Though like other pioneers he erred and erred grievously
in many of his interpretations, his insistence on the theo-
retical significance of associations must rank as one of the
most important points of departure in the study of primi-
tive sociology.

In trying to give some notion of the types of association
it would be easy to group the facts in a series of separate
compartments. But these categories would hinder rather
than promote a synthetic understanding of social organiza-
tion. For example, we might deal with sex dichotomy
under one head and with age-classes under another; yet
such treatment would lead to all sorts of artificiality. It is
true that in Australia the sexes are often rigidly divided in
ceremonies, but feminine disabilities are shared by the
younger males, so that in one sense the real division at a
particular moment is into initiated males and the remainder
of the tribe. Elsewhere there is a tripartite division into
the married couples, the bachelors, and the spinsters, so
that sex dichotomy applies only to the unmarried. Again,
there is little doubt that the bachelors' dormitory and the
men's clubhouse are often genetically connected; the men
after marriage continue to resort for pastime or work to
the dwelling they occupied before wedlock. Yet a logical
classification might easily lead to a separation of these re-
lated institutions. In the present chapter, then, I will
rather select a number of tribes from different geographical
areas and will describe in each case their social organiza-
tion apart from sibs and families. In this selection I shall
be guided partly by the quality of available literature,
partly by the desirability of presenting at least all the main
varieties of associational units. In the following chapter
the treatment will be topical, embracing a number of points
of theoretical interest. ¹

Since the Andaman Islanders are sibless, each of their

-258-