PROCESSES OF CHANGE IN PRONUNCIATION. 189 



a barbarous people, a man of authority and influence, whether by 

 physical defect or from some other cause, commits a fault in articu- 

 lation. It is imitated, in the family first, and then among the rela- 

 tives and neighbors. The peculiarity of pronunciation spreads, and 

 is more marked as it spreads; and if nothing occurs to interfere with 

 it, a phonetic change is accomplished. But is there anything fatal 

 in that ? The change is very like those which take place in costume, 

 or armor, or in the house; a historic fact, having neither more nor 

 less of the character of fatality than other historical facts. It is true 

 that if we go back to the initial cause we find on final analysis a move- 

 ment of the vocal organs; but in what act of our life are not our or- 

 gans the final motive? To assume fatality, it would be necessary to 

 suppose that on a certain day the organs of speech of all the individu- 

 als of a group should be modified in the same manner. 



There is a reason why the phenomena of language should be 

 specially subject to imitation. Being a medium of communication, 

 it would lack its essential condition if it varied as between one per- 

 son and another, and would lose its right to be. Hence the neces- 

 sity of a uniform pronunciation. But this is clearly a matter of 

 social necessity, not of a physical fatality. 



A phonetic change may be adopted; or it may be rejected, after 

 a longer or shorter struggle ; for peoples are composed of individuals 

 who are not all of the same age, or of the same sex, or of equal educa- 

 tion or social position. In the sixteenth century the Parisians were 

 agreed in pronouncing s as r and conversely r as s; Paris became 

 Pasis, and oiseau (bird) became oireau. The poet Marot made this 

 matter the subject of a satire. The usage was contested as a ridicu- 

 lous affectation, and went out, but not without leaving vestiges. The 

 thick utterance of the incredibles of the Directory is another example 

 of a merely passing fashion. 



These fluctuations explain the otherwise incomprehensible vari- 

 ations of geographical maps of dialects. If we make linguistic charts 

 of France in the twelfth, thirteenth, fourteenth, and sixteenth cen- 

 turies, we shall find boundaries changing, one province growing 

 larger, another smaller, and reverse changes perhaps taking place 

 in the following period. Revolutions occur much like those of 

 political power, but the two do not always coincide. 



Phonetic changes may therefore be tracked. The interchange of 

 8 and r is thus traced from its beginning in Koussillon northward 

 through France, till it reached the Norman islands; a second sub- 

 stitution of Germanic consonants, which established a difference 

 between High German on one side and Low Dutch, English, and 

 Scandinavian on the other, passed from south to north to the fifty- 

 first and fifty-second degrees of latitude. The fact is thus explained 



