SONDERDRUCK AUS: AUFSTIEG UND NIEDERGANG DER ROMISCHEN WELT GESCHICHTE UND KULTUR ROMS IM SPIEGEL DER NEUEREN FORSCHUNG HERAUSGEGEBEN VON HILDEGARD TEMPORINI UND WOLFGANG HAASE II PRINCIPAT DREIZEHNTER BAND HERAUSGEGEBEN VON H. TEMPORINI DE WALTER DE GRUYTER BERLIN NEW YORK 1980 Tabulae patronatus: A Study of the Agreement between Patron and Client-Community by JOHN NicoLs, Eugene, Oreg. Contents Introduction ........................................................................................................................... 535 I. The tabulae patronatus and the Designation of the patronus ................................... 537 II. Geographical and Temporal Considerations ............................................................ 538 III. The Community and Its Status .................................................................................. 541 IV. The Individual and His Status .................................................................................. 543 V. The legati ..................................................................................................................... 545 VI. Multiple patrocinia and clientelae ............................................................................... 547 VII. The Formulae ............................................................................................................... 548 1. hospitium facere ..................................................................................................... 549 2. patronum cooptare .................................................................................................. 550 3. in fidem clientelamque recipere ............................................................................ 550 4, Variations of Formulae ........................................................................................ 551 5. The Combination of Formulae ............................................................................ 552 VIII. Varieties and Origins of the tabulae ......................................................................... 553 1. The Varieties ........................................................................................................ 553 2. The Origins and Evolution of the tabulae ............................................................ 554 a) The tabulae and the tesserae hospitales ............................................................ 555 b) Decreta decurionum ........................................................................................ 556 c) Decreta magistratuum ..................................................................................... 556 d) Wooden tabulae ............................................................................................... 558 Summary of Conclusions ..................................................................................................... 559 Appendices ........................................................................................................................... 560 Introduction Although over 1000 civic patrons are known from various, especially epigraphical, sources, only a relatively small number of documents survive 35 ANRW II 13 536 JOHN NICOLS which commemorate the award of the honor'. The purpose of this paper is to discuss the characteristics of 30 of the better preserved tabulae in order to elucidate the origin, meaning and significance of these documents2. There are limits to this discussion which must be established at the beginning. The tabulae considered here are all in Latin, involve agreements between municipalities and individuals and date to the period 50 B.C. to A.D. 250. Consequently, tabulae in one of the various Iberian languages or those involving gentes or collegia as the community, or those dating to a period before or after this time span have been excluded 3 . The reason for the limitation of language is self-evident. The limit of time has been defined by the nature of this volume and by the author's conviction that the nature of patronage of communities changed during the course of the 3rd Century 4 . Non-municipal communities have been excluded because they were not controled by the regulations regarding the cooptation of patrons s . imitation was so prevalent, many of the conclusions reached here apply equally to the excluded categories. The tabulae patronatus considered here are all documents in bronze and sufficiently well-preserved that most of the details are either readily apparent or can be restored with reasonable confidence?. The inscriptions are generally rectangular in shape, the height being somewhat greater than the width (dimensions of .35 X .28 X .03 m. are typical, though nr. 21 measures only .1 x .06 m.). Common variations in shape include the addition of an isosceles triangle to the top of the rectangle, yielding a pentagonal shape (on one tabula, nr. 22, this space is filled by a portrait-head); other tabulae are 'winged' (nrs. 15 and 22) and one tabula is in the shape of a boar (nr. 24). The latter form may Frequently cited works: E. BADIAN, Foreign Clientelae (264-70 B.C.), Oxford 1958 = BADIAN M. GELZER, Die Nobilitat der rOmischen Republik, Leipzig-Berlin 1912 = ID., Kleine Schriften I, Wiesbaden 1962, 17-135 (cited with reference to 'Kleine Schriften') L. HARMAND, .Le patronat stir les collectivites publiques des origines au bas empire, Paris 1957 = HARMAND A. DRs, Epigrafia juridica de la Espana romana, Publ. del Inst. Nac. de Estud. Jurid. Ser. 5a Textos Jurid. Antiguos, Madrid 1953 = DRs 1 I:oee s( of patroni in HARMAND, 501ff. 2 For a list of the 30, see Appendix 1 and n. 6. N.B.: all the `nr.' references are keyed to this list. For a list of all the Spanish tesserae and tabulae, see A. GARCIA Y BELLIDO, Tessera hospitalis del Ano 14 de la Era hallada en Herrera de Pisuerga, Bol. Real Acad. Hist. 159 (1966) 158ff. Cf. B. H. WARMINCTON, The Municipal Patrons of Roman North Africa, Pap. Brit. School at Rome 22 (1954) 54-5. s Cf. J. NicoLs, Zur Verleihung Offentlicher Ehrungen, Chiron 9 (1979) 243ff. For the same reason, agreements between communities have been excluded, cf. D'ORs, 370-1. 6 See Appendix 1. Tabulae which may be relevant to this discussion but which are too fragmentary to be included, are: CIL V 5127, IX, 2464, X 5670, XI, 844 and AE 1954, 260. TABULAE PATRONATUS 537 represent the animal sacrificed on the occasion of the agreement'. Despite the fact that many of the tabulae have nail holes for mounting on a wall, they were clearly intended to be portable. The details will be discussed below, but two categories of tabulae should be noted at the beginning'. There is an 'Italian' form, which was used primarily by Italian municipalities and reproduced the wording of the decretum decurionum of cooptation, and there is a provincial form which employed a significantly different arrangement of formulae. In both cases, however, the intention was to commemorate the agreement reached by a patron and his client. I. The tabula patronatus and the Designation of the patronus There are several indications in the tabulae and in the various municipal charters of how the patron was to be designated and what the role of the tabula was in that process. The charters require that the designation be confirmed by a decretum decurionum and, conversely, expressly forbid the magistrates to act independently 9 . This decretum must be approved by the majority of the decurio- nes voting under oath when at least two-thirds of them are present m . The resolution probably included not only the usual formal aspects of such decre- ta", but also a brief justification for the cooptation in which the beneficia, influence and humanitas (cf. nrs. 27, 29 and 30) of the potential patronus were noted. Thereafter, in accordance with the decree, a number of legati would be selected from the ordo decurionum who were authorized to approach the future patron, to present him with a copy of the decree and to request him to accept the community in his and in his descendants' clientela. When the patronus had done so, and it is by no means clear what form this response took, then a tabula aenea (or patronatus) recording the cooptation would be produced, one copy of which was set up in the house of the patron (cf. nrs. 27 and 29) and another in some public place in the community' 2 . Though some of the tabulae reproduce the wording of the decretum, the great majority of them (27 of 30) simply record the fact of the cooptation'3. 7 The significance of t hese variations in shape is not clear, cf. DRs, Miscelinea Epigrafica, Emerita 28 (1960) 143ff. See below, p. 553ff. 9 Cf. lex Ursonensis (ILS 6087) c. 97 and the lex Malacitana (ILS 6089) c. 61. lc' Lex Malacitana, c. 61. These regulations probably applied to citizen communities, in general. Peregrine communities may not have been subject to the same restrictions. " On the form of such dec-reta, see TH. MOMMSEN, ROmisches Staatsrecht III, Leipzig 3 1888, 1007. " L. HARMAND, 333-4. No such 'pair' has been found and it may well be that, instead of producing a second copy in bronze, the community used wood or simply added the name of the new patron to the list of patroni on the album decurionum, cf. ILS 6121. See also below, p. 539 and 558. 13 All these points will be discussed in detail. 538 JOHN NICOLS II. Geographical and Temporal Considerations In considering the problem of the geographical distribution of these tabulae, a clear distinction must be made between the location of the individual community and that of the corresponding find-spot' 4 . This is an important consideration, for the overwhelming majority of the client-communities were in the African and Spanish provinces, while over one-third of the tabulae were found outside the province of the contracting community. The data are summarized in Table 1. Table 1: The Geographical Distribution (for 'nr.', cf. App. I, below, p. 560) Province Communities (Total) Provenance of Tabula (Total) Gain (Total) Tarraconensis nrs. 17-26 (10) nrs. 17-24, 26 (9) Baetica nrs. 14, 15, 16 (3) nrs. 15, 16 (2) Mauretania Afr. procos. nrs. 1, 2, 3, 4, nrs. 6, 7, 8, 9, 11, 12, 13 5 10, (5) (8) nrs. 1-4, 14 nrs. 11, 12, 13 (5) (3) nr. 14 (1) Sardinia nr. 28 (1) nr. 28 (1) Thrace nr. 27 (1) Italy nrs. 29, 30 (2) nrs. 6-10, 25, 29, 30 27, (9) nrs. 6-10, 25, 27 (9) Numidia nr. 1 nr. 1 (1) I Loss (Total) nr. 25 (1) nr. 14 (1) nr. 5 (1) nrs. 6-10 (5) nr. 27 (1) The first point to be noted is that, although there are 30 inscriptions, they concern only 25 communities (one of which is, in fact, a conventus, nr. 25). This is because Pompaelo and Bocchoris in Tarraconensis are each represented by two tabulae (nrs. 17 and 18), Gurza in Africa proconsularis by two (nrs. 11 and 12) and Banasa in Mauretania by three (nrs. 1, 2, and 3). With due regard for the variations in the rate of survival' s , it is apparent that the comnmnities of Tarraconensis and of Africa proconsularis employed the tabula most extensively, while the Italian communities did so less frequently. That the communities of Tripolitana and of Lusitania are not mentioned on any surviving tabulae may be accidental, but it is likely that they, too, only rarely used tabulae to commemorate the initiation of the patronal relationship''. The same conclusion applies to the other western provinces, though perhaps with more certainty for no tabula at all has survived involving a community from the For the details of this distribution, see Appendix 2. 15 Cf. R. DuNcAN-JONES, The Economy of the Roman Empire. Quantitative Studies, Cambridge, 1974, 360ff., and below, p. 558ff. 16 The patronal relationship was, however, by no means rare, cf. B. H. WARMINCTON, art. cit., 43. TAHULAE PATRONATUS 539 Gallic, British, Rhine or Alpine or upper Danube provinces. Indeed, the only transalpine or eastern community known to have employed a tabula is Colonia Flavin Deultum in Thrace (nr. 27). It is not easy to account for the decision to approve the patronal relationship in a form which is otherwise unknown in the northern and eastern provinces. As the colonists of Deultum were veterans of the legio VIII Augusta, it is possible that they brought the concept with them from their respective patriae. Unfortunately, too little is known about the origins of the legionaries of VIII to venture any conclusions about where these communities might have been". What is important however, is the fact that, though the overwhelming majority of the contracts involve provincial communities (28 of 30), a good number of them (7 of the 28 of provincial origin) have been found in Italy. As these seven tabulae involve men who were imperial administrators in the province of the contracting community, it is likely that they represent the copies which were presented on at least some occasions to the patron and which were intended to be displayed in his home". As with many decreta decurionum, a group of inscriptions to which these documents belong or are closely related 19 , the tabulae generally record the date on which the contract was formalized, mentioning the consular year first and then, though somewhat less often, the precise day. This information usually appears at the beginning of the tabula, but on one occasion the order is reversed; namely, the date appears at the end of the text and the day precedes the year. This inscription (nr. 13) is the earliest of the tabulae, dating to the 40's B.C., and appears to have been modeled on magisterial decress20. Of the 30 tabulae, 26 (including nr. 13) give the consular year. Of the remaining four, three (nrs. 3, 6 and 14) are mutilated at the relevant place, but it is likely that they, too, were dated. The fourth (nr. 15) is well-preserved, but records no date at all, a fact which makes it one of the most unusual of the tabulae. It may be that the lack of a date is due to the fact that the document is not really a decretum decurionum but, rather, the patrons's acknowledgment of his responsibilities to the community. Nevertheless, the mention of the consular year appears to be an integral part of the document. Of the 26 tabulae with consular dates, 19 also give the precise day of the year on which the decretum was resolved. As there does not appear to be any common feature which might explain the lack of the precise day on the remaining seven (nrs. 2, 4, 10-12, 22-3) it would seem likely that the day, in contrast to the consular year, was not reckoned as an essential element of the tabula. As for the four undated documents, two of them (nrs. 4 and 15) can, for 7 On the origins of the legionaries in the legio VIII, see G. FoRm, 1l reclutamento delle legioni da Augusto a Diocleziano, Milano-Roma 1953, 227. The form of the tabula is talianrather t han rovincial 8 Cf. nr. 29; tabula hospitali incisa hoc decreto in domo sua posita. See also DUSSALCS comments to ILS 6100. 19 See below, p. 556-7. 20 See below, p. 556ff. 540 JOHN NICOLS various reasons, be dated with reasonable accuracy 21 . The remaining two (nrs. 3 and 14) present more difficult problems and for that reason have no t been included in Table 2. Nevertheless, the data summarized in Table 2 indicates that this kind of agreement was employed most frequently in the early principate. Indeed, 21 of the datable 28 tabulae are datable to the period between the principates of Augustus and Domitian. Table 2: Temporal Considerations, 50 B.C. to A.D. 250 Province 50-1 BC AD 1-50 50-100 100-150 150-200 200-250 Tarraconensis Baetica Mauretania Afr. procos. Sardinia Thrace Italy 17,21 11,13 18, 22, 23, 24 15,16 6-10 19, 26 2, 4, 5 12 27 29 20 1 28 25 30 The numbers given in the table refer to the relevant tabula, Omitted are nrs. 3 and 14, which cannot be reasonably dated. It is noteworthy that such documents appear to have gone out of fashion in Africa proconsularis at the end of the Julio-Claudian era. The latest tabula from this province (nr. 12) dates to the reign of Nero, while the other six were concluded before A. D. 28. In contrast, the earliest documents from Mauretania date to the years 75, 75 and 55 (nrs. 2, 4 and 5, respectively); that is, tabulae began to be used in Mauretania at the very time that they disappeared in neighboring Africa proconsularis. It is unclear to what extent the degree of urbanization and Romanization may have been a factor in this process. In other parts of the empire the situation was quite different. Tabulae were popular in Spain in the first century A. D., and continued to be employed in Tarraconensis throughout the whole period under discussion. It is noteworthy that Italy, which was long familiar with the related tesserae bospitales and with the formal designation of patrons 22 , does not yield a single tabula of its own before the beginning of the second century and, even then, the tabula is in a form quite different from that used in the provinces. More will be said on this problem below, when the variations in formulae are discussed23. 21 The patronus mentioned in nr. 4 is the same man who was cooped patron in nr. 2. The patronus in nr. 15 is generally considered to be of `Tiberian' date, but there is no direct evidence for the conclusion, cf. H. NESSELHAUF, Zwei Bronzeurkunden aus Munigua, Madr. Mittl. 1 (1960) 147. 22 See below, p. 555 and ILLRP 364, 382, 389, 406, 409, 416, 425, 523, 567-8, 617-8, 1064-9, 1276. 23 Cf. below, p. 548fL TABULAE PATRONATUS 541 In conclusion, these temporal and geographical considerations strongly suggest that the tabula patronatus was not Italian in origin. Rather, it seems to have evolved and flourished in the Iberian and African provinces especially during the first century of the principate. The Community and Its Status Despite the fact that the client communties describe themselves with a variety of terms ranging from the specific to the abstract, it is clear that the client is none other than the whole citizen body including the decuriones (or their equivalent), the commons and all their descendants. This conclusion may be deduced from the common formula: eum posterosque sibi liberis posterisque suis patronum cooptaverunt which appears specifically, though with minor variations, on 20 of the tabulae. In another ten (nrs. 5, 17-21, 24, 27, 29, 30), we may assume that the descendants of the client-community are understood as those of the patron (or simply his domus) are formally mentioned as being heirs to the clientela. But, regardless of whether the community describes itself as an abstract entity (colonia, respublica, civitas, municipium) with its proper name (e.g., civitas Gurzensis, nr. 12) or simply gives its collective name without reference to status (e.g., Baetolonenses, nr. 26), the community is to be understood as the present and future citizenry as a whole. The peregrine communities name themselves variously as senatus populusque or as civitas (e.g., Siagitanus or Themetra, nrs. 8 and 9, respectively). There does not appear to be any pattern in this choice based on either size, location or time. Nor is there any reason to think that those peregrine communities which describe themselves as civitates had not adopted the Roman municipal model implied in the expression senatus populusque. Communities with the status of colonia generally describe themselves in a similar manner, combining the status with the proper name. Hence, the community of Banasa calls itself coloni coloniae Juliae Banasae (nrs. 2 and 3). Provincial municipia, however, tended to use simply their collective name, as did the Baetolonenses (nr. 26) or to combine the collective name with the designation respublica. In three cases the general population (and its descendants) is subsumed under the title colonia nostra (nr. 27) or municipium nostrum (nr. 29). Such tabulae record the petitions, in the form of decreta decurionum, which the ordo made on behalf of the whole community. That the descendants were also included is demonstrated by the reference to the domus of the patron (nrs. 29 and 30) which is to retain the clientela. Hence, despite the considerable variation in the terminology used to describe the contracting community, it is evident that the collective of citizens, present and future, is to be understood. The tabulae discussed here emanate from a variety of communities, from coloniae, municipia, civitates. Indeed, though time and place are important 542 JOHN NICOLS variables, the decision to employ the tabula in order to commemorate the initiation of the connection between individual and community appears to have depended in part on the status of the community. Tables 3 and 4 summarize the evidence. Table 3: The Status of the Community in Time, 50 B. C. to A. D. 250 Status 50-1 BC AD 1? 50 50-100 100-150 150-200 200-250 Peregrine community municipium colonia conventus 11, 17, 13, 21, 7-10, 15, 18, 22, 23 6 16, 24 12, 19 4, 26 2, 5, 27 29 20 1, 28 30 25 Omitted are nrs. 3 arid 14 which cannot be reasonably dated. Table 4: The Status of the Community by Province Status Tarracon Baetica Mauretan Afr. proc. Sardin. Thrace Italy Peregrine community municipium colonia conventus 17-19, 21-22, 24 20, 23, 26 25 15, 16 4 1 ?3, 5 7-13 6 28 27 29, 30 Omitted is nr. 14. It is noteworthy that there is virtually an equal number of citizen-com- munities and peregrine communities, which suggests that there was a considerable degree of institutional conformity despite the differences in status. This picture is, however, misleading. On one hand, the tabulae which emanate from peregrine communites are all dated within the century following the 40's B.C., and not one can be dated later than A.D. 65 (nr. 12). Indeed, the great majority of them (12 of 15) date to the principates of Augustus and Tiberius. On the other hand the earliest tabula from a citizen-community is dated to the principate of Tiberius (nr. 6). And, although tabulae were employed by such communities throughout the period under discussion, over half (7 of 12) of the dated documents were produced in the 60 years between A.D. 40 and 101. It is probably not a coincidence that the earliest and only tabula patronatus referring to a more extended community, namely, to the conventus Cluniensis (nr. 25) dates to the end of the period under discussion, for it was precisely at TABULAE PATRONATUS 543 this time, the beginning of the third century, when references to such extended clientelae of provincial patrons first appear24. A pattern may also be perceived when the status of the community is considered in its provincial context (cf. Table 4)25 . First, citizen communities which concluded such contracts are to be found in every area discussed here. Secondly, and most noteworthy, is that all the peregrine communities are to be found only in the two Spanish provinces (Baetica and the Tarraconensis) and in Africa proconsularis. In summary, two trends may be observed among the tabulae of provincial origin. First, the earliest and the greatest number of them (15 of the dated 26), refer to peregrine communities found only in the Iberian and African provinces. Moreover, this contractual form was particularly popular among the peregrine communities during the principates of Augustus and Tiberius. Beginning with the principate of Claudius, however, there is a shift and it is citizen-communities, instead of peregrine, who employ this form. Although such contracts were employed by citizen-communities regularly from Nero to the end of this period, they seem to have been particularly prevalent during the second half of the first century. Hence, it is probable that the provincial form of the tabulae, which does not have an Italian origin, was `Romanized' by the peregrine communities of Spain and Africa at the beginning of the principate. And, just as these communities were advanced to citizen-status, so too was the provincial tabula accepted among the typically Roman institutions. In Italy by contrast, there was no formal change in the status of the communities during this period and tabulae appear both later and in a different form. IV. The Individual and His Status The discussion of the status of the individuals mentioned on the 30 tabulae involves 26 names due to the fact that Sentius Caecilianus appears on two tabulae (nrs. 2 and 4) and Silius Aviola on four (nrs. 7-10). As different communities are involved, the general conclusions to be made here are not significantly affected by considering these multiple cases individually or collec- tively26 . The data are summarized in Table 5. " The earliest patronus provinciae is datable to the middle of the 2nd Century, but most date to the period after A.D. 230. See HARMAND, 413, for a list of the provincial patrons and I\hcoLs, art. cit., for the significance. 25 The two Italian communities (nrs. 29 and 30) are not included in this discussion because there were no peregrine communities in Italy at this time and because the Italian tabulae have a distinct evolution; cf. below, p. 553ff. Multiple parrocinia and clientele are discussed below, p. 547ff. 544 JOHN NICOLS Table 5: The Individual and his Status Status of the Individu 1 Province of Community Senators Equestrians Status unknown Tarraconensis 17, 25 23, 26 18-22, 24 Baetica 15 14, 16 Mauretania 2, 4, 5 3 Afr. protons. 6, 11 7-10 13 Sardinia 28 Thrace 27 Italy 29, 30 In 19 of the 30 cases, the status of the individual is apparent in his titulature (e.g., nrs. 2 and 7) or can be deduced from other evidence (e.g., nrs. 1 and 17). Of these 19, 11 are of senatorial status including one woman, Nummia Varia, a clarissima femina (nr. 30) 27 . With the exception of Nummia, however, all the individuals of senatorial status became patrons of communities which lay within their sundry administrative spheres. Indeed, it is highly probable that the patrocinium was extended in response to their administrative achievements, which would be consistent with the general pattern of patronage in the principate28. The remaining eight tabulae which refer to patrons of known status involve five equestrians, three of whom certainly had administrative experience in the provinces in which the respective communities lay (nrs. 1, 7-10 and 12). Again, it is likely that the administrative activity is to be associated with the coop- tation. That C. Terentius Bassus C. f. Fab. Mefanas Etruscus (nr. 23) was of equestrian standing may be deduced from his office of praefectus alae Augustae. His previous connection with the town of Clunia is not known; there is no evidence that the ala Augusta was stationed in Spain during the second century and Etruscus' name and tribal affiliation suggest an Italian origin 29 . It may be that he held the post mentioned on the tabula after holding an unknown office in Spain. Particularly interesting is the case of Q. Licinius Silvanus Granianus, patron of his patria of Baetulo (nr. 26). The name could apply to either the father of equestrian status or to his son, a senator. But, as there is no contemporary parallel for a senator becoming patronus patriae and as this connection is common for equestrians", it is likely that the patronus in question was the equestrian father31. 27 Patronae arc not known in the early principate, cf. HARMAND, 281, who does not note Nummia on his list of patrons. 2S On this question, see HARMAND, 290ff. 29 Conclusions based on the name Etruscus alone are too uncertain, but the tribe Fabia is very rare in Spain. 3? Cf, J. Nicms, Pliny and the Patronage of Communities, Hermes, forthcoming. 31 DRS, Sobre la Tabula Patronatus de Badelona, Ernerita 7 (1939) 156ff. and PIR 2 L 247-8. TABULAE PATRONATUS 545 The remaining 11 tabulae make no mention of any office, local or imperial, and the individual is otherwise unknown. In three of these 11 cases, the name is non-Roman (nrs. 21, 22 and 24) and we may readily assume that the individual was of peregrine status and of local or regional importance. Another three of the unknowns (nrs. 3, 13 and 14) may be senatorial governors or legati. The remaining five (nrs. 16, 18-20 and 28) probably belong to the Romanized upper class; they may have been of equestrian status but were, at most, of only regional importance. When one considers the status of the patron in relationship to the geo- graphical location of the community a clear distinction emerges. Communities in Tarraconensis apparently employed this contractual form to bind individuals of local importance. All other areas employed the relationship in order to secure the support of imperial administrators. More will be said about the varieties of formulae below, but if the 'Italian' form be set aside (nrs. 27, 29 and 30), an extremely interesting and suggestive pattern emerges. Communities located in the area of the former Punic Empire (North Africa, Balearic Islands, Sardinia and Baetica) all entered relationships with individuals who were demonstrably or probably imperial administrators in their respective provinces. In contrast, in the non-Punk interior of Tarraconensis, the relationship is generally (in 8 of 10 cases) concluded with men of local importance. This is not to say that the contract had an Iberian or Punic origin, but such traditions may have influenced the manner in which the institution was practiced. V. The legati A universal characteristic of the tabulae is the stipulation that legati be assigned the responsibility of completing the cooptation. This aspect usually appears in summary fashion at the end of the tabula as, simply: egerunt legati followed by the names 32 . The exact duties of the legati are not given in the inscriptions but surely involved the presentation of the deeretum decurionum (cf. nr. 30, 1. 21 ff.) and an unknown degree of formality connected with the completion of the cooptation process. The names of the legati sometimes reflect the status and ethnic constituency of the communtiy. Two of the seven peregrine communities in Africa proconsularis chose legati with Punic names and the legati of two of the five peregrine communities of Tarraconensis have Iberian names. The legati of all the remaining communities have Latin names. For the 14 communities of citizen-status, this is to be expected. That six peregrine communities also designated legati with Latin names (nrs. 12, 15-19) indicates that these 3z There are variations, both grammatical (agentzbus legato, nr. 5) and formulaic (faciundum coeraverunt, nr. 11). 546 JOHN NICOLS communities had not only adopted the Roman municipal model, but indivi- duals had begun to latinize their names. The number of legati designated to complete the cooptation varies between one and six and, as may be seen on Tables 6.1 and 6.2, does not appear to have been influenced by the status of the community. The most frequent number of legati was two, occuring on 13 of the 27 tabulae on which the number is clear. This number is generally in accord with the number of legati appointed for Table 6: The legati 6.1: Peregrine Communities Frequency 7 6 12 5 15 4 17 3 18 9 2 19 11 7 1 8 22 24 16 10 13 1 2 3 4 5 6 Number of legati 6.3: Peregrine Communities Frequency Number of Magistrates who served as !egad 6.2: Citizen Communities Frequency 7 2 6 3 5 5 4 20 3 23 2 25 27 28 1 26 29 30 1 1 2 3 4 5 6 Number of legati (Omitted are nrs. 4, 14 and 21) 6.4: Citizen Communities Frequency 9 2 8 3 7 5 6 6 5 20 4 23 3 25 2 26 1 29 28 30 0 2 Number of Magistrates who served as legati (Omitted are nrs. 4, 14, 16, 21, 27) TABULAE PATRONATUS 547 other purposes 33 . No other factors, whether temporal, geographical, etc., appear to have been significant in the determination of the size of the legation. A pattern can be identified when the number of magistrates who participated in such legations is considered. Of the 25 tabulae which are sufficiently well preserved to allow such distinctions, 13 make no mention of magistrates at all (cf. Tables 6.3 and 6.4). This pattern becomes clearer when the status of the community is considered: 9 of the 12 citizen communities did not appoint a magistrate among the legati, whereas 9 of 13 peregrine communi- ties named at least one. This suggests that the role of the magistrate was, on this question, more important in peregrine than in citizen communities34. Nevertheless, there is no reason to believe, for example, that, when four legati are named, they are the current quattuorviri or other magistrates. In summary, of all the formulae which are employed, only the one which designated legati occurs on all the tabulae patronatus. This fact suggests that the embassy sent to the prospective patronus was an essential feature of the cooptation. VI. Multiple patrocinia and clientelae That communities coopted a number of patrons and that patrons accepted several clientelae is well established from both literary and epigraphical sources. Indeed, Cicero reckoned it a particular honor that he had been selected the sole patron of Capua . me unum patronum adoptavit, pro Sest. 9). The implication is, however, that most communities had several patrons and vice versa35. The tabulae provide some evidence on this phenomenon. Two patrons, one of senatorial and one of equestrian standing, were coopted by two or more communities in the province in which they held imperial offices. Sentius Caecilianus (nrs. 2 and 4) became the patron of Banasa and of Volubilis when he was legatus Augusti pro praetore ordinandae utriusque Mauretaniae and consul designatus in A. D. 75. Silius Aviola was apparently tribunus militum with the legio III Augusta and praefectus fabrum when, in 28, he became patron of four small communities in Africa proconsularis (nrs. 7-10). There is less evidence that communities used tabulae to distinguish between different patronal relationships established at approximately the same time. Banasa may well have done so with Q. Claudius Ferox and L. Labienus (nrs. I and 3), but the chronology is uncertain. The town of Bocchoris on Menorca coopted M. Crassus Frugi and M. Atilius Vernus as patrons in 10 B.C. and in 31 Cf. Diz. ep. IV, 503ff, 34 The possibility cannot be dismissed that magistrates in citizen communities did indeed function as legati and were not so designated on the tabula. But, as these are official documents, this would seem unlikely. 15 Cf. the alba decurionum which list the various patroni, ILS 6121 and HARMAND, 296ff. 548 JOHN NICOLS A.D. 6, respectively. As the former (PIR 2 L 189) was consul in A.D. 14, the two patroni were contemporaries. There is, then, no reason to believe that the tabula was reserved for sole patrons. VII. The Formulae The several formulae to be discussed here occur both alone and in combination with each other. The variations are summarized in Table 7. Table 7: The Formulae Tabula Number Community Individual 1. patrocinium in se recepit 2. FC 3. P (?) FC 4. P (?) FC (?) 5. eosque patrocinio suo tuendos recepit 6. H, rogarunt uti se in F[C] FC 7. HP FC S. HP FC 9. HP FC 10. HP FC 11. HP FC 12. HP FC 13. H(tessera) P H (tessera) 14. HP HFC 15. HFC 16. HFC 17. P C 18. P FC 19. H renovavit 20. H iunxit, civern et P 21. H (tessera) 22. FC H 23. H 24. HFC H 25. 26. HP FC 27. dignetur patrocirtium 28. HP HFC 29. dignetur CPH (tabula) 30. dignetur patroctniton PC P patronum cooptare H hospitium facere FC = in fidern clientelamque rectpere TABULAE PATRONATUS 549 1. hospitium facere The institution of hospitium is mentioned on more than two-thirds of the tabulae. Most commonly (on 16 of the 30) the formula is, simply, that one or the other party hospitium fecit (nrs. 5-12, 14-16, 22-24, 26, 28). In three cases the expression is hospitalem tesseram fecit (nrs. 13, 21 and 29), in another, hospitium iunxit (nr. 20) and, finally, the expression hospitium renovavit occurs once (nr. 19). The only one of these formulas which appears to be more than a minor variation is the last one, hospitium renovavit 36 . More significant, however, is the very frequency of the institution on the tabulae; namely it occurs on 21 of the 30. This frequency, and only the formula patronum cooptare occurs more often, suggests a very close relationship between the notions of patronage and hospitality. Three of the tabulae (nrs. 19, 21 and 23) mention hospitium alone, without further reference to other related institutions; these are, then, true tabulae hospitales and probably evolved from the tesserae hospita1es37 . The earliest of these tesserae commemorated the guestfriendship established by individuals of different states (cf. Plaut. Poen. 958, 1047). It is not clear when they began to be used to symbolize the relationship between an individual and (another) state, but the so-called tessera Fundana (ILLRP 1068), which is not easy to interpret, may be the earliest evidence of such an agreement. Hospitium originally suggested an equality between the contracting parties. BADIAN has argued, however, that when one party was a powerful Roman individual, the institutions of hospitium and clientela merged and the idea of equality was lost 38 . The evidence of the tabulae on this question is ambiguous. In support of BAD/AN'S hypothesis, it is noteworthy that on only 4 of the 21 tabulae which refer to hospitium do both parties expressly state that they have concluded hospitium with one another (nrs. 13, 14, 24 and 28). Indeed, it is generally the case that it was the community, and not the patron who hospitium fecit. On the other hand, there are cases in which the patron concluded hospitium and also accepted the community in his clientela (nrs. 5, 13-16, 28), which suggests that hospitium was not completely identical to patrocinium and clientela. That hospitium had a meaning independent of patronage is also supported by t he provisions of the lex Ursonensis (ILS 6087). In this law there are separate and distinct requirements and penalties affecting the appointment of a patronus and of a hospes (cf. cc. 130 and 131). In general hospitium, whether in the traditional and equal sense or as an equivalent for clientela, was an important aspect of the agreement between the two parties. '6 The significance is unclear. Most tabulae specify that the relationship applies to the descendants of both panics. 37 See below, p. 555 ff, BAD/AN, 12 and 154. 550 JOHN NICOLS 2. patronum cooptare As has been shown above, the function of the tabula was to commemorate the establishment of a particular relationship, in this case, one in which an individual was given the title of patronus or, in some cases, simply that of hospes. According to the lex Malacitana (ILS 6089), the formulae to be em- ployed in the designation of the patron of a community are two: patronum cooptare and patrocinium deferre. And, indeed, these expressions do occur regularly in the tabulae patronatus. The former, patronum cooptare, is employed in 20 of the tabulae (nrs. 1-4, 7-14, 17-8, 20, 25-6, 28-30), and patrocinium deferre, or, more frequently, patrocinium recipere, is employed on four (nrs. 1, 5, 27, 30), and both expressions together on two (nrs. 1 and 30)39 . Altogether, these formulae occur on 22 of the 30 tabulae. Theoretically, the title patronus implies a status or rank superior to that of the other contracting party. In cases when the relationship was initiated by deditio following a military defeat, the superior status of Rome and of the patron would be manifest. When, however, the relationship was initiated by applicatio and the community was a respected ally or even a colonia civiurn Romanorum, the question of superiority becomes clouded 40 . Nevertheless, when a community takes a patron and describes itself as being in his clientele (which occurs on over half of the tabulae), then it admits to an inferiority in status, even if the two ranks are not exactly complementary. Traditionally, communities used the title of patronus in order to secure the good will and the services of a powerful outsider. And, as MOMMSEN has noted, it was probably considered an inappropriate designation when applied to fellow-citizens of a community. During the principate, however, communities began to coopt their own citizens as patroni. Patriae extended this title first to their citizens of equestrian status (in the first century) and, thereafter, to those of senatorial'". The tabulae do not provide much information on this process. In only one document is it implied that the patron, in this case a clarissima femina (nr. 30), achieved the honor in his/her patria. The question must remain open in regards to the Spanish provinces (nrs. 14-26), for nothing at all is known about the origins of most of the patroni. In Africa, however, the situation is clear. All the patroni referred to on tabulae emanating from these communities were imperial administrators, none of whom is known to have come from the respective province. There is, however, no reason to think that such tabulae were employed exclusively to coopt outsiders. 3. in fidem clientelamque recipere When the patron plays an active role in the conclusion of the relationship, which is the case in 23 of the 30 tabulae, he generally acknowledges that he has 39 1LS 6089, c. 61. Cf. j. N[coLs, Patronum cooptare, patrocinium deferre, ZSSt 93 (1980). 4" On these forms, see BADIAN, 2 ff. 41 On this question, see NICOLS, Pliny and the Patronage, cit. TABULAE PATRONATUS 551 received the community in his clientel (ens in fidem clientelamque suam suorumque recepit). This formula is used in 19 of these 23 tabulae. Of the remaining four in which the patron is active, the word patrocinium is employed twice as a clear substitute for fides clientelaque (patrocinium in se recepit, nrs. 1 and 5) and twice hospitium alone is admitted without any mention of patrocinium or clientela (nrs. 13 and 21) 42 . It is difficult to decide whether clientela was implied in these two cases 43 . In nr. 13, this would be a reasonable assumption as the community, for its part, coopted the individual in question as patronus. In nr. 21, however, there is no such statement and it may well be that hospitium alone was intended. Additionally, it should be noted that in three cases, clientela is mentioned without reference to fides (nrs. 17, 29 and 30). In general, 21 of the 30 tabulae record the establishment of clientela and usually do so in combination with fides. The meaning of these words has been discussed at length elsewhere. GELZER and BADIAN have defined fides as a variety of relationships based on trust and on the performance of mutual and continuous services. When the fides-relationship exists between parties of unequal status, then clientela (in fide alicuius esse) is said to exist. In this relationship, the inferior party, or client, performs officia and the superior confers beneficia. It is noteworthy, however, that though communities readily acknowlege that they are in the clientela of a patronus, they never refer to themselves as clientes. Perhaps the stigma attached to this word was too great44. 4. Variations of Formulae In tabulae nrs. 1 and 5, there is a significant variation of formula which suggests much about the meaning of in fidem clientelamque. In these inscriptions, precisely where the patronus usually states that eos in fidem clientelamque suam suorum recepit, the words patrocinium in se recepit and eos patrocinio suo tuendos recepit are inscribed. Here it is clear that patrocinium, on one hand, and fides clientelaque, on the other, are complementary relationships. Another significant variation, not of formula, but of contents, involves nrs. 22 and 24. The first of these two reads: Tillegus . . . hospitium fecit cum Lougeis Caste llanis Toletensibus sibi uxori liberis posterisque suis eumque uxorem liberosque eius in fidem clientelamque suam suorumque in perpetuo Castellanei Tolentensis receperunt .. This tabula appears to record an agreement by which Tillegus and his family came into the clientela of a community. As it would seem a priori unlikely that a community would record such an event, it is tempting to argue that there must 42 See below, p. 555ff. 43 See above, p. 549ff. and below, p. 553ff. 44 GELZER, 76 and BADIAN, 6-7. 36 ANKW 11 13 552 JOHN NICOLS be a mistake here, that perhaps the Toletenses had an imperfect knowledge of Latin and became confused about the vocabulary of patronage. This interpretation is unlikely for two reasons. First, the language of the tabula is clear and consistent. Secondly, this kind of agreement is not unique: nr. 24 records a similar transaction: Amparamus . . . hospitium fecit cum civitate Maggavensium sibi liberos liberiti]sque posterisque Buis. Eumque liberos libertos posterosque eius ornnis Maggaveses in hospitium fidem clientelamque swam suorumque recepe(un)t eademque condicione esset qua civis .. . The language here is not at the same level as in nr. 22, but the intention is clear enough: Amparamus is being taken into the clientela of the `Maggaveses'. It would appear then that communities in the Tarraconensis did on occasion serve as 'patrons' of individuals. It is not easy to account for this reversal of the usual roles, but it may be that these tabulae refer to the extension of special rights to Amparamus and Tillegus which might have provided them with certain privileges in the community under the 'protection' of the local government. Indeed, nr. 24 expressly mentions the connection between hospitium, fides and clientela on one hand, and citizenship on the other. If this interpretation is accepted, then the tendency of communities to take individuals in their clientela may reflect a traditional Iberian idea of hospitium merging, in an uneven manner, with the Roman notion of clientage. This hypo- thesis, namely that the tabulae served a variety of purposes in Tarraconenis, is supported by three considerations. First, such agreements were concluded in the least Romanized parts of Spain in the period following the Augustan conquest. Secondly, though the language of the tabulae is homogeneous in other provin- ces, it is remarkably varied in Tarraconensis. And, thirdly, only the Tarra- conensian tabulae mention hospitium either alone (nrs. 9, 21 and 23) or in con- nection with clientelae (nrs. 22 and 24) but without any reference to patronatus. 5. The Combination of Formulae When the communities are active on the tabulae, which occurs on 26 of the 30 tabulae, they tend to combine formulae. Such combinations, and hospitium facere . . . patronum cooptare is the most common, are found on 15 of these 26 inscriptions. There is, however, considerable regional variation. The com- munities of Mauretania simply coopt a patron and make no mention of any additional honors. In contrast, the communities of Africa proconsularis always combine hospitium facere and patronum cooptare. In Tarraconensis, there is considerable internal variation in which one, two or even three (cf. nr. 20) formulae are employed. In general, the tabulae most frequently associate the complementary notions of patronage and clientela which make the superiority of one party to the other clear. Hospitium is also pervasive, but its implications are not always clear. Theoretically, hospitium was concluded between two parties of equal or TABULAE PATRONATUS 553 comparable status and mutuality was stressed. Hence, when hospitium was offered by one party, we would expect it to be reciprocated by the other. In practice, however, this mutuality was not respected. All eight of the contracting communities of Africa proconsularis, for example, offer hospitium, but only one patron (the earliest and the least representative, nr. 13) responds in kind. This would appear to suggest that the offer of hospitium belonged especially to the duties of the client-community and implied a lesser kind of dientela. Indeed, BADIAN has argued that the Roman nobilis of the Republic tended to treat his hospites and clientes in the same manner 45 . In the principate, however, hospitium facere might better be understood as equivalent to the conferral of the Tura hospitii (cf. Plin. ep. III, 4.5) or rights of public hospitality. In their discussion of the tabulae, CAGNAT and HARMAND consider the following to be the standard formulation concluded by the two parties: hospitium facere, patronum cooptare//hospitium facere, in fidem clientelam recipere46 . This notation is misleading, however, for it occurs on only two of the tabulae under discussion (nrs. 14 and 28). Generally speaking, standardization only developed in Africa proconsularis where six (nrs. 7-12) of the eight communities employed the same combination of formulae. Moreover, there is simply too much variation in the combination of formulae to allow the CAGNAT- HARMAND proposal to be considered the standard formulation. Rather, it should be understood as the summation of the formulaic possibilities. If a basic formulation is to be accepted, it should be: patronum cooptare//in fidem clientelamque recipere, which occurs on 14 of the 19 tabulae involving mutual and reciprocal agreements. Nevertheless, the high degree of irregularity in the combination of formulae suggests that the communities outside Africa proconsularis combined the individual formulae according to the requirements of the particular situation. VIII. Varieties and Origins of the tabulae 1. The Varieties HARMAND, elaborating on the observations of CAGNAT, divides the tabulae patronatus into three groups 47 . The first, in the 'Italian' form, consists of tabulae which reproduce the prods-verbal of the meeting during which the local municipal senate resolved to coopt a patronus (nrs. 27, 29 and 30). Because these are formally decreta decurionum, he does not consider them to be true tabulae patronatus. The second group is made up of tabulae in which two parallel and 45 BATMAN, 154. 46 R. CAGNAT, Cours d'Epigraphie latine, Paris 4 1914 (reprint Rome 1964), 330 and HARMAND, 336. 47 Ibid. 36" 554 JOHN NICOLS reciprocal formulae arc recorded by which one party is coopted patron and the other is received into the appropriate clientela (e.g., nrs. 2, 7 and 14). These are `true' tabulae patronatus. The third group consists of tabulae which record the formula(e) used by only one of the two parties (e.g., nrs. 5, 15 and 21). These formal distinctions are reasonable, but as all three groups serve to commemorate the same event and more or less use the same vocabulary, it is inappropriate to label only one group les ,veritables tabulae. 48 . That the formal appearance varies reflects rather the diverse origins and development of the document in various parts of the empire. In order to understand this evolution, it is necessary to consider what theoretical and legal considerations affected the initiation of the relationship. 2, The Origins and Evolution of the tabulae Of the four means by which a patronal relationship could be established49, only applicatio is relevant to this discussion, for it alone is based on a completely voluntary decision. The degreee to which the applicatio was formally expressed is a matter of some dispute. PREMERSTEIN notes the various formulae used to establish the relationship"; BADIAN, on the other hand, discounts the formal element, observing that a simple request for aid could be sufficient51. Nevertheless, the initiation of the relationship between patroni and their client-communities was, in the principate, accompanied by a certain amount of ritual. Two of the tabulae are explicit about the procedure to be followed once the decuriones had officially, as specified by the municipal charters, voted to coopt a patronus: placere conscriptis legatos ex hoc ordine mitti ad T. Pomponium Bassum clarissimum virum qui ab eo impetrent in clientelam arnplissimae domus suae municipium nostrum recipere dignetur patronumque se cooptari tabula hospitali incisa hoc decreto in domo sua posita perrnittat ... (nr. 29) That is, in accordance with the decretum decurionum, legati were to be chosen from the decuriones who were to be sent to the prospective patron; they were to petition him to accept the community in his clientela. And, when he had done so, to present him with a bronze tablet commemorating the cooptation. This tabula contained no formal response from the patron, but as it was set up apparently in Bassus' house in Rome, it is probable that he did accept the Ferentini in his clientela. The degree to which this process of applicatio and acceptance in clientelam was formalized varies in time and by circumstances, but it is likely that in many, " H ARMAND, 336. 49 BADIAN, 2ff. 5" A. VON PREM FRSTEIN, Art. Clicntcs, in: RE IV (1900) 32. 51 BADIAN, 7-9. TABULAE PATRONATUS 555 if not all, cases, some kind of exchange of tokens took place. The tabulae patronatus are the most obvious, but certainly not the only tokens used. a) The tabulae and the tesserae hospitales52 There is good reason to believe that tesserae hospitales provided an important model in the development of the tabulae patronatus. Tesserae, or symbolae, are essentially tokens which served a variety of purposes in the ancient world. In the Republican era, they usually commemorated some event. Though various materials were employed, it is significant for this discussion that all tesserae hospitales are, like the tabulae, tokens in bronze and that they are both portables. Plautus makes the earliest reference in Latin to such tesserae, observing that they were exchanged by host and guest (Poen. 958, 1047). DEGRASSI included six of them in his collection of Republican inscriptions (ILLRP 1064-9). The earliest of these are in the form of animals, fish and ramheads predominating, with fairly simple texts mentioning little more than the names of the hospites. One early tessera, however, has a longer text recording not only the establishment of hospitium, but also, and this is noteworthy, a deditio m fidem such as is found on the later tabulae patronatus. This connection between tesserae and tabulae is confirmed by two statements in the tabulae. First, nr. 13 mentions that the relationship between patronus and client-community involved the exchange of tesserae hospitales by both parties. Second, nr. 29 refers to the document of cooptation which was presented to the patronus as a tabula hospitalis. If the derivation of the tabula patronatus from the tessera hospitalis is accepted, then a number of the characteristic features of the former can be explained. First, the prominence of hospitium in the tabulae may be accounted for. Second, the fact that the tabulae, like the tesserae, are always in bronze and are portable can be explained. Third, and most significant, the complementary or double-form (what HARMAND calls ,,synallagmatique. 54 ) in which each of the two parties is active, may have derived from the manner in which hospitium was concluded. In the latter case, both parties exchanged tokens commemorating the event or split a token between them (cf. ILLRP 1068). As the statements of the two parties on the tabulae are usually neatly separated, it may be that the double, or complementary, form was originally intended to be split in the same manner. In this interpretation, the one-sided tabulae (cf. Table 6, nrs. 5, 15-16, 19-21, 23, 25) may preserve one aspect of the original practice, while those with mutual and complementary statements may have resulted from the wish of each party to preserve a copy of the complete and reciprocal agreement. sz As the Italian communities gave their patrons tabkaae hearing the decromm of cooptation and as decreta were usually preserved on wood tablets, it is possible that the token was wooden, cf. below, p. 556ff. 53 On the size, see the Introduction. 556 JOHN NICOLS b) Decreta decurionum The charters of Urso and Malaga specify that the cooptation of a patronus civitatis be carried out by a decretum decurionum and mention a number of conditions which had to be met before the cooptation became valid. The influence of such regulations on the form of the tabulae is irregular. Three tabulae (nrs. 27, 29 and 30) are, in fact, copies of these decreta, but, as not one of them is datable earlier than A. D. 82, they could have had little influence on the development of tabulae in general s '''. Of the remaining 27 tabulae, only one (nr. 1, dating to 162) expressly mentions that the cooptation was completed ex decreto splendidissimi ordinis. This should not, however, be interpreted to indicate that the other cooptations were completed without the required decreta, for all the tabulae indicate that the community was acting officially 55 . Indeed, the purpose of the tabula was not to record the wording of the decree, but, rather to commemorate the moment when the clientele was accepted by the patronus 56 (cf. nr. 29). MommsEN in his discussion of senatus consulta, notes that the publication of such decrees on bronze tablets was not usual 57 . Hence, it is likely that the decrees mentioned or implied by the tabulae patronatus belong to the category of ?internationale Urkunde". These decrees, which recorded international agreements as well as privileges for individuals, were published on bronze tablets 58 . It is likely then that the contractual form was, initially, appropriate when peregrine communities coopted powerful Romans, but, in time, came to be employed by citizen communities as well. The influence of the SC or decreta decurionum on the form of the tabulae was then both direct and indirect. First, the form and the position of the consular date of cooptation certainly derived from the usual practice of the SC. Secondly, the tabulae patronatus are closely related to the published form of international documents confirming privileges granted to individuals; here, the manner of publication and the subject matter are relevant. And, finally, the decreta decurionum clearly had an influence on the form of the statement of the contracting community. The origin of the form of the patron's response is, however, more obscure. The tesserae hospitaks provide only a partial explanation, for they concern the establishment of hospitium. Another influence, affecting the statement of the patron in the double-form tabulae or in the one-sided tabulae when the patron alone acts, may be the decretum magistratuum. c) Decreta magistratuum Decrees of magistrates on a number of subjects are known from the periods of the Republic and of the principate (e.g., ILLRP 514 and 1LS 5941, 54 HARMAND, 336. 55 Again, this assumes that wooden tablets were not used for this purpose earlier, cf. n. 51. 56 Cf. nr. 29. MOMMSEN, RStr. 111, 1014. 58 MOMMSEN, RStr. 1, 25511. and especially 256, n. 4. TABULAE PATRONATUS 557 respectively)". What is relevant for this discussion is that they are documents inscribed on bronze tablets and that they begin with the name of the relevant magistrate in a manner which is comparable to that employed for the statement of the patronus on the tabulae. The most dramatic example of this is nr. 13, which in form is very similar to the decree of Aemilius Paulus from 189 B.C. (ILLRP 514). In this interpretation, the municipal decree of cooptation would have been answered, when appropriate, by a magisterial decree. This hypothesis finds some confirmation in practice; four tabulae involve an individual acting alone (nrs. 5, 15, 15, 21) and on two of them, the magistracy of the patronus is mentioned prominently (nrs. 5 and 15). The influence of the magisterial decree on the form of the tabulae would have been strongest at an early (republican) stage of development; namely, at a time when most patrons were powerful senators who came to the honor through their administrative activities. In the late republic and early empire, regulations were introduced which forbade communities to honor their current administrators and at that point, the magisterial decree, as the legal basis of the statement of the patron, was lost 611. The form, however, persisted. In summary, there were a number of factors, theoretical and concrete, which influenced the development of the tabulae patronatus. The theoretical basis was clearly the idea of applicatio, which to a varying decree, had to be acknowledged formally by the patron. A number of bronze instrumenta also influenced the form the tabulae took. In the archaic period of Roman history, the establishment of a friendly relationship between two parties (reckoned individually or collectively) was probably commemorated by the exchange of tokens or tesserae hospitales. Initially, these tesserae had little, if any, verbal content beyond the mention of the name of the party and the honor accorded (cf. ILLRP 1064-7). During the second century B.C., the tesserae began to receive more elaborate statements (ILLRP 1068) and, in consequence of this elaboration, required a more spacious format, namely the tabula (ILLRP 1069 = nr. 13). Though this process was completed by the end of the republican period, the influence of the tesserae bospitales remained strong in the evolved tabulae. This is demonstrated, first, by the continuing prominence of hospitium and, secondly, by the careful division of the statements of the two parties so that the tabulae, like the early tesserae, could (theoretically) be split in the traditional manner, allowing each party to retain his respective ?half". The arrangement of the formulae on the tabulae derives from the decretal forms employed by a senate (for the statement of the community) and by a magistrate (for the statement of the individual). It is important to recall, however, that the 'legal' basis of the latter was cancelled by regulations introduced in the late republic, while that of the former was strengthened by the same rules. Generally speaking, the mutual form, probably under the influence of hospitium, evolved during the Augustan era and was the rule in the western provinces. The alternative form, which reproduced the decretum decurionum of ' 9 On these decreta, see HESKY, Art. Decretum, in: RE IV (1901) 2291. 60 On this question, J. Nrcots, Zur Verleihung Offenthicher Ehrungen, cit. 558 JOHN NICOLS cooptation, developed in Italy during the principate and seems to have had some influence in the Danube area. d) Wooden tabulae One difficult question remains to be considered, namely, the possibility that wooden and not bronze tablets were the usual tokens employed in the cooptation of a patron. Two considerations are relevant to this question. First, decreta decurionum, to which the tabulae are closely related, were usually recorded on such wooden tablets and, second, the otherwise inexplicable lack of such tokens from Italian communities before A.D. 100 can be explained by the perishability of that materia1 61 . Unfortunately, there is no way of testing this hypothesis for no decretum recorded on wood survives and there is no reference, direct or indirect, to wooden tablets being used to commemorate a cooptation. Nevertheless, the theory is an attractive one. Patronage of communities was common throughout the western part of the Empire and it is reasonable to believe that tokens of some kind were used to formalize the agreement and that at least some of them would have been taken to Italy by the respective patrons. That such tokens do not survive from transalpine communities may then be due to the fact that the material employed was more perishable than bronze. If this theory is accepted, two further questions arise. First, why was bronze employed on some occasions and, second, how did the choice of material affect the text of the document. Concerning the first question, it is evident that it cost more to produce a bronze tablet than it did to produce a wooden one. Hence, the choice of bronze might reflect the appreciation of the community following a particularly generous benefaction. This is, however, unlikely for the tabulae are vague about beneficia conferred and there were other, more dramatic ways to recognize special achievements 62 . It is more probable that the choice of material was influenced by other factors, most notably, by a local tradition of using bronze for such purposes. As to the second question, it is likely that there was a relationship between material and text, but there is no reason to believe that it was a causal one. In Spain and North Africa, local tradition, especially the manner in which hospitium was concluded, dictated the arrangement of the formulae and the choice of material. Though the evidence for the following is at best indirect, I would like to suggest that the Italian communities (and also the transalpine ones in imitation) initially provided the designated patron with a wooden tablet bearing the decree of cooptation. During the late Republic and early principate they were introduced (or: re-introduced) to the idea of using bronze as their citizens, men like Silius Aviola (nrs. 7-10), returned from provincial appointments bearing tabulae patronatus of a different sort; namely, with a reciprocal text and one inscribed 61 Cf. VIII, B. 2 and II, respectively. 62 On this question, see Mc()Ls, Zur Verleihung Offentliclier Ehrtingen, cit. TABULAE PATRONATUS 559 on bronze. Though the reciprocal form did not succeed in replacing the decretal form, the idea of using bronze was apparently appealing and seems to have been used with increasing frequency in Italy after A.D. 100. Summary of Conclusions 1. Tabulae patronatus are bronze instrumenta employed by communities to commemorate the cooptation of a patron. The 'Italian' variety preserves a copy of the decree of cooptation; the provincial variety, by contrast, is reciprocal allowing both parties to be active. 2. The geographical and temporal distribution of the tabulae indicates that they flourished particularly in Spain and North Africa during the early principate. 3. The provincial or reciprocal variety was initially employed by peregrine communities and later `Romanizeds. 4. The status of the patron honored by a tabula varies from province to province. In Tarraconensis, the recipient was more likely to be a man of local or regional importance, in Africa he was generally an imperial official. 5. The appointment of legates to complete the cooptation is characteristic of the tabulae. 6. Despite considerable variation, the basic formula appearing on the tabulae was patronum cooptare/in fidem clientelamque reopere. Hospitium facere was frequently employed, but in an inconsistent manner. 7. Tesserae hospitalis and various decreta (of town councils and magistrates) influenced the development of the tabulae in form and content. 8. That wooden tabulae were also employed for this purpose is reasonable, but cannot be proven. 560 JOHN NICOLS Appendices Appendix 1: Key to publication of the tabnlae Tabula Number Publication I. AE 1948, 115. 2. AF, 1941, 79. 3. AE 1954, 259. 4. AE 1969/70, 747 5. CIL VIII, 8837 = ILS 6103 6. AE 1913, 40 7. CIL V, 4921 = ILS 6099a 8. CIL V, 4922 = ILS 6099 9. CIL V, 4919 = ILS 6100 10. CIL V, 4920 11. CIL VIII, 68 = ILS 6095 12. CIL VIII, 69 13. CIL VIII, 10525 = ILS 6094 14. AE 1969/70 746 15. AE 1962, 287 16. CIL 11, 1343 = ILS 6097 17. AE 1957, 317 18. CIL II, 3695 = ILS 6098 19. CIL II, 2958 = ILS 6104 20. CIL II, 2960 = ILS 6108 21. CIL II, 5763 = ILS 6096 22. AE 1961, 96 23. CIL II, 5792 = ILS 6102 24. AE 1967, 239 25. CIL VI, 1454 = II S 6109 26. AE 1936, 66 27. CIL VI, 31692 = ILS 6135 28. CIL X, 7845 = II.S 6107 29. CIL VI, 1492 = ILS 6106 30. CI1, IX, 3429 = ILS 6110 TABULAE PATRONATUS 561 Appendix 2: The Provenance of the tabrslae Tabula Number Date Community Province Where found 1. 162 Banasa Mauretania Banasa 2. 75 Banasa Mauretania Banasa 3. 162 Banasa Mauretania Banasa 4. 75 Volubilis Mauretania Volubilis 5. 55 Tupusuctu Mauretania Numidia (Constantia) 6. ca. 17 Assuras Afr. procons, Italy (Rome) 7. 28 Apisa Maius Afr. procons. Italy (Brixia) 8. 28 Siagitana Mr. procons. Italy (Brixia) 9. 28 Themetra Afr. procons. Italy (Brixia) 10. 28 Thimiliga Afr. procons. Italy (Brixia) 11. 12 B.C. Gurza Afr. procons. ex Africa 12. 65 Gurza Afr. procons. ex Africa 13. ca.45 B.C. Curubis Afr. procons. Tunis 14. ca. 20 Isttirgi Baetica Mauretania (Volubilis) 15. ca. 20 Munigua Bactica Munigua 16. 5 Lacilbula Baetica Lacilbula 17. ca.12 B.C. Bocchoris Tarraconensis Bocchoris 18. 6 Bocchoris Tarraconensis near Bocchoris 19. 57 Pompaelo Tarraconensis Pamplona 20. 185 Pompaelo Tarraconensis near Pamplona 21. 2 Pallantia Tarraconensis near Pallantia 22. 28 Castellum Toletum Tarraconensis Tarraconensis (Caurel) 23. 40 Clunia Tarraconensis Clunia 24. 14 Civitas Maggavensium Tarraconensis Tarraconensis (Pisoraca) 25. 222 Conventus Cluniensis Tarraconensis Italy (Rome) 26. 98 Baetulo Tarraconensis Baetulo 27. 84 Develtum Thracia Italy (Rome) 28. 158 Usellis Sardinia Sardinia (Cagliari) 29. 101 Ferentinum Italia Italy (Rome) 30. 242 Peltuinum Italia Italy (Peltuinum) Inhalt Vorwort ........................................................................................................ VII?VIII RECHT (NORMEN, VERBREITUNG, MATERIEN) Band II. 13: GUARINO, A. (Napoli) Gli aspetti giuridici del principato ............................................................ 3-60 GUARINO, A. (Napoli) La formazione dell'editto perpetuo .................................................. 62-102 BAUMAN, R. A. (Sydney) The `Leges iudiciorum publicorum' and their Interpretation in the Republic, Principate and Later Empire ............................................ 103-233 GIUFFRE, V. (Napoli) `Militum disciplina' e 'ratio militaris' ............................................... 234-277 RADITSA, L. F. (Annapolis, Md.) Augustus' Legislation Concerning Marriage, Procreation, Love Affairs and Adultery .......................................................................... 278-339 GRELLE, F. (Bari) La `correctio morum' nella legislazione flavia .............................. 340-365 NARDI, E. (Bologna) Aborto e omicidio nella civilta classica ............................................ 366-385 AMELOTTI, M. (Genua) Notariat and Urkundenwesen zur Zeit des Prinzipats ................. 386-399 SIMSHAUSER, W. (Augsburg) Untersuchungen zur Entstehung der Provinzialverfassung Italiens 401-452 CAMODECA; G. (Napoli) Ricerche sui curatores rei publicae .................................................. 453-534 NicoLs, J. (Stanford, Cal.) Tabulae patronatus: A Study of the Agreement between Patron and Client-Community ...................................................................... 535-561 WACKE, A. (K?ln) Die potentiores in den Rechtsquellen. EinfluE and Abwehr gesell- schaftlicher Ubermacht in der Rechtspflege der Romer ................. 562-607 BIRLEY, E. (Hexham) Law in Roman Britain ...................................................................... 609-625 MARSHALL, A. J. (Kingston, Ontario) The Survival and Development of International Jurisdiction in the Greek World under Roman Rule ...................................................... 626-661 RABELLO, A. M. (Jerusalem) The Legal Condition of the Jews in the Roman Empire ................. 662-762 WOLFF, H. J. (Freiburg i. Br.) Riimisches Provinzialrecht in der Provinz Arabia (Rechtspolitik als Instrument der Beherrschung) ...................................................... 763-806 KATZOFF, R. (Ramat Gan, Israel) Sources of Law in Roman Egypt: The Role of the Prefect .............. 807-844