Deadly Accountancy Part 2

by Michael Patrick O'Leary

Sri Lanka Eelam War IV

In an address on November 11 2011, under the auspices of the British Scholars Association at the British Council in Colombo[i], Rohan Gunaratna discussed his contention that “only” 1,400 civilians were killed in the north east Vanni pocket in the first five months of 2009. He said that this estimate was based on interviews with Tamil coroners and doctors in the area and with some of the 11,800 Tiger prisoners then held by the government.  He asserted that 1,200 were unintentionally killed by government cross-fire and 200 by LTTE gunfire.

This is in stark contrast to the figures peddled by Channel 4 and in the report by the “Panel of Experts” commissioned by Ban Ki-Moon, known in Sri Lanka as the Darusman Report.[ii] The panel did not include  military or social science experts, or anyone knowledgeable about Sri Lanka.

In Channel 4’s Sri Lanka’s Killing Fields [iii]Jon Snow doomily intoned that ”As many as 40,000, probably more” civilians  had died. Gordon Weiss[iv] , who featured in the Channel 4 programme, started with an estimate of 7,000 which became   15,000, which he then upped to  40,000,  a figure that a whole range of media outlets, including BBC and NDTV, ran with.

Weiss, an Australian who used to work for the UN in Colombo,  went on record as saying the number of civilian casualties was 7,000. This became the official figure quoted by The UN General Secretary’s New York spokesperson,  Michelle Monas, who told Inner City Press reporter Matthew Lee, “We have no way of knowing the exact count”. When Weiss left the UN and returned to Australia he increased the figure to 40,000. Journalists have confused the issue by failing to make clear whether information came from “an employee of the UN” rather than “the UN”.

The UN Hub

The UN left the Wanni at the end of September 2008, but continued to send food convoys deep into LTTE territory, returning to base at Vavuniya after each trip. On January 21 2009, a convoy delivering food to Puthukudiruppu (PTK) returned to Vavuniya after being stuck for four days because of fighting. Two UN staffers stayed back and set up an unauthorised  “UN hub” in Susantirapuram. This was in direct contravention of UN General Assembly Resolution A/RES/46/182 of 1991. The UN hub was deliberately located between two hostile military forces and the UN personnel did not follow basic UN rules for humanitarian workers in conflict zones. The Darusman  report does not name the UN personnel but in his book, The Cage, Weiss blows their cover and makes it clear that the UN officers who provided information to him and to the Darusman team were Chris Du Toit and Harun Khan, although I get the impression that Du Toit was in Colombo and not actually physically with the convoy. Weiss writes that “Du Toit would be the driving force behind the gathering of much of the intelligence revealing that large numbers of civilians were being killed”[v].

The Darusman  report says that a heavy assault on Puthukudiruppu was clearly imminent. The LTTE was firing on the army from the vicinity of the UN hub, thereby inviting the army to fire back. Civilians were encouraged to move into the danger zone by the presence of the UN handing out food. If it had not been for the UN presence the civilians could have been dispersed out of harm’s way. UN members cannot ignore the fact that UN officials took it upon themselves to set up a UN hub in the middle of a war zone, with no authorization from the government.[vi]

Views of The Cage

My own lengthy review of The Cage can be found at

Where did Weiss get his figures? Could it be from Chris Du Toit? Rajiva Wijesinha recalls meeting Du Toit: “Pressed on the number of those seen by the UN, he said it was something like 39, over the previous month.”

When he was  working  for the UN in Colombo,  Weiss  went on record as saying the number of civilian casualties was 7,000. The UN used that figure. In The Cage Weiss recalls a press release by UN Human Rights Commissioner Navi Pillay saying that “as many as 2,800 civilians “may have been killed”. Weiss gives this spin: “Critically, the civilian death toll Pillay quoted finally established a baseline that had some kind of official imprimatur and weakened government efforts to confine solid numbers to the realm of speculation and confusion”. Pillay’s statement did not take us out of the realms of speculation because she said “as many as 2,800 may have been killed”. That is speculation. What does establishing a “baseline” mean? Does it mean that because Pillay says “as many as 2,800 may have been killed” that gives Weiss licence to say 10,000 to 40,000 and Frances Harrison  to say 147,000?

Gordon Weiss’s lower  estimate of 7,000 civilian deaths, made in 2009, was challenged by Sir John Holmes, the UN Under-Secretary-General for Humanitarian Affairs and UN Emergency Relief Coordinator, who stated in New York on 24 March 2009 that this figure could not be verified. In spite of this, Weiss throughout The Cage routinely talks of “between 10,000 and 40,000”,  which is a meaningless mantra and statistically useless.

Weiss was not a witness. Like an urban myth or an internet hoax, a story gets passed around and is treated as legal currency. The neologism “churnalism” has been credited to BBC journalist Waseem Zakir who coined the term in 2008. “You get copy coming in on the wires and reporters churn it out, processing stuff and maybe adding the odd local quote.” Stephen Colbert coined the term “truthiness” – “We’re not talking about truth, we’re talking about something that seems like truth – the truth we want to exist” [vii]

A Sri Lanka Media Watch review[viii] of The Cage quotes other sources which estimated different figures to those provided by Weiss. In February 2009, the US Embassy noted that the pro-LTTE “Tamil National Alliance parliamentary group leader R Sampanthan claimed that 2000 Tamil civilians have been killed and 4500 injured since mid-December….Such reports from Tamil sources cannot be confirmed and are frequently exaggerated.”

The Voice of Tigers, the LTTE’s “official radio”, claimed on 1 March 2009, that the Sri Lankan armed forces had been responsible for the deaths of 2,018 Tamil civilians in January and February 2009 in the Vanni. These figures were repeated by UN High Commissioner for Human Rights, Navanethem Pillay, on 13 March 2009. Sir John Holmes, the UN Under-Secretary-General for Humanitarian Affairs and UN Emergency Relief Coordinator, stated in New York on 24 March 2009 that this figure could not be verified: “The reason we have not come out with this as our figure is because, as I have said before, we cannot verify it in a way that you want to be able to verify, if you put it as your public figure.”

Data compiled by the South Asia Terrorism Portal, data “primarily based on figures released by the pro-LTTE Website Tamil Net”, put the casualty figure for civilians inside Mullaithivu at 2,972 until 5 April 2009.

Michael Roberts

Professor Michael Roberts considers Gunaratna’s statistical estimates “astounding and misleading”. This is mainly because the conditions of battle at that time and the difficulty of distinguishing between “civilians” and “combatants” at a time when the LTTE was rounding up everyone to the cause, makes such precision spurious.

Rajasingham Narendran asked:  “how many coroners were available during the war in the area for recording deaths? “ Narendran had talked to IDPs who had fled the last No-Fire Zone in April 2009 and later with IDPs at Menik Farm and elsewhere.  “My estimate is that the deaths — cadres, forced labour and civilians — were very likely around 10,000 and did not exceed 15,000 at most”

Muttukrishna Sarvananthan of the Point Pedro  Institute told Roberts “[approximately] 12,000  [without counting armed Tiger personnel] “.

Dr. Noel Nadesan: ““roughly 16,000 including LTTE, natural, and civilians”. Note that Nadesan includes fighters and natural deaths. In any population, a number would die from natural causes of ill health or medical misadventure at child birth or operation. Roberts believes that 600 deaths from natural causes would be a reasonable estimate for the area and the time-frame, but it could be more because of stress and shortage of food.

In The Island, [ix]Professor Roberts wrote: “Within their attentiveness to the approximate character of any assessment, there is a striking agreement in their computations. Their evaluations also dismantle Rohan Gunaratna’s estimate on the one hand and, on the other, reveal the exaggerated character of the figures peddled by the Darusman Report, Channel Four and HR bodies abroad. In the latter instance it is both travesty and paradox that moral fundamentalism has encouraged extremism in factual claim in ways that serve the goals espoused by these organisations.”

Someone commented on Transcurrents:

“Dear Professor Michael Roberts, Rohan Gunaratna, Sarvananthan, Noel Nadesan, R. Narendran and you have been living abroad and actually over-estimated the deaths. Your vision far away from Vanni only made you people to come to such an over-estimation. I live a few kilometres from Vanni boundary. I have spoken to a number of people who were trapped in Vanni and escaped death. The majority of them told me that ONLY about 150 (hundred and fifty) civilians WERE KILLED IN THE VANNI WAR. THEY ALSO SAID THAT THE SUPPORTERS OF THE TIGER TERRORISTS ONLY give over estimated figures. We living in the North are not fools to accept your over-estimated figures!”[x]

Lies and Statistics of the Damned

In the documentary Lies Agreed Upon,[xi] it is argued that 40,000 deaths would be physically impossible. The population of the Wanni is estimated as a maximum of 300,000. This figure is based on LTTE records which were probably inflated and many would have left the area.  293,800 people were registered at the receiving centres. That leaves a maximum of 6,200 to be accounted for. Around 5,000 SLA soldiers were killed.

The government maintained for a long time that there were no civilian casualties, expressing moral outrage at the very concept of “collateral damage”. That position has modified over time to a claim that everything possible was done to avoid civilian casualties. Gotabaya Rajapaksa told the Sunday Leader: “What we did say was a ‘zero civilian casualty’ policy! That was what we were aiming at. That was what we told troops. Our goal was to achieve that”. [xii]See the Ministry of Defence’s Humanitarian Operation Factual Analysis .[xiii] However, the government PR machine and the inept foreign office and diplomatic service have allowed western critics to take sole possession of the numbers game by failing to come up with its own numbers. LLRC has recommended further investigation of certain incidents that witnesses say happened.[xiv]

Sense and Census

Gotabaya Rajapaksa said the government has made a proper assessment of the number of civilians killed and missing during the last stages of the conflict. Arbitrary figures of between 10,000 and 40,000, he insisted, had “no basis in reality.” An  assessment was done by the Department of Census and Statistics through Tamil public officials in the relevant districts of the North and East.[xv] The questionnaire specifically addressed the issue of people who died or went missing during the ‘humanitarian operation.’

The government has identified by name all such persons, Rajapaksa said. The results of the census will be released in the near future. He said that some  people died of natural causes and of accidents, some  died whilst fighting as members of the LTTE, some  died as a result of being coerced to fight by the LTTE, some  died as a result of resisting the LTTE and some died because of military action. “It is only for the deaths of people in this last category that the Sri Lankan military can bear any responsibility.” The defence secretary also maintained that if, in future, any substantial evidence is provided about crimes committed by its personnel, the Sri Lankan military will not hesitate to take appropriate action.

The Numbers Game: Politics of Retributive Justice

This is a discussion paper by IDAG-S (Independent Diaspora Analysis Group – Sri Lanka). According to Michael Roberts: “The key hand is a person who wishes to remain anonymous and can be called ‘Citizen Silva.’ Born to Sinhalese parents, raised and educated in the West, he has spent the entirety of his life outside the island. This foreign setting has enabled him to build close personal links with the island’s other ethnic diaspora groups, thus shielding him from the communalistic shadows that overwhelm many of his compatriots back home. As the analysis of the satellite imagery reveals, his engineering background allows him to bring to the examination a range of technical skills not usually associated with the average empirical scientist.”

IDAG-S’s analysis only considers events and the actions of the warring parties (Sri Lankan State and LTTE) leading up to and including May 2009.

The main purpose of the survey is to estimate how many civilians were killed. “Whilst it is widely accepted that the fighting during the last few months was brutal, and that there were potentially many civilian casualties, the aim of this discussion paper will be to examine in detail the accuracy of some of these larger fatality estimates.”

The paper mentions different estimates from UN sources. In June 2010, the UN Secretary General appointed  a Panel of Experts to advise him. In their report  they gave an estimate of 40,000 deaths. Critics in Sri Lanka would quibble at the IDAG-S’s description of this as a UN report. It has generally been referred to in Sri Lanka as the Darusman Report. In November 2012, the UN published a report which pushed the estimate up to 70,000.

The IDAG-S paper quotes some surprising comments from Wikileaks.  Jacques de Maio, ICRC’s (Red Cross) Head of Operations for South Asia,  said that the Sri Lankan military was somewhat responsive to accusations of violations of International Humanitarian Law and was open to adapting its actions to reduce casualties. “He could cite examples of where the Army had stopped shelling when ICRC informed them it was killing civilians. In fact, the Army actually could have won the military battle faster with higher civilian casualties, yet chose a slower approach which led to a greater number of Sri Lankan military deaths.” [xvi]Even Robert O Blake, noted in a confidential embassy cable[xvii] : “The Army has a generally good track record of taking care to minimize civilian casualties during its advances…”.

From available media reports and other sources of information from the conflict zone during this period, it would appear there were no complaints or accusations directed at the Sri Lankan military for causing significant civilian casualties before September 2008. The Government claim that civilian casualties were minimal was widely accepted by the international community as being true, and was not challenged in international forums.

What Is a Civilian?

The IDAG paper explains that defining “civilian” is not easy.  According to Article 1 of the 1938 ILA Draft Convention for the Protection of Civilian Populations against New Engines of War, the phrase “civilian population” within the meaning of this Convention shall include all those not enlisted in any branch of the combatant services nor for the time being employed or occupied in any belligerent establishment as defined in Article 2.”

A ‘civilian’ undertaking any activity that ‘helps/contributes/ advances’ the military ‘goals / objectives’ of the LTTE – like sentry duty, building bunkers / bunds / trenches or transporting military material – cannot then enjoy the protection this category (civilian) is afforded under international law in a conflict situation.

In 2006, the LTTE maintained roughly 25,000 trained cadres. As the conflict progressed, the LTTE escalated its recruitment process, forcibly recruiting and training many more civilians, including child soldiers. At the start of 2008, the Sri Lankan Army estimated that the LTTE had within its ranks approximately 30,000 cadres.

The LTTE abused the No- Fire-Zones created by the Sri Lankan Army, to allow civilians to escape the effects of hostilities. By refusing to acknowledge the protective character of these areas and by deliberately using them for military purposes, their status as a protected space under international law became null and void. As a direct consequence, the LTTE denied the civilian population under its control the best means of shielding itself against the effects of war. The LTTE with increasing regularity fired  from near schools, hospitals and IDP settlement clusters, alongside using hospitals as bases of operation and storing weapons in and around IDP settlement areas. This was all in contravention of clear and specific prohibitions of international law. Armed LTTE operatives routinely mingled with civilians in order to cover their movements and launch attacks against the Sri Lankan Army.

“The Sri Lankan military could not forego a legitimate military objective without undermining its mission and putting at serious risk both its soldiers and the wider Sri Lankan civilian population. In those circumstances, the result of the LTTE approach was to make it difficult, and sometimes impossible, for the Sri Lankan military to avoid harm to civilians and civilian structures.”

The IDAG paper cites Kenneth Watkin (Canadian Judge Advocate General who presided over the workings of the Israeli Tirkel Committee investigating the Israeli attack on  the Gaza Aid flotilla in May 2010)”[a]lthough civilians are not to be directly made the object of an attack, humanitarian law accepts that they may be killed or civilian property may be damaged as a result of an attack on a military objective.”

According to L. Oppenheim’s International Law: Disputes, War and Neutrality “civilians do not enjoy absolute immunity. Their presence will not render military objects immune from attack for the mere reason that it is impossible to bombard [the military objects] without indirectly causing injury to the non-combatants.”

Critics have concluded that there was a deliberate strategy on the part of the Sri Lankan Government to intentionally target and kill civilians. The IDAG-S point out that the Darusman Panel denies  any humanitarian intention despite the Panel’s own account of (a) how soldiers, at the risk to their own lives, had helped countless civilians attempting to escape the war-zone, and (b) their account of how the Army penetrated the second No-Fire-Zone, incurring heavy casualties amongst its troops, to rescue over 100,000 people.”


Eelam War IV  has often been described as a “war without witnesses” on the grounds that foreign reporters and NGOs were discouraged from entering the war zone. The authors of this study have accessed witnesses. Here is how they describe their methodology. Their evidence is gathered from the following sources:

Eyewitness Testimony – Culled from a large body of interviews conducted by the UTHR(J) (University Teachers for Human Rights (Jaffna)) [xviii]team in 2009 with persons from the conflict zone. The UTHR(J) was one of the few organisations to interview people in Sri Lanka, and at the same time cover a broad spectrum of views. They also provide valuable historical data through their informants in the Vanni about the state of affairs in the region from 2006 to 2009. [xix]

Satellite & associated imagery – Images from the conflict zone covering a period from 2007 to 2009.

Wikileaks – Reveals information related to the conflict from sources / contacts maintained by the U.S. Embassy within Sri Lanka and other postings outside the island.

Documentary Information – Covers information gleaned from various reports / articles / statistics.

Media Groups – Information about the conflict from media coverage during the conflict and after.

Information from human rights groups and related organisations – Covers reports / analyses / eyewitness accounts – these sources are usually independent to the ones used by the UTHR(J).

UN – Data from various UN departments in and outside Sri Lanka.

Numbers Game 

Fatality estimates, issued by various international bodies / persons after December 2009, have used  the same empirical data first sampled by the UTHR(J) team during the latter half of 2009. The results are wildly different, ranging from a low of 7,700 by the US State Department to 147,000 by  Frances Harrison[xx].

The UTHR(J) method was to subtract  from the population claimed to have been present in the second No-Fire-Zone in February 2009 – a figure close to 330,000, the final number of registered IDPs in Government camps by May 2009 – roughly +290,000.

According to IDAG-S, the validity of this framework for estimating the number of civilian fatalities depends on:

  • The accuracy of the estimate for the number of people in the second No-Fire-Zone in late February 2009.
  • The ability to differentiate between combatant and non-combatant fatalities.

The reality is that for the most part, the weight of all the evidence of mass-scale fatalities – these alleged numbers of people killed being the primary drivers behind the need for an international investigation; rests on the accuracy of a single figure – 330,000.

That figure was supplied by Assistant Government Agent  Parthipan. He did not do an individual head count. Parthipan’s estimate was based on discussions with village headmen, the Grama Niladharis. The GNs in the Killinochchi and Mullaithivu districts worked very closely with the LTTE: “The government’s social welfare measures, ration cards, identity cards, and voters lists are all routed through the headman. From the 1990s, the LTTE has used the headmen under its control to police the people, force them to attend demonstrations, perform compulsory military service as auxiliaries, impose punitive cuts of rations, diddle government aid and report on those coming in and going out…”[xxi]

The IDAG concludes that number of people in the second No-Fire-Zone towards the end of February was a figure considerably smaller than the 330,000 quoted by AGA Parthipan. This figure would also have included designated combatants like enlisted and forcibly recruited personnel. According to Parthipan, from March 31 to April 29 40,340 people were unaccounted for.

April ICRC data for the number of injured persons transported by sea shows that the mercy missions removed around ~2,800 injured civilians from the No-Fire-Zone. Sources in the conflict zone also claim that only 50% of the total numbers injured were transported by sea. This would mean that for April alone there were potentially 5,600 combat related injuries. for the whole of April 2009, TamilNet reported that there were roughly 2,600 civilian fatalities.

Comparing high-resolution satellite images of the second No-Fire-Zone between February and April 19, indicates that the No-Fire-Zone as a whole did not witness anything like the scale of sustained bombardment required for there to have been more than 40,300 fatalities. The UN estimated that for April there were potentially between 3,000 – 3,900 civilian fatalities. Assuming a worst case scenario where there were 5,000 civilian fatalities, and an additional 2,000 LTTE cadres and conscripts killed in April. This would still leave an unbridgeable deficit of close to a staggering 33,040 unaccounted for.

IADG-S calculates that the number of those who escaped from the conflict zone or detention centres would have ranged between 3,000 and 6,000 and  at least 10,000 LTTE combatants and auxiliaries were killed in this period. Up to 15,000 truly civilian people were possibly  killed in the conflict zone during the last five months, with an additional 2,000 – 3,000 having died by either being shot, shelled or having drowned whilst trying to flee the battle zone. “The respective proportion of civilians killed by the LTTE and the government forces is difficult to work out. Though it is probable that more were hit by government fire than by the LTTE, the latter’s ‘work’ in this sphere was not small.”

“Although the Panel report stated from the onset that there was no authoritative figure for civilian fatalities during the final phases of the war: only assessed evidence – or interpretations of it – that it felt reinforced its primary hypothesis, that there were tens of thousands killed. Whilst at the same time other sources of credible evidence that contradicted these assessments were either by intent or sheer negligence, ignored.”Some reliable witnesses and other IDPs who were present when the Army entered on the 18th May are certain that a large number, perhaps the majority, of those killed in the NFZ during the last 12 hours were killed by LTTE shelling. Shells were falling into them and from the direction they are certain that they were fired by the LTTE”.

The IADG-S report says: “that civilian deaths and injuries from Government Forces firing did occur is indubitable, but one has to be cautious in concluding intentionality from such a result without having studied each incident in detail and taken into account issues like: (a) the conditions ruling at the time of the attacks; (b) whether the commander ordering the attack believed his actions would cause clearly excessive levels of civilian harm in relation to the anticipated military advantage gained; (c) the reasons behind the choice of weapon used in a vast majority of the attacks – mortar as against artillery, rockets and airstrikes; (d) considered the military advantage gained as being part of the overall military objective of which the attack was a part.”

Conclusions are further complicated by the fact that the LTTE killed civilians on several occasions when they sought flight.  Again, computing statistics on fatalities caused by Sri Lankan Army action is complicated by the fact that many LTTE fighters did not wear fatigues and thus deliberately contravened the protocols of war that enjoined the principal of distinction. This in turn makes the identification of a civilian corpse into a questionable issue in a significant number of instances.

Accountancy and Accountability

In 2009, the Banyan column in the Economist said:[xxii] “It is probably too much to hope the government might adopt a fresh approach to these familiar allegations. There were always at least three ways to tackle them. It could, early on, have argued brazenly that the benefits of ending the war outweighed the cost in human life. The Tigers were as vicious and totalitarian a bunch of thugs as ever adopted terrorism as a national-liberation strategy. Or the government could have insisted that its army’s behaviour was largely honourable, but that some regrettable abuses may have occurred, which would be thoroughly investigated.”

 IADG-S consider that some critics , such as Frances Harrison and Alan Keenan have moved “into the realms of statistical fantasy in ways that raise questions about their integrity / morality”. “It would seem that such spokespersons are motivated by moral rage and retributive justice. They seek regime change in Sri Lanka – a form of 21st century evangelism that is imperialist in character and effect.”

Many in Sri Lanka would argue that even if the allegations about these incidents were proved these are small crimes given the context of a long and difficult war which resulted in a peace unknown for thirty years. The incidents  do not seem to fit the Nuremberg criteria. They do not compare in magnitude to the war crimes perpetrated by the USA and UK over the decades and more recently in Iraq and Afghanistan. The USA supports Israel which persistently assassinates Palestinian leaders wherever they can find them. The CIA tried many bizarre methods of assassinating Fidel Castro. Navy Seals succeeded in killing Bin Laden and dumped his body in the sea. The New York Times reports[xxiii] that President Obama presides over Tuesday Terror sessions at which he personally selects victims, including US citizens, for “targeted  assassinations. The USA blatantly ignored the Geneva Conventions and abducted innocents to torture them in foreign countries. Rather than being punished those responsible are still free to sign lucrative book deals for advocating and practising torture.

In Sri Lanka’s case, controversial estimates of civilian deaths were introduced not as irrefutable facts, but as circumstantial evidence to lay the foundation for an international investigation and ultimately regime change.

However, IDAG conclusion states clearly: “Nothing in this survey denies the probability and the evidence that some extra-judicial killings of high-ranking LTTE officers occurred during the last days of the war. These actions need to be impartially investigated by an independent body, and where possible criminal indictments pursued against the perpetrators.”

There is a strong case for accountability and recognition of the loss of life. The current situation does not hold out much hope for  genuine reconciliation. Naming and shaming on the basis  of exaggerated numbers is not the way to persuade the Sinhalese community to recognise  the loss of life amongst the Vanni Tamils.  Bludgeoning them with  inflated numbers could lead to a backlash.

[iv] The Cage: The Fight for Sri Lanka & the Last Days of the Tamil Tigers

[v] The Cage p 107

[xx] Still Counting the Dead: Survivors of Sri Lanka’s Hidden War by Harrison, Frances

[xxi] UTHR(J) Information Bulletin No.39, 1 November 2005.