The boundaries addressed in this volume are of various kinds. They include not only the geographical, political and cultural frontiers that medieval romances imagine and imply, but also more metaphorical demarcations, such as the boundary between reality and fictionality; boundaries between different literary traditions, modes and cultures; and boundaries between different kinds of experience or perception, especially the ‘altered states’ associated with sickness, magic, the supernatural or the divine. All of the essays in this volume demonstrate in one way or another how medieval romances frequently, and perhaps characteristically, capitalize on the dramatic or suggestive possibilities implicit in all such boundaries. Indeed, one conclusion that might be drawn from this collection is that medieval romances both define and challenge boundaries more self-consciously and more provocatively than is usually acknowledged. Yet romance is, as a genre, notorious for its indifference to limits – its apparent readiness to breach the rules both of literary decorum and of literary realism. As a result it has often been interpreted simply as an instrument of – or an excuse for – a form of fantasy that is self-indulgently indisciplined, and therefore intellectually trivial. This has sometimes been seen, in turn, as a reflection of the relatively unsophisticated nature of medieval romance's audiences (and authors). More recently, this has been translated into a more serious allegation, the suggestion that romance implicitly represents a profound form of intellectual failure, dishonesty or aggression. Such criticism perhaps says more about the preconceptions of medieval romance's critics, the intellectual limits in which they choose to confine it as a genre, than about the specific effects or achievements of individual romancetexts – the way in which they themselves employ the various different kinds of boundaries mentioned above. By contrast, the premise that all the essays in this volume share is that romance is a complex and dynamic phenomenon, and considerably less amenable to generalization than it is often taken to be, so that understanding its boundaries is a process that both requires and repays detailed and sympathetic attention to individual texts.

There is, though, some truth in the suggestion that the imagined worlds of medieval romance tend to be, by nature, rather unlocated and/or unlocalizable. One explanation for this is as a form of narrative efficiency, in the sense that it is the consequence of the subordination of setting to plot.