Thursday, November 18, 2010

Costall on Gibson's Attack on Cognitivism

James Gibson engaged in a sustained attack upon cognitivism over many years, from the thirties until his death in 1979 (Costall, 1984, p. 110).
This is interesting, since I think many cognitivists date the cognitive revolution from the publication of Chomsky's review of Verbal Behavior.  I don't want to insist that cognitivism began in 1959, but I do wonder what Costall had in mind by "cognitivism" such that it would have been a view Gibson was attacking in the thirties.  I don't know that much about psychology in those days.   Anyone have references?  Else I might have to do research myself ...  Or, maybe Costall's idea is that Gibson's work began to undermine the cognitivist approach before cognitivism was even born.

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. Ken,

    Edward Tolman is sometimes seen as the first cognitive psychologist. He was definitely doing research in the 30s, but his most well known work (looking at mental modelling in rodents) wasn't published until the late 40s.

  2. Thanks, Chris. That makes sense (at least to my historically uninformed mind). Gibson does not seem to be one just to attack cognitivism in a narrow sense, but instead all of cognitive psychology. And behaviorists are also probably interested in such a broad rejection.

  3. Ken,

    Gibson was highly critical of the "Association psychologies" that were really popular during William James' time and which were still dominating the German experimental psychology scene in the early 20th century. The key idea of an Association psychology is to look at how simple Ideas (sense-data) are manipulated by the Mind (computer) and transformed into complex Ideas (perceptions, judgements) which then cause motor commands, like a centralized controller. If this sounds remarkably like modern day cognitive computationalism, that's because it is. It's the exact same Associationist hypotheses and assumptions, only updated with modern computer metaphors and notions of information (sense data) being "processed" by the brain into meaningful perceptions based on theoretical inferences (Marr is a good example of this approach, following the long tradition of Helmholtz). But the explanatory value is exactly the same. Which means that cutting edge cognitive psychology is been wedded to a preDarwinian psychological theory based on classic faculty psychology and hidden Cartesian-Lockean assumptions.

  4. Gary,
    I had gleaned from some of Gibson that he was interested in beating up on sense datum theories and had suspected that he equates those with computational theories. So, it seems to me to be pretty plausible to think that Gibson is up to what you say.

    Nevertheless, I think that there are important differences between early 20th Century Associationism and contemporary cognitivism. To put things crudely, I think that for the Associationists, ideas are some like "little conscious mental experiences", but for the computationalists representations are often non-conscious mental experiences. Some mental representations are conscious, others are not. So, I think there are mental representations in visual pathway to V1, but I don't think that they are conscious.

    Think of Fodor, 1980. The information processing modules are not subject to introspection. (See, e.g, pp. 55-60.) All one gets out of them is something like distal layout of 3D objects.

  5. Great resources here. I’ll be back for the next your posting. keep writing and happy blogging.

  6. Thanks, Liberation.

    It's surprising to me, but these Costall posts have garnered a lot of attention.