On memory and information: a note

Much of my criticism of the semantic theory of information is informed by what are broadly-speaking social constructivist or post-structuralist worries about the contingency and co-constituative nature of reality. But, any good articulation of contingency and co-constituation must rely on historiography. In common sense terms historians are writing history by employing logic and information, but the referent is typically very distal (even if the object is material and in front of the historian, we do not typically take the historian to be speaking about that present object). And still, historiography is not neutral, it is a form of sense making: “the past invades the present despite historians’ attempts to put it at a distance” (Kasabova).

How can we reconcile historiography with the semantic theory of information? Dretske’s discussion of distal objects, as referring to something elsewhere is an answer, but I worry that it does not give sufficient weight to history or the contested nature of historiography—-of importance for Dretske is the present intentional relationship, not the actual (distal) event. Very memorable, sad, chilling, or joyous events, for example, the Holocaust, election of the first Black American US president, or the East Timor genocide, seem to require a much more subtle account. At the very least, Drestke’s presentism is problematic on purely metaphysical terms. Since Augustine’s formulation of the issue in Confessions XI presentism is coherent only in a loose sense in which the “present” is taken to mean “a short time”. And worse still, if presentism is true, the historian seem to be in a kind of fool, since the “past is no longer” and the “future is not yet” (Augustine, Confessions). Instead, we take it that what is important about history and historiography to be twofold: 1) let us remember these events, commemorate or lament them, so as to direct our current ethical, political, and material orientation and to avoid them or repeat them (this is both presentist and requires extension into the near or distant future), and 2) let us remember these events for their own sake, as events that deserve the respect of existing on their own, without the violence of articulation into current or future being. Following Aristotle (On Memory) we can distinguish between the retentive and the retrieving functions of memory, “the former preserves an event from forgetting and erasure, while the latter recalls and brings it back to the present.” (Kasabova); both forms are necessary.

It seems that the kind of realism required by the semantic theory of information requires a presentism with respect to time and change, yet, if we are to take seriously worries by privacy advocates that a major future issue of computer ethics is that they “don’t forget”, we must understand memory in relation to information. The need for forgetting, rather than remembering or commemorating seems to deal with meaning making (thus, semantics). In order to reason (in the broadest sense) we need to be able to select and evaluate, which means we must be able to forget. As Kasabova states, “to put it another way: if the temporal link memories were continuous, we would have no sufficient criteria for individuating our memories. Only a presentist would have no problems with this situation”. The inability to forget is, quite possibly, one of the reasons why computers are not semantic machines. The ability to extend a gaze through time, take history seriously, is a requirement for semantics.