Framing Agency in the 2011 UK Riots. A Corpus-Based Discourse Analysis of British Newspapers

By the late 1970s, the nation’s first generation of British-born black people (especially of West Indian descent) had started claiming a larger stake in society, which deeply impacted on Britain’s public and political sphere. In addition to the social tensions deriving from increasing cultural and ethnic conflicts and this sort of post-imperial malaise, the 1970s and 1980s were decades of deep recession and widespread unemployment, which obviously affected the less prosperous African-Caribbean community in the first place. Therefore, the combination of poverty, powerlessness, oppressive policing tactics, discrimination and racism led to the riots that sparked in the 1980s, which had remarkably unsettling effects on the whole population, struggling with the fears and uncertainties arising from the proximity with diversity and post-colonial otherness. Accordingly, the Scarman report (that was commissioned by the then Home Secretary William Whitelaw with the aim to address the causes of the 1980s disturbances) identified racial discrimination and racial disadvantage as the root of the riots, concluding that urgent action was needed to prevent such issues from becoming an “endemic, ineradicable disease threatening the very survival of [the British] society” (Scarman 1981: 27). However, still in the 1990s, racist attacks continued to increase; ethnic minorities – especially African-Caribbeans – were persistently and invariably associated to crime, despite the fact that the London Metropolitan Police Service was found to be institutionally racist by the Macpherson Report (1999), a subsequent government inquiry into police conduct. The overview of past events and of a background knowledge on the previous riots in the UK was a necessary step because it provided important insights to understand the most recent disorders. Moving from these assumptions, the research project aims at exploring the extent to which issues of social, cultural, ethnic discrimination could still be said to play a role within the British society, after the violent disturbances that took place in August 2011. By drawing on a variety of sources and studies, this project’s purpose is analysing the ways in which the British press reported the riots, paying special attention to the portrayals of the subjects involved in the events, their linguistic construals, and the different emerging readings of the social unrest foregrounding or downplaying specific aspects, especially those relating to the motivations for the riots. The project starts from a theoretical overview (Chapter Two), presenting the existing literature on media and newspaper discourse. Indeed this field of investigation can prove very revealing as far as political, social and cultural meanings are concerned; it seemed worth exploring since, while shaping public opinions and beliefs, it sets the agenda giving relevance to certain topics and events within the country. News reports, in particular, seen as a practice intervening in the construction of reality, assess the significance of events, providing readers with the frames to make them comprehensible. Among the many ‘critical’ approaches to the study of media and newspaper discourse, ranging from Critical Linguistics (Fowler et al. 1979; Fowler 1991) to Cultural Studies (Hall et al. 1978; Hall et al. 1980), which explored and exposed the ideological significance of media texts (see Chapter Two) – not to mention structural discourse analysis (Bell 1984, 1991) and multimodality (Kress and van Leeuwen 1996, 2001) – Critical Discourse Analysis has certainly given a crucial contribution to the investigation of the role of discourse in the reproduction and challenging of the dominant socio-political order. Within this framework, the scientific research of two scholars in particular, Teun van Dijk and Norman Fairclough, appeared pivotal for the goals of this study. Although from different angles, the former from a socio-cognitive perspective (van Dijk 1988a, 1988b), the latter from a discourse-practice perspective (Fairclough 1995a, 1995b), both have underlined that the media tend to build ideologically-based versions of reality, aiming at persuading their audience that certain events are good or bad, thus determining specific attitudes and affecting the formation of public opinions. The project then moves to further non-linguistic analyses of the riots, taking into account the events from a sociological perspective (Chapter Three). The findings of studies carried out by the Runnymede Trust (an independent race equality think tank), highlighting that the events were too quickly dismissed by the media as sheer and opportunistic looting, together with the findings emerging from the London School of Economics investigation (in collaboration with The Guardian), uncovering a number of political reasons behind the rioters' (mis)deeds, do offer an interesting lens to frame the events. They reflect on a range of questions that appear socially and culturally relevant, but which were given a differing weight in the reporting of the newspapers. As a matter of fact, the riots seemed to represent a critical moment in the UK’s contemporary history, posing a big challenge in the light of the many and recurrent debates on multiculturalism, diversity, and the so-called convivial culture, namely “the processes of cohabitation and interaction that have made multiculture an ordinary feature of social life in Britain’s urban areas and in postcolonial cities elsewhere” (Gilroy 2004a: xv). Concerns over the British failure to explain its post-colonial conflicts and accommodate otherness in relation to a fundamental commonality are still widely present in ongoing discussions on how to envision new conceptions of identity and belonging. This is the reason why the debates on the riots have generally viewed the disturbances from the standpoint of culture and ethnicity. The other major perspective from which (especially) the most recent events were framed was that of consumerism, with rioters reacting to their lack of something that was considered as socially prescribed, but which they could not access. In this context, deprivation would have caused a deep humiliation from which a symbolic and material violence arose. In both cases, these studies provide remarkable insights into a deeper understanding of the riots. The project then proceeds to clarify the methodology chosen and the parameters adopted for the design and collection of the corpus (Chapter Four). While recognizing that most of the research in the field of media discourse was significantly carried out using a qualitative approach, the arbitrary selection of the texts and the very small size of the corpora to be analysed caused some criticism based on the fact that findings seemed less representative and less generalisable. As a consequence, another approach has gained popularity in recent times to investigate media discourse: corpus-based discourse analysis (Baker 2006; Baker et al. 2008; Gabrielatos and Baker 2008; Baker et al. 2013). Generally speaking, corpus-based approaches when doing CDA offer a number of advantages pertaining to the fact that larger amounts of data (based on large corpora) make the findings more credible and reliable. Moreover, the retrieval of keywords, collocates and concordances allows the uncovering of hidden discourses embedded in media texts. In fact, as Fairclough (1989: 54) stresses, “[a] single text on its own is quite insignificant: the effects of media power are cumulative, working through the repetition of particular ways of handling causality and agency, particular ways of positioning the reader, and so forth”. In other words, there is an ideological burden that goes unnoticed and works through unconscious and subtle repetitions, which is where corpus linguistics can be of help, highlighting patterns that could not be detected by manual inspection. For this research project, a corpus of about 1,700 articles (1,112,471 tokens) – including reports, features, editorials and op-eds – was collected from the six British newspapers with the highest circulation rates in August 2011: Daily Mail, Daily Mirror, The Sun, The Telegraph, The Times, The Guardian and their Sunday editions (Chapter Four gives full details on corpus design and collection). This specialised corpus gathers the articles published over a time span ranging from the beginning of August (the riots occurred between August 6th-10th) to the end of December 2011, thus covering the first five months soon after the events, which was regarded as the most salient period. After being refined and annotated, the corpus was ready for analysis, whose data and findings are extensively presented in Chapters Five and Six. More specifically, Chapter Five covers the different stages of investigation: the first, qualitative stage leading to the identification of the main participants and the most recurrent strategies through which they were defined, using van Leeuwen’s framework of social actors (van Leeuwen 1996, 2008); the second, quantitative stage resulting in a series of data concerning frequency information, which allowed the semantic categorisation of the several terms employed in connection to one subject in particular, the rioters. Further analysis of the concordances retrieved for each social actor in each newspaper then gives corpus evidence of the most recurring linguistic representations of Mark Duggan, the rioters and the police. Moving from such findings, Chapter Six focuses exclusively on the evaluative language that was used by the British press when reporting on the three participants under investigation. In fact, the protagonists to the riots can be deemed as important ‘sites’ of evaluation, where the newspapers’ stances and viewpoints appear encoded in the language they employ. Therefore, despite the fact that evaluation may be difficult to spot through corpus techniques – because it is subjective, value-laden and extended over the co-text in which the node words appear (Hunston 1994; Thompson and Hunston 2000) – evaluative statements are noteworthy since they express ideologies that are shared by writers (the newspapers) and readers. Hence, special attention is necessarily paid to the analysis of the nominal, adjectival and verbal collocates co-occurring with the lexical items referring to the participants, and then examined in the light of the basic evaluative parameters of good and bad, looking for their (more or less) positive or negative construals as conveyed by the British press. In the end, Chapter Seven draws the conclusions of the research, summarizing the main findings as emerged from the analysis, elucidating the specific contribution given by this project to the field of newspaper discourse in general, while also presenting the main potential developments for future research starting from this study and its corpus.