Introduction

This chapter sets out the theoretical and methodological framework of the book and briefly describes the data on which it is based. The latter consist of four XML TEI conformant corpora of TV news reporting on the 2003 Iraq war on public and private television in the UK (BBC, 103,806 words), the US (CBS, 59,045 words), and Italy (Raiuno, 89,808 words, and TG5, 108,330). The data were collected from the first day of the war, March 20 to April 17, a week after the so-called ‘fall of Baghdad’. The broadcasts were transcribed and tagged for speaker role (e.g. news presenter, corrispondent, embedded reporter) and speech event (e.g. news presenter’s introduction to a report, news presenter/correspondent live exchange, pre-recorded report with the reporter in voiceover). The discourse represented by the various speaker roles and speech events is quantified for comparative purposes to identify important differences in the ways in which correspondence is constructed and the function of the news presenter is conceived on the different channels in the different socio-cultural contexts. The linguistic analysis of the data is carried out using a Corpus Assisted Discourse Studies (CADS) approach. The quantitative analysis to identify lexical and grammatical patterns in the various subcorpora is performed with the variety of tools available with Wordsmith and Xaira software. The qualitative analysis is conducted with reference to two principal and complementary theoretical frameworks: 1) Iedema, Feez and White’s (1994) objective/subjective dichotomy and its associated set of Authorial Voices, with positioning on the continuum determined by choices made in relation to a set of mutually exclusive linguistic options; 2) Martin and White’s (2005) appraisal theory, in particular with respect to the construction of stance in text (Hunston and Thompson 2000), focusing i.e. on whether key propositions are formulated monoglossically (as ‘bare assertions’) or heteroglossically (with reference to an external source); if monoglossic, whether propostions are presented as ‘taken for granted’ or ‘at issue’; if diaglossic, whether the speaker’s stance is contractive (excluding or constraining alternatives) or expansive (actively allowing for alternative positions). The chapter concludes with a discussion of the realisation of cultural values and identity through the use of linguistic indicators of stance and evaluation, both with respect to the content of propostions and to the construction of a network/audience relationship. Reference is made in this regard to work in media studies on models of journalism as a response to specific socio-political and economic systems (Chalaby 1996; Hallin and Mancini 2004; Hallin 2005).