2 THE AMERICAN JOURNAL OF SOCIOLOGY
other, which in the case of the most diverse purposes may be the same, and in the case of like purposes may be most various these have not as yet been the subject-matter of a particular sci- ence; and yet such a science, when constituted, would for the first time make manifest what it is which makes the society that is, the totality of historical life into society.
I deny myself at this point all further explanation and justi- fication of this program, since it is, after all, less important to propose a program than to show by carrying it out its signifi- cance and its fruitfulness ; and I proceed at once to the special problem, namely, how the form and the inner life of a societary group are determined by the numerical relationships of the same.
It will be conceded at the first glance, without hesitation, that the sociological structure of a group is essentially modified by the number of the individuals that are united in it. It is an everyday experience yes, it is almost to be construed from the most general social-psychological presuppositions that a group of a certain extent and beyond a certain stage in its increase of numbers must develop for its maintenance certain forms and organization which it did not previously need ; and that, on the other hand, more restricted groups manifest qualities and recip- rocal activities which, in the case of their numerical extension, inevitably disappear. A double significance attaches itself to the quantitative determination : first, the negative significance that certain forms which are necessary or possible from the con- tents or the conditions of life can come to realization only before or after a certain numerical extension of the elements ; the posi- tive significance that other forms are promoted directly through definite and purely quantitative modifications of the group. As a matter of course, these do not emerge in every case, but they depend upon other social circumstances in the group. The decisive matter, however, is that the forms in question never spring from these latter conditions alone, but are produced from them only through the accompanying numerical factor. Thus it may be demonstrated that quite or nearly communistic for- mations have up to the present day been possible only in rela- tively small circles, while they have always failed in large groups.
