Thursday, December 1, 2011

Connectionism and the Mind


Reprinted in 2007, original 2nd Edition in 2002, by William Bechtel and Adele Abrahamsen, Blackwell Publishing. ISBN 0-631-20713-9.

A friend recommended three books for me to read if I wanted to pursue my research on Parallel Computing. This was one of them. If you can find a copy (purchase or library), the story on page 12 about the two sisters of AI is really worth repeating. I don't have permission to plagerize but the title page does authorize short passages for review. So, here 'tis (actually taken from Seymour Papert's book on Perceptrons, page 3):
Once upon a time, two daughter sciences were born to the new science of cybernetics. One sister was natural, with features inherited from the study of the brain, from the way nature does things. The other was artificial, related from the beginning to the use of computers. Each of the sister sciences tried to build models of intelligence, but from very different materials. The natural science sister built models (called neural networks) out of mathematically purified neurones. The artificial sister built her models out of computer programs.
In their first bloom of youth the two sisters were quite equally successful and equally pursued by suitors from other fields of knowledge. They got on very well together. Their relationship changed in the early sixties when a new monarch appeared, one with the largest coffers ever seen in the kingdom of sciences. Lord DARPA, the Defense Department's Advanced Research Projects Agency. The artificial sister grew jealous and was determined to keep for herself the access to Lord DARPA'a research funds. The natural sister would have to be slain.
The bloody work was attempted by two staunch followers of the artificial sister, Marvin Minsky and Seymour Papert, cast in the rold of the huntsman sent to slay Snow White and bring back her heart as proof of the deed. Their weapon was not the dagger but the mighter pen, from which came a book - Perceptrons... (1988, p.3)
And that was the extent of humour in the book. Now, about the rest of the book; it is extremely well laid out and covers Parallel Procssing, Dynamics and Evolutions in Networks (as it says on the cover page) from a new and updated outlook and perspective. This is a thoughtful and well-thought-out approach to Connectionism, Modeling and Neural Networks as well as discussing Rules and other related topics. I especially appreciated the complete overview of the entire field of computers versus psychology versus the mind itself. Meaning that if you are doing rulebased systems, psychology, neural networks, AI or anything loosely associated with those fields, it might be helpful to what you are doing. According to others, it is a great improvement on the original book published in 1991. I think that the reprint corrected some earlier misprints in the 2002 edition.

The chapters are:
  1. Networks versus Symbol /systems: Two Approaches to Modeling Cognition
  2. Connectionist Architectures
  3. Learning
  4. Patterns and Recognition
  5. Are Rules Required to Process Representation?
  6. Are Syntactically Structured Representations Needed?
  7. Simulating Higher Cognition: A Modular Architecture For Processing Scripts
  8. Connectionism and the Dynamical Approach to Cognition
  9. Networks, Robots and Artificial Life
  10. Connectionism and the Brain

A short read (350 pages total text plus other stuff) and, while mostly centered on neural networks, it covers other fields as well. Unlike most books on neural networks, the math is simple enough for those who have forgotten most of what they learned in school. Summation is about as hard as it gets. The Rulebased (BRMS) guys will probably turn first to Chapter 5 (as did I) and then discover that they need to read the previous four chapters for it to fit what the author is saying in that chapter. The final chapter is the capstone of the book (as it should be) and brings everything together nicely. Read through the table of contents first and get a good idea of the lay of the land and what you're going to discover. Consider it a map of a new and wonderous adventure.

As with the other books that I have been privileged to read recently, Buy The Book! Even it you have the 1991 or the earlier 2002 editions, this one is an improvement. Also, it makes for a nice addition on your book shelf. If nothing else, it might impress your girl/boy friend or wife/husband that you are an in-depth "thinker" who is entitled to be a little strange sometimes. (Your cherished off-spring probably will never see it and, if they do, they will think you are really weird - which, sometimes, is nice.) And, maybe, you might read it and discover a whole new world. For those who HAVE already read it (or will read it in the future) I would appreciate your comments. And, please: Don't say things like, "Great book." or "Really liked it" only in your comment. Try to be more specific about what you thought of the book. OK, maybe "Really liked it" would be permissible, but, honestly, try to pretend that you actually struggled through the whole thing. :-)