Jeffrey asked:

how does Locke solve the problem with personal identity?

Answer by Eric DeJardin

Hello, Jeffrey!

Let’s try to get clear about what ‘the problem with personal identity’ is; then, let’s look at Locke’s resolution of it.

By ‘identity’ we mean ‘numerical identity’ as opposed to ‘qualitative identity’. The difference could be illustrated like this: two cars that are indistinguishable with respect to model, parts, color etc. are said to be qualitatively identical to each other, while one of those cars at some time t is said to be numerically identical to the same car at some later time t’. And by the modifier ‘personal’, Locke is indicating that he’s concerned with the identity of a delimited class of entities, i.e. of persons. The problem, therefore, concerns the numerical identity of persons.

But what does Locke take a person to be? His stipulated answer is:

“[a person is] a thinking intelligent being, that has reason and reflection, and can consider itself as itself, the same thinking thing, in di?erent times and places; which it does only by that consciousness which is inseparable from thinking, and…essential to it…”

So, what is it about the numerical identity of persons that’s problematic?

There are actually many problems in this area, but Locke is primarily focused on one, viz. the problem of the identity of persons over time. Essentially, Locke’s problem concerns developing a criterion of the numerical identity of persons that clarifies in virtue of what a person at one time and a person at a different time count as the same person.

Here we should draw a distinction between ‘evidential’ and ‘constitutive’ criteria of identity. Evidential criteria provide us with reasonably reliable means of determining the numerical identity of persons; hence, we might conclude that Jones today is numerically identical with the Jones who robbed the local bank years ago because ‘both’ share the same fingerprints, DNA, appearance, etc. Constitutive criteria, however, concern what in fact makes Jones identical with the person who robbed the bank, and whatever that is, it need not be adduced to defend the claim that Jones (today) is the person who robbed the bank. Evidential criteria are thus epistemological (i.e. concern how we know), while constitutive criteria are metaphysical (i.e. concern what something fundamentally is).

We can now clarify Locke’s problem: What is the constitutive criterion of the numerical identity of persons over time?

Before we explain Locke’s answer, it would be helpful to understand what motivates it, for Locke’s proposed constitutive criterion is not merely metaphysical in nature, but forensic as well. That is, Locke is primarily interested in those aspects of persistent personhood in virtue of which one is properly blameworthy or praiseworthy (both morally and legally). So, if Jones today is the same person as the Jones who robbed the bank, then we can say that the today’s Jones is guilty of that crime.

With all the necessary ground clearing out of the way, we can (finally!) present Locke’s response to the problem, viz. it is in virtue of possessing the same continuous consciousness, with its attendant memories and self-awareness and inner sense of continuity, that persons are numerically identified. But what reasons does Locke provide to support this conclusion, and should we accept them?

Locke argues that the criterion of numerical identity will vary according to what it is we’re considering. So, if we take atoms to be indivisible (as he did), then, since any particular atom is identical to itself at any point in time, it will remain the same atom as long as it exists; and, since this holds for individual atoms, it also holds for combinations of atoms, which remain the same as long as no atoms are subtracted or added to the particular combination. But living things seem to exist continuously while undergoing changes in the matter that composes them. Hence, Locke argues that it is their organization and structure, coupled with their continuous life, that serve as the constitutive criteria of the numerical identity of living things.

Since human beings are living things, they’re numerically identified as other organisms are. But can we identify the person (recall Locke’s definition above) with the human being? Locke argues that we cannot, for we can conceive of cases in which the two notions come apart. So, suppose that the consciousness, memories etc. of a criminal are somehow transferred from the criminal’s body (A) to the body of a saint (B), and vice versa; would we say that A is still the criminal, and B the saint? Locke argues that we would not, for A would lack the criminal’s memories and dispositions, and would instead be possessed of a consciousness continuous with that of the saint, with its attendant memories and dispositions; rather, since B would now be possessed of the criminal’s continuous consciousness, we’d want to say that B is now the criminal, and A the saint. But then it follows that the person is distinct from the human being, and that the former, not the latter, is the locus of moral responsibility (for although the crime was committed ‘with’ A, we would not now judge A to be the criminal, but B).

Indeed, Locke argues that the person cannot be identified with any substance, for we can similarly conceive of the same consciousness (say, the criminal’s) moving among different particulars of the same substance (e.g. from one body to another), or even among particulars of different substances (e.g. from a body to a soul). Hence, the person is to be identified with continuity of consciousness alone.

So, did Locke get it right?

Reid famously raised a problem with Locke’s account of personal identity: suppose a man at 80 can remember what he did at 40, and at 40 could remember what he did at 10, but at 80 cannot remember what he did at 10. On Locke’s account, it follows that the 10 year old and the 80 year old are identical with the 40 year old (since, in both cases, they share one continuous consciousness), but that the 10 year old is not identical with the 80 year old (since the latter has no memory of the former). Can Locke’s account handle this objection?

One response (by Quinton) involves supposing that memories count as the memories of the same person if they stand in an ancestral relation to one another, i.e. if the 80 year old remembers what he did at 40, then as long as the 40 year old remembers what the 10 year old did, the 80 year old and the 10 year old are the same person (i.e. the memories of the 10 year old are ancestors of the memories of the 80 year old via the memories of the 40 year old).

Another problem was raised by Butler: If a crucial element of Locke’s criterion of personal identity is memory, then the account seems circular, since one could only identify past memories as the memories of the same person if one presupposes that it is indeed the same person who is in possession of those memories; yet that is precisely the claim — i.e. that is the same person — that the appeal to memory is supposed to support.

Shoemaker’s response is to redefine the relevant notion of memory in an attempt to rid it of the elements that lead to the circularity. He suggests that we instead appeal to ‘quasi-memory’, which could be understood as memory sans one’s awareness of having personally experienced the recalled event. Hence, the memorial element of Locke’s account of personal identity can be redefined in terms of quasi-memory to avoid the circularity objection.

But these are standard objections with their standard responses. Again, did Locke get it right? It seems to me as if there’s one sense in which he did.

Suppose Jones not only (somehow) lost all his memories, but also the dispositions that made the act of robbing a bank possible; instead, he is now, unlike his former self, a kind, caring and law abiding person. Would you be inclined to think that the new Jones is the same person as the old Jones? One way to answer this question is to answer another, decidedly Lockean one: Would you hold the new Jones responsible for the old Jones’s actions? I suspect that you wouldn’t; but then, at least in one relevant sense, you wouldn’t take the new Jones to be the same person as the old Jones. So, while Locke may not have provided us with the necessary and sufficient conditions of persistent personhood, he seems to have developed a sufficient condition, at least in cases that directly concern moral and legal responsibility.