Formal Features Of Verb

It is generally assumed that formal features of functional categories are subject to parametric variations and are the cause of the displacement property of human language. However, some problems are difficult to solve under this assumption, such as in Pseudogapping constructions and multiple wh-movement. Moreover, in MI feature checking is reduced to deletion under the operation Agree and Attact is dispensed Therefore, movement can no longer be regarded to be triggered by formal features of functional categories, which 'attract' a lexical category to it Along these lines, another cause of movement should be found in order to explain the displacement property of human language.In this thesis, we propose mat the cause of movement are formal features of lexical categories. Thus, Subject Raising in subject-initial languages is forced by the strong formal feature of the subject itself while Object Raising in SOV languages is triggered by the strong feature of the object Similarly, Verb Raising is triggered by its strong formal features, which are taken as the focus of this thesis. Since Verb has several types of features, we first make a classification of its features. After a review of the studies of features and parametric variation in the literature in chapter 1, we propose in chapter 2 mat Verb has five types of features: phonological features, semantic features, categorial feature, affix features, and formal features [±v], [±T], [±C]. The former four types of features are represented in the lexicon and are interpretable, while formal features are uninterpretable and must be checked immediately after they enter the computation, if they are strong, i.e. with positive value. These uninterpretable features are crucially the cause of Verb MovementAccording to the Head Movement Constraint (HMC), there are only four possible combinations of these verbal features: [+v] [+T] [+C], [+v] [+T] [-C], [+v] [-T] [-C], [-v] [-T] [-C]. As linguistic variations to a large extent are subject to Verb Movement, human languages can also be classified into these four types. In chapter 3, we exemplify mat German and Romanian belong to the first type, French and Welsh the second one, English and Chinese the third one,SOV languages like Japanese and Tuddsh belong to Ihe fourth type. Since our issue in this thesis is closely related to feature checking, in chapter 4 there is first an introduction of Chomsky's assumptions on feature checking and then an exposition of feature checking, including the fonnal features of Verb, in the languages discussed in chapter 3.Our suggestions of word order variations in those languages are mainly based on the testing of adverb positions and wh-question orders, while our analyses of feature checking processes are based on MI and the assumption proposed in this study.