個體量詞在漢語語族中的分布──晉語的個案研究

Distribution of Sortal Classifiers in Sinitic Languages: A Case Study of the Jin Language

依據亞太地區量詞的「單一起源假設」，僅北方漢語之量詞為原生系統，其他量詞語言都是因語言接觸而發展出量詞（Her and Li 2023）。然而，文獻中普遍認為南方漢語中個體量詞的數量多於北方漢語。這個「南強北弱」的格局成為西方學者反對北方漢語為量詞起源的重要理由，轉而支持壯侗語。Her、Li提出北方漢語的「阿爾泰化」作為反駁，主張其個體量詞的數量乃因此由強轉弱，並非自始即弱。因為晉語是北方漢語中唯一的非官話語言，本文的主旨是實際檢驗晉語中的個體量詞，並比較其與南方漢語之數量。基於數詞與量詞分別為乘數與被乘數的乘法屬性，Her（2012）發展出嚴格區分個體量詞與計量量詞的語意基礎以及語法測試方法。本文在此基礎上，逐一檢視文獻中的晉語量詞，首次明確認定個體量詞的實際數量。依據晉語方言點由南到北的地理分布，本文選取了6個方言點進行檢驗，發現個體量詞較少的約有30個，最多的則超過50個，根據各方言點共用的情況，確定晉語有至少30個常用個體量詞。與南方漢語的比較顯示，漢語個體量詞的分布雖然大體上仍是「南強北弱」，但北方漢語並非積弱不振，這個發現因此有利於亞太地區量詞起源於北方漢語的假設。

According to the“single origin hypothesis”of numeral classifiers in the Asia-Pacific region, classifiers in Northern Chinese are the original system, while those in other languages developed due to language contact (Her and Li 2023). However, the literature generally agrees that Southern Chinese has more sortal classifiers than its northern counterpart. This“strong-South-weak-North”pattern has been a significant argument for Western scholars opposing Northern Chinese as the origin of classifiers and instead supporting Tai-Kadai languages. Her and Li counter this argument with the“Altaicization”of Northern Chinese, claiming that the number of sortal classifiers in Northern Chinese has decreased over time, rather than being“weak”from the beginning. Since the Jin language is the only non-Mandarin Sinitic language in Northern China, the present article aims to closely examine the sortal classifiers in Jin and compare their quantity with those of Southern Sinitic languages. Based on the multiplicative property of the numeral as a multiplier and the classifier as a multiplicand, Her (2012) has developed a semantic thesis and a set of grammatical tests to distinguish between sortal classifiers and mensural classifiers. Building on this foundation, this article examines each classifier put forth in the literature on Jin classifiers to confirm the actual number of sortal classifiers. Six Jin dialects are selected from south to north based on their geographical distribution, and results show that the number of sortal classifiers in these dialects range from about 30 to over 50. Based on the shared classifiers among different dialects, Jin has at least 30 commonly used sortal classifiers. A comparison with Southern Sinitic languages shows that, though the distribution of sortal classifiers indeed matches the“strong-South-weak-North”pattern, Jin Chinese is by no means inherently“weak”in using sortal classifiers. This finding therefore supports the hypothesis that classifiers in the Asia-Pacific region originated in Northern China.