Wednesday, 23 April 2014

Memory and Exemplification

In a just-published paper, Dan Hutto discusses the commonplace interpretation of the brain as a vast storehouse of content, a memory bank of images, ideas and impressions. He finds this characterisation to be as unilluminating and inconsistent as it is antiquated and he suggests that this misconception continues to hinder insights in both philosophy and cognitive science. He remarks that there are better ways to conceive of memory, especially in the case of basic minds and he claims that simple kinds of memory — he cites rats as an example — might only require the ability to "re-enact embodied procedures" by using "local landmarks".

When a rat "re-enacts" its route around a maze, there is no doubt that certain features of the maze have causal influence over the behaviour of the rat — change the features of the maze and you change the behaviour of the rat. And Hutto is right to question whether we need to invoke mental charts, maps or schemas etc. to explain such successful navigation. All that need be assumed is that during the initial investigation of the maze, certain dispositions form in the rat that cause it to respond in certain ways on encountering these features again. The question of whether these dispositions constitute representations is simply more extravagant than is necessary.

Hutto argues that in the case of more sophisticated minds, it is the mastery of socially supported narrative practices that permit the 'reconstruction of past experiences' (i.e. memory) but he makes no indication of the more rudimentary practices that must have preceded narrative. Dance and procedural guidance in particular suggest themselves as obvious antecedents to the skills of narrative description, and it is puzzling therefore that Hutto makes no mention of these. Perhaps his earlier work (2008) on narrative practices and their importance for the development of 'folk psychological' skills — for inferring the thoughts and intentions of others — continues to dominate his thinking. However, applying this same argument to the case of memory is unconvincing to say the least. The engine simply doesn't fit the chassis.

Is it plausible that dance only ever emerged as a consequence of our linguistic powers? And what about tuneful vocalisations? I find it hard to believe that these preeminently social, embodied and enactive behaviours didn't precede the practice of storytelling. And mimicry? Of all the intelligent memory-implicating behaviours that we commonly observe on the part of animals, mimicry must be one of the most obvious. How, I wonder, might a creature successfully mimic a sequence of behaviours enacted by another animal if not by virtue of capacities significantly in excess of the simple behavioural responsiveness of the maze-navigating-rat?

In a 2011 book entitled "Words and Images" Christopher Gauker provides a simple description of the sequence of operations involved in fixing an ordinary tap (though he neglects to mention first switching the water off at the mains). He points out that whilst he is able to find words to describe the procedure, he could just as easily have imagined it without using any words at all. This is surely right — we "find" words to describe our memories and imaginings, not the other way around. If we wish to argue that words precede such imaginings, then infinite regress looms — as Gauker is quick to point out.

Gauker argues that animals and nonlinguistic human infants do not think conceptually but that they do think "imagistically". Imagistic cognition has also been discussed by Hutto in the past so it might be worth examining what is meant by the term. Neither Gauker nor Hutto take imagistic thinking to be literally pictorial, yet precisely what they do take it to be, they give no clear indication. Gauker states that mental imagery is "similar" to the things it represents whilst Hutto claims that it "resembles" (2008, p.81). The two terms (resemblance and similarity) are interchangeable in their vagueness and, like all blunt instruments, they are far more likely to do harm than good in the elucidation of the vexed issues of imagination, memory and consciousness. What I have found to be far more illuminating in this regard is to think of mental imagery in terms of what we could call latent exemplification. So, to imagine or remember something is to engage the cognitive component of skills that allow us to demonstrate, enact, perform, indicate, simulate, picture or select the things with which we are causally engaged. Questions of resemblance or similarity are immaterial in his context. What matters is that such imaginings involve — indeed are constituted by — many of the same brain/body responses (though in diminished form) as would arise as a consequence of actually encountering the thing or circumstance imagined. This also explains why tests of the actual accuracy and detail of imaginings invariably turn out to be disappointing in comparison with their reported vividness and why theorists like Gauker so often assume imagination to be representational. The assumption is understandable but unnecessary (though, admittedly, it is extremely convenient to regard capacities of exemplification as representations). 

We must be careful though. Imagination is not an inner module, theatre or "similarity space" for the private display of representations and resemblances. It is the evolutionary consequence, over countless generations, of a history of exemplification and more recently, as Hutto correctly observes, of linguistic and narrative practices. Moreover, these forms and practices of exemplification have emerged in the context of evolving inhibitory capacities that suppress physical actions of exemplification whilst allowing other associated brain processes to proceed as normal. It is these brain processes — linked inextricably to practices of exemplification — that must be considered as primary candidates in the constitution of thought and consciousness.

Ultimately I think Hutto gets a lot right about memory and I think his rejection of theories of mental content especially is well grounded and well argued but I think his focus upon narrative practices — derived from his earlier research — limits his view of a more extensive history of practices of exemplification.


Post a Comment