There Are Two Cognitive Science Approaches: Only One of Them Leads to AI Minds
Two Kinds of Cognitive Science
As I’ve written about recently, in the mid 1950’s there arose a ‘cognitive revolution’ in psychology (see here and here). This led to the birth of several distinct disciplines including neuroscience, AI, computer science, theoretical linguistics and cognitive science or cognitive psychology.
There are a number of people considered founders of cognitive science. Common names upon a quick google — my google may be biased, but I did get my husband to google the same question and he got the same names — include George Miller, Alan Newell, Herbert Simon, Noam Chomsky, Marvin Minsky, and Urlic Neisser.
I studied Chomsky’s work during my dissertation, so the first time I saw this list of founders of cognitive science, I was surprised to see his name there.
The reason for my surprise was not because he didn’t deserve to be named a founder (he did, after all, found the formal study of language).
The reason was because I know that the way Chomsky does science is very different to how many cognitive scientists do science.
That’s what this article is about.
I think there are at least two very broad methodological ways of doing ‘science of the mind/brain’ research, that came out of the cognitive revolution.
The problem is that one approach (the approach that treats the mind like a computer) is so dominant that the other approach gets systematically dismissed, ignored, or misinterpreted as being part of the dominant approach.
In this article, I’ll lay out the two approaches. I’ll call these approaches the ‘cognitive approach’ and the ‘methodological naturalist’ approach, though I want to be clear that not all work that is called ‘cognitive science’ today falls into the approach I am calling ‘the cognitive approach’. Just a lot of it.
Cognitive science is a fascinating and complex field.
My intention is not to dismiss that complexity.
I also want to make clear that I think much of the results and experiments in cognitive science are interesting and useful. It is the theoretical or philosophical interpretations of the cognitive approach I will discuss that I disagree with.
