Wednesday, November 17, 2010

Costall too fears Cognitivist Homunculi

In his review of Gibson, 1979, Costall also writes,
Interestingly enough, here there is no appeal to the usual argument for the ambiguity of perception. Here we find the uncritical retention of another aspect of the Cartesian scheme, the notion of a mind lurking within the body, in direct contact only with the body and not with the environment itself. This notion, as Reed has recently argued, derives from the Cartesian hypothesis of corporeal ideas (Reed, 1982). Gibson's own criticisms of this assumption-for example, in his discussion of the visual control of manipulation-echo the important arguments Skinner has voiced over many years concerning the persuasive myth of the "inner man" (e.g., Skinner, 1938, chapter I):
The movements of the hands do not consist of responses to stimuli .... Is the only alternative to think of the hands as instruments of the mind? Piaget, for example, sometimes seems to imply that the hands are tools of a child's intelligence. But this is like saying that the hand is a tool of an inner child in more or less the same way that an object is a tool for a child with hands. This is surely an error. The alternative is not a return to mentalism. We should think of the hands as neither triggered nor commanded but controlled. (Gibson, 1979, p. 235)  
(Costall, 1984, p. 114).
Homunculi and "Cartesianism", I've found, are a big concern of EPists.  I guess I have imbided enough cognitivism to think that it doesn't lead to homunculi.  It seems to me that computers do image processing without homunculi, so why not humans?  Indeed, if cognitive processing is a species of computation, then why humans not do image processing without homunculi?

Or, put the matter another way.  I think a cognitivist could be perfectly happy with Gibson's idea that we should think of the hands as neither triggered nor commanded but controlled.  They are controlled by cognitive processes within the brain.

Costall, A. (1984). Are theories of perception necessary? A review of Gibson's The Ecological Approach to Visual Perception. Journal of the experimental analysis of behavior, 41(1), 109.


  1. Computers can only process images because they have been programmed to by humans. This harks back to my previous question: where do representations (programmes) get their content if it's not in the proximal stimulus? This matters because their content is generally about resolving the ambiguity in the proximal stimulus, so the two go hand in hand.

  2. So, it sounds like your objection is that cognitive processes can't be computational, because (unlike the store-bought electronic digital computers) there is no programmer for the brain. Is that it?

  3. The objection is more general; I was just relating to the specific example. The objection is if the representation's job is to enrich the proximal stimulus so as to resolve the ambiguities there, how did the enriching content get in there? If the stimulus is rich enough a neural network kind of system might be able to programme itself; but then the question is why bother, when the proximal stimulus is rich enough?

    The answer for computers is 'people put it there'. The answer for internal representation is...?

  4. Ok. So, your objection is that *at some point* in evolutionary history information from the environment must have been picked up and incorporated into the organism. But, this seems to me to be entirely gratuitous.

    The mechanisms for lateral inhibition, for example, need not have been based on information in the environment. The idea would be simply that those organisms in a population that used lateral inhibition survived better that those organisms in the population that did not. This doesn't have to involve extracting information from the environment.

  5. What about mental representations that we supposedly acquire in developmental time?

  6. Cognitivism allows for both innate knowledge and learned knowledge. Think of language acquisition. That standard cognitivist line is that humans some some innate knowledge of language, maybe that their language is going to contain a genitive construction, but what they have to learn by exposure to a language is how their language forms the genitive.

    Is Gibson, like many other behaviorists, committed to an empiricism in which all information must be developmentally or phylogenetically gleaned from the environment?

  7. Yes, he's very much an empiricist. That's why it was so important to show how the environment furnished adequate information.