C raig Etcheson damns me with faint praise (Nightmare of a History, Post, Dec 17-30).

Our debate would have more zest if he had had the courage of his convictions and given me an honest roasting for taking a position opposite to his own.

The nub of our discord concerns the term "genocide". This is such a fundamental issue, and so intimately connected with any forthcoming trial of the surviving Khmer Rouge leaders, that it is surely worth looking at more closely. Mr. Etcheson writes that the case of the 500,000 strong Vietnamese minority in Cambodia - which was subjected to "extermination and deportation" - is "a textbook example of genocide".

Extermination there certainly was - under Lon Nol in 1970. The French government at that time sought legal advice as to whether the Genocide Convention could be invoked, and was advised that it could. But by then Lon Nol was trying to distance himself from his pogroms, so the matter was quietly dropped.

Under Lon Nol, half the Vietnamese community left Cambodia. Of the remaining 250,000, an unknown but significant number left before 1975, while the remainder returned to Vietnam in the year after the Khmer Rouge came to power. Were they deported, or did they simply depart, thanking their lucky stars that they'd finally found a way out of Pol Pot's dystopia?

The attempts of numerous Khmers and Sino-Khmers to pass themselves off as Vietnamese in order to be permitted to leave suggests the latter. As Mr. Etcheson knows, not all of them succeeded. The Vietnamese authorities of the day, to show solidarity with their Khmer communist brothers, repatriated any Khmers they detected despite knowing that on their return they would be killed by what they would later call "the genocidal regime".

Should the surviving members of Le Duan's administration also therefore be brought to trial for complicity in genocide?

This is all a monstrous red herring. The plight of the Vietnamese minority in Cambodia is not a textbook example of anything, least of all genocide. And even if it were, it would not and could not be the main issue with which a tribunal should concern itself. Do we not all agree that the core of any trial is the responsibility of Pol Pot's regime in enslaving - literally, not metaphorically - the entire Cambodian nation, and causing the deaths of between a sixth and a quarter of them? It should not be too much to ask that at least on that point there be common ground.

With what, then, should the surviving Khmer Rouge leaders be charged in regard to the chief accusation against them: that concerning the Cambodian people?

Mr. Etcheson allows that a charge of genocide against Khmers would be a "difficult proposition in terms of the Genocide Convention". But contradictorily, he insists that "most legal scholars and jurists who have examined the Khmer Rouge crimes" hold the view that genocide was committed, and he supposes that "some readers" (Mr. Etcheson himself, perhaps?) will be dismayed by my rejection of the term.

Do I detect muddled thinking? More to the point, what is it that makes Mr. Etcheson - and, with honorable exceptions, many other scholars - cling to the term "genocide" with such limpet-like devotion, even as, in unguarded moments, they admit that it may not in fact be appropriate?

Some years ago, at the time when the current vogue for genocide studies was being born, Serge Thion wrote that genocide was becoming "a political commodity". Today it is an academic commodity - I am tempted to say, part of an academic food-chain - in which the word "genocide" has become so elastic that the unique horror it is intended to convey has been attenuated and debased. The study of genocide - as perpetrated by the Nazis, and in Armenia and Rwanda - is necessary and important. But when every atrocious regime is described as genocidal, the word loses not only its meaning but, worse, its force.

Genocide studies in their contemporary form - especially as applied to Cambodia - are an American creation, fostered by Congress for political reasons at the instigation of the US Administration at a time when the US was seeking closure for a particularly painful chapter of its policy in Asia and, at the same time, looking for ways to control an emerging system of supranational justice.

That the media should have swallowed the Vietnamese line about genocide in Cambodia, which the US later made its own (politics makes strange bedfellows), is regrettable, but understandable: The word makes good headlines. That other western governments should have gone along with the notion of a Cambodian genocide is also understandable. As one western ambassador in Phnom Penh put it to me a few years ago: "We have enough problems with Washington. For us, this is a small matter. Why make an issue of it?" But that international scholars should unthinkingly lend their authority to such propaganda is unworthy.

A spade is a spade is a spade.

The Khmer Rouge period was an abomination. Its leaders committed crimes against humanity. If that was considered the appropriate charge for the former Nazi leaders, why is it, in some mysterious way, not good enough for the Khmers Rouges? This is not some minor issue that will just go away in the wash, as Mr. Etcheson appears to believe. The prosecutors must frame proper charges and the tribunal must render judgement on those charges. In a fair trial of the Khmer Rouge leadership on charges of genocide, the court would have no choice but to acquit. If we follow Mr. Etcheson's logic to its conclusion, we shall end up being grateful that the chances of a fair trial in Cambodia - and hence of such a verdict - are nil.

What a topsy-turvy world we live in!

On a very different topic - culture vs. communism as the root of Khmer Rouge behavior - Craig Etcheson makes more sense. Cultural generalizations were fashionable in the early part of the last century. Then economic determinism took over and culturalism became anathema. Both approaches are way too extreme. We need what Steve Heder has called a "judicious synthesis". I have attempted - inadequately in Mr. Etcheson's view - to offer a balanced picture of the extremely complex mix of factors that underlay the murderous specificity of Khmer Rouge rule. Cambodian readers will judge for themselves whether or not, or to what extent, I have succeeded. But taking parts of the picture out of context and pretending that they constitute the whole is poor criticism. I wrote early on in the book, in a passage which Mr. Etcheson did not quote: "Cambodians are not biologically more prone to cruelty than Americans or Europeans. The causes are rooted in history. ... in culture ... and in the political and social system."

Cambodia's political and social system remained in practice an absolute monarchy, which engendered the same types of behavior, and the same types of horrific peasant revolt, as did similar regimes in medieval Europe. The French colonial archives contain descriptions of Cambodian peasant atrocities in the 19th century identical to those carried out by the Khmer Rouge. Bunchhan Mol describes the self-same atrocities carried out by the non-communist Issarak in the 1940s. Yet Khmer behavior was not replicated in Vietnam - either in the Khmer Rouge period or before - in the same way that Khmer art has no parallel in Vietnamese or Chinese art.

Culture is important, in life no less than in art. It is not the only element in the mix. But to deny its role - simply because our forebearers exaggerated it - is to throw the baby out with the bath water.

Craig Etcheson's interpretation of what I wrote about reform proves him right on one point. I should have been more careful in my phrasing. The issue is not whether I think that Khmers are reformable; it is that Pol Pot and his colleagues (in contrast to the Chinese and Vietnamese communists) became convinced that they were not. The failure of Son Sen's attempts to re-educate erring soldiers, as related by Deuch to Nate Thayer, confirmed them in that view. I should also have made doubly clear that the views I quoted on Khmers' "innate individualism" are not mine, but those of King Sihanouk and other thoughtful Cambodians. Surely Khmers have the right to discuss their own self-image. For them, it is a matter not of paternalistic generalization but of self-perception.

Mr. Etcheson raises other red herrings. He writes that "Cambodia-watchers have puzzled for years over the circumstances of Pol Pot's death." Maybe Mr. Etcheson has. I have not - for the good and simple reason that Thiounn Thioeunn, who treated him, has described his heart condition in detail and has made clear that a fatal heart attack was only a matter of time. And all those in Pol Pot's immediate circle, from his wife, Meas, to his guards, insist the rumors of suicide are untrue. If there were credible evidence to the contrary, it would be a different matter. But there is none. So why launch into baseless speculation and conspiracy theories? Not to mince words, this is bunkum.

The same applies to his quibbles about the identities of the members of the Communist Party of Kampuchea Security Committee. Is Mr. Etcheson seriously suggesting that Pol Pot was not a member when Pol himself admitted to Nate Thayer that he personally decided the fates of "the important people" who were purged? The mind boggles. Yun Yat took care of the dossiers - which is presumably why Ieng Sary named her - but she was not a committee member.

Then there is the matter of the mysterious old lady to whose car, flying a DK flag, Khieu Samphan and Prince Sihanouk had to yield right of way. Sihanouk himself states that it was not Khieu Ponnary. The only other possibilities are the Khieu sisters' mother or, conceivably, Nuon Chea's mother. Sihanouk's account accords perfectly with Ong Thong Hoeung's description of the former - but not with Nuon Chea's mother, who in any case spent the whole of the DK period near Battambang. If it were not old Madame Khieu, would Mr. Etcheson kindly tell us who he thinks it was? The job of the historian is to make plausible deductions - in other words, informed guesses - not to throw up his hands and say, oh, there's no documentary evidence, therefore we can never know.

Mr. Etcheson's insistence that nothing the former Khmer Rouge leaders say can be believed unless confirmed by "multiple, independent, non-Khmer Rouge sources" falls into the same category. It makes writing the history of the Communist Party of Kampuchea impossible - because, by definition, there are no non-Khmer Rouge sources for decisions and discussions to which only the leaders themselves were privy. To accept such a proposition would leave us with a history written exclusively by the victims - "something clean and perfect in accordance with our stance," as Pol Pot wrote in another context - with no inconvenient discussion of circumstances or context or motives to distract attention from the monstrosity of what the Khmer Rouge did. This would be good propaganda but bad history. I do not believe that is what Mr. Etcheson wants any more than I do.

As regards my "cavalier" and "sloppy" research, mea culpa. I plead guilty to a howler about royal pardons for Khieu Samphan and Nuon Chea. David Chandler pointed it out when the book was still in proof. It was corrected in time for the US edition, but in the UK version was left in, not due to "hurried editing" but to an error in the computerized typesetting, which in this day and age seems beyond human control. Certainly it was beyond mine!

Regarding Hun Sen's links to the grenade attack on the Sam Rainsy Party rally, I spoke three years ago to two sources, one of them directly implicated in the planning of the attack, who both independently confirmed the Prime Minister's role. Craig Etcheson should be sufficiently familiar with the realities of Cambodian politics to understand why they are not named in the notes.

To Mr. Etcheson my conclusions are necessarily unsatisfactory - and are intended to be - for we differ fundamentally in our approach. That is only right and proper. Without academic debate, our understanding of history would be poorer. Cambodian readers will make up their own minds. It is, after all, their country. Their judgement should matter most.