| Special sentence pattern is a hot and difficult topic in linguistics, and also achallenge for syntactic tagging in corpus construction. The essence of syntactic taggingis the formalization of syntactic structure in accordance with the corresponding formalsyntactic theories. At present, two major theories Phrase Structure Grammar (PSG) andDependency Grammar (DG) are adopted in syntactic tagging in equal proportions andeach of them are characterized of their own advantages and limitations. With thecontinuous expansion of corpora and the gradual increasing requirements on corpusapplication, corpus annotation emerges as the core issue in corpus linguistic study. Itconsists of part-of-speech (POS) tagging, syntactic tagging, semantic tagging anddiscourse annotation, within which POS has been developed into a rather mature stageand syntactic tagging has become one of the hot topics as well as difficulties in theresearch of corpus annotation.Information Dependency Language Model (IDLM) put forward by Li Liangyan(2009) is a syntactic tagging oriented theory for corpus construction, which integratesthe methods of both PSG and DG with the reference to conceptual autonomy andconceptual dependency in Cognitive Grammar (CG) to provide language descriptionalong with syntactic and semantic explanation. As IDLM is a new formal syntactictheory, it is essential to conduct a preliminary research on different languagesaccordingly before large-scale corpus construction. On the basis of IDLM, this paper ismainly about the hot and difficult topic in English syntactic tagging—that of Englishspecial sentence patterns. On the one hand, it testifies the feasibility of IDLM in Englishanalysis; on the other hand, it creates a set of syntactic tagging norms of English specialsentence patterns which is the theoretical foundation and knowledge database forlarge-scale English syntactic tagging.The paper is divided into five chapters. Chapter one introduces the motivation,theoretical and practical significance, methodology, research subject and the layout ofthe dissertation. Chapter two is the literature review about current corpus construction,formal syntactic theories, previous study of the English special sentence patterns, and adatabase survey on these sentence patterns. Chapter three is the theoretical foundationfor this paper, which firstly gives an introduction to the major principles and taggingmethods and then further displays how to analyze a specific English sentence pattern with IDLM. Chapter four is the systematic IDLM analysis on five major English specialsentence patterns (middle construction, causative construction, inverted sentence pattern,existential sentence pattern, ditransitive construction). The last chapter is theconclusion.It finds that: firstly, tagmeme transformation such as omission, displacement andcombination and sentence compounding have been applied in IDLM analysis on thefive English special sentence patterns so as to realize the pragmatic purposes—to shiftsentence focus. It is also obvious that some sentence patterns have constructionalmeanings. Secondly, with previous linguistic study as a reference, this paper is adetailed analysis of English special sentence patterns based on IDLM. Therefore, thetagging results are of good interpretability and findability for corpus construction.All in all, IDLM provides more than a set of syntactic tagging solutions to corpusconstruction problems but a new point of view for linguistic study. IDLM analysis onEnglish special sentence patterns can be considered as the perfection and supplement tothe methods of previous syntactic tagging. Through such an analysis, a set of syntactictagging norms for English special sentence pattern has been created as a core databasefor corpus construction, as a comparatively appropriate reference for manual tagging, asthe foundation for designing computer programs of auto-tagging. As a preliminary study,this research is helpful to solve syntactic tagging problems so as to promote corpusconstruction and application. |
PDF Full Text Request