In the early hours of 17 June 1922, an Irish Republican Army (IRA) reprisal party left Dundalk Military Barracks, in the fledgling Irish Free State, and crossed the border into Northern Ireland. After a brief stop in the townland of Ballymacdermott, where some of its number opened fire on the home of a Protestant widow named Mary Thompson, the twenty-strong column arrived in the townlands of Altnaveigh and Lisdrumliska, a small and predominately Protestant farming community on the outskirts of Newry. In the hour that followed, eight Protestant households were attacked with gunfire and incendiaries. The worst of the violence was borne by four families – the Heaslips, the Grays, the Croziers and the Lockharts – who had their homes burned and loved ones shot in their presence. Six people were killed and one other seriously wounded. The dead were mostly men, ranging in age from seventeen to sixty, and, despite republican claims to the contrary, none was a member of any police, military or paramilitary force. One woman was also killed. She recognised a neighbour among the men who had gunned down her husband, and called on him by name. In response, she was shot in the arm and bled to death shortly thereafter. Having completed its operations, the reprisal party returned to Dundalk. A few of its members, however, did so only after splitting from the main group to commit further incendiary attacks in nearby Cloughreagh and Derrymore. These included a failed attempt to burn down the historic Derrymore House, where the Act of Union was drafted in 1800.Footnote 1
Many Irish historians are familiar with this night of violence, since dubbed the ‘Altnaveigh massacre’. An incident that was overlooked for many years – though never forgotten locally – since the late 1990s it has attracted considerable popular and scholarly attention.Footnote 2 As one of the more extreme examples of the violence that occurred amid partition, it has featured prominently in studies by Robert Lynch and Tim Wilson and is regularly cited in the broader historiography.Footnote 3 In 2006 the reprisal also sparked public interest when an Irish television documentary revealed that it had been carried out on the orders of Frank Aiken, who later became one of independent Ireland's most noteworthy statesmen.Footnote 4
Despite the attention the reprisal has received, however, relatively few are familiar with what happened next. Exactly one week later, back across the border in Dundalk, members of the IRA unit responsible for the massacre (the Fourth Northern Division) were now engaged in an apparently sincere effort to deter attacks on the local Protestant community. Their endeavours were prompted by a number of violent and intimidatory incidents in the town – evidently the actions of other IRA volunteers – which purportedly served as reprisals for attacks on the Catholic minority in Northern Ireland. A system of nightly patrols was introduced to deter further incidents. Aiken, who himself had ordered the carnage at Altnaveigh and Lisdrumliska, threatened to shoot anyone caught engaging in such acts. The suspected perpetrators were subsequently seized and tried by IRA court martial, though there is no record of the verdicts. A number of Protestant families preparing to flee the town were also prevailed upon to stay.Footnote 5
This series of events sits uncomfortably amid the prevailing narratives of republican violence during the Irish revolution. The contradictions are not easily dismissed and present an inconvenient anomaly for a historiography that is at times bitterly polarised on the issue of IRA sectarianism during the conflict. Though it risks oversimplification, two broad strands of opinion are discernible. On the one hand, there are those scholars who place a greater emphasis on sectarianism as an explanation for republican violence and thus present a challenge to the traditional nationalist/republican narrative of the conflict. This is most often associated with the work of the late Peter Hart. In his influential study of revolutionary Cork and in subsequent essays, he argued the primacy of sectarianism as an explanation for IRA violence against members of the Protestant community.Footnote 6 His work also popularised the discussion of ‘ethnic cleansing’ in the context of the Irish revolution, though he himself ultimately rejected this interpretation. So too have virtually all professional historians of the conflict, to varying degrees. In her study of Munster during the civil war, for example, Gemma Clark has rejected the ethnic cleansing thesis as the violence involved was not state sanctioned, but her research does suggest that attacks on Protestants and other perceived communal outsiders bore the characteristics of a homogenising purge.Footnote 7 Others have based their rejection of the term on demographic analyses, rather than direct engagement with the violence involved.Footnote 8
On the other hand, there are those scholars who reject the ‘sectarian narrative’ of republican violence and who to some extent adhere to a more traditional interpretation of the IRA's conduct during the conflict. The suggestion that republicans targeted Protestants primarily on the basis of their religious identity is repudiated. Attacks on members of the Protestant community are instead viewed in more individual terms, as acts of ‘controlled military violence’, often driven by necessity, usually guided by firm evidence (or a plausible suspicion) of wrongdoing and conducted via legitimate republican military structures.Footnote 9 It is not disputed that incidents of sectarian violence occurred. But they are not viewed as being representative of the republican movement's conduct during the conflict. This has been a key point of recent critiques of the ‘sectarian narrative’ in articles by John Regan, Andy Bielenburg, John Borgonovo and James S. Donnelly Jr.Footnote 10
It is tempting to think of this schism as a legacy of the ‘revisionist’ debates of the 1970s, eighties and nineties. However, Irish historians roundly reject such labels. As Stephen Howe has noted, moreover, even at the time ‘much research and publication could not readily be assigned fully to either camp’.Footnote 11 As palpable as the division seems, this remains the case. Indeed, it is perhaps more useful to think of it in terms of largely unacknowledged tensions between differing conceptual understandings of sectarianism and its application as an analytical tool for the study of violence.
The resulting historical disputes have largely been confined to the discussion of events in southern Ireland. There is, by contrast, little disagreement as to the sectarian character of the violence in what became Northern Ireland. Granted, over the past fifteen years there has been a wealth of new research on the northern experience of the revolutionary period, and this has encouraged a more nuanced discussion of the dynamics of the conflict and its local contexts. Yet engagements with sectarian violence can still prove problematic.Footnote 12 The ‘Altnaveigh massacre’ is a case in point. As Lynch observed in his 2010 article on the atrocity, it is commonly viewed as ‘one of the few identifiable examples of IRA violence inspired by crude sectarian motivations’. Indeed, despite criticising such an assumption, and drawing much needed attention to the massacre's place in a broader pattern of reprisal and counter-reprisal, he reaches a similar conclusion. Pointing to an intense sectarian environment characterised by ‘frequent attacks on churches and their congregations’ and ‘largely indiscriminate attacks on civilians’, he ultimately describes the episode in terms of ‘sectarian scapegoating’ and suggests its ‘near inevitability’.Footnote 13
The Fourth Northern Division's subsequent efforts to protect the Protestant minority in Dundalk do not necessarily prove otherwise, but they do suggest that matters were decidedly more complex. Accordingly, this article seeks to think afresh about the sectarian dimensions of republican violence on the south-east Ulster frontier, defined here as the Armagh and south Down areas of Ulster and the north Louth area of Leinster.Footnote 14 Drawing loosely on the definition suggested by Robbie McVeigh, it understands sectarianism as attitudes and practices which serve to ‘construct and reproduce the difference between and unequal status of Irish Protestants and Catholics’.Footnote 15 In this sense, it does not dispute that most (perhaps all) of the violence detailed here was sectarian. Yet it keeps an open mind as to what this meant in practice. In particular, it aims to avoid lapsing into the reductionist notions that seem to pervade the discourse around sectarianism – the expectation of largely indiscriminate attacks on civilians, the emphasis on religious identity when explaining victim selection, the attribution of sectarian pathologies to the perpetrators or the implication that visceral sectarian antipathy was a prime motivational force for violence. In doing so, it aims at a more nuanced appreciation of how such sectarian violence was both facilitated and limited at a grassroots level. In this regard it hopes to contribute to filling the ‘analysis gap’ identified by Wilson in his ground breaking study of intercommunal conflict in Ulster and Upper Silesia in this period, within which lies a ‘wide range of local motivations and behaviours that rarely receive sustained academic attention’.Footnote 16
I
On the eve of the Irish revolution, the south-east Ulster frontier had a slight Catholic majority of 54.9 per cent; the various Protestant denominations accounted for a further 43.4 per cent; all others made up a mere 1.7 per cent. The respective strengths of the two main communities varied considerably from north to south, however. The parliamentary constituencies of North and Mid Armagh had Protestant majorities of 67.5 per cent and 55.5 per cent. By contrast, South Armagh, South Down and North Louth had Catholic majorities of 68 per cent, 53.5 per cent and 90.3 per cent respectively.Footnote 17 As might be expected, the region's political and religious geographies were closely bound. North and Mid Armagh were solidly unionist; South Armagh, South Down and North Louth were solidly nationalist. Indeed, this was strikingly reaffirmed by local responses to the Ulster Covenant in 1912, as Ireland's ‘revolutionary decade’ commenced. This was a solemn oath of opposition to home rule, a limited measure of self-government which had been the single-minded goal of Irish nationalists since the 1870s. Locally, it was signed by upwards of 57,270 unionists, the vast majority of whom (around 65 per cent) were based in North and Mid Armagh.Footnote 18
Despite their opposing political goals, and the area's long-established reputation for intercommunal conflict, relations between the nationalist and unionist communities were remarkably stable locally.Footnote 19 In the preceding three decades, violence between the two had been limited to sporadic clashes linked to contentious parades or political demonstrations.Footnote 20 Yet the polarisation of the home rule crisis (1912–14), and the resulting mobilisations of the unionist Ulster Volunteer Force (UVF) and nationalist Irish Volunteers, had remained peaceful. Despite their divisive potential, moreover, the advent of the Great War, the continued negotiations concerning Ulster's exclusion from home rule and the 1916 Easter rising in Dublin failed to stir any violent local discord. Indeed, even the gradual emergence of the republican movement after 1916 was initially more noteworthy for its stimulation of intra-nationalist tensions in the region than it was for its impact on inter-communal relations. This was particularly true across Armagh and south Down, where Sinn Féin and a reinvigorated Irish Volunteers movement struggled to gain ground against the resilient constitutional nationalism of the Irish Parliamentary Party (IPP) and its grassroots organisations, the Ancient Order of Hibernians (AOH) and the United Irish League (UIL).Footnote 21
The situation changed in 1919 when the Irish Volunteers (soon rebranded as the IRA) commenced a violent campaign for independence. The first operations in the area were arms raids, the most promising target for which were the weapons caches of the UVF. It was common knowledge that the latter had imported large quantities of rifles into Ulster at the height of the home rule crisis, and that these remained hidden throughout the province. In pursuit of this elusive bounty, local IRA units forcibly searched a number of suspected arms dumps, among them an Orange hall at Adavoyle, and the country manors of three prominent unionists – one of whom was a UVF regimental commander – at Ravensdale, Rostrevor and Loughgall.Footnote 22 Republicans may also have been responsible for an alleged arms raid on a rectory at Lisnadill in October 1919 which resulted in the non-fatal shooting of a Protestant clergyman. Yet the truth of this incident is elusive.Footnote 23
There is little reason to suspect that these raids were motivated by anything other than the pursuit of arms. Even the incident at Lisnadill – if it was the work of republicans – was evidently aimed at capturing a quantity of UVF rifles that had previously been stored at the property.Footnote 24 As Wilson has observed, however, when analysing sectarian violence the motives or intentions of those involved are often unimportant. What really matters is how their actions are understood by the opposing community.Footnote 25 Ultimately the arms raids could be interpreted by the Protestant-unionist community in sectarian terms. Those targeted represented key elements of the northern Protestant establishment – the Church of Ireland, the Orange Order and the unionist political elite. It is perhaps unsurprising, then, that such incidents inspired some within the unionist community to form defensive patrols. The first appeared in April 1920, with others emerging throughout the year as IRA arms raids were extended to the homes of anyone (unionist or nationalist) suspected of owning a firearm.Footnote 26 The emergence of such resistance was to be expected. There were plenty of suitable leaders and recruits within the local unionist community, many of whom possessed military experience. Though dormant, the paramilitary structures and networks of the UVF remained as a potential means for mobilisation and promised access to an arsenal that far exceeded that of the IRA. Indeed, unionists had mobilised and armed to maintain their place in the union before. Why should they have been reluctant to do so again in response to the separatist threat of the republican movement?
Unionist patrols did not become particularly active locally until the autumn of 1920 and then all but disappeared with the creation of the Ulster Special Constabulary (USC) in November, an almost exclusively Protestant paramilitary police force recruited to re-enforce the struggling Royal Irish Constabulary (RIC).Footnote 27 Nevertheless, their appearance had an immediate effect on IRA decision making. John McCoy, a prominent local IRA officer, later recalled that the decision to attack Newtownhamilton RIC barracks in May 1920 was partly influenced by the suspected presence of ‘Ulster Volunteers’ in the district and a fear that nationalist in-fighting had distracted attention from unionist opposition. This was combined with residual ill-feeling from a brief clash between republicans and unionists in the district during a by-election in 1918 and a determination to give local unionists ‘a chance to prove their mettle when up against a serious local attack on what they had sworn to defend’.Footnote 28
This belligerent attitude was quickly curbed when, in the summer and autumn of 1920, various local and national events illustrated the potential consequences of such conflict. In June a unionist patrol successfully repulsed an attempted IRA arson attack at Lisbellaw in Co. Fermanagh, resulting in at least one casualty.Footnote 29 Closer to home, unionists in Kilkeel also succeeded in thwarting an attempted IRA arson attack in October.Footnote 30 More significantly, however, in July and August, mob attacks and workplace expulsions directed at nationalists in Belfast, Banbridge, Lisburn and Dromore demonstrated the potential for loyalist reprisals in response to IRA operations. These events were sparked by the IRA's assassinations of Lieutenant-Colonel Gerard Smyth (a Banbridge native) in Cork and District Inspector Oswald Swanzy in Lisburn. The violence in Banbridge had a direct impact on the local IRA. The town fell within the operational area of the Newry Brigade – the forerunner to the Fourth Northern Division. The family of a local IRA officer was one of the main targets for aggression.Footnote 31 It was also witnessed first-hand by two of the brigade's highest ranking figures, who inadvertently stumbled into the chaos whilst returning home to Newry after a battalion meeting in Lurgan.Footnote 32
Organised unionist resistance coupled with the threat of reprisals had mixed results as a deterrent for IRA activity. As Charles Day's mapping of violent incidents in the Newry and south Armagh area illustrates, local units appear to have increasingly confined their operations to districts with a pronounced nationalist presence, where the threat of interference or retaliation seemed less likely.Footnote 33 Yet they also remained indifferent to the sectarian significance of their actions. In 1921, for example, volunteers in Mullaghbawn burned down a vacant rectory on the plausible suspicion that it was to be commandeered for use by the army.Footnote 34 Attacks or raids on Orange halls, moreover, which sometimes served as meeting places and equipment stores for the USC's B Specials – part-time reservists who patrolled their home districts one night per week – became increasingly common as the conflict progressed.Footnote 35 Regardless of their intent, such incidents ultimately had a sectarian impact. Again, this was understandable. Occurring alongside unattributable acts of vandalism, such as the defacement of the gateposts at Armagh's Church of Ireland Cathedral, and in conjunction with republican policies such as the Belfast boycott, an effort to exert economic pressure on northern unionist business interests, it is difficult to see how else they could have been interpreted.Footnote 36
Loyalist reprisals also provoked the first republican counter-reprisal in the region. On 27 August 1920, arsonists targeted two Protestant owned department stores in Dundalk. The attacks resulted in the deaths of three Protestant employees. The burnings were interpreted by some as a reaction to unionist attacks on Catholics in Lisburn, Co. Antrim, days earlier. As Natasha Grayson has explained, they were believed to be a warning to local Protestants for their perceived failure to sufficiently condemn the violence of their co-religionists in Ulster. The rationale of the Belfast boycott may also have informed the choice of target. The IRA was believed to be responsible for the attacks, but this was strenuously denied by local leaders. Instead, they attributed the atrocity to a mysterious group of ‘Bolshevik’ ex-soldiers who had fled the town shortly afterwards.Footnote 37 A more plausible (if speculative) explanation may be that the burnings were carried out by IRA volunteers, but that they were neither sanctioned nor supported by local leaders and were thus quickly condemned and disowned amid the widespread public revulsion that the incident provoked. Indeed, the republican denials were first offered at a heated public meeting convened by the council and representative of all shades of political opinion.Footnote 38 Regardless of who was responsible, violence of this nature was a relatively new development.
In December 1920, however, a series of events served to escalate the conflict and created the conditions in which communally charged reprisals became an ever more frequent occurrence in the region. It began with an IRA attack on an RIC barracks in Camlough. Though a total failure, the incident provoked a stern response from the authorities. In its immediate aftermath, the military and a freshly deployed USC platoon burned down a number of houses that had been occupied by the IRA during the attack, an act in accordance with an emerging British policy of ‘official punishments’. In the days that followed, the USC then returned to burn further properties with links to local republicans. Aggressive policing continued in the district for some weeks, eventually culminating in the shootings of two republican suspects near Beleek. One died, the other was seriously wounded. The USC constables responsible claimed that both had been shot whilst trying to escape, but this was contradicted by the testimony of the surviving victim.Footnote 39
Shootings of this nature became a common response to IRA activity in the locality in the six months that followed. For the most part these appear to have been perpetrated by members of the RIC or USC, who were sometimes identified by their uniforms, or their use of police vehicles. Although often perceived as indiscriminate sectarian killings, targeting young Catholic men regardless of their political sympathies, this was not necessarily the case. In Newry and south Armagh, where the majority of such shootings occurred, there is sufficient evidence to suggest that seven of the nine victims were members of the IRA.Footnote 40 Furthermore, in those instances where civilians were killed, there is often reason to suspect that they were secondary targets. The murder of William Hickey near Newry in 1921 is a poignant example. He was abducted after his killers had unsuccessfully raided the homes of two known republicans.Footnote 41
Local IRA units responded to this aggression with reprisals of their own. Initially these were confined to further attacks on the RIC or USC. Shortly after the reprisal burnings in Camlough, for example, Frank Aiken – commandant of the Newry Brigade and later the Fourth Northern Division – ambushed a police patrol from the charred remains of his farm, an attack that left little doubt about its motivation and message.Footnote 42 IRA reprisals soon radicalised, however, and began to draw inspiration from the practices of the police and military, and the precedents established by their comrades elsewhere in the country. The most striking example of this came in April 1921, with a reprisal at Killylea, which was directly influenced by a similar IRA operation at Roslea, Co. Fermanagh, a month earlier. Local leaders had assisted in the planning of the latter attack and subsequently appropriated its methods and logic for use in their own area. This much is suggested by an account of discussions between Aiken and his Fifth Northern Division counterpart, Eoin O'Duffy, recorded in later years by John T. Connolly, captain of the Roslea Company: ‘Aiken did not at first approve of the burnings, as he thought the B.men would retaliate by burning double the number of nationalist houses. O'Duffy struck the table and said: “When you hit them hard they will not strike again.” Aiken then said: “well, burn them and their houses”.’Footnote 43
The details surrounding the Killylea reprisal are worth recounting as an example of how such incidents developed. On 10 April 1921 the IRA ambushed a group of USC constables at Creggan, close to the Armagh-Louth border. The attack took place on a Sunday morning, as the constables made their way to church. In making its preparations the ambush party rounded up local church-goers – both Catholic and Protestant – and held them under armed guard at a nearby pub to ensure that no one would raise the alarm. One constable was killed as a result of the attack, and another – Hans Leeman – was seriously wounded.Footnote 44 The incident provoked a reprisal in Leeman's home village of Killylea, some thirty miles away. The homes of two local nationalists were burned. In the course of the attack one of the householders was also shot in the stomach. Neither of the victims had links to the republican movement. Indeed, as members of the constitutional nationalist Ancient Order of Hibernians (AOH) they were likely antagonistic to both Sinn Féin and the IRA. Significantly, however, republicans and nationalists had recently agreed an election pact. This may help explain why the men were targeted. It may also explain why the IRA responded with a counter-reprisal. The plans called for the destruction of four properties. In the event, however, the reprisal party only succeeded in burning the homes of two local unionists, one belonging to a serving ‘B Special’, the other to Leeman's father. The volunteers involved were also under orders to kill these men, but relented when they provided information regarding those responsible for the earlier attacks.Footnote 45
No further reprisals of this kind occurred in the region prior to the truce. Only a day before the cessation of hostilities was announced, however, the IRA did carry out its first retaliatory killing of a civilian. The victim was a Protestant railway worker named Draper Holmes, and his death ultimately served as a reprisal for the killings of four IRA men who had been rounded up and shot by the USC during late night raids in the townlands of Altnaveigh and Lisdrumliska a week earlier.Footnote 46 Holmes was not the IRA's intended target. He merely stumbled upon the reprisal party as they waited for a group of suspected off-duty ‘B Specials’ and was shot in a panicked effort to ensure that he did not draw attention to their position. Nevertheless, as the later testimony of one of his killers makes clear, his death was subsequently rationalised to fit the purpose: ‘he suffered not for anything he himself had done but for a deadly danger to the lives and freedom of our companions in arms which men of his class represented’. Little regret was ever expressed, aside from the fact that the incident had left ‘our score on the losing side’.Footnote 47 There were no attempts to downplay IRA responsibility, and there is no indication that the incident raised any significant internal criticism. As such, it set a dangerous precedent.
Sectarian reprisals were certainly a more prevalent feature of the local conflict in the first six months of 1921. Yet they were by no means its dominant characteristic, and indiscriminate sectarian killings were not the norm. Of the thirty-three conflict related deaths that occurred in the region between January and June, perhaps five could be described in such terms, although even then the individual circumstances were often more complex than the judgement implies. By contrast, nine policemen were killed in the same period. The IRA also lost as many as fourteen volunteers, most of them at the hands of the USC in Newry and south Armagh. It is fair to suggest sectarian antipathy was not the decisive factor driving these combatant deaths. But such incidents did still have a sectarian significance.
The blurred distinction between civilian and combatant perhaps made this inevitable. The USC is a case in point. Prior to the creation of the force, the RIC had borne the brunt of the local republican campaign. Consequently, police casualties were often Catholics, and were usually not natives to the areas in which they served.Footnote 48 From late 1920, however, the USC superseded the RIC as the most visible policing presence across Armagh and south Down, and thus the most immediate target for republican aggression. Yet attacks on this force had very different reverberations to what had occurred before. USC constables may have represented a logical military target for republicans, but for unionists they represented co-religionists and communal cohorts. Furthermore, attacks on ‘B Specials’ also carried specific local resonances. The same blurring of the civilian-combatant distinction applied for the killings of IRA volunteers, though with one added dimension. Republican casualties were typically recorded as Catholic civilians – particularly in the press – thus fuelling the perception that they had been randomly targeted on the basis of religion. In practice, however, arbitrary violence of this nature was relatively rare prior to the truce.
II
By the summer of 1921, the local conflict had undoubtedly acquired a sectarian dimension. It is worth recognising, however, that this development was neither immediate nor inevitable. Rather, it was something that had emerged gradually – the result of a cumulative process of reciprocal escalation driven largely by the actions of the IRA and the USC. This intensifying cycle was temporarily stalled by the announcement of the truce on 11 July 1921 and the resulting eight months of uneasy peace and political uncertainty that followed as British and Irish leaders conducted negotiations in London. When the fighting resumed in Ulster in the spring of 1922, however, the familiar pattern of reprisal and counter-reprisal quickly regained momentum.
The violence recommenced in the wake of the Anglo-Irish Treaty, which was signed in December 1921. This agreement facilitated southern Ireland's transformation into a self-governing dominion – the Irish Free State – and the continuation of the existing partition of Ireland, as legislated by the Government of Ireland Act in December 1920. This had resulted in the creation of Northern Ireland shortly before the truce, a unionist-dominated entity with a devolved government which was pledged to remain part of the United Kingdom. The compromise proved divisive, however. Many southern republicans refused to countenance the compromise of Irish sovereignty. Their northern counterparts, meanwhile, were disillusioned by its recognition and reinforcement of partition. Consequently, when the Treaty was ratified by Dáil Éireann – the republican revolutionary government assembly – the movement split into opposing pro- and anti-Treaty factions.Footnote 49
In an attempt to placate northern IRA units – and, perhaps, to foster republican unity in the south with the lure of a common cause – the emergent pro-Treaty leadership in Dublin privately signalled their commitment to renewed fighting in Ulster. This was to be coordinated under the aegis of the newly created Ulster Council, a shadowy body composed of northern IRA commandants and heavily influenced by three key pro-Treaty figures: Michael Collins, the chairman of the newly created provisional government; Richard Mulcahy, the Dáil minister for defence; and Eoin O'Duffy, now chief-of-staff of the IRA, whose pro-Treaty faction was being rapidly reorganised to form the new national army.Footnote 50 Accordingly, the renewed campaign in Ulster was initially carefully calibrated in line with the provisional government's broader northern policy. The first operations in February 1922 occurred only after the collapse of the Craig-Collins pact, the first of two conciliatory agreements which (among other things) aimed at improving north-south relations and securing minority safeguards for nationalists in Northern Ireland. This was followed by two months of escalating IRA activity which was promptly restrained by the announcement of a second Craig-Collins pact on 30 March.Footnote 51 When this agreement also proved unworkable, plans were laid for a joint-IRA offensive. Supported by both pro- and anti-Treaty republican factions, this was evidently envisaged as an attempted invasion of Northern Ireland. Poorly conceived and hastily organised, however, the plot misfired and ultimately collapsed amid much confusion and recrimination in late May 1922.Footnote 52
IRA units in the Armagh-Louth borderland were prominently involved in this fresh wave of activity, and as was the case prior to the truce, their aggression was directed primarily at the authorities. This was even true with regard to renewed reprisal attacks, the first of which targeted USC constables.Footnote 53 Indeed, although retaliatory violence had immediately re-emerged as a feature of the renewed fighting in the region, it was initially stifled by the halting progress of the new IRA campaign as it responded to the ups and downs of north-south political developments. Consequently, it was only after the collapse of the May offensive that the familiar pattern of action and reaction regained (and then surpassed) the intensity that it had achieved in early 1921.Footnote 54
The immediate chain of events that led to the ‘Altnaveigh massacre’ began on 22 May, when the Northern Ireland government introduced internment in response to widespread violence in Antrim and Tyrone connected to the abortive joint-IRA offensive. On the south-east Ulster frontier – where last-minute orders had cancelled republican operations – the first round-up netted around thirty republicans.Footnote 55 In retaliation, local IRA leaders ordered the kidnapping (or ‘arrests’) of a number of ‘prominent unionists’ and members of the USC. In all, ten men were abducted and taken across the border to Dundalk Military Barracks. This facility had recently been vacated by the British army in accordance with the Treaty and was now occupied by the Fourth Northern Division with the reluctant agreement of the provisional government. Meanwhile, in Newry, the local Resident Magistrate, James Woulfe-Flanagan, a Catholic, was shot dead by his would-be abductors when leaving mass.Footnote 56 Republicans would later claim that these operations were an attempt to deter USC reprisals, holding hostages as collateral for the latter's good behaviour. In truth, however, they were an attempt to secure the release of their imprisoned comrades. This much was clear from the letters that their captives were compelled to write to loved ones and local notables suggesting prisoner exchanges.Footnote 57 Indeed, in this respect the raids were directly inspired by operations conducted by the Ulster Council further along the border in Tyrone and Fermanagh in February 1922.Footnote 58
The republican operations were not long in receiving a reply. In Newry, two businesses on the mainly nationalist Hill Street were burned in what was perceived locally as a reprisal.Footnote 59 With tensions remaining high, and the situation exacerbated by continued IRA activity along the border, these attacks were followed a week later (on the night of 13–14 June) with two separate outrages in south Armagh. The first occurred in the Camlough district, where two men – Thomas Crawley and Patrick Creggan – were abducted, shot dead and dumped on the road in the townland of Lislea. Neither victim appears to have been a member of the IRA, though Creggan had been linked to the republican movement in the past.Footnote 60 His brother, moreover, was an active IRA volunteer. Members of the USC were widely believed to be responsible for the killings.Footnote 61 The second occurred at Dromintee, close to the border, at a public house owned by a prominent Sinn Féin councillor named James McGuill. Armed men, believed to be disguised members of the USC, raided the premises in the early hours of the morning. McGuill wasn't home at the time. A longstanding target for reprisals, he was staying across the border for his own safety, leaving his family to look after the business. Failing to find him, the men ransacked the building, stealing money and drink, before turning their attention to the female residents. McGuill's heavily pregnant wife was subjected to a violent sexual assault and possibly raped. So too was one of the family's household employees.Footnote 62 Together, these events provided the immediate impetus for the ‘Altnaveigh massacre’, as well as an accompanying ambush of a USC patrol outside McGuill's pub at Dromintee which left one constable dead.
The sexual assaults at McGuill's were not typical of the conflict. This exceptionalism, Wilson argues, explains the severity of the IRA's response. An unspoken convention had been broken, and thus ‘the unstated rules of proportionality no longer applied’.Footnote 63 The point is well made. Indeed, at the very least, the sexual nature of the violence at Dromintee must have played upon more specific anxieties regarding the safety of female relatives, and these were all the more pertinent given that most active IRA volunteers from Armagh and south Down had moved into camps in north Louth, often leaving sisters, wives and mothers behind.Footnote 64
Yet there were other elements in the mix. In many respects the reprisal also represented an archetypal ethnic ‘anxiety reaction’ – as described by Donald Horowitz – where a fear of extinction or ‘swamping’ provokes a response which is disproportionate to the actual stimulus, and which flows from a ‘diffuse danger of exaggerated dimensions’.Footnote 65 Occurring alongside the murders at Lislea, the sexual assaults at Dromintee seemed to confirm local republicans’ worst fears as to the intentions of the northern authorities. These had been expressed only a month earlier in a memorandum to southern leaders: ‘it is unnecessary to give a detailed list of the outrages raids, arrests, looting, robberies etc. of the Specials . . . we are convinced, moreover, that the climax has not yet been reached . . . they are only thirsting to proceed with the extermination of the Nationalist and Catholic population’.Footnote 66 With hindsight we can dismiss the validity of such a pronouncement. The nationalist ‘pogrom’ narrative of northern unionist violence has been convincingly refuted by historians such as Peter Hart and Alan Parkinson on the basis of the conflict's reciprocity, the lack of centralised planning and the limited nature of practices of neighbourhood and workplace expulsion.Footnote 67 Even so, we should not dismiss the sincerity of the fears it conveyed. Events in Europe in the preceding decade had certainly illustrated the terrible potential for such ‘cleansings’. Throughout early 1922, moreover, the Fourth Northern Division had witnessed a steady stream of nationalist refugees fleeing the intense violence in Belfast in search of sanctuary south of the border.Footnote 68
Alongside these collective anxieties, a number of the individuals involved had their own more specific and immediate sense of grievance. Frank Aiken, for example, was a close friend of the McGuill family, and having spoken to Mrs McGuill after her ordeal he was determined to have revenge: ‘I swore that if I could take it out of the skins of the men who did it I would do it’.Footnote 69 Accordingly, he took charge of the ambush at Dromintee and was furious when it did not achieve a more substantial result.Footnote 70 Michael Creggan – brother of the Lislea victim Patrick Creggan – personally participated in the attacks at Altnaveigh and Lisdrumliska.Footnote 71 Indeed, it was this fraternal link with the IRA that determined the selection of Altnaveigh as a target. As one officer later explained, Patrick's death had ‘pinpointed attention’ to the district because Michael had been arrested for his involvement in an arms raid there in September 1920.Footnote 72 There is also strong evidence to suggest that members of the Fourth Northern Division's Corrinshego Company were particularly prominent in the reprisal party, accounting for three of the four known suspects.Footnote 73 This was significant as the four IRA men who had been killed in Altnaveigh and Lisdrumliska in July 1921 had been members of this company. As has already been noted, the attempted reprisal for these killings – which resulted in the murder of Draper Holmes – had not gone as planned. This may have added a particular sense of unfinished business to the night's proceedings.
Personal vendettas and collective anxieties go some way towards explaining the reprisals at Altnaveigh and Lisdrumliska, and most significantly the uncharacteristically aggressive manner in which they were performed. Yet it is their sectarian dimension which has received most attention. All the victims were Protestants. Specific details of the event, moreover, testify to its raw sectarian edge. When one victim pleaded for mercy, for instance, her attacker replied simply that ‘Belfast Catholics got no mercy’ and then fired further shots into the bodies of her already wounded husband and son.Footnote 74 Even so, the massacre does not conform to all that the ‘sectarian’ epithet tends to imply.
It is fair to suggest, for instance, that the immediate desire to avenge wrongs committed against comrades and relations was a more salient emotive motivational force for the massacre than some more obscure sense of sectarian antipathy. Indeed, the same could be said of many retaliatory actions in the region during the conflict. The murder of Draper Holmes and the kidnappings in the weeks before the massacre are pertinent examples. As noted earlier, these were sparked by killings and arrests of comrades. It is also worth noting that the local IRA tended not to engage in reprisals for attacks against members of the broader Catholic-nationalist community who lacked some overt link to the republican movement. Whereas the killings of the four IRA volunteers in July 1921 provoked a reaction, for instance, the killings of two nationalist civilians a week earlier had received no such response. The same was true of many other killings involving nationalists with no clear connection to the movement in early 1922, among them the four men who were killed in the greater Newry area in the immediate aftermath of the ‘Altnaveigh massacre’ itself.Footnote 75 Furthermore, the infrequent exceptions to this broad pattern were decidedly less intense. During the aforementioned reprisal at Killylea in April 1921, for example, the IRA volunteers involved were sufficiently composed to converse with their would-be victims, before deciding to spare their lives in return for information.Footnote 76 This contrasted starkly with the high emotion displayed at Altnaveigh and Lisdrumliska, where the killings and burnings were more frenzied.
The issue of motivation to one side, it is also worth considering whether the underlying logic may have been political rather than religious. In terms of victim selection, that is, it seems plausible to suggest that real or perceived political identity was a more prominent concern for the perpetrators than religion per se. This is not a novel observation in itself. As Bloxham and Gerwarth have argued with reference to violence in twentieth-century Europe more broadly, religion ‘did not play a major role as a motivational force . . . increasingly more important than belief itself was the role of religious adherence as an indicator of national identity or membership of a “civilised” community’.Footnote 77 In the Irish context, however, such a distinction is often lost amid the understandable urge to reinforce both the illegitimacy and moral abhorrence of sectarian violence. Nevertheless, its explanatory significance is suggested by events in Dundalk in the week that followed.
III
Throughout the revolutionary period, the sectarian dimensions of the conflict had been less pronounced in north Louth than in Armagh and south Down. The area was overwhelmingly Catholic and nationalist; the Protestant community was a small minority. Consequently, no organised unionist resistance had emerged in opposition to the IRA. There had also been no comparable attempt to utilise local unionists as a police force, as had been the case in Ulster with the USC. Accordingly, local Protestants had remained somewhat peripheral to the conflict, with the notable exception of the burnings in Dundalk in August 1920. In the spring and early summer of 1922, however, that community was increasingly finding itself subject to violence and intimidation amid the vacuum created by the disbandment of the RIC, the withdrawal of the British army and the absence of any suitably efficient alternative system for the maintenance of law and order.
In March notices appeared in Dundalk threatening reprisals as a response to continuing loyalist attacks on Catholics in Belfast. As in August 1920, the professed justification for this was the view that local Protestants had not expressed sufficient condemnation for the actions of their co-religionists in Northern Ireland. Later the same month, three members of Fianna Éireann – the republican equivalent of the boy scouts – were also found guilty by a republican court for raiding Protestant homes near Ravensdale.Footnote 78 In the days that followed the ‘Altnaveigh massacre’, however, the situation escalated further. In the early hours of 22 June men armed with rifles opened fire on the homes of five Protestant railway workers in Dundalk. The attack was likely informed by a long running dispute concerning the dismissal and eviction of a Catholic employee during the 1920 Irish railway strike, during which transport workers had refused to convey British troops or munitions.Footnote 79 This was followed by a further incident on the evening of 24–25 June, when a Catholic home was ‘fired into by mistake for a Protestant one’.Footnote 80
These were unauthorised attacks carried out by members of the anti-Treaty IRA, which had emerged as a distinct entity in March 1922 when the organisation formally split over the terms of the treaty. Locally, the Dundalk-based First Brigade had shifted its allegiance to the newly formed anti-Treaty IRA executive. In doing so, it parted ways with the remainder of the Fourth Northern Division. The latter adopted a neutral stance on the issue, though it ostensibly remained under the authority of the provisional government. Realising that he could neither trust nor control his own men, however, the local anti-Treaty commander, Patrick McKenna, approached the Fourth Northern Division for assistance. It was in this context that the above detailed efforts to safeguard the local Protestant community were instigated. Aiken, meanwhile, issued a stern circular to all IRA units in the area (regardless of allegiance) to clarify the official position on such attacks:
Acts of aggression by Protestants in the North against Catholics do not give anyone the right to interfere with Protestants in the South. Any such acts are morally wrong and are damnable wrong to the good name of the country, and the men who fought a clean, honourable fight. . . . The North-East Ulster Protestants are quite right to have a parliament of their own if Protestants in the rest of Ireland are not protected. . . . There are only a few at this work, and it must be stopped. We men of the IRA must stop it by shooting if necessary the men who are simply making a gift of Ulster to England, and bring dishonour to the good name of our country.Footnote 81
Overall, the intervention seemed to have the desired effect. The rising tide of sectarian violence and intimidation in the area receded considerably. There would be a few sporadic attacks on Protestants after the outbreak of the civil war, including at least one aggressive raid on a Protestant household for provisions in August 1922 and the burning of two ascendancy homes as part of a larger series of reprisals against real or perceived ‘Free Staters’ following the executions of republican prisoners in early 1923.Footnote 82 Yet these were fewer than might be expected, particularly given Louth's emergence as one of the most violent theatres of the civil war outside the anti-Treaty strongholds of Munster.Footnote 83
Events in Dundalk reinforce the view that the local IRA's pursuit of violence against members of the Protestant community followed a political logic. In north Louth, where Protestants were a politically impotent minority, they did not come to be viewed as legitimate or justifiable targets. As Aiken's circular indicated, this was informed by a complex mixture of moral and political considerations, as well as self-attributed notions of honour. Sectarian violence and intimidation did occur in the area, and at times IRA volunteers were certainly involved, but it was not orchestrated or condoned by the organisation to the same extent as it was in Northern Ireland. By contrast, north of the border in Armagh and south Down, attitudes towards the Protestant community had radicalised to the point that any Protestant – as a real or perceived representative of the dominant unionist community – was liable to be viewed as a legitimate target for reprisal. As one local IRA officer later put it:
By sending our columns into certain areas we would improve the morale of our own civilians and make the unionist civilians (if any unionists could then be classed as civilians) realise that even in their own districts they were not immune from punishment for the misdeeds of their relatives serving in the B Specials.Footnote 84
This was, in essence, the same logic of ‘representative violence’ described by Frank Wright with regard to the more recent ‘troubles’ in Northern Ireland – a means of deterrence in which ‘anyone of a great number of people can be punished for something done by the community they come from’.Footnote 85
In practice, IRA violence in Armagh and south Down in June 1922 retained a modicum of selectivity. There were specific reasons why the townlands of Altnaveigh and Lisdrumliska were chosen as a target above any other unionist district. Nevertheless, the reprisals were guided by the same system of social and religious profiling. The victims were Protestants, thus they were assumed to be unionists. As unionists they were further assumed to be fervent (perhaps fanatical) supporters of the northern government and partition, and thus, by extension, the activities of the USC. Indeed, the menfolk of the district, being of appropriate age, and mostly members of the Orange Order, may also have been assumed to be ‘B Specials’.Footnote 86 It is worth recognising, however, that such assumptions were not necessarily indicative of the exceptional sectarian pathologies of IRA volunteers, though the presence of such personalities cannot be ruled out entirely. Rather they can be considered a reflection of their somewhat one-dimensional understanding of their Protestant neighbours. As Rosemary Harris suggested in her seminal anthropological study of inter-communal relations in an Irish border town, such stereotyping was readily facilitated by the social separation of the two communities and a lack of depth or substance in cross-communal encounters.Footnote 87
Had the conflict in Armagh and south Down continued beyond June 1922 it would be plausible to suggest that IRA attacks would have become increasingly indiscriminate. In the event, however, the reprisal at Altnaveigh and Lisdrumliska marked the zenith of the inter-communal conflict in the region. There was further sectarian violence during 1922. As noted earlier, the reprisal itself was swiftly followed by the killings of four Catholics in the greater Newry area, and sporadic attacks on nationalists in south Armagh and south Down occurred for the duration of the year.Footnote 88 The IRA, meanwhile, was responsible for at least one further sectarian killing – the abduction, murder and secret burial of a Protestant publican named William Frazer.Footnote 89 Yet by the end of June 1922, with the advent of civil war in the Irish Free State, the Fourth Northern Division's attentions had shifted to events south of the border. Consoling themselves in the belief that they had (in Aiken's words) ‘cured the Specials of a tendency to carry out the Belfast pogrom tactics in Armagh & Down’, all operations in the north were abandoned.Footnote 90 In the absence of republican reciprocation, the conflict's sectarian dynamic gradually petered out.Footnote 91
IV
The sense of a contradiction between the Fourth Northern Division's reprisal at Altnaveigh and Lisdrumliska and its endeavour to safeguard Protestants in Dundalk ultimately suggests the limitations of much of the current discourse concerning sectarian violence in the Irish context. The foregoing analysis of IRA violence in the south-east Ulster frontier suggests the prevailing intuitive notions bear little resemblance to the realities of events at a grassroots level. It is clear, for instance, that largely indiscriminate attacks – killings in particular – were not as prevalent as is often believed. Although in the context of Ulster, at least, the sectarian epithet has become synonymous with the idea of ‘representative violence’, such incidents were the exception, not the rule. This reflected the fact that violence of this nature only really became a significant feature of the local conflict in its later stages, the outgrowth of a longer process of escalation and radicalisation spanning almost three years of hostilities.
There is, moreover, no reason to assume that sectarianism was itself a distinctive or decisive emotive motivational force for violence. Rather, it might be more accurate to characterise it in terms of an underlying logic – more political than religious in nature and heavily reliant on stereotyping – that governed acts of retaliation and communal deterrence. This, of course, is not to say that emotions were of no consequence. On the contrary, at both the collective and individual levels vengeance and fear very often appear to have driven the pursuit of retaliatory attacks and could provide a source of heightened aggression when such violence was performed. Yet such emotion often seems to have owed more to an immediate sense of personal grievance, such as the death of a comrade or an attack on a loved one, than it did to sectarian antipathy as such. Acknowledging these complexities is not to deny or excuse the sectarianism of the local IRA campaign, or indeed the sectarianism of horrific acts of violence such as the ‘Altnaveigh massacre’. Rather, it is recognition of the fact that a more thorough appreciation of these complex local behaviours is crucial to understanding how and why such violence occurred, and perhaps more importantly, why it was not more frequent or extreme.
