Let's start with something small. Many people believe that each of the tartan (plaid) patterns worn by Scottish Highlanders corresponds to a particular clan and that kilts made of this fabric have served as the uniforms and emblems of that clan since time immemorial. But, as the historian Hugh Trevor-Roper pointed out in a famous essay titled "The Invention of History: The Highland Tradition of Scotland," that simply isn't true. "Indeed," Trevor-Roper wrote, "the whole concept of a distinct Highland culture and tradition is a retrospective invention," cooked up in the 19th century. Much the same can be said of the customs of the "traditional" wedding (the elaborate church ceremony, the white dress, etc.), which were concocted a the same time. In fact, for most of the history of Christendom, a wedding was a low-key affair conducted at home without the benefit of clergy.

Neither of these interesting tidbits appears in Margaret MacMillan's "Dangerous Games: The Uses and Abuses of History," however. MacMillan, a Canadian historian who heads St. Antony's College at Oxford, believes that historians ought to concern themselves with weightier stuff. Take, for example, the Treaty of Versailles, signed by the Allies and Germany after World War I and -- as I was led to believe in school -- significantly responsible for World War II. For decades the treaty was regarded as, in MacMillan's words "so foolish and vindictive" that it goaded Germany into a renewal of its imperial ambitions. But, MacMillan insists, the terms of Versailles were "never as severe as many Germans claimed and many British and Americans came to believe." The woes, financial and political, suffered by Germany between the wars, and the subsequent rise of Adolf Hitler on a tide of wounded nationalist pride, were really the result of the Great Depression and "a whole series of bad decisions" made by German leaders.

Yet kilts and bouquets have more common with simmering Weimar-era resentment than might initially seem. Even trivial "bad history," as MacMillan would call it, can be driven by profound desires. Trevor-Roper judged the "artificial creation of new Highland traditions, presented as ancient, original and distinctive," to be an attempt to assert a Scottish identity as a kind of protest against "Union with England." The idea of a gallant, free, Scottish tribal past appealed to the sensibility of the Victorian era as, too, did the notion of a very special white wedding dress; the first one was worn by Victoria herself when she married Prince Albert. Just as Scots thrilled to the idea of a rich native culture with deep roots, so we like to believe that the modern vision of wedlock as a union founded in true love is hallowed and eternal. Convincing ourselves that weddings have always been wrapped in sacred and sentimental rituals is like a charm against our suspicion that marriage may not be that romantic after all.

In other words, when it comes to history, we prefer to believe what suits us. "Dangerous Games," based on a series of lectures given at the University of Western Ontario, catalogs the various ways that history has been warped, manipulated and just plain fabricated in order to serve one purpose or another. The Serbs, for example, have organized a good part of their national identity around the story of the tragic Battle of Kosovo in 1389, in which the noble Prince Lazar, betrayed by a traitor, died in the sacred struggle to repel Muslim invaders. That Lazar wasn't really leader of all the Serbs, that the battle wasn't a total rout resulting in the collapse of an independent Serbian state -- these facts sap much of the drama out of his alleged martyrdom, making him a less effective symbol for current Serbian claims in the Balkans. Furthermore, as MacMillan points out, it was the recent revival of those claims that polished up the nationalistic gloss on Lazar story, which had largely been forgotten before the breakup of Yugoslavia.

Some people embrace "bad history" because it reinforces their national, regional or ethnic identity, as in the case of the Serbs or those Japanese conservatives who want archaeologists kept out of the ancient tombs of the royal family for fear that the remains found there will indicate that the emperors have non-Japanese ancestors. People seeking to keep the Irish divided once perpetrated the myth that only Protestants fought alongside the British in World War I, when in fact 210,000 Irish Catholics and nationalists volunteered. Others use the past to deflect attention from their own mischief, like the governing elites in China, who dwell on its history of colonialism, persecution and victimization at the hands of the West in order to invalidate any criticism from outsiders as more of the same.

At times the historical citation is solid but its relevance to a current situation is not. However splendid the imperial reach of Ancient Greece may have been, it hardly justified early 20th-century attempts by modern Greece to extend its boundaries into Turkey; hundreds of thousands of refugees were displaced, and bitter enmity between the two nations has been the result. MacMillan faults the second Bush administration for thinking that the occupation of Iraq could be modeled after the reconstruction of post-World War II Germany and Japan, when the differences between the cases outnumbered the similarities.

"Dangerous Games" calls for "professional historians" (by which I think MacMillan means "academics") to "contest the one-sided, even false, histories that are out there in the public domain. If we do not, we allow our leaders and opinion makers to use history to bolster false claims and justify bad and foolish policies." In recent years, she complains, academic historians have become either unduly "self-referential" or preoccupied with "fun" but ultimately insignificant fluff like culture studies.

Recalling a promising student who chose to pursue a particular topic because it was "undertheorized," and fretting over too many investigations of "carnivals of the French Revolution, the image of the Virgin Mary in the Middle Ages, and the role of the doughnut in the Canadian psyche," she worries that the professionals have abandoned the field of political history to the "amateurs."

I'll give her the doughnuts, but I can't see how studying medieval attitudes toward the Virgin Mary constitutes a sidetrack, given that the church was the most powerful institution of the time and that Marian devotions played a significant and sometimes troublesome counterpart to its official doctrine. It's true that you can't necessarily learn much about top-down policymaking from such scholarship, but what you can learn is in some ways just as important, given our increasingly democratic world: how popular movements and loyalties are built. The Chinese government wouldn't be stifling an eccentric and seemingly apolitical spiritual discipline like Falun Gong if it didn't see folk religion as a potential political force.

According to MacMillan, political and economic history, if properly studied, can offer much-needed guidance to today's leaders and can encourage skepticism in the public when demagogues try to justify themselves using distorted images of the past. Like quite a bit of "Dangerous Games," this statement seems so obvious that it hardly needs to be made in the first place. And surely there must be more to it than simply boning up. After all, George W. Bush reportedly spent most of his presidency reading one historical work after another without gaining much in the way of political wisdom. Instead, he scoured the books in search of validation for his own preset ideas. Reading an account of France's battle against Algerian insurgents, he concluded that it failed not because oppressive French tactics alienated Algeria's population but because "their bureaucracy was not up to the job." He launched his war on terror while reading about Churchill, but he finished his presidency by harping on Truman, an unpopular president whose reputation recovered after he left office.

Churchill is a great favorite with the most common breed of history buff, those who, as MacMillan puts it, seek "simplicity" in the past "when the present seems bewildering and chaotic." The perpetual parade of World War II military documentaries that is the History Channel caters mostly to this crew. Noting that Churchillian hero worship is far more prevalent on this side of the pond than in Britain, where more people have had "direct experience" of the prime minister "in other roles than that of the great World War II leader," MacMillan speculates that the fascination lies in the conviction that WWII was "the last morally unambiguous good war." But lack of ambiguity is precisely what makes history "bad." Bad history dispenses with "nuances in favor of tales that belong to morality plays but do not help us consider the past in all its complexity." The British historian Michael Howard calls this species of fairy story "nursery history."

Real history, which resists being reduced to tales of "heroes and villains," can enlighten policymakers, MacMillan says. The better you understand the meaning of "Munich" -- that is, the policy of appeasement used by European and British democracies trying to contain Hitler in the lead-up to World War II -- the less likely you are to misapply it as an analogy. Munich doesn't teach us that "you should never talk to your enemies and try to find common ground," as some hard-liners have argued with regard to everything from the Soviet Union to Iran. MacMillan is kinder than they are to the British Prime Minister Neville Chamberlain, who signed the Munich agreement of 1938. "There is little evidence that Hitler would have dropped his demands in Europe even in the face of stronger opposition from the democracies," she writes. "He was determined upon war sooner or later." The real lesson is in how to distinguish someone unstoppable like Hitler from a dictator like Saddam Hussein, who was, by contrast, both "weak and deterred."

But there's the rub, for as MacMillan points out, it's much easier to find the right historical analogy after events have played themselves out. Bad history may be more likely to lead to bad analogies, but it's clear from the dozens of examples she offers in "Dangerous Games" that you can also draw a bad analogy from good history. And MacMillan doesn't really explain how to avoid that particular pitfall. Perhaps that's too much to ask. The most important lesson to be gained from history, she writes, is simply "humility ... Knowing that classical Chinese civilization valued scholars above soldiers or that the Roman family was very different from the nuclear one of the modern West suggests other values and other ways of organizing society."

To recognizes this is not, she maintains, merely "relativism," but rather the awareness that the threats, incentives and tactics that work on us won't necessarily persuade people in other cultures and situations. Americans, who often have so little understanding of what it's like not to live in an affluent democratic superpower, are especially prone to mistakes in this department, as our experiences in Iraq and Afghanistan have amply shown.

Which is why seemingly lightweight examples like Highland kilts and wedding veils -- pace MacMillan's concerns about the diversions of culture studies -- remain enlightening. When a contemporary dispute with historical foundations is as explosive as, say, the Arab-Israeli conflict or the debate over whether Turkey perpetrated genocide on the Armenians, it becomes very difficult to acknowledge how history can be rewritten or overlooked in order to satisfy the desires of today. By contrast, we can laugh about the booming business in peddling Highlander trappings to people of Scottish descent or about the truly titanic dimensions of the wedding-industrial complex. When the stakes are lower, we can admit that some of our dearest traditions have very shallow roots indeed, however much we'd like to believe otherwise. The past can't be changed, it's true, but perhaps it can be overcome more easily than we realize.