‘Because their patron never dies’: ecclesiastical freedmen, socio‐religious interaction, and group formation under the aegis of ‘church property’ in the early medieval west (sixth to eleventh centuries)

In the early medieval west, patronate, as adapted from Roman law, was a fundamental category in determining the legal status of freedmen. In many cases it entailed a basic set of obligations. In an increasing number of situations, however, the patron became an ecclesiastical institution, since slaves and freed persons were often given to churches and monasteries. As ecclesiastical institutions regarded their patronal rights over freed persons as part of inalienable church property, the patronal relationship became permanent and inheritable. In Eastern Francia (the Rhineland and beyond) this transformed ecclesiastical freedmen into religiously defined social groups with potentially distinct aims, religious tasks, and organizational structures, and a shared notion of freedom. From the Carolingian period onward, it even became attractive to enter voluntarily into this status. It is argued here that with its underlying network of socio‐religious relations, patronate over ecclesiastical freedmen and censuales can be better understood when considered as an element of a ‘temple society’.

schools, but from the fourth century onwards also acquired huge quantities of land and had innumerable dependants with different obligations towards their ecclesiastical lords. 1 There is no denying that the French revolution took an axe to a centuries-old system of manifold dependencies rooted in ecclesiastical landownership. Yet, when framing ecclesiastical property within a wider perspective that includes both the logic of its acquisition and its function as part of wider social interaction and exchange, one feels increasingly uncomfortable with approaches that neglect the religious implications. Consequently, one can ask to what extent the 'temple society' is a helpful paradigm that allows for a better understanding of religiously motivated social practices involving ecclesiastical property.
By drawing on definitions of the essential characteristics of 'temple societies', 2 four observations can be made at the outset. If (1) rituals of 'temple societies' express an idea of the reigning deity as a sovereign, this reminds us that the inalienability of church property encapsulated a typically medieval notion of 'sovereignty' by defining an ultimate principle beyond human reach, 3 with saints and their cults representing this idea on earth. If (2) in 'temple societies' the sovereign deity stands at the centre of moral and economic transactions that constitute redistributive processes, this might lead us to think of the great variety of transactions so evident in ecclesiastical cartularies and in liturgy. 4 If (3) temple endowments provide the organizational framework for individuals and groups to participate in these redistributive processes and share in its benefits, the early Middle Ages come into mind as an epoch in which numerous collectives were created anewfamilies organizing their memoria, guilds, and communities of clerics, who organized and performed the rituals and networks that connected them. 5 Finally, if (4) conflicts generated by this process are resolved by an outside agency whose mandate is to 'protect' the temple, one can 1 E. Lesne, Histoire de la propriété ecclésiastique en France, 6 vols (Lille, 1910-43). some to be manumitted and to play a particular role as freedmen in the 'temple '.
In what follows, freedmen belonging to churches and monasteries will be addressed to reveal forms of social and religious interaction with early medieval churches. Proceeding from a study I published a decade ago, 12 my current aim is to show how ecclesiastical dependency, as enacted through manumission and defined by patronate, created a social space of its own that embraced manifold types of interaction. Between the sixth and early eleventh centuries, manumission and the status of ecclesiastical freedmen underwent changes that were not merely the result of an evolutionary process that transformed late Roman models on a local level, but were also framed by ecclesiastical and royal legislation.

Manumission as a religiously motivated act in the post-Roman west
Inscriptions offer the first attestations to religiously motivated manumissions in ancient Rome. Slave-owners, among them many former slaves, built mausoleums for themselves and their families and granted access to these spaces to freedmen whom they expected to perform sacrifices and keep their memory alive with lights and incense. 13 If they had no children of their own or deemed their slaves to be more reliable in this respect, these slaves required manumission beforehand so that the manumitter's memory could be eternally preserved. 14 Furthermore, a deceased slave-owner's colliberti as a collective made sacrifices in his memory. 15 Although from early on Christian writers encouraged their co-believers to manumit slaves as a pious deed, 16 there was no immediate fundamental change once Christianity was legal within the Roman 12 S. Esders empire. That the first Christian emperor, Constantine, allowed manumission for religious purposes to be conducted in churches, the manumissio in ecclesia, 17 makes clear that this was a practice shared in both pagan and Christian conceptions of memoria. Some historians have interpreted Constantine's laws as aiming at an 'outsourcing' of administrative functions to the church, 18 but distinct Christian conceptions of sin and its forgiveness through pious acts, and specific ideas of the afterlife, will have had a significant long-term impact on religiously motivated manumissionsnot least in quantitative terms. 19 Thus, a canon of a Gallic synod held shortly after 561 refers to slaves 'who in accordance with the quality of their service have been attributed to the graves of the dead' and who should, according to the will of the deceased, be defended either by the heirs of the manumitter or by a church. In the latter case the freedmen and their descendants should be granted legal protection (defensio) in every respect and pay renders (occursus) to the church. 20 This church not only arranged the defensio of the freedmen, but also took care that they and their descendants performed their commemorative function, which, it was believed, could not be aptly administered by an unfree person. Compiled in the later seventh century, Marculf 's formulary collection, contains several templates for manumission, among them one for drawing up a will according to Roman law. 21 This refers to a variety of written dispositions (iuxta quod ipsas epistolas continent) according to which slaves could be manumitted for the ransom of the soul. Henceforth, as free persons, and followed by their offspring, they not only had to fulfil the obligation to render 'payments and lighting-fee' 17 On Constantine's laws, only partially preserved, see F. Fabbrini, La manumissio in ecclesia (Milan, 1965 (oblata vel luminaria) at the grave of the manumitter (ad sepulchra nostra), but also had to be obedient to his heirs. 22 In his will of 616, 23 Bishop Bertram of Le Mans gave his property to several ecclesiastical institutions including the episcopal church of Le Mans, but in particular to his own foundation, the monastic church of Sts Peter and Paul, where he wished to be buried and have his memoria kept alive. He combined this bequest with an alms foundation for the poor. 24 Bertram manumitted eleven slaves with their wives and children, and likewise wanted a higher number of servants of Roman and barbarian birth (tam natione romana quam et barbara), along with their peculium and conlaboratus, to be subject to the patronal protection (sub tuitione et defensione) of the church. They were all to convene there on the anniversary of Bertram's burial to offer their payments (oblata) in his name in front of the altar, 25 and each was to perform his respective service (ministerium) in the name of God. In addition, they were to offer support (solacium) for an annual meal to be held by the abbot in memory of the manumitter. The abbot, on the threat of eternal damnation, was responsible for commemorating Bertram, celebrating his deposition and lighting his grave (depositionem et lumen sepulturolae). Moreover, Bertram ordered that a group of slaves from his familia who performed administrative functions and held a condoma equally be liberated by epistles and, along with the other freedmen, become part of the community destined to annually celebrate his memory (tam de luminario quam de cineribus) for the rest of their lives. Furthermore, their offspring were to be subject to the abbot's defence (defensio). 26 In his testament of 739, 27 patricius Abbo of Provence 'donated' a large number of freedmen, many of whom he referenced by name, 28 to the monastery of St Peter in Novalesa he had founded some years before. 29 Some liberti are mentioned along with servi, mancipia, coloni, and others as attached to individual pieces of property (colonicae) which some of them held as benefice (in benefitio) sub nomen libertinitatis. 30 Given the number of freedmen and the prominence of patrocinium over them, it is no surprise to find some sort of general statute on the status of the freed persons at Novalesa. This statute emphasized the obligation of all liberti to show obedience to his heir, their new patron St Peter, and henceforth to pay to the monastery the dues (impensio) they had formerly lawfully (iuxta legis ordine) paid to Abbo's parents and him. In case liberti should contumaciously and ungratefully rebel against his heir, the monastery's agents should coerce them cum pietatis ordine, and if they ungratefully and rebelliously declined to pay the impensio or tried to rid themselves of their patron, the judge was to force them 'according to what is contained in the "law about ungrateful and contumacious freedmen"' (quod lex de ingratis et contumacis libertis continet). 31 This lex was almost certainly a law of Constantine allowing a patron to return an ungrateful freedman to servitude. 32 While freedmen were often expected to care for luminaria and wax in the context of liturgical memory, for Paul Fouracre royal immunity charters that granted ecclesiastical beneficiaries fiscal income to provide oil and wax for lighting their churches were situated in a wider context of a 'transition from a fiscal to a moral economy'. 33 In many places, groups of ecclesiastical dependants called luminarii shared in this by providing their 'poll tax' in oil or wax so that the ecclesiastical institution could keep an eternal light and special lighting for memorial 27  purposes. Churches' practical need for lighting materials often caused them to arrange payments to be made by freedmen in wax or oil, but also following precarial grants and other transactions. Those who were obliged to deliver dues in wax could be appreciated symbolically as it placed emphasis on the religious nature of their dependency, preventing their payments from becoming confused with ordinary taxes, rents, etc.
As 'placing a light on the altar put the offering physically as close as possible to God', this practice also encapsulated a more general idea: 'Keeping the flame burning was a metaphor for the maintenance of the church and the protection of its property.' 34 The development of such a 'light-based' logic with its characteristic blending of religious, economic, social, and legal implications can thus be seen as a typically 'medieval' phenomenon, even more so as it entailed processes of group and identity formation. 35 In the post-Roman kingdoms, the duty of venerating the dead, traditionally fulfilled by relatives, freedmen and friends, was transferred on a large scale to churches and monasteries as institutions specialized in performing such tasks. 36 For instance, Abbo's freedmen were not explicitly expected to actively contribute to his memoria, a function that he had most likely transferred to the 'temple' he had founded. 37 In such cases, freedmen were expected to make an annual payment in memory of their manumitter, while the actual memorializing function came to be performed by monastic specialists. In other cases, freedmen's payments as required in cash could be used, for instance, for charity endowments. It could differ accordingly if a payment were given on the manumitter's commemorative day or on the saint's feast day. 38 2 Ius patronatus as a legal concept defining a status of 'limited freedom' These arrangements followed underlying legal patterns that can be described as adaptations of Roman law. According to legal and social 34  practice of the Roman empire, 39 a slave became free through manumission, but his freedom was usually limited by his former owner who legally became his patron. 40 Under the 'law of patronate' (ius patronatus), often commented upon by Roman jurists, the manumitter continued to have certain legal claims over his former slave. For instance, his freedman was obliged to provide specified work-service (operae libertorum), 41 to apportion to his manumitter a share of his inheritance (under the title of bonorum possessio), 42 and to make certain payments, for instance, if he wanted to obtain a licence to marry. 43 Their relationship being defined in legal terms, the patron also had certain obligations, in particular to legally protect his former slave. That Roman jurists did all they could to define a former slave's continuing dependence on his patron is not only illustrated by an oath the slave had to swear before his manumission, 44 but also by the punishment of the revocatio in servitutem first attested in the third century. 45 Things worsened for freedmen in late antiquity, when the bond between freedman and patron became hereditary on both sides, the former slave-owners and their descendants on the one hand, and the freedmen and their offspring on the other. 46 Patronate over freedmen, once a legal device allowing transition from unfree status to almost unlimited freedom within two generations, had now come to define an intermediate and hereditary class of freedmen who in only a limited number of cases would eventually enjoy full freedom. In most cases, however, ius patronatus defined a legal space in which further negotiations concerning functions, obligations, and status could take place.
In late-and post-Roman Gaul, manumissions that conferred full freedom without patronal right were possible, 47 whilst in accordance with Justinianic legislation. 48 Some formularies invoke the authority of Constantine's law to confer upon a slave full Roman citizenship (civitas Romana) and ingenuitas through ecclesiastical manumission, so that patronatus obsequium was excluded and the new Roman was allowed to make a will. 49 Still, it has been rightly pointed out that patrocinium over freedmen played a status-defining role in late Roman societies, 50 and continued to do so in their post-Roman successors. 51 In fact, if we look at early medieval charters referring to manumission, nouns such as defensio, tuitio, mundeburdium, and patrocinium occur quite regularly, often routinely in paired expressions such as defensio et tuitio. While defensio points to legal protection, it is hard to define what these terms meant precisely at a given time; still, their ubiquity justifies speaking of a legally defined patronate. As the iura patronatus continued to be grounded in Roman legal practice, searching in vernacular termssuch as the Latinized mundeburdiumfor a supposed influence of 'Germanic law', appears to be futile. 52 Naturally, the long-term and indeed post-Roman success of a legal concept such as patronate over freedmen can only be explained by assuming its adaptability and openness to external influence. 53 For patronate, be it called mundeburdium or defensio in our sources, constituted, in Jean-Pierre Devroey's memorable words, 'un rapport de réciprocité brutalement hiérarchique'. 54 Synodal decrees show the efforts of early medieval bishops to define both the nature and boundaries of the 'temple' and their control over its dependants. The most explicit post-Roman statement on the status of ecclesiastical freedmen comes from Visigothic Spain, where in 633 the Fourth Synod of Toledo unanimously issued a sequence of no fewer than eight canons on this topic, 55 taking as a point of departure the idea that all ecclesiastical property was inalienable and had a special destination. Bishops who manumitted slaves from among the dependants of a church (ex familiae ecclesiae) should thus be condemned, as it was deemed impious (impium) since their action alienated that which belonged to the right of the church (ius ecclesiae alienare); accordingly, the manumitted person was to be revoked to the ius ecclesiae by the succeeding bishop. 56 Bishops intending to manumit an ecclesiastical slave without reserving for the church patronate (patrocinium) over the future freedman thus had to offer in exchange two slaves of equal merit and worth whom they had privately acquired (iuri proprio adquisivit). 57 Patrocinium over a former slave was considered to be of equal value to the slave himself.
Consequently, freedmen and their offspring were to remain under the patronate of the church (sub patrocinio ecclesiae) and meet their obligations according to their ability; they should lie under the bishop's legal protection (sacerdotali defensione) and be protected in their free status (in statu libertatis) and their possessions. Neither these freedmen nor their posterity could ever rid themselves of the patronate (patrocinium) of this church, 'because their patron never dies' (quia nunquam moritur eorum patrona). Instead, they had to make a solemn assertion (professio) to their bishop declaring that they became free (liberi) from among the ecclesiastical dependants (ex familia ecclesiae), would not abandon the church's patrocinium, and would show obedience and submission (obsequium vel oboedientiam). If not, their manumission should be regarded as invalid, and for their disobedience they were subject to 'legal action against ungrateful freedmen' (ingrati actione tenentur) 58once more a clear reference to Roman law, 59 which also denoted a freedman's patronal church as his vera origo and his status as conditio originis. 60 The claim of the Visigothic bishops that church property was inalienable points to a more recent layer of Roman law. 61 To prevent it from being alienated (alienatio), particularly by clerics, in 470 the emperor Leo I decreed that church property should be regarded as sacrosanct. Ecclesiastical property (ecclesiae patrimonium), it was stated, needed protection, because the church was 'the eternal mother of religion and faith' (religionis et fidei mater perpetua) 62 a striking parallel to the idea of an immortal patron. In a later lengthy novel of 535, Justinian elaborated on the idea that churches were trustees for what belonged to God and specified that an exchange with church property was only legitimate if confirmed by the emperor, since most ecclesiastical property derived from imperial munificence. 63 The bishops at Toledo may thus have been inspired by the implementation of Justinianic law in Byzantine Spain. Enforcing the principle of inalienability benefitted lawful transactions concerning ecclesiastical property for centuries to come, as predominantly only rights of a limited nature could be conferred. 64 Naturally, as with every legal principle, strategies to circumvent it were soon developed. 65 Still, the predominance of usufruct donations, precaria and beneficial grants in ecclesiastical charter collections speaks for itself, as they left the issue of ownership untouched. Exchange was considered the only legitimate way of alienating ecclesiastical property. 66 Most of these provisions and arguments were not an exclusively Visigothic or Spanish feature. In Merovingian Gaul, the inalienability of church property was likewise applied to ecclesiastical slaves and freedmen, 67 with bishops using the same rhetoric. At the Second Synod of Mâcon held in 585, they sought to protect ecclesiastical freedmen against bullying, for they had acquired 'the patronate of the immortal church' (immortalis ecclesiae patrocinium). 68 Considering a patronal relationship as 'immortal' clearly echoed both the inalienability of church property and the fact that a freedman's offspring inherited the parental status. Moreover, the idea that a church as an ecclesiastical freedman's patron never dies, found its complement in the belief that property should be given to the altars of saints. 69 In the post-Roman west, the more general late antique notion of the saint as a patron thus transformed into the idea that the saint relentlessly protected church property by inflicting gruesome punishments upon any offender, either on the spot or later. 70 The saint, who could not die, acted as an equivalent to the idea of a juristic person, 71 with the further advantage that in protecting ecclesiastical property, a saint was imagined to be capable of being simultaneously present at a number of locations. This was an assumption that could be neatly linked to the fact that ecclesiastical holdings were often widely scattered. Since the saint protected the familia of dependants of his church, which the freedmen conceived of as their eternal mother, it is telling that the saint was not classified as pater familias, but as their patronus. The metaphor echoed the saint's power to intercede with the supreme power of God. 72 The Visigothic canons proved to be hugely influential beyond Spain, in particular in the Carolingian empire, where they became widely known through collections of canon law such as the huge Collectio Hispana sistematica, 73 preserved in three Lyonese manuscripts alone and used by the future archbishop Agobard to compile his influential Collectio canonum Dacheriana. 74 Thus in later collections, such as Regino of Prüm's Libri duo de synodalibus causis 75 and the Decretum of Bishop Burchard of Worms, 76 they were considered the standard of canon law. As with their Gallic counterparts, they bore the bishops' stamp on the formation of discourse and rhetoric about their 'temple's' possessions: ecclesiastical property was administered on behalf of God, who was the actual owner. It was accumulated through pious endowments, constituting an irrevocable sacrifice (oblatio), 77 and being inalienable, was destined for special purposes such as feeding the poor and sustaining the clergy. Anyone infringing upon church property was thus defamed as a necator pauperum. 78 Following this change in discourse, a church was considered the immortal patron (patrona) of freedmen and their offspring, with this patronal relationship understood to be perpetual on both sides and sanctioned by the saint.
4 Luminarii, cerarii and cerocensuales: long-term regional and cultural perspectives The need to light churches, as demonstrated by Paul Fouracre, 79 called into being religiously connotated economic circles. In the seventh and eighth centuries, documents attest to groups termed luminarii and cerarii. These terms appear to have resulted from the memorial wax payments made by freedmen manumitted with religious intent, from special mansi devoted to lighting, and from a more general obligation of freedmen to pay a 'poll tax'. Charlemagne, in a general capitulary issued in 779 at his royal palace at Herstal, proclaimed that ancient custom should be observed (sicut a longe tempore fuit) with regard to persons manumitted into the status of 'wax-payers' (cerarii), 'by tables' (tabularii), and 'by charter' (cartularii). 80 These groups would have a long future, too. In a fundamental study of the rich charter evidence from Rhineland, Manfred van Rey traced a connection from the first wax tributes fixed for freed persons in the late eighth century, to the high medieval cerocensuales, a considerable number of wax-payers associated with ecclesiastical institutions and living across the Rhineland and beyond. 81 Indeed, for several regions that once belonged to the eastern part of the Frankish kingdom, a continuous development with its regional variations can be traced over a millennium, from the early medieval cerarii and cerocensuales, to the German Wachszinser who may have numbered in the thousands 82 and eventually disappeared in the early modern period, in some regions only following Napoleon's decree of 1808. 83 Although traceable only from the twelfth and thirteenth centuries onward, addressing the fully developed culture of cerocensualitas can illuminate how these men and women formed an important collective within a medieval 'temple society', wherein 'corporate groups retain their separate identities while being accommodated in the larger ritual and economic process represented by the temple'. 84 Dozens of written iura cerocensualium document their obligations towards their ecclesiastical patron, notably the triad of wax-tribute, a marriage fee, and a post-mortem payment, while their freedom from corporal work and their right to inherit is emphasized. Juristically belonging to the altar of a saint, they had their own judicial court. Sometimes a single obligation, such as paying for the 'dead hand' (mortua manus), could be remitted with a revocation into servitude on negligence of payment being explicitly excluded. Leaving the 'temple' was also possible, regulated by fixed redemption payments. Though their dependency was defined in strictly personal terms, the 'wax-payers' were sometimes given special types of land, the Wachszinsgüter or Lichtergüter, reminiscent of the luminarii's mansi.
The specific group-based religious culture of the cerocensuales is particularly visible. Belonging to a saint, to whom they felt a spiritual closeness, they entertained a direct legal relationship to God. Rendering their tribute in pure beeswax was considered a 'sacrifice of candles', ritually rendered during mass at the altar to display their group identity. As religiously and legally defined communities, the cerocensuales, like guilds, had their own rituals and meals. Their distinct festival culture included, for instance, a candle procession on the feast day of Candlemas (in German, Mariae Lichtmess), while their appreciation of bees became associated with the virgin birth and cult of St Mary, and with St Ambrose who was understood to be a protector of bees. An Easter candle, sometimes weighing more than forty kilogrammes, was made in Passion Week by members of the community. 85 With all these legal rules, customs, and folklore, it is difficult to say when they were precisely conceived. Although a continuous development cannot be denied for the Rhineland, it is clear that in the early medieval period we are only observing modest beginnings. Still, it makes sense to ask why there was such an early and strong development of these features in this region in particular.

The Lex Ribuaria and royal legislation on manumissio in ecclesia in Eastern Francia
Cerocensuales belonged exclusively to ecclesiastical institutions, and thus should not be confused with liti, for instance, a much wider group whose limited freedom was regulated by secular law and who also often figured among ecclesiastical dependants. 86 Although there were several groups of non-slave people whose freedom, when compared with that of free people, appears in various ways to have been 'limited', 87 one should resist the temptation to equate the censuales with these or other groups of ecclesiastical dependants. The censuales' freedom was evidently limited by the patronal power of churches alone. To understand why this group became so prominent in the Rhineland and beyond, one needs to look at the Ribuarian law code, compiled under King Dagobert I, most likely in 633/4. 88 Originally drafted for Cologne, the Rhineland, and its adjacent regions, 89 its aim was, apart from revising norms of the older Lex Salica, to entrench the legal position of the king, royal charters, and the churches in this region. 90 Moreover, the code sought to make the inhabitants adopt a new Frankish identity as Ribuarians (Ripuarii), as its area of application saw a major influx of people coming from elsewhere (advenae). 91 The Lex Ribuaria contains a most detailed law in which we may detect the first piece of legislation on manumissio in ecclesia issued by a ruler since the emperor Constantine. Drawing on regulations issued shortly before by King Chlothar II, who had previously responded to demands by bishops that freedmen be placed under episcopal supervision, 92 a lengthy chapter is devoted to manumissio (in ecclesia) per tabulas and to freed persons who were named tabularii after this procedure. 93 Couched in royal legislative tone (iubemus), it regulated the case of Ribuarian Franks intending to manumit slaves 'in religious motivation or for money' (pro animae suae remedium seu pro pretio), which, despite being Frankish, they were allowed to perform according to Roman law (secundum legem Romanam). In these cases, the manumitter was to hand over his slave with tables (cum tabulis) to the bishop in a church in the presence of the clergy, whose archdeacon would inscribe these tables 'following Roman law, according to which the church lives' (secundum legem Romanam, quam ecclesia vivit). The new freedman (tabularius), along with his offspring, were to henceforth be free (liberi) and subject to the patronate (tuitio) of the church. Payments deriving from his status (omnis reditus status eorum) went to the church in which he had been manumitted, which also became his heir if he had no offspring. Furthermore, judicial hearings (mallum) involving a tabularius had to be held there. Anyone who defended a tabularius against his bishop would have to pay the royal ban of 60 solidi, while the ruler explicitly forbade withdrawing a tabularius from the church to which he legally belonged, which 'we have only recently granted to the churches' (quod dudum ecclesiis concessimus). 94 Roman law is invoked here for the procedure of manumissio in ecclesia in general and more specifically for the use of tables (tabulae) within this ritual, 95 while the subscription of seven witnesses on these tables also points to Roman legal practice. 96 However, the provision sought to implement a practice borrowed from Roman law into a society on the lower Rhine no longer dominated by Roman law. A Ripuarius was a Frank born in Ripuaria, and it was this group that the code addressed specifically, which it did against the background of ethnically defined legal pluralism. It therefore made sense to emphasize as a general rule that the church should live according to Roman law (ecclesia vivit lege Romana). In most parts of Merovingian Gaul the practice of manumitting slaves in a church indeed followed Roman legal practice, allowing for a certain degree of variation and adaptation, whereas the Ribuarian law code imagined a vastly different situation for the Rhineland when exceptionally allowing a Frank to conduct according to Roman law a procedure involving the church. Consequently, any manumission conducted in a church for pious reasons would place the freed person under the patronate of the church where the manumission took place. This differed from other regions, where reserving one's patronal rights remained possible in case of a manumissio in ecclesia. Installing some sort of 'all or nothing' rule for Ripuaria effectively meant that in cases of piously motivated manumissions, the churches in question also claimed the patronal rights over the freedpersons. For this reason, and since the above-mentioned manumission pro pretio suggests an ecclesiastical interest in buying slaves and manumitting them, it is clear that this would result in an increased number of freedmen placed permanently under ecclesiastical patronate.
Moreover, the regulation needs to be interpreted against the backdrop of different types of freedom a slave could achieve according to this code. A person released into the status of a civis Romanus, who, like a tabularius, usually had a wergild of 100 solidi, could even achieve a wergild of 200 solidi if he was manumitted by 'penny-throw' (denariatio) in the presence of the king. 97 Whereas a wergild of 200 solidi equalled that of a freeborn Ribuarian Frank, 98 it was made explicitly clear that such an elevation in status was not possible in case of tabularii. 99 Their status was demonstrably conceived of as a dead-end street: once under the patronate of the church, a tabularius's legal status could no longer be improved, for he belonged to the res ecclesiae and its immortal patron. Chlothar II's royal statute and its incorporation into Dagobert's Ribuarian law code thus layed the ground for the creation of a distinct group of ecclesiastical freedmen, who were clearly far better-off than many slaves and other groups who enjoyed a 'limited' freedom in Eastern Francia. The Merovingian rulers sought to transform episcopal churches into building blocks of power in an area that had lost part of its Romanness and was not fully Christianized. Fittingly, Archbishop Kunibert of Cologne, important adviser of Dagobert I and regent for the under-age King Sigibert III, has long been suspected as one of the driving forces behind the compilation of the Lex Ribuaria, 100 four copies of which were owned by the cathedral library of Cologne alone in the ninth century. 101 Despite these features, Alice Rio has recently cast doubts on the importance of this statute for the emergence of censuales, as she called into question 'whether this law in itself would have the capacity to create such a class'. 102 The impact of the Ribuarian law code and of its royal statute on tabularii thus deserves revisiting. Transcending far beyond the Cologne region and the pagus Ribuarius for which it was originally designed in the seventh century, 103 the geographical area covered by the Lex Ribuaria later included large parts of modern southern Belgium with the Carolingian heartland and cities such as Namur, Mons, and Liège. Further east and beyond the Rhine, parts of the Ruhr area (pagus Rurigowe) including the Westphalian monasteries Essen and Werden, belonged to the enlarged Carolingian 'Ribuarian duchy' (ducatus Ripuarius) and were thus considered to be territories following Frankish law. 104 Southward, Ripuaria reached towards the Moselle region, while the influence of the law-book also extended to the Middle Rhine valley up to Speyer, Worms, and the Alsace. 105 Separate regulations on ecclesiastical freedmen were issued later for Swabia and Bavaria, 106 while most parts of Saxony and Frisia were unaffected by the measures ordained for Ripuaria. Widespread application of the Ribuarian code is suggested by its transmission in almost forty extant manuscripts, many among them from Eastern Francia. 107 Its provisions on 'freedmen by tables' (tabularii) were acknowledged by Charlemagne in Herstal (near Liège) as an ancient custom, 108 who later in his reign confirmed Ribuarian law and even added several novels to it. 109 Louis the Pious, in a memorandum of 819, 110 intended to discuss with his advisers the issue of 'freedmen, who are paying wax or another census to a church and are being treated as slaves'. 111 Thus royal responsibility for ecclesiastical freedmen did not vanish in regions under Ribuarian law. Indeed, Lothar I and Lothar II, who ruled the middle kingdom with the ducatus Ribuarius as its heartland, were in some sources, perhaps belittlingly, called rex Ripuariorum or rex Ripuariae. 112 The Ribuarian code's area of application extended even further. In 826, a nobleman called Germunt donated to the monastery of St Boniface at Fulda three female slaves (ancillae) in the pagus Grabfeld in Thuringia, along with further possessions and seventeen mancipia. Drafted by Fulda's abbot Hrabanus Maurus, a charter detailed that the abbot had decided, upon consent of his community, that the three women should be free from any other servitude and on the feast day of St Boniface should pay two denarii in cash or wax to the church of St Michael in a Thuringian place called Rohr, a proprietory church of Fulda. Following their deaths, their collaboratus were to be used as an elemosinaendowment to celebrate Germunt's memory at Fulda. 113 This all should be done, as was explicitly stated, according to the lex tabularia. As the Fulda cartulary contains several charters documenting donations of slaves into the status of wax-payers, it becomes clear the Ribuarian statute came to be applied in Thuringia, or at least in its Franconian parts, where Fulda held many properties that had been granted immunity. 114 The Ribuarian regulations on ecclesiastical freedmen even entered canon law through the collection of Regino of Prüm, compiled shortly after 900, which amplified their importance. 115 Referring to those manumitted in a church for religious purposes 'according to secular law' (secundum legem mundanam) and thus commended to this church's patronate (patrocinium), Regino inserted Constantine's famous law on manumissio in ecclesia of 321, 116 a provision of unknown provenance, and a large portion of the Ribuarian statute on tabularii, 'since it was written in the Frankish pact' (scriptum quippe est in pacto Francorum). 117 The Lex Ribuaria thus exercised a profound impact on the position of freedmen manumitted in a church, and made this group-forming practice spread widely in larger parts of the East Frankish kingdom. While in other regions of the Frankish kingdom Roman law persisted as customary and adaptable law, the Ribuarian law code, with its provisions on ecclesiastical freedmen, marked a new beginning for those regions in which Roman legal traditions had ceased to be dominant.

'Under a church's tuitio et mundeburdium': some Carolingian transformations
In the Carolingian period, the number of such freedmen participating in 'temple societies' grew significantly, as is documented by numerous charters, but also by lists that kept names for the purpose of collecting their 'poll tax'. 118 Moreover, several interrelated changes contributed to this development. Already before 800, in addition to slaves (mancipia) given to churches for manumission, an increasing number of freeborn individuals voluntarily entered this group. While our earliest example from Cologne is just a short notice, 119 more detailed charters show that protection offered by churches often persuaded individuals to join. Among them were many widows, some of whom also entered religious life. 120 In 853, a widow and deo sacrata named Erkanfrida gave possessions of her deceased husband (senior) along with 95 mancipia to the monastery of St Maximin in Trier, and she released another seven mancipia from the bond of slavery and made them censuales of the church (a iugo servitutis solutos ad ipsam ecclesiam censuales feci). An annual celebratory meal on the feast day of St Martin was to be held for the cultores of this place, as they were to celebrate vigils and masses for her and her senior's soul. 121 The annual 'poll tax' paid to the saint indicated freedom but was often understood as a price to be paid for patronal protection. Voluntarily entering the patronate of a church can be seen as a real game-changer for the definition of the group as a whole, and as part of the 'temple'. These individuals were often in a far better position to negotiate the legal terms of their future life under the saint's patrocinium, while immunity grants put churches in a position to respond more flexibly to their needs and wishes when fixing the conditions of patronate. 122 The steady increase in the number of people entering voluntarily into this status, evident in charters, lists and the later libri censualium, created a self-conscious group that was by no means composed of freed persons alone.
Among those given as mancipia to ecclesiastical institutions, illegitimate children figure prominently. The three ancillae given by Germunt in 826 seem to have been his daughters, 123 while in 827, a man called Eiat gave to Fulda in his elemosina fourteen mancipia along with his three sons and a daughter, presumably born from an illegitimate relationship with a female slave, on the condition that for the lifetimes of these children they were not to be burdened by servitude (vis servitutis), except an annual payment of five denarii in either linen or wax, while the other mancipia should be in full potestas of the abbot. 124 Since children of free-slave marriages usually acquired the lesser status 125 and illegitimacy of children posed a major social problem, 126 placing them (often along with their parent) under ecclesiastical patronate was often deemed a fitting solution that would provide for them and give them a better social status than prescribed by more general legal rules.
In the Carolingian period, the Latin terminology for ecclesiastical freedmen and free people placed under the patronate of churches became vastly differentiated. 127 Some of these terms refer to the payment of the 'poll tax' or tribute paid for the mundeburdium, either in cash (such as censarii, censuales, tributarii, or capite censi), or in wax (such as cerocensuales), in addition to the older terms cerarii and luminarii. Other terms emphasize the patronate as their defining feature (mundiliones, munborati), while still others categorize them as belonging to a saint's altar (sanctuarii). It is remarkable that the ancient term colliberti, which alone carried the implication of manumission 128 and came into use in western and northern France shortly before 1000, 129 is only exceptionally found in Eastern Francia. 130 This diversity seems to reflect regionally or locally diversified customs, with no institution imposing a unified terminology.
The law (lex) which regulated the status of these differently termed groups was now more explicitly referred to. A donation charter of 837 fixed for two ancillae given to the monastery of Wissembourg an annual 'poll tax' of two solidi in cash or wax by reference to other freed persons (sicut et alii tributarii vel censarii seu epistolarii, qui per talem condicionem sunt relaxati ingenui), 131 and the Fulda charter of 826 ostentatiously refers to the lex tabularia. 132 By contrast, charters written after the mid-ninth century first spell out the individual terms of freedom under ecclesiastical patronate. Thus, a free woman called Regnevig, on becoming a tributaria of the abbey St Peter in Ghent around 877, negotiated an annual payment of two denarii for her mundeburdium, six denarii for her marriage, and twelve solidi as an asset payment after her death. 133 Their direct insertion into the charter seems to indicate that the conditions of the patronate somehow became detached from supra-regionally accepted legal customs. These developments effectively contributed to the evolution of local customs (leges) for communities of censuales belonging to a particular saint, and to a finer assessment of their freedom. In addition to clauses stating that these persons should be free from further obligations, 134 several charters now described the rights they should henceforth enjoy in positive terms. A late ninth-century formulary from Cologne for 128  Charters written around 1000 reveal an increasing self-conscience among the censuales. Thus, in a document produced in favour of St Landelin in the abbey of St Crispin (Hainaut) in 1009, a free woman named Bertha emphasized that she 'did not act in the manner of some lords who give their male or female servants' (non ut quilibet domini suos suasque tradunt famulos ac famulas), but followed 'those who as free men or women offer themselves voluntarily to the holy altar of the saints of God' (sed qualiter se sponte offerunt liberi vel libere sanctorum Dei sancto altari). According to law (lex), she and her future offspring would be liable to an annual payment of two denarii on the feast and altar of the saint, six denarii for marriage (pro maritali licentia), and twelve 'for the dead hand' (pro mortua manu). Beyond that, they 'would not be obliged to attend judicial assemblies, provide surety, perform labour work or be subject to any other advocacy' (neque placitum neque vademonium neque servitium nec advocatiam aliquam), except that of the count. 137 Around 980, a free woman called Alcina gave herself along with her future offspring to the altar of the abbey of St Ghislain (near Mons, Belgium), with an annual payment of two denarii by 'eternal law' (lege perpetua), a payment of twelve denarii from her assets, and six denarii for a marriage licence. However, emphasizing that they will be exercising the patronate over themselves (mundeburdem de nobis ipsis habeamus), she would accept no other obligations except those specified (nec placitum nec ullum debitum, exceptis predictis, observabimus). This caused her to insert some reflections on freedom.
Since the Holy Trinity has produced and liberated her (trinitatis, que nos fecit, nos liberavit), and will save her, and 'although she was free and could freely use the licence of secular discretion' (cum essem libera et secularis arbitrii liberaliter possem uti licentia), she 'wanted to exhibit grateful service to the highest king' (volens superno regi grata exhibere famulamina et placita munera), as 'the highest freedom is to be in the service of God, serving whom means to rule as a king' (summa libertas est Dei servitio subiacere et ei servire regnare est). 138 These dependants did not only have something to lose, but could also compare various levels of freedom. From this period, we have the first iura censualium. These were more general statutes that comprehensively fixed the status of these groups, often as part of a religious endowment. When Bishop Burchard of Worms founded a canonical church dedicated to St Paul in Worms in 1016, he detailed the obligations and rights of the numerous mancipia he gave to the altar: no advocate should compel them to perform any servitium, as only a magister in charge of collecting their 'poll tax' could announce two obligatory placita a year. The public fixation of their law was witnessed by eleven clerical functionaries, the fratres of five churches in Worms, forty-three laymen, and pene omnes urbani. 139 The year before, Archbishop Heribert of Cologne gave seventeen mancipia to the monastery of Deutz to absolve them 'from their yoke of servitude', as he specified in slightly less detail their obligations and rights. 140 We are grasping here the legal emancipation of self-conscious groups of censuales, as typical for several episcopal cities along the Rhine and in parts of modern Belgium. 141 While these charters incorporate statutes, 142 most charters refer to overarching norms only in passing, using phrases such as defensio et tuitio. Still, early medieval law was not merely case oriented. Numerous legal texts were circulating, and unwritten custom framed legal expectations. It thus does not seem helpful to play general norms and individual cases against one another, more so as the full repertoire of possibilities also included the exchange of legal ideas. Thus in 1002, Emperor Henry II (1002-24) was approached by Bishop Ansbert of Utrecht to obtain a special privilege. Ansbert asked that those making an annual payment of wax to the episcopal church of Utrecht (homines quoque, qui ceram ad prędictam ecclesiam solvuntur per annos singulos), and who were thus under its patronate (sub mundiburdo et tutione ipsius ęcclesię), should 'enjoy the same law as has been granted to the church of Cologne and the remaining churches that are constituted in our realm' (tali lege sicut Coloniensi ęcclesię concessum est cęterisque in regno nostro constitutis). 143 One can hardly doubt that this is another reference to the statute contained in the Ribuarian law code. Ansbert wished the status of the Cologne ecclesiastical freedmen to be conferred upon the wax-payers of his church. Utrecht was situated in Frisia, which had a legal culture distinct from the regions informed by Frankish law. 144 This did not escape attention at the time, and it was the king alone who, by his privilege, could manage such a 'legal transplantation'. 145 Three weeks after confirming his predecessor's Utrecht charter, 146 on his inauguration circuit King Conrad II (1024-39) was approached by Bishop Walter of Speyer who presented to him private charters for confirmation. Walter had manumitted eleven servile church dependants to make them censuales, who would henceforth live safely forever 'according to the law of censuality' (infra legem censualem) along with their offspring, paying an annual tributum of two denarii or wax on the feast of the birth of St Mary. To compensate for the loss of church property caused by his manumission, Walter donated eleven unfree persons (mancipia) to the church, which he did 'in accordance with both canon law and the authority of capitularies given by Charles, Louis, and Lothar' (canonica capitularisque antecessorum nostrorum Karoli, Luduuici, Lotharii auctoritate fultus). To affirm the procedure's legality, the charter literally quotes 'chapter 65 of the 5th Council of Toledo' stating that 'a bishop is entitled to manumit ecclesiastical slaves and confer freedom upon them provided he gives unfree serfs in exchange'. In addition, it presents yet another quote, this time from a capitulary, that 'without giving unfree in exchange no bishop is allowed to confer liberty to anyone'. Following canon and capitulary law (canonica et capitulari lege), Conrad II therefore gave his assent and ordered the eleven freedmen to enjoy the same legal status as other censuales, fixing for any infringement a fine of ten pounds of gold, one half to be paid to the royal chamber, the other interestingly to the censuales, which underlines their corporate identity as a social group. 147 While the quoted Carolingian capitulary provision is difficult to identify, 148 the Toledo quote provides a clue to the charter's source, as it did not actually derive from the Fifth, but from the Fourth Council of Toledoin fact, it is one of the Visigothic canons referred to above. 149 As this canon is also incorrectly authorized as given by the Fifth Council of Toledo in the famous canonical collection compiled by Bishop Burchard of Worms, the charter's quote must have been borrowed directly from the Decretum, 150 which Burchard had produced, as we learn from his Vita, with the help of his fellow-bishop and friend Walter of Speyer. 151 The charter thus reflects a personal network of two episcopal legal experts. However, in having the king confirm it, Bishop Walter not only obtained the ruler's necessary consensus for a lawful exchange of ecclesiastical property, 152 but also introduced a profoundly new aspect. Effectively dragging him into the business of protecting the legal group of the censuales, Walter held the ruler responsible for maintaining their liberty as a social group of free 'subjects' that would soon play an important part in the