AbstractThe present paper seeks to investigate the characteristics of possessive constructions in Kurdish (the Central variety also known as Sorani) and Hungarian from a cognitive viewpoint. Starting with nominal possessive constructions, which include attenuated possessors and nominal possessees, we will argue that both Kurdish and Hungarian make use of essentially similar typological strategies to encode the possessive relationship. Moreover, the defining characteristics of nominal possessive constructions in both languages will be justifiably accounted for through the same lines of cognitive argumentation in terms of Langacker's reference-point model (2008, 2009). However, a different cognitive treatment is proposed for cases in which the possessor and the possessee are nominal and linked to each other via an Ezafe. We will argue that Ezafe, which links the nominal head to its dependents in a Noun Phrase (Qharib et al. 1971; Moiin 1984; Ghomeshi 1997; Lotfi 2014), evokes an intrinsic asymmetric relationship between the possessor as the landmark and the possessee as the trajector. As the second major type of possessive constructions elaborated on in this paper, the predicative possessive is first classified into topic-possessive and be-possessive categories. Despite their subtle structural differences, it is again shown that both Kurdish and Hungarian employ almost similar clausal patterns to form both categories of predicative possessives, and the same cognitive models can be brought into play to account for their underlying characteristics. The final section of the present paper is devoted to the so-called ergative constructions in Kurdish. Seeking to propose a new cognitive approach to account for the peculiarities of the non-accusative alignment, we will argue that the so-called ergative constructions in Kurdish are conceptually linked to predicative possessive constructions.

中文翻译：

---

认知视角下库尔德语和匈牙利语所有格结构的比较研究

摘要本文试图从认知的角度研究库尔德语（中央变体也称为索拉尼语）和匈牙利语中所有格结构的特征。从名义上的所有格结构开始，包括减弱的占有者和名义上的占有者，我们将争辩说，库尔德语和匈牙利语都使用本质上相似的类型学策略来编码所有格关系。此外，根据 Langacker 的参考点模型（2008，2009），两种语言中名词所有格结构的定义特征将通过相同的认知论证路线得到合理解释。然而，对于占有者和被占有者是名义上的并通过 Ezafe 相互联系的情况，提出了不同的认知处理。我们将争辩说，Ezafe，它将名义头部与名词短语中的附属物联系起来（Qharib et al. 1971; Moiin 1984; Ghomeshi 1997; Lotfi 2014），唤起了作为地标的拥有者与作为轨迹的拥有者之间内在的不对称关系。作为本文阐述的第二种主要所有格结构类型，表语所有格首先分为话题所有格和属格所有格。尽管它们存在细微的结构差异，但再次表明，库尔德语和匈牙利语都采用几乎相似的分句模式来形成这两类谓词所有格，并且可以使用相同的认知模型来解释它们的潜在特征。本文的最后一部分专门讨论库尔德语中的作格结构。