occasionally dallied with it.[1] It is perfectly allowable in the Romance languages, and, as we have seen, is almost the rule in the American vulgate. Now and then some anarchistic student of the language boldly defends and even advocates it. "The double negative," said a writer in the London Review a long time ago,[2] "has been abandoned to the great injury of strength of expression." Surely "I won't take nothing" is stronger than either "I will take nothing" or "I won't take anything." "Language begins," says Sayce, "with sentences, not with single words." In a speech in process of rapid development, unrestrained by critical analysis, the tendency to sacrifice the integrity of words to the needs of the complete sentence is especially marked. One finds it clearly in American. Already we have examined various assimilation and composition forms: that'n, use'to, would'a, them 'ere and so on. Many others are observable. Off'n is a good example; it comes from off of and shows a preposition decaying to the form of a mere inflectional particle. One constantly hears "I bought it off'n John." Sort 'a, kind 'a and their like follow in the footsteps of would' a. Usen't follows the analogy of don't and wouldn't. Would've and should 've are widely used; Lardner commonly hears them as would of and should of. The neutral o-particle also appears in other situations, especially before way, as in that 'a way and this' a way. It is found again in a tall, a liaison form of at all.[3]
§8
Pronunciation—Before anything approaching a thorough and profitable study of the sounds of the American common speech is possible, there must be a careful assembling of the materials, and this, unfortunately, still awaits a philologist of sufficient enterprise and equipment. Dr. William A. Read, of the State University of Louisiana, has made some excellent examinations
