Having just finished rereading Andy Clark's Being There, I want to take time out to clarify where I stand with regard to his position.
[Note that I'm focusing mostly on Being There (1997) because it goes into a little more detail about representation than the more recent Supersizing the Mind (2008). I also personally prefer the previous book. The major claims of the two publications do not seem to differ substantially, at least insofar as they are relevant to the points being made in this blog.]
These days, plenty of cognitive scientists will acknowledge the limitations of old-school philosophical thinking about representations, the idea that abstract symbols inside our heads are manipulated through something resembling deductive reasoning or formal logic. (Fodor's "Language of Thought" is the classic example.) Clark points out that the way internal representation is conceptualized in contemporary neuroscience "constitutes an image of the representing brain that is far, far removed from the old idea of a single, symbolic inner code." (Being There, pages 141-2). I want to be absolutely clear that my arguments apply equally well to the new and improved, partially embodied view of representation that Clark endorses. The following types of statements from Being There are subject to the same fatal flaws as anything Fodor ever claimed:
In "cases that involve reasoning about absent, nonexistent, or counterfactual states of affairs… it is hard to avoid the conclusion that successful reasoning involves creating some kind of prior and identifiable stand-ins for the absent phenomena." (page 167)
"…the common feature is the need to generate an additional internal state whose information-processing adaptive role is to guide behavior despite the effective unfriendliness of the ambient environmental signals… In these representation-hungry cases, the system must, it seems, create some kind of inner item, pattern, or process whose role is to stand in for the elusive state of affiars." (page 168)
Clark has done an admirable job in demonstrating the need to take the whole body and environment into account in order to understand many aspects of human behavior and cognition. His popularization and defense of embodied cognition has opened up useful debates and helped to clear up some pervasive misconceptions about how human beings actually work. Nonetheless, I believe that any true proponent of radical embodied cognitive science must give Clark a Stephen Colbert–style Wag of the Finger, for two reasons:
1 - He ought to know better. It's not a far leap from some of the examples and reasoning he gives in service of embodied cognition to others that clearly support a more radical stance — making it all the more frustrating when in the end he misses the point.
2 - His arguments actively draw attention away from a more radical stance. It's not just that Clark criticizes this stance — for instance, in Chapter 8 of Being There, where he spends a fair bit of time critiquing all the arguments he can think of for radicalism. (Note that this blog relies on none of those.) This I don't have a problem with, since it helps spur debate and keep everyone on their toes. It's not even that Clark pooh-poohs the radical stance. ("The thesis of Radical Embodied Cognition is thus, it seems, a genuinely held view," he says. In other words, "Apparently some otherwise sane people actually believe this crazy idea.")
Much more damaging, I believe, is the way Clark's arguments help to limit the terms of the embodied cognition debate. Although I agree with the major claim he's making (that some resources outside the brain may best be considered part of the extended circuitry of the mind), it is 1) quite limited, compared to a full-fledged radical stance, and 2) essentially a matter of opinion and definition. (I'll discuss that second point in more depth when I address empty questions.) Apparently this limited stance is still quite contentious — in Supersizing the Mind, he puts a lot of effort into defending it. But even if Clark were completely persuasive and everyone came to agree with him, people would be no closer to understanding or embracing the more radical viewpoint. Moreover, I don't think Clark can be completely persuasive (in matters of opinions/definition, that tends not to happen), so this argument could easily go on and on indefinitely; energy that could otherwise be spent considering substantive issues about how aspects of mental life come about will be channeled into debating an empty question.
Perhaps my biggest gripe is that as long as Andy Clark is the most prominent advocate of an embodied viewpoint, it can seem as though his non-radical perspective is a major breakthrough against traditional conceptions, as though the hard work of rethinking is all behind us — when in fact he still leaves the problematic core of representationalist assumptions untouched.