Mind as Machine: A History of Cognitive Science
By Margaret Boden
(Apologies if the link does not work: university administrators mostly fail to understand the importance of preserving links with at least as much care as we give to preserving contents of our libraries.)
1,631 pages in two volumes, one purple one pink,
published by Oxford University Press 29th June 2006
A free chapter is available online at the UK OUP web site, including the 20 page analytical table of contents.
The author is an old friend and collaborator of mine, and I was expecting this to be an outstanding book, as her previous ones have been. (Listed on her web page at Sussex University.)NOTE added 22 Oct 2012:But this is truly a magnum opus, a stunning achievement.
Margaret Boden has amazingly broad and deep knowledge of many aspects of biology, neuroscience, psychology, linguistics, social sciences, philosophy and AI, and in this work she takes readers through many of the detailed strands of development in the last century (some starting further back).
I have not yet read all of it, but so far I have learnt that I am can start almost anywhere in the book, and find myself fascinated by things I did not know (in many cases I wish I had known them earlier) seamlessly interwoven with things I did know, in a story that brings out not merely historical sequences and causes, but also logical interconnections, often unobvious interconnections.
Moreover, she is a wise and penetrating critic as well as historian and synthesiser so she does far more than accept at face value what various authors have written.
In many ways she is ahead of most practitioners in the field, e.g. not assuming that any current popular notion of of computation (e.g. whatever can be modelled on a universal Turing machine) is necessarily the last word, and taking seriously (near the end of Volume 2) doubts raised by a small subset of researchers (e.g. Brian Cantwell Smith and others) regarding our understanding of what computation is and the variety of forms it may take, both in naturally occurring systems and in future man-made machines.
At the end she has even dared to make some predictions about what will happen next.
My deepest regret at first was that because this is a massive two volume book it would always be expensive, so that that the vast majority of people who could benefit from it would never get their hands on it.
If only she, and OUP, could be persuaded to make a much cheaper electronic version of it available -- it would probably be read and used regularly by at least a thousand times as many people. They would also have the benefit of electronic searching, which is often very much better than using an index, even though this book has two extensive indexes of Names and of Subjects. But only at the end of volume 2! So if you expect to be using the index while reading the first volume, you'll need to have the other volume close at hand.
The situation has been partially improved by the availability of a paperback edition of the book, in 2008, mentioned above, and the fact that with careful choice of search phrases it is possible to get google books to display relevant portions of the book.
As I read more of it I may add further comments.
Maintained by
Aaron Sloman
School of Computer Science
The University of Birmingham