Die interne Struktur von Personen-Portmanteaus

人称合成词是将人称作为动词性谓语的主语和宾语的表达捆绑成一个单一形态单位的标记。虽然人称合成词实现了一种非常简单的替代更熟悉的交叉引用口头论点的策略，但它们具有跨语言的高度标记性，并且往往仅限于非常具体的主语和宾语的组合。因此，人称组合可以被视为一种自成一体的参数编码类型，也是对标记头部语言中参数编码的一般属性的试金石。人称组合词的一个重要（虽然有些矛盾）特性是，它们不一定是不可分析的。许多这样的标记可以被理解为同步和/或历时的非合成词，要么是因为它们是由两个不同的人称构筑物连接而成的，要么是因为它们在形式上与非合成词的人称标记相同。基于这些观察，本项目解决了以下密切相关的问题：1)组合词在范式中的位置是否存在系统限制？2)合成词在多大程度上是可分析的？3)混合词的正式表示是什么？4)在密切相关的语言群体中，组合词微变异的程度和分布是什么？5)合成词在历史上是如何发展的（它们又是如何消失的）？为了回答这些问题，我们结合了类型学、形式形态学和历史重建的方法。除了为合成词构建跨语言类型数据库以提取一般属性外，我们还重点研究了三个小语种（尼罗语、易洛魁语和乌拉尔语）中合成词的行为。对于这些语言，我们将对相关的人称组合词进行深入的形式分析，调查这些家庭的微变化范围，并试图在整个代词交叉引用系统的背景下重建人称组合词的历时发展（基于现有的描述性文献）。从不同方法的创新组合到微变化数据，我们期望对人的组合有更好的理解，更普遍地说，在口头论点的交叉引用系统中的融合。

Person portmanteaus are markers which bundle the expression of person for subject arid object of a verbal predicate into a single morphological unit. Although person portmanteaus implement a radically simple altemative to more familiar sttategies of crossreferencing verbal arguments, they are cross-linguistically highly marked, and tend to be restricted to very specific combinations of subject and object. Person portmanteaus may therefore be regarded as a type of argument encoding sui generis as well as a litmus test for general properties of argument encoding in head-marking languages. An important though somewhat paradox property of person portmanteaus is that they are not necessarily unanalyzable. Many such markers can be understood synchronically and/or diachronically as non-portmanteaus, either because they are formed by concatenation from two different person formatives, or because they are identical in form to non-portmanteau person markers.Based on these observations, this project addresses the following closely related questions: 1) Are there systematic restrictions on the location of portmanteaus in paradigms? 2) To which extent are portmanteaus analyzable? 3) What is the formal representation of portmanteaus? 4) What is the extent and distribution of portmanteau microvariation within groups of closely related languages? 5) How do portmanteaus develop historically (and how do they disappear)? To answer these questions, we combine methods from typology, formal morphology, and historical reconstmction. Apart from constmcting a cross-linguistic typological database for person portmanteaus to extract general properties, we focus on the behavior of portmanteaus in three small-scale language families (Nilotic, Iroquoian, and Uralic). For these languages we will provide in-depth formal analyses of the relevant person portmanteaus, investigate the range of microvariation in the families, and attempt to reconstmct the diachronic development of person portmanteaus on the background of the overall pronominal cross-referencing systems (based on the available descriptive literature). From the innovative combination of different methodologies to microvariation data, we expect a better understanding of person portmanteaus, and more generally of syncretism in cross-referencing systems for verbal arguments.