Distorted Memory: Interview with John Sutton

I interviewed John Sutton, Professor of Cognitive Science at the ARC Centre for Excellence in Cognition and its Disorders at Macquarie University, Sydney. John is interested in memory, skill, and distributed cognition, and in his work he seeks to integrate philosophical, psychological, and historical ideas and methods. This is the first in a series of three posts.

ES-B: Do you think that distorted autobiographical memories might have pragmatic benefits insofar as they may function to fill gaps in a person’s recollection of the past, or are distorted in a self-enhancing way?

JS: "I think that’s a very difficult question, and it’s probably going to pan out differently for different kinds of memory pathology. In general, yes, for sure, the mechanisms which give rise to distortions can be adaptive mechanisms, distortions might just be a by-product of mechanisms that work well enough most of the time.

I am uneasy about pushing the line that specific distortions themselves are adaptive too much just because for somebody who stresses the constructive nature of memory, I also want to really underline that remembering is making a claim on the past, memory can be wrong, and that’s not just a philosopher’s point, it’s something that really matters subjectively. I think it’s easy for people like me who talk a lot about constructive processes to lose sight of that fact, that memories are meant to be true, and that phenomenologically when we get some doubt about something we seem to remember that’s something that we worry at, we don’t leave it unresolved."

ES-B: In the project we have been thinking about the potential epistemic benefits of distorted memories. For example, in the clinical domain, subjects with dementia might have memory distortions which are self-enhancing, which might increase wellbeing, reduce anxiety, making them epistemically more able. Do you find this plausible?

JS: "I don’t know if you’d expect there to be any general answer to that question, it’s something where there are important individual difference variables about how much you care about your personal identity being anchored in correspondence with reality, rather than in a good story. With distorted memories, for example, we had a talk from Nicole Muller, a linguist who works with Alzheimer’s patients who spoke of two older women in a dementia care facility in Louisiana. When you read the transcript of their conversation with each other, actually you wouldn’t know there was anything substantial wrong with them. But it turns out that lots of the details in their exchange are known to be false.

So what’s happening in that interaction is something quite different from actually using truths about the past. They are just expressing an intimacy and a sense of friendship and mutual support in the present. So within their social world I think that they probably do have responsibilities to each other as a result of this shared framework which is maybe a partly confabulated framework but it’s tapping into something that’s true of them now."

ES-B: Do you have any thoughts on memory and epistemic innocence?

JS: "I understand that part of the general motivation behind the project is to look at cases in which theorists have depicted certain kinds of behaviour or cognition as irrational and therefore bad and saying hang on, when we look at the processes that are producing these tendencies, can’t we tell a richer story in which, if you isolate the scenario, those processes are actually quite functional? So, for example, in a false memory experiment like Lost in the Mall, where adults are falsely told that they were lost in a mall as a kid, these are not normal scenarios in life. Sure there are analogous things that can happen but in general we live in a relatively benign epistemic world and so it’s not that irrational to trust other people and to trust our environment most of the time from an epistemic point of view.

So I’m interested in seeing how much more liberal one can go. There are lots of cases in which cognition works well enough most of the time even though you can create settings in which it really doesn’t work—in cases like memory where there are normative constraints in play, you want to hang on to the point that it is important to get things right as well. So maybe you need to de-couple the question of whether the process is more or less rational in general from an investigation into the particular ways in which it could go wrong."