Igniting Marx with pomo sparks

I used to think that the profession of history, unlike that of, say, nuclear physics, could at least do no harm. Now I know it can." Eric Hobsbawm is talking here of the misuses of history - of the building of political myth - but the corollary to his point is the desirable, positive, "danger" that history can present. This "danger", this use of history to upset, to revolutionise, has of course been a large part of Hobsbawm's life's work. So my central question, confronted with this collection, is this: is Hobsbawm's history still dangerous?

On History was published earlier this year on the occasion of Hobsbawm's 80th birthday. The book comprises 21 chapters, six published for the first time, dating from the late 1960s to the late 1990s. If one can distil themes from a score of essays written over a similar number of years, the core of On History might be arranged as a triptych. First, Hobsbawm is concerned with the "uses and abuses" of history. Second, he addresses changes in historiography in the past two centuries, with a particular desire to puncture and discard that most misleading of shibboleths, postmodernism. Third - and most importantly - he wants to persuade us of the benefits and strengths of Marxist history, or perhaps "Hobsbawm's history".

In the first thrust of his arguments, Hobsbawm engages with the "history and heritage debate". With customary panache, Hobsbawm puts the central danger thus: "History is the raw material for nationalist or ethnic or fundamentalist ideologies, as poppies are the raw material for heroin addiction ... If there is no suitable past, it can always be invented." Invented pasts are dangerous and self-deluding. It is the duty of the historian, Hobsbawm argues, to lift the blindfold of historical myth, no matter how unpopular this may be, so that society can get to grips with its present state and the direction in which it is headed.

However, this does not mean that historians should sit outside society, providing truths from a position of Olympian detachment. In "Partisanship" Hobsbawm shows us the tricky path a (Marxist) historian must take in order to stay true to the objective principles of his or her craft, while engaging effectively in political combat. The brake on partisanship is the need to abide by the rules of the game: to take due account of evidence and to submit in the face of persuasive argument. However, academics should embrace partisanship through realising that the things we mutter and scribble in our ivory towers do have effects on the world outside. Without this wider view, we fall victim to intellectual solipsism.

This is a powerful argument, which will speak strongly to a younger generation of scholars in need of some goal beyond the next research assessment exercise. However, Hobsbawm goes on to reveal several things about "Hobsbawm history" that give me pause. "Hobsbawm history" is necessarily Marxist, of a particularly patrician brand. There are - as the author points out - parts of Marxism that have been so completely digested by general historiography that their ancestry is no longer noticed. But to accept Hobsbawm's view of history we have to embrace two propositions: first, that Marx's approach "is still the only one which enables us to explain the entire span of human history"; and second, that Marxist history enables us to make predictions.

The first point is uncertain. Others have "explained" the entire span of human history according to their own lights. The weight of the case then rests on whether or not Marxism is better able to provide predictions. To see Marxism as a predictive science requires a belief in objectivity and absolute knowledge that contradicts the earlier points on partisanship and myth. Hobsbawm is careful not to overplay the case: history cannot predict absolute outcomes or specific events. However, he argues, it can and has predicted tides of change. For example, the Russian revolution was widely predicted at the time.

This strikes me as problematic, nudging the historian into the kind of a priori privileged position assumed by the equally silly current interest in "counter-factual" questions. Predictions made after the fact are obviously to be treated with suspicion. Emphasising that predictions (that turned out to be correct) were made at the time is obfuscatory; other predictions in other situations have not come to pass. The problem is not, perhaps, in making predictions, but in how they are presented and understood. Hobsbawm wishes to gloss these rhetorics of utopia and change as both "scientific" and "natural". They are scientific in a rather old-fashioned sense of presenting a hard, objective analysis of "the facts". They are natural in that we cannot survive without them. The alternative, Hobsbawm suggests, is to act as the audience to a horror film - "to assume that anything can happen at any time".

This last analogy gives the game away. The audience for a horror film usually knows exactly when something will happen, and what sort of thing it will be (usually unpleasant). The grammar of the horror film is precisely geared towards audience anticipation. But this is within the confines of that genre; what Hobsbawm is trying to do is to imagine a similar grammar of historical study that allows for predictions. In other words, he is engaged in an exercise in rhetoric.

There is absolutely nothing wrong with this - rhetoric spins the world in the directions we desire. If you can persuade enough people that your prediction is a good one - that, for example, there ought to be a revolution - then perhaps that prediction will come true. But this sits uneasily with Hobsbawm's other strand of argument: the poverty of postmodernism. He is hampered by two things here: he does not really grasp the theories lumped unhappily under that banner; and much of what he says agrees with those theories. Hobsbawm rails against the idea that discourse (language) constructs reality, as he fears that this will undermine the legitimacy of "the facts". However, his refutation cheats by swinging between "event fact" (that Elvis is dead) and "opinion fact" (Hobsbawm's view of historical causation). The final lines of the collection read thus: "Bad history is not harmless history. It is dangerous. The sentences typed on apparently innocuous keyboards may be sentences of death." How so, if not because language constructs reality?

Hobsbawm's history depends on persuading the audience of the authority of its purveyor. The crux of the matter is presented in one, throwaway sentence: "It takes two to learn the lessons of history: one to give the information, the other to listen." Hobsbawm cannot imagine any other power relationship between the historian and the public, or indeed between history and the public. This is not to say that Hobsbawm is "wrong". It is to note that Hobsbawm history - like any history - depends upon rhetoric. Hobsbawm writes elegantly and memorably. When talking about prediction, he notes that "we dream forward", which expresses perfectly (and honestly) the power and beauty of the Marxist project. But as he cannot wholly come to terms with the rhetorical foundation of his discourse, he has to invent bugbears such as postmodernism to legitimate his own position. This renders his rhetoric less persuasive, less dangerous. Marxism was "designed to blow up crucial parts of the fortifications of traditional history". Now that Marxism has become traditional history, it may be the turn of others to provide new dangers and explosive ideas.

But still, there has been wit and wisdom along the way, and I have only engaged briefly with the huge breadth of ideas and arguments presented here. At one point Hobsbawm modestly hopes that he has made some contribution to historical thought. If not, he says, "nobody can possibly deny that I am enjoying myself enormously". For this I am glad; and for the enjoyment of ideas, intellect and historical engagement that he has given to others. A belated but heartfelt "happy birthday" to you, Professor Hobsbawm.

John Arnold is a lecturer, University of East Anglia.