Actionality in Bantu languages: an empirical and typological approach

体的语言学范畴描述事态的内部构成。它与时态不同，时态在时间上定位事态。体通常被理解为二维的：词汇和短语维度（动作性）编码事态的组成阶段，语法维度的设备从中选择特定的阶段（Sasse 2002）。为了理解体，了解动作性的语言具体实例是必要的。然而，对非欧洲语言的声音分析仍然是一个迫切需要（Tatevosov 2002; Bar-el 2015）。班图语提供了特别相关的见解。目前对这一语言家族的研究中一个广为接受的模型假设班图语将复杂的事态词汇化在一个动词中（Botne & Kershner 2000，等等），例如“在愤怒的时候-变得愤怒-生气”。然而，更常见的体理论在先验的基础上排除了这种相位构形。另一方面，在班图语中，似乎也存在着几乎普遍的简单的国家变化（成就）。然而，对更大数据样本的分析却缺失了。此外，很少有诊断测试被发现--例如，结果状态是否实际上是语义编码的，这常常是不清楚的。因此，本研究项目的主要目标是在两个班图语：Nyakyusa（坦桑尼亚）和科萨（南非）的行为的差异化描述。对于Nyakyusa，申请人已经进行了初步研究。科萨语被选中是因为它在遗传学、类型学和地理上的距离，以及可用的文本语料库。在第一步，诊断测试将被开发，以提供词汇化相位配置的见解。这些测试的设计将考虑到其在班图语中的适用性，主要包括评估特定动词或动词短语与特定体词形变化、短语动词（“to begin”、“to stop”、“to finish”）和时间状语的兼容性和可能的阅读。在此基础上，对该合金的抗弯性能进行了研究。将描述Nyakyusa和Xhosa的100个动词和动词短语。数据将通过评估现有语料库中的经证实的话语和通过实地考察相结合来收集。主要方法将是有针对性的启发，结合最近关于语义领域工作的方法论讨论中描述的技术（贡献于Matthewson & Bochnak 2015）。收集的数据将最终导致一个以经验为基础的评价的核心假设，最近的体理论。因此，它将有助于我们理解事态是如何在人类语言中概念化的。研究结果将在同行评议期刊上发表。

The linguistic category of aspect describes the internal composition ofstates-of-affairs. It differs from tense, which localizes states-of-affairsin time. Aspect is commonly understood as bi-dimensional: the lexicaland phrasal dimension (actionality) encodes the constituent phases ofa state-of-affairs, from which the devices of the grammaticaldimension select specific phases (Sasse 2002). In order tounderstand aspect, it is thus essential to know about languagespecificinstantiations of actionality. Sound analyses of non-Europeanlanguages, however, remain a desideratum (Tatevosov 2002; Bar-el2015). Bantu languages promise insights of particular relevance. Amodel well-received in current research on this language familyassumes that Bantu languages lexicalize complex state-of-affairs in asingle verb (Botne & Kershner 2000, among others), for instance ‘tobe at the point of anger – become angry – be angry’. The morecommon theories of aspect, however, exclude such phaseconfigurations on an a priori base. Allegedly universal simplechanges-of-state (achievements), on the other hand, appear to berare in Bantu languages. Yet analyses of bigger data samples aremissing. Moreover, few diagnostic tests are found – it often remainsunclear if, for example, a resultant state is in fact semanticallyencoded. The primary goal of this research project is thus adifferentiated description of actionality in two Bantu languages:Nyakyusa (Tanzania) and Xhosa (South Africa). For Nyakyusa,preliminary research has been performed by the applicant. Xhosa hasbeen chosen for its genetical, typological and geographical distanceand for the text corpora available. In a first step, diagnostic tests willbe developed to provide insights into the lexicalized phaseconfigurations. These tests, which will be devised with their applicability across Bantu languages in mind, will primarily consist inevaluating the compatibility and possible readings of a given verb orverb phrase with specific aspectual inflections, with phasal verbs (‘tobegin’, ‘to stop’, ‘to finish’) and with temporal adverbials. On the basisof these tests, the actional characteristics of approx. 100 verbs andverb phrases in Nyakyusa and Xhosa will be described. Data will becollected through a combination of evaluating attested utterances inthe available corpora and through fieldwork. The main method will betargeted elicitation, incorporating the techniques described in therecent methodology discussion on semantic fieldwork (contributions inMatthewson & Bochnak 2015). The collected data will ultimately leadto an empirically based evaluation of the core assumptions of recenttheories of aspect. It will thereby contribute to our understanding of how states-of-affairs are conceptualized in human language. Theresults will be made available through articles in peer-reviewedjournals.