| Caused-motion construction is a kind of common construction type, and it can be explained as: a causer causes a causee to move along a certain path with some force; the typical structure of caused-motion construction can be presented as [SUBJ [V OBJ OBL]] or [NP1 V NP2 PP]. The caused-motion structure has become a major target for the study of construction grammar and enjoyed long-lasting attention since Goldberg(1995) proposed argument structure from a constructional perspective.Just as what is faced by argument structure construction, caused-motion construction is also confronted with a controversy between the lexical approach and the constructional approach. The lexical approach(e.g. Levin & Hovav, 1991, 1996; Kay,2002, 2005) holds a lexical-centered view, regarding lexical meaning especially verb meaning as the decisive factor for the meaning of an expression. But it sometimes over-emphasizes the polysemy of lexical item which might sacrifice the economy of language; what’s more, it fails to see the influence from construction on semantics in a certain context. The constructional approach(e.g. Goldberg, 1995, 2006; Langacker,2005), on the contrary, opposes this lexical-centered view and claims that the ultimate decision lies on construction. However, sometimes it ignores the influence from the original lexical meaning, and the over-use of constructions might lead to the controversy about its own classification. In order to deal with the problems above,scholars like Iwata(2005), Lin. Y.(2009), Cheng Q. L. & Cheng Q. W.(2014), et al.suggest to take a compromised lexical-constructional approach that combines both of the former two approaches. It efficiently avoids the defects of the two approaches and thus gains popularity. But it still needs empirical evidence to prove the concrete interactive relations between the verb and the target caused-motion construction.Therefore, this thesis aims to analyze a representative structure of caused-motion,[NP1 V NP2 PPinto], by taking the usage-based construction grammar as the theoretical evidence, the lexical-constructional approach as the framework, and the collostructional analysis as the methodology. The target is to further explore the verb-construction relationships and the driving forces in caused-motion construction, to enrich the existing theories, and especially to explore the impact on the practical language usageby checking both the constructional meaning and the lexical meaning.Some important questions for this research are listed below:1) Which verbs show strong collostruction strengths with the target construction?2) What are the characteristics of the verbs entering the target construction?3) How to explain the interaction between verb meaning and construction meaning and their influence on caused-motion construction?In order to answer these questions, this thesis takes the online corpus COCA as the source. Firstly, the caused-motion construction in the form of [NP1 V NP2 PPinto] is chosen automatically. Then the unqualified data are taken away through a manual picking-off procedure. Next, the qualified data are calculated by R software with the function of Fisher-Yates Exact test to get the association strength/collostruction strength of the verb in a given slot and the target construction.The major conclusions in this study are:(1) There exist some interactive relations between the collexemes and the target construction. The raw frequency of caused-motion construction [NP1 V NP2 PPinto] is78058, with 523 verb collexemes occurring in the given slots. The verb collexeme with the highest collostruction strength is inject. The ones following it are plunge, insert,integrate, our, get, make, take, etc..(2) The top-50 verbs with the highest collostruction strengths in the form of [NP1V NP2 PPinto] are observed based on the classification by Luo(2015) as well as the influences like coercion. The classification method for caused-motion verbs is comparatively thorough and comprehensive for it considers both lexical meaning and constructional influence. According to the research result, 9 verbs come from the put-group, including inject, pour, put, dip, pump, set, stuff, dump and load; 7 from the deliver-group, including take, bring, push, pass, pick, drag and slip; 5 from the ownership-shift-group, including give, show, leave, pay and sell; 2 from the throw-group, including throw and shove; 1 from the detach-group, that is stop.Moreover, there are other 26 verbs. The distributions of these verbs are illustrated by citing the corresponding cases from COCA. The results show that the verbs originally without caused-motion sense occupy quite a large part. Therefore, it is not enough to merely pay attention to lexical meaning; rather, the comprehensive perspective from a constructional level shall not be ignored. Construction coerces verb until the latter one has the semantic compatibility with the former one, which can be regarded as a major force to form the constructional meaning for caused-motion.(3) There exists some interaction between lexical meaning and construction meaning. From one part, people’s stored experience is triggered to form the corresponding expression. Usually, the original lexical meaning derives from cognition of speakers in an accumulating way which lays some basis for the interpretation of the caused-motion event. Lexical items can also renew the original meaning, or experience the meaning shift as people’s knowledge accumulates. From another part, construction shows great capability in explaining uncommon language phenomena. By bringing some coercive force, the mismatched components can experience an adapting course until the meaning is reasonable. After this meaning becomes conventionalized, the status of its usage can be acknowledged and thus a novel lexical meaning forms. To summarize it, the relationship is proven to be complimenting rather than separating.Obviously, the factors affecting the formation of construction show a great diversity, e.g. the raw frequencies of the verbs in the target construction, its input frequencies in the daily use, the coercion in some specific context, etc.. Therefore, it can be concluded that caused-motion construction becomes increasingly entrenched in the cognitive system along with the increasing frequency as well as the input in the daily language. It can even influence many linguistic aspects such as the rhetorical skills. In conclusion, this case study can provide much assistance for other researches, such as extending the research coverage about construction and developing the relevant methodology. |
PDF Full Text Request