Show Summary Details

Page of

PRINTED FROM OXFORD HANDBOOKS ONLINE ( © Oxford University Press, 2018. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a title in Oxford Handbooks Online for personal use (for details see Privacy Policy and Legal Notice).

date: 01 December 2021

Introduction: A Diversity of Selves

Abstract and Keywords

This introductory article explains the coverage of this book, which is about developments in research on the topic of the self. This volume introduces the complexity of the concept (or plurality of concepts) of self and to the many different approaches to its (their) analysis. It presents analysis from various fields including analytic philosophy of mind, neuroscience, and postmodernism. Part of the book explores how certain aspects of self are constituted in brain processes, narratives, or actions while other parts investigate how some aspects of the self coming apart in anomalous experiences, experiments, or pathologies.

Keywords: self, research, analytic philosophy, mind, neuroscience, postmodernism, brain processes, anomalous experiences

Research and publications on the topic of the self have increased significantly in recent years across a number of disciplines, including philosophy, psychology, and neuroscience. This increase of interest in the concept of self has been motivated by a number of factors in different disciplines. In philosophy and some areas of cognitive science, the emphasis on embodied cognition has fostered a renewed interest in rethinking mind–body dualism and conceptions of the self that remain too Cartesian. Poststructuralist deconstructions of traditional metaphysical conceptions of subjectivity have led to debates about whether there are any grounds (moral if not metaphysical) for the reconstruction of the notion of self. Recent interest in Buddhist conceptions of no‐self has motivated questions about whether such a thing as self even exists. In light of new understandings of dynamic and distributed processing in the brain, philosophers and neuroscientists are exploring similar questions about whether the self might be an illusion. With respect to the self, understood as an agent, similar questions arise in experimental psychology. Advances in developmental psychology have pushed to the forefront questions about the ontogenetic origin of self‐experience, while studies of psychopathology suggest that concepts like self and agency are central to explaining important (p. 2) aspects of pathological experience. Finally, an increase of interest in narrative has also played a role in generating renewed interest in how we understand, not only ‘the self’, but also how we understand ourselves in social and cultural contexts.

This volume explores a number of these recent developments, and is not limited to any one approach. It is meant to introduce the reader to the complexity of the concept (or plurality of concepts) of self and to the many different approaches to its (their) analysis. It includes essays by leading representatives from areas such as analytic philosophy of mind, phenomenology, pragmatism, Buddhist studies, psychology and psychiatry, neuroscience, feminism, and postmodernism. These various analyses do not necessarily have the same target. Some critically focus on the notion of self as it has been constructed in social and cultural arrangements; others conceive of the self in terms of psychological continuity; others as a bodily manifestation. Some of the authors explore how certain aspects of self are constituted in brain processes, narratives, or actions; others explore how some aspects of self come apart in anomalous experiences, experiments, or pathologies.

In this introduction I will try to provide a map of this broad area of research by summarizing the problems and the conclusions that we find in the following chapters. The details of analyses and arguments are developed in the relevant chapters, and the reader can find them there.

Making a Start

The first part, ‘Self: Beginnings and Basics’, covers a number of large areas, including questions of development and neural underpinnings. We start, however, with the history of the notions of self and person. John Barresi and Raymond Martin, a psychologist and a philosopher respectively, review ancient, medieval, and modern ideas about the self and focus on a central issue of whether the self is something spiritual (an immaterial substance), and therefore beyond any natural scientific analysis, or something that can be explained naturalistically. In the twentieth century the development of this issue is reflected in the fact that, with respect to the self, ‘during the first half of the century [philosophy] labored to separate itself from science and in the last half to reintegrate itself with science’. History also shows us that many of the ideas about the self that we explore today were foreshadowed by past thinkers, from fission examples in the eighteenth‐ and nineteenth‐century writings of Clarke, Collins, Priestley, Hazlitt, and Bradley, to the idea that the self may be a fiction, in Hume and Nietzsche, or that it originates in social relations, as suggested in writings by Hartley, Reid, and Hazlitt. As such ideas have developed in the late twentieth and early twenty‐first century, the challenge, (p. 3) Barresi and Martin suggest, is to provide an integrated theory of selves that makes sense out of their experiential, ontological, and social dimensions.

An important question that may help us to address these different dimensions and to sort out what we mean by self is to ask when and how something like self, or self‐consciousness, emerges. Ontogenetically, can we say that something like a self is present in the newborn, or is the self something that emerges as the child develops? Even to attempt an answer to these questions, we need to distinguish between different aspects of self. While William James (1890) distinguished between the material self, social self, and spiritual self, Ulrich Neisser's (1988) distinctions between ecological, interpersonal, extended, private, and conceptual dimensions of the self are perhaps more directly relevant to the developmental question. Although it may be clear that newborns lack episodic memory required for an extended self, or the ability to entertain a self‐concept, this does not mean that they are necessarily without self or self‐awareness.

Philippe Rochat, from the perspective of developmental psychology, reviews the evidence for such self‐awareness in newborns, specifically a minimal and phenomenal self‐awareness in the context of feeling, perception, and action. In this regard, Neisser's concepts of ecological and, importantly, interpersonal aspects of the self are most relevant. There is good evidence that what some philosophers call the ‘minimal self’ or minimal self‐awareness is already operative in newborn humans, if we understand it to include embodied ecological and interpersonal aspects. Rochat is especially interested to explore the phenomenal and emotional aspects of this self‐awareness, how organized the experience of the body is in infants, and the role of intersubjective interaction. Over the course of the first months of life, however, and especially in contexts of intersubjective interaction, self and self‐awareness develop to the significant point of an objective and conceptually born self‐recognition. The notion of self‐recognition, however, is itself complex, ranging from primitive non‐conscious bodily processes (as in the immune system) to sophisticated aspects of self‐consciousness in human adults.

Gordon Gallup, James Anderson, and Steven Platek focus on one of these more developed aspects of self‐consciousness, namely, mirror self‐recognition, the ability to recognize one's own image in a mirror. Gallup (1970) provided the first experimental report of mirror self‐recognition. He showed that chimpanzees are able to learn that the chimps they see in the mirror are not other chimps, but themselves, as evidenced by self‐directed behavior. There is evidence (some more convincing and some less so) that this phenomenon can be found in elephants, dolphins, magpies, some gorillas, chimpanzees, and human infants starting around fifteen months. Gallup, Anderson, and Platek suggest that this ability correlates to large brain size (relative to the animal's body size), and they review evidence for a neural network for self‐recognition and self–other differentiation. They cite evidence that frontal cortex and cortical midline structures are implicated in self‐recognition tasks.

(p. 4) Again it seems important to ask precisely what aspects of self and self‐processing are being considered in the various experiments and tasks that inform any conclusions about self and self‐consciousness. This is equally so in regard to determining what neural processes and brain areas might be involved. Although a large number of studies point to frontal and cortical midline structures as important for self‐specific experience, Kai Vogeley and Shaun Gallagher reconsider this idea in light of several recent reviews of the neuroscience literature on this point. Depending on the precise nature of the questions being asked, there seems to be overwhelming evidence that the self is both everywhere and nowhere in the brain. A widely cited review by Gillihan and Farah (2005) suggests that there is no specialized or common area responsible for self‐related representations; when the entire survey of self‐related tasks is considered, the entire cortex seems to be involved. As LeDoux suggests, ‘different components of the self reflect the operation of different brain systems, which can be but are not always in sync’ (2002: 31).

In other words, frontal cortex and the cortical midline structure are not the only areas involved in self‐related tasks. In addition, however, these areas may be involved not because the tasks are self‐specific, but because they are tasks that involve a specific kind of cognitive operation, namely, reflective evaluation (Legrand and Ruby 2009). The question is then whether these areas are activated because they are self‐specific, or because the experimental tasks used to test self‐recognition, for example, involve reflective evaluation. The question of how self and brain are related, however, is not localized in one chapter; it is distributed across a number of other essays in this volume (Chapters 3 and 7 especially).

On some conceptions, what we call self may be nothing other than the product of brain processes. On other conceptions, what we call self involves a larger system that includes the whole body and the environment. The second part of this volume, ‘Bodily Selves’, explores questions about how bodily processes contribute to self. Quassim Cassam, in his chapter on the embodied self, focuses on three questions: the metaphysical question about the relation between body and self; the phenomenological question about the nature of our awareness of our own body; and the epistemological question of whether anything is special about the knowledge we have of our own bodies. Although these questions can be treated separately, they are also knotted together and, as such, in various ways they weave through a number of the chapters in this section, and the next two sections, which deal with phenomenological, metaphysical, and epistemological problems. Cassam considers various arguments for and against the claim that I (the person) am identical with my body, demonstrating how unsettled the various answers are. He also considers answers to the question of whether bodily awareness is a form of self‐awareness, showing that it may depend on whether bodily awareness targets the body qua object or the body qua subject. This distinction is central for anyone who considers the question of bodily self‐awareness and its status with respect to the possibility of misidentification. Cassam thus begins a discussion of the principle of immunity to error through misidentification (IEM) that is continued in a number of the following chapters.

(p. 5) IEM

The idea of immunity to error through misidentification can be traced back to Wittgenstein (1958), but was explored more fully by Sydney Shoemaker (1968) and subsequent thinkers. IEM characterizes judgments about self, and the information sources on which such judgments are based, iff the information source can be only about the self as subject. When, for example, I use the first‐person pronoun as subject I cannot be mistaken in regard to whom it refers. If I happen to have a toothache, Wittgenstein suggests, it would be nonsensical to ask: ‘someone has a toothache, is it I?’ I may be wrong about it being a toothache—it may be some kind of referred pain. But I cannot misidentify who it is who is experiencing it. My use of the first‐person pronoun in this case (‘I have a toothache’), my judgment, and the nocioceptive information source that is the basis for my judgment, are IEM. In contrast, uses of the first‐person pronoun as object, that is, in judgments based on sources of information that deliver only objective knowledge about ourselves, are not IEM. For example, when I look at a live video it is possible that I mistakenly identify the person in the video as myself when it is actually someone else.

John Perry, in his chapter on self‐knowledge, considers the possibility that the first‐person pronoun has this characteristic of IEM because it refers to something special, for example, a Cartesian ego. He, in agreement with Elizabeth Anscombe and Sydney Shoemaker, rejects this idea, not because, as Anscombe (1975) famously argued, the ‘I’ does not refer at all, but because the way of referring, rather than the referent itself, is special. ‘I’ is an indexical that follows the rule of a reflexive reference (the token‐reflexive rule): the word ‘I’, without any further intention accompanying the utterance, refers to the speaker of that utterance. It's not clear, however, that this rule reflects the difference Wittgenstein noted between the use of ‘I’ as‐subject, and ‘I’ as‐object. It also motivates questions, pursued by Peter Hobson, about what happens to the use of the first‐person pronoun in some pathologies.

A number of different positions are staked out in various chapters when it comes to the question of extending IEM to proprioception (bodily position sense which allows one to know where one's limbs are), an extension originally suggested by Gareth Evans (1982). Cassam argues, in agreement with Evans, that bodily self‐ascriptions based on proprioception are IEM, but IEM per se is not the thing that guarantees that proprioceptive awareness, or any other form of self‐consciousness is awareness of one's self qua subject, because some judgments (e.g. demonstrative judgments) about things that are not self are IEM. In the case of making a judgement about an object, for example, a new car, when it is perceptually present, and I refer to it as that car, my reference to it cannot involve misidentification. Cassam (p. 6) goes on to make the further point that if a judgment is IEM it is because it is based on a source of information that is IEM.

In contrast, José Bermúdez turns this around and suggests that an information source is only derivatively IEM, based on the fact that judgments based on it are IEM. That is, the property of IEM is primarily a property of judgments rather than information sources. It is not clear, however, that this would reduce the reliability of the self‐experience of the agent, or, as Elizabeth Pacherie suggests, that it implies a high reliability for judgments about self‐agency. Bermúdez agrees, however, that, like introspection, proprioception (including what Gibson calls ‘visual proprioception’, that is, information about the bodily self originating in the visual modality and linked to our bodily position or movement through the environment) is IEM. On the one hand, this is a kind of non‐conceptual awareness that has implications for agency which are constitutive for self‐consciousness and are ‘grounded in the IEM property’. On the other hand, IEM is not sufficient for something to count as self‐consciousness, since some non‐conscious sources of information (e.g. vestibular system processes) are IEM.

IEM, then, is said to apply to the use of pronouns and demonstratives, the judgments that include those uses, and the information sources on which those judgments are based. There are debates about various issues concerning IEM, but there is general agreement that if an information source or a judgment is IEM, then it is highly reliable with respect to the kind of self‐consciousness that may be associated with it. As we'll see later, this may include our awareness of ourselves as agents. It may also apply to the sense of body ownership insofar as it is based on proprioception.

Body Ownership and Some Phenomenological Details

The question of body ownership is explored in the chapters by José Bermúdez and Manos Tsakiris, from two very different perspectives. Bermúdez pursues a philosophical analysis of the concept. He reviews deflationary and inflationary conceptions of the sense of ownership and rejects the strong (inflationary) claim that the sense of ownership is a distinct and phenomenologically salient dimension of bodily awareness. On the deflationary view, ownership is not a first‐order sense (a feeling or experience) at all, but simply a fact about or a label for a certain aspect of bodily experience. Although it seems clear that one can distinguish between a first‐order experience of one's body, and a second‐order judgment about that experience, it is not so clear that there is an important difference between saying (p. 7) that the experience of bodily ownership derives from a particular fact about bodily experience (e.g. the spatial content of bodily sensations), and that the sense of ownership is an implicit feature of bodily experience. On the inflationary view, the issue may simply be a question of how implicit (recessive) or explicit (phenomenologically salient) it is, and that may depend on circumstances and individual differences. On the deflationary view, there is no feeling of ownership over and beyond the judgments of ownership and the facts about bodily experiences that ground them.

Tsakiris follows the more inflationary view and reviews a growing body of empirical research on the sense of body ownership that suggests that the latter depends on the integration of somatosensory signals. The latter may include proprioceptive, kinaesthetic, tactile, visual, and vestibular signals generated as feedback from our movements or actions. In experimental settings, the distinction between a reflective attribution or judgment of ownership and the first‐order experience of ownership is also important. The Rubber Hand Illusion (RHI), which is the focus of much of the research on body ownership, involves the experiential level, and not just the level of judgment. When one sees a rubber hand being tactilely stroked in synchrony with tactile stimulation of one's own unseen hand, one begins to feel the rubber hand to be part of one's own body (Botvinick and Cohen, 1998). In other words, the sense of ownership extends to the rubber hand. Tsakiris argues that this depends not simply on bottom–up somatosensory integration, but also on top–down modulation effected by a pre‐established representational model of the body.

Beyond rubber hand experiments, Tsakiris discusses recent research on out‐of‐body experiences (OBE). He shows that the neurocognitive processes involved in the RHI (and thus for body parts) are also involved in OBE (and thus for the body as a whole). Accordingly, multisensory integration together with the modulation of internal models of the body generate the experience of the body as being one's own as well as the experienced discrimination between one's body and other objects. This experience of bodily ownership may, Tsakiris suggests, be a critical component of self‐specificity as defined by Legrand and Ruby (2009; see Vogeley and Gallagher, in Chapter 4 below).

In her chapter on the phenomenological dimensions of bodily self‐consciousness Dorothée Legrand explores in detail our awareness of the bodily self‐as‐subject. She starts with the central distinction between self‐as‐object and self‐as‐subject, and goes on to show its relationship to the transitivity and non‐transitivity of self‐consciousness. Consciousness of myself‐as‐object, for example, in my reflective evaluation of my posture, is transitive in so far as it considers my body as an intentional object. Consciousness of myself‐as‐subject is intransitive in the sense that this form of self‐awareness does not take my body as an intentional object. When I reach for a hammer, for example, I am transitively aware of the hammer (I may have to consciously look for and locate the hammer to successfully complete (p. 8) the action of grasping it), and, at the same time, intransitively aware of my movement (in a way that I do not have to look for and locate, or keep track of, my hand). Accordingly, my intransitive awareness of my body has the characteristic of experiential transparency within the same consciousness of an object (usually other than my body).

Legrand goes on to show that even this transparent and intransitive awareness of one's own body includes some details about the experienced volume, location, and orientation of one's body. In this regard the experiencing subject is more than a perspectival zero‐point; she is, phenomenologically, something > 0 as perspectival origin, likely because experienced bodily volume and location is something > 0. Moreover, Legrand's analysis shows that the bodily self is self‐conscious, not in spite of the fact that it is intentionally oriented towards the world, but thanks to this intentional orientation. Thanks to this intentional orientation, and the working of kinaesthesis, I experience myself located in a peripersonal space that extends beyond my body's boundaries. Other aspects of experienced embodiment are opened up by different sensory modalities, and even through intersubjective experiences. Without contradiction to the experiential transparency of my body in perception and action, the modality of touch introduces a sense of opacity into the experienced volume of my body. Furthermore, when others direct their gaze at me, the fact that I experience embarrassment or comfort introduces other dimensions into my bodily experience of self. Throughout her analysis, just in such details, Legrand builds up a rich phenomenological description of the body‐as‐self.

Self Versus Non‐Self

A number of chapters turn on questions that run from phenomenological to metaphysical issues concerning the existential status of the self—questions about whether there is a self, and if there is, how much more than some minimal or basic set of characteristics we should bestow on it. On one side some theorists, influenced by Indian and Buddhist philosophies, argue that there is no self (e.g. Albahari 2006, discussed in chapters by Henry and Thompson, Siderits, and Zahavi). This is a view that challenges the view of self as embodied (Bermúdez, Cassam, Legrand), as well as views that emphasize the self as a socially and narratively constituted entity (e.g. Schechtman, Hermans, Gergen). In no case, however, does anyone defend the substantial spirit of the Cartesian self. Indeed, a number of theorists, reacting against just such an idea, posit no more than a minimal self (Strawson, Henry and Thompson), and, in one case, something less than a minimal self (Metzinger).

(p. 9) Aaron Henry and Evan Thompson, in agreement with Legrand that self is a bodily subject, argue against the no‐self or nonegological account. They find in the phenomenological concept of pre‐reflective bodily self‐awareness, which entails a basic experience of boundedness, something more than just the experience of subjectivity, namely a distinction between body and world that constitutes a basic sense of self/non‐self discrimination. This is a minimal bodily and experiential sense of self that involves a phenomenal ‘mineness’, or what Bermúdez and Tsakiris call a sense of ownership. For Henry and Thompson, the mineness of experience is based on an intransitive, non‐observational, pre‐reflective self‐awareness which is part of the structure of intentional consciousness (as such, it is what Bermúdez calls an inflationary sense of ownership, i.e. something that manifests itself phenomenologically). This self‐awareness, at a minimum, helps to account for the first‐person perspective that characterizes our experience of the world, and amounts to a ‘minimal self’ (also see Zahavi, Chapter 13). Although this idea of the minimal self is close to a nonegological conception of the sort found in Albahari (2006), Henry and Thompson argue that the difference hinges on the status of the body, which, for the former view, is constitutive, and for the latter view, merely an accompaniment. On the constitutive view, the body is experienced as subject and as bounded, which involves a fundamental distinction between self and other.

Other conceptions of the minimal self, however, are not so closely tied to bodily experience. Galen Strawson is concerned with both the metaphysics and phenomenology of the self or subject of experience. Although he posits no inconsistency between an understanding of the minimal subject as belonging to the metaphysics of mind or consciousness, on the one hand, and bodily conceptions of the self, on the other, the phenomenological description of the minimal subject requires no reference to body, environment, or social relations. The thought is not that the body is the subject who has experience, but that ‘the subject is identical with its experience’. Rather than a ‘thick’ conception of subjectivity, which would include the body as constitutive, Strawson argues for a ‘thin’ conception which equates the subject with the experience itself. No experience, no subject. Accordingly, on this minimal conception, the subject does not exist if the person (or human animal) is asleep (unconscious). What one can say about the subject is that it exists in a living moment of an indeterminably short amount of time; that it has a unitary singleness amid the rich and complex, but unified, experiential field; that it is a material process (a neural synergy) located in the brain—where or what process that is is open to empirical investigation.

For Strawson the profound metaphysical question about experience, and therefore experiential selves, is whether experience is limited to certain types of physical processes, or is characteristic of all physical processes, which would entail panpsychism. One does not need to answer this question, however, to know, as Strawson suggests at the end of his essay, that as human beings we are more than minimal selves.

(p. 10) On the other hand, is it possible that we are less than a minimal self? Thomas Metzinger further explores the ‘no‐self alternative’ and, as he suggests, takes it not to be an ‘alternative’ but rather the default position since, on his view, there is neither evidence for, nor convincing argument that there is, something that we should call ‘the self’. Even what Strawson calls a minimal subject is a little too much—something that, if not a thing or object or substance, is still, nonetheless, some instance of a real material individual rather than a theoretical entity or conceptual fiction. Metzinger can accept the general parameters of the phenomenological analyses of Legrand, Henry and Thompson, and Zahavi, for instance, that there is a distinct phenomenology of singularity in bodily self‐consciousness, but in contrast to someone like Strawson (1997) who suggested that ‘metaphysics must wait on phenomenology’, for Metzinger, phenomenology does not constrain metaphysics. If science does (i.e. constrain metaphysics), then it doesn't do so in a way that would lead us to think that there is anything that we would call a self. The closest we could come via neuroscience, for example, is to talk about a set of functional mechanisms that would integrate individual property‐representations into a unified ‘self‐representation’ or self‐model, which is not equivalent to a self in any real sense.

Metzinger outlines several different anti‐realist arguments about the self and then goes on to explain why the idea that there are no selves is counter‐intuitive. In the process he shows why the intuitions of phenomenology are traceable to the contingent fact about the causal structure of our brains, which induces in each of us a first‐person perspective that makes it difficult to deny the existence of selves. Even if the phenomenological evidence is overwhelmingly in favor of our intuitions about self, such that it makes it difficult, if not impossible, not to have an intuition about the existence of a self, this is so only because our phenomenology is generated by the kinds of brains that we have. According to Metzinger, this neuroscientific fact pulls the rug out from under phenomenology.

Buddhism is also associated with a no‐self position. As Mark Siderits shows, however, there are various Buddhist versions of this position. Distinguishing between the concepts of self and person, Siderits takes self to mean one part of the psychophysical complex, whereas person is the whole of the psychophysical complex. The Buddhist anti‐realist or reductionist position denies that the self exists, and considers the person to be a conceptual fiction. We may think of the self as something that has permanence over time; but if there is no permanence in the psychophysical complex, then, on that concept, there is no self. Why then does it seem to us that we are in fact enduring entities? It is an illusion, or a fiction, albeit a strong one, and perhaps a practical one. It is practical to think of ourselves as an enduring entity insofar as it may help us to plan for our own future happiness and to avoid practices that may impede our future welfare. Discourse about persons can include concepts such as moral responsibility, self‐interest, and so forth. If understanding ourselves as persons can be beneficial in this respect, it can also lead to (p. 11) problems and the kind of suffering associated with alienation, frustration, and despair. That there can be a valid discourse about the person at this practical or conventional level is not denied by the Buddhist reductionist; but the idea that this kind of discourse reflects something true at the ultimate, metaphysical level is ruled out.

As Siderits points out, however, not all Buddhists are reductionists in this sense. Buddhist personalists hold something close to an emergentist view, which suggests that, although the self does not have real existence, the person does. One objection to reductionist accounts, which we might call the phenomenological objection, seems to support this view. Namely, that if the reductionist holds that it merely seems to us as though we are persons when there are really just psychophysical elements arranged person‐wise, then to whom does it seem to be that there are persons? Another way to put this might be that only a person could be a reductionist. To answer this objection the reductionist appeals to consciousness as an impermanent (apparently unified, but truly not unified) and impersonal (non‐self) subject of experience, itself a concept of practical use, but not one that references something ultimately real.

The phenomenological objection to this Buddhist no‐self view is renewed by Dan Zahavi in his defense of the minimal self. Zahavi, like Henry and Thompson, considers the distinction made by Albahari (2006) between two forms of the sense of ownership: personal ownership and perspectival ownership. Personal ownership, where I identify myself as the owner of my experiences (‘these are my experiences’), on the Buddhist view, is what leads to the illusion of the self and the establishment of a felt boundary between self and other. Perspectival ownership is simply the fact that I experience what I experience from a certain perspective. These different senses of ownership can come apart in pathologies like depersonalization where the subject experiences a thought or feeling, but without the sense that this is his own thought or feeling. On Albahari's Buddhist view, there is no self in reality; the manifolds of thoughts, emotions, perceptions, and so forth, have existence, as a perspectival or ‘witness’ consciousness. This consciousness may produce a sense of self, but there is no reality to self; it is a constructed illusion.

For Zahavi there is an important contrast between this Buddhist conception and the phenomenological conception of nonegological consciousness as found, for example, in the early Husserl and Sartre. For Husserl, as for Sartre, the self (ego) is a product of the self‐constituting, self‐unifying temporal flow of consciousness, but this does not mean that it isn't real. Husserl and Sartre go different ways on this point. For Husserl the self is a transcendental ego; for Sartre, it is nothing more than an empirical ego. One possibility, as Sartre suggested, is that in first‐order experience neither of these selves exists, and that they are more the product of reflection than the product of a temporal synthesis. In that case, there may be little difference between the Buddhist view and the phenomenological view; the ego would have a false reality. But, as Zahavi shows, the difference is more apparent at (p. 12) the pre‐reflective, first‐order level of experience. That level, which is the level of what Albahari calls a perspectival witness consciousness, turns out not to be impersonal. Experiences, as Strawson suggests, always involve an experiencer; and as such they are for‐someone. That is, even the simplest perspectival experience of pain or emotion or perception is, in Sartre's terms, for‐itself. My experience is inescapably for me. Within one stream of consciousness, the difference between one experience and the next experience is not of the same sort as the difference between my current experience and the current experience of somebody else. This quality of ‘mineness’ is not something added on to experience as an extra kind of personal ownership, nor is it part of the content that I experience; rather, Zahavi argues, it is inescapably part of the structure of experience. Experience, in its very structure, involves the first‐person perspective. This brings us back to the minimal notion of self.

A Larger and More Sustained Notion of Self: Personal Identity

Debates about minimal or less than minimal selves cannot be the full story about the self or person. The notion of the minimal self tends to be synchronic or very near synchronic—happening, if not in one or more chronons (see Strawson, Chapter 10) or on the order of the neuroscientist's milliseconds, then in what James calls the specious present, which is on the order of seconds. Because we live our lives in the larger time frames of the everyday, however, we need to be able to say something about our selves in that more extended period. Since the time of John Locke there has been an ongoing discussion of personal identity, conceived as an identity over an extended time. John Campbell's chapter starts the discussion of this more extended self.

Campbell identifies two issues that are central to the question of personal identity: one concerns causality, and the other concerns the notion of first person. Any object, as well as any person, has a kind of internal or what Shoemaker (Chapter 15) calls ‘immanent’ causality: the set of properties it has now will determine how it behaves later under varying conditions, and what sort of properties it will have then. One can understand this in terms of an interventionist conception of causality. An intervention now, changing the properties I have now (whether physical properties or psychological properties), will have an effect on properties that I have later. In this regard there is no quarantine of physical changes from psychological changes: getting a tattoo now may make me happy now and unhappy later. In any case, this kind of causality is measured by the (p. 13) counterfactual. Causality is at play if we can say that, given an intervention on a person's earlier condition, there would have been a difference in the person's later condition. On some interpretations, this internal causality is constitutive of what the object is, or, in this case, what the person is. If we adopt that view, however, as Campbell suggests, it becomes clear that the causal complexity of a person stands in the way of getting a clear conception of the identity of the person. What, for example, does it take for us to have the same human being or person later as we had earlier? (see Parfit, Chapter 18, for more on this question). Since the complexity is both physical and psychological, and may add up to an even more complex psychophysical unit, the question of personal identity is a difficult one, especially when some theorists tend to look for identity in only the biological or embodied existence of the person, and others tend to look for it exclusively in psychological existence.

Issues surrounding the notion of the first person, however, also introduce some uncertainty into our conception of personal identity. Start with the first‐person pronoun. Pragmatically, the problem of reference for ‘I’ is solved by the token‐reflexive rule: any token of ‘I’ refers to whoever produced it. Epistemologically, in some cases, we have immunity to error through misidentification (see discussion of IEM, above). Beyond that, however, problems abound, since, as many of the other chapters make clear, the metaphysical status of the speaker to whom ‘I’ refers is unclear: a body, a mind, a soul, a self, a non‐self? Campbell points out that it is also not clear how the use of ‘I’ can capture or transcend the complex causal happenings that make up the person. Nonetheless, the importance of the first‐person perspective (and the use of the first‐person pronoun) in our understanding of the world seems essential for action. Here Campbell sides with Perry (1979), who shows that an impersonal description (without the ability to use ‘I’) would never motivate action or responsibility, against Parfit (1984; and Chapter 18), who suggests that we could gain in the moral realm (perhaps along the line of an ethics of care) if we were able to drop the use of the first‐person pronoun. These are issues explored in several later chapters.

Sydney Shoemaker takes up one of these questions: should identity be sought only in the biological or embodied existence of the person, or exclusively in psychological existence—or, more generally, what determines the nature of a person? Whatever the answer turns out to be, Shoemaker suggests, it involves causality. Successor states and the diachronic unity of one individual can be explained in terms of the immanent causality of its properties, and especially properties that belong to the thing because it is of a certain kind (Shoemaker calls these ‘thick’ properties). What a thing is, is constituted by what thick properties it is capable of having. These properties are the ones that determine its persistence conditions. So if we are asking what kind of things persons are, we can answer that question by specifying what sorts of properties persons have and, most importantly, what thick properties persons have. Shoemaker argues against (p. 14) the animalist view, and defends the classical neo‐Lockean view by arguing that the thick properties of person are psychological or mental ones, and that a person persists through psychological continuity.

Psychological continuity is defined in terms of immanent causality. This is not to deny that persons also have physical properties, but the physical properties that are relevant to the question of personhood are just those physical properties that instantiate mental properties. Mental properties are primary and defining even for body ownership. My body is the body whose states cause my perceptual and sensory states, and whose behavior is the product of my volitional states.

Campbell indicated two areas of difficulty when it came to personal identity: causality and the first person. While Shoemaker addresses the topic of causality, John Perry starts with the issue of how we use the first‐person pronoun. In thinking about the special way that the word ‘I’ refers, Perry suggests that what we call the self plays a certain role, similar to the role that the present day plays in relation to the word ‘today’ and to our knowledge about today. To know about the present day, the methods I use involve the present day playing a role. For example, I can look out the window to see whether it is raining or not. The methods I use today to find out about today are different from the methods I use today to find out about past or future days. I can't look out the window today to see whether it was raining on 4 July 1776. In this literal way today's methods are always different from tomorrow's when today is the object of knowledge. In knowing about the person I am, the self plays a role similar to the role that the present day plays in knowing about today.

Self‐based inquiry includes reflection, introspection, and perception. Perception gives me knowledge not only about the object perceived, but also about the perceiver. Serving this role, the self presents no difficulty; it is not something that I have independent epistemological worries about. For the present day to function the way it does in my ability to know things about it, I do not have to take today itself as an object (as I might do if I were researching it years from now); and for the self to function the way it does in my ability to know things about it (or about myself), I do not have to take the self as an object.

For example, in my appointment book I might find the following note: ‘SG has an appointment to meet John Perry in Palo Alto on April 7’. For SG to keep the appointment it is not enough for me (SG) to know just the fact that I have the appointment. I have to know when 7 April is, relative to today, and especially if it is today. I have to know that Palo Alto is here, where I am, or there where I am not. And I have to know that I am SG. The latter depends on a first‐person knowledge source, an ability to identify myself. But this ability depends, according to Perry, on a self‐notion—the idea that I have of myself that ‘controls the use of the first person’. Luckily, on the one hand, the self is not difficult to keep track of—indeed I don't need to keep track of it in the way that I have to keep track of what day it is. It is always the same, in the sense that, as Perry suggests, it is a role that is always (p. 15) played by the same person. The concept of role in this case, and as Perry uses it, is a functional one. On the other hand, the idea of a narrative self is just the idea that keeping track of the self across time is what constitutes the self as something more than a minimal self. In this case the role becomes a character role.

Marya Schechtman tells the story of the narrative self. The strongest versions of the narrative approach hold that both a person's sense of self and a person's life are narrative in structure. Schechtman calls this the hermeneutical narrative theory; it takes the self to be a self‐interpreting being. Human behavior is not tied simply to the moment, nor is it understandable completely in its physical facts; insofar as it has meaning it is related to the agent's past and future intentions and to a set of intersecting contexts. These things are best expressed in narrative form, and only in narrative form is the meaning made clear. The self is a character that plays a role in that narrative. Dennett (1992) defends a more abstract view of the narrative self: the self as the ‘center of narrative gravity’. On this account, the self has no existential reality; it is a fictional abstraction generated in the various narratives that we tell about ourselves. The self is no more than the functional role that it plays in these stories; it is the narrated more than the narrator, and the latter role Dennett assigns to the brain. Even in these narrative conceptions, then, the debates between proponents of no‐self, minimal self, and something ‘thicker’ than the minimal self are played out.

Schechtman herself defends the view that the self constitutes itself in narrative. She argues for something less than the hermeneutical view insofar as the narrative is less agency‐oriented and without an overarching thematic unity, and something more than Dennett's view insofar as the self constituted in narrative is something more real than a fictional abstraction. All three of these views contrast with those that simply focus on the fact that as humans we have narrative capacities that we sometimes deploy to make sense of our actions. Narrative capacities make us narrators, but not necessarily products of or characters in our narratives. These learned capacities enable us to understand ourselves as extended over time and to draw more robust distinctions between self and other. Insofar as our capacity to narrate gives us a certain ability to plan and control our lives, however, narrative is not simply a retrospective shadow that follows the self; rather it helps us to shape what we will become and thereby contribute to our self‐identities. As David Velleman (2006) puts it, this narrative capacity makes us agents in a strong sense. This notion of agency, and Schechtman's discussion of narrative as involving strong intersubjective connections, points us in the direction of accounts of self that enter a moral dimension. Narratives enable us to make our actions intelligible and to provide reasons for our actions, to ourselves and to others.

Narrative identity can be distinguished from personal identity following a distinction made both by Paul Ricoeur (1994) and Derek Parfit in this volume. Parfit distinguishes between numerical identity and qualitative identity, which corresponds to Ricoeur's use of the Latin terms idem and ipse, respectively. Idem (p. 16) or numerical identity signifies that a thing (e.g. a person) just is identical with itself, and that there is only one instance of its existence. Most accounts of personal identity, starting with John Locke, are about numerical identity. Ipse or qualitative identity allows for the possibility that something in the thing (or person) changes. Derek Parfit is numerically the same person who published Reasons and Persons in 1984, but qualitatively he is no longer identical with that person since he, like everyone else, has changed. Narrative identity attempts to track this kind of qualitative change in self.

Parfit, like most philosophers who think about personal identity, focuses on numerical identity. Also, like many philosophers who write about personal identity, Parfit follows a tradition started by John Locke, namely, appealing to thought experiments that help to get at the issues of numerical identity. One issue is just this: is the person I am today the same (numerically identical to the) person I was yesterday, or will be next year? Another issue is: does it matter? Parfit is interested in the criteria that we use to answer the first question. The standard criteria, as we see discussed in the chapter by Sydney Shoemaker, are bodily continuity and psychological continuity. While Shoemaker opts for psychological continuity, Parfit defends constitutional reductionism. A person is reducible to but not identical to bodily and psychological events. If I want to know anything about a person I must look to the various bodily and psychological events that constitute her life. Parfit then uses thought experiments (science fiction cases) to show that in many instances we are unable to decide about the identity of the person. The question, then, is whether this matters.

Moral Dimensions of Self

The question of whether identity matters starts us on a set of moral considerations that involve questions about agency, freedom, and responsibility. As Parfit notes, ideas of personal identity come along with implications for living and acting one way rather than another, and for responsibility, and desert. As Thomas Reid once put it, personal identity forms ‘the foundation of all rights and obligations, and of all accountableness’ (Reid 1785: 112). The idea that identity does not matter, then, seemingly challenges our normal concerns with these moral phenomena. Why should I be concerned to deliberate and to act in a certain prudential way today if in fact it is not clear that I will be the one to benefit from the results of those actions tomorrow? Why should agency matter if identity doesn't? The idea that personal identity may not matter challenges us to explain how we can decide moral responsibility. Many narrative theories of the self, we note, meet this challenge by (p. 17) denying the sort of constitutional reductionism that Parfit defends; for such narrative theories, the self is more than the purely physical and psychological events that add up to human behaviors. Rather, narrative accounts for the person in terms of her intentions and actions.

It is often said that what we do makes us who we are, and that action involves self‐discovery. Elisabeth Pacherie explores this close relationship between self and agency. Where focus on the sense of ownership (see Chapters 6 and 7) can tell us something important about how our embodied experience is structured, the sense of self‐agency is important for understanding the relation between self and action. The contemporary debate about agency raises questions about the role of consciousness in controlling action, and considers the possibility that the sense of agency is epiphenomenal or, even worse, an illusion, if we are led to believe that we have some control over our actions (see e.g. Wegner 2002).

Pacherie reviews empirical evidence that supports various theories that explain the sense of agency in either higher level cognitive (and sometimes narrative) terms or lower level sensorimotor terms. Reflecting a recent view in the literature, she argues for a more integrative model that acknowledges the contribution of multiple (higher and lower level) factors to a complex sense of agency, unified by the same principle: that the sense of agency on every level is based on comparisons between intended goal, predicted movement or action, and actual end states. Along these lines, the analysis of the sense of agency is conducted in terms of representational content (e.g. bodily movement, accomplished goal, reasons for acting), mode (experiential or cognitive), structure (directive or descriptive), and temporality. Questions about the reliability of our sense of agency naturally lead into questions about the nature of free will and moral responsibility. For Pacherie, responses to the latter questions will depend on how metaphysically loaded the conception of free will is. That we at least sometimes consciously deliberate and act on our deliberations, however, defeats any overly generalized dismissal of free will.

Alfred Mele takes this fact of conscious deliberation as a starting point to develop a neo‐Aristotelian conception of self‐control, a concept that seems essential to what it means to be a mature human person. That we sometimes succeed and sometimes fail to act in accordance with our best‐laid plans is worth considering. What happens when we lose self‐control, a moral condition known as akrasia? To understand self‐control or the lack of it, one needs to understand the conception of self that underpins it. Aristotle regarded the human self to be primarily rational, where reason is taken in a strong sense. Mele suggests a more holistic conception of the self, where to act out of passion may not mean that one is acting without self‐control. This means that the kind of inner conflict associated with akrasia should not be explained purely in terms of rational versus irrational forces. The evaluations that support our rational decisions are often driven by emotional factors rather than cold logic. For any action there may be both rational and (p. 18) non‐rational motivations involved, but one can find both on the supporting or opposing side of self‐control.

That the self in self‐control may be something other than a purely rational agent conceived in terms of the standard model of mental causation (my mental states directing my body to move in certain ways to accomplish an action) addresses certain worries mentioned by both Pacherie and Mele. On both phenomenological and metaphysical grounds one can object to the standard model. Just as I don't experience my brain as causing my actions, I don't experience my mental states per se as causing my actions; rather I experience myself as controlling my actions. It is not the brain that acts; nor is it the mind that acts; it is the person that I am who engages in action. As Mele and others (Nagel 1986; Velleman 1992) suggest, the person, as agent, tends to disappear in many of the standard causal accounts that draw a line from mental states through bodily movement to worldly action. Mele proposes a place for the more holistic conception of self insofar as someone, not just some brain state, and not just some mental state, must be motivated to exercise self‐control, must care how she conducts herself, must be a practically concerned individual. Without giving up on causal theories entirely, it is this overarching conception of caring about how one's life goes that brings the dimension of personhood back into play.

The practically concerned individual is the one responsible for her own actions. David Shoemaker sets out to show that what makes an action the person's own (what keeps the agent in the picture) with regard to responsibility is better expressed in terms of practical identity than in terms of personal identity. A standard view is that moral responsibility presupposes personal identity. This view would send us back to questions about criteria for personal identity, for example, psychological continuity versus biological continuity (animalism), and to the problem Parfit raises with regard to whether personal identity matters. David Shoemaker shows that all such criteria of personal identity (psychological, biological, narrative) fail to provide a sufficient account of how a person can be morally responsible for an action. Specifically he shows that action ownership, which is important for moral responsibility, does not entail identity. Ownership is a person‐to‐action relation whereas identity is a person‐to‐person relation. Accordingly, identity doesn't matter for moral responsibility.

This does not solve the issue of how self is related to the question of responsibility, however. Shoemaker thus examines the model of the ‘Real Self View’ (RSV) concerning ownership of action and moral responsibility, and various objections that have been raised against it. The RSV states that an action is one's own if it is governed by a will consistent with one's system of valuation, where the idea of having a valuation system is similar to Mele's idea of caring. To maintain the RSV, Shoemaker indicates that it must be open to an approach based on attributable responsibility (where one can be attributed responsibility as a person, often accompanied by significant reactive, and usually emotional, attitudes) rather than (p. 19) an accountable responsibility (where one is held responsible to the extent that one has normative competency). In the former case, an action is judged to be an expression of some aspect (e.g. a character flaw) of the person, and not just caused by the person.

Shoemaker here makes clear the distinction between attributability and accountability using the example of the psychopath. We may want to attribute moral responsibility to the psychopath insofar as we want to judge his actions cruel, or manipulative, even if we don't want to hold him accountable since he lacks the ability to grasp and apply moral reasons. In a number of ways the questions about the moral dimensions of self point to issues concerning normalcy and pathology. Are some pathologies, such as psychopathy, pathologies of the self?

Pathologies of the Self

Part VI of this volume, on pathologies of the self, is very much a fractal image of the previous parts. It covers much the same ground—issues pertaining to development, embodied self, minimal self, personal and narrative identities, and moral dimensions of the self. This time these issues are run against the foil of psychopathology, which allows us to question everything all over again.

The idea that all psychopathologies involve the self can be taken in a thoroughly trivial sense that they all happen to someone. But in a non‐trivial sense there are certain pathologies that are self‐specific insofar as they are devastating or disruptive for the person's self‐experience. Schizophrenia is a clear but complex example of this. Josef Parnas and Louis Sass focus on the phenomenology of schizophrenia and its disruption of ipseity, the basic (minimal) self‐experience. There are two important lessons we can learn from their analysis. First, our self‐experience is not neutral with respect to the metaphysical status of the self. The existence of the self is in some large part constituted in its experience, and if that experience is significantly disrupted, it threatens the self in its existence. The second point follows from the first. To understand schizophrenia, one must carefully attend to the experience of the subject. It is, perhaps, the lack of such attention that explains why the concept of self is not mentioned in the standard diagnostic criteria for schizophrenia (DSM‐IV 2000; ICD 10 n.d.).

Parnas and Sass appeal to large‐scale, clinically based empirical studies, informed by phenomenology, to establish the basic fact that subjects within the schizophrenic spectrum suffer from self‐specific disorders, namely, ‘disorders that affect the articulation and functioning of minimal or core self’. The disturbance includes two main features: (1) hyper‐reflexivity: an exaggerated self‐consciousness, where the subject (p. 20) directs focal, objectifying attention toward processes that would normally remain tacit in experience; (2) diminished self‐affection: a decline in the experienced sense of existing as a living and unified subject of awareness. The result is variable disruptions in the sense of self‐presence, first‐person perspective, the phenomenality (the what‐it‐is‐like) of experience, and intentionality (including the subject's sense of reality). Disruptions of this sort directly affect the minimal self, but as Parnas and Sass note, they also have implications for the narrative self (see Gallagher 2003).

In her account of the narrative self, Schechtman, following Dennett, mentioned Multiple Personality Disorder (recently renamed Dissociative Identity Disorder, DID) as a pathology that involves non‐intersecting narratives. Jennifer Radden takes a closer look at DID in her chapter on multiple selves. She considers recent approaches to the self, including embodied, agentive, and narrative approaches, and suggests that while none rule out the possibility of multiple selves (ranging from non‐pathological to the pathological cases of DID), the idea of rational agency, following the work of Carol Rovane (1998), is perhaps the most appropriate for explaining these types of multiplicity. That is, in the realm of agency, within one human being, one may find separate sets of practical commitments and small‐scale, incommensurable projects. This approach, however, doesn't account for the distinct (and often non‐communicating) centers of apperception in DID, a feature that is explained in Stephen Braude's (1995) account. Radden attempts to introduce this latter feature of multiple centers of apperception into the agentive account, thus offering a more adequate explanation for the possibility of DID. To do this she introduces the idea that personality traits, which easily mesh with agentive patterns, may be a place to start. This construes self as involving dispositions or traits that are manifested, with some degree of continuity, over time. Together with epistemic or amnesic barriers that can lead to dissociated awareness, as discussed by Braude, we can get a better picture of DID. Still, the picture is often blurred by the fact that each of these factors admits of degree.

Discussions of DID may be clarifying for questions about our normative conceptions of self‐unity. It seems possible that we are not as unified as we sometimes think, even if we do not reach degrees of pathology. Whether unity is an ideal to aim for, or, as Radden suggests, the lack of unity, the sometimes contradictory and sometimes confusing agentive roles that we play, is a valuable part of human nature, is a good question to keep in mind as we think about the nature of the self. It also has implications for therapy, and it forces us to consider whether and to what degree unity is to be aimed for in terms of treating pathological subjects. Clearly some relatively strong degree of self‐unity is expressed in the normative preferences of most societies, but whether such norms should be imposed in therapy is a question similar to the one raised by Strawson concerning the moral claim that it is a good thing to experience one's life as a narrative (see Chapter 17).

(p. 21) Peter Hobson, in his chapter on autism, draws on developmental psychopathology to gain insight into the self. This involves, first, focusing on evidence about complementary aspects in the seemingly unified phenomena of the self that may be distinguishable factors in development; and, second, understanding that what we may regard as familiar distinctions in mental life may originate as indissoluble aspects in early experience, for example, cognition and emotion. Hobson is interested in the question of how we, especially as infants, enter into, and sometimes fail to enter into, a shared form of life (Wittgenstein 1958; also see Rochat, Chapter 2). Amid the multitude of problems that autistic individuals have in regard to social relations, there are clues that such problems have an effect on how they relate to themselves. In this regard, Hobson suggests that autism is not only a disorder in self–other relations but also in self‐awareness.

Hobson reviews evidence that autistic children have, among other problems, difficulties in using the personal pronouns, confusing ‘I’, ‘you’, ‘she’, and so forth, one among many things that reflect the lack of coherency in their practices of self‐reference, or in co‐reference between self and other. This problem isn't simply on the grammatical level, of course; it extends deep into their bodily capacities for movement and interaction with others and in some cases (e.g. in some instances of imitating another's action) affects their ability to adopt the appropriate spatial (egocentric) perspective; in other cases their ability to adopt appropriate psychological or emotional perspective vis‐à‐vis others. Autistic children thus have trouble experiencing and understanding themselves as selves in relation to other people with selves of their own. At the conceptual level, Hobson suggests, the notion of self is operative in both self‐ascription and other‐ascription; to learn what ‘self’ means is to learn that it applies to oneself and to others, who are also selves.

The importance of this co‐constitutive relation between self and other is also something to be found in the psychoanalytic relation, as Marcia Cavell makes clear. In the therapeutic relation, the struggle for self‐expression helps create the thing to be expressed, and in some sense the self comes to be in its attempt to be understood by the other. The same can be said about the self in regard to action, as well as in regard to artistic expression. More generally Cavell emphasizes the point that Hobson and Rochat also emphasized: that the self is generated from the beginning in its social relations. For Cavell, however, there is no self at first; it emerges in our relations with others; it then grows, ‘leads a life’, with some continuity; but it can also fragment and come apart in ways that can be either pathological or creative.

The coming apart of the self is not always pathological. One can find it in self‐deception and denial, which are often part of our normal everyday life. Here we find intra‐psychic conflict. Repression and dissociation may lead further toward pathology. The first, repression, Cavell indicates, is vertical—the self coming apart in layers, leaving some important layer of meaning in the dark; the second, dissociation, is horizontal—some set of experiences (part of an idea, an action, a feeling, an experience, a memory) setting up their own shop with a different person (p. 22) behind the counter, as in DID. But some forms of dissociation can be creative and can be integrated—or not integrated—with the self. Art, literature, and our creative encounters with others can count here.

The Self in Diverse Contexts

In the last part of this book the reader will find other fractal elements—repeating and reinforcing themes, yet on different scales and in different contexts, and often reviewing from a critical perspective the positions that have been staked out in previous chapters. We can find, for example, ideas about the formation of self in habits and as developing in social interaction in the work of the pragmatists—Peirce, James, Dewey and Mead—as Richard Menary makes clear in his extended discussion of a pragmatic conception of self. This conception reinforces the agentive idea that what we do makes us who we are: I am what I do. As such, there is no pre‐established certainty in the self; it is marked by fallibility and I come to know myself only through my interactions with physical and social environments. My self‐awareness develops only by attending to the continuities and habitualities in my practices, which are already permeated by the gestures of others, which I internalize to form what I call my self. All of these actions and interactions and self‐formations shape a moral dimension in which we have to negotiate our freedom, since we are neither absolutely free nor absolutely determined.

Menary takes this pragmatic account to support an externalist conception of the self. We construct our selves out in the world, among things, with people, in institutions, all of which operate as scaffolds in this process. To the extent that the environment affords regular structures, including those that are shaped on the bodies of others—gestures, words and facial expressions—our habits become regularized. Yet, they do not become rigid, and the environments themselves are not entirely predictable or stable. The self, then, is not something ‘in the head’, the Cartesian ego, a mental structure, or the brain; it is distributed through embodied practices into the environment. Here we can find good support in John Dewey, noting his proximity to recent conceptions of extended cognition. In complete contrast to what we normally take to be the Cartesian self, Dewey states: ‘Thinking, or knowledge getting, [and here we would insert ‘self’] is far from being the armchair thing it is often supposed to be. The reason it is not an armchair thing is that it is not an event going on exclusively within the cortex.…Hands and feet, apparatus and appliances of all kinds are as much a part of it as changes within the brain.’ (1916: 13–14).

(p. 23) This kind of externalist view opens up the question pursued by Kenneth Gergen—is the self socially constructed? A social constructionist approach to the self is critical insofar as it targets many of the traditional conceptions of self under discussion in this volume. At the same time, a point of departure for the social constructionist is something mentioned numerous times before: that self and self‐knowledge find their origins in human—social, cultural—relationships. This social orientation, however, is set as the compass for truth and falsity, objectivity and subjectivity, the scientific and the mythical, the rational and the irrational, and so forth. That is, these categories, and the power that they have over our philosophical ideas as well as our everyday practices, are brought into being through historically and culturally situated social processes. The social constructionist thus challenges the concept of the individual autonomous self, the rational, self‐directing, morally centered agent of action.

Gergen targets the primary use of the term ‘self’ in psychological and mental discourse, suggesting many ways in which the sources of knowledge about the realm of the mental, for example, introspection, behavioral observation, are open to question. Perhaps more positively, looking at how we use language—or how language allows us to build up conceptions of the self through, for example, the uses of personal pronouns—we can see how a notion of the self is generated. Not grammar alone, however, but conversational and narrative practices, especially, seem to be well‐established sources for the fashioning and fabrication of our sense of self. The analyses of such practices, however, simply reinforce the movement of focus away from purely mental or psychological conceptions of self. Furthermore, the practices themselves have a certain power insofar as they establish vocabularies of the self that legitimize the concept, and that in turn rationalize and sustain these very practices. Here Gergen cites the work of Foucault and others who develop a sustained critique of what may be the ‘dark side’ of the construction of the self as an autonomous, alienated entity differentiated from the other, laying the ground for organizing a world characterized by separation, distrust, and self‐gain. Such relations then get reflected in disciplines like psychology and economics, and psychiatry, which then reinforce the normative divisions that define contemporary conceptions of self in theory and practice.

Alternative conceptions of the self, which acknowledge a certain primacy of the social over the individual, hold some promise, according to Gergen, and can be found in various theorists, including the pragmatism of Mead and the work of Bakhtin on dialogicality. These are just the thinkers that Hubert Hermans takes to be foundational for the theory of the dialogical self, which he construes as a combination of American pragmatism and Russian dialogism. This theory views the other not as external to self, but as part of the self and constitutive of it. The self is intrinsically social, a micro‐society within the larger society to which it belongs.

Hermans introduces the notion of positioning as a further articulation of the dialogical self situated in time and space. The idea is that the self is not organized (p. 24) around one center or core, but distributed over social relations. Insofar as an individual is immersed in dialogical interactions (relations, conversations, cooperative or competitive projects) with others in the social world, the self is distributed in those relations, playing perhaps very different roles or occupying different positions in each one. These different positions are themselves in dialogue with each other within the comprehensive set of relations that we call self. The self fluctuates across these different positions which are sometimes and imperfectly collected together as a set of differences. These differences are given voice as different characters, or different roles to be played out in different times and places, generating a set of complex, intersecting narratives that reflect and are reflected in the self‐structure. These differences can be defined in terms of relative dominance positions of speakers in interaction, or in terms of socially defined power differences tied to class, race, gender, religion, age, etc., all of which can be reflected in a person's self‐perception and self‐understanding. What defines me as an intellectual or a factory worker does not define me as a sexual partner or as a parent. Such positions may come in contact, informing each other, or may not. A self is a unique map of these different positions.

This conception of the dialogical self has implications for a variety of investigations, and Hermans traces some of these out. For example, one can employ the concept in an attempt to understand pathological conditions such as DID and other dissociations discussed by Radden and Cavell. He notes that self‐positions are not constituted in abstract terms, but in embodied practices and spatially situated circumstances. There is also a developmental story to be told about the dialogical self, one that would reinterpret the interactive aspects of development iterated by Rochat. In this regard, for example, one can think that the infant rehearses the positions previously experienced in her embodied relationship with others in her pre‐sleep (proto‐) conversations with herself (see the case of Sarah in Garvey 1984, discussed by Hermans). Hermans's notion of positions can also fruitfully inform feminist or poststructuralist analyses, a possibility that is suggested by Elspeth Probyn's concept of gender positions.

Probyn considers the possibility that young women, with few resources for fashioning a strong sense of self, specifically because of their gender positions, use anorexia as a mode of negotiating societal strictures. Her notion of gender position is closely tied to a psychoanalytic (Lacan)–Marxist (Althusser)–poststructuralist (Foucault) conception of subjectification. The institutional imprinting of a certain image, and a certain kind of third‐person observation (a ‘regime of visibility’) imposed on the girls/patients/anorexics is just such an exercise of subjectification. What is the self like if it is intensively inspected as an object—a medical patient for example—if it is so much the object of inspection that it feels, in the words of one anorexic, like she were made of glass? What happens when the subject becomes a subject of research, which is to say, an object of research?

(p. 25) Furthermore, Probyn asks, what happens to the researcher when certain emotions manifest themselves in connection with the research—something that is not unusual when not only the researchers, but also the research subjects, are human? Such emotion belongs to the subjectivity of the researcher. It belongs to her self. Let's say that the emotional dimension defines an essential aspect of self. Does the researcher simply leave this dimension of herself behind, for example, when the subject matter of research is the self itself? How heavily (i.e. emotionally) invested is the researcher in her research, or the position that she stakes out and defends, such that, even if she fails to mention the emotional dimension of the self, it is lurking there between the lines?

The reader will notice that the questions of power and subjectification are ones that have arisen only in these last three chapters (Gergen, Hermans, Probyn), without having come up at all in the first twenty‐six. The general question posed by these chapters, positioned as they are for critical reflection in this volume, posed one might say against not only the traditional conceptions of self, but against even the most contemporary ones, is the critical question of whether these social, cultural, institutional power structures, these forces of subjectification, can qualify everything that has already been said about the self. Probyn makes this question clear in regard to research on anorexic subjects and the formation of a research discourse that will classify those subjects, and more generally. This motivates a similar question about the theoretical stakes associated with this volume. What does it mean, in terms of practical effect, to inspect the self and to develop a vocabulary, or a set of vocabularies, about the self? This is not, of course, the subjectification of a person. The self is not our patient. But if we are setting a vocabulary, and a system of classification where we can decide between minimal and narrative selves, embodied and psychological selves, and the various categories that have been used to specify self and to talk of action, moral responsibility, pathological behaviour, etc., and if we have any expectations that this will change our way of thinking about selves—about our selves—should we not think critically about what that means in terms of the practical effect that it may have? At the very least, raising such questions motivates, I think, self‐reflection on the set of projects that constitute this volume.

Postmodern thought also poses a set of critical questions about the self. Leonard Lawlor guides us through these questions. For the postmodern, the self is heterogeneous, which means that it is multiple (a theme we also find in Gergen and Hermans). Rather than an ‘I’ or a ‘me’, there is a ‘we’. But even the ‘we’ lacks unity and is thus more like a political collective characterized by differences instead of identities. Rather than a dialogical ‘we’, the ‘we’ is agonistic. Thus Lawlor looks to the political as a helpful metaphor for understanding the self. The postmodern political, in contrast to a social bond, however, is not something that can be brought into unity or under the control of a narrative. Rather, an economic discourse of performance and efficiency (just as it characterizes the political (p. 26) today) seems better able to position the self within the larger system. Here it is not so much that self is observed and disciplined (as Probyn suggests), but that self is measured, commodified, bought and sold, within a system that is dominated by a totalizing (totalitarian), global capitalistic set of values (where values are simply quantifications). What is the self worth; what is its cash value?

This description, which tries to capture the real situation of postmodern selves, is meant as a critique, which at the very least raises the question of what we are doing to ourselves. The political, then, is no longer a metaphor. Rather, the question about the self is just as pressing as (and not disconnected from), for example, environmental politics.

What are we doing to ourselves? Lawlor sets out a response in terms of imagining the right kind of literature that would express and constitute the non‐totalitarian self, without pretending that this would be a ‘we’ that exists in consensus or as otherwise unified. He turns to the analysis of the temporality of experience, in Husserl and Bergson, which then, on a postmodern reading, shows that at base the self is caught in a tension (or is the tension) between an element that goes against time (ana‐chronism), and a powerlessness to escape time. Being against time, however, is being against the efficiency imposed by the economic system (where time is of great value: ‘time is money’). It suggests that the economic discourse doesn't really fit the self, as it tries to fit the self into the efficiency of the system. The alternative to the varieties of totalizing connected with narrative or economic calculation is to accept a powerlessness in the face of the passage of time—to be ‘friends’ of this passage, and to refrain from imposing on it the narrative strictures or the economic structures that characterize our time. This is also to refrain from imposing identity on the self, and is rather to treat the self as a constant possibility, which is never completed.

Lorraine Code carries these kind of postmodern or poststructuralist themes into a feminist analysis of self. The ideas of positioning and being positioned within power structures, or linguistic structures, or the ‘social imaginary’ that define the subject, often as a thin, decontextualized, abstract self, have implications for epistemological, moral, and political philosophies. For Code, a viable feminist analysis of self needs to recognize that selves/subjects are always embodied and situated—which means, always gendered, raced, and identified in various ways. To understand real selves, one needs to understand their particular positions and how, as such, they are thrown together into the complex, rich, and challenging world. This is what she calls the ‘subjectivity’ of the self—subjectivity not in the sense of phenomenological subject, but in the sense of the particular social and political subject that anyone actually is.

Science and philosophy can study what the self is; in doing so, they tend to ignore who we are; the focus on numerical identity fails to recognize the qualitative aspects that define our lives. Narrative accounts seem to have a place for the who, but narrative theorists nonetheless spend much of their time debating about the (p. 27) what: are we narrative animals, are we abstract centers of gravity, are narratives constitutive of the self, and so on? Of course scientists and philosophers are not meant to be biographers, but they should not ignore the strongly situated aspects of their subjects. Is personal identity a universal that does not change across cultures? Across genders? Across economic classes? Perhaps science and philosophy need to be balanced with history and literature in order to get an adequate picture of what selves can be. The danger here, as Code points out, is that the Others will be defined as such, as meeting or failing to meet criteria of personal identity already sorted out in terms of white, male, western, abstract, autonomous, affluently academic standards.

Code cites Annette Baier (1985: 84) as indicating one possible way to shift the center of the discussion: ‘A person, perhaps, is best seen as one who was long enough dependent upon other persons to acquire the essential arts of personhood. Persons essentially are second persons.’ The point is that philosophical analyses that suppress intersubjectivity in favour of the detached first‐person individual self (analyses that are philosophically autistic) conceive of ‘the self’ as quite removed from how human beings actually are. A second‐person orientation opens up possible ways of rethinking autonomy and exploring the kind of ideas that can be found in pragmatist philosophers and dialogical approaches (see Menary and Hermans, respectively), as well as in feminist ethics. Code's conception of understanding selves as differing as they are positioned across differing ecologies helps to break the lines that have been drawn too tightly around standard philosophical conceptions of self.

Where does this leave us? In the various chapters that compose this volume the reader can trace different conceptions of the self, some set in opposition to others—minimal, narrative, real, not real, existing, illusory, reduced, irreducible, embodied, psychological, social, pathological, socially constructed, and deconstructed. This 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, phenomenological analysis, linguistic analysis, the use of thought experiments, empirical research in cognitive and brain sciences, developmental studies and studies of exceptional and pathological behavior, ethical, social and political analyses. One question is whether different characterizations of self signify diverse aspects of some unitary phenomenon, or whether they pick out different and unrelated phenomena. Regardless of how one answers this question, however, the variety of approaches and definitions found in studies of the self productively reinforces the idea that this concept involves complex and varied aspects that are not easily reducible to one set of principles. As a result, it seems imperative to keep methodological options open and to explore different conceptions of self and non‐self in the analyses of philosophy of mind, moral and social philosophy, psychology, psychopathology, phenomenology, neuroscience, psychoanalysis, history, literature, narrative theory, ethnology, cross cultural studies, (p. 28) and so on. What we can be certain about, as we'll see in the following pages, is that we are not all on the same page when it comes to conceptions of self.


Albahari, M. (2006). Analytical Buddhism: The Two‐Tiered Illusion of Self (New York: Palgrave Macmillan).Find this resource:

Anscombe, G. E. M. (1975). ‘The First Person’, in S. Guttenplan (ed.), Mind and Language (Oxford: Oxford University Press). Reprinted in Cassam (1994: 140–59).Find this resource:

Baier, A. (1985). Postures of the Mind: Essays on Minds and Morals (Minneapolis: University of Minnesota Press).Find this resource:

Botvinick, M., and Cohen, J. (1998). ‘Rubber Hands “Feel” Touch that Eyes See’, Nature, 391: 756.Find this resource:

Braude, S. E. (1995). ‘Multiple Personality and Moral Responsibility’, Philosophy, Psychiatry and Psychology, 3: 37–54.Find this resource:

Cassam, Q. (ed.) (1994). Self‐Knowledge (Oxford: Oxford University Press).Find this resource:

Dennett, D. (1992). ‘The Self as a Center of Narrative Gravity’, in F. S. Kessel, P. M. Cole, and D. L. Johnson (eds), Self and Consciousness: Multiple Perspectives (Hillsdale, NJ: Lawrence Erlbaum Associates), 103–15.Find this resource:

Dewey, J. (1916). Essays in Experimental Logic (New York: Dover).Find this resource:

DSM‐IV (2000). American Psychiatric Association, Diagnostic and Statistical Manual of Mental Disorders, 4th edn, text revision (Washington, DC: American Psychiatric Association).Find this resource:

Evans, G. (1982). The Varieties of Reference (Oxford: Oxford University Press).Find this resource:

Garvey, C. (1984). Children's Talk (Cambridge, Mass.: Harvard University Press).Find this resource:

Gallagher, S. (2003). ‘Self‐Narrative in Schizophrenia’, in A. S. David and T. Kircher (eds), The Self in Neuroscience and Psychiatry (Cambridge: Cambridge University Press), 336–57.Find this resource:

Gallup, G. G. (1970). ‘Chimpanzees: Self‐Recognition’, Science, 167(3914), 86–7.Find this resource:

Gillihan, S. J., and Farah, M. J. (2005). ‘Is Self Special? A Critical Review of Evidence from Experimental Psychology and Cognitive Neuroscience’, Psychological Bulletin, 131: 76–97.Find this resource:

ICD 10 (n.d.). World Health Organization, International Statistical Classification of Diseases and Related Health Problems, 10th revision ( icd10online/).

James, W. (1890). The Principles of Psychology (Cambridge, Mass.: Harvard University Press).Find this resource:

LeDoux, J. (2002). The Synaptic Self (New York: Penguin Putnam).Find this resource:

Legrand, D., and Ruby, P. (2009). ‘What is Self‐Specific? Theoretical Investigation and Critical Review of Neuroimaging Results’, Psychological Review, 116: 252–82.Find this resource:

Nagel, T. (1986). The View from Nowhere (New York: Oxford University Press).Find this resource:

Neisser, U. (1988). ‘Five Kinds of Self‐Knowledge’, Philosophical Psychology, 1: 35–59.Find this resource:

Parfit, D. (1984). Reasons and Persons (Oxford: Oxford University Press).Find this resource:

Perry, J. (1979). ‘The Essential Indexical’, Philosophical Review, 86: 874–97.Find this resource:

Ricoeur, P. (1994). Oneself as Another, tr. K. Blamey (Chicago: University of Chicago Press).Find this resource:

(p. 29) Reid, T. (1785). Essays on the Intellectual Powers of Man (Cambridge: John Bartlett, 1850).Find this resource:

Rovane, C. (1998). The Bounds of Agency (Princeton: Princeton University Press).Find this resource:

Shoemaker, S. (1968). ‘Self‐Reference and Self‐Awareness’, Journal of Philosophy, 65: 555–67.Find this resource:

Strawson, G. (1997). ‘The Self’, Journal of Consciousness Studies, 4(5/6): 405–28.Find this resource:

Velleman, J. D. (1992). ‘What Happens When Someone Acts?’, Mind, 101: 461–81.Find this resource:

—— (2006). Self to Self: Selected Essays (Cambridge: Cambridge University Press).Find this resource:

Wegner, D. (2002). The Illusion of Conscious Will (Cambridge, Mass.: MIT Press).Find this resource:

Wittgenstein, L. (1958). The Blue and Brown Books (Oxford: Oxford University Press). (p. 30) Find this resource:


Thanks to research grants from the École Normale Supérieure de Lyon and from CNRS for support of SG's work on this introduction while a visiting professor at the ENS in Lyon, and visiting researcher at the Centre de Recherche en Epistémelogie Appliquée, École Polytechnique, Paris.