He follows the subject from its optimistic beginnings, through its almost disappearance at the hands of behaviourists, to the renaissance of the 1990s tied to new brain scanning techniques.

Decades spent on this subject coupled with acquaintance with the leading researchers has in my opinion produced a remarkable book. He synthesizes earlier research with new findings to produce a fundamentally new picture of how the brain reacts to (or he would say anticipates ) events.

The idea of the brain as a computer is misleading as it suggests an input/program/output model with an ever more complex program to handle new situations. Instead he supports the findings of Donald Hebb in the 1940s showing that nerve fibres make up a network rather than a system of straight lines of stimulus and response, and that there is a high level of feedback.

Brain processing is a competition rather than a calculation and networks evolve their way to stable solutions.

In other words stimuli (including internally generated thoughs), have to compete for attention and are selected or faded depending on the results that they produce. The problem is that all the effort has gone into trying to write more complicated programs for the hardware instead of trying to get it rewire itself and evolve its own processing routines. (top)

This is not to say that the system is completely fluid. The brain reaches a balance between main areas that have a firm structure that will inevitably be needed, and more plastic areas that enable adaption to the unforseen.

Analysis of information (and subsequent neural rewiring) are shown to work on a population voting system, i.e. cells representing different aspects of an experience can fire strongly, weakly, or not at all, creating a picture with shades of certainty and meaning.

Action itself is anticipated. As he says, "......the very fact that we can feel caught out, simply confirms that we must have had a set of expectations in the first place." Confirmed anticipation (for an unfamiliar task) allows its quick automation in the brains circuitry allowing future repetition with much reduced effort. University of Washington PET scans give conclusive evidence and are shown in a series of photos.

Other chapters deal with anticipation leading to a focusing of awareness and the critical role of language as an aid to symbolism and trigger for thought patterns. Try making a mental plan without words!

In my opinion this is a major book, particularly for the signposts that it holds for truly intelligent machines.