Content uploaded by Shaun Gallagher
Author content
All content in this area was uploaded by Shaun Gallagher on Dec 03, 2017
Content may be subject to copyright.
Ever since William James1categorized different senses of
the self at the end of the 19th century, philosophers and
psychologists have refined and expanded the possible vari-
ations of this concept. James’ inventory of physical self, mental
self, spiritual self, and the ego has been variously supple-
mented. Neisser, for example, suggested important distinctions
between ecological, interpersonal, extended, private and
conceptual aspects of self2. More recently, when reviewing a
contentious collection of essays from various disciplines,
Strawson found an overabundance of delineations between
cognitive, embodied, fictional and narrative selves, among
others3. It would be impossible to review all of these diverse
notions of self in this short review. Instead, I have focused
on several recently developed approaches that promise the
best exchange of ideas between philosophy of mind and the
other cognitive sciences and that convey the breadth of
philosophical analysis on this topic. These approaches can
be divided into two groups that are focused, respectively, on
two important aspects of self – the ‘minimal’ self and the
‘narrative’ self (see Glossary).
Philosophical conceptions
of the self: implications
for cognitive science
Shaun Gallagher
Several recently developed philosophical approaches to the self promise to enhance
the exchange of ideas between the philosophy of the mind and the other cognitive
sciences. This review examines two important concepts of self: the ‘minimal self’, a self
devoid of temporal extension, and the ‘narrative self’, which involves personal identity
and continuity across time. The notion of a minimal self is first clarified by drawing a
distinction between the sense of self-agency and the sense of self-ownership for actions.
This distinction is then explored within the neurological domain with specific reference
to schizophrenia, in which the sense of self-agency may be disrupted. The convergence
between the philosophical debate and empirical study is extended in a discussion of
more primitive aspects of self and how these relate to neonatal experience and robotics.
The second concept of self, the narrative self, is discussed in the light of Gazzaniga’s
left-hemisphere ‘interpreter’ and episodic memory. Extensions of the idea of a narrative
self that are consistent with neurological models are then considered. The review
illustrates how the philosophical approach can inform cognitive science and suggests
that a two-way collaboration may lead to a more fully developed account of the self.
91 Traxler, M.J. and Pickering, M.J. (1996) Plausibility and the processing
of unbounded dependencies: an eyetracking study. J. Mem. Lang. 35,
454–475
92 Murray, W.S. and Liversedge, S.P. (1994) Referential context effects on
syntactic processing. In Perspectives on Sentence Processing (Clifton, J.
et al., eds), pp. 359–388, Erlbaum
93 Altmann, G.T.M. et al. (1992) Avoiding the garden path: eye movements
in context. J. Mem. Lang. 31, 685–712
94 Britt, M.A. (1994) The interaction of referential ambiguity and argument
structure in the parsing of prepositional phrases. J. Mem. Lang. 33,
251–283
95 Liversedge, S.P. et al. (1998) Processing arguments and adjuncts in
isolation and context: the case of by-phrase ambiguities in passives.
J. Exp. Psychol. 24, 461–475
96 Spivey-Knowlton, M.J. et al. (1993) Context effects in syntactic ambiguity
resolution. Can. J. Psychol. 47, 276–309
97 Spivey, M.J. and Tanenhaus, M.K. (1998) Syntactic ambiguity resolution
in discourse:modeling the effects of referential context and lexical
frequency. J. Exp. Psychol. 24, 1521–1543
98 Trueswell, J.C. et al. (1994) Semantic influences on parsing: use of
thematic role information in syntactic disambiguation. J. Mem. Lang.
33, 285–318
99 MacDonald, M. (1994) Probabilistic constraints and syntactic ambiguity
resolution. Lang. Cognit. Processes 9, 157–201
100 Spivey-Knowlton, M.K. and Sedivy, J. (1995) Resolving attachment
ambiguities with multiple constraints. Cognition 55, 227–267
101 Traxler, M.J. et al. (1998) Adjunct attachment is not a form of lexical
ambiguity resolution. J. Mem. Lang. 39, 558–592
102 Van Gompel, R.P.G. et al. Unrestricted race: a new model of syntactic
ambiguity resolution. In Reading as a Perceptual Process (Kennedy,
A. et al., eds), Elsevier (in press)
103 Garrod, S. et al. (1994) The role of different types of anaphor in the
on-line resolution of sentences in a discourse. J. Mem. Lang. 33, 39–68
104 Duffy, S.A. and Rayner, K. (1990) Eye movements and anaphor
resolution: effects of antecedent typicality and distance. Lang.
Speech 33, 103–119
Liversedge and Findlay – Eye movements and cognition
14
Review
1364-6613/00/$ – see front matter © 2000 Elsevier Science Ltd. All rights reserved. PII: S1364-6613(99)01417-5
Trends in Cognitive Sciences – Vol. 4, No. 1, January 2000
S. Gallagher is at the
Department of
Philosophy and
Cognitive Science,
Canisius College,
Buffalo, NY 14208,
USA.
tel: +1 716 888 2329
fax: +1 716 888 3122
e-mail: gallaghr@
canisius.edu
Gallagher – Conceptions of the self
The first approach involves various attempts to account
for a ‘minimal’ sense of self. Even if all of the unessential
features of self are stripped away, we still have an intuition
that there is a basic, immediate, or primitive ‘something’ that
we are willing to call a self. This approach leaves aside ques-
tions about the degree to which the self is extended beyond
the short-term or ‘specious present’ to include past thoughts
and actions. Although continuity of identity over time is a
major issue in the philosophical definition of personal iden-
tity, the concept of the minimal self is limited to that which
is accessible to immediate self-consciousness. Certain aspects
of the minimal self are relevant to current models in robotics.
Furthermore, aspects of the minimal self that involve senses
of ownership and agency in the context of both motor action
and cognition can be clarified by neurocognitive models of
schizophrenia that suggest the involvement of specific brain
systems (including prefrontal cortex, supplementary motor
area, and cerebellum) in the manifestation of neurological
symptoms in this disorder.
A second approach to the concept of self involves conceiv-
ing of the self in terms of narrative. This notion was imported
into the cognitive sciences by Dennett4, but it might have a
more complex significance than is indicated in Dennett’s
account. The narrative self is extended in time to include
memories of the past and intentions toward the future. It is
what Neisser refers to as the extended self, and what Dennett
calls a ‘nonminimal selfy’ self. Neuropsychological descriptions
of episodic memory and its loss can help to circumscribe the
neural substrates of the narrative self.
Self-reference and misidentification
There are a number of ways to understand the notion of a
minimal sense of self. In this section, I approach the prob-
lem by discussing how we use the first-person pronoun in a
self-referring way that should never permit a mistake. This
kind of self-reference has a feature that some philosophers call
‘immunity to error through misidentification relative to the
first-person pronoun’5. I will refer to this as the immunity
principle (see Glossary). Once this principle is clarified we
can ask whether, in actuality, it can ever fail and if so, what
this might reveal. In the next section, I will explore this possi-
bility in relation to a neurocognitive model of schizophrenia
that requires us to make a distinction between two aspects
of the minimal sense of self: the sense of self-ownership and
the sense of self-agency.
Wittgenstein distinguished between two uses of the first-
person pronoun in self-reference: ‘as subject’ and ‘as object’6.
Use of the first-person pronoun as subject might best be dis-
cerned by understanding what a speaker could be wrong about,
and the kinds of questions that one could sensibly ask them.
For example, if someone says ‘I think it is raining outside’,
she could be wrong about the rain. It might not be raining.
But it seems that she could not be wrong about the ‘I’. That
is, she could not misidentify herself when she states that it is
she who is thinking. So, according to Wittgenstein, the fol-
lowing question would be nonsensical: ‘Are you sure that you
are the one who thinks it is raining?’ Such use of the first-
person pronoun is immune to error through misidentification.
By contrast, when we use the first-person pronoun ‘as object’
it is possible to misidentify ourselves. For example, in some
experimental situations a subject’s arm may be deafferented
(that is, the subject is deprived of normal proprioceptive
feedback about the position of their limb and therefore can-
not keep track of it without vision). Their visual perception
of arm movements are then manipulated through mirrors or
videotape7,8. In such cases, the subject might be led to say, ‘I
am moving my arm to the left’, when in fact the basis for his
judgment is a videotape of someone else moving their own
arm to the left. In that case, the subject makes a mistake about
who is moving their arm to the left. To say ‘I’ in such a case
involves an objective misidentification of oneself.
Shoemaker5suggests that the immunity principle applies
only to the use of ‘I’ as subject because when we use the first-
person pronoun as subject we are not actually attempting to
identify ourselves. In other words, when I self-refer in this
way I do not go through a cognitive process in which I try to
match up first-person experience with some known criterion
in order to judge the experience to be my own. My access to
myself (my self ) in first-person experience is immediate and
non-observational; that is, it doesn’t involve a perceptual or
reflective act of consciousness. In this sense, the immediate
self that is referred to here is the pre-reflective point of origin
for action, experience and thought.
Are there any exceptions to the immunity principle? Is
there any instance of someone using a first-person pronoun
as subject, and being wrong in their reference? Following sug-
gestions made by Feinberg9and Frith10 about certain schizo-
phrenic experiences (including auditory hallucination, thought
insertion, and delusions of control in which subjects report
that their body is under the control of other people or
things), Campbell has proposed that such experiences might
be counterexamples to the immunity principle11. A schizo-
phrenic patient who suffers thought insertion, for example,
might claim that she is not the one who is thinking a particu-
lar thought, when in fact she is the one who is thinking the
thought. The following example of a schizophrenic’s account
of her own thought processes illustrates this: ‘Thoughts are
15
Trends in Cognitive Sciences – Vol. 4, No. 1, January 2000
Review
Immunity principle: When a speaker uses the first-person pronoun (‘I’) to refer to him
or herself, she cannot make a mistake about the person to whom she is referring.
Philosophers call this ‘immunity to error through misidentification relative to the first-per-
son pronoun’5.
Minimal self: Phenomenologically, that is, in terms of how one experiences it, a con-
sciousness of oneself as an immediate subject of experience, unextended in time. The min-
imal self almost certainly depends on brain processes and an ecologically embedded body,
but one does not have to know or be aware of this to have an experience that still counts as
a self-experience.
Narrative self: A more or 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.
Non-conceptual first-person content: The content of a primitive self-consciousness
that is not informed by conceptual thought. For example, the ecological content of
perception that specifies one’s own embodied position in the environment.
Sense of agency: The sense that I am the one who is causing or generating an action.
For example, the sense that I am the one who is causing something to move, or that I am
the one who is generating a certain thought in my stream of consciousness.
Sense of ownership: The sense that I am the one who is undergoing an experience. For
example, the sense that my body is moving regardless of whether the movement is volun-
tary or involuntary.
Glossary
put into my mind like “Kill God”. It’s just like my mind
working, but it isn’t. They come from this chap, Chris.
They’re his thoughts’ (Ref. 10, p. 66). In such cases the
schizophrenic patient misidentifies the source of the thought
and seemingly violates the immunity principle.
It may be argued whether or not Campbell is correct in
his claim that this is a counterexample to the immunity prin-
ciple12, but the implications of his analysis are quite produc-
tive. His argument implies that a scientific explanation of
schizophrenic phenomena such as thought insertion might
also count as a scientific explanation of how the immunity
principle works. Frith’s neurocognitive model of the break-
down of self-monitoring in schizophrenia turns out to be a
good candidate for explaining immunity to error through
misidentification. If we can identify which mechanisms fail
at the cognitive or neurological level when the schizophrenic
patient suffers from thought insertion, then we also have a
good indication of the mechanisms responsible for (or at least
involved in) the normal immunity to error found in self-
reference, and the immediate sense of self. This insight moves
us from the conceptual and often abstract arguments of phi-
losophy to the more empirical inquiries of neuropsychology
and neurophysiology.
A neurocognitive model of immediate self-awareness
A brief consideration of motor action will help to clarify
two closely related aspects of minimal self-awareness: self-
ownership – the sense that it is my body that is moving; and
self-agency – the sense that I am the initiator or source of
the action. In the normal experience of voluntary or willed
action, the sense of agency and the sense of ownership
coincide and are indistinguishable. When I reach for a cup,
I know this to be my action. This coincidence may be what
leads us to think of ownership of action in terms of agency:
that the owner of an action is the person who is, in a specific
way, causally involved in the production of that action, and
is thus the author of the action. In the case of involuntary
action, however, it is quite possible to distinguish between
sense of agency and sense of ownership. I may acknowledge
ownership of a movement – that is, I have a sense that I am
the one who is moving or is being moved – and I can self-
ascribe it as my movement, but I may not have a sense of
causing or controlling the movement. That is to say, I have
no sense of agency. The agent of the movement is the person
who pushed me from behind, for example, or the physician
who is manipulating my arm in a medical examination. Thus,
my claim of ownership (my self-ascription that I am the one
who is undergoing an experience) can be consistent with my
lack of a sense of agency. Phenomena such as delusions of
control, auditory hallucinations, and thought insertion appear
to involve problems with the sense of agency rather than the
sense of ownership13.
There is good evidence to suggest that the sense of own-
ership for motor action can be explained in terms of ecologi-
cal self-awareness built into movement and perception2,14.
By contrast, experimental research on normal subjects suggests
that the sense of agency for action is based on that which
precedes action and translates intention into action15,16. In
addition, research that correlates initial awareness of action
with scalp recordings of the lateralized readiness potential in
motor cortex, and with transcranial magnetic stimulation of
the supplementary motor area, strongly indicates that one’s
initial awareness of a spontaneous voluntary action is tied to
the anticipatory or pre-movement motor commands relating
to relevant effectors17,18.
It turns out that some schizophrenic patients who suffer
from thought insertion also make mistakes about the agency of
various bodily movements. To explain this, Frith10 appeals to
the notions of efference copy and comparator mechanisms that
were originally used to explain motor control19,20. According to
the most recent version of this model, a comparator mecha-
nism operates as 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
corrections21,22. Such a mechanism is consistent with the
findings cited above. This comparator process anticipates
the sensory feedback from movement and underpins an on-
line sense of self-agency that complements the ecological
sense of self-ownership based on actual sensory feedback12
(Fig. 1). If the forward model fails, or efference copy is not
properly generated, sensory feedback may still produce a
sense of ownership (‘I am moving’) but the sense of agency
will be compromised (‘I am not causing the movement’),
even if the actual movement matches the intended
movement23.
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 feedback24,25. The control based on sensory feedback
is thought to involve the cerebellum21. By contrast, problems
with forward monitoring are consistent with studies of schizo-
phrenia that show abnormal pre-movement brain potentials
associated with elements of a neural network involving sup-
plementary motor, premotor and prefrontal cortexes26. Prob-
lems with these mechanisms might therefore result in the
Review Gallagher – Conceptions of the self
16
Trends in Cognitive Sciences – Vol. 4, No. 1, January 2000
Forward model
Intended state
Motor command
Efference copy
*
Sensory (reafferent) feedback
Movement
Actual state
Predicted state
trends in Cognitive Sciences
Fig. 1. The forward and feedback comparators. This model represents processes that
generate two aspects of the ‘minimal self’ in normal experience (see text). Match at the forward
comparator provides a sense of agency for movement; match at the feedback comparator
provides a sense of ownership for movement.
lack of a sense of agency that is characteristic of these kinds
of schizophrenic experience.
Following a suggestion made by Feinberg9, Frith postu-
lates a similar model for cognition – specifically, for thought
and inner speech10. Phenomena such as thought insertion,
hearing voices, or perceiving one’s own acts as alien, suggest
that something is wrong with the self-monitoring mechanism.
Frith’s model assumes not only that thinking, insofar as it is
intended and self-generated, is a kind of action, but also that
thinking has to match the subject’s intention for it to feel
self-generated, as in the case of a motor action. This suggests
that, although such intentions are not always consciously
accessible, comparator processes that match intentions to the
generation of thought and to the stream of thought bestow,
respectively, a sense of agency and a sense of ownership for
thought, in a similar way to motor action. If the mechanism
that constitutes the forward aspect of this monitoring process
fails, a thought occurs in the subject’s own stream of con-
sciousness but does not seem to the subject to be self-generated
or to be under the subject’s control. Rather, it appears to be
an alien or inserted thought (Fig. 2).
Whether or not this kind of model is adequate to account
for the phenomenon of inserted thought, I would suggest that
the approaches taken by Frith and Campbell promise a way to
explain in specific neurological terms the immediacy involved
in the senses of self-ownership and self-agency, and in the im-
munity principle. Such aspects of the minimal self may find a
neurological explanation in the proper workings of the mecha-
nisms described above and are threatened by their failure.
The minimal self: embodied or disembodied?
Taking the immunity principle as a point of departure, there
are two other directions that one could follow. The first
explores the idea that there is an even more primitive and
embodied sense of self than that involved in the use of the
first-person pronoun. This approach pursues the impli-
cations of what developmental psychologists have recently
discovered about the experience of neonates. The second
involves a more abstract self-reflective access to first-person
experience, which, among other things, leads to issues that
concern AI applications in robotics.
Taking the first approach, are there any aspects of the
minimal self that are more primitive than those identified in
the immunity principle? In speaking about self-reference,
we are already speaking of a self that is capable of linguistic
communication – at the very least, the person is capable of
using the first-person pronoun. If one considers that language
and conceptual capacity develop in parallel, this might mean
that the person’s immediate and pre-reflective access to the
self already involves the mediation of a conceptual framework.
Is it possible to speak of a non-conceptual access to the self –
a more primitive self-consciousness that does not depend on
the use of a first-person pronoun?
Bermúdez explores the many implications of this ques-
tion27. Following on from Gibson’s ecological psychology,
part of what Bermúdez calls ‘non-conceptual first-person
content’ consists of the self-specifying information attained
in perceptual experience. When I perceive objects or move-
ment in the external environment, I also gain information
about myself – information that is pre-linguistic and non-
conceptual. This is what Neisser calls the ecological self2.
The fact that non-conceptual, ecological self-awareness exists
from the very beginning of life can be demonstrated by the
important role it plays in neonatal imitation. Neonates less
than an hour old are capable of imitating the facial gestures
of others in a way that rules out reflex or release mechanisms,
and that involves a capacity to learn to match the presented
gesture28,29. For this to be possible the infant must be able to
do three things: (1) distinguish between self and non-self;
(2) locate and use certain parts of its own body propriocep-
tively, without vision; and (3) recognize that the face it sees
is of the same kind as its own face (the infant will not imitate
non-human objects30). One possible interpretation of this find-
ing is that these three capacities present in neonates constitute
a primitive self-consciousness, and that the human infant
is already equipped with a minimal self that is embodied,
enactive and ecologically tuned31–33.
One can, however, move in a second direction and ask
whether it is possible to capture and make explicit the pre-
reflective minimal self in a reflective, and conceptually in-
formed, introspection. In this case, one may still talk about the
most abstract aspect of what we experience to be ourselves,
even if it is mediated through reflection. Galen Strawson’s
recent essays on the self make it clear that he is seeking the
most basic and stripped-down version of a self that can still
be called self3,34,35. He begins with a reflective description of
his experience of the self. This phenomenologically reflective
approach then naturally leads to a characterization of the self
as a subject of experience. Thus, Strawson is led to define the
self as a subject of experience that is a single (hiatus-free)
mental thing. This is a momentary self without long-term
continuity, and thus, without a history – ‘a bare locus of
consciousness, void of personality’ (Ref. 3, p. 492).
On this view, a human being consists of a series of such
transient selves, each one lasting only as long as a unique period
of experience lasts, coming into existence and going out of
Gallagher – Conceptions of the self
17
Trends in Cognitive Sciences – Vol. 4, No. 1, January 2000
Review
Forward model
Intention
Thought generation
Efference copy
No registration
or match
*
Cognitive feedback
Actual stream
of consciousness
Predicted state
trends in Cognitive Sciences
Fig. 2. Neuropsychological model of inserted thought. This model represents the loss of
a sense of agency in some psychiatric illnesses such as schizophrenia. No match at the forward
comparator deprives the system of a sense of agency for thought; match at the feedback
comparator provides a sense of ownership for thought.
existence, without continuity. Despite the ‘local’ character of
Strawson’s approach, that is, an approach that focuses on his
own experience, the self that he seeks to define is not restricted
to the human case. It would be entirely possible for the im-
mediate self he describes to be instantiated in a non-human
animal that has the right cognitive equipment. It has been
argued that it might even be possible to create the minimal
self in a machine, or more precisely, in a robot (see Box 1),
but this would entail dropping the part of Strawson’s proposal
that defines the self as a conscious subject of experience.
Furthermore, on Strawson’s view, it is not essential that this
minimal self be embodied or enactive within an environment.
The self-consciousness that captures this self is not ecologically
embedded, but is one that operates on a conceptual level, al-
ready in possession of the concept ‘self’. Strawson is nonethe-
less a materialist, and considers the self as ‘mental thing’ to be a
physical entity which, in the human case, is likely to be mani-
fest in terms of brain processes. What is metaphysically the
case, however, is not always revealed in the phenomenological
record: I can be conscious of myself as a minimal subject of
experience that is a single mental thing, without being aware
of the embodiment or brain functions that may (or may not)
generate the self. This is entirely consistent with self-reference
that is immune from error through misidentification, as it speci-
fies an access to the self that does not depend on applying
empirical (in this case physical) criteria of identity. Even if it
is the case that the information that constitutes the minimal
self is generated in ecologically embodied experience, and even
if, in practice, a human being is capable of knowing that this
is the case, one does not gain the self-consciousness that goes
along with the minimal self by knowing this to be true or by
being able to employ empirical criteria to verify it.
The self extended and mediated by narrative
So far we have considered only a minimal self, which is a
concept of self that seems quite at odds with our common-
sense conception of who we are. Surely we think and speak
of ourselves as entities extended in time? Indeed, it seems
undeniable that we have memories and that we make plans,
and that there is continuity between our past and our future.
And do we not, as selves with individual identity, encompass
that continuous experience? What is the nature of this sense of
a continuous self? Is it carried by a succession of momentary
minimal selves that are tied together by real connections? Or
Tani explores the possibility of establishing an artificial version of
Strawson’s minimal self in a machine (Ref. a). He takes up the
challenge of developing an objective definition of this concept
in the context of robot design, although he is still forced to use
terms like ‘subjective mind’ and ‘self-consciousness’ in his objec-
tive account (Ref. a, pp. 150, 173). However, in contrast to
Strawson, Tani makes it clear that the robotic self he is designing
is the result of physical interaction between the robotic body and
its environment. Specifically, its short-term existence is generated
only in cases where the interaction fails to go smoothly. What
Tani retains from Strawson’s model is the idea that the self is only
a momentary phenomenon, called into and going out of existence
from one moment to the next, as the situation requires.
Following theoretical suggestions made by Varela et al. (Ref. b),
Tani constructed a robot (Fig. I) that operates by integrating
information derived from bottom-up (sensory–motor) processes
that represent environmental conditions, with top-down, abstract,
predictive modeling of the world. In general, robotic processes
run smoothly as long as there is a good relative match between
the top-down model and the bottom-up input. Problems occur
when there are inconsistencies between these two pathways.
These problems are difficult to resolve in hybrid systems where
top-down mechanisms are designed to follow the logic of symbol
manipulation, and bottom-up processes are based on analogue
pattern-matching. By contrast, Tani’s design bases both pathways
on dynamic systems, so that their interface takes place in a shared
metric space. During unsteady or conflicting phases of the system
dynamics, an arbitration process takes place in that shared space,
and the robot is required to take its own current state into
account. Specifically it needs to take into account (‘become aware
of’) the conflict in its own system, and its own degree of famil-
iarity with the surrounding environment. This, in Tani’s view,
is the robotic equivalent of self-consciousness. A self comes into
existence when the relationship between top-down ‘mental’ pro-
cesses and bottom-up sensory-motor processes becomes incoher-
ent; that is, in the event of a failure to cope with environmental
demands. This self is not an entity that has continuity over the
long-term, but is a short-term phenomenon (consistent with
Strawson’s view), and in this case one that emerges only on
occasions that motivate self-reference.
References
aTani, J. (1998) An interpretation of the ‘self’ from the dynamical
systems perspective: a constructivist approach. J. Conscious. Stud.
5, 516–542
bVarela, F., Thompson, E. and Rosch, E. (1991) The Embodied Mind,
MIT Press
Box 1. Robotics and the minimal self
Fig. I. Tani’s vision-based robot. (Reproduced, with permission,
from Ref. a.)
Review Gallagher – Conceptions of the self
18
Trends in Cognitive Sciences – Vol. 4, No. 1, January 2000
are momentary minimal selves simply abstractions from a more
substantial continuity that is the more genuine self? The
philosophical traditions are replete with a variety of answers
to these questions.
One famous answer given by Hume suggests that the self
consists of a bundle of momentary impressions that are strung
together by the imagination36. In effect, an extended self is
simply a fiction, albeit a useful one because it lends a practical
sense of continuity to life, but a fiction nonetheless. The nar-
rative theory of self is a contemporary reading of this view.
Dennett offers one version of this theory which he sees as
consistent with recent developments in our understanding of
how the brain functions4,37. The consensus from contemporary
neuroscience is that neurological processing is for the most
part distributed across various brain regions, and it cannot
be said that there is a real, neurological center of experience.
Thus, there is no real simplicity of experience at one time nor
real identity across time that we could label the self. At best,
we might refer to a minimal biological self as something real.
But the latter is nothing more than a principle of organization
involving the distinction between self and non-self. Further-
more, this principle is found throughout living nature, and
is not something sufficient for the purpose of a coherent con-
tinuity or identity over time, such as is found at the level of
human experience. Humans, however, do have something
more than this – we have language. And with language we
begin to make our experience relatively coherent over extended
time periods. We use words to tell stories, and in these stories
we create what we call our selves. We extend our biological
boundaries to encompass a life of meaningful experience.
Two things are to be noted from Dennett’s account. First,
we cannot prevent ourselves from ‘inventing’ our selves. We
are hardwired to become language users, and once we are
caught up in the web of language and begin spinning our own
stories, we are not totally in control of the product. As Dennett
puts it, ‘for the most part we don’t spin them [the stories]; they
spin us’ (Ref. 4, p. 418). Second, an important product of this
spinning is the narrative self. The narrative self, however, is
nothing substantially real. Rather, it is an empty abstraction.
Specifically, Dennett defines a self as an abstract ‘center of
narrative gravity,’ and likens it to the theoretical fiction of the
center of gravity of any physical object. In the case of narrative
gravity, however, an individual self consists of the abstract
and movable point where the various stories (of fiction or
biography) that the individual tells about himself, or are told
about him, meet up (Fig. 3a).
The notion of a narrative self-constitution finds con-
firmation in psychology and neuroscience. In the former,
Neisser’s concepts of the extended and the conceptual self,
initially explained in terms of memory, have been enhanced
by considerations of the role that language and narrative
play in developing our own self-concept38. In the realm of
neuroscience, Gazzaniga has suggested that one function of
the left hemisphere of the brain is to generate narratives, using
what he terms an ‘interpreter’. Gazzaniga proposed this inter-
preting function based on his studies of split-brain patients.
In these patients, the left hemisphere has no internal access
to right-hemisphere experience because the corpus callosum
has been severed. Nonetheless, in properly designed experi-
mental circumstances, the left hemisphere devises interpre-
tations for meanings, actions and emotions produced by the
right hemisphere. Such interpretations show consistency with
the experiential context belonging to the left hemisphere
rather than with the original right-hemisphere context. The
left hemisphere, for example, might remain ignorant of the
content or cause of an emotion generated in the right hemi-
sphere, but the left-hemisphere experience of that emotion
motivates an interpretation of the event in terms relevant to
the content available to the left hemisphere. In Gazzaniga’s
model, the interpreter weaves together autobiographical fact
and inventive fiction to produce a personal narrative that
enables the sense of a continuous self39,40. Gazzaniga, however,
contends that the self, in this regard, is not a fiction because
the normal functioning of the interpreter tries to make sense
of what actually happens to the person. At most, in the non-
pathological case, it may be only ‘a bit fictional’ (Ref. 41,
Gallagher – Conceptions of the self
19
Trends in Cognitive Sciences – Vol. 4, No. 1, January 2000
Review
trends in Cognitive Sciences
Narrative
of being a
spouse Narrative
of being a
spouse
Stories I
tell about
myself
Stories I
tell about
myself
SELF SELF
Narrative
of being a
parent
Narrative
of being a
parent
Narrative
of being a
sports fanatic
Narrative
of being a
sports fanatic
What others
say about me What others
say about me
Narrative
of being a
teacher
Narrative
of being a
teacher
(a) (b)
Fig. 3. Two models of the narrative self. (a) Center of narrative gravity. In this model (formulated by Dennett4), the self is defined
as an abstract ‘center of narrative gravity’, where the various stories told about the person, by himself and others, meet. (b) An extended
and more distributed model of the narrative self, less unified than the one proposed by Ricoeur, but importantly distinct from Dennett’s
model in that the self is not an abstract ‘center’, but rather, an extended self which is decentered and distributed.
p. 713). Perhaps we cannot help but enhance our personal
narratives with elements that smooth over discontinuities
and discrepancies in our self-constitution.
A necessary condition for the non-fictional aspects of a
narrative self is the proper working of episodic memory.
Pribram suggests that this depends on a fronto-limbic system
that includes the anterior poles of the frontal and temporal
lobes, and elements of the limbic formation42. Specifically, this
system is involved in providing a sense of time. The impor-
tance of the proper functioning of episodic memory and time-
sense on the formation of the narrative self is indicated by the
case of a young boy diagnosed with congenital damage to the
right hemisphere and frontal cortex. He suffers from a pro-
found episodic amnesia and because he lacks the ability to
quantify the passage of time or to appreciate the meaning of
temporal units42,43, he is unable to formulate certain essential
structures of narrative, namely, sequential structure and the
demarcations of beginning and end.
Further extensions of the narrative self
In the current context of contentious disagreements on a large
range of issues surrounding the self3, a general consensus
among a diverse group of cognitive scientists concerning the
constitution of the narrative self might seem surprising. How-
ever, it is perhaps even more surprising that there is some
consensus within philosophy on this point, even across the
great divide between continental and analytical philosophers.
What Dennett, Neisser, Gazzaniga and Pribram have to say
about the narrative self echoes in some respects an earlier dis-
cussion among continental philosophers. Perhaps Ricoeur is
the best representative of this earlier discussion concerning the
nature of narrative and the making of the narrative self44,45.
Ricoeur carefully explored these issues in depth, and reached
conclusions that are not inconsistent with the view outlined
in the cognitive science discussion above. In contrast to
Dennett, however, Ricoeur conceives of the narrative self, not
as an abstract point at the intersection of various narratives, but
as something richer, more substantial and concrete. Ricoeur
insists that, importantly, one’s own self narrative is always
entangled in the narratives of others.
We may extend Ricoeur’s model beyond what he takes
to be a unified life narrative and suggest that the self is the
sum total of its narratives, and includes within itself all of the
equivocations, contradictions, struggles and hidden messages
that find expression in personal life. In contrast to Dennett’s
center of narrative gravity, this extended self is decentered,
distributed and multiplex (Fig. 3b). At a psychological level,
this view allows for conflict, moral indecision and self-
deception, in a way that would be difficult to express in terms
of an abstract point of intersection. Furthermore, with respect
to neurological models, this extended model is even more
consistent than Dennett’s abstract center with the concept of
distributed processing, and with what Gazzaniga describes
as the mixing of fact and fiction by the left-hemisphere ‘inter-
preter’. By extending the ideas of a narrative self, we are perhaps
coming closer to a concept of the self that can account for
the findings of the cognitive sciences and neurosciences, as
well as our own experience of what it is to be a continuous,
phenomenological self.
Concluding remarks
In a recent book, Damasio has insightfully captured the dif-
ficulty involved in expressing the interrelations between the
minimal (‘core’) self and the narrative (‘autobiographical’)
self46. The difficulty is due to complexities that are apparent
on both the personal and the sub-personal, neurological levels.
Episodic memory, which is necessary for the construction of
the narrative self, is subject to constant remodeling under
the influence of factors that include innate and acquired dis-
positions as well as social and cultural environments. The
registration of episodic memory as ‘my’ memory of ‘myself’
clearly depends on a minimal but consistently reiterated sense
of self that I recognize, without error, as myself. In some re-
spects, as Damasio insists, this depends on narrowly defined,
embodied capabilities and feelings. In other regards, however,
the core features of the self are constantly being reinterpreted
by the narrative process. In the neurological terms that
Damasio uses this means that there are extremely complex
demands made on the processes that link early sensory cor-
texes that hold information on the minimal or core self, and
convergence or dispositional zones that contribute to the
generation of the narrative self. In this regard, he makes it
clear that at present the neuroscientist, like the philosopher,
can offer, at best, informed speculation on these processes.
In this review, I have tried to show that philosophical ideas
about the self can be aligned with, and can inform, current
ideas in cognitive science. I also believe that philosophers can
learn about the nature of the self from psychologists, neuro-
scientists and other cognitive scientists. Thus, collaborative
efforts between philosophers and scientists promise to open
up more subtle and sophisticated avenues of research, which
will define more fully the concept of the self.
Acknowledgements
I thank S-J. Blakemore, U. Neisser, G. Strawson, and anonymous reviewers
for helpful comments on an earlier version of this paper. Parts of my research
were supported by a fellowship at the NEH Summer Institute on Mind, Self,
and Psychopathology, directed by J. Whiting and L. Sass at Cornell University
in 1998, and by a sabbatical leave from Canisius College in 1999.
Review Gallagher – Conceptions of the self
20
Trends in Cognitive Sciences – Vol. 4, No. 1, January 2000
Outstanding questions
• What relationship exists between the minimal self and the narrative self?
Is one generated from the other? Do they operate independently of
each other?
• Shoemaker5has maintained that immunity to error does not necessarily
extend to episodic memory. What status do truth-claims concerning
episodic memory have?
• Because the sense of self-agency is absent both in the case of involuntary
thoughts and in cases of schizophrenic experiences of inserted thoughts,
the lack of a sense of self-agency by itself cannot fully explain the
schizophrenic patient’s sense that certain of his thoughts belong to
someone else. What else is required to explain this?
• If some aspect of the minimal self depends on a forward model of control,
this complicates Tani’s design for a self-referring robot. To what extent
can feedback mechanisms by themselves provide the requisite minimal
reference to self?
• What are the precise neurological mechanisms involved in the left-
hemisphere interpreter?
• In narrative theory, self-constitution is meant to imply a situation in which
the self is both the narrator and the narrated. If there is more than one
narrator (if we are also constituted by stories about ourselves told by
others), what mechanisms at the psychological level integrate or adjudicate
these diverse constitutions?
References
1James, W. (1890) The Principles of Psychology, Dover Publ. (reprinted
1950)
2Neisser, U. (1988) Five kinds of self-knowledge. Philos. Psychol. 1, 35–59
3Strawson, G. (1999) The self and the SESMET. In Models of the Self
(Gallagher, S. and Shear, J., eds), pp. 483–518, Imprint Academic
4Dennett, D. (1991) Consciousness Explained, Little Brown & Co.
5Shoemaker, S. (1984). Identity, Cause, and Mind, Cambridge University
Press
6Wittgenstein, L. (1958) The Blue and Brown Books, Blackwell
7Jeannerod, M. (1994) The representing brain: neural correlates of
motor intention and imagery. Behav. Brain Sci. 17, 187–245.
8Daprati, E. et al. (1997) Looking for the agent: an investigation into
consciousness of action and self-consciousness in schizophrenic patients.
Cognition 65, 71–86
9Feinberg, I. (1978) Efference copy and corollary discharge: Implications
for thinking and its disorders. Schizophr. Bull. 4, 636–640
10 Frith, C. (1992) The Cognitive Neuropsychology of Schizophrenia, Erlbaum
11 Campbell, J. Schizophrenia, the space of reasons and thinking as a
motor process. The Monist (in press)
12 Gallagher, S. Self-reference and schizophrenia: a cognitive model of
immunity to error through misidentification. In Problems of the Self
(Zahavi, D. and Parnas, J. eds), John Benjamins (in press)
13 Stephens, G.L. and Graham, G. (1994) Self-consciousness, mental agency,
and the clinical psychopathology of thought insertion. Philos. Psychiatry
Psychol. 1, 1–12
14 Gallagher, S. and Marcel, A.J. (1999) The self in contextualized action.
J. Conscious. Stud. 6, 4–30
15 Marcel, A.J. The sense of agency: awareness and ownership of actions
and intentions. In Agency and Self-Awareness (Roessler, J. and Eilan, N.
eds), Oxford University Press, (in press)
16 Fourneret, P. and Jeannerod, M. (1998) Limited conscious monitoring
of motor performance in normal subjects. Neuropsychologia 36,
1133–1140
17 Haggard, P. and Eimer, M. (1999) On the relation between brain
potentials and the awareness of voluntary movements. Exp. Brain Res.
126, 128–133
18 Haggard, P. and Magno, E. (1999) Localising awareness of action with
transcranial magnetic stimulation. Exp. Brain Res. 127, 102–107
19 Sperry, R.W. (1950) Neural basis of the spontaneous optokinetic response
produced by visual inversion. J. Comp. Physiol. Psychol. 43, 482–489
20 Holst, E. von and Mittelstaedt, H. (1950). Das Reafferenzprinzip
(Wechselwirkungen zwischen Zentralnervensystem und Peripherie).
Naturwissenschaften 37, 464–476
21 Frith, C.D. et al. Abnormalities in the perception and control of action.
Proc. R. Soc. Lond. Ser. B (in press)
Gallagher – Conceptions of the self
21
Trends in Cognitive Sciences – Vol. 4, No. 1, January 2000
Review
22 Georgieff, N. and Jeannerod, M. (1998) Beyond consciousness of external
events: a ‘who’ system for consciousness of action and self-consciousness.
Conscious. Cognit. 7, 465–477
23 Spence, S.A. et al. (1997) A PET study of voluntary movement in
schizophrenic patients experiencing passivity phenomena (delusions
of alien control). Brain 120, 1997–2011
24 Frith, C.D. and Done, D.J. (1988) Towards a neuropsychology of
schizophrenia. Br. J. Psychiatry 153, 437–443
25 Malenka, R.C. et al. (1982) Impaired central error correcting behaviour
in schizophrenia. Arch. Gen. Psychiatry 39, 101–107
26 Singh, J.R. et al. (1992) Abnormal premovement brain potentials in
schizophrenia. Schizophr. Res. 8, 31–41
27 Bermúdez, J. (1998) The Paradox of Self-Consciousness, MIT Press
28 Meltzoff, A. and Moore, M.K. (1977) Imitation of facial and manual
gestures by human neonates. Science 198, 75–78
29 Meltzoff, A. and Moore M.K. (1983) Newborn infants imitate adult
facial gestures. Child Dev. 54, 702–709
30 Legerstee, M. (1991) The role of person and object in eliciting early
imitation. J. Exp. Child Psychol. 51, 423–433
31 Bermúdez, J. (1996) The moral significance of birth. Ethics 106, 378–403
32 Gallagher, S. (1996) The moral significance of primitive self-
consciousness. Ethics 107, 129–140
33 Rochat, P., ed. (1995) The Self in Infancy: Theory and Research, Elsevier
34 Strawson, G. (1997) The self. J. Conscious. Stud. 4, 405–428
35 Strawson, G. (1999) Self, body, and experience. Proceedings of the
Aristotelian Society (Supplement) 73, 307–332
36 Hume, D. (1739) A Treatise of Human Nature, Clarendon Press (reprinted
1975)
37 Dennett, D. (1988) Why everyone is a novelist. Times Literary
Supplement Sept.16–22, pp. 1016, 1028–1029
38 Neisser, U. and Fivush, R. (1994) The Remembering Self : Construction and
Accuracy in the Self-Narrative, Cambridge University Press
39 Gazzaniga, M. (1998) The Mind’s Past, Basic Books
40 Gazzaniga, M. (1995) Consciousness and the cerebral hemispheres. In The
Cognitive Neurosciences (Gazzaniga, M., ed), pp. 1391–1400, MIT Press
41 Gazzaniga, M. and Gallagher, S. (1998) The neuronal Platonist.
J. Conscious. Stud. 5, 706–717
42 Pribram, K.H. (1999) Brain and the composition of conscious experience.
J. Conscious. Stud. 6 , 19–42
43 Ahern, C.A. et al. (1998) Preserved semantic memory in an amnesic
child. In Brain and Values: Is a Biological Science of Values Possible?
(Pribram, K.H., ed.), pp. 277–297, Erlbaum
44 Ricoeur, P. (1984) Time and Narrative (3 Vols), University of Chicago Press
45 Ricoeur, P. (1992) Oneself As Another, University of Chicago Press (transl.
from Soi-même comme un autre, Editions du Seuil, 1990)
46 Damasio, A. (1999) The Feeling of What Happens: Body and Emotion in
the Making of Consciousness, Harcout Brace
Trends in Cognitive Sciences online – tics.trends.com
The online version of Trends in Cognitive Sciences contains a simple search and browse facility that will enable
you to view full-text versions of all articles published in the journal since January 1998. Browsing and searching
the articles is straightforward and designed around the need for simple navigation. Articles can be viewed in
three display formats: text-only, text with pop-up figures, or as PDF files for high-quality printing.
Personal print subscribers benefit from online access to Trends in Cognitive Sciences online by accessing the
site using a unique subscription key. For information about how to obtain your key please visit http://www.
trends.com The subscription key is important and should be kept safe as it is your unique access code to your on-
line journal for the remainder of the year. Institutional access is curently available through Elsevier’s
ScienceDirect (http://www.sciencedirect.com) services.
Subscribers and non-subscribers alike will continue to have free access to full content lists, abstracts, search fa-
cilities and the e-mail contents alerting service. It is also possible to download full-text articles on a pay-per-view
basis.
The following recent articles are just some of those available in the fully searchable archive:
• Thornhill, R. and Gangestad, S.G. (1999) Facial attractiveness Trends Cognit. Sci. 3, 452–460
• Perner, J. and Lang, B. (1999) Development of theory of mind and executive control Trends Cognit. Sci. 3,
337–344
• French, R.M. (1999) Catastrophic forgetting in connectionist networks Trends Cognit. Sci. 3, 128–135
