Computation and Cognition: Toward a Foundation for Cognitive Science by Zenon Pylshyn
My rating: 4 of 5 stars
Zenon Pylshyn presents a strong foundation for how cognition should be literally thought in terms of: cognition. He provides a functional approach to understanding how humans can be generic thought machines. We seem to have the ability to parrot the functioning of processes themself. His later chapters in which he provides evidence in cognitive studies as to how people seem to follow a process oriented algorithm is the strongest push for this.
Having said that, looking back, it’s apparent that most of his book was simply describing the inner workings of computation and arguing that cognition needed to work on a symbolic level anyway; that the physical materialism was only the vehicle for the functionality that the functionality is what we want. In some ways I am reminded of Alfred North Whitehead and Bertrand Russell’s attempt at providing a mechanical notion of rationality. Zenon doesn’t go as far, but he wants to provide us a coherent and discrete value system upon which to weigh cognition. This isn’t a bad idea, but it appears that people don’t often make decisions on purely discrete values. We are capable of doing so, but whether or not we do so, especially outside of scientific tests, which always tests for discrete values if possible, is another matter entirely.
This is my second time reading the book. The first time was about ten years ago, and I am not sure I understood the latter parts of it. In a way, Zenon should have probably reversed the order of some of his chapters. He should have provided the cases arguing for cognition as computation with an emphasis on how cognition is process oriented based on functional equivalence, and then provided the chapters on functional architecture’s inner workings. I understand however that he was most likely providing the functional architecture first as a theoretical basis, so that we can understand what we see later on in terms of the theory…but since his book was geared towards arguing for functional architecture as the mode of consideration, it may make more sense to work towards the argument first, rather than having some cognitive proof later on as a BTW, this stuff can be explained by cognition.
In a way, what’s missing about this work is the null state of cognition. That is to say, once we get the system going with equivalences, that’s fine and good. But what mark do we make in order to start it going, where we come to identify as a self? This is perhaps troubling theory but Pylshyn does not get into this. He also seems to think that computation is the literal process of cognition. I would rather consider, in the spirit of functional architecture, that we can’t ever know what the literal process is, we can only denote equivalency in functionality. I guess that makes me fallibilist. It’s suprising that Pylshyn isn’t a fallibist, since he takes some of the question as to how memory works exactly or how processing works exactly as not being important since equivalency trumps actuality when it comes to trying to make sense of how this could work. It doesn’t matter how it works specifically, what matters is that it does in these kinds of algorithmic steps. Perhaps in this sense, this book isn’t that radical after all.
