The Brave Officer Who Does Not Remember

In 1785 Thomas Reid brought on three men who are supposed to be the same. A boy is flogged at school. Years later a young officer captures an enemy standard, and in that moment he remembers the flogging. Decades later again, the officer is an old general; he still recalls the capture of the standard exactly, of the school punishment he has no memory left.

Reid took this for a proof. If a person reaches exactly as far as her backward glance, then the officer is the same as the boy and the general the same as the officer. But the general is not the same as the boy. Identity will not stand for that: if A equals B and B equals C, then A must equal C. A criterion that breaks at this point cannot be identity.

The objection landed. It is still reprinted today whenever John Locke’s theory of the person comes up. Only it has since become contentious whether Reid was refuting Locke or a reading of Locke that Locke never held.

What Locke actually put forward

The chapter people argue over was not in the first edition of Locke’s Essay Concerning Human Understanding at all. “Of Identity and Diversity”, the twenty-seventh chapter of the second book, entered only with the second edition in 1694 — at the prompting of his correspondent William Molyneux.

Locke’s real incision comes before any answer, namely in the question. He separates three things the tradition had thrown together: the substance, the human being as a living organism, and the person. For him the person is neither stuff nor body but a thinking, self-conscious being that can take itself to be the same across time and place. Personal identity, II.xxvii.9 says, reaches back as far as consciousness can apprehend one’s own past actions.

What the separation is good for Locke states a few pages later with remarkable sobriety. “Person”, he writes, is a forensic term:

It therefore fits only beings that fall under a law and can experience happiness and misery (II.xxvii.26). The person is thus not a metaphysical find but the unit to which something can be imputed. Hence too the famous thought experiment of the prince and the cobbler (II.xxvii.15). If the prince’s consciousness wakes in the cobbler’s body, then for all the world the cobbler is the same man — and for purposes of imputation the same prince.

Two objections that became classics

Reid’s brave officer is one. The other is almost fifty years older and cuts deeper still. Joseph Butler, bishop and moral philosopher, appended a short text on personal identity to his Analogy of Religion in 1736. His argument is a circle: consciousness of one’s own identity already presupposes that identity. It can therefore attest to it, but not ground it. I remember an action as mine only if it was mine; the memory thus registers what it was supposed to explain.

Both objections share a tacit assumption. They read Locke’s “consciousness” as what we today call memory — as a glance back into an archive. That assumption is precisely the point at issue.

The quarrel over a single word

Locke scholarship has for decades been split into two camps. One holds to the classical equation: consciousness essentially means memory, and Locke offered a memory theory of the person — with all the consequent problems Reid and Butler named.

The other camp disputes the equation. Margaret Atherton pointed out in 1983 that Locke treats forgetting at length and treats it with open eyes; an author who simply tied identity to memory could hardly have afforded those passages. Kenneth Winkler in 1991 read Locke’s consciousness as appropriation: a person reaches as far as she takes past actions to be her own — and taking on is something other than retrieving. Udo Thiel and Antonia LoLordo have carried that line further, and Galen Strawson has linked consciousness in Locke to the notion of concernment, that is, to the fact that my past is a matter of concern to me.

The difference is not academic fine-tuning; it decides how far the objections reach. If consciousness means appropriation, then Reid’s general does not lose his identity with the flogged boy the moment the scene escapes him — he merely stops imputing it to himself at that moment. And Butler’s circle bites less hard, because an appropriation does not register anything it would have to presuppose; it does something. Anyone who wants to decide whether Locke has been refuted must first decide what he claimed. That question is open.

What memory research contributes

From the psychological side comes no verdict, but a sharpening. Elizabeth Loftus has shown how later information alters recollections; the misinformation effect is among the field’s most robust findings. Ulric Neisser, for his part, compared what people had recorded the day after the Challenger disaster of 1986 with what they recounted two and a half years later. The accounts often diverged considerably — the confidence with which they were delivered stayed high. The flashbulb memory too, the supposedly burnt-in image, is a reconstruction.

For a strict memory theory of the person this is bad news. It rests identity on a material that is full of holes, reassembles itself in the present and takes in foreign matter. We do not remember every earlier self; we forget childhood almost entirely, swap details and adopt other people’s accounts. Even so we go on treating ourselves as the same person, and the courts do too.

In 1984 Derek Parfit drew the most radical consequence from this: perhaps identity need not be an all-or-nothing fact at all. What counts would be overlapping chains of psychological connection — memories, intentions, traits, attachments, embodied habits. No single link has to span the whole distance for the chain to hold. Reid’s general would then not be refuted at one point but connected to the boy in degrees.

With that a circle closes which the debate rarely draws. Reid’s objection draws its force from identity being a strict relation: what is the same is wholly the same, and a criterion that does not hold that is refuted. But a forensic term demands no such strictness. Courts know degrees — contributory fault, diminished responsibility, limitation periods. For anyone who reads Locke’s person from the outset as a unit of imputation, Parfit’s gradation is no break with Locke but the continuation of his move by other means. Reid and Butler then did not strike Locke but the metaphysical reading they imputed to him.

What remains

What is secure is Locke’s incision, not his answer. That person, human being and substance belong apart, and that “person” is a term of imputation, has outlasted the debate; here lies Locke’s enduring achievement. What he meant by consciousness is disputed to this day — and whoever passes over the dispute and briskly credits Locke with the memory theory is refuting a position whose author is unknown. No acquittal is to be expected from psychology, but a dependable report on the material: memory is an apparatus of reconstruction, not an archive.

In everyday life one notices this in an asymmetry everyone knows. We demand of ourselves that we answer for deeds we barely remember, and we recognise ourselves in accounts we have half from other people. In dementia and amnesia this order is violated without a person thereby losing dignity, relationships or her claim to her own past. That is a strong argument that identity never lay in the individual head alone. Other people keep parts of our history for us. We are continuing, revisable connections: stories with a body, a social world and more gaps than their narrators like to admit.