440 BUREAU OF AMERICAN ETHNOLOGY [BULt. 40 



the stem itself, and do not readily admit any preceding suffixes. 

 Most of these either harden or weaken the terminal consonant of the 

 stem, although there is also a considerable number of suffixes of this 

 class which do not produce a»y changes other than those entailed by 

 purely phonetic laws. In a few cases the changes produced by the 

 sufiix are very irregular. It is probable that no verbal or nominal 

 stem ever appears without a suffix of this class. Therefore the 

 terminal sound of a stem can not be determined unless it occurs 

 with a suffix which produces no change. 



§ 7. Changes in the Phonetic Character of the Stem 



Setting aside the secondary changes produced by the action of 

 phonetic laws and by the mutual effect of stem and suffix, we find that 

 reduplication and change of vowel are used to express grammatical 

 concepts. In the verb we find complete duplication of the stem, with 

 assimilation of the terminal consonant of the first repeated syllable 

 with the following consonant; for instance, loq^- to fish halibut, 

 lox'Hoqwa to fish now and again. True reduplication is, on the 

 whole, restricted to the initial consonant. The vowel of the redupli- 

 cated syllable does not always depend upon the stem-vowel, but 

 differs according to the function of reduplication. Vowel-changes in 

 the stem are rare, and consist generally of a lengthening of the stem- 

 vowel. In many cases they may be explained as modified redupli- 

 cation. 



§ 8. Position 



The position of words in the sentence is determined by syntactic 

 particles. The parts of the sentence are held together firmly, and 

 their position is definitely determined by their coalescence with 

 syntactic elements which indicate the relations of subject, object, 

 instrument, and possession. By this means the whole sentence is 

 knit together so firmly that a separation into words is quite arbitrary. 

 The firmness of this word-complex is due largely to the complete 

 phonetic coalescence of the syntactic particle with the preceding 

 word, and to its function as determining the syntactic value of the 

 following word. It is of course impossible to determine whether this 

 is an original trait of the language, or whether it is due to a phonetic 

 decadence of the syntactic elements, similar to the one that may be 

 observed in French in the combinations between verb and pronoun. 



§§7,8 



