A mind-expanding look at how historians go about finding out What Really Happened. Davidson and Lytle, both Yale-trained historians, have no particular methodological or ideological axes to grind. History is a kind of detective-work, they contend, and historical detection is essentially an art that must adapt itself to the investigation at hand--hence the inherent excitement of the process and the luxuriant diversity of its results. A dozen-odd chapters on various events, issues, and individuals from the beginnings of English settlement in America to the present make these points clearly and effectively. The best deal with genuine conundrums: the astonishingly high death rate in the early Jamestown colony: the underlying causes of the witchcraft mania at Salem; what information can be gleaned from Karl Bodmer's paintings of plains Indians; whether a famous Jacob Riis photograph is ""truthful"" or not; how to interpret slave narratives; who killed Silas Deane; where did Woodward and Bernstein get their information for The Final Days. Less persuasive are chapters on the drafting of the Declaration of Independence, John Brown's sanity, the so-called ""Turner Thesis"" about the frontier, Sacco and Vanzetti's guilt, and the decision to use nuclear weapons against Japan--less persuasive, perhaps, because they deal with puzzles created more by historians than by the refractory nature of historical data. Only two could be called real clinkers: a laborious discussion of the passage of legislation to regulate the meat-packing industry; a rather silly inquiry into whether or not Huey Long was a ""great man."" The relation between theory and fact in history is barely probed--but on its own terms the book succeeds abundantly.