What is history? It's not simply gathering facts about the past, it's about searching for truth in them, says Richard Evans in the first of our series on Big Questions in History

The answer to the question "what is history?" seems obvious enough: history is the study of the past. But, of course, it is not quite as simple as that. There are some ways of studying the past that cannot be classified as history. History is, in the first place, the study of the past in order to find out the truth about it. Unlike novelists or film-makers, historians do not invent things that did not happen or conjure up characters that did not exist.

Playwrights and screenwriters can change the raw materials they use when they are dealing, as often happens, with a topic drawn from the past to make the subject more interesting and more exciting. They can make up dialogue, insert words into historical documents that are not in the originals, and generally use their imagination in a manner unfettered by the constraints of the historical evidence. Historians have no such luxury. They deal with fact, not fiction.

This distinction has been made by all historians ever since the first serious historical work to have come down to us from the ancient world, the History of the Peloponnesian War . Its author, the Greek writer Thucydides, rejected the romantic myths purveyed by the poets and checked all his evidence, as he told his readers, "with as much thoroughness as possible".

ÖйúAƬ

But he went on to complain, as historians have done regularly ever since, that the truth was far from easy to discover: "Different eyewitnesses give different accounts of the same events, speaking out of partiality for one side or the other or else from imperfect memories." In the two and a half millennia or so since Thucydides wrote his great work, historians have elaborated a whole battery of sophisticated methods of checking the evidence and dealing with the gaps and partialities of their sources.

But they can never attain perfect or total knowledge of the whole truth. All they can do is establish probabilities - sometimes overwhelming, sometimes less so, sometimes hardly at all - about parts of the past: those parts that can be accessed by means of the remains it has handed down in one form or another to posterity.

ÖйúAƬ

History only ever involves a selection of what is knowable about the past because it has a second essential quality apart from the search for truth: it aims not just at reconstructing and representing the past but also at understanding and interpreting it. This is what makes history different from chronicle, which tells the tale of the years, marking off events as they happened, but does not try to make any connection between them or attempt to explain why they occurred.

The centrality of explanation and interpretation to history also make its approach to the past different from those of religion, morality and the law. Religions seek legitimacy through sacred texts handed down by prophets or their disciples from the distant past. To treat such texts historically, however, means to put their sacrality to one side and to question them just as one would question any other historical source, a procedure undertaken most powerfully by the greatest of the historians of the Enlightenment, Edward Gibbon, in his Decline and Fall of the Roman Empire .

Moral and legal approaches to the past are concerned with judging guilt or innocence and assigning responsibility for actions that are classified as good or evil, lawful or criminal. These, too, are unhistorical ways of dealing with it. In recent times, it has become fashionable to categorise historical figures from a time such as the Third Reich, or the Atlantic slave trade, or the European settlement of Australia, in terms derived from morality and the law: as "perpetrators", "victims", "bystanders", "collaborators" and so on, and to distribute praise and blame accordingly.

This is profoundly alien to the enterprise of history, which is concerned in the first place with explaining why people did what they did, with causes, effects and interconnections, not with issuing arrogant verdicts on complex moral issues based on the luxury of hindsight.

Of course, historians can, do and in many cases have an obligation to provide raw materials, evidence or background briefings to assist institutions such as war crimes tribunals or commissions assessing claims for compensation for legally recognised historic wrongs, just as another important side of their work lies in producing scholarly editions of previously unpublished documents. But such a deployment of expertise, however necessary, is not the historian's main business. The historian's job is to explain; it is for others to judge.

This means, among other things, that historians have to try to understand the past from as wide a variety of points of view as possible, not to see it through the eyes of one particular contemporary or group of contemporaries, still less to study it exclusively in the light of the concerns of the time in which they are writing. History written purely to fulfil a present-day purpose, such as encouraging national pride or showing that one ethnic or national group has been oppressed over the ages by another, is all too likely to degenerate into propaganda unless it is held in check by a willingness to bow to the dictates of the evidence where the evidence runs counter to the historian's purpose.

Nevertheless, at the same time, history also inevitably involves formulating hypotheses on the basis of present-day theories and testing them critically against a thorough review of the evidence. Historical perspectives on the past change not just with growing distance in time but also with the changing ideas and interests of historians themselves and the developing ideas, methods and concerns of the intellectual world and the society within which historians live.

That is one important reason why, over the years, history's scope has been steadily expanding. The days when it was concerned solely or even principally with kings and battles, politics and diplomacy, "great men" and great wars are long gone. In the 21st century, everything is grist to the historian's mill.

ÖйúAƬ

ÖйúAƬ

Big questions involve the history of private as well as public life, of ideas and beliefs, of personal behaviour, even of broad topics such as the environment, geography and the natural world. They can be asked about any part of the world, any era of the past. All with one proviso: research into these areas is history only if it really is undertaken in search of the answer to a "big question". History is not, and never has been, the mere accumulation of facts and knowledge for their own sake: that is better categorised under the heading of antiquarianism.

Of course, historians have always disagreed among themselves about virtually all of these points, as they have about most answers that have been put forward at one time and another to big questions about the past.

Controversy is an indispensable means of advancing historical knowledge, as the rough edges are rubbed off implausible or exaggerated interpretations, and reasoned debate consigns the unsupported argument to the dustbin of discredited hypotheses.

The pervasiveness of controversy among historians is one reason why politicians are always wrong when they claim that "history" will absolve them, judge them or vindicate what they have done. Historians will probably never agree about issues on which national leaders have made such claims, whether it is the Cuban revolution or the second Iraq war.

The historian's training can generate a healthy scepticism with which to puncture the wilder claims of politicians and statesmen. It can, or should, help anyone who undergoes it to spot a fake when they see one, and to demand clear evidence for a statement before they accept it. Before rushing into print to denounce a politician's alleged statement that medieval history is not worthy of state support, for instance, medieval historians should have recollected their training and demanded to see a copy of his speech rather than accepting a tendentious, second-hand account of it from a journalist.

Medieval history is as useful in this respect as modern history is, and in others, too: neither is necessarily more "relevant' than the other.

Training as a historian is essential for a whole variety of jobs in the heritage industry and more than helpful in the wider field of culture, tourism and the arts, which generate a far higher proportion of national income and export earnings than the manufacturing industries do nowadays.

History books, television shows, radio broadcasts, magazine articles and other cultural products have never been more popular.

History in this broad sense is a major national economic earner. But its most important justification lies in its less immediately tangible effects.

History can teach us about other societies, other beliefs and other times, and so make us more tolerant of differences in our world. And it can provide us with a democratic civic education to help us to build a better world for the future.

ÖйúAƬ

Richard J. Evans is professor of modern history at the University of Cambridge.