Negation in Archaic Chinese

The late George A. Kennedy, in his 'Negatives in Classical Chinese',' notes that there are some sixteen forms of the negative 'all loosely defined as no, ', and sets out to discover if some criterion other than 'meaning' can provide a tighter classification and perhaps an explanation for this seemingly random and interchangeable repertory of negatives in Classical Chinese. Kennedy notes that all negatives have a labial initial, which permits a classification of the negatives into two series, one with unaspirated *pand one with *m-. He then seeks for an explanation of the *p-/*mcontrast, arriving at a hypothesis that 'negatives beginning with an initial pare concerned with denying an identity', while those 'beginning with initial mare concerned with denying the existence of something', thus according class-meanings to the *p-/*mmutation. He also notes that a contrast in tone may distinguish occurrence in final and in nonfinal position. Kennedy has, perhaps, taken the problem as far as it is possible to go, working from phonological evidence alone. This evidence permits of a formal two-class classification (by initial) and of the setting up within these classes of subclasses (by tone change). It is possible, however, to arrive at a more rigid and, in practice, more useful classification of the negatives of Archaic Chinese by taking a syntactical approach to the problem. But to do this 'Archaic Chinese' and 'Classical Chinese' need more rigorous definition than grammarians working in the field have hitherto given to them. The corpus of texts that provides the selection of negatives with which Kennedy attempts to deal, for example, comprises the Classical Canon and later texts deliberately written in imitation of the language of the Canon. In the former, texts dating from the 11th and 10th Centuries B.C. and from the 4th and 3rd Centuries B.C. are gathered together indiscriminately. In the latter, the Canon as a whole is drawn upon as a pool of precedent for achieving a style approximating to that of the Classics. The whole is characterised as ku wen, i.e. Archaic or Classical Chinese, and is treated as though it were a homogeneous and consistent whole. This is, of course, not so. In the Canon itself (even when those parts thought to be spurious or interpolated are excluded) authorship differs as to date by as much as eight hundred years. During this time almost the entire repertory of the grammatical auxiliaries of Archaic Chinese undergoes change. In 'Classical writing', usages borrowed from the Canon are indiscriminately drawn from all periods and often used in blunted and imprecise ways. In such writing too, there is an admixture of neologisms, which leads to the according of the status of classical particles to forms that do not appear until Han times. In a corpus of material as disparate as this, it is not surprising that textual authority can be found at some point or other for interchanging the entire repertory of negatives the one