An Argument for Memory Traces

After it has been established that a causal condition – however it might be spelled out – it indispensable for explaining remembering, we can proceed by examining the exact nature of memory causation.

According to the standard view, the causal process connecting a past representation and its subsequent recall involves intermediary memory traces (or engrams). The stipulation of memory traces is motivated by the contention that between any two diachronic mental events there have to be a series of intermediary events, each of which causes the next, and each of which is temporally contiguous to the next. However, in The Analysis of Mind, Bertrand Russell argued that since the physiological evidence for memory traces is not yet quite conclusive, it is prudent to come up with an account of memory causation that manages without the stipulation of memory traces. Russell’s name for this notion of memory causation is “mnemic causation”. According to the theory of mnemic causation, a past experience is directly causally active over a temporal distance; past experiences are proximal causes of states of recalling. The aim of this chapter is to refute the theory of mnemic causation and to establish that the stipulation of memory traces is indispensable for the analysis of memory.

Sections 3.1 and 3.2 motivate the standard view that causality implies contiguity and explicate the notion of a memory trace. Section 3.3 is a discussion of the verifiability of memory traces. Section 3.4 provides a detailed account of Russell’s theory of mnemic causation. The concept of mnemic causation is defended against the widespread objection that direct causal action at a distance in time is impossible because a cause cannot operate when it has ceased to exist. Given that Russell’s notion of mnemic causation does not face any crucial difficulties and given that there is no conclusive empirical evidence for the existence of memory traces, how should we determine whether or not to stipulate memory traces? It is argued that the apparent tie between the theory of mnemic causation and the theory of contiguous causation can be resolved when memory traces are treated as theoretical constructs.