(Lisbon Philosophical Studies - Uses of Languages in Interdisciplinary Fields) Joao Fonseca, Jorge Goncalves (Eds.) - Philosophical Perspectives On The Self-Peter Lang AG (2015)

For the last decade the topic of the Self has been under intense scrutiny from
researchers of various areas spanning from philosophy, neurosciences, and psy-

chology to anthropology and sociology. The present volume addresses the Self
under different and influent philosophical perspectives: from phenomenology
and psychoanalysis to metaphysics and neurophilosophy and discusses several
and distinct problems such as personal identity, the core/narrative self-distinc-
tion, psychopathologies, the mind-body problem and the nature of the relations
between self, consciousness and emotions. The book reflects these different
philosophical problems and approaches and aims to provide a map of current
philosophical perspectives on the topic of the Self.
João Fonseca is a Post-Doc researcher at New University of Lisbon. His

main research interest is philosophy of neuroscience. His area of competence
is philosophy of neuroscience with particular interests in topics such as: the
critical assessment to the mainstream neuro-behavior explanation of instru-
mental fear-conditioning, the quest for the evolutionary origins of self and
consciousness and the links between phenomenology and neuroscience.
Jorge de Almeida Gonçalves was born in Lisbon. He graduated in Psy-
chology (1988) and in Philosophy (1997), both in Lisbon. He has a Master’s
degree in Philosophy (2002) and a Ph.D. (2007) also in Philosophy. Be-
tween 1988 and 1999 he worked in Clinical Psychology. He currently works
at the Institute of Philosophy of Language (New University of Lisbon) and
his current research interests are consciousness and self studies, philosophy
of psychology, philosophy of psychiatry, and philosophy of film.
Philosophical Perspectives on the Self
LISBON PHILOSOPHICAL STUDIES
uses of language in interdisciplinary fields
A P ub lic ation from the Ins t it ut e o f Philo s o phy o f L a ngu ag e at t h e N e w U n i v e r s i t y o f Li s b o n
edited b y
Antón io Marq ues (Ge ne ra l E dit o r )
Nu no Ventu rin ha (Ex e cut ive E dit o r )
E ditorial Board :
Gab riele De A ng elis, Hum be r t o B r it o, J o ã o Fo ns e ca , Fra n c k Li h o r e au , A n t ó n i o M ar q u e s,
Maria Filomen a Molde r, Dio go Pir e s Aur é lio, E r ich R a st , J o ão S àág u a, Nu n o Ve n t u r i n h a
Advisory Board:
Je an- P ierre Cometti ( Unive r sit é de Pr o ve nce ), Lynn Do b s o n ( U n i v e r s i t y o f Ed i n b u r g h ) ,
E rnest L epore (Ru tge r s Unive r s it y), R e na t o L e ssa ( IUPE- R i o d e Jan e i r o ) , A n d r e w Lu g g
(Un iversity of Ottawa ), S t e f a n M a je t s cha k ( Unive r sit ä t K as s e l ) , J e s ú s Pad i l l a Gál v e z
(Un iversidad de Cas t illa - L a M a ncha ), J o a chim S chult e ( U n i v e r s i t ät Zü r i c h )
PETER LANG
Bern · Berlin · Bruxelles · Frankfurt am Main · New York · Oxford · Wien
João Fonseca & Jorge Gonçalves (eds)
Philosophical Perspectives
on the Self
PETER LANG
Bern · Berlin · Bruxelles · Frankfurt am Main · New York · Oxford · Wien
Bibliographic information published by die Deutsche Nationalbibliothek
Die Deutsche Nationalbibliothek lists this publication in the Deutsche National-
bibliografie; detailed bibliographic data is available on the Internet
at ‹http://dnb.d-nb.de›.
British Library Cataloguing-in-Publication Data: A catalogue record for this book is available
from The British Library, Great Britain
Library of Congress Cataloging-in-Publication Data
Philosophical perspectives on the self / João Fonseca & Jorge Gonçalves (eds). – 1st ed.
pages cm. – (Lisbon philosophical studies, 1663-7674 ; Vol. 5)
Includes bibliographical references.
ISBN 978-3-03-431402-2
1. Self (Philosophy) I. Fonseca, João, 1971- II. Gonçalves, Jorge, 1960-
BD450.P472345 2014
126–dc23
2014044012
Financial Support: Project Cognitive Foundations of the Self PTDC/FIL-FCI/110978/2009

(FCT – Fundação para a Ciência e Tecnologia)
ISSN 1663-7674 pb. ISSN 2235-641X eBook

ISBN 978-3-0343-1402-2 pb. ISBN 978-3-0351-0786-9 eBook
This publication has been peer reviewed.
© Peter Lang AG, International Academic Publishers, Bern 2015

Hochfeldstrasse 32, CH-3012 Bern, Switzerland
info@peterlang.com, www.peterlang.com
All rights reserved.

All parts of this publication are protected by copyright.
Any utilisation outside the strict limits of the copyright law, without the permission of
the publisher, is forbidden and liable to prosecution.
This applies in particular to reproductions, translations, microfilming, and storage and proces-
sing in electronic retrieval systems.
Contents
Acknowledgments..................................................................................7
João Fonseca, Jorge Gonçalves

Introduction ...........................................................................................9
Part I. Metaphysics and Personal Identity
Eric T. Olson
Animalism and the Remnant-Person Problem.....................................21
Rui Vieira da Cunha

Will I ever be a Cyborg?......................................................................41
Part II. Epistemology and Phenomenology
Klaus Gärtner
How Consciousness explains the Self..................................................63
António Marques
Self-Knowledge, Introspection and Memory.......................................73
Clara Morando
Imagination as a Bodily Pattern: thinking about
Sartrean’s account of Consciousness...................................................85
6 Contents
Dina Mendonça
Feelings and the Self..........................................................................101
Part III. Cognition, Psychology, Neuroscience
Erich Rast
De Se Attitudes and Semiotic Aspects of Cognition..........................121
Vasco Correia
The Division of the Mind: Paradoxes and Puzzles............................147
João Fonseca
Empirical and conceptual clarifications regarding the
notion of ‘Core-Self’ from Gallagher’s and Merker’s
Behavioural-Neuroscientific Proposals..............................................165
Part IV. Ontology and Taxonomy
Jorge Gonçalves
Core Self and the Problem of the Self...............................................207
Robert Clowes
The Reality of the Virtual Self as Interface to the Social World........221
Alexander Gerner
Conceptual Personae of the “attentional self ”...................................277
Notes on Contributors........................................................................325
Acknowledgments
We would like to thank, first and foremost the Portuguese Foundation for
Science and Technology for their support in funding the Research project
The Cognitive Foundations of The Self (PTDC/FIL-FCI/110978/2009).
We also would like to thank Peter Lang for all the help during the
process and the Instituto de Filosofia da Linguagem for all the support
and facilities.
Finally, we wish to thank the following people: António Marques,
Nuno Venturinha, Inês Hipólito, Daniel Ramalho and all the collabora-
tors in this volume, among whom Peter Olson for his patience and sheer
professionalism.
Lisbon, November 2013

João Fonseca and Jorge Gonçalves
Introduction
I. The Background
The quest for the nature and scope of the human Self has been one
of the most important intellectual tasks in western thought. Neverthe-
less, It was not until Descartes and the rise of modern philosophy, that
the cluster of problems we now associate to the notion of ‘Self ’ were
identified as such (eg.: self-identity, the nature of self-reflection, the
epistemological status of self-evidence, the unity of conscious experi-
ence, among others). What was more, this set of problems were taken
to be among the most crucial philosophical tasks to be addressed in the
upcoming centuries. The work of such diverse authors as Hume, Lock,
Kant, Nietzsche, William James, Husserl, Wittgenstein, or Sartre, to
name just a few, testifies this importance.
In the last 10 to 15 years the topic of the Self has strongly re-emerged.
This renewed interest is illustrated by the number of recent collections
of essays and anthologies (Gallagher, 1998; Kircher, 2003; Gallagher,
2010). One of the main factors holding behind such interest has to do
with the recent burst of different methodologies and approaches adopted
to face the set of problems related to the Self. These methodologies in-
clude but go beyond the more traditional philosophical approaches (like
phenomenology or linguistic analysis) (Dan Zahavi, 2005; Perry, 2002),
into empirical researches in the areas of cognitive psychology (Gallagh-
er, 2005, 2008; Hofstadter 2007) several branches of the neurosciences
(Damasio, 1999; LeDoux, 2002; Kircher, 2003), analysis of psychiatric
pathologies (such as schizophrenia) (Parnas, 2010) and other disciplines
and methodologies related to the interdisciplinary field of current cogni-
tive sciences and even social theorists and cultural analysis (Elliot, 2007).
These are, thus, exciting times in what the studies regarding the
notion the ‘Self’ are concerned: neuroscientists and cognitive psy-
chologists are accessing this notion by providing empirical methods
and scientific tools (redefining and revolutionizing the way the western
10 Introduction
tradition approaches the problem) philosophers and cognitive scientists

from various traditions are rediscovering old methodologies like phe-
nomenology and psychoanalysis and, psychiatric approaches to cases
like schizophrenia are establishing links with philosophical proposals
concerning the nature of the self and its ontological status. Succinctly,
the last decade has seen the reemergence of the interest on the notion of
‘Self’ under a new interdisciplinary umbrella.
As a drawback the adoption of such different approaches and meth-
odologies seems to result in a proliferation of unarticulated and, most of
the time, incommensurable concepts and results. Gallagher and Zahavi
(2008, pp. 197–198) identify this problem:
[T]his disparity, which is both problematic and productive, is directly related to

the variety of methodological approaches taken within philosophy and in related
interdisciplinary studies of the self. They include introspection, phenomenologi-
cal analysis, the use of thought experiments, empirical research in cognitive and
brain sciences, and studies of exceptional and pathological behaviour. One prob-
lem to be posed in this light is whether different characterizations of self signify
diverse aspects of a unitary concept of selfhood, or whether they pick out different
and unrelated concepts.
Therefore, this new transdisplinary approach comes with a cost: a taxo-

nomical confusion and fragmentation inherent to the proliferation of so
distinct methodological approaches. In itself, this consequence consti-
tutes an unfortunate obstacle to the very progress in the field of the stud-
ies about the nature of the ‘Self’. Where it should be expected conceptual
unity there is, instead, incommensurability and lack of communication.
We believe that Philosophy, given its general, far reaching, synop-
tic and conceptual approach is specially suited to overcome this unfor-
tunate scenario by providing conceptual clarification that facilitates the
establishment of links between disciplines. The present volume is an
attempt to a first approximation to different sub-topics and methodolo-
gies about the Self from a Philosophical standpoint.
It should be stressed that, and as stated in the quote above, philosoph-
ical approaches to the problems of the self are, by themselves, everything
but unified and monolithic. This diversification can be understood at two
different levels: the level of the different methodological tools used, and
the level concerning the diversity of topics and problems. Within the first
level; the methodological one, a further division is useful: the distinction
Introduction 11
between, on one hand philosophical methodologies proper (Phenomenol-

ogy, Conceptual/Linguistic analysis, third-person accounts) and, on the
other, the relation established between Philosophy and other scientific
approaches (Neuroscience, Cognitive Psychology, Artificial Intelligence,
psychiatry).
At the level of the different philosophical topics related to the Self
they include: the problem of personal identity, discussions regarding
the ontological status of the Self, the topic of self-knowledge and Im-
munity to Error trough Misidentification, modern assessments to the
Mind-Body Problem and the nature of the relations between Self and
consciousness and emotions.
This collection of essays aims to provide a non-exhaustive map
of this diversity within contemporary philosophy on the Self at both
identified levels. It reflects the pluralism of philosophical perspectives
associated with the problem (or set of problems) of the Self. Besides
providing a general view on such diversity, we aim, at the end of the day
and more implicitly, to suggest possible bridges unifying and relating
apparent protracted and unrelated data and methodologies.
II. The Essays
Part I. Metaphysics and Personal Identity
Eric Olson explores his original proposal according to which per-

sons are animals, i.e., complete organisms. In ‘Animalism and the
Remnant-Person Problem’, Olson discusses a reading of the trans-
planted brains thought experiment. Animalism clashes with the con-
viction that we should go with our transplanted brains. A good reply
is that if animalism were true, we could explain easily enough both
why the conviction is false and why it seems compelling. But another
objection cannot be answered so easily. Animalism seems to imply
that the detached brain would be a person who comes into being when
the brain is removed and ceases to exist when the brain gets into a
new head. And this seems absurd. The article argues that, although
12 Introduction
this is equally problematic for many views besides animalism, it has

no obvious solution.
In ‘Will I ever be a Cyborg’ Rui Vieira da Cunha criticises Ol-
son’s views. Vieira da Cunha states that although very attractive, in no
doubt because of its appeal to a scientific worldview, animalism is not
without its problems. In Olson’s own brand of animalism, the Organ-
ism View, one of those problems is the answer to be given to situations
of inorganic replacement, which this paper explores in a very specific
thought experiment. If indeed Olson is right in saying that animal or
human animal or organism is what best serves as a substance concept
in the case of beings like you and I, then it seems a hard task for the
animalist to account for the intuitions arising from the thought experi-
ment in this paper, at least without changing substantially the concept
of organism.
Part II. Epistemology and Phenomenology
In ‘How Consciousness explains the Self’, Klaus Gärtner asks for an

epistemological access to the Self in order to answer some ontological
questions regarding its own existence. When we talk about the meta-
physics of the ‘Self’ we want to know something about the ‘Self ’s’ na-
ture. Since in Philosophy of Mind it is less than clear if the ‘Self ’ exists
or not, it seems that we need a way to analyze it. A natural suggestion
is that Consciousness might give us the access we are looking for. The
article suggests a way of how a concept of the ‘Self ’ can be tight to a
concept of Consciousness. The key to establish such a conceptual con-
nection is Self-Consciousness. Such a relation has the advantage that
the ‘Self’ is not isolated, it is rather connected to a phenomenon that
is interdisciplinary studied. After testing this idea in a case where the
‘Self’ is compromised (schizophrenia), the article suggests that a con-
cept of the ‘Self’ depends necessarily (but not sufficiently) on a concept
of Consciousness.
In ‘Self-Knowledge, Introspection and Memory’António Marques
claims that self-knowledge by introspection (s-ki) leads to the question
of the status of the content of retrodictions, wich are memory depend-
ent statements. These are in specie different of retrodictions that are
Introduction 13
not memory dependent and therefore s-ki expresses itself in statements

that have not truth value. Furthermore the fact that s-ki has a dual time
structure (the representation at time 2 of an event or experience of a past
time 1) doesn’t mean that any s-ki retrodiction is a kind a of a meta-
representation (a representation of representations). Finally, Marques
states three other claims, namely: 1. s-ki contains a dual time structure
based on memory, 2. it must be direct (non mediated by any exterior
observer) and 3. it must have first-person authority.
Clara Morando’s article aims to identify and analyze the phe-
nomenon of imagination in Sartre’s philosophy, intending at the same
time to clarify some possible connections between imaging skills as a
specific activity of consciousness and the way physical bodies essen-
tially incorporate those kinds of data. A synthetic explanation about
how Sartre sees his theory of imagination is outlined. The article ques-
tions in what way this theory can contribute to a more accurate idea
of psychophysical relations, contending that it makes no sense to em-
ploy the expression ‘psychophysical relations’ simply because the mind
is the body and the body is the mind. Metaphysically, “imaginative
consciousness”, seems to be the very core of the “transcendental con-
sciousness”, since it can be compared to a simple intentional movement
towards ‘objects’, also characterized by an essential “nothingness” and
a great proximity to the “phenomenon of quasi-observation”, which is
not genuine observation.
In ‘Feelings and the Self’ Dina Mendonça starts by noticing that
philosophers of emotion agree that emotions always implicate a self.
However, it is not at all clear within the literature what kind of self, nor
what kind of implication, philosophers have in mind. The article argues
for a situational approach to the nature of emotions by, first, showing
how a situational approach brings to the surface the interesting con-
nections between the self and emotions, and second, by showing how
this approach allows us to understand how emotions contribute to the
constitution of the self. While the first part lays down the situational
approach to emotion and the many ways in which the implicated self
can be understood, the second part looks at some emotions (fear, love,
pride and jealousy) in order to illustrate and elaborate on the conceptual
map constructed in the first part. The article concludes indicating the
open-ended character of both emotions and self.
14 Introduction
Part III. Cognition, Psychology, Neuroscience
In ‘De Se Attitudes and Semiotic Aspects of Cognition’ Erich Rast

explores the connection between so-called de se puzzles that are well-
known in the Philosophy of Language to Jackson’s Knowledge Ar-
gument. De se puzzles attempt to show that references of an agent to
herself formulated in a 3rd-person perspective do not necessarily have
the same explanatory power for the agent’s behavior than corresponding
1st-person self-ascriptions, while the Knowledge Argument attempts to
refute physicalism by showing that the way a certain color feels, its phe-
nomenal character, cannot be explained or emulated by mere physical
knowledge no matter how exhaustive it may be. According to Rast, both
puzzles need to be addressed from the perspective of the semiotics of
cognition. Drawing from a computational metaphor he suggests a triv-
ializing interpretation of the puzzles according to which actual thought
tokens of one sort cannot be replaced by tokens of another sort because
these play different roles in cognition. From this perspective the irreduc-
ibility asserted by both kinds of puzzles is ultimately the result of the
(trivial) difference between actually cognating and explaining cognition.
In ‘The Division of the Mind: Paradoxes and Puzzles’ Vasco Cor-
reia, starts by the claim according to which divisionist models of the
mind argue that irrational phenomena such as akrasia and self-decep-
tion can only be understood if one assumes that the mind is somewhat
differentiated in relatively autonomous sub-systems. Yet, the divisionist
postulate seems to be intrinsically paradoxical in many regards. The ar-
ticle reviews some of the most influential divisionist models and argues
that each of them leads to specific inconsistencies. It is not to suggest,
however that the mind cannot possibly suffer any sort of partitioning,
and even acknowledges this possibility in pathological cases of mental
dissociation. Instead, Correia claims that the divisionist hypothesis is
not necessary to account for ordinary cases of irrationality. This anal-
ysis relies on a unitary account of the mind which maintains that irra-
tional attitudes typically derive from conflicts that take place between
individual mental states (e.g., a desire and a belief), and not between
differentiated parts of the mind.
In ‘Empirical and conceptual clarifications regarding the notion of
‘Core-Self’ from Gallagher’s and Merker’s Behavioural-Neuroscientific
Introduction 15
Proposals’ João Fonseca addresses the problem of conceptual frag-

mentation brought about by the methodological and disciplinary diver-
sity concerning current scientific studies on the Self. Fonseca focuses
specifically on two different proposals regarding the nature of Core (or
Minimal)-Self: Merker’s evolutionary perspective and Gallagher’s phe-
nomenological one. Each one corresponding to very different propos-
als for the neural implementation of Core-Self: Merker’s brainstem/
sub-cortical and Gallagher’s pre-motor cortex suggestion. At a first
approximation these two proposals seem mutually incompatible. By
deploying a model-theoretic framework for theoretical concepts of be-
havioral neuroscience, Fonseca tries to uncover some of the underlying
principles sustaining both proposals. Using the instrumental and fun-
damental notion of ‘Nested Concept’ within such framework, he shows
how the two proposals can relate to each other. Fundamentally he shows
that Merker’s and Gallagher’s suggestions of Core-Self (including their
different proposals for its neuronal implementation) are related in a com-
mon nested conceptual relation formally defined. Such nested relation
provides conceptual clarity and empirical unification into a hitherto frag-
mented and confused scenario. By bringing both proposals under the
same formal conceptual framework Fonseca shows how to gain concep-
tual and taxonomic clarification, explanatory richness and bridges, both
conceptual and empirical, between different disciplines and practices
regarding current scientific studies on the Self.
Part IV. Ontology and Taxonomy
The concepts of consciousness and self have been central in contem-

porary philosophy of the mind. Inevitably, this lead to the recuperation
of a few conceptions from classical Phenomenology, starting with Hus-
serl. It is the case of the concept of “pre-reflective self-consciousness”.
The approach of these philosophers is not existential, but what could be
called “biological” in the sense that they considered consciousness and
self as natural phenomena, explained scientifically. One of the problems
that these philosophers intended to resolve is the renowned problem of
the self that was initially formulated by David Hume and more recently
by Metzinger, among others. In his article Jorge Gonçalves, departing
16 Introduction
from scientific data regarding the developmental origins of the self, ar-
gues that the concept of pre-reflective self-consciousness does not solve
the problem. In spite of the facts not being conclusive, Gonçalves states
that there are good reasons to reject the idea that the entire form of phe-
nomenal consciousness assumes a feeling, no matter how small, of self.
In ‘The Reality of the Virtual Self as Interface to the Social World’
Robert Clowes explores the idea and implications of the virtual self.
This idea has so, up until now, been most associated with the philoso-
pher Thomas Metzinger phenomenal self model hypothesis (Metzinger,
2004, 2009). Metzinger takes the idea to imply a “no self ” thesis
(Metzinger, 2011). Clowes’ paper puts the idea against a background
of virtualist representation (Clowes & Chrisley, 2012) and from here
questions the metaphysical implications which are often drawn from the
idea of the virtual self.
Clowes’ paper starts with a review of the perplexing place of vir-
tuality more generally in theorizing about mind. His paper then focuses
in on the arguments Metzinger employs to argue for a virtual self, ask-
ing whether Metzinger’s approach can be cast into a broader virtualist
framework. Arguing that indeed it can, Clowes then attempts to show
that while Metzinger gives us an interesting way to think about the self
that the conclusion he draws from them about the non existence of self
are unwarranted.
In this context, the final part of the paper attempts to show that the
virtual self may in fact be a useful way of making the self theoretically
tractable for further scientific investigation including in the context of
psychopathology. The paper concludes that the concept of the virtual
self may not only be a useful theoretical tool but it may be real enough
to supply the conceptual roles required in much theorizing around the
self. Concluding that the virtual self does not need imply a “no self ”,
Clowes concludes it may in fact be a useful way of unifying several cur-
rent ideas about self. The virtualist view of self may be best thought of
as a fruitful scientific reduction rather than an elimination as Metzinger
argues.
In his paper “Conceptual Personae of the ‘attentional self’” Alex-
ander Gerner explores conceptual personae of the self in relation to
the phenomenon of attention. Philosophical concepts, as the concept
of the self, can be personae that we think with and that emerge from
Introduction 17
a constitutive plane of philosophy in which they create concepts on

this plane that are more complex than a specific model -of the self-
or a specific mode of symbolic representation. While Gerner takes
up Thomas Metzinger’s claim that the strong first person perspective or
“subjective self ”, defined as the centre of awareness, is the possibility
of being able to manipulate the focus of attention in order to stabilize
subjective experience, he proposes attention as a constitutive ground
of the self that exceeds the self-model theory. For Gerner the self as
attentional self is less about stabilization of subjective experience but
rather responsible for the constitutive imbalance of the self. Hence Ger-
ner puts Metzinger’s thesis of the “control of the focus of attention” and
the resulting complex notion of an attentional self into a new perspec-
tive through an approximation of two possible conceptual personae of
the self: a) the impossible attentional self in Paul Valéry’s ‘Monsieur
Teste’ and b) the ‘heautoscopic’ attentional self in Italo Calvino’s ‘
Mister Palomar’.
João Fonseca
Jorge Gonçalves
References
Clowes, R. W., & Chrisley, R. (2012). Virtualist Representation. Inter-

national Journal of Machine Consciousness, 04(02), 503–522.
Damásio, A. (1999). The Feeling of What Happens: Body and Emotion
in the Making of Consciousness. San Diego: Harcourt Brace.
Elliott, A. (2007). Concepts of the Self. Malden: Polity Press.
Gallagher, S. (ed.) (2010). Oxford Handbook of the Self. Oxford: Ox-
ford University Press.
Gallagher, S. & Shear, J. (eds.) (1999). Models of the Self. Exeter: Im-
print Academic.
Gallagher, S. & Zahavi, D. (2008). The Phenomenological Mind: An
Introduction to Philosophy of Mind and Cognitive Science. New
York: Routledge.
18 Introduction
Hofstadter, D. (2007). I Am a Strange Loop. New York: Basic Books.

Kircher, T. (ed.) (2003). The Self in Neuroscience and Psychiatry. Cam-
bridge: Cambridge University Press.
LeDoux, J. (2002). Synaptic Self: How Our Brains Become Who We Are.
New York: Viking Adult.
Metzinger, T. (2004). Being No One: The Self-Model Theory of Subjec-
tivity. Cambridge, MA: Bradford Book.
Metzinger, T. (2009). The Ego Tunnel: The Science of the Mind and the
Myth of the Self. New York: Basic Books.
Metzinger, T. (2010). The no-self alternative. In Gallagher, S. (ed.),
Oxford Handbook of the Self. Oxford: Oxford University Press.
279–296.
Parnas, J. & Sass, L. (2010). The Structure of Self-Consciousness in
Schizophrenia. In Gallagher, S. (ed.), Oxford Handbook of the Self,
Oxford: Oxford University Press. 521–546.
Perry, J. (2002). Identity, Personal Identity, and the Self. Cambridge:
Hackett Publishing Company.
Zahavi, D. (2005). Subjectivity and Selfhood: Investigating the First-Per-
son Perspective. Cambridge: MIT Press.
Part I
Metaphysics and
Personal Identity
Animalism and the Remnant-Person Problem
Eric T. Olson
Introduction
Animalism clashes with the conviction that we should go with our

transplanted brains. A good reply is that if animalism were true, we
could explain easily enough both why the conviction is false and why it
seems compelling. But another objection cannot be answered so easily.
Animalism seems to imply that the detached brain would be a person
who comes into being when the brain is removed and ceases to exist
when the brain goes into a new head. And that seems absurd. The paper
argues that, although this is equally problem for many views besides
animalism, it has no obvious solution.
1.
Animalism is the view that you and I are animals. That is, we are an-
imals in the straightforward sense of having the property of being an
animal, or in that each of us is identical to an animal – not merely in the
derivative sense of having animal bodies, or of being “constituted by”
animals. And by ‘animal’ I mean an organism of the animal kingdom.1
Sensible though it may appear, animalism is highly contentious. The
most common objection is that it conflicts with widespread and deep
1 Many philosophers say that we “are” animals, but mean only that we are nonani-
mals constituted by animals. And some deny that human animals are organisms.
For an example of both views, see Johnston (2007, pp. 49, 56).
22 Eric T. Olson
beliefs about our identity over time. These beliefs are brought out in re-
actions to fictional cases. Suppose, for instance, that your brain is trans-
planted into my head. The being who ends up with that organ, everyone
assumes, will remember your life and not mine. More generally, he will
have your beliefs, preferences, plans, and other mental properties, for the
most part at least. Who would he be – you, me, or someone else?
Animalism implies that he would be me. That’s because the oper-
ation does not move a biological organism from one head to another.
It simply moves an organ from one animal to another, just as a liver
transplant does. One organism loses its brain and remains behind as an
empty-headed vegetable; another has its brain removed and replaced
with yours. (Or perhaps, as van Inwagen (1990, pp. 172–181) propos-
es, the naked brain would itself be an organism, and the empty-headed
thing left over would be a mere hunk of living tissue, like a severed arm,
owing to the brainstem’s role in directing a human organism’s life-sus-
taining functions. In that case, the operation would pare down an animal
to the size of a brain and move it to another head, and the being who
ended up with your brain and the rest of me would be you, even accord-
ing to animalism. If this is right, the objectors must replace ‘brain’ with
‘cerebrum’. No one thinks an organism could be pared down to a naked
cerebrum. I will ignore this complication in the sequel.)
So if you and I are animals, I could swap my brain for yours. In that
case I should suddenly acquire your knowledge, skills, and interests,
and lose my own. I should lose all the memories of my past. In their
place I should acquire memories of your past: of holidays I never took,
people I never met, experiences I never had. My head would be filled
with false beliefs: I should be convinced that I lived in your house,
worked at your job, and was married to your spouse. I should think I
was you. I should be systematically mistaken about who I am and how I
fit into the world. As for you: if the operation didn’t kill you outright, it
would cause you to lose all your memories, knowledge, plans, abilities –
everything that matters. Unless, that is, you too got a new brain. If you
got my brain, your lost memories would be replaced by memories of my
life. You would be convinced, mistakenly, that you lived in my house,
worked at my job, and were married to my spouse.
That is what animalism implies about the transplant story. And it
is easy to be unhappy with this description. In my experience, those
Animalism and the Remnant-Person Problem 23
presented with the thought experiment have an almost irresistible

tendency to say that the one who got your brain would be you, not me.
The operation would not give me your brain, but would give you my
body. Transplanting your brain amounts to transplanting you. A brain
transplant is not at all like a liver transplant. Call this conviction the
transplant intuition. (What about me? Well, the intuition implies when
the surgeons remove my brain to make way for yours, they remove me
from my own head, just as they remove you from yours.)
The objection, then, is this: Even if our brains are never actually
transplanted, each of us has the capacity to go with one’s transplanted
brain: to be pared down to a naked brain and moved to another head. But
no animal has that capacity. Transplant an animal’s brain and the animal
stays behind. It follows (by Leibniz’s Law) that we are not animals. To
put the point another way: if your brain were transplanted, then accord-
ing to the transplant intuition you would go one way, while the animal
we call your body – the animal you would be if you were any animal
at all – would go another. But it is impossible for a thing and itself to
go their separate ways. If it is possible for you and your animal body to
go your separate ways, it follows that you are one thing and the animal
is another thing, numerically different from you. And there is no other
animal or organism that you could be. It follows that you are not an
animal, or any other a biological organism. You may still relate in some
intimate way to an animal: you might have an animal as your body, or be
constituted by an animal. But you are not yourself an animal.
2.
Is this a strong objection? Well, here is a way of defending animalism

against it. Suppose for the sake of argument that we were animals (hard-
ly a wild conjecture). That would not only entail that the transplant intu-
tion was false, but it would explain in an unmysterious way why it was
false. If we were animals, we should not go with our transplanted brains
because no animal would go with its transplanted brain. An animal,
even a human organism, simply loses an organ and gets an empty head
24 Eric T. Olson
when its brain is removed. (At any rate that is a premise of the transplant
objection.) And if we were animals, that is what would happen to us.
Animalists can also explain why we should nonetheless find the
transplant intuition attractive, even compelling, if it were false for
this reason. Why might the transplant intuition seem true if we were
animals? Why does it seem so obvious that the person who ended up
with your brain would be you? How could generations of philosophers
and their students have got it so badly wrong? Well, there would be
strong grounds for accepting the transplant intuition even if it were false
because we are animals. In fact, human animals would have the same
reasons for supposing that they would go with their transplanted cere-
brums as we have to believe that we should.
There are two main grounds supporting the transplant intuition.
First, the psychological and behavioural evidence that supports judg-
ments about personal identity in familiar cases – judgments that are
compatible with animalism – also supports the transplant intuition.
Why do I suppose that the man who will wake up in my bed tomor-
row (in normal circumstances) is me, and not a new person who came
into being during the night? One reason is that he will have my memo-
ries, beliefs, preferences, and plans, or at least memories, beliefs, and
plans that are causally dependent in a special way on my current mental
states. In other words, he will be psychologically continuous, then, with
me as I am now. What’s more, this psychological continuity will be (as
Unger puts it) continuously physically realized in my brain, which will
remain intact overnight. And no one in any real situation is ever psy-
chologically continuous with someone without being that person: no
one is ever psychologically continuous with someone else. In real life,
someone’s being psychologically continuous with you at some time in
the past or future is powerful evidence – probably conclusive evidence
– for his being you.
The person who ended up with your brain in the transplant story
would likewise be psychologically continuous with you, and this conti-
nuity would be continuously physically realized. And there would be no
psychological continuity between that person and me. In real life, this
would be conclusive evidence for his being you. So it would hardly be
surprising if this led us to believe strongly that the brain recipient would
be you rather than me. And we should be no less inclined to think so
if in fact he would not be you because (unbeknownst to us) you are an

animal.
The second sort of evidence supporting the transplant intuition
is practical. The person who ends up with your brain may have “what
matters in identity” for you. Before the operation you might have the
same reason to care about the welfare of that person as you normally
have to care about your own welfare. He might be morally responsible
for the things you did before the operation. Everyone might be morally
obliged to treat him just as if he were you; and he might be obliged to
treat your children, spouse, and friends as if they were his own. And
maybe the empty-headed being left behind would not have what matters
in identity for you. Even if that being were subsequently given a new
brain from some third party, you might have no more reason, before the
operation, to care about his or her welfare then than you have to care
about the welfare of any other stranger. In all actual circumstances, a
person bears these relations of practical importance only to herself: it’s
never the case that someone else has “what matters in identity” for you.
So again, it would hardly be surprising if the fact that the person who
got your transplanted brain would have what matters in identity for you
inclined us to suppose that he would be you. And this inclination would
be in no way diminished if the supposition were false because we are
animals.2
These are all grounds for supposing that you and I and every other
human person would go with his transplanted cerebrum. And they are
grounds that we should have even if we were animals and it were false.
In that case our mistake would be entirely understandable. Nor would
there be the slightest mystery about why the transplant intuition was
false. So it looks as if we have little right to any great confidence in the
transplant intuition unless we already have a good reason to suppose
that we are not animals.
2 “Fission” is another imaginary case where what is actually conclusive evidence

for someone’s being you might be consistent with her not being you. If each half
of your brain were transplanted into a different head, resulting in two beings psy-
chologically continuous with you, it is easy to suppose that both would be you. Yet
even many opponents of animalism concede that this cannot be, as there is only
one of you, and one thing cannot be numerically identical to two things.
26 Eric T. Olson
3.
Even if this is a good defence of animalism, however, the brain-transplant

story raises another objection that cannot be answered so easily. Think
about your brain in mid-transplant, detached from your skull but kept
alive. It is of course not alive in the way that an organism is alive. But it
can be alive in the sense that its cells remain alive, like a liver awaiting
transplant.
It seems possible for the brain, in this condition, to support thought
and consciousness. At any rate this is almost universally assumed, and I
am not going to challenge it here. “Brain-in-a-vat” thought experiments
are a staple in the diet of philosophy students everywhere, and it is part
of the story that a detached brain could produce thought. (It may be that
mental life is possible only if the cerebrum is stimulated in some special
way by the brainstem, so that a naked cerebrum merely kept alive in a vat
would have little or no mental life. In that case, if we are transplanting
only the cerebrum, we must imagine it provided artificially with whatev-
er stimulation it needs to support mental life.) But if the brain supports
thought and consciousness, then there is a thinking, conscious being there
– a being psychologically more or less like you. In other words, your
brain, kept alive in a vat or whatever, would be a person. Or if it would
not be a person itself, it would at least “realize” or “constitute” a person.
Mark Johnston, to whom I owe this objection, calls it a remnant person.3
The possibility of remnant people is trouble for animalism. Ani-
malism seems to imply that this remnant person would not be you. Oth-
erwise it would be possible to pare down a human organism until it was
nothing but a brain. But if an organism could be reduced to a detached
brain (or cerebrum), then it could be transplanted from one head to an-
other, and we have already ruled that out. (I will revisit the ruling in the
next section.) The problem is not that we are convinced that the rem-
nant person would have to be you, contrary to animalism. That would
be the transplant intuition again, or at any rate an intuition epistemically
3 We might say that something is a remnant person at a time t just if she is a wholly
organic person at t, she is not herself an organism or a thing constituted by an
organism at t, and this condition is a result of cutting away a large portion of a
normal human person at some time before t.
indistinguishable from it, and animalism could be defended against it

in the same way. Nor is the problem that the remnant person would not
be an organism. That is perfectly compatible with animalism, which
does not say that necessarily all people are organisms, but only that we
are – we normal human people. (For all animalism says, there might
be entirely inorganic beings who count as people in the sense of being
rational, self-conscious, and so on: angels, for instance.)
The trouble comes when we ask where the remnant person could
have come from, if he or she could not be you. Surely he did not exist
before the operation. Otherwise there would have been two people with-
in your skin – the organism, who became an empty-headed vegetable,
and the remnant person, who became a naked brain; and that is absurd.
It looks as if animalists must say that the operation brings the remnant
person into being. But that looks absurd.
For one thing, it is evident that there are just two people in the
transplant story, you and I, even if there is dispute about what happens
to us. If the remnant person were someone new, there would be three.
Or rather four: if removing your brain from your head creates one new
remnant person, then removing my brain to make room for yours cre-
ates another. There are you, I, and the two remnant people created when
our brains are removed. That’s two too many.
More seriously, it is impossible to believe that removing someone’s
brain from her head could create a new person. As Johnston puts it:
You can’t bring a person into being simply by removing tissue from something…,
unless that tissue was functioning to suppress mental life or the capacity for men-
tal life. A developing fetus might have a massive tumor in its developing brain,
which suppresses its mental life, and perhaps even its capacity for mental life.
Given that, we can understand how removing the tumor could allow a person in
Locke’s sense to be present for the first time. But how could removing a sustaining
[head and] torso bring this about? (2007, p. 47).
If animalism implies that the transplant operation would bring a rem-

nant person into being, it violates what we might call the creation prin-
ciple: that you cannot bring a person into being merely by cutting away
sustaining tissues.
Animalism would violate a second and equally attractive principle
as well, namely that you cannot destroy a person merely by surrounding
28 Eric T. Olson
him with sustaining tissues: the destruction principle. Suppose we im-

plant our remnant person into a new head – or, for that matter, into your
own head again – hooking up the utilities in such a way as to make the
resulting person more or less normal. According to animalism, this
normally embodied person would be an organism. And no organism was
ever a detached brain. The result of implanting a brain into an empty head
is not that the brain comes to be an organism; rather, the organism simply
acquires a new organ, just as it might acquire a new liver or kidney. Yet the
remnant person does not cease to be a person when he or she is implanted.
Nor does it come to be the case that there are two people within the same
skin. It follows that the remnant person must cease to exist when he is put
into a head. But providing a maimed person with the parts he was missing
is a funny way of destroying him!
So the new objection to animalism is that it is incompatible with the
creation and destruction principles. You cannot create a person just by
cutting away sustaining tissues, or destroy one just by providing them.
Because animalism implies that you can create and destroy a person in
these ways, it must be false.
This is not just the original transplant objection put differently, and
it demands a different response on the part of the animalist. Our being
animals cannot explain why the creation and destruction principles are
false. It may explain why none of us can come to be a remnant person,
and thus why the remnant person who results from removing your brain
from your head would not be you: the explanation would be that you are
an animal, and no animal can become a remnant person. But our being
animals cannot explain how the operation could have brought a person
into being. The remnant person’s not being you is one thing; his coming
into existence when your brain is removed is another. Likewise, our
being animals may explain why the remnant person is not the normally
embodied person who would result from putting your brain into my
head: the explanation is that no animal could once have been a remnant
person. But it cannot explain how implanting the remnant person into
a new head could destroy him. His ceasing to exist is different from his
not being identical to any normally embodied person.
To explain how removing your brain from your head could bring
someone into existence, and how replacing that organ could destroy him,
we need an account of the metaphysical nature of remnant people. We need
an account that would tell us, among other things, why remnant people
come into being and pass away in such a surprising way – why they have
such funny persistence conditions. But animalism offers only an account
of our metaphysical nature: that of normally embodied human people. So
even if animalism entails that the creation and destruction principles must
be false, it cannot explain why they are false. The strange behavior of rem-
nant people would remain a mystery. Call this the remnant-person problem.
4.
I can offer no account of why removing someone’s brain from her head
would bring a remnant person into being, or why that person would
cease to exist if the brain were put back. If animalism really did imply
that this would be so, it would be a serious objection. I think animalists
should accept the creation and destruction principles, and deny that a
brain transplant would create and then destroy a remnant person. How
could that be? Who would the remnant person be, and how would he
relate to the animal people, you and me?
One suggestion is that despite appearances, the remnant person is
you, the donor organism. Removing your brain or cerebrum from your
head would not remove an organ from a human animal, leaving that
animal with an empty head. And putting your brain or cerebrum into
my head would not supply this animal – my body – with a new organ.
Rather, the operation would pare an organism down to a naked brain
and later supply it with new peripheral parts to replace the ones cut
away. Of course, the brain or cerebrum in mid-transplant is not an or-
ganism, at least not then. But perhaps an organism is not an organism
essentially, and can exist for a while as a nonorganism, just as a student
can take a leave of absence and exist for a while as a nonstudent. This is
logically consistent with animalism. Animalism is the view that we are
organisms, not the view that we are organisms essentially. So maybe a
human animal really would go with its transplanted brain. In that case,
no remnant person would be created or destroyed in the operation. This
would answer not only the remnant-person objection, but the transplant
30 Eric T. Olson
objection as well. Both objections would be based on a false assumption

about what it takes for a human animal to persist. Because this proposal
implies that we are animals accidentally and not essentially, we might
call it accidental animalism.
Accidental animalism raises a number of worries. Perhaps the most
obvious is that the empty-headed thing left behind after your brain is
removed may still be alive – especially if the operation transplants only
the cerebrum. That is, it might be a living organism. And if it were, it
would seem to have the same life, as Locke would say, as the original
organism had: the original organism’s life-sustaining functions would
have continued uninterrupted throughout the operation in the large
thing left behind, just as they would had the surgeons removed the liver
rather than the cerebrum. And if an organism’s biological life carries on,
how could it not continue to be the life of that same organism? Even if
an organism could continue existing after its life comes to an end, how
could an organism be outlived by its own life?4 Yet accidental animal-
ism implies that this living, empty-headed organism would not be the
original animal. That looks obviously false. Surely it is possible for a
human animal to have an organ not essential to the maintenance of its
life-sustaining functions removed, and to continue existing without that
organ. If a human animal could not possibly exist even for a moment
without its cerebrum, then we ought seriously to wonder whether any
animal could exist even for a moment without its liver or kidney or ap-
pendix. It would be a real epistemic possibility that an appendectomy
reduces an animal to the size of an appendix, while the living organism
left behind is not the survivor of the operation, but something new.
For that matter, accidental animalism creates a new version of the
problem it was meant to avoid: a “remnant-animal problem”. It implies
that removing your brain would create a new animal – the one with an
empty skull. And replacing your brain with a new one, or even rehousing
it in its original head, would cause this remnant animal to cease to exist,
as the resulting “whole” animal would be the former brain rather than the
former brainless animal. But surely you cannot bring an organism into
being simply by removing tissue from something, unless (as Johnston
4 For a defence of Locke’s view that the continuation of an organism’s life is both
necessary and sufficient for the organism to persist, see van Inwagen (1990,
pp. 142–158).
would say) that tissue was functioning to suppress life or the capacity
for life. Perhaps there could be tissues that were capable of developing
into or coming to make up a living organism but for the presence of a
tumour or other foreign body. In that case removing the tumour might
bring an organism into being. But removing the cerebrum of a normal,
healthy human being cannot do this. Nor can you destroy an animal living
without a cerebrum or other nonvital organ simply by providing it with
such an organ. These principles, surely, have the same force as the origi-
nal creation and destruction principles. Accidental animalism avoids the
objection that remnant people would have absurd persistence conditions
by proposing that organisms have them. Nothing is gained.
5.
Here is a better proposal: The remnant person who would result from
removing your brain from your head would be simply your brain – that
is, the thing that is now your brain. The remnant person is never an
organism, and the operation doesn’t make him any bigger or smaller.
Removing your brain from your head does not bring a person or any
other material thing into being, and putting it into a new head does not
destroy anyone or anything.
Plausible thought this may be, it raises an obvious pointed ques-
tion. We are supposing that the remnant person can think and be con-
scious while he is detached from the rest of you. If the remnant person
is your brain, this means that your brain could think and be conscious
while detached. The pointed question is this: does your brain think now,
in its normal surroundings? Is it now conscious? I don’t mean whether
it “thinks” in some attenuated or derivative sense – in the sense of being
the organ responsible for your thinking, say. The question is whether it
thinks in the strictest and most straightforward sense. There is no very
attractive answer to this question.
Suppose your brain does now think in the strictest sense. But we
are supposing that you are not your brain. You are not, as things are, a
three-pound, yellowish-pink organ located entirely within your skull.
32 Eric T. Olson
Rather, you are the animal your brain is a part of. And you also think
in the strictest sense. Surely it couldn’t be the case that you think only
in a derivative sense: that you think only insofar as you have a part that
thinks strictly speaking. As Chisholm said, if there are now two things
thinking your thoughts, one doing it on its own and the other such that
its thinking is done for it by something else, you are the one that thinks
on its own (1976, p. 104).
Or maybe Chisholm was wrong about this, and there is no prob-
lem in saying that we think only in the derivative sense of having a
part that thinks strictly speaking. In that case, we really can solve the
remnant-person problem by saying that the remnant person would be
your brain. Your brain thinks now, and is the only real thinker there. If it
were removed from your head and kept alive in a vat, it would continue
to think – though it would then think only for itself, and not for you.
Presumably your brain would count as a person while it was detached,
but not now while it remains a part of you, even though there would
be no change in its mental capacities. Many philosophers will have no
objection to this. Orthodox four-dimensionalism – the view that all per-
sisting things including ourselves are composed of temporal parts – has
a similar consequence: it implies that you think now only insofar as a
part of you – your current stage – thinks strictly speaking. And although
that stage is not in fact a person, it would be were it not surrounded
by other stages psychologically continuous with it, even though there
would be no difference in its mental capacities. Yet almost no one takes
this to be an objection to four-dimensionalism.5
But let us suppose that Chisholm was right: if anything thinks your
thoughts in the strictest sense, you do. If your brain now thinks, then
so do you. It would follow that every normal human person is accom-
panied by another being psychologically indistinguishable from her.
This article would have two authors, I and my brain, and there would
be at least two conscious beings now sitting there reading it. On the
Lockean assumption that a person is by definition a being with certain
mental properties – rationality and self-consciousness, say – your brain
would be a person. You would be one of two people now thinking your
thoughts. The transplant operation would separate them, turning one
5 Olson (2007, pp. 122–125) is an exception.

into a remnant person and the other into a brainless vegetable. You
ought to wonder which person you are. Any grounds you might have for
supposing that you are the animal person rather than the brain person
would seem to be grounds for the brain to suppose that it is the animal
person rather than the brain person. How could you know that you’re
not making this mistake? Likewise, you ought to wonder whether you
are the one who would go with your brain if it were transplanted, or the
one who would stay behind with an empty head.
Well, suppose your brain does not now think in the strictest sense.
But it would think if it were detached and suitably cosseted. This seems
to imply that your brain is now prevented from thinking by its fleshy
surroundings. Normally it is a mere brute organ, no more sentient or in-
telligent than a kidney; but remove it from its natural habitat in the right
way and it will blossom instantly into a mature philosopher. And putting
it back where it belongs would deprive it of these new-found intellectual
capacities and restore it to its former state of total oblivion (something
that is normally a serious crime). So the sustaining tissues surrounding
the brain really do “suppress mental life or the capacity for mental life”.
They may not suppress mental life altogether: they don’t suppress it
in the organism. But they suppress it in the brain, by preventing the
brain from having its own mental life. Yet surely, we want to say, you
can’t give something the capacity for thought and consciousness merely
by cutting away sustaining tissues; nor can you deprive something of
that capacity just by surrounding it with such tissues. That looks just as
compelling as the original creation and destruction principles.
So if your brain thinks now, there are too many thinkers; if it doesn’t,
things can gain or lose mental capacities in an utterly baffling way. That’s
the trouble that comes of saying that the remnant person would be your
brain. We might call this trouble the remnant-brain problem.
6.
However grave the remnant-brain problem may be, it is no reason to

doubt whether we are animals. It is not a problem for animalism in
34 Eric T. Olson
particular, but applies equally to almost any view about what we are.
I say almost any view because you could avoid the problem by saying
that we are brains: that each of us is literally a three-pound lump of
tissue housed within the skull. In that case your brain thinks even now,
and removing it from your head would do nothing to enhance its mental
capacities. The operation would merely change your surroundings.
What about the animal – your body? Doesn’t it think, giving us too
many thinkers? Well, if you are your brain, that is presumably because
your brain is the only thinking being there. Anything bigger than a brain
has mental properties only in the derivative sense of having a brain that
has them strictly speaking. This would solve the remnant-brain prob-
lem. But no philosopher that I know of thinks that we really are brains.6
If you are not your brain, then our pointed question returns: does
your brain now think – really think, in the strictest possible sense? If it
does, then there are now two beings thinking your thoughts, your and
your brain. On the assumption that a person is a being with such men-
tal properties as rationality and self-consciousness, you are one of two
people now thinking your thoughts. You ought to wonder which one you
are, and how you could ever know. Or maybe your brain doesn’t think.
But it would if it were removed from your head and suitably cosseted.
This means that your brain is now prevented from thinking by its fleshy
surroundings. And putting it back where it belongs would presumably
deprive it of its power to think and restore it to its former state of total
oblivion. The remnant-brain problem does not arise only if we are ani-
mals. It arises if we are anything bigger than brains.
Someone might propose that your brain could never think, even
when removed from your head. At most it might constitute a thinker: its
matter would make up a thinking being other than your brain itself. Your
body – an organism – constitutes you now, but if your brain were removed
from your head, that organ would then constitute you, or at any rate it
would do so for as long as it continued to realize your psychology. (This is
Johnston’s view, and presumably that of most other “constitutionalists”.)
6 Though Hudson (2001, p. 143) says that each of us is a temporal part of a brain.
I discuss the view that we are brains in Olson (2007, pp. 76–98). I suppose one
might also avoid the remnant-brain problem by saying that we are Humean bun-
dles of impressions, or immaterial substances.
But this only relocates the problem. Constitutionalists do not think

that your brain now constitutes you, or indeed any other thinking being.
You do not weigh three pounds. You – not merely your body, but you
yourself – extend all the way out to your skin. It is your animal body that
now constitutes you, not your brain. So although removing your brain
from your head would not give it consciousness or the power to think, it
would give it the power to constitute a conscious, thinking being. Your
brain is now prevented from constituting a thinker by its fleshy sur-
roundings, and putting it back where it belongs after its removal would
prevent it from doing so once more. Those sustaining tissues “suppress
mental life or the capacity for mental life” insofar as they prevent the
brain from constituting a psychological being. And that seems absurd.
This is just the remnant-brain problem in a different form. As far as I
can see, any explanation of why a brain could constitute a thinking being
when isolated but not in its natural habitat would serve equally well as an
explanation of why a brain could think when isolated but not when em-
bodied. It may be, for instance, that a brain cannot in its normal surround-
ings constitute a thinker because it is a then part of a larger being that
constitutes a thinker, namely the animal. In that case, animalists can say
that, although the brain could think when isolated, it cannot think in its
normal surroundings because it is then a part of a larger being that thinks.
7.
So the remnant-brain problem is not a worry for animalism in particular,

and simply denying that we are animals does nothing to solve it. If consti-
tutionalists could explain why your brain would constitute a thinker when
detached from the rest of you but not in its normal surroundings, then an-
imalists could explain in the same way why your brain would think when
detached but not otherwise. If animalists cannot explain why removing
your brain would enable it to think, then constitutionalists cannot explain
why removing your brain would enable it to constitute a person. If the rem-
nant-brain problem is a reason to deny that we are animals, it is equally a
reason to deny that we are material things constituted by animals.
36 Eric T. Olson
For all that, I have done nothing whatever to solve the remnant-brain
problem, or the more general remnant-person problem. Can it be solved –
other than by saying that we are brains, or some other desperate ploy?
I haven’t much hope for a “psychological” solution – that is, an
account of why human brains should have the power to think (or to
constitute a thinker) when detached but not otherwise. Another possi-
bility is to deny that there are such things as undetached brains. There
are of course particles “arranged cerebrally” within your skull; but
they don’t compose anything. Nor would they compose anything when
outside your head. There are no undetached brains, and no detached
ones either. There are no remnant people. There would be no remnant
people even if brain transplants really occurred. So there is nothing in
the transplant story whose surprising inability to think needs explaining
and the question of how the transplant operation could create and then
destroy a person does not arise. We might call this brain eliminativism.
If you think it sounds crazy, I don’t blame you. But it may be no worse
than the alternatives, including rejecting animalism. In any case, I will
devote the rest of this paper to exploring it.
The obvious question it raises is why particles arranged cerebrally
never compose anything. Why should there be human animals but no
human brains? The only way to answer this question is to work out
when any particles compose something. How, in general, do smaller
things have to be arranged and situated for them to compose or add up
to something bigger? There are two “extreme” answers to this ques-
tion. One is compositional universalism: any things, no matter what
their nature or arrangement, compose something. The other is compo-
sitional nihilism: no things ever compose anything. There are no com-
posite objects, but only mereological simples. Nihilism is obviously
incompatible with animalism, since no organism is a simple. And for
reasons I have given elsewhere (2007, pp. 229–232), animalism is not
easily combined with universalism. Animalists need to say that some
things compose something and others don’t – which is what most of us
probably thought anyway. But which ones, and why? Very few answers
to this question have been proposed. The best answer I know of is van
Inwagen’s (1990, pp. 81–97): those things compose something if and
only if their activities constitute a biological life. This implies that the
only composite objects are living organisms. I cannot defend this view
here. But it would explain why there are no remnant people: because a
remnant person is by definition not an organism.
8.
Whatever its merits, brain eliminativism raises deep issues. Even if the
particles arranged cerebrally in the vat would not compose anything,
they might still manage to produce thought. At any rate, what went on
inside the remnant person (to speak loosely as if there were such a thing)
would be indistinguishable from what goes on within a normal human
brain. So I have been assuming, anyway. Since you are a philosopher,
the remnant person resulting from removing your brain would also be
a philosopher. Or rather, those particles arranged cerebrally would col-
lectively produce philosophical thought. This would be thought without
a thinker. That possibility might be epistemically troubling.
Suppose the particles in the vat were collectively to produce the
following line of argument:
I could be wrong in thinking that I have hands, or that there are other people, or
that anything existed five minutes ago. But even so, I can be sure that I am now
thinking. Nothing could mistakenly think that it was thinking. And if I am think-
ing, then surely I must exist. I could never be mistaken in thinking that I exist. So
my own existence, at least, is certain.
According to brain eliminativism, the conclusion of this reasoning

would be false. Even if it is logically valid, and anyone who thinks that
she exists (or who thinks anything at all) really does exist then, it would
not be true in this case that anyone was thinking. The word ‘I’ would
not refer to anything. Or at least it would not refer to anything that
was thinking. At most it might refer in the plural to the particles jointly
producing the thought. But no particle thinks. Of course, in this case no
one would be mistaken in giving the argument, since no one would be
giving it. But the argument would be unsound, because the premise ‘I
am thinking’ would be false.
38 Eric T. Olson
If this is possible, then I ought to wonder whether I exist. I feel pretty

certain that I do. Why? Well, if I didn’t exist, how could I be sitting here
writing this? How could I even raise the question of whether I exist? But
the particles in the vat could produce the very same reasoning, yet the
conclusion would be false. What reason have I got, then, to suppose that
my reasoning is not unsound for the same reason – because my thoughts
are produced by particles that don’t compose anything? However obvious
it may appear to me that I am now thinking, particles arranged cerebrally
in a vat could (collectively) find the thought ‘I am thinking’ equally obvi-
ous, though it would be false. So for all I know it’s false in my case too.
The worry does not require us to take seriously the possibility that
I might be a remnant person – that evil surgeons might have snatched
Olson from his bed last night and put his particles arranged cerebrally in
a vat, and that those particles are the ones producing these thoughts. That
scenario is farfetched, to say the least. Maybe we can legitimately ignore
it. Even so, brain eliminativism seems to imply that it is possible, in the
brain-transplant case, for particles collectively to produce thought just
like mine without thereby composing any thinking being. That raises the
worry that it might be possible in other cases too, even in ordinary ones.
It deprives brain eliminativists of what seems the best argument for sup-
posing that there are any thinking beings, namely that there is thought
and that thought requires a thinker. Presumably it would undermine any
other argument for the existence of thinking beings in the same way. If
this is the way to solve the remnant-person problem, it comes at the cost
of undermining our belief in our own existence: hardly a welcome result.
9.
Or maybe there could not be thought without a thinker, and particles ar-
ranged cerebrally in a vat would be unable to produce thought. (This is van
Inwagen’s view [1990, p. 118f.].) Even if they could produce something
intrinsinsically indistinguishable from thought, it might not be genuine
mental activity: no beliefs, desires, reasoning, or conscious states. Maybe
there is nothing that it would be like to be a remnant person, and nothing
it would be like to be in the states that particles arranged cerebrally in a

vat could produce. That might enable me to know that I exist, and that my
thoughts are not produced collectively by particles that compose nothing.
For I can know that thinking is going on. There seems to be thinking go-
ing on, and this seeming is itself a sort of thought. If my particles did not
compose anything, as the particles arranged cerebrally in the vat would
not, then they could not collectively produce any seeming, and so it would
not seem to me that I was thinking. That it seems to me that I am thinking
(or that there seems to be thinking going on) would entail that it seems
that way to something, and hence that something thinks: me. That might
enable me to know that I exist.
This would mean that remnant people, or more precisely particles
arranged cerebrally that are not parts of an organism, would be able to
produce only pseudo-thought and pseudo-consciousness, though these
states might be intrinsically identical to real thought and consciousness,
and would have similar causes and effects to those of real thought and
consciousness. Remnant people would be able to have conversations and
write philosophical essays – or at least they could interact in ways out-
wardly indistinguishable from conversations, and collectively produce
philosophical essays. There could be many remnant people hooked up
by wireless links to robotic “bodies” whose movements were indistin-
guishable from those of human beings. There could be a vast society (as
it were) of such remotely controlled robots, behaving and interacting in a
way indistinguishable from real people, or at least as much like it as their
inorganic anatomy allowed: working, chatting, quarrelling, publishing
scholarly books, devising arguments for their own existence… It would
look for all the world like a community of thinking, conscious beings,
yet there would be no thought or consciousness at all. What enables me
to know that I exist may make it hard to know whether anyone else does.
The possibility that unthinking beings might produce every out-
ward appearance of thinking is nothing new. It seems also to follow
from the claim that artificial intelligence is impossible – that no inor-
ganic digital computer, no matter how powerful or cleverly programmed
and no matter how it interacted with its environment, could have real
consciousness or intentionality. For all anyone knows, it is physical-
ly possible to build self-contained inorganic robots – not controlled
by remnant people in vats – whose outward behavior and interactions
40 Eric T. Olson
would be indistinguishable from those of human beings. There could

be a vast society of such beings. If they could not really be intelligent,
then here too no amount of behavioural evidence – and no amount of
“intelligent” behaviour or dispositions so to behave – would guarantee
the existence of thought or consciousness.
We might even doubt whether there is any importance difference
between real thought and consciousness, which may require a subject,
and pseudo-thought and pseudo-consciousness, which don’t. Who
cares whether the particles arranged cerebrally in the vat, or the robots
that behave just like ourselves, could think or only pseudo-think? For
that matter, how could I ever know whether I am now thinking or only
pseudo-thinking? I proposed that if it seems to me that I am thinking (or
that thinking is going on here), then I really am thinking, because seem-
ing is itself a kind of thinking. But can I distinguish pseudo-seeming
from genuine seeming? That would again cast doubt on my knowledge
of my own existence.
Perhaps a good question to ask is whether pseudo-thought and pseu-
do-consciousness would have any normative status. I have a reason to
promote the satisfaction of my desires and to avoid pain, and I have a
moral duty to do the same for others. Have I got a reason or a duty to pro-
mote the satisfaction of pseudo-desires and to minimize pseudo-pain?
I wish I knew the answers to these hard questions, but I don’t. Brain
eliminativism may well raise more problems than it would solve.
References
Chisholm, R. (1976). Person and Object. La Salle, IL: Open Court.

Hudson, H. (2001). A Materialist Metaphysics of the Human Person.
Ithaca: Cornell University Press.
Inwagen, P. van (1990). Material Beings. Ithaca: Cornell University Press.
Johnston, M. (2007). “Human beings” revisited: My body is not an an-
imal. In D. Zimmerman (ed.), Oxford Studies in Metaphysics 3.
Oxford: Oxford University Press.
Olson, E. (2007). What Are We? New York: Oxford University Press.
Will I ever be a Cyborg?1
Rui Vieira da Cunha*
1. Introduction
Eric Olson’s animalist view relies on the premise that person is not a fit
candidate to be a substance concept, in Wiggins’s terminology. Instead,
he claims, animal is what best serves as the answer to what we most
fundamentally are and what determines our persistence conditions. Pro-
posing a thought experiment concerning inorganic replacement, I aim to
show that Olson’s animalist view cannot accommodate our very strong
intuitions about such cases. My claim is then that animalism either fails
on its own grounds or requires some tuning regarding what exactly an
organism is and its persistence conditions. I will examine Matthew Li-
ao’s attempt to accommodate such intuitions within an animalist view. I
will also describe the basics about Olson’s animalist view but let us first
look at the rough outline of my thought experiment. The essentials are
as follows, the rest I will provide later, as we go along.
1 For comments on an earlier draft of this paper, I am grateful to Arto Laitinen, So-
fia Miguens, Mikko Yrjönsuuri, and, most of all, Eric Olson, whose patience and
detailed comments were more than I could hope for.
* Researcher in the Mind Language and Action Group – MLAG – of the Institute
of Philosophy, University of Porto – FCT grant number FCT – SFRH/BD/45701/
2008. Jyväskylä, November 2009.
42 Rui Vieira da Cunha
2. A Cyborg2 Thought Experiment
Suppose the year is 2020. Tom has just been born. He is a human animal
(or just plainly an animal or a member of the species Homo sapiens or
an organism). Now suppose, for argument’s sake, that during his career
Tom will undergo transformations of the kind that transhumanists en-
visage as possible (and even likely and desirable) in the future. And let
us leap further into the future to describe such transformations.
In 2060, Tom has become an enhanced human. He has nanobots
running through his bloodstream that fight infections and aging pro-
cesses and monitor his life functions. Most of the matter that was ever
part of his body has been replaced by inorganic parts, far more efficient
and enduring. In short, he has technology of the sort some futurists
claim to be possible in the very near future and more sceptical scientists
assume possible in a more distant future (Bostrom, 2003). We can say
that Tom has almost become a full cyborg. Almost every part of his
body has been replaced by artificial parts. But not all the parts: Tom’s
brain (including his brainstem) has not been replaced.
In 2090, Tom is about to undergo another transformation. There are
now inorganic brains available and these come with inorganic brainstems,
that function as control and coordination centres much in the same way
organic brainstems have always functioned for human animals: they reg-
ulate the body’s metabolism, the capacity to breathe and circulate blood,
etc. The procedure is by then common to everyone: Tom’s brain and
brainstem will be gradually replaced, bit by bit and not all at once. At
no point in the procedure will Tom’s vital functions ever be interrupted.
Whatever memories or psychological features Tom may have, they will
be reproduced exactly in the inorganic brain. Nevertheless, Tom’s friends
are worried. They wonder if he will survive in the process.
2 I would like to thank Mimosa Pursiainen for pointing out that I am not using
the word cyborg in a technical sense. If we take a cyborg to be something partly
biological and partly inorganic, then we can only consider Tom a cyborg in the
first stage of our experiment, that is, in 2060. By 2090, after the last surgery, Tom
will be fully inorganic and he won’t be a cyborg. Still, for practical and stylistic
reasons, I chose to go with the word cyborg when I refer to Tom, both in 2060 as
in 2090.
Will I ever be a Cyborg? 43
Tom’s friends are worried because they believe Olson’s animalist

view to be true. They believe he has persisted through such a long time
as the same animal because his life functions have been assured by that
brainstem. And they now fear that undergoing the procedure may cause
him to cease to exist.
In order to fully grasp everything I am pursuing with this thought
experiment, let us stop for a while and recall Olson’s animalist view and
consider how this thought experiment and the resulting intuitions would
be explained on that view.
3. Olson’s Animalist View
Eric Olson’s defence of the animalist view on matters of personal identi-

ty rests on the claim that animal is a more adequate concept than person
to be a Wigginsian substance concept. Based on his reading of Wiggins
(1980, p. 15), Olson assumes that
every particular object falls under some kind or concept that tells us, in a special
sense, what the object is, and not merely what it does or where it is located or some
other accidental feature of it. And that concept determines persistence conditions
that necessarily apply to all (and perhaps only) things of that kind. Concepts of
this sort are substance concepts. (1997, p. 28)
Thus, a substance concept 1) tells us what the object is, as opposed to

merely telling us what it does, and 2) determines persistence conditions
that necessarily apply to all things of that kind. Moreover, we can infer
that two objects falling under the same substance concept will share
the same persistence conditions and that any object that falls under a
substance concept will always have those persistence conditions, i.e., “a
thing cannot change its criterion of identity partway through its career”
(Olson, 1997, p. 29).
Substance concepts, according to Olson, can be distinguished
from what Wiggins calls phase sortals, or sorts, such as child or ath-
lete or philosopher. These concepts do not tell us what the object most
fundamentally is but rather what it is at some point of its existence.
Furthermore, an object can undergo a change during its existence such

that it continues to exist but no longer falls under that phase sortal.
To take the example provided, something that is an athlete comes into
existence before he or she is an athlete and can cease being an athlete
without ceasing to exist. As Olson puts it, “there are such things as for-
mer athletes and potential athletes. Athletes don’t have the criterion of
identity that they have by virtue of being athletes.” (1997, p. 30).
According to the Biological Approach supported by Olson, our
substance concept, the concept that best answers the question about
what we most fundamentally are, is not person but rather Homo sapiens
or animal or living organism. In his own words: “Animal (or organism
or human animal) is a paradigm case of a substance concept, and so
is an ideal candidate for determining a thing’s persistence conditions.”
(1997, p. 36).
Olson’s rejection of person as a substance concept is based both
on the fact that personhood “is merely a capacity or ability of a thing”
(1997, p. 35) and not something that is closely connected with that
thing’s “internal, structural, or intrinsic features” (1997, p. 34)3 and on
the fact that biological entities like us can persist under the same condi-
tions as many non-people, such as human embryos and human vegeta-
bles. A human embryo or a human foetus or a human vegetable cannot
be counted as persons but they are still human animals and so, Olson
argues, that is what we most fundamentally are: animals. Saying that we
are animals is an excellent answer to the question what something is.
Person, on the other hand, is but a phase sortal, in the sense that to say
that something is a person doesn.t tell us what something is but rather
what it does and also in the sense that something (in our case, a human
3 One may of course wonder why a thing’s capacity or ability isn’t closely connected
to its intrinsic features. Moreover, one can doubt, along with Nichols (2010), the
grounds for the distinction between a thing’s structure and its capacities. In fact,
Nichols (2010) has objected to Olson’s line of reasoning on the grounds that (1)
human animal, animal, and organism are all functional concepts, and (2) the dis-
tinction between what something is and what something does is illegitimate in the
reading that Olson’s argument needs. Since my point here is to take the animalist
view at face value and argue that it faces a problem concerning inorganic replace-
ment, I will not argue against such assumptions here.
animal) can exist before coming to be a person and can continue to exist
(as a human animal) even after ceasing to be a person4.
One aspect of Olson’s account that should be evident by now is his
interchangeable use of the terms animal, human animal, and organism.
One could also include living organism in this list but it is less often
used and, considering that Olson deems life to be a fundamental charac-
teristic of organisms5, it would be redundant to do so, since there could
be no non-living organisms. Or so he claims. For the time being, I shall
also use those terms interchangeably, although later on some specifica-
tion will need be made.
When attempting to accurately describe the Biological Approach,
Olson is clear on a number of issues that concern us. First, it does not
exclude other kind of persons – wholly immaterial, like Gods or angels
or Cartesian Egos, or even material persons, made “out of nuts and bolts,
or wires and diodes” or of other biological species (1997, p. 124). And,
of course, it does not exclude that these different kinds of persons will
have different persistence conditions (1997, p. 27, p. 124f.); in fact, it
implies so. It does exclude, however, that an organism could ever come
to be a non-organism or a non-biological organism (1997, p. 125) –
and that is something to be explored when we return to my thought
experiment. Another important reminder when it comes to Olson’s an-
imalist view is that, as he warns us, it must not be confused with the
Bodily View that states that “we are identical with our bodies, or that
we persist just in case our bodies continue to exist” (1997, p. 19) or with
any modified version of it that would rely on some sort of physical cri-
terion of personal identity, that would somehow focus on the brain and
answer any question just by looking where the brain is6.
4 Olson further illustrates the inadequacy of person as substance concept by means

of an analogy with a “Locomotive Criterion” of identity (1997, p. 31).
5 Olson (1997, p. 136): “This proposal entails that an animal necessarily ceases to
exist when it dies. In that case there is no such thing as a dead animal, strictly so
called. We may call something lying by the side of the road a dead animal, but
strictly speaking what is lying there are only the lifeless remains of an animal that
no longer exists”.
6 Olson (1997, p. 144) gives a number of reasons against the idea that body could
be a substance concept – the fact that an object could be someone’s body for a
while and then someone else’s body and continue to exist, for instance. Or the
fact that an object can cease to be a human body and continue to exist. The main
Olson’s Biological Approach makes two main claims: first, that

we are animals, members of the species Homo sapiens. Second, that
psychological continuity is neither necessary nor sufficient for a hu-
man animal to persist through time. Now, the first claim does not
seem something that anyone would want to deny. At least, in a weak
sense, everybody agrees that we are animals or that we are constituted
by an animal or that we stand in some special relation to an animal.
What some may be tempted to deny is the stronger sense of that claim,
which in fact is exactly the one Olson endorses. In that stronger sense,
when we say that you are an animal, we are not just saying that you
are constituted by an animal or that you have the body of an animal
but that you are numerically identical with an animal, that you are
essentially an animal. To be accurate, this should actually count as a
further claim, or as a special qualification of the claim that we are an-
imals, since this is in fact the claim that we are essentially an animal.
The is in “Tom is an animal” is not the is of constitution but an is of
identity. And if that is correct, then we have our persistence conditions
by virtue of being animals, members of the species Homo sapiens.
Once the first claim is correctly understood in its stronger sense, in
this qualified form, the second claim – that psychological continuity is
neither necessary nor sufficient for a human animal to persist through
time – seems more easily acceptable.
Now, for Olson to enunciate those persistence conditions, he needs
to provide a more detailed account of what an animal is. This is the
point where it seems to me that Olson’s interchangeable use of the terms
animal, human animal, and organism requires some specification. In
the only two direct statements of identity conditions Olson provides
us with, one refers to animal7 and the other to organism8. However, it
seems clear to me that he is considering the same question in both of
them and although he constantly shifts between these terms, he seems
issue, however, is put more clearly by Olson (2007, pp. 25–26): “I have never seen
a good account of what makes something someone’s body […]. I am unable to
complete the formula ‘necessarily, x is y’s body if and only if…’”.
7 “If x is an animal at t and y exists at t*, x=y if and only if the vital functions that
y has at t* are causally continuous in the appropriate way with those that x has at
t.” (1997, p. 135).
8 “For any organism x and any y, x=y if and only if x’s life is y’s life” (1997, p. 138).
to show some preference for organism. I will then assume that, even
though Olson uses either of those terms interchangeably, it is organism
that somehow best expresses what he has in mind9 and it’s an organism’s
persistence conditions he is looking for. In doing that, ultimately he
relies on biology and so his account is scientifically informed. Metabo-
lism, teleology, and organized complexity are the relevant features of a
living organism. What about the identity condition of these organisms,
then?
In the process of explaining what an organism life consists in, Ol-
son constantly refers to the brainstem, “the organ that is chiefly respon-
sible for directing your life-sustaining functions” (1997, p. 140). That is
the reason I mentioned it when I presented my thought experiment and
since it is now time to get back to that, I will leave the considerations on
the brainstem’s importance for the next section of this paper.
Given this description of Olson’s animalist view, I think it can now
be asked what it is about my thought experiment that may counter this
view. You might be asking yourself what is the problem of a human
animal turning into a cyborg. Since the animalist only cares if there
is still a human animal living, even if the cyborg were to retain Tom’s
psychological characteristics (his memories, habits, and so on), the only
question would be whether he is still a living animal or not. Perhaps the
animalist can simply reply that the enhancements Tom underwent in
2060 have simply changed some of his features but his life functions
have persisted and so has the organ that directs them. And if Tom’s brain
and brainstem were to be replaced by an inorganic device in 2090, then
the human animal known as Tom would cease to exist. The fact that the
cyborg resulting from the procedure could have the same psychological
properties Tom had is not a problem for the animalist view, you could
9 I find some further comfort for this view in Olson’s later statement that “[…] an-
imals, including human animals, have more or less the same metaphysical nature
as other biological organisms. This is not to deny that some animals may have
properties of considerable metaphysical interest – rationality and consciousness,
for instance – that no plant or fungus could ever have. But if we ask what organ-
isms are made of, what parts they have, whether they are concrete or abstract,
whether and under what conditions they persist through time, and the like, I be-
lieve that the answer will be more or less the same for human organisms as it is for
plants and fungi. So we need an account of the metaphysical nature of organisms
generally.” (Olson, 2007, p. 27).
be tempted to reply. The resulting being would simply be something like

Tom’s copy. It would not have his biological life.
Perhaps now is a good time to disclose one essential fact of my
thought experiment. If you read it again, you will notice I never men-
tioned Tom’s choices or thoughts. In fact, nothing even remotely close
to Tom’s actual mental life. That is because in 2059 Tom fell into a per-
sistent vegetative state. Does that make a difference?
4. Back to the Cyborg
I think it is clear by now why I decided Tom should have such an un-
fortunate event in his future. Had I presented the thought experiment
in the usual manner, granting that Tom would still have higher mental
functions such as memory and reasoning even after becoming a cyborg,
the focus would be on the necessity of psychological continuity or con-
nectedness in the matter of personal identity. However, I have stated
that my aim is to show that the animalist view either fails on its own
grounds or requires some tuning regarding what exactly an organism is
and its persistence conditions. Unfortunately for Tom, this purpose is
best served if he is in a persistent vegetative state, maintaining all of his
vital functions throughout our experiment. So, what can we say about
the events in Tom’s imagined future and the way they could – if indeed
they could – affect our judgement concerning his numerical identity?
What plausible intuitions, if any, arise from those events? And, more
importantly, can the animalist view accommodate such intuitions?
Consider Tom in 2059, when he lapses into a persistent vegetative
state. Let’s assume it is an irreversible one: Tom’s cortex is damaged
beyond repair, even if his brainstem is untouched. After the enhance-
ment surgery, we could practically call him a full cyborg, even if his
brain (still damaged) and his brainstem (untouched) are not replaced.
The most plausible intuition would be to say that Tom survived such
a process. That is, of course, if we grant that Tom still persisted after
lapsing into a vegetative state. Even if some of us would say that he
did not, I believe they would probably say this because of the lack of
psychological features. But in that case it would be the audience who

would reject Olson’s view to begin with. Olson’s approach grants abso-
lutely no importance to those features. Its main characteristic is that it
puts biology in the place of psychology (Olson, 1997, p. 16). That is not
the audience I am seeking. I want to show that we can be persuaded by
the animalist view and still have some very strong and plausible intui-
tions about this thought experiment that that view cannot explain. And
I also believe one of the strongest points for the animalist view is that
it grants that in a case such as Tom’s in 2059 (Olson’s Human Vegeta-
ble Case) you would persist. The human animal that you are would not
cease to exist just because its higher mental functions are gone. I think
that is a very strong intuition most of us will share with the animalist
view.
Let us now turn to Tom’s state in 2090. Could the severely mentally
debilitated Tom survive the last surgery? Would it still be Tom after that
procedure? I believe we can say so. You may object that such a pro-
cess of inorganic replacement is neither possible nor conceivable. These
procedures do not strike me as impossible or inconceivable. To put it
as Parfit does (1984, p. 219)10, I would deem them technologically im-
possible but not deeply impossible. Regardless of what our position on
thought experiments and their validity in drawing conclusions in philos-
ophy, it seems this one is not more far-fetched than any other commonly
described in addressing personal identity: Olson himself considers the
possibility of inorganic replacement – and even in a situation where
consciousness and thought is maintained during the process, as we shall
see. If we can consider that, my thought experiment must be a lot easier
10 “It may be impossible for some of these cases to occur, whatever progress may
be made in science and technology. I distinguish two kinds of case. Some cas-
es contravene the laws of nature. I call these deeply impossible. Other cases are
merely technically impossible. Does it matter if some imagined case would never
be possible? This depends entirely on our question, on what we are trying to show.
Even in science it can be worth considering deeply impossible cases.” I cannot be
so bold as to claim that inorganic replacements of the kind envisaged here do not
contravene any law of nature. However, even if they do, there is still the possibility
(also sustained by Parfit 1984, p. 219) that we can derive some intuitions from
them, as long as we pose the right questions. And I am fairly confident that the
scenario imagined here includes and specifies all relevant conditions as Wilkes
(1988, p. 9) would demand.
to consider. And I argue that the intuition it elicits is that Tom or you and
I or any other human animal could survive such a process.
Remember the story once more: Tom had already lapsed into a per-
sistent vegetative state when his transformations began. If not having
any of the other parts removed made any difference, why would the
brain and brainstem be more important? His mental life has been gone
since the beginning of the experiment, so that is not interfering. If that
did not keep us from saying he ceased to exist, why would it deter us
now? But perhaps it is not the psychological aspects. Perhaps by now
you have become so persuaded by the animalist view that the idea of
a brainstem replacement is holding you back. The animalist view is
clearly attractive in that it can account not only for our persistence con-
ditions but for those of other animals as well. And it not only coincides
with our intuitions about numerical identity in most of our ordinary life
but it also seems scientifically informed, as far as possible. So, since
you have figured out the importance Olson’s animalist view attributes
to the brainstem, you might feel tempted to say Tom would die if he
had his brainstem replaced with an inorganic one. But why should the
brainstem matter all that much?
I believe the explanation of the brainstem’s importance in the ani-
malist view is best given if I present what I take to be two possible read-
ings of Olson’s account of an organism’s persistence conditions. These
are not absolutely incompatible readings, I think, but they do contain
some discrepancies and might render different results when addressing
Tom’s situation. To be thorough, I will begin with another reading, one I
wouldn’t even consider a possible reading at all but which, even if I con-
sider it incorrect, I must mention, since there seems to be some room
to interpret Olson in that sense. In his explanation of what is an organ-
ism’s life, Olson’s writing is increasingly analogical and metaphorical.
He compares life to thunderstorms, for instance, to account for the fact
that a life permanently integrates new matter into the organism which
life it is (“A life is a sort of storm of particles in constant motion” 1997,
p. 136). He states at a certain point that “Every organism has a life,
and it is hard to see how there could be a life without there being an
organism whose life it was. And an organism cannot be animated by two
lives, at least not at once.” (1997, p. 137). Some readers might take this
as Olson engaging in some sort of vitalist path, some sort of inhabitancy
of the body by a life, somewhat like a breath of life or something al-

most as mystical. That, I think, couldn’t be further away from Olson’s
intentions and, in fact, from his achievement and I hope it is clear why I
dismiss this. Olson explicitly mentions biology and biological issues so
many times that it would simply be unbelievable to assume that, not to
say downright inconsistent. If it means anything at all, I would venture
to say it can be no more than Olson’s – and any philosopher of biology
or even any biologist, for that matter – inability to scientifically convey
what life is in a biological sense.
So then, what are the true possible readings of Olson’s account
of an organism’s persistence conditions? The first is the reading that
focuses on the brainstem’s importance directly or just by itself. Olson
seems to endorse this reading several times: when describing the Hu-
man Vegetable Case, for instance, he states that “Neither is the animal
“brain-dead”, for those parts of its brain that direct its vegetative func-
tions remain fully intact.” (1997, p. 8). Later on, when considering a
case where the brainstem is missing, he states:
But a detached cerebrum is not an animal […] because its parts do not coordinate
their activities in the way that the parts of an organism coordinate theirs. […] The
reason is not just that many life-sustaining organs –heart, lungs, digestive tract,
and just about everything else – have been removed from the cerebrum, but also
that those organs that once coordinated the life-sustaining functions that went on
in the arm or cerebrum have been cut away. (1997, p. 115)
And, more emphatically, he also states “I have suggested that your

brainstem, as the organ that is chiefly responsible for directing your
life-sustaining functions, is essential to you, for without it there is no
Lockean life and no living human organism at all.” (1997, p. 140). There
are many other passages of this sort11, which may lead one to think the
brainstem’s importance is so that one cannot live without it: in fact, we
have just seen Olson stating this. And one may be led to assume that it is
absolutely impossible for anyone to have their brainstem replaced and,
11 See, for instance, (1997, p. 132): “That is because the organs that once directed
those activities – the pons, medulla oblongata, and hypothalamus, among others –
are missing.”); or 1997, p. 10: “Imagine, then, that our surgeon leaves the rest of
you intact when she removes your cerebrum, so that your brainstem continues to
do its job of directing your heartbeat, circulation, breathing, and digestion […]”.
therefore, that Tom is no exception. On this reading, then, Olson’s view

would accommodate our intuitions about Tom in 2060. At that time
Tom is still a human animal. He just happens to have had most of his
parts replaced by inorganic ones. He is reduced to his brain and brain-
stem, that much is obvious. But in Olson’s view that would probably be
only an extreme case of amputation, of an animal being pared down to
the bare minimum that ensures his survival12. His brainstem, the organ
responsible for his life function, has suffered neither any sort of injury
nor intervention whatsoever. It still plays the same role as in any other
human who has not been enhanced the way Tom was. True that most of
Tom’s body is made up of inorganic attachments which are not part of it.
But that is not problematic for the animalist view: attaching a prosthesis
to Tom cannot bring about his death. And it is definitely not a matter of
how many prostheses or how many organs and parts are being replaced
but of which organs and parts. Now, Tom in 2090 would be a complete-
ly different matter for the animalist view. Remember Olson stated that
there would be no living organism without its brainstem. If this reading
of Olson’s view were right, then we would have to say that Tom dies in
2090, when he has his brainstem replaced.
However, that is not what I would consider Olson’s truest reading,
and even those evidences of it are, it seems to me, pretty clear concern-
ing what Olson actually means when he refers to the brainstem. It is not
that the brainstem matters per se but that it matters in as much as it con-
trols the vital functions of the organism. In all of the above statements,
every reference of the brainstem includes a reference to its role, its job,
its activity, within the organism: it coordinates the vital functions, it
directs those life-sustaining functions, it is chiefly responsible for those
functions. So, the important thing is not the brainstem but the brain-
stem’s job, the functions it coordinates. In case the quotations above are
not convincing enough, we may consider these:
Your life-sustaining functions are not disrupted when you lapse into a persistent
vegetative state (1997, p. 11);
12 It would be pretty much like the case of a detached head connected to a life-sup-
port system Olson considers (1997, p. 133).
[…] the human vegetable in the story is biologically continuous with you – that
your life-sustaining functions continue on in that animal […] (1997, p. 12);
What it takes for us to persist through time is what I have called biological con-
tinuity: one survives just in case one’s purely animal functions – metabolism,
the capacity to breathe and circulate one’s blood, and the like – continue. (1997,
p. 16).13
So, if we suspected the first reading to be not quite wrong but not quite
accurate either, we can now look at this second reading as the more rig-
orous one, the one Olson himself would certainly endorse. It is not the
case that the brainstem matters by itself but that it matters in a derivative
sense, because of its functions14.
So, is Olson’s view on this second and more accurate reading
now able to accommodate our intuitions about Tom in 2090? It would
seem that there would be nothing to object, from Olson’s view, to a
replacement of the original brainstem by an inorganic brainstem. If the
13 Other examples could be the following: “[…] you are an animal, and an animal
ceases to exist when it dies – when its vital functions cease and its tissues decay
beyond the point where they can be reanimated. […] you could be immortal only
if it is possible for the life-sustaining functions of a biological organism to con-
tinue forever.” (1997, p. 71); “On the Biological Approach, what it takes for us to
survive remains the same throughout our careers: like other animals, we persist as
long as our life-sustaining functions remain intact.” (1997, p. 89); “[…] its organic
functions are continuous with yours: your metabolism and other life-sustaining
functions have continued on without interruption and are now the life-sustaining
functions of the brainless animal.” (1997, p. 116); “Consider the biological con-
cept of death. […] most biologists would agree that it has something to do with the
irreversible cessation of those metabolic and other activities that distinguish living
organisms from non-living things.” (1997, p. 119).
14 Note that on the first reading the organ is being enthroned on itself, because of
being what it is and not of doing what it does. My stress in this last sentence is
merely to draw attention to the fact that Olson’s most suitable reading, the one
which is closer to his intentions and which has on its side the overwhelming tex-
tual evidence, is one in which the basics of his substance concept are defined on
the grounds of what something does rather than what something is. And that was
exactly Olson’s reason to reject person as a substance concept in the first place. Of
course it can be argued that these are different levels of explanation. But it has also
been argued – and very convincingly so, I may add – that the distinction between
what something is and what something does is illegitimate on the reading Olson
needs to prefer animal over person as a substance concept – see Nichols, op. cit..
brainstem is granted such importance only because of its functions,

one might think that it is irrelevant whether it is something organic
or inorganic, as long as it does its job, as long as it coordinates the
life-sustaining functions.
Unfortunately, even though it may seem unproblematic, the idea of
an inorganic brainstem is something the animalist view cannot accept.
At a certain point, considering the possible existence of other kinds of
persons different than human animals, Olson is very clear on stating
that no organism can survive a process of inorganic replacement of its
brainstem:
It may be possible to replace all of your parts, including your brain, gradually and
piece by piece, with inorganic prostheses in such a way that your mental capaci-
ties were preserved throughout […] The result would be a wholly non-biological
person – with rationality, consciousness, free will, the works – who was both
psychologically and materially continuous with you. Nevertheless, according to
the Biological Approach that being would not be you, for you are a biological
organism, and no organism could come to be a non-organism (or so I shall argue).
(1997, p. 125)
The argument Olson is referring to in the end of that statement is his

account of what an organism is, as I presented it supra, with the rele-
vant features being metabolism, teleology and organized complexity.
On Olson’s view, Tom would survive just until 2090, because his brain-
stem is controlling his life-sustaining functions. The moment it would
be replaced, the organism would perish. However, it’s not the case that
Tom would perish just because his vital functions were gone or because
there was nothing there to direct them. Tom would also cease to exist
because he would cease to be an organism and, given Olson’s insistence
that organism is a substance concept, Tom could never cease to be an
organism without ceasing to exist.
Olson is perfectly aware, I think, of the problem that inorganic re-
placement poses to the animalist view and so, after he has presented his
account of what an organism is, he addresses the issue once more:
… imagine that your brainstem is replaced by an inorganic substitute gradually, bit

by bit, rather than all at once […] there is never a period when your life-sustaining
functions are left without an organ to coordinate them, or when your cerebrum is
not aroused and activated in the normal way by the brainstem. As a result, there
need be no interruption in consciousness throughout the operation (suppose the

surgeons use only a local anaesthetic). The result would be a rational, conscious
being with your mind. Isn’t it obvious that you would be that being? My view,
however, entails that you could not survive this … For something with an inorgan-
ic brainstem, I argued, could not be an animal at all (1997, pp. 141–142).
In this statement, Olson is focusing on a situation where consciousness

is maintained throughout the whole process and so it seems that the only
reason why we intuitively say that the being resulting from the operation
would be you is due to psychological factors. In Tom’s case, however,
there are no such factors to consider and, I argue, the intuition remains:
we still think Tom is Tom and has not ceased to exist just because he has
that last surgery in 2090. So, given that Tom can become a cyborg and
given that he started out as a human organism, there seems to be only
two possible answers (apart from denying that Tom has ceased to exist):
either to conclude that Tom was not, after all, a human organism, at least
not essentially; or to conclude that an organism can survive inorganic
replacement15. Either animalism fails on its own grounds, given that
this thought experiment was “untainted” with psychological features,
or animalism requires some adjustments concerning what exactly an
organism is and what its persistence conditions are. The first answer
amounts to giving up the animalist view altogether. The second answer
would allow us to accommodate our intuition within the animalist view.
But how could this be done?
One way of going about it is Liao’s way. Liao (2010) argues for
the possibility of accommodating a process of inorganic replacement
within the animalist view – or the Organism View, as he puts it. Liao
considers that a being, X, is essentially an organism, if
a) X begins to exist when the capacity to regulate and coordinate its metabolic and
other life processes is there; b) X persists as long as there is what may be called
“organismic continuity”, which is the continuing ability to regulate and coordinate
its metabolic and other life processes; and c) X ceases to exist when the capacity
to regulate and coordinate its metabolic and other life processes is permanently
gone. (2006, p. 337)
15 Here I am particularly indebted to Eric Olson, who has shown me the need to
clarify this point.
This account of what an organism is, is not all that different from
Olson’s. The main difference is that Liao thinks that “there could be
non-carbon-based life forms that have non-carbon-based interdepend-
ent parts that are used to regulate and coordinate various life processes
such as absorption, assimilation, metabolism, and so on, in order to
process certain material into fuel so that they would be able to func-
tion. If so, it seems that these non-carbon-based life forms would also
qualify as organisms.” (2009, p. 17). If this were correct, then inorgan-
ic replacement would no longer pose a problem to the animalist (or
organism) view and our very strong intuition that Tom would survive
the surgery in 2090 may sit easily with that view. Liao (2010, p. 68)
even presents a thought experiment similar to Tom’s except that con-
sciousness is maintained throughout the process and the experiment is
in reverse: you would begin your existence as a non-carbon-based life
form and then have all your parts gradually replaced by carbon-based
functional equivalents.
I said in the beginning that my thought experiment poses a problem
to the animalist view. Given a case like Tom’s, I think an animalist is left
with two options: he can bite the bullet and deny Tom survives the 2090
replacement of his brainstem or he can choose Liao’s solution. If he de-
nies that Tom survives after 2090, he cannot accommodate our intuition
that Tom does not cease to exist just because a small organ is replaced. In
this case, I argued, my thought experiment is more damaging to the ani-
malist view because they cannot dismiss our intuition just by saying that
we are biased or influenced by psychological considerations. That option
is not available for an animalist in Tom’s case and so he must admit this is
a strong blow to the animalist view. If we somehow think Tom can survive
a gradual process of inorganic replacement, it seems we can no longer
say that Tom falls under the substance concept “organism”. After all, Tom
began to exist as an organism but has ceased being an organism without
ceasing to exist. So, is being an organism just a phase in Tom’s existence,
just like being a child, an athlete, or a philosopher? It certainly seems so.
As for Liao’s solution, I believe it comes at a very high cost for the
animalist (or organism) view. The cost of redefining what an organism
is in such a broad way that it may cease to be a substance concept. Of
particular note, for instance, is Liao’s case for the possibility or conceiv-
ability of non-carbon-based life forms:
But while it may be the case that organisms that are most familiar to us are all
carbon-based life forms, there is no reason to suppose that all organisms are nec-
essarily carbon-based life forms. Strictly speaking, organisms are just entities that
have interdependent parts. (2009, p. 17, my emphasis16)
You can object that there is nothing wrong with Liao’s solution, that
it just broadens the concept of organism. In doing so, you would say,
Liao is just providing a more detailed account of what an organism is.
Isn’t that what I said my thought experiment was meant to show about
the animalist view? That in order for it not to fail on its own grounds, it
would need some finer tuning concerning what exactly an organism is
and its persistence conditions? That is true. I do not think, however, that
Liao’s tuning can be made while still maintaining organism as a sub-
stance concept17. Remember, a substance concept in Wigginsian termi-
nology is something that tells us what a thing fundamentally is. It tells
us that thing’s most essential metaphysical nature. It helps to set that
thing apart from other things which do not share its nature. If we take
Liao’s solution, even though his explicit definition of what an organism
is seems very close to Olson’s, we are broadening the concept to such an
extent that I believe too many different things in the world could count
as organisms.
Olson himself is very aware of all the problems this poses for the
animalist view, I think. Right after addressing the topic of inorganic
replacements, in the above excerpt, he states that there is little that he
can say to reply to our intuition that we could survive such a process.
16 My emphasis here should by no means be understood as saying that this is all there
is to Liao’s definition of organism – which I have already presented earlier. What I
want to point out is that given that definition, and the fact that an organism could
be non-carbon-based in his account, that aspect – being an entity that has interde-
pendent parts – is one of the main features of an organism. And that is something
perhaps too many things in the world possess.
17 I thank Arto Laitinen for suggesting that perhaps Liao’s solution would require
us to think of organism as the substance concept and of carbon-based or non-
carbon-based as a phase sortal. Though I am tempted to say that would be an
accurate rendition of Liao’s proposal, I am not sure if it is. And I am also not sure
if that explains away the fact that we do tend to think of organisms and its being
carbon-based as one essential feature of it, as something an organism could not
cease to be without ceasing to exist. Talk of a non-carbon based organism may be
acceptable but doesn’t that force us to count many different things as organisms?
And he is clear regarding the options that are left: “If you want to insist
that you could survive such a thing, you must either deny that you are a
living organism, or come up with an alternative account of what it takes
for an animal to persist through time” (1997, p. 142). Liao’s solution is
probably an attempt at such an alternative account. However, the way
I understand it, it can also be seen as falling terribly close to being an
outright denial that we are essentially living organisms, that is, that or-
ganism is the substance concept under which we fall.
I have tried to show how Olson’s animalist view fails to explain our
strong intuition about Tom’s survival in my thought experiment. Even
after trying two possible readings of Olson’s account of an organism’s
persistence conditions, there is no plausible answer the animalist view
can offer us. Since I resorted to no psychological or mental continuity or
connectedness of any sort to bring about such intuitions, there is no way
we can say that we were influenced by a psychological approach nor, I
dare say, by practical attitudes about “what matters in identity” (Olson,
2007, p. 43). My claim that animalism fails on its own grounds thus
means that it cannot offer plausible answers even to an audience willing
to consider only biological factors when accounting for our persistence
conditions. I see no other alternative than to conclude that animalism
requires some adjustments regarding what exactly an organism is and
what its persistence conditions are. Whether these adjustments can be
satisfied with something like Liao’s solution, I am not quite sure, con-
sidering the possible implied consequences of such a solution on the
central premise of animalism: that organism is a substance concept.
5. Conclusion
The main premise of Eric Olson’s view on personal identity issues is

that animal or human animal or organism is what best serves as the
answer to what we most fundamentally are and what determines our
persistence conditions. That that is, in short, our substance concept.
Resorting to a thought experiment concerning inorganic replacement,
I have tried to show that things like you and I can become totally
inorganic. If organisms cannot become such inorganic things, then you

and I are not (essentially) organisms. If organisms can become inor-
ganic things, then we need a better definition of organism. As I claimed
at the beginning of this paper, Olson’s animalist view cannot accom-
modate our very strong intuitions about such cases and the only op-
tions left for an animalist, apart from trying to deny that premise – that
Tom and you and I can become totally inorganic – are either to give up
animalism altogether or come up with a better account of what exactly
an organism is and what its persistence conditions are. Liao’s solution,
even though cleverly argued, is one that in attempting to achieve the
latter, does so in such a broad way that it comes too close to achieving
the former.
References
Bostrom, N. (2003). Human Genetic Enhancements: A Transhumanist

Perspective. The Journal of Value Inquiry, 37, 493–506.
Liao, M. (2006). The Organism View Defended. The Monist, 89, no. 3,
334–350.
Liao, M. (2010). Twinning, Inorganic Replacement and the Organism
View. Ratio, 23, 59–72.
Nichols, P. (2010). Substance Concepts and Personal Identity. Philo-
sophical Studies: An International Journal for Philosophy in the
Analytic Tradition, 150, no. 2, 255–270.
Olson, E. (1997). The Human Animal: Personal Identity Without Psy-
chology. New York: Oxford University Press.
Olson, E. (2007). What are We? A Study in Personal Ontology. New
York: Oxford University Press.
Parfit, D. (1984). Reasons and persons. Oxford: Clarendon Press.
Wiggins, D. (1980). Sameness and Substance. Cambridge, MA: Har-
vard University Press.
Wilkes, K. (1988). Real People: Personal identity without thought ex-
periments. Oxford: Clarendon Press.
Part II
Epistemology and
Phenomenology
How Consciousness explains the Self
Klaus Gärtner
What is the ‘Self’? This question has a long tradition in Philosophy of

Mind. For some Philosophers, like Descartes, it is the entity which holds
our consciousness, while others, like Hume, deny its existence. What-
ever it is – whether it exists or not – there seems to be no doubt about
how we can access something that we are inclined to call a ‘Self’. Since
Descartes, this access has been our Consciousness. While trying – but
still failing – to obtain an explanatory concept of Consciousness we are
still in the state of determining whether the ‘Self’ really exists, and if so
in what form. I therefore propose the following hypothesis: A concept
of the ‘Self’ depends necessarily (but not sufficiently) on a concept of
Consciousness. What will be the aim? Well, to establish a conceptual
relation between Consciousness and the ‘Self’ and if possible to obtain
a tool able to handle a concept of the ‘Self’. What we can gain from this
approach is that our concept of the ‘Self’ will depend on the choices,
experiments, ideas and conceptual background that also account for the
concept of Consciousness. I will propose the following methodology: we
have to introduce an epistemological relation, where Consciousness can
be seen as the access to whatever it is we call a ‘Self’. This epistemolog-
ical approach will depend in part on phenomenological reasoning since
this kind of reasoning is able to show limits and problems regarding
certain concepts. Furthermore, this approach seems to be integrative and
therefore connects a variety of related aspects in Philosophy of Mind.
The key to establishing such a relation will be Self-Consciousness.
64 Klaus Gärtner
Justifying the Connection
What is Consciousness?
I do not want to discuss all the different theories of Consciousness.

Rather, I want to show which is the relevant one to support my argu-
ment. Basically there are three general categories of Consciousness:
Creature Consciousness, State Consciousness and Consciousness as
an entity.1 I will focus on Creature Consciousness only. Creature Con-
sciousness is concerned with how an “[…] animal, person or other
cognitive system may be regarded as conscious in a number of different
ways.”2 I decided to focus on Creature Consciousness because this is
the specific set of characterizations used where we actually describe
in what way someone or something is conscious, that is to say what
attributes someone has to have to be considered a conscious creature.
To explore the ‘Self ’ three criteria seem to be interesting: the Wake-
fulness –, the What it is like – and of course the Self-Consciousness
criterion. There are three more criteria mentioned within the context
of Creature Consciousness which are either criticisable, because they
can be considered an understatement (Sentience), depend on conscious
states (Subject of conscious states) – which were already excluded –,
or which involve how consciousness is directed at an object (Transitive
Consciousness), but this does not tell us anything about accessing the
‘Self ’.
The Wakefulness criterion tells us about the capacity of an organ-
ism to respond to its surrounding environment or better one can only be
conscious when one is awake and alert. What this means is something
or someone cannot be conscious when asleep or in a coma. The What
it is like characterization gives us a more subjective notion. When we
are conscious it seems to be the case that there is a certain ‘something
what is like to be that organism’. This is famously described in Thom-
as Nagel’s article ‘What is it like to be a bat?’3. The third and most
important tool to access our ‘Self’ is the Self-Consciousness criterion.
1 See Van Gulick (2009).
2 Van Gulick (2009).
3 See Nagel (1974).
How Consciousness explains the Self 65
This is the awareness that one is aware. Those are the three criteria to
be used describing access to the ‘Self’. Again, we have to keep in mind
that this is not everything there is to say about Consciousness. These
ideas seem to be the critical ones when we are talking about accessing
the ‘Self’ and when we are trying to show how a concept of the ‘Self ’
depends on a concept of Consciousness.
What is the ‘Self’?
This question cannot be answered here. It is important, however, to

show a relation to Consciousness so that perhaps in the end there might
be a solution to this problem. Nevertheless we have to at least charac-
terize the ‘Self’ to see what we are dealing with.
The problem of the ‘Self ’ is characterized by the subject of
mental phenomena. It is the entity which bears sensations, percep-
tions, thoughts, desires and actions.4 Usually it is referred to as a
first-person perspective. This is not to say that the ‘Self ’ can be called
an object, it does not seem to be a thing such as something we refer
to when we say ‘it is sunny’. However short and problematic this
characterization may be, it seems to be more interesting how one can
think about the ‘Self ’.
According to Colin McGinn three mature theories of the ‘Self ’ can
be identified:
theories which identify the self with the body; theories which explain the self in
terms of various mental relations; and theories which take the concept of the self
to be primitive and not to be explained in terms of anything else.5
I think we have to add a fourth theory: denying the ‘Self’ completely6.

The first theory explains that our self-reference refers to the body, an
object with mental attributes. The second theory states that our self-ref-
erence refers to complex mental states which belong to the ‘Self ’. The
third theory claims that the ‘Self’ is a substance different from the
4 See McGinn (1998).

5 McGinn (1998, p. 144).
6 As an example see Metzinger (2003).
66 Klaus Gärtner
body – non-reducible to mental states – which is simple and basic. The

fourth theory denies the ‘Self’ in all of its forms, which means there are
multiple reasons to reject a theory of the ‘Self’ (for example, identity
disorders).
Those approaches to the ‘Self’ are very general. It is not my in-
tention to go into further detail. This small introduction has only been
given to keep in mind that there are different approaches. Now we can
get to the real problem: why is a concept of Consciousness necessary
(but not sufficient) for any concept of the ‘Self’?
The relation between a concept of Consciousness

and a concept of the ‘Self’
The questions are now the following: How can we connect a concept of
the ‘Self’ to a concept of Consciousness? Furthermore, we also have to
ask ourselves: Why does a concept of the ‘Self’ depend on a concept of
Consciousness?
Well, the answers depend on the three criteria of Consciousness
which have something to do with our access to the ‘Self’. So, how
can we connect the above criteria – Wakefulness, What is it like and
Self-Consciousness – to the ‘Self’? Intuitively, we would say that, well,
our Consciousness has to have a certain Subject, Consciousness does not
seem to exist for itself and classically that structure has been the ‘Self’.
That, however, might not be true anymore, but for a certain form of In-
trospection or Phenomenology it seems to be necessary. Since at least in
these terms all three criteria need a relation to the ‘Self’. Let us start here.
It seems to be obvious that the relation between the ‘Self ’ and
Self-Consciousness is the strongest. What we have to do are two things:
first, we have to find the relation between Self-Consciousness and the
‘Self’ and second, explain how the other two characteristics can be in-
cluded in that relation. The first part does not seem too complicated. In
most of the literature that introduces the ‘Self’, the ‘Self’ is somehow
connected to Self-Consciousness (to talk about her ‘Self’ she needs a
certain self-awareness). This awareness can be explained by “[…] dif-
ferences between self-awareness and the awareness of the objects of
experience that enter into the content of experience […].”7 Ascribing

this capacity to someone seems to fulfil our notion of the ‘Self ’. A sub-
ject, for example an animal, that has a certain living body but lacks the
idea of a psychological subject appears to be different in nature from
this idea – at least that is what we think. That means an animal is a
subject that has mental states but only in a living body without access
via Self-Consciousness. This Self-Consciousness gives us a special kind
of unity which is a unity from the inside. So we can say the following:
What this position implies is that the kind of unity a self exemplifies is bound up
with self-consciousness; without self-consciousness the mind of a creature has
no more unity than that conferred body – from the inside it is just a collection or
succession of mental states.8
As we can see, the relation between the ‘Self ’ and Self-Consciousness

is established.
It is now possible to talk about the second problem. How can we
include the Wakefulness criterion and the What is it like criterion into
this relation? We can connect both criteria to Self-Consciousness. When
we talk about phenomenological approaches to Self-Consciousness we
can claim that two kinds of Self-Consciousness exist. It is important that
there is a pre-reflective – apart from the reflective – Self-Consciousness
which means that “[…] (1) it is not an explicit or thematic form of
self-consciousness, and (2) reflective self-consciousness is possible
only because there is a pre-reflective self-awareness that is an on-going
and more primary self-consciousness.”9 This notion incorporates the
idea of What it is like because it is a subjective feeling of self-awareness
and therefore implies certain qualities. Thus, this is the notion defend-
ed by Thomas Nagel10. So the problem of including the What it is like
criterion is basically to give this criterion a Self-Consciousness base.
What is suggested is “[…] that first-person experience presents me
with an immediate and non-observational access to myself, and that
7 Morrison (1996, p. 550).

8 McGinn (1998, p. 143).
9 Gallagher and Zahavi (2009).
10 See Nagel (1974).
68 Klaus Gärtner
(phenomenal) consciousness consequently entails a (minimal) form of

self-consciousness.”11
This solves our problem of how to entail the What it is like criterion,
but still does not give us a solution for the Wakefulness criterion. That
is a bit trickier, but can also be solved by the phenomenological notion.
What we have to claim here is that the pre-reflective self-awareness has
to be distinguished from an explicit self-awareness. This seems to be the
case, since we are not talking about an awareness that is observational;
rather, it is a non-observational self-acquaintance. We can put it this
way:
To have a self-experience does not entail the apprehension of a special self-object;

it does not entail the existence of a special experience of a self alongside other
experiences but different from them. To be aware of oneself is not to capture a
pure self that exists separately from the stream of experience; rather it is to be
conscious of one’s experience in its implicit first-person mode of givenness.12
This contradicts the idea of some Philosophers who claim that children
do not have a Self-Consciousness, while in psychology, the evidences
seem to point in a different direction13. This is the basic idea of mini-
mal Self-Consciousness, which is also defended by some contemporary
analytical Philosophers14. Minimal Self-Consciousness is primitive and
already present from birth. It is prior to the learning of language and the
ability to conceptually form judgements and seems to be the basis of a
more advanced form of Self-Consciousness. With this minimal notion
of Self-Consciousness it is now possible to include Wakefulness in our
relation. We just have to claim that every form of consciousness entails
this primary pre-reflective Self-Consciousness. Object experiences do
not enter our mind in a purely conscious form; there is always some sort
of ownership of an experienced object. Excluding this ownership seems
to be implausible.
Of course phenomenological reasoning cannot give the complete
answer. It seems merely impossible to justify our constitution as hu-
mans – especially when we talk about the mind – only through certain

13 See Gallagher and Zahavi (2009).
14 See for example Goldman (1976).
problems and limits we have but, at the same time, to establish a relation
of certain problems and concepts seems to be necessary.
We can now try to explain why a concept of the ‘Self’ depends on
a concept of Consciousness. The idea is simple and has already been
mentioned above. We cannot know anything about the ‘Self’ without
accessing it and the way we access our ‘Self’ is via Consciousness. It is
therefore necessary to have a concept of Consciousness first. Depending
on that concept we can analytically access the ‘Self’. At this point we
normally say that is not what is happening: Whatever Consciousness is
and how it is grounded, we usually think of it as just a tool for intro-
spection. That might not be entirely right. I argued above that a certain
pre-reflective Self-Consciousness forms part of all sorts of Conscious-
ness; there is, at least on a pre-reflective level, a direct phenomenological
link to our ‘Self’. Since that is the case and there are possibilities to an-
alyse Consciousness – but the ‘Self’ only via Consciousness – we need
a concept of Consciousness first. That turns a concept of Consciousness
into a necessary condition for a concept of the ‘Self’.
Is a concept of Consciousness also sufficient for a concept of the
‘Self’? Here the answer has to be no. It seems to be an overstatement to
claim: accessing the ‘Self’ via Consciousness may constitute the ‘Self ’.
The ‘Self’ is only the content or character of our Consciousness15 and
we can therefore be wrong. Even if we try to argue that there is a di-
rect link of acquaintance at the pre-reflective level, that is, a relation of
sensing a certain ownership within the What it is like, in a way these
sorts of sensations may exist, but this does not mean that their content
has to be true. It might be that in the end there is nothing more than our
Self-Consciousness, which in its pre-reflective form, has a certain What
it is like character and that this character is sensed with an ownership,
but this still leaves the possibility open that a ‘Self ’ might be an illu-
sion. It is true that it has been argued that this very basic relation forms
part of being a certain creature, like a bat16, but even though we need
this sensation and this sensation might be different in every life form
and subject, it does not follow that the world is the way we sense it.
This would leave us with no more than Self-Consciousness and a certain
15 This depends on what theory of Consciousness one defends.

16 See Nagel (1974).
70 Klaus Gärtner
sensation, but this does not directly imply a ‘Self ’. Therefore, it seems
clear that the sufficiency condition cannot be fulfilled.
Conclusion
This basically leads to the idea of the hypothesis: a concept of Con-

sciousness is necessary (but not sufficient) for a concept of the ‘Self ’.
This implies certain things: first, there is no claim about ontological
existence. We cannot say what is first: Consciousness or the ‘Self ’. Sec-
ond, we cannot decide whether the ‘Self’ does exist or not. What we
can argue though is: if we suppose that the ‘Self ’ exists (some people
seem to think that it has to) its conceptual form depends on the choices,
experiments, ideas and conceptual background that we make for the
concept of Consciousness.
References
Block, N. (1996). Consciousness. In Guttenplan, S. (ed.), A Companion

to the Philosophy of Mind. Oxford and Malden: Blackwell Publish-
er, 210–219.
Brook, A. & Raymont, P. (2009).The Unity of Consciousness. In Edward
N. Zalta (ed.), The Stanford Encyclopedia of Philosophy (Winter
2009 Edition), <http://plato.stanford.edu/archives/win2009/entries/
consciousness-unity/>.
Gallagher, S. & Zahavi, D., (2009). Phenomenological Approaches
to Self-Consciousness. In Edward N. Zalta (ed.), The Stanford
Encyclopedia of Philosophy (Spring 2009 Edition), <http://plato.
stanford.edu/archives/spr2009/entries/self-consciousness-phe
nomenological/>.
Goldman, A. (1976). A Theory of Human Action, Princeton: Princeton
University Press.
McGinn, C. (1996). The Character of Mind. An Introduction to the Phi-

losophy of Mind, New York: Oxford University Press.
Metzinger, T. (2003). Being No One. The Self-Model Theory of Subjec-
tivity, Cambridge, MA: MIT Press.
Morrison, K. (1996). The Self. In Guttenplan, S. (ed.), A Companion to
the Philosophy of Mind., Oxford and Malden: Blackwell Publish-
ers, 550–558.
Nagel, T. (1974). What is it like to be a bat? Philosophical Review, 83,
435–450.
Van Gulick, R., (2009). Consciousness. In Edward N. Zalta (ed.),
The Stanford Encyclopedia of Philosophy (Winter 2009 Edition),
<http://plato.stanford.edu/archives/win2009/entries/conscious
ness/>.
Self-Knowledge, Introspection and Memory1
António Marques
Self-knowledge (S-K) has a variety of dimensions and philosophers

have proposed, and continue to propose, a host of different methods to
explore or to establish the principles of a very particular form of knowl-
edge such as this. What is quite impressive in the philosophical liter-
ature (by which I mean not only that of analytical inspiration but also
that of the so-called continental philosophy) is the fact that the usual
relationship between S-K, introspection and reflection leaves no space
for the essential role of memory. Hence, it is memory’s role in S-K that
I wish to look at here.
Traditionally, empiricist and rationalist approaches have dominated
the epistemic views of S-K. The latter starts with Descartes, or even
earlier, and continues until Kant; the former includes Locke and Hume2.
1 A first draft of this paper was presented at the seminar about “Knowledge and Abil-
ity” coordinated by Franck Lihoreau. I’m grateful for his kind invitation and for the
stimulating discussion with the researchers who participated in that session.
2 This is a too simple if not erroneous scheme taking into account the multiple and
often contradictory positions that exist even inside the same field. It is enough to
stress that Descartes (or what can be designated as Cartesianism), who is point-
ed out as the classical paradigm of self-knowledge by introspection (S-KI) and
the concept of reflection that operates in it, is submitted to a devastating critique
in Kant’s first Critique (notably in the “Paralogisms of Pure Reason” of the first
Critique). I do not think there is enough evidence in Descartes’ main works to
support the picture referred to above, but I shall not follow up this discussion. The
fact is that it is in Locke that a clear concept of reflection as an essential piece of
introspection and self-knowledge is to be found. The following quotation from
Locke’s Essay Concerning Human Understanding illustrates a relatively common
concept of reflection which prevails in the literature on these themes: “The mind
receiving the ideas mentioned in the foregoing chapters from without, when it
turns its view inward upon itself, and observes its own actions about those ideas
it has, takes from thence other ideas, which are as capable to be the objects of
its contemplation as any of those it received from foreign things” (Essay …, II,
74 António Marques
In fact, if one thinks of Hume and Kant as the most relevant modern
representatives and the culmination points of both approaches to S-K,
it would seem perfectly acceptable to suggest that introspection plays
a determinant methodological role. In the first case, let us remember
what Hume famously said regarding his own mind and what he finds
there: “For my part, when I enter most intimately into what I call my-
self, I always stumble on some particular perception or other, of hot or
cold, light or shade, love or hatred, pain or pleasure. I never can catch
myself, at any time without a perception, and never can observe any
thing but the perception” (D. Hume, 1984, 300). In the case of Kant, in
the section “The paralogisms of pure reason”, because he is occupied
with the deconstruction of all dogmatic metaphysics about the ‘I’, the
introspective tool is also used: “Now in inner intuition there is nothing
permanent, for the ‘I’ is merely the consciousness of my thought. So
long, therefore, as we do not go beyond mere thinking, we are without
the necessary condition for applying the concept of substance, that is,
of a self-subsistent subject, to the self as a thinking being. And with
the objective reality of the concept of substance, the allied concept of
simplicity likewise vanishes…” (Kant, Critique of Pure Reason, B 413,
transl., N. Kemp Smith).
With these quotations from these two classical authors I only wish
to show that knowledge of the self, which should be acquired by intro-
spection, is relatively poor: a particular kind of perception in the case
of Hume, and the mere consciousness of the ‘I’ in the case of Kant. I
make this remark in order to show that skepticism about the possibility
of genuine knowledge already has, to my mind, its roots in classical phi-
losophy, be it empiricist or rationalist. When I refer here to skepticism
in relation to the knowledge of one’s own mind (or self), I mean our
capacity to know our mind by setting up a dual structure, that is, a mind
that looks at itself, a mind that sets itself up as a target. Surprisingly, if
one takes the examples of both Hume and Kant (for different reasons),
chapter VI). Objects “as capable to be the objects of its contemplation as any of
those it received from foreign things” are precisely what seem to be at the core of
any introspective move, but this is precisely what makes it so problematic.
Self-Knowledge, Introspection and Memory 75
this target seems to offer us little: a perception of feelings or awareness

of an ‘I’3.
Nonetheless, in order to approach the notion of self-knowledge,
one needs to understand as clearly as possible the link between reflec-
tion and memory, and my argument below is based on the view that
the dual structure of reflection has to be revised by the introduction of
the role of memory. Consequently, what I suggest is that clarification
of what can be considered self-knowledge and introspection requires
a consideration of the role that is played by memory and the evalua-
tion of the epistemic nature of retrodictions (that is, statements about
past events, in this case, statements about past mental events). Then my
argument, which for the moment is only designed to be a draft, goes
through the following points:
– I: Introspection is an inescapable tool of S-K and is an instance

of direct knowledge of my mental or psychological/mental states,
then S-K is in most cases S-K by introspection (S-KI).
– II: S-KI is memory dependent (there is a structural time factor in
all S-KI that is virtually unrecognized by many philosophers).
– III: if S-KI is dependent on memory and at the same time is direct
(unmediated) knowledge, it cannot be treated as an analogy with
displaced perceptual knowledge (displaced knowledge: for exam-
ple, to know that person A was in this room by seeing his hat).
– IV: S-KI is expressed in retrodictions, that is, statements about
my own past mental states (not to be confounded with statements
about myself in the past).
– V: S-K that is not S-KI does not depend on memory. Typically, it
corresponds linguistically to expressions or avowals in the pres-
ent tense which possess a different epistemic nature, namely they
3 In the same vein, Wittgenstein quoting William James makes a nice (and inten-
tionally over-simplified) picture of introspective operations: “Here we have a case
of introspection, not unlike that which gave William James the idea that the ‘self’
consisted mainly of ‘peculiar motions in the head and between the head and the
throat’. And James’ introspection showed, not the meaning of the Word ‘self’ (so
far as it means something like ‘person’. ‘human being’, ‘he himself’, ‘I myself’),
or any analysis of such a being, but the state of a philosopher’s attention when he
says the Word ‘self ’ to himself and tries to analyse its meaning. (And much could
be learned from this.)” (Wittgenstein, PI §413).
76 António Marques
are not only authoritative in the first person but are also infallible.
Typically, expressions like “I’m cold”, “I’m furious”, “I’ve the in-
tention to do this”, etc., which are expressions of mental/psycho-
logical states, require only sincerity conditions. Yet I shall suggest
that even expressions or avowals are not totally memory independ-
ent if one accepts the working memory function introduced by
some psycholinguists. But this issue is not to be discussed here.
– VI: S-KI does not have a reflexive structure, in the sense that it is
not a second-order thought targeting a first-order one.
– VII: S-KI has first-person authority with respect to the contents of
one’s own mind although it is fallible precisely because the retrod-
ictions are linguistically memory dependent.
– VIII: Statements like “I was sad in Paris last week”, which possess
first-person authority, are fallible and are pieces of S-KI unlike the
correspondent statement “I was in Paris last week”, which does
not have any authority of that kind and is consequently not S-KI.
Nevertheless, let us mention that both statements are fallible.
Below I will not deal with each of the above points separately; I only
wish to identify some transversal guidelines against the background of
the views of some of main authors. Let us begin with Fred Dretske,
whose book Naturalizing the Mind (1995) has an entire chapter entitled
“Introspection”. An intuitive definition of it would be, for example, “the
mind’s direct knowledge of itself ” (Dretske, 1997, p. 39) although he
adds that introspection is an “instance of displaced knowledge”.
Introspective knowledge is knowledge of the mind – i.e., mental facts. Since

mental facts (according to the thesis I am promoting) are representational facts,
introspective knowledge is a (conceptual) representation of a representation – of
the fact that something (else) is a representation or has a certain representational
content. It is, in this sense, meta-representational. Meta-representations are not
merely representations of representations. They are representations of them as
representations (1997, p. 43).
Since mental facts (according to the thesis I am promoting) are representational

facts, introspective knowledge is a (conceptual) representation of a representation
– of the fact that something (else) is a representation or has a certain representa-
tional content (ibid.).
Then the examples Dretske gives are analogies with representa-

tions of external objects: I can see a photo as a piece of paper, as an
object with 2 grams, as a photo of Clyde, and so on. I suggest that this
feature of introspection as representation of one’s own mental states as
something else seems quite hard to accept. Normally I do not represent
me as sad, but I am or I was or I will be simply sad. Cases in which
I represent me as sad are typically situations where I say that I represent
my face as sad (either in a picture or in a photo or in my imagination).
But these situations are not pieces of S-K.
Of course a little more introspective attention on my recalling can
raise doubts in me about my sadness and then I will ask myself whether
what I felt was a genuine feeling of sadness, about the reason I was
sad, and so on. Anyway in such cases it would be better to speak about
continuous and associated memory operations. However, these cases do
not have the double form of a first-order representation, which is set as
a target, and another second-order one, that is, a meta-representation of
that target. Even a retrodiction of some complexity such as “I was sad in
Paris because of this and that, and so on” does not have a dual structure
like a meta-representation of first-level representations. It is much more
analogous to my direct representation of myself in a mirror. The situa-
tion that just simulates the meta-representation would be: I am seeing
myself in a second mirror, which mirrors me in a first mirror. But this
does not adequately simulate S-KI: remembering my past mental state,
when I was sad, does not have any kind of intermediate representation;
it is direct knowledge (occurred at a time 2) that essentially consists of
the recalling of a past event (occurred at time 1). Recognizing this dif-
ficulty Dretske claims that:
It may seem as though this account of introspective knowledge – as a species

of displaced perception – makes it into a form of inferential, and thus indirect
knowledge. If introspective knowledge of oneself – that one represents the world
thus and so – has the same structure as knowing that the postman has arrived by
hearing the dog bark, then there is an intermediate ‘step’ in the reasoning that
makes knowledge of the target indirect. There are two ways of knowing that the
postman has arrived: by seeing or hearing him arrive and by seeing or hearing
something else (the dog) that ‘tells’ you he has arrived. If, on a representational
theory, introspective knowledge is more like the latter than the former, then the
representational account fails to give self-knowledge the immediacy that we know
it to have (Dretske, 1997, p. 60).
78 António Marques
Then Dretske adds that this is a relevant objection to the analogy

that he claims between introspection and displaced perceptual knowl-
edge, but he insists that in an introspective statement like “I represent
k as blue” the introspective quality comes (and this is the fundamental
difference between displaced introspective knowledge and displaced
perceptual knowledge) from the fact that the truth of the information of
my representation of blue is not relevant, if what I see/represent is or is
not blue; the relevant fact in that introspective statement is about I am
representing k (ibidem, p. 61). To my mind Dretske wishes to preserve
the traditional dualistic structure of introspection and self-knowledge
as reflection.
This leads us to another point about the reflexive nature of intro-
spection and S-K. Like Dretske, T. Burge, or more recently P. Jacob4,
accepts this reflexive structure as part of S-K. I have already comment-
ed on Dretske’s view of displaced knowledge, on his wish to preserve
the reflexive feature of introspection and the implicit difficulties he rec-
ognized. In the case of Burge, it is possible to find difficulties regarding
S-K as a dual reflexive structure although of another kind to those that
are to be found in Dretske. The case of Tyler Burge is quite specif-
ic in the sense that he considers that, on the one hand, at least some
S-K statements seem not to require a dual structure since the targeted
thought coincides in time and content with the subject (second-order)
thought, but on the other hand S-K involves necessarily the difference
between meta- and first-level thoughts on which is based all reflection.
Let us observe in some detail Burge’s approach to this topic.
Burge sees the differences between S-K and perceptual knowledge
in the personal status of S-K. Namely in S-K acts, one is facing state-
ments that are made simultaneously from and about one first-person
point of view. He identifies this feature in the following terms: “the
point of view and time of the judgment must be the same as that of
the thought being judged to occur” (Burge, 1988, p. 651). And he adds
“When I judge: I am thinking that writing requires concentration, the
time of the judgment and that of the thought being judged are the same”
4 For a detailed discussion of Dretske’s model of displaced knowledge applied to S-K,

see P. Jacob, 2010. The theoretical framework of the discussion is Jacob’s wish to
show that “our picture of introspective self-knowledge derives from two traditional
sources: rationalist epistemology and empiricist epistemology” (ibid., p. 3).
(ibid.). This also means, as Burge makes clear in another passage of

the same paper, that “The reflexive judgment simply inherits the con-
tent of the first-order thought” (ibid., p. 650). The essential difference in
relation to perceptual knowledge consists of the personal status of S-K,
or in other words the use of first-person pronouns, and as he notes: “In
all cases of authoritative self-knowledge, even in those cases which are
not ‘basic’ in our sense, it is clear that their first-person character is fun-
damental to their epistemic status” (ibid. p. 651). Thus it is first-person
authority and not so much its reflexive structure that confers a special
epistemic status on S-K if compared with perception. Or perhaps in
other words, the correct interpretation of Burge’s view is that it is the
reflexive structure of S-K plus the first-person authority that gives it its
peculiar epistemic nature. Yet if one accepts Burge’s description of S-K,
when one compares the epistemic nature of S-K and that of perceptual
knowledge, one feature of the former (the thought I have has the same
content and coincides in time with what is thought about) does not place
both knowledge types far from each other. After all, in reflexive knowl-
edge (as all S-KI should be) more than one content is not to be found.
This means that the first-level (targeted) thought does not have a content
on its own5 5. What makes the difference is the place of first-person au-
thority in S-K. Yet does this difference produce drastic epistemic conse-
quences? In Burge’s view it does, but I think it is a mistake: first-person
authority does not imply infallibility in the process of introspection. Let
us make the nature of this process clearer.
5 This is a remarkable result in order to re-design the traditional concept of reflec-
tion and introspection. It is interesting to consider another feature of this redesign
since in S-K each targeted event can be seen as a “recognizing” of what happens or
of what is happening in me. Wittgenstein’s approach to the nature of recognition is
striking inasmuch as it describes the misunderstandings surrounding recognition
when memory is taken into account. The result is also “fusion” into one content,
as is the case in Burge’s approach. “It is easy to misconceive what is called “rec-
ognizing”; as if recognizing always consisted in comparing two impressions with
one another. It is as if I carried a picture of an object with me and used it to iden-
tify an object as the one represented by the picture. Our memory seems to us to
be the agent of such a comparison, by preserving a picture of what has been seen
before, or by allowing us to look into the past (as if down a spyglass)” (PI §604).
Yet this is a mistaken picture, “Indeed, it is not so much as if I were comparing the
object with a picture set beside it, but as if the object coincided with the picture.
So I see only one thing, not two” (PI §605).
80 António Marques
Using such criteria, it seems hard to maintain that S-K is clearly

distinct from the usual perceptual knowledge since, after all, in both
there is one content and there is no time gap that I can be aware of.
My concern lies not so much in the claim for the existence of a unique
content in the case of reflexive knowledge (“reflexive judgment sim-
ply inherits the content of the first-order thought”), with which I
agree, but much more with the claim that in S-K there is a typical
reflexive structure with a dual composition based on a targeted object
by a second-level thought/representation. As I have tried to show, the
reflexive operation does exist in all S-K yet as an introspective move
which is memory dependent6 S-K is S-KI with a dual time structure
where memory is necessarily involved. By dual time structure I mean
the awareness that a past t1 event is targeted by an actual thought/
representation typically expressed in retrodictions at t2. This is exact-
ly why Burge does not see in basic S-K a necessary time gap, which
requires the work of memory. As I said above, he sees S-K as pos-
sessing the classical dual structure plus first-person authority and it
is this characteristic alone that gives S-K a special epistemic nature. I
would say that he means something near to infallibility. The following
lines illustrate, in my opinion, this conviction: “The source of our
strong epistemic right, our justification, in our basic self-knowledge
is not that we know a lot about each thought we know we have. It is
that we can explicate its nature and its enabling conditions. It is that
we are in the position of thinking these thoughts in the second-order,
self-verifying way” (Burge, 1988, p. 654).
But this is already problematic for Davidson to whom “of course
people have beliefs, wishes, doubts, and so forth; but to allow this is
not to suggest that beliefs, wishes, and doubts are entities in or before
the mind, or that being in such states requires there to be correspond-
ing mental objects” (D. Davidson, 2001, pp. 35–36). So if there are no
mental states as objects in the mind that one could grasp in any sense,
there is no such a thing as introspection in the usual sense for Davidson
6 The fact that in S-K there is one content (first-order thoughts do not have any con-
tent on their own) does not imply that the targeted content that occurred in the past
coincides in time with the actual statement/thought. On the contrary, it is essential
to S-K that the awareness of a time gap (even the narrowest) is preserved. Yet if I
am not wrong, Burge’s position erases any time gap whatsoever.
I guess. But Davidson’s point is that such a thing is not needed in order
to have an attitude. In Davidson’s words, “having an attitude is not hav-
ing an entity before the mind; for compelling psychological and epis-
temological reasons we should deny that there are objects of the mind”
(ibidem, p. 36). I believe Davidson has a fundamental point here, but I
think he could go deeper in the consideration of these “compelling psy-
chological and epistemological reasons”. It is at this moment that the
problem we face can be formulated in the following terms: it is counter-
intuitive to simply deny the existence of introspection and the corre-
spondent S-K, but one should not count on dual reflexive structures,
on target objects in the mind and similar sorts of objects to identify
introspective knowledge. It is at this point that it seems to me necessary
to introduce the time factor, or in other words memory. S-KI needs a
time distance, even a minimal one: the statement about my own mental/
psychological state made at t2 always refers to a t1 event. There is, I
agree, a dual structure, but it is a present/past dual one not a reflexive
one. Even I “reflect” at t2, for example, as a result of mental states that
I felt at t1. There is always a dual time structure that still requires at
least two things that are usually required for S-KI, which are 1) to be
direct, that is, not mediated like any displaced knowledge, and 2) to
have first-person authority.
Recent studies on memory and especially short-term memory elu-
cidate us not only about the complex neuro-chemical processes involved
and what distinguishes this kind of memory from long-term memory.
They also allow us to understand the role that short-term memory plays
as it works when representations/thoughts coincide (in time) with their
correspondent content ones or when we report something to other peo-
ple or to ourselves. Psychologists call this very short-term memory
“working memory” and it can be defined in the following terms: “Work-
ing memory can be thought of as a low capacity information reservoir
that is always full, sensations flowing into it continuously at about the
same rate that they are forgotten […]. Working memory is an indispen-
sable form of transient memory; it is a moving window of comprehen-
sion that allows us to understand the present in terms of the very recent
past” (O’Shea, M., 2005, p. 85). This is the case whenever the grammat-
ical form is equivalent to present tense statements. It is enough to think
about statements like “I’m understanding this text”, “I’m telling you
82 António Marques
how much I am sad”, “I’m seeing this sunset as a Rothko picture”, etc.
These are statements that do not seem to have the same structure as sim-
ple avowals or expressions. Precisely they involve a working memory
that eanables cognitive performances and are present in all language
learning7.
This way of seeing S-KI leads us to the question of the status of
the content of retrodictions, which are memory dependent statements
and are different from retrodictions that are not memory dependent, or
rather that have a truth-value that does not depend on my memory: the
statement that I was in Paris last week is a retrodiction of this type. It
is not a piece of introspection, and even if it possesses the dual time
structure, it lacks the other essential component, that is, the authority
of first-person8.
My past sadness is not a target of a meta-representation but sim-
ply the content of a thought/ representation that in this case is a recall
of a past state of mind, experience or whatever. What happens is this:
the so-called first level content, i.e. the event that at time1 I was sad, is
7 For many psycholinguists working memory is implicated in syntactic compre-

hension or in keeping track of syntactic dependences. Furthermore, the working
memory is what enables cognitive performances such as attention being focused
on targets at a particular time. For example, the judgment, “Now I’m understand-
ing this argument” involves working memory, and even implicit elements learned
in the past that enable comprehension are, so to say, linguistically hidden. If we
consider that judgment is a piece of S -K, then the targeted element is memory
dependent and the statement itself possesses a dual time structure.
See C. Philip Beaman, “Working Memory: Beyond Language and Sym-
bolism”, Current Anthropology, Vol. 51, June 2010, pp. S27–S38, and Conway,
A. R. A., M. Kane, M. Bunting, Z. Hambrick, O. Wilhelm, and R. W. Engle, 2005.
“Working memory span tasks: a methodological review and user’s guide”, Psy-
chonomic Bulletin and Review 12, pp. 769–786.
8 First-person authority must fulfil the following condition: a total asymmetry in
relation to the truth/satisfaction/conditions of the correspondent third-person
statement’s content. Then “I (A.M.) was in Paris last week” has the same truth/
satisfaction condition of “He (A.M.) was in Paris last week”.
They are not asymmetric. But “I (A.M.) was sad last week in Paris” has not
the same truth/satisfaction condition of “He (A.M.) was sad last week in Paris”.
So “I was sad last week in Paris” has a first-person authority which is lacking in
the first statement “I was in Paris last week”.
recalled at time2. Is the t2 representation a meta-representation of the

t1 representation?
I would say that it is not and I shall illustrate this claim with a re-
mark of Wittgenstein’s from the Philosophical Investigations.
“Does it make sense to ask ‘How do you know that you believe
that?’ – and is the answer: ‘I find it out by introspection’? In some
cases it will be possible to say some such thing, in most not. It makes
sense to ask, ‘Do you really love her, or am I only fooling myself ?’,
and the process of introspection is the calling up of memories, of im-
agined possible situations, and of the feelings that one would have
if…” (PI §587).
The remarkable thing in this quotation is that for Wittgenstein in-
trospection, so to speak, achieves its target in most cases by an opera-
tion of recalling, of memory. In other words, in most cases the structure
of introspection requires a time factor, and without it there is no target
whatsoever. (In this case, the target element is the quality of a feeling.)
On the other hand, when memory does not play a role, it seems that
Wittgenstein would reject introspection. Without memory, no target, no
introspection. Possibly the cases where Wittgenstein would say that it
does not make sense to speak of introspection are such as: believing p
and observing my belief seems to be absurd. Let us note that even some
statements like “Now I’m believing p” implicate that I perhaps did not
believe p before. But this contrast only needs the time factor. In fact,
what is more plausible according to the quotation above is that Witt-
genstein considered that in some cases where time did not play a role
there was no sense in the question. Is memory dependent inquiry truly
self-knowledge and are other forms, which exclude memory, pseudo
self-knowledge? Wittgenstein does not make this explicit, but one can
suspect that it is what he means.
Thus, as a non-definitive conclusion, S-KI has at least 3 features: a
dual time structure based on memory, it must be direct and it must have
first-person authority.
84 António Marques
References
Burge, T. (1988). Individualism and Self-Knowledge. Journal of Phi-

losophy, 85, 649–663.
Burge, T. (2007). Foundations of the Mind, Oxford: Clarendon Press.
Davidson, D. (2001). Subjective, Intersubjective, Objective (“Knowing
One’s Own Mind”), Oxford: Clarendon Press.
Dretske, F. (1995). Naturalizing the Mind, Cambridge, Mass.: MIT
Press.
Hume, D. (1984). A Treatise of Human Nature, London: Penguin.
Jacob, P. (2010). Do we know how we know our own minds yet?
<http://jeannicod.ccsd.cnrs.fr/docs/00/05/34/50/HTML/index.
html>.
Kant, I. (1992). Critique of Pure Reason, trans. N. Kemp Smith, Lon-
don: Macmillan.
Locke, J. (1690). Essay Concerning Human Understanding.
O’Shea, M. (2005). The Brain, Oxford: Oxford University Press.
Wittgenstein, L. (2009). Philosophical Investigations, (revised 4th edi-
tion), Oxford: Wiley Blackwell (quotations from PI are made in the
usual form referring to the number of the section in the case of the
first part of the work).
Imagination as a Bodily Pattern: thinking about
Sartrean’s account of Consciousness
Clara Morando
Sartre’s study called L’Imaginaire. Psychologie Phénoménologique de

L’Imagination (1940) is a follow-up to another book named L’Imag-
ination (1936), where we can find a) a clarifying list of the imagina-
tive consciousness’s distinctive traits, b) an extensive characterization
of various kinds of images (eidetic, hypnagogic), c) long-detailed de-
scriptions about certain psychopathologies, and finally, d) a perspicuous
categorization of the remaining consciousness modes as perceiving and
conceptualising. Concerning what draws the specificity of imagination,
this book adds to a historical view a phenomenological deepening into
a peculiar type of intentionality, ascertained by an eidetic treatment.
The first part of the book, called “Le Certain”, deals with the im-
age’s intentional structure where Sartre proposes a functional meta-
psychology claiming that when consciousness works based on an
irrealizing scheme connected to non-existent or absent objects, we are
then allowed to call it imagination. This particular way of intending
things goes on by building up, step by step, the imaginary consistence
whose very function is to establish another strand of knowledge consti-
tuted by several noetic acts which have in common the imaging move-
ment as a different type of intentionality beyond perceptive knowledge.
In order to make the image’s description as an image possible, it is
necessary to embrace a second-order-thought, just because it is imperi-
ous to understand how the very same object, the imaged object, occurs
in our consciousness. Although the Cartesian view considers that only by
a conscious act (an intentional act) can imagination be cleared, we have
to recognise that Sartre gives us more than we can simply find in Carte-
sianism as a philosophical doctrine, historically rooted on the distinction
between mind and body and on the assumption that consciousness is a
86 Clara Morando
mind’s indispensable attribute. It is true that only by a reflective act is it

possible to achieve a fuller comprehension of what imaging means. How-
ever, we cannot forget that Sartre does not say that that same act is incor-
rigible, i.e., that all we get by reflection is absolutely clear and distinct.
The French author also never appeals to an obscure anthropological
dualism, believing contrarily that the body is a psychic substance, and
the psychic is bodily instantiated. What we get here is that the imaging
activity can only be a special corporeal operation, inserted in a vaster
field called consciousness, although is not just that simple.
He tears apart imaginative consciousness from all other forms of
intentionality, formulating suggestive hypothesis about its very nature
and putting in practice the Husserlian phenomenological method. How-
ever, he does not follow it strictly and refuses to suspend the existence
of the natural world. He decomposes the images to what matters in
terms of their distinctive traits, making a sort of reduction but without
the respective phenomenological epoché that he considers to be a mir-
acle we cannot understand very well. Thus, we have now an open gate
to assert that every image is a form of consciousness, although this
particular consciousness differs from merely perceptive consciousness.
We can also say now that consciousness has nothing to do with a closed
and sterile immanence, incapable of reaching the world.
Gregory McCulloch stresses that “consciousness is nothing else
than a relationship”, and this applies also to a special form of con-
sciousness – the imagination.
When I have a perception it means that I have a certain (pre-re-
flexive) consciousness of it, and, for instance, a perceived chair can
be viewed as the object of that particular movement of consciousness.
What Sartre greatly ensures goes on to say that when we have an im-
age of the very same chair, what specifically happens is that we are
allowed to believe in a sort of identity between the chair as the object
of perception and the chair as the object of imaging. According to
this, consciousness relates itself with the referred chair in two differ-
ent ways, and in both ways of dealing with the object ‘chair’ we see as
constant the directive movement to its corporeity. In the specific case
of imagination, Sartre adds that consciousness establishes a kind of
synthetically organized intentional movement directed to the existing
chair.
Imagination as a Bodily Pattern 87
We can say that the image’s object does not simply correspond to
a mere image of it in terms of what we consider a reproducing copy
of the real object. In his analysis he vehemently insists that “imaging
experiences do not have images as intentional objects”, they have the
very objects themselves. So, an image, as Sartre says, is first of all a
relation entailed between the conscious subject and its object, which
results after conceiving an image through the so-called consciousness’s
intentional movement. The image is the very intentional movement.
We must have present the fact that perceiving and conceptualising,
with imagining, represent three types of complex structures linked to
conscious activity. There are, thus, three possible ways of being con-
scious of the same object.
a) In perception, I am really facing the object in its very existence,
although I am not capable of perceiving it in its entirety, i.e., in all of
its material richness. I am obliged (structurally) to make, through my
perceptive consciousness, an always incomplete adding operation re-
lated to the available object’s data. This means that there is a kind of
perceptive resistance in the objects that opens to the possibility of a suc-
cessive annihilation onto their existing characteristics, and the only way
we have to partially avoid it (the annihilating process) is to successively
produce a series of profiles about the same object. We establish, then, a
crucial synthesis that surrounds objects and shows most of all (or only
some of) their traits or properties.
The idea of the perceived object is represented in a whole-think-
ing of it, and presupposes a sort of concrete knowledge, firstly without
words and images and attached corporeally to what is perceived. This
happens also in imaging activity, which appears to locate in the very
same side of perception, marked in its constitution by a synthesis of
multiple appearances. The only but great difference among them relies
on the fact that if perception consists of a slow-building knowledge as
Sartre points out, imaging is, on the other hand, an immediate form of
knowing that. The latter is then directly connected with the object by an
intriguing and instantaneous way that stands for us like a sort of enig-
ma. In perception, contrarily, nothing appears without having a complex
branch of connections with other objects that gives us the special feel-
ing that when we perceive things we just only get on them half of the
picture from them.
88 Clara Morando
On image we just find an essential poverty in terms of what we

call the synthetical object’s richness. That specific way of intending/
understanding an object leads us to a particular kind of isolation from it
among all other objects, imaged or not. The different elements that com-
pose an image and the way they connect each other are not sufficient to
establish an internal coherence that could rival the one we encounter on
perceived objects.
As we said, the worldly existing objects and the impossibility of
knowing them in their entirety constitute a strong proof of the infinity
of the perceptive process. Consciousness is always inferior to the real
world’s richness because it is incapable to get the complexity under-
stood. The perception of the real things is really menaced by the incom-
pleteness signal.
b) In imaging, on the other hand, we see a strict correspondence
between consciousness and the very image produced by its intentional
movement. On image we have nothing to learn, because the object’s
imaging intuition seems to happen very rapidly, like lightning, and is
not also compelled, by its very nature, to respect physical laws (or just
some of them). It is reasonable to assert now that only by imaging can
we be totally sure of what we are imaging, just because here we do not
need to know if there is a proper adequation between existing objects
and imaged objects. In Sartre’s philosophy of imagination we are never
allowed to say that images represent mere copies of the existing ob-
jects. Images are, instead, just an intentional movement that envisages
non-existing objects (absent, unreal) showing us, on the other hand, that
they have a certain scale of reality, but now another type of reality, with
different powers, and not less intense.
To clarify this, the French author suggests that all contents pertain-
ing to an image point out, in a delusive way, to the empirical world. In-
deed, on that specific pointing out to a presumable material plan is found
to be an almost-collapsing “irrational quality”, simply because in its
movement towards a non-existing object, imagination does not commit
itself to strictly respect as a whole the physical laws engaged in percep-
tion. It remains, in all the image’s contents, a sort of opacity in relation
to their structural meaning in a latter sense, due not only to the fact that
they are not obliged to follow physical laws, but also to the fact that in
the very intentional moment directed to a particular non-existing object
we are confronted with an emerging power of consciousness entirely un-

suspected. We do not find here regulating essences nor generating rules.
We simply find a kind of assessment of the image that is greatly linked
to the feeling of certainty we encounter in perception. That is why we
are authorized to assume that images are objects of quasi-observation,
and we know now this is not because they copy real objects, but because
they produce the irreflective belief of a peculiar existence, although we
rationally maintain their effective non-existence. This powerful belief,
incoherent and outstanding in a minimal first stage, is the very core of
imagination, equivalent to intuition. That is why it is impossible to see
on images a décalage between objects and consciousness, representing
only the result of an intense intuition. We can also say that they (images
and consciousness) are just the same thing.
It follows then that there is another characteristic one can focus on
the image’s status, and it consists exactly of that the imaging conscious-
ness posits its objects all-embedded in nothingness. Or, more rigorously,
it grasps them in a progressive movement towards nothingness. We have
seen that the precise instant in which we almost believe in the image’s em-
pirical existence is a sort of paradoxical step in relation to what it follows.
What comes after is then precisely the stage where we assure the image is
a simple image. But is also on that very first step (and we cannot forget it)
that relies the imagination’s intriguing and powerful trait.
We cannot get erroneously drifted by this descriptive and critical
set of statements to the very assumption that in the irreflective level
of consciousness we are dealing with a certain type of unconscious.
Nothing can be more wrong than this. Even in the irreflective level,
that Sartre prefers to call pre-reflective, consciousness possesses of it-
self merely an immanent consciousness or a non-thetically rooted con-
sciousness. This means that consciousness in its lower-level is simply
consciousness of ‘the’ thing; for instance, in the case of perception,
it is consciousness of the ‘chair’. More rigorously, it means percep-
tive consciousness of the existing ‘chair’. Just in the upper-level, the
second-order-thought-level (the reflexive level), is consciousness able
to posit itself as consciousness of an existing ‘chair’. So, consciousness,
in its various categories, is omnipresent.
What happens then in the specific case of imagination? How is it
possible for the imaging consciousness that originally pertains solely to
90 Clara Morando
the pre-reflective level of consciousness to posit itself as an image? How

does that specific type of consciousness appear to itself as an image if
being an image would initially presuppose a certain mode of reflexiv-
ity? This last question seems only to make sense in terms of what we
understand as the traditional image theory. Sartre objects to it by saying
that the image is not a sort of miniature in our minds in relation to ob-
jects (he calls this view ‘the illusion of immanence’), but is in the very
image’s producing act where there occurs a special way of constituting
a particular object as an image in the wide world of things. Thus, is
precisely the transcendent consciousness connected with a determined
thing (or object), turned into an image, that posits the very same object
as an image, which essentially seems to be out of our consciousness.
Is uniquely the astonishing intuition of transcendence that is conveyed
to all imaged objects, belonging now apparently to the world of things,
which leads to the compelling impulse of consciousness to hunt them as
if they were exterior to their own producing act of consciousness.
This explains why the positional act taken by all imaging con-
sciousnesses in relation to a specific object involves necessarily (and
specially) not only a) a character of negation by positing the object in
the plan of quasi-observation, as we above said, but involves also b)
a kind of “positive element” in that very act of positing. This “pos-
itive element” is paradoxically ascertained by a strange and intuitive
feeling about what we could name the always-running evanescence of
the image. We have now clearly understood that images are themselves
objects among other objects, including the perceived objects, and that
an image can be defined by its constant negation as “nothingness of be-
ing”, although it starts through a belief in a sort of intensity of its proper
existence. However, this primary belief is not what a superficial critique
would call an unconscious belief, functioning merely as an obscure in-
centive for the rest of the process, but represents abstractly the belief in
the image as a pure and simple object and nothing else.
When I say that I have Pierre’s image I also want to say that I do
not see anything at all. But it does not mean that I am not envisaging
Pierre in an intuitive way just because I am not seeing him. Sartre says
then that what my actual intention intends to grasp is precisely Pierre
in his corporeity, “that Pierre’s image is a way of not touching him,
a way that he has of not being”. So: Pierre’s image’s main trait is to
be a kind of “intuitive-absent” object of consciousness, more precisely,

an intuited given absent to intuition. How can we have an intuition of
nothingness? It is precisely this immediate consciousness of a particu-
lar type of nothingness which constitutes that specific corporeity I was
pointing to before.
Another characteristic figured in imaginative consciousness in its
peculiar movement towards emptiness and whose proper mode is to be
at the same time and in-itself a non-thetically consciousness involving
the object is, precisely, spontaneity. This spontaneity or, more properly,
this intriguing spontaneous emergence described as a specific kind of
intentionality that progressively denies the existence of a certain object,
does not have in its core, as a primary goal, the urgency to be reported
to a specific object, i.e., to this or that object. What happens is just that
intentional movement represents first of all a type of consciousness that
is simply transversal. Sartre uses a metaphor saying that what imagina-
tive consciousness specifically does is to arrange a sort of crepuscular
light glued to itself in order to spontaneously produce and maintain the
object as an object-in-image, as a nothing-object. Consciousness seems
then to create several objects-in-image but abjures simultaneously its
creative power on them. This tells us that consciousness simply takes
off through negation of the positive element that would affirm the mod-
elling gift of existence. It substitutes then this particular gift by another
type of positivity, a paradoxical positivity, essentially linked to the act of
empirical negation whose intensity feels sometimes to be so great that
can even rival that of objects constituted by perceptive consciousness.
To a descriptive static of imaginative consciousness there follows
a dynamic way of talking about it, just because it is not possible to phe-
nomenologically study consciousness without seeing it as a temporal
synthesis organized between other forms of consciousness. Those other
forms of consciousness come after or before imaginative conscious-
ness, composing among all of them a continuum that metaphorically
evocates a melody.
The pattern followed by this particular type of consciousness seems
to have in common with perceptive consciousness the fact that also the
first one looks at its object, analogically, through perceptive territory,
having special care when determining the sensible elements constitut-
ing it. McCulloch stresses that Care is in order here, since there are, for
92 Clara Morando
instance, two ways in which one might image ‘a chair’. The difference
here is that whether at perceptive processes we just find a passive ac-
ceptance taken by consciousness in relation to sensible objects, simply
representing them as they are (although this is always a partial rep-
resentation), at imaging processes consciousness adopts towards things
on their representations a dynamic posture, over-flowed by an unstop-
pable creative will.
Hypnagogic Images
This astonishing creative will can be so dramatic that it produces some-

times a sort of pathological feeling of over-reality in certain images.
These images seem nervously to be more real, more concrete and more
palpable, than empirical objects. Once they are hypnotising objects they
also seem to have great powers on other things, including us, fulfilling
a special kind of hierarchical superiority that totally submits their very
creator subject. Therefore, these special images constitute a transcend-
ent menace over us. Sartre calls these hypnagogic images menacing
visions saying also that our consciousness posits them dangerously
among all other objects, images or not, simply because when we per-
ceive things, our consciousness does not primarily ascertain objects into
the perception level. We just put the representation as simply being a
representation. To perceive a thing is solely to put it in the middle of
all other things, not having the concern to know if they are real or not.
Although this is done with a specific concern of internal coherence,
what we only achieve is a particular coherence that cannot be said to be
empirical. So hypnagogic images rely on a positive belief about the like-
ness of their very structure, engaging themselves also with the quasi-
observation phenomena; however they are not obliged subject to the
exactly same requirements than in perceptive objects or plausible
images (according to the world of perception).
It is true that we can talk about an internal coherence on hypnagog-
ic images; however, that does not imply a sense of full respect concern-
ing physical laws. As we said before, for Sartre those images contain
a fantastic character, caressing concise frontiers on their forms, ghost-

ly-like. The strict law of individuation is not applicable to them, neither
are to other laws of perception like, for instance, the “perspective law”.
Sartre says that there is something inexplicable in them, although they
reflect only contents of the specific knowledge we possess. What hap-
pens is that in hypnagogic images, the subject abstains from making
judgements that would rectify, organize and stabilize the image con-
tents. We just do that in perception, and to do that we need more or less
some distance in relation to the perceived object. It is true that in plau-
sible images sphere (which are much likened with sensible objects) that
specific judging distance is smaller, but that does not mean that it fails
to appear. What we see now is that in hypnagogic images the essential
décalage is inexistent and consciousness is then able to feel some of the
elements that compose its knowledge, multiple-versions-combined, in a
way so intense that they seem to constitute sensible evidences.
The appearing of the object and the certainty of its existence make
just one. And we are not talking about a slow movement of conscious-
ness, as occurs in perception, but about a dazzling and “fatal” appearance
resulting from an instantaneous sum of several past consciousnesses
that form between them a temporal unity of a longer-persisting type
of consciousness – the consciousness of hypnagogic images. What in-
trigues us here is not the process of summing up through multiple and
finished consciousnesses. What perhaps constitutes motive for reason-
able pre-occupation is precisely the arbitrary and devastating way in
which this particular form of consciousness passes through all over the
rest. It’s like a big hole in the middle of perceptive and imaging plausi-
ble representations, using strangely their own materials to compose in
an unpredictable way new things, completely bizarre or even monstrous.
Ontologically speaking, it seems that nothingness opens a breach
through our consciousness but disguises itself in a total contradicto-
ry appearance – as an all-powerful being, in this case, the hypnagogic
image. It is not its content that relates to nothingness, it is precisely
the movement of consciousness producing it that relates to nothingness
through an interior process of negation, though very intense.
Consciousness here is captured by hypnagogic images (it is equiv-
alent to the hypnagogic image), and in this very equivalence we can
say that consciousness is captured by itself. Consciousness is then a
94 Clara Morando
nothingness that produces paradoxically several forms of nothingness.

The producer of judgements about things, the I, taken as transcendent
to consciousness no longer means a transcendence, an indispensable
fiction created by consciousness in order to be reflexive, and is now
plunged in the territory of the transcendental consciousness, i.e., the
only form of consciousness.
Aron Gurwitsch describes it as non-egological or completely
anonymous. The I matches then with the pre-reflective level of con-
sciousness, and this consists of a longer-type of consciousness, as we
explained above, simply because consciousness is fascinated by the pre-
sumable contents of that specific image. It is ultimately fascinated by
its own movement as consciousness in a sort of self-suggested paralysis
in-itself, not being able to create properly the (transcendent) I as the
equivalent to the reflective moment of consciousness. We stay fasci-
nated by the hypnagogic image in a sort of physical and mental trap.
Other psychologists have studied these very phenomena and they are
explicitly referenced in Sartre’s work.
The object has then no longer any type of exteriority in face of
consciousness and its presence coincides totally with the surface of
transcendental consciousness. Consciousness “superficiality” is due to
the fact that it is, as Sartre says in BN, pure wind empty of contents – a
mere translucent movement of intentionality towards objects with no in-
habitants within. Thus, what happens when we have hypnagogic images
is that consciousness, as stated before, does not possess a contemplative
power in itself – does not find a way to keep at distance the influence
of those very images. Consciousness operates then, “miraculously” or
instantaneously, a bizarre synthesis over accumulated materials (mem-
ories, feelings and thoughts) that functions similarly to an absurd cat-
egorical imperative. All of this is done with a “sense of despair” taken
by consciousness about its impossibility to guarantee a decent infer-
ential unity. The only way seen to keep something likened to reality
is to confer an inedited meaning to the hypnagogic image in order to
make clear that there is, maybe, some reasonability, although those im-
ages disrespect several laws of physics. When my body is paralysed and
my thoughts are fascinated by the hypnagogic image, then there ap-
pears, in Sartre’s words, a new structure of consciousness: the captured
consciousness. And this happens often in moments that appear before,
for example, sleep (in a light form), or that happen in schizophrenia and
other mental disturbances.
Human consciousness is left with no alternatives in ontological
terms; it no longer has the power of negation to deny what is intended,
i.e., the possibility of the very annihilating possible becomes unreacha-
ble. The self as a distance born through the creative power of conscious-
ness as a strategy has to avoid being hypnotized by objects (whatever
they are: images, concepts, perceptions, memories) ceases completely
its efficacy.
Moran (Introduction to Phenomenology) accuses Sartre of being
all-Cartesian due to his constant refusal about the possibility of the un-
conscious. Consciousness is always conscious of itself, including in its
pre-reflective level, even if it is there only consciousness of its objects.
It seems that we can sometimes be, or better yet to say that we are most
of the time, caressing beings of reflective awareness and that essential
caress must be, in a certain way, or if we prefer, in pathological cases, a
recurrent state or a permanent state.
Although Sartre’s book departs from Husserl’s theory of imagina-
tion, relating the latter with the perception sphere and recognising in it
inferiority in face of perceptive processes, the French author rethinks it
and agrees to say that imagination is an independent type of conscious-
ness that cannot be reduced to perception. As Dermot Moran stresses,
and appealing to Sartre’s words, specific activity of consciousness –
the imagination activity – generates sui generis objects. In order to get
through this Sartrean statement Moran recommends that we are more
attentive about what is meant with the puzzling activity of imagination.
He focus his reasoning on what is said about the very act of imagin-
ing (which is also imaging). More specifically he states that the act of
imagining is a magical act, i.e., is a sort of enchantment intending to
produce the object of thinking that each one of us has, and that imag-
ination process must be exactly like this, because we need, when we
imagine, to literally possess the imagined object. We are the imagined
object. I would add to Moran’s observation to the fact that this magical
act depends upon an “immediate awareness of its nothingness”. That is
why imagined objects are never totally coincident with perceived ob-
jects. That is why we also can talk of a transversal intentionality that
goes through images and through all objects of imagination. And that
96 Clara Morando
is why it is possible for us to be compossible with those very kinds of

images – through that nothingness which is paradoxically a positivity
as we have seen.
In the case of images, what happens is that when our conscious-
ness produces them it does not do so in a way that we could totally
believe to be disconnected from a material sense. Moran adds that the
particular operation of recalling the material content of an image in our
minds involves a certain type of corporeal presentation, although this
presentation is not really or actually present. When we imagine we do
it in a corporeal way, without any kind of platonic separation between
our bodies and our minds, without any artificial distance within our
all-beings. Sartre stresses in L’Imaginaire that when I think about Pe-
ter I think about him corporally, although I cannot see, touch or hear
him. And I also know at the precise moment I recall him that he is not
materially present.
We are then allowed to believe that the body engraves on imagi-
nation its marks. This unavoidable fact is visible, for instance, in the
case when I see an arbitrary bunch of lines and I find there a face.
This happens often when we look at clouds, as everybody knows. The
referred example gives me reasons to think that when we imagine we
do it under powerful influence of our bodies, i.e., that we follow strict-
ly on images-building a sort of generality concerned with corporeal
movements. In this sense, another example given by Sartre also shows
the very ocular movement as the provider of rules to find (to imagine)
in chaotic or meaningless drawings any kind of familiar knowledge,
whatever it is and whatever it costs. This explains why we have to hunt
(firstly) for the object-in-image in the world, as said before, knowing
now that that imaginative hunt must obey the requisites of an embodied
consciousness. Our judgement about what is seen is determined by the
physical apparatus of our eyes, and this is a judgement that is not merely
representative but means essentially a corporeal way of sensing objects
traditionally referred to the internal arena of our minds.
As Merleau-Ponty recognises, in his Phénoménologie de la Per-
ception, Sartre sees that the body’s-spatiality implies that all things are
observed in a relational of the build-over human’s physical disposi-
tions, and everything what we deal with is precisely accounted for in
the interior of that same net. Sartre calls this the hodological space. In
a radical version it also includes the products of imagination, and it is

only in this sense that we can say that imagination has something to do
with perception.
The problem here is that it is not possible to have any point of view
about our bodies without making at the same time an infinite regression.
Sartre underlines an absolute incapacity of the body to be simultane-
ously self-known and transcendent, because the very body is always the
instrument that cannot be used by another instrument, i.e., the particular
point of view about which I cannot have another point of view. So: all of
the stated arguments lead to a conclusion establishing the pre-reflective
level of consciousness (spontaneous, a first-order-thought awareness)
as being no longer the body’s self-consciousness, although the body be-
longs permanently to the non-thetic consciousness structures board. If
we admit imagination as a specific form of consciousness belonging in
a primary order to irreflective consciousness, then we have to ask how is
imagination not a way of escaping from the world, but an essential way
of living more sensible in it.
In Being and Nothingness Sartre makes explicit that if is true
that our consciousness is not able to recognise at the same time its
body’s-being without making of that specific assumption a transcend-
ence, it is no less true that in all irreflective moments of consciousness
(which always and necessarily occur when we are in a reflexive state)
the usage of the verb to exist should be done in a transitive mode, and
that consciousness exists in its body. The type of relation established
here is an objective relation between the body as a point of view and
things, whether between consciousness and its body we can find dif-
ferently an existential relation. What we cannot forget at this point is
that they are simultaneous (the two types of relation – objective and
existential), so, knowing about the Sartrean-Cartesian consciousness
Omni-presence commitment, we also conclude for this double-relation
in images sphere.
The body is a conscious structure of my own consciousness, and
body’s-consciousness means, more particularly, a lateral and retro-
spective way of sensing the world. We face here a strong equivalence
between body and consciousness. The mind is its own sensations, but
it never gets totally identified with them, simply because in that very
movement of self-identification with a corporeal object – the body,
98 Clara Morando
consciousness needs to open a distance within itself in order to be

sophisticatedly aware of it, of what it is. Thus, to have consciousness
of our own body we need to face it as if it were another person’s body.
We have to reflexively deal with it just inserted in an objective relation
which means precisely that we never reach it reflexively through an
existential relation. But the latter exists and the only way we have to
prove it is to assume indirectly its effects severely induced in corpo-
reality. Kathleen Wider states that “Sartre’s development of this bodi-
ly self-presence to oneself as subject as well as object is minimal and
inadequate. This is because of his adherence to a division between the
lived body and the body as an object in space, in the world, open to
the look of another”.
As Phyllis Sutton Morris says, we are unable to introspect the sub-
ject, objecting strongly to Russell’s claim that experiencing consists of a
relation, that we find there an experienced object and who goes through
the experience – the subject. The body is only an implicit term in the
analysis of our own conscious experience. However, as Sartre claims,
the same body is a centre through which objects are intentioned, and
in this very claim we cannot put the body as an additional object of
experience. The sensation is a hybrid notion between the subjective and
the objective, conceived as an exterior object but then applied to the
subject. The Being-for-itself has to be all body and all consciousness; it
cannot be united to a body.
This is very familiar with what we have seen about images. We
could add here the example given by Sartre when he talks about sche-
matic drawings. He states that what is interesting about the scheme is
the very fact of being an intermediary between the image and the sign.
However we can place it on the side of an image. The intention as initially
perceptive becomes imaged, but for that it is necessary for the subject to
interpret the scheme in a way that allows consciousness to create. On that
very creation it is necessary for the body to adopt a certain attitude, i.e.,
to play a symbolic pantomime in order to animate the bunch of traits con-
stituting the scheme. We can see it, for instance, in the different moments
of a bird’s flight or in a ballerina’s movements. Not forgetting also that the
object’s relation with the retina is neutral and the only way of giving an
answer to the movement of the real subject is to have an existential rela-
tion with it. And imagining is another form of doing it.
We can now finish these very lines about what in Sartre’s philos-
ophy the activity of imagination means and implies by pointing out
several suggestions:
If the body is implicitly perceived within its relation with other per-
ceived objects (including “its own body” in just an objective relation),
indicating a distance as a necessary condition to be aware of the objects,
then what happens with the objects we call images? It seems that we can
discern in the very activity of imagination an attempt of consciousness to
grasp absent or non-existent objects in an always-corporeal way, because
imagination lends life to what is feeble, to what is surrounded by nothing-
ness, conferring to a negative trait (the object’s inexistence or absence) a
positive character which is much stronger than its material concreteness.
That is why imagination is also much more powerful than conceptualis-
ation; because it is a dynamic form that operates having as reference the
possibilities of the existence, of the corporality, although both of them are
negated. In hypnagogic images this very process reaches such a higher
level that there is no distance between our embodied consciousnesses and
images, and we can say here that we are pathologically our images.
So, in conclusion:
1) If the body can provide a unified subject in single acts of con-
sciousness, as Morris sees in Sartre’s proposal, because in perception
the same body is at the centre of different situations, spatially, we could
say the same of imagination. Also in it there is consciousness unity,
even in hypnagogic images – that represent, as we said, a desperate
attempt to give some (corporeal) reasonability to the appearing objects.
2) Summing up these traits which were pointed out, we just have to
say, as Sartre did, that consciousness is not a black box, but a continuous
relation that a particular body has with objects – perceived, concep-
tualised or imagined (the hodological space). In order to be coherent
with Sartre’s perspective we also have to assume that in imagination we
similarly relations with objects; the only difference is that they are not
real. But that does not mean there is an evanescent way in dealing with
images. To prove against that prejudice we take the schematic drawings
case, for instance, and we are compelled to see that the body-subject
inexorably puts its marks in images.
3) So, if consciousness has a spatial origin referring itself per-
manently to that same origin, an identical situation must happen in a
100 Clara Morando
special form of consciousness as the imagination, which is a form of

consciousness, probably the most important.
References
Gurwistch, A. (1964). Field of Consciousness. Duquesne: Duquesne

University Press.
Merleau-Ponty, M. (1945). Phénoménologie de la perception. Paris:
Gallimard.
McCulloch, G. (1994). Using Sartre. An Analytical Introduction to Ear-
ly Sartrean Themes. New York: Routledge.
Moran, D. (2005). Husserl. Founder of Phenomenology. Cambridge:
Polity Press.
Morris, P. S. (1976). Sartre’s Concept of a Person. An Analytic Ap-
proach. Massachussetts: University of Massachussetts Press.
Sartre, J. P. (2003). La transcendance de l’Ego et autres textes
phénoménologiques. Paris: Vrin.
Sartre, J. P. (1993). O Ser e o Nada. Ensaio de ontologia fenomenológi-
ca. Lisbon: Círculo de Leitores.
Feelings and the Self
Dina Mendonça
All philosophers recognize how crucial the notion of self is for emo-
tional processes – that is, to be emotionally aroused is to feel the self
implicated in some way. Both self and emotions are at the center of
much philosophical debate and there are on-going debates and disa-
greements and many unclear issues regarding both topics. However,
both concepts seem to share a common trait: there is a sense that both
emotion and self are such that their design is never fully complete – that
is that they are open-ended entities. Even though both the concept of
emotions and the concept of self are continuingly being analyzed and
discussed in philosophical debates, this open-ended character of both
may be one way to better grasp their nature. In this paper I want to
look upon their relationship by focusing as much as possible on their
open-ended nature.
The paper argues for a situated approach to the nature of emotion
by showing first how a situated approach can bring to the surface in-
teresting connections between the self and the emotional world, and
second how such an approach allows us to understand how emotions
design the self. The first part of the paper lays down the situated ap-
proach to emotion and the many ways in which the implicated self can
be understood. The second part of the paper looks at some emotions
(fear, love, pride and jealousy) in order to illustrate and elaborate on the
conceptual map constructed in the first part of the paper, and indicate
some of the open-ended nature of both concepts.
102 Dina Mendonça
Part I: Situations and the Self
i. Situated approach to emotions
One way to grasp the intentionality of emotions and avoid attaching

them solely to the sentient subject is to take a situated approach to the
nature of emotions. De Sousa takes such approach in his book The
Rationality of Emotions (1987) when he introduces the notion of para-
digm scenarios. De Sousa writes,
My hypothesis is this: We are made familiar with the vocabulary of emotion by

association with paradigm scenarios. These are drawn first from our daily life
as small children and later reinforced by the stories, art, and culture to which we
are exposed. Later still, in literate cultures, they are supplemented and refined by
literature. Paradigm scenarios involve two aspects: first, a situation type providing
the characteristic objects of the specific emotion-type (where objects can be of the
various sorts identified in chapter 5), and second, a set of characteristic or “nor-
mal” responses to the situation, where normality is first a biological matter and
then very quickly becomes a cultural one (De Sousa, 1987, p. 182).
I think it is accurate to state that we become acquainted with the vo-

cabulary of emotion through paradigm scenarios but I also think that
the dynamics of paradigm scenarios is more complex. First, the story
of how these paradigms are drawn is more complex than De Sousa de-
scribes them. It is not simply that stories reinforce paradigm scenarios,
though I’m sure that experience before story telling is crucial for the
emotional relevance of stories. However, I think they also can introduce
new paradigms that are reinforced by daily life existence (or not), that
is, stories point out possibilities of paradigms as well as complexity of
paradigms. There is probably a creative process between stories and
daily events that form these paradigm scenarios. And only this explains
that, as De Sousa writes, “A paradigm can always be challenged in the
light of a wider range of considerations than are available when the case
is viewed in isolation.” (De Sousa, 1987, p. 187)
Second, the assumption of normal responses to a situation requires
detailed explanation avoided by De Sousa. How is this normal deter-
mined? De Sousa writes that it is first a biological matter and then a cul-
tural one, but the question remains: how is it determined? Biologically
Feelings and the Self 103
men and women procreate; does that mean that it is normal for couples
to have children in situations of love and abnormal not to have children?
What is the status of the abnormality? No doubt biological and cultural
structures make some general boundaries for responses to emotional
situations: biologically, one cannot feel like expressing love with kisses
if one does not have a mouth to kiss with; and culturally yawning is to
be avoided (or disguised) to prevent offending people in certain situa-
tions in countries where it is taken as a sign of boredom. Nevertheless,
given that there are issues about the normativity of emotional processes,
as well as questions regarding the source and function of normality,
much more needs to be said about the status of these biological and
cultural structures and their role in establishing boundaries and norms
for emotional processes.
Finally, in De Sousa’s description the person who experiences is
placed outside the scope of the situation, and this forces De Sousa to
end up dividing the world of paradigm scenarios in two: objects and
normal responses. Such division obscures one of the greatest advan-
tages of the situated approach to emotion, namely the movement and a
complex dynamics of sequences of events in emotional processes.
In order to maintain the richness of De Sousa’s paradigm scenarios
without the problems identified above I want to complement it by using
John Dewey’s conception of situation (Logic: The Theory of Inquiry,
1938) in order to indicate a different understanding of the connections
between self and emotion, and the genealogy of such a connection.
The notion of “situation” is crucial for the understanding of Dew-
ey’s philosophical work. For the purpose of this paper, it suffices to
point out that Dewey’s redefinition of experience, as primarily an affair
of doing/undergoing, shifts the focus of attention from the functioning
of our sense organs to the situated, complex structure of our actions as
living creatures. That is, we base our analysis of experience on the com-
plex structure of situations, instead of basing it solely upon the func-
tioning of our sense apparatus. Another important consequence follows
from this description of experience as an interaction: experience is no
longer primarily cognitive. Stating that experience is no longer primar-
ily cognitive means that experience is now primarily an active-passive
affair (MW 9, p. 147), and there are more activities than the cognitive
ones. Ultimately, this provides a reason to show that when we reflect
104 Dina Mendonça
upon reality we are not reflecting merely upon an epistemological ac-

count of experience, and therefore knowledge is no longer the only ruler
available to measure reality.
The idea of being situated allows us to clarify that organism and
environment are not understood as independent entities. In fact, we can
only understand organism as part of its environment, and we can only
understand environment as a dynamic process that involves the organ-
ism. This means that we cannot fully understand the implicated self of
emotional processes without seeing it as partly defined by the emotional
processes it undergoes, and similarly the emotional process is defined
by the character of the self. Consequently, self and emotional processes
are never totally taken to be given and complete once and for all for, as
Tallisse writes, in his simple and clear book on Dewey, “the term ‘envi-
ronment’ does not denote some permanent, independent entity; rather, it
is shorthand for an array of interrelated and active forces or factors that
collectively constitute the conditions within which we, at some particu-
lar time and in some particular place, live.” (Tallisse, 2000, p. 19) Like-
wise, organism does not stand as something permanent, an independent
entity; rather it is shorthand for an array of interrelated active forces or
factors that collectively constitute modes or ways in which one lives at
some particular time and in some particular place.
A common objection that is raised to Dewey’s philosophy and which
is of particular interest for the philosophy of emotions is the charge of
subjectivism. The claim is usually supported by the fact that Dewey’s
description of experience always leaves unanswered the question “who
is the subject of the experience?” Dewey’s view of experience in terms
of situations does not seem to completely cut across the subject-object
distinction, for the subject necessarily characterizes situations. For ex-
ample, the notion of a problematic situation requires Dewey to accept
that a problematic character be identified by a concrete and determinate
subject who has the experience (Mackay, 1942, p. 394). The difficulty
may be formulated in the following question: “Would the situation still
be problematic without a subject to disclose it?” If we answer yes, we
are forced to wonder why Dewey chooses such terminology; for the
word “problematic” implies someone to find it such. However, if we
answer negatively the problematic character of the situation is subjec-
tive because the subject of the experience necessarily determines it. In
addition, it may be argued that within Dewey’s description the problem-

atic character of situations is dependent on the abilities of the subject to
identify problems.
According to Dewey, the objection relies on a confusion regarding
the importance of the subject in the experience and on our bias towards
objectivity. Dewey writes:
Experience shows that as a matter of fact objective reference precedes subjective

reference. Reference to a subject instead of an object is extrinsic and reflective. It
is indeed only another mode of objective reference; that is, some tediousness of
the object is accounted for in terms of an unusual state of the object. Otherwise to
say “I am bored” and “It is tedious” are merely two phrases to express exactly the
same fact. (LW 2, p. 91)
No doubt, with Dewey’s terminology, we can establish a distinction be-

tween the sentences “I am bored” and “It is tedious.” In fact, it is pre-
cisely because we can translate experience in terms of situations that we
can draw such a distinction and imagine, for instance, both a scenario
where one is bored without facing a tedious situation, and one where
someone is not bored within a tedious situation. Therefore, Dewey’s
terminology is not subjectivist but, on the contrary, allows us to escape
the type of subjectivism that invalidates any exchange of information
and makes criticism impossible. Dewey claims that the objection arises
mainly because we take for granted that one gets bored when and only
when a situation is tedious. We assume that when the situation is of a
certain kind it accordingly affects the subject who experiences: if the
situation is tedious the subject feels bored; if the situation is pleasant,
the subject is pleased; if the situation is confusing the subject feels puz-
zled; etc. We may add that it is precisely this that makes us think that
something is wrong when, for some reason, the subject does not have a
certain disposition in a given situation; for instance, when a scary situ-
ation does not frighten the subject.
For Dewey, asking the question “Whose experience?” carries an-
other implication, namely that the nature of ownership is such that
rational discourse about what is owned is unattainable. As Dewey ex-
plains, this implication is absurd for “it would be to infer from the fact
that houses are usually owned, are mine and yours and his, that the
possessive reference so permeates the properties of being a house that
106 Dina Mendonça
nothing intelligible can be said about the latter” (LW 1, p. 179). We

can talk about experiences and situations just as we can talk about the
existence and the properties of a house independently of being owned.
The quality of belonging is an additive that is important to note, both
concerning a house or a situation, with respect to certain consequences.
It can be argued that this additive is not as simple as it sounds. Hocking,
who in general accepts Dewey’s description of experience, asserts that
the addition of the subject is both a gain and a loss. He argues that,
This addition of the psychological “I-think” gives us a more com-
plete view of the whole situation. But – and here is the puzzle – with the
addition something seems to have escaped. To assert “John is a rascal”
is less complete than to say, “I think John is a rascal,” but it is more
forceful: the prefix “I think” may relieve me from an action for libel,
because in telling the truth more circumspectly it withdraws the force of
assertion from the outer world. (Hocking, 1940, p. 412).
Hocking’s objection allows us to state the importance of ownership
more clearly. The notion of situation requires a more detailed descrip-
tion than one given by a mere sentence “I think John is a rascal.” In
order to fully develop such an example we would have to describe what
prompts such a statement: what John did, why he did it, when he did it,
etc. The notion of “situation” helps us to see not only that ownership is
not a private and exclusive affair, but also, that ownership is capable of
denotation when the situation is fully described. For, as Dewey writes,
“the quality of belonging to some one is not an all-absorbing maw in
which independent properties and relations disappear to be digested
into egohood,” (LW 1, p. 179) and “in some specifiable respects and for
some specifiable consequences, these selves are capable of objective
denotation just as are sticks, stones, and stars.” (LW 1, p. 179) What
this discussion reveals is that Dewey’s notion of situations transforms
the sense of subjectivity of experience into possible objective reference
because the subject can be observed, named and described. When we
need to ask “whose experience?” we have simply not described the sit-
uation accurately.
Accordingly, experience is no longer something private, only ac-
cessible to the subject who has the experience. And Dewey’s notion of
situation is a rich way to grasp the dynamics of emotional processes for
it gives us the connection to the self without limiting such a connection
to a private and inaccessible realm. In order to fully profit from the in-
sights of Dewey into experience requires a more complex description of
paradigm scenarios. The pattern of emotional activity that I will present
in a very brief form aims at providing such theoretical vocabulary such
that we may be able to focus our attention on the lively activities of
mind instead of its results, and enable us to understand that the clear and
compact words we have for some emotions do not exhaust the thought
provoking reality of emotional life (Dewey LW 10, p. 49).
ii. Self and The Pattern of Emotional Activity
Before exploring the way in which we can map out the pattern of emo-
tional activity it will be helpful to consider which way we should under-
stand the notion of self so that we can more clearly understand its role
in emotional activity and see how emotional activity contributes to the
notion of self.
In an overview article on the several notions of self, Gallagher
(2000) summarizes the continuous ongoing reflection by philosophers,
psychologists and other cognitive science experts, focusing on two
important aspects of the self: the minimal self and the narrative self
(Gallagher, 2000, p. 14) The Minimal Self is considered phenomeno-
logically, that is, in terms of how one experiences the sense of self and
how consciousness of oneself one takes oneself as an immediate subject
of experience, unextended in time. As Gallagher writes, “the minimal
self almost certainly depends on brain processes and an ecologically
embedded body, but does not have to know or be aware of this to have
an experience that still counts as a self-experience.” (Gallagher, 2000,
p. 15) One of the aspects of the minimal self is that it is the home of the
first-person pronoun “I” in its self-reference form such that it does not
allow a mistake. That is, it has the feature of immunity to error principle,
such that when a person refers to herself she can never make a mistake
about this (Gallagher, 2000, p. 15). The other important conception of
self that Gallagher identifies, the Narrative Self, is defined as “A more
of less coherent self (or self-image) that is constituted with a past and
a future in the various stories that we and others tell about ourselves.”
(Gallagher, 2000, p. 15) The narrative self grants not only an abstract
108 Dina Mendonça
entity of narration as an identity extended in time, but also a sense in

which this abstract narrative is in a sense open (not only because future
events transform past history) but also because, as Gallagher writes,
“we cannot prevent ourselves from ‘inventing’ ourselves.” (Gallagher,
2000, p. 19)
The design of the pattern of emotional activity, made in analogy to
the Deweyan pattern of inquiry, provides a theoretical background to
seeing how emotions and self (both minimal and narrative conceptions)
interact with one another. After an explanation of the pattern of emo-
tional activity it is possible to illustrate in which ways we can consider
the self to be implicated, granted that there is considerable difficulty in
clarifying the interrelations between the minimal self and the narrative
self as Damasio’s work testifies (Gallagher, 2000, p. 20).
I will go through the pattern quickly, pointing out the different mo-
ments with an illustration.
Pattern of Emotion (Mendonça)
1. Settled adjustment
2. Indeterminate situation; disturbance
3. Identification of a Tension
a. Feel
b. Create families of emotions
c. Compare/Contrast (other situations)
d. Construct a Narrative
e. Work out possible sequences
4. Assert Judgment (label situation)
The first moment of the pattern (settled adjustment) aims to point to

how things stand before the situation occurs, but this moment is some-
what false. That is, in order to conceive a situation one needs to be
able to separate it and distinguish it from the flow of occurrences and,
consequently, one needs to assume some type of settled flow of events
where nothing is disturbed. However, as every moment of life is full of
uncertainties, tensions etc, such that we can say that there are always
several situations involved. Yet, when we think of a situation, for exam-
ple the death of a close friend, this appears separated from the rest of
experience. The moment before receiving the bad news (for example,
imagine someone sitting in her living room reading a book) would be
taken as the settled adjustment before the situation, and it is in this

sense a settled adjustment is conceived. Nevertheless, the separation of
something as a situation depends on being of relevance for the subject
of the situation, though it does not require that the subject, the self, be
conscious of that as a situation.
The situation becomes indeterminate when this previous flow of
events such as sitting in the living room reading a book is disturbed
when the telephone rings and the person is told: “Roger died in a mo-
torbike accident.” From the moment the person hears the words to feel-
ing tension there is a space, that is, the tension is not immediate. So,
the situation is indeterminate with respect to its issue (LW 12, p. 110),
consequently the situation is not one of doubt (or fear, or sadness) in a
subjective sense (LW 12, p. 110): at that point the situation is simply
disturbed, confused, obscure (LW 12, p. 109).
The pattern of emotion is felt as beginning with the recognition of a
tension, that is, the indeterminacy is formatted in terms of a situation of
tension. The identification of a tension is not a quick process and as the
picture of the pattern illustrates it involves a process with many stages,
and the sense of identification may vary in degree and may not require
full consciousness. That is, the items “a-e” should not be seen as a rigid
sequence or one that must be fulfilled completely in order for a situa-
tion to be identified as an emotional process. Instead, the identified items
should be taken as a dynamic occurrence where one may return to the first
step given (a: feel) after having gone through the process of (i) through
(iii). In our example, it may be that from the moment she hears the words
to feeling the tension there is a space – that is, the tension is not immedi-
ate. But once tension settles, she feels sad, wrapped up in the news. She
begins to form a family of emotions: sadness, bewilderment, anger, pity,
curiosity, etc. She compares and contrasts the news of his death with oth-
er situations of loss, and with memories of him (situations of no loss). She
works out possible consequences of tension (future family events, future
possible loses, what she will do next time she sees her aunt and uncle,
etc.). By doing this she constructs a narrative; his death slowly becomes a
story that points to future action: calling her aunt and uncle, remembering
him more than she has done over the past year, etc.
In Dewey’s pattern of inquiry the end is reached when the indeter-
minate situation becomes settled, but the pattern of emotion ends when
110 Dina Mendonça
the initial tension is integrated. This requires some explanation. Integra-

tion is the assimilation of the emotional situation as an experience (to
use Dewey’s terminology). Using my previous illustration, we can un-
derstand the notion of integration by seeing that though her cousin died
a few days ago it may take her months, years to feel that grief is inte-
grated. One acknowledges this sense of integration of grief by realizing
that it takes time for people to be able to talk about someone they have
lost. This is not only valid for the case of grief but one can also think
of other emotional processes that sometimes take years for the person
who lived through them as an experience to be able to talk about them.
The pattern of emotion was made with the startling conclusion that
within Dewey’s description emotions behave like ideas. But why would
thinking through situations be helpful for a theory of emotions? There
are many advantages to the philosophy of emotions (Mendonça, 2012)
but here I simply want to concentrate on those advantages that allow a
richer description between the self and the emotional process felt.
Given the description of the pattern of emotion we can see that first,
the pattern allows for the emotional character of a situation to change
with reflection on it. This is because the construction of a narrative may
transform the initial family of emotions of a certain situation such that
a situation that is initially labeled as anger taken from the perspective
of the minimal self, can be later labeled as sad taken from the narrative
self perspective. For example, I may feel initially angry about a harsh
reply from a friend but label the situation as sad by attaching it to a lack
of communication between us months after the event occurred. This
means that the person who forms part of the situational whole is also a
potential source of change of the emotional situation. Consequently, we
can consider personhood in its complexity allowing for continuity and
simultaneously accepting the possibility of change.
Second, the ability to identify change in situational processes in the
pattern of emotion opens the way for an explanation of emotional matu-
rity, since experiencing sad situations will hopefully change the impact
of sadness when sadness is part of the family of an emotional situation
of grief. That is, the narrative self-perspective will have an impact on
the minimal self.
Third, it provides a more complex way to explain how emotions
resonate. Feeling like crying because I see someone else crying varies
in intensity to the extent that the situation resonates with me. And the
situation resonates with me depending on the person I am (have I felt
similar things, does my “self ” incorporate this other person) and the
person I see in an emotional situation.
Given the interconnectedness that the pattern of emotion allows us
to establish within a situational whole it may be interesting to consider
that the way we conceive personhood, among other things, will be more
accurate when thought of as the sorts of situations that the person has
undergone and what type of expectations one has of future experiences.
In fact, there is a part of our common sense description of other people
that captures the connection made clear by the pattern of emotion: we
refer to people saying, “Joanna is a solid person, someone you can trust
in a situation involving danger,” or “John is at loss in situations involv-
ing change,” etc.
What I hope the previous words have offered is that there is a dual
open-ended connection between the self and person and the emotions.
On the one hand, the emotional situations a person undergoes modify,
change, fortify the person contributing to the ongoing creation of who
we are both at the minimal and narrative level of the self; on the other
hand, while someone undergoes an emotional situation they contribute
towards the modification and strengthening of the paradigm scenario at
hand. That is, while people are building and creating selfhood they are
simultaneously constructing and perfecting paradigm scenarios.
Part II: Some Emotions and the Implicated Self
In the second part of the paper I want to consider a few emotions and
point to how their open-character is connected to the open character of
the narrative self. This reflection does not aim to give a complete analy-
sis of each of these emotions but to point out how emotional complexity
implies certain types of openness. Ultimately this means that there are
ways in which emotional processes are open to modification, and that
these transformations are incorporated into the narrative self in such a
way as to contribute to the ongoing invention of the self. Underlying
112 Dina Mendonça
all the descriptions given in this second part there is the attempt to es-
tablish the open-ended character indicated above. Firstly, a short look
at fear will show that even the emotions considered as simple may car-
ry complexity and when we acknowledge them we realize that their
open-endedness is only visible while we consider the narrative self and
it is the self that is implicated. Secondly, when we look upon love we
will recognize that part of this open-endedness is linked to contrary
elements within the same emotional process. Thirdly, the look at pride
and jealousy aims to point out that what we take as part of our self or
not changes the emotional processes available to us, and that taking
something as part of ourselves can result in different formats.
i. Fear
Now you may ask how it is that people are constructing these paradigm
scenarios for fear is fear no matter what people undergo. Even if one
accepts that certain emotions are so complex that they are constant-
ly under change one would have to admit that the basic emotions are
not changeable like that. Typically, the emotion of fear is portrayed as
a quick response to a perceived threat: we see a snake and we flee,
or shake, or scream. It is considered a basic survival mechanism that
identifies dangerous situations. Yet we are scared of things like dogs,
fire, snakes but also of bureaucracy, speaking in public and engaging in
relationships, and though there is clearly something in common among
these examples they also raise the insight that there is an element of
complexity that is not grasped by the description of fear as a quick re-
sponse to danger.
I want to take the complexity of fear a step further by using a book
published by the Portuguese philosopher José Gil entitled Portugal,
Hoje. O Medo de Existir (Portugal, Today. The Fear of Existing) where,
among other things, the writer considers the inheritance of fear such
that this emotional state is no longer something felt in situations of dan-
ger but as part of the character of a person (in his book Gil refers to the
character of the Portuguese people). This heritage of fear passes from
parents to parents, generation to generation, similar to the way we have
learn to react fearfully to sight of a snake. Gil claims that, “today, thirty
years after the regime of fear, we continue to live with it. Portuguese
society and the Portuguese have not lost the fear, even though (or maybe
because of that) the new generations know little about the past of the
Salazar period”1 (Gil, 2004, p. 78).
The descriptions of fear that Gil provides currently existing in the
Portuguese are not the things that typically come to mind in the philos-
ophy of emotions, but they testify to how even in a basic emotion like
fear, the way we learn is crucial for future experience of emotional situ-
ations, and how complex an apparent simple emotion can be. However,
I do not want to deny the difficulty in changing and modifying the ex-
perience of fear, but to point out that fear may be a much more complex
emotion if considered within the scope of the narrative self as José Gil’s
book does. This complexity points to an open-ended aspect of emotion-
al processes, namely that the inheritance of emotional processes may
be such that emotions become part of an identity (self or nation) as a
background emotional mood, but this is only visible when we consider
the self in its narrative format.
ii. Love
It is arguable whether love can qualify as an emotion but it clearly is

an emotional process that is felt by the minimal self, such as when a
mother looks at her children, and felt by the narrative self as when the
mother looks at how she has loved her grown up children. The myste-
rious nature of love is a witness to the complexity that surrounds the
emotional process of love. Maternal love is unquestionable, yet I once
heard a son tell his mother about the issue of hitting his children: “It’s
okay to discipline them, as long as they know you love them.” In fact
some people actually believe that it is not possible to discipline without
doing this. Fortunately, this method has slowly become unacceptable.
But what I want to consider here is that the love of a mother, such an
unquestionable type of love, has been created as an image associated
1 My translation of the original quotation: “Hoje, trinta anos depois do fim do re-
gime do medo, convivemos ainda com ele. A sociedade portuguesa, os portu-
gueses não perderam o medo, ainda que (ou talvez por isso) as novas gerações
pouco saibam do passado salazarista.”
114 Dina Mendonça
with a type of behavior that one would classify as opposed to loving.

That is, when someone thinks of the love of parents, in the abstract or
concretely, it is seldom the case that one thinks of times or moments of
hitting, or speaking violently, etc. Yet, the relationships with parents are
the unavoidable proof that this sentiment of so-called love holds much
more other feelings than loving ones.
There is no doubt that the way we have first learned to experience
such complex emotional situations such as living with the ones we love
and share routines with, combine wishes and desires, and negotiate ten-
sions and problems, must determine the way we, as adults, form and
continue relationships with our children. The importance of identify-
ing this inescapable characteristic of the emotional process of love is
that it makes us realize that while we are not free to change the past,
the future lies open-ended and its open-endedness is related to how we
view what has happened to us. That is, the character of the emotional
process changes when the person of the situation changes her perspec-
tive – that is, the one that is able to identify the hidden contradictory
feelings within an emotional process may be capable of renewing their
emotional processes (for instance, avoiding hitting as a means of pro-
tection in love and finding other ways to deal with difficulties). It is not
a problem in itself that situations of love include elements of power
struggles, fights, irritations, etc. for it is not necessarily problematic that
emotions frequently incorporate contradictory elements even when we
are not aware of this. In sum, the contradictory feelings that may lurk
within emotional processes grant us ways of modify the feeling of those
same processes.
iii. Pride
Pride requires a lot of cognitive processing related to the self and overall
results from a positive self-evaluation (Lewis, M., Takai-Kawakami,
K., Kawakami, K., & Sullivan, M. W., 2010). It is considered one of
the self-conscious emotions and is associated with behavioral cues
that have innate elements that are automatically perceived by others
(Shariff AF, Tracy JL., 2009). The positive self-evaluation of pride
results in seeing the self and the object as separate. As Michael Lewis
writes,
The emotion I have labeled “pride” is the consequence of a successful evalua-

tion of a specific action. The Phenomenological experience is joy over an action,
thought, or feeling well done. Here the focus of pleasure is specific and related
to a particular behavior. In pride the self and the object are separated, as in guilt.
Unlike shame and hubris, where subject and object are fused, pride focuses the
organism on its action; the organism is engrossed in the specific action that gives
it pride. (Lewis, 2004, p. 630)
However, the separation between subject and object is connected to the

self in a relevant way such that what is considered the object, standing
in relation to the subject belongs to it in a narrative sense. A person
who is proud of her country’s achievements reveals a sense of self that
requires a connection to her country, as opposed to a person who does
not feel part of her country. In this sense the variety of things that can be
the object of pride (oneself, an action, a member of the person’s family,
a person’s children, an action by a person’s children, other people, a per-
son’s possessions such as a house, history, etc., a person’s connections
such as her country, her friends, etc.) vary to the extent that one takes
something as part of one’s sense of self, and reveals the open character
of emotional processes through the change of valence it holds. While
it is mostly considered something to feel and there is also a tradition of
seeing it as a virtue, it stands for a sin and negative emotion in many
religious traditions. The point of this last remark is not to make a judg-
ment on how to consider pride but to point out that the valence given to
it clearly depends on the type of narrative self at hand.
iv. Jealousy
Finally, I want to take a very short peek at jealousy because in a way it

stands at the opposite side of pride concerning the relation of the self
with something other than the self. While with pride subject and object
stand apart, in jealousy subject and object are united. In Jealousy what
stands in relation to the self is taken as interior and belonging to the self,
not through a relevant connection but it is taken to belong to the self,
116 Dina Mendonça
such that a threat to this belonging anticipates a sense of loss of some-

thing the person values as part of herself, such as friendship, or love, or
a working relation. Similar to the case of pride, the way the narrative
self is understood and the way the emotional situation is understood
will change the existence of jealousy, its impact and its design.
Conclusions
What the paper here presents is a suggestion that the notion of an im-
plicated self should be examined in more detailed so as to find some of
the ways people free themselves from certain implications or how they
create new ways of being implicated. Though it may be clear that the
boundaries between exterior and interior to the self are not fixed, the
very existence of emotional processes seems to indicate that patterns of
establishing borders exist and that these may be transformable.
In addition, if the first part of the paper holds true then the connec-
tion of being a self and learning to speak about emotional processes may
hold more interesting insights than the ones that have so far been offered
concerning the connection between language and emotional processes.
Finally, hopefully the paper has made clear that it would be most
suggestive to produce a map of different conceptions of the self with
the connections to their different theoretical consequences regarding
the experience of emotional processes (for example, the Stoic per-
spective of self admits the possibility of not feeling grief, yet for
Rousseau that would possibly be the mark of a cruel self). It may be
the case that our philosophical heritage may provide some keys for the
connection between the contemporary Minimal self vs. the Narrative
Self, for somewhere in our past reflection may lie the answers for our
present models.
References
De Sousa, R. (1987). The Rationality of Emotion. Cambridge, MA:

MIT Press.
Dewey, J. The Early Works, (1882–1898). Edited by Jo Ann Boydston.
5 vols. Carbondale and Edwardsville: Southern Illinois University
Press, 1969–1972.
Dewey, J. The Middle Works, (1899–1924). Edited by Jo Ann Boydston.
Press, 1976–1988.
Dewey, J. The Later Works, (1925–1953). Edited by Jo Ann Boydston.
Press, 1988–1991.
Faber, A. & E. Mazlish, (1982), How To Talk So Kids Will Listen & Lis-
ten So Kids Will Talk, New York: Avon Books.
Gallagher, S. (2000). Philosophical conceptions of the self: implications
for cognitive science. Trends in Cognitive Science, 4(1), 14–21.
Gil, J. (2004). Portugal, Hoje. O Medo de Existir. Lisboa: Relógio
d’Água.
Hocking, W. E. (1991). Dewey’s Concepts of Experience and Nature. In
John Dewey The Later Works, 1925–1953. Edited by Jo Ann Boyd-
ston, 14, Carbondale and Edwardsville: Southern Illinois Universi-
ty Press, 411–426.
Lewis, M. Takai-Kawakami, K., Kawakami, K., & Sullivan, M. W. (2010).
Cultural differences in emotional responses to success and failure.
International Journal of Behavioral Development, 34(1), 53–61.
Lewis, M. (2004). Self-Conscious Emotions: Embarrassment, Pride,
Shame and Guilt. In The Handbook of Emotion, Third Edition, Mi-
chael Lewis, Jeannette M. Haviland-Jones and Lisa Feldman-Bar-
rett (eds.), New York, NY: Guilford, 623–636.
Mackay, D. S. (1991). What Does Mr. Dewey Mean by an ‘Indeter-
minate Situation’? In John Dewey: The Later Works, 1925–1953.
Edited by Jo Ann Boydston, 15, Carbondale and Edwardsville:
Southern Illinois University Press, 393–401.
118 Dina Mendonça
Mendonça, D. (2012). Pattern of Sentiment-following a Deweyan sug-

gestion. In Transactions of Charles Peirce Society, Vol. 48, No. 2,
Spring 2012, 209–227.
Mendonça, D. (2003). “When Heads Happen – Exploring the Deweyan
Conception of Mind”, presented at SIFA October.
Mendonça, D. (2003). Human Nature in the Making. Cadernos de Filo-
sofia, IFL, ed. Colibri, nº 13, 29–41.
Shariff A. F., Tracy J. L. (2009). Knowing who’s boss: implicit percep-
tions of status from the nonverbal expression of pride. Emotion.
9(5), 631–9.
Talisse, R. B. (2000). On Dewey: The Reconstruction of Philosophy.
Belmont, USA: Wadsworth/Thomson learning.
Part III
Cognition, Psychology,
Neuroscience
De Se Attitudes and Semiotic Aspects of Cognition
Erich Rast1*
Overview
In this article, I will re-examine some of the classical puzzles for de se

attitudes that have been laid out by Hector-Neri Castañeda, David Lew-
is and John Perry in various articles and compare them with Jackson’s
Knowledge Argument. The origins of these puzzles go further back to
work by Russell on egocentric particulars, by Frege in ‘Der Gedanke’,
Wittgenstein’s considerations on subject-uses of I in the Blue Book, and
work by Roderick Chisholm. Nevertheless, it is fair to say that de se
puzzles gained widespread popularity only later due to publications by
Castañeda (1967), Perry (1977, 1979), and Lewis (1979).
The article starts with a survey of well-known de se puzzles: Per-
ry’s supermarket example, his Rudolf Lingens example, and David
Lewis’s Two Gods thought experiment. I will then discuss Jackson’s
Mary example, which bears a striking similarity to de se puzzles. After
this primarily exegetical part, I will address the question regarding what
these puzzles have been taken to show and what they really show. My
central thesis is that typical de se puzzles reveal an important and epis-
temically irreducible aspect of thinking, but do not allow for any con-
clusions regarding physicalism and the Mind-Body problem. As I will
argue, there is a special kind of introspective knowledge, the existence
of which is fully compatible with physicalism and this special kind of
* An earlier version of this paper was presented at the conference ‘Metaphysics of

the Self ’ at the Institute for the Philosophy of Language (IFL), Lisbon, in Decem-
ber 2009. I would like to thank Klaus Gärtner, Jorge Gonçalves, Franck Lihoreau,
Daniel Ramalho, António Marques and Dina Mendonça for fruitful suggestions
and commentaries. This research was conducted under a postdoctoral fellowship
from the Portuguese Fundação para a Ciência e Tecnologia.
122 Erich Rast
knowledge results from the fact that all sorts of episodal thoughts plays
a particular role in thinking and cannot be replaced by other kinds of
thoughts. On the basis of this insight, I will suggest a trivializing inter-
pretation of de se puzzles: A thought of a certain type, say α, is a nec-
essary condition for the occurrence of a corresponding α-behaviour or
action, simply because episodal thought tokens are divided into distinct
classes according to the role they play in cognition. Correspondingly,
it is highly unlikely that thoughts of type α, which present subjective
experiences in cognition to the one currently thinking, could play the
same role in that person’s thinking as thoughts of another type, say β,
which present physical knowledge to the one currently thinking, and
vice versa. This does not mean, however, that physical knowledge can-
not explain thoughts of type α or that instances of α and β belong to
different ontological categories.
Before going on, some terminology must be clarified. Talk about
thinking is often ambiguous between thinking in an episodal sense and
dispositional thinking in the sense of having the ability to entertain cer-
tain thoughts or having a dispositional belief that p for some embedded
proposition p. In what follows, I have the former in mind when talking
about thinking here; this kind of thinking might also be called cognition
in order to set it apart from the dispositional reading. Cognition is time-
bound and actual. In contrast to this, in what follows belief and other
propositional attitudes should be understood in the common disposi-
tional sense. When an agent thinks (viz. cognates) that p this means that
he currently entertains a p-thought or is having a p-thought. In contrast
to this, when someone believes that p he has a disposition to act as if p
were the case. Without further qualification the terms episodal thinking
and cognition leave the question open whether the respective agent hav-
ing the thought endorses the embedded proposition or not, i.e. whether
she considers the embedded proposition true or not. I shall understand
these terms in their non-philosophical sense in what follows, according
to which the agent indeed takes the embedded proposition to be true.
Understood in this sense, when someone thinks that John is 32 years
old he also dispositionally believes that John is 32 years old and does
not just ‘think the thought’ without being committed to its truth. The
converse does not hold. From the fact that someone dispositionally be-
lieves that p it does not follow that he thinks that p at a certain time. I
De Se Attitudes and Semiotic Aspects of Cognition 123
will often talk about dispositional knowledge instead of belief, since the
kind of beliefs about one’s own thoughts I will discuss result in some
form of introspective knowledge as long as the person in question does
not suffer from serious mental illness. My main points could be made
entirely in terms of belief instead of knowledge, though, and so not too
much weight should be given to this terminology.
Puzzles of De Se Attitudes
The supermarket example laid out by Perry (1979) is one of the clearest
cases for the de se attitudes:
I once followed a trail of sugar on a supermarket floor, pushing my cart down the
aisle on one side of a tall counter and back the aisle on the other, seeking the shop-
per with the torn sack to tell him he was making a mess. With each trip around
the counter, the trail became thicker. But I seemed unable to catch up. Finally it
dawned on me. I was the shopper I was trying to catch. (Perry, 1979, p. 3)
This scenario has two different, though related aspects. On one hand,
Perry might cognate in various ways about himself without realizing
that his episodal thoughts are about himself. For example, he might
think that someone/the only bearded philosopher in a Safeway store
west of Mississippi/John Perry/this man (+pointing gesture) in the mir-
ror is making a mess without realizing that it is he himself whose sack
of sugar is torn. As long as John Perry is sufficiently amnesiac, doesn’t
remember his name or that he hasn’t shaved himself in the morning, and
doesn’t recognize himself in a mirror, none of these ways of thinking
about himself seems to explain his behaviour, until he starts to think:
It is me, who is making a mess, I am producing a trail of sugar! On the
other hand the example also has a linguistic aspect. The different ways
in which John Perry might linguistically realize his episodal thoughts
don’t seem to have the same explanatory power for his action than Per-
ry’s utterance of I am making a mess. For example, in order to express
the same thought using his proper name, John Perry would have to ad-
ditionally believe that he is called John Perry, which he doesn’t believe
124 Erich Rast
in the given scenario. This has led Perry and others to the conclusion
that there is an essential reading of the first-person indexical that is
irreducible with respect to its power for explaining certain changes in
behaviour and actions.
Another well-known example can be found in Perry (1977):
An amnesiac, Rudolf Lingens, is lost in the Stanford library. He reads a number

of things in the library, including a biography of himself, and a detailed account
of the library in which he is lost. He believes any Fregean thought you think might
help him. He still won’t know who he is, and where he is, no matter how much
knowledge he piles up, until that moment when he is ready to say, ‘This place is
aisle five, floor six, of Main Library, Stanford. I am Rudolf Lingens.’ (Perry, 1977,
p. 492).
Again, the problem in this example is that it always seems to be con-

ceivable that the epistemic agent in question gathers all kind of knowl-
edge about himself from external evidence, but only when he realizes
that this information is about himself will he be able to act accordingly.
So it appears as if, from an epistemic point of view, the agent learns
something when he realizes that he is John Perry or Rudolf Lingens
respectively, he himself is making a mess on the floor or he himself
is located at aisle five, floor six, of Main Library, Stanford, and so on.
Lewis (1979) has provided a famous variant of these examples that, al-
beit being highly contrived, is rather instructive from a logical point of
view, because it is spelled out in terms of possible worlds semantics for
propositional attitudes like dispositional belief:
Consider the case of two gods. They inhabit a certain possible world, and they
know exactly which world it is. Therefore they know every proposition that is true
at their world. Insofar as knowledge is a propositional attitude, they are omnisci-
ent. Still I can imagine them to suffer ignorance: Neither one knows which of the
two he is. They are not exactly alike. One lives on the top of the tallest mountain
and throws down manna; the other lives on top of the coldest mountain and throws
down thunderbolts. (Lewis, 1979, pp. 520–521)
The idea behind this thought experiment is as follows. In epistemic log-

ic based on normal modal logic an agent’s epistemic state is represented
by the set of possible worlds, i.e. maximally truth-making doxastic al-
ternatives, that are accessible from the actual world by an accessibility
relation for that agent and the respective kind of propositional attitude.
Suppose the belief set Γ is the set of possible worlds accessible by an

agent’s current belief relation from the actual world. Consider now a
case when the agent is in doubt whether p is the case or not. In a possi-
ble worlds setting, this boils down to saying that there are some w1, w2
in Γ such that p holds at w1 and ¬p at w2.When the agent learns that p
is the case, the revised accessibility relation will yield a set of possible
worlds from which all ¬p worlds have been removed. The more an agent
learns about the universe, the smaller becomes his belief set. If only one
world is left and as in Lewis’s scenario that world is the actual world,
then the agent is omniscient (see Figure 1 for illustration). Whatever he
can learn about the state of nature he has already learned. Nevertheless,
Lewis claims that for example Zeus in this scenario still doesn’t know
that he himself is Zeus, that he himself is sitting on the tallest mountain
and is throwing down manna, and so on.
Lewis has devised this scenario to illustrate a limitation of the usual
truth-conditional semantics for dispositional attitudes based on modal
logics with Kripke or Scott-Montague semantics. He concluded from the
thought experiment that possible worlds alone do not suffice to represent
what an agent might learn, hence do not suffice to represent epistemic
states and corresponding dispositional attitudes in general, and for this
reason he proposed a more fine-grained model in which having a relational
attitude is modelled as the self-ascription of a property by the respective
agent. Since properties in Lewis’s view are more fine-grained than sets
of possible worlds, more attitudes can be distinguished by the property-
ascription view and so his approach can adequately represent the
assumption that the two gods in the scenario might have different epis-
temic attitudes despite being omniscient about all the ‘external’ facts.
Lewis’s solution was not very appealing to many philosophers, and
rightly so, because he did not present enough details about the prop-
erties he had in mind. Lacking a detailed and positive metaphysical
account of properties his suggestion remains unsatisfactory. Fortunate-
ly, many other solutions to the Two Gods puzzle have been proposed
during the past few decades, and, technically speaking, any way to make
dispositional belief more fine-grained will do as long as one merely
strives for descriptive adequacy with respect to de se puzzles. One of the
simplest solutions, which was also discussed by Lewis (1979), is using
centred possible worlds. Instead of bare possible worlds ordered pairs
126 Erich Rast
consisting of an agent and a possible world are taken as semantic base

entities. With this small change the lack of omniscience of the two gods
can be easily represented. For example, Zeus’s belief set {‹w, Zeus›, ‹w,
Jahwe›} would contain two tuples, one containing himself and the other
one containing Jahwe as long as he hasn’t realized that he himself is
Zeus. When he learns that he himself is Zeus he removes the pair con-
taining Jahwe from his belief set (see Figure 2). There are many other
approaches to hyperfine-grained attitudes that may be used for dealing
semantically with Lewis’s scenario: structured propositions from von
Stechow (1982) and Cresswell (1985), approaches based on substitut-
ing reified propositions for truth-values as in Thomason (1980), using
impossible worlds as in Hintikka (1975), and property theory of Bealer
& Mönnich (1989). To make a long story short, it is fair to say that de se
puzzles nowadays no longer pose any particular technical challenge to a
logician who is interested in a descriptively adequate truth-conditional
semantics for belief and similar attitudes.
However, none of these approaches give a satisfactory answer to
the question of what exactly the agent learns when he realizes that he
himself is called by a certain name or satisfies a certain property. The
power of Lewis’s example can be illustrated further by thinking of pos-
sible worlds for a moment in an old-fashioned way as possible constel-
lations of matter and the forces that hold it together. By assumption,
in the example both gods know everything there is to say about the
constellation of matter in the universe and the forces that hold between
matter. Still, so it is claimed, they can learn something. Consequently,
whatever there is left for them to learn cannot be an aspect of matter and
the forces between that matter. As Stalnaker (2004) remarks, there is a
striking similarity between de se puzzles of this kind and the Knowl-
edge Argument by Jackson (1982, 1986), which was originally intended
to show that qualia are not physical and therefore physicalism is false.
Jackson (1986) has laid out this example in the following, much cited
passage:
Mary is confined to a black-and-white room, is educated through black-and-white

books and through lectures relayed on black-and-white television. In this way
she learns everything there is to know about the physical nature of the world.
She knows all the physical facts about us and our environment, in a wide sense
of ‘physical’ which includes everything in completed physics, chemistry, and
neurophysiology, and all there is to know about the causal and relational facts
consequent upon this, including of course functional roles. If physicalism is true,
she knows all there is to know. […] It seems, however, that Mary does not know
all there is to know. For when she is let out of the black-and-white room or given
a color television, she will learn what it is like to see something red, say. (Jackson,
1986, p. 291)
Just like the gods in Lewis’s scenario Mary is omniscient about the ex-
ternal world, yet she doesn’t know what it feels like to experience red,
before she has been exposed to a red object. In de se puzzles the agent
still needs to learn that it is he himself/she herself who has a certain
property or is called by a certain name. According to the Knowledge
Argument, the agent is also supposed to learn something in addition to
her physical knowledge about the world, namely the way in which a cer-
tain experience feels to her. So do these thought experiments indicate
that there is something nonphysical like an irreducible I-thought or a
particular red quale? As I will argue in the following section, the answer
to both questions is No. Neither de se puzzles nor corresponding argu-
ments for qualia show that the physical world is not closed or that cer-
tain phenomena in episodal thinking are not reducible to physical states.
Both kinds of puzzles show, however, that there is a special sort of in-
trospective knowledge which is fully compatible with physicalism and
based on remembering past experiences or thoughts. In order to have
this kind of knowledge, an agent must store an experience or thought in
a way such that the retrieval of this memory fulfils a certain role in his
thinking; when something else is stored instead, it might not fulfil the
same role when it is retrieved. Investigating this kind of knowledge in
combination with some general semiotic considerations will pave the
way for a trivializing interpretation of de se puzzles.
Assessment of the Thought Experiments
It is worth noting that the Two Gods thought experiment does not strict-
ly speaking present a positive argument for the thesis that there are de
se attitudes. Instead of acknowledging that each of the gods lacks some
128 Erich Rast
knowledge one could also take the example as illustrating that the gods
do not lack any knowledge, since there are no epistemic alternatives
left for them to consider. By assumption Zeus knows everything there
is to know about the physical world, hence one might argue that he also
knows who he is, what his name is and which properties he has, that it is
himself who is currently thinking, whether he is throwing down manna
or thunderbolts, and so on. If the mental supervenes on the physical, the
defender of physicalism might argue, then Zeus also knows everything
there is to know about his mental life and consequently does not lack
any knowledge about his own episodal thoughts either. From this point
of view Lewis’s premise that the two gods lack knowledge could be
regarded as false. Let us call this the denial reply. It does not seem to be
very convincing. On the one hand, the thought experiments have an in-
tuitive power that lures us into believing that the agent can indeed learn
more. Using thought experiments in this way has rightly been criticized
by Dennett (1991) as intuition pumping in an area where our intuitions
may be utterly misleading. On the other hand, the denial reply neglects
an important distinction between the static theoretical knowledge an
agent has at a certain time and the knowledge an agent might gain when
he or she has an insight about him/herself. Nothing in the description of
the scenario warrants that one of the gods actually has the insight that it
is he himself/she herself who has such-and-such properties.
Jackson’s Mary example has been criticized by Dennett (1991)
in a similar fashion as in the denial reply to the Two Gods example.
According to Dennett, Mary does not learn anything new when she is
exposed to a red object for the first time and we only find this claim
counter-intuitive because we cannot adequately imagine what it means
to have exhaustive knowledge about the physical world. Many other re-
plies have been given to the original argument, ranging from a critique
of ‘intuitions’ to the denial of one or more of its premises, and Jackson
himself has by now converted to physicalism. For example, according to
the well-known ability hypothesis Mary does not gain new knowledge
but an ability – see Lewis (1983, 1988/1990), Nemirow (1980, 1990),
and Churchland (1985, 1989), cf. Coleman (2009) for a comprehensive
critique. Others such as Conee (1994) and Tye (2009) have defended
the so-called acquaintance hypothesis according to which acquaint-
ance with certain objects or experiences cannot be explained in terms
of knowing that, or abilities. I will in the following paragraphs lay out

a position according to which (a) in the classical de se puzzles and in
Jackson’s Mary example the agent in question learns something new in
the sense of gaining new knowledge, but (b) neither de se scenarios nor
Jackson’s Mary example speak against physicalism, (c) the examples
are described incorrectly, since the world changes in each of them, and
(d) episodal thoughts of a certain kind can trivially not be substituted by
episodal thoughts of another kind and dispositional knowledge cannot
replace episodal thinking. In a nutshell, what I claim is that in order to
know the episodal thought that you’ve been thinking (or are just think-
ing), you need to think it first, and thinking a certain episodal thought
is trivially different from an explanation of that thought or an episodal
thought that explains that thought.
To back up these claims let me start with an analogy. Imagine a
little robot toy that contains sensors that react to light. A sensor func-
tion s: T→{l, r, f} from time to three values works as follows: When
there is more light to the left than to the right, the sensors return l;
otherwise if there is more light to the right than to the left they return
r; otherwise they return f. The robot automatically stores each sensor
input in its memory and can also perform three actions: move forward
F, move forward left L, or move forward right R at a time t. The device
is programmed as follows: (1) Store s(t) in memory. (2) If s(t)=r then
R, else if s(t)=l then L; otherwise F at discrete times t measured by an
internal clock. If designed correctly, this little device will follow the
strongest light source until it bumps against something, malfunctions,
or its battery runs out. In its memory a certain sequence is stored, and
which one depends on its environment. If, for example, the robot is
never exposed to light from its left, it may go in circles to the right and
the symbol l will never be stored in its memory. Now suppose your
colleague from the local A.I. department wanders by and laments that
this device is much too simple. He upgrades the robot’s software and
hardware to allow for running complex programs on it, including condi-
tional actions based on the contents of the memory and devising simple
plans such as ‘If the last sequence of actions was llrll, then R.’ Despite
this gain in abilities and ‘cognitive powers’, as long as the robot is never
exposed to light from the left none of its routines that introspect l states
and act conditionally upon them will do anything. Suppose further that
130 Erich Rast
your colleague from the A.I. department becomes famous by inventing

a rather complicated program that, when hooked up with sensors and
running on suitable hardware, becomes conscious and self-aware. (As
a philosopher, you remain rather sceptical about this conjecture but the
dispute ends in a stalemate: neither can you falsify his claim that the
running program is conscious nor can your colleague falsify your claim
that it is not conscious.) Out of generosity your colleague upgrades
your robot toy, uploads the program, and immediately once the program
is started the robot begins to explore the world. Being equipped with
sophisticated A.I., this robot bears some resemblance to RoboMary de-
scribed by Dennett (2005, pp. 122–129), but it illustrates a different
point here. The point is this: Even though our robot is now conscious,
self-aware, and capable of sophisticated planning and other higher cog-
nitive abilities, what it can experience, learn, and do is restricted by
its environment. If you mount a strong light on its right side, the robot
will never store the symbol l. Since the robot is conscious, this means
that it will never have an l experience, or, to put it in other words, the
robot will have no possibility of knowing how it feels like to have an l
experience.
This example shows two things. First, it is obvious that the robot
does not lack any ability when it is never exposed to light from the left.
It has the same abilities when it is never exposed to light from the left as
it has when it is exposed to light from both sides and does not lack any
know-how in either case. So the so-called ability reply to the Knowledge
Argument does not seem to be convincing if the above example illus-
trates basically the same problem as Jackson’s Mary example. Secondly,
there does seem to be a sense in which the robot learns something new
when it is exposed to light from the left for the first time, even though its
blueprint and the software running on it doesn’t change. It has a certain
experience that it has never had before and can now store the fact that
it has had this experience in its memory. It acquires memory of a past
experience, which is, by definition, a learning process.
Keeping the robot in mind, let us return to the Mary example. Does
Mary learn something new when she is exposed to a red object? The an-
swer is clearly Yes. She can now remember an experience that she didn’t
have before. Does that mean that qualia are not physical or that there is
some irreducible phenomenal knowledge? The answers which one gives
to these questions depends on the stance one takes about red experiences
viz. l-experiences. Someone who thinks that the robot can only have an
l-experience if it is conscious, i.e. consciousness is a necessary condition
of having an l-experience, would without doubt be inclined to say the
same about Mary. In this view, which has been criticized by Dennett and
others for its reliance on unstable intuitions, Mary learns something new
when she is exposed to red, but only if she is conscious. In this point of
view, machines cannot be conscious as they merely follow a fixed set
of instructions. So when our robot acquires knowledge and eventually
becomes just as brilliant a scientist as Mary, it will still not have an
i-experience even if it is exposed to light from the left. Notice, however,
that neither of the thought experiments supports this view or its opposite.
Certain types of computational structures, when running on some hard-
ware, or certain types of physical structures could be conscious without
anyone ever knowing for sure that they are. Suppose this were the case.
Then there would still be a difference between Mary’s epistemic state be-
fore she has been exposed to red and afterwards and the robot’s epistem-
ic state before it has been exposed to light from the left and afterwards.
Moreover, suppose the robot was omniscient about the physical
world and not conscious. Still, the robot’s epistemic state would be dif-
ferent before and after it has been exposed to light to the left, because
it can store l-input from the sensors, which is a learning process, and
being able to remember a past input is a kind of knowledge. But in order
to be able to remember a past input or experience the respective agent
needs to have had it. So there is indeed a special kind of knowledge
at play that one might label as ‘phenomenal’, but the conditions for
acquiring this knowledge are entirely independent from the question
whether qualia are physical or not and from the question whether an
agent is conscious or not. If qualia are nonphysical and exist, then the
occurrence of a nonphysical phenomenon is a necessary condition for
having knowledge of qualia experiences, but the Knowledge Argument
does not show this. The Knowledge Argument merely shows that cer-
tain knowledge can only be acquired by having certain experiences, be
these reducible to physical structures or not.
A very similar point can be made about the Two Gods example.
Since each of the gods is omniscient about the physical universe, they do
know who they are and are able to identify themselves. However, there is
132 Erich Rast
still a sense in which each of them lacks knowledge, as long as they do

not make use of their ability. For instance, unless Zeus thinks the episodal
thought I am Zeus he cannot get to know that he has been thinking that
thought. Regarding his dispositional belief this means that the occurrence
of the episodal thought I am Zeus is a necessary condition for his having
the dispositional de se belief that he himself is Zeus. The thought has
to occur in actual cognition first, and only after it has occurred can the
agent have the corresponding dispositional de se attitude. If this is so,
then there are two ways to interpret the puzzle. First, one may claim that
as a consequence of having complete theoretical knowledge about the
universe an agent will invariably come to episodically think correspond-
ing thoughts of the form I am F. Since we are talking about arbitrary
properties F that the agent possesses, this view does not seem to be very
compelling. If at all, it only makes sense under very strong rationality
assumptions and when the agents in question are highly idealized like the
two gods. Since no actual, resource-bound agent can have infinitely many
episodal thoughts, and thus every actual agent lacks knowledge about
herself, another response seems to be more appropriate. According to this
view, the complete theoretical knowledge of the agent in question does in
itself not warrant that she has corresponding thoughts of the form I am
F. If the agent had such a thought, then she would be able to recognize
that she has had it even from a purely 3rd-person perspective, provided
that episodal thoughts are entirely reducible to certain kinds of physical
structures, but since having the dispositional knowledge does not cause
or otherwise warrant that the agent actually has an episodal thought of
the kind needed for having a corresponding de se belief, she might not
recognize that she herself is F. If, on the other hand, episodal thoughts
are not entirely reducible to certain kinds of physical structures, then an
agent that is omniscient about the physical world might not recognize
from a 3rd-person perspective that she has had that thought. In both cases,
however, the agent can only recognize that she has had such a thought
after she has actually had the thought. Again, the puzzle does not decide
anything about the ontological status of the mental. All it says is that you
need to actually think certain thoughts in order to be able to retrieve from
memory that you have thought them, and, as in case of the robot, the abil-
ity to retrieve memories of past mental events, be they ultimately physical
or not, can be considered a form of knowledge. This sort of knowledge

may be called introspective knowledge.
Semiotic Aspects of Cognition
From what has been said so far, a number of conclusions can be drawn.
Firstly, the Two Gods and Jackson’s Mary example are fully compatible
with physicalism and neutral about the question whether physicalism
holds or not. Secondly, the descriptions of the examples are incom-
plete. Suppose physicalism is true and Mary is experiencing something
red. Then her brain state changes, too, and consequently her previous
knowledge about the physical world is outdated. Likewise, if physical-
ism is true and an agent has the episodal thought I am F, then hav-
ing this episodal thought corresponds to a change in the agent’s brain
state, rendering his previous knowledge incomplete. A static picture of
the universe, as is presumed by using some simple, tenseless possible
worlds semantics for propositional attitudes, is not entirely adequate for
describing these scenarios. When for example Mary’s knowledge is up-
dated according to the changes in the world after she has been exposed
to a red object for the first time, she would be able to deduce from her
physical knowledge alone that she has had a red experience. Neverthe-
less, as simple as this may sound, the fact that she has had a red experi-
ence is a necessary condition for her to realize, on the basis of physical
knowledge about the world only, that she has had a red experience. As I
will lay out in more detail below, even under the premise that physical-
ism is true, recognizing the red experience by means of measurement
and physical knowledge and subsequently forming knowledge about its
recognition by remembering it is not the same as remembering the red
experience itself.
This position is similar to what has been proposed by defenders of
what Nida-Rümelin (2009) calls the new knowledge/old fact view but
with one rather crucial difference: in the present view, the fact is new,
although it will be possible to fully explain it in physical terms if phys-
icalism is true. If on the other hand physicalism happens to be false,
134 Erich Rast
then having a red experience trivially amounts to a new phenomenal

fact, namely the fact that the person in question had this particular red
experience. In both cases, the world changes and so the knowledge of
the omniscient agent in question needs to be updated. What about the
updated knowledge then? Does the updated dispositional knowledge
suffice to explain a particular red experience or an agent’s insight that
he himself or she herself has a certain property? The correct answer is:
Why not? Why should the updated knowledge not suffice to explain the
respective experiences provided that physicalism is true? Explaining
in this sense means nothing more than knowing the conditions under
which the agent in question has the respective experiences and there-
fore is able to form introspective knowledge about them. However,
there is no reason whatsoever to believe that dispositional knowledge
of an agent or someone else about an agent’s experiences or thoughts
can substitute or replace in any way the agent’s having certain episodal
thoughts that present experiences, insights or memories thereof to him/
herself in his/her cognition. Even under the assumption that certain
episodal thoughts directly correspond to dispositional, physical knowl-
edge about a particular red experience this red, i.e. when we would
confine further considerations to whatever corresponds to dispositional
knowledge about this red in actual cognition – like the episodal thought
that the episodal thought this red is such-and-such –, there is no reason
whatsoever to believe that these kinds of episodal thoughts could or
should play the same role in cognition as the thought this red that an
agent has when he is confronted with a red object or the thoughts he
might have when he imagines or remembers a red object. Likewise,
various ways of identifying oneself in actual cognition in 3rd-person
ways might not play the same role in cognition as the thought circum-
scribed as being of the form I am F that occurs in cognition when an
agent realizes that he has the property F.
The resulting perspective on de se attitudes is trivializing. Certain
episodal thoughts are necessary conditions for a certain kind of behav-
iour, whereas other sorts of episodal thoughts are not – no matter what
we know about episodal thinking and how it should be described. An
explanation of our robot’s inner workings does not substitute the ma-
chine’s actual having and processing certain sensor inputs or fetching
data from its memory even if the robot itself processes this explanation
and accurately measures its own state. Correspondingly, Mary’s knowl-

edge about the physical world cannot substitute her having a red ex-
perience. Generally speaking, episodal thoughts of a certain sort play
a certain role in thinking that thoughts of another sort might not. This
thesis should be relatively uncontroversial, yet it has been overlooked
consistently in the assessments of respective thought experiments.
Episodal thinking has another important property that is relevant
for the assessment of the examples: it does not work like a language.
Too see why, consider the robot example again. In a well-functioning
robot an l symbol occurs when the left sensor receives more input than
the right sensor, and correspondingly for the processing of r and f sym-
bols. The symbols are stored in a memory module of a given capacity.
Suppose now that the robot’s program says ‘When the memory contains
the sequence lll, remove it from memory and execute R for the next
two time units.’ Another part of the program could possibly use another
symbol, say a, to refer to a sequence lll. However, in order for the sym-
bol a to have the same effect as in the above programming instructions
it has to be assigned in an appropriate way to sequences of lll’s stored
in memory; no matter how this connection looks like in detail the robot
must remove three l’s from memory and give the R signal for the motors
to go right for the next two time units in order to run the same program
on the basis of the alternative representation a. In this respect, although
the choice of the symbol l is completely arbitrary, a sequence of three
l’s in memory is the most basic representation of three inputs from the
left sensor on this machine with respect to the particular program run-
ning on it. While it would be possible in this example to replace three l
symbols with a single one such as a, which is stored whenever the left
sensor has received three subsequent inputs in a row, this strategy only
works because the program only acts upon three received l symbols. It
does not work in general. Suppose the robot gets the following, more
complicated instructions: ‘Whenever there is a sequence of n symbols
l followed by a sequence of n/2 symbols r, remove the whole sequence
and drive right for n time units.’ This rule involves the counting of sen-
sor inputs, and for this purpose at some level of description the robot
must store the number of subsequent l symbols in memory. To store the
number of subsequent l inputs, one may for example store n symbols
l for n inputs from the left sensor. Alternatively, binary states could be
136 Erich Rast
used. The number 3 is then represented by setting two states to on (see

Figure 3). No matter which representation for n inputs of a kind is cho-
sen, though, in order to run the program successfully it must be able to
produce motor output for n time units. Taking into account the complete
device, including the software running on it, we may thus say that the
desired representation of n sensor inputs encodes the number n.
Most symbols do not encode what they stand for. Consider for ex-
ample the Arabic digit 3. In order to be able to represent three things
this symbol must be mapped to three signals, objects, or other entities
of a suitable kind. The symbol 3 does not itself exemplify what it stands
for and it may thus be called a purely representational symbol, or, in the
terminology of Langer (1951), a discursive symbol. From these sym-
bols Langer distinguishes presentational symbols such as the three dots
in Figure 3. Presentational symbols exemplify what they stand for in
addition to representing it. In other words, they encode directly at least
certain aspects of what they represent. Notice that the two binary gates
are a tricky case. They seem to be more presentational than the digit 3,
yet without a rule that maps the on-state of one of them to two objects
or signals and the on-state of the other one to one signal or object, the
gates cannot for themselves present three objects directly. If the map-
ping mechanism is included in the description of the symbol, then it
may be considered presentational but otherwise it is purely representa-
tional just like the symbol 3.
I do not claim that this distinction can be generalized easily to cases
that are more complicated than the encoding of natural numbers. A sim-
ple isomorphism will most likely not suffice as an explanation of more
complex presentational symbols and iconicism is one of the most vex-
ing problems of semiotics. Instead of grounding the desired distinction
on the notion of an isomorphism or making similar attempts to give a
precise account that would require much more support, I will only as-
sume some less stringent and more tentative definitions in what follows.
Let us speak of a purely representational sign if the connection between
sign and the signified is conventional and (in this sense) arbitrary. Let
us, following H.N. Castañeda to some extent, assume in contrast to this
that a presentation encodes finitely many aspects of what it represents in
a way that makes the connection between the sign and the signified not
just conventional and arbitrary. (The reason for allowing such a vague
definition here is that an explanation of the exact meaning of the phrase

‘not just conventional and arbitrary’ should be left to neurophysiology if
physicalism is true and psychology if physicalism is false. That at least
some such presentations exist has been illustrated by the example of fi-
nite presentations of natural numbers.) Given this distinction, it is clear
that languages are purely representational, for it is well-known that even
allegedly iconic expressions like onomatopoeia in natural languages are
conventionalized. Cognition, on the other hand, cannot be purely rep-
resentational. If cognition worked like a language and were also purely
representational, then each thought token would stand for something
else. As I briefly laid out in Rast (2007, pp. 270–275), this view would
lead to an infinite regress, since an agent would then have to check to
which entity a particular thought token refers in order to ‘understand’
the token, and so on for the symbols the token refers to, for the symbols
the referent of the token refers to, etc., until a presentational symbol is
encountered. This modern Homunculus problem shows that episodal
thinking cannot in general be only representational; it must at some
point involve presentations of objects, experiences, and so on. More-
over, as I’ve mentioned before and as H.N. Castañeda has emphasized
throughout his work (see for example Castañeda, 1989, 1990), humans
are finite and therefore presentations of objects or experiences in epi-
sodal thinking must be finite, too. Consequently, when an agent thinks
a thought that may be circumscribed in public language as I am F for
some property F, the I-presentation of herself in her cognition encodes
finitely many aspects of herself and she attributes the property F to this
presentation. Likewise, a presentation of a particular red experience in
cognition presents finitely many aspects of a particular red object or a
region of an agent’s visual field in her cognition. If physicalism is true,
these presentations are encoded and processed directly by an agent’s
brain – an open, analogue, massively parallel, and presumably non-de-
terministic computational system – and the result of such a computation
is itself finite. But even if physicalism is not true, there does not seem
to be any alternative to the popular view that episodal thinking is finite.
Although not the central tenet of this article and somehow inde-
pendent of the previous considerations, I would briefly like to discuss
the popular yet sometimes misunderstood thesis that episodal thinking
is also computational in nature, which further closes the gap between
138 Erich Rast
the robot and the Mary example. By ‘computational’ I mean that the
mind or brain can within the limits of its finite resources compute only
functions that are representable by terms of the λ- or π-calculus and
their reduction to canonical form, or another suitable sequential or par-
allel Turing complete formalism. While there are alternatives to this
view, they are hard to reconcile with what we know about the physical
world. First of all it must be noticed that the brain or mind can compute
certain computable functions within its physical limits. It is therefore
not a good idea to claim that the brain or mind is not at all a compu-
tational device. So to reject the computational model one has to claim
that the mind/brain is higher than a Turing machine in the hierarchy
of computational systems. Hypothetical devices that can solve prob-
lems that a Turing machine cannot solve are called hypercomputers, and
there is a whole hierarchy of them. A relatively weak hypercomputer
would for example be the ‘accelerated Turing machine’, which is based
on ideas already considered by Russell. An accelerated Turing machine
completes its first step during computation (such as moving the head
on the tape, printing a symbol to tape, etc.) at time 1, the second step
at 1+1/2 time units, the third one at time 1+1/2+1/4, and so on. When
it reaches the limit at time 2 it might have solved a task that provably
no ordinary Turing machine can solve. There are more credible descrip-
tions of hypercomputers, but none of them seems to be fully compatible
with current physical knowledge (see Lokhorst, 2000). Other alterna-
tives such as Penrose (1989, 1997) are more elaborate, but don’t of-
fer any convincing explanation of how the mind might work. Yet other
alternatives such as claiming that the mind is neither computational
nor hypercomputational nor based on quantum-physical phenomena
amount to sheer mysticism. Now the fact that the computational model
seems to be the best explanation of how episodal thinking works does
not mean that it is the right explanation. The issue is clearly undecided.
Still, the best explanation is the best explanation unless someone comes
with a better one, and we should prefer the computational model unless
someone comes up with direct counter-evidence to it. Notice that even
if physicalism were false a computational model of the mind would still
be more attractive than any of the alternatives mentioned; in that case,
however, some mysticism could not be avoided due to the problem of
explaining mind-matter interaction.
If the computational model is basically the right picture of how

episodal thinking works, then invariably certain episodal thoughts
fulfil a different role than others within the computational system as
a whole. Although some episodal thoughts may in principle fulfil the
same role as others if they are ultimately linked to the same sensor
inputs and motor outputs, as the a versus lll example has illustrated, it
is not very likely that a particular this red presentation or a particular I
am F thought in human cognition could be replaced by some radically
different thoughts, such as those that represent some realization of dis-
positional knowledge in cognition, and still fulfil the same role within
the given computational system.
What conclusions can be drawn from these general considerations
concerning de se puzzles? Let us return to Perry’s supermarket example
mentioned at the beginning of this article. In the thought experiment cer-
tain kinds of thoughts are connected to a certain kind of behaviour. For
example, John Perry looks at himself in the mirror without recognizing
himself, thinks what may be circumscribed in public language as This
guy is making a mess and attempts to follow the person in order to tell
him that his sugar package is damaged. Suddenly he thinks I am making
a mess, which must here be understood as a rough and deficient natural
language circumscription of what actually goes on in his brain (or mind,
if physicalism is false) when he has the insight and correspondingly
starts to clean up the sugar. According to the view I have suggested,
this is so because a particular finite presentation in John Perry’s cogni-
tion lead to a particular behaviour, whereas another presentation leads
to another kind of behaviour. The connection of this phenomenon to
natural language is loose, though. Take any double-vision puzzle like
Quine’s Ortcutt example in Quine (1956) Perry’s ship example in Perry
(1979, p. 483), or Richard’s phone booth example in Richard (1983).
Being forced to base their judgments on finite presentations of objects,
people use certain expressions for some presentations and others for
other presentations without realizing that these presentations present
the same object in different ways to them. Hyperfine-grained belief can
be used to model such cases, but outside of attitude ascriptions the re-
spective natural language expressions retain their public language ref-
erence and refer to objects with infinitely many properties. As countless
discussions about propositional attitudes have shown, even when the
140 Erich Rast
respective natural language expressions occur inside attitude ascrip-

tions it is controversial whether hyperfine-grained attitudes should be
used to truth-conditionally encode an agent’s tendency to use certain
expressions as opposed to others for certain presentations in cognition.
Language is purely representational and public, whereas cognition is
predominantly presentational, private, and its intricacies are mostly un-
known at the time of this writing. For that reason, mappings between
these symbolic systems remain inaccurate and leave room for variations
due to different theoretical goals and modelling purposes.
Things look different, however, when one is interested in logical
representations of epistemic states of thinking or computing agents
independently of natural language semantics. Certain thought tokens
form equivalence classes on the basis of the role they play within the
computational system as a whole, including the program currently run-
ning on it, and tokens from one equivalence class can by definition not
be substituted by tokens from another one. Thus, when instances of cer-
tain classes of tokens are stored for later retrieval they play an essen-
tial role in the constitution of introspective knowledge of a certain sort
that can trivially not be replaced by knowledge based on instances of
another class of thought tokens within the computational system. De
se attitudes and the special status of subjective experiences within an
agent’s thinking as opposed to 3rd-person knowledge about them are
symptoms of this general property of computational systems. So from
an epistemic point of view hyperfine-grained attitudes seem to be una-
voidable. However, being closely tied to purely representational formal
languages, sometimes even to their syntax, existing accounts of hyper-
fine-grained attitudes lack a certain explanatory adequacy concerning
these phenomena even when they are descriptively adequate. An ex-
planatory satisfying account of hyperfine-grained attitudes for humans
would have to be based on a comprehensive theory of presentational,
episodal thinking in general, and no such theory is available at the time
of the writing of this paper. If what has been said above is correct, ac-
cording to such a theory episodal thinking is radically different from
what we are used to calling a language.
Finally, some things have to be said about the talk about ‘necessary
conditions’ in the previous sections. I have not laid out any principal
reasons why thought tokens that represent physical knowledge about
the world could not substitute thought tokens for subjective experiences
or tokens that present thoughts like I am F for a property F in the sense
that they might, under certain circumstances, fulfil the same role in epi-
sodal thinking as the thoughts they are considered to replace. There is no
principal reason why within a computational system one sort of signals
could not play radically different roles dependent on the overall state of
the system, and I have only claimed that assuming such a dual role for
certain thoughts in human thinking is an implausible attempt to explain
away introspective knowledge. My reply to Dennett and the denial reply
to the Two Gods puzzle is not that it is infeasible that certain theoreti-
cal knowledge could necessarily not play the desired role in thinking;
my reply is rather that it actually doesn’t. How then, as one might ask,
does this support the much stronger thesis that certain thoughts are a
necessary condition for certain actions? The answer is that this depends
on the way thoughts are categorized. The way typical de se puzzles are
set up no thought other than I am F plays the same role for subsequent
I-behaviour, even though another thought such as John Perry is F could
in principle result in the same behaviour. I-thoughts lead to I-behaviour
by definition, and, as one might continue, John Perry-thoughts lead to
John Perry behaviour, this guy-thoughts lead to this guy-behaviour, and
so on. If someone would claim that in the supermarket example a John
Perry-thought might lead to the I-behaviour, a defendant of de se atti-
tudes could reply that this is not possible, because thoughts that result
in the I-behaviour, i.e. in the agent having the insight about themselves
respectively described by the scenarios, are I-thoughts. They fulfil a role
in episodal thinking and memory thereof which manifests itself as the
I-behaviour. Understood in this way, having an I-thought is a necessary
condition for the occurrence of an I-behaviour, because thoughts have
been classified in this way. This is the trivial aspect of the suggested
interpretation of de se puzzles, but as I have shown above they also
exemplify the non-trivial formation of introspective knowledge. Using
the same trivial ‘cognitive a priori’ definition of thoughts in the Mary
example would render it rather incomprehensible, because in this way
of talking we would have to say that if Dennett was right then thoughts
that represent an omniscient scientist’s knowledge about colour vi-
sion would just be thoughts that represent a certain colour experience.
Not even Dennett talks this way. We usually talk in another way about
142 Erich Rast
cognition, namely in the way that thoughts are not individuated by their
role; so in this way of talking two different thoughts might (at least in
theory) play the same role, yet remain different from each other.
Summary and Conclusions
Regarding the Mary example, I have outlined an approach that Dennett

(2005) calls ‘thick materialism’. Mary learns something new and this
new, introspective knowledge is compatible with physicalism. The par-
ticular insight of an agent about himself that is described in typical de
se examples and an agent’s memory of this insight likewise result in in-
trospective knowledge. In both cases the agent stores a certain thought
token in memory and by having had the particular thought and being
able to remember having had it in some way the agent acquires intro-
spective knowledge. Within a computational system certain signals or
thought tokens play a certain role in the system as a whole that other
thoughts might not play. If t is the symbolic presentation of a certain
sensory input within a given computational system, then a presentation
t’ of knowledge about t can only play the same role within the compu-
tational system as t if it is suitably linked to other parts of the system
in the same way as t. For example, when t consists of three symbols or
signals lll that represent three inputs from a certain sensor and within
the system at that time there are also some actual presentations t’ of the
system’s dispositional knowledge about lll symbols, their storing and
retrieval, and the various roles they can play in relevant subroutines of
the computational system then under usual circumstances t’ cannot sub-
stitute t in the sense that t’ cannot play the same role as t within the com-
putational system as a whole for the roles that are fine-grained enough
to be of interest for a role-based explanation of the system. Concerning
human thinking it seems even less likely that thoughts presenting senso-
ry experiences could be replaced by thoughts that encode dispositional
knowledge while retaining the same role within the system as a whole.
If phenomenal concepts are taken as a condition for the ability of
an agent to form introspective knowledge of the kind laid out in this
article, then it follows from what has been said above that these can be
understood in purely physicalist terms without resorting to ontologi-
cally or metaphysically irreducible phenomenal properties. However,
there does not seem to be any need for any special conditions of this
sort, since the fact that humans have thoughts of different kinds in com-
bination with the fact that they have the ability to remember them (to
some extent) suffices as a general explanation as to why humans have
the ability to form introspective knowledge.
Figures
Figure 1: An agent’s belief set is shrinking when he accepts new propositions.
Figure 2: Using centered propositions for representing de se belief.

144 Erich Rast
Figure 3: Different ways of representing the number 3 and their respective decoding:
(a) Arabic digit, (b) three dots, (c) two active binary states.
References
Bealer, G. & Mönnich, U. (1989). Property Theory. In Dov Gabbay (ed.),

Handbook of Philosophical Logic. Dordrecht: Kluwer, 133–251.
Castañeda, H.-N. (1967). Indicators and Quasi-Indicators. American
Philosophical Quarterly, 4, 85–100.
Castañeda, H.-N. (1989). Thinking, Language, Experience. Minneapo-
lis, Minn.: University of Minnesota Press.
Castañeda, H.-N. (1990). Indexicality: The Transparent Subjective
Mechanism for Encountering A World. Noûs, 24(5), 735–749.
Churchland, P. (1985). Reduction, Qualia and the Direct Introspection
of Brain States. Journal of Philosophy, 82, 8–28.
Churchland, P. (1989). A Neurocomputational Perspective: The Nature
of Mind and the Structure of Science. Cambridge, MA: MIT Press.
Coleman, S. (2009). Why the Ability Hypothesis is Best Forgotten.
Journal of Consciousness Studies, 16(2–3), 74–97.
Conee, E. (1994). Phenomenal Knowledge. Australasian Journal of
Philosophy, 72, 136–50.
Cresswell, M. J. (1985). Structured Meanings. Cambridge, MA: MIT
Press.
Dennett, D. C. (1991). Epiphenomenal Qualia? In Consciousness Ex-

plained. Boston: Little, Brown and Company.
Dennett, D. C. (2005). Sweet Dreams. Cambridge, MA: MIT Press.
Hintikka, J. (1975). Impossible possible worlds vindicated. Journal of
Philosophical Logic, 4(3), 475–484.
Jackson, F. (1982). Epiphenomenal Qualia. The Philosophical Quarter-
ly, 32(127), 127–136.
Jackson, F. (1986). What Mary Didn’t Know. The Journal of Philoso-
phy, 83(5), 291–295.
Langer, S. (1951). Philosophy in a New Key. Cambridge, MA: Harvard
University Press.
Lewis, D. K. (1979). Attitudes De Dicto and De Se. Philosophical Re-
view, 88(4), 513–543.
Lewis, D. K. (1983). Postscript to ‘Mad Pain and Martian Pain. In Da-
vid Lewis. Philosophical Papers Vol. I. New York: Oxford Univer-
sity Press, 130–32.
Lewis, D. K. (1988). What Experience Teache. In Proceedings of the
Russellian Society. Sidney: University of Sidney.
Lewis, D. K. (1990). What Experience Teaches. In William Lycan (ed.),
Mind and Cognition. Oxford: Blackwell, 499–518. (Reprint)
Lokhorst, G.-J. C. (2000). ‘Why I am Not a Super Turing Machine’.
Manuscript of a talk given at University College London on 24.
May 2000. Retrieved in December 2010 from <http://homepages.
ipact.nl/~lokhorst/hypercomputationUCL.pdf>.
Nemirow, L. (1980). Review of ‘Mortal Questions’ by Thomas Nagel.
Philosophical Review, 89, 473–77.
Nemirow, L. (1990). Physicalism and the Cognitive Role of Acquaintance. In
William Lycan (ed.). Mind and Cognition. Oxford: Blackwell, 490–99.
Nida-Rümelin, Martine (2009). ‘Qualia: The Knowledge Argument’.
Published online in Edward Zalta (ed.): Stanford Encyclopedia of
Philosophy. Retrieved on 12.2.2010 from <http://plato.stanford.
edu/entries/qualia-knowledge/>.
Penrose, R. (1989). The Emperor’s New Mind. New York: Oxford Uni-
versity Press.
Penrose, R. (1994). Shadows of the Mind. New York: Oxford University
Press.
146 Erich Rast
Perry, J. (1977). Frege on Demonstratives. Philosophical Review, 86,

474–497.
Perry, J. (1979). The Problem of the Essential Indexical. Noûs, 13, 3–21.
Quine, W. V. O. (1956). Quantifiers and Propositional Attitudes. The
Journal of Philosophy LIII (5: March), 177–187.
Rast, E. H. (2007). Reference and Indexicality. Berlin: Logos.
Richard, M. (1983). Direct Reference and Ascriptions of Belief. Jour-
nal of Philosophical Logic (12), 425–452.
Stalnaker, R. (2004). ‘Knowing Where We Are, And What It Is Like’. Man-
uscript of a talk given at NYU’s La Pietra Conference on Conscious-
ness, Florence 2004, version of 2006 retrieved from <http://www.nyu.
edu/gsas/dept/philo/ faculty/block/lapietra/Stalnaker.pdf>.
von Stechow, A. (1982). Structured Propositions. Technical report of
the SFB 99. Konstanz: Universität Konstanz.
Thomason, R. (1980). A Model Theory for Propositional Attitudes. Lin-
guistics and Philosophy 4, 47–70.
Tye, M. (2009). Consciousness Revisited. Cambridge, MA: MIT Press.
The Division of the Mind: Paradoxes and Puzzles
Vasco Correia
1. Beyond divisionism
The hypothesis that the mind is divided is often put forward as a means
to make sense of irrationality, whether in the cognitive sphere (delusion-
al beliefs) or in the practical sphere (weak-willed actions). In essence,
the “divisionist” argument states that we cannot understand irrational
actions and beliefs without assuming that the mind is composed of dif-
ferent sub-systems. Donald Davidson (1985b, p. 353), one of the most
prominent proponents of divisionism, explicitly endorses this method-
ological assumption:
I have urged in several papers that it is only by postulating a kind of compartmen-

talization of the mind that we can understand, and begin to explain, irrationality.
In a similar vein, Aristotle (2009, p. 20) introduced the distinction be-

tween the “rational” and the “irrational” part of the soul to be able to ac-
count for the phenomenon of akratic action (or “lack of self-control”),
in which the agent deliberately acts against his or her best judgment.
And it was also in an attempt to solve the problem of irrationality that
Plato argued in The Republic that there are three components of the soul
(nous, thumos and epithumia). Irrational actions, he argues, stem from
a conflict between those components. Plato (1992, p. 115) evokes the
case of Leontios, a man who was unable to resist the temptation of star-
ing at a bunch of dead bodies, despite deeming such voyeurism morally
wrong. What seems puzzling about this sort of behaviour is that the
agent is aware that it would be in his best interest to perform the course
of action A, all things considered, but chooses to do B nonetheless.
148 Vasco Correia
Philosophers have always questioned the very possibility of such at-

titudes. Can one see the best option, and approve of it, and nevertheless,
in full awareness, choose the worst option? How is that possible? Al-
though a number of philosophers deny the possibility of akrasia thus
described, the proponents of the divisionist hypothesis argue that it is
indeed possible to act contrary to one’s best judgment, insofar as there
are different instances operating within the mind. In the case of Leon-
tios, for example, Plato argues that a part of him wanted to stare at the
gruesome scene, while another part of him repudiated such an attitude1.
Many authors make a similar claim with regard to specific cases
of cognitive irrationality, such as self-deception and wishful thinking,
arguing that this type of phenomenon also requires some sort of differ-
entiation in the mind2. In this view, there is a formal analogy between
self-deception and akrasia, to the extent that both phenomena seem to
imply some degree of inconsistency between the agent’s attitudes: the
akratic agent is someone who believes that A is the better option and
nonetheless decides to do B; and the self-deceiver, likewise, is someone
who believes that p is the most likely hypothesis and nonetheless de-
cides to believe that not-p. After all, if the self-deceiver did not initially
hold the unwelcome belief that p, he or she would have no reason to
make the effort of embracing the opposite belief. For example, if John
did not know “deep inside him” that he is an alcoholic, he wouldn’t go
to great pains to persuade himself that he doesn’t have a problem with
alcohol. To that extent, as Davidson suggests, it seems reasonable to
suggest that the initial belief that p, in concert with the desire that not-p,
causally sustains the contradictory belief that not-p: “Self-deception is
notoriously troublesome, since in some of its manifestations it seems
to require us not only to say that someone believes both a certain prop-
osition and its negation, but also to hold that the one belief sustains
1 Hence Plato’s metaphor in the Phaedrus (2003, p. 28) depicting the psyche as a
“composite pair of winged horses […] one of them noble and of noble breed and
the other ignoble and of ignoble breed”, constantly torn between the impulses of
desire and the recommendations of reason.
2 In particular, Donald Davidson, Robert Audi, David Pears, Herbert Fingarette and
Sebastian Gardner.
The Division of the Mind: Paradoxes and Puzzles 149
the other”3. According to the so-called “intentionalist” account of

self-deception, at any rate, the self-deceiver must hold two contradicto-
ry beliefs simultaneously: he or she must belief that p is true, while also
believing that p is false.
And this is precisely why the divisionist hypothesis is so appealing
to many authors, since it solves the paradox of contradictory attitudes
coexisting in the same mind. Even at a cognitive level, it doesn’t ap-
pear paradoxical to hold two contradictory beliefs simultaneously, so
long as the beliefs in question belong to different sub-systems of the
mind. In this sense, divisionists seem to avoid what Alfred Mele (1998,
p. 38) calls the “doxastic paradox” of self-deception, i.e. the possibility
that the same individual believes that p and that not-p simultaneously.
According to Robert Audi (1982, p. 137), this is possible because con-
tradictory beliefs coexist at different levels: the subject believes that p
at a conscious level, to be sure, but he or she “unconsciously knows that
not-p (or has reason to believe, and unconsciously and truly believes,
not-p)”. To that extent, it seems possible not only that the subject holds
two contradictory beliefs simultaneously, but also that he or she remains
unaware of that contradiction (the two beliefs “failing to clash”, so to
speak).
There are nevertheless two crucial difficulties that undermine the
divisionist postulate. The first was pointed out by Sartre in his famous
analysis of the mauvaise foi (or self-deception) and concerns primarily
Freudian-like accounts of the divided mind. Sartre’s argument is pre-
sented as a dilemma, known as the “paradox of repression”, and ques-
tions the idea that a part of the mind (supposedly unconscious) could
somehow perform the complex task of preventing another part of the
mind from contemplating harmful representations. The second difficul-
ty concerns all kinds of divisionism, Freudian-like or not. It is common-
ly referred to as the “homunculus fallacy” and was initially stressed by
Wittgenstein, who argued that the hypothesis that there are relatively
autonomous sub-systems coexisting in the mind compromises the very
notion of personal identity.
3 Davidson, “Deception and Division”. In Le Pore and McLaughlin (eds.), Actions

and Events. Perspectives on the Philosophy of Donald Davidson, Oxford, Basil
Blackwell, 1985, p. 138.
150 Vasco Correia
In this paper I shall examine some of the most influential versions

of the divisionist account, namely the models developed by Freud, Fin-
garette, Pears and Davidson. I argue (1) that each of these accounts leads
to a specific paradox or set of paradoxes, and also (2) that the divisionist
hypothesis is NOT necessary to explain ordinary cases of irrationality,
such as self-deception, denial, rationalization, wishful thinking, and
the like. This is not to say that it is impossible for our minds to suffer
any sort of division (as indeed happens in pathological cases of mental
dissociation), but simply that we do not need to presuppose such divi-
sions to be able to account for ordinary cases of irrationality. My view
is that most cases of irrationality, whether practical or cognitive, stem
from the influence that desires and other emotions exert upon our cog-
nitive processes, and thereby upon our judgments. If this hypothesis is
correct, irrational thought and irrational behaviour are typically caused
by a conflict between individual mental states (for example, a conflict
between a desire and a belief), rather than from a conflict between dif-
ferentiated parts of the mind.
2. The paradoxes of the Freudian account
We may begin by examining Freud’s influential model of divisionism,

since it admittedly inspired Fingarette’s, Pears’ and Davidson’s more re-
cent versions of this approach. As Freud (1923, p. 19) often stresses, the
cornerstone of Psychoanalysis is the claim that a significant number of
our representations remain unconscious or repressed, forming what he
initially calls the Unconscious (das Unbewusste) and later the Id: “the
division of mental life into what is conscious and what is unconscious is
the fundamental premise on which psycho-analysis is based”. However,
he also specifies that unconscious contents remain excluded from the
consciousness, not because our consciousness accidentally neglects to
contemplate them, but because they cannot become conscious, insofar
as a given strength represses them: “Analysis of these examples of for-
getting reveals that the motive of forgetting is always an unwillingness
to recall something which may evoke painful feelings” (Freud, 1901,
p. 332). In fact, the process of repression needs to be understood in the

light of the fundamental principles that rule the psychic apparatus, and
particularly the so-called “pleasure principle”, which implies that the
psyche spontaneously seeks pleasure and avoids pain (caeteris pari-
bus). Thus, a representation is repressed either because it is deemed too
painful to be made conscious or because it constitutes a threat, directly
or indirectly, to the equilibrium of the psychic apparatus as a whole.
In L’être et le néant Sartre famously argues that the Freudian ac-
count is intrinsically paradoxical. The kernel of Sartre’s argument is that
the so-called Unconscious would have to behave as a sort of “second
consciousness” to be able to repress harmful thoughts (Sartre, 1943,
p. 89). After all, Sartre observes, in order to withhold dangerous infor-
mation from the consciousness, the Unconscious would have to per-
form the highly complex task of assessing the potential impact of each
given representation in the economy of the psyche. Yet such a task could
not be achieved successfully without the unconscious being aware of
the representations in question, for otherwise it would lack the means to
decide which information is suitable to be made conscious, and which
information needs to be repressed. But if that were indeed the case, the
unconscious would have to know the negative representations in order
not to know them, which seems paradoxical4.
The second problem with Freud’s account is that it undermines the
concept of personal identity, since it attributes propositional attitudes
such as beliefs, desires, representations, memories and emotions to
mere parts of the mind, despite the fact that these attitudes only seem
appropriate to describe the person as a whole (the in-dividual, strictly
speaking). As Wittgenstein (1953, p. 281) points out, “only of a human
being and what resembles (behaves like) a living human being can one
say: it has sensations; it sees, is blind; hears, is deaf; is conscious or
unconscious”. Freud’s description of the mind’s components often falls
prey to what Anthony Kenny calls the “homunculus fallacy”, in other
words the tendency to describe the alleged parts of the mind as though
4 Simon Boag (2007, pp. 421–447) has argued that the resolution of this paradox
hinges upon the recognition that the process of repression inhibits knowledge of
knowing that which is repressed. The problem with this solution, however, is that a
third degree knowledge would then be required to determine which of the knowl-
edge is to be inhibited, and which is not.
152 Vasco Correia
they were different persons within the person. In The Ego and the Id, in
particular, Freud (1923, p. 146) develops an anthropomorphic model of
the psyche which depicts the ego, the super-ego and the id as different
“homunculi” coexisting and competing in accordance with their specif-
ic set of demands. Regarding the ego, for example, he writes that “we
can see [it] as a poor little creature subjected to servitude in three dif-
ferent ways, and threatened in consequence by three different dangers –
one posed by the external world, one by the libido of the id, and one by
the harshness of the super-ego”. Likewise, in An Outline of Psychoa-
nalysis, Freud (1940, p. 173) goes as far as to suggest that the psycho-
analyst ought to make an alliance with one of the mind’s components,
namely the ego, in an attempt to protect it from the demands set by
its two rivals: “The analytic physician and the patient’s weakened ego,
basing themselves on the real external world, have to band themselves
together into a party against the enemies, the instinctual demands of the
id and the conscientious demands of the super-ego. We form a pact with
each other”. But this raises of course the problem of personal identity,
given that the components of the psyche do not seem to have much in
common. Under such a description, in effect, one may wonder what
sort of principle would be able to ensure the overall unity of the several
sub-systems. Sartre (1943, p. 89) also stresses this point: “By rejecting
the conscious unity of the psyche, Freud cannot but presuppose every-
where a magic unity linking distant phenomena across obstacles”.
Irvin Thalberg (1984, pp. 253–254) makes a similar objection by
raising the “who’s who?” question regarding the alleged constituents of
the psychic apparatus: “Is our ego awake or asleep when we sleep?”;
“Which ‘self’ does my ego have the duty of protecting?”; “Whom does
my super-ego watch when it engages in ‘self-observation’ – me, my
ego, itself?”; “Whose enjoyment do [my instincts] ‘strive’ for?”; and
“Whose interests is the ego trying to protect in its repressions?”. Far
from solving this serious difficulty, Freud only seems to aggravate it in
his later writings, bringing forward the hypothesis of a so-called “split-
ting of the Ego” (Ichspaltung). In the posthumous text “The splitting of
the Ego in the process of defence”, remarkably, he claims that the ego
itself is liable to suffer an inner division. When the conflict between
the ego’s desires and what reality demands becomes unbearable, Freud
(1938, p. 276) writes, “The two contrary reactions to the conflict persist
as the centre-point of a splitting of the ego […] The synthetic function

of the ego, though it is of such extraordinary importance, is subject to
particular conditions and is liable to a whole number of disturbances”.
Although Freud initially confines this hypothesis to the specific cas-
es of fetishism and psychosis, he later extends it to neurosis in gener-
al5. But, of course, if the components of the mind may themselves be
sub-divided into sub-constituents, the paradoxes of divisionism become
all the more difficult to surmount.
3. Fingarette: “ego” and “counter-ego”
In spite of these difficulties, Herbert Fingarette’s analysis takes up

Freud’s idea of a splitting of the ego. However, he gives it an ingenious
twist by suggesting that the ego’s inner conflicts lead to the formation of
what he calls a “counter-ego”. According to Fingarette (1984, p. 224),
this instance is formed during the emergence of the person’s self and
coexists with the ego in the mind throughout life: “The defensive out-
come, then, is to establish what we may call a counter-ego nucleus, this
nucleus being the structural aspect of counter-cathexis”. Aware of the
dilemmas that Freud’s theory seems to face, the author explains that
“such paradoxes arise because both earlier and later versions [of Freud’s
theory] are parallel in insisting correctly on the fact that there is a split
in the psyche, but fail in defining the nature of the split adequately”
(Fingarette, 1984, p. 223). The adequate way to define the nature of
the split, in his view, would be to suggest that every ego is essentially
divided (and not just the neurotic or the psychotic ego), and also that the
upshot of such a division is the formation of two basic sub-systems, the
ego and the counter-ego.
In addition, Fingarette contends that the origin of such a split is not
the process of repression, but a process of “disavowal” through which
the counter-ego is split off from the ego. He illustrates this process by
5 Cf. Freud, Outline of Psycho-Analysis, The Standard Edition, Vintage, The Ho
garth Press, London, vol. XXIII, p. 202.
154 Vasco Correia
evoking the example of a man who resents his employer’s attitude to-
ward him, but is unable to accept (or avow) his own resentment because
he considers it immoral to feel such an anger toward someone else. As a
result, he argues, the subject refuses to identify himself with the attitude
in question and claims: “It is not ‘I’ who am angry; from henceforth I
will dissociate myself from it; it is repugnant to me” (Fingarette, 1984,
p. 218). The process of splitting off would begin at an early age, through
the constraints of education and social interactions. Negative feelings such
as shame and guilt, in particular, would play an important role in causing
the person to “disavow” certain emotions and desires, progressively induc-
ing the creation and development of a counter-ego nucleus in the mind.
The advantage of Fingarette’s account is that it seems to avoid the
paradox of repression, given that the phenomenon of disavowal, un-
like Freud’s mechanism of repression, is a conscious and deliberate act
performed by the person’s Ego: “the defensive process is a splitting of
the ego which is not something that ‘happens’ to the ego but something
the ego does, a motivated strategy” (Fingarette, 1984, p. 224). There is
however a fundamental dilemma in Fingarette’s account: assuming that
the counter-ego is not repressed (nor unconscious) but merely “split-
off” from the ego, it follows that the same consciousness must serve
the purposes both of the ego and of the counter-ego. But how could the
Ego and the counter-ego remain effectively dissociated if they share
the exact same consciousness? Here lies the paradox: on the one hand,
the subject’s consciousness is meant to deny the set of beliefs and emo-
tions which seem unacceptable to the ego; but on the other hand, as part
of the counter-ego, the consciousness must be aware of such beliefs and
emotions. In the previous case, for example, the subject has to be aware
of his anger toward his boss, for otherwise he wouldn’t bother to deny
it. But if the subject’s ego knows about the unacceptable emotions, how
can it not recognize them as its own? And what sense does it make to
say that the subject’s consciousness acknowledges certain attitudes as a
counter-ego, but not as an ego?
An additional problem with Fingarette’s account lies in the as-
sumption that people are capable of disavowing inconvenient realities
both consciously and voluntarily. This philosophical hypothesis, known
as doxastic voluntarism, is in fact ruled out as a psychological impos-
sibility by most philosophers and psychologists, who argue that no one
seems to be able to decide at will, hic et nunc, what to believe or not6.

As Jonathan Bennett (1990, p. 96) points out, although it may not be
conceptually impossible to cause oneself to believe something at will, it
seems impossible de facto to achieve such a task:
There could be simpler, quicker, more reliable means for causing beliefs in people
without giving them evidence. I passionately want to spend the evening in a state
of confidence that the weather will be fine tomorrow (I have my practical reasons),
so I give myself the thought of tomorrow’s weather being fine while snapping my
fingers in a certain way, and sure enough I end up convinced that the weather will
be fine tomorrow. We have no such fast, reliable techniques for producing belief
without evidence, but they are not conceptually ruled out.
Finally, one could also question the very existence of a so-called “counter-
ego”, much like Freud’s opponents have questioned the existence of an
unconscious. Even assuming that the process of denial may occur in-
tentionally, it seems doubtful that it should result in the formation of a
counter-ego which most of us are unaware of.
4. Pears: the rival “centres of agency”
In a sense, one could describe David Pears’ version of divisionism as

a compromise between Freud’s founder model and Fingarette’s sophis-
ticated account. On the one hand, Pears agrees with Fingarette in say-
ing that the divided mind is essentially composed of two sub-systems,
instead of the three Freud describes, which he generically refers to as
the “main” and the “secondary” sub-system. But on the other hand,
Pears (1982, p. 269) adopts the Freudian assumption that one of those
sub-systems must remain unconscious:
There is then one centre of activity in the subject’s contemporary consciousness

and another in the reservoir […] We have to suppose that each of these two cen-
tres includes any information needed to give its desire a line of action […] This
6 See for example Pascal Engel, “Volitionism and Voluntarism About Belief ”. In A.
Meijers (ed.), Belief, Cognition and the Will, Tilburg University Press, 1999.
156 Vasco Correia
extremely economic hypothesis is based on Freud’s fundamental concept of a

boundary dividing the conscious from the preconscious and the unconscious.
Pears (1982, p. 268) suggests that this is “the most economical form of
the hypothesis of the divided mind”, not only because it refrains from
engaging in an anthropomorphic description of the mind’s sub-systems,
but also because it does not postulate that such a division is an essential
feature characterizing each and every mind. According to Pears, we only
need to assume that the mind suffers this kind of differentiation in the case
of subjects who are prone to some degree of irrationality, whether cogni-
tive or practical. In principle, the cause of the division is the emergence
of a strong desire which turns out to be incompatible with the subject’s
predominant preferences. For example, the desire to have an extra-marital
affair with an attractive person is of course incompatible with the aspira-
tion to remain a faithful partner and a moral person7. If strong enough, the
desire in question may eventually form what Pears calls a “rival centre of
agency”, which is able to compete with the main sub-system and liable to
induce irrational behaviour and irrational thinking.
The interest of Pears’ account is that it avoids the homunculus fal-
lacy, given that the secondary sub-system of the mind is described as a
purely functional entity. To begin with, it is not an essential feature of
the mind, but a characteristic which only applies to irrational minds;
secondly, it is not defined idiosyncratically, given that any type of de-
sires could in principle characterize the rebellious sub-system; and fi-
nally, for that very reason, it may actually differ from one person to the
other (unlike Freudian sub-systems, which are supposedly universal).
Nevertheless, Pears’ model seems unable to avoid the paradox of
repression. After all, how could the secondary sub-system, which is un-
conscious, be able to assess the impact of each mental state on the psy-
chic apparatus? As a matter of fact, it seems difficult to conceive that an
unconscious instance should possess the awareness and the discernment
required to predict the positive or negative impact of every belief, every
desire and every representation upon the subject’s mental equilibrium.
7 It is worth noting, however, that the rebellious desire could in principle be of any
type: not just a sexual drive, as in many of Freud’s analysis, but also professional
ambition, desire for fame and glory, desire to get rich, and so forth, depending on
the subject’s specific set of preferences and values.
Aware of this difficulty, Pears (1982, p. 275) argues that the act through
which the mind excludes a given content from consciousness is not only
unconscious but also “self-reflective”: “The only possible reply is that
the suppression of the consciousness that normally accompanies a par-
ticular type of mental event may itself be unconscious without having
succumbed to a previous act of suppression. To put the point in another
way, suppression can be self-reflective”.
Ingenious as it may sound, Pears’ solution seems insufficient to
neutralize Sartre’s criticism, for even if we accept the idea that there
are self-reflective acts of repression, the question of how exactly such
mental acts are able to perform the elaborated task of separating the
wheat from the chaff, so to speak, sorting out what might be beneficial
and what might be harmful for the mind, remains to be answered. As
Mark Johnston (1989, p. 82) rightly points out, this would imply that an
unconscious sub-system of the mind should be able to manipulate the
main system in the best interest of the mind as a whole:
The question arises how the protective system could do all this without being
conscious of (introspecting) its own operations. After all, it has to compare the
outcome it is producing with the outcome it aimed for and act or cease to act
accordingly. Any consciousness by the protective system of its own operation is
“buried alive”, i.e., is not acceptable to the consciousness of the main system.
An additional problem is that Pears’ definition of the secondary

sub-system remains somewhat vague. He writes that, for the mind
to be divided, “there must be two conflicting desires which set up
rival centres of activity” (Pears, 1982, p. 268), but neglects to specify
which other conditions must be combined for the division to occur. In
particular, the reader is left wondering whether there are as many sec-
ondary sub-systems of the mind as there are strong conflicts between
desires. In the book Irrationality Alfred Mele (1987, p. 143) stresses
this point with an amusing analogy: “Explaining doxastic irration-
ality in Pears’ fashion is rather like explaining how a football team
held another scoreless by saying that the former strategically rendered
all of the latter’s scoring attempts ineffective. The football fan wants
much more than this. He wants to know how team A rendered team B’s
attempts ineffective”.
158 Vasco Correia
5. Davidson: the mind’s “compartments”
Although Donald Davidson extensively acknowledges the extent to

which his account of irrationality is inspired by Freud’s analysis, the
divisionist hypothesis he brings forward is in fact weaker than Freud’s,
and arguably even more minimal than Pears’. To begin with, Davidson
doesn’t assume that there is such a thing as an unconscious part of the
mind allegedly inaccessible to the subject’s consciousness. He doesn’t
go as far as to deny the existence of an Unconscious, but suggests that
we do not need to presuppose the existence of contents inaccessible
to the consciousness to be able to make sense of irrationality. All we
need to presuppose, he argues, is that there are different “territories”
or “compartments” in the mind which the consciousness cannot survey
simultaneously.
In addition, while Pears and Fingarette suggest that the second-
ary sub-system emerges in the mind as an organized centre of action,
competing with the main system and aiming at distinct goals, Davidson
(1982, p. 182) refuses to speculate about that aspect, thereby avoiding
the homunculus fallacy. All that is required, Davidson suggests, is that
“within each [department of mind] there is a fair degree of consisten-
cy, and where one element can operate on another in the modality of
non-rational causality”. Thus, if an agent holds contradictory beliefs,
as in the case of self-deception, one should assume that there is a sub-
system of mental states consistent with the belief that p, and another
sub-system of mental states consistent with the belief that not-p. More
specifically, Davidson suggests that the partition of the mind occurs
whenever a non-rational causality arises, i.e. whenever a mental state A
causes a mental state B without there being a (good) reason for B. For
example, Jack’s belief that Jane is in love with him is irrational in the
case where this is caused by his feelings for Jane (rather than objective
information), inasmuch as his desire to be loved by Jane is surely not
a good reason to believe that Jane is in love with him. If Jack’s irra-
tional belief stems from a genuine process of self-deception, and not
from sheer wishful thinking, this would mean that a compartment of his
mind believes that Jane is in love with him, while another compartment
of his mind believes the opposite. And in a sense, Davidson explains,
it’s precisely because Jack believes that Jane probably doesn’t love him
(and because he cannot bear that reality) that he is motivated to embrace
the belief that she does – the former belief causally inducing the later.
Despite the advantages of Davidson’s account relative to other di-
visionist models, it seems to encounter a new set of problems. The first
difficulty is that it undermines Davidson’s own holistic theory of mind,
which contends that an individual mental state can only be understood in
the light of the whole set of attitudes constitutive of the subject’s mind,
and that there must be a large degree of consistency between those atti-
tudes. Davidson (1982, p. 184) himself highlights the difficulty:
There is no question but that the precept of unavoidable charity in interpretation is

opposed to the partitioning of the mind. For the point of partitioning was to allow
inconsistent or conflicting beliefs and desires and feelings to exist in the same
mind, while the basic methodology of all interpretation tells us that inconsistency
breeds unintelligibility.
A second insufficiency lies in the fact that Davidson’s description of the

mind’s sub-systems remains too vague. As Pears (1991, p. 395) points
out, “[t]he drawing of a fault-line through a point at which internal
irrationality occurs is only the beginning of the theory, and it is not
enough to identify a sub-system”. As a matter of fact, Davidson states
that whenever irrational causality occurs in the mind one must assume
that a correlative partition is consumed; but he doesn’t specify how that
partition comes about, nor the terms in which the sub-systems co-exist
and interact. Besides, as Pascal Engel (1991, p. 16) rightly suggests, the
vagueness of Davidson’s analysis extends to the very dynamic underly-
ing the mind’s alleged differentiation: “[Davidson’s model] would need
the equivalent of what Freud calls the dynamic aspect of the uncon-
scious to account for the interactions between the sub-systems”. In oth-
er words, what Davidson’s model seems to lack is a descriptive account
of irrationality, rather than a merely normative one.
And, finally, many authors have questioned Davidson’s claim
that ordinary cases of irrationality, such as weakness of will and
self-deception, necessarily involve some sort of inconsistency between
the agent’s mental attitudes (which, in turn, would justify the claim
that such cases necessarily entail a partition of the mind). With regard
to weakness of will, authors such as George Ainslie and Jon Elster
160 Vasco Correia
convincingly argue that one needs not to conceive the phenomenon as

the result of an action contrary to the agent’s own judgment, as if the
action and the intention to act were strictly opposed. More plausibly,
what seems to happen when an agent yields to temptation in spite of a
previous decision is that he or she revises his or her initial judgment in
a sudden and ill-considered fashion under the pressure of an urging de-
sire. Thus, for example, my strong desire to eat a dessert is liable to af-
fect my judgment and lead me insidiously to revise the decision to lose
weight. Even if, all things considered, the option of losing weight might
be the one which maximizes my well-being, the craving for a more
immediate reward induces a biased perception of my preferences and
causes me to believe, for a short and decisive instant, that it is preferable
to eat the dessert. Ainslie (2001, p. 32) calls this cognitive effect “hyper-
bolic discounting bias”, which he defines as a sort of myopia regarding
preferences over time: caeteris paribus, people tend to overrate the val-
ue of immediate rewards and to underrate the value of future rewards.
According to this hypothesis, the agent who chooses to go on a diet and
eventually fails to act accordingly does not act against his or her own
judgment, strictly speaking, for at the time he or she eats the dessert his
or her (biased) judgment assesses that option as being preferable. In this
sense, there is no inner inconsistency in the agent’s mind (no simulta-
neously contradictory intentions to act), but simply an evolution in time
of the agent’s preferences; in other words, a revised decision, however
irrational and temporary it may turn out to be. The upshot of this analy-
sis is that divisionism is not required as a presupposition to understand
irrational action, given that a diachronic inconsistency of preferences
does not raise the paradox of contradictory attitudes.
Likewise, several philosophers and psychologists have argued that
self-deception does not imply some sort of inner inconsistency in the
subject’s mind either. Alfred Mele and Ariela Lazar, in particular, have
argued that self-deception does not entail the subject holding two con-
tradictory beliefs at the same time, but simply a hiatus between what the
available evidence clearly suggests, on the one hand, and what the sub-
ject is led to believe, on the other hand. In their view, the self-deceived
subject does not initially know the true proposition that p and at some
point decides to believe the false proposition that not-p. Instead, what
happens is that the subject’s judgment is biased by some desire (or other
emotion) which induces a distorted perception of reality. If Jack be-

lieves that Jane is in love with him in the teeth of evidence, for example,
the problem is not that Jack intentionally decides to believe what seems
more convenient, as Davidson suggests, but rather that his feelings for
Jane surreptitiously distort his interpretation of her attitudes toward
him. A simple sign of friendship may then appear to be an evidence of
love. But, again, if there are no contradictory beliefs in the agent’s mind,
if the irrational belief simply stems from a biased perception of reality,
there is no compelling reason to assume that his or her mind is divided
and that one part of the mind is aware of the truth while the other one is
being deceived. Furthermore, as Mele (1987, p. 121) observes, it would
be paradoxical to suggest that one part of the mind is able to employ a
deceptive strategy upon the other, given that a “potential self-deceiver’s
knowledge of his intention and strategy would seem typically to render
them ineffective”.
6. Conclusion: toward a unitary solution
In essence, our critical analysis brings about two significant conclu-

sions. First, that each version of the divisionist account leads to a cer-
tain number of paradoxes. This is not to say that the divisionist view is
intrinsically paradoxical, but simply that, to my knowledge, no coherent
model of divisionism has been developed so far. Second, and perhaps
more importantly, it would seem that it is not necessary to postulate
the division of the mind to be able to account for ordinary cases of ir-
rationality. As we have seen in the last section of this paper, most cases
of cognitive and practical irrationality can be accounted for in unitary
terms, without appealing to the divisonist postulate.
Crucially, my claim is that both irrational behaviour and irrational
thinking can be explained as the result of a conflict between individual
mental states, rather than the result of a conflict between differentiated
parts of the mind. We have seen this regarding cognitive cases of irra-
tionality, such as self-deception, denial, rationalization, wishful think-
ing, and the like, which arguably stem from the influence our desires
162 Vasco Correia
and emotions are liable to exert upon our judgments. And we have seen
it regarding practical cases of irrationality, and particularly akrasia (or
weakness of will), which can also be understood as the result of a bi-
ased judgment, given that the phenomenon of motivated irrationality is
also liable to affect the evaluative judgments upon which we base our
choices. Here too, it’s simply an individual mental state, generally a de-
sire, which induces an irrational assessment of the feasible options and
subsequently an irrational action.
Even though the hypothesis of mental partitioning is not required to
explain ordinary cases of irrationality, however, it may apply to extreme
cases of irrationality, and particularly to pathological cases of “disso-
ciative identity” (DSM IV) or “multiple personality disorder” (ICD 10).
To that extent, it would be wrong to rule out the divisionist hypothesis
as a paradox per se. Instead, it seems fair to suggest that divisionism
remains a puzzle to be solved. It would be interesting, in particular,
to examine what exactly makes the difference between “normal” and
“pathological” cases of irrationality with regard to identity disorders.
References
Ainslie, G. (2001). Break-Down of Will, Cambridge: Cambridge Uni-

versity Press.
Aristotle, (2009). The Nicomachean Ethics, transl. D. Ross, New York:
Oxford University Press.
Audi, R. (1982). Self-Deception, Action and Will. Erkenntnis, 18, 133–
158.
Bennet, J. (1990). Why is Believing Involuntary. Analysis, 50, 87–107.
Bird, A. (1994). Rationality and the Structure of Self-Deception. Euro-
pean Review of Philosophy, 1, 19–37.
Boag, S. (2007). Realism, Self-Deception and the Logical Paradox of
Repression. Theory and Psychology, 17(3), 421–447.
Bouveresse, J. (1991). Philosophie, mythologie et pseudo-science,
Combas : L’éclat.
Davidson, D. (1982). Paradoxes of Irrationality. In R., Wollheim, &

J., Hopkins (eds.), Philosophical Essays on Freud, Cambridge:
Cambridge University Press.
Davidson, D. (1985a). Deception and Division. In Actions and Events,
E. Lepore and B. McLaughlin (eds.), New York: Basil Blackwell.
Davidson, D. (1985b). Incoherence and Irrationality. Dialectica, vol. 39
(no. 4), 345–353.
Davidson, D. (1998). Who is Fooled? In Dupuy (ed.), Self-Deception
and Paradoxes of Rationality, Stanford: CSLI Publications, 1–19.
Elster, J. (ed.), (1985). The Multiple Self, Cambridge: Cambridge Uni-
versity Press.
Engel, P. (1991). Avant propos. In Davidson, Paradoxes de l’irrational-
ité, Éclat, Combas, 1–20.
Engel, P. (1999). Volitionism and Voluntarism About Belief. In A. Mei-
jers (ed.), Belief, Cognition and the Will, Tilburg: Tilburg Univer-
sity Press, 9–25.
Fingarette, H. (1984). Self-deception and the splitting of the ego. In
R., Wollheim, & J., Hopkins, (eds.), Philosophical Essays on Freud,
Cambridge: Cambridge University Press, 212–226.
Freud, S. (1920). Beyond the Pleasure Principle, trans. J. Reddick, Lon-
don: Penguin classics, 2003.
Freud, S. (1901). Psychopathology of Everyday Life, trans. A. Brill,
T. Fisher, London: Unwin, 1914.
Freud, S. (1923). The Ego and the Id, transl. J. Riviere, London, Hoga-
rth Press and Institute of Psycho-analysis, 1927.
Freud, S. (1938). The Splitting of the Ego in the Process of Defense,
trans. James Strachey, The Standard Edition, vol. XXIII, London:
Vintage, 1964, 271–279.
Freud, S. (1940). An outline of Psycho-analysis, The Standard Edition,
Vintage, London, 2001.
Johnston, M., (1989). Self-Deception and the Nature of Mind. In Rorty
and McLaughlin (eds.), Perspectives on Self-Deception, Berkeley,
L.A., London: University of California Press, 63–91.
Kenny, A., (1984). The Homunculus Fallacy. In The Legacy of Wittgen-
stein, Oxford: Blackwell, 125–136.
Mele, A. (1987). Irrationality: An Essay on Akrasia, Self-Deception and
Self-Control, N.Y., Oxford: Oxford University Press.
164 Vasco Correia
Mele, A (1998). Two Paradoxes of Self-Deception. In J.P. Dupuy (ed.),

Self-Deception and Paradoxes of Rationality, CSLI Publications,
Stanford, California, 37–58.
Pears, D. (1982). Motivated irrationality, Freudian theory and cognitive
dissonance. In Wollheim R. and Hopkins J. (eds.) Philosophical Es-
says on Freud, Cambridge: Cambridge University Press, 279–288.
Pears, D. (1984). Motivated irrationality, Oxford: Oxford University Press.
Plato (2003) Phaedrus, transl. S. Scully, Newburyport: Focus Philo-
sophical Library.
Plato (1992). Republic, transl. G.M.A. Grube, Indianapolis : Hackett
Publishing Company.
Sartre, J.-P. (1943). L’être et le néant, Paris: Gallimard.
Talbott, W. (1995). Intentional Self-deception in a Single, Coherent
Self. Philosophy and Phenomenological Research, 55, 27–74.
Thalberg, I. (1984). Freud’s Anatomies of the Self. In R., Wollheim,
& J., Hopkins (eds.), Philosophical Essays on Freud, Cambridge:
Cambridge University Press, 241–263.
Wittgenstein, L. (1953). Philosophical Investigations, trans. Anscombe,
Oxford: Blackwell.
Empirical and conceptual clarifications regarding
the notion of ‘Core-Self’ from Gallagher’s and
Merker’s Behavioural-Neuroscientific Proposals
João Fonseca
I. Introduction: Conceptual Confusions

and Methodological Fragmentation
The notion of Core-Self’ has been used and formulated by several au-
thors in different contexts and with different meanings, using diverse
methodologies and approaches ranging from traditional conceptual
analysis, Phenomenology, Cognitive Psychology and Neurosciences
(Damasio, 1999; Gallagher, 2000; Northoff & Bermpohl, 2004; Pank-
seep & Northoff, 2009). This conceptual and methodological fragmen-
tation threatens the very ambitions of the scientific task at stake.
As a result, the current scenario concerning the studies on the nature
of the (Core) Self presents us with a confusing profusion of concepts
that tends to obscure the intellectual endeavour itself. This problem has
recently been identified as such by some of the most prominent authors
in the field. (Gallagher & Zahavi, 2008; Praetorius, 2009). I share the
authors’ worries and also their diagnosis regarding the main causes of the
problem: the proliferation of different methodological and disciplinary
approaches to the subject. For instance, Gallagher & Zahavi state that:
This disparity, which is both problematic and productive, is directly related to

the variety of methodological approaches taken within philosophy and related in-
terdisciplinary studies of the self. They include introspection, phenomenological
analysis, linguistic analysis, the use of thought experiments, empirical research in
cognitive and brain sciences, and studies of exceptional and pathological behav-
iour. One problem to be posed in this light is whether different characterizations of
166 João Fonseca
self signify diverse aspects of a unitary concept of selfhood, or whether they pick
different and unrelated concepts. (2008, pp. 197–198)
Although originally restricted to the more general notion of Self, this

same observation applies neatly to what currently happens concerning
the more specific notion of ‘Core-self’.
In the recent literature, ‘Core-Self’ is loosely identified with the
more basic and primitive form of self-consciousness present in all con-
scious experiences (such as basic emotions or conscious cognitive func-
tions). But this is where the consensus among authors ends – there is
not one consensus even at this basic and loose level of definition. Some
authors suggest that the notion of Core-Self is a basic primitive (Za-
havi, 2009) where others consider it to be a complex-layered concept
(Panksepp & Northoff, 2009). To complicate this issue even more, some
authors go as far as to propose the elimination of the very notion of
‘Core-Self’ (Dennett, 1991; Metzinger, 2003, 2009).
The fundamental aim and motivation of this article is to bring some
conceptual clarification to the notion of ‘Core-Self ’ taking into consid-
eration what is stated in the last paragraph of the quotation above, i.e.,
“One problem to be posed in this light is whether different characteriza-
tions of self signify diverse aspects of a unitary concept of selfhood, or
whether they pick different and unrelated concepts.” I intend to develop
this line of questioning concerning the notion of ‘Core-Self ’. In order
to achieve this goal, and also for matters of simplicity, I will confront/
compare two proposals regarding the notion of ‘Core-Self’ by adopting
different methodologies: Bjorn Merker’s neuro-evolutionary perspec-
tive (2007) and Shaun Gallagher’s cognitive/phenomenological stance
(2000). I will try to show that, in relation to the aforementioned prob-
lem, that the answer consists of a conciliatory move where we can state,
without contradiction, that Merker’s and Gallagher’s characterizations
both point towards a unitary concept and, at the same time, select differ-
ent concepts of ‘Core-Self’. My strategy will fundamentally consist in
redefining these two proposals within my own taxonomical framework
for Behavioural Neuroscience and reformulating the notion of ‘Core-
Self’ as a Theoretical Concept in Behavioural Neuroscience within that
same framework.
Empirical and conceptual clarifications 167
II. Fundamentals of a Model-Theoretical Framework

for Behavioural Neuroscience
I will adopt a model/semantic approach to neuroscientific practice

(specifically, in the present context, that part of neuroscientific prac-
tice dealing with the explanation of behaviours, hence I will call this
kind of activity behavioural neuroscience – or BN for short). The se-
mantic view I adopt here has some connections with Ronald Giere’s
approach to physics (Giere, 1988, 1999). The semantic approach to
science endorsed by Giere focuses on scientific activity as a prac-
tice and how this practice is achieved and carried out by scientists as
human cognisers. Likewise other model/semantic approaches to sci-
ence are an explicit reaction against the Deductive-Nomological con-
ception. In particular, Giere is quite sensitive to the way mainstream
physics is academically communicated through textbooks (Giere,
1988). According to him, the main media of information displayed
by those books are idealised physical entities and systems that do not
exist in the world (e.g. frictionless motion). Given this, Giere’s pro-
posal is to consider those objects and systems as abstract entities that
constitute theoretical models of Classical Mechanics. These models
have the status of abstract cognitive entities capable of being repre-
sented in several ways: equations, linguistic descriptions, graphic rep-
resentations deployed by physicists. As Giere states: “[T]hey function
as ‘representation’ in one of the more general senses now current in
cognitive psychology. Theoretical models are the means by which sci-
entists represent the world – both to themselves and for others.” (1988,
p. 80). In this sense, models are interpreted intentional objects and not
just structures as sometimes proposed.1
1 Current literature on the model-theoretical approach to science is vast and diverse.

The word ‘model’ can mean very different things depending on the approach ad-
opted. Sometimes the term ‘model’ is used to refer to entities or practices that
help in the process of constructing explanation (as in the case of ‘scale models’).
Differently from this perspective, my usage of the term ‘BN models’ refers to the
explanations themselves. In this view, animal experimental models are part of the
evidence in constructing BN models conceived qua explanations and so are not
covered by my intended usage of the term ‘BN model’.
168 João Fonseca
Just like models of classical mechanics, models of behavioural neu-

roscience (BN models) are intentional objects, cognitive representations
used by scientists in several ways (here too, they can be deployed ver-
bally, graphically or in other ways). More precisely, BN models state
general explanations of behavioural phenomena. These models are
achieved by inductive abstraction from previous neuroscientific empir-
ical results. Usually, these intentional models are generalizations from
animal experimental findings. For instance, results from protocol exper-
iments of spatial memory in mice serve as a basis for constructing the
explanatory content of a BN model for spatial memory covering all the
animals that display this mental state, assuming the conservation across
species of the relevant structures (in this case, homologues of the CA1
area of the hippocampus in vertebrates). This important topic will be
addressed in more detail in section III. In structural terms, a BN Model
(Mbn) is a triple: Mbn = <B, f, M>, where:
‘B’ is a target behaviour or cluster of behaviours (e.g. the ac-
quisition of different behaviours that different animals adopt in or-
der to avoid a threat). B can group several behavioural phenomena
(different – and exhibited by different species – but probably simi-
lar under a certain explanatory perspective) under a unique general
concept – that fills ‘B’. For example, ‘The Acquisition of Avoidance
Behaviour’.
‘f ’ states the functional profile of the explanation (imposing con-
straints on M). f states that there must be a neural mechanism M such
that M explains B by stating the causal relation ‘M causes B’. Other
constraints stated on f directly or indirectly qualify the nature of this
causal relation. For instance, and depending on the explanatory target
described in B: M is domain specific on B, M is multimodal, etc. In gen-
eral, the causal claim, in conjunction with other constraints, qualifies
the causal relation ‘M causes B’ as being necessary and sufficient (with-
in the explanatory goal determined by the model itself). For instance,
taking B as ‘The Acquisition of Avoidance Behaviour’, f could state
that M is a neural structure such that M causes the acquisition of avoid-
ance behaviour, M is domain specific concerning avoidance behaviour
acquisition (i.e., its physical manipulation would only affect avoidance
behaviour acquisition – at least in the intended, and therefore, relevant
experimental protocols)2, M is input multimodal (there is one struc-

ture for all kinds of avoidance acquisition, independent of the sensory
modality involved in each particular case) , M is a memory/learning
structure, etc. This set of constraints states the functional profile of the
psychological concept ‘Fear Conditioning’ that explains ‘The Acqui-
sition of Avoidance Behaviour’. Assuming the principle according to
which ‘mental states cause/explain behaviours’ we can, more or less
safely, establish that f states the mental/psychological concept that sup-
posedly causally explains B.
‘M’ is a description of a neuronal mechanism satisfying the con-
straints expressed in f. Differently from f, M states a componential and
not just functional explanation of B. In more formal terms, a mechanis-
tic explanation M of a phenomenon B is a description of the (relevant
aspects of the) physical mechanism that produces/causes B. More spe-
cifically (and adopting Carl Craver’s terminology used in several places.
e.g. 2007) this explanation depicts a (finite) set of components (Φs) and
activities (Ωs) organised in such a way as to produce the role or effect
(Ψ). For example, concerning the functional profile of ‘Fear Condition-
ing’, a good candidate for a neural mechanist structure satisfying that
profile is the Central Amygdala, since a causal connection between this
structure and ‘the Acquisition of Avoidance Behaviour’ has been exper-
imentally established and it also satisfies the other constraints (i.e. it is
multimodal and domain specific).
It should be noted, however, that mechanistic explanation M can
assume different kinds of abstraction depending on the available knowl-
edge we have regarding a specific explanation. In this case, and follow-
ing Darden’s suggestion, we have different mechanism schemas which
are “truncated abstract descriptions of a mechanism that can be filled
with more specific descriptions of component entities and activities”
(Darden, 2002, p. 356). Darden calls this process ‘Schema Instantia-
tion’ and calls attention to the fact that these schema can come in differ-
ent degrees of abstraction.
2 This ‘Domain-Specificity Constraint’ is required in order to avoid attribution of

functions to any structure capable of affecting the target behaviour. For instance,
it would be an error to attribute the function of ‘fear conditioning’ to the eyes,
even if is true that a blind mice would not be capable of acquiring the avoidance
behaviour based on visual cues.
170 João Fonseca
Extensionally, Mbn range over a domain D of terrestrial multi-

cellular animals i.e. the objects of Behavioural Neuroscience research
are not counterfactually conceived extra-terrestrial or ‘possible world’
forms of life or even actual artefacts such as robots or any kind of arti-
ficial intelligent ‘device’.
As an important ‘final statement’, BN Models reveal the structure
of Theoretical Concepts of Behaviour Neuroscience, i.e. Psychological/
Mental Concepts. Psychological Concepts are Theoretical Concepts of
Behavioural Neuroscience because they refer to the putative underly-
ing/neuronal explanation/cause of B, where B is a Phenomenological
Concept of Behavioural Neuroscience, grouping several observable be-
havioural phenomena. We can summarize this by stating that: for any
Psychological Concept/Theoretical Concept of BN there is one corre-
sponding BN Model (with the general form ‘<B,f,M>’). BN Models
simultaneously reveal the structure of Theoretical Concepts of BN and
the meaning (and reference) of the corresponding theoretical terms.
Henceforth I will use the terms ‘BN Models’ and ‘BN Concepts’ inter-
changeably.
III. Hierarchical taxonomy of psychological concepts:

introducing ‘Nested Concepts’
A general and trivial feature of the family of concepts about a certain

topic or cognitive domain concerns the fact that the concepts can, in
a very intuitive and non-technical sense, include the cognitive content
of another concept. Take the non-problematic example of the concepts
‘Dog’ and ‘Animal’ respectively within a zoology framework. It is ob-
vious, in this case, that all the informational or cognitive content of
Animal is included in the concept ‘Dog’ and the same with the latter and
the concept ‘Labrador’. The extensional counterpart of this intensional
relation consists in the fact that all objects satisfying the open sentence
‘x is a dog’ also satisfy ‘x is an animal’ (that is: the set of dogs is a proper
subset of the set of animals). In this section I will claim that something
similar is also true for Psychological/BN Concepts.
Consider, for instance, what we could call the Psychological Do-

main Memory Consolidation. A model/concept ‘Declarative Mem-
ory Consolidation’ includes all the information of the concept/model
‘Memory Consolidation Simpliciter’. Accordingly, the set of creatures
satisfying ‘Declarative Memory Consolidation’ is a proper subset of
those satisfying ‘Memory Consolidation Simpliciter’. One interest-
ing consequence of this is a suggestion of a taxonomic hierarchy of
psychological concepts varying in their intended extensional scope and
informational content. Sometimes and in certain circumstances, psy-
chological/BN theoretic concepts are ‘nested’ in each other satisfying
the intentional and extensional conditions just suggested. That is the
case, I would suggest, with the concepts ‘Memory Consolidation Sim-
pliciter’ and ‘Declarative Memory Consolidation’: the first one is nest-
ed in the second since its informational content is included in the latter.
Being so, Memory Consolidation can be considered as an example of a
Psychological Domain determined by the nested relation between the
concepts ‘Memory Consolidation Simpliciter’ and ‘Declarative Mem-
ory Consolidation’ (and perhaps other concepts). All these notions of
‘informational content inclusiveness’, ‘subset extensional relation’ and
‘Psychological Domain’ are clearly in need of a more technical refine-
ment. In what follows I will explore the idea of ‘Nested Concept’ and
related notions in greater detail.
I should start with the fundamental claim that the proposed mod-
el/conceptual framework for Behavioural Neuroscience is constrained
by evolutionary considerations. Two of the most important are: (i) The
conservation of phenotype traces across taxa through homology and (ii)
(stated verbally) The ‘fact’ that Evolution does not start from scratch
(in other words: adaptive – if not other – requirements of transforma-
tion, modification or ‘specializations’ of certain biological functions are
achieved by ‘working upon’ the already existing phenotype structures
responsible for that function). (i) and (ii) are closely connected (in par-
ticular, (ii) assumes (i)). By assuming these principles we should be
able to answer the question: How do we establish that two different BN
concepts satisfy the ‘nested relation’?
Firstly, we should know how to settle the extension of a certain BN
concept. Regarding this question, neuroscientists (implicitly or explic-
itly) assume principle (i) as a rationale in their induction from targeting
172 João Fonseca
experimental animals models to class generalisations. Commenting on

an inference proposed by Squire concerning the neural structures un-
derlying ‘declarative memory’, John Bickle states that:
This approach forges a connection between human neuropsychological data and
experimental mammalian research. The ‘particular structures and connections’
namely the hippocampus proper, entorhinal cortex, perirhinal cortex, and peri-
hippocampal gyrus, have homologs across the mammalian class. Since declara-
tive (or explicit) memory is coextensive with hippocampal-requiring memory, the
term is applicable to memory research on humans, other primates and rodents.
(2003, p. 78)
In evolutionary theory, the term ‘Homologue’ means precisely a pheno-

typic trait conserved through species that share a common ancestor. The
answer for our question is that scientists extensionally induce f (in par-
ticular, the claim that ‘M causes B’, by both assuming structural homol-
ogy (e.g. the hippocampus) and functional homology (e.g., Declarative
memory) through a class of species belonging to a single evolutionary
clade (i.e., sharing the same common ancestor). In the case mentioned in
the quotation, the function Declarative Memory is inductively presumed
to exist in all the animals (mammals) that share the Hippocampus as a
common phenotypic trace conserved in all the members of the clade.
Taking this into consideration we should now ask: what principles gov-
ern the hierarchical relations comprising Nested Concepts? Taking again
the examples of ‘Memory Consolidation simpliciter’ and ‘Declarative
Memory Consolidation’, consider the following quotation from Squire and
Kandel that adopts evolutionary principle (ii) as an explicit rationale:
The simplest memory capabilities, and those that seem to have appeared earliest
in evolution, seem to be nondeclarative memories related to survival, feeding,
mating, defense and escape. As a variety of additional types of nondeclarative
memory and then declarative memory evolved, the new memory processes re-
tained not simply a set of genes and proteins, but entire programs for switching on
and stabilizing synapse connection. Moreover, these common mechanisms have
also been conserved through the evolutionary history of species: they are found in
both simple invertebrates such as Drosophila and Aplysia and complex mammals
such as mice. (1999, p. 155)
Two important and related questions arise here: the process by which the
information is shared (in a nested/inclusive relation) among BN concepts,
and how that relation is ‘translated’ in extensional terms. The sharing of

information in an inclusive way among models follows the answer to the
question of how the extension of a concept is inferred (conserved pheno-
type and functional traces through species) and adds to that which is stat-
ed in principle (ii): modifications (‘specializations’) of a certain function
are achieved by modifying the already existing structures responsible for
that function. This means that a certain structure responsible for a cer-
tain function is conserved in the more specific functional modifications
(this process is well attested in the above quotation). Regarding again
the above example concerning ‘Memory Consolidation simpliciter’ and
‘Declarative Memory Consolidation’, we can propose (in a simplistic
way) the current proposed neural mechanisms corresponding to each of
the two different types of memory as follows: ‘Memory Consolidation
simpliciter’ = ‘cAMP-PKA-CREB Molecular Pathway’ (see, for example,
Bickle, 2003); ‘Declarative Memory Consolidation’= Early to Late-Long
Term Potentiation switch in the Synapses of Hippocampus Cell’s. Fol-
lowing Squire and Kandel (1999), ‘Declarative Memory Consolidation’
is a function that was ‘built upon’ the evolutionarily conserved structures
corresponding to the more general and basic form of ‘Memory Consol-
idation simpliciter’, which underlies all kinds of Memory Consolidation
(be them ‘declarative’, ‘non declarative’, or any other specification on
what kind of memory is being consolidated – more on this below).
Therefore, the mechanistic abstraction M of <B,f,M> relative to
‘Declarative Memory Consolidation’ (i.e. E-LTP to L-LTP Switch at
the level of hippocampal synapses) has as one of its ones components
the ‘cAMP-PKA-CREB Molecular Pathway’ (specifically, the latter
corresponds to the ‘E-LTP to L-LTP Switch’ ‘part’ of the more complex
mechanism of ‘Declarative Memory Consolidation’). Simply stated, the
Mechanistic Schema of ‘Memory Consolidation simpliciter’ ‘is part of’
or ‘is filled in by’ the (in comparison) less abstract Mechanistic Sche-
ma of ‘Declarative Memory Consolidation’, i.e., the information stated
in the model of the former is included in that of the latter. Regarding
the behavioural component B of <B,f,M> for ‘Memory Consolidation
simpliciter’ it abstracts consolidation of behaviours regardless of being
declarative or non-declarative. The B for ‘Declarative Memory Consol-
idation’, on the other hand, retains the fundamental information from
the former and adds to it more informational specifications.
174 João Fonseca
As a matter of clarification, it should be emphasized that the Psy-

chological Concept/BN Model ‘Memory Consolidation Simpliciter’
does not correspond to a particular very simple kind of memory con-
solidation deployed by a very ancient and simple creature. On the con-
trary, the model/concept corresponds to an abstraction (a very abstract
Schema) that is satisfied by many different specific kinds of memory
consolidation (e.g.: Declarative) deployed by a large variety of species.
As Bickle says regarding the molecular mechanisms corresponding to
‘Memory Consolidation Simpliciter’:
[T]he molecular and gene composition and the intracellular pathway interactions
are shared across species. These shared features obtain despite vast differences
in brain size, organization, site of principal effect (presynaptic or postsynaptic),
behavioural repertoire, and even ‘cognitive logic’ of the distinct types of memory
being consolidated (declarative versus nondeclarative). (2003, p. 148)
The concept ‘Memory Consolidation Simpliciter’ is abstract to the

point that it can encompass a vast array of target behaviours and does
not specify if the proposed mechanism ‘cAMP-PKA-CREB Molecular
Pathway’ occurs pre or postsynaptically. This information can be added
to and made more fine-grained when we specify the particular kind of
memory consolidation we are seeking to understand (e.g.: Declarative
Memory Consolidation). Thus, we can state something like a general
intensional condition (GIC) regarding BN Nested Concepts:
GIC: if a concept C1 is nested in concept C2 and C1= <B,f,M> and

C2 = <B*,f*,M*>, then all the information stated in M and B is
included in M* and B*.
It is not difficult to translate the above line of reasoning to the question

of the extensional relation between nested models/concepts of BN. From
what was stated above, we can easily envision that the class of species
which are true of the open sentence ‘x has Declarative Memory Consoli-
dation’ is a sub-set of the class of species which are true of the open sen-
tence ‘x has Memory Consolidation Simpliciter’. Sustaining this inference
is the evolutionary condition stating that a class A of living creatures is a
subset of another class B, if the evolutionary clade corresponding to class
B is Basal in relation to the evolutionary clade corresponding to class A.
A clade is basal in relation to another clade if it contains that other clade

as a branch within it (both clades share a common ancestor).
For instance, whereas ‘Memory Consolidation Simpliciter’ is shared
by a vast number of species including invertebrates and vertebrates –
probably all multicellular animals (let us assume for the sake of the
argument), – ‘Declarative Memory Consolidation’ is possessed only by
mammals (and presumably other non-mammal vertebrates). In more
formal terms, the clade corresponding to ‘Memory Consolidation sim-
pliciter’ (i.e., all species possessing this trait are descendent from a
common ancestor also possessing the trait) is basal in relation to that
corresponding to ‘Declarative Memory Consolidation’ and, therefore,
the class of Mammals is a sub-set of the more general and comprehen-
sive class of Multicellular Animals.
We are now in position to state something like a general extension-
al condition (GEC) regarding BN Nested Concepts:
GEC: if a concept C1 is nested in concept C2, then the class of objects
of the domain satisfying ‘x is C2’ is a proper subset of the class
of objects of the domain satisfying ‘x is C1’.
From GIC and GEC we can now introduce the notion of ‘Psychological
Domain’ that mixes extensional and intensional parameters regarding
nested BN concepts. A Psychological Domain refers to the sets of actual
neural mechanisms and behaviours referred to by the abstractions stated
in B and M. A general definition for ‘Psychological Domain’ states:
Def.: A Psychological Domain PD is a set consisting of the union of a

set of actual neural mechanisms AM and a set of actual behaviours
AB such that, if C1= <B,f,M> and C2 = <B*,f*,M*>, and C1 is
nested in C2, then AM is the extension of M, AB is the extension
of B, AM* is the extension of M* and AB* is the extension of B*
(where AM* is a proper subset of AM and AB* is a proper subset
of AB – given GEC).
Departing from the definition of a Psychological Domain (PD) the Gen-

eral Intensional Condition (GIC) and the General Extensional Condition
(GEC), we are now in a position to formulate 3 structural conditions
regarding the nested relation between BN concepts. These conditions
176 João Fonseca
are structural in the sense that they imply empirical claims regarding the
causal structure of reality (in particular those parts of reality denoted by
the sets AM, AB, AM* and AB*). This means that for a particular case of
two BN concepts being nested there are empirical claims those concepts
have to be capable of verifying (at least in principle).
If C1 and C2 are BN concepts, if C1= <B,f,M> and C2 = < B*,f*,M*>
and C1 is nested in C2 defining an extensional Psychological Domain
PD, then C1 and C2 must conjunctively fulfil the following three struc-
tural conditions:
1. Necessary Structural Condition: M (of C1) states necessary struc-

tural (neural-mechanistic) conditions for AB (i.e., all the behaviours
comprising PD).
Empirical claim (regarding 1): if (by means of an ideal intervention)

all elements of neural mechanisms AM are physically inhibited, then all
the elements of the set of behaviours AB of PD will be compromised.
Example: Assuming that the concept ‘Memory Consolidation Sim-
pliciter’ (MCS) is nested within the concept ‘Declarative Memory Con-
solidation’ (DMC), then (by (1) and its empirical claim) if the M of MCS
(‘cAMP-PKA-CREB Molecular Pathway’) is inhibited, then all behaviours
of the Psychological Domain Memory Consolidation will be compromised.
In fact, several studies have shown that this is the case. Take, for instance,
the important and influential work by Bourtcholadze et al. (1994) where
two distinct kinds of learning (dependent on declarative/hippocampus and
non-declarative/amygdala memory respectively) were tested regarding
CREB knockout mice. John Bickle summarizes their results as showing
that “the loss of CREB function disrupts long-term memory for both con-
textual (declarative) and cued (nondeclarative) conditioning.” (2003, p. 90)
2. Sufficient Structural Condition: M* (of C2) states sufficient struc-

tural (neural-mechanistic) conditions for the set difference AB\AB*
(i.e., all elements/behaviours of AB that are not members of AB*).
Empirical claim (regarding 2): if (by means of an ideal intervention) all

elements of neural mechanisms AM* are physically inhibited, then all
the elements of the set of behaviours from the set difference AB\AB* of
PD will be compromised.
Example: Assuming again the MCS and DMC nested relation. This
condition states that if the M* of DMC (i.e., ‘E-LTP to L-LTP Switch
at the level of hippocampal synapses’) is disrupted, then only a sub-set
of behaviours from the Memory Consolidation psychological domain
will be compromised. Considering the general role of the hippocam-
pus in memory (assuming here for simplification that the consolidation
process is included in this more general picture) it is well known, since
the seminal observations of Patient H.M., that the hippocampus role in
memory is selective to what was then called Declarative (or Explicit)
Memory. As Squire (1992) puts it: “in the absence of the hippocampus
several other kinds of learning can still be accomplished, including the
learning of skills and habits, simple conditioning, and the phenomenon
of priming” (Squire, 1992, p. 196).
3. Unitary Mechanism Condition: Given GIC, all the information stat-

ed in M (of C1) must be included in M* (of C2) (i.e., M must be a
submechanism of M*). As a matter of causal mechanistic explana-
tion, if M is a submechanism of M*, then all members of AM* have
to be physically/causally linked to some members of AM described
by a mechanism schema M** in a way not stated in M or M*.
Empirical claim: we should be able to empirically detect the physical

components and activities regarding the neural mechanism M**.
Example: in the Memory Consolidation example this condition
states that there must be a physical causal link between the M of MCS
(‘cAMP-PKA-CREB Molecular Pathway’) and the M* of DMC (‘E-LTP
to L-LTP Switch at the level of hippocampal synapses’). The nature and
structure of such a link has been profusely studied as a matter of ex-
plaining what are the molecular and sub-cellular mechanisms underly-
ing the E-LTP L-LTP link. It is more or less well established how the
cAMP-PKA-CREB mechanism affects the nucleus of pyramidal cells
(via phosphorylation of CREB) and produces proteins that ultimately
affect the structure of the cell at the level of synapses. A detailed descrip-
tion of such a mechanism lies beyond our purposes here. All we need to
acknowledge is the existence of such a physical link between M of MCS
and M* of DMC as predicted by the application of this third condition.
We are now in a position to state the conditions two BN concepts
must satisfy in order to establish between them a nested relation and,
178 João Fonseca
henceforth, determine a Psychological Domain PD. A BN concept C1 is

nested in another BN concept C2 (i.e., they satisfy both the extensional
and intensional conditions stated in GEC and GIC and define a Psycho-
logical Domain PD) only if they satisfy the three structural conditions
stated above (Necessary Structural Condition, Sufficient Structural
Condition and Unitary Mechanism Condition).
IV. R
edefining Merker’s and Gallagher’s
proposals for Core-self
Having presented the fundamentals of the framework, my purpose now

is to consider both Merker and Gallagher’s proposals for ‘Core-Self ’.
Hence, I will rephrase both proposals, considering Core-Self as a The-
oretical Concept in Behavioural Neuroscience using the basic tools
summarized in the previous sections. The basic aim of this conceptual
clarification is to answer the question mentioned in the introduction,
posed by Gallagher & Zahavi (2008), according to which: “[o]ne prob-
lem to be posed in this light is whether different characterizations of self
signify diverse aspects of a unitary concept of selfhood, or whether they
pick different and unrelated concepts”.
IV.1 ‘Core-Self’ as a Theoretical Concept in

Behavioural Neuroscience
According to the framework, one could try to understand the notion

of ‘Core-Self’ as a Theoretical Concept in Behavioural Neuroscience.
Therefore, we can assume initially that ‘Core-Self’ corresponds to a BN
Model assuming the general structure: <B, f, M>. Given the particulari-
ties of both Merker and Gallagher’s proposals, I will start by addressing
the functional description f in the model of Core-Self.
Following on from what has been stated above, f expresses the
functional profile of the psychological concept in question. Besides the
‘mandatory clause’ stating that ‘M causes B’, other more specific clauses
are introduced. The most important property of Core-Self emphatically

stated by both Merker and Gallagher concerns the cognitive ability of
basic ‘Decision Making’. The most important trait of Core-Self, ac-
cording to these authors, consists of a basic subjective substratum capa-
ble of discriminating objects and events in the world and judging them
accordingly in terms of the actions to take on each occasion. Important
for this is what both authors call ‘sense of agency’ as a fundamental
characteristic of Core-Self; Merker says that “[Core-self] enters con-
sciousness only in the form of motivated agency” (2005, p. 96), whereas
Gallagher & Zhavi equate Core (or Minimal)-Self with an “experiential
sense of agency” (2008, p. 160). This fundamental property of Core-
Self is shared by other authors and is encapsulated in the following
quotation from Panksepp and Northoff: “We envision the core-self to
allow organisms to be active agents” (2009, p. 160).
We will describe later how both Merker and Gallagher approach
this idea in a more detailed way. For now, we can try to define the func-
tional profile f of ‘Core-Self’ as stating the following clause (in addition
to the ‘mandatory’ clause stating: ‘M causes B’): ‘M allows purposive
voluntary non-reflexive action through a subjective sense of agency (M
is the substratum for basic – immediate – decision making)’.
Given the task of ‘building’ a BN Model for ‘Core-Self’ based
upon Merker and Gallagher’s proposals, I have to risk proposing what
the ‘bare’ target behaviour B should be, since neither Merker nor Gal-
lagher are explicit concerning this point. Assuming that (as stated
above) the basic functional trait of ‘Core-Self’ is the capacity for basic
decision-making (through a subjective sense of agency), I risk suggest-
ing that the target behaviour B of the model for Core-Self consists of
‘Operant/instrumental conditioned behaviour’. The following quotation
commenting on the nature of ‘Operant Conditioning’ is clear enough in
establishing the link between the fundamental functional trait of core-
Self (decision making) and this kind of behaviour:
The distinction between Pavlovian and operant conditioning therefore rests on

whether the animal only observes the relationships between events in the world (in
Pavlovian conditioning), or whether it also has some control over their occurrence
(in operant conditioning). Operationally, in the latter outcomes such as food or
shocks are contingent on the animal’s behaviour, whereas in the former these occur
regardless of the animal’s actions […] The scientific study of operant conditioning
180 João Fonseca
is thus an inquiry into perhaps the most fundamental form of decision-making. It

is this capacity to select actions that influence the environment to one’s subjective
benefit, that marks intelligent organisms. (Staddon & Niv, 2008, Emphasis added)
So, it is safe enough to suggest that, in the BN Model for ‘Core-Self ’,

the respective target behaviour B states: ‘General (i.e., all kinds of) op-
erant/instrumental conditioned behaviours’. The functional profile of
this model states that these behaviours are caused by a neural mecha-
nism M that allows purposive voluntary non-reflexive action through a
subjective sense of agency. As such, Merker and Gallagher’s proposals
agree (by assumption) both on the target behaviour B and the functional
profile f of the putative model for Core-Self. Where they disagree sharp-
ly is on the third member of the ordered triple <B,f,M> of the model for
Core-Self. That is, they propose very distinct neural mechanisms M for
implementing what is stated in f. A closer look at each proposal is now
necessary with regard to the neural implementation of Core-self.
IV.2 Merker’s upper brainstem proposal
What follows is an over-simplification of Merker’s detailed proposals and

suggestions. The aim of this section is to briefly present Merker’s per-
spective concerning the neural mechanism implementing ‘Core-Self’.
Merker’s view regarding Core-Self is an adaptative-evolutionary
one. In his view, the emergence of consciousness (and Self) was nature’s
response to two different (although deeply related) evolutionary pres-
sures: 1- The Stability Problem and, 2- The Decision Making Problem.
In what follows I will briefly present each pressure separately and show
what evolution’s solution to each one has been.
Let us focus on the first evolutionary pressure The Stability Prob-
lem. This problem is well illustrated by what Merker calls The ‘Earth-
worm Dilemma’:
Earthworms display a swift withdrawal reflex to cutaneous touch. […] Consider

the worm’s initiation of a crawling movement. Such a movement will produce
sudden stimulation of numerous cutaneous receptors, yet no withdrawal reflex is
released to abort the movement. (2005, p. 90)
The solution to this dilemma (how to distinguish self-produced

from other-produced sensory inputs) has a very simple solution in the
case of worms by simply distinguishing excitatory vs. inhibitory syn-
apses related to self-produced movement. The solution to this problem
has, of course a high survival value. The real problem is not with worms
but rather concerns how more complex bodies solve this same problem/
dilemma. Here is the problem as stated by Merker:
The evolution of higher animals leads not only to increased complexity of sin-
gle sensory and motor systems but produces a diversification of such systems
in the equipments of a given species. Vision, hearing, touch, smell, taste, entro-
ception, proprioception and vestibular systems are some of those on the sensory
side, while a great variety of locomotors, orienting, grasping, and manipulatory
appendages – often paired in sets and with multiple, independently moving joints –
proliferate on the motor side. Such diversity brings special problems in its train.
On the sensory side, for example, the receptor arrays of different modalities are
often disposed in different parts of the body. They are therefore, not affected in
the same way by self-motion and so cannot be subject to the same compensatory
remedies, nor be integrated directly. (2005, p. 92)
In a nutshell, there is a complexity problem at the level of sensorimotor

integration in vertebrates given the several distinct sensorial modalities,
the myriad of different possible body actions or the fact that the different
sensorial transducers are located in different parts of the body, there-
fore being affected differently by the same movements (the ‘orienta-
tion problem’). There appears to be too many scattered sensorial-motor
variables that, if kept separately, would prevent the distinction between
self-produced from other-produced sensory inputs. What is needed here
is an Integration (for action) of different sensorial and motor ‘local solu-
tions’. We need a single coordinate reference system in order to frame
all this proliferation and diversification.
So, what was evolution’s ‘solution’ to this problem? According to
Merker, nature’s structural answer relates to the phylogenetically very
old and widely conserved structures corresponding to the basic Verte-
brate Brain Plan, namely the roof/tectum of the midbrain/Superior Col-
liculus (S.C.). Merker stresses the peculiar anatomical and functional
roles of this structure:
182 João Fonseca
[The superior colliculus] is the only site in the brain in which the spatial sens-
es are topographically superposed in laminar fashion within a common, pre-
motor, framework for multieffector control of orienting. Its functional role
appears to center on convergent integration of diverse sources of information
bearing on spatially triggered replacement of one behavioural target by anoth-
er (2007, p. 67).
The S.C. laminar structure is such that some information registered in

one layer (say a mapping of auditory space) translates directly to another
layer (say, a mapping of oculomotor orientation) by means of coordinate
transformation. This is why when we hear a sound, our eyes ‘know’ im-
mediately where to look in the search for the source of the sound.
What this means is that the S.C. laminar structure in fact brings the
various local sensory-motor solutions under a common referential system
framework. The end result is that when someone moves his or her own
body (the eyes for instance) the world seems to remain stable. Consider
now what happens if you slightly tap your eyeball. You will notice that the
world seems to be moving independently. This is so because your eyes
move slightly without you intentionally doing this. The layer in the SC re-
sponsible for a retinotopic map of stimulus registers change whereas the
layer responsible for the oculomotor map does not. There is a mismatch
in the coordinate transformation of information from one layer to another.
When there is a coordinate transformation from a motor map layer
to a sensorial map layer, the result is a sense of an external stabilized
world. According to Merker this is, therefore, nature’s solution to the
evolutionary pressure of the stabilization problem in more complex
(vertebrate) bodies.
Let us turn now to the second selection problem faced by creatures
with a complex body and movement; The Decision Making Problem.
According to Merker, the evolution of real time decision making is the
result of the merging of three distinct biological selection domains: 1-
Needs/Action Selection: ‘time axis’ – what action should be taken now
and next; 2- Opportunities/Target Selection: ‘space axis’ – which targets
in the world should be ‘chosen’; 3- Motivational Ranking: some targets
should be avoided while others should be approached.
Again, how did evolution ‘solve’ the problem of integrating these
three selection domains into one single decision space? According to
Merker, it was by answering this question that consciousness arose
in biological systems, since it provides an optimal solution to the

decision-making problem. The question, of course, is ‘how’.
The answer is somewhat complex. Therefore, in order to explain
Merker’s views more easily I will divide them into three different el-
ements. The first element for the answer relates directly to the previ-
ous solution to the first evolutionary pressure, i.e., the multimodal
sensory-motor common reference framework at the level of SC. Accord-
ing to Merker, as we have just shown, this provides a phenomenological
contrast between an external stabilized world and my own intended
actions. But this is an Abstract/Formal contrast. It should be stressed
that this ‘Common Reference System Framework’ does not depend on
particular content – it is, so to speak, a built-in formal framework. n a
Kantian vein, we can conceive this as providing a kind of pure intuition
of sensibility, i.e., a spatio-temporal formal aspect of conscious expe-
rience. However, as Kant taught us, forms without content are empty.
So, if the first element deals with the formal aspect of consciousness,
the second element regards the content aspect of that formal frame.
More particularly, this is the domain of conscious ‘representations’ ly-
ing within the formal frame. Two major domains of representation are
of interest here: the representation of the external world in terms of its
objects and events (what Merker calls an ‘analogue representation of
the world’) corresponding to the target selection axis of the decision
problem. The other pole to take into consideration regarding conscious
content representations is a particular and special object in the world:
the body (what Merker calls the ‘analogue representation of the body’)
that corresponds to the action selection side of the decision problem.
Finally, the third element relates more directly to the phenomeno-
logical aspect of consciousness. It consists of sensations/feelings/emo-
tions signalling the motivational ranking of targets in the world, serving
as bias in the decision process.
We have, therefore, the merging, through consciousness, of the
three selection domains: 1 action selection – perceived (analogue) body;
2- target selection – perceived (analogue) world: and 3- motivational
ranking- emotional feelings, within a single frame conceived for opti-
mal real time decision-making.
Considering now the neural implementation of this functional-
ly described solution to the Decision Problem, Merker’s suggestion is
184 João Fonseca
that biologically evolved structures at the level of the upper brainstem

constitute the structural ‘side’ of this solution: close interconnection
between 1- Superficial layers of SC (the multimodal shared integrated
spatial coordinate system – ‘space-time sensory world’); 2- Substancia
Nigra/Basal Ganglia (motor commands – Body); 3- PAG/Hypothalamus
(motivation – feelings/emotions). These three poles: the multi-modal/
reality, the hedonic/motivational and the motor commands, functionally
and structurally interconnected, provide the basics for a (phenomenic)
conscious decision making process.
According to Merker, this complex structure is capable of imple-
menting a ‘virtual’ Core-Self in the form of an implicit ‘ego-center’
subjective ‘decision-maker’:
There is reason to believe that the implicit “ego-center” origin of this coordi-
nate space is the position we ourselves occupy when we are conscious […] The
ego-center places the conscious subject in an inherently “perspectival”, view-
point-based, relation to the contents of sensory consciousness. (2007, p. 72)
However, Merker suggests a possible specific neural implementation

of (Core) Self, turning it, so to speak, more ‘real’. Merker’s hypothesis
concerns the possible role of the Zona Incerta (or Thalamus Ventralis
in non-mammals) in the resolution of remaining decision making prob-
lems not ‘completely solved’ by the ‘superior culliculus – PAG / hypo-
thalamus – Substancia Nigra/Basal Ganglia’ complex:
[a] sophisticated control system needs some means for resolving residual conflict
among alternatives left unsettled by routine mechanisms because of stochastic
happenstance in a complex multicomponent system or because of exceptional
combinations of contingencies encountered in a lively and unpredictable world.
[…] At the level of the core control system the Zona Incerta […] seems optimally
placed to play such a role (2005, p. 103).
The Zona Incerta is also an (sub-cortical) upper-brainstem structure

massively connected with the other structures (sensorial and emotion-
al) of the upper brainstem. According to Merker, this structure and its
above mentioned putative functions could ’supply consciousness with
that subjective presence of a tacit “witness” to its contents which is at
the heart of the very concept of consciousness. These characteristics
would amount, in other words, to a subjective sense of self, not in the
sense of self-image – which is a content of consciousness – but in the

sense of the nodal monitoring function presupposed by all contents of
consciousness’ (2005, pp. 104–105).
Independently of the two possibilities; a ‘virtual’ arising from the
‘superior culliculus – PAG /hypothalamus – Substancia Nigra/Basal
Ganglia’ interaction, or a ‘real’ one directly implemented in the Zona
Incerta, what matters for the present purposes is that Merker’s proposal
concerning the neural implementation of the ‘Core-self’ points to an
upper-brainstem structure, evolutionarily conserved across vertebrates,
predating the emergence of neo-cortical structures in evolutionary more
recent creatures (namely mammals).
Therefore, with regard to the BN Model for Core-Self and accord-
ing to Merker’s proposal, the neural mechanism ‘M’ could be stated,
generally and loosely, as consisting of the Upper-brainstem complex
(more specifically in the ‘superior culliculus – PAG /hypothalamus –
Substancia Nigra/Basal Ganglia – (and possibly) zona incerta’ complex.
IV.3 Gallagher’s cortical motor forward model proposal
Shaun Gallagher’s departure concerning the neural implementation of

core-self rests on his phenomenological definition of this concept. For
Gallagher, the notion of Core-self or, as he prefers, the minimal-self,
is defined firstly, at a more general level, as “consciousness of oneself
as an immediate subject of experience” (2000, p. 15) and secondly, at a
more specific level as the conjunction of two phenomenic experiences:
the sense of agency and the sense of ownership. Both ‘senses’ (of agency
and ownership) are conjunctively necessary conditions for Intentional
Action and Decision Making. More specifically, the sense of agency cor-
responds to the experience that I am in control (I am responsible) of my
action; this makes the action in question to be considered a voluntary
and intentional one. On the other hand, the sense of ownership corre-
sponds to the experience that a certain movement is mine, be it voluntary
or not: even in a pure reflex movement I can report it as being mine.
Gallagher deploys a control engineering schematic approach cur-
rently used in contemporary cognitive neuroscience (see, for instance
Grush, 2004; Wolpert, Mial and Kawato, 1998). The phenomenon under
186 João Fonseca
consideration is motor control and Gallagher’s aim is to locate the roles

of the senses of ownership and agency in the schema. Although complex,
the fundamentals of this system are encapsulated in the following: “[A]
comparator mechanism operates as a part of a non-conscious premotor
or ‘forward model’ that compares efference copy of motor commands
with motor intentions and allows for rapid, automatic error corrections
[…] This comparator process anticipates the sensory feedback from
movement” (Gallagher, 2000, p.16).
Two different control models make up the overall mechanism: the
forward model and the feedback model. This distinction is important
since Gallagher suggests that the sense of agency corresponds to the
forward model – by means of a sense of anticipation of the movements
comprising an intentional action – and the sense of ownership corre-
sponds to the feedback model – the feedback from actual movements
allows the agent to identify it as movements from his or her own body.
Given this, the question arises what about the neural implementation
of the mechanism responsible for intentional action and, in particular, re-
garding the two component models (forward and feedback)? Departing
from cognitive neuroscience empirical data, Gallagher suggests that
the main areas relating to the forward model and hence, to the sense of
agency, correspond to the Motor, Pre-Motor and Prefrontal areas of the
neo-cortex. On the other hand, the feedback model (responsible for the
sense of ownership) seems to be implemented in the Cerebellum.
Gallagher illustrates this hypothesis with empirical studies con-
cerning schizophrenia:
Schizophrenic patients who suffer from thought insertion and delusions of control
also have problems with this forward, pre-action monitoring of movement, but
not with motor control based on a comparison of intended movement and sensory
feedback. The control based on sensory feedback is thought to involve the cerebel-
lum. By contrast, problems with forward monitoring are consistent with studies
of schizophrenia that show abnormal pre-movement brain potentials associated
with elements of a neural network involving supplementary motor, premotor and
prefrontal cortexes. Problems with these mechanism might therefore result in the
lack of a sense of agency that is characteristic of these kinds of schizophrenic
experience. (2000, pp. 16–17)
In fact, schizophrenics report that their actions are not under their com-
mand and will; they seem to lack a sense of agency. The phenomenological
failure in feeling their actions as caused by themselves promotes de-

lusional reasoning concerning the real element ‘responsible’ for their
actions: ‘I’m not causing my movements, someone else is’. Since the
sense of agency is a necessary condition for a ‘normal’ core or minimal
self to arise, this illness can be considered, under these current assump-
tions, as resulting from a ‘truncated’ core/minimal-self.
Thus, if according to Gallagher the core/minimal-self can be iden-
tifiable with the conjunction of the sense of agency with the sense of
ownership and if we accept the proposed suggestion for the neural iden-
tification of each one of these phenomenic experiences, we can identify
the ‘M’ in the BN Model for ‘Core-Self’ as: ‘Motor, Pre-Motor and
Prefrontal areas of the neo-cortex plus Cerebellum’.
V. Conciliating Merker and Gallagher’s proposals:

‘Core-Self’ as a Theoretical Nested Concept in
Behavioural Neuroscience.
V.1 The ‘Explanatory Dilemma’
We have just testified Merker and Gallagher’s disagreement concerning

the proposed neural implementation of Core-Self. If we start by assum-
ing that both proposals are empirically correct, then, as I claim, we face
a kind of ‘Explanatory Dilemma’: either 1. the two proposals select the
same concept and the claim of Multiple Realization of ‘Core-Self ’ is
vindicated, or 2. the two proposals select different and unrelated con-
cepts. The explanatory ‘dilemma’ (so to speak) consists of the fact that
either of the two possibilities implies a fragmentation in our explanato-
ry practices where some kind of unity would be preferable. In this sec-
tion, I will address this problem and show how it is possible to suppress
this. In particular, I will try to show how, in the present context, the two
proposals relate to each other in a way not predicted by the rationale
implicit in the formulation of the dilemma.
188 João Fonseca
The first step in addressing the dilemma consists of challenging

the Multiple Realization (MR) claim. The MR claim assumes that the
two proposals have exactly the same functional profile as stated (within
the framework adopted here) by the ordered pair <B,f>. The fact that
both proposals (assuming that they are both empirically correct) only
disagree in the neural realization of the same functionally described
profile of ‘Core-Self’ seems to support the MR claim that this particular
concept is differently realized in different species or individuals. I will
try to show that this claim, in fact, is not vindicated; while maintaining
the assumption that both proposals are empirically correct, I will show
that the two proposals differ in their functional profile and, as a natural
consequence, they select different concepts.
The question that naturally arises concerns in What and How the
two proposals differ in their functional claims. In order to answer this
crucial question we must look closely at what both authors claim con-
cerning the functional properties of their respective proposals.
As we described in section III, both proposals address the mecha-
nisms (cognitive and neural) underlying basic purposive/voluntary mo-
tor actions. In this particular, and at this general level, both proposals
seem strikingly similar. It is my contention that the key for detecting a
fundamental difference between the two proposals consists in looking at
their proposed solutions rather than their (apparently similar) problems.
Recalling Merker’s proposal, we have seen that, fundamentally, his
suggestion consists of the integration of various sensory modalities,
body representation and emotional values within a common coordinate
system at the upper brainstem. At a more abstract/functional descrip-
tion, Merker’s solution consists of the evolved common state space in-
tegrating all the other domains (body, world and emotions) otherwise
kept separate. In other terms, and assuming the risk of great simplifica-
tion, his proposal can be understood as a dynamic system where some
variables change their values while others remain constant when some
motor action takes place. According to some authors, this constant in-
teraction of values is sufficient for the emergence of a subjective sense
of an external world distinct from a self responsible for body agency in
that world (see, for instance, Philipona et al, 2003). In its fundamentals,
Merker’s solution is computationally very simple – for instance, the
‘sensorial multimodal – body’ interaction taking place at the roof of
the superior culliculus assumes the form of a simple direct mapping

look-up table resulting in a pre-motor framework for (among other
functions) body orientation in space.
If we take Merker’s proposed computational solution as a refer-
ence point of departure, we can testify that Gallagher’s proposal is, from
a computational/control theory point of view, much more complex.
Whereas Merker’s suggestion is not, at its core, much more complex
than a simple look-up table mapping input-output, Gallagher introduc-
es a more complicated scenario with looping and re-entrance informa-
tion-processing culminating with the idea of a motor forward-model.
Basically, the role of a motor forward-model (FM) consists of the re-
finement of voluntary movements, in particular rapid ones. At its essence,
the FM compares the relation between the intended state of the body (e.g.
The position of my arm in relation to a certain target) with its actual state.
‘Running’ an FM allows error correction before the actual execution of the
movement. The end result is a smooth and precise rapid voluntary move-
ment. The resulting ‘sense of agency’ is a by-product of this FM.
There are other cognitive features of the FM worth mentioning not
addressed previously but important within the current context. A very
important cognitive property of systems with the architecture of an FM,
as several authors have noticed, consists of the fact that, when running
off-line (independently of actual motor movements) the FM is capable
of produce motor imagery (i.e., the capacity of imagining movements
without actually making them) (see Churchland, 2002; Grush, 2004).
This imaginary capacity is, nevertheless, important for actual behav-
iour in the environment since “[o]ff-line imagining yields an advanced
peak at the likely consequences of pondered actions, which permits
undesirable consequences of contemplated actions to be foreseen and
avoided” (Churchland, 2002, p. 81).
How do these cognitive features and properties of an FM relate
to Merker’s upper-brainstem proposal? An important clue is given in a
note in Merker’s article for Behaviour and Brain Sciences. In that note
Merker states the following:
To avoid possible misunderstandings of this key point, note that the analogue “re-
ality simulation” proposed here has nothing to do with a facility for simulating
things such as alternate course of action by, say, letting them unfold “in imagi-
nation” or any other versions of an “inner world,” “subjective thought,” “fantasy”
190 João Fonseca
or the like. Such capacities are derivative ones, dependent upon additional neural
structures whose operations presuppose those described here. (2007, p. 81)
This remark is crucial in the present context since it establishes a fun-

damental cognitive difference between the two proposals. From what is
stated in this quotation, it becomes clear that Merker’s proposal is, at a
cognitive level, simpler than Gallagher’s. Gallagher proposes a cogni-
tive scenario, presupposed by the FM, where ‘imagery’, ‘off-line plan-
ning’ and other cognitive properties are the fundamental traits. These
traits are, as we have described, absent in Merker’s perspective, which
is much simpler, and provides the fundamentals for these more complex
cognitive tasks to arise. If we take this cognitive difference seriously,
then we can conclude, after close scrutiny, that the cognitive profile of
both proposals is not the same after all.
Things being so, I feel comfortable enough to address the MR claim
as initially proposed. It is clear, or so I hope, that the MR claim regarding
the notion of ‘Core-Self’ is thus seriously called into question. Recall
that, in addition to the assumption that both proposals are empirically
correct, the MR claim rests on the supposition that those two proposals
have exactly the same functional profile as stated by the ordered pair <B,
f>. This is precisely the conclusion I have shown to be false. So, the MR
claim is not vindicated. However, another consequence of this conclusion
is that each proposal picks a different concept of ‘Core-Self’. But this
leads us to the second horn of the ‘dilemma’ since this ‘solution’, just like
the MR one, has the unwelcome consequence of an explanatory fragmen-
tation regarding the notion of ‘Core-Self’. The splitting into two different
concepts does not seem to help us. We are back to the original problem
(this is what constitutes the ‘explanatory dilemma’ after all).
V.2 Core-Self as a ‘Psychological Domain’ of BN. Introducing

the ‘Nested Concept Hypothesis’
My suggestion in order to avoid the hindrance described above (what

I have dubbed, for want of a better term, the ‘Explanatory Dilemma’)
consists of proposing the following hypothesis: Merker’s and Gallagh-
er’s proposals correspond to two different concepts of ‘Core-Self’ but
related to each other in a ‘nested ‘way that corresponds to a single Psy-

chological Domain. This hypothesis, if confirmed, would solve the ‘ex-
planatory dilemma’ since it would grant us, at the same time, two differ-
ent concepts of ‘Core-Self’ but inter-related in a way which is capable
of providing the desired explanatory unification.
Having stated the hypothesis, we are faced now with the challenge
of testing it. Following what was established in section II, there are
several characteristic properties and constraints that two BN concepts
have to satisfy in order to be related in a single nested way. These sev-
eral principles and constraints are encapsulated in the three structural
conditions stated at the end of section II.
As a start, we have to assume that the two concepts of Core-Self,
corresponding to each of the two proposals, are related in such a way that
one of the concepts (call it ‘Core-Self1’ provisionally) is nested in the
other (‘Core-Self2’). What we do not know is the following: which pro-
posal (Merker’s or Gallagher’s) corresponds to which concept (‘Core-
Self1’ and ‘Core-Self2’). As we have seen in section II, there are two
basic conditions two BN concepts must satisfy in order to be considered
nested in each other: the General Intensional Condition (GIC) and the
General Extensional Condition (GEC). For several technical reasons
(not addressed here), an extensional difference is easier to determine and
quantify than an intensional one. Therefore, we should explore the re-
quired extensional difference between the two proposals first.
Let us recall what was stated in the general extensional condition
(GEC) regarding BN Nested Concepts:
GEC: if a concept C1 is nested in concept C2, then the class of objects
of the domain satisfying ‘x is C2’ is a proper subset of the class
of objects of the domain satisfying ‘x is C1’.
In this respect, it is not very difficult to establish and justify the exten-
sional difference. Although Gallagher is not very explicit about the real
intended extensional scope of his proposal, nevertheless, taking into
consideration the fact that the neural implementation implies a relative-
ly developed neo-cortex (including pre-frontal areas) we could, reason-
ably, establish the scope of his proposal as varying from: only humans
at minimum, or: all mammals at maximum. On the other hand, Merker
is very explicit concerning his proposed domain of ‘application’: all
192 João Fonseca
vertebrates. As he says in relation to the neural structures implementing

his proposal of core-Self: “[t]hese structures together are proposed to
form the essential scaffolding of the mechanism of consciousness in
vertebrates.” (2005, p. 106) Merker justifies this statement by relying
on evolutionary structural conservation (homologies) across taxa. The
upper brainstem complex comprising Merker’s proposal, like other neu-
ral structures, belongs to a common ‘brain plan’ homologous across
vertebrates (2005, p. 68). Things being so, the set corresponding to the
intended scope within domain D of Merker’s proposal is wider than
Gallagher’s; even assuming the maximal perspective, presumably the
set of all mammals is a proper subset of all vertebrates.
Let us recall the evolutionary criterion for subset relation among
animal classes: a class A of living creatures is a subset of another class
B if the evolutionary clade corresponding to class B is Basal in relation
to the evolutionary clade corresponding to class A. In accordance with
this rule, the clade corresponding to the extensional domain of Core-
Self1 should correspond to a clade basal in relation to the clade corre-
sponding to the extensional domain of Core-Self2. Things being so, we
can establish that Core-Self1 corresponds to Merker’s proposal whereas
Core-Self2 corresponds to Gallagher’s, given that Vertebrates corre-
spond to a basal clade in relation to Mammals. What is more, therefore,
Merker’s concept/proposal is (in principle) nested in Gallagher’s.
For taxonomic and explanatory reasons we have to classify each
proposal/concept by ‘naming’ them. Taking into consideration that
Merker’s proposal (‘Core-Self1’) appears to address some more funda-
mental aspects of Core-Self ranging over a wide class of vertebrate spe-
cies, I suggest naming his proposal as ‘Core-Self Simpliciter’. On the
other hand, since we have seen that Gallagher’s proposal (Core-Self2),
based on the implementation of a forward-model, sustains a series of
more elaborate cognitive abilities, my personal suggestion is to call it
‘Cognitive Core-Self’.
So, for now, the nested concept hypothesis gains some strength
since ‘Core-self Simpliciter’ and ‘Cognitive Core-Self’ satisfy the GEC
condition for being considered related in a ‘nested’ way. It remains to be
shown that the Three Structural Conditions are also satisfied.
Starting with the Necessary Structural Condition, it should be re-
called that it stated that the M (in this case, of ‘Core-Self Simpliciter’)
comprises necessary structural (neural-mechanistic) conditions for the

set of all the behaviours of the related Psychological Domain (in this
case of the Core-Self domain). In the present context this means that,
if ‘Core-Self Simpliciter’ is nested in ‘Cognitive Core-Self ’, the neural
structures corresponding to the former necessarily underlie the later.
In fact, we were given just a glimpse of this situation when we were
searching for a cognitive difference between Merker’s and Gallagher’s
proposals in sub-section IV.1 above. There, a quote from Merker’s ar-
ticle form Behavioural and Brain Sciences, announced that his views
were cognitively different and simpler than certain cognitive competen-
cies such as ‘imagery’ and ‘off-line planning’ supposedly sustained by
a more complex Forward-model. But, in addition to the cognitive/func-
tional difference, Merker stated that: “[s]uch capacities are derivative
ones, dependent upon additional neural structures whose operations
presuppose those described here.” (2007, p. 81, emphasis added)
According to the empirical consequence associated with this con-
dition, if the neural structure corresponding to the concept which is
nested in another is disrupted, all the behaviours within the psycho-
logical domain are completely disabled. There is disparate empirical
evidence that an analogous situation is true in the present context of
‘Core-Self’. A particular line of evidence is important here: the situa-
tion where the upper-brainstem (the neural structure corresponding to
‘Core-Self Simpliciter’) is physically impaired. An important contribu-
tion favouring this line of evidence is given by the neuroscientists Jaak
Panksepp and Georg Northoff when presenting a proposal of (Core)Self
very similar to the one suggested by Merker. Panksepp and Northoff
(2009) depart from the seminal experiments conducted by Bailey and
Davies in the early nineteen forties to stress how fundamental the role
of some structures of the upper-brainstem complex is, namely the PAG,
on sustaining any purposive or intentional behaviour. This conclusion is
driven by the results of Bailey and Davies showing that:
complete damage of the lowest midbrain integrator for emotionality, namely the
PAG, compromises all world directed action tendencies in animals. Such animals
lose all intentionality, and seem to be barely conscious; this highlights how se-
verely compromised are both agency and ownership of experience when the most
fundamental emotional-affective SELF-generating functions of the brain are de-
stroyed. (2009, p. 204, emphasis added)
194 João Fonseca
Merker arrives at the same conclusion, although through different

empirical sources:
Cases of human collicular damage are rare, but in one such case whose damage
extended into the midbrain tegmentum (Denny-Brown, 1962) the symptoms were
severe, consisting of profound apathy with “no evidence of recognition of people
or any event in her surroundings.” This points to the potential importance of the
extended midbrain damage in determining outcomes – a point also underscored
by the findings on macaques in the same report – and to the significance of spon-
taneous, self-initiated behaviour and reactivity in this regard. (2005, p. 101)
Generally, as Parvizi and Damasio point out: “damage to the upper-

brainstem is a known cause of coma and persistent vegetative state”
(2000, p. 1351, emphasis added).
What can be concluded from these several sources is that a selec-
tive impairment in the upper-brainstem complex (sometimes at specific
structures such as the PAG or the Superior Culliculus), and maintaining
the cortical apparatus intact, seem to disable any kind of purposive be-
haviour (the behavioural target of ‘Core-Self’. Panksepp and Northoff
even mention explicitly that, in particular, the senses of agency and
ownership, the main phenomenological characteristics of ‘Cognitive
Core-self’, are disabled when those more fundamental structures com-
prising ‘Core-Self Simpliciter’ are physically impaired.
All these data and results point towards a very probable vindication
of the Necessary Structural Condition regarding the supposed nested
relation between the concepts ‘Core-Self Simpliciter’ and ‘Cognitive
Core-Self’. In accordance with this, it has been shown that the set cor-
responding to all behaviours of the psychological domain Core-Self
(i.e., all purposive behaviours) are compromised if the set of all the
neural mechanisms corresponding to ‘Core-Self Simpliciter’ (i.e., the
upper brainstem) is physically disabled.
Regarding the Sufficient Structural Condition, in the present context
its empirical claim basically states that if the neural structures corre-
sponding to ‘Cognitive Core-Self’ (namely the cortical areas implement-
ing a motor forward model) are physically inhibited, then just a subset
of the set of all behaviours of the psychological domain Core-Self (the
elements of the set of behaviours from the set difference AB\AB* of Core-
Self) will be compromised. It is time now to explore the line of evidence
resulting from damage to the cortex while sparing the upper-brainstem.
We could start with Gallagher’s own point regarding situations where

the cortical implementation of a forward-model is impaired. Following
Gallagher, in these cases what is observed is that the phenomenological
‘sense of agency’ is disabled and, as a consequence, the subjects report
that it is as if ‘alien’ or ‘outside agents’ were the causes of their own vol-
untary actions. This kind of phenomenological report is typical of sub-
jects diagnosed with schizophrenia. Nevertheless, in these cases, what we
are describing is not a complete loss of the self but a sort of a truncated or
distorted experience of it. There is a self which is disabled in some of its
properties, not a complete displacement (this is well attested in some ver-
bal reports by schizophrenic patients: “I’m not causing my movements,
someone else is”). Even more impressive are the various empirical results
showing a total functional independence of the upper-brainstem from the
neo-cortex. In this respect, Merker notes that:
All of the behaviors just mentioned have also been exhibited by experimental
animals after their cerebral cortex has been removed surgically, either in adult-
hood or neonatally. Best studied in this regard are rodents (Wishaw 1990; Woods
1964). After recovery, decorticate rats show no gross abnormalities in behavior
that would allow a casual observer to identify them as impaired in an ordinary
captive housing situation, although an experienced observer would be able to do
so on the basis of cues in posture, movement, and appearance (Whishaw 1990, on
which what follows relies, supplemented by additional sources as indicated). They
stand, rear, climb, hang from bars, and sleep with normal postures (Vanderwolf
et al. 1978). They groom, play (Panksepp et al. 1994; Pellis et al. 1992), swim, eat,
and defend themselves (Vanderwolf et al. 1978) in ways that differ in some details
from those of intact animals, but not in outline. Either sex is capable of mating
successfully when paired with normal cage mates (Carter et al. 1982; Whishaw
& Kolb1985), though some behavioral components of normal mating are missing
and some are abnormally executed. Neonatally decorticated rats as adults show the
essentials of maternal behavior, which, though deficient in some respects, allows
them to raise pups to maturity. Some, but not all, aspects of skilled movements
survive decortication (Whishaw & Kolb 1988), and decorticate rats perform as
readily as controls on a number of learning tests (Oakley 1983). Much of what is
observed in rats (including mating and maternal behavior) is also true of cats with
cortical removal in infancy. They move purposefully, orient themselves to their
surroundings by vision and touch (as do the rodents), and are capable of solving a
visual discrimination task in a T-maze. (2007, p. 74)
The above extensive citation (justified by the profusion of important

examples and reference sources) suggests that only a subset of all
196 João Fonseca
purposive behaviours (i.e., a subset of the set of behaviours corre-

sponding to the psychological domain Core-Self) are disabled by
the physical impairment of the neural structures corresponding to
‘Cognitive Core-Self ’ (the cortical areas implementing a motor for-
ward model). Just like the necessary structural condition, the sufficient
structural condition seems to be vindicated in this particular case.
Finally, the Unitary Mechanism Condition, applied to the present con-
text, states that there should be a neural mechanism neither stated in ‘Core-
Self Simpliciter’ (upper brainstem complex) nor in ‘Cognitive Core-Self’
(cortical implementation of a Forward Model) that causally links both.
We have just described above how the upper-brainstem complex
(the neural mechanism corresponding to ‘Core-Self Simpliciter’) seems
to consist a structural necessary condition for more specific and fine-
grained concepts/models of ‘Core-Self’ (namely ‘Cognitive Core-
Self’). It remains, however, to address the question concerning How the
neural mechanisms of ‘Core Self Simpliciter’ and ‘Cognitive Core-Self ’
relate to each other in a physical way.
Panksepp and Northoff (2009), although departing from a slightly
different perspective from the one adopted here, address this point di-
rectly. They do this by introducing additional cortical structures into the
picture, namely: the cortical midline structures (such as the posterior
cingulated cortex, the dorsomedial prefrontal cortex, or the orbitomedi-
al prefrontal cortex). Regarding Gallagher’s proposal of the implemen-
tation of a motor forward-model in the motor, pre-motor and prefrontal
areas of the cortex, they note (based on several empirical findings) that
these cortical areas “are closely connected with midline regions” (2009,
p. 204). According to Panksepp and Northoff, the forward-model cor-
tical areas are physically connected with the sub-cortical upper-brain-
stem region via these midline structures (2009, p. 204).
Therefore, according to Panksepp and Northoff’s suggestion,
the neural mechanism M* described in the model/concept ‘Cognitive
Core-Self’ would consist of the ‘filling’ of the schematic M deployed
in the model for ‘Core-Self Simpliciter’ (upper-brainstem complex) via
the inclusion of the information stated in the mechanism schema M**
(projections to certain cortical midline structures plus its connections
to the areas of motor and pre-motor cortex implementing the forward-
model). This means that the Unitary Mechanism Condition is satisfied.
At the beginning of this section I suggested that the concepts of

Core-Self involved (Merker’s and Gallagher’s) could relate in a nested
way and, therefore, define a single Psychological Domain in order to
overcome the Explanatory Dilemma described in the previous section.
In order to vindicate this relation the two concepts had to conjunctively
satisfy the three structural conditions presented earlier. At this point,
therefore, the two concepts can be said to be related in a nested way
(Core-Self Simpliciter is nested in Cognitive Core-Self) since they sat-
isfy the three structural conditions. If they are indeed related in a nested
way, they define a single Psychological Domain, then we could simply
classify this as the Core-Self domain. It remains to show how this nested
relation between the two concepts is capable of solving the so-called
Explanatory Dilemma.
V.3 Explanatory Unification and the Overcoming of the Dilemma
Before directly addressing the solution of the explanatory dilemma, it

is important to consider the important claim concerning the promised
‘explanatory unification’ of the ‘Nested’ concept of Core-Self, since, as
stated above, the satisfaction of this claim is instrumental in the over-
coming of the dilemma. Here I will just scratch the surface of this topic
and advance some suggestions for further empirical research.
In the previous section it was shown that ‘Core-Self Simpliciter’
and ‘Cognitive Core-Self’ satisfy the three structural conditions in or-
der to be able to be considered as nested concepts. At the same time, the
notion of a single Psychological Domain covering Core-Self was also
vindicated. This means that a certain level of unification was gained,
in particular a structural/neuronal explanatory unification. This unifica-
tion, among other things, provides us with a clearer view regarding the
neural structure of the notion of Core-Self as a whole (i.e., as a single
Psychological Domain).
In parallel to this unification at the structural/neuronal level, one
at the functional/cognitive level is also suggested. The very functional
characterization of both notions is in need of further clarification along
with their mutual dependence (as predicted by the ‘nested hypothesis’).
We have already shown in section IV.1 that Merker’s proposal (renamed
198 João Fonseca
‘Core-Self Simpliciter’) and Gallagher’s (‘Cognitive Core-Self ’) dif-

fer in some of their basic cognitive properties. In particular, it was
established that whereas Gallagher’s proposal, based on the implemen-
tation of a forward-model, is capable of more complex cognitive fea-
tures such as imagery, Merker’s is simpler and lacks that specific kind of
sophistication. What remains is to understand these differences in more
detail within a context that frames both.
The above three structural conditions are not only useful in clar-
ifying the neural mechanistic relation between the two concepts of
Core-Self. A more clear and fine-grained determination of what both
concepts are really about (i.e., what precisely their cognitive content
is) also comes about as an important consequence brought about by the
nested conceptual relation between the two. For instance, by applying
the Sufficient Structural Condition we can state with greater precision
which behaviours (B*) are supposed to be explained by ‘Core-Self Sim-
pliciter’ (<B,f,M>) and ‘Cognitive Core-Self’ (<B*,f*,M*>) respective-
ly. According to the Sufficient Structural Condition there is a difference
between B and B* (formally captured by the set difference AB\AB* of
the Psychological Domain Core-Self). The question is: what exactly is
this difference? If the target behaviour of the initially proposed unique
model of ‘Core-Self’ corresponded to ‘purposive behaviour’, then the
B of ‘Core-Self Simpliciter’ is supposed to correspond to that same tar-
get of ‘purposive behaviour’ generally (without further specification).
On the other hand, ‘Cognitive Core-Self’ is expected to causally ex-
plain only a particular sub-set (B*) of the more inclusive set of general/
non-specified purposive behaviour (B). It is hard to state firmly what
specific behaviour ‘Cognitive Core-Self’ is supposed to explain since
Gallagher is far from clear regarding this issue. Nevertheless, recasting
the idea of ‘explanatory unification’, some progress is legitimately to
be expected. In fact, Panksepp and Northoff once again provide some
interesting clues within the context of explanatory unification. In the
following quotation, they start by addressing the functional difference
and dependence and end by extrapolating to the behavioural level:
Functionally, higher cortical regions might represent the functions that are pri-
marily represented subcortically in a more detailed, specific and refined way. [...]
Though both complex sensory and motor functions are already represented in
subcortical regions like the PAG, basal ganglia, mesencephalic locomotor system,
those primitive sensory-motor functions are again represented in cortical regions,

e.g., somatosensory and motor cortices. This may allow behavioural functions to
be elaborated in more detail, with specific types of flexibility, especially with much
greater regulation of and inhibition over instinctual outputs, than it is possible on
the subcortical level. (2009, p. 201, emphasis added)
So, an obvious dependence relation between the behavioural/cognitive

profiles of ‘Core-Self Simpliciter’ and ‘Cognitive Core-Self ’ is again
suggested and vindicated. At the same time, that relation is inferred and
clarified by linking the supposed behavioural differences to both cogni-
tive and neuronal differences, interactions and dependencies. A unified
research programme into motor functions, as suggested by Panksepp
and Northoff, could clarify in greater detail, which specific target be-
haviours (B and B*) correspond to ‘Core-Self Simpliciter’ and ‘Cogni-
tive Core-Self’. This clarification, within a unified framework, would
also reformulate both the functional profiles (f and f*) and the neural
mechanisms (M and M*) of both concepts. These considerations show,
at the very least, how explanatorily fruitful it is to consider ‘Core-Self ’
as a single Psychological Domain where the concept ‘Core-Self Sim-
pliciter’ is considered to be nested within ‘Cognitive Core-Self ’.
Given the above considerations we can address, and easily solve,
the initial ‘Explanatory Dilemma’. Recall the initial formulation of the
‘dilemma’: assuming both proposals for the neural implementation of
‘Core-Self’ (Merker’s and Gallagher’s) there are two distinct and incom-
patible possibilities: 1. The two proposals choose the same concept and
Multiple Realization (MR) is vindicated, or 2. The two proposals pick
different and unrelated concepts. The ‘explanatory dilemma’ consisted
in the fact that, independently of which possibility one adopts, the end
result would consist of an explanatory fragmentation over a desired uni-
fication. The nested concept solution for ‘Core-Self’ avoids this dilemma
by, at the same time, overcoming MR and radical fragmentation implicit
in solution 2. The ‘nested solution’ overcomes MR since it claims that,
contrary to the assumption in the argument for MR, the two proposals
select two different concepts (with different grains of explanation and
different extensions in the domain D). On the other hand, it avoids radical
explanatory fragmentation since the two concepts are inter-related and
partially overlap each other both intensionally and extensionally (they are
related in the sense that they belong to the same Psychological Domain
200 João Fonseca
Core-Self). This explanatory unification between the two concepts was

clearly shown in the preceding paragraphs of this section.
Therefore, by adopting the nested concept solution, we overcome
the explanatory dilemma and achieve explanatory unification and clari-
fication regarding the notion of ‘Core-Self’ by gathering the two initial
proposals under a common (and empirically vindicated) explanatory
framework.
Conclusion
At the beginning of this paper we showed the methodological fragmen-

tation brought about by the current profusion of scientific approaches
concerning ‘Self’/‘Core-Self’ (neuroscience, psychology, philosophy,
among others). The main concern was that a methodological fragmen-
tation could lead to (and sometimes does) a related conceptual frag-
mentation. As Gallagher and Zahavi state: “One problem to be posed
in this light is whether different characterizations of self signify diverse
aspects of a unitary concept of selfhood, or whether they pick different
and unrelated concepts.” (2008, p. 198). I tried to address this question
directly by focusing on two different proposals: Merker’s evolutionary
perspective and Gallagher’s phenomenological one. By deploying my
proposed model-theoretic framework for theoretical concepts of behav-
ioural neuroscience, I tried to uncover some of the underlying principles
sustaining both proposals. By using the instrumental and fundamental
notion of ‘Nested Concept’ I tried to show how the two proposals could
relate to each other. Finally, I suggested a possible answer to Gallagher
and Zahavi’s anxiety: Merker’s and Gallagher’s proposals point, simul-
taneously and without contradiction, towards a unitary concept of Core-
Self and to two different concepts. This ‘nested solution’ proved to be
both empirically and conceptual vindicated, and explanatorily fruitful
(even suggesting avenues for further empirical research).
I hope I have shown that, for explanatory and conceptual reasons, it is
useful to address the notion of ‘Core-Self’ as a theoretical concept of be-
havioural neuroscience. By assuming such an approach we gain conceptual
and taxonomic clarification, explanatory richness, and bridges, both con-

ceptual and empirical, between different scientific disciplines and practices.
References
Bard, P. & Rioch, D. M. (1937). A study of four cats deprived of neocor-

tex and additional portions of the forebrain. Bulletin of the Johns
Hopkins Hospital, 60, 73–148.
Bickle, J. (2003). Philosophy and Neuroscience. Dordrecht: Kluwer Ac-
ademic Publishers.
Bickle, J. (2006). Reducing mind to molecular pathways: explicating
the reductionism implicit in current cellular and molecular neuro-
science. Synthese, 151, 411–434.
Bjursten, L.-M., Norrsell, K. & Norrsell, U. (1976). Behavioral rep-
ertory of cats without cerebral cortex from infancy. Experimental
Brain Research, 25, 115–30.
Bourtcholadze, R., Frenguelli, F, Blendy, J., Cioffi, D., Shultz, G., &
Silva, A. (1994). Deficient long-term memory in mice with a tar-
geted mutation of the Camp-responsive element binding protein.
Cell, 79, 59–68.
Carter, C. S., Witt, D. M., Kolb, B. & Whishaw, I. Q. (1982). Neonatal
decortication and adult female sexual behaviour. Physiology and
Behavior, 29, 763–66.
Churchland, P. S. (2002). Brain Wise, Cambridge MA: MIT Press.
Craver, C. (2007). Explaining the Brain. Oxford: Oxford University Press.
Damasio, A. (1999). The Feeling of What Happens: Body and Emotion
in the Making of Consciousness. New York: Harcourt Brace.
Darden, L. (2002). Strategies for discovering mechanisms: schema
instantiation, modular subassembly, forward/backward chaining.
Philosophy of Science, 69, S354–S365.
Dennett, D. (1991). Consciousness Explained, Boston MA: Back Bay
Books.
Denny-Brown, D. (1962). The midbrain and motor integration. Pro-
ceedings of the Royal Society.
202 João Fonseca
Gallagher, S. (2000). Philosophical conceptions of the self: Implica-

tions for cognitive science. Trends in Cognitive Science, 4, 14–21.
Gallagher, S. & Zahavi, D. (2008). The Phenomenological Mind: An
Introduction to Philosophy of Mind and Cognitive Science. New
York: Routledge.
Giere, R. (1988). Explaining Science. Chicago IL: University of Chi-
cago Press.
Giere, R. (1999). Science Without Laws. Chicago IL: University of Chi-
cago Press.
Grush, R. (2004). The emulation theory of representation: motor con-
trol, imagery, and perception. Behavioral and Brain Sciences, 27,
377–442.
Kogan, JH., Frankland, PW. & Silva, AJ (2000). Long-term memory
underlying hippocampus-dependent social recognition in mice.
Hippocampus, 10, 47–56.
Merker, B. (2007). Consciousness without a cerebral cortex: a challenge
for neuroscience and medicine. Behavioral and Brain Sciences,
30(1), 63–81.
Merker, B. (2005). The liabilities of mobility: A selection pressure for
the transition to cortex in animal evolution. Consciousness and
Cognition, 14, 89–114.
Metzinger, T. (2003). Being No One: The Self-Model Theory of Subjec-
tivity. Cambridge MA: MIT Press.
Northoff, G. & Bermpohl, F. (2004). Cortical midline structures and the
self. Trends in Cognitive Sciences, 8, 102–107.
Oakley, D. A. (1983). The varieties of memory: A phylogenetic ap-
proach. In Mayes (ed.) Memory in Animals and Humans, 20–82.
Van Nostrand Reinhold.
Panksepp, J., Normansell, L., Cox, J. F. & Siviy, S. M. (1994). Effects of
neonatal decortication on the social play of juvenile rats. Physiolo-
gy and Behavior, 56, 429–43.
Pankseep, J. & Northoff, G. (2009). The trans-species core SELF: The
emergence of active cultural and neuro-ecological agents through
self-related processing within subcortical-cortical midline net-
works. Consciousness and Cognition, 18, 193–215.
Parvizi, J. & Damasio, A. (2001). Consciousness and the brainstem.

Cognition, 79, 135–159.
Pellis, S. M., Pellis, V. C. & Whishaw, I. Q. (1992). The role of the cor-
tex in play fighting by rats: Developmental and evolutionary impli-
cations. Brain, Behavior and Evolution, 39, 270–84.
Philipona, J. K. & O’Regan, J. P. (2003). Is there something out there?
Inferring space from sensorimotor dependencies. Neural Compu-
tation, Vol. 15, No. 9, 2029–2049.
Praetorius, N. (2009). The phenomenological underpinning of the no-
tion of a minimal core self: a psychological perspective. Conscious
Cognition, 18(1), 325–38.
Squire, L. & Kandel, E. R. (1999). Memory: From Mind to Molecules,
New York: Scientific American Library.
Staddon, J. & Niv, Y. (2008). Operant Conditioning, Scholarpedia, 3(9),
2318. <http://www.scholarpedia.org/article/Operant_conditioning>.
Vanderwolf, C. H., Kolb, B. & Cooley, R. K. (1978). Behavior of the rat
after removal of the neocortex and hippocampal formation. Jour-
nal of Comparative and Physiological Psychology, 92, 156–75.
Whishaw, I. Q. (1990). The decorticate rat. In B. Kolb & R. C. Tees
(eds.), The Cerebral Cortex of the Rat, 239–67. Cambridge MA:
MIT Press.
Whishaw, I. Q. & Kolb, B. (1985). The mating movements of male de-
corticate rats: Evidence for subcortically generated movements by
the male but regulation of approaches by the female. Behavioural
Brain Research, 17, 171–91.
Wolpert, DM., Mial, RC. & Kawato, M (1998). Internal Models in the
Cerebellum. Trends in Cognitive Science, vol. 2.
Woods, J. W. (1964). Behavior of chronic decerebrate rats. Journal of
Neurophysiology, 27, 635–44.
Zahavi, D (2009). Is the self a social construct? Inquiry, 52(6), 551–573.
Part IV
Ontology and Taxonomy
Core Self and the Problem of the Self
Jorge Gonçalves1
The concepts of consciousness and self have been central in contempo-

rary philosophy of the mind. Inevitably, this has led to the recuperation of
a few conceptions from classical Phenomenology, starting with Husserl.
This is the case with the concept of “pre-reflective self-consciousness”.
The approach of these philosophers (Zahavi, 2007; Gallagher, 2007;
Parnas, 2010) is not existential, but what could be called “biological”
in the sense that they considered consciousness and self as natural phe-
nomena, explained scientifically. One of the problems that these philos-
ophers intended to resolve is the renowned problem of the self that was
initially formulated by David Hume and more recently by Metzinger,
among others. In this lecture, I intend to argue, using the facts regarding
the origin of the self in the early stages of life as a foundation for the
argument, that the concept of pre-reflective self-consciousness does not
resolve this problem. In spite of the facts not being conclusive, I will
affirm that there is good reason to reject the idea that the entire form of
phenomenal consciousness assumes a feeling, no matter how little, of
self.
According to these authors, all consciousness is simultaneously
self-consciousness. For example, when I look at a landscape, in the mo-
ment I am looking there is always a feeling that it is I who is looking.
This feeling could be reflective, when I carry out an introspection of my
mental states with the view that I am studying consciousness. However,
for this feeling of self to exist, it is not necessary to be in a thematic
reflective state. Even without it being reflective, consciousness always
implies a sense of self.
We could succinctly divide phenomenal consciousness into two
dimensions. One of them is known as “what is it to be like”, a term
1 Fellowship funded by the Portuguese Foundation for Science and Technology.

208 Jorge Gonçalves
that was popularized which comes from a celebrated article by Thomas

Nagel. We can roughly say that it describes the states that imply to
feel.2
The other dimension is of the first person. The phenomenological
states always imply, according to the referred authors, a subject who is
minimally conscious. There are no phenomenological states that can be
felt being disconnected from a feeling of the self.
An example referred to by Sartre, is of a person who is concen-
trated on reading. She is not thematically conscious that she has a self,
because she is absorbed in the object she is reading. However, if she is
asked what she is doing, she immediately responds that she is reading.
This fact demonstrates, according to the defenders of phenomenal con-
sciousness, that she was always conscious of herself reading, but in a
pre-reflective way, the proof of this being in the immediacy with which
she responds to the question.
This minimal feeling of self has been designated as the core-self or
minimal self. In Gallagher’s words, “Minimal [core] self: phenomenologi-
cally, that is, in terms of how one experiences it, a consciousness of oneself
as an immediate subject of experience, unextended in time. The minimal
self almost certainly depends on brain processes and an ecologically em-
bedded body, but one does not have to know or be aware of this to have an
experience that still counts as a self-experience.” (Gallagher, 2000).
This core self is always elusive, it never really becomes an object,
and is never thematic. The Buddhists use metaphors to describe it: it is
like a flame that illuminates objects, but never illuminates itself. The
authors maintain it is impossible to present the self before ourselves as
an object, that could be observed.
However, they also claim that the I/self is not a pure empty place,
as Kant conceived it to be. The core self is endowed with a certain mass
of a “phenomenological material”. This “phenomenological material”
is invisible, meaning that it is not observable using the methods of the
third person.
One of the philosophical problems the authors wish to see resolved
with the notions of self-reflective pre-consciousness and core-self is, as
2 For example, “What is it to be like on the road and getting a new bed almost every
night?”
Core Self and the Problem of the Self 209
we referred to earlier, the so-called problem of the self, as was famously

presented by David Hume.
But self or person is not any one impression, but that to which our several impres-
sions and ideas are supposed to have a reference. If any impression gives rise to
the idea of self, that impression must continue invariably the same, through the
whole course of our lives; since self is supposed to exist after that manner. But
there is no impression constant and invariable. Pain and pleasure, grief and joy,
passions and sensations succeed each other, and never all exist at the same time.
It cannot, therefore, be from any of these impressions, or from any other, that the
idea of self is derived; and consequently there is no such idea.3
When I show my own self to myself, I represent it as if it were an entity

with a given unity and continuance in time. This representation raises
the problem of knowing what kind of entity it is. For many authors the
self does not exist; it is nothing more than an illusion.
Metzinger (2009) recently elaborated a theory of the self where
he argues that we are an illusion, an illusion for no one, in fact, the
“we” of the previous phrase does not exist. There is no entity with these
described characteristics. “We” are no one. Despite these statements
seeming counter-intuitive, Metzinger defends them in a convincing way
combined with philosophical perspectives and scientific data. He does
not argue that the feeling of the self does not exist. He considers it as a
phenomenological datum. What he defends is that phenomenology it-
self is an illusion, just like the other illusions he explains with his model
of the mind. It the case of the rubber-hand illusion (Metzinger, 2009).
The subject is put in a position such that his hand is hidden and he
looks at an artificial hand as though it were his real one. His real hand is
touched at the same time as the artificial one. In a few seconds, this pro-
duces the illusion for the subject that his real hand is the artificial one.
Phenomenologically the subject feels that his hand is the artificial one,
but this is an illusion because he has a real hand of his own flesh and
blood. In the same way, the experience that a self exists, an independent
entity capable of existing on its own, having an essence, or rather a set
of invariable intrinsic properties, and is provided with individuality, is
an illusion.
3 In A Treatise of Human Nature, by David Hume (1739), Book I, Part 4.

The difference is that in the case of the partial illusions we are

conscious, while in the case of the self we are, ourselves, the illusion,
and in this sense we do not have the possibility of perceiving that we are
illusions. Metzinger explains this situation by stating that the self has
a transparent property, in the phenomenal sense. The self understands
itself as a reality because it does not know the processes from which
it originates; it does not know the reality where it is an illusion itself.
In order to understand this better, we can think about another case of
transparency: visual perception. When I look at a pen, I think I am look-
ing directly at an object just as it is in the objective exterior world. I do
not comprehend this object as a result of representational and neuronal
mechanisms. If I do not give in to a reflective process, I am in the posi-
tion of naïve realism or the belief that reality is just as it appears to me.
According to Metzinger, organisms are equipped with representa-
tional mechanisms that result from the course of biological evolution due
to the diverse advantages that they bring. Some of these representational
mechanisms represent the organism to itself. Metzinger develops an en-
tire complex theory, from which here I will present only the conclusions.
What interests me here refers to what the phenomenological self, what
we feel, is the momentary content of an auto-representational mechanism
of the human body. This auto-representation has evolutionary advantages
and is not, because of this, an epiphenomenon. To be conscious of our
own self, we are completely unaware of the representational mechanisms
from which we originate. We are transparent in this sense: we comprehend
ourselves directly without paying attention to the mediatory mechanisms.
Metzinger draws out the following results from this biological situation:
we are no one, we think that we exist as a reality but in truth we do not
exist because we are like shadows in Plato’s cave. Moreover, Metzinger
himself uses this thousand-year-old image4. Neurology is reality, the self
is a shadow of this reality that we take to be reality. The ontological status
of the self is that of the illusion. From the scientific point of view, the
self is a concept to be eliminated because it does not notice the reality it
is trying to describe. The experience that I have of the self is not the real
thing, and thus I live in a state of auto-illusion.
4 Metzinger, T., Being No One. The Self-Model Theory of Subjectivity, MIT Press,

Cambridge, MA, 2003 p. 547.
Metzinger’s ontological conclusion that the self does not exist has
been criticized in the field of the philosophy of mind by various au-
thors5 who assert the Phenomenology tradition of Husserl and other
philosophers, his followers such as Heidegger, Sartre, Merleau-Ponty,
and Michel Henry. This new Phenomenology has a naturalist charac-
ter, different from that of the aforementioned authors whose problems
were existential more than anything else. The basic critical idea is that
Metzinger, stating that the self does not exist, has a specific concep-
tion of self. Despite Metzinger reporting the most recent scientific dis-
coveries in neuroscience, he continues to hold a traditional conception
of the self, which is conceived of as an image of material things, as
being something equipped with spatiality and permanence. Another
conception defended by these authors is that of calling the self some-
thing more basic and not differentiating it from the flow in phenomenal
consciousness. In this conception, the self is reduced to its minimal
level. According to Gallagher’s definition, “phenomenologically, that is,
in terms of how one experiences it, a consciousness of oneself as an
immediate subject of experience, unextended in time.”6 The concept of
self as we just defined it is broadened in time. However, these authors
maintain that if we take away all of the non-essential characteristics of
the self, we will still remain with a nuclear self, a minimal self. This
pure consciousness of the self exists only in the so-called “specious
present”, where there are controversies regarding its real duration. It
deals with a minimum point in time when we are conscious. The core
self is connected to the flow of consciousness as it is an integral part of
it. Whatever experience – pain, for example, the perception of a color,
a sudden rage, a brilliant idea, is always felt by a subject as something
happening to him and not to another. The experience is always given in
a certain way and felt as a happening “to me”, as if I were its possessor.
The phenomenality of the experience and its first person character are
not separable. You cannot speak of an experience in an impersonal way.
The experience always happens, by definition, to somebody.
5 Zahavi, 2005; Zahavi, D., Grünbaum, T. & Parnas, J. 2004; Gallagher, 2000,
2006.
6 Gallagher, S. (2000). Philosophical Conceptions of the Self. Trends in Cognitive
Sciences, 4(1), pp. 14–21, p. 15.
In this way, when we speak of the self it is unnecessary to conceive

it as a separated entity. The experience itself has a personal character.
The self is not an essential center that groups experiences around itself.
It is implicitly present in any experience. This idea that the self is im-
plicitly present is very important for the phenomenologists’ argument.
In order for the self to exist, a conscious auto-reflection is unnecessary.
It is unnecessary that I am thinking about myself in order for the feeling
of mine-ness to arise. One of the ways of verifying that this pre-reflec-
tive consciousness of self is always present consists of, for example,
asking a person who is absorbed in reading what he or she is doing.
Without hesitation he will respond, “I’m reading”. In the moment that
he is absorbed in reading, the subject is not conscious of himself explic-
itly. He is completely absorbed in that which he is reading. Neverthe-
less, why he responds without hesitation to a question shows (according
to the phenomenologists) that his consciousness of himself was always
tacitly present. This consciousness of self is not purely empty; there is
a specific depth that constitutes the core self. In this way, the self is al-
ways implicitly present in all of the stream of consciousness and exists
not just in a few privileged moments of reflection.
What implications does this have for the ontological thesis that the
self is an illusion? The self will only be an illusion when we see it in
analogy with physical objects, with the objects of the world. If we allow
ourselves to think of the self in this way and we see it as an auto-ref-
erential process, the problem of knowing whether the self exists or not
becomes different. In this case, questioning the existence of a core self
will be like questioning the existence of consciousness and even of the
world itself, as the consciousness in Phenomenology is not opposed to
the objective world. Consciousness is where the world is given. The
concept of the core self does not stop the phenomenologists from hav-
ing abandoned the concept of the broader or narrative self. Merely that
the core self is assumed in these most diverse selves and that once it is
admitted that the core self is part of the stream of consciousness, it does
not make more sense to say that “we are nobody” as Metzinger and oth-
ers claim, unless they wish to deny the existence of consciousness itself.
The phenomenologists thus claim to reduce the problem of the ex-
istence of the self to the problem of the existence of consciousness. If
there exists a mind-body problem, it includes the problem of the self
because once consciousness is explained there is no longer an addi-

tional problem of the self. Of course empirically the formation of the
broader self from the core self remains to be explained, but this would
not be a hard problem. The so-called “hard problem” is a problem of
understanding how the connection between materia and conscious-
ness is possible. It is a problem that many consider to be insoluble. The
problem of the relation of the nuclear self with the narrative self is not
a hard problem; it is an easy problem because there isn’t a difference in
nature, there is no explanatory gap.
Pre-reflective self-consciousness is not a consciousness of a pre-
existing self, but a consciousness that contains a minimal sense of self,
a core self, as an integral part within itself. If there is experience, there
is a core self. This assumes that any being to which we can attribute
phenomenal and sensorial states will always have a minimum sense of
self. Now, an objection to this train of thought is that the self has not
existed since the beginning of life, even though there are experiences,
such as for example, sensations of pleasure and pain. If babies have ex-
periences but no feeling of self, the thesis of the self as an integral part
of consciousness is questioned. There can be consciousness without a
feeling of self, therefore they are two different things although intimate-
ly related. This question is difficult to investigate because, remember,
what interests us here is the phenomenology of the child and not only
his observable behavior. We can infer the former from the latter, how-
ever this generates different interpretations and it is difficult to produce
evidence for the theory to be proven. The most traditional research-
ers (Henri Wallon7, Merleau-Ponty8, Lacan9) have defended the thesis
that the first months of human existence are selfless. Nevertheless, the
most recent research has claimed to demonstrate that that is not the
case. Gallagher and Meltzoff give an account of this research10. Both
the traditional researchers and the most recent ones agree regarding the
proprioceptive system of the body schema as responsible for the feel-
ing of self and of others, during the first months of life. Gallagher and
Meltzoff confirm meanwhile that there has been a confusion regarding
7 Wallon, H. (1934).
8 Mearleau-Ponty M. (1990).
9 Lacan, J. (1966).
10 Gallagher, N. & Meltzoff, A. (1996).
the body schema and the body image. We can define the body schema as
a system that is unconscious of its processes (motor capacities, abilities,
habits) and that constantly regulates posture and movement. The body
image is a system that can be conscious of perceptions, attitudes, and
beliefs directed at the body itself. The body schema is connected to do-
ing, moving, while the body image is the perception, analysis or control
of these movements. The ambiguous use of these two terms complicates
the understanding of the theory of traditional authors, because they are
not conscious of the fact that they are speaking about different things.
Everyone accepts, in a pacificatory fashion, that a child has an innate
body schema; the same does not apply to that of a body image. Howev-
er, the body image is what gives the feeling of self in childhood. When a
child acquires a body image, it is able to feel experiences as its own and
differentiate them from the experiences of others. According to tradi-
tional authors, this does not happen before recognizing one’s own image
in the mirror. This does not mean that the mirror, as a physical object,
is the cause of this mental evolution, but the mirror translates it and
measures it. Before this there is a phase designated by Merleau-Ponty
as pre-communication and by Wallon as social syncretism where there
is no differentiation between the I and others: as Merleau-Ponty says,
“There is no one individual over against another but rather an anony-
mous collectivity, an undifferentiated group life”11. In early stages of
life, there is a “chaos in which I am submerged” (Merleau-Ponty), a
“blooming, buzzing confusion” (William James). A child does not have
a unified vision of its body. This unified vision of the body depends,
in part, on learning. A child’s perception starts as being interoceptive,
since a newborn does not have the capacity to understand exterior in-
formation, as exterior. There is no distinction between self and environ-
ment. When a child recognizes the image in the mirror as his own this
signifies that he can distinguish the other because it looks at him in the
same way as another. His body appears to him as unified (in anticipated
imagination because in reality it isn’t yet, according to Lacan) and like
one among many others. This conquest is not yet total and definitive
because it is here that the sketch of the self begins, which did not exist
before. The evolution of the no-self phase to the self phase is described
11 Mearleau-Ponty, M. op. cit. p. 83 (my own translation).

by Wallon and Merleau-Ponty as an evolution of an interior that exte-

riorizes itself. Lacan is more radical in indicating that this pertains to
an identification process in the psychoanalytical sense. There is, in the
identification of the image of another, a process from outside to inside.
Without the other, the child cannot construct the self, which is wholly
realized in the image of this other.
Whatever the explanation, what is important here is that these au-
thors make a clear distinction between a period in which the self does
not exist and one where it begins to exist. This has been criticized by
more recent conceptions that attribute more capacities to newborns
than those authors considered as traditional. According to the new
investigators, a newborn already has an innate feeling of self that be-
gins working right away. A newborn already has the feeling of its own
experiences – for example, the sensations of pleasure and pain, that are
his own and not another’s. Of course it would be a minimal feeling, but
justifiably if these authors are right, their data will serve as support for
the thesis that there is a core self where there is consciousness.
The most important empirical fact that these authors have estab-
lished is the observation of imitation by newborns. Merleau-Ponty wrote
in “The Child’s Relations with Others” that “The child executes a gesture
to the image that he sees made by another: he smiles because it smiled
at him. It is necessary that the perceived image is translated into motor
image; since he cannot have an image of himself smiling nor the mo-
tor feeling of another. The transfer of the other to him is impossible by
analogy.”12 He shows a model here of translation that states perception
and motor skills are different by nature considering that it is necessary
to transform visual stimulus into motor stimulus. However, a child does
not have an image of itself nor the feeling of the other before 6 months
of age and because of this it cannot imitate. The integration of perception
and motor skills will be progressive through maturity and learning. Gal-
lagher and Meltzoff argue against this, based on various research, that
there is imitation among children between 0 and 6 years of age, the main
empirical fact being in reference to tongue protrusion. In agreement with
their interpretation of the facts, a child recognizes the adult’s gesture of
sticking out a tongue at him and rapidly transforms this perception into
12 Mearleau-Ponty, M. op. cit. p. 70 (my own translation).

a motor perception. To explain this fact, they argue that a supramodal

system/device exists where the visual and motor systems speak, from the
beginning, in the same language. A child does not have any difficulty in
relating the actions he sees in others, that are in fact others to him, with
the invisible actions within himself because innately, without requiring
any training, he already has a body image that makes this connection.
Obviously training will play an important role in his development, but
Gallagher and Meltzoff argue that the child already has the cognitive bag-
gage necessary in order to carry out elementary imitation. If imitation is
possible, then we have evidence that a minimal sense of self and of others
exists from birth. However, this interpretation of the empirical data has
been opposed by researchers who carried out other experiments and of-
fered alternative interpretations. Talia Welsh refers to these results13 and
as a philosopher assumes the freedom to propose a global interpretation
of the other results. According to her perspective, intelligent and com-
municative behavior does not necessarily imply a self-consciousness. If
it did then we would also have to consider that bees have consciousness
of self, which does not seem to be the case. Another interpretation of the
facts related to sticking out the tongue is possible. The child sticking out
his tongue is not imitating an older child. His behavior seems identical,
but the phenomenology is different. In an older child’s imitation, there
is, as we have seen, a mental picture, consciousness of himself and the
other. Assuming that the same behavior exhibited by the newborn has the
same meaning is adapting the facts to the theory. Talia Welsh argues this
is valid both in the strong imitation thesis according to which a newborn
does imitate, and in the weaker thesis according to which a newborn’s
imitation is not properly imitation but is a preform whose development
will result in imitation. The second possibility seems more plausible, but
different experiments seem to deny it. The alternative theories that Talia
Welsh refers to basically claim that the initial behavior does not reveal
self-consciousness. It is merely exploratory reflex and auto-regulation,
that exists to avoid discomfort. It is because of adult influence – the par-
ents – that this exploratory behavior and auto-regulation transforms into
a self-consciousness. I think that we could say that if we were left alone
to ourselves (believing this were possible), the exploratory behavior and
13 Welsh, T. (2006).
auto-regulation would not make us auto-conscious. In the case of tongue

protrusion, it is an exploratory behavior just like any other, which is re-
inforced by an adult and is repeated as such. Sticking the tongue out does
not signify that the child understands the behavior of another as an other
and then reproduces it in himself. This implies knowing how to move an
invisible part of the body, knowing that it is his tongue that is moving and
not the other’s. Nevertheless, various experiments, that I cannot enumer-
ate here, go in the direction of showing that tongue protrusion is an ex-
ploratory behavior just like any other and does not have a posterior sense
of imitation. It cannot even be said that it deals with a proto-imitation in
the sense of a very basic mimicry that will be developed. It does not deal
with any type of imitation, but with a different behavior, an exploratory
reflex like any other.
Although it is not proven that there are no feelings of self in the ini-
tial stages of life, even of a core-self, I think that there is good reason to
speculatively concede that it does not exist. If it were true, what would
the consequences be for the problem we are studying, the problem of
the self?
If the feeling of self is not present from the start, will there be a period
of time in which the feeling of self and simple phenomenal consciousness
are in different states? There will be phenomenal consciousness without
the feeling of self. Thus the concept of pre-reflective self-consciousness
cannot resolve the problem of the self. The feeling of self is a phenom-
enological fact and its reality depends on the reality of phenomenology.
Metzinger considers that all phenomenology is an illusion, and in this
manner the self does not exist. He seems to share the idea that phenom-
enology implies a self, but here I consider that phenomenology can exist
without a self, thus the two problems are distinct from one another.
If we accept that phenomenology is not a hallucination but a reality,
we will find ourselves with yet another problem of knowing from where a
self forms from no self. Through what mechanisms does the state of cha-
os and indistinctness becomes the feeling of being one, of continuance
in time, of being a separate entity. Through simple nervous maturity?
Through social interaction? From the phenomenological point of view, as
a process of integration of the various states in one unified center?
I do not intend to explore these questions, but only to demonstrate
that the notion of core-self does not resolve the problem of the self. In
conclusion, given that one can phenomenologically exist without a mini-

mum feeling of self, the problem of its nature and origin remains unsolved.
References
Gallagher, S. (2007). The phenomenological Mind: An Introduction to

Philosophy of Mind and Cognitive Science, New York: Routledge.
Gallagher, S. (2000). Philosophical conceptions of the Self: implica-
tions for cognitive science. Trends of New Science, vol. 4, Nº 1,
January.
Gallagher, S. (2006). How the Body Shapes the Mind, Oxford: Oxford
University Press.
Gallagher, S. (2000). Philosophical Conceptions of the Self. Trends in
Cognitive Sciences, 4(1): 14–21.
Gallagher, N. & Meltzoff, A. (1996). The Earliest Sense of Self and
Others: Merleau-Ponty and Recent Developmental Studies. Philo-
sophical Psychology, 9(2), 211–233, 1.
Lacan, J. (1966). Le Stade du Mirroir comme Formateur de la Fonction
du Je, Écrits, Paris: Seuil.
Mearleau-Ponty M. (1990). As relações com o outro na criança. In
Sorbonne: Resumo de Cursos, Papirus, S. Paulo: Éditions Cynara.
Metzinger, T. (2009). The Ego Tunnel – The Science of the Mind and the
Myth of the Self, New York: Basic Books.
Metzinger, T. (2003). Being No One. The Self-Model Theory of Subjec-
tivity, Cambridge, MA: MIT Press.
Parnas, J. & Sass, L. (2010). The Structure of Self-Consciousness in
Schizophrenia. In Gallagher, S. (ed.) Oxford Handbook of the Self,
Oxford: Oxford University Press, 521–546.
Wallon, H. (1973). Les origines du caractère chez l’enfant. Les préludes
du sentiment de pesonnalité, Paris: Boisvin.
Welsh, T. (2006). Do Neonates Display Innate Self-Awareness? Why
Neonatal Imitation Fails to Provide Sufficient Grounds for In-
nate Self and Other-Awareness. Philosophical Psychology, 19(2),
221–238.
Zahavi, D. (2005). Subjectivity and Selfhood: Investigating the first-

person perspective, Cambridge, MA: The MIT Press.
Zahavi, D. (2005). Being Someone. Psyche, 11(5), 1–20.
Zahavi, D. (2007). Subjectivity and Selfhood: Investigating the First-
Person Perspective. Cambridge, MA: The MIT Press.
Zahavi, D., Grünbaum, T. & Parnas, J. (2004). The structure and devel-
opment of self-consciousness John Benjamins.
The Reality of the Virtual Self as Interface
to the Social World
Robert Clowes
Introduction
This article argues for a virtual notion of self built around the idea of
social self- positions. It begins by analysing several virtual-reality mod-
els of mind and noting the different ways the ideas of virtuality, virtu-
al reality, and presence, can work in explaining mental phenomena. It
looks particularly at Metzinger’s idea that the self is a form of virtual
body representation. It then explores his contention that the self, if vir-
tual in this way, is strictly speaking inexistent and should therefore be
eliminated from our scientific conceptual vocabulary. It finds this
conclusion is premature and ill-motivated.
Metzinger’s notion of phenomenal self models (PSMs) opens the
way to make the case that selves are actualizing virtualities. These are
projective virtual entities that play a central role in organising our ac-
tions and constituting us as beings that act, take decisions and realize
ourselves in consideration of an ongoing sense of ourselves. These par-
ticular actualizing virtualities can be viewed as instruments that we use
to help maintain and realize ourselves as coherent beings.
The paper then changes tack to look at one area where a notion
of the self appears to be indispensible, namely, psychiatric science and
practice and specifically the theorisation of schizophrenia. This exami-
nation reveals that selves are necessary in much psychiatric theorisation
perhaps especially the Ipseity Disturbance Hypothesis (Cermolacce,
Naudin & Parnas, 2007; Parnas, 2003; Sass & Parnas, 2003). I find that
there are strong empirical implications that we will need to explain to
222 Robert Clowes
understand how self can become distorted in pre-reflective experience

in order to achieve better theorisations and explanations of certain as-
pects of schizophrenia. I then turn to a related, phenomenologically
led, analysis of schizophrenia: the dialogic theory of self (Lysaker &
Lysaker, 2008) which leads me to conclude that the virtual notion of self
plays an important role in reconstructing our notion of pre-reflective self
experience. One implication of this theory is that, in important ways,
the self cannot be identified with the body, real or virtual. Rather self,
even the pre-reflective self, has an irreducibly social and interpersonal
aspect, albeit based in an extended form of body representation. An-
other implication is that we may not be able to fully explain “disorders
of the self ” such as schizophrenia purely in terms of the embodied
self, but may require a more extended and richer idea of self and self-
experience.
Having made this leap, I return to, and attempt to reconstruct the
idea of the virtual self, that, I argue, needs to be refocused around the so-
cial roles we inhabit, namely Lysaker & Lysaker’s idea of self-positions.
We look at a reconstructed version of the virtualist representation of the
self, arguing that while the self may be based in bodily representation,
this representation is not targeted on the body. It should rather be under-
stood as a sort of body-based projection into the social world, managing
the more or less coherent projection of appropriate self-positions. This
gives us the basis to contend that while the self may be virtual, it is not
necessarily illusory. Rather the virtuality of self plays a central role in
constituting us as the sorts of beings we are. Somewhat reconceived,
the self concept nevertheless continues to do an important job both in
our scientific and folk-psychological theorizing and potential expla-
nation. Indeed, achieving a proper understanding of how the self may
be grounded both in neuroscience and the social world may be needed
if we are ever to make progress in understanding conditions such as
schizophrenia and what it means to be a coherent human being.
The Reality of the Virtual Self as Interface to the Social World 223
Virtual Reality and the Experience of Presence
The idea that virtual reality might be a good metaphor for the ar-
chitecture of our minds persistent in the literature on mind from the
early 1990s. Several philosophers of cognitive science (Clark, 2008;
Metzinger, 2004a; Noë, 2004; Revonsuo, 2006) have taken up the idea
seriously although there has been little systematic attempt to look at
the variety of sometimes conflicting ways that the metaphor is used1.
Metaphors are important in scientific and philosophical model building
(Lakoff & Johnson, 1999) although there is always a danger that they
can be misleading if not examined properly, so this section will discuss
some of the ways in which the ideas of virtual reality, and virtuality
have been used in recent theoretical cognitive science before we return
to its potential application to the self.
Ancestors of the idea of virtual reality can be found in philosophy
as far back as Plato’s cave and, in early modern thought, Descartes idea
of a malicious demon. An influential recent version of these ideas was
Revonsuo’s analysis of consciousness, dreaming and especially the then
nascent work in virtual reality around the concept of presence. This
led him to claim that perception, and our sense of reality more gener-
ally, could be considered a sort of total simulation (Revonsuo, 1995).
Revonsuo thus uses virtual reality and telepresence technologies as
a metaphor for normal or everyday experience. Today the concept of
some virtualistic working out of perception has become rather wide-
spread among philosophers of cognitive science, who ostensively hold
very different ideas about what perception is, how it works and its
broader place in the mind (Clark, 2008; Metzinger, 2004b; Noë, 2004;
Revonsuo, 2006)2.
One common way of theorising what virtual reality and telepres-
ence technologies do is to hold that they project presence (Lombard
& Ditton, 1997). That is to say they make us feel present in a scene, a
1 However the job of working out some of the possible positions was begun in the
recent paper Virtualist Representation (Clowes & Chrisley, 2012).
2 A detailed discussion of the strengths and weaknesses of a number of these views
and how they relate to a broader virtualist picture can be found in (Clowes &
Chrisley, 2012).
224 Robert Clowes
locale, a place which may or may not be the one our bodies are actually
located in. In virtual reality our bodies are typically located at a com-
puter screen or, more exotically, using VR goggles or some other piece
of technology. Yet we feel present, that is to say spatially and temporally
situated and embodied in some projected scene. Behind any perceptual
seeming is typically a computer simulation.
The sense of presence – at least in virtual reality scenarios – always
appears to depend on the mediation of an interface. An interface con-
nects some actual or simulated world via technological device (such as a
remote video-camera controlled by a joystick, or a head-mounted video
display into a virtual world) to our sensory apparatus (our eyes, our ears).
It is through the interface that we view the perceptual world and are able
to act in it. All interfaces both make available and constrain certain pos-
sibilities for action. It is through the interface that the sense of presence
is conveyed. When presence is spoken of as the sense of being there, or
being now there in such scenarios it is relatively clear what is meant. We
feel – to a greater or lesser extent – as if we were actually located in the
virtual world being projected by the computer, or (in telepresence scenar-
ios) in the actual location of the video camera we are controlling, rather
than wherever our bodies are physically located.
Some hold that the concept of presence as understood in virtual real-
ity research can be understood to have a deeper significance than the ways
in which we interact with computers. It has been argued that presence
should really be understood as a central aspect of phenomenal experience
itself (Seth, Suzuki, & Critchley, 2011). It has been linked to certain psy-
chiatric conditions – e.g. depersonalisation disorder – supposedly marked
by a deficit in this sense of presence (Radovic & Radovic, 2002; Sierra,
Baker, Medford, & David, 2005). Another possibility is to link it to a wid-
er set of existential feelings (Ratcliffe et al., 2008). Using the concept of
presence – or the sense of presence – to refer to the something more basic
about our apprehension of the world outside of VR scenarios, stretches
the use of VR as a metaphor. The idea is that our basic cognitive architec-
ture itself projects a sense of presence3.
Anti Revonsuo (2006) has perhaps gone further than anyone in de-
veloping ideas around virtual reality as a sort of total metaphor both for
3 I return to this idea later in this article in the discussion of ipseity and the ipseity
disturbance hypothesis about schizophrenia.
mind and our knowledge of the world. Revonsuo uses the term “world
simulation” to convey the idea than the world as experienced is itself a
sort of simulation run by our brains. He writes:
in world simulation, the brain creates for us only virtual presence or telepresence
in the surrounding world, but we ourselves, and our consciousness never in actual
fact escape the confines of the brain. (Revonsuo, 2006, p. 109)
On this analysis the use of telepresence or virtual reality technol-

ogies can be understood as promoting a perceptual illusion that we are
somewhere – out in the world – when we are not. Our intuition that we
are truly in touch with the world, on this analysis, is a sort of naïve-real-
istic illusion created by the brain as part of its representational activities.
The idea that perceptual experience can be understood as a form of
virtual reality which projects a world has become widespread. Indeed
it can be taken as the new way of working out traditional philosophical
ideas about indirect perception. For instance Barry Dainton recently
describes the projectivist approach to perception:
We take ourselves to be directly perceiving the world; we do not (while in the

natural attitude) detect the presence of an experiential medium which lies between
ourselves and the things in the world we see, hear and touch. Yet if projectivism
is true, there is a sense in which we are all enclosed in spheres of virtual reality,
phenomenal spheres somehow produced by activities within our brains: all we are
directly aware of are contents within these spheres. But this is not how it seems.
Even when I am dwelling on the absurdities of naive realism I do not seem to be
enclosed in a virtual world. I seem to be surrounded by ordinary material things,
tables, chairs, walls not experience. (Dainton, 2006)
Yet one does not have to be a projectivist or anti-realist in the way that
Dainton discusses here in order to consider that the metaphor of virtual
reality has some use. Implications of the virtual reality metaphor for
mind are not in fact univocal. They can and have been used in ways to
conceptualise the mind without implying that we are out of contact with
the world beyond our perceptual processes. In fact, the idea of virtuality
has also been used to theorize the way that our (especially) perceptual
systems are in constant contact with, even including, the proximal world.
On these views our perceptual system does not need to reproduce infor-
mation internally because our cognitive systems operate as if absences
were not there. (See Noë’s idea of enactive perception discussed below).
226 Robert Clowes
One early and influential development of such ideas can be found

in Chapter 11 of Dennett’s (1991) Consciousness Explained. This is de-
veloped around his critique of the then popular idea of filling in. Den-
nett’s discussion began from the close analysis of a number of phenomena
about visual perception. His starting point was the well-known blind-spot
in our visual field. It is well established that the part of our retina, where
it connects to the optic nerve, does not in fact have light receptors. Yet as
Dennett (and others) pointed out we do not notice any gap in our visual
field. The question became – in much neuroscience discussion of the time
– how is the gap filled in? Dennett believed the question to be misguided.
On the contrary he argued the brain does not need to fill-in, it simply
ignores the lack of information. In a related example Dennett imagines a
viewer entering a gallery and coming across a wall of Marilyn Monroes
in the style of Andy Warhol (ibid, pp. 354–355). Dennett notes that the
art lover experiences a seemingly immediate impression of the visual si-
militude of all of those Marilyns and yet her visual system does not have
enough time to pick-up and represent internally all of the information that
would seem to be necessary for her to enjoy the apparent rich visual detail
of the scene. Dennett’s explanation of this apparently rich phenomenolo-
gy again turns on the brain not having to fill in detail but, on the contrary,
its ability to selectively ignore the lack of certain apparently needed infor-
mation to complete the visual field. Dennett imagines what our art lover’s
brain might have to do on entering the gallery: “Having identified a single
Marilyn, and received no information to the effect that the other blobs are
not Marilyns, it jumps to the conclusion that the rest are Marilyns, and
labels the whole region ‘more Marilyns’ without any further rendering of
Marilyns at all.” (ibid, p. 355).
According to the filling-in hypothesis the brain must synthesize
content for those missing parts of the visual field in the blind-spot or
produce extra Marilyns. Dennett argued instead that the brain practiced
a benign form of neglect simply ignoring much absence which was only
signalled to higher brain areas if some problem was noticed. Dennett
formulated the problem of the tradition view and his response like this
“the fundamental flaw in the idea of ‘filling in’ is that it suggests that the
brain is providing something when in fact the brain is ignoring some-
thing” (ibid, p. 356). What the brain is ignoring for Dennett is the ab-
sence of a lot of detailed information about the environment.
Nevertheless the brain does produce a sense of virtual presence.

Dennett introduces the idea with a citation from Marvin Minsky: “We
have the sense of actuality when every question asked of our visual
system is answered so swiftly that it seems as though those answers
were already there.”4 For Dennett the Marilyns were virtually present in
consciousness in the sense that we had access to highly-detailed visual
information when and as needed. Dennett and others have taken this
line of thought to indicate that perception is not all that it seems (Den-
nett, 2001). Some even take it that the richness of the perceptual world
itself is a kind of Grand Illusion (O’Regan, 2002).
Other responses to the putative Grand Illusion, that hold that we do
not misperceive the world. Alva Noë, Kevin O’Regan (2001) and others
have developed an enactivist (or Sensory Motor Contingency) view of
perception that in many ways builds upon Dennett’s ideas about virtual
presence. They contended that the reason we enjoy rich experience of
the world is not because of the construction of a detailed internal rep-
resentation, but because we are constantly in touch with the world and
able to pull in detail on a need-to-know basis. Noë refers to the idea in
his more recent writings where he claims we enjoy virtual presence in
perceptual consciousness because we are always able to draw on the
actual presence of the world. He writes:
We have the impression that the world is represented in full detail in conscious-
ness because, wherever we look, we encounter detail. All the detail is present,
but it is only present virtually, for example, in the way that web site’s content is
present on your desktop. (Noë, 2004)
For Noë the detail of the visual scene is virtually present in the sense
that it can always be effectively drawn upon with an – often uncon-
scious – glance or a movement. However unlike Dennett, Noë argues
this does not mean we are mistaken about perception (Noë, 2004). It
is our active engagement (not our out-of-touchness) with the objects
of our perceptual processes that makes a visual phenomenology come
out correct. We believe we are in touch with the world and indeed we
are, in virtue of our ongoing contact with it, and action within it. On
this view virtual presence is not an illusion but a way of making sense
4 Cited in (Dennett, 1991) p. 359, from (Minsky, 1985) p. 257.

228 Robert Clowes
of both phenomenology and the findings of perceptual psychology. For

Noë, it is not that we are persistently out of touch with the world, but
rather our cognitive architecture continually factors in the presence of
that world. It is because we are continually drawing on our contact with
it, that it is (virtually) present to us. On Noë’s interpretation we are not
operating under any illusion that we are experiencing the rich detail of
the world, because we are indeed experiencing that detail in the only
way we could. Not because of the construction of a detailed internal
representation, but because of our ongoing contact with the world.
This article seeks to extend Noë’s general approach to the use of the
virtualist metaphor, and specifically his critique of the idea of a Grand
Illusion to dealing with the problem of self. Just as Noë argues we are
in continual contact with the world and therefore not experiencing an
illusion, I will argue we are similarly always in contact with the bases
of self. For now it is enough to say that endorsing and using the virtual
reality metaphor, in some form, need not commit us to the sorts of global
anti-realism that at least Metzinger and Revonsuo think follow from it.
On the contrary, it may be that a form of virtualism is actually one of the
best ways of rebuilding a concept of veridical perception that does justice
to psychological findings. As I will attempt to show, it can help rebuild a
concept of the self that might allow us to ground some of the more im-
portant roles self plays in our current conceptual schemes. In order to do
this, however, we need to do some conceptual groundwork.
The virtual reality metaphor is easily misused such as when it is
suggested that we are ‘imprisoned in the brain’. However, I shall at-
tempt to show here that it has real purchase in helping us understand
some of the ways in which situated agents like us experience our social
presence in the world. It offers new possibilities for developing a refined
analysis of that sense of presence and even how having a sense of pres-
ence links up with some important themes in contemporary psychiatry.
This essay will aim at the reconstruction of the virtuality metaphor in
such a way that it preserves some of its important theoretical advantag-
es while some of its more paradoxical implications are overcome.
The Virtual Self and Pre-Reflective Self Experience
The idea that the self is in some way virtual can be found in several
places in the literature including Dennett’s idea that selves are Centres
of Narrative Gravity (Dennett, 1991). For Dennett, selves are narrative
constructs created by organisms like us to make sense of themselves,
but their apparent coherence is greater than its reality. We spin selves,
and they do a job for us, but they are rather fictional: The best stories we
can come up with, rather than a true story. It is Thomas Metzinger, how-
ever, who has done most to develop the idea of a virtual self in the recent
discussion (2000; 2004a; 2009). Unlike Dennett, Metzinger’s notion of
the virtual self is not a high-level narrative construct but is grounded
in the way organisms represent and control their bodies. Metzinger’s
idea of virtual selves can be seen as a way of resolving a major problem
implied by Revonsuo’s (internalist) account of virtual presence, namely
that our inner (Cartesian) selves are out-of-touch with the world beyond
its representational media. But – it is natural to ask – if we concede that
we are never in touch with reality itself, only our own representational
media, what makes the self so special? Might not the self itself just be
another form of representational content? The boldness of Metzinger’s
(2004a) account is to do away with the confined or imprisoned self as
posited by Revonsuo by proposing that the self is itself virtual.
What Metzinger appears to primarily mean by this is that the self
is a form of simulation produced by our brains in order to situate our-
selves in the world. The benefit here is that with both the self and the
world considered as artefacts of brain simulation, the radical disconnect
is closed. On one level this is a nice way out of the cage that Revonsuo
has built for the self. If there is no real self to begin with it makes no
sense to think of it as being imprisoned in the brain. The cost is that, by
Metzinger’s lights, the self as typically understood becomes itself a type
of illusion. Whether it is worth bearing this cost, and indeed whether a
well-developed theory of the virtual self requires us to bear this cost is
one of the central questions of this paper.
Representational systems on Metzinger’s analysis – in terms which
derive from the mental models theory of classical cognitive science
(see Johnson-Laird, 1983) – are just a certain class of brain models.
230 Robert Clowes
Metzinger’s basic notion of the Phenomenal Self Model (or PSM) is

a brain model that an organism builds of itself in order to plan its ac-
tions and allow the coordination of its parts. PSMs are necessary insofar
as systems (organisms) like us require an approximately total coher-
ent sense of themselves in order to organise their activities as a whole
system. The evolutionary value of PSMs accrues in virtue of the fact
that, for complex multi-limbed organisms like us, much everyday and
planned action requires not just a response to environmental stimulus,
but a nuanced, whole-body response that requires the overall coordina-
tion of body-parts with many degrees of freedom.
Nevertheless, Metzinger frequently emphasizes how the self/world
distinction is merely a somewhat arbitrary boundary in representa-
tional space; as Metzinger says at various points, it is a partition (e.g.,
Metzinger, 2004a, p. 87). One part of representational space is coded
as the organism itself, the rest its environment. The sense that we have
a self arises from this basic (internal and fallible) partition our brains
make between self and world. The sense that this partition is in some
sense arbitrary appears to be another reason Metzinger wants to say it
is virtual, i.e., the self partition does not respect the existence of real
entities in the world; it is only a form of projection that the organism
finds useful.
Metzinger claims that the self itself is an artefact of this sort of in-
ternal informational organisation. Selves arise in organisms that need to
plan, reason, and act upon the world in sophisticated ways. In large part
the job of our brains is not just to produce models of the environment,
but, by using a variety of motor models, to simulate, model and situate
our place in the world. According to Metzinger complex organisms like
us build models of not just their environments but also of themselves,
they (we) are self-simulators.
The root of Metzinger’s argument that the self is an illusion hinges
on his idea that the self is really a sort of artefact produced by representa-
tional systems like us in the act of self-control. This arrangement, for
Metzinger, simply falls out of the nature of our basic computational
architecture. Crucially the Phenomenal Self Model (PSM) is not
merely a way of navigating the environment but the basic means by
which we become aware of ourselves as unified beings. PSMs are con-
scious self models. PSMs thus unite two important aspects of the sorts
of beings we are. They are simultaneously a sort of content produced by

certain complex agents for self-control, but they are also the means by
which those agents become aware of themselves. As Metzinger writes:
Conscious experience, too, is an interface, an invisible, perfect internal medium
allowing an organism to interact flexibly with itself. (Metzinger, 2009, p. 104)
PSMs are also thus, in terminology we have already discussed, to be

regarded as interfaces. They exist in order to mediate certain types of
action which require the coordination of the organism as a whole. In-
sofar as these models are interfaces they present particular modes of
action and possibilities and obscure others. Those that are presented are
ultimately oriented to the organism’s self-control. They do not however
present, as Metzinger often notes, some deep ontological truth about
the body or its relationship with the world. Rather they are tools with
which to produce coherent action and ways of regarding and interacting
with oneself.
For Metzinger the PSM is a model that complex organisms build
of themselves in order to produce and integrate coherent action. The
content of a PSM is the organism’s current best guess about how it
is concurrently poised and organised. This is another sense in which
Metzinger holds that selves are virtual, i.e., they do not reflect any ob-
jective truth about the world, but, like the red arrow on a subway map,
they allow us to find ourselves in the world and orient and act within it.
Metzinger writes:
This self-model is the little red arrow that a human brain uses to orient itself with-
in the internal simulation of reality it has generated. The most important difference
between the little red arrow on the subway map and the little red arrow in our
neurophenomenological troglodyte’s brain is that the external red arrow is opaque.
(Metzinger, 2004a, p. 552)
This metaphor of the little red arrow is important to Metzinger because

it establishes a central phenomenological role for the PSM. The PSM is
the means by which organisms are able to transparently and pre-concep-
tually register their presence in the world. But the PSM, rather than be-
ing a content of experience, is given transparently in experience. When
Metzinger writes the red arrow is opaque, the notion of opacity being
used is a phenomenological one to be contrasted with phenomenological
232 Robert Clowes
transparency. The red arrow on the subway map is a perceptual object

that allows us to establish our position in the world. But the PSM is not
encountered as a content or object of perception but as part of the way
the world is given to us in experience. Put another way, through the
PSM, the self is encountered, phenomenologically speaking, not as a
content but as an aspect of experience. In this way – and perhaps unex-
pectedly5 – Metzinger’s ideas connect up those neo-phenomenological
writers trying to explicate the basic structure of experience; especially
with those writers who contend that there is minimal sense of self to be
found pre-reflectively in experience (Gallagher, 2005a; Legrand, 2007;
Sass & Parnas, 2003; Zahavi, 2005b). Metzinger’s notion of a transpar-
ent PSM thus endorses much recent and classical phenomenological
discussion which finds a basic apprehension of self in the structure of
phenomenal experience itself. What Metzinger refers to (citing Frank,
1991) as pre-reflexive self-intimacy6 appears to be the very same con-
cept that Zahavi and others call pre-reflective self intimacy. What more
ontologically realistic writers are inclined to call the minimal self is
thus very closely connected to what Metzinger calls a PSM.
The way that the PSM operates within the organism is supposed
to explain certain aspects of phenomenology, and crucially for our
discussion, why it is misrepresentational. It is important to remember
that Metzinger claims that the pre-reflective self experience presents
us with an illusion. Metzinger thinks that insofar as the basic structure
of subjective experience presents us as having a form of pre-reflective
self-consciousness, it is also presenting us with a distorted picture of
ourselves and our place in the world. For Metzinger our pre-reflective
sense of self is merely the projection of a virtual reality producing sys-
tem, that is, a system manufacturing a sense of presence by continually
attempting to model its ongoing interactions with the environment in
the most coherent way possible.
5 As Metzinger is often taken as being anti-realist about phenomenology.

6 Metzinger writes the following, in terms that seem to patently echo what others
call pre-reflective self-consciousness: “There seems to be a primitive and prere-
flexive form of phenomenal self-consciousness underlying all higher-order and
conceptually mediated forms of self-consciousness and this nonconceptual form
of selfhood constitutes the origin of the first-person perspective.” (Metzinger,
2004a, p, 158).
What we experience then is not ourselves or the world but a com-

plex simulation. Thus Metzinger argues that “phenomenal first-person
experience and the emergence of a conscious self are complex forms
of virtual reality” (Metzinger, 2009, p. 106). Metzinger argues that in
a deep and metaphysically necessary sense, PSMs only present a best-
guess about how things really are.
At any given time, the content of bodily experience is the best hypothesis that
the system has about its current body state. The brain’s job is to simulate the
body for the body and to predict the consequences of the body’s movements, and
the instrument it uses is the self-model. (Metzinger, 2009, p. 114)
This sense in which PSMs are only best-guesses is for Metzinger an-
other, indeed the central reason that they should be considered virtual.
To recap then, for Metzinger, selves are virtual, or are analogous to
virtual reality devices, in at least five distinct senses.
1. Selves are virtual because they are simulations; or more precisely,

simulational content.
2. Selves are virtual in that they are constructs arising from partitions
in representational space that do not necessarily respect any divi-
sions in reality.
3. Selves (PSMs) are virtual – by analogy with virtual reality sys-
tems – because they can be regarded as control interfaces that make
available (and obscure) certain possibilities for action.
4. Like the interfaces in virtual reality devices, selves (PSMs) convey
a sense of presence.
5. The contents of PSMs are also virtual because they tend not to
correspond to, or depict, actual states of affairs in the world, but
best-guesses or mere possibilities about the state of the body.
Insofar as PSMs show up for us as pre-reflective and transparent aspects

of our experience, they also tend to foster certain misapprehensions or
‘illusions’ about what we are and our place in the world. As Metzinger
likes to remark:
“there is no thing like “the self.” Nobody ever had or was a self.”
(Metzinger interviewed in Taft, 2012).
While such conclusions are clarifyingly provocative, I do not be-
lieve they follow from anything we have discussed so far. It is possible
234 Robert Clowes
to endorse most, or indeed all, of what Metzinger has to say about the
virtuality of self, without drawing his conclusions about the self ’s inex-
istence or unreality. In order to better understand how Metzinger comes
to the conclusions he does, and why they do not follow, we need to
narrow our focus to his ideas about the body and body representation.
The Virtual Body and the Minimal Self
Metzinger develops several examples to help illustrate his case that bod-
ily representations, or more properly representations of a self grounded
in the body, are virtual and thus illusory. One rather exotic case study
comes from the observations of some problems of astronauts during the
course of space-travel (Metzinger, 2004b). While in orbit or otherwise
placed in a condition of weightlessness, it appears that human beings
can develop a confused sense of presence, i.e. the sense of how their
body fits into a wider spatial frame of reference. Alongside this, they
sometimes undergo quite bizarre illusions about the current state of
their bodies. This condition, known as “Space Adaptation Syndrome”,
can also cause severe nausea and appears to affect around 50% of those
who participate in space flight. Interestingly it is not merely the sense
of spatial orientation which is affected but also the disposition and inte-
gration of one’s body. Here are the words of one astronaut:
The first night in space when I was drifting off to sleep,” recalled one Apollo as-
tronaut, “I suddenly realized that I had lost track of [...] my arms and legs. For all
my mind could tell, my limbs were not there. However, with a conscious command
for an arm or leg to move, it instantly reappeared – only to disappear again when
I relaxed. (Barry & Phillips, 2001)
This condition seems to be created by the lack of certain contingent

stimulations or sensory information that would that appear to help our
brains construct a fused image of the body, along with its current dispo-
sition, composition and spatial configuration within a reference frame.
The particularly common aspect of the deficit appears to be a disruption
of the subject’s ongoing sense of where the ground is and hence how up
and down should be organised in egocentric space. Metzinger (2004b)

writes that in such cases other astronauts can “help their buddy” by
striking the soul of his or her foot. This seems to help with restabilising
a more normal reference frame and with it a coherent representational
fusion for the body.
It is perhaps surprising that a coherent and stable body image turns
out to depend on such apparently contingent matters as the impact of
one’s feet on the ground. However, it is not clear why one should think
that, because body-image and the sense of presence can be disturbed by
the absence of a normal gravitational field, they are systematically defi-
cient or misrepresenting. What we might admit, however, is that insofar
as the systems that the brain uses to represent and control the body are
considered to give a fully contemporaneous and veridical snapshot of
the body, they are not fully successful.
One further hypothesis, in the case of the astronaut, is that it is
not so much that the egocentric body image is impoverished in itself,
but that some bodily representations are poorly integrated into one’s
overarching subjective experience. This line of thought fits nicely with
some theoretical accounts that claim that the brain systems that underlie
our sense of body schema and peripersonal space are the same system
(Cardinali et al., 2009). In space sickness the sense of how the body is
embedded or situated in a wider space becomes, for the organism, con-
fused and alongside it the occurent sense of the body. On this analysis
the partition in representational space appears to be intact but its em-
bedding in a more global space can no longer be adequately fused. The
condition is in part interesting because it reveals how under standard
conditions, the representational systems of our body do invisibly per-
form this integration. Space sickness is the phenomenological correlate
of the inability to fuse a stable sense of the body embedded in space.
The discussion of space sickness also helps us understand what
Metzinger means by a virtual model. Regarding the astronaut example
he notes, “[t]his shows that the self-model of human beings is a virtual
model which, if underdetermined by internally generated input (i.e., from
gravitational acceleration affecting the maculae utriculi and sacculi in
the vestibular organ), is highly context dependent: Its content is a possi-
bility and not a reality.” (Metzinger, 2000, p. 290) This helps convey part
of what might make the PSM a virtual model in Metzinger’s sense, and
236 Robert Clowes
presumably in some sense ontologically dubious. First the model goes

beyond the data, insofar as it is made to fit into some pre-existing struc-
ture. Second, the model does not convey and probably does not seek to
convey the actual current information from the body, but is rather a sort
of most-coherent best-guess about that body’s actual state as embedded
in a spatial reference frame. But why should being under-determined by
the data in this way make body representation illusory?
Insofar as the brain’s model of the body goes beyond the actual state
of the body by, for instance, creating various forward models to predict
action, this can be argued to be – far from necessarily problematic –
the best way to actualize an overall and coherent view of the body.
Indeed this picture fits nicely into some current attempts to under-
stand representation along Bayesian lines (Clark, 2012) or in other
ways that see mental representation grounded in anticipatory systems
(Grush, 2004; Pezzulo, 2008). Is there some other – non-problematic
– way that we could grasp the state of our body, say by creating some
instantaneous internal picture? Is the fidelity on an internal ‘snap-
shot’ of the body really a good success criterion for a felicitous rep-
resentational architecture? Who or what would be looking at such
a model anyway? Whether our sense of ourselves is compromised
insofar as it relies on such models is surely something which needs
to be shown. Nevertheless, I would like to agree that Metzinger is
importantly on the right track about the way in which our represen-
tational processes might be considered virtual. But whether or not
this is inherently problematic rather depends on what we think the
desiderata might be for a proper or realistic sense of self and indeed
successful representation.
A second set of illustrative examples, Metzinger uses, involves
phantom limbs. It is well known that when a human being suffers the
loss of a limb, a side-effect of the corporeal damage is often the appear-
ance – for the patient – of a phantom limb (Ramachandran & Blakeslee,
1998). Phantom limbs are experienced by some patients as highly
present and real even to the point that they will appear to forget they
do not really have the phantom, by falling over as they attempt to stand
on a missing leg, or attempting to reach for an object with an arm that
doesn’t exist. On reflection however subjects with phantom limbs are
generally aware that the phantom is unreal although they cannot shake
off its felt presence. Phantoms thus appear to be relatively immune from
higher-level knowledge or standard rational inference. For Metzinger,
this is another illustration of how the body is virtual. One hypothesis
about such cases is that the body’s best guess about itself is constrained
by certain pre-suppositions about the form the body must take. This
causes the brain to fairly systematically miss-represent the current re-
ality of the body.
Figure 1: Taken from Brugger, P., S. Kollias, et al. (2000).
Strong evidence for this hypothesis is the fact that subjects who have
never possessed their actual organic counterparts can experience phan-
tom limbs. Such aplasic phantoms are demonstrated in the well-known
case by AZ: a middle aged woman born without forearms and legs,
who nevertheless consistently and continuously experienced phantoms
of these ‘missing’ limbs. The diagram (from Brugger et al., 2000)7
7 This diagram is reproduced with kind permission of the authors.

238 Robert Clowes
shows her rating of the reality of her limbs on a seven point scale; with
seven signifying a bodily part feels as real as actual parts of her body.
As can be seen, parts of AZ’s virtual body appear to her as almost fully
realised.
The systems that produce this sense of bodily presence appear some-
times limited in their ability to reconfigure to meet unusual or abnormal
body configurations, or simply unable to track changes to the plan of the
body after injury. While explanations of this phenomenon are tentative8,
it seems that body image does have a sort of default topography. But the
pre-reflective sense of the presence, disposition and organic composition
of our bodies seem difficult and sometimes impossible to shake up by
rational inference. When these systems go wrong there can be jarring
inconsistencies between what we see, believe and what we feel. Even
insofar as we know at some level we do not have certain body parts they
can still be felt as present, indeed as almost fully real.
The sub-personal mechanisms of the astronaut with space sick-
ness cannot seem to integrate a stable body configuration in relation to
the immediate environment (a body-world relation). The sub-personal
mechanisms that present the body of the phantom limb patients have
trouble squaring residual (or expected) image with the body’s current
configuration. From these examples it is easy to see why Metzinger
claims that body representation is virtual in the sense that it is not a
straight-forward mirroring of reality but a sub-personally produced
best-guess about the body’s current state and its location in a spatial
frame. But it is less clear why he thinks this implies that our conscious
experiences are of a body that is in some sense unreal.
Take Metzinger’s remark that: “Strictly speaking, and on the level
of conscious experience alone, you live your life in a virtual body not a
real one.” (Metzinger, 2009, p. 114). On the face of it, given Metzinger’s
ultimate argumentative strategy, this comment is difficult to interpret.
Who is this you that lives its life in a virtual body? Metzinger seems
to be simultaneously denying the validity of a concept while trading
on it. Yet there is a deeper problem with what is being said here. Is the
8 There is an interesting discussion of whether the findings about aplasic phantom

limbs should be taken to indicate body-image is innate (2005a). The question goes
beyond the scope of this paper but Gallagher still appears to think it an open one,
awaiting more compelling empirical research.
virtual not part of the real? It has been remarked that counter posing
the real with the virtual is an ontological mistake, a better opposition
may be between the virtual and the actual (Deleuze, 1991). In the case
of the experience of presence it is not so much that what we experience
is unreal but that there is a virtual element to our experience of the ac-
tual. The real question is then, what role do virtual constructs play in
our cognitive lives? Statements that imply we ‘live our lives’ in unreal
bodies elicit more questions than they answer.
To see why, consider how the PSM is a production of bio-rep-
resentational systems in the act of producing coherent action. In the
case of body-representation there is a something, an “image” or mod-
el, that is produced as part of the process of an organism taking con-
trol of itself. But this certainly does not make either the processes
that produce this representational content or the content itself unreal.
Insofar as a subjective sense of presence is the correlate of this fusion
of an adequately coherent representation of our bodies in space, this
also is not obviously illusory. Insofar as PSMs are used as part of a
bodily control system they are as real as all of the neural structures
and bodily structures that realize them. Insofar as PSMs also make
available our capacities for pre-reflective self-awareness, they are as
real and non-illusory as any other aspect of phenomenology. In terms
of content they are a real reflection of a set of processes that seek
to achieve coherent self-control through projecting an overarching
self-unity.
Body image may be virtual in Metzinger’s sense: that it is the brain’s
best guess about the state of the body, yet it is not clear why this should
make body-image, or other self-representational properties, illusory in
any general sense. It merely indicates that they operate within certain tol-
erances. Yet while phantom limb and the astronaut examples are striking
evidence of the variability and sometimes non-veridical character of body
image, any implications this has for our understanding of self requires
further analysis. It is clear that our brains operate given certain assump-
tions, not all of which are fulfilled in all circumstances. The lack of grav-
ity and actual limbs may be the limits of the condition under which our
good representational system can fuse a coherent and reasonably veridical
sense of the body. It is also clear that they have a predisposition to present
240 Robert Clowes
a coherent sense of body, even under circumstances of great difficulty in

integrating that information into a coherent whole.
One way of developing this is to draw an analogy with Noë’s
response to the Grand Illusion. For Noë we are mislead about the
nature of our perception only if we practice bad phenomenology, e.g.
by assuming that the way the world is presented to us is like some pho-
tographic snapshot. But the world is not presented in this way and our
apparent richly detailed access to it is not similarly an illusion (Noë,
2002). By analogy we might argue that if we require for selves to exist
that we need to have a snapshot-like impression of some underlying
reality, then there is an illusion. However, if we relax the constraint and
merely require we are in contact with the bodily basis of self, as and
when needed, then there is no illusion. It is just the way that the brain
fuses a body image, a sense of presence and the bodily basis of self – is
by virtual means. But none of this implies our bodily sense of self is
typically illusory.
In What Sense is the Self Illusory?
Metzinger’s notion of a PSM and his use of the metaphors of virtuality

has not especially drawn positive comment from those working in the
contemporary phenomenological tradition (Gallagher, 2005b; Zahavi,
2005a)9. But it is important to recognise that Metzinger’s theorisation
of the subjective structure of experience has many points of contact with
that tradition and perhaps especially the contemporary approaches to
neuroscience-inflected phenomenology being developed by Gallagher
and Zahavi (2012). In fact, Metzinger’s PSM is supposed to offer an ex-
planation of pre-reflective self-consciousness by grounding its analysis in
terms of functional and neuroscientific analyses.
But this can make the basic point about the self being illusory or
unreal all the more perplexing. If the self is grounded in the body and
9 On the subject of Metzinger’s virtuality metaphors, Shaun Gallagher wonders
whether “Metzinger has become tangled in a matrix of his own self-generated
metaphors” (Gallagher, 2005b).
virtual, why should this imply the scientific elimination of the concept of
the self? In fact, the ‘partition’ in representational space that makes pos-
sible the PSM reflects the way our body is a relatively distinct part of the
world, an organism in fact. The idea that embodied self is unreal merely
because it depends on virtualistic body representation seems, in this light,
bizarre. Nevertheless it is clear that Metzinger thinks the various cases of
body illusion – and therefore in his view, the virtuality of the body – are
supposed to convince us that we should eliminate the self.
I have found it difficult to fully reconstruct the argument in
Metzinger’s work on this point as it often seems that he just takes it that
if the self can be shown to be virtual, in the several senses previously
detailed, this just implies it is ripe for scientific elimination10. Never-
theless, I detect the following lines of reasoning all of which Metzinger
takes to imply the elimination of the self in theoretically interesting
contexts that seek to explain subjectivity and how our minds work.
I. As PSMs are virtualist in the sense they are only best-guesses or

hypotheses about actual states of affairs (5th sense of virtual above),
they are strictly misrepresentations.
II. As PSMs are optimised to be control devices (virtualist in sense 3)
they are not just simplifications or reductions of reality, but tend
to systematically misrepresent reality in terms of an agent’s action
possibilities.
III. As PSMs are pre-reflective modes of accessing reality they tend
also to pre-dispose us to trust the way they represent reality and just
take their presentations as actual. These presentations are relatively
immune from higher level knowledge.
Which Metzinger takes to imply
IV. As PSMs are our basic mode of encountering ourselves, and be-
cause they necessarily misrepresent what we really are, we operate
under incorrigible and systematic illusions about ourselves.
Let us examine these points in a little more detail.
10 In fact Metzinger is very committed to the “no-self ” theory on independent

grounds. See Metzinger (2011).
242 Robert Clowes
Even some of Metzinger’s favourite examples seem to work against

his case for I. The cases of phantom limbs and the astronaut, rather
than demonstrating why we are deluded about the body, seem to show
merely that body representation is contingent on some unexpected but
lawful characteristics of how the body is in fact situated in the world
or internally configured, and the way the mind brings structure to those
representations. The astronaut gets space sickness not because the
connection between bodily content and the body is random but rather
because the relation is contingent on worldly interactions such as the
sensation we normally have of our feet hitting the ground, and indeed a
host of other sensory impressions that are created merely by moving our
bodies around in gravity. When this normal information is absent it may
or may not be surprising that bodily representation cannot be fused in
the usual way, but this hardly indicates we are usually misled.
Similarly in cases of phantom limbs there are number of expla-
nations but all seem to turn on the usual or expected information
the brain would receive from the body being interrupted and what
remains being misinterpreted11. Phantoms are produced when or-
ganic bodily damage (or congenital conditions) produce bodily con-
figurations that cannot in some sense be integrated veridically into
the brain’s pre-conceptual sense of its body. These findings actually
undermine hyper-representational and internalist pictures of mental
representation that picture us experiencing only “world simulation”
or “living our lives in a virtual body”. They also indicate problems in-
herent in internalist ideas of presence where simulations of the body
float entirely free of the actual body’s worldly interactions with its
immediate environment. They rather serve to demonstrate that these
simulations are being fed by contact with the extra-dermal world and
a constant attempt to make the best sense possible of this ongoing
contact.
If the experienced body is virtual in the sense that it is a best guess,
isn’t this a reflection that all such knowledge can never be more than
11 Clearly this point is tricky in the case of aplasic phantoms. Nevertheless if we

assume that at least the core representational systems involved in the representa-
tion of the body are developed to deal with a four-limbed creature, the absence
of limbs, even if congenital, can still be regarded as a case where the absence of
information which would have been normal in the evolution of the system.
provisional? The claim that our representational equipment is based

on best guesses elicits questions about what the successful condi-
tions for our perceptual apparatus are supposed to be: Godlike knowl-
edge? What an examination of body representation does suggest –
contra some influential claims that representational fusion is unneces-
sary for intelligence (Brooks, 1991) – is that the various component
systems that represent the body generally are unified in some way. Nev-
ertheless the various component systems can come into conflict in a va-
riety of ways and these can result in odd perceptual configurations and
senses of bodily self. None of this obviously implies the systems that
produce a bodily sense of self, or underly our experience of presence,
generally misrepresent the way things actually are with our bodies. If
our pre-conceptual sense of presence, including the configuration of our
body, the body’s organic composition, and the way our body fits into a
wider reference frame, appear fallible, this is no different from the fal-
libility of any other sort of perceptual content.
Metzinger’s first argument (I) turns on the series of examples around
malfunctioning body representation that are supposed to indicate the
PSM’s non-veridical contents. While, the examples of the astronaut and
phantom limbs may indeed convince us that body representation is in
various ways fallible and often has non-veridical contents, it is unclear
how this is supposed to indicate such representational content is typi-
cally misrepresenting.
A second line of argument (II) is based around the claim that be-
cause our sense of self is really given by an instrument the primary
function of which is self-control, it misrepresents the way things really
are with us. This line of thought is particularly evident in his recent
book The Ego-Tunnel (Metzinger, 2009), indeed the name of the book
reflects the way in which Metzinger holds that our representational ar-
chitecture does not present a veridical picture of reality but a literally
self-centred construct: An Ego Tunnel.
Sometimes Metzinger runs this argument around the more general
claim that all representation is a sort of simplification and reduction of
reality beyond our perceptual and representational apparatus. On this
analysis representational contents misrepresent insofar as they neces-
sarily reduce and limit the true complexity of their worldly objects. The
problem is that this is not just something that faces the representation
244 Robert Clowes
of self, but all noumena more generally. The representational process is

always in a certain sense a reduction of reality, or so Metzinger claims.
But then, what is special about self-representation? Aren’t the targets
of all presentations similarly approximate? Our grip on them only ever
partial and reductive? All presentations are just ‘tunnels’, i.e. a kind of
collapsing of the world into a reduced form that is mentally graspable.
In fact, as we have seen, there is an important sense in which PSMs
are indeed special, i.e., in the third sense of virtuality noted above:
because they are control systems. The actual content produced by the
PSM is optimised for certain specific types of self-control. It is not an
objective representation, especially in the “view from nowhere” sense
of objectivity. But, this glass can just as easily be portrayed as half-full
as half-empty. Rather than aiming to reconstruct a snapshot of external
reality or a god’s eye perspective on ourselves, it is better to recognise
that, at least at the pre-reflective level, our brains produce a sense of
ourselves and our place within our immediate surroundings – essential-
ly a sense of presence – which is optimised towards the production of
a coherent scene and time-frame in which action is coordinated. This
physical and temporal space of presence can be viewed as much a way
of going beyond the merely actual, as it is a reduction of reality.
To take just one example let’s take the apprehension of “thick”
moments of time in consciousness. What James called “the specious
present”, and Husserl described in terms of the protentive and reten-
tive aspects of consciousness, both gesture toward how our subjectivity
creates or projects a sort of extended now. The example often used to
illustrate this is how we listen to music. When listening to a tune we
do not merely hear the note being played at some exact time-slice of
that instant but have a sense of it being the continuation of a melody
from a few seconds ago, and tending toward a future now being real-
ized. For Husserl (and James) retention was not memory and protention
not foresight but rather – phenomenologically speaking – part of the
moment of consciousness itself. In fact Metzinger’s notion of presenta-
tionality12 is exactly a notion of how consciousness presents the world as
necessarily temporally extended. This extended now is but one way that
12 Metzinger discusses what he call “the presentationality constraint” extensively in

Being No-One as a necessary aspect of conscious representation, see especially
sections 3.2.3 and 6.2.7 (Metzinger, 2004a).
the representational process can be considered an enhancement, rather

than a reduction of reality. Other forms of representational enhancement
make it possible to form impressions not just of how the world is, but
how it might be and how we might affect the world in order to make it
more like we would might like it to be.
It is thus very one-sided to claim that the conscious representation-
al process is simply one of diminishment as it is just as easy to charac-
terize it as one of enhancement. More problematic still is that Metzinger
offers us no standard as to what would be an accurate and veridical
process of representation. For these reasons it is difficult to conclude
that either I or II have demonstrated that we necessarily misrepresent
ourselves in consciousness.
Third, Metzinger argues the PSM forces us to confound ourselves –
presumably the real organismic us – with a type of content produced by
a simulation process in our brains. Or, put differently, from our experi-
ence we take ourselves to be an entity in the world, but this we cannot
be, for all we can experience is a form of content produced by the brain
when involved in certain modelling activities. As Metzinger writes:
In other, more metaphorical, words, the central claim of this book is that as you
read these lines you constantly confuse yourself with the content of the self-model
currently activated by your brain (Metzinger, 2004a, p. 1).
But are we really ‘confusing’ a type of representational content with

an actually existing being. It is odd to view representational content as
things that can be mistaken for worldly entities. If the content of our in-
ternal representations is – qua Metzinger’s hyper-representationalism –
the only things we apprehend, what is the comparison class with which
we are mistakenly comparing them? The natural response here is to
say that we apprehend the world through our representational media.
Metzinger has not shown us at that this way of thinking is a mistake.
Metzinger’s critique of our supposed naïve realism appears to have a
more general problem, however. If we are making a mistake about the self,
aren’t we making similar mistakes about everything else we perceive? But if
this is so, then selves are no more problematic than all of the other medium-
sized targets of our representata. We do not just mistake representations
for ourselves, but also representations for trees, rocks and automobiles.
But Metzinger does not call for the elimination of all of these. Either we
246 Robert Clowes
are global anti-realists or moderate realists about the putative targets of

representation. There is nothing special about selves here. Metzinger does
flirt with a more global anti-realism in The Ego-Tunnel, but this does not
much help underwrite the supposed special unreality of the self. More
over if, as I have argued, Metzinger’s arguments I and II fail to demonstrate
that the PSM does misrepresent reality, then argument III does not follow
in any case. For, if PSMs do not systematically misrepresent, then it has
not been demonstrated we are systematically operating under illusions. On
these grounds at least, there is then no reason to conclude IV: that we are
operating under systematic illusions about ourselves.
Presence, the Minimal Self, and Ipseity Disturbance
One further problem with Metzinger’s apparently radical goal of elimi-

nating the self is, in the field of psychiatry, and clinical psychology, it is
kicking at an open door. Parnas and Handest, writing on schizophrenia
and arguing for the necessity of self: “The notion of a self is deleted from
the terminology of DSM-IV and ICD-10. It is rarely used in psychiatric
literature, and usually in a colloquial or psychoanalytic sense” (Parnas
& Handest, 2003, p. 122)13. As the Diagnostic and Statistical Manual of
Mental Disorders (DSM) and the International Classification of Diseas-
es Manual (ICD) represent the most important and widely recognised
diagnostic manuals used by psychiatrists, clinical psychologist and the-
oreticians alike it is important to realise here that the self as theoretical
category – at least in psychiatric practice – is a category under threat.
Yet there has been a far from universal positive reception of these
moves. DSM’s focus on describing symptoms and presentations of illness
13 In fact it is not strictly true that DSM IV and ICD-10 were entirely successful in
deleting the self. It may not show up in the diagnostic criteria for schizophrenia,
but, as we shall see in a moment, it still appeared necessary in to use the concept
in the diagnostic criteria for both Depersonalisation Disorder (DPD) and Disso-
ciative Identity Disorder (DID); (the latter once known as Multiple Personality
Disorder). Some residual of the self as a theoretical category it would seem is not
so easily annihilated.
rather than providing underlying explanation may be one reason the cat-
egory of the self can apparently be discarded. Parnas, Sass and their col-
leagues have been at the forefront of those arguing that we need, as a mat-
ter of urgency, an adequate rethinking of the concept of self that would
allow us to do justice to the reality of several types of mental illness,
(Parnas, 2000, 2003, 2012; Parnas et al., 2005; Sass, 2003). Historically
speaking, some notion of the distortion of self, self experience, and the
unity of consciousness was central to the very first formulations of the
disorder undertaken by Bleuler and Kraepelin, along with related themes
such as the unity of consciousness.
Commenting on this absence of the self in the contemporary diag-
nostic manuals, Louis Sass and Josef Parnas write:
It has long been recognized that schizophrenia involves profound transformations

of the self. Eugen Bleuler (1911, p. 143) noted that the patient’s ego tends to un-
dergo “the most manifold alterations,” including splitting of the self and loss of the
feeling of activity or the ability to direct thoughts. Kraepelin (1986) considered
“loss of inner unity” of consciousness (“orchestra without a conductor”) to be a
core feature of schizophrenia. (Sass & Parnas, 2003, p. 427)
Thus since its clinical recognition, schizophrenia has often been un-
derstood as a disorder of the self but this has arguably been somewhat
obscured by the clinical concentration on other aspects of the disor-
der, especially on the set of positive (hallucinations and delusions of
control) and negative symptoms (inexpressive faces, monosyllabic or
absent speech, lack of interest in the world) codified by the DSM IV.
Nevertheless, several groups of contemporary researchers have con-
tinued to develop the theoretical links between a series of refined notions
of self and self-experience, and how they can be compromised in a va-
riety of psychiatric illnesses. Arguably, an adequate conceptualization of
several types of altered mental state and psychiatric illness appears to
be virtually impossible to even characterise without making some use of
the notion of the self. Even in the super-positivistic new version of DSM
(DSM-5) it is still apparently necessary to make use of the concept of self
in describing the symptoms of Dissociative Identity Disorder (DID)14.
14 The draft of DSM 5 includes the following explanation as part of the definition
of DID: “Disruption of identity characterized by two or more distinct personality
states or an experience of possession. This involves marked discontinuity in sense
248 Robert Clowes
Another important disorder in which reference to some notion of

the self appears to be a conceptual necessity – and indeed the word
self is still used in both of the main standard diagnostic manuals is15 –
Depersonalisation Disorder (DPD). Patients with DPD often find their
way to a specialist complaining of difficult to articulate “feelings of
unreality”. Patients find it hard to explain or put into words their strange
experiences they encounter but they often fall back on one or both of
two metaphors:
A first kind makes reference to a sense of being cut-off, alienated from oneself and
surroundings. For example, patients would often talk about ‘being in a bubble’,
or being ‘separated from the world by an invisible barrier such as a pane of glass,
a fog, or a veil’. A second group of metaphors emphasise instead a qualitative
change in the state of consciousness, and the feeling as if in ‘a dream’…‘stoned’,
‘not awake’ or an indescribable ‘muzzy feeling’, etc. (Sierra & David, 2011, pp.
99–100).
Two signature features of DPD are depersonalisation: The disturbing

feeling of being separate from oneself, observing oneself from the out-
side, or feeling like an automaton or robot; and derealisation: the threat-
ening sense of the unreality of the environment, a perceptual sense of
unreality, and sense that others are actors in a play. But in fact deper-
sonalisation and derealisation look to be closely interrelated in that it is
the problematic embedding of body (or bodily self) in the world which
appears to be at the root of the disorder:
Depersonalisation disorder involves an unpleasant, chronic and disabling alter-

ation in the experience of self and environment [my emphasis]. In addition to
of self [their italics] and sense of agency, accompanied by related alterations in af-
fect, behavior, consciousness, memory, perception, cognition, and/or sensory-motor
functioning. These signs and symptoms may be observed by others or reported by
the individual.” <http://www.dissociative-identity-disorder.org/DSM-5.html>.
15 DSM IV characterises depersonalisation as ‘alteration in the perception or experi-
ence of the self so that one feels detached from and as if one is an outside observ-
er of one’s mental processes or body’ (American Psychiatric Association, 1994:
pp. 488–490). ICD-10 characterises depersonalisation-derealisation syndrome as
‘in which the sufferer complains that his or her mental activity, body, and/or sur-
roundings are changed in their quality, so as to be unreal, remote, or automatized’
(World Health Organization, 1992: p. 171). Both cited in (Medford, Sierra, Baker,
& David, 2005).
these classic features of depersonalisation and derealisation, symptoms may also

encompass alterations in bodily sensation and a loss of emotional reactivity.16
(Medford et al., 2005)
Depersonalisation and derealization appear to turn, for their conceptual

coherence, on the problematisation of the self/world boundary or, to put
this another way, the pre-reflective experience of how we are embedded
and embodied in the world. This is often referred to as the sense of
presence.
Despite DSM and ICD, today several initiatives are underway to
attempt to understand the onset and underlying aetiology of schizophre-
nia that ties the disorder directly to the basic structure of conscious-
ness, presence and especially the categories of pre-reflective or minimal
self. Much of this work starts from a recognition of the importance of
the phenomenology of schizophrenia (Parnas, 2003; Parnas & Hand-
est, 2003). An important line of this contemporary work represents
the attempt to seriously take account of the subjective experience of
patients with schizophrenia in understanding the disorder; hence the
phenomenological orientation. One influential set of studies has focused
especially on the preonset or prodromal stages – that is the pre-psychotic
stages – of the disorder where patients who will go on to develop full-
blown schizophrenia often report a variety of disturbed and altered ex-
periences of self, presence, ability to think, and agency (Parnas, 2000,
2003; Parnas & Handest, 2003). Some of the most important of these
current theorisations explore the link between anomalous self-experi-
ence, or self/world boundaries in the prodromal stages of schizophrenia
(de Vries et al., 2013; Henriksen & Parnas, 2012; Sass & Parnas, 2003).
Much of this work depends on theorizing schizophrenia, or at least
the prodromal stages of schizophrenia, as a disturbance of the pre-reflec-
tive sense of self, (Parnas, 2003; Parnas, Bovet, & Zahavi, 2002; Sass,
2003; Sass & Parnas, 2003), or, what has been termed ipseity (Zahavi,
2005b). The term ipseity originally comes from the phenomenological
tradition and was especially developed by the French phenomenologist
16 The experience of distorted presence here clearly recall the phenomenology of

space-sickness discussed above.
250 Robert Clowes
Michel Henry (1963)17. Parnas clarifies the term: “We may speak of
ipseity or prereflective self-awareness whenever we are directly, nonin-
ferentially or nonreflectively conscious of our own ongoing thoughts,
perceptions, feelings or pains; these always appear in a first-person
mode of presentation that immediately tags them as our own, i.e. it en-
tails an automatic self-reference” (Parnas, 2003, p. 224). Parnas contin-
ues that it is this dimension of ipseity that make experiences subjective.
Thus, when ipseity is compromised in our experiences, thoughts and
sensations, they can take on a variety of morbid symptoms which can be
characterised as a distorted or diminished unity in our thoughts, body
or sense of presence. It is sometimes also emphasized, that it is the self
pole of presence that is particularly disturbed, or that the pre-experience
of self (ipseity) fails to saturate experience (Parnas, 2003, p. 255).
The notion of ipseity is sometimes used more or less interchange-
ably with the idea of a minimal (or core) self. The minimal self con-
cept was first developed, or at any rate codified, in Gallagher’s (2000)
paper although it was significantly influenced by Strawson’s ‘The Self’
(1999). Strawson had attempted to theorize the most basic sense of self
conceivable, thus the minimal self. The minimal self – in Gallagher’s
sense – is to be understood as hiatus-free, sense of self that is immediate
and non-observational. Importantly, it shouldn’t involve “a perceptual
or reflective act of consciousness” (2000, p. 15) and thus be definition
is pre-reflective. For this reason the term “minimal self ” is often under-
stood as a putative explanation of, or just treated as synonym of pre-re-
flective self consciousness. As will become clear I do not agree with
this analysis in that it seems to be possible and indeed likely that more
extended senses of self can also be treated as pre-reflective dimensions
of experience.
According to Cermolacce, Naudin & Parnas (2007) this minimal self
is no mere conceptual entity but it also confers our basic sense of self
or ipseity, and indeed it is claimed that it is this sense of self – perhaps
more properly of self-world boundaries – which is interrupted in the early
preonset (or prodromal) stages of schizophrenia. In light of this, schizo-
phrenia is understood as a disorder where this basic sense of self, the role
it plays in structuring subjectivity and in unifying mental life, becomes
17 For an excellent discussion of terms development, usage and application to the
understanding of schizophrenia see (Zahavi, 2005b).
altered in one of several ways. It is partly because it is the self pole, or the
self / world relation that is generally thought to be affected in pro-dromal
schizophrenia, that this approach is referred to as the ipseity disturbance
(Sass, 2003), or self disorder, hypothesis (Sass, 2013, In Press)18.
In fact there is now a great deal of research that shows that prodro-
mal patients seem to have very marked and particular symptoms that up
until now have been largely ignored by DSM and ICD, as indeed has
the more general attempt to achieve a phenomenological characterisa-
tion of schizophrenia (Parnas, 2003). Prodromal patients often report a
variety of altered and distorted experiences – especially self experience
– from feelings that body parts are changing, to the feeling that aspects
of the outside world are entering into oneself, from feelings of lack of
agency, to a lack of bodily presence, to feelings that one’s thought is
doubled, or somehow delayed (de Vries et al., 2013; Parnas, 2003; Par-
nas & Handest, 2003).19
These feelings and reports have marked similarities with the feel-
ings of depersonalisation and derealisation that we have already dis-
cussed around DPD. However in addition to these alterations in the
sense of presence, there are variety of other altered experience experi-
ences including: altered sense of the body or its parts feeling heavier or
lighter, smaller or larger, longer or shorter; alterations to the stream of
consciousness especially where one’s thoughts appear to take on quasi-
autonomous character, or one’s thoughts, especially inner speech take
on an object-like character; other phenomena include feelings of disso-
lution of the self or that the self is somehow doubled20; another problem
18 It should be noted however that as the disorder is theorized as routed in an ex-

periential problem of the self world boundary, it may be better to think of it as a
disorder of presence rather than of ipseity per se. See a related discussion in (Sass,
2013, In Press).
19 A recent attempt to catalogue and distinguish the different types of anomalous
self-experience in prodromal schizophrenia and related conditions resulted in
the Examination of Anomalous Experience or EASE scale, “a symptom checklist
for semi-structured, phenomenological exploration of experiential or subjective
anomalies that may be considered as disorders of basic or ‘minimal’ self-aware-
ness.” (Parnas et al., 2005, p. 236).
20 “The most prominent feature of altered presence in the preonset stages of schizo-
phrenia is disturbed ipseity, a disturbance in which the sense of self no longer satu-
rates the experience. For instance, the sense of myness of experience may become
252 Robert Clowes
is the apparent increased subjectivisation of the world, i.e. that reality

is increasingly mind dependent; lack of volition and asociality (Parnas,
2003).
One way of generalizing many of these experience is to say that
“The most prominent feature of altered presence in the preonset stages
of schizophrenia is disturbed ipseity, a disturbance in which the sense
of self no longer saturates the experience.” (my emphasis: Parnas, 2003,
p. 225) or at least those aspects of experience that should be saturated.
One way of interpreting the problem is as a lack of coherence or sim-
ply a lack of – as Metzinger might say – a proper partion of self and
world. That the disorder is typically theorized around the ipseity pole
of experience perhaps reflects the fact that subjects are left feeling that
aspects of their mental life are not really their own: the typical problem
then appears to be one of under saturation. However, prodromal patients
maintain a sense that these are ‘as if’ experiences, that is to say, their
reality testing is relatively intact and they do not believe – as patients
with chronic schizophrenia in full blown psychosis – that they “are the
world”, or that “someone else is causing my thoughts”.
Apart from these various disturbed senses of presence it is clear
that other aspects of subjectivity, whether by direct implication, or as
forms of compensation, also are transformed. One of the most interest-
ing of these is what was generally referred to as hyperreflexivity (Sass,
2003, 2013, In Press; Sass & Parnas, 2003); and sometimes hyperre-
flectivity (Parnas, 2012)21. This is a sort of exaggerated self conscious-
ness, or intense attention to one’s inner life, thoughts, or inner speech
that interferes with one’s ability to think coherently and act in a timely
manner. However, it should not be understood as merely an excessive
form of introspection or reflection on one’s own thoughts (this is some-
times referred to as hyperreflectivity), but typically involves a “pop-
ping out” of internal phenomena, typically not attended to by healthy
subjects (Sass et al., 2013). Patients with hyperreflexivity frequently
subtly affected: one of our patients reported that his feeling of his experience as
his own experience only ‘appeared a split-second delayed’” (Parnas & Handest,
2003, p. 225).
21 The terminology difference turns on Sass’s insistence that hyperreflexivity is not
merely a sort of enhance introspection of self-reflection (Sass, Pienkos, Nelson, &
Medford, 2013). For further discussion of the distinction see (Seigel, 2005).
complain that their normal thinking processes are derailed by this ex-
cessive focus. It may be that hyperreflexivity is a sort of compensation
for feelings of estrangement from one’s body and mind that are central
to ipseity disturbance more generally.
One analysis of hyperreflexivity is that typically transparent mech-
anisms of thought are actually becoming the subject of thinking process
in their own right (hence opaque). What these subjective reports in the
prodromal stages of schizophrenia seem to indicate is that something of
the basic structure of thinking and consciousness is rupturing, or break-
ing apart. It may be that the problem here then is not simply with the
saturation of the relevant parts of experience with ipseity, but with how
the mind typically renders parts of its thinking processes as transparent.
However we analyse this, hyperreflexivity tends to exacerbate the ap-
parently more basic problem of the disruption in the sense of presence
by interrupting and disrupting a patient’s ability to maintain a coherent
and directed stream of consciousness.
As the patient develops more full-blown schizophrenia, these dis-
torted or altered pre-reflective senses of self “may be, and usually are,
followed by the changes in the reflective I awareness and in the social
self, thus resulting in the profound disturbances of identity, clinically
manifest on all […] levels of selfhood.” (Parnas, 2003, p. 219). The
idea that thought, like the body, may increasingly be less “saturated by
self ” may offer a way into making sense of the basic phenomenology
of schizophrenia, and perhaps can be extended to deal with some of the
more social aspects of the disorder (we shall return to this point below).
Such findings around the prodromal stages of schizophrenia are
now crying out for a refined analysis of pre-reflective self conscious-
ness, the integration and unity of the self and the role of self-conscious-
ness in subjectivity and thought. Further attempts to purge the notion of
self from our psychiatric taxonomy seem only likely to harm rather than
aid the development of research into conditions like these conditions.
Metzinger’s contention that the self concept should be eliminated is of
no obvious help to those attempting to study and develop treatments for
such conditions, especially when it is “the phenomenology of disturbed
self-experience” that appears to be at the centre of what clinicians are
trying to treat and theoreticians explain. Further moves to elimination
appear far from “useful to a psychiatrist, engaged in describing mental
254 Robert Clowes
states of his patients” (Parnas & Handest, 2003, p. 122). Conversely,

Metzinger’s ideas of phenomenal self models and the virtual self may
offer us new resources to productively rethink selves, self experience
and their conditions of possibility. The development of refined tools
to think about these issues is surely one of the main contributions of
Being No-One, which makes Metzinger’s ontological orientation all the
more puzzling. Indeed if the ipseity theorists are correct, what we need
is a deeper understanding of ipseity, presence, their interrelations, and
their role in theorising the structures in which thought and body-alien-
ation can occur and Metzinger’s multi-level analysis give us powerful
resources in this regard.
Schizophrenia and Social Self Diminishment
Lysaker and Lysaker’s (2008) book Schizophrenia and the Fate of the
Self offers an account of schizophrenia and its phenomenology. Their
analysis links to the ipseity disturbance hypothesis in that the authors
give great importance to the phenomenology of schizophrenia including,
although not centring on, the transition from prodromal to the advanced
psychotic phases of schizophrenia. They also attempt to understand the
phenomenology and aetiology of schizophrenia in terms of disruptions,
they use the term ‘diminishment’, of our pre-reflective sense of self,
although, crucially not as a minimal sense of self as described above.
Finally, they draw heavily upon the European phenomenological tradi-
tion in developing their analysis, although their main reference point is
Heidegger, rather than Sartre, Henry and Merleau-Ponty22. This said,
the account of self-diminishment they propose is quite different from
the ipseity account, although as I will explore here, it may be possible
to propose a more general account that captures some of the advantages
of each.
The starting point of the book is an attempt to pay careful attention
to schizophrenia from the point of view of how it is experienced by
22 The main reference points of ipseity theorists, see (Zahavi, 2003).

those who suffer with the condition, but while the phenomenological
approach employed has many points of contact with the ipseity ap-
proach discussed above, Lysaker and Lysaker focus especially on how
our social sense of self is compromised by schizophrenia. They focus
on patients that speak of their sense of self being diminished because
they find themselves unable to mentally inhabit aspects of their lives in
relation to others, especially their inability to mentally inhabit custom-
ary social roles such as are necessary at work, in their everyday social
interactions and in their intimate relations to others.
Lysaker and Lysaker argue this diminishment of the sense of self
can at once be understood as a failure to meet and inhabit the social
roles that are expected of normally functioning human beings and the
accompanying inner diminishment that to go along with this, or possi-
bly precedes as the ipseity disturbance hypothesis holds. On the face
of it, this is a very different analysis from disruptions of the self-pole
in distortions of presence, or the disruption of inner life as basic men-
tal unity become decomposed. Before discussing this matter further let
us look in a little more detail about exactly what Lysaker and Lysaker
claim about what they call the dialogic self.
The concept of the dialogic self is developed by Lysaker and
Lysaker to address one aspect of schizophrenia that afflicts many peo-
ple in the advanced stages of the disorder, as patients: find themselves
no longer able to function among others. Their approach aims to give
a phenomenological analysis of changes that take place as the disor-
der develops from the prodromal phases of schizophrenia into its full-
blown psychotic manifestation. They are especially interested in how
patients apparently can no longer relate to the social roles and interests
they once inhabited. Lysaker and Lysaker theorize these problems as an
inability to inhabit what they call self-positions. Self-positions are ba-
sically situation or role specific existential orientations, or ways of be-
ing in the world in particular social roles. Self-positions simultaneously
arise from the social roles we come to inhabit but also are means of
disclosing the world through those social roles, or as Lysaker and Ly-
saker write: “self-positions arise out of the very life that they disclose.”
Self-positions, they theorize, come in three types: character-
positions, meta-positions and organism-positions. The first kind char-
acter-positions, can be understood as social roles which we inhabit
256 Robert Clowes
and through which the world is disclosed to us. Examples Lysaker and
Lysaker use to illustrate this include: self-as-citizen; self-as-daughter;
and self-as-teacher. It is not that all social roles constitute a self-position,
but rather that: “A character-position involves a recurring action-orien-
tation characteristic of one’s being in the world, e.g., citizen and lover.”
(Lysaker & Lysaker, pp. 48–49). In each character position we inhabit,
the sense of self we enact and the world that these positions disclose will
be rather different. Character-positions, while open to reflection, gener-
ally are manifested pre-reflectively: they are “pre-reflective disclosures
concerning action-orientations derived from social roles” (Lysaker &
Lysaker, 2008, p. 54). That it is to say they structure our action and inter-
actions with others without the need for us to reflect on them. Different
possibilities and constraints of action will be afforded by each particular
self-position we pre-reflectively inhabit. As we encounter a situation that
implies we inhabit the self-position of self-as-teacher, or self-as-lover,
we act and indeed perceive and interpret the world from the stance that
particular self-position discloses.
As normally functioning human beings we are able to inhabit and
shift seamlessly and unconsciously between a variety of self-positions,
essentially projecting and inhabiting them as we move between different
social contexts. What should I do as a father, as a lover, as a teacher, as a
clarinettist? Self-positions often provide us with an answer to such ques-
tions before they consciously arise. The particular, cares, joys and action
possibilities that are disclosed to me as a father are distinct but they are
also distinctly part of me and disclose my presence in the world in one of
my various social aspects. Character-positions are situation and role spe-
cific pre-reflective ways of encountering the world from those disclosed
by my other. But they are not minimal ones in the sense of Gallagher’s
minimal self. They imply our worldly roles and they can be as socially
rich as the roles which we inhabit. Crucially if Lysaker and Lysaker are
right, they are also primary ways in which we pre-reflectively experience
our sense of self in the world23.
23 If this line of thinking is correct it implies the ipseity theory as it is currently being
developed may have a crucial lacuna. However it opens possibilities for the theory
to explain what happens as prodromal stages of schizophrenia move to the more
advanced stages. Unfortunately, a full development of the implications of these
The second type of self-position is the meta-position. These are

essentially character-positions plus some evaluative content. Examples
would be self-as-good-teacher, self-as-mediocre-lover. Meta-positions
also generally imply a reflective dimension so we might think of them
as the elevation of our basic ways of modulating social interaction to a
level where we can reflect and tune them. A third type of self-position
is the organism position. Lysaker & Lysaker link these positions to
Damasio’s (2000) core consciousness: “the pre-reflective monitoring of
an organism’s being-in-the-world” (Lysaker & Lysaker, 2008, p. 53).
It virtue of them being pre-reflective and concerned with the body’s
self-monitoring it is natural to make the link here to the minimal self
and the sense of ipseity discussed above24.
Self-positions are thrown up by typical situations we face in the
social world and the problems of action they throw out. They arise out
of the striving within those roles but they are not inner essences, rather
they “unveil time-slices in a life that is still unfolding” (Lysaker &
Lysaker, p. 49).
As Lysaker and Lysaker develop their theory it is central that we
are not identical with any particular self-position but instead exist at
the intersection, or dialogue, of different self-positions. Dialogue for
Lysaker and Lysaker is crucial for it is what essentially generates us.
Their understanding of self – the dialogic self – is so-named because
it is to be understood that we exist and are disclosed as the dialogue
between different self-positions. This idea hints at explanations to why
we are sometime beset by, and cannot answer such as questions such
as “Who am I really?”. According to the self-position theorist, you are
the congruence, or dialogue between the different self-positions you are
able to inhabit. That such questions are thrown-up will always evidence
not just an inner struggle but the problems inherent in being somewhat
theoretical formulations for our understanding of schizophrenia, are beyond the

scope of this discussion.
24 Another possibility is to think of organism-positions, not as actual self-positions
but more as the embodied basis from which other self-positions are projected. Ly-
saker and Lysaker suggest that “self-as-threatened” is an example of an organism
position and why this makes sense as a particular temporary existential orientation
it is less clear why it should be consider a self position. We do not need, I think, to
posit a type of self for every situation or existential orientation the organism faces.
258 Robert Clowes
plural beings in this way. We are not defined by an inner essence but by
a plurality of sometimes competing self-positions.
The central problem for the schizophrenic, according to self-position
theory, is not that the self cannot form a stable self-position, but rather
that it (typically) has become inflexibly fixed in a single self-position.
The theoretical account of Schizophrenia on this analysis helps ex-
plain why it is so often been described in terms of a diminishment of
self, for if the self is precisely the nexus of dialogues between different
self-positions, then to the extent that the self becomes reduced to a sin-
gle position, dialogue is reduced to monologue and our usual abilities to
shift self-positions is curtailed or altogether undermined.
Importantly, phenomenological reports of diminishment are ex-
plained in the following way. Self-positions endow us with a substan-
tive pre-reflective sense of self that links what we do, and how we
behave with who we take ourselves to be. When this everyday ability
to inhabit self-positions becomes inhibited for whatever reason, we
find ourselves diminished. Lysaker and Lysaker hold that the essential
problem faced by the schizophrenic patient is of trying to relate to a
variety of social situations for which they cannot find, or more appo-
sitely inhabit, appropriate self-positions. This, is according to Lysaker
and Lysaker explains the phenomenology of diminishment; they do
not feel themselves adequate to be in the situations in which they find
themselves. It is also why so many schizophrenics are unable to oper-
ate in the social contexts, that prior to the onset of the disorder they
faced unproblematically.
One of the great advantages of Lysaker and Lysaker’s position is
that it helps make sense of a variety of the claims made by schizophren-
ic patients and the phenomenology of schizophrenia, in particular the
widely agreed phenomenology of diminishment which is a central – al-
though under-discussed – aspect of the condition. Dialogic theory is in
many ways constructed to, and deals with, the particular problems faced
by schizophrenics in the social world. It focuses on how schizophrenia
is not only a catastrophic break-down of an inner unity but of the ability
to project a sense of self in the world. One might further link the ap-
proach to ipseity theory by postulating that just as ipseity has become
unstable in other aspects of the subject’s experiential life, so problems
have developed in the ability to invest self-positions with their typical

ipseity25.
Self Positions: An Extended Virtual Self ?
I want now to make an additional claim that while developed within the
spirit of Lysaker and Lysaker’s work, also marks an important theoret-
ical departure. Lysaker and Lysaker hold that the phenomenology of di-
minishment can be explained in terms of the failure of dialogue among
self-positions. They more or less assume that healthy subjects just possess
self-positions which exist in dialogue with each other. While I certainly
would not deny that there can be dialogue between self-positions, it may
be our feelings of being a coherent self, or having a coherent sense of self,
are not grounded in any particular inner unity that requires dialogue. It may
rather be that our everyday experience of the plenitude of self, like the more
general plenitude of perceptual consciousness is a rather virtual one. It is
grounded in our generally fluid ability to engage and operate within so-
cially adequate self-positions as the need arises in order to inhabit socially
adequate roles. Pre-reflective self-experience is however not based on any
deeper inner unity. In order to make this case in this and the next section I
will draw together ideas about self-positions as just outlined with the ideas
of virtual self we discussed in the earlier part of this paper.
On this analysis the self-positions we (typically) unconsciously
adopt, and the pre-reflective sense of self they engender, is self enough
without further unification. While sometimes the potentially contradic-
tory nature of the multiple self-positions we inhabit, may require us
to engage in processes of further unification we might call dialogue –
whatever that is – the worldly roles we inhabit, are generally enough
to provide the needed degree of stability without the need for outright
25 This is just one very schematic attempt to unify self-position theory and ipse-
ity theory. Really developing the points of contact between these theories and
whether they are really compatible in explaining schizophrenic phenomenology
is beyond the scope of this paper although it would be a highly valuable area to
research.
260 Robert Clowes
unification. We are relatively unified, or unified just enough in order slip

between self-positions, but we are not in general very sensitive to the
way that this movement implies that we face the world as rather differ-
ent beings depending on the different self positions we inhabit26.
This suggests a way of approaching the self-position account of
self in virtualist terms. It is central to Lysaker and Lysaker that some
unification of self beyond self-positions take place in dialogue. And yet
the theory is rather sketchy about how this happens. One way of dealing
with this problem is to argue that the unification of self-positions is it-
self a rather worldly matter. We seamlessly move between self-positions
because the social world we inhabit is generally there to allow and sup-
port the transitions. This is not to say that there is no dialogue between
self-positions, nor, when problems arise that there is not a need to me-
diate between our various social roles. However, insofar as we have a
pre-reflective sense of self which is engendered by these self-positions
it is not necessary that they strictly span any incoherencies, merely that
we can continue to employ them as needed.
By analogy with the enactive theory of perception that we discussed
earlier, it may simply be that we are typically able to unproblematically
apply, that is, inhabit the relevant self-positions as and when required
and simply do not worry much that all of the roles we are able to inhabit
do not overlap in entirely coherent ways. Dialogue between self-posi-
tions, insofar as it happens, may occur much more implicitly than the
word dialogue seems to imply. That is to say that while we – who are
not living through serious psychiatric difficulties – may pre-reflective-
ly feel a sort of plenum of self, but that experience is based on our
ability to project and inhabit self-positions as and when required. One
theoretical advantage of this position is that we no longer have to ex-
plain any further unified self beyond self-positions, nor develop a much
fuller account of what dialogue between self-positions actually entails;
(something Lysaker and Lysaker’s formulation still appears to owe us).
Our pre-reflective self-experience may appear to be a plenum in virtue
of our ability inhabit different self positions as and when required, not
26 There are of course important similarities between this proposal and the proposals
of Erving Goffman in The Presentation of Self in Everyday Life (Goffman, 1959).
Unfortunately there is no space to pursue the various similarities and differences
of approach here.
on the manufacture of inner dialogues27. The apparent coherence of self

may be a largely virtual matter depending on our ability to summon and
deploy, as and when needed, a coherent self-position.
This approach may also offer an important novel insight into some
symptoms of schizophrenia. It may be that the development of full-
blown schizophrenia happens as patients become increasingly aware of
the typically ignored absences and incoherencies between self-positions.
This may link to what Sass (2003) calls hyperreflexivity, i.e., a sort of
exaggerated focusing on self as an object of attention rather than as
a transparent medium through which we experience and engage with
the world. Hyperreflexivity on this analysis is engendered precisely be-
cause the normally ignored virtuality of the self has become salient and
a problem for the subject.
This reconstruction of Lysaker and Lysaker’s idea of self-positions
in terms of the grounded virtualist theory introduced above, allows us
some useful theoretical innovations. Perhaps the most important is that
it allows us to link the idea of a pre-reflective sense of self with our
social nature. It is often taken (although this is by no means theoreti-
cally necessary) that the pre-reflective sense of self is identified with a
minimal, typically rather biological sense of self. Such a view perhaps
turns on the theoretical treatment that sees minimal selves and narra-
tive selves as two theoretical poles. What I want to propose here is that
our pre-reflective sense of self, at least phenomenologically speaking,
is typically socially rich and inflected with a sense of who we are in the
social world. It may be possible to draw out a conceptually separate
minimal pre-reflective sense of self but this is not how experience con-
fronts us.
Let us take one example which may give some grounding to this
theoretical contention. Imagine you are a teacher who has just made a
terrible discovery. Your star student, whose work you have been showing
off to colleagues for months, turns out to have plagiarized everything.
You find this out after another colleague has raised a doubt and you
run an essay through a plagiarism database. As the confirmation comes
back you feel a sickness in the pit of your stomach. The feeling here is
27 It must be pointed out that the manner or mode in which dialogue is implemented
intra or inter-subjectively is decidedly under theorized in Lysaker and Lysaker’s
book.
262 Robert Clowes
crucial and connects to our pre-reflective sense of self. But the feeling
may not come from what Lysaker and Lysaker would call an organ-
ism position, but a character or more properly a meta-position: Self-as-
proud teacher. The feeling here is distinctly embodied – the pit in the
stomach feeling – but entirely targeted at and suffused with one’s sense
of self in a particular self-position. At the phenomenological level at
least, the feelings are entirely fused and it appears as if it is impossible
to at least phenomenologically dissociate a minimal pre-reflective sense
of self from the wider, socially inflected senses of self which are given
in the social roles we inhabit.
Rethinking the Virtual Self: Actualizing Virtualities
There are many different types of virtual objects in our minds, impli-
cated in our perception of the world, and even in the way things appear
to us. Reflections in mirrors for example can be considered a type of
virtual object, albeit in this case, they are mere appearances. However,
according to Metzinger’s own proposal, the self, the PSM, is virtual in
a special sense that it different from a mere appearance. The PSM oper-
ates as an interface that makes available a multimodal sense of the body
as an integrated whole. In so doing it allows us to operate as a more
coherent entity in space and time. PSMs, and thus the self, are what we
might call an actualizing virtuality28. Actualizing virtualities are forms
of projection that present coherent possibilities for the agent which can
be made actual by taking the actions they imply or simply by inhabiting
a stance that makes those action possible.
If this is right, and insofar as they play a central role in our con-
trol architecture and also in the way we pre-reflectively experience our-
selves in the world, they play a crucial role in our existence as agents.
But PSMs are also central constituents of the control structure of crea-
tures like us that are self-aware and have a sense of themselves enduring
28 NB., The terminology is mine, not Metzinger’s, although I think the idea is implic-
it in his concept of the PSM.
over time. In virtue of this they must be considered a part of the real.
The content of PSMs is virtual but the production of this virtual content
plays a central role in making us what we are for ourselves, i.e., in con-
ferring ipseity. But this I think is a clear problem, not an argument for
Metzinger’s case that the selves are especially unreal, or their content
illusory. Rather, PSMs are virtual in a special way. They are, or are at
least at the core of, what makes it possible for us to know ourselves and
to realize ourselves through our actions in the world.
Metzinger himself is uncompromising about what this virtual na-
ture of the PSM means for the existence of selves. He writes:
No such things as selves exist in the world. At least their existence does not have
to be presupposed in any rational and truly explanatory theory. Metaphysical-
ly speaking, what we called “the self ” in the past is neither an individual nor a
substance, but the content of a transparent PSM. There is no unchanging essence,
but a complex self-representational process that can be interestingly described
on many different levels of analysis at the same time. (Metzinger, 2004a, p. 626)
But isn’t this enough? It is really only Metzinger’s prior assumption that
the self should be an “unchanging essence” or nothing at all? While there
is certainly precedent in the Western philosophical tradition for such a
claim, from either the standpoints of scientific or folk psychology there
is no real reason the self needs to be considered an unchanging essence
in order for it to play the main conceptual roles required of it. Indeed we
have seen above how the idea that ipseity, developed in the phenomeno-
logical tradition, in many ways play the sorts of conceptual roles that have
been required of the self in much philosophical discussion.
In making this claim, it is of course open to Metzinger to show
that scientific and folk psychology require the self to be an individual,
substance, or unchanging essence. But this he has not undertaken. The
real question is whether we can plausibly reconstruct folk and scientific
psychology around the idea of the virtual self and whether it is worth
doing this. It certainly seems premature to abandon the idea altogether.
In fact I think there are several plausible ways of reconstructing the
concept. One resolution is offered by Marcello Ghin (2005) to the effect
that we should identify the self with the organism that projects a PSM.
I think this suggestion has much to recommend it, including offering
a plausible way of reconstructing folk psychology. In fact Metzinger
264 Robert Clowes
concedes as much in his (2006) response to Ghin’s critique, and also his
responses to (Hobson, 2005).
Another plausible (and to me more interesting) way of reconstruct-
ing the self is to agree that it is indeed virtual in many of the senses that
Metzinger claims, but this is enough for the conceptual purposes we re-
quire of the concept. The self then, at its root, is a form of pre-reflective
experience, very much as is claimed by those accounts of pre-reflective
sense of self developed in neo-phenomenological accounts by Zahavi and
Gallagher, but capacious enough to include the forms of pre-reflective
self-experience discussed by Lysaker and Lysaker. I want to embrace this
analysis while claiming that in important respects our pre-reflective sense
of self is indeed virtual. Selves on the revised analysis presented here are
a form of content produced by an interface that determines central dimen-
sions of our presence in the world, especially, the social world.
On this analysis it is true we are not selves in a strict sense but
we possess, or better, are in touch with ourselves at a pre-reflective
level. My claim here – contrary to the idea that our pre-reflective
sense of self need be minimal – is that it is rather a more expansive
and socially inflected sense of being. We do not have to identify our
pre-reflective sense of self with the mere actual registration of the
body. We can instead think of the pre-reflective self as virtual, pro-
jective, and while perhaps based in body representation, going signif-
icantly beyond the body. Insofar as self is composed of virtual social
(self-position) projections over bodily content, it can link up the idea
of a pre-reflective and indeed embodied sense of self, with the idea
that our pre-reflective sense of self is not minimal at all, but irreduc-
ibly social. The virtualist idea of self as developed here through the
prism of self-positions, gives us an alternative which links our social
nature to the structure of our experience.
The Virtuality of the Self and its Ontological Status
Metzinger’s search for the self is hardly exhaustive. He more or less

assumes a very classical – even Cartesian – notion of the self as a sort
of inner essence. He then finds that this notion is unsustainable in part

because the virtual self generated by phenomenal self-models cannot
be reduced to such a hypothetical inner essence. But given the account
just offered we can note that there are other possibilities. On the vir-
tualist version of self-position theory, selves are not inner essences
but rather a pre-reflective capacity to inhabit multiple self-positions.
And self-positions are crucially worldly things. They come from our
abilities to interact with – in the terminology to interface – with affor-
dances of the social world. Crucially this does not render selves unreal
in any sense that should cause a problem for science nor for much of
folk-psychology. Rather it gives us a way of making sense of these at
both a neuroscientific and social level of important aspects of self-
experience.
The self can be identified with certain forms of virtual content;
forms which can be understood as projections from an embodied basis
and that we typically apprehend pre-reflectively in experience. But while
our pre-reflective sense of self may be body based and body inflected it
is also substantially composed, in any particular case, of self-positions,
i.e., particular social roles that we inhabit as persons. These come with
attitudes, responsibilities and are enforced by social expectations. We
inhabit these self-positions pre-reflectively and so they can be linked –
like our embodied sense of self – to an intrinsic sense of subjectivity.
Self-positions are worldly, but they are not minimal.
I claim that the self-positions are another form of virtual content,
but that Theorizing pre-reflective self-experience in this way allows us to
make sense of our worldly experience. That is, not just as bodies attempt-
ing to integrate coherent motor action, but as social beings attempting to
live coherent lives. The virtual self on this analysis is not best identified
with a projected coherent body image, but with projected whole social
being or presence which we use to inhabit the social world. When our
ability to inhabit this social presence fails, as may happen for a variety
of reasons, our ability to function as coherent human beings itself can
become deeply undermined.
Let us then summarise how we have rethought the idea of a virtual
self. To recap Metzinger argued the sense of self was virtualist in the
following respects:
266 Robert Clowes
1. Selves are virtual because they are simulations; or more precisely,

simulational content.
in representational space that do not necessarily respect any divi-
sions in reality.
available (and obscure) certain possibilities for action.
4. Like the interfaces in virtual reality devices, selves (PSMs) convey
a sense of presence.
5. The contents of PSMs are also virtual because they tend not to
correspond to, or depict, actual states of affairs in the world, but
best-guesses or mere possibilities about the state of the body.
I believe self can be considered virtual but in the following revised

senses:
1. Selves are virtual because they are actualized projections into the
social world.
in representational space and arise from respecting the (also vir-
tual) social roles that we inhabit as individuals.
available (and obscure) certain possibilities for action. The par-
ticular possibilities for action they track, however, are primari-
ly in the social world of interpersonal intercourse.
4. Like the interfaces in virtual reality devices selves (PSMs) convey a
sense of social presence disclosed by the self-positions we inhabit.
5. The content of PSMs are also virtual because they tend not to cor-
respond to, or depict, actual states of affairs in the world, but ide-
alisations and projections of the roles we inhabit in the social
world.
The hypothesis here is that our ability to inhabit and project self-positions
provides much of the pre-reflective sense of self which allow us to operate
as coherent social beings. Yet the ability to project an overall coherent sense
of self is nevertheless virtual. This is not to deny that dialogue between
self-positions might sometimes be necessary, nor that there are not other
internal and worldly processes that help us integrate a coherent sense of
self 29. But in general we have a rich pre-reflective sense of self because of
our abilities to fluidly inhabit the social world, not because we can always
access the very same sense of bodily self anchored in the body.
There are two central implications for the theory of self as thus
conceived. First, the theory offered expresses continuity with traditional
theorizing that made the explanation of self, in this case reconceived
around the idea of virtual self-positions, central to the understanding
of certain forms of psychiatric illness. There is nothing unique here as
many other psychiatric theories also posit the self as a central explana-
tory unit. Insofar as Metzinger argues that there will be no explanatory
loss if we jettison the self we need to understand how its replacement
offers us any explanatory advantage. Metzinger, although he gives ex-
tended analysis of many psychiatric illnesses, is practically silent on the
question of how the PSM would help explain a central aspect of schiz-
ophrenia, namely the phenomenology of diminishment. And yet if the
account here is on the right lines we have a good way of dealing with
this problem. If the projection of the PSM is virtual in the sense of us
experiencing a plenum when what we really do is inhabit self-positions,
a reconceived PSM can help us explain this aspect of schizophrenia.
Schizophrenics feel diminished because their abilities to express, and
inhabit self-positions are diminished, because their interface to the so-
cial world through the projection of multiple appropriate PSM is, for
whatever reason, curtailed and crucially this must compromise both
their ability to operate as social beings and their sense of who they are.
In terms of psychiatric explanation, if on no other grounds, the elimina-
tion of self seems premature.
The second implication is how congenial a revised self-position
based theory is to non-eliminativist approaches to self. Let us cast our
minds back to the discussion of Dennett’s virtual presence discussed
at the beginning of this paper. On this approach perceptual richness
can be considered to be based on a lack of epistemic hunger about
certain internal informational absences. Many of these absence can be
29 NB – We should not overlook here how our sense of self is constrained by an array
of folk-psychological practices which help to produce being that are more unified
and have the greater appearance of being more unified (Zawidzki, 2008).
268 Robert Clowes
made good by the ability to lean on the world when necessary in order
to bring in information as required; information that is typically left
in the ambient environment. In similar ways we experience our social
presence in the ways it is disclosed to us through the self-positions we
inhabit.
For we human beings at least, the body may be the ground for pro-
jecting a self among others. But if the claims here are correct the expec-
tations that synthesize a self are derived from a social realm. The self
is centrally an instrument produced by our brains in order to minimally
integrate our actions as an individual agent and more expansively in the
multiple social relationships in which we are involved. Its unity may
be more projected and inhabited as a result of our abilities in the so-
cial world rather than based on any particular unities in the brain. This
conception has several predictions. The self as such can be dissolved
or disturbed in several ways, but to roughly dichotomise, we could say
this can happen in top-down or bottom-up ways. Bottom-up ways in
which the self may be compromised include when bodily content is
distorted, attenuated or otherwise no longer able to be integrated or syn-
thesized into a coherent construct (e.g. Parnas, 2003). But it may also
be compromised in a top-down manner where bodily content cannot
be integrated into to the currently active set of social expectations. Our
sense of self may become problematic for us in multiple ways and there
may be multiple routes to the forms of diminishment of self we find in
schizophrenia.
But does the virtuality of self imply its non-existence, or the ne-
cessity of an elimination of its use in science? Only if one adopts an
extremely traditional – namely Cartesian – view of the self as an inter-
nal substance or immaterial spirit. Even reducing the self to a PSM-like
partition within the representational space and grounded in the home-
ostatic processes of the body is not obviously a way of rendering the
self inexistent. Rather it is a way of reducing the self to a particular sort
of representational content / process involved in maintenance of the
agent30. As I have begun to argue, such a picture is radically incomplete
30 As this content is playing an ongoing role in controlling the action profile of the
system it might be regarded as an actualising virtuality. A content produced by
certain bio-systems to project presence in order to play a particular role in the
unification of action; or as Metzinger puts it, “a real-time world-model that can be
in that it does not supply an explanation of the apparent richness of

pre-reflective self experience. The version of self-positions defended
here is both able to make sense of how embodiment enabled self, but
also why the idea of the minimal self never seems to have a enough sub-
stance to explain central problems of the self such as personal identity
or even the question ipseity, that is, what really makes us ourselves. But
as I have argued, the self is best conceived of as an intrinsically social,
or worldly, virtual projection. The virtualist view of self presented here,
far from implicating the need to eliminate the self, invites us to inves-
tigate further how it might be grounded in neurologically describable
systems and the social processes we inhabit.
So is this a reduction or elimination of the self as traditionally con-
ceived? The self as conceived here is a socially targeted construct –
a virtual projection – anchored in bodily representation but projected
through a rich set of expectations about our social roles and synthesized
by an underlying sense of coherence and structure. Our pre-reflective
sense of self, it is in large part not constituted by our basic sense of pres-
ence, but in our social self-positions that largely give us our identity.
Actualizing Virtualities and the Conceptual Role of Self
In this article I have claimed: that the self can be understood as a vir-
tual construct whose core is bodily content but the principle function
of which is to allow us to inhabit the social world. On this model the
pre-reflective experience of self derives from the way the experienced
body is projected into a social world. The character of this experienced
self is not limited to natively endowed expectations and modes of inte-
gration. Rather it is always an interactive construct, drawing much of
viewed as a permanently running online simulation allowing organisms to act and

interact” (2004a, p. 104). This looks like a way, as Damasio argued, of giving the
self some substance rather than suggesting its elimination, as Metzinger contends.
There is little reason to endorse elimination if ‘virtual selves’ can confer the right
sorts of unity and integrity that we would expect from the traditional use of the
concept. The contention is that the virtual self can play the roles required.
270 Robert Clowes
its character from the social world, especially the self-positions we are
required to articulate in order to exist as social beings. A great advan-
tage of this version of the virtualist view is that it does not imply any
identity between the self and the body, or imply that minimal self is
somehow the real or more authentic self 31.
Rather we can – following Lysaker and Lysaker – highlight how
the retreat into a more fixed or minimal sense of self can be a dramatic
problem for the human subject. As Lysaker and Lysaker’s analysis of
schizophrenia makes abundantly clear, the failure of an individual’s
ability to project self-positions consistently and appropriately under-
mines or diminishes one’s sense of self.
The virtualist scheme requires the projection of socially refined
senses of self around an embodied core. Thus we can understand these
projections as extensions of the core rather than narrative fictions
(Dennett, 1992). A further potential advantage of this approach is that
self is identified neither with the body, nor with groundless narratives,
but that the two poles are unified32. This is not to claim the virtual self
is not embodied. The self may be grounded in the body in multiple
ways including the production of homeostatic content that is at the
heart of one important embodied theory of self (Damasio, 2000)33. But
it is given structure and definition through the process of virtuality
projection.
Does this virtualist conception of the self imply compelling rea-
sons to eliminate the self-concept as we inherit it from the Western
tradition? I think not. Instead the self, somewhat conceptually revised
perhaps, still seems to be a necessary concept in understanding the
constitution of agents like us. On the contrary, the revised account
helps us see how the self can be virtual, while still being a proper
part of the real and central to the operation of agents like us. Selves
31 It is worth noting in passing that there may be, indeed almost are, other ways in which
we can be said to have or be selves. This article is really groundwork in a theory of self
which would have to be incorporated into a wider network of theories and explanation
(Gallagher, 2013). There may be other forms of self than discussed here.
32 The basic tension between these two ways of seeing things is concisely developed
in (Gallagher, 2000).
33 In fact I think aspects of Damasio’s approach highly congenial to the idea of the
virtual self presented here. Unfortunately making these links goes beyond the
scope of the current discussion.
are best not conceived of as Cartesian essences but this does not im-
ply their inexistence. Once the content of selves is thought of as tied
up with inhabiting the social realm their necessity becomes apparent.
Selves are real, indeed they are actualizing virtualities and that is good
enough – at least provisionally – for them to play the required role in
our conceptual scheme.
Acknowledgements
I want to express my heartfelt thanks to the Cognitive Foundations of Self

Group at the IFL, New University of Lisbon especially the long-suffer-
ing editors of this volume Jorge Gonçalves and João Fonseca for their
extraordinary forbearance in allowing me to revise this text long after it
was due. Special thanks also go to Shaun Gallagher who first suggested
I should look at Lysaker and Lysaker’s book as a way of developing my
ideas about schizophrenia. Dina Mendonça and Michael Baumtrog also
deserve special thanks for commenting on advanced drafts of this paper.
Many thanks also to other members of the Cognitive Foundations of Self
Group to whom I presented several versions of this paper, especially Vera
Pereira, Inês Hipólito and Caio Novais, and to Sean Bell for helping me
copy-edit several versions of it. All remaining problems are of course my
own. Although this paper was begun while I was a tutor at the Universi-
ty of Sussex its development was supported by a grant from Portuguese
Foundation for Science and Technology grant (SFRH/BPD/70440/2010).
References
Barry, P. & Phillips, T. (2001). Mixed up in Space. Science@ NASA,

National Aeronautics and Space Administration.
Botvinick, M. & Cohen, J. (1998). Rubber hands ‘feel’ touch that eyes
see. Nature, 391(6669), 756–756.
272 Robert Clowes
Brooks, R. (1991). Intelligence without Representation. Artificial Intel-

ligence, 47, 139–160.
Brugger, P., Kollias, S. S., Müri, R. M., Crelier, G., Hepp-Reymond,
M. C. & Regard, M. (2000). Beyond re-membering: Phantom sen-
sations of congenitally absent limbs. Proceedings of the National
Academy of Sciences, 97(11), 6167.
Cardinali, L., Brozzoli, C., & Farne, A. (2009). Peripersonal space and
body schema: two labels for the same concept? Brain topography,
21(3–4), 252–260.
Cermolacce, M., Naudin, J. & Parnas, J. (2007). The “minimal self ”
in psychopathology: re-examining the self-disorders in the schizo-
phrenia spectrum. Consciousness and cognition, 16(3), 703–714.
Clark, A. (2008). Supersizing the Mind, Oxford: Oxford University
Press.
Clark, A. (2012). Whatever next? Predictive brains, situated agents, and
the future of cognitive science. Behav. Brain Sci, 1–86.
Clowes, R. W. & Chrisley, R. (2012). Virtualist Representation. In-
ternational Journal of Machine Consciousness, 04(02), 503–522.
Dainton, B. (2006). Stream of consciousness: Unity and continuity in
conscious experience: Taylor & Francis US.
Damasio, A. R. (2000). The Feeling of What Happens: body, emotion
and the making of consciousness,London: Vintage.
de Vries, R., Heering, H. D., Postmes, L., Goedhart, S., Sno, H. N. & de
Haan, L. (2013). Self-Disturbance in Schizophrenia: A Phenome-
nological Approach to Better Understand Our Patients. The Prima-
ry Care Companion for CNS Disorders, 15(1).
Deleuze, G. (1991). Bergsonism (H. Tomlinson & B. Habberjam,
Trans.). New York: Zone Books.
Dennett, D. C. (1991). Consciousness Explained, Boston: Little Brown.
Dennett, D. C. (1992). The Self as a centre of Narrative Gravity. In
F. Kessel, P. Cole & D. Johnson (eds.), The future of folk psycholo-
gy, Cambridge: Cambridge University Press.
Dennett, D. C. (2001). Surprise, surprise. Behavioral and Brain Scienc-
es, 24(05), 982–982.
Frank, M. (1991). Selbstbewußtsein und Selbsterkenntnis: Essays zur
analytischen Philosophie der Subjektivität. Stuttgart: P. Reclam.
Gallagher, S. (2000). Philosophical conceptions of the self: Implications

for cognitive science. Trends in Cognitive Sciences, 4(1), 14–21.
Gallagher, S. (2005a). How the body shapes the mind. New York: Ox-
ford University Press.
Gallagher, S. (2005b). Metzinger’s Matrix: Living the Virtual Life with
a Real Body. PSYCHE, 11(5), 5.
Gallagher, S. (2013). A Pattern Theory of Self. Frontiers in Human
Neuroscience, (7).
Gallagher, S., & Zahavi, D. (2012). The phenomenological mind. New
York: Routledge.
Ghin, M. (2005). What a self could be? PSYCHE, 11(5).
Goffman, E. (1959). The Presentation of Self in Everyday Life. Garden
City, NY.
Grush, R. (2004). The emulation theory of representation Motor control,
imagery, and perception. Behavioral and Brain Sciences, 27(3).
Henriksen, M. G. & Parnas, J. (2012). Clinical manifestations of
self-disorders and the Gestalt of schizophrenia. Schizophrenia bul-
letin, 38(4), 657–660.
Henry, M. (1963). L’Essence de la Manifestation. Paris: Presses Univer-
sitaires de France.
Hobson, A. (2005). Finally Some One. PSYCHE, 11(5), 5.
Johnson-Laird, P. (1983). Mental models: Towards a cognitive science
of language, inference and consciousness. Cambridge MA: Har-
vard University Press.
Lakoff, G. & Johnson, M. (1999). Philosophy in the Flesh. New York:
Basic Books.
Legrand, D. (2007). Pre-reflective self-as-subject from experiential and em-
pirical perspectives. Consciousness and cognition, 16(3), 583–599.
Lombard, M. & Ditton, T. (1997). At the heart of it all: The concept of
presence. Journal of Computer Mediated Communication, 3(2).
Lysaker, P. H. & Lysaker, J. T. (2008). Schizophrenia and the fate of the
self: Oxford University Press, USA.
Maravita, A. & Iriki, A. (2004). Tools for the body (schema). Trends in
Cognitive Sciences, 8(2), 79–86.
Medford, N., Sierra, M., Baker, D. & David, A. S. (2005). Understand-
ing and treating depersonalisation disorder. Advances in Psychiat-
ric Treatment, 11(2), 92–100.
274 Robert Clowes
Metzinger, T. (2000). The subjectivity of subjective experience:

A representationalist analysis of the first-person perspective. In
T. Metzinger (ed.), Neural Correlates of Consciousness (285–307).
Cambridge, MA: MIT Press.
Metzinger, T. (2004a). Being No One: The Self-Model Theory of Subjec-
tivity. Cambridge, MA: Bradford Book.
Metzinger, T. (2004b). The Subjectivity of Subjective Experience:
A Representationalist Analysis of the First-Person Perspective.
Networks, 3(4), 33–64.
Metzinger, T. (2006). Reply to Ghin. In Self-Sustainment on the Level of
Global Availability. PSYCHE, 12(4).
Metzinger, T. (2011). The no-self alternative. In The Oxford Handbook
of the Self, 279–296.
Minsky, M. (1985). Society of mind. New York: Simon and Shuster.
Noë, A. (2002). Is the Visual World a Grand Illusion? Journal of Con-
sciousness Studies, 9(5–6), 1–12.
Noë, A. (2004). Action in Perception. Cambridge, MA: Bradford Books,
MIT Press.
O’Regan, J. K. (2002). The Grand Illusion and Sensorimotor Approach-
es to Consciousness. Journal of Consciousness Studies, 9(5–6).
O’Regan, J. K. & Noë, A. (2001). A sensorimotor account of vision
and visual consciousness. Behavioral and Brain Sciences, 24,
939–973.
Parnas, J. (2000). The self and intentionality in the pre-psychotic stages
of schizophrenia. Exploring the self: Philosophical and psycho-
pathological perspectives on self-experience, 115–147.
Parnas, J. (2003). Self and schizophrenia: a phenomenological perspec-
tive. The self in neuroscience and psychiatry, 217–241.
Parnas, J. (2012). The core Gestalt of schizophrenia. World Psychiatry,
11(2), 67–69.
Parnas, J., Bovet, P. & Zahavi, D. (2002). Schizophrenic autism: clinical
phenomenology and pathogenetic implications. World Psychiatry,
1(3), 131.
Parnas, J. & Handest, P. (2003). Phenomenology of anomalous self-

experience in early schizophrenia. Comprehensive psychiatry,
44(2), 121–134.
Parnas, J., Møller, P., Kircher, T., Thalbitzer, J., Jansson, L., Handest,
P., et al. (2005). EASE: examination of anomalous self-experience.
Psychopathology, 38(5), 236–258.
Pezzulo, G. (2008). Coordinating with the future: the anticipatory na-
ture of representation. Minds and Machines, 18(2), 179–225.
Radovic, F. & Radovic, S. (2002). Feelings of unreality: A conceptual
and phenomenological analysis of the language of depersonaliza-
tion. Philosophy, Psychiatry, & Psychology, 9(3), 271–279.
Ramachandran, V. S. & Blakeslee, S. (1998). Phantoms in the brain.
New York: Harper Collins.
Ratcliffe, M., Slaby, J., Manzotti, R., Stephan, A., Holzhey-Kunz, A.,
et al. (2008). Feelings of being. Oxford: Oxford University Press.
Revonsuo, A. (1995). Consciousness, dreams and virtual realities. Phil-
osophical Psychology, 8(1), 35–58.
Revonsuo, A. (2006). Inner Presence: Consciousness as a Biological
Phenomenon. Cambridge MA: MIT Press.
Sass, L. A. (2003). Self-disturbance in schizophrenia: hyperreflexivity
and diminished self-affection. In T. Kircher & A. David (eds.), The
self in neuroscience and psychiatry (Vol. 870539117): Cambridge
University Press Cambridge.
Sass, L. A. (2013, In Press). Self-Disturbance and Schizophrenia:
Structure, specificity and pathogenesis (Current Issues, New Di-
rections). Schizophrenia Research.
Sass, L. A. & Parnas, J. (2003). Schizophrenia, consciousness, and the
self. Schizophrenia Bulletin, 29(3), 427–444.
Sass, L. A., Pienkos, E., Nelson, B. & Medford, N. (2013). Anomalous
self-experience in depersonalization and schizophrenia: A compar-
ative investigation. Consciousness and cognition, 22(2), 430–441.
Seigel, J. (2005). The idea of the self: Thought and experience in Western
Europe since the seventeenth century: Cambridge University Press.
Seth, A. K., Suzuki, K. & Critchley, H. D. (2011). An Interoceptive
Predictive Coding Model of Conscious Presence. Frontiers in Psy-
chology, 2.
276 Robert Clowes
Sierra, M., Baker, D., Medford, N. & David, A. S. (2005). Unpacking

the depersonalization syndrome: an exploratory factor analysis on
the Cambridge Depersonalization Scale. Psychological Medicine,
35(10), 1523–1532.
Sierra, M. & David, A. S. (2011). Depersonalization: a selective im-
pairment of self-awareness. Consciousness and cognition, 20(1),
99–108.
Strawson, G. (1999). The Self. In S. Gallagher & J. Shear (Eds.), Mod-
els of The Self ( 1–24). Exeter: Imprint Academic.
Taft, M. (2012). An Interview with Thomas Metzinger: What Is the Self?
Retrieved 04 October 2013, from <http://www.beinghuman.org/ar-
ticle/interview-thomas-metzinger-what-self>.
Zahavi, D. (2003). Phenomenology of self. The Self in Neuroscience
and Psychiatry, 56–75.
Zahavi, D. (2005a). Being someone. Psyche, 11(5), 1–20.
Zahavi, D. (2005b). Subjectivity and Selfhood: Investigating the
First-Person Perspective. Cambridge, MA: MIT Press.
Zawidzki, T. W. (2008). The function of folk psychology: mind reading
or mind shaping? Philosophical Explorations, 11(3), 193–210.
Conceptual Personae of the “attentional self ”
Alexander Gerner1
1. Introduction
“Everything has to be rethought from the beginning” Mr. Palomar (Calvino 1985, 115)
In this paper I take attention as a constitutive ground of the self. I crit-

ically survey the claim of Metzinger that a subjective self, defined as
the centre of awareness, is the possibility of being able to manipulate
the focus of attention, thereby stabilizing subjective experience. Thus I
propose an attentional self in which I will put Metzinger’s thesis of the
“control of the focus of attention” and the resulting notion of the “atten-
tional self ” critically into perspective by approximating the concept of
the self by means of conceptual personae of the “impossible” attention-
al self in Paul Valery’s dyadic conceptual personae “Monsieur Teste”/
“Émilie Teste” and the “heautoscopic” attentional self in Italo Calvino’s
“Mister Palomar”.
Philosophical Concepts, according to Gilles Deleuze & Felix
Guattari in “What is philosophy?”(WP) can be personae – as ex-
plicitly as the appearing of “Socrates” for Plato or “Dionysus” and
“Zarathustra” for Nietzsche- that act from a constitutive plane of phi-
losophy and create concepts on this plane2 at the same time. What
1 Affiliation: University of Lisbon, Center for Philosophy of Science (CF-

CUL), FCT Post-Doc SFRH/BPD/90360/2012, <amgerner@fc.ul.pt>, <http://
cognitiveenhancement.weebly.com/>. Member of the Research Project:
“Cognitive Foundation of Self ” IFL FCSH-UNL Project Fund: PTDC/FIL-
FCI/110978/2009.
2 “There are innumerable planes, each with a variable curve, and they group togeth-
er or separate themselves according to the points of view constituted by personae.
Each persona has several features that may give rise to other personae, on the same
or a different plane: conceptual personae proliferate.” (WP, p. 76).
278 Alexander Gerner
Deleuze & Guattari call “A conceptual personae […] thinks in us”3

(WP, p. 69) is thus more complex than a specific representational
model of thinking or of a represented narrative self in symbolic rep-
resentation. Therefore we have to underline the importance of con-
ceptual personae for the creation of different points of view by these
personae that direct thoughts concerning concepts such as “self ” and
“attention” by giving these concepts not just certain names but also
providing concrete concepts to consider, putting these conceptual
personae into theoretical and practicable perspectives and into coher-
ent conceptual constellations with the phenomenon of the self. Going
towards the “self ” and back over conceptual personae, the “self ” be-
comes a diagrammatic feature on superimposed conceptual maps4 of
concrete personae of the “self ”.
Thus to state the first point:
1) Complex concepts such as the “self” have personae with
which we think. With these personae of thinking we create plural
conceptual maps with which we actively explore the complexity of
the phenomenon.
The personae of the self may be extended into a conceptual
space, but are always present as concepts that show their perspectives,
connecting lines, strengths and weaknesses as well as their lines of
flights by being superimposed on to and by others. The concept of the
3 Hereby the expression “personae” is used in the sense of different instances of one
being and “personas” as different beings.
4 “Maps […] are superimposed in such a way that each map finds itself modified
in the following map, rather than finding its origin in the preceding one: from
one map to the next, it is not a matter of searching for an origin, but of evaluating
displacements. Every map is a redistribution of impasses and breakthroughs, of
thresholds and enclosures, which necessarily go from bottom to top. […] Maps
should not be understood only in extension, in relation to a space constituted by
trajectories. There are also maps of intensity, of density, that are concerned with
what fills space, what subtends the trajectory. […] A list or constellation of affects,
an intensive map, is a becoming. […] We see clearly why real and imaginary were
led to exceed themselves, or even to interchange with each other: a becoming is
not imaginary, any more than a voyage is real. It is becoming that turns the most
negligible of all trajectories, or even the fixed immobility, into a voyage; and it
is the trajectory that turns the imaginary into a becoming. Each of the two types
of maps, those of trajectories and those of affects, refer to each other” (Deleuze,
1998, p. 63.)
Conceptual Personae of the “attentional self ” 279
“self ” is seen here as a relational entity and as such can be oriented

within a diagrammatic map5 of concepts.
The self might be superimposed on to the concept of the “other”,
or in distinction to the “Non-self ” (as in Buddhist philosophy and psy-
chology; see: Olendzki 2010, Engler/Fulton 2012) or even be integrat-
ed or opposed towards a representational “self-model” (Metzinger) or
a virtual, “disembodied” entity of a “mind”. Another way to think the
self is to think of it as basically embodied in between the body-im-
age and the body schema (Gallagher, 1995) or described by a double
aspect of an basic bodily embodied self and a personal, reflexive or
extended self6 (see: for example Fuchs 2012b). In the terms of Thom-
as Fuchs7 there is a fundamental second-person perspective in the de-
velopment of the self and thus in the development of the attentional
self (from simple8 to dyadic attention) as well, suggesting a central
5 On diagrammatic thinking and the map see: Gerner 2010a.

6 The extended, reflexive or personal self, according to Fuchs (2012b) is char-
acterized by the aspects of (a) capacity of taking on the perspective of the other
(Perspektivenübernahme), (b) the capacity of introspective or reflexive self-con-
sciousness, (c) a capacity of verbalization of its own experience, (d) the creation
of coherent narrative Identities and (f) a self-concept by obtaining conceptual and
autobiographic self- knowledge.
7 The primary experience of self -in Fuchs account- is not a pure self-experience,
but also includes a sense-motor relationship between the experiencing subject
and the environment, – often expressed in spontaneous agency and in action-per-
ception cycles – which is mediated by the body and its habitual capacity and
later on leads to social inter-corporeity. Available through the senses, members
and capacities the body is embedded in the surrounding space that presents it-
self to it as a field of possibilities and valences. By this structural pairing of
the lived body and its complementary surrounding, the basic self becomes an
embodied spatial “ecological self” (Neisser, 1988). This is expressed by the sit-
uatedness and attunement of the lived body as well as by the feeling of body
related to potentialities of a situation, and specifically by bodily sensations im-
plicit in world-directed attention. Thus a basic feature of the attentional self is
a) the body as zero-point of spatial orientation and attentional coordination,
b) sensory-motor cycles, the sensory-motor coupling with the environment and
a basic bodily being-towards-the world and c) the experience of flexible enacted
and transgressed boundaries of the lived body.
8 “From birth on, infants are attentive to external entities (call this ‘simple atten-
tion’) and are engaged in dyadic self-other interactions which involve dyadic
attention where subjects are mutually attending to each other. Later, when the
ecological conceptual map (I-me [1st PP] & you-me [2nd PP]) of the self
and its conceptual personae that then is able to socially share attention
and intentionality (Tomasello & Carpenter, 2007). This developmen-
tal account of an attentional self, based on early mutual attentional
engagements, are crucial for later shared triadic attentional relations
(Me-You-object), which then become important for developing com-
plex joint actions (Frith & Gallagher, 2012). These views will later on
be explored in slightly different way in the responsive account of the
dyadic self in between the conceptual personae of “Monsieur Teste”
and his wife Emilie.
Each superimposed conceptual map of the self has to show how
productive it acts and engages in the clarification of the phenomenon
of the self. I wish to propose to superimpose the concept of the “self ”
on to the concrete working of “attention” (see Gerner PhD thesis 2012
upcoming) and propose to show how conceptual personae of attention,
although being realistically “impossible” as in the case of a purely and
isolated episodic attention function or a merely disembodied heauto-
scopic attention, nevertheless can help to elucidate how to conceive the
“self ” by its relations to the phenomenon of attention.
The officially “non-philosophical” fictional cases of “Monsieur
Teste” (Paul Valéry) and “Mister Palomar” (Italo Calvino) can produce
insights into the complex philosophical concept of the “self ” – in our
case of the self as an “attentional self ” – before reviewing the phil-
osophical position in Thomas Metzinger’s account of the self as “at-
tentional self ” that in his account has the main function of stabilizing
subjective experience by controlling the attentional focus.
The possible/impossible and aesthetic/ heautoscopic notions of the
self relate to the constitution of modes of appearance and disappearance
of a bodily episodic self, that seems more fundamental than a narrative
or reflexive conceptual self. Thus an “attentional self ” seems impossi-
ble as a permanently and constantly alert “hunter” that is permanently
“on“ as in the ideal of a perfectly and permanently alert man. Attention
as I see it – pertains to a double movement in between “waking up”
infant begins to follow the gaze of the other person, it may occasion a new form of
attention (call this ‘shared attention’): the infant is aware of the adult being atten-
tive towards the object and of herself being attentive towards the object” Fiebrich
& Gallagher (2012, p. 7).
and “falling asleep” and therefore is exposed to modes of fatigue and

change, and bound with it seems the notion of the self in the sense of
an attentional self, that neither might be permanently “on”, although
it might as well not be entirely “episodic” as conveyed by Strawson
(2004) also make part of a diachronic temporal instances.
Alertness has been studied in empirical psychology and cognitive
(neuro-) science, being termed “vigilance9“ or “sustained attention”, to
hold on to something for a longer amount of time, as distinct from the
shorter rhythmic structure of spatial orientation in turning towards and
turning away in attention. This vigilance function of attention has to
be distinguished from the conceptual personae of the “hunter” as the
attentively alert self, that – as Ortega y Gasset (2007) underlines – is not
necessarily programmed for successful hunting, by catching the prey10,
9 Vigilance was first studied on the empirical level by the neurologist Henry Head
(1923) in patients with brain injuries in the 1920s. Norman Mackworth (1948,
1950/61) started the “systematic study of vigilance during World War II in human
factors”(Warm et al., 2008, 433). Later Parasuraman (Parasuraman, 1976; Para-
suraman & Davis, 1977; Parasuraman, 1986; Parasuraman et al., 1986) dedicated
his studies to the taxonomy and utility of vigilance (Parasuraman et al., 1987) and
sustained attention (Parasuraman, 1984a) in detection and discrimination (Para-
suraman, 1984b), in human factors and automation (Parasuraman & Reiley, 1997)
as well as monitoring and search (Parasuraman, 1987) and the topics of the brain
systems of vigilance (Parasuraman et al., 1989) of how the brain is put to work
in the sense of “neuroergonomics” (Parasuraman & Rizzo (eds.), 2007; Parasura-
man & Wilson, 2008). Vigilance may not be automatic and “effortless”, but may
involve stress and “hard mental work” (see: Warm et al., 2008).
10 It should seem almost “natural” to start any kind of exposition about human atten-
tion with the traditional figure of the attentive “hunter” that besides the monitoring of
other animals seems a classical conceptual persona of creature attention as different
from machine attention and thus contains consequences for the self in its develop-
ment. New et al. (2007) recently proposed an ancestral animate monitoring hypoth-
esis that is confirmed by imaging studies (see Böttger et. al., 2010) by comparison
with stimuli of moving animals in comparison to optokinetic stimuli. Attention is
captured more easily by ancestral animals than for instance by equivalent contem-
porary moving computer stimuli: “For ancestral hunter-gatherers immersed in a rich
biotic environment, non-human and human animals would have been the two most
consequential time-sensitive categories to monitor on an on-going basis. As family,
friends, potential mates, and adversaries, humans afforded social opportunities and
dangers. Information about non-human animals was also of critical importance to our
foraging ancestors. Non-human animals were predators on humans; food when they
strayed close enough to be worth pursuing; dangers when surprised or threatened by
but might give us hints to understand how attention and the self could
have co-evolved and differentiated over time.
2) Moreover, we can state that conceptual personae of the self give
us the possibility of accessing something subjectively phenomenal
even if the phenomenon in question shows itself to be difficult to be
measured empirically and thus has to be combined and superimposed
virtue of their venom, horns, claws, mass, strength, or propensity to charge; or sources
of information about other animals or plants that were hidden or occluded; etc. Not
only were animals (human and non-human) vital features of the visual environment,
but they changed their status far more frequently than plants, artefacts, or features of
the terrain. Animals can change their minds, behaviour, trajectory, or location in a
fraction of a second, making their frequent reinspection as indispensable as their initial
detection. For these reasons, we hypothesized that the human attention system evolved
to reliably develop certain category-specific selection criteria, including a set designed
to differentially monitor animals and humans. These should cause stronger sponta-
neous recruitment of attention to humans and to non- human animals than to objects
drawn from less time-sensitive or vital categories (e.g., plants, mountains, artefacts).
We call this the animate monitoring hypothesis.” New (2007, 16598).
In animals the application of attention is crucial for survival. “Although atten-
tion is an efficient filtering mechanism, limited attention may be a major cause of
mortality in nature.” (Dukas & Kamil, 2000, 502). As hunting or predation as well as
predator avoidance (for example through the fear of snakes that catches attention in
mammals) are principles of the animal kingdom, attention can be seen as a strategic
biological mechanism that has evolved also in relation to survival in between preda-
tion and prey. On predation see Langerhans, R. (2006) ”Evolutionary consequences
of predation: avoidance, escape, reproduction, and diversification.” In A.M.T Elewa
(ed.). Predation in organisms: a distinct phenomenon. Heidelberg: Springer, 177–220.
See Isabell (2006, 2009) on the category of the sensitive evolution of the mammal
brain (including vigilance or strategic attention) towards those animals that have pro-
vided a recurrent survival threat from an evolutionary perspective (such as snakes)
and those which have not: “Snakes have a long, shared evolutionary existence with
crown-group placental mammals and were likely to have been their first predators.
Mammals are conservative in the structures of the brain that are involved in vigilance,
fear, and learning and memory associated with fearful stimuli, e.g., predators. Some
of these areas have expanded in primates and are more strongly connected to visual
systems. However, primates vary in the extent of brain expansion. This variation is
coincident with variation in evolutionary co-existence with the more recently evolved
venomous snakes.”(Isbell, 2006, p. 1). On the fear of snakes as attention “commander”
in an evolutionary perspective see Soares (2010). In “Meditations on Hunting” Ortega
y Gasset meditated, however, not from a biological evolutionary standpoint, but from
a philosophical point of view on “hunting” as a fundamental human attentional diver-
sion phenomenon, that we should have in mind as well besides its biological function.
on other concepts or only appears in the moment of its failure, as is

documented by the large literature on pathological situations of the self
as in schizophrenia or the missing body image (see: Gallagher, 1995),
and is parallel to the case of “attention” that can be called a philosoph-
ical embarrassment (Waldenfels, 2004) that accompanies the literature
of attention and its many non-unified theories and incoherent theoreti-
cal metaphors (see: Gerner 2012; Fernandez- Duques & Johnson 1999,
2002) through which it is described.
The idea of proposing a necessary relation between attention and
the self by means of an “attentional self ” shares difficulties in the de-
scription of “attention” and “self ” as independent concepts. And one
might say that through the introduction of attention into the concept of
the self increases the complexity in the understanding of what the self
could be. Nevertheless, it is our aim to show that looking tangentially at
the self from the perspective from attention we obtain concrete insights
that can be made fruitful for the concept of the “Self ”.
1.1 Conceptual personae of the impossible attentional self:

Monsieur Teste as the impossible “man” of attention
Let us assume here for a few short paragraphs that the actual philo-
sophical debate of the self would be fundamentally enriched if we took
the philosophical personae of writers into consideration – as we will do
in the first place with the conceptual personae Monsieur Teste of Paul
Valéry as an attentional self in its conceptual constellations.
The “impossible man of attention” Teste is considered by Valéry as
being “no one” (Valéry, 1989, p. 131), and as such has “no personality
that is unique and his own, no inseparable attributes.”(Ibid.) He is also
characterized as well as the “Tête (Teste, Head)” (Valéry, 1989, 90) or
“The game played with oneself ” (Valéry, 1989, p. 106) and its living
self observation and will still always be “some stranger to me” (Valéry,
1989, p. 102), thus is fundamentally related to the “other” in its proper
doubled definition as other-self. The Self is thus seen as a strange habit
of interchange of two different perspectives in a kind of “self-game” be-
tween “me” (Teste himself) and “he” (Teste narrated or the narrative self
and put into perspective by the narrator/author of Teste) – a higher-order
“narrative self ”: “Monsieur Teste had taken the strange habit of think-
ing of himself as a chessman in his own game. He could see himself.
He could push himself across the table. At times he lost interest in the
game. /The systematic use of Me and He.” (Valéry, 1989, p. 100). What,
however, happens if the interest in the distinction of he/me is dropped
and the self-game stops, seems a fundamental question that we should
answer with the concept of a “self ”. “He” is also characterized as a
plural episodic multitude of Selves “Teste. / The home of Selves/ The
islands of Selves” (Valéry, 1989, p. 148) or man’s potential of a “pure”
or virtual attentional self “beside his body” (ibid.) and as such will be
helpful to understand better what – as we will see later- Metzinger in
“Being no one” understands as an “attentional self ” as a precondition of
stabilized subjective experience.
But let us stick to Monsieur Teste, who Valéry paradoxically de-
scribes as between The “Everything” in the sense of all particular sen-
sations (bodily needs, pain or caution) before any reflection and before
a “pure reflexive self ” in which “Self ” and the sensational “Everything”
would be merely polar properties of an attitude or disposition that can
be turned into an automatic routine (body schema or body image). In
the end for Valéry the maximum consciousness of the self is turned into
a routine behavior and as such in Valery’s perspective is seen as animal-
ized by the biological fact of being sensitive to typical biomechanical
processes of animal life.
Firstly, we will look at the case of the Paul Valery’s conceptual
personae in “Monsieur Teste” in which not only Monsieur Teste shows
itself as an a) impossible pure attentional self, but as well as part of an
extended concept of the self by b) “Monsieur Teste’s” wife Émilie Teste
and who is accompanied by a c) partial observer of Monsieur Teste, a
friend of Monsieur Teste who observes him and describes him in an at-
tentional scene at the “theatre” of consciousness as well as by an d) au-
thor-narrator-self of Paul Valéry that gains self-knowledge by reflecting
on his own conceptual personae Monsieur Teste that he properly created
in odd moments of his life11.
11 “Teste was conceived – in a room where August Compte spent his early years – at
a period when I was drunk on my own will and subject to strange excess of con-
sciousness of my self.” (Valéry, 1989, p. 3).
Teste is defined from the outset as a concept of the impossible12,

a “man of attention” that shows different notions about the conceptual
personae of the attentional self related by its author:
[A] Monsieur Teste – alias the “attentional self ” – is regulated by an
equilibrium of possibility and impossibility
[B] Monsieur Teste – alias the “attentional self ” – is a mechanism that
sets up and regulates the relation between the known and the un-
known
[C] Monsieur Teste – alias the “attentional self ” – deals with the expe-
rience of the finite.
All three notions of Monsieur Teste as a thinking image of the atten-

tional self are concepts of a multitude of attentional selves that are in-
terrelated and connected and none of them alone can provide a realistic
account of a unified “attentional self ” or the one and only persona of
Monsieur Teste, which also means each conceptual personae being tak-
en on its own and as a materialization of the “Self ” is impossible, as
they would be neglecting necessary layers of the self (from the minimal
to the narrative self). Even so, considering the minimal self we have
to listen to what Bernhard Waldenfels states about the complex work-
ing of attention. According to Waldenfels (2004) there is not one but a
double-event of attention, and thus – I would propose – also a double
concept of – for instance – the minimal “self ”: as attention is not just
“Auffallen” the “what happens” side that could be conceptually mapped
by physiological processes alone, but is as well an “Aufmerken” – that
is an event that shows “what happens to me”, and thus implies the per-
spective of an attentional self to which through my body/ a “me” is
12 “Why is Monsieur Teste impossible? That question is the soul of him. It changes
you into Monsieur Teste. For he is no one other than the demon of possibility.
His concern for the whole range of what he can do rules him. He observes him-
self, he manoeuvres, he will not allow himself to be manoeuvred. He knows only
two values, two categories- those of consciousness reduced to its own act: the
possible and the impossible. […]. Its brief and intense life is spent in watching
over the mechanism, which sets up and regulates the relation between the known
and the unknown. It even applies its obscure and transcendent powers obstinately
to simulating the properties of an isolated system in which the infinite plays no
part.”(Valéry, 1989, pp. 6–7).
affected and may respond before any conceptual epistemic noting of

x, and any kind of narrative or narrative self is implied. According to
Waldenfels, attention should first be approached through a double-view
or a double-grasp of the two single movements (Auffallen- Aufmerken),
in addition to the third part, the taking notice side of epistemic atten-
tion, which is when I attend to x and note or recognize that it is x (for
example, a book on my desk).
Thus, the “self ” for Valery’s thinking personae Monsieur Teste
deals correspondingly to our view on attention with the incoherence
of the mind that for example a new stimulus (It happens) creates
that happens first as an event that is followed by the event of the self
(something happens to me) – that what happens, is actually happening
to me, before I note what it is that happens and what this me as such
involves. The “self ” thus first appears as a mental pre-conceptual
response to the instability of what happens and what can thus pos-
sibly happen to me. In this line of thought Paul Valéry has already
stated:
– I am the unstable.
– The mind is maximum possibility- and maximum capacity for incoherence
– The SELF is the immediate response to each partial incoherence which is a
stimulus (Valéry, 1989, p. 73).
Thereby we can note three of Monsieur Teste’s notions of the attentional

self:
[1] The self is part of responsive attention, an direct or “immediate re-

sponse” to a situation involving the incoherence event
[2] The self is stimulus driven, and as such we can easily call it an em-
pirical attentional bottom-up self
[3] The self is both an actuality attentional mode, or a capacity/bodily
mechanism and a possibility attentional mode, or a mere possibility
of the mind.
The self is seen here as a consequential response to the maximum of

incoherence of the mind. The incoherence of the mind makes it alive,
as does the coming to the point in its most exactness. This also means
that Monsieur Teste would be not “mortal” but dead, if he would be
fixed as a “pure absolute” or “pure transcendental” self. This impossible
purely transcendental notion of the concept of the “self ” leads us to

several key questions of Monsieur Teste as the dynamic centre of
awareness:
[…] how does one choose a character to be oneself? How does that centre take
shape? Why, in the theatre of the mind are you You? You and not me? But is that
not precisely Monsieur Teste’s research: to withdraw from the self, the ordinary
self, by trying constantly to diminish, combat, compensate the irregularity, the
anisotropy of consciousness. (Valéry, 1989, p. 109)
Hence Valéry that turns the self into a paradoxical concept makes an
essential double notion of the self clear:
a) The ordinary functional self-mechanism creating coherence in a
sea of chaotic episodic impressions and stimuli is located in a perma-
nent struggle with an implicit b) “extraordinary” self that in extremis
properly not only compensates irregularity but even creates the incoher-
ence of the mind by taking each episodic extraordinary attentional event
of the self at face value.
Let’s stick just for a moment to the second extraordinary notion
of the self that breaks with diachronic self-habits by considering the
relation of the conceptual personae of Monsieur Teste together with the
conceptual personae of his “wife” Émily Teste:
Émilie Teste achieves the possibility of self-formulation while reflect-
ing on her husband, self- designating herself as being more than a mere
witness, but already an “organ” (Valéry, 1989, p. 24)- though external-
of Monsieur Teste. By reducing herself to an organic and organizational
inter-dependence with Teste, however, being a “non- essential” (ibid.) or
external organ in his life-world, she describes the essence of her character
as longing “to be surprised” (ibid.) by Monsieur Teste’s episodic self and
the relation she has to Teste as being strongly bound by “the uncertainty
of his moods”(ibid.). Their conceptual relation is double a) of “husband”
and “wife” and their distinct activities and necessities are said to be in
tune b) in the sense of interdependence of lived organization. Émilie Teste
and her “husband” are two conceptual personae of one attentional self and
as such are interrelated13. The practical organ of sensible stimulus-driven
13 “But I am somewhat more than the witness of his life. I am a part, almost an
organ of it, though non-essential. Husband and wife as we are, our actions are
harmonized in marriage and our temporal necessities are well enough adjusted,
responsivity to the world and the ideal episodic or “epochal” attentional

self of Monsieur Teste are two inseparable aspects of the attentional self.
The way of investigating in the impossible autochthonous pure attention-
al self of Monsieur Teste finds its necessary responsive structure in the
pragmatic function of a stabilizing attentional self – that here is called
Madame Émilie Teste- and that in Thomas Metzinger could be described
as the function of focusing the centre of attention in order to stabilize
subjectivity. This, however coming from the double constellation of Paul
Valéry’s Monsieur Teste/Émilie Teste is however just a simplistic picture
of the attentional self that in its foundation is dyadic and contains a dou-
ble movement of focus and distraction.
Thus coming back to the key question of how the centre of the self
gains form, and how Monsieur Teste becomes himself, we have to look
at the description of Émilie Teste.
In the metaphoric “conceptual marriage” between Edmond and
Émilie Teste, her preoccupation about his episodic self is defined by the
fear of his disappearance through absorption, or a purely “episodic” self
that would be equaled with change itself:
…A bit more of such absorption, and I am sure that he would be invisible!…

(Valéry, 1989, p. 26).
Émilie shows herself as “almost” a bodily organ regarding the possi-

bility of an episodic attentional self of Monsieur Teste in her function
of self-preservation, for instance in regulating the “social life”14 of the
attentional self of Monsieur Teste. This is exemplified by her actions
of responding to the environment as in the case of her responding to a
despite the immense and indefinable difference of our minds. So I am obliged to

tell you incidentally about her, who is now telling you about him” (Valéry, 1989,
p. 24).
14 In relation of a merely “spiritual” part of the attentional self alias Monsieur Tes-
te and its social organ Émilie Teste one should consider the following quote by
Valéry in which the vital function of self-preservation is attributed to the social
“organ” of the ego while the pure spiritual Teste kills in its epochal way: “Ego./
There are things I could “clarify” and don’t wish to…/ I keep them in condition.
The social keeps the spiritual in condition. /Social life does not want Monsieur
Teste to discharge his duties. We must admit that the spirit kills and the letter gives
life… or at least, preserves it.” (Valéry, 1989, p. 125).
letter to a friend of Monsieur Teste, that is, questioning and observing

the “life” of Monsieur Teste from a distance.
By being a conceptual personae or a figure of thought of sensibility
that understands little but responds to the necessity of the attentional
self to work, she would sometimes act instead of or as part of Monsieur
Teste, the “man of attention”. The relation of Monsieur Teste and Émi-
lie Teste shows itself through in-between events15 (Waldenfels 2002) of
attention because “in the phenomenon of attention the mundane con-
sciousness fights with transcendental consciousness.” (Blumenberg,
2002, p. 200), and in this perspective one has to see what could be
formulated as a complex attentional self between constitutional stability
and instability at the same time: The self is a marriage! Valéry gives us
the most rich and thought-provoking conceptual personae of an atten-
tional self: Monsieur Teste and his wife Émilie with whom we are able
to explore the working of related but distinct complementary functions
of the self that are in constant (dis-) equilibrium.
Equipped with the pragmatic function of her “reading” Monsieur
Teste, Émilie Teste describes herself as acting complimentary to the
extraordinary, transcendental attention based on interior observation of
Monsieur Teste:
Kind sir (and friend) […] So I read your letter to Monsieur Teste. He listened
without showing what he thought of it, nor even that he was thinking about it. You
know that he reads almost nothing with his eyes, but uses them in a strange and
somehow inner way. I am mistaken – I mean a particular way. But this is not it at
all. I don’t know how to put it; let’s say inner, particular…, and universal!!! His
eyes are beautiful; I admire them for being somewhat larger than all that is visible.
One never knows if anything at all escapes them, or, on the other hand, whether
the world itself is not simply a detail in all that they see, a floating speck that can
15 “Zwischenereignisse” or “In-between-events” for Waldenfels are events in which

something appears [auftritt], while they connect to something else, without hav-
ing being connected or related to it before. (Waldenfels, 2002, p. 174; see also
Waldenfels, 1987, pp. 46–48). According to these in-between-events, Waldenfels
inscribes in-between-events of attention in his philosophy of the “strange” or the
“other”. In-between-events of attention happen either a) between me and the oth-
ers or b) between a group (us) and another group or c) in general between the
“proper” [dem Eigenem] and the strange [dem Fremdem] that unfolds in the phe-
nomenon of attention consisting of a double event in auffallen-aufmerken and is
irreducible to one of its double events.
besiege you but does not exist. Sir, in all the time I was married to your friend I
was never sure of what he sees. (Valéry, 1989, p. 22, translation altered, A.G.)
Monsieur Testes attention is not only particular, but also universal, and
above all an inner principle. The eyes of Teste are not just organs to
perceptually pick up visual information. They are more actors of trans-
formations in which a world densifies or vanishes, in which by attentive
observation a detail of the world can become a whole world or in which
the world expands or shrinks in an instant: zim-zum. What stays are the
riddles of not understanding exactly what it is that the eyes of Monsieur
Teste actually see: What is this inner attentive self?
As Valéry says, the “proper soul” of Monsieur Teste is his impossibil-
ity of existing as a “real person”. On the other hand M. Teste is thinkable
as a “proper demon of possibility”, a “personnage de fantaisie” (Valéry,
1946, p. 7) that extends in an episodic self that is measurable in excep-
tionally few “quarters of an hour”16. Paul Valery called “Monsieur Teste”
in his introductory preface the “impossible man of attention”.
Despite the rhetoric of failure and the impossibility of the model
character Monsieur Teste postulated by Valéry as the constitutive im-
possibility of M. Teste as a real person, paradoxically he is taken as
someone that could transform the one thinking Monsieur Teste into
the proper Monsieur Teste – “[Elles change vous en Monsieur Teste]
It changes you into Monsieur Teste” (Valéry, 1989, p. 6 [1945, p. 11]).
This change into Monsieur Teste from the one thinking Monsieur Teste
is central. At the same time it is even more central to note that there
is another change that is excluded in the moment in which the reader
or thinker of Monsieur Teste changes herself into Monsieur Teste. By
becoming a “man of attention”: the change into change itself is exclud-
ed. This means per definition of an attentional self there can never be
an absolute identity between the episodic change and the one experi-
encing the change. Thus this difference constitutes the self-variance or
16 “Coming back to Monsieur Teste, and observing that a character of his kind
could not survive in reality more than a few quarters of an hour, I say that the
problem of that existence and its duration is enough to give it a sort of life. The
problem itself is a seed. A seed is a living thing; but there are some that could
never develop. These make an effort to live, become monsters, and monsters
die. In fact we know them only by this remarkable property of being unable to
endure. (Valéry, 1989, p. 5).
instability of the attentive self. This also shows the openness of the self
as unequal of the changing episode: The attentional self acts as the dif-
ference of “what happens” and “what happens to me”.
Monsieur Teste, as described by his author Paul Valéry, shares with
his author one basic intention regarding an episodic self: “to extend the
duration of certain thoughts” (Valéry, 1989, p. 3)17.
Parallel to this temporal limitation we find the neurobiological re-
search of the Munich Neuroscientist Ernst Pöppel and the temporal em-
pirical window of three seconds18 of attention. This limited capacity in
time or load of an episodic attentional self that Valéry in his “Extracts
from Monsieur Teste’s logbook” calls the centre of elasticity, “Centre
de ressort” (Valéry, 1946, p. 65) can ideally prolong the attentive ob-
servation until a temporal maximum of a quarter of an hour. Recent-
ly research on the praxis of meditation has shown that this temporal
plasticity differs and increases when carrying out a continuous praxis
of attentive meditation19. The “centrality” into which these different ep-
isodic attentional periods come together and gain coherence while not
dissolving into change and the episodic itself is exactly one of the most
difficult philosophical problems of the self, from perceptual binding to
the coherent identity of selves over time:
What he calls: The Central Problem. To be a center. This problem is one of agree-
ments – coincidences, stabilities, time intervals of various kinds in combination
17 “Monsieur Teste? / If reflection on one point could be prolonged, it is likely that

the content of thought would allow its substance to appear, and that whatever it
might represent, one would see the canvas, the panel, the texture- the true element
of time. There would be an expansion of time during which a sort of pure sensation
would arise in its place, as the substance of the meaningful illusion. But such an
extension is impossible. Attention ends in the perception of attention itself, which
is fatigue […]” (Valéry, 1989, p. 121).
18 Pöppel, E. (1985). cit. In Assmann, A. (2001, p. 17).
19 See for example: Lazar, S. et al. (2005). Meditation experience is associated
with increased cortical thickness. Neuroreport 28; 16(17), 1893–7; Lutz, A.
et al. (2008). Attention Regulation and monitoring in meditation. Trends Cogn
Sci 2008, 12(4), 163–9; Manna, A. et al. (2010). Neural correlates of focussed
attention and cognitive monitoring in meditation. Brain Res Bull. 82 (1,2),
46–58; MacLean, K. et al. (2010). Intensive Meditation Training Improves Per-
ceptual Discrimination and Sustained Attention. Psychological Science, 21(6),
829–838.
or contrasts […] The Contrasts between specific times, various functions- that
is to say, what keeps us from merging with change itself, and makes us aware
of divergence, […] and gives us the Capacity of the moment […] its instability
(self-variance). (Valéry, 1989, p. 120)
For Bernhard Waldenfels, Paul Valéry marks a hybrid between writing

and thinking in a singular way that is preoccupied with attention in the
Cahiers- but also in Monsieur Teste- that opens up a “new form of po-
etics” (Waldenfels, 2004, p. 28) as well as formulating a concept of the
self by following the self’s traces of strangeness [Fremdheit], absences
[Absenzen], deviations [Abweichungen] and “delays” [Retardandis],
and as such considers this worrisome phenomena of an attentional self,
that could here be called unbalanced attentional dives.
As the diver never can maintain himself completely in his or her
“element”, and as such needs to come back to the surface to take a
breath, the man of attentional self – Monsieur Teste – feels like he is a
fish out of water, that needs to prolong his being in a longer duration
and thus stretch the episodic attentional self and the episodically ap-
pearing objects of attention. For Valéry the condition of something as
an object of attention, of a thing, is only able to be understood when
one is shown a rule of a medium value in the moment of its observa-
tion: “The things are only what they are as average duration and depth
of observation.” (Valéry, Cahiers II, p. 262, cit. in Waldenfels, 2004,
p. 34). Things are gazes directed towards them for a certain amount
of time in comparison and relation to other possible things. The func-
tioning of attention, according to Valéry, is explained in relation to a
regular value that only shows up at the moment of a dislocation and thus
the imbalance in relation to this medium value is constituted through a
process of finding a re-equilibrium, or we could say: attention is what
deviates from what should be expected as a medium value and thus the
attentional self is the gesture between the balance and imbalance of this
medium value of experience. This medium value seems already implicit
in expectation regarding things and their givenness and by a possibly
notable shift away from a pre-given expectation towards the actually
attended thing in observation. To be attentive means – in this sense of
Valéry – “to find or go looking to find a value of X that differs from its
medium value X_” (Valéry, ibid.) and that both values show themselves
only in their dyadic relation.
That is how the attentional self shows itself as an imbalance of the

mind that calls for (re-) equilibrium. The moment at which attention
happens – Valéry notes – can be minimal: “a tooth aches – a disquiet-
ness [Sorge] – a necessity” etc. are all seeds, luminous signs- of atten-
tion” (Waldenfels, 2004, p. 34). This is why the conceptual personae of
a pure “unstable” episodic self that is devoid of any bodily multisensory
experience – such as Monsieur Teste – is still a conceptual personae of
an attentional self.
2. Heautoscopic “attentional self ”
2.1 Italo Calvino’s Mister Palomar’s development from an aesthetic to

an heautoscopic “attentional self”
Calvino’s text “Mr. Palomar” (1985 [1983]) shows three steps of a phe-
nomenology of attention – firstly, phenomenological description, sec-
ondly, narration, and thirdly, meditation. Moreover, Mr. Palomar deals
with the self in three types of experience20 as Calvino explicitly remarks
at the end of his book: a) the self of Mister Palomar in visual experience
b) the self in experience involving language/meaning and symbols and
c) the self in speculative experience (cosmos, time, infinity, the relation
of self and world and the dimensions of mind).
The name of the conceptual personae of the self, Mr. Palomar,
is derived from a famous astronomical observatory21 (Calvino, 1985,
p. 37), which is located in the north of San Diego County operated
20 Those marked “1” generally correspond to a visual experience, whose object is

almost always some natural form; the text tends to the descriptive/ Those marked
“2” contain elements that are anthropological, or cultural in the broad sense; and
the experience involves, besides visual data, also language, meaning, symbols.
The texts tend to the form of the story./Those marked “3” involve more speculative
experience, concerning the cosmos, time, infinity, the relationship between the
self and the world, the dimensions of the mind. From description and narrative we
move into meditation” (Calvino, 1985, p. 128).
21 See <http://www.astro.caltech.edu/palomar/>.
by the California Institute of Technology, Palomar. Palomar in Calvino

is the second conceptual personae of the attentional self we want to
explore and is seen here as an exemplary conceptual personae of the
attentive gaze of the self in observation of the self and the world.
At the end of Calvino’s book, Mr. Palomar comes to the phenom-
enological conclusion of his own being, and that his task is “to look at
things from the outside” (Calvino, 1985, p. 113) while all the observ-
ables (from the wave to the stars, animals or daily objects) “ask” him
to prolong his attention towards them. His attentive gaze is caught by
a multitude of details ready for his observation of his attentive gaze
by “running all over the details” and is “then unable to detach itself ”22
(Calvino, 1985, p. 113) from these observable details. That is why in
the end Palomar sees just one path that he could go and that is to extend
his attention:
Mr. Palomar has decided that from now on he will redouble his attention: first, by
not allowing these summons to escape him as they arrive from things; second, by
attributing to the observer’s operation the importance it deserves. (Calvino, 1985,
p. 113)
But the following questions remain beyond the magnification of atten-

tion and thus the extension of the self: can Palomar actually observe
everything that he wants or ought to observe in a universal and ob-
jective manner, and how long can he keep the empirically perceptible,
observable details or run over all of them in an attentive gaze?
Is it actually possible to run over all the details if they are for ex-
ample all the stars in the universe, or all the crumbs of bread on a table,
or all the atoms and neuronal relations of single neurons in a single
brain? How is the complexity of the world and reality manageable with
Palomar’s attitude?
22 “After a series of intellectual misadventures not worth recalling, Mr. Palomar has
decided that his chief activity will be looking at things from the outside. A bit
near-sighted, absent-minded, introverted, he does not belong temperamentally to
that human type generally called an observer. And yet it has always happened that
certain things- a stone wall, a seashell, a leaf, a teapot- present themselves to him
as if asking him for minute and prolonged attention: he starts observing them
almost unawares, and his gaze begins to run over all the details and then is unable
to detach itself ” (Calvino, 1985, p. 113).
The perfect universal observer shows himself in Mr. Palomar as

being impossible and Palomar becomes more and more the expression
of this failure of universal observation, as his self is located in between
the world and its pure observation “as it is”:
At this point he faces a critical moment; sure that from now on the world will
reveal to him the wealth of things, Mr. Palomar tries staring at everything that
comes within eyeshot; he feels no pleasure, and he stops. A second phase follows,
in which he is convinced that only some things are to be looked at, others not, and
he must go and seek the right ones. To do this, he has to face each time problems
of selection, exclusion, hierarchies of preference; […] (Calvino, 1985, 113–114).
After the failure of looking at all the things with the same importance or
giving all things the same relevance (as he already tried while observing
the naked bosoms of a woman at the beach along with everything else)
his pleasure of observation dissolves and he finally stops. His difficul-
ties are due to the involvement of the self in observation. However, in
the second attempt Palomar gets involved in one of the basic problems
extremely relevant to the theme of attention and thus the attentional
self: the problem of selection, exclusion, creation of hierarchies of at-
tention and preference – why Palomar is attentive to this and not that, as
not everything can be attended to at the same time and there is a basic
non-simultaneity in attention. All that we are attentive to has already
happened in a double way: it has happened and it has happened to me
and while I start noting my experience of what has happened it is al-
ready “too late” for the attentional double event.
This leads for Palomar into the next question of how we can exclude
the self, here called the “ego”, of the observer from the observed and
impossibly, but desirably, let the “world look[s] at the world” (Calvino,
1985, p. 113) itself. Mr. Palomar therefore questions the “who” in ob-
servation, or as he says: “Who’s eyes are doing the looking?”:
[…] he soon realizes that he is spoiling everything, as always when he involves

his ego and all the problems he has with his own ego./ But how can you look at
something and set your own ego aside? Whose eyes are doing the looking? As a
rule, you think of the ego as one who is peering out of your eyes as if leaning on
a window sill, looking at the world stretching out before him in all its immensity
(Calvino, 1985, p. 114).
Calvino’s Mr. Palomar does not just want to suspend the participation
of the human “ego” in observation, but also wants to shift radically the
perspective of attention constituting an observer or a self as if “leaning
on a window sill” observing the world, in which the world observes the
world without the blur or noise that the egoic bodily centered observer
always seems to introduce.
Palomar’s intent seems the opposite of a classical Husserlian phe-
nomenological approach that by its program wants to bracket all state-
ments on the existence or non-existence of the world, in order to reach
phenomenological consciousness of things by epoché- the suspension
of the world (bracketing) by eidetic variation. Palomar uses the image
of the “window” through which the “ego” observes the world and that
changes position as a thinking image of the attentional self (that has
another image for instance in the “frame” ([Der Rahmen23] Pöppel,
2010[2006]) or the tunnel in the “Ego-Tunnel” (Metzinger, 2009) met-
aphor.
However, Mr. Palomar is actually seen as the conceptual personae
of an “attentive observing Ego” that observes and besides being linked
to a certain egoic perspective, tries to change its perspective from look-
ing from “inside out” to observing from “outside out” -as happens for
example in Out-of Body experiences, as we will see later. In terms of
Italo Calvino we can state a shift of perspective. From the perspective of
subject-object relation in “the ego looks at the world” he shifts the gaze
from one object to another. This means we get a radical new perspective
in “The world looks at the world” (Calvino ibid.), in which a non- ego-
ic self represents the looking world and the objects are the looked-at-
world: The gaze becomes autonomous and seemingly independent of
the ego.
Palomar doubts that this radical shift brings about an expected
“general transfiguration” (Calvino, 1985, p. 114) of the observed ob-
jects of the world, as he “casts his eyes around” and states that no such
transformation of worldly objects occur:
23 The subtitle of the book of the German neuroscientist Ernst Pöppel on the brain
and the self shows a similar inverted shift in which the observational ego or self is
now looked at from the perspective of the brain as the title shows: “The Frame. A
view of the brain onto our ‘I’” [Der Rahmen. Ein Blick des Gehirns auf unser Ich].
No such thing. The usual quotidian grayness surrounds him. Everything has to be
thought of from the beginning. Having the outside look outside is not enough: The
trajectory must start from the looked-at-thing, linking it with the thing that looks
(Calvino, 1985, p. 115).
This higher attention shift, not just from one observed object A to an-
other object B, but an radical change of perspective in observation –
in which the proper frame or window of the self is pushed outside
the proper sensual limitations of the body – seems to fail and always
bounces back on the ego of Palomar, his eyes and glasses that the world
“needs” to be perceived:
But how can you look at something and set your own ego aside? Whose eyes are
doing the looking? As a rule, you think of the ego as one who is peering out of
your own eyes as if leaning on a window sill […] So then: a window looks out
on the world. The world is out there; and in here, what do we have? The world
still- what else could there be? With a little effort of concentration, Mr. Palomar
manages to shift the world from in front of him and set it on the sill, looking out.
Now, beyond the window, what do we have? The world is also there, and for that
occasion the world has been split into a looking world and a world looked at.
And what about him, also known as “I”, namely Mr. Palomar? Is he not a piece
of the world that is looking at another piece of the world? Or else, given that
there is world that side of the window and world this side perhaps the “I”, the
ego, is simply the window through which the world looks at the world. To look
at itself the world needs the eyes (and the eyeglasses) of Mr. Palomar. (Calvino,
1985, p. 114)
The understanding of the “I” or ego here as the window through which
the world looks at the world could be called the attentional self that can be
shift its perspective from an aesthetic bodily-centered perceptive point of
view towards an heautoscopic point of view among others. The basic met-
aphor of the self therefore seems the window through which the world is
observed. Palomar’s undertaking seems derived from the transformation
of a window scene by Caspar David Friedrich famous drawing “Window
looking over the park” to the surreal visual experience of the window of
René Magritte and his painting “The human condition”. Can we imagine
the window through which we look as being put outside – beyond the
frame of the window of subjectivity – and even turned around towards
the window?
Palomar is confronted with the dilemma that he wants to look at

things “from the outside” (Calvino, 1985, p. 113) and not from the in-
side. This means he wants to bracket and thus suspend his ego in the
moment of observation.
Palomar is trying to change the anchor of his attention from a bodily-
centered aesthetic point of view (inside-out) towards a heautoscopic
attention point of view (outside-out) in which even his body is disem-
bodied and his gaze and regard can start not from himself but from
the observed object. This absolute bracketing of the aesthetic attention
seems, however, impossible. The attentional self can be seen as the play
of perspectives and gazes in between aesthetic and heautoscopic poles.
Similarly we will see that attention purely without a body in out-of-
body experience and heautoscopy (Metzinger, 2009) seems impossible.
Breyer24 (2011) also notes that the theoretical transformation of pure
aesthetic towards pure heautoscopic attention in which the own lived
body [Leib] can be seen as an objectified body [Körper] among others
in the thing world and scrutinized from the point of view of a neutral
objective observer is a “liminal phenomenon” that shows the two poles
of attention between the pure immanence of inwardness of the sensing
living body and the pure outward directedness of the neutral objective
observer as impossible, as in reality it can never be reached by phenom-
enological experience permanently suspending one of these poles.
Through this meditation Palomar reaches the problem of identity
and differentiation in attentional observation, as for him observation is
linked to the observation of a self-signifying activity (a sign, a sum-
mons, a wink), and the signification of a self, an “itself ” that signifies
something and thus catches and attaches attention to both poles- of the
observer and to the observed- by separation or distancing one thing
from the other:
From the mute distance of things a sign must come, a summons, a wink: one thing
detaches itself from the other things with the intention of signifying something…
24 “Die Kraft der Aufmerksamkeit reicht […] nicht zu ihrer eigenen Verabsolutierung
in die eine oder andere Richtung hin. Zwar kann man sich den Polen, zwischen
denen sich das attentionale Spektrum erstreckt, je nach den Umständen und den
eigenen Vermögen, die durch Übung modifiziert werden können, relative stark
annähern, doch bleibt stets die anziehende bzw. rückbindende Kraft des jeweils
anderen Pols bestehen” (Breyer, 2011, p. 161).
what? Itself, a thing is happy to be looked at by other things only when it is

convinced that it signifies itself and nothing else, amid other things that signify
themselves and nothing else (Calvino, 1985, p. 115).
Mr. Palomar is, however, also a conceptual personae of failure in the

attentive gaze in observation, as he- from the outset of an naïve phe-
nomenological approach- fails to observe all that there is to see in
a single observation with his bare naked eye: His plan is to main-
tain the duration of observation as long as possible to simply “see a
wave” (Calvino, 1989, p. 4) until he would notice “that images are
being repeated, he will know he has seen everything he wanted to see
and he will be able to stop” (Calvino, 1989, p. 4) his phenomenologi-
cal observation of the single wave. The succession of singularities in
which he wants to notice a pattern of repetition seems a difficult task
and time-consuming as well as tiring for his attention, but also his
disposition as being “impatient to reach a complete, definitive conclu-
sion of his visual operation, looking at waves […]” (Calvino 1989, p.
6). Palomar wants to simplify his observation through the method of
“mastering the world’s complexity by reducing it to its simplest mech-
anism.” (Calvino, 1989, p. 6) in this case a single wave. However, the
observation of everything there is to see with “scientific attention” – a
single wave that then may become a universal of seeing waves – does
not seem at all easy and brings us to the paradox of seeing the reg-
ularity of the singular that in the case of the wave is a very complex
dynamic phenomenon, thus becoming another conceptual personae of
impossible attention.
In another episode regarding the conceptual personae of impos-
sible attention Mr. Palomar rehearses four modes of attentional gaze
(Bernhard Waldenfels, 2006, pp. 96–98) in being attentive to the na-
ked breasts of a woman lying on the beach (the empty gaze, the fixed
gaze, the overflying gaze and the universal all-embracing gaze) that
show the important relation between attention and the ethos of the
senses. Attention and responsibility, amenity, or courtesy of how to
be attentive is closely related. This means: not paying heed and thus
overhearing or overseeing someone can create a socially conflicting
situation as well as being too attentive or (over-) attentive. This also
means that attention is not just programmable but has to be offered
and can also be withhold, can be voluntary given or involuntary
imposed, attention can be withdrawn or denied and as such is cultur-

ally and historically codified by values concerning how to be attentive
to certain situations or not. Attention is valued and socially shared
and thus is joined by social claims of validity. This also means that the
attribution of attention can in a social situation become too much for
the “normal” claim of validity of a certain situation. Therefore paying
close attention – for example to the breasts of a naked woman on the
beach- can be perceived as penetrating over-attention, intrusiveness
or pushiness. At the same time a lack in giving attention for instance
to the beauty of a new haircut or a new dress someone wears may be
being seen as rudeness.
From aesthetic to heautoscopic attention
In the end Mr. Palomar comes to the phenomenological conclusion of

his own being, and that his task is “to look at things from the outside”
(Calvino, 1985, p. 113) while all the observables (from the wave to the
stars, animals or daily objects) “ask” him to prolong his attention to-
wards them. His attentive gaze is caught by a multitude of details ready
for the observation of his attentive gaze by “running all over the details”
and is “then unable to detach itself ” (Calvino, 1985, p. 113) from these
observable details. That is why in the end Palomar sees just one path
he could go and that is to increase his attention, even if the following
questions remain:
1. Can Palomar as conceptual personae of the attentional self ac-

tually observe everything that he wants or ought to observe in a
universal and objective manner, and how long can he stay with
the details or run over all of them in an attentive gaze?
2. Can Palomar anchor his attention not from an bodily centered per-
ceptive (aisthetic25) point of view, but instead from a point of view
“out of his body” and from “inside out of the body of the thing” that
he wants to observe? A heautoscopic attentional self as a scientific
observer is proposed in which even his body is disembodied and his
gaze and regard can start not from himself but from the observed
object. This means that the theoretical transformation of a pure
aesthetic towards a pure heautoscopic attentional self in which the
own lived body [Leib] can be seen as an objectified body [Körper]
among others in the thing-world and scrutinized from the point of
view of a neutral objective observer is a “liminal phenomenon” that
shows the two poles of attention between the pure immanence of
inwardness of the sensing living body and the pure outward direct-
edness of the neutral objective observer in an impossible split. This
means in reality it can never be reached by phenomenological ex-
perience. This absolute bracketing of the real aesthetic attentional
self seems impossible; however, it could be imagined and thought
of, as a metaphysical form of virtual attention, that would, however,
never be independent from a self in a body.
Now52I will concentrate on the implicit “attentional ghost” of Thomas

Metzinger’s virtual and attentional self that in out-of-(real)-body expe-
riences can shift the self-localization and viewpoint of attention away
from an intra-or extramissionist attention, and thus becomes autoscop-
ic from a viewpoint beyond embodiment. All conceptual personae are
fundamental diagrammatic features of the concept of attention which, if
taken on its own as the only necessary feature or function, show them-
selves to be impossible, and as such would provoke a strange and im-
possible phenomenon in which an episodic attentional self would gain
no dynamic of experience in between stability and imbalance.
25 Aisthesis (perception) refers to a form of aesthetics that is based on perception.

But this does not mean that one could reduce aesthetics to attentive perceptual
routines. (See: Gerner, 2010b).
2.2 Metzinger’s “attentional ghost”: the virtual self floating out of the

body and the impossible full body illusion
Let’s look at a non-orthodox experience of attention that lies in between

sleep and wakefulness. Here attention seems to do the impossible and
follow a virtual self, leaving the material body or the attentional focus,
going up in the air and floating ghostlike over one’s body. This figure
of impossible attention is referred to as Out of Body experiences (OBE)
and has been recently investigated as an empirical phenomenon in the
Cognitive Neurosciences as we will see. Metzinger makes the point
that non-standard cases of self experience are an epistemological im-
portant entrance door for verifying and researching into his theory of
the minimal phenomenal self and its model (even though non-standard,
deviant or pathological, extraordinary cases of selfhood might not au-
tomatically lead to a correct description of having a self or “selfhood”.
For Metzinger what is important in these non-standard, altered and/or
pathological experiences is what fails and shows up and this gives us ac-
cess to in comparing 1st PP reports (of patients or research subjects) with
neuro-scientific 3rd PP studies (trying as well to understand the “neural
correlate” to the phenomenological description of what appears in expe-
rience, which leaves the difficult questions of causality still open for de-
bate and the philosophical foundation of correlates in general) into how
a permanent body-“model” is transformed. The phenomena he deals
with in the third chapter of The Ego Tunnel (2009) called “Out of the
Body and into the Mind: Body-image, Out-of-Body Experiences, and
the Virtual Self ” are non-orthodox phenomena such as the Out-of-body
experience which for a long time where treated as and in para-/pseudo-
science, but which gained momentum in psychology in the 1980s (see
Blackmore’s research) although they have long been “career obstacles”
for researchers choosing these topics. However, since the end of the late
1980s and 1990s they have been picked up by neuroscientists, (neuro-)
psychologists and psychiatrists such as Devinsky (1989) or have been
treated in relation to seizures and autoscopic phenomena as well as con-
cerning felt presences, multiple perspectives and duplications of one’s
body (phantom body); Botvinick & Cohen (1998) regarding the rubber
hand illusion, and which recently entered a new phase of research in
which Metzinger became involved with as did the Swiss neuroscientist
Olaf Blanke (2002; 2004; 2008) and the Swedish researcher Henrik
Ehrsson (2007; 2012). The latter actively induced these phenomena in
test persons through brain-area stimulations, that of course also rais-
es lots of ethical questions that Metzinger is also aware of (in the last
chapter of his book (2009) he asks what a good state of consciousness is
(and what are non-desirable or eliminable mental states or experiences;
also Metzinger tries to induce certain abnormal body experiences in
test persons by stimulating certain brain areas, swapping bodies or to
evoke a full body illusion as in Blanke and Metzinger’s test case (2008).
Other phenomena in which our attention is working in between sleep,
dream and wakefulness are for example trance-like states in between
dreaminess and wakefulness and out-of-body-experiences which are an
example, that through attention guide our body image into an unusual
or non-orthodox experience of our body-selves. It seems in this case
that our attention is directed in a way that a regular perceptual state
would not enable us to. We can say that our perceptual map of ourselves
is dislocated, our attention is placed on a mapping of a virtual body self
in an out-of-body experience, which means that OBE form part of being
attentive to ourselves in a non-orthodox way, and are less grounded by
our surroundings than in perceptual attending.
The in-between event between wakefulness and dream attending of
ourselves from an out-of-body point of view had happened to myself,
when I personally experienced something that I had never felt before and
that was not a mere dream. At the same moment that I had broken my
leg and had been prepared to be operated on after being anaesthetized
in Germany over 20 years ago, I had a proper out-of-body experience in
which my body left my fixed body, which was not able to move on the
operation table and the virtual body floated as an “out-of-body” ethereal
body trying to start to play football but my body was floating across the
operation room looking down on me and I couldn’t control it myself,
and then the OBE stopped as the anesthesia had put me in a sleep state.
2.2.1 The concept of disembodied attentional self in

an out-of-body experience (OBE)
It seems evident: attention and an attentional self both are embodied.
But having said this, is it actually always necessarily put into perspective
from one unique and unified fixed body position? Can attention actually
wander its centeredness or attentional self-location, and can it be dislo-
cated, willingly or not, can the perspective with what we feel, perceive,
reason or image be experienced subjectively out of a, my or the body?
The heautoscopic and autoscopic, as well out of body experi-
ence will now be considered and what this can tell us concerning the
“impossible” attention out of a living body and how this relates to the
attentional self in general:
Let’s consider this phenomenal relation of attention in the instance
of a body by taking a closer look at Thomas Metzinger’s research into
the seemingly impossible but actually experienced out-of-body experi-
ences (OBE) (see Metzinger, 2009, chapter 3, Metzinger, 2005). Even
though I will not come to the same representionalist conclusions about
a permanent “virtual body model” as Metzinger does, his observations
and research is valuable in respect of this phenomenon in between
wakefulness and dream in which attention plays a major role in non-or-
thodox self-experience such as Out-of Body experiences.
Metzinger refers to his own OBE in his recent “The Ego Tun-
nel” book (2009) and refers to several other researchers, specifical-
ly the OBE and consciousness researcher Susan Blackmore (1982;
1984; 1987, see also: 1993) whom he met for the first time in 1995.
Moreover, he draws on the work of the Swiss biochemist Ernst Waelti
and also research into the pathological conditions of OBE, for exam-
ple OBE induced by seizures. In the domain of empirical psychology
Blackmore carried out several review studies and came up with a sci-
entifically sound empirical theory for these strange body phenome-
na. What is interesting in her book on near death experiences (1993)
is how she collects data from phenomenological 1st person reports
and structures their “grammar”. One of the characteristics she found
out was the jump-like movement of the “ethereal” body movement.
Another point Metzinger picks up from Blackmore’s26 research is the
26 Blackmore (1993) shows how subjectively experienced OBE as “religious limit

phenomena” of a “soul” flouting out of the body or having “tunnel” visions
of light at the end of the tunnel merely are phenomenological evidence for an
injured and altered self (under the influence of drugs or anaesthetics) or even a
“dying” brain state of consciousness, such as triggered by the fact that the brain
in certain moments (dying) undergoes lack of oxygen (and as we might all know
following idea: Out-of-Body experiences are “models of reality creat-

ed by the brains, that are cut off from sensory inputs during stressful
situations and have to fall back on internal sources of information”
(Metzinger, 2009, p. 87). Here a conflict of two processing sources
(perceptual or observational feedbacks) and internal models could
create different non-unified body representations and thus two types
of attentional selves. This could explain why during the OBE I was
angry that my body would not move towards the ball but floated over
the ball in the operation room. Metzinger also underlines in Black-
more the idea that memory reconstructing visual cognitive maps of
past experiences – say for example movement maps of walking on the
beach – are most often organized from a geometrical 1st PP that is not
our actual real but motionless body perspective – that is, a bird’s eye
perspective27 rather than in real life (online) human- height perceptive
situation.
But what exactly are Out-of-Body experiences (OBE) and what
have they in common with other autoscopic phenomena? And especial-
ly why are they important for understanding the self?
We know that the representation of the self as bodily self is complex,
and if we take the idea of a basic bodily self serious we are confronted
by a multitude of bodily systems that have to be integrated into one more
or less coherent body schema and a body image that include motor sig-
nals as well as multisensory signals (somato-sensory, vestibular, visual,
auditory, visceral). First of all OBE are autoscopic body phenomena. Au-
toscopic phenomena as Bolognini et al (2012) put it, “refer to complex
experiences involving the illusory reduplication of one’s own body”. One
could add that OBE not only refer to the dislocation of body parts, but that
introduce a global shift in the body image. According to Mohr & Blanke
(2005) autoscopic phenomena (AP) are rare, illusory visual experiences
how the excess of oxygen and breathing rate is a hallucinogenic “natural drug”
experience, we can imagine how – on the contrary – a lack of oxygen in the
brain can trigger specific experiences while the brain and the “one” that experi-
ences, dies (if – of course – the people coming close to death, survive and can
still report their experiences).
27 “Close your eyes and remember the last time you were walking on the beach. Is
your visual memory one of looking out at the scene itself? Or is it of observing
yourself, perhaps from above, walking along the coastline? For most people the
latter is the case” (Metzinger, 2009, p. 87).
during which the subject e.g. has the impression of seeing or feeling a
second own body in extrapersonal space. AP – in their view – consist
in “out-of-body experience, autoscopic hallucination, and heautoscopy”.
For Brugger et al (2006), Heautoscopy is the encounter with one’s dou-
ble (the reduplication of a single body and self and thus a breakdown of
integrative processes that let me identify with my body or my self), in the
sense of a multimodal “illusory” reduplication of one’s own bodily self.
Aspell et al (2012) express it as following:
An individual undergoing an OBE usually experiences a dissociation between his

self-location and his first-person visuospatial perspective with respect to the seen
location of his own body – in other words, he perceives his own body (and the
world) from a spatial location and perspective that does not coincide with the seen
position of his body.
The phenomenon of polyopic heautoscopy (a multiplication of body and

self) according to Brugger et al (2006) “points to the multiple mappings of
the body, whose disintegration may give rise to the illusory experience of
multiple selves.” Autoscopic phenomena deal with a) viewpoint changes,
b) “illusionary” self-identification, c) altered or abnormal self-location(s)
and d) changes in the first- person perspective (see: Aspell et al., 2012)
as studies of Ehrsson 2007 and Langenegger et al. 2007 show by the ma-
nipulation of multisensory cues that the brain is supposed to use in order
to “create a representation of self-location and identity” (Aspell et al.,
2012). Blanke (2005) defines out-of-body experiences in the following
manner:
In an out-of- body experience, subjects feel that their “self ”, or centre of awareness,
is located outside the physical body and somewhat elevated. (Blanke, 2005, p. 173)
OBE can be idiopathic, self induced or induced by non-invasive tech-

nological aid (Blanke & Metzinger, 2009) using for instance video
(Lenggenhager et al., 2007), virtual reality (Ehrsson, 2007) or robotic
devices (Ionta, 2011), inducing changes in the self-location, self-iden-
tification or first person perspective in healthy subjects. Moreover,
recent research has not only described phenomenologically these
strange doubling, mirroring or shadowing phenomena of a “disrupt-
ed” self (Mishara, 2010) but has shown as well that invasive manipu-
lation of the brain can even induce a “illusory shadow person” (Arzy
et al., 2006) by artificial brain stimulation for instance in the paradigm

of asynchrony stimulation of different modes of perception (e.g. tac-
tile and visual).
*
Metzinger, following Blanke calls 1) OBE, as well as 2) autoscopic
28
hallucination and also 3) Heautoscopy “illusionary global own body

perceptions” (Blanke/Metzinger, 2008, pp. 9–10) or autoscopic phenom-
ena “leading to alterations of the minimal phenomenal self ” (ibid, p. 10).
Diagram 1, © Gerner.
This diagram (adopted from: Blanke & Metzinger, 2008, p. 10) shows the dynamics of the
attentional self-location [SL] in autoscopic experience: Three cases of direction of the atten-
tional point of view, either from the hallucinated body towards the somatic body in Out-of-body
experience (virtual observer perspective) or from the somatic body towards the hallucinated
virtual body (somatic 1P perspective) in Autoscopic hallucination or both ways in Heautoscopy
(switching between somatic 1P perspective and virtual observer perspective). The somatic and
the virtual body in these three cases of observer perspectives always face each other on the
contrary to a “felt presence” from behind (existential feeling perspective of autoscopy), another
form of autoscopic experience.
These altered self phenomena are defined by “seeing a second own

body in extracorporeal space” (see: Metzinger 2005; Blanke 2008 my
underlining) and are phenomenologically distinct in the sense of the
28 Blanke (2005) in a very interesting article on the relation of heautoscopic

experience and self-portrait in artists resumes this phenomenon as follows: “To
conclude, one might say that autoscopic hallucinations and AH [autoscopic hal-
lucination] self-portraits reflect that the self perceives its body from the outside
and as a mainly visual body. The reduplicative character of the double is only
implicitly present and the mechanisms of ‘seeing’ predominate over spatial and
body-perceptual mechanisms. The self ‘sees’ its body as the body of somebody else.
a) self-identification, b) the direction of the visual weak 1st PP and

c) the self-location either in the physical body as in autoscopic hallu-
cination, or in the floating above body looking down onto the physical
body as in OBE, or in the either/or direction of the weak 1st PP in
the physical or illusionary body in Heautoscopy. All these illusion-
ary global body perceptions can be caused by different pathological
conditions in specific brain areas. In general heautoscopy and OBE
are abnormal minimal phenomenal self-extensions towards a strong
1st PP in which by self-identification with the incorrect content of a
global body representation attention is misled by wrong global body
perceptions. This seems to be the difference of heautoscopy to the
state of dreaming in which attention cannot be specifically directed.
In an OBE self-location is never at the position of the physical body
as represented, therefore OBE are associated with the feeling of “dis-
embodiment (the experience that the subject of conscious experience
is localized outside the person’s bodily borders)” (Blanke/Metzinger,
2008, p. 9)
There is another non-standard phenomena of a feeling of a (non-
visual!) presence of another body behind the actual one, not seen but
felt as a presence behind, that can also be induced by stimulation of
certain brain areas (in this case the left angular gyrus, while in the same
study by Blanke (2002), the stimulation of the right angular gyrus re-
sulted in an OBE, “as if the patient were floating from the ceiling and
looking down and being directed from the floating body onto herself ”
(Metzinger, 2009, p. 96).
Out-of-body experiences and OBE self-portraits also reflect the self, perceiving
its body from the outside, but not only as a visual body but as a body in a complex
spatial scene of which the body is only a part. The reduplicative character of the
double is explicit and the painter’s self is distanced and elevated from its body. The
self is ‘out of touch’ with its somato-sensory perceptions and ‘sees’ its body with
the eyes of somebody else. Finally, heautoscopy and heautoscopy self-portraits
reflect the perception of the body from the inside via somatosensory and motor
mechanisms. The self is in touch not only with her or his bodily feelings but also
with the bodily feelings of the double. The self does not ‘see’ but mainly ‘feels’ its
body and the body of the double to be the same body.”
Diagram 2, A felt presence behind, © Gerner.
In recent studies Blanke et al have discovered other brain areas that are
supposedly involved in the causation of OBE when leisured brain areas
are stimulated (as in seizure patients that report OBE): the extrastriate
body area (on the right half of the temporo-parietal junction) and the left
half of the temporo-parietal junction (Blanke et al. 2004; 2005). The brain
region of the temporo-parietal junction (TPJ) is also activated in “less
than half a second” (Metzinger, 2009, p. 97) if a test person is asked to
imagine himself as if in the position of an OBE watching his body from
above. “If this brain region is inhibited by a procedure called transcranial
magnetic stimulation, this transformation of the mental model of one’s
own body is impaired.” (Metzinger, 2008, p. 97) Metzinger, as Blanke,
expresses two pathological conditions for OBE: 1) disintegration on the
level of the self (or self-model in the words of Metzinger) 2) the conflict
between the visual space and the vestibular sense of balance:
Two separate pathological conditions may have to come together to cause an OBE.
The first is disintegration on the level of the self-model, brought about by a failure
to bind proprioceptive, tactile and visual information about one’s body. The sec-
ond is conflict between external, visual space and the internal frame of reference
created by vestibular information, i.e., our sense of balance. In vertigo or dizzi-
ness, for example, we have problems with vestibular information while experienc-
ing the dominant external, visual space. If the spatial frame of reference created
by our sense of balance and the one created by vision come apart, the result could
well be the conscious experience of seeing one’s body in a position that does not
coincide with its felt position. […] Finally when an epileptic patient whose OBEs
were caused by damage to the temporo- parietal junction was asked to stimulate
mentally an OBE self-model, this led to a partial activation of the seizure focus.
Taken together these observations point to an anatomical link among three differ-
ent but highly similar types of conscious experiences: real, seizure-based OBs;
intended mental simulations of OBs in healthy subjects, and intended mental sim-
ulations of OBEs in epileptic subjects (Metzinger, 2009, p. 96).
Important in my view is what Metzinger puts at the end of his neuronal

explanation of OBE, and that is that in deviant or pathological condi-
tions, as in the case of seizure patients, the phenomenology in these pa-
tients in comparison to healthy subjects differs in the first seconds of an
OBE, and again show a different phenomenology in those you learned
to influence mental states by meditation. Metzinger does not mention
examples such as those of attentional absorption into a task or a situa-
tion such as in post-voluntary flow-experiences (See Bruya, 2010; Dor-
mashev, 2010) which makes his assumption of getting closer to what the
minimal form of selfhood is by examining non-standard cases of self
experience epistemologically at least a more complex issue:
The initial seconds clearly seem to differ between spontaneous OBEs in healthy
subjects and those experienced by the clinical population, such as epileptic pa-
tients. The onset may also be different in followers of certain spiritual practices.
Moreover, there could be a considerable neuro-phenomenological overlap be-
tween lucid dreams and OBEs as well as body illusions in general (Metzinger,
2009, p. 98).
Metzinger proposes – and indeed has to do so if OBE were to be of val-

ue to discovering the minimal phenomenal self – that OBE are general
human experiences, and do not only appear in deviant or pathological
conditions or are merely culturally based narrative inventions. As fur-
ther examples he refers to situations such as meditation and I would add
flow experiences (skilled experiences that had to be learnt with effort
and become habit in a way that they are experienced as post-voluntary

effortless flowing, such as skiing down a ski slope after having learnt
how to ski well), and the example he gives (but not designates as such)
of a Marathon runner, where after some kilometers the woman goes
into a trancelike state in which her body floats above her and her vision
is of a bird-eye’s view watching the landscape from above. This unor-
thodox change in the perspective or the center of projection of the “I”
I would designate as unorthodox attentional self maps, in which the
dimensionality, the projection type and the material reference towards
a real world scene is altered with attention occurring in an unusual way
that we have no habitual experience of. It is a new map that occurs to us
as an unusual or non-habitual orientation, but where we still can alter
and move around the focus points we want to reach attentively. Seeing
myself, however, floating out of my material body position is, however,
an altered non-orthodox experience of my self.
This, however, brings us deeper into the concept of a virtual
“self-model” that attention supposedly controls.
For Metzinger no such “things” as selves as substance exist, as he
explained in the Introduction to his book “Being No one. The self-model
theory of subjectivity”:
[…] no such things as selves exist in the world: Nobody ever was or had a self.
The phenomenal self is not a thing but a process – and the subjective experience
of being someone emerges if a conscious information processing system operates
under a transparent self-model. (Metzinger, 2004, p. 1)
The fiction of the “self ” and its transparent model is said to have sub-
jective selective advantages to form, for example, “higher forms of
thoughts” (Ware, 2006, p. 1). In conscious experience Metzinger iden-
tifies three characteristics related to the self: a) Mineness, the notion
that something belongs to the self (i.e. my thought, my feeling, my legs,
my choice) b) selfhood, the idea that “I am a person”, the maintenance
of an identity over time c) Perspectivalness a certain understanding of
orientation in a representational space (giving a self a positioning and a
relation to other representations) implying experience of a global atten-
tional self and the relation of the self and some other object. The self,
here defined as the “center of awareness”, is defined by Metzinger as
the possibility of being able to manipulate the focus of attention. This is

the reason why with Metzinger we can call it an attentional self.
In Metzinger’s representational view we have no direct access to
the world, but the senses and their underlying neurobiological complex
systems and functions create internal objects and contents that can be
observed and manipulated by conscious “processing”. Metzinger pro-
poses the elimination of an ontological self not for the sake of reduc-
tionism but for the sake of a more sound explanation of what conscious
experience is said to be:
It has now become clear that we will never solve the philosophical puzzle of
consciousness- that is, how it can arise in the brain, which is a purely physical
object- if we don’t come to terms with this simple proposition: that to the best of
our current knowledge there is no such thing, no indivisible entity, that is us, nei-
ther in the brain nor in some metaphysical realm beyond this world. (Metzinger,
2009, p.1)
Thus Metzinger tries to address exactly the next question that arises
from a “being no one” perspective of the self in conscious experience:
So when we speak of conscious experience, what is the entity having these expe-
riences? (Metzinger, 2009, p. 1)
Metzinger tries to answer this question of the first-person perspective

through the intention to explain a) the theory of an apparent self with-
in his naturalist “phenomenal Self-Model (PSM)” a multimodal plastic
structure that will in principle for Metzinger be explained by neuro-sci-
entific findings probably involving distributed processing and the b) the
conceptual metaphor of the “Ego-Tunnel” as the “Consciousness tunnel
that has evolved the additional property of creating a robust first person
perspective, a subjective view of the world” (Metzinger, 2009, p. 11,
my underlining). For him these two are the explanatory challenges in
describing how “[…] a genuine sense of selfhood appears.” (Ibid, my
underlining)
He continues as follows: “We have to explain your experience of
yourself as feeling the tactile sensation in the rubber hand, of yourself
as understanding sentences you’re reading right now. This genuine con-
scious sense of selfhood is the deepest form of inwardness, much deeper
than just being “in the brain” or in a “simulated world in the brain”. This
non-trivial form of inwardness is what the book is about.” (Metzinger,

2009, p.12, my underlining). He evokes the “transparency” of the un-
derlying (neuroscientific) mechanisms of conscious experience as an
argument for the process of a self- modeling its existence. In general
Metzinger’s studies want to “contribute to a deeper understanding of
non-conceptual, pre-reflective layers in conscious self-representation”
(Metzinger, 2010, p. 25), in the embedded bodily self, that are distinct
from for example the contents and the relations of visual perceptive ex-
perience in the “seeing self ” and what is perceived, in which he assumes
the following theses:
1) Metzinger conceptually defends the claim that agency is not
“part of the metaphysically necessary supervenience- basis for bodily
self-consciousness.”(Metzinger ibid, see also: Blanke/Metzinger, 2008).
But what are the constitutive conditions for selfhood after Metzinger,
what is “truly necessary and not only sufficient to bring about an ego”
(Metzinger, 2009, p. 101)? For Metzinger agency (see chapter 4 of his
book and the phenomenon of the alien hand) is not a necessary condi-
tion, as it only selectively manipulates two dimensions: a) self-identifi-
cation or “[…] the degree to which an organism identifies with the con-
tent of a global body representation” (Blanke/Metzinger, 2008, p. 7) or
with the “content of a conscious body image” (Metzinger, 2009, p. 101)
and b) self-localization as the embedding principle of the self in space.
Self- localization is necessarily spatial-temporal including the feeling
of “nowness”. For Metzinger the target phenomenon of consciousness/
conscious experience lies in basic minimal phenomenal embodied self-
hood that can be “causally controlled by multisensory conflict alone”
(ibid.) without any meta-control or author of action. For this minimal
phenomenal self a “seeing” self for him is neither necessary as an “emo-
tional” self nor a self-reflective or narrative self. The seeing self is not
necessary because
[…] you can shut down the windows in front of the little man behind your eyes.
The seeing self disappears. The Ego remains. You can be a robust, conscious self
even if emotionally flat, if you do not engage in acts of will, and also in the ab-
sence of thought. Emotions, will and thoughts are not necessary to the fundamen-
tal sense of selfhood (ibid.)
2) On the level of the empirical and phenomenological foun-

dation of his theory he wants to show how several phenomena of a)
non-standard partial (body-) self experience such as the rubber hand
illusion and b) “illusionary own global body perceptions” in which he
includes “Out of Body experiences”, (virtually induced) “full body il-
lusions”, “autoscopic hallucinations” and “Heautoscopy” that can be
seen as an interesting “entry point” for researching into “the nature
of the phenomenal self ” (Metzinger, 2010, p. 25) in general and the
“Minimal Phenomenal Self ” (MPS) in particular. The MPS is related to
embodiment and the simplest form of self-consciousness. Embodiment
deals with the sense of “ownership” and identification with the body as
a whole (the phenomenological sense of “mineness”) and is specifically
important as it anchors the self, situating cognition through the subjec-
tive experience of having a body. One can, however, interpret the rubber
hand phenomenon not merely as an “illusory” phenomenon, but as an
extension or better an addition of an extra co-present object (the rubber
hand) to the bodily self as Thomas Fuchs underlines:
Functionally the rubber hand connects temporarily to the body, because by that the
best possible coherence of different sense-modalities is installed […] this does not
turn the sensed body into an illusion. (Fuchs, 2012, p. 33)
3) Metzinger tries to install a new research target in the investigation

of the attentional self, which he calls the Self Model Theory of Subjec-
tivity (SMT) in which the most important research entity – besides his
questionable general constructivist “Phenomenal Self-Model” (PSM) –
is what he calls the “Minimal Phenomenal Selfhood” Hence he differ-
entiates epistemologically between different 1st Person perspectives (see
Blanke/Metzinger, 2008, p. 7):
He calls this first and basic state of a minimal phenomenal self,
“selfhood-as-embodiment” (Metzinger, 2009, p. 102) or “weak 1 PP”
(Blanke/Metzinger, 2008, p. 7), which he describes as a “calm, emo-
tionally neutral state, deeply relaxed and widely alert, a state of pure
observation, without any thought, while a certain elementary form of
bodily self-consciousness remains” (Metzinger, 2009, p. 102). On his
immanent bodily account of the self it is not a possessive self in the
sense of “global ownership or “The feeling of this body being mine”,
because global ownership already presupposes three entities: a) an
invisible self b) the body and c) a relation of body and self. On the
contrary this basic selfhood for Metzinger is “the body that possesses
itself ” (ibid.) The question of ownership is put into perspective in the
following way:
Owning means to be able to control it, and selfhood is intimately related to the very
moment in which the body discovers that it can control itself- as a whole”(ibid.)
The example he gives is the initial moment of “coming to yourself ” in the moment
of waking up. The self is however not about control as such but about the enabling
conditions for control. The minimal phenomenal self thus includes a bodily sense
of space (spatial frame of reference) and time (the feeling of now), a body image
and the transparency of this body image (the organism creating this image does
not recognize it as an image). Moreover, the Here- and Nowness has to be accom-
panied by a “visual (or auditory) perspective originating within the body volume,
a centre of projection, embedded in the volume of the body. (Ibid.)
*
Metzinger proposes a three levels of the self in which attention plays
a significant role, however, only on the second and third level, as for
him phenomenal “attentionality” would be only linked to a weak 1st PP
in which in his view still there is no attention (in the sense of an active
control mechanism of the focus of attention).
Firstly, for Metzinger there is the weak 1st person perspective (1PP)
a “purely geometrical feature of an egocentric model of reality” that is a
spatio-multimodal model of reality (spatio-visual, spatio-auditory etc.).
This includes a) a spatial frame of reference and b) a global body
representation. Hence the body projects c) a perspective “originating
within this body representation”. The weak 1st person perspective orien-
tates the body through a centre of projection, “[…] which functions as
the geometrical origin of seeing (or hearing etc.) Organism’s perspec-
tive.”
I would call this the initial orienting, attentional bodily self.
Secondly, and important in relation to attention in the sense of
Metzinger is the “selfhood-as-subjectivity” (Metzinger, 2009, p.
102) or “strong 1 PP” (Blanke/Metzinger, 2008, p. 7), or strong-
er selfhood. The transition from the minimal phenomenal self to the
self-as-subjectivity is marked -for him- by the supplementary self-
representation of the minimal selfhood as being directed towards an
object that is by being able to change the location and duration of
attention in attentional agency or how he calls this by the “controlling

the focus of attention”. Thus the stronger -not just minimal but also
subjective phenomenal- self that first of induces a perspective after
Metzinger’s hypothesis happens
[…] when we first discover that we can control the focus of attention. We under-
stand that we can draw things from the fringe of consciousness into the centre
of experience, holding them in the spotlight of attention or deliberately ignoring
them – that we can actively control what information appears in our mind. Now
we have a perspective, because we have an inner image of ourselves as actual-
ly representing, as subject directed at the world. Now we can, for the first time,
also attend to our own body as a whole – we become self-directed. Inwardness
appears. […] when we attend to the body itself […] Consciousness is the space
of attentional agency. Selfhood as inwardness emerges when an organism for the
first time actively attends to its body as a whole. If a global model of the body is
integrated into the space of attentional agency, a richer phenomenal self emerges.
It is not necessary to think. It is not necessary to move; the availability of the body
as a whole for focal attention is enough to create the most fundamental sense of
selfhood-as-inwardness – that is, the ability to become actively self-directed in
attention. The body model now becomes a self- model […]. The organism is now
potentially directed at the world and at itself at the same time. It is the body as
subject (Metzinger, 2009, p. 102).
Thus the strong 1PP seems to be a global representation of the organ-

ism as a whole, given through a) Minimal Phenomenal Self and b) as
being “directed” towards something, an “object component” (including
the body itself). This direction of the 1PP towards a thing of experience
for Metzinger necessarily includes the ability “to control the focus of
attention”. For me this could be called the noting self.
Thirdly, Selfhood as cognitive 1PP: The cognitive 1PP occurs
“[…] when a system possesses a concept of the strong 1PP and is able
to apply this concept to itself (i.e. it has an abstract and active mental
representation of itself as a subject of experience which includes a spe-
cial form of cognitive self-reference).” The cognitive 1PP presupposes
the capacity of self-reference in the sense of using concepts such as “I”
in “I (myself)-thoughts”. “I myself am a subject of experience”. This
cognitive 1PP for Blanke/Metzinger (2008, p. 7) might be exclusively
human “Many organisms might have phenomenal self-models, but per-
haps only humans have self-concepts.”(ibid.). This could be called the
self-attending (and self-observing) self.
2.2.2 Virtual Out- of-body experiences and impossible

full body illusions
Metzinger and Blanke (2008) amplify the notion of a partial body il-
lusion as in the rubber hand illusion that I will sketch out here briefly.
The rubber hand illusion takes up a strange phenomenon that Botvinick
& Cohen (1998) called “Rubber Hand ‘Feels’ Touch That Eyes See”
regarding an artificial body part lying on the table (the rubber hand)
while the real left hand is hidden under the table (as the right arm and
hand). Then a probe will repeatedly simultaneously stroke the visible
rubber-hand, as will the real corresponding body part under the table.
After a while (about a minute) the body image is redirected towards the
visible rubber hand (body part), even creating a link (an illusory arm)
that normally is accompanied with a sound or a crack felt when redirec-
tion of the virtual body part is active in this illusion.
Metzinger and Blanke (2008) now distinguish between partial vs.
global ownership of a body and carried out an experiment within a vir-
tual reality setting to simulate a full body illusion and a swapping of
the notion of one’s 1st PP towards an illusionary body in front of the test
person.
A head mounted display consisting of goggles that showed two separate images
to each eye, creating the three-dimensional illusion of being in a virtual room.
Subjects were able to see their own backs, which were filmed from a distance of
two meters and projected into the three dimensional space in front of them with
the help of a 3D decoder. When I acted as the subject of the experiment I felt as
if I had been transposed into a 3D-version of René Magritte’s painting La Repro-
duction interdite. Suddenly I saw myself from the back standing in front of me.
While I was looking at my back Bigna Lenggenhager was stroking my back, while
the camera was recording this action. As I watched my own back being stroked,
I immediately had an awkward feeling, I felt suddenly drawn towards my virtual
body in front of me, and I tried to “slip into” it. This was as far as things went.
(Metzinger, 2009, p. 99)
The most “interesting” outcome of these experiments (even though

Metzinger’s conclusion of a reductive “geometrical view” of the self –
and thus separated from biological matter and body structures – does
not convince me) seems to be the conceptual conclusions Metzinger
proposes after a series of tests with one’s own body being stroked (own
body condition) or a fake body being stroked (fake body condition) and
an object being stroked synchronically and out of synch (control condi-

tion) the “embedding principle”:
The bodily self is phenomenologically represented as inhabiting a volume in

space, whereas the seeing self is a point without extension- namely, the centre of
projection for our visio-spatial perspective, the geometrical origin of our perspec-
tival visual model of reality. Normally this point of origin (behind the eyes, as if
a little person were looking out of them as one looks out of a window) is within
the volume defined by the felt bodily self. Yet as our experiments demonstrated,
seeing and bodily self can be separated, and the fundamental self of selfhood is
found at the location of the visual body representation. (Metzinger, 2009, p. 100)
3. Outlook
We are confronted with lots of questions – that we cannot answer for

now – but it is still important to clarify what the results of this research
are, even if we do not have all the final answers but are rather con-
fronted with posing future questions to be investigated. In this bold but
strange account of the very interesting phenomenon of the so called
“full body illusion” and the out of body experience the following ques-
tions and critical points, and the approximation towards the phenome-
non of the “attentional self ” serve not as a conclusion but as an outlook
within the impossible and the more possible fields of attention. It seems
not so clear that epistemological questions concerning standard vs.
non-standard cases of experience are easily compared. Does a dying
brain always inform us about the living brain in all its functions or does
a dying brain simply give us more insight into how a brain with impaired
blood-flow or lack of oxygen alters our states of mind. Another point
remains unclear: Why is it that the visual geometrical self should be the
basic condition for the minimal phenomenal self (what about time con-
sciousness, what about the basic feeling of being affected? What about
touch and the skin as a compass and map of the body, that is not to be
dislocated and touched elsewhere, as it is the proximate sense and not,
as with the case of vision, the sense for approximating what is in the
distance). Attention can arouse what is affected involuntarily through
sensory stimulation of one’s own body skin, though not from the stim-
ulation of merely seeing another body stimulated. This question can be
linked to Metzinger’s representative account of the self as a fiction of a
transparent model in general: Metzinger himself asks
[…] The organism is now potentially directed at the world and at itself at the same
time. It is the body as subject./ But again- who controls the focus of attention?
In our Video Ergo sum study, who is the entity misidentifying itself? (Metzinger,
2009, p. 103)
There logically cannot be a deception happening in Metzinger’ s illu-

sionary self-model if there is “no one” or “no entity” to be deceived in
the first place by the “self ”. And still there is that feeling to know what
it is like to be myself today this morning different from yesterday night
when I finished this text in the year 2012, and still the one I am referring
to and comparing the two notions of myself from yesterday and now,
as well as Valéry’s question of what it is that makes up the center, what
forms the center of attention? How are the episodic attentional self –
Monsieur Teste – and the stabilizing and responding attentional self –
Emilie Teste – interconnected, how are they “husband” and wife” inside a
unified attentional self? These are deeper philosophical questions of as-
cribed identity that Metzinger leaves out in his “illusionary” account of
the self. Again we can pose another question to be clarified in the future:
Does the search for a minimal phenomenal self help in clarifying the
question of conscious experience and the different modes of experience
of an attentional self? To deeply link the different modes of attention
such as in aesthetic and heautoscopic attention, attention in distraction
(self-absorption, meditation etc.) and self-consciousness in relation to
the non-self is beyond the scope of this paper but should be developed
in the near future when describing a complete account of an attentional
self. Thus, what the attentional self is and how it works in all its variety is
an important future task of a philosophy of cognitive science. Is an atten-
tional self an epistemological construct that joins separate research are-
as, but does not clarify what a non-dualist, or immanence account of the
self should be about? This remains unanswered as well as the following
question: How can OBE exactly help to clarify what the self is as a centre
of projection and how does attentional control of focus actually stabilize
the first person perspective? What role does distraction play in the
formation and the working of the self? Does attention therefore, besides
attending with a specific sense modality, introduce a specific modus, in
between virtuality (self) and the actuality of attending perception? If at-
tention can follow the virtual body floating and look from above, then it
should not necessarily be seen as a perceptive phenomenon, but as a vir-
tual phenomenon that should also combine important conflicting inputs
(as for instance visual and tactile) in a mereological account of a self.
This also includes not only the notion of the visual body but the whole
range of the multimodality of senses (somato-sensory, vestibular, visual,
auditory, visceral) and especially the question how these sensible parts
of the self are interconnected synchronically and even diachronically as
well as synthesized in a global embodied self experience and linked to –
what Metzinger calls – the orientation aspects of perspective (for him the
weak 1st PP). The question remains however: Are these aspects merely
representational and just “illusionary” or are they real and imminent in
the bodily senses? These important questions that have arisen in our re-
search show that it is fruitful to think more about attention and the self
together in a mutual close relation and discover more about what can be
called an “attentional self ” that we modestly started to explore with two
of its conceptual persona.
References
Aspel, J. et al. (2012). Multisensory Perception and Bodily Self-Con-

sciousness. From Out-of-Body to Inside-Body Experience. In
Murray, M., Wallace, M. (eds.). The Neural Basis of Multisensory
Processes. CRC Press: Boca Raton online: <http://www.ncbi.nlm.
nih.gov/books/NBK92848/>.
Assmann, A. & Assmann. J. (eds.) (2001). Aufmerksamkeiten. Archäo
logie der literarischen Kommunikation VII. München: Fink.
Blackmore, S. (1982). Beyond the body: An Investigation of Out-of-
the-body Experience. Granada: London.
Blackmore, S. (1984). A psychological theory of the Out-of-Body Ex-
perience. Jour. Parapsychol. 48, 201– 218.
Blackmore, S. (1987). Where am I? Perspectives in Imagery and the

Out-of-Body Experience. Jour. Mental Imagery, 11, 53– 66.
Blackmore, S. (1993). Dying to live. Near Death Experiences. New
York: Prometeus.
Blanke, O., et al. (2002). Stimulation of Illusionary own body percep-
tions. Nature, 419, 269–270.
Blanke, O. (2004). Out-of-body experiences and autoscopy of neuro-
logical origin. Brain, 127, 243–258.
Blanke, O. (2005). Perception and experience of the self in autoscopic
phenomena and self-portraiture. Schweizer Archiv für Neurologie
und Psychiatrie, 156(4), 173–188.
Blanke, O., Mohr, C. (2005). Out-of-body experience, heautoscopy, and
autoscopic hallucination of neurological origin. Implications for
neurocognitive mechanisms of corporeal awareness and self con-
sciousness. Brain Research Reviews, 50(1), 184–199.
Blanke, O., Metzinger, T. (2008). Full body illusions and minimal phe-
nomenal selfhood. Trends in Cognitive Science, 13(1), 7–13.
Blanke, O. et al. (2008). Illusionary Reduplication of the human body
and self. Handbook of Clinical Neurology, 88, 429–458.
Blumenberg, H. (2002). Zu den Sachen und Zurück. Frankfurt a. M.:
Suhrkamp.
Bolognini et al. (2011). Spatial perspective and Coordinate Systems in
autoscopy. A Case Report of a “Fantôme de Profil” in Occipital
Brain Damage. Journal of Cognitive Neuroscience, 23(7), 1741–
1751.
Botvinick, M. & Cohen, J. (1998). Rubber hands “feel” touch that eyes
see. Nature, 391, 756.
Breyer, T. (2011). Attentionalität und Intentionalität. Grundzüge einer
phänomenologisch-kognitionswissenschaftlichen Theorie der Auf-
merksamkeit. München: Fink.
Brugger et al. (2006). Polyopic heautoscopy: Case report and review of
the literature.Cortex, 42, 666–674.
Bruya, B. (ed.) (2010). Effortless Attention. A new perspective in the cog-
nitive science of attention and action. Cambridge MA: MIT Press.
Calvino, I. (1985 [1983]). Mister Palomar. Orlando: Harcourt Books.
Deleuze, G. & Guattari, F. (1992 [1990]). What is philosophy? New
York: Columbia University Press (WP).
Deleuze, G. (1998). What children say. In G. Deleuze, Essays critical

and clinical. London: Verso, 61–67.
Devinsky, O. (1989). Autoscopic Phenomena With Seizures. Arch. Neu-
rol. 46(10), 1080–1088.
Dormashev, Y. (2010). Flow experience explained on the grounds of
an activity approach to attention. In Brian Bruya (ed.), Effortless
Attention. A new perspective in the cognitive science of attention
and action. Cambridge MA: MIT Press, 287–333.
Ehrsson, H. (2007). The Experimental Induction of Out-of-Body Expe-
riences. Science, 317, 1048.
Ehrsson, H. (2012). The concept of body ownership and its relation to
multisensory integration. In Barry E. Stein (ed.), The New Hand-
book of Multisensory Processes. Cambridge MA: MIT Press,
775–792.
Engler, J. & Fulton, P. (2012). Self and No-Self in Psychotherapy. In
Christopher Germer & Ronald Siegel (eds.), Wisdom and Compas-
sion in Psychotherapy. New York: Guilford Press, 176–188.
Fernandez-Duque, D. & Johnson, M. (1999). Attention metaphors: How
metaphors guide the cognitive psychology of attention. Cognitive
Science, 23, 83–116.
Fernandez-Duque, D. & Johnson, M. (2002). Cause and effect theories
of attention: The role of conceptual metaphors. Review of General
Psychology, 6, 153–165.
Fiebrich, A. & Gallagher, S. (2012). Joint Attention in Joint Action. Phil-
osophical Psychology, 1–17, DOI:10.1080/09515089.2012.690176.
Fuchs, T. (20124). Das Gehirn- ein Beziehungsorgan. Eine phänome-
nologisch-ökologische Konzeption. 4. Akt. u. erw. Aufl. Stuttgart:
Kohlhammer.
Fuchs, T. (2012b). Selbst und Schizophrenie. DZPhil, 60(6), 887–901.
Gerner, A. (forthcoming). Philosophical Investigations of Attention.
PhD Thesis in History and Philosophy of Science at the University
of Lisbon (2012).
Gerner, A. (2010a). Notes on Diagrams and Maps. In Olga Pombo, Al-
exander Gerner (eds.), Studies in Diagrammatology and Diagram
Praxis. (=Logic and Cognitive Systems, Studies in Logic 24). Lon-
don, College Publications.
Gerner, A. (2010b). Critical Notes on Two Forms of Reductionism:

Some Shortcomings of Both Conceptual and Strong Neuroaethetic
Accounts of Aesthetics”. In Olga Pombo, Silvia Di Marco, Marco
Pina (Org.), Neuroaethetics. Can Science Explain Art? Lisboa: Fim
de Século, 139–163.
Head, H. (1923). Speech and cerebral localization. Brain, 46, 355–528.
Lazar, S. et al. (2005). Meditation experience is associated with in-
creased cortical thickness. Neuroreport 28, 16(17), 1893–7.
Lutz, A. et al. (2008). Attention Regulation and monitoring in medita-
tion. Trends Cogn Sci 2008, 12(4), 163–9.
Manna, A. et al. (2010). Neural correlates of focussed attention and cog-
nitive monitoring in meditation. Brain Res Bull, 82(1, 2), 46–58.
MacLean, K. et al (2010). Intensive Meditation Training Improves Per-
ceptual Discrimination and Sustained Attention. Psychological
Science, 21(6), 829–838.
Mackworth, N. H. (1948). The breakdown of vigilance during pro-
longed visual search. Quarterly Journal of Experimental Psychol-
ogy 1, 6–21.
Metzinger, T. (2004). Being No-one. The Self-Model Theory of Subjec-
tivity. Cambridge MA: MIT Press.
Metzinger, T. (2005). Out-of-Body Experiences as the Origin of the
Concept of a “Soul”. Mind & Matter, 3(1), 57–84.
Metzinger, T. (2009). The Ego Tunnel. The science of the mind and the
myth of the self. New York: Basic Books.
Metzinger, T. (2010). The Self-Model Theory of Subjectivity: A brief
Summary with Examples. Humana.Mente Journal of Philosoph-
ical Studies, Vol. 14, 25–52. retrieved online 17.12.2010: <http://
www.humanamente.eu/PDF/Issue14_Paper_Metzinger.pdf>.
Neisser, U. (1988). Five kinds of Self-Knowledge. Philosophical Psy-
chology, 1, 35–59.
Olendzki, A. (2010). Unlimiting Mind: The radically experiential psy-
chology of Buddhism. Boston: Wisdom.
Ortega y Gasset, J. (2007[1943]). Meditation on Hunting. Belgrade,
CA: Wilderness Adventures Press.
Parasuraman, R. (1976). Consistency of individual differences in hu-
man vigilance performance: An abilities classification analysis.
Journal of Applied Psychology, 61, 486–492.
Parasuraman, R. & Davies, D. R. (1977). A taxonomic analysis of vigi-

lance. In R. R. Mackie (ed.), Vigilance: Theory, operational perfor-
mance, and physiological correlates. New York: Plenum, 559–574.
Parasuraman, R. (1987). Human computer monitoring. Human Factors,
29, 695–706.
Parasuraman, R. (1984a). The psychobiology of sustained attention.
In J. S. Warm (ed.), Sustained attention in human performance.
Chichester, UK: Wiley, 61–101.
Parasuraman, R. (1984b). Sustained attention in detection and discrim-
ination. In R. Parasuraman, R. & Davies, D. (eds.), Varieties of at-
tention. Orlando, FL: Academic Press, 243–271.
Parasuraman, R., Warm, J. S. & See, J. E. (1998). Brain systems of vig-
ilance. In R. Parasuraman (ed.), The attentive brain. Cambridge,
MA: MIT Press, 221–256.
Parasuraman, R. & Wilson, G. (2008). Putting the brain to work: Neuro-
ergonomics past, present, and future. Human Factors, 50, 468–474.
Peirce, C. S. (1994). “Application to the Carnegie Institution, July 15,
1902”. Manuscript L75, Analytical reconstruction and editorial
work by Joseph Ransdell, Indiana University.
Pöppel, E. (2006). Der Rahmen. Ein Blick des Gehirns auf unser Ich.
Stuttgart: DTV.
Strawson, G. (2004). Against Narrativity. Ratio (new series) XVII, 428–452.
Tomasello, M. & Carpenter, M. (2007). Shared intentionality. Develop-
mental Science, 10(1), 121–125.
Valéry, P. (1989[1973]). Monsieur Teste. The Collected Works of Paul
Valéry, ed. by Jackson Mathews (=Bollingen Series XLV, 6).
Princeton: Princeton University Press
Valéry, P. (1946). Monsieur Teste. Paris: Gallimard.
Waldenfels, B. (2006). Grundmotive einer Phänomenologie des Frem-
den. Frankfurt a. M.: Edition Suhrkamp.
Waldenfels, B. (2004). Phänomenologie der Aufmerksamkeit. Frank-
furt a. M.: Suhrkamp.
Waldenfels, B. (2002). Bruchlinien der Erfahrung. Phänomenologie,
Psychoanalyse, Phenomenotechnik. Frankfurt a. M.: Suhrkamp.
Warm, J. S. et al. (2008). Vigilance requires hard mental work and is
stressfull. Human factors, 50, 3, 433–441.
Notes on Contributors
Eric T. Olson – Professor of Philosophy at University of Sheffield.

Rui Vieira Da Cunha – Member of MLAG (Mind, Language and Ac-
tion Group) and Research Fellow at the Institute for Philosophy, Uni-
versity of Porto.
Klaus Gärtner – Doctoral Research Fellow of Institute for Philosophy
of Language, New University of Lisbon.
Clara Morando – Doctoral Research Fellow at the Institute for Phi-
losophy, University of Porto.
Dina Mendonça – Post-Doctoral Research Fellow of Institute for Phi-
losophy of Language, New University of Lisbon.
António Marques – Full Professor of Philosophy of Knowledge and
Communication at the New University of Lisbon.
Erich Rast – Post-Doctoral Research Fellow of Institute for Philoso-
phy of Language, New University of Lisbon.
Vasco Correia – Post-Doctoral Research Fellow of Institute for Phi-
João Fonseca – Post-Doctoral Research Fellow of Institute for Philos-
ophy of Language, New University of Lisbon.
Jorge Gonçalves – Post-Doctoral Research Fellow of Institute for Phi-
Robert Clowes – Post-Doctoral Research Fellow at the University of
Sussex.
Alexander Gerner – Post-Doctoral Research Fellow of the Center of
Philosophy of Science, University of Lisbon.

(Lisbon Philosophical Studies - Uses of Languages in Interdisciplinary Fields) Joao Fonseca, Jorge Goncalves (Eds.) - Philosophical Perspectives On The Self-Peter Lang AG (2015)

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

(Lisbon Philosophical Studies - Uses of Languages in Interdisciplinary Fields) Joao Fonseca, Jorge Goncalves (Eds.) - Philosophical Perspectives On The Self-Peter Lang AG (2015)

Uploaded by

Copyright:

Available Formats

For the last decade the topic of the Self has been under intense scrutiny from

researchers of various areas spanning from philosophy, neurosciences, and psy-

João Fonseca is a Post-Doc researcher at New University of Lisbon. His

A P ub lic ation from the Ins t it ut e o f Philo s o phy o f L a ngu ag e at t h e N e w U n i v e r s i t y o f Li s b o n

Library of Congress Cataloging-in-Publication Data

Financial Support: Project Cognitive Foundations of the Self PTDC/FIL-FCI/110978/2009

ISSN 1663-7674 pb. ISSN 2235-641X eBook

This publication has been peer reviewed.

© Peter Lang AG, International Academic Publishers, Bern 2015

All rights reserved.

João Fonseca, Jorge Gonçalves

Part I. Metaphysics and Personal Identity

Rui Vieira da Cunha

Part II. Epistemology and Phenomenology

Part III. Cognition, Psychology, Neuroscience

Part IV. Ontology and Taxonomy

Lisbon, November 2013

I. The Background

tradition approaches the problem) philosophers and cognitive scientists

[T]his disparity, which is both problematic and productive, is directly related to

Therefore, this new transdisplinary approach comes with a cost: a taxo-

between, on one hand philosophical methodologies proper (Phenomenol-

II. The Essays

Part I. Metaphysics and Personal Identity

Eric Olson explores his original proposal according to which per-

this is equally problematic for many views besides animalism, it has

Part II. Epistemology and Phenomenology

In ‘How Consciousness explains the Self’, Klaus Gärtner asks for an

not memory dependent and therefore s-ki expresses itself in statements

Part III. Cognition, Psychology, Neuroscience

In ‘De Se Attitudes and Semiotic Aspects of Cognition’ Erich Rast

Proposals’ João Fonseca addresses the problem of conceptual frag-

Part IV. Ontology and Taxonomy

The concepts of consciousness and self have been central in contem-

a constitutive plane of philosophy in which they create concepts on

Clowes, R. W., & Chrisley, R. (2012). Virtualist Representation. Inter-

Hofstadter, D. (2007). I Am a Strange Loop. New York: Basic Books.

Animalism clashes with the conviction that we should go with our

presented with the thought experiment have an almost irresistible

Is this a strong objection? Well, here is a way of defending animalism

if in fact he would not be you because (unbeknownst to us) you are an

2 “Fission” is another imaginary case where what is actually conclusive evidence

Even if this is a good defence of animalism, however, the brain-transplant

indistinguishable from it, and animalism could be defended against it

If animalism implies that the transplant operation would bring a rem-

him with sustaining tissues: the destruction principle. Suppose we im-

objection as well. Both objections would be based on a false assumption

5 Olson (2007, pp. 122–125) is an exception.

However grave the remnant-brain problem may be, it is no reason to

But this only relocates the problem. Constitutionalists do not think

So the remnant-brain problem is not a worry for animalism in particular,

According to brain eliminativism, the conclusion of this reasoning

If this is possible, then I ought to wonder whether I exist. I feel pretty

it would be like to be in the states that particles arranged cerebrally in a

would be indistinguishable from those of human beings. There could

Chisholm, R. (1976). Person and Object. La Salle, IL: Open Court.

Rui Vieira da Cunha*

2. A Cyborg2 Thought Experiment

Tom’s friends are worried because they believe Olson’s animalist

3. Olson’s Animalist View

Eric Olson’s defence of the animalist view on matters of personal identi-

Thus, a substance concept 1) tells us what the object is, as opposed to

Furthermore, an object can undergo a change during its existence such

– I: Introspection is an inescapable tool of S-K and is an instance