, human souls do not and cannot think;
only the persons to which they belong can think. But these Corruptionists
add that our souls can think after our deaths, once separated from a body
and (according to the Corruptionist) no longer part of a human person. I
address this Corruptionist position in the next section (§4). In this section,
I frst address those Corruptionists who maintain that our souls don’t just
think when separated but think even now, while informing our bodies and
parts of us human persons.
Is there any basis for claiming that our souls can think even now, prior
to our deaths? Some Corruptionists would appeal here to the fact that
our souls can think while separated. The idea is that if our souls can think
while separated, then surely they can also think now. Kendall Fisher, for
example, a Corruptionist interpreter of Aquinas, writes, “if the soul is
the understander [i.e., the thinker] after death, there seems to be no prin-
cipled way to rule it out as an understander [i.e., a thinker] during life”
(emphasis in original).12
Now if one’s aim were the exegetical one of showing that Aquinas
thinks that our souls can think even now, before our deaths, such an argu-
ment might work, or at least might work if one could show that Aquinas
thought that our souls could think after our deaths. (I take no stand on
the correct interpretation of Aquinas here). But if the Corruptionist’s aim
is the one under discussion here, viz., to provide a reason or basis to
believe that our souls can think prior to our deaths, then this argument
can succeed only if we already have a reason to believe that our souls can
think after our deaths. But this won’t do since, as I’ll argue in §4, the order
of justifcation must go the other way around: without a prior reason to
believe that our souls can think before our deaths, there is no reason to
believe that our souls can or will acquire the power to think after our
deaths, when they cease to inform a body.
Could Corruptionists invoke Aquinas’s claim that the power of thought
“is in the soul as its subject”? In a well-known passage in his Summa
Theologiae, Aquinas argues,
Some of the operations of the soul are performed without a corpo-

real organ, as in the case of thinking (intelligere) and willing (velle).
Hence, the powers which are the principles of these operations are in
the soul as their subject.13
Commenting on this passage, Patrick Toner, a Corruptionist inter-

preter of Aquinas and a contemporary advocate for Corruptionist
Thomistic Hylomorphism, claims, “It is plain that St. Thomas holds
that the subject of the operations of knowing [i.e., thinking] and will-
ing is the soul.”14
Again, such an argument might work to show that Aquinas thought
that our souls could think before being separated (though, again, I take
no stand here on the correct interpretation of Aquinas). But, again, if the
aim is the one under discussion here, viz., to provide a reason to believe
that our souls can think (before or after our deaths), then the argument
fails. It fails because the claim that the relevant kind of thinking can be
performed without a corporeal organ in no way suggests that there is an
incorporeal part of me that can engage in such thinking, or that the form/
soul by virtue of having which I can engage in such thinking can itself
engage in such thinking.
Some Corruptionist authors seem to believe that what thinks must be
incorporeal if thinking does not require the involvement of any corporeal
organ. Thus, for example, Kendall Fisher writes,
If intellective operations cannot be carried out in or through some-
thing corporeal, they must be carried out by the soul alone. Thus
the soul must be . . . the part of the human being that performs the
intellective acts.15
If the claim that “intellective operations cannot be carried out in or

through something corporeal” is taken to mean that nothing corporeal
can perform these operations, then this claim implies that whatever per-
forms these operations is incorporeal, which may in turn be taken to
imply that we must have an incorporeal part which thinks, a part which
could then be identifed as one’s soul. But the problem is that Thomistic
Hylomorphists hold that we human persons are corporeal and that, as is
evident to all of us, we can and do perform the intellective operations in
question. Given this, no Thomistic Hylomorphist can consistently claim
that our souls can think on the grounds that nothing corporeal can think,
for such a claim would imply that we, whom the Thomistic Hylomorphist
believes are corporeal, cannot think. On the other hand, if the claim that
“intellective operations cannot be carried out in or through something
corporeal” is taken to imply only that the relevant kind of thinking does
not necessarily involve bodily changes of one sort or another, we’re back
to where we were in the preceding paragraph: this claim in no way sup-
ports the idea that our souls, the constituents by virtue of which we have
such a power, can themselves think.16
Again, Corruptionists may claim that the nature of thought is such that
only something that is not wholly material, that is, does not have only
material subsistent parts, can think. Such a claim is arguably implicit in
certain kinds of Thomistic arguments for the thesis that the exercise of
our power of thought does not necessarily involve any corporeal organ
or body-involving process. Such arguments invoke the idea that, just as
seeing essentially involves a certain kind of event (viz., the “reception of a
sensible form”) that takes place in one’s body (in particular, in one’s visual
organ), thinking essentially depends on a certain kind of event (viz., the
“reception of an intelligible form” or storage of a concept) that cannot
take place in anything corporeal and thus must take place in something
incorporeal, specifcally an incorporeal subsistent part of oneself that is
identifed as one’s soul. If this claim were true, it would imply that, since
human persons can think, human persons must have an incorporeal sub-
sistent part. But, even granting all of this, it still provides no more reason
to believe that this incorporeal subsistent part can think than the claim
that my hearing essentially depends on certain events in my ear implies
that my ear can hear (i.e., have auditory experiences).
It is impossible to anticipate every argument that might be invoked to
justify the claim that we have souls (understood in the Aristotelian way)
which can think. Nevertheless, enough has been said to show that it is
doubtful that any such argument will succeed. That is, it is doubtful that
there is any reason for Thomistic Hylomorphists, or any other theorists
who posit Aristotelian souls, to believe that, unlike our sensory powers,
which we possess in virtue of having the souls that we do without our
souls themselves having such powers, our power to think is not just a
power which we have in virtue of having the souls that we do but also a
power which our souls themselves have.17
In fact, the situation is even worse for Corruptionists who make
such a claim, for it is a claim that is not only baseless but also deeply
counterintuitive. The problem is analogous to the one that Thomistic
Hylomorphists would face if they were to claim that our souls can
see, hear, imagine, or exercise any of our other, nonintellectual men-
tal powers. The souls in question, Aristotelian souls, are not us but
constituents of us; we human persons are not identical to these souls,
which are not us but rather forms possessed by us. If one were to claim
that we have such souls and that these souls can see or hear or imag-
ine, then there would be two entities, ourselves and our souls, which
can see, hear, or imagine where we pre-theoretically think that there
is only one (viz., just ourselves). Likewise, Corruptionists who claim
that our souls can think prior to our deaths are committed to there
being two entities which can think, ourselves and our souls, where we
pre-theoretically believe that there is only one (viz., just ourselves). For
this reason, such Corruptionists face a version of the much-discussed
too-many-thinkers problem.18
But is there really a problem here? Some authors have contended that
there isn’t. Richard Swinburne, for example, writes in response to a
too-many-thinkers objection advanced by Eric Olson,
[I]f we (embodied on earth) are not mere souls although our souls
think, then there are two thinking things – I and my soul. . . . I argue
that this is unparadoxical, since there is only one act of thinking
going on – I think, in virtue of my soul thinking. Olson admits (his
p. 76) that “there are some properties we have in a derivative sense.
We are tattooed insofar as our skin is tattooed”, but he seems to
think this unimportant. However, innumerable similar examples can
be adduced . . . The reason these examples do not have paradoxical
consequences is because the diferent descriptions (“I being tattooed”
and “my skin being tattooed”) are descriptions of the same event,
since the descriptions mutually entail each other.19
Though Swinburne is not a proponent of Thomistic Hylomorphism but

of a certain kind of substance dualism, nonetheless what Swinburne says
here in defense of his substance dualist view could be adopted by Cor-
ruptionist Thomistic Hylomorphists who claim that we have souls under-
stood in the Aristotelian way (i.e., substantial forms) which can think
prior to our deaths.
And yet this won’t do, for there are two serious problems with Swin-
burne’s claims. First, Swinburne seems to believe that there aren’t too
many thinkers if the person’s act of thinking can be identifed with her
soul’s act of thinking. But even if one grants that this identifcation can
be made, this is not enough to claim that there is only one thinker, for
even if a person and her soul were to perform all the same acts of think-
ing, each act of thinking would still be performed by two nonidentical
things, the person and her soul, that is, two thinkers. Second, a related
problem confronts Swinburne’s claim that the event of my thinking a
thought and the event of my soul’s thinking that thought are “the same
event” if I think the thought in virtue of my soul thinking the thought.
Since I am not my soul, even if we think the same thought, the event of
my thinking that thought has a diferent constituent than the event of my
soul’s thinking that thought, and so these events are not the same event.
Moreover, it doesn’t help for one to defne a notion of “same event”
according to which two events can be “the same event” even if they have
diferent individuals as constituents (as Swinburne does elsewhere in his
book): this does nothing to solve the problem since my thinking a certain
thought’s being “the same event” in this sense as my soul’s thinking that
thought doesn’t imply that there is only one individual (i.e., one thinker)
thinking that thought.
But perhaps these criticisms refect a misunderstanding of Swinburne’s
position. Perhaps his claim is not that there is really only one thinker but
instead that, while there really are two thinkers, “this is unparadoxical,
since there is only one act of thinking going on – I think, in virtue of my
soul thinking.” This judgment is allegedly supported by the tattoo anal-
ogy. In other words, the idea is that the claim that I think in virtue of hav-
ing a soul which thinks no more creates a problem of too many thinkers
than the claim that someone is tattooed in virtue of having skin which is
tattooed creates a problem of too many tattooed things.
Before criticizing this idea, it is worth noting that Kendall Fisher defends
a similar position in response to the objection that Corruptionist Thomis-
tic Hylomorphists would face a too many thinkers problem if they claim
that human persons have souls that can think. Fisher writes,
In order to have too many thinkers, we must have two distinct think-
ers. To have two distinct thinkers, we must have two distinct things.
But it is misleading, at best, to characterize soul and human being
as two distinct things. . . . When one thing is part of a larger whole,
the two are nonidentical, but they are only distinct to the extent that
the whole exceeds the part. My arm is nonidentical with my elbow,
but it is not distinct from my elbow. So if my elbow is bruised or
scraped, my arm is bruised or scraped. While there are two nonidenti-
cal bruised or scraped subjects – arm and elbow – to the extent that
they are bruised or scraped, they are numerically one . . . [In the same
way,] while a human being is not identical with her soul, or indeed
with any incomplete part of herself, she is not utterly distinct from
it either . . . to the extent that the part operates, it is identical with
the whole. Since the human being is partly her soul, soul and human
being are not distinct candidate understanders [i.e., thinkers].20
Like Swinburne, Fisher argues that the claim that a person thinks in virtue
of having a soul which thinks is no more problematic than other cases
(like her scraped elbow case) in which we claim that a whole has a certain
property in virtue of having a part with that property. But Fisher also
goes further than Swinburne insofar as she asserts that not only is there
just one act of thinking going on but also the person and her soul are
not really numerically two thinkers but rather numerically one thinker,
insofar as to the extent that it is thinking the person’s soul is not distinct
from the person.
However, there are two problems with the ideas proposed here by Swin-
burne and Fisher. First, Swinburne’s and Fisher’s analogies don’t show
that the claim that we have souls which think is any less counterintuitive
than it initially appears. While there are cases (like the tattooed skin and
scraped elbow cases) where we do pre-theoretically believe that a whole
has certain property in virtue of having a part with that property, our
pre-theoretical views about thinking are not like this: none of us pre-
theoretically believes that there is some part of him or her that thinks
(or senses or imagines or engages in any of other mental activities we
pre-theoretically think we alone, and no part of us, engage in). Thus,
Swinburne’s and Fisher’s drawing attention to the fact that there are cases
of wholes having properties in virtue of having parts with those properties
does nothing to make the claim that we think in virtue of having a soul
which thinks any less counterintuitive.
A second and, in my view, even more decisive problem for Swin-
burne’s and Fisher’s positions is that there are some thoughts (including
thoughts involving the application of universal concepts) which human
persons think but which nothing nonidentical to those persons can
think. For this reason, in such cases it cannot be maintained that the
person thinks the relevant thoughts in virtue of his or her soul’s thinking
those thoughts.
I have in mind here frst-personal thoughts. Consider, for example, my
thought I am a human being. According to both Swinburne’s substance
dualism and Thomistic Hylomorphism, I am not my soul, and so it is not
possible for my soul to think that very frst-personal thought; if I have a
soul that thinks, it can only think frst-personal thoughts about itself, not
me. In general, any frst-personal thought which I think is not one that it
is possible for my soul (or anything else nonidentical to me) to think. Con-
sequently, none of the frst-personal thoughts I think are thoughts that
I could be thinking in virtue of my soul’s thinking those very thoughts.
Conversely, any frst-personal thoughts that one’s soul can think are
not thoughts that oneself can think. If, for example, a person’s soul were
to think the frst-personal thought What am I?, a thought which is about
that soul, the person could not also think that very frst-personal thought,
for the only frst-personal thoughts the person can think are about herself,
not her soul. For this reason, in addition to its being untenable to claim
that a person thinks all the thoughts he or she thinks in virtue of having a
soul (nonidentical to her) that thinks those thoughts, it is also untenable
to claim that a person can think any thought which her soul can think if
she has a soul (nonidentical to her) that can think.
These observations allow me to make clearer just how strange it is to
claim that we have proper parts that can think (regardless of whether
these proper parts are Aristotelian souls, Swinburnian souls, or anything
else). To posit such parts truly does introduce additional thinkers, thinkers
that do not and indeed cannot think all of the thoughts that the human
persons to which they belong think, and that can (and presumably do)
think thoughts that the persons to which they belong cannot think.21 I
count myself among those who fnd any view that has such an implication
to be incredible.22
But though it is important to show (against those who, like Fisher
and Swinburne, would claim otherwise) that the claim that we have
souls that can think really does confict with our pre-theoretical beliefs,
this doesn’t get to the heart of the issue. In the course of developing a
theory, we sometimes realize that there are good reasons to renounce
one or more of the pre-theoretical beliefs with which we began. I grant
that this is so and, for this reason, maintain that the real issue is not
whether it is counterintuitive to believe that we have souls that can
think but rather whether there is any good reason to believe that we
have such souls. Of course, I argued above that, in the case of the
Aristotelian souls posited by Thomistic Hylomorphists, there is no
such reason.
I add here that the points just made about frst-personal thought allow
for a further sharpening of that argument. Once it is granted that some
of our thoughts (viz., our frst-personal thoughts) are not thoughts that
we think in virtue of having souls that think those thoughts, it becomes
even harder to see what reason there could be to maintain that the rest
of our thoughts are thoughts that we think in virtue of our having souls
that think those thoughts. If we can think frst-personal thoughts “on our
own,” that is, not in virtue of having souls that think these thoughts, why
can’t we also think non-frst-personal thoughts “on our own,” that is, not
in virtue of having souls that think these thoughts? And if the Thomis-
tic Hylomorphist doesn’t have any reason to suppose that we think our
thoughts in virtue of having souls that think those thoughts, what other
reason could there be to suppose that we have souls that can think? None,
I claim.
4. Our Souls Can Think Only After Our Deaths:
Another Baseless Claim
In §2, I noted that some Corruptionists maintain that human souls can
think both when separated and when unseparated, whereas other Corrup-
tionists maintain that human souls can think only when separated: prior
to our deaths, we human persons, not our souls, can think; however, once
separated at our deaths, our souls come to have a power they did not have
before, namely, the power to think. I turn now to this latter position and
argue that it too is baseless: if there is no reason to believe that our souls
can think prior to our deaths, then there is no reason to believe that they
can think after our deaths.
Consider frst a proposal ofered by Patrick Toner, a leading contem-
porary defender of Corruptionist Thomistic Hylomorphism. Adopting an
idea from Aquinas, Toner claims that our souls can think once separated
but not before because in becoming separated our souls come to have a
diferent “mode of being” and hence “a diferent mode of action.”23 For
now, let it be granted that whether a thing has the power to think is (or
can be) determined by its “mode of being.” The problem with Toner’s pro-
posal is that he ofers no reason to believe that human souls can change
from having a mode of being according to which they cannot think to
having a mode of being according to which they can think. We have seen
that there is no reason to believe that our souls have the latter mode of
being prior to our deaths. What reason then is there to suppose that they
can or will acquire it when they become separated at our deaths? It is, of
course, true that if our souls can continue to exist in a separated state after
our deaths, then they can change from informing a body to not informing
a body, and if one likes, one can describe this change as a change in the
soul’s “mode of being.” But claiming that our souls can undergo such
a change in no way supports the claim that our souls can undergo the
relevant change in “mode of being,” that is, a change from not being able
to think to being able to think.
What advocates of this position need is some reason to think that
our souls changing from informing a body to not informing a body can
result in our souls changing from not being able to think to being able to
think. John Haldane, another contemporary proponent of Corruptionist
Thomistic Hylomorphism, may be thought to provide such a reason when
he writes,
Now the question is whether in the case that A/B ceases to exist cer-
tain of its powers might be transferred to A, which, though hitherto
merely virtual, might now emerge as actual. To fx this idea think of
compound pigment colours such as brown, and the claim that red,
say, exists virtually but not actually in this compound. What that
means is that, certain conditions obtaining, the brown pigment might
be destroyed but red pigment is precipitated out. Might this provide a
model for the post-mortem existence of a subject of abstract thought?
. . . [Thomistic Hylomorphists maintain that] there are powers of the
living human body (the power of abstract thought), and other pow-
ers which are subservient to it, which are not necessarily exercised
through any part of the body, i.e. they have no bodily organ. That
being supposed it cannot be objected that such powers could not be
exercised by some precipitated non-physical substance because they
require a material embodiment. So if thought is an immaterial activity
exercised ante-mortem through the operation of A/B [the human per-
son, composed of (A) a human soul and (B) matter], there is no contra-
diction involved in supposing that it might continue post-mortem but
be transferred from the hitherto actual substance A/B to the hitherto
merely virtual but then actual substance A [the person’s soul].24
Patrick Lee and Robert George, two other prominent proponents of

Corruptionist Thomistic Hylomorphism, cite Haldane’s discussion with
approval and sum up the position thus:
given that the human act of understanding is not itself an act per-
formed with a bodily organ . . . [the] powers of understanding (con-
ceptual thought) and willing could be transferred to a nonphysical
substance that “precipitated out” at death, namely, the separated
human soul.25
Let it be granted that in some cases a power can be transferred from a

whole to a former part of that whole that survives the whole’s destruc-
tion. The question is whether there is any reason to believe that a human
person’s power to think is such a power. The problem is that, for all that
they say, Haldane, Lee, and George ofer no reason to believe that this
particular transfer of powers can occur. Even if there are certain cases
where a power can be transferred from a whole to a surviving former
part when the whole is destroyed, this alone provides no reason to sup-
pose that any power of a whole can be transferred to a surviving former
part or, more importantly, that such a transfer can happen in the specifc
case under discussion, viz., a case involving human souls and the power
of thought.
Here the Corruptionist might think it relevant to note, as Haldane, Lee,
and George do, that thinking (according to Thomistic Hylomorphists)
does not necessarily involve a corporeal organ. But while this claim indi-
cates the absence of one kind of obstacle to the power to think being
transferred from a person to his soul at death, it doesn’t yet provide a
reason to believe that such a transfer is possible. The mere fact that one
kind of obstacle to such a transfer is absent does not show that the trans-
fer is possible.
What about Haldane’s pigment case? Might that provide some
argument by analogy for believing that the power to think could be
transferred from a person to her soul upon the person’s death and the
separation of her soul from her body? One problem here is that it’s not
clear that the pigment case is a case where a power is being transferred
from a whole to a former part/constituent of the whole: what power is
it exactly that the brown pigment has that is transferred to the surviv-
ing red pigment? But even setting this worry aside, there is a crucial
disanalogy between the two cases. In the case of the red pigment, we
have independent knowledge of samples of red pigment that are not
constituents of brown pigment. In particular, we know of such samples
that they have certain powers (e.g., the power to appear red). Given
this, even if we’ve never observed any red pigment that has been precipi-
tated out from some brown pigment, we can nonetheless have reason to
believe that, if the red pigment in some brown pigment were “precipi-
tated out,” it would gain certain powers (e.g., the power to appear red)
that it did not have (or at least could not manifest) while a constituent
of the brown pigment. In the case of human persons and their souls,
however, we have no independent knowledge of human souls that are
not constituents of human persons. A fortiori, we have no independent
knowledge (or reason to believe) concerning such separately existing
souls that they have, or are such that they could have, the power to
think. For this reason, unlike in the case of the red pigment, we have no
basis for thinking that if a human soul were to become separated from
the human person to which it presently belongs, it could or would gain
certain powers, like the power to think, that it did not have while a
constituent of the person.26
This brings me to a further problem with both Toner’s proposal and
Haldane, Lee, and George’s idea: Corruptionists cannot appeal to the
claim that our souls can exist separately to support their claim that our
souls can think once separated. To see this, note that the Corruptionist’s
claim that a human person’s soul can exist separately depends for its
justifcation on the claim that something can exercise the relevant power
of thinking even while the soul is not informing a body (see claim 3 in
§2). Since Corruptionists deny that human persons can exist without their
souls informing a body, the only candidate for this separated thinking
thing is the soul itself (see claims 4 and 5 in §2). Thus, the Corruptionist’s
claim that a human person’s soul can exist while not informing a body
depends for its justifcation on the claim that the soul can think while
not informing a body. Given this, the Corruptionist cannot appeal to the
former claim to support the latter claim; such a move would be viciously
circular.
In short, I don’t see any way for Corruptionists to justify the claim that
even though our souls cannot think before our deaths, they can think
after our deaths.
Edward Feser, a prominent proponent of Survivalist Thomistic Hylo-
morphism, has argued that Corruptionist Thomistic Hylomorphism is
incoherent on the grounds that it is incoherent to suppose that a soul,
understood in the Aristotelian way as a substantial form, can persist after
the substance to which that soul belonged has ceased to exist. Feser writes,
[T]he human being is the substance of which the soul is a form. So,
if the human being ceases to exist at death, then that means that the
substance of which the soul is the form ceases to exist at death. And
in that case, how could the soul carry on? How could a form exist
apart from the substance of which it is the form?27
Such an argument is not likely to persuade any Corruptionists, for Cor-

ruptionists will say in response that a form can exist apart from the sub-
stance of which it is the form if the form can act apart from the substance
of which it is the form. All Thomistic Hylomorphists agree that a thing
exists in the manner in which it acts: if a thing can act under certain
conditions, then it can exist under those conditions. Thus, Feser’s argu-
ment won’t move Corruptionists as long as they can justifably claim that
human souls can act (and, in particular, think) independently of informing
a body.
It is this Corruptionist claim that I have challenged earlier. I have argued
that there is no basis for thinking that the Aristotelian souls Thomistic
Hylomorphists believe in are things that can think, either now (while
informing a body) or after our deaths (when they have ceased to inform
a body). And yet the Thomistic Hylomorphist’s belief that these souls
can continue to exist after our deaths depends for its justifcation on the
claim that these souls are each a principle of a power, a power to think,
which can be exercised even after our deaths. For this reason, if Thomistic
Hylomorphists are to maintain their belief that a human person’s soul
can continue to exist after that person’s death, they must believe that
something else, besides the separated soul, can exercise the relevant power
to think after the person’s death. The only other candidate for such a
post-mortem thinker is the person to which that soul formerly belonged.
Thus, if Thomistic Hylomorphists are to maintain their belief that our
souls can persist after our deaths, they must reject Corruptionism and
instead maintain, as Survivalists do, that we too can persist (and think)
after our deaths, with souls but without bodies informed by those souls.
Notes
1. For a good comparative overview of these diferent theories, see Loose,
Menuge, and Moreland (eds.), The Blackwell Companion to Substance Dual-
ism (2018).
2. For two discussions of these questions informed by developments in contem-
porary science, see Chapters 1 and 2 of this volume.
3. For recent arguments for a Corruptionist interpretation of Aquinas, see Toner,
“Personhood and Death in St. Thomas Aquinas” (2009); Toner, “St. Thomas
Aquinas on Death and the Separated Soul” (2010); Nevitt, “Aquinas on the
Death of Christ” (2016); and Nevitt, “Survivalism vs. Corruptionism: Whose
Nature? Which Personality?” (2020). For recent arguments for a Survival-
ist interpretation of Aquinas, see Stump, “Resurrection, Reassembly, and
Reconstitution: Aquinas on the Soul” (2006); Stump, “Resurrection and the
Separated Soul” (2012); and Brower, Aquinas’s Ontology of the Material
World (2014), ch. 13. References to other authors engaged in this exegetical
dispute can be found in these works.
4. For a prominent articulation of this objection, see Toner, “On Hylemorphism
and Personal Identity” (2009). For one prominent Survivalist response to
this objection, see Oderberg, “Survivalism, Corruptionism, and Mereology”
(2012).
5. For a clear articulation of this objection, see Nevitt, “Survivalism vs. Cor-
ruptionism: Whose Nature? Which Personality?” (2020). In response to this
objection, Survivalists often claim that while having a body is “normal” for
human beings, human beings’ surviving the loss of their entire bodies is no
more impossible than human beings’ surviving the loss of signifcant parts
of their bodies (e.g., all their limbs). See, e.g., Brown, “Some Advantages
for Thomistic Solution to the Problem of Personal Identity Beyond Death”
(2017), 249–256; Eberl, The Nature of Human Persons (2020), 206–208;
Eberl, “Surviving Corruptionist Arguments: Response to Nevitt” (2020); and
Feser, “Aquinas on the Human Soul” (2018), 138–144.
6. For an example of an author pushing this objection, see Brown “Some Advan-
tages for a Thomistic Solution to the Problem of Personal Identity Beyond
Death” (2017). For Corruptionist rejoinders, see Toner, “Personhood and
Death in St. Thomas Aquinas” (2009), 131–132, and Lee and George, Body-
Self Dualism in Contemporary Ethics and Politics (2008), 74–81.
7. For authors pushing this objection, see Hershenov and Koch-Hershenov,
“Personal Identity and Purgatory” (2006), 440–443; Stump, “Resurrection,
Reassembly, and Reconstitution: Aquinas on the Soul” (2006); Stump, “Res-
urrection and the Separated Soul” (2012); Feser, “Aquinas on the Human
Soul” (2018), 144; and Brown, “Some Advantages for a Thomistic Solution
to the Problem of Personal Identity Beyond Death” (2017). For a Corruption-
ist response, see Toner, “St. Thomas Aquinas on Punishing Souls” (2012).
8. Thomistic Hylomorphists, following Aquinas, typically claim not only that
our power to think can be exercised without a body but also that our power
to will can be exercised without a body. Throughout this chapter, I focus
on the power to think. But I note here that the problems for Corruptionism
discussed in this chapter are only made worse when one brings into consid-
eration not only the power to think but also the power to will. For some
indications of this, see notes 17 and 18.
9. See, e.g., Eberl, The Nature of Human Persons (2020), 34–35; Feser, “Aquinas
on the Human Soul” (2018), 144; Oderberg, “Survivalism, Corruptionism,
and Mereology” (2012), 8; Stump, “Non-Cartesian Dualism and Material-
ism Without Reductionism” (1995), 512; Brower, Aquinas’s Ontology of the
Material World (2014), 285–286; and Brown, “A Thomistic Solution to the
Problem of Personal Identity Beyond Death” (2017), 252–253.
10. See Fisher, “Saint Thomas Aquinas and the Too-Many-Thinkers Problem”
(2020); Fisher, “Operation and Actuality in St. Thomas’s Argument for the
Subsistence of the Rational Soul” (2019); Klima, “Aquinas’ Balancing Act:
Balancing the Soul Between the Realms of Matter and Pure Spirit” (2018);
Klima, “Aquinas on the Union of Body and Soul” (2020); and Toner, “Per-
sonhood and Death in St. Thomas Aquinas” (2009).
11. See Davies, The Thought of Thomas Aquinas (1993), ch.11; Haldane, “The
Examined Death and the Hope of the Future” (2001); Haldane, “A Thomist
Metaphysics” (2002), 104–106; Haldane, Reasonable Faith (2010), ch.11;
Lee and George, Body-Self Dualism in Contemporary Ethics and Politics
(2008), 66–74; and Toner, “St. Thomas Aquinas on the Problem of Too Many
Thinkers” (2012).
12. Fisher, “Saint Thomas Aquinas and the Too-Many-Thinkers Problem”
(2020), 107.
13. ST I.77.5.corpus, my translation.
14. Toner, “Personhood and Death in St. Thomas Aquinas” (2009), 128.
15. Fisher, “Operation and Actuality in St. Thomas’s Argument for the Subsis-
tence of the Rational Soul” (2019), 195.
16. Gyula Klima defends a Thomistic view according to which “strictly speaking
only the intellective soul does the activity of thinking” (“Aquinas’ Balanc-
ing Act: Balancing the Soul Between the Realms of Matter and Pure Spirit”
(2018), 37). According to Klima’s proposed view, human persons do not
really think but can be said to think, or at least can be said to think if one
accepts the scholastic idea that “a whole is properly denominated by any
attribute of its part which can apply only to the part in question” (37). The
problem here is that it is clearly false to claim that human persons do not
really think. We’re all aware of our own thinking, and yet we could not be
aware of our own thinking if we did not really think but only could be said
to think because we have souls that think. For this reason, any view which
denies that we really think (and maintains that we can only be said to think)
should be rejected.
17. The same point can be made concerning our power to will. There is no reason
to think that, unlike our sensory powers, our power to will is not only a
power that we have in virtue of having the souls that we do but also a power
that our souls themselves have. Again, such a claim in no way follows from
the idea that the power to will is a power whose exercise does not require the
involvement of any corporeal organ.
18. In the same way, those who would claim that our souls can will things face
a possible too-many-willers problem.
19. Swinburne, Mind, Brain, and Free Will (2012), 236.
20. Fisher, “St Thomas Aquinas and the Too-Many-Thinkers Problem” (2020),
120.
21. A referee suggested that the theorists criticized here could claim that a
human person’s frst-personal thoughts ambiguously refer to either the per-
son or her soul. For example, according to this suggestion, the frst-personal
pronoun “I” in a human person’s thought “I am a human being” would
ambiguously refer to either the person or her soul, just as the noun-phrase
“the person in the next room” can, in a situation where there is more than
one person in the next room, ambiguously refer to one of those persons.
One can think of it this way: the frst-personal pronoun denotes the one
thinking this thought, and so if there are two things thinking this thought,
the person and her soul, the pronoun is ambiguous in its reference. This
suggestion would resolve the problem discussed in the earlier main text if
indeed it were true that human persons’ frst-personal thoughts are ambigu-
ous in this way.
The problem with this is that it is evident to all of us that our frst-personal
thoughts are not ambiguous in this way. The proposed solution only works
by denying that we have an ability we all believe we have, namely, an ability
to think unambiguous frst-personal thoughts about ourselves (and to do so
without employing disambiguating noun-phrases such as “I, the person” or
“I, the human being”). We all ordinarily believe that when we think frst-
personal thoughts like “I am hungry,” “I am a human being,” or “I am not a
soul,” these thoughts are unambiguously about ourselves. Those who adopt
the referee’s suggestion must pay the steep cost of denying this. Beyond this,
they must accept even stranger consequences, such as that a human being who
thinks “I, the soul thinking this thought, am not a human being” is thinking
a true thought; given the view in question, this is a perfectly legitimate disam-
biguation of the human being’s allegedly ambiguous frst-personal thought “I
am not a human being.”
Could one maintain that only persons and not their souls can think frst-
personal thoughts? One should not be misled by the term “frst-personal”
into thinking this is a plausible position. On the contrary, this position is
hopelessly ad hoc: anything that can think can surely think about its own
thinking and hence can think frst-personal thoughts. (It is worth mentioning
here that Aquinas would agree with this, for he (rightly) holds that one way
that the capacity for thought difers from other cognitive capacities is that it is
a self-refexive capacity: whereas, for example, one cannot hear one’s hearing,
anyone who can think can think about her own thinking. See Summa Contra
Gentiles, bk.2, c.49).
22. Contemporary theorists often use the fact that a rival theory implies the exis-
tence of too many thinkers as grounds for rejecting that theory. The most well-
known example of this is Eric Olson’s thinking-animal argument for animalism
(see Olson, What Are We? A Study in Personal Ontology (2007), ch. 2).
23. Toner, “St. Thomas Aquinas on the Problem of Too Many Thinkers”
(2012), 220.
24. Haldane, Reasonable Faith (2010), 158. See also Haldane, “The Examined
Death and Hope for the Future” (2001), 254–255, and Haldane, “A Thomist
Metaphysics” (2002), 104–106.
25. Lee and George, Body-Self Dualism in Contemporary Ethics and Politics
(2008), 73.
26. The same disanalogy features in the molecule example used by Lee
and George in Body-Self Dualism in Contemporary Ethics and Politics
(2008), 72.
27. Feser, “Aquinas on the Human Soul” (2018), 144.
Bibliography
Aquinas, Thomas. Summa Theologiae. In Sancti Thomae de Aquino Opera omnis
iussu Leonis XIII P.M. edita t. 4–12. Rome, 1888–1906.
Aquinas, Thomas. Summa Contra Gentiles: Book Two: Creation. Translated by
James F. Anderson. Notre Dame, IN: University of Notre Dame Press, 1975.
Brower, Jefrey. Aquinas’s Ontology of the Material World: Change, Hylomor-
phism, and Material Objects. Oxford: Oxford University Press, 2014.
Brown, Christopher. “Some Advantages to a Thomistic Solution to the Problem
of Personal Identity Beyond Death.” In Paradise Understood: New Philosophi-
cal Essays about Heaven, edited by T. Ryan Byerly and Eric Silverman, 228–
262. Oxford: Oxford University Press, 2017.
Davies, Brian. The Thought of Thomas Aquinas. Oxford: Oxford University
Press, 1993.
Eberl, Jason. The Nature of Human Persons: Metaphysics and Bioethics. South
Bend, IN: University of Notre Dame Press, 2020.
Eberl, Jason. “Surviving Corruptionist Arguments: Reply to Nevitt.” Quaestiones
Disputatae 10, no.2 (2020): 145–160.
Feser, Edward. “Aquinas on the Human Soul.” In The Blackwell Companion to
Substance Dualism, edited by Jonathan Loose, Angus Menuge, and J. P. More-
land, 131–147. Oxford: Wiley Blackwell, 2018.
Fisher, Kendall. “Operation and Actuality in St. Thomas Aquinas’s Argument for
the Subsistence of the Rational Soul.” The Thomist 83, no.2 (2019):
185–211.
Fisher, Kendall. “Saint Thomas Aquinas and the Too-Many-Thinkers Problem.”
Quaestiones Disputatae 10, no.2 (2020): 106–124.
Haldane, John. “The Examined Death and the Hope of the Future.” Proceedings
of the American Catholic Philosophical Association 74 (2001): 245–257.
Haldane, John. “A Thomist Metaphysics.” In The Blackwell Guide to Metaphys-
ics, edited by Richard Gale, 87–109. Oxford: Blackwell Publishers Ltd, 2002.
Haldane, John. Reasonable Faith. New York: Routledge Press, 2010.
Hershenov, David. “Soulless Organisms? Hylomorphism vs. Animalism.” Ameri-
can Catholic Philosophical Quarterly 85, no.3 (2011): 465–482.
Hershenov, David and Rose Koch-Hershenov. “Personal Identity and Purgatory.”
Religious Studies 42, no.4 (2006): 439–451.
Klima, Gyula. “Aquinas’ Balancing Act: Balancing the Soul Between the Realms
of Matter and Pure Spirit.” Bochumer Philosophisches Jahrbuch für Antike
und Mittelalter 21 (2018): 29–48.
Klima, Gyula. “Aquinas on the Union of Body and Soul.” Quaestiones Disputa-
tae 10, no.2 (2020): 31–52.
Lee, Patrick and Robert George. Body-Self Dualism in Contemporary Ethics and
Politics. Cambridge: Cambridge University Press, 2008.
Loose, Jonathan, Angus Menuge, and J.P. Moreland. The Blackwell Companion
to Substance Dualism. Oxford: Wiley Blackwell, 2018.
Nevitt, Turner. “Aquinas on the Death of Christ: A New Argument for Corrup-
tionism.” American Catholic Philosophical Quarterly 90, no.1 (2016): 77–99.
Nevitt, Turner. “Survivalism vs. Corruptionism: Whose Nature? Which Personal-
ity?” Quaestiones Disputatae 10, no.2 (2020): 127–144.
Oderberg, David. “Survivalism, Corruptionism, and Mereology.” European
Journal for the Philosophy of Religion 4, no.4 (2012): 1–26.
Olson, Eric. What Are We? A Study in Personal Ontology. Oxford: Oxford Uni-
versity Press, 2007.
Stump, Eleonore. “Non-Cartesian Dualism and Materialism Without Reduction-
ism.” Faith and Philosophy 12, no.4 (1995): 505–531.
Stump, Eleonore. “Resurrection, Reassembly, and Reconstitution: Aquinas on the
Soul.” In Die menschliche Seele: Brauchen wir den Dualismus?, edited by
Bruno Niederberger and Edmund Runggaldier, 152–171. Frankfurt: Ontos
Verlag, 2006.
Stump, Eleonore. “Resurrection and the Separated Soul.” In The Oxford Hand-
book of Aquinas, edited by Brian Davies and Eleonore Stump, 458–466.
Oxford: Oxford University Press, 2012.
Swinburne, Richard. Mind, Brain, and Free Will. Oxford: Oxford University
Press, 2012.
Toner, Patrick. “On Hylemorphism and Personal Identity.” European Journal of
Philosophy 19, no.3 (2009): 454–473.
Toner, Patrick. “Personhood and Death in St. Thomas Aquinas.” History of Phi-
losophy Quarterly 26, no.2 (2009): 121–138.
Toner, Patrick. “St. Thomas Aquinas on Death and the Separated Soul.” Pacifc
Philosophical Quarterly 91, no.4 (2010): 587–599.
Toner, Patrick. “St. Thomas Aquinas on the Problem of Too Many Thinkers.”
The Modern Schoolman 89, no.3–4 (2012): 209–222.
Toner, Patrick, “St. Thomas Aquinas on Punishing Souls.” International Journal
for Philosophy of Religion 71, no.2 (2012): 103–116.
10 Angels, Principalities, and
Powers
Travis Dumsday
1. Introduction
Historians of mediaeval philosophy readily grant that angelology was
often the locus for debates on issues of broad philosophical consequence,
debates pertaining to free will, sin, motivation, faculty psychology, indi-
viduation, spatial location, motion, time, and a good deal more. That is
one reason why the secondary literature dealing with Scholastic angelol-
ogy is massive and ever-growing. The present chapter will draw on that
literature without seeking to add to it; its purpose rather is to illustrate
the ongoing relevance of angelology to contemporary neo-Aristotelian-
ism and analytic Scholasticism (movements which, if distinct, are still
functionally joined at the hip). Specifcally, I will argue that one impor-
tant lesson to be gleaned from mediaeval angelology is that if angels are
taken seriously as metaphysically possible beings (whether or not they
are believed actually to exist), then today’s neo-Aristotelian will either
(a) be prompted towards one or another variety of ontological plural-
ism (either pluralism in the realm of substance ontology or pluralism
about the nature of matter) or (b) in the course of resisting such plural-
ism will be prompted to abandon what is (arguably) a commitment of
Aristotelian philosophy of nature, namely the essential role of prime
matter in the individuation of substances.1 To state the argument a bit
more formally:
Premise 1 If angels (understood as fnite, naturally bodiless sub-

stances) are metaphysically possible, then either (a) some form
of ontological pluralism obtains or (b) prime matter doesn’t play
an essential role in the individuation of substances.
Premise 2 Angels (understood as fnite, naturally bodiless substances)
are metaphysically possible.
Conclusion Therefore, either (a) some form of ontological pluralism
obtains or (b) prime matter doesn’t play an essential role in the
individuation of substances.
DOI: 10.4324/9781003125860-13
268 Travis Dumsday
The focus will be on explaining and defending P1, since that is the key
premise for present purposes (namely, illustrating the relevance of ange-
lology to neo-Aristotelianism). While interesting and important, a treat-
ment of P2 is beyond the scope of this discussion.
The remainder of the chapter is structured as follows: in the next sec-
tion I clarify what is intended here by ‘ontological pluralism,’ focusing on
its use within the recent substance ontology literature. Then in Section 3
I review some of the major moves made in the thirteenth-century debate
surrounding universal hylomorphism, emphasizing the part played by the
relatively novel consensus in Western theology regarding the status of
angels as naturally bodiless entities. In the fourth section I sum up the
case for the aforementioned contemporary lesson in pluralism that is to
be drawn from that mediaeval debate and conclude with some quick sug-
gestions regarding other areas of angelology that neo-Aristotelians might
wish to look to by way of fuelling further refection.
2. Ontological Pluralism in the Context of

Substance Ontology
Within the current analytic metaphysics literature are four main compet-
ing models concerning the ontological makeup of substances. These will
be familiar to most readers, but a refresher might still be useful:
A. Substratum Theory A substance is a compound of two ontological

ingredients, namely a bare substratum (AKA a ‘thin particular’
or ‘bare particular’) and a property or set of properties (which
some advocates of the theory take to be tropes, others universals).2
The substratum functions as the ground of the properties; it is
that in which they inhere, that on which they depend for present
existence (no free-foating concrete3 properties allowed), and that
by which they are unifed (in the case of fundamental properties
inherently separable from each other). Those who view properties
as universals typically also assign to the substratum the explana-
tory role of individuating and/or concretizing those universals.
Most substratum theorists maintain that the substratum, while
devoid of properties considered in and of itself (except perhaps
for a few formal properties like self-identity), can only exist when
instantiating some property or properties (i.e., substrata may be
inherently bare but they are too shy ever to leave the house actu-
ally nude). The substratum also renders possible radical change,
such that a substance can switch out and replace all of its proper-
ties while retaining some continuity with its previous self. Advo-
cates of substratum theory include Armstrong,4 Connolly,5
LaBossiere,6 Martin,7 Moreland,8 Pickavance,9 Vallicella,10 and
Wildman.11
Angels, Principalities, and Powers 269
B. Bundle Theory A substance (or what we tend to think of as a
‘substance’)12 is simply a collection of properties (tropes for some,
universals for others) unifed by some sort of relationship (whether
co-location or primitive compresence or mereological fusion . . .).
Properties do not need to inhere in or be grounded by any sort of
non-property but can instead exist freestanding, whether singly or
(the more common view) in clusters. Defenders of bundle theory
include Campbell,13 Denkel,14 Ellis,15 Giberman,16 Keinänen,17
O’Leary-Hawthorne and Cover,18 Paul,19 Robb,20 Shiver,21 and
Simons.22
C. Primitive Substance Theory On this view ‘substance’ is an irreducible
ontological category, and an individual substance is not compounded
out of any more foundational ontological ingredients (like a sub-
stratum + properties); rather an individual substance (a particular
electron, say) is simply the instantiation of a substance-universal/
kind (‘electronhood’). See Broackes,23 Ellis,24 Fales,25 Hofman,26
Loux,27 Lowe,28 Macdonald,29 and Wiggins.30 A rather diferent,
nominalist formulation of the theory is defended by Heil.31
D. Hylomorphism Historically there have been multiple versions of this
theory, difering from each other in more or less important ways.
Today the most commonly defended version is Thomistic hylomor-
phism, so that is the one I will summarize here.32 In common with
primitive substance theory, hylomorphism afrms the reality of
substance-universals/kinds. Unlike primitive substance theory, it
maintains that an individual substance is still a compound of onto-
logical ingredients, namely the substance-universal (AKA ‘substantial
form’) and the prime matter which it actualizes. This binary onto-
logical structure renders hylomorphism akin to substratum theory
(especially in its afrmation of the possibility of radical change, even
change in a thing’s basic natural kind), although diferences remain.
Thus prime matter is not exactly bare but instead can be character-
ized in dispositionalist terms, namely as a pure potency – a potency
specifcally for receiving new substantial form. (In the language of
the contemporary metaphysics of powers, prime matter counts as a
disposition, but not as a dispositional property – rather it is a
substantial disposition, a power to become a diferent kind of sub-
stance by receiving a new substantial form.) While prime matter
plays an important role in individuating the substantial form, it is
not quite the same role as that played by the substratum. Prime
matter is essential for individuation because of its role in circum-
scribing/delimiting the presence of substantial form and rendering
the resultant compound substance apt to receive determinate spatial
extension and divisibility, thereby allowing it in principle to be
parsed out and distinguished spatially from other substances belong-
ing to the same natural kind. A substratum, by contrast, is supposed
270 Travis Dumsday
to individuate all by itself, with no assistance from properties. Also,
prime matter plays no direct role in unifying the properties of a
substance (unlike a substratum), since unifcation of properties is
the job of the substantial form. Prime matter cannot exist on its
own, without being actualized by some substantial form or other.
Advocates of hylomorphism include Bobik,33 Brower,34 Clarke,35
Feser,36 Goyette,37 Madden,38 Oderberg,39 and Toner.40
Some proponents of primitive substance theory cited earlier (Hofman,

Loux, and Lowe) label it as neo-Aristotelian. They do so partly on the
basis of its allegedly refecting the substance ontology Aristotle employs in
his Categories, with hylomorphism viewed as a misstep characteristic of
his later works. More commonly, self-identifed neo-Aristotelians afrm
some version of hylomorphism.41
Advocates of one or another of these theories generally count as monists
with respect to substance ontology; that is, they maintain that their pre-
ferred theory is true of all substances in our universe, and necessarily
so. For instance, the substratum theorist will argue that her theory isn’t
just true of substances in our world but in all possible worlds in which
substances exist, such that bundle theory and its other competitors are
metaphysically impossible. And the same can be said for the other theo-
ries, with the possible exception of hylomorphism.42
Given the prevailing monist sentiment within substance ontology today,
what would it mean to be a pluralist in substance ontology? Elsewhere
I’ve sought to answer this question by ofering a tripartite taxonomy of
pluralisms:
One might distinguish between at least three kinds (or degrees) of plu-
ralism in substance ontology: mild pluralism, according to which only
one substance ontology is true of our world, but there is at least one
other possible world in which a diferent substance ontology obtains;
moderate pluralism, according to which only one substance ontology
is true of our world, but there is at least one other possible world in
which multiple diferent substance ontologies obtain (e.g., a possible
world where some substances are hylomorphic compounds and oth-
ers are compounds of substrata + properties); and extreme plural-
ism, where it is maintained that our actual world contains substances
belonging to diferent substance ontologies (emphasis in original).43
I go on to explore the potential advantages of a version of extreme plural-

ism combining substratum theory with primitive substance theory. Still,
pluralism of any sort remains very much an outlier in the recent substance
ontology literature.
Should it though? In Section 4 I will argue that taking angels seriously
will tend to push neo-Aristotelians either towards some sort of pluralism
(moderate or extreme, depending on whether one accepts the actual reality
of angels or merely their metaphysical possibility) or towards a less-than-
Aristotelian account of individuation. To see why, we need to examine
some of the key moves made in the thirteenth-century debate over univer-
sal hylomorphism, a debate in which angels play a starring role.
3. Angelology and the Mediaeval Debate Over

Universal Hylomorphism
At this point a bit of additional historical background will be useful, by
way of an entry point into the mediaeval debate. Aristotle himself was of
course unaware of the specifcally Judeo-Christian conception of angels,
yet his system included entities that can plausibly be seen as angel-ana-
logues or (more charitably) angel-aspirants: namely, the many incorporeal
conscious agents that function as movers of the heavenly spheres within
Aristotelian astronomy. Copleston summarizes:
As we have seen, God moves the universe as Final Cause, as being the
object of desire. Apparently God is conceived as moving directly the
frst heaven, causing the daily rotation of the stars round the earth.
He moves by inspiring love and desire (the desirable and the intelligible
are the same in the immaterial sphere), and so there must be an Intel-
ligence of the frst sphere, and other Intelligences in the other spheres.
The Intelligence of each sphere is spiritual, and the sphere desires to
imitate the life of its Intelligence as closely as may be. Not being able
to imitate it in its spirituality, it does the next best thing by perform-
ing a circular movement. In an earlier period Aristotle maintained the
Platonic conception of star souls . . . [according to which] the stars
themselves possess souls and move themselves; but he abandoned the
conception in favour of that of the Intelligences of the spheres.44
The picture here is one in which fnite non-physical intelligent substances

efect local motion in heavenly spheres (constructs surrounding the stars
and planets) with which those agents are somehow paired; the spheres
are bodily (composed of the aethereal ffth element distinctive of heavenly
bodies) but their associated Intelligences are not. The bodily spheres have
an imitative telos that directs the manner of their own motion, while the
incorporeal Intelligences paired with the spheres prompt the occurrence of
their motion, in accordance with each Intelligence’s knowledge and desire
pertaining to the highest Unmoved Mover (God). Such at least is a com-
mon and historically infuential interpretation of this portion of Aristotle’s
system; like nearly every portion of that system, its details remain subject
to interpretive dispute.45
Copleston immediately proceeds to discuss a closely related area of
debate: whether Aristotle posits a single Unmoved Mover who functions
272 Travis Dumsday
as the fnal cause for all the Intelligences associated with the various
heavenly spheres, or whether there are instead multiple Unmoved Mov-
ers, perhaps one for every Intelligence. Insofar as Aristotle considered
an Unmoved Mover divine, this issue immediately gets into the wider
debate over whether Aristotle’s philosophy of religion is best classed as
monotheist or polytheist. Copleston also draws a connection between this
interpretive question and later mediaeval Christian angelology:
It is a curious fact that Aristotle does not seem to have had any very
defnite conviction as to the number of unmoved movers. . . . Aristo-
tle speaks of the One Unmoved Mover. But in chapter eight [of the
Metaphysics] the ffty-fve transcendent movers make their appear-
ance. Plotinus afterwards objected that the relation of these to the
First Mover is left wholly obscure. He also asks how there can be a
plurality of them, if matter is the principle of individuation – as Aris-
totle held it to be. Now, Aristotle himself saw this last objection, for
he inserts the objection in the middle of chapter eight without giving
a solution. Even in Theophrastus’ time some Aristotelians clung to
one Unmoved Mover – not seeing how the independent movements
caused by the plurality of movers could be harmonised. It was ulti-
mately due to this notion of a plurality of movers that mediaeval
philosophers supposed there were Intelligences or Angels that move
the spheres (emphasis in original).46
The upshot for our purposes is that Aristotle’s inclusion of multiple dis-
tinct incorporeal substances (whether multiple unmoved movers + the
multiple Intelligences associated with the various heavenly spheres, or
merely the latter) within his cosmology raised a number of difcult meta-
physical questions, questions which he arguably did not fully resolve him-
self but which were to be picked up by early commentators and were later
to play a role in Western Christian angelology.
Regarding that role: Aristotelian infuence in this realm of Christian
thought was a relatively long time in coming, partly because early Chris-
tian angelology (and, more broadly, early Christian theology) was so
diverse. It is well known that the patristic era hosted a variety of difer-
ent and often conficting understandings of the ontology of angels and
demons, both in the East and in the West. Likewise Christian theology
exhibited diversity in its appropriation of earlier pagan philosophical
systems, with diferent Christian thinkers drawing freely and eclectically
from Aristotelianism, Neoplatonism, and Stoicism.
One area of disagreement amongst the Church fathers pertained to
angelic corporeality: do the angels naturally possess bodies? If they don’t,
then how is their purely incorporeal, spiritual status to be understood in
relation to the incorporeal status of God and human souls? Alternatively,
if they do, then what sorts of bodies do they naturally possess? And if
they don’t naturally possess bodies, can they nonetheless take on bodies
temporarily for certain purposes? If so, how does that work and (once
again) what sorts of bodies do they assume? And what sorts of things are
demons? If they are fallen angels, then can the answers to the aforemen-
tioned questions simply be imported wholesale into demonology? And if
they are not fallen angels, what sorts of things are the demons, and how
do we answer these questions as applied to them? Etcetera.
The patristic-era diversity regarding the nature (and more specifcally
the corporeality) of angels and demons persisted in the Christian East,
and with the Eastern Orthodox Church never having laid down formal
dogmas on the subject it remains an occasional topic of discussion there.47
The Roman Catholic Church is a diferent story; despite an equally diverse
range of views present in the Western fathers (Augustine for instance
generally held to corporeal angels and demons), which diversity persisted
up through the early portion of the high middle ages (Peter Lombard’s
twelfth-century Sentences remains ambivalent on the question), in the
thirteenth century the view of angels as naturally bodiless won out in the
West. This was partly due to the growing infuence of the Latin transla-
tion of the Dionysian corpus, which clearly afrms the idea. The Catho-
lic Church made angelic incorporeality a dogma at the Fourth Lateran
Council in 1215, the relevant portion of which informs us that God is
“creator of all things visible and invisible, spiritual and corporeal, who
by his almighty power together from the beginning of time formed out
of nothing the spiritual creature and the corporeal creature, that is, the
angelic and the terrestrial.”48
Even after 1215, however, discussion persisted about what precisely
angelic incorporeality meant. This became particularly pressing given
the need to determine how best to integrate Christian angelology into
the newly dominant Aristotelian philosophy of nature, and in particular
with its hylomorphic substance ontology.49 The various attempts to do so
ended up giving rise to three main sorts of theory, which can be initially
sketched as follows:
A. Universal Hylomorphism Only one substance ontology naturally

obtains of fnite substances in our world (or indeed any possible
world), namely hylomorphism. Consequently, there must be some
way in which both terrestrial creatures and incorporeal angels can
rightly be considered composites of form and matter (in some sense
of ‘matter’). This is the route pioneered by the eleventh-century
Jewish philosopher Avicebron50 and later championed within Chris-
tian thought by a number of fgures, including the Dominican Rich-
ard Fishacre and prominent Franciscans like Alexander of Hales,
his pupil Bonaventure, and Roger Bacon. This seems (emphasis on
seems) to entail a pluralistic understanding of the nature of matter,
according to which there are at least51 two sorts of matter: a spiritual
274 Travis Dumsday
matter present in the bodiless angels and a corporeal matter present
in bodily things.
B. A Version of Extreme Pluralism (Primitive Substance Theory +
Hylomorphism) This the solution championed by Aquinas and the
later Thomistic school. He opts simply to give up the claim that a
single ontology holds true of all fnite substances. Instead, two
distinct substance ontologies naturally52 obtain within our world:
primitive substance ontology for both the fallen and unfallen incor-
poreal angels (each angel is devoid of any sort of matter, being
instead the simple instantiation of a unique and non-repeatable
substantial form, with each angel thereby constituting the sole pos-
sible member of a distinct natural kind), and hylomorphism for
every other fnite substance.
C. Retain Aspects of Aristotelianism but Drop Its Account of Individu-
ation There are several ways of pursuing this third option,53 but the
one that achieved most prominence is that of John Duns Scotus,
whose abandonment of universal hylomorphism led to its downfall
as a doctrinal distinctive of the Franciscan order.54 Scotus maintains
that although prime matter exists in the physical realm, it plays no
essential role in individuation there and is wholly absent from angels.
There is no need to posit a distinct kind of matter for the bodiless
angels (as his Franciscan forebears were wont to do), nor any need
to claim that each angel constitutes a species unto itself (as Aquinas
claimed). Instead, there are many bodiless and matterless angels all
belonging to a single natural kind, ‘angel,’ distinguished from one
another ontologically by their distinct individual essences or ‘haec-
ceities’ – Scotus’s innovative and thoroughly non-Aristotelian con-
tribution to the metaphysics of individuation.55
I will not attempt to summarize all of the many arguments that lie
behind and motivate these three positions, but I should present at least a
few lines of reasoning by way of clarifcation. Position (A) was prompted
in part by a desire to safeguard divine omnipotence. Prima facie it seems
like God ought to be able to create many individual members of a single
angelic species should He wish to do so (there appears no obvious con-
tradiction in the idea), moreover there are reasons to think that He would
wish to do so. Alexander of Hales for instance suggests that creating
multiple individual angels of the same species would serve to manifest
God’s goodness and glory especially well:
For in spirits a plurality of individuals does not exist for that reason
[referencing an earlier point about animal species needing many indi-
vidual members for the sake of preserving the species over time, since
all individual animals die], but instead for the manifestation of divine
goodness, which aims to extend to many a participation in beatitude:
so the number of individuals under any species is multiplied for the
sake of the fruition of beatitude. Moreover, a temporal king is able
more efectively to display a ftting order in his palace with many
servants serving in much the same station, than if only one servant
occupied one ofce: like if there were multiple cup-bearers rather than
only one cup-bearer. Since therefore the power of the king of heaven
be infnitely disproportionate to the power of a temporal king and the
order of the heavenly palace more ftting than that of the earthly, it is
appropriate that angels, who are the highest ministers, be multiplied
in one species.56
But how are multiple bodiless entities to be individuated from each other,
within a broadly Aristotelian framework? (Recall Plotinus’s allegation
that Aristotle failed to address this very question.) Alexander of Hales
does not delve into this issue in much detail, and while Roger Bacon
supports universal hylomorphism, his reasons for doing so centre around
issues other than individuation (chiefy a concern to ensure that angels
remain contingent beings, naturally susceptible to annihilation by God).57
By contrast, this is an important part of Fishacre’s case for universal hylo-
morphism, and Long concisely summarizes one of his arguments pertain-
ing to individuation:
Likewise, every individual is an individual through matter. If the angel

is the same in this and in that property, what will you add to it that
it might be this particular angel? Either an accident or a (specifc)
diference or matter. If an accident, then individuals difer by acci-
dents alone. And if substances can be imagined without accidents
(the implication is that they can), they will be the same substance.
If, however, angels difer by specifc diference, then each angel will
belong to a distinct species – an option Fishacre thanks so little of
that he does not even bother to refute it. By process of elimination he
decides that individuation – including angelic individuation – has to
be through matter (emphasis in original).58
Bonaventure similarly supports the notion that the individuation of fnite

substances always requires matter; for him, it is the conjunction of these
two metaphysical principles that serves to individuate a thing, and that
point applies to any fnite substance. Matter, considered in and of itself,
is defned in dispositionalist terms: it is pure potency for the reception
of substantial form. What kind of substantial form? Any kind, whether
spiritual or corporeal. When it receives a spiritual substantial form the
resultant compound substance is an individual angel, and we can then
reasonably label the angel’s matter ‘spiritual matter.’59 Likewise, when
matter receives a corporeal substantial form (one naturally apt for quan-
titative accidents, chiefy spatial extension and division), the resultant
276 Travis Dumsday
compound substance is an individual body, and we can then reasonably
label that body’s matter ‘corporeal matter.’ But at bottom there is just
matter simpliciter, with the divisions between spiritual and corporeal mat-
ter arising from the diversity of substantial forms with which matter can
be conjoined. As Cullen puts it, “The essence of matter is identical in all
substances, and its being is conferred and diferentiated by its form. Mat-
ter is one, even under the being of diferent forms, whether spiritual or
corporeal.”60 Given the Aristotelian principle that matter is essential for
the individuation of substances belonging to the same natural kind, the
importation of matter into the composition of angelic substance would
seem to address the individuation question. Pini summarizes Bonaven-
ture’s conclusion:
Bonaventure concluded that all creatures are composed of matter and

form. This does not entail, however, that any thing composed of mat-
ter and form is also composed of quantitative parts (i.e., parts with a
certain extension), for Bonaventure carefully distinguished between
matter, on the one hand, and extension, on the other. Quantity, not
matter, makes something extended, i.e., corporeal. Accordingly, some
things may be composed of matter and not be extended. This is the
case with angels. Thus angels are both material and spiritual, i.e.,
not corporeal. Bonaventure’s position concerning the composition of
matter and form in any created thing, angels included, allowed him to
give a unitary solution to the problem of individuation for all created
things. Even though angels are diferent in kind from corporeal crea-
tures, they are individuated in the same way, i.e., as a result of their
being composed of matter and form. Neither matter nor form, taken
by themselves, accounts for the individuality of a certain thing. It is
because of their union that a certain individual exists. Accordingly,
Bonaventure could easily account for the numerical multiplication of
angels within the same species.61
Bonaventure also thinks that matter is required in order to serve as an

ontological ground for the variety of changes that an angel is capable of
undergoing. For him, a principle of potency is needed to help ground the
possibility of change – another very Aristotelian claim – and for Bonaven-
ture, potency = matter (in contrast to corporeal matter, which is matter
+ extension).62
The upshot of adopting option (A): pluralism about substance ontology
is avoided (hylomorphism for all!). But does the avoidance of pluralism
in that realm end up imposing it elsewhere, like an annoying bulge in the
rug which, pressed down, inevitably pops up elsewhere? As noted ear-
lier, universal hylomorphism, upon initial viewing, does tend to give the
impression of entailing a pluralism about the nature of matter, resulting
in the positing at least two (and, for Bacon, three) distinct sorts of matter.
I have just suggested that Bonaventure’s version of universal hylomor-
phism forestalls that accusation, since, for him, matter in and of itself has
nothing to do with ‘body’ in the ordinary sense (three-dimensional mid-
sized dry goods); bodily/corporeal entities result from matter (potency)
being actualized by certain sorts of substantial form, namely those with
an aptitude for spatial extension and divisibility. The resultant compound
substance can then be said to have ‘corporeal matter’ (rather than ‘spiri-
tual matter’), but the origin of the dividing line is really to be found in
the associated substantial forms. To claim that Bonaventure’s view entails
a pluralism about matter is just to confuse the technical sense of materia
he is employing with the ordinary English ‘matter,’ which does of course
have a diferent sense.
However, while I fnd that line plausible, I am not sure it is wholly
satisfactory. Aristotle’s original hylomorphism (so far as I understand
it) accords prime matter a key role in individuation precisely because his
prime matter is a potency for the reception of those substantial forms
which are naturally apt for determinate spatial extension and division.
Aristotle had no notion of ‘prime matter’ receptive of substantial forms
incompatible with spatial extension and division (i.e., purely spiritual
substantial forms). Maybe he should have (thus allowing for a prime
matter that would in some cases function as spiritual matter, individu-
ating his many spiritual Intelligences and/or Unmoved Movers), but he
didn’t. For better or worse, Bonaventure’s materia is not the prime mat-
ter of Aristotle; Bonaventure’s matter can do things (receive substantial
forms incapable of spatial extension and division) that Aristotle’s can’t.
A modern-day Bonaventurean might defend the account as a necessary
improvement on Aristotle’s model, but the fact remains that Aristotle’s
prime matter matches Bonaventure’s corporeal matter but not Bonaven-
ture’s matter simpliciter (a generic potency susceptible of receiving either
spiritual or corporeal substantial forms). Today’s neo-Aristotelian is in
fact liable to question whether a single power could really be receptive
of such radically diferent sorts of inputs (spiritual forms + corporeal
forms), or whether instead it makes more sense to posit two distinct sorts
of powers here, and correspondingly two fundamentally distinct sorts of
potency/matter. Analogous disputes about the status of certain powers
arise all the time; consider the debate over whether various aptitudes
(say, for mathematics and music theory and second-language learning)
arise from distinct cognitive faculties or whether they are instead bet-
ter understood as specifc manifestations of a single overarching faculty/
power like ‘intelligence.’ There is an argument to be made that in fact
what initially might have seemed like wholly distinct powers are indeed
rooted in one multi-track disposition (i.e., a single power with many dis-
tinct but related manifestations). By contrast, there is no plausible argu-
ment to the efect that an aptitude for mathematics and an aptitude for
powerlifting are rooted in a single multi-track disposition – to explain
278 Travis Dumsday
their joint presence in a single substance, ultimately we have to posit
at least two wholly distinct dispositions. These sorts of debates are not
merely the product of the higher-level special sciences but crop up also in
the metaphysics of physics.63 At any rate, the worry is simply that given
the radical distinction between spiritual and corporeal substantial forms
(angels and earthworms, say) it makes more sense to suppose that there
must be distinct potencies for reception of such distinct forms, just as the
power for powerlifting must be distinct from the power for trigonometry.
Consequently, from a neo-Aristotelian perspective at least, Bonaventure’s
position might still be viewed as plausibly entailing the positing of mul-
tiple matters (i.e., a pluralism about the nature of matter), regardless of
Bonaventure’s own intent.
Let’s turn now to consider part of the Thomist case for option (B). It is
well known that Aquinas rejects any notion of spiritual matter, opting for
purely spiritual angels that are incorporeal and immaterial in every sense.
Across multiple works produced over a period of decades he provides a
variety of both a posteriori and a priori arguments for the existence of
angels understood in this manner. Here is a very Aristotelian example of
the former sort (supplemented by an appeal to dubiously apostolic-era
authority), from chapter 91 of book II of the Summa Contra Gentiles:
Movement that is continuous, regular, and in its own nature unfailing

must be derived from a mover which is not moved, either through
itself or by accident. . . . Moreover, a plurality of movements must
proceed from a plurality of movers. The movement of the heaven,
however, is continuous, regular, and in its nature unfailing; and
besides the frst movement, there are many such movements in the
heaven, as the studies of the astronomers show. Hence, there must be
several movers which are not moved, either through themselves or
by accident. But . . . no body moves unless it is itself moved; and an
incorporeal mover united to a body is moved accidentally in keeping
with the movement of the body, as we see in the case of the soul.
Hence, there must be a number of movers which neither are bodies
nor are united to bodies. Now, the heavenly movements proceed from
an intellect. . . . We therefore conclude to the existence of a plurality
of intellectual substances that are not united to bodies. With this con-
clusion Dionysius is in agreement, when, speaking of the angels, he
says that “they are understood to be immaterial and incorporeal.”64
For an example of an a priori argument (one that directly conficts with

the previously cited argument by Alexander of Hales), consider the fol-
lowing from chapter 93 of the same work:
Furthermore, in each individual that which belongs to the species is

superior to the individuating principle, which lies outside the essence
of the species. Therefore, the universe is ennobled more by the mul-
tiplication of species than by the multiplication of individuals of one
species. But it is in separate substances, above all, that the perfection
of the universe consists.65 Therefore, it is more consonant with the
perfection of the universe that they constitute a plurality, each diverse
in species from the other, rather than a numerical multiplicity within
one and the same species.66
In the same chapter Aquinas also argues from the Aristotelian account
of individuation (as he understands it) to the falsity of Bonaventure’s
plurality-of-same-species-angels idea:
Moreover, things specifcally the same, but numerically diverse, pos-

sess matter. For the diference that results from the form introduces
specifc diversity; from the matter, numerical diversity. But separate
substances have no matter whatever, either as part of themselves or
as that to which they are united as forms. It is therefore impossible
that there be several such substances of one species.67
This last argument might be stated a bit more formally as follows:
Premise 1 If matter is essential to the individuation of fnite sub-

stances belonging to the same species, then since there are multiple
immaterial angels they cannot belong to the same species.
Premise 2 Matter is essential to the individuation of fnite substances
belonging to the same species.
Conclusion Therefore, since there are multiple immaterial angels
they cannot belong to the same species.
It is similar both in structure and content to the aforementioned argument

for the opposing side:
Premise 1 If matter is essential to the individuation of fnite sub-

stances belonging to the same species, then since there are multiple
incorporeal angels belonging to the same species they must possess
matter of some sort.
Premise 2 Matter is essential to the individuation of fnite substances
belonging to the same species.
Conclusion Therefore, since there are multiple incorporeal angels
belonging to the same species they must possess matter of some
sort.
The divide is obviously over the consequent of P1, which perhaps indicates
that settling the dispute in favour of Bonaventure and company will involve
showing that they have the better overall a priori/axiological case, and that
280 Travis Dumsday
any a posteriori arguments ofered by Aquinas are either faulty or that the
explanatory work for which he invokes angels can be accomplished just
as well by incorporeal rather than wholly immaterial angels. Settling it in
favour of the Thomist will of course involve countering such claims or
showing that universal hylomorphism is metaphysically impossible.
At any rate, for our purposes the key point is that by rejecting universal
hylomorphism and positing that each angel is the sole possible member
of its very own species, Thomists avoid any possible slide into a plural-
ism about the nature of matter (there is unambiguously just one kind of
matter, namely the prime matter of corporeal things) and adopt in its
place a pluralistic approach to substance ontology, in which primitive
substance theory holds true for some fnite substances (the angels) and
hylomorphism for all the rest.
Finally, option (C) is most famously adopted by the Scotists. They are
in the position of being able in this context to avoid pluralism about the
nature of matter and pluralism about substance ontology, all the while
retaining a unifed account of individuation. As Lang observes, Scotus
“had argued that angels are pure spirit containing no matter whatever;
nevertheless, because they are created, angels must be fnite beings and
fnite causes.”68 Scotism’s answer to the problem of angelic individuation
is the same as its answer to the individuation of any fnite substance: the
haecceity. Every fnite substance is a compound of (at least) a common
nature and an individuating haecceity. Prime matter exists in material
things, but it fulfls explanatory roles other than individuation (notably,
it still functions as the source of continuity through changes in substantial
form). In its functional role as inherent individuator the Scotist’s haecceity
is analogous to the substratum of contemporary substratum theory, yet
difers from the substratum in a number of important respects (e.g., it
plays no independent role in grounding or unifying properties and is not
to be understood as that in which properties inhere), which diferences
arguably allow it to avoid some of the major criticisms facing substratum
theory.69 I will not try to summarize the case for the Scotist account of
individuation, which, in Scotus’s own works, consists largely of argu-
ments from elimination (i.e., haecceities are metaphysically possible and
must be real because prime matter cannot individuate and neither can
properties and neither can existence. . . .).70
So option (C) appears to have at least a prima facie advantage in terms
of parsimony (one substance ontology applying to all fnite things +
one nature of matter + one model of individuation applying to all fnite
things), though at the cost of introducing a new ontological principle,
the haecceity. From the perspective of the neo-Aristotelian, however, that
addition is liable to seem a large theoretical cost, perhaps constituting a
hefty departure from traditional Aristotelianism.
Indeed, the degree of departure perhaps warrants additional empha-
sis. Arguably the Scotist is not only supplying a diferent solution to the
problem of angelic individuation than that supplied by the Bonaventurean
or Thomist, but is also conceiving the problem itself rather diferently.
We are here skirting around the edges of deep and turbulent conceptual
waters, but I would briefy note the following: for the Thomist (and argu-
ably also Aristotle himself), the problem of individuation is at bottom
the problem of what distinguishes two clearly distinct members of one
and the same natural kind. You and I are both humans but we’re clearly
diferent humans, so what grounds the distinction between us? It can’t be
our humanity (that’s what’s shared), so it must be something else, and the
chains of deduction and elimination eventually hit upon signate matter
as the ground of distinction (i.e., prime matter actualized by substantial
form and marked by quantitative accidents, notably including determi-
nate spatial extension). From that Thomist (and probably Aristotelian?)71
perspective, if there were only one human then there would be no problem
of individuation with respect to humanity. Likewise, there is no problem
of individuation when it comes to entities belonging to distinct natural
kinds. When I and my octopus sit contentedly side by side on the park
bench we provide passers-by with no ingredients for pondering the prob-
lem of individuation, since we clearly belong to diferent kinds. Hence,
the Thomist solves the problem of angelic individuation by positing that
every angel is the sole member of its natural kind. Gabriel and Michael are
as diferent from one another (in terms of kind membership) as I and my
octopus are, so no problem of individuation arises for them. For the Sco-
tist, by contrast, there would still be a pressing problem of individuation
even in scenarios where there is only one human being or only one angel.
The key question is what delimits/contracts/circumscribes any universal72
such that it is realized as a particular entity. ‘Humanity’ could in principle
be shared by many, even if, as a matter of contingent fact, it is actualized
only by one, and the same can be said for ‘angelness.’ In the hierarchy of
universals, more generic kinds are circumscribed by and realized in more
specifc kinds – for example, ‘substance’ gets circumscribed by ‘living’
which then gets circumscribed by ‘animal’ which then gets circumscribed
by ‘octopus.’ (No doubt a biologist would cringe at the oversimplifca-
tion, but leave that to one side.) Assume that ‘octopus’ here counts as
the infmic species of this hierarchy (i.e., the lowest-level kind, indivisible
into more specifc kinds). Well, what circumscribes that infmic species
‘octopus’ such that it can be realized as an individual in the actual world?
For Scotists, within the corporeal realm this can’t be prime matter, even
prime matter qua combined with determinate spatial extension. (On their
view, prime matter itself needs individuating.) Rather, what circumscribes
‘octopus’ must be inherently individual and individualizing, a primitive
ontological component of a thing that serves only to mark it out and
render it as that thing. This is the haecceity. For Scotists, the haecceity
is what delimits/contracts/circumscribes ‘octopus’ such that we get my
particular benchmate. The same applies in the incorporeal realm, such
282 Travis Dumsday
that Gabriel’s haecceity is what delimits/contracts/circumscribes ‘angel-
ness’ such that we get Gabriel. The haecceity is an objectively real (if
inseparable) component of the resultant entity, just as its infmic species
and all the higher-level, more generic kinds are also objectively real (if
inseparable) components73 – though the haecceity is a unique, sui generis
and non-formal sort of component. (I.e., the haecceity is not technically
among the formal causes of a substance, even while these components are
all referred to as ‘formalities’ within Scotism.)
Whatever one says about the strengths and weaknesses of this Scotist
manner of conceiving the problem of individuation (and its solution),74
that there is a divide between it and that of the Thomist is undeniable.
Historically undeniable too is the productive role played by angelology
(and more specifcally the later Catholic notion of wholly spiritual, incor-
poreal angels) in helping to spur on the clarifcation and elaboration of
these diferent perspectives, by providing fuel for thought experiments
about individuation in the absence of any sort of matter.
Having covered the relevant historical material, let’s turn to a brief
consideration of what all this means for neo-Aristotelians interested in
how their philosophy of nature relates to Christian theology.
4. Some Lessons for Neo-Aristotelians

The goal of this chapter has been to illustrate the utility of angelology
for neo-Aristotelianism. The strategy for achieving that goal has been to
explain and defend the tidy modus ponens laid out in the Introduction:
Premise 1 If angels (understood as fnite, naturally bodiless sub-

stances) are metaphysically possible, then either (a) some form
of ontological pluralism obtains or (b) prime matter doesn’t play
an essential role in the individuation of substances.
Premise 2 Angels (understood as fnite, naturally bodiless substances)
are metaphysically possible.
Conclusion Therefore, either (a) some form of ontological pluralism
obtains or (b) prime matter doesn’t play an essential role in the
individuation of substances.
The preceding section constituted the bloated justifcation for P1. There,
we saw how the mediaeval debate over universal hylomorphism was posi-
tively steeped in angels; their mere metaphysical possibility sufced to
prompt original refections on the scope of the Aristotelian metaphysics
of fnite substance. Bonaventure and others made the case that hylomor-
phism should be extended from the corporeal into the incorporeal realm,
arguably resulting in a pluralistic ontology of matter (though Bonaven-
ture himself would have denied this); it is a moderate pluralism if angels
are seen as merely possible, extreme pluralism if they are viewed as real
(to recall the terminology deployed in Section 2). Alternatively, Aquinas
rejects universal hylomorphism but in doing so still ends up with a form
of pluralism, this time in the realm of substance ontology (again, either
moderate or extreme, for him certainly extreme). As a third prominent
option, those pluralisms are avoided but at the cost of abandoning the
traditional Aristotelian account of individuation (or at least one common
interpretation of that account), which accords an essential role to prime
matter in the individuation of substances.
Obviously this defence of P1 is hardly airtight. As noted, the Bonaven-
turean will resist my claim that she is pushed into pluralism about matter,
and not unreasonably; moreover the consequent of P1 may not present
an exhaustive list of options – perhaps there are ways to entertain angels
as possible beings without having to accept any sort of pluralism and
without having to drop prime matter’s individuating function. Thank-
fully, however, the argument needn’t be airtight in order for it to do its
assigned job: to illustrate one way in which Christian angelology holds
interest for neo-Aristotelian theorizing about nature.
In fact there are plenty of other ways that could have been used for
illustrative purposes, among them: (a) as an advocate of universal hylo-
morphism, Alexander of Hales holds that angels are compounds of
spiritual matter and substantial form, yet still count as naturally incor-
poreal: they naturally lack bodies consisting of ordinary matter (the
fve elements). However, Alexander also holds that angels can assume
ordinary bodies for providential missions, and his model of such bodily
assumption is much more ontologically robust than that of some later
Scholastics (notably Aquinas). Elsewhere I’ve argued that Alexander’s
model of such voluntary, temporary angelic corporeality is of potential
interest for contemporary substance dualists, supplying as it does a pic-
ture of how the contingent embodiment (and resultant spatial location)
of a naturally incorporeal entity might work (which would in turn be
of relevance to those neo-Aristotelians who think their view compatible
with substance dualism).75 (b) Toner shows the relevance to contempo-
rary virtue ethics (a popular ethical theory amongst neo-Aristotelians)
of the duelling Thomist versus Scotist accounts of the origins of angelic
sin.76 (c) Lang’s work demonstrates how Scotus’s innovative account of
angelic location challenges traditional Aristotelian conceptions of place
while at the same time laying down new concepts forming historical
bridges to ideas in Newtonian physics.77 (d) I make the case that the
distinctive role of hylomorphism in the Halensian, Bonaventurean, and
Thomist accounts of demonic agency (in particular their working of
deceptive ‘signs and wonders’) can contribute to current discussions
in philosophy of religion concerning the problem of natural evil.78 (e)
Guthrie draws attention to some of the ways in which refecting on
the possibility of angels and demons interacting causally with material
objects could shed light on the interaction problem facing substance
284 Travis Dumsday
dualism (which, again, could be of interest to neo-Aristotelians opting
for that account of mind).79
Further examples could be enumerated, but hopefully that sufces to
show that angels are a neo-Aristotelian’s friend. (Read that as you will.)
Notes
1. Some analytic Scholastics will already be on board with that latter option,
namely neo-Scotists – but then the Scotist model of individuation has often
been counted as one of that school’s more signifcant departures from Aris-
totle’s original system. More on this later.
2. For Scholastic readers: I am using ‘property’ here in the sense common within
the analytic literature, namely as a synonym for ‘characteristic/attribute/
trait.’ Unless otherwise indicated, I am not using it in the Scholastic sense
of a ‘proprium’ as distinguished from a contingently possessed accident. A
related terminological clarifcation: when seeing ‘universals’ here, Scholas-
tics should read ‘common natures.’ On many Scholastic systems (including
both Thomism and Scotism), a common nature like ‘humanity’ is something
multiply realizable, having at least two potential ways of being. One man-
ner of being is as actualized by real particular people (you and me and Avril
Lavigne), and the other is as a universal concept abstracted from particulars
by the human mind (the object of your thought when you’re thinking about
‘humanity’). The common nature, considered in and of itself, is neither uni-
versal nor particular – transcending that divide is what allows it to have either
manner of existence. (One major bone of contention between Thomists and
Scotists concerns the existential status of the common nature and whether any
sort of being or unity can rightly be attributed to it considered apart from its
realization as a particular or universal.) Among the disconnects in terminol-
ogy between historical Scholasticism and contemporary analytic metaphysics
is ‘universals,’ precisely because the universals/common natures distinction
is not employed in the analytic literature. In what follows I will adhere to
customary analytic usage, but Scholastic readers should understand my ‘uni-
versals’ as common natures. (I do this by way of imperfect compromise in
addressing two sorts of audiences; I do not thereby presuppose any specifc
answer to the question of whether ‘common nature’ can be fully equated
with ‘universal’ as the latter is typically understood in the analytic realm.
Relatedly, I make no claim about where Thomism and Scotism should fall
when plugged into contemporary analytic taxonomies of nominalists versus
realists versus moderate realists etc. Such matters could use further attention
by analytic Scholastics.) For analytic readers unfamiliar with this Scholastic
distinction, consult Joseph Owens, “Common Nature: A Point of Compari-
son Between Thomistic and Scotistic Metaphysics,” Mediaeval Studies 19
(1957): 1–14, and also his “Thomistic Common Nature and Platonic Idea,”
Mediaeval Studies 21 (1959): 211–223. Useful too is the concise discussion
by Francis Clarke, “St. Thomas on ‘Universals’,” Journal of Philosophy 59
(1962): 720–725.
3. Some substratum theorists (like J.P. Moreland) are platonists about abstracta,
and so would allow for independently existing abstract properties. But they
draw the line at admitting freestanding concrete (i.e., instantiated) properties –
concrete properties need to be paired with substrata.
4. David Armstrong, A World of States of Afairs (Cambridge: Cambridge Uni-
versity Press, 1997).
5. Niall Connolly, “Yes: Bare Particulars!” Philosophical Studies 172 (2015):
1355–1370.
6. Michael LaBossiere, “Substances and Substrata,” Australasian Journal of
Philosophy 72 (1994): 360–370.
7. C.B. Martin, “Substance Substantiated,” Australasian Journal of Philosophy
58 (1980): 3–10.
8. J.P. Moreland, “Theories of Individuation: A Reconsideration of Bare Par-
ticulars,” Pacifc Philosophical Quarterly 79 (1998): 251–263; Universals
(Montreal and Kingston: McGill-Queen’s University Press, 2001); “Exempli-
fcation and Constituent Realism: A Clarifcation and Defense,” Axiomathes
23 (2013): 247–259.
9. Timothy Pickavance, “In Defence of ‘Partially Clad’ Bare Particulars,” Aus-
tralasian Journal of Philosophy 87 (2009): 155–158; “Bare Particulars and
Exemplifcation,” American Philosophical Quarterly 51 (2014): 95–108.
10. William Vallicella, “Three Conceptions of States of Afairs,” Noûs 34 (2000):
237–259.
11. Nathan Wildman, “Load Bare-ing Particulars,” Philosophical Studies 172
(2015): 1419–1434.
12. Some bundle theorists are eager to remove ‘substance’ from the list of our
fundamental ontological categories.
13. Keith Campbell, Abstract Particulars (Oxford: Blackwell, 1990).
14. Arda Denkel, Object and Property (Cambridge: Cambridge University Press,
1996); “On the Compresence of Tropes,” Philosophy and Phenomenological
Research 57 (1997): 599–606.
15. Brian Ellis, The Metaphysics of Scientifc Realism (Montreal and Kingston:
McGill-Queen’s University Press, 2009).
16. Daniel Giberman, “Tropes in Space,” Philosophical Studies 167 (2014):
453–472.
17. Markku Keinänen, “Tropes – The Basic Constituents of Powerful Particu-
lars?” Dialectica 65 (2011): 419–450.
18. John O’Leary-Hawthorne and Jan Cover, “A World of Universals,” Philo-
sophical Studies 91 (1998): 205–219.
19. Laurie Paul, “Logical Parts,” Noûs 36 (2002): 578–596.
20. David Robb, “Qualitative Unity and the Bundle Theory,” Monist 88 (2005):
466–492.
21. Anthony Shiver, “Mereological Bundle Theory and the Identity of Indiscern-
ibles,” Synthese 191 (2014): 901–913.
22. Peter Simons, “Particulars in Particular Clothing: Three Trope Theories
of Substance,” Philosophy and Phenomenological Research 54 (1994):
553–575; “Farewell to Substance: A Diferentiated Leave-Taking,” Ratio 11
(1998): 235–252.
23. Justin Broackes, “Substance,” Proceedings of the Aristotelian Society 106
(2006): 131–166.
24. Brian Ellis, Scientifc Essentialism (Cambridge: Cambridge University Press,
2001).
25. Evan Fales, Causation and Universals (New York, NY: Routledge, 1990)
26. Joshua Hofman, “Neo-Aristotelianism and Substance,” in: Contemporary
Aristotelian Metaphysics, ed. Tuomas Tahko (Cambridge: Cambridge Uni-
versity Press, 2012), 140–155.
27. Michael Loux, “Kinds and the Dilemma of Individuation,” Review of Meta-
physics 27 (1974): 773–784; Substance and Attribute: A Study in Ontology
(Dordrecht: Reidel, 1978); Metaphysics: A Contemporary Introduction, 2nd
ed. (New York, NY: Routledge, 2002).
286 Travis Dumsday
28. E.J. Lowe, “Primitive Substances,” Philosophy and Phenomenological
Research 54 (1994): 531–552; “Form Without Matter,” Ratio 11 (1998):
214–234; The Four-Category Ontology: A Metaphysical Foundation for Nat-
ural Science (Oxford: Oxford University Press, 2006); “A Neo-Aristotelian
Substance Ontology: Neither Relational Nor Constituent,” in: Contemporary
Aristotelian Metaphysics, ed. Tuomas Tahko (Cambridge: Cambridge Univer-
sity Press, 2012), 229–248; “Neo-Aristotelian Metaphysics: A Brief Exposi-
tion and Defense,” in: Aristotle on Method and Metaphysics, ed. Edward
Feser (New York, NY: Palgrave Macmillan, 2013), 196–205.
29. Cynthia Macdonald, Varieties of Things: Foundations of Contemporary
Metaphysics (Oxford: Blackwell, 2005).
30. David Wiggins, Sameness and Substance Renewed (Cambridge: Cambridge
31. John Heil, From an Ontological Point of View (Oxford: Oxford University
Press, 2003); The Universe as We Find It (Oxford: Oxford University Press,
2012).
32. We will encounter other historical versions in the next section. Also, it is
worth noting that there has been a good deal of controversy over whether
or to what extent the distinctive commitments of Thomistic hylomorphism
accord with the hylomorphism outlined in Aristotle’s own texts. For a recent
defence of the claim that they difer considerably, see Christopher Byrne, Aris-
totle’s Science of Matter and Motion (Toronto, ON: University of Toronto
Press, 2018).
33. Joseph Bobik, “St. Thomas on the Individuation of Bodily Substances,” in:
Readings in the Philosophy of Nature, ed. Henry Koren (Westminster, MD:
The Newman Press, 1961), 327–340; Aquinas on Being and Essence: A
Translation and Interpretation (Notre Dame, IN: University of Notre Dame
Press, 1965); Aquinas on Matter and Form and the Elements: A Translation
and Interpretation of the De Principiis Naturae and the De Mixtione Elemen-
torum of St. Thomas Aquinas (Notre Dame, IN: University of Notre Dame
Press, 1998).
34. Jefrey Brower, Aquinas’s Ontology of the Material World: Change, Hylomor-
phism, and Material Objects (Oxford: Oxford University Press, 2014).
35. William Norris Clarke S.J., The One and the Many: A Contemporary Thomis-
tic Metaphysics (Notre Dame, IN: University of Notre Dame Press, 2001).
36. Edward Feser, Scholastic Metaphysics: A Contemporary Introduction (Heu-
senstamm: Editiones Scholasticae, 2014).
37. John Goyette, “Substantial Form and the Recovery of an Aristotelian Natural
Science,” Thomist 66 (2002): 519–533.
38. James Madden, Mind, Matter, and Nature: A Thomistic Proposal for the
Philosophy of Mind (Washington, DC: Catholic University of America Press,
2013).
39. David Oderberg, “Hylomorphism and Individuation,” in: Mind, Metaphys-
ics, and Value in the Thomistic and Analytic Traditions, ed. John Haldane
(Notre Dame, IN: University of Notre Dame Press, 2002), 125–142; “Hyle-
morphic Dualism,” Social Philosophy & Policy (2005): 70–99; Real Essen-
tialism (New York, NY: Routledge, 2007).
40. Patrick Toner, “Hylemorphic Animalism,” Philosophical Studies 155 (2011):
65–81.
41. Confusingly, Michael Rea defends a version of bundle theory, which he opts
to label ‘hylomorphism.’ See his “Hylomorphism Reconditioned,” Philo-
sophical Perspectives 25 (2011): 341–358.
42. Because most hylomorphists writing today are advocates of Thomistic hylo-
morphism, the picture in their case is arguably more complicated, given
Aquinas’s own pluralism. Yet the issue of pluralism rarely comes up in the
contemporary discussion of hylomorphism, precisely because the ontology
of angels isn’t on the radar in the contemporary debates. More on this in the
next section.
43. Travis Dumsday, “How to Be a Pluralist in Substance Ontology,” Erkenntnis
85 (2020): 995–1022, with quote at 999.
44. Frederick Copleston, S.J., A History of Philosophy Volume I: Greece & Rome
– Part II (New York, NY: Image Books, 1962), 57–58.
45. That is the case not just for contemporary Aristotle scholarship, but also for
the ancient commentators. For a sketch of some of the early disputes sur-
rounding the Intelligences and their explanatory role in planetary motion see
Miira Tuominen, The Ancient Commentators on Plato and Aristotle (Berke-
ley, CA: University of California Press, 2009), 150–155.
46. Copleston, A History of Philosophy Volume I, 58.
47. Consider for example the treatments of the issue by Vladimir Lossky, Ortho-
dox Theology: An Introduction, trans. Ian and Ihita Kesarcodi-Watson
(Crestwood, NY: St. Vladimir’s Seminary Press, 1978), 81; Michael Poma-
zansky, Orthodox Dogmatic Theology, 3rd edition, trans. Seraphim Rose
(Platina, CA: St. Herman of Alaska Brotherhood, 2005), 116–117; and Sera-
phim Rose, The Soul After Death, 4th edition (Platina, CA: St. Herman of
Alaska Brotherhood, 2004), 22–31.
48. Quoted in R. James Long, “Roger Bacon on the Nature and Place of Angels,”
Vivarium 35 (1997): 266–282, with quote at 266. It’s perhaps worth not-
ing that the early Protestant denominations largely took over this Catholic
understanding of angelic ontology without alteration, though over time a
greater diversity of views began to appear. So far as I am aware, there is as yet
no in-depth scholarly history of Protestant teachings on angelic corporeality.
For a recent Evangelical defence of angelic corporeality, see Austin Freeman,
“Celestial Spheres: Angelic Bodies and Hyperspace,” TheoLogica: An Inter-
national Journal for Philosophy of Religion and Philosophical Theology 2.2
(2018): 168–186.
49. Long goes so far as to say of this period that the “metaphysical testing ground
par excellence for the compatibility of these two world views [Aristotelian
and Christian] was the doctrine of the angels, spiritual substances that were
at the same time limited and mutable.” See R. James Long, “Of Angels and
Pinheads: The Contributions of the Early Oxford Masters to the Doctrine of
Spiritual Matter,” Franciscan Studies 56 (1998): 239–254, with quote at 239.
50. On the origins of universal hylomorphism in Avicebron and a delineation
of the lines of infuence he exercised on the mediaeval Scholastics, see James
Weisheipl, “Albertus Magnus and Universal Hylomorphism: Avicebron – A
Note on Thirteenth-Century Augustinianism,” Southwestern Journal of Phi-
losophy 10 (1979): 239–260.
51. Long observes that Bacon is even more generous, positing three irreducibly
distinct matters: “[S]piritual matter, as for example the matter of the intel-
ligences; intermediate matter (materia media), as the matter of the heavenly
bodies; and fnally corporeal and sensible matter, as that of generable and
corruptible substances.” Long, “Roger Bacon,” 268.
52. Aquinas’s extreme pluralism could arguably be viewed as embracing bundle
theory as well, though this third substance ontology would not obtain natu-
rally, but rather only by way of miraculous divine intervention – specifcally,
with the independently existent accidents of the consecrated Eucharistic ele-
ments. It is unclear how best to classify this idea, however; Aquinas himself
certainly did not label an independently existent accident a ‘substance,’ and so
presumably would not consider acceptance of the existence of such a thing as
288 Travis Dumsday
entailing a commitment to an additional sort of substance ontology. But then
some contemporary bundle theorists would, in a way, agree – namely, those
who wish to drop ‘substance’ altogether as a fundamental ontological category
and who thus commit to the claim that an independently existent accident
is real but is not strictly a substance. Such bundle theorists are still viewed
as presenting an option within the debate over ‘substance ontology’ broadly
construed (even if a reductionist or eliminativist option). These are deep waters
and I will say no more, except to note that the belief in independently existent
accidents within the supernatural context of the Eucharistic transformation
becomes a generally agreed-upon doctrine throughout the otherwise compet-
ing Scholastic schools, so if its acceptance sufces to render someone a pluralist
in substance ontology then nearly all Scholastics from the early thirteenth
century onward will have to count as pluralists. For an overview of some of
the relevant history here, with an emphasis upon the interplay of Eucharistic
doctrine with Aristotelian philosophy, see Marilyn McCord Adams, “Aristotle
and the Sacrament of the Altar: A Crisis in Medieval Aristotelianism,” in: Aris-
totle and His Medieval Interpreters, eds. Richard Bosley and Martin Tweedale
(Calgary, AB: University of Calgary Press, 1992), 195–249.
53. Another interesting version of this third approach, quite diferent from Scotus’s,
is the account of angelic individuation pursued by Henry of Ghent. For a sum-
mary of his model (and Scotus’s critique of it) see Giorgio Pini, “The Individua-
tion of Angels From Bonaventure to Duns Scotus,” in: A Companion to Angels
in Medieval Philosophy, ed. Tobias Hofmann (Leiden: Brill), 79–115.
54. A fact noted by David Keck, Angels and Angelology in the Middle Ages
(Oxford: Oxford University Press, 1998), 99.
55. Innovative at least in the context of western Scholasticism – arguably the
same basic idea is present in earlier fgures, notably including Maximus the
Confessor, via his understanding of the creaturely logoi (which seem to func-
tion much like individual essences). On Maximus’s system see Torstein Tollef-
sen, The Christocentric Cosmology of St. Maximus the Confessor (Oxford:
56. Alexander of Hales, Summa Theologica, Tomus II, Inq. II, Tract. II, Quaest. I,
Art. II (Quaracchi: Collegii S. Bonaventurae, 1928), 156. My own translation;
the Latin reads as follows:
In spiritibus enim pluralitas individuorum non est propter hanc causam,
sed est ad manifestationem divinae bonitatis, quae vult communicare plu-
ribus participationem beatitudinis: ut sicut multiplicatur numerus indi-
viduorum sub unaquaque specie ad fruitionem beatitudinis. Praeterea,
posse regis temporalis et decens dispositio in aula eius ostenditur, cum
multi consimilis dispositionis serviunt, magis quam si unus in uno ofcio
serviret: ut si multi pincernae quam si unus solus pincerna, et ita de aliis
ofciis. Cum ergo posse regis caelestis sit improportionale in infnitum
super posse regis temporalis et decentior dispositio aulae caelestis quam
terrenae, conveniens est angelos, qui sunt summi ministeriales, plures
esse sub una specie.
57. Long, “Roger Bacon,” 268–269.
58. Long, “Of Angels and Pinheads,” 242–243.
59. This is one factor that serves to distinguish Bonaventurean hylomorphism
from the Thomistic hylomorphism summarized in the previous section:
Thomistic hylomorphism grants matter an essential role in individuation but
also grants such a role to the accident of determinate spatial extension, such
that individuation is strictly accomplished not by prime matter working alone
but by so-called signate matter (prime matter + determinate quantity); by
contrast, for Bonaventure matter plays the individuating role (upon conjunc-
tion with form) with no assistance from accidents.
60. Christopher Cullen, S.J., Bonaventure (Oxford: Oxford University Press,
2006), 46.
61. Pini, “The Individuation of Angels,” 83.
62. For more on this aspect of Bonaventure’s reasoning, see Keck, Angels and
Angelology, 95–98, and also Cullen, Bonaventure, 44–45. This argument
from change is not part of Bacon’s case for universal hylomorphism, given
his unusual view that angels are incapable of change (though their spiritual
matter does render them naturally subject to destruction by God). See Long,
“Roger Bacon,” 268–271 and 275–278.
63. For example, if we assume that mass is to be understood as a single disposi-
tional property (a causal power), how are we to understand the distinction
between inertial mass and gravitational mass? Is there really a single power
here, ‘mass,’ with radically diferent sorts of efects? Or are there actually two
distinct powers here, which distinct powers are nevertheless contingently com-
present? Some argue that mass is a single property (a multi-track disposition)
while others argue that we are dealing here with two distinct properties. Alex-
ander Bird for instance opts for the second option in his Nature’s Metaphysics:
Laws and Properties (Oxford: Oxford University Press, 2007), 211–212.
64. Thomas Aquinas, On the Truth of the Catholic Faith: Summa Contra Gen-
tiles Book II, trans. James F. Anderson (New York, NY: Image Books, 1956),
315. His quote from pseudo-Dionysius is from the fourth book of On the
Divine Names.
65. A nice reminder that mediaeval Christians were not as anthropocentric as
many have claimed.
66. Aquinas, On the Truth, 321.
67. Aquinas, On the Truth, 321.
68. Helen Lang, Aristotle’s Physics and its Medieval Varieties (Albany, NY:
SUNY Press, 1992), 173.
69. Such at least is the claim of Todd Bates, Duns Scotus and the Problem of
Universals (London: Continuum, 2010), 122–125.
70. For a concise summary see again Bates, Duns Scotus, 87–104.
71. I remind the reader that the Thomist understanding of Aristotle on individu-
ation (and much else) remains controversial within Aristotle scholarship. See
again the valuable recent study by Christopher Byrne, Aristotle’s Science of
Matter and Motion (Toronto, ON: University of Toronto Press, 2018).
72. For Scholastic readers: recall my earlier clarifcation regarding ‘universal’
versus ‘common nature.’ In Scholastic terms, the Scotist is worried about
what contracts the common nature.
73. This is in marked contrast to the standard Thomist line of thinking about
higher-level generic forms (or ‘determinable universals’ in analytic parlance),
all of which are viewed by the Thomist merely as products of mental abstrac-
tion rather than as irreducibly real ontological components of creatures.
74. For a more thorough treatment of the Scotist account of individuation, see
again Todd Bates, Duns Scotus and the Problem of Universals (London:
Continuum, 2010). For a briefer but still accessible discussion see Timothy
Noone, “Individuation in Scotus,” American Catholic Philosophical Quar-
terly 69 (1995): 527–542.
75. Travis Dumsday, “Alexander of Hales on Angelic Corporeality,” Heythrop
Journal 54 (2013): 360–370.
76. Christopher Toner, “Angelic Sin in Aquinas and Scotus and the Genesis
of Some Central Objections to Contemporary Virtue Ethics,” Thomist 69
(2005): 79–125.
290 Travis Dumsday
77. Lang, Aristotle’s Physics, 173–187.
78. Travis Dumsday, “Natural Evil, Evolution, and Scholastic Accounts of the
Limits on Demonic Power,” Pro Ecclesia: A Journal of Catholic and Evangeli-
cal Theology 24 (2015): 71–84.
79. Shandon Guthrie, Gods of This World: A Philosophical Discussion and
Defense of Christian Demonology (Eugene, OR: Pickwick, 2018), 264–267.
Part 3
God and Nature

11 Grounding and
Participation in God
Ross D. Inman
This chapter aims to explore the intersection of Christian theism, a neo-

Aristotelian gloss on metaphysical grounding, and creaturely participa-
tion in God. In Section 1, I aim to develop several core tenets at the heart
of a theistic participatory ontology as it is found in the Christian tradi-
tion, what I call minimal participatory ontology. In Section 2, I examine
the contemporary notion of metaphysical grounding, namely the formal
and structure features of the grounding relation, and ofer a grounding-
theoretic framework for understanding a minimal participatory ontology.
Finally, in Section 3, I put forward a neo-Aristotelian account of meta-
physical grounding in particular, one that is uniquely suited to capture the
central tenets of minimal participatory ontology.
1. Minimal Participatory Ontology in the Christian

Tradition
At its most basic level, a Christian participatory ontology pertains to the
notion that all creaturely being is received being; every aspect of non-God
reality fnds its ultimate source, foundation, and proper end in the pleni-
tudinous life of the Triune God (Rom. 11:36). That individual creatures
exist at all and are what they are is, in some way or other, metaphysi-
cally derivative on God. Of course, a fully orbed doctrine of creaturely
participation in God will include much more besides the mere creaturely
reception of being and essence from the Triune Creator. A fully developed
Christian doctrine of creaturely participation will identify and develop the
various ways in which the Triune God is the distinctively (efcient) causal
source of all creaturely reality, as well as participatory themes throughout
the wider theological and philosophical landscape with respect to value
(natural law, virtue, evil, beauty), theological anthropology (imago dei),
Christology, and soteriology (atonement, consummation, salvifc and
sanctifying grace, etc.), to name a few.1
Nevertheless, getting clear on the precise ontological commitments of
what I will call a “minimal participatory ontology” is vitally important
for at least two reasons. First, it is arguably the case that more fne-grained
DOI: 10.4324/9781003125860-15
294 Ross D. Inman
notions of soteriological participation in God (e.g., deifcation, union with
or participation in Christ) are predicated upon a minimal participatory
ontology, the view that creatures derive their very existence and natures
from God. While soteriological participation in God cannot plausibly be
reduced to minimal participatory ontology, it is arguably founded upon
it. Thus, before one builds upon a minimal participatory ontology across
the various theological loci, it is important to frst get clear on the minimal
metaphysical commitments of creaturely participation in God.
Second, among friends of creaturely participation in God, there has
been a call for “greater precision” regarding the metaphysics of participa-
tion.2 In Section 2, I aim to heed these calls for “greater precision” regard-
ing participatory metaphysics by employing a neo-Aristotelian conception
of metaphysical grounding at play in contemporary analytic metaphysics.
While much work remains to be done concerning the various dimensions
of a robust, theistic participatory ontology, my hope is that a contem-
porary neo-Aristotelian framework can help elucidate and shore up the
philosophical foundations of such an idea.
What, exactly, is the minimal metaphysical content of a participa-
tory ontology at work in the classical Christian tradition? Towards this
aim, it may be helpful to briefy rehearse what such an ontology is not,
according to classical Christian theism. First, creatures do not partici-
pate in God by way of parthood, by becoming proper parts of the divine
nature.3 From the perspective of classical Christian theism, there is one
rather obvious reason why creatures do not participate in God by way
of becoming proper parts of the divine nature. Classically understood,
the divine nature is absolutely indivisible, and thus metaphysically and
mereologically simple; the divine nature is not distributed or divided out
into separable ontological bits. Thus, divine simplicity rules out what has
been called “mereological panentheism,” the view that the created order
is “in” God by way of being a proper part of God (or even contained in
God in a more general sense).4
Second, one might appeal to divine aseity or ultimacy in order to pre-
clude a second sense in which creatures participate in God, by God’s
being a proper part of creatures. Just as I might participate in the activity
of hand-clenching by way of the activity of one of my proper parts (my
hand), so too I might participate in God by way of the activity of one of
my proper parts, God. Against this, Thomas Aquinas (2012: 35) argued
(quoting Dionysius), that “‘There can be no touching Him’, i.e. God,
‘nor any other union with Him by mingling part with part.’” In other
words, God in no way enters into the composition of any creaturely being,
whether as form, matter, or any other type of metaphysical constituent. To
do so would, on Aquinas’s view, compromise the absolute metaphysical
primacy of God as the prima causa of all creaturely reality.5
With these two mereological notions of creaturely participation in
God set aside, what positive content is there to a minimal participatory
Grounding and Participation in God 295
ontology according to the classical Christian tradition? For one, crea-
turely participation in God minimally involves the view that creatures
depend for their very existence on God; all non-God reality is in virtue of
the existence and activity of the Triune God. This minimal claim of crea-
turely existential dependence is richly attested throughout the span of the
Christian tradition, East and West alike. Here, a representative sampling
from the Christian tradition will sufce. John of Damascus (1958: 202),
for instance, underscores this aspect of creaturely participation in God
when he says, “For, toward Him all things tend, and in Him they have
their existence, and to all things He communicates their being in accor-
dance with the nature of each.”6 Moreover, in his Monologion 3, Anselm
(2007: 9) argues for the existence of a supreme nature through which all
created things exist. He notes, “[T]here must be one thing through which
all existing things exist. Therefore, since all existing things exist through
that one thing, undoubtedly that one thing exists through himself. So
all other existing things exist through another; he alone exists through
himself.” For Anselm, it is characteristic of all creatures that they exist
through another, in this case, through God as the sole, ultimate reality
that exists through Himself.
Not only do all creatures exist through the one supreme nature,
Anselm (2007: 40) further emphasizes in Monologion 31 the radically
qualifed and attenuated nature of creaturely existence by saying, “the
truth of existing is understood to be in the Word, whose essence exists so
supremely that in a certain sense it alone exists, whereas in the things that
by comparison with him in a certain sense do not exist.” In a similar man-
ner, the Angelic Doctor (Aquinas 2012: 451–452) underscores in Summa
Theologiae I.44.1 the derivative nature of creaturely existence as follows:
It must be said that every being in any way existing is from God.
For whatever is found in anything by participation, must be caused
in it by that to which it belongs essentially, as iron becomes ignited
by fre . . . God is the essentially self-subsisting Being . . . Therefore
all beings apart from God are not their own being, but are beings by
participation.
In short, the existence of creatures is wholly metaphysically derivative;

God, by contrast, is wholly metaphysically underived or fundamental in
His Being. We can formulate this frst tenet of a minimal participatory
ontology simply as follows:
(CED) Creaturely Existential Dependence: creatures metaphysically

depend on God for their existence.
Ultimately, according to CED, the existence of all creaturely being is

continuously and graciously received from the hand of God (Rev. 4:11);
296 Ross D. Inman
creaturely being per se is being by participation in or through the supreme
being.
While CED captures a crucial dimension of creaturely participation in
God, it does not exhaust the distinctive ways that creatures metaphysi-
cally participate in God according to the Christian tradition. Tradition-
ally, creatures not only derive their existence from God, they also derive
what they are, their natures, from God. According to minimal partici-
patory ontology, both the existence and content of creaturely natures
are metaphysically explained, in some fashion, by God. Each and every
dimension of created reality is derivative on God in that the rich plenitude
of the divine nature itself provides the intelligible content of all creaturely
natures.
This additional dimension of creaturely participation as it pertains
to the content of creaturely natures in particular is also amply attested
in the Christian tradition. Divine exemplarism, the idea that the divine
nature itself is the exemplar or pattern after which all creaturely natures
are produced, has a longstanding precedent in the Christian tradition.7
Again, consider a few representative examples.8 According to Origen
(1973: I.2–3), in Christ, the Wisdom and eternally begotten Word of God,
“there was implicit every capacity and form of the creation that was to
be . . . she [Wisdom] fashions beforehand and contains within herself the
species and causes of the entire creation.”9 Likewise for Augustine (1982:
80–81), all creaturely natures are “created in accord with reasons unique
to them,” reasons that “must be thought to exist nowhere but in the very
mind of the Creator.” Indeed, Augustine goes on to underscore, “it is by
participation in these that whatever is exists in whatever manner it does.”
As Andrew Davison (2019: 97–98) points out concerning Augustine’s
exemplarism, “the origin and archetypes of human beings, or of trees or
beds, or of goodness or justice, do not lie in some eternal and free-foating
forms of human beings, trees or beds, goodness or justice. According to
Augustine, their origin is in God.”10
Additionally, Anselm (2007: 17), in underscoring the rationale for
divine exemplarism with respect to its connection to creation notes, “there
is no way anyone could make something rationally unless something like a
pattern (or, to put it more suitably, a form of likeness or rule) of the thing
to be made already existed in the reason of the maker.” Indeed, Anselm
(2007: 40, 128) says elsewhere that when it comes to created things that
“have been made something through him and in accordance with him,
there is judged to be an imitation of that supreme essence” and “what-
ever is, truly is, insofar as it is what it is in the supreme truth.” Thomas
Aquinas (2012: 180) also echoes this rich divine exemplarist tradition by
claiming,
In the same way natural things are said to be true in so far as they
express the likeness of the species that are in the divine mind. For
a stone is called true, which possesses the nature proper to a stone,
according to the preconception in the divine intellect.
While, for Aquinas, particular creaturely natures are expressions of a

mere likeness of the mode of existence those natures have in the divine
nature, such natures truly imitate or resemble the rich, qualitative pleni-
tude of the divine nature itself. Lastly, one can fnd clear articulations
of divine exemplarism in a host of sixteenth- and seventeenth-century
Protestant divines.11 Consider one such Protestant divine, William Ames
(1968: 95):
The idea of all things is the divine essence, meaning that essence
understood by God himself and imitable by his creatures – at least
the image or vestige of that perfection may be expressed in some way
in creatures. That is, the creatures themselves, so far as they are con-
ceived in the divine mind, are the idea or image of that nature which
they have in themselves . . . therefore all things are frst in his mind
before they are in themselves. In us the things themselves are the pat-
tern and our knowledge is the image of them. But in God the divine
knowledge is the pattern and the things themselves are the image or
express likeness of it.
Consequently, given the centrality of the notion that God is the exemplar
or pattern for all creaturely natures in the Christian tradition, we can add
the following additional tenet to minimal participatory ontology:
(CND) Creaturely Nature Dependence: creatures metaphysically

depend on God for their natures.
With respect to both CED and CND, then, God is the ultimate metaphysi-
cal fount of created being and essence; all chains of creaturely existential
and essential dependence ultimately terminate in the Triune God.
A minimal participatory ontology consisting of CED and CND argu-
ably gives rise to the following additional tenet of minimal participatory
ontology regarding God’s metaphysical ultimacy and priority to all cre-
ated reality:
(DF) Divine Fundamentality: God is absolutely fundamental and

metaphysically prior to all created reality.
To say that God is absolutely fundamental in this sense is to say that God
is the ultimate ontological terminus for all chains of creaturely existential
and essential dependence. Ultimately, both the existence of the creaturely
order itself and its diverse qualitative essential structure obtain in virtue
of the nature and activity of God. As such, the kind of metaphysical
298 Ross D. Inman
dependence relation operative in CED, CND, and DF is asymmetric in
that it carves out a distinctive metaphysical ordering relation between
creatures and God; God alone exists and is what God is in an unpartici-
pated manner, we might say. As metaphysically fundamental in every way,
God is metaphysically prior to creatures; as metaphysically derivative in
every way, creatures are metaphysically posterior to God.
Lastly, it is important to identify several theistic explanatory theses that
follow quite naturally from the conjunction of CED, CND, and DF, theses
that could be included either as an essential tenet of minimal participa-
tory ontology or perhaps as a natural corollary to it.12 Note that all three
tenets of a minimal participatory ontology (CED, CND, and DF) carve
out the necessary metaphysical structure needed to back metaphysical
explanations concerning why creatures ultimately exist and are what they
are, explanations that ultimately terminate in facts about the existence,
nature, and activity of God.13
Intuitively, metaphysical explanation holds between at least two propo-
sitions or facts and is thought to be a noncausal variety of explanation;
arguably, not all ontological explanatory structure in reality is carved out
by (efcient) causal relations. Some fact F, say the apple’s exemplifying a
redness trope or the universal redness, metaphysically explains some fact
G, say the apple’s being red, only if G obtains because of F and not vice
versa, where the sense of “because” here is commonly believed to track
something deeper than an (efcient) causal explanatory relation (though
it may share similarities with such a notion).14
If all creaturely dimensions of reality are ontologically dependent on
God along the lines of CED, CND, and DF, it is a natural step to think
that all metaphysical explanations concerning facts about creaturely exis-
tence and the content of creaturely natures also terminate in theistic facts
about God. Call an “existence-fact” ([x]) a fact pertaining to the existence
of a creature x, and an “essential-fact” ([□x]) a fact concerning the nature
or essence of x. We have, then, the following expansive, theistic explana-
tory thesis:
(TEB) Theistic Explanatory Breadth: Necessarily, for any creature x,

and for any [x] and [□x] that obtains, there is some theistic fact (or set
of facts) [G] that obtains such that the obtaining of [G] metaphysi-
cally explains the obtaining of [x] and the obtaining of [□x].
As a thesis about the order of metaphysical explanation between facts

about God and facts about creaturely existence and creaturely natures,
TEB fts quite naturally within a minimal participatory ontology that
includes objective, metaphysical ordering relations between God and
creatures along the lines of CED, CND, and DF. If asymmetric relations
of metaphysical dependence hold between creatures and God, and God is
metaphysically prior to creatures (and not vice versa), then such structure
plausibly supports metaphysical explanatory relations between God and
creatures as well; facts concerning the existence and natures of creatures
are explained in terms of facts about God, whether facts about His nature
or activity (or some combination thereof). For instance, the fact that tigers
and tulips exist and have the biological natures they do (including the
diverse, qualitative content of those natures) is explained by some fact or
facts about God.
Since all creaturely beings and natures are derivative on God according
to CED and CND, theistic facts about God’s nature and activity serve to
metaphysically explain the existence and qualitative nature of each crea-
ture. We might even go so far as to say that on a robust theistic participa-
tory ontology, perhaps all “explanation-eligible”15 facts about creatures
are metaphysically explained by facts about God; there are no non-divine
facts that fall outside of the explanatory scope of the existence, nature,
and activity of the Triune God (Rom. 11:36).
To be clear, TEB is a thesis about explanatory scope, not explanatory
completeness.16 The aforementioned claim at the heart of TEB is that
among the explanatory grounds for why some creaturely fact obtains will
be some theistic fact or set of theistic facts. As far as I can tell, TEB in
no way rules out the additional claim that some creaturely facts are also
explained by other creaturely facts. For example, one might think that
among the explanatory grounds for the obtaining of facts concerning free
creaturely actions are also facts about the existence and activity of free
agents.17 Thus, the theistic explanatory thesis in question does not entail
that facts about God are the sole explanatory grounds for each and every
creaturely fact.18
We can supplement TEB, as a thesis concerning explanatory scope, with
a thesis of theistic explanatory immediacy that also fts well within a mini-
mal participatory ontology: the way in which theistic facts metaphysically
explain all existence-facts and essential-facts concerning creatures is in
no way mediated. More specifcally, we can supplement TEB to generate
the following:
(TEI) Theistic Explanatory Immediacy: Necessarily, for any creature

x, and for any [x] and [□x] that obtains, there is some theistic fact (or
set of facts) [G] that obtains such that the obtaining of [G] metaphysi-
cally explains the obtaining of [x] and the obtaining of [□x], and there
is no creaturely fact or set of facts [S] (not including [x] or [□x]) that
obtains such that God metaphysically explains [x] or [□x] by way of
metaphysically explaining the obtaining of [S].
On TEI, for each and every existence and essential fact about creatures
that obtains, God is immediately explanatorily relevant to its obtaining.
TEI helps ground the notion that God plays an up-front, non-instru-
mental ontological role in explaining facts concerning the existence and
300 Ross D. Inman
qualitative natures of creatures. Kathryn Tanner (1988: 84), in the spirit
of both TEB and TEI, puts this emphatically as follows, “[E]verything
non-divine must be talked about as existing in a relation of total and
immediate dependence upon God.”19
2. Metaphysical Grounding and Minimal Participatory

Ontology
We have thus far seen how a minimal participatory ontology can be stated
in terms of relations of existential and essential metaphysical dependence
(CED, CND, and DF), as well as two metaphysical explanatory theses
backed by these relations (TEB and TEI). I now want to aim to show how
a contemporary, neo-Aristotelian account of the nature of metaphysical
grounding in particular can help support each of these ontological claims
at the heart of a minimal participatory ontology.
In recent years, the notion of metaphysical grounding has received
renewed attention by analytic metaphysicians and considered by many
to be a helpful piece of metaphysical machinery. It has been argued that
an appeal to grounding as a form of noncausal, metaphysical dependence
improves our understanding of truthmaking (Schafer 2009; Lowe 2009;
Inman 2012), physicalism (Schafer 2009; Loewer 2001: 39), intrinsicality
(Witmer et al. 2005), objective similarity (Sider 2012), perfectly natural
properties (Schafer 2004), the nature of noncausal explanation (Audi
2012), trope inherence (Lowe 2006), and an overall realist approach to
metaphysics (Fine 2001; Schafer 2009).
As a result, grounding has been hailed as a unifed and theoretically
fruitful notion that undergirds a variety of concepts in metaphysics. As the
literature on metaphysical grounding is expansive, ever expanding, and
fraught with disagreement, it is beyond the scope of this chapter to argue
for the specifc account of metaphysical grounding I ofer later. Rather,
my aim is to unpack a neo-Aristotelian view of grounding that can be
put to use in illuminating the minimal participatory ontology outlined
in Section 1.
2.1. The Nature of Metaphysical Grounding

In the broadest sense, I am inclined to think that the relation of “meta-
physical grounding” picks out a general type of structure-making relation,
one that serves to carve out the distinctively metaphysical structure of
reality. Along these lines, I adopt a version of what is known as ground-
ing monism, the view that there is a single, fundamental grounding rela-
tion.20 Following Bradley Rettler (2017: 13–14), this single, fundamental
grounding relation, call it Grounding for short, is essentially defned by
its structure-making role of carving out the distinctively metaphysical
(non-causal) structure of reality. According to Rettler (2017: 13–14),
Grounding “relates the fundamental to the non-fundamental, relates the
relatively more fundamental to the relatively less fundamental, lays out
the structure of the world, says which things depend on which other
things, explains why something exists, and explains why something has
a property.”21
As such, Grounding is a specifcation relation that admits of various spe-
cies (or determinates); there are plausibly many ways of being Grounded.
Relations such as ontological dependence, truthmaking, and metaphysical
explanation are all plausibly considered to be structure-making relations
that carve out the metaphysical structure of reality. What unifes each of
the previous relations and makes them all species of Grounding is that
each is characterized by its role of carving out metaphysical structure in
some way. As Rettler (2017: 12) puts it,
The essence of truthmaking is that it’s the grounding of a thing’s

truth in an object. The essence of metaphysical explanation is that it’s
the grounding of a thing’s truth in other, usually more fundamental,
truths. The essence of ontological dependence is that it’s the ground-
ing of a thing’s existence in another thing’s existence.
Consequently, “grounding” is multivocal in that it picks out diferent spe-

cies of Grounding in diferent contexts, whether ontological dependence,
truthmaking, or metaphysical explanation.
If “grounding” is multivocal in this way, it is important to specify
which particular species of Grounding one has in view when it comes
to explicating a minimal participatory ontology in grounding-theoretic
terms. As we have seen in Section 1, CED and CND together carve out
several deep metaphysical dependence relations that hold between crea-
tures and God. Consequently, as a particular species of Grounding that
holds between entities in particular, the relation of ontological dependence
seems to nicely ft the bill when it comes to explicating CED and CND
within a contemporary grounding-theoretic framework.
Moreover, in Section 1 we outlined two explanatory theses, TEB and
TEI, that are founded upon CED and CND and are perhaps corollaries
of a minimal participatory ontology. Since TEB and TEI are distinctively
explanatory theses about creaturely facts (truths or states of afairs) and
their relation to facts about God, it is natural to think that the species of
Grounding involved here is that of metaphysical explanation. Accord-
ingly, the created order is metaphysically grounded in God along the lines
of TEB and TEI in the sense that facts about the existential and essential
structure of the created order immediately obtain in virtue of theistic facts,
and not vice versa. Along these lines, God is explanatorily ultimate in that
the class of absolutely fundamental facts or truths are coextensive with
theistic facts; there are no non-theistic facts that are absolutely, meta-
physically fundamental or ungrounded.22
302 Ross D. Inman
By my lights, a comprehensive account of a minimal participatory
ontology in grounding-theoretic terms should invoke both species of
Grounding, ontological dependence and metaphysical explanation. But,
for my limited purposes here, I’ll aim to unpack a minimal participatory
ontology within a grounding framework exclusively in terms of ontologi-
cal dependence. Thus, when I speak of “metaphysical grounding” in what
follows, I mean to pick out the relation of ontological dependence whose
relata are entities, broadly construed.
2.2. The Structure of Metaphysical Grounding

According to our minimal participatory ontology outlined in Section 1,
creaturely beings exist (CED) and are what they are (CND) ultimately in
virtue of the existence, nature, and/or the activity of the Triune God.23
That creatures exist at all, which creatures exist, and what individual
creatures are fundamentally is metaphysically grounded in God. More-
over, creaturely beings and the content of their creaturely natures are
metaphysical grounded in God, and not vice versa; God is the ultimate
metaphysical ground of all non-God reality, according to DF. In ofering
a grounding-theoretic framework as a helpful guide to unpacking a mini-
mal participatory ontology, it is absolutely vital to identify a particular
species of metaphysical grounding that supports this crucial ontological
asymmetry between God and creatures.24 As we will see in Section 3, not
all varieties of this particular species of grounding (again, understood
in terms of ontological dependence) support the needed asymmetry for
creaturely participation in God.
What might we say about the general formal structure of the type of
metaphysical grounding needed to support a minimal participatory ontol-
ogy along the lines of CED, CND, and DF? To start, I take the domain of
entities that can be relata of the grounding relation in this context to be
maximally general; entities of any ontological category are able to stand
in metaphysical grounding relations:
G1: Topic Neutrality: entities of any ontological category can serve

as the relata of grounding.
Intuitively, created beings of diverse ontological categories such as sub-

stances (living organisms), events (baseball games), kinds (horseness),
modes (redness), and relations (taller than) are potential relata of the
grounding relation.
Moreover, in light of the topic-neutrality of grounding as per G1, the
logical form of grounding in this context is best expressed by a two-place
predicate “x  y” (which stands for “x is grounded in y”):
x  y = x is grounded in y
As noted previously, grounding is thought to be a type of structure-mak-
ing relation that carves out (in various ways) the metaphysical structure of
reality. Thus, it is plausibly construed as a metaphysical ordering relation
that generates a strict partial order over a domain of entities. More precisely,
G2. Irrefexive: ¬(x  x)

G3. Asymmetric: (y  x) →¬(x  y)
G4. Transitive: (y  x ∧ z  y) → (z  x)
G2–G4 are formal features of metaphysical grounding that are vital for
preserving the core tenets of minimal participatory ontology, in particular,
God’s ultimate metaphysical priority with respect to creaturely reality.
However, as we will see shortly, arguably not all varieties of this particular
species of grounding carve out strict partial ordering relations. Sufce it to
say at this point that in order for metaphysical grounding to help elucidate
a minimal participatory ontology it is important to preserve grounding as
a partial-ordering relation in terms of G2–G4.
Metaphysical grounding can also be either total or partial. To briefy illus-
trate the diference between total and partial grounding, consider a rather
common analysis of the nature of events in terms of the triple [o,P,t], where
“o” stands for some object or objects, “P” a property (whether monadic or
polyadic), and “t” a time. On this account of events, the event [o,P,t] exists
just in case o has P at t. The total grounds for the event of the collision of the
Titanic with the iceberg on April 14, 1912, for example, include the objects
of the Titanic and the iceberg, the dyadic relation colliding with, and the par-
ticular time of April 14, 1912. The event is totally grounded in the items that
are among its constituents in the sense that no additional creaturely being
needs to be added to these three items to fully explain what grounds the
event in question. Partial grounding, by contrast, is often defned in terms of
total grounding in that while some item, say o, does not fully ground [o,P,t]
on its own, o, together with other entities (in this case P and t) provide full
grounds for the triple [o,P,t]. In this way, as a partial ground, o contributes
to (but does not exhaust) the grounds of [o,P,t].
By my lights, a commitment to minimal participatory ontology along
the lines of CED, CND, and DF (together with the auxiliary explana-
tory theses of TEB and TEI) lends credence to the following claim about
God’s being at least a partial ground for the existence and nature of any
creaturely being:
(TPG) Theistic Partial Grounding: For any existing entity x that is

not identical to God, x is partially grounded in God.
TPG afrms that God is, at the very least, among the metaphysical grounds
for any being that is not-God. Of course, God, as universal primary cause,
will be the principal ground among the partial grounds of x.25
304 Ross D. Inman
Whether one opts for TPG or the much stronger thesis that God is the
total ground for the existence and nature of any creaturely being will
largely depend on one’s wider metaphysical and theological commitments.
For instance, if one goes for the stronger claim that every being that is
not identical to God is totally grounded in God, then this would seem to
preclude the view that any creature, understood in the broad sense as any
existing entity that is not identical to God, has another creature among
its total or partial grounds. Consequently, on this stronger view, there
would be no creature-to-creature grounding relations. This would seem
to preclude the idea that creaturely agents are at least among the partial
grounds for the existence and explanation of events brought about by free
creaturely acts of will. But if there are creature-to-creature grounding rela-
tions and creatures can serve as the partial grounds for other creatures (as
seems plausible), then TPG preserves this possibility while ascribing God
metaphysical pride of place as the chief partial ground of every creaturely
being; every creaturely being has, at bottom, a theistic ground.26
In addition to being total or partial, I consider grounding to be both an
existence entailing relation and one that holds of necessity. Taking “E”
as the existence predicate,27 grounding is:
G5. Existence Entailing: □(x  y → (Ex ∧ Ey))
as well as a relation that obtains of necessity (if it obtains at all):
G6. Necessity: (x  y →□(Ex → x  y))
G5 states that, necessarily, if x is grounded in y then both x and y exist. If

creatures are metaphysically grounded in God, then both God and crea-
tures exist. Regarding G6, if x is grounded in y, then it is necessarily the
case that if x exists, then x is grounded in y. G6 tracks the intuition that an
entity’s depending on another entity for its existence is a non-contingent
feature of that entity. The very concept of creatureliness, we might say,
essentially involves the notion of received or participated being, both the
existence and the natures of creatures are metaphysically derivative on
God. On G6, if the existence and nature of Fido the dog is metaphysically
grounded in God, then it is necessarily the case that if Fido exists, then
Fido is grounded in God in such a way.
Let us now turn to the question of the well-foundedness of ground-
ing, which has immediate bearing on our proposed grounding-account
of a minimal participatory ontology. We frst need to get clear on the
relationship between grounding and metaphysical fundamentality. For
our purposes here, it will be important to distinguish between the notions
of relative and absolute fundamentality. Something, x, is relatively fun-
damental to y just in case x grounds y, and not vice versa. In this case,
since x grounds y, x is more metaphysically fundamental than y; likewise,
since y is grounded in x, y is less metaphysically fundamental than x.
The notion of relative fundamentality carves out a hierarchically (asym-
metrically) ordered conception of reality where some beings are more
or less metaphysically fundamental than others. In general, the notion
of metaphysical fundamentality should march in step with the direction
of metaphysical grounding. With this notion of relative fundamentality
in hand, we can go on to defne absolute fundamentality such that x is
absolutely fundamental if and only if there is nothing distinct from x, y,
such that y is relatively fundamental to x. Absolutely fundamental beings
are those that are metaphysically ungrounded, full stop.
Now, to say that grounding is well-founded is to say that for any non-
empty grounding domain D there must be at least one metaphysically
ungrounded entity in D.28 To unpack this idea, call a minimal element, e,
of a non-empty grounding domain D one such that there is no y in D such
that e is grounded in y. A minimal element of D is one that is not grounded
by anything else in D; e is absolutely metaphysically fundamental and
foundational with respect to D. With this notion of a minimal element
for a domain in hand, we can state the well-foundedness of grounding
as follows:
G7: Well-Foundedness: for any non-empty grounding domain D

there is, of necessity, at least one minimal element in D.
G7 is the metaphysical analogue of the axiom of foundation in set theory,

that every non-empty set contains a minimal element (as a member), as
well as the metaphysical analogue of foundationalism in epistemology,
that every chain of epistemic justifcation terminates in a set of basic
beliefs that are non-inferentially justifed. The denial of G7 amounts to
the possibility that the exhaustive inventory of reality consists entirely of
grounded entities: it’s just one grounded thing after another such that a
exists in virtue of b, b exists in virtue of c, c in virtue of d, and so on ad
infnitum. On the denial of G7, there is no minimal grounding element
and thus no ultimate metaphysical foundation or bedrock to D. By my
lights, a minimal participatory ontology entails the metaphysical impos-
sibility of what Bohn (2018: 8) calls “infnite descent” in this context,
the view that “it is possible that there is no fundamental ground at all.”29
Of course, G7 is directly relevant to a core tenet of minimal partici-
patory ontology, that God is absolutely metaphysically fundamental
and prior to all created reality (DF).30 How might we think about DF in
grounding-theoretic terms along the lines of G7? Suppose we consider
the maximally inclusive grounding domain, DI, to include not only the
creature-to-creature grounding relations that obtain but also the (asym-
metric) grounding relations that obtain between God and creatures. We
can then formulate DF in grounding-theoretic terms along the lines of G7
as follows: God is the minimal element of the most inclusive grounding
306 Ross D. Inman
domain, DI. God is the absolute metaphysical foundation of and meta-
physically prior to all non-God reality; God alone is the sole (absolutely)
ungrounded reality.
In addition to the aforementioned theistic reason for afrming G7 as
part and parcel of a grounding-theoretic account of a minimal participa-
tory ontology, there are additional theoretical reasons why one might
endorse G7 as a constraint on metaphysical grounding. Many metaphysi-
cians who feel the pull of G7 justify their acceptance of it based on its
naturalness or intuitiveness; G7 strikes them as a reasonable thesis that, in
the absence of overriding considerations to the contrary, is more plausible
than its denial.31 Some proponents of G7 take its underlying motivation to
be that if there were no lower bound to the grounding domain of reality,
then nothing would exist in reality. Leibniz, for instance, in his June 30,
1704, letter to de Volder stated thus: “Where there is no reality that is not
borrowed, there will never be any reality, since it must belong ultimately
to some subject.”32
Jonathan Schafer (2010: 62) shares the Leibnizian intuition in that
“endless dependence conficts with the foundationalist requirement
that there be basic objects . . . Being would be infnitely deferred, never
achieved.” Others simply report their inability to comprehend the denial
of G7. On this score, Lowe (1998: 158) candidly states, “all real existence
must be ‘grounded’ or ‘well-founded’. Such an ‘axiom of foundation’ is
quite probably beyond conclusive proof and yet I fnd the vertiginous
implications of its denial barely comprehensible.”
Wholly apart from a commitment to a minimal participatory ontology,
then, one can argue that accepting G7 has a certain theoretical utility
in that it ofers a unifed explanatory ground for the existence of each
grounded entity in a domain.33 In a domain deprived of a minimal ele-
ment, the existence of each grounded entity is explained in terms of a
distinct (albeit immediate) ground or collection of grounds, which are
themselves grounded entities. Accepting the well-foundedness of ground-
ing, on the other hand, allows the ungrounded entity to serve as one and
the same explanation for the existence of each grounded entity in that
domain. This preserves the theoretical principle that it is better to have a
single explanatory ground for each phenomenon (the phenomenon here
is the existence of each individual grounded entity) than to have a distinct
explanatory ground for each phenomenon.
Even more, positing at least one ungrounded entity in a grounding
domain lends an explanation for not only the existence of each grounded
entity in that domain, but also the existence of grounded entities per se
in that domain. It is one thing for there to be an explanation for the
existence of each grounded entity in a domain, quite another for there
to be an explanation for why the class of grounded entities exist in that
domain in the frst place.34 Accepting G7 afords both theoretical sim-
plicity and explanatory power with respect to the existence of grounded
entities. Of course, such theoretical simplicity and explanatory power
provide independent, theoretical warrant for the core tenets of a mini-
mal participatory ontology (CED and CND) and its natural corollaries
(TEB and TEI).
3. Neo-Aristotelian Grounding for Creaturely

Participation
We have been concerned up to this point with getting clear on the for-
mal and structural features of a particular species of Grounding, viz.
ontological dependence, and how exactly it relates to minimal partici-
patory ontology. However, it is important to point out that ontologi-
cal dependence, as a particular species of Grounding, is also a genus
that itself admits of a variety of diferent species, not all of which are
equally suited to secure CED, CND, and DF as core tenets of mini-
mal participatory ontology. And since our immediate context in this
chapter is the claim that creatures participate in God by way of both
their existence and natures being asymmetrically grounded in God, it
is important to identify a particular species of metaphysical ground-
ing that fts the bill in question.35 What we are looking for, then, is a
species of metaphysical grounding that is both fne-grained enough to
account for both existential and essential dependence of creatures on
God (CED and CND), as well as one that supports the metaphysical
priority of God with respect to creatures, thus securing the crucial
asymmetry between God and creaturely being at the heart of a theistic
participatory ontology. By my lights, only a neo-Aristotelian account
of grounding is suited for the task.
3.1. Modal-Existential Grounding

It is commonplace in contemporary metaphysics to fnd ordinary ground-
ing claims such as “ordinary composite objects are grounded in their
proper parts” and “tropes are grounded in their bearers.” These ordinary
grounding claims naturally give expression to the idea that, at the very
least, the existence of the one is grounded in the existence of the other(s).
Ordinary composite objects exist in virtue of the existence and structure
of their proper parts; or particularized properties or tropes exist in virtue
of the existence of their bearers. These grounding claims are very often
explicated in modal-existential terms such that it is metaphysically neces-
sary that composite objects (or tropes) exist only if their proper parts (or
bearers) exist.
One common way of formulating this variety of existential grounding
is in the following modal terms:36
(RG) x is rigidly existentially grounded in y = def □(Ex → Ey)

308 Ross D. Inman
As a rigid form of existential grounding, RG captures the insight that one
thing (x) may depend on another specifc entity (y) for its existence: it is
metaphysically impossible for x to exist unless y – that very entity – exists.
It would seem that RG naturally fts the bill when it comes to elucidat-
ing CED as a core tenet of minimal participatory ontology, the claim that
creatures metaphysically depend on a specifc entity for their existence,
namely God. Surely, it is metaphysically impossible for creatures to exist
unless God Himself existed. Yet, there are several good reasons for think-
ing that RG fails to adequately account for the species of grounding at
work in a minimal participatory ontology.
First, recall that creaturely participation in God is asymmetric; crea-
tures exist and are what they are in virtue of God, not vice versa. Yet,
there are plausible instances of RG that are symmetric.37 If so, then x’s
being rigidly existentially grounded in y doesn’t guarantee that y fails to
be rigidly existentially grounded in x, which would be needed in order to
secure the claim that y is more fundamental than x.
Consider the following examples of mutual or symmetric existential
grounding along the lines of RG. On one particular brand of trope-bundle
theory defended by Peter Simons (1994), objects consist of a two-tiered
bundle of tropes. On the one hand, the essential features of an object are
determined by a particular bundle of mutually dependent tropes (what
Simons calls the “nucleus”), while the non-essential features are deter-
mined by a further bundle of tropes that inhere in the nucleus. Or consider
the grounding relations that obtain between Socrates and his biological
life, taken as a temporally extended event. Arguably, the existence of
Socrates’s life is rigidly existentially grounded in Socrates (necessarily, it
exists only if Socrates exists), and Socrates is rigidly existentially grounded
in his biological life (necessarily, he exists only if his life exists). Conse-
quently, it would seem that RG is unable to secure the asymmetry of
creaturely existential dependence on God as per CED.
A second, more fatal worry is that RG, as a purely modal-existential
species of grounding, is ill-suited to capture an integral dimension of crea-
turely participation in God, namely, the essential dependence of creatures
on God along the lines of CND. Recall that minimal participatory ontol-
ogy consists of the view that creaturely natures are patterned after God
along the lines of divine exemplarism; both the existence and the content
of creaturely essences are metaphysically grounded in God.
However, following the contemporary infuential work of Kit Fine
(1994, 1995) and other neo-Aristotelian-minded philosophers, the essence
or nature of a thing is arguably not reducible to what is modally required
for its existence. In other words, what is true of a creature in every pos-
sible world in which it exists does not adequately capture its essence,
what it is fundamentally. If so, then a purely modal-existential account
of grounding along the lines of RG will not sufce to adequately capture
the way in which the essences of creatures are grounded in God.
While a neo-Aristotelian defense of the irreducibility of essence to
modality is well beyond the scope of this chapter, let me ofer one Finean
counterexample to a purely modal account of essence.38 While modal
requirement for existence may be necessary for a thing’s essence (after all,
how could a creature be essentially F if it didn’t have F in every possible
world in which it exists?), it is not sufcient to adequately capture what
a particular creature is fundamentally.
Consider Fine’s (1994) now well-known counterexample involving
Socrates (s) and his singleton set {Socrates}, the set whose sole member
is Socrates. On standard modal set theory, necessarily, if Socrates exists
then he is a member of his singleton set:
(a) □(Es → s ∈ {Socrates})
If modal requirement for existence were both necessary and sufcient

to capture the essence of Socrates, then (a) would be equivalent to the
following essentialist thesis, where “□s” is to be read as the essentialist
operator “it is part of the essence of s that:”
(b) □s(s ∈ {Socrates})
But, argues Fine and others, (b) seems deeply implausible. It is difcult
to see how Socrates’s essence, what Socrates is fundamentally, involves
reference to any set-theoretic entity whatsoever. As Fine (1994) puts it,
“There is nothing in the nature of a person, if I may put it this way, which
demands that he belongs to this or that set or which demands, given that
the person exists, that there even be any sets.”
More precisely, the worry here is that there is an important modal sym-
metry between Socrates and his singleton across possible worlds where
Socrates exists; necessarily, if Socrates exists then his singleton exists (and
vice versa). Yet, there is arguably an essential asymmetry between Socrates
and his singleton regarding their fundamental nature or identity; while
the essence of {Socrates} is what it is in virtue of Socrates, the essence
of Socrates is not what it is in virtue of his singleton, {Socrates}. What
{Socrates} is fundamentally seems wholly irrelevant to what Socrates is
fundamentally. Thus, to reduce what a creature is fundamentally to what
is true of a creature in every possible world in which it exists is wholly
inadequate. In the place of a purely modal account of essence, Fine recom-
mends a primitive, non-modal account of essence that is best unpacked in
terms of the more traditional notion of Aristotelian real defnition.
3.2. Essential-Existential Grounding

If RG is ill-suited to account for the core tenets of minimal participatory
ontology, CED and CND in particular, what might we ofer in its place?
310 Ross D. Inman
I’m inclined to think that it is only if we take the neo-Aristotelian turn
and opt for a non-modal or essentialist notion of existential grounding
that we can adequately make progress in our attempt to formulate a
grounding-theoretic framework of a minimal participatory ontology in
terms of CED, CND, and DF.
With the notion of essence as primitive, we can formulate the following
non-modal species of existential grounding, call it “essentialE grounding”:
(REG) x is rigidly essentiallyE grounded in y = def □x(Ex → Ey)
REG states x is rigidly essentiallyE grounded in y if and only if it is part

of the essence of x such that if x exists, then y exists. In important ways,
REG improves on RG as a species of metaphysical grounding that aims
to capture the notion of creaturely participation in God along the lines of
CED and CND. For one, it is crucial to note that REG is more fne-grained
than RG precisely because it entails (but is not entailed by) its respec-
tive modal counterpart (RG); that is, while every case of rigid essentialE
grounding is a case of rigid existential grounding, the converse does not
hold. More specifcally, if it is part of the nature or essence of Socrates’s
singleton that it exist only if Socrates exists (REG), then it is necessary that
if Socrates’s singleton exists then Socrates exists (RG). In general, since
the essence or real defnition of a singleton set involves reference to the
existence of its sole member, the existence of the singleton both essentially
and existentially necessitates the existence of its sole member.
Likewise, as a non-modal species of grounding, REG does not entail
that the essence of the ground is wholly irrelevant to the existence of
that which is grounded. Recall that according to RG, since it is neces-
sarily the case that in every world in which Socrates exists his singleton
exists, Socrates is rigidly existentially grounded in his singleton; Socrates,
the grounded, necessitates the existence of his singleton, the ground. Yet
intuitively, as we noted earlier, the existence of a set-theoretic entity seems
wholly irrelevant to whether Socrates exists.
In contrast to RG, the neo-Aristotelian move to REG has no such impli-
cation. According to REG, from the fact that it is necessarily the case that
in every world in which Socrates exists his singleton exists, it does not
follow that Socrates is rigidly essentiallyE grounded in his singleton; it is
no part of the essence of Socrates that he exists only if such a set-theoretic
entity exists. Again, the relevant grounding ordering intuitively runs from
Socrates to Socrates’s singleton, not the other way around. Similarly, from
the fact that it is necessarily the case that in every world in which Socrates
exists the temporal event that is his life exists, it does not follow that
therefore Socrates is rigidly essentiallyE grounded in his life; arguably the
event exists and is what it is (and which particular event it is) in virtue
of Socrates, not the other way around. Insofar as REG entails but is not
entailed by RG, it rightly models the order of grounding as Socrates is
arguably more fundamental than either his singleton or the temporally
extended event that is his life.
There is, however, reason to think that even the aforementioned non-
modal species of grounding in terms of REG is neither fne-grained
enough nor suited to support all three core tenets of a minimal partici-
patory ontology (CED, CND, DF). First, some have argued that there
are plausible cases of essentialE grounding that are symmetric, and that
arguably all forms of existential grounding, whether modal or non-modal,
are best characterized as non-symmetric. For instance, on certain (realist)
structuralist mathematical ontologies that explicate numbers as nodes in
a structural network, it is part of the essence of each number that it exists
in virtue of the other nodes in the network. If so, then structuralist views
in the philosophy of mathematics ofer plausible examples of symmetric
essentialE grounding.39 Thus, REG doesn’t appear to be suited to support
the crucial asymmetry of creaturely participation in God, nor does the
core claim outlined in DF that God is absolutely fundamental and meta-
physically prior to creatures.
Second, as Kit Fine (1995: 274) and Kathrin Koslicki (2013: 51–60)
have argued, all forms of existential grounding, whether modal or non-
modal, are too weak insofar as they characterize grounding solely in terms
of what an entity requires for its existence. There certainly seems to be
more to the nature or essence of a thing than what that thing requires for
its existence. Perhaps, for instance, there’s more to the essence of a lion
than what a lion modally requires for its existence (think Aslan).
Some philosophers, for instance, have argued that there are plausible
cases where x is essentiallyE grounded in y, where nevertheless the identity
of x fails to be grounded in the identity of y. Lowe (2006: 199–200),
for instance, has argued, “[v]ery plausibly, an entity can, for example,
depend essentially for its existence on one or more other entities, without
necessarily depending essentially for its identity upon those other enti-
ties” (emphasis in original). Lowe points to an Aristotelian immanent
universal, roundness, as an example. As an immanent universal, round-
ness is such that it is part of its essence that it exists only if some instance
or other exists and is round; thus immanent universals are entities that
are essentiallyE grounded in a more broad sense than REG. However, the
essence of the universal roundness – what it is fundamentally and which
entity it is – does not involve reference to any single round object nor to
the totality of all actually existing round objects.
Along similar lines, Fine (1995: 274) presses the point further,
The present examples [viz. impossible objects and identity properties]

highlight a problem that besets any existential account of dependence,
whether it be modal or essentialist in form. For, it does not seem
right to identify the “being” of an object, its being what it is, with
its existence. In one respect, existence is too weak; for there is more
312 Ross D. Inman
to what an object is than its mere existence. In another respect, exis-
tence is too strong; for what an object is, its nature, need not include
existence as a part.40
It would seem, then, that any form of existential grounding, whether

modal or non-modal, is ill-suited to capture the full range of ways in
which creaturely natures are grounded in God, not merely in terms of
what such natures require for their existence. If so, then a neo-Aristote-
lian-grounding framework for minimal participatory ontology may want
to carve out an even more fne-grained notion of essential grounding
strictly in terms of what the thing requires for its identity and not merely
its existence.
3.3. Essential-Identity Grounding

For neo-Aristotelians like Lowe (2005) and Koslicki (2013), essentialist
(non-modal) species of grounding admit of further classifcation to include
identity grounding, cases where the identity of x and which thing of its
kind x is are grounded in the identity of y and which thing of its kind
y is. Mark Johnston (2006: 676) captures this more robust connection
between essence and grounding nicely,
Associated with the ideas of real defnition and essence is the idea of
the ontological dependence of one item on another; where an item x
is ontologically dependent on an item y just when y features at some
point in the full account of the essence of x (the real defnition of x),
but not vice versa.
In the same vein, Fine (2010: 582) notes, “One object may be (onto-
logically) prior to another in the sense that it is possible to provide an
explanation of the identity of the one object, to explain what it is, with
the help of the other object” (emphasis in original).
Following Lowe (1998: 149) we can formulate this further species of
essential grounding in the following manner, call it “essentialI grounding:”
(IG) x is essentiallyI grounded in y = def there is a function ϕ such that

□x(x = ϕ(y)).41
IG moves beyond purely existential grounding, whether modal (RG) or

non-modal (REG), in the sense that it specifes what is required not simply
for x’s existence per se but what is required for x’s identity, what x is and
which thing of a certain kind x is. The set {Fido, Wilber}, for example,
is essentiallyI grounded in its members, Fido and Wilber, in that there is
function, the set-formation function, and it is part of the essence of {Fido,
Wilber} that it be formed by Fido and Wilber as its members; the identity
of the set {Fido, Wilber}, what the set is and which set it is, is grounded
in its specifc members.
Along with Lowe (2012), I take essentialI grounding to be a relation of
individuation; if x essentiallyI grounds y, then the individuality of y is fxed
by the individuality of x, where x serves as the individuator of y. Since
sets are at least partially essentiallyI grounded in their members, Fido and
Wilber individuate the set {Fido, Wilber}; which set {Fido, Wilber} is fxed
by Fido and Wilber.42 Or, consider once again an ontology of events that
construes events as a triple [o,P,t], where “o” stands for some object or
objects, “P” a property (whether monadic or polyadic), and “t” a time.
If [o,P,t] is essentiallyI grounded in o, P, and t as its constituents, then the
latter ground not only the essence of [o,P,t] generally qua event, but also
which individual event it is, say the sinking of the Titanic (as opposed
to the Battle of Hastings). Which event [o,P,t] is fxed by its essentialI
grounds, its individual constituents.
As before, IG entails but is not entailed by the broader species of
essential grounding in terms of REG. If a set {Fido, Wilber} is essentiallyI
grounded (IG) in its individual members Fido and Wilber, then it is rigidly
existentially (RG) as well as rigidly essentiallyE grounded (REG) in Fido
and Wilber as its members. Since IG moves beyond mere modal and non-
modal existential grounding, it is uniquely situated to account for the
core tenet of CND, that each and every aspect of creaturely essences are
metaphysically grounded in God. Moreover, since IG entails both forms
of existential grounding, both modal (RG) and non-modal (REG), IG also
secures CED, that the existence of creatures is metaphysically grounded
in God.
We are left with the question of whether this neo-Aristotelian concep-
tion of grounding in IG secures the needed asymmetry at the heart of a
minimal participatory ontology. As a species of non-modal grounding,
essentialI grounding moves beyond a view of grounding in purely exis-
tential terms (whether modal or non-modal) and arguably provides the
requisite asymmetry we are looking for when it comes to God’s supreme
existential and essential priority over creatures. While two entities may
be mutually essentiallyE grounded in one another, this is arguably not the
case for essentialI grounding.43
As a strict partial ordering relation (G2–G4), IG is well suited to carve
out relations of metaphysical priority and posteriority and support the
notion of one concrete entity’s being more or less fundamental than
another. For instance, an event’s being metaphysically posterior to or less
fundamental than its constituents is understood in terms of the fact that
the event is essentiallyI grounded in its constituents, and not vice versa.
Similarly, an occupant of spacetime is metaphysically prior to its spa-
tial boundary precisely because the former metaphysically fxes what the
boundary is in general, as well as which boundary it is in particular, and
not vice versa. Lastly, since tropes or modes are essentiallyI grounded in
314 Ross D. Inman
their bearers, the very identity of a mode – what a redness trope of an
apple is in general, as well as which redness trope it is in particular – are
grounded in the identity of its bearer and not vice versa. We could mul-
tiply examples.
The idea that all creatures, great and small, are essentiallyI grounded
in God seems to encompass everything we want in a minimal participa-
tory ontology. In virtue of being essentiallyI grounded in God, the very
existence (CED) and content of all creaturely natures (CND) is what it is
in virtue of being so grounded.
Let me at least attempt to ofer a preliminary sketch of a more fne-
grained theistic gloss on how all creaturely natures are asymmetrically
grounded in God (CND) along the lines of IG, using the example of a
particular lion named Rory. The complete essential profle of Rory the
lion – those leonine features that constitute the identity of a lion in gen-
eral (its being living, mammalian, carnivorous, etc.) as well as Rory in
particular (Rory being the very lion that she is) – are all grounded in and
resemble the rich plenitude of the divine nature as per divine exemplarism.
More specifcally in terms of IG, suppose we take the primitive “crea-
turely resemblance to God” function as a function (φ) from existing crea-
tures to divine ideas of existing creatures, and let “IR” stand for God’s
idea of Rory the lion in particular (y), which includes all general leonine
features that are constitutive of the identity of a lion (living, mammalian,
carnivorous, etc.) as well as the particularity that is defnitive of Rory as
an individual instance of the general leonine kind. On divine exemplarism,
each and every existing creature resembles a particular intentional aspect
of the divine mind in precisely this way. Thus, Rory the lion is essentiallyI
grounded in IR along the lines of IG insofar as there is a function φ – “the
creaturely resemblance to God” function – such that it is part of the
essence of Rory that she (and she alone) resembles God with respect to IR
in particular (y). Rory, we might say, is the unique creaturely expression
of that particular intentional aspect of the divine mind.44 Another way of
saying this is that it is part of the essence of Rory that she is the unique
creaturely being she is in virtue of resembling IR, God’s idea of Rory. It is
not part of the essence of a distinct lion, say Leo the lion of MGM fame,
that it resemble God by way of IR, just as it is in no way part of the essence
of the set {Fido, Wilber} that it resemble God by way of God’s idea of the
set {Larry, Moe, Curly}.
Since essentialI grounding is arguably a relation of individuation, not
only is the general essence of Rory (what it is to be a lion) essentiallyI
grounded in IR, but so is Rory’s creaturely particularity, which particular
lion Rory is of the general leonine kind. God himself is (at the very least)
a partial essentialI ground for the existence and essence of each existing
creature (as per TPG in Section 2.2). The asymmetry of IG here is crucial
in the case of creaturely participation: Rory exists and is the way she is
essentially – indeed is the very lion she is – ultimately in virtue of the
particularity and plenitude of God’s mental life, and not vice versa. Echoes
of a broadly similar conception of creaturely participation in God along
the broad lines of IG can be found throughout the Christian tradition.
Recall once again Augustine’s previous remark that all creaturely natures
are “created in accord with the reasons unique to them,” reasons that
“must be thought to exist nowhere but in the very mind of the Creator.”
Along the same lines, Aquinas notes, “all creatures are nothing but a kind
of real expression and representation of those things which are compre-
hended in the conception of the divine Word.”45
In conclusion, I have tried to show how a distinctively neo-Aristotelian
gloss on metaphysical grounding is best suited to provide the most prom-
ising grounding-theoretic framework for a minimal participatory ontol-
ogy along the lines of CEN, CND, and DF. The neo-Aristotelian move
from a modal (RG) to a non-modal species of grounding (REG), which
assumes the irreducibility of essence to purely modal notions, provided
the proper ontological framework to account for the dependence of both
the existence and essence of creatures on God (CND). Second, the fur-
ther neo-Aristotelian move from a mere essential-existential variety of
grounding (one that limited essential grounding to what a thing requires
for its existence by nature) to the notion of essential-identity grounding
(IG) provided a fuller conception of essential dependence as well as the
requisite asymmetry at the heart of a minimal participatory ontology; God
alone is absolutely metaphysically fundamental in that, for God, there is
no distinct entity in which God is essentiallyI grounded, and all non-God
reality is essentiallyI grounded in Him.
Notes
1. An excellent introduction to a participatory framework applied to a broad
range of theological and philosophical areas is Davison (2019).
2. See in particular the work of Gavrilyuk (2009).
3. Where x is a proper part of y if x is a part of y and x is not identical to y.
4. This would exclude certain forms of panentheism. See Crisp (2019).
5. For more on the bearing of divine simplicity on contemporary mereological
metaphysics, as well as this second line of reasoning developed in more detail,
see Inman and Pruss (2019).
6. I owe this citation to Davison (2019: 43).
7. For outstanding treatments of divine exemplarism in the Christian tradition
see Boland (1996), Doolan (2014), Te Velde (1995), and Ward (2020). See
Davison (2019: ch. 4) and Ward (2020: section 7) for the explicit connection
between divine exemplarism or the theory of divine ideas to a participatory
ontology in particular.
8. My aim in ofering these samples from the Christian tradition is to merely
point the reader to a longstanding and deeply entrenched pattern in Christian
refection on the nature of God and creation. I make no attempt to be com-
prehensive here.
9. I owe this citation to McIntosh (2012: 368).
10. Davison (2019: 97–98).
316 Ross D. Inman
11. The tradition of divine exemplarism spans the post-Reformation and Enlight-
enment periods, and extends well into the nineteenth century. See Ames
(1968: 95), Muller (2017), van Mastricht (2019: 258), Crisp (2020), and
Bavinck (2004: 206), respectively.
12. If one thinks that facts are the primary (perhaps sole) relata of the grounding
relation, then the explanatory theses to follow will play a more central role
in outlining a minimal participatory ontology. On my own view, as facts are
less fundamental than their constituent entities, it is the grounding relations
that hold between entities that determine which facts obtain.
13. That is, assuming what Jaegwon Kim (1988) has called “explanatory realism,”
the view that explanations track objective dependence relations in the world.
14. For those who gloss metaphysical grounding or explanation in terms similar
to causation, that is, “metaphysical causation,” see Schafer (2016) and Wil-
son (2018).
15. I owe this fne phrase to Murphy (2011: 62).
16. This distinction is helpfully emphasized by Murphy (2011: 62).
17. For more on the claim that both God and creatures provide dual, non-
competitive metaphysical explanations for certain facts regarding creaturely
action, see Davison (2019: ch. 9), Grant (2019), and Koons (2002).
18. This, perhaps, is enough to safeguard against the charge of occasionalism. See
Grant (2019).
19. For a defense of something similar to TEI see Philip Quinn (1988: 87, 98). I
owe these citations to Robert Garcia (2015: 114).
20. See Fine (2012).
21. Bennett (2017) takes a genus-species approach to what she calls “building
relations.” This view also has precedent in Rodriguez-Pereyra (2015). Rodri-
guez-Pereyra (2015: 519) notes,
Grounding is the noncausal generic relation of being F in virtue of (or,
equivalently, the generic relation of being F non-causally in virtue of).
Thus a relation is a case or species of grounding if it is a specifcation
of the non-causal generic relation of being F in virtue of. Therefore,
truthmaking is a case or species of grounding since it is the non-causal
relation of being true in virtue of: the proposition <Socrates is white> is
true in virtue of the fact that Socrates is white. Similarly, being right in
virtue of, being blue in virtue of, existing in virtue of, and many other
such relations are also cases or species of grounding.
22. Of course, one might take one particular species or determinate of Grounding
to be more fundamental than another, say grounding relations between facts
or truths as more fundamental than grounding relations between entities.
23. Additionally, all facts pertaining to the existence, nature, and activity of
creatures is in some way grounded in the being, nature, and activity of facts
about God and his causal activity. I will leave aside this aspect of creaturely
participation in God at the level of facts.
24. Davison (2019: 148) helpfully unpacks this asymmetry in the context of what
I am calling Creaturely Nature Dependence:
Here, we can also usefully note that the relation between exemplar and
likeness runs diferently in tis two directions. It is asymmetrical. We say
that the image is like the exemplar, which is what likeness means. We
only perversely say that the exemplar is like the image: my portrait is
like me; I am not “like my portrait”. The world bears some of God’s
likeness but God is not like the world, not least since the act of creation
is constitutive of the creature, but not to the creator.
25. See Kenneth Pearce (2017) for an interesting grounding proposal that God is
the “foundational ground” for the entire sequence of causal events he calls
“History.” Pearce contends that primary causation in this sense is not a causal
relation at all (except perhaps analogically). In a similar vein, Caleb Cohoe
(2013: n. 4) remarks, “Aquinas’s account of causality is therefore closer to
contemporary theories of ontological dependence than it is to the predomi-
nant contemporary theories of causality.”
26. TPG, as well as minimal participatory ontology in general, can also serve
to provide the ontological scafolding for a theistic metaethical theory that
accounts for creaturely goodness (and perhaps the convertibility of being and
goodness) in the fact that all creaturely existence and essence is metaphysi-
cally grounded in God. As Mark Murphy (2011: 165–166) points out along
these lines,
For on this concurrentist view, all created goodness is merely a participa-
tion in, a resemblance to, God’s goodness. One cannot properly contrast
the love appropriately given to God and the love appropriately given
to created goods if the goodness of created goods is thus derivative of
the divine goodness; all love of created goods is, in a way, a love of the
divine goodness. This does not entail horrifying instrumentalization or
nihilism with respect to created goods . . . it does, rather, acknowledge
that created goods have their goodness through another.
27. Where “Ex” is defned as (∃y)(x = y).
28. Where a grounding domain D is non-empty just in case there are at least
two existing entities in D (as per G2 and G5) that stand in the grounding
relation with respect to one another (remaining neutral as to which grounds
which).
29. In contrast to Bohn (2018: 8) who employs the possibility of infnite descent
as a potential defeater to the necessity of what he calls “(strong) divine
foundationalism” (the view that “anything distinct from God is existentially
grounded by God”), I employ the (conditional) necessity of DF as evidence
against the possibility of infnite descent.
30. In fact, several recent versions of the cosmological argument, as in the work
of Deng (2020) and Pearce (2017), have appealed explicitly to God as meta-
physically foundational in this sense akin principle G7.
31. See Fine (1991: 267).
32. Quoted in Adams (1994: 335).
33. See Cameron (2008) for instance.
34. See Pearce (2017) for further argumentation on this point.
35. My discussion of the varieties of metaphysical grounding follows closely
the work of Lowe and Tahko (2015). See Koslicki (2013) for an excellent
overview of the varieties of ontological dependence in the neo-Aristotelian
literature.
36. As per Section 2.2, here I use the sentential operator “E” for the existence
predicate and defne it in terms of the existential quantifer: Ex = def (∃y)
(x = y).
37. See Barnes (2018) and Thompson (2016).
38. For further detail see Fine (1994), Oderberg (2007), and Inman (2018).
39. Barnes (2018). Although see Lowe (2012) for a critical discussion of alleged
cases of symmetrical identity grounding or dependence (see later) in structur-
alist ontologies.
40. As quoted in Koslicki (2013).
41. I owe this formulation of Lowe’s account of essential-identity dependence to
Koslicki (2013: 51).
318 Ross D. Inman
42. However, if one opts for TPG as stated in Section 2.2 (for any existing entity
x that is not identical to God, x is partially grounded in God), God is also a
partial ground of {Fido, Wilber}; in fact, on TPG, no existing entity that is
not identical to God has wholly non-theistic grounds.
43. See Lowe (2012) for further defense of the asymmetry of essentialI grounding.
44. For discussion of whether a view like this is compatible with a full-blooded
doctrine of divine simplicity see Pruss (2011: 274–276) and Ward (2020:
33–38).
45. I owe this citation to Davison (2019: 103). For a similar account see Ames
(1968: 95) and Anselm (2007: 40).
References
Adams, Robert. 1994. Leibniz: Determinist, Theist, Idealist. Oxford: Oxford
University Press.
Ames, William. 1968. Marrow of Theology. Grand Rapids, MI: Baker Book
House.
Aquinas, Thomas. 2012. Summa Theologiae, trans. and eds. Fr. L. Shapcote, O.P.,
J. Mortensen, and E. Alarcn. Lander: The Aquinas Institute.
Audi, Paul. 2012. “A Clarifcation and Defense of the Notion of Grounding.” In
Fabrice Correia and Benjamin Schnieder (Eds.), Metaphysical Grounding:
Understanding the Structure of Reality. Cambridge: Cambridge University
Press.
Barnes, Elizabeth. 2018. “Symmetric Dependence.” In Ricki Bliss and Graham
Priest (Eds.), Reality and Its Structure. Oxford: Oxford University Press.
Bavinck, Herman. 2004. Reformed Dogmatics Vol 2: God and Creation, ed. John
Bolt, trans. John Vriend. Grand Rapids: Baker Academic.
Bennett, Karen. 2017. Making Things Up. Oxford: Oxford University Press.
Bohn, Einar Duenger. 2018. “Divine Foundationalism.” Philosophy Compass 13
(10), pp. 1–11.
Boland, Vivian, O.P. 1996. Ideas in God According to Saint Thomas Aquinas:
Sources and Synthesis. Leiden, Netherlands: Brill.
Cameron, Ross. 2008. “Turtles All the Way Down: Regress, Priority and Funda-
mentality.” The Philosophical Quarterly 58 (230), pp. 1–14.
Cohoe, Caleb. 2013. “There Must Be a First: Why Thomas Aquinas Rejects Inf-
nite, Essentially Ordered, Causal Series.” British Journal for the History of
Philosophy 21 (5), pp. 838–856.
Crisp, Oliver. 2019. “Against Mereological Panentheism.” European Journal for
Philosophy of Religion 11 (2).
Crisp, Oliver. 2020. God, Creation, and Salvation: Studies in Reformed Theol-
ogy. New York: T&T Clark Bloomsbury.
Damascus, John. 1958. On the Orthodox Faith in John of Damascus, Writings,
trans. Frederic H. Chase. Washington, DC: Catholic University of America Press.
Davison, Andrew. 2019. Participation in God: A Study in Christian Doctrine and
Metaphysics. New York: Cambridge University Press.
Deng, Duen-Min. 2020. “A New Cosmological Argument from Grounding.”
Analysis 80 (3), pp. 418–426.
Doolan, Gregory. 2014. Aquinas on the Divine Ideas as Exemplar Causes. Wash-
ington, DC: Catholic University of America Press.
Fine, Kit. 1991. “The Study of Ontology.” Nous 25, pp. 263–294.
Fine, Kit. 1994. “Essence and Modality.” Philosophical Perspectives 8, pp. 1–16.
Fine, Kit. 1995. “Senses of Essence.” In Walter Sinnott-Armstrong, Diana Raf-
man, and Nicholas Asher (Eds.), Modality, Morality, and Belief. Cambridge:
Cambridge University Press.
Fine, Kit. 2001. “The Question of Realism.” Philosophers Imprint 1 (1), pp. 1–30.
Fine, Kit. 2010. “Towards a Theory of Part.” Journal of Philosophy 107 (11),
pp. 559–589.
Fine, Kit. 2012. “A Guide to Ground.” In Fabrice Correia and Benjamin Schnie-
der (Eds.), Metaphysical Grounding: Understanding the Structure of Reality.
Cambridge University Press, pp. 37?80.
Garcia, Robert. 2015. “Tropes as Divine Acts: The Nature of Creaturely Proper-
ties in a World Sustained by God.” European Journal for Philosophy of Reli-
gion 7 (3), pp. 105–130.
Gavrilyuk, Paul. 2009. “The Retrieval of Deifcation: How a Once-Despised
Archaism became an Ecumenical Desideratum.” Modern Theology 25,
pp. 647–659.
Grant, Matthews W. 2019. Free Will and God’s Universal Causality. New York:
Bloomsbury.
Inman, Ross D. 2012. “Essential Dependence, Truthmaking, and Mereology:
Then and Now.” In Lukas Novak, Daniel D. Novotny, Prokop Sousedik, and
David Svoboda (Eds.), Metaphysics: Aristotelian, Scholastic, Analytic. Heusen-
stamm, Germany: Ontos Verlag, pp. 73–90.
Inman, Ross D. 2018. Substance and the Fundamentality of the Familiar. New
York: Routledge.
Inman, Ross D. and Alexander Pruss. 2019. “On Christian Theism and Unre-
stricted Composition.” American Philosophical Quarterly 56 (4),
pp. 345–360.
Johnston, Mark. 2006. “Hylomorphism.” Journal of Philosophy 103 (12),
pp. 652–698.
Kim, Jaegwon. 1988. “Explanatory Realism, Causal Realism, and Explanatory
Exclusion.” Midwest Studies in Philosophy 12, pp. 225–240.
Koons, Robert. 2002. “Dual Agency: A Thomistic Account of Providence and
Human Freedom.” Philosophia Christi 4 (2), pp. 397–410.
Koslicki, Kathrin. 2013. “Ontological Dependence: An Opinionated Survey.” In
Benjamin Schnieder, Miguel Hoeltje, and Alex Steinberg (Eds.), Varieties of
Dependence: Ontological Dependence, Grounding, Supervenience, Response-
Dependence. Munich, Germany: Philosophia Verlag, pp. 3164.
Loewer, Barry. 2001. “From Physics to Physicalism.” In C. Gillet and B. Loewer
(Eds.), Physicalism and Its Discontents. Cambridge: Cambridge University
Press, pp. 37–56.
Lowe, E.J. 1998. The Possibility of Metaphysics: Substance, Identity, and Time.
Oxford: Clarendon Press.
Lowe, E.J. 2005. “Ontological Dependence.” Stanford Encyclopedia of Philoso-
phy <http://plato.stanford.edu/entries/dependence-ontological/>.
320 Ross D. Inman
Lowe, E.J. 2006. The Four-Category Ontology. A Metaphysical Foundation for
Natural Science. Oxford: Oxford University Press.
Lowe, E.J. 2009. “An Essentialist Approach to Truthmaking.” In E.J. Lowe and
A. Rami (Eds.), Truth and Truth-making. Stocksfeld, UK: Acumen.
Lowe, E.J. 2012. “Asymmetrical Dependence in Individuation.” In Fabrice Cor-
reia and Benjamin Schnieder (Eds.), Metaphysical Grounding: Understanding
the Structure of Reality. Cambridge: Cambridge University Press.
Lowe, Jonathan and Tuomas Tahko. 2015. “Ontological Dependence.” In
Edward N. Zalta (Ed.), The Stanford Encyclopedia of Philosophy <https://
plato.stanford.edu/archives/win2016/entries/dependenceontological/>.
McIntosh, Mark. 2012. “The Maker’s Meaning: Divine Ideas in Salvation.” Mod-
ern Theology 28, p. 3.
Muller, Richard. 2017. “Calvinist Thomism Revisited: William Ames (1576–
1633) and the Divine Ideas.” In Kathleen M. Comerford, Gary W. Jenkins, and
W.J. Torrance Kirby (Eds.), From Rome to Zurich, between Ignatius and Ver-
migli: Essays in Honor of John Patrick Donnelly, SJ. Leiden, Netherlands: Brill,
pp. 103–120.
Murphy, Mark. 2011. God and Moral Law: On the Theistic Explanation of
Morality. Oxford: Oxford University Press.
Oderberg, David. 2007. Real Essentialism. New York, NY: Routledge.
Origen. On First Principles, Book I, chaps. 2–3, trans. G. W. Butterworth. Lon-
don: SPCK, 1936; reprint, Gloucester, MA: Peter Smith, 1973, p. 16.
Pearce, Kenneth. 2017. “Foundational Grounding and the Argument from Con-
tingency.” In Oxford Studies in Philosophy of Religion, Volume 8, pp. 245–
268. Oxford, UK: Oxford University Press.
Pruss, Alexander. 2011. Actuality, Possibility, and Worlds. New York, NY:
Bloomsbury Academic.
Quinn, Philip L. 1988. “Divine Conservation, Secondary Causes, and Occasional-
ism.” In Thomas V. Morris (Ed.), Divine and Human Action: Essays in the
Metaphysics of Theism. Ithaca: Cornell University Press, pp. 50–73.
Rettler, Bradley. 2017. “Grounds and ‘Grounds’.” Canadian Journal of Philoso-
phy 47 (5), pp. 631–655.
Rodriguez-Pereyra, Gonzalo. 2015. “Grounding is Not a Strict Order.” Journal
of the American Philosophical Association 1 (3), pp. 517–534.
Schafer, Jonathan. 2004. “Two Conceptions of Sparse Properties.” Pacifc Philo-
sophical Quarterly 85, pp. 92–102.
Schafer, Jonathan. 2009. “On What Grounds What.” In David J. Chalmers,
David Manley, and Ryan Wasserman (Eds.), Metametaphysics: New Essays on
the Foundations of Ontology. Oxford: Oxford University Press, pp. 347–383.
Schafer, Jonathan. 2010. “Monism: The Priority of the Whole.” Philosophical
Review 119, pp. 31–76.
Schafer, Jonathan. 2016. “Grounding in the Image of Causation.” Philosophical
Studies 173, pp. 49–100.
Sider, Theodore. 2012. Writing the Book on the World. Oxford: Oxford Univer-
sity Press.
Simons, Peter. 1994. “Particulars in Particular Clothing: Three Trope Theories of
Substance.” Philosophy and Phenomenological Research 54, pp. 553–576.
St. Anselm. 2007. Anselm: Basic Writings, trans. and ed. T. Williams. Indianapo-
lis: Hackett Publishing Company.
St. Augustine. 1982. Eighty-Three Diferent Questions, trans. David L. Mosher.
Washington, DC: Catholic University of America Press, pp. 79–81.
Tanner, Kathryn. 1988. God and Creation in Christian Theology: Tyranny and
Empowerment? Minneapolis: Fortress Press.
Te Velde, Rudi A. 1995. Participation and Substantiality in Thomas Aquinas.
Leiden, Netherlands: Brill.
Thompson, Naomi. 2016. “Metaphysical Interdependence.” In Mark Jago (Ed.),
Reality Making. New York: Oxford University Press, pp. 38–56.
Van Mastricht, Petrus. 2019, Theoretical-Practical Theology: Faith in the Triune
God. Volume 2, eds. Joel R. Beeke and Michael T. Spangler. Grand Rapids:
Reformation Heritage Books.
Ward, Thomas. 2020. Divine Ideas. New York: Cambridge University Press.
Wilson, Alastair. 2018. “Metaphysical Causation.” Nous 52 (4), pp. 723–751.
Witmer, D. G., B. Butchard, and K. Trogdon. 2005. “Intrinsicality without Natu-
ralness.” Philosophy and Phenomenological Research 70 (2), pp. 326–350.
12 God and Hylomorphism
Anne Siebels Peterson
1. Introduction
God is the paradigm non-hylomorphic being in Aristotle’s universe. Fur-
ther, the ontological dependency relationship between divine and natu-
ral beings for Aristotle is entirely asymmetric: natural beings depend on
the divine for their existence and their fulfllment, but not vice versa.
Despite these diferences, Aristotle’s views on divine unity are helpfully
understood by comparison and contrast with his views on the unity of
natural hylomorphic substances, since one of Aristotle’s epistemological
principles is that we must begin with things closer to us. Moreover, a
comparative study of divine and hylomorphic unity is necessary to help
us understand the relationship between divine and natural unities for
Aristotle. I argue that common criteria for the genuine unity of natural
hylomorphic substances imply that Aristotle’s God counts as a unity in
precisely the same sense, or in precisely the same sense as the forms of
hylomorphic substances; this result is problematic, since unity for Aristo-
tle is non-univocal between God and natural beings. In addition to being
divided into the ten categories, it is also divided into potentiality and
actuality; and actuality itself has many senses related via analogy. Thus,
a perspicuous articulation of the unity of God requires a perspicuous
articulation of the precise sense of actuality that defnes God. Shifting the
focus toward this way of illuminating the unity of God uncovers a new
(or rather, old!) way of understanding the relationship between divine and
natural in Aristotle’s universe.
2. Two Models of Unity in Hylomorphic Beings

What is the nature of unity? In particular, how should the unity of the
two components of a hylomorphic unity – matter and form – be under-
stood in relation to the unity of the whole? These two questions, the
generic question about unity and the specifc question about hylomorphic
unity, are interconnected, and we fnd proponents of diferent accounts
DOI: 10.4324/9781003125860-16
God and Hylomorphism 323
of hylomorphic unity defending those accounts by reference to diferent
accounts of the requirement for unity more generally.
Although many nuances diferentiate scholarly views on the answer
to the question about hylomorphic unity, these views can be helpfully
divided into two general categories. The frst takes the unity of each of
matter and form to be distinct from the unity of the hylomorphic compos-
ite (taken as a whole). On this option, hylomorphic composites are unities
made up of further unities – namely, matter and form. The second denies
any such distinction between the unity of the hylomorphic whole and the
unity of its matter or form. Proponents of the latter option tend to main-
tain that matter and form are no more than conceptually, defnitionally,
or intentionally distinct from the hylomorphic whole – not ontologically
distinct from it.
Consider frst those views that deny any ontological distinction between
matter, form, and composite. We can understand proponents of these
views as adhering to a model of hylomorphic unity on which matter and
form are no more than conceptually distinct (from each other as well
as from the composite whole), and we can refer to this model as the
conceptual distinction model of hylomorphic unity. Aryeh Kosman pro-
vides a helpful characterization of one specifc way of articulating this
general view, committing to “identity between a substance’s form and its
matter . . . an essential identity between things diferentiated only as an
ability and its exercise difer” (2013, 105).
The central goal of views that adhere to the conceptual distinction
model is to secure the unassailable unity of hylomorphic composites. If
matter and form are not ontologically distinct from the composite in
the frst place, then worries about the unity of that composite fall away.
According to this family of views, true and unassailable unity thus requires
the lack of ontological distinction between the components of that unity.
Anna Marmodoro, for example, explains that form brings unity to hylo-
morphic composites by “stripping” the initially separate elements of their
distinctness to yield a unifed whole – not by being a further distinct part
within the composite (2013, 17). “Matter and form for Aristotle are not
parts of the substance they compose,” Marmodoro concludes, “because if
they were . . . the substance in question would lack unity, and thus would
not be a new thing over and above the sum of its parts” (2013, 6).1
Mary Louise Gill’s view falls into this general family as well. She main-
tains that although the pre-existent material survives potentially, it does
not survive actually, and so there is actually no ontological distinction
between the matter and form of a hylomorphic unity (1989, 146–163).
She frames the issue in terms of the unity of defnition for hylomorphic
composites. Living hylomorphic beings are “defnable with reference to
their form alone” (163), rather than through something else that is prior
to them (such as pre-existent type of material), a criterion that, in her
324 Anne Siebels Peterson
view, any “genuine or primary substance” (145) must meet. Theodore
Scaltsas, like Marmodoro, argues for the view in directly ontological
terms: “[t]he paradigmatic unity, namely, substantial unity, is not com-
patible with being a plurality of many” (1994, 110).
We can see from these passages that arguments for the conceptual dis-
tinction model of hylomorphic unity tend to employ as a premise a more
generic requirement about the nature of unity: that any genuine unity
cannot depend on or be built out of ontologically distinct parts. The
relationship between any items that serve to explain the unity of a genuine
being can be nothing short of identity. Whether they frame this require-
ment directly in ontological terms or argue for it via certain constraints
on the unity of defnition (as Gill does), the commitment is the same:
a genuine hylomorphic unity cannot harbor an ontological distinction
between matter and form.2
By contrast, the alternate model of hylomorphic unity maintains an
ontological distinction between matter and form within the hylomorphic
unity; here I will refer to it as the ontological distinction model of hylo-
morphic unity. Kathrin Koslicki uses mereological hylomorphism as a
term for “any account which not only views wholes as compounds of
matter . . . and form . . . but which also takes both of these components
(and, in particular, form) to be themselves parts, strictly and literally
speaking, of the whole they compose” (2006, 717).3 Using part language
as Koslicki does is one way of committing to the ontological distinction
model, but there are alternative ways. Robert C. Koons rejects the idea
that form is a part of the hylomorphic composite (2014, 158), maintain-
ing instead that form is a “process” – one that has “material participants”
such that “A composite substance exists at time t because its material
components participated in an appropriately formal process in some inter-
val of time immediately before t” (2014, 159). Koons thus maintains that
only the material components of a hylomorphic whole are strictly and
literally parts of it; nonetheless, he clearly rejects what I am calling the
conceptual distinction model. Matter and form are non-identical to each
other and to the hylomorphic whole. Frank Lewis likewise maintains
that for Aristotle, each of matter and form are non-identical with the
composite (and with each other). For Lewis, this claim is motivated by
an interpretation of Aristotle’s views on change from Physics I according
to which any change – even substantial changes such as the coming into
being of a living thing – requires a pre-existent material that continues to
persist within the fnal product (1994, 247–248). Koons’s commitment
to the ontological distinction model is also motivated by the need for a
persistent matter (2014, 162), as well as by the need for the material parts
to retain their own powers and activities within the whole (2014, 167).
Earlier in this volume, Koons argues for a version of hylomorphism on
which a compound’s matter, at lower levels, includes virtual parts – parts
that have purely potential existence within the compound. However, since
he also maintains that the higher levels of a compound’s matter do include
actually existing parts (such as, in the case of living things, organs), his
view remains opposed to the conceptual distinction model of hylomor-
phism. Likewise, earlier in this volume William M. R. Simpson draws on
the contextual wave function collapse model of quantum mechanics to
argue for a hylomorphic pluralist view, one which recognizes the reality
of multiple hylomorphic composites at varying levels of reality. On this
view, substances “are metaphysically composed of matter and substantial
form” (52). A hylomorphic substance possesses a “parcel of gunk” as its
matter, and this parcel does possess genuine causal powers of its own,
though its powers are determined by the form of the whole hylomorphic
composite (53). The attribution of genuine causal powers to the hylomor-
phic substance’s matter suggests that this matter does not have merely
conceptual, but genuinely ontological, distinction from the whole.
Obviously, proponents of the ontological distinction model cannot adopt
the generic view of unity that commonly motivates proponents of the con-
ceptual distinction model, according to which genuine unity requires the
lack of any ontologically distinct components (at least, not unless they are
willing to deny the genuine unity of hylomorphic composites altogether).
Instead, proponents of this model defend the unity of hylomorphic com-
posites by appealing to distinctive features of the matter/form relationship
as the source of genuine unity, often focusing on the role of form.
Koslicki, for example, argues that hylomorphic unities are unities
because form belongs to a distinct ontological category from matter; as
Aristotle puts the point in Metaphysics VII.17, form is not an element but
a principle (1041b31). Further, as Koslicki interprets Aristotle, the reason
that form secures the unity of hylomorphic composites is that form is
the “ultimate mereological atom” – something “indivisible relative to all
conceivable measures” of unity (2006, 728–729). On this account, hylo-
morphic composites are unities because they incorporate within them-
selves a principle – form – which is simply indivisible. Indeed, Koslicki’s
account of the unity of form is deeply similar to the account of the unity
of hylomorphic composites put forward by proponents of the conceptual
distinction model of hylomorphic unity. But whereas Koslicki views form
as the principle whereby a hylomorphic composite secures its unity (so
that the composite has its unity derivatively or dependently) due to the
indivisibility of form, proponents of the conceptual distinction model of
hylomorphic unity view the hylomorphic composite as meeting the indi-
visibility criterion immediately, directly, or in its own right.
Sally Haslanger likewise defends the ontological distinction model of
hylomorphic unity, maintaining that “matter and form are proper parts of
the substance” (1994, 169). She goes on to address the point about unity:
[G]iven this we cannot explain the unity of the sensible substance

by treating the matter and form as dependent or posterior parts of a
substance, unifed by reference to the substance itself. Instead a sen-
sible substance is unifed by virtue of an intrinsic integration principle
that places substantial form in the role of privileged individual; matter
and form are unifed by their respective relations to the form that is
itself a member of the unity. . . . The unity of sensible composites is
grounded in the more basic (or prior) unity of substantial forms.
(169–70)
Whereas proponents of what I have called the conceptual distinction

model can explain hylomorphic unity by “treating the matter and form
as dependent or posterior parts of a substance” (Haslanger 1994, 169),
proponents of the ontological distinction model such as Haslanger and
Koslicki cannot – and Haslanger’s appeal to form as the ultimate source of
hylomorphic unity broadly aligns with Koslicki’s account of hylomorphic
unity grounded in the form, where form is taken to be a principle the unity
of which requires no further grounding.
Michael Loux highlights not just the ontological distinction between
matter and form but also the defnitional distinction between them that
undergirds their ontological distinction, claiming that “the relationship
between the matter and the predicated from is merely kata sumbebekos
[accidental], that the matter has an identity independent of the form so
that the form is predicated of it merely accidentally” (1995, 262–263).
Focusing on the text of Metaphysics VIII.6, a text that has played a central
role for many proponents of the conceptual distinction model of hylomor-
phic unity, Loux ofers a diferent account of hylomorphic unity than we
see in Koslicki and Haslanger. Instead of tracing the composite’s unity to
the ideal unity of the form, Loux argues that it is due to the special nature
of the relationship between matter and form:
[M]atter and form have categorial features such that their compres-
ence is necessarily both necessary and sufcient for the existence of
the composite entity whose constituents they are. . . . To be for a
bronze sphere is just to have the appropriate predicative structure, to
be a thing such that its existing consists in the fact that its matter, the
bronze, is actually spherical; and to be one for a bronze sphere is just
to be one bronze sphere. . . . And for each distinct kind of complex,
there will be a distinctive predicative structure, involving a matter and
form related as the bronze and sphere . . . are related. Accordingly,
there will be a diferent kind of being and unity in each case.
(271)
The unity of the hylomorphic composite is not simply derivative of the

more perfect unity of form but rather arises from the distinct categorial
structure of matter and form together – from the way these ft together
to make a unity of a given kind. Loux’s analysis, though adhering to the
ontological distinction model also followed by Koslicki and Haslanger,
denies the premise that genuine unity must come – immediately or
derivatively – from something which is a pure, non-complex unity.
That is, he ofers an alternative explanation for hylomorphic unity on
the same ontological distinction model accepted by Marmodoro and
Koslicki.
I have argued for another way of reading VIII.6, one that supports the
ontological distinction model and defends the unity of the hylomorphic
composite in a diferent way. I argue that due to the non-univocality
of unity in Aristotle, the objection that a unity (simpliciter) cannot be
composed of a plurality (simpliciter) simply fails to articulate a genuine
criterion for unity in the frst place. To get a genuine criterion, we must
supplement the equivocal terms with whatever the category of being at
hand is – in this case, substance: “no substance can be composed of sub-
stances.” But once we have done this, we see that hylomorphic unity on
the ontological distinction model was never in violation of this criterion;
for hylomorphic substances are composed of matter and form, but matter
and form are not themselves substances. Thus, no genuine threat to what
I am here calling the ontological distinction model has even been articu-
lated. We can still ask questions about what matter and form are and how
they relate to substance, but no genuine problem about hylomorphic unity
arises (Peterson 2018). On this view, it is the fact that neither component
of a hylomorphic substance is itself a substance (but rather, matter is a
potential substance and form is the actuality of a substance) that allows
the composite whole to count as a unifed substance.
In short, proponents of the ontological distinction model of hylomor-
phic unity maintain that matter and form are non-identical and ontologi-
cally distinct within the unifed hylomorphic composite, while proponents
of the conceptual distinction model maintain the identity of matter and
form within the unifed composite. For proponents of the latter model of
unity, the distinction between matter and form is not ontological – it is
purely conceptual. While proponents of the conceptual distinction model
tend to maintain that matter and form cannot be distinct within a hylo-
morphic unity, proponents of the ontological distinction model deny this
claim. They argue that genuine unities can obtain their unity derivatively
(from a component that itself lacks any distinction), or that a certain
feature of the categorial distinction between matter and form is itself the
source of unity for hylomorphic beings, or that no threat to hylomorphic
unity genuinely arises because the whole is a substance while matter and
form are not substances.
3. Comparing Divine and Hylomorphic Unity

According to Aristotle, no potentiality is involved in God’s being. And
since matter is, for Aristotle, the locus of potentiality, it follows that God
has no matter. Stephan Herzberg summarizes Aristotle’s argument to this
efect in Metaphysics XII.6–7:
In order to guarantee the eternal and uninterrupted motion of the frst

heaven, the frst principle must be eternal and, in addition, in actual-
ity. More precisely, it must be in actuality in the special way that it is
not the actuality of an underlying potentiality. If it were the actuality
of an underlying potentiality, then it would also be possible that its
potentiality might, at times, not be actualized and would, therefore,
not generate any motion. Yet since the motion has always existed,
Aristotle reaches the conclusion that this principle has to be actuality
according to its own essence, i.e. its essence is actuality. So it must
not possess any kind of potentiality.
(2016, 158)
Although Aristotle’s argument for the existence of God as an unmoved

mover, outlined in XII.6, is not my subject here, it is the reason he identifes
for the lack of potentiality in God. It is thus relevant background informa-
tion to highlight why, for Aristotle, God must be non-hylomorphic. Often
Aristotle’s God is described as being pure actuality (opposing objects
which are combinations of actuality and potentiality, for example natu-
ral objects). As Sylvia Fazzo has argued, it is closer to the Aristotelian
text to describe God as being in actuality (rather than as being actuality),
due to Aristotle’s use of the dative (ἐνεργείᾳ) moreso than the nominative
(ἐνέργεια) in discussing the nature of God (2016); but it is uncontroversial
that no potentiality and thus no matter is to be found in God.
Diferent accounts of the nature of hylomorphic unities – and in par-
ticular, diferent accounts of the background accounts of unity more gen-
erally employed by those accounts – imply diferent commonalities and
contrasts between those unities and the ultimate unity of God. These com-
monalities and contrasts have key implications for both how we under-
stand the overall nature of unity for Aristotle and how we understand the
divine/natural relationship for Aristotle.
We have seen that the conceptual distinction model of hylomorphic
unity is motivated by the view that a true unity cannot be built out of
distinct ontological components. On this account, what is shared between
hylomorphic and divine unities is just this ontological simplicity, that is,
the lack of any ontologically distinct components. This understanding of
unity implies that something is more unifed just to the extent that that its
unity is not composed of distinct parts. In short, God meets this account’s
standard for genuine unity in just the same sense that hylomorphic unities
meet it: both alike cannot be divided into ontologically distinct parts.4
We have also seen that certain proponents of the ontological distinction
model of hylomorphic unity (e.g. Koslicki and Haslanger) do not require
that a genuine unity itself lack any ontologically distinct parts. Instead,
they allow for something to count as a genuine unity in virtue of having
a part that lacks any ontologically distinct parts – that is, as Koslicki puts
it, in virtue of having a part that is the “ultimate mereological atom”
(2006, 728). In other words, such proponents of the ontological distinc-
tion model require that genuine unity ultimately (if only indirectly) be
rooted in something lacking any division. By contrast, proponents of the
conceptual distinction model require that any genuine unity must itself
meet that criterion. But the two views share the premise that the ultimate
criterion for genuine unity is lack of divisibility – they simply difer on
whether a genuine unity must itself (i.e. directly) meet this criterion or
whether it can meet the criterion in an indirect or mediated way.
God’s unity is clearly unmediated on Aristotle’s view, so that God does
not derive God’s unity from a further principle in the way that schol-
ars like Marmodoro and Koslicki maintain the hylomorphic composite’s
unity to be derived from form. Despite this disanalogy, however, God
does meet the criterion for unity that form is said by proponents of this
account to meet – and as we have seen, it is in virtue of form’s meeting
this criterion that it sufces to confer genuine unity on the hylomorphic
composite. The upshot is that although, on this account, God does not
meet the criterion for unity defnitive of hylomorphic objects (since their
unity is mediated or derivative), God does meet the same criterion for
unity met by that principle from which hylomorphic objects derive their
unity: form. God, on this account, is unifed in precisely the sense in which
the forms of hylomorphic objects are unifed.
Although Loux follows the same ontological distinction model (terming
Aristotle’s hylomorphism a “constituent ontology” [2012]), we have seen
that he ofers an altogether diferent criterion for genuine unity. His crite-
rion for genuine unity does not require either that an item be itself onto-
logically indivisible or that it derive its unity from something ontologically
indivisible. Instead, his criterion requires that the parts of the whole be
of complementary categorial kinds. Clearly, God simply cannot meet this
criterion of unity; for there is no distinction at all (much less a categorial
distinction) in God. The upshot is that on this account of hylomorphic
unity, God cannot be a unity in the same sense as hylomorphic wholes
or as that from which hylomorphic wholes derive their unity (since, on
this account, they do not derive their unity from a single principle at all,
but rather from the precise mode of interaction between form and mat-
ter). On the view I have argued for, hylomorphic wholes count as unifed
substances because the distinct parts that compose them, matter and form,
are themselves not substances; rather, one is potentially a substance and
the other is the actuality of a substance. Clearly, God does not count as a
unity in this sense, since God lacks potentiality.
Summing up the results so far, we can see that the conceptual dis-
tinction model and its criterion for unity can be met by God and by
hylomorphic objects in the very same sense – on this account, though
God and hylomorphic unities certainly difer in fundamental ways, they
do not difer qua unities. On the ontological distinction model, by con-
trast, God and hylomorphic unities do not count as unities in precisely
the same sense. However, if the hylomorphic composite derives its unity
from the form, and if the form in turn counts as an ideal unity in virtue
of being indivisible, then God counts as a unity in precisely the same
sense as the principle whereby hylomorphic composites get their unity
(i.e. form). If, by contrast, the substantial unity of an ontologically com-
plex hylomorphic substance comes from the relationship between matter
and form as ontologically distinct items, or from having parts which are
not substances, then God does not count as a unity in the same sense as
the hylomorphic composite itself (or in the same sense as either of its
components).
Thus far, then, three diferent pictures of the relationship between the
divine and the natural emerge when it comes to the issue of unity: (1) God
is a unity in the same sense in which hylomorphic composites are unities;
(2) God is a unity in the same sense as that principle whereby hylomorphic
composites attain their unity, namely form; or (3) God is a unity in a dif-
ferent sense altogether. Which of these results fts best with the details we
fnd on divine unity in the Metaphysics? Answering this question is the
task of the next section.
4. Unity, Analogy, and the Divine

In Metaphysics IV and VII Aristotle explains that being has many senses,
captured by the ten categories; these categories are related by having a
single, focal sense (that of “substance”). At the very beginning of Meta-
physics IX he reminds us of this point, adding, however, that being is
“in another way distinguished in respect of potentiality and actuality”
(1045b32–24, emphasis added).5 Actuality and potentiality are also dis-
tinct senses of being.
We likewise know from Metaphysics IV that unity and being go hand
in hand: where being is said in many ways, unity also is. Given this back-
ground, the reader of Metaphysics IX can already conclude that since
being is divided according to actuality and potentiality, unity must be
likewise divided into these two senses.
At this point, however, Aristotle takes the issue one step further, argu-
ing that the terms “actual” and “potential” themselves cannot be given
a univocal defnition:
Actuality [ἐνέργεια] means the existence of the thing, not in the

way which we express by “potentially”; we say that potentially, for
instance, a statue of Hermes is in the block of wood and the half-
line is in the whole, because it might be separated out, and even the
man who is not studying we call a man of science, if he is capable of
studying. Otherwise, actually. Our meaning can be seen in particular
cases by induction, and we must not seek a defnition of everything
but be content to grasp the analogy – as A is in B or to B, C is in D
or to D; for some are as movement to potentiality, and the others as
substance to some sort of matter.
[1048a30–1048b9]
There is no universal defnition of an actuality. However, rather than

claiming that the senses of “actuality” are connected by having one focal
sense on which all the other senses depend (in the way that all the acci-
dental categories depend on the focal category of substance), he instead
claims that the senses of actuality are connected only by analogy. No
single defnition is shared by all actualities qua actualities.
We have already seen that Aristotle argues for a God who lacks poten-
tiality and thus lacks any matter; actuality exhausts the essence of Aris-
totle’s God. Now because of the connection between unity and being, it
follows that to be one God is to be one actuality (or, following Fazzo’s
way of putting the point, to be in actuality in one way). However, this
still does not give us a full defnition of the unity of God; for as we have
just seen, Metaphysics IX tells us that “actuality” just as such lacks a
defnition. We thus need to specify further which of the analogically
related senses of actuality is at issue; only then will we have genuinely
defned divine unity. And clearly, God will not meet either of the def-
nitions of actuality laid out in the earlier passage (1048b9); for God
involves neither movement nor matter. Indeed, as we will now see, it
turns out that no earthly sense of actuality will do; to capture the nature
of God, a new sense of actuality must be articulated, and articulating
it is just Aristotle’s project in Metaphysics XII.7–9. Thus, Aristotle’s
God is an actuality in a thoroughly diferent sense from hylomorphic
substances (and their parts).
In Metaphysics XII.7 Aristotle delves further into which sort of actual-
ity the unmoved mover is. The frst more determinate point he makes
about it is to claim, “its life (διαγωγὴ) is such as the best which we enjoy,
and enjoy for but a short time” (1072b14–15). He next identifes which
sort of life this is as that of thought, or contemplation (νοῦς): “For it is
ever in this state (which we cannot be). . . . And thought in itself deals with
that which is best in itself, and that which is thought in the fullest sense
with that which is best in the fullest sense” (1072b15–20). However, the
divine thought is not simply a more permanent version of our own con-
templative thought. After all, the human intellect is by nature in potential-
ity, depending on external objects to engage in active thought (De Anima
III.4, 429a21–24); the divine thought, by contrast, lacks any potency.
From this contrast Herzberg concludes that Aristotle discusses human
thought not as a direct connection to the divine thought, but rather “as
a suitable candidate for an analogy to the divine mode of being” (2016,
169, emphasis added); for divine thinking “cannot be identical in kind
with human thinking” (2016, 175).
In XII.9 Aristotle’s focus is to lay out the precise nature of the divine
thought – that is, its object and the way in which it is purely activity with-
out potentiality. Perhaps the key point is that the divine thought, unlike
human thought, does not depend on any objects external to itself in order
think; rather, the divine thought has itself as its object. Thus, it lacks any
potentiality since it depends on nothing outside itself for its actuality. As
Aristotle famously and cryptically puts it, “its thinking is a thinking on
thinking” (ἡ νόησις νοήσεως νόησις) (1075a34).
If to be is to be one, and actuality is one of the senses of being, and
actuality in turn remains unspecifed until one of its analogous senses is
articulated, it follows that the unity of any actuality likewise remains
unspecifed until its precise sense of actuality is articulated. This result
suggests that if God is an actuality in a diferent sense from natural actu-
alities such as hylomorphic composites or forms, then it also follows that
God is a unity in a diferent sense from natural actualities such as hylo-
morphic composites or forms. The sense in which God is a unity will be
only analogically related to the sense in which natural actualities are such.
We can now return to our two models of hylomorphic unity and
explore the implications of Metaphysics XII.7 for those models. First, we
are now in a position to see that the conceptual distinction model of hylo-
morphic unity, at least as it is commonly defended in terms of the lack of
ontological distinction, or ontological indivisibility, rendering something
a genuine unity, draws too simplistic a connection between the natural
and the divine. Given this model and this criterion for unity, hylomorphic
composites and God will count as unities in precisely the same sense – not
in a merely analogous way.6
A similar result applies to those arguments for the ontological distinc-
tion model of hylomorphic unity that take the ultimate criterion for unity
to be the lack of divisibility, and go on to apply this criterion to hylo-
morphic forms rather than to hylomorphic wholes themselves. Given this
criterion for unity and this understanding of natural forms, natural forms
and God will count as unities in precisely the same sense – not in a merely
analogous way.
To be clear, my argument is not a direct case against these two models
of hylomorphic unity. Rather, it has the structure of a reductio against
implementations of those models that employ either the premise that to
be a genuine unity is to be something that itself lacks divisibility or the
premise that to be a genuine unity is to be something that itself derives its
unity from a part that lacks divisibility.7 If either of these premises truly
revealed the nature of unity for hylomorphic objects, then God would
count as a unity in precisely the same sense as hylomorphic unities or
as the forms of hylomorphic unities that serve as the principle of their
unity (depending on which of these premises is at play). But as I have
argued in this section, key claims from the Metaphysics tell against this
result. God is an actuality in a diferent sense from natural actualities
such as hylomorphic composites and forms; therefore, because actuality
is a sense of unity, God is also a unity in a diferent sense from natural
actualities such as hylomorphic composites and forms. Indeed, Aristotle
does not ofer a generic, univocal criterion for unity even among natural
actualities; even less, then, does he ofer a univocal criterion for unity that
bridges divine and natural actualities. To understand God qua unity is to
understand God qua the distinctive sort of actuality that God is. And that,
as I have shown, is a certain kind of life: specifcally, “a thinking thinking
on thinking” (1074b34). In short, to be one God is to be one thinking
thinking on thinking.
These considerations suggest that even in the context of questions about
hylomorphism, we should not use either the premise that to be a genuine
unity is to be something that itself lacks divisibility or the premise that
to be a genuine unity is to be something that derives its unity from a part
that lacks divisibility. Just as there is more to the nature of genuine unity
in the divine case, so there is more to the nature of genuine unity in the
hylomorphic case and in the case of the forms of hylomorphic objects.
The nature of unity, for Aristotle, can in no case be illuminated by so
generic a criterion.
5. Divine Unity and Divine Simplicity: New Work for

an Aristotelian God?
Before delving into the details of this section, let me take a moment to
clarify what I will not argue. The point of this section is not to make any
argument against the applicability or importance of divine simplicity to
Aristotle’s own views (the doctrine certainly does have important ties to
certain aspects of Aristotle’s view of God, given his claim that God is not
just a being that thinks, but rather a being that is identical to God’s think-
ing). My goal is to show that despite its applicability and even importance
in other respects, appealing to divine simplicity as a (or even the) way to
illuminate the unity of God – that is, as the key to understanding God
qua unity – difers deeply from Aristotle’s own methodology in explaining
the unity of God. After highlighting this diference, I will explore some
of its implications for our understanding of the relationship between the
natural and the divine.
Let us begin with an exemplar whose discussion of divine simplicity was
deeply infuential in the ongoing tradition of Aristotelian scholarship. In
the Prima Pars of the Summa Theologiae, Thomas Aquinas frst argues
for the existence of God, drawing directly on Aristotelian principles. But
once this basic task is accomplished, his very next task – his frst foray into
the nature of God – is to defend the “simplicity or lack of composition”
of God (q. 3 a. 7).8 In doing so he clarifes that God is not composed of
matter and form (q. 3 a. 7 resp.). Most notably for my argument, he does
this even before arguing for the claim that there is one God; indeed, when
he does address the unity of God (q. 11), he argues from God’s simplicity
or lack of parts to God’s unity (q. 11 a. 3 resp.).
Aquinas’s strategy here in using simplicity as a key premise for unity
is reminiscent of the arguments we have seen for the conceptual distinc-
tion model of hylomorphic unity, in which proponents use ontological
simplicity as the criterion for genuine unity. Though less directly, it is also
reminiscent of the strategy used by those proponents of the ontological
distinction model who defend the unity of hylomorphic composites by
appealing to the ontological simplicity of their forms. In either case, the
nature of genuine or ultimate unity is illuminated in terms of ontological
simplicity – either immediately or mediately. In sum, as we have already
seen, these accounts maintain that hylomorphic unities are unities in vir-
tue of being, or being properly related to, an entity that is incomposite or
indivisible; but as I have argued in the previous section, if this were the
criterion that illuminated the nature of genuine unity, then God would
count as a unity in precisely the same sense as that in which hylomorphic
unities or their forms count as unities.
Aquinas’s explanation of divine unity in questions 3 and 11 of the Prima
Pars of the Summa Theologiae may seem at frst to similarly emphasize
an overly generic criterion for God’s unity, since he too explains that
unity in terms of God’s lack of composition. However, his overall view is
not subject to the same objection, since elsewhere he ofers a caveat that
allows him to avoid it. Aquinas warns us that when we make a certain
claim about God, the terms in that claim should not be understood to
apply to God in the same sense that we apply them to creatures: “univocal
predication is impossible between God and creatures” (q. 13 a. 5 resp.). If
we apply this point to Aquinas’s explanation of the nature of divine unity
in terms of divine simplicity, the implication that the criterion for divine
unity could apply in the very same sense, or univocally, to something in
the natural world (such as hylomorphic beings) is thereby blocked.
Despite Aquinas’s clear avoidance of this objection, however, there
remains a notable diference between Aristotle’s discussion of divine
unity and Aquinas’s: Aquinas does make the choice to emphasize these
points about simplicity and lack of composition in his initial endeavor to
illuminate the unity of God (if only under the constraints of analogous
predication); that is, he introduces God’s unity in terms of God’s simplic-
ity or lack of composition in question 3, going so far as to argue from
the simplicity of God to the unity of God. And although Aquinas denies
the possibility for his descriptions of God to apply univocally between
God and natural beings, thereby blocking the aforementioned objection
that faced proponents of certain views of hylomorphic unity, at the same
time Aquinas also afrms that our terms do not apply purely equivo-
cally between God and natural beings. Instead, like Aristotle, he afrms
that our terms can apply to God and creatures “analogically” (q. 13 a. 5
resp.). Thus, although our terms will never ofer full illumination of God’s
nature, Aquinas still takes himself to be going some way toward illumi-
nating the nature of God when he argues from divine simplicity to divine
unity, since our claims do apply to God in an analogous sense.
By contrast, I have argued that even with the constraint of merely
analogous predication in place, Aristotle himself would not see indivis-
ibility or simplicity of this sort as the way to illuminate God qua unity.
Although Aristotle agrees with Aquinas that our terms apply neither
univocally nor purely equivocally between God and natural beings
(since these have distinct senses of actuality) but rather analogously,
Aristotle does not choose to directly illuminate the unity of God in terms
of God’s lack of composition. Rather, he highlights the specifc features
of divine life that constitute the specifc sense of actuality, and thus of
unity, that God possesses. In more detail, he maintains that unity is
not itself a determinate concept, but one that has many senses, one of
which is actuality – thus to understand God qua unity is to understand
God qua actuality (not qua incomposite being). Further, since actuality
itself has multiple analogous senses, to understand God qua actuality
is to understand the distinctive sort of actuality that God is. Putting
all of this together, it follows that to understand God qua unity is to
understand God qua the distinctive sort of actuality that God is. And
that, as I have shown, is a certain kind of life: specifcally, “a thinking
thinking on thinking” (1074b34). Again, to be one God is to be one
thinking thinking on thinking.
The upshot is that Aristotle illuminates God qua unity not in terms of
indivisibility or lack of parts but rather in terms of that specifc sense of
actuality (among those familiar to us) that is most analogous to God’s
specifc type of actuality: the life of intellect. Although both Aristotle and
Aquinas agree that our terms apply only analogically between God and
creatures, they still make diferent choices about the best way to illumi-
nate the unity of God under that constraint. Although Aquinas does not
directly face the objection faced by proponents of the claim that to be one
for Aristotle is ultimately to be partless or incomposite, Aquinas’s use of
divine simplicity to illuminate divine unity in question 3 still privileges
that claim in the project of explaining unity in a way that Aristotle would
not privilege it.
Aristotle certainly agrees that God is partless and indivisible (Metaphys-
ics XII.7, 1073a3–7). However, for him these terms do not serve as the
central way to illuminate the unity of God, and we can see why this is so
in further detail by looking to Metaphysics IV.6, where Aristotle tells us
that there is no such thing as just being indivisible simpliciter:
For in general those things that do not admit of division are one
insofar as they do not admit of it, e.g. if something qua man does not
admit of division, it is one man; if qua animal, it is one animal; if qua
magnitude, it is one magnitude.
(1016b3–6)
He likewise claims in Metaphysics VII.10, “‘part’ is used in several senses”

(1034b34). The upshot is that it is not parthood or indivisibility as such
that serve to illuminate the unity of anything; on the contrary, in order to
apply parthood and indivisibility to any particular case we must already
have in hand a specifc sense of unity! To argue from partlessness or indi-
visibility to unity therefore gets things backwards, according to Aristotle;
parthood and indivisibility must be illuminated in terms of unity (which is
non-univocal) rather than the other way around. According to Aristotle,
the specifc sense of actuality that best illuminates the unity God is that
of thought.
What implications might these diferences between Aristotle and Aqui-
nas have for our understanding of the relationship between the natural
and divine? The non-univocity of unity for Aristotle suggests that it is the
activity of life itself, rather than (for example) the forms that characterize
living hylomorphic composites or those composites themselves, to which
God should be most directly connected in the natural world. A focus on
divine simplicity as such, by contrast, suggests a connection with whatever
in the natural world is incomposite or lacking in parts, notably form.
Indeed, Aquinas winds up calling God a form (ST I q. 3 a. 2). Aquinas is
certainly not alone here; he is following in the tradition of Alexander of
Aphrodisias, who in his Quaestiones called God “a form without matter
and separate” (Alexander of Aphrodisias, Quaestio I.25, 39.9–12). Even
in the later Aristotelian tradition it is not uncommon to refer to God as
a form, due to the incomposite nature of God and form alike. Indeed,
according to Michael Frede,
it seems reasonable to suppose that Aristotle could have thought that

the idea of substance applies primarily to pure substantial forms,
like God, then to substantial forms of natural objects, then to these
objects and to their matter, insofar as it is potentially these objects.
(1983, 71)
But Aristotle chooses to emphasize God’s likeness to the actuality of

the intellectual life, not to form. Though God and natural forms alike are
actualities, for Aristotle other actualities (namely those that constitute life
in its various senses) are closer to the divine than form is. Even with the
qualifcation that merely analogous, not univocal, predication is in place,
Aristotle still would not frst connect the divine unity with a form but
rather with life of a certain sort. As we have seen, Aristotle spends plenty
of time in Metaphysics XII.7 analogizing God’s actuality frst and fore-
most to our thinking, inferior though our rational life may be to God’s.
In Aristotle’s view, our thought is the part of the natural world that most
(though still imperfectly) images the divine, to which the divine should be
likened (though not identifed, due to the constraints of analogy). Central
claims from the De Anima allow us to see what would be next in line in
the string of analogies for Aristotle: other life activities beyond that of
the rational life, such as the activities of sensation and then of nutrition
(De Anima II.2). Indeed, as Aristotle puts it in De Anima II.4, even plants
and animals through reproduction “participate in the eternal and divine”
(415a30–31). We have seen that the very frst descriptor of the divine
activity which Aristotle gives is to call it life (διαγωγὴ) (1072b15). In this
way, we fnd in Aristotle a very diferent picture of what within the natu-
ral world best images the divine than the picture Frede lays out when he
privileges natural forms. Even given the qualifcation that our predicates
apply to the divine only via analogy, the best way to understand God qua
unity is to understand God qua a specifc sort of actuality rather than to
understand God qua simple or incomposite (despite the fact that those
descriptors do apply to Aristotle’s God), and the implication is that it is
life (most of all the life of thought), rather than the forms of hylomorphic
objects, that best images the divine.
We can further illuminate Aristotle’s connection between God and our
life of thought (and other sorts of life in the natural world) by highlighting
the fact that God and life alike, for Aristotle, are actualities in the special
sense of energeia – that is, in the sense of an actuality that contains its
goal (telos) in itself. Such actualities are contrasted with actualities that
do not contain their own goals:
[T]hat in which the end is present is an action (ἐνέργεια). E.g. at the

same time we are seeing and have seen, are understanding and have
understood, are thinking and have thought: but it is not true that at
the same time we are learning and have learnt, or are being cured and
have been cured . . . we are living and have lived.
(Metaphysics IX.6, 1048b22–28).
Aristotle’s everyday examples of this special type of actuality are seeing,

understanding, thinking, and living in general. We do not engage in the
activities of thinking or, more generally, any sort of living in order to
attain some goal beyond the process of thinking or living; these processes
are their own goals. Whereas while we are being cured we cannot yet say
that we have been cured (i.e. that we have attained the ultimate goal of
the process), while we are living we can already say that we have lived. In
other words, we attain the goal of life continuously while we live; thus liv-
ing is an energeia while curing is not. From these criteria we can conclude
that a natural form such as soul (the form of living creatures) will not
count as an energeia. For Aristotle tells us in De Anima II.1 that the soul
is an actuality merely of the frst kind, that is, analogous to knowledge
possessed rather than to fully active refecting (412a24–26). As such, soul
is an actuality that still reaches out toward a further actuality beyond
itself as a goal. The actuality toward which it reaches is life in a fully
active sense (corresponding to refecting rather than to knowledge). Life
itself, by contrast with soul, contains its own goal and is thus, like God,
an energeia. The concept of energeia thus highlights not form, but rather
life – most of all the life of thought – as that which, within the natural
world, best images the divine.
To sum up the argument of this section, divine simplicity, of course,
has played a signifcant role both in historical fgures before and after
Aquinas and in more recent discussions in philosophy of religion, and for
good reason. Indeed, divine simplicity is implied by parts of Aristotle’s
own views. However, I have argued that if we want to remain fully true
to Aristotle, we should not use divine simplicity as an explanation of
divine unity – not even with the constraint of analogical predication in
place. Even if we agree that our terms apply only analogically to God,
there is still a question of which of our terms deserves central emphasis in
illuminating the nature of God, including the nature of God’s unity (to the
extent that it can be illuminated). On that question Aristotle and Aquinas
diverge, and that divergence afects precisely what connections we might
be tempted to draw between the natural and divine realms.
Taking divine simplicity as the core criterion for divine unity can tempt
us to connect God frst and foremost with other indivisible, simple objects
in the natural world (such as forms); if we follow Aristotle, by contrast,
we are invited frst to connect divinity with those items in the natural
world that are most analogous to the fully specifed nature of the divine
actuality – not incomposite, immaterial forms but rather a certain life
activity, the life of thought. In Aristotle’s picture divine simplicity is not so
much the reason for or explanation of divine unity; it is an implication of
divine unity. It is life itself, moreso than indivisibility or lack of composi-
tion, that best illuminates the nature of divine unity.
6. Conclusion
Early in the Metaphysics Aristotle warns us,
[t]he hardest inquiry of all, and the one most necessary for knowledge
of the truth, is whether being and unity are the substances of things,
and whether each of them, without being anything else, is being or
unity respectively, or whether we must inquire what being and unity
are, with the implication that they have some other underlying nature.
(III.4, 1001a4–9)
We have seen that his own answer, of course, is that being and unity just
as such are not the substances of things; rather, these terms are said in
many ways. Unity itself is divided into potentiality and actuality, with
these terms themselves having further senses. Since actuality is prior to
potentiality (Metaphysics IX.8), a full understanding of unity will there-
fore require identifying the senses of actuality related to each other via
analogy, including which of them is primary (God’s life of thought think-
ing itself), and mapping out the analogies between them.
I have argued that if lack of ontological composition (mediate or imme-
diate) were the source of genuine unity, then God would count as a unity
in the same sense as hylomorphic beings themselves (on the conceptual
distinction model) or in the same sense as the forms of hylomorphic beings
(on a certain way of implementing the ontological distinction model).
But for Aristotle, I have argued, no criterion for unity applies univocally
between divine unity and unities in the natural world; unity is not univo-
cal between actual and potential unities, and actuality itself in turn is
not univocal.9 The nature of unity for Aristotle is thus not captured by a
generic formula (e.g. concerning indivisibility or lack of parts), but rather
by mapping relationships (of priority, similarity, etc.) between the fully
determinate senses of unity and their connection – near or remote – to
God’s life of active thought.
This implication tells against those accounts of hylomorphic unity that
defne hylomorphic unity in terms of a feature that is univocally shared
with the divine: being either ontologically simple or properly related to
something (form) that is ontologically simple. If we retain the focus on
simplicity or indivisibility as a (mediate or immediate) criterion for genu-
ine unity, as for example in the premise that hylomorphic composites
are unities because of their lack of ontological composition, we can then
attribute unity to God in precisely the same sense as we attribute it to
hylomorphic composites: God is a genuine unity because God is simple
and indivisible, as highlighted by the concept of divine simplicity. Simi-
larly, if we maintain that hylomorphic composites are unities because they
have a part that lacks ontological composition, namely form, we can then
attribute unity to God in the same sense as we attribute it to the forms of
hylomorphic composites. Thomas Aquinas avoids this objection, despite
the fact that he does argue for divine unity by appeal to divine simplicity
or lack of composition, due to his insistence that all our predicates apply
to God only via analogy. However, his chosen method for illuminating
the unity of God still focuses on God’s simplicity or lack of composition,
leading to emphases that difer from Aristotle’s on the question of what,
within the natural world, most closely images the divine. Like Aquinas,
Alexander of Aphrodisias as well as Michael Frede in the more recent
Aristotelian tradition can be found characterizing the divine unity as a
form.
Divine simplicity is of course an important component of Aristotle’s
own view about God; however, even with the constraint of analogous
predication in place, Aristotle would not illuminate the nature of divine
unity by reference to or in terms of divine simplicity. Rather, he does so
by asking what specifc kind of unity in the natural world is most directly
analogous to the divine unity or best images it. He ultimately connects it
frst and foremost to the life of thought, and thereafter to life in its various
senses, rather than to forms. For Aristotle, what most likens something to
the divine unity is not its lack of composition or indivisibility but rather
its life.10
Notes
1. Marmodoro’s project here is to articulate Aristotle’s own views; however she
has, in later work, rejected certain aspects of Aristotle’s views and defended
the view that the unity characteristic of substances is something that results
from conceptual individuation (2018).
2. Of course, proponents of this view do not deny any distinction whatsoever
within hylomorphic beings; they afrm, for example, that for Aristotle such
beings do have distinct bodily organs and other parts. After all, the distinc-
tion between bodily traits is posterior to the hylomorphic unity of the whole
living thing. As we know from Aristotle’s De Anima, the distinct bodily parts
of a hylomorphic being are dependent upon the unifed life of that being,
rather than the other way around. The point, then, is not to outrightly deny
any ontological distinction within hylomorphic beings, but rather to deny
that any ontological distinction is to be found between the matter and form
defnitive of their unity.
3. I have argued that this general view need not commit to the commonly held
claim that the parts of the hylomorphic composite, matter and form, have
their identity and diversity prior to the composite. One could maintain that
hylomorphic composites incorporate distinct matter and form within them-
selves, while maintaining that either or both of these components attain their
unity and diversity either derivatively from or at least coeval with (rather than
prior to) the hylomorphic composite (Peterson 2017).
4. As we have seen, hylomorphic substances do have ontologically distinct parts
(e.g. multiple bodily organs) while God has none; however, these parts are
not components of the unity as such, in the way that matter and form are.
Rather, they presuppose that unity. More specifcally, the conceptual distinc-
tion model holds that once we have a hylomorphic unity in place (even if
distinct matter pre-existed it), no distinction remains between the unity of its
matter and the unity of its form.
5. Translations of Aristotle’s texts are from Barnes (1984), with slight modif-
cation. Barnes uses “fulfllment” to translate ἐντελέχεια and “actuality” to
translate ἐνέργεια. Here I use actuality for both terms, simply because the
distinction between them will not concern us here, and it is more common
in other translations to use “actuality” for ἐντελέχεια. Aristotle connects the
two Greek terms at 1047a30.
6. Proponents of these criteria for unity might, of course, still maintain that God
is a unity in a higher sense or to a greater degree than natural objects (and
they usually do make such a claim; for example, even those who maintain
the ontological indivisibility of hylomorphic composites can still claim that
God is a unity in a higher degree because God lacks defnitional as well as
ontological composition, while hylomorphic composites have defnitional
composition). My point is just that the criterion for ultimate unity applied
to hylomorphic objects on these accounts applies univocally to God; it does
not apply to God in purely analogous senses, as it should, given Aristotle’s
non-univocal account of unity.
7. Therefore, it calls into question the implementations of the conceptual dis-
tinction model described by Scaltsas, Marmodoro, and Gill, and the imple-
mentation of the ontological distinction model described by Koslicki and
Haslanger, but not the implementation of the ontological distinction model
described in Loux (1995) or the alternate one described in Peterson (2018).
Since the views discussed in detail here by no means canvas all the possibili-
ties, my goal is to have laid out a constraint or standard that can be used in
evaluating diferent ways of applying criteria for genuine unity to hylomor-
phic composites.
8. Citations from Aquinas are taken from the Summa Theologiae, trans. Domin-
ican Fathers of the English Province (2008).
9. Rather than being related to each other by one primary sense on which all
others depend (in the way that the accidental categories depend on the cat-
egory of substance), the senses of actuality are instead related via analogy.
10. I would like to thank the editors of this anthology for their very helpful
feedback in revising central claims of my argument. I would also like to thank
Brandon R. Peterson for his very insightful feedback on framing my argument
and clarifying key parts of it.
References
Alexander of Aphrodisias. (1992) Quaestiones. Translated by R.W. Sharples.
Ithaca, NY: Cornell University Press.
Aquinas, Thomas. (2008) Summa Theologiae, vol. 1. Translated by Dominican
Fathers of the English Province. Scotts Valley, CA: NovAntiqua.
Aristotle. (1984) The Complete Works of Aristotle: The Revised Oxford Transla-
tion, 6th edition (two vols.). Jonathan Barnes (Ed.). Princeton: Princeton Uni-
versity Press.
Fazzo, Sylvia. (2016) “Unmoved Mover as Pure Act or Unmoved Mover in Act?
The Mystery of a Subscript Iota,” in Christoph Horn (Ed.), Aristotle’s Meta-
physics Lambda – New Essays (pp. 181–205). Boston/Berlin: De Gruyter.
Frede, Michael. (1983) “Individuals in Aristotle,” in Essays in Ancient Philoso-
phy (pp. 49–71). Minneapolis: University of Minnesota Press.
Gill, Mary Louise. (1989) Aristotle on Substance: The Paradox of Unity. Prince-
ton: Princeton University Press.
Haslanger, Sally. (1994). “Parts, Compounds, and Substantial Unity,” in Theo-
dore Scaltsas, David Charles, and Mary Louise Gill (Eds.), Unity, Identity, and
Explanation in Aristotle’s Metaphysics (pp. 128–170). Clarendon: OUP.
Herzberg, Stephan. (2016). “God as Pure Thinking. An Interpretation of Meta-
physics Λ7, 1072b14–26,” in Christoph Horn (Ed.), Aristotle’s Metaphysics
Lambda – New Essays (pp. 157–180). Boston/Berlin: De Gruyter.
Koons, R. (2014). “Staunch vs. Faint-hearted Hylomorphism: Toward an Aristo-
telian Account of Composition,” Res Philosophica, 91(2):151–177. http://doi.
org/10.11612/resphil.2014.91.2.1
Koslicki, Kathrin. (2006). “Aristotle’s Mereology and the Status of Form,” The
Journal of Philosophy 103(12):715–736.
Kosman, Aryeh. (2013). The Activity of Being. Cambridge, MA: Harvard Univer-
sity Press.
Lewis, Frank. (1994). “A Thing and its Matter,” in Theodore Scaltsas, David
Charles, and Mary Louise Gill (Eds.), Unity, Identity, and Explanation in Aris-
totle’s Metaphysics (pp. 247–277). Clarendon: OUP.
Loux, Michael. (1995). “Composition and Unity: An Examination of Metaphys-
ics H.6,” in May Sim (Ed.), The Crossroads of Norm and Nature: Essays on
Aristotle’s Ethics and Metaphysics (pp. 247–279). Lanham: Rowman and
Littlefeld.
Loux, Michael. (2012). “Substances, Coincidentals, and Aristotle’s Constituent
Ontology,” in Chris Shields (Ed.), The Oxford Handbook of Aristotle (pp. 372–
399). Oxford: OUP.
Marmodoro, Anna. (2013). “Aristotle’s Hylomorphism without Recondition-
ing,” Philosophical Inquiry 36(1–2):5–22.
Marmodoro, Anna. (2018). “Whole, But Not One,” in J. Heil, A. Carruth, and
S. Gibb (Eds.), Ontology, Modality, and Mind: Themes from the Metaphysics
of E.J. Lowe (pp. 60–72). Oxford: OUP.
Peterson, Anne Siebels. (2017). “The Primacy of the Organism: Being, Unity, and
Diversification in Aristotle’s Metaphysics,” Review of Metaphysics
70(4):645–661.
Peterson, Anne Siebels. (2018). “Unity, Plurality, and Hylomorphic Composition
in Aristotle’s Metaphysics,” Australasian Journal of Philosophy 96(1):1–13.
Scaltsas, Theodore. (1994). “Substantial Holism,” in Theodore Scaltsas, David
Charles, and Mary Louise Gill (Eds.), Unity, Identity, and Explanation in Aris-
totle’s Metaphysics (pp. 107–128). Clarendon: OUP.
13 Natural and Supernatural
Edward Feser
The intersection of metaphysics and theology is perhaps nowhere clearer

than where questions about the natural and the supernatural arise. And
it is in its understanding of the natural and the supernatural that the Aris-
totelian-Thomistic (A-T) tradition in metaphysics and theology (which I
will be commending in this chapter) most clearly difers not only from
philosophical naturalism, but also from the assumptions of many contem-
porary theologians and analytic philosophers of religion. In this chapter,
I ofer an exposition of the A-T position that lays bare these diferences,
treating, along the way, notions closely allied to that of the supernatural,
such as creation, miracles, and grace.
Some False Starts

What is it for something to be “natural”? To answer this question, let’s
begin by seeing why some common contemporary answers are inadequate.
Sometimes it is proposed that the term can be defned precisely by contrast
with supernatural phenomena, examples of which are said to be things
like “ghosts, goblins, gods, and other spooky entities . . . tarot-card read-
ings and séances, and other strange ways of fnding out about the world.”1
This won’t work. For one thing, this defnition of “natural” isn’t very
helpful unless we already have a defnition of “supernatural,” and this
proposal doesn’t give us that. It merely ofers some purported examples
of “supernatural” phenomena. Nor are the examples good ones. For
instance, suppose it turned out that goblins are real. Exactly why would
they count as “supernatural,” as opposed to being unusual and poorly
understood but still perfectly natural creatures? To say that they are
“spooky” or “strange” is no answer. There are purely natural phenomena
that once seemed strange or spooky (such as gorillas, once regarded as
mythical, or gravitation, once regarded as an “occult” force) or that still
do seem strange or spooky (such as quantum phenomena). Indeed, the
conception of the “natural” that some philosophical naturalists operate
with is so broad that “there are many immaterial things that [they] are
DOI: 10.4324/9781003125860-17
344 Edward Feser
perfectly happy to countenance: for example, concepts and numbers.”2
Moreover, even if the strangeness or spookiness of some phenomenon was
sufcient to keep it from being natural, that would not sufce to make it
supernatural as opposed to preternatural (a notion we’ll return to later).
For another thing, the literal meaning of “supernatural” is, of course,
“above or beyond nature.” Hence, however we fesh out the notion of the
supernatural, it is bound to presuppose the concept of the natural. So, it
puts the cart before the horse to try to defne “natural” by reference to
“supernatural.”
Another common approach is to defne “natural” by contrast with
“artifcial.”3 One problem with this is that in order for it to be useful, we
need to know what “artifcial” means, and here too we have a notion that
in fact seems to presuppose the concept of the natural. True, a thing is
said to be artifcial when it is made by human beings. But that is at most a
necessary rather than sufcient condition – otherwise, babies, saliva, tears,
and other things that human beings routinely produce just by virtue of
being the kinds of organisms we are would count as “artifcial,” and they
don’t. Something is said to be artifcial when it is made by human beings
as opposed to occurring naturally. Hence the concept of the “natural” is
prior to the concept of the “artifcial,” just as it is prior to the concept of
the “supernatural.” It simply gets things the wrong way around to try to
defne “natural” in terms of either of them.
There is also the problem that human beings are themselves parts of
the natural world, and that in making the artifcial things they do they are
just doing what comes naturally to them. Hence, without some prior and
independent characterization of the “natural,” it isn’t even clear exactly
how the artifcial is distinct from the natural.
A third proposal is to characterize the “natural” by way of contrast
with the normative.4 This refects the Humean thesis that there is a sharp
divide between facts and values, between the descriptive and the prescrip-
tive, and that to suppose otherwise is to commit a “naturalistic fallacy.”
The “natural,” on this view, is what is inherently non-normative. One
problem with such a proposal is that it tells us at most what the natural
is not, but not what it is. Another is that it is tendentious fatly to assert
that the natural is inherently non-normative; certainly the Aristotelian
tradition would disagree.5
But that brings us to what is probably the most widely accepted con-
ception of the “natural” in contemporary philosophy, a conception that
might seem to solve both of these problems with the proposal that the
natural is inherently non-normative. This is the idea that “it would do best
to defne ‘natural’ as ‘what is recognized by natural science.’”6 This might
seem to give us a positive characterization of the “natural” insofar as sci-
ence certainly makes positive existential claims rather than merely telling
us what does not exist. It might also seem to justify the supposition that
the “natural” is inherently non-normative, insofar as it is widely thought
Natural and Supernatural 345
that science excludes inherently normative notions from its description of
the world. For example, Alex Rosenberg tells us, “ever since physics hit
its stride with Newton, it has excluded purposes, goals, ends, or designs
in nature.”7 To be sure, even many philosophers who share Rosenberg’s
naturalism would hold that sciences other than physics reveal truths about
nature that cannot be entirely captured in the language of physics, and
thus would disagree with Rosenberg’s additional thesis that “physics is
the whole truth about reality.”8 Still, most of them would also hold that
any purportedly normative notions that feature in these other sciences
(such as the notion of function in biology) can and ought to be analyzed
in non-normative terms.
However, this won’t work either. For one thing, as it stands, the pro-
posal that “it would do best to defne ‘natural’ as ‘what is recognized
by natural science’” is obviously circular, since to understand the phrase
“natural science” presupposes that we already know what “natural”
means. For another, even if we simply deleted the modifer and suggest-
ing defning “natural” as “what is recognized by science,” the proposal
is still a nonstarter, since we could have the concept of what is “natural”
whether or not we had the concept of “science” (indeed, the notion of
what is “natural” predated the rise of science, certainly as Rosenberg and
other naturalists understand what counts as science). Even if it turned out
that the referents of the expressions “natural” and “what is recognized by
science” were identical, this would no more show that “natural” means
“what is recognized by science” than the fact that the morning star is
identical to the evening star entails that the phrases “the morning star”
and “the evening star” have the same sense.
Another way to see the problem is to note that the proposal in question
appears to confate metaphysical and epistemological questions. Whatever
else a “natural” phenomenon is, it is supposed to be a kind of reality.
Science, meanwhile, is a method for coming to know about reality. Once
we know what it is for something to be “natural,” science can no doubt
help us determine which specifc things are in fact natural and which are
not. But that doesn’t mean that to be natural just is to be the sort of thing
known or “recognized” by science. To suppose otherwise is like supposing
that it would be best to defne “star” or “planet” as “the sort of thing
typically studied with a telescope.”9
Aristotle on Phusis
It might seem ironic that contemporary philosophers’ attempts to defne
“natural” should prove so inadequate, given that as contemporary phi-
losophers they tend to be committed to naturalism, and as philosophers
they are engaged in an enterprise that got its start with the Pre-Socratics’
investigations into nature. But from an Aristotelian point of view, their
naturalism is precisely part of the problem and has led them to recapitulate
346 Edward Feser
mistakes made by the Pre-Socratics. Aristotle’s account of nature in Book
II of the Physics avoids the difculties with the proposals considered so
far, while capturing what is plausible in them. It begins as follows:
Some things exist by nature, others are due to other causes. Natural
objects include animals and their parts, plants and simple bodies like
earth, fre, air and water; at any rate, we do say that these kinds of
things exist naturally. The obvious diference between all these things
and things which are not natural is that each of the natural ones
contains within itself a source of change and of stability, in respect of
either movement or increase and decrease or alteration. On the other
hand, something like a bed or a cloak has no intrinsic impulse for
change – at least, they do not under that particular description and to
the extent that they are a result of human skill, but they do in so far
as and to the extent that they are coincidentally made out of stone or
earth or some combination of the two.10
So, what exists by nature (phusis) is, frst of all, something that “con-
tains within itself a source of change and of stability,” something which
possesses an “intrinsic impulse for change.” For example, consider a liana
vine, which is the kind of vine Tarzan famously swings around on. Such a
vine will sink roots into the ground, grow a very long stem, climb up a tree
toward the forest canopy, and exhibit other characteristic attributes and
tendencies. These attributes and tendencies are intrinsic to it in the sense
that nothing outside the vine has to make it exhibit them. It will simply
do so if left to itself. Contrast this with a hammock that Tarzan might
construct by tying together several living liana vines. The hammock will
also exhibit certain characteristic attributes and tendencies, such as being
large, stable, and comfortable enough to allow Tarzan to take a nap in it.
But these attributes and tendencies are not intrinsic to the vines. For one
thing, the vines would never have gotten into the confguration typical
of a hammock had Tarzan not make them do so. For another, they are
not likely to stay in that confguration unless Tarzan periodically reties
them, prunes them, and so forth. Common sense would say that when
the vines sink roots, grow toward the forest canopy, and so on, they are
doing what comes naturally to them, whereas when they function as a
hammock they are not doing what comes naturally but rather being made
to do so artifcially. Aristotle’s account both captures this commonsense
distinction and explains it as grounded in the distinction between intrinsic
versus extrinsically imposed attributes and tendencies.11
That there are things that are natural in this sense is just obvious, Aris-
totle says, and not in need of argument:
It would be absurd, however, to try to prove that nature exists, since

it is evident that there do exist many things of this sort. To rely on the
non-obvious to establish the obvious is a sign of being incapable of
distinguishing between what is and what is not intelligible in itself . . .
Inevitably, then, people in this kind of situation argue only at the
verbal level, but do not understand anything.12
I would suggest that what Aristotle is saying here is that there is a kind
of incoherence in denying the reality of things that are natural in the
sense in question. For if nothing is natural in this sense, then everything
is artifcial. Yet the artifcial itself presupposes the natural. For example,
you can make hammocks (which have extrinsically imposed tenden-
cies toward the changes and stability characteristic of hammocks) only
because there already exist liana vines or other natural materials (which
have intrinsic tendencies toward the changes and stability characteristic
of them). Things can be non-natural or artifcial only because there are,
more fundamentally, things that are not – things that are just naturally a
certain way. A skeptic might verbally contend otherwise, but his words
will be without sense. The reality of nature in this sense is thus simply a
bedrock part of the conceptual framework within which we make sense
of the world.
We can, however, say much more about the metaphysical structure of
natural phenomena. Aristotle says that what have natures are substances
rather than attributes. An attribute “is not a nature, and does not have a
nature either, but is due to nature.”13 For example, in the case of a liana
vine, what has a nature or intrinsic source of change and stability is the
vine itself. The characteristic growth pattern of the vine does not itself
have a nature in the strict sense, but rather is due to the nature of the vine,
in the sense of being a consequence of the vine’s having the nature it does.
Aristotle also says, “form is a more plausible candidate for being nature
than matter is.”14 The nature of the liana vine, for example, is to be found
primarily in the distinctive way its matter is organized rather than in the
matter itself (albeit the fact that a vine is a material thing also refects its
nature, so that Aristotle does not deny that “‘nature’ refers . . . both to
form and to matter”).15 Moreover, “a thing’s nature involves purpose.”16
Indeed, “the nature of a thing is its end and its purpose” and “what a
thing is and its purpose are the same.”17 To be the kind of thing that is
naturally directed toward ends like sinking roots, developing a very long
stem, growing toward the forest canopy, etc. just is to be a liana vine.
As this indicates, when fully articulated, Aristotle’s account of nature
is bound to incorporate such classic Aristotelian notions as those of
substance, proper attribute, active and passive powers, substantial form,
prime matter, and fnal cause. Qua substance, a natural object (such as
a liana vine, an animal, water, etc.) is going to be a compound of sub-
stantial form and prime matter. In this it will difer from an artifact (such
as a hammock, a bed, a cloak, etc.), which has only an accidental form
rather than a substantial form. Certain properties or proper attributes
348 Edward Feser
will fow or follow from its nature, and this will include active causal
powers (such as a liana vines’ ability to sink roots into the ground) and
passive powers or liabilities (such as the vines’ susceptibility to being
tied together). By virtue of its powers, a natural object will be intrinsi-
cally directed toward the realization of certain purposes or fnal causes
(such as the vines’ taking in of water through the roots and growing
upward toward sunlight). This intrinsic fnality difers from the extrinsic
or externally imposed fnality of an artifact (such as the end of function-
ing as a hammock, which is not intrinsic to the liana vines but exists
only relative to Tarzan’s intentions).18
Aristotle takes his account to correct some errors of his Pre-Socratic
predecessors. For example, he says, “if we look to the ancient thinkers,
natural science would seem to be concerned with matter [alone] (for it
was only very slightly that Empedocles and Democritus touched on form
and essence).”19 He is also very keen to criticize the suggestion that appeal
to fnal causes is unnecessary and that efcient causal necessity alone is
sufcient to account for the behavior of natural objects.20 In other words,
Aristotle insists that formal and fnal causes, and not just material and
efcient causes, are essential to natural phenomena.
Now, the mechanical world picture that displaced Aristotle’s concep-
tion of nature in the early modern period was defned in part by a rever-
sal of this judgment of Aristotle’s and a return to the atomist project of
explaining all of nature in terms of causal relations between fundamental
particles of matter.21 And as the remarks from Rosenberg quoted earlier
indicate, contemporary naturalists are essentially committed to the same
project, however diferent in points of detail their explanations are from
those of the ancient and early modern atomists. Naturally, then, the Aris-
totelian will reject the proposals, considered previously, that the “natu-
ral” is to be identifed with the non-normative, or that it is exhausted by
what modern physics describes in a non-teleological way.
As Aristotle’s distinction between natural objects and things like beds,
cloaks, and the like indicates, the Aristotelian will agree with the proposal
that the “natural” can be understood by contrast with the “artifcial.”
However, Aristotelians will nevertheless insist that the concept of the
natural is prior to that of the artifcial and can be defned independently
of it. The contrast between natural and artifcial objects elucidates the
former, but it is not essential. What is essential is the idea that natural
objects have an intrinsic source of activity.
It goes without saying that the Aristotelian account of nature, whatever
its merits compared to the contemporary conceptions of the “natural”
criticized earlier, is controversial.22 But my aim here is not to defend that
account. I have done that at length elsewhere.23 The present aim is simply
to explain what the Aristotelian account is, so that we might understand
the role it plays in the Aristotelian-Thomistic theology of the natural and
the supernatural.
Divine Conservation and Concurrence
We have seen that while a thing’s nature is, for Aristotle, primarily to
be understood by reference to its form, there is also a sense in which its
matter is part of its nature. Now, the term “essence” is often used as a syn-
onym for “nature.” Aquinas, building on Aristotle, insists that the essence
of a physical object includes both its form and its matter.24 Stones, trees,
and dogs are, after all, essentially material substances, so that their matter
can hardly be less a part of their essences than their substantial forms are.
Aquinas goes beyond Aristotle, though, with his famous real distinction
between a thing’s essence and its existence.25 A lion, a Tyrannosaurus,
and a unicorn too (we’ll suppose for the sake of argument) each has an
essence, and in each case that essence includes matter, since they are all
physical substances. But while lions exist, Tyrannosauruses no longer do,
and unicorns never did. Existence must be added to a thing’s essence in
order for that thing to be part of the real world.
Now, Aquinas makes of this thesis the starting point of an argument for
the existence of God as frst cause of the existence of things.26 He under-
stands this to be a matter, not of God merely having gotten the natural
world going at some time in the past, but rather of keeping it in being at
every moment. At any time at which any natural object exists, it does so
because God is imparting existence to its essence, and if he were not doing
so the object would cease to be. This is known as the Doctrine of Divine
Conservation, and for Aquinas, it reveals to us the fundamental way in
which God is creator of the natural order. While Aquinas believes that the
world had a beginning in time, he does not think that that supposition
is essential to the idea of creation. Even if the natural world had always
existed (as Aristotle supposed), it would still have to have been always
conserved in being by God.
Aquinas is also committed to the principle “agere sequitur esse” (or
“action follows being”), according to which the way a thing acts refects
what it is. An implication of this is that, if a thing is of the kind that
cannot persist in being without divine assistance, neither can it act with-
out divine assistance. This is the Doctrine of Divine Concurrence. For
Aquinas, not only can a lion not continue in existence even for a moment
without God conserving it in being, but also it cannot walk, or breathe,
or hunt, or do anything else, without God concurring with these activities
by imparting causal power to it at every moment.27
For Aquinas, then, the entire natural order depends at every moment
for its continued existence and operation on divine causality. That does
not entail that natural objects lack causal power. The notion of divine
concurrence is not to be confused with the occasionalist thesis that God
alone acts and natural objects are inert. Natural objects do have genuine
causal power, but it is secondary causal power, borrowed or derived from
the divine primary cause.28 When you use a stick to push a stone, the
350 Edward Feser
stick really does move the stone, and its distinctive nature makes a real
contribution to the efect. For example, if the stick is solid and durable,
it will yield a smooth and continuous motion in the stone as long as you
keep pushing it, whereas it would not do so if it were limp or brittle. All
the same, the stick would not be able to move the stone at all if you were
not using it as an instrument and thereby imparting to it the power to
move the stone. Similarly, natural objects in general are genuine causes
and their natures make a real contribution to the natures of the efects
they produce, but according to the Doctrine of Divine Concurrence, they
would produce no efects at all if God were not continuously imparting
causal power to them.29
Now, for Aquinas, these doctrines apply not only to physical objects,
but to incorporeal things as well. In particular, they apply to angels,
understood as immaterial intellectual substances – entities that think
and will but have no corporeal attributes or activities at all. To be sure,
because an angel is devoid of matter, it is not susceptible of perishing in
the way that a physical object is. Since a physical object is a composite
of prime matter and substantial form, and matter is always capable of
taking on some other form, it is possible for a physical object to go out
of existence if something acts on its matter in a way that causes it to
lose its substantial form. Angels, lacking matter, are not subject to such
destructive forces. Still, there is in an angel too a real distinction between
its essence and its existence, so that it too requires a divine cause to
impart existence to it. Thus, were God to cease imparting existence to it,
an angel would be annihilated no less than a physical object would be.
Moreover, given that agere sequitur esse, an angel, no less than a physi-
cal object, could not act at all without the divine frst cause imparting
secondary causal power to it. What is true of angels is also true of human
souls, which Aquinas takes to be incorporeal and thus capable of surviv-
ing the death of the body as incomplete substances. Though immortal,
they too nevertheless depend for their existence and operation on divine
conservation and concurrence.
So, the created order is not exhausted by the order of physical objects
but includes immortal souls and angelic intellects. And the latter, given
their immateriality, are by nature immortal or imperishable. Moreover,
since agere sequitur esse, these incorporeal substances, since they can exist
apart from matter, can also operate apart from matter. For example, an
angel needn’t have sense organs or bodily limbs in order to know about
the world or act on it. This gives it a broader range of action than a mate-
rial substance has. For example, a nonhuman animal cannot know about
or act upon an angel, but an angel can know about and act upon nonhu-
man animals (as in the incident of the possession of the Gadarene swine
in the Gospel of Matthew 8:28–34). All the same, no part of the created
order can exist or operate even for a moment without divine conservation
and concurrence.
The dependence of the created order on God involves a causality that
is so radical that it cannot, in Aquinas’s view, be delegated. That is to
say, only God can create, in the strict theological sense. Other kinds of
making involve acting on what already exists and modifying it in some
way. Creation is not like that. It is not a matter of making something that
is already there to be this way rather than that way (as when preexisting
prime matter is made to lose one substantial form and take on another).
Rather, it is a matter of making it the case that there is anything there
at all rather than nothing. And Aquinas holds that only that which has
existence in an absolutely unlimited and underived way can do that.30
Preternatural Versus Supernatural

It is only in light of these ideas that we can correctly understand the super-
natural as Aquinas and other Thomistic theologians conceive of it. Let’s
begin, for purposes of contrast, with some examples of phenomena that
are not supernatural. When a liana vine grows toward the forest canopy,
it is obviously acting according to its natural inclinations and in no way
supernaturally. Notice that this is true even given divine conservation and
concurrence. God keeps the vine in being and sustains its causal power,
but in doing so he is not making it act in any way contrary to or even
other than the way it is natural for a vine to act. On the contrary, he is
precisely making it possible for the vine to act naturally.
Now, when the vine functions as a hammock, it is not acting naturally,
but no one would call this “supernatural” either. The reason is that it
obviously involves one part of the natural order (namely Tarzan) using
his own natural powers in a way that interferes with those of another
part. The phenomena are all internal to the natural order, understood
as a system of entities having natures in the Aristotelian sense. And once
again, the fact that God is conserving this order in being and concurring
with the causal activity taking place within it does not make it any less
natural. Rather, divine conservation and concurrence is a precondition of
there being any natural order at all.
Suppose that goblins, werewolves, vampires, and the like existed. Here
too there is no reason to think of such things as supernatural in the sense
of being beyond nature as the Aristotelian understands “nature.” On the
contrary, each of these things would “contain within itself a source of
change and of stability,” would have a substantial form, causal powers
directed toward the generation of certain characteristic efects, and so on.
Moreover, each would be a material substance. Though unusual, their
attributes and activities would be as natural as those of liana vines, lions,
and other more familiar physical objects.
But what about cases of ordinary physical substances being caused by
fallen angels or demons to act in a way they otherwise wouldn’t have (as
in the case of the Gadarene swine), or even in ways that would not be
352 Edward Feser
possible in the natural order of things (as in the case of a demon-possessed
person levitating)? These too would not be supernatural phenomena, as
Thomists understand that notion. For fallen angels too are part of the
broader created order, so that the efects they have on material things
are analogous to the efect Tarzan has on the liana vines, albeit far more
dramatic. For example, causing a person to levitate is analogous to caus-
ing a set of liana vines to function as a hammock. In both cases, natural
objects are being diverted from their natural tendencies by outside forces.
True, the levitation of the person difers from the making of vines into a
hammock insofar as nothing else in the order of physical objects could
bring it about. Still, something else in the larger created order (which
includes incorporeal angelic intellects) can do so.
The traditional way of describing such interferences with the order
of material objects is to characterize them as preternatural rather than
supernatural. Because phenomena like those associated with demonic pos-
session are not possible for a physical substance or even the whole system
of physical substances given the natures they actually have, such phe-
nomena are in an obvious sense beyond the natures of these substances.
At the same time, they do not require any special divine action over and
above God’s ordinary conservation of and concurrence with the created
order – which, again, includes the fallen angels who, given their natures,
are capable of bringing about the unusual phenomena in question. So,
such phenomena are not supernatural in the strictest sense. Calling them
“preternatural” is a useful way of capturing their intermediate status
between the natural and the supernatural.
What is truly supernatural is what would not be possible for anything
in the created order, including even the angelic order, to bring about.
It would be something that only God can bring about by acting in a
way over and above his ordinary conservation of and concurrence with
creation. Now, miracles would be the most obvious example, and I will
discuss those presently. But frst, note that so far, we have been talking
mostly about the order of efcient causes, and what sorts of efects it is
or is not possible for created things to bring about. But there is also an
order of fnal causes or teleology, which concerns the ends toward which
a thing might aim. Here too we can draw a distinction between natural,
preternatural, and supernatural.
Drawing water from the soil through its roots is an end toward which
a liana vine is directed by its nature. Nothing needs to make it aim at
this end. It aims at it just by virtue of being a liana vine. By contrast,
functioning as part of a hammock is not an end toward which a vine is
naturally directed. It has to be imposed on it from outside, by Tarzan. But
since Tarzan is himself part of the same order of physical substances and
imposes this end on the vine in the course of exercising his own natural
powers, there is nothing supernatural or even preternatural about the
end of functioning as a hammock. By contrast, when a demon imparts
preternatural powers to a person it possesses, the ends toward which
these powers aim (for example, levitation) can themselves be said to be
preternatural.31
Now, a strictly supernatural end would be one toward which a created
thing could not possibly be directed apart from special divine action over
and above ordinary conservation and concurrence. The classic example
is the beatifc vision, which involves immediate or non-inferential knowl-
edge of the divine essence. Human beings are by nature capable of know-
ing God in an indirect and inferential way. For example, that is what is
achieved when we develop philosophical arguments for the existence of
God as uncaused cause of the world, and then go on to argue that such a
cause would have to be one, simple, immaterial, immutable, and possess
the other divine attributes. But direct knowledge of God’s nature is not
something we can have given the limitations on our own nature, any more
than liana vines can all on their own aim at functioning as a hammock.
We cannot aim at this end, let alone realize it, unless God directs us to
it as an additional act over and above merely conserving us in being and
concurring with the operation of our natural powers. To posit such an
additional act is what it means to say that grace builds upon nature.
In twentieth-century Catholic theology there arose a famous debate
about the adequacy of this traditional Thomistic conception of nature
and grace, with Henri de Lubac being its most infuential critic.32 The
critics held that the account makes the operation of grace objectionably
“extrinsic” to nature, and nature objectionably self-sufcient and closed
of to God. We should, in the critics’ view, instead regard grace as operat-
ing in a way that is intrinsic to nature, so that human beings are naturally
directed toward the supernatural end of the beatifc vision.
From the traditional Thomistic point of view, the critique is incoherent
and aimed at a straw man. It is incoherent insofar as, if we are naturally
directed toward an end, then of necessity that end is no longer super-
natural. Or if the end really is supernatural, then it is not one to which
we are naturally directed. The critique implicitly (even if unintentionally)
collapses the distinction between natural and supernatural. Now, one way
to do this would be to dissolve the supernatural down into the natural. De
Lubac clearly does not want to do that, since his objection to the notion
of a self-contained “pure nature” is motivated precisely by the worry that
it locks us into nature and blinds us to what is beyond it. The other way
would be to absorb the natural up into the supernatural. But the problem
with that is that the very concept of the supernatural presupposes that of
a natural order that the supernatural is above or beyond, and therefore
distinct from.33
The critique is aimed at a straw man insofar as the Thomistic account
of nature by no means makes it closed of to God. For one thing, the very
existence and operation of the natural order depends at every moment
on divine conservation and concurrence, and this is knowable even by
354 Edward Feser
purely philosophical arguments apart from special revelation. Hence
“pure nature” is precisely the opposite of closed of to God either meta-
physically or epistemically. For another, though human beings do not
naturally have an active power to achieve the beatifc vision, they do
have a passive power or liability to achieve such a supernatural end. That
is to say (and to use the Scholastic jargon), they have an “obediential
potency.” Compare once again the example of the liana vine. It has no
active tendency to become a hammock, but it does have a liability to being
made into one, in a way that other materials do not. (You cannot make a
hammock out of smoke or water, for example.) Similarly, human beings
qua rational animals have a liability or obediential potency for attaining
the beatifc vision, in a way that a nonrational creature does not. Hence
our nature makes a real contribution to our achieving the beatifc vision,
even though the latter is entirely supernatural.34
Miracles and Laws of Nature

At least since David Hume, miracles are often characterized as “viola-
tions” of laws of nature. The conceit is that modern science has revealed
the natural world to be a self-contained machine, with the facts about
its microstructural parts at any time together with the physical laws that
govern them being sufcient to fx all the facts about it at any later time.
A miracle, on this picture, would be a kind of violent interference with
the operation of this machine, something contrary to its natural workings.
Some have taken this to make the very notion of a miracle theologically
problematic, insofar as it is thought to attribute to God incompetence
comparable to that of a human mechanic who periodically has to tinker
with a machine he couldn’t get right the frst time. It also seems to make
God redundant to the ordinary course of things, insofar as the machine
carries on on its own when he is not interfering with it. Naturally, athe-
ists draw the lesson that God’s existence is a superfuous hypothesis that
Ockham’s razor justifes us in discarding.
Michael Dodds helpfully summarizes fve approaches that modern
theologians have taken to the problem.35 Interventionism simply bites
the bullet and embraces the idea of God as an occasional “lawbreaker.”
Deism holds that though God set the world in operation at its beginning,
it has operated independently of him ever since and he does not interfere
with it by causing miracles. Liberal theology has tended to refrain from
committing itself even to as thin a metaphysics of creation as deism, pre-
ferring instead to ground theology in religious experience rather than in
philosophical arguments for a First Cause. Process theology accommo-
dates divine causality by bringing God into nature rather than interfering
with it from without. Divine limitation theologies take God to limit his
control over and knowledge of the world for the sake of preserving the
freedom of human agents within it.
There are other approaches, too, such as those that look for some
gap in natural causality that God can be said to fll. For example, Alvin
Plantinga suggests that we can fnd such gaps at the quantum level
of nature, and in particular where wave-function collapses occur. He
posits “divine collapse-causation,” by which “God could thus control
what happens at the macroscopic level by causing the right microscopic
collapse-outcomes.”36
From a Thomistic point of view, both the purported problem and the
proposed solutions to it refect deep misunderstandings of nature, divine
causality, physical law, and miracles. First, the natural world is not in the
frst place like a machine on autopilot. Rather, and again, it could not per-
sist in being or operation even for a moment without divine conservation
and concurrence. Hence even the ordinary causal course of things – let
alone miracles, and purported gaps in the causal order – already requires
divine action. But this action is not a matter of interfering with natural
causes, any more than an author interferes with the actions of the char-
acters in a story he has written. Rather, just as the author’s writing is the
precondition of the characters’ acting, divine action is the precondition
of there being any natural causality in the frst place.
Physical laws, meanwhile, are on an Aristotelian analysis simply a
description of the ways physical substances tend to operate given their
natures. As David S. Oderberg puts it, laws of nature are laws “of the
natures of things,” understood as intrinsic sources of things’ activity in the
sense described earlier.37 When God brings about a miracle, this involves
his either adding to physical substances powers the natural order would
not sufce to give them or subtracting from physical substances powers
they would naturally have, by simply temporarily ceasing his ordinary
conservation of and concurrence with these substances and their powers.
For example, when the apostles healed the sick, this involved God impart-
ing to them powers they would not naturally have, and when Daniel
and his friends survived the fery furnace, this involved God temporarily
ceasing to concur with the fames’ power to destroy fesh. In this way, as
Oderberg points out, a miracle is not a violation of the natural order, but
rather a “suspension” of it.38 It is analogous to a government temporar-
ily revoking a law, rather than leaving the law in place and breaking or
violating it.39
Grace and Faith

Adding to nature rather than interfering with it or violating it is, for the
Thomist, also what grace involves. We have already seen how this is so in
the case of God directing us toward the supernatural end of the beatifc
vision. It is true also of his provision of the means to realizing this end,
which involves both remedying the damage of original sin that prevents
our perfect acquisition of the natural virtues (such as justice, courage,
356 Edward Feser
temperance, and wisdom), and adding to them the supernatural or theo-
logical virtues (faith, hope, and charity).
As Aquinas argues, human beings in an uncorrupted state would be
capable, just by virtue of the natural powers that God conserves and
concurs with, of acquiring the natural virtues and of having the indirect
knowledge of Him that our nature directs us toward.40 We would in that
case have needed supernatural assistance only to realize the supernatural
end of the beatifc vision. Compare a pristine liana vine, which is per-
fectly capable on its own of taking in water through its roots, growing up
toward the forest canopy, and so on even if it cannot on its own take on
the form of a hammock.
As a consequence of original sin, however, we are like a liana vine that
has gotten so tangled and knotted that it is choking itself of, preventing
normal growth patterns, keeping water from being efciently distributed
along the length of the vine, and so forth. While such a vine retains its
natural powers, it is manifesting them in a perversely self-stultifying way
that makes it impossible for it to fourish, and it cannot free itself from
this condition. Something from outside would have to untie the vine and
remedy the damage it has done to itself. In a similar way, human beings
are in a damaged state that requires a further operation of grace, one
that restores our ability to manifest our natural powers in a normal and
healthy way.41
The theological virtue of faith deserves special attention in the context
of an account of the natural and the supernatural. As we have seen, for
the Thomist, God’s existence and, to some extent, his nature, are know-
able through purely philosophical arguments and thus by virtue of the
operation of our natural rational powers. However, properly to under-
stand the arguments requires a degree of knowledge and philosophical
acumen that not everyone has, and even those who do have it often fall
into error because of the general damage to the operation of our natural
powers that has resulted from original sin. Moreover, there are truths
about the divine nature, knowledge of which is beyond our natural pow-
ers but would facilitate the realization of our supernatural end.
For these reasons, special divine revelation is necessary. Typically this
involves God sending a prophet and giving him the power to perform mir-
acles so as to confrm the divine provenance of the prophet’s message. What
is revealed might include truths that are in principle knowable via pure rea-
son, but which in actuality have been forgotten, or have gotten mixed with
serious errors, or are known only to a few. Examples would be truths about
divine attributes such as oneness, omnipotence, and omniscience, which are
knowable by philosophical arguments but also revealed in scripture. But the
revelation might also include theological mysteries, which are truths that
are beyond the power of human reason to discover through natural means
or fully to understand even after they are revealed. An example would be
the Trinitarian claim that God is three Persons in one nature.
The notion of a theological mystery might superfcially appear obscu-
rantist, but on refection it is not. For one thing, we should expect there to
be some truths that are not naturally knowable to us just given the limits
of the human intellect. One hardly needs to be a theist to see this. Philo-
sophical naturalists like Noam Chomsky, and Colin McGinn following
his lead, have drawn the same conclusion, even using the term “mystery”
for truths that are not in principle knowable via human reason.42 Sensory
organs are suited to taking in certain specifc kinds of information, but the
very aspects of their natures that make them capable of doing so might
preclude them from being sensitive to other kinds of information. For
example, the structure of the ear both enables it to register sound but ill-
suits it to have the sensitivity to light that the eye has. For Chomsky and
McGinn, there is, especially given the naturalistic assumption that human
beings are entirely the products of random mutation and natural selection,
no reason to rule out the possibility that the human mind is comprised of
special-purpose cognitive organs that suit it to learning some things while
ruling out its learning others. As McGinn puts it, we might well be subject
to “cognitive closure” with respect to some truths. Thomists too afrm
such closure with respect to certain truths, albeit their reasons for doing
so are not exactly the same as those of Chomsky and McGinn.
For another thing, we should also expect the divine nature not to be
entirely knowable to us given what we do know about it through philo-
sophical arguments. For example, Thomists hold that the reasoning that
leads us to afrm the existence of an uncaused cause of the world also
shows, when followed out consistently, that this cause must be one, both
in the sense of being utterly unique and in the sense being absolutely
simple or non-composite. And given his simplicity, there is in God no
distinction between essence and existence or between genus and difer-
entia. Hence he belongs to no species or indeed to any class of things
and is not a mere instance of some essence, not even a unique instance.
But these very facts about God make it difcult for us to form a positive
conception of him – for our mode of understanding a thing is precisely
a matter of identifying the essence of the species it instantiates, which in
turn requires knowing the genus under which that species falls and what
diferentiates the species from others within that genus. Given his simplic-
ity, God’s Trinitarian nature in particular is bound to be opaque to natural
reason even after it is revealed. Our knowledge of the divine nature is,
accordingly, largely negative or apophatic and heavily dependent on an
analogical use of language which stretches it to its limits.43
This is something that Thomists might be better placed to see than many
other philosophers and theologians. William Wainwright notes a “surpris-
ing . . . neglect of mystery” in contemporary philosophy of religion.44
William Alston observes, “one fnds in contemporary Anglo-American
analytic philosophy of religion what looks to be a considerable degree of
confdence in human powers to determine what God is like . . . in some
358 Edward Feser
detail.”45 Richard Swinburne regards even God’s Trinitarian nature to
be something knowable by natural reason.46 I would suggest that this
confdence stems from the tendency of many contemporary analytic phi-
losophers of religion to adopt what Brian Davies has called a “theistic
personalist” conception of God, which rejects the doctrine of divine sim-
plicity.47 Theistic personalists are so labeled because they tend, at least
implicitly, to think of God as the unique member of a species falling
under the genus person, alongside other species such as human persons
and angelic persons. This entails attributing to God an essence distinct
from his existence, an essence of which he is the sole instantiation. Given
our mode of understanding, this makes God much more comprehensible
to us, but – the Thomist holds – at the expense of essentially putting him
within the natural order, broadly construed. In particular, it makes of him
just one further item in the order of things with a nature or essence distinct
from its existence, which would (even if the theistic personalist does not
realize this) therefore require him to have a conserving and concurring
cause from outside this natural order. The theistic personalist’s neglect of
mystery is thus a consequence of this implicit denial of the supernatural
character of the divine (in the Thomistic sense of “supernatural”).
To acknowledge that character is to acknowledge the possibility of
mysteries, which must be specially divinely revealed if we are to have any
knowledge of them. Now, to afrm something based on faith is, for the
Thomist, precisely to believe it because it has been divinely revealed. In no
way is this contrary to reason. For one thing, in order to know that some-
thing really has been specially revealed by God, we need to rely on our
natural powers of reason. In particular, we need to know the “preambles
of faith” – matters of natural theology that establish the possibility of a
special divine revelation, such as arguments concerning the existence and
nature of God, divine conservation and concurrence, and the possibility of
miracles. We also need the “motives of credibility” – empirical evidence,
interpreted through sound philosophical principles, for the conclusion
that a revelation backed by miracles really has in fact occurred. Argu-
ments of both sorts must appeal to premises that need not presuppose
anything revealed, or they would be unable rationally to establish that
anything has in fact been revealed (since they would beg the question).
Faith comes in only after this work is done. Once we know by reason,
independently of faith, that something has in fact been revealed by God,
faith is a matter of afrming it because it has been revealed by him.
Now, sometimes this account of the relationship between faith and rea-
son is criticized on what purport to be Thomistic grounds. For example,
John Jenkins raises three objections to it.48 First, for Aquinas, faith is a
supernatural virtue requiring grace, but the account represents the objects
of faith as the deliverances of our natural rational powers. Second, Aqui-
nas takes the assent of faith to be more frm than any deliverances of those
powers can be. Third, for Aquinas, articles of faith (such as the doctrine
of the Trinity) are precisely not the sort of thing for which arguments can
be given, which is why they have to be revealed.
But rightly understood, the account of faith and reason that I have
described is in no way incompatible with the points Jenkins raises. The last
of his points is based on a confusion. The account of faith I’ve described is
not saying that an article of faith can itself be argued for, and thus needn’t
be known via revelation. Rather, it is saying that though an article of faith
cannot itself be argued for, the thesis that that article of faith has been
divinely revealed can be argued for.
Nor is the account inconsistent with the supernatural character of faith.
For one thing, it does not entail that the object of faith is something
yielded by our natural powers. What our natural powers yield is knowl-
edge of a proposition such as: God has revealed the doctrine of the Trin-
ity. But this (naturally knowable) proposition is not the object of faith.
Rather, the doctrine of the Trinity itself (which is not naturally knowable)
is the object of faith. And it is afrmed because God has revealed it, not
because our natural powers tell us that God has revealed it. The latter is
merely a precondition of faith and not itself the object of faith. It is also
the judgement that God has revealed a doctrine that grounds the frmness
of faith; it is not the judgement that our natural powers tell us that God
has revealed it that grounds it.
Grace operates, frst, by clearing away the obstacles lying in the way
of our natural powers’ yielding the knowledge that God has specially
revealed some doctrine. For example, God might lead a person to become
acquainted with certain powerful arguments for the preambles of faith
and the motives of credibility, might enable the person to overcome preju-
dices and moral vices that make him reluctant to give such arguments a
fair hearing, and so forth. Grace operates, second, by bringing the intel-
lect frmly to assent to a doctrine once natural reason shows that it has
in fact been revealed and to maintain such frm assent in the face of the
vicissitudes of ordinary human cognitive life.49
At no point, however, does faith involve afrming something either
contrary to reason or in the absence of any rational grounds. Faith builds
on and completes the deliverances of our natural rational powers, rather
than subverting them. And this is so precisely because it is a product of
grace, which, as Aquinas famously insists, does not destroy nature but
perfects it.50
Notes
1. Jack Ritchie, Understanding Naturalism (Stocksfeld: Acumen, 2008), p. 2.
Ritchie does not endorse this defnition, but merely notes it as one way the
term “natural” is often understood.
2. Mario De Caro and David MacArthur, “Introduction: The Nature of Natu-
ralism,” in their anthology Naturalism in Question (Cambridge, MA: Har-
vard University Press, 2004), at p. 2.
360 Edward Feser
3. This too is a proposal considered by Ritchie in Understanding Naturalism,
at pp. 3–5. Ritchie develops the idea not only in terms of everyday examples
(such as “orange juice [that] announces that it is 100 per cent natural”) but
also by reference to Arthur Fine’s notion of the “natural ontological atti-
tude,” which is to “try to take science on its own terms, and try not to read
things into science” (quoted from The Shaky Game: Einstein, Realism, and
the Quantum Theory, second edition [Chicago: University of Chicago Press,
1996], p. 149). What is common to the everyday examples and the case that
Fine is interested in seems to be the idea that we grasp what it is for something
to be in its “natural” condition when we consider it “on its own terms” rather
than by reference to some artifcial state human beings put it in or artifcial
use they put it to. I think this is plausible, but then it is the phrase “on its own
terms” that is doing the real work here, rather than the word “artifcial.” The
Aristotelian conception of the “natural” that I will be expounding presently
can be seen precisely as an elucidation of what it is to understand something
“on its own terms.”
4. Once again cf. Ritchie, Understanding Naturalism, at pp. 6–7. The purported
divide between the natural and the normative is also the theme of the essays
in Mario De Caro and David MacArthur, Naturalism and Normativity (New
York: Columbia University Press, 2010).
5. For recent neo-Aristotelian defenses of the view that goodness is a natural
property, see Philippa Foot, Natural Goodness (Oxford: Clarendon Press,
2001); Edward Feser, “Being, the Good, and the Guise of the Good,” in
Daniel D. Novotny and Lukas Novak, eds., Neo-Aristotelian Perspectives in
Metaphysics (London: Routledge, 2014); and David S. Oderberg, The Meta-
physics of Good and Evil (London: Routledge, 2020).
6. Frederick F. Schmitt, “Naturalism,” in Jaegwon Kim and Ernest Sosa, eds., A
Companion to Metaphysics (Oxford: Blackwell, 1995), at p. 343.
7. Alex Rosenberg, The Atheist’s Guide to Reality (New York: W. W. Norton
and Company, 2011), p. 40.
8. Ibid., p. 25.
9. There is also the fact that, as a referee points out, the notion of being “recog-
nized by” natural science could use clarifcation. Presumably it is not intended
to include only what is in fact known to natural science, but also what is in
principle knowable to it. But is it also meant to entail being something know-
able only via natural science? Or might something “recognized by” natural
science also in principle be knowable by other means too?
10. Aristotle, Physics II, 1, 192b8–19. I quote from the translation by Robin
Waterfeld (Oxford: Oxford University Press, 1996), p. 33.
11. Cf. E. J. Lowe’s view in More Kinds of Being (Oxford: Wiley-Blackwell, 2009)
that what distinguishes natural kinds from artefactual kinds is that considered
as an instance of a natural kind, a thing is governed by laws of nature, but
it is not so governed considered as an artifact (pp. 5–6). I discuss the A-T
conception of laws of nature later.
12. Physics II, 1, 193a2–8, at p. 34 of Waterfeld.
13. Physics II, 1, 192b38, at p. 34 of Waterfeld. Emphasis added.
14. Physics II, 1, 193b6, at p. 35 of Waterfeld.
15. Physics II, 2, 194a12, at p. 37 of Waterfeld.
16. Physics II, 7, 198b4, at p. 49 of Waterfeld.
17. Physics II, 2, 194a28 and II, 7, 198a25, at pp. 37 and 49 of Waterfeld.
Emphasis added.
18. In contemporary neo-Aristotelian metaphysics, causal powers have by vir-
tue of their intrinsic fnality or directedness been variously characterized as
nonmental “intentional states” or as exhibiting “physical intentionality” or
“natural intentionality.” Cf. respectively U. T. Place, “Dispositions as Inten-
tional States,” in D. M. Armstrong, C. B. Martin, and U. T. Place, Disposi-
tions: A Debate, ed. Tim Crane (London: Routledge, 1996); George Molnar,
Powers: A Study in Metaphysics (Oxford: Oxford University Press, 2003),
chapter 3; and John Heil, From an Ontological Point of View (Oxford: Clar-
endon Press, 2003), pp. 221–222.
19. Physics II, 2, 194a19–20. Here I quote from the translation by C.D.C. Reeve
(Indianapolis: Hackett Publishing Company, 2018), at p. 23.
20. Physics II.8.
21. See William M. R. Simpson’s chapter in this volume for discussion of the
mechanical world picture.
22. Though, as the references in note 18 indicate, recent metaphysics has seen
renewed interest in Aristotelian teleology. There has also been renewed inter-
est in the Aristotelian notion of causal powers, as evidenced by the work of
contributors to Ruth Grof and John Greco, eds., Powers and Capacities
in Philosophy: The New Aristotelianism (London: Routledge, 2013), and
in Aristotelian essentialism, as evidenced by works such as Brian Ellis, The
Philosophy of Nature: A Guide to the New Essentialism (Chesham: Acu-
men, 2002); Crawford L. Elder, Real Natures and Familiar Objects (Cam-
bridge, MA: The MIT Press, 2004); E. J. Lowe, The Four-Category Ontology
(Oxford: Clarendon Press, 2006); and David S. Oderberg, Real Essentialism
(London: Routledge, 2007).
23. Cf. Edward Feser, Scholastic Metaphysics: A Contemporary Introduction
(Heusenstamm: Editiones Scholasticae, 2014) and Aristotle’s Revenge: The
Metaphysical Foundations of Physical and Biological Science (Neunkirchen-
Seelscheid: Editiones Scholasticae, 2019). See also the chapters in this volume
by William M. R. Simpson and Robert C. Koons.
24. Thomas Aquinas, De Ente et Essentia, chapter 2.
25. Ibid., chapter 5.
26. Ibid. I expound and defend this argument in chapter 4 of Edward Feser, Five
Proofs of the Existence of God (San Francisco: Ignatius Press, 2017).
27. For exposition and defense of these ideas, see Feser, Five Proofs, especially
pp. 174–176 and 232–238.
28. See Ross D. Inman’s chapter in this volume for discussion of the relationship
between divine and creaturely causality.
29. For discussion of issues surrounding divine concurrence, see Alfred Fred-
doso’s invaluable series of articles on the subject: “Medieval Aristotelianism
and the Case Against Secondary Causation in Nature,” in Thomas V. Morris,
ed., Divine and Human Action: Essays in the Metaphysics of Theism (Ithaca,
NY: Cornell University Press, 1988); “God’s General Concurrence With Sec-
ondary Causes: Why Conservation Is Not Enough,” Philosophical Perspec-
tives 5 (1991): 553–585; and “God’s General Concurrence With Secondary
Causes: Pitfalls and Prospects,” American Catholic Philosophical Quarterly
67 (1994): 131–156.
30. Cf. Summa Theologiae I.45.5, Summa Contra Gentiles II.21, and De Potentia
Dei III.4.
31. A referee raises the interesting question whether the preternatural is contrary
to the natural. I would say that what is preternatural need not be contrary to
what is natural, though it can be. To see what I mean, consider once again
the analogy with artifacts. If Tarzan incorporates a certain liana vine into the
structure of a hammock but does so in a manner that does not frustrate the
vine’s realization of its natural tendencies (drawing in water, growing toward
362 Edward Feser
the canopy, etc.) then his making of it part of an artifact is not contrary to
its nature. On the other hand, if he ties the vine in a way that makes it dif-
fcult for water to pass through it, or lops it of altogether, then he would be
using it in a way that is contrary to the realization of its natural tendencies.
Similarly, if an angel were to cause someone to levitate in order to avoid an
oncoming attacker, this would not be contrary to the person’s realization of
the ends toward which his nature directs him, and indeed would facilitate
their realization. But if a demon were to cause a person’s perceptual faculties
to malfunction so as to generate hallucinations, this would be contrary to the
end toward which those faculties are naturally directed.
32. Cf. Henri de Lubac, The Mystery of the Supernatural (New York: Herder and
Herder, 1967).
33. Of course, the supernatural is metaphysically prior to the natural insofar
as God is the cause of the natural order and would exist whether or not he
chose to create it. But the natural is nevertheless epistemically and conceptu-
ally prior to the supernatural, insofar as we know the natural order frst and
come to know the supernatural only by inference from, and contrast with,
the natural.
34. To be sure, the debate between de Lubac and his critics is much more com-
plicated than this brief summary lets on, but the issues are largely theologi-
cal rather than philosophical. John Milbank, The Suspended Middle (Grand
Rapids: Eerdmans, 2005) is a recent book sympathetic to de Lubac, whereas
Steven A. Long, Natura Pura (New York: Fordham University Press, 2010),
and Lawrence Feingold, The Natural Desire to See God According to St.
Thomas Aquinas and His Interpreters, second edition (Naples, FL: Sapientia
Press of Ave Maria University, 2010), are two recent books critical of de
Lubac.
35. Michael J. Dodds, O.P., Unlocking Divine Action: Contemporary Science and
Thomas Aquinas (Washington, DC: Catholic University of America Press,
2012), chapter 3.
36. Alvin Plantinga, Where the Confict Really Lies: Science, Religion, and Natu-
ralism (Oxford: Oxford University Press, 2011), p. 116. Infuential recent
discussions of quantum models of divine action include George F. R. Ellis,
“Ordinary and Extraordinary Divine Action: The Nexus of Interaction”
and Robert John Russell, “Divine Action and Quantum Mechanics: A Fresh
Assessment,” both in F. Leron Shults, Nancey Murphy, and Robert John
Russell, eds., Philosophy, Science and Divine Action (Leiden: Brill, 2009).
37. Oderberg, Real Essentialism, p. 144.
38. Ibid., p. 148. See pp. 148–149 and 155–156 of Oderberg for discussion of
some of the metaphysical questions that may be raised about this account of
miracles.
39. See Oderberg, Real Essentialism, pp. 143–151, and Feser, Scholastic Meta-
physics, pp. 63–72, for further discussion of the Aristotelian interpretation
of laws of nature. For further discussion of the Thomistic conception of
miracles, see Feser, Five Proofs, pp. 238–246; Brian Davies, An Introduc-
tion to the Philosophy of Religion, third edition (Oxford: Oxford University
Press, 2004), chapter 11; and Anselm Ramelow, “The God of Miracles,” in
Ramelow, ed., God, Reason, and Reality (Munich: Philosophia Verlag, 2014).
40. Summa Theologiae I–II.109.3.
41. Summa Theologiae I–II.109.2. For a useful overview of Aquinas’s account of
grace, see Brian Davies, The Thought of Thomas Aquinas (Oxford: Claren-
don Press, 1992), pp. 262–273.
42. Cf. Noam Chomsky, Rules and Representations (New York: Columbia Uni-
versity Press, 1980), pp. 6–7, and Colin McGinn, Problems in Philosophy:
The Limits of Inquiry (Oxford: Blackwell, 1993), pp. 1–8.
43. Cf. Feser, Five Proofs, pp. 229–232.
44. William J. Wainwright, “Theology and Mystery,” in Thomas P. Flint and
Michael C. Rea, eds., The Oxford Handbook of Philosophical Theology
(Oxford: Oxford University Press, 2009), at pp. 79–80.
45. William Alston, “Two Cheers for Mystery!,” in Andrew Dole and Andrew
Chignell, eds., God and the Ethics of Belief (Cambridge: Cambridge Univer-
sity Press, 2005), at p. 99.
46. Richard Swinburne, The Christian God (Oxford: Clarendon Press, 1994).
47. Davies, An Introduction to Philosophy of Religion, chapter 1.
48. John Jenkins, CSC, “Faith and Revelation,” in Brian Davies, OP, ed., Phi-
losophy of Religion: A Guide to the Subject (Washington, DC: Georgetown
University Press, 1998), at pp. 216–217.
49. To be sure, the precise nature of the act of faith is a disputed matter among
Scholastic theologians. For a more detailed treatment of the specifc account
I have been defending and a discussion of rival accounts, see G. Van Noort,
Dogmatic Theology, Volume III, translated by John J. Castelot and William
R. Murphy (Westminster, MD: The Newman Press, 1961), pp. 335–349.
50. Summa Theologiae I.I.8. I thank two anonymous referees for helpful com-
ments on an earlier version of this chapter.
14 God, Chance and Evolution
In Memory of Benjamin Arbour
Alexander R. Pruss
1. Introduction
In the Western monotheistic religions, God isn’t just our frst cause. He is
our creator and designer. God’s design extends not just to general char-
acteristics like intelligence but to a signifcant amount of physical detail.
There is, thus, a fairly detailed set of intentions for features that one might
designate as God’s “blueprint” for us. Therefore, Western monotheism
accepts this thesis:
BLUEPRINT: When frst created, humans’ physical structure exactly

matched God’s blueprint, and humans had the structure they did
because of God’s blueprint.
The restriction “When frst created” allows for the possibility of a fall of
humankind, for instance as described in the book of Genesis.
Evolutionary explanations are both naturalistic and statistical. They
are naturalistic in that they do not invoke any supernatural causes like
God. They are statistical in that they do not make predictions that follow
with certainty from the theory, but rather describe processes that have
certain probabilities of producing the efect that is to be explained. These
probabilities are likely often low (cf. Gould 1991) but are nonetheless
explanatory. If a coin is loaded so that it has probability 2/3 of coming
up heads, we can give a high-probability statistical explanation of why
it came up heads if it did so, but we could also give a low-probability
statistical explanation of why it came up tails if it did that instead. The
low-probability case, further, is no harder to understand than the high-
probability case (Jefrey 1969). Likewise, even if the probability of ran-
dom mutations producing something like a girafe is low, at least as long
as that probability isn’t astronomically low (cf. Dembski 1998) we can
give a statistical explanation. I take it that evolutionary theory ofers such
statistical explanations for many important features of human beings,
such as our warm bloodedness, social intelligence, and so on.
Many people think, then, that the following thesis is true:
DOI: 10.4324/9781003125860-18
God, Chance and Evolution 365
EVOLUTION: There is a full evolutionary story – currently not
known in all its detail – that is true and explains the important fea-
tures of human physical structure entirely in naturalistic statistical
terms, and there were no supernatural interventions in evolutionary
history.
In Pruss (2011), I have argued that EVOLUTION is a speculative

philosophical claim that is not a part of current biological science (e.g.,
because “naturalistic” is itself not a scientifc concept), but in this chapter
I won’t address the question whether EVOLUTION or BLUEPRINT is
true. Rather, the question for this chapter is to see whether BLUEPRINT
and EVOLUTION can be held together. For there is a tension between
the two claims. How could God use evolution to produce human beings
in a way that guarantees that we ft the blueprint when evolution relies
on randomness?
I will begin by sketching criticisms of fve approaches to reconciling
BLUEPRINT with EVOLUTION, drawing largely on Pruss (2007), where
I had argued against the compatibility of the two theses. Then I will make
an excursus on the interpretation of probability, ofering both a liberal
ecumenical view of probability as anything that satisfes the axioms of
probability and a theistic version of best-ft probabilities connected with
the divine attribute of omnirationality (Pruss 2013). Finally, I will show
how it is that theistic best-ft probabilities can undergird statistical expla-
nations in a way that could make both BLUEPRINT and EVOLUTION
true. Furthermore, along the way, and independently of considerations
of evolution, we will see that theistic best-ft probabilities are a promising
alternative that solves problems with the non-theistic version.
2. Five Reconciliations
Either the initial physical conditions that the evolutionary process starts
with causally guarantee the fulfllment of the blueprint or they do not. If
they do, then we have the DETERMINISM reconciliation. If they do not
causally guarantee fulfllment of the blueprint, they may still epistemically
guarantee it for God: that is the GENERALIZED MOLINISM solution.
If they neither causally nor epistemically guarantee it, it may be that the
natural initial conditions end up guaranteeing fulfllment of the blueprint
when combined with supernatural causes. The difculty here is to retain
EVOLUTION, and a THOMISTIC story on which God’s activity works
on a level that does not compete with natural causation is the best account
available. There is, however, one last option left, and that is that the natu-
ral conditions and supernatural causes do not guarantee the satisfaction
of the blueprint, but it so happened that the natural conditions did result
in the blueprint being satisfed. This comes in two variants: either the
attempt to satisfy the blueprint happened once or a few times, the SKILL
366 Alexander R. Pruss
AND LUCK story, or it happened many times, of which the most preva-
lent account is the MULTIVERSE account.
Each of these ways has problems. In the frst three cases, I will focus
primarily on the problems that each way faces as a solution to the prob-
lem of evolution and creation, rather than its well-known independent
difculties, such as the ft between determinism and free will, the ground-
ing and circularity problems for Molinism (Adams 1977 and 1991), and
the difculty of distinguishing Thomistic levels of causation.1
2.1. Determinism
On the DETERMINISM solution, God chooses the initial conditions that
causally guarantee the fulfllment of the blueprint. The causal processes
whereby the plan enfolds, culminating in human beings, do not involve
any further special intervention from God but only the kind of divine
sustenance or cooperation that all natural processes involve.
It may seem that determinism cannot be reconciled with the statistical
element of evolutionary explanation, but that is too quick an objection.
Thermodynamic explanations are statistical in nature, and yet thermo-
dynamics developed in the 19th century at least in part in the context
of deterministic particle physics. It is a matter of continuing research in
philosophy of science just how such statistical explanations work given a
deterministic physics, but it is widely held that they do work.
However, determinism is only half of the DETERMINISM solution. The
other half is the thesis that God chooses the initial conditions so as to guar-
antee a particular outcome. And that throws a wrench into the statistical
explanations. To see this, suppose that you sequentially throw 1000 coins
of a roof in a perfectly normal way, and 485 of them land heads. Then
there is a fne statistical explanation of why about half of them landed
heads, namely that by the law of large numbers we expect convergence of
the proportion of heads to be half when tossing coins, and more precise
calculations using the binomial distribution can tell us that the probability
that the proportion of heads is between, say, 48% and 52% is about 0.79.
But suppose now that I toss 1000 coins of a diferent roof, but I have
a special skill: I can choose initial parameters like position, orientation,
rotational rate and velocity of a coin so as to guarantee a particular out-
come of the toss. I then employ that skill in each throw to ensure that
exactly 485 of my coins land heads. Then the statistical story that was
given in your case is no longer applicable. The law of large numbers and
the binomial distribution calculation both assume that the coin tosses are
independent. But they aren’t independent if I exercise the kind of control
over initial conditions that is indicated.
Even though the physical process between the initial conditions of
the toss and the landings of the coins could be exactly the same when
I threw them as when you threw them, my deliberate rigging of these
initial conditions to produce the particular outcome prevents the statisti-
cal explanation that applied in your case from applying in mine. This
seems to be true even if the rigging was done in order to precisely mimic
a random result: it seems that mimicry of the random isn’t random.
To make the point clearer, it helps to think about a paradigmatic expla-
nation of how randomness arises in processes like coin tossing that occur
in the realm of classical, and hence deterministic, physics: the microequi-
probability account (Strevens 2013). Take a case simpler than the fip of
a coin, namely the rotation of a spinner inside a circle divided up into ten
equal pie slices, alternating in color between white and black. Idealizing
away minor issues such as the spinner’s wobble, we can take the fnal
position of the spinner to be a simple function of the initial rotational
speed ω0, the initial angular position ϕ0, as well as the fxed parameters
of the frictional torque and the moment of inertia.2 Given that we have
two adjustable parameters in each spin, angular position and rotational
speed, we can now do a two-dimensional graph where the two axes rep-
resent the two parameters varying in some reasonable range, and where
we draw a black point in the chart if the spinner stops in a black sector.
The result is a number of relatively closely spaced diagonal black stripes
of approximately equal width. And this ensures microequiprobability:
chunks of the graph that aren’t too small will be about half black and half
white, and so if the person spinning the spinner does not have very fne
control over the spinning speed, the probability of the spinner landing in
Figure 14.1 Spinner results with frictional deceleration 1/2

the black is about a half. Likewise, if the initial parameters come from
some other physical process that can put out a somewhat broad range of
these parameters, we get roughly equiprobable outcomes.
However, if the initial position and speed of the spinner are calibrated
by the person doing the spinning so as to produce a particular result – for
instance, calibrated to lie in the small black triangular region in the lower
left corner of the graph – then it is no longer right to talk of black and
white being equiprobable outcomes.
In the coin-toss example, the parameter space is much more complex
than in the spinner example. For instance, there is the height of the coin
relative to the foor, the initial orientation of the coin, the velocity vector,
the axis of rotation and the speed of rotation. Moreover, these parameters
need to be adjusted for each of the 1000 coins tossed. The full state space
would then have thousands of dimensions. But the basic point remains.
If the person doing the tossing does not have fne control, we should treat
all possible heads/tails sequences as equally likely and hence we can give
the explanation in terms of the law of large numbers or the binomial
distribution for why the proportion of heads is close to a half. But if the
person doing the tossing does have fne control over the parameters and
exercises that fne control to get a particular result, we should no longer
treat all the sequences as equally likely.
Thus, when the initial conditions of a coin toss are preset in order to
result in a particular outcome, the statistical features of the situation are
irrelevant to the obtaining of the outcome. And it is precisely the point of
the DETERMINISM solution that God carefully chooses the initial condi-
tions to ensure particular features of the outcome, namely the satisfaction
of the blueprint. In that case, we can no longer give statistical explanations
of those features that lie in the blueprint. But while we may not know
exactly which features of humans lie in God’s blueprint – does God intend,
for instance, that we have ten fngers? – it is very plausible that there
is signifcant overlap between features that evolutionary theory purports
to explain and features that are in the blueprint. Social intelligence, for
instance, is something that evolutionary theory ofers an explanation of
and yet is very plausibly a central part of God’s plan for the human race.
Note an important point here. Given DETERMINISM and the previous
argument, it may very well be that all the causal processes in the evolu-
tion of the human race are exactly the way a complete evolutionary story
would say they are. There are mutations, some organisms die of, some
survive, and so on. But features like social intelligence can no longer be
given the statistical explanation that evolutionary theory ofers for them.
2.2. Generalized Molinism

Molinism holds that if C is a sufciently precise specifcation of the ante-
cedents of an indeterministic free action F, then either God knows that
if C were to hold, F would take place, or God knows that if C were to
hold, F would not take place, even though C does not determine whether
F does or does not take place. This enables God to exercise meticulous
providence over creatures despite their having a free will incompatible
with determinism, since his knowledge of which antecedents would result
in which actions allows him to choose these antecedents so as to get the
actions he wants. These actions, then, are not determined by God’s actu-
alization of the antecedent circumstances, but nonetheless God knew that
they would follow these circumstances.
It is natural to generalize Molinism to the case where F is any unde-
termined event, be it an action or a quantum indeterminacy. Given such
a generalized Molinism, if God wants, say, a particular indeterministic
mutation to occur, he need only think about the infnitely many possible
antecedents and fnd one such that the mutation would occur given that
antecedent. After all, it would be massively improbable that none of the
possible antecedents would yield the desired result. The GENERALIZED
MOLINISM solution, then, holds that God set up the initial conditions of
the evolutionary system in such a way as to ensure that both BLUEPRINT
and EVOLUTION would hold.
Standard objections to Molinism about free agency (e.g., Adams 1977
and 1991) carry over to the generalized version. However, there is another
problem in the statistical explanation context. Imagine that the 1000 coins
being tossed of the roof spin in an indeterministic way, but nonetheless I
have the ability to predict how a given coin would land were I to toss the
coin with any possible set of initial conditions, and that this allows me
to control how the coin lands simply by choosing the right initial condi-
tions. Then just as in the DETERMINISM case, the statistical explanation
of the outcomes is trumped by an explanation in terms of my choosing
the initial conditions so as to guarantee the outcome. For, once again,
the probabilities of various outcomes are irrelevant given my decision
to ensure the outcome. Even if heads were very unlikely, I could choose
initial conditions that would ensure that 485 of the coins land heads.
Like DETERMINISM, the GENERALIZED MOLINISM story is com-
patible with the particular events in an evolutionary account happening
without divine intervention, but it undercuts the statistically explanatory
aspect of the theory.
2.3. Thomism
The THOMISM solution ofers a theological compatibilism. Whenever a
creature causes some efect E, God causes the creature to cause E. More-
over, when the creature causes E indeterministically, then God causes the
creature to indeterministically cause E. The divine causation, however,
is of a diferent order from creaturely causation: it is primary causa-
tion, while the creaturely causation is secondary causation. This primary
causation does not in any way hamper indeterminism at the creaturely
level but is fully under God’s control.
Once again, I suspect my main objection to the DETERMINISM and
GENERALIZED MOLINISM solutions to the evolution-creation prob-
lem carries over. If the person throwing the dice controls how each indi-
vidual indeterministic outcome goes, then the statistical explanation in
terms of the propensities of the dice is trumped. But this point may be
less clear in this case because it is not clear how primary causation works.
Further, there is an objection specifc to THOMISM. According to
THOMISM, it is necessarily true that a fnite cause C causes E if and
only if God chooses to cause C to cause E. But if two propositions p and
q are such that necessarily p is true if and only if q is true, then they have
the same probability. Consequently:
P(God chooses to cause C to cause E) = P(C causes E).
But it is implausible that God’s decisions are subject to numerical prob-

abilities. It is even less plausible that numerical probabilities of divine
choices can be discovered by science and that these numerical probabili-
ties correspond exactly to the probabilities of events in the physical world.
For suppose I toss an indeterministic coin, and you bet on it. If a per-
fectly good God is choosing whether to cause the coin to come up heads
or to cause the coin to come up tails, the probabilities of God’s choice will
surely be tied to the values of the two outcomes – is it good for you to
win the bet, and so on – rather than just to the physics of the coin throw.
It might, of course, be that the two probabilities happen to coincide in
some cases. For instance, the coin might be fair and the reasons for God
choosing each of the outcomes might be equally strong and so God might
be equally likely to choose them. But it is unlikely that such a correla-
tion would always be present among the vast numbers of indeterministic
events in evolutionary history and outside it. And it is quite implausible
that such a correlation be a matter of metaphysical necessity.
It is more plausible that, apart from the presumably rare cases where he
has reason to do a miracle, God simply concurs in the coin’s indetermin-
istic causation rather than deciding how the coin will go as THOMISM
claims. But mere concurrence rather than theistic determination will not
be enough to guarantee that the blueprint will be satisfed.
2.4. Skill and Luck

The next option is that there are genuinely indeterministic processes
behind evolution and that God does not determine the outcomes of these
processes. Instead, God skillfully chooses initial conditions that are as
likely to lead to the satisfaction of his blueprint as he can make them
(consistently with his pursuit of other goals). Then, whenever God sees
that events are not according to his plan, God intervenes in the process
to put it back on track.
This will indeed yield BLUEPRINT. But to get EVOLUTION we need
the further thesis that God had the good luck that his intervention wasn’t
in fact needed. Things evolved on their own to satisfy the blueprint.
This account has an advantage over the preceding three: it is looks
clearly logically possible, and it is clear that if it is true, it reconciles
BLUEPRINT and EVOLUTION. The serious faw, however, is that while
the account is looks logically possible, it is unlikely to work.
For it is plausible that God’s plan for the human race has a fair amount
of specifcity. For instance, there are many combinations of sociality and
intelligence that a species could have, and the degree of sociality in beings
that God creates is something that we would expect to matter to a perfect
being. In light of convergent evolution, it might be that it’s likely for
evolution to produce something roughly like God’s blueprint, but it is
also quite plausible that the blueprint is sufciently precise that the kinds
of initial conditions that obtained at the beginning of the evolutionary
process were unlikely to result in the obtaining of the blueprint. It seems
much more likely that if God simply set initial conditions for an evolution-
ary process without any help from DETERMINISM or GENERALIZED
MOLINISM, he would have had to intervene at some point, and hence
EVOLUTION would be false.
So for SKILL AND LUCK to hold, God would have to be very lucky,
and we should avoid theories that posit great amounts of luck.
2.5. Multiverse
God could, however, decrease the amount of luck needed to get the blue-
print satisfed if he created a large – perhaps infnite – collection of sites
for evolution, either in a very large universe or in a multiverse. With a
large enough fnite number of sites, the probability of getting creatures
that satisfy the blueprint is close to one, and with an infnite number of
sites, it can be actually one by the law of large numbers.
One difculty with this is theological. We would be impressed by Shake-
speare’s power if he produced Hamlet by simultaneously typing on an
infnite number of typewriters at random, but we wouldn’t be impressed
by his artistry. A signifcant part of the motivation for BLUEPRINT is that
creation expresses divine art rather than just divine power.
This difculty can be solved. Perhaps at the sites where things don’t go
according to plan, God miraculously intervenes, for instance by prevent-
ing intelligent life from arising. This may solve the Shakespeare problem.
For imagine that Shakespeare is typing on infnitely many typewriters,
but destroying all the typescripts that are not masterpieces. Selection by
Shakespeare may take as much artistry as creation, and so this leaves
room for Shakespeare to be a real artist.
However, the hypothesis that God intervenes selectively at the evo-
lutionary sites where things weren’t going according to plan endangers
the statistical character of the explanations. Imagine that a man simul-
taneously tosses 1000 coins of the roof of each of 21000 buildings, but
whenever the number of coins landing heads is looking like it’s going to
be other than 485, he scoops up the coins just before they hit the ground.
Suppose further that for some reason we can only observe a location
where the coins hit the ground. Then the natural statistical explanation of
why we see close to 500 coins heads up is incorrect. The correct explana-
tion has to take into account the selection. We no longer have the kind of
statistical explanation found in EVOLUTION.
The best bet so far for reconciling EVOLUTION and BLUEPRINT may
turn out to be the multiverse approach without the widespread interven-
tion. This does undercut divine artistry to some degree, but perhaps that
degree isn’t as great as one might worry. After all, there will be artistry
in choosing the laws of nature, at least. Still, it would be good to have an
alternative. Developing that alternative will require a brief excursus into
the philosophy of probability theory.
3. What Is Probability?
3.1. Towards a Liberal View

There are four classical interpretations of probability theory. The indif-
ference interpretation, perhaps the oldest of the four, partitions the set
of possible outcomes of a process in a naturally symmetric way and dis-
tributes the probabilities equally between the outcomes. Thus, a cubical
die has equal probability, and hence probability 1/6, of landing on each
side. Unfortunately, the applicability of the indiference interpretation is
limited to cases where there are enough natural symmetries to ensure the
success of the procedure. It is difcult, for instance, to handle a loaded die
with the indiference interpretation.
The frequency interpretation aligns probabilities with the actual or
hypothetical frequencies of observations: a coin has probability 1/2 of
landing heads provided that in repeated independent experiments the fre-
quency of heads is 1/2. This sufers from a number of serious problems.
With a fnite sequence of experiments, we don’t expect the probability to
match the frequency exactly (if you toss a million coins, you should be
surprised to see exactly 500,000 of them landing heads), while defning
probabilities in terms of the frequency in an infnite sequence of experi-
ments will in some cases violate the axioms of probability.3 Third, the
frequency interpretation just gets explanatory facts wrong. That a coin
landed heads six times out of ten seems to be explained by the fact that the
coin had a chance 1/2 of landing heads in each experiment, but it is not
explained by the frequency fact that in a much, much longer sequence it
landed heads about half that time (cf. Hájek 1996, 79–80). And indeed we
are searching for probabilities that enter into evolutionary explanations.
The objective version of the epistemic view makes the conditional prob-
ability P(E|K) of E given K be the degree of belief that a perfectly rational
agent would assign to E if her total evidence were K. There is a subjective
version on which the values of these probabilities can difer from agent to
agent. Neither version is helpful when our interests are explanatory. The
degree of belief that a rational agent would assign to a proposition does
not yield an explanation of that proposition.
Finally, the propensity interpretation starts with conditional prob-
abilities defned by causal connections: P(E|K) is the degree to which the
situation described by K tends to lead causally to E being true, in the case
where K describes the conditions causally antecedent to E. There are then
technical challenges in extending the propensity interpretation to yield
conditional probabilities P(E|K) where K does not describe the condi-
tions antecedent to E, but at least probabilities defned by the propensity
interpretation connect neatly to explanatory concerns.
One lesson from the previous all-too-quick survey is that there are
at least two diferent uses to which we put probabilities: epistemic and
explanatory. For epistemic purposes, we surely need epistemic probabili-
ties. For explanatory purposes, we need a diferent interpretation. What
to do?
One suggestion is to adopt a liberal view. A plausible necessary condi-
tion on an adequate interpretation of probabilities in some domain of
discourse is that it makes the probability function P satisfy the classical
axioms of probability. One version (admittedly, more used in philoso-
phy than mathematics, since mathematicians prefer to work with sets
rather than sentences) of the axioms is that probabilities are defned for
a non-empty set of sentences closed under conjunction, disjunction and
negation, and:
i. P(A) ≥ 0 for any A

ii. P(T) = 1 for any necessary truth T
iii. P(A ˅ B) = P(A) + P(B) if A&B is impossible.
The third condition is often extended for countably infnite disjunctions,

but I will stick to the fnite case for simplicity.
Each of the four classical interpretations satisfes our necessary con-
dition at least in some contexts. Each is inadequate as a view of what
probability in general is, but each is useful in a limited domain. We can
then ecumenically say that any function that satisfes the classical axioms
of probability in a domain of discourse simply is a probability. There are
infnitely many probabilities, then, most of them not very useful. But some
are useful. And this is how mathematicians talk about probability: they
defne a probability as any function P that satisfes the relevant axioms.
At this point, we will consider a ffth and somewhat more recent inter-
pretation of probabilities. It will satisfy the axioms of probability, so
by the liberal view it will be a probability whenever it is defned. I don’t
claim it to be a view of what probability in general is, but, like the four
classic views, I simply take it to be a probability that is useful in some
contexts. And in particular it will be useful for reconciling EVOLUTION
with BLUEPRINT.
3.2. Best-Fit: Lewis’s New Frequentism

On the deterministic and hence simpler version of David Lewis’s best-ft
account, the laws of nature are defned as follows. Start by considering all
true (but typically incomplete) descriptions of the world. Informativeness
is the amount of information conveyed by the description, while simplic-
ity can be measured by the brevity of expression in a language whose
terms all carve the universe at the joints. Some descriptions are more
informative than others, and some are simpler than others. It is easy to
make a description more informative at the cost of simplicity by conjoin-
ing more information. But there are also ways of making a description
more informative while gaining simplicity. For instance, the statement
that the momentum of all horses, dogs, parrots and sheep is preserved
is less informative than the simpler statement that the momentum of all
material objects is preserved.
Among the descriptions of the world, there may be one that is opti-
mized to provide the best balance of informativeness and simplicity. If
there is one, then any proposition that is a consequence of such a descrip-
tion counts as a law. And Lewis believed that in the actual world there is
such a description, and hence there are laws.
For instance, one can say that E = mc2 quite briefy in a joint-carving
language, and saying this conveys an enormous amount of information
about stuf all the world. We can thus expect “E = mc2” to be either a
part of the optimal description or at least a consequence of it. On the
other hand, to say in joint-carving language that I am writing a paper on
a computer will take many words – we have to specify in joint-carving
ways what I am, what writing is, what a paper is and what a computer
is. And these many words will convey a limited amount of information
about a very small portion of the cosmos. It will be neither brief nor very
informative. Thus, it’s a law that E = mc2 but not a law that I am writing
this paper on a computer.
An account of best-ft probabilities can be obtained by following Lewis
(1994) in generalizing beyond determinism and extending the idea to
probabilistic descriptions. Instead of requiring a probabilistic description
of the world to be true, we require it to be a unique description of a func-
tion that coheres with the axioms of probability theory, and then instead
of just optimizing for a balance of informativeness and simplicity, we
optimize for a balance of informativeness, simplicity and ft of the prob-
ability function to the world’s actual frequencies of events.
We can then say that if there is such an optimal description, and at
least if that balance is good enough (Lewis does not have this condition,
but perhaps we should include it), then the described probability function
embodies the global best-ft probabilities.
This is a new frequentism. Like in classical frequentism, the world’s
actual frequencies of events enter into the defnition of probabilities. But
unlike in classical frequentism, neither are the frequencies of events the
only thing that enters into the defnition nor need the probabilities defned
by the account exactly match the frequencies. For instance, if some funda-
mental physical process occurs a million times in the history of some small
universe, and 500,002 times outcome A occurs, classical frequentism will
either have to dubiously reach for the counterfactual frequency given an
infnite number of experiments, or will have to say that the probability
is exactly 0.500002. On the other hand, best-ft probability will sacrifce
a slight bit of ft for the sake of much greater simplicity and assign prob-
ability 1/2, which is presumably what an experimenter given the data
would actually assign (cf. Lewis 1994, 481).
The new frequentism guarantees that probabilities satisfy the axioms
and does not require an infnite sequence of events, thereby solving two
of the problems we saw for classical frequentism. And, by the liberal
account of probability, it is a probability. It also solves another problem
pointed out by Alan Hájek (2009, 220–225): classical frequentism ensures
by fat that probabilities are always defned when there are frequencies.
But that is risky: what if in fact some event does not have a meaningful
probability? The new frequentism does not guarantee that every event E
has a probability. It might be that the probability function specifed by
the optimal description is defned over a set of sentences that does not
include a report of E. Or we might be in a world where no description has
an optimal balance of informativeness, simplicity and ft, or perhaps the
one where the ft isn’t between the optimal description and the frequencies
just is not good enough.
Finally note that while the best-ft account of probabilities evolved from
an account of laws, it need not be seen as an account of laws. One could
adopt any other account of laws and still use the best-ft apparatus to
defne best-ft probabilities. And these best-ft probabilities would still be
probabilities according the liberal view, simply because they satisfy the
axioms of probability.
Moreover, it is possible to defne best-ft probabilities relative to par-
ticular contexts. A context can be defned by (a) specifying the non-math-
ematical predicates of a language L and (b) restricting the scope of the
quantifers in L. The language L will then specify the set of sentences to
which we are assigning probabilities, and simplicity will be defned by
brevity in L. Then the context-relative best-ft probability will be obtained
by optimizing informativeness, ft in the restricted scope, and L-relative
simplicity.
One might, for instance, specify the language to have only the predi-
cates Heads(x) and Tails(x), and then restrict the quantifers to unambigu-
ous coin toss events on earth. As a result we will get a best-ft probability
for heads in coin tosses, presumably 1/2. In doing this, one needn’t worry
about the fact that the two predicates do not cut nature perfectly at joints.
More interestingly, to get probabilities useful for evolutionary theory we
might restrict the language to non-gerrymandered terms of earth biology
and restrict the quantifers to things and events on earth. We then do not
need to worry whether the terms of biology turn out not to be perfectly
natural.
3.3. A Well-Motivated Theistic Variant

The explanatory problem that frequentism faced remains both for global
best-ft probabilities and especially for contextual ones. If in a million
instances we had A occurring 500,002 times, that fact does not explain
why in the frst ten A occurred approximately half the time (six times,
say). Rather, pace Lewis, both facts should be explained by the fact that
A had a chance of 1/2 of occurring each time.
This problem is analogous to a well-known circularity problem faced
even by the deterministic version of Lewis’s account of laws. On the one
hand, laws are explained by the patterns of particular occurrences. On
the other hand, laws explain these occurrences. Some of the discomfort
may be alleviated when we note that the types of explanation are difer-
ent: the patterns of particular occurrences ground the laws, while the laws
scientifcally explain these occurrences. But discomfort still remains. And
it really looks like one is using the more obscure, namely global patterns
of occurrences, to explain the less obscure, namely local occurrences.
Call those propositions that are generated by Lewis’s account of laws
“Lewisian laws,” and those generated by the contextualized version “con-
textual Lewisian laws.” Even someone who accepts a competing account
of laws can agree that there are such things as Lewisian laws but may
deny that Lewisian laws are the same thing as laws of nature.4 Now, (con-
textual) Lewisian laws are generated by an optimal description of (a part
of) the world. Let us call the conjuncts in that description “fundamental
(contextual) Lewisian laws.”
Suppose now that it so happens that God intends the fundamental
(contextual or not) Lewisian laws to be true, so that these laws’ being true
is a part of God’s reason for creating this world or acting in the world as
he has. Then the fundamental Lewisian laws will fgure in a fnal causal
explanation of events in the world: the particular events in the world are
so arranged (by God) as to ensure that the fundamental Lewisian laws
are true. Final causal explanations of this sort do not involve any vicious
circularity. That this paper gets read fgures in a fnal causal explanation
of my writing it, but that I write it is a partial efcient causal explanation
of its getting read.
One reason there is no vicious circularity is that in such agential cases
we can arguably reformulate the fnal causal explanation. That this paper
gets read fgures in the explanation of why it gets written, but it doesn’t
fgure as a conjunct in that explanation. Rather, it fgures within the scope
of a report of my intentions: that I intend that this paper get read explains
why I write it.
When the Lewisian laws are intended by God, theism thus ofers a
solution to the explanatory weakness of Lewisian laws.
Unfortunately, there is a complication in the indeterministic case. Sup-
pose that some event type B occurs 10100 times, and that almost exactly
half of the time it has a property Q, and that it so works out that it is a
fundamental Lewisian law that events of type B have property Q with
best-ft probability 1/2. But now suppose that 1000 instances of B have
been observed, and 485 of them turn out to have Q. What explains this
statistical fact? God’s intending that B-type events have best-ft probabili-
ties that make each one have an independent probability 1/2 of having
property Q does not require him to make these 1000 instances be roughly
evenly balanced between Q and non-Q. Indeed, these 1000 instances are
statistically swamped by the 10100 − 1000 other instances. Unless we have
a borderline case (and we can suppose we don’t), given the other instances
it makes no diference to what the fundamental Lewisian laws are whether
zero, 485 or 1000 of these 1000 Bs have Q. So it seems that although we
can give an explanation in terms of God’s intentions for why Q is instanti-
ated about half the time among the full 10100 instances of B, we cannot
give an explanation for why it so happens in the observed 1000 times.
There is, however, a solution to this problem. I earlier suggested that
we might want to say that we have Lewisian laws only as long as there
is a description whose balance of informativeness, simplicity and ft is
good enough. In some worlds Lewisian laws will have a better balance
of informativeness, simplicity and ft than in others. Call that balance the
“lawlikeness” of the Lewisian laws. Suppose now that God is not only
trying to make the Lewisian laws be true, but that he also wants them
to (a) be Lewisian laws and (b) be as lawlike as they can be while ftting
with God’s other plans.
Now although the fact that a given set of 1000 Bs has the expected dis-
tribution of Q may not make any diference as to what the Lewisian laws
of the world are, it will make a diference as to how lawlike these laws are.
While much technical work would be needed to defne ft, intuitively the
statement that each B-type event has an independent best-ft probability
of 1/2 of having Q better fts with the hypothesis that approximately
half of these observed 1000 Bs have Q than with the hypothesis that
they don’t. Given God’s desire to make the Lewisian laws as lawlike as
possible subject to other constraints, we thus do have an explanation for
the observed statistics.
Of course all this leads to the question whether God cares about the
Lewisian laws of this world and their lawlikeness. But I contend that he
does. A perfect being cares about everything that is good. And lawlikeness
is good for two reasons. First, lawlikeness of Lewisian laws in a world
containing agents like us is instrumentally good: it makes it possible for
the agents to predict and understand the natural world around them.
Second, a world’s being describable with a high degree of informative-
ness, simplicity and ft is aesthetically valuable. Physicists are particularly
sensitive to this aesthetic value.
If lawlikeness is good and so God cares about it, then it’s not a mere
coincidence that he created a world with lawlike Lewisian laws. God is
perfectly rational – indeed, omnirational (Pruss 2013) – and the fact that
a given world with particularly lawlike Lewisian laws has both epistemic
and aesthetic value is going to be among God’s reasons for actualizing
that world, and hence the facts about lawlikeness will be explanatorily
relevant.
If theism is true, then Lewisian laws, both deterministic and indetermin-
istic, have explanatory force. This is true even if the Lewisian laws turn
out not to be (or not to all be) laws of nature. Patterns in the world are
valuable, both epistemically and aesthetically, even if these patterns are
not actually laws of nature (think, for instance, of all the patterns humans
love to impose in architecture and other arts).
Moreover, the existence of contextual Lewisian laws in cases where
either the context is important to rational agents like us or is not too
badly gerrymandered will also have epistemic and aesthetic value relative
to humans to the extent that these laws are lawlike in that context. And
hence they will also be explanatory. If the universe is created by a being
who cares about humans’ epistemic and aesthetic goods, such an anthro-
pocentrism is not a fatal faw in an explanatory story. So we need not even
worry about how objectively valuable the patterns embodied in Lewisian
laws are (though I believe that they are likely to be objectively valuable).
Thus, theistic best-ft probabilities, whether global or contextual, are
genuine probabilities by the liberal view of probability and escape a num-
ber of the standard difculties facing frequentism.
4. Evolutionary Statistical Explanation
4.1. The Trick

We can now leverage theistic best-ft probabilities to overcome the dif-
fculties facing DETERMINISM, GENERALIZED MOLINISM and
THOMISM as reconciliations between EVOLUTION and BLUEPRINT.
This approach won’t overcome all the difculties facing GENERALIZED
MOLINISM and THOMISM, such as the grounding problem or the dif-
fculty in explaining the diference between primary and secondary causa-
tion. But it will overcome the difculties I pointed out that are specifc to
reconciling creation and evolution.
The trick to give us the solutions is simply this. Evolutionary explana-
tions are statistical in nature. Statistical explanations involve probabili-
ties. And now we take the relevant probabilities to be theistic Lewisian
best-ft probabilities, global ones if that turns out to work for evolutionary
purposes, or, if not, best-ft probabilities in a restricted context such as
earth biology. We will next examine how this trick solves the difculties
for the aforementioned three reconciliations.
4.2. Determinism and Generalized Molinism

Both DETERMINISM and GENERALIZED MOLINISM faced an
explanatory problem that I illustrated through a story about coin tosses.
If the coins were tossed skillfully by me so as to ensure that about half
of them land heads, it seemed that the fact that probabilistic claims
about the coin tosses no longer provided a statistical explanation of the
outcome.
That argument was very plausible when the probabilistic claims about
the coins were understood in terms of something like indiference or pro-
pensity probability. But suppose that they are instead understood using
Lewisian best-ft probabilities contextualized to a language that can report
only whether a given toss results in heads or tails and with the domain
restricted to my coin tosses. Suppose further that I always skillfully toss
the coins with the intent to promote the lawlikeness of the contextual-
ized probabilistic Lewisian laws. Then the fact that the contextualized
Lewisian probabilistic laws predict that about half of the coins will land
heads becomes explanatorily relevant, for the further I depart from half
of the coins landing heads in the particular situation, the poorer the ft
of the contextualized Lewisian laws would be. Hence I try to avoid such
departures, and my attempt to avoid them is explanatory of the statistical
features of my coin tosses. And so the predictions of the probabilistic laws
are central to this explanatory story.
Similarly, then, we can suppose that God can skillfully choose the initial
conditions for the evolutionary process in such a way as to ensure, either
by means of determinism or knowledge of relevant counterfactuals, that
both (a) the blueprint is satisfed and (b) the Lewisian best-ft probabilities
(either global or contextualized to earth biology) are as they are and ft
as well as they can keeping other desiderata fxed (or perhaps balancing
the ft against them). In such a case the probabilistic predictions of the
Lewisian best-ft probabilities are explanatorily relevant, and we have
both bona fde statistical evolutionary explanations (with respect to the
best-ft probabilities) and intricate divine design.
One may have a worry that this smacks of deceit. God is making
the frequencies of events look like the ones predicted by evolution-
ary theory, but these frequencies are explained by divine skill rather
than the natural propensities of objects in the world. But there is no
deceit here. We learn from the liberal view that there are many dif-
ferent kinds of probability, and thinking about theistic best-fit prob-
abilities shows that those are among the explanatory ones as well.
Some thinkers think the frequencies of biological events are explained
by natural propensities rather than theistic best-fit probabilities. But
others, including all Lewisians, read the data as explained by best-fit
probabilities. On the view I am defending, the non-theists in the lat-
ter group are mistaken as to how best-fit probabilities explain – for,
indeed, without theism they don’t explain at all. But they are not
mistaken that these probabilities explain.
In any case, the error here is probably metascientifc rather than scien-
tifc. Evolutionary biology should surely be neutral on the foundations of
probability. There is no deceit within the scope of biology. And the avail-
ability of a wide variety of metascientifc views means that anyone who
adopts a particular interpretation of probability does so at their own risk.
4.3. Thomism
The difculty with THOMISM was the equation:
P(God chooses to cause C to cause E) = P(C causes E).
It seemed impious to suppose that numerical probabilities govern divine

decisions and implausible to suppose that science discovers probabilities
of divine choice.
Once again, however, the plausibility of the objection depends on the
interpretation of probability. If we take P to represent propensity-based
probabilities, then indeed to ascribe numerical propensities to God does
seem to limit God, and it is implausible that science would discover God’s
propensities of choice.
On the other hand, if P is understood in a frequentist way, the equation
is entirely innocent. The frequency of divine choices of events matches the
frequency of these events, simply because on THOMISM God exercises
meticulous control. The frequentist probabilities are consequent on God’s
choice. Nor is it surprising to suppose that science discovers frequencies of
divine choice when these frequencies are expressed in the natural world.
But frequentism undercuts the explanatory force of the probabilities.
However, if we move to theistic best-ft probabilities, we can have the
benefts of the frequentist view without the undercutting of explanation. If
the language in which we express the best-ft probabilities includes terms
for describing divine choices regarding events in this world, then best-ft
probabilities of divine choices of events will match the best-ft probabili-
ties of the events.5 But, just as with frequentism, this does not mean that
God’s decisions are in some objectionable way governed by numerical
probabilities. Rather, God selects the Lewisian laws and plans to make
them as lawlike as reasonably possible, and then his further choices ft
with his plan.
It is true that we would expect the probabilities of divine choices to be
proportioned to the strength of the reasons for these choices. But that God
has opted to make the Lewisian laws be as lawlike as he reasonably can
gives him reason to act in ways that ft with these Lewisian laws.
5. Conclusions
The theistic best-ft account of probabilities allows one to remove the
special objections to using DETERMINISM, GENERALIZED MOLIN-
ISM or THOMISM to reconcile evolutionary explanations with divine
blueprints. For these special objections have signifcant force only if
we understand probabilities in terms of causal propensities. As long as
propensity-based probabilities are the only explanatorily relevant prob-
abilities, the objections are fatal. But theistic best-ft probabilities can
be explanatory, given an omnirational God who seeks to promote the
epistemic and aesthetic goods involved in Lewisian laws.
This account is neutral on whether Lewisian laws are actually laws, as
well as on what other kinds of probabilities there may be in the world
beyond the best-ft ones.
Notes
1. However, it is worth noting that one of these three independent difculties is
easy in this context: we need not worry about the ft between determinism and
free will, because we only need determinism within the evolutionary history,
which is compatible with indeterminism in the minds of humans once humans
are on the scene.
2. The angular position at time t will be ϕ0 + ω0t − κt2 / 2I, where κ is the frictional
torque and I is the moment of inertia, as long as t does not exceed the stopping
time. The stopping time is tfnal = ϕ0I / κ, and the fnal angular position can be
seen to be ϕfnal = ϕ0 + ω02I / 2κ.
3. Suppose two coins are tossed repeatedly. The frst yields the sequence TH10
T100H1000T10000 . . . (superscripts indicate repetition). The second yields
THTHTHTH . . . . Consider three events: A, the frst coin landing heads; B, the
second coin landing heads, and C, both coins landing the same way (i.e., both
heads or both tails). It is easy to verify that the limiting frequency of events of
type B is 1/2, as is the limiting frequency of events of type C. But now notice
that events of type A have no limiting frequency. After one toss, the frequency
of heads is 0; after 11, it is 10/11; after 111, it is 10/111; after 1111, it is
1010/1111; and so on. The frequency of heads keeps on oscillating between a
low number and a high number, and has no limit. According to the frequency
interpretation, then, P(A) cannot be defned. But A is logically equivalent to
(B & C) ∨ (~B & ~C), and since the sentences for which probability is defned
need to be closed under conjunction, disjunction and negation, P(A) needs to
be defned according to the axioms.
4. There are two possibilities here. It could be contingently true that Lewisian
laws are coextensive with the laws of nature, but that being a law is diferent
from being a Lewisian law, or it could actually be true that there are Lewisian
laws that aren’t laws or laws that aren’t Lewisian laws.
5. Either because it’s necessarily true, given THOMISM, that God chooses some-
thing just in case it happens, or simply because it will be a Lewisian law that
God chooses something just in case it happens.
References
Adams, Robert M. “An Anti-Molinist Argument”, Philosophical Perspectives 5
(1991) 343–353.
Adams, Robert M. “Middle Knowledge and the Problem of Evil”, American
Philosophical Quarterly 14 (1977) 109–117.
Dembski, William A. The Design Inference: Eliminating Chance Through Small
Probabilities (Cambridge: Cambridge University Press, 1998).
Gould, Stephen. Wonderful Life: The Burgess Shale and the Nature of History
(London: Penguin, 1991).
Hájek, Alan. “Fifteen Arguments Against Hypothetical Frequentism”, Erkenntnis
70 (2009) 211–235.
Hájek, Alan. “‘Mises Redux’ – Redux: Fifteen Arguments against Finite Fre-
quentism”, Erkenntnis 45 (1996) 209–227.
Jefrey, R. C. “Statistical Explanation vs. Statistical Inference.” In Essays in
Honor of Carl G. Hempel, ed. N. Rescher (Dordrecht: Reidel, 1969)
104–113.
Lewis, David. “Humean Supervenience Debugged”, Mind 103 (1994) 473–490.
Pruss, Alexander R. “How Not to Reconcile the Creation of Human Beings with
Evolution”, Philosophia Christi 9 (2007) 145–163.
Pruss, Alexander R. “A New Way to Reconcile Creation with Current Evolution-
ary Science”, Proceedings of the American Catholic Philosophical Association
85 (2011) 213–222.
Pruss, Alexander R. “Omnirationality”, Res Philosophica 90 (2013) 1–21.
Strevens, Michael. Tychomancy: Inferring Probability from Causal Structure
(Cambridge, MA: Harvard University Press, 2013).
15 Teleology, Providence,
and Powers
Simon Maria Kopf
Introduction
What would happen if one were to rerun the tape of life? Is the emer-
gence of human life an inevitable product of biological evolution? Or are
humans a mere product of Lady Luck in disguise – an improbable event
resulting from the concurrence of various circumstances and occurrences,
all of which have such a low probability that any replay of evolution
would, for all that we know, yield wildly diferent results. Consider the
following thought experiment:
You press the rewind button and, making sure you thoroughly erase
everything that actually happened, go back to any time and place
in the past. . . . Then let the tape run again and see if the repetition
looks at all like the original. [1] If each replay strongly resembles life’s
actual pathway, then we must conclude that what really happened
pretty much had to occur. [2] But suppose that the experimental ver-
sions all yield sensible results strikingly diferent from the actual his-
tory of life.1
This oft-quoted thought experiment captures the essence of a famous

controversy in evolutionary biology between Harvard palaeontologist
Stephen Jay Gould and his colleague at Cambridge, Simon Conway
Morris. The debate concerns the repeatability of biological evolution,
raising in particular questions about the purported goal-directedness of
evolution, or its teleology. In this debate, Gould takes a stance against
the idea of evolutionary progress, arguing instead that each evolution-
ary replay would yield substantially diferent results. Conway Morris, by
contrast, famously disputes Gould’s thesis that evolution is an entirely
undirected and unpredictable process.2 The trigger and empirical basis
for this controversy was the reinvestigation of the Burgess Shale of British
Columbia – one of the most important fossil localities enabling palaeon-
tologists to study the aftermath of the Cambrian explosion.3
DOI: 10.4324/9781003125860-19
384 Simon Maria Kopf
The thought experiment of replaying the tape of life, initially considered
by Gould to be empirically untestable,4 has gained further prominence in
recent years due to new scientifc developments showing that the biological
sciences are, in fact, making some progress towards an empirical adjudica-
tion.5 Although this new evidence is far from conclusive, it suggests that
“there are predictable portions within life’s tape and that evolution might
not be as unpredictable as once thought 25 years ago, when Stephen Jay
Gould formulated his original question.”6 The question is, in particular,
would such evolutionary replays, at present testable only in microevolu-
tion, (1) resemble the factual course of evolution and thus indicate, at least
to some extent, the probability of the actual macroevolutionary outcomes,
including human beings? Or would these replays (2) look substantially
diferent, suggesting that evolution would probably take a diferent path
each time because it is a contingent and undirected matter?
In evolutionary debates, such as the one just outlined, the topic of
teleology, or end-directedness, remains a controversial issue, especially in
the feld of science and religion. In this context, it is often assumed that
theologians are in need of an end-directed account of evolution, that a
robust doctrine of divine providence calls for an overarching teleology
of nature.
This chapter challenges this assumption, arguing instead that on nature’s
part teleological natures and causal powers of substances are sufcient
for a traditional account of providence. To this end, I shall frst discuss,
in Section 1, a recent trend suggesting that a Christian view requires an
end-directed evolution. On this approach, providence is factually equated
with and substituted by progress and predictability. Then I shall turn, in
Section 2, to a traditional theory of providence to show that at least on
the infuential Thomistic account, God’s providence operates primarily
through the teleological natures and powers of creatures, which sufce in
an important sense for a robust account of providence.
1. Does Teleology Matter?

An Evolutionary Controversy
Among many scholars in the feld of science and religion there seems
to be a common conviction that a lack of evolutionary directionality,
understood as scientifcally predictable progress towards a set end or
goal, might render obsolete the traditional doctrine of divine providence.
Providence here means God’s universal guidance of and loving care for
all of his creation. On this common view, God’s providential guidance of
evolution implies an end-directionality of evolution in the scientifc sense.
Michael Ruse, for instance, writes:
[F]or the Christian, human beings are necessary. Their arrival in this
universe is not a matter of chance or whim – might have been, might
Teleology, Providence, and Powers 385
not have been. We cannot paint God as an aspiring parent, trying
desperately to have kids but with no frm guarantees. If God wanted
to have kids, God was going to have kids. And here’s the rub: evolu-
tion through natural selection makes all this very problematic.7
In line with option (1) in the earlier quote from Gould, where he empha-
sises that if the replays of life’s tape were to resemble each other, then
what happened would have had to happen, Ruse suggests in this passage
that within an evolutionary framework the human species had to evolve
according to the Christian view. He states that the arrival of the human
species is inevitable and not by chance, on a providential view, but that
our contemporary accounts of evolution do not show, and in efect do
not allow for, the necessary directionality. To the extent that the scien-
tifc account of biological evolution exhibits no directionality towards
human beings, then, the theological doctrine of providence appears to
lack warrant.8
a. Providence and Necessity

Ruse believes that we have “a theological problem with the necessity
of humankind”9 and aims for a scientifc solution to this problem. On
this view, the non-directionality of evolution is a problem for theology
that needs a scientifc resolution, namely, evolutionary directionality. But
since evolutionary biology cannot at present fulfl the theological demand,
Ruse suggests a multiverse theory with an infnite number of universes as
remedy. If the number of attempts is infnite, then somebody will even-
tually win the lottery ticket no matter how low the probability of this
specifc outcome is. Given an infnite number of tries, humans will evolve
inevitably in some universe at some point.10
As a frst step of analysing the implications of the outlined view, let us
look more closely at the necessity posited in Ruse’s statement: “human
beings are necessary,” he says, for, “If God wanted to have kids, God
was going to have kids.”11 In other words, if God wills human beings, as
Christians believe, then necessarily human beings will emerge. And if we
add, as Ruse suggests further, that they come about by way of evolution,
we can restate: if God wills human beings, then necessarily human beings
will evolve.
Now the necessity posited by Ruse could be taken either as a neces-
sity of the consequent (necessitas consequentis, NCtis) or as a necessity
of the consequence (necessitas consequentiae, NCtiae) – the diference
being whether the necessity attaches to the consequent (‘p’) or rather the
consequence (‘if God wills p, then p’):
(NCtis) If God wills p, then necessarily p: God wills p → □ p

(NCtiae) Necessarily, if God wills p, then p: □ (God wills p → p)
In the frst case, the necessity attaches to the consequent, or the apo-
dosis. Applied to the question of the evolution of humans, this would
mean that human beings evolve necessarily; that is, they evolve in a neces-
sary manner. Ruse comments – and note that this is the same question as
posed by Gould in his thought experiment of replaying life’s tape – that
“since humans did evolve as they did, they could have evolved as they
did. But did they have to evolve as they did?”12 So the fact that humans
have evolved shows its possibility, but not its necessity. Like Gould, as we
shall see later, Ruse then denies that humans had to evolve, arguing that
this puts pressure on theology:
In the frst place, natural selection is relative. . . . Then in the second

place, the raw building blocks of evolution – the mutations – are
random, not in the sense of being uncaused (we know a lot about the
causes) but in the sense of not occurring according to need. There is
no direction and so evolution apparently can go whichever way.13
According to Ruse, then, on the current theory of biological evolution

with its two pillars of natural selection and random genetic mutation,
humans did not have to evolve; their emergence is a matter of contin-
gency, not of necessity.
But if human beings did not evolve necessarily, Ruse argues, theol-
ogy faces a problem with the posited ‘necessity of humankind.’ This
problem consists presumably in the fact that no necessity attaches to
the consequent of Ruse’s rending of the Christian providential claim,
which would eventually make it a false statement: ‘If God wills human
beings, then human beings will evolve necessarily’ is a false statement if
the protasis, or the antecedent, namely, that ‘God wills human beings,’
is true, as Christians assume, but the apodosis, or the consequent, that
‘human beings evolve necessarily,’ that is, evolve by biological necessity,
is in fact wrong. Therefore, if one assumes the necessity resulting from
providence to be a necessity of the consequent, then the demands of
theology are not met.
b. Providence and Necessity: An Objection

Some theologians and philosophers object to this use of necessity, arguing
instead that providence entails only a necessity of the consequence, on a
traditional theological account. Among them is Thomas Aquinas, who
states that, as a consequence of their view of divine providence, “some
have posited that the series or the disposition of [natural] causes is in itself
necessary, so that all [things] would happen by necessity. . . . But this is
clearly false.”14 By contrast, Aquinas, among others, claims that the neces-
sity resulting from providence is rather a necessity of the consequence, or
a conditional necessity:
And so it has to be said that when considering secondary causes,
fate [the disposition of secondary causes under God’s providence] is
changeable, but when considering its being subject to divine provi-
dence, it derives an unchangeableness, not of absolute necessity but
of conditional [necessity]: accordingly, we have to say that this con-
ditional is true and necessary, ‘If God foreknew that this will be, then
it will be’ (si Deus praescivit hoc futurum, erit).15
And likewise Aquinas states,
it does not follow that if God wills something, it will come about by
necessity: but that this conditional is true and necessary, ‘If God wills
something, then it will be’ (si Deus aliquid vult, illud erit). Yet it is not
necessary that the consequent is necessary.16
If the necessity in question is a necessity of the consequence, then the

necessity attaches to the consequence (‘if God wills p, then p’), not to the
consequent (‘p’). This alternative position, where the necessity concerns
the conditional as such, leaves open whether the consequent is necessary:
whether or not the evolution of the human species is necessary in this
instance, it is necessary that if God wills human beings to evolve, then
they will evolve.
The point of this alternative approach is that the necessity here quali-
fes the conditional: on the condition that God wills p, p is necessary.
But in and of itself, p may either be necessary or contingent. In scholastic
terminology, the consequent, p, is conditionally necessary. But condi-
tional necessity is arguably compatible with contingency. For example,
if Socrates runs, it is necessary that Socrates runs. Yet such a conditional
necessity does not necessitate his running, which is a contingent act of
choice. As Eleonore Stump observes, “Although it is necessary that if
Socrates is running he is running, it is not necessary that Socrates is
running.”17
If providence implies a necessity of the consequence, where the conse-
quent is conditionally necessary, therefore, the consequent can be either
(absolutely) necessary or contingent. Although the conditional ‘if God wills
p, then p’ is necessary, by a necessity of the consequence, the consequent, p,
can be either absolutely necessary or merely conditionally necessary – the
latter of which is compatible with contingency. On Aquinas’s account,
providence does not imply a necessity of the consequent but merely a
necessity of the consequence, for reasons we will see later.
This brief discussion of the linking of necessity and providence might
sufce to show that Ruse, as an exemplar of the view under discussion,
seems to suggest that evolutionary processes would themselves have to
be necessary in order to meet the theological demands of the doctrine
of providence. While this view is in principle shared by others, some of
these scholars, as we shall see next, would advocate a moderated ver-
sion. According to this moderated version, an evolutionary directionality
towards the emergence of the human species is sufcient in case this end
is attained with a very high probability – a view Ruse might concur with,
on a more amicable reading. To bolster this assertion, it will prove helpful
to examine the outlined debate between Stephen Jay Gould and Simon
Conway Morris more closely.
c. The Gospel of Contingency

In his acclaimed book Wonderful Life (1989), Harvard palaeontologist
Stephen Jay Gould challenges the conventional iconographies of the lad-
der of progress, or the idea of evolution as predictable progress towards a
goal. He argues that the reinvestigation of the fossil record of the Burgess
Shale calls for a substantially altered view of the history of life, accord-
ing to which the emergence of life is radically contingent rather than
progressive and predictable. Taking these fossils as an exemplary case
study, Gould makes a case for the signifcance of contingency in biological
evolution. He states:
[The Burgess Shale fossils have] confronted our traditional view about
progress and predictability in the history of life with the historian’s
challenge of contingency – the ‘pageant’ of evolution as a stagger-
ingly improbable series of events, sensible enough in retrospect and
subject to rigorous explanation, but utterly unpredictable and quite
unrepeatable.18
The new view of the history of life Gould advocates in this quotation is
based on contingency rather than the two conventional pillars of progress
and predictability; as he famously puts it, “almost every interesting event
of life’s history falls into the realm of contingency.”19
Gould’s endorsement of contingency, by which he refers both to events
that are possible not to be and to those that are dependent upon some-
thing else,20 was dubbed ‘the gospel of contingency’21 by biochemist and
Nobel laureate Christian de Duve because, as he puts it, in Gould’s writ-
ings contingency becomes a ‘creed’ spread to uproot the Christian belief
that the evolution of humankind “refects some kind of directionality
in biological evolution.”22 Indeed, Gould concludes his thought experi-
ment of replaying life’s tape with a prediction that has stirred quite some
controversy:
Wind back the tape of life to the early days of the Burgess Shale; let it
play again from an identical starting point, and the chance becomes
vanishingly small that anything like human intelligence would grace
the replay.23
So on Gould’s view the emergence of the human species is extremely
improbable because the history of life is radically contingent – a view
Gould explicitly turns against conceiving of “life’s history as the fulflment
of a divine purpose,”24 a phrase, incidentally, that shows the way in which
he confates evolutionary directionality with God’s providential care.
In Gould’s writing, then, we encounter once again the assertion that
for the doctrine of providence to prevail, replays of the tape of life would
have to (1) yield similar results. For recall that only in this case, as Gould
puts it in the opening statement, ‘what really happened pretty much had
to occur.’ Consequently, providence implies a necessity of the consequent,
on his view: human beings would ‘pretty much’ have to evolve necessar-
ily. But in actual fact, Gould continues, evolution is full of contingencies
indicating that (2) each replay of life’s tape leads “evolution down a path-
way radically diferent from the road actually taken.”25 Thus the factual
evolutionary outcomes, including human beings, did not have to evolve,
but evolved rather contingently, both in the sense that the emergence of
the human species was contingent upon, or dependent on, events that
were themselves contingent, or possible not to be. As in the case of Ruse,
the verdict is that the evolution of the human species is all but necessary.
d. The Gospel of Inevitability

By contrast, Cambridge palaeontologist Simon Conway Morris ofers a dif-
ferent take on the debate, one many Christians concerned with providence
and contingency have welcomed and endorsed, not least because it renders
contingency largely irrelevant. But when it comes to their fundamental
claim about the inevitability of evolution, the assumption of providence
implying a necessity of the consequent reappears also on this approach.
In The Crucible of Creation (1998), Conway Morris challenges Gould’s
presentation of the history of life as utterly contingent and undirected,26
arguing instead that the real lesson to be drawn from the Burgess Shale
fossils is the ubiquity of evolutionary convergence, being defned as “the
recurrent tendency of biological organization to arrive at the same ‘solu-
tion’ to a particular ‘need.’”27 These convergences show, according to
Conway Morris, that the types of organisms that can evolve are severely
constrained. Biological evolution is quasi-directed, or constrained,
towards certain recurrent biological properties.28 Thus Conway Morris
contests Gould’s thesis: “Rerun the tape of life as often as you like, and
the end result will be much the same.”29
The directionality towards certain biological properties exhibited in
various convergences incorporates, as it were, contingency but renders it
irrelevant for the evolutionary outcome. Conway Morris explains:
[T]he constraints of evolution and the ubiquity of convergence

make the emergence of something like ourselves a near-inevitability.
Contrary to received wisdom and the prevailing ethos of despair, the
contingencies of biological history will make no long-term diference
to the outcome.30
On this alternative view, then, humans are not an improbable event but
an inevitable product of evolution. In contrast to de Duve’s ‘gospel of con-
tingency’ label, Eörs Szathmáry terms positions stressing the inevitability
of human life ‘the gospel of inevitability.’31
But despite their apparent opposition, the gospel of contingency,
when applied to the topic of providence, supports, at least implicitly,
the assumption that providence implies a necessity of the consequent, as
presupposed by the gospel of contingency, precisely by emphasising the
inevitability of the emergence of human beings. For in reply to the gospel
of contingency, what is challenged is the denial of the necessity of the
consequent rather than the fact that providence implies a necessity of
the consequent.
Hence both the gospel of contingency and the gospel of inevitability
appear to agree, in one form or another, on the presupposition that for the
requirements of the doctrine of providence to be met, biological evolution
has to be directed, or substantially constrained, towards the emergence
of specifc traits rendering human beings a necessity or at least a near-
inevitability. They disagree, however, on whether this is the case. However
likely the outcome has to be in order to fulfl the theological demands,
however high the probability of the emergence of human life needs to be,
the likelihood concerns in each case the consequent, namely, that human
beings will evolve necessarily, inevitably, or most likely.
e. Secularised Providence and Teleology Revisited

When belief in providence, traditionally interpreted as God’s guidance
of his creation, is substituted by belief in progress, predictability, and
evolutionary directionality, as seen to varying degrees in the previous dis-
cussion, one might call this transforming of providence stripping it from
its original religious or theological content, to adopt Alister McGrath’s
terminology, a ‘secularization of providence.’32 Such a secularised notion
of providence is at times merged with or turned against the traditional
theological doctrine of providence, on the assumption that when applied
to the context of biological evolution, divine providence too would imply
or amount to belief in evolutionary directionality. According to these
authors, the emergence of the human species has to be necessary or at least
has to be given a very high probability, to fulfl the supposed theological
demands. Evolution has to be end-directed or teleological for theological
reasons.
We can express this conviction also in teleological terms, by distinguish-
ing two sorts of teleology. For this purpose, we frst have to distinguish
the directionality towards an end, to which we commonly refer by the
term ‘teleology,’ from the attainment of the end. If the end is attained with
certainty, we may speak of a ‘necessary teleology,’ that is, a directional-
ity that reaches its end by necessity. But if the end is attained with some
probability, then we could describe this end-directionality as ‘contingent
teleology,’ that is, a directionality that reaches its end with some probabil-
ity, where the attainment of the end is neither impossible nor necessary; it
is a matter of contingency. Note that the former might not be considered
a form of teleology by some wanting to reserve the term for the latter
case. Be that as it may, on the Thomistic account, to which we shall turn
next, processes or agents directed towards their end by necessity are also
counted among the agents acting for an end (propter fnem).33 In fact, the
scholastic axiom states that every agent acts for an end, where agency
includes all kinds of efcient causes.
Accordingly, we can conclude that various authors in the contemporary
debate, including Ruse, Gould, and to some extent Conway Morris, seem
to posit for a Christian providential view a necessary teleology or at least
a contingent teleology with a high probability coming close to a necessary
teleology, due to their shared assumption that providence implies a neces-
sity of the consequent: ‘If God wills human beings, then human being will
evolve necessarily,’ where evolution itself, or the evolutionary processes
by which human beings come about naturally, has to be necessary. Or at
least the ends of evolution have to come about with a very high probabil-
ity, in which case providence would presumably imply no certainty, but
at least a high probability. Henceforth I shall label this end-directionality
of the biological evolution as such ‘teleology of nature,’ where ‘teleology’
stands for a necessary or contingent directionality towards an end, and
‘nature’ for the efect of the act of God’s creation, both synchronically and
diachronically, as a whole.
On the outlined view, therefore, teleology matters to the extent to which
the directionality of evolution matters. The argument goes as follows. If
there is no directionality towards the emergence of the human species in
evolution, a necessity or high probability that humans will evolve – in
short, if there is no teleology of nature – then theology is in trouble. Yet
the context in which contingency can possibly thwart or at least under-
mine providence is one where the posited necessity or quasi-necessity of
providence is taken to be a necessity of the consequent. Otherwise the
necessity would attach to the consequence leaving open, as we shall see
in more detail later, the possibility of a contingent consequent. Generally
speaking, then, positions in the dialogue between science and religion tak-
ing the contingency of evolution to speak against providence presuppose,
in one way or another, that the necessity associated with providence is a
necessity of the consequent.
With this discussion of a secular account of providence, I shall now
turn to a traditional theological account of providence. In order to show
that on a traditional theological view providence presupposes teleological
natures rather than a teleology of nature, on the natural level, I will make
use of the theory of providence of Thomas Aquinas, who objects to the
view that providence implies a necessity of the consequent.
2. Divine Government Through Natures and Powers:

Aquinas’s Account of Providence
On a theological account, providence is generally taken to refer to God’s
guidance of and care for his creation. The doctrine of providence is thus
usually distinguished from the doctrine of creation, which concerns the
bringing into existence of creation – and at times also from conservation,
or the upholding of creation in being, although on the Thomistic account,
conservation as the continuation of the act of creation is counted among
the efects of divine government and hence falls within the doctrine of
providence.34
a. Divine Providence and Government

According to Thomas Aquinas, providence has to do with the ordering
of things to an end (ordinare res in fnem).35 God orders in his providence
all created things to an end.36 Two aspects of this divine ordering have to
be distinguished.
First, ‘providence’ (providentia) in the narrow sense of the word denotes
the eternal plan of, or reason for, the ordering of all created things to an
end. Aquinas defnes providence as the “reason of the order of things to
an end” (ratio ordinis rerum in fnem). This reason of the end-directed
ordering of all creatures is immediate to God; it is the eternal plan of the
providential ordering as existing in the divine mind.37
By contrast, the second element of Aquinas’s theory of providence con-
cerns the temporal “execution of this order” (executio huius ordinis),
to which Aquinas refers by the name ‘government’ (gubernatio). Divine
government is the execution of the eternal divine plan in space and time,
the temporal realisation and causal manifestation of the divine ordering
of all creatures to their end and perfection as eternally existing in God.
Contrary to providence as the divina ratio, government as the executio
allows for a mediation through creaturely, or secondary, causes.38
Aquinas distinguishes two forms of government. God can bring about
the execution of the providential ordering either (i) by himself without any
created causes or (ii) mediated through secondary causes. The frst and
extraordinary way we call miracles, the second and ordinary way, which
Aquinas at times terms ‘fate,’ is essential to understanding his vision of
divine government.39 Secondary causes administer God in the execution;
they are ministers of divine government.40 As Aquinas stresses time and
again, “secondary causes are the executors of divine providence.”41
That God has immediate providence over all things does not exclude
secondary causes, which are the executors of this order [ST I.22.3
ad 2]; they are also called ministers, insofar as by their operation
they execute . . . the order of divine providence [SCG III.79]; divine
providence does not exclude other causes: rather, it orders them so
that the order which is determined by it is imposed on things; and so
secondary causes are not incompatible with providence, but instead
they execute the efect of providence [SCG III.96]; divine providence
executes its efects through mediating causes [ST I.116.2]; the opera-
tion of providence, whereby God operates in things, does not exclude
secondary causes but is executed through them, insofar as they act
through the power of God [SCG III.72]; the perfection of divine
providence therefore requires that there are mediating causes as its
executors [SCG III.77]. Therefore, God himself disposes all [things]
immediately through his providence: but whatever [secondary causes]
are said to be providers under him are executors of his providence
[SCG III.76].
The notion of providence presented by Aquinas, both as providence and

as government, is essentially teleological, or end-directed. “Order,” Aqui-
nas writes, “is the proper efect of providence,”42 and the order provi-
dence implies is clearly end-directed. Providence concerns God’s ordering
of all things to their ends, where ‘providence’ denotes the divine reason
of this end-directed ordering and ‘government’ the execution thereof. The
question that emerges in this context is as follows. Are these ends to
which God orders and directs creatures purely external to them, ends
merely imposed on creatures, or are these ends also in a sense internal
to creation; that is, are these ends in any relevant sense the ends of the
creatures themselves? Put diferently, is the end-directionality resulting
from God’s providential guidance purely externally imposed on or also
immanent to creatures?43
In the frst case, God would direct created things to ends, by externally
imposing these ends. In the second case, God would direct internally
directed things to ends which creatures seek themselves. In the latter case,
the providential ends will also be immanent in the sense that creatures
tend towards these ends of themselves, wherefore creatures and hence
creation will become end-directed itself. Let us call the externally imposed
directionality ‘extrinsic teleology’ and the immanent directionality ‘intrin-
sic teleology.’44
With regard to divine government, Aquinas endorses intrinsic teleol-
ogy.45 He states, “It would be contrary to the notion of providence, if
things subject to providence were not to act for an end (propter fnem):
since it is [the function] of providence to order all to an end.”46 So Aquinas
believes that things subject to providence, that is, due to the universal
nature of divine providence in efect everything created, act for an end.
They are end-directed objects exhibiting an intrinsic teleology, namely,
by receiving their nature, or essence, from God – what Aquinas calls
an ‘impression from God.’47 Note that this applies to all created things,
including not only voluntary but also involuntary agents exhibiting a nat-
ural inclination only. Due to this impression from God – their imparting
of nature – all agents, whether voluntary, sensitive, or natural, “attain to
that [end] to which they are divinely ordered, as if of their own accord.”48
b. Divine Government and Teleological Natures:

God’s Execution of His Providence
Providence and government are teleological notions – they denote a divine
end-directed ordering – and so are the natures through which God ordi-
narily governs and directs creatures to their ends, on Aquinas’s view.49
In this context, Aquinas explicitly identifes the impression from God
by which he directs all things to an end with their nature: “that which
creatures receive from God is their nature.”50 More precisely, God directs
towards ends by imposing (substantial) form on creatures, thus endowing
them with a nature. Aquinas discusses the link between form and nature
in a representative passage on imparting natures in the natural world thus:
The impression of an agent does not remain in the efect after the
action of the agent has ceased, unless it merges into the nature of the
efect. For the forms of things generated, and their properties, remain
in them after generation until the end, because they become natural
to them.51
Agents generating efects by imposing substantial forms on things, thus

constituting them, leave a lasting impression on their efects. Similarly,
what Aquinas calls the ‘impression from God’ on creatures remains in
them, as indicated in the passage, because it is their nature. As Edward
Feser highlights, “The act of ordering a natural cause to its typical efect
just is the imparting to it of a certain nature or substantial form.”52 Crea-
tures, then, are directed towards their end by virtue of their natures, or
substantial forms.
These natures or essences of things are dynamic, or teleological, and
God-given. I mean here primarily things that have an intrinsic principle of
motion and rest.53 Aquinas states, “nature is nothing other than the ratio
of a certain art, namely, the divine [art], endowed to things, by which
these things are moved to a determinate end.”54 This theological add-on,
or the embedding of natures in a theological context, does notably not
change the Aristotelian notion of nature. These natures through which
God governs and directs all creatures to their due ends are intrinsically
teleological: creatures act for an end, by an intrinsic teleology. Therefore,
an essential metaphysical or natural-philosophical element of Aquinas’s
theory of government are teleological natures in the Aristotelian sense, a
divinely guided ‘teleology of natures.’
In particular, Aquinas names three prerequisites for something naturally
to attain an end: (1) a nature proportionate to the end; (2) an inclination
to the end consequent upon the nature; and (3) an operation or movement
toward the end.55
First, as we have seen, natures are created by God and teleologically
ordered to an end. Aquinas reiterates that in establishing nature (insti-
tuente naturam), God “has given each nature its proper inclination con-
venient to it.”56 This is due to a thing’s substantial form. John Wippel
observes that on the Thomistic account, “what ultimately causes the
form of a natural agent is also responsible for that agent’s inclination to
its given end.”57 For example, Aquinas explains, “What gives something
some principle gives it all that results from it: as the cause that gives to an
elemental body gravity gives it downward motion.”58
Second, Aquinas states that “every inclination of any [sort of] thing
. . . is nothing other than some impression from the frst mover,”59
and that “there is no natural inclination, unless from God who gives
the nature.”60 On Aquinas’s theory, then, inclinations are grounded in
natures and follow upon forms: upon (i) natural forms follow purely
natural inclinations; and upon apprehended forms, if apprehended (ii)
by the senses, follow sensitive inclinations, but if apprehended (iii) by
the intellect, intellectual inclinations follow.61 Aquinas calls the frst
‘natural appetite,’ the second ‘sensitive appetite,’ and the third ‘rational
appetite,’ or will.
Third, natural operations result in an important sense from, and are
specifed by, this tripartite of inclinations. As analysed by Aquinas, they
have the following structure: forma (natura) – inclinatio – operatio. From
each specifc form follows a specifc inclination, and from each inclination
a particular operation or movement. The various appetites are specifc
forms of the inclination linking form and operation.62 “For in natural
things,” Aquinas observes, “the intention of the end belongs to the agent
according to its form, by which the end is convenient to it: hence it is
necessary that the natural thing tends to the end according to the power
of the form.”63 Thus, he who gives creatures their nature or substantial
form and hence, as we shall see next, their powers rooted in their form,
gives them their directionality.
In summary, in arguing that creatures are themselves teleologically
ordered towards their providentially ordained ends, by God’s imparting of
natures in the Aristotelian sense, Aquinas posits a divinely guided intrinsic
teleology as regards God’s government. On this view, creatures, through
which God executes his providential ordering, have natures that are teleo-
logical; they are directed towards an end. To see how these natures, or
essences, by their form ground powers and how, on Aquinas’s account,
these powers are the intrinsic principle of the operation of creatures and
yet also due to God’s action, we need to turn to the doctrine of primary
and secondary causation.
c. Divine Government and Intrinsic Powers:

Primary and Secondary Causes
The doctrine of primary and secondary causation is a traditional theory
about the relation, interaction, or connectedness of divine and creaturely
causation. Aquinas teaches that agents act by virtue of their powers. In
scholastic terminology, power is the principle of operation. But there is a
diference between the agency of God and creaturely agents, also called
the ‘primary cause’ (causa prima) and ‘secondary causes’ (causae secun-
dae). For Aquinas, power (virtus), which is the principle of operation
(operatio), proceeds from an essence (essentia); and operation, which is
on this view the actualisation of a power, proceeds in turn from its prin-
ciple power. Aquinas states, “Each power proceeds from the essence, and
operation from power; whence whose essence is from another, necessarily,
that power and operation is from another.”64 In other words, if an essence
is dependent on something else, so are the power and operation resulting
from that essence.
On Aquinas’s view, then, God is his power, whereas creatures have
power. That is to say, they have intrinsic powers grounded in their
essence – or substantial form, for that matter. Yet since the essence is
dependent on something else, so are the powers rooted in this essence.
Thus creaturely powers are derived from the power of God, by which
creatures receive their nature and hence their powers, in ways to be shown
later. This is because for Aquinas God is his essence and hence, according
to the earlier line of argument, his power and operation, but the essences
of creatures depend on God – their essence is not to be existence, as
Aquinas would say – and so do their powers and operations.65 In this
sense, therefore, creatures act not only by virtue of their power, but also
by virtue of the power of God: “every agent acts by the power of God.”66
Aquinas adds elsewhere:
But because secondary causes do not act unless by the power of the
primary cause (per virtutem primae causae) . . . it is necessary that all
other agents through which God carries out the order of his govern-
ment act by the power of God himself.67
God’s providential order is executed through secondary causes acting by

the power of God, but it is important to qualify this claim, not to dimin-
ish the genuine causation of creatures.
On Aquinas’s view, secondary causation is ‘caused causation.’68 What
this means Aquinas explains in his doctrine of divine application. Accord-
ing to Aquinas, God (a) creates and (b) conserves in being all creaturely
powers, by virtue of which creatures act and are real agents. But God also
and necessarily (c) applies these created and conserved intrinsic powers to
act, and does so also (d) in an instrumental manner – that is to say, God
the principal cause makes use of the intrinsic powers of the instrument
beyond what it could achieve on its own accord and through its own pow-
ers.69 In a representative discussion of divine application, Aquinas states:
The power of the inferior agent, however, depends on the power of

the superior agent, inasmuch as the superior agent [a] gives the power
itself (dat virtutem ipsam) to the inferior agent by which it acts; or
[b] conserves it (conservat eam), or also [c] applies it to act (applicat
eam ad agendum).70
And he summarises his position elsewhere thus:
In this manner, therefore, God is the [primary] cause of each action

inasmuch as he [a] gives the power to act (dat virtutem agendi), and
inasmuch as he [b] conserves it (conservat eam), and inasmuch as he
[c] applies the action (applicat actioni), and inasmuch as [d] every
other power acts [instrumentally] by virtue of his power.71
With his doctrine of divine application Aquinas argues that God’s primary
causality is not restricted to the creation and conservation of creatures
and their operations, but God also operates in creaturely operations. In
(a) giving to creatures and (b) upholding in being creaturely powers, God
bestows on creatures intrinsic powers, by which creatures can operate
naturally. That is to say, they have an intrinsic principle of their opera-
tion. So although these powers are derived from God, they are intrinsi-
cally grounded in their (substantial) form. In fact, Aquinas stresses that
the form is also a principle of operation, namely, as he puts it, the form is
that which the agent applies to action, by the power grounded in its own
form.72 When applied to God, Aquinas expressly teaches, “God not only
[a] gives forms (dat formas) to things, but also [b] conserves them (conser-
vat eas) in being, and [c] applies them to act (applicat eas ad agendum).”73
Secondary causes always act in virtue of the power of the primary cause,
and this means, according to Aquinas, that God moves secondary causes
to act in accordance with their mode of causation grounded in their sub-
stantial form. On Aquinas’s view, each operation proceeds from a power
grounded in, or existing in virtue of, form. Yet all operations of creatures
and hence natural causation as such is the exercise of an intrinsic power
created, conserved, and applied to act by God.
Consequently, secondary causes are real causes because they act by vir-
tue of their intrinsic powers. But they are secondary in the sense that their
causation, including their directionality, constantly and actively depends
on the agency of God the primary cause. Creatures are agents but have
powers – powers which are intrinsically grounded in their substantial form
but ultimately derived from God. We can therefore ascribe real agency to
creatures insofar as they act by their own power, but only to the extent to
which this power is enacted by God’s primary causation. For this reason,
each manifestation of a power, each operation of creatures, requires (c) a
divine application of these intrinsic powers to act. This divine application
moving the creaturely power to action accounts, in scholastic terms, for
the diference between posse agere and actu agere, or the being able to act
yet not actually acting, and actually acting.
In short, divine causation is primary in the sense that secondary causa-
tion only counts as genuine causation due to God’s (a) causing and (b)
upholding the powers of creatures, as well as his (c) applying these thus
established, intrinsic powers to act. In this way, on Aquinas’s theory,
secondary causes presuppose the causing of a primary cause for their
very causation.
d. From a Teleology of Natures to a Teleology of Nature:

God as Universal Cause
In contemporary debates about the directedness of biological evolution,
various scholars assume, as shown earlier, that the Christian view of
nature and the doctrine of providence in particular requires an overall
directionality of biological evolution as such, or of nature as a whole – a
teleology of nature. In contrast to this view, I have been arguing that on
a traditional account of providence as presented by Aquinas, teleological
natures and powers in the Aristotelian sense sufce on the natural level.
Put diferently, on the theological account, providence presupposes a tele-
ology of natures and powers rather than a teleology of nature.
On this alternative view, God executes his providential order primar-
ily through creatures, which act by their intrinsic powers grounded in
their substantial form, or teleological natures. As in contemporary powers
ontology, these intrinsic powers (created, conserved, and applied to act
by God) tend naturally towards certain manifestations.74 These tendencies
are, as we have seen, expressed through the notion of inclinations. Thus,
both the natures and powers through which God governs and directs cre-
ation are teleological. On the proposed account, these teleological natures
and powers are sufcient on the natural level for a robust notion of divine
providence, for the following reasons.
To use traditional language, frstly God is a universal cause. A univer-
sal cause in this context is a cause whose causality extends to each and
every other cause and its causal manifestations.75 Because on Aquinas’s
view every secondary cause depends for its very causality on God the
primary cause, God’s causality extends to every secondary cause and all
possible manifestations of each and every natural power, as a primary
cause giving, upholding, and applying to act all powers of every creature.
Hence all created causes fall within the causal order of the primary cause.
Consequently, not only the causes that are naturally related to each other,
but also those that are naturally unrelated are causally related in God’s
providential order. And since natures and their powers are teleological on
the outlined view, in operating in all agents under God’s providence – by
giving them powers, by upholding them in being, and by applying them
to act – God not only causes their causation but also directs in a universal
manner all creatures to their intrinsic ends.
What is more, in applying these intrinsic powers rooted in a thing’s
form instrumentally, God orders and directs creatures to ends beyond
their nature. Thus, God can achieve, as a principal agent using natures
instrumentally, an ordering of nature as a whole – a teleology of nature –
through directing the various teleological natures to ends beyond what
they could achieve of their own, beyond their natural order and yet
through their natures.
Thus, on Aquinas’s view, creatures can fall under God’s providence
in a twofold sense. Things created with a nature, or natures, can be
subject to providence (1) as an end (fnis) to which something, whether
something else or they themselves, are providentially ordered (ad quod
aliquid ordinatur); or alternatively, if they fall short of attaining their
end, merely (2) as to an end (ad fnem), that is, as something that is
providentially ordered to some other end (quod ad alterum ordinatur).76
Only the former is directly intended and immanently ordered to a par-
ticular end by God’s providence; the latter goes beyond the directional-
ity of the agent endowed with a nature and indicates cases where God
governs creatures beyond the teleology of individual natures. The latter
case can in this instance be used to argue for and secure a teleology of
nature on the proposed account of providence operating on teleologi-
cal natures and powers, even if this teleology of natures falls short of a
teleology of nature on a natural level. This is the case precisely to the
extent that God makes use of the powers of creatures instrumentally,
directing them to ends beyond what they are ordered and could achieve
of their own accord.
In summary, then, there is what I called the teleology of natures. Things
like substances and biological organisms are ordered to particular ends,
by God and through their natures and powers. In addition to these teleo-
logical natures and powers, there is also something like a teleology of
nature. The various tendencies of individual substances might also be
ordered to some further end, as parts to a whole. Whatever the details of
this second aspect might be, we can say that on Aquinas’s account, God
orders in his providence and through his government not only creatures,
by their natures and through their powers, to their individual ends, but
also creatures and their ends to other, extrinsic ends. So in an important
sense we have a teleology of nature on this account as well, but as a
second-order ordering. This second-order ordering is itself a directing of
teleological natures and powers in the Aristotelian sense, by God in his
universal providence.
e. Providence and Necessity Revisited:

God as Transcendent Cause
The reasonable extent to which any potential lack of a teleology of nature
can be accounted for by God’s ordering and directing these natures to fur-
ther ends beyond their intrinsic powers and ends, therefore, bolsters the
thesis that teleological natures sufce on the level of secondary causes. In
God’s government there emerges a second-order teleology of nature with
regard to God the primary cause, which bring us fnally back to the topic
of providence and necessity and the question of whether this teleology is
a necessary teleology.
As Bernard Lonergan has pointed out in a watershed publication, on
Aquinas’s account, “providence was [not] certain in all cases because it
was certain in each”; no necessary teleology is demanded here on the level
of secondary causes:
But to St Thomas providence was certain in each case because it was

the cause of all cases: the [secondary cause] moves the moved if the
pair are in the right mutual relation, disposition, proximity [(divine
application)]; the [secondary cause] does not, if any other cause pre-
vents the fulflment of this condition; but both the combinations that
result in motion and the interferences that prevent it must ultimately
be reduced to God who is universal cause, and therefore divine provi-
dence cannot be frustrated.77
So Aquinas endorses a holistic view, where providence is certain not because

each event is necessitated, implying a necessity of the consequent (‘if God
wills p, then necessarily p’), but rather because God is active in every mani-
festation of a causal power. According to Aquinas, as shown earlier, all
secondary causes, whether voluntary or involuntary, are causally related to
God the primary cause. In his providence God orders and directs all things
to ends, including those which are causally unrelated on the level of second-
ary causes. These ends can be divinely ordered even if no teleology and in
efect no causal relation at all emerges on the natural level, that is, on the
level of secondary causation abstracting from God’s ordering of things as a
primary cause. Thus God is the primary cause not only of the necessities of
nature, but also of the various contingencies of the created world, ordering
both to their respective ends, in the mode proper to their causality.
And this brings us to the second unique characteristics of primary cau-
sation. God is not only a universal cause but also a transcendent cause,
on Aquinas’s view. A transcendent cause in this instance is a cause whose
causality extends to and determines the causal modality of secondary causes.
As a transcendent cause, God causes not only (a) some efect but also (b) the
causal mode of this efect; that is, he determines whether the efect comes
about necessarily or contingently.78 This means for Aquinas that although
God’s efcacy cannot be impeded, he can bring forth something contingent.
As Aquinas puts it, the incompatibility of irresistible cause and contingent
efect holds in efect only within the causal nexus of secondary causes, to
which the causal modalities and corresponding specifc characteristics of
contingent and necessary causation properly applies. As primary cause tran-
scending this order, thus, God’s irresistibly willing an efect does not neces-
sitate the efect. Harm Goris comments, “A created necessary cause excludes
the contingency of the efect, but the uncreated necessary cause does not. On
the contrary, it constitutes the modality of the efect.”79
In scholastic terminology, as the cause of being, God is outside the
order of beings (ordo entium), and as such, God causes not only being
but also the essential diferences of being (diferentia entis), including
necessity and contingency. Thus he constitutes the very causal order
of necessity and contingency among secondary causes. But whatever
causes the order of beings itself and hence the corresponding law of
necessity and contingency, as Aquinas puts it, can irresistibly bring forth
a contingent efect, if the incompatibility of unfailing cause and con-
tingent efect holds only within the order of beings, where the law of
necessity and contingency applies, as Aquinas argues.80 God as primary
cause is the only agent with these specifc characteristics.
Accordingly, Aquinas stresses, “the efect of divine providence is not
only that something happens in whatever mode, but that something hap-
pens either contingently or necessarily.”81 God’s willing and causing some-
thing as primary cause does not make the efect come about by necessity;
but rather, every efect happens necessarily the way God wills and causes
it to be. The causal modality of secondary causes, including contingency
and necessity, is part and a function of God’s providence, on Aquinas’s
view. God orders in his providence and through his government creatures
to their ends through necessary and contingent secondary causes alike:
necessarily, if God wills something to happen necessarily, then it will hap-
pen necessarily; and likewise, necessarily, if God wills something to happen
contingently, then it will happen contingently.82 As Aquinas puts it,
consequents (posteriora) have necessity from their antecedents (prior-

ibus), according to the mode of the antecedents. Hence, those [things]
which are efected by the divine will have that [kind of] necessity that
God wills them to have, that is, either absolute [necessity] or merely
conditional [necessity]. And so not all [of those things efected by the
divine will] are necessary absolutely.83
So on this traditional theological account of providence, the fact that

the conditional ‘if God wills human beings to evolve, then human beings
will evolve’ is true and necessary from a Christian point of view does not
imply that what God wills, namely, the evolution of the human species, is
absolutely necessary. Rather, the human species can evolve contingently,
if willed by God. The necessity providence implies here is a necessity of
the consequence, where the consequent can be either necessary or contin-
gent. Since contingency is an efect of providence and an essential part of
government, therefore, on Aquinas account, providence does not imply
a necessity of the consequent but merely a necessity of the consequence.
The doctrine of divine providence does then not so much demand the
necessity or near-inevitability of evolution or that alternative replays
would resemble the present tape of life, but rather that the actually exist-
ing tape of life is guided to particular ends, including, but not limited to,
the emergence of human beings, by God. What accounts for the teleology
of nature, on this account, is not a necessary teleology on the natural
level, a necessity of the consequent, but a conditionally necessary and
contingent teleology conjoined with the belief that all is subject to God’s
universal and transcendent providence.
To conclude, apart from miracles, God’s providential action presupposes
and works through creaturely operations, on this account. Creatures have
teleological natures and intrinsic powers through which God directs them
to their respective ends and perfections. In order to analyse the resulting
end-directionality of nature, we have to distinguish the intrinsic workings
of natural causes by virtue of these natures and powers, and the workings
of God in and through these natures and powers.
A traditional way of doing so is in terms of primary and secondary
causation. By showing that God is a universal cause giving, upholding,
and applying the intrinsic powers of each and every creature to act, we
can causally trace back everything to God as primary cause. Since God
causally connects – as primary cause – all secondary causes, which are
true and directed causes of themselves, precisely to the extent that God
works in them, every unconnected or thwarted secondary cause falls
within the causal nexus of the primary cause and is hence part of his
divine government. Providence is certain not because each and every
case is certain considered on its own but rather because God is the
primary cause of each and every constellation of power manifestations.
Moreover, in instrumentally applying creaturely powers to act, God can
bring about efects beyond the intrinsic directionality of natures and
thus providentially direct creatures to ends beyond their own intrinsic
ends and teleology.
3. Conclusion
The traditional doctrine of divine providence, at least as presented by
Thomas Aquinas, presupposes teleological natures in the Aristotelian
sense rather than a teleology of nature, on the level of natural causes.
This view advocated in this chapter goes contrary to the claim, common
among scholars in the feld of science and religion, that the doctrine of
providence requires nature as a whole to be teleological. Despite their
opposing views on the question of the repeatability and directedness of
biological evolution, both Stephen Jay Gould and Simon Conway Morris
seem to assume, in one way or another, that a providentially guided world
implies a directedness towards an end which it attains either necessarily
or at least with a very high probability. What is more, this necessity or
quasi-necessity appears to be interpreted as a necessity of the consequent,
as a necessary teleology of nature.
By contrast, on the traditional theological account discussed in this
chapter, the necessity of providence is viewed as a necessity of the conse-
quence, not of the consequent. What God wills cannot not happen, and
in this sense is necessary. But since God – as transcendent cause – wills
not only things to happen but also their mode of coming to be, what God
wills will come about either necessarily or contingently, according to his
will. By implication, then, on this view, if God wills human beings to
evolve, they will evolve precisely in the manner God wills them to evolve.
Necessarily, if God wills human beings to evolve, then human beings
will evolve. But the consequent, the way they evolve, might in turn be
either necessary or contingent, according to his will. Hence God’s willing
human beings to evolve does not imply that humans evolve necessarily
rather than contingently. In God and only in God, the necessity of the
consequence of his willing and acting providentially is compatible with
the contingency of the consequent.
The providence of God is executed through necessary and contin-
gent causes alike. As primary cause, God is in his divine government
causally related to all (secondary) causes in nature, to each and every
power and their manifestations in a universal and transcendent way.
What accounts for the teleology of God’s government of all things,
including biological evolution, is thus, on God’s part, his universal and
transcendent causation as primary cause and, on the part of creatures,
their teleological natures and powers making them secondary causes,
through which God orders and directs all creatures to their ends and
perfections in his providence. On this traditional theological account,
then, teleological natures in the Aristotelian sense and intrinsic causal
powers would appear to sufce on nature’s part for a robust account
of providence.84
Notes
1. Stephen Jay Gould, Wonderful Life: The Burgess Shale and the Nature of
History (New York: W.W. Norton & Company, 1989), 48.
2. Gould, Wonderful Life; Simon Conway Morris, The Crucible of Creation:
The Burgess Shale and the Rise of Animals (Oxford: Oxford University Press,
1998).
3. The term ‘Cambrian explosion’ refers to the rapid diversifcation of animal
life during the period roughly between 550 and 485 million years ago. See
Conway Morris, The Crucible of Creation, 31–32.
4. Gould, Wonderful Life, 48.
5. Recent review articles discussing new evidence from studies of convergent
evolution, evolutionary genetics, especially empirical studies of ftness land-
scapes, and experimental evolution, including parallel and analytic replay
experiments, and historical diference experiments testing microevolution,
include Zachary D. Blount, Richard E. Lenski, and Jonathan B. Losos, “Con-
tingency and Determinism in Evolution: Replaying Life’s Tape,” Science 362
(2018): 5979; Zachary D. Blount, “A Case Study in Evolutionary Contin-
gency,” Studies in History and Philosophy of Biological and Biomedical
Science 58 (2016): 82–91; Virginie Orgogozo, “Replaying the Tape of Life in
the Twenty-First Century,” Interface Focus 5 (2015): 20150057; Alexander
E. Lobkovsky and Eugene V. Koonin, “Replaying the Tape of Life: Quan-
tifcation of the Predictability of Evolution,” Frontiers in Genetics 3 (2012):
246.
6. Orgogozo, “Replaying the Tape of Life.”
7. Michael Ruse, “Darwinian Evolution and a Providential God,” in Abraham’s
Dice: Chance and Providence in the Monotheistic Tradition, ed. Karl W.
Giberson (Oxford: Oxford University Press, 2016), 312.
8. Ruse, “Darwinian Evolution,” 314–316.
9. Ruse, “Darwinian Evolution,” 323.
10. Ruse, “Darwinian Evolution,” 323–325.
14. ST I.116.3. All texts cited from Thomas Aquinas are taken from S.
Thomae de Aquino, Opera Omnia: Recognovit et instruxit Enrique Alar-
cón automato electronico Pompaelone ad Universitatis Studiorum Navar-
rensis aedes a MM A.D. Pamplona 2019. All translations from the Latin
are mine.
15. ST I.116.3.
16. SCG I.85.
17. Eleonore Stump, Aquinas (New York: Routledge, 2003), 123. Stump goes on
to explain: “Aquinas’s conditional necessity in God is thus like the necessity
of the present [illustrated by Socrates’s running], except that the present in
question is the timeless present which characterises all of God’s life at once.”
20. As John Beatty points out, the fact that contingency can either mean that
diferent outcomes are possible from the same initial state (contingency per
se) or that diferent prior states lead to diferent outcomes (contingency upon)
complicates matters in Gould’s argumentation. See John Beatty, “Replaying
Life’s Tape,” The Journal of Philosophy 103 (2006): 336–362; John Beatty,
“What Are Narratives Good For?,” Studies in History and Philosophy of
Biological and Biomedical Sciences 58 (2016): 33–40.
21. Christian de Duve, “The Gospel of Contingency,” Nature 383:6603 (1996):
771.
22. De Duve, “The Gospel of Contingency,” 771.
26. Conway Morris, The Crucible of Creation, vii–viii.
27. Simon Conway Morris, Life’s Solution: Inevitable Humans in a Lonely Uni-
verse (Cambridge: Cambridge University Press, 2003), xii.
28. Conway Morris, The Crucible of Creation, 202.
29. Conway Morris, Life’s Solution, 282.
30. Conway Morris, Life’s Solution, 328.
31. Eörs Szathmáry, “The Gospel of Inevitability: Was the Universe Destined to
Lead to the Evolution of Humans?” Nature 419:6909 (2002): 779–780.
32. Alister E. McGrath, “The Secularization of Providence: Theological Refec-
tions on the Appeal to Darwinism in Recent Atheist Apologetics,” in The
Providence of God: Deus Habet Consilium, ed. Francesca Aran Murphy and
Philip G. Ziegler (London: T&T Clark, 2009), 194–207.
33. SCG III.1. What does not qualify as teleology on this view, by contrast, are the
exceptions to the rule, as it were, or to the directionality (ut in paucioribus)
that account for the diference between something attaining its end always
(semper et ex necessitate) or merely most of the time (ut in pluribus).
34. ST I.104.
35. ST I.22.4.
36. ST I.22.1–2. Note that divine providence in this sense can only be God’s
ordering of things other than himself, that is, of all creatures, for God has his
end in himself.
37. ST I.22.1.
38. ST I.22.3.
39. ST I.105; ST I.116. The topic of how to categorise God’s action in the subse-
quent Thomistic tradition is also discussed by Edward Feser in this volume.
40. More precisely, Aquinas draws a distinction between executors and ministers,
attributing the latter phrase specifcally to intellectual beings: “for a minister
is like an intelligent instrument; but an instrument is moved by another, and
its action is ordered to another” (ST I.112.1).
41. SCG III.77.
42. SCG III.77.
43. On this topic, see also Edward Feser’s contribution to this volume.
44. Note that by adapting this terminology I refer to the (1) end-directionality and
not to (2) the ends themselves. The ends can in turn either be benefcial for
the substance exhibiting the end-directionality, at times called ‘intrinsic tele-
ology,’ or benefcial for something else, at times called ‘extrinsic teleology.’
On this view, intrinsic teleology means that substances directed to these ends
beneft from them, whereas extrinsic teleology means that the ends to which
these substances are directed are not benefcial for the substances themselves
but rather for something else.
45. This is not to say that God’s providential ordering of creatures can solely
happen through intrinsic teleology – God can also direct, in various ways,
creatures externally (see ST I.105) – but that the doctrine of providence gener-
ally implies an immanent ordering of creatures. On Aquinas’s view, then, God
ordinarily directs creatures to their ends by way of intrinsic teleology, that is,
through their teleological natures.
46. SCG III.74.
47. ST I.103.1 ad 3.
48. ST I.103.8.
49. The topic of teleological natures and causal powers in the Aristotelian tradi-
tion is also discussed by William M. R. Simpson (in Chapter 1), Robert C.
Koons (in Chapter 2), and Edward Feser (in Chapter 13) in this volume.
50. ST I.103.1 ad 3.
51. SCG III.65.
52. Edward Feser, “Between Aristotle and William Paley: Aquinas’s Fifth Way,”
Nova et Vetera 11 (2013): 736. Aquinas comments on the issue: “For if some
agent is not the cause of the form as such, neither will it be the per se cause of
being which results from that form, but it will be the cause of the efect only
in its becoming” (ST I.104.1).
53. For more on nature in the Aristotelian-Thomistic tradition, see Feser’s con-
tribution to this volume.
54. In II Phys. 14. And likewise, Aquinas holds ‘nature’ “to signify the essence of
a thing as it has an order to the proper operation of the thing” (De Ente 1).
55. De Ver. 27.2.
56. De Ver. 25.1.
57. John F. Wippel, The Metaphysical Thought of Thomas Aquinas: From Finite
Being to Uncreated Being (Washington, DC: The Catholic University of
America Press, 2000), 484–485.
58. SCG III.69.
59. ST I.103.8; see also SCG III.88.
60. ST I.111.2.
61. ST I.80.1–2; ST I-II.8.1.
62. De Malo 6; In II De Anima 5; De Ver. 27.2; Super II Cor. 5.2; Super Gal. 5.6;
De Virt. 4.3; Quodl. IV.11.1.
63. SCG II.30.
64. In II Sent. 37.2.2.
65. See, for instance, ST I.3.3–4.
66. SCG III.67.
67. CT I.130.
68. Bernard Lonergan, Grace and Freedom: Operative Grace in the Thought of
St Thomas Aquinas, ed. Frederick E. Crow and Robert M. Doran, Collected
Works of Bernard Lonergan 1 (Toronto: University of Toronto Press, 2000),
88.
69. SCG III.70; De Pot. 3.7; ST I.105.5.
70. SCG III.70.
71. De Pot. 3.7.
72. ST I.105.5.
73. ST I.105.5 ad 3.
74. For example, the late George Molnar argues that directedness is an essential
feature of powers: “Powers, or dispositions, are properties for some behav-
iour, usually of their bearers. These properties have an object towards which
they are oriented or directed,” namely, their manifestation. Molnar concludes,
“Having a direction to a particular manifestation is constitutive of the power
property” (George Molnar, Powers: A Study in Metaphysics, ed. Stephen
Mumford [Oxford: Oxford University Press, 2003], 60). More generally, Ste-
phen Mumford and Rani Lill Anjum point out that on dispositionalist views,
“Causation . . . involves a tendency . . . towards a certain type of outcome”
(Stephen Mumford and Rani Lill Anjum, Causation: A Very Short Intro-
duction [Oxford: Oxford University Press, 2013], 108). On their own view,
“Powers tend towards their manifestations” (Rani Lill Anjum and Stephen
Mumford, What Tends to Be: The Philosophy of Dispositional Modality
[London: Routledge, 2018], 24). Similarly, Robert C. Koons stresses in this
volume, “Causal powers correspond to teleological structures, since each
causal power is intrinsically ordered to a particular result.” And he adds, “In
fact, whenever a thing acts according to its intrinsic power and potentialities,
immanent teleology exists. Fundamental causal powers as Aristotle conceives
of them are inherently teleological.”
75. In VI Meta. 3; see also ST I.22.2.
76. De Ver. 5.4.
77. Lonergan, Grace and Freedom, 79.
78. De Ver. 23.5; Quodl. XI.3; SCG I.85; SCG III.94; ST I.19.8; CT I.140; Peri
Herm. I.14; In VI Meta. 3.
79. Harm J. M. J. Goris, Free Creatures of an Eternal God: Thomas Aquinas on
God’s Infallible Foreknowledge and Irresistible Will (Leuven: Peeters, 1996),
298.
80. Peri Herm. I.14; In VI Meta. 3.
81. ST I.22.4 ad 1.
82. As Eleonore Stump points out, “On Aquinas’s view, any divine act that is an
instance of free choice . . . is necessitated conditionally, but not absolutely.”
That is to say, “for Aquinas, what is conditionally necessary for God is what
cannot be changed over time in God in consequence of something God wills,
whereas what is absolutely necessary is what cannot be changed across pos-
sible worlds” (Stump, Aquinas, 123 and 124). See also SCG I.83.
83. ST I.19.8 ad 3.
84. This research was supported by the University of Oxford project ‘New Hori-
zons for Science and Religion in Central and Eastern Europe’ funded by
the John Templeton Foundation. The opinions expressed in the publication
are those of the author and do not necessarily refect the view of the John
Templeton Foundation.
16 Divine Lawmaking
Powers and Kinds
James Orr
The intuition that the natural sciences track objective distinctions between
lawful and accidental regularities in the natural world has proven difcult
to dislodge. Anglophone philosophy cannot be faulted for failing to try.
For more than century, the dominant orthodoxy of Humean positivism
has insisted that the diference between law and accident is stipulated and
not discovered by our best scientifc theories. The historical pedigree of
this sceptical stance stretches back to the sharp reaction against scholastic
philosophy and the wide array of Aristotelian ideas that permeated it,
frst among which was its doctrines of substantial form, causal powers,
and natural kinds. A second factor motivating scepticism towards nomic
realism was its perceived complicity with theological commitment. A.J.
Ayer once claimed, ‘[o]ur present use of the expression “laws of nature”
carries traces of the conception of Nature as subject to command. . . .
[T]he sovereign is thought to be so powerful that its dictates are to be
obeyed . . . [T]he commands which are issued to Nature are delivered with
such authority that it is impossible that she should disobey them.’1 More
recently, Jonathan Schafer has characterised one version of nomic realism
as the ‘remnants of a dubious theology.’2 It is often overlooked that these
two sources of nomic scepticism emerge from two opposed metaphysical
and theological frameworks. On the scholastic view, regularities in nature
emerge ‘from below’ in virtue of the intrinsic properties of the entities
whose actions and interactions law-statements describe. Nature’s patterns
exemplify the creative blueprint in the divine mind towards which all things
were ultimately oriented. But when scholastic philosophies of nature came
to be displaced in the course of the seventeenth century, nature began to be
construed more as a modally inert mechanism manipulated ‘from above’
by a divine legislator whose volitions were treated as laws.
This chapter explores two connected realist neo-Aristotelian accounts
of lawhood, one based on causal powers and the other on natural kinds.
Defenders of both approaches typically insist that they do not ofend natu-
ralistic scruples: the disagreement over powers and kinds, it is claimed,
lies not between naturalists and non-naturalists, but rather between
an austere neo-Humean rendition of naturalism and a more expansive
DOI: 10.4324/9781003125860-20
Divine Lawmaking 409
neo-Aristotelian one. My aim in what follows is to argue that if a theory
of laws based on powers or kinds is to deliver the explanatory goods,
non-naturalistic options should be considered. In particular, I suggest that
accommodating powers and kinds forces a confrontation between the two
leading non-naturalistic alternatives – Platonism and theism; and that
there is a signifcantly stronger case for preferring one kind of theistic
proposal than is generally acknowledged. Such a conclusion is likely to
be grist to the mill of those neo-Humeans suspicious that neo-Aristote-
lian mischief-making splinters the naturalistic consensus that metaphys-
ics should only invoke spatiotemporal entities. It will certainly not suit
those whose ontological preference is for Quine’s ‘desert landscapes.’ But
theism has the resources to address sufciently many of the theoretical
difculties of rival alternatives to warrant serious consideration as the
framework that best provides the supporting metaphysical structure of
the powers network. On a comparative basis, there is at least one way
of making sense of power-laws and kind-laws within the framework of
divine exemplarism that avoids the difculties of more orthodox theories
and their Platonic and quasi-Platonic alternatives.
1. Lawmaking Powers
David Lewis is well-known for deploying the metaphor of a mosaic to
capture the neo-Humean metaphysic of perfectly natural intrinsic cat-
egorical properties.3 One can imagine this mosaic as a work of art in the
abstract style: the connections and regularities that emerge from staring at
the tiles of the mosaic make up images that the spectators conjure in their
minds as they stare at the tiles. Laws of nature, for example, are simply
those axioms in a deductive system of generalisations about natural regu-
larities that achieve an optimal combination of simplicity and ft.
Powers holism can also be imagined as a mosaic, albeit in a rather dif-
ferent sense.4 In contrast to the abstract style of the neo-Humean mosaic,
one can think of the powers mosaic as adopting the artistic style of real-
ism or naturalism, since it depicts a single image in which the tiles of
the mosaic are integrated in an interconnected and intelligible way, so
that the viewer simply grasps the image that is there rather than projects
interpretations onto the disconnected tiles of the abstract mosaic. The
intelligibility of the powers mosaic emerges from an intricate network of
active and passive powers that are reciprocally connected. Unlike the neo-
Humean mosaic, whose tiles represent spatiotemporal events, it depicts
abstract nomological space. The identity of a property is what it is in
virtue of its position as a ‘tile’ in the overall mosaic: every property has a
causal profle that constitutes what efects it will bring about under certain
stimulus conditions and its connections to other properties. On a mosaic
depicting (say) a forest, the diferent colours of each tile – the yellows, the
greens, and the browns – ‘depend’ upon each other to render the image
410 James Orr
intelligible, much as semantic holists argue that the words of a sentence
depend on each other to convey the sense of the sentence.
On the one hand, the resurgence of interest in causal powers triangulates
a dispute about lawhood between neo-Humean regularity theories and
eforts to ground lawmaking in second-order causal necessitation relations
between types of states of afairs.5 Defenders of powers-based accounts
of laws typically side with neo-Humeans in rejecting governing laws in
favour of a ‘lawless’ account, since they hold that powers do all the modal
work that nomic realists invoke laws to explain.6 But in contrast to the
neo-Humean view, powers theorists also insist that real modal work is
done: law-statements pick out objective necessities in nature, and those
necessities are grounded in the irreducible intrinsic properties of things.
Dispositional properties are lawmakers in the sense that they are the truth-
makers for scientifc statements of law.7
On the other hand, defenders of powers reject the shared commitment
of neo-Humeans and necessitarians to quidditism, the view that proper-
ties can be individuated independently of the causal role they play. Active
and passive powers can meet all the explanatory desiderata that overlay-
ing categorical properties with laws construed as necessitation relations,
which until recently was the leading realist alternative to the defation-
ary approach. Intrinsic powers explain the diference between law and
accident, they support inductive inference, they ground counterfactual
resilience, and they can explain the peculiar modal weakness of natural
necessity, a modal status that seems to lie between brute contingency and
strict logical necessity. While contemporary work on powers diverges in
important ways from Aristotle’s own view,8 the revival of powers-based
analyses of laws of nature is one of the key factors driving the renewal
of confdence in neo-Aristotelian metaphysics and philosophy of science.
Since it was hostility to causal powers that played a signifcant role in
the demise of Aristotelian metaphysics in the seventeenth century, this
development marks a signifcant and striking moment in the history of
philosophy.9
2. Lawmaking Kinds
Some theorists sympathetic to powers resist construing dispositional
properties as lawmakers, preferring instead to explain the modal strength
of lawlike regularities in terms of natural kinds to which objects belong
and the essences that anchor those properties. To borrow David Oder-
berg’s convenient distillation of this position, the laws of nature are what
they are in virtue of the laws of natures.10 Aristotle held that patterns of
behaviour are explained by the particular constitution of an object: the
nature of a thing is the source or cause of non-accidental change or rest
in anything to which it belongs,11 a view that Aquinas endorses in the
familiar scholastic idea that every agent intends an end by its action.12
The intentionality involved in the orientation of an agent to an end is not
necessarily cognitive in character, since an agent can act by nature and
not only by intelligence.13 For Aquinas, just as a likeness of the end of an
intelligent act is apprehended beforehand in the mind of an agent, there
is a likeness of the efect that pre-exists in a natural agent: ‘fre begets fre,
and an olive produces an olive.’
In the scholastic idiom that Oderberg favours, the nature of an object is
its substantial form, the intrinsic actualising principle in a substance that
informs its material constituent parts, and the principle in virtue of which
an object is the kind of thing that it is. The nature of an object is logically
prior to the active and passive powers that it grounds.14 On this view,
law-statements describe regularities in the interaction of substances exer-
cising their powers. But the lawmakers are not the properties of an object
as such; rather, the metaphysical basis for lawhood lies in the nature of
objects. Natures crystallise causal powers into a modally robust confgu-
ration in an object, and it is these natures that account for the membership
of their bearers in a natural kind.15 Natural kinds are typically construed
as a fundamental ontological category that carves reality at its joints. The
constellation of necessary and sufcient properties that constitute a kind is
not a stable cluster of properties, as the leading version of conventional-
ism about kinds has argued,16 but is prior in the order of explanation to
the properties it confgures and ontologically irreducible to them. One
defender of this approach parses kinds as substantial universals. On E.J.
Lowe’s view, the laws of nature are grounded not in regularities in the
behaviour of dispositional properties; rather, ascriptions of dispositional
predicates are fxed by their truthmakers, which for Lowe are natural
kinds. Kinds are construed universals that are instantiated by objects, as
opposed to properties and relations, which are instantiated by the modes
that characterise objects.17
Part of what motivates the preference for kinds over powers as law-
makers is the worry that powers are modally untethered to their bearers
and are therefore objectionably mysterious entities. Lowe insists that not
all powers are directed in any event: water’s power to dissolve salt is
reciprocally matched to salt’s liability to be dissolved by water; but ben-
zene’s fammability of a substance is a monadic property whose identity
conditions do not include reference to any extrinsic powers or liabilities.18
He distinguishes relational laws, where the law consists in the existence
of a relation (understood as a universal) between two substantial kinds,
and non-relational laws that consist in the possession by a substance of
a monadic property.19
There are other reasons for preferring to construe kinds as lawmak-
ers rather than individual dispositional properties. It is true that parsing
powers as dispositional universals explains why a particle exemplifying
<negative charge> is disposed to attract subatomic particles exemplifying
<positive charge>. But that does not explain why so many subatomic
412 James Orr
particles of the same type are disposed to interact causally with other
subatomic particles of the same type in uniformly regular ways. Ground-
ing law-statements in kinds addresses that worry.20 Law-statements pick
out diachronic regularities in the collective behaviour of entities. If after
picking a random card from a pack thirteen times I fnd I have thirteen
cards from the same suit in my hand, it would be possible to explain,
with respect to each selection, how I ended up with a card from one suit
rather than any other. There is a sense, then, in which I could ofer a
complete explanation of why I have these thirteen cards in my hand by
pointing to the aggregated set of explanations for each selection. But that
set of explanations would not explain what needs to be explained, since
it would not explain why the thirteen cards collectively exemplify the
regularity of belonging to the same suit. Considered separately, causal
powers can explain the behaviour of their bearers in this limited way, but
they do not explain why so many discrete objects possess an identical (or
nearly identical) confguration of causal powers, and they do not explain
why such objects, in exercising their causal powers, collectively exem-
plify nomologically signifcant regularities. There are countlessly many
fundamental entities endowed with countlessly many confgurations of
causal powers, but the truthmakers for law-statements are those patterns
of resemblance between these confgurations, patterns that distinguish the
behaviour of one kind of fundamental entity from another. Lowe claims
that the reason only some confgurations of powers are found in nature
when so many other confgurations are possible is that law-statements
pick out fundamental entities that belong to one and the same substantial
kind.21 It is not a brute fact that the essential properties of electrons are
confgured in the regular and uniform ways that they are, as regularity
theorists would propose. Nor is it in virtue of an extrinsic law of nature,
as the defenders of the necessitation account propose, since that gets the
order of explanation the wrong way round.
3. Objections to Lawmaking Powers and Kinds

A perplexing feature of lawmaking powers is that determining their iden-
tity involves reference to unrealised or unmanifested states of afairs. If
we analyse the lawlike regularity that water dissolves salt as grounded in
the power of water to dissolve salt, it is hard to avoid the intuition that
water is somehow directed towards salt in a generalisable way – that is,
that one of water’s identity conditions is that it has the power to dissolve
salt even if it does not ever come into contact with it and even in possible
worlds in which salt does not exist. That in turn suggests that passive
powers are reciprocally directed towards active powers: salt’s liability
to be dissolved in water means that salt is reciprocally directed towards
water’s disposition to dissolve it. Active and passive powers are necessary
correlates for the manifestation events involving powers. They are, in C.B.
Martin’s phrase, ‘mutual manifestation partners.’22 The basic problem for
naturalistic powers theorists is that naturalism is committed to the claim
that for every property there exists a particular that instantiates it: there
are no uninstantiated properties. The insistence on instantiation makes
it difcult to make sense of the idea of what ontological status pow-
ers have independently of their manifestation.23 But it also undercuts the
central claim of powers realists that many powers belong to their bearers
intrinsically: water’s lawmaking power to dissolve salt is intrinsic to it,
since it is a power that water possesses independently of the existence or
non-existence of any particulars discrete from water. And yet one can
readily imagine ways in which the exercise of a power could be afected
by altering conditions extrinsic to it: it is easy, for instance, to imagine
water in a saltless possible world in which its power to dissolve salt would
not be anchored in anything.24 I have argued elsewhere that the prospects
of elucidating these identity conditions within a naturalistic framework
are dim.25 The upshot is that accommodating these identity conditions
confronts powers theorists with a dilemma between reverting to a cat-
egoricalist analysis of lawmaking (whether contingentist or necessitarian)
or abandoning the ontological parameters of naturalism altogether.
Lowe departs from standard formulations of the dispositionalist posi-
tion by identifying natural kinds as lawmakers and denying the existence
of dispositional properties in favour of dispositional ascriptions. More
precisely, he takes the truthmakers of law-statements to describe activities
that characterise a kind, which he parses as a substance. But Lowe’s view
leaves it unclear how we are to understand the reciprocity that a powers
ontology seems to presuppose. It is only intelligible to posit a power to
bring about a state of afairs if there is some corresponding liability (or
liabilities) on which the power acts. Water has the power to dissolve salt
if and only if salt possesses the corresponding liability to be dissolved
in water. Now on the monadic view it would of course be the case that
salt is analysed as possessing the liability to be dissolved, given certain
stimulus conditions. But those conditions will include contact with water
because its liability to be dissolved is reciprocally directed to salt’s power
to dissolve it. Thus even if all fundamental lawmaking properties were
monadic, it would still be the case that they would make reference to
properties involved in stimulus and manifestation conditions. An obvi-
ous solution to the conundrum is to abandon monadicity and to accept
that both passive liability and active power are irreducibly polyadic and
directed towards each other. But even if we do not, it remains the case
that Lowe’s lawmaking kinds do not entirely overcome worries about
the involvement of unmanifested states of afairs that confronts standard
dispositionalism. What it is to be a member of kind is to be character-
ised by an activity that still involves reference to universals that are not
instantiated and may never be instantiated. It is difcult to avoid Travis
Dumsday’s conclusion that Lowe’s lawmaking kinds cut sharply against
414 James Orr
naturalism and that the most plausible way of resolving this tension is to
adopt Platonism.26
Including unmanifested states of afairs in the identity conditions of
powers presents power-based theories of lawmaking with special dif-
culties. But there are also more general difculties for powers theorists
wedded to the naturalistic intuition that ontological inventories should be
confned to instantiated entities. It is far from clear, for example, how the
powers theorist committed to the principle of instantiation could explain
the omnitemporal character of laws. For suppose that at some stage in
the early history of the universe there was a range of causal powers or
liabilities that were not yet possessed by any entity and therefore no regu-
larities in the exercise of powers. In that scenario, there would be nothing
that could feature in the content of the relevant statements of fundamental
laws that describe regularities in the exercise of those powers now that
they have emerged. Those fundamental laws would be phantom laws.
Powers theorists could insulate their position against this sort of objection
by taking up an eternalist ontology of time, since that would place each
instantiated property on a par ontologically regardless of when it hap-
pened to be instantiated. Nevertheless, it still seems plausible to suppose
that there are nearby nomologically possible worlds branching of from
the actual world in which lawmaking properties instantiated in the actual
world are forever uninstantiated and would therefore have no ontological
standing even given eternalism.27
The scenario also generates a bootstrapping objection. If powers ground
laws, a question arises: which law-statements could have described the ini-
tial instantiation of powers that hitherto no object possessed? Such pow-
ers could not have been specifed by any law-statement, since according
to the instantiation principle they did not exist.28 On the standard view of
powers, laws describing regularities in the exercise of hitherto uninstanti-
ated (and so non-existent) powers come into existence ex nihilo together
with the powers that ground them. That possibility plainly undercuts the
realist intuition that, in the order of explanation, the lawlike regularities
obtain omnitemporally and exist prior to their instances. Conversely, it
motivates powers theorists to treat the network of lawmaking powers as
real and to abandon the instantiation principle.
Other worries arise for friends of naturalistic lawmaking powers.
There are, for instance, categories of laws that are difcult to explain if
only instantiated powers are posited. There are laws that do not make
any reference to the causal powers and do not appear to be reducible
to causal powers, but merely state functional or mathematical relation-
ships of co-variation. There are functional laws that are expressed in
terms of determinable properties. Such laws quantify over values that are
never instantiated as determinate values and must therefore account for
absent determinate values by inferring what they would be based on the
determinate values that have been instantiated. Moreover, mathematical
quantities feature in the specifcation of the laws that immanent powers
are supposed to ground, in particular in the stimulus conditions and mani-
festation conditions involved in their exercise. Finally, there is the problem
of uninstantiated laws.29 D.H. Mellor has noted that positing such laws
brings with it ‘a very contentious consequence,’ namely that of accepting
the existence of universals that do not depend on their instances.30 To
his credit, Armstrong is explicit about the problem: if one endorses the
principle that only instantiated entities exist, he concedes, one must accept
that ‘strictly there are no uninstantiated laws.’31 Accepting the possibility
of uninstantiated laws means accepting the possibility of laws that do
not depend on the powers of existing objects.32 It is difcult to see how
naturalistic lawmakers could account for such laws.
4. The Turn From Naturalism

The difculties confronting naturalists who seek to press powers into the
service of a realist theory of lawhood are substantive. But which alterna-
tive frameworks could accommodate them? A number of solutions have
been proposed that strive at least for a degree of consistency with natural-
ism. Drawing on Franz Brentano’s infuential recovery of the scholastic
doctrine of psychological intentionality and its famous axiom that inten-
tionality is the mark of the mental, George Molnar has proposed that we
think of powers as endowed with physical intentionality towards existent
or non-existent objects.33 As Molnar recognises, this solution comes close
to panpsychism.34 More recently, Philip Gof has argued that the problems
involved in specifying powers motivates precisely this kind of alternative.35
Alexander Bird suggests that there are ‘unrealized possibilia’ that are ‘akin
to Platonic abstract objects’ but are contingently abstract, since they could
be realised they could be concrete.36 Anjan Chakravartty has suggested
that the problem of unmanifested states of afairs can be resolved once
it is grasped that the identity of a single instantiated lawmaking power
is exhausted by its position in the network; that one instantiated power
entails the existence of all the active and passive powers that constitute its
identity, together with those powers that constitute their identity, and so
on throughout the entire nomological network of powers and their cor-
responding stimulus and manifestation conditions. Chakravartty insists
that the fact that part of this network is not instantiated need not present
any special concern and does not require a non-naturalistic ontology to be
adopted. On his view, powers are actual even if they are not actualised and
the spokes of the network are to be construed ‘standing by to be realized
under appropriate circumstances.’37
What is to be made of these sorts of attempts to accommodate the
identity conditions of powers? In my view it is difcult to distinguish
them from the kind of extreme Meinongian solutions that positing objects
that ‘subsist’ in an ontological twilight zone between existence and
416 James Orr
non-existence. In the case of Molnar’s solution, the difculties involved
in appealing by analogy to a doctrine as contested as the intentionality of
mental states are forbidding; it seems to be, in fact, a solution more opaque
than the problem it is supposed to solve. The solutions advanced by Bird
and Chakravartty seem indistinguishable from Platonism and therefore
make it hard to avoid the conclusion that the most attractive solution
to the various difculties is to abandon naturalistic or quasi-naturalistic
strictures altogether. A more orthodox alternative to these proposals is
to take up a Platonic ontology, a strategy sometimes deployed to shore
up weaknesses in the necessitarian conception of laws as second-order
relations between immanent universals.38 Indeed one way of interpreting
Molnar’s own account of powers is that they just are abstract objects.39
Still, since these ways of trying to accommodate powers within an
ontology broadly consistent with naturalism are not perspicuous or prom-
ising, it is tempting to turn instead to a fully-fedged Platonic ontology
of powers to render the identity conditions of powers more intelligible.
In the frst instance, Platonising powers makes the otherwise puzzling
identity conditions of powers much more intelligible. Powers are time-
lessly directed towards the universals that feature in all the stimulus and
manifestation conditions involved in their exercise, which can in turn
be construed as being reciprocally directed towards the active powers.
Platonising powers also addresses the puzzle of the ontological status of
powers when they are not manifested. As Matthew Tugby has argued
in a defence of Platonic dispositionalism, it also deals with the intuition
that powers are intrinsic to their bearers, since the identity conditions
of a Platonic power would not be afected by changes in spatiotemporal
conditions extrinsic to it.
For all the theoretical benefts that it brings, however, Platonism saddles
a theory of laws with a number of difculties. Apart from familiar regress
arguments directed at Platonic theories of attributes more generally –
which may not be insuperable but are not easily overcome – the benefts
come at a signifcant cost to metatheoretical parsimony. In particular, by
introducing abstracta Platonism cuts sharply against qualitative ontologi-
cal parsimony; and, by invoking inordinately many fundamental lawmak-
ing entities that are both instantiated concrete entities and uninstantiated
abstract entities, it ofends against qualitative ideological parsimony,
which bids us to minimise kinds of primitive posits that are fundamental
to a theory.
One feature that a successful theory of lawhood needs to accommodate
is the peculiar modal strength of nomological necessity. Realists want
to preserve the intuition that regularities of law-statements are modally
robust: that just is what distinguishes nomic regularities from accidental
generalisations, that is why laws underwrite inductive reasoning, and that
is what grounds the truth of counterfactual statements. But they must
also accommodate the countervailing intuition that the nomic regularities
might have been otherwise. Armstrong simply stipulates that his law-
making second-order relations are neither logically necessary nor purely
contingent but ‘quasi-causal.’ On his view, the puzzle over the weakness
of nomological necessity – midway between strict necessity and brute
contingency – has to be resolved by treating it as a primitive posit, a move
that most of his critics insist is unduly ad hoc and insufciently explana-
tory.40 But a Platonic approach implies a set of metaphysically necessary
laws, since a Platonic network of powers or Platonic kind of universals
would exist necessarily if they existed at all. That would undercut our
intuition that the pattern of nomic regularities in nature could have been
otherwise than they are. Tugby anticipates this objection by suggesting
that the Platonic dispositionalist can claim that while the powers network
exists necessarily, there could have been diferent nomic regularities, since
there could have been a diferent pattern of spatiotemporally instanti-
ated powers than the ones there actually are.41 The reason for this is
that the Platonic dispositionalist must point to a selection mechanism to
account for how lawmaking powers came to be spatiotemporally instanti-
ated in the frst place. But it is hard to see what this could be if not the
sort of transcendent, concrete entity that God is supposed to be. We can
accept, perhaps, that some abstract objects exist as the result of a causal
process. Plausibly, Beethoven’s Third Piano Concerto is now an abstract
structure,42 but that composition is plainly both the causal outcome of
concrete events and a contingent entity (it might, for example, have ended
diferently). But it is much harder to make sense of causal trafc fowing
in the other direction – that is, from abstract to concrete domains – and
harder still to imagine how trafc from the Platonic network to natural
regularities could result in a pattern of natural regularities that could have
been otherwise. If that is correct, Platonic dispositionalism cannot accom-
modate the distinctive modal character of these regularities, namely that
they are naturally necessary but metaphysically contingent.43
5. The Turn to Theism

Taking a step back for a moment, here is one way of framing the problem
of laws in general terms. Identifying powers or kinds as lawmakers along
naturalistic lines ruled out metaphysical ingredients without which they
could not be intelligibly explicated. A special set of concerns arises from
various problems involving manifestation: powers exist independently of
being exercised; powers are intrinsic to their bearers yet capable of being
afected by extrinsic changes; powers point to existing but unmanifested
states of afairs; and those states of afairs point reciprocally to those
powers. More general difculties arose when it came to accommodating
the range of diferent sorts of law-statements, difculties that any realist
theory of lawhood faces if it conforms to naturalistic strictures. Chief
among these are functional laws and uninstantiated laws, but similar
418 James Orr
worries arise in relation to exclusion laws and disjunctive laws.44 None
of the various intermediate solutions that fall short of fully-fedged Pla-
tonism – whether Chakravartty’s actual but unactualised entities or Bird’s
unrealised possibilia – seem perspicuous enough to do the explanatory
work.
Yet yielding to the temptation to adopt Platonism drives a metaphysical
wedge between lawmaking powers or kinds and the concrete regularities
they are supposed to ground. To say that a regularity conforms to a law
is to say both that the law is logically prior to the instance it governs and
that there is a relation of formal identity between token and type.45 If law-
makers do ground nomological necessity, then they provide a universal
standard to which the relevant entities conform their regular behaviour;
yet the lawmakers are also immanent to the instances they govern. The
chief appeal of positing lawmaking powers or kinds is that they satisfy
this second criterion better than realist theories of lawhood that intro-
duce primitive quasi-causal relations that govern categorical properties
extrinsically. But to the extent that it insists on the instantiation principle
and, with it, a partially instantiated fragmented network of powers, the
analysis is much less successful in satisfying the frst criterion, since only
if the powers network is posited in its entirety would it be possible to
account for the universal scope of laws.
What then are the theoretical benefts of theistic approaches to power-
laws and kind-laws? Not all theists take a robustly realist stance towards
powers irreducible to subvening categorical properties or kinds that are
irreducible to aggregates of their members. Richard Swinburne and John
Foster have advanced two of the most developed theistic theories of law-
hood, but they explain lawlike regularities in powers reducible to cat-
egorical properties and take a broadly conventionalist approach to kinds.
Swinburne does not accept that either power-laws or kind-laws generate
a pressure to make ontological room for unmanifested entities,46 nor does
he see any tension between the directed character of causal powers and the
claim that they are intrinsic to their bearers.47 Foster is equally untroubled
by the ontological puzzles of positing powers.48 I have argued elsewhere
that overlooking these difculties weakens the explanatory power of each
of these theistic accounts of lawhood.49 There is a much more prom-
ising theistic strategy for grounding lawmaking powers and kinds, one
that can be found in a contemporary formulation of divine exemplarism,
a tradition that boasts a historical pedigree stretching from Maximus
the Confessor and Augustine to Descartes and Leibniz via Aquinas and
Bonaventure.50
In its classic formulation, divine exemplarism states that divine ideas
are the ways in which God understands his essence as capable of being
imitated in fnite ways.51 As Aquinas distils this conception of nature,
the knowledge of God is the cause of things.52 Contemporary variants of
divine exemplarism have yielded theoretical tools for dealing with modal
truths,53 mathematical objects,54 and propositions.55 More controversially,
it has also been used to formulate a theory of properties.56 Important
diferences separate these approaches, but there has been a signifcant
revival of confdence among theistic philosophers that various theoretical
puzzles in metaphysics can be fruitfully resolved in terms of what God, as
the originating and sustaining source of all reality distinct from himself,
conceives the denizens of reality to be in creating them. The renewal of
these approaches – which, for the sake of simplicity, we can describe as
variants of divine conceptualism – is perhaps unsurprising: it is an axiom-
atic principle of Abrahamic monotheism that all creatures depend on God
both as the source of its existence and for their natures.
God’s mental life supplies all the metaphysical resources necessary to
account for lawmakers, be they powers or kinds. In particular, divine
conceptualism makes it less tempting for nomic realists to turn to univer-
sals as the most promising candidate for lawmaking. What makes it the
case that water has the active power to dissolve salt is that God’s concept
water includes the active power to dissolve salt. Conversely, the reason
that salt is endowed with the passive power to be dissolved in water is that
God’s concept salt includes the passive power to dissolve in water. The
substantial kind salt is what it is in virtue of its members satisfying the
kind-concept guiding God’s initiating the particular causal processes that
led to the emergence of sodium chloride as a chemical kind. Classifying
chemical kinds involves correctly tracking the scheme of chemical reality
that God’s kind-concepts make up. All instances of salt belong to the same
kind because each one depends on a process that originates in the same
constituent of God’s mental life. But they also share a common ground in
virtue of the fact that God sustains them in existence and timelessly intends
them to be instances of salt. If God has kind-concepts, and if lawmaking
regularities supervene on the kinds of substances involved in the relevant
nomic behaviour, then an ultimate truthmaker for law-statements emerges.
The fact that created particular x is a member of kind K is determined by
the fact that God conceives that <x is a K> and wills this to be so, such that
x is a member of K. What grounds the fact that particular x is an electron
is God conceiving that <x is an electron> and God willing that <x is an
electron> be causally satisfed.
An obvious attraction of divine conceptualism is that it does away
with the exemplifcation of universals and the various puzzles associated
with it. Active and passive causal powers are not parsed as properties
that instantiate immanent or Platonic universals but in terms of the more
familiar mechanism of concept-satisfaction. It delivers explanatory gains
comparable to what Platonic solutions deliver relative to naturalistic
options and at a reduced metatheoretical cost. As I have argued elsewhere,
this approach allows unrealised states of afairs that appear to be included
in the identity conditions of powers and kinds while ofering a plausible
way of preserving the intuition that powers are intrinsic to their bearers.57
420 James Orr
It also yields a more unifed metaphysical picture, since the explanatory
appeal to divine concepts integrates the intentionality of psychological
states and the directedness of inorganic physical states that are the causal
outcome of God’s creative intentions.
Still, as a solution it is likely to strike most powers theorists as rebarba-
tively mysterious. If so, we might consider a more familiar secular version
of it. Although he left some of the details unclear, U.T. Place attempted
a conceptualist reduction of universals by characterizing dispositions as
mental abstractions:
Universals on this view are abstracted . . . on the basis of resem-

blances between concrete particulars, their features and the situations
in which they are involved. A relation of resemblance exists between
two or more particulars in so far as they both possess what, when
viewed in the light of the system of universals incorporated in human
language, is the same property or set of properties, though, needless
to say, each possesses a diferent instance of that property.58
But secular conceptualism of this kind seems circular: if universals are

abstracted on the basis of resemblance relations between particulars,
resemblance could not also be grounded in the ‘properties’ incorporated
in a conceptual-linguistic scheme. And other worries loom. First, the
approach cannot account for the possibility that there are nomologically
relevant concepts that human concept-users either have never acquired
or could not in principle acquire. Second, it cannot explain how concepts
could adequately substitute for the unifying function of universals. Secular
concepts could not account for objective similarities between members of
a naturally unifed class: if anything does constitute the natural unity of a
kind, its ontology could not be abstractions in the minds of beings that it
preceded. Third, our conceptual schemes are fallible: conceptual schemes
that included ether or phlogiston failed to capture reality.
But these objections to secular conceptualism do not apply to its the-
istic counterpart. First, the vicious circularity afecting Place’s account
falls away: the resemblance relations between particulars about which
we form concepts track the way those particulars are related because
they fall under the relevant confguration of divine concepts. On divine
conceptualism, moreover, there is no aspect of reality that does not feature
in the intentional content of God’s mental states, including all constituents
of the nomological structure of reality.59 Many suppose universals to be
indispensable to a successful theory of laws and for good reason: since
they underwrite the spatial and temporal scope as well as the objectivity
of lawlike regularities, they are obvious candidates for lawmakers. But
positing universals present well-known difculties independently of elu-
cidating powers and kinds, in particular with respect to exemplifcation:
worries arise over Bradley’s Regress on Platonic realism or in relation
to the immanent realist’s claim that universals are capable of passing
in and out of existence or are multiply instantiated at diferent spatial
and temporal locations. So a theory that does away with them without
undercutting their explanatory benefts may tempt nomic realists to take
up George Bealer’s claim that classical theism provides the only reason for
preferring conceptualist (or nominalist) attribute-theories to universals.60
One immediate worry with divine conceptualism is that it breaches
Russell’s injunction against philosophical solutions that secure ‘the advan-
tages of theft over honest toil.’ Theistic solutions strike many as a stipu-
lative ad hoc response to the metaphysical puzzles of lawmaking. The
degree to which divine conceptualism appears ad hoc will depend heavily
on prior metaphysical assumptions. Some naturalists and Platonists may
view the proposal as a stipulative ad hoc solution introduced only for
patching up difculties that could be tolerated without adopting it. Those
who begin with theistic intuitions are unlikely to be moved by that com-
plaint: on theism, God, divine concepts, and divine powers are of course
part and parcel of a theoretical architecture developed independently of
the explanatory power it might yield for the metaphysical puzzles that
laws present. A version of divine conceptualism is hardly an ad hoc solu-
tion for the theist: if one factors in the doctrine that God creates all reality
distinct from himself, as well as the theoretical parameters arising from
classical commitments to divine sovereignty and aseity, the fundamen-
tal metaphysical explanation for attributes and lawmaking must involve
some appeal to God’s creative intentions.
Another concern with divine conceptualism is that it tends to stress
the role of God’s mind to the exclusion of God’s will. Michael Loux,
for instance, suggests that our justifcation for thinking that properties,
kinds, and relations are exemplifable is that God’s beliefs apply to them.
Predicates about God’s strong beliefs provide the basis for essentialist
analyses of substances and their membership of kinds. The claim that
God strongly believes that <if salt exists, then salt dissolves in water> is
the truthmaker for the claim that salt necessarily behaves in that lawlike
way.61 But that approach does not explain why lawmaking facts involving
powers or kinds obtain in the frst place, since God’s belief that electrons
are negatively charged cannot ground the fact that electrons are negatively
charged, because logically prior to the existence of electrons that belief
would have been false.62 To address this problem, then, divine conceptual-
ists must therefore make room for the role of divine volition. God not only
conceives of every way that electrons could be characterised as electrons
but also wills the causal processes that terminate in the satisfaction of
his concept of an electron by particulars. If negative charge is part of the
essence of an electron, there is only one metaphysically possible way in
which God conceives of electrons once he has created them. God wills
whatever is necessary for his conceiving that <electrons are negatively
charged> be satisfed. Since God is an omnipotent being, the relevant
422 James Orr
divine volition entails whatever causal process is required to bring it about
that electrons are negatively charged.
Any attempt to ground attributes – whether powers or kinds – in God
must overcome a familiar bootstrapping objection: logically prior to God
grounding attributes, God must have already have at least those attributes
without which he could not ground them.63 Some theists might bite the
bullet here, perhaps by accepting that there are uncreated entities that
are ontologically distinct from God.64 Others might attempt to appeal to
the doctrine of divine simplicity by insisting that since God is identical
to his essential attributes he does not create them, but only creates those
attributes with which he is not identical.65 Alternatively, they might soften
the doctrine by treating God and the attributes necessary for creation
as a concrete state of afairs that obtains prior to creation.66 Still more
contentiously, they might advance a bifurcated attribute-theory, accord-
ing to which God possesses the attributes necessary for creating every
other attribute while creating every other attribute.67 Another approach,
which cleaves more closely to the classical divine exemplarism, construes
attributes as ways that God knows himself as containing the archetype
of each lawmaking power or kind.68 A comprehensive assessment of all
of these approaches lies beyond the scope of this chapter, but each is
vulnerable to objections and as yet there is no clear consensus about how
the bootstrapping problem can be resolved.69 It seems to me that the least
unsatisfactory way of anchoring lawmaking attributes in God is to substi-
tute divine concepts for the universals that nomic realists typically invoke
as lawmakers and to deny that there are attributes prior to God creating
them. Attributes are neither ingredients in created objects nor entities that
exist prior to creation but independently of God: they are divine thoughts.
God’s ability to bring lawmaking powers and kinds into existence is what
it is prior to him doing so. It does not depend on God instantiating the
attribute of having the ability to them into existence, since God does not
derive what he is from any prior conception of what he is.
A theistic account of lawmaking is then, by no means free of its own
internal tensions. But it seems reasonable to suggest that these difcul-
ties are not so acute for the theist that they are not outweighed by the
more serious problems afecting naturalistic and Platonic analyses of
power-laws and kind-laws. It does not seem unreasonable to conclude
that there is ample motivation for turning to theistic accounts as the
leading non-naturalistic alternative to Platonism. Some neo-Humeans
will no doubt be tempted to chalk up the pressure on defenders of
power-laws and kind-laws to avail themselves of theistic resources as
confrmation of nomic contingentism by way of a reductio ad theolo-
giam. But as David Lewis once noted, incredulity cannot be used as
a dialectical roadblock without examining the explanatory payof of
the explanation provoking incredulity.70 The costs of theism may seem
high in the currency of incredulity, but that is in part because it is a
currency backed by naturalism, one that rules out a priori powers or
kinds in the robust sense that nomic realism requires. The question is
not how much incredulity is towards theistic solutions to the puzzle
of lawmaking, but whether the theoretical benefts theism brings can
be bought for less. Naturalistic powers and kinds can be bought rela-
tively cheaply, even if many naturalists baulk at the ontological impli-
cations of powers holism, and the explanatory gains are not trivial.
Conversely, while its explanatory payof is considerable, Platonism
charges an exorbitant price in terms of parsimony. Even if we ignore
quantitative ontological parsimony, by positing abstract lawmakers,
Platonic powers and kind-universals ofend against qualitative onto-
logical parsimony by multiplying kinds of ontological entities.71 On
the assumption that its explanatory power is comparable to rivals,
theism ofers a remarkably lean metaphysical framework in terms of
qualitative ideological economy. With a single, minded concrete as its
fundamental explanatory primitive, this account of lawmaking satisfes
what Jonathan Schafer has called the principle of Ockham’s Laser, an
injunction not to multiply fundamental entities beyond necessity.72
Conclusion
The congruence between divine exemplarism and neo-Aristotelian
accounts of natural necessity should not be taken to imply that theism
could not be assimilated to rival theories of lawhood. Theists might pre-
fer necessitation accounts. If lawmaking were analysed as second-order
necessitation relations between typed states of afairs, a theistic transposi-
tion would envisage God selecting which states of afairs are confgured
and impart causal power to the nomic relations. If a neo-Humean best-
system analysis is preferred, then God fxes the tiles of the mosaic and
laws supervene on the divine mosaic. David Lewis defects the charge the
neo-Humean mosaic leads to an implausibly psychologistic conception of
laws. Instead, he appeals to nature’s benevolence: ‘If nature is kind, the
best system will be robustly best . . . We have no guarantee that nature
is kind in this way, but no evidence that it isn’t. It’s a reasonable hope.
Perhaps we presuppose it in our thinking about law.’73 Given Lewis’s start-
ing assumptions, though, it is not obvious why naturalism should expect
this hope to be a reasonable one. Theists, by contrast, have principled and
independent reasons for expecting nature to be kind – that is, for expect-
ing the tiles of the mosaic to display a universe whose regularities can be
codifed by axioms that underpin the true system and that those axioms
could be identifed as laws in the objective sense. Theism explains the
otherwise puzzling ft that best-system theories take for granted between
standards of theoretical elegance that human inquirers contingently apply
and the fact that nature can be described by reference to those standards
in heuristically fruitful ways.
424 James Orr
The trouble with theistic versions of the necessitarian and best-system
accounts to the extent that any theistic theory of lawhood treats lawmak-
ers as is that animating an otherwise causally inert landscape of categori-
cal properties, there may be theological reasons for avoiding the picture of
divine agency that would be difcult to diferentiate from occasionalism.
After all, a theistic account of lawlike regularities that does not treat
causal powers as intrinsic to their bearers is more likely to conceive of
God as sustaining regularities directly.74 For many theologians, such an
account yields a picture of divine agency as an exercise in cosmic puppe-
teering. To the extent that a theistic analysis of natural regularities along
the lines described in this chapter difers dramatically from a ceaselessly
interventionist deity, it may commend itself to theologians who prefer a
theology of nature that is at once metaphysically distinct from God and
refective of his intentions and volitions for it.
It has not been the aim of this chapter to endorse a realist analysis of
lawhood in terms of powers or kinds but rather two claims, one modest,
one more contentious. First, it has suggested to defend that powers and
kinds are not easily assimilated to a naturalistic framework, whether one
that supports an austerely defationary approach to laws as purely contin-
gent regularities, or one that combines a modally robust necessitation rela-
tions with quidditism, or even one that incorporates the more expansively
naturalistic ontology of neo-Aristotelianism. The second claim is that a
theistic alternative to naturalistic and Platonic strategies for a theory of
lawhood based on powers and kinds is well-motivated. While some theistic
approaches are more promising than others, substituting divine concepts
for lawmaking universals – whether construed as ‘power’ universals or
‘kind’ universals – does bring explanatory advantages and non-trivial gains
in metatheoretical parsimony relative to rival alternatives.
Insofar as the perceived advantages of pressing powers or kinds into the
service of nomic realism sufciently outstrip necessitarian theories, it is
tempting to suggest that the pressure on powers theorists to adopt a the-
istic triangulation strategy between naturalistic and Platonic approaches
represents a novel sort of argument for theism from powers and kinds.
But what sort of argument might it be? Classifying it correctly presents an
interesting challenge for natural theology’s standard dialectical taxonomy.
It is proper to describe it in the frst instance as a nomological argument
for God’s existence in that it purports to advance the best explanation
of the modal weakness of natural necessity and to distinguish it from
a law-specifc version of the cosmological or teleological argument. Yet
since it turns on the claim that powers and kinds should be understood
in terms of an inorganic teleology – that is, as directed towards states of
afairs best accounted for in a theistic way – it is also possible to frame it
as a teleological argument. Third, the strategy can be developed to include
features typically associated with cosmological argument, since one way
of construing the distinction between occurrent and dispositional states
is to analyse it as a ‘real’ distinction between act and potency, one that
can only be resolved without setting of an infnite regress by positing a
purely actual being.75 But it might also be characterised – fourth – as a
conceptualist argument on the basis that it takes the best explanation for
natural regularities to involve the ingredients and activities of a divine
mind. Finally, in satisfying a number of metatheoretical criteria better
than rival accounts of lawmaking do, especially in relation to qualitative
and ideological parsimony, one can also construe it as a meta-ontological
argument, at least to the extent one accepts that its explanation of natu-
ral regularities is comparably powerful to those rival accounts.76 There
is every reason, then, for both theistic philosophers and theologians to
welcome the rehabilitation of realist stances towards powers and kinds as
indispensable ingredients in elaborating a coherent and convincing meta-
physical account of nature. More generally, there is also every reason to
suppose that the synthesis of classical theism and neo-Aristotelian ideas
will continue to open up promising avenues for philosophical work in
the future that will illuminate each of these traditions in equal measure.
Notes
1. Ayer 1956: 146.
2. Schafer 2008: 95.
3. Lewis 1986: ix.
4. For a comprehensive recent defence of the powers metaphysic, see Williams
2019.
5. The classic versions of the regularity theory are Mill 1843; Ramsey 1978; and
Lewis 1983; 1986.
6. Mumford 2004 is the most well-known advocate of powers eliminating the
need for laws. Bird 2007, on the other hand, argues that while powers ground
laws, laws of nature do the distinct role of describing relations between events.
They supervene on powers and do not govern them. For the purposes of this
chapter, I prescind from this internal debate by deploying the term ‘lawmaker’
to refer to whatever it is that makes a law-statement true.
7. For the notion of a ‘lawmaker’ we can adapt the defnition David Armstrong
ofers for a truthmaker to describe whatever it is that makes law-statements
true. Thus, ‘a [lawmaker] for a particular [law-statement] is just some exis-
tent, some portion of reality, in virtue of which that [law-statement] is true’
(Armstrong 2004: 5). I adopt this terminological practice to avoid the ques-
tion of whether powers metaphysic is ‘lawless’ (as Mumford 2004: 195–200
argues) or whether it includes laws as well as powers (as Bird 2007: §9.3
claims).
8. Clark 2014.
9. For a lucid historical account, see Lamont 2009.
10. Oderberg 2007: 144.
11. Physics, 2:1.
12. Summa Contra Gentiles, III, §2.1.
13. Summa Contra Gentiles, III, §2.8.
14. For a classic statement, see Wuellner 1956: 62.
15. Ellis 2001; Ellis 2002.
16. Boyd 1999.
426 James Orr
17. Lowe 2006: 39
18. Lowe 2006: 129. It might be possible to preserve Molnar’s analogy to inten-
tionality by noting that Brentano, who was the frst to retrieve the scholastic
notion of esse intentionale, treated it not as a relation between an intentional
mental state and its object but as a non-relational monadic property – that
is, having the property of thinking about the intentional object. For further
discussion, see Taieb 2018: §3.3.
19. Lowe 2006: 130. For similar approaches, see Ellis and Lierse 1994 and Ellis
2001: 6: ‘[C]ausal laws all ultimately depend on the causal powers of things
belonging to natural kinds, and the manners and circumstances of their dis-
plays . . . causal laws can only be found in areas concerned with natural kinds
of objects, properties, or processes.’
20. Lowe 2006: 127: ‘[I]ndividual objects possess their various “natural” powers
in virtue of belonging to substantial kinds which are subject to appropriate
laws – these laws consisting in the possession by such kinds of certain proper-
ties (in the sense of universals), or in the standing of kinds in certain relations
to one another.’ Inman 2018: 38–39 describes this as the ‘Kind-Power Con-
nection’: a substance necessarily has the causal powers that it does in virtue
of being an instance of a kind, though given that they can be impeded by (say)
masks, the powers dispose rather than necessitate their bearer towards their
manifestation.
21. Lowe 2006: 135.
22. Martin 2008: 3. Diferent labels have been used for the ontology that these
identity conditions imply – e.g. Martin 2008: 22 (‘a web or net of readiness’);
Mumford 2004: 182 (‘an interconnected web’); Armstrong 1997: 255 (‘a
power net’).
23. For this objection against powers, see Psillos 2006.
24. Molnar 2003: 102; see also Williams 2010: 85–86.
25. Orr 2019; for critiques in a similar vein, see also Tugby 2013 and Dumsday
2014.
26. Dumsday 2016: 92–93.
27. I am grateful to Robert C. Koons for raising this objection and proposing this
solution.
28. For similar objections to Armstrong’s theory of laws as necessitation relations
construed as immanent universals, see Fales 1993: 134.
29. Armstrong 1983: 111.
30. Mellor 1995: 201.
31. Armstrong 1983: 117.
32. Mellor 1995: 201.
33. Molnar 2003: 62–81.
34. Molnar 2003: 71.
35. Gof 2017: 137–140.
36. Bird 2007: 113.
37. Chakravartty 2007: 147.
38. Tooley 1977.
39. Ingthorsson 2012: 539.
40. For example Lewis 1983: 366 and van Fraaseen 1987.
41. Tugby 2013: 26–27.
42. Dodd 2002.
43. For further discussion of the modal issues and a defence of natural necessity
as relative necessity, see Fine 2005: 238–248.
44. For discussion of the problems arising from exclusion laws and disjunctive
laws, see Orr 2019: §2.4.4; Mumford 2004: 91; and Tooley 1977: 677.
45. Peterson 1996: 89.
46. Swinburne 1995: 389: ‘The power to exert such and such a force is a property
possessed by an individual whether or not it ever exerts it or whether there
are any other actual individuals on which it can exert it.’
47. Swinburne 1994: 33: ‘The power to lift a 100-pound weight, although it in
some sense concerns a relation which I might have to some other thing, is on
my defnition a monadic property, because it does not concern a relation to
any actual thing – my having this causal power is to be distinguished from
my exercising it.’
48. Foster 2004: 162–163: ‘[T]he presence of a disposition does not depend on
there being occasions when the disposition is exercised; nor, when there are
such occasions, is the presence of the disposition exhausted by the totality
of events that occur on them. Thus if something is fragile, it is so even if it is
never subjected to the kind of impact that would make its fragility manifest.’
49. Orr 2019: 123–157.
50. For convenient overviews of the historical background in late antique and
scholastic philosophy, see Boland 1996: 17–192. A lucid discussion of the
fate of divine intellectualism into early-modern philosophy can be found in
Craig 1987: ch. 1 and Hight 2008.
51. For a classic statement, see Aquinas, Summa Theologiae, Ia, q. 15, a. 2,
resp.; and for detailed discussion of the (shifting) approaches Aquinas adopts
towards divine ideas, see Boland 1996: 200–225 and Doolan 2008: ch. 1.
52. Aquinas, Quaestiones Disputatae de Veritate, q. 6, a. 6 (‘scientia dei est causa
rerum’).
53. Plantinga 1982.
54. Menzel 1987.
55. Adams 1994: ch. 7; Welty 2014.
56. Morris and Menzel 1986.
57. Orr 2019: 423–424.
58. Place 1996: 26.
59. As Leftow 2006: 339 puts it, ‘there is nothing a thing could be of which God
would not have a concept.’
60. Bealer 1993: 27–28:
My arguments [against nominalism and conceptualism] would fail if, for

every logically possible circumstance, everything object there – in par-
ticular, every object that is not among the objects that actually exist – is
nevertheless such that God right now has an actual idea that would apply
uniquely and rigidly to that object. Thus, a necessary condition for the
truth of nominalism or conceptualism is not only that God exist but also
that God have ideas of a kind that in principle we could not have and,
indeed, that we could not know what it is like to have.
61. Loux 1986: 510.

62. Leftow 2012: 432–434.
63. For a sophisticated formulation of this objection, see Bergmann and Brower
2006: 366–372.
64. van Inwagen 2009.
65. Davis 2001: ch. 4; Leftow 1990; Bergmann and Brower 2006: 383–385.
66. McCann 2012: 127–129.
67. Gould and Davis 2014.
68. Ward 2020: §6.
69. Craig 2017: 139–158 ofers a comprehensive and insightful discussion.
428 James Orr
70. Lewis 1986: 135.
71. Nolan 1997 argues with some force that, from a historical point of view,
quantitative economy has played an important role in the selection of some
scientifc theories, but there seems to be a broad consensus in contemporary
metaphysics that it should count less than qualitative economy.
72. Schafer 2015: 647–651.
73. Lewis 1994: 479.
74. An account of divine causation in terms of concurrence or conservation is not
strictly incompatible with a defationary stance towards powers and kinds;
but given such a stance, occasionalism is certainly a more tempting option and
one that seems harder to reconcile to a powers-based account (for a dissenting
view on this point, see Schultz and D’Andrea-Winslow 2017).
75. For this approach, see Oderberg 2007: 125:
[B]y grasping the real distinction [between essence and existence] we

make room for the very idea of contingency, for in contingent beings
no essence must be actualised. But at the same time we make room for
the idea of a being whose essence is its existence, in the sense that there
is no mixture of potency in it. Theists identify this being as God, and
it is traditionally held that God is pure actuality, i.e. a being that has
no potentiality in its constitution, this absence being the root cause of
its unlimited and infnite nature (emphasis in original).For a critical
analysis of the internal and external features of powers in terms of
potency and act that draws on Aristotle, Aquinas, and Suárez, see also
Psillos 2021.
76. Cf. Leftow 2011;Orr 2019: 165–167.
References
Adams, Robert, Leibniz: Determinist, Theist, Idealist (Oxford: Oxford University
Press, 1994).
Armstrong, David, What is a Law of Nature? (Cambridge: Cambridge University
Press, 1983).
———, A World of States of Afairs (Cambridge: Cambridge University Press,
1997).
———, Truth and Truthmakers (Cambridge: Cambridge University Press, 2004).
Ayer, A.J., ‘What is a Law of Nature?’ Revue Internationale de Philosophie 36
(1956), 144–165.
Bealer, George, ‘Universals,’ Journal of Philosophy 40 (1993), 5–32.
Bergmann, Michael, and Brower, Jefrey E., ‘A Theistic Argument against Pla-
tonism (and in Support of Truthmakers and Divine Simplicity),’ in D. Zimmer-
man (ed.), Oxford Studies in Metaphysics, Volume 2 (Oxford: Oxford
University Press, 2006), 357–386.
Bird, Alexander, Nature’s Metaphysics: Laws and Properties (Oxford: Oxford
Boland, Vivian, Ideas in God According to Saint Thomas Aquinas: Sources and
Synthesis (Leiden: E.J. Brill, 1996).
Boyd, Richard, ‘Homeostasis, Species, and Higher Taxa,’ in R. A. Wilson (ed.), Spe-
cies: New Interdisciplinary Essays (Cambridge, MA: MIT Press, 1999), 141–185.
Chakravartty, Anjan, A Metaphysics for Scientifc Realism: Knowing the Unob-
servable (Cambridge: Cambridge University Press, 2007).
Clark, Errin D., ‘How Aristotelian is Contemporary Dispositionalist Metaphys-
ics? A Tale of Two Distinctions,’ Proceedings of the American Catholic Philo-
sophical Association 88 (2014), 85–99.
Craig, Edward, The Mind of God and the Works of Nature (Oxford: Clarendon
Press, 1987).
Craig, William Lane, God and Abstract Objects – The Coherence of Theism:
Aseity (Dordrecht: Springer, 2017).
Davis, Richard Brian, The Metaphysics of Theism and Modality (New York: Peter
Lang, 2001).
Dodd, Julian, ‘Defending Musical Platonism,’ British Journal of Aesthetics 42
(2002), 380–402.
Doolan, Gregory T., Aquinas on the Divine Ideas as Exemplar Causes (Washing-
ton, DC: Catholic University of America Press, 2008).
Dumsday, Travis, ‘Lowe’s Unorthodox Dispositionalism,’ Res Philosophica 93(1)
(2016), 79–101.
———, ‘A Cosmological Argument from Moderate Realism,’ Heythrop Journal
61 (2014), 732–736.
Ellis, Brian, Scientifc Essentialism (Cambridge: Cambridge University Press, 2001).
———, The Philosophy of Nature: A Guide to the New Essentialism (Chesham:
Acumen, 2002).
Ellis, Brian, and Lierse, Caroline, ‘Dispositional Essentialism,’ Australasian Jour-
nal of Philosophy, 72:1 (1994), 27-45.
Fales, Evan, ‘Are Causal Laws Contingent?,’ in J. Bacon, K. Campbell, and L.
Reinhardt (eds.), Ontology, Causality, and Mind: Essays in Honour of D.M.
Armstrong (Cambridge: Cambridge University Press, 1993), 121–144.
Fine, Kit, ‘The Varieties of Necessity,’ in Modality and Tense: Philosophical
Papers (Oxford: Oxford University Press, 2005), 235–260.
Foster, John, The Divine Lawmaker: Lectures on Induction, Laws of Nature, and
the Existence of God (Oxford: Clarendon Press, 2004).
Gof, Philip, Consciousness and Fundamental Reality (Oxford: Oxford University
Press, 2017).
Gould, Philip, and Davis, Richard Brian, ‘Modifed Theistic Activism,’ in P. Gould
(ed.), Beyond the Control of God? Six Views on the Problem of God and
Abstract Objects (London: Bloomsbury, 2014), 51–79.
Hight, Marc A., Idea and Ontology: An Essay in Early Modern Metaphysics of
Ideas (University Park, PA: Pennsylvania State University Press, 2008).
Ingthorsson, R.D., ‘The Regress of Pure Powers Revisited,’ European Journal of
Philosophy 23(3) (2012), 529–541.
Inman, Ross D., Substance and the Fundamentality of the Familiar: A Neo-Aris-
totelian Mereology (New York and Abingdon: Routledge, 2018).
Lamont, John, ‘The Fall and Rise of Aristotelian Metaphysics in the Philosophy
of Science,’ Science and Education 18 (2009), 861–884.
Leftow, Brian, ‘Is God an Abstract Object?,’ Noûs 24 (1990), 581–598.
———, ‘God and the Problem of Universals,’ in D. Zimmerman (ed.), Oxford
Studies in Metaphysics, Volume 2 (Oxford: Oxford University Press, 2006),
325–356.
430 James Orr
———, ‘One Step Towards God,’ Royal Institute of Philosophy Supplement 68
(2011), 67–103.
———, God and Necessity (Oxford: Oxford University Press, 2012).
Lewis, David, ‘New Work for a Theory of Universals,’ Australasian Journal of
Philosophy 61 (1983), 343–377.
———, On the Plurality of Worlds (Oxford: Blackwell, 1986).
———, ‘Humean Supervenience Debugged,’ Mind 103 (1994), 473–490.
Loux, Michael, ‘Towards an Aristotelian Theory of Abstract Objects,’ in
P. French, T. Uehling, and H. Wettstein (eds.), Midwest Studies in Philosophy
11 (1986), 495–512.
Lowe, E.J., A Four-Category Ontology: A Metaphysical Foundation for Natural
Science (Oxford: Oxford University Press, 2006).
Martin, C.B., The Mind in Nature (Oxford: Oxford University Press, 2008).
McCann, Hugh J., Creation and the Sovereignty of God (Bloomington, IN: Indi-
ana University Press, 2012).
Mellor, D.H., The Facts of Causation (Oxford: Oxford University Press, 1995).
Menzel, Christopher, ‘Theism, Platonism, and the Metaphysics of Mathematics,’
Faith and Philosophy 4(4) (1987), 365–382.
Mill, John Stuart, System of Logic (London: J.W. Parker, 1843).
Molnar, George, Powers: A Study in Metaphysics (Oxford: Oxford University
Press, 2003).
Morris, Thomas V. and Menzel, Christopher, ‘Absolute Creation,’ American
Philosophical Quarterly 23(4) (1986), 353–362.
Mumford, Stephen, Laws in Nature (London and New York: Routledge, 2004).
Nolan, Daniel, ‘Quantitative Parsimony,’ British Journal for the Philosophy of
Science 48 (1997), 329–343.
Oderberg, David, Real Essentialism (Abingdon and New York: Routledge, 2007).
Orr, James, The Mind of God and the Works of Nature: Laws and Powers in
Naturalism, Platonism, and Classical Theism (Leuven: Peeters, 2019).
———, ‘No God, No Powers: Classical Theism and Pandispositionalist Laws,’
International Philosophical Quarterly 59(4) (2019), 411–426.
Peterson, John, ‘Law and Thomistic Exemplarism,’ The Thomist 60(1) (1996),
81–108.
Place, U.T., ‘Dispositions as Intentional States,’ in T. Crane (ed.), Dispositions: A
Debate (London and New York: Routledge, 1996), 19–32.
Plantinga, Alvin, ‘How to be an Anti-Realist,’ Proceedings and Addresses of the
American Philosophical Association 56 (1982), 47–70.
Psillos, Stathis, ‘What Do Powers Do When They Are Not Manifested?,’ Philoso-
phy and Phenomenological Research 72 (2006), 137–156.
———, ‘The Inherence and Directedness of Powers,’ in H. Lagerlund, B. Hill, and
S. Psillos (eds.), Reconsidering Causal Powers: Historical and Conceptual Per-
spectives (Oxford: Oxford University Press, 2021), 45–67.
Ramsey, Frank, ‘Law and Causality: Universals of Law and of Fact,’ in D. Mellor
(ed.), Foundations: Essays in Philosophy, Logic, Mathematics and Economics
(London: Routledge, 1978), 128–132.
Schafer, Jonathan, ‘Causation and Laws of Nature: Reductionism,’ in T. Sider, J.
Hawthorne, and D. Zimmerman (eds.), Contemporary Debates in Metaphysics
(Oxford: Blackwell, 2008), 82–107.
———, ‘What Not to Multiply Without Necessity,’ Australasian Journal of Phi-
losophy 93(4) (2015), 644–664.
Schultz, Walter J., and D’Andrea-Winslow, Lisanne, ‘Causation, Dispositions,
Physical Occasionalism,’ Zygon 52(4) (2017), 962–983.
Swinburne, Richard, The Christian God (Oxford: Oxford University Press, 1994).
———, ‘Thisness,’ Australasian Journal of Philosophy 73 (1995), 389–400.
Taieb, Hamid, Relational Intentionality: Brentano and the Aristotelian Tradition
(Dordrecht: Springer, 2018).
Tooley, Michael, ‘The Nature of Laws,’ Canadian Journal of Philosophy 7 (1977),
667–698.
Tugby, Matthew, ‘Platonic Dispositionalism,’ Mind 22 (2013), 451–480.
van Fraaseen, Bas C., ‘Armstrong on Laws and Probabilities,’ Australasian Jour-
nal of Philosophy 65 (1987), 243–259.
van Inwagen, Peter, ‘God and Other Uncreated Things,’ K. Timpe and E. Stump
(eds.), Metaphysics and God: Essays in Honor of Eleonore Stump (London:
Routledge, 2009), 3–20.
Ward, Thomas, Divine Ideas (Cambridge: Cambridge University Press, 2020).
Welty, Greg, ‘Theistic Conceptual Realism,’ in P. Gould (ed.), Beyond the Control
of God? Six Views on the Problem of God and Abstract Objects (London:
Bloomsbury, 2014), 81–96.
Williams, Neil E., ‘Puzzling Powers: The Problem of Fit,’ in A. Marmodoro (ed.),
The Metaphysics of Powers: Their Grounding and Manifestation (Abingdon
and New York, NY: Routledge, 2010), 106–132.
———, The Powers Metaphysic (Oxford: Oxford University Press, 2019).
Wuellner, S.J.B., Dictionary of Scholastic Philosophy (Milwaukee, WI: Bruce Pub-
lishing Company, 1956).
Epilogue
Anna Marmodoro
This book raises and aims to address two questions: How does the current
resurgence of ‘neo-Aristotelian’ metaphysics impact on our philosophical
understanding of nature? And how does this in turn impact on theology,
insofar as theology is concerned with questions regarding God’s nature,
God’s relation to nature, and the place of humanity within nature? My
comments here concern how the important discussion that the editors
have initiated might be taken forward in the future.
The overall argument which runs through the book’s individual chapters
is that ‘neo-Aristotelian’ metaphysics gives us an understanding of nature
as contemporary science reveals it to us (which is implicitly assumed to
be better than what alternative metaphysical systems have to ofer), and
thus, it needs to be taken into account by a scientifcally engaged theology
of nature. Put diferently, contemporary science reveals to us a natural
world which neo-Aristotelian metaphysics equips us to understand. Thus,
insofar as the theology of nature is supposed to be engaged with nature as
much as with the divine, the resulting conception of nature, revealed by
contemporary science and conceptualized via ‘neo-Aristotelianism’, will
be a game-changer for theological discussions too (but in which specifc
ways, the authors do not attempt to say in this book).
This is one thought-provoking conclusion that emerges from the book,
but it is not the only one. Another argumentative strand to be found in the
frst part of the book, even if not the overall argument as such, goes in the
direction of ‘rehabilitating’ Aristotle’s metaphysics from being unpalat-
able for the scientifcally minded thinker of today to being an essential
tool to develop a scientifcally based conception of nature. This conclu-
sion, too, will spark discussion among the book’s readers.
The crux then is to understand what is distinctive of ‘neo-Aristotelian’
metaphysics, in relation to other alternative metaphysics; for instance,
that of David Lewis, which many are developing at present, sometimes
under the name of ‘neo-Humeanism, in direct opposition to ‘neo-
Aristotelianism’. What is it that ‘neo-Aristotelian’ metaphysics brings to
the table in terms of conceptual tools for our understanding of nature that
alternative approaches like neo-Humeanism do not?
Epilogue 433
Addressing this question in full would be a very difcult task to man-
age within the limits of one single book; so here the authors do well to
focus on some philosophical ideas of Aristotle’s, which they take to be
key to ‘neo-Aristotelian’ metaphysics. Among these ideas is Aristotle’s
hylomorphism, his conception of the substantial form, the priority he
gives to substances as primary beings in his system, his theory of powers
and causation, and more. The book’s overall strategy is to show how
explanatorily successful it is to draw on such ideas to account for a selec-
tion of debated issues today, in the domains of the philosophy of nature
(Part 1), the philosophy of mind (Part 2), and philosophical theology
(Part 3). Any selection of issues would have been bound to make only a
partial case for the large-scale claims that the book, as a whole, aims to
argue for. What is missing in the book, however, is an explicit justifcation
of why the cases selected are supposed to be particularly salient in relation
to the book’s overarching argument. One consideration may be that the
editors have focused on issues where the Abrahamic theological traditions
have come under pressure from science, as understood through the lens of
(non-Aristotelian) philosophy – such as free will, divine action, creation
and evolution, the immortality of the soul, and so on – to motivate us
to reconsider those issues in the light of a neo-Aristotelian conception of
nature. A forward-looking consideration to make here is that the readers
persuaded by the cases made in this book might want to extend the range
of case studies, and further, to strengthen the value of the cases made
by developing broader and more explicit criteria by which the cases are
selected.
Some of the chapters (those in Part 1) draw on Aristotle’s metaphys-
ics to ofer a philosophical understanding of current scientifc theories
of nature. Within this group, William M. R. Simpson shows that the
Aristotelian doctrine of hylomorphism allows for a cogent interpretation
of the most fundamental stratum of reality science has discovered, quan-
tum mechanics, both as it has been understood in the de Broglie-Bohm
interpretation, and more recently, as it is being understood in an inter-
pretation put forward by the cosmologist George Ellis and the physicist
Barbara Drossel. Robert C.Koons argues that we can apply a version of
Aristotle’s holism (here understood as dependence of a substance’s powers
on its substantial form) to understand the holistic character of quantum
chemistry and thermodynamics. Stephen Boulter discusses the evolution-
ary transitions that led to the emergence of life on this planet, as involv-
ing entities with new powers and operations arising from a causal base
lacking those powers and operations. He argues that such transitions are
underpinned by a challenging ‘principle of the conservation of perfec-
tion’. Each transition in the history of life involves both an increase and a
decrease in operational powers (of Aristotelian descent), allowing for an
overall conservation of perfection in an evolving universe. David S. Oder-
berg ofers arguments for making room in current thought for the idea of
434 Anna Marmodoro
a hierarchy of being, according to which all existing things are ordered
according to a classifcation of the powers they have – an idea he fnds
rooted in Aristotle and more generally in ancient thought, but which had
lost appeal in the modern world. As mentioned earlier, one might raise
the question of whether the selection of topics addressed in this section of
the book ‘hang together’ in the sense of making collectively the case that
‘neo-Aristotelian’ metaphysics gives us a better understanding tout court
of nature than we would have if we did not lean on it, although it is fair
to recognise they do serve well the narrower goal of showing how certain
classical tenets of the Abrahamic-Aristotelian tradition can be coherently
afrmed in the light of contemporary science. This remark is not how-
ever a critique of the book’s strategy but rather a pointer toward future
research to be pursued by those who endorse the book’s general thesis
and wish to expand upon it. Physics and biology are obvious directions
to suggest for further exploration, but it might be both adventurous and
possibly rewarding to consider cosmology too, for instance: might ‘neo-
Aristotelian’ metaphysics enable us also to understand better the universe,
its origins, and its evolution than how we know (or hypothesise) them to
be on the basis of current science?
Another group of chapters in the book (those in Part 2) concerns human
beings in nature, with focus on our mental life and agency. Timothy
O’Connor uses a broadly Aristotelian theory of causal powers (in con-
junction with a constituent ontology of substances, and a theory of time
according to which the future is not real) to defend (indeterministic) free
will. Janice Tzuling Chik argues that the paradigm for conceptualising
human action should refer essentially to animal powers, conceived in
an Aristotelian manner as embodied in animate corporeality. Daniel D.
De Haan focuses on a particular set of human powers, those to perform
scientifc experiments. This chapter argues that only a neo-Aristotelian
conception of human being, premised on embodied intentional action,
can explain how we can perform experiments; and it shows this by con-
trasting the view with ‘standard Humean’ philosophical conceptions.
Along broadly similar lines, Antonio Ramos-Díaz argues that we engage
in mathematical and logical activity, such that it can be explained only
by the relevant intellectual powers we have, and we have features such
that are not explicable solely in terms of material properties and powers.
Taken together this group of chapters defend the view that we would not
be who we think we are, in terms of our willing, thinking, and generally
acting rationally, unless we conceived ourselves in an Aristotelian way.
Forward-looking, to ‘round-of’ this line of argument, one may want to
discuss what’s essential to being a human being, and whether for all that
is essential, ‘neo-Aristotelianism’ provides the apt conceptual framework
for understanding it. In this connection, it may once again be adventurous
and potentially rewarding to consider the relation that humans have to
technology, and whether that relation (taken to the extreme, as it were, of
Epilogue 435
the creation and increasing development of artifcial intelligence) may or
may not be explained within a ‘neo-Aristotelian’ framework. Or one may
think that the chapters here concerned with showing that our abilities to
act and think rationally, grounded on our human corporeality, are a line
of argument against functionalism, and therefore indirectly against the
possibility that intelligence may be ‘artifcial’.
The remaining two chapters in this part of the collection ofer non-
mainstream approaches to the long-standing mind-body problem. Chris-
topher Hauser argues that ‘neo-Aristotelians’ who defend a hylomorphic
view according to which human beings are composed of matter (i.e. their
body) and a certain kind of substantial form (i.e. a human soul) have to
embrace one of two options: either it is the case that neither a human
person nor her soul can exist without matter; or it is the case that both a
human person and her soul can exist without matter. Travis Dumsday dis-
cusses the place of angels, understood as bodiless/purely spiritual beings,
which nonetheless may exercise causal powers within nature, within a
neo-Aristotelian philosophy of nature. The counter-intuitive consequences
following from Hauser’s and Dumsday’s arguments may be taken as a
reductio ad absurdum of ‘neo-Aristotelianism’, or else one may think that
their counter-intuitiveness ofers ways out from the stalemate of the mind-
body problem; the readers will want to make up their minds about this.
The chapters that follow (in Part 3) are specifcally about God: God’s
nature, and action upon the created world. Anne Siebels Peterson puts
Aristotle’s account of the unity of substance qua hylomorphic composite
to use to explain divine simplicity. Ross D. Inman and James Orr on the
other hand tackle another aspect of God’s nature, namely God’s being
the metaphysical foundation of all creaturely being; Inman arguing for
an account of creaturely participation in God that explains a non-inter-
ventionalist, causal powers account of divine action in the world; and
Orr, more broadly, arguing that all the power in the world depends on
the existence of God. Edward Feser investigates the relationship between
the natural and supernatural (order and powers), and their metaphysical
interdependencies. Alexander R. Pruss discusses ways of reconciling God’s
design with statistical evolutionary explanations of nature and argues
that while none of the existing accounts is adequate, there is a theistic
version of David Lewis’s ‘best ft’ account of probability that can allow for
a reconciliation between randomness and design, avoiding explanatory
difculties. Simon Maria Kopf argues that what a classical account of
providence requires is not so much an overall teleology of nature, but tele-
ological natures in the Aristotelian sense. The conclusion this last group
of chapters taken together suggest enriches with one additional dimension
the overall picture canvassed by the book: we see ‘neo-Aristotelianism,’
not only here, in service as it were of our understanding of nature – human
beings included – but also of our understanding of the divine. So if the
book’s aim was to show that the scientifcally engaged theologian ought to
436 Anna Marmodoro
think about the impact that the resurgence of neo-Aristotelian metaphys-
ics is having on our understanding of the natural world, the position that
emerges at the end of the book is actually stronger: the theologian would
do well to think along the lines of neo-Aristotelian metaphysics, not only
for achieving a better understanding of nature, but of God too. The latter
stance may be what will spark most discussion among its readers.
About the Contributors
Stephen Boulter is Reader in Philosophy at Oxford Brookes University.

He is author of The Rediscovery of Common Sense Philosophy (2007),
Metaphysics From a Biological Point of View (2013), and Why Medi-
eval Philosophy Matters (2019).
Janice Tzuling Chik is Associate Professor of Philosophy at Ave Maria
University and Associate Member of the Aquinas Institute at Blackfri-
ars Hall, Oxford. She publishes on issues relating to the metaphysics
of action and mind. She was previously a Visiting Research Scholar at
University of Oxford (2017 and 2019), and was the 2019–2020 John
and Daria Barry Foundation Fellow and Lecturer in Philosophy at the
University of Pennsylvania. She holds degrees in philosophy, public
policy, and music performance from Princeton University, the Univer-
sity of Texas at Austin, and the University of St Andrews in the UK.
Daniel D. De Haan is a Research Fellow in the Ian Ramsey Centre for
Science and Religion at the Faculty of Theology and Religion, Univer-
sity of Oxford. Before coming to Oxford, De Haan was a postdoc-
toral fellow in the Faculty of Divinity at the University of Cambridge
working on the neuroscience strand of the Templeton World Charity
Foundation Fellowship in Theology, Philosophy of Religion, and the
Sciences Project, directed by Sarah Coakley. He received his doctorate
in philosophy from Catholic University of Leuven and the University of
St Thomas. He is the author of Necessary Existence and the Doctrine
of Being of Avicenna’s ‘Metaphysics of the Healing’ (Brill, 2020).
Travis Dumsday is an Associate Professor of philosophy at Concordia
University of Edmonton. He received his PhD in philosophy from the
University of Calgary in 2010. After a postdoctoral fellowship at the
University of North Carolina Chapel Hill, he was Assistant Profes-
sor of Religious Studies at Livingstone College. He then returned to
Canada to join the philosophy department at Concordia University of
Edmonton. There he also held the Canada Research Chair in Theology
and the Philosophy of Science from 2015 to 2020. He has published
438 About the Contributors
widely in analytic philosophy of science, philosophy of religion, applied
ethics, and Eastern Orthodox theology. He would very much like to
own an octopus one day.
Edward Feser is Professor of Philosophy at Pasadena City College in
Pasadena, California. He is the author of many academic articles and
books, including Scholastic Metaphysics: A Contemporary Introduc-
tion, Five Proofs of the Existence of God, and Aristotle’s Revenge: The
Metaphysical Foundations of Physical and Biological Science.
Christopher Hauser is an Assistant Professor in the Philosophy Depart-
ment at the University of Scranton, where he teaches and conducts
research on topics in metaphysics, philosophy of religion, ancient
philosophy, and medieval philosophy. Prior to joining the philosophy
faculty at the University of Scranton, Christopher earned a PhD in phi-
losophy from Rutgers University and a BA in philosophy and history
from Dartmouth College. He has published scholarly journal articles
in Metaphysics and Faith and Philosophy.
Ross D. Inman is Associate Professor of Philosophy at Southeastern Bap-
tist Theological Seminary, USA. He is a former research fellow at the
University of Notre Dame, Center for Philosophy of Religion and Saint
Louis University. He was awarded the 2014 Marc Sanders Prize in
Philosophy of Religion. His research has appeared in Philosophical
Studies, American Philosophical Quarterly, Oxford Studies in Philoso-
phy of Religion, Metaphysica, and Philosophia Christi.
Robert C. Koons is a Professor of Philosophy at the University of Texas at
Austin, where he has taught for 33 years. He is the author or co-author
of four books, including Realism Regained (Oxford University Press,
2000) and The Atlas of Reality: A Comprehensive Guide to Metaphys-
ics, with Timothy H. Pickavance (Wiley-Blackwell, 2017). He is the
co-editor (with George Bealer) of The Waning of Materialism (Oxford
University Press, 2010) and co-editor (with William M. R. Simpson
and Nicholas Teh) of Neo-Aristotelian Perspectives on Contemporary
Science (Routledge, 2018). Koons has authored over 50 peer-reviewed
articles in metaphysics, philosophical logic, theology, and philosophy
of science.
Simon Maria Kopf is Principal Investigator on a Templeton-funded
project titled ‘A Virtue-Based Approach to Providence: Bridging the
Analytic/Continental Divide in the Central-European Divine Action
Debate’, based at Humboldt University of Berlin (2021–22), a visiting
research fellow at King’s College London (2020–23), and a visiting lec-
turer in systematic theology at the International Theological Institute,
Austria. Previously, he was a research fellow on the Templeton project
‘Divine and Human Providence’ at Universidad Austral, Argentina
About the Contributors 439
(2018–2020), and a research assistant on the ERC project ‘Authority
and Innovation in Early Franciscan Thought (c. 1220–45)’ at King’s
College London (2017–2019).
John Marenbon is a fellow of the British Academy, senior research fel-
low of Trinity College, and honorary professor of medieval philoso-
phy, as well as a visiting professor at the Philosophy Department of
Peking University. His interests cover the whole breadth of philosophy
in the Long Middle Ages (c. 200–c. 1700), in the Latin and Greek
Christian, Islamic, and Jewish traditions. He has written both general
books (especially Medieval Philosophy: An Historical and Philosophi-
cal Introduction (2007) and (as editor) the Oxford Handbook of Medi-
eval Philosophy (2012)), as well as more specialized studies of Boethius
and Abelard. His most recent book is Pagans and Philosophers. The
Problem of Paganism From Augustine to Leibniz (2015).
Timothy O’Connor is the Mahlon Powell Professor of Philosophy and a
core member of the Cognitive Sciences Program at Indiana University.
He has taught there since 1993 (serving as chairman 2006–12), apart
from visiting research fellowships at the Universities of Notre Dame,
St. Andrews, and Oxford and an appointment as Distinguished Profes-
sor at Baylor University. His main areas of scholarship are metaphys-
ics, philosophy of mind/cognitive science, and philosophy of religion.
He has published over 90 scholarly articles and reviews and two mono-
graphs, Persons and Causes: The Metaphysics of Free Will and Theism
and Ultimate Explanation. He has also edited and contributed to six
other books on the philosophy of mind and action; ‘emergent’ or ‘top-
down’ forms of explanation in the sciences; and the epistemology of
religious belief.
David S. Oderberg is Professor of Philosophy at the University of Reading,
England. He is the author of several books as well as many articles in
metaphysics, ethics, philosophy of religion, and other subjects. His
most recent book is The Metaphysics of Good and Evil (Routledge,
2020). He is also the editor of several collections on ethics, logic, and
metaphysics. David edits Ratio, an international journal of analytic
philosophy.
Anna Marmodoro is a Professor of Metaphysics at the University of Dur-
ham, and an associate member of the Philosophy Faculty at the Uni-
versity of Oxford. She has published monographs, edited books, and
journal articles in metaphysics, ancient, late antiquity, and medieval
philosophy, the philosophy of mind, and the philosophy of religion. She
is the co-founder and co-editor of the peer-reviewed journal Dialogoi:
Ancient Philosophy Today, published by Edinburgh University Press,
whose mission is to promote mutual engagement between ancient and
current philosophy.
440 About the Contributors
James Orr is University Lecturer in Philosophy of Religion at Cambridge.
He was formerly the McDonald Postdoctoral Fellow at Christ Church,
Oxford. James holds a PhD and MPhil in philosophy of religion from
St. John’s College, Cambridge, and a Double First in Literae Humanio-
res (Classics) from Balliol College, Oxford.
Anne Siebels Peterson is Associate Professor of Philosophy at the Uni-
versity of Utah and received her PhD from the University of Notre
Dame. She researches Aristotle’s metaphysics and natural philosophy,
specifcally the way in which Aristotle’s hylomorphic understanding of
living things is supported and infuenced by his views on the nature of
being and unity in his Metaphysics. She also explores how Aristotle’s
methods in philosophy of science and philosophy of biology complicate
concepts like those of body, soul, change, and life, perhaps in ways that
are still relevant for us as we continue to explore similar concepts in
contemporary metaphysics and philosophy of science, especially phi-
losophy of biology.
Alexander R. Pruss is Professor of Philosophy at Baylor University. His
research focuses on metaphysics, philosophy of religion, applied eth-
ics, formal epistemology, and the philosophy of mathematics. He is
the author of several monographs, including Infnity, Causation, and
Paradox (Oxford University Press, 2018) and The Principle of Suf-
fcient Reason: A Reassessment (Cambridge University Press, 2006).
Antonio Ramos-Díaz is an Assistant Professor of Philosophy at the Pon-
tifcia Universidad Católica de Puerto Rico (Ponce Campus). His areas
of specialization are logic, philosophy of mind, metaphysics, and philo-
sophical anthropology, with competence in the philosophy of mathe-
matics, the philosophy of computation, and the philosophy of cognitive
science. He received his BA in history, with a minor in philosophy,
from the University of Puerto Rico (Río Piedras Campus). He holds an
M.Litt. degree in philosophy from the University of St. Andrews and
an M.Phil. degree in philosophy from the University of Leuven (KU
Leuven), where he also obtained his PhD in philosophy.
William M. R. Simpson is a Junior Research Fellow of Wolfson College,
Cambridge, and an Honorary Research Fellow of the University of St
Andrews, Scotland. He holds doctorates in both philosophy and theo-
retical physics. He received the international ‘Springer Thesis Prize’ for
his doctorate in physics at St Andrews (2014) and the international
‘Expanded Reason Award’ for his doctorate in philosophy at Cambridge
(2020). He has wide-ranging interests across metaphysics, the philoso-
phy of physics, the philosophy of science, and philosophical theology.
He is co-editor of Neo-Aristotelian Perspectives on Contemporary Sci-
ence (Routledge 2017) (with Robert C. Koons and Nicholas Teh).
Index
ability 193–194, 196, 204–205 providence 392–395; transcendent

accident 71–72, 73–74, 80, 275 causality 400–402
accidental form see form, accidental Aristotle: agency 169, 172, 174–175;
action: Aristotelian account of astronomy 271–272; desire 177–178;
173–174; disjunctivism 180–185; divine unity 330–333; hierarchy
intentional vs. unintentional 181; of being 102; hylomorphism
as psychophysical process 176–180; 24–28, 69–72; living organisms 67;
see also agency; agent; causalism mathematics 242; nature of unity
actuality 327–328, 335, 339 330–333; nature (physis) 346–349;
afection 177–178, 188n20 perfection 132–133, 139; unmoved
agency 24; Aristotelian account of mover 327–329
174–175; control 198–201; human Armstrong, David 417, 435n7
67–68, 161–164, 176–180, 215; of artifact 344, 347–348
organisms 84–86; as psychophysical artifcial 344, 347–348; see also
process 176–180; reductive account natural
of 206–207 atomism 67–68, 348; see also
agent 170–172, 176–177, 183–184 microphysicalism
agent causation 171, 184, 187n4 Augustine 144n25, 273, 296
Ames, William 297
analogy 331–332, 335–338 Bacon, Roger 275–276, 287n51,
angelology 267, 271–283 289n63
angels: and constitution 268, 271–283; beatifc vision 353–354
and the hierarchy of being 97, belief 171, 173, 175, 198
105–106, 111–112, 114–118; and Bonaventure 275–277
the supernatural 350–352 bundle theory 153–154, 159, 269,
animalism 169, 245 270, 285n12, 286n41, 287n52,
animals: as agents 171, 176, 186–187; 308; see also substance
hierarchy of being 105, 110–113; as
substances 105, 131, 347 Cartesian dualism see dualism,
Anselm 131–132, 295–296 substance
Aquinas, Thomas 80, 102–103, Cartwright, Nancy 54, 193
410–411; divine unity 333–335, categoricalism 208
295–297; essence vs. existence causal closure 186; see also
349–351; hylomorphism 27–28, physicalism
245–246, 274, 278–279; necessity causal exclusion, principle of 209
386–387; perfection 132–133, 139, causalism 171–172, 175–176, 182,
143n18; power of thought 252; 187n4, 212; see also noncausalism
primary and secondary causation causal overdetermination, problem of
386–400; proportionality 135; 206, 210
442 Index
causal pluralism 209 Davison, Andrew 296, 316n24
causal powers 24, 159, 207–211; death 246–247, 252, 258, 260–261
active vs. passive 84, 208–209, deism 354
348, 412; agent vs. patient demon 272–273, 283, 352
209; of animals 186–187; and dependence, ontological 31, 42–43,
choice 162–164, 175; collective 52–54, 73, 153, 157, 301–302,
84; co-manifestation 152, 157, 307; essential 297; existential 295;
208–209, 413; directed 412–413; see also grounding, metaphysical;
emergent 51–52, 84–85, 161–164; order, of dependence
of humans 162–164, 186–187; Descartes, René 30, 74; see also
irreducibility of 25, 85; natural vs. Cartesian dualism
preternatural 352–353; one-way vs. desire 173, 175, 177, 186, 198
two-way 210–211; of organisms determinacy, formal 221–222,
68, 84–86; to perform experiments 230–232
207–211; platonic 415–417, 419; determinism 155, 162–163, 366, 369,
primary vs. secondary 349, 369, 374, 379–380
392, 396–400, 403 deviant causal chains, problem of 181,
causation: efcient 29, 42, 74; fnal 206, 211–214
347–348, 352–354; formal 29, Dewhurst, Joe 242n43
42, 43, 70; material 70, 76; and Dionysius the Areopagite 278
the principle of proportionality disappearing agency, problem of
128–130; and relation 156–158; 173–174, 176, 206
and time 154–156 disembodiment 246–247
chain of being 94–97; see also dispositionalism 40–41, 355; kind-
hierarchy of being based 410–415; platonic 416–417,
change 25–26, 174, 182 422–423; standard 410, 412–415;
Charles, David 176–180, 188n22, theistic 417–424
189n25 dispositions 151–152, 156–157, 165n4
choice 162–164, 175; see also intention divine application, doctrine of
Chomsky, Noam 357 397–398; see also government,
cognitive science 220–221, 236–237 divine; providence
common nature 280, 284n2; see also divine conceptualism 419–421
universals divine concurrence, doctrine of
compatibilism, theological 369–370, 349–350
380–381 divine conservation, doctrine of
computationalism 220, 237–238 349–350
conservation of overall perfection, divine exemplarism 296–297,
principle of 139–141; see also 418–419, 422–423
perfection divine simplicity 294, 357, 422
content, mathematical see mathematical divine unity 328–330, 339; in Aquinas
and logical form/content 333–335; in Aristotle 335–338
content fallacy 235–236 dualism: property 51–53; substance
contingency 387–390, 401–402 67, 170–171, 185–187, 237,
continuity, principle of 96, 103–104 283–284
Coope, Ursula 173–175, 182 Duve, Christian de 388
Copleston, Frederick 271–272
counterfactual 225–226, 239n17, Eastern Orthodox Church 273
368–369, 379 Eleatic principle 25, 48, 205
creation 349–350 eliminativism 192–193
embodiment 25, 27, 196–198
Darwin, Charles 97, 125 emergence: ontological 85; strong 50,
Davidson, Donald 172, 175, 177, 85, 161–162, 166n19; weak 47,
187n5 161–162, 166n19
Index 443
emotion 178, 198 grace 353, 355–356, 359
empiricism 47, 193 gradation, principle of 97, 99, 103
Enlightenment 98–99 grounding, metaphysical 41, 43, 215,
error-theory 202–203, 208 251; modal 307–309; necessity
essence 68–69, 88, 100, 297, 304; non-modal 310–314; species
308–313, 349–350 of 301–302; strict partial ordering
event 171–174 303; topic-neutrality 302; total vs.
evolution 69; and divine creation partial 303–304; well-foundedness
364–372, 378–381; evolutionary 305–306; see also dependence,
transitions 125–126, 128, 141; and ontological; fundamentality,
the principle of proportionality 141; metaphysical
and teleology 383–385, 388–391, 403 growing block theory of time 155–156
evolutionism 97
explanation, statistical 366–369, 372, Haddock, Adrian 185–186
376, 379 haecceity 274, 280–282
Haldane, John 258–260
Feser, Edward 261, 394 Hales, Alexander of 274–275, 283
Fine, Kit 308–309, 311–312 Harré, Rom 199
frst-personal thought 256–257 Haslanger, Sally 325–326
Fischer, Kendall 252–253, 255–256 Heil, John 152–153, 156–158
Fishacre, Richard 273, 275 hierarchy of being 27, 94–96; defence
form: accidental 26, 29, 53, 146n44, of concept 106–118; demise of
347; in Aristotle 25–27; and concept 91–101; formalized
hylomorphic unity 323–327; in 105–106; recovery of concept
Plato 25; in quantum mechanics 101–104
41–43, 52–56; in scholasticism Humean supervenience 23–24, 31;
28–29; in the Scientifc Revolution and quantum mechanics 32
29–30; substantial 26–31, 42, hylomorphic unity: conceptual
52–56, 70, 74–77, 87–89, 245–246, distinction model of 323–324,
269–270, 275–277, 347, 349–351, 327, 329–330, 339; ontological
394–395 distinction model of 324–327,
free will 161–164 329–330, 332, 339
frequentism see probability, hylomorphism: and action 178–179;
frequentism Aristotelian-Thomistic 25–27, 69–72;
function, pure 221–223, 238n6 and evolution 87–89; in quantum
functionalism 86, 194, 221, 238n3, mechanics 41–43, 52–56, 72–79; in
238n8, 240n31 scholasticism 28–29; vs. substratum-
fundamentality, metaphysical object view 165n4; Thomistic
197–198, 301, 304–305, 315; see 245–246, 248–250, 269–270,
also grounding, metaphysical 278–280; unity (see hylomorphic
unity); universal 273–278
George, Robert P. 259–260 Hyman, John 213
Gill, Mary Louise 323–324
God: and evolution 364, 478–481; idealism 67, 201
as fundamental ground of reality immateriality 220, 233
293–299, 315; and the hierarchy of indeterminism 157, 163, 167n22,
being 101–106, 111–112, 116–118; 368–370
and law-making 419–424; unity indispensability 201, 204, 214
of 333–339; as universal cause 398– individuation 70, 81–84, 268–269,
400; as transcendent cause 400–402 274–277, 279–281, 313
Gould, Steven Jay 383–384, 388–389 indivisibility 325, 329, 333–339
government, divine 392–393, infnite descent 305, 317n29; see also
394–395; see also providence grounding, metaphysical
444 Index
instantiation 269, 358, 413–414 materialism 220, 229, 235–236;
intellective operation 253 see also functionalism;
intellective power see Aquinas, microphysicalism; physicalism
Thomas, power of thought mathematical and logical form/
intellectualism 170 content 233–234; see also logic;
intention 162–164, 170, 180–181, mathematics
186, 194, 206, 213–214 mathematics 220, 227; and
Aristotelianism 237
Jaworski, William 205 matter: in Aquinas 27; in Aristotle
Jenkins, John 358–359 26; and hylomorphic unity
John of Damascus 295 323–327; prime vs. proximate
70; in scholasticism 28–29; in
kinds, natural 25, 85, 146n44, 248, the Scientifc Revolution 29–30
360n11, 408–409, 410–413, (see also microphysicalism);
417–423 spiritual vs. corporeal 274–278;
Klima, Gyula 263n16 substrate 28–30, 153
Koslicki, Kathrin 240n33, 311–312, Maximus the Confessor 288n55
324–329 McGinn, Colin 357
Kripke, Saul 221–222 McGrath, Alister 390
Kripke-Ross Argument (KRA) mental state 171–172, 174–176,
221–224, 238n3 181, 186; see also belief; desire;
psychological state
laws of nature 193; disjunctive 418; as mereology 323–328, 333–338;
dispositions (see dispositionalism); see also parthood
exclusion 418; functional 414, 417; microphysicalism 23, 29–31, 74–75,
Humean account of 39–40, 157, 170; see also physicalism
410, 412; Lewis’ best-ft account of Milikan experiment 200, 205
374, 376–378, 409, 423; theistic mind-body problem 170, 180; see also
best-ft account of 376–378; dualism, substance; functionalism;
uninstantiated 415, 417 materialism; physicalism
Lee, Patrick 259–260 miracle 354–355, 392
Lewis, C. S. 22–23 Molinism 366, 368–371, 379, 381
Lewis, David 14, 23, 165n6, monism 48, 68, 102, 209; see also
374–375, 376–377, 409, pluralism
422–423, 432, 435 Morris, Simon Conway 383, 389–390
Lewis, Frank 324 motion 174–176, 180, 183–184, 199
logic 220, 227; and motives of credibility 358–359
Aristotelianism 237 multiverse 371–372, 385
logical form/content 233–234 mystery, theological 356–358
Lonergan, Bernard 400
Loux, Michael 326–327, 329, 421 natural 343–348; see also artifcial;
Lovejoy, Arthur 95–99 preternatural; supernatural
Lowe, E. J. 152–153, 158, 163–164, naturalism 126–127, 141, 193,
306, 311–313, 411, 413, 426n20 204–205, 343–345, 357, 408–409,
Lubac, Henri de 353 413–415, 422–423; see also
luck 370–371 physicalism
necessity: of the consequence
Macpherson, Fiona 185–186 385–387, 402–403; of the
manifest image of humans 192, 196, consequent 385–386, 389–390,
198, 202, 214–215, 217n15; see 402–403; of humankind 385–390,
also scientifc image of humans 401–402
Marmodoro, Anna 58n10, 157–158, neo-Platonism 96, 101, 103, 137,
208–209, 323 142, 272
Index 445
noncausalism 163–164, 177; see also potentiality 327–328, 330–333, 339
causalism powers 193–194; causal (see causal
powers); of choice 162–164;
object 152–154, 159; see also essential 102; generic 102, 106,
substance 108–111; and intentionality 85;
occasionalism 349–350, 424 mathematical and logical 227, 229,
Ockham, William of 133 236–237; rational 67, 110
Oderberg, David S. 355 preambles of faith 358–359
Olson, Eric 254, 264n16 presentism 156
omnirationality, divine 378, 381 preternatural 352–353, 361n31;
ontological dependence see see also natural; supernatural
dependence, ontological primitive substance theory 269;
ontology, constituent 152–153, see also substance
323–327 probability: axioms of 372–375; and
order: of dependence 132, 134–136; causation 157; and choice 164;
of eminence 132, 134, 137–141; classical interpretations of 372–372;
see also dependence, ontological; frequentism 14, 372–373, 375,
grounding, metaphysical 376–378, 380–381; Lewis’ best-ft
organism 68–69, 72, 84–87, 163 account of 374–376; liberal view
overdetermination see causal of 365, 373–374, 375, 380; in
overdetermination, problem of quantum mechanics 31–32, 34, 77;
theistic best-ft account of 376–371
parsimony 424; ideological 416; process theology 454
ontological 416, 423; quantitative property 409; monadic 411, 413;
vs. qualitative 416 platonic 415–417, 419; polyadic 413
parthood 328–329, 335–337; see also proportionality, principle of
mereology 128–130, 141
participation 295–300, 315; see also providence 384, 402–403; Aquinas’
universals divine ordering account of
perfection: negative 133–141; 392–394; and causation 396–402;
privative 133–135 God’s execution of 394–396; and
personhood 246, 254 necessity 385–390; secular account
philosophy of mind see of 390–392; see also government,
computationalism; dualism, divine
substance; functionalism; Pruss, Alexander 86
physicalism psychological state 177–180; see also
physical computation 233–236, mental state
234, 236, 239n20; see also psychophysicality 170, 179
computationalism psychophysical processes 178–179, 185
physicalism 192–193, 206–207, 228– pure nature 353–354
229, 235; see also microphysicalism
physiological processes 177–178, 180 quantum mechanics: Bohmian theory of
Place, U.T. 420 37–39; CWC theory of 49–50; GRW
plant 105–106, 337 theory of 35–36; indeterminacy
Plantinga, Alvin 355 34–35, 73, 78; quantum chemistry
Plato 25, 67, 131; see also and thermodynamics 72–79;
neo-Platonism quantum entanglement 31–34;
plenitude, principle of 96, 99, 103–104 superposition 32, 34–35; wave
Plotinus 131, 272 function 34–38, 42–43
pluralism: neo-Aristotelian 68–69,
77, 89; ontological 270, 278–280, rational embodied control 194, 204,
287n52 214; see also control; embodiment;
potency 269, 275–276; obediental 354 rationality
446 Index
rationality 194–196 28–29; and substantial form 28–30,
realization 222, 226–227, 235 42–43, 52–56; substratum theory
reason 172, 181, 185–186, 187n8, 194 of 153, 165n4, 165n6, 268, 284n3;
reductionism see microphysicalism; thermal 53–54, 71–74, 81–84; unity
naturalism; physicalism of 26–28, 80–81
Romanticism 98 substantial form see form, substantial
Ross, James 221, 230–231, superiority, metaphysical 106–118;
240n32, 241n41 see also fundamentality,
rule 231, 233–235, 241n35 metaphysical
Ruse, Michael 384–386, 387–388 supernatural 343–344, 351–354;
see also natural; preternatural
Scaltsas, Theodore 324 Swinburne, Richard 254–256,
Schafer, Jonathan 68, 70, 306, 358, 418
408, 423 Szathmáry, Eörs 390
scholasticism 28–29, 128–130
scientifc experiment 195–196, Taylor, Charles 203–204, 217n14
202–204, 214 teleology 68, 77, 84–85, 352, 361n22,
scientifc image of humans 192–193, 384–385, 390–395, 398–400, 402,
202–203, 214–215; see also 424; contingent 391; extrinsic 393;
manifest image of humans intrinsic 393; necessary 391
scientifc practices 193, 198, 205, 207 theism: explanatory breadth 198–199;
scientifc realism 43; pluralistic explanatory immediacy 299–200;
approach to 46–47 and laws of nature 419–424
Scientifc Revolution, the 29–30, 67, 98 Thomism 380–381
Scotus, John Duns 132, 136–137, time 154–156
144n24, 280–282 Toner, Patrick 252, 258, 260
semantics 233–234; see also too many thinkers, problem of
mathematical and logical form/ 254–257
content transcendental argument 201–203
Sellars, Wilfred 192 trinity 356–357, 359
skepticism 202–204 triviality argument 222, 238n6
soul 179, 189n24, 337–338, 350; trope 152–153, 158, 268–269
corruptionism 246–247, 249–250, truthmaker 155–156, 194, 410–414
257, 260; survivalism 246–247, 261 Tugby, Matthew 416–417
Stabler, Edward 223, 225, 238n9
standard causal theory of action 212; umwelt 199
see also causalism underdetermination 227–229
Steward, Helen 160, 210 unity: divine 328–339; hylomorphic
structure 222–223, 226, 233–234 (see hylomorphic unity)
Stump, Eleonore 387, 404n17, universals 5, 25, 39, 58n7, 152–153,
407n82 165n3, 166n12, 248, 268–269,
Suarez, Francisco 133, 143n8, 146n45 281, 284n2, 311, 411, 413,
substance: in Aristotle 25–27; and 415–417, 419–424, 426n20
causes 159–160; and hylomorphic univocality 327, 330–331,
unity 324–327; integral vs. virtual 334–335, 339
parts 71–73; ontology 268–271,
273–274; in quantum mechanics Williams, Porter 46–47
41–43, 48, 52–56; in scholasticism Wittgenstein, Ludwig 173

(Routledge Studies in Metaphysics) William M.R. Simpson, Robert C. Koons, James Orr - Neo-Aristotelian Metaphysics and The Theology of Nature-Routledge (2021)

Uploaded by

Copyright:

Available Formats

You might also like

(Routledge Studies in Metaphysics) William M.R. Simpson, Robert C. Koons, James Orr - Neo-Aristotelian Metaphysics and The Theology of Nature-Routledge (2021)

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

(Routledge Studies in Metaphysics) William M.R. Simpson, Robert C. Koons, James Orr - Neo-Aristotelian Metaphysics and The Theology of Nature-Routledge (2021)

Uploaded by

Copyright:

Available Formats

“This book ofers a fresh, bold, stimulating take on some of the hardest

questions philosophers and theologians alike are engaged with, arguing

“Neo-Aristotelianism is one of the important movements in contemporary

“While scholasticism has been on the margins of Christian theology for

“Despite its long neglect in mainstream philosophy of science, neo-

“When the Scientifc Revolutionaries of the seventeenth century left

This book explores the relationship between a scientifcally updated Aristotelian

William M. R. Simpson is Junior Research Fellow at Wolfson College, Cambridge,

Robert C. Koons is Professor of Philosophy at the University of Texas at Austin.

James Orr is University Lecturer in Philosophy of Religion at the University of

Eliminativism, Objects, and Persons

Metametaphysics and the Sciences

Authority and the Metaphysics of Political Communities

Language and Word

Neo-Aristotelian Perspectives on Formal Causation

A Powerful Particulars View of Causation

Neo-Aristotelian Metaphysics and the Theology of Nature

For more information about this series, please visit: www.routledge.com/

Edited by William M. R. Simpson,

Introduction: Refections on Science, Theology, and the New

1 From Quantum Physics to Classical Metaphysics 21

2 Essential Thermochemical and Biological Powers 66

3 Restoring the Hierarchy of Being 94

4 Evolution and the Principle of Proportionality 125

5 Free Will in a Network of Interacting Causes 151

6 Animal Powers and Human Agency 169

8 Logical and Mathematical Powers 220

9 Persons, Souls, and Life After Death 245

10 Angels, Principalities, and Powers 267

11 Grounding and Participation in God 293

12 God and Hylomorphism 322

13 Natural and Supernatural 343

14 God, Chance and Evolution: In Memory of

15 Teleology, Providence, and Powers 383

16 Divine Lawmaking: Powers and Kinds 408

About the Contributors 437

The editors owe a debt of gratitude to many colleagues and professional

I was asked to write this Prologue as an historian of (medieval) philoso-

Decline and Fall of the Old Aristotelianism

Rise of the New Aristotelianism

Modern Theology and the New Aristotelianism

For Further Reading

Naturalism and Nature

2.1. Modern Microphysicalism

all there is to the world is a vast mosaic of local matters of particular

Proposition 2.1. The whole truth about nature supervenes upon

2.2. The Doctrine of Hylomorphism

Whilst the ontological status of the forms in Aristotle’s metaphysics is

Proposition 2.3. The physical parts of a (macroscopic) substance do

2.3. Matter and Form in Late Scholasticism

The aggregation of multiple faculties or accidental forms in a simple

2.4. The Rise of Microphysicalism

3. Form in Standard Quantum Mechanics

3.1. Quantum Entanglement

According to this formalism, when a system comprised of two particles

3.1.1. Non-Local Correlations

P°a ,°b ( A | B,˛ ) = P°a ( A | ˛ ) (2)