Perhaps we will all agree that the main trend of present day social psychology is in the direction of elucidating everyday social events and concrete human social behavior. American psychologists who always maintain their pragmatic standpoint, have succeeded in obtaining objective data from their inquiries into all kinds of social phenomena, using their highly developed techniques of survey and measurement. However, most of them are so engrossed in specific phenomena as to neglect, consciously or unconsciously, other no less important tasks of social psychologits, although we admit that they have every good reason to do so.In course of our concentration upon individual specific phenomena we generally are apt to overlook its relation with other phenomena and think less of, if not completely neglect, the effort of systematising them. Public opinion survey as conducted at present may be cited, among others, a pertinent example of the case.For all their highly developed techniques of investigation of public opinion and their extensive application in recent years we can hardly get a satisfactory answer to the questions of where public opinion is based, how it originates, and how it changes. Most students who are interested in it are contented only with noting and describing the distribution of opinion. The utmost they have done is to report how such and such particular change in external conditions has effected such and such particular change in the distribution. Thus we can barely hope for any further progress of the theory of public opinion, so long as the external conditions and a particular opinion are dealt with in their raw state, namely, in terms of common sense. There is no wonder that the public opinion survey has hardly anything to say in the province of theory, and we are obliged to turn to the more or less abstract speculations of the former days for an answer to our afore-said questions. The only way which seems to be open for us to get rid of their difficulty in order to attain a truly scientific elucidation of the phenomena of public opinion is likely to be in substituting dynamic concepts for common-sencial ones, in such a manner as to make us find the dynamic structure underlying the phenomena.Arguing in this way, we are naturally reminded of the experimental psychological researches made by Lewin and his followers which have been directed to the elucidation of group phenomena and especially of the dynamic structure of the group itself.It is true that their researches have yet been confined to their laboratories and so those results, of theirs can not be at once put on the same level with concrete everyday social phenomena, but the theory of “social field” which has been advanced by them must occupy our attention in its principle. It is by the suggestion of this theory that we can reduce our everyday life to the communication between man and man, and man and his environment. Such a reduction will enable us to treat every social phenomenon in terms of dynamic concepts. Applying this to the phenomena of pubic opinion, the so-called external conditions may be reduced to the “potential” acting upon individuals, and the communication between individuals may be described as a “force” acting between them, enabling us in this way to understand and measure the whole phenomenon as a dynamic process conducted in the dynamic field of the group.The level of social psychology being as it is, we must admit that there are many difficulties in substitutiing at once quantitative dynamic concepts for various dimensionally different factors. Some intermediary concept is necessary to be introduced, and we think we can find a suitable one for the purpose in the concept of “attitude.” Attitude means to us an action-tendency of an individual which is formed in social field in the past, an potential which may be discharged as an acting force by communicating with other