III. MEMORY AND RESPONSIBILITY
The discussion has thus far focused on the ontological aspect of our coping with the past. The primary foil has been the radical and indeed paradoxical suggestion that in order to escape the shadow of the past, a subject must undergo a change of identity, and so, in effect, cease to exist. The notion of temporal boundaries and their revision offers a less extreme and more appealing alternative. This notion has corresponding attractions at the epistemological level as well. Here the foil is the suggestion that in order to rid ourselves of a nasty past we must forget it, thus forfeiting important knowledge. The notion of a change in temporal boundary offers an intermediate possibility here as well. In discussing it, I reverse the order of proceeding, looking first at the individual case and next at the collective case.
Memory and Identity
At first sight, a change in the offender’s boundary that excludes the wrongful act renders the appeal to forgetting superfluous. Whereas identity change is offered as primarily an account of repentance, and so concerns the offender, forgetting supposedly accounts for forgiveness, and concerns the victim. A change in the offender’s temporal boundary allows the victim to retain full knowledge of the past events, consistent with an appreciation that reactive attitudes toward the offender, as currently constituted, are no longer appropriate. The victim may accordingly still benefit from the lessons of the past while associating those lessons with a version of the offender that is no longer viable. But although forgiveness is in the first place bound up with the victim’s epistemological stance, the offender’s epistemological position is relevant too, complicating matters. If asked point blank whether she remembers the wrongful act, the offender would have to avow that she does, despite having repented the act and having been forgiven for it. To be sure, under these conditions the matter is unlikely to arise. The truly repentant ought not dwell on the past misdeed, and the hypothetical interrogator’s bringing up the nasty event would be deemed insensitive and obtuse. Still, the mere possibility of such an avowal seems to belie the claim that revisionary practices accomplish anything approaching the removal of the misdeed from the ambit of the wrongdoer’s self. So even if the preceding arguments establish that revisions of the self’s temporal boundary are conceivable and metaphysically possible, it may be felt that as long as the wrongdoer herself remembers the wrongful act, such revisions do not in fact occur.
One way to defuse this objection requires that we take a closer look at the facts. These, as I had the objector recount them, are complex. One fact, the one that gives the objection its bite, is the supposed avowal. But an equally salient fact is that the avowal is prompted by an inappropriate question likely to provoke indignation and disdain. If the misdeed were forever inscribed in the offender’s past, as the objector implies, this aspect of the situation would be puzzling. Admittedly, an inquiry into the truth may sometimes be considered tactless, but why would it seem to be a serious transgression? How can dredging up the past strike us not just as uncouth but also as unfair? Obviously, important norms are here at play, colloquially rendered by expressions such as “let bygones be bygones,” designed to discourage an appeal to the past by rendering such an appeal highly inappropriate. Barring a rude violation of these norms, the offense will not come up, and the various attitudes and other ramifications ordinarily associated with it will have been effectively extinguished. These are powerful facts that need to be accounted for no less than what happens when the norms are breached and the past is exhumed.
The proposed account offers a straightforward explanation for all these facts. A serious wrongdoing invariably casts a long shadow over the offender’s life in the form of lasting negative attitudes and other consequences. When, due to the operation of the revisionary practices, the shadow disappears, we ought to conclude that its source in the wrongdoer’s life has been removed. But this is only part of the story. The possibility, no matter how remote, that the past misdeed may still surface must be accounted for as well. The key is a point I made earlier concerning the indeterminacy of a border change. Various parties can draw a state’s border in different ways without there being a single overriding authority to choose among them or a fact of the matter as to which is correct. A similar indeterminacy, the product of a plurality of competing versions, pertains to the temporal boundary as well. Revisionary practices give rise to a new version of the state, and correspondingly of the individual self, from which the wrongful act is excluded. When this version is enacted, it replaces the older one as superior or more authoritative. As long as this version is adhered to, as by and large it will be, the misdeed is indeed excluded and does not cast a shadow. But the new version does not obliterate the other one; it only supersedes it. So if one insists, cruelly or obtusely, on unearthing and scrutinizing the older version, one is not strictly speaking mistaken, but merely cruel or obtuse.
Still, this appeal to discursive norms may seem to be an evasion. Jogging the offender’s memory in the hypothesized circumstances, though inappropriate, is possible. And this possibility is unsettling. Doesn’t the offender’s memory, even when dormant and merely dispositional, disprove the claim that the offensive act is no longer hers? Moreover, if recognizing a change in a temporal boundary requires that the wrongful action not be mentioned (as in “don’t mention the war”), the practical result is akin to forgetting, since it makes the lessons garnered from the experience publicly unavailable for future use. If a change in temporal boundaries is to provide a viable model for escaping the past, such a change must be less precarious, and in particular less vulnerable to recollection.
To allay these worries we must identify their source. Suppose that Gertrude is charged with battery for hitting Kim. An eyewitness is found. Plainly, for the purpose of testimony, having observed the event is not enough; the witness must remember it. Memory thus serves as a crucial conduit that makes available the initial information acquired by watching the event. Notice, however, that though the testimony is highly relevant to the charge against Gertrude, it does not settle it. Witnesses don’t allocate responsibility; they only provide information that bears on such allocations. The allocation of responsibility requires a judgment, and the authority to make it. Testimony regarding Gertrude’s alleged attack is one thing; a decision concerning her responsibility, another. Now Gertrude’s own memory can play a similar evidentiary role to the witness’s. After all, she too was present at the scene of the alleged offense, and so can draw on her memory to provide a first-hand account. But if Gertrude’s own memory is treated as equivalent to the witness’s memory and as playing the same evidentiary role, it would not be able to settle the question of Gertrude’s responsibility any more than the witness’s. The objection we’re considering (that an agent’s remembering a wrongdoing inescapably saddles her with responsibility for it, thus disproving my account of revisionary practices) must assume that first-person memory is more than evidence relevant to responsibility: to remember performing an action is to reveal oneself as the author of that action. This view of the significance of biographical memory reflects a conception of memory as the key to personal identity and as the thread that links the disparate elements of a human life.
Although this relationship between memory and personal identity is highly contested, we can reduce the area of contestation, and canvass at least some of the issues involved, by focusing on a particularly prominent advocate of this view, John Locke, and by giving the objection under consideration a Lockean gloss. Locke’s version is particularly congenial here, since he famously distinguishes the identity of a human being conceived as a biological organism (“an animal of such a certain form”21) from the identity of a “person” or “self,” which is “a forensic term appropriating actions and their merit.” Since Locke further maintains that “in this personal identity is founded all the right and justice of reward and punishment,” we are at present interested exclusively in his views regarding personal identity, or (what amounts to the same thing) the identity of a self rather than that of a human being, biologically conceived.
Now at first sight (and according to much of the Lockean tradition, at second sight too), Locke would indeed be in agreement with our objector, who considers Gertrude’s memory of the attack on Kim to be sufficient to settle the issue of her responsibility. According to Locke, personal identity is indeed constituted by first-personal memory—“whatever has the consciousness of present and past actions is the same person to whom they both belong”—and so is one’s responsibility: “I being as much concerned and as justly accountable for any action was done a thousand years since, appropriated to me now by this self-consciousness, as I am for what I did the last moment.” Call this the sufficiency thesis. In light of this thesis, the persistence of memory even after revisionary practices have run their course would indeed disprove my claim that these practices absolve the wrongdoer of responsibility by removing the offense from the ambit of her self. Is the sufficiency thesis sound?
In addressing this question, we should note that Locke goes beyond the sufficiency thesis, insisting that memory is also necessary for responsibility: just as Gertrude’s remembering the attack would ground her responsibility, so her forgetting it would negate responsibility. This necessity thesis, however, is patently implausible. Consider a robber who sustains a concussion during a heist, and when he comes to, no longer remembers the entire affair. Would that amount to denying that he’d participated in the crime or absolve him of liability? A negative answer is, if anything, even clearer in a parallel case in which credit rather than blame is meted out. Think of an intrepid firefighter, who suffers a memory-erasing injury while performing a heroic act; would she be denied the laurels that would otherwise be her due? Locke recognizes that the law does not in fact conform to the necessity thesis, as it convicts defendants of crimes they do not recall, and tries to account for the disparity between his theory and the law by pointing to evidentiary exigencies faced by law. Memory, Locke argues, is subjective, and so forgetting easy to feign. Expediency thus mandates that the law dispense with memory as a condition of liability, despite the resulting injustice. However, Locke’s attempt to reconcile legal practice with his own view is unconvincing, for two reasons. First, the evidentiary worry does not arise in the case of reward: no one is likely to feign forgetting in such a case. So Locke would have to acknowledge, counterintuitively, that by forgetting her heroic acts, the firefighter must forfeit all credit for them. Second, criminal liability routinely depends on establishing a defendant’s intent or other state of mind; such subjective elements of crime do not present an insurmountable obstacle to law enforcement. Memory raises no special difficulties in this regard.
These considerations only show that memory is not necessary for responsibility, whereas the question before us is whether memory is sufficient. But though, in principle, the sufficiency thesis might be independent of the necessity thesis, that principle would not be the one actually espoused by Locke. That much is suggested by the very fact that Locke deems it necessary to defend not just the sufficiency thesis but the necessity thesis as well. When he insists in the face of pervasive legal practice that memory is necessary for responsibility, he bites a hard bullet indeed; and his appeal to expediency to explain this practice is ad hoc and patently unconvincing. Why doesn’t Locke follow the more inviting road of allowing that memory is sufficient but not necessary for responsibility? A likely answer follows from Locke’s main objective, which is to advocate a particular theory of personal identity. To understand why Locke insists that memory is necessary for responsibility, and not just sufficient, and why the two theses stand or fall together, we need turn to this broader objective.
In a thumbnail sketch, Locke’s theory of personal identity consists in an alleged alignment between three key terms: memory, identity, and responsibility. I am responsible for the actions I remember since memory is what links me to them, what makes them mine. It is crucial to this theory that the three terms be coextensive, and Locke’s defense of the necessity thesis is part of an attempted demonstration that they are. Conversely, rebutting the necessity thesis drives a wedge between memory and responsibility, thereby breaching this tripartite equation. This leaves two options: retaining the equation between memory and identity (by adhering to the view that memory is the criterion of personal identity) while cutting responsibility loose, or retaining the equation between identity and responsibility and cutting memory loose. In making the choice we need to bear in mind that as applied in the present context neither the notion of identity nor that of memory is a unitary one. Regarding identity, recall that Locke distinguishes the identity of a human being from that of a person or a self.
Remembering an action or experience does, plausibly enough, establish the continued existence of a human being, that is of the organism bearing the memory, between the time of the action or other experience and the present. But this is not the identity that concerns Locke. His main point in distinguishing human being from person or self is to mark off those aspects of the human condition that he designates as “forensic.” Responsibility is a defining aspect or characteristic of the forensic, and so indispensable to Locke’s concept of self. Since, as the rejection of the necessity thesis shows, memory cannot be convincingly seen as constituting responsibility or as coextensive with it, and the three-way equation (between memory, identity, and responsibility) cannot be maintained, we must conclude that the self’s identity (unlike that of a human being) coincides with the self’s field of responsibility, whether or not memory persists.
This conclusion is reinforced by the realization that memory isn’t a unitary notion either. When we contemplate autobiographical memory, we tend to focus on a paradigm case. In such a case, memory is not just a matter of the path by which knowledge is acquired and retained. In addition, the bit of knowledge possessed by way of autobiographical memory is experienced as self-regarding or reflexive. In the typical case, such reflexivity has an affective side as well as a cognitive side. Specifically, to remember an action as one’s own is to entertain toward it such self- regarding attitudes as pride, or satisfaction, or remorse. But though this is the paradigm case, not all instances of what we would ordinarily describe as memory fit this description. We often possess knowledge that in its modes of acquisition and retention accords with those of memory, while lacking the full-fledged phenomenology. Once we separate the purely cognitive aspect of memory from the affective aspect, memory is no longer univocal, and no longer appears convincingly as constitutive of responsibility. Just as forgetting an action does not by itself remove it from the ambit of the self and does not absolve the agent from responsibility for it, so remembering it in this attenuated sense does not encompass it within the self nor secure responsibility.
The possibility of affectless memory also casts light on the paradigm case, in which memory does involve self-regarding emotions. Emotions such as pride and shame are not brute stirrings in the heart. To entertain such emotions with respect to an action involves a judgment that the action is a suitable source of them. The relationship between the possession of first-personal knowledge of a past action and such emotions is accordingly a normative relationship. Since in the paradigm case the knowledge and the emotions are linked, the link may appear inexorable, inducing the impression that memory is constitutive of both identity and responsibility. But once the two factors that are conjoined in full-fledged memory are pried apart, it turns out that the factor that is constitutive of responsibility is not the possession of first-hand knowledge of the past affair, but rather the appropriateness of entertaining such attitudes as pride or shame with regard to it. And given Locke’s forensic conception of the self, the same factor must be seen as defining the boundaries of the self and its identity even in cases of paradigmatic memory, in which the relevant attitudes are tied to the possession of first-hand, internal knowledge of the predicate actions. Moreover, given the complexity of what memory designates, its negation cannot be perspicuously rendered by the single term “forgetting”; this dichotomy between remembering an action and forgetting it turns out to be too restrictive. There is conceptual room for a stance toward performing a past action in which one retains the purely cognitive aspect of memory, but forgoes or overcomes the affective aspect. We can say about such a person that she unremembers the action without forgetting it. Revisionary practices can lead to this intermediate stance.