Sunday, 6 January 2013

Tim Tyler: Calvin: The Cerebral Code (review)

Transcript:

Hi. I'm Tim Tyler, and this is a review of this book:

The Cerebral Code: Thinking a Thought in the Mosaics of the Mind by William H. Calvin

I read this book because William Calvin coined the term "Darwin machines" - named after Turing machines - in 1987. He was a pioneer in studying Darwinian evolution within the brain. The evolution of mental structures is now an important field, which relates to understanding the mind and emulating its properties using machines. Calvin also attempted to formulate basic principles of Darwinian evolution - in an area now widely known as Universal Darwinism.

The book continues work which was pioneered by Gerald Edelman in Neural Darwinism which was based on work dating back to the late 1970s.

The book introduces the idea of mental evolution through the idea of cultural evolution. Calvin says that:

Dawkins’s real contribution has turned out to be on the copying side, not the selection side, of mental Darwinism. In his 1976 book, The Selfish Gene, he extended the notion of copying genes to copying memes (cultural entities such as words and tunes). It took awhile before anyone realized its implications for copying inside a single brain.

This is a pretty accurate summary of what happened back in that era. The blurb for this book describes Calvin as a theoretical neurophysiologist. I don't consider myself an expert in neurophysiology - but I think I know enough about brains to say that most of this this book is misguided and outdated. The main problem is with Calvin's geometrical obsession. Calvin is the Buckminster Fuller of the brain - but not in a good way. He sees the brain's layers as composed of large numbers of hexagonal units, less than a millimeter across. When he considers the evolution of brain structures, these hexagonal units are what he considers to be evolving. He broadly equates them to concepts and has them evolving on the surface of the brain - rather like a two dimensional cellular automaton.

The book presents this vision, but does a miserable job of presenting evidence for it. Calvin cites lateral inhibition, and recurrent excitation between equidistant neurons to argue that neurons will tend to form triangular assemblies of interlinked neurons which fire together. He argues that such linked neurons will fire synchronously - like fireflies - rather than fire in cyclic chains.

It is true that loops of neural activity in the brain are probably important to it's function. However, Calvin doesn't produce much evidence for his vision of hexagonal cerebral mozaics, and I think it is fair to say now that not much evidence of them has been uncovered since this book was published. This wouldn't matter so much, except for the fact that the hexagonal patches seems to be the main theme of the book - and it is completely saturated with them.

It is pretty easy to see what is copied at a low level in the brain. Neural impulses are copies as they travel down branching axons, and axon and dendrite tips are copied as they grow and divide. Apparently there's also some gross electrical copying as electrical waves propagate through the brain and neurons directly affect their neighbours. These basic copying processes in turn go on to support higher-level copying processes - as memories are recalled, as habitual behaviours are performed and as ideas interact.

However, these days, Calvin's search for a geometry of the brain seems rather misguided. The problem seems to be the lack of a proper scientific methodology. Calvin's brand of armchair philosophy is reminiscent of a mixture of numerology and phrenology. It pays insufficient heed to scientific evidence. I think Calvin's axiomitization of Darwinism suffers from a similar problem. He didn't convince me that we needed six axioms of Darwinism - or the eleven that he later went on to present. I don't think axiomitizing things is one of Calvin's strong points.

In summary, I thought this book was pretty disappointing. It's hard to recommend it to anyone interested in the evolution of brain structures or mental entities - it just has too much speculative nonsense in it. On the positive side, the contents of this book are available free of charge on Calvin's web site - which makes it good value for money.

Enjoy,

No comments:

Post a Comment