September 29, 1911] 



SCIENCE 



395 



recorded forty years ago by Morgan and 

 Hyde, and I obtained evidence that the 

 system is still deeply interwoven with the 

 intimate mental life of the people. 



If, then, social structure has this funda- ■ 

 mental and deeply seated character, if it is 

 the least easily changed and only changed 

 as the result either of actual blending of 

 peoples or of the most profound political 

 changes, the obvious inference is that it is 

 with social structure that we must begin 

 the attempt to analyze culture and to as- 

 certain how far community of culture is 

 due to the blending of peoples, how far 

 to transmission through mere contact or 

 transient settlement. 



The considerations I have brought for- 

 ward have, however, in my opinion, an im- 

 portance still more fundamental. If social 

 institutions have this relatively great de- 

 gree of permanence, if they are so deeply 

 seated and so closely interwoven with the 

 deepest instincts and sentiments of a people 

 that they can only gradually suffer change, 

 will not the study of this change give us 

 our surest criterion of what is early and 

 what is late in any given culture, and 

 thereby furnish a guide for the analysis of 

 culture? Such criteria of early and late 

 are necessary if we are to arrange the cul- 

 tural elements reached by our analysis in 

 order of time, and it is very doubtful 

 whether mere geographical distribution it- 

 self will ever furnish a sufficient basis for 

 this purpose. I may remind you here that 

 before the importance of the complexity of 

 Melanesian culture had forced itself on my 

 mind, I had already succeeded in tracing 

 out a course for the development of the 

 structure of Melanesian society, and after 

 the complexity of the culture had been es- 

 tablished, I did not find it necessary to 

 alter anything of essential importance in 

 this scheme. I suggest, therefore, that 

 while the ethnological analysis of cultures 



must furnish a necessary preliminary to 

 any general evolutionary speculations, 

 there is one element of culture which has 

 so relatively high a degree of permanence 

 that its course of development may furnish 

 a guide to the order in time of the different 

 elements into which it is possible to analyze 

 a given complex. 



If the development of social structure is 

 thus to be taken as a guide to assist the 

 process of analysis,' it is evident that there 

 will be involved a logical process of consid- 

 erable complexity in which there will be 

 the danger of arguing in a circle. If, how- 

 ever, the analysis of culture is to be the 

 primary task of the anthropologist, it is 

 evident that the logical methods of the sci- 

 ence will attain a complexity far exceeding 

 those hitherto in vogue. I believe that the 

 only logical process which will in general 

 be found possible will be the formulation 

 of hypothetical working schemes into which 

 the facts can be fitted, and that the test of 

 such schemes will be their capacity to fit in 

 with themselves, or, as we generally express 

 it, ' ' explain ' ' new facts as they come to our 

 knowledge. This is the method of other 

 sciences which deal with conditions as com- 

 plex as those of human society. In many 

 other sciences these new facts are discov- 

 ered by experiment. In our science they 

 must be found by exploration, not only of 

 the cultures still existent in living form, 

 but also of the buried cultures of past ages. 



And here is the hopeful aspect of our 

 subject. I believe our present store of 

 facts, at any rate on the less material sides 

 of culture, to form but a very small part 

 of that which is yet to be obtained, and will 

 be obtained unless we very wilfully neglect 

 our opportunities. Waiting to be collected 

 there is a vast body of knowledge by means 

 of which to test the truth of schemes of the 

 history of mankind, not only of his migra- 

 tions and settlements, but of the institu- 



