It would be as wrong to restrict the study of villainage to legal documents as to disregard them. The jurisprudence and practice of the king's courts present a one-sided, though a very important view of the subject, but it must be supplemented and verified by an investigation of manorial records. With one class of such documents we have had already to deal, namely with the rolls of manorial courts, which form as it were the stepping-stone between local arrangements and the general theories of Common Law. So-called manorial 'extents' and royal inquisitions based on them lead us one step further; they were intended to describe the matter-of-fact conditions of actual life, the distribution of holdings, the amount and nature of services, the personal divisions of the peasantry; their evidence is not open to the objection of having been artificially treated for legal purposes. Treatises on farming and instructions to manorial officers reflect the economic side of the system, and an enormous number of accounts of expenditure and receipts would enable the modern searcher, if so minded, to enter even into the detail of agricultural management248. We need not undertake this last inquiry, but some comparison between the views of lawyers and the actual facts of manorial administration must be attempted. Writers on Common Law invite one to the task by recognising a great variety of local customs; Bracton, for instance, mentioning two notable deviations from general rules in the department of law under discussion. In Cornwall the children of a villain and of a free woman were not all unfree, but some followed the father and others the mother249. In Herefordshire the master was not bound to produce his serfs to answer criminal charges250. If such customs were sufficiently strong to counteract the influence of general rules of Common Law, the vitality of local distinctions was even more felt in those cases where they had no rules to break through. It may be even asked at the very outset of the inquiry whether there is not a danger of our being distracted by endless details. I hope that the following pages will show how the varieties naturally fall into certain classes and converge towards a few definite positions, which appear the more important as they were not produced by artificial arrangement from above. We must be careful however, and distinguish between isolated facts and widely-spread conditions. Another possible objection to the method of our study may be also noticed here, as it is connected with the same difficulty. Suppose we get in one case the explanation of a custom or institution which recurs in many other cases; are we entitled to generalise our explanation? This seems methodically sound as long as the contrary cannot be established, for the plain reason that the variety of local facts is a variety of combinations and of effects, not of constitutive elements and of causes. The agents of development are not many, though their joint work shades off into a great number of variations. We may be pretty sure that a result repeated several times has been effected by the same factors in the same way; and if in some instances these factors appear manifestly, there is every reason to suppose them to have existed in all the cases. Such reflections are never convincing by themselves, however, and the best thing to test them will be to proceed from these broad statements to an inquiry into the particulars of the case.
Terminological classification.
The study of manorial evidence must start from a discussion as to terminology. The names of the peasantry will show the natural subdivisions of the class. If we look only to the unfree villagers, we shall notice that all the varieties of denomination can easily be arranged into four classes: one of these classes has in view social standing, another economic condition, a third starts from a difference of services, and a fourth from a difference of holdings. The line may not be drawn sharply between the several divisions, but the general contrast cannot be mistaken.
Terms to indicate social standing.
The term of most common occurrence is, of course, villanus. Although its etymology points primarily to the place of dwelling, and indirectly to specific occupations, it is chiefly used during the feudal period to denote servitude. It takes in both the man who is personally unfree and stands in complete subjection to the lord, and the free person settled on servile land. Both classes mentioned and distinguished by Bracton are covered by it. The common opposition is between villanus and libere tenens, not between villanus and liber homo. It is not difficult to explain such a phraseology in books compiled either in the immediate interest of the lords or under their indirect influence, but it must have necessarily led to encroachments and disputes: it has even become a snare for later investigators, who have sometimes been led to consider as one compact mass a population consisting of two different classes, each with a separate history of its own. The Latin 'rusticus' is applied in the same general way. It is less technical however, and occurs chiefly in annals and other literary productions, for which it was better suited by its classical derivation. But when it is used in opposition to other terms, it stands exactly as villanus, that is to say, it is contrasted with libere tenens251.
Villains personally unfree.
The fundamental distinction of personal status has left some traces in terminology. The Hundred Rolls, especially the Warwickshire one252, mention servi very often. Sometimes the word is used exactly as villanus would be253. Tenere in servitute and tenere in villenagio are equivalent254. But other instances show that servus has also a special meaning. Cases where it occurs in an 'extent' immediately after villanus, and possibly in opposition to it, are not decisive255. They may be explained by the fact that the persons engaged in drawing up a custumal, jotted down denominations of the peasantry without comparing them carefully with what preceded. A marginal note servi would not be necessarily opposed to a villani following it; it may only be a different name for the same thing. And it may be noted that in the Hundred Rolls these names very often stand in the margin, and not in the text. But such an explanation would be out of place when both expressions are used in the same sentence. The description of Ipsden in Oxfordshire has the following passage: item dictus R. de N. habet de proparte sua septem servos villanos. (Rot. Hundr. ii. 781, b: cf. 775, b, Servi Custumarii.) It is clear that it was intended, not only to describe the general condition of the peasantry, but to define more particularly their status. This observation and the general meaning of the word will lead us to believe that in many cases when it is used by itself, it implies personal subjection.
The term nativus has a similar sense. But the relation between it and villanus is not constant; sometimes this latter marks the genus, while the former applies to a species; but sometimes they are used interchangeably256, and the feminine for villain is nieve (nativa). But while villanus is made to appear both in a wide and in a restricted sense, and for this reason cannot be used as a special qualification, nativus has only the restricted sense suggesting status257. In connection with other denominations nativus is used for the personally unfree258. When we find nativus domini, the personal relation to the lord is especially noticed259. The sense being such, no wonder that the nature of the tenure is sometimes described in addition260. Of course, the primary meaning is, that a person has been born in the power of the lord, and in this sense it is opposed to the stranger—forinsecus, extraneus261. In this sense again the Domesday of St. Paul's speaks of 'nativi a principio' in Navestock262. But the fact of being born to the condition supposes personal subjection, and this explains why nativi are sometimes mentioned in contrast with freemen263, without any regard being paid to the question of tenure. Natives, or villains born, had their pedigrees as well as the most noble among the peers. Such pedigrees were drawn up to prevent any fraudulent assertion as to freedom, and to guide the lord in case he wanted to use the