I understand bodily experience in very general terms as the sum of the first personal ways that we have of finding out about our bodily properties. These include
• proprioceptive experience of how our limbs are disposed;
• experience of how one's body is oriented and in balance (or imbalance);
• kinesthetic experience of movement, both of the body as a whole and of individual body parts;
• perspectival information about one's body and bodily location derived through vision, primarily, and through the other senses to a lesser extent; and
• experience of one's own homeostatic states obtained via bodily sensations
These different types of bodily experience seem not just to be ways of finding things out about our own bodies. They are also, one might reasonably think, ways in which the bodily self is directly encountered in experience—ways in which we are conscious of the bodily self. Once that is granted, it seems a short move to think of bodily experience as a form of self-consciousness. The viability of that short move is the subject of this article.
To help structure the discussion, here are four theses about self-consciousness and bodily experience:
Bodily self-consciousness
Bodily experience is a form of self-consciousness. We are conscious of ourselves through experiencing our own bodies.
Subjective self-consciousness
Nothing can count as a genuine form of self-consciousness unless it is consciousness of oneself as a subject.
Objective bodily self-presentation
In bodily experience we are presented with ourselves as physical objects.
Bodily distinctiveness
The way we encounter ourselves in bodily experience is fundamentally different from how we encounter non-bodily physical objects in outward-directed, exteroceptive perception.
As the title of this article suggests, these four theses have distinguished proponents in the history of philosophy. The first we find in Descartes, the second in Kant, the third in Locke, and the fourth in Merleau-Ponty. No doubt, each can be found in many other places, also, but I take these four philosophers as canonical illustrations of the four theses
Some of the four theses go naturally together. So, for example, if bodily experience is a form of self-consciousness, then that would seem to be one reason why the way we encounter ourselves in bodily experience is different from the way we encounter non-bodily physical objects in vision and other sensory modalities. But there are also significant tensions, the most striking being between objective bodily self-presentation and the conjunction of bodily self-consciousness and subjective self-consciousness. If we are presented with ourselves as physical objects in bodily experience, then it would seem that we cannot be presented to ourselves as subjects in bodily experience. No single episode of bodily experience can present us to ourselves both as an object and as a subject, one might reasonably think.
Still, the principal claim of this article is that these four theses about bodily experience and self-consciousness are not just consistent. They are also all true. Bodily experience provides a form of self-consciousness in which we are presented to ourselves as subjects even while we experience ourselves as physical objects, albeit differently from how we experience non-bodily physical objects.
Here is the plan of attack. Section I spells out the four theses in more detail, focusing on their historical antecedents in Descartes, Kant, Locke, and Merleau-Ponty. In Section II, I focus on the apparent tension between bodily self-presentation and bodily self-consciousness. In his book Self and World (Reference Cassam1997), Quassim Cassam proposes two ways in which he thinks we are aware of ourselves both as physical objects and qua subjects. In particular, he emphasizes how bodily experience presents us to ourselves as the point of origin of egocentric space, and how in bodily experience we experience the spatial boundaries of the body as the boundaries of the self. These are both important aspects of the phenomenology of bodily experience. However, I think that they do not do full justice to subjective self-consciousness, in ways that are brought out in Béatrice Longuenesse's interpretation of Kant's ‘I think’ in her book I, Me. Mine: Back to Kant and Back Again (Reference Longuenesse2017).
But that does not mean that we should reject bodily self-consciousness. In Sections III and IV, I explain why I disagree with Longuenesse's own claim that only ‘I think’ judgments express a consciousness of oneself qua subject. Even if we accept her basic premise that to be conscious of oneself as subject is to be conscious of oneself as a thinker, we need to accept a broader conception of what it is to be a thinker—and, correlatively, a broader conception of what it is to be conscious of oneself as a thinker. Thinking is not a purely ratiocinative exercise. It is an active process. And relative to that broader conception, bodily experience is indeed an integral part of consciousness of oneself as subject.
In Section V, I turn to bodily distinctiveness. In bodily experience, we are presented with ourselves as physical objects, but in a very different way from how we are presented with non-bodily physical objects. This is often put by saying that we are aware of our bodies in bodily experience ‘from the inside’. I unpack this metaphor through the idea that the non-Cartesian frame of reference through which we experience the space of the body is fundamentally different from the Cartesian frames of reference that structure vision and the other modalities through which we experience non-bodily objects. This way of experiencing the space of the body makes manifest to the subject their bodily agency and potentiality for bodily movement.
I
Descartes's Thesis: Bodily Self-Consciousness
One of the most quoted passages in discussions of the body and the self comes from Descartes's Meditations in First Philosophy. In Meditation Six, Descartes is reviewing the things in which he can have confidence, now that he is sure that his clear and distinct ideas are reliable guides to the nature of reality. He writes,
[t]here is nothing that my own nature teaches me more vividly than that I have a body, and that when I feel pain there is something wrong with the body, and that when I feel hungry or thirsty the body needs food and drink, and so on. So I should not doubt that there is some truth in this.
Nature also teaches me, by these sensations of pain, hunger, thirst, and so on, that I am not merely present in my body as a sailor is present in his ship, but that I am very close and, as it were, intermingled with it, so that I and the body form a unit. If this were not so then I, who am nothing but a thinking thing, would not feel pain when the body was hurt, but would perceive the damage solely by the intellect, just as a sailor perceives by sight if anything in his ship is broken. (Descartes [Reference Descartes, Cottingham, Stoothoff and Murdoch1641] 1984, 56.)
The simile that Descartes is rejecting, that the mind inhabits the body as a sailor inhabits a ship has a venerable ancestry in the history of philosophy. It appears for the first time in a puzzling passage at the end of the first chapter of Book II of Aristotle's De Anima and then appears to have become a standard trope. (Puzzling because Aristotle appears to use it to introduce as a live possibility the separability of psuchē [soul] and soma [body], which the preceding discussion seems conclusively to have ruled out. The difficulties are summarized in Shields's translation and commentary [Aristotle Reference Aristotle2016: 179–81], with a more detailed discussion in Tracy Reference Tracy1982.)
The famous passage from the Sixth Meditation where Descartes rejects the sailor/ship simile is focused more on phenomenology than on metaphysics. This is fortunate, because interpreting Descartes's views on the metaphysics of the mind is fraught with difficulty, since he simultaneously maintains that body and mind are distinct and in substantial union, which on the face of it seem inconsistent with each other. But from a phenomenological perspective, the basic ideas are both tolerably clear. For current purposes, I emphasize that Descartes is affirming, first, that the self is embodied, and second, that bodily experience offers ways of being directly conscious of states of the embodied self. In other words, bodily experience is a form of self-consciousness.
Kant's Thesis: Subjective Self-Consciousness
In the Critique of Pure Reason, Kant discusses self-consciousness, primarily in the Transcendental Deduction of the Categories and in the Paralogisms of Pure Reason, both of which differ in the A and B editions. His discussion is far from perspicuous, but one theme that does emerge fairly clearly is a distinction standardly phrased (although not by Kant himself, I believe) as the distinction between consciousness of oneself as subject and consciousness of oneself as object. Kant himself does not use that precise terminology, to the best of my knowledge, but it is widely attributed to him, for example Cassam (Reference Cassam1997: 1–2) and Longuenesse (Reference Longuenesse2017: 26–32), both of whom I discuss further below.
Here is a passage from the beginning of the A version of the Paralogisms, where Kant makes clear what he means by consciousness of oneself as object:
Meanwhile, however pure from the empirical (from impressions of sense) it may be, it still serves to distinguish two kinds of objects through the nature of our powers of representation. I, as thinking, am an object of inner sense, and am called ‘soul’. That which is an object of outer sense is called ‘body’. (Kant [Reference Kant, Guyer and Wood1781/1787] 1998: A342; emphasis original)
So, consciousness of oneself qua object occurs through both inner sense and outer sense. By ‘inner sense’ Kant means what we would call introspection, while by ‘outer sense’ I am pretty confident that he means ordinary, outward-directed perception (and there is no evidence that he takes this to include what I am calling bodily experience).
In contrast, self-consciousness qua subject is limited to the proposition I think (which in the Transcendental Deduction he called apperception). The self-consciousness yielded by the proposition I think does not contain any element of intuition (through either inner or outer sense). It is a condition of the possibility of all thought and it expresses the thinker's consciousness of himself qua thinker. But the recurring theme in the Paralogisms is that nothing can be concluded from this about the nature of the self qua object. Consciousness of self as subject is purely formal. It yields only
the simple and in content for itself wholly empty representation I, of which one cannot even say that it is a concept, but a mere consciousness that accompanies every concept. Through this I, or He, or It (the thing), which thinks, nothing further is represented than a transcendental subject of thoughts = x. (Kant [Reference Kant, Guyer and Wood1781/1787] 1998: A345/346; emphasis original)
So, relative to this Kantian distinction, it is widely held that genuine self-consciousness needs to be consciousness of oneself as subject, because only that type of self-consciousness reveals, as it were, the essence of subjectivity. To be a self is to be a subject, and hence one can only be properly self-conscious if one is conscious of oneself qua subject. This is Kant's thesis of subjective self-consciousness.
Locke's Thesis: Bodily Self-Presentation
Locke tends to feature in discussions of the self primarily as a proponent of a psychological model of personal identity—that psychological continuity is what matters for the survival of the person, to the point that bodily continuity is not even required. Lockean body-swapping cases continue to be much discussed, although they now tend not to feature princes and cobblers. What I am calling Locke's thesis represents a much less frequently discussed aspect of Locke's thinking about self-consciousness. However, it is articulated in the very chapter (book 2, chapter 27) of the Essay Concerning Human Understanding where Locke offers his account of personal identity, and, in fact, it arguably underpins it.
Locke's account of personal identity is grounded in his general theory of self-consciousness (although he actually uses the term consciousness). This is because he starts from a distinctive account of what a person is:
[T]o find wherein personal Identity consists, we must consider what Person stands for; which, I think, is a thinking intelligent Being, that has reason and reflection, and can consider it self as it self, the same thinking thing in different times and places; which it does only by that consciousness, which is inseparable from thinking, and as it seems to me essential to it. (Locke [Reference Locke and Nidditch1689] 1975: 335; emphasis in original)
Paraphrasing anachronistically, for Locke a person is characterized by diachronic and synchronic self-awareness (since I take ‘considering it self as it self’ to be a form of self-awareness). Looking at this passage in isolation, one might think that this self-awareness is purely psychological, in the sense of having to do only with our occurrent perceptions and thoughts. But actually, his notion of self-awareness is much more expansive than that. A little later in the same section (§11) Locke makes plain that ‘consciousness’ extends to awareness of one's actions and of oneself as acting. And then, in the next section (§12), he justifies his claim that consciousness unites ‘distant Actions into the same Person’ by saying
[t]hat this is so, we have some kind of Evidence in our very Bodies, all of whose Particles, whilst vitally united to this same thinking conscious self, so that we feel when they are touch'd, and are affected by, and are conscious of good or harm that happens to them, are a part of our selves: i.e. of our thinking conscious self. Thus the Limbs of his Body is to every one a part of himself: He sympathizes and is concerned for them. Cut off an hand, and thereby separate it from that consciousness, we had of its Heat, Cold, and other Affections: and it is then no longer a part of that which is himself, any more than the remotest part of Matter. (Locke [Reference Locke and Nidditch1689] 1975: 336; emphasis in original)
Locke offers this observation as part of an argument that sameness of person does not necessarily require sameness of substance. But I offer it as an illustration of what I call Locke's thesis, which is that bodily experience presents us to ourselves as physical objects. (For a discussion of Locke that highlights this aspect of his thinking about the self and self-consciousness see Ayers Reference Ayers1991, particularly vol. 1, part 3 and vol. 2, chap. 22. I have addressed some of the metaphysical aspects of Locke's thinking about personal identity in Bermúdez Reference Bermúdez1996.)
Merleau-Ponty's Thesis: Bodily Distinctiveness
In Phenomenology of Perception Merleau-Ponty explores the ways in which, as he puts it, “the body is the vehicle of being-in-the-world” ([Reference Merleau-Ponty and Smith1945]1968: 82). A recurrent theme in the book is the contrast between two different ways in which we can engage with the body—which might be termed the first-person and third-person perspectives, although he tends to use the terminology of the for-itself and the in-itself. (The contrast is explored further in Bermúdez Reference Bermúdez, Smith and Thomasson2005.)
From a first-person perspective we experience the body qua for-itself, as something that structures and gives meaning to our engagements with the physical world. From a third-person perspective, in contrast, we treat the body qua in-itself, as a complex of muscles, bones and nerves that enters into causal interactions with other objects and that can in principle be studied and understood as one object among others, albeit a distinctive and highly complex object.
Although Merleau-Ponty engages deeply with scientific psychology at various points in Phenomenology of Perception, most prominently in fact in his discussions of the body, he imposes clear limits on what he thinks can be learned about the body from the third-person perspective of science. Science can tell us only about the objective body (the body as it is in-itself, made up of muscles, nerves, and bones). It cannot tell us about the phenomenal body, the body as it is for-itself: ‘It is never our objective body that we move, but our phenomenal body, and there is no mystery in that, since our body, as the potentiality of this or that part of the world, surges towards objects to be grasped and perceives them’ (Merleau-Ponty [Reference Merleau-Ponty and Smith1945]1968: 106).
It is the job of phenomenology to elucidate the lived experience of the phenomenal body. For Merleau-Ponty, it is a fundamental principle that the phenomenal body is experienced as a very distinctive type of object—completely differently from how we experience ordinary physical objects in the world. He writes,
[t]he body is not one more among external objects. It is neither tangible nor visible in so far as it is that which sees and touches. The body, therefore, is not one more among external objects, with the peculiarity of always being there. If it is permanent, the permanence is absolute and is the ground for the relative permanence of disappearing objects, real objects. The presence and absence of external objects are only variations within a field of primordial presence, a perceptual domain over which my body exercises power. Not only is the permanence of my body not a particular case of the permanence of external objects in the world, but the second cannot be understood except through the first: not only is the perspective of my body not a particular case of that of objects, but furthermore the presentation of objects in perspective cannot be understood except through the resistance of my body to all variations in perspective. (92)
This passage clearly illustrates what I am calling Merleau-Ponty's thesis of bodily distinctiveness: the way we encounter ourselves in bodily experience is fundamentally different from how we encounter non-bodily physical objects in outward-directed, exteroceptive perception.
II
On the face of it, the four theses are inconsistent. If bodily self-consciousness is true, then bodily experience is a form of self-consciousness. If subjective self-consciousness is true, then any bodily self-consciousness must incorporate consciousness of oneself as a subject. But this seems to be ruled out by objective bodily self-presentation, which states that in bodily experience we are presented with ourselves as physical objects, albeit (by bodily distinctiveness) in a fundamentally different way from how we encounter non-bodily physical objects in outward-directed, exteroceptive perception. The basic problem is that the very possibility of bodily self-consciousness seems to require simultaneously being conscious of oneself as a subject and as an object. As Schopenhauer remarked in a widely quoted passage from an unpublished manuscript, ‘that the subject should become an object for itself is the most monstrous contradiction ever thought of’ (cited in Janaway Reference Janaway1989: 120).
One of the people who have quoted Schopenhauer's words is Cassam, on the second page of Self and World. In that book, Cassam defends what he calls a materialistic conception of self-consciousness, which is the thesis that ‘we are aware of ourselves as abiding material substances, as physical objects among physical objects’ (Reference Cassam1997: 2). This is, in effect, the conjunction of bodily self-consciousness and objective bodily self-presentation. Cassam explicitly recognizes that defending any such materialistic conception of self-consciousness requires defusing Schopenhauer's ‘monstrous contradiction’, and he offers several proposals for showing that one can experience oneself both as a subject of experience and as a physical object among other physical objects. His view is that the bodily self is experienced as a subject-object—what he terms the ‘presented subject of thought and action’ (Cassam Reference Cassam1997: 142).
Cassam points to two salient features of the content of bodily experience. He emphasizes, first, that the body is experienced as the point of origin of egocentric space. As he points out (for example, in chapter 2), an important part of what it is to be a subject of experience is to experience the world from a particular perspective. And part of what it is to experience the world from a particular perspective is that one perceives it and acts within it on the basis of a uniquely individuating point of view that is centered in the body—either the body as a whole or relevant body parts. And that our point of view on the world is centered in the body is something that is actually given to us in bodily experience. Second, he stresses (particularly at 136–142) that we experience the boundaries of our body as our own boundaries—as the boundaries of our self. He writes, ‘[i]ntuitive awareness of one's own shape provides one with a sense of one's own boundaries qua subject; to have a sense of oneself as determinately shaped is therefore to have a concrete sense of where one ends and where the rest of the world begins’ (Cassam Reference Cassam1997: 142). This intuitive awareness of one's own shape is typically derived from bodily experience, particularly proprioception and the sense of touch.
In my own book The Paradox of Self-Consciousness, published around the same time as Cassam's, I also adopt a content-based strategy (Bermúdez Reference Bermúdez1998: particularly chapters 6 and 7). I highlight these two dimensions of bodily awareness (among others) and argue along the same lines that they count as genuine forms of self-consciousness, in the sense of making available a consciousness of oneself as subject.
Still, even if these two points are granted, someone might object that they fall short of explaining why bodily experience counts as providing consciousness of the self qua subject, and hence satisfying subjective self-consciousness. To experience the body as the geometrical origin of egocentric space and as circumscribing the limits of the self is not to experience oneself as a subject. It is simply to experience one's body as a distinctive and unique object. So, objective bodily self-presentation and bodily distinctiveness are both satisfied, but not subjective self-consciousness—which means that bodily experience cannot count as a form of self-consciousness.
In responding to this type of objection, both Cassam and I were persuaded by a line of reasoning originating with Sydney Shoemaker (Reference Shoemaker1968) and developed more fully by Gareth Evans in The Varieties of Reference (Reference Evans1982: particularly chapter 6) This line of reasoning emphasizes the importance of a certain type of judgment as expressive of genuine self-conscious thought. These are judgments that have the, by now much-discussed, property of being immune to error through misidentification relative to the first-person pronoun. Judgments have the immunity-to-error property relative to particular sets of grounds or evidence. They are judgments such that there is no gap between having information that a particular property is being instantiated and having information that that property is instantiated in oneself. Since each person's bodily experience provides information only about that person's own body, it follows that judgments made on the basis of bodily experience have the immunity-to-error property.
Both Cassam and I thought that judgments with the property of immunity to error through misidentification are the key to elucidating how bodily experience can provide consciousness of the self qua subject. Here is how he puts the point: ‘for a form of awareness to count as awareness of the self qua subject, it must be a form of awareness on the basis of which it is possible to make first-person statements that are, in this sense, immune to error through misidentification’ (Cassam Reference Cassam1997: 60).
So, putting it all together, the content of bodily experience reveals an awareness of the body as a distinctive type of object (as the geometrical origin of the subject's point of view and as circumscribing the limits of the self). And because bodily experience can ground judgments with the property of immunity to error through misidentification, it counts as a consciousness of the self qua subject. In this way, bodily self-consciousness, subjective self-consciousness, objective bodily self-presentation, and bodily distinctiveness are all satisfied.
I must admit to having been convinced by this general, content-based strategy until quite recently. However, I now think that simply appealing to the property of immunity to error through misidentification is by no means as effective as I thought it was. I came to see this thanks to some observations on the immunity-to-error property made by Longuenesse (Reference Longuenesse2017) in the context of Kant's ‘I think’. Longuenesse is happy to accept that the Kantian ‘I think’ has the property of immunity to error through misidentification. That is to say, there is no gap between judging (on the basis of introspection) that there is an occurrent act of thinking going on and judging that one is oneself thinking. To be (introspectively) conscious of thinking is just to be conscious that one is oneself thinking. What is particularly interesting, though, is the explanation she offers for this (on Kant's behalf):
Kant's view is that the capacity to refer to myself via the concept and word ‘I’ and the process of connecting my representations in such a way that I can form thoughts (combinations of concepts) and account for my reasons, are mutually conditioning. In other words, having available the concept ‘I’ and being engaged in the kind of unifying activity that makes possible reason-giving thought are mutually conditioning. For this reason it is nonsensical to suppose that I could wonder ‘Someone is thinking [that this proof is valid, that this is a tree, that it will rain] but is it me?’ In this case, and in this case alone, the reason there is no place for this question is not that the judgment is justified by a particular kind of awareness. Rather the reason is the very nature of thinking. (Longuenesse Reference Longuenesse2017: 29)
So, for Longuenesse, the fact that Kant's ‘I think’ has the property of immunity to error through misidentification is really a relatively superficial reflection of a more fundamental connection between the nature of thinking and the capacity for self-reference. Moreover, and this is the key point for present purposes, it is only when this more fundamental connection holds that Longuenesse thinks it is appropriate to talk of being conscious of oneself as a subject. She concludes that only ‘I think’ judgments qualify as expressing consciousness of oneself as subject sensu stricto.
Her reasoning is based on a persuasive account of what it is to be conscious of oneself qua subject. That account, in turn, starts from the nature of the rule fixing self-reference. The first-person pronoun I, and its equivalent at the level of thought, refers to the speaker of the sentence in which I is used or to the author the relevant thought. Longuenesse uses this reference rule for I to define what it is to be conscious of onself qua subject. Here is a formulation of her definition:
Self-Consciousness qua subject
To be conscious of oneself qua subject is to be conscious of oneself as engaged in the activity (or activities) in virtue of which something counts as the referent of I.
This is an appealing thought, which might be justified along the following lines. To be a subject is to be capable of self-reference. Self-reference is available only to creatures able to engage in the activity or activities in virtue of which something counts as the referent of I. Putting those together, the activities in virtue of which something counts as the referent of I are constitutive of subjecthood. Since it is natural to think that to be conscious of oneself qua subject is to be conscious of oneself as engaged in the activities that constitute subjecthood, Longuenesse's definition follows immediately.
Now, with this definition in view, I go back to Longuenesse's earlier discussion of the ‘I think’. Her basic claim there is that the activities that constitute subjecthood in virtue of being what makes one the referent of I are precisely the activities of thinking that are expressed in ‘I think’ judgments. Since the first person concept refers to the thinker of this episode of thinking, it is solely in virtue of the activity of thinking that one succeeds in referring to oneself. And so, the judgments that express self-consciousness qua subject are all and only those that express the activity of thinking—the Kantian ‘I think’. Judgments that are based on bodily experience fail to qualify, even if they have the property of immunity to error through misidentification. Thus,
[s]omething counts as the referent of ‘I’ (= the current thinker of the thought and speaker of the sentence, ‘I think the proof is valid’, ‘I think this is a tree’) precisely in virtue of being engaged in the activity that is predicated of it in the proposition ‘I think that p’: the activity of thinking, premised on an activity of binding for thinking. In contrast, it is not of course the case that something counts as the referent of ‘I’ (= the current thinker of the thought ‘I am sitting cross-legged’) just in virtue of her sitting cross-legged. (Longuenesse Reference Longuenesse2017: 31)
Longuenesse concludes that bodily experience is properly described as consciousness of oneself qua object. The fact that judgments based upon bodily experience have the property of immunity to error through misidentification is irrelevant. And so subjective self-consciousness is not satisfied.
III
To take stock, in the previous section I looked at one strategy for reconciling the four principles and doing justice to the thought that bodily experience is a form of self-consciousness. That strategy highlights aspects of the content of bodily experience and how that content presents the bodily self—as the geometrical origin of the self's perceptual point of view on the world and as fixing the boundaries of the self. As discussed above, though, on one plausible way of thinking about what it is to be conscious of oneself qua subject, the content-based approach fails to satisfy Kant's thesis of subjective self-consciousness.
One possible response would be to reject the constraints that Longuenesse places upon self-consciousness. As discussed above, her account is grounded in the reference rule for the first person pronoun I and she clearly takes self-consciousness to be coeval with the capacity for self-reference. That is certainly a mainstream view, but it can be challenged. I challenged it myself in The Paradox of Self-Consciousness (Bermúdez Reference Bermúdez1998), where I drew attention to multiple forms of nonconceptual and prelinguistic self-consciousness, arguing that those more primitive forms of self-consciousness are the foundation for full-fledged linguistic self-consciousness. If we locate those aspects of bodily experience discussed in the previous section at the nonconceptual level, then Longuenesse's points cease to be directly applicable. The conditions for successful self-reference are not relevant to nonconceptual self-consciousness.
The nonconceptual approach still seems to me correct as an analysis of the structure of self-consciousness in general. However, shifting from the conceptual level to the nonconceptual level is at best question begging in the present context. The issue is not whether there is such a thing as nonconceptual self-consciousness. Nor it is whether the two aspects of the content of bodily experience that I have been discussing are indeed forms of nonconceptual rather than conceptual content. Rather, the issue is whether, simply in virtue of its content, bodily experience yields consciousness of the self qua subject. And saying that that content is not conceptual but nonconceptual does not really address that issue. Accordingly, a more promising strategy would be to accept Longuenesse's strict constraints on what it is to be conscious of oneself qua subject and then to try to show how bodily experience actually satisfies them. That is what I attempt in the remainder of this discussion.
Here is a formulation of Longuenesse's definition again:
Self-Consciousness qua subject
To be conscious of oneself qua subject is to be conscious of oneself as engaged in the activity (or activities) in virtue of which something counts as the referent of I.
As Longuenesse herself understands the matter, that in virtue of which something counts as the referent of I is engaging in the activity of thought. In particular, she thinks it both necessary and sufficient that the subject engage in the activities of combining concepts and binding representations that Kant called synthesis. There is, she maintains on Kant's behalf, a two-way dependence between the activity of synthesis and the capacity self-referentially to ascribe thoughts to oneself. This two-way dependence reflects the fact that both are grounded in ‘the transcendental unity of apperception: the unity of self-consciousness that makes possible both our synthesizing representations into conceptualizable wholes (what I call here “binding”) and our ascribing thoughts to ourselves in the proposition “I think”’(Longuenesse Reference Longuenesse2017: 31).
While the necessity claim seems very plausible, the sufficiency claim is less so. Below I sketch out an alternative way of thinking about the conditions of successful self-reference. On this alternative way of looking at the matter, bodily experience will satisfy subjective self-consciousness, even by Longuenesse's criterion for what it is to be conscious of oneself qua subject.
A brief aside before pressing on. Kant's own thinking about the self and self-consciousness is ineliminably constrained by his transcendental idealism. As he emphasizes throughout the Paralogisms of Pure Reason, we cannot know the self in-itself, and to think that we can is to fall into trap of rational psychology. It is not very surprising, therefore, that Kant's perspective on what it is to be a thinking self should be so narrow, confined as it is to the transcendental unity of apperception (the ‘transcendental subject of thought = x’). Plainly, however, for those of us unimpressed by any form of antirealism, there is no ex ante need to follow Kant in thinking about the self in such an etiolated way. We can, in fact, engage in a characteristically Kantian activity that Kant's transcendental idealism made it impossible for him to apply in thinking about self-consciousness in a more substantial sense. We can explore what are effectively transcendental arguments about the necessary conditions for successful self-reference, and for self-consciousness in general. Since Kant's transcendental idealism forced him to such a formal conception of the thinking self, he could not but have an equally formal conception of self-reference and self-consciousness, which presents somewhat impoverished starting-points for transcendental argumentation. But we do not have to follow him in this. There are good precedents for the general strategy of pursuing transcendental arguments without transcendental idealism, which has been most prominently promoted by Peter Strawson, both in The Bounds of Sense (Reference Strawson1975, his commentary on the Critique of Pure Reason), and in Individuals (Reference Strawson1959), where he puts the theory into practice.
IV
According to Longuenesse, to be conscious of oneself qua subject is to be conscious of oneself as engaged in the activity (or activities) in virtue of which something counts as the referent of I. How this criterion is to be applied depends, of course, on which activity (or activities) are deemed to be those in virtue of which something counts as the referent of I. Longuenesse takes the following to be the reference rule for I: ‘“I” is a word or concept that refers, in any instance of its use, to the author of the thought or the speaker of the sentence in which “I” is being used’ (Reference Longuenesse2017: 23). Since Longuenesse takes thought to be primary and language use secondary, she concludes that the activity of thinking is that in virtue of which something counts as the referent of I, and hence that we are conscious of ourselves qua subject to the extent that we are conscious of ourselves as engaged in the activity of thinking.
As I said earlier, I am happy to accept Longuenesse's basic characterization of what it is to be aware of oneself qua subject. However, I would put pressure on two aspects of the way she applies that basic characterization:
1. I think that her account of what it is to be a thinker is too narrow. I develop below a much more expansionary conception of what it is to be a thinker, such that bodily experience counts as a way of being conscious of oneself as a thinker, and hence qua subject. This more expansionary conception emphasizes the role of bodily agency in thinking.
2. It seems to me also that her account of the requirements of successful self-reference leaves out a crucial dimension of the reference rule governing the first person in thought. In particular, successful self-reference requires the thinker to understand the first person concept (to grasp the sense of I, as I prefer to put it). When that additional requirement is factored in, bodily experience comes into the picture as a form of consciousness of oneself qua subject.
To elucidate the first point, I start with a passage where Longuenesse explains her own narrow conception of what it is to be the current thinker of a thought. She considers the thought ‘I think this is a tree’. Here is what she says about it:
The predicate ‘––think that this is a tree’ is a concept that refers to the activity of reviewing the reasons for asserting that this is a tree. The term in the argument-place, ‘I’, is a concept that refers to the current thinker of the thought ‘I think this is a tree’, namely, the current agent of the activity of reviewing the reasons for asserting ‘this is a tree’. Nothing further is needed for ‘I’ to meaningfully refer to an individual: the current thinker of the thought ‘I think this is a tree’. (Reference Longuenesse2017: 27)
The key claim is about how to characterize what it is to be the current thinker of the thought. If all that there is to being the current thinker of the thought is being the current agent of the activity of reviewing reasons for thinking that some presented object is a tree, then all that there is to being conscious of oneself qua subject is being conscious of oneself as a mental agent in the way that she describes. Longuenesse accepts, of course, that one cannot think (demonstratively) that something is a tree without being a bodily agent—one has to perceive it, after all. But that has nothing to do with the deployment of I. It is, she claims, a function of the predicative part of the thought, not the self-referential part.
The real issue, however, is what the requirements are on being a mental agent. Longuenesse thinks that mental agency is disembodied (even though the mental agent herself is actually an embodied subject, considered in the round, as it were). I am not convinced by this, however. Longuenesse herself talks about reviewing the reasons for the judgment as a paradigm example of mental agency, and she accepts that reviewing reasons is not a purely passive process (with the reasons parading by on some sort of inner screen, while the mental agent assigns scores). But she does insist that the activity of reviewing reasons is purely mental. The example to which she returns on several occasions is reviewing a proof, evaluating each inferential step, filling in any gaps, and so forth. This is the most plausible example to make her case, in the sense that mathematics is about as disembodied a form of thinking as it is possible to get. But even here things are far from clear-cut. Reviewing proofs often involves writing formulas down, sketching figures, and performing various other types of mathematical construction. Surely being able to do these things is part of the mental agency involved in being able to review mathematical proofs.
And actually, these time-honored techniques may just be the tip of the iceberg. There is a thriving movement in mathematics education studying how different types of action can enhance mathematical understanding. This in turn opens up opportunities to use technologies such as touch-screen devices, 3D printers, location sensors, gyroscopes, and so on. The aim is to allow students actively to construct mathematical experiences that help them evaluate and assess mathematical argumentation. (For a review, see Tran, Smith, and Buschkuehl Reference Tran, Smith and Buschkuehl2017.) For what it is worth, I have come to appreciate the pedagogical utility of this general approach through my own experience teaching logic online using the highly visual and interactive software that comes with Language, Proof, and Logic (Barker-Plummer, Barwise, and Etchemendy Reference Barker-Plummer, Barwise and Etchemendy2011).
In any event, mathematical thinking is probably not the best paradigm for thinking about what it is to be a thinker, so I go back to demonstrative thoughts about trees. If I see something that I think might be a tree, but I am not sure, then there are many things that I might do. I might move closer to get a better look. I might change position to get away from the direct sunlight. Or if I happen to be in an area known for its papier-mâché trees, I might have to go right up to it to check that what I am looking at is really organic. It might seem a little ponderous to describe these as ways of reviewing my reasons for thinking that I am seeing a tree, but that is really what I am doing. I am exercising mental agency through exercising bodily agency.
The active nature of visual perception is nicely conveyed in a passage from the computer scientist Ruzena Bajcsy. Bajcsy is bemoaning the dominant approach to robotics and computer vision at the time she was writing:
Most past and present work in machine perception has involved extensive static analysis of passively sampled data. However, it should be axiomatic that perception is not passive, but active. Perceptual activity is exploratory, probing, searching; percepts do not simply fall onto sensors as rain falls onto ground. We do not just see, we look. And in the course, our pupils adjust to the level of illumination, our eyes bring the world into sharp focus, our eyes converge or diverge, we move our heads or change our position to get a better view of something, and sometimes we even put on spectacles. This adaptiveness is crucial for survival in an uncertain and generally unfriendly world, as millennia of experiments with different perceptual organizations have clearly demonstrated. (Reference Bajcsy1988: 996; see also Bajcsy, Aloimonos, and Tsotsos Reference Bajcsy, Aloimonos and Tsotsos2018)
The phrase ‘static analysis of passively sampled data’ is telling. Leaving aside the question of how well it describes research in machine vision in 1988, I think it accurately characterizes a broadly Kantian way of thinking about perception, and cognition more generally. The perceiver/thinker is presented in receptivity with the manifold of intuition, which is then interpreted and has order imposed upon it through various acts of synthesis and combination (this being the dimension of spontaneity).
Against this Kantian picture, I propose taking seriously not just the idea of active perception, as so eloquently described by Bajcsy but also its extension to thought in general, so that we think about thinking as an exercise of agency that goes beyond purely mental agency. A more sustained presentation of this idea would emphasize the following points:
• Much of our thinking is at the service of, and directed toward, our goals and ends as practical reasoners. As Aristotle pointed out, practical reasoning eventuates in bodily action.
• To be a thinker is to be capable of, and committed to, seeking out new information relevant to the propositions one is considering. The key point for present purposes is that this active information seeking is something that one does as an embodied agent.
• For creatures capable of self-reference, being a thinker is inextricably linked to being a speaker and being a listener, both of which are embodied activities. Think about the role of gesture in communication, for example, and how spoken language is complemented by body language.
• The canvas of thought includes emotions and moods, as well as ‘pure’ thinking. Both emotions and moods are embodied states that involve distinctive patterns of action.
Once we take this broader perspective on what it is to be a thinker, it seems plausible that there will have to be a correlatively broader and embodied perspective on what it is to be conscious of oneself as the current thinker of a particular thought. And that, in turn, will create a space for bodily experience to count as consciousness of oneself qua subject.
This line of argument does not fall prey to the objection already considered from Longuenesse—namely, that embodiment comes in to the picture only as a function of the predicative component of the thought, not the self-attributive I component. The argument is not that we need to think of ourselves as embodied when we think of ourselves as engaged in practical reasoning, or as in a particular mood or emotion. Rather, it is that we cannot think of ourselves as engaged in any type of thinking unless we think of ourselves as thinkers in a rich sense that includes affectivity and embodied agency. This has to do not with the predicative component of the thought, but with the I component.
We can reach the same result, I think, by reflecting on the requirements of successful self-reference (and this is the second of the two arguments referenced above). Here again is the fundamental reference rule for I, as Longuenesse formulates it: ‘“I” is a word or concept that refers, in any instance of its use, to the author of the thought or the speaker of the sentence in which “I” is being used” (Reference Longuenesse2017: 23). On the face of it, however, this characterization leaves out something important. One of the lessons to be drawn from G. E. M. Anscombe's rather tortured example in ‘The first person’ (Anscombe Reference Anscombe and Guttenplan1975) of the people who have a self-referential name ‘N’ that only they can see on the inside of their wrist is that successful self-reference using I requires a (self-conscious) intention to refer to oneself—plausibly, that is the key difference between ‘N’ and I.
One consequence of this, which I emphasized in The Paradox of Self-Consciousness, is that the capacity for self-reference cannot be used to provide a reductive explanation of the capacity for self-conscious thought. In the present context, though, I want to make a different point. If successful self-reference requires a (self-conscious) intention to refer to oneself, then it requires an understanding of the first person concept. It requires a mastery, in other words, of the sense of I. I claim, however (contra Longuenesse), that one cannot grasp the sense of I without thinking of oneself as embodied.
To cut a long story short, I agree with Gareth Evans that self-referential thought and language rest upon capacities for self-location. This holds, not because of any metaphysical claims about the essentially embodied nature of human existence (true though those claims certainly are), but rather as a requirement upon being able to grasp thoughts about oneself where what are being self-ascribed are predicates for which we do not have distinctive ways of knowing that they apply to us. Here is a well-known passage where Evans makes the point:
So long as we focus on judgements which a person might make about himself on the basis of the relevant ways of gaining knowledge, the inadequacy may not strike us. A subject's knowledge of what it is for the thought ‘I am in pain’ to be true may appear to be exhausted by his capacity to decide, simply upon the basis of how he feels, whether or not it is true—and similarly in the case of all the other ways of gaining knowledge about ourselves. However, our view of ourselves is not Idealistic: we are perfectly capable of grasping propositions about ourselves which we are quite incapable of deciding, or even offering grounds for. I can grasp the thought that I was breast-fed, for example, or that I was unhappy on my first birthday, or that I tossed and turned in my sleep last night, or that I shall be dragged unconscious through the streets of Chicago, or that I shall die. (Evans Reference Evans1982: 208–9)
Evans's point is that in order to grasp a certain class of thoughts about oneself, one needs to be able to think of oneself from a third-person point of view—as an element in the objective order. And that in turn requires being able to think of oneself as an enduring physical object following a single, continuous trajectory through space-time.
To see what he is getting at, consider the contrast between these two thoughts. The first is the introspective thought ‘I am happy now’. The second is Evans's own example: ‘I was unhappy on my first birthday’. Since judgments based on introspection have the property of immunity to error through misidentification, there is no gap between knowing that happiness is instantiated (as it were), and knowing that it is instantiated in oneself. For that reason, thinking this thought requires only a very thin conception of oneself. In contrast, for me to think that I was unhappy on my first birthday is for me to be thinking about a particular thing, historically continuous with the thing that I now am. The only way for me to think about myself as I was on my first birthday is as a physical object. Since I have no epistemic access to my state of mind then, I am thinking about myself third-personally. What I conceptualize is a ‘mewling and puking babe’ historically continuous with the physical object I now am—and my manner of thinking about it is little different from how I might think, for example, about the mewling and puking babe sitting behind me on an aircraft.
I emphasize again that the issue here is to do with the self-attribution component of the thought, rather than the predicative component (despite what Longuenesse says in her discussion of Evans [Reference Longuenesse2017: §2.14). The argument is transcendental in form. Somewhat compressed, it goes like this. In order to be able to refer to oneself successfully in thought one needs to have a self-conception that will allow one to think about oneself from a third person perspective. This self-conception must be of oneself as embodied. (For more details, see Evans Reference Evans1982: chap. 6; Bermúdez Reference Bermúdez2017a).
Each of these two lines of argument leads, therefore, to the conclusion that the capacity for self-reference requires conceiving of oneself as embodied. This puts us in a position to see why bodily experience qualifies as self-consciousness qua subject. Here is the reasoning. According to Longuenesse, to be conscious of oneself qua subject is to be conscious of oneself as engaged in the activity (or activities) in virtue of which something counts as the referent of I. Something counts as the referent of I in virtue of engaging in the activity of thinking and in virtue of self-consciously referring to itself. So consciousness of oneself qua subject will have to include consciousness of oneself as a subject capable of thinking and capable of self-conscious self-reference. This is where bodily experience comes into the picture. For most people, bodily experience is a fundamental way of being conscious of oneself as a thinker (in the broad sense sketched out above), and of being conscious of oneself as a physical object persisting over time. And for those very reasons bodily experience counts as a form of consciousness of oneself qua subject.
I should emphasize what I am not saying, however. My claim here is not that bodily experience is the only way of being conscious of oneself in these two basic ways. Nor even that it is a necessary way of being conscious of oneself in those ways. My point is simply that, to the extent that bodily experience is a way of being conscious of oneself as a thinker and as capable of self-reference, to that extent it counts as a form of consciousness of oneself qua subject.
V
In the previous section I argued that bodily experience qualifies as a form of consciousness of oneself qua subject, thus demonstrating the consistency of three of the four principles with which I began: bodily self-consciousness, subjective self-consciousness, and objective bodily self-presentation. This final section turns to the fourth principle, bodily distinctiveness. In the space remaining I will speak to one aspect of how the way we encounter ourselves in bodily experience is fundamentally different from how we encounter non-bodily physical objects in outward-directed, exteroceptive perception.
In previous work (most recently Bermúdez Reference Bermúdez, de Vignemont and Alsmith2017b, which explicitly addresses important recent experimental work on illusions of bodily awareness) I have stressed two basic features of the phenomenology of bodily awareness:
Boundedness
Bodily events are experienced within the experienced body (a circumscribed body-shaped volume whose boundaries define the limits of the self).
Connectedness
The spatial location of a bodily event is experienced relative to the disposition of the body as a whole.
These two principles jointly illustrate how individual bodily experiences incorporate an ongoing awareness of the body's limits and moment-to-moment layout, and the relation is reciprocal, since individual bodily experiences are an important element in generating that ongoing awareness.
Boundedness and connectedness each emphasize different aspects of the spatial dimension of bodily experiences. Boundedness emphasizes the relation to the experienced spatial bounds of the body, while connectedness emphasizes the relation to the body's experienced layout. Since there are not analogs of either for exteroceptive perception in any sensory modality, there seems to be something very distinctive about how the spatial location of bodily events is experienced.
In order to capture this distinctiveness, I propose, we need to model how the space of the body is experienced in bodily awareness very differently from standard models of how the space of the external world is experienced in vision and other exteroceptive modalities. When the experienced spatiality of the body is discussed, it is standardly conceptualized in terms of a Cartesian frame with three axes, corresponding to the frontal, saggital, and transverse planes, with the body's center of mass serving as the origin of the coordinate system. This way of thinking about the space of the body, however, cannot do justice to either boundedness or connectedness. There is no sense in which spatial locations that fall within the limits of the body are privileged (as boundedness requires), and since each point is completely independent of every other there is no way in which it can accommodate connectedness.
My proposed alternative model distinguishes two aspects of experienced bodily location, which I term A-location and B-location. Both presuppose a general map of the body as a relatively immoveable torso connected by joints to moveable body parts.
A-location
To experience a bodily event (a pain, say) at a particular A-location is to experience it relative to such a map of the body, without taking into account the body's actual position. For this reason experienced A-locations can only be experienced within the boundaries of the body. So, A-location corresponds most closely to boundedness.
B-location
The B-location of a bodily event is its A-location calibrated relative to the position of the rest of the body. A-location specifies a bodily event within a particular body-part, while B-location fixes the location of that body-part in terms of the angles of the joints that lie between it and the immoveable torso. So, the B-location of the pain in the ball of my foot is its A-location within my foot, supplemented by specifying the angles of the foot relative to the lower leg, and the lower leg relative to the upper leg.
This model of A-location and B-location as non-Cartesian, because the spatial coordinate system that I am envisaging does not have a Cartesian frame of reference. As there is no origin, certain standard metric concepts do not apply. With respect to the frame of reference of vision, for example, it makes perfectly good sense to think of one thing being nearer (to the origin) than another. But there is no ‘nearer’ and ‘further away’ in the space of the body. As Merleau-Ponty evocatively puts it (on somewhat different grounds), ‘the outline of my body is a frontier that ordinary spatial relations do not cross’ (Merleau-Ponty [1945]Reference Merleau-Ponty and Smith1968: 98).
The fact that we experience the spatiality of the body in this fundamentally non-Cartesian way is part of the explanation of bodily distinctiveness. Moreover, to connect back to the discussion in the previous section, this way of thinking about the experienced spatiality of the body ties it very closely to the subject's awareness of the body's potentiality for action. Muscles move limbs by changing joint angles, and so the experienced B-location of bodily events conveys critical information for action-planning and motor control. This means that bodily locations are experienced in a way that makes manifest to the subject their own bodily agency, which I argued above is part of what it is to be aware of oneself as a thinker.
VI
By way of conclusion, therefore, here again are the four theses I have been discussing.
Bodily Self-Consciousness (Descartes's thesis)
Bodily experience is a form of self-consciousness. We are conscious of ourselves through experiencing our own bodies.
Subjective Self-Consciousness (Kant's thesis)
Nothing can count as a genuine form of self-consciousness unless it is consciousness of oneself as a subject.
Objective Bodily Self-Presentation (Locke's thesis)
In bodily experience we are presented with ourselves as physical objects.
Bodily Distinctiveness (Merleau-Ponty's thesis)
The way we encounter ourselves in bodily experience is fundamentally different from how we encounter non-bodily physical objects in outward-directed, exteroceptive perception.
I claim that they are all true.
