The Intelligence War against the IRA

by Thomas Leahy

  During the Smithwick tribunal, numerous security-force members from both Northern and southern Ireland made similar points. Retired British Army brigadier Mike Smith operated in south Armagh during the 1980s. He saw the area’s IRA unit as ‘amongst the most capable and experienced of the terrorist groupings’ primarily because ‘they were very much risk averse’. In other words, they did not carry out attacks without plenty of intelligence beforehand. Ian Liles, another retired British Army brigadier who served in the area in the early 1990s, also described the South Armagh IRA as ‘extremely professional and extremely risk averse’. According to Smithwick’s report: ‘[Liles] emphasised that … the South Armagh Brigade … survived intact for so long … because “they were ultra-cautious.” … if they had seen one vehicle out of place, they would simply have called an operation off.’ Peter Maguire, a long-serving Garda officer, also believed that the South Armagh Brigade was ‘the most efficient’ IRA unit because it was ‘security conscious’.51 These factors explain why the South Armagh IRA was still at the forefront of the republican armed campaign in August 1994.

  Fermanagh

  In terms of ‘intended targets’ killed per year, the IRA’s campaign in Fermanagh remained intermittent after 1975.52 Appendix A.1 illustrates that three ‘intended targets’ were killed in both 1976 and 1977, declining to one in 1978, before increasing to five and eight in 1979 and 1980. This inconsistent pattern continued in Fermanagh up to 1990. For example, the IRA were responsible for the deaths of no personnel linked to the security forces in 1983, whereas they killed four in 1984. At least two killings per year occurred until 1990, when the level again dropped to zero. The Fermanagh units only targeted three security-force members between 1990 and 1994. The activities of agents and informers do not appear to explain these statistics. Lost Lives records no suspected agents and informers from the Fermanagh area being executed there or elsewhere by the IRA between 1976 and 1994. Admittedly, some agents and informers from Fermanagh only emerged after 1998. In February 2006, for example, Séan Lavelle, a local Sinn Féin helper, confessed in An Phoblacht to having worked as a RUC Special Branch agent there for many years. At the time, Tom O’Reilly, a local Sinn Féin representative, said: ‘[t]he knowledge … Lavelle had on republicans in Fermanagh isn’t worth talking about’, since Lavelle was not a member of the republican movement.53 As we shall see, there is evidence to support O’Reilly’s argument.

  In 2000, Rob Lewis, a former FRU member, released his account of agent-handling in Fermanagh during the 1980s. Lewis claims that the FRU’s best source was a woman he calls ‘Brenda’, a Sinn Féin member. The FRU tasked Brenda with forming a relationship with a person that the FRU believed was a leading republican in Fermanagh. Another low-level agent, ‘Declan’, informed the FRU that this republican attended clubs, and so a plan was hatched to see if Brenda could befriend the individual. The idea apparently worked and they began dating. Her photos of this person were crucial, Lewis says, because the FRU’s older photographs looked nothing like him, meaning that ‘he had been travelling back and forth into Northern Ireland for a long time completely unknown’. But the suspect stopped crossing the border after a particular incident, minimising opportunities to arrest him. Nevertheless, he did ask Brenda eventually to receive his mail at her address so he could avoid detection. She agreed; it was handed over to the FRU and, in turn, MI5 in London, since it supposedly discussed Libyan weapons. To cover Brenda, the Army raided all the houses in her street, allowing her to tell the leading republican that she had burned the letter for security purposes. On other occasions, she used a post box down the street, which allowed the intelligence services to read the letters before returning them to the post box. Brenda’s informing ended once the republican suspect apparently became tired of her refusals to marry him.54

  The Fermanagh IRA faced similar surveillance threats to the IRA in other rural areas, including checkpoints installed with the Vengeful system and undercover operations by special forces such as the SAS. In 1984, Antoin Mac Giolla Bhride from Magherafelt was killed by the SAS in Fermanagh following the detection of a planned IRA attack on 2 December. A small unit of volunteers including Mac Giolla Bhride were allegedly caught planting a bomb on a Fermanagh border road. Mark Urban reports that the IRA left the bomb near a restaurant so that they could phone and warn the police, who would be lured into the path of the explosion. But the SAS arrived on the scene before the attack could take place. Urban suggests an informer provided details about the forthcoming IRA attack, although there is no corroborating evidence. Accounts differ in their explanations of why Mac Giolla Bhride was shot alongside Alistair Slater, a member of the SAS. The IRA claimed they had only fired at the SAS because they had been attacked first, whilst preparing the device. In contrast, the SAS claimed that they only shot Mac Giolla Bhride after attempting to arrest him, when the detainee allegedly reached for a weapon. This attack also cost the IRA another volunteer, Kieran Fleming from Derry City, who drowned following an attempted escape across the River Bannagh.55 There clearly was some intelligence on IRA activities in the area, although it is worth noting that IRA volunteers from other regions conducted this attack. The lack of persistent SAS operations against IRA members from Fermanagh could suggest that the latter were not under surveillance to the same extent as IRA volunteers from Derry.

  These examples do not indicate high-level infiltration of the IRA in Fermanagh. Those informers noted were loose acquaintances of the organisation, occasionally gathering ‘loose talk’. In Brenda’s case, her information could not lead to the arrest of the republican since he did not enter the North frequently. The republican suspect was also very security-conscious, which meant that he would disappear for weeks without Brenda knowing where he had gone. It proved extremely difficult to recruit spies within republican ranks in Fermanagh. Lewis records attempted recruitments of a former IRA prisoner and a female close to a suspected volunteer. Both refused. Lewis commented: ‘[t]he border areas … were a desert for intelligence-gathering … the communities were tightly knit affairs that were a real nightmare scenario for source-recruiting’. The cross-border nature of some IRA operations in Fermanagh made trying to recruit republicans to inform tricky because they tended to reside in the Irish Republic.56 The British area review for Fermanagh in January 1981 concurred: ‘[t]he collection of intelligence presents difficulties, as few terrorists in the region live north of the border. The supporters and sympathisers who do are so intensely hostile as to make the recruitment of sources very difficult.’ The report notes a ‘dearth of available intelligence’ on the Fermanagh and Monaghan IRA volunteers, meaning that the British forces ‘had very few successes’ against the Fermanagh units.57 Lewis agrees that: ‘[t]he information we extracted from our sources remained … fairly low-level.’58

  As little intelligence emerged for British forces in Fermanagh, IRA activity there remained a persistent threat to the security forces, even if it was sporadic. Whilst ‘intended-target’ killings by the IRA in Fermanagh declined in the 1990s, the Fermanagh units remained active. Richard Latham, a former RUC officer who patrolled the area in the 1990s, certainly did not sense a trajectory of decline for the Fermanagh IRA before 1994. They continued to attack Roslea, Tempo and Newtownbutler RUC and British Army barracks right up until the 1994 ceasefire; as a result, according to Latham, ‘the whole talk of a ceasefire hardly seemed credible’.59

  What made the IRA potentially so dangerous in the rural areas was their capacity to inflict multiple deaths in single attacks. At times, the Fermanagh IRA certainly posed a threat of this nature as these examples demonstrate: on 15 May 1976, an IRA landmine killed three RUC officers during their patrol in Belcoo; similar incidents occurred in April 1977 and April 1982, when on each occasion a British soldier died after their vehicle struck a landmine in Belleek; the IRA killed a UDR soldier in February 1986 by detonating a bomb in a wall as a security patrol passed in Belcoo; and in November 1992, a sniper shot an RUC officer at a Belcoo chec
kpoint from across the border.60 The typical IRA method of attack in Fermanagh, which involved the shooting of off-duty security-force members living in the border farmlands, also continued. These tended to be ‘soft targets’ for the IRA, who sometimes operated out of Monaghan in order to hide from the British security forces before and after the attacks.61 This variety of IRA operation in Fermanagh meant that the security forces were unable to lower their guard until 1994.

  The intermittent nature of the IRA’s campaign in Fermanagh compared to that in south Armagh can be explained by factors other than infiltration. According to the British area review, on the one hand, the local security-force community living on the border in Fermanagh provided ‘soft targets’ for the IRA. On the other, the larger border Protestant population had two inhibiting effects on the Fermanagh IRA. One was that ‘the presence of Protestants within the community gives PIRA the perception that it is more difficult to operate unobserved and unreported’. The other problem was that:

  [t]he Protestant community … is seen to be under attack and this alters the public perception of the threat. If a policeman is killed in Belfast he is seen as a policeman, while if a policeman is killed in Fermanagh he is seen as a Protestant … the Protestant perception of this threat … leads to local demands for reassurance by the Security Forces.62

  Henry Patterson’s work on Fermanagh concurs. The IRA argued that they targeted police officers because of their uniform. But the fact that the majority of officers in Fermanagh tended traditionally to be Protestant meant that killing them provided ammunition for local unionists to claim that the IRA was conducting ‘ethnic cleansing’ in the border areas.63

  Certain incidents only added fuel to the unionists’ argument. During a Remembrance Day service in Enniskillen in 1987, a bomb killed eleven Protestant civilians. The West Fermanagh unit was eventually stood down in the late 1980s after they also shot dead a civilian in March 1988 near Belleek, whose family the IRA claimed was connected to the security forces.64 The temporary disbanding of the West Fermanagh Brigade could account for the decline in attacks during that period. Civilian killings had to be addressed by republican leaders, as they could erode Sinn Féin’s support. After the Enniskillen bombing, for instance, the SDLP dropped their support for Sinn Féin candidates in favour of the Ulster Unionists for the chairperson positions on the Fermanagh district council.65

  Laurence McKeown provides an alternative explanation as to why IRA attacks in Fermanagh declined in particular periods:

  A lot of the time in these places it depended on whether people could come together who had leadership qualities. In Fermanagh things really revived once Seamus McElwaine came out of jail … Perhaps the activities died off because a few were killed or imprisoned leaving nobody there to grab the initiative.66

  With fewer volunteers to recruit from a smaller population, finding leading figures with ‘initiative’ in small rural areas was vital to sustaining a campaign.67 Kieran Conway found whilst operating for IRA general headquarters that ‘rural areas were variously strong or weak depending on the presence there of … one or two strong personalities who would galvanise the rest’.68 For the Fermanagh IRA, Seamus McElwaine from County Monaghan was crucial; hence the IRA selected him to escape from the Maze prison in 1983. According to Ken Maginnis, a former UDR officer and an Ulster Unionist MP for the area, McElwaine was responsible for at least a dozen killings. The IRA’s killings did decline in Fermanagh after his arrest in 1981. Once he escaped in 1983, they increased to five in 1984, before declining thereafter; this was particularly noticeable after the SAS killed McElwaine in 1986. Even McElwaine’s death at the hands of the SAS on 26 April 1986 was not the result of prior intelligence. His unit had placed a bomb on the Lisnaskea to Roslea road. By pure chance, a soldier spotted the trip wire during a routine patrol, allowing the SAS to wait for the IRA to recover the bomb after it failed to explode.69 Further evidence of the lack of significant informers within the Fermanagh IRA can be seen in the British area review of 1981. It described how intelligence-led operations using the SAS had been ‘successful’ against the IRA elsewhere and called for ‘a similar operation in … Fermanagh’. Sufficient intelligence ‘does not exist to enable SAS operations to commence’, the authors commented at the time.70 A sustained SAS assault on the Fermanagh IRA never emerged, so it seems likely that there was a lack of adequate intelligence.

  South Derry

  Activity on the part of the IRA in other predominately rural areas maintained a smaller degree of pressure on the British state at different times across the period. The South Derry IRA was particularly active in the late 1970s under the leadership of Francis Hughes, who died on hunger strike in 1981. Hughes is a further example supporting McKeown’s argument that local initiative was crucial in driving republican armed activity in rural areas. Hughes and Dominic McGlinchey – who became a member of the INLA in 1982 – led South Derry volunteers in attempting to dominate the area. In 1976, they killed seven intended targets, followed by another four in 1977.71 Hughes and McGlinchey were key protagonists. Indeed, the RUC issued a ‘wanted’ poster including their images in 1977. In a gun battle in 1978, Hughes was wounded and captured.72 His departure – and that of McGlinchey in 1982 – had a detrimental effect on the South Derry IRA. Conway argues: ‘south Derry was never the same after the break-up of the unit centred on Francis Hughes and Dominic McGlinchey’.73 From 1979 to 1988, very few IRA targets were killed in the region.74

  Yet by the 1990s, the IRA had become more active outside of Derry City again. Examples include the destructive bombings of Coleraine and Magherafelt town centres in 1992 and 1993.75 It is not entirely clear whether South Derry volunteers were involved in these incidents, because Derry City volunteers had also contributed to recent attacks, such as at Coshquin in 1990. Whatever the case, the attacks show that the IRA evaded operational difficulties in one area by switching their focus to other areas. This argument is not to say that South Derry units posed a serious threat to the security forces across the period, but they did assist the IRA in maintaining their campaign. In addition, local republican groups were difficult to eradicate entirely because of a sizeable republican support base. Council elections in Magherafelt saw republicans frequently gain four or five seats from the 1980s, the same number as the SDLP.76

  Down

  Another area where the IRA remained active until the 1990s was that encompassing north Down and mid-Down. Local IRA units there did experience infiltration. Martin Dillon records that James Young from Portaferry allegedly became an informer in 1981. The IRA claimed that Young ruined a local bombing offensive in the mid-1980s, and that he bugged weapons.77 Nonetheless, the IRA maintained a persistent low level of activity in Down from 1983 onwards (this does not include south Down, as a separate unit covered that region). The organisation killed five intended targets – UDR and RUC officers – across County Down in 1983, and another seventeen between 1984 and 1989. A further five killings took place in 1990, which included four UDR soldiers killed in a single attack in April 1990 when their vehicle struck a landmine near Downpatrick.78 Although there were very few killings in the 1990s, IRA bombs devastated Bangor and Newtownards town centres in 1992 and 1993.79 The North Down and Mid-Down units created additional pressure on local security forces, although republicans did not pose a political threat there as the SDLP won the electoral battles by a substantial margin.80

  North and Mid-Armagh

  North and Mid-Armagh units shared some similarities in their experience with those of Belfast and Derry City volunteers, as some of the units operated in towns such as Lurgan and Portadown. They had a lower number of volunteers, however, because of the smaller population available to recruit. This point partly helps to explain why they escaped significant infiltration. Nevertheless, the IRA claimed that it did take place. The IRA shot David McVeigh, an IRA volunteer from Lurgan, near Carlingford in September 1986. The IRA claimed that he had become a police informer after being arrested in relation to a
bombing at Lurgan Golf Club in 1982. McVeigh’s family deny these accusations.81 The IRA killed a further three volunteers from Portadown in July 1992, in a particularly disturbing case. Republicans allege that Gregory Burns, Aidan Starrs and John Dignam informed on IRA activity in the area in the 1980s. Martin Ingram says that Gregory Burns applied to join the UDR in 1980, but the FRU approached him to inform. The information he provided allegedly included informing the police on the whereabouts of his brother Sean Burns, who was shot by the HMSU in November 1982. Ingram also suggests that these three men profited from IRA robberies. They were eventually caught out. According to Ingram, the FRU decided not to relocate them after Gregory Burns revealed to a girlfriend, Margaret Perry, that he worked for the intelligence services. The story then took a sinister twist. The three men supposedly agreed to kill Perry and leave her body in an unmarked grave, since they feared that she would expose them. Ingram’s account concludes with the IRA eventually investigating the matter and luring Gregory Burns, Dignam and Starrs to their deaths at the hands of Stakeknife’s internal security unit.82

  A series of IRA deaths at the hands of the HMSU at the beginning of the 1980s also represented setbacks and demonstrates the existence of good sources of intelligence for the security forces. On the 11 November 1982, the HMSU shot Sean Burns, Eugene Toman and Gervaise McKerr at a police roadblock near Lurgan. The RUC claimed that the HMSU unit only opened fire after the republicans failed to stop, which the IRA disputes. In subsequent inquiries, it emerged that Sean Burns and Toman were suspected of killing three RUC officers in previous weeks, and that they ‘had been under observation for some time’. Another controversial killing occurred a few weeks later, on 24 November 1982. According to various sources, Michael Tighe from north Armagh was shot dead, and Martin McCauley was wounded, following a HMSU surveillance operation on a hay shed near Lurgan. The intelligence services believed that the hay shed was an arms dump. MI5 had bugged the hay shed beforehand. There are allegations that the three volunteers killed as informers in the 1990s set up this ambush.83 This incident became part of the investigations carried out in May 1984 by John Stalker, the Deputy Chief Constable of Greater Manchester Police, into alleged shoot-to-kill attacks by the HMSU. Stalker was replaced in June 1986 in controversial circumstances.84 The crucial point here is that there were a variety of intelligence sources available in the early 1980s on which to base ambushes of republicans in north Armagh.