Some models of the morphology-phonology interface take (certain aspects of) morphology and phonology to be computed in the same component of the grammar, simultaneously, including many instantiations of Optimality Theory (McCarthy and Prince 1993a,b, Kager 1996, Hyman and Inkelas 1997, Mascaro 2007, Wolf 2008, i.a.). On the other hand are models that separate morphology from phonology, including Distributed Morphology (Halle and Marantz 1993, 1994) and related models (e.g., Trommer 2001, Bye and Svenonius 2012, Dawson 2017, Rolle 2020), as well as “subcategorization”-based approaches (Paster 2006, 2009, Yu 2007, i.a.). I undertake a careful study of the order of operations needed to derive the form of finite verbs in the Neo-Aramaic language Turoyo (Jastrow 1993). Two morphophonological phenomena found in Turoyo verbs provide evidence for a separation of morphology from phonology: (i) phonologically-conditioned suppletive allomorphy that is anti-optimizing and surface opaque (reaffirming the findings of Paster 2006); and (ii) phonological displacement of an affix (a la infixation) that is also anti-optimizing and surface opaque, and even more surprisingly, counterbleeds morphological operations in the verbal complex but feeds/bleeds phonological ones. The main conclusion from Turoyo is that exponent choice precedes, and is oblivious to, the regular phonology of the language and considerations of phonological optimization. Turoyo also provides a more general window into a number of issues at the morphology-phonology interface, including cyclicity, the timing of infixation, and constraints on allomorphy.

中文翻译：

---

语音之前的形态：以图罗约（新阿拉姆语）为例

形态学-语音学接口的某些模型需要同时在语法的同一部分中计算形态学和语音学（的某些方面），包括最优性理论的许多实例化（McCarthy和Prince 1993a，b，Kager 1996，Hyman和Inkelas 1997年，Mascaro 2007年，Wolf 2008年，ia）。另一方面，是将形态学和语音学分开的模型，包括分布式形态学（Halle和Marantz 1993，1994）和相关模型（例如Trommer 2001，Bye和Svenonius 2012，Dawson 2017，Rolle 2020）以及“子分类”。 （Paster 2006，2009，Yu 2007，ia）。我仔细研究了导出新亚拉姆语Turoyo（Jastrow 1993）中的有限动词形式所需的操作顺序。在Turoyo动词中发现的两种形态学现象为形态学与语音学的分离提供了证据：（i）语音学条件下的反变同质异形是反优化的且表面不透明（重申Paster 2006的发现）；（ii）词缀（la固定）的音位置换，该词缀也是反优化且表面不透明的，更令人惊讶的是，它抵消了言语复合体中的形态操作，但喂入/排除了音位。Turoyo的主要结论是，指数选择在该语言的常规音系学和音系优化考虑之前，并且被忽略。Turoyo还提供了一个更通用的窗口，可以了解形态学-语音学界面上的许多问题，包括循环性，固定时间和同种异形约束。