Anatol Lieven's article "Qu'est-ce qu'une nation?" (Fall 1997) is, like the curate's boiled egg in the old Punch cartoon, good in parts. The good parts are the generalities: his account of the difference between "primordialist" and "constructionist" theories of national identity is admirably lucid, and so too is his general argument that the "construction" of such identities is far from being the mechanical or arbitrary process that the term is sometimes taken to imply. No doubt the full explanation of the development of national identity must involve, where most nations or peoples are concerned, both the material of pre-existing realities (of language, culture, religion, political history, et cetera) and the more active workings of imagination and ideology. Doubtless, too, the mix will vary from case to case; local factors are never quite the same, and each national identity will be, like Tolstoy's unhappy families, national in its own different way. These are general truths, and Lieven's exposition of them is, surely, generally correct.

It is in the detail of his argument, however, and in its application to particular cases, that Lieven goes wrong. Much of the problem lies in his method of refutation. First he sets up a straw man, a dogmatic ultra-constructionist; then he identifies with that straw man the authors with whom he disagrees on particular cases (the commentators he singles out are Johnathan Sunley, Norman Stone, and myself); and then, instead of presenting evidence that would disprove what we have claimed about those particular cases (claims about Moldova, for example, in Sunley's writings, or about Bosnia in mine), he relies on a priori arguments to do his work for him. But those arguments simply fail. Neither Sunley, nor Stone, nor I, nor anyone else I can think of, has claimed on a priori grounds that all national conflicts must necessarily be caused by artificial manipulation from above. We have merely observed, from empirical evidence, that this is what has in fact happened in many cases, and that, in the particular political circumstances of post-communist Eastern Europe, such instances amount to a general pattern.

These observations have nothing to do with any theoretical commitment to ultra-constructionism. A rough analogy might clarify this point. Sexual identity and sexual desire are usually based, like national identity and national feeling, on some pre-existent realities--in this case, biological ones. Sunley and I, observing several episodes of mass rape, find reasons to believe that these incidents were organized from above. Lieven, instead of providing evidence to the contrary, merely identifies us with some Foucault-style straw man who says that sexual identity is nothing more than an arbitrary social construction.

What is really needed to refute our claims about Eastern Europe is empirical evidence to overturn our interpretations of the causes of particular conflicts. This Lieven fails to supply. He seems to agree that manipulation from above has been a key factor in some conflicts; his disagreement with Sunley is merely about how long the list of such conflicts is. The best way to prove Sunley wrong would be to give some detailed case histories of conflicts that Sunley believes were manufactured from above, but which Lieven can demonstrate to have been caused from below. But apart from a few assertive remarks made in passing about the Nagorno-Karabakh war, Lieven does not even attempt this.

Instead, paradoxically, his most detailed examples are of cases in the Baltic states where he agrees that there were definite attempts to manipulate and manufacture conflicts from above, but where, he notes, the attempts failed. If Lieven thinks that these examples overturn Sunley's argument, then he is making the elementary logician's confusion between a necessary cause (necessary, that is, in these particular cases) and a sufficient one.

Where the particularities of the Yugoslav or Bosnian case are concerned, Lieven is doubly wrong. He is mistaken in what he says, or at least implies in a negative sort of way, about the origins of the war there, and he is also quite mistaken about the position he thinks I have taken on this issue. While I realize that the second of these mistakes is, sub specie aeternitatis, of minuscule importance, let me deal with it first.

Lieven begins his article by suggesting that the only alternative to the policy actually pursued by Western governments over the wars in Croatia and Bosnia was to intervene militarily by sending in large numbers of ground troops. He then puts my name forward as, apparently, a prime advocate of such all-out intervention. At the end of his article he appears to return to this theme, implying that NATO soldiers will be endangered if their masters are "foolish" enough to listen to the advice of people such as me.

Lieven is badly misinformed, both about Yugoslav affairs and about my views. I never advocated the sending of Western ground troops to intervene in the war in Bosnia; on the contrary, I argued consistently against it, and pointed out repeatedly that the Bosnian government itself was not asking for such an intervention. I also pointed out, again and again, that to present this kind of intervention as the only alternative to the existing policy was to misrepresent the situation. In Lieven's article that misrepresentation is perpetuated.

Broadly speaking, there were not two but three policy options: anti-Serb, neutral, or pro-Serb. (When I say "Serb" here I refer, of course, not to Serbs in general but to their political leaders and those leaders' military forces.) Western governments claimed to be following the second of these policies, but in crucial ways they were in fact pursuing the third. By actively enforcing an arms embargo, they locked in a massive strategic advantage on the Serb side, which, from the start of the war, already possessed huge stockpiles of arms. By sending in token forces, the Western powers created a potential or actual hostage situation that gave the Serbs extra leverage over their policymaking. I could extend this list of implicit, and in some ways perhaps unintended, pro-Serb aspects of Western policy. (I did in fact develop this argument at greater length in the Spring 1995 issue of The National Interest.)

As a British citizen, I do not want British soldiers to be committed in a full-scale ground war for any purpose except the defense of vital, immediate British interests--which were not at stake in Bosnia in 1992. I would expect American citizens to have similar views about their own troops. This does not mean, on the other hand, that European NATO powers had no long-term strategic interest in preventing the outbreak of such a massively disruptive war in their own backyard; I believe that they could have done that through the use of airpower during the first days of Milosevic's military operation. Given their complete reluctance to act in such a way, the best policy would have been a genuinely non-interventionist one, allowing the Bosnian government to buy arms on the world market and defend its state against unprovoked attack. As the war developed, however, Western involvement there--of the sort I had constantly argued against--created an ever-rising level of Western responsibility for the outcome of events. To correct this, I advocated a more active policy that would help the Bosnian government in more positive but subsidiary ways. These included some use of airpower, which, I believed, could be deployed at low levels of risk to NATO pilots. In August and September 1995 this belief was proved, belatedly, to be correct.

As for the origins and causes of the war in Bosnia, Lieven does not even attempt an explanation of his own; so to controvert him is as difficult as grappling with thin air. His only method of argument on this point is to set up a parody of what he disagrees with: his "Iago" image of people being persuaded by the malign whispers of propaganda to "rush out madly and start strangling their neighbors." I do believe that propaganda and psychological manipulation played an important role, but I do not think, and have never said, that the war was created merely by persuading ordinary Bosnians to go out and kill their neighbors. The war was created in the way that wars are usually made: by a political commander-in-chief (Miloÿsevi«c) using his military forces to achieve political purposes. As I said in my Bosnia: A Short History (p. 238): "It is quite clear that the conquest was mainly achieved by federal army forces directed from Belgrade, and paramilitary groups from Serbia. In other words . . . this was predominantly an invasion of Bosnia planned and directed from Serbian soil."

Propaganda was churned out relentlessly by the state-controlled Serbian media before and during the war, and certainly had an effect. But the most crucial propaganda was the propaganda of the deed, not the word. For example: When, in April 1992, ordinary Serbs from small towns in eastern Bosnia awoke to find the streets littered with the bodies of their Muslim neighbors, murdered by newly arrived Serb paramilitary units, this was a very powerful way of reinforcing the message (hitherto only a verbal one) that the apparent normality of their town life up until then had concealed dreadful hidden threats and dangers--typically, the danger of some secret Muslim conspiracy to murder local Serbs in their beds. How could such a terrible sight--a bloody, sticky reality in your own street--not have somehow an equally terrible reality as its justification?

The deed thus completed the job that the word had merely begun. Paradoxically, the fact that it was the Muslims who were murdered did not cancel out the claim that they were the murderous ones: it gave it a much-needed psychological reinforcement. And for those who then went along with the murder or expulsion of their Muslim neighbors (whether because they accepted those claims, or because they just wanted to keep their heads down), a further psychological mechanism came, as time passed, into play: the mechanism of acquired complicity. Many Bosnian Serbs today do not want to believe that there was no Muslim conspiracy to kill them, not merely because they have heard the reality of this conspiracy affirmed so often that it no longer sounds strange to them, but also because to admit that it was unreal would make their own actions hard to live with in retrospect.

I have frequently heard it said that propaganda, on principle, cannot be held responsible for anything, because if it works it does so only by locating pre-existing prejudices in its target audience. Lieven goes along with this, to some extent, in his article: he writes that Milosevic and his counterparts "would not have been so successful in their manipulations if they had not had a mass of suitably conditioned human material to work on." But how successful is "so successful"? There are differences in degree here that amount to differences in kind: day-to-day prejudices are one thing, but an acceptance of, or participation in, mass-murder is very much another. All through his article, Lieven's terminology glides effortlessly up and down a slippery sliding scale, from "national identity" to "national feeling", to "nationalism", to "national hatreds", to "national hostilities", and finally to "national conflicts." At times he seems to take it for granted that if you have a national (or ethnic) identity, it will sooner or later lead you to kill people of a different identity. But it is precisely the transition from prejudice to war that needs to be explained, and, in the Yugoslav case, receives no proper explanation in his account.

He criticizes me for trying to deny "the truth, depth, and antiquity of national feelings and national hostilities." But I have never denied that the Bosnian Serbs, for instance, have a sense of national identity resting on pre-existent cultural realities. (I do deny the "antiquity" of this identity; elsewhere I have given reasons for thinking that the Orthodox Bosnians only started to gain a sense of Serb national identity in the nineteenth century. If Lieven has evidence to the contrary, he should present it.) But, most importantly, what I deny is that a sense of national identity led people naturally or inevitably to take part in the mass murder of Bosnian Muslims in 1992-95.

Again, let me propose a rough analogy. The last time I said in a public lecture in America that the war in Bosnia was not caused by "ancient ethnic hatreds", someone came up to me afterwards and said that he did not believe me, because twenty years ago his family once had a Croatian au pair who confided in them that she didn't like Serbs. From this not untypical example of day-to-day national prejudice, it seemed to him entirely natural that the people of the former Yugoslavia should have slid, of their own accord, down that slippery scale to national conflict and mass murder. My analogy would be as follows: If I traveled around America today, I could probably collect many examples of whites confiding in me that they did not really like or trust blacks, would not want their daughter to marry one, and so on. But if I returned to America five years later and found that black-inhabited districts had been bombarded with heavy artillery, and that many tens of thousands of blacks had been murdered or put in concentration camps, would I then be justified in saying, "Oh well, it doesn't surprise me; I always knew that the whites didn't really like the blacks"?

Would it not make more sense to start asking what political measures had brought about this extraordinarily rapid transition from one end of the sliding scale to the other? And if I found evidence of such political manipulation at work, and presented it, would it be an adequate refutation to announce that I was just a dogmatic constructionist who denied that blackness or whiteness was based on anything real?

It was kind of Noel Malcolm to approve the general principle set out in my essay--but his own position, or pose, as an empirical historian leaves him with little choice. It would be better, however, if he were to apply this principle more rigorously in his own work.

Professor Malcolm alleges that in my essay I set up an artificial "straw man" as an easy target. It is true that part of my essay was focused on one target, but it was a real one: Johnathan Sunley, and in particular, his essay for The National Interest (Summer 1996). Furthermore, after a careful reading of some of Malcolm's own writings, I am convinced that Sunley's essay was also a fair representation of the Professor's own approach, albeit perhaps rather crudely expressed.

What is this approach? At first sight, it is almost identical to my own. Malcolm claims to be following an empirical approach, shunning rigid a priori principles. He admits that every national experience involves a different mixture of factors, and therefore implies his acceptance of the principle that every historical event must be examined in its specificity, and in detail. He deliberately ignores my statement that manipulation of national sentiments from above has often played an important role. In my essay, I mentioned attempts at such manipulation by Milosevic in Yugoslavia, Iliescu in Romania, and the Soviet authorities in the Baltic States, and gave some of the various reasons why these attempts had succeeded or failed.

It is quite true, as Malcolm writes, that Transdniestria is also a good example of such manipulation. My treatment of that dispute had to be dropped from the essay for lack of space, but it is discussed at length in the relevant chapter of my forthcoming book on contemporary Russia. In any case, Malcolm can hardly complain that in the limited space available I was not able to deal with all the post-Soviet conflicts, since he himself ignored all the examples I did give.

Much more seriously, Malcolm fails to address directly what is after all my key argument in this context: that "while both communist and nationalist manipulators have indeed been at work in many countries, the readiness of different nations to respond to their provocations has differed enormously according to local circumstance, local history, local culture, and yes, to 'national character.'" The reason that he does not try to "refute" this argument is presumably that he knows he cannot. It is so deeply rooted in both evidence and sheer common sense as to be irrefutable.

Malcolm therefore has to claim to be working on empirical principles, but his own letter reveals the falsity of this claim. It is surely by now blindingly obvious that the "post-communist Eastern Europe" of which he speaks (especially if extended to the former Soviet Republics which I discuss) covers a multitude of political cultures and tendencies; it is equally clear from the vast historiography of national conflicts that in no case, anywhere, is it possible to assign one single cause to these phenomena--which is exactly the mistake that Malcolm makes once again in his letter. It is of course legitimate to seek common patterns in post-communist behavior, but to believe that exactly the same pattern is dominant from Azerbaijan to Hungary is ipso facto evidence of a superficial and biased approach. In any event, outside the Bosnian case neither Malcolm nor Sunley provided detailed evidence in support of this argument.

A real historian, rather than an ideologue, knows that for every event, and every war, it is necessary to seek out a hierarchy of causes, with their relative importance differing from case to case. Malcolm's failure in this regard is also revealed in his remark about how "wars are usually made." A cursory acquaintance with either anthropology or history should remind him that the modern Western pattern of warfare is by no means standard for all peoples and times.

Which brings me to Bosnia. First of all, I owe Malcolm an apology on one point. He did indeed call not for ground intervention but for air attack. However, this mistake was in a sense a compliment--the compliment of confusing his position with my own.

The reasons for our difference are summed up in Malcolm's statement that in Western policy toward Bosnia there were three options: "anti-Serb, neutral, or pro-Serb." Malcolm has himself refuted this on a number of occasions, as for example when he wrote at the height of the war that, "The only way the West can save Bosnia now is by a major policy initiative to persuade the Croatian president [my italics], Franjo Tudjman, to abandon his own expansionist aims in Bosnia."

To assign chief responsibility for the partition of Bosnia to the Serbs should not be to deny the important role played by the Croats and Croat nationalism. I opposed "lift and strike" because I thought it would lead to a Croat, not a "Bosnian" victory--which is pretty much what has happened. If however there remains any chance of binding Bosnia back together into one functioning society, it lies not in blind backing of one nationality, but in a continuation and intensification of the present policy of military occupation by outside forces backed by economic sticks and carrots. This is not altogether unlike the policy of the Habsburgs or Tito's Yugoslavia--or the British in Northern Ireland, for that matter.

Finally, a word on the Professor's style of argument. His letter makes an analogy--quite inappropriate and in very poor taste--between my arguments concerning the varied origins of national conflict on the one hand and the reasons for mass rape in Bosnia on the other. He thereby suggests to the reader that I would deny that mass rape by the Serbian forces was organized from above. I mentioned this nowhere in my essay, but if I had done so, I should certainly have stated that there is indeed overwhelming evidence for the organization of mass rape by the high Serbian command as a weapon of "ethnic cleansing", and that the leaders concerned should be severely punished for this and their other innumerable atrocities.

That said, the fact that the Serbs adopted this hideous weapon so quickly and on such a widespread scale surely has much to do with the fact that it had been used frequently as a deliberate tactic both by and against the Serbs in the Balkan and Yugoslav wars of 1912-18 and 1941-44, and by Ottoman forces long before that. Stating this in no way diminishes either the evil of what was done by the Serbs in Bosnia or the responsibility of the Serbian leaders. It does however suggest once again the importance of local history. But then, for a long time now history has seemed to make Malcolm uncomfortable. Perhaps he should become a foreign correspondent instead, and learn something about collecting evidence at first hand.