Featural Affixes: The Morphology of Phonological Features

特征词缀是通过在基础词的词段上添加音韵特征来表达形态句法类别的词缀（如德语Bruder brother/ br<e:1> der brothers，复数词缀完全由语音前发组成，即重读元音上的音韵特征[-back]的词缀）。虽然特征性词缀是交叉语言学中最常见和最广泛的形态学类型之一，但我们对特征性词缀的理论理解受到两个实质性问题的限制：(1)对该主题的系统研究几乎仅限于它们的音系特性，因此我们对特征词缀作为形态学对象知之甚少；(2)特征词缀研究的经验基础是基于极小的语言样本，并且在关键方面依赖于经验上有缺陷的数据。本项目通过对四个中心领域的特征词缀形态进行详细的实证和理论研究来解决这两个问题：它们的线性化、共现、异型和它们在世界语言中的融合。这些领域一直是最近对分段词缀进行密集的实证和理论研究的主题，但在特征词缀方面却很少受到关注。特征词缀作为最像片段的非连接形态学类型之一，因此为在更广泛的经验领域中检验理论形态学中心假设的经验有效性提供了理想的试验场。更具体地说，我们追求以下目标：(i)对代表性特征词缀进行广泛的跨语言调查，以确定其分布的系统模式；（ii）将描述性和理论范畴从片段词缀扩展到特征词缀；（iii）对特征词缀的片段词缀假设进行测试；（iv）将特征词缀理论整合到限制性语法形式主义中。总之，对特征词缀的深入研究有望大大加深我们对词缀的一般原理的理论理解，而且，由于特征词缀与基本词的音系结构的相互作用比段音词缀更大，音系计算对形态结构的访问，更一般地说，是语法结构中模块的相互作用。

Featural affixes are affixes which express morphosyntactic categories by adding phonological features to segments of a base word (cf. German Bruder brother/Brüder brothers, where plural morphology consists exclusively in vocalic fronting, i.e. the affixation of the phonological feature [-back] on the stressed vowel). Although they are among the most frequent and widespread types of morphology crosslinguistically, our theoretical understanding of featural affixes is limited by two substantial problems: (i) Systematic research on the topic has been limited almost exclusively to their phonological properties, hence we know little about featural affixes as morphological objects, and (ii) the empirical foundation of research on featural affixes is based on an extremely small sample of languages, and relies in crucial respects on empirically flawed data. This project addresses both problems by a detailed empirical and theoretical study on the morphology of featural affixes in four central areas: their linearization, cooccurrence, allomorphy, and their syncretism in the languages of the world. These areas have been the subject of intensive recent empirical and theoretical research for segmental affixes, but have received little attention with respect to featural affixes. Featural affixation as one of the most segment-like types of nonconcatenative morphology thus provides an ideal testing ground for examining the empirical validity of central assumptions of theoretical morphology in a broader empirical domain. More specifically, we pursue the following goals: (i) the creation of an extensive crosslinguistic survey of representative featural affixes to identify systematic patterns of their distribution, (ii) the extension of descriptive and theoretical categories from segmental affixation to featural affixes, (iii) the testing of hypotheses developed for segmental affixation on featural affixes, and (iv) the theoretical integration of featural affixation into a restrictive grammar formalism. Altogether, the in-depth study of featural affixation promises to substantially deepen our theoretical understanding of general principles of affixation beyond segmental affixes, but also, since featural affixes interact more massively with the phonological structure of base words than segmental affixes, the access of phonological computation to morphological structure and, more generally, the interaction of modules in the architecture of the grammar.