Introduction

Since the beginning of the twenty-first century, both UK laws on terrorism and those regulating the media have been gradually reinforced and intensified. Whilst the Broadcasting Ban introduced by the British government in 1988 was described as “the furthest the British government ha[d] gone to control the media directly” and a measure of direct-censorship (Miller 1995, p. 72), since 9/11 British counter-terrorism has created new offences and introduced measures that, according to law professor Cram, “go considerably further than the previous bans on the direct broadcasting of Sinn Féin’s representatives and their sympathisers” (2006, p. 336). For instance, the most recent Counter-Terrorism and Border Security Act (2018) has been harshly criticised by human rights and civil liberties organisations for posing a “significant threat” to freedom of expression, freedom of thought, and freedom of information (Liberty 2018; Article19 2018).

Despite this growing trend in media control policies, and although connections between the media and political violence have been an ongoing concern for British political elites, a review of the current literature reveals a lack of research on political elites’ (historical) understandings of the role of the media in relation to political violence. Whilst there is a considerable body of work on elite and expert discourses on terrorism, such as research that investigates the construction of terrorism and the terrorist threat in political elite and expert discourses, there is a lack of research on how political elites interpret the relationship between the media and terrorism and, more specifically, how they envision the effects of terrorism-related media content on audiences.

This article aims to fill this gap through a historical comparison of British political elite discourses on terrorism, the media, and audiences. Analysing key parliamentary debates (PDs) in which British MPs discussed the introduction of far-reaching media control measures (i.e. the Broadcasting Ban in 1988 and the Counter-terrorism and Border Security Bill in 2018), the article investigates how British MPs conceive the relationship between the media, political violence, and audiences, and how this understanding plays a role in the legitimation of policies. PDs represent a rich source of linguistic data to research and historically compare influential discourses in a number of social and political contexts (van Dijk 1993). Moreover, the PDs in which the above policies were discussed and negotiated have not been studied in depth, let alone compared; this material has been largely overlooked by terrorism and media scholars, and literature has certainly failed to examine the two debates in conjunction. Thus, this article makes an important contribution by analysing and contrasting under-researched primary materials (PDs transcriptions) and exploring an equally under-researched topic: the construction of relationships between the media, terrorism, and audiences in British political elite discourse in 1988 and 2018.

The historical comparison was conducted through a discourse-historical approach (DHA) (Reisigl and Wodak 2016; Reisigl 2017), employing a concept that has not previously been used to investigate elite political discourses and the development of public policies: the “imagined audience”. The concept of the imagined audience has been used since the 1970s to explore the ways in which various communicators (writers, journalists, performers, and more recently social media content creators) imagine their public (i.e. the different ways audiences are perceived) and the implications of these “mental conceptualisation[s]” (Litt 2012, p. 331). This article argues that the concept of the imagined audience is a useful tool for analysing fictionalisations of the public in political elite discourses on terrorism and the media. The concept offers an original perspective to investigate how political elites legitimate and/or challenge media control measures through their (discursive) constructions of media audiences. Thus, this article also contributes to contemporary research on the imagined audience, broadening the scope of the concept by employing it to research PDs and the introduction of media control measures.

Drawing on a DHA, I formulated the following discourse-related research questions:

RQ1:

How is the relationship between terrorism and the media represented and discussed in 1988 and 2018 British PDs?

RQ2:

How is media influence (on audiences) represented and discussed in these discourses?

RQ3:

How are audiences constructed in the two PDs?

RQ4:

How does the imagined audience impact on the introduction of media control measures?

Terrorism and the media: analysing the “imagined audience” in elite discourse

The first part of this section highlights the value of PDs for analysing and comparing political elite discourses and stresses the lack of previous scholarly attention to British PDs in which the relationship between the media, terrorism, and the audiences is discussed. The second part introduces the concept of the imagined audience and argues that this represents an innovative tool for dissecting and comparing elite discourses on terrorism and the media and examining their implications.

Political elite discourses on terrorism: analysing parliamentary debates

Parliamentary debates (PDs) have rarely been studied despite their crucial role in legislation and influence on citizens' everyday lives (van Dijk 1993; Wodak and van Dijk 2000). Nonetheless, an important body of literature in the field of discourse analysis has demonstrated the value of analysing PDs to examine the influential role of political elite discourses in the (re)production of meanings and the formation of public opinion. For example, van Dijk (1993) analyses PDs on immigration legislation between 1983 and 1990 to explore the construction of racist notions in the speeches of British political elites. Similarly, the volume edited by Wodak and van Dijk (2000) investigates the role of political elites in the reproduction of racism (and antiracism) in PDs, and examines dominant discourses on migrants and refugees and the (re)production of prejudices and stereotypes in six European countries including the UK. Also based on the analysis of PDs, Huysmans and Buonfino (2008) investigate the role of British political elite discourse in securitising migration and asylum by linking immigration to counter-terrorism in the post-9/11 debates.

Research on PDs acknowledges that political elites play a significant role in the (re)production of particular assumptions and/or ideologies that reflect and at the same time are implemented “in other domains, for instance, in the media, academic research, education, or in corporate business and employment” (van Dijk 1993, p. 49). As a result, whilst research on elite discourse might include the discourse of actors who have privileged access to various forms of public discourse (i.e. “symbolic elites”, van Dijk 1993, p. 17), the analysis of PDs still provides a notable way to study dominant discourses that have the power to formulate, legitimise, and introduce policy and laws (Wodak and van Dijk 2000). Moreover, the analysis of PDs also enables comparative research (Wodak and van Dijk 2000) and historical analysis since it offers a “strong institutional locus for researching political positioning among the political elite over time” (Huysmans and Buonfino 2008, p. 766).

Whilst an important corpus of research has analysed how terrorism and terrorist threats are constructed in elite political speech and how these discourses have contributed to the implementation (and public acceptance) of particular counter-terrorism strategies (Stampnitzky 2013; Silberstein 2004; Zulaika and Douglass 1996), fewer studies have focused specifically on analysis of PDs. Some notable exceptions include Swinhoe (2021), who examined political elite discourses on terrorism and religion in British PDs, and Legrand and Jarvis (2014), Jarvis and Legrand (2016), and Onursal and Kirkpatrick (2021), who analysed PDs on the proscription of terrorist organisations in the UK. This literature demonstrates how the analysis of PDs provides a valuable tool for identifying dominant discourses and frameworks on terrorism and security that are (re)produced and applied both inside and outside parliament. For instance, Jarvis and Legrand’s (2016) research reveals the role of parliamentary discourses on the construction of an antagonistic relationship in which the UK represents a liberal, open, and responsible “Self” and the terrorist is constructed as the illiberal and irrational “Other”, a construction which is then reproduced in other contexts.

With the exception of Miller’s (1994, 1995) and Moloney’s (1991) work, which briefly summarises the main arguments employed by the British government in 1988 when introducing the Broadcasting Ban, there is a lack of research on British PDs discussing terrorism and media control. Although researchers have studied the introduction, representation, and consequences of the 1988 Broadcasting Ban (Curtis 1998; Edgerton 1996; Miller 1994; Moloney 1991; Pettigrew 2017) and the impacts of recent counter-terrorism legislation on freedom of speech and the media generally (Awan 2012; Cram 2006), there is an absence of research investigating how British political elites understand the complex relationship between terrorism, the media, and audiences.

The imagined audience in elite discourse

The concept of the imagined audience is not new and can be traced back to literary studies and communication sociology research in the 1970s (Atwood 1970; Gans 2004 [1979]; Ong 1975). The term was developed to explore how various communicators, such as writers and particularly journalists, imagine their public. As Ong puts it, the audience “is always a fiction” (1975, p. 17), in the sense that authors can never have a complete knowledge of who their audience is. Thus, the audience is rather a “vague abstraction” and/or a “mental conceptualisation with whom we are communicating” (Litt 2012, p. 331). The concept of the imagined audience was coined to explore these different imaginings and/or constructions of the audience and their impacts. Early literature on the imagined audience demonstrated the material power of these fictionalisations by revealing how imagined audiences have the capacity to alter both the format and content of the news (Gans 2004 [1974]), and how different imaginaries and/or conceptions of audiences not only cast individuals in a particular role but also expect them to play that role (Ong 1975).

With the coming of digital media, there has been a resurgence of interest in the concept of the imagined audience as a way to explore the perceptions that content creators and/or social media users (Litt 2012; Litt and Hargittai 2016; Marwick and boyd 2010) have of their audiences. This renewed interest in the imagined audience has been accompanied by a new wave of research on journalists’ constructed audiences (Ferrucci et al. 2020; Coddington et al. 2021). These recent studies explore the variations in imagined audiences between different individuals and how diverse conceptions of audiences influence what and how journalists and social media users communicate. For instance, whilst some picture their spectators as rational and/or smart actors, others envision them as irrational and only interested in sensationalistic news, and still others perceive their audiences to be similar to them, or “homophilous” (Coddington et al. 2021, p. 1029). Research has also explored the different mental images that Twitter users have of their audiences; whilst some users imagine their audience as their friends, others picture an ideal person as their reader, and others imagine their followers to be a “broad audience with disparate tastes” (Marwick and boyd 2010, p. 121). Although website metrics may reveal some characteristics of audiences in the digital world, media creators are still largely left to imagine whom they may be communicating with, and various sources, including personal relationships and intuition, play a major role in the construction of audiences (Coddington et al. 2021). In short, research on the imagined audience has largely focused on investigating constructions of audiences by different actors, as well as on the origins of these constructions and on their manifold implications.

The concept of the imagined audience has predominantly been limited to literary studies, media research and performance studies; however, this article argues that the concept can be a valuable analytical tool for studying political discourses and the development of public policies.Footnote 1 Research on how political elites envision their audiences and the public in general can shed light on the rationalities behind the implementation of particular public policies. Thus, this article employs the concept of the imagined audience to explore how members of the British parliament have applied different constructions of media audiences to support and/or legitimise (or to challenge/oppose) various media control measures. Specifically, an analysis of PD discourses can reveal the significant transformations that have occurred in elite actors’ perceptions of (British) audiences, which have far-reaching implications. The following section introduces the discourse-historical approach (DHA) as a useful method for analysing variations in imagined media audiences constructed by British political elites during different historical periods.

A discourse-historical approach for analysing British parliamentary debates on the media and terrorism

The discourse-historical approach (DHA) is an interdisciplinary approach within critical discourse studies (CDS)—previously referred to as critical discourse analysis (CDA) (Wodak and Meyer 2016)—and has been “acknowledged as one of the main approaches to CDA” since the 1990s (Reisigl 2017, p. 45). All CDS modalities see language as a form of social practice and/or action (Fairclough and Wodak 1997; Forchtner 2011) and understand the role of discourse as signifying and constituting social realities (Barker and Galasiński 2001) rather than neutrally representing and describing them. All the approaches also share the aim of studying society, including social processes, structures, and social changes through discourse (Flowerdew and Richardson 2017, pp. 1–2). Nonetheless, within the “heterogeneous movement” (Forchtner 2011, p. 1) of CDS, different approaches have “different theoretical backgrounds, oriented towards different data” (Wodak and Meyer 2016, p. 5) and “distinctive, but also overlapping methods” (Flowerdew and Richardson 2017, p. 2). The DHA stands out for its interdisciplinary orientation and its historical analysis (Reisigl and Wodak 2016, p. 57) and, as its name indicates, the approach seeks to “stress [its] strong historical research interest” (Reisigl 2017, p. 44). This approach draws attention to “comparisons over time” (Reisigl and Wodak 2016, p. 39) and puts “weight on historical subjects and on the historical anchoring, change and echo of specific discourses more than other CDA approaches” (Reisigl 2017, p. 49).

My research employs a DHA as the most suitable method to analyse the discourses of political elites on terrorism, the media, and audiences through the study of British PDs. A DHA makes it possible to explore changes over time in the discourses used by British political elites to characterise relationships between the media and terrorism, and to critically explore the discursive construction of audiences in PDs at two points thirty years apart. In other words, a DHA is a valuable approach for identifying historical shifts in political elites’ fictionalisations of audiences in key PDs. Drawing on a DHA, I formulated (discourse-related) research questions that directed the analysis towards representations of the relationship between terrorism and the media in the 1988 and 2018 PDs (RQ1); how the media’s influence on audiences is represented and discussed (RQ2); how audiences are constructed in the two PDs (RQ3); and finally the impact of these conceptualisations of audiences on the legitimisation of (or opposition to) media control measures (RQ4).

Empirical data were collected from the Hansard website, which is a highly reliable official source providing transcripts of UK Parliamentary proceedings. The PDs selected include three sessions in the Commons chamber on Broadcasting and Terrorism, two of which took place in 1988 before the ban came into force (19th October and 2nd November) and one on 17th December 1992 when the ban was already in place (40,000 words in total), and six sittings in the Commons chamber debating the Counter-terrorism and Border Security Bill which took place between 26th June and 10th July 2018 (approximately 110,000 words in total). The debates on these two bills were selected because they both led to unprecedented media control measures, and because they represent the most relevant PDs in recent British political history discussing the relationship between the media, terrorism, and audiences.

Drawing on the discourse-analytical tools proposed by the DHA, this analysis focused on three distinct features of the discourse used by speakers in the PDs: nomination, predication, and argumentation strategies (see Wodak 2001; Reisigl and Wodak 2016). These strategies pose specific questions that provide a useful basis for rigorous and systematic analysis of the data. Nomination refers to how audiences (i.e. the viewers and/or media receivers) are “named and referred to linguistically” (Reisigl and Wodak 2016, p. 32). Predication refers to the “traits, characteristics, qualities and features” attributed to these audiences (Wodak 2001, p. 72) by speakers in the PDs. Lastly, argumentation strategies analyse are the various arguments employed in these discourses to endorse (or challenge) the implementation of new media control measures. Unlike many DHA analyses, my research goes beyond the examination of “typical content-related argument schemes […] carried out against the background of the list of topoi” (Wodak 2001, p. 74) and reconstructs the arguments from the original texts to enable an in-depth comparison of the two debates. Thus, this article does not limit itself to identifying the topoi; rather, it explores in detail the arguments employed by MPs in order to illustrate the dramatic shift in elite discourse over a period of thirty years, which correlates with a transformation in how audiences are constructed.

My analysis of the PDs, guided by focusing on the three strategies mentioned above, was executed in three main phases. The first stage involved a careful reading of the PDs, which together constitute a corpus of approximately 150,000 words, accompanied by the writing of analytic or “CDA memos” (Ziskin 2019). During the second stage, I coded the data in line with the nomination, predication, and argumentation strategies. This involved coding the fragments of text in which audiences were both explicitly and implicitly nominated and predicated (i.e. segments in which audiences were referred to and/or described, or in which particular features were attributed to them) and identifying and classifying the various arguments employed to support (or challenge) the introduction of particular media control measures. In the analysis of argumentation strategies, I paid special attention to claims about the media’s relationship with and impact on the imagined audience. The final stage involved creating tables to classify the data, enabling a systematic comparison of the discursive strategies employed in the two sets of debate. The following section shows how analysis of the PDs revealed a serious (historical) shift in the construction audiences, which was marked by a transformation in nomination, predication, and argumentation strategies.

Analysis and discussion: conflicting audiences in the discourses of British political elites

The discussion of the results begins with a brief description of the historical contexts of the two sets of PDs, followed by a description of and commentary on the main discourse strategies employed by MPs in each case. The analysis explores the four RQs and discusses the implications of different historical conceptualisations of media audiences.

The Broadcasting Ban (1988) and its imagined audience

Historical context

Whilst British policy had long aimed “to push Ireland to the margins of British politics” (Miller 1994, p. 2), and British broadcasters had been ignoring the existence of serious grievances and divisions in Northern Ireland since the 1921 partition (Moloney 1991), this situation changed rapidly when the Civil Rights Association took to the streets in 1968. With the intensification of the NI conflict and the escalation of violence at the end of the 1960s, national and international media attention proliferated, and British media reporting of NI became an everyday phenomenon (Pettigrew 2017; Miller 1994). Scholars refer to the British government’s battle to exercise control over the media coverage of NI during the “Troubles” as the “propaganda war” (Curtis 1998; Miller 1994; Pettigrew 2017). Media coverage of the conflict became increasingly important for the subsequent British governments as they sought to defend their role in NI and win support from the British public (Curtis 1998). During the 1970s, even though broadcasting companies had introduced “internal controls and rules largely to obviate government and public criticism” (Moloney 1991, p. 11), tensions between the British government and broadcasters increased in response to media coverage of events such as Bloody Sunday in 1972 and the broadcasting of programmes such as three-hour current affairs special The Question of Ulster, also in 1972.

Before the Broadcasting Ban was introduced, other media control measures, including legislation (some of the seventy pieces of “emergency” legislation passed between 1970 and 1986 introduced new controls over the media), intimidation, political pressure, and self-censorship were active (see Miller 1994). For example, between 1970 and August 1989, “a total of 76 TV programmes on NI, [including] documentaries, plays and even church services were either banned, refashioned, cut, or postponed because of either internal or external pressure” (Moloney 1991, p. 11). Although some journalists and outlets made notable efforts to resist government pressure and maintain their journalistic independence, scholars note that by the 1980s, “an increasingly hostile relationship had developed between the British government and the broadcast media” (this was linked to coverage of the Falklands War in 1982 and the miners’ strike in 1984/85 as well as coverage of NI) (Pettigrew 2017, p. 229). For example, Margaret Thatcher’s request that the Attorney General consider taking legal action against the BBC over their interview with the INLA on Tonight in 1979 represented a “completely new departure in the relationship between broadcasting and the state” (Schlesinger et al. 1983, p. 127 cited in Miller 1994:34). Ultimately, it was outcry from the Thatcher government over programmes such as the 1995 BBC documentary Real Lives: At the Edge of the Union, which included an interview with Martin McGuinness, and the 1988 documentary Death on the Rock which contradicted the official version of events by examining the deaths of three IRA members in Gibraltar at the hands of the British Special Air Service on 6th of March 1988, as well as the rise of the Sinn Féin, that tipped the scales in favour of the introduction of more direct media control methods (Curtis 1998; Miller 1995). In October 1988, Home Secretary Douglas Hurd announced the so-called Broadcasting Ban, which would restrict the broadcasting of statements by representatives or supporters of eleven Irish republican and Ulster loyalist political and paramilitary groups.

Nomination, predication, and argumentation strategies

In the debates on the Broadcasting Ban, there are no major differences in how speakers supporting the ban and those opposing it refer to media audiences. MPs construct audiences as an abstract majority, employing terms such as “the British people”, “most people throughout the United Kingdom”, “the general public”, “viewers”, “the overwhelming majority of men and women”, etc. (HC Deb. 1988a, b, 1992). On many occasions, MPs opposing the ban refer to audiences using the first person plural, “we” and “us”, whilst some MPs emphasise the citizenship of the audience by referring to them as “fellow citizens”, “constituents”, “the individual” and “any human being” (Ibid.). In terms of predication strategies, it is notable that throughout the parliamentary discussions, all speakers assume audiences to be rational, capable, and informed. Whilst opposition MPs stress the capacity of individuals to make “correct judgements”, “outface evil with argument”, deal with “unacceptable” opinions, spot “a moral cretin”, and cope with the “evil banalities of the supporters of terrorism”, the government (and its supporters) emphasise the British public’s anger, arguing that the appearance on television of terrorists and their supporters causes “offence” to viewers (HC Deb. 1988a, b, 1992). They construct audiences as “decent”, as not wishing to watch terrorists/criminals on their screens, and/or as uninterested in listening to their views. Nonetheless, British MPs backing the broadcasting restrictions also characterise audiences as likely to feel intimidated or afraid of (in)direct threats from terrorists on TV. This is contested by the opposition speakers, who argue that British citizens have great “intelligence” and judgement and do not need the protection of censorship, and that they are individuals with the right to free speech and a free press. These MPs argue that British citizens are genuinely interested in understanding the underlying causes of the violence in NI, and that they must understand it as they are all responsible for finding a solution.

Those supporting the broadcasting restrictions offer four main arguments as to why the ban was necessary: 1. Broadcasting the voices of terrorists or their supporters causes offence to the viewers; 2. Terrorists and their supporters do not have the right to appear on TV/radio (supported by the idea that audiences have the right not to see terrorists on their screens); 3. The ban would protect audiences from the fear caused by watching or listening to advocates for terrorism; 4. Media outlets were acting irresponsibly and the government needed to introduce clear guidelines and limits. These arguments are contested by four main points made by the opposition: 1. The ban was an act of censorship which would violate the rights and liberties of citizens and a free society, including freedom of information, freedom of expression, and freedom of publication, and set a dangerous precedent. Many MPs described the ban as authoritarian and illiberal, and also argue that it would promote self-censorship; 2. The ban would be counter-productive in the fight against terrorism. Some MPs propose alternative counter-terrorism/conflict resolution measures, including negotiating a peace settlement and including all sides in the democratic process; 3. Individuals are rational and play an important role in democracy (i.e. people need to face the facts). MPs stress the capacity of the British public to make correct judgements, as well as the power of the democratic debate (i.e. “the public are entitled to hear any point of view”; “opinions should compete in a free market” HC Deb. (1988b pp. 28–29)). 4. The media plays a crucial role in democracy, it builds the public sphere by informing citizens, scrutinising activities, examining facts, “penetrat[ing] the truth”, and exposing terrorism/terrorists (HC Deb.1988b, p. 48). These MPs argue that a free and impartial press is in the public’s interest.

Analysis

In the 1988 and 1992 PDs, the relationship between terrorism and the media (RQ1) is represented in contradictory terms: whilst the government and their supporters represent the media as a tool/weapon that terrorists employ to intimidate and spread fear, the opposition constructs it as the best tool to counter-terrorism, insisting that in the light of the day (by obtaining media representation) terrorists and their supporters will lose the battle to non-violent actors. According to this view, terrorism is weakened by media exposure, when viewers are able to listen to terrorists’ arguments. When it comes to the media influence on audiences (RQ2), there is an overall consensus among MPs participating in the debates that the media does not increase support for terrorism, and that it does not lead individuals to participate in political violence. The ban’s supporters stress the fear and anger that the media causes to viewers when it provides access to terrorists and their supporters, whilst those opposing the restrictions characterise media influence on audiences as positive and beneficial; the media’s role in informing the public is emphasised, and this in turn enables the public to participate in politics/democracy.

The audience is unanimously constructed (RQ3) as a majority, and often referred to as the British/general public. Without relying on ‘hard facts’, MPs characterise audiences as rational, responsible, ethical, and intelligent, with a capacity for judgement, and not easily influenced and/or deceived. However, whilst a traditionally liberal view of the individual and the society prevails in the discourse, there is no consensus over the characterisation of media outlets. The government (and others backing the ban) portray the media as irresponsible, sensationalistic, indecent, struggling to set appropriate limits, and/or only interested in increasing their audience share. Meanwhile, those challenging the ban who describe the media as transparent, and its democratic role as essential. Thus, the main difference in the PDs is not in political elites’ construction of audiences, but in their conflicting representations of media outlets.

This construction of the imagined audience plays a crucial role in the rhetorical defence of (or opposition to) the introduction of the media control measures (RQ4). Arguing in favour of media restrictions, the government focuses on both the “outrage” and fear apparently felt by the British public when they see terrorists on their screens, and the media’s irresponsibility/ inability to make judgements about appropriate content. In contrast, the opposition employs a liberal depiction of both audiences and the media to oppose what they conceive of as the “illiberal” and authoritarian/undemocratic ban. Interestingly, whilst there is an overall consensus in how audiences are constructed in the PDs, each side attributes different features to them (being interested vs. not being interested in political violence; feeling fear vs. being fearless) in order to discursively back or challenge the ban.

The Counter-Terrorism and Border Security Bill (2018) and its imagined audience

Historical context

The historical context in which the Counter-Terrorism and Border Security Bill (2018) was discussed in parliament was marked by an important transformation in the focus of British counter-terrorism and a radical shift in the construction of the “suspect communities”. This period was also characterised by several violent attacks carried out in the UK in 2017, and a gradual introduction of media control measures focused on countering terrorism and radicalisation. In 2002, as a result of the peace process in NI and the emergence of the global “war on terror” after 9/11, resources for countering “international terrorism” exceeded those directed to NI (considered “domestic”) for the first time in British history (Hewitt 2011, p. 94). British counter-terrorism targeted “Islamic terrorism” as the most significant threat to national security and focused on countering so-called homegrown terrorism after the terrorist attacks in London in 2005. Scholars have noted how Muslims tended to be negatively represented in the media during this period, often being actively associated with terrorism (Poole and Richardson 2006), and a “popular perception of a new ‘enemy within’, namely, ‘homegrown’ Muslim terrorists, especially young, male Muslims” emerged in the UK (Edmunds 2010, p. 216). Employing Paddy Hillyard’s (1993) term “suspect community” (originally conceived to discuss the discrimination and harmful treatment experienced by Irish individuals in the UK with the introduction of Temporary Provisions in 1974), scholars (Hickman et al. 2011) have noted how, since 2005, political discourse, the media, and legislative measures (such as the Prevent Strategy) have constructed Muslims as the “new” suspect community, and “Muslim minority groups have been subjected to pervasive scrutiny in the United Kingdom” (Mythen et al. 2009, p. 736). Research also demonstrates how British mainstream politicians have fostered Islamophobia and engaged in Orientalism in their speeches (Moosavi 2015; Singh 2023). The rise of Islamophobia in the UK during this period has been widely documented (see, e.g. Tell MAMA, n.d.).

Since 9/11, and especially after 7/7, a large amount of counter-terrorism legislation has been passed in the UK, which has significantly increased media control. For example, the 2001 Anti-Terrorism, Crime and Security Act (ATCSA) imposed “a new set of duties on journalists and others to assist police investigation by disclosing information”, including their sources, “in the interests of justice or national security” (Cram 2006, p. 348), and the 2006 Terrorism Act introduced new offences including the “encouragement of terrorism” and distributing terrorist publications (or providing a service that enables others to access them) (see Awan 2012). The introduction and development of the Prevent Strategy (the part of the British Counter-Terrorism Strategy CONTEST which focuses on terrorism prevention) in the following years further increased media control measures. The Prevent Strategy increasingly focused on de-radicalisation interventions (Elshimi 2017), preventing (self-)radicalisation and the dissemination of “extremist” and “online terrorist content” under the premise that “vulnerable” individuals can be drawn into terrorism through the consumption of extremist content (Rodrigo-Jusué 2022). Furthermore, in 2010, the Counter-Terrorism Internet Referral Unit (CTIRU) was created to remove terrorist propaganda and information online (Counter-Terrorism Policing 2018). The aim of preventing (self-)radicalisation by exerting greater control over the media was reflected in the 2019 Counter-Terrorism and Border Security Act, which introduced new terrorism offences related to the speech and the use of the media, including expressing an opinion or belief that is supportive of a proscribed organisation, publishing images in such a way or in such circumstances as to arouse reasonable suspicion that the person is a member or supporter of a proscribed organisation, obtaining or viewing material over the Internet likely to be useful to a person committing or preparing an act of terrorism, and encouraging terrorism or disseminating terrorist publications (HO 2019).

Nomination, predication, and argumentation strategies

Throughout the transcript of the six sittings in 2018, MPs describe media audiences in ambivalent terms; on some occasions, audiences are characterised as a large, abstract group through the use of terms such as “people”, “individuals”, “the general public”, and “the audience” (HC Deb. 2018). However, most often, the audiences are referred to represent small sections of the population and/or minoritised groups. For instance, during the first sitting, Assistant Metropolitan Police Commissioner Neil Basu, a key witness backed by the Minister for Security and Economic Crime Ben Wallace, refers to “a section of our society”, “the malleable vulnerable people”, “individuals”, and “minors” (HC Deb. 2018, pp. 8–10). Government representative Wallace refers to “individuals who mean us harm” and “vulnerable people”, and the construction of media audiences as “vulnerable” is echoed by opposition MPs (idem, p. 72). Whilst the general public are rarely associated with specific attributes and characteristic in these debates, particular “individuals”—“the most vulnerable”, “those who are vulnerable”, “vulnerable adults”, and/or “vulnerable people”—are characterised as easily influenceable, “unaware”, easily exploited, susceptible to being “groomed” to undertake violent acts, be drawn into “extremism”, “become involved in terrorism”, or be brainwashed, and as victims of “radicalisers” and targeted by those “seeking to spread their poison […] and cause the maximum harm” (HC Deb. 2018, pp. 94–97).

Vulnerable individuals are on some occasions discursively constructed as “young”, although most often they are referred to as distinct from children. MPs use phrases such as “the vulnerable and the young”, “a vulnerable adult or a child”, and “children and other vulnerable people” (HC Deb. 2018, pp. 94–95). Interestingly, such depictions of the vulnerable viewer are common to MPs from different political parties (notably Labour and the Conservatives), and are not challenged by anyone throughout the debates. Whilst the elements and/or characteristics that make someone “vulnerable” are not directly explained during the sittings, MPs describe vulnerable individuals as, for example, those who “have become disaffected with friends or traditional sources of authority or guidance—whether that is their local mosque or their family” (Stephen Doughty, Lab); as “converts” (Rupa Huq, Lab); “people with learning disabilities and mental health problems” (Neil Coyle, Lab); individuals with “significant mental health issues” and “lonely, often damaged, young individuals sitting in their bedrooms” (Ben Wallace, Con) (HC Deb. 2018, pp. 95–143). Although no ‘hard facts’ were mentioned during the sittings, on some occasions, MPs and witnesses’ discourses contained few examples of so-called vulnerable individuals. Examples given by Commissioner Basu and various MPs of vulnerable individuals who were “radicalised” include Safaa Boular, Khalid Masood, and “a young man […] with knives and an ISIS flag” for whom there is no evidence that “had ever met a Muslim, was from a Muslim family, or had been to a mosque”, whilst Anjem Choudary and Umar Haque are given as examples of the terrorist and/or “radicaliser” who targets vulnerable individuals (HC Deb. 2018, p. 92). The dominant characterisation of the vulnerable viewer as a Muslim, a Muslim convert, and/or someone with mental health issues is extended several times by Wallace and other MPs, including Stephen Doughty, who argues that people from different backgrounds are “being radicalised” by terrorist propaganda and/or extremist messages, including “neo-Nazis and the far right” (Wallace) and “other types of extremism” (Doughty) (HC Deb. 2018, pp. 95–140).

Wallace, as a government minister, offers four main arguments in support of new restrictions on the media. 1. “Vulnerable people” are being self-radicalised online by terrorist propaganda and “extremism content”, or are being “radicalised” by extremists and/or “radicalisers” who “brainwash” and “groom” them (HC Deb. 2018); 2. The new legislation simply updates outdated terrorism legislation to account for new technological developments, including streaming and the emergence of the social media; 3. The creation of new offences is necessary to give the police and prosecutors additional powers to prosecute “radicalisers”, but will not affect innocent individuals; 4. New measures will protect the general public from those dangerous individuals who have been groomed/brainwashed/radicalised. The lack of any opposition to these arguments is the most significant feature revealed by analysis of the 2018 PDs. Apart from two witnesses (Corey Stoughton, advocacy director of Liberty, and Peter Carter, a member of the Criminal Bar Association) who oppose a number of the proposed measures and vocalise their concerns in similar terms to those used by opposition MPs in 1988, there is a notable lack of opposition to the introduction of the new measures. In short, the main arguments employed by Wallace (and other MPs and witnesses who support the bill) are met with a general lack of resistance, or, at most, some (mainly technical) concerns dispassionately raised by a few MPs.Footnote 2

Analysis

The relationship between terrorism and the media (RQ1) is represented in negative terms throughout the 2018 PD. According to MPs, terrorists use the media to “radicalise” “vulnerable” individuals and draw them into terrorism. Throughout the debates, terrorism is depicted as a phenomenon propagated through the media, and thus terrorism’s relationship with the media can be understood as a relation of dependence. Unlike the 1988 PD, in the 2018 discussions terrorism is conceived, to a great extent, as a technological phenomenon. In turn, media’s influence on audiences (RQ2) is represented as absolute and dangerous. According to the discourse used by MPs, the media are powerful/effective enough to turn “vulnerable” individuals into terrorists and/or terrorist supporters.

An analysis of nomination and predication strategies reveals how audiences, constructed as vulnerable, lost, and made up of “damaged” individuals, individuals with learning disabilities and/or mental health issues, Muslims, particularly Muslim converts and/or young Muslims, are supposedly drawn into terrorism (even if individuals have no other off-line influences) through the influence of terrorist propaganda. The existence of this effect is supported through the construction of audiences (RQ3) as easily influenced, weak, irrational, uninformed, unknowledgeable, and not capable of making their own judgements and/or participating in debates. Such constructions of audiences as weak, easily influenced, and in need of guidance and/or protection form the basis of the main arguments employed by participants who support the introduction of new media control measures (RQ4). In other words, since audiences are to a great extent, unknowledgeable and incapable of making good judgements (for instance, when they encounter extremist messages), restrictive measures, including censorship, are legitimised and welcomed. In short, media control measures are defended as necessary actions to protect vulnerable audiences from themselves and from the rest of the population.

Discussion and summary

The audiences constructed in the 2018 PD differ greatly from those constructed in MPs’ discourse three decades earlier. In the earlier debate, the audiences are universally envisioned as rational, well-informed, and responsible, whilst thirty years later, irrational, vulnerable, and easily influenced audience are discursively deployed by MPs to support highly restrictive media control measures. This shift in the audiences corresponds with significantly different depictions of the media and terrorism in the two sets of debates. Whilst in the 1988/1992 debates, the media was constructed both as irresponsible/offensive and as an essential element for democracy; in 2018, the media was regarded as a mere tool that could be used by terrorists to recruit vulnerable and easily influenced individuals to their cause. In this way, the significant role of the media in the public sphere and democracy is degraded by the vulnerability discourse that strictly focuses on the capacity of the media to brainwash and/or radicalise individuals. This view of audiences and the media resonates with the so-called contagion theory (also known as “contagion hypothesis”) developed during the 1970s, according to which terrorism depends on the media and the media produces the effect of increasing terrorist activity (see Picard 1986; Dobkin 1992; Cottle 2006). Despite the lack of scientific evidence, by stressing an interdependence between terrorism and the media, many exponents of this perspective embrace restrictive and censorship measures (Picard 1986; Kingston 1995; Wilkinson 1997). Similarly, regardless of the lack of consensus among scholars on online radicalisation (so-called self-radicalisation) and the absence of evidence to support this thesis (see Hussain and Saltman 2014; Meleagrou-Hitchens and Kaderbhai 2017), the vulnerability discourse demands and legitimises highly illiberal media control measures. For this reason, it is crucial to critically engage with the impacts of the vulnerability discourse in counter-terrorism and beyond (for example, in wider debates about censorship and online harm—which could also be relying on imaginations of sub-sections of the public as infantile, irrational, and in need of protection/care).Footnote 3

DHA analysis also reveals how the 1988/1992 consensus among political elites that audiences were capable, logical, and prudent has been replaced with a consensus that audiences are vulnerable and unreasonable. A fundamental difference in the two sets of PDs is revealed by analysis of nomination and predication strategies: whilst audiences are assumed to be homogeneous during the first debate, the vulnerable audiences of the second debate are often characterised as representing only part of the population rather than all British citizens. As the discourse analysis shows, the vulnerable, irrational, and, ultimately, dangerous audience is often imagined as Muslim and/or composed of individuals with mental health issues/learning disabilities. The 2018 portrayal of Muslims, particularly young males, as unwise, easily influenced, and potentially dangerous corresponds with academic research showing an intensification of Islamophobia and the creation of Muslims as the new suspect community in the UK, particularly after 7/7 (see Hickman et al. 2011). As a consequence, this article suggests that the construction of a vulnerable, irrational, and highly susceptible audience, in contrast to the rational, capable, and responsible audience that MPs envisioned in 1988, could be influenced by Orientalist and/or racist notions that mark Muslim individuals as inferior, untrustworthy, and essentially unfit for Western liberal democracies (Abbas 2021; Moosavi 2015; Mythen et al. 2009). Thus, the oppositional construction of these two audiences (the 1988 imagined audience and the 2018 imagined audience) resembles the distinction between the Occident (white) “Self” and the Eastern/Oriental (racialised) “Other” (Said 2003; Lau and Mendes 2012).

Along these lines, it could be argued that the rational and capable audience identified in the 1988/1992 debates has not been (completely) replaced by the racialised irrational and vulnerable audience of the 2018 PD, because this latter audience is not assumed to represent the whole population. Nonetheless, it is important to stress that (problematically) the idea of an enlightened, well-informed, and autonomous audience has indeed disappeared from the 2018 elite discourse. In other words, the traditionally liberal and democratic characteristics and features consensually attributed to audiences in the 1988/1992 debates are not merely contrasted with a newly emerged vulnerable audience, but are entirely missing from MP’s discourse in 2018. Whilst opponents of the Broadcasting Ban energetically argued in favour of public debate, praising the power of intellectual political argument and citizens’ capacity to discern between the good and the “evil”, the more recent debates, deeply rooted in contagion theory, displace and silence the democratic and liberal principles that oppose censorship and celebrate public deliberation by endorsing the rational capacity of audiences to play an active part in public affairs, including (violent) conflicts. Thus, this discourse-historical analysis of British PDs on terrorism and the media shows a clear illiberal turn in elite political discourse, revealed by the dramatic shift in the construction of imagined audiences in British elite political discourse. This illiberal/authoritarian shift is also reflected in the consensus on the adoption of media control policies that are dismissive of the media’s role in the public sphere and seek to curtail individual civil rights (including freedom of expression and access to information, see Liberty 2018; Article19 2018).

Research on imagined audiences also illustrates how MPs strategically employ different constructions of the audience to endorse or challenge the introduction of media regulations and, thus, raise the question of whether particular imaginaries of the audience are instrumentalised/rhetorically exploited in political discourse. This article not only contributes to a body of academic research that stresses the relevance of analysing PDs in order to understand the broader social and political processes, but makes an original contribution by comparing media control measures introduced in two distinct historical contexts in the UK. The analysis also demonstrates the usefulness of the concept of the imagined audience to analyse and compare elite political discourse and media policies. Thus, this article also makes an important conceptual and interdisciplinary innovation by employing the concept of the imagined audience to evaluate and compare the development of terrorism discourse and media control policies in the UK. This opens up an interesting line of research that could be applied to other countries and/or historical contexts. In other words, future research could analyse and compare the construction of audiences and their correlation with the introduction of (il)liberal measures internationally. Thus, research on the construction of audiences offers original and appealing tools to investigate the so-called illiberal and/or populist turn in democracies across the world (Diamond 2021).