<p>In this paper, we provide an in-depth case study of properties of the verbal domain in the Indo-Iranian language Sinhala across different modules of grammar. Specifically, we investigate two seemingly independent grammatical phenomena (i) the phonological (under)application of umlaut and its relation to morphological structure and (ii) the syntactically conditioned choice of negation. As we argue, these phenomena strongly suggest respective analyses in terms of locality domains. Crucially, we find that locality domains in the syntax and the ones in phonology are essentially isomorphic. The set of constructions that we analyse as bi-domainal in the syntax corresponds exactly to the set of constructions that we analyze as bi-domainal in the phonology. This cannot be treated as accidental and we devise a model that derives these parallels in a systematic way. The model rests on the assumptions that (i) the syntax creates locality domains and inherits them to the morphology and the phonology and (ii) specific processes in these modules can manipulate the locality domains in a limited way. This allows us to maintain a restricted model of the syntax-morpho-phonology mapping that proves flexible enough to incorporate systematic exceptions. Such an approach paves the way to a more nuanced understanding of mismatches between different modules.</p>

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Matching domains: The syntax, morphology, and phonology of the verb in Sinhala

  • Paula Fenger,
  • Philipp Weisser

摘要

In this paper, we provide an in-depth case study of properties of the verbal domain in the Indo-Iranian language Sinhala across different modules of grammar. Specifically, we investigate two seemingly independent grammatical phenomena (i) the phonological (under)application of umlaut and its relation to morphological structure and (ii) the syntactically conditioned choice of negation. As we argue, these phenomena strongly suggest respective analyses in terms of locality domains. Crucially, we find that locality domains in the syntax and the ones in phonology are essentially isomorphic. The set of constructions that we analyse as bi-domainal in the syntax corresponds exactly to the set of constructions that we analyze as bi-domainal in the phonology. This cannot be treated as accidental and we devise a model that derives these parallels in a systematic way. The model rests on the assumptions that (i) the syntax creates locality domains and inherits them to the morphology and the phonology and (ii) specific processes in these modules can manipulate the locality domains in a limited way. This allows us to maintain a restricted model of the syntax-morpho-phonology mapping that proves flexible enough to incorporate systematic exceptions. Such an approach paves the way to a more nuanced understanding of mismatches between different modules.