Flage on Hume's Account of Memory

166, FLAGE ON HUME'S ACCOUNT OF MEMORY In the Treatise Hume writes that an impression which "has been present with the mind" may "make its appearance there as an idea," and that it can appear either through the faculty of memory or the faculty of the imagination. Memory and imagination each produces its own species of idea. In "Hume on Memory and 2 Causation" Daniel Flage addresses Hume's carving out of these two species. Flage's main purpose is to show that Hume has a criterion for distinguishing memory ideas from others. I will argue that Flage has not located necessary and sufficient conditions for memory in Hume's discussion of memory. I will show that if Hume intended to provide a criterion at all, it is not the criterion Flage attributes to him, and that the criterion Flage has formulated is not an improvement over what appears to be Hume's criterion. Before looking at the details of Flage's interpretation I want to mention briefly an underlying assumption of the paper. The assumption is that Hume believes that there must be a clear distinction between memories and other perceptions. The assumption is supported by the fact that Hume speaks of the species of memory. However, there is textual evidence against this assumption. In the Treatise Hume emphasizes the ways in which memories are like other perceptions. For example, Hume calls memories impressions, while in distinguishing impressions from ideas he introduces 3 memories as paradigm examples of ideas . The reason Hume calls memories impressions, I believe, is that for the purpose of inference, he takes memories to be equivalent to impressions of sense. Since Hume is willing to put memories in two seemingly incompatible categories, perhaps his primary concern is with the role they play in belief; we make causal inferences 167. from memory and sense, not from the imagination. That Hume's primary concern is with inference, and not with memory per se, is supported by the fact that both Treatise sections devoted to memory are followed by 4 sections which deal with causal inference. Hume thinks that we cannot always distinguish memories from other perceptions "in their operation." It is not clear that he thinks that there must be a sharp distinction at all, since the traditional epistemological purpose for such a distinction, to partition the veridical from the non-veridical, is not obviously Hume's purpose. However, Flage claims that there is a clear criterion of memory in the Treatise. Let's take a look at what he has found. Flage maintains that Hume has both a phenomenal criterion based on vivacity which allows us in a rough and ready way to distinguish memories from imaginings "in their operation," and a formal criterion which "provides one with insight into the complex structure of an idea of memory." (p. 169) The phenomenal criterion does not distinguish memory ideas from others, since, as Hume himself allows, non-memories (e.g. beliefs, delusions, even ideas of the imagination) can have the same vivacity as memories. (Cf. T86) The marking off of memory as a distinct species is accomplished, on Flage's view, by the formal criterion. Hume writes "The chief exercise of the memory is not to preserve the simple ideas, but their order and position." (T9) This is what Flage calls Hume's formal criterion, and Flage takes it as both necessary and sufficient for being a memory idea. Flage takes memory ideas to be what he calls relative ideas, ideas which pick out or refer to a unique item. A memory idea, on this view, is a complex idea which uniquely picks out the antecedent complex 168, impression which resembles and is the cause of the idea. Flage writes: The formal criterion is basically a causal thesis: if a particular positive idea (mental image) is a genuine idea of the memory, then at some point in the past there was a complex impression that was the cause (or the original cause) of and exactly (or closely) resembles that positive idea as an idea. (p. 172) In this formulation the criterion is a necessary condition of memory. In a subsequent formulation Flage claims its sufficiency as...