It is a pity that he takes this valid and important observation too far, thereby falling prey to vagueness, inconsistency and error.

Vagueness results from the different meanings he gives to the concept of “stateless zone”. The core meaning of such a zone is a territory in which sovereignty has been lost and in addition where the apparatus of government and the rule of law have disintegrated. A territory may or may not become a “stateless zone” following occupation by a foreign power. It depends on what happens to its machinery of government and law. Following its defeat in 1940, France suffered occupation but did not become a zone of lawlessness. The analytical problem is that the criteria of statehood and statelessness become unclear. States may have varying degrees of statehood. Thus, in explaining why a higher percentage of Jews were murdered in occupied Holland than in occupied France, Snyder is forced to explain that the Netherlands “were, for several reasons, the closest approximation to statelessness in Western Europe”. The SS, he explains, had greater dominance in the Netherlands. The problem is that Snyder’s reasoning risks becoming circular and thus loses much of its explanatory force. If the degree of statelessness depends on that of Nazi dominance, why not merely conclude that the Nazis killed Jews most easily where they enjoyed the greatest dominance — true but banal?

There are further difficulties. It is not accurate to assert that German murders of Jews all took place in stateless zones. German Jews who failed to emigrate or to hide were liquidated; Germany certainly was not a stateless zone as defined by Snyder at the time it murdered its remaining Jews. That most of them were deported to the East before being killed there hardly affects the issue. In Hungary, the government of Admiral Horthy agreed in March to June 1944 to deport nearly a half million of its Jewish citizens, most of whom were then gassed at Auschwitz. Though under pressure from its senior ally in Berlin, Hungary was not a stateless zone at that time.

Snyder’s analysis also fails to accommodate mass murders by governments of their own citizens, which do not occur in “stateless zones”. Stalin’s starvation of kulaks in the 1930s, the deaths by starvation and violence during Mao Zedong’s “Great Leap Forward” in China from the late 1950s, the Indonesian regime’s murders of Communists in the 1960s, Pol Pot’s killings in Cambodia as well as all too recent numerous events in Africa are examples.

His finding that the breakdown of nation states was a prime causal factor of the Holocaust leads him to conclude that nation states are indispensable. This, of course, is inconsistent with his avid support for the European Union. The essence of its “European Project” is to weaken the role of nation states. Indeed, when interviewed in Bratislava in June 2015, he accused Russia under the “tyrannical” Putin of wishing to destroy the EU by turning it “into a big mass of nation states”.

Seventh and finally, Snyder’s concept of the character and role of historical explanation is questionable. This is especially the case with application to the Holocaust. He argues (again in his Stanford lecture) that historical explanation “can only work if its arguments apply both to the past and to the present . . . if it brings an event of the past into our own present understanding . . . in the sense that it enlightens our present and may give us a hint about what’s going to happen in the future”.