Research Findings – The Irish Revolution

The results of our research make it possible to tabulate the number of deaths in each month of the conflict, and thus for the first time we can move away from the more aggregate picture provided by Hart’s fatality figures towards a more temporally-sensitive assessment revealing the level of fatalities on a monthly basis. The results are set out in Tables 1 and 2 below. The figures reveal dramatic variations from month to month in the intensity of violence throughout the conflict.

The first phase of the War of Independence in County Cork exhibited a very limited level of killing. When intentional killings of adversaries are compared to all deaths connected with the War of Independence, it transpires that self-inflicted fatalities and accidents were of greater consequence in 1919, most notably within the British army. Poor handling of weapons, munitions, or explosives also accounted for the first two Volunteer fatalities in that year. In many respects the first third of the conflict could reasonably be described as a ‘phony war’.

In terms of intentional killing it was the IRA who drew first blood in County Cork on 7 September 1919, when Private William Jones was fatally shot while about ten Volunteers ambushed some sixteen members of the King’s Shropshire Light Infantry Regiment as they were about to enter the Methodist chapel in Fermoy. Even in this instance the objective of the attack (led by Liam Lynch) was to secure rifles (rather than to kill soldiers⁴) at a time when the IRA was engaged in trying to develop its exceedingly limited arsenal. This remained a major challenge throughout the conflict. Conversely, the extent of the loss and leakage of weapons and munitions by crown forces to the IRA was an important factor in the failure to defeat this relatively ill-equipped guerrilla army, which in Cork at least depended to a large degree on crown-force weaponry and munitions.

A notable feature of all killing episodes in 1919 was that they involved just one victim, and this pattern continued down to the end of April 1920, when two RIC men were killed at Ballinspittle in the West Cork Brigade area; shortly afterwards, on 10 May, three RIC men were killed at Ahawadda Cross near Timoleague. While there were altogether five episodes in 1920 in which three people were killed, it is noteworthy that there were only two other episodes that year in which the death toll was higher. The first was the killing of RIC Sergeant James O’Donoghue on the night of 17-18 November, together with the three reprisal killings carried out by crown forces that followed hours later. The second was the Kilmichael ambush of 28 November, an altogether exceptional and untypical episode resulting in twenty deaths.⁵ The dominant pattern in the majority of 1920 fatal incidents, however, typically involved just one death, sometimes two, but seldom more than this.

IRA Military Campaign

A significant finding about the initial phase of the conflict (before July 1920), if we focus on intentional killings alone, is the lack of IRA fatalities. In the first six months of 1920 the IRA increasingly took the conflict to the RIC, targeting police stations, patrols, and individual RIC men in a brisk campaign to nullify the primary bulwark of crown law-and-order in Ireland and its primary source of local intelligence. In January 1920, IRA GHQ permitted greater local Volunteer initiative in attacking police stations. At this stage police consolidation was already underway, with heavier fortifications in more defensible barracks and with higher numbers of RIC men, along with corresponding closures of smaller and more isolated stations. But official policy was greatly accelerated by IRA barrack attacks. The assault on Carrigtwohill RIC barracks on 3 January 1920 resulted in the capture of arms and munitions, and this was followed up by five barrack attacks in County Cork in February 1920; withdrawals by police promptly led to the destruction of the empty barracks by the IRA. About half of all the police stations in Ireland had already been vacated by August 1920.

This intense campaign (resulting in many RIC fatalities) terminated the RIC presence in many parts of the county, opening up a vacuum partially filled by IRA courts and policing. Individual RIC men and their families came under greater pressure from threatening letters, physical intimidation, and boycotts in Cork than in any other county, and with at least 90 RIC men killed there between 1919 and 1921, the menacing letters and other forms of intimidation were clearly not to be taken lightly.⁶ Overall, the strategy of targeting the RIC and its resulting geographical contraction and removal from many areas (apart from motorised patrols) marked a major British reversal in the Irish Revolution; the initiative was never entirely regained in many localities, which effectively passed under IRA control.

It is apparent that the IRA instigated by far the greater number of intentional fatal attacks, revealing that the Volunteers had an advantage down to the autumn of 1920 at least, especially as crown-force intelligence as yet remained constricted. A prominent intelligence operative was identified and eliminated; the killing of this first suspected informer, Timothy Quinlisk, an ex-soldier, made spectacular headlines in the national press when his body was found in a field at Ballyphehane (just south of Cork city), riddled with bullets and labelled as a spy. He was easily unmasked through intercepted communications between Dublin and Cork city British intelligence networks.⁷ But a more systematic campaign against local suspected spies and informers got underway in the summer of 1920, and it was waged on a far greater scale than in other counties. This determined campaign contributed to the higher number of civilians killed by the IRA than by crown forces in County Cork, which was the reverse of the national pattern revealed in Eunan O’Halpin’s figures.

Response of Crown Forces

The crown-force response to the IRA was fairly muted in the first half of 1920. But following the shooting of Constable Joseph Murtagh on Pope’s Quay on 19 March 1920 (a detective who had been successfully extracting information from an IRA prisoner by using extreme methods), the RIC carried out its first major reprisal, assassinating the republican Lord Mayor of Cork city and commandant of the Cork No. 1 Brigade, Tomás MacCurtain, who as a public figure could be readily identified. His death, however, was the first intentional killing by crown forces, which continued to suffer far higher levels of intentional fatalities in most months of the conflict. Indeed, the only months when the crown forces inflicted higher or equal fatality levels on the IRA were September 1920 and the first three months of 1921. Of these four months, it was only February 1921 in which the number of IRA men killed by crown forces far exceeded those of their adversaries, and this was the peak month for Volunteer deaths. Over the conflict as a whole, we have identified 186 crown-force fatalities inflicted by the IRA as opposed to 114 IRA fatalities. The IRA also killed 83 civilians as compared to 66 civilians who died at the hands of crown forces.

Fluctuations in Intensity of Conflict

The figures in Table 1 reveal an intensification of conflict beginning in July 1920 and reaching a temporary peak in November of that year, when 44 persons were killed intentionally (the data include all those for whom the killers could be identified). This number eased back slightly in the following month, but deaths in December 1920 remained at a high level. Fatalities in the conflict rose to their peak in February 1921, when the offensive by the crown forces sharpened dramatically (they killed 33 IRA men and in turn suffered 23 fatalities). IRA attacks on suspected civilian spies also peaked in this month, when Volunteers killed as many as 22 civilians, probably in direct response to the intensification of the crown offensive, which had become far more militarised. Overall, intentional fatalities in March diminished only slightly from this peak but fell back considerably in April. This lull proved temporary as fatalities soared again in May 1921 to the second highest monthly peak of the entire conflict, while diminishing thereafter.

Hart’s data on all casualties in County Cork (including woundings) for the War of Independence follow a much more linear upward pattern throughout the conflict and right down to the Truce. Our focus on deaths alone reveals a somewhat more variable and cyclical picture, with peaks in November 1920, February and March 1921, and May 1921. Woundings dominate Hart’s chart of casualties simply because the incidence of this form of often extreme violence dramatically exceeded the number of deaths. Of course, many cases of wounding were simply failed attempts at killing, but there was evidently a difference between the ever-increasing effort to inflict casualties throughout the conflict and those delineated periods of intense violence when the results are refined to include deaths only. The reportage on woundings is much more difficult to capture consistently than that on deaths, partly because woundings were underreported and partly because the reports blurred the differences between more and less serious injuries.

Taken together, these figures do not support the contention that the IRA campaign in Cork was over by the summer of 1921, despite the undoubted intensification of the British campaign. William Sheehan’s argument that in 1921 the IRA ‘shifted their attention to softer targets than the army and police in an effort to retain some pressure on the British forces’ needs significant qualification.⁸ While it is true that the IRA in Cork certainly killed more civilians in 1921 (70 up to the Truce whom we have identified, compared to merely 13 in all of 1920), the IRA in 1921 (up to the Truce) killed 101 crown forces as compared to 91 in all of 1920. Thus British military claims that during 1921 the rebel offensive was ‘directed entirely’ in a murderous campaign against loyalists and Protestants should not be understood as meant literally,⁹ but instead needs to be examined in terms of the general escalation of violence in 1921. Even in that year up to the Truce the IRA evidently posed more of a threat to the crown forces than it did to civilians, if fatalities can be taken as a measure. The figures also reveal the greater deployment of the British army in 1921, when soldiers suffered higher fatalities than the RIC, whereas in the preceding period the RIC had accounted for a much larger share of crown-force fatalities in County Cork. These figures do not lend support to the thesis that the IRA was shifting in 1921 to softer targets. This intensification of the conflict from the end of 1920 also resulted in the crown forces killing far more civilians intentionally.

Civilian Deaths, Official Executions, and Extrajudicial Killings

The Cork Fatality Register has helped to illustrate the significant extent of civilian killings, which have drawn less attention than combatants’ deaths known through republican commemorations or through published research on crown-force fatalities.¹⁰ Though intentional crown-force killings of civilians (66) did not match the level achieved by the IRA (83), the crown record was not too far behind, and this aspect of the conflict has drawn less attention from historians.

County Cork also accounted for over half of all official executions carried out in Ireland. These 13 British executions of IRA prisoners at Victoria Barracks in Cork city in turn resulted in republican reprisals. The most extreme of these took place on 14 May 1921 under a general order from the First Southern Division, when fifteen members of the crown forces were killed across County Cork.¹¹ Collectively, this coordinated set of IRA attacks (which included actions in Kerry and Limerick) accounted for more crown-force casualties (including woundings) than any other engagements since the 1916 Rising; it also resulted in crown-force counter-reprisals adding further fatalities. Some weeks later, at the end of May, eight deaths resulted from a land-mine explosion targeting the Hampshire Regiment band near Youghal and from a subsequent crown-force reprisal. This explosion was also possibly a reprisal for the IRA disaster at Clonmult, in which the Hampshire Regiment had been directly involved.

Extrajudicial killings by crown forces and the occurrence of numerous deaths in British custody feature with some prominence in the Cork Fatality Register, which lists 41 Volunteers who were shot or died in custody (including the 13 official executions¹²); some of these Volunteers were shot dead following their surrender to crown forces. Another significant feature of crown-force killings was the shooting dead or mortal wounding of 24 Volunteers and 17 civilians for allegedly failing to halt or trying to escape (largely in 1921). While this crude strategy removed a number of Volunteers, it also resulted in the deaths of innocent civilians, alongside other civilians killed in curfew killings, random-fire episodes, and other fatal incidents.

Special Features of County Cork Killings

A striking feature of the War of Independence in County Cork was the small scale of most attacks, with the majority of episodes of killing involving just one victim, and less frequently two or more. Over the conflict at large, episodes that resulted in more than 10 deaths could be counted on one hand. They mostly occurred after the general escalation from mid-February 1921 onwards; these were the abortive Upton train ambush mounted by the IRA (15 February), resulting in 11 fatalities; the annihilation of the IRA’s East Cork Flying Column at Clonmult (20 February), resulting in 14 deaths; the successful IRA ambush at Crossbarry (19 March), also resulting in 14 fatalities; those 15 members of the crown forces killed in coordinated reprisals on 14 May; and the already mentioned 20 fatalities in the Kilmichael ambush of 27 November 1920. Less than 28 percent of total fatalities took place in incidents where over three people were killed. Thus episodes that could even be described as ‘skirmishing’ were not the general rule despite the higher incidence of such episodes starting in mid-February 1921.

Another conspicuous aspect of Cork fatalities for this period is the far higher number of documented disappearances carried out by the IRA than in other counties—largely within the Cork No. 1 Brigade area. We have been able to document 49 disappearances. A significant finding here is the sizeable minority who were members of the crown forces. (Arthur Harrison was counted as a civilian ex-RIC constable in our returns, but he had just resigned and was tracked down by the IRA for his police work in bringing about IRA convictions.) In at least two ways our evidence provides much-needed correctives to Gerard Murphy’s more speculative research on disappearances. First, our documented results indicate that more attention deserves to be given to crown-force disappearances, while a significant segment of the civilians disappeared were ex-soldiers, thus underlining the significance of crown-force networks among those suspected and fatally attacked by the IRA using this tactic. Second, the documentary evidence strongly suggests that disappearances by the IRA took place to a greater extent between the summer of 1920 and the Truce of 11 July 1921 rather than between the Truce and the summer of 1922. We therefore contend that possible undocumented disappearances are more likely to have occurred before the Truce rather than afterwards, following the wider pattern of all killings in County Cork, in which those during the War of Independence far outweighed the number during the Truce and the Civil War.¹³

We have recorded military accidents, notably fatal shootings or fatalities involving military vehicles. We include conflict-related deaths that occurred only in County Cork.¹⁴ This record captures some victims who were fatally injured outside the county but who died within it; conversely, the record excludes those who may have been mortally wounded in County Cork but who died elsewhere. We have also excluded some fatalities because the sources were inadequate for us to be certain of their deaths or because we deemed the deaths as not sufficiently related to the conflict. Neither have we included those who died long after the end of the conflict, though as a consequence of it.¹⁵

The Cork Fatality Register will be revised on a continuing basis; it is very much a work in progress. The authors welcome criticisms or new information that might add to the number of deaths properly identified or might improve the context or classification of individual fatalities already recorded. Our readers may be able to provide additional sources verifying or supplementing particular entries, justifying new ones, or perhaps even prompting the cancelation of existing ones.