Historians inevitably look at what was, but Niall Ferguson also asks what if ... Sian Griffiths talks to him

Niall Ferguson's London flat is like an Oxford undergraduate's set of rooms. True, it is in the heart of Soho, surrounded by strip joints and flashing neon signs and directly above the film-makers Merchant Ivory's offices - but wander up the rickety stairs to his door and you are in a different world.

The rooms are old-fashioned, wood-panelled, sparsely furnished. Basic, an intellectual's space with little to distract the mind. Ferguson's relatively youthful looks are dated by a waistcoat and starchy white shirt, oddly in keeping with his surroundings.

Ferguson, 33, fellow in history at Jesus College, Oxford, is one of a band of youngish Eurosceptic historians who successfully mix academia with journalism, making their names with a torrent of right-wing polemic. His columns appear in newspapers such as The Daily Mail and The Daily Telegraph, opining on issues as one-sided as whether Cherie Blair will be Britain's Hillary Clinton and whether a stock market crash could yet deny Tony Blair a general election victory.

³ÉÈËVRÊÓƵ

But today his topic is academic, and, he says firmly, apolitical. Virtual History, his forthcoming book, is a collection of historians' essays indulging in the ultimate historical parlour game - asking, of nine key moments in history, not "What did happen, and why?" but "What might have happened if..?"

Thus, J. C. D. Clark speculates on "What if there had been no American revolution?"; Michael Burleigh on the nightmare scenario "What if Nazi Germany had defeated the Soviet Union?" and Mark Almond argues that the collapse of communism in 1989, far from being inevitable, was due largely to Gorbachev's naivety and stupidity. The "dyed-in the wool" Democrat Diane Kuhn demolishes the Camelot myth by painting a bleak picture of the possibilities had Kennedy not been assassinated in 1963: "There would have been no early withdrawal from Vietnam. There might have been no Great Society".

³ÉÈËVRÊÓƵ

Ferguson's own contribution speculates on the consequences of Britain "standing aside" in August 1914. He summarises provocatively: "Germany dominates Europe anyway in 1997 - so wouldn't Britain have been better staying out and preserving its own strength and wealth? It might have been more of a contender in the late 20th century."

At first sight it appears to be history as fiction, pure invention. Even more so when you read the afterword, a skip through world history from 1646 to 1996, not as it was, but as the sum of the preceding nine what-if? scenarios. Britain stays out of the first world war, Germany wins the second. Michael Foot is elected in 1983 and again in 1987, following Thatcher's humiliating loss of the Falklands. In 1989-90 Britain breaks up along national boundaries and the Anglo-American empire declines. The Communist bloc, meanwhile, remains intact and rather gleeful.

Is this history? Isn't it rather a political allegory, a use of snippets of evidence to support personal antipathies and passions? No, says Ferguson. "I dare say that people will argue that my chapter about 1914 is a political subtext because people know that my views about Europe are broadly sceptical about Maastricht but I really did not go into that chapter in that way, although I do think it would have been better for Britain not to have intervened in the first world war."

Virtual History is, he says, a serious historical study, its seriousness bolstered by an introduction which is nothing less than a manifesto for a different approach to the discipline - "a new chaostory - a chaotic approach to history". From which one might infer that the state of the subject in universities is something of a mess.

The book was born during a family conversation. Ferguson and his wife, Sue Douglas, former editor of the Sunday Express, were talking about computers and virtual reality. Somehow the conversation got on to alternative outcomes to historical situations; "counterfactual history" as Ferguson terms it. It is an approach which has found little favour in universities. "It has been almost illegal for historians to think in this way," argues Ferguson. "It was banned by E. H. Carr in that awful book What is History?, a book taken as Holy Writ, but, in my view, a load of rubbish."

Douglas is equally unimpressed. "My wife has the advantage of not being a professional academic," he says, a trifle wistfully, an advantage which results "in an impatience with some of the intellectual limits historians impose on themselves".

Such limits are explored in his introduction, which concludes that most of the methodologies that have informed the practice of historians are sadly inadequate. Under the influence of postmodern philosophical theories, the "narrative method" is in vogue. Trendy historians stress the subjectivity of the scholar and the impossibility of objective, evidence-based history. According to this methodology, excavating the facts to write an account of "what really happened" is an impossible task - all the historian can hope to do is put together a literary narrative, a story from his or her own viewpoint.

³ÉÈËVRÊÓƵ

The narrative method is "nonsensical", says Ferguson. All literary narratives have predetermined endings. Comedies end happily, tragedies sadly; a novel's ending must develop logically from the plot. History is not like that.

³ÉÈËVRÊÓƵ

Nor, apparently, is it determinist. Ferguson is tough on determinism, especially Francis Fukuyama's The End of History, which argued that the triumph of liberal democracy across the world following the collapse of communism was the inevitable outcome of historical development. "You can't argue that the unfolding of events has an inevitable outcome," Ferguson argues. Forget E. H. Carr. Instead he urges historians to look to science for inspiration. "Historians lag behind scientists by around a century," he says. "They have no scientific education, they still think science is deterministic - they don't understand that scientists in the 20th century are constantly rethinking the rules of the natural world".

Reading a book on chaos theory given him by his biochemist sister, he concluded that the laws of the natural world are so complex that it is virtually impossible for us to make accurate predictions. Witness the butterfly effect - the flapping of a single butterfly wing today could notionally determine whether a hurricane will hit southern England next week. "Chaos theory rescues us from the nonsensical notions of determinism and of idealism. It reconciles causation and contingency," he writes.

Acceptance of the unpredictability and chaotic nature of the world justifies, in Ferguson's view, the construction of alternative historical scenarios. After all, if the smallest events can have vast unforeseen consequences, the past might easily have turned out differently. Why, then, shouldn't we study how the past might have been? "I am not suggesting that alternative outcomes have a physical existence somewhere else. It is a purely imaginary exercise - to imagine an alternative possible world where Kennedy is not killed at Dallas. But taking that imaginary step is an extremely useful way of understanding our real world." To understand what did happen, we need to understand what might have happened.

The what if ...? game has one inviolable rule. This is that you can "consider as probable only those alternatives which you can show, on the basis of contemporary evidence, that contemporaries actually considered." As Ferguson says: "If you really want to understand what it was like, you have to realise that people, at the time, did not know what was going to happen but they had a number of plausible scenarios in their heads. You have to go back to the documents to work out what the options were and you can see that the option which actually happened was very often not the one thought most plausible at the time."

The rule brings a proper sense of scholarly enquiry to the book and subverts the charge that this is history as fiction. You can see how the method might prove useful in teaching undergraduates, stirring them to question received wisdom and challenge established theories. Indeed, Ferguson admits that is how he uses it. "My students in Oxford would probably say - yes, it's hard to get through a tutorial with Ferguson without having a counterfactual thrown at you. But I am only suggesting that it is one way of approaching historical questions."

It also recognises that historians invariably ask themselves such questions as they sift sources and evidence. "Lots of people who saw draft chapters said 'You can't really do this, it's not history'. But actually they just need to acknowledge that in practice they ask counterfactual questions too," says Ferguson.

Will chaostory emerge as a new school of history? It seems unlikely. What is more plausible is that Ferguson will use the counter-factual approach as one of several historical techniques in his latest project - a history of the Rothschild family, which he describes as an "opportunity to get inside one of the most amazing archives of 19th-century history". Inevitably, once again he is challenging the accepted theory - "this enormous myth that exists about the power of the Rothschilds; the myth of the family who came from the ghetto in Frankfurt and became the richest family in the world. Already classic stories are being straightened out - like the tale of Nathan Rothschild speculating on the outcome of the Battle of Waterloo and making millions as a reward." Pure fiction.

Virtual History, edited by Niall Ferguson, Picador Books, April 21, Pounds 20.

³ÉÈËVRÊÓƵ