Andrew Naylor

Most of our memories are inferential, so says Sven Bernecker in Memory: A Philosophical Study. I show that his account of inferentially remembering that p is too strong. A revision of the account that avoids the difficulty is proposed. Since inferential memory that p is memory that q (a proposition distinct from p) with an admixture of inference from one’s memory that q and a true thought one has that r, its analysis presupposes an adequate account of the (presumably non-inferential) memory that q. Bernecker’s account of non-inferentially remembering that is shown to be inadequate. A remedy lies in strengthening the account by requiring the rememberer to have had prima facie justification to believe that q, any defeaters of which were misleading.

One difference between traditional and contemporary nontraditional theories of memory is that the former would affirm, whereas the latter would deny, that a person can be correctly described as having remembered that p solely in virtue of having knowledge the certainty of which is grounded upon the person’s present remembering. I argue that there cannot be such a case, and that what may appear to be such a case—as presented in Don Locke’s book Memory—can be explicated by a contemporary nontraditional theorist without making any concessions to the traditional theorist

The definition of memory knowledge that p put forward in this paper is nontraditional in that the justification for the belief that p which constitutes that knowledge is not located in any memory-impression or other present state of the subject. Rather it is the subject's actual past justification for p, or a proper part thereof, that justifies this present belief that p. It is argued (1) that the notion under definition is that of knowing straight from memory, (2) that an adequate definition here must take into account a difference, as to conflicting evidence one does not possess, between evidence one has forgotten and evidence one has never had, (3) that compared to Ginet's traditional definition (1975), the definition has several advantages, and (4) that the definition handles at least one type of situation where there can be memory knowledge that p without previous knowledge that p.

This article sets forth a view about how epistemic justification figures in the ongoing justification of memory belief, a view that I call moderate justificational preservationism . MJP presupposes a nontraditional notion of memorial justification according to which what makes one's present belief that p prima facie justified is that which provided one with prima facie justification to believe that p originally . The article offers support for MJP by examining a series of cases that involve forgetting, and in doing so, criticizes views of Jennifer Lackey, David Owens, Michael Huemer, and George Pappas

Abstract: A person who remembers having done something has a belief that she did it from having done it. To have a belief that one did something from having done it is to believe that one did the action on the (causal) basis of having done it, where this belief (in order for one to have it) need not be (causally) based even in part on any contributor to the belief other than doing the action. The notion of a contributor to a belief (as opposed to a mere facilitating cause of the belief) is explicated through a series of examples. The account of having a belief that one did something from having done it is then deployed in criticising Ginet's account of ‘memory connection’, in assessing Martin and Deutscher's causal theory of remembering, in indicating how diachronic justification functions in a nontraditional theory of memory, and in setting forth one type of psychological connectedness which, according to advocates of a psychological continuity theory of personal identity, may be employed (noncircularly) in formulating the theory, and which, according to opponents of the theory, provides a target for criticising the theory

Against Russell’s skeptical conjecture, that the world and its entire population came into existence five minutes ago, it is argued that any one of the following is logically incompatible with the conjunction of the other two: ostensible memories of certain events, records of such events, and the non-occurrence of these same events. This conclusion is reached through a critical examination of (1) the arguments advanced by Norman Malcolm in trying to show that Russell’s “hypothesis” does not express a logical possibility, and (2) the counterarguments by which James W. Cornman tries to show that it does.

The paper presents considerations that weigh against one or another version of the psychological continuity theory of personal identity over time. Such Locke-like theories frequently go wrong, it is argued, in not formulating precisely how the psychological states of an individual person are related diachronically, in failing to capture a truly appropriate causal connection between later and earlier psychological states, and in claiming support from particular cases. In addition, the paper offers examples and other considerations that support an alternative, biological continuity theory according to which you and I are each identical with a human organism.

For cases in which to remember that p is to have (strict) nonbasic, unmixed memory knowledge that p; in which there is at most one prior time, t, from which one remembers; in which one knew at t that p; and in which there can arise a sensible question whether one remembers that p from t — a person, B, remembers that p from t if and only if: (1) There is a set of grounds a subset of which consists of (i) only those grounds B has at both t and the present for B to be sure that p, and (ii) enough such grounds to make it reasonable at both t and the present for B to be sure that p (I call any such subset a set of “adequate original grounds dating from t”), and (2) there is no time prior to t such that B has a set of adequate original grounds dating from that time. The way in which the crucial terms in this explication are being used is explained. And the explication is defended by showing how it can deal with cases that are counterexamples to explications recently offered by Malcolm and by Munsat.

This paper examines a leading traditional account of memory knowledge. (A “traditional” account of memory knowledge locates whatever positive justification there may be for the belief which constitutes that knowledge in a present memory-impression.) The paper (1) presents a pair of cases designed to show that Carl Ginet’s four-part defeasibility-type definition of memory knowledge that p is either too weak or too strong, and (2) suggests how these cases could be handled by one sort of non-traditional account.

A theory of occurrent factual memory is sketched out. The theory represents an alterative to the traditional theory in John L. Pollock’s Knowledge and Justification, in that it analyzes occurrently remembering that p without employing the notion of ostensible recollection that p. The latter notion, it is argued, can be understood in terms of occurrently believing (or being inclined to believe) that p. In defending his theory against nontraditional alternatives, Pollock employs arguments that conflict with his own principle of implicit reasons. That principle, it is shown, sanctions cross-temporal justification of the sort presupposed by the alternative nontraditional theory. [CORRIGENDUM: Line 6 at the top of page 137 should read: memory, then if S occurrently remembers that p at t, then S, at t, has a justified (occurrent) belief that p such that (a) this belief that p constitutes, at t, S’s In lines 7–9 at the top of page 140, the sentence beginning with “For” and ending with “way” should be deleted, and in line 11 “fact” should be “face”.]