Bringing the Feudal Law Back Home: Social Practice and the Law of Fiefs in Italy and Provence (1100-1250)

Attilio Stella. Journal of Medieval History. Volume 46, Issue 4, 2020.

The Problem: A Disjuncture Between Practice and Legal Literature?

What else? I gave them horses, garments, arms, infinite money, gold and silver plate. When there was nothing left to give, they seduced my wife with great shrewdness, brought her to Canossa, and kept her and the castle by force (Marquis Albert I of Verona, c. 1132-5)

Around the year 1125, when the followers of the House of Canossa elected Albert of S. Bonifacio, marquess of Verona, as their lord after Matilda of Tuscany’s death, they could hardly ignore the fact that he was not a paragon of loyalty. He is indeed known to have pursued his own interests at any time he sensed the opportunity. Although he had repeatedly supported the Canossa against the Holy Roman Emperor during the Investiture Controversy, Albert betrayed them at least twice. Once he had been chosen as Matilda’s successor, he swore allegiance to Emperor Lothair III; soon after, however, he broke his oath and secretly backed the emperor’s rival, Conrad Hohenstaufen. Nonetheless, when he had to face an uprising by the former Canossa followers (‘capitanei, valvasores et cuncti satellites domus comitissae Mathildis’), he did not think twice before crawling back to Lothair. In his letter to Lothair—a display of opportunism and realpolitik—Albert strategically depicted the rebellion as a disgraceful act of treason and himself as a good, generous lord, who had humbly and almost reluctantly accepted his leading role. These greedy noblemen, he wrote, had consumed all his wealth by reclaiming gifts of all kinds and, when he had nothing left to give, they kidnapped his wife and occupied Canossa. When Albert seemed to have recovered the loyalty of these followers, one of them, Rainerius de Saxo, who apparently had held Canossa as a fief, retook the castle in an act of great treachery (‘cum magna prodicione’) and imprisoned Albert’s two sons, causing incalculable damage to his honour.

The picture provided by the noblemen in their response to Lothair presents a wholly different state of affairs. Their betrayal was now described as a legitimate reaction against a greedy and treacherous lord: Albert, they wrote, had conspired to hand Canossa to Conrad Hohenstaufen, who had usurped the royal title (‘invasor nominis regis’), in return for an immense amount of money. The kidnapping of Albert’s wife was now depicted as her voluntary call for help to prevent Albert’s betrayal. The noblemen complained that after Albert persuaded them to return the castle, he ceased to behave as a lord and retired to Verona, and they never saw him again. His two sons, whom he left in Canossa and who were now described as illegitimate, committed a series of offences, not least the imprisonment of an innocent nobleman, the son of Rainerius de Saxo, who in return took the castle and incarcerated the two felons.

The subtext of these letters helps us understand how the bonds tying some of the great nobility in early twelfth-century Italy were not subject to fixed rules but, on the contrary, needed to be renewed through a constant exchange of gifts and counter-gifts—sometimes in the form of expected behaviours or services. Interestingly, these letters were inserted in a Tractatus de dictamine composed in Lombardy about a decade later, around 1135, at the same time and place as the first tracts of the earliest known collection of feudal law, the Libri feudorum [LF], were written. If one is to believe this law book, treason was itself sufficient cause to lose a fief (LF 1.16); yet not only did the parties refrain from invoking any specific penalty, but they seemed much more focused on depicting themselves, or feigning so to depict themselves, as keen to compromise: Albert wanted the restitution of the castle and to win back his followers’ loyalty; the followers declared themselves willing to obey him only if he behaved as a good lord. The emperor, in an attempt to keep peace in the kingdom and secure military aid for his forthcoming campaign, promised to judge the case in person and appraise whether Rainerius had held Canossa as a fief or as castle-guard (‘pro feudo an pro custodia’)—an important distinction, since, according to LF 1.2, a castle-guard fief (feudum guardie) was not a proper fief because it could be recovered by the lord at the end of the agreed term. However, it would be futile to look for further correlations between the strategies deployed by these parties and other norms contained in the LF. On the contrary, informal norms, status, timely violence and rhetorical skills contributed to shaping such negotiations and the very nature of political relationships. In other words, they did not necessarily conform to written law or any other set of objectified rules. Equally, the LF and later literature on fiefs are utterly silent on all such matters, which so often remained implicit, untold, even though in practice they sustained this kind of relationship.

The main idea underlying this article is that comparative analysis of the sources recounting real practice—including archival sources—and the rules collected and ‘codified’ in law books and legal literature can help us bridge their disjuncture. The argument challenges the idea that legal literature, including the LF, was a product of a scholarly milieu fundamentally unconcerned with actual social developments. In other words, it challenges a notion that informs most of the debates that have followed Susan Reynolds’ deconstruction of modern models of feudalism. One of the main achievements of Reynolds’ works is the discovery of the influence of the LF on such models. The history of the book, indeed, would seem to speak in her favour. The first known version of the LF, conventionally called antiqua, appeared in Lombardy shortly after 1150 as an assemblage of seven earlier tracts. Five of them are dated to c.1100-30, while the last two were written c.1150 by the Milanese judge and politician Obertus de Orto. All these tracts related to an ordinance issued in 1037 by which Emperor Conrad II secured the possessions which the Milanese military elite held of the archbishop of Milan. The authors of the LF interpreted this decree in light of local customs and provided partial, occasionally inconsistent, insights into the substantive and procedural law to be followed in disputes over fiefs. They implicitly deemed fiefs to be property held by the military class. After some decades of textual sedimentation and augmentation, mostly taking place in Milan, a second phase of codification concluded by c.1180. At this stage, the LF drew the attention of lawyers trained in Bologna: the first was Pillius de Medicina, who produced a glossae apparatus and a brief commentary on the code. In the early thirteenth century, other jurists produced commentaries and recompilations of the LF. The tradition commented upon by Pillius was supplemented with new texts, including legislation of Frederick I and Lothair III, and was accepted as a standard version by Accursius, the most influential law professor of the time. About 1250, Accursius completed the apparatus of Pillius’ gloss and had the text copied in the new editions of the Corpus iuris civilis produced in Bologna. Therefore, a text originally anchored in the local custom of Lombardy now had the potential to reach law schools across Europe.

As Reynolds suggested, the generalisation of a ‘feudal’ vocabulary was made possible by the influence exerted by this text on lawyers. The spread of properties or tenures called fiefs would be the outcome of imposing the learned terms inspired by the LF on different customary realities. Her book, Fiefs and Vassals, however, tends to imply such influence rather than analysing the way it developed. In what follows, therefore, this article assesses the relationship between the LF and the literature associated with it on the one hand, and the sphere of practice and of ‘customary’ norms on the other. To do so, it attempts to set aside traditional narratives depicting the rise of the ius commune as an intellectual construct developing in contrast with an allegedly more natural customary law. Rather, it is necessary to escape the narrow field of property law and envisage the norms of fief-giving not as the execution of formalised rules, but in terms of the use of cultural models by interested actors. In this way, the LF and the literature associated with it may be reinterpreted as products of specific social and political contexts. This operation, of course, entails several risks. It should not assume an a priori divide between legal science on the one hand and the values and customary norms embedded in the social fabric on the other. Neither ought it to imply the existence of shared social practice whenever codified norms emerge nor expect ‘the highly ambiguous vocabulary of rules … to express a social practice that in fact obeys quite different principles’. One last risk concerns terminology: to avoid possible confusion, the term custom is used here to refer to a source of law and legal decisions. The expressions customary rules and customary norms fit Max Weber’s definition of legal norms: regularities in social behaviour connected to sanctions enforceable by specialised staff. When the sources employ the terms consuetudo or usus with ambiguous meanings, however, their Latin forms are retained.

The article focuses on northern Italy and the French Midi, two areas whose role in the making of feudalism—whether a medieval reality or a learned construct—has been long acknowledged. The first section deals with twelfth-century Lombardy and describes how the earliest stages in the codification of the LF developed in close contact with shifting practices and power relationships—the decline of archiepiscopal power and the rise of the city commune—in which fief-giving seems to have been governed by norms that were often dependant on the relative status of the parties. The legal experts who wrote the LF were not external observers of this reality. Together with the noblemen who were at the centre of the exchange of fiefs, they were part of the ruling class, interested actors in the establishment of the municipal government, of its institutions and of its Romanising legal procedures.

The second section considers how the works of three lawyers were influenced, and to some extent even shaped, by different socio-institutional backgrounds, thus stressing the relationship between the sphere of practice and the making of learned law. Firstly, it examines the circumstances in which the Italian glossator Pillius de Medicina developed the notion of dominium utile, a legal device that distinguished the rights retained by the lord from those transmitted to the fief-holder, a theory that was of paramount importance in post-medieval representations of power. Secondly, it shows how practices of rural lordship contributed to shaping the notion of custom in the work of another Italian glossator, Iacobus de Ardizone. Finally, the example of Jean Blanc, a learned lawyer and a diplomat in Provence, highlights a particular, generally undervalued aspect of the inner logic of the ius commune, namely the abstraction of general rules from specific cases, their transmission and re-contextualisation. This last example connects directly to the third section, which puts in perspective the possible impact of the LF on learned law and practice before 1250.

Fiefs and the Libri Feudorum in Milan

This section appraises the relationship between the LF and the social context where the texts composing this book were produced. The five earliest tracts are by unknown authors—except for one, Ugo de Gambolado, a judge and consul from Pavia—and were all probably written between 1100 and the early 1130s. They all discuss, in the light of local practice, how a fief could be acquired, held, lost or transmitted to heirs, who ought to judge disputes about fiefs, and how. These texts were either written in Lombardy, perhaps Pavia, or in other milieux deeply influenced by the Lombard law school, which had flourished in the eleventh century. All their authors, indeed, show some interest in an old decree that was still, and would be later, deemed of great importance with regard to fiefs and which was included in the collection of imperial laws known as Lombarda. It was enacted in 1037 by Emperor Conrad II, who was faced with an uprising by the military clientele of the archbishop of Milan, the main political figure in Lombardy and ruler of the wealthiest and most powerful city in north Italy. Conrad satisfied the requests of the knights who held archiepiscopal land and who were complaining about the insecurity of their possessions. He pronounced that these possessions had to be heritable and that holders could be deprived of them only for wrongs proved before their equals. He eventually accorded to holders the right to appeal: from high noblemen to the emperor himself, from lesser men to the imperial missi.

When the authors of the LF wrote their commentaries on this decree, the political situation had radically changed. In the eleventh century, the Milanese military elite at the centre of this tumult was a restricted group of families divided into two composite categories, capitanei and valvasores, respectively greater and lesser knights. In the early twelfth century, some of their successors, namely those who had managed to build strong connections with the city, were regularly elected consules, the rotating office that is seen as the first expression of communal government. At that time, the archbishop, who was also the main landowner in Milanese territory, could not afford to lose their backing without facing potentially disastrous consequences. Between 1050 and 1150 five archbishops were deposed, one of whom was even imprisoned after opposing the anti-imperial policy enforced by the Milanese elite.

This consular elite, however, was quite heterogeneous in the early twelfth century. Alongside the military elite, it included an increasing number of legal experts (iudices), who drove the process of institutionalisation of the commune. The primary function of these officials was judicial: groups of consuls presided collectively over the city court and their judgements, usually on ‘rural land, and often feudal and signorial rights as well’, were pronounced by one consul, generally a judge, with the consent of the others. The legal experts who were regularly elected consuls, including some of the authors of the LF, were not, or not just, academic or professional lawyers. They were interested agents in the making of the city-state. Ugo de Gambolado, the author of one of our tracts, was in 1112 a judge and consul of Pavia. The Milanese judge Obertus de Orto (d. c.1171), who wrote the two last tracts of the antiqua version of the LF, was elected consul of Milan at least eight times between 1140 and 1171. The other two Milanese lawyers mentioned in texts attached to the LF in the late twelfth century, Stephanardus and Gerardus Cagapistus, were appointed consuls respectively five (1138-49) and 14 times (1141-1180). They were not just accomplished legal experts, but skilled politicians whom the commune entrusted with very delicate tasks and whose counsel was sought well beyond Milan. In 1154, Obertus and Gerardus were chosen to represent the city before Frederick I at the diet of Roncaglia and to negotiate with him on behalf of the commune.

Within this context, the exchange of fiefs was a significant means of sustaining the traditional political bonds between the military component of the urban elite and the archbishop, but now it served to a much larger extent to cement the clienteles of the military elite. Along with this development, the fief started to be seen no longer as remuneration for service but as property over which grantees of any social level had enforceable rights—a ius in re. In another way, however, fief-giving was still seen as a peculiar form of gift that followed some basic rules and triggered reciprocal expectations. This practice of gift-giving had already undergone a good deal of institutionalisation by the time the antiqua was composed. Some of its features are important for contextualising the making of the LF in Milan. There is, in the first place, evidence of fiefs held of the emperor: in 1140 Obertus himself judged a dispute about sovereignty over two Alpine communities that were claimed as an imperial fief (feudum) by both the counts of Seprio and the nobleman Locarnus de Besocio, of a powerful capitaneal family. The majority of the fiefs that emerge in charters, however, derived from older archiepiscopal grants. These grants concerned either shares of rural lordships or rights to collect tithes. Archbishops had enfeoffed most of these resources, at once both important politically and lucrative, mostly in 980-1035, to establish and sustain their political control over Milan. Enrica Salvatori has shown how this worked out for two capitaneal families, the de Porta Romana and de Porta Orientale, whose social distinction was mainly due to the support, also military, they had offered to the archbishop. The former’s fief consisted of lands and rights in several rural districts, but its core was the lordship over a castle located just outside the gates of Milan (Villamaggiore) which allowed the family to become part of the urban elite in the late eleventh century and fulfil a central role in the communal institutions. The de Porta Orientale family, whose fief consisted mostly of tithes in the Varese area, even though one of their members was elected archbishop (Arnulf III, 1093-7), was never involved in urban government. In both cases, however, fiefs had been treated as family assets, passed on to heirs, alienated to churches or sub-enfeoffed to followers, to the point that in the early twelfth century the two families only held residual shares of them.

That the same fief could be used at different social levels and hence trigger different expectations and obligations emerges clearly from three charters from the archive of Morimondo Abbey. The capitanei de Besate and de Setara—very influential in the city—held a fief from the archbishop which included land, tithes and most likely jurisdictional rights in the parish of Coronate, not far from the territory of Pavia. The charters attest to the acquisition of this fief by the monks of Morimondo partly from the capitanei themselves, partly from the people to whom these capitanei had sub-enfeoffed it. In the first charter, written in Pavia (9 January 1137), one Lanfranc and his two sons, for £13 (of Pavia), leased in perpetuity to one of the monks of Morimondo all the land and tithes they held from their lords (seniores) of the de Besate family. The same day, Wido de Besate renounced to the abbot portions of the fief that he had sub-enfeoffed to two other men. One year later, another Lanfranc, a capitaneus from the de Setara, a consul of Milan in 1145 and 1156, gave up a portion of the same fief, which he had sub-enfeoffed to a Milanese citizen. The fief that once had configured a political alliance between two noble families and the archbishop was now used to build these noblemen’s clienteles in a relationship overtly described as lordship. There is no direct evidence of how this specific lordship worked in practice, but studies of Lombard rural communities confirm how such grants contributed to the creation of hierarchy in rural societies, by according immunity from a lord’s jurisdiction.

Besides land, jurisdictional rights and the right to collect tithes, fiefs could also consist of pecuniary (de camera) or in-kind (de caneva) incomes—fiefs-rente and warehouse-fiefs. In 1131-3 the church of Velate, an area of the diocese of Varese politically subject to Milan, performed at least five transactions to obtain the waiver of a fief de camera of at least £2 10 solidi (of Milan) per annum—the original amount was certainly higher. The charters mention explicitly that it was an archiepiscopal fief-rente originally granted to the capitanei of the de Besocio family, who had divided it among heirs and sub-enfeoffed it to followers. The earliest warehouse fiefs appear in two different situations. The first is a beneficium de caneva (1116) consisting of 16 modia (about 2340 litres) of rye per annum held from the canons of Sant’Ambrogio by two Milanese citizens, probably capitanei of the Grassi family, advocati of that church. The second one is a grant of 1137 (‘investivit per feudum’) by Musso de Concorezo, a powerful Milanese citizen, of just 4 staria (about 73 litres) of rye and millet per annum, an amount that suggest this was a sub-enfeoffment of a larger fief. The charter specifies that the grantee, another inhabitant of Milan, could pass on the fief to both male and female heirs and that the fief was granted in return for his oath of fidelity (‘debeat … facere fidelitatem suprascripto Mussoni’).

These cases outline some of the main features of fief-giving in central Lombardy during the period when the first codification of the LF was made. Whatever their nature—lordships, lands or income in money or kind—fiefs were heritable resources; these grants were usually intended to extend the political patronage of the grantor, whether a man or a church, over the grantee and his descendants. They were exchanged, possibly sub-enfeoffed or sub-leased, at different social levels and thus they triggered different expectations, entitling the donor to services commensurate to the receiver’s status and honour—holders of high status were expected to provide political backing and occasionally military aid with their men; peasant elites were instead presumed to guarantee good estate management and serve as squires. Fiefs were conveyed through a public ritual of investiture, perhaps in a way not so different from other types of transactions, after which the recipient was generally required to take an oath of fidelity. These ‘feudal’ oaths did not normally mention the provision of specific services but, in the context of the lay aristocracy, they certainly served to sustain aristocratic clienteles. This was all the more important if one considers that clienteles were a fundamental element in the blood feuds and private wars which were used as a means to establish and exert power in twelfth-century Italy.

The Making of the Libri Feudorum

In light of all these elements, the texts of the antiqua can be reappraised as products of a specific social and political context. The 1037 ordinance had sanctioned (or confirmed) the heritability of archiepiscopal fiefs which, although fragmented, in the early twelfth century remained a valuable tool in the hands of a military elite (capitanei and valvasores) who constituted an important part of the urban ruling class and used these fiefs as a means to cement their clientele. The rise of this elite and the decline of the archbishop’s power are manifest in the latter’s gradual eclipse in the texts of the antiqua. In the first section of the book (c.1100-20), the archbishop appears at the top of an imagined hierarchy of tenures (LF 1.1). About 1150, however, Obertus did not mention the archbishop in his tracts and when he set out the hierarchy of tenures once more he placed the emperor (princeps) at its top (LF 2.10). This should not induce us, however, to believe that Obertus was a champion of imperial authority. Even though he was appointed missus imperialis by Lothair III and managed the relationships between Milan and Frederick I, his letters rather seem to acknowledge and support the increasing agency of the Milanese capitanei and valvasores. An early tract (c.1100-20), following the letter of the 1037 ordinance, stated that fiefs granted by high nobles (‘marchiones, comites et … capitanei’) ought to be forfeit only for evident crimes whereas those fiefs granted by lesser or ‘the smallest’ valvasores might be seised at the lord’s will; but its author added, not without a hint of bitterness, ‘the moderns [i.e. present-day writers], who do not distinguish so subtly, say that what is said for the greater valvasores ought to be observed for the smallest valvasores.’ Obertus, in LF 2.10, held that this was common practice in the Milanese courts, but felt it necessary to draw a boundary between the traditional aristocracy and these newcomers by pointing out that ‘they who do not hold a benefice from long ago, although they recently acquired it from the capitanei or valvasores, are nonetheless commoners (plebeii).’ Both authors were witnessing, and to some extent prompting, the decline of archiepiscopal authority and the increased agency of the Milanese elite. Obertus, in particular, was a protagonist in the political life of Milan, part of a restricted group of legal experts who ran the highest civic court and shared the rule of the city with the same military aristocracy which he liked to depict as a closed social group.

It has been argued that these early texts are ‘not actually very close to what the 1037 ordinance had said’ and that they were the outcome of scholarly discussions. But our evidence shows that they may be better envisaged as formulations by interested actors directly involved in the developments of their society. Divergence of opinion seems not to reflect different intellectual views about the text of the ordinance, more than that it is a manifestation of the variety of practice in different places, times and social levels—as we have seen in our examples. This appears in the treatment of succession and obligations. Whereas Ugo de Gambolado (writing before 1136) thought that fiefs consisting of counties, marquisates or other high imperial offices ought not to be heritable (LF 1.13), and another early author implied likewise (LF 1.1), a third one, perhaps commenting on Ugo’s passage, admitted that ‘no succession lies in a fief of a county, a marquisate or any other high offices according to reasonable usage, but today this [succession] is taken into use (usurpatum).’ As we have seen, the bulk of the Milanese fiefs were subject to partible inheritance. All the authors of the antiqua acknowledged this principle (LF 1.1, 8, 13, 18, 24; 2.11). A notable exception was the exclusion of daughters: in one passage it was made clear that this exclusion was allegedly due to helplessness of females to raise feuds or wage war. Succession to daughters, however, was allowed if it was agreed between grantor and grantee, on the condition that no other male heir was left (LF 1.8.1, 1.14), no matter if, as Obertus admitted, such an agreement was against the ‘custom of fiefs’. These elements explain why the clause allowing succession to daughters was generally inserted in infeudations of minor portions of revenues (as in our 1137 example) and not of entire lordships or castles, which implicitly required the receiver to be able to exert coercive power or possess military skills.

Probably because mutual obligations often remained unspoken, encoded within the reciprocal status of the parties, the authors of the LF decided not to commit themselves to any strict definition. The earliest tracts provide generic mentions of serving (servire) and refer the matter to the specific promises sworn in the oaths (pacta or sacramenta), which were deemed binding and could over-ride nearly any principle governing the customs of fiefs—as we have now seen in the succession to daughters. Chapters on how a fief ought to be forfeit mainly express ‘negative’ obligations—not to serve the lord (LF 1.20); not to wound him or besiege his castle; not to sleep with his wife nor harass her; not to kill the lord’s closest relatives; not to reveal to anyone his secrets to his detriment (LF 1.5, 16). The only positive obligation is to support the lord on the battlefield, but this is subsumed in another negative obligation: not to abandon him in the field if he is not yet dead. Obertus, in an attempt to define fiefs in more precise terms, rephrased Seneca by defining the beneficium as ‘a well-intentioned action that confers joy on the receivers and in so doing, the receiver, who is well inclined and willingly prepared for what he does, confers joy’. He continued:

Within this category (genus) of actions, there is one that is given out of someone’s benevolence so that the ownership of an immovable good that is given as a beneficium remains in the hands of the giver, while its usufruct is transferred to the receiver in a way that it will belong to him and his heirs … in perpetuity, and thereby he and his heirs ought to serve the lord faithfully, whether it is explicitly expressed what the service (servitium) should be, or it is promised indeterminately.

Obertus thus admitted that the servitium could be either explicit or implicit, although he added that ‘there is no more just cause to take away a benefice than when [a vassal] refuses to provide the service in return for which the benefice is given.’ Indeed, rather than offering a taxonomy of all the possible services or promises of services that often remained implicit, he and the authors of the LF opted for creating a supple framework into which it was possible to fit actual cases—a framework that was informed by thorough awareness. This is all the more evident in the progressive Romanisation of feudal law in which Obertus was a leading actor and which developed in the mid twelfth century. The process was closely related to the increasing use of legal actions derived from Romano-canonical procedure in the Milanese judicial system. This shift induced plaintiffs, lawyers and notaries to define possession in more precise terms both in the formulation of complaints and in the compilation of charters to be used as proof in court.

This transition is evident both in Obertus’ letters and in the texts that were added to the antiqua throughout the second half of the twelfth century. The letters convey greater precision in terminology than the earlier texts and an explicit, though rudimentary, use of Roman law notions. At the start of the first, Obertus declared that Milanese disputes were ‘solved by either Roman law, Lombard law or the custom of the kingdom’ (LF 2.1). In the following statement, a subject of much debate, he inverted the wording and the meaning of a famous rescript by Constantine, transmitted through the Codex, which sanctioned the supremacy of law over custom and long established uses. Obertus noted that in disputes concerning fiefs ‘the authority of the Roman laws is not negligible, but it does not extend its force so far as to over-ride usage and practice (mores).’

A new procedure based on ancient Roman legal actions had by then become predominant in the Milanese courts. Plaintiffs were required to present their claims according to certain formulae which would then induce the judge to take specific ‘actions’ to conclude the case. Most aspects of fief-giving and fief-holding, however, were alien to the Roman law tradition and had no precedent in the Corpus iuris civilis, which made it difficult to specify claims on fiefs in an effective way. Lawyers, therefore, had to proceed by analogy, and Roman law provided cutting-edge tools for this. It was only at this stage that a specific aspect of fief-giving was framed in terms of property law. Obertus was the first lawyer to define fiefs in terms of ius in re. He tells us that fiefs were things (res) whose possession could be defended against unlawful occupants by asking the judge to enforce restitution by means of an action known as the rei vindicatio, a claim that, according to Roman law, was allowed to full owners only. More precisely, he held that a vassal could bring an action analogous to the vindicatio (‘possit … quasi vindicare’) as if he were the owner (‘tanquam dominus’). This categorisation was reformulated in more precise terms later in the twelfth century, and not in Milan. Here it is only necessary to demonstrate how Obertus stretched the notion of res to fit the varied nature of resources that he saw enfeoffed: a fief, he wrote, may consist of a piece of land (res soli), things attached to land (res solo coherentes) such as buildings and crops, or ‘what can be reckoned as immovable things’ (‘quae inter immobilia computantur’) (LF 2.1.1). The last category, he explained, also included the fiefs de camera and de caneva, annual payments of pecuniary or in-kind revenues that were generally held by persons of non-noble status.

This formalisation, however, informed by a rudimentary knowledge of Roman law, was not an academic exercise unconcerned with social practice. It stemmed directly from practical problems and was aimed at solving them within a precise institutional setting. It attests to a regularisation of the treatment of the possessory implications of fief-giving in the new procedures followed in the Milanese courts. The later recensions of the LF confirm the ways in which the treatment of feudal disputes in court became increasingly entwined with these procedures in a changed political landscape. The inclusion in the law book of legal opinions given by Milanese consuls and deliberations of the city council, as well as more than 40 examples of textual convergence between Obertus’ and post-Obertine texts on the one hand, and the Liber consuetudinum Mediolani (Book of customs of Milan) of 1216 on the other, stand as evidence of this.

The Literature Associated with the Libri Feudorum

The most noticeable effect of the Romanisation of feudal law was an emphasis on the tenurial aspects of fiefs—an attempt to regularise their treatment within known categories of property law. If issues emerging from practice had played a fundamental role in shaping the texts of the LF, they nonetheless continued to compel later lawyers to reinterpret a variety of shifting characteristics of fief-giving in different times and contexts. These efforts did not build only on the interpretation of the fief as ius in re. They also tackled matters that were not directly connected with property law, such as personal obligations and service, that continued to stir up issues in practice and court, as well as debates among lawyers. Although the definition of the fief as ius in re was of great importance for later articulations of power, as the case of Pillius will show, it is far from representing a comprehensive definition of the entire set of practices of fief-giving. It rather suggests that the tenurial aspect was the most likely to give rise to litigation—not a surprising fact if one considers the volume of exchange and fragmentation of fiefs. The following sections turn to the circular relationship between practice and legal theory, to the ways in which some practical aspects of fiefs could shape legal theory and how theory might then be re-contextualised in different settings.

Glossae and Quaestiones by Pillius de Medicina in Modena

Pillius de Medicina (d. c.1212) was one of the most talented and innovative lawyers of his time. A professor in Roman law at Bologna, he moved about 1180 to a new law school founded in Modena, where he used the LF as a text book for the first time in its history, implicitly admitting it alongside the traditional texts of Roman law—Justinian’s Code, the Digest, Institutes and Novels. Pillius’ teaching based on the LF has come down to us as a commentary in a version heavily reworked by later scholarship, and an apparatus of glossae that created a standard reference tool for future teaching and interpretation of the LF. The importance of this apparatus in terms of legal doctrine cannot be stressed enough: by building a first, structured system of connections between the LF and the Roman law texts, Pillius ‘opened the path for the Libri feudorum to be quoted by the glossators, and within a few decades to make their triumphal entry in the standard version of the Corpus iuris civilis’.

One of the major accomplishments attributed to this apparatus is the formalisation of the notion of ‘double ownership’, one of the most successful legal devices to frame fief-holding in later centuries. While working for the commune of Modena, Pillius, like Obertus, was concerned with formulating a connection between a holder’s claim on the possession of a fief and a proper legal action that could be accepted in the civic courts. The question was: if the vassal can act as an owner of a fief, how can we distinguish his ownership from that of the lord who granted it? The solution he proposed was a distinction between two types of ownership: the dominium directum retained by the lord and the dominium utile transferred to the fief-holder. While acknowledging a separation and a hierarchy of ownership, this distinction solved an urgent problem: since in practice a lord and a fief-holder held different rights of real property in the same thing, it was not possible to allow them the same legal action in court. The subtlety was to resort to the notion of utilitas, which in ancient Roman law was used to connect by analogy rights and claims that were not covered by existing law (in this case, rights and claims relating to fief-holding) to known standard formulae—the rights of direct owners and the legal actions available to them. Obertus had already suggested in elementary terms that the vassal could quasi vindicare the possession of a fief; Pillius devised a more refined solution: by virtue of his dominium directum, a lord could use a rei vindicatio directa while the fief-holder could use a rei vindicatio utilis.

This fundamental distinction appears in two glossae: ‘rebus suis’ (LF 2.3) and ‘vindicare’ (LF 2.8.1). The former tackles the question of whether an heir who is forbidden to alienate an inheritance may enfeoff it; it represents the earliest description of fief-giving as a form of transfer of dominium utile. In the second glossa, Pillius affirms that since the vindicatio utilis is allowed for any perpetual or long-term lease of land, ‘even more so it is commonly allowed (concedi solet) to a fief-holder’. Pillius was facing a fundamental issue that had already been addressed by Obertus: how to define rights and claims relating to practices that had no precedent in Roman law in order to regularise their treatment within the Romano-canonical procedure adopted by the city-communes, in particular in Modena, where he lived and taught. One important document outlines the context in which Pillius came to formulate this theory. In October 1182, a dispute was settled between the commune of Modena and several great prelates, including the bishop: an earlier municipal statute had allowed long-term lessees and fief-holders of the main Modenese churches to alienate the tenures located in the city or in its surroundings. Under this statute, holders were permitted to alienate these tenures in any way except sale, even without the lord’s approval; any earlier sub-lease or sub-enfeoffment of these tenures was confirmed; lords were prevented from recovering them even when holders died without heirs, to the benefit of sub-tenants. This means that the rights of long-term lessees and fief-holders had been de facto acknowledged as very similar, yet not identical, to those of lessors and grantors. This is precisely the situation that Pillius had in mind in his glossa ‘vindicare’ when he suggested that all these holders could bring a vindicatio utilis. He framed a contingent configuration of socio-political relationships within a precise legal notion. The creation of one of the most important distinctions in the history of the law of real property was shaped by tangible, practical issues.

Ardizone and the Summa Feudorum

A very similar approach to practice is recognisable in the treatment of personal obligation and service by Iacobus de Ardizone, one of the most cited authors in feudal law. A member of a leading Veronese family of bankers and legal experts, Ardizone studied law at Bologna in 1224-8 and worked as a public official and legal expert in Verona around 1229-54. At this time in Verona, urban and rural institutions were intimately connected, and it was not uncommon to find notaries from the contado working at the communal palace or urban judges attending seigniorial courts. On 2 January 1230, a young Ardizone counselled the prior of the canons regular of S. Giorgio in Braida, a suburban church of Verona, in the seigniorial court held in the village of Sabbion. The community was dominated by a group of fief-holders of the church, whose fiefs were tax-free farms (mansi) and shares of seigniorial incomes held in exchange for the annual render of a horse. Although, in theory, fief-holding distinguished local elites from lesser villagers who were subject to heavy taxation and demeaning services, marriage practices blurred social distinctions. The trade in land and partible inheritance, furthermore, led to the fragmentation of these small fiefs among male successors, generation after generation.

When disputes about these fragmented fiefs arose among holders, they first tried to settle them by private arbitration or by force; only as a last resort were the cases brought before their lords, the canons of S. Giorgio. On the other hand, disputes between lords and their fief-holders were usually judged by the latter’s peers—peers only in theory since S. Giorgio’s vassals encompassed powerful capitanei as well as urban nobles, lesser knights and squires. In the early thirteenth century, however, when the jurisdiction of the commune of Verona over the entire contado was unchallenged, the city courts became a regular place for setting disputes even in a small rural hamlet such as Sabbion. In 1223, Aimo Tebaldi, a notary public from this village who possessed a small share of an old fief held of S. Giorgio, was commanded by his lords to prepare a written description of the entire family fief, which consisted of both land and shares of seigniorial income. In a shrewd attempt to avoid providing services, or perhaps to buy time and thus delay the lords’ inquiry, he produced a complicated document which the canons struggled to interpret. The case reappeared on 14 February 1229 in the city court before the judge and consul Martinus da Cerea, where the parties exchanged opposing statements about the length of time Aimo and his family had failed to provide a horse in return for their fief—S. Giorgio’s attorney maintained it had been five years, whereas Aimo stated that it had been only three.

The case was discussed at the seigniorial court held in Sabbion before Ardizone just a few months later, in January 1230. It has recently been demonstrated that the manuscripts of Ardizone’s Summa feudorum transmit different versions of the treatise: a first incomplete draft was circulated soon after he came back from Bologna, about 1229-30, whereas at least two extended and updated versions appeared after 1234. In the extended versions, Ardizone added a considerable number of references to local practices labelled as ‘the custom of Verona’ (consuetudo Veronensis). One of these additions dealt with horse-fiefs:

One may say that, by custom (consuetudine), when there is mention of horse service, the vassal ought to keep a horse and offer it to his lord at his will. … The application (usus) [of this custom] in Verona is that … the vassal ought to provide a horse of medium size. … Should the horse die accidentally during the service, by the custom of Verona the vassal ought to buy a new one after five years. Even though some courts hold [that this should happen] after seven or eight years, here we prefer to retain it as five, as it is counselled by experts (a prudentibus consultum).

The chapter describes a particular usus and consuetudo of Verona and frames it within a more general customary rule according to which the verbal agreement about the provision of a horse was binding on fief-holders. Indeed, the offering of horses as a service in return for fiefs seems to belong to a broader and shared repertoire of models available to lords in different regions of medieval Europe. Ardizone was perhaps aware of it, but he certainly derived particular details of this customary rule from direct observation. The connection between Aimo’s case and Ardizone’s argument could not be clearer, and in light of the latter, we can understand how, by urging Aimo to admit that he failed to provide a horse for five consecutive years, the canons aimed at obtaining service immediately. It is even possible that the municipal official who judged Aimo’s case was one of the prudentes to whose judgement Ardizone referred in the summa.

Obertus had already acknowledged the binding force of verbal agreements on the matter of service, even though he did not pronounce on its content, the entire matter remaining mostly outside written tradition. Ardizone was concerned with the validity of such verbal agreements and their binding force. He made clear in his summa that unwritten norms had the same value as written ones and the LF was no exception. The ‘custom of fief’, he wrote, was contained only partially in the LF and did not necessarily acquire greater authority from the mere fact of being written down. On the contrary, unwritten custom had its own intrinsic legal force.

Jean Blanc of Marseilles and the Provençal Context

Pillius and Ardizone were the first glossators to produce structured arguments based on the LF. It is useful to follow how these arguments were received and possibly reused in different geographical settings. By 1237, a young scholar from Marseilles, Jean Blanc, had concluded his legal studies at Modena; he went back to Provence, taking with him a copy of Ardizone’s Summa. Here he built a successful career as a private legal counsellor and a diplomat for the commune of Marseilles (1241-1263), Charles I d’Anjou and King James I of Aragon, for whom he received the oaths of fidelity of the crown’s vassals. In 1260 he wrote an Epitome feudorum, the first work based on the LF written on the other side of the Alps, which was evidently inspired by Ardizone’s Summa. Blanc copied entire chapters of the Italian treatise, including some passages dealing with the ‘Veronese customs’. Quite remarkably, however, he sometimes omitted the adjective ‘Veronese’: in this way, arguments initially shaped by local usage came to describe what eventually seemed like general norms.

To provide a possible explanation for this process, it is helpful to understand how Blanc’s work related to the teaching of law at Modena, in particular the disputations—one of the most widespread teaching methods—introduced by Pillius decades earlier. In a disputatio, students were divided into two groups, which had to deploy opposing arguments pro and contra a given case. The written outcomes of such disputations, called quaestiones disputatae, were then used as teaching material and Pillius’ collection quickly became a standard for this genre. The Epitome reveals how Blanc was heavily influenced not only by the pattern of reasoning underlying these exercises, but also by specific quaestiones he had learned in Modena. In many arguments, his method was to start from a quaestio he knew, derive a principle from it, and produce new quaestiones by applying that principle to cases he observed in practice. The transmutation of local usage into general principles seems to have followed the same logic. As a result, the Epitome appears as a mixture of teachings and readings received in Italy, implemented through practical knowledge accumulated in Provence, all painted onto the canvas of Ardizone’s Summa.

A practical example of how this process of abstraction and reuse of legal knowledge worked in Blanc’s case illuminates the unsettled dialectics of law and practice. In 1186, the bishop of Ferrara enfeoffed to the local capitaneus Taurellus a number of villages and castles in the region around Ferrara. The jurisdiction over one of these castles, Massa, a small centre on the River Po, was soon disputed and the case was concluded only in 1204. The matter drew the attention of Pillius, who set up a quaestio using these actors and places—the bishop of Ferrara, Taurellus and Massa—asking his pupils whether jurisdiction was granted together with the enfeoffment of the castle. To avoid his students becoming distracted by unfamiliar elements and to focus on the crucial points of the argument, Pillius substituted fictional actors for the real ones—with the bishop of Modena enfeoffing Bazzano, a castle not far from Modena, to ‘someone’. Blanc’s Epitome reports this quaestio and supplements it with the opinion of Blanc’s master, Homobonus, omitting all references to specific actors, whether real or fictional.

These intentional omissions—whether of the adjective Veronensis or a quaestio’s real actors—unveil a fundamental mechanism in the construction of the glossators’ legal culture. To some degree, the regularisation of procedures and the nature of the teaching methods drove the abstraction of cases and social facts. The translation of social norms and strategies into written traditions often came at the cost of changing their nature, to adapt them to the encoded system that jurists were building, a process of continuous decontextualisation and recontextualisation. On the one hand, scholastic methods invited lawyers to focus on questions rather than on answers—and thus on how to argue cases rather than to seek or provide rigid solutions—and to connect species to genera, specific cases to general principles. On the other hand, these methods provided a solution for a strictly practical issue: to provide an audience of students and practitioners with tools to be used in the judicial arena. The logic underlying both contexts required legal arguments to be exposed in their purity. All the superfluous details, the potentially distracting background noise, had to be removed. Such mechanisms were evident to thirteenth-century lawyers, but in many cases later interpreters failed to understand the historical roots of legal arguments precisely because of them.

Once a species was connected to its genus, the logic informing the connection could be applied to different arguments and realities. In fact, the trajectory of Pillius’ quaestio did not end with its inclusion in Blanc’s Epitome. To determine whether jurisdiction was granted together with the castle, Pillius answered that one ought to conform to whatever the verbal agreement between the parties had set out (‘quid actum fuit inter eos, ut illud servetur’); should this point not be clear, the judge ought to consider the parties’ relative status and the size of the fief—one of the features of fief-giving we have observed in Milan. Blanc, following Homobonus, dissented: jurisdiction did come together with a castle, especially when there was no explicit agreement on the matter for ‘what is not explicitly excepted is openly granted.’ This is exactly the argument Blanc applied to a famous case he attended in the 1240s, between the bishop of Apt and the nobleman Bertrand Rainaud of Agoult-Simiane, concerning jurisdiction over the men of Simiane. The bishop had received Simiane’s donjon by imperial grant and was trying to recover the possession of the castle and the rights that belonged to it, long held by Bertrand’s family. Blanc favoured the bishop’s claim by resorting to a bodily analogy: he played on the fact that the infeudation concerned the donjon—in Latin, caput castri, the ‘head of the castle’. Since the head is useless without limbs, the grant of the former should always imply the grant of the latter; thus, when the emperor granted the bishop of Apt the castle’s head, the inclusion of the rest of the castle, its appurtenances, and hence the jurisdiction over its district, was implicit in the infeudation.

Feudal Law Without the Libri Feudorum?

The Simiane case not only provides a thought-provoking insight into the complicated two-way relationship between practice and law, it also offers some details that allow us to reappraise the practical use of the LF further. The dispute concluded on 24 March 1247 with Bertrand’s oath of fidelity and homage to the bishop. The charter suggests that the oath reproduced that contained in a much studied letter of c.1020, in which Fulbert of Chartres described to Duke William V of Aquitaine the duties of a fidelis toward his lord. The letter was included in the LF (2.6: De forma fidelitatis) and by the 1220s had become the principal formula for feudal oaths in the French Midi. It has been argued that its widespread use was a consequence of its inclusion in the LF, which is believed to have had the force of law in the Midi. The arguments in favour of this hypothesis, however, are not fully convincing. They rest on the assumption that feudal cases could be resolved only by rules derived from an established corpus of feudal law, that is, the LF, and the evidence has then been forced to fit this mould and, in turn, to reinforce this assumption.

The circumstances of the inclusion of Fulbert’s letter in the LF suggest that it is very unlikely that the oath was derived from this book in the 1220s. It is well known that Fulbert’s letter circulated in canon law compilations, including Gratian’s Decretum. All extant manuscripts of the LF dating to the early and mid thirteenth century, as well as Ardizone’s Summa, present the letter not as a part of the book, but as an extravagans, one of the texts recurrently copied as an appendix to it. In the vulgata version of the LF (c.1250), the opening of the letter states its position in the Decretum (LF 2.7: ‘in decretis causa XXII’), and both Odofredus (c.1240) and Blanc (c.1260) envisaged it as part of the same canon law collection. Before then, the content of the letter and its application to feudal cases had already been the subject of commentary by canonists such as Huguccio (d. 1210) and Tancred (d. c.1236). The use of this formula in feudal oaths, therefore, does not prove the statutory force of the LF in the Midi. It might prove—though the matter should be investigated further—the influence of canonists on the formulation of feudal oaths and feudal relationships. This hypothesis might find support in studies suggesting that fief-holding was widespread in Provence as a compromise to protect ecclesiastical land—a means devised by lawyers to convert into fiefs the tenures usurped by local nobles. Be that as it may, there is no direct proof, besides the circulation of Fulbert’s letter, of any use of the LF in Provençal practice until the 1260s.

It has also been argued that references to ‘double possession’ and ‘double ownership’ would be indirect proofs of the influence of the LF, but both arguments seem weak. The distinction between possessio civilis and possessio naturalis—that is, possession justified by a lawful title and possession de facto, without legal grounds ‒ appears in archival records from 1224 onwards. This notion, however, does not derive from the LF and there is no reason why we should take for granted that its source was Pillius’ Summa feudorum—a work whose authorship is much in doubt. After all, the acknowledgement of possessio civilis to fief-holders might already have been envisaged by Rogerius, who taught in Provence in the mid twelfth century. This position was rejected by orthodox legal opinion in Bologna, embodied by Iohannes Bassianus and his pupil Azo, who only acknowledged that fief-holders might have possessio naturalis. The notion of duplex dominium could have reached Provence through Pillius’ apparatus to the LF. Its first occurrence is in a donation of 1227, with no reference to fiefs though. It re-emerges in relation to fiefs only in 1235, in a notarial formulary compiled in Avignon. Even if this is reasonable proof of the circulation of commentaries on the LF in the Midi, it is far from demonstrating that the LF was an authoritative source of law. Rather, it attests to the lawyers’ need to frame in legal terms resources called fiefs, which entailed the transfer of power and over which grantors and grantees maintained, or aimed at maintaining, specific rights. In particular, they had to find solutions that could be accepted as part of new procedures, and while looking for these solutions they met with the same problems we have observed in the examples from Milan and Modena.

Lawyers could frame different aspects of fiefs—ius in re, service, oath—by interpreting practice through the lens of Roman law and canon law. From 1250 onwards, after the inclusion of the LF in the Corpus iuris civilis, French lawyers gradually accepted the authority of this book as a source of law. Although they might have found the LF useful in developing legal arguments about fiefs, they nonetheless tended to deny it statutory force in practice. In 1256, Jean de Blanot, the famous Burgundian lawyer, educated in Bologna, distinguished between the consuetudo feudorum contained in the LF, which he scarcely mentioned, and the customs of the kingdom of France or other regions, including Burgundy, which to him were the real authoritative custom. A century later, the lawyer Petrus Jacobi, from Aurillac in the Auvergne and a teacher in Montpellier, whilst he acknowledged the LF as a text that was useful in developing legal arguments, he deemed it, from the first to the last line, as inapplicable in French territory, where disputes about fiefs were determined by local customs which lawyers framed within Roman law categories.

The situation was undoubtedly different in Italy, but not so much as one might imagine. Lawyers and court practitioners framed matters relating to fiefs within Roman law categories before the LF became the subject of interpretation by jurists. Two surviving feudal quaestiones by Bulgarus (d. 1166) specified what legal actions were or were not available in disputes over fiefs, tackling once again the strictly practical problem of how to fit such cases into court procedure. When a court practitioner asked John Bassianus (d. 1197) to provide a consilium about a dispute over a fief, the former was perfectly capable of framing the case within a set of detailed Roman law arguments, without resorting to the LF. Pillius himself constructed all his quaestiones on fiefs relying on Roman law alone, citing the LF but once. Similarly, Roffredus Beneventanus, in the early 1230s, described procedure in feudal cases through Roman law actiones, without mention of the LF. The interest of canonists in feudal tenures and oaths also preceded their utilisation of the LF: when they first discussed the section—or title—concerning fiefs that was inserted in the collections of papal decretals, they rarely mentioned the book, and they came to use and cite it systematically only in the late thirteenth century.

These contexts all suggest that the influence of the LF on legal practice needs a thorough reappraisal. Despite the dramatic increase of explicit citations of Roman law in charters, there is no clear evidence for any direct use of the LF in scribal practice in this period. When litigants, scribes or attorneys thought it useful to resort to a sense of authority in the regulation of fiefs, they referred to either usage held in local courts (usus curiae) or the ‘custom of the kingdom’. From the late twelfth century onwards, the latter expression was often replaced by a more technical usus or ius feudi. These notions, however, very often covered matters that were not treated by the LF at all. While the LF slowly became a source of authority for jurists in constructing their arguments, it was itself an incomplete and historically determined expression of a broader and changing repertoire of models, norms and potentially normative practices that underpinned the day-to-day exchange of fiefs and services. All the cases examined in this article show, to different extents, how similar practices of fief-giving and fief-holding developed independently of the availability of the LF and written law in general. Lords—whether lay or ecclesiastical, integrated into civic government or isolated in their rural castles—did not seem to need the LF to understand what a fief was. On the contrary, these shifting practices and strategies of fief-giving could influence—sometimes even shape—the early learned law of fiefs to an extent larger than the learned law seems to have been able at first to influence them.

Bringing the Feudal Law Back Home

The alleged disjuncture between practice and learned law ought not to be either an obstacle to our interpretation of medieval societies or a source of mistrust in the learned law. It should rather be an encouragement to reappraise our interpretation of the role of practice both in the creation of learned law and the making of scholarly traditions. The historical value of the LF is at once augmented and reduced. The book increasingly stimulated the attention of learned lawyers and from the late thirteenth century onwards it came to be known by scholars across Europe. Nonetheless, the slow development of this process and the limited influence of the LF on scribal practice require a re-examination of the part the book played in shaping the learned vocabulary of fiefs. Classic models of feudalism have gone too far in their insistence on interpreting its institutions as the embodiment of military values shared across medieval society: the evidence analysed in this article confirms this overemphasis. It may also be true that modern interpretations of medieval societies as ‘feudal’ have been led, either directly or indirectly, by notions and terms contained in the LF. The arguments set out here, however, cast reasonable doubt on the fact that a high medieval academic milieu was the source for this. Fiefs and the logic that informed their exchange did not need learned or professional law to exist in the real world.

This reappraisal of the creation of the learned law of fiefs has demonstrated some aspects of the nature of legal work and writing, its uses and modalities—in other terms, its practical nature and hence its intellectual dimensions. As scholars trained in scholastic methods, the authors commenting upon the LF often argued about the law’s subtleties, but they never lost sight of the practical purpose of their activity. They were primarily concerned with two orders of problem: shaping legal devices with the tools available to them to fit the practices of fief-giving in the new Romanised legal procedure; and providing their audience with adequate legal training. They were less focused on creating substantive, abstract rules for fief-holding: they were more interested in posing new questions and elaborating sets of arguments pro and contra to be used in the judicial arena to defend any claim—or the opposing counterclaim. From this perspective, it is the more apparent that notions related to fiefs and vassals were not the by-product of a scholarly tradition.

If daily practice continued to exert influence on learned lawyers, in a more or less direct way which it is sometimes possible to trace, it seems insufficient to insist only on theoretical analyses of fiefs as the way in which the law of property was shaped. It might be equally fruitful to investigate the shifting values and models underlying bonds of patronage in communities that were undergoing processes of bureaucratisation and the centralisation of power. For that analysis, we must look beyond the aristocratic elites and the evidence connected to them—which has thus far been the main focus of attention. To bring the feudal law back home, we need to include in the picture a broader range of actors—lesser clienteles, castellans, rural knights, peasant elites, ministeriales—all the forgotten agents in the historians’ arguments pro and contra feudalism.