PHONOLOGY. 227 



may also be pronounced patatclia, patatsa, patadsa, but padatcha or ba- 

 datsa is scarcely ever heard from natives. Some terms, as pipa paper, 

 ndaui three, undergo no vocalic or other changes whatever, while others 

 cannot assume ceitain alternations without a change of signification. Cf. 

 Homonymy. 



All these conversions of cognate sounds often impart to certain words 

 a quite different appearance, which renders them unrecognizable to the 

 unexperienced. Still the interchange of sounds is more extensively devel- 

 oped in some dialects of the Caril) oi- Galibi, as well as in Kayowe, Hi- 

 datsa, and other languages spoken on the Mississippi plains. 



, Like all phenomena in nature, this interchangeability is not produced 

 Ijy the fancy or ()[)tion of the natives, Init is based on natural laws, and as 

 language is one of the efiects of nature, we must look to physiology and 

 not to psychology to discover its latent causes. One of these is the tend- 

 ency of rendering pronunciation easier; this we perceive, e. g., in the 

 dropping of the laryngeal sound h in: nji-ut for mi hilt, atunk for (\i hunk, 

 n'unk for ni {or nn) luuik, and also in a/ut for a hu't. It will be remem- 

 bered that h can be drop}jed even when belonging to the body of the word. 

 In U7, 1, hunk kiuliga has probably been nasalized into hunk n^iuliga to 

 avoid the collision of two identical sounds. Another cause of these permu- 

 tations is the laryngeal utterance of the Indians, which I have discussed 

 under that heading (pp. 215-21 7); it also accounts for the circumstance that 

 permutation among sounds originating in the rear mouth aie much more 

 frecpient than those produced by the action of the lips and the fore part of 

 the vocal tube. 



PHONETIC FIGURES. 



Besides the phonetic changes spoken of in the foregoing section, there 

 are other alterations in the sounds of words which generally affect the 

 body of the words more thoroughh , and occur in all tlie languages ex- 

 plored. These alterations are produced by various causes, as the shifting 

 of the accent from one syllable or word to another, the attenuation or 

 increase in quantity, the habit of fast speaking, etc., and chief of all, the 

 desire of .saving vocal exertion. The tendency for retrenchment is more 



