>> Attention and Memory

	Interview with Moris Moscovitch, University of Toronto From Studying The Mind, VHS © 2003, W. W. Norton

What is the relationship between attention and memory?

Attention seems to be a crucial component if the medial temporal lobes are going to pick up information that's in consciousness. What gets into working memory may be largely determined by what the frontal lobes are doing, and what their working memory capacities are. So, the prefrontal cortex is crucial for allocating attention, directing it, and holding the information in working memory so that the hippocampal system may be able to pick it up. Anything that divides your attention will reduce how much information about the target task is in working memory; the hippocampus won't pick it up, and memory will suffer.

At the retrieval end, things seem to be much more interesting. Retrieval seems to be an effortful process. Yet, if you divide attention at retrieval it seems to have absolutely no effect on memory. Why not? One reason is that at retrieval the cues are already available, they're minimal, and they immediately activate the hippocampus and related structures, so memory comes out automatically. Dividing attention at retrieval doesn't have very much of an effect unless the information that's dividing your attention is competing for the cortical structures that represent the memory. In that case, the function of those structures is taken up by the divided attention task. So, at retrieval there has to be a lot of specificity between the memory task and the divided attention task to get interference—interference occurs only if they share the same domain.

Autobiographical memory is another way of talking about episodic memory, but it's that part of episodic memory where you actually re-experience an event. Not just that you know the event happened, but that you could actually re-experience it. When patients with a lesion to the prefrontal cortex are given the cue "Describe an autobiographical event" they may not be able to develop a good search strategy, because that cue is very general. Even cues such as "Describe an accident you had when you were young" or "Describe an embarrassing incident" are still pretty general, and may give access to some information, but not all of the rich detail. The prefrontal cortex helps query the medial temporal lobe system to get those details out, and so patients with damage to prefrontal cortex will be impaired in recovering that detail. If, however, you hints at what might have been happening in that situation, then you as the interrogator are acting as the frontal lobes for that person. And then the detail comes out.

How do neuropsychological studies help us understand memory function?

Imaging studies provide information about the structures that are correlated with particular functions. What they don't do very well, at least not yet, is tell us which structures are crucial for performing a particular function. And in those cases, we still have to rely on neuropsychological evidence from patients with lesions to determine whether the structure that's been activated is essential for carrying out the particular function that we're investigating, or whether it's along for the ride. And while the structure might contribute somewhat, it may not be as essential as imaging studies suggest. For example, consider the ubiquitous activation of the prefrontal cortex on tests of memory. What we find, at least from neuropsychological investigation of patients with lesions, is that lesions in even large areas of prefrontal cortex seem not to impair performance on a variety of memory tasks, yet, based on imaging studies, one would not expect this. I should point out that one other technique that's coming into fashion is transcranial magnetic stimulation which, depending on the stimulation, can either mimic a lesion on a particular site, or enhance the activity and performance of a particular site. So, coming around full circle, I think you need converging evidence from all these different techniques—and others—to get at what the brain is doing.