4. Names of persons and classes of persons, and of places, such as /Buckeye/, /Cracker/, /Greaser/, /Hoosier/, /Old Bullion/, /Old Hickory/, the /Little Giant/, /Dixie/, /Gotham/, the /Bay State/, the /Monumental City/.
5. Words which have assumed a new meaning, such as /card/, /clever/, /fork/, /help/, /penny/, /plunder/, /raise/, /rock/, /sack/, /ticket/, /windfall/.
In addition, Thornton added a provisional class of "words and phrases of which I have found earlier examples in American than in English writers; ... with the /caveat/ that further research may reverse the claim"--a class offering specimens in /alarmist/, /capitalize/, /eruptiveness/, /horse of another colour/ (/sic!/), /the jig's up/, /nameable/, /omnibus bill/, /propaganda/ and /whitewash/.
No more than a brief glance at these classifications is needed to show that they hamper the inquiry by limiting its scope--not so much, to be sure, as the ridiculous limitations of White and Lounsbury, but still very seriously. They meet the ends of [Pg034] purely descriptive lexicography, but largely leave out of account some of the most salient characters of a living language, for example, pronunciation and idiom. Only Bartlett and Farmer establish a separate category of Americanisms produced by changes in pronunciation, though even Thornton, of course, is obliged to take notice of such forms as /bust/ and /bile/. None of them, however, goes into the matter at any length, nor even into the matter of etymology. Bartlett's etymologies are scanty and often inaccurate; Schele de Vere's are sometimes quite fanciful; Thornton offers scarcely any at all. The best of these collections of Americanisms, and by long odds, is Thornton's. It presents an enormous mass of quotations, and they are all very carefully dated, and it corrects most of the more obvious errors in the work of earlier inquirers. But its very dependence upon quotations limits it chiefly to the written language, and so the enormously richer materials of the spoken language are passed over, and particularly the materials evolved during the past twenty years. One searches the two fat volumes in vain for such highly characteristic forms as /would of/, /near-accident/, and /buttinski/, the use of /sure/ as an adverb, and the employment of /well/ as a sort of general equivalent of the German /also/.
