So, if you’re someone who pays attention to the former Yugoslavia at all, then unless you’ve been living under a rock for the past few months, you’ll be aware of the award of the Nobel Prize in Literature to Peter Handke and that he is an unapologetic Bosnian genocide denier. I don’t intend to re-invent the wheel here with regard to Handke – the excellent work of Peter Maass for The Intercept speaks for itself. But inevitably, the Handke scandal has resulted in the dredging up of other genocide deniers and their claims on social media. Just to cite my own experience, I ended up in a rather unproductive thread with someone citing the writings of the American author Edward S. Herman.
Herman, who died in 2017, is probably best known for his collaboration with Noam Chomsky on the 1988 book Manufacturing Consent. Which I suspect, for most people who’ve heard of his name, is all they know about him. But Herman is also a hardcore denier of the Bosnian and Rwandan Genocides. What makes him different from Handke, however, is the difference in style and effort involved. Handke’s denial might be called casual. It is usually quick and brief, consisting largely of assertions that certain atrocities did not happen, which he makes little to no effort to substantiate. He is writing, in other words, for the choir – those who already agree with his conclusions and are already primed to accept any assertions he makes supporting them. Herman is different, writing entire long articles with significant details and citations. To compare them to Holocaust deniers, if Handke is a random Nazi sympathiser on Stormfront, Herman is the Institute for Historical Review, attempting to build up an entire pseudohistorical counter-claim. Rather than preaching to converts, Herman wrote for those who know little about the former Yugoslavia, but who do know him and the publications he wrote for, and so may be convinced by the deceptive appearance of serious research.
But it is not serious research. Herman’s use and interpretation of sources is not merely poor, but actively dishonest. I don’t intend here to go through every line and claim Herman has ever made about the Yugoslav Wars, but I would like to walk through a handful of his more egregious abuses of the source base, both to show how bad historical methodology works, and of course to debunk some common denier talking points.
NB: Some of the articles and works cited in this article are authored by Herman alone, while some are co-written with a regular collaborator, David Peterson. Because Peterson is virtually unknown outside his collaborations with Herman – see Monthly Review’s corresponding descriptions of the two below for a rather striking contrast! – and the claims and style change little regardless of whether Herman is writing alone or not, I’m going to largely ignore him in this article, rather than specify each and every piece he was a co-writer on each time I mention it. Just be aware that he did also contribute to some of them.
1. The number of Serbs killed in raids during the Siege of Srebrenica
I’ll begin with a claim Herman makes in a book to which he both contributed and edited, The Srebrenica Massacre: Evidence, Context, Politics (PDF in link). A passage on p.14 of the Preface, written by Herman himself, more or less establishes his basic thesis:
“We certainly believe that there were a significant number of executions
at Srebrenica following the evacuation of the designated Srebrenica “safe
area” in July 1995. But we also believe that they may not have been
more numerous than the Serb civilians killed in the towns near Srebrenica by Bosnian Muslim forces operating out of that “safe area” in the
prior three years (well over 1,000, with one estimate as high as 3,287)”
To provide some minimal context here, I will briefly note that prior to the Srebrenica Massacre in July 1995, Srebrenica had been besieged by Serb nationalist forces (primarily the Republika Srpska Army, or VRS) for three years. Herman incorrectly portrays the UN designation of Srebrenica as a “safe area” as lasting this whole time – in reality, this was only introduced in April 1993. During this period, from time to time the local Bosnian Army division (the 28th Division) in Srebrenica (referred to by Herman as “Bosnian Muslim forces”) undertook raids into the surrounding VRS-controlled territory. A very common denialist talking point has been to exaggerate the level of death and destruction inflicted on local Serbs by these raids, while also playing down the number of Bosniaks killed in July 1995 in order to create the appearance of two equal atrocities.
There are a vast array of reasons why Herman’s framing of these events, and what he chooses to emphasise and omit, is in itself misleading. But for today, I simply want to focus on the numbers he claims, as they are a prime example of how Herman uses sources.
Herman’s source for the figure of 3,287 Serb dead is a relatively obscure Serbian author, Milivoje Ivanisevic and his work Srebrenica July 1995—in search of truth. Herman repeatedly cites Ivanisevic for this – in pretty much every article he has ever written on Bosnia – and no-one else. Despite his reference to Ivanisevic’s figure as merely “one estimate” and providing a lower bound of 1,000 (thus implying there are other estimates closer to 1,000), no other estimates are ever cited.
There is an obvious reason for this – there are no such other estimates. An ICTY Press release in July 2005 provides two internal Republika Srpska sources providing much lower numbers of 641 and 995 respectively. The only source they note which supports Herman’s claim of “well over 1,000” is a book claiming a 1,200 figure (total, both military and civilian) by… Milivoje Ivanisevic. The book Herman cites is in fact Ivanisevic’s second claim in which he makes another attempt to big up the numbers. It’s also worth pointing out that the ICTY press release already notes that even the 1,200 estimate is based on erroneous information – he claims that the Kravica attack by the 28th Division in January 1993 resulted in the massacre of virtually the entire village of Kravica, around 350 people, whereas internal RS records show that the actual death toll of the attack was less than 50.
Ivanisevic is the only source available to Herman (who had no topical expertise on Bosnia, and appears to have not read Bosnian/Croatian/Serbian at all) which he can use to claim such high numbers of Serb deaths. But a generous interpretation could still say that Herman is simply lazy rather than actively dishonest – that he simply doesn’t know of any other sources, and never bothered to look for them, right? Not so – Herman makes it clear that he is aware of other sources. He cites on multiple occasions (see also p17n8 of The Srebrenica Massacre PDF) perhaps the most authoritative source on Bosnian War casualties, The Bosnian Book of the Dead by Mirsad Tokaca and the Research and Documentation Center in Sarajevo. Yet apparently it never occurs to him to check what Tokaca and the RDC have to say about Serb deaths around Srebrenica. Probably because they also support a much lower number – 870 Serb civilians killed across the whole region of Podrinje (i.e. most of Eastern Bosnia, not just the area in the immediate vicinity of Srebrenica) in the whole war.
Even when we limit our analysis purely to Ivanisevic, Herman plays fast and loose with the information he relates from the source. For instance, in the initial quote, Herman describes the figure of 3,267 deaths as all being civilians. Elsewhere, however, he cites this figure (altered to 3,262, but I’ll put this down to a typo or minor variation from Ivanisevic) as the total Serb dead regardless of combatant status, with 2,383 being civilians. But even here, Herman misrepresents what Ivanisevic says. Compare Herman’s description:
“Serb historian Milivoje Ivanisevic has produced detailed lists of Serbs from the Srebrenica and Bratunac municipalities killed during the years 1992-1995; they total 3,262 victims in all, including 2,382 civilians (73%), and 880 police (27%).”
to that of Ivanisevic himself:
“The given list contains 3262 Serbian victims. According to previous data about 27%, or 880 killed persons, were the members of military and police forces. The rest of them, or 73%, i.e. 2383 victims, were civilians.”
Now ignoring the fact that Ivanisevic’s maths is off by one, there’s a blatant sleight of hand here – he claims that 880 victims were either soldiers or police, but Herman omits the former, and claims that all 880 were police. Even in a source cherry-picked to support his argument, he can’t resist the temptation to “civilianise” military victims.
One final note on Ivanisevic before moving on. As noted already, in the initial quote from The Srebrenica Massacre that began this section, Herman claims that the killings of Serbs in question took place during the period when the Srebrenica enclave was dubbed a “safe area” by the UN. He repeats this elsewhere in an interview (where he doesn’t explicitly mention Ivanisevic, but is clearly citing his figures) that “all those killings of Serbs took place coming out of an area that was supposed to be a “safe haven”.” Herman is fully aware that the “safe area” policy only began in April 1993, given his demand that “the surviving relatives of Serb victims of Bosnian Muslim attacks launched from the “safe area” from late April 1993 to July 1995 should also have been able to sue the Dutch for damages.” Apart from the fact that Herman doesn’t seem to know a crucial fact about the Siege of Srebrenica – that the infamous Dutch battalion of UN soldiers didn’t arrive until early 1994, and that during 1993 the “safe area” had been the responsibility of a Canadian battalion instead – Ivanisevic, far from supporting this claim, contradicts it; he says that the vast bulk of Serb deaths in the region – 71% of those in Srebrenica municipality, and 64% of those in Bratunac municipality – took place in 1992, before the start of the “safe area” policy. And these numbers don’t include the January 1993 Kravica attack which, in Ivanisevic’s aforementioned inflated numbers, accounts for about ~10% of the total Serb dead.
Having thoroughly gone through this particular claim, we can now, mercifully, move on to something else, where Herman’s dishonest is even more on display….
2. The number of Serb villages around Srebrenica destroyed by the 28th Division
This is the same focus as the previous point, just emphasising the number of villages rather than deaths among the Serbs around the Srebrenica enclave. Herman, on numerous occasions, quotes the following from Dutchbat commander Thomas Karremans:
“We know that in the area surrounding the Srebrenica enclave alone, 192 villages were razed to the ground and all the inhabitants killed. That’s what I mean when I say ‘no good guys, no bad guys’.”
He cites this to p. 2259 of the Dutch Srebrenica Report. If you open it and turn to page 2259, true enough, there the quote from Karremans is, right at the bottom of the page. However, if you read on even a few sentences, you’ll notice the second paragraph near the top of the next page, p. 2260, which reads:
“There had indeed been attacks on Serb villages and settlements in 1992 and 1993, led by Naser Oric, in which many people died. However, the figure of 192 mentioned by Karremans was on the high side. Since his source is unclear, it is not possible to work out what he meant by the term ‘villages’. The pattern of settlements in the area around Srebrenica was characterized by a large number of small settlements, often consisting of only a few houses but which did have their own names. The most reliable estimates indicate that in 18 months of attacks, around thirty Serb villages and seventy of these settlements had fallen victim to Oric’s troops.”
In other words, the very source for the Karremans quote explicitly says that Karremans is wrong almost immediately following the quote. By even the most expansive and generous definition of “village” (i.e. any collection of houses which was designated by a separate name, no matter how small in number) the number destroyed come to around 100, barely half of the number claimed by Karremans. It would be almost impossible for Herman to have actually checked the relevant page of the report and not noticed this subsequent passage. Either he knew about it and deliberately ignored it, or else he never actually read the report itself and plagiarised another author’s citation rather than crediting them.
3. American claims of the number of Kosovo Albanians killed by Serbian forces
Changing gears for a moment here, let’s head over to the War in Kosovo in 1999. Here Herman’s goal is rather different. Rather than playing up Serb deaths, he is now seeking to play down Albanian deaths – or, failing that, exaggerate how many American and other Western sources claimed were dead.
In accordance with this goal, Herman claims on p.29 of The Srebrenica Massacre that “State Department claims of Serb killings ran up to 500,000.” A similar claim appears on p.50 of his 2010 book The Politics of Genocide: “Hysterical NATO and KLA estimates of the missing and presumably slaughtered Kosovo Albanians at times ran upwards of one hundred thousand, reaching 500,000 in one State Department press release.”
When he actually does give something of an actual quote for this claim, however, the dishonesty behind the above claims starts to become apparent:
“Actually, the death toll in Kosovo turned out to be low; perhaps when Power was writing her book, she still took as truth the hugely inflated estimates of her government (i.e., “from a low of 100,000…up to nearly 500,000”).”
Ah, so it turns out this is actually a range of numbers, which Herman has elsewhere chosen to cite only the upper bound of, not a figure given as a likely estimate. But there’s more. When checking the reference, someone reading between the lines may think there’s something odd about the figures being attributed to a subsection of a documented titled “Detentions” – an odd place to mention death tolls.
When the original document itself is checked, however, the context of the quoted section says something rather different:
“Refugees have claimed that Serb forces are systematically separating military-aged men from the groups, and the vast majority of refugees crossing international borders out of Kosovo, especially into Albania, have been women and children. We are gravely concerned about the fate of the missing men. Their number ranges from a low of 100,000, looking only at the men missing from among refugee families in Albania, up to nearly 500,000, if reports of widespread separation of men among the IDPs within Kosovo are true.
The following locations within Kosovo have been reported as the sites of mass detention facilities:”
Thereafter follows a bullet point list of suspected detention facilities, and short description including estimated numbers of detainees. The 100,000-500,000 figure refers to missing men, not dead. Nowhere are they referred to as dead, nor does the source imply they are – quite the contrary it clearly indicates that they are alive and being held in detention camps. Once again, this is something that someone who reads the source cannot fail to notice.
4. The number of Croatian Serbs missing in Operation Storm
A minor and particularly odd one. Regarding the retaking (“Operation Storm”) by the Croatian Army of the Krajina region controlled by Serb nationalist rebels in August 1995, Herman claims the following:
“The Srebrenica massacre took place in the month before Operation Storm, Croatia’s devastating attack and ethnic cleansing of some 250,000 Serbs from the Krajina, with over 1,000 civilians killed, including over 500 women and children—no women and children were bussed to safety by the perpetrators, as they were at Srebrenica—and more than 2,000 missing.”
I will ignore for the moment the 1,000 civilians killed claim, as it is sourced to a website that seems to be no longer active, and I’m unable to find the source anywhere else, but I will note that it refers to the Croatian attack as an “aggression” when Krajina was in fact internationally recognised as legally part of Croatia. However, the source given for the 2,000 missing is particularly odd:
“At tenth anniversary ceremonies in Belgrade, various survivor groups that represent former Krajina Serbs estimated as many as 2,627 Serbs had gone missing in the Krajina between 1991 and 1995. See “Patriarch Pavle holds memorial service for Serb victims of operation Oluja,” August 4, 2005”
Here, there isn’t even a need to check the source – Herman has, remarkably, given the evidence of the dishonesty of his claim in his own footnote. He claims the over 2,000 missing figure refers specifically to Operation Storm in August 1995. The source, however, is clearly referring to the entire four-year war from 1991 to 1995!
5. The composition of the ICMP Staff
Several years ago, Herman engaged in a rather public exchange with George Monbiot over the former’s genocide denialism. Monbiot’s writings on this are well worth reading, and I would encourage you to do so, as he’s covered a lot of the same themes as this article. But in particular, I want to pick up on a claim made by Herman in reply to a piece by Monbiot:
“Monbiot also takes the International Commission on Missing Persons (ICMP) as an unchallengeable authority on the body count at Srebrenica, even though its staff is 90 percent Bosnian Muslim and operates under U.S. sponsorship.”
The ICMP, founded to investigate and determine the identities and causes of death of missing persons and discovered remains, is, despite having been founded in 1996 in the direct aftermath of the Bosnian War, is not a solely Bosnian body, but also has offices in Albania, Iraq, Mexico, Lebanon and Colombia, and has engaged in work in plenty of other countries. But assuming Herman is referring solely to their two Bosnian offices in Sarajevo and Tuzla, what’s the evidence that their staff are 90% Muslim? Well, Herman’s footnote reads as follows:
“In late 2007, the Financial Times reported that the ICMP’s “staff . . . are 93 per cent Bosnian [Muslim]. . .” (Christian Jennings, “Forensics: DNA Fills Gaps of History,” December 11, 2007).”
The use of [Muslim] in square brackets should be a dead giveaway. The full sentence from the cited original article reads as following:
“The staff, who are 93 per cent Bosnian, demonstrate the former war-ravaged country’s mostly untapped capabilities, the two directors say.”
Not 93% Bosnian Muslim. Just 93% Bosnian – no ethnic breakdown mentioned. The article mentions the word “Muslim” only once, in a completely different place and context (to refer to the victims of the Srebrenica massacre).
In some of the other examples, I’ve tried to allow for a more generous interpretation of Herman’s motives – purely for the sake of argument, not because I believe it to be the case – such as maybe he was just lazy, maybe he lied about reading the source at all rather than omitting what it contains, etc. There is no more generous interpretation here. Herman straightforwardly lied about what the source said. He inserted a detail that is not in the article he cites.
Conclusion
I don’t want to bore you with endless examples. I think the ones I’ve listed so far speak for themselves. They are merely a taste of Herman’s denial, an indeed the wider bizarre phenomenon of Western denial of genocide and other atrocities (and apologetics for the Serb nationalist project) in the former Yugoslavia. These few are the most egregious and blatant cases of dishonesty I could find in Herman’s writings, but the same misrepresentation of sources and facts, often blatantly and deliberately, remains throughout his writings. I cannot help but be reminded of some Richard Evans’ concluding comments in his report on the pseudohistory of David Irving for the Irving-Lipstadt trial on Irving’s Holocaust denial:
“Not one of his books, speeches or articles, not one paragraph, not one sentence in any of them, can be taken on trust as an accurate representation of its historical subject. All of them are completely worthless as history, because Irving cannot be trusted anywhere, in any of them, to give a reliable account of what he is talking or writing about.”
Replace “Irving” with “Herman” or indeed “Handke”, and the sentence remains just as true.
