Friday, 16 April 2010

Reading Group - Heft (2001) William James and Radical Empircism

This book got heavy, fast! The first full chapter discusses the metaphysics of William James (pictured in his younger, cooler days) and how it is the philosophical groundwork for Gibson. Assuming Heft's analysis of James is right, this is precisely what it is.

I won't review all the detailed argument because there is a lot of material in this chapter. What I will do is summarise the key points.

The goal of James' metaphysics (theory of the way the universe is composed) is to replace the dualism he and others like Dewey knew was (and still is) lurking at the heart of psychology. Metaphysical dualism is the assumption that there are two kinds of stuff in the universe - for Descartes it was soul and world, for psychology it was mind and body. Dualism is a major problem, if true, because of one important point - if there are indeed two types of thing in the universe, how can something of one type (e.g. a mental state) come to have knowledge about something of another type (e.g. an event in the world?). It is, in fact, impossible by definition, and for this reason any line of reasoning that entails a dualism is generally frowned upon in philosophy.

There have been two basic solutions to dualism in psychology. Behaviourism (beginning with Watson) made the claim that there was only one kind of stuff (observable, physical events) and that there was no mental life. Cognitive psychology recognised that this was absurd, but then has felt the need to reduce psychological states to neural ones in order to make naturalise them (and hence make them an object for scientific study). Both solutions treat 'mental states' as problematic, things that smack of 'soul' and must be made to reflect something more 'real', such as a physically observable state.

Some philosophy of science
James' solution is grounded in the notion that mental states are already perfectly natural things, perfectly part of the natural world. He, unlike modern psychology, had learned the lessons of evolution; organisms are entirely natural, with no intelligent designer around to introduce any metaphysically dodgy mechanisms. We are part of this natural world, so it seems safe to assume (at least on a first pass) that all our attributes are also part of the natural world. The attributes must all be the same kind of thing (the metaphysics must be a monism). James was also an empiricist, which metaphysically means the only kinds of entities you can posit are those available via experience (contrast to the idealism of Kant, who claimed that there must be structure imposed on experience in order to constrain the possible interpretations of experience; this was a reply to Hume, who noted that empiricism as generally practiced in philosophy was not constrained enough).

Hume and Kant both assume that the world of experience is, initially, composed of separate elements from which order must be regained. Kant required this order to be imposed (the kind of work mental representations often do), while the British Empiricist movement wanted that order to be recovered via association (which these days shows up in  the form of fancy statistical inference, using Bayesian analyses).  Hume notes that the association process needs constraint; Kant wants imposed form. James simply notes that all their problems go away if you simply stop thinking that the primary material of experience is elemental. Physiology be damned: no system that has evolved would evolve perceptual systems that create such a metaphysical catastrophe.

The metaphysics of William James: relations are real
James' metaphysics is radical empiricism: it's empiricism because you can only posit entities that are experienceable, and it's radical because William James allows relations to be primary metaphysical, experienceable entities. This move cannot be overstated: it is an incredibly powerful move that opens everything up, but it really is a radical move.

What are relations, then? Relations entail two things (made of the same stuff, though) doing something with respect to one another. My coffee cup is sitting on my table - I can't define this relation without reference to both things. Every philosopher up until this point would have to cope with the fact that there were two 'real things' (objects) with a relation that must be imposed; James simply claims that the relation itself is exactly as real, if not more so, that the objects themselves. If it's real, then it must be experiencable; we experience relations as well as objects. William James therefore invents, right here, the notion of direct perception.

The world, then, is composed of pure experience, and pure experience contains things and the relations between things. All things stand within numerous constellations of relations: my coffee cup is on the table, and below the ceiling, and within my reach, and so on. Many of these transient relations leave imprints: my coffee cup has coffee in it from when it entered the right relation with my coffee pot.

We come to know about the world by an act of selection: knowing is the selection of some relations from all the relations currently available to us. This set is constrained by the way things are: pure experience is a 'quasi-chaos' (a dynamic flux that ebbs and flows as time passes and things change) but it is not random - indeed, it is a 'textured manifold' of relations and objects, and knowing is the selective process of following some but not others of these lines of texture.

If you know your Gibson, then the previous two paragraphs should have made you think of affordances and optic flow. James had the metaphysics nailed; what Gibson is going to give him is a theory of information, namely the rules by which an organism interfaces with the relations that exist in pure experience. (I'm so excited right now it's embarrassing :)

Two more pieces: percepts and concepts. James is trying to account for our experience of the world, and we are not slaves to perception. Percepts, then, are the immediate objects of perception that ebb and flow as we move around the world. Concepts are abstractions from percepts; they derive from percepts, are the same kind of stuff, but serve as a tool to fix a percept in time. They have verisimilitude (nearness to the truth) to the extent that they remain grounded in percepts, and this truth value is revealed to an organism by whether or not using them or relying on them works (James is also a pragmatist and a functionalist, philosophically).

There's a lot here, and there's still a little more to talk about, but I will digest more and bring it up as we go. A lot of what James is saying sounds, at a first past, like pseudo-science babble - the world is pure experience? What the hell? But Heft conveys clearly the fact that James spent a lot of time and effort making it rigourous and reasonable, and I have to say that knowing what I do about Gibson and seeing how it relates to James' metaphysics really helps keep it all grounded. Relations seem like such nebulous things, too nebulous to serve as an object of perception. But an affordance is a relation*, and the act of perceiving an affordance is the process of entering into a relation with the constituents of the affordance; we also got there by following the textured flow of relations between where we started and where we are now.

Gibson apparently wasn't a major James scholar or anything; his connection is via his mentor, James' student Edwin B Holt. His contribution is the topic of the next chapter.

Dualisms are bad. Radical empiricism replaces dualism by claiming that relations between objects have the same metaphysical/ontological status as the objects themselves; possibly more. The world, as far as any organism can ever know, is composed of pure experience. Pure experience contains structure (objects and their relations) and these are directly perceived (percepts).  The essence of knowing is 'being there', i.e. in the relation you selected from all the possible ones that pure experience provided; "The word 'I', then, is primarily a noun of position, just like 'this' and 'here'" (James, 1912). Radical empiricism is the metaphysics required for Gibson to be right; ecological psychology is then a theory of the information, the interface, the mechanism by which we can select and enter a relation.

Summary of the summary
William James is all that and then some.

*Actually this is a little complicated; more soon


  1. Great summary. Thanks so much. I remember somewhere that Heft talked of James' philosophical stance as 'neutral monism'. I read a bit on that in the Standford Encyclopedia of Philosophy and was surprised at the careful definition and of its marginal status in philosophy.

    I think the question of how the concept evolved from the percept is key. This is why and where I appreciate Mead so much. I also make a connection between direct perception as a conscious state associated with the percept and knowledge of the 'world' as a function of the concept. Where I am working is in that relation and its effect on my own perceptions, awareness, and beliefs.

    Again, thanks for all your work (and interest) with this.

  2. The Encyclopedia entry:

    knowledge of the 'world' as a function of the concept
    I don't think James gives this job entirely to the concept. In fact, the concept is a clearly secondary entity, derived from the percept. You can know via either, but the possibility always exists that a concept can be a false belief, because it's been separated from any percepts.