162 SECTIONAL ADDRESSES. 
regarded as offering us in its culture system a special variety of a general 
type. By studying these variations as minutely as possible we can carry 
out a process of generalisation which enables us to give a general definition 
or description of the type itself. By this process we are often able to 
discover correlations between one element of culture and another. Further, 
this procedure is almost essential in any attempt to discover the meaning 
and the function of any element. For by it we are able to determine, in 
any institution or custom or belief, what remains constant and what varies 
as between one part of a culture area and another. 
This study of culture types and varieties in comparative sociology is 
quite different from the study of culture-areas in ethnology. The latter 
aims, above all, at providing material for the hypothetical reconstruction 
of movements of culture diffusion. The former is essentially a process of 
generalisation, a means of discovering general features or principles which 
remain constant throughout the type while taking different forms in 
different parts of the area. 
In this study of variations of a single culture type we should aim at 
comparing the whole culture of one tribe with that of another. But that 
is often impossible ; in fact, in the present state of our knowledge, almost 
always. We may proceed, therefore, by making a comparative study of 
variations in some particular aspect of the culture. But we must be 
careful how we isolate one part of the culture from another for the purposes 
of study. Thus, a good deal of misunderstanding has resulted from 
dealing with some particular aspect of the social organisation of Australian 
tribes, instead of dealing with that organisation as a whole. 
There is perhaps no other region which is quite the same as Australia 
in the opportunities it offers for the study of many variations of a single 
culture type. In other regions, therefore, our procedure must be somewhat 
different. Thus, if we wish to deal with the Bantu cultures of Africa we 
must begin by dividing the whole region into suitable units. One such 
unit would be composed of the Basuto-Bechuana tribes, while the Zulu- 
Kaffir tribes would provide us with another. Our first step will consist of 
a careful study of the variations within the unit region. We then compare 
the one region with the other, and may proceed in this way to explore the 
whole Bantu area in such a way as to be able to give a sound description 
of the general characters of Bantu culture as a whole. Only when we 
have carried studies of this kind a certain distance does it become really 
profitable to make comparisons between Bantu culture and Polynesian or 
North American. 
Thus, for the new anthropology the comparative method is a method 
of obtaining generalisations. Amongst the variations of institution and 
custom in one region we seek to discover what is general to the whole 
region or type. By comparing a sufficient number of diverse types we 
discover uniformities that are still more general, and thus may reach to 
the discovery of principles or laws that are universal in human society. 
A word, the constant use of which has been a great obstacle to scientific 
thinking in anthropology, is the word ‘ primitive.’ It conveys the sugges- 
tion that any society to which we apply it represents for us something of 
the very beginnings of social life. Yet if culture had, as we may well 
assume, a single origin some hundreds of thousands of years ago, then any 
a 
Rieter» mire ria Aid ee 
