Marching arm in arm with the credulity of US journalists still swallowing nonsense from the State Department and Pentagon about "exposing the central front" to Soviet blitzkrieg, is their willingness to take at face value almost anything they are told by intellectuals in Moscow, so long as it is somehow associated with "reform" (which to American journalists means, in the last analysis, restoration of capitalist relations).
And in these heady days in Moscow it seems that some of these Soviet intellectuals will say anything they think Americans want to hear. A friend of mine recently returned from a conference in Hawaii, where American historians associated with "revisionist" readings of the origins of the Cold War were increasingly irked to hear Soviet participants piously echoing all the most hawkish American constructions of Soviet behaviour in the postwar period, to the tremendous pleasure of the American right-wingers in attendance. At this stage in the process of _glasnost_, these Soviets plainly felt that any defence of Soviet deeds in the pre-Gorbachev years amounted to a betrayal of the new thinking.
At the start of February, the tabloid _Argumenti i Fakti_ reported that the Soviet historian Roy Medvedev had proposed that Stalin's victims amounted to some 20 million. From Moscow, the _New York Times'_ correspondent Bill Keller relayed this to his newspaper which on 4 February ran a front-page headline announcing "Major Soviet Paper Says 20 Million Died as Victims of Stalin", with the lead paragraph reiterating Medvedev's claim that "about 20 million died in labour camps, forced collectivisation, famine and executions".
To me, the total figure seemed to have an insouciant roundness and also a suspect symmetry with the total - also 20 million - normally reckoned for Soviet losses in the war against Hitler. Looking through Medvedev's breakdown one could perceive that the word "million" really meant "a lot", with no substantive precision beyond the vague imputation of multitude. As relayed by Keller these volumes were expressed as "one million imprisoned or exiled from 1927 to 1929", or "nine or 10 million of the more prosperous peasants driven from their lands", and so on. In the end we were left with an overall figure of 40 million who, on Medvedev's account, had an awful or terminal time of it between 1927 and 1953, with 20 million actually killed.
All US reports of Medvedev's estimates told their readers that his was the most "precise" accounting thus far. No newspaper or TV programme that I saw, rang up any of the relevant scholars to get their reaction. When I started to do so myself, I was interested to find well-qualified historians and demographers in the US who regard Medvedev's claims as absurd. Sheila Fitzpatrick, professor of history at the University of Texas at Austin, told me there was "no serious basis for his calculations" and that privately some Soviet demographers and historians find Medvedev's work in this area embarrassingly bad. She gave me a couple of examples to explain why she thought Medvedev's numbers ridiculous.
Medvedev concludes that nine to 11 million prosperous peasants were driven from their lands with another two to three million arrested or exiled in the forced collectivisation of the early 1930s. But, Fitzpatrick says, Medvedev makes no distinction between those who left their villages voluntarily and those who left by force. This was the era of industrialisation and many of Medvedev's millions were moving to the town. Medvedev also bases his figures on the assumption that the average peasant family in the late 1920s had eight members, whereas in fact five was the normal size. Fitzpatrick also cited the famous conversation between Churchill and Stalin as another flimsy source, often used by some to claim that Stalin told Churchill that ten million peasants died in collectivisation. The actual passage in _The Hinge of Fate_ makes it clear Stalin was talking about the total number of peasants he was dealing with, not those who died. According to Fitzpatrick, a respected estimate, concurred with by several historians in the West, is that of the historian Victor P Danilov, who recently wrote in _Pravda_ that approximately three to four million died in the famine of the early 1930s. But where does that leave us on the matter of the purges?
In his 1946 survey, _The Population of the Soviet Union_, the demographer Frank Lorimer studied data from the Soviet census of 1925 and 1939 and all available information on fertility and mortality between these two dates. He calculated that what demographers call "excess deaths", that is, in Lorimer's method, a comparison of the reported total population in 1939 with the expected population at that date - given the count in 1925 and everything known about fertility, mortality and emigration between those years - amounted to somewhere between 4.5 million and 5 million, though this total included perhaps several hundred thousand emigrants, such as those Central Asian nomads moving into Sinkiang to avoid collectivisation.
In their 1979 volume _How the Soviet Union is Governed_, Professors Jerry Hough and Merle Fainsod generally supported Lorimer's calculation and concluded that the more extreme western estimates "cannot be sustained". Rather, "a smaller - but still horrifying - number" of "maybe some 3.5 million" emerges as the direct or indirect result of collectivisation in the early 1930s. With respect to the purges of 1937 and 1938, Hough and Fainsod again criticise excessive Western estimates, and report that on the evidence of extant demographic data, "the number of deaths in the purge would certainly be placed in the hundreds of thousands rather than in excess of a million". Indeed, "a figure in the low hundreds of thousands seems much more probable than one in the high hundreds of thousands, and even George Kennan's estimate of 'tens of thousands' is quite conceivable, maybe even probable."
At the far end of the spectrum from Hough and Fainsod, is Robert Conquest who had reckoned in excess of 20 million deaths under Stalin before 1939. In his essay _The Stalin Question Since Stalin_, Bukhtain's [sic -Bukharin?] biographer, Stephen Cohen cites Conquest's figure as "conservative", without mentioning lower numbers by other scholars and concludes by saying: "Judged only by the number of victims, and leaving aside important differences between the two regimes, Stalinism created a holocaust greater than Hitler's".
In this decade the most significant scholarly battle on the subject has been waged in the pages of the _Slavic Review_ between Stephen Wheatcroft and Steven Rosefielde, with Wheatcroft writing in 1985 that "these wildly unscholarly estimates [such as Cohen's] serve neither science nor morality", and "It is no betrayal of thm [the victims] nor an apologia for Stalin to state that there is no demographic evidence to indicate a population loss of more than six million between 1926 and 1939, or more than three to four million in the famine. Scholarship must be guided by reason and not emotion."
In a widely noted essay, also in _Slavic Review_ for 1985, the demographers Barbara Anderson and Brian Silver supported Wheatcroft, reckoning "excess deaths" between 1926 and 1939, to those alive in 1926, at a median figure of 3.5 million. Conquest, now at the Hoover Institution in Stanford, says that Medvedev's numbers are "obviously in the right range" though "perhaps he spread
them wrong", and "I'm not sure where he gets them from". He slighted Anderson and Silver's work as the product of demography rather than sovietology, and derided Hough and Fainsod's figures as "improbable". From the University of Michigan, Anderson responds that: "Conquest wouldn't know a number of it bit him." She thinks Medvedev's computations "ludicrous".
No doubt some will be eager to conclude that the foregoing is somehow an attempt to exonerate Stalin, dismiss the purges as got up by western propaganda. The following observation by Hough and Fainsod is salutary: "Some persons seem instinctively to object to [our] figures on the grounds that the Great Purge was so horrible that the number of deaths cannot have been so 'low'. We must become so insensitive to the value of human life, however, that we dismiss tens of thousands of deaths as insignificant and need to exaggerate the number by ten, 20, 30, 40 times to touch our feelings of horror."
The task is obviously to arrive at truth, but many such estimates evidently have a regulatory ideological function with an exponential momentum so great that now any computation that does not soar past ten million is somehow taken as evidence of being soft on Stalin. One can find an analogy in current writing on the French Revolution, where the passionately anti-Jacobin Rene Sedillot has produced a book addressing the matter of the Revolution's (and Counter-Revolution's, though it's never quite put like that) human cost where he boils up, by very questionable means, a casualty figure far in excess of all previous estimates.
The symmetry that calculations, such as Medvedev's, seek to establish between Stalin and Hitler performs similar injury to history. Hitler wanted to exterminate the Jews and the gypsies, and though accuracy is important it does not alter the moral scale of this horror one iota to propose that, in pursuit of this design, Hitler may have, in reality, killed a million less or a million more than the conventional estimate.
Evil though he was, Stalin did not plan or seek to accomplish genocide, and to say that he and Hitler had the same project in mind (or as right-wing German historians now argue, that somehow Lenin and Stalin put Hitler up to it) is to do disservice to history and to truth.