A Corpus-based Cross-cultural Study On Metaphor

Corruption has long been a social problem that the whole world has to deal with and has received much attention on media,especially in China for the past three years since President Xi started the anti-corruption campaign in 2013.It is interesting to find that metaphors are widely used in corruption media discourse,for corruption is an abstract social phenomenon that needs more concrete concepts to understand.Corpus-based methods to researching metaphor use in media discourse have become increasingly prevalent,but metaphor use in corruption discourse has received little attention.Besides,previous comparative metaphor studies are basically cross-language studies.With the development of globalization and the promotion of China's international status,China has started to pay more attention to its national image in the eyes of westerners.The use of English as lingua franca(ELF)for China to show the world what is happening at home and express its view on international affairs has become increasingly important.Therefore,within the framework of Conceptual Metaphor Theory(CMT),this paper sets out to do a corpus-based cross-cultural study on metaphor use in Chinese and American corruption media discourse,which are both written in English.The thesis is based on two self-built corpora: China Daily corpus(around 58,965 words)and CNN corpus(approximately 46,611 words),both made up of 90 media reports about corruption.The corpus-tool Wmatrix is chosen as the research instrument,because its embedded semantic annotation system(USAS)can provide for every word in the corpus a semantic tagset,which crudely conform to the source and target domain of conceptual metaphor.By conducting a thorough quantitative and qualitative analysis on the linguistic data with the help of Wmatrix,the study manages to find the source domains in the two corpora,the shared conceptual metaphors,the significant differences,as well as the underlying cultural reasons.36 source domains are retrieved in the China Daily corpus and 29 in CNN.Out of the 28 same source domains,those conceptually-linked ones are grouped together,constructing eight dominant conceptual metaphors.These recurring metaphors that exist in both corpora are(ANTI-)CORRUPTION IS A WAR,CORRUPTION IS A DISEASE,ANTI-CORRUPTION IS HUNTING ANIMALS,(ANTI-)CORRUPTION IS PLANT,(ANTI-)CORRUPTION IS JOURNEY,(ANTI-)CORRUPTION IS AN EXPANDING PHENOMENON,(ANTI-)CORRUPTION IS A DIFFICULT THING,(ANTI-)CORRUPTION IS COOPERATION.Based on the two lists of key domains,significantly overused semantic domains in one corpus than in the other(LL?6.63 for P<0.01),quantitative and qualitative analysis are conducted to ensure that the differences the study finds are significant ones.It shows that in S7.4+ [Allowed] and L2 [Living creatures: animals,birds,etc.] that roughly correspond to the source domain of conceptual metaphor(ANTI-)CORRUPTION IS LIVING ORGANISM,and in S8-[Hindering],A1.7+ [Constraints],M3 [Vehicles and transport on land] that roughly correspond to conceptual metaphors(ANTI-)CORRUPTION IS WAR and(ANTI-)CORRUPTION IS JOURNEY,China Daily corpus has a bigger number of and a higher variety of metaphorical lemmas than CNN corpus.Also in terms of metaphor conventionality,China Daily uses more novel metaphors than CNN.Moreover,the significant differences of corruption-related metaphor use in Chinese and US media discourse can be attributed to cultural factors.It is the unique characteristics of Chinese culture,including rule of man,large-country mindset and collectivism that causes the differences in metaphor use.