Book review
On Becoming a Personal Scientist By M. L. G. Shaw. 1980. Academic Press, pp. 322. Dr Shaw's book provides a valuable, often unique, contribution to the literature on man-machine processes. As such, the volume is rightly subtitled "Interactive Computer Elicitation of Personal Models of the World". Eleven types of computer program are described in sufficient detail for ready implementation in Chapters 4-8 (pp. 32-105) and specific applications, together with original and very interesting data, are dicussed in the next chapter (pp. 108-147). Apart from a brief review of psychometric issues (pp. 156-167) the lengthy appendices (pp. 167-307) consist in exemplary runs of the "Elicitation" programs with annotated printout. Tlae ingenuity and utility of these programs as well as methods upon which they are founded is unquestionable. But, although much of the space is occupied by technical matters, the book is not primarily a technical manual. It is also a profound theoretical essay, succinct because it is clearly illustrated by the programs, applications, and data.
This essay is mostly contained in the first few chapters (pp. 1-32) and between p. 105 and p. 147. The following comments are addressed to it almost exclusively without in any way derating the technical material and the empirical results, the value of which is henceforward taken for granted.
Dr Shaw is one of the few people who grasp the meaning and the importance of Personal Construct Theory, who does not try as Adams Webber, for example, does so valiantly try (Personal Construct Theory, Wiley, 1979) to fit it in a mould acceptable to mainstream opinion. She has taken pains to express it with great clarity and to set it in the context of other theories; for example, those of Jahoda and Thomas, of Hanson and my own conversation theory. Shaw also provides an historical context for Kelly's personal construct theory: not only (as others have done, notably Bannister, Fransella and Mair) of contemporary psychology, but also of scientific and philosophical thinking, for example, (McCulloch, Jeans, Schroedinger, Wittgenestein, Bartlett and "Effort after Meaning").
This degree of perspicuity is not, perhaps, so surprising since Shaw worked for many years at the Centre for Human Learning, especially with Laurie Thomas, one of the most brilliant innovators in the area of Personal Construct theory. Her indebtedness and mutualism, both at the theoretical and technical level, is fully acknowedged in this book. The book is, however, the first generally available publication which spells out a genuinely kinetic as well as rigorous exposition of Kelly's work. There are other accounts of like kind, for instance, Bussis, Chittenden & Amarel, 1976, Beyond the Surface Curriculum, but this is specific to education (the outcome of a Ford Foundation project) and lacks technical sophistication; some (in press) work by Boxer and the publications of the Centre for Human Learning (mostly available in Report Form). Further, Shaw goes beyond an exposition and integration of personal construct theory and furnishes original theses and methods that exhibit them, impressively.
The "Repertory Grid" is well known as a metrical device. For a period after the first presentation of Kelly's theory (Kelly, 1955) the "Repertory Grid" became degraded into an expensive substitute for Osgood's "Semantic Differential" in quantifying otherwise qualitative studies or an unwieldy adjunct to psychotherapy: unwieldy, because in this application, it was recognized that the adjectival predicates (polar constructs) really are individual and that the values assigned to them, once elicited, are individual values. Until computers and, in practice, microprocessors, became widely available a genuine use of the "Repertory Grid" or any comparable device was necessarily cumbersome and well intentioned use of factor-analytic techniques served only to further obfuscate feedback of the element-construct-implication model which is, as Shaw points out, the main reason for building such a personal model. This book describes a variety of conversational systems based upon computer elicitation programs which, contrary to the degenerate practices of a few years ago, render "elements" as processes, 167