French (1956) 's paper on the “formal theory of social power” and its development by Harary (1959) were reexamined and criticized. It was made clear from the comparison of French-Harary model (FH model) with Umino (1981a) 's “decision-making process model (U model) ” that FH model was a special case of U model. Expressing interaction process in terms of U model, which includes several parameters such as the coefficient of unconformity, the coefficient of the perception of mutual distance, and etc., U model enabled us to explain such kinds of interaction as repulsion and over-conformity. It also specified the difference between pattern of relations and the intensity of relations. Several incomplete points and/or defects of FH model were also pointed out. In the cases of French's paper, complete proofs of theorems 1, 2, and 7 were given. The incompleteness in theorem 6 was corrected, with the proof of the revised theorem. According to Harary, FH model was represented in terms of Markov chain. The present paper made clear the differences between the two models; FH model as well as U model was difference equation system of deterministic process. Inconsistency in axiom 3′ was also pointed out.
A panic experiment using rats as experimental subjects demonstrated that the grouped rats in a kind of modified maze box showed a serious panic behavior, on the occasion of escaping from a real imminent fire in the box. None of grouped rats could not escape alive from the fire, although one of them had ever have the same experience and he could have performed very well (Abe, 1978), Abe suggested those different results between the individual and the grouped condition were due to the heavy conformity of grouped rats caused from the panic situation. The aim of the present study was to examine strictly the effects of psychological interference upon the conformity in grouped subjects, without the contaminated effects from the physical interference Abe mentioned. The experimental procedure used in the present study was a kind of escape training situations. Two homogeneous groups of five rats of Wistar strain each were used, and those were not different substantially in their experimental procedures (Table 1., and Fig. 2.). The apparatus used was amodified shuttle box (Fig. 1.), and each rat of the same group was observed together in that box. After some preliminary training sessions, the experimental sessions were carried out. Those consisted of two phases of 30-sec sessions each. In the first phase (choice phase) with tone signal presentation, five rats of two groups could freely move about both compartments of the box, without physical interference. At the end of that phase, the tone stopped and then another phase (outcome phase) started. Each animal was confined forcedly within either of the compartments according to his own choice. The illumination in one compartment was changed into low level (5 lux), but another was still high level (500 lux or 5000 lux) at an even chance, respectively. Each rat could escape at the chance level from the high illumination, regardless of his behavior as “stay-in-the-same-compartment” or “move-to-theadjacent-compartment”. The main findings were as follows. (1) A higher degree of behavioral conformity in grouped rats than the chance expected was observed (Fig. 2., Fig. 3., & Table2.). (2) The behavioral conformity was not affected by the success or the failure of their choice. (3) The behavioral conformity was furthermore enhanced with introducing the unexpected and occasional increase of the highly aversive illumination.