Unified Consciousness and the Self
Director, Cognitive Science PhD Programme
Ottawa, Ontario K1S 5B6
I am in virtually complete sympathy with Galen Strawson's conclusions in 'The Self'. He takes a careful, measured approach to a topic that lends itself all too easily to speculation and intellectual extravaganzas. The results he achieves are for the most part balanced and plausible. I even have a lot of sympathy with his claim that a memory-produced sense of continuity across time is less central to selfhood than many philosophers think, though I will argue that he goes too far in the opposite direction.
Thus my purpose in these comments for the most part is not to criticise his conclusions. Instead, I want to look at certain aspects of the framework of argument and observation that he uses to reach these conclusions. Perhaps additional precision in some aspects of it would add strengthen to his analysis.
In particular, I want to look at elements (3) and (4) in his list of features that we conceive a self to possess. (3) concerns synchronic singularity, i.e., being one mental being at a time, and (4) concerns diachronic singularity, i.e., being one mental being over time. I will argue that the spirit of Strawson's claims about (3) and (4) is supportable but that the letter of them is flawed, due mainly to a failure to distinguish singleness of self from a self being unified. The feature relevant to Strawson's overall analysis is being unified, not being singular. The relationship between unity and singularity is interesting. Before we take up these issues, I must do some stage-setting..
1. Phenomenology and what the self is like
Strawson puts phenomenology, that is, how the self appears to itself, at the centre of his picture of what the self is like. Should we let how things appear to us influence our view of how things are in the way that he does? The move is unusual. It is not used in science and if it had been, we would never have discovered all those laws where how things appear is quite different from how they are. ("Feathers and cannonballs fall at the same rate in a vacuum? Nonsense!" - that is what appearances would suggest. Yet for all that, they do.) Nonetheless, I think that Strawson is right to put phenomenology at the centre of what the self is like. Here is why.
What is 'the self'? In my view, my 'self' is simply me, specifically, me as I am aware of myself. Even more specifically, what I call my self is myself as I appear to myself in those representations of myself in which I am aware of myself 'from the inside', to use an old but still suggestive metaphor of Sydney Shoemaker's. Here is the distinction I have in mind. To be aware of the colour of my hair, I have to use my eyes (or, in extreme cases, hear or read what someone else's eyes have seen). But to know what I want to say in this sentence - and also what it feels like in my arm muscles to be typing this sentence into the keyboard of my computer - I do not need to use any external senses. These are things that I could be aware of in myself even if I were in a sensory deprivation chamber.
What is the difference between awareness 'from the inside' (merely a useful metaphor, of course) and awareness by using an external sense. This is a complicated matter and we can only rough out the appropriate differences here. Take my desires for the just mentioned sentence that I am typing, that it come out clearly, that it say what I want it to say, etc. There are two ways in which I could become aware of these desires:
1. by having the desires, and,
2. by encountering evidence of various kinds indicating that I have them.
The kind of evidence envisaged in (2) might include: how I am behaving, others' judgments about what is motivating my behaviour, (perhaps someday) readouts from brain scanning devises, and so on and so forth. However, it is (1) that is relevant to the self.
More specifically, it is when I am experiencing my thoughts by thinking them, my desires by feeling them, my perceptions by having them, my actions by doing them, etc., that I am directly aware of my self (or my Self if you prefer). I could and others do infer that I have a self by observing what I say and do, but no one can be directly (i.e., non-inferentially) aware of my self via such observations, not anyone else and not me either. To be directly aware of my self, I must have the kind of awareness of it that I get when feeling desires, thinking thoughts, having perceptions, etc., makes me aware of myself doing those things. That is to say, my self is what I am aware of when I am aware of myself and my mental states and activities by having those states and doing those activities. We called this 'awareness from the inside'. Awareness from the inside is what gives me direct awareness of my self. Other ways of encountering myself do not.
If so, phenomenology has to be central to what the self is, more central than it is to the nature of virtually anything else. If the self is what I am aware of when I am aware of myself from the inside, then what each of us is aware of as the self will be absolutely central to delineating what the self is: to a large degree, a self simply is what one is aware of when one is aware of oneself from the inside. For this reason, Strawson is also right when he puts the self's mental properties at the centre of determining what it is like. Even though for a materialist, the self has to be brain states of some particular kind, it is the self's introspectible properties that will define what it is like and properties described in or requiring description in the language of the brain will be peripheral.
2. The self - just results or also activities?
If the self is simply what one is aware of as oneself when one is aware of oneself from the inside, a number of features of the self become clear. First, what the self is will be largely an empirical question. The reason is that what any given person actually happens to be aware of when aware of his or her own self from the inside will be an empirical question, indeed, something that is likely to vary and even vary enormously from person to person. A completely task-oriented athlete is likely to be aware of very different and probably many fewer things in herself than a psychologically-minded novelist who has undergone years of psychoanalysis.
If the 'contents of the self' simply is an empirical matter in this way, that would hold implications for any attempt such as Strawson's to lay down a list of essential features of selfhood (p. 408).(1) There are apt to be severe limitations on such a project, limitations arising from the large element of sheer variability in the contents of the self from person to person. Strawson makes some allowance for this feature of selfhood in (8): selves are things that have a certain character or personality. I fear, however, that the limitations on the possibility of any essentialist analysis may reach further than he recognizes.
If we add a further element to Strawson's picture of the self, however, we may be able to save his essentialist analysis. The additional element would be this: as well as the self being what I am aware of when I am aware of myself from the inside, we need to add that it also includes this and other acts of awareness. On this proposal, the self would be both what one is aware of as oneself and the activities of this being of which one is aware. This suggestion has some plausibility. In our conception of the self, we view it as not just an inert residue of psychological properties. We also view it as being active - its actions can be forceful, inauthentic, kind, mean, and so on and so forth. If so, it is plausible to urge that the activities of the self of which we are aware, including its activities of being aware of itself, are part of the self.(2)
Now, if the self is activities as well as psychological states (by 'psychological states', I mean something like Strawson's (8), having a personality - i.e., things like traits, capacities, dispositions to behave, etc.), then we can resuscitate Strawson's essentialist analysis. For even if the contents of the self can vary widely from person to person, it is likely that the activities characteristic of selves will take much the same general form from person to person. And these common form is likely to be essential to them being the activities they are. In particular and turning to the details of Strawson's analysis, such a self would have to be:
3. The self: unity versus singularity
We have looked at three of Strawson's eight features (with a side glance at a fourth, namely, , personality). In addition, as Kant argued already over 200 years ago, the activities characteristic of the self and the contents which these activities generate will have a certain kind of unity - or perhaps even a number of kinds of unity. This brings us to his (3) and (4):
A self is,
Strawson accepts that (3) is an essential feature of selves but has doubts about (4). He recognizes that (4) is very commonly found in selves as we know them, of course, but doubts whether singularity over time is essential to being a self (p. 409).
Here is my view. Strawson is right to accept (3) though my reasons for accepting it are a bit different from his (as we will see shortly). He is also right to reject (4) as he words it. However, there is a different though related version of (4) that he should not reject.
Return to the distinction mentioned earlier between singularity and unity. Strawson spends a lot of time on principles of unity and I agree with everything he says about them but he may not have noticed that there is a particular kind of unity that is distinctive to conscious cognition (and perhaps to some nonconscious forms of cognition, too, though I cannot go into that here). Let us call it, after Kant, unity of consciousness:
The unity of consciousness =df. an act of representing in which a number of objects and/or representation of them are combined in such a way that to be aware of any of these items is to be aware of others of these items as connected to it.
I am using the broadest possible notion of representation here, one in which a representation is any state or activity of a person(3) that gives that person epistemic access to an object, where: 'object' is also used in the broadest possible sense to mean physical things, ideas, non-existent entities (e.g., Santa Claus), abstract objects (e.g. numbers) and so on; the access may or may not be conscious; and, the object is experienced as a fully presented instance of something (no further representations are needed to be aware of it as a full object).
Put more simply but less accurately, unity of consciousness consists in being aware of a number of objects and/or the representations of them at the same time: I am aware, not just of A, and, separately, of B, and, separately, of C, but of A-and-B-and-C, as we say, 'all together'. That is to say, to be aware of any of them is to be aware of others of them. I think that this is what we mean, or one of the things that we mean, when we talk about being aware of a number of items as the contents or object of a single representation.
I have heard it objected that it is not clear that we can be aware of more than one item at a time, something that my account of the unity of consciousness requires. In fact, this seems very clear: right now I am consciously looking at a screen, monitoring my fingers on the keyboard, thinking about what I want to say, feeling the heat and humidity of the current heat wave, wishing that these comments were finished, and a number of other things - all simultaneously. So there is no objection to my account of the unity of consciousness here. This unity, delineated as I have delineated it, is, and so far as I can see must be, a feature of the activities of awareness that are part of what we take a self to be (see Section 2).(4)
A similar notion is at work in Strawson's account, though perhaps not so precisely articulated (p. 416). Such unified awareness is closely related to our sense of what a single self or mind consists in. Here is evidence for a link between unity and singularity. When we suspect a duality or plurality of selves (as in brain bisection and dissociative identity disorder cases), it is precisely the absence of such unity that gives rise to the suspicion. In these cases, the evidence suggests that the representations in such persons can be divided into two groups such that there is a subject of experience for whom it is true that: the items in one group are unified in a single representation, but the items of the other group are not part of that unified consciousness. Yet the items in the second group often make up a substantial group and there is often evidence for a second subject, 'secondness' here being based on the notion that items in the second group are unified in a single representation but items in the first group are not included in it.
Strawson himself claims that "explicitly self-conscious thought" must experience the self as synchronically single. (p. 416) and I think that he is right. I have examined this issue (and the others under discussion) elsewhere but let me note just one argument.(5) Suppose I am told that I am to undergo fission into two persons (two minds, two selves, two souls - the terminology does not matter). I am given a detailed account of how the fissioning machine works and I set out to imagine myself going through the process. I have no difficulty with imagining myself lying down on the table, the table moving into the fissioning machine, the machine starting to work to divide me into two, .... and then what? I can not go any further. I cannot imagine myself, i.e., me as I am aware of myself from the inside, becoming two selves (Brook 1994, p. 172). To think of anything happening to myself, as Bennett noted many years ago, "pre-requires an undivided me" (Bennett 1974, p. 83). If so, Strawson is right that a self is synchronically a single thing. The reason is that synchronically a self must have unified consciousness and unity of consciousness requires singularity in the thing that is conscious.(6)
4. Diachronic unity and singularity
I further agree with Strawson that diachronic singularity is not needed for selfhood. However, diachronic unity is another matter. Diachronic unity is achieved by remembering past experiences, etc., and anticipating future experiences, etc., from the inside, in such a way that I can combine the representation of the remembered earlier or anticipated future states and events with current experienced states and events. This process of combination (or as Kant called it, synthesis) produces a form of synchronic unity, i.e., the unity that results when I combine a number of current states with one with another. If so, diachronic unity is simply synchronically unified consciousness where the representations being unified represent past and future states and events, the past ones in memories, the future ones in anticipations, etc.
This analysis gives rise to two questions:
1. Does diachronic unity require diachronic singularity of mind or person?
2. Is diachronic unity an essential feature of selfhood?
Strawson is inclined to answer 'no' to both questions. I agree with him about (1) but am not so sure about (2).
I agree with Strawson that diachronic unity does not require personal singularity across time (what philosophers call personal identity). All that is required for diachronic unity backward is that something of earlier experiences, actions, etc., be carried forward in such a way that representations from the inside of the earlier experiences, actions, etc., are available to me now. As Kant showed 200 years ago, such memories (or q-memories, to use the philosopher's fastidious term(7)) and the unity across time that they make possible could well survive transfer of whatever carries the memory being transferred from one person to another over time. (Strawson quotes a relevant passage from Kant and offers a related analysis on p. 415.) Strawson calls this picture the Pearl (string of pearls) view of diachronic unity (p. 424) and I agree with him entirely that it is a possibility. If so, unity of self backwards across time does not require that the self persists across time, paradoxical as that may sound.
The same is true and even more true of unity of self forward into the future. Since the self does not yet exist in the future, all unity forward in time can consist in is some present state. And all this present state consists in is the anticipation of certain future states as states of myself, i.e., 'projecting' myself into the future. If the unity achieved by projection into the future does not require that the self already exists in the future, it would seem likely, by parity of reasoning, that unity backward over time does not require that the self actually existed earlier (though perhaps some self had to exist earlier, the self that actually had and did the q-remembered experiences and actions). Anticipations of the future are one kind of experience in the present and memories are merely another. If so, Strawson is right to hold that what counts for determining the nature of the self is how we appear to ourselves in the present moment, not how we were or how we will be.(8)
So let us turn to (2): If diachronic singularity is not essential for selfhood, what about diachronic unity? Do we need to have memories of having and doing some past experiences and activities? Do we need to be able to at least some degree to imagine a future conceived of as our own future? Here, I think, Strawson's story may need to be qualified. On the one hand, I am inclined to agree with him that one would not lose one's selfhood by losing most of one's sense of having existed, of having experienced things and done things in the past. (As I analyse this sense, it really consists in unifying current awareness of having had and done the earlier things with awareness of one's current life and world, of course.) Likewise if one lost or perhaps even never had most of the normal ability to anticipate a future as one's own.
However, I am not so sure about a complete lack of any diachronic unity, a complete lack of any sense of personal continuity from the past to the present and into the future. I think Strawson may be inclined to understate how very, very different creatures who lost all sense of personal continuity would be from us. Oliver Sacks has explored a few cases in which virtually all sense of continuity with any earlier person has been lost (for example, cases where the patient has to introduced to the medical staff anew every morning). In the sense that matters for forming interpersonal relationships of any significance, there is no person there in these patients. A person who could not think of him- or herself as existing in the future, who could not, for example, plan what to have for dinner, wonder what tomorrow will hold, would be even more uncanny than these patients. Strawson may tend to understate the importance of a sense of continuity from pass to present to future to the question of having a self. As he tells us, his own sense of personal continuity with any past or future being is quite limited. I suspect that there is a huge difference between this sense being limited and having no sense of personal continuity at all.
For one thing, beings with no sense of personal continuity would also and for that very reason be cognitively impaired in a certain way. The ability to retain the contents of earlier experiences and, via remembering them, synthesize them with current experiences is an unsubstitutable feature of virtually all forms of cognition of any complexity. But when I remember an earlier experience from the inside, I also remember myself having that experience. (Strictly, I remember the experience as though it was me earlier who had it. It need not have been me who had it so long as something goes from that person to me that results in it seeming to me as though I had it.) If so, a measure of diachronic unity is essential for virtually all cognitive activity of any complexity.
This argument for diachronic continuity is not enough to undermine Strawson's most radical doubts completely. All it shows, strictly speaking, is that selves of any cognitive complexity must be unified over time. It is still possible that cognitively very simple selves do not have to be. Because of what I just said about Sacks' cases, I inclined to resist even this weaker claim but I am also inclined to think that it is not very important how we react to it. As Strawson articulates in his (8), selves have a certain personality, when we think of selves, we think of beings fairly richly endowed with personal characteristics. Indeed, we think of beings rather like us. If I have provided reason to think that selves of this sort must have significant diachronic unity, then I am content.
5. Remaining issues
I have not said anything about two of the features that Strawson identifies in his account of selves, namely, (1), that a self is a thing of a certain kind, and (7), that a self is an agent. (1) is easy: if the self is simply me as I am aware of myself from the inside, then, since I am a thing of a certain kind, the self is a thing of a certain kind. (7) is trickier. I find it hard to imagine a subject of experience, a thing able to be aware of itself, that had no volitional, action-initiating abilities of any kind whatsoever. Surely it must have at least the ability to initiate certain mental actions, e.g., directing and changing the focus of attention. However, I do not know how to prove, against Strawson's suspicions, that selves must have volitional abilities. (We have also said relatively little about (8), that selves have personalities, but that may not be very important. As Strawson suggests, having a certain personality may well be a less important feature of selves than any of  - .)
To sum up: if we push the notions of the unity of consciousness and singularity of person a bit
further than Strawson pushes them, some of his suggestions appear in even a more convincing light -
but at least one of them needs to be qualified, , namely, the suggestion that unity across time may not
be required for something to be a self.(9)
1. All page references are to Strawson 1997.
2. And when we are not aware of those activities? That is a difficult and tricky question that for reasons
of length I will not attempt to resolve.
3. I define 'representation' in terms of states of people to exclude the kinds of things external to persons
that also may have a representational function. The aim is to avoid taking a stand on a number of large
issues, externalism about representations in particular. More needs to be said about every aspect of
representations, of course, than I can say here.
4. Unity of consciousness as I have defined it here is only one of the relevant kinds of mental unity.
Others include: cognitive unity (ability to bring a vast range of cognitive resources to bear on a single
problem), unity of focus (ability to focus resources, including attention, on an objective), and unity of
behaviour (ability to keep all our vast repertoire of bodily motions and dispositions coordinated). Even
within unity of consciousness a distinction needs to be made between unity of simple consciousness and
unity of consciousness of self (Brook 1997).
5. Brook 1994, esp. Ch's 7 and 8.
6. Unity certainly goes with singularity here at least generally. I won't examine whether such unity
always or necessarily requires such singularity. As we will see, such a link is clearly not necessary in
the case of diachronic unity and singularity.
7. Starting with Shoemaker (1970), philosophers have used the term 'q-memory' in context like the
current one because of an objection that we can remember having and doing only our own earlier
experiences and actions. Response: switch to talk of q-memories, where q-memories are exactly like
memories except that this definitional link is absent. It then becomes an open question whether I can
remembering having and doing experiences and actions that, when they were done, were done by
someone else (Parfit 1984).
8. The term 'present moment' needs clarification. The experienced present is not the infinitesimal point
of modern physics. What we experience as "now" is in fact a stretch of time - a very small stretch, but a
stretch nonetheless. Philosophers call this phenomenon the 'specious present'.
9. My thanks to an anonymous referee for JCS for helpful questions and challenges.
1. All page references are to Strawson 1997.
2. And when we are not aware of those activities? That is a difficult and tricky question that for reasons of length I will not attempt to resolve.
3. I define 'representation' in terms of states of people to exclude the kinds of things external to persons that also may have a representational function. The aim is to avoid taking a stand on a number of large issues, externalism about representations in particular. More needs to be said about every aspect of representations, of course, than I can say here.
4. Unity of consciousness as I have defined it here is only one of the relevant kinds of mental unity. Others include: cognitive unity (ability to bring a vast range of cognitive resources to bear on a single problem), unity of focus (ability to focus resources, including attention, on an objective), and unity of behaviour (ability to keep all our vast repertoire of bodily motions and dispositions coordinated). Even within unity of consciousness a distinction needs to be made between unity of simple consciousness and unity of consciousness of self (Brook 1997).
5. Brook 1994, esp. Ch's 7 and 8.
6. Unity certainly goes with singularity here at least generally. I won't examine whether such unity always or necessarily requires such singularity. As we will see, such a link is clearly not necessary in the case of diachronic unity and singularity.
7. Starting with Shoemaker (1970), philosophers have used the term 'q-memory' in context like the current one because of an objection that we can remember having and doing only our own earlier experiences and actions. Response: switch to talk of q-memories, where q-memories are exactly like memories except that this definitional link is absent. It then becomes an open question whether I can remembering having and doing experiences and actions that, when they were done, were done by someone else (Parfit 1984).
8. The term 'present moment' needs clarification. The experienced present is not the infinitesimal point of modern physics. What we experience as "now" is in fact a stretch of time - a very small stretch, but a stretch nonetheless. Philosophers call this phenomenon the 'specious present'.
9. My thanks to an anonymous referee for JCS for helpful questions and challenges.
Bennett, J. 1974. Kant's Dialectic Cambridge and New York: Cambridge University Press
Brook, A. 1994. Kant and the Mind New York and Cambridge: Cambridge University Press
Brook, A. "Unity of Consciousness and Other Mental Unities", Proceedings of the 19th Annual Conference of the Cognitive Science Society, 1997 New York: Ablex Press, 1997, p. 875
Parfit, D. 1984. Reasons and Persons Oxford and New York: Oxford University Press
Shoemaker, S. 1970 "Persons and Their Pasts" American Philosophical Quarterly 7, pp. 269-85
Strawson, G. 1997. "'The Self'" Journal of Consciousness Studies 4, 5/6, pp. 405-28