Grammaticalization Of Functional Categories: A Formal Linguistic Perspective

Grammaticalization, a very common phenomenon of languageevolution, is in nature the creation of new functional categories eitherfrom lexical categories or from existing functional ones throughcategorical reanalysis associated with structural simplification.Grammaticalization research is an important issue in the field of historicallinguistics since it is quite enlightening and helpful for exploring thenature and the law of language change.So far, most researches on grammaticalization are conducted from afunctional or typological angle, but those from a formal linguisticperspective are very few. Sometimes, grammaticalization is evenregarded as a challenge to the Formal Linguistics. Hence, the thesis isattempted to analyze grammaticalization of three main functionalcategories in English i.e. auxiliaries (take modals as an example),complementizers (take that as an example) and determiners (take the andan as examples) under the theoretical framework of the FormalLinguistics (mainly the Principles and Parameters theory (PPT), theLanguage Acquisition Hypothesis and the Interface Interpretability) andprovide their grammaticalization with a formal account so as to retain theuniversal explanatory force of the generative grammar. The paper is based on the following assumptions put forward byRoberts & Roussou (2003): in the light of implication from PPT, anassumption is fashioned that language change results from the change ofparameter setting; in terms of the Language Acquisition Hypothesis,another assumption is posed that the change of parameter setting isclosely related to the ambiguity of the input linguistic data. When theinput becomes ambiguous or obscure, parameter setting may change;according to the Interface Interpretability, the third assumption is putforward that creation of new functional categories is due to the change ofthe way in which the functional heads are realized in PF.In accordance with these assumptions, the paper analyzes severaltypical instances of grammaticalization, and develops a formal account asfollows: when a functional head requires an overt realization in PF andthe lexicon cannot provide it with a corresponding lexical item, then thePF-interpretability of this functional head has to be realized by moving ormerging other lexical items. Compared with move, merge is much moreeconomical since it can establish a one-to-one matching between a lexicalitem and a functional feature so as to avoid feature syncretism (i.e. alexical feature spells out more than one features) and produce a simplersyntactic structure. Therefore if triggering experience, which supports themove of a lexical item, becomes ambiguous, the merge is alwayspreferred. When the lexical item merges with the functional head, it is reanalyzed as a new functional category. And this is also the basic processof grammaticalization.Taking a formal approach not only provides grammaticalizationresearch with a novel angle and method, but also offers the generativegrammar a new empirical support. Hence, it is a win-win combination,which is beneficial for both the development of the Formal Linguisticsand the further research on grammaticalization.