Abstract

In lieu of an abstract, here is a brief excerpt of the content:173. PERCHANCE TO DREAM: A REPLY TO TRAIGER1 In "Hume on Memory and Causation" I argued that Hume took ideas of the memory to be relative ideas corresponding to definite descriptions of the general form "the complex impression that is the (original) cause of a particular positive idea m and which exactly (or closely) resembles m, " where 'm' is a variable ranging over positive ideas (mental images). My argument was based primarily upon what I called Hume's formal criterion, i.e., Hume's contention that, unlike the imagination, memory preserves "the same order and form with the original impressions" (T9), it preserves "the original order and position of its ideas" (T85). Saul Traiger has recently taken exception to my interpretation of Hume, suggesting that my interpretation suffers from at least four defects. First, Traiger maintains that there is evidence that Hume took some ideas of the memory to be simple, while my account requires that they are all complex. Secondly, he maintains that my interpretation will not allow Hume to remember dreams or the contents of his imaginings, and Traiger contends that this provides at least prima facie evidence against my interpretation. Thirdly, he maintains that my account of the formal criterion is not sufficient to distinguish ideas of the memory from ideas of the imagination. Finally, he raises several questions regarding the temporal reference of ideas of the memory. I do not believe that any of Traiger's points require a significant modification of my account. I shall begin by discussing his third and fourth points and then focus 2 on his first and second points. 1. Memories and fictions. Traiger suggests that if my account of the formal criterion is correct, the formal criterion is not sufficient to distinguish 174, ideas of the memory from ideas of the imagination, "because there are complex ideas which are caused by resembling complex impressions but which are not memories" (FHAM, p. 170). He illustrates this by appealing to a writer of fiction who might appeal to some of his memories in constructing his fictitious world. Why, then, would not the resulting ideas of fiction count as ideas of the memory, since they meet both the causal conditions and the order and position requirements? The answer is that in the case of an idea of fiction one does not conceive of one's positive idea as a representative of some other perception. It is only because one conceives of the positive idea of which one is aware in memory as a representative of some other particular perception or perceptions that questions of form and order come into play. 2. Temporal reference. At several places in his paper, Traiger raises questions concerning the temporal reference of ideas of the memory. He suggests, first, that Hume's concern with "order" was with temporal order, rather than the order and arrangement of simple ideas in the complex positive component of an idea of the memory (FHAM, p. 168), and, secondly, that "when we remember, we place an impression in its temporal context, its place among a temporal sequence of perceptions" and this "temporal reference is generally not secured by the causal and resemblance relations an idea bears to an antecedent impression" (FHAM, p. 171). Regarding the first of these claims, it is clear that an idea of the memory can be "true" only if it represents a past impression or event, and since Hume construed representation in terms of resemblance (T233), this indicates that at least part of his concern with "order" must be with the arrangement of simple ideas in the complex positive component of an idea of the memory. Were it not for 175. such a correspondence between the "order" of simple ideas in the positive component of one's relative idea of the memory and the "order" of the simple impressions in the original impression, one's idea of the memory could not represent the original impression. Resemblance, however, plays no role in determining what the temporal position of the original impression was, and this brings us to Traiger's second point. Since Hume's account of time would seem to require that one locate a past impression within a series...