OUTLINES OF A CONSTRAINT ON SYNCRETISM
OUTLINES OF A CONSTRAINT ON SYNCRETISM
复制标题
限制综合主义的概述
DOI：
10.1515/flin.1984.18.1-2.73
复制
发表时间：
1984
期刊：
影响因子：
--
通讯作者：
A. Carstairs
中科院分区：
文献类型：
--
作者：
A. Carstairs
近几十年来，一些语言学家断断续续地关注着合音现象。他们集中在三个主要方面：（a）一个波段上的形态和词汇的同音异义与另一个波段上的语音“中性化”之间的平行关系;（B）对融合可能产生的歧义的限制，以及这种限制是如何实施的;（c）形态句法属性之间的关系，这些属性独立于它们的语义实现，它们要么有利于要么抑制融合。第一个问题是30年前巴黎语言研究所向40多位语言学家发出的一份问卷调查的主题，但没有得出结论（TIL 1957）。第二个问题在卡洛（1968）对凯森的中和作用的研究中处于突出地位，在派克（1965）对德语形态学“矩阵”的研究以及普朗克（1979; 1980）最近的一些研究中也很突出。最后一个是由雅各布森在他的经典文章“Beitrag zur allgemeinen Kasuslehre”（1936）中探讨的。在这篇文章中，我将关注融合的一个方面，这是不同于所有这些方面-一个方面涉及约束的关系之间的形态句法属性和它们的双曲型指数。在性质和它们的指数之间可以得到的最简单的关系是一对一配对。这是在VBNNEMANN（1972）称之为“洪堡宇宙被严格遵守的地方获得的关系;它也是COMBIE（出现）称之为”规范凝集“的特征。但我所遇到的所有屈折语言都表现出一些
Syncretism, or inflexional homonymy, has received attention intermittently from a number of linguists in recent decades. They have concentrated on three main aspects: (a) the parallel between morphological and lexical homonymy on the one band and phonological 'neutralisation' on the other; (b) the limits to the ambiguity which syncretism potentially engenders, and how such limits are enforced; (c) relationships between morphosyntactic properties äs such, independently of their inflexional realisations, which either favour or inhibit syncretism. The first of these issues was the subject of a questionnaire circulated thirty years ago to more than forty linguists by the Institut de Linguistique in Paris, with inconclusive results (TIL 1957). The second issue is to the fore in CALLOW'S (1968) study of neutralisation in Käsern, and is also prominent in PIKE'S (1965) work on morphological 'matrices' in German and in some more recent work by PLANK (1979; 1980). The last of the three was explored by JAKOBSON in his classic article 'Beitrag zur allgemeinen Kasuslehre' (1936). In this article I will be concerned with an aspect of syncretism which is distinct from all of these — an aspect involving constraints on the relationship between morphosyntactic properties and their inflexional exponents. The simplest relationship which can obtain between properties and their exponents is that of one-to-one pairing. This is the relationship which obtains where what VBNNEMANN (1972) calls 'Humboldt's Universar is strictly observed; it is also characteristic of what COMBIE (to appear) calls 'canonical agglutination'. But all inflecting languages that I have encountered exhibit some