視野闘争の研究 : 行動の力学への一探索

Abstract

The behavior was regarded here as a whole system of dynamical interactions or interrelations of various systems. If an entity could be defined through a set of converging operations each specifying a certain stimulus-response relationship, it was called a system. As an example of these systems, the binocular vision system was characterized as a subsystem of the whole, and at the same time as a system of interaction of two monocular systems. In the present study, the phenomenon of binocular rivalry was considered as a limiting situation of such dynamical interactions, and, first of all, an approach to the determination of dynamical law of rivalry was intended. A number of published research findings were critically reviewed and some of them were reexamined experimentally. The important points were predominancy or prevalency on the one hand and rate of alternation on the other. It was decided that the former should be the most relevant variable for the purpose of the present study. As the most adequate measure of predominancy of one figure over the other in rivalry, total time of appearance of one figure in a specified period of observation was the main concern of our experiments, and when necessary, appropriate indices of predominancy were derived. Various aspects of dynamics of interaction in rivalry were revealed through experiments where determination of functional relations between predominancy and stimulus intensity were intended. Predominancy was found to be basically an S-shaped increasing function of log stimulus intensity. This, along with several other findings, was interpreted as showing the unique property of the binocular interaction system. Next, it was intended to relate the binocular interaction system to another, probably higher-orded, system and to find some higher-ordered interaction between these systems. Specifically, in one series of experiments, effects of the set which were determined by preceding observation of a figure upon succeeding rivalry between that figure and another one were examined, and in the other, binocular rivalry was controled voluntarily to counteract the stimulus intensity function described above. Although the results were neither complete nor conclusive, they suggested some possibility and fruitfulness of such methodology. Finally, adequacy of indices and measures used through these experiments was reconsidered and the question was raised on the methodology of determining the properties of the sensory systems through various, especially verbal, indicator responses. It was insisted that psychophysical data for sensory systems should be considered as manifestation of interaction of sensory response and some kind of verbal reference system. Some preliminary attempts to analyze such interaction experimentally were reported.