The Language of Film: Corpora and Statistics in the Search for Authenticity. Notting Hill (1998) - A Case Study

Abstract

While it is well known that film scripts generally fall short of capturing the varied and subtle characteristics of spontaneous dialogue, it is less clear exactly how and to what extent such language fails to ‘ring true’. Extensive investigation into the components of both spontaneous talk and film discourse over a wide range of film and television material have proved a solid basis on which to extend research in this field. With the aid of corpus linguistics and the assistance of statistics experts, some interesting phenomena have been uncovered relating directly and indirectly to the original aims of the research.
For example, comparisons of British and American film and TV scripts with spoken language corpora such as those within the Bank of English project and the San Diego spoken language corpus, have already provided material to prove the hypotheses about the lack of authenticity in film material. But observing the changes that take place between an original script and the transcription of the final version of a film text is more illuminating, as are the statistical analyses that show how particular language features occur in clusters and bundles (Biber et al 2004), and how combinations of word and word groups seem primed (Hoey 2004) for different film genres. This paper therefore sets out to illustrate the methodology employed in this particular study of film language, explain the results obtained and present the serendipitous findings that the research threw up, with particular reference to the film Notting Hill.