Like most other examples of "bad grammar" encountered in American the compound negative is of great antiquity and was once quite respectable. The student of Anglo-Saxon encounters it constantly. In that language the negative of the verb was formed by prefixing a particle, /ne/. Thus, /singan/ (=/to sing/) became /ne singan/ (=/not to sing/). In case the verb began with a vowel the /ne/ dropped its /e/ and was combined with the verb, as in /naefre/ (never), from /ne-aefre/ (=/not ever/). In case the verb began with an /h/ or a /w/ followed by a vowel, the /h/ or /w/ of the verb and the /e/ of /ne/ were both dropped, as in /naefth/ (=/has not/), from /ne-haefth/ (=/not has/), and /nolde/ (=/would not/), from /ne-wolde/. Finally, in case the vowel following a /w/ was an /i/, it changed to /y/, as in /nyste/ (=/knew not/), from /ne-wiste/. But inasmuch as Anglo-Saxon was a fully inflected language the inflections for the negative did not stop with the verbs; the indefinite article, the indefinite pronoun and even some of the nouns were also inflected, and survivors of those forms appear to this day in such words as /none/ and /nothing/. Moreover, when an actual inflection was impossible it was the practise to insert this /ne/ before a word, in the sense of our /no/ or /not/. Still more, it came to be the practise to reinforce /ne/, before a vowel, with /n[=a]/ (=/not/) or /naht/ (=/nothing/), which later degenerated to /nat/ and /not/. As a result, there were fearful and wonderful combinations of negatives, some of them fully matching the best efforts of Lardner's baseball player. Sweet [Pg233] gives several curious examples.[73] "N[=a]n ne dorste n[=a]n thing [=a]scian," translated literally, becomes "/no/ one dares /not/ ask /nothing/." "Thaet hus n[=a] ne feoll" becomes "the house did /not/ fall /not/." As for the Middle English "he /never/ nadde /nothing/," it has too modern and familiar a ring to need translating at all. Chaucer, at the beginning of the period of transition to Modern English, used the double negative with the utmost freedom. In "The Knight's Tale" is this:
He /nevere/ yet /no/ vileynye /ne/ sayde In al his lyf unto /no/ maner wight.
By the time of Shakespeare this license was already much restricted, but a good many double negatives are nevertheless to be found in his plays, and he was particularly shaky in the use of /nor/. In "Richard III" one finds "I never was /nor never/ will be"; in "Measure for Measure," "harp not on that /nor/ do /not/ banish treason," and in "Romeo and Juliet," "thou expectedst not, /nor/ I looked not for." This misuse of /nor/ is still very frequent. In other directions, too, the older forms show a tendency to survive all the assaults of grammarians. "/No/ it /doesn't/," heard every day and by no means from the ignorant only, is a sort of double negative. The insertion of /but/ before that, as in "I doubt /but/ that" and "there is no question /but/ that," makes a double negative that is probably full-blown. Nevertheless, as we have seen, it is heard on the floor of Congress every day, and the Fowlers show that it is also common in England.[74] Even worse forms get into the /Congressional Record/. Not long ago, for example, I encountered "without /hardly/ an exception" in a public paper of the utmost importance.[75] There are, indeed, situations in which the double negative leaps to the lips or from the pen almost irresistibly; even such careful writers as Huxley, Robert Louis Stevenson and Leslie Stephen have [Pg234] occasionally dallied with it.[76] 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,[77] "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."