Did Proto-Mandarin Exist?
「原始官话」存在过没有?
William H. Baxter 白一平

Abstract 摘要
From a synchronic point of view, the 'Mandarin' dialects of Chinese have a certain commonality: they lack the quanzhuo 全浊 'fully muddy' (voiced obstruent) initials; they generally have few tonal distinctions; they either lack the rusheng 入声 (entering tone), or have only a glottal stop coda in rusheng, not distinguishing final -p, -t, -k; etc. But was there ever a 'Proto-Mandarin'? It is generally recognized that all the Chinese dialects are part of a single language family: that is, that they have the same origin, and are all descended from a common parent language. But whether the Mandarin dialects themselves are descended from their own parent language, distinct from the other dialects, is not clear. To put it another way, if, using the techniques of comparative linguistics, we reconstruct a 'Proto-Mandarin', based on the modern Mandarin dialects, we do not know what time period this 'Proto-Mandarin' would represent, or whether its descendants might include some non-Mandarin as well as Mandarin dialects. The present paper uses several modern Mandarin dialects to reconstruct some of the phonological characteristics of 'Proto-Mandarin', and shows that some of these characteristics were rather un-Mandarin-like: 1. The ping 平 (level), qu 去 (departing), and ru 入 (entering) tones were all divided into upper and lower registers, so that there were at least seven tones: or alternatively, one could count them as four tones, and assume that the voiced obstruent initials were preserved (as in the modern Hangzhou 杭州 dialect). 2. Although final -p was probably not distinguished from final -t, final -t (including original final -p) was rather clearly distinguished from final -k (as in the modern Nanchang 南昌 dialect). 3. It is possible that the palatal finals of the Zhi 止 and Xie 蟹 rhyme classes were distinguished in more phonological environments than in modern Mandarin dialects, so that 基 jī 'foundation' (from the Zhi class) and 鸡 jī 'chicken' (from the Xie class) were distinct (as in the Guangzhou 广州 dialect). Thus it is possible that the descendants of 'Proto-Mandarin' are not limited to the dialects now considered 'Mandarin'. The modern Mandarin dialects may all belong to a more broadly defined 'Northern dialect group', but they themselves are not necessarily a separate genetic unit.

从共时语言学的观点来看，汉语的「官话方言」（或称「北方方言」）有一定的共同性：没有全浊声母，声调多半比较少，入声或者没有，或者只有喉塞音韵尾，不分 -p、-t、-k，等等。可是，在汉语的历史上，到底有没有「原始官话」？大家都承认，所有的汉语方言都属于一个语系，那就是说，它们都有同一的来源，都是一个共同的始源语的派生语。但官话方言是不是来自一种独自的，不同于其它方言的始源语，这个问题还没弄清楚。换句话说，假如用比较语言学的方法，根据现代官话方言来构拟「原始官话」，这种「原始官话」到底代表什么时代的语言，它的派生语，除了现在的官话方言以外，是不是也包括一些非官话方言，这个问题我们还不知道怎么回答。本文根据几个现代官话方言，构拟「原始官话」的一些语音特点，指出这么一种「原始官话」又有一些不太像现代官话的语音特点：（一）平声、去声、入声都分阴阳，至少有七个声调（或者可以算四个声调而仍保存全浊声母，像现在的杭州话）。（二）虽然入声韵尾大概已经不分-p、-t了，可是 -t（包括原来的 -p）和-k多半分得相当清楚（像现在的南昌话）。（三）止摄和蟹摄的齐齿音韵母可能分得比现代官话多，可能「基」≠「鸡」（像现在的广州话）。这漾，「原始官话」的派生语很可能不限于现在认为「官话」的方言。现在的「官话方言」可以说都属于一种广义的「北方方言系」，可是它们本身不一定应该看做一个独立的谱系单位。

<< Back 返回

Readers 读者