A Cognitive Approach To Hypallage Production

Figure of speech plays an important role in literary works as well as our daily life. Hypallage, as one figure of speech, is also frequently used. Thus it attracts attention from linguists, psychologists and estheticians. Further research into hypallage can help us have a better understanding of language and human thoughts.Previous researches on hypallage are mainly concerned with the definition, classification, structure, meaning, mental basis, aesthetic function of hypallage and its comparison with other rhetoric devices. Those researches have contributed to people's better understanding of hypallage, and lay a solid foundation for the current study. With reference to former studies, the classification of hypallage into synaesthesia, empathy and transferred epithet is adopted and the cognitive approach to hypallage is followed in this thesis. Compared with developments of researches on other aspects of hypallage, the explanation of hypallage generation leaves too much space for exploration. Wang Lirong (2005, pp. 239-248) has made a trial to apply Conceptual Integration Theory (CIT) to hypallage, and demonstrates that CIT is powerful in explaining hypallage. His study is meaningful, but he doesn't make a critical review of CIT itself. Thus he doesn't solve the problem why some elements in the input spaces are projected into the blend while others are not.This thesis manages to build up a new analytical framework on the basis of CIT and Autonomy-dependency Analysis Framework to reveal the generation mechanism of hypallage. CIT provides a good approach for us to understand the meaning construction of hypallage, but it has its own problem explaining the generation mechanism of hypallage. The problem is that it doesn't make clear the determining factor deciding the choosing of one element over others in the input spaces to be projected into the blend. This problem can be solved by resorting to Intentionality and Proximity/Similarity Relations in Autonomy-dependency Analysis Framework. Autonomy-dependency Analysis Framework is convincing in explaining utterance generation for three reasons: a) It interprets the process of utterance generation to be an alignment, autonomy vs. dependency; b) It stresses the important role of intentionality in utterance generation; c) It shows clearly the approach through which the source of language materials can be accessed to generate utterances. But this theoretical framework has problem of its own: it doesn't give good explanation to meaning construction of the utterance generated. So Autonomy-dependency Analysis Framework can be integrated into CIT to form a unified analytical framework (CIT+) to explain the generation mechanism of hypallage. According to this framework, hypallage generation can be described as: in the process of hypallage generation, the input spaces and the generic space are autonomous, while the blend, with its emergent structure, is dependent; the cross-space mapping between the input spaces is projection mapping or schematic mapping; elements in the input spaces assemble with proximity/similarity Relations; guided by the intentionality of the language user, the elements in the input spaces are intentionally selected and projected into the blend; then, through conceptual blending, the emergent structure, in the form of hypallage, arises from the blend. In this thesis, the framework CIT+ is built up and applied to the explanation of hypallage generation. With satisfactory explanation of the generative process of three examples of hypallage, the practicability of CIT+ is proven.