Page:The American Language.djvu/204

188 children above twelve. Its examination threw a brilliant light upon the speech actually employed by children near the end of their schooling in a typical American city, and, per corollary, upon the speech employed by their parents and other older associates. If anything, the grammatical and syntactical habits revealed were a bit less loose than those of the authentic Volkssprache, for practically all of the written evidence was gathered under conditions which naturally caused the writers to try to write what they conceived to be correct English, and even the oral evidence was conditioned by the admonitory presence of the teachers. Moreover, it must be obvious that a child of the lower classes, during the period of its actual study of grammar, probably speaks better English than at any time before or afterward, for it is only then that any positive pressure is exerted upon it to that end. But even so, the departures from standard usage that were unearthed were numerous and striking, and their tendency to accumulate in definite groups showed plainly the working of general laws.

Thus, no less than 57 per cent of the oral errors reported by the teachers of grades III and VII involved the use of the verb, and nearly half of these, or 24 per cent, of the total, involved a confusion of the past tense form and the perfect participle. Again, double negatives constituted 11 per cent of the errors, and the misuse of adjectives or of adjectival forms for adverbs ran to 4 per cent. Finally, the difficulties of the objective case among the pronouns, the last stronghold of that case in English, were responsible for 7 per cent, thus demonstrating a clear tendency to get rid of it altogether. Now compare the errors of these children, half of whom, as I have just said, were in grade III, and hence wholly uninstructed in formal grammar, with the errors made by children of the second oral group—that is, children of grades VI and VII, in both of which grammar is studied. Dr. Charters' tabulations show scarcely any difference in the