230 michelson: American Indian languages 



hand, sounds, morphology, syntax, and vocabulary have no 

 innate connection (vide supra) then we would have a series of 

 various borrowings instead of "stock" in its present sense. It 

 may be well however, to state that no matter how "stock" 

 is defined at present, it certainly is not used in that sense alone, 

 but in a more loose way, almost according to the whim of the 

 author. 



The most decisive proof that two or more languages belong 

 to the same stock, used in a somewhat free sense, is numerous 

 and detailed resemblances in their structures. A large per- 

 centage of vocabulary held in common is a welcome additional 

 proof. The most decisive proof that a single language constitutes 

 a special stock is numerous unique morphological features. No 

 amount of purely lexical resemblances between languages, no 

 matter how far apart geographically, would prove that they 

 belonged to a single stock. For, experience has shown us that 

 vocabulary is very often borrowed in large amounts, and hence 

 is not a good criterion. If the tribes were far apart geograph- 

 ically, that would not preclude the possibility that in prehistoric 

 times they had been in contact, and at that time extensive 

 borrowing had taken place. Another reason why vocabulary is 

 not a good criterion is that the number of words in even distinct 

 stocks that superficially resemble each other is really consider- 

 able. An example is Sanskrit (Indo-European stock) dsan-, 

 Fox (Algonquian stock) asem, both meaning "stone." Compari- 

 sons of vocabulary are only valuable when we know that the 

 morphology of the languages compared are the same, or at least 

 very similar. Otherwise we should not know whether we were 

 dealing with comparable elements, even if the words in their 

 totality resembled each other. For example, the comparison 

 of the pronoun Avestan cis t Greek Hs, Latin quis, Oscan pis, 

 all meaning "who," is entirely justifiable, because the structure 

 of all four languages is fundamentally the same; and, which is 

 also important, it has been shown that though these words 

 apparently resemble each other only slightly, yet as a matter of 

 fact the correspondence of the various sounds forming these 

 words is precisely what we should expect from our knowledge 



