critical discourse analysis, London, population, corpus, Discourse Studies, Discourse & Society, Home Office, corpus linguistics, HMO, the modern state, populations, quantitative analysis, government, violent extremism, governmentality, refugees and asylum seekers, UK, Michel Foucault, UK government, New Labour, Edinburgh University Press, Oxford University Computing Services, Critical Studies, Edinburgh, nation state, violent extremists, EXTREMISM, corpus-based approach, singular noun, Countering Terrorism, Irish terrorism, corpus studies, discursive constitution, discursive practice, qualitative analysis, national security policy, UK security, Counter-terrorism, electronic documents, Relevant documents, UK New Labour Government, N. Fairclough, British National Corpus, community, government websites, Foucault, George W. Bush
Security, Population and Governmentality: UK Counter-terrorism Discourse (2007-2011) Copyright © 2013 Critical Approaches to Discourse Analysis across Disciplines http://cadaad.net/journal Vol. 7 (1): 123 140 ISSN: 1752-3079 MALCOLM N. MACDONALD University of Warwick [email protected]
DUNCAN HUNTER University College
Plymouth St Mark & St John [email protected]
Abstract Over the past decade, governments worldwide have taken initiatives both at a national and supra-national level in order to prevent terrorist attacks from militant groups. This paper analyses a corpus of policy documents which sets out the policy for UK national security. Informed by Foucault's (2007) theory of governmentality, as well as critical discourse analysis
and corpus linguistics, this paper analyses the ways in which the liberal state in late modernity realizes security as discursive practice. A corpus of 110 documents produced by the UK government relating to security in the wake of the 7/7 attacks between 2007 and 2011 was assembled. The paper analyses the discursive constitution of the Foucaultian themes of regulation, knowledge and population, though carrying out a qualitative analysis of relevant key wards, patterns of collocation, as well as features of connotation and semantic prosody. Keywords: CDA, corpus, counter-terrorism, governmentality, security 1. Introduction Since the end of the eighteenth century, security has emerged across Europe as the `essential technical instrument' of governmentality; and it persists as the principal apparatus through which power is exercised upon populations by modern governments (Foucault 2007). Islamicist attacks which took place between 2001 and 2007 on the World Trade Centre
(`9/11'), the Madrid Cercanнas (`11-M'), the London Transport network (`7/7'), and Glasgow Airport, have led to the incremental production of a discourse of security and counter-terrorism in Europe and the USA. This paper analyses a corpus of documents produced by UK government departments between 2007 and 2011, which set out national security policy. It will investigate how UK security is realized as discursive practice, as an exemplar of counter-terrorism discourse within the liberal state in late modernity. In so doing, it will analyze
124 | P a g e
the discursive and linguistic features
of counter-terrorism documents in order to uncover the principles which underwrite the constitution of security in the UK. It will therefore address: first, what tactics of governmentality are realized by the discourse of UK security between 2007 and 2011; and secondly, what patterns of language are used in UK security documents between 2007 and 2011?
2. literature review
For modernist historians, political theorists and legislators, the state is a type of government which arrived at its present form in the eighteenth century and persists worldwide as a form of objective reality up to the present day. In his final lectures (2007), Michel Foucault
problematised this essentialist description of the state. While he certainly agreed that a rupture began to emerge within sixteenth century Europe between the absolutist medieval rule of sovereign justice and the modern `administrative' state, he differs about what is at core the ontology of the modern state. On his argument, the role of the state is conventionally either exaggerated as some preeminent `cold monster' which confronts us, or is reduced to an account of reproductive forces and relations of production (2007). By contrast, the term `governmentality' is adopted in order to avoid describing the state as an essence, having a `unity, individuality ... [and] rigorous functionality' (ibid: 109). At its most conventional, governmentality is concerned with the `mentalities of government' (Miller and Rose 1990; Rose and Miller, 1992), or `different rationalities' of government (Dean 2010). On this argument, the way we think about government is `explicit and embedded in language and other technical instruments...and relatively taken for granted...by its practitioners'. At its most rational, the authority of government is located in the shared ideas, theories and knowledge which are derived in modernity from economics and political science. Less rationally, these `mentalities' of government might also be drawn from the more emotive discourses of politics and the political media (Dean 2010: 25). However, governmentality is described more specifically by Foucault (2007) as being the means of exercising of power upon populations which is informed by a principal form of knowledge, political economy. The ostensible goal of government is to ensure the wellbeing of the population, expressed in its longevity, health and wealth. The techniques which exercise this power over populations are lodged in the `apparatuses of security'. From Foucault's historical purview, security is maintained through two arms: a militarytechnological wing whose function up to 1945 was to maintain a balanced distribution of territory within Europe, and a police force whose function within the modern state is to maintain order within the population. Thus Foucault understands governmentality to be: ... the ensemble formed by institutions, procedures, analyses and reflections, calculations and tactics that allow the exercise of this very specific, albeit very complex, power that has the population as its target, political economy as its major form of knowledge, and apparatuses of security as its essential technical instrument. (2007: 107-8)
MacDonald & Hunter
P a g e | 125
While the relations between disciplinary power and knowledge (Foucault 1977, 1980) have for some time informed critical discourse analysis (e.g. Chouliaraki and Fairclough 1999; Pennycook 2001), for Foucault discipline in fact only constitutes one of three aspects of power - sovereignty, discipline and security. Sovereignty was the dominant form of power in feudal times through which the monarch exercised power over the subjects within his territory. This power was exercised by the judiciary and the executive through law and constitution. Disciplinary power was exercised upon individuals (`singularities'), within clearly delineated spaces such as the clinic, the barracks, the school or the prison. Thus discipline is focused on restrictive spaces in order to regulate specific aspects of the individual, and especially the body (Foucault 1973, 1977). By contrast, security is exercised upon entire populations (`multiplicities') within wider ranging territorial spaces such as the nation state and `milieus' within them, such as the town. Thus, security is expansive and laisser-faire, and its essential function is to regulate the components of `effective reality' and the relations between them (Foucault 2007: 47). The setting in place of...mechanisms or modes of state intervention whose function is to assure the security of those natural phenomena, economic processes and the intrinsic processes of population... is what becomes the basic objective of governmental rationality itself. (Burchell, Gordon and Miller 1991: 3) While security is distinct from the relations of sovereign and disciplinary power, it does not supercede them. Thus the conceptualization of governmentality combines these three elements, but it also `departs from them and seeks to reinscribe and recode them' (Dean 2010: 29). Most recently, in the wake of the 9/11 and 7/7 attacks the constitution of governmental rationality through the discourse of security has expanded exponentially in the UK, not least through successive iterations of the counterterrorism policies set out in CONTEST (Home Office 2006, 2009a, 2011a) and Prevent (Home Office 2003, 2009b, 2009c, 2011b), and the associated documents which are analysed below. Various combinations of critical discourse analysis (CDA) and corpus linguistics (CL) have been used to analyse texts in the public sphere which relate to the discourse of modern government. However, despite the recent publication of Foucault's final lectures in English (2007, 2008), the discourses of the modern state not yet been widely investigated by our field. Fairclough describes how the New Labour administration between 1997 and 1999 had already adopted a strategy of `governing by shaping and changing the cultures of public services, claimants and the socially excluded, and the general population'(2000: 61). This `cultural governance' operated not least through the crucial role that language plays in the ever-increasing mediatisation of politics and government in advanced capitalist societies (ibid: 4). More recently (2003, 2011a, 2011b) in relation to the discourse of the same administration's education policy
, Mulderigg has revealed a `type of verbal process which discursively enacts[s] a more subtle or `soft' coercive force in contemporary discourse' (2011a: 63). Other studies have analysed US presidential speeches, legislative hearings, congressional sessions, committee reports, and bills (Gales 2009); books on Wahhabi Islam and Wahhamism
126 | P a g e
(Salama 2010); and a series of `briefs' in French and English submitted to a commission assessing the prospects of Quebecois independence (Freake, Gentil and Sheyholislami 2011). Without adhering strictly to any singular CDA framework (e.g. Fairclough 2003; Wodak, De Cillia, Reisigl and Liebhart 1999; Van Djik 2001), these studies use CL methods to engage in the critical analysis
of documentary evidence. Recently, a qualitative study has also considered how the members of different populations were `labelled' in four counter-terrorism documents produced by the UK government between 2005 and 2007 (Appleby 2010: 427-430). Here, the labels `extremist' and `terrorist' appeared to be linked with the idea of Islam, and polarized against categories of Britishness. Paradoxically, those labelled as `extremist' were constituted as living outside the boundaries of the nation state. The documents also created a homogenizing label to constitute `the Muslim community' as a separate social group within British society (Appleby 2010: 427-430). Research more specifically into security discourse has been carried out from a more CDA perspective. The US-led invasion of Iraq triggered a plethora of studies which focused on the verbal rhetoric and policy documentation of the Bush administration in the run-up to the war, e.g. the negative representation of Saddam Hussein (Bhatia 2009; Chang and Mehan 2008; Meadows 2007); the discursive polarisation of us (the `West' and/or the `American people') and them (the `terrorists' and/or `Iraqis') (Bhatia 2009; Graham, Keenan and Dowd 2004; Johnson 2002; Meadows 2007). Critiques of the representations of the UK and US legal responses to the WTC, Madrid and London attacks in the public sphere have also been carried out (e.g. Preston 2009), with particular reference to: the discursive construction of the extraordinary categories of `terrorism' and `terrorists' in both Patriot Acts (De Beaugrande 2004); the US Patriot Act revoking of many of the civil liberties which were simultaneously being asserted in the White House rhetoric (Graham et al. 2004); the syllogistic logic of the discourse of the US Justice Department's website promoting the Patriot Act (Simone 2009); the greatly increased government surveillance of the US population (Simone 2009); and in the UK the extension of the minimum period detention of terrorist suspects in the without charge (Gillborn 2006: 8186; Preston 2009). However, to date no discourse studies have systematically applied Foucault's (2007) theory of governmentality to a corpus of texts, neither have they used corpus analysis
techniques to investigate the discourse of a sizeable corpus of documents concerned with national security.
3. Methodology This paper combines techniques of critical discourse analysis (CDA) and corpus linguistics (CL) to analyse a collection of texts in the public sphere (cf. Baker 2010; Baker and McEnery 2005; Baker et al. 2008; Gabrielatos and Baker 2008). A corpus of documents produced by the UK government relating to security and counter-terrorism was assembled between 1 January 2007 and 31 December 2011. This embraces the later years of the UK New Labour Government (2006-2009), including the premiership of Gordon Brown, in which a policy of `deradicalisation' in schools, FE colleges and universities was implemented in the wake of 7/7; and the early years of the Conservative-
MacDonald & Hunter
P a g e | 127
Liberal Democrat coalition (2010-11), most recently featuring the third revision of the CONTEST and Prevent strategies in June 2011. Preliminary readings suggested that the policy impact from the 7/7 attack took at least one year to filter through to official documentation. Our corpus comprises 110 documents (c. 2 million words). Documents were downloaded from five government websites: the Cabinet Office, the Home Office, the Department of Education, the Department for Communities and Local Government (DCLG), and the Department for Innovation, Universities, and Skills. Websites were searched for documents produced between January 2007 and December 2011 in Adobe Acrobat PDF format amenable for analysis, using the terms `security', `terrorism', `citizenship' and `radicalisation/radicalization'. Since it was not viable to carry out statistical sampling procedures for documents distributed over different websites, the corpus was assembled opportunistically. Relevant documents were selected from a direct trawl of the government websites above, and then augmented by using links from the iCoCo website (iCoCo 2011). Relevance was determined by prominence of key terms in the title and by a preliminary reading of electronic documents for the frequency and salience of the search terms. While several of the corpus studies above and elsewhere have focused principally upon the quantitative measurement of statistically significant keywords, this study employs an innovative, mixed-methods approach, which prioritised quantitative analysis (QAN) and qualitative analysis (QAL) differently in each of two phases. The first phase principally used qualitative analysis (qan+QAL) in which a combination of interpretive reading and keykeyword analysis was carried out to identify a core sample of six texts in which corpus themes were most densely concentrated. First, the documents in the corpus were converted from their original, varied formats to a uniform text format amenable to machine analysis, and the entire corpus was machinesearched for salient lexical items using Wordsmith Tools (Scott, 2008). Words were then identified as `key', where their difference in ranking was statistically significant when compared with the British National Corpus (BNC) - or `reference corpus' - according to the log-likelihood (LL) algorithm (Scott, 2008). Those keywords relating to the topic of security were identified and assigned a preliminary coding. Key-key-words were then ascertained by creating batch files for each of the 110 documents in order to ascertain the degree of dispersion of keywords, and downloaded in Excel format for further analysis. These words were then colour-coded to enable the identification of core texts where the grouping of the lexis of security was particularly dense. Finally, six texts where then selected for qualitative analysis, intensively treated and coded for linguistic features and preliminary themes (Table 1). The second phase prioritised quantitative analysis (QAN+qal). Here, corpus tools were again applied intensively using a combination of concordance, collocation, and cluster data to reveal cross-corpus variations in linguistic phenomena identified previously in the sample. Selections made were also cross-checked via themes suggested by the keyword and keyword distribution data.
128 | P a g e
CONTEST: The United Kingdom's Strategy for Countering HMO Terrorism
Prevent strategy, part 2
Delivering the Prevent Strategy: An Updated Guide for Local OSCT Partners. London: Home Office.
Stronger together: Prevent agenda and community cohesion
Preventing Violent Extremism: Sixth Report of Session 200910 DCLG
2011a 2009b 2009c 2009 2009 2010
Table 1: Core security documents, 2007-2011 (n=6)
4. Results Three principal themes relating to governmentality (after Foucault, 2007) were uncovered in our analysis of this sample of UK security discourse: regulation, knowledge and population. 4.1 Regulation One of the aims of UK counter-terrorism policy within the period under investigation is to deter radical forms of Islamic fundamentalism from being promulgated within the UK, particularly with regard to the recruitment of new members (HMO 2009a, 2009b, 2011a, 2011b). Two regulatory tactics are realized discursively across the corpus in order to achieve this: constituting the police and other security agencies as being embedded in the population; and using language which medicalises subjects who might potentially be engaged in terrorist activities. The following extract from Prevent, part 2 sets out an elaborate network of agents, agencies, and procedures who are engaged in the practice of security. The overarching discursive strategy which is realized in this extract constitutes the security agencies as being embedded within the population. In this respect the population and the police appear to be melded together in joint pursuit of each others' interests. Local police have a critical role to play in working with local communities
to build their resilience to violent extremism and intervening to support individuals at risk of violent extremism. Prevent community engagement will be delivered locally through local policing units basic/borough command units (BCUs) supported, in the areas of highest priority, by dedicated Prevent policing resources. Local forces will work to mainstream the Prevent agenda across all existing engagement activities, including neighbourhood mapping, support for those individuals in the community most at risk of becoming involved in violent extremism (through the Channel Scheme), schools liaison and community intelligence. (HMO 2009c:47) Here, as elsewhere in our corpus, the impression is being built up of a panoply of agencies (collective) and agents (individual) which are involved in the security apparatus. This strategy of generating lists of disparate practices and processes, gives the impression of a `hyper-complexity' associated with the
MacDonald & Hunter
P a g e | 129
practice of security, although the precise relations between the disparate elements are often left unspecified. The keyword LOCAL (ranked 3rd, LL 41301,) occurs six times in this short extract, similarly premodifying both `police' and `communities'. The density of this occurrence reflects the massive presence of this word throughout the corpus (n=14,079). An L1 search reveals that noun phrases occurring here also occur elsewhere in the corpus. This is especially the case with `local communities' (n=458), and `local community
' (n=384), where `communities' and `community' both rank as top collocates of LOCAL; `local police' also occurs 77 times. LOCAL appears to be used here and elsewhere in the corpus to convey positive semantic prosody to the `police' and to the `community' alike. This insistent usage of LOCAL to premodify both `police' and `community' also discursively allies both these parties, and sets them off against `violent extremism', via the antagonistic phrases `resilience to' and `at risk of'. Both the collective (`communities') and the singular (`individuals') securitized subject(s) are interpolated within the subsidiary clause `working with local communities to build their resilience to violent extremism and intervening to support individuals at risk of violent extremism'. `Build' is a top collocate of `resilience', occurring with this metaphorical usage elsewhere in the corpus in the L2 (n=37), as well as in the L1 position (n=34). `Resilience' is invariably in an antagonistic relation - realized by `to', and occasionally 'against' - with `violent extremism' (n=47). A second regulatory tactic is realized in the third objective of Prevent (HMO 2009b): `supporting vulnerable individuals'. This objective continues: Apologists for violent extremism very often target individuals who for a range of reasons are vulnerable to their messages. Although there is presently less evidence on vulnerability in relation to violent extremism compared with other forms of crime, local partners will recognise relevant factors... A range of existing structures and programmes are already in place to support people exhibiting many of these vulnerabilities (e.g. helplines, mentoring programmes) ... (2009b: 27) Here, the person who is amenable to the influence of Islamicist doctrines is described as `vulnerable'. Across the corpus this epithet is a keyword (LL 1880), and most frequently premodifies `individuals' (n=154). An R1 concord search revealed that it also frequently premodifies `people' (n=134) often `vulnerable Young People
', `communities' (n=47), `institutions' (n=27), households, (n=13) as well as `workers' (n=12). An R1 concord search of the adjectival phrase `vulnerable to' also revealed that it is most frequently postmodified by `radicalisation' (n=34), `recruitment' (n=20), `violent extremism' (n=10), as well as `their messages' (n=5), as above. L2 and L3 searches of the `radicalisation' and `violent extremism' also reveal that both noun phrases are also pre-modified by the near synonyms `at risk of'/`at risk from' (n=29 and 16, respectively) and `susceptible to' (n=3). All three of these adjectival phrases usually exhibit some form of negative collocation, not least through their association with the medical register, e.g. `vulnerable to disease', `susceptible to the Aids virus', `at risk of a heart attack
' (BNC 2007). Thus the medicalisation of `vulnerable individuals' is achieved in this document and throughout the corpus through the recontextualisation of
130 | P a g e
medical lexis and linguistic patterning (c.f. Bernstein 2000, MacDonald, Badger and O'Regan, 2009). In the extract above, the patterning of the phrase `evidence on vulnerability' appears to be suggestive of epidemiological discourse, while the expression `exhibiting many vulnerabilities' also implies an association with the symptoms of a disease. While the phrase `evidence on vulnerability' does not appear elsewhere in the L2 position, the equally medicalised `signs of vulnerability' does (n=4), with the most frequently occurring word in the R2 position being `reduce' (n=15), as in the pattern [reduce(s) * vulnerability]. Later in this document, the expression `coming into contact with vulnerable individuals' (27) refers literally to `social contact', but there is also an underlying metaphorical suggestion that `radicalisation', again like a disease, is contagious. The document goes on to specify that, as with the sick and the insane, `those who are vulnerable to... violent extremism' alone should not be `put through the criminal justice system' (28). Instead, here and elsewhere in the corpus these quasi-medical subjects are offered `support' (n=97), and `interventions' (n=10). 4.2 Knowledge The production, maintenance and reproduction of forms of knowledge about the population is another aspect of the exercise of governmental power by the modern state (Foucault 2007). On this argument, the emerging discipline of political economy, fuelled by the concomitant techniques of statistical analysis
, have become the principal way in which modern governments gain knowledge about their populations. However, our data suggests the forms of knowledge transmitted by current modalities of security discourse differ in certain respects from those described by this conceptualisation of governmentality. At the level of discursive practice, there appears to be a shift from forms of knowledge which operate as a process of objectification towards knowledge which operates as a process of subjectification. A qualitative reading of our core document Delivering the Prevent Strategy (OSCT 2009) indicates the knowledge realized in this security document certainly concerns the entire population, e.g.: `programmes need to focus on individuals, communities and places' (2009: 11). However, knowledge also intrudes upon the individual, e.g.: Partners may consider sharing personal information with each other for Prevent purposes, subject to a case by case assessment which considers whether the informed consent of the individual can be obtained and the proposed sharing being necessary, proportionate and lawful (HMO 2009c: 28). This suggests that the discourse of contemporary security discourse exercises governmental power (over the nomothetic, the multiplicity) alongside disciplinary power (over the idiopathic, the singular). However, the documents in our corpus also use distinctive patterns of language to describe the process of disseminating knowledge and the ways in which this dissemination serves to constitute social relations within the security apparatus. What is striking in these documents is the way in which the dissemination of knowledge is realized as a form of social action. With regard to counter-
MacDonald & Hunter
P a g e | 131
terrorism, knowing is doing. One potent indicator of the socialisation of knowledge in this document (HMO 2009c) and elsewhere across the corpus is the regular occurrence of `share', in relation to `values' as well as `information'. The following extract illustrates how one discursive strategy operates to realize the social relations of `Prevent work' as begin coterminous with the sharing of `information', or `knowledge' (as in Foucault 2007). Prevent work needs to be grounded in information about local communities and local risks... [T]wo additional sources of information are now available which can be shared with Prevent delivery partners using new information sharing protocols: Counter Terrorism Local Profiles (CTLPs) ...and Central Prevent Analysis (CPA)... These products provide deeper background information on issues related to radicalisation and violent extremism. (OSCT 2009: 8) This present participial form, SHARING, is key across the corpus (LL 960). Top collocates appear to confirm our argument set out above: `information' (n=169),' practice' (n=60), `good' (n=38), ` local' (n=33),' partners' (n=31), `data' (n=30), `best' (n=22) and 'intelligence' (n=22). An R1 concord analysis also identifies the patterns: `sharing protocol(s)' (n=9), `sharing knowledge' (n=5), `sharing best practice' (n=5) and `sharing personal information' (n=4). A further R1 concord analysis of the phrase `sharing of' also revealed the patterns: `sharing of information' (n=13), `sharing of best practice' (n=9)m `sharing of good practice(s)' (n=9), `sharing of personal information' (n=9) `sharing of knowledge' (n=5), `sharing of data' (n=3) and `sharing of intelligence' (n=3). The past participle form SHARED also occurs in this document and is key across the corpus (ranked 61, LL 4221). The Government is committed to promoting cohesion and our shared values more clearly and strongly across society. The Government regards the promotion of shared values including fairness, respect and tolerance, democracy and the rule of law as a key element of building strong, empowered and resilient communities; tackling all forms of hate crime; and promoting equal opportunities. (HMO 2009c: 21) However, as suggested by this extract, the past participial form regularly cooccurs not with `information' or `knowledge' but with `values'. This is confirmed by reviewing the top collocates of SHARED across the corpus: `values' (n=380), `future' (n=235), `sense' (n=150), `cohesion' (n=86), `promoting' (n=73) and `understanding' (n=69). Additional patterns revealed by an R1 concord search are also more associated with cohesion and Britishness, e.g. `sense of shared belonging (n=5), e.g. `shared experience(s)' (n=18), `shared citizenship' (n=16), `shared identity(es)' (n=15) and `shared British values' (n=9). Thus, the more active participial form SHARING appears to be associated with the social relations of knowledge that are inherent in the active dissemination of knowledge; while the past passive participial form SHARED is associated with the more passive experience of being part of a collectivity. This active engagement with the dissemination of knowledge and information is encapsulated in the key word PARTNERSHIPS (LL 2788). Indeed, the word PARTNERSHIPS appears central to Delivering the Prevent Strategy OSCT
132 | P a g e
2009). Here, PARTNERSHIPS are described as being `key to the successful local delivery of Prevent': ...the breadth of the Prevent challenge ... requires a multi-agency approach. Local partnerships with responsibility for Prevent should include... police, local authorities
..., education, probation, prisons, health and the UK Border Agency
(OSCT 2009: 7). The singular form PARTNERSHIP is also key across the corpus (LL 4226). Top collocates of the plural form which lend it positive semantic prosody are LOCAL (n=260), STRATEGIC (n=79) and COMMUNITY (n=55), mostly occurring in the L1 position. A concord search also reveals that `working' occurs frequently in the R1 position to PARTNERSHIP (n=297); it is also the top collocate. A cluster analysis also suggests something of the significance and dynamism which is accorded to the phrase `partnership working', e.g.: `the importance of' (14), `partnership working between the' (10), `partnership working between local' (7), `through local partnership working' (7), partnership working for delivery' (6). Other top collocates lend the term positive semantic prosody across the corpus: `local' (n=254), `strategic' (n=137) and `community' (n=114). The usage of the word PARTNERS, which features in the title of the text under analysis - An Updated Guide for Local Partners and is salient across the corpus (ranked 42, LL 5285) is also striking here. Again, `local' occurs as top collocate with remarkable frequency (n=666). The widespread usage of these terms serves to establish the tacit presupposition that agents and agencies across the corpus are already working for a common cause even before grounds are established explicitly in the text. A number of distinctive metaphors are used across the corpus to describe this `Prevent work' as a material process, even though it is essentially the dissemination of knowledge. The compound noun TOOLKIT is used metaphorically to refer to other documents in the battery of texts compiled here to deter `violent `extremists'. This features in the titles of several documents such as Learning together to be safe: a toolkit to help schools contribute to the prevention of violent extremism (Department for Innovation, Universities and Skills 2009). It is also a generic term, as in: ...the start of the toolkit gives background information on the threat from violent extremist groups and on what might make young people vulnerable. (HMO 2009c: 15) The stolid practicality connoted by this keyword (LL 965) conveys positive semantic prosody to counter-terrorism practices and also resonates with three metaphoric verbal processes which appear across the corpus. First, TACKLE is used metaphorically in this document and features as a keyword (LL 1718). Concordance data (Figure 1) illustrates how this verbal process is principally used to describe the countering of `violent extremism'.
MacDonald & Hunter
P a g e | 133
374 colleges · Adult and youth justice 15 · Tackling violent extremism on the
375 noMs Extremism Unit: 020 7217 2727; Tackling violent extremism on the
local partners evidence how they are tackling Prevent and identify areas for
377 , empowered and resilient communities; tackling all forms of hate crime; and
378 of all available resources, with action to tackle violent extremism embedded
and communities to take the lead on tackling violent extremist influences.
partners to take informed action to tackle an identified threat. This will
381 the Government's long term strategy to tackle this threat. However the
Figure 1: Concordance data from Delivering the Prevent Strategy (OSCT 2009)
Another material verb which is used figuratively to describe a verbal process is found in the phrase `myth-busting'. The participial form BUSTING is key throughout the corpus (LL 659), almost exclusively appearing in combination with 'myth(s)' in both the L1 and the R1 position (n=106). The word `myth' is also recontextualised in this discursive formation as a consistently negative term. Across the corpus, MYTHS is a keyword (LL 869); and an L1 concord search reveals the variety of combative verbs which co-occur to position the word in an antagonistic frame, e.g.: `challeng(ing) (n=17), `dispel(ling)' (n=16), `address(ing)' (n=8), `counter(ing)' (n=7), `tackl(ing)' (n=7), `rebut' (n=6), `bust(ing)' (n=5) and `combating(ing)' (n=4). Finally, the verb `peddle' is also used to import negative semantic prosody to Islamicist rhetoric, as in `... allow us [to] challenge myths often peddled by violent extremists...' (HMO 2009c: 18). A concord search for the lemma `peddl(e)' (Figure 2) reveals the derogatory usage of all three forms `peddle', `peddled' and `peddlers'.
the myths and half- truths being peddled by violent extremists and to
and allow us challenge myths often peddled by violent extremists in their
3 to improve the prosecution of those who peddle extremist messages and commit
enough the preachers of hate and the peddlers of simplistic, yet nihilistic,
of the freedoms in this country to peddle their messages without debate
of the freedoms in this country to peddle their messages freely. If they are
Figure 2: Concord data for the lemma `peddl(e)'
These examples illustrate that metaphorical language with both positive and negative connotations is regularly used across the corpus. This imports a sense of wholesome materiality to the idea of `Prevent work', and colours the social relations underlying the production, transmission and reproduction of knowledge by the members of a complex security apparatus.
134 | P a g e
4.3 Population As well as being objectified in the epistemological systems of political economy and the human sciences (Foucault 1970), the population of the modern state is the focus of the technologies of the police, army and other regulatory forces (Foucault 2007). One of the less prototypical documents selected for qualitative analysis Stronger Together: A new approach to preventing violent extremism (Turley, 2009) - is produced by the `independent think-tank', the New Local Government Network (NLGN). Here, it positions itself antithetically to the stance of Prevent (2009b): The origins of Prevent and much of the supporting guidance makes clear that this approach is clearly targeted as a direct response to the threat of Al Quaeda-inspired terrorism, rather than being a tool for enabling wider community cohesion.... Experience has shown that violent extremism can emerge from even the most cohesive communities, but extremist messages are less likely to find support, and are more easily isolated in a cohesive environment. (Turley 2009: 10-11) The discursive constitution of `us' and `them' has been explored by earlier research on security discourse (e.g. Graham et al 2004; Meadows 2007); and across this corpus considerable discursive effort is expended in order to define in-group and out-group categories. Here `community' and by implication its members is positioned as the in-group. However, it is telling that the `community' has to be brought into being (`enabling...cohesion') by some discursive `tool', such as `Prevent'. This apparent manufacturing of community cohesion as a recent tactic of the governmental regulation of population - emerges as an overarching concern of our entire corpus, with COHESION and COMMUNITY occurring massively (n=10,139, n=13,563) as mutual collocates and the two strongest keywords (LL 71201, LL 54794), with COHESIVE also being key (LL 3937). Both nouns share some top collocates, such as `local' (n=599, n=1,087), `promote' (n=555, n=403), `promoting' (n=361, n=305) and `building' (n=233, n=252). In particular, the association of affirmative verbs such as `promote', `promoting' and `building' lend both words positive semantic prosody, with the root `build' also frequently occurring in the L1 position. Two other top collocates of COHESION are `integration' (n=1,003) and `migration' (n=269), with the phrases `integration and cohesion' (n=720) and `cohesion and migration' (n=164) occurring frequently throughout the corpus. These combinations highlight the fact that that `building community cohesion' chiefly concerns the flow and absorption of new populations into particular milieu within the nation state. Indeed, the proximal co-occurrence of these referents suggest that `cohesion' may well be actually be being used as a euphemism for `integration'. However, the nub of Turley's argument is that `cohesion' is coterminous with security: `It is right that local authorities are at the heart of building safe, secure and cohesive communities' (2009: 5). Yet this state of affairs cannot be arrived at without the intervention of government albeit at a local level.
MacDonald & Hunter
P a g e | 135
All the words in the extract above which constitute the out-group are statistically significant across the corpus: TERRORISM (LL 18787), VIOLENT (LL 13563), EXTREMISM (LL 21327) EXTREMIST (LL 4502); with TERRORIST (LL 7743) and EXTREMISTS (LL 1947) also being keywords. A comparative concord analysis of EXTREMISM, EXTREMIST and EXTREMISTS reveals that all three words share `violent' as a top collocate (n=1,955, n=665, n=168). However, there were also differences in their patterns of collocation. `Preventing' (n=789) `prevent' (n=157) `tackling' (n=117) are the top verbal collocates of EXTREMISM, with the keyword often occurring as object in the phrase `violent EXTREMISM'. However, just as the abstract noun EXTREMISM lacks a clear sense of agency, an L2 concord analysis reveals that there is little sense given by the participial verb forms `tackling' or `preventing' of precisely what agent is carrying out the action; likewise there is hardly ever any designated subject for the root form `prevent', which occurs most frequently as an infinitive. EXTREMIST never occurs as a singular noun, but always as an adjective. An R1 concord analysis reveals that EXTREMIST premodifies two types of noun: one indicates some form of communication - often in a declamatory vein, giving rise to patterns such as `extremist ideology(ies)' (n=126), `extremist views' (n=45), `extremist message(n=s)', (n=32); the other is used to indicate some form of `out-group' collectivity, giving rise to the patterns `extremist group(s)' (n=71), `extremist organisations' (n=53) and `extremist networks' (n=9). This out-group is realized more synoptically by the plural form of the noun EXTREMISTS. One common pattern revealed by an L3 concord search suggests that the plural form is regularly used as an `othering' device in post-modifying prepositional phrase
s to describe some form of activity or communication, e.g.: `messages of violent extremists' (n=10), `activities of violent extremists' (n=4), `influence of violent extremists' (n=4); also, `recruited by violent extremists' (n=12), and 'promoted by violent extremists' (n=4). TERRORISM and TERRORIST are the other strong indicators of out-group membership across the corpus. An R1 concord analysis reveals that TERRORIST also occurs most frequently in combinations which indicate either collectivities - such as `terrorist group(s)' (n=137) and `terrorist organisation(s)' (n=104), or hostilities such as `terrorist attack(s)' (n=176) and `terrorist threat(s)' (n=152). An R1 concord analysis also brings us as close as we get to being able to identify the origins of the terrorist activities to which they refer. As well as the vaguer `international terrorism' (n=225), we also find the more specific descriptions ascribing terrorism to three broad fronts: first, `Al-Qaeda inspired terrorism' (n=14), `Islamist terrorism' (n=10), `Jihadist terrorism' (n=4) `Islamic terrorism' (n=2); second, `NorthernIreland-related terrorism' (n=16), `Irishrelated terrorism' (n=7) `Irish terrorism' (n=3); third, `extreme right-wing terrorism' (n=15). It is intriguing, given the events of recent history that Northern Ireland
is mentioned as frequently as Islamicist groups as a potential source of terrorism. This reluctance to explicitly ascribe the origins of terrorist activity brings us to an important area of problematisation within this discursive formation regarding in-group and out-group identification. This is hinted at in the extract above, and is explicitly set out later in this document, i.e.: `... by focusing on Al Quaeda-inspired extremism... it [Prevent] actively encourages prejudice towards Muslim communities...' (Turley 2009: 11). Both MUSLIM
136 | P a g e
and MUSLIMS are strong keywords in our corpus (ranked 4th, LL 23840; ranked 20, LL 10102); and top collocates of MUSLIM are `communities' (n=904) and `community' (n=449). The phrase `Muslim community(ies)' is also often premodified by the adjective `British' (n=142), as is the keyword MUSLIMS. Yet the continual foregrounding of Muslim communities, however positive, can be interpreted as implying that this social group is in some way separate from the mainstream UK population. A concern regarding the stigmatisation of the `Muslim population', expressed throughout our corpus, is summed up forcefully in our sixth core document Preventing Violent Extremism, Sixth Report of Session 200910: The single focus on Muslims in Prevent has been unhelpful. We conclude that any programme which focuses solely on one section of a community is stigmatising, potentially alienating, and fails to address the fact that that no section of a population exists in isolation from others. (DCLG 2010: 3) Although positioning itself contra Prevent, the strength of problematisation realized in this government report only serves to reinforce the Foucaultian thesis with respect to the concern of the modern state with the integrity of its multicultural population.
5. Discussion The intensified production of counter-terrorism documents in the UK in the wake of 7/7 suggests that the issue of national security remains a central concern of the modern state, although continuously being reconfigured to meet the less tangible conditions of late modernity. Our analysis has uncovered distinctive ways in which the three aspects of governmentality proposed by Foucault (2007) regulation, knowledge and population - are realized throughout the analysis of a sizeable corpus of UK counter-terrorism documents. For Foucault, population emerged at the beginning of the modern period as the object of the `science of government' and the calculations of political economy (2007: 103-110). Since then, government has been as much about population as population has been about government. Yet this gives the impression of two characteristics that our evidence suggests are being superseded in this contemporary discursive realization of governmentality: first, the unitary nature of the population of the nation state; and secondly, the objective nature of the knowledge about that population. The keywords COMMUNITY and COHESION emerged from our corpus as indicators that population remains a central focus of the modern state. However, our analysis would suggest that the nature of that concern has now broadened from the population being constituted as demographic `problem' to be addressed quantitatively through statistics and economics (ibid: 104). Instead, our documentary analysis suggests that population is now being addressed by government as a more of a qualitative issue. Given the fact that the 7/7 attacks on the London transport system were carried out by UK citizens, one concern regards the homogeneity of the population of the modern state in a period of unparalleled global flows of economic migrants
and political refugees. Therefore our analysis identified a range of top keywords in the corpus which appear to be discursively constituting an `in-
MacDonald & Hunter
P a g e | 137
group' and an `out-group' within the nation state. However, despite the findings of earlier research in the rhetoric leading up to the 2003 Iraq invasion that the political discourse in the public sphere appeared to identify an enemy in order to demonise it (Bhatia 2009; Graham, Keenan and Dowd 2004; Johnson 2002; Meadows 2007), our analysis indicated a curious absence of precise specification of the source of the threat to the nation state. Moreover, while the consistent identification of a discreet `Muslim community' can be seen as stigmatising or `unhelpful' (Appleby 2010; DCLG 2010), we have also found evidence of considerable discursive work being carried out to constitute this social group as being an integral part of the population of the nation state. In another respect, if - in contradistinction to the focus of the ancient sovereign administration with the family - concern with population has become one of the hallmarks of modern government (Foucault 2007: 103), our documentary analysis would also suggest that the level of engagement of government with population is changing in late modernity. This emerges not only through the salience of the keyword LOCAL, and its regular occurrence as a collocate across the corpus, but also from our core texts, e.g.:. The evolution of local government's role as `place-shaper' means that it is no longer just a deliverer of services but has a key role to play in leading and shaping the way we live our lives with one another. (Turley 2009: 7) The principal agent of government which regulates both the `shaping' and the security of `places' is the police. Foucault sums up the role of the police as the `set of interventions and means that ensure that ...coexistence will be effectively useful to the constitution and involvement of the state's forces' (2007: 327). In its earliest manifestation, the police were principally concerned with this in terms of utility: the `number of citizens', the `necessities of life', `public health', the `regulation of the professions' and the `circulation of goods' (ibid: 322-325). While the police force remains central to the contemporary realm of security, our documents construct both the role and the constitution of the police rather differently. First, the police appear to be engaged in the project of `community cohesion' to the extent that their concerns are constituted as increasingly indistinguishable from those of the `community' itself. Furthermore, what is in essence an ideological war is constituted throughout our corpus though the use of metaphorical language as rather humdrum artisanal labour. This `Prevent work' is also no longer being carried out by the police alone but is marked across our corpus as being SHARED by a panoply of different agencies, working together collegially in PARTNERSHIP. The final discursive tactic that has been revealed in our corpus is to constitute the `vulnerable' subject though a novel modality of `therapeutic discourse' (c.f. Ecclestone and Hayes, 2009). A number of adjectival phrases are regularly deployed to convey negative semantic prosody to those who might be in danger of subscribing to Islamicist beliefs and practices, i.e. `vulnerable to'/'at risk of'/'at risk from'/'susceptible to' X, where X is a negative event or problem. The semantic impact of these phrases is not only to assign negative prosody to the noun phrases which they pre-modify - thereby ascribing a negative value to `radicalisation' and `violent extremism' - but also in part to
138 | P a g e
invoke a register which enmeshes this `imaginary' subject in the technologies of medicalisation (Bernstein, 2000; Foucault, 1973). The effect of this is to diminish the claim of the medicalised subject that his/her behaviour is autonomous and self-determined. In conclusion, the semantic configurations constituted through the particular combinations, regularities and frequencies of the words analysed above are not the realization of some set of a priori policies, or constellation of strategies and tactics that already `exist' in some way pre-defined by the government of the day. Rather, the production of these words and statements are coterminous with the very strategies and tactics of governmentality itself (Foucault 1972, 2007). On this analysis, what is commonly regarded as the `state' no longer appears as a unitary, substantive phenomenon but rather as a de-essentialised, plethoric network of lines of engagement which is realized through the discursive practice of governmentality, within which evidence would suggest security remains central.
References Appleby, N. (2010). Labelling the innocent: how government counter-terrorism advice creates labels that contribute to the problem. Critical Studies on Terrorism 3 (3): 421-436. Baker, P. (2010). Representations of Islam in British broadsheet and tabloid newspapers. Journal of Language and Politics 9: 310-338. Baker, P. and T. McEnery (2005). A corpus-based approach to discourses of refugees and asylum seekers in UN and newspaper texts. Journal of Language and Politics 4: 197-226. Baker, P., C. Gabrielatos, M. KhosraviNik, M. Krzyzanowski, T. McEnery and R. Wodak (2008). A useful methodological synergy? Combining critical discourse analysis and corpus linguistics to examine discourses of refugees and asylum seekers in the UK press. Discourse & Society 19: 273-306. Bernstein, B. (2000). Pedagogy, Symbolic Control and Identity: Theory, Research, Critique. (rev. edn.). Lanham: Rowman & Littlefield. Bhatia, A. (2009). The discourses of terrorism. Journal of Pragmatics 41 (2): 279289. British National Corpus, version 3 (BNC XML Edition). (2007). Distributed by Oxford University Computing Services on behalf of the BNC Consortium. URL: http://www.natcorp.ox.ac.uk/ Burchell, G., C. Gordon and P. Miller (eds.) (1991). The Foucault Effect: Studies in Governmentality. Chicago: University of Chicago Press. Chouliaraki, L. and N. Fairclough. (1999). Discourse in Late Modernity: Rethinking Critical Discourse Analysis. Edinburgh: Edinburgh University Press. Chang, G.C. and H.B. Mehan. (2008). Why we must attack Iraq: Bush's reasoning practices and argumentation system. Discourse & Society 19 (4): 453-482. De Beaugrande, R. (2004). Critical discourse analysis from the perspective of ecologism. Critical Discourse Studies 1 (1): 113- 145. Dean, M. (2010). Governmentality: Power and Rule in Modern Society. London: Sage. Department for Innovation, Universities and Skills. (2009a). Learning Together to be Safe: A Toolkit to Help Colleges Contribute to the Prevention of Violent Extremism. London. Department of Communities and Local Government. (2010). Preventing Violent Extremism: Sixth Report of Session 200910. London.
MacDonald & Hunter
P a g e | 139
Ecclestone K. and D. Hayes (2009). The Dangerous Rise of Therapeutic Education. Abington: Routledge. Fairclough, N. (2000). New Labour, New Language. London: Routledge. Fairclough, N. (2003). Analyzing Discourse: Textual Analysis for Social Research. London: Routledge. Foucault, M. (1970). The Order of Things. London: Tavistock. Foucault, M. (1972). The Archaeology of Knowledge. London: Tavistock. Foucault, M. (1973). The Birth of the Clinic. London: Tavistock. Foucault, M. (1977). Discipline and Punishment. Harmondsworth: Penguin. Foucault, M. (1980). Power/Knowledge: Selected Interviews and Other Writings, Michel Foucault, 1972-1977 (Colin Gordon, Ed.) Brighton: The Harvester Press. Foucault, M. (2007). Security, territory, population; lectures at the College de France, 197778 (trans. G. Burchell). London: Palgrave Macmillan. Foucault, M. (2008). The Birth of Biopolitics: Lectures at the Collиge de France 19781979. Basingstoke: Palgrave. Freake, R., G. Gentil and J. Sheyholislami (2011). A bilingual corpus-assisted discourse study of the construction of nationhood and belonging in Quebec. Discourse & Society 22 (11): 127. Gabrielatos, C. and P. Baker (2008). Fleeing, sneaking, flooding: a corpus analysis of discursive constructions of refugees and asylum seekers in the UK Press 1996-2005. Journal of English Linguistics 36: 5-38. Gales, T. (2009). Diversity as enacted in US immigration politics and law: a corpus-based approach. Discourse & Society 20 (2): 223240. Gillborn, D. (2006). Rethinking white supremacy: Who counts in whiteworld. Ethnicities 6 (3): 318340. Graham, P., T. Keenan and A.-M. Dowd (2004). A call to arms at the end of history: A discourse-historical analysis of George W. Bush
's declaration of war on terror. Discourse & Society 15 (23): 199222. Home Office (2003). Prevent Strategy. London. Home Office (2006). CONTEST: The United Kingdom's Strategy for Countering Terrorism. London. Home Office. (2009a) CONTEST: Countering International Terrorism: The United Kingdom's Strategy. London. Home Office. (2009b) Prevent Strategy. London. Home Office. (2009c). Prevent Strategy (Part 2) London. Home Office. (2011a) CONTEST: The United Kingdom's Strategy for Countering Terrorism. London. Home Office. (2011b) Prevent strategy. London: Home Office. House of Commons Communities and Local Government Committee (2010). Preventing Violent Extremism. London: DCLG. Institute of Community Cohesion (iCoCo) (2011). Available at: http://www.cohesioninstitute.org.uk/ Johnson, R. (2002). Defending ways of life: The (anti-)terrorist rhetorics of Bush and Blair. Theory, Culture and Society 19 (4): 21131. MacDonald, M.N., R. Badger and J. O'Regan (2009). The social cognition of medical knowledge: With special reference to childhood epilepsy. Critical Inquiry in Language Studies 6 (3): 176 204.
140 | P a g e
Meadows, B. (2007) Distancing and showing solidarity via metaphor and metonymy in political discourse: A critical study of American statements on Iraq during the years 20042005. Critical Approaches to Discourse Analysis across Disciplines 1 (2): 1-17. Miller, P. and N. Rose (1990). Governing economic life. Economy and Society 19 (1): 1-31. Mulderigg, J. (2003). Consuming education: A critical discourse analysis of social actors in New Labour's education policy. Journal of Critical Education Policy Studies 1 (1). Available at: http://www.jceps.com/index.php?pageID=article&articleID=2. Mulderrig, J. (2011a). The grammar of governance. Critical Discourse Studies 8: 45-68. Mulderrig, J. (2011b). Manufacturing consent: A corpus-based critical discourse analysis of New Labour's educational governance. Journal of Educational Philosophy and Theory 43: 562-578. Office of Security and Counter Terrorism (2009). Delivering the Prevent Strategy: An Updated Guide for Local Partners. London: Home Office. Pennycook, A. (2001). Critical Applied Linguistics: A Critical Introduction. Chicago: L. Erlbaum. Preston, J. (2009). Preparing for emergencies: Citizenship education, whiteness and pedagogies of security. Citizenship Studies 13 (2): 187- 200. Rose, N. and P. Miller (1992). Political power behind the state: Problematics of government. British Journal of Sociology 43 (2): 173-205. Salama, A.H.Y. (2011). Ideological collocation and the recontexualization of Wahhabi-Saudi Islam post-9/11: A synergy of corpus linguistics and critical discourse analysis. Discourse & Society 22 (3): 315342. Scott, M. (2008). Wordsmith Tools Version 5. Liverpool: Lexical Analysis Software Ltd. Simone, M.A. (2009). Give me liberty and give me surveillance: a case study of the US Governments discourse of surveillance. Critical Discourse Studies 6 (1): 1-14. Turley, A. (2009). Stronger Together: A New Approach to Preventing Violent Extremism. London: NLGN. Van Dijk, T.A. (2001). Multidisciplinary CDA: A plea for diversity. In R. Wodak and M. Meyer (eds.), Methods Of Critical Discourse Analysis. London: Sage. pp. 95-120. Wodak, R., R. De Cillia, M. Reisigl and K. Liebhart (2009). The Discursive Construction of national identity
. Edinburgh: Edinburgh University Press.
MN MacDonald, D Hunter