Margaret Macmillan is a renowned professor of history at Oxford University, and everything she writes is worth reading. Some more than others: her Paris 1919 (2003) is a shining descriptive account of the 1919-1920 peace conferences after World War I. Someone who knows, understands, and writes about history will—perhaps predictably—give some thought to the conditions of historical knowledge and to its consequences. That is what MacMillan does in Dangerous Games.

Her first chapter, entitled “The History Craze,” is the most important one; it is all there in her first sentence: “History, and not necessarily the sort that professional historians are doing, is widely popular these days, even in North America, where we have tended to look toward the future rather than the past.” She suggests that the popular demand for history “can be partly explained by market forces.” Yes and no. The appetite for history is one thing; its nutrition another. They are connected, perhaps even inseparable, but they are not identical. The craze for history exists, and will continue I suspect, and I wish that MacMillan had extended her reflections on it beyond this nine-page, introductory chapter.

Her succeeding chapters, alas, are not so good. Most of them are directed to misuses of history—by politicians and others—in recent times. But there is nothing very new in that. I wish she had written more about how professional historians have used or abused their acquired knowledge of the past; about the dependence of “facts” on the words with which they are expressed and indeed perceived; about the ultimate inseparability—and this goes beyond historianship—of the knower and the known.

Here is an example—very, very close to me. It is about the crucial days in May 1940. Churchill, she writes “as David Reynolds has so convincingly shown” (not at all convincingly: trust me) “glossed over many awkward issues. Churchill said little, for example, of the debates within the British cabinet in those dark days of May 1940. France had fallen to the Nazis….” (France had fallen in June, not May, but that is not the point.) “In fact, as the record shows, the cabinet properly considered alternatives, most notably to see if Benito Mussolini, the Italian dictator, could broker a peace.” No: it was not the cabinet, not even the five-member war cabinet; it was Lord Halifax, the Foreign Secretary, who was not even wholly supported by his former ally Chamberlain. So her “example” about the history of those days is not “in fact,” nor “shown” thus “by the record.” I wrote an entire book about those five days: a descriptive account with which I believe that Margaret MacMillan would not disagree.

The more valuable parts of the book give telling examples of how entire peoples tend to construct legends about their past, often exaggerated and sometimes largely false, which cannot be separated from their ideas about their identity in the present. She quotes the bromide of L.P. Hartley: “The past is a foreign country; they do things differently there.” That is not even a half-truth, because the past is big and present and forms our entire mind: as Kierkegaard said, the past is the only thing we know.