H.—ANTHROPOLOGY. 145 
of whole cultures with the migration of peoples from one region to another, 
or spread of particular elements of culture through the interaction of 
neighbouring peoples. The present situation of the peoples of the world, 
or the situation at any moment of history, is the result of the total series 
of changes that have taken place over some hundreds of thousands of years. 
The aim of the ethnologist is to make hypothetical reconstructions of some 
of these processes. 
Ethnology, as thus defined, is a historical and not a generalising science. 
It is true that in making their historical reconstructions the ethnologists 
often assume certain generalisations, but as a rule little or no attempt is 
made to base them on any wide inductive study. The generalisations are 
the postulates with which the subject starts, not the conclusions which it 
aims to attain as the result of the investigations undertaken. 
Social anthropology, in the sense I am giving to that term, has con- 
cerned itself with a different type of problem. It has interested itself in 
the development of institutions in human society. From its earliest 
beginnings it attempted a sort of compromise between the two different 
scientific methods, the historical and the generalising. Undoubtedly one 
_of the aims of Social Anthropology has been to understand the nature of 
human institutions and, if I may use the phrase, how they work. But 
instead of adopting outright the methods of the generalising sciences, social 
anthropology was dominated by the conception of history, of historical 
explanation and the historical method. And since historical records were 
insufficient it endeavoured to make a hypothetical history of institutions 
and of the development of human society. It discussed such matters as 
the origin of language and of religion, the development of marriage and 
of property, the origins of totemism and exogamy, or the origin and 
development of sacrifice or of animistic beliefs. 
Social anthropology frequently sought the origins of social institutions 
in purely psychological factors, 7.e. it sought to conjecture the motives in 
individual minds that would lead them to invent or accept particular 
customs and beliefs. Its explanations were frequently, or even usually, 
historical in one sense, but psychological in another, almost never socio- 
logical. This point will be returned to later. 
Throughout almost the whole of the last century this historical-psycho- 
: ogical method so dominated anthropological study that it was hardly 
2 ossible for any one to escape from it. Thus, when Robertson Smith laid 
the foundations of the scientific study of religions and took up the problem 
of the nature of sacrifice, (for that, as we should now see it, was really the 
problem,) he was not content to isolate and classify the different varieties 
_of sacrifice, and show their relation as different forms of a widespread type 
of religious rite—that would be the method of the modern sociologist, as 
esented in the essay of Hubert and Mauss—but the strong tradition 
his time made him attempt to fit the different varieties of sacrifice into 
a scheme of historical development whereby one variety was supposed to 
bs ave had its origin in another. 
_ The compromise that social anthropology made between the historical 
aoa the generalising methods was one impossible to maintain. As a result 
iere have been in the last few decades two movements, one towards 
et ethnology and the other towards sociology, and the traditional social 
193] Ts 
