DoctorateOpen Access

Giving in Turkish: A corpus-driven frame semantic description

Is this your thesis?

This record came from a bulk archive import. If it’s yours, link it to your profile.

2023
0 views
0 downloads

Abstract (EN)

The lexicons have become one of the focus areas of lexical semantics and other linguistic studies in terms of providing key information in both grammatical structures and language processes. Particularly, the category of verb plays a significant role in the identification of sentence patterns, and generally in the formation of linguistic theory and framework, so the studies on semantic descriptions of verbs take a large place in the literature (Jackendoff, 1983; Levin, 1993). In this study, the semantic and syntactic features of the verb ver- 'give' in Turkish is analysed on the basis of frame semantics theory (Fillmore, 1982, 1983, 1992; Fillmore & Atkins, 1992) and following a corpus-driven approach. The semantic and grammatical structure of sentences are investigated capturing the relationship between semantic roles (agent, patient) and grammatical roles (subject, object). The study is significant in terms of compiling the first corpus-driven and frame-based dictionary of the actual realizations of the verb through a large data set. The lexical decription of the verb is based on the theory that speakers of a language know the meaning of a word only if they first understand the background frames that motivate the concept that the word encodes. Therefore, word senses are not related to each other directly; they are related to each other by way of their links to common background frames and the manner in which word senses highlight the particular elements of these frames (Fillmore & Atkins, 1992: 77). The database of our study is constructed out of the concordace lines of the verb ver- 'give' in 10-million-word subcorpus of Turkish National Corpus, so each concordance line is categorized and classified adopting a frame-based (Fillmore & Atkins, 1992) and corpus-driven approach making reference to the classification methods in the Berkeley FrameNet project (https://framenet.icsi.berkeley.edu) (Baker et al., 1998; Fillmore et al., 2003; Fillmore & Baker, 2010; Ruppenhofer et al., 2010), which is an online lexical database of English lexical units built on the principles of frame semantics. Based on the comprehensive and multi-directional corpus-driven analyses of the examples of actual language use, in the present study, a) the sentences including the verb ver- as predicates are annotated with the aim of forming a detailed syntactic and semantic description; b) the semantic frames that the verb evokes, their descriptions, and the relation between these frames are identified; c) the frame elements representing the semantic roles in frames are listed and categorized as core or non-core for each sense; d) the sentences exhibiting common collocations and showing all major syntactic contexts are selected, so the constituents of sentences that express these frame elements are labelled and a specific syntactic role category is assigned to each frame element; e) the algorithm of the underlying syntactic and semantic features of the frames and their schematic constructions are suggested. Keywords: frame semantics, corpus linguistics, FrameNet, the verb ver-.

Author

Yasemin Erköse

How to Cite

Yasemin Erköse (Doctorate thesis). Giving in Turkish: A corpus-driven frame semantic description, 2023, Mersin University.

License

Tüm Hakları Saklıdır

This work is shared under the specified license terms.

More theses from Mersin University