I like to think about myself as a neuroscientist. Quite often, while reading philosophy books, I feel a gap between my daily scientific life and contemplations of the author. However, this time I have by accident grabbed a book that tries to leverage neuroscience data to fit in the hypotheses about the mind. It was Minds without Meanings: An Essay on the Content of Concepts by J.A. Fodor and Z.W. Pylyshyn. Or as I thought. In the end, the gap is still there.

The beginning of the book is devoted to assumptions and definitions, which is very helpful. The authors state right away that they believe that physical behavior is caused by mental events, namely beliefs, desires, etc., and that mental is part of the physical world. The authors also seem to be supporters of representational theory of mind and computation theory of mind. In particular, representation theory of mind (RTM) would suggest that propositions (for example, 'all shops are closed on Sunday') in the mind exist in the form of mental representations, and, in general, cognition is a formation, manipulation, and transformation of mental representations. Meanwhile, cognitive processes are computations over mental representations—a statement supported by the computation theory of mind (CTM). The notion of mental representations is rather obscure to me, but as I understand it, they are physical or not really (for instance, functional) vehicles that the mind manipulates. The main issue addressed in this book is how mental representations acquire meaning (or 'concepts').

Thoughts consist of concepts. One popular stance in philosophy is that concepts can be explained by causal or functional relations with the outside world. For example, ‘dog’ is a concept, and ‘barks’ is a concept, and ‘dog barks’ is a complex concept consisting from two less complex concepts. Both complex and less complex concepts relate to actual things or processes in the external world. However, and it is the critics by the authors, some concepts are not images, not definitions, and not stereotypes. There could be abstract concepts that are impossible to image; for instance, you can image chair, but not the property of being a chair. There could be empty concepts that do not apply to things, such as the concept ‘and’. There could be strereotypical complex concepts, such as ‘pet fish’, that are not a mere sum of two less complex stereotypical concepts, here stereotypical ‘pet’ and stereotypical ‘fish’. Thus, not all concepts relate to things in the external world.

The set of things, that concept is applied to, is its extension. Extension is determined by intension. Intension is the content of the concept. For example, the extension of 'cat' includes all actual cats that exist. The intension of 'cat' might include features like 'a small, furry, carnivorous animal that meows'. If two concepts are not equivalent from the point of view of their extensions, then they should be as well distinct from the point of view of their intensions. The authors show that this doesn’t hold. There could be objects with the same extension that have different intension for different people. For example, two terms 'Clark Kent' and 'Superman' for someone who doesn’t know that Clark Kent is Superman would seem distinct. Moreover, there could be extensions that cannot define the meaning because of different intensions. For some, Jupiter may be a celestial body of gas, while for others it is a Roman god. Thus, there is no one-to-one mapping from extension to intension.

The authors conclude that meaning is a myth. Meaning is hardly about causal relation to the external world, but rather, according to authors, is determined by a structured syntax of symbols that constitute mental representations. Furthermore, because this syntactic structure is specific to individuals, the only naturalistic psychology that can be is individual psychology.

This book for me was like a running train: I jumped in in the middle of the journey, when everyone on the train already met each other and had many conversations. Although I was able to understand particular examples and definitions, I was completely oblivious to where the discussion is coming from. I had to spend some time reading other texts to understand the context that we are in. On the other hand, there are many examples of experimental evidence from neuroscience and cognitive science that I was able to relate to. However, the experimental set-ups in the past as well as nowadays allow many possible interpretations of results, so I wouldn't say I was convinced. Otherwise, if you want to know more about mental representations and the idea of syntactic structure, I suggest starting with the earlier books of Jerry Fodor. And this is what I myself will do.

Interesting idea: “percepts are initially represented as ‘indexed objects’ or as ‘this’ and ‘that’ ”

November, 2024