Here’s an easy question: What is thought?
This follows on from an earlier blog-post but since then I’ve finished reading Lawrence Shapiro’s survey of ‘Embodied Cognition’ and its relationship to conventional cognitive studies. (Shapiro’s book is great: balanced, thoughtful, clearly written – I should have read it ages ago.) The question is more problematic than it seems as different researchers in the field take different positions about what constitutes thought. On one level it seems like semantics but the semantics have profound implications.
The conventional view, as Shapiro portrays it, is of thought as a computation-like handling of representations of the external world: The external world is presented to the individual through their senses (etc..), the individual’s brain then translates the input into neural representation which it can process to deliver an output which is initially in the form of neural representations but these become translated into actions of the individual in the world. In this view the proper objects for concern and study are: What constitutes the neural representations? (and) How are these processed? (This is position is defended in William Bechtel’s paper, ‘Investigating Neural Representations: The Tale of Place Cells’). From this perspective, the body and the environment are not objects for concern per se because they are part of the input and have been translated into representations.
Shapiro reviews various challenges to this position, not merely from embodied cognition. One interesting challenge, for me, concerned gesture (p173-175). Whilst exploring the ‘Consitution Hypothesis’ (i.e. the hypothesis that some of the constituents of cognitive processes extend beyond the brain) he highlights the role of gesture in spatial communication and reasoning. Studies have indicated that, when asked to described situations involving spatial content, “[the] speech of subjects who were prevented from gesturing showed dysfluencies: it is slower than normal and filled with pauses within rather than between grammatical clauses.” Another study indicated that gender-based differences in spatial reasoning may be attributable to the greater use of gesture in spatial communication and problem solving amongst males than females. Shapiro draws from these and other studies to suggest that gesture has a role “beyond the merely communicative. Gesture, at least in some cases, seems bound to thought.” (p174)
In what follows from this, Shapiro goes on to show how the Constitution hypothesis moves from considering cognition to considering cognitive processes. He writes, “[Constitution] is a thesis about the location of the constituents of cognitive processes. The question to ask is whether constituents of a cognitive process occur outside the brain; not whether a cognitive process occurs outside the brain.” (p187-188) It is clear that this is an area of on-going and heated debate but Shapiro himself suggests (slightly surprisingly, given the nature of the debate) that “[constitution] does not compete with standard cognitive science, but pushes it to extend its boundaries further than many of it’s practitioners would have expected.” (p210)
Once you allow for cognitive processes to extend beyond the brain, then it opens up the field for study. This is quite liberating as it seems (at least to me) to allow for a reconsideration of tool-use (here thinking of tools in quite a broad sense) as part and parcel of the thinking process. This is both good, as it gives a firmer validation for the incorporation of tools into learning than a rather fuzzy invocation of ‘kinaesthetic learning’. That, in turn, requires an even-handed consideration of what is both gained and lost when certain tools are deployed. (I’m reminded of a recent story I heard about a community of story-tellers in an oral culture who forgot their stories once writing was introduced into that community.)
However, it also creates problems. One of the criticisms of the constitution hypothesis, highlighted by Shapiro, is what he called the Motley Crew Argument (p189ff). One of the points in the Motley Crew Argument is: “Processes within the brain are well-defined, in the sense that they can be an object of scientific observation, whereas processes cross the bounds of the brain are not well-defined – are a motley crew – and so cannot be an object of scientific observation.” Later on, tool-use is cited as just such a ‘motley’ situation. Amusingly enough, this is precisely what Lambros Malafouris is trying to do in his book, ‘How Things Shape the Mind’ – more on this later.