Applying Corpus Linguistics in Discourse Analysis

The present study employed corpus analytical tools to study and compare BP Company’s CSR reports before and after the 2010 oil spill in the Gulf of Mexico. Corpus linguistics served in the research as a methodological tool since it ensured a more qualitative research methodology. And Fairclough’s three-dimensional conception was employed to explore how an organization uses justificatory discursive strategies to legitimize its operations after perceived disasters.

Key words： Corpus linguistics； Discourse analysis； CSR reports

Corpus linguistics （CL） focuses on “the study of language based on examples of real life language use” （McEnery & Wilson， 1996）. There are two kinds of CL researches. The first kind is called descriptive corpus research， where the researchers work hard to find the linguistic patterns of a language. In the second kind of research， the researchers still work hard to find the linguistic patterns of a language， but in order to explain how the language people use in certain interactional context help constructing the reality they are in. This paper belongs to the second kind of research， applying CL as a methodological tool to understand the construction of discourses. Here we call CL a methodological tool because CL ensures a more qualitative research methodology with the help of large quantity of naturally occurring language data and various corpus software and statistics analytical tools. However， just as Biber （1998） suggests， corpus-based researches（applied corpus researches） have to depend on both the quantitative techniques and qualitative interpretation frameworks. The following are the examples of this kind of research. O’Halloran （2010） analyzed a corpus of British newspaper articles about immigrants， and the researcher employed the theory of Critical Discourse Analysis （CDA） as his interpretative framework. The other studies which are corpus-based and take CDA as the interpretative include Fairclough’s （1995） study of media discourse， Hajer’s （1997） study of environmental discourse， Chen & Lam’s （2012） study on Western perceptions on Hong Kong a decade after the reversion of the sovereignty from Britain to China in 1997， Bhatia’s（1997） study on public discourse in Hong Kong， Fang’s（2001） study on Chinese print news media discourse， Flowerdew’s （2004） study on globalization discourse， etc. The above-mentioned researches conducted quantitative analysis with the help of large general corpora and corpus processes （keyword， frequency， dispersion， concordance）. At the same time， the analytical framework of critical discourse analysis was employed to interpret how language， whose linguistic patterns have been found in CL analysis， served as a form of social practice. In the next part， I will take my research on CSR reports as an example to show how to conduct a corpus-based critical discourse analysis.

The present study is about how to conduct a discourse analysis by using corpus analytical techniques， among which， two techniques， keyword lists and concordance lines are employed. Keyword list derived from the comparison of the two little texts help us generalize significant lexical differences before and after the environmental disaster. By sorting the concordance lines of those keywords， we are exposed to the contexts where the keywords occur. But the patterns we found from the contexts are still subject to the researchers’ interpretation. The quantitative results derived from the corpus analytical software help counter the bias of researchers’subjective interpretation. All the keywords that occur more frequently in the 2011 CSR report are employed to help the BP Company to legitimate its managerial and operational behaviors. Adopting Fairclough’s （1992） three-dimensional conception， discourse is a text， a discursive practice and a social practice. In this sense， a CSR report is itself a text that can be analyzed in a linguistic method. Researcher can talk about its linguistic patterns， semantic prosodies， and rhetorical styles. A CSR report is also a discursive practice whereby a company discloses the social， environmental， legal impacts of their operational practices. At the same time， a CSR， especially when released after an environmental accident， is a social practice whereby a company addresses the concerns of potential readers and defuses the criticisms. In face of a major environmental disaster， BP deploys interpersonal resources to portray itself as an indispensable provider of energy for the whole world， a survivor from the disaster， and a student who is good at learning from a lesson.

REFERENCES

Baker， B. （2006）. Using corpora in discourse analysis. Continuum： London.

Bhatia， V. K. （1997）. Democratizing government decisionmaking： A study of public discourse in Hong Kong. Journal of Pragmatics， 28， 515–532.

Biber， D.， Conrad， S.， & Reppen， R. （1998）. Corpus linguistics： investigating language structure and use. Cambridge： Cambridge University Press.

Chen， W.， & Lam， P. （2012）. Western perceptions of Hong Kong Ten years on： a corpus-driven critical discourse study. Applied Linguistics， 33（5）， 1-19.

Fang， Y. J. （2001）. Reporting the same events？ A critical analysis of Chinese print news media texts. Discourse and Society， 12（5）， 585–613.

Fairclough， N. （1992）. Discourse and social change. Cambridge： Polity Press.

Fairclough， N. （1995）. Media discourse. London： Hodder Arnold.

Flowerdew， J. （2004）. Identity politics and Hong Kong’s return to Chinese sovereignty： Analysing the discourse of Hong Kong’s first chief executive. Journal of Pragmatics， 36（9）， 1551–1578.

Hajer， M. （1997）. The politics of environmental discourse： ecological modernization and the policy process. Oxford： Oxford University Press.

McEnery， A.， & Wilson， A. （1996）. Corpus linguistics. Edinburgh： Edinburgh University Press.

O’Halloran， K. （2008）. Fleeing， sneaking， flooding： a corpus analysis of discursive constructions of refugees and asylum seekers in the uk press. Journal of English Linguistics， 36（1）， 5-38.

Scott， M. （2008）. Oxford wordsmith tools 5.0 Manual. Oxford： Oxford University Press.