Vicinage and the Antecedents of the Jury

MIKE MACNAIR

The origins of trial by jury have been the subject of an extensive debate. 1 The traditional approach to the creation of the "palladium of liberty" saw the jurors as lay judges and located their origins in the lay judges of Anglo-Saxon England, continuing through the survival of the "ancient constitution." An alternative approach, that of Heinrich Brunner, found wide acceptance from the end of the last century and until recently. Brunner detected the origins of the jury in fiscal inquiries imposed by strong monarchs, reversing the constitutional politics of the older view. At present, the wheel has turned back toward judicial character and origins in general early medieval practice. The purpose of this article is to take a new look at the issue by approaching it from a different angle: the requirement that jurors should come de vicineto, from the locality. This approach has produced the following observations.	1
First, specially convened groups of vicini, locals or neighbors, were quite widely used as a mode of proof for "local" facts (like boundaries) in early medieval normative sources, following late Roman antecedents. The gist of the idea is not that locals will necessarily have direct knowledge of the facts, but that local reputation about them—including hearsay—is itself probative. These uses will explain much of the evidence for the use of groups of locals as a mode of proof in late Anglo-Saxon and Anglo-Norman law.	2
Second, early medieval uses of local reputation as a form of evidence were incorporated into the Roman-canon law of proof, and the perceived value of this sort of evidence was at a high point in the later twelfth century, when the use of jury-like bodies began to be generalized and systematized in the English royal courts. I suggest that a part of this process of generalization is compromises made between the royal courts, which at this time were committed to the judgment of God in the form of trial by battle and by ordeal, and the contemporary ecclesiastical courts, which rejected trial by battle and sought to use documents, witnesses, and procedure per notorium on the basis of the personal knowledge of the ecclesiastical judge. Local reputation was an acceptable canonical mode of proof (unlike trial by battle), which had already been practiced in England for more limited purposes and met the objections of the king and lay landowners to the procedural claims of the church. I suggest that Glanvill's characterization of the jurors as witnesses, and the persistence of this idea and the idea that the jury was only appropriate to facts of a certain type, reflect this context.	3
Third, in relation to the political history of the jury, this view implies that the role of the jury as the palladium of liberty was neither foreordained by Anglo-Saxon or early medieval roots, nor produced by a dialectic of strong government bringing forth self-government. The jury was produced in intimate, if antagonistic, contact with the canon law of proof, and down to the time of Bracton, if not later, it was still possible that either the separate examination of jurors, or the use of individual witnesses, would develop. The fact that it did not develop within the common law jurisdictions is a matter of later medieval political and legal history. 2 I would suggest, however, that the internal conceptual contradictions of the idea of the jurors as witnesses to local reputation made it highly suitable for compromise solutions to the problem of the distribution of power in the lawsuit.	4
The first part of this article gives a summary review of the Brunner thesis on the origins of the jury, the objections that have been made to it, and the currently preponderant alternative view. My primary objection to the recent view of the jury as a survival of collective judgment is its failure to account for a good deal of later medieval common law doctrine about the jury, which thought of the jurors as witnesses, as distinct from judges, and of the verdict as a mode of proof appropriate only to certain types of fact. I will therefore also lay out some of this doctrine. The second part considers early medieval normative sources on the use of vicini and local reputation and how this normative material relates to the Anglo-Norman uses of panels of locals in litigation. The third part looks at some of the uses of local reputation in the later medieval Roman-canon law of proof and at the possible relationship between the canon law and the legislation of Henry II, which systematized the use of panels of locals in presentment of crime and land cases. The concluding section considers briefly the larger issue of the implications of this view for the political history of the jury. The range of the issues posed by the question of the origins of the jury inevitably means that I will paint with a fairly broad brush; but I think that this is justified by the interpretive possibilities opened up by this approach.	5
I. Theories
1. Inquisitio: The Brunner Thesis and Its Critics
Traditionally, the jury was seen as a bastion of the liberties of the subject, and in consequence it was associated from the seventeenth century with the "Anglo-Saxon liberty" that distinguished England from the Continental absolutist Roman-law regimes. The origin of the jury was therefore located in the courts of Anglo-Saxon England, whose judges were (like jurors) laymen, rather than professionals; in the early Middle Ages, all courts were characteristically composed of groups of laymen, who decided both factual and normative questions, or at least decided the form of proof (testimony, documents, compurgation, ordeal, or battle) that would settle the dispute. 3 This view indicated that the jury was from the outset a judicial body, as it is today.	6
In the nineteenth century, it was suggested by Sir Francis Palgrave that the origin of the jury should be separated from the lay judges and from the jury's later judicial role. Rather, in its origin the jury was a piece of fiscal machinery for the compulsory interrogation of the inhabitants of a locality in the interests of the Crown, which was later extended to private litigants. It was introduced by the Norman kings following the fiscal practices of the Carolingian Empire. This suggestion was taken up and very considerably elaborated by Heinrich Brunner in his book Die Entstehung der Schwurgerichte (1872). This view sees the origins of the jury as testimonial, rather than judicial. It also has large implications about the political history and legal politics of the jury. Instead of the palladium of liberty descending from Anglo-Saxon or Germanic liberty, in the Brunner thesis the jury is in origin the creation of the fiscality of an aggressive centralizing monarchy, and not even "national" in origin. Strong royal government, as it were, trains the subjects for self-government.	7
A clear example of the Carolingian materials Brunner relied on for this derivation is a capitulary of 819:	8
Volumus autem ut de his libertatibus et rebus reddendis quae in nostra vestitura sunt primo per optimos quosques inquiratur; et si per illos inveniri non possit, tunc per eos qui post illos in illa vicinia meliores sunt; et si nec per illos rei veritatis inveniri potest, tunc liceat litigantibus ex utraque parte testes adhibere; et si discordaverint secundum constitutionem a nobis promulgatam examinentur. (However, we wish that in relation to liberties and things which are in our possession whose return is claimed inquiry is first as far as possible to be made from the best and if [the truth] cannot be found from them, then through the next best persons in the locality. And if the truth cannot be found from them either, then the litigants on either side may call witnesses. And if they should have disagreed, they should be examined according to our constitution.) 4
Brunner reinforced the relation to royal interest, which is apparent in this and some other capitularies, by suggesting that there might be a relationship with the late Roman law of fiscal claims to bona vacantia and forfeited goods. 5
The sequence of jury use in England is then: use in the Domesday inquiry for fiscal or near-fiscal purposes; use in the later eleventh and twelfth centuries by special concession; 6 regularization under Henry II, while retaining a royal monopoly 7 and a limitation to title to land, which under the English feudal regime had fiscal implications; subsequent generalization. Brunner's account obtained very wide support. It was followed by J. B. Thayer, by F. W. Maitland (with some reservations), and by W. S. Holdsworth. J. P. Dawson's book Lay Judges (1960) similarly takes Brunner's account as the starting point for the history of trial by jury, as does J. H. Baker's standard textbook on English legal history. 8	9
However, substantial objections have been made to Brunner's argument. In the first place, the argument that the inquisitio continued to be used in pre-Conquest Normandy and was imported from there to England was weakened by the silence of the pre-Conquest Norman sources and by the doubtful continuity between the Carolingians and the Normandy of William the Conqueror. 9 R. C. van Caenegem opposes to this objection the thinness of the Norman sources in general and the consequent danger of an argument from silence. 10 However, fiscal institutions are prima facie those most likely to leave records behind, 11 and conversely the least likely to survive across disorderly periods like the tenth and eleventh centuries. The latter point could be met by Goebel's argument that aspects of Carolingian criminal law and procedure countinued in use through their adoption by the "immunists," the holders of private jurisdiction, particularly ecclesiastical. But while Goebel found evidence for other aspects of Carolingian practice from the immunity jurisdictions, the fiscal inquisitio is not among them. 12	10
The second line of objection is that Brunner's derivation is over-specific. At first this argument was largely focused on the jury of presentment. Here prior Anglo-Saxon communal duties to present crime could be found in the provision of Ethelred's Wantage Code of a.d. 996 for twelve thegns who were to swear that they would accuse no one wrongly: an argument strongly developed by N. D. Hurnard in 1941, though since criticized by R. C. van Caenegem. 13 Recently Patrick Wormald has carried this line of argument further, linking the twelve thegns to general public duties to report crime found elsewhere in Anglo-Saxon law and in Carolingian sources—incidentally weakening the link between the Carolingian sources and the specific form of the jury of presentment offered by Brunner. 14 These suggested Anglo-Saxon antecedents, however, remained testimonial rather than judicial bodies; and the strength of royal government had merely been pushed back into Anglo-Saxon, rather than Anglo-Norman, England. 15	11
Another possible source for the jury of presentment is the use of jury-like bodies to present crime in the ecclesiastical courts; this was noticed by Maitland, C. H. Haskins and Hurnard, and extensively developed by R. C. van Caenegem, 16 while R. H. Helmholz has shown that jury-like bodies continued to be used for the establishment of fama, public suspicion of crime, in the later medieval English church courts. 17	12
The core of Brunner's argument, however, was the fiscality of the use of the jury in land title matters. It was a substantial blow to this that van Caenegem, in his English Writs from the Conquest to Magna Carta (1958), found considerable numbers of "recognitions" of locals in local, feudal, and church courts and arbitrations in the late eleventh and twelfth centuries. He also found an associated phenomenon of perambulations of land to establish boundaries. The appearance of the jury in local and feudal jurisdictions after it had become routinely available in the royal courts would be only natural. Its appearance there alongside its earliest appearances in the royal courts suggests that what we are concerned with is a mode of proof for facts of a certain type, rather than a royal privilege. Having found that the earliest example of the use of locals to establish boundaries appeared more than ten years before the Conquest, van Caenegem suggested that there must be deeper Anglo-Saxon (or Danish) roots. D. M. Stenton went further to argue that van Caenegem's evidence showed that a Norman derivation was wholly unnecessary. 18 On the other hand, further study has led van Caenegem to argue that institutions of lay fact finding seem to have been common in Europe, and particularly in northern Europe, at this period. 19 As might be expected, the institution closest to the English jury is that found in Normandy more or less contemporaneously (which gradually decayed after the destruction of the Angevin empire). 20 Institutions of lay fact finding also existed far more widely. Examples can be found in wider areas of France, Norman Sicily, Spain, parts of Germany, Scandinavia, the Low Countries, Hungary, and Serbia. 21 The lay decision of disputes of fact was thus quite widespread in high medieval Europe.	13
These two lines of argument are further developed by Susan Reynolds and by the contributors to the book The Settlement of Disputes in Early Medieval Europe (1986). These authors point to quite widespread use in the early Middle Ages of sworn panels of neighbors—not restricted to summons by royal authority, or to fiscality. Reynolds says that "Legal historians have seen these inquisitions as intrinsically different in legal principle from other collective judgments, but to contemporaries they may have been just a way of discovering the truth which was most readily available to a president of high authority and most appropriate for issues which required special knowledge." 22 Similarly, the Settlement of Disputes authors comment that judicial practice "put oath-helping, witnessing and the inquest all in the same spectrum," and that "it is not at all clear that it was the Carolingians who 'invented' the inquest, or even that they used it to any great extent." 23 What is distinctive about the English jury, these authors argue, is merely its regularization, due to the precocious development of royal law in later twelfth-century England, before the Roman-canon inquisitorial use of witnesses was developed. 24	14
Paul R. Hyams has found an analogous but not identical link between the jury and the absence of a distinction between testimony and judgment in early medieval judicial procedure. He argues that the early common law jury functioned as a substitute for the "judgment of God" (trial by ordeal and by battle). Hence it shared with these procedures the early medieval characteristics of being a formal proof, which substitutes for detailed enquiry into evidence, and a proof that closes the dispute, concluding both factual and normative debates. 25 But Hyams's account stresses the conceptual centrality of the judgment of God to early medieval judicial procedure, while The Settlement of Disputes is critical of this view. 26	15
In effect these views reinstate the idea of the jury as a survival of early medieval "liberty," in the sense of communal, as opposed to official, judgment—albeit without the nationalistic overtones of the original thesis of Anglo-Saxon liberty. However, the survival of the jury into the later Middle Ages is still made to depend on strong monarchy, as it is explained by the precocious strength of the English Crown.	16
These views are based on studies of practice; it can be added that the normative sources cited by Brunner do not, in fact, support an intimate link between fiscality and the judicial summons of locals as a mode of fact finding. In the first place, the sources for the late Roman fiscal inquisitio palatina cited by Brunner from the Theodosian Code provide for an official enquiry in the locality in which bona vacantia or forfeited goods are found. But they make no provision whatever for how this enquiry should be carried out, for instance, by asking a group of locals, which is, in contrast, the focus of the relevant Carolingian capitularies. 27 Second, the capitularies themselves contemplate—the capitulary of 819 quoted above is a good example here—that inquisitiones have previously been carried out using witnesses produced by the parties. Inquisitio, in fact, prima facie merely means an enquiry, and is not tied to any particular mode of proof: a proposition that remains true in the twelfth and thirteenth century sources. 28	17
If, however, we are forced back on the idea that "inquests" are just another mode of early medieval testimony, selected pragmatically, we are left without an explanation of some significant facts about the later medieval conception of the jury—in particular, its doctrinal restriction to specific classes of facts, and the extent to which bits of Roman-canon learning about witnesses are thought to apply to jurors. These, therefore, now need to be considered.	18
2. The Testimonial Theory of the Jury in Later Medieval Common Law Doctrine
Though the dominant tendency of recent literature has been to emphasize the judicial aspects of the role of the later medieval jury, 29 there is much later medieval common law legal-doctrinal evidence for the view that the jurors were in some sense witnesses, or that the jury verdict was a type of proof tendered to the court, which rendered judgment on the basis of one of a number of types of proof presented to it. First, there is direct evidence for treatment of the jurors as a species of witness: the challenges and the attaint. Second, the jury were supposed to have knowledge of the facts, a principle reflected in the requirements of vicinage and venue. Third, the doctrinal relationships between verdicts and some other modes of proof point to a definite conception of the jury as one among several modes of proof, and one apt to facts of a particular type.	19
(a) Jurors as witnesses: The challenges and the attaint. Glanvill says of the grand assize that it is to be preferred to battle, among other reasons, because "in proportion as the testimony of several suitable witnesses in judicial proceedings outweighs that of one man, so this constitution relies more on equity than does battle; for whereas battle is fought on the testimony of one witness, this constitution requires the oaths of at least twelve men." 30 When he comes to identify possible exceptions (challenges) to jurors, he says merely that "the grounds for taking exception to these jurors are the same as those for rejecting witnesses in an ecclesiastical court," and does not elaborate further. 31 Bracton on the jurors in novel disseisin removes the direct reference to the ecclesiastical courts, saying that "they may be kept from taking the oath for the same reasons that witnesses are kept from giving testimony"; he goes on to list some of the reasons, drawing mainly though not exclusively on Roman-canon procedural texts. 32 If the grounds for objecting to judges had not been specified in the canon law procedural sources, or if they had been the same as the grounds for objecting to witnesses, these references would be neutral. In fact, the grounds for recusation of a judge are specified in contemporary Roman-canon procedural literature, and they are not the same as those for objecting to witnesses, though there is some overlap. 33 If the authors of Glanvill and Bracton had thought that jurors were, as a matter of legal doctrine, lay judges, there is no reason for them not to have used the grounds of recusation of a judge. To this extent, therefore, jurors were clearly thought of in law as witnesses.	20
Similar reasoning applies to the procedure for attacking a verdict by way of attaint, invented for novel disseisin in 1201 and extended to juries in real actions generally in 1275. 34 The attaint is unmistakeably a proceeding to convict the trial jury of perjury, and the prior form of proceeding in relation to the grand assize is explicitly described as such by Glanvill. 35 Here, if the jury were considered as lay judges, the obvious approach would be that proceedings for false judgment would lie against them. This possibility appears to be considered obliquely in Bracton's treatment of the attaint, but is not followed up. 36 If they were judges of fact and the canon law proceduralists were to be followed, some sort of procedure by way of appeal would be appropriate. 37 Jurors share with witnesses, but not with judges, perjury prosecution as a primary mode of attacking what they say.	21
(b) Seeking informed jurors: Vicinage, venue, and special types of jury. The jury writs in Glanvill characteristically required that the jurors should come from the locality of the alleged facts, identified by vill (village or township). 38 By the later thirteenth century, this requirement was causing difficulty in impaneling juries, and it was progressively diluted and finally abolished by statutory intervention. 39 The rules as to venue subserved the vicinage rule, requiring the issuable facts alleged in pleadings to be laid in some specific place; they continued to be important after the vicinage rule had been substantially diluted. 40 The rules persisted in spite of significant inconveniences, as, for example, that in the fifteenth century the requirements of venue meant that events overseas were nontriable. 41	22
If the venue rules had meant that trial took place in the venue, this would be consistent with a rule of convenience for the attendance at trial of witnesses to give evidence to the jury. But trial could perfectly well take place in Westminster. If the venue rules had directed trial to the place of residence of the plaintiff or defendant, this would be consistent with the jurors as lay judges, the idea being that I ought to be judged by locals, comparable with the idea of trial by peers. But the venue was determined by the facts in issue. This gave rise to problems where, for example, a contract was alleged to be made in one county and broken in another. 42 The rules are therefore seeking informed jurors—even where, as in the contract cases, informed jurors are fairly unlikely.	23
This is corroborated by the various uses of special panels of "experts," summoned by venire facias, found episodically through the later Middle Ages; these range from foreign merchants to cooks, court officials, and surgeons. 43 A systematic instance was the writ de ventre inspiciendo under which a jury of matrons was summoned to determine whether a woman was genuinely pregnant. 44 The uses of special panels are evidence, along with the vicinage and venue rules, of a deep-rooted idea in the doctrine that jurors should be drawn from people who at least in some token sense had knowledge of the facts in issue.	24
(c) The verdict as a proof: Verdicts and other proofs. The Statute of Westminster I, 1275, in two places provides that matters may be averred by jury "or as the king's court may decide." 45 The jury was one among a range of possible proofs that could be tendered to support a pleading. Both this fact, and the alternative proofs and their scope, tell us something about later medieval doctrine about the jury. The fact that there were alternative proofs points to the verdict being a type of proof rather than a type of judgment; some of the rules about the boundaries between other proofs and verdicts tell us something about the perceived nature and scope of the verdict as a type of proof.	25
(i) Proof by documents. The use of documents (at least royal documents) was already a normal mode of proof in Glanvill, where jury trial was still exceptional, 46 and it remained doctrinally separate through the Middle Ages. A royal charter could be used to prevent the assize proceeding, that is, to oust the jury, in some petty assizes from an early date. 47 In the later medieval law, if documents were to be relied upon, they were supposed to be pleaded. They fell into two classes. Records were those documents of which a central copy was kept enrolled—the decisions of the royal courts, final concords, letters patent, and so forth. Here disputes as to the existence or terms of the document would be settled conclusively by calling up the original from the central record. 48 Deeds were private documents under seal (and with other formal requirements). Deeds could be proved to be forged or ineffective by inspection in court, without jury trial of the issue. According to Glanvill and Bracton they could also be proved to be genuine by comparison of seals on inspection in court. 49 The production of deeds thus did not automatically lead to jury trial, though it could. 50 In fact, the adverse party was not generally allowed to plead that a pleaded record or deed inaccurately stated the terms of the transaction it purported to record, or (for example) that he had paid money stated on the face of the deed to be owing, unless he could tender another deed. The reason given is that the writing ousts "parol averments" to the jury. 51 The characterization of such averments as "parol" is a clear indicator that the line drawn is between writing on the one hand, and oral testimony, that is, the verdict, on the other.	26
In practice in land litigation parties tended not to plead their documents, but to go to the general issue and then lead the documents in evidence to the jury. In the fourteenth and fifteenth century this gave rise to a problem if the jury returned a special verdict that was dependent on a disputed document. As long as the jury found a general verdict, the court had no official notice that its decision was not based on the knowledge of the jurors. When a special verdict was found, however, and the jury recited a document, it was arguable that the document ought to have been formally pleaded or at least produced in evidence, and that to find its terms was beyond the competence of the jury, so that the special verdict was bad. There are contradictory authorities on the point in the mid-fourteenth century; there is an early fifteenth-century decision that the verdict is bad, but Littleton, in the middle of that century, says that it is good. 52 These arguments point to juries being expected (and competent) to find a special type of fact: facts that are, or are capable of being, public knowledge in a locality.	27
(ii) Suit followed by wager of law. Witnesses produced by the party, who did not wage battle, began in the time of Glanvill to be described as secta or sequela, usually translated as suit. 53 Suit continued to be used as a form of proof in the thirteenth century outside the contexts in which wager of battle or a jury verdict was required, but it was seen as a weak type of witness evidence; 54 exceptionally in cases of personal status, "suit of kin," or the production of relatives was the required and normally conclusive form of proof. 55 Suit other than suit of kin could be defeated by a successful wager of law or compurgation, in which the proof was the defendant's oath with the support of eleven oath-helpers. In the early fourteenth century, suit became a mere matter of form, and by the fifteenth compurgation had become in effect (whatever it had been before) a decisory oath by the individual party, through the use of hired oath-helpers. 56	28
In the fourteenth century, there are identifiable conceptual boundaries between jury trial and compurgation in the contractual context. In the action of account, compurgation is denied where the liability depends on receipt from a third party. 57 The boundary here, also found in some debt cases, is between, on the one hand, transactions to which in principle the only witnesses are the parties, and on the other, transactions to which there may be other witnesses. The jury was appropriate where the facts were "within the knowledge of the country." 58 Wager of law is only appropriate to secret transactions.	29
When we go back to the early thirteenth century, any conceptual distinction is less clear because compurgation was more generally available. On this basis Maitland argued that the survival of compurgation in the old contractual actions was due merely to the antiquity of these forms of action. 59 But early fourteenth-century lawyers were capable of reshaping the rules of proof—as they did by reducing suit to a formality and by imposing the requirement of a deed in covenant—so that it is prima facie to be supposed that they allowed compurgation to survive in debt and detinue because they did not think jury trial appropriate to secret transactions. 60 In any case, the relationship between suit/ compurgation and trial by jury in these arguments is between types of proof, not between a type of proof and a type of adjudication.	30
(iii) Other modes of proof. The remaining alternative modes of proof are inspection by the court, trial "per proves," and variants on proof by witnesses. Inspection by the court as a proof is attested in two areas: the inspection of a party to determine age, and the inspection of wounds to determine whether they amounted to maims. 61 Allegations that a specific person was alive or dead fell in several cases to be tried "per proves" rather than by jury. 62 "Proves" here may mean witnesses, but could have let in anything that could be used to prove life or death in the canon law. 63	31
Minor officials and persons involved in procedural acts such as summons or distress could be joined to the jury or separately examined to prove facts in their knowledge. 64 Glanvill tells us that on an issue of family relationship in the writ of right, proof is by judicial summons of the parentela, the common blood relatives, the practice that developed into "suit of kin"; Bracton tells us that where the husband's right to curtesy is disputed and the issue is on the live birth of a child, this is to be proved by suit (meaning here witnesses) who heard the child cry. These bring us back to the thirteenth-century use of suit as witnesses who could be examined by the court as an optional mode of proof (which Maitland thought might, but for the entrenched position of jury trial, have developed into proof by witnesses along continental lines). 65	32
These proof variants illustrate the fact that jury trial is conceived as one among several modes of proof. Not even in the possessory assizes (where summoning the assize was specified in the original writ) does the jury wholly oust other methods of proof, whether by documents or "per proves." The availability of alternative modes of proof depends not on forms of action, but on factual issues arising on the pleadings.	33
There is, therefore, a considerable body of later medieval doctrinal evidence, starting with Glanvill, which shows a view of jurors as witnesses, not lay judges, and the verdict as a proof of a specific kind available for facts of a specific type. If the jury was unique in these features, we could attribute them to an original development of legal thought in England. However, there are several instances of "testimonial" lay fact finding by the production of panels of locals, restricted to specific types of disputed issue, from elsewhere in Europe in the twelfth century and later Middle Ages.	34
Besides the obvious Norman case, 66 the probatio per turba, French preuve par tourbe, was used to establish matters of custom in canon law and in France down to the early modern period; 67 inquisitions were used to fix boundaries in eleventh-century Norman Sicily; 68 in eleventh- to fourteenth-century Spain the pesquisa was used in cases of title to land and for some criminal cases; 69 in southern Germany in the fourteenth to fifteeenth century the Kundschaft was used in land litigation; 70 in thirteenth- and fourteenth-century Hungary the inquisitio was used in land, status, and fiscal cases, but only as a mechanism to allocate the duty of compurgation; 71 and in fourteenth-century Serbia some variants of the porota represented the use of local panels in boundary cases and to assess damage in cattle trespass. 72 Possible additions are early twelfth-century Aquitaine and Tuscany. 73 These cases are, like the early medieval "inquests" found by the Settlement of Disputes authors, wider than Brunner's narrowly fiscal conception. But equally, like the testimonial conception of the jury, they are not generally available modes of proof, but are restricted to certain types of factual issue. To borrow terminology from the later common law, the issues in question are generally, though not entirely, "local," rather than "transitory": 74 they concern persistent facts of which there may reasonably be expected to be a local reputation, like local custom, boundaries, possession, and personal status.	35
Now it is possible that all these are just instances of judicial pragmatism and hence convergent evolution of legal techniques—which is, in effect, Reynolds's view. This argument is, however, less attractive than it at first appears. A real instance of convergent legal evolution is discussed below, in the area of conveyancing techniques. Frankish law offers us a description of a form of conveyance of property by public ceremonious transfer before witnesses, often involving a symbolic object. 75 The function of this form of conveyance is much the same as the Roman mancipatio, 76 so that this is an instance of convergent legal evolution. We can see from this case that convergence in function can produce fairly radical differences in form, while what we see with jury-like modes of proof is closer to being a common form. It is therefore worth considering the possibility that the use of panels of locals for certain types of proof has a basis in late Roman and/or early medieval normative sources, before resorting to the hypothesis of convergent evolution.	36
II. Local Reputation in Early Medieval Law and Anglo-Norman Practice
1. Vicini in Late Roman and Early Medieval Normative Sources
The use of vicini, neighbors, as a special source of testimony, is primarily found in early medieval sources in the related contexts of conveyancing and boundaries. Besides these two are a number of minor uses.	37
(a) Conveyancing. The starting point in the late Roman sources for an early medieval concept of vicini as a special type of witnesses is two laws of Constantine the Great, which altered the law in relation to conveyances of land, so as to require a ceremonious transfer of the land executed on the property itself and in the presence of vicini (corporalis traditio). This was a change to the previous law, which allowed conveyance in the case of Italic land by ceremony of mancipatio, or generally by documents, executed away from the land itself. 77 The changes were motivated by concerns about tax liability and double conveyancing; 78 but their effect was inter alia to make the neighbors a required class of witnesses in relation to title. Thus, in the law on conveyance on sale,	38
... neminem debere ad venditionem rei cuiuslibet adfectare et accedere, nisi eo tempore, quo inter venditorem et emptorem contractus solemniter explicatur, certa et vera proprietas vicinis praesentibus demonstretur.... (No one may contract for or execute a sale of any property unless at the time of the formal conveyance | a certain and true property is shown in the presence of the neighbors.) 79
The conveyancing context, as Honoré explains, is that before these laws on a conveyance by traditio a description of the boundaries of the land by view or perambulation was theoretically required and the use of neighbors as witnesses was normal, in order to avoid or at least identify boundary disputes; however, conveyance by traditio had become "constructive" so that documentary conveyances could be used; and the use of neighbors was not required. 80 Under the new laws, the presence of the neighbors was required and, as G. G. Archi has pointed out, they become an added source of information about title. 81	39
Both laws were incorporated in the Code of Theodosius II in 438. 82 In the Code of Justinian, however, the law on sales disappears, and that on gifts loses the corporalis traditio provision. 83 It is possible that this is because the use of vicini is displaced (though still existing as a subsidiary mode) by the registration of transfers in local public records. This can be seen from a law of the Emperor Zeno dating from 478 in the Code of Justinian:	40
In donationibus, quae actis insinuantur, non esse necessarium iudicamus vicinos vel alios testes adhibere; nam superfluum est privatum testimonium, cum monumenta publica sufficiant. (In gifts, which have been inserted in the records, we do not judge it necessary to call the neighbors or other witnesses; for private testimony is superfluous, as the public records are sufficient.) 84
A variant form of the law on gifts was included in the Edictum Theoderici (which has been attributed either to the Visigoth Theoderic II [453-66] or to the Ostrogoth Theodoric the Great [489-526]) 85 and that on sales in the Breviary, dated to 506, of the Visigoth Alaric, where it received an interpretatio or explanation:
... quia iubetur, ut vicini rei quae venditur testes esse debeant et praesentes, in tantum, ut etiam de mediocribus rebus si quid in usum venditur, ostendi vicinis placeat, et sic conpareri, ne aliena vendantur. (... because it is ordered that the neighbors of the property which is sold must be witnesses and present, to the extent that even in the case of things of slight value, if anything is sold for use, it is our pleasure that it be shown to the neighbors and thus purchased, in order that the property of others may not be sold.) 86
The Breviary was extensively circulated in early France, and more widely in the form of epitomes. Haenel's edition reprints five of these epitomes, and in three of these, the Epitome Guelphyterbiani, the Epitome Monachi, and the Epitome of St. Gall, the requirement that sales be in the presence of the neighbors is present. 87 The same provision is found in the eighth-century Swiss Lex Romana Raetica. 88
The Burgundian Code of Gundobad, which is roughly contemporary with the Breviary, contains a similar requirement that sales of land must be in writing and attested by seven or five witnesses "loci illius consistentibus" (dwelling in that place) or at least three witnesses "loci illius consistentibus" who are of good reputation. 89 This requirement is not, however, present in the contemporary Lex Romana Burgundionum, which refers only to traditio and following possession. 90 The late fifth-century Visigothic Code of Euric requires either a document, or witnesses to the price, following a different provision of the Theodosian Code, and this approach is followed in the seventh-century Spanish Visigothic code. 91 Considerably later, the late eleventh- or early twelfth-century Expositio ad Libram Papiensem (Commentary on the Book of Pavia) says that a traditio must be on the land and in the presence of witnesses "sicut in lege Romanorum precipitur" (as is required in the law of the Romans), presumably referring back to the requirement of the Breviary. 92	41
The Salic Laws are uninformative on transfers of property; 93 the early seventh-century Frankish Lex Ribuaria, however, does give a definite mode of conveyancing using witnesses that is clearly not dependent on the use of vicini. The key is a ceremony of traditio and receipt of price, and the witnesses are to be children and to be slapped and have their ears pulled to fix the event in their memory. 94 This model of a ceremony off the land is also followed in the Leges Alamannorum and Lex Baiwariorum from the same period. 95 There is a fairly substantial reason to suppose that this concept had wider application, as it also appears in the eleventh-century Norman conveyancing practice studied by Emily Zack Tabuteau. The ceremony did not have to be on the land, and the witnesses do not seem to have been required to be vicini, as some of them were potential claimants selected to create, as it were, estoppels against themselves. 96	42
We can therefore see two broad early medieval conveyancing models. The first is sub-Roman and requires a ceremony on the land with neighbors to witness it. The second, the ceremony of transfer of a symbolic object off the land, with child and quasi-grantor witnesses, has a function analogous to the Roman mancipatio but is clearly unrelated to any Roman source. When we come to consider English practice, there is a striking movement between the two models at the time of the regularization of jury trial of land title under Henry II. The work of S. E. Thorne, M. T. Clanchy and, most recently, John Hudson has shown that much Anglo-Norman conveyancing practice was, like that of eleventh-century Normandy, based on the model of ceremonious transfer not executed on the land itself. 97 But, as Thorne showed in 1936, this practice was displaced in the later twelfth century by a requirement of a ceremony on the land and the execution of acts of possession by the transferee. 98 This is a striking coincidence if it is not more; Thorne drew the inference that the movement to a ceremony on the land was necessitated by jury trial. 99	43
(b) Boundaries. Probably linked to the conveyancing rule is a rule about boundaries. Honoré explains that the use of vicini before the legislation of Constantine was related to boundary issues. Laws of Constantine and later on the establishment of boundaries used expert surveyors, the agrimensores. 100 This was, however, of little relevance under early medieval conditions, and several normative texts required the use of locals. Thus, the Visigothic Code of Euric provides in the first place that boundaries that have been wrongfully altered are to be restored in the presence of the neighbors; 101 and, second, that where there are no clear boundaries,	44
... tunc Gothii [in-]/grediantur in loco hospitum et ducan[t, ubi]/ terminum fuerit ostensus. Tunc iud[ex], /quos certiores agnoverit, faciat eos s[a]/cramenta praebere, quod terminum s[ine]/ ulla fraude monstraverint.... (... then let the Goths enter into the place of the hosts and follow the boundary where it is shown. Then let the judge cause those whom he knows to be well informed to give oaths that they have shown the boundary without any fraud.) 102
In the main Visigothic codification, this method is reproduced in a slightly different form:
Cum autem proprietas fundi nullis certissimis signis aut limitibus probatur, quid debeat observari, eligat inspectio iudicantium, quos partium consensus elegerit, ita ut iudex, quos certiores agnoverit vel seniores, faciat eos sacramenta prebere, quod terminos sine ulla fraude monstraverint. (When the ownership of the land is not proved by any certain signs and limits, which ought to be observed, [the judge] may choose an inspection by arbiters, chosen by agreement of the parties, so that the judge should make those whom he knows to be the better informed, or elder, take oath [etc.].) 103
Similarly, the later Lombard kings made use of an inquisitio procedure in boundary cases. 104 In contrast, in the Alaman and Bavarian laws the procedure for determining boundary disputes involves trial by battle. 105
The use of local "seniores" to establish boundaries is also found in canon law sources. In the Irish collection of canons, the Hibernensis, it appears as:	45
De duabus ecclesiis contendentibus agrum unum. Synodus: Ager inquiratur in scriptione duarum ecclesiarum; si in scriptione non inveniatur, requiratur a senioribus et propinquis quantum temporis fuit cum altera et si sub jubileo certo mansit, sine vituperatione maneat in aeternum.... (About two churches disputing a piece of land. An Irish synod: the land should be looked up in the written evidences of the two churches; if it is not found there, inquiry should be made of elders neighboring [on the property] how long it has been held by the present holder; if it has remained unchallenged for longer than the period of legal warranty, let it remain [with the holder] for ever.... ) 106
At the Synod of Toul in 838 a dispute as to the boundaries of parishes was settled by "investigantes testimonio multorum, tam clericorum quam laicorum, qui antiquiores videbantur" (investigating by the testimony of many, clergy and lay, who appeared to be the older), who set out the parish bounds on oath. 107 Burchard of Worms attributes to one of the Synods of Aachen a diluted version that mixes this procedure with the use of the judgment of God, as in the Alaman and Bavarian laws:
... si contentio fuerit de terminatione duarum matricum, plebes utrarumque discernat; si non convenerit, lis Dei judicio discernatur. ( ... if there was dispute as to the boundaries of two mother churches, let the people of both decide it; if this is not possible, let the suit be decided by the judgment of God.) 108
Ivo of Chartres eliminates the use of the judgment of God, but Gratian retains it. 109 As will be seen below, the use of locals to prove boundaries was incorporated in the learned laws of the twelfth and thirteenth centuries.
(c) The assessment of damages to land. The use of vicini to assess damage to land in cases of cattle trespass can be found in the Visigothic laws 110 and, following them, the Bavarian laws, 111 and in the mid-seventh-century Lombard Edictus Rothari. 112 The Edictus Rothari also prescribes that neighbors are to estimate damages in the case of civil liability for fire. 113	46
(d) Questions of status. Early canon law sources make passing reference to the use of vicini or similar phrases in relation to two questions of personal status: proof that an individual was baptized, and proof of family relationship for the purposes of the prohibition of marriage within the prohibited degrees. The first is found in a letter of Pope Leo I (440-461): "Si nulla exstant indicia inter propinquos et familiares, nulla inter clericos aut vicinos quibus, hi de quibus quaeritur baptizati fuisse doceantur.... " (If there is no evidence among relatives and household members, none among clergy or any neighbors, from which it may be inferred that the subject of the enquiry has been baptized.... ) This text appears in the same terms in numerous canon law collections down to and including Gratian. 114	47
The case of family relationship is slightly more complex and more directly relevant. The acts of the Synod of Friuli (796/97) contain a provision requiring espousal and a delay before marriage, to allow enquiries from "vicini vel majores natu loci illius, qui possint scire lineam generationum utrorumque" (the neighbors and elders of the place, who may be able to know the ancestry of both of them). 115 Burchard of Worms attributes to St. Fabian (pope 236-251) the following:	48
Consanguineos extraneorum nullus accuset, vel consanguinitatem in synodo computet: sed propinqui, ad quorum notitiam pertinet, ... Si autem progenies tota defecerit, ab antiquioribus, et veracioribus, quibus eadem propinquitas nota sint, episcopus canonice perquirat.... (No unrelated person should accuse [married persons] of consanguinity, or compute consanguinity in synod: but the relatives, to whose notice it belongs ... If, however, the whole family stock is unavailable, the bishop should carry on a canonical inquiry among the older and more reliable who may have notice of the relationship.... ) 116
He also gives the form of the oath to be tendered in such a consanguinity inquiry, which requires the witnesses to tell on interrogation what they know or have heard from their neighbors or ancestors about the relationship: "quidquid inde scis, aut audisti a tuis vicinis, aut a tuis antiquioribus propinquis." 117
This body of law was collected by Gratian, 118 and the basic ideas are followed by Glanvill in discussing proof of family relationships. 119 The parentela, the members of a common lineage, are the preferred witnesses, but if they fail to produce an answer, then resort is had to the vicini; in either case, hearsay is acceptable.	49
The uses of vicini as a source of evidence discussed so far are fairly straightforward. The late Roman law of land transfer and, with it, the use of vicini to determine boundary issues, evidently survived in southern European normative materials, which is unsurprising as this was the area of greatest Roman legal influence in this period; 120 and the use of vicini in boundary cases passed into the canon law, which is equally unsurprising as the church was said to live by Roman law. 121 The derivation of the damages and status rules is less clear, but the concept is evidently the same: local reputation is an appropriate source of evidence for both local values and family relationships (though in the latter case the evidence of the family is preferred).	50
The Carolingian inquisitio capitularies, which are by no means limited to matters of royal interest, 122 should probably be understood in the light of this background, especially as they date after Charlemagne's coronation as emperor. Indeed, most date from the reign of Louis the Pious (814-29), an emperor notoriously influenced by the Church. 123 In addition, instances of the application of the procedure seem to be predominantly southern; the "southern" or "ecclesiastical" character of the other normative sources for the use of vicini therefore seems to persist in the inquisitio. 124 A link to land and status matters is apparent in a capitulary of 816:	51
Testes vero de qualibet re non aliunde nisi de ipso comitatu in quo res unde causa agitur posita est congregentur; quia non est credibile, ut de statu hominis vel de possessione eius per alios melius cognosci rei veritas possit quam per illos qui vicini sunt. (However, witnesses in any matter should not be assembled except from the county, in which the matter that is the subject of the case is alleged; because it is not credible, that in relation to the status of a man, or someone's possession, the truth of the matter can be better known than through those who are neighbors.) 125
In all these areas, of course, the evidence is too limited to show consistency of practice. What it shows is a normative tradition that was certainly available to judges, disputants, and writers about law in the twelfth century.
2. Panels of Locals and Other Forms of Collective Testimony in Anglo-Norman Judicial Practice
The early medieval normative sources discussed above imply that there may have been normative claims that parties could make in litigation about appropriate forms of proof. 126 This, in turn, would create patterns in the evidence of use: we would expect panels of neighbors, as distinct from party witnesses, compurgation, and direct determination by the court, to be commonly used in boundaries and status cases, and possibly in other areas of "continuing" facts in which local reputation was analogously probative. We would not expect to find panels of vicini used in relation to transitory facts like crimes and torts. In relation to title to land made under a conveyance, the normal mode of conveyance is also significant. Only the mode of conveyance by ceremony on the land would imply production of vicini in later litigation. But private ceremony off the land would imply production of charters or attesting witnesses, and conveyance by act in court would imply an appeal to the memory of the court itself.	52
How far does this hypothesis fit the evidence for collective testimony in Anglo-Norman England? R. C. van Caenegem's collection of lawsuits between 1066 and 1199 provides a convenient body of material. 127 It is a limited sample, from which any conclusions are impressionistic. There is, however, enough material to get some idea of the relative frequency of the use of different proof modes for different types of issue.	53
I have approached this data in two ways. The first looks at the nature of the disputed issues in those cases van Caenegem identifies as involving neutral collective testimony of some sort (inquests, recognitions, and so on). 128 It appears that there were two distinct modes of neutral collective testimony. One is the use of a special panel of locals; the other is the testimony of a local or feudal court. When we analyze the disputed issues, it becomes clear that there were also differences in the type of subject matter dealt with by the two modes. Table 1 shows the breakdown of disputed issues numerically and as a percentage of the total issues for that proof mode. 129	54
Table 1. Cases Indexed as "Inquest," etc.; Subjects     a This category includes claims characterized both as claims "to" churches and as claims to advowsons. In one of the earlier cases, no. 296 (1138 x 1139), the issue is characterized as lay fee or free alms, but the substance of the dispute is as to the possession of the church and its lands. b This category includes claims to all sorts of consuetudines, including rent, tolls, monopolies, jurisdictions and immunities, rights of common, rights to water, other easements, etc., as well as custom in the narrower sense of local norms. c This category includes all cases where a claim is said to be to land and not identifiably a dispute about boundary disputes or consuetudines.
The figures for cases up to 1164 tell us first that forms of neutral collective testimony were at this period used for practical purposes entirely to deal with local, as opposed to transitory, matters. The one apparent exception, a case of disseisin of a ship, is, in fact, part of a wider dispute about customs. 130 The nature of the sources inevitably means that the collection is heavily biased toward land disputes where one or more of the parties was an ecclesastical body. 131 This probably hides the use of locals to present crime 132 and may well hide other uses. However, in the case of trial by battle, van Caenegem's index gives ten cases before 1154, nine concerning land and one treason, and thirty-two cases between 1154 and 1198, seventeen concerning land, fourteen crime, and one an unidentified subject. We may therefore expect that if there was wide use of panels of locals to determine questions of fact other than those seen, the collection would provide some evidence of it.	55
Second, before 1164, special panels appear particularly likely to be used to decide boundary issues and questions of customs. The figures may somewhat overstate the extent to which at this period special panels were used to determine general land claims and claims to churches. 133 These features of the use of special panels are broadly consistent with the early medieval rule about boundaries and with its apparent conceptual basis: special panels were used primarily to determine boundaries and other questions, such as customs, that lie in local reputation.	56
The second angle of approach is through the cases that appear to involve claims to land, but are not identifiable as concerning boundaries or customs. Here the 111 Domesday cases, Lawsuits, nos. 21-131, form a distinct group. 134 Table 2 shows the analysis.	57
Table 2. Land Claims; Proof Modes     a There is one case, no. 408, the Anstey case (1158-63), of reference of an issue of bastardy to the ecclesiastical courts; since this is unique in the sample it is not tabulated.
These figures corroborate the impression from the "inquest, etc." cases analysed in Table 1. They show that before 1164 proof by a local court was a significant mode of proof in general land cases, significantly greater than the use of panels. In the Domesday cases it is absolutely dominant. 135 Some of these cases also give us some idea of what it is that the local court speaks to in general land cases. This seems to be a ceremony of "seising" by act in court, either by the authority of a writ, or by some representative. Thus, for example, in no. 45 one party is said to hold land unjustly "nemo enim eorum vidit brevem regis vel ex parte ejus hominem qui eum inde saisisset" (for none of them has seen a royal writ, nor on his side a man who might have given him seisin); in no. 171 the Riding say that "nunquam viderunt antecessorem episcopi fuisse saisitum neque per brevem neque per legatum" (they never saw that the predecessor of the Bishop was seised, either by writ or by messenger). 136	58
The claims to churches produce an analogous picture (see Table 3). 137 Here a relatively even distribution of proof modes in the earlier cases is replaced by an absolute dominance of the use of panels. The shift may be sharper, if slightly earlier, as three of the cases of the use of panels before 1164 are in the early part of the reign of Henry II. 138	59
Table 3. Claims to Churches and Their Advowsons