Abstract

In this paper some applications and methodological developments of mechanical models in psychology in the 1950s are examined. During that period, a new conception of the theory-model relationship in psychology become evident, which had been proposed earlier by the mechanistic trend in psychology in the 1930s. Such a conception allowed psychologists a new approach to many problems in theoretical psychology, such as the role of hypotheses and neurophysiology in psychological explanation and the positions of psychologists concerning neobehavioristic theories of behaviour and the operational method. The ideas of psychologists who accepted mechanical model (and had also been influenced by the advancement of cybernetics) were not homogeneous in that period. Two trend are examined in this paper: the first (in turn not a homogeneous one), which prevailed during the previous decade and was more influenced by cybernetics, was developed during the first half of the 1950s (Deutsch’ machine with insight, Wyckoff model for learning, Broadbent’s model for attention); the second, which originated around the middle of the 1950’s and was dominant in in the next decades, is Newell and Simon’s Information Processing Psychology. Several methodological principles that characterize those two trends are discussed in this paper, along with their similarities and above all their differences.