Monday 26 April 2010

Reading Group - Heft (2001) on EB Holt

In Chapter 2, Heft examines the life and work of James student EB Holt, who links Gibson to James by being the student of the latter and the graduate advisor of the former. Heft discusses Holt's work and frames the discussion to show how the key ideas from radical empiricism come through to feature so heavily in Gibson's psychology.

A lot of the fundamentals should now be familiar: Holt follows James in denying any type of dualism and in considering cognition to be a completely natural phenomenon. Holt is a realist, considers relations to be real and where the action is, and (eventually) considers all psychological phenomena to be the result of dynamic, time extended interactions between an active organism and a structured environment.  It's no coincidence that Gibson described himself as a 'Holtian philosophical behaviourist'!


A lot of Holt's language would fit perfectly into modern dynamical systems accounts of perception, action & cognition (although Holt lacked the mathematical framework and had therefore not fallen prey to the common error of getting overly fascinated with the fancy toys). Perception, says Holt, is a selective process of establishing an external relation between two orders of being (not types of thing), the psychological and the physical. This relation is the intersection of the two manifolds (a surface; he's trying to convey that the relation has extent over space and time, and isn't just in the instant). He demonstrates what he means with an analogy, of a boat shining a spotlight on a shore (Heft, pp 75). The shoreline is one manifold that has properties and reality independent of being viewed, as do the boat and spotlight. The shoreline does not specify what gets illuminated, although it does contribute by having features of more or less interest to the boar/spotlight manifold. The latter is engaged in a selection process (James' sine qua non of cognition). The intersection (the relation of interest) is fully described as 'the parts of the shoreline illuminated by the spotlight from the boat'; it is defined entirely as a relation (a transaction) and requires all the parts to be a correct description.

This is a key theme for Holt: all perception and cognition occurs relationally within nested events, and if your focus is too molecular (here meaning a focus on the parts, rather than the molar whole) then you will simply not be able to comprehend what is happening and why, and you will have to enrich (via representations) your account in order to provide the organism with all the required information. Holt insists that you must always first ask 'what is an organism doing?', and only then will the correct answer become apparent:
...a fairly accurate description of this activity will invariably reveal a law (or laws) whereby this activity is shown to be a constant function of some aspect of the objective world (Holt, 1915, pp 166)
This is a wonderful description of what Gibson (1979) is up to: he's trying to provide a 'fairly accurate description' of the environment an organism exists in so that invariant laws become apparent. It's also a succinct summary of the modern perception-action research programme - a correct characterisation of the task at hand is the essential first step (the 'outfielder problem' is a fine example of this programme in action).

Holt notes that a primary drive towards gaining more of an informative stimulus (a drive he terms adience; talk of drives is pretty typical of this time in psychology) will progressively push an organism to become more and more discriminating in what it responds to and hence more and more specific in the behaviours it produces. These behaviours become more and more nested, with the 'stimulus' actually 'causing' the behaviour becoming more and more distant in space and time. This recession of the stimulus allows behaviour to be driven by distant goals, and therefore behaviour must be analysed at this higher order in order to make any kind of sense. This idea will feature heavily in EJ and JJ Gibson's discussions of what perceptual learning is like.

An example (Holt, 1915; Heft, pp85-86):
The man is walking past my window; no I am wrong, it is not past my window that he is walking; it is to the theater; or am I wrong again? Perhaps the man is a journalist, and not the theater, nor yet the play, but the "society write-up" it is to which the creature's movements are adjusted.
Holt's point is simple: at each level of analysis the man's behaviour has different causes; but only the analysis that includes the distant goal of the "society write-up" is sufficiently comprehensive to make the man's behaviour fully understandable. Gibson will take this lesson to heart, and intention and meaning will feature heavily in his psychology, as will the extended-in-time nature of perception-action transactions.

Summary
EB Holt is the link from William James to James J Gibson. Holt taught Gibson the key lessons from James' radical empiricism; dualism is wrong & relations are real and where all the action is. He also taught Gibson about task analysis: you must ask 'what is the organism doing?', specifically 'with respect to what feature of the environment is action directed?'. Your account of the task space must be broad enough in both space and time to make the behaviour explicable without enrichment by representation, and you must keep one eye on the developmental process that might have allowed the stimulus to recede so far and yet still be a part of an organism's experience.

No comments:

Post a Comment