When I first became involved in the study of the Irish past, some thirty years ago, the disciplines of history and literary scholarship coexisted harmoniously. Historians of the eighteenth or the early twentieth century saw no problem in enlivening their texts with quotations from Swift or Yeats. And literary scholars were happy to take the historical background to their reading of these and other authors from the standard works of J.C. Beckett or FSL Lyons.

More recently, all this has changed. The two disciplines have moved in opposite directions. The main reason has been a marked divergence of perceptions. The main thrust of much Irish historical writing over that period has been, in the disapproving term of one literary scholar, to 'normalise' the Irish past: to argue that what was long seen as exceptional in its history was in fact not all that far from the mainstream of European experience. Its wars were punctuated with long periods of peace and stability. Its economic history included periods of growth as well as episodes of crisis. Its ruling classes were not, in comparison to elsewhere, particularly cruel or exploitative. All this contrasts with a body of work from literary scholars which continues to start from the assumption that Irish writing from the early modern period to the recent past must be read against a history of exceptional violence, oppression, and cultural deprivation.

The most common response of historians to this divergence is to complain that the literary scholars have not done their homework: that they are working with a model of the Irish past that is decades out of date. But that is to trivialise the problem. For the point about the most significant literary scholars today is that they have read what the historians have been writing, and they have rejected it. Explicit in this dismissal, moreover, is a rejection, not just of particular conclusions by the historians, but of their pretensions to authority.

Historians claim to set popular myth, or literary depictions of a particular period, against evidence of what it was 'really' like. In doing so, of course, they set themselves up for a charge of extreme epistemological naievty. Their repeated revisions of accepted interpretations are themselves evidence that no one version of the past can claim to be authoritative. And we can go further. The very act of creating a narrative of the past is inescapably a process of imposing an order not inherent in the events themselves. And it will be a process shaped by the conventions of a particular rhetoric and particular narrative modes. Instead of a one true narrative of the past we must accept the existence of a range of alternative narratives, each valid within the frame of reference of those who create it.

This is a formidable critique: indeed at one level it is unanswerable. Many working historians, I think, pursue their trade with an awareness of the unavoidably limited scope of their vision, in much the same way that scientists are aware that whole dimensions, literally, are bound to be missing from any model of the universe that they construct. But there are two points to emphasise. The first is that to admit that more than one narrative of the Irish past is possible, is not the same as saying that any narrative is as good as any other. That a glass can be described as both half full and half empty does not mean that the amount it contains is a purely subjective matter.

Secondly, if we do go down the road of saying that there is no one true history of the past, that all we can offer a variety of equally valid literary constructions, then it follows that we have to abandon altogether the idea of linking literary criticism to historical analysis. We cannot be selective in our scepticism. In other words, it is not legitimate to brush aside a historian's inconvenient findings, about the level of rents in the 1850s or the operations of the penal laws, as merely one narrative among many, and yet go on to give a privileged status to an alternative narrative, in which the traditional themes of oppressive landlordism or parasitic Protestant ascendancy are restated as simple fact.

This is not to say that the historians are off the hook. Anyone committed to normalising the Irish past does have to ask serious questions about the dominant themes in the writing coming out of Ireland across the period concerned. Is it a coincidence that so much of this work, from Gulliver's Travels, Tristam Shandy and Bishop Berkeley's meditations on the nature of the material world, to Sheridan Le Fanus's Irish Gothic and the work of Samuel Beckett and Flann O'Brien, is concerned with the irrational and the grotesque? There is also, in works as far apart as those of Sheridan and Wilde, what seems to be a pervasive obsession with the difference between appearance and reality. And, of course, there is the much discussed failure of Irish writers across a long period to produce a fully achieved novel of social analysis, an Irish Middlemarch. All this can convincingly enough be cited as evidence that we are after all dealing with an unusually fractured society.

On this basis, I would propose a truce. Literary scholars have to accept that any analysis based on data from rent rolls or mortality statistics cannot simply be dismissed as just one narrative out of many - at least not unless one wants to abandon meaningful discussion of the past altogether. They also have to recognise that the writings of a Berkeley or a Swift are a dangerous source from which the reconstruct the social and political conditions of the past. Historians, on the other hand, have to consider the possibility that the sources with which they work are by their nature likely to direct them towards the mundane and the routine. They can also concede that the work of a Swift or a Berkeley can contribute alternative and equally valid insights - that what makes figures like these significant as artists is precisely a sensitivity that allows them to perceive fault lines of which others were less conscious, but which nevertheless were there and were influential. On that basis both enterprises, history and literary scholarship, could perhaps move forward more productively than they have managed to do in recent years.