Susan Reynolds's work is part of the urgent process of reassessment which has astonishingly turned medieval history into exciting new ground. Historians are casting off the word feudalism which Spelman invented in the 17th century and which has so distorted our view of the Middle Ages. Although European scholars have been working on this new approach for almost a generation some English historians have been laggardly.

Reynolds reiterates the salutary demand that those who write - or think - about the Middle Ages should be careful of the words they use. Our accustomed vocabulary must be scrupulously scrutinised for its validity and meaning at the time that is being discussed. This is true not just of feudal terminology but also words like knight and free which underwent social and legal evolution. This basic scholarly requirement has extra bite because so much of medievalist terminology had acquired anachronistic and even mythical meaning. Many words have taken on a life of their own. Hence the too-common habit of placing them in inverted commas, apparently to exonerate the author from responsibility and leave the reader to supply whatever meaning is considered appropriate. Worse, in the growing body of source translations such words can too often be found to have slid into the translations without authority from the text.

Reynolds's enemy is modern ideas of medieval feudalism based, she suggests, on early European academic lawyers who created elaborate schemes of regulation which have been wrongly taken by historians for a reality from which they were utterly remote. Reynolds might have added that the myth-making started even earlier - in the 12th century with the chansons de geste which presented an image of the knight as unreal as the Hollywood image of the cowboy, though with greater social consequences.

But Reynolds has deliberately omitted knights from her considerations. Instead she has picked the words fief and vassal to rescue from misuse and misapprehension. One can say that feudalism never existed but fiefs and vassals did. Concern over them is of a different order, whatever false images may or may not prevail they can be traced in detail. Like the words thegn and knight, vassal changed its meaning and came up in the world. It began as one of many words to describe a subordinate, deriving ultimately from a Celtic word for slave. Reynolds's conclusion that there was a fundamental change in fiefs in the 11th and 12th centuries chimes with current chronology. But her book has too many vague denunciations of what "people" have thought taking up space which might otherwise have been devoted to what actually was.

ÁñÁ«ÊÓƵ

Reynolds makes the important point that the problems of the great churches in managing their lands, their tenants, and the necessary protective force, differed from those of lay landholders so that the great body of evidence which the churches have left cannot necessarily be transposed to lay lords. This indeed is why the recent analyses of vernacular literary sources such as those of Jean Flori have made such a profound contribution to social history. However, although the church records were more elaborate and better kept it would be difficult to argue convincingly that lay lords were a softer touch than churchmen. It may be moreover that those who did not have the security of writing would rely more heavily on ritual, symbolic submissions, loyalty myths and the application of the strong right arm. But the institutions of an oral or vernacular culture are even less likely to be adequately recorded.

Most difficult to follow is the book's thesis that there was no hierarchy of property rights. The statement that there was no difference between tenure and ownership would have startled those obliged to pay hugely extortionate reliefs in order to enter into their inheritances.

ÁñÁ«ÊÓƵ

We need to know what we mean by the words we use, but we have also to recognise that meaning could vary in medieval usage. There was no force to impose uniformity even in a single country. Time as well as location brought change. Medieval Europe and the widely differing entities of which it was composed was very different in 1400 from what it had been in 900. These were living societies, not developing as fast as in times of mass communication and mass production but nevertheless changing, trying out new things. Reynolds's villains, the lawyers and academics, were trying to keep up with a society bursting at its intellectual seams. Reality changed, and real people dealt with change as best they could and with what words they knew. So the meaning of individual words had to encompass that change, and stretch to the new circumstances until a better one was found. But how to get a new term into circulation? King Alfred tried using "vassus", but it didn't catch on - probably because a vassus arriving somewhere would have to try to explain what a vassus was. It was easier to keep on using King's Thegn even if it could mean practically anything. A quick look at the furs, arm rings, horses and number of servants would provide the necessary definition, and distinguish a King's Thegn to be given the best of everything from a king's thegn who could bed down with his horse.

Few would argue with Reynolds's plea to treat the Middle Ages as a real world not a formulaic parlour game with rigorous rules enforced by unseen powers.

This is not a book for the faint-hearted, the non-specialist, or the impatient. It is an interesting stimulus likely to send working historians back to examine time-honoured traditional texts to see whether or not the time-honoured traditional terms are actually there - and what their presence or absence implies.

Jean Scammell is a former research fellow of Clare Hall, Cambridge and lectures in the history faculty, Cambridge University.

ÁñÁ«ÊÓƵ