A Study Of The Constraints On The Dative Shift

In a transformational approach to dative shift, dative verbs have a single meaning, giving rise to two derivationally-related syntactic structures: the dative flame of NP₁ V NP₂ to/for NP₃ and the double object flame of NP₁ VNP₂ NP₃. The derivational relation between the two frames has, in linguistic theory, commonly referred to as dative shift or dative alternation. Facts show that some verbs will allow for both frames, but some verbs may be restricted to a realization with only one of the frames. This suggests that there are constraints or restrictions on the choice of the dative variant or the double object variant.In the study of the constraints on dative shift, researchers have come up with a number of different proposals. According to the color-blind conservatism proposed by Pinker (1989), Gropen et al (1989) and Jackendoff (1990), it is assumed that the grammar is only relevant to a certain specific elements of meaning, i.e. that most of a verb's conceptual representation is linguistically irrelevant, is proposed. In Pinker (1989) and Gropen et al (1989), verbs are classified into the dativizable and the non-dativizable in accordance with the broad-range rule and narrow-range rule. The broad-range rule specifies that a verb with the thematic core 'to cause X to go to Y' can alternate with the thematic core 'to cause Y to have X'. Narrow-range rules then categorize verbs into those alternate and those that do not. In fact, the narrow-range rule can't exhaustively accommodate all the verbs that do and do not undergo dative alternation. Therefore, Pinker (1989) has to introduce morphophonological rules into the system as a remedy, because his model is color-blind. This constraint states that native-stem verbs do allow the double object construction, whereas Latinate class verbs do not. As a matter of fact, not all the subclasses of verbs are sensitive to the constraint. In addition, Jackendoff (1990) holds that dative shift is restricted by Thematic Hierarchy. He proposes that the two syntactic forms for dative verbs mapping onto two conceptual representations which only differ in that the conceptual representation of the double object frame marks the first object as Beneficiary. However, this conclusion is not well grounded empirically. Following Pinker's proposal and according to the idiosyncratic information of dative verbs, Groefsema (2001) proposes the Unique Effect Constraint which says that whether a verb can occur with the specific construction is dependent on the specific ways in which the affected entity is affected in the events expressed by individual verbs. To illustrate his view, he classifies dative verbs into three types: the verbs used in only one of these two constructions; the verbs used in both of these constructions. Then he gives a detailed analysis of each kind of verbs in terms of their conceptual representations.In a sense, all these accounts are semantic in nature. It can be seen that these views have certain explanatory power, but fail to fully capture the essence and the nature of dative shift. A closer look at the data in alternations reveals the influence of pragmatic factors.In this study, after a critical review of the previous analyses, I propose that a full and principled account of dative shift should take into consideration. Form my point of view, the information structure will restrict dative shift pragmatically. Under this analysis presented here, some of the dative verbs traditionally considered to occur in both syntactical forms practically appear only in a certain form, and this largely depends on the arrangement of information in the syntactic form. To be specific, in the organization of information, the principles of end focus and end weight have a role to play. These principles work together and contribute to the presentation of the content of a clause in one particular order rather than another. The principle of end focus requires that the newest and the most important information should be placed at the end of the clause. So, in dative alternation, if the direct object is a pronoun, the dative construction is supposed to be adopted, as a pronoun refers to something that is mentioned before. The principle of end weight tells us that the weightier parts should be placed towards the end of the clause. Thus, in dative altemation, if the direct object is longer and more complicated than the indirect object, we'd better choose the double object construction. Both end-focus and end-weight principle restrict the choice of dative construction jointly. Besides, dative shift depends on the context of utterance and the implications the speaker wants to transfer. In addition, some dative verbs have preference for one variant.What I put forward in my proposal aims at a more adequate explanation of the dative shift so that a better understanding of its essence can be achieved.