H.—ANTHROPOLOGY. 149 
“explain ’ the phenomena to which they refer. A science of this kind, as 
I conceive it, still remains descriptive, but in place of descriptions of 
particulars and their particular relations, such as the historical sciences 
give us, it provides general descriptions. 
The older social anthropology did not follow this method, at any rate 
consistently. We have seen that it devoted most of its attention to 
formulating hypotheses about the origins of social institutions. Never- 
theless social anthropology, by its comparative study of institutions, made 
possible the development of comparative sociology. I could, if I had time, 
show you how the new anthropology, i.e. comparative sociology, grew 
gradually out of the older study; how the first tentative movements 
towards this science began in the eighteenth century ; how the work of 
such men as Steinmetz, Westermarck and others, and particularly that 
of Emile Durkheim and his followers, led step by step to the present 
position in which we can claim that there is now in existence a comparative 
sociology which demands recognition as something radically different in 
important respects from the social anthropology out of which it has 
Town. 
: The essential difference between the older social anthropology and the 
new lies in the kind of theories that one and the other seek to establish 
by the study of the facts. As I see it, comparative sociology rejects, and 
must reject, all attempts at conjecturing the origin of an institution when 
we have no information based on reliable historical records about that 
origin. 
Sr only hope to make my meaning in this matter clear to you if 
you will permit me to refer to a particular example. We may take as 
our example totemism, which has received a good deal of attention in 
social anthropology. Totemism is a name which we apply to a large 
_ number of different kinds of institutions in different cultures, all having 
in common the one feature that they involve some special relation between 
social groups and natural species, usually species of animals or plants. 
It is to be noted, first of all, that totemism is not a simple concrete thing ; 
it is an abstraction, a name applied to a number of distinct and diverse 
things which have something in common. What is or is not included 
under the term depends on the definition we adopt, and different writers 
choose different definitions. 
The older social anthropology concerned itself with the question of the 
origin of totemism. Even supposing that we have settled what we are 
and what we are not to include under the term, our question is still not 
specific. If we try to make it specific we must recognise that there are 
three possibilities. One is that all the things we call totemism in Asia, 
Africa, America and Oceania are historically derived from some one 
rticular institution which had its origin in a particular region at a 
particular time. A second is that some one particular form of totemism 
may have arisen independently in two or more regions at different times 
_ as the result of similar historical processes, and that existing varieties of 
temism are all derived from this. The third is that different forms of 
_totemism may have had their origin independently in different regions at 
- different times by different historical processes. If I had to decide which 
of these three possibilities seemed to me the most likely, I should select 
é 
