Please use this identifier to cite or link to this item:
|Title:||Spatial Expressions and Case in South Asian Languages|
|Authors:||Tafseer, Ahmed Khan|
|Abstract:||This dissertation investigates the relation of case and spatial expressions in South Asian languages. The dissertation surveys South Asian languages with respect to synchronic case usage, and identifies interesting and/or unusual patterns and proposes a lexical semantic explanation for the patterns in terms of an underspecified feature-based model for spatial relations and an identification of metaphors for the extension into the nonspatial domain. The data for this dissertation were obtained by a survey of case marker usages in ten South Asian languages. South Asia is considered as a “linguistic area” or Sprachbund (Emeneau 1956) leading to the convergence of linguistic features due to language contact persisting over centuries. Because of this reason, six Indo-Aryan (Haryani, Nepali, Punjabi, Saraiki, Sindhi and Urdu/Hindi), two Iranian (Balochi and Pashto), one Dravidian (Malayalam) and one Tibeto-Burman (Manipuri) language were selected for the study. The survey confirms the observation that there are areal features that are found in common among the languages of different families. The study of case marker usages mainly focuses on synchronic issues. However, it also investigates the origin of case markers. It is found that many modern core case markers (e.g., ergative and accusative markers) originate from spatial terms. Some of these forms e.g., Haryani ergative/accusative nae and Urdu/Hindi accusative/dative ko have spatial usages in today’s language as well. This observation shows the deep connection between case and spatial expressions. This connection is explored in more detail from two perspectives. For one, the dissertation tries to understand the patterns of usage and multiple senses across the surveyed languages just within the domain of spatial relations. For another, the dissertation identifies new verb classes within South Asian. This identification of verb classes represents a novel contribution to the field of lexical semantics. In trying to understand the pattern of case marking on these verb classes, the dissertation further investigates how spatial markers can be extended into a non-spatial case marking domain. With respect to just the spatial domain, it is shown that the theoretical frameworks developed in earlier studies on spatial markers such as Ostler (1979), Jackendoff (1990) and Kracht (2002) cannot explain all the problems of fine-grained differences and polysemy of spatial case markers found in South Asian languages. For a coherent and systematic explanation of these problems, this dissertation therefore proposes an alternative underspecified feature-based model. With respect to the issue of how spatial marking can extend into a non-spatial domain, the dissertation explores the non-spatial usages of spatial case markers. Some of the forms that are used as spatial markers also mark instrument, addressee and noncanonical second arguments. The study of the constructions with non-canonically marked second arguments provides six verb classes that are common in most of these languages. This method of verb classification is parallel to Levin (1993), who claims that the verbs sharing the same syntactical structure form a coherent semantic class. The establishment of verb classes with a systematic use of non-canonical second argument (NCSA) leads us to try to understand the semantic factors behind the choice of the case marker for NCSAs. The dissertation provides the following major findings for the South Asian languages. Some South Asian languages distinguish between static and dynamic sources, e.g., the Nepali ablatives dekhi vs. baaTa. The spatial domain supplies metaphors to the nonspatial domains. This is the reason for the use of spatial markers in non-spatial domains. There can be more than one spatial metaphor for a non-spatial usage. Different languages may select different metaphors for the same usage, e.g., the addressee of the verb ‘ask’ can be considered as a (dative marked) recipient or a (ablative marked) potential source. The choice of the case marker on an argument does not solely depend on its semantic features. The semantic features of the whole clause may influence on the choice of the case marker. The identified NCSA verb classes were not predicted by any earlier theory. Non-canonical, e.g., dative, marker for the experiencer subject is a known phenomenon (Verma & Mohanan 1990), but the non-canonical marking on the second argument of an experiencer subject construction was not systematically studied earlier. Hence both the diachronic development of modern case markers and synchronic nonspatial usage of spatial markers show the deep relation between spatial expression and case. The non-spatial domain borrows its metaphors from the spatial domain and spatial markers can therefore be extended to mark non-spatial usages in a very regular manner. Previous models of spatial usage and NCSAs were shown to be inadequate for accounting for the South Asian data. This dissertation therefore proposes an alternative feature-based model for the spatial domain and extends our understanding of the semantic factors involved with respect to the non-spatial domain.|
|Appears in Collections:||Dissertations (restricted access)|
Files in This Item:
|tafseer_dissertation.pdf||11.69 MB||Adobe PDF||View/Open Request a copy|
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.