IMAGINE the problem confronting a man in the early ninth century, trying to determine the differences in legal status among a group of assorted human beings in, say, the Frankish state. He might be a high official of the Palace on a mission in the provinces, a bishop counting his flock, a lord taking a census of his subjects. There is nothing fanciful in the situation; we know of more than one actual attempt of this kind, and the impression conveyed is that there was much hesitation and disagreement. In the same region, at more or less the same date, we almost never find two manorial surveys (censiers) employing the same criteria. Evidently, to contemporaries the structure of the society in which they lived did not possess clear-cut contours. The fact was that very different systems of classification cut across each other. Some, borrowing their terminology indifferently from Roman or from Germanic traditions-traditions that were themselves in conflictwere now very imperfectly adapted to the present; others tried their best to express the reality but did it clumsily.