This is essentially a philosophical treatment of human induction, or synthesis, using chess and backgammon methodology from which conclusions may be drawn. The general method of human synthesis has been known [1]. In synthesis, the term assembly is to be avoided, since it infers normally that known parts or elements are combined. Thus, the author’s term assembly, used in dealing with “. . . the chunking and reassembly of parts into higher level entities . . . ,” is regrettable.
However, the chapter does offer suggestions to the AIer dealing with serial and parallel processing, which can be of value. The reader should be knowledgeable of chess terminology from a mechanistic viewpoint. The entire chapter approaches computer problem solving from the position that if we can find out how a human processes information elements, then it should be possible to emulate this on a machine. The reviewer would have liked to have read a little about human error correction by use of feedback loops since it may well be that the induction/synthesis process involves micro-level feedback paths. Wh- ile the author states that “. . . intelligence arises out of a series of mappings . . . ,” this procedural series may actually consist of minute feedback signal paths and may not be truly linear. The concept of feedback within the series of mappings has not been discussed. The chapter should appeal to strategists in Artificial Intelligence who seek to design systems emulating the human thought process--and that probably includes everyone in this field.