Knowledge Structures (PDF)

This book is a sign of its times. Each one of the chapters - papers written by European authors of various backgrounds- illustrates a departure from the style of theorizing that has been prominent in the behavioral and social sciences for most of the century. Until very recently, models for behavioral phenomena were chi~fly based on numerical representations of the objects of concern, e. g. the subjects and the stimuli under study. This was due in large part to the influence of nineteenth century physics, which played the role of the successful older sister, the one that had to be imitated if one wished to be taken seriously in scientific circles. The mystical belief that there could be science only when the objects of concern were susceptible of measurement in the sense of physics was a credo that could not be violated without risks. Another, more honor­ able justification was that the numerical models were the only ones capable of feasible calculations. (In fact, these models were typically linear. ) An early example of such theorizing in psychology is factor analysis, which attempted to represent the results of mental tests in a real vector space of small dimen­ sionality, each subject being represented by a point in that space. A dimension Wa£ interpreted as a scale measuring some mental ability. The analysis was simple, and only required an electrical desk calculator (with spinning wheels), and a suitable amount of determination.