Mnemonic Confabulation

We’re continuing our series of posts on “Philosophical Perspectives on Confabulation” - our special issue in the journal Topoi this week. In today’s post, Sarah Robins, Associate Professor of Philosophy at the University of Kansas, introduces her paper “Mnemonic Confabulation”.

The motivation for this paper was the following question: How are discussions of confabulation in the philosophy of memory related to discussions of confabulation in empirical and clinical work? At first pass, it’s easy to suppose that they’re closely related. After all, both focus on confabulatory remembering. For philosophers of memory, confabulation is one of many memory errors (alongside misremembering, forgetting, relearning, etc.) that needs to be distinguished from successful remembering. 

In clinical work, interest in confabulation began with Korsakoff (1885) and Wernicke’s (1906) observations of bizarre false memory reports in patients with amnesia and dementia. Despite the shared focus on memory, the two have always struck me as distinct and difficult to put in direct conversation with one another. 

And so, in this paper, I am trying to articulate the differences I see between mnemonic confabulation on the one hand and broad confabulation on the other. Ultimately, I conclude that—as an error—mnemonic confabulation has more in common with perceptual hallucination than with the confabulatory phenomena included in standard accounts of broad confabulation.

As a philosopher of memory, my focus is on mnemonic confabulation and so the first half of the paper provides an account of mnemonic confabulation, as situated within a more general account of remembering. My own account is situated within a more broadly causal theory of remembering, but shares several points of consensus with other accounts of mnemonic confabulation (Michaelian 2016; Bernecker 2017).

Mnemonic confabulation, as I define it, occurs when there is no relation between a person’s seeming to remember a particular event or experience and any event or experience from their past—either because there is no such event in their past or because any similarity to such an event is entirely coincidental.

Then, in the second half of the paper, I turn to the ways that mnemonic confabulation is different from broad confabulation. I identify three features common to accounts of broad confabulation and argue that, for each, mnemonic confabulation lacks it. Mnemonic and broad confabulation accounts differ in the ways that they allow for veridicality; broad confabulations are ill-grounded, whereas mnemonic confabulations are not, and finally, recovery from broad confabulation is possible, at least in principle, while there is no such possibility for mnemonic confabulation.

The use of “confabulation” in discussions of broad and mnemonic confabulation are both apt, and the ensuing discussions of each have been interesting and illuminating. The best way to ensure that this continues, on both fronts, is to recognize that the two phenomena are distinct.