Keeping track, autobiography, and the conditions for self-erosion

Michael Luntley

in Dementia

Published on behalf of Oxford University Press

Published in print December 2005 | ISBN: 9780198566151
Published online February 2013 | e-ISBN: 9780191754418 | DOI:

Series: International Perspectives in Philosophy & Psychiatry

Keeping track, autobiography, and the conditions for self-erosion

Show Summary Details


In this chapter I want to apply some ideas from recent work on self-consciousness to explore the following metaphysical issue: under what conditions would it make sense to say that severe dementia gives rise to a loss of self? I shall argue that on a broadly Kantian conception of the self there is a case for saying that the erosion of certain basic cognitive capacities central to autobiographical memory amounts to an erosion of the self. I suggest that it is an empirical matter whether or not severely demented patients display the relevant erosion of cognitive capacities. The argument shows, however, that the very idea of loss of self makes sense.

The basic cognitive capacities at issue concern the capacity for integrating ideas into an autobiographical unity, a unity that underpins our capacity to self-narrate. This requires an ability for temporal binding—treating an Idea at one time as bearing upon an Idea at another time. The simplest way in which this occurs is when a subject retains an Idea—a way of thinking of a thing, property or event—through time.1 When a subject does this they keep track of the thing picked out by the Idea. Keeping track of things is one of the basic achievements of the self and, on the conception I exploit, it is a condition for the possibility of self-reference. Genuine self-reference is, on this conception, explained by self-consciousness conceived in terms of a set of cognitive capacities. The possibility I want to outline concerns the sort of breakdown in these cognitive capacities that amounts to a loss of self-consciousness and, thereby, a loss of self-reference. The breakdown at issue is the loss of the capacity to keep track of things.

This might seem to reverse the obvious order of dependence, for you might think that the ability to think of oneself is prior to the ability to think of other things. But on the conception of the self I exploit, the self is that which keeps track of things. More precisely, the self is the ground for the possibility of keeping track of things. The ability to think of oneself—to self-refer—depends upon the array of cognitive capacities for thinking of other things, for it is only with those in place that one has the capacity to think of oneself as an object, one thing among others.2 The capacities for thinking of other things include, centrally, a capacity for temporal binding. On this conception, such capacities are conditions for the possibility of self-reference. The capacity for self-reference is explained in the light of these other capacities. When these capacities slip, especially the capacity for temporal binding, so too does the self. When the capacity for keeping track of things has been eroded there is nothing left to constitute a self. At that stage, self-consciousness and, thereby, self-reference is lost. The point is not that under such circumstances the subject fails to keep track of their own self, for on the conception I employ, the self is not something that is ever tracked. The self is always the tracker—the ground of the possibility of keeping track of things. So the erosion at issue in the case of severe dementia is not a failure to keep track of the self, but the failure to keep track of outer things. The self is not, on this conception, an inner thing that is tracked on analogy with outer things; it is the condition for the possibility of keeping track of outer things. When that goes, the self goes too.

The conclusion is counter-intuitive, for the sort of cases I envisage will include cases of patients who have some capacity for ‘I’ vocalizations. For example, a patient may have a capacity to utter sentences like, ‘I am thirsty’ but nevertheless have lost the capacity to keep track of things in the way required to manifest the cognitive capacities constitutive of self-consciousness and self-reference. Under such circumstances I suggest that despite the apparent self-reference involved in the language use, we can make sense of the idea that no genuine self-reference is available for such a language user. Whether it is right to think that is ever the case is an empirical matter.3

I shall outline the argument for this conclusion in the context of two competing conceptions of the self, which I shall call the Lockean and the Kantian models.4 The Lockean model is a reductionist view of the self; for it is a conception on which the individuation of Ideas is taken as primitive and selves are defined over collections of Ideas. In contrast, the Kantian model treats the self as a condition for the possibility of certain basic cognitive skills, skills that critically underpin the achievement of keeping track of things. The self is, thereby, partly constitutive of the account of the ideas involved in such skills. On this conception, the self is not itself something that one tracks. For the Kantian, the first-person pronoun is not an expression the reference of which has to be achieved. It does not function like a demonstrative, let alone a name or description. There is such a thing as self-reference for the Kantian, but unlike all other forms of reference, it is not an achievement. For the Kantian, self-reference is fixed by the token-reflexive rule: any token of ‘I’ refers to whoever produced it. But that only fixes reference against a background of conditions for being a language user, someone capable of achieving, amongst other things, reference to objects. It is a condition for the possibility of achieving reference to other things that a language user manifests the cognitive skills that make up self-consciousness. When those skills are absent, the subject who utters sentences containing ‘I’ has lost the cognitive surround that makes it compelling to say that the token-reflexive rule alone suffices to fix reference. A laptop computer that was programmed to say ‘I need a mains power source’ when its battery ran low does not count as making self-reference just because it has produced a sentence with the first-person pronoun in it. The appropriate surround of cognitive skills is absent. The central hypothesis of this chapter is that we can make sense of circumstances under which things could be so bad for a severely demented patient that, notwithstanding their use of the first-person pronoun in sentences such as, ‘I am thirsty’, the appropriate cognitive surround, which warrants treating this as a case of self-reference, is absent.

Chapter.  8349 words.  Illustrated.

Subjects: Psychiatry

Full text: subscription required

How to subscribe Recommend to my Librarian

Buy this work at Oxford University Press »

Users without a subscription are not able to see the full content. Please, subscribe or login to access all content.