Sketches of CULTURE from the web. A Preliminary Study

This paper reports on a work in progress whose aim is to explore how some tools, methods and data sets recently developed within corpus linguistics can be used to obtain not only information concerning language use but also fresh insights into culture. The background for the research can be found in a growing body of work in which the corpus linguistics approach has been profitably used to attain a deeper understanding of discourse and society, starting from Stubbs’s (2001) study of such words as “heritage”, “racial”, ”tribal”, to more recent research by Mahlberg (2007) and Pearce (2008), to mention a few contributions in this field. It is also important to mention the increasing interest in studies which integrate Discourse Analysis and Corpus Linguistics (e.g. Baker 2006). However, while supporting Corpus Linguistics as a research domain that is increasingly meeting the challenge of tackling issues relating to discourse and society, this paper also aims at exploring the specific potential and limitations – in this respect - of tools and resources that have been developed within Corpus Linguistics under the impact of the web. In order to achieve its aim the paper investigates the behaviour of culture, a notoriously “difficult” word (Williams 1983, 87), using the Sketch Engine (Kilgarriff et al. 2004), a corpus query tool which is able to extract and sum up linguistic information from large corpora. The “sketch” obtained for culture from ukWaC, a 1.5 billion word web corpus of English, is briefly analysed and discussed, and attention is focused on areas for further investigation.