34 H. L. RROOMALL : 



tation has been modified in the more progressive centres of 

 English-speaking- people. So, too, the farmer's cattle call coo 

 is old English cu, while the standard coiv is the modified pro- 

 nunciation of ci(, just as i]}us, nt and Iius have become mouse, 

 02it and house. In vandue and anvelopc, for vendue and enve- 

 lope, there is still left an imitation of the French nasal sound 

 of e)i belonging to the words when adopted by English. 



Errors of grammar arising under the same process are 

 innumerable. One of the commonest is the use of the double 

 negative without the forms neutralizing one another and leav- 

 ing the phrase affirmative. In the Anglo-Saxon Gospels of 

 about 995 is written: " Ne geseah naefre nan man God," 

 literally " Not saw never none man God," now worded (John 

 i-i8) " No man hath seen God at any time." The negative 

 idea is here attached to each negatived word in the phrase, 

 just as, for illustration, the grammatical gender of a French 

 noun appears also in its accompanying article and adjective, 

 " le bon homme " and '-'la bonne femme." Our present 

 vulgar double negative is the survival of this old mode of 

 negation, becoming an eiTor now because the standard has 

 developed a new method. The forms hern, ourn, his'n, are 

 modern instances of the use of the old English adjectival 

 suffix ;/. It is still represented among connect forms by mine 

 and thine as contrasted with the abbreviated my and thy. So, 

 too, them horses preserves the same grammatical suffix /;/ that 

 the standard still holds proper in him and iciiom . 



Just as the foregoing eiTors arise from maintaining the Imi- 

 tation too long or too well, so, conversely, all other kinds of 

 errors result from modifying the Imitation too much or too 

 soon. This erroneous modification, however, is always along 

 lines or directions of development already operating among 

 standard forms. Just as the former errors are Survivals of past 

 correct forms,. so these are Anticipants of future coirect forms. 



Anticipant errors are modifications. In order that we may 

 consider them comprehensively, we may generalize the process 

 of language as primarily — 



