Self-evident interfaces and perceivable possibilities

In attempting to understand Durrell Bishop's research into self-evident design with more clarity I thank Fusewire and Alec for their comments to my previous post. Cultural mediation and perception appear to be two aspects that need considering with Bishop's thesis. With further reading around the subject I have read highlights of an interview with Bill Gaber, professor of design at Goldsmiths College in London. In this interview he cited the work of perception theorist J.J. Gibson, in particular Gibson's 1979 book on The Ecological Approach to Visual Perception.

As I have an interest in human evolution and the development of human creativity (in the words of Douglas Adams "the secret is to keep banging the rocks together guys!") Gaber's thesis was understandably of interest. If, as Gaber hypothesizes, our evolution has allowed our species to benefit from perceiving what actions would be useful to our survival over those that wont, we are able to perceive, validate and combine complicated actions together such as our ancestors' stone tool-making. This outcome of perceiving the possibilities when flint is struck with another stone, mammoth meat can then be butchered for the tribe back home is disseminated through perceptual experimentation and then cultural-mediation. Each stone age tribe makes stone tools with their own twist on the technology, evolving it through an ongoing iterative process of constant development. This perceptual process is the same as we use to evolve interactive design.

Gaber in his research and work when at Xerox EuroPARC has tried to demonstrate that people can deal with graphical elements within interface designs as if they are self-evident physical objects by "thinking about the interface being a physical environment rather than being a kind of command-line-driven conversational metaphor". This is more than just creating an icon of something from the physical world and placing it into the visual design of an interface. A rewind button on a tape deck does not function in the same way as a backwards button on a physical or digital CD player. A scrub wheel does not function in exactly the same way on an iPod as it does on an analogue device. The physical function may be similar but the perceived outcome may not match the outcome. Gibson's theory refers to these perceivable possibilities for action as "affordances" and Gaber has applied Gibson's affordance theory to interactive design by attempting to peel back "the notion of affordances to its real essentials".

Has anyone read Gibson's work? I would be interested to hear if his theory has been advanced further or has been dismissed in favour of a more robust thesis. Any further reading on this matter would be greatly appreciated by myself.