Anaphora vs. Agreement: investigating the Anaphor Agreement Effect

照应一致效应（Anaphor Agreement Effect，以下简称AAE）描述了一个最初由Rizzi（1990）提出的观察结果，即没有一种语言能够证明照应词（即表达，如 *herself* our *ourselves*，必须受到它们出现的句子或从句中的先行词的约束）能够触发限定动词上的常规人称、数和/或性别一致。简而言之，这意味着回指语可能永远不会成为常规口头协议的来源。例如，在一种表现出宾语一致性的语言中，像“我们喜欢自己”这样的句子中的动词永远不会接受第一人称复数宾语一致性标记。目前还不清楚AAE是否反映了自然语言的普遍禁令，或者只是语言之间的强烈趋势。也不清楚在避免违反AAE的策略中，确切的变化范围是什么。这是语言不同的一点.例如，在某些语言中，如阿尔巴尼亚语，动词上的一致标记，当被照应词控制时，会采用默认形式（第三人称单数，而不是第一人称复数与第一人称复数回指）;在其他语言，如因纽特人，回指必须出现在一个不规则的，协议阻塞的情况下，例如与格，当它出现在一个位置，它会控制协议;甚至其他语言也有特殊的同意标记，与回指语一起使用，以防止它们触发常规的同意。该项目是对AAE的第一次大规模研究，旨在解决这些不明确之处，一方面通过实证研究AAE的普遍性和语言为避免违反AAE而表现出的策略变化范围，另一方面通过为AAE及其跨语言表现提供理论解释。两名博士生将由PI监督，一个位于莱比锡，一个位于哥廷根，将对一系列语言进行深入的类型学研究，以全面调查AAE。本文的研究不仅有助于对AAE做出更全面的解释，而且有助于我们更好地理解回指约束和言语同意的本质。特别是，这个长期存在的问题，即是否回指约束反映了一种形态句法协议（其中的回指人，数量和性别特征必须同意其先行词），最终可以解决在这个项目中直接检查回指约束和口头协议如何相互作用。

The Anaphor Agreement Effect (AAE henceforth) describes an observation, originally due to Rizzi (1990), that no language is attested where anaphors (i.e. expressions, such as *herself* our *ourselves* that must be bound by an antecedent in the sentence or clause they occur in) are capable of triggering regular person, number, and/or gender agreement on finite verbs. In simple terms, this means that an anaphor may never be the source of regular verbal agreement. For instance, in a language that exhibits object agreement, the verb in a sentence like *we like ourselves* could never receive a 1st person plural object agreement marker. It remains as yet unclear whether the AAE reflects a universal ban in natural language or rather simply a strong tendency among languages. Nor is it clear what the exact range of variation is in strategies for avoiding violations of the AAE. This a point with respect to which languages differ. For instance, in certain languages like Albanian, the agreement marker on the verb, when controlled by an anaphor, takes on a default form (3rd person singular instead of 1st person plural with a 1st person plural anaphor); in other languages like Inuit, the anaphor has to appear in a non-regular, agreement-blocking case, for instance a dative, when it appears in a position where it would have controlled agreement; and even other languages have special agreement markers that go with anaphors to prevent them from triggering regular agreement. This project, the first large-scale examination of the AAE, aims at resolving these unclarities, on the one hand by empirically investigating the universality of the AAE and the range of variation in strategies that languages exhibit to avoid violating it, and on the other hand by providing a theoretical explanation for the AAE and its cross-linguistic manifestations. Two PhD students to be supervised by the PIs, one located in Leipzig and one in Göttingen, will carry out in-depth typological studies on an array of languages to investigate the AAE in its full extent. This research will not only result in a more complete explanation of the AAE, but also yield a better understanding of the nature of anaphoric binding and verbal agreement in general. In particular, the long-standing question as to whether anaphoric binding reflects a type of morphosyntactic agreement (where the anaphors person, number and gender features must agree with that of its antecedent) can finally be addressed in this project by directly examining how that anaphoric binding and verbal agreement interact.