This article offers a new analysis of the Northern Ireland peace settlement through an examination of the pivotal relationship between two key actors: the British state and the Provisional Republican movement that included Sinn Féin and the IRA. It traces the negotiating relationship between these key parties and argues that the ending of violent conflict in the 1990s can best be understood as the outcome of a long bargaining process between these two actors that was conducted both tacitly and explicitly over a span of more than two decades. It concludes that the development of a cooperative relationship between the British state and the Provisional leadership and the active coordination of British policy and republican strategy were the crucial elements in securing an end to violence in the 1990s.

This article offers a reassessment of the pivotal relationship between two key actors in the Northern Ireland peace settlement: the British state and the Provisional Republican movement that included both Sinn Féin and the Irish Republican Army (IRA). It argues that the final cessation of the IRA campaign in the 1990s and the inclusion of the Provisional Republicans in a subsequent peace settlement can best be understood as the outcome of a long bargaining process which was conducted both tacitly and explicitly over a span of more than two decades. From the beginning this process necessitated a degree of choreography and partnership, most importantly in the ongoing cooperative project to conceal secret contacts from other parties.

The focus on these parties is justified on the basis that only these two actively and persistently contested the question of sovereign control in Northern Ireland. Despite the huge gap between them in terms of resources, organisational complexity, democratic mandate and international and domestic legitimacy, they had in common the fact that both asserted claims to a monopoly of the legitimate use of violence, claims that were mutually exclusive. The eventual establishment of stable government in Northern Ireland was dependent on the successful conclusion of negotiations between all significant parties – and it was the two main unionist parties rather than the Provisional Republicans that exerted the dominant shaping power on British policy throughout the Troubles and the peace process. Nonetheless, the ending of violent conflict was crucially dependent on reconciling these two competing assertions of legitimacy and sovereignty.

The changing relationship between the British state and the Provisionals over the course of 30 years of conflict is crucial to understanding the peace settlement in Northern Ireland in the 1990s. This relationship is poorly understood and is characterised quite inaccurately in much of the academic literature. This is partly because it was in the shared interests of both parties to obscure key elements of that relationship and of their respective policies for propaganda purposes (Dixon, 2001, p. 364).

Much work has been published on British policy (Cunningham, 2001; Dixon, 2001; Neumann, 2003a; O’Leary, 1997) and on Provisional Republican strategy and politics (Bew and Patterson, 1985; English, 2004; Moloney, 2002; Mulholland, 2007; O’Leary, 2005; Smith, 1997) but there is relatively little work that focuses directly on the relationship between the two parties (Bean, 2008; Bew et al., 2009; McIntyre, 1995). Many accounts of the Northern Ireland conflict focus primarily on the British state or on the British and Irish states, as the key actors in the peace settlement of the 1990s (Bew et al., 2009; Kerr, 2006; O’Donnell, 2007; O’Duffy, 1999; 2000; O’Kane, 2004; Tannam, 2001). This reflects the blunt reality that the two states, but above all the British state, were the key power actors in the situation and had more influence over the outcome than any other actor. It tends however to obscure the agency of the Republican movement in the ending of conflict. In most of these state‐centred accounts the Republicans are ‘enticed’ into politics (O’Kane, 2004, p. 78), they are led away from violence by a ‘carrot and stick’ approach (Bew et al., 2009) or they are ‘cultivated … into the constitutional fold’ (O’Duffy, 2000, p. 411). The peace process is characterised as a process of ‘bringing in the rogues’ (Neumann, 2003b). When it comes to explaining violence the Republicans are credited with agency, but when it comes to the pursuit of peace they are represented as acted upon – pushed and pulled into peace.

The British state becomes the key agent in some of these accounts not only because it is far and away the most powerful actor, but because we are invited to view the conflict through the eyes of the state and to ground our analysis in an uncritical acceptance of state legitimacy (see, e.g., Dingley, 2009). The issue of violent conflict is repeatedly characterised from the perspective of state actors as a policy dilemma, rather than a contest over sovereignty in which the state’s claim to legitimacy was not entirely unproblematic. Writing from a position of close identification with the perspective of one key actor makes it difficult to discern some of the crucial dynamics operating at the intersection between these two key actors.

One reason why this intersection is misunderstood is that both parties had strong motivations to ensure that aspects of their positions remained obscure during the course of the conflict. As Jonathan Powell, one of the key British negotiators in the 1990s has noted, ‘Because of the secrecy of the IRA and the republican movement, it was pretty much impossible to know what their bottom line was’ (Spencer, 2010, p. 440). Parties to negotiation carefully guard information about their negotiating strategy and their positions, and in particular about the agreement they would find minimally acceptable, their ‘resistance point’, because this information can provide an advantage to the other party. It is simply bad negotiating practice to let it be known how much you are willing to concede (Walton and McKersie, 1991, p. 54, p. 61, p. 63, p. 83). It has frequently been argued that the Provisionals did not understand the political dynamics of the situation and the constraints on the British state and that the British state did not understand Irish republicanism. The argument is made in this article that in certain respects these two parties understood each other much better than anyone else did. Key British officials were able to see past the distorted public image of the Provisionals that the British state itself did much to sustain while the Provisionals had a relatively well‐developed understanding of the constraints under which the British state operated. Focusing on the intersection between the parties allows us to develop a much better understanding of both British policy and republican strategy.

How then were the positions of these two key actors eventually reconciled? This article analyses the negotiated termination of the IRA campaign in the 1990s as the outcome of a long‐term bargaining process between these parties that stretched from 1971 onwards, a process in which short bursts of direct negotiation were separated by long periods of tacit bargaining. The article challenges existing explanations of the ending of the conflict and argues that it can best be understood as the outcome of the active coordination of British policy and republican strategy to arrive cooperatively at a peace settlement. It emphasises the agency exercised by these two key actors. This is not to suggest that we can understand the settlement as a mechanistic process predetermined by the rational interests of these actors. The pattern and pace of this bargaining was decisively shaped by contingent events, by ideology, by shifting power relationships within the parties, by the pressures generated by other key actors and by the medium‐term shifts ‘in the communal power balance’ identified by Joseph Ruane and Jennifer Todd (2007, p. 453).

The article draws on interviews with key actors on both sides who were active at this intersection between the two parties in both the 1970s and the 1990s, and on extensive interviews with and the private papers of Brendan Duddy, the key intermediary between the parties in the 1970s and again in the early 1990s. It draws too on recently released papers from the UK National Archives and on the private papers of key Republican leaders Ruairí Ó Brádaigh and David O’Connell.1

Long War, Long Negotiation

Understanding the engagement between the British state and the Republican leadership in the 1990s requires first a reassessment of the ‘long war’ strategy that the IRA adopted in 1976 and the parallel policy of normalisation adopted by the British state. The Provisionals’ long war is almost invariably portrayed as a strategy aimed at securing victory through a sustained campaign that would eventually produce British withdrawal (English, 2004, pp. 212–5; Patterson, 1990, p. 11; Smith, 1997). It is also useful, however, to analyse it as a bargaining move aimed at pressuring the British government to re‐engage in negotiations with the Provisionals. To understand this aspect of the long war we need first of all to reject the existing academic consensus on the IRA ceasefire of 1975 which states that the IRA called the ceasefire because they were fooled into believing that Britain was about to withdraw. This interpretation obscures the fact that the Republican leadership attached a high priority at that stage to achieving a peaceful settlement of the conflict and were willing to make major compromises to that end. Two points are of particular importance in this regard. The first is that the Provisional leadership was extremely reluctant to resume its campaign. The ceasefire lasted as long as it did primarily because of this fact rather than because the Provisionals were ‘duped’ (Ó Dochartaigh, 2011a). The second is that the British government was fully aware of this reluctance and withdrew from substantive engagement partly because of this knowledge. If the Provisionals were weakening and if they were reluctant to restart their campaign, then it did not make sense for the British government to incur the costs involved in negotiating a settlement with them. These two points are captured succinctly in the account by Brendan Duddy, the intermediary, of a conversation with his British interlocutor, Donald Midleton (DM) as the British government withdrew from engagement with the Provisionals in early 1976:

I said ‘the IRA want peace, but we need a way out of this war. Let us talk and somehow we will end all of this killing’. DM agreed but said ‘if the IRA is losing support, why should we aid them by giving them a way out?’ (1976 Diary, 8 May 1976, POL 35/131, Duddy papers).

The 1975 ceasefire negotiations had demonstrated to the Provisionals that they did not provide a sufficiently serious threat to the British state to achieve a negotiated compromise settlement, given the countervailing pressures from loyalists, unionists, the security forces, the parliamentary opposition and the Irish government. The experience indicated the very limited extent of the leverage the movement enjoyed. It also indicated that conveying a strong Provisional willingness to end violence had the perverse effect of providing incentives for the British government to minimise movement towards the Republican position, given the pressures against engagement that could be exerted by other actors. The reluctance with which the Provisionals ended the 1975 ceasefire is evident in British reports of the secret talks, most strikingly in their account of the attitude of the top Belfast IRA commander Billy McKee at the final secret meeting between the parties:

In an emotional (but not angry) outburst, McKee said that the violence in Northern Ireland, from which people of both communities had suffered for over 6 years, was destroying both the physical and spiritual qualities of life; the people had nothing to live for, and they looked in dread at the prospects ahead for their children (Spun Sugar 10: Meeting with O’Brady [Ó Brádaigh], McKee and McCallion, 10 February 1976, PREM 16/960, UK National Archives).

Noting that ‘The Provisional Sinn Féin representatives spoke calmly, … pointedly avoided making provocative statements … and appeared to welcome the opportunity to meet our men’, they paint a picture of a leadership that did not wish to go back to a war of any kind, let alone embark on a twenty‐year war. The long war was a strategic plan but it was also a bargaining position aimed at pressurising the British government to return to substantive engagement despite the relative military weakness of the Provisionals. The British would return to engagement not because the IRA would or could intensify its violence to the point necessary to defeat the British militarily but because it would persist.

One of the earliest pieces of documentary evidence for the long war strategy is an IRA planning document from 1977 which asserts that ‘We must gear ourselves towards Long Term Armed Struggle’ (O’Brien, 1999, p. 109). Talk of victory or of forcing Britain to withdraw is entirely absent from this document which notes instead that new security policies were ‘contributing to our defeat’. Far from being a recipe for victory it was aimed at ensuring the survival and persistence of the IRA (Moloney, 2002, p. 150). The optimum outcome of the long war was a return to talks after a few years. This was the hope of the intermediary Brendan Duddy, who says that for two years after the closing of the back‐channel he expected the British to re‐establish contact. Writing under the pen name ‘Brownie’ in August 1976, Gerry Adams, the most important figure in the emerging leadership, talked of a long conflict that might continue until perhaps 1983, rather than for twenty years (Moloney, 2002, p. 150). The appointment of Roy Mason, a hard‐line former Minister of Defence as Northern Ireland Secretary of State in 1976, the election to power of Margaret Thatcher in 1979 and the escalation surrounding the 1981 hunger strike were among the most important factors that worked against an early return to engagement. In addition, a series of particularly provocative IRA attacks, including the killing of the British Ambassador in Dublin in 1976 and of Lord Mountbatten in 1979, targeted the British political establishment directly. They made it significantly more difficult to argue internally for any move to re‐engage with the Provisionals.

Against this it might be argued that the Provisionals repeatedly proclaimed their determination in the late 1970s and 1980s to achieve military victory through a long struggle that would ‘sicken’ the British government into withdrawal. They also took up a more rigid bargaining position, hardening up their demand for British withdrawal (Moloney, 2002, pp. 170–1). While not discounting these moves entirely, we need to be aware that this rhetoric does not provide an accurate guide to the strategic thinking and expectations of the leadership, even if some within the movement were convinced that eventual victory was possible. The rhetoric of a long struggle towards victory was necessary not only to convince the British government of the movement’s willingness to continue on an open‐ended basis, but also to motivate volunteers to renew the struggle after the experience of a ceasefire. In addition, there was little to be gained politically from moderating their bargaining position in the absence of any possibility of engagement.

The old leadership had learned in the course of their 1975 talks just how extraordinarily difficult it was to secure movement from the British state, even on minor, peripheral issues such as transfer of prisoners or release of internees. The emerging leadership that would replace them in the 1980s was heavily involved in attempts to negotiate a resolution of the prison dispute from 1978 onwards and found it similarly difficult to secure any movement (Adams, 2003, p. 11; Clarke, 1987, pp. 103–7). Given their shared experience of the difficulties of securing gains of any kind in negotiation with the British state, it seems unlikely that either the existing or the newly emerging leadership ever envisaged a moment when they would finally impose their demands on the British government.

As with many negotiating threats, there is the great drawback that carrying them out carries huge costs for those who make them (Schelling, 1980, pp. 35–6). The costs of the proposed long war were so high for the Provisionals that it is understandable that the British government would test them on their capacity and willingness to carry out this threat, particularly since the IRA had seemed so eager for peace in 1975 (Cowper‐Coles, 2012; Kerr, 2011, p. 280). The British government’s termination of all communication with the IRA in 1977, as part of the broader policy of normalisation, can also usefully be interpreted as a bargaining move. As Thomas Schelling emphasises in his classic text, The Strategy of Conflict, making oneself unavailable for negotiation can be an effective bargaining strategy:

Threats are no good if they cannot be communicated … When the outcome depends on coordination, the timely destruction of communication may be a winning tactic … If the threatened person can be unavailable for messages, or can destroy the communication channels, even though he does so in an obvious effort to avert threat, he may deter the threat itself (Schelling, 1980, pp. 38–9, p. 146).

Disengagement from contact sent the message that the British government would refuse to receive any threats or offers communicated by the Republicans. They would not engage with them at any level, would concede nothing and would take the necessary steps to counter their campaign in military terms. The government was calling the IRA’s bluff, so to speak. This was a hardball negotiating position that tacitly required the IRA to end its campaign unilaterally, an outcome for which there was a recent precedent in the IRA’s abandonment of the border campaign in 1962. There are indications that the British government very nearly succeeded in achieving this and that IRA leaders seriously considered unilaterally bringing an end to the campaign in 1975/6 (Taylor, 1998, p. 198).

If we think of the IRA’s long war strategy as a bargaining move, aimed not at achieving victory but at pressuring the British government to re‐engage, the political moves made by the Provisionals in the 1980s can be understood, not as a major change of direction, but as attempts to find additional or alternative ways of exerting pressure towards the same goal. Republican initiatives to engage with the Social Democratic and Labour Party (SDLP) and Fianna Fáil in the 1980s (O’Donnell, 2007) can usefully be interpreted as attempts to alter the negotiating balance, as bargaining moves aimed at generating British engagement just as the long war strategy aimed to do. To those who argued that all that was necessary to generate such engagement was for the IRA to end its campaign of violence, the Republicans could point to the experience of 1975 and argue that the ending of the IRA campaign had generated neither British movement towards a negotiated compromise nor an end to violence. The experience of 1975 taught them instead that ending their campaign removed pressure for movement towards the republican position by the British government while unionist and loyalist pressure against compromise was not only maintained but intensified.

And we can begin to think of the growth of Sinn Féin in the 1980s not primarily as an attempt to augment the armed struggle that ultimately had to give way to the logic of electoral politics because of the contradictions between the two, but as a major innovation that strengthened the bargaining position of the Provisionals. One of the few writers to identify correctly the strategic logic underlying the expansion of Sinn Féin in the 1980s has written:

The thinking of the leadership was that, whatever else, the movement had to remain strong enough to become part of the ultimate political solution when the time came … that meant getting into elections, maximising their political support north and south, to arrive finally, at the negotiating ‘table’ with the strongest possible mandate. But this in itself was a significant admission, indicating that the IRA on their own were not able to beat the British out of Ireland (O’Brien, 1999, p. 119).

Ultimately these innovations in the 1980s augmented the pressure on the British government to negotiate a settlement that the Republicans could live with and that included them. This is not to say that there was unanimity over the conditions under which they would enter negotiations. The precise circumstances under which the leadership re‐engaged with the British government in the 1990s was shaped by intense internal struggles and many within the movement were deeply disappointed with the settlement that was eventually reached.

We can only understand the engagement between Republicans and the British government in the 1990s if we see it in the context of a long‐term negotiation between the parties that was dominated by long periods of tacit bargaining, in which, as Schelling (1980, p. 21) puts it, ‘adversaries watch and interpret each other’s behaviour, each aware that his own actions are being interpreted and anticipated, each acting with a view to the expectations that he creates’.

A major dilemma for the Provisionals in the late 1980s and 1990s was how to maintain a British expectation that the IRA campaign could and would continue in the absence of substantive engagement, while simultaneously signalling willingness to accept a compromise agreement. Republican manoeuvring in the early 1990s has to be understood as that of a movement whose bargaining position depended to a great degree on its ability to communicate a willingness to continue the IRA campaign indefinitely, even if it was its firm intention to bring it to an end.

The Wrong Project: British Policy towards the IRA

As the British government closed off channels of communication with the Provisionals after the 1975 ceasefire and moved away from attempts to secure an agreed settlement, security measures expanded to fill the policy vacuum. British policy from 1976 onwards was aimed at pressing the military advantage and pushing the IRA towards a unilateral abandonment of its campaign. This sense that victory was close was vividly expressed by Margaret Thatcher in her comments during the 1981 hunger strike that the IRA might be playing its last card. The struggle over the prisons in the late 1970s which culminated in the hunger strike of 1981 can be usefully understood from this perspective, not as an inevitable consequence of the simple enforcement of the rule of law but as part of a policy of intensifying the pressure on the Provisionals on all fronts.

Before the orthodoxy of the new security approach had become firmly established there was a last flurry of debate about the wisdom of excluding the Provisionals. This debate was associated with the MI6‐dominated office at Laneside which had overseen back‐channel communication with the IRA and with loyalists in the 1970s. This office would shortly be abolished by a hostile new secretary of state. The debate illustrates the strong awareness of some within the state apparatus of two underlying factors that would be crucial in informing the turn that British policy took in the late 1980s. The first was that the Provisionals were actively seeking to bring an end to their campaign and were willing to negotiate a compromise solution which would involve major concessions on their part. The 1975 negotiations had taught this lesson. The second was that, contrary to the British state’s propaganda, the Provisionals enjoyed sufficient support and were sufficiently deeply embedded in the nationalist community that they would continue to be a significant force. They would not fade away. The implication of this latter understanding was that a settlement without the Provisionals was no settlement at all.

In an analysis that pushed back against the gathering consensus on defeating the Provisionals, one civil servant wrote in a paper on the republican movement in May 1976:

Unless we take more determined steps to involve the leaders of the Republican tradition in political life, the formation and execution of a coherent long‐term political strategy will fail … if we are to give real encouragement to the republican movement to pursue their aims politically – and now is as good a time as any in view of their reported disillusion over the lack of success of their military campaign – then some such statement [a declaration of some kind] is required … such a policy could finish the SDLP but that would be a small price to pay for peace … a politicised Provisional Sinn Féin would be more likely to produce political stability throughout Ireland as a whole than the continuation of a terrorist movement, however isolated. It is in our interest to see a strong Provisional Sinn Féin, if at the expense of the SDLP, so that the extremists are brought into the mainstream of politics and are forced to act politically and in due course responsibly (‘The Republican Movement’, 5 May 1976, CJ4/1427, UK National Archives).

It is important to note that this was a ‘losing paper’, illustrating a perspective that would be almost completely marginalised within a few months. In the context of the broader shift away from engagement in 1976 this looks like the dying kick of a fading policy of engagement with paramilitary groups that was being criticised as a form of ‘appeasement’. From our vantage point in 2013, however, it has the ring of prophecy, a policy document that was far ahead of its time. The document indicates two clear understandings: that the Provisionals were prepared to accept a compromise settlement, and that a settlement without them would ‘fail’. The prognosis that a politicised Provisional movement would eclipse the SDLP also indicates an early recognition that the Provisionals were in tune with a far broader section of the Catholic population than was publicly acknowledged.

In 1976 these arguments were being made in the teeth of a dominant consensus that the struggle against the Provisionals could be won and that it was not necessary to engage with them. We get a flavour of this new consensus in a letter from the incoming Royal Ulster Constabulary (RUC) Chief Constable, Kenneth Newman to the head of the Northern Ireland Office and the General Officer Commanding (GOC) at around the same time in which he wrote: ‘My first priority must be to create a strong and efficient crime fighting machine designed to erode and ultimately overthrow the power of the PIRA’. He went on to envisage the IRA being ‘weakened and ultimately eliminated’ and looked forward to ‘the defeat of the PIRA’.2 The document is significant not only as an exemplar of the emerging consensus but also for the light it sheds on the intensely political character of the drive towards ‘normal’ policing.

For those within the state apparatus arguing for engagement, a policy based on excluding the Provisionals was just kicking the can down the road. One of the most powerful arguments underlying the policy shift in the late 1980s was this long‐established argument that any solution required inclusion of the Provisionals – that shutting them out was a recipe for permanent conflict. It was this understanding that underlay the shift in British policy in the late 1980s.

Changing Course

In late 1989 and early 1990 Northern Ireland Secretary of State Peter Brooke, working closely with a core group of senior civil servants in the Northern Ireland Office, made a carefully thought out and calculated move to re‐engage with the Provisionals. It marked a significant shift in British government policy, a clear and calculated decision to explore, for the first time since the mid‐1970s, the possibility of a political settlement that included the Provisional Republicans. The fact that the Republican leadership had been taking various steps and making statements in the late 1980s that suggested it was interested in a compromise settlement provided an important opening for some in the state apparatus to argue that it was worth re‐engaging. The first important text to be generated by this decision was Brooke’s ‘Whitbread speech’ of November 1990, a copy of which was passed in advance to the Provisionals. The final text was the outcome of extensive collective discussion. As Quentin Thomas puts it, ‘Peter Brooke was in charge and masterminded it and took all the key decisions but it was a team effort to work out how we should present things and what we should say and what our pitch should be’ (interview, Sir Quentin Thomas). John Chilcot reinforces this impression of a very carefully constructed attempt to engage the Provisionals for the first time in many years: ‘Yes that [speech] was very much the concern of all of us and carefully chosen adjectives and carefully omitted commas! (interview, Sir John Chilcot).

Brooke and his team embarked on this significant step with the full awareness and consent of Prime Minister Thatcher. As Brooke says, ‘There is no way in which I could deliver the Whitbread speech without Mrs. Thatcher being wholly aware of it. I mean absolutely total clearance’ (interview, Peter Brooke). The passing of this text to the Provisionals in advance of the speech marked the first high‐level back‐channel contact on the question of a political settlement since the mid‐1970s.

The fact that re‐engagement with the Provisionals began under the stewardship of Margaret Thatcher, the Prime Minister most associated with hard‐line rejection of the IRA, emphasises the pragmatic basis for this shift. During her early period in office Thatcher had expressed confidence that victory over the Provisionals was within reach. The conflict had then persisted for a further decade and, according to one Conservative MP, Thatcher admitted to him privately in the late 1980s that ‘she [has] no solution to the Irish problem, no glimmer of light – the first time he had heard her admit defeat on any issue’ (Lees‐Milne, 2008, p. 151, cited in Bew et al., 2009, pp. 104–5). The fact that current policy was simply not delivering made it easier to argue that it was worthwhile exploring the possibility of an inclusive settlement, even while Thatcher remained in office. Enthusiasts for the security‐first policy that aimed at the defeat of the IRA frequently pointed to the decline in killings in Northern Ireland since the mid‐1970s as evidence of the success of this approach, but it was a success that came at great cost. The violence might have been reduced but the commitment required to achieve this in the 1990s was just as high as it had been at the peak of the Troubles in 1972. In the early 1990s 30,000 armed soldiers and police officers were permanently employed in pinning down this small territory, almost exactly as many as at the height of the violence in the early 1970s (McGarry and O’Leary, 1995, p. 85). From the point of view of the British state the conflict was draining as much time and resources as it had ever done. The IRA carried out several high‐profile bomb attacks in London from 1990 onwards that inflicted huge financial costs (O’Brien, 1999, p. 162), including one bomb that was estimated to have caused more financial damage than the entire IRA campaign up to that point (Tonge, 2006, p. 117). The annual death toll might have been reduced but the financial cost of the IRA campaign remained as high as ever.

Underlying the moves to engage with the Provisionals was the argument that the existing policy of excluding them was simply not delivering. It had not worked. Quentin Thomas, a key senior civil servant involved in the formulation of policy at the time, articulates this position:

[T]here was an important shift from that position that involved marginalising Sinn Féin, demonising them even, to one which recognised that, well firstly that that project was unlikely to succeed and secondly, that it was the wrong project anyway … the project of marginalising Sinn Féin was unnecessary and probably doomed to fail (interview, Sir Quentin Thomas).

Excluding them would not work, but it was also ‘unnecessary’ because the Provisionals were amenable to a negotiated compromise. As Thomas puts it:

We for our part didn’t have any particular wish to marginalise the sort of politics represented by Sinn Féin which had its own distinctive characteristics, a certain millenarianism, a certain quasi socialism, a certain fascism if you like, anti clericalism and what we objected to and had difficulty with was the violence (interview, Sir Quentin Thomas).

That is, the political ideology of the Provisionals, despite its distinctive aspects, was not an insurmountable obstacle to their inclusion in a compromise settlement.

Two additional elements created the circumstances in which this realist argument became ‘consensual knowledge’ (Haas, 1990; Tannam, 2001). The first was the rise of Sinn Féin as an electoral force in the early 1980s. This presented a genuinely novel policy dilemma: how to deal with this newly strengthened political party. Engagement with the Provisionals could be represented not as a belated return to the policies of the early 1970s but as a strategic response to a new problem. The rise of Sinn Féin also made it easier in practical terms to consider engagement. Including the Provisionals in the 1970s had required finding a way to convert the IRA’s support into a political form. In the 1990s inclusion simply required inclusion of an existing political party. The second factor was dissatisfaction with the failure of the 1985 Anglo‐Irish Agreement to deliver a decisive military advantage against the IRA (Thatcher, 1993, pp. 406–7).

As John Chilcot, Permanent Undersecretary at the time and a centrally important figure in this initiative puts it:

[T]he security cooperation which was one of the elements of the deal with Garret3 [the Anglo‐Irish Agreement of 1985] hadn’t come through in any particularly noticeable or practical way and I think that was what forced the re‐think in the late ’80s, that that was not a track that was going to resolve the problem from a British perspective by itself (interview, Sir John Chilcot).

The Agreement was a joint project of the British and Irish states with a much wider remit than security cooperation but it had been sold to Thatcher very much on the basis of this element. It is somewhat ironic that dissatisfaction with the security aspects of the Agreement formed an important part of the subsequent argument for engagement with the Provisionals. Engagement could be advocated as a way to make progress on the broader strategic goal of moving beyond the Agreement. Explaining the move to engage with the Provisionals, Thomas emphasises that ‘the project underpinning the Agreement, which was an Irish initiative rather than a British one, was to marginalise Sinn Féin. So it was their project not ours’ (interview, Sir Quentin Thomas). Thus, the aim of marginalising and excluding the Provisionals could be viewed as a relatively recent and problematic position associated with the Anglo‐Irish Agreement rather than as a foundation stone of British policy. The failure to implement a security solution to the conflict by defeating the IRA could be presented as a regrettable failure of an Agreement that was strongly associated with the Irish government.

Much of the literature treats British government steps to engage with the Provisionals from 1989 onwards as a relatively peripheral development, very much secondary to the efforts to reach an agreement between the ‘constitutional’ parties. Brendan O’Duffy, for example, suggests that the preference of the British government at this point was for an agreement between the other parties that would exclude and marginalise the Provisionals, following the model of previous initiatives (O’Duffy, 2000, p. 409): ‘While using the secret dialogue to explore the republican movement’s willingness to accept a negotiated settlement, the British government was more hopeful that constitutional parties could reach agreement which would effectively marginalise the IRA’. But there is evidence to suggest that this dialogue represented a much deeper shift in British policy, a clear and calculated decision to attempt to reach a negotiated settlement that included the Provisionals for the first time since 1975.

The way in which John Chilcot, the Permanent Undersecretary at the time describes it, talks with the constitutional parties and engagement with the Provisionals constituted two parallel strands in government activity, both of them important:

[T]he thing was pretty much ordered and articulated. We had two main streams, small p political activity going on called political development … for working with the Constitutional parties and the Irish government but the other was political movement which was trying to engage with the Republicans and, to the extent it was useful or possible, with Loyalism. So there were quite ordered streams of activity going on (interview, Sir John Chilcot).

Thomas’ account of the relationship between these two streams indicates that the inclusion of the Provisionals in a settlement was the preferred outcome of these parallel strands of activity. Rather than aiming to marginalise Sinn Féin, the talks were seen as a way of exerting pressure on Sinn Féin:

[A]longside the project of … talks was the project of trying to encourage Sinn Féin and the IRA … to abandon violence and enter the political process. Now that involved, as we saw it, putting both positive and negative pressure on Sinn Féin. The positive pressure was to say this will be an open agenda, there’s nothing ruled out save achieving agreement. … Now the negative pressure was obviously to say this is an important political and constitutional process we’re launching and if you’re out of it you will miss something very important (interview, Sir Quentin Thomas).

Thomas emphasises that the inclusion of Sinn Féin in the process was preferable to a settlement without them:

[A] deal with the constitutional politicians while the violence continued would be much less strong and robust than one which embraced those who had previously thought it right to deploy force to advance their political interests … when the talks started in March 1991 involving constitutional politicians [only] that was not our choice, our wish was to bring in all those who would commit themselves to constitutional means (interview, Sir Quentin Thomas).

The ultimate aim was a successful talks process involving all of the parties. There was no point in including Sinn Féin if the other parties were lost to the process. But inclusion of Sinn Féin was such an important objective that when the talks from which they were excluded collapsed Chilcot regarded it as a broadly positive development: ‘all‐party talks minus Sinn Féin would present a hell of a problem if they succeeded … to get Sinn Féin in subsequently would be far too difficult. You’d have to renegotiate the whole thing’ (O’Kane, 2004, pp. 80–1).

The dynamics of this policy shift provide some support for the thesis that policy ‘learning’ was crucial to the resolution of the conflict (O’Leary, 1997, pp. 675–6; Tannam, 2001). The discussions within the state apparatus in 1989 and 1990 show clear evidence of ‘critical reassessment and re‐evaluation of underlying assumptions’ (Tannam, 2001, p. 494). Ernst Haas (1990) notes that crises generate fundamental rethinking. The growth of Sinn Féin and the failure of the Anglo‐Irish Agreement to deliver a security solution in the 1980s served as a kind of ‘crisis’ that generated this deep re‐evaluation and that permitted this knowledge to become ‘consensual’.

The contemporaneous indications by Sinn Féin that the IRA was open to a negotiated settlement of the conflict were crucial in facilitating this policy shift. Back‐channel contact between the parties, beginning in 1990, facilitated a progression from tacit bargaining to an embryonic if uneasy partnership and a limited coordination of action, most notably in preventing leaks. Those involved emphasise the importance of maintaining consistency with earlier British policy statements and of working within the parameters long laid down for British government policy. Above all they insisted on the primacy of the principle that changes to the constitutional status of the North required majority consent within the North. These parameters had not excluded engagement with the Provisionals in the 1970s, however, and they did not exclude it in the 1990s.

If a realist goal of ending the violent conflict rather than living with it was the main driving force behind this policy shift, a number of other factors provided strong incentives. Permanent exclusion of the Provisionals provided increased political leverage to unionists, to the SDLP and to the Irish government in different ways, and constituted a significant restriction on the British government’s political freedom. Explaining in 1860 why he refused to rule out an alliance with the ideological enemy, France, that grand master of realpolitik, Otto Von Bismarck, explained that this would be like playing chess without being able to use all of the squares on the board (Steinberg, 2011, p. 133, p. 313). Engaging Sinn Féin brought these other squares on the board into play for the British government and increased its room for manoeuvre, ultimately decreasing the leverage that other parties could exert on its policy. We can see an analogous logic underlying the Israeli decision to engage with the Palestine Liberation Organization (PLO) in Oslo in the early 1990s, a move that reduced the leverage that the US government and emerging Palestinian political forces could exert on Israel (Waage, 2004).

Global patterns of change also influenced the termination of the conflict just as they had influenced its outbreak. In the late 1960s the American civil rights movement and the guerrilla war in Vietnam provided powerful models for the Northern Irish civil rights movement and the IRA campaigns, respectively, suggesting models for action and providing hope of success. In the late 1980s and early 1990s South African and Israeli–Palestinian peace talks similarly provided models that informed moves towards a negotiated settlement in Ireland. Crucially, they provided a new frame for interpreting, explaining and legitimating engagement on the part of governments with armed opponents. The dominant rhetoric of counter‐terrorism was partially displaced by the language of peace processes in public discourse.

The reopening of back‐channel contact between the parties is presented by some scholars as a relatively trivial adjustment in British government policy, ‘a shift in style, rather than substance’ (Bew et al., 1999, p. 116). But if we take it that the Provisionals had been aiming for a negotiated compromise settlement for many years, then the crucially important political development of the late 1980s and early 1990s was the decision of the British government to offer engagement that might lead to such a settlement. For the intermediary at the intersection between the parties this policy shift was crucial:

[T]he choice was made by the British to end it … England had decided the time had come to alter the structure in Ireland. And frankly all of us … were bit players in that bigger picture. That is how I see it (interview, Brendan Duddy, 27 November 2009).

In 1989 and 1990 the ship of state changed course. It may have shifted course by only a single degree, but the effects of that calculated policy change increased exponentially as time went on.

The Provos Need a Victory: Republican Strategy

If the British state returned to engagement with the Provisionals in 1989 after a gap of many years, driven by the realist goal of achieving a stable settlement, why then did the Provisionals respond to these approaches, given the unlikelihood of achieving their goals? Was it because they had effectively been defeated and were ready to surrender? Let us begin with a few bald statements that provide a foundation for explaining IRA strategy in relation to the peace settlement of the 1990s but which go against the grain of conventional understandings of the Provisional IRA. In the first place, the IRA was a realistic, rationally calculating organisation, inasmuch, that is, as any organisation can be described as such. It was focused on achievement and capable of adapting to changed conditions, as Richard English (2004) has emphasised. Second, the IRA correctly understood that the power to effect substantial political change rested with the British government. The unionist majority in the North constituted the most important and powerful force opposing republican objectives but unionist positions were strongly shaped by British government policy. Third, after a brief period of euphoric mobilisation in the early 1970s, the IRA leadership realised that it could not achieve a military victory and that the most it could achieve was a negotiated settlement that would require difficult and fundamental compromises on its part. That is, the Provisional leadership understood from a very early stage that the struggle would ultimately produce a settlement that fell significantly short of the aspirations of their supporters in important ways. It follows from this that managing those expectations and avoiding a split would be one of their key challenges when the conflict was terminated. The internal dynamics of the IRA are crucial to understanding the leadership’s approach to negotiation. The movement was always heavily dependent on strongly localised networks from which it drew both material support and legitimacy (McIntyre, 2003, pp. 192–3). The national leadership was not in a position simply to command ground‐level activists to pursue a course of action that they strongly opposed. This was evident during the 1975 ceasefire during which British officials repeatedly noted the difficulty the leadership had in imposing control on local units. Jonathan Powell has spoken of the persistent importance of these internal dynamics into the 1990s:

When you think about it from the point of view of Republicans Adams and McGuinness, had they set out their aims of what they were going to settle for in the Good Friday Agreement and afterwards at any time in the ’80 s or ’93 or ’94 when the ceasefire happened, they’d have never got support from the republican movement. They had to get there very crablike through a series of tactical moves (Spencer, 2010 p. 439).

Fourth, the Provisional leadership understood that, despite the loud denials to the contrary, violence and the danger of violence decisively shaped British government policy in the North, but it did so in complex ways. Most importantly, they understood by the early 1970s that the need to avert large‐scale violence supported by a large proportion of the unionist majority was the single most important determinant of British policy in Northern Ireland. Britain would not make any concessions to the Republicans that might generate such violence. To condense: the IRA was a politically astute actor focused on achievement, ready to make major concessions in order to achieve a settlement and quite logically focused on the British state as the key actor it was necessary to influence in order to achieve this. Engagement with the British state and an imperfect compromise settlement were what the IRA had been aiming for since the early 1970s.

The politics of Provisional violence were cogently distilled by the intermediary, Brendan Duddy, to his British interlocutor as the 1975 ceasefire came under pressure in the summer of that year. Seeking to explain the willingness of the Provisionals to reach a compromise settlement he told him:

Their policy was directed initially to the military defeat of the English, but it became immediately apparent that this was impossible. More slowly it became apparent that the British Army couldn’t militarily defeat the Provos. From the time that they accepted that, Provo violence was aimed at producing a negotiating situation (‘The Collected Works of Mr. C Vol 2’, 25 June 1975, POL 35/92, Duddy Papers).4

This analysis still held good by the early 1990s. The Provisionals would end their campaign in order to enter negotiations but they would not end it unilaterally. This explains the central place of engagement with the British government in the strategic thinking of the IRA. As the intermediary puts it, explaining why the Provisionals sought to maintain communication with the British government in 1976, even after they had supposedly been strung along and duped, ‘Whether or not they would have succeeded [in their political aims], half succeeded, maybe succeeded, any version of that really depended on keeping the line open’ (interview, Brendan Duddy, 29 July 2009). That is, any political achievement made by the Provisionals would be secured through engagement with the British government as the key power player.

But what kind of negotiated settlement would be acceptable to the Provisionals? While contemporary critics focused on the details of the movement’s demands and argued that these could not be met, the intermediary repeatedly emphasised to his British interlocutors not the details of a settlement but its general character. For a variety of reasons, the Provisionals needed a ‘victory’, a settlement that delivered visible gains. The intermediary repeatedly made this point during the 1975 contacts. When the British representative told him in early 1976 that the Secretary of State Merlyn Rees was about to announce the withdrawal of a substantial number of troops from the North, for example, the intermediary wrote in his diary that he requested that they:

Delete a sentence in Rees’ speech, saying ‘reduction of troops to peace‐time levels’. This would give the Rep.[ublican] Mov.[ement] no option but to declare a ‘fight’. I asked that the sentence read; – It is our intention to reduce our troops in N. I. or something similar. … to state, in advance, the final position leaves no room for Provo’s victory, which is a prerequisite for ending the Rep. necessity to fight the British (1975 Diary, 10 January 1976, POL 35/62, Duddy papers, emphasis added).

That is, peace could be achieved through implementing changes that the British government was content to make, and that even served its interests, such as the withdrawal of troops. But these changes had to be made as part of an agreed process in which the Provisionals secured visible gains. The British government had to be willing to coordinate these changes with the Provisionals, allow them to influence these changes, and acknowledge that they were being made partly in response to the Provisionals. This was a crucial element in any settlement, whether in the 1970s or the 1990s.

The point is made, from a different perspective, by John Chilcot:

I think, by 1990 … we had learned what both the history, the ideology, Sinn Féin, the Republican movement needed if they were to be responsive … If you talked about defeat you were just making it more difficult for yourself because you’re making it too difficult for them (interview, Sir John Chilcot).

But why would the Provos need a victory, as Duddy put it? The answer lies with the movement’s strength. The movement was sufficiently well resourced and well supported and its members had made sufficiently large sacrifices by the early 1970s that no leadership could abandon the campaign without any visible achievement and hope to bring the bulk of the movement with them. Surrender was not in the gift of any Republican leadership. Second, the role of leadership brought with it a responsibility to the movement and its activists, which had a force of its own regardless of a willingness to scale back political ambitions in order to achieve peace. Leaders had a responsibility in terms of their organisational role to try to achieve as much as was politically possible. Explaining what he saw as a driving motivation underlying the political position of key IRA leader David O’Connell during the 1975 talks, Duddy argues that O’Connell did not want to lead his movement to a terrible defeat. If the movement was to stop short of its goals it had at least to secure ‘an honourable settlement’, a term used by the Provisionals both in 1975 and the 1990s. The point made repeatedly by the intermediary in 1975, that the Provisionals would accept a compromise peace settlement, but that they needed to secure visible gains, had direct implications for the role of the British government. Peace would require active cooperation and coordination between the two parties. It would require a kind of partnership.

Conclusion

The intersection between the British state and the Provisional Republican movement is crucial to understanding the ending of violent conflict in Northern Ireland. To understand the peace settlement of the 1990s it is necessary first to understand the importance of republican agency in seeking a peace settlement. This agency was constrained by the need to secure sufficient gains to be able to carry the bulk of the movement in accepting a compromise that the leadership had understood from a very early stage would fall well short of expectations. As a consequence, republican moves towards peace required a cooperative relationship with the British state as only the British state had the capacity to deliver those gains.

It is necessary too to understand British decisions to engage and disengage from contact with the Provisionals at various stages as hard bargaining moves driven by realpolitik calculations. There was an awareness within the British state apparatus from the early 1970s that the Provisionals were willing to accept a compromise settlement but also that the political price of such a settlement for the British government could be high, given the resistance and suspicion of Ulster loyalists and unionists. We should cease thinking of the British state as the sole agent in this relationship, enticing the Provisionals into politics by a combination of carrot and stick. We might think instead of the peace settlement as the outcome of the conscious, if bumpily uneven, coordination of British policy and Provisional strategy.

The growth of Sinn Féin in the 1980s created what was in many senses a novel problem for the British state. Exclusion of the Provisionals now required the ongoing exclusion of a well‐supported political party with representation in a variety of elected fora. It added a layer of difficulty to any policy based on exclusion. The British government decision in 1989/90 to begin working towards a settlement that included the Provisionals was, on the whole, a much more significant new development than the rethinking of republican policy in the 1980s. The decision to return to engagement after a gap of many years was driven by understandings that had been circulating within the state apparatus for many years: that a settlement without the Provisionals would not resolve the conflict and that the Provisionals were willing to settle for a compromise that fell well short of their ideal.

Much debate over British government policy in Northern Ireland centres around the concept of policy learning and the issue of whether the state incrementally developed a more sophisticated understanding of the dynamics of conflict and the measures necessary to resolve it (Dixon, 2001; O’Leary, 1997). Given that the key understandings underpinning the engagement with Sinn Féin had been in circulation in the mid‐1970s, the key question is why they did not become ‘accepted’ knowledge. Why were they ‘forgotten’ for the duration of the long war of the late 1970s and 1980s? Crucial to the shaping of policy was not the simple availability of knowledge but the struggles involving power, politics and ideology in which certain kinds of learning were marginalised and forgotten and others became ‘consensual’. These struggles are vitally important in generating apparent discontinuities in policy. Haas notes that crises generate fundamental rethinking. The parallel growth of Sinn Féin and the failure of the Anglo‐Irish Agreement to deliver a security solution in the 1980s served as a kind of ‘crisis’ that generated a re‐evaluation and that permitted this knowledge to become ‘consensual’. Within high‐level decision‐making and military circles there was apparently no significant resistance to this move towards engagement. There were no loud voices at the highest levels arguing that the state should instead push on and attempt finally to secure a military victory over the Provisionals (interview, Peter Brooke).

The ending of the violence in the 1990s required the development of a cooperative relationship between the British state and the Provisional leadership. This cooperative relationship was first established through back‐channel communication in the early 1970s, was suspended after the 1975 ceasefire and was renewed in the early 1990s (Mumford,2011; Ó Dochartaigh, 2009; 2011a; 2011b; Ó Dochartaigh and Svensson, 2013). This cooperative process required the development of a high degree of mutual understanding of the limitations within which the other party operated, as well as limited trust and active coordination. John Chilcot puts it well:

[I]t’s a process of mutual understanding, where, and it’s not two static positions of course, is it. It’s a process. Can we come together and can we build enough trust and confidence about the other side’s intentions, the limits around their freedom to move, pace of movement, direction of movement. Both sides are doing it (interview, Sir John Chilcot).

Reading back over the messages sent through the back‐channel in the 1970s and again in the 1990s, one is struck above all by the tone of the communication, a tone of reasoned discussion and even occasionally of solidarity. It is a tone that contrasts sharply with the contemptuous tenor of the bulk of the many thousands of official documents generated by the British state in relation to the Provisionals during twenty years of conflict. Crucial to the success of this process was the building of a partnership in which, ironically, the British government and Sinn Féin in certain ways and on certain occasions came to cooperate and coordinate more closely with each other than either of them did with any of the other parties to negotiation (Powell, 2008).

Both sides learned well the constraints within which the other party was operating, and gradually became willing to make the moves and concessions that would allow the other party to move in turn. They also came to rely on the judgement of the other party as to how far they could move. Ultimately this partnership between the Provisionals and the British government, a partnership that first developed during the face‐to‐face talks of 1975, was the single most important element in ending the violent conflict.

• Niall Ó Dochartaigh is a Lecturer in the School of Political Science and Sociology at the National University of Ireland Galway. He has published on conflict, negotiation, territory and new technologies in a range of journals and is the author of two books: From Civil Rights to Armalites: Derry and the Birth of the Irish Troubles (Cork, 1997; Palgrave, 2005) and Internet Research Skills (Sage, 2002; 2007; 2012). He is a founding convener of the ECPR Standing Group on Political Violence. Further information is available at niallodoc.wordpress.com. Niall Ó Dochartaigh, School of Political Science and Sociology, Aras Moyola, National University of Ireland Galway, University Road, Galway, Ireland; email: niall.odochartaigh@nuigalway.ie