528 ANNUAL EEPORT SMITHSONIAN INSTITUTION, 1931 



disappearing. Civilized man is far older than the merely literary 

 scholar has dreamed. 



But the archeologist also must be careful not to exceed his evi- 

 dence, and, above all, not to make assumptions which belong to a 

 prescientific age. When I was young, the assumption that language 

 and race were interchangeable terms was still widely prevalent, and 

 one of my first incursions into linguistic science was an article pro- 

 testing against it in the Journal of the Anthropological Institute. 

 To-day, every scientist would acknowledge its falsity, but neverthe- 

 less the assumption sometimes appears even in quarters where it 

 might least be expected. It can not be too often made clear that all 

 linguistic science can do is to indicate geographical contact; where 

 we have allied languages we have evidence of social intercourse, but 

 nothing more. 



The old fallacy, however, which confused language and race to- 

 gether, has been succeeded by another fallacy, which is unfortunately 

 not infrequent in modern anthropological books. Similarities in 

 technique are assumed without question to indicate relationship or 

 contact in race and history. It is especially when dealing with 

 pottery that the archeologist is tempted to assume without further 

 evidence that such contact or relationship exists. But it is clear that 

 mere similarity in form proves nothing of the sort when standing 

 alone. The number of possible forms, for instance, belonging to 

 vessels intended for use is necessarily limited; man is an inventive 

 animal, and the same form could have been devised independently 

 in different parts of the world. Coloration and ornament are more 

 evidential, but even here there is plenty of room for the existence 

 of accidental similarities. Moreover, we have to allow for primitive 

 barter, which implies, not actual 'trade between two widely sepa- 

 rated communities, but the passage of certain objects through a 

 number of intervening hands. Thus, the painted aeneolithic pottery 

 of China does not prove that there was intercourse between its 

 makers and the early inhabitants of Susa and Babylonia; all that 

 can be inferred is that at a particular period in the history of Asia 

 there was a trade which passed slowly through a multitude of sepa- 

 rate communities and races, generally assuming on its way peculiari- 

 ties of its own.^- Even the megalithic monuments erected to the 



^ In an article on the oarly pottery of Ur (Antiquaries' .Tourn., vol. 9, No. 4, p_ 344), 

 Mr. Frankfort justly remarks : " Generalizations are of no avail. Thinness and thicknes.s 

 of ware has been taken to illustrate differences in period as long as Susa only was 

 considered ; ' Musyan ' has disqualified this criterion. Polychromy and monochromy of 

 decoration no longer provide chronological indications, since Sir Aurel Stein's discoveries 

 in Seistan and Professor Langdon's work at Jemdet Nasr suggest that certain monochrome 

 fabrics survive and overlap the polychrome wares. Yet all those isolated qualities are 

 still used as criteria for classification. But only if we insist on a many-sided resemblance, 

 a resemblance affecting technique, shape, and decoration alike, before claiming the 

 existence of relationship between wares from different sites, is there any likelihood of 

 our not combining heterogeneous elements." 



