MING DYNASTY 'MOUNTAIN SONGS'
103
written by the literati rhymed often by academic fiat rather than in accordance with actual dialect pronunciations and the conversational styles which we know must have been spoken at the time.
This is not entirely the case since in the verses of the classical poets we often find deviation from the patterns of the rhyming dictionaries. Still the norm held true to the accepted versions, and as time passed the accepted version remained relatively stable while the living language went through a series of sound changes. There is even reason to assume that the earliest rhyming dictionaries may have preserved archaisms or dialect pronunciations, or otherwise mixed the information in a way that would complicate Karlgren's Ancient Chinese. For example, we know from the preface to the Ch'ieh Yün that this important rhyming dictionary was the product of an informal committee composed of members who represented several regional dialects. Presumably a situation like that might lead to a levelling process and the final results might be to some extent an overall pattern of several speech forms rather than a consistent recording of a single dialect.
In summary, the first proposition is that Ancient Chinese as now reconstructed should be paired with a proto-Chinese developed by the comparative method of modern linguistics. One can look forward to the time when the necessary spoken language data will be gathered and the preliminary reconstructions of individual Chinese proto-dialects will be completed. The second proposition is that the standard rhyming dictionaries can be expected to diverge in greater or lesser degree from any standard spoken language of their time. This second point suggests to the linguist that an ideal target for research might be poetry outside the intellectual, classical tradition. In other words, we can look to folk poetry since in that genre we will more likely be dealing with colloquial rhymes having no reference to the educated patterns of the rhyming dictionaries. This type of poetry would provide rhymes which are so useful in reconstructing earlier forms of Chinese, yet it would be much less likely to present some of the problems of the more artificial rhyming dictionaries. It is a safe assumption that original folk poetry would represent the everyday speech of the area from which it comes rather than any prestige second language of the educated class.