Memory without identity

ABSTRACT

I defend the view that episodic memory judgments do not depend on any kind of identification of oneself as the person whose past is being remembered, and are therefore logically (rather than merely de facto) immune from error through misidentification relative to “I”. There are two challenges to this view that have been pressed in the literature. One appeals to the idea of background presuppositions of identity and says that “I am the person from whom my memory impression derives” is a background presupposition of any memory judgment. The other appeals to wh-misidentification and says there are possible cases in which memory goes astray that should be counted as cases of error through wh-misidentification. Although the details are different, the core thought behind both challenges is the same: Shoemaker’s thought that there could be a memory-like relation that one stands in to another’s past. I think this thought is a mistake.

KEYWORDS:

• IEM
• memory
• background presupposition
• epistemic dependence
• semantic presupposition

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1. See Shoemaker (1968) and (1970).

2. It may be helpful to provide a clarificatory note on “logically”. One might think that the logically necessary truths are a subset of the necessary truths and that “Superman is Superman” is logically necessary, because it’s necessary in virtue of its logical form in a way that “Superman is Clark” is not. In the debate about IEM, the point of “logically” is different. It’s not to point at a particular explanation of why a judgment is IEM, e.g., to say that it is IEM in virtue of its logical form. It’s essentially contrastive, pointing toward “de facto”.

How then does “de facto” work? If a judgment is merely de facto IEM that means that it is not IEM (it is VEM), but the errors of misidentification in virtue of which it is not IEM only occur in worlds very different from ours. Saying that a judgment is de facto IEM is thus more like saying that the Moon is de facto immune from the spread of C-19 (which does not imply that the moon is immune from the spread of C-19) than like saying that someone is de facto in charge (which does imply that they are in charge).“De facto IEM” means something like nearly IEM or ignoring far-flung cases IEM.

Someone might prefer to say that “de facto IEM” really is a kind of IEM, and “logical IEM” is just another more demanding kind of IEM. This wouldn’t affect my argument, which will focus on how we test for the higher status, “logical IEM”, whether or not we think that “logical IEM” and “IEM” pick out the same status.

3. See, for example, Coliva (2006), García-Carpintero (2018), and Wright et al. (2012).

4. See Pryor (1999).

5. See Morgan (2019) and Smith (2006). Smith imagines that the person with the pain, whenever they are in pain, presses a button that causes the other person to feel an itch. Obviously, there are lots of ways in which the causal connection could be implemented.

6. The Simple Explanation says that immunity to error through misidentification is explained by a judgment not being grounded in an identity, as opposed to by some claim about the nature of the self or the nature of mental contents. One challenge Coliva and Palmira identity (2024) for the Simple Explanation in their introduction is whether it can account for the distinction between logical and de facto IEM, which they treat as a desideratum on any explanation of IEM. To anticipate, my picture of memory involves denying that there is such a distinction. If this denial is correct, that desideratum disappears. Another challenge, which I don’t offer a response to but Coliva and Palmira do, is that the Simple Explanation struggles with wh-misidentification. For this criticism, see Merlo (2017).

7. See Morgan (2019), and McGlynn (2021).

8. “Shot in the dark” is Evans’s phrase, used in his discussion of Shoemaker’s case, at (Evans, 1982, p. 246)