72 BUREAU OF AMERICAN ETHNOLOGY [bull. 40 



block which has made it difficult to reach accurate conclusions. The 

 same words may be used with different significance, and by assum- 

 ing the word to have the same significance always, erroneous con- 

 clusions may be reached. It may also be that the word expresses 

 only part of an idea, so that owing to its use the full range of the 

 subject-matter discussed may not be recognized. In the same man- 

 ner the words may be too wide in their significance, including a 

 number of distinct ideas the differences of which in the course of the 

 development of the language were not recognized. Furthermore, we 

 find that, among more primitive tribes, similarities of sound are 

 misunderstood, and that ideas expressed by similar words are con- 

 sidered as similar or identical, and that descriptive terms are mis- 

 understood as expressing an identity, or at least close relationship, 

 between the object described and the group of ideas contained in 

 the description. 



All these traits of human thought, which are known to influence 

 the history of science and which play a more or less important role 

 in the general history of civilization, occur with equal frequency in 

 the thoughts of primitive man. It will be sufficient to give a few 

 examples of these cases. 



One of the most common cases of a group of views due to failure 

 to notice that the same word may signify divers objects, is that 

 based on the belief of the identity of persons bearing the same name. 

 Generally the interpretation is given that a child receives the name 

 of an ancestor because he is believed to be a re-incarnation of the 

 individuality of the ancestor. It seems, however, much more likely 

 that this is not the real reason for the views connected with this 

 custom, which seems due to the fact that no distinction is made 

 between the name and the personality known under the name. The 

 association established between name and individual is so close that 

 the two seem almost inseparable ; and when a name is mentioned, not 

 only the name itself, but also the personality of its bearer, appears 

 before the mind of the speaker. 



Inferences based on peculiar forms of classification of ideas, and 

 due to the fact that a whole group of distinct ideas are expressed 

 by a single term, occur commonly in the terms of relationship 

 of various languages; as, for instance, in our term uncle, which 

 means the two distinct classes of father's brother and mother's 



