The (de)legitimation of torture: rhetoric,shaming and narrative contestation in two British cases

Interrogation and torture in 1970s Britain

The British Army and the Royal Ulster Constabulary (RUC) police force used coercive interrogation against many detainees in Northern Ireland during the 1970s, while some police forces in mainland Great Britain were also accused of abusing terrorist suspects in their custody. In 1971, the British Army introduced five interrogation techniques for use in Northern Ireland, which it had developed in previous colonial counter-insurgency campaigns. These were: (a) hooding, (b) standing for hours at a wall with legs apart and arms high, (c) subjection to noise and sensory isolation, (d) a diet of bread and water and (e) sleep deprivation. Fourteen detainees were selected for exposure to the five techniques in 1971 (European Court of Human Rights, 1978: para. 96). It was usually police officers who carried out these coercive interrogations at centres operated jointly by the army and the RUC (ECtHR, 1978: paras. 92–115). A report by Amnesty International (1972: 41) concluded what the British government itself later stated in private – that the use of these techniques against detainees had amounted to torture (Carver, 2016: 110).

Although the government declared in 1972 that the five techniques had been discontinued, coercive interrogations and abuse of detainees continued in practice. Between 1971 and 1974 alone, approximately 2000 complaints of ill-treatment or assault were made against the police and the army in Northern Ireland. In total, 218 members of the security forces were prosecuted for assault between 1972 and 1977 and 155 were convicted (ECtHR, 1978: para. 140). The report of a detailed investigation by Amnesty International (1978) concluded that RUC mistreatment of suspects took place on a frequent basis. A public inquiry followed, which substantiated the allegations of mistreatment in a number of cases and recommended reforms (Bennett, 1979). A number of former RUC officers later admitted that coercive interrogation by the police had been widespread in Northern Ireland during the 1970s. They outlined the techniques used, which included beatings, punches to the centre of the stomach and forcing detainees’ wrists or elbows into painful positions, sometimes for long periods of time (Cobain, 2012: 182–183).

In Great Britain, meanwhile, there was substantial evidence that some English police forces coerced confessions out of terrorist suspects during the 1970s, particularly in notorious miscarriages of justice cases. Some of the Birmingham Six – wrongly convicted and later acquitted of carrying out deadly bombings in 1974 – alleged that they had been beaten while in custody. Doctors verified that the men had sustained extensive injuries, some of which were visible when they appeared in court. The judge at their trial also acknowledged that they had been ‘quite outrageously assaulted’ while in custody (Faul and Murray, 1976). Other victims of miscarriages of justice – the Guildford Four – also made credible allegations that members of an English police force had beaten and forced false confessions out of them (Ewing, 2006: 47–54). While not on the same scale as the abuse in Northern Ireland, such cases indicated that some police forces in Great Britain also carried out coercive interrogation of terrorist suspects during the 1970s.

Britain’s record on torture in the post-9/11 era

In contrast to the 1970s, by all accounts there have been no coercive interrogations of detainees within the UK in the period since 9/11. Jihadist terrorist suspects have not faced the intimidating and violent cells that many Irish detainees endured in the 1970s. The ‘vast majority’ of terrorist suspects arrested in Britain during the 2000s ‘maintain[ed] their right to silence’ according to Peter Clarke, a former head of counterterrorism at the Metropolitan Police. ‘The walls of Paddington Green police station rarely reverberate with outpourings of guilt and contrition’, he wrote (Clarke, 2008). However, there was no question of the police resorting to coercive interrogation in order to get suspects to talk.

Beyond the domestic sphere, Britain did become involved in torture after 9/11 through its collusion with the prisoner abuse of many foreign intelligence services, including those of the United States, Pakistan, Libya and Egypt. A parliamentary committee inquiry into the period 2001–2010 found evidence that in 232 cases the UK’s intelligence agencies continued to supply questions or intelligence to foreign services after they knew or suspected that a detainee was being mistreated. In 13 cases, British officials witnessed a prisoner being mistreated by others (ISC (Intelligence and Security Committee), 2018: 2–3). Indeed, UK officers allegedly interviewed some detainees after – or in between – periods of torture by a foreign security service. Britain was also complicit in other states’ rendition operations – the transfer of prisoners, outside of any legal process, to countries where the risk of torture was high. The parliamentary inquiry found evidence that in 28 cases UK intelligence ‘suggested, planned or agreed to rendition operations proposed by others’. In a further 22 cases, the UK ‘provided intelligence to enable a rendition operation to take place’ (ISC, 2018: 3). In 2004, for example, Britain’s foreign intelligence service, MI6, provided intelligence which enabled US agents to abduct at least two Libyan dissidents and fly them into the hands of the Gaddafi regime, which subsequently tortured them. A senior MI6 official, Mark Allen, wrote to his Libyan counterpart to ‘congratulate’ him on one of these operations and request access to the resulting intelligence (quoted in Blakeley and Raphael, 2017: 255). Overall, as the High Court later put it, the UK had ‘facilitated’ abusive interrogations by foreign security services through providing questions, intelligence and other cooperation (UK High Court, 2009: para. 88). Britain colluded in these coercive interrogations with a view to gaining the intelligence produced by them.

Alternative explanations

While this article emphasises the role of anti-torture narratives, Blakeley and Raphael’s extensive analysis of UK involvement in torture after 9/11 offers an alternative explanation, which depicts Britain as a ‘petty sovereign operating. . . under the aegis of US exceptionalism’. There was ‘no need’ for the UK to carry out torture itself because the United States had established an extensive torture infrastructure across the world (Blakeley and Raphael, 2017: 243, 251, 261). Britain could simply outsource torture to others and receive intelligence from them. This, it may be argued, reflects the nature of contemporary international terrorism. One could also extend this argument to conclude that, in contrast, when facing the Irish terrorism of the 1970s, Britain carried out torture because it could not outsource its dirty work on this domestic threat to any other country.

While this argument tells an important part of the story, it does not provide a full understanding of the British case because contemporary jihadist terrorism which targets the UK is as much a domestic phenomenon as it is international. Attack planning has been carried out mostly by British nationals based in the UK (Foley, 2013: 29–31). US or other countries’ intelligence is very useful against such threats – but it is far from sufficient. The need for the British government to generate its own intelligence on suspects’ activities within the UK far outweighs its need for information from foreign partners. During the 1970s, it used torture as a means of gaining such intelligence. In the 2000s, it did not. On 21 July 2005 for example, soon after the 7/7 London bombings, a second plot came very close to causing explosions and loss of life in London once again. The five men responsible escaped and a desperate manhunt for them ensued. When the first of these men were captured, British police officers did not use coercive interrogation to find out the locations of their accomplices. A focus on outsourcing does not help us to understand why.

A second explanation would focus on specific changes to domestic law. Undoubtedly, this did play a role in the evolution of the UK’s approach to interrogation. The Police and Criminal Evidence Act 1984 (PACE) introduced the audio tape recording of detainee interviews and strengthened suspects’ right to have a lawyer present with them in police stations. Over time, PACE created a context in which English lawyers were able to defend detainees’ rights effectively in the initial detention period after arrest and provided practical protections against coercive interrogation. (Carver, 2016: 118–119). Notwithstanding its importance, domestic legislation can be changed and it does not offer an adequate explanation of the UK’s record on interrogation after 11 September 2001. After all, the government pushed through legislation in 2000 and again after 9/11 that departed from ordinary criminal justice procedures in some respects, such as extending the length of maximum pre-charge detention in terrorism cases (Foley, 2013: 183). There was no legal impediment to the government extending this special regime so as to also deprive terrorist suspects of the PACE protections covering the interrogation process. Indeed, countries such as Spain do precisely this in terrorism cases. I argue below that to understand why the UK did not take this step, we need to consider the influence of certain anti-torture narratives on the British government during the 2000s. Let us now turn to consider the theoretical expectations which underpin this argument.

Narrative contestation

While rhetorical coercion can have significant effects, it is often deeply contested as target states make extensive efforts to resist such attempts to influence them (Friman, 2015). ⁵ In this process, accusers’ messages and state counter-messages often take a narrative form. Scholars in the field of IR are increasingly bringing this concept into their research on a wide range of topics (Krebs, 2015). In essence, a narrative is a framework that allows humans to connect various phenomena around some kind of causal chain. Causality and causal claims are central to any narrative. The end point of the causal chain – the claimed outcome – is crucial for giving meaning to the previous links in the chain (Miskimmon et al., 2013: 5). A desirable outcome casts a positive light on the decisions, actions and events that are said to have led to it. A negative outcome can delegitimate many if not all of the actions that preceded it. Narratives are also ‘necessarily selective’, as Krebs (2015: 11) puts it. To sustain its thread, a narrative must highlight some elements and leave others out of the story. This opens an opportunity for actors who seek to deploy narratives to achieve a political effect. The state and its opponents thus deploy rival narratives in an attempt to sway relevant audiences. This narrative contestation involves contrasting interpretations of what the relevant setting is, which actors should be included in the story, how their purposes should be characterised, how their actions and means should be depicted and what the causal effects of their actions are (Krebs, 2015: 12).

What conditions shape the outcome of narrative contestation? Which factors have the most important influence on whether rhetorical coercion is effective in producing the desired outcome? (i.e. a situation where the target refrains from or significantly reduces its involvement in the ‘illegitimate’ behaviour concerned). Two factors, emphasised in the literature, are the authority and credibility of the accuser(s) and the relative clarity of the standard of legitimacy used by them. I outline these factors below and show how they are illustrated well by the two British cases examined here. The IR literature on rhetorical effectiveness also emphasises that we need to pay attention to the content of the rhetoric being transmitted in order to understand the outcomes of political struggles (Goddard and Krebs, 2015: 29). When analysing rhetorical coercion on the issue of torture, it is useful to pay attention to the multiplicity of narrative content.

Anti-torture narratives: ethical, utilitarian and shaming rhetoric

Opponents of torture in democracies tend to rely a great deal on four types of narratives. Though they have varying effects, all four are considered here as potential instances of rhetorical coercion. First, ethical claims have been central to the case against torture. The argument that torture is morally wrong is often developed through narratives which focus on the pain and suffering caused by coercive interrogation (Del Rosso, 2015: 140–142; Wisnewski, 2010).

A second prominent type of anti-torture narrative has been one which uses utilitarian rhetoric to depict coercive interrogation as a tactically ineffective method of gaining intelligence. Detainees may say whatever they believe will make the torture stop, including falsehoods, and in any case standard interrogation techniques offer a more effective way of gaining accurate information (Del Rosso, 2015: 138–144; Wisnewski, 2010: 153–157). The third type of narrative also takes a utilitarian approach, this time to depict torture as strategically ineffective and counterproductive. It highlights the broader effects of torture on a counterterrorism campaign, suggesting that if the state is known to torture, that will alienate key sections of the population, increase support for the terrorist group and help it to recruit more people into violence (Del Rosso, 2015: 145–148; Ralph, 2013: 118–130). Claims to effectiveness form an important part of governments’ efforts to legitimate their policies. Consequently, narratives which cast doubt on effectiveness need to be engaged with and are an important contestation ground for rhetorical struggles on countless issues, including torture.

The fourth type is a narrative on how torture contradicts states’ international commitments and damages their reputation. By drawing attention to violations of international legal commitments, this narrative is related to the practice of shaming, which is the subject of an extensive literature covering torture and many other issue areas. Shaming is commonly understood as the public exposure and condemnation of a target state for illegitimate actions which violate its own international commitments, done with a view to damaging its reputation and thus motivating a change in behaviour (Friman, 2015: 143–144; Schimmelfennig, 2001: 64). Scholarship on the politics of shame relies on a number of premises. On the one hand, it is assumed that international norms or standards of legitimacy have some influence on states, which have publicly committed themselves to upholding these standards. As members of regional and international communities, political actors care about their reputation and the perceived legitimacy of their behaviour. On the other hand, however, actors are only ‘weakly socialised’ into the community. International norms are not necessarily internalised by actors; norms do not directly motivate their behaviour and state officials may regard such norms as mere external constraints (Schimmelfennig, 2001: 48, 58, 62–63, 65).

This is where shaming comes in. It is through being shamed that the targeted state may be brought to focus on, and perhaps come closer to fulfilling, its own professed standards of conduct. The organisations or governments, which drive this shaming effort, direct their messages to several audiences. They focus on those who are responsible for the illegitimate behaviour, such as political leaders and senior government and security officials of the target state. At the same time, they also direct their messages to other elements of the political class, media or civil society in the targeted country, along with key constituencies in allied countries and broader international audiences (Vennesson, 2014: 31). The resulting process of shaming can work in at least two ways. Some of the state officials directly responsible for the objectionable deeds may feel ashamed following this public exposure and condemnation, leading them to push for changes in their government’s behaviour. Yet feelings of shame are not necessary for shaming to work. Even officials who see little or nothing wrong in their government’s actions may still be concerned about the negative effect of this public shaming on their reputation among the various domestic and international audiences where the message is resonating (Schimmelfennig, 2001: 64). That, by itself, can be enough to stimulate changes in behaviour.

As noted in the introduction, this article provides the first qualitative analysis of the effects of shaming on a state’s degree of involvement in torture. The analysis, offered below, indicates that the literature needs to pay more attention to the interaction between shaming and other types of rhetorical coercion. Busby and Greenhill (2015: 106) examine how shaming can be combined with material sanctions to influence target states. They suggest that ‘shaming by itself is. . . a weak tool’. If this is true, it is equally important to analyse how shaming can be combined with other kinds of rhetoric to have an impact on target governments.

Dynamic responses to rhetorical coercion

One contribution of this article is to develop a theoretical analysis of the rhetorical constraints on democracies’ involvement in torture. To understand these constraints fully, we need to focus attention on the target state. After all, research on shaming and socialisation has been criticised for treating these phenomena as a one-way process in which the accuser impacts on the target state (Adler-Nissen, 2014: 150–151). Building on this critique, this article speaks to the literature which analyses shaming effectiveness primarily from the perspective of the target (Adler-Nissen, 2014; Vennesson, 2014). This focus gives rise to two more specific points on the dynamic, ‘socio-strategic’ response (Mantilla, 2018: 319) of the target state to rhetorical coercion and constraint. First, I argue that greater attention needs to be paid to the multiplicity of narrative content and the resultant complexity of contestation. The previous section illustrated how accusers propagate multiple narratives simultaneously. However, this opens opportunities for rhetorical manoeuvre by the target government – to accept some critical narratives, while simultaneously contesting others in an attempt to circumvent the full implications of the overall norm at stake. Many different narrative combinations could result from this multifaceted contest. However, as some critical or shaming messages fail and others prove effective, the result is a specific configuration of prevailing narratives, which enable and constrain target governments in particular ways. The specific configuration of narratives accepted by the British government during the 2000s, for example, is crucial for understanding the particular way in which it sought to both avoid torture and collude in torture.

The second point is to show how target states react dynamically to shaming and rhetorical coercion – even when they do not resist them. Petrova (2016) analyses NGOs’ positive altercasting of a target state, showing how a government may in part be praised into meeting NGO demands. Her analysis indicates how social influence rarely works through pure shaming or disapproval alone. The analysis developed in this article explores another way in which a ‘positive’ narrative can be developed to legitimate government action. It shows how a target government can creatively recontextualise the critical or shaming narratives of its accusers. Here, the target accepts some of its opponents’ criticism – but recasts these points in the context of a positive overall narrative for itself. ⁶ While this manoeuvre allows the target government to ‘look good’ to relevant audiences, it may simultaneously reinforce its own rhetorical entrapment and constrain its subsequent behaviour, as will be illustrated below.

Typology and methods

The preceding discussion drew on secondary literature to identify four key anti-torture narratives. In the empirical research for this article, I examined the rhetoric of opponents of British torture, observing the presence or absence of these four narratives across a range of sources from the 1970s and 2000s, including reports by NGOs and other activists, court proceedings and public inquiries. Thus, a theoretically informed but empirically grounded typology of anti-torture narratives was constructed. Though they usually take a narrative form in public discourse, the four parts of the typology can be stated briefly as follows (with each one then classified according to the type of rhetoric used):

The following sections show how these four narratives have been deployed by various actors for the purposes of influencing the British government’s position on torture. I analyse the effects of this rhetorical coercion on government discourse, tracing the varying salience of the four anti-torture messages in UK government narratives across the two cases. The result was a different configuration of prevailing narratives in the two time periods, which I argue is key for understanding the British government’s varying degrees of involvement in torture in the 1970s and after 9/11.

The methods used for this analysis were twofold. First, a qualitative analysis of government documents was carried out. Relevant sources were selected according to three criteria that they contain ‘clear articulations’ of the position of the British government, have the ‘formal authority’ to define this position and are ‘widely read and attended to’ (Hansen, 2006: 76). I identified patterns in the ways in which involvement in torture was legitimated and delegitimated in government sources, including speeches and interventions in Parliament, strategy documents, official reviews of operations and internal communications. ⁷ On this basis, I develop an account of government narratives (I do not seek to demonstrate here that these are broader elite or societal narratives). While this first method was a qualitative analysis of heterogeneous documents, the second method used was a content analysis of a homogeneous data source. A research assistant coded speeches and interventions in Parliament by government ministers during 10 key debates on torture from the 1970s and 2000s in order to observe how many times the 4 key narratives appeared in each time period. This resulted in 52 discursive observations, allowing us to compare the salience of the narratives across the 2 cases (see online supplemental file for further detail on both methods).

1970s: narrative contestation

As allegations of torture by British security forces in Northern Ireland emerged during 1971, those who expressed condemnation or criticism included not only civil rights and Irish nationalist activists in Northern Ireland but also the Irish government in Dublin, elements of the British media, some opposition Members of the UK Parliament and Amnesty International (Newbery, 2015: 86–90). The British government established a three-person committee to conduct a review of UK interrogation policy – the Parker Inquiry. The committee’s majority report, authored by Lord Parker and John Boyd Carpenter, recommended permitting the security forces to use their ‘five techniques’ of coercive interrogation. However, the third member of the committee – Lord Gardiner – came out strongly against the five techniques. He and the other critics mentioned above deployed four main narratives in their efforts to shame or influence the UK government to move away from coercive interrogation.

Post-9/11: narrative reconfiguration

While the British authorities had contested all efforts to shame or criticise them during the 1970s, their response to these narratives evolved in the following decades. When the debate about torture once again intensified in the years after the September 11 attacks on America, important parts of the UK government and security establishment continued to disregard two of their accusers’ main narratives, regarding the ethics and tactical effectiveness of torture. However, the other two narratives, concerning international reputation and strategic effectiveness, were essentially accepted by that time across most parts of the British government and its security agencies.

Rhetorical manoeuvres: avoiding torture, colluding in torture

The British case illustrates well the dynamics of rhetorical manoeuvre in a context of multiple narratives. As outlined above, opponents of torture did succeed in manoeuvring the government onto more favourable rhetorical terrain for them. Thus, the broad acceptance within government of two anti-torture narratives – concerning strategy and reputation – explains much about why UK intelligence and security officials generally refrained from perpetrating torture after 9/11. Unlike their American counterparts, there were no circumstances in which British officials could publicly defend an ‘enhanced interrogation’ programme of their own. They had accepted that the public effects of such a move would be ‘dire’, as Omand put it.

However, British officials resisted the narratives which emphasised the ethical impermissibility and tactical ineffectiveness of torture. Their rhetorical manoeuvre was to accept two anti-torture narratives while simultaneously contesting two others in an attempt to circumvent the full implications of the international norm against torture. As the shaming literature would expect, this norm was not internalised by UK officials. Instead, they varied their behaviour according to their perception of the likelihood of exposure. They did not perpetrate torture themselves – but they did collude in the violations of other states.

Collusion flowed directly from the particular configuration of narratives accepted by the UK government. They told themselves that torture could be morally acceptable and tactically beneficial in gaining intelligence. These two narratives were concerned with the private realm of torture – the ethics and ‘secret’ tactical ‘benefits’ of what happens inside the interrogation room. In the narrative world of British officials, the downsides of torture were only in the public realm (if it was exposed to the public) – its negative strategic effects on certain constituencies and how it would damage the UK government’s international reputation. The question for officials therefore was which course of action was more likely to lead to public exposure? If the state perpetrates torture, it places itself very close to the ‘scene of the crime’. And the UK’s coercive interrogations of the 1970s did lead to the government being condemned in court. In this context, collusion was an attractive option for UK officials because it allowed them to increase the distance between themselves and the act of torture, while still being able to reap some of its perceived tactical benefits. They believed that it would be possible to maintain a covert collusion in torture, which was carried out not by them but by other governments in distant lands, without this involvement becoming public knowledge (Aldrich, 2011: 162–163; Cobain, 2009). If the fact of collusion never became publicly known, the British government would never have to suffer the negative strategic or reputational effects of its involvement in torture.

In the event, however, British collusion did not remain covert. It was uncovered by investigative journalists and activists who sought to shame the UK government for its links to torture by foreign security services. It has been the subject of parliamentary committee inquiries, police investigations, and an aborted judge-led inquiry (Cobain, 2012: 264–271; ISC, 2018: 8–11). In the context of this shaming and controversy, in 2010 the government issued a new, more restrictive policy governing the intelligence agencies’ liaison with foreign services. Since 2010, there have been far fewer allegations that British officials have colluded in torture, compared with the 2001–2010 period (ISC, 2018a: 1, 25).