Dtoken

ARISTOTLE O N I NQUIR Y
Aristotle is a rarity in the history of philosophy and science – he is

a towering figure in the history of both disciplines. Moreover, he
devoted a great deal of philosophical attention to the nature of
scientific knowledge. How then do his philosophical reflections on
scientific knowledge impact his actual scientific inquiries? In this
book, James Lennox sets out to answer this question. He argues
that Aristotle has a richly normative view of scientific inquiry, and
that those norms are of two kinds: a general, question-guided frame-
work applicable to all scientific inquiries, and domain-specific norms
reflecting differences in the target of inquiry and in the means of
observation available to researchers. To see these norms of inquiry in
action, the second half of this book examines Aristotle’s investigations
of animals, the soul, material compounds, the motions of heavenly
bodies, and respiration.
james g. lennox is Professor Emeritus of History and Philosophy of

Science at the University of Pittsburgh. He has published widely on the
history and philosophy of biology, with a focus on Aristotle, William
Harvey, Charles Darwin and Darwinism. His books include Aristotle’s
Philosophy of Biology (Cambridge, 2001) and a translation, with com-
mentary, titled Aristotle: On the Parts of Animals (2001) in the Clarendon
Aristotle Series.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6
C 1D19 12 1 8 31 2 947 73 8 4 9 7
ARISTOTLE ON INQUIRY
Erotetic Frameworks and Domain-Specific Norms
JAMES G. LENNOX
University of Pittsburgh
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
University Printing House, Cambridge cb2 8bs, United Kingdom
One Liberty Plaza, 20th Floor, New York, ny 10006, USA
477 Williamstown Road, Port Melbourne, vic 3207, Australia
314–321, 3rd Floor, Plot 3, Splendor Forum, Jasola District Centre,
New Delhi – 110025, India
79 Anson Road, #06–04/06, Singapore 079906
Cambridge University Press is part of the University of Cambridge.

It furthers the University’s mission by disseminating knowledge in the pursuit of
education, learning, and research at the highest international levels of excellence.
www.cambridge.org
Information on this title: www.cambridge.org/9780521193979
doi: 10.1017/9781139047982
© Cambridge University Press 2021
This publication is in copyright. Subject to statutory exception
and to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without the written
permission of Cambridge University Press.
First published 2021
A catalogue record for this publication is available from the British Library.
Library of Congress Cataloging-in-Publication Data
names: Lennox, James G., author.
title: Aristotle on inquiry : erotetic frameworks and domain-specific norms /
James G. Lennox, University of Pittsburgh.
description: Cambridge, United Kingdom : New York, NY, USA : Cambridge University
Press, 2020. | Includes bibliographical references and index.
identifiers: lccn 2020034989 (print) | lccn 2020034990 (ebook) | isbn 9780521193979
(hardback) | isbn 9781139047982 (ebook)
subjects: lcsh: Aristotle. | Inquiry (Theory of knowledge)
classification: lcc b485 .L46 2020 (print) | lcc b485 (ebook) | ddc 121/.6–dc23
LC record available at https://lccn.loc.gov/2020034989
LC ebook record available at https://lccn.loc.gov/2020034990
isbn 978-0-521-19397-9 Hardback
Cambridge University Press has no responsibility for the persistence or accuracy of
URLs for external or third-party internet websites referred to in this publication
and does not guarantee that any content on such websites is, or will remain,
accurate or appropriate.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
Dedicated to the memory of Allan Gotthelf
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6
C 1D19 12 1 8 31 2 947 73 8 4 9 7
Contents
List of Figures page ix

Preface xi
List of Abbreviations xv
Introduction 1
part i erotetic frameworks and domain-specific

norms 9
1 The Goal of Knowledge and Norms of Inquiry 11
2 An Erotetic Framework: The Posterior Analytics on Inquiry 37
3 A Discourse on μέθοδος 65
4 Natural Science: Many Inquiries, One Science 99
part ii norms of natural inquiry 119

5 The Methodos of Nature 121
6 Shaping the Methodos of Animals: The Purpose of Parts
of Animals i 141
7 The Soul: One Subject, Two Methods? 174
8 The Order of Inquiry i: Right and Left in De caelo
and De incessu animalium 200
9 The Order of Inquiry ii: The Debt of Zoological Inquiry to
Meteorology iv 227
vii
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
viii Contents
10 Aristotle on Respiration: Framework Norms Meet
Domain-Specific Norms 264
Bibliography 293
Index Locorum 307
General Index 314
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
Figures
6.1 Division of Wings, after HA I.5, 490a5-16. page 151

6.2a Forms of bird feet, indicating differentiation for different 155
ways of life.
6.2b Forms of bird beaks, indicating differentiation for different 156
ways of life.
6.3a Eledone cirrhosa (the ‘lesser’ or ‘curled’ octopus) with a single 158
row of suckers on each arm.
6.3b Sepia officinalis (cuttlefish) with feeding tentacles extended. 158
ix
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7
D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
Preface
This book has had a long gestation period, during which I have accumu-
lated many debts. Throughout the process, Patricia Lennox has supported
my work in every way imaginable – not the least of which is filling life
beyond my work with joy. Early on in the process I invited Alan Code to
Pittsburgh for an extended visit for the sole purpose of discussing this
project, and those discussions were critical in helping me frame it.
Discussions over meals and long walks with my former colleague James
Allen and with Allan Gotthelf, David Charles, Philip van der Eijk, and
István Bodnar also were helpful at this early stage. At a somewhat later
stage, Mariska Leunissen and Joe Karbowski read through a preliminary
draft and discussed difficulties of overall structure with me – the two-part
structure of the final version I owe to them. In a similar vein, Aryeh
Kosman and Joel Yurdin visited Pittsburgh and met with Jon Butacci
and me for discussions of how Aristotle sees experience in relation to
inquiry, which helped to alter the trajectory of my work at a crucial point.
While the manuscript was developing, a number of people did me the
(helpful) honor of organizing workshops where I could present and discuss
the ideas here presented. Klaus Corcilius organized one such event at the
University of California, Berkeley, March 6–7, 2015, during which
I presented an overview lecture and took part in four seminars organized
around specific chapters. I received helpful comments from Timothy
Clarke, Lucas Angioni, Michel Curbellier, and of course Klaus Corcilius
on that occasion.
Victor Caston organized a similar event at the University of Michigan,
Ann Arbor, April 16–17, 2015, during which I received helpful suggestions
from Victor, Richard Janko, Emily Katz, and James Joyce. During that
visit I also had two memorable conversations over coffee about the project
with Victor. My dear friend Giovanni Camardi and his colleague
Giovanna Giardina invited me to give two lectures based on material
from three chapters at University of Catania, Sicily, May 11–12, 2015,
xi
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
xii Preface
during which I had memorable conversations with them and their students
about the book’s themes. Marko Malink organized a workshop at
New York University on April 22–23, 2016, on Chapters 1–7, with chapter
presentations by Josh Mendelsohn (Chapter 1), Marko (Chapter 2), Joel
Yurdin (Chapter 3), Pieter Sjord Hasper (Chapter 4), Greg Salmieri
(Chapter 5), Claire Bubb (Chapter 6), and Jessica Moss (Chapter 7). The
discussions during each of these sessions helped me identify assumptions
that needed defense and gaps in the argument that needed to be filled.
David Charles and Verity Harte invited me to present some of the ideas
discussed in Chapters 2 and 3 to the Ancient Philosophy Colloquium at
Yale University on October 14, 2016. Besides doing a wonderful job of
organizing that visit, Emily Kress met with me to discuss her work on
Physics ii and how it might relate to the ideas expressed in Chapter 5. On
April 13–14, 2017, besides delivering a lecture on Aristotle and Darwin to
the Rotman Institute of Philosophy, my friend and fellow Canadian Devin
Henry organized a seminar around themes developed in the book.
I benefited enormously from these events, probably even more than I am
now aware.
During these years, I also had the pleasure of presenting the themes of
the book in seminars (various iterations of “Aristotle’s Natural Science”
and “Aristotle on Method”) attended by outstanding graduate students in
Philosophy and History of Philosophy of Science and colleagues Allan
Gotthelf, Jessica Gelber, and Kathleen Cook. Working with Keith Bemer,
Jon Buttaci, Tom Marré, and Greg Salmieri during those years, as they
researched and wrote dissertations on themes closely connected to mine,
was especially valuable, but all the participants in these seminars contrib-
uted in different ways to the final product.
In addition to these many helpful live interactions, I received valuable
written comments from Christina Hoenig and Jacques Bromberg in our
Classics Department at the University of Pittsburgh, as well as from Jon
Butacci, David Charles, Alan Code, Andrea Falcon, Jessica Gelber, Allan
Gotthelf, Mariska Leunissen, Joe Karbowski, Emily Katz, Emily Kress,
Tom Marré, Anne Peterson, Diana Quarantotto, Stasinos Stavrianeas, and
Joel Yurdin.
I also thank Cambridge University Press’s two readers for their
extremely helpful questions, suggestions, and corrections. One of these,
Sarah Broadie, revealed her identity to me, so I am able to personally thank
her for her detailed, constructive questions and suggestions. For their help
at the production stage, it is a pleasure to thank Michael Sharp and Mary
Bongiovi for editorial guidance, Bret Workman for his help with
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
Preface xiii
copyediting, and Akash Datchinamurthy for overseeing production. And
last but very far from least, I owe special thanks to Christopher Kurfess,
who not only helped with the preparation of the indexes but also offered
valuable advice on many other questions that came up during the copy-
editing process.
This book is dedicated to my dear friend Allan Gotthelf. He joined me
in my department in 2003, shortly after I began working on this book, and
discussions with him about it were a constant source of inspiration.
I deeply regret that Allan passed away before work on it was completed.
While all of the above-mentioned people (and no doubt many others)
have helped to make this a better book than it would otherwise be, I doubt
any of them will find the final product fully convincing. I take full
responsibility for the result.
Certain ideas and themes in Chapter 3 were first presented in April 2010
in a paper entitled “Aristotle on Norms of Inquiry” as part of a Biggs
Residency at Washington University, St. Louis, and then at the History of
Philosophy of Science meetings in Budapest in June; it was published in
2011 in the first issue of the society’s journal, HOPOS 1(1): 23–46. Chapter 4
grew out of ideas first presented as “Aristotle’s Natural Science: the Many
and the One” at the 2009 Duke-UNC-Chapel Hill Conference on Ancient
Philosophy, where I received helpful comments from Gisela Striker; and as
the 2010 Rosamond Kent Sprague Lecture in Ancient Philosophy,
University of South Carolina, published in From Inquiry to
Demonstrative Knowledge: New Essays on Aristotle’s Posterior Analytics,
J. H. Lesher, ed., Kelowna, BC: Academic Printing and Publishing,
2010: 1–24; also published in 2010 in Apeiron 43(2/3): 1–24. Chapter 5
relies heavily on ideas first published as “How to Study Natural Bodies:
Aristotle’s μέθοδος” in Aristotle’s Physics: A Critical Guide, Mariska
Leunissen, ed., Cambridge University Press, 2015: 10–30. Chapter 6 devel-
ops some of the ideas presented in “The Unity and Purpose of On the Parts
of Animals i,” in Being, Nature, and Life in Aristotle, James G. Lennox and
Robert Bolton, eds., Cambridge University Press, 2010: 56–77. Chapter 7
is based on a lecture presented at Flinders University in 2009 and published
as “Aristotle on Mind and the Science of Nature” in Marietta Rossetto,
Michael Tsianikas, George Couvalis, and Maria Palaktsoglou, eds., Greek
Research in Australia: Proceedings of the Biennial International Conference of
Greek Studies, June 2009, Flinders University Department of Languages,
2009: 1–18. Chapter 8 develops ideas first presented in “De caelo ii.2 and Its
Debt to De incessu animalium,” in Alan C. Bowen and Christian Wildberg,
eds., New Perspectives on Aristotle’s De caelo (Philosophia Antiqua, Vol.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
xiv Preface
117), Leiden and Boston: Brill, 2009: 147–214; and Chapter 9 builds on
“Aristotle on the Emergence of Material Complexity: Meteorology iv and
Aristotle’s Biology,” presented as part of a 2012 HSS/HOPOS symposium
on Meteorology iv organized by Tiberiu Popa, published in 2014 in
HOPOS 4(2): 272–305. Finally, Chapter 10 is based on material first
presented at University of Patras in June 2011, and in subsequent iterations
at the University of Toronto, University of Rochester, Oxford University,
University of Pisa, University of Notre Dame, Aristotle University of
Thessaloniki, Humboldt University (Berlin), Catholic University of
America (Washington), and Charles University (Prague). Some of the
ideas presented in this chapter appear as “Why Animals Must Keep
Their Cool: Aristotle on the Need for Respiration (and Other Forms of
Cooling),” chapter 10 in Heat, Pneuma and Soul in Ancient Science and
Philosophy, Hynek Bartos and Colin King, eds., Cambridge, 2020.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
Abbreviations
Aristotle
APo. Posterior Analytics (Analytica Posteriora)
APr. Prior Analytics (Analytica Priora)
Cat. Categories (Categoriae)
Cael. On the Heavens (De caelo)
de An. On the Soul (De anima)
EE Eudemian Ethics (Ethica Eudemia)
EN Nicomachean Ethics (Ethica Nicomachea)
GA On the Generation of Animals (De generatione animalium)
GC On Generation and Corruption (De generatione et corru-
ptione)
HA History of Animals (Historia animalium)
IA On the Locomotion of Animals (De incessu animalium)
Juv. On Youth and Old Age (De juventute et senectute)
Long. On Length and Shortness of Life (De longitudine et brevi-
tate vitae)
MA On the Movement of Animals (De motu animalium)
Mem. On Memory (De memoria et reminiscentia)
Metaph. Metaphysics (Metaphysica)
Mete. Meteorology (Meteorologica)
PA On the Parts of Animals (De partibus animalium)
Ph. Physics (Physica)
PN Parva naturalia
Po. Poetics (Poetica)
Pol. Politics (Politica)
Pr. Problems (Problemata)
Resp. On Respiration (De respiratione)
Rh. Rhetoric (Rhetorica)
Sens. On Perception and Perceptibles (De sensu et sensibilibus)
xv
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
xvi List of Abbreviations
Somn.Vig. On Sleep and Waking (De somno et vigilia)
Top. Topics (Topica)
Plato
Phlb. Philebus
Plt. Statesman (Politicus)
R. Republic
Sph. Sophist
Ti. Timaeus
Hippocratic Writings
Carn. On Fleshes (De carnibus)
VM On Ancient Medicine (De vetere medicina)
Other Ancient Authors
D.L. Diogenes Laertius, Lives of the Philosophers
in de An. Philoponus, Commentary on Aristotle’s De anima
in Cael. Simplicius, Commentary on Aristotle’s De caelo
in Mete. Alexander of Aphrodisias, Commentary on Aristotle’s
Meteorology
in Ph. Philoponus, Commentary on Aristotle’s Physics
Simplicius, Commentary on Aristotle’s Physics
Modern Authorities
DK H. Diels & W. Kranz, Die Fragmente der Vorsokratiker,
6th ed., 1951.
LSJ H.G. Liddell, R. Scott, & S. Jones, Greek–English
Lexicon, 9th ed., 1968.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
Introduction
Dante refers to Aristotle as “il maestro di color che sanno,” the master of
those who know.1 But Aristotle typically refers to the works that have come
down to us as ‘inquiries’ or ‘investigations,’ and in the pages that follow,
I will make a case for Aristotle as “il maestro di color che cercano,” the
master of those who inquire. The chapters to follow attempt to get clarity
on Aristotle’s conception of inquiry, insofar as the goal of inquiry is
scientific knowledge (ἐπιστήμη).2 Does Aristotle see inquiry, as he clearly
sees explanation, as a process constrained by epistemic norms – norms of
inquiry, as I am calling them? That is, given that Aristotle has clearly
articulated ideas about what the goal of scientific inquiry looks like, does he
also have clearly articulated norms that must be adhered to if one is to
achieve that goal?
Typically, when scholars ponder where to find Aristotle’s views on that
topic, they turn to the second book of the Posterior Analytics (APo.), which
begins by characterizing four different objects of inquiry “equal in number
to things we know” (ἴσα τὸν ἀριθμὸν ὅσαπερ ἐπιστάμεθα, 89b23-24). That
discussion operates, however, on a rarefied plane of abstraction – so
rarefied that Aristotle is at times happy to exemplify the same philosophical
thesis by means of examples drawn from geometry and botany. What that
discussion provides, I argue in the chapters that follow, is an erotetic
framework – a ‘logic of questions and answers,’ as it were – for any inquiry
aiming at achieving scientific knowledge. Every such inquiry must operate
within the general guidelines provided by this framework, which in turn is
1
Inferno, Canto 4, line 131 (Durling 1996, 77).
2
How best to translate this term is a matter of longstanding dispute. A tradition going back at least to
Kosman 1973, and endorsed and further defended by Burnyeat 1981, would render it ‘understanding’
and stress the centrality of the capacity to explain or demonstrate to its possession. Since Aristotle
uses a variety of terms to designate weaker and stronger forms of knowing, another option,
challenged by Burnyeat, is ‘scientific knowledge,’ to indicate that ἐπιστήμη, at least in its unqualified
form, is the sort of knowledge that one who has mastered a science has.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
2 Introduction
shaped by the detailed characterization of scientific knowledge provided in
APo. i. However – and this is among my central themes – this turns out to
be the wrong place to turn if one wants to understand Aristotle’s views
about the norms of inductive inquiry that must guide research in distinct
domains. Many, if not most, of these norms, Aristotle holds, are domain-
specific – local, not global.
Where then does one look for Aristotle’s thoughts about substantive
norms of inquiry? Historians of philosophy and science face an obvious
difficulty in attempting to answer that question: we have no written reports
from those who observed Aristotle the inquirer, telling us how he carried
out different inquiries; nor do we have (as we do with so many researchers
from the seventeenth century on) his ‘lab’ or ‘field’ notebooks either,
assuming some fourth century bce analogues of such things actually
existed.3 All we have are treatises reporting the results of his inquiries.
Does that mean it is impossible for us to investigate his views about the
norms that ought to guide actual research and the methods by which it
should be carried out?
Fortunately, no: there are at least three fruitful sources of information to
which we can turn in order to answer this question. First, Aristotle wrote
a great deal, at different levels of abstraction, about how inquiry in general,
and different kinds of inquiry, ought to be carried out.4 In fact after you
finish this book, you may well conclude that he was obsessed by this
subject! Thus, if we make the charitable assumption that Aristotle at least
attempted to practice what he preached, then these normative discussions
of inquiry ought to be a rich source of information about how he actually
carried out his inquiries – or perhaps about what he learned about the
nature of inquiry while engaging in them. Much of what Aristotle has to
say on this topic is to be found in self-consciously methodological intro-
ductions to his various inquiries – De partibus animalium (PA) i, De anima
(de An.) i.1, and Nicomachean Ethics (EN) i.1–6 are three familiar examples.
A second source of information that is helpful in exploring which
norms were in play in distinct Aristotelian inquiries involves a sort of
reverse engineering process – starting with the written results we possess
3
To anticipate a thought some readers might be having at this point, the Historia animalium (HA)
decidedly is not an exception. On the place of the HA in Aristotle’s zoological inquiries, see Balme
1987a, Gotthelf 1988 [2012, ch. 14], and Lennox 1991 [2001b, ch. 2].
4
One could also take the attitude that this is all that is of philosophical significance anyway – whether
Aristotle followed his own norms is perhaps an interesting historical question, but irrelevant to their
epistemological virtues. For reasons that I will allow to gradually reveal themselves as the argument
progresses, that is not the attitude I adopt here.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
Introduction 3
and then reflecting on the means by which those results could have been
achieved. Now this would be a hopeless task if the treatises were written
as axiomatically organized presentations of demonstrations – but they
are not. Aristotle invariably refers to them as inquiries or investigations,
and they are often organized in such a way as to model the investigation
on which they are reporting – moving from what is clearer to the
beginning investigator and more obvious to perception toward what is
clearer in itself (i.e., toward the natures and causes of the phenomena in
the domain under study). They often begin by reviewing previous
research on the subject and the puzzles that emerge from subjecting
that previous research to critical scrutiny. Aristotle’s own considered
answers are typically only presented late in the narrative, and are often
presented in preliminary ways first – after which he will often declare
that it is time for a fresh start!
I suspect that this mode of presentation is pedagogical in intent – but the
primary focus is not on teaching scientific content, but on providing object
lessons in the methods and norms appropriate to specific inquiries. And, as
Part ii of this book will show by means of a number of case studies, the
lessons will vary from one domain of investigation to the next.
Finally, there is one other feature (already alluded to) of virtually all his
treatises that is helpful in determining what Aristotle deems the proper
norms of inquiry – the number of pages devoted to identifying the errors
made by previous thinkers in investigating each domain; more often than
not the errors are fundamentally errors of method. In many cases the errors
are due to applying norms or standards that are appropriate in one domain
to another, where they are inappropriate. In a well-known passage near the
beginning of the EN, for example, Aristotle comments that “it is just as
mistaken to demand demonstrations from a rhetorician as it is to accept
[merely] persuasive arguments from a mathematician” (EN i.3,
1094b25-27).
Here again it is rewarding to look at these critical discussions of the
errors of his predecessors as pedagogical in character – not, again, primarily
focused on teaching scientific content (i.e., the results of an investigation),
but rather on instructing future investigators in the methods and norms
appropriate (and inappropriate) to specific inquiries.
It is thus no accident that every treatise of Aristotle’s opens with
a methodological discussion of the norms that ought to govern inquiry
into the subject to which it is devoted. It is to these discussions, I argue,
that one needs to turn in order to understand Aristotle’s epistemology of
inquiry, the norms that must be followed if one is to stay ‘on track’ in the
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
4 Introduction
quest for knowledge – and Part ii of this book will be a study of a selection
of those discussions.
But why, it is reasonable to ask, would Aristotle come to adopt this
domain-specific approach to inquiry? After all, one could believe, as he
does, that knowledge is hierarchically organized, differentiated into
domains and sub-domains, and that at least some of the principles in
each domain and sub-domain are distinctive to it, without concluding
that the modes of inquiry that lead to those principles are also distinctive.
As I see it, there are three grounds for Aristotle insisting that norms of
inquiry are domain-specific.
1. He rejects two forms of reductionism that would encourage one to
think that one global set of norms should be sufficient. One of these is
mathematical, stemming either from a broadly Pythagorean perspective or
from Platonism – or from some combination of the two.5 The other
alternative is some form of materialism: insofar as true knowledge is
possible, it is knowledge of fundamental material elements – atoms, one
or more of the four elements, and so on – which are eternal, changeless, and
underlie and explain the appearances. If one supposes the objects of
knowledge are all of one kind, or at least explicable by reference to one
kind of principle, as either form of reductionism encourages you to believe,
it is plausible to suppose there is a single path to knowledge. Aristotle,
however, denies that our knowledge is restricted to mathematical or formal
entities; and as far as natural objects are concerned, while they have certain
features in common,6 that turns out to be insufficient information to
determine how they ought to be investigated, because nature is, to use
a metaphor of Nancy Cartwright’s, “dappled.”7 Some natural substances –
namely, the heavenly spheres – are constituted of a material that does not
partake of any change other than eternal, circular locomotion, and thus the
objects studied by cosmology and astronomy are natural and eternal.
Others – animals and plants – have souls, are constituted of uniform and
nonuniform parts, undergo a complex process of development that is goal-
directed and governed by a distinctive kind of necessity, are perishable, and
must constantly engage in a complex and coordinated set of activities in
order to remain alive. Still others – the subject matter of meteorology – are
more unstable and ephemeral, such as rain, hail, frost, clouds, thunder,
lightning, or rainbows. In short, the subjects to be investigated differ in
5
These alternatives are discussed in some detail in Metaphysics (Metaph.) Α.6 and 9, and Μ–Ν.
6
For example, they are composite unities of matter and form, they have within them their own sources
of motion and rest, and they are perceptible. I discuss general norms of natural inquiry in Chapter 5.
7
Cartwright 1999.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
Introduction 5
fundamental ways that require distinctive norms of inquiry. And when we
leave the natural realm, either in the direction of mathematics and theology
or in the direction of ethics, politics, rhetoric, or poetics, the norms become
even more distinctive.
2. Our epistemic access to the objects of investigation will vary from one
domain to the next. A little reflection on the examples I just reviewed will
make this apparent. In De caelo (Cael.) Aristotle regularly laments the
limited nature of the phenomena he must depend on in reaching conclu-
sions about the nature of the heavenly bodies or the principles governing
their motions, a point he reiterates in PA i.5 when he notes that while the
objects studied by the cosmologist and astronomer are noble, animals and
plants offer the natural inquirer far more in the way of opportunities for
study, and if contemplated philosophically, “take the prize with respect to
scientific knowledge (ἐπιστήμη)” (645a1–4). The objects investigated by the
meteorologist provide obvious problems of accessibility of a different sort,
as do those of mathematics or ethics. Such limitations must be taken into
account in deciding how an inquiry should be conducted and what counts
as staying on track. One can only imagine what Aristotle would have said
about Spinoza’s attempt to apply the methods of geometry to ethics.
3. Finally, there is an ineliminable perspectival aspect to every scientific
inquiry. To take just one example: consider the rainbow. Aristotle argues
that certain features of rainbows can only be understood by taking into
account the natural interactions between water droplets in the air and light
from the sun. This is to take the perspective of the natural scientist, and the
norms in play would be those of the meteorologist. To understand certain
features of the shapes of rainbows, however, one must adopt the perspective
of geometric optics, as he does in Mete. iii.5–6. The same object is being
investigated from two very different perspectives, and different norms will
be in play depending on the perspective one takes.
These, then, are three reasons for Aristotle insisting on certain norms of
inquiry being domain-specific. A concept that is central in his presentation
of such norms, and to the discussion to follow, is μέθοδος. An oddity of
Aristotelian scholarship over the last few decades is that, while there has
been a great deal of discussion about Aristotle’s philosophical and scientific
methods, that discussion has all but ignored Aristotle’s deployment of this
concept. Of course, there is no a priori reason why a Greek word that
happens to be the etymological root (via Latin transliteration) of our
English word ‘method’ should have a close semantic connection to it –
but in this instance the connections are in fact complex and rich. The
concept’s distribution in the corpus attests to its importance for my
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
6 Introduction
argument. It appears in the first few lines of virtually every treatise report-
ing on one of Aristotle’s inquiries – that is, in those methodological
introductions that discuss the norms of inquiry important for each work.
In stark contrast, it is entirely absent from the Posterior Analytics. I will
make the case in Chapter 3 that it is Aristotle’s concept for the domain-
specific manner in which one ought to inquire – or for an inquiry when
considered from the perspective of its distinctive mode of inquiry.
Chapter 4 concludes Part i by considering the consequences of Aristotle
insisting on domain-specificity within the science of nature. The evidence
is reasonably conclusive that Aristotle thought of all these inquiries as
contributions to a single, integrated science of nature. Much of the litera-
ture on Aristotle’s science (or philosophy) of nature does not face the
difficult consequences of the fact that there are nevertheless many self-
consciously autonomous natural investigations, at several levels of abstrac-
tion and approached from a variety of different perspectives. To under-
stand these various inquiries as contributions to a single epistêmê, we must
consider what sort of integration Aristotle thinks is possible and how
successful he is at achieving it. In this chapter we look at what Aristotle
has to say about what differentiates the science of nature from that of the
two other theoretical modes of inquiry, first philosophy and mathematics.
By a careful study of Aristotle’s discussions of that subject, we can develop
a picture of what every natural inquiry must have in common, qua natural.
That picture will turn out to be a valuable asset to have in hand as we turn
to thinking about the domain-specific norms governing the inquiries to be
investigated and how Aristotle conceives of the interconnections among
the various natural investigations, the subject of Part ii.
Part ii thus opens, in Chapter 5, with a study of what Aristotle refers to
as the methodos of nature – that is, the general norms that must guide any
natural inquiry qua natural. That natural objects are unities of matter and
form – that their natures are ‘dual,’ as he sometimes puts it, and that this
dualism grounds distinctive inherent sources of change, has more far-
reaching implications for natural inquiry than is typically acknowledged,
and drawing out those implications will be a primary task of this chapter.
One theme of Chapter 4, however, was that even within the study of
nature, there is considerable methodological variability, and the remaining
chapters are explorations of the distinctive norms that are to be found in
a number of Aristotle’s natural inquiries. Chapter 6 explores the distinctive
norms of zoological inquiry outlined in the first book of De partibus
animalium and shows those norms at play in a number of specific inquiries.
One question raised in PA i.1 is whether it is appropriate for the natural
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
Introduction 7
scientist to investigate all aspects of the soul, or only some – and Aristotle
appears to conclude that an inquiry into the capacity of reason and its
activity is outside the scope of the science of nature, which in turn raises
questions about whether the De anima in whole or in part, should be
considered part of natural science or not. In Chapter 7, principally through
a careful investigation of de An. i.1, we consider what Aristotle has to say
about the form of inquiry appropriate for the soul, and how the norms
governing that inquiry are related to those governing zoological inquiry.
In Chapters 8 and 9, I rely on two pieces of previous research to shed new
light on an old question: how did Aristotle (as opposed to later editors and
commentators) conceive of the interrelationships among his various
inquiries? Are the many cross-references simply editorial additions, do
they tell us about Aristotle’s preferred order in which treatises should be
read or studied, or does Aristotle have discernible views about the order in
which inquiries ought to be carried out – a distinctive category of norm of
inquiry? In Chapter 8, I explore this question by looking at the dependence
of a particular inquiry reported in Cael. ii.2 on conclusions reached in De
incessu animalium (IA) 2–6. In the following chapter the same question is
explored by investigating the dependence of his zoological investigations
on the generation and existence of uniform (homoeomerous) parts on his
theory of the emergence of differential powers of uniform materials pre-
sented in Meteorology (Mete.) iv. These chapters provide compelling
evidence that for Aristotle it is not just that the treatises should be studied
in a certain order, but that the inquiries on which they report have
a preferred order as well. Of course, Aristotle may only have come to
realize this after the fact – that is, while engaged in a biological investiga-
tion of uniform parts he may have decided he needed to have a better
understanding of the differences between uniform materials and embarked
on such an inquiry. But that could very well lead him to a conclusion about
the dependence of one inquiry on another, and to formulate views about
how inquiries ought to be ordered based on such dependence relations.
I’ve chosen a rather unorthodox strategy for the conclusion of this
volume. I apply the lessons learned from the previous chapters to a quite
focused and delimited scientific inquiry, one that had a very significant
impact on the history of anatomy, physiology, and biochemistry:
Aristotle’s attempt to understand why respiration is (obviously) so vitally
important to those animals that breathe. The conclusion serves to demon-
strate the value of Aristotle’s epistemology of inquiry by showing that
epistemology of inquiry in action.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6
C 1D19 12 1 8 31 2 947 73 8 4 9 7
part i
Erotetic Frameworks and Domain-Specific
Norms
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6
C 1D19 12 1 8 31 2 947 73 8 4 9 7
chapter 1
The Goal of Knowledge and Norms of Inquiry
It has been said, then, what is the nature of the knowledge being
sought, and what is the object toward which our inquiry and our
entire methodos should aim.
Aristotle, Metaph. A.2, 983a21-231
Chapter Summary. Book i of the Posterior Analytics articulates necessary

and sufficient conditions for the achievement of unqualified scientific
knowledge, among which is being in possession of first principles of
demonstration that themselves are indemonstrable. Book ii, an investiga-
tion of inquiries leading to knowledge of essences and causes, describes
stages on the way to that goal, each stage looked at from two different
perspectives: definitions expressing knowledge of essences and demonstra-
tions expressing knowledge of causes. It has proven difficult to read
this second book as a unified discussion, and that has led to a great deal
of literature focused almost exclusively on its last chapter, on the assump-
tion that this is where Aristotle provides his answer to the question of how
we achieve knowledge of those indemonstrable first principles. But the
description of that process in ii.19 is widely viewed as insufficiently robust
to explain how first principles arrived at by such a process could possess the
epistemic authority Aristotle claims for them. Some of that lack of robust-
ness can be overcome by reading chapter 19 as an abstract summation of the
lessons of the previous eighteen chapters, and over the last few decades
some fine research has helped in that respect. Nevertheless, it has not been
1
Τίς μὲν οὖν ἡ φύσις τῆς ἐπιστήμης τῆς ζητουμένης, εἴρηται, καὶ τίς ὁ σκοπὸς οὗ δεῖ τυγχάνειν τὴν
ζήτησιν καὶ τὴν ὅλην μέθοδον. The entirety of Chapter 3 will be devoted to exploring the meaning
and uses of methodos in the Aristotelian corpus, and until that exploration is completed, I will simply
transliterate the word. However, to aid the reader, I note here that the term has two distinct but
closely related uses: to refer to an investigation or inquiry in pursuit of knowledge (the apparent
meaning here), or to refer to the way of carrying out such an inquiry, a method. These two uses
mirror two distinct uses of the root term ὁδός, which can refer either to a path along which a journey
proceeds or to the journey itself.
11
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
12 The Goal of Knowledge and Norms of Inquiry
sufficiently appreciated that those first eighteen chapters of APo. ii, while
providing an invaluable general framework for scientific inquiry, have very
little to say about the process by which an inquiry moves from one stage to the
next, and especially to its last stage, when the researcher is in the process of
acquiring a domain’s first principles.
The view to be defended in the following pages is that this absence
is intentional. The substantive norms that govern specific inquiries, as
they move from everyday experience to unified, causally grounded
knowledge of a clearly defined and demarcated subject, are articulated
and defended in metalevel discussions within the inquiries themselves.
What the Posterior Analytics (and, I will argue, the Metaphysics)
provide is a highly abstract erotetic framework within which any
inquiry should be approached. Characterizing this erotetic framework
will be the task of Chapter 2. In this chapter I will review other
approaches to this topic and contrast them with a dogmatic summary
of the approach to be defended here.
1.1 Introduction
One theme that unifies the first book of Aristotle’s Posterior Analytics is
that of specifying necessary and sufficient conditions for having know-
ledge in the strongest sense, unqualified scientific knowledge (ἐπιστήμη
ἁπλῶς; APo. i.1, 71a28-29; 2, 71b9, 15), the knowledge we have of some-
thing through demonstration, characterized as a syllogism productive of
scientific knowledge (συλλογισμὸν ἐπιστημονικόν).2 Aristotle does not
deny that there are less robust cognitive states that can lay legitimate
claim to knowledge (γνῶσις), and he insists that knowledge in the robust
sense must somehow arise out of cognitive states that fail to satisfy the
conditions for unqualified scientific knowledge (71a17-30). But he pur-
posefully puts off discussion of a question that is an immediate conse-
quence of that insistence: How does one move from less robust, prior
knowledge to those principles that are needed for demonstrative,
2
APo. i.2, 71b18. The question of the extent to which the theory of demonstration in APo. is dependent
on the theory of categorical syllogistic developed in APr. has long been debated (on which see Barnes
1981, which does a fine job of reviewing the history of this debate as well as defending one specific
answer). Since συλλογισμός just means ‘reasoning’ outside the context of the theory of the ‘syllogis-
tic’ developed in APr., how one should translate this expression depends on where one stands in that
debate. Barnes 1993 translates “scientific deduction,” but in his notes (93) expresses a preference for
‘deduction productive of understanding’; Irwin and Fine (1995, 40) have opted for “deduction
expressing knowledge”; Ross (1949, 507) and Mignucci (2007, 7) go respectively for “scientific
syllogism” and “sullogismo scientifico” without comment.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
1.1 Introduction 13
scientific knowledge?3 To put it in the terms of this chapter’s epigraph: If
unqualified scientific knowledge is the target of our inquiry, what is the
method the inquiry ought to follow in order to achieve it? He alludes to
a “source of scientific knowledge” (ἀρχὴ ἐπιστήμης; i.3, 72b23-24)
through which we know the definitions that are among the basic prin-
ciples of scientific knowledge; and what that source is becomes clear later
near the end of book i (i.33, 88b36-37): “by reason (νοῦς)4 I mean a source
of scientific knowledge.” This reminder of the need for such a ‘source’ or
‘principle’ for knowledge in the unqualified sense occurs within a series of
chapters dealing with the nature and role of first principles in scientific
knowledge and the relationship between perception and scientific know-
ledge. These chapters thus lead rather naturally to the sweeping statement
that opens the second book of the Posterior Analytics:
The things about which we make inquiries (Τὰ ζητούμενα) are equal in
number to those we know scientifically. We make inquiries (ζητοῦμεν)
about four things; the fact (τὸ ὅτι), the reason why (τὸ διότι), whether it
is (εἰ ἔστι), and what it is (τί ἐστιν). (89b23-25)
A unifying thread running through APo. ii is an exploration of how each of
these four inquiries is related to the others (they are intimately, and com-
plexly, related, as it turns out). But it is also an exploration of a number of the
presuppositions of the theory presented and defended in book i. Notice how
the second book emerges as a unity from this perspective.
1. According to the theory articulated in book i, demonstrations are valid
deductions in which the middle term identifies the cause of the attribute
belonging to the subject identified in the conclusion. But it also says that
definitions are among the most important starting points of
3
Thus in the first chapter he reminds his readers that if this is not in some sense true, the puzzle raised
in the Meno will arise, that we will either learn nothing or what we already know (APo. i.1, 71a29-30).
As McKirahan (1992, 23) notes, this follows Aristotle’s assertion that we think we have ἐπιστήμη of
a fact when we know its cause as its cause and that it could not be otherwise than it is – and thus it
entails that the fact to be known must be the necessary conclusion of a deduction and that the
premises must somehow identify, and be known to identify, the cause of the fact. In context, then, it
is a more significant claim than it may at first appear.
4
Any translation of νοῦς presupposes answers to at least two interpretive questions nicely distin-
guished by Barnes 1993, 259–271 (these two are in fact related to his first and third puzzles about the
chapter stated on page 259): Does Aristotle use it to refer abstractly to a faculty for acquiring
knowledge, or to the state of being in possession of principles? And does Aristotle take what it
possesses to be propositional or conceptual in character? I take νοῦς, at least here, to refer to a faculty
that holds knowledge of principles in conceptual form, and take ‘reason’ to be a preferable translation
to other ‘nonintuitionist’ options such as ‘thought,’ ‘intellect,’ ‘mind,’ or ‘comprehension.’ A more
complicated defense of this choice will emerge as we proceed.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
demonstration. What is the relationship between inquiries leading to
scientific definitions, on the one hand, which provide us with accounts
of what the entities in our domain are, and inquiries leading to scientific
demonstrations, which provide us with causal explanations of the attri-
butes that belong per se to those entities? Understanding this relationship is
a central organizing topic of chapters 1–13 of book ii.5
2. Book i also presupposes that the explananda of a demonstrative science will
meet certain conditions that are required if there are to be causal demon-
strations of them, but only hints at how one might determine what facts
within a domain of investigation meet those conditions, or how one might
pursue inquiries within a domain to help insure that propositions for
which we are seeking explanations (which throughout the Topics and
Prior and Posterior Analytics he refers to as ‘problems’) are appropriate for
demonstration. This is a central organizing theme of chapters 14–18.6
3. Book i twice alludes to the possibility of ‘another kind of ἐπιστήμη’
besides demonstrative ἐπιστήμη. In fact, that there is such seems
required by the theory defended in book i. That theory requires that
the starting points or ‘primaries’ of a science are not themselves demon-
strated, and yet that they ground the necessity of the conclusions. How,
then, do we come to know them? There is an obvious Meno problem
here, alluded to in the first chapter. So what is the epistemological status
of our starting points – what grounds them, if not demonstration? This
appears to be the topic of the infamously difficult ii.19.
This last chapter of book ii self-consciously looks back to the beginning of
book i, where we were told that “there is not only scientific knowledge, but
also some source (ἀρχή) of scientific knowledge by which we get to know
the definitions” (i.3, 72b23-25). In ii.19 he describes a process, which begins
in perception and leads to universals coming to rest in the soul, which, he
tells us, produces a state that is “a source (ἀρχή) of art or of scientific
knowledge, of art if it deals with generation, of scientific knowledge if it
deals with being” (ii.19, 100a8-9). And indeed, the very last sentences of the
treatise (100b5-17) provide an argument (a thoroughly conditional one, it
should be noted)7 that this source “would be” reason (νοῦς), the faculty
5
For different attempts to clarify this relationship between essence-identifying definitions and causes
within the Posterior Analytics see Charles 2000, Ferejohn 2013, and most recently Bronstein 2016, esp.
Part ii.
6
Lennox 2014a provides a full defense of that claim, though the outlines of that defense can be found
in Lennox 1987, 1991, and 1994 (reprinted as chapters 1–3 of Lennox 2001b).
7
“If, then, we have no other kind besides scientific knowledge [that grasps] truth, reason (νοῦς) would
be the principle of scientific knowledge” (APo. ii.19, 100b14-15).
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
1.1 Introduction 15
that knows the principles of demonstration, and that this would be the
principle of scientific knowledge alluded to in APo. i.
In Chapter 2, I will present a positive view of what is accomplished in
APo. ii. The purpose of the rather sweeping overview just offered is to
highlight the fact that Aristotle shows an equal concern for questions
relating to the nature or conditions of unqualified scientific knowledge
and for questions relating to how we go about acquiring it (i.e., to scientific
inquiry). It is possible for us to have unqualified knowledge of
a proposition that is not a primary, for Aristotle, only on the condition
that it follows necessarily from something that is primary. Much hinges,
then, on that ‘source of scientific knowledge,’ reason (νοῦς), and on the
principles that it has in its possession, and on Aristotle’s views about how
the conceptual starting points we possess are acquired. Perhaps the central
epistemological challenge facing Aristotle is whether his account of the
acquisition of the sources of scientific knowledge, which he consistently
claims is by induction (ἐπαγωγή),8 ensures that they will have the creden-
tials they need to ground the demonstrations that deliver unqualified
scientific knowledge.
The position I will be defending in the chapters to follow is that Aristotle
has a robust response to this challenge, but the robustness of his answer is
not apparent from the Posterior Analytics. That work, as is generally
recognized, operates on a plane of high abstraction, presenting a view
about the nature of theoretical knowledge and of the forms of inquiry
that lead to it, that is, as far as possible, completely general. And yet
Aristotle consistently defends the view that each investigation operates
with principles that are for the most part distinctive to it. To quote the
best known of many passages that make this point:
The majority of principles for each science are distinctive (idiai) to it.
Consequently, it is for experience to provide the principles concerning
each subject. I mean, for instance, that it is for astronomical experience to
provide the principles for the science of astronomy (for when the appear-
ances had been sufficiently grasped, in this way astronomical demonstra-
tions were discovered; and it is also similar concerning any other art or
science whatsoever). (APr. i.30, 46a17-22)
8
“It is quite clear that it is necessary for us to come to know the primaries (τὰ πρῶτα) by induction
(ἐπαγωγῇ); for it is in this way that perception implants the universal” (APo. ii.19, 100b3-5). “For we
are persuaded in all cases either through deduction (διὰ συλλογισμοῦ) or from induction (ἐξ
ἐπαγωγῆς)” (APr. ii.23, 68b13-14).
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
This passage is not, primarily, a general endorsement of empiricism about
first principles (though I think it implies that).9 Its primary message,
I believe, is that most of the causal principles and definitions that inquiry
seeks are specific to distinct areas of knowledge and are arrived at by having
sufficient experience with the appearances of that specific domain. One of
the tasks of the following chapters is to make the case that the norms and
standards governing the search for those principles are, in an important
sense, also domain-specific.10 We will need to explore what can be accom-
plished at the level of abstraction on which the Analytics operates; but the
primary argument here is that, whatever can be accomplished at that level,
it will be insufficient to provide guidance for the sorts of inquiry that make
up the body of Aristotle’s science of nature. For the substantive standards
and norms that Aristotle argues are needed to insure successful scientific
inquiry, we must turn to the methodologically normative passages in his
scientific works, especially, though not limited to, the introductions to
these works.
1.2 Inquiry and Grounding for Principles:

My View and Four Others
I begin by presenting a very brief and dogmatic sketch of what I will argue
in the remainder of this book is Aristotle’s theory of scientific inquiry. I will
then provide similar sketches of four quite distinct positions that have been
taken on this question of the relationship between Aristotle’s views about
scientific inquiry and his views about the warrant or grounding for the
principles that are a precondition of unqualified scientific knowledge in
recent discussions of the subject. After these brief sketches, the remainder
of the chapter will be devoted to a critical examination of each of them. We
then turn, in Chapter 2, to the full presentation and defense of my
alternative.
9
By ‘empiricism’ here I do not mean adherence to the philosophical views of the ‘empirical school’ in
Hellenistic medicine and philosophy, let alone to those of the British Empiricists. I have in mind
only the idea that perceptually grounded experience is the sole source of our knowledge of the world.
10
There are certain parallels to this idea in John Norton’s forthcoming The Material Theory of
Induction, a preprint of which is available at www.pitt.edu/~jdnorton/papers/material_theory/mat
erial.html. The norms of inquiry I discuss are not formal or algorithmic, and thus ‘material’ in
Norton’s sense; but in the view I am defending, Aristotle has both general, epistemic norms of
inquiry shaped by his understanding of the goal of scientific inquiry, demonstrative knowledge; as
well as domain-specific norms, which reflect differences in the objects of investigation and our access
to those objects.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
1.2 Inquiry and Grounding for Principles 17
1. Induction as norm-governed inquiry.11 As the epigraph for this chapter
indicates, Aristotle sees inquiry as a goal-directed process; its goal is
knowledge. Aristotle recommends that one begin an inquiry with
a proper conception of its goal, and he thinks most of his predecessors
either had nοt thought clearly about that goal (i.e., about what knowledge
is); or (like Plato) had thought deeply about what knowledge is, but came
to the wrong conclusion. Beginning with a proper conception of the goal
provides one with a standard for articulating appropriate norms that can
serve as guides for the process of inquiry targeted on that goal.
We do nοt, however, aim for knowledge in the abstract, but for scientific
knowledge of certain, delimited domains. The majority of the norms that
govern inquiry are, for that reason, domain-specific. These norms provide
inquirers with regular feedback on whether they are on course toward their
goals or not, and thus at each stage of inquiry grounds or justification are
provided for knowing that that stage has been achieved. Critical to this
perspective is that this is not a process that begins with belief that at some
step is magically converted into knowledge. It is a process of moving from
perceptual knowledge of particulars through general knowledge of facts to
causal explanations grounded in having determined which feature or
features of the objects of inquiry are essential to its being what it is – that
is, to scientific knowledge.12 What these substantive, domain-specific
norms of inquiry provide are criteria for the achievement of various stages
along the way and for recognizing when you are off course.
The opening lines of APo. ii.1 claim that there are four things we know
scientifically, corresponding to the four objects of inquiry specified there;
as we will discuss in more detail in Chapter 2, the inquiries consist of two
parallel tracks – one from seeking to establish a fact to seeking the reason
why or cause of that fact, the other from seeking to establish that some
entity exists to seeking knowledge of what it is. APo. ii.2 complicates that
analysis considerably; nevertheless, the message of APo ii.1 is that you can
know (and the verb is ἐπίστασθαι) that the moon is in eclipse, or that
elephants exist, prior to going on to inquire about the causes of eclipses or
the nature of elephants. And as we get deeper into APo. ii, it appears that
11
A view distinct from those I review here, and with certain affinities to my own, is defended by Marc
Gasser-Wingate. Unfortunately, it only came to my attention after my manuscript had been
submitted to the press. See Gasser-Wingate 2016, 1–20, and Gasser-Wingate (2021).
12
I am using ‘knowledge’ here to refer to all those states captured by the concept of γνῶσις; for
perception as a form of γνῶσις, see Metaph. Α.1, 981b10-13, GA i.33, 731a30-33. For a defense of the
idea that γνῶσις is the generic term that covers perceptual knowledge of particulars as well as reason
(νοῦς) and scientific knowledge (ἐπιστήμη), including textual evidence for it, see Salmieri 2014 (in
Lennox ed. 2014b). I thank an anonymous press reader for requesting clarification on this point.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
each of these four inquiries consists of identifiable stages – one might
know that the moon is suffering eclipse because something is blocking its
source of light without yet knowing what is doing the blocking or the
precise relative motions that are involved in the process; you might have
identified a number of distinctive, differentiating features of the elephant
without having decided which of those features is most basic, or perhaps
what its function is. (The trunk, after all, is puzzling – answering the
“What is it for?” question is nontrivial, since it appears to be for many
things.)13
A major part of this book will be devoted to identifying norms
governing the process of inquiry that are specific to each domain.14
These domain-specific norms, I will argue, are entirely complementary
to the erotetic framework provided by the Posterior Analytics. For
example, that framework tells us that ultimately we are seeking answers
to four intimately related questions: “Is there an S (or P)?,” “What is S (or
P)?,” “Does P belong to all S?,” and “Why does P belong to all S?”
Knowing, from the start, that one’s goal is achieving answers to these
questions is important. But if S is (say) Soul, your inquiry will be
governed by very different norms than if S is Plane Figure, The Good
Life, or Animal Locomotion – even though in all four cases you will be
seeking answers to all four framework questions that are intertwined in
analogous ways.
That, in brief, dogmatic outline, is the view that will be defended in the
chapters that follow. But before that defense begins, I will outline four
alternatives and then discuss them in some detail.
2. Coherentism. According to this view, the process or processes Aristotle
describes as leading from perception to a grasp of universals and,
finally, of indemonstrable principles of knowledge is merely
a ‘psychological’ description of the process of discovery (Irwin 1988),
and therefore does not provide epistemic grounding or justification for
13
That example is chosen for two reasons: (i) one can compare the many factual claims about elephants
in HA with the very long explanation (focused on the trunk, but covering pretty much everything
Aristotle knew) of many differentiating features of the elephant in PA ii.16; and (ii) there is a rich
paper by Allan Gotthelf (Gotthelf 1997b, reprinted in Gotthelf 2012, ch. 8) that unravels the
complicated explanatory threads in PA ii.16 and identifies which features are causally/explanatorily
fundamental.
14
I am using ‘domain’ rather than ‘science’ because, as we will see, there are distinct norms governing
inquiries within what Aristotle thinks of as a single science: the central theme of Chapter 4, in fact, is
that Aristotle thinks of the inquiries presented in his Physics, Meteorology, De caelo, De partibus
animalium, De generatione animalium, and a number of other treatises as contributions to a single
science of nature.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
1.2 Inquiry and Grounding for Principles 19
the knowledge achieved. In Aristotle’s First Principles, Terence Irwin
argued that Aristotle presents, in the end, an incoherent defense of first
principles in APo., and thus “we ought to expect him to reconsider his
views on dialectic, and with them his views on coherence as a source of
justification and knowledge” (Irwin 1988, 150).
3. Rationalism. In “Aristotle’s Rationalism,” Michael Frede argued that
“[w]hat makes it [the grasp of a first principle] an insight is not the
support it gets from observations or considerations, but that one
finally sees in a way which fits how the features in question are related
to each other and to other relevant features” (Frede and Striker 1996,
172). It would seem this too is a sort of ‘coherentism’ – one grasps
a universal as a first principle once one sees how it fits within a network
of relationships among ‘features.’ Frede sees this as providing the basis
for a sort of ‘insight,’ and Frede rightly concludes his essay by noting
that on his account Aristotle ends up not differing significantly from
Plato (at least on one common understanding of Plato’s epistemol-
ogy): perception and experience prepare you in some psychological
way for a grasp of first principles, but they are not what certifies those
principles as the principles of knowledge.15
4. Justification through Explanation. Aryeh Kosman (1973), Myles
Burnyeat (1981), David Charles (2000), and David Bronstein (2016),
while again providing quite different views about the Posterior
Analytics in many respects, share in common an approach to the
question of how inquiry leads to knowledge of first principles. All of
them reject the ‘causal/epistemic’ and ‘discovery/justification’ distinc-
tions – deployed in different ways by Irwin and Frede – as useful tools
for understanding the project of APo.,16 yet they too accept the claim
that induction (at least as it is ordinarily understood) is unable to
certify the status of a first principle as a first principle. On their view,
that certification comes as part of the process of using the principles
successfully in scientific explanation. In that respect, this is what
15
In this respect both Irwin and Frede appear to assume a view about the impossibility of
grounding knowledge in perception like that articulated by Donald Davidson): “The relation
between a sensation and a belief cannot be logical, since sensations are not beliefs or other
propositional attitudes. What then is the relation? The answer is, I think, obvious: the relation
is causal. Sensations cause some beliefs and in this sense are the basis or ground of those
beliefs. But a causal explanation of a belief does not show how or why the belief is justified”
(Davidson 2009, 127).
16
See Burnyeat 1981, 131–132 and note 59. In that note, Burnyeat describes Irwin as providing “a very
clear statement of the interpretation of the An. Post. programme in terms of knowledge and
justification which I am opposing.”
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
would today be referred to as an ‘inference to the best explanation’
justification of scientific principles. One can argue, of course, that this
is a form of inductive reasoning, as at least some of the scholars
mentioned do.17 But even if one accepts that claim, it is questionable
whether Aristotle uses the term epagôgê for any such process.
5. Finally, Robert Bolton has consistently defended Aristotle as an empiri-
cist and inductivist, while also providing a robust but limited place for
certain of the dialectical practices that Aristotle claims will have a role
to play in grounding the first principles of science. As he says, “This
shows that the account in [APo.] ii.19 is not, as some have supposed,
merely a genetic account of the psychological preconditions for the
generation of knowledge of first principles. It is an account of the
inferential process by which such knowledge is reached from previous
knowledge” (Bolton 1991, 5). Bolton has not, however, provided
a convincing account of how Aristotle conceives of the inductive
process, such that it reliably guides us toward principles of the kind
that can serve as the foundations for scientific knowledge.18
Taking a closer look at each of these approaches will provide us with
a clearer picture of what options are currently available for interpreting
Aristotle’s views about inquiry in search of scientific principles and what
their strengths and limitations are.
1.3 From Intuitionism to Strong Dialectic

Terry Irwin, in Aristotle’s First Principles, has provided a classic critique of
those who attempt to portray Aristotle as arguing that the first principles
that serve as the foundations of demonstrative knowledge are grounded in
an inductive process that begins with perception.
Here is the essence of the argument. At its heart is a popular current view
about epistemic justification:
The knower must grasp self-evident principles as such; for if they are grasped
non-inferentially, without any further justification, they must be grasped as
17
That is, it is inductive insofar as it is an inference from the explananda of a science to the best
explanans. But contemporary philosophers who defend this viewpoint see induction as
a confirmation procedure rather than as a discovery procedure, which is not how Aristotle understands
induction.
18
Recently, in as yet unpublished work, Bolton appears to be moving toward those who see Aristotle
adopting a form of justification of first principles akin to inference to the best explanation (Bolton,
in Salmieri forthcoming).
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
1.3 From Intuitionism to Strong Dialectic 21
true and necessary when considered in themselves, with no reference to
anything else. (Irwin, 1988, 134)
This remark contains two undefended assumptions: the noetic grasp of
principles must be noninferential; and a noninferential grasp can have no
further justification, which I take to mean that the intuitive grasp of the
principle must be ‘self-certifying.’ Irwin feels these assumptions are justi-
fied, given the positions Aristotle rejects in APo. i.3. There, Aristotle argues
that attempts to provide any sort of deductive justification of first principles
must involve either reasoning in a circle (72b16-18; 72b25-73a6) or an
infinite regress (72b7-16). In either case, any allegedly first principle will
not be first at all, and our trust in it will be only as good as our trust in those
premises upon which it rests.
But what, one might ask, of induction? Aristotle treats it as a form of
reasoning that takes one from perception of particulars to knowledge of
universals, so why can it not provide the inferential justification that Irwin
is demanding? The answer, as we will see and as the above quotation
presupposes, is that, whatever role induction may play in Aristotle’s
views about science, it cannot play the role of justification for first prin-
ciples. Here is an explicit statement of Irwin’s position.
Experience and familiarity with appearances are useful to us as a way of
approaching the first principles; they may be psychologically indispensable
as ways to form the right intuitions. But they form no part of the justifica-
tion of first principles. . . . The acquisition of nous is not meant to be
magical . . . . Nor, however, is it simply a summary of the inquiry, or
a conclusion that depends on the inquiry for its warrant. (Irwin 1988, 136)
I take it that Irwin thinks Aristotle has painted himself into a corner, from
which the only escape is a kind of Platonism. Aristotle himself appears to
rule out deductive justification for principles in APo. i.3; Irwin now argues
that induction, if that is Aristotle’s word for acquiring universals based on
inquiry grounded in experience, while it may have heuristic value, also plays
no role in justifying first principles – as he puts it here, the grasp of first
principles does not depend on the inquiry for its warrant. That is, Irwin has
quietly slipped into the exegesis of Aristotle the twentieth-century ‘context
of discovery/context of justification’ dichotomy, along with its philosoph-
ical devaluation of the first, and its view that all justification must come
about deductively.19 Thus it appears that Irwin has now ruled out the only
19
This is occasionally explicit: “Aristotle . . . argues that dialectic and empirical inquiry are simply
means of discovery, not essential for justification” (141). He cites no text where Aristotle says this,
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
two possibilities open to Aristotle; for more than once Aristotle asserts, as
in Prior Analytics ii.23, that “all conviction is produced either by deduction
or induction” (68b13-14; cf. EN vi.3, 1139b26-29; APo. i.18, 81b6-9).
Nevertheless, Irwin feels justified in supposing that the only view
Aristotle can legitimately adopt is some form of intuitionism.
For these reasons . . . it is reasonable to ascribe to Aristotle a doctrine of
intuitive nous that grasps first principles. If we refuse to ascribe this to him,
we leave him with unanswered questions; his foundationalist views about
justification and his objections to Plato will create demands that he does not
try to satisfy. (Irwin 1988, 141; cf. 140)
In the conclusion to chapter 7 of Aristotle’s First Principles, Irwin provides
a summative argument of the following form.
1. The Analytics and Topics present views about the justification of scien-
tific principles that are inconsistent: scientific principles must be
noninferentially known to be true and primary; but they must be
grounded either in induction or dialectic.
2. “In describing what we ought to expect [Aristotle to do to resolve the
inconsistency] we have described the programme of the Metaphysics.”
(Irwin 1988, 150)20
A governing assumption of Irwin’s argument appears to be that the first
principles of a science must either come with their status as first principles
intuitively clear (or self-evident) or must acquire that status through some
sort of proof. Aristotle never states the former view, and has clear and
powerful arguments against the latter. The one view he does state clearly, as
we’ve seen, is the view that we must come to know the first principles
through induction, beginning with perceptual awareness. Why would
Irwin not take Aristotle at his word on that?
I believe he has good reasons. The final paragraph of APo. ii.19 goes
beyond what may well be a quasi-psychological description of the percep-
tual grounding of universals in the soul earlier in the chapter, to an
argument that the faculty that knows first principles must have a grasp of
them that is superior to the grasp we have on anything that follows from
and footnotes Sidgwick’s Methods of Ethics. An excellent way into the historical and philosophical
background to this distinction is Schickore and Steinle (2006).
20
Another contemporary philosophical assumption plays a role in Irwin’s argument, the assumption
that ‘perception is theory-laden’: “Aristotle should have seen that this second objection to dialectic
[that there is no assurance of the objective correctness of appearances] works equally well against
empirical inquiry . . .” [137] “. . . alleged non-inferential justifications are either inferential or not
justifications” [142].
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
1.3 From Intuitionism to Strong Dialectic 23
them. But the description of the process leading to first principles that we
are given just prior to this is not nearly rich enough to give us any reason to
think the state we will be in at the end of this process will have the requisite
superiority. For there is nothing in that description that provides for even
the most rudimentary inductive controls and tests that humans (including,
as we will see, Aristotle) typically use in guarding against erroneous
inductive inferences. As Myles Burnyeat put it in a classic discussion of
the theory of knowledge of the Posterior Analytics, “[h]is treatment of this
process in B 19 is by our standards perfunctory in the extreme” (1981, 133).
Nor has anything in the preceding eighteen chapters on the stages of
inquiry and the relationship between definition and demonstration at
each stage provided that (though I think Aristotle’s awareness of the nature
of inductive inference is implicit in a number of his examples).
Now Irwin has one other fact about Aristotle’s practice to rely on:
in the Analytics Aristotle sometimes asserts as a norm that it is not for
the practitioner of a science to defend his first principles against those
who deny them, and this is a norm we occasionally see invoked in
practice.21 So how can those principles be justified by the inductive
inquiries of the practicing scientist? Irwin wants to argue that first
philosophy steps in at this point: “[the arguments of first philosophy]
are prior to empirical inquiry, in so far as they defend the assump-
tions taken for granted in empirical inquiry” (Irwin 1988, 276). But in
certain cases, most notably in his dealings with the Eleatic denial of
natural change in Ph. i.2, Irwin sees Aristotle relying on nothing more
than simple dialectical arguments based on commonly held beliefs.
There is no argument against the Eleatics that does not rely on them
accepting common beliefs that contradict their static monism (Irwin
1988, 66–70).
I think the point of Aristotle’s theory and practice in these cases
can be understood differently, however. The scope of the application
of the above norm is restricted to situations in which a thinker either
mounts an attack on a science’s principles, or at any rate attempts to
practice the science while ignoring its principles or importing alien ones.
In these circumstances, the practitioner of the science will be unable
to argue with such a thinker from the principles of that science, for the
obvious reason that it is precisely those principles that the opponent
has already tacitly or explicitly rejected. If you are to argue with such
a person, it must be in some other manner, perhaps by means of
21
For example, Ph. i.2, 184b25-185a20; cf. APo. i.12, 77a40-b15.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
arguments analogous to what we see in Metaph. Γ.4–6 or in Ph. i.2.22
Aristotle is not saying that it is for a more general philosophical
discipline to justify the principles of geometry or physics. Those
principles are grounded, I shall argue, in the inductive investigations
of practitioners of particular sciences, once we properly understand
what that entails. But to argue with someone who denies your funda-
mental starting points requires a different kind of argument.
1.4 Frede’s Rationalist Aristotle

As I noted earlier, there are certain background assumptions that Michael
Frede’s paper “Aristotle’s Rationalism” shares with the view presented in
chapters 6 and 7 of Irwin 1988. Frede happily acknowledges the many
passages, such as the following, in which Aristotle lays stress on the
importance of experience to the proper investigation of nature:
Lack of experience (ἡ ἀπειρία) is a cause of the relative inability to
comprehend the admitted facts. Wherefore those who are more at home
among natural things (ἐνῳκήασι μᾶλλον ἐν τοῖς φυσικοῖς) are better able to
postulate principles of the sort that can connect many things together;
while those who, from engaging in many arguments, have failed to study
things as they are, readily show themselves capable of seeing very little.
(GC i.2, 316a5-10)
This is not merely a description of how one acquires principles “that can
connect many things together” – it is a passage shot through with norma-
tive language. Lack of experience accounts for a failure of comprehension;
the more “at home among natural things” you are the better able you will be
to identify appropriate principles. It sounds as if Aristotle has some view
that long experience of the subject matter for which you seek understand-
ing is necessary to guide you to appropriate first principles.
But necessary in what sense? Frede, who was perfectly familiar with texts
such as this one, denies that first principles are grounded in, or warranted
by, such experience, even while acknowledging that Aristotle thought it
was necessary. Here is his response:
[T]o the extent that this [acquisition of concepts] is a natural process based
on perception, the relation between our perceptions and our knowledge of
first principles . . . is a natural, a causal, rather than an epistemic relation. And
22
A point argued in detail in Bolton 1991, 13–22.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
1.4 Frede’s Rationalist Aristotle 25
this is how Aristotle can be an extreme rationalist and still constantly insist on
the fundamental importance of perception for knowledge. (Frede 1996, 172)
This portrait of Aristotle as extreme rationalist is based primarily on
a reading of APo. ii.19 and Metaph. A.1. What leads him to this ‘anti-
empiricist’ reading of Aristotle’s views about grounding first principles?
First of all, it is clear that Frede shares with Irwin two widespread assump-
tions of modern epistemology:
1. Justification is not part of the process by which one forms concepts or
beliefs. That process may be of interest to psychology or biography,
but epistemology is concerned with the warrant or justification for our
beliefs and with what sort of warrant bestows the honorific ‘know-
ledge’ on some of them.
2. Norms and standards are denizens of ‘the space of reasons’, not of the
natural, law-governed world.23
From the standpoint of a person wedded to these assumptions, any
proposal that epistemic norms are built into the inductive, ‘principle
discovering’ process of scientific inquiry is ill conceived from the start.
Frede refuses to attribute such a view to Aristotle, reading passages that
sound like such a view as characterizing a “natural, a causal, rather than an
epistemic relation.”
The problem is not merely one of reading the Posterior Analytics
through these assumptions of twentieth-century analytic epistemology,
however. Frede is also genuinely puzzled by the position Aristotle aims to
defend.
[W]e would like to know why Aristotle insists that all our knowledge has its
origins and its basis in perception, if he assumes that all that can be known,
strictly speaking, is known by reason. (Frede 1996, 158)
The phrase “strictly speaking” in that sentence is very important. As Frede
carefully details, the message of Metaph. A.1 is that there is a strong sense of
‘knowing’ in Aristotle that goes beyond experience. ‘True’ or ‘real’ know-
ledge is both of universals and of causes; mere experience acquaints us with
neither: “to know, properly speaking, is to have insight, to understand, to
know why, rather than to merely know that” (Frede 1996, 159). Experience
by itself cannot deliver the goods – at least, not these goods. For Frede
23
There is evidence that Frede thought this distinction was an innovation of Plato’s, and that Aristotle
follows him in this innovation (Frede 1987b, 3–8). For a classic statement of the problem, see
McDowell 1994.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
acknowledges that, as far as the account of Metaph. A.1 is concerned, for the
sort of common-sense reasoning that we use in our day-to-day decision-
making, mere experience is sufficient (Frede 1996, 162–163). But the sort
of grasp that reason has of causes and essences is of an entirely different
order:
[W]hat makes a notion an adequate notion is that it is appropriately related
to other concepts, and thus imports a whole system of related concepts in
terms of which one can understand and explain the objects falling under the
given concept. (Ibid., 164)
It is giving an account of this sort of understanding that Frede takes to be
the central task of APo. ii.19. That chapter argues, on Frede’s reading,
that we acquire a disposition, reason, to know first principles, and that
our particularly human abilities of perceptual discrimination and mem-
ory “in the course of ordinary development, give rise to concepts and
ultimately to first principles” (Frede 1996, 170). But he takes two features
of the account to tell in favor of the view that Aristotle is, nevertheless,
a rationalist: first, the metaphor of the battle, and second, the thorough-
going naturalism of Aristotle’s account of concept acquisition. The
message of the battle metaphor, on Frede’s telling, is that arriving at
first principles involves “a complex process in the course of which our
notions again and again are readjusted until they finally fit into a coherent
and appropriately structured system of notions . . . in terms of which we
finally can make sense of what we know from experience” (171). It is only
in the thoroughly rational construction (and reconstruction) of this
coherently structured conceptual system that certain of our universals
are grasped as first principles.
What is unquestionably sound in Frede’s picture of Aristotle’s epistem-
ology is that experience (ἐμπειρία) is a far cry from the universal grasp of
causal natures that is fundamental to scientific knowledge. And it is also
true that Aristotle discusses the advent of that knowledge in thoroughly
naturalistic terms. What I find wanting in Frede’s discussion, and what is
sorely needed, is an understanding of APo. ii.19 informed not only by
Metaph. A.1 but by the earlier chapters of APo. ii. Frede notes that,
according to Metaph. A.1, experience gives us, at best, knowledge that,
while to know scientific principles is to have knowledge of what things are,
knowledge that grounds understanding of why things have the other
features they do, including why they behave and change as they do. But
APo. ii.1, as I noted earlier, opens with a declaration about four kinds of
inquiry, corresponding to four kinds of knowledge. Two of these – ‘if it is’
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
1.5 Justification as Inference to the Best Explanation 27
and ‘that it is’ inquiries – are of the sort Frede supposes are adequately
grounded in experience; the other two – ‘what it is’ and ‘reason why’
inquiries – are of the sort that, when successful, give us knowledge of the
essences and causes of the things discovered by means of the first two
inquiries. One plausible way of seeing the unity of the rest of APo. ii is as
a discussion of the processes by which an inquiry passes from the first,
weaker form of knowledge, much closer to the way ‘experience’ is charac-
terized in Metaph. A.1,24 to knowledge of essences and causes, that is,
knowledge of items among which will be found scientific principles. If
that discussion provides us with at least the framework for norms that
govern such inquiries, inquiries that begin with perceptual discrimination
and end with causal understanding, then Aristotle in APo. ii gives us
precisely what Frede denies that he does – a richly normative account of
inductive inquiry that provides epistemic warrant for its results. Frede’s
strongly ‘coherentist’ Aristotle, in which scientific principles gain their
epistemic warrant only by being seen as part of a web of other concepts,
lacks textual support and is worth considering only if there is no such
normative theory of inquiry to be found.
1.5 Justification as Inference to the Best Explanation

As I noted, Michael Frede’s reading of APo. ii.19 rests on certain assump-
tions he shares with Terence Irwin. The ‘coherentism’ he ends up attribut-
ing to Aristotle in turn shares certain assumptions with what I will
characterize very broadly as the ‘Inference to the Best Explanation’ (IBE)
reading, a tradition of understanding how Aristotle thinks first principles
come to be identified as such that goes back at least to a classic paper of
Aryeh Kosman’s published in 1973. In that paper, Kosman acknowledges
Aristotle’s consistent claim that induction (epagôgê) is the process by which
we come to know the principles of a science, but asks whether it is also the
process by which we come to recognize their status as scientific principles
(Kosman 1973, 386–389).
Crucially for the view I will defend, he also directs our attention to
a number of texts that suggest that different subject matters will require
different inductive methods.25 Kosman takes texts that are often cited to
underscore Aristotle’s empiricist credentials (Cael. iii.7, 306a5-17 and
24
Although there are significant tensions in Aristotle’s account of the cognitive content referred to as
‘experience’ (ἐμπειρία), even within Metaph. A.1. For a recent attempt to resolve those tensions, see
Hasper and Yurdin 2014.
25
NE i.7, 1098b2; vii.8, 1151a15; Top. i.3, 101a36, viii.3, 158b4.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
GC i.2, 316a6-12, both critical of Plato’s approach to the study of nature)
to be making a more subtle point: that ‘induction’ (epagôgê) is not just
grasping things that are in fact principles, but grasping them qua
principles, and how the particular phenomena being studied guide us
to that grasp. Up to this point, the view being articulated by Kosman is
very much in the spirit of the interpretation to be defended in the
present volume.
However, Kosman appears to reach this result by widening the scope of
the term ‘induction’ to such an extent that it becomes questionable
whether it has much to do with the Aristotelian notion of epagôgê.
[T]he process by which we explain is the process by which we acquire and
grow to understand principles; it is by employing them in the act of
explanation that we come to see their truth, recognize their explanatory
power, and thus understand them qua principles. Νοῦς therefore is related
to an activity of the intellect in roughly the same way as ἐπιστήμη is related
to ἀπόδειξις. This activity is ἐπαγωγή; but it is not an activity radically
independent of ἀπόδειξις. (Kosman 1973, 389)
He takes Aristotle’s description of epagôgê at 71a8 as “showing the universal
through the particular being clear” as sanctioning the claim that it is “an act
of insight” (389). He lends credibility to this reading by translating
δεικνύντες τὸ καθόλου διὰ τοῦ δῆλον εἶναι τὸ καθ’ ἕκαστον (APo. i.1,
71a8) not as I just did, but as “revealing the general through the particular’s
being clearly understood.” But δείκνυμι in Aristotle is not a verb of
revelation (i.e., of uncovering something hidden), but of pointing, show-
ing, and (sometimes) proving; and δῆλον doesn’t imply ‘understanding,’ at
least not in the sense of that term being discussed in Kosman’s paper.
Kosman’s translation thus minimizes the difference between induction and
demonstration, because on the view he wishes to defend, when Aristotle
claims that induction is the path to first principles, that path has to include
the explanatory use of causal universals.26 It is only when they have proven
26
There is a sense in which a weak form of demonstration (which I dubbed A-explanations in Lennox
1987) does play a role in induction, as I have argued before and will rely on here. One step, discussed
extensively in APo. ii.14–17, toward identifying ‘convertible’ or ‘coextensive’ universals is to search
for middle terms that connect nonconvertible universals: for example, by seeking a ‘reason why’ all
pigeons, sparrows, and ducks are feathered, you might notice that they are all forms of bird and
explain that they have feathers because all birds do. This aids you in identifying the convertible
relationship between feathers and birds – and, crucially, any other features that are convertible with
birds and thus with feathers (cf. Lennox 1991 and 1994 [reprinted in Lennox 2001b, chs. 1 and 3];
Lennox 2014a). But this is clearly not what Kosman has in mind, since he is discussing the
explanatory use of what are in fact first principles.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
themselves in the explanatory arena that a universal achieved by induction
is grasped qua first principle. As he puts it in the very next sentences:
Ἐπαγωγή and ἀπόδειξις are thus related as teaching and learning, or seeing
and being seen. They are one and the same thing, but their εἶναι is different;
that is, they are one and the same thing under different descriptions. For it is
the same ἐνέργεια, scientific activity, in which we do ἀπόδειξις and
ἐπαγωγή, by which we explain phenomena and come to acquire the
understanding of principles necessary to explanation. (Kosman 1973, 389)
Given that induction is the road to the principles and demonstration is the
road from the principles, it is all too easy to fall prey to the metaphors here,
but temptation must be avoided. First, epagôgê is often contrasted, not with
apodeixis but with sullogismos, which is why it can be used as often in the
context of dialectic and rhetoric as it is in properly scientific contexts.
Second, unlike the road that leads from Athens to Thebes and back, the
way to scientific first principles is not at all the same as causal, deductive
demonstrations from them – or so I shall be arguing. In particular, I shall be
making the case that the norms that govern the inductive procedures that
justify our confidence in our causal accounts of the objects of investigation
are very different from the norms we use to evaluate a scientific
demonstration.
Like Kosman’s 1973 paper, Burnyeat’s “Aristotle on Understanding
Knowledge” stands strongly opposed to a rationalist or coherentist reading
of the Posterior Analytics and insists that we take seriously Aristotle’s
consistent claim that it is by means of induction that we come to know
first principles (Burnyeat 1981, 131–133). There are, moreover, strong hints
of an IBE stance in that paper, though this is far less clear than in Kosman
(for the hints, see 129 and 128, n. 53).27 But there is a tension between these
hints and Burnyeat’s insistence that concern about justifying our first
principles as first principles is simply not something about which
Aristotle is concerned.
It is remarkable how little interested Aristotle is in the central concepts of
that enterprise [epistemology] as it is carried on today. Concepts like evi-
dence and justification, the Humean problem of induction—all this belongs
in Aristotle’s terms to the process by which we make something γνώριμον to
us. His treatment of this process in B 19 and its companion, the first chapter
27
And the contrast in the following remark is also suggestive: “Hence, as he sees the problem of our
grasp of first principles, the difficulty is not a lack of evidence to transform inductive belief into
certain knowledge. That inductive belief is already knowledge (γνῶσις). What it is not yet is
understanding” (1981, 131).
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
of the Metaphysics, is by our standards perfunctory in the extreme. (Burnyeat
1981, 133)
About APo. ii.19, I wholeheartedly agree. But Burnyeat goes on to assert
that it is mistaken to try to get more robust answers to ‘our’ epistemological
questions out of APo. as a whole. Here I am willing to go halfway. I think
much has been done since 1981 to outline a quite robust theory of inquiry in
APo. ii. There is a normative epistemology of inquiry there: which ques-
tions need to be asked as you attempt to develop knowledge of a domain,
and in which order? How are ‘what is it?’ inquiries, aimed at essence-
identifying definitions, related to ‘reason why?’ inquiries, aimed at dis-
covering causes that can serve as middle terms in demonstrations? Does
division play any important role in either process? Is there a proper
procedure to be used in formulating scientific ‘problems’? In the end, as
I will argue in Chapter 2, I don’t think this provides more than a framework
for particular scientific investigations. But it establishes clearly that
Aristotle was concerned about the proper way to proceed from knowledge
‘if’ and ‘that’ to knowledge of ‘what’ and ‘on account of what’ and
concerned about how these inquiries were related, and not merely describ-
ing the way humans ordinarily go about inquiring.
In a similar vein, David Charles, in Aristotle on Meaning and Essence,
rejects the ‘intuitionist’ or ‘Platonist’ understanding of APo. ii.19. As with
Kosman, Charles draws a distinction between the process by which we come
to know universals from that by which we know a universal as a principle
(Charles 2000, 155–161, 266–272). The process characterized in ii.19, 100a4-
b4, he insists, is a description of the former, not the latter, process. The latter
requires that nous be imbedded in “an explanation-involving procedure of
induction” (Charles 2000, 271). His fullest statement of the position is this:
[N]ous can operate in a non-discursive manner without undermining induc-
tion’s claim to be a process essentially involving discursive thought and infer-
ence. Nous can see certain objects in a given way provided that its operation
arises out of and its results are constrained by a preceding inductive process. It
sees what it sees because it is grounded in the preceding inductive stages.
(Ibid., 271)
But how exactly does the “preceding inductive process” constrain the
operation of nous, and how exactly does its operation arise out of this
process (or these processes)? This is what the present investigation aims to
describe in some detail.
Charles also appears to see that inductive process as insufficient for
grasping principles as principles, in a manner reminiscent of Kosman.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
Nous, by contrast, may enable us to grasp those things that are the starting
points of explanation, but we will not grasp them as starting points. We do the
latter as the result of our study of the relevant pattern of demonstration.
(Charles 2000, 268; emphasis added)
And similarly, a few pages later, we are told that “[a]lthough induction does
not itself directly involve demonstration, it is permeated by our explana-
tory concerns” (ibid., 271). That could mean that inquiry is all along
motivated by explanatory questions, questions about what things are and
why they have the properties they do and change as they do – and if so, this
claim is in the spirit of the position that will be defended in the chapters
that follow. In fact, as we will see in the following chapters, the inductive
process will often involve formulating preliminary explanations – perhaps
without realizing their preliminary nature. We may, having compared
many varieties of trees with the question in mind of whether they shed
their leaves seasonally or not, conclude that only those with broad, flat
leaves shed their leaves. At that stage of investigation, we might, in response
to a question about why grape vines shed their leaves, answer, “Because
grape vines are broad-leafed.” We could give that answer while being in
two different epistemic states: we might be fully aware that all we’ve
actually done is advance inductively to the right level of universality to
search for the cause; or we might mistakenly think that it is the property of
being broad and flat that causes the leaves to fall.28
1.5.1 Interlude
Before considering the next and final option,29 I would like to pause here
and draw attention to something common to many of the positions looked
at so far. Prior to Charles 2000, there was a tendency to elide the distinc-
tion between Aristotle’s views about the ascent from perception of particu-
lars to knowledge of universals and his views about how we grasp first
principles; and with that elision, a tendency to focus an inordinate amount
of attention on the very last chapter of APo. ii. Though there was literature
28
I say “mistakenly” because a scientifically minded individual would already have good reason to
reject that answer, given that leaves have those properties during the seasons when they are not being
shed. “Seasonally” becomes important, for the question is not why such trees shed their leaves, but
why they do so at certain times and not others. And that leads a curious investigator to inquire into
what else changes at the precise time that the leaves drop.
29
This interlude was added after conversations with David Charles and Marko Malink convinced me
that the norms of inquiry that are my focus are of epistemic significance whether or not they are
sufficient for the identification of first principles qua first principles.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
devoted to earlier chapters (esp. 8–10), they were treated as if they had little
to do with chapter 19. And prior to three papers of mine (reprinted as
chapters 1–3 in Lennox 2001b), there was very little attention paid to the
five chapters directly preceding chapter 19, again as if they had little to do
with it.30 David Charles’ Aristotle on Meaning and Essence was, from the
standpoint of my interests here, a game changer. A central theme of that
book is the presentation of a “Three Stage Theory of Inquiry” grounded in
a systematic and detailed analysis of all of APo. ii. Moreover, Charles reads
ii.19 in the light of that analysis, so that the focus shifts to understanding
the distinction made there between experience and the whole universal,
and between that and the grasp of a universal as a scientific first principle.
What is missing, however, even in Charles, is a sustained investigation of
the epistemic or normative constraints on the inquiry process. There is, of
course, at the latest stages of inquiry, an issue about how you know you
have first principles. What I shall insist, however, is that if you have a story
about norms that govern the process of inquiry at each stage, norms that
are based on a view about what the goal of scientific knowledge looks like,
the issue of when nous grasps first principles as first principles is a much less
significant issue than previous scholars have thought. Because they lacked
a view about norms of inquiry, they had to put all the umph at the end of
the process – that is where justification, and epistemology, and norms
finally kick in. Up until then it is all armchair psychology. Part of the
reason for this, I believe, is the prevalence of a ‘justified true belief’ account
of knowledge that fits poorly with Aristotle’s view that all learning and
inquiry is based on preexistent knowledge. Along with that account of
knowledge is a (no longer as) prevalent dichotomy between ‘discovery’ and
‘justification,’ such that justification of a belief as knowledge was typically
assumed to have little or nothing to do with the process or processes by
which the belief was formed. We saw this dichotomy explicitly playing
a role in the work of Irwin and Frede – and I believe it plays an implicit role
that extends more widely. I propose an account of inductive inquiry in
which normative guidance, based on a prior understanding of what the
goal of scientific knowledge looks like, is present throughout the inquiry
process.
30
And it is telling that in an otherwise systematic study of APo. (Bronstein 2016) David Bronstein
jumps from a detailed and enlightening discussion of ii.13 directly to ii.19 – chapters 14–18 are the
only chapters in book ii that are not discussed.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
1.6 Foundationalist Empiricism 33
1.6 Foundationalist Empiricism
As I noted in introducing this consideration of alternative takes on
Aristotle’s views about the acquisition of scientific knowledge, Robert
Bolton has consistently insisted on reading Aristotle as prepared to provide
empiricist justification for his first principles.31
[T]he account in [APo.] ii. 19 is not, as some have supposed, merely
a genetic account of the psychological preconditions for the generation
of knowledge of first principles. It is an account of the inferential process
by which such knowledge is reached from previous knowledge. (Bolton
1991, 5)
In the paper from which that quotation comes, Bolton gives a careful,
section-by-section reading of Ph. i, and contrary to the expectations with
which one would emerge from reading Irwin 1988 (or, before him, Owen
1970), Aristotle’s candidates for the “causes and principles of things due to
nature” (i.7, 190b17-18) are, as Bolton reads it, justified primarily by appeal
to experience. Bolton also reads the claims of Ph. i.1 about moving from
the undifferentiated and confused universals to the elements and principles
against the background of APo. ii.19, noting that these initial universals are
of the sort that Aristotle claims we reach initially on our way to knowing
the causes in APo. ii.8. There our initial accounts of thunder, eclipse, and
human being are statements such as “noise in the clouds,” “a sort of
privation of light,” or “a sort of animal” (93a22-24; cf. Bolton 1991,
10–11). And he insists that the inductive reasoning to the principles in
Ph. i.4–8 is a form of justification for those principles, not merely
a description of the path by which principles are typically reached.
As with Aristotle’s earlier proof [in Ph. i.5] that the principles are
contraries, here also his proof that every case of coming into being
involves three things is based on a review of particular cases. It is the
fact that those cases are obvious (φανερά) cases of the general conclu-
sion that leads Aristotle to draw that conclusion as something which is
evident. So Aristotle’s argument is again inductive in form. (Bolton
1991, 28)
Unfortunately, the account of induction as the way to first principles that
Bolton provides, based as it is on APo. ii.19, is strikingly devoid of what any
self-respecting modern inductivist is looking for. In particular, it is devoid
31
However, in “Aristotle on Knowledge” (Bolton and Code 2012, 55–78) the view defended sounds
more like an IBE account. An alternative defense of Aristotle’s empiricist foundationalism is to be
found in Marc Gasser-Wingate’s forthcoming book, entitled Aristotle’s Empiricism.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
of one feature that, at least since Francis Bacon, everyone expects to be the
centerpiece of a theory of induction: a set of norms or standards to keep your
inquiry ‘on track’ and ‘self-correcting’.32 Aristotle is almost obsessively
concerned with the errors of his predecessors, and often the errors he
identifies are errors of method, and so he ought to have been concerned
to articulate such norms. But where, in APo. ii.19 for example, does one
find them? Norms for such things as proof, explanation, or proper defin-
ition abound – but what about inquiry, the subject of the entire second
book? Even those who wish to paint Aristotle as an empiricist of some sort
make excuses for him on this point.33
I shall argue that they need not do so, however. Both those who want to
depict him as a rationalist of one sort or another and those who depict him
as an empiricist have been looking in the wrong place for his views on
norms that need to be followed in order for an inquiry to be successful.
Now is a good time to be reminded of the words from the Prior Analytics
that I quoted in this chapter’s introduction.
The majority of principles for each science are distinctive to it.
Consequently, it is for experience to provide the principles concerning
each subject. I mean, for instance, that it is for astronomical experience to
provide the principles for the science of astronomy (for when the appear-
ances had been sufficiently grasped, in this way astronomical demonstra-
tions were discovered; and it is also similar concerning any other art or
science whatsoever). (APr. i.30, 46a17-22)
As I noted earlier, there are two primary messages we should take away
from this text:
1. Most of the causal principles and definitions that inquiry seeks are
domain specific.
2. The norms and standards for searching for them are likely also domain
specific, due to fundamental differences in the objects of inquiry, our
perspective on those objects, and our epistemic access to them. These
norms derive from extensive experience with subject matter. Very little
can be said in the abstract about this topic.
32
A convenient list of such inductive norms (or ‘rules’ as he calls them), with historical examples of
their application, can be found in chapter 6 of John Herschel’s wonderful A Preliminary Discourse on
the Study of Natural Philosophy (1830). An inexpensive facsimile of the first edition was produced by
The University of Chicago Press in 1987.
33
For example, here are the closing words of Bolton’s essay: “However crude his description or use of
the method may be his endeavours belong, in spirit, with those which we now think of as clearly
scientific” (Bolton 1991, 29).
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
1.7 Rethinking Norms, Warrant, and Inquiry 35
The implication I have drawn from these messages is that Aristotle’s views
on the question of what standards and norms are needed to insure success-
ful inquiry will be found in the methodologically normative passages in his
scientific works, especially, but not limited to, the introductions to these
works.
1.7 Rethinking Norms, Warrant, and Inquiry:

The Position in Outline
The basis of the Sellars/Davidson views that lie behind the arguments of
Frede and Irwin is that the physical processes involved in perception are
of a different ‘logical’ type than beliefs – something like ‘no physical
process can justify a belief, only other beliefs can justify beliefs.’ Such
a view is not necessarily opposed to inductive justification, provided
a belief about particular facts is being used to ground a belief about
a universal.
But suppose Aristotle’s view is that the quest for knowledge begins with
conscious, volitionally guided, perceptual awareness of particulars. That is,
just as some sort of neurological activity is necessary for belief formation
and use but doesn’t justify the beliefs, so some sort of sensory stimulation is
necessary for perceptual awareness but doesn’t justify the awareness. As he
puts it in De motu animalium (MA) 6:
Indeed, imagination and perception hold the same position as reason; for all
are discriminative faculties, but they differ according to the differences
discussed elsewhere. (700b19-22)34
Perception and thought are in the same logical category, namely,
‘conscious, discriminative activities.’ Aristotle can therefore avoid the
dilemma formulated by Davidson and Sellars if he believes, as he
does, that perception is a form of knowing, of being aware of the
world. In that case, inductive justification comes in the form of norms
governing the process that takes an inquirer from perceptual awareness
of particulars to universal claims about all relevantly similar
34
καὶ γὰρ ἡ φαντασία καὶ ἡ αἴσθησις τὴν αὐτὴν τῷ νῷ χώραν ἔχουσιν· κριτικὰ γὰρ πάντα,
διαφέρουσιν δὲ κατὰ τὰς εἰρημένας ἐν ἄλλοις διαφοράς. As Nussbaum notes (1978, 334), the
reference is likely to de An. iii.3, but this view of perception as allied with reason as
a ‘discriminative faculty’ (κριτικόν) is expressed or implied throughout the corpus, including the
opening paragraph of the Metaphysics, which notes that we are especially fond of vision among the
senses because “it reveals many differences” (980a27). It is also found in de An. iii.9, 432a15-16,
where the soul of animals is said to be defined by two capacities, one of which is the ‘discriminative’
capacity, which is the function of both thought and perception.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
particulars, and finally to knowledge of the essential features that
define those similar particulars and that explain their similarities. As
we will see, these norms include recommendations about how actively
to use that perceptual evidence in various comparative and contrastive
ways to avoid erroneous generalizations and increase confidence in the
inductive inferences we make, including those about which features of
a kind are essential and about causal relationships. They also include
recommendations regarding asking the appropriate questions, given
the subject being investigated and the perspective from which it is
being investigated. These, and many more, will be taken up in the
ensuing chapters.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
chapter 2
An Erotetic Framework
The Posterior Analytics on Inquiry
Regarding each methodos, accounts given philosophically differ from

those given nonphilosophically. Hence one should not, even when it
comes to politics, regard as superfluous the kind of study that makes
clear not only what something is (τὸ τί) but also that on account of
which it is (τὸ διὰ τί). For such is the philosophical approach
concerning each methodos.
(EE i.6, 1216b35-39)1
Chapter Summary. The second book of the Posterior Analytics is a sustained

investigation of different modes of inquiry aimed at achieving scientific
knowledge. So understood, it has far more unity than is typically claimed
for it, and its last chapter is a fitting conclusion. But it does not present, let
alone defend, substantive norms to guide distinct scientific inquiries. Those
norms are domain-specific and as such are presented and defended in the
introductions to Aristotle’s scientific treatises themselves. What, then, is
accomplished by APo. ii? In this chapter, I introduce and defend the idea
that APo. ii provides an erotetic framework for any inquiry, provided only
that its goal is epistêmê, demonstrative knowledge. As such, it can be thought
of as providing metalevel norms, for example, about how different inquiries
ought to be staged, or about how inquiries aimed at formulating definitions
are related to those aimed at achieving causal demonstrations.
2.1 Introduction: The Notion of an Erotetic Framework

The second book of the Posterior Analytics begins with a bold and sweeping
pronouncement: “The objects of inquiry are equal in number to those we
1
διαφέρουσι δ’ οἱ λόγοι περὶ ἑκάστην μέθοδον, οἵ τε φιλοσόφως λεγόμενοι καὶ μὴ φιλοσόφως. διόπερ
καὶ τῶν πολιτικῶν οὐ χρὴ νομίζειν περίεργον εἶναι τὴν τοιαύτην θεωρίαν, δι’ ἧς οὐ μόνον τὸ τί
φανερόν, ἀλλὰ καὶ τὸ διὰ τί. φιλόσοφον γὰρ τὸ τοιοῦτον περὶ ἑκάστην μέθοδον.
37
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
38 An Erotetic Framework
know scientifically” (89b23-24).2 The remainder of chapter 1 and all of
chapter 2 are concerned with two questions implicitly posed by this
opening statement: what is that number, and what are these objects of
inquiry?
Aristotle places no restrictions on the range of applicability of this claim
other than that implied by the concepts of inquiry and knowledge – see the
epigraph at the head of this chapter. But the answers to the two implicit
questions could take two quite different forms: they could be content
answers or framework answers. By a ‘content’ answer I have in mind an
answer of the following sort:
We know about different kinds of things: numbers, figures, virtues, social
organizations, animals, plants, materials . . . and we inquire about precisely
the same things.
By a ‘framework’ answer I have in mind the kind of answer Aristotle
actually provides and exemplifies in the remainder of the chapter, and
then modifies and restricts in various ways in chapter 2:
We inquire into four things: the ‘that,’ the ‘reason why,’ ‘whether it is,’ and
‘what it is.’ (89b24-25)
What Aristotle is referring to are not four different objects of inquiry but
four different modes of inquiry. Referring to this framework as ‘erotetic’
highlights the fact that these modes of inquiry are characterized as answers
to four different kinds of question, questions that can be asked about any
subject domain whatsoever. You will not arrive at these four distinct modes
of inquiry by starting with different objects of interest and then character-
izing them more and more abstractly. These four modes of inquiry seek
answers to four different kinds of questions a curious human being might
ask about the objects in any domain of inquiry. Exactly the same series of
questions could arise about anything at all, provided that at first you did
not know there was such a thing, eventually confirmed that there was, and
then became curious about its nature and its cause.3
2
Τὰ ζητούμενά ἐστιν ἴσα τὸν ἀριθμὸν ὅσαπερ ἐπιστάμεθα. A translation that stays closer to the roots
of ζητούμενα would be “The things we seek”; the suffix -περ added to the quantitative relative
pronoun ὅσα stresses that the numerical equivalence is exact.
3
Many episodes in the history of science exemplify the process Aristotle seems to have in mind. To
take one familiar example: based on a careful series of experiments on a variety of species of pea
plants, Gregor Mendel postulated the existence of what came to be called genes or alleles; but for
a long time researchers questioned their existence. By 1925, no one any longer questioned their
existence, but there were serious debates about their chemical structure and mode of functioning well
into the 1950s, when the work of people like James Watson, Francis Crick, and Seymour Benzer
provided satisfying answers to these questions.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
2.1 The Notion of an Erotetic Framework 39
Throughout the remainder of this book, whenever I say that APo. ii
provides an erotetic framework for inquiry, it is a framework of this sort –
and its ramifications throughout book ii – that I have in mind.4 This
chapter provides a broad brushstroke picture of APo. ii as providing an
erotetic framework for inquiry. But before turning to that task, it is worth
recalling how characteristic of Aristotle it is to begin a philosophical
investigation in this manner.
Consider, for example, a familiar erotetic framework also related to
different ways in which we can be said to know something: the mislead-
ingly named ‘four causes’ framework. I say ‘misleadingly named’ not
because I have doubts about the items referred to being suitably designated
as ‘causes,’ but because this characterization hides the fact that Aristotle
thinks of the four causes as four fundamentally different answers to the
question ‘Why?’ or (more accurately) ‘On account of what?’. Note how the
subject is introduced in the Physics ii.3:
Since our undertaking is for the sake of knowing, and we do nοt think
we know each thing until we grasp ‘the on-account-of-what’ (τὸ διὰ τί)
about each (and this is to grasp the primary cause), it is clear that we
must also do this regarding generation, destruction, and all natural
change. (194b17-20)
The four different ways of answering this question are also often identified
in the form of nominalized questions: the from-which-something-comes-to-
be, the what-is-it, the whence the source of change, and the for-the-sake-of-
which. In fact, in Ph. ii.3, the answer to this last question – the ‘final
cause’ – is defended as a legitimate cause by noting that it is one common
answer to the διὰ τί question:
For <we ask> ‘why does he walk?’, and we reply ‘in order to be healthy’ and
saying this, we think we have provided the cause. (194b33-35)
And when it comes time to make the case that the natural scientist must be
prepared to search for answers to all four questions in Ph. ii.7, he makes the
connection explicitly:
4
This language is borrowed from work in erotetic logic (the logic of questions and answers; see Prior
and Prior 1955, Belnap and Steel 1976). That work was motivated by the idea of a distinctive logic of
inquiry reflecting the peculiar syntax and semantics associated with the language of asking and
answering questions, and derives its name from the Greek verb ἐρωτάω. The extent to which
Aristotle’s epistemology is shaped by his view that all knowledge arises in the form of answers to
questions is, I think, underappreciated.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
That there are causes and that they are as many in number as we say is clear;
for such is the number of things τὸ διὰ τί embraces. (198a14-16; emphasis
added)
2.2 APo. ii: Providing an Erotetic Framework

With this reminder, then, let us return to APo. ii and consider it as
providing an erotetic framework for knowledge-seeking inquiry. As
noted above, it opens by distinguishing four forms of inquiry, correspond-
ing to four sorts of things we know: (i) inquiries into whether something
exists, which, if successful, lead to (ii) inquiries into what a thing is; and (iii)
inquiries into whether an attribute belongs to some subject, which, if
successful, lead to (iv) inquiries into the reason why it does. If successful,
the first pair of inquiries eventuates in a definition, the second pair in
a demonstration, in which the middle term refers to the cause of that state of
affairs.
In the next chapter, however, a more ‘interwoven’ picture of the rela-
tionship among these inquiries emerges:
But we are inquiring, when we inquire into the ‘that’ (τὸ ὅτι) or the
unqualified ‘if it is’ (τὸ εἰ ἔστιν ἁπλῶς),5 whether there is a middle for it
or not; while when, having come to know either the ‘that’ or ‘if it is,’ either
partially or without qualification, we inquire into the reason why (τὸ διὰ τί)
or the what-it-is (τὸ τί ἐστι), we are then inquiring what the middle is. (APo.
ii.2, 89b37-90a1)
This passage intentionally blurs a number of distinctions drawn in chapter
1. First, Aristotle is now downplaying the distinction between the two
‘preliminary’ inquiries, the distinction between ‘that it is’ and ‘if it is’
inquiries; there is now an unqualified and a partial (epi merous) version of
the ‘if it is’ inquiry, the latter of which appears to be identified with the
‘that it is’ inquiry (a move already hinted at in chapter 1 at 89b33).
Moreover, both are seen as attempts to find out whether there is a middle.
As I understand that claim, both are, though preliminary, guided by the
goal of causal demonstration. Likewise, the differences between the second
two inquiries, between ‘what it is’ and ‘why it is’ inquiries, are also
downplayed in this chapter; both are now described as attempts to discover
5
How should one read the “or” (ἢ) in this phrase? Is it the “or” of explication (“i.e.”), of qualification
(“or rather”), or of alternative (“this or that”)? In fact, this passage forces us to face this question not
once but four times. This is not the place to defend an answer, but I am reading all of them as
presenting alternatives, though not always exclusive alternatives.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
2.2 APo. ii: Providing an Erotetic Framework 41
what that middle is.6 This is of great interest, since, after reading chapter 1,
it would have been reasonable to suppose that the search for a middle term
would be appropriate only if one were searching for the explanation of
a previously established predication. But here, in chapter 2, as he puts it
a few lines later, in all our inquiries we are asking either whether there is
a middle or what the middle is: “for the middle is the cause, and in all cases
this is what is sought” (90a5-7). But that statement amounts to a third
blurring, of the distinction between a preliminary, purely fact- or exist-
ence-establishing stage of inquiry and a later stage of inquiry in search of
causes – those preliminary inquiries now seem to be aimed at establishing
that there is a cause. If they are to be thought of as in some sense ‘precausal’
inquiries, they are such in a different sense than is usually imagined.
Thus, while Aristotle has not abandoned his claim that there are four
distinct inquiries (in fact, the chapter opens by referring back to them), he
is now stressing the way in which they are interwoven: even as we inquire
whether some kind of thing exists or whether a subject has some attribute
or other, we are involved in some sort of causal, or anyway explanatory,
inquiry. Moreover, to assert that in either of these cases we are seeking
whether there is a middle stresses that the existential ‘if it is’ inquiry and the
predicative ‘that it is’ inquiry, are, at the very least, intimately related. And
by insisting that all such inquiries are causal, he is suggesting that, in
seeking answers to ‘what is it?’ questions, we are also engaged in the search
for answers to questions about why the objects of inquiry have the attri-
butes they do.
In order to elaborate on the idea that APo. ii provides an erotetic
framework for inquiry, I want to highlight certain features of the rest of
the book that point in this direction. Chapters 3–7 systematically explore
aporiai related to understanding how the search for definitions and the
search for demonstrations (i.e., for answers to the τί ἐστι and the διότι
questions) are related to one another. He introduces that discussion as
follows:
So then, it is clear that every inquiry is the search for a middle; but we must
now discuss how the what-it-is (τὸ τί ἐστι) is established and in what way it
is grounded, and what a definition is and of what, reviewing the puzzles
about these things first. (APo. ii.3, 90a35-38)
The puzzles about the answer to the ‘what is it?’ question and about
definitions are said to rest on the premise that every inquiry is the search
6
Cf. 90a14-15: “In all these cases it is apparent that ‘what it is’ and ‘why it is’ are the same.”
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
for a middle (i.e., a cause) – presumably including inquiries about what
things are. Indeed, this is the source of the puzzles. Chapter 3 considers and
rejects the possibility that we can know the same thing in the same respect
both by definition and by demonstration; chapter 4 reconsiders, making
systematic use of the machinery of the syllogistic, the idea that there can be
a demonstration of what something is (i.e., its definition), and concludes
that any attempt to do so begs the question; chapter 5 considers and rejects
division as a method that can prove a definition, making reference back to
the arguments against viewing division as a form of proof in APr. i.31;
chapter 6 considers the idea of providing a sort of demonstration ex hypothesi
of what something is, again showing that any such attempt is question-
begging; and chapter 7 considers and rejects the possibility that some subset
of existence proofs could be considered demonstrations of the ‘what it is.’7
Based on these aporetic explorations, chapters 8–10 consider how these
two pairs of inquiries are in fact related, concluding, as we would expect
from chapter 2, that they are tightly interwoven. The discussion opens by
making it clear that the previous discussion was not the last word on the
subject at hand:
Let us inquire again into which of these things are well said and which not,
and into what definition is and whether there is in some way demonstration
and definition of the what it is (τὸ τί ἐστι), or in no way at all. (APo. ii.8,
93a1-3)
The following three chapters argue that one type of definition prevalent in
science can be reformulated as a demonstration.8 These chapters have been
the subject of intense scrutiny over the past fifty years,9 and it is not my
purpose here to add substantively to those discussions. What I shall do,
however, is point to a number of features of these chapters that highlight
both their framework character and their focus on inquiry.
Chapter 8, in fact, reminds us of the erotetic framework established in
chapters 1 and 2:
Just as, grasping the fact (τὸ ὅτι) we inquire into the reason why (τὸ διότι) –
in some cases they become clear at the same time, but it is certainly
7
For a detailed consideration of these chapters, see Bronstein 2016, 144–147; Byrne 1997, 147–156;
Charles 2000, ch. 7.
8
As he puts it, such a definition “will clearly be like a demonstration of what something is, differing in
arrangement from a demonstration” (Barnes trans.; APo. ii.10, 94a1-2). Charles 2013 explores the
metaphysical underpinnings of the relationship between definition and causal explanation in APo. ii.
9
Ackrill 1981; Bolton 1976; Bolton 1987; Bronstein 2016, chs. 9–10; Charles 2000, chs. 2, 10; Goldin
1996; McKirahan 1992.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
2.2 APo. ii: Providing an Erotetic Framework 43
impossible to know the reason why prior to the fact – in the same way it is
clear that it is not possible to know the essence (τὸ τί ἦν εἶναι) without
having the fact (τὸ ὅτι). For it is impossible to know what something is (τί
ἐστιν) when one is ignorant whether it exists (εἰ ἔστιν). (APo. ii.8, 93a16-20)
There are in this brief passage two subtle reminders of the lesson of chapter
2, the intertwined nature of the four inquiries mentioned in chapter 1. The
first pair of inquiries mentioned here are to hoti and to dihoti, but when the
discussion then turns to inquiry into the essence, that inquiry is also paired
with to hoti inquiry – and then, in explicating that pairing, Aristotle
substitutes ei esti for to hoti. What is at issue in this passage is the order of
inquiry, and by this substitution Aristotle is sending the message that
determining whether some kind of thing genuinely exists is closely tied
to determining whether it possesses certain attributes nonaccidentally.
He goes on to provide a list of examples of having such a nonaccidental
grasp of to ei estin, and that list provides further evidence of the ‘framework’
nature of APo. ii:
Sometimes we grasp whether something is accidentally, and sometimes by
grasping something of the thing itself e.g., thunder that (ὅτι) it is a certain
noise in the clouds; eclipse, that it is a certain privation of light; human
being, that it is a certain animal; soul, that it is self-moving. (APo. ii.8,
93a21-24)
The evidence for the framework status of this discussion lies in the
extraordinary range of the examples: thunder is a meteorological phenom-
enon; eclipses are objects of cosmological and astronomical inquiry; human
beings are subjects of natural/zoological, medical, and ethical inquiries; the
soul, it will be argued in chapter 7, is in part the object of a natural and in
part of a metaphysical inquiry. That is to categorize these examples from the
standpoint of domain; but they are also ontologically diverse – thunder and
eclipses are ephemeral processes, human beings are natural substances, and
souls are forms of certain natural substances. That is, the examples empha-
size that this argument about how inquiries are to be ordered applies to any
inquiry concerning any sort of being that is a suitable object of scientific
investigation.10 As we will see, this is a prominent feature of the discussion
from this point on. Aristotle is here establishing what I think of as
‘framework’ or ‘domain-neutral’ norms of inquiry: inquiries that establish
10
It is worth noting, however, the total absence of mathematical examples in these chapters. There is
a parenthetical reference to the practices of arithmeticians in chapter 9, but without an example.
Indeed, from this point on the ratio of examples from natural science to those from mathematics is
quite high, in contrast to book i. See Wians 1996b, 131–150 (in Wians ed. 1996a).
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
that a certain object of inquiry exists or that a certain attribute belongs in
a nonincidental way to a certain subject must be settled before moving on
to attempt to answer questions about the essences and causes of such
facts.
He concludes this thread of argument by insisting that “to inquire what
<something> is when not having grasped that it is, is to inquire about
nothing” (93a26-27). That is, he concludes by reminding his readers that he
has been making a point about the order of inquiry, stressing the import-
ance of establishing that there really is a legitimate object of investigation
before going in pursuit of its nature.
The remainder of chapter 8 primarily works with the example of coming
to understand eclipses in order to spell out the process of moving from
knowing the fact to knowing the reason why. He asks us to imagine a point
at which we grasp that the moon is being eclipsed because it is unable to
cast a shadow though there is nothing between us and it (93a35-39). At this
stage, he says, it is clear that the moon is eclipsed, but not why; and that
there is an eclipse, but not what it is – note, yet again, the interweaving of
our four framework questions (93b1-3). He goes on:
When it is clear that A [eclipse] belongs to C [moon], to then ask on account
of what (διὰ τί) it belongs is to inquire what B is, whether an obstruction or
rotation or extinction of the moon; and this is the account (ὁ λόγος) of the
other extreme [term], in these cases A; for eclipse is obstruction by the earth.
(APo. ii.8, 93b3-7)
He then goes through the same process, this time using thunder as his
example. Notice, again, that though there is a ‘syllogism of the fact’
illustrating the grasp of the fact of a lunar eclipse, this entire discus-
sion is about inquiry – and by offering three alternative explanatory
middles he is stressing that identifying the actual cause will likely be
a difficult process.11
2.3 Problems and a Framework for Causal Search

Among the concerns that lie behind this insistence is avoiding the Meno
paradox – often presented as an epistemic paradox, it is explicitly a paradox
11
Twice in earlier chapters he has used a thought experiment to make this point: “If we were on the
moon, we would not have to inquire either whether the eclipse was occurring or why – both would
be clear simultaneously. For from the perceiving of it the universal would become known to us”
(APo. ii.2, 90a26-29; cf. APo. i.31, 87b39-88a4). Recall that chapter 8 begins by noting that it is
possible for the fact and the reason why to be discovered simultaneously in certain cases (93a17-18).
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
2.3 Problems and a Framework for Causal Search 45
about inquiry.12 Aristotle sees inquiry as a process with distinguishable
stages, beginning with our perception of distinct particulars and ending
with knowledge of first principles that are universal, indemonstrable
truths, and which identify fundamental causes constituting the essences
of things. It is reasonable to ask whether the characterization of this process
in APo. ii is sufficiently rich to provide an investigator with the norms
needed to guide actual inductive inquiry in specific domains of inquiry.
The answer I defend in the following chapters is ‘No’: but what this
discussion does provide is an erotetic framework, and that framework
establishes its own norms of inquiry. In fact, a strong case can be made
that the last seven chapters of APo. ii provide a framework for inquiry
targeted on the goal of finding causes of already established universal
relationships. In the remainder of this chapter, I outline that case.
In particular, chapters 14–18 explore an important intermediate stage of
inquiry on the path to demonstrative knowledge, which involves moving
from what Aristotle refers to as ‘partial’ universals to coextensive universals,
sometimes designated as primary universals, the kinds of propositions that
Aristotle thinks form the core of any science, properly so called.13 Here he is
exploring, albeit in a somewhat preliminary and ‘aporetic’ way, the process
of advancing from early, factual (τὸ ὅτι) stages of inquiry to the point
where we have knowledge of commensurate relationships among universals.
This leads him to discuss an issue familiar to contemporary philosophers of
science: among robustly correlated attributes of a kind, how does one
identify those that are causally fundamental and provide answers to inquir-
ies into the reason why (τὸ διὰ τί) and the essence (τὸ τί ἐστι)? As I read
them, these chapters provide important context for the discussion in APo.
ii.19. They are exploring the question of how to move to knowledge of
commensurate universals, and then beyond that knowledge to determining
12
As emphasized by Kleiner 1988. The ‘aporia of the Meno’ is explicitly referred to at APo. i.1, 71a29 (as
is the doctrine that learning is recollection at APr. ii.21, 67a21). But it is also clearly in the
background of the aporia that introduces APo. ii.19. Cf. Bronstein 2016, 229–234. Charles 2010a,
ch. 3 (2010b), 8 (2010c); Ferejohn 2013, ch. 3.
13
By ‘partial universal’ I refer to attributes that Aristotle describes as ‘belonging to all (of a kind) but
extending beyond’ (e.g., the attribute ‘interior angles equal to two right angles’ belongs to all scalene
triangles, but extends to all triangles). The important defense (at APo. i.24, 85a13-86a30) of the
superiority of universal (katholou) to partial (kata meros) demonstration is much easier to compre-
hend once one realizes that the Greek often translated ‘universal’ (katholou) literally means ‘accord-
ing to the whole,’ and that the Greek for ‘partial’ (kata meros) literally means ‘according to the part.’
The argument in APo. i.24 is in defense of the superiority of demonstrating that an attribute belongs
commensurately to its subject versus demonstrating that it belongs to all the members of a subset of
that subject.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
which per se attributes of a kind are causally basic (i.e., to grasping first
principles).
2.3.1 Preparing Problems for Demonstration: APo. ii.14–18

APo. ii.13 begins with a highly condensed but accurate summary of what
Aristotle has accomplished by the end of ii.10, after which he announces
a discussion of “hunting down what is predicated in [accounts of] what
something is” (96a22-23). Often commentators talk as if this refers only to
chapter 13, but it accurately describes chapters 13–18.14 These are the
chapters where Aristotle has an extensive discussion of how to formulate
scientific problems as a preliminary stage in the search for causes.
Valuable context was provided in APo. i for these chapters during
Aristotle’s sustained argument that the explanandum of a proper demonstra-
tion should be an attribute that belongs per se to its subject and qua that
subject. The theory of demonstration that Aristotle defends places severe
constraints on the premises of a demonstrative syllogism: that they be true,
primary, and immediate (i.e., not themselves conclusions derived from more
basic premises) and that they be better known than, prior to, and causally
explanatory of the explanandum identified in the conclusion (APo. i.2, 71b20-
22; these conditions are then explained and defended from 71b26-72b3). This
will guarantee that the principles will be appropriate to what is to be proven.
But what does it mean for scientific principles to be appropriate? It
means that the premises of a science that meet the first three conditions
mentioned above are related to its conclusions in the way specified by the
last three conditions. In fact, the concept is first introduced in APo.
immediately after those six conditions are presented:
demonstrative understanding in particular must proceed from items that are
true, primitive, immediate, more familiar than, prior to, and causes of the
conclusion. In this way the principles will also be appropriate (αἱ ἀρχαὶ
οἰκεῖαι) to what is being proved. (i.2, 71b20-23)15
14
And thus, while I will note disagreements with the discussion of these chapters in Charles 2000,
chapters 8 and 9, we agree that chapter 13 provides a context for the discussion of problems in
chapters 14–18. It is useful to be reminded that the chapter divisions were introduced in the
Renaissance and have no ancient authority. What is puzzling about these summary remarks is
that they do not mention chapters 11 and 12, the chapters that explore causality in the context of
syllogistic demonstration. Interestingly, while Bronstein 2016 has an excellent discussion of chapter
13, there is no sustained discussion of chapters 14–18.
15
And compare: “For it is not what is reputable that we count as a principle, but rather what is primary
in the kind with which the proof is concerned – and not every truth is appropriate (oikeion)” (APo.
i.6, 74b24-26).
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
One way to frame the concerns people have about how Aristotle grounds
his first principles is in terms of the question of how you could start with
relatively unsystematic experience of a subject and end up confident that
you have knowledge of premises that actually meet the above-mentioned,
very robust conditions. A necessary condition that any candidate for such
a premise must meet is that it be a “primary universal” – that the subject
and predicate be commensurate.16
Two related passages that close APo. i.4 and 5 respectively explore the
question of how an inquirer might ‘test’ or ‘try’ whether a feature holds
both primarily and as such of some subject. This is the mark of a predicative
relationship ‘ready’ for scientific demonstration.17 The first of these pas-
sages comes near the end of chapter 4:
Something holds universally when it is proved (δεικνύηται) of an arbitrary
and primary case; e.g., having two right angles [2 R] does not hold univer-
sally of figures – you may indeed prove of a figure that it has 2 R, but not of
any arbitrary figure; for quadrangles are figures but do not have 2 R. An
arbitrary isosceles does have 2 R – but it is not primary; triangles are prior.
Thus if an arbitrary primary case is proved to have 2 R (or whatever), then it
holds universally of this primarily, and the demonstration applies to it
universally and in itself. To the other items it applies in another way, not
in themselves – it does not apply to the isosceles universally, but extends
further. (APo. i.4, 73b32-74a3)
Two features of this passage are of note: first, it describes a procedure for
finding the kind to which an already grasped attribute belongs primarily
and in virtue of the kind it is; second, it notes that belonging to all instances
of a kind is not sufficient for belonging as a primary universal – a primary
universal must belong commensurately to its subject. The word I have
translated ‘prove’ above is often used by Aristotle to refer to proof in
a technical sense, though its root meaning is to point to or show. Here,
though, in the context of geometry and where demonstration (apodeixis) is
a special kind of deixis, ‘proof’ is clearly what Aristotle intends.
But what is it to prove in this context? The procedure described would,
it seems, involve drawing or constructing various figures – triangles with
various kinds of sides and angles, other rectilinear figures with more than
16
There is an excellent discussion of these passages in Ferejohn 2013, 81–97. He refers to these ‘primary
universals’ as Cathólic Predications. My only significant disagreement with that discussion is
Ferejohn’s insistence (84–86) that the “qua itself” condition is extensional, like the “belongs to
all” condition, rather than intensional, like the per se condition. Cf. Hasper 2006.
17
The connection between being prepared for demonstration and these two properties is made
especially clear at APo. i.9, 75b37-76a9.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
three sides – and then carrying out the appropriate test to establish that the
interior angles of that figure do or do not sum to 180 degrees. The test
would be based on certain assumptions about parallel lines and angle
equivalences, of course; but the testing would be recognizably inductive
in Aristotle’s sense – something universal about triangles as such being
made clear by a study of concrete instances of particular kinds of figures.
Chapter 5 is primarily a discussion of the ways in which we mistakenly
believe we have grasped such a commensurately universal relationship
when in fact, we have not. It opens with the following concern:
It must not escape our notice that we often make mistakes – that which is
being proven does not belong as a primary universal, though it seems to be
proven as a primary universal.18 We make this error either when there is
nothing higher we can grasp apart from the particular case, or when there is
but it is nameless and covers objects of different forms, or when the proof
applies to something that is in fact a partial whole. (74a4-10)
Notice that this is a very sophisticated form of error that is being discussed.
It is the error of imagining that a partial demonstration is a universal
demonstration: for example, imagining that a proof that every scalene
triangle has 2 R or every vine suffers leaf loss is a demonstration of
a commensurately universal proposition. Are there ways to avoid making
this sort of mistake? Aristotle considers this question near the close of the
chapter.
Does it [2 R] belong to them as triangle or as isosceles? And when does it
belong in virtue of this and primarily? And to what does the demonstration
apply universally? Clearly that to which it belongs primarily when features
are being removed (ἀφαιρουμένων); for example, 2 R will belong to bronze
isosceles triangle – and also when being bronze and being isosceles have been
removed (ἀφαιρεθέντος), but not when figure or limit have been. But
neither of these is the primary feature. Then what is primary? If triangle,
it is in virtue of this that 2 R belongs to the other items, and it is to this that
the demonstration applies universally. (APo. I.5, 74a36-b4)19
18
I am avoiding the common practice of translating πρῶτον καθόλου, which appears twice here (and
again at 74a11 and a12), conjunctively (“primitively and universally,” as do Barnes and Tredennick),
since it obscures the fact that Aristotle is characterizing a type of universal predication that is primary
and distinguishing it from what he sometimes refers to as what holds κατὰ παντός (cf. i.4, 73b25-28;
74a10-11) – that is, distinguishing convertible from nonconvertible universal predication. The
common mistake, for which Aristotle is identifying three distinct sources, is of thinking that
a merely extensional, nonconvertible universal predication is a primary, convertible universal
predication.
19
Hasper 2006 provides a detailed analysis of this chapter. For a good discussion of problems in
translating this passage see McKirahan (1992, 172–173); I don’t find Barnes’ emendation at 74a38
necessary (1994, ad loc.). It will be noted that the passive participle of the verb ἀφαιρέω appears twice
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
Here, we learn a bit more about the procedure implicit in chapter 4. You
(correctly) believe that the figure before you, which is a bronze, isosceles,
triangular figure, has 2 R. But in virtue of what, qua what, does this
figure have that attribute? In virtue of being bronze? If so, a triangle
drawn in the dirt should not have it, while any random bronze object
should. Is it then in virtue of being isosceles? Then a scalene triangle should
not have it. Is it in virtue of being a plane figure? Then nontriangular
geometric figures should have it. Eventually, by a sort of pincer movement,
you zero in on ‘triangle’ as the kind to which the attribute belongs as such –
it is in virtue of being triangular that this, or any, figure has the property
2 R. There are clear procedures for testing each of these options and for
finding the subject to which the property in question belongs in itself and
as such. These procedures can be seen as domain-neutral, or ‘framework’
norms of inquiry – for as we will see, these procedures are as useful in the
study of plants as they are of geometric figures.
From the first passage, we see that Aristotle’s characteristic way of
describing the ‘isosceles’ case is to say that the 2 R property belongs to
every member of the kind but ‘extends beyond’ it. We also see there explicit
recognition of a self-correcting procedure for zeroing in, as it were, on
a subject/attribute relation that is per se and as such – you can test whether
triangles have this property because they are figures by looking at an
arbitrary selection of plane figures other than triangles, and you can test
whether it is in virtue of being the kind of triangle it happens to be by
looking at other kinds of triangles to see whether they too have the
property.20 The procedure is necessarily both empirical and comparative/
contrastive.
In these passages, as commentators have noted, Aristotle is adopting
a strong notion of ‘universality,’ primary universality, which he stipulates
in chapter 4: “I call ‘universal’ what holds of every case and in itself and as
such” (73b25-26).21 As he points out earlier in the chapter, you might have
proven on different occasions that scalene, isosceles, and equilateral tri-
angles each have interior angles equal to two right angles and thus have
to describe the process of cognitively ‘removing’ features to identify which is the first that, being
removed, gives you a subject to which 2R no longer belongs. That is significant, since the expression
often translated ‘abstraction’ (ἐξ ἀφαιρέσεως), based on the abstract noun formed from this verb, is
used by Aristotle to refer to the process by which mathematicians ignore or ‘subtract’ the materials
and processes of nature in order to focus on a natural object’s mathematical features, qua
mathematical.
20
People familiar with Bacon’s Novum Organum will see obvious affinities to the description in that
work of procedures for creating ‘tables of presence and absence’ (NO ii.11, 12).
21
See notes 17 and 18, above.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
(extensionally) proven it of all triangles – but unless you did so knowing
that it is qua triangle that they have this property, your proofs fall short of
providing unqualified scientific knowledge.
There is a puzzling feature of these discussions that is rarely noted,
however: the critical middle term that would be required to demonstrate
that 2 R belongs per se to Triangle and qua Triangle is never mentioned.22
The proposition that is in focus and under discussion here, therefore, is the
conclusion of an imagined demonstration that is never revealed to us. What
the procedures used in these passages aid in achieving, then, are proposi-
tions that are prepared to serve as the conclusions of demonstration, not the
premises of demonstration. This point is driven home in chapter 6, when
Aristotle turns to making the case that the middle term, which identifies
the cause in virtue of which these per se features belong to a kind, must also
belong to the kind per se and as such.23
These two passages in book i show that Aristotle has reflected on the
problems of identifying features that belong per se and as such to the kinds
being investigated by a science. As we are about to see, chapters 14–18 of
APo. ii explore how to identify primary universal predications through
a different, though related iterated procedure involving ‘why’ questions;
and these same chapters raise a number of puzzles about how to identify
causal primaries among the primary universals thus achieved. This too is
part of the erotetic framework that APo. ii provides for any inquiry at all –
a point Aristotle again drives home by using examples from a remarkably
diverse range of domains.
2.3.2 In Search of Appropriate Problems

A ‘problem’ (problêma) in the context of Aristotle’s Analytics is
a proposition about which one is seeking, but does not yet have, a proof.
Once one does have a proof, the same proposition is a conclusion
(sumperasma).24 This explains why the concept appears, in APo., only in
chapters 14–18 of book ii, for in those chapters Aristotle is discussing
difficult questions about the search for universal propositions that are
prepared for demonstration and about identifying the causal middle
term needed for a demonstration. These chapters, in turn, are applying,
22
This fact is rarely taken note of by commentators; but see Ferejohn 2013, 115–120.
23
cf. APo. i.6, 74b5-12, 74b27-32.
24
On the concept of a problem in Aristotle, see Lennox 2001b, 13–15, 48–53, 76–91; and Byrne 1997,
55–80. I discuss the question of how Aristotle’s use of the concept of problêma in the Analytics and
the Aristotelian Problemata Physica are related in Lennox 2015, 36–58 in Mayhew (ed.) 2015.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
within the context of scientific inquiry, methods for identifying proper
middles relative to syllogistic problems, developed in APr. i.27–30. This is
only to be expected, for the APr. is the first part of the investigation “about
demonstration and . . . of demonstrative science” (περὶ ἀπόδειξιν καὶ
ἐπιστήμης ἀποδεικτικῆς; i.1, 24a10-11). The transition to a discussion of
problems, in the more abstract domain of the theory of the syllogism,
occurs at APr. i.26–27:25
So then, it is clear from what has been said how every syllogism comes
about, both through how many terms and premises, and how these are
related to each other; and furthermore, what sort of problem (ποῖον
πρόβλημα) is proved in each figure, and what sort [is proved] in more and
what sort in fewer figures. But how we may ourselves always be well supplied
with syllogisms related to what is taken up, and through which route (διὰ
ποίας ὁδοῦ) we can grasp the principles (ληψόμεθα τὰς ἀρχάς) about each
thing, we must now state; for it is not only necessary to study the coming to
be of syllogisms but also to have the capacity to produce them. (APr. i.26,
43a16-24)
In a demonstrative context, problems are stated as why-questions for which
we seek causal explanations26 – recall how the Physics introduces the four
kinds of cause: “we think we have scientific knowledge of a thing only
when we can answer the question ‘on account of what?’, and that is to grasp
the primary cause” (Ph. ii.3, 194b18-21).
The concept πρόβλημα is not introduced in APo. until ii.14. This
chapter sketches a means of grasping problems by “selecting” information
from “dissections and divisions.” This is language explicitly borrowed from
the discussion of selecting middle terms for premises relative to a specific
problem in APr. i.27. Moreover, the techniques sketched for selecting flow
quite naturally from the long discussion in the previous chapter of using
division in the hunt for definitions (i.13, 96b15-97b6).
Aristotle begins with an example that reflects already familiar subjects
and attributes, then moves on to cases where the universals lack common
25
The chapter division, which occurs halfway through this passage, makes very little sense. Bekker
quite rightly does not have a paragraph break in his text.
26
Indeed, in the “Problems” tradition, which may have been inspired by these chapters of the APo.,
this becomes a trope: virtually every chapter of the 38 books of Problems (now generally assumed to
be composed by later members of Aristotle’s school rather than by Aristotle), which run from 859b1-
967b26 in the Bekker edition of the Aristotelian Corpus, begin with the expression dia ti (why?).
I have made some tentative suggestions about how these books may be connected to the distinction,
in Greek geometry, between theorems and problems in Lennox 1995 (reprinted in Lennox 2001b,
chapter 3).
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
names, cases drawn from his own biological investigations. The familiar
example is used to identify a procedure:
1. Posit a higher kind common to all the objects being studied; for
example, animal (98a2-4).
2. Select from your dissections and divisions what belongs to all members
of that kind;27 for example, select what belongs to every animal
(98a1-4).
3. Then select what follows (= belongs to)28 all of the “first remaining
things” – from his examples, he means to select attributes that belong
to all the groups you come to by dividing animal into the most
immediate sub-kinds. He mentions “bird,” which is one of two
examples he consistently mentions as extensive kinds (megista genê)
of animals appropriately identified in the Greek language (the other
being fish)29 (98a5-7).
He then explains why we are doing this:
For it is clear that we will already be able to state on account of what the
attributes that follow belong to things that fall under the common kind, for
example, why they belong to man or horse. Let A be assigned to animal, B to
what follows every animal, and C, D, E to specific animals; surely it is clear
why B belongs to D, for it is on account of A. (APo. ii.14, 98a7-11)
Here we have what amounts to a domain-neutral, or ‘framework,’ norm for
constructing preliminary explanations: the norm involves the combined
use of selection and division. It is important to note that the ‘explanation’
formalized at the conclusion of this passage is not a demonstration accord-
ing to the standards we saw articulated in book i: if the conclusion here
27
Pace Ross (1949, 663–664), there are compelling reasons to consider the expression τὰς ἀνατομὰς
(‘the dissections’) to be a reference to the lost collections of anatomical diagrams, collections referred
to 28 times in Aristotle’s zoological works and referred to in the ancient catalogues of Aristotle’s
writings (e.g., D.L. v.25; cf. references to Divisions at D.L. v.23, 24). First, this is the only use
Aristotle makes of this term outside of the context of dissection, so that if it had a different sense here
it would be unprecedented; second, as we will see, unlike other chapters in the APo., all the examples
in this chapter are zoological, and the second and third are technical examples that require
knowledge of animals that can be acquired only by wide-ranging dissections on many different
kinds of animals. The first example, by contrast, draws much more obviously on the language and
techniques of division discussed in chapter 13 (96b35 ff.). In fact, the use of “dissection” here may
provide some insight into what the works Aristotle refers to as “the dissections” (typically in
conjunction with the “histories”) were, since the usage here suggests that the information embedded
in these works was helpful for determining the extension of one part relative to other parts.
28
The expressions “select what follows S” and “select what S follows” are borrowed from the Prior
Analytics chapters referred to earlier, and are roughly equivalent to finding what “belongs to every S”
and what “S belongs to every one of” (cf. 43b1-37, 44a37-b6, 44b20-36).
29
See PA i.2, 642b14-16, i.4, 644b1-8; HA i.6, 490b7-9.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
predicates of one specific kind of animal a property that is common to all
animals, it follows that it does not belong to the subject of the conclusion
“in virtue of itself and as such.” The property may belong per se to animal,
but even that is not guaranteed by the method outlined here, since
something could belong to every animal because it is a property that
belongs to every organism.30
If the method here outlined is not intended to direct us toward a proper
demonstration of the proposed problem, what is its intent? W. D. Ross,
following the sixteenth-century Paduan Aristotelian Zabarella, gave the
following gloss on what is going on: “this chapter is concerned with advice
not as to the solution of προβλήματα . . . but as to their proper formula-
tion” (1949, 662). Recall that the chapter opens with the claim that the
method to be described would aid in “grasping problems,” not in formu-
lating demonstrations.31 Moreover, as we will see, the proper formulation
of problems is clearly one of the primary goals of the chapters following this
one. As those chapters clearly demonstrate, Aristotle is interested in iden-
tifying procedures that will guide investigation originating with problems
that are universal but noncommensurate toward commensurate, primary
universals of a higher level of abstraction. And as we will see by means of
the range of examples he uses, these are domain-neutral procedures,
framework norms that will be useful in the pursuit of knowledge in any
domain at all.
On this line of interpretation, then, Aristotle is indeed outlining
a method for grasping problems that are prepared for unqualified demon-
stration. The recommendation is to posit the kind common to the various
things that are being studied (τὰ τεθεωρημένα), in this case animals, and
then to select features common to every animal. Suppose after consulting
the result of dissections you find that hearts are common to every animal.32
There is now a sort of answer to your initial problems: Birds and fish have
hearts because they are animals, all of which have hearts. But that implies
that hearts do not belong to birds qua bird or to fish qua fish. This is an
exact analogue of the example we looked at in APo. i.4–5: by coming to
recognize that isosceles or scalene figures each have 2 R because they are
triangles, you are recognizing that 2 R is a proper attribute, not of these
types of triangle, but of triangle as such. It is about this convertible
30
I refer to these as “A-explanations” in Lennox (1987 [2001b, ch. 1]); McKirahan as “application
arguments” (1992, 177–178). The kind (γένος) is in the middle term position, so that you explain
why some form of the kind has a property by noting that it is a property common to the kind.
31
See Lennox (1994, 64 n. 21 [2001b, 82 n. 20]) and (1987, 90–99 [2001b, 7–15]).
32
Aristotle actually would say “heart or heart analogue.”
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
predication, Heart–Animal, that one may ask the properly formulated
problem. And by the search procedure outlined in the first example in
APo. ii.14, you should be given the materials and methods to get hold of
such problems.
What Aristotle has not even hinted at, however, is a procedure for
discovering the answer to a question of this sort: supposing that hearts
belong to animals qua animals, why do they do so? By using the techniques
Aristotle discusses here, we should eventually find problems prepared for
strict demonstration, about properties (e.g., parts) that belong (in this case)
to animal qua animal. But what about the search for causal middle terms?
As we will see shortly, this is the problem on which Aristotle focuses in
chapters 15–17.
Before leaving chapter 14, however, there are two more examples that
further clarify Aristotle’s intentions. The previous example, Aristotle
acknowledges, tapped into familiar concepts drawn from everyday experi-
ence: bird, animal, horse, human being – he did not even bother to identify
the attribute to be selected from our ‘divisions and dissections’ that belong
to all animals – it is some value for the variable β. But he now recommends
carrying investigation well beyond such familiar cases.
At present we are speaking according to commonly assigned names, but one
ought to investigate not only in these cases, but also if some other thing
belonging in common has been observed – selecting it, we should investigate
what it follows and what follows it, for example, having the third stomach and
lacking upper incisors follows having horns and again having horns follows
something. So it is clear why (διὰ τί) the identified feature will belong to
these, for it is because having horns will belong. (APo. ii.14, 98a13-19)
Notice that these are normative recommendations for inquiry or investi-
gation, and specifically for inquiry that goes beyond the realm of things
that have established names. There are no ‘kind terms’ for that which is
common in this case;33 rather we have three verbal phrases – literally “that
which has the third stomach,” “that which lacks both upper and lower
teeth,” “that which has horns.” The Greek expressions are even ambiguous
as to whether they are referring to groups of animals with these features or
to the features themselves. But (probably not coincidentally) there is an
extensive discussion of the relationship among these three features in
Aristotle’s On the Parts of Animals.34 What that discussion stresses is that
33
Our term for the group under discussion is ‘ruminant,’ but the Greeks had no common name for
this group.
34
Cf. Gotthelf (1987, 178–185 [2012, 164–171]) and Lennox (1991, 275–276 [2001b, 49–50]).
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
these three properties – multiple stomachs, the lack of upper incisors, and
horns – are (i) co-extensive and (ii) causally connected. In brief, the
explanation offered there is as follows: the production of horns limits the
amount of bony material left available for forming teeth, causing the lack
of upper incisors; that in turn limits the extent to which food can be
masticated, a problem dealt with by means of multiple stomachs (cf. PA
iii.2, 662b35-663a7, 663b28-664a2; iii.14, 674a30-b17). Without him say-
ing so, then, what Aristotle has provided here are norms to aid us in
identifying a problem formulated appropriately for a proposed explanation.
The possession of horns is identified as the feature on account of which
having multiple stomachs belongs to all animals lacking upper incisors –
but in our passage, he offers no justification for that choice. That, as we will
see, is significant: a major theme running through chapters 15–18 is pre-
cisely the question of how to identify, once you have established that
a number of properties are coextensive, which (if any) of them is causally
fundamental.35
There is a third example in this chapter that is also suggestive and helpful
for framing a discussion of the following chapters. It is brief enough to
quote in full:
Yet another way [to grasp problems] is to select [predicates] according to
analogy. For it is not possible to grasp one thing that is the same, which
sêpion,36 fish spine, and bone should be named; but there will be things
following these as well, as if there were some one nature of this sort. (APo.
ii.14, 98a20-23)
Because each of these internal, bony structures is found in different great
kinds (cephalopods, fish, quadrupeds) these structures will only be ana-
logues. But if one searches among the features predicated of each of them,
one may find features they have in common and consider whether all these
features are due to a common cause. In fact, Aristotle thinks these parts are
35
This is precisely the issue in terms of which certain counter-examples were posed to the deductive-
nomological account of explanation initially proposed by Carl Hempel and Paul Oppenheim in
1948 (for a nice summary of the examples and the issue at stake, see Salmon 1990, 46–50). For
example (one due originally to Sylvan Bromberger [1962], though according to Salmon, 189 n. 12,
this precise example never appeared in publication), given sunlight shining on a flagpole of height
h at a given angle a, there will be a shadow of length l. You can deduce the length of the shadow from
the height of the flagpole and vice versa. But since the interaction of the light and flagpole determine
the length of the shadow, while the shadow is not similarly implicated in determining the height of
the flagpole, there seems to be more to explanation than mere deducibility. Aristotle, as we will see,
discusses a number of similar examples.
36
This is an evolutionary remnant of the shell possessed by the common ancestor of present-day
cuttlefish; cf. HA iv.1, 524b23-30, iv.7, 532b1.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
materially alike and that the material likeness provides them with the
ability to play a ‘skeletal’ role in the three different kinds.37 And thus
there may be a common explanation for their presence in each of the three
kinds, even though they are only one by analogy. That is, beginning with
problems such as “On account of what do cuttlefish have a sêpion?” or “On
account of what do land animals have bones?” you may find a common
answer, even though there is no common kind to be identified as there was
in the first example.
Chapter 14, then, appears to assume, as its point of departure, that
a certain problem has been identified, and that the investigator is in search
of an explanation. Its focus is on identifying a set of procedures for finding
a middle term relative to that problem, procedures that build on and
deploy the technical machinery of APr. i.27–30, chapters devoted to
finding middle terms for valid syllogistic proofs. In some cases, the subject
of the problem will belong to an already identified kind and you can take
that as your starting point. Making use of knowledge embedded in cata-
logues of divisions and dissections, you search for attributes that belong to
every member of that kind and for attributes to which every member of
that kind belongs, relying on the method of division for “hunting down the
what is it” discussed in chapter 13. If no kind has been identified relative to
the problem you are working on – say, “On account of what do animals
with no upper incisors have multiple stomachs?” – you can still use the
method of asking what the subject belongs to all of and what belongs to all
members of that subject class. Even when the objects in need of explan-
ation are only analogously alike, one should look to see if they have
attributes in common, for they may have a common explanation. This
chapter, in short, provides a set of framework norms to aid in identifying
problems that are appropriate for causal demonstration.
Chapter 15 is focused on characterizing ways in which different problems
may be related to each other.38 Aristotle distinguishes three types of case,
and in each case how the middle terms proposed as the explanantia are
related is the key.
1. Once one identifies a middle term for initially distinct problems, it
may turn out that the middle term is the same. His ‘example’ here is
37
“There is even an analogue to the spines of fish in the soft-bodied animals; for example, in the
cuttlefish there is a part called the ‘sêpion,’ in the squids what is called the ‘sword’” (PA ii.8,
654a19-22).
38
One thing that is relevant to determining whether these chapters of APo. set the agenda for the
tradition represented in the thirty-eight-book collection of problems is whether there are consistent
attempts in those books to discover these sorts of relationships among initially distinct problems.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
unhelpful because he doesn’t identify the distinct problems, but only says
that one might discover that the explanatory middle in all cases is ‘mutual
displacement.’ Should this turn out to be true, it could lead you to the
conclusion that these apparently distinct problems are actually instances of
one, more abstract problem. Though the example here is not spelled out
enough to be helpful, this is identical in form to the examples Aristotle
discusses in APo. i.5, in which, for a variety of reasons, an inquirer may be
deluded into thinking that the appropriate universal subject for an
unqualified demonstration has been identified when it has not. One of
those reasons is that there is a universal feature common to the distinct
problems, but this common feature lacks a name and so has not been
identified. The example in that chapter was a property common to
proportions:39
That what is proportional alternates, though at one time proved separately
as numbers, as lines, as solids, and as times, can be proven of them all by one
demonstration; but due to all these things – numbers, lengths, times, solids –
not being given some one name and differing in form from one another,
they were grasped separately. But now it is proved universally; for it is not
qua lines or numbers that alternation belonged, but qua this, which is
posited to belong universally. (APo. i.5, 74a17-25)
The four kinds to which this property belongs are said to ‘differ in form.’
That is not sufficient for them to lack a name, of course, since they could be
forms of a commonly recognized kind – but by saying “qua this, which is
posited to belong universally,” Aristotle is acknowledging that this is not
the case here. In fact, Aristotle uses this very example to illustrate how
problems may be alike in one respect and different in another in ii.17:
For example, why do proportions alternate? For the cause in the case of lines
and of numbers is different, and yet the same – qua line, different, but qua
having such and such an increase, the same. (APo. ii.17, 99a8-10)
Here we see the alternando theorem treated as a problem – why do propor-
tionals alternate? Aristotle is suggesting that there are, in fact, problems at the
specific level that will have specific causal explanations appropriate to that
level, but a more general problem, and a way of characterizing these causes
such that it is the same for both problems. Many distinct problems at the
lower level become unified at the higher level.40
39
A scholium to Euclid, Elements 5 says that the general theory of proportion presented there is due to
Eudoxus. See Heath (1931, 190–191).
40
For a detailed discussion of this example, both in APo. i.5 and ii.17, see Hasper 2006, 262–269.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
2. In some cases the problems may be genuinely different, but generically
the same. Here the focus is not on the identity of the cause, but of the
problems.41 Here his example is helpful. Suppose you are concerned to
solve the following three problems:
a. Why are there echoes?
b. Why do images appear in mirrors?
c. Why do rainbows occur?
Each of these is genuinely different, but Aristotle notes that they all involve
reflection – “all these are the same problem in kind (for all are [cases of]
reflection), but are different problems in form” (APo. ii.15, 98a28-29).
I take it that what distinguishes this sort of case from the first is that here
the causal processes underlying reflection in the three examples are really
different: there is no sense here that we mistakenly thought these were
different phenomena but an advance in knowledge has shown us they are
in fact all the same. Rather, sound, sunlight, and patterns of shape and
color can be reflected, but the form of reflection is very different.42
3. Finally, he discusses cases where distinct problems are related because,
while they are different, the middle term that is the answer to one is not
“immediate” but subordinate to the middle term that answers the other.
Again, he provides an example, one that is related to a complex and
fascinating history:43
Why does the Nile flow more when the month is ending? Because there are
more storms when the month is ending. But why are there more storms
when the month is ending? Because the moon is waning. (APo. ii.15,
98a31-33)
Here the two problems are: “Why is there increased flow in the Nile at
month’s end?” and “Why are there more storms at month’s end?” For the
41
However, it remains unclear how different these cases are – it may be nothing more than a matter of
whether one is focused on the identity of the middle term or of the proposed problem. Hasper 2006
suggests that in the ‘proportionals’ case the general demonstration is applicable to the various cases
without the cases being generically identical, a solution that may be applicable here as well.
42
As I will discuss in Chapter 6, PA i.1, 639a19-b3 distinguishes attributes, such as forms of locomo-
tion, that are generically alike in animals but that “differ by a difference in form” from features, such
as sleep and respiration, that are the same across kinds.
43
A work attributed to Aristotle on this subject is preserved only in a thirteenth-century Latin
translation, Liber Aristotelis de inundatione Nili. For a thorough discussion of the history of this
text, who the original translator may have been, and its authorship, see Beullens (2014). The
connection between this manuscript and the passage we are discussing is by no means direct. The
problem that De inundatione Nili discusses and presents an explanation for is why the Nile floods in
the summer rather than the winter, whereas in our passage the problem is why the Nile increases in
flow level at the end of each month.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
sake of illustration, Aristotle asks us to imagine that the increase in flow is
due to an increase in storms, and that the increase in storms is due to the
waning of the moon, presuming (I suppose) that the moon’s waning has
some effect on the atmosphere. It is noteworthy that what is a premise in the
explanation of the first problem becomes the problem to be explained in
the second. Though Aristotle’s language of one middle term being ‘under’
another may imply this, he never states that what this example displays is
a search for a more fundamental or immediate cause. In fact, it seems likely
that in this case the moon’s waning would be appropriate as an explanation
for the periodic occurrence of storms, but inappropriate, because too
remote, as an explanation for the flooding of rivers generally, let alone
a particular river.
2.3.3 Primary Universals and Causal Primacy

What unites the discussions in chapter 15 of the ways in which distinct
problems may be related is the methodology by which one goes about
discovering that distinct problems are related and how they are related.
These discussions all presuppose a context of inquiry, that is, a process
of question-driven causal search. And that context leads directly to the
fundamental philosophical issue that opens chapter 16:
Concerning a cause and that of which it is a cause one might be
puzzled whether, when the effect (τὸ αἰτιατόν) is present, the cause (τὸ
αἴτιον) is also present (so if there is a shedding of leaves or an eclipse,
the cause of there being an eclipse or a shedding of leaves will be
present; for example, if this is the possession of broad leaves or, of the
eclipse, the earth being in the middle; for if it is not present, some
other thing will be the cause of these things); and likewise, if the cause
is present at the same time the effect also [is present] (e.g., if the earth
is in the middle an eclipse occurs, or if broad leaves, leaf shedding
occurs). (APo. ii.16, 98a35-b4)
This passage nicely highlights the ‘erotetic framework’ character of this
discussion: the same philosophical difficulty is illustrated with examples
from domains that could hardly be more different, botany and astron-
omy. And what, precisely, is the aporia about which Aristotle is con-
cerned? The puzzle is Aristotle’s version of the classic ‘length of shadow/
height of flagpole’ problem referred to earlier – if the two properties are
truly commensurate, how do you determine which is explanatory of
which? One can see that this is the concern by looking to the next
sentence in the text: “If this is how things are, [cause and effect] would
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
be simultaneous and could be proved through one another” (98b4-5).44
This is a concern that had already been discussed in APo. i.13, in the
context of distinguishing two different kinds of demonstration, one
proceeding “through the fact”45 and another “through the reason why.”
Aristotle’s concerns, in a nutshell, are two:
1. Is there something in the nature of the relationship between cause and
effect that requires that they be commensurate?
2. Once one has arrived at the stage of an inquiry achieved by the
previously described ‘problems’ method, where you have identified
a number of properties that are appropriately commensurate, how do
you take the next step of identifying which of these terms refers to the
property that is causally fundamental and explanatory of the others?
He first illustrates his concerns by means of a demonstration in which
the problem is at the lower, noncommensurate level of universality: Why
do grape vines lose their leaves? The example is:
Shedding Leaves: A; Broad-Leafed: B; Grape vines: C
A belongs to all B, B belongs to all C, therefore A belongs to all C. (ii.16,
98b5-10)
Here, while B is assumed to be the middle term, A and B are commensurate
with one another, and both are of wider extension than C. He then notes his
worry:
But you can also demonstrate that vines are broad-leafed by way of the fact
that they shed their leaves. . . . Hence every vine is broad-leafed and
shedding is explanatory. (ii.16, 98b10-19)
That is, since A and B are commensurate, why are we assuming B to be the
causal middle term? Notice that, once again, the problem arises because we
have been given no criterion by which to determine, once we have identi-
fied a commensurately universal connection, whether it is a causal
44
Barnes translates the verbs as simple futures, but since this in the midst of stating an aporia and this
is the apodosis of a conditional, it is preferable to make the modal force of the potential optative
explicit. It is also important to note that the verb is δείκνυμι, to show or prove, not ἀποδείκνυμι, the
verb he uses for scientific demonstration. It is possible to construct a proof with either the cause or
the effect in the middle term position, but only one of those proofs will be a demonstration, in which
the middle term identifies the cause.
45
The proof that has the more familiar, noncausal middle term does not meet the criteria for
unqualified demonstration as he has defined it in APo. i.2–6; but there are a number of such uses
of ἀπόδειξις in the Posterior Analytics (e.g., i.13, 78a30, a36; i.24, 85a13-17, a30), and they often seem
to refer to cases where, at a certain stage of inquiry, one could think that one had produced an
unqualified demonstration.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
connection – nor, if it is a causal connection, which of the related attributes
is the cause and which is the effect.46 But this passage does assume that the
connection between being broad-leafed and shedding leaves seasonally is
a commensurate one, and the example takes it for granted that one of the
two commensurate terms is the cause of the other. Moreover, the ‘demon-
stration’ here is not an unqualified, or B-type, explanation – whichever of
the commensurate terms you use for your middle term, it is wider in extent
than the subject of the conclusion. So, in effect, this example has only
illustrated a case where we have moved the inquiry to the right level of
universality to search for the cause of the robust connection between being
broad-leafed and leaf-shedding. At the end of chapter 16, picking up on the
leaf-shedding example once more, Aristotle acknowledges precisely that,
and begins to move toward a resolution of these issues. He asks: “If
problems are always universal and the cause is something whole, then
must that of which it is the cause also be universal?” (98b32-33). He urges
us to an affirmative answer:
For example: shedding leaves is determined to some whole, even if this has
forms, and it holds of these universally (either of plants or plants of a certain
form). Hence in these cases the middle term and that of which it is the cause
must be equal [in extension] and convert. For example, why do trees shed
their leaves? If it is because of solidification of their moisture, then if a tree
sheds its leaves solidification must hold, and if solidification holds – not of
anything whatever but of a [sort of] tree – then the tree must shed its leaves.
(APo. ii.16, 98b33-37)
Without fanfare, the ‘middle’ that had provided an interim, A-explanation
for why various forms of broad-leafed trees shed their leaves – being broad-
leafed – is now (implicitly) the minor (subject) term in the conclusion, and
a material/efficient cause takes up residence as the causal middle. As in all
the examples in these chapters, Aristotle has sketched a cognitive ascent to
a level where subject and predicate in the conclusion are commensurate
and a new middle term is identified as the cause. What had been taken as
the cause when this example first appeared – being broad-leafed – is now
identified with the sort of tree to which the predicate ‘leaf-shedding’
belongs per se and as such, and solidification of the moisture (presumably
sap) is identified as the cause of such ‘trees’ losing their leaves.47
46
Or, as he imagines in the next chapter, whether the connection between them is due to a common
cause.
47
For a very different interpretation of the relationship between these two passages, see Ferejohn 2013,
147–155; for an interpretation more in line with that offered here, see Charles 2000, 204–209. I raise
questions about Ferejohn’s account in Lennox 2014c.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
Near the end of his discussion of the search for explanations relative to
specific problems, he fleshes out this example considerably, comparing it to
a geometric example of a property (external angles equal to four right
angles) that belongs to both triangles and quadrangles but extends beyond
them.
The cause, that of which it is the cause, and that for which it is a cause are
interrelated in this way. If the subjects are taken case by case, the effect has
a wider extension (e.g., having external angles equal to four right angles
extends further than triangle or quadrangle); but if they are all taken
together, the effect is equal in extent. . . ; and likewise with the middle
term [i.e., the cause]. And the middle is the definition of the first extreme, which
is why all sciences come about through definition. For example, shedding of
leaves follows every grape vine and extends beyond, and follows every fig tree
and extends beyond, but it does not extend beyond all. Indeed, if we grasp the
primary middle, it will be an account of leaf shedding. For there will be a first
middle in the one direction (that all are such and such); and then a middle
for this (that sap solidifies, or some such). What is the shedding of leaves? The
sap solidifying in the connection with the seed. (APo. ii.17, 99a16-29; emphasis
added)
This passage is reminiscent of the passages discussed earlier in APo. i.4–5.
One could imagine the daughter of a Greek vintner growing up on
a vineyard and watching the vines lose their leaves year after year, and
mistakenly concluding that this is a property distinctive of grape vines. At
a certain point in her life, she visits neighboring farms with fig trees and sees
that they also shed their leaves seasonally; and she gradually learns that many
plants do this. She begins to search for something that all these plants have in
common and which differentiates them from plants that do not lose their
leaves, and notices that they all have broad leaves.48 At this point, she may
posit this as the cause of shedding. But then, a new question quite naturally
arises: Why do all broad-leafed trees lose their leaves? Is that simply
a primitive, immediate connection, or is there something more basic about
these trees that is responsible?49 That question leads her (if she is curious and
scientific) to search for what precisely happens to the broad-leafed trees at the
48
It is important to note here that near the end of this discussion, at 99a30-b7, Aristotle considers
whether the same effect could have a different cause in different kinds – for example, could the
attribute of being long-lived be due to different causes in four-legged land animals versus birds
(99b4-7)? Though it appears he leans toward supposing that they are not really the same effect if the
kinds are truly distinct, he leaves the question open.
49
As Charles stresses (2000, 206–207), Aristotle’s question is whether there is something about broad-
leafed trees as such that produces leaf loss. Even if there is, that does not rule out other ‘incidental’
causes of leaf loss, such as disease or poisoning.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
2.4 Conclusion 63
time of the year when they lose their leaves. When Aristotle says that there
will be “a first middle in the one direction,” I take it he is referring to the
identification of the appropriate subject of shedding leaves – being broad-
leafed – that initially served as the middle term in accounting for why specific
kinds of broad-leafed plants shed their leaves. This interpretation is con-
firmed by him next saying that there will be a middle “for this” – that is,
a new, more fundamental middle, for the commensurately universal prob-
lem of why broad-leafed trees lose their leaves. In this passage, Aristotle also
ties his discussion of finding problems appropriate for causal demonstration
back to the account of the relationship between definition and demonstra-
tion in APo. ii.8–10: he concludes by simply stating that the causal middle of
the demonstration will also be a definitional account of the explanandum. As it
was with eclipse and thunder in those earlier chapters, so here it is with leaf-
shedding.50
2.4 Conclusion
The primary purpose of this chapter has been to make the case for APo. ii
providing what I’ve been calling an erotetic framework for scientific
inquiry, that is, inquiry aimed at scientific knowledge. The framework
provides norms for how inquiries ought to be ordered and for how we
ought (and ought not) to think of the relationship between definition- and
demonstration-oriented inquiries – that is, between the search for essences
and the search for causes. I’ve focused special attention on the ‘Problems’-
oriented chapters (ii.14–18), highlighting the way in which they give us
a glimpse into Aristotle’s views about the process of establishing the sort of
factual knowledge that is poised for causal understanding, including the
role that weaker forms of explanation can play in that process.
Throughout, Aristotle highlights the framework status of the discussion
by repeatedly exemplifying his philosophical point with examples drawn
from widely disparate areas of knowledge. But in the very generality of its
results lie the book’s limitations.
The first principles of a science, we are told in the last chapter of the
Posterior Analytics, are achieved by a process called induction (ἐπαγωγή).51
I conclude this chapter by drawing attention to what is not discussed in
APo. ii.14–18 that one might expect to be, especially in chapters leading up
50
This interdependency of definition and causal demonstration is one of the primary themes of
Charles (2000); it is discussed in the context of these examples on 209–214.
51
Cf. EN vi.3, 1139b25-31.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
to ii.19: there is no discussion of the process by which we get from
a knowledge of a commensurate relationship between such things as 2 R
and triangle, or shedding leaves and being broad-leafed, or having multiple
stomachs and lacking upper teeth, to the identification of the causes that
explain these relationships. Take the last example I discussed in the previ-
ous section, solidification of fluid as the cause of leaf-shedding (presumably
sap, which ordinarily provides fluid nutrition to the leaf): Aristotle identi-
fies it as the causal middle, but he says nothing about how one might arrive
at that identification. Similarly, Aristotle never mentions what it is about
triangles that explains their having 2 R, or why it is possession of horns that
explains few teeth or multiple stomachs, rather than the other way around.
While we get some hints about how Aristotle imagines that an inquiry
moves from ‘close to perception’ experience of some subject (‘Vines shed
their leaves in Autumn’) to the kinds of commensurate universal truths
that a science seeks to understand, there is no hint of how one finally
identifies the causal primaries at that commensurate level. The causes that
are identified in these discussions, when they are, come, as it were, out of
the blue.52
This may appear to lend credence to the ‘insight’ reading APo. ii.19: the
mind is now prepared for that final flash of recognition. My aim in what
follows is to present a more attractive alternative: In order to understand
Aristotle’s views about proper norms for causal search, we need to look
beyond the Posterior Analytics, to the many methodological reflections in
his scientific treatises. Those texts suggest that Aristotle thinks the search
for causes is governed by norms and methods that are specific to specific
domains and thus impossible to characterize adequately in a work as
abstract as the Posterior Analytics. That work provides the erotetic frame-
work for such investigations, and that is of fundamental importance. We
should not expect more of it than that.
52
It is easy to miss this because we fall into the habit of supplying plausible answers from common
sense or from his other treatises.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
chapter 3
A Discourse on μέθοδος
Every art and every methodos, and likewise every action and decision,
seems to aim at some good: hence it has been well said that the good is
that at which all things aim.
(EN i.1, 1094a1-3)1
Chapter Summary. Aristotle uses the word μέθοδος in a variety of

crucial passages in the works that constitute his studies of nature, and
in two distinct, yet overlapping, ways: it can refer to the norms of
research appropriate to a domain-specific investigation, and it can
refer to such an investigation carried out according to such domain-
specific norms. It thus forms a conceptual bridge between the
domain-neutral norms of APo. ii discussed in the previous chapter
and the domain-specific norms that will be discussed in the remain-
ing chapters. In this chapter, I investigate the uses of this concept in
Aristotle’s writings on nature. By comparing them with his use of the
term in other fields, and the uses of ὁδός and μέθοδος in Plato and in
the Hippocratic On Ancient Medicine, it emerges that Aristotle has
rich and sophisticated views about ‘proper ways of proceeding’ in an
investigation. The norms that are appropriate to a specific inquiry are
shaped in part by Aristotle’s general views about the goal of inquiry,
scientific knowledge; but they are also, and crucially, differentiated to
take three factors into account: (i) differences in the objects being
investigated, (ii) differences in our epistemic access to those objects,
and (iii) differences in the perspectives we are able to take on the
same objects.
1
Πᾶσα τέχνη καὶ πᾶσα μέθοδος, ὁμοίως δὲ πρᾶξίς τε καὶ προαίρεσις, ἀγαθοῦ τινὸς
ἐφίεσθαι δοκεῖ· διὸ καλῶς ἀπεφήναντο τἀγαθόν, οὗ πάντ’ ἐφίεται.
65
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
66 A Discourse on μέθοδος
3.1 Introduction
The title of this chapter is a triple entendre: the chapter will be, to a large
extent, an exploration of Aristotle’s use of the term μέθοδος in relationship
to the context of inquiry – a discourse on μέθοδος. This concept may also
have referred to the subject of a now lost work by Aristotle entitled
“Methodics” – his ‘discourse on method,’ as it were. And finally, the title
is intended to call to the reader’s mind a well-known treatise published in
1637 entitled Discours de la méthode, often read on its own in philosophy
courses, but written by its author, René Descartes, as an introduction to
a number of essays on natural philosophy and mathematics. A generation
prior to its publication, Jacopo Zabarella wrote his De methodis, one of the
documents that helped shape the Aristotelianism taught in Padua when
Galileo Galilei and William Harvey were there. Descartes may well have
been thinking of his work as providing an alternative to Zabarella’s highly
influential work, in something analogous to the way Francis Bacon
thought of his Novum Organum as an alternative to the Organon of the
Aristotelians. It is clear that Aristotle’s views on this topic are a significant
part of the historical background for both Zabarella and Descartes.
3.2 The Word

The term μέθοδος is a compound formed from the preposition μετά and
the noun ὁδός. The root noun can refer either to a path along which
a journey takes place (a road, course, or route) or to the activity of following
that path (a journey, voyage, or expedition) – and thus, by metaphorical
extension, to a way of proceeding toward a cognitive goal, such as know-
ledge. Used in the dative, ὁδός can also have adverbial force, about which
we will have more to say later. The exact force provided by the prepos-
itional prefix is best decided by seeing the word in its contexts, but it is
reasonable at this point to suggest that the compound conveys the idea of
proceeding along or according to a defined route, with the implication of
their being a goal or end of the journey that is best achieved by proceeding
in a certain way or along a certain course.
3.3 Philosophical Background i: Plato

The first appearance of methodos in a philosophical context is in Platonic
dialogues, and Plato’s use of this concept provides an important part of the
context for Aristotle’s. Book vii of the Republic opens with the metaphor
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
3.3 Philosophical Background i 67
of the Cave, which introduces the thought that, for those who have lived all
their lives in a world of appearances, even the best minds among them will
need to be properly trained and educated to contemplate Forms, and in
particular, the Form of the Good. Eventually (at 521d), a discussion of the
proper education for the Guardian class ensues: it should begin with
arithmetic, move on to geometry and solid geometry, and finally to
astronomy and harmonics.2 At the end of this curriculum outline, at 531c7-
d3, Socrates comments that if the methodos we have used to progress
through these studies is such as to reveal their commonality and kinship
to each other, then it is worthwhile, but otherwise worthless. With that
comment, the “power of dialectic” is introduced, later referred to as “the
dialectical method” (ἡ διαλεκτικὴ μέθοδος, 533c7). Glaucon then poses the
critical questions:
Tell me what is the modality (τρόπος) of the power of dialectical thinking,
and into what forms it is distinguished, and again what are its routes (ὁδοί).
For these, as I suppose, would lead us to that thing, by reaching which we
would be, as it were, at a resting place from our journey and an end of our
travels. (R. vii, 532d8-e3)
Socrates then insists that no one would argue that “there is any other
methodos that endeavors systematically (ὁδῷ) and about everything to
grasp about each thing what it is” (533b2-3).
Here then, dialectic is characterized in terms of a metaphor of a route
and journey toward the goal of knowledge of Forms and in particular, the
Form of the Good, a mode of characterizing the philosophical quest for
true understanding that goes back at least to Parmenides. Parmenides uses
ὁδός in the Proem in its literal sense (though in a mythical context) to refer
to a journey, and in its metaphorical use to characterize two “ways” or
“routes,”3 one being the way of Truth, the way “that it is,” the path down
which, according to the Goddess, right and justice have sent him. And
certainly, the context in the Republic echoes this Parmenidean back-
ground – for both Parmenides and Plato, there is a deceptive world of
2
This progression is somewhat surprising, since on certain epistemological assumptions you might
expect the progression to move from less to more abstract, and Socrates recommends beginning with
pure mathematics and moving toward what appears to be the application of the theory of ratios and
proportions to music and of solid geometry to the heavens. This does allow him, however, to
disparage a particular way of conceiving harmonics and astronomy as being about appearances.
3
The Proem (DK 28 B1) of Parmenides’ poem depicts a chariot’s journey on a “celebrated road” (ἐς
ὁδὸν . . . πολύφημον), maidens leading the way (κοῦραι δ’ ὁδὸν ἡγεμόνευον) to a meeting with
a goddess, a road upon which right and justice sent him. DK 28 B7 famously warns against traveling
the way of much experience (πολύπειρον ὁδὸν) and DK 28 B8 begins by asserting that only one story
remains about a way (μόνος δ’ ἔτι μῦθος ὁδοῖο): that it is.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
appearances and opinion that can lead inquiry in the wrong direction and
carry us away from truth and understanding, a path that must be avoided.
For both thinkers, behind the need to ‘stay on course’ is a fact about human
cognition: reason can be misled by perception and needs guidance and
direction to stay on the right track toward the truth. Discovering the truth
about things as they are is not automatic, and it is the philosopher’s task, or
at least one central one, to identify the goal, the path to that goal, and
norms for staying on the path to knowledge. Besides an implicit agreement
on that point, these two thinkers explicitly agree that it is the deceptive
nature of sensory experience that throws us off track.
Determining the precise epistemic character of the clearly normative
language here is difficult. Plato states quite clearly that it is not merely the
study of these mathematical disciplines that prepares the mind for grasping
Forms – this study has to take place in a certain order, and it needs to be
done in such a way as to reveal what is shared and common to all these studies.
But precisely how the teaching of these disciplines in this order prepares
the mind for grasping Forms is not clear.
In the Sophist, Plato conjoins the concept of a methodos with another
metaphor that Aristotle adopts for discussing inquiry, the metaphor of the
hunt.
Judging the Sophist to be a very troublesome sort of creature to hunt down,
let us first practice the method (τὴν μέθοδον) of tracking him down on some
easier quarry – unless you have some other, readier path (ὁδόν) to suggest.
(Sph. 218d1-6)
The mixing of the two metaphors continues once the Eleatic Stranger has
moved on to practicing the method of division in pursuit of the true nature
of the Sophist.
But if he [the Sophist] should find some hiding place among the subdivi-
sions of this art of imitation, we must follow hard upon him, constantly
dividing the part that gives him shelter, until he is caught. In any event there
is no fear that he or any other kind shall ever boast of having eluded this
method, consisting of the powers to pursue in this way each and every
quarry (τὴν τῶν οὕτω δυναμένων μετιέναι καθ’ ἕκαστά τε καὶ ἐπὶ πάντα
μέθοδον). (Sph. 235c1-6; Cornford with modifications)4
In the Sophist and Statesman, it is the systematic use of division that the
Eleatic Stranger is recommending as the method for pursuing an answer to
4
Note too the periphrastic construction at 243d7, δεῖν ποιεῖσθαι τὴν μέθοδον ἡμᾶς, likely modeled on
the quite common expression ὁδὸν ποιεῖσθαι, to make a journey.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
3.3 Philosophical Background i 69
the ‘what is it?’ question in every case – here, about the Sophist.5 Similarly,
at the conclusion of a passage in which the Eleatic Stranger proposes
a division that requires creating a category for which there is no name,
he reminds Young Socrates that “our inquiry is for the sake of the ruler
(τοῦ γὰρ ἄρχοντος ἕνεκα ἡμῖν ἡ μέθοδος ἦν) and not the opposite” (Plt.
260e8-261a1; compare 266d4-9). In these later dialogues, then, there is
a clear suggestion that Plato has concluded that division is sufficient for
acquiring knowledge of what any particular object of pursuit is. This will
be significant for our discussion of Aristotle’s use of methodos, since in
a number of cases Aristotle is explicitly critical of division as an appropriate
or sufficient methodos for the cognitive task at hand, typically while offering
his own alternative.
A well-known passage in the Phaedrus will provide a segue into one other
text that provides valuable historical context for Aristotle’s use of the
concept of methodos, On Ancient Medicine. In this passage, Socrates begins
to draw a parallel between medicine and rhetoric: if we are seeking an art in
each case and not mere routine and experience, we need to grasp natures –
the nature of the body in the case of medicine, of the soul in the case of
rhetoric. In fact, he goes on, it is “the nature of the whole” of the soul that
rhetoric needs to understand, just as Hippocrates claimed that “without
this method (μέθοδος) it will not be possible to understand the body”
(270c3-5). There is a notorious difficulty with this passage both in deter-
mining what method Socrates goes on to describe and to which, if any,
Hippocratic text it is comparable.6 Those difficulties, thankfully, I can
avoid. For my purposes, the importance of this passage lies in the fact that
it uses the concept of methodos to refer to a detailed set of steps that must be
taken in the investigation of an object if our goal is knowledge and not
mere know-how or experience – key components of Aristotle’s deployment
of this concept.
After a preliminary outline of the appropriate way to begin such an
inquiry (270d1-7), Socrates remarks that “the inquiry (μέθοδος) without
these steps would seem to be like the meandering of a blind man” (270d9-
e1). On the view articulated in this passage, then, it is not sufficient for
5
Compare Plt. 286d5-9: “Again in relation to the inquiry into something problematic, reason
commands that finding the answer easily and quickly should be regarded as secondary, not primary;
while it commands us to honor most of all and primarily the procedure itself of being able to divide
according to forms (πολὺ δὲ μάλιστα καὶ πρῶτον τὴν μέθοδον αὐτὴν τιμᾶν τοῦ κατ’ εἴδη δυνατὸν
εἶναι διαιρεῖν).”
6
On the endless debate about which, if any, work in the extant Hippocratic corpus is being referred to
in the Phaedrus, cf. Smith 1979, 44–50, Joly 1983, 28–47.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
inquiry aimed at knowledge that one have experience of the subject being
investigated, though such experience may well be necessary. In addition,
the inquiry must follow a prescribed set of steps if its goal is to be achieved.
In its first appearance in this passage, methodos could be referring either to
the norms needed to guide the inquiry into the soul or to an inquiry
governed by those norms, an ambiguity mirrored, as we will see, in
Aristotle’s use of the term.7 In the second use, Socrates at least allows for
the possibility of a methodos without such norms, but then characterizes it
as “the meanderings of a blind man.” I suspect that if Phaedrus were to
have asked, at this juncture, whether it then really deserved to be called
a methodos, Socrates would have responded, “Certainly not, my good
Phaedrus.” For in Plato’s usage, to act in a norm-governed manner appears
to be at least partially constitutive of the concept. So is it, as we shall see, in
Aristotle.
3.4 Philosophical Background ii: On Ancient Medicine

This Platonic reference to the views of Hippocrates about a norm-governed
route toward knowledge points in the direction of another text that,
though it does not make use of the term μέθοδος, does provides further
insight into the use of this metaphor of a proper “route” or “path” in
pursuit of knowledge at the time Aristotle was coming to philosophical
maturity – On Ancient Medicine (VM), perhaps the most methodologically
self-conscious text in the Hippocratic corpus.8 The text begins by arguing
that the sorts of “empty hypotheses” that are typical in writings about the
heavens or the center of the earth, where there are no touchstones for
determining how things stand regarding the truth, are unnecessary in
medicine (VM 1, ll. 16–21).9 In medicine, the author claims, from ancient
times everything we need has been present, its early practitioners having
discovered both a starting point (ἀρχή) and a path (ὁδός) in virtue of which
many wonderful discoveries have been made over a long time (VM 2, ll.
7
And it is an ambiguity that is rooted in this concept’s original context: recall that a ὁδός can refer
either to the path or road one takes to a goal or to the activity of traveling along that path (LSJ s.v.
ὁδός, i and ii).
8
There is no direct evidence that Aristotle was familiar with this work; in fact, the only direct evidence
that Aristotle was familiar with any work in the Hippocratic corpus is a lengthy quotation in HA
iii.3, 512b13-513a8, of a text which appears twice in Hippocratic works, once in chapter 9 of De natura
ossium and once in chapter 11 of De natura hominis. In some manuscripts, Aristotle attributes the
quotation to Πολύβιος, which is the name of Hippocrates’ son-in-law; in the majority, however, it is
attributed to Πόλυβος.
9
The line numbers follow Heiberg 1927.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
3.4 Philosophical Background ii 71
1–4). There is a proper route to follow, and anyone who ignores this and
“attempts to search by means of a different path and plan (ἑτέρῃ ὁδῷ καὶ
ἑτέρῳ σχήματι)” has been misled (VM 2, 4–9).10
In laying out “what the art is” (VM 2, 8), the author of VM is not
focused primarily on proof, explanation, or exposition. The focus is on
inquiry and how properly to proceed in discovering the nature of health, the
causes of sickness, and the means of maintaining and restoring health. To
be in possession of the art of medicine is, then, to be in possession of the
proper method of inquiry – again, a key feature of Aristotle’s concept of
methodos. The method the author goes on to outline begins with reflec-
tion on the ancient practices of food preparation, aimed at helping the
human body assimilate nutrients more easily. The author notes that it
was eventually discovered that foods that are appropriate for healthy
people are often harmful for those who are sick, and vice versa; and the
same goes for other aspects of daily life such as exercise – if this were not
the case, he says, the art of medicine would never have been sought (VM
5, 9–11). But that means that a method for discovering appropriate
treatments is more complicated than simply discovering which foods
are appropriate for a healthy person.
For this reason, matters are more complex and require greater precision
(ἀκριβείης). For a certain measure is needed at which to aim. But no other
measure, neither number nor weight, can be discovered, by reference to
which precision is achieved, than the perception of the body. Wherefore the
task is to learn with sufficient precision so as to err only slightly here and
there. Indeed, I would heap strong praise on the doctor who makes only
small mistakes. (VM 9, 19–25)11
The measure used in medicine cannot be an absolute one, since what
treatment is appropriate will depend on the condition of the particular
patient’s body – much as Aristotle would argue that determining the mean
10
Cf. VM 8, 9: αὕτη ἡ τέχνη πᾶσα ἡ ἰητρικὴ τῇ αὐτῇ ὁδῷ ζητεομένη εὑρίσκοιτο ἄν.
11
For a fine, detailed comparison of this passage with similar discussions of number, weight, and
measure in relation to goal-directed activity in Plato’s dialogues, see Festugière 1948, 41–43, note 41.
Perhaps the most interesting comparison drawn by Festugière is to the discussion of different forms
of knowledge in Philebus 55c-56e, where crafts achieve greater precision and are more scientific to the
extent that they make use of calculation, measure, and weighing (ἀριθμητικὴν . . . καὶ μετρητικὴν καὶ
στατικήν), in the absence of which there is nothing but “conjecture and the training of our senses
through experience and routine” (Frede 1993). Socrates goes on to say it is the latter that is typical of
medicine (56b1), but building, through the use of measures and instruments, achieves greater
precision and craftsmanship (b4-6). See Schiefsky 2005, 13–25 for a valuable discussion of the
concept of ἀκρίβεια in relation to τέχνη in this work and its connection to the discussion of these
concepts in the Philebus, as well as an interesting discussion of the varying attitudes in the
Hippocratic corpus toward natural inquiry and Socrates’ attitude as described in the Phaedo.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
with respect to justice in a particular case will depend on particular details
of that case. It is not that justice is subjective, nor that demanding precision
in deciding what to do is inappropriate; it is that acting justly demands
attention to context – and the precisely just action will vary from one
context to another. The author of On Ancient Medicine appears to be
arguing that the same will be true of medicine in determining the appro-
priate treatment to restore a patient to health.
3.5 Aristotle’s Use of μέθοδος

The concepts methodos and hodos in Plato’s dialogues and in On Ancient
Medicine build on the idea, going back at least to Parmenides, that inquiry
is a goal-oriented quest. The goal is knowledge, and its achievement
depends on following standards appropriate to that quest, and one task
of philosophy is articulating those standards or norms. A much-debated
issue, as one can see clearly in the Philebus and VM, is whether those norms
will vary from one discipline to another – whether, for example, the degree
of precision appropriate for one discipline is appropriate for another. As we
are about to see, Aristotle is deeply concerned with these issues and
articulates a unique and sophisticated approach to resolving them.
Aristotle may actually have composed a discourse on method. Early
in the Rhetoric, in the context of explaining that rhetoric makes use of
techniques that have their counterparts in dialectic, he remarks that:
It is also apparent that each form of rhetoric has its own good; for what has
been said in the discourses on method (ἐν τοῖς μεθοδικοῖς [λόγοις?]), applies
equally well here – some forms of rhetoric appeal to exemplars, others to
persuasive arguments, and the same goes for rhetoricians. (Rhetoric i.2,
1356b18-22)
And in its catalogue of Aristotle’s works, Diogenes Laertius’ Lives of the
Philosophers (D.L. v. 23) refers to Μεθοδικὰ α΄β΄γ΄δ΄ε΄ς΄ζ΄η΄.12 The possibil-
ity that this title is simply an alternative way of referring to the Topics is
ruled out by the above-quoted reference in the Rhetoric, since both the
Topics and the Analytics are explicitly referred to by name in the very same
passage. In any case, we can tell from the context the sort of thing that was
apparently discussed in this work: The use of exemplars and enthymeme in
rhetoric corresponds to the use of induction and deduction in dialectic.
12
There is also a reference to this work in Simplicius’ commentary on the Categories. Cf. Rose 1886,
xix, fr. 116, l. 11.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
3.5 Aristotle’s Use of μέθοδος 73
That is, this work apparently discussed, among other things, the ways in
which different disciplines proceed by different, though related methods –
in this case, different though related methods of argument. That is, such
a work was apparently focused on what I will refer to as ‘domain-specific
norms’ governing inquiry in different disciplines.
The word methodos appears with remarkable frequency in the
opening sentences of Aristotle’s major treatises, a fact that is rarely
noted, but I believe is quite important. Take the following nonex-
haustive set of examples, a few of which were discussed in Chapters 1
and 2:
1. Ph. i.1, 184a10-11. “Knowledge, and in particular scientific knowledge
about every methodos . . . ”
2. PA i.1, 639a1-2. “Regarding every study (theoria) and every methodos,
the more humble and the more valuable alike . . . ”
3. EN i.1, 1094a1-3. “Every craft (technê) and every methodos, and likewise
every action and decision (proairesis), seems to aim at some good.”
4. Mete. i.1, 338a25-26. “It remains for us to study a part of the same
methodos, which everyone prior to us has called meteorology . . . ”
5. Top. i.1, 100a18-20. “The theme proposed for this work is to discover
a methodos by which we will be able to reason from accepted opinions
about any proposed problem . . . ”
6. Ph. iii.1, 200b12-13: “Since nature is a source of motion and change,
and our methodos is about nature, we must not overlook the question,
‘What is motion?’”
One reason for the failure on the part of commentators to take note of
this potentially revealing fact about the language with which Aristotle
chooses to introduce so many investigations is suggested by the lack of
agreement on how methodos is to be translated. To take one example from
the above list, consider the following sample of translations of the
opening words of Ph. i.1, 184a10 (the relevant Greek phrase being περὶ
πάσας τὰς μεθόδους): “in all disciplines” (W. Charlton); “toutes les
recherches” (P. Pellegrin); “in any subject” (R. Waterfield); “in any
department” (Hardie and Gaye); “in every line of inquiry” (Irwin and
Fine, with an explanatory note); “in every inquiry” (R. Bolton). There
appears to be no general agreement among translators, judging from this
sample, about what Aristotle has in mind by the term in any particular
application. Even regarding a single usage, there is no agreement as to
whether the term refers to disciplines, lines of inquiry, researches, depart-
ments, or just subjects!
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
Part of the problem stems from the fact that Aristotle, in line with what
we have seen in Platonic dialogues and in On Ancient Medicine, deploys the
concept in two different, though related, ways:
(a) In one usage, the corresponding English word ‘method’ is a perfectly
acceptable translation, for Aristotle seems to be referring to procedures
that, if followed, will insure (or at least increase the likelihood of)
success in achieving the goal of an investigation – very much in the
spirit of the use of methodos in Plato and hodos in VM reviewed in the
previous two sections. This seems to be the sense carried, for example,
by the term in passage 5 in the list of opening texts, from the
beginning of the Topics.
(b) The term is also used with some frequency in a way that suggests that
the reference is to a subject of inquiry or area of study. Passage 4, from
the Mete., can be taken that way; and most of the translations of the
opening phrase in Ph. i.1 just reviewed take it that way, although they
differ about whether it refers specifically to an inquiry or to a field of
knowledge more generally. In this usage, the term is often conjoined
with terms having a similar or overlapping reference. So, for example,
in passage 2, the opening words of PA are “Concerning every study
and every methodos . . . ” (Περὶ πᾶσαν θεωρίαν τε καὶ μέθοδον . . .,
639a1); and the opening words of passage 3, from the EN, are “Every
art and every methodos, and similarly [every] activity and [every]
decision . . . ” (Πᾶσα τέχνη καὶ πᾶσα μέθοδος, ὁμοίως δὲ πρᾶξις τε
καὶ προαίρεσις . . ., 1094a1-2).
By looking at a wide range of occurrences and considering those that most
naturally appear to refer either to a course of investigation or to a way of
investigating, one can see how intimately these two modes of reference are
related to each other, just as ‘path’ and ‘journey’ are. Indeed, we will look at
a number of key texts where it is impossible (as we saw that it sometimes is
in Plato) to make a principled decision between the two modes of reference
as a guide to understanding the semantic field for this term. It is perhaps
telling, given his attention to the fact that a term can be “said in many
ways,” that Aristotle never suggests that this term has two distinct mean-
ings, nor does he comment on how these two different uses are related.
It is significant, I believe, that the term methodos never appears in the
APo., and only rarely, and in a highly delimited context, in the APr. At the
end of this chapter, I will be in a position to frame a hypothesis for this
initially surprising fact. But what is most striking, as I am about to
demonstrate, is that, in virtually every case, the word is used in the context
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
of a discussion of how one who is investigating a certain subject ought to
proceed with an inquiry into that specific subject.
3.5.1 Methodos and zêtêsis

In the passage I selected as the epigraph for Chapter 1, we find the
expression τίς ὁ σκοπὸς οὗ δεῖ τυγχάνειν τὴν ζήτησιν καὶ τὴν ὅλην
μέθοδον (Metaph. A.2, 983a22-23). That expression encapsulates two
closely related themes that will concern us in this and later chapters: (i)
Inquiry is a norm-guided and goal-directed process; and (ii) There is an
intimate relationship between Aristotle’s concepts of inquiry (zêtêsis) and
methodos which is, nevertheless, not synonymy. Through the remainder of
this chapter, we will survey the interplay between the concept methodos and
the idea that different inquiries must be conducted according to certain,
domain-specific norms if they are to achieve their goal.
In Chapter 4, we will look at Aristotle’s attempt to distinguish the
norms of natural inquiry from those in two other domains of theoretical
investigation, mathematics and metaphysics (or first philosophy), and we
will begin to investigate the relationship between Aristotle’s various
natural inquiries and their goal, scientific knowledge of nature. In Part
ii, we will turn to specific natural inquiries with a focus on two issues:
how are the general norms of natural inquiry related to (i) the domain-
neutral erotetic framework developed in APo. ii and (ii) the domain-
specific norms of distinct natural inquiries. The first question will be
a central task of Chapter 5, while Chapters 6, 7, and 10 will be focused on
the methodological recommendations that Aristotle makes for three
distinct domains of scientific investigation: the investigation of animals,
the investigation of the soul, and the investigation of respiration, one of
those processes ‘common to body and soul’ that are the focus of the so-
called Parva naturalia.13
There are, then, three features of the contexts in which Aristotle deploys
this concept that are of critical importance for this investigation:
1. The contexts are ones of inquiry.
2. They are normative.
3. The norms are domain-specific.
13
Chapters 8 and 9 will examine Aristotle’s deliberate practice of borrowing the results of certain
inquiries to aid in advancing others. It turns out that such borrowings are also constrained and
guided by epistemic norms.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
The remainder of this chapter is devoted to two tasks: first, specifying these
three contextual features in some detail; and second, examining a selection
of texts in which these features are prominently on display. I begin by
identifying in general terms, the three contextual features I aim to bring
into focus.
1. The Context of Inquiry. Suppose, just based on the term’s etymological
roots, we think of methodos as referring either (a) to a path leading to
a certain goal or (b) to an activity characterized by reference to its manner of
proceeding along a path toward a certain goal. One can then ask whether
Aristotle restricts his use of the term to a delimited class of endeavors or
procedures, or whether he uses it in a broad and indiscriminate way. In
some characteristic uses, including many of those in which it is used in the
introductory sentence of a treatise, it is hard to say. When, in the opening
lines of the EN, for example, Aristotle says “Every art and every
methodos . . ., ” it is difficult to determine what might or might not count
as a methodos, and whether technê and methodos are contrasting, comple-
mentary, or co-referential terms.14 But in the majority of cases, Aristotle
employs the term in order to identify (a) the manner in which an inquiry or
investigation is to be carried out or (b) the inquiry itself, looked at as a goal-
directed process, and thus it is often closely associated, as in that epigraph
from Metaph. Α, with terms such as zêtêsis (inquiry) or skepsis (investiga-
tion) and their associated verbs zêtein (to seek, to inquire) and skepsasthai
(to investigate). Indeed, the context is typically more specific than that: the
context is usually that of a question at issue about whether one ought to
inquire or investigate something according to a certain methodos. This
should make us skeptical of translating μέθοδος in a way that treats it
merely as a synonym for these other terms, as it often is.
2. Normativity. Implicit in that way of conceiving of the term’s semantic
field, however, is the thought that the concept methodos is often introduced
as part of an explicitly normative recommendation, in which the issue
under consideration is how an investigation or inquiry ought to proceed
or must proceed if a certain goal is to be achieved. Thus, we often find it in
association either with verbal adjectives (ending in -τέος) or impersonal
verbs (typically δεῖ) carrying the normative force of ‘should’ or ‘ought’. The
reference, then, is not just to a way of proceeding to a goal, but to the proper
or appropriate way of doing so. As we have seen, this normative dimension
14
One central purpose of this chapter is to have, by its conclusion, a number of concrete examples
before us of what he has in mind.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
is a feature of this concept’s use in Plato’s discussions of dialectic, and
especially division.
3. Domain Specificity. Finally, there are a number of passages that stress
that a particular kind of investigation, or a particular domain of investiga-
tion, requires a domain-specific norm or set of norms. In fact, though this is
harder to determine with certainty than the other two features, I will
defend the thesis that when methodos is used to refer to a domain or subject
of investigation, it is precisely because the question of what is distinctive
about the mode of inquiry in that domain is either in the background or
explicitly under consideration.
Let us now turn to the task of filling out the picture of these three
contextual features of Aristotle’s deployment of the concept of methodos.
Often, all three features are in play in the same passage, so that to attempt
to classify the passages according to which feature is in play is impossible.
Rather, I am going to select some passages that will begin to give us
a concrete sense of the philosophical work that this concept does for
Aristotle in different settings. I will begin with a quite detailed discussion
of one passage in On the Generation of Animals, and will then turn first to
a number of other examples from the context of natural inquiry and then
to a number of examples from other domains, including some that might
appear to be counterexamples.
3.5.2 GA ii.6: Thinking about Embryological Method

This passage is not at the beginning of a treatise, nor of a new book;
however, a critical new stage in the investigation of animal generation
does begin here, one that requires new methods of investigation and ways
of thinking. A summary of the argument up to this point will be helpful in
seeing why this is the case. Abstracting fairly dramatically from the
details,15 to this point the inquiry into animal generation has followed
a well-organized plan: in the first sixteen chapters of book i, Aristotle has
described and explained the instrumental, nonuniform parts (organs)
involved in reproduction of all the blooded and bloodless kinds of animals;
chapters 17–23 discuss the (fluid) uniform parts of males and females that
contribute to generation, and a theory of the material makeup and causal
contributions of each gradually emerges. These are parts, as he explicitly
15
More detailed discussions of the flow of argument of the first two books can be found in Bolton 1987,
120–166; and Gotthelf and Falcon 2018, in Falcon and Lefebvre, eds. 2018, 15–34. For a somewhat
different understanding, focused especially on book ii, see Leunissen 2018, also in Falcon and
Lefebvre, eds. 2018, 56–74.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
told us in PA ii and iv, about which he put off discussion until the study of
generation commenced. As book i draws to a close, Aristotle begins to
focus on the fact that, unlike most plants, in sexually reproducing animals
the contributions of male and female are provided by distinct organisms,
and that therefore they must engage in some form of coitus if reproduction
is to occur. Book ii opens with a sweeping explanation of why there is
reproductive generation at all, argues that the male and female principles
exist for the sake of generation, and explains why it is better that the male
principle, to which the logos16 and form belong, be separate from the female
principle, which is the material contribution to development. Over the
next four chapters, a complicated argument unfolds about the precise way
in which the capacity for, and source of changes that eventuate in, a new
organism, one-in-form with its parents, derive from the male parent and
are transmitted to the female material, which is at that point the potential
offspring. The male semen, as it turns out, makes no material contribution
to the process at all, but rather contributes only a special heat within its
pneuma to the female material (a portion of the menstrual blood in live
bearing animals), related to the nutritive soul of the male parent. This heat
is sometimes described as a power or potential (δύναμις) and sometimes as
a source of movement or change (ἀρχὴ τῆς κινήσεως)17 – its first product in
blooded animals is the heart,18 and once formed, the heart is the source of
heat and becomes the agent that directs development. Chapter 5 raises an
aporia, based on the fact that in many species eggs that have not been
fertilized can develop up to a point, about why there is need of a male
principle; this aporia is used to generate further discussion about the
different contributions of males and females to the generative process.
Though very much a sketch, this overview provides sufficient context for
the fresh start that occurs in chapter 6, one that requires a lengthy discus-
sion about the proper way to proceed from that point on. It is here that
Aristotle begins to discuss the coming to be of the parts of the developing
embryo, a discussion that continues to the end of book iii (and, in a sense,
to the end of the treatise).19
16
In this context, the term logos seems to refer to the plan or organization of development.
17
For examples of δύναμις used in this way, see 729b6, b28, 730a2, a14; for ἀρχὴ τῆς κινήσεως, see
729a10-11, 729b13-14, 28, 730a27-28.
18
“The source comes to be first. And this is the heart in the blooded animals, while in others its
analogue, as has been said repeatedly” (GA ii.5, 741b15-17).
19
For further discussion of the argument in this text, see Leunissen 2018, 69–74. Book iii, which
discusses development in egg-laying and grub-producing animals (and finally those generated
spontaneously), focuses most of its attention on the development of the egg and its parts (where
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
After a highly abstract discussion stressing the primary role of the heart
(ii.5, 741b18-24) and of pneuma in the actualization of the parts that are
first present potentially (ii.6, 741b37-742a16), Aristotle begins a long,
methodologically rich discussion of the question of the order in which
parts come to be. As is often the case, the discussion of methodology is
provoked by errors made by earlier natural investigators – they attempted
to say something about this question but “without being sufficiently
experienced with the facts.”20 Some parts are prior to others, certainly,
but “there are many ways of being prior.”21 There follows a long and
complex discussion of the relationship between what is prior in the order
of coming to be and what is prior in being, intertwined with a discussion of
the closely related distinction between what is or comes to be for the sake of
some goal and that for the sake of which it is coming to be (i.e., the goal).
We will need to look closely at the entire passage; but for our purposes, it
will be helpful to begin with its conclusion.
Wherefore it is not easy to distinguish which things are prior, those things
which are for the sake of something else, or that for the sake of which these
other things are. For the motive parts, being prior in generation to the end,
intrude, and it is not easy to distinguish the motive parts relative to the
instrumental parts. Yet it is necessary to inquire what comes to be after what
according to this methodos; for the end of some things is posterior, while the end
of other things is prior (καίτοι κατὰ ταύτην τὴν μέθοδον δεῖ ζητεῖν τί
γίγνεται μετὰ τί· τὸ γὰρ τέλος ἐνίων μὲν ὕστερον ἐνίων δὲ πρότερον).
(GA ii.6, 742b6-12)
The three contextual features that provide clues to the meaning of the
concept methodos that were introduced earlier – inquiry, normativity,
domain specificity – are all on display in the single clause “it is necessary
to inquire what comes to be after what according to this methodos.” First, the
particular methodos under discussion is a methodos of inquiry – we are
engaged in a specific sort of search. We are trying to understand the orderly
development of a complex sequence of stages in the continuous develop-
ment of a living thing, specifically a goal-directed sequence, one with
a τέλος. Aristotle is recommending that we seek or inquire after answers
to the question of what comes to be after what according to a certain
methodos. Second, the recommendation is explicitly normative – it is
eggs are involved) rather than on the development of the parts of the animal itself. I thank an
anonymous referee for insisting on this.
20
οὐ λίαν ἐμπειρικῶς ἔχοντες τῶν συμβαινόντων (742a17).
21
τὸ δὲ πρότερον ἤδη πολλαχῶς ἐστιν.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
necessary (or “one ought”) to inquire (or search; δεῖ ζητεῖν) in a specific way.
Why? To answer that question requires that we shift attention to the third
contextual feature of the use of methodos, domain specificity. There is an
important feature of biological coming to be, a feature that makes it
a distinctive kind of natural change and which, in Aristotle’s view, all of
his predecessors failed to appreciate: temporal order does not always reflect
priority in being, particularly in the case of goal-directed processes. Thus,
the distinctive nature of the subject matter under investigation and the fact
that it can easily confuse and mislead us in our search for causal under-
standing both require us to inquire according to a set of norms that are
specific to that subject matter.
In addition to these three features that govern the use of the concept
methodos here, there are also three different features of domain specificity,
discussed in chapters 1 and 2, that impact the proper way to carry out the
investigation: (a) differences in the nature of the objects and processes
being investigated, (b) differences in our epistemic access to those objects,
and (c) differences in our perspective on those objects. We can again see all
three features coming into play and leading to the need for a distinctive
mode of inquiry if we are to acquire scientific knowledge. I take up each
feature in turn.
(a) The nature of what is being investigated. The inquiry in which
Aristotle is about to engage is focused on the process of the coming to be
of an actual, ensouled being, consisting of multiple organs constituted out
of a multitude of different kinds of tissues, capable of performing
a coordinated set of activities constituting its way of life. The process
begins with a small bit of uniform fluid residue that has none of the
complexity or capabilities of the end product, and the continuous emer-
gence of that complexity typically takes place either inside a female organ-
ism’s womb or inside an egg. These facts about the subject matter of the
investigation place specific, distinctive demands on how the inquiry needs
to proceed, if scientific knowledge is to be achieved.
(b) Epistemic access. Aristotle believes that the search for knowledge is
fundamentally dependent on perceptual experience with the object of
inquiry. But there are at least three facts about biological development
that place limitations on our access to such experience. First, as noted
above, embryological development typically takes place within an enclos-
ure, which raises a question about how you will observe the developing
entity at all. Moreover, supposing you decide that such knowledge
demands somehow observing what is going on inside a womb or an egg
throughout the process of development, how does a researcher do that
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
without destroying the very thing to be observed? As anyone who has
performed dissection, and especially vivisection, knows, it takes great skill
and expertise to observe in such situations – observation is far from passive,
and the activities involved must be carried out with considerable skill if
they are to serve the aim of acquiring knowledge.22
Second, there is the problem of temporal unfolding, and how a human
observer deals with that problem. To concretize this point, consider how
Aristotle decided to solve this problem with birds: He selected a number of
chick eggs equivalent to the number of days from laying to hatching, and
opened one egg each day, comparing what he observed on each day with
what he had observed on the previous day.23 That is a methodological
move aimed at overcoming an epistemic limitation imposed by the fact
that the subject being investigated changes and develops slowly over a long
period of time – indeed, it is that very development that is the object of
inquiry.24 Third, there is the ‘priority’ problem on which Aristotle is
explicitly focused in the above passage: as you observe development,
different parts become visible that had not been earlier. But scientific
knowledge of such a process would require that you understand the causes
of the process unfolding as it is, why the parts are appearing where and
when they do, and in the way they do. And that takes us to the third
feature.
(c) Perspective. The most obvious way in which the methods involved in
this investigation owe their distinctive features to the perspective of the
researcher is the very fact that the focus is on coming to be. Aristotle
explicitly marks this as a distinctive investigation: one can focus inquiry on
the fully developed beings that animals are, as PA and IA do, and one can
focus inquiry on their coming to be, as GA does; and as we will see in
Chapter 6, these distinctive perspectives occasion one of the first meth-
odological questions Aristotle raises about animal investigation in PA i.1
(640a10-b3): which perspective should take priority in the investigation of
animals? As I will discuss in detail in that chapter, one of Aristotle’s most
consistent complaints about the investigations of his predecessors is that
they have accepted precisely the wrong answer to that question.
It is beyond the scope of this chapter to consider every aspect of what is
distinctive about the norms governing this sort of investigation, and how
22
For more on this topic, see Lennox 2018 in Falcon and Lefebvre, eds. 2018, 249–272.
23
The day-by-day development is described at HA vi.3, 561a4-562a22.
24
It is worth noting that in order to overcome the problem of temporal unfolding he also had to
overcome the observational problem – in this case opening the egg in such a way as not to destroy
the embryonic chick inside.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
the nature of the phenomena being investigated informs those norms. But
in order to help us visualize the ways in which local norms are, in one
direction, shaped by the broader erotetic framework provided by Aristotle’s
metaphysics and epistemology and, in the other direction, how they in
turn shape the actual investigative tools and methods of a specific inquiry,
it is instructive to look at the immediately preceding discussion that leads
into the above methodological detour.25
From the beginning of GA ii.6 down to 742a16, Aristotle has been
describing the earliest stages of embryological development at a very
abstract level, such that much of what he says applies to almost all
animals – in all animals except certain insects, the upper body begins to
develop before the lower (741b27-32); the distinction between upper and
lower is clear in all locomotive animals except the cephalopods (741b33-
37); pneuma of some sort plays a critical role in this process (741b37-
742a16). He then turns to the subject of the order in which parts develop –
and to understand the recommendations for inquiry that follow,
a lengthy quotation will be necessary. I will divide it into seven sections
for ease of discussion.
[1] Some of the early natural philosophers attempted to say which parts
come to be after which, without sufficient experience with the facts (οὐ λίαν
ἐμπειρικῶς ἔχοντες τῶν συμβαινόντων). [2] For with the parts, as too with
other things, one thing develops naturally (πέφυκεν)26 prior to another. But
there are actually many ways of being prior. [3] For that for the sake of which
and what is for the sake of it differ, and the latter is prior in its generation (τῇ
γενέσει), while the former is prior in its substantial being (τῇ οὐσίᾳ). [4]
Moreover, that which is for the sake of [the end] has two different senses; [i]
the source of motion, and [ii] what is used by that-for-the-sake-of-which.
I mean, for example, both that which is capable of generating and that
which has the instrumental capability in what is being generated; [5] for one
of these – that which is capable of producing – must be present first, for
example the teacher [must be present] prior to the one who is learning, and
the pipes [must be present] later than the person learning to play them; for it
is useless for pipes to belong to those who do not know how to play them.
[6] Since there are three beings [being distinguished] – [a] first, the end
we call ‘that for the sake of which,’ second, the motive and generative source
of those things for the sake of this [end] (for the productive and generative,
25
For an illuminating and more detailed discussion of this passage, see Leunissen 2018, 69–74.
26
Here I suppose the use of this (present sense) perfect form of the verb φύω is tied more closely than
usual to its semantic roots, where it refers to processes of budding, growth, and development. Cf.
743a21-23 where the warmth that produces flesh and bone during embryogenesis is said to produce,
not any chance part at any chance place or time, but τὸ πεφυκὸς, ᾗ πέφυκε, ὅτε πέφυκεν – that
which is natural, where it naturally belongs, at the naturally appropriate time.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
qua such, is relative to that which is produced and generated), and [c] third,
the useful, what is made use of by the end – [7] first it is necessary that there
be some part in which the source of motion is present (for straight off this
part is one part – and the most authoritative part – of the end), next after this
the whole and the end, and third and last the instrumental parts for these in
relation to various uses. (GA ii.6, 742a16-36)
Let us now consider this long and complicated passage from the standpoint
of its implications for the modes of inquiry appropriate to the study of the
development of the parts of animals.
[1] The passage begins by placing responsibility for the errors of his
predecessors concerning the order of the development of parts on a lack of
experience – we’ll come back to the question of what sort of experience
they were lacking, and what you would need to do to acquire that experi-
ence, later.
[2] Aristotle next insists that certain parts arise naturally prior to others,
but since priority is said in many ways – a central thesis of his metaphysics,
and in particular his account of substantial being – this claim needs to be
disambiguated.27 The ‘for’ (γὰρ) linking sections [1] and [2] suggests that
the right sort of experience would at least have helped those inquiring
before him to think more clearly about the natural order of the develop-
ment of parts. [3] Though not defended here, Aristotle next insists that the
final cause or goal of coming to be is prior in being to what happens for its
sake, while what happens for the sake of the goal is prior in generation,
temporally prior. To fully understand Aristotle’s defense of this claim, we
would need to dig into his argument for the priority, in both account and
being, of actuality (ἐνέργεια) over potentiality (δύναμις) in Metaphysics
Θ.8. Here I will extract only one thread from that argument (1050a3-15),28
one that provides the metaphysical grounding for our GA ii.6 passage.29
This is a clear case where Aristotle’s distinctive metaphysics provides part of
the erotetic framework for investigations in which the object of inquiry is
a substantial being, which is the goal of a continuous process of generation.
27
See, for example, Metaph. Δ.11; Θ.8; and Γ.2, 1005a14-18, which lists priority and posteriority among
attributes to be studied by the science of being qua being.
28
For contrasting accounts of this argument see the essays by Charles and Broadie in Lennox and
Bolton, eds. 2010, chapters 8 and 9; and for arguments on how to understand this notion of priority
in being, see Peramatzis 2011, 278–286; Beere 2009, 285–304; and Witt 1994, in Scaltsas, Charles, and
Gill 1994, 215–228.
29
There is in fact Platonic background as well: twice in the biological works (PA i.1, 640a15-19 and GA
v.1, 778b2-7, which is apparently referring back to the PA i.1 passage) Aristotle echoes (though
without explicit acknowledgment) Phlb. 54a5-11, c9-11, stating that “generation is for the sake of
being, rather than being for the sake of generation.”
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
But [actuality is prior] in being (οὐσίᾳ) as well, first of all because what is
later in generation is prior in being . . . for example, father to child and
human being to seed, for the one already has form while the other does not;
and second, because in every case that which comes to be proceeds to an
origin and goal (ἐπ’ ἀρχὴν . . . καὶ τέλος); [I say ‘origin’] because that-for-the
-sake-of-which is an origin, and generation is for the sake of the goal. And
the actuality is a goal, and the potentiality is acquired for the sake of this.
(Metaph. Θ.8, 1050a4-10)
I take the first argument for the priority in being of the actual to be parasitic
on the second. It assumes that the adult is prior in being to the seed or
child, but that assumption depends on the causal – specifically final causal –
priority of the actual over the potential, which is only defended in
the second argument. Generation is for the sake of its goal, and that goal
is the actuality – the full realization – of that which is in the process of
coming to be. But (a point made via a different but closely related line of
reasoning) this is also an origin, for the goal of coming to be is that for the
sake of which generation occurs, and that for the sake of which is an
origin.30 It is unclear whether archê in this passage refers to a starting point
of demonstration or to a causal source. But that may be a distinction
without a difference in this context, since Aristotle holds that identifying
the goal of a generation in fundamental terms is to grasp both causal and
definitional starting points.31 Addressing the question of causal priority in
PA i.1, for example, his initial response is as follows:
[I]t is apparent that the primary cause is the one we refer to as ‘that for the
sake of which’; for this is an account (λόγος), and the account is an origin
(ἀρχή) alike in things composed according to art and in things composed by
nature. (PA i.1, 639b18-19)
The logic is impeccable: The final cause is the account, the account is the
origin, ergo the final cause is the origin. But what grounds does he provide
for believing the premises? He defends the claim that the account is also an
origin or starting point by noting that craftsmen begin with definitions
(i.e., accounts) of the goal of their activity and then “provide the accounts
and the causes of each of the things they produce, and [provide] the reason
why it must be produced that way” (639b17-18). In this passage, then, the
30
The last few lines in this passage are essential, since in Greek this claim would be as paradoxical as it
would be in English if I were to say “the end is a beginning.” I’ve softened the paradoxical feel by
translating archê as ‘origin,’ but the word can be used for a first principle in mathematics, a primary
thing, the beginning of a process, and an authority (the masculine nouns ἀρχός and ἄρχων refer to
military and political leaders).
31
Recall, this is a critical feature of the erotetic framework of the Posterior Analytics.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
definition of the goal object is an archê in the sense of where we begin in
providing an explanation of the process by which it is produced. Aristotle
says nothing here, however, about how that translates to the goal of
a natural process being the explanatory starting point of a natural process
of coming to be.
[4] The next step in our GA ii.6 passage is to introduce a distinction
between two quite different items in the category of ‘things that are for the
sake of something’: [i] sources of change or generation and [ii] things that
are made use of by the goal of generation. The items in the second category
are not present for the sake of the generator, qua generator, but for the sake
of the end or goal of the generation. This follows quite directly from the
fact that both these useful items and the generator are in the category of
things present for the sake of the end, but in very different ways – and
Aristotle is insisting that it is critical that the embryologist recognizes that
difference.
To get a clearer view of this distinction, it will help to recall where we are
in the argument of GA. The male provides only a special power or source of
generation, in the form of an ‘inborn’ or ‘psychic’ warmth, conveyed by the
pneuma of the semen to the female material. Once that warmth is present
within the material provided by the female, it is the primary goal-directed
source of production or generation – and its first task is to produce the
heart, which then begins to produce a network of blood vessels and to
convert material into nutritive blood, which is transported through this
network for the production of parts. This is all so carefully orchestrated
that Aristotle feels justified in assuming that this warmth instantiates
a logos – a ‘program’ we might anachronistically say – so that all the changes
are the right changes and take place at the right places and times so that, in
the normal course of events, the result is an organism formally identical
with its parents.32 This special heat bears an ‘irreducible (active) potential
for form,’ to use Allan Gotthelf’s expression.33 Moreover, ordinary heating
and cooling are causally involved in this process as ‘tools’ used by the
generative power or nature (cf. ii.1, 734b19-735a4, ii.4, 740b12-741a3, ii.6,
743a36-b5).34 So the generative source is acting directly for the sake of the
32
Cf. GA ii.6 743a18-27.
33
For the details see Gotthelf 2012 14–19, 90–100. The potential is ‘irreducible’ in the sense that the
goal-directed process is not merely the result of interactions between the four potentials (hot, cold,
moist, and dry) of the elements that materially constitute the developing organism. Notice that in
the passage quoted in the next footnote, Aristotle equates this productive capacity with a capacity of
the nutritive soul, which is “also that which generates.”
34
Here is one such passage, arguing that the capacity of the nutritive soul responsible for growth is also
responsible for generation and uses ordinary heating and cooling as tools, insuring that they produce
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
ensouled organism to be (and indeed in one sense it is the ensouled
organism to be, 742b2-3), and the nutrients, heatings and coolings, hard-
enings and softenings, are used instrumentally for the same end.
[5] Based on this distinction between generator and instruments for
generation, Aristotle argues that it is necessary for the generative power to
be present before (i.e., to be temporally prior to) what is useful, since if the
latter were present first it would in effect be useless. Even with the
assumption in the background, as it surely is, that nature does nothing in
vain, this is a questionable conclusion – why could materials to be used
later not be present before they are put to use? Indeed, is that not the case
with the menstrual discharge that is prepared for generation prior to the
generative power being transmitted to it via the male semen? Aristotle may
well be looking toward a view he argues for later, that certain parts are
formed out of residues of prior (and primary) formations, and certainly
these residues will not be available until those prior formations have
occurred (744b21-28); but there is no hint of that argument here.
Section [6] summarizes the argument to this point by distinguishing
three beings (ὄντα) – the goal for the sake of which the development is
occurring, the generator, and things useful for achieving the end. The final
section, section [7], draws an inference about the priority relationships
among parts from the previous discussion:
1. There must be some part in which the source of generative or product-
ive motion is present.
2. This part is the most authoritative part of the goal of development.
3. Next in order of priority is ‘the whole and the end,’ presumably the
whole, actual animal, which is the goal for the sake of which gener-
ation is taking place.
4. Third in order of priority are the instrumental parts serving these in
relation to various uses.35
according to a determinate program: “in this way the capacity of the nutritive soul, just as it produces
growth from the food in the animals and plants themselves later on, using as tools heat and cold (for its
motion is in these, and each thing comes to be by a certain logos), so too that which comes to be by
nature is constituted in this way from the beginning. For it is the same matter by which it grows and
from which it is first constituted, so that the productive capacity <for growth> is also the same as the
initial one (though greater in strength). So if this is the nutritive soul, it is also that which generates
(740b30-37; Gotthelf translation with modifications; emphasis added).
35
It is unclear both what the reference of “these” (τούτοις) is and how to interpret the dative case
ending. Peck 1942, Platt 1912, and Lanza & Vegetti 1971 all translate it as referring to the two previous
items in the list. Peck (1942, 213, note d) has a footnote that shows a concern about the plural,
perhaps because Aristotle had previously (742a32-33) connected the useful only to the end.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
What then follows is the concluding text, with which we began. Having
these distinctions in hand is critical, but careful observation of the early
stages of development teaches the experienced anatomist just how hard it is
to apply them in practice. The most authoritative part – the heart or its
analogue – qua generator must be formed first, and yet throughout the
animal’s life it is the central organ of perception, locomotion, and nutrition
(i.e., of the soul) – and so it comes to be as part of the goal and “along with
the whole” (μετὰ τοῦ ὅλου, 742b3). So those instrumental parts that are
generative in nature, while they are for the sake of the goal, must be present
prior to others that are also for the sake of the end, but which are not
generative.
Now immediately after insisting that despite these difficulties we must
investigate the development of the embryo according to this methodos,
Aristotle states, as if by way of explication:
For of some [parts] the end is posterior, while of other [parts] it is prior.36
The point appears to be this: the goal of some things occurs later (in
generation) than them, while the goal of other things occurs earlier (in
generation). For example, the goal of lungs is breathing, and breathing
begins to take place sometime after the lungs are formed; while the goal of
blood vessels is to serve as the container and conveyor of blood and thus to
subserve the function of the heart, yet the heart and blood come to be prior
to the blood vessels. Why is this? Aristotle has given us the tools for an
explanation in the passage we have been discussing – in biological gener-
ation, there is a complex ontological/causal relationship between the end
for the sake of which and two distinct kinds of entities that come to be and
are present for the sake of the end. The heart, as the part in which resides
the generative capacity and in which the blood is concocted into the
primary nutrition for all the parts, must be present before parts that are
present in order to subserve the nutritive (and other) functions of the heart
and blood. So indeed, this brief remark does explicate why, in gaining
knowledge of animal generation, it is critical that we keep these different
kinds of priority relationships clearly in view.
This complex teleological conception of coming to be, then, grounds
a set of norms that must govern a natural scientist’s thinking – and acting –
in coming to understand “what comes to be after what” during animal
36
τὸ γὰρ τέλος ἐνίων μὲν ὕστερον ἐνίων δὲ πρότερον. Peck translates: “For the End, though it comes
after some of them, is prior to others.” But so translated, it is unclear what the antecedent of “them”
and “others” is.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
generation. What are these norms? For to a great degree to characterize
which methodos is to govern such an inquiry is a matter of identifying the
norms specific to it.
1. Because animal generation is a single, goal-oriented process, priority
must be given to understanding the goal of the process; for it is because
birds must have beaks while cats must have teeth that the development
of the mouths of birds and cats takes such a different course.
2. This norm is grounded in a fundamentally metaphysical principle, that
there is more than one mode of priority – and in particular, priority in
being should not be equated with priority in the order of coming to be.
3. As a consequence, Aristotle gives causal priority in the investigation of
animals to the final cause over the efficient cause (PA i.1, 639b14-16).37
4. This requires that the investigator first distinguish the end for the sake
of which the generative process is taking place from the efficient causes
directed to that end; and then between two sorts of efficient cause that
are sources of the generative process and that are made use of in
achieving the goal.
5. Finally, this has implications about how to prioritize the parts that are
coming to be, the actual phenomena under investigation at this
point. The part that is the source of the generative process must
both be present first in the order of generation and at the same time
be the most authoritative part of the goal of that process. Aristotle
ranks the whole, actual animal, second. This at first seems odd, since
this in one sense is the goal of the entire process. But Aristotle is
discussing the coming to be of the parts here – what he has in mind,
I suppose, is all the other parts that constitute the actual animal, all of
which come to be through the actions of that logos-bearing heat
residing primarily in the heart and distributed, through the network
of blood vessels, in the process of constituting the other parts. Third
in order of priority are parts that are made use of by the heart in the
process of producing the parts out of which the body of the animal is
constructed. If my understanding of the items in the second priority
ranking is correct, the parts ranked third would include blood, blood
vessels, perhaps all of the uniform parts used in the construction of
organs, and perhaps the pneuma that conveyed the heat that initiated
the processes.
37
On which see Code 1997, esp. 134–136; Charles 2012, 227–266 (in Shields, ed. 2012); and the
upcoming discussion in Chapter 6.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
And this leads us back to what Aristotle had in mind in saying that earlier
natural investigators attempted to discuss such matters without sufficient
experience of the facts. Acquiring sufficient experience with the facts about
the generation of animals is extremely difficult. It either takes place inside
an egg or inside a womb (or in certain cases in both). Thus to acquire
experience of such things one must first acquire a certain kind of expertise in
dissection and in peeling the shells off developing eggs – and in looking at
what is thus exposed with a philosophically trained eye. Knowledge of the
functional relationships among the parts of fully developed animals must
be acquired prior to studying embryological development, because what
you are seeing only makes sense in light of understanding the teleological
relationships that exist among the various parts.38 Thus one of the norms of
inquiry that is a central theme of PA i.1, and which will be discussed in
Chapter 6, is critically important to the methodos being insisted upon
here – one must study the way of life of the actual, fully developed
organism performing its vital functions prior to studying its generation.
Only by doing so will you understand the complex order in which the parts
actually come to be. It looks as if the inquiry into the coming to be of the
parts of animals can only be carried out productively once there is
a somewhat advanced understanding of the goal-directed activities and
ways of life of actual organisms.
3.5.3 Methodos and Paideia

In the opening paragraph of PA i, a passage we will look at in more detail in
Chapter 6, Aristotle declares that regarding every theoria and methodos,
however noble or humble, there are two distinct ways of being cognitively
disposed: scientific knowledge of the subject (τὴν ἐπιστήμην τοῦ
πράγματος) and a state that is some sort of discernment (τὴν οἷον
παιδείαν τινά)39 (639a1-5). I here only want to draw attention to
38
On the relationship between this kind of experience and the knowledge of the craftsman, see Charles
2000, 151–161.
39
This phrase has been variously translated: Balme 1992, “as it were a sort of educatedness”; Lennox
2001a, “a certain sort of educatedness”; Lanza and Vegetti 1971, “una sorta di cultura”; Ogle 1882, “a
kind of educational acquaintance”; Peck 1961, “what is roughly described as an educated person’s
competence.” There are two distinct problems: first, there is no word in English that captures all of
the nuances of the Greek word παιδεία; second, the above translations show the various options of
combining the relative οἷον with the enclitic indefinite τινά. Balme (“as it were”) and Peck (“roughly
described”) take it adverbially (LSJ s.v. οἷος v.2.d); Lennox, Vegetti, and Ogle take it as a general
relative with τινά, indicating an unspecified type of παιδεία. Given the meaning specified here and
in related passages throughout the corpus, the adverbial rendering is inappropriate: it is a form of
genuine discernment that Aristotle has in mind, which some people have regardless of the domain
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
a number of passages that are suggestive of a very close connection between
the sort of ability to which Aristotle is referring here and being in posses-
sion of a domain-specific methodos – the full case for this connection will be
presented in Chapter 6.
In broad outline, what Aristotle goes on to say in the opening of PA i.1 is
that being in possession of this sort of paideia – discernment – gives one the
ability to discriminate between what is well presented on a given topic and
what is not, and he further distinguishes between those who have a general
discriminative ability of this sort and those who have it concerning
a determinate nature – and in particular, he is insisting that those who
are in possession of this sort of ability with respect to the investigation of
nature (τῆς περὶ φύσιν ἱστορίας) will possess certain standards (τινὰς ὅρους
τοιούτους) in accordance with which natural inquiries and explanations
should be carried out (639a5-16).
Those who are familiar with Aristotle’s practical philosophy will no doubt
hear echoes in this discussion of the πεπαιδευμένος in possession of the
appropriate standards of judgement for a particular field in Aristotle’s ethical
and political writings. Early in EN i, for example, a person with this discrim-
inative sort of ability seeks the right degree of precision (akribeia) for each kind
under investigation and does not demand mere persuasive arguments in
mathematics nor for demonstration in rhetoric (i.3, 1094b24-28). In that
same passage Aristotle distinguishes between a person with a general form of
this discriminative ability and one having it with respect to a particular subject
matter (1094b28-1095a2). Similarily, in the Politics, using medicine as an
example, we find a related distinction we’ll have reason to revisit in Chapter 6.
Hence just as a court of physicians must judge the work of a physician, so also all
other practitioners ought to be called to account before their fellows. But
‘physician’ means both the ordinary practitioner, the master of the craft, and
thirdly, the person of discernment (ὁ πεπαιδευμένος) about the craft (for in
almost all the arts there are some such people, and we assign the right of
judgment (τὸ κρίνειν) just as much to people of discernment (τοῖς
πεπαιδευμένοις) as to those with knowledge (τοῖς εἰδόσιν)). (Pol. iii.11,
1282a1-7)
Here too, in the concluding sentence, we get a contrast between the person
with paideia and one with knowledge of the subject, but in a way that
makes clear that a person can have the general form of paideia without
having what would count as knowledge of the subject.
under consideration, and others have about a specific subject. Thanks to an anonymous reader for
pressing me about the meaning of this phrase.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
3.6 Methodos across the Corpus Aristotelicum 91
And while the word is not used, surely in the following passages from Ph.
viii.3, Aristotle is describing a natural investigator who lacks the sort of
discriminative expertise that, in the previously discussed passages, he refers
to as paideia.
But to investigate about this [whether nothing moves] and to seek an
account of things which we know better than to need an account, is to be
a poor judge of better and worse, of what is plausible and implausible, of
what is a starting point and not a starting point. . . . In relation to all these
views, conviction is enough – we see that some things are at one time moving
and at another time are at rest. (Ph. viii.3, 254a30-b1)
What is described here might be called a meta-norm of inquiry: don’t
engage in an inquiry into what is so obvious to perception that no further
inquiry is necessary! But in any case, this special form of discriminative
ability that is sometimes referred to as a special sort of paideia is distinct
from having knowledge of the subject matter, but also appears to be
intimately related to possession of the norms needed for successful inquiry
leading to knowledge.
3.6 Methodos across the Corpus Aristotelicum

The word methodos appears eighty-two times in the Corpus Aristotelicum,
almost all in works of unquestioned authenticity.40 The largest concentra-
tion is in the Topics (Top.) and Sophistici Elenchi (SE), which together
account for nineteen occurrences; the next highest concentration is in the
Politics (ten). In the remainder of the corpus the word tends to appear, as
we have seen, at the beginning of the discussion of a new subject – and, as
we saw in Section 3.4, in the very first sentences of many treatises, for
reasons that, at least in broad outlines, should now be clear.41 In this
concluding section, I will survey the use of the term methodos across the
Aristotelian corpus generally, drawing attention to those features of its use
outlined previously in Section 3.5.1: that it typically appears in discussions
of how an inquiry or investigation is to be carried out, specifically discus-
sions of the norms that should govern an inquiry, and in contexts where the
norms being discussed or recommended are specific to a certain domain of
40
There is one occurrence in the Physiognomica, one in a fragment attributed to a work called On
Divisions, and three in fragments of On Philosophy (which also contains a reference to the Μεθοδικά
that, as we’ve already seen, is referred to in the Rhetoric).
41
It is not always clear that what our corpus treats as a single treatise was so considered by Aristotle.
A good example, to be discussed in the Chapter 4, is the occurrence of μέθοδος at the beginning of
Physics iii.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
study, that is, based on distinctive features of the particular subject matter
being investigated, or of our access to it or the distinctive perspective being
taken toward it, as the grounds for the norms that are being recommended.
3.6.1 Methodos and Ιnquiry in the Topics and Sophistical Refutations

The fact that there are so many occurrences of the term methodos in a work
devoted to formalizing techniques to be used in the context of dialectic
might seem to be a challenge to the argument I’ve been making to this
point. Dialectic, after all, is supposed to be, in at least one sense, domain-
neutral – the techniques of argument and counterargument are designed to
work regardless of what subjects are under dispute. However, when one
looks at the uses of the concept in the Topics, they fall into two broadly
defined categories: (i) references to modes of questioning or arguing that
are appropriate to dialectic because of features unique to it; and (ii)
references to the use of techniques of argumentation in a domain where
they are inappropriate.
A nice example of this second use appears in Top. viii.12, in which
Aristotle is laying out four different ways in which a logos may be said to be
‘false.’ He describes the third in this way:
. . . or [the argument] comes to a conclusion related to what was proposed,
but yet not in accordance with the appropriate method (μέθοδος).
[Examples of] this are, when [the argument] though not medical, seems to
be medical, or not being geometrical [it seems to be] geometrical, or not
being dialectical [it seems to be] dialectical, both in cases where the result is
false and where it is true. (162b7-12)
Here the term methodos is used, modified by oikeia, to designate a mode of
inquiry with standards appropriate to a certain specific domain – and the
fallacy is using that mode of inquiry in a context where that mode is
inappropriate. It does not matter whether the resulting conclusion turns
out to be true or false – it was not established in the appropriate way.42
The vast majority of occurrences of methodos in the Topics appear in the first
five chapters, and in many of those occurrences we see Aristotle identifying his
goal of finding, as he says in the treatise’s very first sentence, “a methodos with
which we shall be able to construct deductions from acceptable premises
concerning any problem that is proposed” (100a18-20). So, the problems
proposed for debate and questioning can be on any topic you like – but
42
For similar uses in Top. and SE, see Top. i.6, 102b36-103a1; SE 11, 171b11-22.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
that does not mean dialectic does not have a domain-specific set of standards
and norms of its own. In the inquiry the above sentence initiates, Aristotle is
searching for, and aiming to systematize, modes of proposing questions and
answers related to problems that are based on endoxa, commonly held beliefs
either in general or at least among ‘the wise.’ Almost immediately he distin-
guishes between demonstrations from premises that are true and primary and
dialectical deductions from endoxa, and begins then to distinguish various
fallacious forms of argument, including those that are specific to certain
sciences. That is, the norms and standards that are going to govern inquiry
in dialectic are distinct from those in demonstrative sciences and may include
being familiar with fallacious arguments that depend on premises that, though
false, are of a form appropriate to a certain domain, such as geometry. In
chapter 2, Aristotle discusses the various ways in which the inquiry he is
engaged in may be useful (101a25-b3): mental exercise (πρὸς γυμνασίαν),
encounters (πρὸς τὰς ἐντεύξεις), and the philosophical sciences (πρὸς τὰς
κατὰ φιλοσοφίαν ἐπιστήμας); and methodos appears twice in characterizing
the first and third uses. Regarding the first, we are told that participants in
dialectical disputation will be able to challenge a proposed subject more easily
when they have a methodos (μέθοδον ἔχοντες, 101a29); that is, having
a systematic set of the techniques appropriate to dialectical disputation will
improve one’s ability as a dialectician. The sense of methodos here is identical
to that in the treatise’s opening sentence.43
The second appearance of the term is in the very last line of the chapter
and is of special interest for two distinct reasons. First, Aristotle connects it
directly to hodos, which helps us to explore differences in the connotations
of these two related expressions; second, methodos here appears to refer not
to modes of specific inquiries but to the inquiries themselves.
Furthermore, [the methodos we are seeking] is useful in connection with the
first of the principles concerning each science. For starting from the prin-
ciples that are appropriate for a given science, it is impossible to say anything
about those principles (since these principles are the first of all); and it is
through the accepted opinions (διὰ τῶν ἐνδόξων) about each that it is
necessary to discuss them. But this is either a distinctive or a most appropri-
ate task of dialectic; for being fitted for examination, [dialectic] possesses
a path toward44 the principles of all the methodical inquiries (πρὸς τὰς
ἁπασῶν τῶν μεθόδων ἀρχὰς ὁδόν ἔχει). (Top. i.3, 101a36-b4)
43
Cf. Top. i.3, 101b5 (the first line of ch. 3) and Top. i.4, 101b11 (the first line of ch. 4) for other
examples of the term used in this way.
44
The word rendered here as ‘toward,’ πρὸς, can also be translated ‘in relation to.’ On the controver-
sies surrounding these sentences, see the judicious remarks in Smith 1997, 52–54.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
This passage begins by raising a problem about how one can investigate the
first principles of a science – such an investigation cannot be carried out
starting from the principles of that science, since it is those very principles
that are being investigated. But since dialectic does not start from prin-
ciples appropriate to specific sciences, it is suitable for doing so. In fact,
Aristotle identifies what makes it distinctively useful for this purpose – it is
‘fitted for examination’ (ἐξεταστική). It is on the basis of this characteriza-
tion of dialectic that he claims it is a path or a route toward the principles –
but note, not the first of the principles concerning each science τὰ πρῶτα
τῶν περὶ ἑκάστην ἐπιστήμην ἀρχῶν, 101a37) but those of all the methodoi,
all the norm-governed inquiries.
For now, the directly relevant fact about the use of methodos in this passage
is that rather than referring to the methods or ways of inquiry of dialectic
that the project of the Topics is aiming to discover (as in the treatise’s opening
lines), here methodos is referring to the various domain-specific inquiries with
their own distinctive starting points. However, the move in this passage from
discussing the principles of specific sciences to referring to the principles of
specific methodoi raises an important question about how Aristotle under-
stands the relationship between the concepts of ἐπιστήμη and methodos.
That question will be explored in Chapter 6, in our discussion of PA i.1 and
what Aristotle refers to as the methodos of the animals.
3.6.2 Methodos and Ιnquiry in Aristotle’s Practical Philosophy

Perhaps more than any other subject area, Aristotle’s ethical treatises have
been the focus of significant debates about his method or methods of
investigation. Typically, however, those discussions do not consider
Aristotle’s use of the concept of methodos.45 Doing so sheds considerable
light on Aristotle’s views about how an inquiry into human good and
human excellence should proceed, and what norms ought to govern that
inquiry. Looking at a selection of these passages reveals how differences not
only in subject matter but in the purposes for which an inquiry is pursued
decisively impact the norms that shape it.
A good place to begin is Eudemian Ethics (EE) i.6, 1216b35-1217a10.46 In
this passage, Aristotle distinguishes philosophical and nonphilosophical
45
Allan 1961; Anagnostopoulos 2009b; Barnes 1980; Burnyeat 1980; Irwin 1981, 1988; Owen 1961;
Kraut 2006b; Reeve 2012; Salmieri 2009; Whiting 2001. A recent exception is Karbowski 2019, esp.
88–106.
46
Joe Karbowski first drew this passage to my attention and I have benefited from numerous
conversations with him about it. On this passage in particular, see Karbowski 2019, 110–120.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
modes of argument “concerning each methodos.” The philosophical mode
is aimed at grasping ‘the what’ (τὸ τί) and ‘the reason why’ (τὸ διὰ τί), and
Aristotle insists that even in political inquiry, the philosophical mode is to
be preferred. Aristotle goes on to warn, however, that there are some in this
domain who “make arguments that are foreign (ἀλλοτρίους) to the inquiry
and empty,” and thus even people experienced in political matters can be
deceived through lack of discernment (ἀπαιδευσία):
for it is lack of discernment to be unable to distinguish, in regard to each
subject, between those arguments that are appropriate (οἰκείους) to it and
those that are foreign (ἀλλοτρίους). (EE i.6,1217a4-10)
This passage stresses the domain-specific nature of the arguments in each
distinct methodos. There are arguments, Aristotle tells us, that are appropriate
(oikeios) to specific, norm-governed inquiries and those that are inappropriate
or ‘foreign’ (allotrios), and being properly educated in a subject is having
acquired standards that allow a researcher or student to tell the difference. It
also stresses that it is a lack of proper training or education in a given domain of
inquiry that leaves people unable to judge whether an argument is appropriate
to it or not. This is the same connection between domain-specific paideia and
the proper methodos for a specific inquiry that we found in the opening
paragraph of PA i.1 and which will be examined in detail in Chapter 6.
There are objective, though contextual, criteria that determine the appropri-
ateness of a logos to the subject matter being investigated – and that is a concern
with many dimensions. The inability to make such judgments is here attrib-
uted to a lack of discernment (ἀπαιδευσία); the ability to do so is, in PA i, said
to be “a certain sort of discernment” (οἷον παιδείαν τινά), an acquired state
(ἕξις) distinct from having knowledge (ἐπιστήμη) of the subject (639a1-5) but
which appeals to standards in discriminating what is well and poorly said.
EE I.6 also makes an explicit reference to the general erotetic framework
that I have claimed is common to all inquiries aimed at knowledge, and which
I see as the primary achievement of APo. ii. Aristotle insists that even when the
subject is political affairs, if one is approaching the subject philosophically, one
will seek knowledge not only of ‘the what’ but also of ‘the reason why.’
Answering both the ‘what is it’ and ‘why’ questions is the philosophical
approach in every methodos – the term here referring to a domain-specific
inquiry carried out by appropriate standards and methods.47
47
With this passage compare Pol. iii.8, 1279b12-15: καὶ γὰρ ἔχει τινὰς ἀπορίας, τῷ δὲ περὶ ἑκάστην
μέθοδον φιλοσοφοῦντι καὶ μὴ μόνον ἀποβλέποντι πρὸς τὸ πράττειν οἰκεῖόν ἐστι τὸ μὴ παρορᾶν
μηδέ τι καταλείπειν, ἀλλὰ δηλοῦν τὴν περὶ ἕκαστον ἀλήθειαν.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
One important distinction between inquiries aimed at providing guid-
ance for practical affairs and more purely theoretical inquiries is stressed in
the introductory chapter of EE:
There are many studies which present difficulties about each thing and each
nature and which need investigation – some of these studies extend only to
knowledge, but others are concerned with the acquisitions and actions of the
thing. Now we should speak about those subjects that are the concern of
theoretical philosophy alone on the appropriate occasion, and about what-
ever is proper to the methodos. But first we ought to examine in what living
well consists and how to achieve it. (EE I.1, 1214a8-15)
This passage divides inquiries broadly into those directed to knowledge
alone and those also concerned with acquisition and practical affairs. He
sets aside the first category for another time, at which point we must discuss
it in whichever way is “proper to the inquiry” (οἰκεῖον . . . τῇ μεθόδῳ,
1214a14). Aristotle likely chose this wording to emphasize that there is
a distinctive mode of inquiry appropriate to theoretical philosophy.
While the referent might well be to a ‘discipline,’ οἰκεῖον . . . τῇ μεθόδῳ
emphasizes once again the need for a mode of inquiry appropriate to it.48
The uses of methodos in the Politics are, as I noted earlier, numerous, but
a number of them (iv.2, 1289a26-29, iv.10, 1295a1-7, vi.2, 1317b33-34) are
simply references back to earlier methodological remarks. There are two
(i.1, 1252a18, ii.1, 1260b36), however, that are especially important for the
project of this book – and it is not surprising that they are found near the
beginning of books i and ii. Three other passages (iii.8, 1279b12-15, vii.1,
1324a1-3, vii.2, 1324a19-24) are of value in that they reinforce themes that
we have already explored in the use of methodos in the ethical treatises.
Let us begin by looking at Pol. i.1, 1252a17-23.
What I am saying will be clear by investigating according to the prescribed
methodos (κατὰ τὴν ὑφηγημένην μέθοδον).49 For just as in other investiga-
tions it is necessary to divide the composite until one reaches the incompo-
sites (these are the least parts of the whole), so too in examining that out of
which a polis is constituted – we shall see more clearly both how they differ
from each other and whether it is possible to acquire some expertise about
each of the things mentioned.
48
The Solomon translation in the Revised Oxford Translation, “appropriate to that study,” has more
of the flavor I am suggesting. For similar uses of μέθοδος in Aristotle’s ethical treatises, see EN i.1,
1094a1-3; i.2, 1094b9-12; i.7, 1098a25-29; v.1, 1129a3-6.
49
This expression, and one with a virtually identical sense, κατὰ τὸν ὑφηγημένον τρόπον (cf. Mete. i.1,
339a6-7), appear a number of times in the corpus. See, for example, GA iii.9, 758a28-29 (discussed in
Leunissen 2018, 62) and my discussion in Chapter 4.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
3.7 Conclusion 97
What is referred to as “the prescribed method” here is one that is relatively
domain-neutral – when dealing with a composite thing, one must divide it
into its least parts. There are similar recommendations for natural inquiry
at Ph. i.1, 184a21-b3 and for investigating poetry at Poetics i.1, 1447a8-14,
and it is a norm put into practice a number of times in the Politics (see iv.2,
1289b12-26, and iv.4, 1290b21-38 – where the method is explicitly modeled
on the multi-axis method of division outlined and defended in PA i.3).
When all of these passages are taken together, it becomes clear that this is
not material division that Aristotle is describing, but a search for the
indivisible forms of the kinds under investigation.
At the beginning of Pol. ii., however, before embarking on a systematic
study of the various forms of constitution as the natural starting point of
his search of the best polis possible, Aristotle feels the need to explain that
he is taking this methodos upon himself (ἐπιβαλέσθαι τὴν μέθοδον) not out
of a wish to appear clever, but because the constitutions that currently exist
are not well-established (τὸ μὴ καλῶς ἔχειν) (1260b34-36). That is, the
norm-governed inquiry includes a study of current forms of constitution
for a very domain-specific reason: none of those that currently exist are
ideal, but by studying them we will learn something about the virtues and
vices of those that do exist, as a suitable starting point for an inquiry aimed
at framing the best possible polis for a flourishing human life.
3.7 Conclusion
Understanding the dual roles of the concept of methodos in the Aristotelian
corpus is a key to understanding Aristotle’s views about scientific inquiry,
and specifically his views about the way in which distinctive scientific
inquiries are governed by norms that are domain-specific, appropriate for
and distinctive to that domain. By exploring its use across the corpus as
a whole we have seen how it is deployed and why it so often appears in the
introductory ‘methodogical’ passages of his investigations. Moreover, it
appears to be part of a network of concepts, such as paideia, zêtêsis, and
episkepsis, related to the question of how one ought to pursue one’s inquiry
so as to optimize the chances of acquiring scientific knowledge. From this
point forward, I will use ‘method’ to translate it when Aristotle is discuss-
ing the proper way to carry out an inquiry, and ‘methodical inquiry’ when
Aristotle is referring to a domain of inquiry, viewed from the perspective of
the norms governing its practices.
The focus of the remaining chapters of this book will be on Aristotle’s
scientific inquiry into the natural world. I will conclude Part i by
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
considering two related questions: Does Aristotle think that there is
a single, unified science of nature? And, does Aristotle think that there
are many distinct inquiries with distinctive norms of inquiry, contributing
to that science of nature? The evidence, I will argue, points to the conclu-
sion that the answer to both questions is “Yes.” It is in part the mandate of
Part ii of this book to provide that evidence.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
chapter 4
Natural Science
Many Inquiries, One Science
Wherefore one needs to have been educated about how one ought to
accept each subject, since it is absurd to inquire at the same time about
scientific knowledge and the way of acquiring it; but neither one is
easy to grasp. One ought not to demand mathematical accuracy in all
cases, but only in those subjects without matter; for which reason the
[mathematical] approach is not a natural one, for presumably nature
in all cases possesses matter. For which reason one ought first to
investigate what nature is; for in this way it will also be clear with
which things natural science is concerned, and whether it is for one
science or more than one to study the causes and principles.
(Metaph. α.3, 995a12-20)1
Chapter Summary. Metaph. Ε.1 provides a deceptively clean map of

scientific knowledge, differentiating three forms of theoretical knowledge
(first philosophy, mathematics, natural science) from each other and
theoretical knowledge as a whole from practical and productive know-
ledge. The focus of this chapter will be on natural science (ἡ φυσικὴ
ἐπιστήμη, 1025b19), considered as the ultimate goal of natural inquiry.
Given the results established up to this point about Aristotle’s general
account of inquiry, my aim in this chapter is to answer two related, more
specific questions about natural inquiry:
(i) Given that there are a number of natural inquiries with their own
domain-specific norms, to what extent does Aristotle see them as
contributions to a single science – and, to the extent that they are, to
what extent and in what respects is that science unified?
1
διὸ δεῖ πεπαιδεῦσθαι πῶς ἕκαστα ἀποδεκτέον, ὡς ἄτοπον ἅμα ζητεῖν ἐπιστήμην καὶ τρόπον
ἐπιστήμης· ἔστι δ’ οὐδὲ θάτερον ῥᾴδιον λαβεῖν. τὴν δ’ ἀκριβολογίαν τὴν μαθηματικὴν οὐκ ἐν
ἅπασιν ἀπαιτητέον, ἀλλ’ ἐν τοῖς μὴ ἔχουσιν ὕλην. διόπερ οὐ φυσικὸς ὁ τρόπος· ἅπασα γὰρ ἴσως
ἡ φύσις ἔχει ὕλην. διὸ σκεπτέον πρῶτον τί ἐστιν ἡ φύσις· οὕτω γὰρ καὶ περὶ τίνων ἡ φυσικὴ δῆλον
ἔσται καὶ εἰ μιᾶς ἐπιστήμης ἢ πλειόνων τὰ αἴτια καὶ τὰς ἀρχὰς θεωρῆσαί ἐστιν.
99
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
100 Natural Science: Many Inquiries, One Science
(ii) How does Aristotle understand the relationship between natural
inquiry and its goal, scientific knowledge of nature?
The abovementioned Metaph. Ε.1 ‘map’ leads naturally to two further
questions that will be explored in more detail in later chapters:
(iii) What sorts of interactions can we expect at the boundaries between
the various natural inquiries? This question will be explored in
Chapters 8 and 9.
(iv) Are there works of Aristotle’s, unified on their own terms, which do
not fit into any one of the demarcated domains? This is an issue to be
explored in Chapter 7, with the focus being on the De anima.
An ancillary goal of the present chapter is to motivate the need to explore
these questions.
4.1 Introduction
Much like Aristotle’s use of the concept methodos, our current concept of
‘science’ is used in two closely related but distinct senses: to refer to the
structured knowledge that has been achieved about a domain of investiga-
tion, and to refer to the norm-governed investigation of that domain. This
duality naturally leads to two distinct ways of conceiving of the unity of
a science. Thinking of a science as a body of knowledge, we can ask if the
knowledge we have of a domain constitutes a structured unity, the unity
perhaps provided by explanatory relationships among propositions or
models, or classificatory relationships among concepts or theories. Of the
biological sciences, for example, we might ask whether there is one bio-
logical theory that unifies (perhaps by means of a single pattern of explan-
ation, or by means of a unifying concept or network of concepts) all our
knowledge of the living world. Thinking of a science as an inquiry carried
out according to certain distinctive practices, on the other hand, we might
ask whether there is a single set of methods or norms specific to the domain
in question that, if followed, will continuously advance our understanding
of that domain.2
I have been insisting that it is important to keep a distinction of that sort
in mind in considering Aristotle’s corpus as well, although the conceptual
map is drawn in a very different way. The Greek term that I have been
2
There is an extensive literature in philosophy of science on just these questions. To select a few,
staking out different positions: Cartwright 2000, Dupré 1993, Kitcher 1993, Mitchell 2003,
Rosenberg 1994, Schaffner 1993.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
4.1 Introduction 101
translating as ‘scientific knowledge,’ epistêmê, is used by Aristotle first and
foremost to refer to the cognitive state of a knower with a special kind of
structured knowledge of a domain – knowledge structured by causal
demonstrations based on fundamental, indemonstrable principles.3 He
does not, however, use that term to refer to the inquiries that lead to
such knowledge. As we saw in Chapter 3, Aristotle also has a rich and
complex vocabulary for discussing inquiry – zêtêsis, historia, skepsis, meth-
odos, and theoria, just to mention the most common terms.4 Each of these
terms has a distinct etymology and thus each carries a different set of
connotations5 – but Aristotle will occasionally conjoin them or use them
interchangeably. Attending carefully to their use in Aristotle’s presentation
of different areas of inquiry, a process begun in Chapter 3, will help us to
understand how he conceives of the relationship between the cognitive
states that make inquiry possible, the processes involved in active inquiry,
and the ultimate goal of those processes, scientific knowledge.
This distinction – between structured scientific knowledge of a domain
and a norm-governed inquiry or investigation directed toward achieving
that knowledge – is particularly helpful in considering the two books of the
Posterior Analytics. As I argued in Chapters 1 and 2, APo. i is primarily
devoted to exploring the structure of scientific knowledge, while book ii is
devoted primarily to an exploration of the modes of inquiry that are
involved in proceeding toward the goal of scientific knowledge, and in
particular, toward knowledge of causal essences that can serve to explain
why it is that the subjects in a domain have the attributes they do. As
I noted in Chapter 2, Aristotle telegraphs this way of understanding the
relationship between the two books when he opens book ii with these
words:
The objects of inquiry (τά ζητούμενα) are equal in number to those we
know scientifically (ἐπιστάμεθα). (APo. ii.1, 89b23-24)
3
In EN vi it is one of a number of cognitive virtues, all of which are ‘possessions’ (hexeis), that is habits
acquired through practice or teaching. Strictly speaking, the account given here only applies to
unqualified scientific knowledge, (epistêmê haplôs, APo. i.2, 71b9), since epistêmê is used, even within
the Analytics, in a less restricted fashion – for example, when he distinguishes epistêmê of the fact from
epistêmê of the reasoned fact (APo. i.13, 78a22).
4
In works such as Burnyeat 1981 and Salmieri 2013, there have been valuable studies of Aristotle’s
network of concepts referring to knowledge (especially gignoskein, gnorizein, epistasthai, and eidenai).
One consequence of the research in which I’ve been engaged in writing this book is the realization
that there needs to be a similar exploration of this network of concepts referring to inquiry.
5
This is a topic that will be explored in a variety of ways in later chapters. But to point to what I have in
mind, some of these words come from the context of seeking and searching (zêtêsis, skepsis), some
from that of a route or path to a goal (hodos, methodos,), and some from that of observing or studying
(historia, theoria).
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
He then immediately identifies four distinct sorts of things to be known by
four distinct modes of inquiry and begins to explore how they are related to
one another, and in Chapter 2, I made the case that this exploration is the
central theme that unifies the rest of APo. ii.
Just above, I described APo. ii as “an exploration of the modes of inquiry
that are involved in proceeding toward the goal of scientific knowledge.”
But as we saw in Chapter 2, “modes of inquiry” is open to a number of
interpretations. My claim is that the two pairs of questions or “things we
seek” around which APo. ii is structured do not relate to different domains
of inquiry, but rather to different modes and stages of any inquiry, insofar as
its aim is scientific knowledge. They provide an erotetic framework for
knowledge-seeking inquiry – and as we saw, an oft-noted feature of that
work is strong evidence for this claim: its examples are drawn from an
extremely wide range of subjects, from arithmetic and geometry to meteor-
ology and astronomy to biology and medicine – even ethics and warfare;
and often the same point about inquiry will be simultaneously illustrated
by seemingly unrelated inquiries, such as botany and astronomy (98a35-
b24) or botany and geometry (99a16-29).
In this chapter, I will begin to narrow our focus to natural inquiry, but
with a specific question in mind: how does Aristotle understand the
relationship between natural science viewed as a variety of activities of
search or inquiry employed in the pursuit of knowledge, on the one hand,
and natural science viewed as the goal of such inquiries, scientific know-
ledge of nature, on the other? Aristotle took the view that, even within
natural science itself, differences in the objects being investigated, in our
epistemic access to those objects, and in the perspective taken on them
demand different methods of investigation governed by different norms.
But his admirable refusal to give in to the temptation of unification by
reduction (to which so many of his predecessors and contemporaries
succumbed) created a problem for him and for our understanding of
him: Given the many, disparate forms taken by natural inquiry, is there
sufficient underlying unity to consider them forms of a single, common
endeavor? In order to answer that question in the affirmative, one needs to
identify the nature of that unity: is it a unity of purpose and method of
investigation, a unity in the structure of the resulting scientific knowledge,
or some combination of the two?6
6
A different approach to the question of the unity of Aristotle’s diverse explorations of nature can be
found in Falcon 2005, driven more by ontological concerns about causation than the epistemological
issues on which I am focused.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
4.2 The Epistemic Map 103
4.2 The Epistemic Map
Aristotle took the first explicit step toward subdividing the study of the
natural world into a variety of distinct, and somewhat autonomous,
investigations (and distinct works, logoi or pragmateiai, devoted to those
investigations).7 Prior to Aristotle, there may have been works on nature
(περὶ φύσεως),8 but it does not appear to have been assumed (to take one
example more or less at random) that the study of animal generation was
a study distinct from that of the transformation of the elements or of the
nature and motions of the heavenly bodies – let alone distinct from the
study of animal parts or animal motion.9 By contrast, Aristotle’s investiga-
tion of nature is an articulated investigation, and the articulation is self-
conscious: almost every work begins by introducing its subject as distinct
from others, and usually by indicating how, though distinct, it is related to
other natural investigations.10
This division of the investigation of nature is not sufficient to create
a concern about the unity of scientific knowledge of nature, however. For
Aristotle might have concluded that the results of each natural investiga-
tion constitute an autonomous epistêmê, in which case the concern about
how they are all to be unified would simply not come up. But for better or
worse, this appears not to be the route he took.11 As will be discussed in
more detail in Section 4.3, the program or plan (προαίρεσις, 339a9) for the
7
A point nicely stressed in the opening pages of Wilson 2000.
8
Though there is reason to be skeptical about the writings of Empedocles, Anaxagoras, οr
Democritus having titles in our sense; cf. Schmalzriedt 1970. The reference to the Timaeus in
Diogenes Laertius (D.L. iii.60: Τίμαιος ἢ περὶ φύσεως) suggests it too was considered
a contribution to this tradition, and it does share the attribute of being a single, all-encompassing
treatise on every aspect of the natural world, from the elements to the heavenly spheres. Moreover,
Aristotle treats it in the same way as previous investigations of nature. For example, at de An. i.3,
406b25-28 he says: “And in the same way [as Democritus] the Timaeus also gives a natural account
(φυσιολογεῖ) of how the soul moves the body; for it is by itself being moved that it moves the body,
on account of being intertwined with it” (cf. Sens. 2., 437b9-14, where the theory of vision in the
Timaeus is compared to that of Empedocles). And Ph. i.4 opens with a general categorization of
earlier investigators of nature, and at 187a12-23, Plato is cited as an example (albeit an odd example)
of the first group discussed.
9
I am setting aside here the development, recognized by both Plato and Aristotle, of astronomy,
optics, mechanics, and harmonics as distinct mathematical sciences, as well as the developments in
medicine reflected in the earliest material in the Hippocratic corpus. I do discuss Aristotle’s views
about how the science of nature is related to mathematics in Lennox 1986, and to medicine in
Lennox 2005.
10
For paradigmatic examples, cf. PA i.1, 639a1-12; MA 1, 698a1-10; GA i.1, 715a1-18; de An. i.1, 402a1-
11 – and the passage we will consider in some detail shortly, Mete. i.1, 338a20-339a10.
11
And, in this respect, the science of nature is in stark contrast with mathematics. See, for example,
APo. i.14, 79a18-20: αἵ τε γὰρ μαθηματικαὶ τῶν ἐπιστημῶν . . . οἷον ἀριθμητικὴ καὶ γεωμετρία καὶ
ὀπτική. Cf. Metaph. Γ.1, 1003a25-6; Ε.1, 1026a25-28; Μ.3, 1078a33, b2.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
study of nature presented at the beginning of Meteorology (Mete.) clearly
treats the study of nature as some sort of unity. That program begins with an
inquiry into the general principles about natural motion and change (338a20-
21), which is to be followed by a study of the heavens (a21-22), of the elements
and their transformations (a22-24), and of coming to be and passing away in
general (a24-25). The subject matter of meteorology is said to be “part of this
inquiry,”12 and its subject matter is described in some detail (338a25-339a5),
after which we need to consider whether an investigation of animals and
plants can be carried out in the same manner (339a5-8) – for having done so,
he declares, the goal initially set out will have been achieved (a8-10). Whereas
references to mathematical knowledge throughout the corpus typically refer
in the plural to a number of distinct disciplines (see references in note 11),
references to scientific knowledge of nature do not.13 Metaph. Ε.1 identifies
natural science (φύσικη ἐπιστήμη, 1025b18-19) as knowledge of a certain
kind of being, namely of substantial unities of matter and form with their
own sources of change and rest (1025b19-21). And a concern about the unity
of natural science motivates much of the discussion in Ph. ii.2: since nature
is twofold (διχῶς, material and formal), does this threaten the possibility of
nature being studied in a unified way (cf. 194a15-27)?
4.3 Demarcating the Science of Nature

In Metaph. Ε.1, Aristotle aims to clarify further the special sort of wisdom
that has been the goal of his inquiry from the beginning, by differentiating
it from other kinds of inquiry. He is seeking the starting points and causes
of being qua being, in contrast to disciplines such as medicine or mathem-
atics that are concerned with principles and causes of a delimited class of
being. Yet the chapter has a great deal to say about what is distinctive about
scientific knowledge of nature, and the reason why becomes clear near the
end of the chapter:
If then there is no other substantial being (οὐσία)14 apart from those
constituted naturally, the science of nature would be primary; but if there
12
μέρος τῆς μεθόδου ταυτῆς, 338a25. This is a good example of the way in which methodos can serve to
pick out a domain of inquiry in a way that stresses the mode of inquiry specific to that domain, as
discussed in Chapter 3. This passage says very little about what unites these as parts of this methodos;
but these are all natural inquiries, and we will explore what is involved in the methodos of nature in
Chapter 5.
13
See, for example, Metaph. Ε.1, 1025b19, 26, 1026a6, 12; Cael. i.1, 268a1, on which see Falcon, 2005,
31–48.
14
Throughout I will translate οὐσία as ‘substantial being,’ the adjective in recognition of the time-
honored, though in my view confusing, habit of translating it ‘substance,’ the nominal participle in
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
4.3 Demarcating the Science of Nature 105
is some changeless substantial being, this would be prior and <the science of
it> would be primary philosophy, and universal in this way, because
primary. And it would then be for this science to study being qua being,
both what it is and the things that belong to it qua being. (1026a27-32)
Thus, while the entire range of intellectual pursuits are at least briefly
discussed in this chapter, one of Aristotle’s chief concerns is to determine
whether the science of nature is the primary science, or whether there is
a study prior to it. There will be a study prior to natural inquiry only if it
turns out that, among the things that are, there are not only entities that are
material and undergo change, but at least one that is changeless (which for
Aristotle would imply immateriality).
At first glance, it is puzzling that Aristotle raises this concern here.15 This
chapter opens by reiterating the distinction, introduced at the beginning of
Metaph. Γ.1, between the investigation of being qua being and investiga-
tions focused on some one part of being (1003a21-25); and the science of
nature is introduced in Metaph. Ε.1 within that very framework:
But since scientific knowledge of nature happens to be about a certain kind
of being (for it is concerned with the sort of being in which there is an
inherent source of change and rest) . . . (1025b18-21)
This would appear to disqualify natural science immediately as a candidate
for first philosophy, since, in the language of this very chapter, it is one of
those disciplines that “circumscribe a certain being and a certain kind, and
treat of it, and not of being without qualification nor qua being” (1025b8-
10). How, then, is it circumscribed, such that it is not an inquiry into being
qua being?
One clear mark of physikê epistêmê has already been identified in the
process of distinguishing it as a theoretical rather than a practical or product-
ive activity: the particular kind of being which the natural investigator takes
as his subject matter consists of objects that have inherent sources of change
and rest. It is “[theoretical knowledge] about the sort of being that is capable
of undergoing change, and for the most part about substantial being accord-
ing to account (οὐσίαν τὴν κατὰ τὸν λὸγον), only not separable.”16
Commentators typically assume that οὐσίαν τὴν κατὰ τὸν λὸγον is a way
recognition of the word’s actual roots. It is an abstract noun derived from οὖσα, the feminine
participle of εἶναι, to be. A literal but inelegant translation would be ‘beingness.’
15
For a valuable discussion of just how puzzling, see Shields 2012, 343–371. See Frede 1987a, 81–95,
following up an earlier suggestion by Patzig, for one sort of solution.
16
Cf. Ross 1924: 1, 354. Ross prints ὡς οὐ χωριστὴν μόνον following mss. E and T and omits the
preceding comma of the Oxford Classical Text. But the addition of ὡς is not necessary. I am
following Alexander’s understanding of the passage, taking μόνον adverbially (LSJ, 1145 cites this
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
of referring to substantial being as form.17 But caution is called for; as de An.
ii.1 reminds us, οὐσία can refer to three things: matter, form, or the
composite of the two. Aristotle may well be leaving the question of what
to substitute for οὐσία here open, and thus what sort of account will be
given. And in fact, immediately after this puzzling sentence concerning the
kinds of things natural inquiry is to investigate, Aristotle issues a caution:
It is necessary not to lose sight of the manner of the-being-what-it-is and the
account, because without this the inquiry is destined to be unproductive.
(Metaph. Ε.1, 1025b28-30)18
The caution is a reminder that what one identifies as ‘the being-what-it-is’
of something and what the account (τὸν λόγον) should include will differ
as between mathematics and natural science, and it introduces a passage in
which Aristotle elaborates on what is distinctive about the objects to be
investigated by the natural scientist.
Among things defined and among the ‘what-it-is’ of things, some are in the
manner of the snub and others are in the manner of the concave. These differ in
that the snub is comprehended with the matter (for the snub is concave nose)
while the concave is <comprehended> without perceptible matter. And if all
natural things are spoken of in like manner to the snub, for example, nose, eye,
face, flesh, bone – generally, animal – leaf, root, bark – generally, plant (for the
account of none of these things is without change, but always has matter), then
it is clear in what way one ought to inquire into and define the what-it-is in
natural things, and the reason why it is also for the natural inquirer to study
certain things about the soul, as many as are not without matter. (1025b30-
1026a6)
Thus, after being cautioned about attending to the distinctive ways that
different inquiries approach ‘the account’ and ‘the being-what-it-is,’
Aristotle introduces, conditionally, the thought that all natural things are
spoken of like ‘the snub.’ These are things the accounts of which are “not
without change and always include matter.” Unfortunately, although he
claims that on the assumption that natural things are spoken of like the
snub, “it is clear in what way one ought to inquire into and define the what-
it-is in natural things,” he does not actually spell out the implications for
inquiry and definition of this assumption.
passage for this use of the neuter μόνον). On its own it is difficult to understand, but the next few
lines make Aristotle’s point clear.
17
Ross 1924: 1, 354; Kirwan 1971, 185; Peramatzis 2011, 100.
18
δεῖ δὲ τὸ τί ἦν εἶναι καὶ τὸν λόγον πῶς ἐστὶ μὴ λανθάνειν, ὡς ἄνευ γε τούτου τὸ ζητεῖν μηδέν ἐστι
ποιεῖν.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
4.3 Demarcating the Science of Nature 107
One obvious question that needs to be answered is whether the material
aspect of a natural substance should be referenced in its definition or not,
and if so, how. Determining the answer to this question is directly relevant
to the question of what a methodical inquiry concerning nature will look
like, since arriving at definitions that identify the what-it-is of natural
substances is a central part of the goal of such an inquiry. If the material
nature of the object of inquiry is somehow to be included in a scientific
definition of that object, then there are further questions about how one
inquires into that material nature and into the relationship between the
material and formal aspects of different natural objects – and as we will see
in Chapter 5, this is a primary concern of Physics ii.2, and indeed of Ph. ii
generally. I will simply flag these questions here, since a primary goal of
Part ii of this book is to explore Aristotle’s actual natural inquiries with
a focus on methods of inquiry – and that exploration ought to provide
a good deal of evidence to aid in answering such questions.
Before leaving Metaph. Ε.1, however, I would like to highlight three
additional features of the above passage that are important for the themes
of this book:
1. All the examples of ‘snub-like’ natural things are biological, and both
nonuniform19 and uniform parts of animals and plants, as well as
animals and plants themselves, are mentioned. Are there features of
living things that make the questions being raised here especially
challenging or important?
2. Aristotle claims that the snub-like character of natural objects not only
has implications for how one ought to define what natural things are
and explain their attributes and changes, but also for how one ought to
inquire into what they are.20 The literature on ‘the snub’ has focused
almost exclusively on the question of definition:21 in this study our
focus is on inquiry, keeping in mind that at least one primary goal of
such inquiries is achieving a proper, essence-identifying definition of
these ‘snub-like’ objects.
3. The last line in the above passage has important implications for how
one understands the project of de An. and its relationship to Aristotle’s
animal investigations – this will be a central theme of Chapter 7. The
19
We may, perhaps, consider the fact that he mentions noses first in his list of nonuniform animal
parts that are ‘snub-like’ as an instance of Aristotelian humor.
20
δῆλον πῶς δεῖ ἐν τοῖς φυσικοῖς τὸ τί ἐστι ζητεῖν καὶ ὁρίζεσθαι (1026a4-5).
21
Among more recent discussions, see Balme 1987b, Code 2010, Ferejohn 1994, Frede and Patzig 1988,
Gill 1989, Lennox 2008a, and Peramatzis 2011.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
message of this sentence is clear, however: the fact that all animals,
plants, and their parts are investigated “like the snub” will have
implications for how the soul is to be investigated – but only insofar
as the soul is “not without matter.”
While there are many questions that are left unanswered, then, the
message of this chapter is, in broad outlines at least, clear: the account
we are aiming for in natural inquiry, though theoretical, must include
reference to matter and change; and therefore, the natural scientist, while
acknowledging the priority of form, must make the material makeup of
natural things and the changes they undergo central to his investigations.
Thus, the sort of abstraction that is appropriate in mathematics, where
the proper account of (say) a parallelogram or rhombus does not mention
perceptible material or change at all, would be entirely inappropriate.
However, little or nothing is said in this chapter about the definition
referring as well to form, nor about the nature of the relationship between
matter and form in natural objects, nor about how that relationship
should be investigated or represented in a proper definition or a causal
demonstration. Of primary importance for our concerns, nothing is said
about the implications of this characterization of the objects of natural
science for how one ought to inquire into them. I now turn to that
question, in part using the results of our general exploration of the
concept of methodos to characterize what is distinctive, at the most
general level, about a natural inquiry.
4.4 Demarcating Natural Inquiries

While Metaph. Ε.1 clearly indicates that Aristotle considers all investi-
gations of natural objects as parts of a single science of nature, it is also
clear from a variety of other texts that he sees this science demarcated
into a number of relatively autonomous domains of study. In Section
4.4.2, we will investigate Aristotle’s views about what is common to all
natural inquiries that distinguishes them both from other theoretical as
well as from practical and productive inquiries. And to complicate
matters still further, though autonomous in ways I will spell out in the
chapters in Part ii, these natural investigations are self-consciously
interdependent. This interdependence will be the focus of Chapters 8
and 9 – here I will simply highlight a number of clear examples of this
interdependence and the manner in which Aristotle takes great pains
to point it out.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
4.4 Demarcating Natural Inquiries 109
4.4.1 Highlighting Dependence
Mete. i.2 opens by referring to a number of conclusions reached in De caelo
(Cael.) that will serve as assumed starting points for meteorology – and
then, in chapter 3, he explains that in that chapter he has “recalled our
initial posits and previously stated definitions (τὰς ἐξ ἀρχῆς θέσεις καὶ τοὺς
εἰρημένους πρότερον διορισμούς)” (339a33-34).22 Similarly, De motu ani-
malium (MA), after a brief but important introduction, begins its argu-
ment as follows:
So, then, that a source of the other motions is that which moves itself, and
a source of this [self-motion] is what is unmoved, and that the primary
mover must be unmoved, [these matters] were determined previously,
precisely when we determined, regarding eternal motion, whether or not
it exists, and if it exists, what it is. (MA 1, 698a8-11)23
This is a quite precise reference to Ph. viii.6, and the inquiry that ensues
relies in a number of respects on the conclusions reached in Ph. viii.24
Similarly, as we will examine in detail in Chapter 9, PA ii and GA ii and
v are systematically dependent for their accounts of the constitution and
development of the uniform parts of animals on the theory of the consti-
tution of uniform materials in Mete. iv – as the last lines of Mete. iv.12
would lead one to expect. And there are clear references to that background
in both texts, for example:
But what the solidifiable things are and due to what causes they are solidi-
fied, these matters have been determined more clearly in other places. (PA
ii.2, 649a33-34)
The generation of the uniform things is by cooling and heat; for some
things are solidified by cold and some by heat. I have spoken previously in
other places about the difference between these things – what is dissolvable
by moisture and by fire, and what is not dissolvable by water and is
unmeltable by fire. (GA ii.6, 743a3-8)25
22
For more detail on the dependence of the biological account of uniform parts on Mete. iv, see
Lennox 2014d, 272–305. Chapter 9 further develops the ideas in that paper.
23
It will be noted that the characterization of the accomplishment of Ph. viii.6 is in terms of the
erotetic framework set out in APo. ii – first it was established that there is eternal motion, and then it
was determined what it is. A number of the chapters in Primavesi and Rapp, eds. 2020 discuss the
many interdependencies of MA with other natural inquiries. Besides the Physics, the questions taken
up in the last six chapters are intertwined with those of de An. iii and PN – in fact, there are good
reasons (beginning with its last sentence) to think MA was conceived as part of the sequence of
studies that make up PN, on which see the Introduction to Laks and Rashed 2004, and the
Introductions to Primavesi and Rapp, eds. 2020.
24
Wilson 2000, 47–52. 25 Cf. GA v.4, 784b8-9.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
And – to continue narrowing the scope – within the works devoted to animals
there are not only regular cross-references,26 but a careful demarcation of
tasks. In PA iv.10, for example, Aristotle self-consciously puts off defending
the residual nature of seed and menstrual discharge with the following words:
This [the fact that the same organs are used for reproduction and for the
elimination of useless residues] is because the seed is something moist and
a residue – let this be assumed (ὑποκείσθω) for now; later it will be proven
(δειχθήσεται). And of the same character too are the menstrual discharges in
females and that by which there is an emission of seed. (These things too will
be defined (διορισθήσεται) later; for now, let it only be assumed (ὑποκείσθω)
that menstrual discharges in females are a residue.) (PA iv.10, 689a9-14)
Here again, there is a sustained investigation aimed at establishing that the
male and female contributions to generation are both useful residual fluids
in GA i.17–19, and every reason to think the above passage is a (forward!)
reference to that discussion. The future tense of the references to proving
and defining what must be assumed in the PA iv.10 discussion also points
to Aristotle having a preferred order in which these topics are to be studied;
and the parts contributing to generation, including the uniform parts, are
to be studied later as part of the study of generation. Moreover, we are told
that it will there be proven that seed is ‘moist and a residue’ and that seed
and menstrual discharge will be defined there, both of which tasks are
accomplished in those chapters of GA i.27
Though many more examples could be provided,28 these are, I trust,
sufficient to show that Aristotle’s various natural investigations are both to
some extent autonomous and yet carefully and self-consciously integrated.
4.4.2 Autonomy and Integration

I turn now to an exploration of the nature and extent of that autonomy and
integration. This question could be explored in two very different ways. We
could ask the question: Does this self-conscious stress on both autonomy and
integration pose a problem for the model of the unity of the scientific
knowledge that one finds articulated in APo. i? The references to mathematics
in the Metaphysics and Analytics show that it would be reckless to assume that
any domain of theoretical philosophy (φιλοσοφία θεωρητική) has the unity
26
A reasonably complete list can be found in Bonitz, Index Aristotelicus (Aristotelis Opera v): Where
the list starts depends on what one counts as works devoted to animals, but it should run at least
from 99a21 to 100b42.
27
On all of which see Bolton 1987, 151–166.
28
For other examples, see Burnyeat 2004, 13–24; Lennox 2009 (in Wildberg and Bowen eds., 2009).
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
demanded by APo. i simply because there is sufficient unity for there to be
a single generic name for it. Aristotle typically talks of mathematical sciences in
the plural – and more to the point, geometry and arithmetic are his prime
examples in the Analytics of domains of scientific knowledge between which
the borrowing of definitions and proofs to serve as assumptions would be
illegitimate (e.g., APo. i.7, 75a38-b5). That is, they are his prime examples of
distinct and autonomous sciences, even though they are both referred to as
forms of mathematics.
In the remaining sections of this chapter, I will consider previous
attempts to deal with this issue that fail, but do so in instructive ways. In
Part ii of this book, beginning in Chapter 5, I will then spell out what is,
I believe, a compelling case for the following thesis: Aristotle has a well-
defined picture of what he refers to as a methodos of nature, that is, a single
domain of inquiry with general norms that are common to all natural inquiries
and distinguish them from other theoretical, practical, and productive inquir-
ies. Nevertheless, these norms are themselves insufficient to insure the
success of the various distinct natural inquiries. The case for this last
point will rest on the explorations in the remainder of Part ii, though
the outlines of my argument are no doubt already tolerably clear.
This is an issue of some importance for Aristotle’s philosophy of science,29
and it can be stated in the form of the following dilemma: Aristotle holds
either that there is one science of nature, or that there is more than one. If he
holds that there is a single, unified science of nature, then the many
differentiated investigations we have just been reviewing might be thought
to raise questions about how they can constitute a unity. If, on the other
hand, we take those differentiated inquiries to be evidence that Aristotle
holds that there are many natural sciences, then we need to ask how the
clearly identified interconnections among these inquiries can avoid the
sanctions against ‘kind-crossing’ in APo. To return to our point of departure,
how much variation can there be in the methods and norms of distinct natural
inquiries before it threatens the goal of a single science of nature?
4.4.3 Advocates for the One and the Many

In order to situate the ensuing investigation, Aristotle’s Meteorology begins
with what appears to be the entire course of natural inquiry as Aristotle sees
29
And not only Aristotle’s: one of the most distinctive features of recent science, despite dreams of
a ‘theory of everything,’ is the simultaneous subdivision of advanced sciences such as biology and
chemistry into multiple subdisciplines and attempts at ‘cross-fertilization’ between disciplines (e.g.,
neuroscience and psychology, developmental and evolutionary biology).
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
it. A number of authors have taken it to be a kind of guide to the structure of
Aristotle’s science of nature. But what guidance it may provide depends on
what sort of structure it represents, and in the end, I find the passage more
puzzling than illuminating on that subject. It is of some importance that
Aristotle never uses the term epistêmê in Mete. i.1. Rather, in discussing
meteorology in relation to prior topics listed that correspond, at least in
a general way, to the topics covered in our Physics, On Generation and
Corruption, and De caelo, he refers to it, as we saw, as “a part of this methodos”
(338a26) – this way of proceeding, if one wants a neutral rendering. Based on
the results of Chapter 3, it is reasonable to suppose this refers to inquiry based
on the same norms and standards, though Aristotle does not offer any
explication of the method he has in mind here. After providing a list of topics
meteorological investigation will take up, Aristotle goes on:
Having dealt with these subjects, we will study whether we are somehow
able, according to the recommended manner,30 to give an account of
animals and plants, both in general and separately. For having given an
account of these things we would pretty much have reached the goal of our
original plan in its entirety. (Mete. i.1, 339a5-9)
Perhaps by ‘recommended manner’ he means ‘according to the same
methods used in the previously mentioned inquiries,’ but again he makes
no effort to explain what he means. What is clear, however, is that, when it
comes to animals and plants, Aristotle is not asserting that the researcher
who is following his plan should next go on to study animals and plants
according to the recommended manner. Rather, he is saying that the next
thing to be investigated is a question: are we able, somehow, to give such an
account of animals and plants following the same methods? How Aristotle
answers this question is of some importance for our current concerns about
the autonomy and unity of his investigations of nature. If the study of
animals and plants is taken to complete an ordered investigation into
nature, as he here claims, what will it mean for the unity of the investigation
of nature if a significant portion of it – indeed its culmination – is to be
conducted according to a different set of norms from the rest of that
investigation?31
30
The Greek is κατὰ τὸν ὑφηγημένον τρόπον; a comparison with Pol. i.1, 1252a17-18 and GA iii.8,
758a28-29, where the expression κατὰ τὴν ὑφηγημένην μέθοδον is used, suggests that in such
contexts the two phrases are more or less interchangeable. The verb ὑφηγέομαι has the general
sense of leading, guiding, or showing the way, and the participial form thus needs to be rendered in
a way that captures the implicitly normative force it has in these passages.
31
Burnyeat hints at this concern in a footnote to his 2004 essay, 13 n. 16: “By saying ‘a different set of
norms’ I don’t mean to imply that the norms are incompatible in any way.”
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
Myles Burnyeat has argued that the many temporally qualified cross-
references in the natural treatises confirm that the remarks opening Mete.
i.1 represent “the order of argument and exposition” in natural science
(2004, 20); earlier (Burnyeat 2001, 113–120) he refers to a “didactic” and
“systematic” order. His main point, with which I wholeheartedly agree, is
that the opening chapter of the Mete. should not be assumed to reflect the
order in which Aristotle actually investigated or wrote about nature, nor
even a recommendation of such an investigative order.32 The many cross-
references in the extant natural treatises are, however, to be taken as
reflecting Aristotle’s views about explanatory or conceptual relationships
among these works.33 Andrea Falcon suggests, in addition, that this passage
reflects causal priorities, priorities that will secure the unity required for an
Aristotelian genos, and thus for an Aristotelian epistêmê (Falcon 2005, 1–16).
Many years previously, Martha Nussbaum, in one of the essays appended
to her text and translation of MA, argued that the Mete. introduction
represents “a certain logical order that ought to be followed in presentation
and study” (Nussbaum 1978, 108). Both Burnyeat and Nussbaum were
assuming that the opening of Mete. i lays out a plan for a single, unified
science of nature.
Nussbaum, however, discusses this passage in the context of her argu-
ment that MA does not fit into the Analytics mold of a unified natural
science. She claims that, in virtue of its mix of physics, geometry, cosmol-
ogy, practical syllogisms, and animal physiology, it “contravenes a basic
tenet of [Aristotle’s] philosophy of science” (Nussbaum 1978, 109), namely,
that each science must start from true and necessary first principles peculiar
to that science, and those principles must be grounded inductively on that
science’s own appearances. “Valid deduction may not pass from one genus
32
Burnyeat’s discussions do, however, make one questionable assumption: that the closing reference to
a study of animals and plants implies that De anima is part of this organized plan for the study of
nature. I discuss this further in Chapters 5 and 7.
33
The issue of how to make use of cross-references in investigating such questions is complex, but
I think it is helped by making a number of distinctions. First, one needs to distinguish references that
are imbedded in discussion and argumentation (such as the forward references in PA iv.10 to the
arguments and definitions provided by GA i.17–23, discussed on p. 110) from those that are
essentially isolated and which could be removed from the text without significant damage to
argument or exposition (e.g., PA i.5, 646a2-4; iv.14, 697b27-30; MA 11, 704a3-b3; IA 18, 714b20-
24). It is less likely that the former references were added by a later editor, more likely with the latter,
especially in light of occasional inconsistencies about the order of inquiry in the latter references.
Nevertheless, some of the references forward and backward at the beginnings and endings of texts
explicitly point to relations of conceptual or explanatory dependency (e.g., those at the beginning of
GA i and PA ii) and for that reason are potentially valuable for the current project. With these texts,
an interesting test of their usefulness as evidence of Aristotle’s views is considering whether the
beginning of the text would make sense in their absence.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
to another” (ibid., 110). Nussbaum acknowledges that the discussions, in
APo. i.7–13, of the subordination of one science to another, could provide
an option, but claims, mistakenly, that these passages only sanction using
materials from one science as “a source of illustrations and models” for the
other.34 Notice that, unlike Burnyeat and Falcon, Nussbaum’s argument
assumes that the different natural studies allegedly combined in MA are
distinct sciences, not parts of one science.35
In an interesting response to Nussbaum’s essay, Joan Kung made the same
assumption,36 but argued that the role played by the appeals at the beginning
of MA to the conclusions of Physics viii – as well as its apparent references to
cosmological principles – are validated by his theory of subordinate sciences
in APo. i.7–13. We have here, she argued, a case of subordination: “the
science of motion in general and the science of animal motion.”37
Nussbaum and Kung share, then, the view that the various natural
treatises (or at least the MA, Cael., and Ph. viii) are contributions to distinct
sciences,38 but disagree over whether this poses a problem for the author of
APo. I want to leave open the possibility that their shared assumption is
mistaken and (positively) that Aristotle considers all of the investigations
mentioned at the beginning of the Meteorology to be contributions to a single
epistêmê, but that he nevertheless has serious concerns about the sufficiency of
methods of the other investigations of nature for the investigation of animals
and plants. If that is the case, and I will make the case that it is in Chapters 5
and 6, then the questions Nussbaum raises about the relation of the MA to
34
She cites Owen’s British Academy lecture, “The Platonism of Aristotle” (1966/1986), as support for
this claim, but Owen’s only explicit reference to the issue in that lecture is this: “[Aristotle] allows
that sometimes one science may take over and apply the arguments of another; but these are the
exceptions” (Owen 1986, 213).
35
An assumption that is explicitly acknowledged: “[MA] will not only use one science to get
knowledge about another, but also claim that the second provides a necessary part of any valid
justification of the principles of the first” (Nussbaum 1978, 112). The issue of interest to us is not so
much her concern that the MA seems to constitute an unwarranted mixture of sciences, but the
assumption lying behind that concern, that the different natural investigations will necessarily be
contributions to separate sciences.
36
Kung 1982, 65–76. Barnes’ report of this dispute (Barnes 1993, 131) is somewhat misleading. Cf.
Wilson 2000, 47–52, who endorses Kung’s response.
37
Ibid., 68. In fact, MA, after noting that the different movements found in different kinds of animals
and their causes have been studied in other places (an apparent reference to the investigations
reported in IA and perhaps HA), refers to itself as investigating “in general the common cause of being
moved with any motion whatsoever” (698a4-5).
38
This is not an uncommon assumption; for example, note in the following remark the assumption
that for Aristotle, physics and biology are distinct sciences: “The same thing may be treated by more
than one science. For example, man is treated qua a thing-that-is by first philosophy, qua a substance
whose principle of motion and rest is in itself by physics, and qua animal by biology” (McKirahan
1992, 63).
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
the Physics take on a different shape. The tenets of the APo. that Nussbaum
assumed to be violated are proscriptions on the appeal to the starting points
of one science in constructing demonstrations of facts investigated in another
science. Suppose, however, that Aristotle imagines each of these investiga-
tions to be contributing to a unitary science of nature – in that case, there is
no violation of the proscription against kind-crossing with which she was
concerned.39 On the other hand, Aristotle clearly thinks of these as distinct
investigations, and so a different question arises that is no less difficult: How
can Aristotle conceive of these many distinct inquiries as contributions to
one science, given what he says about how one secures unity for an
epistêmê in the Analytics?40
4.4.4 Why Subordination Is Not the Answer

The proscription that Nussbaum worried about is stated clearly at the
beginning of APo. i.7:
Thus you cannot prove by transferring [a proof] from another kind, for
example something geometrical by an arithmetic proof. (75a38-39)
As Barnes notes, the chapter division is particularly inept in this case, since
this statement follows directly from, and is grammatically linked to, the
closing sentences of the previous chapter. Aristotle there insists that the
terms in the demonstrations of a given science must be, and be known to
be, in one way or another, related per se to the kind (genos) studied by that
science. That is the basis for the proscription on ‘kind-crossing’ stated
above – the predicates of geometry are not appropriately related to the
subject of arithmetic (number, or perhaps discreet quantity). In a number
of critical passages, however, Aristotle notes exceptions, and Kung thought
this was where to look in understanding what was going on in MA. About
these exceptions, Aristotle says:
39
Problems would, of course, still arise if there were a serious role played by moral psychology or
geometry in a natural treatise, but for now I will restrict my focus to the role of principles from Ph.
viii in MA, and its discussions about the movement of the whole cosmos. Interestingly, in the
sixteenth and seventeenth centuries, a number of researchers in the Aristotelian tradition built on
the various hints toward geometric models of limb motion in MA and IA to work on creating a sub-
subaltern science of biomechanics – that is, applying geometrical mechanics beyond the realm of
simple machines to animal locomotion. Examples of such works include Hieronymus Fabricius ab
Aquapendente, De motu locali animalium secundum totum and an unpublished work by William
Harvey, recently edited and published also with the title De motu locali animalium (which appears to
be the title of the first chapter rather than of the entire work). For details about this creative
extension of Aristotelianism, see Distelzweig 2014a and 2014b.
40
This is simply to be skewered by the other horn of the dilemma posed on p. 111.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
[O]f those things the kind of which is different, as with arithmetic and
geometry, it is not possible to apply the arithmetic demonstration to the
[proper] incidentals of magnitudes.41 . . . The arithmetic demonstration
always possesses the kind about which the demonstration is concerned,
and other demonstrations do likewise. Therefore the kind must be the
same either without qualification or in a way, if the demonstration is to
transfer. (APo. i.7, 75b2-9)
There are two interesting implications of the conclusion here, and they pull
in opposite directions. First, there is apparently a qualified way in which
one kind can be the same as another that would allow some sort of kind-
crossing (e.g., being the same by analogy). As a general claim, this seems
unexceptionable: for example, the kind consisting of fish and that consist-
ing of birds are distinct, but are the same qua being egg-layers.42 In that
case, a demonstration regarding egg-laying in fishes might be transferable
to birds, provided the demonstration was restricted to attributes that each
has qua oviparous, as opposed to those attributes possessed qua bird or
qua fish.
A second implication of the last sentence in the quoted passage is that it is
possible to ‘transfer’ a demonstration even if we are discussing scientific
knowledge of a kind that is without qualification the same. It would seem,
then, that transferring demonstrations need not imply ‘kind-crossing.’ One
reason why Aristotle might want to leave open the possibility of demonstra-
tive transfer within a kind is that, within a science comprised of relatively
autonomous investigations, that might be the most appropriate way to
describe the relationships between inquiries, such as those reviewed in
Section 4.4.1. I will provide much more evidence for this possibility in
Chapters 8 and 9.
41
The verb used here for ‘apply,’ ἐφαρμόττειν, often expresses the relationship between a principle of
demonstration that is more ‘abstract’ or more general and one that is more ‘concrete’ or specific (cf.
APo. i.7, 75b42-76a7; i.9, 76a22-25; i.32, 88a32, 33). Passages in which the term seems to be playing
a similar semantic role (Cael. iv.2, 308b2; Resp. 7, 474a10-11; Ph. iii.1, 201b14; v.4, 228b25; Pol. iii.1,
1275a33-35, b32; iii.2, 1276b25) suggest translating it ‘to apply.’ Especially important for fixing its
meaning is MA 1, 698a11-14: “One should grasp this [the existence and nature of the first unmoved
mover] not only by a universal account, but also in the presence of the particulars and perceptibles,
on account of which we search for the universal accounts, and to which (ἐφ’ ὧν) we think these
accounts should apply (ἐφαρμόττειν).” Though it cannot be reflected in a readable translation, the
prepositional phrase ἐφ’ ὧν and the infinitive ἐφαρμόττειν are contiguous in the Greek, indicating
that the preposition is simply emphasizing the verb’s prefix. Given its use in the passages we are
looking at, I think Nussbaum’s ‘to harmonize’ is not precise enough; Aristotle likely has in mind the
application of universal accounts via proof to the particulars, since it was in order to understand
them that the search for the universal began.
42
At one level; Aristotle does, of course, see significant differences in the eggs of birds and fish (cf. GA
ii.1, 732b1-8; iii.1, 749a10-34).
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
This interesting possibility notwithstanding, in this section of
APo. i, the qualified sameness in kind on which Aristotle is focused is
that found in, to use the language of Ph. ii.2, “the more natural of the
mathematical sciences” (194a7-8): that is, optics, mechanics, astronomy,
and harmonics, the so-called ‘subordinate’ or ‘mixed’ sciences.
Demonstration does not apply to another kind, other than, as has been said,
geometrical demonstrations to mechanical and optical facts, and arithmet-
ical demonstrations to harmonic facts. (APo. i.9, 76a22-25)
However, leading up to this conclusion, he has apparently delimited the
class of exceptions in a way that prevents ‘subordination’ helping us with
the problem in dispute between Nussbaum and Kung:
Otherwise it will be like proving something in harmonics by arithmetic.
Such things are proven in a like manner, but the proof differs; for the fact is
from one science (for the underlying kind is different), while the reason why
is from the science above, to which the attributes belong per se. Hence from
these cases too it is apparent that it is not possible to demonstrate each thing
without qualification other than from the principles of that thing. But the
principles in these cases have something common. (APo. i.9, 76a9-15)43
It may be possible to extend this model beyond the cases of the subordinate
mathematical sciences; but if so, the extension must be to cases of two
distinct sciences where the causal demonstrations or principles from one
find application in the other.44 In the APo. these ‘subordinate’ sciences are
without exception treated by Aristotle as branches of mathematics, and
I read the lines just translated to be saying that the attributes that are
proven to belong to their subjects are, in these cases, those which belong
per se to the higher science (i.e., they are mathematical attributes) – and that
is why the geometer or arithmetician is the appropriate person to prove
them. That those attributes are present in a delimited natural domain is
a fact established by a different science, namely, natural science.
The extended discussions of these ‘exceptions’ to the model of the unity
of scientific knowledge in APo. i.7, 9, and 13, then, deliver a clear message:
they concern situations in which certain mathematical demonstrations can
be mobilized to explain why a certain range of mathematical attributes
belong to objects that fall under the domain of natural science. These
43
The thought here is followed out in much more detail in APo. i.13, 78b34-79a16.
44
Elsewhere I have argued that three passages in Aristotle (in the opening and closing chapters of the
Parva naturalia – Sens. 1, 436a18-b2, Resp. 21, 480b22-31 – and PA ii.7, 653a1-10, suggest that he did
see the relationship between medicine (a productive science) and the science of nature on the
subordinate science model (see Lennox 2005, 55–72, esp. 66–68).
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
discussions are not intended to answer the question of how to understand
the relationships among the various investigations of a single science,
though there is at least one hint in that direction.
4.5 Epistêmê, Zêtêsis, and Methodos: A Hypothesis

We are told in APo. i.28 that a single science is a science of one kind, and
that the kind must be composed from primary things and their parts and
per se attributes (87a38-b1). Metaph. Ε.1 apparently identifies such a kind to
which the science of nature is devoted, a kind consisting of substantial
beings that have both material and formal natures and their own inherent
principles of change and rest. Certainly, animals and plants fall within that
kind, and so there would appear to be no doubt that their study will be part
of a unified science of nature. However, at some point it seems that
Aristotle realized that the goal of a unified science of that kind was not to
be achieved by means of a single, undifferentiated method of investigation.
At that point, he became quite self-conscious of the differences in prin-
ciples and methods required for the pursuit of knowledge (i.e., inquiry)
about different aspects of the natural world, and from that point on the
question of the unity of natural science became his concern, as it is ours.
In this chapter, I have focused on a number of philosophical problems
that arise from the fact that Aristotle conceives of the science of nature as
unified at a fundamental level and yet argues that natural inquiry should be
carried out as a number of somewhat autonomous, yet self-consciously
integrated, investigations. Aristotle was no reductionist: there is just too
much diversity in the natural world for a single uniform set of norms to do
the trick.
Our next step is to consider how Aristotle conceives his various natural
inquiries to be both autonomous and integrated, and whether the form of
integration he imagines can satisfy his own requirements for a single,
unified science. This step will involve a careful study of the many meth-
odologically reflective passages in Aristotle’s scientific treatises themselves.
That is the task of Part ii.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
p a r t ii
Norms of Natural Inquiry
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6
C 1D19 12 1 8 31 2 947 73 8 4 9 7
chapter 5
The Methodos of Nature
Since the nature [of a natural thing] is a source of motion and change,
and our methodical inquiry is concerned with nature, [the question]
what is motion must not escape our notice; for necessarily when we
are ignorant of this we are also ignorant of nature.
(Ph. iii.1, 200b12-15)1
So then, about the being of sensible things, what it is has been stated,
in the methodical inquiry of the naturalists about the matter, and later
about the <being> in accordance with activity; but since the investi-
gation is whether there is some being besides the sensible beings that is
immoveable and eternal or not, and if there is, what it is, we must first
consider what is said by others.
(Metaph. Μ.1, 1076a8-14)2
Chapter Summary. Part i made a case for seeing the Posterior Analytics
ii in a new light. In that new light, it does not provide poor answers to
questions people bring to it about the path by which inquiry takes us from
perceptual experience to first principles of demonstration. Rather it pro-
vides very good answers to questions about what questions ought to shape
any inquiry aimed at knowledge and about how the answers to those
questions are related to one another – it provides, that is, an erotetic
framework for inquiry. Many of the norms that guide scientific inquiries,
however, are domain-specific, and are not to be found in APo., though they
are best seen as different ways of specifying the framework provided by the
Analytics. Part ii of this book tests these ideas in the context of specific
1
Ἐπεὶ δ’ ἡ φύσις μέν ἐστιν ἀρχὴ κινήσεως καὶ μεταβολῆς, ἡ δὲ μέθοδος ἡμῖν περὶ φύσεώς ἐστι, δεῖ μὴ
λανθάνειν τί ἐστι κίνησις· ἀναγκαῖον γὰρ ἀγνοουμένης αὐτῆς ἀγνοεῖσθαι καὶ τὴν φύσιν.
2
Περὶ μὲν οὖν τῆς τῶν αἰσθητῶν οὐσίας εἴρηται τίς ἐστιν, ἐν μὲν τῇ μεθόδῳ τῇ τῶν φυσικῶν περὶ τῆς
ὕλης, ὕστερον δὲ περὶ τῆς κατ’ ἐνέργειαν· ἐπεὶ δ’ ἡ σκέψις ἐστὶ πότερον ἔστι τις παρὰ τὰς αἰσθητὰς
οὐσίας ἀκίνητος καὶ ἀΐδιος ἢ οὐκ ἔστι, καὶ εἰ ἔστι τίς ἐστι, πρῶτον τὰ παρὰ τῶν ἄλλων λεγόμενα
θεωρητέον.
121
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
122 The Methodos of Nature
inquiries within the science of nature. However, I argued in Chapter 4 that
Aristotle sees each of these somewhat autonomous natural inquiries as
contributions to a single science of nature. Not surprisingly, throughout
the Physics, and in other natural investigations as well, Aristotle refers to
a methodos of nature. Before turning in later chapters to these more specific
natural inquiries, then, I will here explore what Aristotle has to say about
a methodical inquiry into nature generally, building on some of the results
from Chapter 4 of Part i. We will then turn to various specific natural
inquiries to see how their different norms and consequent methods of
inquiry conform to the general methodos of nature.
5.1 Introduction
As I will discuss in some detail shortly, the opening sentences of the Physics
strongly suggest that the erotetic framework for inquiry in the pursuit of
scientific knowledge presented in the second book of the Posterior Analytics
shapes Aristotle’s views about natural inquiry. Assuming that is the case,
there are a number of features one would expect to find. In light of the
framework discussed in Chapter 2, there ought to be two intimately related
goals of natural inquiry: knowledge of what natural beings are, ideally to be
formulated in definitions; and knowledge of causes used to demonstrate
why natural beings have the necessary but nonessential attributes that they
do; and fact-establishing stages on the way to each of those goals.3 The
intimate relationship between these two paths of inquiry stems from
Aristotle’s conviction that the essences of things are, in various ways,
causally responsible for their nonessential features – if one has scientific
knowledge of what something is, one will thereby be able to explain why it
has the other nonaccidental features it has.
Nevertheless, notwithstanding the generality of this framework, I have
been arguing that Aristotle sees different subject matters or domains as
governed by norms that are quite specific to them. This specificity derives,
from (at least) three sources: (i) differences in the objects being investi-
gated, (ii) differences in our epistemic access to those objects, and (iii)
differences in the perspective we take on those objects. Thus, while the
Posterior Analytics provides what I have called an erotetic framework for
scientific inquiry, it does not provide the domain-specific norms needed to
guide actual research.
3
See APo. ii.1, 2. The view is, of course, much more complicated than this brief summation suggests.
For detailed discussion of the complications, see Charles 2000, 2010; Ferejohn 2013; Lennox 2005.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
5.1 Introduction 123
What, then, are the domain-specific norms that Aristotle thinks guide
natural inquiry, qua natural, and distinguish it from the other two forms of
theoretical investigation? The Physics opens, like so many of Aristotle’s
inquiries, with a general claim about every methodical inquiry and then
applies what is said generally to natural inquiry:
Since, concerning every methodical inquiry (μέθοδους) of which there are
starting points, causes, or elements, knowing, that is, scientific knowing (τὸ
εἰδέναι καὶ τὸ ἐπίστασθαι) comes about from knowing (γνωρίζειν) these
things (for we think we have knowledge (γιγνώσκειν) of each thing when
we know the primary causes and the primary starting points and even to the
elements), it is clear that, concerning the scientific knowledge (ἐπιστήμης)
of nature too we must first attempt to determine the facts about the starting
points. (Ph. i.1, 184a10-16)
The very next sentence highlights the idea that the discussion to follow will
be concerned, in part, with the proper way to proceed:
And the path is by nature (ἡ ὁδος . . . πέφυκε) from things more knowable
(γνωριμωτέρον) and clear to us to the things that are by nature (φύσει)
clearer and more knowable; for what is knowable to us and what is knowable
without qualification are not the same. (184a16-18)
What is being sought is a path toward knowledge of the starting points,
causes, and elements of natural things; by achieving these we can
achieve scientific knowledge of nature. And the natural way to proceed
is to start with what is better known and clearer to us and proceed
toward what is better known and clearer ‘by nature’ or without qualifi-
cation, that is, without epistemic relativization to the state of the
knower.
This opening, then, establishes that the framework of the Analytics is in
place, and natural inquiry will certainly take place within that framework.4
At the same time, however, it tells us nothing specific about what distin-
guishes natural inquiry from other theoretical pursuits.
It is, then, still an open question what is the proper method for natural
inquiry. I shall approach an answer to this question by considering what
Ph. ii.2 says about distinguishing the inquiries of the mathematician and
the natural scientist and about the unity of natural science. I will then
consider the rest of book ii in light of what it says about how one will need
to proceed in order for one’s inquiry to be physikôs (i.e., appropriate for an
inquiry into nature).
4
Cf. APo. i.2, 71b33-72a5; EN i.4, 1095b2-4.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
As we saw in Chapter 4 (Section 4.2), Aristotle has a good deal to say in
Metaph. E.1 about what is distinctive about scientific knowledge of nature
(hê physikê epistêmê, 1025b18-19). We were told that “it is concerned with
the sort of being in which there is an inherent source of motion and rest”5
(1025b18-21). This is the feature of natural things already alluded to in this
chapter’s epigraph from Ph. iii.1 – we need to investigate what motion is
because natures are sources of motion in natural things.
Metaph. Ε.1 also insists that natural things are among those spoken of “in
the manner of the snub” rather than “in the manner of the concave.” The
accounts of such things always include references to change and to matter,
and in that chapter Aristotle alleges that it should be clear both how one
ought to inquire into natural things and how to define them – as well as why
natural inquiry will include investigating as many features of the soul as are
not without matter (1026a1-5). Unfortunately, apparently because he
thinks it ought to be clear how to inquire into natural things so under-
stood, he does not actually spell out the implications of this ‘snub-like’
characterization of the objects of natural science for how one ought to
inquire into them. That is, however, the central focus of Ph. ii.2, and the
‘snub-like’ character of the objects of natural investigation is Aristotle’s
starting point in that chapter, to which we now turn.
5.2 Physics ii.2 and ‘Snub-like’ Inquiries6

Ph. ii.2 begins with the assumption of one major result of Chapter 4,
namely the number of ways in which nature is spoken,7 and turns to
a question he apparently thinks follows more or less directly from that
result – in what way does the mathematician differ from the person who
investigates nature. Aristotle generates a concern about this by highlighting
two related but distinct points: first, the objects investigated by the math-
ematician, such as points, lines, planes, and solids, are features of natural
bodies; and second, it is apparently a goal of the natural scientist to grasp
the nature of the sun and moon, and to determine whether the earth or the
5
For the sake of consistency, I will render κίνησις as ‘motion’ and μεταβολή as ‘change,’ but ‘motion’
will thus be referring to a variety of processes besides local motion. Cf. Ph. iii.1 and v.1–3 for
Aristotle’s distinction between μεταβολή and κίνησις and discussions of the various kinds of change.
6
I will, to some extent, be assuming the results of Lennox 2008a here, which is almost entirely devoted
to an interpretation of this chapter.
7
Reading λέγεται at 193b22. Later in the chapter, at 194a12, we are told that nature is διχῶς, τό τε εἶδος
καὶ ἡ ὕλη, a result Aristotle argues for from 193a9-b5, after which he makes the case for the form being
a composite’s nature ‘more than’ the matter.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
5.2 Physics ii.2 and ‘Snub-like’ Inquiries 125
cosmos is spherical or not. But these would appear to be the concerns of the
astronomer, who is a type of mathematician (193b22-30).
As he proceeds to respond to these problems, it becomes clear that
a major concern at this point in the Physics is to begin to make a case for
the unity of the objects being investigated by the natural scientist, as
a precondition for there being a single science of nature. This becomes
critical once Aristotle has reached the conclusion of chapter 1 that the
nature of a natural thing is, and is spoken of, in two ways, as matter and as
form.8 Almost immediately he points to the errors of “those who speak of
Ideas” (193b36), indicating his concern to be clear about what it means to
say that the natural investigator should be pursuing knowledge of the forms
of natural things. He precisely criticizes the friends of Ideas for attempting
to separate the natural things,9 which are “less separable than the mathe-
maticals” (193b36-194a1). He explains:
This would become clear if one were to attempt to state the definitions of
each – both of the objects and of their attributes. For the even and odd and
the straight and curved, as well as number, line, and figure will be [defined]
without motion, while flesh, bone, and human being will not – these are
spoken of as snub nose is (ῥὶς σιμὴ), not as the concave. (Ph. ii.2, 194a1-7)10
As in Metaph. E.1, here the contrast with mathematical definitions is
intended to highlight the fact that the definitions of natural things make
reference to change – and once again, they are said to be referred to in the
way one refers to a nose as snub, not as concave. But what exactly are we
referring to when we refer to a natural object in that way? This is spelled
out a bit more fully in Metaph. Z.11.
And the comparison that Socrates the younger used to draw between an
animal and a circle is not sound; it misleads one into supposing that there
might be a man without his parts, as there can be a circle without bronze.
But the cases are not the same. For an animal is a perceptible object, and
cannot be defined without motion, nor therefore without reference to the
state of its parts. (For it is not the hand in any state that is a part of a man,
but the one capable of fulfilling its function, and thus is ensouled. And what
is not ensouled is not a part of a man.) (1036b24-32)11
8
I make the case for this understanding of the argument in Lennox 2008a.
9
That is the precise language he uses, but I assume he means the natures of natural things (i.e., the
forms of natural things).
10
Cf. Metaph. Z.10, 1035a4-5, Z.11, 1036b21-31.
11
This passage is controversial, of course. For a review of one central controversy with a valuable
discussion of the parallels I am relying on between Metaph. Ε.1 and Ph. ii.2, see Peramatzis 2011,
Appendix 1, 312–316; cf. Bostock 1994, 162–165; Gill 1989, 130–134.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
It is of the very essence of a living thing, and more generally of any natural
object, to move, behave, act, change in specific ways – in fact in specifying
what it is to be a hand or a leaf – or even air – one must mention capacities
to function or change in specific ways.12 But it is also the case that those
capacities are grounded in the material nature of the living thing or its
parts, and thus any such definition must refer to matter constituted in
precisely the way it must be to have the capacities to move and change as it
does.
And this has implications for how one carries out a natural inquiry. A bit
later in Z.11 Aristotle alludes to these implications in the process of
explaining why it is that first philosophy is bothering to discuss natural
composites.
It is in fact for the sake of this [preparing ourselves to consider whether there
is another sort of matter or substantial being besides natural ones] that we
are attempting to define perceptible beings too, since the study of percep-
tible beings is in a way the task of the study of nature, or second philosophy.
For the natural investigator must have knowledge not only of the matter of
things but also, and more especially, of the being according to the account.
(1037a13-17)
This is a discussion of why the person interested in being qua being should
be thinking about natural substantial beings, but in doing so it underscores
that there is a special, dual focus to natural inquiry. With that in mind, let
us return to the argument of Ph. ii.2. Aristotle once again invokes the
analogy of natural substantial beings and the snub, but now as an implica-
tion of the fact that natural things have a twofold nature.
Since the nature [of natural substantial beings] is in two ways (διχῶς), both
the form and the matter, we should study as if we were investigating, about
snubness, what it is; that is, we should study such things neither without
matter nor in accordance with matter. (194a12-15)13
The message appears to be this: natural things have a nature in two ways –
as he says in a number of places, such beings have a material nature and
a formal nature.14 For that reason, we need to investigate the ‘what is it’ of
12
Cf. Mete. iv.12, 390a10-12: “Everything is defined by its function. For things able to perform their
own function are truly what they are in each case, for example, an eye if it sees; whereas one that
cannot is an eye in name only, for instance, a dead eye or one made out of stone.”
13
The interpretation of this chapter is defended more fully in Lennox 2008a; for a different reading see
Peramatzis 2011, 59–64, 130–137; our disagreements are highlighted at 62 n. 5, 65, and 130.
14
Aristotle identifies the two natures in relation to one another in various ways: at Ph. ii.1, 193a28-31,
the contrast is of the primary underlying matter with the shape and form according to the account
(cf. Ph. ii.1, 193b3-5); at PA i.1, 640b28-29 it is of the nature in respect of form with the material
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
natural things while keeping their twofold nature in focus; we cannot
ignore their matter, but we cannot focus only on their matter, for they
have a formal nature as well – indeed, Ph. ii.1 gives a series of (more or less
compelling) arguments for giving more attention to the formal nature. It is
somewhat surprising, then, that he immediately raises an aporia about
which of the two natures the natural scientist should study! Have we not
just been told that he should study both? The full statement of the aporia is
worth reviewing:
For indeed one might also raise a problem about the following question:
since the natures of things are two, which of the two is it for the natural
scientist to study – or is it rather for him to study that which derives from
both natures? But if [it is for the natural scientist to study] that which is
derived from both, [then it is] also [for him to study] each of them. (Ph. ii.2,
194a15-17)
It is prima facie odd that this is treated as an aporia, but I think we can
understand why it is by asking why he might think it would be a problem
that the natural scientist, though not treating form as separate from matter,
would still be studying the two natures in separation. The question at issue
is, can these two very different natures, even if not ontologically separate, be
studied by a single science (epistêmê), or do they require distinct inquiries
involving distinct methods? That question continues to have purchase
because, within an Aristotelian framework, the objects of mathematics
are also not ontologically separate from natural objects – and yet they are
‘separable in thought from change’ and are studied by distinct sciences,
sciences that use very different methods guided by very different norms
from those of the natural scientist.
To this point, Aristotle has defended the following claims: (i) natural
substances have two natures; (ii) their attributes are spoken of like snub
nose and are not defined without reference to change; and (iii) we are to
study them ‘not without matter yet not according to matter.’ He has not
yet made a case for why we should study them in this way, nor given us any
sense of how a single science operating according to a single method would
do so. Finally, however, after noting that certain of his predecessors
stipulated that the natural scientist was simply to look to the matter of
natural things, and grudgingly acknowledging that Empedocles and
nature; at PA iii.2, 663b22-24, the nature in respect of the account is contrasted with the necessary
nature; and at GA iv.4, 770b9-18, the nature in respect of the form is contrasted with the nature in
respect of the matter.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
Democritus “touched to a small extent on the form and essence,” he
presents his solution, albeit in conditional form:
If art imitates nature, and it is for the same science (ἐπιστήμη)15 to know the
form and the matter to a certain extent (e.g., for the doctor to know health
and bile and phlegm, in which health is present, and similarly also for the
house builder to know both the form of the house and the matter, that it is
bricks and wood, and likewise in other cases), then it would also be for
natural science to know both natures. And further: it is for the same science to
know that for the sake of which and the goal and as many things as are for the
sake of these. (Ph. ii.2, 194a21-27; emphasis added)
The knowledge that the natural scientist seeks is of both the matter and the
form; and they can be the subject of one science, on the model of the crafts,
if the materials are both present for the sake of, and defined by reference to,
the formal nature. It is because of the teleological unity of matter and form
that, though natural things have two natures, they are properly the subject
matter of a single science.
Aristotle picks up on the qualification ‘to a certain extent’ in the above
passage, near the end of Ph. ii.2:
To what extent is it necessary for the natural scientist to know the form and
the what-it-is? Is it just as [it is necessary] for the doctor [to know] sinew or
the sculptor bronze, to the extent of knowing what each thing is for the sake
of, and concerning those things which are separable in form, yet in matter?
(194b9-13)
It appears that the issue of the extent to which the two natures are to be
studied by the natural scientist is only explicitly addressed with respect to
form, a point I will return to in a moment. Before doing so, it is important
to recall the wider context of this question, specifically issues concerning
the demarcation of natural philosophy from first philosophy. Charlton
translates the chapter’s closing sentence thus: “What it is which is separ-
able, and how things are with it (πῶς δ’ ἔχει τὸ χωριστὸν καὶ τί ἐστι), it is
the work of first philosophy to determine” (194b14-15). However one
decides what Aristotle means by the Greek here, one thing is certain: he
is making this point because he has just made a claim about the limited
nature of the separability of the forms to be investigated by the natural
scientist.16 And this passage surely looks back to the close of book i, where
15
Here ἐπιστήμη is being used in a ‘big tent’ sense to include the crafts.
16
Some other options for translation are Waterfield (1996): “Questions remain – in what sense is
anything separable? What is it that is separable? – but it is the job of first philosophy to answer
them”; Hardie and Gaye (1930): “The mode of existence and essence of the separable it is the
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
Aristotle says that the discussion to follow (in book ii) will investigate the
forms of natural and perishable things,17 while it will be “the work of first
philosophy to discuss with precision the principle in respect of form,
whether it is one or many and what it is or they are” (Ph. i.9, 192a34-36).
In any case, what this passage does tell us is that we are to study the
formal nature in so far as the formal nature is that for the sake of which the
materials that make up the material nature are present. Thus it is also
telling us the extent to which the natural scientist is to study the matter –
to the extent that it is for the sake of the form. This chapter, then, looks
forward to Ph. ii.8 and its defense of the teleological dependence of
material nature on formal nature, succinctly summed up in the follow-
ing sentence:
And since the nature [of a thing] is twofold, on the one hand as matter and
on the other as form, and the nature as form is an end, while the others are
for the sake of the end, this [nature as form] would be the cause for-the-sake-
of-which. (199a30-32)18
This passage in ii.8, moreover, looks forward to the discussion of the
relationship of conditional necessity between materials and what they are
for in ii. 9, an aspect of this way of understanding the task of the natural
scientist to which I will return later.
We now begin to make contact with norms that govern Aristotle’s actual
practice as a natural investigator, such as this one, presented while review-
ing his predecessors’ attempts to understand the nature of respiration:
The most significant reason for [previous thinkers] not discussing these
[questions about respiration] well is a combination of (τε) their being
inexperienced with the internal parts and (καὶ) not grasping that nature
makes them all for the sake of something; for seeking what respiration
belongs to animals for, and examining this question in the presence of the
parts (e.g., in the presence of gills and lungs), they would have discovered
the cause more easily. (Juv. 9/Resp. 3, 471b23-29)19
business of the primary type of philosophy to define”; Pellegrin (2000): “Ce qu’il en est du separable
e ce qu’il est, c’est l’affaire de la philosophie première de le determiner”; Angioni (2006): “Mas
delimiter como se comporta e o que é o separável é tarefa da filosofia primeira.” Certainly, the nature
of the separability of the objects of mathematics, Platonic Forms, the forms of natural substances,
and a first unmoved mover is a theme that runs through the entirety of the Metaphysics.
17
This reading is defended in Lennox 2008a. It is usually translated ‘natural and perishable forms,’ but
since Aristotle consistently denies that forms (as opposed to composites) are perishable, I prefer to
follow the Λ mss. and translate as above.
18
καὶ ἐπεὶ ἡ φύσις διττή, ἡ μὲν ὡς ὕλη ἡ δ’ ὡς μορφή, τέλος δ’ αὕτη, τοῦ τέλους δὲ ἕνεκα τἆλλα, αὕτη
ἂν εἵη ἡ αἰτία, ἡ οὗ ἕνεκα.
19
We will return to this passage in Chapter 10.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
That is, one must carry out a natural investigation of a living activity like
respiration on the assumption that it, and the structures that perform it, are
present in animals for the sake of something; and one must carry out such
an investigation making use of comparative dissection so that you have
experience with the internal parts that are involved in performing the
activity of respiration. Recalling an oft-cited passage from the beginning
of Aristotle’s investigation of animal locomotion, we should not be sur-
prised by what Aristotle says about the study of respiration, for he there
claims that this is an assumed starting point of natural inquiry generally:
The starting point of our investigation is achieved by supposing principles
we are accustomed to use often in natural inquiry (πρὸς τὴν μέθοδον τὴν
φυσικήν) – assuming this is the way things stand in all the works of nature.
One of these principles is that nature does nothing in vain, but always, given
the possibilities, does what is best for the substantial being of each kind of
animal (τῇ οὐσίᾳ περὶ ἕκαστον γένος ζῴου τὸ ἄριστον) . . . (IA 2,
704b12-17)
The assumed starting points Aristotle is introducing here are starting
points of investigation (not of proof) and are said to be used often in the
methodos of nature. Yet the first one that is stated, the nature-does-nothing-
in-vain principle, is stated in a way that is specific to animals. There is, of
course, absolutely nothing inconsistent about that – a principle that is
often used in the investigation of animals will, by that very fact, be used
often in natural inquiry. But he also says that it is appropriate to assume
that these principles (he lists two others) are applicable to “all the works of
nature.” He could mean that they collectively apply to all works of nature,
or he could simply be allowing that, at the beginning of an investigation of
nature, we have no grounds for restricting these principles to a limited part
of nature.20
5.3 Implementing the Physics ii.2 Program for Natural Inquiry

In Physics ii.2, the primary focus is on the question of whether, given that
natural substances have both material and formal natures, they could be
investigated by a single science – where the goal of such an investigation is
specified in terms of defining and knowing the essence of such things.
However, the possibility of a unified science of natural, ‘snub-like’ beings
turned on there being a teleological dependence of their material nature on
20
Cf. Lennox 2009, 187–214, and Leunissen 2009, 215–237.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
5.3 Implementing the Physics ii.2 Program 131
their formal nature. Indeed, the natural scientist’s knowledge of the matter
and form of a natural substance is restricted to knowing what the material
constituents of a natural being are for, which means knowing the form in so
far as it is that for the sake of which the matter is as it is. The chapter thus
concludes by implicitly grounding the ‘what is it?’ inquiry in a causal
inquiry, the sort that is in focus in chapter 9, an inquiry into the causal
relationship between ends, on the one hand, and materials and their
movements, on the other. But that is precisely what the Erotetic
Framework provided by APo. ii would have led us to expect – that the
search for an answer to the ‘what is it?’ question will be intertwined with
causal inquiry in precisely this way.
5.3.1 Physics ii.3: A Causal Framework for Natural Inquiry

The implication of the solution to the final aporia of Ph. ii.2, then, is that
the investigator of nature will need to investigate matter, form, and end in
an integrated way – and nature is also an inherent source of change or rest
as well. With that in mind, recall how chapter 3 begins, echoing the opening
sentence of the Physics (which in turn echoes APo. i.2 and ii.11).
Having determined these things, we must investigate concerning the causes,
both of what sort and how many in number they are. For since our study is
for the sake of knowing, and we do not think we know each thing until we
grasp that on account of which (τὸ διὰ τί) about each thing (and this is to
grasp the primary cause), it is clear that we must do this concerning
generation, destruction, and all natural change, so that knowing their
starting points we may try to bring each of the objects of inquiry (τῶν
ζητουμένων) back to them. (194b16-23)
However, though it is introduced as important for knowledge of nature,
the discussion of the four types of cause in this chapter,21 and as far as that
goes the next three chapters on chance as well, has little to say that is specific
to natural inquiry; and in fact, chapter 3 is, as Ross puts it, “a doublet which
corresponds in all except small details with Met. Δ.2.”22 Virtually all the
examples used to illustrate the four causes are drawn either from mathem-
atics or the arts. It is not until chapter 7 that Aristotle explicitly relates the
idea that there are four distinct kinds of cause to the context of natural
21
For a valuable overview of Aristotle’s account of the four causes and the various attempts to make
sense of it, see Johnson 2005, 40–63.
22
Ross 1936, 511. Ross’ claim applies only to the discussion of the causes; the introduction of the
chapter quoted above, relating the APo. i.2 assertion that scientific knowledge is knowledge of causes
to the investigation of nature, is unique to Ph. ii.3.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
inquiry – and what he says there is of great importance for understanding
how he distinguishes the norms that govern the inquiry into nature from
other inquiries.
Before turning to that chapter, however, there is one other feature of the
discussion of causes in chapter 3 that needs to be highlighted – its stress on
the importance of properly identifying the cause in light of the way in
which the effect is identified. He introduces this issue by distinguishing
how many causes there are differing in form (the familiar four) from the
many modes (polloi tropoi) of causes, which he spends the next thirty or so
lines of text bringing under fewer heads (195a27-b21). He sums up the
results in the following terms:
One ought always to inquire into the uppermost cause of each thing
(δεῖ δ’ ἀεὶ τὸ αἴτιον ἑκάστου τὸ ἀκρότατον ζητεῖν), just as with other
things (for example, a man builds because he is a builder, and the builder in
virtue of the building art; so then, this is the prior cause, and that is the way
it is in all cases). Furthermore, one should inquire into the kinds of causes of
kinds of things, and particular causes of particular things (for example,
a sculptor is the cause of a statue, but this sculptor of this particular statue);
and potential causes of possible things, and active causes of things being
activated (τὰς μὲν δυνάμεις τῶν δυνατῶν, τὰ δ’ ἐνεργοῦντα πρὸς
τὰ ἐνεργούμενα). So then, let these distinctions suffice regarding how
many causes there are and their modes of being causes. (Ph. ii.3, 195b21-30)
Notice that this summary opens with a sentence worded in the now familiar
language of a norm of inquiry: we are not yet sure what Aristotle means by
‘the uppermost cause of each thing’ (τὸ αἴτιον ἑκάστου τὸ ἀκρότατον), but
whatever he means by it, it is this cause which we ought to search for in our
inquiries. We do learn, by virtue of his presentation of the examples,
something about what he has in mind: after running through the first
example, he characterizes the relationship among the ways of identifying
the causes in terms of priority:23 since it is the cause of a building that is being
sought, identifying the cause simply as a human being does not highlight
that in virtue of which the human being is a cause of a building; identifying
him as a builder is a step in the right direction – but he is a builder in virtue of
possessing the art of building. Identifying the building craft as the cause is
prior in the sense of being that in virtue of which the human being who is
a builder is the cause of the building.
Aristotle next stresses the need for a match between the level of univer-
sality or particularity of the cause and effect: sculptors produce statues,
23
Cf. 195a29-30.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
Auguste Rodin produced The Gates of Hell and The Thinker. Presented
this way it can seem like a somewhat trivial (because obvious) point –
but in Aristotle’s scientific writings he regularly notes that natural
scientists have made serious mistakes by mismatching the level of uni-
versality of the cause with that of the thing to be explained. As one
example, in GA ii.8 he notes that Democritus made just such an error by
arguing that mules are sterile because they are hybrids. No, Aristotle
insists, many hybrids are not sterile; the cause must be something
distinctive to mules that explains their sterility. He then captures the
philosophical point in terms that lead directly to a normative recom-
mendation about inquiry:
This account is too universal and empty (καθόλου λίαν καὶ κενός). For
accounts that do not derive from appropriate starting points (μὴ ἐκ τῶν
οἰκείων ἀρχῶν) are empty – they seem to be about the subject matter, but
they are not. For accounts from geometric starting points are geometric, and
the same is true in other domains as well: that which is empty seems to be
something, but is nothing at all. [The account of Democritus] is not true,
because many of those animals that come to be from animals not the same in
form are fertile, just as we said previously. So then: one should inquire (δεῖ
ζητεῖν) in this way neither about other things nor about natural things; one
studying from facts about the kind consisting of horses and asses is more
likely to grasp the cause. (748a7-16)
From here to the end of the chapter, he presents a long list of facts about
these kinds that he thinks are much more likely to point us to the cause of
the infertility of mules.
I have translated archê in this passage as “starting point” since Aristotle
seems to be focused at least as much on the process by means of which we
arrive at appropriate starting points from investigating facts at the right
level of generality as he is on proofs that derive from such starting points.
And in fact, from 748a17-749a7 we see Aristotle engaged in finding the
cause of the sterility of mules by considering facts peculiar to the sires and
mares that give birth to them.
The summation of this discussion of the modes of causality in Ph. ii.3
draws a parallel between matching levels of universality of cause and
effect and matching the modalities of cause and effect: seeking potential
causes for potential effects and actual causes for actual effects (195b27-28;
cf. 195b6-7, 16-21). This entire chapter, in sum, is a discussion of norms
that should govern the search for causes – but a discussion with virtually
no mention of norms that are specific to natural inquiry. Virtually all
the examples of causality are drawn either from craftsmanship or
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
mathematics – odd, for a treatise devoted to the foundations of natural
science.
It will be recalled, however, that searching for attributes that are com-
mensurately universal (which Aristotle occasionally refers to as primary
universals) is a recurring theme in the Posterior Analytics, and the aporia
that governs the discussion of APo. ii.16–18 is “whether, concerning a cause
and that of which it is the cause, when the effect is present, the cause is also
present (just as, if shedding leaves or eclipsing is present, the cause of being
eclipsed or of shedding leaves also will be)” (98a35-38). In the above
example from GA ii.8, we see on display just how important this norm is
in practice. But the discussion in Ph. ii.3 is as domain-neutral as that in
APo. ii.
5.3.2 Physics ii.7: Investigating Physikôs

Chapter 7 begins by reminding us of the four kinds of answers to the ‘on
account of which?’ (to dia ti) question that were discussed in chapter 3 –
and once again, with no particular specificity to nature (198a14-21). He
then proceeds to make the connection:
So then, that the causes are these and this many is apparent; and since the
causes are four, it is for the natural scientist to know about them all, and by
bringing [accounts] back to all of them – the matter, the form, the mover,
and that for the sake of which – he will display that-on-account-of-which
(τὸ διὰ τί) in a way appropriate to natural science (φυσικῶς). (198a21-24;
emphasis added)24
Notoriously, he goes on to say the last three causes often come to one – and
then, substituting ‘the what-it-is’ (τὸ τί ἐστι) for ‘the form’ (τὸ εἶδος), he
explains what he has in mind. In cases of substantial generation, the what-it-
is and that-for-the-sake-of-which are one, while the primary source of
motion is the same as these in form (198a24-26). The goal of generation
is also the form, so in these cases there is referential (and numerical)
identity between them; while the parent, though not numerically (and
24
Cf. Metaph. H.4, 1044a32-b20. Importantly, in this passage, after citing the four causes and saying
that in the study of generated natural substances it is necessary to discuss all the causes, he states that
we should proceed by ‘another logos’ when discussing the natural and eternal substances (1044b2-6)
and that certain natural processes, such as eclipses, perhaps do not have final causes (1044b9-13). In
fact, no final cause is ever provided in the Meteorology for any of the objects and processes discussed
there. Cf. Wilson 2013, 93–107. Despite this, Scharle 2015 makes a valiant effort to argue that “there
is nonetheless an important explanatory role for teleological processes to play in the treatise [i.e.,
Meteorologica]” (79).
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
thus not materially) identical with the offspring, is the same in form. There
are a number of possible referents of the expression “primary source of
motion” – Aristotle immediately restricts its reference to movers that
“move while being moved” (a27) and goes on:
[F]or as many as are not like this are not natural; for they move without
having motion or a source of motion in themselves, but are immoveable; for
which reason there are three subjects of study – that concerned with the
immoveable, that concerned with what moves but is imperishable, and that
concerned with perishable things. (Ph. ii.7, 198a28-31)
This way of dividing up theoretical investigations is quite different from that
in Metaph. E.1; if the last two subjects mentioned correspond to a study of
the heavenly bodies (which, while imperishable, partake of eternal circular
motion) and a study of sublunary natural objects, these are both appropriate
objects of natural inquiry.25 The first-mentioned subject area is described in
such general terms that it could refer to a number of quite different things,
including objects of desire and the heavenly unmoved movers, both of which
are sources of motion for other things while being themselves unmoved.26 It
is best, I think, to read this as having the limited purpose of clarifying what
sort of first mover the natural scientist can investigate.
Aristotle notes that it is common practice to display or exhibit that on
account of which by tracing the question back to the matter, what a thing
is, and what first produced motion or change (198a31-35). But he seems to
be concerned that the way in which this is typically done is inappropriate
for the study of substantial generation.
For concerning generation people usually investigate the causes in this way,
[investigating] what comes to be after what, what acted first and what was
acted on, always in this way investigating what comes next. But the sources
that move things naturally (φυσικῶς) are of two sorts, one of which is not
natural; for it does not have a source of motion in itself. (Ph. ii.7, 198a33-b1)
Things that are by nature have a source of motion or rest within themselves
(192b13-15, 20-23, 32-34). He postulates two sorts of things that fall into the
category of things that are sources of natural motion without being natural
in that sense: one of these is not to be investigated by the natural scientist,
25
Cf. PA i.5, 644b22-24: “As many substantial beings as are constituted by nature, some are
ungenerated and imperishable for all eternity, while others participate in generation and perishing.”
See too Cael. i.1, 268a1-6, Metaph. H.4, 1044b3-6.
26
The discussion in Metaph. E.1 depends on comparing the objects of mathematics, natural science,
and a still hypothetical first philosophy along two axes, separability and changeability, whereas here
separability is not discussed. Recall that in Ph. ii.2 he refers that topic to first philosophy.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
presumably the prime unmoved mover;27 the other is the form viewed as
final cause – “the what-it-is and form, for it is an end and for the sake of
which” (198b3-4). He goes on:
So since nature is for the sake of something, and it is necessary to know this,
it is also necessary to display that-on-account-of-which (τὸ διὰ τί) in all
ways, for example that this comes from this necessarily (and from this either
without qualification or for the most part); <what must be> if this is to be (as
the conclusion from premises); that this is being-what-it-is and for what
reason it is better in this way, not better without qualification but in relation
to the being of each thing. (Ph. ii.7, 198b4-9)28
These concluding remarks build on the idea that there are two different
kinds of things that are sources of natural motion while not themselves
changing, and that one of those, while not itself a natural object, is an
appropriate subject of natural science, namely that which nature, under-
stood as an inherent source of motion, acts for the sake of. Chapter 7 has
thus prepared us for the subjects to be taken up in the final two chapters of
Ph. ii.8, the cause for-the-sake-which and its relation to necessity.
5.3.3 Physics ii. 8–9: Teleology, Matter, and Necessity29

Echoing the conclusion of chapter 7, the opening sentence of chapter 8 says
that nature is “among the causes for the sake of something” (compare 198b4
and b10); it pointedly does not, initially, say that nature is ‘that for the sake
of which.’ The concept of nature is likely being used initially to refer to the
inherent source of motion delineated in Ph. ii.1–2. And here we see the
importance of the distinction between form understood as that for the sake
of which and form understood as source of motion.30 For the defense of
natural science as a unitary science in Ph. ii.2 depends on the idea that the
two natures of a natural substance are teleologically unified – and that
unity depended on the formal nature of a natural substance being that for
the sake of which its material nature comes to be and persists. This
27
One central conclusion of Ph. viii is that there must be such a thing, but the investigation of its
nature is the topic of Metaph. Λ.
28
There are a number of puzzling aspects of this remark I will not be able to deal with here; but see
Ross 1936, 527–528; Philoponus (in Ph. 303.8-306.12); Simplicius (in Ph. 366.30-369.13).
29
The details of the defense of the teleological character of natural processes and their outcomes in this
chapter are hotly disputed. My primary focus here is not to enter into those disputes, but in recent
years there have been a number of publications that not only enter into these disputes but do an
excellent job of surveying the issues and the various alternative interpretations on offer: Gotthelf
2012, esp. chs. 1–4; Johnson 2005; Quarantotto 2005, esp. chs. 3–4; Leunissen 2010, esp. ch. 1.
30
Cf. de An. ii.4, 415b8-28.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
conception of natural science depends, then, not only on nature being
among the causes that act for the sake of something – it must also be ‘the
cause for-the-sake-of-which.’ This is in fact asserted in Ph. ii.2, but only
comes to the fore in chapters 8 and 9, most clearly in the following, densely
written, sentence.
And since the nature [of a thing] is twofold, on the one hand as matter and
on the other as form, and the nature as form is an end, while the others are
for the sake of the end, this [nature as form] would be the cause for-the-sake-
of-which. (Ph. ii.8, 199a30-32)31
This teleological way of conceiving the relationship between the materials
that constitute a natural being and the form of the thing so constituted
provides the ontological underpinnings for a new, conditionalized, way of
conceiving of necessity, the subject of the last chapter of Ph. ii – given that
a certain goal is to be realized, certain materials, with appropriate disposi-
tional properties, must come to be in a certain temporal and spatial order.32
And this, in turn, provides yet another norm that is specific to natural
inquiry, to the methodos of nature.
It is apparent, then, that the necessary in natural things is that which is
spoken of as matter and its motions. And it is for the natural scientist to discuss
both causes, but more the cause for the sake of what; for this is a cause of the
matter, rather than the matter being the cause of the end; and the end is that for
the sake of which, and the starting point is from the definition and the
account . . . (200a30-35; emphasis added)33
Again, questions of causation and definition are integrated tightly. As so
often when teleological causation is under consideration, Aristotle stresses
that the goal, that for the sake of which, has priority because in these cases
the account that specifies the identity of the goal must be understood in
order to properly understand the materials and processes that are needed to
achieve that goal.34 The prioritization of ‘that for the sake of which’ in
accounts of natural substances is clear; however, Ph. ii ends by considering
whether ‘the necessary’ will also be mentioned:
Perhaps the necessary is also in the account. For in defining the work of
sawing, that it is a certain sort of dividing, surely this will not exist unless the
31
For the Greek, see p. 129, note 18.
32
Cf. Charles 2012, 230–232; Quarantotto 2005, 178–212. 33 Cf. PA i.1, 639b12-16; 642a14-18.
34
To put the point in a less epistemic mode: the nature of the goal determines what the appropriate
materials and patterns of changes are to be, and that causal relationship needs to be reflected in our
accounts of natural things.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
saw has teeth of a certain sort; and such teeth will not be, unless they are
made of iron. Thus even in the account some parts of the account are as
matter. (Ph. ii.9, 200b4-8)35
Once more, then, but now stated within the context of explanations by
conditional necessity, form, understood as the end for the sake of which
materials are as they are, is prioritized, but matter is not ignored. Indeed, it
is argued that the necessary materials should be brought into the account
by way of conditional necessity.
As with the teeth of saws, so too with the teeth of animals: In discussing
the uniform materials out of which hard parts are made in PA ii, Aristotle
considers the material out of which teeth are made:
The nature of teeth is . . . in some cases present for a single function, the
preparation of nourishment, while in other cases it is present both for this
and defense (e.g., in all those with saw-like teeth) . . . . Of necessity all of
these parts have an earthen and hard nature; for this is the defensive
potential (ὅπλου δύναμις). (PA ii.9, 655b8-13)
When, in book iii, he discusses teeth as nonuniform parts, he assumes the
material nature of teeth as established and focuses on differences in their
shape and organization in relationship to the different lives animals lead:
Some animals have teeth . . . for the sake of nutrition alone. But those that
have them for protection as well as for strength in some cases have tusks . . .
and in other cases have sharp, interlocking teeth, for which reason they are
called saw-toothed. For since their strength lies in their teeth, and this comes
about because of their sharpness, teeth that are useful for strength fit
together in an alternating pattern, so as not to be worn down by being
rubbed against one another. (PA iii.1, 661b17-22)
Here we see in his actual biological practice exactly the sort of teleological
account defended in Ph. ii.9: from the fact that certain teeth are present for
the sake of both nutrition and defense, it follows that they must have
35
This example should be compared to PA i.1, 642a9-13: “for just as, since it is necessary for the axe to
split, it is a necessity that it be hard, and if hard, then either bronze or iron, so too, since the body is
an instrument (for each of the parts is for the sake of something, and likewise too the whole), it is
therefore a necessity that it be a certain sort and from such materials, if this [end] is to be.” The
differences are as important as the similarities: in PA i.1 the analogue to the living body, the axe, is
explicitly mentioned as the object being defined. In Ph. ii.9, however, it is unclear whether the
definiendum is the saw or the activity of sawing. The example begins with reference to defining the
work of sawing (perhaps analogous to defining respiration or sleep); however, this might be
a reference to the form of the saw, since it is that work that then determines that the saw must
have teeth and that they must be made of a specific material. And by listing two material options, the
PA example makes it clear that it is the dispositional property (hardness) of the material that is
conditionally necessary, not a specific uniform material.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
5.4 Conclusion: A Quick Glance Forward 139
a certain structure and organization and be made of a suitably hard
material.
5.4 Conclusion: A Quick Glance Forward

At the beginning of Ph. iii (200b12-14), Aristotle explains that “[s]ince the
nature of a natural thing is a source of motion and change, and our
methodos is concerned with nature, the question ‘what is motion?’ must
not escape our notice.”36 And indeed, for the next three chapters of book
iii he pursues that very question and comes up with a notoriously compli-
cated answer.37 But before doing so, he notes that, since motion seems to
be continuous, there are a number of other questions – about the unlim-
ited, place, void, and time – that must be taken up after the question of
what motion is. He then makes a comment that sheds light on how this
treatise and its method differ from the rest of Aristotle’s more narrowly
focused natural investigations. The lines I am especially interested in might
be translated as follows:
It is clear, then, that, both on account of these things and on account of
these things being common to all and universal (κοινὰ καὶ καθόλου), we
must investigate by getting hold of each of them (for the study of the
distinctive things (τῶν ἰδίων) follows the study of the common things
(τῶν κοινῶν)). (Ph. iii.1, 200b21-25)
In order to proceed with any particular investigation of nature, whether of
meteors, rainbows, minerals, animal parts or activities, or the motions of
the heavens, you will need to have a clear conception of motion and change
in general, and you will need to determine whether the unlimited, place,
void, and time exist and, if so, what they are. These concepts refer to things
that are common to any domain of the natural world.38 The Physics,39 then,
36
The term methodos appears four more times in the Physics – once in iii.5, 204a34-b11, and three times
in book viii (251a5-8, 253a32-b10, 261a27-31). As one would expect if the position I am defending is
correct, all are concerned with delimiting the approach to the topic at hand so that it is appropriate
to natural inquiry, as opposed to mathematics or first philosophy.
37
These are chapters that have received a great deal of attention. For a clear review of the issues and
options, see Coope 2009. For defenses of various options, see Gill 1989, ch. 6; Graham 1988;
Heinaman 1994; Kosman 1969; Kostman 1987.
38
If they exist at all, that is. Aristotle denies the existence of void and actual infinity. This is one way in
which the erotetic framework of APo. ii structures the inquiries in Ph. iii–iv – the inquiries proceed
by first considering if the purported object of inquiry exists and then, if it does, moving on to
consider what it is. Cf. iii.4, 202b35-36 (re infinity); iv.1, 208a27-30, iv.4, 210b32 (re place); iv.6,
213a12-14 (re void); iv.10, 217b29-32 (re time).
39
I remain neutral on how the various books of our Physics came to be related to one another. Relevant
to answering that question are the following points: (i) The opening of book iii both looks back to
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
is a work in the conceptual foundations of natural science; not in the sense
that its methodology is any more ‘conceptual’ than any other, but in the
sense that the concepts about which it aims to get clear – nature, cause,
chance, motion, place, time, and so on – are presupposed by any of the
more specialized natural inquiries in which Aristotle was engaged.40 In Ph.
i.2–3 Aristotle makes it clear that he does not think it is appropriate for the
natural scientist, qua natural scientist, to question whether change exists –
obviously it does. At the beginning of Ph. iii, however, he makes it equally
clear that what change is is anything but obvious,41 and to fail to answer
that question is tantamount to being ignorant of nature itself. Moreover, in
the case of the unlimited, void, time, and place, it is even legitimate to
consider the question of whether they exist, and part of the exploration of
what they are involves exploring in what sense they exist, and that is indeed
a big part of the discussion of those topics. This is work that is presup-
posed, and is thus conceptually prior to, the more particular natural
investigations we will be considering in the following chapters.
the earlier books, noting that the account of nature in those books assumes that there is motion in
nature, but it remains to investigate what it is; and it then previews all of the topics covered in books
iii and iv; (ii) book v can be viewed as a natural appendix to the general discussion of change; and
(iii) since all of the topics in books iii and iv arose out of the undefended assumption that motion
‘seems’ (δοκεῖ, 200b16) to be continuous, the discussion of continuity in book vi fills an obvious gap
in the earlier discussion as well.
40
I am here distancing myself from a thesis defended by David Bostock (Bostock 2006, 4–7 and
passim) that Aristotle’s Physics is largely an exercise in ‘conceptual analysis’ and that Aristotle
regularly confuses such an exercise with empirical inquiry. (See Lennox 2008b.)
41
And as is clear from the literature referenced in note 37, Aristotle’s account of what motion is is
among the most subtle and complicated discussions in the entire corpus. This is one of many cases in
the Aristotelian corpus where the distinction between an inquiry into τὸ ὅτι and τὸ διότι, a key
feature of the erotetic framework provided by APo. ii, is of fundamental importance.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
chapter 6
Shaping the Methodos of Animals

The Purpose of Parts of Animals i
This argument [about the infertility of mules], then, is too universal

and empty. For arguments that are not from appropriate starting
points are empty – they seem to be about the subject but they are not.
For accounts from geometric starting points are geometric, and it is
likewise in the other sciences as well; that which is empty seems to be
something, but is nothing at all.
(GA ii.8, 748a7-11)1
Chapter Summary. Both Aristotle’s investigation of animals and his

investigation of soul begin with long and detailed methodological discus-
sions, and both of those discussions eventually take the form of a series of
questions that the investigation needs to answer. The two discussions are
quite different, and the questions that are highlighted are also different –
and yet both are recognizably framed in terms of Aristotle’s Analytics and
Metaphysics. In the next two chapters we will focus on PA i and de An. i.1 as
case studies of how Aristotle’s views about the way differences in the target
of inquiry, as well as our epistemic access to, and our qua-perspective on,
the target, must influence the norms that govern that inquiry. These case
studies will also be an opportunity to investigate the ways in which the
‘erotetic shape’ of an inquiry is provided by Aristotle’s metaphysics and
epistemology and to explore the way in which Aristotle’s methodological
approach to an investigation is influenced by the ‘state of the art’ in a given
domain. All of Aristotle’s predecessors had views about the soul, views
which typically reflected ontological commitments of various kinds. That
historical fact results in a lengthy and detailed aporetic investigation of his
predecessors’ views in de An. i. Animals qua animals, on the other hand,
had not been considered as an independent subject of inquiry before
1
οὗτος μὲν οὖν ὁ λόγος καθόλου λίαν καὶ κενός· οἱ γὰρ μὴ ἐκ τῶν οἰκείων ἀρχῶν λόγοι
κενοί, ἀλλὰ δοκοῦσιν εἶναι τῶν πραγμάτων οὐκ ὄντες. οἱ γὰρ ἐκ τῶν ἀρχῶν τῶν
γεωμετρικῶν γεωμετρικοί, ὁμοίως δὲ καὶ ἐπὶ τῶν ἄλλων· τὸ δὲ κενὸν δοκεῖ μὲν εἶναί τι, ἔστι δ’ οὐθέν.
141
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
142 Shaping the Methodos of Animals
Aristotle. For that reason, previous views are treated very differently in his
animal investigations – more typically referred to and critically rejected
during his own investigation of a part or process, rather than as
a preliminary aporetic review of previous thought. This is a rarely noticed
way in which the subject under investigation will dictate the shape of the
investigation itself. First, then, we turn to PA i.
6.1 Introduction
In Chapters 3 and 5, I surveyed the ways in which Aristotle deploys the
concept methodos generally and in the context of the general investigation
of nature. There are a number of more or less invariant features of its use,
as we saw: (i) It is nearly always invoked in the context of inquiry; (ii) it
refers specifically to how one ought to proceed toward the goal of
knowledge, or to an actual inquiry viewed from the perspective of its
investigative methods; and (iii) these ‘methodic’ norms are invariably
domain-specific. These three features of the concept’s use lead to an
interesting distinction in its reference: often it refers to the methods
that are or ought to be used in a specific domain of investigation; but
occasionally it refers to the domain of investigation itself, looked at from
a methodological perspective. In this chapter and the next, I look in detail
at two distinct introductions to two distinct inquiries: De partibus ani-
malium i, generally agreed to be a philosophical introduction to
Aristotle’s various investigations of animals; and De anima i.1, the intro-
duction to Aristotle’s investigation of the soul. The concept methodos
plays a critical role in both introductions, and attending to that role will
provide additional insight into the ways in which different epistemic/
methodological norms govern different inquiries.
6.2 How to Proceed with Animals

As we saw in Chapter 4, Aristotle’s Meteorology opens with a review of what
the course of study of the natural world has covered so far (338a20-25): the
primary causes of nature, all natural motion, the movement of the ordered
system of stars, the elements of bodies, how many there are and what sort,
their changes into one another, and generation and decay of the most
common kind. He then reviews the subjects to be covered in the investiga-
tion to follow, “a part of this methodos, which all previous investigators
have called meteorology” (338a25-26). After outlining what is involved in
that investigation, he goes on:
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
6.2 How to Proceed with Animals 143
Having dealt with these subjects, we will study whether we are somehow
able, according to the recommended manner (κατὰ τὸν ὑφηγημένον
τρόπον),2 to give an account of animals and plants, both in general and
separately. For having spoken of these things we would pretty much have
reached the goal of our original plan in its entirety. (Mete. i.1, 339a5-9)
In Chapter 2, I argued that the question posed here, about whether we are
able to give an account of animals and plants according to the same method
adopted for the previously recounted branches of natural science, is
a genuine concern for Aristotle. Moreover, it is not in the first instance
a question about how to study animals and plants; rather, it is a meta-level
question about whether the methods used elsewhere in the study of nature
suffice for living nature. It is, as we shall see, the question that lies behind
the need for PA i.
PA i, which is generally recognized as a philosophical introduction to
the study of animals,3 begins with the following words:
Regarding every study and every methodical inquiry (Περὶ πᾶσαν θεωρίαν τε
καὶ μέθοδον), the more humble and more valuable alike, there appear to be
two sorts of state (δύο . . . τρόποι τῆς ἕξεως), one of which may properly be
called scientific knowledge of the subject matter, the other a certain sort of
discernment (τὴν δ’ οἷον παιδείαν τινά). (PA i.1, 639a1-5)
PA i can usefully be considered as a guide to acquiring the second state,4 this
state that is a certain sort of discernment, which provides one with the
ability to discriminate between what is well and what is poorly expressed.5
Or rather, it is a guide to acquiring a specific form of this state – for he goes
2
Compare Pol. i.1, 1252a17-18: κατὰ τὴν ὑφηγημένην μέθοδον. As I noted in Chapter 3, these two
phrases are very close in meaning.
3
See Ogle 1882, 141; Düring 1943, 35; Le Blond 1945, 51–52; Vegetti 1971, 485; Balme 1992, 69; Lennox
2001a, 119.
4
The Greek term here, ἕξις (literally ‘possession’), is typically restricted in reference by Aristotle to
states acquired by practicing some activity, often under the guidance of a teacher. Cf. EN ii.5, which
identifies ἕξις as the genus of moral virtue by distinguishing such states from power (dunamis) and
affection (pathos), neither of which we acquire by deliberate choice (1105b20-28; 1106a10-13). EN vi.5
refers to practical intelligence (phronêsis) as “a state (ἕξις) grasping the truth, involving reason,
concerned with action about what is good and bad for a human being” (1140b4-6; Irwin trans.)
and says it is the same state as political intelligence (1141b23-24); he likewise refers to art (1140a6-10)
and scientific knowledge (1139b31-32) as ἕξεις; and of course scientific knowledge is the other ἕξις
referred to here.
5
Compare the following comment in Pol. iii.11, 1282a1–7: “Hence just as a court of physicians must
judge the work of a physician, so also all other practitioners ought to be called to account before their
fellows. But ‘physician’ means both the ordinary practitioner, the master of the craft, and thirdly, the
person of general discernment (ὁ πεπαιδευμένος) about the craft (for in almost all the arts there are
some such people, and we assign the right of judgment (τὸ κρίνειν) just as much to discerning people
(τοῖς πεπαιδυμένοις) as to those with knowledge (τοῖς εἰδόσιν)).” As I noted in Chapter 3, in the
concluding sentence here we get a contrast between the person with discernment and one with
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
on to distinguish people with a very general form of this ability from people
who have it about a specific discipline, such as the art of medicine.6 Thus,
at the end of this paragraph, he makes a transition to the business at hand
with the following words: “So it is clear, for natural inquiry too (καὶ τῆς περὶ
φύσιν ἱστορίας), that there is need of some such standards (ὅρους) . . . ”
(639a12-13). The ability to make such judgments requires certain standards,
and if one’s discernment is about a specific field, then it will be standards
appropriate to that field that one needs to acquire. It is with identifying such
standards that the five chapters of PA i are primarily concerned.7 This is
critical to inquiry: as Aristotle puts it in EE i.6:
Lack of discernment (ἀπαιδευσία) about each subject is to be unable to
discriminate appropriate reasoning (τούς οἰκείους λόγους) about the subject
from that which is foreign to it. (1217a8-10)8
Let us now return briefly to the appearance of the concept of methodos in
the first sentence of this book. I want to draw attention to the fact that
theoria and methodos are conjoined not simply by ‘and’ (kai) but by
‘both . . . and’ (te kai), which indicates that these terms are conveying
different ideas, and are not merely synonyms. It could be that Aristotle has
two different categories of cognitive endeavor in mind. However, it could
also be – and I will make a case for this reading – that he wants to
distinguish two different aspects of a single cognitive endeavor, aspects
that are distinguished when he identifies two different states associated
with them – scientific knowledge (epistêmê) and that general discriminative
knowledge of the subject, but in a way that makes explicit that a person can have the general form of
discernment without having what would count as knowledge of the subject.
6
I here flag a number of features of these passages that make it somewhat difficult to nail down precisely
what sort of expertise he is referring to. The PA i.1 passage appears to contrast a transdisciplinary
discriminative ability with one limited to a specific domain, while the Pol. iii.11 passage distinguishes
three sorts of physician, one of which apparently does not actually practice medicine and yet is
discerning enough about the practice to be called a doctor. Tentatively, I suggest that the third sort of
doctor is akin to what today might be called a philosopher of medicine; and the person referred to in
the PA i.1 passage as discerning about the inquiry into nature is akin to a philosopher of science.
Given the almost exclusive focus on animals of PA i, we might, as David Balme (1992, 69) does, think
of it as a philosophy of zoology. For a contrasting viewpoint, see Tipton 2014, 19–29.
7
Notice that it is natural inquiry (ἡ περὶ φύσιν ἱστορία) that is said to be in need of certain standards.
I cannot, therefore, agree with David Balme’s assertion that this introduction “makes it clear that he
[Aristotle] is not setting out to discuss scientific method . . . [nor] considering how to arrive at an
explanation . . . but how to judge an explanation when made” (Balme 1992, 69; for a contrasting view,
see Le Blond 1945, 52). In fact, as we shall see, many of the norms eventually established in this
discussion are explicitly identified as norms for guiding inquiry along the right lines. Cf. Kullmann
1997, 43–62 (which builds on Kullmann 1974, ch. 3), who sees PA i generally as propaedeutic to, and
as a study of the presuppositions of, the scientific study of animals.
8
Cf. Karbowski 2019, 116–117.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
6.2 How to Proceed with Animals 145
ability he identifies as a certain sort of discernment (paideia). This ‘two
aspects’ reading gains support from the last paragraph of chapter 4, which
summarizes what has been accomplished.
We have said, then, how the methodical inquiry concerning nature
(τὴν περὶ φύσεως μέθοδον) should be appraised, and in what way the
study (ἡ θεωρία) of these things might proceed on course and with greatest
ease (ὁδῷ καὶ ῥᾷστα). Further, about division we have said in what way it is
possible by pursuing it to grasp things in a useful manner, and why dichot-
omy is in a way impossible and in a way vacuous. (PA i.4, 644b15-20)
The repetition of methodos and theôria here is almost certainly a conscious
echo of the opening five words of PA i.1, and here the two words are used to
emphasize different aspects of a single inquiry. Based on the conclusions of
Chapter 3, I suggest that to refer to a methodos of some subject matter is to
identify an inquiry by reference to the distinctive manner in which the
inquiry is carried out;9 while to refer to it as a theôria is to identify the
inquiry by reference to the subject matter on which it is focused.10 A proper
study of the subject requires that one proceed in a proper manner – the use
of the adverbial dative hodôi here, in close proximity to methodos, provides
further support for this thought.11 What Aristotle has been discussing for
the previous four chapters are the norms that must guide an inquiry into
animals, if it is to stay ‘on course’ or ‘on track’ toward the goal of scientific
knowledge.
This passage, then, supports the hypothesis put forward in Chapter 3
about the connection between these two uses of the term: To ask, as Mete.
i.1 does, ‘Is x part of the same methodos as y?’ is to ask whether the same
norms or standards of inquiry are appropriate for the study of x and y. If so,
9
Since Aristotle holds that the same objects may be studied from a variety of distinct perspectives, and
that each perspective may be governed by different normative constraints, the subject matter will be
determined intensionally rather than extensionally. As he says in Metaph. M. 3: “It would be best to
study each thing in the following way – take separately that which is not separated, which is what the
arithmetician and geometer do. For human being, qua human being, is one and indivisible, and the
arithmetician takes him as one indivisible thing and then considers whether something is attributed
to man qua indivisible. But the geometer does not study man qua man nor qua indivisible, but qua
solid” (1078a21-26). Note that this in itself is a norm of inquiry. He had previously suggested that
standards of precision will change depending on whether or not one prescinds from considering
objects as having magnitude or undergoing change (1078a9-14).
10
And though I am less sure about this (given the range of uses to which Aristotle puts this word), it
may also stress that the study is theoretical, as opposed to productive or practical, in nature (cf.
Metaph. Ε.1, 1025b18-28 and EN vi.3, 1139a14-22 on this distinction).
11
As noted in Chapter 3, the noun ἡ ὁδός refers to a road, path, or track and is used metaphorically in
much the same way as those English expressions are. The dative form used here often has adverbial
force, conveying the idea of staying on the road to one’s destination, thus my ‘on course.’
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
then you are licensed to refer to them as instances of the same methodos.
When the term is used in this way, its rules for application will be similar to
those for ‘inquiry,’ ‘investigation,’ or even ‘discipline.’ However, it will
identify the investigation by attending to the norms and standards that are in
play rather than to the object and theoretical nature of the investigation.12
6.3 Three Norms for Animal Inquiry

During the introduction to PA i.1, Aristotle discusses the special skill of
critical judgment we need to acquire in a way that suggests it will be
primarily of use in judging what others have said or written. However, as
he turns to the questions that need to be considered in acquiring this
critical ability, as we saw, he refers to certain standards that need to be in
place for the inquiry into nature, strongly suggesting that these are stand-
ards that ought also to be referred to by one engaged in an ongoing
investigation.
The lines introducing the first question do not rule that out, but do not
clearly confirm it either. For while they refer to standards appropriate to
natural inquiry, the passage goes on to say that these are standards
by reference to which one can appraise the manner of [the inquiry’s] proofs
(τὸν τρόπον τῶν δεικνυμένων), apart from the question of what the truth is,
whether thus or otherwise. (PA i.1, 639a13-15)
Since the word I have translated “proofs” is the participial form of a verb
Aristotle often uses to refer to deductive proof,13 this passage could be
referring to standards for the evaluation of the way explanations are
presented. But its most common meaning is ‘to show’ or ‘display.’ Still,
at this point, it is an open question whether the standards being referred to
are standards for carrying out an investigation as well as for appraising the
results of an investigation. What is important to realize is that when he
turns to the question of what these standards are, it is explicitly by way of
illustrating the special sort of paideia he has just characterized. It is time,
then, to look carefully at the way in which Aristotle introduces and
discusses those norms.
12
Though of course the subject matter, and the fact that the investigation is theoretical rather than
practical or productive, will need to be taken into account in articulating those norms and standards.
13
That is, the participle is formed from δείκνυμι, not ἀποδείκνυμι, the verb cognate to ἀπόδειξις
(demonstration). However, later in the chapter Aristotle claims that the goal of the inquiry he has in
mind is a different mode of demonstration, grounded in a different mode of necessity (ἀλλ᾽ ὁ τρόπος
τῆς ἀποδείξεως καὶ τῆς ἀνάγκης ἕτερος; 639b30-640a1).
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
6.3 Three Norms for Animal Inquiry 147
6.3.1 Normative Question 1: Universality and the Beginning of Inquiry
The very first question Aristotle raises is actually introduced simply as an
example of the point he has just made:
I mean, for example, should one take each substantial being singly
(μίαν ἑκάστην οὐσίαν) and define it independently, for example taking up
one by one the nature of mankind, lion, ox, and any other animals as well; or
should one first establish, according to something common (κατά τι κοινὸν),
the attributes common to all? (PA i.1, 639a15-19)
Clearly Aristotle is treating this as an example of the standards of appraisal
just mentioned, and it is a normative question about which of two possible
ways of approaching a subject of investigation is best. The examples used to
illustrate the issue are animals, but it may be important that the alternative
being considered is not specifically limited to animals, and the standards
referred to in the previous sentence were standards for an inquiry about
nature, not specifically about animals.
Nevertheless, the examples indicated that Aristotle’s focus is clearly on
a question about where to start an investigation into animals, and in
particular, at which level of universality or specificity. After stating the
alternative abstractly, he notes (at 639a19-22) that there are affections and
dispositions, like sleep and respiration, that belong in the same manner to
many different kinds of animals, and others, like modes of locomotion,
that belong to animals in formally differentiated ways (flying, swimming,
walking, creeping; 639a29-b3). And this fact means that if one adopts the
first alternative method mentioned, you will be forced to repeat yourself
over and over. Yet he leaves the dilemma unresolved, saying that matters
are too “unclear and indefinite” (ἄδηλον καὶ ἀδιόριστον, 639a22-23) at this
point to decide between the alternatives. Before going on to the next
question of concern, however, he reformulates the issue at stake concerning
the appropriate level to enter the investigation of animals – and when he
does so, he makes it explicit that these standards are intended to govern
how we engage in investigation, and not merely how we appraise the results
of completed investigations.
Wherefore how we ought to carry out the investigation (πῶς ἐπισκεπτέον)
should not be overlooked; I mean, whether one should study things in
common according to kind (κοινῇ κατὰ γένος) first, and then later their
distinctive characteristics, or whether one should study them one by one
straight away. (639b3-5; emphasis added)
I have highlighted three features of importance in this formulation:
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
1. The question is about how one ought to investigate: that is, it is
a normative question about inquiry.
2. This is now reflected in an explicit question about the proper order of
investigation. That is, Aristotle appears to be concerned that if one
does not proceed in the correct order, the investigation may never be
successfully completed.
3. To be specific, the question is one about the level of universality that
should be targeted at the beginning of such an inquiry (639a15-19):
should we study individual species one at a time, or should we begin
searching for attributes possessed in common in virtue of or according
to kind?14 In the passage just quoted, he has converted the initial either/
or question into an interesting alternative:
Alternative 1. Pursue a temporally staged, two-tiered investigation, in which
one begins by investigating features that belong in common to kinds,
and then later moves to investigating those characteristics that are
distinctive to specific forms of those kinds.
Alternative 2. Begin investigation immediately with specific forms, studying
each of them one by one. (He neither mentions, nor rules out, the
possibility of a later investigation that looks for universals that emerge
from comparing these case-by-case investigations.)
That change in the way in which these alternatives are posed reflects the
results of the intervening discussion (639a25-b3), in which Aristotle has
articulated the distinction I mentioned earlier: between two different kinds
of commonly possessed features, those that are undifferentiated across
many kinds and those differentiated “according to form.” Though he
does not make the point here, this is a distinction you are far more likely
to notice if you begin at the more general level. These alternatives are
directly reflected in the way in which sleep is discussed in the PN, on the
14
There is a legitimate worry about whether, if Aristotle ends up recommending starting, wherever
possible, with what is universal and common, this could be a norm about how to begin an
investigation, since on Aristotle’s general epistemic view one begins the quest for knowledge with
perceptual experience of particulars. This passage does not conflict with his perceptual foundation-
alism, however. The issue being raised in this passage in PA i.1 is how inquiry into a domain of
experience (say, animals) should proceed. As we will see, Aristotle’s answer (which is a view defended
by many cognitive psychologists today) is that the right way to begin, when possible, is with
categories I refer to as ‘entry-level kinds,’ categories like ‘bird’ and ‘fish’ – and as I am about to
argue, he arrives at that conclusion by reflecting on the practices that led ordinary Greeks to
successfully grasp those general categories. Aristotle is arguing that in most cases, this is the
appropriate first step in inductive inquiry. I thank Victor Caston for pushing me on this point.
As he pointed out, in a way this is the same puzzle that bothers commentators about the claim in Ph.
i.1 that inquiry starts with the undifferentiated universal, and epistemic progress is achieved by
making the appropriate differentiations.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
one hand, in contrast to the way in which Aristotle introduces the investi-
gation to be carried out in MA, distinguishing its task from that of the IA,
on the other:
Concerning all those movements of animals that belong to specific kinds, what
their differences are and what are the causes of the features attributed to each of
them, this we have investigated elsewhere; however, we must now investigate
(ἐπισκεπτέον νῦν) generally the common cause (περὶ τῆς κοινῆς αἰτίας) of their
being moved by any motion whatever (for some animals move by flying, some by
swimming, some by walking, some by other such means). (MA 1, 698a1-6)
Note that this difference is taken to be significant enough that Aristotle
presents the two investigations separately, and makes a point of stressing
their difference. By contrast, the only clear differentiation of sleep dis-
cussed in De somno et vigilia is between those animals that sleep and those
that do not – the account of sleep, briefly summarized in its conclusion (3,
458a25-32), applies, without any essential differences, to all animals that
sleep.15
6.3.2 Answering the First Question: Determining a Norm

Though Aristotle spends a great deal of time introducing and then fine-
tuning the question of the proper level of abstraction at which to enter into
a study of animals, he concludes this discussion (639b6-7) by noting, with
regret, that neither this question, nor the next one to which he is about to
turn, can at this point be adequately answered. But he returns to this first
question in PA i.4, presenting it as an aporia about which of the alterna-
tives to choose. At that point, it can be adequately answered, because he
has, in the previous two chapters, defended a new method of division that
grows out of a scathing demolition of Platonic dichotomous division. That
new method is designed to identify levels of differentiation of multiple,
coextensive features that belong in common to general groups of animals
(as, for example, feathers and beaks are common to all birds), groups he
will elsewhere refer to as ‘great kinds’ (megista genê).16 That attack begins as
15
There are, of course, variations in things like how long and how often animals sleep, and on whether
they dream or not, but these are not formal differences in sleep per se. On the methodological norms
of the inquiry into sleep, see Code 2015, 11–45.
16
Three notes of caution on this point are necessary: (i) The expression μέγιστον γένος is not used
here; (ii) in PA iv.8 he identifies four μέγιστα γένη of soft-shelled animals – that is, he uses the
expression μέγιστον γένος to refer to sub-kinds of one of the μέγιστα γένη mentioned in HA i.6
(490b10); and (iii) even in HA the term is used rarely, and never in a rigid and fixed way. See
Gotthelf 2012, 293–306.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
a list of negative results that flow inevitably from using iterated dichotom-
ous division (PA i.2-3, 642b5-643a23); but there is a gradual transition to
a list of positive norms for the proper use of division,17 introduced by
a strongly anti-Platonic statement about the right way to think about the
relationship between differentiae and the form of a living animal:
The difference is the form in the matter (ἔστι δ’ ἡ διαφορὰ τὸ εἶδος ἐν τῇ
ὕλῃ); for no part of an animal exists without matter, nor is it matter alone;
neither will a body in any condition whatever be an animal, nor will any of
its parts, as has been said repeatedly. (643a24-27)18
Notice that the phrase “for no part of an animal exists without matter, nor is it
matter alone” is, in essence, a specification for the study of animals of the claim
about natural inquiry generally in Ph. ii.2: “such [natural] things should be
studied neither without matter nor according to matter” (194a14-15). In
addition, Aristotle introduces norms recommending that the researcher
should not divide by accidents, should divide by contraries, and should aim
to insure that each differentia introduced in each axis of division is
a determinate form of that which is being divided – which is, in fact, one
way to avoid ‘accidental’ or arbitrary differentiation. For example, if we were
to start with the general differentia ‘locomotion,’ then each determinate
differentiation should be a form of locomotion. Suppose one of those deter-
minate forms is flying: then the next division should be into forms of flying, as
in the following passage early in HA i, intended to introduce the reader to the
methods to be used more systematically in the rest of the treatise:
Of those that fly, some are feather-winged (e.g., eagle and hawk), some are
membranous-winged (e.g., bee and cockchafer), and some are dermatous-
winged (e.g., flying fox and bat). (HA i.5, 490a5-8)
After showing correlations between each of these forms of wing and other
attributes, and noting that among the animals that have these three forms
of wing only the one with feathered wings has a single name (i.e., bird), he
focuses in on the flying insects and continues down the axis of division:
Of those that fly but are bloodless, some are sheath-winged (for their wings are
in a casing, e.g., the cockchafer and dung beetle), some are without a sheath –
and of these, some have two wings and some have four . . . (HA i.5, 490a13-16)
Were this presented diagrammatically, it would look like this:
17
For more detailed discussion, see Balme 1992, 114–119; Lennox 2001a, 158–165.
18
Notice that this also links the discussion of division with the earlier argument that only a functioning
body or part is the body or a part of an animal (640b30-641a6).
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
Winged Flyer
Feather- Dermatous- Membranous-

winged winged winged
Sheathed Non-sheathed
Membranous- Membranous-
winged winged
Nonsheathed Nonsheathed
Diptera Tetraptera
Figure 6.1 Division of Wings, after HA i.5, 490a5-16.
But such a method of iterated, multiple-axis differentiation presupposes

starting at the very general level and searching for coextensive features at each
level of differentiation – and PA i.4 lays out a set of standards to be used in
identifying such correlations at this general level.19 Aristotle begins with
a puzzle20 about why people have not named a single kind that embraces
swimmers and flyers, given that there are features common even to these (see
the list at 639a20-22). He explains why they were right not to do so:
Nevertheless, they are correctly defined in this way: as many of the kinds21 as
differ according to degree and by the more and the less (καθ’ ὑπεροχὴν
καὶ τὸ μᾶλλον καὶ τὸ ἧττον) have been brought together under one kind,
while those that are analogous have been kept apart; I mean, for example, that
bird differs from bird by the more or according to degree (for one has long
feathers, another short feathers), while fish differs from bird by analogy (for
what is feather in the one is scale in the other). (PA i.4, 644a17-22)
19
I have discussed these in Lennox 2005, 87–100; a different, more ‘theory-laden’ reading of the same
text can be found in Charles 2000, 312–316.
20
Ἀπορήσειε δ’ ἄν τις (644a12).
21
I am here correcting my published translation in two important respects: first, by explicitly translating
τῶν γενῶν in a17, so it is clear that Aristotle is saying that kinds that differ more and less are brought
together under one kind; second, I now read τοῦτον τὸν τρόπον at a16 as referring forward, and thus
take the γὰρ in the next sentence as ‘explanatory’ or ‘prefatory,’ picking up on τοῦτον.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
There is very little explication of what he has in mind by kinds that differ
by more and less as opposed to kinds differing by analogy – the burden of
explanation is carried entirely by the examples of each kind of difference he
provides. But what he has already accomplished is nevertheless significant:
first, by means of his examples he has identified familiar groups, bird and
fish, as occupying a certain level of generality; second, he has said there is
something right about grouping kinds of bird and kinds of fish into
distinct, higher kinds, but not grouping birds and fish together into one
kind;22 third, he has introduced the idea of continuous variation (by
degree, by more and less) as the sort of likeness that underwrites uniting
many forms of animal into one kind and analogous likeness permitting
comparisons between kinds but not unification into one kind.
The next section of the chapter introduces yet another aporia, namely the
very problem mentioned at 639a19-29 – if we focus our investigation on the
most determinate forms (on grounds perhaps that these are ontologically prior
to everything predicated of them),23 we will be endlessly repeating ourselves. As
I’ve argued elsewhere, this is not merely a concern about the impracticality of
repetition. For it is one of Aristotle’s most profound epistemological insights
that failure to recognize in virtue of what a particular subject possesses an
attribute implies an inability to grasp its cause and therefore an inability to have
demonstrative, scientific knowledge of it – and one way to fail in this respect is
to identify the subject to which an attribute belongs either too specifically or
too generally. As he puts the point in APo. i.5:
even if you prove of each [sort of] triangle, either by one or by different
demonstrations, that each has two right angles – separately of the equilat-
eral, the scalene, and the isosceles – you do not yet know of triangles that
they have two right angles, except in the sophistical sense; nor do you know
it of triangles universally, not even if there are no other triangles apart from
these. For you do not know it of triangles qua triangles . . . (74a25-30)
In precisely the same way, even if you grasped that each kind of bird has
a beak, but you failed to grasp that it was qua bird that each of them has
a beak, you would not be prepared to search for the reason why birds have
beaks. And that is the state you will be in if you don’t carry out your
22
The aporia with which this discussion opens is stated with reference to ‘flyers’ and ‘water dwellers.’
This could be a preliminary (and rough) way of introducing ‘bird’ and ‘fish’ – rough, since Aristotle,
as we have seen, recognizes two kinds of flyers other than birds, and discusses in great detail water
dwellers (such as cephalopods, crustaceans, and cetaceans) that are not fish. Another possibility is
that he is withholding the terms ‘bird’ and ‘fish’ until he has introduced the idea of the appropriate-
ness of uniting into one kind those animals that differ only by more and less.
23
Ἐπεὶ δ’ οὐσίαι μέν εἰσι τὰ ἔσχατα εἴδη (644a23-4).
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
investigation with a set of methods designed to determine the appropriate
level of universality on which to focus.
After reviewing the problem, he is prepared for a solution:
Perhaps, then, the right course (ὀρθῶς) is this. In some cases – whenever
kinds are spoken of by people in a clearly defined manner (καλῶς
ὡρισμένων) and (καὶ)24 have both a single common nature (μίαν φύσιν
κοινὴν) and forms in them not too distant – we should speak in common
according to kinds, like bird and fish and any other [group] there may be
that, though it is unnamed, embraces, like a kind, the forms within it. But
whenever they are not such as this, we should speak one by one (καθ’
ἕκαστον), for example about mankind and any other such kind. (644b1-7)25
Here we are finally provided with a solution to the first question raised in
chapter 1, and thus with a normative recommendation both for judging when an
investigation has been carried out properly and for how to carry out an inquiry
as it proceeds. In cases where kinds are referred to in a clearly defined manner,
we should speak of them (in the context of an investigation) “in common
according to kind.” Notice too that Aristotle countenances cases where we
should speak in common according to kind, even though the groups are
unnamed, and yet embrace many forms like those that have been properly
designated.26
But how do we go about determining the kinds that should be investi-
gated “in common according to kind”? Aristotle gives us two criteria to
appeal to here, both of which are initially puzzling:
1. They share one common nature.
2. They have forms within them that are not too distant.
What is initially puzzling about criterion 1 is that Aristotle is here discussing
norms that will guide us in knowing when to investigate many forms of a kind
24
Charles 2000, 312, explicitly treats the καὶ as epexegetic; Gotthelf (2012, 276) follows Balme’s
translation but in his analysis of the passage takes it as “more or less epexegetical” (277). This may
well be right, but I am leaving open the possibility that Aristotle is distinguishing the way the words
‘bird’ and ‘fish’ are used in everyday Greek from the criteria that need to be explicitly identified in
order to determine when we “should speak in common according to kind.” That is, when, in
addition to people speaking of kinds in a well-defined manner, animals meet two other criteria, we
should then identify them as kinds. People may speak in a well-defined way of groups of animals
that do not meet these other two criteria.
25
Throughout I am reading the text of Bekker, not Peck, who for reasons that are not clear to me
makes a number of emendations with no manuscript support.
26
As Gotthelf notes (2012, 277, 299) Aristotle’s formulation allows for two different cases of groups
that embrace forms in the way that kinds do, one in which the kind is recognized but not designated
by a proper name (but perhaps by a ‘name-like expression’), and one in which there is a group that
meets the criteria but has not been recognized.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
as a kind, discussing them “in common according to kind.” Surely, we do not
know their common nature as we begin to investigate them.27 So how can this
be used as a criterion for determining when to discuss birds qua birds or for
delineating groups not previously identified as deserving of investigation
according to kind?
What is initially puzzling about criterion 2 is that Aristotle in this
passage gives us no help in determining “distance” between forms, which
we must do if we are to apply this criterion.
In my view, both of these problems are settled by the end of chapter 4,
especially in light of the discussion of division in chapters 2 and 3. Here is
the critical passage:
Roughly speaking, it is by the figures of the parts and of the whole body that
kinds have been defined, when they bear a likeness – for example, members
of the bird kind are so related to each other, as are those of the fish kind, the
soft-bodied animals, and the hard-shelled animals. For their parts differ not
by analogous likeness (οὐ τῇ ἀνάλογον ὁμοιότητι), as bone in mankind is
related to fish-spine in fish, but rather by bodily affections, for example, by
large/small, soft/hard, smooth/rough, and the like – speaking generally, by
the more and less (ὅλως δὲ τῷ μᾶλλον καὶ ἧττον). (PA i.4, 644b8-16)
Following David Charles’ analysis of ‘stages of inquiry’ identified by Aristotle in
the APo. ii, I begin by distinguishing knowing that a group of animals has
a common nature from know what that common nature is.28 The above
passage, in my view, outlines the procedures for having good reasons to think
many forms of animal have a single common nature (i.e., that there is
a common nature), prior to knowing what that nature is. All cephalopods
have an overall bodily configuration more akin to one another than to the
members of any of the other groups, and the same holds true for all crustaceans,
all fish, all birds, and so on – and likewise for their parts (and though Aristotle
does not say this explicitly, these two facts are related; the arrangement and
location of the tentacles of cephalopods, for example, is part of what gives
cephalopods their distinctive overall appearance).
27
Charles 2000, 313–316, worries about this problem as well, but resolves it by reading ideas from
chapter 5 – in which Aristotle is discussing the ways in which, once we know the proper function of
a part, we can explain its presence teleologically – back into this chapter, which I take to be aimed at
answering “entry level” questions (Lennox 2005), questions about how to initiate an inquiry into
animals, and which level of generality is appropriate for different groups. I have an alternative
reading that depends only on evidence from the context and from parallel passages in HA. Cf. the
careful, comparative analysis of our views on this issue in Gotthelf 2012, 275–287, 296–298; and in
the fine dissertation of Keith Bemer (Bemer 2014).
28
Charles 2000, ch. 2.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
But – and this takes us to the second puzzle – what does it mean to say that
the forms embraced by the kind are “not too distant” from each other? Here
I draw on my own previous work on the central role that Aristotle’s analysis of
variation or difference by “excess and defect” or “more and less” plays in his
understanding of kinds.29 The answer is that along each axis of similarity, the
qualitative and quantitative features of their parts differ within a range along
continua, as opposed to having completely different axes of similarity. An
example will be helpful: as different as bird feet and beaks are, they are all
recognizable as the feet and beaks of birds because they can be compared with
one another along a variety of continua – of length, breadth, depth, curvature,
hardness, color, and so on. Beaks cannot be so compared with mouths
constituted of lips and teeth, or insect proboscises.30 The mouths of birds
are simply not constituted of those parts, though they may well be analogous
(perhaps with respect to function or location).
JAÇANA
SHAG (WALKING ON
(SWIMMING) FLOATING PLANTS)
PTARMIGAN JUNGLE FOWL

(FEATHERED) (WALKING, SCRAPING)
COOT
(SWIMMING)
CROW SEA EAGLE

(PERCHING, LIFTING) (RAPTORIAL)
Figure 6.2a Forms of bird feet, indicating differentiation for different ways of life.
29
Lennox 1987a (reprinted in Lennox 2001b, ch. 1), 2005. The relationship between Aristotle’s use of
the expression ‘the more and the less’ and Plato’s use of it in the Philebus is important and needs
further exploration. There is a more extensive discussion of these different sorts of sameness or
identity in the first chapter of HA.
30
Lest the reader think the hummingbird beak is an exception: it is actually structurally just like other
beaks, though it is longer, thinner, and more flexible than most. The hummingbird opens its beak
and extends a long tongue with grooves and a feathery tip to extract the nectar from flowers.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
TOUCAN
HAWK
SHOVELER
PETREL CURLEW
SPOONBILL
AMERICAN ROBIN GOOSANDER
COCKATOO
GROSBEAK
Figure 6.2b Forms of bird beaks, indicating differentiation for different ways of life.
Thus, if there are many distinguishable specific kinds of animals in

which both their overall bodily configuration and their parts vary only
by more and less along their various qualitative and quantitative dimen-
sions, they will, by that very fact, be “not too distant.”
The narrative structure of PA i 31 explains why Aristotle introduces this
question about the level of universality or specificity at which to enter into
an investigation of animals early in chapter 1, but does not fully answer it
31
I defend the idea of PA i as a unified document with a narrative structure in Lennox 2010, in Lennox
and Bolton, eds. 2010, 56–77.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
until much later. It is a question that can receive a proper answer only after
a proper method for simultaneously tracking divisions of multiple features
that ‘belong in common according to kind’ has been described and
defended.
Above, I displayed the divisional aspect of this method in the sketch
form in which it appears in the first few pages of HA. To put a little flesh on
the bone (or perhaps to replace bone with flesh!), it will be helpful to look
at this method at work in presenting the results of his investigation of the
‘soft-bodied animals,’ our cephalopod mollusks, in the first chapter of
book iv.
Among the animals called ‘soft-bodied’, these are the external parts: 1. the
so-called feet; 2. the head, continuous with the feet; 3. the sac, containing the
internal organs, which some mistakenly call the head; 4. the fin, which
encircles the sac. In all of the soft-bodied animals the head turns out to be
between the feet and the belly. Moreover, all have eight ‘feet,’ and all have
two rows of suckers, except for one kind of octopus.32 The cuttlefish and the
large and small calamary have a distinctive feature, two long tentacles, the
ends of which are rough with two rows of suckers, by which they capture
food and convey it to their mouth and fasten themselves to a rock when it
storms, like an anchor.33 (HA iv.1, 523b21-33; cf. PA iv.9, 685a33-b2)
Here we see Aristotle capturing as much external anatomy as possible at
the level general to the kind as a whole first, and then gradually moving to
features that are peculiar to sub-kinds. He notes the peculiar ‘extra
tentacles’ of cuttlefish and calamari here (see figure 6.3b), and at
524a25-32 he discusses their other differences. He also notes here one
exception to the claim that there are two rows of suckers on all eight of the
octopuses’ legs, an exception he focuses on during his discussion of “the
many kinds of octopus” (525a13-19). Indeed, this methodology governs
the entire discussion, which goes on for two more Bekker pages, as it does
for the remainder of the discussion of the parts of the bloodless animals
(525a30-532b29).
32
Eledone cirrhosa, the ‘lesser’ or ‘curled’ octopus. See figure 6.3a.
33
Aristotle’s account and those found on contemporary websites are very similar: “Their [cuttlefish]
preferred diet is crabs or fish, and when it [sic] is close enough it opens apart its eight arms and out
shoots two deceptively long feeding tentacles. On the end of each is a pad covered in suckers that
grasp hold of the prey” (www.tonmo.com/articles/basiccuttlefish.php). One principal difference is
that Aristotle consistently mentions two functions for these tentacles, feeding and mooring during
storms, while modern texts I have consulted only mention feeding. One audience member in
attendance at a lecture in which I mentioned this discrepancy claimed to have repeatedly observed
the mooring behavior described by Aristotle while snorkeling in the Mediterranean.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
Figure 6.3a Eledone cirrhosa (the ‘lesser’ or ‘curled’ octopus) with a single row of
suckers on each arm.
Figure 6.3b Sepia officinalis (cuttlefish) with feeding tentacles extended.

The opening question in PA i.1, then, initiates a discussion aimed at
articulating a set of norms about how to initiate and proceed with an investiga-
tion of animals so as to properly organize information about them. These
appear to be norms that govern the preliminary, ὅτι stage of inquiry. These
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
norms are gradually articulated and defended in PA i, and are consistently
implemented in the presentation of the results of his empirical investigation of
animals in HA. In order to defend his views about the appropriate level of
universality at which to begin an investigation, views finally presented in
chapter 4, he needs to articulate and defend another set of standards for (a)
identifying kinds that share a large number of features in common and (b)
producing divisions by means of noting differentiations of those general
features within the many forms of those kinds. Concretely, for example:
cephalopods are a good entry-level kind because they all share an overall
distinctive bodily organization (head, eight legs, visceral sac, circular fin, etc.);
many of those features are found to be differentiated in distinct ways in different
sub-kinds (cuttlefish, squid, octopuses); and these differentiations can be
tracked and correlated by the multiaxial method of division characterized in
PA i.2–3 (e.g., very slim body and tentacles correlated with a single row of
suckers). Moreover, these are norms that both arise out of extensive experience
with the animal world and are designed with an investigation of that world
specifically in mind.34 There is no reason to think such norms would make
sense in studying the heavens or meteorological phenomena. Finally, these
norms are designed with a specific conception of scientific knowledge in mind.
Only by adhering to these norms, Aristotle insists, will you reach the goal of
causal explanation rooted in an identification of the essential natures of animals.
6.3.3 Normative Question 2: Establishing Facts and Searching for Causes

APr. i.30, which I have referenced a number of times in earlier chapters,
asserts that the distinctive principles of a domain arise from the specific
34
An important issue that I cannot properly address here, but which I want to flag, is the difficulty of
distinguishing between Aristotle’s actual methods of empirical investigation and his written reports
of the results of those investigations. When historians of science investigate many post-Renaissance
natural philosophers/scientists, they have access to such things as field notes, letters, laboratory
notebooks, autobiographical accounts of investigations, and so on. This allows the historian to make
a principled distinction between the published book or journal article, on the one hand, and the day-
to-day struggles of information collection, analysis, interpretation, and integration of new findings
with old, on the other – aspects of science often suppressed in the publications that result. There is
no clear demarcation of this kind in the extant Aristotelian corpus – certain works seem more like
a published book, others more like an investigator’s notes on an ongoing investigation. (The ancient
catalogues mention a number of works, no longer extant, that may well have reflected early stages of
investigation: Divisions [seventeen books], Dissections [eight books], Selections from Dissections [one
book] along with the [mostly?] post-Aristotelian Problems.) This poses an intriguing meta-level
problem for an inquiry like the one in which I am currently engaged – is it possible to distinguish, in
Aristotle’s texts, norms for presenting the results of an inquiry from norms for carrying out an
empirical inquiry? This footnote is intended to signal only an awareness of the problem, not
a solution to it.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
experiences of the objects of that domain, and uses astronomy as an
example.
[I]t is for astronomical experience to provide the principles for the science of
astronomy (for when the appearances had been sufficiently grasped, in this
way astronomical demonstrations were discovered; and it is also similar
concerning any other art or science whatsoever). (APr. i.30, 46a19-22)
At that level of generality, perhaps the discovery of principles is similar in
any other art or science. But in PA i.1, Aristotle’s next question centers on
this very issue:
. . . whether, just as the mathematicians explain (δεικνύουσιν) the phenom-
ena that concern astronomy, so too the investigator of nature (τὸν φύσικον),
having first studied the phenomena regarding animals and the parts of each,
should then state the reason why and the causes, or whether he should
proceed in some other way. (639b7-10)
Now one might think that if anything is an axiom of Aristotle’s philosophy
of science it is that you must first study the phenomena before going on to
determine the reason why and the causes – what other way could an
scientist proceed? The investigation of animals and their parts is, however,
complicated and very different from that undertaken by the mathematical
astronomer. Though it may not be immediately obvious, there are two
normative questions embedded in this passage:
1. Should the natural scientist proceed in the same way as the astronomer?
2. Should the natural scientist, in studying animals, proceed from a study
of animals and their parts to a study of their causes?35
Concerning the first question, the precise contrast in this passage is
interesting and problematic: the verb used here to express what mathemat-
icians do with respect to astronomical phenomena can have the sense of
‘prove’ or ‘explain,’ but in many contexts it is closer in sense to ‘show,’
‘display,’ or ‘represent’; and in any case, we have here a classic ‘subordinate
science’ situation where the natural phenomena are being accounted for by
mathematical proofs. For that reason, it is by no means obvious that the
natural scientist should proceed in the same way. Moreover, the compari-
son relates specifically to the order of investigation: but questions about the
35
Note that the answer to this question will be very different depending on whether you think that the
material and generative conditions are the only or primary causes of animals and their parts or you
think that the form of the completed animal determines the nature of its parts and the order in which
they must come to be. I believe it is Aristotle’s sensitivity to this difference that leads him to delay
answering it.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
order of inquiry are much more complicated when we are thinking about
animals and plants.
In considering why this question is being raised in the form that it is, it is
helpful to recall that all of PA i is structured around a distinction between
two sorts of natural beings: those that are eternal and governed in their
movements by unconditional necessity, and those that come to be and pass
away and are governed in their changes by conditional necessity (PA i.1,
639b21-27, 640a1-2, 642a2-17; PA i.5, 644b22-645a11). In PA i.5, as Aristotle
looks forward to the study of animal parts that will commence in book ii,
he begins with just such a reminder: Among natural beings, some are
eternal and some undergo generation and perishing (644b22-25).
Knowledge of the former, in virtue of their divinity, is valuable but hard
to come by; while knowledge of the animals and plants around us is much
easier to come by (644b31-645a4). He then speaks as if the study of the
eternal natural bodies, such as it is, has been completed:
Since we have completed stating the way things appear to us about the
divine things, it remains to speak about animal nature (περὶ τῆς ζωικῆς
φύσεως), omitting nothing in our power, whether of lesser or greater esteem.
For even in the study of animals disagreeable to perception, the nature that
crafted them (ἡ δημιουργήσασα φύσις) likewise provides extraordinary
pleasures to those who are able to know their causes and are by nature
philosophers. (PA i.5, 645a4-10)
In considering the comparison being made in the second normative
question about the proper order of inquiry, then, it is important to keep
in mind Aristotle’s views about the differences between the objects studied
by astronomy and zoology, our cognitive access to those objects, and (as
a consequence) the available modes of investigation and explanation. The
objects of astronomy are, Aristotle argues in De caelo, eternal – the only
change they engage in is movement in place, and their apparent motions
can be reduced to endlessly repeating circular patterns, for which the
mathematician can provide geometric explanations. We can be sure of
very little about their material constitution or the physical causes of their
motions, since our access to these objects is extremely limited.36
Animals, on the other hand, are a special class of natural substances that
come into being in a particularly complex and yet coordinated manner that
36
We will revisit these issues in Chapter 8. For a full discussion of the passages in Cael. in which
Aristotle expresses the epistemic limitations imposed on such research, see Bolton 2009, 51–82.
Falcon 2005 does a fine job of pointing out how unusual the views defended in Cael. were, which
helps explain the widespread skepticism with which they were greeted.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
is evidently goal-directed. Moreover, there are a vast number of differ-
ent animals, and the number of ways in which they differ is yet more
vast. Among their differences are a wide range of coordinated activities
that constitute their distinctive ways of life. While this makes under-
standing them much more arduous, we are able to study them in great
detail to our heart’s content – and if we do so with the eye of the
philosopher, we can come to understand the causes of their structures,
habits, and activities.
Thus, the prescription to study the phenomena first and then the
reason why and the causes will likely have very different methodological
implications for mathematical astronomy and for the methodical inquiry
concerning the animals (ἡ περὶ τῶν ζῴων μέθοδος; Long. 6, 467b7-9). To
cite one class of problems on which we will focus shortly: in studying the
phenomena first, should one first study the phenomena related to the
process of animal generation, and then proceed to study their causes, and
then move on to investigate the phenomena related to the anatomy,
physiology, and behavior of fully developed animals? Might the process
of coming to be actually be the cause of the fully developed animal and its
features? If so, should we study the phenomena related to fully developed
animals first, or investigate their development first? Moreover, how are
we to compare the way in which mathematicians explain the observed
motions of the heavens with the sorts of causal explanations that are
appropriate in coming to understand animals? And once we introduce
the possibility that the end of development, the fully developed animal, is
the cause of coming to be, rather than the other way around, questions
about the order of inquiry become intertwined with questions about
different kinds of priority.
These concerns, I would suggest, lie behind the fact that the answer to
this question about the proper way to order the investigation of animals,
like the answer to the first question about whether to begin with an
investigation of specific kinds or features common to many kinds, is also
postponed. Before this second question can be answered, the nature of
causal explanation in the study of animals needs to be addressed.
6.3.4 Normative Question 3: The Priority of Goal-Causation

Thus, after asking, and not answering, the question about whether the
zoologist’s investigations should be patterned on those of the mathematical
astronomer, Aristotle asks another question about the order of
investigation.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
[S]ince we see more than one cause of natural generation, for example both
the cause for the sake of which and the cause whence comes the origin of
motion (τήν θ’ οὗ ἕνεκα καὶ τὴν ὅθεν ἡ ἀρχὴ τῆς κινήσεως), we need also to
determine, about these causes, which is naturally first and which second.
(PA i.1, 639b11-14)
Unlike the first two questions, however, Aristotle turns immediately to the
task of answering this one. From 639b14-21, he defends the causal priority
of the ‘cause for the sake of which’ (the so-called ‘final cause’) over the
‘causal source of motion’ (the so-called ‘efficient cause’) in things com-
posed by nature and by art. The argument runs as follows:
1. The cause for the sake of which things come to be is the logos.37
2. In things constituted either by art or by nature, the logos is the starting
point.
3. [Evidence for number 2] In defining health or a house, either in
thought or in perceiving, the doctor and builder give the accounts
and causes of the things they produce, and the reason why it must be
produced in a certain way.
4. That for the sake of which and the noble is present more in the
products of nature than in the products of art.
There are a number of puzzles in this argument. The first premise is
critical, but it is undefended. Does logos here mean simply ‘account’ or
(given ὁρισάμενος at 639b17) ‘defining account’? Or does it refer to the
object of the definition, as logos often seems to? The evidentiary support
provided by number 3 does not help in answering these questions about
natural coming to be, for that support is restricted to evidence drawn from
the crafts – accounts, reasons why, and causes are all given by the doctor
and builder. Most fundamentally, what does Aristotle mean by saying that
the cause for the sake of which is the logos both for things constituted by art
and by nature? One can see the need for asserting number 4, since the
evidence in number 3 only supports giving priority to the final cause in the
case of artistic production: without number 4, that evidence provides no
support for number 2. But again, number 4 is simply asserted without
justification.
The evidence provided by number 3 is perhaps to be understood in the
following way: a doctor, in treating a patient with (say) liver disease, must
first have an understanding of what a healthy liver is, and that, Aristotle
(quite correctly) thinks, begins with understanding what the liver is for,
37
A claim repeated at GA i.1, 715a8-9: ὅ τε γὰρ λόγος καὶ τὸ οὗ ἕνεκα ὡς τέλος ταὐτόν.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
what it contributes to the life of an animal with a liver. This in turn will
guide the doctor in determining what he must do to transform the diseased
liver into a healthy one – and the same goes for a builder of houses. The
plan for the house to be must be the starting point in providing the reason
why he must take certain actions in a certain order with certain materials
and tools to achieve his goal. That, however, only provides evidence for the
claim that ‘the cause for the sake of which’ is by nature first in the case of
natural generation if the telos plays the same role in nature that it does in
these two arts. Aristotle asserts (in number 4) that it does, but at this point
provides no evidence or argument for the claim. And it is hard to see how
the telos could play the same role in natural generation, given the import-
ance in the craft cases to the craftsman having a conception of the end
to be.
The picture so far, then, is that in nature, as in the case of healing or
building, the goal of development, the fully developed animal, is actually
the starting point of an animal of a certain sort coming to be. In the case of
the art of medicine or house-building, this is explained by reference to the
doctor or builder having a defining conception of the final product that
guides what he does, and thus the changes that the body or the building
materials must undergo to achieve the end. We are not provided with any
parallel explanation of how the nature of the developing animal could
begin with something analogous to a defining conception of the fully
developed organism. As one reads on in PA i.1, a sketch of Aristotle’s
view is finally provided at 640a20-27, but his thoughts on natural goal-
directed development are only fully articulated in GA ii.1–6. Thus, the
question of which, among the variety of causes of natural generation that
we see (ὁρῶμεν, 639b11), is by nature first or primary is answered in this
passage, but the answer is not so much defended as asserted.
The next step in the argument perhaps picks up on the fact that the
other cause he gives as an example of the numerous causes of natural
generation is the motive cause. He notes that an error in all previous
thinkers is that they refer their accounts (λόγοι) back to necessity without
distinguishing the different ways in which we speak of necessity. Some
natural things are apparently eternal and do not come into being or pass
away, and these, Aristotle tells us, are necessary without qualification
(ἁπλῶς). Other natural things, such as animals, come into being and
pass away. In things that come to be by nature, as with things that are
the products of art, what is to be, the end of the developmental process,
necessitates that certain materials and motive causes be present. If a certain
house is to be, bricks, mortar, and timber must be present and must be acted
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
upon in specific ways by builders. Again, Aristotle gives no analogous
example of natural generation, simply concluding that “[i]t is the same
way too with things that come to be by nature.” His central point is that the
necessities governing the material and efficient causes are conditional on
the goal to be achieved – the material and efficient causes do not necessitate
the production of the goal, but they are necessary for the production of the
goal.38
With these key ideas in place, he then argues, from 640a1-9, for a distinct
manner of demonstration for contexts in which conditional necessity is
operative, and where the starting points and definitions will identify the
goals of the processes that are productive of those goals.39
6.3.5 Answering Normative Question 2: Investigating Animals

Is Complicated
And, finally, with the machinery of teleology integrated with conditional
necessity in place, Aristotle has prepared the ground to return to his second
question: whether we should proceed as the astronomers do, first studying
the phenomena and then the causes. But he proceeds to answer this
question in what appears to be, at least if read without that preparation
in mind, a puzzling manner. He reminds us that those before him (he cites
Empedocles in particular) have sought to understand the way things are by
investigating how they came to be that way, without asking whether that is
the appropriate way to proceed. Then, as if it were a response to such
people, he says:
It seems we should begin, even40 with generation, precisely as we said
before: first, one should get hold of the phenomena (τὰ φαινόμενα
ληπτέον) concerning each kind, and then state their causes (τὰς αἰτίας
τούτων λεκτέον). (PA i.1, 640a13-15)
And he immediately goes on, as if he thinks it will help us to understand
this claim, with yet another appeal to craftsmanship:
38
PA i.1, 639b21-640a1; cf. Ph. ii.9, 200a30-34: “Plainly, then, the necessary in things which are natural
is that which is given as the matter, and the changes it undergoes. The student of nature should state
both causes, but particularly the cause which is what the thing is for; for that is responsible for the
matter, whilst the matter is not responsible for the end” (Charlton trans., 1992). Note the explicitly
normative conclusion about how the student of nature should present his causal understanding.
39
Again, it is important to read this passage together with Ph. ii.9, 200a15-b8 where he compares such
teleological demonstrations with those in geometry.
40
The καί translated as ‘even’ could also be rendered ‘also.’ I’m inclined to read this sentence to be
making a point about the importance of observing generation in a context where it is typically not
studied carefully or perhaps not at all.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
For even with house-building, it is rather that these things [i.e., what
happens during coming to be] happen because the form of the house is
such as it is, than that the house is such as it is because it comes to be in this
way. For generation is for the sake of being (ἡ γὰρ γένεσις ἕνεκα τῆς οὐσίας
ἐστίν); being is not for the sake of generation. (640a15-19)
But how does this reminder of the conditionally necessary nature of goal-
directed generation help us to understand that in studying generation too
we should first grasp the phenomena and then search for their causes? This
appears to be a non sequitur.
The solution to this puzzle is to place the argument in its context. What
stands between Aristotle’s statement of the question and its answer is
Aristotle’s insistence that natural generation has more than one cause,
that goal causation has priority over motive causation,41 and that the
necessity of the materials and motive causes in generation is conditional:
they do not necessitate the goal, but they are necessary conditions for its
achievement. Given that context, then, it is natural that, in the passage we
are currently discussing, he criticizes his predecessors for explaining the
attributes of the developed animal as a necessitated (and coincidental)
outcome of the motive causes that produced them – and that, against
them, he insists that coming to be happens as it does because of the goal
toward which it is directed. Thus, after reviewing what Empedocles had
missed by assuming that one accounts for (e.g.) the segmented backbones
of vertebrates by explaining how they happen to get twisted as they come to
be (640a19-22), Aristotle summarizes what he takes to be the appropriate
method.
Hence it would be best to state that, since this is what it is to be a human
being (τοῦτ’ ἦν τὸ ἀνθρώπῳ εἶναι), on account of this it has these <fea-
tures>; for it cannot be <a human being> without these parts. If one cannot
say this, one should say the next best thing, that is, either that in general it
cannot be otherwise, or that it is at least good <that it is> thus. And certain
other things follow. And since it is such <as it is>, its generation necessarily
happens in this way and is such as it is. (This is why this part comes to be
first, then that one.) (PA i.1, 640a33-b3)
Notice that in this passage, the causal relationships among the features of the
actual members of a kind must be understood in order to explain the patterns
that one sees in generation. Herein lies the complexity in studying animals and
41
There is a considerable literature around the question of the nature of this priority. For a recent
survey of the issues at stake that takes a somewhat different view from that defended here, see
Leunissen 2009, 99–108.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
their generation that prevented Aristotle from simply affirming that inquiry
should proceed here in exactly the same manner as in mathematical astron-
omy, where one observes repeated regularities and tables cataloguing these
regularities, and then appeals to the mathematicians for explanations. In order
to understand biological generation, one actually needs to have a systematic
understanding of the facts about the fully developed animals one is studying
and the lives they live, in order to move on to a causal understanding of their
coming into being – in Aristotle’s view, in fact, that fully developed animal is
the cause for sake of which the developmental process is taking place. To
know what that animal is – its anatomy, physiology, activities, and way of
life – is to know the primary cause of that process. Thus, the order of inquiry
would be: (i) a fact establishing and organizing study of the parts of animals;42
(ii) an attempt to establish the causes of the universal relationships at various
levels of generality and specificity established at stage (i); (iii) a fact establishing
and organizing study of animal generation; and (iv) a causal investigation of
animal generation, seeing how the efficient cause is controlling its instruments
(heating and cooling) such that an end product one in form with the efficient
cause comes about.
At the beginning of GA, after reviewing the four causes,43 Aristotle
makes the following programmatic statement:
Concerning the other causes, then, we have spoken (for the account and that for
the sake of which as goal are the same (ὅ τε γὰρ λόγος καὶ τὸ οὗ ἕνεκα ὡς τέλος
ταὐτόν), and the parts are matter for animals – in every case the nonuniform
parts are matter for the whole animal, the uniform parts for the nonuniform,
and the so-called elements of bodies matter for the uniform parts);44 it remains
[to discuss] the parts that contribute to the generation of animals, about which
nothing was previously determined, and, concerning the motive cause, [to
discuss] what its source is (περὶ αἰτίας δὲ τῆς κινούσης τίς ἀρχή). And the
investigation about the motive cause and about the generation of each [animal]
are in a way the same; wherefore our account has united them, putting these
parts last in the account of the parts, and the account of the starting point of
generation next in order after them. (715a7-18)45
42
Complicating the picture presented here, Aristotle explicitly puts off the causal investigation of the
parts related to generation in PA until the first book of GA. Cf. PA ii.7, 653b13-18; iv.10, 689a4-20;
GA i.1, 715a11-18.
43
Of note in this review is that ‘form’ is not mentioned – in its place is ὁ λόγος τῆς οὐσίας, the account
of the substantial being.
44
To this discussion of how the elements, uniform parts, and nonuniform parts are related, compare
PA ii.1, 646a13-b28.
45
David Balme in his translation brackets the GA i.1 introduction and in his notes says that it may be
post-Aristotelian (1992, 127). His only grounds are that it is ‘stylized’ and is ‘so different’ from the PA
i introduction. That is true: but there are striking differences between the introductions of every
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
This introduction stresses the need to investigate the actual parts and what
they are for first, postponing the discussion of the parts concerning generation
until last, and uniting that discussion (finally) with the discussion of the cause
of generation. Now when he says that this will be a discussion of the motive
cause, and that he has already given an account of the other three causes, this
does not mean that the other three causes will not be discussed (indeed, book
ii opens with a long discussion of the final cause of the separation of male and
female principles in sexually reproducing species). What is true is that the
primary focus of the account of the beginning and process of development in
books ii and iii is on determining the nature of the generative capacity or
potential and tracing its effects through the early stages of development
(concluding with an analogous account of asexual, ‘spontaneous’ generation);
that the primary focus of book iv is on the efficient causes of the differentiat-
ing characteristics of male and female, of inherited likeness, and of such
phenomena as how many offspring are born together, the mixture of males
and females, and various sorts of deformities; and that the primary focus of
book v is on the efficient causes of ‘more and less’ variations within kinds (hair
color and texture, eye color, voice, baldness, etc.).
A great deal of thought, then, needs to go into parsing the difference
between investigating the phenomena and investigating their causes when it
is an investigation of animals and their generation that we are considering,
principally because when we are investigating both the process of coming to
be and the lives of fully developed animals as well, it turns out there are two
very different kinds of phenomena to be studied, and getting a proper
understanding of the facts about animal generation requires that we already
have an understanding of what the parts that are coming to be are coming to
be for.
As with our discussion of question 1 concerning the proper level of
abstraction to initiate our investigation, it is important to see that these
methodological norms reflect, or are reflected in, Aristotle’s practice – in
this case, in GA ii.6, a passage I used earlier (in Chapter 3)46 to illustrate the
use of the concept methodos in practice to refer to the norm-guided
practices of distinct domains of inquiry. Having presented his theory of
the causal contributions of male and female to sexual generation, Aristotle
work in the corpus; and PA i, as Balme himself stresses, is not the first book of the investigation of
animal parts but a philosophical discussion about how the investigation of living nature in its
entirety ought to be carried out. I will therefore treat the opening lines of GA as a genuinely
Aristotelian introduction to the study of generation, comparable to the opening pages of PA ii, to
which it clearly alludes.
46
See section 3.4.2.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
6.4 Preliminary Results 169
prepares his readers for the task of accounting for the formation of the parts
in development. As so often at such transitional points in his scientific
writing, however, he opens by worrying over the best way to proceed,
about the norms that should govern the inquiry. We can now look at this
passage in light of the norms regarding causal inquiry of PA i.1.
Thus, as many of the instrumental parts as are generative in nature must be
present first (for such <parts> are for the sake of others as their origin), while
as many <parts> as are among those present for the sake of another, but not
generative in nature, must be later. Wherefore it is not easy to distinguish
which things are prior, those things which are for the sake of something else,
or that for the sake of which these other things are. For the motive parts,
being prior in generation to the end, intrude, and it is not easy to distinguish
the motive parts relative to the instrumental parts. Yet it is necessary to
inquire what comes to be after what according to this methodos; for the end
of some things is posterior, while the end of other things is prior (καίτοι
κατὰ ταύτην τὴν μέθοδον δεῖ ζητεῖν τί γίγνεται μετὰ τί· τὸ γὰρ τέλος ἐνίων
μὲν ὕστερον ἐνίων δὲ πρότερον). (GA ii.6, 742b3-12)
Because biological generation is goal-directed, the order in which the parts
appear is dictated by teleologically established priorities; if a part plays
a crucial role throughout the generative process – as the heart most
certainly does – it must come to be early in development even if it is an
essential part of the telos, as the organ responsible for nutrition, perception,
and locomotion. On the other hand, if some part has a teleological role to
play only in the completed animal (such as the lung), it will develop later.
One cannot simply read off teleological priority from the temporal
sequences that one observes taking place during development.47
6.4 Preliminary Results

It is now time to step back from these first three questions, which set the
agenda for much of the rest of PA i, and reflect on the way in which they
establish a set of norms of inquiry for the study of animals – norms that,
47
As an aside that underscores Aristotle’s point here – William Harvey takes Aristotle to task in his
Exercitationes de Generatione Animalium for thinking that the first appearance of a pulsing bloody
spot in the developing chick embryo is the heart. Harvey held that the blood carries the causal
agency (in the form of its heat) guiding generation, and therefore thought it must be present prior to
the heart in order to produce it – because he gives causal priority to the blood, he insists that
Aristotle must be mistaken and that what he was actually observing was not a tiny heart beating but
blood pulsing with the power of life. Aristotle, on the other hand, thought the blood was merely
instrumental, serving to distribute the generative heat throughout the organism, while the heart was
the source of that generative heat. On this disagreement, compare Aristotle, HA vi.3, 561a4-15 with
Harvey’s Exercitationes de Generatione Animalium (Whitteridge 1981, 96, 241).
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
while fully consistent with Aristotle’s framework for scientific inquiry in
APo. ii, and the general norms for natural inquiry established in the
Physics, are specific to zoological inquiry.
1. The first point of note is that these norms take the form of answers to
a set of domain-specific, normative questions – questions about how an
inquirer, or inquiry, should proceed, if knowledge of the specified domain
is the goal. They assume an investigator of nature is faced with alternatives –
and indeed, Aristotle often has in mind methodological errors of his prede-
cessors that inhibited and limited their progress. These questions establish an
erotetic structure for the ensuing inquiry – which questions need to govern
our inquiry, and in what order should these questions be answered. Such
erotetic structures implement, in a domain-specific way, the erotetic frame-
work that is provided by the APo. ii, and they are clearly specifying and
enriching the more general norms of natural inquiry that are found in Ph. ii
for a zoological inquiry. Those more general norms of natural inquiry, as we
saw in Chapter 5, arise from Aristotle drawing out the implications of the
erotetic structure provided by the four distinct answers to the dia ti question
spelled out in Ph. ii.3 and 7.48 In Chapter 7, we will see similar features in the
methodologically oriented first chapter of de An.
2. It is also noteworthy that the three questions are, among other things,
about the order of inquiry: Should we start by grasping attributes common
to many kinds first and then proceed to the more specific?; should we, in
investigating animals, study the phenomena before inquiring into their
causes in the same way that the astronomers do?; should we give priority in
our investigation to determining the final cause or the efficient cause?49
3. Finally, question 1 concerns inquiry into the defining natures of things,
while questions 2 and 3 concern the order of inquiry aimed at causal under-
standing. As I noted, these two themes dominate PA i and are intertwined in
just the way we would expect if the framework for inquiry of APo. ii is
48
And recall the way in which that set of questions is introduced by first reminding the reader of the
Analytics framework: “For since our study is for the sake of knowing, and we do not think we know
each thing until we grasp τὸ διὰ τί about each thing . . . ’ (Ph. ii.3, 194b17-19); see the discussion in
Section 5.2.1.
49
For reasons of length I have limited the discussion to the three initial questions in PA i.1. It is already
clear, however, from looking at the norms recommended for answering those questions, that Aristotle
also develops a series of norms for the proper use of division in a zoological context (PA i.2-4, 642b5-
644a11); for applying necessity to teleological processes (639b22-640a9; 642a1-13); for the study of the
material and formal nature of living things, especially in light of the formal nature of a living thing
being its soul (640b4-641a32); and for delimiting the parts of the soul to be investigated by the natural
scientist (641a33-641b23). Finally, in chapter 5, there are a number of complex norms about how best to
integrate the methodology of division that dominates the discussion in chapters 2–4 with the
teleological and hylomorphic perspective articulated in chapter 1.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
operative. The overarching picture of the epistemic quest laid out in the
Analytics has not changed; but how to keep such a quest on track toward
achieving the goal of scientific knowledge of animals requires norms that are
largely domain-specific, influenced by distinctive features of the subject matter
under investigation and our epistemic access to it. In the case of animals, we are
dealing with a distinctively complex, dynamic, and multilayered investigation
of animals in all their vast variety, and of their amazingly complex, orchestrated
development. We cannot ignore these distinctive features of living beings if we
are to keep such an inquiry on track. Living beings are unities of matter and
form – but the relationship of matter and form is to be understood as that
between complex structures and their functional capacities. Equally important
is the fact that living beings are the end product of an astoundingly compli-
cated, goal-directed generative process, which makes it an especially difficult
subject of inquiry. However, there is an upside: we have virtually unlimited
observational access to the subject matter, though this will involve developing
the expertise of dissection, especially difficult in the context of studying
embryological development. Aristotle opens the encomium to the study of
animals in PA i.5 by remarking on both aspects of this particular methodos.
Among the substantial beings constituted by nature (Τῶν οὐσιῶν
ὅσαι φύσει συνεστᾶσι), some are ungenerated and imperishable through-
out all eternity, while others partake of generation and perishing. Yet it has
turned out that our studies of the former, though they are valuable and
divine, are fewer (for as regards both those things on the basis of which one
would examine them and those things about them which we long to know,
the perceptual appearances (τὰ φανερὰ κατὰ τὴν αἴσθησιν) are altogether
few). We are, however, much better provided in relation to knowledge
(πρὸς τὴν γνῶσιν) about the perishable plants and animals, because we live
among them. For anyone wishing to labor sufficiently can grasp many
things about each kind. Each study has its attractions. Even if our contact
with the eternal beings is slight, nonetheless because of its surpassing value
this knowledge is a greater pleasure than our knowledge of everything
around us, even as a chance, brief glimpse of the ones we love is a greater
pleasure than seeing accurately many other great things. Perishable beings,
however, take the prize in respect of scientific knowledge (λαμβάνει τὴν
τῆς ἐπιστήμης ὑπεροχήν) because we know more about them and we know
them more fully. Further, because they are more of our own nature, they
provide a certain compensation compared with the philosophy concerned
with divine things (πρὸς τὴν περὶ τὰ θεῖα φιλοσοφίαν). Since we have
completed stating the way things appear to us about the divine things, it
remains to speak about animal nature (περὶ τῆς ζωϊκῆς φύσεως), omitting
nothing in our power, whether of lesser or greater esteem. For even in the
study of animals disagreeable to perception, the nature that crafted them
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
(ἡ δημιουργήσασα φύσις) likewise provides extraordinary pleasures to
those who are able to know their causes and are by nature philosophers
(τοῖς δυναμένοις τὰς αἰτίας γνωρίζειν καὶ φύσει φιλοσόφοις). (644b22-
645a10)
In this passage, Aristotle distinguishes the study of the ‘divine’ natural
beings, the heavenly bodies, from the perishable ones along both metaphys-
ical and epistemological dimensions: their distinctive natures – eternal and
imperishable versus generated and perishable – and our perceptual access
to them as objects of investigation. He freely acknowledges the honorable
nature of any knowledge we can achieve about ‘the divine things’ but also
acknowledges that it is the perishable things that ‘take the prize’ epistemo-
logically. The perspective of the investigator is also critical: Those who
study the perishable things in the right way – as natural philosophers who
focus on the nature that crafted them and their causes – will derive great
pleasure from studying even the most disagreeable animal. As he explains
a few lines later:
One should approach the inquiry about each of the animals without
disgust, since in every one there is something natural and beautiful
(φυσικοῦ καὶ καλοῦ). For what is not haphazard but rather for the sake
of something is in fact present most of all in the works of nature; the
end for the sake of which each animal has been constituted or comes to
be takes the place of the beautiful (τὴν τοῦ καλοῦ χώραν εἴληφεν).
(645a21-26)50
It is due to these two kinds of difference that investigating these two kinds
of natural beings requires different norms of inquiry: each is a distinct
methodos (a methodical inquiry) because each is governed by distinct
norm-governed practices – distinct because they study radically distinct
kinds of beings, and because our access to those beings is also radically
different. The use of methodos to refer to the investigation itself in PA
i reflects its focus on articulating and defending domain-specific norms for
the investigation. The intimate relationship between the two senses of the
term is particularly clear in the closing sentence of this introduction to the
study of animals.
Enough said about our mode of inquiry (περὶ τοῦ τρόπου τῆς μεθόδου); we
must attempt to state the causes both of the common and of the distinctive
50
I depart from the translation in Lennox 2001a in two respects: I now translate ζήτησις consistently as
inquiry, and I now think that the use of τὸ κάλον in this text has a decidedly aesthetic aspect. I owe
the latter change of heart to reflecting on conversations with Richard Kraut.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
attributes, beginning first, as we have determined, with those that are first.
(646a1-4)
In Chapter 7, I will turn to the methodological introduction to Aristotle’s
investigation of the soul, De anima i.1, a chapter in which the concept of
methodos again signals Aristotle’s concerns about the norms that are
appropriate for such an inquiry.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
chapter 7
The Soul
One Subject, Two Methods?
It remains then for reason alone to enter from without and alone to be
divine; for bodily actuality does not participate at all in its actuality.
(GA ii.3, 736b27-29)1
Chapter Summary. The long, introductory chapter of the De anima

provides a detailed and complicated erotetic structure for a norm-governed
inquiry into the soul. Answering the complex list of questions laid out in that
chapter provides the norms that govern Aristotle’s critical discussion of his
predecessors in de An. i and his own positive account of the soul in de An. ii–
iii. In this chapter, I will begin with a careful study of that introductory
chapter from the standpoint of what we can learn about Aristotle’s methodos
for an inquiry into the soul, and then turn to a study of the way in which the
norms derived from answering the questions laid out in de An. i.1 govern the
inquiry that ensues. One surprising result of this study is that, contrary to a
widespread assumption, the de An. is, for good reason, not simply one of a
number of contributions to natural science.
7.1 Introduction
Let us return, yet again, to the beginning of the Meteorology. It will be
recalled that Aristotle recites an appropriate course of study for an investi-
gator of nature (338b20-339a10) in order to locate the investigation of
meteorological phenomena. For the topic of this chapter, two things are
significant about this course of study: (i) nothing corresponding to the
subject of the De anima, the investigation of the soul, is mentioned; and (ii)
the course of investigation of nature is said to culminate in a study of
animals and plants. It is sometimes assumed that the investigation of the
1
λείπεται δὴ τὸν νοῦν μόνον θύραθεν ἐπεισιέναι καὶ θεῖον εἶναι μόνον· οὐθὲν γὰρ αὐτοῦ
τῇ ἐνεργείᾳ κοινωνεῖ <ἡ> σωματικὴ ἐνέργεια.
174
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
7.2 De anima i.1 on the Proper Method for Investigating Soul 175
soul reflected in the de An. is directly implied by the claim that the study of
nature is completed when we have concluded our investigation of animals
and plants. I have already questioned that assumption. Here I shall argue
that the lack of mention of an investigation of soul in this recounting of the
course of natural inquiry reflects a tension at the heart of Aristotle’s
philosophy of nature. The question of the extent to which the soul is a
proper object of inquiry for the natural scientist is a difficult and compli-
cated one, and there are good reasons to set it aside in this overview of the
different natural inquiries. But before turning to that subject, we need to
begin by looking at the very first chapter of the de An. from the standpoint
of the way in which it provides the erotetic structure for soul inquiry, in the
way that PA i provided the erotetic structure for animal inquiry: a series of
related questions about how to investigate the soul is introduced, the
answers to which will provide the norms for an inquiry into the soul.
7.2 De anima i.1 on the Proper Method for Investigating Soul

De anima opens with praise for the inquiry into the soul (τὴν περὶ τῆς
ψυχῆς ἱστορίαν, 402a3-4; compare τῆς περὶ φύσιν ἱστορίας, PA i.1,
639a12). We can evaluate the knowledge of beautiful and honorable things
either in terms of accuracy or in terms of the intrinsic value of the objects
known. On either count, Aristotle tells us, it is with good reason that the
inquiry into the soul is placed in the first rank. He claims that such an
inquiry contributes greatly to all truth, but especially truth related to
nature. He justifies this last thought by saying that the soul is . . . well,
what precisely does he say that it is? The Greek is ἔστιν γὰρ οἷον ἀρχὴ τῶν
ζῴων (402a6-7). Ross paraphrases it as “since soul is the inner principle of
animal life”;2 Hamlyn takes οἷον a bit more seriously: “for the soul is as it
were the first principle of animal life”;3 Hicks is expansive: “the soul being
virtually the principle of all animal life”;4 Irwin and Fine take a minimalist
approach: “the soul is a sort of principle of animals”;5 as does Johansen: “the
soul is like a principle of animals.”6 Shields, like Hamlyn, supposes ἀρχὴ to
refer to a first principle: “a sort of first principle of animals.”7
Among these various renderings, Johansen and Irwin and Fine acknow-
ledge the plural, τῶν ζῴων, and their deflationary translation of οἷον ἀρχὴ,
‘like a principle,’ ‘a sort of principle,’ is also a precise reflection of the Greek
lacking in some other translations. Hicks supports his expansive translation
2
Ross 1961, 163. 3 Hamlyn 1968, 1. 4 Hicks 1907, 3. 5
Irwin and Fine 1995.
6
Johansen 2012, 85 and n. 18. 7 Shields 2016, 1.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
176 The Soul: One Subject, Two Methods?
by appealing to a passage in book ii, in which Aristotle argues that the soul
is a principle and cause of the living body (415b8-416a18; Hicks 1907, 176).
But given the long and daunting list of open questions Aristotle is about to
review at the very beginning of the inquiry, it is questionable whether he
would here be assuming at the outset something that will take considerable
argumentation to establish much later. It is more plausible to suppose that
his wording here is intended to leave open options that will eventually be
under critical scrutiny.
The desire to leave options open may well extend to the term that all the
above translators suppose refers to animals. As Andrea Falcon has noted
(Falcon 2005, 6 and n. 11; and see Hicks 1907, 184), in the Timaeus Plato had
already defended the use of the term τὸ ζῴον to refer to plants (77a5-6) as
well as to beings superior to animals and humans, and earlier had used it to
refer to the perceptible cosmos and the stars (38e7, 39e1-2). Falcon reminds us
of this Platonic background in support of reading the term here with an
‘open’ reference, not restricted to animals. In these opening remarks, all
Aristotle is insisting on is that, whatever things turn out to be ‘ensouled’ once
the dust settles, soul will be rightly venerated as an object of investigation
because it will be, in some way or other, a source or principle relative to those
things. Thus, the stress that all the above translations put on animal life may
not be justified, even though Aristotle often uses this term to contrast
animals with plants, and even sometimes (e.g., GA ii.1, 732a1-2) with
humans. His own, considered view is that plants, in virtue of having a
nutritive/generative capacity, have souls, and De caelo states a number of
times that the heavenly spheres are ensouled (e.g., Cael. ii.2, 285a29-30; 12,
292a20-228) – and Aristotle’s unmoved mover is a necessarily active living
thing that has its own activity as its object (Metaph. Λ.7, 1072b25-30).
I think it is reasonable to see Aristotle’s use of the expression οἷον ἀρχὴ
as a way of not insisting that soul is the principle (let alone the inner, or first,
principle) of living things. In fact, this phrase minimizes Aristotle’s starting
commitments in two ways: by leaving room for more than one principle
and by leaving the sense in which the soul is a principle open.
A non-question-begging translation, suitable to its place in this intro-
ductory passage, then, is “for the soul is a sort of starting point9 of living
things.” Since this is the beginning of an inquiry into a subject about which
8
The latter passage states that “it is necessary to suppose that they (the spheres of the stars and planets)
participate in action and life.”
9
I am going to use ‘starting point’ as a translation for ἀρχή here, since at the very least Aristotle is not
merely saying that the concept of soul is a scientific first principle. The possessors of the ἀρχή here are
living things. ‘Starting point’ is sufficiently generic that it can refer both to the principles of a science and
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
there is general agreement on almost nothing, the exact extension of beings
with soul should properly be left open, as should the questions of whether
there is one or more than one principle of living things and what sort of
principle or source a soul is.10
Aristotle chose his words carefully here: In the end, while he will insist
that all soul is an archê of living things in some sense, he will deny that all
soul is an inherent archê of motion (i.e., a nature). Specifically, he denies
that nous is a source of the sorts of changes investigated by the natural
scientist. This does not, however, imply that nous is not an archê in some
other way. Moreover, insofar as the life of the unmoved mover is to be
identified with the activity of thought, it can be an archê of living things in
a different, but equally legitimate, sense of the term, though not as their
nature.
Aristotle goes on in de An. i.1 to specify what it is about the soul that
needs to be investigated: its nature and substantial being, and then its
attributes, some of which seem to be properties of the soul itself, others
properties of living things that they possess in virtue of being ensouled
(402a7-10). We will have cause to return shortly to these two distinctions,
between essence and attribute, and between attributes of the soul itself and
attributes that belong to living things on account of the soul. For it is the
possibility of the soul having attributes not shared with the body that gives
rise to a tension that has implications for the norms that govern an inquiry
focused on the soul.
Given that Aristotelian souls, at least in the case of animals and plants,
are their formal natures, it is initially surprising to read that the investiga-
tion concerns the nature of soul – for this immediately points toward a
regress. Surely the soul is not the sort of thing that can have a nature, a
starting point of motion within itself; rather, it is such a starting point. In
fact, that it is some sort of starting point for living things was our initial
assumption.
Perhaps we can avoid worrying about a possible regress by recalling that,
in his lexical entry on φύσις (nature) in Metaph. Δ.4, Aristotle explains that,
starting from the sense of ‘nature’ that refers to a natural thing’s being, the
the sources of motion or being in natural things, and by adopting it I can allow for the possibility that
Aristotle is either thinking of soul as an ἀρχή in both senses, or intentionally suggesting that such
questions should be left open at the beginning of the inquiry. Of course, in de An. ii.4 Aristotle argues
that the soul is a cause in three ways: (i) as source of motion, (ii) as the source of being for living things,
and (iii) as that for the sake of which the instrumental body exists (415b8-27).
10
Given that, according to Metaph. Δ.1, there are perhaps six or seven distinct sorts of ἀρχή, it is
perhaps surprising that, despite all the other differences, all of the translators referred to earlier settle
on ‘principle.’
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
term has been extended to refer to the being of anything.11 Thus, if it is
appropriate to say that one is investigating the οὐσία of soul (i.e., what it is
essentially), then in this ‘extended’ sense one can also say one is investigat-
ing its ‘nature.’ This reading fits well with the fact that when Aristotle
begins to list the questions that must be answered in our investigation into
soul, the first question is to which category soul should be assigned. At this
point the ‘what is it?’ inquiry is open to the consideration of answers such
as, ‘Soul is a sort of bodily affection’ or ‘Soul is a harmonic ratio’ – that is,
the sorts of answers that are considered and rejected throughout the
remainder of book i.
The introduction to the de An. next warns the reader that, while it is an
inquiry of great importance, it is also an inquiry into a subject that is “in
every way and in all respects the most difficult about which to get hold of
any secure belief” (402a10-11). How one reads the rest of this chapter turns
on how one interprets what Aristotle says about the character of this
difficulty. Let us turn next to that question.
The difficult nature of the task at hand turns not so much on the fact
that we are investigating the soul as on the fact that we are investigating its
‘substantial being and essence.’ The issue he spends considerable time on at
the very outset is whether, as with the method of demonstration for gaining
knowledge of the attributes of things, there is a single, common method to
be used when we are investigating essence. Even if there is, however, there
are still many problems. Aristotle reminds us that while we might not, in
that case, need to seek a new method of study for each new investigation, we
will still need to search for unique starting points: “For there are different
starting points for different things, as in the case of numbers and of planes”
(402a21-22).
The chapter’s first question, then, is about the proper mode of inquiry:
For in fact since the inquiry is one which is common to many different
things – I mean the inquiry concerning the substantial being and what it is
(τοῦ περὶ τὴν οὐσίαν καὶ τὸ τί ἐστι) – perhaps it might seem to someone
that there is some one method of inquiry (μία τις μέθοδος) regarding
everything about which we wish to know its being, just as there is demon-
stration (ἀπόδειξις) of the distinctive attributes of things, in which case we
ought to inquire into this method (ζητητέον ἂν εἴη τὴν μέθοδον ταύτην).
However, if there is not a single common method of inquiry concerning
what a thing is, our task becomes yet more difficult; for we will need to grasp
what the mode of inquiry is in each case. But if it should be apparent,
11
Metaph. Δ.4, 1015a12-13.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
whether it be demonstration, division, or some other method of inquiry
(ἀπόδειξίς . . . ἢ διαίρεσις ἢ καί τις ἄλλη μέθοδος), still there are many
puzzles and distractions – from what basis ought we to inquire (ἐκ τίνων δεῖ
ζητεῖν); for different inquiries have different starting points, as for example
those of numbers and of planes. (402a11-22)
Before discussing the substantive issues raised by this passage, it is worth

recalling the lessons of Chapter 3 with this passage before us. There are four
uses of μέθοδος in a mere eleven lines, in a discussion focused on the
question of how one ought to carry out a specific inquiry.
It is helpful, in considering the options presented, to recall the dialect-
ical development of APo. ii.1–10.12 Aristotle is there asking about how
inquiries seeking definition (i.e., knowledge of what things are) are
related to inquiries seeking causal knowledge (i.e., knowledge of the
reason why). The options on the table for proceeding from existence to
essence are demonstration, which he discusses and rejects in chapters 3 and
7; and division, which he has already ruled out as a proper form of proof
in APr. i.31, and which he rejects as a proper method of inquiry leading to
definition in APo. ii.5.13 From chapter 8 on, he articulates another method
for this process, described, however, at such a level of abstraction that
each domain of study will have to enrich its recommendations by means
of a domain-specific set of norms. Moreover, having apparently rejected
both demonstration and division as methods for establishing definitions,
in chapters 10–14 both are revived as having a role to play in this other
method.
The inquiry that is about to ensue, then, into the nature of the soul, is
firmly within the erotetic framework established in APo. ii.14 In any
domain of knowledge, the necessary attributes of the objects in that
domain are to be demonstrated from a set of starting points that are
indemonstrable (APo. i.3, 72b19-25; i.6, 75a29-37; cf. EN vi.5, 1139b29-
32). Among these are principles that identify the being and essence of the
primary objects in each domain (APo. ii.10, 94a10-12; ii.13, 96b7-14). But
since these are not themselves demonstrated, the question arises, how one
investigates and discovers these principles (APo. i.2, 71b17-19; i.3, 72b23-24;
ii.13, 96a20-23). This is precisely the issue being identified as problematic
here on the opening page of De anima; for however we understand it,
12
This in turn has an Academic background, since division, in dialogues such as the Phaedrus, Sophist,
and Statesman, is viewed as the primary tool for inquiry into being and essence.
13
Though in chapter 13 he acknowledges that it can be helpful in hunting for the essence.
14
A point properly stressed in Sisko 1999, 249–267. As Johansen puts it: “This passage takes a page, or
rather several pages, out of the APo” (Johansen 2012, 10).
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
Aristotle’s baseline assumption is that to investigate the soul is to
investigate some sort (or sorts) of archê of living things. Knowledge of
its essence will certainly count as a scientific starting point if anything
will.
In the remainder of de An. i.1, Aristotle provides the outlines of an
answer to this first question, regarding the proper mode of inquiry into the
nature of the soul; but before doing so, he presents in outline a series of
other, essentially metaphysical questions, questions that provide the struc-
ture for the inquiry in books ii and iii, and for the critical review of the
alternatives he finds inadequate.
7.3 The Erotetic Structure of Inquiry into the Soul

As in the opening pages of PA i, Aristotle is establishing the erotetic
structure of the investigation by asking a set of questions that will structure
the inquiry to follow. As with our discussion of PA i, we will look carefully
at only the first few questions Aristotle considers and the striking way they
serve to structure his positive account of the soul in book ii. My goal here is
not only to explore these questions and answers on their own, but also to
draw attention to how different they are from the questions that shape the
inquiry into animals in the PA i.1, and to think about their source. We have
already discussed the first question regarding whether there is a common or
a distinctive method of inquiry for investigating the nature of the soul, a
question to which we will return later. Let us now turn to the second and
third questions.
Question 2. 402a22-25
First, perhaps, it is necessary to determine in which of the kinds <the soul
is found>, that is, what it is (ἐν τίνι τῶν γενῶν καὶ τί ἐστι); I mean, whether
soul is a this and a substantial being (τόδε τι καὶ οὐσία), a quality, a quantity,
or even some other of the categories that have been marked off.15
Question 3: 402a25-26
. . . and further, is soul in potentiality (ἐν δυνάμει) or rather some kind of
fulfillment (ἐντελέχειά τις)?
15
This may seem an odd question. However, as one reads through his careful critique of previous
accounts of the soul in de An. i, it becomes clear that, apart from a number of materialist theories, the
other alternatives are that the soul is a ‘self-moving number’ (404b30), a ‘ratio,’ or a ‘harmony’ of
some sort (407b28ff; i.e., in the category of quantity), or that it is movement of some sort (see note
17). In this context, determining how to categorize the soul is an important first step and in fact, as
I’ll discuss shortly, this is a specific application of a general norm of inquiry specified at APo. ii.8.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
7.3 The Erotetic Structure of Inquiry into the Soul 181
There are three features in Aristotle’s formulation of these questions about
the metaphysical status of the soul that indicate the care with which he is
framing the investigation. (i) There is a norm of inquiry implicit in the
wording: it is necessary to answer this question about the soul’s categorical
status first. (ii) Aristotle has two terms that often overlap in reference and
are often both translated as ‘actuality’ – ἐντελέχεια and ἐνέργεια.16 The
latter often simply refers to activity or movement, while ἐντελέχεια con-
notes a state of fulfillment or realization. The contrast here is, then,
between something in a state of potentiality and that same thing in a
fully realized state. (iii) The nonspecific qualification on the term
ἐντελέχεια, τις, is also noteworthy. As we will see, Aristotle has a very
specific worry that is signaled by that qualification.
All three features of the way this sentence is worded point toward the
way in which Aristotle will answer these metaphysical questions at the
beginning of book ii.
The general inquiry, we have already been told, is into the nature and
essence of the soul; but Aristotle is now claiming that the proper place to
start such an inquiry is to ask, at the most general, metaphysical level, what
the soul is – that is, to inquire into its categorial status, and then its modal
status.
As I discussed in Chapter 2, APo. ii.8, 93a21-27 identifies two ways in
which we can establish whether some purported object of investigation
exists or not, and one of those ways is by grasping in an unspecified way the
wider kind into which it falls:
Sometimes we grasp whether something is accidentally, and sometimes by
grasping something of the thing itself, for example thunder, that (ὅτι) it is a
certain noise in the clouds; eclipse, that it is a certain privation of light;
human being, that it is a certain animal; soul, that it is self-moving.
(93a21-24)
In Chapter 2, I emphasized the wide range of examples in this passage to

stress the ‘domain neutrality’ of the Analytics. Here I want to draw atten-
tion to two other features of these examples: (i) their variable categorial
status and (ii) the indefinite qualifiers (‘a certain sort,’ τις). These are all
intended as examples of grasping whether something is by “grasping
16
Aristotle comments on the relationship between these terms at Metaph. Θ.3, 1047a30-1047b2, also
discussed in Blair 1992, Menn 1994, and Graham 1995. In this same passage, Aristotle contrasts
investigating being from a categorical perspective and from the perspective of potency, fulfillment,
and function – the two perspectives reflected in Questions 2 and 3 about the ontological status of the
soul in de An. i.1.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
something of the thing itself”; that is, they are all cases of getting a
nonspecific handle on the kind of thing it is. The kind mentioned in APo.
ii.8 when citing the inquiry into the soul as an example is “self-moving” –
one of the views to be critically evaluated in de An. i.17 The thought
outlined in APo. ii.8, and instantiated here, is that an important first
step in inquiry is to try to locate the object of inquiry, in a preliminary
way, in the right general category or kind; but to do that successfully is
already to “grasp something of the thing itself.”
Famously (at least among students of Aristotle), the categorical status of
the soul is the very first question he turns to answer at the beginning of his
“most common account (κοινότατος λόγος) of the soul,” in the opening
paragraph of de An. ii.1, 412a6-9:
Now we call one kind (γένος ἕν) of existent substantial being (οὐσία), of
which one sort is as matter, which in itself is not a this, another sort is as
shape and form, in virtue of which it is already called a this (τόδε τι), and
third that which is composed of these.
That is, he begins by identifying the category in which the soul is found,
but then notes that there are a number of senses in which something can be
in the category of substantial being. Note the precise correspondence with
the first metaphysical question of de An. i.1, 402a24-25, about “whether it is
a this and a substantial being” (τόδε τι καὶ οὐσία) – Aristotle points to the
role of substantial being qua form as the source of ‘thisness.’ He later goes
on to argue against seeing the soul as either the matter or the compound,
and concludes:
The soul must, then, be substantial being as form (οὐσίαν ὡς εἶδος) of a
natural body that has life potentially. (de An. ii.1, 412a19-21)
That is, the very first positive step in providing his own account of the soul,
in opposition to the views he has reviewed and rejected in book i, is to
answer the first metaphysical question, about the soul’s categorial status,
presented in his methodological introduction. And, by identifying it as
belonging in the category of substantial being in the sense of form, he has
affirmed that it is substance in the sense of that which makes the ensouled
being a this, a τόδε τι.
17
Soul as a self-moving number, perhaps a view held by Xenocrates, is mentioned at 404b27-30 and
critically examined in chapter 4 (408b32-409a31); and the view that it is the οὐσία of soul to move
itself is noted at 406a17, various versions of which are critically examined in chapter 3 (406a16-b25).
At i.4, he argues that the soul, properly speaking, cannot move, which leads him to point out at
408b30-32 that one implication of this is that it cannot move itself.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
But that initial account also identifies soul as the form of a body that has
life potentially – that is, there is a reference to the modal status of the body.
But then, what is the modal status of soul? Returning to the list of questions
in de An. i.1: Question 3 (402a25-26) asks whether the soul is a potency or
some sort of fulfillment (ἐντελέχειά τις). And once more, following pre-
cisely the metaphysical questions shaping the inquiry in de An. i.1, having
noted that substantial being (οὐσία) can refer to matter, form, or a
compound of the two, Aristotle continues:
. . . and matter is potentiality (δύναμις), while form is fulfillment
(ἐντελέχεια) – and that in two ways, first as knowledge is, and second as
studying is. (de An. ii.1, 412a9-11)
Just as Aristotle enriches the category of substantial being (οὐσία) in ii.1

by means of his metaphysical analysis of that category into matter, form,
and composite, so here he enriches his account of fulfillment (ἐντελέχεια)
by indicating two different ways in which a potency can be realized. In de
An. i.1, it was left open whether soul is to be understood as a potency or
fulfillment of the body, but also the precise way in which it would be a
fulfillment,18 if that is what it turns out to be. Following that framework,
the very next claim Aristotle seeks to establish in book ii is that the soul is
a first fulfillment, which, viewed in relationship to the body of the animal
is its fulfillment (indeed, what it is for, de An. ii.4, 415b15-21; PA i.5,
645b19-20) – it is that which makes it a living being and not merely a
body. At the same time, viewed in relationship to the animal’s living
activities, soul is the living thing’s fully realized capacity for those activ-
ities. Having the ability to fly, see, or solve equations is different from
actually doing those things.
Those phenomena associated with soul that might incline one to classify
it on the potential side of the potential/actual distinction are captured by
Aristotle’s distinction between two ways that a capacity can be realized. A
person who has the expertise to repair an internal combustion engine has,
in a perfectly legitimate sense of the term, fully realized his skills as an auto
mechanic (compared to me, for example); but that same person actually
repairing an engine is, in an equally legitimate sense, fully realizing his
capacity as an auto mechanic (as compared to when he is at home playing
with his children). All of this is captured in the conclusion of this first
attempt to provide a common definition of the soul, which summarizes his
answers to questions 1 and 2 of de An. i.1:
18
ἐντελέχειά τις.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
But substantial being (οὐσία) is fulfillment (ἐντελέχεια). Soul, then, will be
the fulfillment of a body of this kind. But fulfillment is spoken of in two
ways, first as knowledge and second as study. It is clear then that the soul is
fulfillment as knowledge is. (de An. ii.1, 412a21-23)
This idea of the soul as the living body’s fully realized ability to perform a
kind of activity accounts for a vast range of facts about ensouled beings,
such as that they remain fully capable of the rapid mobilization of coordin-
ated actions needed to, say, successfully elude a predator, even while they
are at rest and doing no such thing. A living thing in a state of rest is neither
merely a body capable of life nor is it fully engaged in living, and Aristotle’s
‘first’ fulfillment aims to capture this important and fundamental fact
about life as it had not been captured previously, and has rarely been since.
This opening paragraph, then, is following precisely the normative
demands of de An. i.1 about which questions ought to be answered first in
an inquiry into the nature of the soul. Questions about its categorical and
modal status are among the first mentioned, but there are more to come.
Aristotle next raises questions about whether the soul has parts and, if so,
in what sense;19 about whether or not there is a general account of soul as
there is of animal;20 whether there are many souls or just many parts of
soul, and in either case, whether there is a proper order in which they are to
be studied and whether they differ in form or in kind from each other; and
whether we should investigate the parts first or their functions, or perhaps
the objects correlated with each function. Every one of these questions is
stated in the form of a norm (σκεπτέον, 402b1; εὐλαβητέον, 402b5), and
each plays an important role in structuring the inquiry to follow in de An.
ii–iii. As noted, Aristotle refers to the account of the soul with which ii.1
opens as “the most common account” and a bit later refers to “first
fulfillment of a natural body with organs (ὀργανικοῦ)” as a statement
common to every soul, noting that he has stated “what soul is generally
19
de An. i.1, 402b1-5: “We must also investigate whether the soul has parts or is without parts, and
whether every soul is alike in form or not; and if they are not alike in form, whether they differ in
form or in kind (εἴδει διαφέρουσιν ἢ γένει). For at the moment, those speaking and inquiring about
the soul seem to investigate the human soul alone.”
20
402b5-9: “And one should take care not to overlook whether there is one account of the soul (εἷς ὁ
λόγος αὐτῆς ἐστι), just as there is of animal, or a different account for each, as there is of horse, dog,
human being, god, the universal ‘animal’ being either nothing or posterior; and similarly even if any
other common thing were predicated.” Note the similarities and differences between this question
as formulated at the beginning of the de An. and the first methodological question raised in PA i.1.
Both are questions about the level of universality at which an investigation is to focus, but here the
question is about whether there is a common account of soul or a different one for each (kind or part)
of soul; and there is also apparently a question raised about the ontological status of whatever it is
that corresponds to the universal account.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
(καθόλου, 412b10).” As he noted in raising the question about the need for
a universal account, his predecessors have focused their inquiries almost
exclusively on the human soul; the definition he comes up with, as he also
points out, applies not only to animals but to plants as well.
The remainder of de An. ii.1 introduces a number of issues surrounding
the question of whether the soul has parts by noting how different soul
capacities are intimately related to specific organs, and he draws an analogy
to the living body: “[I]f the eye were an animal, sight would be its soul, for
this is the substantial being (οὐσία) of an eye according to the account (and
the eye is matter for sight)” (412b9-10). Perception in general, he adds, is so
related to the entire perceptible body (412b23-25).
Moreover, when de An. ii.2 makes a fresh start, it is by noting that
defining accounts of the soul currently only make the fact clear but not the
cause of the fact (413a13-16),21 and after stating the somewhat obvious fact
that what differentiates animate from inanimate beings is the possession of
life, he notes that ‘being alive’ is said in many ways – a being is alive if it
possesses any one of the capacities of reason, perception, locomotion or
nutrition, growth and decay. The chapter also draws our attention to many
interesting relationships among these capacities, after which he once more
makes a clear allusion to his framework questions in i.1:
Whether each of these is a soul or part of a soul, and if a part, whether it is
such as to be separable only in account or also in place – about some of these
questions it is not difficult to get clear, but some pose a problem. (de An.
ii.2, 413b13-16)
And in particular, a few lines later we read:

About reason and the theoretical capacity nothing is yet clear, but it would
seem to be a different kind of soul (ψυχῆς γένος ἕτερον), and it alone is
capable of being separated, as the eternal is separated from the perishable.
(413b24-27; emphasis added)
It is not my purpose here to add to the many detailed interpretations of

these passages on offer. Rather, I am highlighting the way in which the
questions being raised, and even the order in which they should be taken
up, in the opening pages of de An. i.1, provide a domain-specific, erotetic
structure for the ensuing inquiry. Moreover, I want to draw attention to
21
Here I agree with Johansen 2012, 36 and Charles 2000, 168, as against Bolton 1978 that the
reference to ‘conclusion-like’ definitions is not to the common definitions in chapter 1 but to
the starting point of the discussion that follows, that a common property of ensouled things is the
possession of life.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
the similarities and differences between the opening questions of the
inquiry into the nature of the soul and those that set the agenda for the
study of animals – Aristotle sees these as very different inquiries.
Not only are the questions being raised in de An. i.1 different from those
in PA i.1; the way in which those questions are related to the actual
investigations is also different. It seems clear that the list of questions in
de An. i.1 was penned with a clear conception of where the investigation is
headed. That is, there is a much greater degree of precise coordination
between the questions presented at the outset and the design of the inquiry
to follow than is the case in PA.22 As I have stressed a number of times, each
treatise is presented in such a way as to give readers who are being
introduced to its subject matter a rich sense of how they ought to proceed
so as to maximize their chances of achieving knowledge of the domain.
While the treatises that have come down to us do not provide a glimpse
into Aristotle at work, so to speak, they are structured in such a way as to
give us insight into how he thinks an inquiry into a specific subject should
be organized.
On the other hand, while the questions and subsequent norms that
emerge are quite different, de An. i.1, like PA i.1, at a certain point makes a
gradual transition from discussing a series of unanswered questions to a
lengthy discussion of two related methodological recommendations: (i)
just as an understanding of essence can provide an understanding of
nonessential attributes (via demonstration), so a study of the latter can
be helpful in coming to know a thing’s essence (τὸ τί ἐστι γνῶναι; 402b16-
22); and (ii) in studying attributes that are common to body and soul, the
proper method to adopt is that of the Aristotelian natural scientist dis-
cussed in Chapter 5: that is, to study the soul as an aspect of a matter/form
unity.
In de An. i.1, that is, Aristotle makes an advance on the question of
whether there is a methodos for achieving knowledge of the essence – the
first, methodological question, raised at 402a15-20. But it does so in a way
specific to the study of ensouled beings. The approach he recommends is to
start by attempting to grasp, via what is available to perception, all or most
22
As one pointed example of the lack of such coordination between PA i and PA ii–iv: in PA i.1 there
are repeated discussions of a special kind of necessity, conditional or hypothetical necessity, that is in
play in investigating animals and how it is distinguished from unqualified necessity; but in PA ii–iv,
though the language of necessity is a constant feature of the explanations on offer, Aristotle never
distinguishes different kinds of necessity. Similarly, while much is made about a distinction between
material and formal natures in PA i, that distinction is rarely mentioned in his actual study of animal
parts.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
of a thing’s attributes, which will in turn allow us to speak best about the
ousia, the source of demonstration (402b22-25).23 He goes on to urge that
there should be preliminary definitions that at best give us ‘acquaintance
knowledge’ of the attributes and at the least allow us to make probable
inferences – definitions that do not provide at least this much will be stated,
he declares, “dialectically and vacuously” (διαλεκτικῶς καὶ κενῶς).24 And
in the discussion of the aporia that follows these general remarks, we see
Aristotle providing an extended example of the method he has outlined –
beginning with observations about various attributes (pathê) of the soul, he
reaches some insight as to what the soul is.25
This aporia, which returns us once more to the treatise’s prologue,
concerns whether all the pathê of the soul are shared by the soul and the
ensouled thing, or whether any of them are properties of the soul by itself;
and as virtually all commentators on this chapter note, the expression ta
pathê is employed in a somewhat slippery manner here. A case can be
made that throughout the discussion the term refers to the attributes of
soul generally, not simply to the emotions, notwithstanding that the
examples of the affections of the soul (τὰ τῆς ψυχῆς πάθη) at 403a16-18
are all emotions. If that were so, it would also refer to perceiving
(αἰσθάνεσθαι), and perhaps even thinking (νοεῖν), both listed along
with being angry (ὀργίζεσθαι), being confident (θαρρεῖν), and being
appetitive (ἐπιθυμεῖν) a few lines earlier, at a7-8.26 I’m not confident
about this, but taking pathê to refer generally to the soul’s affective
attributes, or even more generally to all of its attributes, serves the
point I want to make well. What I see Aristotle doing in this passage is
starting with familiar cases of actions and passions that are shared – are
common (κοινά) as that term is used here – and reasoning to a general
conclusion about the nature of soul, and thus about the proper way to
investigate it, and thus about the proper science to investigate it. And from
23
This recommendation has parallels in a number of other texts; for example, HA i.6, 491a9-14: “After
[we have grasped the differences and attributes of every animal] we must attempt to discover their
causes. For to pursue the study in this way is natural, once there is a body of systematic information
[ἱστορία] about each kind, since from these both the facts to be demonstrated and the principles
from which demonstration ought to be performed become apparent.” Here the focus is on finding
causes after a study of differences and attributes has been carried out, but from the standpoint of
APo. ii this is a matter of focus and emphasis, since scientific definitions identify causes. Compare
APr. i.30, 46a3-28.
24
This may be an example of “knowing that it is accidentally” (APo. ii.8, 93a24-25), in which case,
Aristotle tells us, “necessarily we have no fix on what it is, for we do not know that it is” (a25-26).
25
Though I am not sure, it appears that Philoponus may have seen the connection that I am
discussing; cf. in de An. 44.22–26 Hayduck.
26
Hicks 1907, 198.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
the beginning the question of whether reason (νοῦς) is an ἴδιον, a
distinctive property, of the soul or should be included among those
attributes common to body and soul is central to the investigation. For
example, when the above list is provided, τὸ νοεῖν (reason) is introduced
in a highly qualified way:
. . . it would seem that reason (τὸ νοεῖν) is most of all distinctive [to the soul];
but if this is a certain sort of imagination (φαντασία τις), or cannot be
without imagination, even this would not be present without body. (de An.
i.1, 403a7-10)27
To list the multiple ways in which Aristotle qualifies the possibility that
reason is among the affections common to body and soul: first, the
statement is conditional; second, we are offered two very different condi-
tional options without being given any reason for preferring one or the
other; and third, not being present without body is not quite the same as
being an affection common to body and soul.
Before proceeding, it is worth reflecting on this contrast between what is
distinctive (ἴδιον) and what is common (κοινόν). Aristotle refers to the objects
under discussion as affections of the soul, and yet one of the ontological
options under consideration throws that characterization into doubt. For if
fear, say, is shared in common by soul and the ensouled body or subject, then
we might think it is no more nor less a pathos of one than of the other. On the
other hand, were fear a feature distinctive to the soul, we might likewise think
that it will not be something ‘shared in common’ with the body. And if any
functions or affections of the soul (τι τῶν τῆς ψυχῆς ἔργων ἢ παθημάτων,
403a11) are distinctive to it, Aristotle appears ready to draw the immediate
conclusion that soul could be separate – and investigated separately.28
The direction of inference, underscored by the illustration that follows,
is from being something of the soul’s own to the possibility of soul’s
27
On νοῦς as a sort of imagination (φαντασία), cf. de An. iii.10, 433a9-12; MA 6, 700b17-20; and on
φαντασία as a necessary condition of thought, cf. de An. iii.3, 427b14-16.
28
Both Hicks and Hamlyn translate χωρίζεσθαι at 403a11 as “separated from the body.” For reasons
given in the text above, I do not think this is correct even as interpretation, and there is nothing in
the Greek corresponding to ‘from the body.’ It is true, however that the feminine pronouns and
adjectives here (αὐτὴν . . . αὐτῆς . . . χωριστή) make it clear that the implication he is at this point
drawing is that if any functions or affections are not shared with the body, the soul could be separate.
Later in the chapter, it is the separability of the functions/affections themselves that is at stake
(403b10-19). I am not sure why he appears to draw a conclusion about soul’s separability from the
possibility of the separability of one function or affection, but one possibility is this: even after
Aristotle’s own account of the soul is in place, he is willing to say, about the rational faculty, that it
seems to be a different kind of soul, one that may admit of being separated as what is everlasting is
from what is perishable (cf. ii.2 413b25-28). At this point, Aristotle may be leaving open the
possibility that a soul is separate if, for example, intellect is.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
separation. That is, if something is not a function or affection shared by
soul and its possessor, then soul may be separated in some sense – there has
as yet been no clarification of what kind of separability is under consider-
ation.29 That we are considering the possibility of a function or affection
not shared with a body is an assumption built into the protasis – the kind of
separability this entails, however, is for the time being left unexplored.
We are now ready to tackle the question of what this text has to say
about the nature of the investigation, and investigator, of the soul. Aristotle
thinks that consideration of all the pathê of the soul gives a presumption in
favor of the view that they are with a body (μετὰ σώματος, 403a16-17).
There is a subtle priority here. These affections are of the soul but ‘partici-
pate in’ or ‘are together with’ the body.30 They are accounts in matter
(λόγοι ἐν ὕλῃ),31 a phrase I will return to momentarily. This argument
moves from consideration of something that seems to be the case with all
the soul’s pathê to the general character of an inquiry into the soul itself –
that is, this is an instance of reasoning from a consideration of the attributes
of the soul to a general conclusion about its nature, what it is. Beginning
with the above characterization of these pathê, he draws a conclusion about
their definitions: these definitions will have to include a reference to a
certain change of a kind of body (or anyway a part or capacity of a body) as
well as to the efficient and final causes of that change. And from that
description of the character of their definitions, he concludes:
And on account of these considerations the study of the soul – either of all
soul or of the sort we are discussing – is at once [a study] for the natural
scientist. (403a27-28)
Thus, from a consideration of the content of the definitions of the pathê of

soul a conclusion is reached about which of the theoretical sciences is to
study soul itself – either all soul or a certain sort of soul. Aristotle remains
uncommitted regarding extending this conclusion, reached by focusing on
the pathê under discussion, to ‘all soul.’ And he remains uncommitted well
after he has given his own positive general account of the soul as substance
qua form and first fulfillment of a natural body with organs.32
29
See Miller 2012, 306–339 for a good review of the options.
30
Hicks 1907: “conjoined with”; Hamlyn 1968, Smith 1931: “involve”; Irwin and Fine 1995: “require”.
31
Reading with mss. C and E. For reasons that are opaque to me, Ross defends ἔνυλοι on the basis of
the assertion, at Metaph. Ζ.7, 1033a4, that “the bronze circle has matter in the logos.”
32
As I noted earlier, de An. ii.2 asserts that theoretical intellect seems to be a different kind of soul, one
that can be separated, “just as the eternal from the perishable” (413b24-27). Cf. Shields 2016,
100–101.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
7.4 Souls and Natural Inquiry
We are now ready to look at the concluding section of de An. i.1, which
faces directly the question of whether the norms governing natural inquiry
generally are appropriate for an inquiry into the nature of the soul. This
topic is obviously intimately connected to the question of whether all or
only some of the attributes of the soul are “common to body and soul.”
The introduction to the collection of essays known as Parva naturalia
opens, after all, by differentiating itself from de An. in the following terms:
Since a determination has been made about the soul by itself (περὶ ψυχῆς
καθ’ αὑτὴν) and about each of its capacities in accordance with its part, our
next task is to make an investigation about animals and all things having life,
which are their distinctive and which their common activities
(τίνες εἰσὶν ἴδιαι καὶ τίνες κοιναὶ πράξεις αὐτῶν). Let what has been said
about the soul be assumed, and let us speak about what is left, and first about
those things that are primary. And it is apparent that the most important
[activities], both those common to all animals and those that are distinctive,
are those common to soul and body (κοινὰ τῆς τε ψυχῆς ὄντα καὶ τοῦ
σώματος), for example perception, memory, spirit, appetite, and generally
desire, and in addition to these pleasure and pain; for these belong to
practically all animals. (Sens. 1, 436a1-11)33
Here Aristotle first clearly distinguishes the inquiry into the soul by itself
from an inquiry into animals (and plants, presumably). The distinction
between distinctive and common activities is clarified later in the passage –
some of the activities considered are shared by all, or practically all,
animals; others, such as respiration (and probably waking and sleeping)
belong to some but not all; still others (youth and old age, life and death)
belong, not just to animals, but to everything with a share in life
(436a12-17).
But he uses ‘common’ (κοινά) in a different sense here as well, and one
we are familiar with – for the activities to be considered in PN are ‘those
that are common to soul and body.’ Consequently, as Thomas Johansen
has pointed out, it is “a point of continuity between the de An. and Sens.
that affections that are common to body and soul are to be approached as
such, in the manner of the physikos” (Johansen 2006, 142). And I agree
wholeheartedly with Johansen’s conclusion that the distinction Aristotle is
making here
33
There are especially helpful discussions of this introduction and the way it distinguishes the inquiries
in PN from the de An. in Morel 2006 and Johansen 2006 in King (ed.) 2006, 121–139, 140–164.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
7.4 Souls and Natural Inquiry 191
is meant to mark off the subject-matter so as both to exclude certain topics
that were also discussed in the de An. and to include certain topics, not
touched on in de An., which we may not know whether to include in the
inquiries of a physikos. What is excluded are the affections of the soul
referred to in de An. as “peculiar” (idia) to the soul, that is, by the antonym
of “common” (koina). (Johansen 2006, 144)34
The first chapters of both of these inquiries, then, aim to specify conditions
that need to be met in order for attributes of the soul to be appropriate
subjects for investigation by natural science. That task involves consider-
ation of what it means to study something as a physikos; which in turn
requires a discussion of how the investigations of a physikos, and his objects,
are to be distinguished from other related investigations (and their objects)
with which it may be confused. Aristotle turns to that topic near the end of
de An. i.1.
Aristotle begins by distinguishing between a supposed account of anger
to be given by the physikos and another to be given by the dialektikos.35 The
first would give a statement about the matter, the second would give ‘the
form and the account.’ At this point, he briefly notes that this logos must be
in a certain sort of matter if it is to exist. Using a house as his example, he
now characterizes three distinct accounts, one that talks about a house by
reference to its sheltering function, a second that catalogues the materials
from which it is built, and a third that talks about the form in these
materials for the sake of certain functions (403b3-7). In the end, the account
appropriate for an Aristotelian natural science is the third, the account that
is ‘from both’ (i.e., an account that draws on knowledge of the form
present in the matter for the sake of an end).36
34
There are many other points in this perceptive discussion that support the argument I’m making in
this chapter.
35
403a29-b2. Initially it is surprising that, immediately after he has given a model definition of an
embodied pathos, he introduces such a distinction. But while movement, a source of movement, and a
goal of movement are mentioned in these accounts, and thus objects of such accounts will be
appropriate for natural investigation, a couple more steps are needed before Aristotle can conclude
that the natural scientist’s definition must involve essential reference to both form and matter. There
is an interesting parallel here with the structure of the Physics. The form of the definition given at
403a25-27 is what one might expect based on the discussion of Ph. i. But I argued in Chapter 5 that
Ph. ii enriches the discussion of the ‘principles of change’ in book i with the idea that natural things
have natures in two respects, as form and as matter, and that natural science is engaged in a search for
four different kinds of cause.
36
Translating is tricky. I take the ὁ δὲ at 403b5 to answer to the ὁ μὲν at b4. Thus, when Aristotle asks at
403b7 τίς οὖν ὁ φυσικὸς τούτων, the most natural way to take it is “Which of these, then, is the
natural <logos> (i.e., the logos of the natural investigator)?” Following through on that suggestion,
however, leads one to see why the majority of translators (e.g., Shields 2016) rather assume that ὁ
φυσικός is being used substantively in the same way as it was at 403a29, to refer to the natural
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
These accounts, those appropriate to natural science, refer to functions
and attributes of natural bodies and natural matter – that is, Aristotle is
arguing that the general program for the unity of natural science in Ph. ii
that I discussed in Chapter 5 (especially Sections 5.1 and 5.2) will be
appropriate for the inquiry into the soul, insofar as all the attributes of the
soul are embodied. However, throughout this discussion, the possibility of
attributes that are proper to the soul itself and are not shared with a body has
been repeatedly mentioned, a possibility that in turn opens up the possibility
of soul being separable in some manner.37 If, and apparently only if, there are
such attributes could soul be separated (403a10-12). But what exactly is
meant by ‘separation’ and ‘being separable’ is left unexplored – exactly as
it was in Ph. ii.2, where we were told that this is a topic more appropriate to
first philosophy (194b14-15). A properly ‘natural’ account needs to be con-
trasted both with accounts in the crafts, which, while they too refer to
functions and affections of bodies (403b12-14; cf. PA i.1, 639b14-21), do not
refer to attributes of natural body; and with accounts that involve separabil-
ity from body and matter. The unstated implication is most interesting:
neither member of the original contrast – the natural scientist focused purely on
matter and the dialectician focused only on form – is anywhere to be found.
Matter that is not the matter of something more interesting appears to be
ruled out as an object of study (403b9-10). An inquiry that studies enmat-
tered attributes as if they were not attributes of natural bodies (i.e., which
treats them in abstraction) will fall to the mathematician; while the investi-
gation of attributes “insofar as they have been separated” is the task of the
first philosopher (403b15-16). If these are our options, and reason (νοῦς)
turns out to be a property of the soul itself and not involved with matter, it
would seem to be something for the first philosopher to investigate.
investigator. For on the alternative reading, a natural way to continue translating is: “Is it the <logos>
about the matter, ignoring the logos, or the <logos> about the logos alone? Or is it rather the <logos>
derived from both? And which of these is <the logos> of each of the other two?” In other words, we
must suppose that the meaning of logos in this passage keeps shifting between ‘account’ and ‘formal
aspect of natural thing’. There are problems in either case. In favor of reading logos all the way
through is that it is suggested by the opposition with which the passage begins, and it makes best
sense of ὁ ἐξ ἀμφοῖν (b8-9). (To see the hoops one must jump through otherwise: the normally lean
and textually precise Irwin and Fine translation translates that phrase, “Or is the <real> student of
nature more properly the one who mentions both form and matter?”. They are aware of the problem
with their choice; when they translate ὁ δὲ at 403b5 as ‘someone else,’ a footnote reads: “Or perhaps
‘another account’” (Irwin and Fine 1995, 172 n. 15). I lean toward supposing the definite articles pick
up on logos throughout. I can even imagine Aristotle smiling at the fact that there will be logoi that
ignore the logos, logoi that are of the logos alone, and logoi that derive from each of these other logoi.
37
And it must be separable in a way stronger than in account, since the forms of natural things are “not
being separable, other than in account” (Ph. ii.1, 193b4-5; cf. ii.2, 194b12-13).
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
7.5 PA i.1 Meets de An. i.1 193
However, all of this is stated as a conditional. All that is positively affirmed in
the concluding sentence is:
We said that the affections of the soul (τὰ πάθη τῆς ψυχῆς), insofar as they
are present as anger and fear are and not as line and plane are, are in this way
inseparable from the natural matter of animals (ἀχώριστα τῆς φυσικῆς ὕλης
τῶν ζῴων). (403b17-19)38
7.5 PA i.1 Meets de An. i.1

Though the first chapter of de An. certainly lays out questions the answers
to which shape (and are shaped by) the nature of the inquiry in book ii, it
also works toward a resolution of the question of what sort of an inquiry
the study of the soul is. Insofar as they are akin to anger and fear, the
affections of the soul are inseparable from their natural matter. Read in the
light of his general program for the study of nature, Aristotle has already
concluded that, with the possible exception of reason, the soul is to be
studied in a way contemplated neither by Plato and his followers nor by
traditional natural philosophy (presumably exemplars of the two kinds of
accounts referred to at 403a29-b2 and then rejected in favor of an inte-
grated, ‘hylomorphic’ account). Broadly speaking, it is to be studied
according to the program for a natural methodos of the sort discussed in
Chapter 5, with the important caveat that reason and activities associated
with it may well require a distinctive mode of inquiry. Nevertheless, an
inquiry focused on the soul is unlike other natural inquiries in a variety of
important respects, and the questions that structure the inquiry, and the
norms of inquiry that flow from Aristotle’s answers to those questions, are
distinctive to it.
The parallels between the closing sections of Ph. ii.2 and the de An. i.1
mentioned in Section 7.4 are important. Both passages take seriously the
question of whether the formal and material components of a composite
are sufficiently unified to be studied by a single science or not; and both
discussions revolve around questions about the different types of separabil-
ity characteristic of the objects to be investigated by the natural scientist,
the mathematician, and the first philosopher (cf. 403b9-19). Ph. ii.2 aims
to insure a place for a distinctive, and unified, theoretical science of nature,
reducible neither to mathematics nor to first philosophy (let alone to an
38
The text, unsurprisingly, is corrupt on various counts; I’m following the text printed by Hicks rather
than by Ross, but not with any strong conviction.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
investigation of matter by itself); and de An. i.1 aims to specify what needs
to be true about ‘parts’ or ‘attributes’ of soul for them to be the subject of
such a science.
There is, then, a rich background to the passage in PA i.1 with which we
began. Much ink has been spilled in recent years39 over the so-called
‘correlatives argument’ in PA i.1, a conditional argument that concludes
that if reason were to be studied by the natural philosopher, there would be
no other philosophy – it would have all theoretical objects as its domain.
This is not a conclusion Aristotle accepts, and it is clear from the passage
that follows that he rejects it by denying the antecedent. What makes this
passage so important, as the Renaissance commentators realized, is that it
states a position on the question about whether reason is to be studied by
the science of nature without any of the hesitation and qualification that
one finds elsewhere in Aristotle.
However, it is not the case that all soul is an origin of motion, nor all its
parts; rather, of growth the origin is the part which is present even in plants,
of alteration the perceptive part, and of locomotion some other part, and
not the rational (τὸ νοητικόν); for locomotion is present in other animals
too, but thought (διάνοια) in none. So then, it is clear that natural science
should not speak of all soul; for not all of the soul is a nature, but some part
of it, one part or even more. Further, none of the abstract objects (τῶν ἐξ
ἀφαιρέσεως) can be objects of natural study, since nature does everything for
the sake of something. (PA i.1, 641b4-12)
Commentators have tried to limit the scope of this claim in various ways.
In “On Aristotle’s Conception of the Soul,” Michael Frede, though with-
out explicitly citing this passage, surely has it in mind when he says:
I think it is with reference to this kind of intellect [active intellect which is
immaterial and not dependent on the body in any way] that Aristotle
sometimes says that not every kind of soul falls within the province of the
study of the natural scientist. (Frede 1992, 105)
Sarah Broadie, however, and I think correctly, notes that when one reads
through this passage in its context, it is hard to imagine that Aristotle has in
mind “an active intellect [that] is not a human intellect, that is not an
39
And of course, not only in recent years. I’ve already had reason to note Philoponus’ (mis)use of the
‘correlatives’ argument, and this passage was of central concern to the Renaissance Italian commen-
tators on PA i since both Averroists and their critics saw its relevance to their debate over the unity
and separability of the intellect, on which see Perfetti 1999, 297–316; Mahoney 1982, 264–282. See
Broadie 1998, 163–176, and Caston 1998, 177–192; cf. Charlton 1987, 408–423; Lennox 1999, 1–16;
and Lennox 2002, 139–145.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
integral part of the human soul,” as Frede suggests. Frede goes on, however,
to face the fact that even when considering the human intellect Aristotle
seems to be wavering on the issue of whether there are attributes of the soul
that are peculiar to it. He points to the fact that “the exercise of the intellect
. . . unlike the exercise of the other so-called mental faculties, does not
involve the use of a bodily organ” (1992, 105; presumably a reference to
de An. iii. 4, 429a25-27). Thinking of this sort may be a second-order
processing of the results of perception and imagination, directly dependent
on those results, and thus (and in this his view is akin to Broadie’s) only
indirectly dependent on their bodily organs.40
I want to focus on the obvious – Aristotle’s emphasis, consistent with
every passage in which he discusses what distinguishes the science of nature
from other theoretical sciences, is on nature as a source (or principle) of
motion in the changing thing itself. His restriction here on which parts of
soul can be investigated by the science of nature is based primarily on an
answer to the question, “Are all of the soul’s capacities sources or principles
of natural motion?” He considers, and rejects the idea that the rational
capacity is a source of locomotion. The grounds of this denial initially seem
weak, namely that many locomotive animals lack the rational capacity. But
Aristotle may well want to restrict the natural investigation of animal
locomotion to features that are common to all locomotive animals.41
Charlton cites de An. iii.10, 433a9ff. against Aristotle’s conclusion in PA
i.1, but that passage restricts the claim that thought may be a source of
locomotion to practical thought, and explicitly denies that theoretical
thought is such a source. Moreover, even that restricted claim about
practical thought is in the end called into question.42 Here is the key
passage, with some important context.
40
See Frede 1992, 106; Broadie 1998, 163–166. A full consideration of this topic would need to include
careful consideration of GA ii.3, 736b5-30, a passage that puzzles over how best to understand the
temporal emergence of the different capacities of the soul and concludes with the passage I have
chosen as the epigraph for this chapter, referring to reason entering ‘from without’. Burnyeat asserts
that this may “perfectly well refer . . . to a second potential νοῦς acquired through the agency of a
teacher” (2003, 70 n. 111), a suggestion already hinted at in Charlton 1987, 411–412.
41
On which, see the opening lines of MA 1, 698a1-7.
42
See Charlton 1987, 411; Lennox 1999, 4–5. Mary Louise Gill in her comment on an early draft of
Lennox 2009 challenged me to look at this material, for which I thank her. She also argued that there
was a tension between the two passages, and offered a most ingenious reading of the PA i.1 passage
on the hypothesis that it was an aporetic consideration of a view Aristotle in the end rejects. Part of
her argument rested on Aristotle’s remarks about order being more apparent in the cosmos than in
animals, but the sentiment in that passage is exactly mirrored at De caelo ii.8, 290a29-35. In the end,
since I do not see a significant tension between the PA i.1 text and de An. iii, I do not feel the need to
press a different interpretation.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
At least it appears that the moving source is one of these two things, desire or
reason, if one were to take imagination as a sort of reasoning (εἴ τις τὴν
φαντασίαν τιθείη ὡς νόησίν τινα); for many people follow imagined things
contrary to knowledge, and in other animals there is neither reason nor
calculation (οὐ νόησις οὐδὲ λογισμὸς), but there is imagination. Therefore,
both of these things, that is, both reason and desire, are capable of moving
things, but by reason here I mean the one calculating for the sake of
something, that is, practical reason (νοῦς δὲ ὁ ἕνεκά του λογιζόμενος καὶ ὁ
πρακτικός); and it differs from theoretical reason in virtue of its end. And
desire is always for the sake of something; for the object of desire is the
starting point of practical reason (ἀρχὴ τοῦ πρακτικοῦ νοῦ), while its end is
the beginning of action. (de An. iii.10, 433a8-17)
In the previous chapter he had stated the idea more baldly: “neither is the
mover the faculty of calculation (τὸ λογιστικόν) nor what is called reason;
for the theoretical faculty thinks nothing practical nor speaks about what is
to be pursued and avoided, while [animate] motion is always in pursuit or
avoidance of something” (432b26-29); while later in chapter 10, he reiter-
ates that “it is always the object of desire that causes motion” and thus “it is
this sort of capacity, that which is called desire, that moves” (433a30-31).
And when he finally summarizes these two chapters on the faculties of soul
involved in locomotion, the summary leaves out mention of reason
altogether:
Since [animate] locomotion involves three things, first the mover, second
that by which it moves, and third that which is moved, while the mover is
twofold, one of which is unmoved while the other is both a mover and is
moved – the unmoved mover is the practical good, that which is moving
and moved is the faculty of desire (for what is moved is moved insofar as it
desires, and desire qua activity is a sort of motion), and that which is moved
is the animal. (433b13-18; emphasis added)43
Aristotle’s position, then, is tolerably clear. The mandate of the natural
scientist is to explain change by searching for its sources and causes. In a
highly qualified way, such explanations may involve consideration of
practical reasoning, but even in that case the search for the sources and
causes of locomotion will point the natural scientist toward a study of the
objects of desire and imagination and away from a study of reason.
These are surely the thoughts in the background of our passage in PA i.1.
Natures are inherent sources of motion (κίνησις). Properly speaking, apart
from substantial generation, there are only three categories of natural
43
Compare MA 6, 701a4-6: “For the animal moves and progresses by desire or choice, when some
alteration has taken place in accordance with perception or imagination.”
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
motion: growth (quantitative change), alteration (qualitative change), and
locomotion (change in place). He mentions all three here, pointing out
that the part of soul that is a source of growth is shared even by plants,
while the perceptive part, associated with alteration, and ‘some other part’
(τὸ ὀρεκτικόν?), associated with locomotion, are shared by all animals.44
He denies, however, that the part of the soul associated with locomotion is
the rational part, and the passages from the later chapters of De anima
argue for this very conclusion.
As a look at the opening chapters of EN make clear, Aristotle sees goal-
directed activity as a central aspect of human life, and he is equally clear
that such a life is one lived in accordance with reason. But unless those
activities are properly conceived of as changes in one of the above three
categories, they are going to be the subject of another science, not the
science of nature.45 In those passages in which he taxonomizes knowledge,
the study of human action is characterized as a practical science, not, as
natural science is, a theoretical science. Investigating theoretical reason,
however, seems to be the task neither of practical science nor of natural
science – where does its investigation belong?
This question brings us to the last sentence I quoted from PA i.1.
Mathematics and first philosophy study their respective objects in separ-
ation from change. But, while the natural scientist in his inquiries about
nature will think abstractly, what he will be thinking about will be particu-
lar, changing things and the sources and causes of those changes – these are
the objects of natural inquiry. As Aristotle says a few lines later, we say that
one thing is for the sake of another whenever there is apparent a goal
toward which a change proceeds when nothing prevents it (641b24-26).
Reasoning in mathematics and first philosophy begins only when the kinds
of change investigated by natural science are put aside. An inquiry into the
essence of abstract thought, such as the investigation reported in de An.
iii.3–7, is more akin to these two theoretical sciences. At any rate, it is not
studying nature as Aristotle understands it.
PA i.1, then, does clearly deny that thought and its objects are, as such,
suitable subject matter for natural science, and at least in part on grounds
44
The passage thus generally comports with the opening of de An. ii.3, which lists θρεπτικόν,
αἰσθητικόν, ὀρεκτικόν, κινητικὸν κατὰ τόπον, and διανοητικόν as the capacities of soul to be
discussed (414a31-32), but considers the capacity for movement in place in relation to those of desire
and perception.
45
This is, of course, a central theme of Burnyeat 2002. He argues that in the end the precise wording of
the qualification about perception – regarding whether it is an alteration or not – is driven by
Aristotle’s background thesis that the study of the soul is the province of the natural scientist. My
concern is that Aristotle seems not to be driven to extend this thesis beyond perception.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
of the ‘separability’ of thought and its objects. But the crucial separability
in both cases is separability from natural change. Theoretical reason and its
objects are not proper objects for natural science, since they are neither
engaged in, nor the sources of, the sorts of change that are the proper
subjects of the natural scientist. And even when Aristotle considers prac-
tical reason in relationship to human locomotion in de An. iii.8–10, he
places much more emphasis on desire and its object as its sources of
locomotion. Whether the intellect is separate from the body in any more
robust sense is not a question that even arises in the context of PA i.1. And
at any rate the argument against reason being a proper subject for natural
science does not turn on its ontological separability from the body, but
rather on whether or not it is source or cause of natural change.
7.6 Conclusion
In the end, de An. i.1 leaves the issue of whether natural science studies all
soul or not undecided; but it is quite clear that if there is a capacity of the
soul that is not to be studied by natural science, it is reason (νοῦς). That
question, like many others raised in this preliminary, problem-setting
chapter, is not there answered. But some advances are made. First, we
learn something about the nature of soul from a study of its pathê, namely
that most, if not all, of its attributes are associated with the body and are
not attributes distinctive to the soul, meaning that neither the approach of
the traditional dialectician nor that of the traditional (materialist) natural
scientist alone can study the soul. There needs to be a way of inquiring into
the soul modeled on the form of inquiry defended in Ph. ii, one that
studies teleologically unified composites of matter and form. Whether reason
can be studied in this way remains, however, an open question in book ii,
well after Aristotle has presented his own theoretical account of the soul as
the first fulfillment of a natural, instrumental body; and however mysteri-
ous the discussion of nous in book iii.3–6 is, it gives us every reason to
think he found the question of the proper way to inquire into its essence a
profoundly difficult one.
PA i.1, on the other hand, is straightforward on the issue of whether such
an inquiry should be considered a natural one. Of the various sorts of
change that are the proper subject of natural science, the capacity for
thought is not a proper source or principle for any of them. Neither it
nor its activities are a proper subject of natural investigation. It is not a
nature (i.e., not an inherent source or principle of natural motion). As far as
one can tell from that passage, however, its activities could well be
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
7.6 Conclusion 199
dependent, in various ways, on other activities carried out by bodily
structures. Its ontological status is not part of the discussion.
The backdrop, the context that explains what is going on in both of
these passages, I have argued, is Ph. ii.2. It lays down the conditions for an
entity to be investigated by a unified science of nature, and only certain
parts of the soul, according to PA i.1, meet those conditions. In the end,
while Aristotle need not be a ‘spiritualist’ with respect to human reason, he
is not a straightforward naturalist either. As with the study of consciousness
today, for Aristotle the question of the proper norms for an inquiry into
reason was perhaps the most difficult question he had to face. The seeds for
an autonomous study of reason and its object are sown in these texts. But
unless Aristotle is prepared to add to his list of theoretical sciences, it seems
the investigation of the soul is to be shared by the natural scientist, the
moral philosopher, and the metaphysician. And that has implications
regarding which norms will govern such an investigation – as we can see
clearly in EN i.13. After noting that, since happiness is an activity of the
soul, political inquiry will require investigation of the soul, Aristotle adds
an important qualification:
And while it is for the political scientist to study the soul, he should do so as
a means to inquiring about virtue and to the extent required for the subject
of inquiry. For a study of greater precision would perhaps be more work
than is necessary for the subject set before us. Some things said about the
soul in our public lectures are adequate for our purposes, for example that
one part of the soul is without reason and one part possesses it. Whether
these two elements are separate, like the parts of the body or any other
divisible thing, or whether they are logically separable though in reality
indivisible, as convex and concave are in the circumference of a circle, is
irrelevant for our present purposes. (1102a23-32)
It might be worthwhile for the reader to now return to Section 7.4 with this
methodological passage from EN fresh in mind, before moving on to
Chapter 8.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
chapter 8
The Order of Inquiry i

Right and Left in De caelo and De incessu animalium
We begin the investigation [of the reason why] with presuppositions

we are accustomed to use frequently in natural inquiry, assuming that
this is the way things are in all of nature’s works.
(IA.2, 704b12-14)1
Chapter Summary. In this chapter and the next I explore the ways in
which Aristotle uses the results of one natural inquiry as starting points for
inquiry in another area. This is a practice in which Aristotle routinely
indulges, and yet there is, at least prima facie, a problem with him doing so,
given his views about how we arrive at first principles and the propriety of
those principles to specific domains. In APo. i.7–13, he allows that geomet-
ric and arithmetic premises can be used in a range of fields of inquiry he
refers to as ‘subalternate’ branches of mathematics, which he elsewhere
describes as ‘the more natural of the mathematical sciences’ (i.e., optics,
astronomy, mechanics, and harmonics). However, whether and, if so, how
this practice might apply within the science of nature is never explicitly
addressed. In this chapter, I address this question by exploring the depend-
ence of Aristotle’s discussion of the application of the concepts ‘right’ and
‘left’ to the motions of the heavens on his discussion of directional dimen-
sions in De incessu animalium.
8.1 Introduction
In the previous two chapters we have seen that, though the inquiries into
animals and into the soul are, in whole or in part, natural inquiries
structured according to the erotetic framework of APo. ii and the causal
framework of Ph. i–ii, they were each governed by norms specific to their
1
Ἀρχὴ δὲ τῆς σκέψεως ὑποθεμένοις οἷς εἰώθαμεν χρῆσθαι πολλάκις πρὸς τὴν μέθοδον τὴν φυσικήν,
λαβόντες τὰ τοῦτον ἔχοντα τὸν τρόπον ἐν πᾶσι τοῖς τῆς φύσεως ἔργοις.
200
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
8.2 De caelo as a Natural Inquiry 201
own domains, norms that were self-consciously articulated as part of the
inquiry itself. In this chapter, I turn to a very common practice in
Aristotle’s scientific works, and a practice he discusses in language quite
reminiscent of the Posterior Analytics: the practice of using results of one
natural inquiry in the process of engaging in another. This has important
implications for his account of inquiry, since it suggests that there is an
order of inquiry that is not merely a matter of pedagogy. Over the next two
chapters, I will look at two extended examples of this practice in Aristotle’s
scientific work. In this chapter, we will focus on his use of principles
articulated in his study of animal locomotion (De incessu animalium) in
his discussion of ‘cosmic directionality’ in Cael. ii.
8.2 De caelo as a Natural Inquiry

Aristotle’s De caelo begins2 by declaring that the science of nature appears
most of all to be concerned with bodies, magnitudes, their affections, and
their changes, but it also deals with the first principles3 of such substantial
beings. He rests this claim on ontological grounds – things constituted by
nature either are bodies and magnitudes, have body and magnitude, or are
the first principles of things having body and magnitude (Cael. i.1, 268a1-6).
A dimension of magnitude is defined in terms of continuous divisibility,
and a natural body is said to be complete insofar as it has all three of the
dimensions of magnitude (πάσας τὰς διαστάσεις, 268b6-7) – that is, it is
divisible in length, breadth, and depth.4 Here, then, διάστασις is
2
Though it is not the primary focus of this chapter, the opening lines of the De caelo raise questions
that are directly relevant to the topic of this book. There is no connecting particle in the first
sentence, and the way in which the science of nature is identified is not based on, nor is there any
reference to, anything said in the Physics about the objects of natural investigation. Here ‘body’ is
simply treated as magnitude in three dimensions, and the principles eventually identified are
principles of magnitude. And while continuity is mentioned at the beginning of Physics iii, it is
mentioned simply as a property of motion that points to the need to discuss other things that are said
to be continuous (infinity, time, space). For an excellent discussion of the difficulties posed by De
caelo’s introduction, see Falcon 2016, 423–436, esp. 428–432. On its status as natural science, see
Leunissen 2010, 156–158.
3
As in previous chapters, where ἀρχή clearly refers to the starting-points of a science, I translate this
term as ‘principle’ or ‘starting point’ (keeping in mind that Aristotle distinguishes starting points of
inquiry from principles of demonstration). However, where it appears to refer to the source or origin of
a change of any kind, I translate ‘source’ or ‘origin,’ in this case keeping in mind that a nature is
defined as an inherent archê of motion, and what is often referred to as his ‘efficient’ cause is described
as ‘whence comes the archê of motion’ (but, as noted in ch. 6, he often insists that the telos or that for
the sake of which is also an archê). As the entries in Metaphysics Δ.1 (archê), 2 (aition), and 4 (physis)
make abundantly clear, Aristotle is fully aware of this problem.
4
Timaeus argues that the Demiurge chose the form of the sphere for the cosmos on grounds that it is
most complete and like itself (Ti. 33b-c), and introduces the three pairs of directional dimensions at
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
202 The Order of Inquiry i
appropriately rendered ‘dimension’ and refers to precisely what we refer to
by that English word.5
What is interesting about this way of specifying the science of nature is
its stress on magnitude and its dimensions.6 In other treatises that discuss the
subject matter of the science of nature, the stress is typically on two features
peculiar to the objects of the science of nature – that they are material/
formal composites and that they undergo one or more of four kinds of
change in virtue of their material and formal natures (Ph. ii.1–2; iii.1;
Metaph. Ε.1; Cael. iii.1). Nevertheless, at the beginning of chapter 2,
Aristotle does make it clear that this is a natural inquiry:
Concerning the nature of the entire cosmos, whether it is unlimited in
magnitude or is limited in its total bulk should be investigated later; but we
must now discuss its parts in accordance with form, making our beginning
here. That is, we state that all natural bodies and magnitudes are, in virtue of
themselves, capable of motion in place; for we say nature to be a source of
motion in them [i.e., in natural bodies]. (Cael. i.2, 268b11-16)
The stress is still on magnitude, both in specifying the topic that will be
taken up in chapters 5–7 (i.e., whether the cosmos is limited or unlimited),
and in specifying that we begin by recalling that all natural beings have
their own source of motion. For while that reminder is in one respect
imported from the Physics, there he refers to the three kinds of motion
natural beings are capable of; while here, besides twice stressing magnitude
in the above passage, he specifically limits the focus to bodies capable of
motion in place. Aristotle is carefully delimiting the scope of inquiry, while at
the same time stressing that it is a natural inquiry.
43b-c, but in a manner essentially unconnected to Aristotle’s. Excellent discussions of the connection
between the two works may be found in Falcon 2005, 31–36; and Johansen 2009, especially 18–24.
5
This passage also includes (i) an odd endorsement of the Pythagorean dictum that ‘the all and
everything are defined by threes’; (ii) a claim that we take the number three from nature and make use
of it in worship of the gods; and (iii) a claim that the completeness of three is shown by the fact that
the first time we say ‘all’ (rather than ‘both’) is when we refer to three items. I will take up later the
question of why Aristotle stretches in these various ways for support of the idea that objects with these
three dimensions of magnitude are complete objects. For now, however, it is enough to note that it is
part of a set of wider epistemological questions raised by the method and style of the De caelo, on
which see Bolton 2009 and Falcon and Leunissen 2015, 217–240.
6
Interestingly, Cat. 6 distinguishes discrete and continuous quantity (πόσον) but does not mention
magnitude (μέγεθος) (4b20-22, 5a1-6); while in Metaph. Δ.13, Aristotle immediately divides quantity
into plurality (πλῆθος) and magnitude (μέγεθος) based on whether the quantity is countable or
measurable, and whether it is divisible into discontinuous or continuous parts (1020a7-11). In line
with Cael. i.1, he goes on: “Of magnitudes those continuous in one dimension are lengths; in two,
breadths; in three, depths. Of these, limited plurality is a number, [limited] length a line, breadth
a surface, depth a body” (1020a11-14).
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
What he means by saying that the immediate topic for discussion is the
parts of the entire cosmos in accordance with form (κατ’ εἶδος) becomes
clear immediately – for he first produces a division of natural bodies into
simple and compound (i.e., two forms of natural body) and then divides the
simple bodies into two that move naturally toward and two that move
naturally away from the center of the cosmos. He next raises the question
whether, given that there are heavenly bodies apparently moving with
a circular motion, they are constituted of one of these four or whether
there needs to be a first body prior to these that moves by nature with
a circular motion. That is, the remainder of the discussion up to chapter 5 is
governed by a division of natural body into its various forms based on how
they move by nature.
Based on premising that all motions are either circular or straight or
a mixture of the two; that simple motion is either straight or circular, and
either toward, away from, or around the center; and that simple bodies
have a principle of natural motion (268b20-269a2), he reaches the follow-
ing conditional conclusion:
If there is simple motion (ἁπλῆ κίνησις), and motion in a circle is simple,
and simple motion is the motion of a simple body, then it is necessary that
there be a simple body which moves by nature in a circle in virtue of its own
nature (κατὰ τὴν ἑαυτοῦ φύσιν). (Cael. i.2, 269a2-7)
The stress, here in Cael., on dimensions of magnitude as central to the
science of nature is another instance of the domain specificity of scientific
inquiry for Aristotle. The dimensions of magnitude play a number of
central roles in the study of the cosmos – in differentiating the motions
of the spheres from those of the four elements; in using results from
astronomy in support of cosmological conclusions, and for the inquiry
on which we are going to focus, concerning whether directional concepts
such as right and left have application to cosmological motion. For all these
reasons, there is a heavy emphasis in the opening chapters of De caelo on
bodily dimensions.
8.2.1 Correcting the Pythagoreans

At the beginning of Cael. ii, not long after reviewing weaknesses in the
arguments of those who deny that the heaven is eternal,7 Aristotle
7
Cael. i.1–4 argues that there is a “primary body” that is ungenerable and indestructible and whose
natural motion is circular, out of which the heavenly bodies are constituted. (For an illuminating
discussion of how unorthodox this move was and Aristotle’s reasons for taking it, see Falcon 2005.) In
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
considers whether the right and the left are among the principles to be
assigned to ‘the body of the whole cosmos,’ an idea defended, we are there
told, by the Pythagoreans.
Since there are some, such as those called Pythagoreans (for this is one of
their statements), who claim there to be a certain right and left to the
heaven, one should investigate whether this is so in the way they claim, or
in some other way – if indeed one ought to apply these principles to the
body of the whole cosmos. (Cael. ii.2, 284b6-10)
Aristotle is raising two distinct, normative questions here. The more
fundamental one is whether it is appropriate to apply these principles to
the heavens at all, and in a moment we will see that this is a well-motivated
question. Assuming the answer to this question is that it is legitimate to
speak of a cosmic right and left, the second question is whether it is
legitimate to do so in the way the Pythagoreans do, or in some other
way. Aristotle responds immediately to both questions.
For first off, if right and left are present in something, then one should posit,
prior yet to these, the prior principles to be present in it. Now
a determination was made regarding these matters in our studies of the
movements of animals, on account of these principles being proper to the
nature of animals; for at least in animals it is readily apparent that all such
parts (I mean, for example, right and left) are present in some animals, some
of them are present in some animals, while only above and below are present
in plants. (284b12-18)
Aristotle claims that these principles are proper, or appropriate –
oikeia – to the nature of animals, and for that reason a determination
about them was made in the discussions of animal movements.8 One
might suppose this to be a reference to De motu animalium (MA); but
it is clearly to De incessu animalium (IA), and in fact to chapters 2–6 of
that work.9
chapters 5–7 Aristotle mounts an extended argument against the idea that the cosmos is infinite; in
chapters 8–9, against there being more than one cosmos; and in chapters 10–12, against the cosmos
being generable or destructible.
8
ἐν τοῖς περὶ τὰς τῶν ζῴων κινήσεις. The standard formula Aristotle uses when cross-referencing
another investigation (i.e., ἐν + τοῖς + περὶ + the subject of investigation) is used here. The phrase
used for the subject of investigation, τὰς τῶν ζῴων κινήσεις, is the same as that used in the first
sentence of MA to refer to a prior inquiry reported on in IA.
9
Cf. Simplicius, in Cael. 384.9-20; Leggatt 1995, 224; Stocks 1922, 284b14 n.1; Guthrie 1939, 138. While
all these commentators and editors correctly identify this as a reference to IA, none note the
identification, in IA 2, of these directional principles as among the archai frequently posited as
starting points of natural investigation, which, from the viewpoint of understanding the dependence
of Cael. on the discussion in IA, is critical.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
IA 2 provides a methodological discussion of the starting points of
natural inquiry as important as any in the corpus and provides the
framework for Aristotle’s defense of the application of these ‘directional’
principles to the heavenly spheres in De caelo. We can see that framework
at work in one of his criticisms of the Pythagoreans in Cael. ii.2.
It is important for the theme on which I am focused in this chapter to
highlight the fact that Aristotle is raising questions both about the
propriety of applying the concepts of ‘right’ and ‘left’ to the heaven and,
supposing it is appropriate, about the proper method for doing so.
Aristotle initially remains noncommittal on the question of propriety,
but he insists that, regardless of how one answers that question, the
Pythagoreans have made a methodological mistake. Among these ‘direc-
tional’ principles some are prior to others, and in particular, some are
prior to ‘right’ and ‘left.’ Thus, if the Pythagoreans are committed to
attributing right and left to the heaven, they ought to justify the applica-
tion of those prior principles first.
And if it is necessary to apply any of these to the heaven, it would be
reasonable (εὔλογον) that the one present first in animals be present in
the heaven. For there are three, and each is a sort of principle. The three
I am talking about are the above and below, the front and its opposite,
and the right and the left. For it is reasonable (εὔλογον) that all these
dimensions (διαστάσεις) are present in complete bodies. And ‘above’ is
a principle of length, ‘right’ a principle of width, and ‘front’ of depth.
(284b18-25)
There are two clear links in this passage to the framework established in
Cael. i.1: first, Aristotle insists that if we are investigating complete bodies, it
is reasonable to expect that all three dimensions will be present;10 second,
recalling the stipulation of the treatise’s very first sentence, he notes that the
three directional pairs are the principles of the three dimensions of magni-
tude, which according to Cael. i.1 define complete body. Both of these
links, however, are rendered deeply puzzling on account of the same term –
διαστάσεις – used earlier to refer to the standard three dimensions of
10
Recall Cael. i.1, 268a22-24: “of magnitudes body alone [vs. line or plane] would be complete; for
body alone is defined by the three [dimensions], and this is all [the dimensions there are].” An
anonymous reader for the press has suggested translating διάστασις as ‘interval’ as a way of lessening
the tensions I discuss in this section. I have decided to follow the standard English translation
(‘dimension’ is used by both Guthrie and Stocks) in part because I think ‘interval’ simply changes
the tension without eliminating it, and because it does seem that dimensions are what are under
discussion. Nevertheless, it is important to remember that it is, along with διάστημα, a term used in
harmonics (cf. Ti. 36a6-7) for musical intervals, and less technically for gaps or separations.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
bodily magnitude, now being used to refer to the three pairs of directional
orientation. Moreover, as puzzling as this duality of reference might be, it is
clearly intentional. For in the last sentence in the above-quoted passage,
one half of each pair of directional orientations is said to be the principle
(ἀρχή) of one of the three bodily dimensions – ‘above’ of length, ‘right’ of
width, and ‘front’ of depth.11 Now if the IA discussion was not explicitly
referred to just prior to this claim, it might be just possible to gloss over this
apparent ambiguity in the use of the term διάστασις by supposing that
‘these dimensions’ here refers, admittedly in a sloppy and confusing
manner, forward to the dimensions of magnitude mentioned in the next
sentence. However, IA 2 introduces, as the second of three fundamental
‘presuppositions’ of natural inquiry, the following:
Next it is necessary to grasp how many and what sorts of dimensions of
magnitude (διαστάσεις τοῦ μεγέθους) belong to what kinds of things. For
while the dimensions (διαστάσεις) are six, there are three pairs: first, above
and below; second, front and back; and third, right and left. (704b18-22; cf.
705a26-29)
Here Aristotle refers to the three pairs of directional dimensions immedi-
ately after saying that it is necessary to grasp the number and kinds of
dimensions of magnitude; indeed, he does so in such a way as to encourage
us to think that the reference of διαστάσεις has not changed. This admit-
tedly is puzzling: but the discussion in IA 2 is explicitly about archai of
natural inquiry – and as we have seen, Aristotle considers the directional
pairs to be the archai of the dimensions of magnitude.
There is no doubt, then, that Aristotle refers to both the three dimen-
sions of bodily magnitude (width, breadth, depth) and the three pairs of
directional orientation (right/left, front/back, above/below) with the same
term, and is explicitly connecting these two modes of reference in Cael.
ii.2. Moreover, since IA 2 introduces them as among presuppositions that
often serve as archai in the investigation of nature, these two discussions are
clearly closely related to each other.
Before turning to the discussion in IA, however, there is another
puzzling feature of the passage quoted above from Cael. ii.2 (284b18-25)
to be noted: Initially, three pairs of directional dimensions are identified
and each (pair) is said to be a principle; but in the last sentence, above is
referred to as a principle of length, right as a principle of width, and front as
11
Immanuel Kant takes the opposite stance and derives the directional distinctions from the spatial
dimensions; cf. Kant 1768/1992, 2:375–383.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
a principle of depth. Are the three pairs of directions the principles of
dimensionality, or only one half of each pair?
In confronting this puzzle, it helps to remember that for Aristotle, pairs
of contraries are seldom ‘equals.’ Cold is the absence of heat (GA ii.6,
743a36, Mete. iv.8, 384b26-27); the female is female due to an incapacity,
the male due to a capacity (GA i.20, 728a18-21), and being cultured is both
the opposite of, and better than, lacking culture (Ph. i.7, 189b34-190a13).
Indeed, in Ph. i.7 one contrary involved in a change may be referred to as
privation, the other as form.12 This inegalitarianism holds for directional
pairs as well, as Aristotle makes perfectly clear:
For the principle (ἀρχὴ) is honorable, and above is more honorable
(τιμιώτερον) than below, front than back, and right than left. (IA 5,
706b12-13)
Thus, while the three directional pairs are ‘in a way principles’ with respect
to length, breadth, and depth, one member of each pair is more honorable
than the other, and thus ‘more a principle’ than its opposite.
Having resolved these puzzles, let us return to Aristotle’s consideration
of the Pythagorean idea that there is a cosmic right and left. There are two
appeals to what is ‘reasonable’ (εὔλογον) in that discussion. The first makes
a direct appeal to the idea that these directional concepts originate in
a zoological context: if one of these dimensional pairs has priority over
the others in the case of animals, it is reasonable to suppose that it does in
the case of the heaven – one more reason to look at IA. The second appeal to
what is ‘reasonable’ relies on the argument in Cael. that the heaven, in the
sense of the whole encompassed by the outermost sphere,13 is complete
because it is composed of three-dimensional bodies as its parts.14 This
appeal not only provides further evidence that these two uses of
διάστασις are intimately related. It also provides evidence of the use of
εὔλογον in natural science to refer to support for a claim derived from its
agreement with principles established elsewhere.15
12
Although on at least one occasion, he allows that “the privation is in a way form” (Ph. ii.1,
193b19-20).
13
The three meanings of ‘heaven’ (οὐρανός) distinguished in Cael. i.9 are (i) the substantial being of
the outermost circumference of everything; (ii) the body continuous with the outermost circumfer-
ence that includes the sun, moon, and planets; (iii) the entire body surrounded by the outermost
circumference (278b11-22).
14
Cael. i.1, 268a19-b10.
15
For recent discussions of Aristotle’s quite common appeal to what is εὔλογος, see Falcon and
Leunissen 2015, 217–240 in Ebrey 2015; Bolton 2009, 51–82 in Bowen and Wildberg 2009; and
Karbowski 2014, 25–38. A classic and still valuable discussion is LeBlond 1938. Both occurrences in
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
It is time, then, to look carefully at the account of these ‘directional’
dimensions in the domain where Aristotle tells us their application is ‘most
appropriate,’ the study of animal locomotion.
8.3 De incessu animalium on Directions as Presuppositions

of Natural Inquiry
In IA 2, Aristotle introduces three different presuppositions of natural
investigation.
A starting point of the investigation [of the reason why]16 (Ἀρχὴ δὲ τῆς
σκέψεως) is by presupposing those things which we are accustomed to
use frequently in natural inquiry (τὴν μέθοδον τὴν φυσικήν), assuming
(λαβόντες) that this is the way things are in all of nature’s works. One of
these [presuppositions] is that [i] nature does nothing in vain, but
always does the best for the being of each kind of animal given the
possibilities; which is why if it is better in a certain way, that is also how
things are by nature. Again, [ii] it is necessary to assume (δεῖ λαβεῖν) [as
presuppositions] the dimensions of magnitudes, both how many and
what sort belong in what way. There are in fact six dimensions, but
three pairs, the first above and below, the second front and back, and the
third right and left. And in addition to these [presuppositions it is
necessary to assume] [iii] that the sources (ἀρχαὶ)17 of movements in
place are pushing and pulling.18 Now these are motions per se, while that
which is transported by another is moved per accidens; for that which is
transported by something does not seem to move itself but to be moved
by another. (IA 2, 704b12-705a2)
The reference to the methodos of nature in the opening sentence tells us that
the presuppositions about to be articulated are specific to the domain of
natural inquiry. Nevertheless, these presuppositions of natural investiga-
tion are characterized in language reminiscent of the framework established
in APo.19 These presuppositions, however, are presuppositions of natural
this passage support the view that a claim is εὔλογος if it is in agreement with a more general or more
fundamental proposition that Aristotle accepts as true.
16
The closing lines of chapter 1, to which this sentence is syntactically linked, say “For that (ὅτι μὲν)
these facts about animal locomotion happen to be so is clear from the inquiry into nature (τῆς
ἱστορίας τῆς φυσικῆς); we must now investigate the reason why (διότι δέ, νῦν σκεπτέον)” (IA 1,
704b9-10). Yet another link is that between the verbal adjective in this sentence (σκεπτέον) and the
opening words of chapter 2, Ἀρχὴ δὲ τῆς σκέψεως. Note here the clear echo of the ὅτι/διότι
characterization of the two stages of inquiry in APo. ii.1.
17
A good example of the ambiguity in the use of this term is discussed in note 2 above.
18
Referred to as the functions (τὰ ἔργα) of motion at MA 10, 703a19-20.
19
That is, these are three archai that are hupotheseis often used in natural investigations.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
8.3 De incessu animalium on Directional Dimensions 209
investigation, not of demonstration.20 Elsewhere,21 I have argued that, in his
studies of animals, the ‘nature does nothing in vain’ principle, the first
presupposition discussed here, is the sort of principle Aristotle refers to as
a hypothesis in APo.22 The same case can be made for the second presuppos-
ition of natural inquiry listed, the three pairs of directional dimensions. The
use of both of these starting points in Cael. ii.,23 however, introduces another
sense of ‘presupposition’ (ὑπόθεσις). Aristotle says these archai are appropri-
ate (οἰκεῖα) to the nature of animals, and that is why establishing them as
archai belongs to their investigation rather than cosmology. They are to be
used in the study of the heavens, but they are to be properly defined and
established as principles in his discussion of modes of animal locomotion.24
In the passage just quoted, the three presuppositions referred to are said
to be “used frequently in natural inquiry, assuming that this is the way
things are in all of nature’s works.” He does not explicitly restrict them to
the study of animals. Rather, it appears that Aristotle has selected three
general principles of natural investigation that are to be immediately put to
use in his investigation of the different forms of locomotion.25 This raises
the question of what he means in Cael. ii.2 when he claims the directional
principles are appropriate to the nature of animals.
The first step in answering this question must be to look at the actual
account of the directional dimensions given in IA 4 and 5, in order to allow
us to determine how the principles are defined there and what he does and
does not appropriate from that discussion for use in Cael. We will see that
20
For similar language in a very similar context, see Plato, Meno 86e4-5, 87b2-5.
21
Lennox 1997, 199–214 (repr. Lennox 2001b, 205–224).
22
See APo. i.2, 72a18-24. There are two uses of the first, negative half of this presupposition in Cael. as
well, at 271a25 (or a32 if one adopts Moraux’s suggested reordering of the text) and 290a30-33. Later
in this chapter we will be discussing a use of the positive version of the principle at 288a2-12, in
defending the applicability of ‘front’ and ‘back’ to the heaven, an especially problematic move. In
this case, uniquely, the principle is ‘God and nature do nothing in vain,’ a form of the principle
adopted by Galen.
23
The first appears a bit later in Cael. ii.5: εἰ γὰρ ἡ φύσις ἀεὶ ποιεῖ τῶν ἐνδεχομένων τὸ βέλτιστον
(288a2-3). For a valuable discussion of Aristotle’s invocation of this principle in Cael., see Leunissen
2009, 152–175; a different approach is defended by Johnson 2005, 136–140, though this particular text
is not discussed.
24
For an exact parallel, see Aristotle’s discussion of the organs that function both for excretion and
generation at PA iv.10, 689a4-b1, and the note to this passage in Lennox 2001a, 323–324. (τοῦτο δὲ
νῦν μὲν ὑποκείσθω, ὕστερον δὲ δειχθήσεται περὶ αὐτοῦ, 689a10-11; διορισθήσεται δὲ καὶ περὴ
τούτων ὕστερον, νῦν δὲ ὑποκείσθω μόνον . . ., 689a12-14). That the seed and the menstrual
discharges are moist residues is for now to be assumed, and their nature is to be defined and proven
later – presumably a reference to the discussion in GA i.
25
Oddly, however, while chapter 3 is a general discussion about the need in locomotion for something
against which limbs must press, the concepts mentioned as ἀρχαί in chapter 2, ὦσις and ἕλξις, are
never mentioned again in IA, and appear only once in MA!
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
far and away the most attention is directed to right and left, and this is
exactly as one would expect, given the nature of the inquiry. IA is a causal
investigation of animal locomotion, and of our three pairs of concepts only
right and left are restricted to animals that move from place to place, and are
in fact defined by reference to animal movement. IA 4 begins, nevertheless,
by tying all six concepts immediately to living things:
Although26 the dimensions (διαστάσεις) by which animals are naturally
bounded are six in number – above and below, front and back, and again
right and left – all living things have the above and below part. For above
and below are present not only in animals, but also in plants. It [viz. above
and below] is delineated by function and not only by position in relation to
the earth and the heaven. For whence the distribution of nourishment and
growth comes in each of these things is ‘above,’ and the extremity toward
which this extends is ‘below.’ The first is a sort of origin (ἀρχή τις),
the second a limit (πέρας), and above is an origin. (IA 4, 705a26-b2)
There are three subtly different messages conveyed by the contrast being
drawn here between ‘functional’ and ‘cosmic’ above and below, as I shall
call them.
1. Above and below can be delineated in two distinct ways, either by
reference to biological function or by reference to the center and
periphery of the cosmos.
2. A full delineation of above and below requires both a reference to
function and to cosmic orientation.
3. Different modes of delineation are appropriate in different contexts, or
with reference to different kinds of things.
Aristotle first notes that the above and below are positioned alike in plants
and animals, but then qualifies this claim by saying that “relative to the
whole [cosmos] they are not positioned alike, but relative to function they
are” (705b4-5), thus differentiating the two ways of defining the oppos-
ition. That is (as Aristotle goes on to explain), according to the functional
26
ἐπεὶ δ’ εἰσὶν. The syntax of this and the next paragraph is complex. The use of ἐπεὶ with the
indicative likely has the concessive force of ‘although.’ Aristotle wants to begin with a contrast
between most animals, which are bounded by six directional dimensions, and a wider class, which
includes plants, that is captured if we limit the directional dimensions to “the above and below part”
(τὸ μὲν ἄνω καὶ κάτω μόριον). The μὲν at 705a28 at first looks to be solitarium, adding emphasis to
the dimensions that belong to all living things. But the next paragraph begins ὅσα δὲ μὴ μόνον, and
I suspect the δὲ here loosely marks a contrast with the μὲν clause at a28. The contrast is very loose,
however, since a group of living things that has directional orientations in addition to above and
below is being contrasted not with a wider group of animals but with the above and below possessed
by that wider group.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
account of above, the roots of plants are above, but according to the
cosmological account they are below.27 Moreover, that there are two
distinct accounts is restricted to above and below – the accounts Aristotle
provides of front, back, left, and right are entirely by reference to function
and make no direct reference to cosmic orientation. Thus, at 705b8-12 he
states:
Both front and back, however, belong to such as are not only alive but are
also animals. For all of these have perception, and front and back are defined
according to this (ὁρίζεται δὲ κατὰ ταύτην). For those things in which
perception is naturally present and whence it derives in each of the animals is
front, while their opposites are back. (IA.4, 705b8-13)
It is settled doctrine for Aristotle that sense perception is the distinguishing
feature of animal life, and here he simply asserts that front and back are
defined by reference to the orientation and position of the sense organs. In
chapter 6, however, we are given more by way of a justification and account
for this claim.
The directional orientation marked by the concepts right and left is
found only in animals that “act on their own so as to move in respect of
place” (705b14-15). In this case he says not only that these are delineated by
‘a certain function,’ but also that they are not delineated by position
(705b17-18). How then, is this done?
In animals with organs of locomotion, these come in pairs, and it is in
these animals with limbs that right and left are clearest.28 In all of these
pairs one half of the pair of organs must be the source of motion, and from
705b31-706a9 Aristotle provides evidence that this side is the same for all
such animals and is what is designated as ‘the right.’ He concludes:
And on account of the same cause, the right sides of all animals are the same;
for whence the source of motion is in every case the same and is positioned
27
This passage should be supplemented by PA ii.10, 656a11-14; iv.7, 683b21-25; iv.10, 686b31-36, and
Juv. 1, 467b13-468a12. In fact, in PA ii.10 Aristotle not only says that human beings have their
functional above and below aligned with cosmic above and below – he says that their natural parts
are according to nature (τὰ φύσει μόρια κατὰ φύσιν ἔχει), where κατὰ φύσιν pretty clearly means
according to the nature of the cosmos.
28
See PA ii.1, 647a21-24; 8, 653b19-27; de An. ii.3, 414b1-10. Aristotle was definitely onto something
here. A recent discussion of “animal architecture” notes that “[m]ost familiar animals are bilaterally
symmetrical in that they have matching right and left sides with a central axis of symmetry running
down the middle of the long axis of the body. This design also imposes a front/rear orientation to
animals and has enabled the evolution of many efficient modes of locomotion” (Carroll 2005,
33–34). The author goes on to mention other forms of symmetry, especially among echinoderms
such as sea urchins and starfish. He would no doubt appreciate Aristotle’s long discussions of sea
urchins at PA iv.5, 680a4-681a9 and HA iv.5, 530a32-531a7.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
in the same place according to nature; and the right is whence the source of
motion (ὅθεν ἡ ἀρχὴ τῆς κινήσεως). (IA 4, 706a9-14)
If Aristotle were simply recommending labeling whichever side originates
motion ‘right,’ this argument would not be necessary. He must, there-
fore, be asserting the stronger claim that for any two animals, if they are
oriented in the same direction, they will by nature originate locomotion
from the same side. In principle, such a claim can be refuted by some
members of the same kind originating motion on one side, some on the
other, or by different kinds with the same perceptual orientation origin-
ating motion on different sides. But to some extent Aristotle can resist
such evidence by claiming that individuals (or groups) within a kind that
do not initiate motion on the right are acting contrary to nature. That he
would be willing to make that move is suggested by the way in which he
concludes chapter 4: “Of all the animals, humans have things arranged
most in accordance with nature; and the right is both better than the left
and separated from it. And that is why the right in human beings is most
right” (706a18-22). He then generalizes the argument to all of the direc-
tional orientations, concluding that the other sources, the above and
front, are also most in accordance with nature in mankind (706a24-
26).29 That is, things can be more or less aligned with what is taken to
be natural – all animals are natural, but some have things arranged more
naturally than others!30
The opening of chapter 5 continues this argument by applying these
abstract remarks about ‘above’ and ‘front’ to animals differentiated by
reference to how many feet they have.
So, then, those animals in which above and front are distinct (διώρισται), as
they are in human beings and birds, are bipeds (of the four points,31 two of
them are wings in the one case and hands and arms in the other). But those
that have the front and above in the same [orientation] are either four-
footed, many-footed, or footless. (IA 5, 706a26-31)
29
On the idea that the better of two related principles should be separated from the worse, recall GA
ii.1, 732a3-7: “Since the nature of the primary cause of motion, to which the logos and the form
belong, is better and more divine than the matter, it is also better that the superior be separated from
the inferior. On this account in those cases where it is possible, and to the extent it is possible, the
male has been separated from the female.”
30
Cf. Clark 1975, 28–40; Lloyd 1983, 26–43; Gregoric 2005.
31
Aristotle is explaining that, though blooded and therefore in possession of four limbs, two of them
do not qualify as ‘feet.’ The very next lines explain that by foot he means a part that has a ‘point of
contact’ (σημείον) with the ground for movement in place. An anonymous press reader correctly
noted that wings appear to initiate flight simultaneously – apparently it is only their walking that is
under consideration here.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
The rest of the chapter elaborates on these claims, and provides more
content to Aristotle’s axiology of direction, as we might call it. He notes
that in bipeds, and especially in human beings (birds have a pelvic articu-
lation that tips them forward slightly), the above and below are ‘in relation
to the whole cosmos’ (πρὸς τὸ τοῦ ὅλου, 706b4). However, since those
with four or many feet, as well as those with no feet, are not upright, their
‘above’ is on the same axis of orientation as their ‘front,’ and thus front and
above are not (directionally) distinct (cf. 706b2-9). This chapter concludes
with a statement I have already had reason to discuss – that it is reasonable
(εὐλόγως) that the origins (ἀρχαί) are from these parts because the origin is
an honorable thing, and the above, front, and right are more honorable
than the below, back, and left (706b11-13).32 There is, however, another
way to think about the relative honor of a part:
But there is beauty in also saying the reverse (καλῶς δ’ ἔχει καὶ τὸ ἀνάπαλιν)
about these matters, that it is because the origins are in these places that
these parts are more honorable than the opposing parts. (IA 5, 706b14-16)
Aristotle does not insist that one of these options is correct and the other is
wrong. The first option treats the greater honor of the above, the front, and
the right over their opposites as an unexplained premise, from which you
conclude that that is where the origins of motion will be. The second
option derives the greater honor of these parts that are above, front, or right
from the fact that these are where the origins of motion are located. The
premise common to the two options is that origins (ἀρχαί) are honorable.33
Aristotle does not think it is wrong to attribute greater honor to one of each
of these oppositions; the only issue is whether such attributions are to be
made a priori, or whether they follow from an empirical determination of
which of the opposites is the origin of motion. It is likely that Aristotle sees
Pythagoreans, and perhaps Platonists of a Pythagorean hue, as adopting
the former position.34 He is arguing for the latter.
32
A clear example where Aristotle is using ἀρχή to mean both the starting point or origin of motion
and a starting point or principle of reasoning.
33
For an extended discussion of this passage and a comparison of it with PA iii.3, 665a9-21,
Theophrastus, Metaphysics 11a8-13, and Plato, Timaeus 45a-b; cf. Lennox 1985, 143–164 (repr. in
Lennox 2001b, 266–272). I am still in substantial agreement with that discussion and review critical
reactions to its original publication in Lennox 2001b, 226–228. For two alternative readings of these
passages see Lloyd 1991, 27–53 (a reprint of his classic “Right and Left in Greek Philosophy” (Lloyd
1962) with a valuable introductory update); and Carbone 2016, 1–31.
34
In light of his criticisms of the Pythagoreans in Cael. ii.2, it is of interest that in his presentation of
the Pythagorean ‘table of opposites’ in Metaph. Α.5 (986a22-b2), the only directional dimensions that
appear in the table are right and left. For further consideration of that passage, see Schofield 2012, in
Steel (ed.) 2012, 141–166.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
IA 6–7 (706b17-707a23) argue that there must be a common source of
control over locomotion continuous with the locomotive parts, and that
these parts will be distinguished by being on the right or the left and (when
there are at least four parts) by being above or below. He rejects as
irrelevant to locomotion the distinction between front and back on
grounds that there is no natural backward motion (706b28-32).
8.3.1 Norms of Inquiry from De incessu animalium

Our concern now is the way in which Aristotle uses the results of his
discussion of directional dimensions in IA 2–6 as normative constraints on
inquiring into the applicability of the concepts ‘right’ and ‘left’ to the
heaven, and in particular to the motion of the spheres that constitute the
heaven. Part of that exploration will include a consideration of whether this
sort of ‘borrowing’ is consistent with Aristotle’s views about inquiry in
APo. Before embarking on that exploration, it is worth recalling that the
question of inappropriate borrowing of premises from one science for the
purposes of demonstrating in another science is a different question from
whether results from one scientific inquiry can be used to guide or constrain
an inquiry in another science, or in a distinct domain within a science. It is
the latter question that is of interest here.
I will briefly summarize the results of our examination of Aristotle’s
account of these ‘directional dimensions’ in IA, prior to turning to its use in
Cael. First and foremost, the claim in Cael. ii.2 that matters about these
directional dimensions had been determined in the discussions of animal
locomotion because these dimensions are appropriate to the natures of
animals is fully vindicated. It is only in IA 2–6 that we find an argument
that the three pairs of directions are starting points, as well as arguments for
priority both among and between the pairs.
Directionality is defined in IA by reference to biological function, and
two of the three pairs are specifically defined by reference to animal
locomotion, another reason for Aristotle to insist that these principles are
proper (οἰκεῖα) to that investigation. ‘Above’ and ‘below’ can be defined
both by reference to organic function and by reference to position relative
to earth and heaven. But only with respect to bipeds, and most completely
with respect to humans, is there agreement between the results of applying
cosmic and functional concepts of ‘above’ and ‘below.’ In all other loco-
motive organisms functional ‘above’ and ‘below’ are, to a greater or lesser
degree, aligned with ‘front’ and ‘back’ and on the opposing axis to cosmic
‘above’ and ‘below.’
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
Right and left are concepts applicable only in the case of locomotive self-
movers and are delineated only by function and not by position. And
finally, all natural locomotion is forward, not backward.
In the case of all three pairs, Aristotle argues that one member of the pair
is more of an ‘origin’ or ‘principle,’ and more honorable, than the other. In
the case of right and left, the source of movement, and thus the more
honorable side, is the right.
Aristotle had good reason, then, to insist in Cael. ii.2 that the appropri-
ate place to work toward definitions of these principles is in the inquiry
into animal locomotion. The question before us now is: In what way do the
results of that discussion serve as norms that shape the application of
directional concepts in the domain of cosmology?
We can begin by reminding ourselves of the context of Aristotle’s appeal
to IA for support of his cosmological argument. That appeal was made at
a point in Cael. ii when it is still an open question whether ‘right’ and ‘left’
have any legitimate use in cosmology. Whether the answer be yes or no, an
appropriate first step is to appeal to a previously worked-out account of
these directional principles, to see what, if anything, is applicable to the
inquiry into the movements of the heavenly spheres. This is a legitimate
move even if the ultimate answer is that there is no legitimate application of
these concepts in cosmology, since that worked-out account should help us
to answer this question of legitimacy.
More specifically, recall, there is a criticism of the Pythagorean method-
ology on the table – they have applied the concepts ‘right’ and ‘left’ to the
cosmos independently from the other two pairs of directional principles.
Aristotle appeals to IA specifically to defend a norm for their application:
there are directional concepts prior to ‘right’ and ‘left,’ so that unless it is
appropriate to apply these prior principles to the heavens, it cannot be
legitimate to apply ‘right’ and ‘left.’
For this reason, one might well wonder that the Pythagoreans spoke of only two
of these principles, the right and left, leaving aside four that are no less
important; for above and front are no less differentiated from below and back
than right is from left in all animals. I say ‘in all animals’ because in some cases
these directional principles differ by functional capacity (τῇ δυνάμει) only,
while in other cases they also differ by their configuration (καὶ τοῖς σχήμασι),
and while above and below are present alike in all the ensouled animals and
plants, right and left are not present in plants. And again, as length is prior to
breadth, if above is the principle of length and right of breadth, and the principle
of what is prior is prior, above would be prior to right – prior in generation, that
is, since ‘prior’ is said in many ways. (Cael. ii.2, 285a10-22)
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
Aristotle is following the IA account here in every detail.35 He begins by
noting the difference between above and below and the other directional
principles, that they can be defined both functionally and cosmologically,
and notes that they are the only pair that apply to all living things and not
only animals, specifically noting that right and left have no application to
plants. He then ties the directional dimensions directly to the three dimen-
sions of magnitude with which Cael. began by first reminding us of the
priority of length to breadth, and then invoking the idea from IA that
‘above’ is the principle of length and ‘right’ is the principle of breadth to
argue that above is the prior principle. He might also have mentioned, with
reference to the distinction between above and below, that human beings
are organized most according to nature – but in fact, this is not used to
justify the priority claim in either work.
There are three methodological points upon which to insist in consid-
ering the background in IA of the argument in Cael. ii. First, the depend-
ence here is not the dependence of cosmological demonstrations upon
premises that are only appropriate to zoological demonstration. Rather,
Pythagoreans are being criticized for ignoring four of the six dimensional
principles, and it is that criticism that relies on details of the discussion of
IA – in particular, they are criticized for ignoring a principle that has been
established in IA to be prior to ‘right,’ namely ‘above.’ The purpose of this
discussion is to establish the norms of inquiry for answering the questions of
whether and how these directional principles should be applied to the
cosmos. And given the relationship of these directional pairs to the three
dimensions of magnitude, and given the centrality of magnitude to the
subject matter of De caelo established at the very beginning of the inquiry,
there must be a strong presumption in favor of their application.
Second, the fact that Aristotle appeals to the IA account does not
commit him to the claim that the entire theory of IA is relevant here.
Just as a person studying optics need not be committed to the claim that all
of geometry is relevant to optical investigation, so Aristotle need not be
committed to the claim that all or even most of his theory of animal
locomotion is applicable in cosmology.
Third, the claim about the discussion of these principles being appro-
priate to the works on animal locomotion should not blind us to the fact
that there is an important dependence that runs in the other direction.
Throughout the discussion in IA 2, 4, 5, and 6, orientation of animals with
35
Simplicius (in Cael. 385.15–22 Heiberg) rehearses Aristotle’s general views about priority, but it is
clear that here he is relying directly on the arguments of IA.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
8.4 Domain Specificity of Inquiry in De caelo ii.2 217
respect to cosmic above and below is a constant reference point. In chapter
5, for example, organisms are compared by reference to “three places,
above, middle, and below” (706b2-3):
. . . bipeds have their above in relation to the above of the whole, many-
footed and footless animals in relation to the middle [of the whole], and
plants [have their above] in relation to the below [of the whole]. (IA 5,
706b3-6)
After he makes these comparisons, moreover, Aristotle goes on to give the
causes for these differences relative to cosmic orientation. The connection
to cosmic dimensionality is, however, even more subtle than this suggests.
Recall that what makes humans special is not just that our upright posture
aligns us with positional above and below – that alignment also accounts
for the fact that our three pairs of dimensions are distinct, whereas they are
confused to a greater or lesser extent in all other creatures. Thus, there is
a constant reminder in IA that above and below are prior principles to right
and left, and that the most natural arrangement for functionally defined
above and below is in alignment with cosmically defined above and below.
Thus, while Cael. ii.2 appeals to the discussion of directional dimension-
ality in IA 2–6, the latter discussion also explicitly depends on a concept of
cosmic above and below that derives from Aristotle’s general understand-
ing of the cosmos taken as a whole. Given how different the objects and
methods of these inquiries are, the number and variety of interconnections
is remarkable.
8.4 Domain Specificity of Inquiry in De caelo ii.2

In Cael. i.1, Aristotle asserts that it is reasonable to say all three dimensions
of magnitude belong to ‘complete’ bodies: indeed, this appears to be the
criterion by which the completeness of bodily magnitude is judged.36
Then, in Cael. ii.2, he treats one member of each directional pair as
a principle relative to one of the dimensions of magnitude: above to length,
right to breadth, and front to depth. Up to this point there is no need to
seek support from IA, since none of these ideas explicitly depends on the
study of animal locomotion. But he then notes, and this is where the
36
Pace Leggett 1995, 225, who rules out this meaning because it would make the above claim
‘tautological.’ Certainly, Aristotle does say that to call a body complete in this sense is to say that
it is extended in width, depth, and breadth, so that to say ‘complete body’ is, since bodies are three-
dimensional, strictly redundant. But here Aristotle is reminding the reader of his grounds for calling
bodies complete.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
relevance of referring to IA arises, that these are also starting points in
respect of motions – “starting points (ἀρχάς) whence the movements, in
those having them, first begin” (284b26-27).37 And here it is worth quoting
his precise words:
Growth is from above, movement in place is from the right, and movement
in accordance with perception is from the front – for I call front that toward
which perceptions are directed. (Cael. ii.2, 284b27-30)
For this reason, Aristotle tells us, we should only appeal to these principles
in bodies that “being ensouled, have an inherent source of motion; for in
none of the things without soul do we see that from whence comes the
source of motion” (284b32-34).38 It is then a straightforward matter to
argue for the use of these suppositional starting points in the investigation
of the motion of the heavens.
Since we determined previously39 that such capacities (αἱ τοιαῦται
δυνάμεις) are present in those bodies which have a source of motion, and
that the heaven is ensouled (ἔμψυχος) and has a source of motion, it is clear
that it also has above and below and right and left. (285a27-31)
It is notable that forward and backward are not mentioned in this conclu-
sion. One might suppose that this silence is due to the fact that the heaven
lacks perceptual organs, but on those grounds, one might think that above
and below, which in IA are defined functionally in terms of the intake of
nutrition and the direction of growth, would be equally lacking. But as we
have seen, Aristotle has an entirely nonbiological account of above and
below, which he invoked regularly in IA 2–5 – and indeed Aristotle relies
on this cosmological understanding of above and below in the argument
37
Note this is the precise formula that Aristotle uses for one of his causes, the misleadingly labeled
‘efficient’ cause.
38
Here we confront a widely noted difficulty with Aristotle’s fundamental claim that all natural
substances have within themselves a source of motion and rest: by occasionally listing the four
sublunary elements as natural substances (e.g., Cael. i.2, 268b16, Ph. ii.1, 192b9-12), he appears to
contradict the claim that only ensouled beings are self-movers; while by sometimes denying that
even animals can, strictly speaking, move themselves without an external mover (Ph. viii.2, 253a11-
21; 6, 259b1-16), he seems to deny self-motion even to ensouled beings! For a variety of discussions of
this problem, see the papers by Furley, Gill, Freeland and Sauvé Meyer in Gill and Lennox 1994;
Kelsey 2003, 59–87; Graham 1996, 171–192; and Waterlow 1982, ch. 5. Simplicius (in Cael. 387.5–24
Heiberg) provides us with Alexander’s view on this issue (that the elements have only a passive but
not an active source of motion), a view with which Simplicius takes issue.
39
If this is a reference to Cael., the only plausible candidate would be i.9, 279a17-28, on which see the
discussions of Kosman and Judson, chapters 7 and 8 in Gill and Lennox 1994. But the passage
currently under consideration is the first in which Cael. explicitly claims the heaven to be ensouled,
so this reference may well be to another work.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
8.4 Domain Specificity of Inquiry in De caelo ii.2 219
we are currently considering (cf. 284b34-285a1, 285b8ff.). Moreover, while
Aristotle argues in IA 6 that animals have a front and a back, he denies that
these concepts apply to their motions:
There is no such distinction according to forward and backward concerning
that which moves itself, on account of there being nothing to which natural
motion backwards belongs, nor is that which is moving defined in accord-
ance with it having made a motion in respect of these two directions; but
there is [such a distinction] in accordance with right and left and above and
below. (IA 6, 706b28-33)
Once again, the discussion in Cael. ii.2 hews closely to the account in IA.
The explanation for the absence of ‘forward’ and ‘backward’ here is
provided by the immediate purpose of this argument. Eventually, and
problematically, Aristotle does argue for the applicability of ‘forward’
(but not the forward/backward pair) to heavenly motion at Cael. ii.5,
288a3-12, an argument we will consider shortly. Here, however, Aristotle
is in the midst of criticizing the Pythagoreans for attributing right and left
to the cosmos while ignoring ‘prior’ principles. Among the three pairs, the
only pair that is prior to right and left, according to IA, is above and below.
Thus, following the argument in IA closely, he needs only to show that
right and left are posterior to above and below.
So what, then, is his positive argument for the attribution of the right
and left to the heaven? He begins with the claim that “it is not necessary to
puzzle”40 about these concepts applying just because the shape of the whole
cosmos is spherical, all its parts are identical, and it moves eternally (Cael.
ii.2, 285a32-b1)! Is this irony? Given that IA is quite clear that right and left
must be given a functional account, why should we not be puzzled? For that
functional account depends on being able to distinguish two sides (easily
done with most locomotive animals) and being able to determine which of
the two sides initiates motion. The eternal rotation of a sphere would seem
to be precisely the sort of motion to which ‘right’ and ‘left’ could not apply!
Nevertheless, the argument proceeds:
[I]t is necessary to think about it as though something, in which the right
parts differ in relation to the left also by configuration, were then sur-
rounded by a sphere; for it will have a different capacity, but will seem
not to have it, on account of the uniformity of its configuration (διὰ τὴν
ὁμοιότητα τοῦ σχήματος). And it is necessary to think in the same manner
about the origin of motion; for even if it never began to move, nonetheless it
40
οὐ δεῖ ἀπορεῖν.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
must have an origin (ἔχειν ἀναγκαῖον ἀρχήν) whence it would have begun if
it started moving and whence, if it were to stop it would start to move again.
(Cael. ii.2, 285b1-8)
I am not about to defend this argument; I find it weak.41 What is important
for the purposes of this chapter, however, is the way in which this inquiry
into whether there is an appropriate application of ‘right’ and ‘left’ to the
motions of the celestial spheres is dependent on the prior inquiry in IA. He
accepts that these concepts have primary reference to locomotive animals
and are defined by reference to function and not to position; and it appears
that reliance on the views established in IA is sanctioned by assuming that
the heavenly spheres are ensouled, and thus, “being ensouled, have an
inherent source of motion; for in none of the things without soul do we see
that from which comes the source of motion” (284b32-34).42
The nature of this dependence, however, is problematic. In the above mix
of counterfactual statements and thought experiment, Aristotle asks us to
imagine a thing that will still have right and left differentiated by functional
capacity, but in which, due to it spherical shape, this will not be apparent. And
we are further to imagine that this locomotive capacity is present even in
a thing that is eternally in motion – in material substances every actuality is the
actuality of a capacity, even if that capacity is always activated. It is in virtue of
that capacity that it would have begun to move, had it needed to, and would
begin again, if it were to ever cease moving. And given the results of the
inquiry in IA, we must assume that it would begin motion from the right!
The argument proceeds, famously (or infamously), to argue that cosmic
length extends from pole to pole along the axis of revolution, that the
direction around the equator is breadth, and thus that right and left for the
heaven will be the motive sources associated with movement of the sphere
of the fixed stars, as indicated by their passage from rising to setting.
Therefore, it is the pole invisible to those living in the northern hemisphere
that is in fact above, insofar as we are considering the revolution of the
sphere of the stars.43 On the other hand, since the spheres of the planets
41
As indeed may have Aristotle; see the discussion of this passage in Bolton 2009, 56–63. Notice the
uncharacteristic appeal to thought experiments and counterfactuals to make the case that we need
not be puzzled. For more on thought experiment in Cael., see Leunissen 2010, 165–166, and Rapp
2019, 79–88.
42
Aristotle may have in mind either that the elements do not have within them a source of initiating
motion from rest (see Ph. viii.4, 255b13-256a3) or (as a reader suggested) that there is no distinct part
of an element that can be isolated as the initiating part.
43
So that motion from the East will be ‘from the right.’ Yet another peculiarity of this discussion: as we
have seen repeatedly, above and below used with reference to the cosmos typically refer to the outer
periphery of the cosmic sphere and its center, respectively, not to the poles.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
8.5 Conclusion: What Sort of Inquiry Is De caelo? 221
rotate in the opposite direction, their right (i.e., their source of motion) is
on the opposite side, and thus the opposite pole would be up in relation to
their motion (285b8-286a1).
In his note on this passage, Stuart Leggatt comments that “in conceding
this point, Aristotle may be indicating the relative unimportance of the
issue for him” (Leggatt 1995, 226). I draw a very different conclusion: By
indicating that right and left will be defined relative to the source of motion
as determined by the direction of rotation, he is once again making use of
conclusions established in IA to guide the inquiry on the issue first raised by
Pythagoreans: whether and, if so, in what way ‘right’ and ‘left’ apply to the
cosmos as a whole. Moreover, the inquiry proceeds on the assumption that
the directional dimensions are the principles of the spatial dimensions of
length, breadth, and depth. For cosmology, a natural inquiry, issues of the
source and direction of motion are primary.
8.5 Conclusion: What Sort of Inquiry Is De caelo?

We have now seen that the reliance of Cael. ii.2 on the results of the
discussion aimed at defining directional dimensions in IA 2–6 is systematic
and self-conscious, and this reliance raises a number of important ques-
tions about Aristotle’s methods of inquiry. As we have seen, Aristotle treats
the three directional pairs (or the more honorable member of each) as
principles, in some sense, of the dimensions of magnitude. Now APo. i.7–13
lay down some rather strict guidelines regarding using principles appropri-
ate to one domain in proofs in a distinct domain; but that discussion is not
obviously relevant to using the results of one inquiry as normative con-
straints on another inquiry – and that is what is going on here.
Aristotle says clearly in Cael. ii.2 that the heaven is ensouled and thus
has its own source of motion;44 and since the dimensional directions are
principles appropriate to ensouled, locomotive beings and should only be
sought in the bodies of such beings (284b32-33), it is entirely appropriate to
treat them as suppositions of his inquiry into the motions of the heavenly
bodies. Yet while it may be ensouled, the heaven45 apparently does not
perceive, nourish itself, procreate, or initiate motion from a state of rest.
44
Cf. i.9, 279a18-30; ii.3, 286a7-12; ii.6, 288a28-b8; ii.12, 292a15-28. Admittedly, these passages are
often accompanied by those discussed in Bolton 2009, where Aristotle acknowledges the difficulties
of the questions and the relative lack of empirical support for the needed answers (e.g., ii.3, 286a3-8;
ii.5, 287b29-288a2; ii.12, 291b24-28).
45
It seems that some of the arguments in Cael. ii demand that ‘the heaven’ must be used in its ‘second’
sense (i.e., to refer not only to the outermost sphere but also to the spheres of the planets).
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
Moreover, being spherical, it has no obviously distinguishable organs. That
is, if it is ensouled, it is not ensouled in a way that fits easily with the account
of the soul that is familiar from De anima, nor with the account of living
bodies in De partibus animalium. In that case, one might fairly ask why he
works so hard to rely on the principles of directionality found in IA?
The answer, I believe, lies in certain epistemological constraints distinct-
ive to cosmological inquiry that become explicit in Cael. ii. These con-
straints, combined with the methodological commitments of the Analytics
(e.g., in APr. i.30) imply domain-specific norms for his inquiry into cosmic
locomotion that lead him to turn to his account of animal locomotion in
IA for guidance.46
Shortly after concluding his discussion of “the parts in respect of the
dimensions (κατὰ τὰς διαστάσεις) and such as are determined in respect of
place (κατὰ τόπον),” Aristotle takes up the questions of why it is that the
motions of the sun and moon are opposed to those of the stars. That
discussion opens with a cautionary warning:
Since circular movement is not opposed to circular movement, we must
investigate why (σκεπτέον διὰ τί) there are many motions, even if we are
attempting to make the inquiry (ποιεῖσθαι τὴν ζήτησιν) from a great
distance – distant not in respect of location, but much more by being able
to perceive far too few of the attributes (τῶν συμβεβηκότων) that belong to
these things. Nevertheless, we must discuss [this question]. (Cael. ii.3,
286a3-7)
Similarly, in Cael. ii.5, reflecting on the fact that the planetary spheres,
though made of the same natural substance, move in the opposing direc-
tion to that of the stars, Aristotle asks whether it is appropriate to even
attempt an answer.47
Perhaps, then, the attempt to make some statement about certain things, or
even about all things, passing nothing by, might indeed seem to be the mark
of great simplemindedness or of great zeal. Yet it is by no means right to
censure all people alike; rather one ought to look to their reason for saying
something, as well as the credibility of what they say, whether ‘human’
credibility (ἀνθρωπίνως)48 or something more secure. So then, when some-
one happens upon the more accurate necessities, it is necessary to be grateful
to those who made the discoveries, but at the moment we must state what
appears to be the case (νῦν δὲ τὸ φαινόμενον ῥητέον). (287b28-288a2)
46
And I here find my conclusions largely confirmed by Robert Bolton’s study of the underlying
epistemology of the De caelo in Bolton 2009, 51–82. Cf. Burnyeat 2004, 15–16, Lloyd 1996, 171.
47
Cf. Leunissen 2010, 160–164.
48
On the specific force of ἀνθρωπίνως here, see Bolton 2009, 68–70.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
Cael. ii.12 opens (291b24-31) with a similar ‘disclaimer,’ noting that over-
zealous attempts to state what appears to be the case (λέγειν τὸ
φαινόμενον), if done out of philosophical thirst to make even a slight
advance about subjects where we have great puzzlement, should be greeted
with indulgence. Slightly later in that same chapter, Aristotle is clearer
about why he is being so tentative. He has just raised a puzzle about the
different apparent motions associated with the spheres of the fixed stars,
the sun, the moon, and the planets. He pauses:
About these things, then, it is well to seek after ever greater understanding,
even though we have too few resources and are at such a great distance
from their attributes; nevertheless, studying that about which we are now
puzzling from such resources would not seem to be unreasonable. We
contemplate such things as if studying their bodies alone, and as though
they are units possessed of order yet completely without soul. But it is
necessary to suppose that they participate in action and life (δεῖ δ’ ὡς
μετεχόντων ὑπολαμβάνειν πράξεως καὶ ζωῆς); for in this way the attribute
we want to understand will seem to be nothing unreasonable. (Cael. ii.12,
292a14-22)
For Aristotle, the understanding of perceptual phenomena comes with
knowledge of the causes that produce them, and that knowledge
derives from experience of the phenomena (APr. i.30, 46a17-27).
Alas, cosmology is problematic in this respect. Nevertheless, he insists,
even though our resources (ἀφορμαί) are few and we are at a great
distance from the attributes (τὰ συμβαίνοντα) of the heavenly bodies,
it is not unreasonable to seek further understanding. He then suggests
that the motions of the heavens will be ‘less paradoxical’ if we assume
in our inquiry that they participate in action and life. Here is yet
another recommended norm for inquiring into the nature of the
heavens and their movements.
All these passages indicate that Aristotle, a committed inductivist,
recognizes the thin empirical basis for his causal theorizing in cosmology.
Yet they also indicate that he felt compelled, to the point of begging
indulgence, to pursue such theorizing. One may reasonably ask, why?
Why not simply resist temptation and say we are unable to reach any well-
grounded conclusions on this question, since we lack the data necessary for
reasonable causal inferences?
There are, I believe, two plausible answers, and they reinforce one another.
The first is stressed repeatedly in De caelo, but its relevance to the epistemo-
logical dilemma we are discussing is most clearly on display in PA i.5, in
a passage we have looked at previously.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
Among the substantial beings constituted by nature, some are ungenerated
and imperishable through all eternity, while others partake of generation
and perishing. Yet it has turned out that our studies of the former, though
they are valuable and divine, are fewer (for as regards both those things on
the basis of which one would examine them and those things about them
which we long to know, the perceptual phenomena are altogether few). We
are, however, much better provided in relation to knowledge about the
perishable plants and animals, because we live among them. For anyone
wishing to labor sufficiently can grasp many things about each kind. Each
study has its attractions. Even if our contact with eternal beings is slight,
nonetheless because of its surpassing value this knowledge is a greater
pleasure than our knowledge of everything around us, even as a chance,
brief glimpse of the ones we love is a greater pleasure than seeing accurately
many other and great things. (PA i.5, 644b22-645a1)49
Here the relative lack of perceptual acquaintance with the heavenly bodies
is stressed, and is explicitly contrasted with the ready availability of infor-
mation about animals and plants – for those willing to make the effort to
obtain it. But the two domains of investigation are also contrasted along
a different axis – the relative value of the knowledge gained. On that axis,
Aristotle contrasts the surpassing value of even a little knowledge of the
eternal things with the plentiful knowledge of the perishable things around
us. The extent of one’s willingness to draw causal inferences on a slender
database is directly proportional to the value of the prize at the end of the
inferential quest. This, in part, accounts for his willingness, in Cael., to
offer rather speculative explanations while simultaneously highlighting
their somewhat shaky foundations.
Yet Cael. provides us with evidence for another motive behind
Aristotle’s willingness, with apologies, to skate inferentially on thin evi-
dential ice. The evidence for this other motive lies in the constant attempts
to counter the arguments of the Timaeus and of unnamed Pythagoreans,
even when Aristotle agrees in some general way with their conclusions.50
Granted, in many cases observation provides scant grounds for reaching
conclusions – but for Aristotle, that must always be our starting point and
the measure of their acceptability. Thus, though the Pythagoreans turn out
49
See Chapter 6, 161, 171–173.
50
For a thorough account of the complex impact of the Timaeus on the De caelo, see Johansen 2009,
9–29. And the attack on the Timaeus goes beyond the discussion of the heavenly motions:
a significant portion of book iii is devoted to criticizing the geometric account of the elements in
the Timaeus; cf. iii.1, 300a1; 2, 300b17; iii.7–8, 306a1-307b18. Pythagorean attempts to generate
material elements out of numbers are also under attack there (e.g., iii.1, 300a17).
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
to be correct in applying the concepts of right and left to the cosmos, the
a priori ways in which they arrive at this application render it baseless.
And thus, Aristotle starts from an epistemically more secure base. The
directional dimensions have been defined by reference to “the animals and
plants around us.” As best we can, we must attempt to extend these
concepts, formed on this secure empirical basis, to the heaven – in pointed
contrast to the way the Pythagoreans do so.
In fact, in Cael. ii.5, as I briefly alluded to earlier, Aristotle not only
appeals to the discussion of directional principles in IA 2–6 – he also
appeals to the first postulate discussed there – that nature does what is
best among the possibilities – as a premise in his problematic attempt to
apply the concepts ‘front’ and ‘back’ to the heaven.
For if nature always does the best among those things that are possible
(Εἰ γὰρ ἡ φύσις ἀεὶ ποιεῖ τῶν ἐνδεχομένων τὸ βέλτιστον), and just as the
movement toward the upward place is, among rectilinear motions, more
honorable (for the upper place is more divine than the lower), in the same
way too motion forward is more honorable than motion backward, then it
possesses, if both right and left, just as we said before (and the puzzle just
stated testifies that it does), the prior and the posterior. In fact this cause
resolves the puzzle. For if it possesses the best state possible, this would be
a cause of the stated conclusion. For it is best to be moved in a simple and
ceaseless motion, and this [forward] is the more honorable direction. (Cael.
ii.5, 288a2-12)
Once again, Aristotle is relying on one of the starting points of natural
inquiry identified in IA 2. On a number of occasions Aristotle claims this is
a starting point that is posited based on what we observe, but in every case
that observational basis is the anatomy and behavior of animals.51 To use it
here to argue for the use of concepts defined by reference to animal
locomotion is clearly a stretch, and Aristotle is aware of it.
Nevertheless, he is providing an object lesson in empirical cosmology,
countering the approach found in Pythagorean doctrine. His appeal to the
discussion of above, below, right, left, front, and back in IA is an integral
part of this strategy of recommending a different methodological approach
to cosmological questions. In opposition to the application of those
concepts to the heavens in an a priori fashion by the Pythagoreans,
51
This is most fully spelled out in the midst of criticizing Democritus’ account of the growth and
shedding of teeth in GA v.8: “Since we posit, doing so based on what we see (ἐπεὶ . . . ὑποτιθέμεθα,
ἐξ ὧν ὁρῶμεν ὑποτιθέμενοι), that nature neither defaults nor does anything in vain given the
possibilities for each thing, then it is necessary that they have instruments for working on nutrients
so that they will be able to grasp nutrients once they stop suckling” (788b20-24).
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
Aristotle is recommending that investigation should begin, whenever and
to the extent that it can, by drawing on the results of observationally
grounded inquiry. Evidence for this application may be thin; but given
the value of the knowledge sought, we should appreciate the effort and try
to enrich the evidence and go further.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
chapter 9
The Order of Inquiry ii

The Debt of Zoological Inquiry to Meteorology iv
Both sinews and bones come to be by means of internal heat, as the

moisture evaporates. For this reason too, bones are indissoluble by
fire, just like earthenware; for as if in an oven, they are baked by the
heat during generation.
(GA ii.6, 743a17-21)1
Chapter Summary. In this chapter, we turn to Aristotle’s inquiry into the

formation of uniform bodies in Mete. iv and explore the ways in which the
investigations reported in PA ii and GA ii and v are dependent on it.
Aristotle often insists on the paramount importance for natural inquiry of
searching for ends as causes of natural processes, and thus of teleological
explanations in understanding the natural world. Yet, as we are about to
see, in Mete. iv he develops independent, material-level explanations of the
formation of uniform bodies, including the uniform parts of animals.
Looking carefully at this text will thus further our understanding of the
normative constraints on inquiry into nature. What are the limits on
material and teleological explanation in an Aristotelian science of nature?
Moreover, looking at the interplay between Mete. iv and the biological
study of the coming to be and being of uniform parts provides a second
rich case study by which to explore the nature of the dependence of one
natural inquiry on the results of another.
9.1 Introduction
In Chapter 8, we focused on the dependence of Aristotle’s discussion, in
Cael. ii, of whether the concepts of ‘right and left’ have application to the
movements of the heavenly spheres, on his discussion of the three pairs of
1
Ὑπὸ δὲ τῆς ἐντὸς θερμότητος τά τε νεῦρα καὶ τὰ ὀστᾶ γίγνεται ξηραινομένης τῆς ὑγρότητος. διὸ καὶ
ἄλυτά ἐστι τὰ ὀστᾶ ὑπὸ τοῦ πυρὸς καθάπερ κέραμος· οἷον γὰρ ἐν καμίνῳ, ὠπτημένα ἐστὶν ὑπὸ τῆς
ἐν τῇ γενέσει θερμότητος.
227
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
228 The Order of Inquiry ii
directional concepts in the opening chapters of IA. In that case, as we saw,
Aristotle explicitly signals this dependence (284b12-18). The discussion in
IA identifies these directional concepts as principles or starting points used
often in natural investigation – that is, they are being considered as (a)
presuppositions of investigation and (b) they may have application in
natural inquiry more widely, even if the appropriate place to define them
is in the investigation of animal locomotion. Aristotle draws systematically
on his discussion of directionality in IA 2–6, in part to reject the
Pythagorean attempt to apply the concepts of ‘right’ and ‘left’ to the
motions of the heavenly bodies, and in part to articulate his own alterna-
tive. As we saw, however, in using principles defined in the domain of
animal locomotion to make progress in a cosmological inquiry, Aristotle
had to import a number of questionable assumptions into that inquiry.
In this chapter, we will explore a very different sort of dependence
relationship between two other domains of Aristotle’s natural science:
the detailed dependence of the material and efficient causal explanations
of the nature and formation of uniform parts in PA ii and GA ii and v on
the conclusions reached in his ‘chemical’ treatise, Mete. iv. As we will see,
Mete. iv constitutes an independent, material-level inquiry into the for-
mation of uniform materials that possess a large constellation of emergent
powers or dispositions not found at the elemental level, and found more
widely in nature than just biological parts (e.g., in various kinds of
minerals). We will focus on the ways in which Aristotle’s biological
inquiries depend on a prior inquiry of the sort reported in Mete. iv,
a prior inquiry he references on a number of occasions. In order to
understand that dependence relation, however, we will need to explore
the limits of an investigation like that in Mete. iv and the limited role of
teleological explanation in Aristotle’s science of nature. Such limits are
consequences of the norms that govern inquiry in the two domains, as are
the relations of dependence between them.
Another feature of Aristotle’s method of inquiry will be highlighted in
this chapter: the way in which the norms that govern inquiry are not just
domain-specific, but level-specific. The inquiry into uniform bodies in
Mete. iv is not only shaped by the norms that govern all natural inquiry,
but by the more specific norms that govern inquiry into unqualified
generation defended in On Generation and Corruption (GC). This allows
the inquiry to assume conclusions of prior, and more fundamental, inquir-
ies as starting points for the more specific inquiry. It also implies
a methodologically grounded priority ordering among inquiries – which
is fundamentally different from either a pedagogical or an explanatory
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
9.2 The Emergence of Material Complexity 229
priority ordering (though of course there are interesting relationships
among these).
9.2 The Emergence of Material Complexity:

Meteorology iv.1–11
Mete. iv takes up in detail an inquiry only briefly sketched in On
Generation and Corruption (GC ): to find causal explanations for the
dispositional properties that are present in uniform materials but absent
at the level of the four elements – including, but not limited to, those
found in the uniform parts of animals, such as flesh, blood, and bone. It is
a book that has perplexed its readers from the very beginnings of scholarly
commentary. Alexander of Aphrodisias, for example, opens his commen-
tary on Mete. iv with a proclamation about its legitimacy and its place in
Aristotle’s natural philosophy:
The book entitled ‘the fourth’ of Aristotle’s Meteorology does belong to
Aristotle, but not to the treatise on meteorology, for the matters discussed in
it are not proper to meteorology. (in Mete. 179.3-5; tr. Lewis)
That Alexander opens his commentary in this way shows that there were
already questions both about the authenticity and editorial placement of
this work in the second century ce. Such worries expressed by such an
astute Aristotelian are one clear indication that this treatise can raise
questions both about whether its methods and conclusions conform to
those found in other treatises in the corpus, as well as more specific
questions about how its methods and conclusions relate to other natural
investigations.
The goal of this chapter is to demonstrate the dependence of the study of
the development and nature of the parts of animals on Mete. iv; but to
achieve that goal, we need first to understand Mete. iv on its own terms. It
is aimed at developing a material-level explanatory theory for the emer-
gence of the dispositional properties (δυνάμεις) of uniform compounds,
none of which are properties of the four elements. Some, but not all, of
these compounds are biological uniform parts, such as the uniform parts of
animals;2 and most, but not all, of these biological parts come to be for the
sake of constituting the nonuniform parts, or organs, which in turn
constitute living bodies. The coming to be of just the right uniform parts
2
As we saw in Chapter 4, Metaph. Ε.1, 1025b30-1026a5 distinguishes uniform and nonuniform parts of
plants as well.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
in just the right order and arrangement so that the end product is a highly
integrated, fully functional living being of a specific kind requires that the
material-level causes, which are explanatorily sufficient in Mete. iv, now
operate for the sake of producing and maintaining a functional animal of
a specific kind, under the guidance and control of the formal nature – that
is, the soul – of that animal. We will explore this topic more fully in
Section 9.3.
Mete. iv, then, is seeking to provide an explanatory framework – and
specific explanations for – the emergence of the differential, dispositional
properties of uniform bodies, including, but not limited to, the uniform
parts of animals and plants.3 The inquiry it outlines presents a problem for
two extreme views about Aristotle’s account of uniform bodies: (i) that no
‘bottom up’ account of such bodies is possible within an Aristotelian
philosophical framework; (ii) that there is an independent materialistic
account sufficient to fully explain the existence of all these parts.4
More positively, a careful study of Mete. iv provides insight into what
Aristotle thinks we are and are not able to understand about something
complex by reference to its material ingredients or constituents and their
efficient causal interactions.
As we are about to see, Mete. iv is an investigation that seeks explanations
of how a variety of dispositional contraries laid out along continua such as hard/
soft or pliant/brittle, which are not present in their elemental constituents,
emerge in the various uniform compounds of this world, from the metals mined
from the earth to the bark of a tree, to the flesh and blood of an animal. These
explanations are not primarily mechanistic, but thermodynamic.5 A given
3
Victor Caston (Caston 1997) argues that Galen is the first “emergentist” in the history of philosophy.
If my interpretation of Mete. iv is correct, Galen was anticipated by Aristotle. For the purposes of this
chapter, I’m following the account of emergence in Humphreys 1997, according to which the fused
elements may lose certain dispositional properties and will no longer exist as distinct elements, while
the emergent entity generated by their fusion will be characterized by novel dispositional properties
not found at the lower level.
4
The following well-known remark of Myles Burnyeat exemplifies the first extreme: “Aristotle simply
does not have our task of starting with the existence of matter as physics and chemistry describe it and
working up to the explanation of secondary qualities” (Burnyeat 1992, 22). The claim that Aristotle
“takes himself to have given material conditions sufficient to account for the generation and
continued existence of homoeomerous bodies” (Charles 1988 33) exemplifies the second extreme,
even given his claim that such an account is insufficient to fix the essence of homoeomerous parts.
A similar position is defended in Freudenthal 1995.
5
By saying they are not ‘mechanistic,’ I mean that Aristotle rarely makes explanatory appeals to
contact interactions such as pushing and pulling; by referring to the explanations as ‘thermo-
dynamic,’ I am drawing attention to the fact that the primary causal power or δύναμις involved in
these explanations is heat, either internal or environmental – that is, I am appealing to the Greek roots
of the term. From that perspective, the title of Furley 1983, “The Mechanics of Meteorology iv,” is
misleading – even though, by the criteria for ‘mechanistic explanations’ defended in, for example,
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
uniform body comes to have the dispositional differentiae it has by virtue
of the way that the agencies of heating and cooling have acted upon the
elemental ingredients from which it is constituted (or more precisely,
upon their primary passive dispositions, moist and dry). Such compounds
acquire and lose these differential dispositions of necessity whenever
acted upon by the appropriate sorts of heating and cooling under the
appropriate conditions.
In order, however, for parts like flesh or sinew to develop in such a way
as to be biologically functional, they must differentiate at precisely the right
place and time, and in precise and complicated relationships to each
other.6 Furthermore, each of them must be maintained in this dynamic
equilibrium throughout the life of the organism if it is to be preserved as
the specific living being that it is. Thus, there is a sense in which the
explanatory tools developed in Mete. iv are insufficient to provide an
account of what flesh is, even though they are sufficient to explain the
emergence of a certain earthen/watery compound with the dispositional
properties, the material nature, of flesh. It will, that is, allow you to grasp
the sort of uniform material that flesh is and to explain the affections it has
in virtue of being that sort of material. The distinction is clearly spelled out
in GA ii.1:
So, then, hardness, softness, viscosity, brittleness, and as many other such
affections as belong to ensouled parts, these may be produced by heat and
coldness – not, however, the formula (λόγος) by which the one part is
already flesh and another bone; rather [that is produced by] the motion
derived from the generator, which is in complete actuality what that from
which [the offspring] comes to be is potentially . . . (734b31-36)
That “motion derived from the generator,” as we will see, is a special
capacity of the nutritive soul, transmitted by the semen in sexually gener-
ated animals, which is responsible for the temporal and spatial organization
of the heating and cooling during biological development, so that an
organic system comes to be and not just a random collection of uniform
materials. More on that topic in Section 9.3.
Machamer, Darden and Craver 2000 or Machamer 2004, the explanations in Mete. iv would count
as mechanistic.
6
As Aristotle points out explicitly: cf. GA ii.1, 734b31-735a26; ii.4, 740b12-741a3; ii.6, 742a10-16,
743a26-b5. This remains the central explanandum of developmental biology today, as this very clear
statement of the problem from a contemporary advanced textbook shows: “This generation of
cellular diversity is called differentiation. Since each cell of the body . . . contains the same set of
genes, how can this identical set of genetic instructions produce different types of cells? How can the
fertilized egg generate so many different cell types?” (Gilbert 2006, 4; emphasis in original).
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
In Mete. iv.8 Aristotle presents a list of eighteen opposed pairs of
dispositional properties.7 These properties are not present in the four
elements, yet they are present in the uniform bodies that are predomin-
antly compounds of earth and water (or ‘the dry and the moist’) – indeed,
not only are they present, but it is these properties that give the uniform
materials their identity and differentiate them from one another. These
dispositional properties are essential to the uniform parts playing their
distinctive functional roles in the activities of living things – and yet they
are also found in many inanimate materials, and the early chapters of Mete.
iv account for their emergence without any appeal to biological function.8
What we find instead are complex, material-level explanations of their
emergence, framed entirely in terms of the consequences of materials with
various degrees of moistness or dryness undergoing various sorts of heating
and cooling under different, specifiable conditions.
9.2.1 The Problem of Emergence in On Generation and Corruption (GC)

GC ii.7 begins with an aporia for those who deny that the elements can be
transposed into one another, but then notes a second aporia for those, like
Aristotle himself, who allow inter-transposition:
In what way does something different, besides these elements, come to be
from them; I mean, for example, water comes from fire and fire from this,
because there is a certain shared substratum. But clearly flesh and marrow
also come to be from these – how do they come to be? (GC ii.7, 334a22-26)
A bit later in the chapter he puts the problem more precisely:
. . . how will there be something <else> from both, for example, from cold
and hot or from fire and earth? For if flesh is from both but it is neither of
them, nor again is it a composite of preserved elements (σύνθεσις
σωζομένων),9 what is left for what is composed of these to be, except matter?
7
παθητικά spoken of κατὰ δύναμιν καὶ ἀδυναμίαν (Mete. iv.8, 385a11-12) – affective properties spoken
of in virtue of their capacities to undergo or resist specific kinds of change (e.g., the capacity or
incapacity to harden or to melt). For the full list of eighteen pairs, see 385a12-18.
8
There are, it should be noted, a number of references to ‘completion’ (τελείωσις) and ‘end’ (τέλος)
during the discussion of concoction (πέψις) and its forms – for example, 379b18-22, b25-28, 380a13-
16, a31 – and inconcoction and its forms are characterized as incompletions (ἀτέλεια) at 380a6-9, a31-
33. This language is understandable given that Aristotle is discussing processes of cooking and
ripening and failures of such processes. What is completely absent is the language of ‘final causality,’
that is, τὸ οὗ ἕνεκα or ἕνεκα τοῦ.
9
The view that they are mixtures of preserved elemental constituents he attributes to Empedocles –
the idea being that uniform bodies consist of (say) actual quantities of earth and of water, combined
according to a certain ratio.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
For the passing away of the one <element> either produces the other
<element> or the matter (334b3-7).
The answer to this aporia depends on the theory of μῖξις developed in GC
i.10. According to that theory, two bodies at different points on the hot/
cold and moist/dry continua, but sharing a common substratum, interact
with one another and as a result come to occupy a point within an
intermediate range on at least one of these continua.10 Rather than one
contrary on each continuum being converted to the other extreme, these
contrary powers are ‘equalized,’ held at a mean point, in a sort of dynamic
tension. Things like bone or copper are the products of such an interaction:
they are uniform bodies in that they can be divided forever but will never,
even in principle, be reduced to bits of earth and water; by subjecting them
to suitable conditions, you could, however, reduce them to earth and
water.11
GC ii.7 presents a highly schematic solution to its aporia, but there is no
actual attempt to explain the coming to be of the dispositions that serve to
distinguish a uniform composite, such as flesh, from its elemental ingredi-
ents. Other than pure theory, here is all we are told:
. . . and first the elements change, and from them flesh and bones and such
things arise, the hot coming to be cold, the cold hot, when they approach
the mean (πρὸς τὸ μέσον ἔλθῃ);12 for there neither one [of the elements]
exists, though the mean is of some extension and is not indivisible (τὸ δὲ
μέσον πολὺ καὶ οὐκ ἀδιαίρετον). Likewise too, the dry and moist and such
things produce flesh and bones and the other <uniform bodies> in accord-
ance with the mean. (334b24-30)
Since flesh and bone are the products in each case, and since both moist
and dry and hot and cold designate contrary ends of continua, it is safe to
10
That there is a range of intermediates is clear from 334b28 (τὸ δὲ μέσον πολὺ καὶ οὐκ ἀδιαίρετον)
and from the idea that the λόγος (here, certainly ‘ratio’) of the contraries in the compound can be
2:1, 3:1, or some other (334b15-16). See Frede 2004 289–314. In John Cooper’s contribution to the
same volume (Cooper 2004, 315–326), he poses the question of whether the theory of mixture
defended in GC i.10 and slightly elaborated in ii.8 implies that every part of a mixture has
(potentially) precisely the same balance of elements as every other part. His view is that the theory
elaborated in GC does not imply this. While I cannot explore the issue here, Mete. iv must be of
considerable value in deciding the question.
11
I defend this way of understanding the notion of potential presence in the exchange between Sorabji
1989 and Lennox 1989, published in Cleary and Shartin, eds. 1989, 35–76. Cf. Popa 2009.
12
πρὸς τὸ μέσον is a very common prepositional phrase in Aristotle (used, with very different
references, in discussing the virtues (EN 1106b9, 1108b28; EE 1222a26, 1234b7), the middle term of
a syllogism (APr. 28a17, b5, 29a1, 31a19, b11), and the center of the cosmos (Cael. 269a27, 270a7,
295a13, a34, b21, b25 – and many more; Ph. 212a26, 214b15), but this is the only instance of its use with
this verb.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
assume on the basis of this passage that these uniform parts are the products
of some sort of interaction between the means of both. But we can only
guess because we are not further enlightened by this discussion.
Aristotle can do, and does do, much better in Mete. iv.13 There are,
however, three theoretical points imported from the discussion of GC ii.7
into Mete. iv, and this is likely an important factor in Alexander’s thought
that Mete. iv would more properly be treated as a sort of appendix to GC.
These points constitute a domain-specific framework, more specific than,
but consistent with, that provided by the Physics, for the inquiry into the
formation of uniform bodies that we find in Mete. iv. That framework
includes the following theses:
1. The four contrary elemental powers denote extremes on two continua.14
2. The ‘mean’ refers to a range along such continua, and thus the resulting
emergent powers of uniform bodies may also vary continuously – for
example, bone can be more or less hard, flesh more or less soft, and
so on.
3. The resulting uniform body will also be observably somewhere on the
hot/cold and moist/dry continua, though not at the extreme end
represented by, say, earth or water – in the biological works, as we
will see, bone is often said to be ‘more earthen’ than flesh.
GC ii.2 also provides its own list of differential tactile contraries – hot,
cold, dry, moist, heavy, light, hard, soft, viscous, brittle, rough, smooth,
coarse, and fine (329b18-20) – and argues that “all the other differences are
reduced to the first four, while these are not reduced to fewer” (330a24-6).
Those dispositional contraries said to be reducible to the primary four
include some, but by no means all, of the eighteen catalogued in Mete. iv.8.
The idea here expressed in terms of ‘reduction’ may well be the converse of
the idea expressed in PA ii.1, that certain dispositional properties ‘follow
from’ the four primary ones that are constitutive of the four elements. All
of the ‘secondary’ dispositional properties15 on the list are said to be ‘of the
moist’ or ‘of the dry.’
13
Much of the literature on this topic restricts itself to a discussion of GC. Bogaard (1979, 11–29)
provides a valuable discussion of mixis in GC, but refers to Mete. iv as a “collation of empirical data”
and more or less ignores it. So too Needham 2009, 149–164. Frede 2004 does devote a few pages to
Mete. iv.
14
Compare the discussion of contrariety in Metaph. Ι.4–7 (thanks to Devin Henry for reminding me of
this background).
15
I realize the expression ‘secondary property’ might be misleading, given its use in John Locke and
modern empiricist literature generally, but it seems the most straightforward way to denote the
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
While providing a general framework for Mete. iv, GC fails to answer
the following questions:
1. Supposing that these secondary differential dispositions all ‘follow
from’ (or are ‘reducible to’) moist and dry, how do such things as
viscosity or hardness emerge from a ‘dynamic tension’ – an equalizing
at the mean – of the primary dispositions, as GC ii.7 demands? It is
never explicitly mentioned here, though it is obvious by implication,
that the secondary dispositional properties mentioned are not proper-
ties of the elements, while the primary ones are constitutive of the
elements, on a number of occasions said to be their matter.16
2. What role do ‘hot’ and ‘cold’ play in the emergence of the uniform
bodies – are they merely efficient causes acting on more or less moist
and dry bodies, or are they also in some way constitutive – material
causes – of the emergent product?
3. What does it mean to say that the ‘secondary’ contraries are reducible
to, and/or follow from, the primary?
These questions are all within the general erotetic framework established
by the Physics for the study of natural beings, but they are left unanswered
by the general study of coming to be and passing away in GC. Mete. iv.4–9
sheds light on all of these questions.
9.2.2 The ‘Thermodynamics’ of Meteorology iv.4–7 17

I begin with a very brief overview of the theoretical framework provided in
Mete. iv.1–3. The first chapter opens by reminding us of the general
framework established in GC ii,18 of four elements constituted by
a ‘yoking together’ of one each of the two paired, contrary powers, hot/
contrast with the primary dispositions. Aristotle does make an interesting distinction in Mete. iv, in
introducing his list of emergent dispositions, between those that differ in relation to the senses (πρὸς
τὰς αἰσθήσεις) such as smells and tastes, and those, like solidity and flexibility, that are more
appropriately termed affections; and those he goes on to list all fall into the latter category (Mete.
iv.8, 385a1-11).
16
For example, PA ii.1, 646a16-17; Long. 5, 466a21-3; Mete. iv.1, 378b32-34, 10, 388a21-22. For two
different attempts to make philosophical sense of this idea, see Gill 1989 78–82, 243–252; Furth 1988,
76–83. Claiming that the four contraries are the matter of the composites is at odds with the idea that
they are qualities inherent in some underlying ‘prime’ matter.
17
Frede 2004 unfortunately only looks at chapters 8–12, and thus not at the causal theory of the
production of the dispositions that differentiate uniform compounds.
18
The reference need not be directly to GC, however, since the same doctrine is stated in Mete. i.2–3 as
well.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
cold, moist/dry – fire being hot/dry, water being cold/moist, and so on
(378b10-14). But it adds something important to that framework as well. In
GC ii.1 Aristotle tells us it is well said that the four primary powers are
sources and elements (ἀρχὰς καὶ στοιχεῖα, 329a5-6) of the “so-called
elements” (τὰ καλούμενα στοιχεῖα, 328b32). Mete. iv.1 opens, however,
by referring to these four primary differential powers as the causes of the
four elements (378b10). One of these pairs, hot and cold, is said to be active
or productive, and the other, moist and dry, is said to be passive: moist and
dry are ‘given definition’ through the productive actions of the hot and
cold (378b21-22). These opposing powers are causes, then, in two senses: as
in the passage from PA ii.1 discussed earlier, all four are the matter out of
which the elements are constituted;19 but they are also causes in that it is
efficient causal interactions between the actions of hot and cold and the
passive capacities of moist and dry that give rise to the uniform bodies that
are the primary explananda of Mete. iv. As Aristotle puts it much later, in
chapter 10:
. . . the matter from which these [composite bodies] are constituted is the
dry and moist, and therefore water and earth (for each of these has the very
conspicuous potency of one or the other of those), while the productive
potencies are heat and cold (for heat and cold constitute and solidify bodies
from the dry and moist) . . . (388a21-25)
Moreover, Aristotle occasionally (382b5-7, 389a29-32) points out that since
cold is the shared potency of earth and water, there is a sense in which it
should also be viewed as a material constituent of composite bodies, along
with moist and dry.
The remainder of the first chapter is devoted to a general account of the
role of these potentials in natural generation and destruction. At 378b31-34,
simple and natural generation is defined as “a change from the underlying
matter for each nature by these potencies, when they stand in a proportion
(ἔχωσι λόγον).” Once again, Aristotle is framing the inquiry in terms of
the principles established in Ph. ii. The investigation to follow is described
as a study of “the operations (τὰς ἐργασίας), which the active causes
perform and the forms (τὰ εἴδη) of the passive causes” (378b26-28).
Chapters 2–3 proceed to discuss the operations of the active causes, while
19
Or perhaps they are the matter of the uniform bodies. As we will see, in discussing the causal
processes that give rise to the uniform bodies, Aristotle’s preferred language is that of heat and cold as
agents transforming bodies on the moist/dry continuum – though he often describes these bodies as
earthen or ‘of earth’ and watery or ‘of water.’ This is apparent in 388a21-25, quoted below.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
chapter 4 announces that it will discuss “the forms of the passive <poten-
cies, causes>, that is, of the moist and the dry” (381b23-24).
Given the aims of this chapter, I want to highlight two points from these
‘framework’ chapters that are crucial for understanding Aristotle’s account
of the emergent differential powers of uniform bodies. First, Aristotle
characterizes the agents of the changes he is about to discuss not, at least
not directly, as hot and cold, but rather as the ‘operations’ (τὰς ἐργασίας)
by means of which hot and cold act. In chapters 2 and 3 these operations are
generically described as ‘concoction’ and ‘inconcoction,’ which are then
differentiated into three paired contraries – a classification that is, puz-
zlingly, generally ignored throughout the rest of Mete. iv.20 The other side
of the investigation is described as a discussion of the forms of the passive
powers, the dry and moist themselves.
Second, the active causes are said to produce their results “from under-
lying things already constituted by nature” (379b10-11), that is, things
already possessed of both the ‘primary’ differential powers and various
‘secondary’ ones. And indeed, the investigation does not proceed by
starting with pure elements and describing the effects of (say) fire on
(say) water, or earth. Rather it starts with already naturally constituted
uniform bodies and moves to the investigation of the differential effects of
the operations by which hot and cold affect them.21 Take, for example,
Aristotle’s description of how one uncovers the elemental character of
blood (a uniform fluid) by investigating what happens to it when it cools.
Blood too behaves similarly; for it solidifies (πήγνυται) when it cools by
being dried. Those kinds of blood that do not solidify, for instance that of
the deer, consist more of water (ὕδατος μᾶλλον) and these are cold. Hence
such blood has no fibers. For fibers consist of earth (γῆς) and are solid.
Therefore blood, when the fibers have been removed, also does not solidify.
This will be because it does not become dry, for the remainder is water, like
milk when the cheese has been removed. And here is an indication: diseased
20
For valuable discussions of the range of uses for these terms, see Düring 1944, 68–74 and Lloyd 1996,
83–103. Neither comments, however, on the virtually complete absence of this taxonomy of forms of
concoction and inconcoction throughout the remainder of Mete. iv. (Inconcoction is mentioned
once in ch. 7 at 384a33 and concoction once in ch. 11 at 389b8.) In the biological works, concoction
and inconcoction and their cognates are used repeatedly to describe the way heat operates in
nutrition and generation, but again there is no systematic use of the elaborate classification found
in Mete. iv.2–3.
21
An implication of this description, tellingly supported by the way Aristotle grounds a number of his
explanations, is that he thinks you can ‘get at’ the underlying nature of these bodies by means of
experiment. Needless to say, this runs counter to a common view about Aristotle’s science: however,
see Newman 2004, especially the discussion of ‘Aristotelianism and Experiment’ that opens chapter
5, 238–242; cf. Newman 2006.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
blood is not disposed to solidify; for it is serous, and serum is phlegm and
water because of inconcoction (διὰ τὸ ἄπεπτον) and lack of mastery by the
animal’s nature (ὑπὸ τῆς φύσεως). (Mete. iv.7, 384a25-33)
The discussion begins with a fluid uniform body, blood. Two different
kinds of blood, with and without fibers, behave differently when they
undergo cooling outside the animal’s body, and on the basis of that
difference an inference is made about their distinctive material natures –
one is more watery, the other more earthen. The evidence that backs up
these inferences is of considerable interest: Aristotle notes that as some
blood cools and dries out, no solid residue remains, while with blood of
a more fibrous nature, a residue does remain; he then claims that if the
fibers are removed from fibrous blood, it does not solidify.22 Next, he
draws an analogy to the process of removing curds from milk in the
production of cheese. And finally, he notes that diseased blood is serous,
consisting of water and phlegm due to a lack of proper concoction.23
With this understanding of the nature of the investigation as back-
ground, it is time to take a detailed look at the investigation itself. The
first emergent dispositional properties (or powers),24 he argues, are hardness
and softness, since they are necessarily present in bodies with their own
proper boundary:
22
HA iii.6, 515b30-516a6 also discusses a kind of fiber that comes to be in blood. Fiber does not arise in
the blood of deer, antelope, roe, or hares (b33-35), but blood solidification in these kinds ranges along
a more/less continuum. Deer blood solidifies like milk without rennet; antelope’s blood solidifies
better, somewhat like that of sheep. For help in thinking about what Aristotle might have in mind
by ‘fiber’ and removing it from blood, I consulted with Dr. Resia Pretorius, director of the Applied
Morphology Research Centre in the Department of Physiology, University of Pretoria. She informs
me that if blood is allowed to sit undisturbed in a container, the components will separate in the
same way they do when centrifuged, though she wondered if it would need to be mixed with an
anticoagulant (any liquid high in citric acid would do). When this is done, the fibrin is concentrated
in a yellow layer above the red blood cells. If Aristotle extracted that layer, the blood would then not
clot – or so Dr. Pretorius speculated, since she had not tried it.
23
Thin blood due to abnormally low platelet production is a symptom of a wide and diverse class of
illnesses. PA ii.7, 652b33-653a8 explains that fluxes begin in the head because when the blood around
the brain is cooled this region has “the capacity to produce phlegm and serum.” Compare the
following, from the Hippocratic text On Fleshes, on the formation of the liver: “When you slaughter
a sacrificial animal, as long as its blood remains warm it is liquid, but when it becomes cold, it
congeals. If, however, you shake it, the blood does not solidify, for its fibers are cold and gluey”
(Carn. 8 [Potter trans.], 1995).
24
In recent philosophy, there have been a number of attempts to develop a taxonomy for the latent or
‘nonoccurrent’ attributes of things: powers, potencies, capacities, dispositions, abilities, capabilities,
and so on. I will let context determine which of these terms I use to render Aristotle’s δύναμις, which
has a generic sense (which encompasses nature (φύσις)), referring to any source of change, and
a specific sense referring only to a source of change in another thing (or in the thing itself qua other)
or to a source of being changed by another (cf. Metaph. Θ.1, 1046a9-36).
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
Among the bodily affections, these must belong first to a definite thing:
hardness and softness. For what is from moist and dry must be either hard or
soft. A thing is hard which does not yield into itself along its surface; a thing
is soft which yields without being displaced. (Mete. iv.4, 382a8-13)
‘Hard’ is defined as that property in virtue of which the surface of a body
resists ‘yielding into itself,’ ‘soft’ as that in virtue of which a surface does
yield, and in a way that does not involve mutual displacement (382a11-13).
This last clause is crucial, since it leads Aristotle to point out that for that
reason water is not soft – that is, the hard-soft continuum is emergent, it is
a property of compounds that is necessarily absent in their ingredients.
What have we learned, then, about the emergence of this most basic
contrary-continuum on the ontological scene? It is dependent on the appear-
ance of a certain sort of compound, because it is only with a true compound of
earth and water – something that is actually neither – that the ‘easily
bounded’ (i.e., the moist, 381b29) takes on actual boundary. Hardness and
softness are properties of bodies with surfaces, for the distinction is in fact
along a ‘yield’ continuum. What is unqualifiedly hard is what does not yield
at all to applied pressure, and what is unqualifiedly soft is what does so easily,
while not displacing the surrounding medium.25
To this point we have not been told anything about how bodies come to
be bounded by a surface – only that when they do, their surfaces will vary
along the hard-soft continuum. Mete. iv.5 aims to fill this void. It begins by
reminding us of the modal conclusion of the previous chapter, and draws
an implication from it.
A body that is made definite by its proper boundary26 must be hard or soft
(for either it yields or it does not). Furthermore, it must be solidified (for it is
made definite by this). Therefore, since everything definite (πᾶν τὸ
ὡρισμένον) and constituted is either soft or hard, and these are due to
solidification, no composite and definite body can exist without solidifica-
tion. (Mete. iv.5, 382a22-27; emphasis added)
Though the argument is sound, the premise that becoming a body with
a defined boundary is due to solidification is in need of justification, which
Aristotle goes on to provide, though not before reminding us of the causal
framework governing natural inquiry that is being assumed.
25
He speaks of these properties as universally present in uniform materials, but sometimes appears to
restrict them to things that are at least to some extent solid, as we shall see presently. For doubts
about whether all the uniform bodies on the hard/soft spectrum are compounds, see Popa 2009.
26
οἰκείῳ ὅρῳ. I take the contrast to be with fluids or piles of sand, which have a defined boundary only
by being placed in a container.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
Now the causes besides the matter are two, the agent and the affection – the
agent is a cause as whence comes the motion, whereas the affection is a cause
as form (τὸ μὲν οὖν ποιοῦν ὡς ὅθεν ἡ κίνησις, τὸ δὲ πάθος ὡς εἶδος).
Therefore, these are causes of solidification and liquefaction, and also of
drying and moistening. An agent acts by means of two potencies (δυσὶ
δυνάμεσι), and a thing is affected by means of two processes, just as we have
said. A thing acts by means of hot and cold, whereas the affection is due to
the absence or presence of hot or cold. (Mete. iv.5, 382a27-b1)
This causal framework is interesting because of what it does not include –
there is no mention of an end for the sake of which these processes are
occurring.27 Moreover, the ‘forms’ under consideration are affections, such
as solidity or liquidity. These ‘forms’ are not the forms of natural substances,
such as the souls of animals, but affections brought about by the agency of
heating and cooling giving rise to certain processes (evaporation, melting).
The focus in chapters 4–7 is on earthen/watery compounds being dried,
moistened, or solidified, due to the operations of internal and external heat
and cold. If these are the materials, agents, and processes at work, what are
the affections qua form? They are, as we will see, the emergent differential
dispositions – such things as solidity, hardness, viscosity, malleability, and
their opposites.28 That is, the theory of Mete. iv.4–7 is a theory of how the
operations of internal and external heat (or their absence), acting on the
matter (i.e., on elemental compounds of varying degrees of moistness and
dryness), necessitate changes along the solid/liquid and hard/soft continua –
and do so primarily through the processes of drying things out or moistening
them. The elemental compounds are, then, the matter underlying these
emergent dispositions, and those dispositions constitute the affective forms of
composite materials. It is important to note, however, that of the eighteen
pairs of emergent dispositional properties identified in chapter 8 only the
first three pairs – solidifiable/unsolidifiable, meltable/unmeltable, soften-
able/unsoftenable – are discussed in chapters 4–7. These are, Aristotle is
insisting, the primary passive dispositions (i.e., dispositions to be acted upon
or to resist being acted upon).
27
Nor is there in Mete. i–iii; see 134, note 24. Famously, Mete. iv.12 notes this absence; on which see
Gill 2014.
28
When considering the use of the concept of ‘form’ (εἶδος) here, it helps to keep in mind that the
entities under consideration are uniform materials (copper or clay, bone or blood) being considered
in themselves, and not as parts of organisms or artifacts. As elsewhere, ‘form’ here refers to what it is
to be something – in this case, those dispositional properties that, once acquired, differentiate
a specific uniform body from all others, qua uniform body. For the use of εἶδος to refer to bodily
affections, see Ph. i.7, 190b28-29; ii.1, 193b19-20.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
The processes involved, on the other hand, can be described both actively
and passively, but Aristotle typically characterizes them passively, as pro-
cesses undergone by the compound – being dried or moistened. Doing so
is to be expected if Aristotle is working within the general framework of his
account of change at Ph. iii.3, 202a13-21 – the locus of change, he there
argues, is in the patient, not in the agent.
Note that these basic processes refer to alterations in one or the other of
the primary passive dispositions, the dry and the moist. The ‘emergence’
question can thus be more precisely focused by asking how such processes
give rise to properties other than moist and dry. In Mete. iv, then, Aristotle
does set himself “the task of starting with the existence of matter as physics
and chemistry describe it and working up to the explanation of secondary
qualities.”29
In fact, Mete. iv.5–7 provide us with a list of an Aristotelian equivalent
of laws of nature.30 Their form is:
As many compounds as are <K> undergo φ when <H, C>
K is typically an elemental description of the kind of compound in
question – which can be represented symbolically in the following way:
<E/W> = balanced compound of earth and water31
<E/w> = compound of earth and water with earth
predominant
<e/W> = compound of earth and water with water
predominant
<W> = of water, or ‘form of water’
<E> = of earth, or ‘form of earth’
The values for φ are more or less lengthy descriptions of the changes
undergone. These changes are all said to be due, in one way or another,
to the actions of heating and cooling, either internal or external. Here are
three examples from chapter 6:
• As many as are <E/W> are solidified by fire or cold, and made dense by
both. (383a13-15)32
29
Quoting Burnyeat 1992, 22, who denies that Aristotle sets himself this task; see note 4, above.
30
What I have in mind by this claim is that the statements typically have the following properties: (i)
they are universal in scope; (ii) they are taken by Aristotle to identify natural necessities; and (iii)
they are intentionally worded so as leave the extension of the relationships identified open. This
gives these statements the ‘projectability’ expected of scientific laws.
31
It must be remembered that these are true compounds, however – ‘earth’ and ‘water’ are present
potentially, not actually, as explained earlier.
32
ὅσα δὲ κοινὰ γῆς καὶ ὕδατος, καὶ ὑπὸ πυρὸς πήγνυται καὶ ὑπὸ ψυχροῦ, παχύνεται δ’ ὑπ’ ἀμφοῖν.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
• As many as are <E/w> and solidify by cooling due to the departure of
their internal heat, melt when heat returns. (383a26-29)33
• As many as are <E/w> become solidified by cooling, when all the heat
has evaporated, are indissoluble except by excessive heat, but are soft-
ened. (383a29-31)34
One with a slightly different form and content is:
• Of those solidified by dry heat, some are insoluble, others are water
soluble. (383b9-10)35
The plural quantitative relative pronoun ὅσα is over and over again the
subject of these propositions, leaving their scope open to anything with the
appropriate material nature reacting to the agency of heating and cooling
in the specified way. The behaviors of familiar materials are regularly cited
as examples of things that fall under these causal universals, and may well
have served as the initial inductive evidence for them.36
The causal theory in these chapters eventually issues, in chapter 10, in
a theoretical reconstruction of the ‘constituted bodies.’ At the most general
level, all are either ‘liquid’ or ‘solid,’ softness and hardness being applicable
only to things somewhere along the solid band of the continuum (388a25-
29). As we will see in Section 9.3.5, this classification underlies the classifi-
cation of the uniform parts of animals in PA ii.2, and indeed the items
categorized in Mete. iv.10–11 include most of the uniform biological parts,
discussed along with various man-made and ‘mineralogical’ compounds.
Once ranked into one or the other of these categories, each of the liquids
and solids will be one or the other of the five kinds that can take the subject
position (K) in the ‘laws’ mentioned earlier.
There are two important background principles implicit in the entire
discussion, the first of which is identified explicitly in chapter 10: (i) all the
uniform bodies are of three kinds: forms of water, forms of earth, and
33
ὅσα μὲν οὖν ὑπὸ ψυχροῦ πήγνυται τῶν κοινῶν γῆς καὶ ὕδατος, πλέον δὲ ἐχόντων γῆς, τὰ μὲν τῷ
τὸ θερμὸν ἐξεληλυθέναι πηγνύμενα, ταῦτα τήκεται θερμῷ εἰσιόντος πάλιν τοῦ θερμοῦ.
34
ὅσα δὲ διὰ ψύξιν, καὶ τοῦ θερμοῦ συνεξατμίσαντος ἅπαντος, ταῦτα δὲ ἄλυτα μὴ
ὑπερβαλλούσῃ θερμότητι, ἀλλὰ μαλάττεται.
35
ὅσα δ’ ὑπὸ ξηροῦ θερμοῦ πήγνυται, τὰ μὲν ἄλυτα, τὰ δὲ λυτὰ ὑγρῷ.
36
These two prominent features of the presentation in Mete. iv are also prominent features of the
presentation of information about animal differentiae in HA. For the prominent role of ‘open
scope’ universals in the presentation of information in HA, see Lennox 1991 (repr. Lennox 2001b,
ch. 2). However, there is one crucial difference: the propositions in Mete. iv cited above often
identify causal relationships, while those in HA identify universal correlations without mentioning
causation – and its method is explicitly said to be pre-causal (HA i.6, 491a7-14; PA ii.1, 646a8-12).
See Popa 2014, 320–323 for a discussion of this contrast.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
‘compounds’ (κοινά) (388a25-26).37 The second is stated clearly in chapter
7: (ii) As many as are compounds of earth and water are properly spoken of
“in accordance with the amount of one or the other” (384a3-4). In practice
this means that bodies that are in fact compounds may properly be referred
to as ‘of earth’ or ‘of water’ (or ‘earthen’ or ‘watery’), if their behavior
clearly indicates that they are predominantly one or the other. Moreover, at
the beginning of chapter 8, Aristotle insists that because these compounds
are all produced by the agency of heat or its absence, all will be found
somewhere along the hot/cold spectrum as well as the dry/moist spectrum:
“Since heat and cold are present because they act, and moist and dry
because they are affected, compounds partake of all of them.” (384b28-
30) He also there reminds us that in thinking about the ores and metals we
have to remember the role of the ‘exhalations enclosed in each,’ noted with
a backward reference to the end of Book iii. (Cf. 384b24-34; the produc-
tion of ores and metals by exhalation is discussed at iii.6, 378a15-378b5.)
We can gain further insight into the inquiry into the generation of
uniform bodies by looking at how he deals with a material that initially
appears to violate the ‘laws’ discussed above, such as olive oil. It is discussed
in chapter 7, which is devoted to investigating the conditions under which
<E/w> and <e/W> compounds increase in density and undergo solidifica-
tion. The chapter opens with the following assertion:
Those things that contain more water than earth are thickened by fire alone,
but those that contain more earth are solidified. Hence both soda and salt
consist more of earth, and likewise stone and potter’s clay. (Mete. iv.7,
383b18-20)
However, Aristotle immediately notes that olive oil presents a grave prob-
lem (ἀπορώματα, 383b20-21): according to his principles, if it is an <e/W>
compound, then cold should solidify it; if it is an <E/w> compound, then
fire should solidify it. Olive oil, however, is solidified by neither – rather,
his investigations have established that, while it increases in density both
when heated and when cooled, it is solidified by neither. His solution is to
argue that it is infused with air, and so is an <A/w> compound.38
37
These principles are implicit in chapters 4–7, but not highlighted. Aristotle regularly notes that
compound bodies are combinations of moist and dry, or earth and water, and are called by one or
the other name depending on which one is predominant (cf. 381b24-27, 382a4-6, 383a27-28). But the
language of uniform bodies (τὰ ὁμοιομερῆ σώματα), literally ‘bodies all the parts of which are alike,’
is used for the first time in chapter 8 and is not used frequently until chapters 10–12.
38
383b24-25, 384a15-16, 385b4-5, and 388a32 all seem to say that it is a compound of air and water with
air dominant; 388a8-9 says it is ‘more of moist’ than dry, suggesting there may also be a small amount
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
Initially this appears ad hoc; but Aristotle, like all self-respecting experi-
mentalists, cites a number of lines of independent evidence for the presence
of air and then goes on to argue – with mixed results – that this theory can
account for all of the anomalous behavior of olive oil:
The reason is that it is full of air. Hence too it rises to the surface in water,
since air moves upward. So the cold thickens it by producing water from the
air present in it – for always, when water and olive oil have been combined,
the combination becomes thicker than both. On the other hand, oil is
thickened and whitened by fire and lapse of time – whitened when the water
evaporates (if there was any present in it), and thickened because of water
coming to be from the air when the heat dies away. So, then, the same effect
occurs in both cases, and for the same reason, but not in the same way. Now
while oil is thickened by both, it is dried by neither, since neither the sun nor
the cold weather dries it, not only because it is viscous, but also because it
consists of air. And the water is not dried or boiled by fire, because it is not
vaporized owing to its viscosity. (Mete. iv.7, 383b25-384a2)39
The causal theory that emerges in these chapters is thus complex. The basic
causal processes are operations of heating and cooling, but the source of
heating or cooling may be external while that which is acted upon will not
only be somewhere on the dry/moist continuum, but will also be some-
where on the warm/cold continuum. Chapter 6 distinguishes two forms of
becoming moist, condensation and melting (382b28-29); and he notes that
both things that are predominantly of water or that are compounds of earth
and water can undergo solidification. The variety of observed changes is
accounted for in part by reference to variations in the relative amounts of
the elemental ingredients, and in part by changes in the ‘internal tempera-
ture’ of the compound. Heating of a relatively cool, moist compound
causes evaporation, and this, on Aristotle’s account, produces solidifica-
tion. The dry is, after all, what resists being shaped, the moist (or as we
might say, the fluid) what is easily shaped. On the other hand, if the
internal heat departs from a naturally warm, moist body by its being
cooled, it ‘draws off’ moisture and also leads to evaporation; as a result,
an object becomes less and less fluid, more and more solid. This always
produces a change from a softer to a harder surface, but not always a more
of earth present – and the other passages are worded so that they allow for the presence of earth,
provided the moist, or water, predominates.
39
Interestingly, at GA ii.1, 735b1-37, a number of the same observational claims are cited in establish-
ing that there is pneuma in semen, and at one point (735b13-16) the behavior of olive oil is cited as
evidence.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
dense material. That depends on the amount of moisture in an <e/W>. If there
is excess moisture, its reduction produces compacting (i.e., greater density).
This idea, that when the innate heat of a compound departs, if there is
also moisture present in it, that “moisture is in all cases drawn off either by
the internal or external heat” (πάντα θερμῷ ἢ τῷ ἐντὸς ἢ τῷ ἐκτὸς
συνεξατμίζοντι τὸ ὑγρόν, 382b24-25), reappears often. An implication of
it is that cooling as well as heating a moist body can dry (and thus solidify)
it, since as the heat departs it apparently somehow takes moisture with it.40
9.2.3 The Differential Dispositions of Uniform Bodies

We are now ready to consider the eighteen differential dispositional
continua listed in Mete. iv.8, insofar as they are products of the processes
characterized in these ‘thermodynamic laws.’ Here is the paragraph in
which they are introduced:
Let us first enumerate all those [properties] that are spoken of by capacity
and incapacity (κατὰ δύναμιν καὶ ἀδυναμίαν). They are the following:
solidifiable, unsolidifiable; meltable, unmeltable; softenable, unsoftenable;
absorbent, nonabsorbent; capable or incapable of being bent; breakable,
unbreakable; fragmentable, unfragmentable; capable or incapable of taking
an impression; moldable, unmoldable; squeezable, unsqueezable; elastic,
inelastic; malleable, unmalleable; splittable, unsplittable; cuttable, uncutta-
ble; viscous, crumbly; compressible, incompressible; combustible, incom-
bustible; capable or incapable of emitting fumes. (385a10-18)
Ten of these are explained directly by reference to the causal laws previously
discussed. Indeed, the first two, as he notes, just are the dispositions to
undergo the processes of becoming solid or liquid discussed in the previous
chapters. Solid bodies are softenable by fire (385b6-12) if, though predom-
inantly earthen, they retain some moisture and are either elastic or malle-
able – he cites iron and horn. He discusses meltability and absorbency
together (385b12-20), noting that copper is nonabsorbent, meltable, and
not water-soluble.41 In general, absorbency is a feature of earthen materials
with pores that can admit water, whereas those things are water-soluble
40
This principle of heat ‘drawing’ moisture also underlies the theory of moist and dry exhalations,
which is crucial in his account of the formation of the metals and minerals discussed in book iv; see
iii.6, 378a17-b6.
41
In all these cases, though Aristotle rarely goes into detail, it is clear that in establishing that
a particular uniform body has a particular disposition he is relying on evidence that may be thought
of as experimental: for example, what happens to copper when it is submerged in water or subjected
to intense heat? This in turn leads him to draw inferences about the underlying material nature of
the body, that is, where it is located on the moist/dry (or watery/earthen) continuum.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
that are porous throughout. So too, a body’s impressibility, elasticity,
malleability, and compressibility have to do with the way its surfaces
yield to contact; while a hard and solid body has the capacity or
incapacity to split, divide, or crumble depending on whether its pores
all run lengthwise, widthwise, or alternate. A number of these disposi-
tions – squeezability, ductility, and viscosity – are explained by refer-
ence to a body’s hardness, softness, malleability, porosity, and so on.42
The two exceptions, combustibility and fumability, are reactions of
predominantly dry or moist bodies to intense heat. Unlike the causal
agents of uniform bodies coming to have these dispositional properties,
a number of these are dispositions to react in various ways to physical
pressures such as being squeezed, bent, cracked, compressed, molded,
shaped, and so on.
Though it is occasionally noted that some of the uniform materials
mentioned are parts of animals (e.g., 388a16-20) there is no reference
to the potential biological importance of these dispositional properties
of the uniform parts. This is clearly not part of the project of Mete.
iv, which raises the question of the goals of this inquiry, and the ways
in which those goals dictate what norms govern it. We can, I think,
identify at least five questions that Aristotle is aiming to answer in this
inquiry:
1. How do specific forms of heating and cooling impact and transform
different earth/water compounds?
2. What affective changes do different uniform materials undergo when
heated and cooled in specific ways?
3. What dispositional properties emerge, in which bodies, due to these
affective changes? These dispositions are then used to differentiate one
uniform material from another.
4. What is the relationship between the emergent, dispositional proper-
ties of the different uniform compounds and their material, elemental
constitution?
42
The pores mentioned here, and in the cognate biological passages, are not, by the way, a non-
Aristotelian idea, as has occasionally been claimed (e.g., by Gottschalk 1961) as an argument for the
inauthenticity of Mete. iv. Aristotle’s notion of a uniform body in no way excludes that body having
a porous conformation. The compound may be porous, and if so it will have some other material in
its pores; but Aristotle is extremely careful to stress that what is inside the pores is no part of the
compound. The theory of pores he is arguing against in GC i.8 (326b8-29) is not the theory
defended here. On this topic, see the thorough discussion in Düring 1944, 74–78; cf. Furley,
73–93; Lee 1962, xiii–xxi; Lewis 1996, 3–15; Pepe 1978, 503–523; and the discussion in Popa 2014,
6–12 and note 26.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
9.3 Putting Dispositions to Work 247
5. What are the relationships among these dispositional potentials – that
is, which are primary, which secondary, which come packaged
together, and which do not?
Why, we might ask, would Aristotle expend so much care and effort trying
to answer these questions? One way to answer that question is to see what
use, if any, he makes of the results of this investigation in the biological
works. As I will argue in the following sections of this chapter, the
dispositions of uniform parts to act and react in various ways is a central
feature of Aristotle’s concept of conditional necessity, a distinctive explana-
tory concept in his biological practice. Thus, the results of the inquiry
reported in Mete. iv play a critical role in Aristotle’s biological program;
those results account for the emergence of the dispositional properties that
uniform parts must have if they, and the organs they constitute, are to
perform the biological functions necessary for living things to survive and
flourish.
9.3 Putting Dispositions to Work: Mete. iv

and Aristotle’s Biology
The final chapter of Mete. iv makes it clear that at least one of this book’s
functions is to serve, in the words of David Furley, as a “prolegomenon to
biology.” As Gill 2014 points out, the closing sentences of Mete. iv.12
(390b14-22) explicitly point forward to his biological studies. That chapter
opens with very similar language in order to clarify what Mete. iv has – and
has not – accomplished up to that point:
Since we have determined these things, let us state individually (καθ᾽
ἕκαστον) what flesh, bone, and each of the other uniform bodies are. For
we get hold of the things from which the nature of the uniform stuffs (ἡ τῶν
ὁμοιομερῶν φύσις) is constituted, their kinds (τὰ γένη αὐτῶν), and to what
kind each belongs, through their generation. For from the elements are
constituted the uniform bodies, and from these as matter are constituted
whole works of nature (τὰ ὅλα ἔργα τῆς φύσεως). (389b23-28)43
43
Cf. Gill 2014, 339–344. This passage tells against the claim of Mirus 2006 that the differential
dispositional properties discussed in Mete. iv provide true, nonbiological definitions of the uniform
bodies. Aristotle is quite clear here that we have yet to determine what each of these uniform bodies
is, though we have learned, by studying how they come to be, the material kinds to which they
belong. The language of “getting hold” (ἔχομεν) of these bodies by grasping their kinds is evocative
of APo. ii.8, where we come to know that there is, say, thunder, by coming to know that there is
a certain noise in the clouds, or that there is man by coming to know that there is a certain sort of
animal (93a20-24).
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
Both of these passages make it clear that the eleven preceding chapters have
accomplished something, but also stress that, concerning uniform mater-
ials that are parts of “the whole works of nature,”44 something more needs
to be done, if our goal is to understand what each of those uniform parts is.
In the previous two sections, I have described what has been accomplished,
accomplishments that are nicely summarized in the following passage in
chapter 10. The previous chapters have provided the differentiae by which
we can distinguish the uniform bodies from one another. These bodies fall
into two broad categories – things that can be extracted from the earth and
their by-products, and those found in plants and animals, out of which
their nonuniform parts are made.
By these affections and by these differences, as we have said, the uniform
bodies differ from one another according to touch, and further, by tastes,
smells, and colors. I mean by uniform bodies, for instance, things that are
mined – copper, gold, silver, tin, iron, stone, and other such things – and
those things that come to be separated out of these; and things in animals
and plants, for instance, kinds of flesh, kinds of bone, sinew, skin, viscera,
hair, fibers, veins, from which the nonuniform bodies are constituted – for
example, face, hand, foot, and other such things; and in plants, wood, bark,
leaf, root, and all such things.45 Since the nonuniform bodies are constituted
by another cause, whereas the matter from which these are constituted is the
dry and moist, and therefore water and earth (for each of these has the very
conspicuous potency of one or the other of those), while the productive
potencies are heat and cold (for heat and cold constitute and solidify bodies
from the dry and moist), let us grasp which sorts of uniform bodies are forms
of earth, which of water, and which are compounds. (Mete. iv.10,
388a10-26)
44
While certainly open to dispute, I take this reference to “whole works of nature” to be to animals
(and perhaps plants), and that this provides the context for the rest of the chapter. Mirus (2006) puts
much stress on the fact that, amidst a long discussion of how animals and their parts are what they
are in virtue of their capacities to perform living activities, Aristotle once mentions, in conjunction
with the parts of plants, inorganic metals (bronze and silver, 390a17-18). But (i) this actually
reinforces the point that the chapter is primarily about animals, since he stresses that he is extending
the point he is making about animals to these cases; they are “just as flesh and sinew,” and (ii) the
point he is making here is generalizable – “all these things are what they are by a capacity of acting or
being acted upon” (390a18-19). That does not imply that all capacities are the same, and it allows that
some things may be defined primarily by passive capacities. It is also worth noting that when
Aristotle talks about the processes of heating and cooling being able to produce the affections of
uniform bodies, but not nonuniform parts like a hand or foot, he goes on to make an analogous
point about bronze or silver – heating and cooling may be a cause of their generation, but “not of the
generation of a saw, a cup, or a box: in this case art [is the cause], whereas in the other case nature or
some other cause” (390b12-13).
45
Recall the virtually identical list of parts at Metaph. Ε.1, 1025b30-1026a5, to make the point that the
natural scientist studies both the matter and form of natural objects.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
He here alludes to “another cause” by which the nonuniform parts (organs)
are constituted, the very issue he turns to in chapter 12.46 But nothing more
is said about that other cause here: rather, he goes on to note that the
matter of uniform things is moist and dry – but since each of these powers
is most manifest in water and earth, he often substitutes the latter pair for
the former – and the productive causes are the activities of hot and cold.
Having established all that, we are now ready to say which of these uniform
bodies are forms of earth, which are forms of water, and which are
compounds. In doing so, he pays no attention to whether they are animate
or inanimate – for example, iron, bone, sinew, and bark are grouped
together as mixtures in which earth predominates (389a11-13). The focus
is on inferring their material makeup from the way they react to various
kinds of agents, especially heating and cooling.
From this perspective, understanding them ‘as matter,’ focusing on
dispositional contraries such as hardness/softness, meltability/non-
meltability, elasticity/rigidity, and so on, it may turn out that earthenware
and bone are the same kind of uniform material. But as Aristotle goes on to
say in chapter 12: “Everything is from the things previously discussed as
from matter, but is in formula as according to substantial being (389b28-
29).”47 What does Aristotle have in mind by this unusual turn of phrase?
One possibility, which will be reinforced by looking at the biological
works, is that an account of the formal nature of a uniform part of an
animal requires identifying its functional contribution to the life of the
organism of which it is part.
9.3.1 The Explanatory Dependence of PA ii on Mete. iv

PA ii provides the context for understanding this twofold nature of
uniform parts by opening with a bottom-up ‘composition’ of the organ-
ism. Rather than starting with the ‘so-called’ elements – earth, water, air,
fire – Aristotle recommends starting with four powers or potencies
(δυνάμεις) that constitute the elements, which he here calls “the matter of
the composite bodies.”48
46
I think this is the most plausible way to read this text; for an alternative reading, see Mirus 2006.
47
One of those Aristotelian sentences where an attempt to capture the nuances of the original in
readable English is a challenge: Ἔστιν δ’ ἅπαντα ὡς μὲν ἐξ ὕλης ἐκ τῶν εἰρημένων, ὡς δὲ κατ’ οὐσίαν
τῷ λόγῳ.
48
The assertion that these four contrary powers are the matter of bodies (presumably in virtue of being
‘that out of which’ the ‘so-called elements’ are constituted) is repeated again in On Length and
Shortness of Life (Long.) 5, 466a20-22. In Mete. iv.1, 378b31-379a3 and 10, 388a20-26, only moist and
dry are identified as matter, in virtue of their being what is acted upon by hot and cold.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
That is to say, moist, dry, hot, and cold are matter of the composite bodies,
while the other differences, for example, heavy and light, dense and rare,
rough and smooth, and the other such bodily affections, follow from
(ἀκολουθεῖν) these ones. (PA ii.1, 646a16-20)
The four primary powers are here said to be the matter of composite bodies,
and other powers are said to ‘follow’ them. Follow, in what sense, one
might ask? Mete. iv, as we saw, provides the answer: these emergent
dispositional properties are due to a variety of interactions between pro-
cesses initiated by heat and cold and material compounds that fall on the
moist/dry, watery/earthen continuum.49
After presenting this ‘bottom-up’ compositional hierarchy, Aristotle
immediately contrasts the way things are ordered in generation (ἐπὶ τῆς
γενέσεως) with the way they are ordered in substantial being (τῆς οὐσίας):
“for things posterior in generation are prior in nature, and the final stage in
generation is primary” (646a25-27). The implication for animals and their
parts is clear: the living being, which is the end product of generation, is
metaphysically primary.
Hence, as Mete. iv.12 leads us to expect, PA ii.1 also ranks dispositional
properties based on the teleological priorities inherent in the nature of
living things.
And since the actions and motions present in animals as a whole and in such
[nonuniform] parts are complex, the things from which they are composed
must have distinct potentials; softness is useful for some things, hardness for
others; certain things must be elastic, others flexible. Now in the uniform
parts such potentials are distributed part by part (one is soft, another hard,
one moist, another dry, one viscous, another brittle), while in the nonuni-
form parts they are distributed to many and in combination with one
another; for a different potential is useful to the hand for squeezing than
for grasping. That is why the instrumental parts have been constituted from
bones, sinews, flesh, and the other parts of this character, not the latter from
the former. (646b14-27)50
49
An anonymous referee commenting on Lennox 2014 noted the oddity of heaviness and lightness
being mentioned here. They are not among the list of emergent dispositions in Mete. iv.8, and Cael.
iv defines heavy and light in terms of a natural capacity to move either toward the center or the
periphery of the cosmos. The two elements that possess natural heat (fire and air) are the ones that
tend upward, so Aristotle may have in mind that the presence of heat is essential to lightness – but
I am not aware of any passage where he actually says that.
50
The Greek of the opening phrase here is critical: Πολυμόρφων δὲ τῶν πράξεων καὶ τῶν κινήσεων
ὑπαρχουσῶν τοῖς ζῴοις ὅλοις τε καὶ τοῖς μορίοις τοῖς τοιούτοις, ἀναγκαῖον ἐξ ὧν σύγκεινται, τὰς
δυνάμεις ἀνομοίας ἔχειν.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
Moist and dry are here said to be present in the uniform parts, and when
Aristotle later makes his initial division of the uniform parts (647b10-18) it
is based on some being moist (but also soft), and others dry (though also
hard and solid), in agreement with the generic classification of uniform
compounds of Mete. iv.10.51 Thus, while new dispositional properties
emerge at the level of these uniform parts, old ones may, but need not,
be lost or suppressed. Softness and hardness, viscosity and brittleness, on
the other hand, while closely associated with moist and dry, are, Aristotle
claims, only found at the level of uniform compounds.52
The message in this passage is that if a hand is to be a hand in more than
name, it must be able to perform a complicated variety of actions –
squeezing and grasping, releasing – it is an instrument of instruments.53
This is the ‘must’ of conditional necessity – if a hand is to perform its
complex variety of functions, it must have fingers, fingers constituted of
a number of different uniform parts – bones, sinews, flesh, nails – each
with different dispositional properties and arranged in just the right way.
In explicating his notion of conditional necessity in PA i.1, he uses a craft
analogy to make this very point.
This is, as it were, conditionally necessary: for just as, since the axe must
split, it is a necessity that it be hard, and if hard, then made of bronze or iron;
so too, since the body is an instrument (for each of its parts is for the sake of
something, and likewise also the whole) it is therefore a necessity that it be of
such a character and constituted from such things, if that is to be. (642a9-13)
Whether illustrated by axes or hands – and this point is seldom commented
on – it is noteworthy that the necessitation goes from entity (axe) to
function (splitting) to specific differential disposition (hardness), and only
then to some uniform part or other with that disposition (bronze, iron).
These differential dispositions are thus located at a philosophically
crucial juncture – they (rather than the uniform parts per se) are required
by (i.e., are conditionally necessary for) the functional nature of the
instrumental parts; while they somehow ‘follow’ those primary powers
(hot, cold, moist, and dry) that characterize the material nature of the
uniform parts.
51
See Popa 2014, 306–334.
52
In modern chemistry ‘compound’ refers to the product of a chemical bonding of different elements,
which gives rise to properties not present in the elements, while ‘mixture’ refers to a physical
intermingling of different elements or compounds that retain many of the properties of the elements.
Given those options, ‘compound’ is closer to what Aristotle has in mind by μῖξις than ‘mixture’ is.
53
The account in PA iv.10, 687a3-b25 includes a good deal of discussion of how well constructed the
hand is for grasping and squeezing.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
Nevertheless, Aristotle does not put aside the investigation of the
material processes that are necessary steps (albeit conditionally) in the
process of generation.
As being for the sake of something, then – on account of this cause – (ὡς μὲν
οὖν ἕνεκά τινος διὰ ταύτην τὴν αἰτίαν) these parts [uniform and nonuni-
form] are related in the way stated; but when one also seeks how it is
necessary that they be thus, it is apparent that they were antecedently so
related to one another from necessity. (PA ii.1, 646b27-30)
The debt of the investigation of the uniform parts in PA ii to Mete. iv is
explicitly acknowledged once and obvious throughout. That inquiry
begins with a lengthy preliminary exploration of the many ways in which
the terms hot, cold, moist, and dry are used, an exploration replete with
vocabulary familiar to us from Mete. iv.54
The hot seems both to solidify (πηγνύναι) and soften (τήκειν). Thus cold
solidifies as many things as (ὅσα) consist only of water, while fire solidifies
those consisting of earth; and among hot things, as many as (ὅσα) are more
earthen solidify quickly by means of cold and are insoluable (ἀλύτως), while
as many as (ὅσα) are watery can be dissolved (λυτῶς). What sorts of things
are capable of solidification, and the causes owing to which they are solidi-
fied, have been determined more clearly elsewhere. (PA ii.2, 649a29-34)
No less than four of the ‘material-level laws’ identified in Section 9.2.2 are
repeated here and are used as material-level starting points in the accounts that
follow of the uniform parts. The last sentence in the quoted passage is likely
referring us to the inquiry reported in Mete. iv – at least, that is the only place
in the corpus where things capable of solidification and its causes are discussed.
The dependence of the PA ii account of the uniform parts of animals on Mete.
iv is also attested by the way these laws are expressed – the repeated use of the
quantitative relative pronoun ὅσα; the precise vocabulary used for the effects of
heating and cooling; and the contrast between bodies that are “more of earth”
or “more of water.” Aristotle assumes that these causal dependencies have been
established elsewhere – there is no attempt to defend them here. Here he
assumes their truth and makes extensive use of them in attempting to explain
various features of the uniform parts, as I will now briefly illustrate.
Blood is, not surprisingly, given its fundamental role in his physiology,
the first uniform part taken up for specific examination. Aristotle begins by
discussing its material makeup (649b20-650a2) and then turning to its
biological function. A discussion of the concoction of other nutriments
54
These were already said to be “the matter of the composite bodies” at 646a17.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
into blood (650a2-32) leads to the conclusion that blood is the final
nutriment for blooded animals (650a34), and therefore that it is present
in animals for the sake of nourishing the other parts (650b2-13). Chapter 4
then discusses materially different kinds of blood and how these material
differences are responsible for a number of other differences. It begins with
what appears to be a summary of the material analysis of Mete. iv: 55
Some blood has what are called fibers, but some, such as that of deer and
gazelles, does not. It is for just this reason that this sort of blood does not
solidify (οὐ πήγνυται) – the part of the blood that is watery is colder, which
is also why it does not solidify; while the part of the blood that is earthen
solidifies when the moisture evaporates (πήγνυται συνεξατμίζοντος τοῦ
ὑγροῦ), and the fibers are made of earth. (650b14-18)56
While the material differences in the constitution of blood of different
animals, discussed in Mete. iv, are repeated here, the primary focus of PA
ii.4 is on the importance of this difference in blood, something never
mentioned in Mete. iv. Differences in the material composition of blood
appear to underlie differences in the characters of animals.57 Thick, warm
blood is productive of strength and spirit; thinner and cooler blood is
productive of greater perceptiveness and intelligence; warm, thin, and clear
blood leads to a combination of courage and cunning, and so on (647b33-
648a13; 650b19-651a19). He concludes:
The nature of the blood is the cause of many features of animals with respect
to both character and intelligence, as is reasonable, since blood is the matter
of the entire body; for nourishment is matter, and blood is the last stage of
nourishment. It therefore makes a great difference whether it is hot or cold,
thin or thick, turbid or pure. (PA ii.4, 651a12-17)
The material differences in blood are characterized as productive of various
emotional, perceptual, and intellectual differences, though no attempt is made
to argue that these differences are also present for the sake of these character
differences.58 There is no reason to doubt that Aristotle thought they might
be; but it is clear that he felt he could offer a material-level account of these
differences independently of providing a parallel teleological explanation.59
55 56
Cf. 237–238, above. Cp. HA iii.6, 515b27-516a7; iii.19, 520b10ff.
57
Cf. de An. i.1, 403a24-b9; and for systematic discussion of this aspect of Aristotle’s natural
philosophy and its relevance for his ethics and politics, see Leunissen 2017.
58
Cf. Freudenthal 1995, 48–56; Laberrière 1984, 405–428; Leunissen 2012, 507–530.
59
Cf. Lennox 2001a, 187, 188, and 201–203; Leunissen 2012, 517–520 and 2017. As we will see shortly,
many of the explanations of pathêmata of organic parts in GA v have the same character.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
But it is not only differences in character that are dependent on
differences of blood. Hard and soft fat (i.e., suet and lard) are the end
result of the concoction of the excess blood of a well-nourished organism,
and which one is produced is accounted for by reference to the material
nature of the blood that undergoes concoction (651a20-36; cf. HA iii.19,
521a18; GA i.19, 727a32-36). Aristotle uses the complex method of Mete.
iv to establish which sort of blood, fibrous or serous, underlies these
differences in fat (651a26-36), and cites the fact that animals with horns
and hoofs – which are independent evidence of an animal’s earthen
nature – also have suet, while those with no horns and no hoofs have
lard (651a30-36).
Moreover, a number of explanations throughout PA ii–iv depend on
this one.60 As one example: the kind of omentum an animal has is said to
be a direct product of the kind of blood from which it is concocted.
The omentum is a membrane; and in animals with suet it is hard, while in
those with lard it is soft; and what sort of thing each of these [suet and lard]
is has been stated previously. (PA iv.3, 677b14-16)
The generation of this structure, like many others, is said “to happen from
necessity due to this sort of cause” (677b21-22, b29-30),61 though nature
makes use of it in aid of the concoction of nutrients. The animal’s nature
does that because the omentum has the appropriate material dispositions:
. . . in order that the animals may concoct their nutriment easily and
quickly; for that which is hot is able to concoct, and that which is fat is
hot, and the omentum is fat. (677b31-33)62
Likewise, some animals have lard around their kidneys, others suet,
again depending on the nature of the blood that is filtered. That there
is fat around the kidneys is due to the blood in the kidneys being
further concocted there (PA iii.9, 672a1-9) – whether it is lard or suet
is dependent on the material differences of the blood in each case
(672a10-12).
60
Gotthelf 1987b 167–198, esp. 172–178 [reprinted in Gotthelf 2012, 153–185, esp. 158–164] makes
a compelling case that PA ii–iv is structured in such a way that the premises needed for later
explanations are established earlier in the treatise.
61
This is one of many examples of what Leunissen refers to as ‘secondary teleology’. Cf. iii.7, 670a30-
32, iii.9, 672a14, iv.12, 694a22-23, 694b5; cf. Leunissen 2010, 95–97, 140–141.
62
As Popa 2014 stresses, these material-level explanations often have a transparently syllogistic
character.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
At times PA ii uses the theories of Mete. iv to explain the nature of
a uniform part that is never explicitly discussed there; for example, the
brain.
That the brain is a combination of water and earth (κοίνὸς ὕδατος καὶ γῆς)
is clear from the following facts about it: boiling it makes it dry and hard,
and with the evaporation of the water by heat the earthen material remains.
It is just like what happens with the boiled mashes produced from legumes
and other fruits once the moisture mixed in them departs, because they are,
for the most part, constituted of earth; that is, these too become completely
hard and earthen. (PA ii.7, 653a20-27)
Here, two different results from Mete. iv are in play. Boiling is one of the
three forms of concoction there discussed. Its distinctive feature is that it is
a process of concocting by external moist heat (380b13-16). It can have its
‘setting’ effect only on bodies with moisture capable of being acted on by
moist heat (380b27-28). It does so by the heat of the surrounding moisture
extracting the body’s own moisture from it (380b20-24). PA ii’s discussion
of the brain uses these ‘material-level laws’ of Mete. iv as a basis for
inferring the brain’s elemental constitution.63 But there is far more discus-
sion of why this material constitution provides the appropriate disposi-
tional properties to enable the brain to perform the function for which it is
present in blooded animals.
On account of this cause nature has devised the brain in relation to the
heart’s location and heat. And it is for the sake of this that this part, with the
combined nature of earth and water, is present in animals. (PA ii.7,
652b19-23)
For if it becomes overly moistened or dried it will not perform its function,
and either will not cool the blood or will solidify it, so as to produce diseases,
mental derangements, and death. (653b3-5)
It is, then, the animal’s nature that has devised a structure with the
emergent dispositional properties appropriate for its role in relationship
to the heart, with its opposing dispositions. They are the dispositions of an
earth/water compound, as Aristotle establishes by attending to how it
behaves when boiled.
The brain’s material constitution is also appealed to in explaining
certain facts about development. Aristotle explains the development of
the cranium around the brain by reference to heat evaporating the moisture
63
Note again that the technical vocabulary is that of Mete. iv: the verb ἐξατμίζω for the evaporation
process, and the use of κοινός with the genitives ὕδατος and γῆς for the compound nature of
uniform bodies.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
on the brain’s surface. As the human brain is large in bulk, thus containing
a great deal of moisture, this process occurs only slowly. This accounts for
the softness at birth of the human cranium (653a32-36; cf. GA ii.6,
744a24-32).
These explanations presuppose an already established theory of the
constitution of uniform bodies that accounts for them having the material
dispositions they do. These material dispositions (solidity, flexibility, melt-
ability) in turn allow the uniform parts to play the roles they do, including
constituting the nonuniform parts so they can perform their functions. But
Aristotle’s prior understanding of these material dispositions also gives him
the resources, as in the discussion of the brain, to infer the underlying
material nature of a uniform part. The dependence of the discussion of the
uniform parts of animals in PA on the inquiry reported in Mete. iv is
extensive and multidimensional – and it is not the only biological inquiry
where we find such dependence.
9.3.2 Mete. iv and the Developmental Thermodynamics of GA

Perhaps the most sustained and extensive use of the explanatory patterns
developed in Mete. iv to be found in Aristotle’s biology is in GA ii.6 and
GA v. Having, in GA ii.4, established that the heart is the first organ to
come to be, and that the blood vessels containing blood produced by the
heart grow out from it to provide nourishment for the production of the
other parts, in GA ii.6 Aristotle begins to discuss the formation of the
uniform parts in precisely the correct order, place, and arrangement to
constitute properly structured and organized nonuniform parts. Compare
the following passage with those we considered earlier in Mete. iv.
The blood vessels that stretch themselves out from the heart are just like the
skeletal figures drawn on walls;64 for the parts, inasmuch as they have come
to be from the blood vessels, are [formed] around them. The generation of the
uniform parts is by means of cold and heat; for some are constituted and
solidified by cold, while others are by heat. The differences among them have
been stated previously in other places; which ones are soluble by cold and which
by fire, and which ones are neither soluble by cold nor meltable by fire.65 So as
the nutrient oozes through the blood vessels and through the channels in
each just like the water in unbaked earthenware,66 fleshes or their analogue
64
For other comparisons of embryogenesis to figure drawing, cf. HA iii.6, 515a35; PA ii.9, 654b29-32;
GA iv.1, 764b30-31.
65
As Peck notes (Peck 1942, 219 note e); cf. Düring 1944, 83. This must be a reference to Mete. iv.
66
Cf. Mete. iv.3, 380b8-11.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
come to be, being constituted by the cold, wherefore they can also be dissolved by
fire. But as much of the nutrient as is welling up is exceedingly earthen, since
it has but little moisture and warmth, and undergoes cooling when the
moisture evaporates off with the heat and becomes hard and earthen in character
(τὴν μορφήν), for example, nails, horns, hoofs, beaks. This is why these
become malleable by fire, though none can be melted; some, however, can be
dissolved by moist things, such as the shells of eggs. The sinews and bones
come to be by means of the internal warmth through the drying out of the
moisture. And this is why bones, just like earthenware, cannot be dissolved by
means of fire, for it is like being baked in an oven by the generative heat. (GA
ii.6, 743a1-21; emphasis added)
This is not simply an example of material-level explanation of the forma-
tion of uniform parts; as the phrases I have italicized demonstrate, these
explanations make a direct appeal to those ‘thermodynamic laws’ formulated
in Mete. iv that we discussed in Section 9.2.2.67 But throughout the
discussion in GA ii.6, Aristotle stops periodically to remind the reader
that these changes are taking place in an amazingly complex and orches-
trated manner, which demonstrates that the nature or soul of the animal is
somehow in control of the heating and cooling that is taking place.
Nature uses both heating and cooling, having a power from necessity so that it
makes one thing to be this and another to be that, so that in things that are
coming to be the occurrence in the one case of cooling and in the other of
heating is for the sake of something, and each of the parts comes to be, flesh soft,
sinew dry and elastic, bone dry and brittle, each being made such in one respect
from necessity and in another respect for the sake of something. (743a36-b5)
All these things, just as we said, must be declared to come to be in one way
from necessity but in another not from necessity but for the sake of something.
(743b16-18; emphasis added)
We see the material-level explanations, in precisely the form they were
presented in Mete. iv, here being mobilized for explanatory purposes in
embryology. Yet the animal’s generative soul is using these material causal
processes for the sake of producing a fully functional organism. This is the
two-level explanatory strategy defended theoretically in Mete. iv.12 (and
more fully in PA i.1) put to work in embryological inquiry. And it again
adds support to the view that, while in organic nature Aristotle thinks
explanation by appeal to all four causes is the norm, he is open to
67
As a reminder, by using the term ‘law’ here I only mean to stress that these are universal propositions
identifying necessary causal relations between forms of heating and cooling, on the one hand, and
the changes they bring about, on the other – it is not a historical claim about Aristotle originating
the notion of a ‘law of nature.’ For a discussion of this issue, see Popa 2014.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
explanations that only appeal to the material level or to the formal level.
And as we are about to see, even in the organic realm, material-level
necessitation may be the full explanation of certain phenomena.
GA v presents the results of inquiry into the coming to be of the
παθήματα by which the parts of animals differ ‘by more and less’
(778a16-18).68 In such cases, we needn’t think of the cause in the same
way as before (778a30-32). If a feature is not universally present in a kind, it
may neither have come to be nor be for the sake of something (a31-33). In
certain cases, the feature in question may not extend (συντείνω) to the
account of the animal’s being (λόγος τῆς οὐσίας); rather, it may (merely)
come to be from necessity, so that we must refer back to the matter and the
moving source for our understanding (778a34-b1).
Among those which are not such <i.e., not for something>, but of which
there is a generation, it is necessary to seek the cause in the movement and
generation itself, grasping the difference in the composition itself. For the animal
will have an eye from necessity (for being such an animal is posited), while it
will have this sort of eye from necessity as well, though not the former sort of
necessity, but another sort – because its nature is to act or be acted on in this
or that way. (GA v.1, 778b13-19; emphasis added)
The contrast is between a part that must be there as part of what it is to be
the animal in question, and an affection of a part that arises during
development because of the active and passive capacities of the materials
out of which the part is constituted and which may even undergo changes
after birth. These latter features may well not come to be for the sake of
something – they are not conditionally necessary, but are materially
necessitated.69
This claim at first appears to fly in the face of everything APo. tells us
about scientific explanation. These features (say baldness or a high-pitched
voice in human beings) are not universally present in the kinds of animals
that have them, nor proper to specific kinds, and they may even be
transient – nor are they, either directly or indirectly, ‘proven through the
68
For a detailed analysis of the explanatory project of GA v and its place in the overall project of GA,
see Gotthelf and Leunissen 2012 (in Gotthelf 2012, 117–141).
69
Cf. PA i.1, 640a33-640b4 on the distinct forms of explanation referenced here. There is no
implication that none of the affections discussed in this book are for the sake of something. The
last chapter of this book (GA v.8), for example, discusses differences in whether animals shed a first
set of teeth that are later replaced, and Aristotle strongly criticizes Democritus both for providing
too general an explanation for this phenomenon (788b9-19) and for failing to investigate what it is
for the sake of (789b2-15). On this, and GA v generally, see Gotthelf and Leunissen, in Gotthelf 2012,
117–141; on precisely what is and is not implied by the opening discussion (778a29-b19) see the
appendix to their chapter, 139–141.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
essence.’ In what sense, then, are they the proper subjects of a scientific
inquiry?
The answer lies in the fact that there are things that happen during
organic development that are due to material necessity, and where the
universals appealed to are those found in Mete. iv. The kinds and essences
appealed to are those of the materials that constitute the uniform parts –
the basic active and passive potentials and the active and passive processes
that must take place when certain kinds of materials undergo certain kinds
of heating and cooling. The discussion of variations in the hair of animals
will serve as an example. The discussion begins as follows:
For the sake of what nature made hair as a kind in animals, that we discussed
previously in our work on the causes of the parts of animals. It is for the
present study to make clear what affections result and on account of what
necessities each of these <differences> results. (GA v.3, 782a20-24)
The variations to be considered are of the sort explained in Mete. iv (of
density, rarity, thinness, thickness); the material causal factors invoked are
the density of the skin (out of which hair is produced), whether its moisture
is of a greasy or watery nature, the size of the skin’s pores, or the relative
balance of earth and water in the skin’s nature. The following discussion of
differences in hair brought on by cold climates shows a thorough indebt-
edness to Mete. iv.
Cold climates have opposite effects on sheep and on humans. Thus, the
Scythians have soft hair, while the Sarmatian sheep have hard hair. The
cause of the latter fact is the same as in all wild animals. The coldness
hardens because the drying process hardens things; for when heat is expelled the
moisture evaporates with it, and the hair and skin both become earthen and
hard. (GA v.3, 783a12-18; emphasis added)70
Later in the discussion a somewhat tentative explanation for curly hair
appeals to the impact of a dry environment. Hair, Aristotle explains, is
mostly composed of dry, earthen material with little moisture present “so
that the hair contracts when dried out by the environment. For the straight
bends if it undergoes evaporation, and shrinks up just like hair burned by
fire . . .” (782b25-28, emphasis added).
70
As a reminder of the relevant conclusions at work here: “Therefore since everything definite and
constituted is either soft or hard, and these are due to solidification, no composite and definite body
can exist without solidification” (Mete. iv.5, 382a24-27). “All things are dried either by being heated
or by being cooled, but in both cases by heat – that is, by internal or external heat. For even things
that are dried by cooling . . . are dried by the internal heat that causes the moisture to evaporate
along with it, the heat being driven off by the surrounding cold” (iv.5, 382b16-22).
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
These characters are often variable within a kind because they are not
functionally important to the life of that kind of animal – nevertheless,
there is a common causal explanation for (e.g.) curly hair wherever it
occurs, at the material level. As we have seen, the theory and vocabulary
of these material-level explanations once more mirror the details of the
account of the effects on different kinds of uniform materials of processes
such as evaporation secondary to loss of internal heat in Mete. iv. No
further explanation for these affective differences by reference to a goal is
provided, though as in PA ii, that does not imply that there could not be
one. What it does show is that Aristotle is willing to offer such material-
level, causal explanations in the absence of teleology. Moreover, the pro-
logue to GA v demonstrates that in at least some cases the explanation
derived from material necessitation will be the full explanation. But as our
study of GA ii.6 indicates, material-level explanations often occur as one
part of the teleological explanation of a feature coming to be for the sake of
an end, where the animal’s nutritive/generative soul controls the timing,
location, and intensity of the heating and cooling so that just the right
uniform parts needed to constitute (say) the leg of a horse arise at the right
time, in the right order, and in the right relationships to one another.
9.3.3 Mete. iv.12 and GA ii.1 in Conflict?

Against all this evidence of a tight dependence of Aristotle’s biology on
Mete. iv, it is occasionally argued that there is a clear inconsistency between
a passage in Mete. iv.12 (390b3-14) and one in GA ii.1 (734b24-735a4).71
Initially, the former passage seems to say that uniform parts can come to be
simply by the actions of hot and cold, while the nonuniform parts require
“nature or some other cause” (cf. 388a21); while the latter passage from GA
ii.1 clearly claims that flesh and bone have biological functions just as
nonuniform parts do, and thus require an appeal to a formal/final explan-
ation just as much as face and hand do.
This conflict is, however, only apparent. Recall that the explanation of
the generation of the uniform parts of the blooded animals in GA ii.6 relies
heavily both on material-level principles from Mete. iv and on a formal
nature that is ensuring that heating and cooling are being applied at just the
right times, and in just the right places, to just the right degree, and to just
the right materials so that a functioning organism one-in-form with its
71
For example, Cohen 1989. The apparent inconsistency is also discussed in Gill 2014, who like me sees
no conflict, though she resolves it in a different way than I defend here.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
parents comes to be.72 Aristotle does not say that heating and cooling
always produce the requisite dispositional properties in earthen/watery
compounds for the sake of something – they produce such properties for
the sake of something only when these processes are directed by an animal’s
nature. GA ii.6, then, makes it prima facie unlikely that there is any actual
conflict between the accounts of Mete. and GA. And as Gill has shown, the
argument of Mete. iv.12 makes it equally implausible, for the argument in
that chapter both begins and ends with passages that tell us that Mete. iv
has not given us a full account of what it is to be any of the uniform parts, at
least not one that takes account of their distinctive being (389b23-28).73
Thus, if 390b3-14 were taken to be claiming the opposite, we would not
have to go to GA for a contradiction – it would be staring us in the face in
Mete. iv itself. Indeed, both Mete. iv and GA ii would be internally
inconsistent.
The solution is clear if we focus on what Aristotle says in Mete. iv.12
about nonuniform parts: that not even what answers to a material-level
account of, say, a hand can be generated by heat and cold alone, though
things answering to the material-level account of flesh and bone, in terms of
their emergent dispositional properties, can. That is, putting aside the
functions for which hands and arms come to be, mere heating and cooling
acting on moist/dry compounds “without a logos” are simply insufficient
for the production of anything with such structure and complexity. The
material nature of bone is not unlike that of various rocks and earthenware;
the material nature of a hand or arm, however, involving the formation of
a number of different uniform parts into a precisely coordinated system, is
quite another matter. Similarly, the flesh and bone of such a system are not
defined solely by the dispositional properties discussed in Mete. iv.1–11:
they are defined more by their functional role in that system.74
72
Thus, after explaining how bone is formed by the action of heat like potter’s clay baked in an oven,
he adds: “But this heat does not make any chance thing flesh or bone nor in any chance place nor at
any chance time; rather [it makes] that which is naturally so, where it is naturally so, and when it is
naturally so” (GA ii.6, 743a21-22; trans. after Gotthelf 2012, 104).
73
See Gill 2014.
74
Cf. PA ii.9, 654a32-b6, esp.: “a bone on its own is nothing; rather, it is a part either as part of
something continuous or through contact and binding, in order that nature may use it both as one
and continuous and, for bending, as two and divided. . . . And indeed, if any bone were separate, it
would not perform the function for the sake of which bones exist (for if it were not continuous but
disconnected, it would not be a cause either of bending or of straightening).” This is fully consistent
with the following claim: “Such parts, then, can come to be by heat and coldness and the motions
caused by these, since they are solidified by heat and by cold. I mean those that are uniform, for
instance, flesh, bone, hair, sinew, and such things; for all these differ by the differentiae previously
mentioned – tension, elasticity, fragmentability, hardness, softness, and other such properties.
These things are generated by heat and cold and by their combined motions” (Mete. iv.12,
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
This understanding of the passage coheres well with the closing lines of
Mete. iv.12, with which Gill opens her paper.
So if we have a hold (ἔχομεν) of what kind each of the uniform bodies is, we
need to grasp individually what each one is (e.g., what blood is, or flesh or
semen, and each of the others). For we know in each case on account of
what, and what, it is (διὰ τί καὶ τί ἐστιν), if we have a hold (ἔχομεν) either of
the matter or of the account (ἢ τὴν ὕλην ἢ τὸν λὸγον), and especially when
we have both <the matter and the account> of its generation and destruc-
tion, and also whence the source of the motion. When these things have
been clarified, we must likewise investigate the nonuniform bodies, and
finally the things composed of these, such as man, plant, and other such
things. (390b14-22)
Thus, one can attain a certain grasp of the uniform bodies by grasping their
matter and the motive source of their generation and destruction. This is
the understanding described in the first sentence here, and in the first
paragraph of the chapter, as “getting hold of the kind to which each of
uniform parts belong . . . through their generation” (cf. 389b25-28).
According to the first paragraph of this concluding chapter, that is what
the previous eleven chapters have accomplished. The above concluding
paragraph tells us that, while this is not everything, any more than a formal
account alone would be, it is one way of grasping the reason why and what
it is (διὰ τί καὶ τί ἐστιν) of flesh or bone, grasping the dispositional
properties that make them the kind of uniform materials they are, and,
as his biological investigations will show, make them suitable to play
specific biological roles as well. Given what a complex and difficult task
it can be to get this far in a scientific investigation, Aristotle is willing to say,
in concluding the Meteorology, that we actually achieve, as he puts it at the
beginning of chapter 12, a sort of grasp of those things from which the
nature of the uniform materials has been constituted and their kinds (ἐξ ὧν
ἡ τῶν ὁμοιομερῶν φύσις συνέστηκεν,75 τὰ γένη αὐτῶν) in such a study –
and we have done so by focusing on their coming into being (διὰ τῆς
γενέσεως) (389b23-26). GA ii.1 adds to this that what makes flesh flesh and
bone bone is not only these material-level dispositions, but also the bio-
logical functions of flesh and bone – functions that they cannot perform, of
course, unless they have the emergent dispositional properties that they
390b2-9). The dispositions that differentiate bone from the other uniform materials are brought
about by the actions of heating and cooling and the associated motions discussed in Mete. iv.1–11,
but not in such a way as to organize a properly functioning skeleton.
75
Cf. 388a21-22.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
9.4 Conclusion 263
have in virtue of the actions of heating and cooling acting on the sorts of
material compounds that they are.76
9.4 Conclusion
In this chapter and the previous one, my goal has been to highlight the
complex interdependencies among Aristotle’s many and varied inquiries
into the natural world. These dependencies point to a preferred, if not
necessary, order of inquiry; and while this order of inquiry is likely related
to the order that Aristotle might have considered best from a pedagogical
point of view, or the order dictated by his views regarding explanatory
priority, it is distinct from either of those orderings. What is revealed in the
two case studies that make up this chapter and Chapter 8 is that progress in
certain aspects of inquiry into one domain of nature (e.g., into the
directionality of the motions of the heavenly bodies, or the material nature
of the uniform parts of animals) depends on having already reached at least
preliminary conclusions in another domain. This need not be reflected in
the pedagogical order, where questions such as what is clearer to the novice
versus the expert and what can be treated as locally unsupported assump-
tions for the purposes of teaching will play an important role. Similarly,
given Aristotle’s view that, in the study of living things, explanatory appeals
to form (living capacities and activities) and end (what various parts and
processes are for) take priority over efficient causes based in the material
natures of living things, the dependence of PA ii and GA ii and v on the
results of the explorations reported in Mete. iv will not necessarily be
reflected in the explanatory priorities of those treatises. But what I hope has
become clear in these chapters is that the norms of inquiry that we’ve been
exploring extend to norms about the order in which inquiries need to be
carried out.
76
Some of the material in this section of Chapter 9 is based on Lennox 2014d, which in turn was based
on a presentation at an HSS/HOPOS symposium in November 2012, organized by Tiberiu Popa –
which also included the papers by Popa and Mary Louise Gill on which I have relied in this
discussion. The translations of texts from Mete. iv in this chapter are based on one prepared by the
three of us for publication.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
chapter 10
Aristotle on Respiration
Framework Norms Meet Domain-Specific Norms
One should explain in the following way, for example, breathing

exists for the sake of this, while that comes to be from necessity
because of these. But ‘necessity’ sometimes signifies that if that –
that is, that for the sake of which – is to be, it is necessary for these
things to obtain, while at other times it signifies that things are thus in
respect of their character or nature. For it is necessary for the hot to go
out and enter again upon meeting resistance, and for the air to flow in.
This is directly necessary; and it is as the internal heat retreats during
the cooling of the external air that inhalation and exhalation occur.
This then is the manner of methodical inquiry, and it is concerning
these things and things of this sort that one should grasp the causes.
(PA i.1, 642a31-b4)1
Chapter Summary. The principal distinction at the core of this book’s

argument, between erotetic framework norms and domain-specific norms,
gives rise to the following questions: What role do these distinct kinds of
norms play in a particular inquiry, and how do they interact? If the erotetic
framework norms are operative in any domain of truth-oriented inquiry,
then even if a specific inquiry has its own domain-specific norms, that
inquiry should also be governed by those framework norms. How are we to
understand the respective roles of these distinctive kinds of norms of
inquiry? Moreover, we have only tangentially addressed the issue of what
counts as a ‘domain’ – as we have seen, De caelo, Meteorology, and the
various biological treatises are all contributions to the science of nature and
in that respect are governed by norms common to all natural inquiries; but
1
Δεικτέον δ’ οὕτως, οἷον ὅτι ἔστι μὲν ἡ ἀναπνοὴ τουδὶ χάριν, τοῦτο δὲ γίγνεται διὰ τάδε ἐξ ἀνάγκης.
ἡ δ’ ἀνάγκη ὁτὲ μὲν σημαίνει ὅτι εἰ ἐκεῖνο ἔσται τὸ οὗ ἕνεκα, ταῦτα ἀνάγκη ἐστὶν ἔχειν, ὁτὲ δ’ ὅτι
ἔστιν οὕτως ἔχοντα καὶ πεφυκότα. τὸ θερμὸν γὰρ ἀναγκαῖον ἐξιέναι καὶ πάλιν εἰσιέναι ἀντικροῦον,
τὸν δ’ ἀέρα εἰσρεῖν. τοῦτο δ’ ἤδη ἀναγκαῖόν ἐστιν. τοῦ ἐντὸς δὲ θερμοῦ ἀντικόπτοντος, ἐν τῇ ψύξει
τοῦ θύραθεν ἀέρος ἡ εἴσοδος καὶ ἡ ἔξοδος. ὁ μὲν οὖν τρόπος οὗτος ὁ τῆς μεθόδου, καὶ περὶ ὧν δεῖ
λαβεῖν τὰς αἰτίας, ταῦτα καὶ τοιαῦτά ἐστιν.
264
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
10.1 Introduction: From On Inquiry to On Respiration 265
we have also seen that the study of the heavens is very different from that
of earthbound living things and is explicitly said to be governed by
distinctive norms. This concluding chapter is devoted to exploring
these questions in detail; and in order to do so, I’ve selected a limited
inquiry within the Parva naturalia, Aristotle’s inquiry into the nature and
causes of respiration. Since the publication of the Prussian Academy
edition of Aristotle’s corpus, under the editorship of Immanuel Bekker,
the pages devoted to this inquiry have been treated as a distinct treatise.
As we will see, this is a mistake, and it is a misleading one in a number of
respects. But as we will also see, there are hints, beginning from the
opening page of the PN, that Aristotle sees it as a distinctive inquiry and
one that poses difficult methodological questions of an especially intri-
guing kind.2
10.1 Introduction: From On Inquiry to On Respiration

APo. ii opens by telling us that the things about which we inquire are equal
in number to those about which we have scientific knowledge, and that we
inquire about four things: the ‘that,’ the ‘reason why,’ the ‘if it is,’ and the
‘what it is’ (89b23-25). There are four different kinds of questions that can
be asked about any subject of inquiry, and Aristotle sees these questions as
paired and sequential: the paradigmatic ‘that’ is a predicative relationship
between a subject and attribute, and the paradigmatic ‘reason why’ is some
more fundamental fact about the world that explains that predicative
relationship; the paradigmatic answer to an ‘if it is’ question is an affirm-
ation of something’s existence, and once affirmed we will naturally seek to
know what it is. But as I argued in Chapter 2, APo. ii.2 paints a more
‘interwoven’ picture of these inquiries:
But we are inquiring, when we inquire into the ‘that’ or the unqualified ‘if it
is,’ whether there is a middle for it or not; while when, having come to know
either the ‘that’ or ‘if it is,’ either partially or without qualification, we
inquire into the reason why or the ‘what-it-is,’ we are then inquiring what
the middle is. (89b37-90a1)
As a reminder that will help us understand what is going on in the inquiry
into respiration: this passage intentionally blurs the distinctions so clearly
drawn in APo. ii.1 in four respects.
2
And see PA iii.6, 669a4-5, which refers to what was said ἐν τοῖς περὶ ἀναπνοῆς; cf. de An. ii.8, 421a6;
iii.9, 432b10-11.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
266 Aristotle on Respiration
1. Aristotle downplays the distinction between the two preliminary ‘that it
is’ and ‘if it is’ inquiries; the former is now described as a partial (ἐπὶ
μέρους) version of the latter (a possibility already hinted at in chapter 1
at 89b33).
2. Moreover, both these preliminary inquiries are seen as attempts to find
out whether there is a middle – that is, though preliminary, both are
guided by the goal of demonstrative knowledge.
3. The differences between the second two inquiries, between ‘what it is’
and ‘why it is’ inquiries, also are downplayed; both are now described
as attempts to discover what that middle is!3 Thus, while chapter 1
might have suggested that the search for a middle term would be
appropriate only if one were searching for the causal explanation of
a predication that had been established, but not if one was searching for
the essence of some kind of thing, in chapter 2 we are told that in all
our inquiries we are asking either whether there is a middle or what the
middle is: “for the middle is the cause, and in all cases this is what is
sought” (90a5-6).
4. And finally, this last-quoted statement amounts to a more profound
blurring of the distinction between a preliminary, purely fact- or
existence-establishing stage of inquiry and a later stage of inquiry in
search of causes – those preliminary inquiries now seem to be prelim-
inary stages in the search for causes. This once again points to the view
that even preliminary inquiries are aimed at the goal of unqualified
scientific knowledge.4
Two features of APo. ii.2, then, complicate the account given in APo. ii.1:
While Aristotle has not abandoned his claim that there are four distinct
inquiries, he is nevertheless claiming that even as we inquire whether some
kind of thing exists or whether a subject has some attribute or other, we are
involved in some sort of causal, or anyway explanatory, inquiry. Moreover,
to assert that in either of these cases we are seeking whether there is a middle
would seem to imply that the existential ‘if it is’ inquiry and the predicative
‘that it is’ inquiry are, at the very least, intimately related. And finally, in
saying that all such inquiries are causal, he is suggesting that, in seeking
3
Cf. 90a14-15: “In all these cases it is apparent that ‘what it is’ and ‘why it is’ are the same.”
4
And it is this view that accounts for a number of otherwise puzzling aspects of APo., among these the
fact that he is willing to discuss ‘partial demonstrations’ (i.24); to distinguish demonstrations of the
fact from demonstrations of the reason why (i.13); and to argue that in at least some cases a causal
demonstration of the ‘reason why’ can be reconfigured as an account of what the subject of inquiry is
(ii.10). At one point in ii.8, we are told that in order to grasp that something is non-incidentally, we
need to grasp something of the thing itself (93a22-28).
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
10.2 Framework Norms in the Study of Breathing 267
answers to “What is it?” questions we are also engaged in trying to answer
questions about why the kind of object being investigated has the attributes
it has. I turn now to Aristotle’s investigation of respiration to explore this
interweaving of the four inquiries in actual practice.
10.2 Framework Norms in the Study of Breathing

Chapter 1 of On Respiration5 begins with an introduction that is highly
critical of investigations of the subject by previous thinkers, some of whom
claim all animals breathe.
A few of the earlier natural philosophers have spoken about respiration; yet
for what purpose (τίνος . . . χάριν) it belongs to animals, some have said
nothing, while others, though they have spoken, have done so poorly, due to
lack of experience with the facts (ἀπειροτέρως τῶν συμβαινόντων). Further
they say all animals breathe, but this is untrue. So, it is necessary first to
review their claims, in order that we not appear to level an empty accusation
in their absence. (470b6-12)
This is a passage we will return to in Section 10.3. For now, I want to draw
attention to the complicated nature of Aristotle’s complaints here. The first
has to do with teleology: those of his predecessors who have discussed
respiration either did not consider the question of what it is for or they did
so badly. Why? Because they were inexperienced with the facts. But which
facts, exactly – what sort of experience were they lacking? This complaint is
fleshed out a bit more precisely at the end of chapter 3, after a lengthy
criticism of the views of Anaxagoras, Democritus, and Diogenes:
The most significant reason for [these thinkers] not discussing these things
well is both (τε) their being inexperienced with the internal parts and (και)̀
not grasping that nature makes them all for the sake of something; for
seeking what respiration belongs to animals for, and examining this ques-
tion in the presence of the parts (e.g., in the presence of gills and lungs), they
would have discovered the cause more quickly. (Juv. 9/Resp. 3, 471b23-29)
Here the nature of the inexperience is spelled out: it is inexperience of the
relevant internal parts – experience of which would require skilled
5
I am using the title On Respiration simply for convenience’s sake, since it does delimit the inquiry on
which we are focused. Ross (1955, 2–3) presents the evidence against treating this as a separate work.
The opening lines of Juv. 1, 467b10-13 show that this inquiry is continuous with that into the causes of
youth and old age and the later inquiry into the causes of life and death. See too Hannon 2011, 5–14
and Morel 2000, 51. In Section 10.3, I discuss one especially telling piece of textual evidence for this
continuity.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
dissection of many different kinds of animals. But Aristotle is insisting that
gaining the relevant experience of the internal parts needs to take place as
a preliminary step in an inquiry into ‘that for the sake of which,’ a special
kind of causal inquiry. Anaxagoras, Democritus, and Diogenes were
unsuccessful in their inquiry because they failed to study the internal
organs involved in respiration while asking what respiration is for.
Moreover, determining what respiration is for will at the same time provide
us with an account of what respiration is (i.e., with its proper definition) –
and as a bonus prize we will have learned what parts such as gills, lungs, and
windpipes are for! This is the sort of recommendation for scientific inquiry
we should expect from the author of APo. ii 2.
To put this in Analytics terms, Aristotle’s predecessors in the study of
respiration dealt with that causal question poorly because they had not
done a good job with the factual inquiry. He follows this up with a second
complaint: that they mistakenly claim that all animals breathe. He then
begins an extended critical review of his predecessors by setting the record
straight on that point: “It is apparent that as many animals as have a lung,
all these breathe” (470b12-13). A simple statement of a highly general,
universal fact, indeed one that appears at the beginning of the discussion
of the internal parts of blooded animals in HA ii.15. It will turn out to be
of critical importance for the entire inquiry. For as it happens, not all
animals have lungs – and of equal importance, not even all blooded
animals do.
Notice that the ‘subject’ of this universal fact is not a ‘natural kind.’ As
with so many of the correlations specified in the HA, the subject is
identified in such a way as to leave its extension open: “as many animals
as have lungs.” So, if it turns out, for example, that there are animals with
lungs that live their entire lives in the sea, such as whales, dolphins, and
porpoises, you can infer, surprisingly, that they breathe. But, and this is
Aristotle’s immediate point here,6 fish do not have lungs and thus do not
breathe – and this fact poses the fundamental puzzle around which the
6
He takes up the cetaceans in Juv. 18/Resp. 12, which opens: “Among the water-dwellers, one might be
puzzled about the cetaceans, that is about the dolphins, whales, and as many as have the so-called
blowhole – but in fact these are in accordance with our account” (476b13-16). The problem someone
might raise, he says, is that these creatures are footless, take in sea water, yet have lungs (b16-17). He
assures the reader that they do, nonetheless, manage to cool themselves by breathing and are thus
κατὰ λόγον (b14), in accord with the account he has been giving. Note again the open extension
characterization of the subject – the relevant group is ‘as many as have a blow hole.’ Aristotle wants to
restrict the term κητώδη to a group of sea creatures united by a system of related similarities, but he
does so in a way that allows for the discovery of other forms of cetacean that have not yet been
identified.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
10.2 Framework Norms in the Study of Breathing 269
entire inquiry revolves. The complex interplay between existential inquiry
(are there lunged sea creatures?), factual inquiry (do all blooded animals
breathe, or have lungs?), causal inquiry (what is the purpose of respiration,
or lungs?), and essence inquiry (what is respiration?) is already on display in
the opening lines of this inquiry.
Before following these interwoven inquiries as they play out, however,
I want to flag a concern one might have about whether such an inquiry
would be in tension with some well-known descriptions of inquiry in the
biological works. For, while the characterization given above of the inves-
tigation of respiration sits comfortably with the interwoven nature of
inquiry that I argued, back in Chapter 2, was the message of APo. ii.2, it
might seem to conflict with a well-known passage in HA i.6 that suggests
a more discrete boundary between these two stages of inquiry, akin to the
initial description of the four inquiries in APo. ii.1. Here is the problematic
passage:
These things have now been stated in outline, in order to provide a taste of
which things, and about which things, we need to study in order that we
may first grasp in all cases the differences that are present and the attributes.
(We will speak with greater precision later.) After this we must attempt to
discover their causes. For this is the natural way (κατὰ φύσιν) to carry out
the methodical inquiry (τὴν μέθοδην), after the factual research (ἱστορίας)
about each thing is in place; for it is from these things that both the things
about which there needs to be demonstration and the things from which
there needs to be a demonstration become apparent. (HA i.6, 491a7-14)
As has been noted,7 there is a remarkable similarity in the mode of
expression here to the closing lines of Prior Analytics i.30, where ἱστορία
again refers to being in possession of the facts that truly hold about each of
the things being studied, which then prepares an investigator to find the
demonstration of what is by nature demonstrable (46a22-27). The same
distinction is referred to at PA ii.1 (646a8-11); and (as we discussed in
Chapter 8)8 IA 1 concludes its review of the questions about animal
locomotion to be taken up in its inquiry in language that clearly suggests
the same distinction:
About all these things, and others akin to them, we must study the causes.
For that (ὅτι μὲν) these things happen to be so is clear from the factual
7
Kullmann 1974, 81 n. 8; Lennox 1987, 102 n. 26 (reprinted in Lennox 2001b, 35 n. 25); Lennox 1991,
267–270 (reprinted in Lennox 2001b, 43–45).
8
See page 208, note 16.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
research (ἱστορίας), but the reason why (διότι δέ) we must now investigate.
(704b8-11)
The possible tension between these passages and the method one finds in
Resp. arises because in them the preliminary ‘fact gathering/establishing’
stage of inquiry appears not only to be preliminary to a search for causes,
but to be completed independently of the search for causes. It mirrors, in
practice, the very same tension that was noted in Chapter 2 between the
way the four modes of inquiry are characterized in APo. ii.1 and in APo.
ii.2. In the remainder of this chapter, we will take a close look at the
inquiry into respiration with this tension in mind. Doing so will give us
a fresh perspective on how to understand the recommendations of HA i.6
and APo. i.30 and how to integrate them with the more ‘interwoven’
account of inquiry that emerges from the later chapters of APo. ii.
10.3 Domain-Specific Norms: Where Aristotle’s Predecessors Went

Wrong
I noted earlier that it is misleading, in informative ways, to think of 470b5-
480b31 as a separate treatise entitled On Respiration. We can begin to get
a sense of why this is so by looking at the conclusion of On Length and
Shortness of Life, in which Aristotle looks forward to the topics ‘common to
body and soul’ remaining to be covered:
It remains for us to study youth, old age, life, and death; for having defined these
things, our methodical inquiry concerning the animals would have achieved its
end (τέλος ἂν ἡ περὶ τῶν ζῴων ἔχοι μέθοδος). (Long. 6, 467b6-9)
Respiration, it will be noted, is not mentioned. And while the lines that
immediately follow these, traditionally marked as the opening lines of On
Youth and Old Age, do mention respiration, it is almost as an
afterthought:
We must now discuss youth, old age, life, and death; and presumably at the
same time it is also necessary to discuss the causes of respiration; for in some
animals living and not living are dependent on this. (Juv. 1, 467b10-13;
emphasis added)
As he puts the point in a later criticism of one of his predecessors that we
will look at shortly (p. 273): “And yet surely we see these things to be
sovereign over (κύρια) life and death; for when animals that breathe are
unable to do so, at that moment they begin to pass away” (Juv. 11/Resp. 5,
472b27-29).
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
10.3 Domain-Specific Norms 271
This much is an unquestionable datum of common sense, something
familiar to us: animals that breathe, if prevented from doing so for any
length of time, die. This gives the researcher good reason to investigate it
from a teleological perspective – it clearly is doing something upon which
life depends; but what, precisely? Aristotle’s eventual answer is that it helps
to moderate the heat in the region around the heart. Thus, the investiga-
tion into the question of why those animals that breathe do so is trans-
formed gradually into an investigation of methods of cooling across the
animal kingdom, of the need for a variety of methods of cooling, and of the
importance of effective cooling for the preservation of life. That discussion,
in turn, will depend heavily on the central topic of Juv., the intimate
relationship between the nutritive soul and internal, natural, or psychic
heat – and the physical changes in old age that lead to natural heat no
longer serving as an effective agent of nutritive soul.9
All of this suggests that Resp. is a discussion embedded within these
wider concerns. Indeed, when one considers what are, according to the
Prussian Academy edition, the closing lines of De juventute et senectute with
the opening lines of De respiratione, one realizes just how embedded in the
wider investigation it is! Juv. closes by setting up the coming discussion of
respiration with the following words:
Since some of the animals are water-dwellers, while others pass their time in
the air, and they accomplish the task of cooling from and by means of these
(i.e., some by water and some by air), we must discuss in what manner and
how they do this, by becoming better acquainted with the debate about it
(τὸν λόγον). (Juv. 6, 470b1-5)
While the opening words of Resp., cited earlier, are:
For (γὰρ) a few of the natural philosophers have spoken previously about
respiration; yet for what purpose (τίνος . . . χάριν) it belongs to animals,
some have said nothing, while others, though they have spoken, have done
so poorly, due to lack of experience with the facts (ἀπειροτέρως τῶν
συμβαινόντων).10 Further they say all animals breathe, but this is untrue.
So, it is necessary first to review their claims, in order that we not appear to
level an empty accusation in their absence. (Juv. 7/Resp. 1, 470b6-12)
The γὰρ not only makes it clear that these remarks are continuous with the
previous sentence, as Ross notes; it also clarifies the otherwise somewhat
9
As an aside, this is another example of the methodological importance of order of inquiry and the
dependence of one inquiry on the results of another, the focus of the previous two chapters.
10
These two criticisms, as we’ll soon see, correspond to two important normative recommendations
Aristotle makes at the end of Resp. 3.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
puzzling reference to becoming acquainted with a logos at the (alleged)
close of Juv. The sentence that opens Resp. explains which logos it is about
which we must become better acquainted. There is an ongoing discussion –
a logos – about what respiration is, which animals breathe and how they do
so, a debate into which Aristotle is entering, and we need to become better
acquainted with that logos before moving forward.
Ross is certainly right to conclude, then, that the evidence supports the
conclusion that the discussion from the beginning of Juv. on is one
continuous discussion, though with clearly demarcated subsections.11
Moreover, the introduction to De sensu, which is more accurately described
as the introduction to our Parva naturalia, lists a sequence of studies to
come, of attributes common to body and soul, a sequence that begins with
an investigation of perception and concludes with an investigation of four
paired attributes: sleeping and waking, youth and old age, breathing in and
out, life and death (436a1-17). It is noteworthy that in this preview of
coming attractions, the discussion of life and death is listed where it
actually appears, after a long discussion of respiration (and other forms of
cooling).12
Nevertheless, it is worthwhile considering Aristotle’s discussion of res-
piration on its own – he does distinguish it from the related topics, and it
contains a rich discussion of how inquiries into nature can go wrong, and
what you need to do to stay on track in the study of a biologically
important process and the organs that accomplish it.
It will come as no surprise that a recurring theme of his review of his
predecessors’ accounts of respiration is their failure to ask “What is respir-
ation for?,” since he mentions this as a common failing in the opening lines
of Resp. quoted above; and in the focal text with which we began in Section
10.2, Aristotle concludes a review of the errors of Democritus, Diogenes,
and Anaxagoras by pointing out their failure to study respiration with this
very question in mind. That is not by any means the last time he takes his
predecessors to task for ignoring the ‘what for’ question in their attempts to
understand respiration. In chapter 4, for example, he laments:
Democritus says that something results from respiration for those that
breathe, when he claims that it prevents the soul being squeezed out; yet
he said nothing about nature doing this for the sake of something, for generally,
11
See Ross 1955, 2–3.
12
It should also be noted that On Length and Shortness of Life is not mentioned, though Aristotle
claims to be mentioning only ‘the most important’ (τὰ μέγιστα, 436a13) pairs of topics.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
10.3 Domain-Specific Norms 273
just like the other natural investigators, he does not touch at all on this sort
of cause.13 (Juv. 10/Resp. 4, 471b30-472a3)
And in chapter 5, he criticizes the account in the Timaeus in similar terms:
Again, those14 who discuss the matter in this way have said nothing about
what these things (by which I mean both inhaling and exhaling) are present
in animals for the sake of (τὸ τίνος ἕνεκα ταῦθ’ ὑπάρχει τοῖς ζῴοις), but
make proclamations as if speaking about something merely coincidental.
And yet surely we see these things to be sovereign over life and death; for
when animals that breathe are unable to do so, at that moment they begin to
pass away. (Juv. 11/Resp. 5, 472b24-29)
Empedocles, not surprisingly given passages in Ph. ii.8 and PA i.1 we
have had reason to discuss previously, is given the same treatment in
chapter 7.
Empedocles also speaks about breathing, yet certainly does not discuss what
it is for the sake of; nor does he make it clear, about all the animals, whether
they breathe or do not breathe. (Juv. 13/Resp. 7, 473a15-17)
But why would these investigators carry out their investigations with
this question in mind? If they were of the conviction that what these
parts do as part of the animal’s ‘machinery’ has nothing at all to do with
why these parts come to be, they would consider an investigation
governed by this question wrong-headed. Aristotle insists that, had
they asked the teleological question while investigating the parts, they
would have discovered the cause quickly. But to start investigating with
that question in mind, you must at least be open to the idea that
answering that question is relevant to understanding whatever it is you
are investigating. The claim that those who were not looking for the
final cause of respiration would have found it sooner had they decided to
look for it is trivially true, assuming that there is one. Is this all Aristotle is
saying? Decidedly not: His message is that you will not succeed in
figuring breathing out at all, you will not even get the material or
mechanical story right, if your investigation is not guided by questions
regarding what the process – and the parts required for it – are for. Or so
I aim to show.
13
The Loeb translator, W. S. Hett, makes a pair of glaring missteps here, translating: “Democritus
states that respiration serves a certain purpose in animals that respire . . . but he never says that this is
why nature evolved respiration.” This comes close to stating the exact opposite to the view Aristotle
attributes to Democritus and suggests that Aristotle is looking for an evolutionary alternative!
14
An earlier comment makes the reference clear: “The circulation written about in the Timaeus”
(Juv.11/Resp. 5, 472b6).
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
By the end of chapter 7, he has completed his consideration of previous
accounts of respiration:
So then, such and so many are the troubling difficulties (δυσχερείας) to be
found in the things said by others about breathing. (Juv. 13/Resp. 7,
474a23-24)
10.4 The Need for Comparative Anatomy: The Function

of the Heart
As we have seen, throughout the critical discussion Aristotle has repeatedly
stressed that almost all previous thinkers ignored the question of what
respiration is for. And the one (unidentified) author who does consider the
question of function suggests breath is a sort of fuel for the internal heat.
He now begins his own account:
We have said previously that living and the disposition of soul is conjoint
with a certain warmth; for concoction, through which nutrition comes
about in animals, occurs neither in the absence of soul nor in the absence
of warmth; for everything works (ἐργάζεται) by means of fire. (474a25-8; cf.
τὸ ἐργαζόμενον, de An. ii.4, 416a13)
The reference to what was said previously is almost certainly to the
discussion of the role of the heart in nutrition in chapters 4–6 of
Juv., which immediately precedes his investigation of respiration.15
He first reminds us of the reasons why the nutritive soul and the
primary nutritive part must be in the middle region of the body
(474a28-b1; cf. Juv. 2, 468a13-18). That primary nutritive part, he
asserts, is nameless in bloodless animals but is the heart in blooded
animals (474b1-3). He then recapitulates the argument from Juv. 3–4
for this claim:
The nourishment from which the parts come to be in animals is the nature
of the blood. And there must be the same source (ἀρχήν) of both the blood
and the blood vessels, for the latter are for the sake of the former, as [its]
vessel and receptacle. (474b3-7: cf. PA iii. 5, 667b17-18)16
He concludes:
15
Hannon 2011, 75, titles this chapter “Reiteration of the First Principles of the Theory.”
16
Note, by the way, the claim that one part is present for the sake of another part. This is a form of
causal dependence explicitly defended in PA i.5: “if some actions are in fact prior to, and the end of,
others, it will be the same way with each of the parts whose actions are of this sort” (645b30-32).
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
10.4 The Need for Comparative Anatomy 275
But the heart is the source of the blood vessels in blooded animals, for they
do not run through it, but all in fact extend from it. And this is clear to us
from the dissections. (474b7-9)17
This is the first reference to dissections in Resp., and is a textbook example
of how dissection, if done with a teleological focus, can reveal answers
about function. What is the ‘this’ that is clear from ‘the dissections’? It is
the fact that the blood vessels are the receptacle for blood and that the heart
is their origin. Why would this claim need support from dissections?
Because of a point Aristotle makes regularly in arguing for the heart’s
function: it is by a comparative anatomical study of many different internal
organs and their connections to the blood vessels in many different
blooded animals that you learn that the relationship of the blood vessels
to the heart is unique and unlike their connections to any other visceral
organ.
PA iii.4–5 return over and over again to the fact that dissection will
reveal to you that, while the blood in all the other organs is within blood
vessels, and that in all other organs blood vessels permeate the organ, in the
case of the heart the blood is stored in the heart’s own cavities, and the aorta
and vena cava grow out of those cavities. Readers of PN had already been
referred to this background in the discussion of youth and old age:
That the heart is the source of the blood vessels has been said previously in
our works on the parts of animals (ἐν τοῖς περὶ τὰ μέρη τῶν ζῴων). (Juv. 3,
468b31-469a1)
The PA iii discussion of the heart begins by stressing the differences
between the heart and the other viscera in relation to the blood vessels
(I’ve placed in italics all of the references to what would be seen by means of
dissection):
And as we said [at 665b14-16], while the blood vessels run through the other
viscera, no blood vessel extends through the heart; whence it is also clear that the
heart is a part, in fact an origin, of the blood vessels. And this is reasonable;
for the middle of the heart is a body which is naturally dense and hollow; and
further, it is full of blood, inasmuch as the blood vessels originate there; it is
hollow to serve as the receptacle for blood, and dense in order to protect the
source of heat. For in this part, alone of the viscera and of the body, is there
blood without blood vessels, while each of the other parts has blood in its blood
vessels. This is also reasonable; for the blood is conducted from the heart and
into the blood vessels, but not to the heart from elsewhere; for this is the
17
HA iii.2–4 quotes extensively from earlier thinkers’ accounts of the blood vessels and discusses in
detail problems in observing the vascular system via dissection.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
origin and spring of blood or its first receptacle. These things are more
manifest (κατάδηλα μᾶλλον) with the help of the dissections and the
generations; for the heart, which comes to be first of all the parts, is immediately
blooded. (PA iii.4, 665b31-666a11)
The highlighted clause is added as explication of what specifically the study
of generation adds by way of evidence; two additional points you learn by
studying animal generation closely are that the heart is constituted directly
from blood and that it is the very first organ to appear.
Wherefore the heart becomes apparent first as a distinct part in all blooded
animals; for it is the source in all blooded animals of the uniform and
nonuniform parts. . . . And this is clear (δῆλον δὲ τοῦτο) from the histories
and the dissections. (GA ii.4, 740a17-24, selections; cf. Juv. 25/Resp. 20,
480a6-7)
It is worth spending some time over the references to dissections in these
passages. Aristotle has discovered that, in all blooded animals, the blood
vessels permeate the visceral organs, and he argues that this is because those
organs are constituted out of blood, as their ‘final nutrient’;18 and in
support of that, dissection reveals that as the vessels get smaller and smaller,
they virtually merge with that which they feed. As he explains during his
discussion of the blood vessels in PA iii.5:
. . . just as during irrigations the largest of the trenches remain, while the
smallest are first quickly obliterated by the mud, but when it is removed they
once again become evident, so in the very same way the largest of the blood
vessels remain, while the smallest become in actuality flesh, though poten-
tially they are blood vessels no less. For this reason too, when the flesh is in
any respect preserved, blood flows when it is cut; and though without blood
vessel there is no blood, yet no blood vessel is manifest, just as in aqueducts
the trenches are not manifest until the mud has been removed. The blood
vessels always proceed from the greater to the lesser, until the channels have
become smaller than the thickness of the blood. (668a27-b3)
As we’ve seen, Juv. 3 makes a clear reference to this discussion as Aristotle
begins his argument for the primacy of the heart, and the middle part of the
body, as the seat of nutritive, locomotive, and sensitive soul.19 He concludes:
So it is necessary for the source (ἀρχήν) of the perceptive and nutritive soul
in the blooded animals to be in the heart; for the functions (ἔργα) of the
18
“It is, then, apparent from these and like considerations that blood is present in blooded animals for
the sake of nourishment” (PA ii.3, 650b2-3).
19
See the discussion of 468b31-469a1 on the previous page.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
10.4 The Need for Comparative Anatomy 277
other parts concerned with nourishment are for the sake of (χάριν) its
function; for the authoritative part should strive toward that for the sake
of which, as a doctor is related to health, and should not be found in those
things for the sake of this. (Juv. 3, 469a5-10)
On the basis of what must have been a quite extensive and systematic series
of dissections, Aristotle has apparently noticed the following patterns
emerging across the entire range of blooded animals:
• In each case, the blood vessels are very large at the heart and get smaller
and smaller as they reach the viscera.
• The vessels essentially permeate and disappear in all the visceral organs
except the heart, while in the case of the heart, they seem to be
extensions of its hollow cavities, which contain blood.
• The material of the blood vessels is similar to that of the heart but not to
that of any of the other organs.
• In contrast to the heart, the blood in the other viscera is always inside the
blood vessels that permeate those organs.
• During embryogenesis, the heart comes to be before the other viscera,
and the blood vessels seem to grow out from the heart, after which the
viscera gradually form at the extremities of systems of these vessels.
Only by doing careful, comparative dissection across a wide range of
animals will you become confident that this essential difference in the
way the blood vessels are connected to the heart in comparison to the way
they are connected to the other viscera is true of blooded animals qua
blooded. But it is Aristotle’s view, as we have seen, that an inquiry
involving systematic comparative dissections should be guided, from the
beginning, by teleological questions about these differences between the
heart and the other viscera. Indeed, it is likely that iterated functional
questions of this kind would guide every step of the investigation. In the
end, he concludes that all of these differences are present because in all
blooded animals the heart is the source of the nutritive blood for all the
other visceral parts.
That was a long digression away from respiration, I realize, but it
provides us with two different pieces of background that we need in
order to deal with his investigation of it. First, the texts we have been
looking at in PA and GA give us a rich idea of how Aristotle draws
conclusions about biological function by using dissection to carefully
study anatomical similarities and differences between parts, and the
anatomical connections among them. Second, it will turn out that it is
how structures are related to the heart (related both ‘physically’ and
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
‘extensionally’), against the background of what Aristotle feels he has
already established about the function of the heart and how it accomplishes
that function, that will be central to Aristotle’s account of respiration.
Based on background in de An. ii.4, reviewed quickly in Juv., Aristotle
begins his investigation of respiration already confident that the hearts of
all blooded land animals are a source of heat that is used by the nutritive
soul in preparing blood for distribution to the rest of the body as its
nourishment. And that critical role played by the heart, and the need for
psychic heat in playing that role, underlies the need for respiration.
10.5 Framework Norms Again: Getting the Extensions Right

Before proceeding, it is worth recalling the account provided in Chapter 2
(Section 2.3 in particular) of the erotetic framework for causal search in
APo. ii.14–18. In those chapters Aristotle presents a domain-neutral set of
questions that will move an inquiry from an early stage of empirical inquiry
where we have a grasp of partial universals (isosceles triangles have 2 R,
hawks have curved beaks, grape vines lose their leaves) to the stage where
we have a grasp of ‘primary universals,’ predications that are commensur-
ately universal (triangles have 2 R, raptors have curved beaks, broad-leafed
plants lose their leaves). As I noted then, using the leaf loss example that
figures prominently in APo. ii.16–17, once you get to that stage of inquiry,
a new question naturally arises: Why do all broad-leafed trees lose their
leaves? Is that simply a primitive, immediate connection, or is there
something more basic about these trees that is responsible for this com-
mensurately universal truth? That question might well lead an investigator
to search for what else happens to broad-leafed trees at that time of year
when they lose their leaves. That inquiry might lead to the discovery that
the fluid nutrient coagulates, perhaps due to the cold, at that time of year.
That is, it might lead to the discovery of the efficient cause of leaf loss in
such trees.
That domain-neutral framework discussion in APo. ii concludes, in
chapters 17–18, with Aristotle raising the question of whether it is always
going to be the case that the cause, and that of which it is the cause, are
commensurate. But what is missing entirely is a discussion of the methods
of inquiry that one uses, once you have identified relationships at the
commensurate level of universality, for discovering whether you have
identified a causal connection and, if not, how you go about doing so. In
the case in point, once one has established the commensurate connection
between broad leaves and leaf loss, how does one discover that “the sap
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
10.5 Framework Norms Again 279
solidifying in the connection with the seed” is both the cause of leaf-
shedding and what leaf-shedding is (APo. ii.17, 99a26-29)? We are about
to see how one goes about causal search of this kind, using methods specific
to the zoological domain of inquiry.
Beginning in Juv. 15/Resp. 9, the argument proceeds to make the case
that the heat required by the organism for nutrition gives rise to
a subsequent need for cooling. Interestingly, this chapter is focused almost
exclusively on the bloodless animals, arguing that they have relatively little
heat and can therefore be adequately cooled by their surroundings. That is,
it is governed by the implicit question: ‘Why do these creatures not have
organs of respiration?’ But the chapter also looks forward, making
a number of comparative comments about animals that cool themselves
by means of gills or lungs. Chapter 10 then opens20 by drawing a line on
that discussion, and begins the work of specifying as precisely as possible
the extension of the relevant attributes, in a form reminiscent of HA. The
discussion opens, for example, with the following claims:
In those that are blooded and have a heart, as many as have a lung all21 take in
air and cool themselves by inhaling and exhaling. And those that are
internally, and not merely externally, live bearing have a lung (for selachians,
while live bearing, are not so internally); as do, among the egg layers, both
those that are feathered, such as birds, and those that are scaled, such as
tortoises, lizards, and snakes. (Juv. 16/Resp. 10, 475b16-23)
The care with which groups are correlated here, identified by differences
relevant to how cooling is accomplished, is remarkable. The widest back-
ground group of interest is now the blooded animals, all of which have been
established to have a heart. Among that group, the group with a lung is
coextensive with those that cool themselves by inhaling and exhaling air.
And the subgroup with a lung consists of those that are truly live bearing,
feathered egg-layers or scaly22 egg-layers. Aristotle has carefully specified
20
“Concerning the bloodless animals, then, we have stated that they are aided in relation to living in
some cases by the surrounding air and in others by the surrounding water” (475b15-17). Once again,
we are reminded of the unreliable nature of these chapter breaks that were introduced in the
Renaissance; this reads much better as the conclusion of the previous discussion.
21
The use of the double quantifier ὅσα . . . πάντα is common in Aristotle and particularly common in
the Historia animalium, and appears to be used to identify coextensive or counter-predicated
attributes such as lungs and the taking in of air for cooling and (in the next quotation) gills and
the taking in of water for cooling. For some discussion of its use cf. Lennox 2001b, chapters 1 and 3;
Gotthelf 1988 (reprinted in Gotthelf 2012, 307–342). This passage should be compared with HA ii.
15, 505b32-506a8.
22
The point is not as clear in English as in Greek, which has distinct words for ‘fish scale’ (λεπίς) and
for the scale of a reptile (φολίς).
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
the egg-layers with lungs so as to exclude fish, without ever mentioning
them (other than an aside to stress that the selachian fishes are not truly
viviparous). He segues into discussing fish by first considering animals with
lungs that spend a good deal of time in the water, yet either birth their
young on land or take measures to insure their newborn can breathe. He
then goes on:
But23 as many as have gills, all cool themselves by taking in water; both the
kind consisting of the so-called selachians and the other footless animals.24
And all the fish are footless; and indeed what part they have [for locomotion]
is named for its likeness to wings.25 (476a1-5)
He then notes that of those so far studied (τῶν τεθεωρημένων, 476a6) only
one animal with feet also has gills, “the so-called kordulos.”26 But he notes
that it has no lung,27 and in fact, “no animal has yet been seen (οὐδὲν
ὦπταί πω) possessing both lungs and gills (476a6-7).” ‘Why not?,’ one
might ask, and (of course) Aristotle has an answer:
A cause is that the lung is for the sake of (ἕνεκα) cooling by breath . . . while
the gills are related to (πρὸς) cooling by means of water; and there is one
instrument for one use, and one is sufficient for cooling in all cases. So since
we see nature doing nothing in vain, while having two, one would be in
vain, on this account some animals have gills, some lungs, but none have
both. (476a7-15)28
There are a variety of interesting features of this argument for our topic.
First, Aristotle goes out of his way to stress the inductive basis of the
23
The δὲ at 476a1 is coordinate with the μέν at 475b17. The background class is the animals with blood
and a heart, and within that class he is distinguishing those that have a lung and cool themselves with
air from those that have gills and cool themselves with water; and then within those two classes he is
distinguishing various subdivisions. The use of ὅσα . . . πάντα in those two cases is to stress the
coextensive relationship between the organs of cooling and the medium used (air or water).
24
In case you are wondering about snakes or cetaceans, the background class here is animals with gills,
so it is only footless animals with gills that are under consideration.
25
The Greek for ‘wing’ is πτέρυξ and for ‘fin’ is πτερύγιον, the plural of which is πτερύγια.
26
Most salamanders have a larval stage during which they breathe with external, feathered gills and
lack lungs. In an attempt to be clearer about the sort of organism Aristotle was observing and the
data behind the claims he makes about the kordulos I have been greatly aided by Dr. J. W. Arntzen of
the Netherlands Centre for Biodiversity Naturalis in Leiden. On the basis of our email correspond-
ence he conjectures that Aristotle is referring to a so-called paedomorphic newt, regularly found in
mountain populations of Ichthyosaura alpestris, in Greece. Paedomorphism is the retention into
adulthood of features of the larval stage, in this case including the presence of gills and lack of lungs,
which occurs in some but not all members of this species. Aristotle would have had no reason not to
think that the paedomorphs were a distinct species.
27
And how is he sure about that? Another instance where dissection is clearly playing a role.
28
And he is seriously mistaken: there are numerous species of lungfish around the world that have, and
make good use of, both.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
10.6 From Breathing to Cooling 281
explanandum: none has yet been seen with both lung and gills. Indeed, until
now only one animal equipped for walking on land has been seen with
gills.29 Given the plausible conjecture of J. W. Arntzen mentioned in note
26, just above, it is possible that Aristotle had seen two otherwise similar
adult animals, one with lungs and the other with gills. Second, there are
subtle distinctions between the functional characterizations of the lung and
gills: the former is present for the sake of (ἕνεκα) cooling by (ὑπό) breath;
the latter are related to (πρός) cooling by means of (ἀπό) water. Aristotle is
not yet committing himself to the claim that gills are for the sake of cooling.
Up to this point, it has been mentioned a number of times without
justification that fish cool by means of gills, but there has not been any
argument that cooling is the function for the sake of which fish have gills.
What has been clearly established is that the class of animals with
a cardiovascular system, and the heat generated by it, is of wider extension
than breathing and the parts that are necessary to perform the activity of
breathing; and that the class of animals with a cardiovascular system is
coextensive with the animals that breathe plus the animals with gills. He
has established the primary universals and has carefully narrowed down the
causal search space. The ‘that it is’ stage of causal inquiry is about to
transition into a deeper inquiry into the reason why breathing occurs.
But the fact that not all blooded animals with hearts have lungs, and some
water dwellers have lungs, makes the causal inquiry more complicated.
10.6 From Breathing to Cooling: The Value of Functional

Anatomy
At the beginning of Resp. 13, Aristotle announces that he must speak “about
cooling, in what way it comes about both for those that breathe and for
those having gills” (477a11-12). This is a central theme of the remainder of
the treatise, though it is interwoven with discussions of the intimately
connected topics of generation, life and death, and the movements of the
heart and blood vessels. Since our focus in this chapter is the interaction
between the Analytics-based framework norms and the zoologically specific
norms in this inquiry, we will zero in on the discussion in chapter 16 where
this interaction is most clearly on display.
In this chapter, Aristotle refocuses the inquiry in a number of closely
related respects:
29
The same stress can be found at HA vii(viii).2, 589b26-7: ἓν δὲ μόνον νῦν ὦπται τοιοῦτον, ὁ
καλούμενος κορδύλος· οὗτος γὰρ πνεύμονα μὲν οὐκ ἔχει ἀλλὰ βράγχια.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
1. The inquiry has shifted to a more universal level, from an inquiry into
respiration to one into methods of cooling in all the blooded animals.
If the Analytics framework of inquiry discussed in Chapter 2 is in place,
this is a shift that is to be expected.
2. But in this case, that shift is a consequence of having focused, from the
beginning of the investigation, on the purpose of respiration – that should
be one’s priority in a zoological inquiry, as we discussed in Chapter 6.
3. Once the purpose of respiration is properly understood, it follows that
cooling must be shared by all blooded animals.
4. But since fish are blooded and therefore have hearts, their hearts must
use heat for the same nutritive purposes as all blooded animals and
distribute that blood to the other parts of the body by means of blood
vessels. This conclusion is a direct result of the inquiries reported in PA
iii.4–6 and in earlier chapters of Juv.
5. And given all that, they have the same need for cooling as all blooded
animals. This is the proper level of universality for the causal inquiry.
By focusing on the function of respiration then, recognizing that it is
a function that must be shared by all blooded animals, and realizing that
lungs are not coextensive with the entire class of blooded animals, Aristotle
is poised to grasp the analogical identity between lungs and gills. This shift
of level of inquiry is due to a seamless integration of norms deriving from APo.
and from PA i. That is, what led Aristotle to shift the inquiry to this higher,
functional/analogical level was holding a constant teleological focus while
he was carrying out his comparative anatomical studies. We have in this
inquiry an illustration of the value of following the normative recommen-
dation that closes Resp. 3, while searching for the proper level of universality
for this inquiry, as called for by the erotetic framework of APo. ii. And
I will make the case that these chapters may well have been structured
precisely in order to illustrate proper method for such an inquiry.
10.7 Proving Gills Are for the Sake of Cooling

Juv. 22/Resp. 16 makes the case for gills being present for the sake of
cooling. But it does so in a way that is rich in hints about the methodology
to be followed in reaching this conclusion, and it is on these hints that
I shall focus. Let us begin with a sentence that in Bekker, followed by most
modern editors, closes Juv. 21/Resp. 15.
And for what reason those breathe most that have a bloody lung is clear from
these facts. For those that are warmer have need of greater cooling, but at the
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
10.7 Proving Gills Are for the Sake of Cooling 283
same time the breath is transported easily to the source of the heat in the
heart. And the way in which the heart has a passageway to the lung needs to
be studied from the things that have been dissected (ἐκ τῶν ἀνατεμνομένων;
cf. ἐκ τῶν ἀνατομῶν, 478a35-b1) and from the written inquiries about
animals (τῶν ἱστοριῶν τῶν περὶ τὰ ζῷα γεγραμμένων). (478a21-28)
The opening line of this passage refers back to facts established in the
immediately preceding discussion, regarding the large network of blood
vessels (connected ultimately to the heart) and the large network of tubes
filled with air (and ultimately connected to the windpipe) lying side by side
one another in the lungs of all lunged animals.30 These facts, he claims,
help us understand why animals that breathe must also have more blood in
their lungs – Aristotle infers that this is because they are warmer and have
a greater need for cooling. In these animals there is both a system that
allows for easy transport of air to the lung and, as he is now going to
discuss, a passageway to the lung for the blood.
There are many references to dissections scattered throughout Aristotle’s
zoological works.31 Here he recommends that the student/investigator look
at dissected animals to observe the connections between the lung and the
heart.32 This reference is unusual in that it uses the present participle of the
verb ἀνατέμνω, while Aristotle typically refers the reader to what can be
observed “in the dissections” using the plural of the noun ἀνατομή. This
suggests that we are being directed to what would be observed by performing
dissections rather than to collections of diagrammatic representations, which
often seems to be the implied reference of the noun.33 Both here and in the
reference to dissections we are about to consider in Juv. 22/Resp. 16 (where
the dative plural noun is used), there are also references to ‘histories’ (plural),
that is, written reports of his research. In the first reference there is an implicit
contrast in the restriction of what has been written to the histories; but in
30
Resp. 15, 478a12-14: “the lung is sponge-like and filled with tubes, and of all the so-called viscera this
part is the most filled with blood (ἐναιμότατον).” There is another detailed discussion of the vascular
connections between the heart and the lung in HA iii.3, 513a35-b1, 513b11-25. Unfortunately, it does
not have a parallel discussion of the vascular connections between heart and gills.
31
Of the explicit references to dissection in Aristotle, the two in this chapter, along with those in the
PA iii and GA ii discussed above in Section 10.3, and one previous one in PN (Som. 3, 456b2), are all
there are about the heart in relation to the blood vessels. These, and those referring to the male and
female reproductive organs, account for fifteen of twenty-eight such references. For a review of all of
the passages in which Aristotle refers to his dissections, and their implications for his scientific
method, see Lennox 2018 in Falcon and Lefebvre 2018, 249–272.
32
Since Aristotle has no concept of a circulatory system and thus of a functional difference between
our ‘arterial’ and ‘venous’ systems, I suppose he would be referring to both the network of blood
vessels stemming from the pulmonary artery and those leading to the pulmonary vein.
33
For example, GA ii.7, 746a15 refers to παραδείγματα, HA i.17, 497a32, iv.1, 525a9 refer to
a διαγραφή, and HA iii.1, 511a12-13 refers to σχήματα; cf. Lennox 2018, 262–266.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
the second we are explicitly told that there are distinct epistemic reasons for
consulting the histories and the dissections. Recommending consulting both
histories and dissections for distinct reasons is not uncommon. Surprisingly,
however, on a number of occasions it is the dissections rather than, or along
with, the histories that are recommended to improve precision (ἀκρίβεια).34
To this point, then, functional anatomy has established that, in animals
with lungs, there is a system of blood vessels within the lung, originating in the
heart, that allows the blood in the lung to be contiguous with a system of
tubes, originating in the windpipe, that transport breath to and from the
lungs. He goes on:
So then: generally speaking (ὅλως) the nature of animals has need of cooling on
account of the setting of the soul aglow in the heart (διὰ τὴν ἐν τῇ καρδίᾳ τῆς
ψυχῆς ἐμπύρευσιν).35 As many animals as have not only a heart but a lung as
well, produce this <cooling> by respiration. However, those having a heart but
not a lung, as with the fish, on account of their nature being aquatic, produce
cooling by the water passing through the gills. (Juv. 22/Resp. 16, 478a28-34)
Once again there is a strong emphasis on getting the extensions right: as
many as have both hearts and lungs cool by breathing; fish, however, have
hearts, but not lungs – they produce cooling by water passing through their
gills. There are echoes here of the framework established in Juv. 14/Resp. 8,
especially of 474b7-13. In that passage, after stating that it is clear from the
dissections that the heart is the starting point of the blood vessels, which
contain the blood, Aristotle refers to de An. as having established that the other
capacities of the soul depend on the nutritive soul, which in turn depends on
‘natural fire’ – and “in this nature has set the soul aglow.” The heat thus
generated by that natural fire must be moderated, and in all animals with
lungs as well as hearts that is accomplished by means of respiration. But fish
too are blooded animals with hearts, and so need cooling as well. And here
Aristotle asserts, without argument, that “on account of their nature being
aquatic, [fish] produce cooling by the water passing through the gills.” As so
often with Aristotle, the evidence in support of this very broad and bold causal
claim follows its assertion. He next reviews what one learns by doing a careful
dissectional study of the relationship between hearts and gills in fish. The
review begins with a point about the orientation of the heart in fish.
34
For example, at HA iii.1, 511a14-15, PA iii.4, 666a9, and iv.13, 696b15-16.
35
The majority of manuscripts have ἐμπύρωσιν, but Ross defends ἐμπύρευσιν (which is found in two
manuscripts) on grounds that the spelling of the verb used by Aristotle is ἐμπύρευω. Ross notes
another piece of evidence supporting his recommendation, namely the use of the phrase ἡ φύσις
ἐμπεπύρευκεν αὐτήν earlier in Resp. 8, 474b13, which uses the verb, as opposed to the noun based on
it, to express exactly the same thought.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
Now how the heart is positioned in relationship to the gills, one ought to
study for purposes of visualizing from the dissections, and for purposes of
precision from the histories;36 but to speak summarily and for present
purposes, they are related in following manner. While it might seem that
the position of the heart in the footed animals and in the fishes is not alike,
in fact it is. For the direction in which their heads face is where the apex of
the heart is situated. But since the heads in the case of walking animals and
fish do not point in the same way, the heart <in the case of fish> has its apex
toward the mouth. (Juv. 22/Resp. 16, 478a34-b7)
It is the visual evidence from dissection, as descriptively interpreted in the
historia that is based on it, that will be critical for his case that the gills play
the same functional role in fish that lungs do in all other blooded animals.
As I have noted, the texts that refer to both the dissections and the histories
sometimes stress that the former provide a visualization and the latter
a written description.37 It is easy to assume that Aristotle thinks of the
first as providing a rather vivid, concrete illustration of the point he is
making while the historia captures the universal nature of the phenomenon
in question; but caution is needed. We do not know what was being
provided for “visualization,” but the words that are used to characterize
what one sees “in the dissections,” such as διαγραφή and παράδειγμα, can
refer to quite abstract representations. Just as a geometric illustration might
be drawn in such a way as to represent a truth about all triangles, so too
a diagram of the vascular connections between heart and lung or heart and
gills can be illustrated in a way that prescinds from all details specific to
particular kinds of animals – indeed, that is what textbook or website
illustrations typically do today.38 In our passage the crucial point Aristotle
is making is that the orientation of the body of a fish might mislead you
into thinking there is a fundamental difference in the location of the apex
of the heart in fish compared to other blooded animals. This is a first step in
Aristotle’s case that the cardiovascular system in fish is fundamentally like
36
There is one very brief account of the connection between heart and gills at HA ii.17, 507a3-10, but it
more or less duplicates what is said here.
37
Cf. HA i.17, 497a32 (ὦν ἡ μὲν ὄψις θεωρείσθω ἐκ τῆς διαγραφῆς ἐν ταῖς ανατομαῖς); PA iv.5, 680a1-4
(ὅν δὲ τρόπον ἔχει τούτων ἔκαστον, ἔκ τε τῶν ἱστοριῶν τῶν περὶ τὰ ζῷα θεωρείσθω καὶ ἐκ τῶν
ἀνατομῶν· τὰ μὲν γὰρ τῷ λόγῳ τὰ δὲ πρὸς τὴν ὄψιν αὐτῶν σαφηνίζειν δεῖ μᾶλλον); and GA ii.7,
746a14-16 (Δεῖ δὲ ταῦτα θεωρεῖν ἔκ τε τῶν παραδειγμάτων τῶν ἐν ταῖς ανατομαῖς καὶ τῶν ἐν ταῖς
ἱστορίαις γεγραμμένων).
38
And Aristotle is well aware of this fact about diagrams, as the following passage makes clear: “The
same affection comes about in thinking as in diagramming (ἐν τῷ διαγράφειν); in the latter, though
it is of no use that the triangle is a definite quantity, nevertheless we draw (γράφομεν) one of
a definite quantity; and likewise one who is thinking, even though he does not think of a quantity,
places a quantity before his eyes, but thinks of it not qua a definite quantity” (Mem. 1, 450a1-6).
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
that in all other blooded animals. More detail on the similarities comes
next.
A tube, sinewy and vascular (φλεβονευρώδης) in character, extends from the
top of the heart to the midpoint where all the gills connect to one another.
So this is the largest tube, but on either side of the heart others too stretch to
the uppermost point of each of the gills, through which the cooling comes
about in relation to the heart, by means of the water continuously flowing
through the gills. In like manner, in those that breathe the chest frequently
moves up and down when breath is taken in and expelled, like the gills in the
fish. (Juv. 22/Resp. 16, 478b7-15)
Two important analogical identifications are noted as the result of func-
tionally focused dissection: first, in fish there is a tube extending from the
top of the heart,39 composed of the same sort of uniform substance as the
blood vessels in lunged animals, which extends out from the heart to
a point between the gills on the two sides of the fish; second, there are
other blood vessels extending to each of the gills so that water can cool the
blood as it flows over them;40 and finally, there appears to be a reference to
the ‘fanning’ of the gills, since he says “the chest frequently moves up and
down when breath is taken in and expelled, like the gills in the fish.”41
Aristotle provides a bit more detail to the analogy in Juv. 27/Resp. 21,
once again drawing attention to the reason why respiration should be
discussed as part of an investigation of youth, old age, life and death:
And the air enters into many ducts in the lungs, like channels, and alongside
each [of these ducts] blood vessels are extended, so that the whole lung seems
to be filled with blood. And we call the inhaling of the air ‘respiration’ and
its exhaling ‘expiration.’ And this always takes place continuously, and the
animal lives so long as this part moves continuously. And for this reason, life
depends on respiration and expiration. And in the same way motion arises in
the gills of fish. For when the heat in the blood throughout the parts
increases, the gills also rise up and water flows through [them]. And when
39
Today referred to as the conus or bulbus arteriosis.
40
It is difficult to be certain what he is referring to here; Ogle conjectures that he is generalizing from
examining a ray (1897, 131, n. 136) in which the aortic arches are idiosyncratic, but that seems
unlikely. Ogle assumes that Aristotle is only referring to the aortic arches stemming from the conus
arteriosis to the gills. But since Aristotle does not make our functional distinction between veins and
arteries, and it seems clear he is discussing vessels that are not connected to the conus arteriosis, it
occurs to me that he may be referring to the blood vessels that we think of as taking oxygenated
blood from the gills to the body.
41
This is more explicit in the discussion of gills in PA iv.13 (which refers to our discussion): “A cause of
the number of gills being larger or smaller is a larger or smaller amount of heat in the heart; for the
movement must be more rapid and stronger for those with more heat. And those with more gills and
double gills have a nature of this sort more than those with simpler and fewer gills” (696b16-20).
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
water is allowed to flow in relation to the heart through these channels, the
gills are also cooled and collapse, and the water is expelled. Thus, whenever
the heat of the heart increases water is always taken in, and when it is cooled
again the water flows out. For this reason too, the final authority42 concern-
ing living and not living for those with lungs rests with breathing, and for
those with gills with the taking in of water. (480b6-20)
Engaged as he is in a comparative anatomical study governed by teleo-
logical norms, Aristotle is searching for analogies between gills and lungs
because, given that fish are blooded animals with hearts, but lack lungs,
they must have a way of cooling the heart region. It has been determined
that that is what lungs are for, and since everything with a cardiovascular
system needs that cooling, and fish lack lungs, they must accomplish it by
some other means. But Aristotle is not satisfied with imagining or supposing
that fish accomplish this by means of gills – he makes his case based on
a careful comparative anatomical study of the vascular connections
between heart and lungs and those between heart and gills, and finds
that, appearances notwithstanding, they have much in common.
Returning to Juv. 22/Resp. 16: Aristotle concludes his discussion of
respiration with one final analogy between gills and lungs, which provides
a clever segue into the topic of life and death.
And breathing animals are suffocated in a small amount of air that remains
the same; for each of them rapidly becomes hot since contact with the blood
heats each animal. And being hot, the blood prevents cooling; and when,
owing to sickness or old age, the lung in the case of breathing animals, or the
gills in the case of water-dwelling animals, is unable to move, at that point
death results. (478b15-21)
This, it will be recalled, was the initial justification, at the beginning of On
Youth and Old Age, for including respiration in this discussion: for animals
that breathe, it is a matter of life and death that they do so. But there is an
important difference between what was said at the beginning of the inquiry
and what is said here, because we now know why animals that breathe need
to do so – that is, we know what it is for. Moreover, we now have learned that
the one class of blooded animals that lacks lungs accomplishes the same
necessary cooling function by means of an analogous activity of the gills.
At the beginning of the inquiry, it will be recalled, Aristotle stressed that
it was necessary to discuss respiration at the same time as youth, old age,
42
τὸ τέλος, rendered “final authority” here, as recognized by LSJ (s.v. τέλος i.2), has the sense of the
final authority over decision-making (indeed, LSJ’s recommended translation for its use here is
“decisive difference between”).
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
life, and death because in some animals, whether they live or not is
dependent on respiration (467b10-13); and a few pages later, as he initiates
his discussion of respiration, one of his very first complaints about previous
accounts is that they claim all animals breathe, and in particular that fish
do – which, he insists, is not true (470b9-10). But by keeping his focus on
the question of the goal of respiration, while engaged in his comparative
anatomy, he has established that its function, cooling the region where the
heart concocts the blood into its final, nutrient form, is a function shared
by all blooded animals, including fish. And that has led us to the discovery
that the gills, in animals with gills – in their structure, in their vascular
connections to the heart, in their movements, and finally in their age-
related pathology – are the same by way of analogy with the lungs in all the
other blooded animals. That puzzling comment about the need to study
lungs and gills while inquiring into what respiration is for, even though
gills are not for respiration, is no longer puzzling: the fanning of water over
the gills, as it turns out, serves the same function as inhaling and exhaling
air, of breathing: both activities, performed by appropriate organs, serve
the same cooling function.
10.8 Framework Norms and Zoology-Specific Norms

Aristotle began his discussion of respiration by noting the obvious import-
ance of breathing to life for all animals that breathe – that breathing is of
life-and-death importance is obvious; but why it is so important is not at all
clear.43 Aristotle approaches that question having already determined to his
satisfaction that the nutritive soul acts by means of heat in the heart to
produce the blood that is the final nutrient in blooded animals. And we
have seen that he repeatedly stresses the crucial importance of dissection of
the visceral organs in his having reached that conclusion. The heart, then,
performs its nutritive function by means of a ‘natural fire’ or an ‘inborn
heat.’ But that ‘natural fire,’ without constant moderation, would simply
burn itself out. As we have seen, he reaches the conclusion that in animals
with heart and lungs, that moderation is achieved by respiration, while in
animals with heart and gills, it is achieved by the gills bringing in and
expelling water.44 My concern has been whether Aristotle is right to insist
on the importance for a natural inquiry that the anatomical study of the
43
And in fact, remained unclear until the nineteenth century!
44
There is an interesting question, tangential to this discussion, about why the brain is never
mentioned in this discussion. For, while it is not involved in respiration, according to PA ii.7 its
primary function is to moderate the temperature of blood, the same function performed by
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
10.8 Framework Norms and Zoology-Specific Norms 289
parts be governed from the outset by teleological questions about their
functions. But I have also been at pains to stress the importance of
comparative anatomy in such a teleologically motivated investigation.
Aristotle reaches an answer to his initial question, “Why do animals
breathe?”, by starting with an already established theory about the heart
as the source of the nutritional blood and the blood vessels and then
carefully examining, by means of dissection, the internal anatomy of the
lungs and their complicated connections to the windpipe and to the heart.
But then, by noting comparative facts such as that many kinds of blooded
animals with cardiovascular parts do not have lungs and spend most or all
of their time in water, and that many kinds of water animals do have lungs
and breathe, he realizes two things: (i) certain animals with hearts, and thus
with the same need for moderating the heat generated by the heart, cannot
do so by breathing, since they do not have the appropriate equipment to do
so; and (ii) it is not because it is impossible for a footless, aquatic animal to
be properly equipped to do so – Aristotle is well aware of, and impressed by
the fact that, whales, dolphins, and porpoises both live a fully aquatic way
of life and have lungs for breathing. Thus, armed with an answer to the
question “What is breathing for?” and with these revelations from com-
parative dissection about which viscera are and are not coextensive with
which, and details about physical and physiological connections among
parts, he is able to answer the question about how blooded animals without
lungs accomplish that cooling function, so important to the maintenance
of life for blooded animals. It is this constant interplay that we see in these
texts, between comparative anatomy and teleology, that is, as I now see it,
the key to the success of his inquiry.
Prioritizing the ‘What is it for?’ question is, as we saw in Chapter 6, one
of the very first norms of zoological inquiry defended in PA i.1. But the
defense of that priority is very abstract and relies heavily on examples
drawn from craftsmanship. In looking at the role of that question in
Aristotle’s inquiry into breathing, we have seen the way in which it set
the terms for his anatomical study of the parts by which certain blooded
animals take in and expel air – because of that norm being in play, the
inquiry with which Aristotle begins is taken to a level of universality and
causal depth that it would not have achieved otherwise.
But we have also seen the way the Analytics erotetic framework plays an
equally fundamental role in this inquiry: in particular, in the complicated
breathing and by the actions of gills. Equally puzzling, PA ii.7 makes no mention of the lung as
performing a similar function to the brain.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
relationship between the four inquiries we were introduced to at the begin-
ning of APo. ii, into the ‘that,’ the ‘reason why,’ the ‘if it is,’ and the ‘what it is’
(89b23-25). On the ‘that’ side of the inquiry are figuring out the precise,
predicative relationships between heart, lung, gills, and breathing, as well as
the anatomical relationships between the air passages and the blood vessels
in the lungs, and the analogous relationships between blood vessels leading
from the hearts of fish to their gills. Figuring these relationships out is,
however, closely interwoven with answering a number of ‘if it is’ questions:
• Are there animals with both gills and lungs?
• Are there fully aquatic animals that have lungs and breathe?
• Are there land animals with gills?
• Do fish have organs for cooling?
Similarly, the same sort of complicated relationships described using eclipses
and thunder as examples45 in APo. ii.2, 8, and 10, between ‘why it is’ and
‘what it is’ inquiries, are also vividly on display in this inquiry into the
function of respiration. In the case of both breathing and the organs used to
accomplish it, the answers to ‘reason why’ inquiries are, in the first instance,
teleological; but since the ultimate reason why in both cases is to provide
cooling in order to moderate the heat generated by the heart, and since fish
accomplish the same function by means of gills and the intake and expulsion
of water, stating that function cannot provide a fully adequate answer to the
‘what is it?’ question. The means by which cooling is accomplished is
a necessary component of the definition of respiration, which means it
must include reference to the material structures involved and the efficient
causal processes involved in the cooling process. It is by reference to efficient
causal processes and material, instrumental details that one form of cooling is
differentiated from another. This is true not only at the level of differentiat-
ing the way fish accomplish this goal from that employed by other blooded
animals – it is also true in differentiating one form of respiration from
another.46 The cetaceans (to take an extreme example) certainly do breathe
and therefore have lungs; but they also take in and expel water, which
Aristotle (mistakenly) claims is the reason they have the “pipe” (aulos) or
blowhole. The process of inquiry in this case is structured by the goal of
answering the four kinds of questions that introduce APo. ii, which leads to
a process as complex as APo. ii.2 would lead us to expect. But this is a natural
45
But it is worth recalling that in APo. ii.14, 16, and 17 biological examples, some quite technical, are
also used to make similar points.
46
See PA iii.6, 669a22-b7; discussed in Lennox 2001a, 265–268.
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
10.8 Framework Norms and Zoology-Specific Norms 291
inquiry, and thus the sought-for answers to the ‘what is it?’ questions will
involve determining material structures and capacities, efficient causal pro-
cesses, and functions. Moreover, it is a zoological inquiry, and thus Aristotle
repeatedly stresses that the overarching nature of the inquiry must be
teleological: even when practicing anatomy in order to understand the
material structures involved in respiration, your work must be governed by
the ‘what is it for?’ question. Norms at multiple levels are in play in this
inquiry, leading to a complex process of inquiry.
appendix
Schematic Outline of Aristotle’s Inquiry into Respiration
Ch. 1. Others have spoken about breathing, but not well. Among
other errors, many seem to claim all animals breathe. Clearly
all with lungs do, but there are differences even there.
Chs. 2–3. Anaxagoras and Diogenes. What is wrong with their ideas
about fish breathing. Their failure to ask teleological questions
while performing dissections.
Ch. 4. Democritus. Like others, Democritus also does not consider
what respiration is for, and there are lots of more specific
problems with his account.
Ch. 5. Plato in the Timaeus gets things all wrong anatomically and
also does not say what respiration is for.
Ch. 6. Some say it nourishes the intrinsic fire. Lots of problems with
that idea.
Ch. 7. Empedocles has much to say about breathing, much of it is
quoted, and many problems in his account are identified. He
also fails to consider the question of what breathing is for.
Ch. 8. A reminder of the views about heat, soul, and nutrition in On
Youth and Old Age. Final nutrient is blood; blood is produced
in the heart, which is the source of blood vessels as well. (On
this point, the Dissections need to be consulted.) The produc-
tion of nutritive blood requires inherent heat, which is the
instrument of the nutritive soul.
Ch. 9. Given the conclusion of chapter 8, there is a subsequent
need for moderation of heat by cooling. Bloodless animals
are sufficiently cooled by their surroundings. A few
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
comparative comments are made about those who cool by
gills or lungs.
Ch. 10. A review of the universal generalizations/correlations among
parts and activities in the blooded animals related to lungs and
gills.
Ch. 11. Nourishment is needed for being, cooling for preservation –
how the organs for feeding and cooling work together.
Ch. 12. The cetaceans present an aporia: they are footless, water dwell-
ing, and take in water, yet have a lung – how that integrates
with the rest. There is also a discussion of cooling in the
cephalopods and crustaceans.
Ch. 13. How exactly do lungs and gills bring cooling about?
Ch. 14. Empedocles is wrong about this as well – but his error serves to
introduce us to the relationship between an animal’s constitu-
tion, where it lives, and thus how it must cool itself.
Ch. 15. Why does the part that receives and expels the air have so
much blood, as lungs do; and the rate of breathing seems
related to the amount of blood. (This receives further devel-
opment in chapter 21.)
Ch. 16. Cooling and the communication between heart and lungs,
heart and gills. The importance of dissection for understand-
ing this. The important parallels between the heart/lung rela-
tion and the heart/gills relation.
Ch. 17. Old age, death, and their relation to respiration and condition
of the lungs and gills. A number of references to death due to
disease.
Ch. 18. Genesis defined. And “What genesis, life, and death are, and due
to what cause they belong to animals, has been stated.”
Ch. 19. Why fish cannot live for long without water, and lunged
animals cannot live for long without air.
Ch. 20. How are pulse, heartbeat, heart palpitations, and breathing
related?
Ch. 21. Breathing and the nutritive principle, how breathing works
like a bellows, the interface of air and blood, water and blood
in lungs and gills. A most interesting closing passage about the
relevance of all of this to the subject of health and disease (i.e.,
to medicine).
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
Bibliography
Ackrill, J. 1981. “Aristotle’s Theory of Definition: Some Questions on Posterior

Analytics ii. 8–10.” In Berti ed. 1981: 359–384.
Allan, D. J. 1961. “Quasi-Mathematical Method in the Eudemian Ethics.” In
Mansion ed. 1961: 303–318.
Anagnostopoulos, G. ed. 2009a. A Companion to Aristotle. London: Wiley-Blackwell.
2009b. “Aristotle’s Methods.” In Anagnostopoulos ed. 2009a: 101–122.
Angioni, L. 2010. Aristóteles: Física i–ii. Campinas: Editora Unicamp.
Balme, D. M. 1987a. “The Place of the Biology in Aristotle’s Philosophy.” In
Gotthelf and Lennox, eds. 1987: 9–20.
1987b. “Aristotle’s Biology Was Not Essentialist.” In Gotthelf and Lennox, eds.
1987: 291–312.
1992. Aristotle: De partibus animalium i and De generatione animalium i (with
passages from ii. 1–3). Oxford: Clarendon Press.
2002. Aristotle: Historia animalium Volume i: Books i–x: Text. Cambridge
University Press.
Barnes, J. 1980. “Aristotle and the Methods of Ethics.” Revue Internationale de
Philosophie 34(133/134): 490–511.
1981. “Proof and the Syllogism.” In Berti ed. 1981: 17–59.
1993. Aristotle: Posterior Analytics. 2nd ed. Oxford University Press.
Beere, J. 2009. Doing and Being: An Interpretation of Aristotle’s Metaphysics Theta.
Oxford University Press.
Bekker, I. ed. 1831. Aristotelis Opera (rev. ed. 1960, O. Gigon). Berlin: De Gruyter.
Belnap, N. D., and T. B. Steel. 1976. The Logic of Questions and Answers. New
Haven, ct: Yale University Press.
Bemer, K. 2014. “A Philosophical Examination of Aristotle’s Historia animalium.” PhD
dissertation, University of Pittsburgh. http://d-scholarship.pitt.edu/id/eprint/
22674
Berti, E. ed. 1981. Aristotle on Science: The Posterior Analytics. Proceedings of the
8th Symposium Aristotelicum. Padua: Editrice Antenore.
Beullens, P. 2014. “Facilius sit Nili caput invenire: Toward an Attribution and
Reconstruction of the Aristotelian Treatise De innudatione Nili.” In De
Leemans ed. 2014: 165–202.
Blair, G. A. 1992. Energeia and Entelecheia: ‘Act’ in Aristotle. University of Ottawa
Press.
293
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
294 Bibliography
Bogaard, P. A. 1979. “Heaps or Wholes: Aristotle’s Explanation of Compound
Bodies.” Isis 70(1): 11–29.
Bolton, R. 1976. “Essentialism and Semantic Theory in Aristotle: Posterior
Analytics ii. 7–10.” Philosophical Review 85(4): 514–544.
1978. “Aristotle’s Definitions of the Soul: De anima ii, 1–3.” Phronesis 23(3):
258–278.
1987. “Definition and Scientific Method in Aristotle’s Posterior Analytics and
Generation of Animals.” In Gotthelf and Lennox, eds. 1987: 151–166.
1991. “Aristotle’s Method in Natural Science: Physics i.” In Judson ed. 1991: 2–29.
2009. “Two Standards for Inquiry in Aristotle’s De caelo.” In Bowen and
Wildberg, eds. 2009: 51–82.
forthcoming. “Epistemology and Psychology in Aristotle’s Posterior Analytics
ii.19.” In Salmieri, ed. forthcoming.
Bolton, R., and A., Code. 2012. “Aristotle on Knowledge.” In Hetherington ed.
2012: 50–71.
Bonitz, H. 1870. Index Aristotelicus. Berlin: Reimer.
Bostock, D. 1994. Aristotle: Metaphysics Books Ζ and Η. Oxford: Clarendon Press.
2006. Space, Time, Matter, and Form: Essays on Aristotle’s Physics. Oxford
University Press.
Bowen, A. C. ed. 1991. Science and Philosophy in Classical Greece, New York:
Garland.
Bowen, A. C., and C. Wildberg, eds. 2009. New Perspectives on Aristotle’s De caelo.
Leiden: E. J. Brill.
Broadie, S. 1998. “Nous and Nature in De anima iii.” In Cleary and Wians, eds.
1998: 163–176.
Bromberger, S. 1962. “An Approach to Explanation.” In Butler ed. 1962: 72–105.
Bronstein, D. 2016. Aristotle on Knowledge and Learning. Oxford University Press.
Buchheim, T., D. Meissner, and N. Wachsmann, eds. 2016. Soma: Körperkonzepte
und körperliche Existenz in der antiken Philosophie und Literatur. (Archiv für
Begriffsgeschichte, Sonderheft 13). Hamburg: Felix Meiner.
Burnyeat, M. 1980. “Aristotle on Learning to Be Good.” In Rorty ed. 1980: 69–92.
1981. “Aristotle on Understanding Knowledge.” In Berti ed. 1981: 97–139.
1992. “Is an Aristotelian Philosophy of Mind Still Credible? A Draft.” In
Nussbaum and Rorty, eds. 1992: 15–26.
2001. A Map of Metaphysics Zeta. Pittsburgh: Mathesis Publications.
2002. “De anima ii.5.” Phronesis 47(1): 28–90.
2004. “Aristotle and the Foundations of Sublunary Physics.” In Mansfield and
de Haas, eds. 2004: 7–24.
Butler, R. S. ed. 1962. Analytic Philosophy – Second Series. Oxford: Basil Blackwell.
Byrne, P. 1997. Analysis and Science in Aristotle. Albany: State University of
New York Press.
Carbone, A. 2016. “The Axes of Symmetry. Morphology in Aristotle’s Biology.”
Apeiron 49(1): 1–31.
Carroll, S. 2005. Endless Forms Most Beautiful. New York and London:
W. W. Norton.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
Bibliography 295
Cartwright, N. 1999. The Dappled World: A Study of the Boundaries of Science.
Cambridge University Press.
Caston, V. 1997. “Epiphenomenalism, Ancient and Modern.” Philosophical
Review 106: 309–363.
1998. “Aristotle on the Relation of the Intellect to the Body: Commentary on
Broadie.” In Cleary and Wians, eds. 1998: 177–192.
Charles, D. 1988. “Aristotle on Hypothetical Necessity and Irreducibility.” Pacific
Philosophical Quarterly 69: 1–53.
2000. Aristotle on Meaning and Essence. Oxford University Press.
ed. 2010a. Definition in Ancient Philosophy. Oxford University Press.
2010b. “The Paradox in the Meno and Aristotle’s Attempts to Resolve It.” In
Charles ed. 2010a: 115–150.
2010c. “Definition and Explanation in the Posterior Analytics and Metaphysics.”
In Charles ed. 2010a: 286–328.
2012. “Teleological Causation.” In Shields ed. 2012: 227–266.
Charlton, W. 1987. “Aristotle on the Place of Mind in Nature.” In Gotthelf and
Lennox, eds. 1987: 408–423.
1992. Aristotle: Physics Books i and ii. Oxford: Clarendon Press.
Clark, S. 1975. Aristotle’s Man: Speculations upon Aristotelian Anthropology. Oxford:
Clarendon Press.
Cleary, J. J., and D. C. Shartin, eds. 1989. Proceedings of the Boston Area Colloquium in
Ancient Philosophy, Vol. iv (1988). Lanham, md: University Press of America.
Cleary, J. J., and W. Wians, eds. 1998. Proceedings of the Boston Area Colloquium in
Ancient Philosophy, Vol. xii (1996). Lanham, md: Rowman & Littlefield.
Code, A. 1997. “The Priority of Final Causes over Efficient Causes in Aristotle’s
PA.” In Kullmann and Föllinger, eds. 1997: 127–144.
2010. “An Aristotelian Puzzle about Definition: Metaphysics Z.12.” In Lennox
and Bolton, eds. 2010: 78–96.
2015. “The Matter of Sleep.” In Ebrey ed. 2015: 11–45.
Cohen, S. M. 1989. “Aristotle on Hot, Cold, and Teleological Explanation.”
Ancient Philosophy 9: 255–270.
Coope, U. 2009. “Change and Its Relationship to Actuality and Potentiality.” In
Anagnostopoulos ed. 2009: 277–291.
Cooper, J. 2004. “A Note on Aristotle on Mixture.” In Mansfeld and de Haas eds.
2004: 315–326.
Cornford, F. M. 1935. Plato’s Theory of Knowledge: the Theaetetus and the Sophist.
London: Kegan Paul, Trench, Trübner & Co.
Davidson, D. 2009. “A Coherence Theory of Truth and Knowledge.” In Sosa,
Kim, Fantl, and McGrath, eds. 2009: 124–133.
De Leemans, P. ed. 2014. Translating at the Court: Bartholomew of Messina and
Cultural Life at the Court of Manfred, King of Sicily. Leuven University Press.
Diels, H. 1882. Simplicii in Physicorum libros quattuor priores commentaria. Vol. 9
of Commentaria in Aristotelem Graeca. Berlin: Reimer.
Distelzweig, P. 2014a. “Fabricius’s Galeno-Aristotelian Teleomechanics of
Muscle.” In Nactomy and Smith, eds. 2014: 65–84.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
296 Bibliography
2014b. “‘Meam de motu & usu cordis, & circuitu sanguinis sententiam’:
Teleology in William Harvey’s De motu cordis.” Gesnerus: Swiss Journal of
the History of Medicine. 71(2): 258–270.
Dunlap, C. 2003. Cuttlefish Basics. www.tonmo.com/articles/basiccuttlefish.php
Dupré, J. 1993. The Disorder of Things: The Metaphysical Foundations of the
Disunity of Science. Cambridge, ma: Harvard University Press.
Düring, I. 1944. Aristotle’s Chemical Treatise: “Meteorologica iv,” with Introduction
and Commentary. Göteborg: Elanders.
Durling, R. M. 1996. The Divine Comedy of Dante Alighieri, Volume I: Inferno.
Ebrey, D. ed. 2015. Theory and Practice in Aristotle’s Natural Science. Cambridge
University Press.
Falcon, A. 2005. Aristotle and the Science of Nature: Unity without Uniformity.
2016. “The Subject Matter of Aristotle’s Physics.” In Buchheim, Meissner, and
Wachsmann, eds. 2016: 423–436.
Falcon, A., and D. Lefebvre, eds. 2018. Aristotle’s Generation of Animals: A Critical
Guide. Cambridge University Press.
Falcon, A., and M. Leunissen. 2015. “The Scientific Role of Eulogos in Aristotle’s
Cael. ii 12.” In Ebrey ed. 2015: 217–240.
Feigl, H., M. Scriven, and G. Maxwell, eds. 1958. Concepts, Theories and the Mind-
Body Problem: Minnesota Studies in the Philosophy of Science, vol. 2.
Minneapolis: University of Minnesota Press.
Ferejohn, M. 1994. “The Definition of Generated Composites in Aristotle’s
Metaphysics.” In Scaltsas, Charles, and Gill, eds. 1994: 291–318.
2013. Formal Causes: Definition, Explanation, and Primacy in Socratic and
Aristotelian Thought. Oxford University Press.
Festugière, A.-J. 1948. Hippocrate: L’Ancienne Médecine. Paris: Librairie
C. Klincksieck.
Fortenbaugh, W. W., and R. W. Sharples, eds. 1988. Theophrastean Studies (Rutgers
Studies in Classical Humanities 3). New Brunswick, nj: Transaction Books.
Frede, D. 1993. Plato: Philebus. Indianapolis, in: Hackett.
2004. “On Generation and Corruption i.10: On Mixtures and Mixables.” In
Mansfeld and de Haas, eds. 2004: 289–314.
Frede, M. 1987a. Essays in Ancient Philosophy. Minneapolis: University of
Minnesota Press.
1987b. “Observations on Perception in Plato’s Later Dialogues.” In Frede
1987a: 3–8.
1992. “On Aristotle’s Conception of the Soul.” In Nussbaum and Rorty, eds.
1992: 93–108.
1996. “Aristotle’s Rationalism.” In Frede and Striker, eds. 1996: 157–174.
Frede, M., and G. Patzig. 1988. Aristoteles Metaphysik Ζ: Text, Übersetzung und
Kommentar, 2 vols. Munich: Beck.
Frede, M., and G. Striker, eds. 1996. Rationality in Greek Thought. Oxford:
Clarendon Press.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
Bibliography 297
Freeland, C. 1994. “Aristotle on Perception, Appetition, and Self-Motion.” In Gill
and Lennox, eds. 1994: 35–64.
Freudenthal, G. 1995. Aristotle’s Theory of Material Substance: Heat, Pneuma, Form
and Soul. Oxford: Clarendon Press.
Furley, D. 1983. “The Mechanics of Meteorology iv: A Prolegomenon to Biology.”
Reprinted in Furley 1989: 132–148.
1989. Cosmic Problems: Essays on Greek and Roman Philosophy of Nature.
1994. “Self-Movers.” In Gill and Lennox, eds. 1994: 3–14.
Furth, M. 1988. Substance, Form and Psyche: An Aristotelian Metaphysics.
Gasser-Wingate, M. 2016. “Aristotle on Induction and First Principles.”
Philosopher’s Imprint 16 (4): 1–20.
2021. Aristotle’s Empiricism. Oxford University Press.
Gilbert, S. 2006. Developmental Biology. 8th ed. Sunderland, ma: Sinauer.
Gill, M. L. 1989. Aristotle on Substance: The Paradox of Unity. Princeton University
Press.
1994. “Aristotle on Self-Motion.” In Gill and Lennox, eds. 1994: 15–34.
2009. “The Theory of the Elements in De caelo 3 and 4.” In Bowen and
Wildberg, eds. 2009: 139–161.
2014. The Limits of Teleology in Meteorology iv.12. HOPOS: The Journal of the
International Society for the History of Philosophy of Science. Vol. 4 (2): 335–350.
Gill, M. L., and J. G. Lennox, eds. 1994. Self-Motion from Aristotle to Newton.
Princeton University Press.
Goldin, O. 1996. Explaining an Eclipse: Aristotle’s Posterior Analytics 2.1–10. Ann
Arbor: University of Michigan Press.
Gotthelf, A. 1987. “First Principles in Aristotle’s Parts of Animals.” In Gotthelf and
Lennox, eds. 1987: 167–198. (Reprinted in Gotthelf 2012: 153–185.)
1988. “Historiae I: Plantarum et Animalium.” In Fortenbaugh and Sharples, eds.
1988: 100–135. (Reprinted in: Gotthelf 2012: 307–342.)
1997. “The Elephant’s Nose: Further Reflections on the Axiomatic Structure of
Biological Explanation in Aristotle.” In Kullmann and Föllinger, eds. 1997:
85–96. (Reprinted in Gotthelf 2012: 186–196.)
2012. Teleology, First Principles, and Scientific Method in Aristotle’s Biology.
Gotthelf, A., and J. G. Lennox, eds. 1987. Philosophical Issues in Aristotle’s Biology.
Gotthelf, A., and A. Falcon. 2018. “‘One Long Argument’? The Unity and
Structure of Aristotle’s Generation of Animals.” In Falcon and Lefebvre,
eds. 2018: 15–34.
Gotthelf, A., and M. Leunissen. 2012. “What’s Teleology Got to Do with It?
A Reinterpretation of Aristotle’s Generation of Animals v.” In Gotthelf 2012:
117–141.
Gottschalk, H. B. 1961. “The Authorship of Meteorologica Book iv.” Classical
Quarterly 11: 67–79.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
298 Bibliography
Graham, D. W. 1988. “Aristotle’s Definition of Motion.” Ancient Philosophy 8:
209–215.
1989. “The Etymology of Ἐντελέχεια.” American Journal of Philology 110: 73–80.
1995. “The Development of Aristotle’s Concept of Actuality: Comments on
a Reconstruction by Stephen Menn.” Ancient Philosophy 15: 551–564.
1996. “The Metaphysics of Motion: Natural Motion in Physics ii and Physics
viii.” In Wians ed. 1996: 171–192.
Gregoric, P. 2005. “Plato’s and Aristotle’s Explanation of Human Posture.” Rhizai
2(2): 183–196.
Guthrie, W. K. C. 1939. On the Heavens. Loeb Classical Library. Cambridge, ma:
Harvard University Press.
Hamlyn, D. W. 1968. Aristotle’s De anima Books ii and iii. Oxford: Clarendon
Press.
Hannon, J. F. 2011. “Aristotle’s Conception of Science: The Case of On Youth and
Old Age, and Life and Death, and Respiration.” University of Chicago, PhD
dissertation.
Hardie R. P., and R. K. Gaye. 1930. Physica. Volume 11 of The Works of Aristotle
Translated into English. W. D. Ross and J. A. Smith, eds. Oxford: Clarendon
Press.
Hasper, P. S. 2006. “Sources of Delusion in Analytica Posteriora 1.5.” Phronesis 51
(3): 252–284.
Hasper, P. S., and J. Yurdin. 2014. “Between Perception and Scientific Knowledge:
Aristotle’s Account of Experience.” Oxford Studies in Ancient Philosophy 47
(Winter 2014): 117–147.
Hayduck, M. 1897. Philoponi in De Anima commentaria. Vol. 15 of Commentaria
in Aristotelem Graeca. Berlin: Reimer.
Heath, T. 1931. A Manual of Greek Mathematics. Oxford University Press.
Heiberg. J. L. 1894. Simplicii in De Caelo commentaria. Vol. 7 of Commentaria in
Aristotelem Graeca. Berlin: Reimer.
1927. Hippocratis Opera, Vol. 1. Leipzig: Teubner.
Heinaman, R. 1994. “Is Aristotle’s Definition of Change Circular?” Apeiron 27:
25–37.
Hempel, C., and P. Oppenheim. 1948. “Studies in the Logic of Explanation.”
Philosophy of Science 15: 135–175.
Herschel, J. F. W. 1830/1987. A Preliminary Discourse on the Study of Natural
Philosophy. The University of Chicago Press.
Hetherington, S. ed. 2012. Epistemology: The Key Thinkers. Continuum: New York
and London.
Hett, W. S. 1957. Aristotle: On the Soul, Parva naturalia, On Breath. Loeb Classical
Library. Cambridge, ma: Harvard University Press.
Hicks, R. D. 1907. Aristotle: De anima. Cambridge University Press.
Humphreys, P. 1997. “How Properties Emerge.” Philosophy of Science 64: 1–17.
Irwin, T. 1981. “Aristotle’s Methods in Ethics.” In O’Meara ed. 1981: 193–224.
1988. Aristotle’s First Principles. Oxford University Press.
Irwin, T., and G. Fine,. 1995. Aristotle: Selections. Indianapolis, in: Hackett.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
Bibliography 299
Jardine, L., and M. Silverthorne. 2000. Francis Bacon: The New Organon.
Johansen, T. 1997. Aristotle on the Sense-Organs. Cambridge University Press.
2006. “What’s New in De sensu? The Place of De sensu in Aristotle’s
Psychology.” In King ed. 2006: 140–164.
2009. “From Plato’s Timaeus to Aristotle’s De caelo: The Case of the Missing
World Soul.” In Bowen and Wildberg, eds. 2009: 9–28.
2012. The Powers of Aristotle’s Soul. Oxford University Press.
Johnson, M. R. 2005. Aristotle on Teleology. Oxford University Press.
Joly, R. 1983. “Hippocrates and the School of Cos.” In M. Ruse, ed. Nature
Animated. Dordrecht: D. Reidel: 29–47.
Judson, L. 1994. “Heavenly Motion and the Unmoved Mover.” In Gill and
Lennox, eds. 1994: 155–171.
Kant, I. 1768/1992. Concerning the Ultimate Ground of the Differentiation of
Directions in Space. In David Walford and Ralf Meerbote, eds.,The
Cambridge Edition of the Works of Immanuel Kant. Theoretical Philosophy,
1755–1770. Cambridge University Press.
Karbowski, J. 2014. “Empirical Eulogôs Argumentation in Aristotle’s Generation
of Animals iii.10.” British Journal for the History of Philosophy 22.1: 25–38.
2019. Aristotle’s Method in Ethics: Philosophy in Practice. Cambridge University
Press.
Keil, G., and N. Kreft, eds. 2019. Aristotle’s Anthropology. Cambridge University
Press.
Kelsey, S. 2003. “Aristotle’s Definition of Nature.” Oxford Studies in Ancient
Philosophy 25 (Winter 2003): 59–87.
King, R. A. H. 2001. Aristotle on Life and Death. London: Duckworth.
ed. 2006. Common to Body and Soul. Berlin: Walter de Gruyter.
Kirwan, C. 1971. Aristotle’s Metaphysics, Books Γ, Δ, Ε. Oxford: Clarendon Press.
Kitcher, P. 1993. The Advance of Science. Oxford University Press.
Kleiner, S. A. 1970. “Erotetic Logic and the Structure of Scientific Revolution.”
British Journal for the Philosophy of Science 21: 149–165.
Kosman, L. A. 1969. “Aristotle’s Definition of Motion.” Phronesis 14(1): 40–62.
1973. “Understanding, Explanation, and Insight in the Posterior Analytics.” In
Lee, Mourelatos, and Rorty, eds. 1973: 374–392.
1994. “Aristotle’s Prime Mover.” In Gill and Lennox, eds. 1994: 135–154.
Kostman, J. 1987. “Aristotle’s Definition of Change.” History of Philosophy
Quarterly 4(1): 3–16.
Kraut, R. ed. 2006a. The Blackwell Guide to Aristotle’s Nicomachean Ethics.
Malden, ma: Blackwell.
2006b. “How to Justify Ethical Propositions: Aristotle’s Method.” In Kraut ed.
2006a: 76–95.
Kress, E. 2018. “Aristotle on Chance, Causation and Teleology.” PhD dissertation,
Yale University.
Kuhn, T. 1961. The Structure of Scientific Revolutions. University of Chicago Press.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
300 Bibliography
Kullmann, W. 1974. Wissenschaft und Methode: Interpretationen zur aristotelischen
Theorie der Naturwissenschaft. Berlin: Walter de Gruyter.
1997. “Die Voraussetzungen für das Studium der Biologie nach Aristoteles.” In
Kullmann und Föllinger, eds. 1997: 43–62.
2014. Aristoteles als Naturwissenschaftler. Berlin: Walter de Gruyter.
Kullmann, W., and S. Föllinger, eds. 1997. Aristotelische Biologie. Intentionen,
Methoden, Ergebnisse. Stuttgart: Franz Steiner Verlag.
Kung, J. “Aristotle’s ‘De motu animalium’ and the Separability of the Sciences.”
Journal of the History of Philosophy 20 (1982): 65–76.
Laberrière, J-L. 1984. “Imagination humaine et imagination animale chez
Aristote.” Phronesis 29: 405–428.
Laks, A., and M. Rashed, eds. 2004. Aristote et le mouvement des animaux: Dix
études sur le De motu animalium. Villeneuve d’Ascq: Presses Universitaires
du Septentrion.
Lanza, D., and M. Vegetti, eds. 1971. Opere Biologiche di Aristotele. Turin: Unione
Tipografico-Editrice Torinese.
LeBlond, J.-M. 1938. Eulogos et l’argument de convenace chez Aristote. Paris: Les
Belles Lettres.
1945. Logique et Methode chez Aristote. Paris: J. Vrin.
Lee, E., A. Mourelatos, and R. Rorty, eds. 1973. Exegesis and Argument: Studies
Presented to Gregory Vlastos. Phronesis Supplementary Vol. 1. Assen: Van
Gorcum.
Lee, H. D. P. 1962. Aristotle: Meteorologica. 2nd ed. Loeb Classical Library.
Cambridge, ma: Harvard University Press.
Leggatt, S. 1995. Aristotle On the Heavens i and ii, with an Introduction,
Translation and Commentary. Warminster: Aris and Phillips.
Lennox, J. G. 1986. “Aristotle, Galileo, and the Mixed Sciences.” In Wallace ed.
1986: 29–51.
1987. “Divide and Explain: The Posterior Analytics in Practice.” In
Gotthelf and Lennox, eds. 1987: 90–119. (Reprinted in Lennox
2001b: 7–38.)
1989. “Comment on Sorabji.” In Cleary and Shartin, eds. 1989: 64–76.
1991. “Between Data and Demonstration: the Analytics and the Historia
animalium.” In Bowen ed. 1991: 261–295. (Reprinted in Lennox 2001b:
39–71.)
1994. “Aristotelian Problems.” Ancient Philosophy (Special Issue: Logic,
Dialectic, and Science in Aristotle, R. Bolton and R. Smith’ eds.) 14: 53–78.
(Reprinted in Lennox 2001b: 72–97.)
1999. “The Place of Mankind in Aristotle’s Zoology.” Philosophical Topics 27(1),
1999: 1–16.
2001a. Aristotle: On the Parts of Animals. Oxford: Clarendon Press.
2001b. Aristotle’s Philosophy of Biology: Studies in the Origins of Life Science.
2005. “The Place of Zoology in Aristotle’s Natural Philosophy.” In Sharples ed.
2005: 55–71.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
Bibliography 301
2008a. “‘As if We Were Investigating Snubness’: Aristotle on the Prospects for
a Single Science of Nature.” Oxford Studies in Ancient Philosophy 35 (Winter):
149–186.
2008b. “Review of D. Bostock. Space, Time, Matter, and Form.” Mind 117
(January): 170–174.
2009. “De Caelo 2.2 and Its Debt to De incessu animalium.” In Bowen and
Wildberg, eds. 2009: 187–214.
2010. “The Unity and Purpose of On the Parts of Animals i.” In Lennox and
Bolton, eds. 2010: 56–77.
2014a. “Preparing for Demonstration: Aristotle on Problems.” In Lennox ed.
2014b.
ed. 2014b. Episteme, Demonstration, and Explanation: A Fresh Look at
Aristotle’s Posterior Analytics. (Papers by D. Bronstein, D. Charles,
G. Salmieri, and J. G. Lennox). Metascience (2014) 23: 1–35. DOI 10.1007/
s11016–013-9815–1
2014c. Review of Michael Ferejohn, Formal Causes: Definition, Explanation, and
Primacy in Socratic and Aristotelian Thought. Notre Dame Philosophical Review:
http://ndpr.nd.edu/news/53060-formal-causes-definition-explanation-and-
primacy-in-socratic-and-aristotelian-thought/
2014d. “Aristotle on the Emergence of Material Complexity: Meteorology iv and
Aristotle’s Biology.” HOPOS: The Journal of the International Society for the
History of Philosophy of Science 4(2)(Fall 2014): 272–305.
2015. “Aristotle’s Posterior Analytics and the Aristotelian Problemata.” In
Mayhew ed. 2015: 36–60.
2018. “Aristotle, Dissection, and Generation: Experience, Expertise and the
Practices of Knowing.” In Falcon and Lefebvre, eds. 2018: 249–272.
Lennox, J., and R. Bolton, eds. 2010. Being, Nature, and Life in Aristotle.
Leunissen, M. 2009. “Why Stars Have No Feet: Explanation and Teleology in
Aristotle’s Cosmology.” In Bowen and Wildberg, eds. 2009: 215–237.
2010. Explanation and Teleology in Aristotle’s Science of Nature. Cambridge
University Press.
2012. “Aristotle on Natural Character and Its Implications for Moral
Development.” Journal of the History of Philosophy 50(4): 507–530.
2017. From Natural Character to Moral Virtue. Oxford University Press.
2018. “Order and Method in Aristotle’s Generation of Animals 2.” In Falcon and
Lefebvre, eds. 2018: 56–74.
Lewis, E. 1996. Alexander of Aphrodisias: On Aristotle Meteorology 4. London:
Duckworth.
Lloyd, G. E. R. 1962. “Right and Left in Greek Philosophy.” Journal of Hellenic
Studies 82: 56–66.
1983. Science, Folklore and Ideology. Cambridge University Press.
1991. Methods and Problems in Greek Science: Selected Papers. Cambridge
University Press.
1996. Aristotelian Explorations. Cambridge University Press.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
302 Bibliography
Machamer, P. 2004. “Activities and Causation: The Metaphysics and
Epistemology of Mechanisms.” International Studies in the Philosophy of
Science 187: 27–39.
Machamer, P., L. Darden, and C. Craver. 2000. “Thinking about Mechanisms.”
Philosophy of Science 67: 1–25.
Mahoney, E. 1982. “Neoplatonism, the Greek Commentators, and Renaissance
Aristotelianism.” In O’Meara ed. 1982: 264–282.
Mansfeld, J., and F. J. de Haas, eds. 2004. Aristotle. On Generation and
Destruction, i. Proceedings of the 15th Symposium Aristotelicum. Oxford
University Press.
Mansion, S. ed. 1961. Aristote et les problèmes de méthode. Publications
Universitaires de Louvain.
Mayhew, R. ed. 2015. The Aristotelian Problemata physica: Philosophical and
Scientific Investigations. Leiden: Brill.
McDowell, J. 1994. Mind and World. Cambridge, ma: Harvard University Press.
McKirahan, Jr., R. 1992. Principles and Proofs: Aristotle’s Theory of Demonstrative
Science. Princeton University Press.
Menn, S. 1994. “The Origins of Aristotle’s Concept of Ἐνέργεια: Ἐνέργεια and
Δύναμις.” Ancient Philosophy 14: 73–114.
forthcoming. Physics ii and Its Relation to Physics i–iv.
Mignucci, M. 2007. Aristotele, Analitici secondi: Organon iv. Rome: Editori Laterza.
Miller, Jr., F. 2012. “Separability.” In Shields ed. 2012: 306–339.
Mirus, C. V. 2006. “The Homogeneous Bodies of Meteorology iv.12.” Ancient
Philosophy 26: 45–64.
Mitchell, S. 2003. Biological Complexity and Integrative Pluralism. Cambridge
University Press.
Morel, P.-M. 2000. Aristote: Petits traités d’histoire naturelle. Paris: Flammarion.
2006. “Common to Soul and Body.” In King ed. 2006: 121–139.
Needham, P. 2009. “An Aristotelian Theory of Chemical Substance.” Logical
Analysis and History of Philosophy 12: 149–164.
Newman, W. 2004. Promethean Ambitions: Alchemy and the Quest to Perfect
Nature. University of Chicago Press.
2006. Atoms and Alchemy: Chymistry and the Experimental Origins of the
Scientific Revolution. University of Chicago Press.
Norton, J. D. “The Material Theory of Induction.” www.pitt.edu/~jdnorton/pa
pers/material_theory/material.html
Nussbaum, M. 1978. Aristotle’s De Motu Animalium. Princeton University Press.
ed. 1986. Logic, Science, and Dialectic: Collected Papers in Greek Philosophy by
G. E. L. Owen. Ithaca, ny: Cornell University Press.
Nussbaum, M., and E. Rorty, eds. Essays on Aristotle’s De anima. Cambridge
University Press.
Ogle, W. 1882. Aristotle on the Parts of Animals. London: Kegan, Paul, Trench
& Co.
1897. Aristotle on Youth and Old Age, Life and Death, and Respiration. London:
Longmans, Green and Co.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
Bibliography 303
O’Meara, D. ed. 1981. Studies in Aristotle. Washington, dc: Catholic University of
America Press.
ed. 1982. Neoplatonism and Christian Thought. Albany, ny: SUNY Press.
Oppenheim, P., and H. Putnam, 1958. “Unity of Science as a Working
Hypothesis.” In Feigl, Scriven, and Maxwell, eds. 1958: 3–36.
Owen, G. E. L. 1961. “Tithenai ta phainomena. ” In Mansion, ed. 1961: 83–103.
1986. “The Platonism of Aristotle.” In Nussbaum ed. 1986: ch. 11.
Pasnau, R. ed. 1999. Thomas Aquinas: Commentary on Aristotle’s De anima,
translation with introduction and notes. New Haven: Yale University
Press.
Peck, A. L. 1961. Aristotle: Parts of Animals. Loeb Classical Library. Cambridge,
ma: Harvard University Press. (Original edn. 1937)
1963. Aristotle, Generation of Animals. London and Cambridge, ma: Loeb
Classical Library. (Original edn. 1942)
1965. Aristotle: History of Animals, Books i–iii. Loeb Classical Library.
1970. Aristotle: History of Animals, Books iv–vi. Loeb Classical Library.
Pellegrin, P. 2000. Aristote: Physique. Paris: GF Flammarion.
Pepe, L. 1978. “A proposito del iv libro dei Meteorologica de Aristotele.” Atti
dell’Academia di Scienza Morale e Politica, Napoli 89: 503–523.
Peramatzis, M. 2011. Priority in Aristotle’s Metaphysics. Oxford University
Press.
Perfetti, S. 1999. “Three Different Ways of Interpreting Aristotle’s De partibus
animalium: Pietro Pomponazzi, Niccolo Leonico Tomeo and Agostino
Nifo.” In Steel, Guildentops, and Beullens, eds. 1999: 297–316.
Platt, A. 1912. De generatione animalium. Volume v of The Works of Aristotle
Translated into English, J. A. Smith and W. D. Ross, eds. Oxford: Clarendon
Press.
Popa, T. 2009. “Aristotle on Pure and Simple Stuff.” Rhizai 6.2: 127–160.
2014. “Scientific Method in Meteorology iv.” HOPOS: The Journal of the
International Society for the History of Philosophy of Science 4: 306–334.
Prior, A. and Prior, M. 1955. “Erotetic Logic.” Philosophical Review 64: 43–59.
Quarantotto, D. 2005. Causa Finale Sostanza Essenza in Aristotele. Napoli:
Bibliopolis.
Rapp, C. 2019. “The Planetary Nature of Mankind: A Cosmological Perspective
on Aristotle’s Anthropology.” In Keil and Kreft, eds. 2019, 77–96.
Rapp, C., and O. Primavesi, eds. 2020. Aristotle’s De Motu Animalium.
Proceedings of the xixth Symposium Aristotelicum. With an edition of the
Greek Text by O. Primavesi. Oxford University Press.
Reeve, C. D. C. 2012. “Aristotle’s Method of Philosophy.” In Shields ed. 2012:
150–170.
Rorty, A. ed. 1980. Essays on Aristotle’s Ethics. Berkeley: University of California
Press.
Rose, V. 1886. Aristotelis: Qui Ferbantur Liborum Fragmenta. Leipzig: Teubner.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
304 Bibliography
Rosenberg, A. 1994. Instrumental Biology or the Disunity of Science. University of
Chicago Press.
Ross, W. D. 1924. Aristotle’s Metaphysics: A Revised Text with Introduction and
Commentary. (2 vols.). Oxford: Clarendon Press.
1936. Aristotle’s Physics: A Revised Text with Introduction and Commentary.
1949. Aristotle’s Prior and Posterior Analytics: A Revised Text with Introduction
and Commentary. Oxford University Press.
1955. Aristotle: Parva naturalia (Revised Text with Introduction and Commentary).
Oxford: Clarendon Press.
1961. Aristotle: De anima. Oxford University Press.
Salmieri, G. 2007. “Aristotle and the Problem of Concepts.” PhD dissertation,
University of Pittsburgh.
2009. “Aristotle’s Non-dialectical Methodology in the Nicomachean Ethics.”
Ancient Philosophy 29: 311–335.
2014. “Aristotelian epistêmê and the Relation between Knowledge and
Understanding.” In Lennox ed. 2014d.
ed. Forthcoming. Knowing and Coming to Know. Pittsburgh: University of
Pittsburgh Press.
Salmon, W. 1990. Four Decades of Scientific Explanation. Minneapolis: University
of Minnesota Press.
Sauvé Meyer, S. 1992. “Aristotle, Teleology, and Reduction.” Philosophical Review
101: 791–825.
1994. “Self-movement and External Causation.” In Gill and Lennox, eds. 1994:
65–80.
Scaltsas, T., D. Charles, and M. L. Gill, eds. 1994. Unity, Identity and Explanation
in Aristotle’s Metaphysics. Oxford University Press.
Schaffner, K. 1993. Discovery and Explanation in Biology and Medicine. University
of Chicago Press.
Scharle, M. 2015. “Teleology, Necessity, and Explanation in Aristotle’s
Meteorologica.” In Ebrey ed. 2015: 79–99.
Schickore, J., and F. Steinle, eds. 2006. Revisiting Discovery and Justification:
Historical and Philosophical Perspectives on the Context Distinction. Berlin:
Springer.
Schiefsky, M. 2005. Hippocrates, On Ancient Medicine. (Studies in Ancient
Medicine, vol. 28) Leiden and Boston: Brill.
Schmalzriedt, E. 1970. Peri Phuseos: Zur Frühgeschichte der Buchtitel. Munich:
Fink.
Schofield, M. 2012. “Pythagoreanism: Emerging from the Presocratic Fog.” In
Steel ed. 2012, 141–166.
Sharples, R. W. ed. 2005. Philosophy and the Sciences in Antiquity. Aldershot:
Ashgate Publications.
Shields, C. ed. 2012. The Oxford Handbook of Aristotle. Oxford University Press.
2016. Aristotle: De anima. Oxford: Clarendon.
Sidgwick, H. 1874. The Methods of Ethics. London: MacMillan & Co.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
Bibliography 305
Sisko, J. 1999. “On Separating the Intellect from the Body: Aristotle’s De anima III
4. 429a10-b5.” Archiv für Geschichte der Philosophie 81: 249–267.
Smith, J. A. 1931. De anima. Volume III of The Works of Aristotle Translated into
English. W. D. Ross ed. Oxford: Clarendon Press.
Smith, W. D. 1979. The Hippocratic Tradition. Ithaca, ny: Cornell University
Press.
Sorabji, R. 1989. “The Greek Origins of the Idea of Chemical Combination: Can
Two Bodies Be in the Same Place?” In Cleary and Shartin, eds. 1989: 35–63.
Sosa, E., J. Kim, J. Fantl, and M. McGrath, eds. 2009. Epistemology: An Anthology,
2nd ed. Oxford: Blackwell.
Steel, C., ed 2012. Aristotle’s Metaphysics Alpha: Symposium Aristotelicum. Oxford
University Press.
Steel, C., G. Guildentops, and P. Beullens, eds. 1999. Aristotle in the Middle Ages
and Renaissance. Leuven University Press.
Steinfartz, S., S. Vicario, J. W. Arntzen, and A. Caccone. 2006. “A Bayesian
Approach on Molecules and Behavior: Reconsidering Phylogenetic and
Evolutionary Patterns of the Salamandridae with Emphasis on Triturus
Newts.” Journal of Experimental Zoology. (Molecular Developmental
Evolution) 306(2): 139–162.
Stocks, J. L. 1930. De caelo. Volume II of The Works of Aristotle Translated into
English,. W. D. Ross ed. Oxford: Clarendon Press.
Tipton, J. A. 2014. Philosophical Biology in Aristotle’s Parts of Animals. Dordrecht:
Springer.
Tredennick, H. 1960. Aristotle: Posterior Analytics. Loeb Classical Library.
Vitelli, H., 1887. Philoponi in Physicorum libros tres priores commentaria. Vol. 16 of
Commentaria in Aristotelem Graeca. Berlin: Reimer.
Wallace, W. ed. 1986. Reinterpreting Galileo. Washington, D.C.: Catholic University
of America Press.
Waterfield, R. 1996. Aristotle: Physics. Oxford University Press.
Waterlow [Broadie], S. 1982. Nature, Change, and Agency in Aristotle’s Physics:
A Philosophical Study. Oxford: Clarendon Press.
Whiting, J. 2001. “Strong Dialectic, Neurathian Reflection, and the Ascent of
Desire: Irwin and McDowell on Aristotle’s Methods of Ethics.” Proceedings
of the Boston Area Colloquium in Ancient Philosophy 17: 61–116.
Whitteridge, G. 1981. William Harvey: Disquisitions Touching the Generation of
Animals. London: Blackwell Scientific Publications.
Wians, W. ed. 1996a. Aristotle’s Philosophical Development: Problems and Prospects.
Lanham, md: Rowman & Littlefield.
1996b. “Scientific Examples in the Posterior Analytics.” In Wians ed. 1996a:
131–150.
Wilson, M. 2000. Aristotle’s Theory of the Unity of Science. University of Toronto
Press.
2013. Structure and Method in Aristotle’s Meteorologica: A More Disorderly
Nature. Cambridge University Press.
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
306 Bibliography
Witt, C. 1994. “The Priority of Actuality in Aristotle.” In Scaltsas, Charles, and
Gill, eds. 1994: 215–228.
Woods, M. 1992. Aristotle: Eudemian Ethics, Books i, ii and viii. Oxford:
Clarendon Press.
Zabarella, G. 1582. In duos Aristotelis libros Posteriores Analyticos comentarii,
Venice. (Reprinted in W. Risse ed. Opera logica. Hildesheim: Olms, 1966)
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
Index Locorum
Alexander of Aphrodisias I.4, 73b32-74a3, 47

Comm. on Aristotle’s Meteorology (in Mete.) I.5, 74a4-10, 48
179.3-5, 229 I.5, 74a10-11, 48 n. 18
I.5, 74a11-12, 48 n. 18
Aristotle I.5, 74a17-25, 57
Categories (Cat.) I.5, 74a25-30, 152
6, 4b20-22, 202 n. 6 I.5, 74a36-b4, 48
6, 5a1-5, 202 n. 6 I.6, 74b24-26, 46 n. 15
Prior Analytics (APr.) I.6, 75a29-37, 179
I.1, 24a10-11, 51 I.7–13, 114–118
I.26, 43a16-24, 51 I.7, 75a38-b5, 111
I.27-30, 51, 56 I.7, 75a38-39, 115
I.27, 51 I.7, 75b2-9, 116
I.27, 43b1-37, 52 n. 28 I.9, 75b37-76a9, 47 n. 17
I.28, 44a37-b6, 52 n. 28 I.9, 75b42-76a7, 116 n. 41
I.28, 44b20-36, 52 n. 28 I.9, 76a9-15, 117
I.30, 46a17-22, 15, 34 I.9, 76a22-25, 116 n. 41, 117
I.30, 46a22-27, 269 I.12, 77a40-b15, 23 n. 22
I.30, 46a19-22, 160 I.13, 78a30, 60 n. 45
II.21, 67a21, 45 n. 12 I.13, 78a36, 60 n. 45
II.23, 68b13-14, 15 n. 8, 22 I.14, 79a18-20, 103 n. 11
Posterior Analytics (APo.) I.18, 81b6-9, 22
I.1, 71a8, 28 I.24, 85a13-86a30, 45 n. 13
I.1, 71a17-30, 12 I.24, 85a13-17, 60 n. 45
I.1, 71a28-29, 12 I.24, 85a30, 60 n. 45
I.1, 71a29, 45 n. 12 I.31, 87b39-88a4, 44 n. 11
I.1, 71a29-30, 13 n. 3 I.33, 88b36-37, 13
I.2, 71b9, 12 II, 13–15, 37, 40–46
I.2, 71b15, 12 II.1–13, 14
I.2, 71b18, 12 n. 2 II.1–10, 179–180
I.2, 71b20-23, 46 II.1, 26
I.2, 71b26-72b3, 46 II.1, 89b23-25, 13
I.3, 72b7-16, 21 II.1, 89b23-24, 38, 101
I.3, 72b16-18, 21 II.1, 89b24-25, 38
I.3, 72b19-25, 179 II.2, 265–267
I.3, 72b23-25, 14 II.2, 89b37-90a1, 40,
I.3, 72b23-24, 13 265
I.3, 72b25-73a6, 21 II.2, 90a5-7, 41
I.4–5, 53 II.2, 90a26-29, 44 n. 11
I.4, 73b25-28, 48 n. 18 II.3–7, 41–42
I.4, 73b25-26, 49 II.3, 90a35-38, 41
307
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
308 Index Locorum
Posterior Analytics (APo.) (cont.) I.7, 189b34-190a13, 207
II.8-10, 42–44 I.7, 190b17-18, 33
II.8, 93a1-3, 42 I.9, 192a34-36, 129
II.8, 93a16-20, 43 II.1, 193a28-31, 126 n. 14
II.8, 93a17-18, 44 n. 11 II.2, 124–130
II.8, 93a21-27, 181 II.2, 193b22-30, 125
II.8, 93a21-24, 43 II.2, 193b36-194a1, 125
II.8, 93a22-24, 33 II.2, 194a1-7, 125
II.8, 93a24-25, 187 n. 24 II.2, 194a12-15, 126
II.8, 93a26-27, 44 II.2, 194a14-15, 150
II.8, 93a35-39, 44 II.2, 194a15-27, 104
II.8, 93b1-3, 44 II.2, 194a15-17, 127
II.8, 93b3-7, 44 II.2, 194a21-27, 128
II.10, 94a1-2, 42 n. 8 II.2, 194b9-13, 128
II.13, 46–63 II.2, 194b14-15, 128
II.13, 96b15-97b6, 51 II.3, 131–134
II.14–18, 14, 45, 50–59, 278–279 II.3, 194b16-23, 131
II.14, 98a1-11, 52–53 II.3, 194b17-20, 39
II.14, 98a13-19, 54 II.3, 194b18-21, 51
II.14, 98a20-23, 55 II.3, 194b33-35, 39
II.15, 98a28-29, 58 II.3, 195a27-b21, 132
II.15, 98a31-33, 58 II.3, 195b21-30, 132
II.16, 98a35-b24, 102 II.3, 195b27-28, 133
II.16, 98a35-b4, 59 II.7, 134–136
II.16, 98b4-5, 60 II.7, 198a14-16, 40
II.16, 98b5-10, 60 II.7, 198a21-24, 134
II.16, 98b10-17, 60 II.7, 198a24-26, 134
II.16, 98b32-33, 61 II.7, 198a28-31, 135
II.16, 98b33-37, 61 II.7, 198a33-b1, 135
II.17, 99a8-10, 57 II.7, 198b3-4, 136
II.17, 99a16-29, 62, 102 II.7, 198b4-9, 136
II.17, 99a30-b7, 62 n. 48 II.8–9, 136–139
II.19, 11, 14, 26, 30 II.8, 198b4, 136
II.19, 100a4-b4, 30 II.8, 199a30-32, 129, 137
II.19, 100a8-9, 14 II.9, 200a30-35, 137
II.19, 100b3-5, 15 n. 8 II.9, 200a30-34, 165 n. 38
II.19, 100b14-15, 14 n. 7 II.9, 200b4-8, 138
Topics (Top.) III.1, 200b12-15, 121
I.1, 100a18-20, 92 III.1, 200b12-14, 139
I.1, 100a18, 73 III.1, 200b12-13, 73
I.2, 101a25-b3, 93 III.1, 200b21-25, 139
I.2, 101a29, 93 III.3, 202a13-21, 241
I.3, 101a36, 27 n. 26 III.4, 202b35-36, 139 n. 38
I.3, 101a36-b4, 93 III.5, 204a34-b11, 139 n. 36
VIII.3, 158b4, 27 n. 26 IV.1, 208a27-30, 139 n. 38
VIII.12, 162b7-12, 92 IV.4, 210b32, 139 n. 38
Physics (Ph.) IV.6, 213a12-14, 139 n. 38
I.1, 33 IV.10, 217b29-32, 139 n. 38
I.1, 184a10-16, 123 VIII.1, 251a5-8, 139 n. 36
I.1, 184a10-11, 73 VIII.3, 253a32-b10, 139 n. 36
I.1, 184a16-18, 123 VIII.3, 254a30-b1, 91
I.1, 184a21-b3, 97 VIII.7, 261a27-31, 139 n. 36
I.2, 184b25-185a20, 23 n. 22 De caelo (Cael.), 5
I.4–8, 33 I.1, 268a1-6, 201
I.4, 187a12-23, 103 n. 8 I.2, 268b11-16, 202
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
Index Locorum 309
I.2, 269a2-7, 203 IV.7, 383b18-20, 243
II.2, 205–207, 217–221 IV.7, 383b25-384a2, 244
II.2, 284b6-10, 204 IV.7, 384a3-4, 243
II.2, 284b12-18, 204 IV.7, 384a25-33, 238
II.2, 284b18-25, 205 IV.8, 232
II.2, 284b27-30, 218 IV.8, 384b26-27, 207
II.2, 284b32-34, 218, 220 IV.8, 384b28-30, 243
II.2, 285a10-22, 215 IV.8, 385a1-11, 235 n. 15
II.2, 285a27-31, 218 IV.8, 385a10-18, 245
II.2, 285a29-30, 176 IV.10, 388a25-26, 243
II.2, 285a31-b1, 219 IV.10, 388a10-26, 248
II.2, 285b1-8, 220 IV.10, 388a21-25, 236
II.2, 285b8-286a1, 221 IV.11, 389a29-32, 236
II.3, 286a3-7, 222 IV.12, 389b23-28, 247
II.5, 287b28-288a2, 222 IV.12, 389b28-29, 249
II.5, 288a2-12, 225 IV.12, 390b3-14, 260–263
II.12, 291b24-31, 223 IV.12, 390b14-22, 247, 262
II.12, 292a14-22, 223 De anima (de An.)
II.12, 292a20-22, 176 I.1, 175–182
III.7–8, 306a1-307b17, 224 n. 50 I.1, 402a1-11, 103 n. 10
III.7, 306a5-17, 27 I.1, 402a3-4, 175
On Generation and Corruption (GC) I.1, 402a6-7, 175
I.2, 316a5-10, 24 I.1, 402a7-10, 177
I.2, 316a6-12, 28 I.1, 402a11-22, 178–179
II.2, 329b18-20, 234 I.1, 402a21-22, 178–180
II.2, 330a24-26, 234 I.1, 402a22-25, 180
II.7, 232–234 I.1, 402a25-26, 180
II.7, 334a22-26, 232 I.1, 402b1-5, 184 n. 19
II.7, 334b3-7, 233 I.1, 402b5-9, 184 n. 20
II.7, 334b24-30, 233 I.1, 402b22-25, 187
Meteorology (Mete.) I.1, 403a7-10, 188
I.1, 112–113 I.1, 403a16-17, 189
I.1, 338a20-339a10, 103 n. 10, 103–104, I.1, 403a27-28, 189
174 I.1, 403b10-19, 188 n. 28
I.1, 338a20-25, 142 I.1, 403b12-14, 192
I.1, 338a25-26, 73 I.1, 403b15-16, 192
I.1, 339a5-9, 112, 143 I.1, 403b17-19, 193
I.2, 109 I.3, 406b26-28, 103 n. 8
I.3, 339a33-34, 109 II.1, 412a6-9, 182
IV.1–3, 235–238 II.1, 412a9-11, 183
IV.1, 378b26-28, 236 II.1, 412a19-21, 182
IV.1, 378b31-34, 236 II.1, 412a21-23, 184
IV.2, 379b18-22, 232 n. 8 II.1, 412b18-20, 185
IV.2, 379b25-28, 232 n. 8 II.1, 412b23-25, 185
IV.2, 380a7-9, 232 n. 8 II.2, 413a13-16, 185
IV.3, 380a13-16, 232 n. 8 II.2, 413b13-16, 185
IV.3, 380a31-33, 232 n. 8 II.2, 413b24-27, 185
IV.3, 380b13-16, 255 II.2, 413b25-28, 188 n. 28
IV.3, 380b20-24, 255 II.4, 415b8-27, 177 n. 9
IV.3, 380b27-28, 255 II.4, 415b15-21, 183
IV.4, 382a8-13, 239 II.4, 416a14, 274
IV.5–7, 238–245 III.3, 427b14-16, 188 n. 27
IV.5, 382a22-27, 239 III.9, 432a15-16, 35 n. 35
IV.5, 382a27-b1, 240 III.9 , 432b26-29, 196
IV.5, 382b5-7, 236 III.10, 433a8-17, 196
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
310 Index Locorum
De anima (de An.) (cont.) I.6, 491a7-14, 269
III.10, 433a9-12, 188 n. 27 I.6, 491a9-14, 187 n. 23
III.10, 433a29-30, 196 I.17, 497a32, 285 n. 37
III.10, 433b13-18, 196 III.2–4, 275 n. 17
Parva naturalia III.3, 512b13-513a8, 70 n. 8
De sensu (Sens.) III.3, 513a35-b25, 283 n. 30
1, 436a1-17, 272 III.6, 515b30-516a6, 238 n. 22
1, 436a1-11, 190 IV.1, 523b21-33, 157
1, 436a12-17, 190 IV.1, 524a25-32, 157
1, 436a18-b2, 117 n. 44 IV.1, 524b23-30, 55 n. 36
2, 437b10-14, 103 n. 8 IV.7, 532b1, 55 n. 36
De memoria (Mem.) VI.3, 561a4-562a22, 81 n. 23
1, 450a1-6, 285 n. 38 VI.3, 561a4-15, 169 n. 47
De somno (Somn.Vig.) VII(VIII).2, 589b26-27, 281 n. 29
3, 458a25-32, 149 De partibus animalium (PA)
On Length and Shortness of Life (Long.) I.1 , 146–149
6, 467b6-9, 270 I.1, 639a1-12, 103 n. 10
6, 467b7-9, 162 I.1, 639a1-5, 89, 95, 143
On Youth and Old Age (Juv.) I.1, 639a1-2, 73
1, 467b10-13, 270 I.1, 639a5-16, 90
3, 468b31-469a1, 275 I.1, 639a12-13, 144
3, 469a5-10, 277 I.1, 639a13-15, 146
6, 470b1-5, 271 I.1, 639a15-19, 147
On Respiration (Resp.) I.1, 639a19-b3, 58 n. 42
1, 470b6-12, 267, 271 I.1, 639a25-b3, 148
1, 470b12-13, 268 I.1, 639b3-5, 147
3, 471b23-29, 129, 267 I.1, 639b7-10, 160
4, 471b30-472a3, 272–273 I.1, 639b11-14, 163
5, 472b24-29, 273 I.1, 639b14-16, 84, 88
5, 472b27-29, 270 I.1, 639b18-19, 84
7, 473a15-17, 273 I.1, 639b21-640a1, 165 n. 38
7, 474a23-24, 274 I.1, 639b22-640a9, 170 n. 49
8, 474a25-28, 274 I.1, 640a1-9, 165
8, 474a28-b3, 274 I.1, 640a10-b3, 81
8, 474b4-9, 275 I.1, 640a13-15, 165
8, 474b7-13, 284 I.1, 640a15-19, 83 n. 29, 166
9, 279 I.1, 640a20-27, 164
10, 475b16-23, 279 I.1, 640a33-b3, 166
10, 476a1-5, 280 I.1, 640b4-641a32, 170 n. 49
10, 476a7-15, 280 I.1, 640b28-29, 126 n. 14
12, 476b13-14, 268 n. 6 I.1, 641a32-b10, 194–198
15, 478a12-14, 283 n. 30 I.1, 641a33-b23, 170 n. 49
15, 478a21-28, 283 I.1, 641b4-12, 194
16, 281–288 I.1, 642a1-13, 170 n. 49
16, 478a28-34, 284 I.1, 642a9-13, 138 n. 35, 251
16, 478a34-b7, 285 I.1, 642a31-b4, 264
16, 478b7-15, 286 I.2–3, 642b5-644a11, 170 n. 49
16, 478b15-21, 287 I.2–3, 642b5-643a23, 150
21, 480b6-20, 286–287 I.2, 642b14-16, 52 n. 29
21, 480b22-31, 117 n. 44 I.3, 643a24-27, 150
Historia animalium (HA) I.4, 149–157
I.5, 490a5-8, 150 I.4, 644a17-22, 151
I.5, 490a13-16, 150 I.4, 644b1-8, 52 n. 29
I.6, 269–270 I.4, 644b1-7, 153
I.6, 490b7-9, 52 n. 29 I.4, 644b8-16, 154
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
Index Locorum 311
I.4, 644b15-20, 145 De incessu animalium (IA)
I.5, 644b22-645a10, 1, 704b8-11, 270
171–172 2–5, 214–215
I.5, 644b22-645a1, 224 2, 208–209
I.5, 644b22-25, 161 2, 704b12-705a2, 208
I.5, 644b31-645a4, 161 2, 704b12-17, 130
I.5, 645a1-4, 5 2, 704b12-14, 200
I.5, 645a4-10, 161 2, 704b18-22, 206
I.5, 645a21-26, 172 4–5, 209–213
I.5, 645b19-20, 183 4, 705a26-b2, 210
I.5, 645b30-32, 274 n. 16 4, 705b8-13, 211
I.5, 646a1-4, 173 4, 705b31-706a10, 211
II.1, 646a8-11, 269 4, 706a10-12, 212
II.1, 646a16-20, 250 4, 706a18-22, 212
II.1, 646b14-27, 250 4, 706a24-26, 212
II.1, 646b27-30, 252 5, 706a26-31, 212
II.2, 647b10-18, 251 5, 706b2-9, 213
II.2, 649a29-34, 252 5, 706b3-6, 217
II.2, 649a33-34, 109 5, 706b11-13, 213
II.3, 649b20-650a2, 252 5, 706b12-13, 207
II.3, 650a2-32, 253 5, 706b14-16, 213
II.3, 650b2-13, 253 6–7, 706b17-707a23,
II.4, 650b14-18, 253 214
II.4, 651a12-17, 253 6, 706b28-33, 219
II.5, 651a20-36, 254 De generatione animalium (GA)
II.5, 651a26-36, 254 I.1, 715a1-18, 103 n. 10
II.7, 652b19-23, 255 I.1, 715a7-18, 167
II.7, 653a1-10, 117 n. 44 I.20, 728a18-21, 207
II.7, 653a20-27, 255 I.33, 731a30-33, 17 n. 12
II.7, 653a32-36, 256 II.1, 734b19-735a4, 85
II.7, 653b3-5, 255 II.1, 734b24-735a4, 260–263
II.8, 654a19-22, 56 n. 37 II.1, 734b31-36, 231
II.9, 655b8-13, 138 II.1, 735b1-37, 244 n. 39
II.16, 18 n. 13 II.3, 736b5-30, 195 n. 40
III.1, 661b17-22, 138 II.3, 736b27-29, 174
III.2, 662b35-663a7, 55 II.4, 740a17-24, 276
III.2, 663b22-24, 127 n. 14 II.4, 740b12-741a3, 85
III.2, 663b28-664a2, 55 II.5, 741b15-17, 78 n. 18
III.4–5, 275 II.5, 741b18-24, 79
III.4, 665b31-666a11, II.6, 77–89, 256–258,
275–276 260–261
III.5, 668a27-b3, 276 II.6, 741b27-32, 82
III.9, 672a1-9, 254 II.6, 741b31-742a16, 79
III.14, 674a30-b7, 55 II.6, 741b33-37, 82
IV.3, 677b14-16, 254 II.6, 741b37-742a16, 82
IV.3, 677b31-33, 254 II.6, 742a16-36, 83–86
IV.5, 680a1-4, 285 n. 37 II.6, 742b3-12, 169
IV.10, 689a9-14, 110 II.6, 742b6-12, 79
De motu animalium (MA) II.6, 743a1-21, 257
1, 698a1-10, 103 n. 10 II.6, 743a3-8, 109
1, 698a1-6, 149 II.6, 743a21-23, 82 n. 26
1, 698a8-11, 109 II.6, 743a17-21, 227
1, 698a11-14, 116 n. 41 II.6, 743a36-b5, 85, 257
6, 700b17-20, 188 n. 27 II.6, 743a36, 207
6, 700b19-22, 35 II.6, 743b16-18, 257
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
312 Index Locorum
De generatione animalium (GA) (cont.) VI.3, 1139b26-29, 22
II.6, 744b21-28, 86 VI.5, 143 n. 4
II.7, 746a14-16, 285 n. 37 VIII.8, 1151a15, 27 n. 26
II.8, 748a7-16, 133 Eudemian Ethics (EE)
II.8, 748a7-11, 141 I.1, 1214a8-15, 96
III.8, 758a28-29, 112 n. 30 I.6, 1216b35-1217a10,
IV.4, 770b9-18, 127 n. 14 94–95
V, 258–260 I.6, 1216b35-39, 37
V.1, 778a16-18, 258 I.6, 1217a4-10, 95
V.1, 778a30-32, 258 I.6, 1217a8-10, 144
V.I, 778a34-b1, 258 Politics (Pol.)
V.1, 778b2-7, 83 n. 29 I.1, 1252a17-18, 112 n. 30
V.1, 778b13-19, 258 I.1, 1252a17-23, 96
V.3, 782a20-24, 259 II.1, 1260b34-36, 97
V.3, 782b25-28, 259 III.8, 1279b12-15, 96
V.3, 783a12-18, 259 III.11, 1282a1-7, 90,
V.8, 258 n. 69 143 n. 5
Problems (Pr.), 51 n. 26 IV.2, 1289b12-26, 97
Metaphysics (Metaph.) IV.4, 1290b21-38, 97
A.1, 25 VII.1, 1324a1-3, 96
A.1, 980a27, 35 n. 35 VII.2, 1324a19-24, 96
A.1, 981b10-13, 17 n. 12 Rhetoric (Rh.)
A.2, 983a21-23, 11 I.2, 1356b18-22, 72
A.2, 983a22-23, 75 Poetics (Po.)
α.1, 995a12-20, 99 I.1, 1447a8-14, 97
Γ.1, 1003a25-26, 103 n. 11
Δ.4, 177 Diogenes Laertius
Δ.13, 1020a7-11, 202 n. 6 Lives of the Philosophers (D.L.)
Δ.13, 1020a11-14, 202 n. 6 III.60, 103 n. 8
Ε.1, 104–108 V.23, 52 n. 27, 72
Ε.1, 1025b8-10, 105 V.24, 52 n. 27
Ε.1, 1025b18-21, 105, 124 V.25, 52 n. 27
Ε.1, 1025b19-21, 104
Ε.1, 1025b28-30, 106 Hippocratic Writings
Ε.1, 1025b30-1026a6, 106 On Ancient Medicine (VM),
Ε.1, 1026a25-28, 103 n. 11 70–72
Ε.1, 1026a27-32, 105 1, 16–21, 70
Ζ.11, 1036b24-32, 125 2, 1–4, 70
Ζ.11, 1037a13-17, 126 2, 4–9, 71
Η.4, 1044a32-b20, 134 n. 24 2, 8, 71
Θ.8, 1050a3-15, 83 5, 9–11, 71
Θ.8, 1050a4-10, 84 9, 19–25, 71
Λ.7, 1072b25-30, 176 On Fleshes (Carn.)
Μ.1, 1076a8-14, 121 8, 238 n. 23
Μ.3, 1078a9-14, 145 n. 9
Μ.3, 1078a21-26, 145 n. 9 Parmenides
Μ.3, 1078a33-b2, 103 n. 11 Fragments (DK 28)
Nicomachean Ethics (EN) B1, 67 n. 3
I.1, 1094a1-3, 65, 73 B7, 67 n. 3
I.3, 1094b24-28, 90 B8, 67 n. 3
I.3, 1094b25-27, 3
I.3, 1094b28-1095a2, 90 Philoponus
I.7, 1098b2, 27 n. 26 Comm. on Aristotle’s De anima
I.13, 1102a23-32, 199 (in de An.)
II.5, 1105b20-28, 143 n. 4 44,22–26, 187 n. 25
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
Index Locorum 313
Plato 235c1-6, 68
Philebus (Phlb.) Statesman (Plt.)
54a5-11, 83 n. 29 260e8-261a1, 69
54c9-11, 83 n. 29 266d4-9, 69
55c-56e, 71 n. 11 270c3-5, 69
Republic (R.) 270d1-7, 69
VII, 521d, 67 270d9-e1, 69
VII, 532c7-d3, 67 286d5-9, 69 n. 5
VII, 532d8-e3, 67 Timaeus (Ti.)
VII, 533c7, 67 38e7, 176
Sophist (Sph.) 39e1-2, 176
218d1-6, 68 77a5-6, 176
. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

D 6 2 2: 23 6 2C 9CC , 42 3 :586 8 4 6 C6 9CC , 5 : 8
General Index
abstraction heart as source of, 274–278, 289

levels of, 1, 2, 6, 15–16, 53, 147–149, 152–153, 168, need for cooling of, 288, 290
179, 278, 282 with and without fibers, 237–238, 252–253
mathematical, 48–49, 108, 192 blood vessels, 85, 87–88, 256–257
Ackrill, J. L., 42 as containers of blood, 274–275, 277, 282
Alexander of Aphrodisias, 105, 218, 229, 234 as permeating viscera, 275–276
anatomy, comparative, 155, 162, 167, 273–277, carrying blood to gills, 285–286, 290
287–289 heart as source of, 275, 277, 289
and dissection, 274–275, 277, 283–286 in the lungs, 282–284
of blood vessels, 275–277 bodies, 232, 233–235, 236–237, 245–246
of gills, 282–287 elemental vs. compound, 228–229, 240–241
of heart, 87, 274–277 uniform vs. nonuniform, 247–249, 259,
of lungs, 279–281 260–263
functional, 281–284 Bogaard, P. A., 234
need for, 273–277 Bolton, R., 20, 24, 33–34, 77, 110, 161, 185, 202,
Anaxagoras, 103, 268, 272, 291 207, 220, 221, 222
animal locomotion, 18, 87, 115, 130, 147, 150, 185, brain, 288
211–212, 214–215 material make-up and function of, 255–256
in relation to movements of heavenly bodies, Broadie, S., 83, 194–195
218, 222–225 Bronstein, D., 14, 19, 32, 42, 45, 46
inquiry into, 210, 214–215 Burnyeat, M., 1, 19–20, 23, 29–30, 101, 113–114,
proper principles of, 208–209, 216–217 195, 197, 230, 241
source or cause of, 194–198, 214
archê, translation of, 84, 133, 175–177, 201, 213 capacity (dunamis)
Arntzen, J. W., 280, 281 active vs. passive, 236, 248, 258–259
autonomy and functions/activities, 126, 171, 220
of natural inquiries, 103, 108, 116, 118 and incapacity, 232, 245–246
of sciences, 6, 103, 112 generative, 85, 87, 168, 176
of study of reason, 199 of soul, 35, 78, 183, 185, 190, 195, 197, 198,
related to integration, 110–111, 118, 122 231, 284
Cartwright, N., 4
Bacon, Francis, 34, 49, 66 Caston, V., 148, 230
Balme, D. M., 2, 89, 144, 150, 153, 167–168 causal framework, 200, 239–240
Barnes, J., 12, 13, 48, 60, 114, 115 causes, 3, 101, 104, 168–169, 171–172, 196–198, 242,
blood 244–245, 248–249, 284
and animal character, 253 active vs. passive, 236–237, 240–241, 246
and fat, 254 and demonstrations, 11–12, 13–14, 29, 40, 46,
and omentum, 254 63, 101, 122, 266
as a uniform compound, 229, 230, 262 and effects, 59–62, 132–134
as final nourishment, 85, 87, 253, 274, 276, 288 and middle terms, 13, 40–42, 44, 50–51, 54,
as matter for parts, 78, 88 61–64, 266
314
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
General Index 315
as answers to dia ti? (‘on account of what?’, Dante Aligheri, 1
‘why?’), 39–40, 45, 51, 179 Davidson, D., 19, 35
as goals of inquiry, 11, 16, 17, 27, 30, 34, 41, definition, 34, 40, 163, 215
44–45, 57, 59–60, 64, 101, 122–123, as archê, 109, 137, 165
131–132, 133, 159–162, 266–267, 268, domain-specific, 34
269–270, 278–282 of natural substances, 107–108, 122, 125–126
efficient, 88, 163–165, 167–168, 189, 211–212, of respiration, 268, 290
228, 230–231, 236–237, 258, 290–291 of soul, 183–185, 186–187, 189
final, 39, 83–84, 88, 129–130, 136–137, 162–168, relation to cause, 16, 62, 84–85, 137, 179
189, 225, 252, 255, 268, 273, 280–281 relation to demonstration, 13–14, 16, 37, 41–42,
four kinds of, 39, 131–132, 134–136, 257 62–63, 179
knowledge of, 26–27, 33, 81, 122–123, 223–224 relation to division, 51, 179
material level, 230, 235–236, 253–254, 257–260 relation to essence, 11, 30
modes of, 132–133 Demiurge, in Plato’s Timaeus, 201
cephalopods (malakia, soft-bodied animals), 55 Democritus, 103, 128, 133, 225, 258, 267,
as a ‘great kind’, 159 272–273, 291
distinctive anatomy of, 82, 154–159 demonstration (apodeixis), 11–15, 84, 93, 115–118,
in HA, 157–159 165, 265–266, 269
sub-kinds of, 157 causal, 101, 108
change (metabolê), 140, 238, 240–241 definition and, 11, 40–42, 63, 178–180, 186
vs. motion (kinêsis), 124 induction and, 15, 27–29, 30–31
Charles, D., 14, 19, 30–32, 45, 46, 61, 62, 63, 88, partial (vs. unqualified), 28, 48, 152–153, 266
89, 122, 137, 151, 154, 185, 230 problems (problêmata) and, 46–63
Charlton, W., 195 Descartes, René, 66
coherentism, 18–19, 27 diagrams, Aristotle’s scientific use of, 52, 285
concoction (pepsis), 237 dialectic, 29, 187, 192, 198
and inconcoction, 232, 237, 238 in Plato, 66–67, 77
and nutrition, 252–255, 274 in the Topics, 92–94
of blood, 237–238 strong (Irwin), 19, 20–24
cooling directional dimensions (up/down, right/left,
and heating (as tools), 85–86, 109, 167, forward/backward), 200, 203, 205–205
230–232, 240–242, 244, 246, 252, applied to the cosmos, 214–228
256–257, 259–263 as presuppositions of natural inquiry, 206,
and respiration, 264, 271, 279–282 209–214, 228, See also presuppositions of
and the brain, 255–256 natural inquiry, directional
as function of lungs and gills, 282–291 dimensions as
Cooper, J., 233 in relation to dimensions of magnitude,
cosmology, 114 205–207
and astronomy, 43, 203 discovery
norms of inquiry for, 202, 215–221 and justification, 18–21, 32
right and left in, 203–221 of analogy of gills and lungs, 288
vs. zoology, 4–5, 209–211 of principles and causes, 30, 160, 278
cosmos dispositional properties (dunameis)
dimensions of, 203, 210, 213, 214–215 differential nature of, 230–231, 235, 245–247
motions of, 202–203 emergence of, 229–231, 238, 241
parts of, 202–203 list of 18 pairs of, 232, 245–246
cross-references, 7, 113 role in biology, 137, 232, 234, 249–251, 255–259,
between Cael. and IA, 204–205 261–262
between MA and IA, 114 dissection(s)
between MA and Ph. VIII, 109 and experience, 81, 89, 130, 171, 268
between PA and GA, 110 and grasping problems, 51–52, 53–54, 56, 80–81
between PA/GA and Mete., 109, 252–253, and study of generation, 276
256–258 and teleology, 275–276, 277–278, 286,
from PA/GA/PN to HA/Dissections, 274–277, 289, 291
283–285 Aristotle’s practice of, 283–284
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
316 General Index
dissection(s) (cont.) in relation to inquiry, 12, 47, 64, 79–80, 83
diagrams of, 283, 285 in relation to scientific principles, 15–16, 19, 21,
lost books of, 52, 159, 283–284, 291 24–27, 32, 33–34
division, 178 Plato’s view of, 67–70
and definition, 30, 42, 51–52, 56 explanation, 55–56, 100, 256
and inquiry, 51–54, 202–203, 251 A- vs. B-explanation, 52–53, 61
dichotomous, 149–150 and scientific problems, 57–59, 62
in Plato, 68–69, 77 causal, 51, 57, 85, 159, 196, 259–260, See also
multi-axis, 97, 149–151, 157–159 causes, and demonstrations
norms of, 150 domain-specific, 161–162, 166–167, 219, 224
dunamis. See capacity; dispositional properties; material level, 227, 228, 231, 253–254,
power/potential 257–258, 260
translation of, 238 mechanistic vs. thermodynamic, 236–245, 257
Düring, I., 237, 246 teleological, 87, 138, 162–164, 260–261
eclipse, 33, 43, 181 Falcon, A., 77, 102, 104, 113, 114, 161, 176, 201, 202,
as object of inquiry, 43–44, 59–60, 290 203, 207
definition and demonstration of, 63 Ferejohn, M., 14, 45, 47, 50, 61, 107, 122
knowledge of, 17–18, 43–44 first philosophy, 6, 23
no final cause of, 134 and separability, 128–129, 192, 197
of moon, 17–18, 44, 59, 134 and the soul, 192–194
elephant, trunk of, 18 vs. natural/second philosophy, 75, 99, 105,
embryology, 77–89, 169, 171, 256–257, 277 126, 128
emergence, 230 form(s), 78, 84, 85, 236–237, 240, 242
during embryogenesis, 80 as nature, 118, 125, 126–128, 136, 177, 249, 260
in GA, 231, 256–260 as that for the sake of which, 129, 134–138
in GC, 232–235 as what a thing is, 128, 134
in Mete. IV, 229–232, 236–247 in relation to matter, 104, 106–107, 108, 128,
in PA, 249–256 150, 171, 263
of dispositional properties, 230, 231 of kinds, 57–58, 61, 97, 147–149, 150–159, 184,
Empedocles, 103, 127, 165, 166, 232, 273, 291, 292 209, 255
empiricism, 16, 20, 25, 33–34, 234, See also Platonic, 66–68
experience (empeiria) soul as, 43, 106, 177, 182–184, 186, 189, 230
entelecheia. See fulfillment studied by natural science, 128–131, 134–135,
epistemic access, 5, 34, 65, 80–81, 102, 122, 141, 171 191, 193, 198, 202
epistemic map, 103–104 framework. See causal framework; erotetic
epistemology, 19, 24–27, 29–30, 141 framework
of inquiry, 30, 32, 82 Frede, D., 233, 234, 235
erotetic framework, 1, 12, 18, 37–40, 95, 102, 179 Frede, M., 19, 24–27, 32, 35, 105, 107, 194–195
and causal inquiry, 44–64, 278–279 Freudenthal, G., 230
and domain-specific erotetic structure, 180, fulfillment (entelecheia), 181
185–186, 235 in relation to activity (energeia), 181
and domain-specific norms, 82, 122–128, soul as, 183–184
169–170, 264–265 vs. potentiality, 181, 183
as provided by APo. II, 40–44, 121–122, 139, Furley, D., 218, 230, 247
200, 282 Furth, M., 235
eternal natural objects, 4, 109, 135, 203
and circular motion, 220 Galileo Galilei, 66
and unconditional necessity, 161, 164 Gilbert, S., 231
vs. perishable natural objects, 4, 161, 171–172, Gill, M.-L., 107, 195, 218, 235, 240, 247, 260,
185, 224 261–262, 263
exhalations, 243, 245 gills, 267–268, 279–280
experience (empeiria) analogical relation to lung, 280–281, 282
and expertise, 88–89, 267–268 function of, 282–288, See also cooling, as
and inexperience, 79, 267–268, 271 function of lungs and gills
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
General Index 317
goal-causation, 162–165, See also causes, final; Plato on, 67–70
priority, of final cause to efficient cause; relation to methodos, 11, 73–77, 79–80, 92–96,
teleology 97, 118, 130, 141–142, 178–179, See also
Gotthelf, A., 2, 18, 77, 85, 136, 149, 153, 154, 254, methodos
258, 279 relation to paideia, 89–91, 95, See also paideia
Gottschalk, H. B., 246 zoological, 6–7, 81–83, 87–89, 146–170, 172,
great kinds (megista genê), 52, 55, 149–159 279, 281–282, 289–291
intuitionism, 20–22
hair, explanation of variation in, 259–260 Irwin, T., 12, 18–25, 27, 32, 33, 35, 143,
Hannon, J. F., 267, 274 175, 189, 192
Harvey, William, 66, 115, 169
Hasper, P. S., 27, 47, 48, 57, 58 Kant, Immanuel, 206
heart, 53–54, 78, 255, 277–278 Karbowski, J., 94, 207
analogue of, 87 kind-crossing, 115–116, See also subordinate
as source of blood vessels, 256, 275–277 sciences
functions of, 87, 274–278 knowledge (gnôsis), 12, 29
in relation to lungs and gills, 279–290 perception as a form of, 17
role in biological development, 79, 87, 169, kordulos, 280–281
275–277 Kosman, A., 1, 19, 27–29, 30, 218
Hippocratic texts, 65, 69–72 Kung, J., 114, 115, 117
horns
correlated with multiple stomachs and few LeBlond, J.-M., 207
teeth, 54–55 Lloyd, G. E. R., 213, 237
correlated with suet, 254 locomotion. See animal locomotion
earthen nature of, 254, 257 logos (account), 85, 88, 92, 163–164, 271–272
Humphreys, P., 230 as a starting point, 95, 163
in relation to form, 78
induction (epagôgê), 15, 17, 48, 63, 223, 242, 280 of goal or end, 163
and justification, 35–36 lung(s)
Bolton on, 33–34 extension of, 268–269, 279–281
Burnyeat on, 29 function of, 290, See also cooling, as function
Charles on, 30–31 of lungs and gills
Irwin on, 20–23 relation to gills, 267–268, 279, 280–281, 282
Kosman on, 27–29 relation to heart, 282–284
norms for, 45
inquiry (zêtêsis), 1–4 Machamer, P., 231
causal, 59–61, 169–170, 268, 281–282 magnitude, 201–203
erotetic framework for, 41, 45, 49–50, 63, principles of. See directional dimensions
131–132, 140, 170, 180, 185, 264–267, 282 three dimensions of, 201–202, 206, 216, 217
four modes of, 26–27, 38–40, 94–95, 101–102, Mahoney, E., 194
265–270, 278–279, 289–291 materials. See bodies
goal of, 1, 13, 15–17, 63, 101–102, 104–105, 118, matter
246–247, See also scientific knowledge, as as cause, 137, 258
goal of inquiry as nature, 6, 125–126, 129, 137
in On Ancient Medicine (VM), 71 blood as, 253
inductive, 27, 30–31, 33–34, 45 hot/cold, moist/dry as, 233, 235, 236, 240,
into the soul, 6–7, 175–198 248–250
natural, 5–6, 99–100, 102, 103–110, 118, in relation to form, 104, 127–128, 138, 171,
123–124, 126, 129–130, 134–139, 144, 182–183
190–193, 197–199, 206–209 studied by natural science, 106–108, 128–131,
norms of, 1–6, 17–18, 34, 49, 54, 87–88, 137–138, 150, 192–194, 202, 262
148–149, 160–162, 169–171, 181–182, teleological dependence on form, 127–129, 138,
216–221 191, 198
order of, 7, 40–45, 161–162, 167–168, 198–207, McDowell, J., 25
214–217, 228–230, 250–260, 263, 271 McKirahan, R., 13, 48, 53, 114
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
318 General Index
metaphysics. See first philosophy included in definition, 137–138
meteorology, 104 material, 230–231, 259
relation to biology, 142–143, 246–260 Needham, P., 234
relation to cosmology, 109, 111–112 Newman, W., 237
methodos, 65–98, See also paideia, relation to norms (of inquiry). See also inquiry, norms of
methodos; inquiry, relation to methodos domain-neutral, 49, 65, 91
Aristotle on meaning of, 72–77 domain-specific, 16, 17–18, 65, 70, 73, 81–82,
lost treatise on, 72–73 87–88, 91–92, 94, 99–102, 122–123, 129,
of animals, 87–89, 141–173 137, 145–171, 181, 184, 190, 193, 204,
of nature, 112, 121–140, 193, 208 278–289
of soul, 174 framework, 49, 53, 56, 63–64, 268–270,
Plato on meaning of, 66–70 278–282, 289–291
middle term regarding order of inquiry, 228–229, 247–263
as object of inquiry, 42, 54, 56–59, 61–62, Nussbaum, M., 35, 113–117
265–267
in demonstrative syllogism, 13, 30, 40–41, 50 olive oil, aporia concerning, 243–244
referring to cause, 30, 40–41, 61, 63
relative to a problem, 50–51 paideia
Mirus, C. V., 247, 248, 249 general vs. specific, 90, 143–144, 146
mixed sciences. See subordinate sciences relation to knowledge, 143, 144–145
more and the less, the, 151–152, 258 relation to methodos, 89–91, 95, 97, 145
and excess and defect, 155 Parmenides, 67, 72
as continous variation, 155–156 Perfetti, S., 194
forms of kinds differing by, 151–152, 154–156 Philoponus, 136, 187, 194
in Plato’s Philebus, 155 Plato, 4, 19, 21–22, 28, 65, 66–70, 74, 77, 83,
vs. analogy, 151–152, 154–155 149–150, 155, 176, 193, 291
motion (kinêsis), 103–104, 114–115, 139–140, Popa, T., 242, 246, 254, 257, 263
204, 250 power/potential (dunamis), 78
eternal, 4, 109, 135, 219–220 for form, 85
source of, 78, 82–83, 134–136, 163, 201–203, of elements or compounds, 236, 250
211–213, 218–220, 240 vs. actuality/fulfillment, 79, 83–84, 132, 133,
nature as a, 121, 124, 195–196, 198 180–181, 184
precision (akribeia), 71–72, 90, 199, 284
natural beings. See also definition, of natural presuppositions, 13
substances of natural inquiry, 208–209
as ‘snub-like’, 106–108, 124–131 directional dimensions as, 208–214, See also
natural inquiry. See inquiry, natural directional dimensions, as
natural laws, Aristotelian equivalents of presuppositions of natural inquiry
in GA, 257 primary powers
in Mete. IV, 241–245 as matter of composites, 249–250
in PA, 252, 255 emergent properties follow from, 234–235,
natural science. See also methodos, of nature 250, 251
demarcation of, 102–108 hot/cold, moist/dry as, 234, 235–236
program for (in Mete. I.1), 109–115, 142–143, 145 priority
unity of, 110–115 among directional dimensions, 207,
nature. See also form, as nature; matter, as nature; 214–217, 219
methodos, of nature; natural science; among inquiries, 228–229, 263
as source of motion or change, 85–86, 105, among parts, 86–88, 169
118, 124 of actuality/fulfillment to potentiality, 83–84
does nothing in vain, 86 of being to coming to be, 80–81, 83, 162
spoken of in two ways, 104 of final cause to efficient cause, 83–88, 137,
necessity 162–164, 166, 263, 282, 289
and teleology, 136–139, 257–258 of form to matter, 107–108
conditional vs. unqualified, 136–137, 161, problems (problêmata), 29–30, 51–56
164–165, 247, 251–252, 254, 258 and coextensive predication, 56–57, 60–61
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
General Index 319
and search for causes, 44–46, 55, 59, 62–63 as knowledge of causes, 26, 81, 131, 152, 159, See
and why-questions, 51, 57–59, 63 also causes, knowledge of
in APo. II.14–18, 14, 46–51 as knowledge of essences, 122, 159, 179–180,
in APr. I.26–30, 50–51 186–187
in Topics, 92–93 demonstrative, 12–13, 152, See also
puzzle (aporia) demonstration
about a single science of nature, 127, 131 domain-specific, 17
about affections (pathê) of soul, 41–42 nous as source of, 13–15
about demonstrability of definitions, 41–42 of nature, 103–105, 123–124, See also natural
about emergence of uniform materials from science
elements, 232–233 unity of, 103–104, 110–111
about need for males in generation, 78 vs. paideia (discernment), 89, 143–145, See also
about olive oil, 243–244 paideia
about reciprocity of cause and effect, Sellars, W., 35
59–60, 134 separability
about study of animals by kinds, 149, 151, 152 as a subject for first philosophy, 128–129, 192
Pythagoreans, 4, 202, 203–205, 207, 213, 215–216, of reason (nous), 188–189
219, 221, 224–226 types of, 128, 135, 192, 197–198
sêpion (hard part of cuttlefish), 55–56
Quarantotto, D., 136, 137 Simplicius, 72, 136, 216, 218
snubness/the snub
rationalism, 19, 24–26, 29, 34 and ‘snub-like’ inquiries, 124–127, 130–131
reason (nous) and ‘snub-like’ natural objects, 106–108
as a capacity of soul, 185, 187–188, 192–193 how to define, 106–107
as a principle/source of scientific knowledge, Sorabji, R., 233
13, 21–22, 30–31 soul
in Plato, 68 affections (pathê) of, 186–189
not a subject for natural inquiry, 174, 177, as form of living thing, 78, 85–86, 186
192–199 as principle of living things, 175–177
translation of, 13 as substantial being (ousia), 182
reduction, 102, 118 categorical status of, 178, 180–182
materialist, 4, 232–235 methodos of/inquiry into, 75, 106–108, 124,
mathematical, 4 141–142, 174–175, 180–181, 186, 189–193,
respiration, 264–292 198–199
and cooling the heart/blood, 279–282, 284 modal status of, 180–181, 183–184
as subject of inquiry, 129–130, 147, 190, nature of, 177–178
269–271, 289–291 of heavenly bodies, 218, 220, 221–222
function/final cause of, 265–268, 274, 287–288 parts/capacities of, 78, 184–185, 231, 257, 260,
no separate treatise on, 265, 267, 271–272 271, 274
parts related to, 267–268, 279–280, reason as distinctive property (idion) of,
282–284, 286 187–188, 194–196
right and left, 200–226, See also directional relation to heart, 274, 276–277, 278,
dimensions 288–289
in animal locomotion, 209–215 Spinoza, Baruch, 5
in cosmology, 203–205, 207, 215–217, 218–221 subordinate sciences, 114, 117, 160
Rodin, Auguste, 133
Ross, W. D., 52, 105, 131, 136, 189, 193, teeth
267, 284 function of, 88, 137–138
lacking, 54–55, 64
Salmieri, G., 17, 101 shedding of, 225, 258
Schiefsky, M., 71 teleology. See also causes, final
scientific knowledge (epistêmê), 5 and biological generation, 82–89, 169, 256–263
and kind-crossing, 116–118 and necessity, 136–138, 165
as goal of inquiry, 1, 32, 63, 65, 75, 80, 100–102, and relation of matter to form, 128–131,
145, 171, 265–266 136–139, 198
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7
320 General Index
teleology (cont.) commensurate (coextensive), 45–46, 57,
limits to, 228, 253, 260 59–63, 278
norms related to, 87–89, 227, levels of, 147–148, 152–153, 156–159, 167,
250 278, 282
that for the sake of which. See causes, final primary, 45, 47–50, 53, 105, 278, 281–282
theoretical knowledge vs. particular, 28, 31, 33, 35–36, 139
distinguished from practical and productive,
96, 105, 108, 111, 195–198 whence the source of motion. See causes, efficient
forms of, 75, 99, 103–106, 108, 110, 123, 135, 189, wings
193–195, 198–199 division and forms of, 150–151
thermodynamics
developmental (in GA), 256–260 Yurdin, J., 27
of Mete. IV, 235–245
Zabarella, Giacomo (Jacopo), 53, 66
universal(s) zoology
and causes, 25–27, 28–29, 60–62, 132–134, 167, differences from astronomy, 161–162
242, 258–259, 289 differences from cosmology, 171–173
as principles, 30, 32, 45 norms for, 161–163, 169–173, 288–291
14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

C 1D19 12 1 8 31 2 947 73 8 4 9 7

Dtoken

Uploaded by

Document Information

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

Dtoken

Uploaded by

Copyright:

Available Formats

ARISTOTLE O N I NQUIR Y

Aristotle is a rarity in the history of philosophy and science – he is

james g. lennox is Professor Emeritus of History and Philosophy of

14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

Cambridge University Press is part of the University of Cambridge.

14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

List of Figures page ix

part i erotetic frameworks and domain-specific

part ii norms of natural inquiry 119

. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

6.1 Division of Wings, after HA I.5, 490a5-16. page 151

. 2565 7 9CC , 42 3 :586 8 4 6 D3 64C C C96 2 3 :586 6 C6 7

14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

The Goal of Knowledge and Norms of Inquiry

Chapter Summary. Book i of the Posterior Analytics articulates necessary

14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

1.2 Inquiry and Grounding for Principles:

14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

1.3 From Intuitionism to Strong Dialectic

14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

1.4 Frede’s Rationalist Aristotle

14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

1.5 Justiﬁcation as Inference to the Best Explanation

14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

1.7 Rethinking Norms, Warrant, and Inquiry:

14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

14 4 6 8 31 2 947 73 C2: 3 8 ,1 2 947 , 6

Regarding each methodos, accounts given philosophically differ from