Lecture 15: Gestaltism III: Experience error, CNS error, Psych-neural Isomorphism, Behavioural Environment (Turvey, 2019, Lectures on Perception)

The two previous lectures reviewed some of the changes happening in related fields to psychology that were showing up in Gestalt psychology, specifically the notion of fields, the understanding that behaviour could emerge from dynamical processes, and the limits on mechanical approaches to living systems. This chapter finally actually engages with Gestalt psychology, and discusses how it was applying these changes.

The first step is to note that Gestalt moves away from the 'man in the inner room'/homuncular analysis of cognition. There is nothing doing inference, no separation of mind from body. How we perceive is a function of our self-organising dynamics. However, they are not going to be ecological in the sense we will eventually develop. Gestaltists attribute order in perceptual experience to the nervous system; things look the way they do is because the nervous system processes are what they are. They claim that if you attribute the order in perception to the world or to the proximal stimulus, you are making the experience error. Turvey in turn calls this move the CNS error; attributing order in perception to the nervous system instead of to information-L. 

The next key element to Gestalt is the claim that the whole is not the sum of it's parts, but that it is different and prior to the parts. The example is the phi-phenomenon, where two alternating flashing lights will cease to look this way when the duration, spacing and frequency parameters meet certain values. Then, the perceptual experience shifts to seeing a single moving dot, even though nothing is moving. Motion (the 'whole') is not therefore made up of parts, it is a thing in itself. 

Even with these moves and more modern ideas, Gestalt is still missing information-L. The proximal stimulus does not have the structure of the world in it, and so even wholes such as motion are neural things imposed on sensations. Even though they tried to understand the environment in behavioural terms, they were also missing affordances, and ended up proposing that organisms could experience the environment in behavioural terms via a neural isomorphism. This effort effectively means they are relying on mental representations, although they would not have been happy with that (actually inevitable) conclusion. 

Summary: Foundational Concepts

This is the final chapter in the Foundational Concepts section. The major thrust of this section has been to lay out the Cartesian program, and to show that every attempt to engage with the problem of perception falls foul of numerous errors, fallacies, and fatal problems (the major one being unrepayable loans of intelligence). Even the Gestaltists, who were working at a time when physics was starting to develop new and potentially valuable ideas (such as fields), were unable to reject the key problem: the assumption that the raw material of perception is meaningless sensation. 

The next four chapters will focus on the computational-representational perspective; the modern version of the Cartesian program. Again, it will fail, for all the expected reasons, but Turvey will spend time here on some of the specifics, because some of the failures can now be couched in computational terms. This is worth noting as the problems are being expressed in the very language that this approach to perception adopted to solve the problems.