Semantic and phonological correlates of affix order

词缀顺序对循环理论和优化扩展到形态学的发展都至关重要，但也提出了一些实质性的问题，特别是所谓的“templatic”或位置类形态学中的平面任意顺序的证据，以及导致经典词汇形态学消亡的形态语义和语音词缀属性之间的明显差异。从类型学的角度来看，这个项目的动机是词缀顺序的研究留下了巨大的经验空白。虽然有很好的个案研究的具体方面，单一的语言或语言的家庭，还没有系统的类型学研究的基本语音属性的词缀（如词缀大小和它们触发的变化）与词缀顺序。在理论方面，从词汇音系学到层次优选论的转变揭示了如何将词缀排序整合到优化方法中的核心理论问题。在越来越多的情况下，语音至少部分地决定了词素线性化，这一挑战尤其紧迫。我们的期望是，考虑到语音和语义相关的词缀顺序串联，不仅导致原则性的解决方案，显然任意模板词缀排序模式的情况下，而且还允许诊断前缀和后缀之间的层次关系，这不能依靠它们的线性位置来完成。该项目将为可能的词缀顺序汇编不同的跨语言语言样本，包括其整体排序系统和特定的子系统（例如，代词词缀和保持范畴的派生词缀），并系统地研究词缀排序与语义（例如特异性和范围）和语音属性（例如词缀的韵律大小和语音变化）之间的相关性。我们的工作假设是，词缀顺序和它的语音效果可以捕获在一个理论上的保守和最低限度的方式追求最小的扩展到一个经典的词汇主义的语法架构，采用层优选理论的语音和词汇分解语法的增量结构建设的语义表示。在我们的正式分析中，我们将解决四种类型的数据，这是潜在的问题，这个限制性的整体框架：（1）语义任意词缀顺序，（2）多个向心域，（3）梯度相关的排序，（4）不连续的语音和语义依赖。

Affix order has been crucial for the development of both, theories of cyclicity and the extension of optimization to morphology, but raises also substantial problems for both, especially the evidence for flat arbitrary ordering in so-called "templatic" or position class morphology, and the apparent discrepancies between morphological-semantic and phonological affix properties which have led to the demise of classical Lexical Morphology. Typologically, the motivation for this project is that research on affix order has left huge empirical gaps. While there are excellent case studies on specific aspects for single languages or language families, there have been no systematic typological studies relating basic phonological properties of affixes (such as affix size and the alternations they trigger) with affix order. On the theoretical side, the shift from Lexical Phonology to Stratal OT has opened up the central theoretical question how to integrate affix ordering into an optimization approach. This challenge is especially pressing in face of a growing body of cases where phonology at least partially determines morpheme linearization.Our expectation is that taking into account phonological and semantic correlates of affix order in tandem, not only leads to principled solutions for cases of apparently arbitrary templatic affix ordering patterns, but also allows for diagnosing hierarchical relations between prefixes and suffixes which cannot be done by virtue of their linear position. The project will compile different cross-linguistic language samples for possible affix orders, both for their overall ordering systems and for specific subsystems (e.g., pronominal affixes and category-preserving derivational affixes), and systematically investigate correlations between affix ordering and semantic (e.g. specificity and scope) and phonological properties (e.g. prosodic size and phonological alternations of affixes). Our working hypothesis is that affix order and its phonological effects can be captured in a theoretically conservative and minimalist way by pursuing minimal extensions to a classical lexicalist architecture of grammar by adopting Stratal Optimality Theory for phonology and Lexical Decomposition Grammar for the incremental structure building of semantic representations. In our formal analyses, we will address four types of data which are potentially problematic for this restrictive overall framework: (1) semantically arbitrary affix orders, (2) multiple endocentric domains, (3) gradient correlates of ordering, and (4) discontinuous phonological and semantic dependencies.