Double dissociations (Dan Reisberg) ch. 5

The chapter describes a number of results showing that implicit memories are different from explicit ones. Thus, for example, certain forms of brain damage disrupt explicit memory but leave implicit memory intact. Likewise, we know that explicit memory is strongly influenced by level of processing during encoding, whereas implicit memory may not be. How should we think about findings like these?

It turns out that these results are ambiguous, and the ambiguity leads us to an important methodological issue: On the one side, we can read these results as suggesting that implicit memory is fundamentally different from explicit memory-governed by its own principles and served by separate portions of the brain. But on the other side, perhaps implicit and explicit memory are not different types at all. Perhaps they are fundamentally the same, obeying the same rules and principles. In that case, the memories we call "explicit" might simply be a more fragile version of this single type of memory, and hence more easily influenced by external factors, such as brain damage or level of processing.

The issue at stake here is a crucial one, and often arises in science: What is being asked, in essence, is whether the difference between the memory types is "qualitative" or "quantitative," and it's important to get this straight. Claiming a qualitative difference is equivalent to claiming that the two are different "species" of memory. In that case, we will need different theories for implicit and explicit memory, and we'll simply confuse ourselves by classing the two types of memory together, seeking principles that apply to both. In contrast, claiming a quantitative difference is equivalent to saying the two are fundamentally similar, differing only in some "adjustment" or "parameter." In this case, we'd be wasting our time if we search for separate governing principles, because the same principles apply to both.

How can we resolve this ambiguity? The key lies in realizing that the facts we have mentioned so far (the effects of brain damage or of level of processing) are concerned with ways we can influence explicit memory with no effect on implicit. To demonstrate a qualitative difference, we also need the reverse result-cases in which we can influence implicit memory but not explicit. This provides what the chapter calls a double dissociation and allows us to rule out the possibility that explicit memories are simply more fragile than implicit (and hence more readily influenced by any manipulation). This in turn allows us to conclude that the difference is truly a difference in type, and not just a quantitative difference.

As the chapter mentions, we do have the double dissociation for implicit and explicit memory: Certain factors (including forms of brain damage) influence explicit memory but not implicit; other factors (and other forms of brain damage) influence implicit memory but not explicit. This makes it clear that neither type of memory is, in general, easier to influence than the other. This allows us to set aside the hypothesis that one of these types of memory is simply a weaker version of the other, or a more fragile version. Instead, the evidence tells us that each type of memories is indeed affected by its own set of factors. That's what tells us that the two do follow different rules and are governed by different principles, and so truly are qualitatively different from each other.