¹ On the Lex Irnitana, see González, Julián, ‘The Lex Irnitana: a new copy of the Flavian Municipal Law’, JRS 76 (1986), 147–243 and Pls v-xxii, with an edition of the Latin text, a commentary by himself, and a translation into English and two Appendixes by Prof. M. H. Crawford. Besides the papers of A. d'Ors and T. Giménez-Candela quoted at length in his paper (p. 147 n. 1), we must expect many more discussions (cf. D. Johnston, above pp. 62–77.Google Scholar

² Cf. González, 181 for the text and 237–8 for the commentary. He gives no explanation as to why he grouped the text of the ‘letter’ and of the ‘Faciendum curaverunt’ formula in a single paragraph. The fact that they are separated by quite a wide space on the bronze tablet can be confirmed from Pl. II ( = González pl. xxii)

³ González and Crawford have finally chosen to print late, but even if, as is probable, such was the spelling on the bronze tablet, it must be a mispelling, for late is incomprehensible: the expression late scire does not exist in Latin, neither is it possible to find any equivalent for scire with an adverb indicating extent. As kindly suggested by Dr Lintott, the expression late comprehensum would be a little easier, and would make good sense (‘I know that some marriages are extensively covered by the law’). However, not only was I unable to find any other example of qualification of the very common expression lege comprehensum (cf. Sen., Contr. 10. 1. 9, Marcell., Dig. 18. 1. 60, Papin., Coll. Mos. 4. 8. 1, CIL v. 7637. 7 and viii. 24616, etc.), but the grammatical construction is awkward: it is very difficult to explain why late would have been put so far away from the verb to be qualified. Therefore one is obliged to follow Millar and Honoré, and to consider late as a mispelling for lata: such an expression may be abundantly found in the Digest (1. 2. 2. 2, 1. 2. 2. 16, 1. 22. 3, 4. 5. 5. 1, 48. 8. 7. pr., 49. 15. 2. 8, 50. 7. 18. pr., etc.) and can be justified on palaeographical grounds, if one considers the problem of transcription from a text written in the bureaucratic cursive of the imperial chancery; cf. Mallon, J., De l'écriture (1982), 188Google Scholar col. 10 for a presentation of the writing used by the only remaining exemplar of the original of an imperial rescript. Further, this restoration makes good sense, since the Lex Irnitana describes itself as a lex rogata in ch. xxxi, 1. 43.

⁴ Cf. González (1986), 238, who, certainly wrongly, relates this remark to his hypothesis that the law and the letter were granted during the same embassy. But see n. 34.

⁵ This arrangement seems to indicate that the rubric LXXXXVII and the litterae of Domitian had been conceived as a unity in the inscription, separate from both the sanctio rubric (LXXXXVI) and the ‘Faciendum curaverunt’ formula. Cf. below/p. 85.

⁶ On this point, see W. Williams, ‘Epigraphic Texts of Imperial Subscripts: a Survey,’ ZPE 66 (1986), 181–207, esp. 195, where he defends his opinion on the basis of the heading (sacrae litterae) of the Severan constitution of 204 (quoted and analysed below, n. 12) or the petition of the Skaptopareni in 238 (Syll. ³ 888. 1. 102) calling the imperial subscript θεία γράμματα. As regards edicts, cf. Oliver, J. H., Aspects of the Civic and Cultural Policy in the East (Hesperia Suppl. xiii, 1970), 35–7Google Scholar; and on the notion of litterae in this context, dell'Oro, A., ‘Mandata’ e ‘litterae’ (1960), who declares, p. 79Google Scholar: ‘Litterae in senso generico è ogni documento scritto destinato dall'autore ad essere portato a … conoscenza di altri’.

⁷ Cf. Wilcken, U., ‘Zu den Kaiserreskripten’, Hermes 55 (1920), 1–42Google Scholar; Wenger, L., Die Quellen des römischen Rechts (1933), 428Google Scholar; S. Brassloff, RE vi (1900), 204 S.V. ‘Epistula’ and recently Honoré, A. M., Emperors and Lawyers (1981), 34–5Google Scholar.

⁸ The change in practice is well known: during the first century A.D. a perpetuation of the republican formula Ἐρρωσθε (cf. Sherk, R. K., Roman Documents from the Greek East (1969), 35Google Scholar); and from the second century onwards the adoption, certainly under Hadrian, of a new standard translation, Εὐτυχεῖτε. See, still, Lafoscade, L., De epistulis aliisque titulis imperatorum magistratuumque Romanorum quas ab aetate Augusti usque ad Constantinum graece scriptas lapides papyrive servaverunt (1902)Google Scholar; Martin, F., La documentación griega de la cancilleria del emperador Adriano (1982)Google Scholar, does not seem to realise that there is a problem.

⁹ For such extracts with the heading pars epistulae, see CJ 10. 5. 1 (Alexander), 8. 40. 13 (Gordian), 9. 41. 4 (Caracalla) and 10. 61. 1 (pars edicti Antonini). It is true that in the Digest—for example, Florentinus, Dig. 1. 24 (Trajan to Statilius Severus) and Callistratus, Dig. 21. 2 (Hadrian to the concilium of Baetica)—the quotations are less accurate. But one must not forget that such a work is only a compilation from the second- and third-century jurists, who could—but did not need to—adopt in their writings the formal system of abbreviations usual in their time. However, see Modestinus, Dig. 27. 1. 6. 2, who has adopted nearly the same heading (κεφαλαῖον ἐπιστολῆς) as the Cyrenaican inscription recently re-edited by Reynolds, J., JRS 68 (1978), 113–14, ll. 13, 25, 69.Google Scholar On the notion of exemplum, see Kübler, B., ‘Pariculum, Exemplum’, Studi Riccobono 1 (1936), 437–53Google Scholar, and Palazzolo, N., ‘Le modalità di trasmissione dei provvedimenti imperiali nelle province (II–III sec. d.C.)’, Iura 28 (1977), 40–94, esp. 68–72.Google Scholar

¹⁰ On this point, cf. W. Williams, ‘Formal and Historical Aspects of Two New Documents of Marcus Aurelius’, ZPE 17 (1975), 37–56 criticizing on this basis the proposals of Oliver, op. cit. (n. 6), 35–7.

¹¹ On the libellus procedure, beside the items already quoted, cf. Williams, W., ‘The Libellus Procedure and the Severan Papyri’, JRS 64 (1974), 86–103Google Scholar; ‘Individuality in the Imperial Constitutions: Hadrian and the Antonines’, JRS 66 (1976), 67–83, and ‘The Publication of Imperial Subscripts’, ZPE 40 (1980), 283–94 (in response to the paper of d'Ors, A. and Martin, F., ‘Propositio Libellorum’, AJPh 100 (1979), 111–24)Google Scholar. See also D. Nörr, ‘Zur Reskriptenpraxis in der höhen Prinzipätszeit’, ZSS 98 (1981), 1–46, and Millar, F., The Emperor in the Roman World (1977)Google Scholar (= ERW), esp. 240–52 and 537–49. Further, see now Coriat, J. P., La législation des Sévères et les méthodes de création du droit imperial à la fin du principat. i. La technique législative des Sévères. ii. La palingénésie de la législation des Sévères (thèse d'Etat en droit, Paris, 1985), 379–405Google Scholar (on the libellus procedure) and 417–551 (its objects and sense in Roman imperial law).

¹² The text quoted here is that of the Parian inscription, which has a second version in Greek (Syll. ³ 881 =CIL III. 14203. 8–9). Other copies in Latin: Mirtatz in Pentapoli (Drew-Bear, Th., Chiron 7 (1977), 355Google Scholar); Satala (TAM v. i. 607); two from Ephesus (I.Ephesos ii. 207–8); Pisidian Antioch (Jones, C. P., Chiron 14 (1984), 93–9)Google Scholar; and one of unknown provenance (L. Robert, BCH 102 (1978), 435–7). Copies in Greek: together with their Latin original text, in Pisidian Antioch (Jones, loc. cit.), and another one of unknown provenance (Robert, loc. cit.), quite different in its formulation, and therefore important for its demonstration that there never existed any official translation of the subscripts, as wrongly believed by Martin, op. cit. (n. 8), 324–7. Since there are so many exemplars of the same text, the restorations on the Parian inscription must be considered as completely sure.

¹³ For the Severan rescript, see most recently Williams, art. cit. (n. 6), 197–8.

¹⁴ For full discussion of this point, see below, p. 83.

¹⁵ The new subscript must be inserted between the subscript of Octavian to the Samians (Reynolds, J., Aphrodisias and Rome (1978), no. 13Google Scholar), and the three subscripts collected on the bronze tablet from Vardacate (see n. 31 below).

¹⁶ For discussions of the imperial communication system, see Riepl, W., Das Nachrichtenwesen des Altertums mit besonderer Rücksicht auf die Römer (1913), 193–240Google Scholar; Ramsay, W. M., ‘The Speed of the Roman Imperial Post’, JRS 15 (1925), 60–74Google Scholar; M. Amit, ‘Les moyens de communication et de défense de l'Empire romain’, PP 20 (1965), 207–22. For a full discussion of the evidence and its implications, see Millar, F., ‘Emperors, Frontiers, and Foreign Relations, 31 b.c. to a.d. 378’, Britannia 13 (1982), 1–23CrossRefGoogle Scholar, esp. 7–11, and ERW, 28–40. See now too Halfmann, H., Itinera principum. Geschichte und Typologie der Kaiserreisen im Römischen Reich (1986)Google Scholar. For the speed of legislative communications, see Bacchi, M., ‘La rapidità delle communicazioni legislative sotto l'Impero’, Scritti (1925), 149–54Google Scholar, and Aru, L., ‘Osservazioni sulla rapidita delle communicazioni legislative nell'Impero’, Studi economici e giuridichi della Facoltà Giuridica dell' Università di Caglìarì 17 (1929), 127–30.Google Scholar

¹⁷ It is in itself surprising to see used in the dating formula of a Flavian subscript an expression well attested, in this context, only in the Codex Iustinianus (but see n. 9), in some constitutions of Justinian himself, ‘recitatae in novo consistorio palatii’ (CJ 1. 2. 22. 1, 1. 14. 12. 5, 2. 55. 4. 7, etc.), the only exception being CJ 6. 60. 2 (a letter of Constantine in a.d. 319 to the magistrates and the senate of Rome, read solemnly in the curia). Nevertheless, it is not unknown under the early Empire, where it seems to have all the functions of an oral publication: it is only after their recitatio, for example, that the clauses of a testament may be put into effect; and for the imperial legislation, as rightly pointed out by Coriat, op. cit. (n. 11), 726–48, together with the propositio, it has all the senses of a promulgation— only with its recitatio does the product of the imperial will acquire an official value. Pace González (1986), 238, this has nothing to do with the inscription Syll. ³ 883, where a quasi-private epistula of Caracalla (with the unofficial, at these times, valedictory formula Ἐρρωσο, cf. n. 8) to the great notable Aurelius Iulianus is publicly read to honour him in the theatre.

There is a difference between publication by recitatio and that by propositio, which always took place at the imperial residence, while recitatio necessarily took place in the provinces of the addressees. From this point of view, its equivalent in the Codex Iustinianus is to be found in the expression acceptum/accepta, quite often used in the dating formulas: most of the time it is the only dating element (as in CJ 2. 12. 2, 2. 20. 1, 5. 75. 1, etc.), but in three cases—CJ 1. 15. 1 (Gratian, Valentinian, and Theodosius), 9. 17. 1 and 9. 47. 16 (both Constantine)—as in our inscription, jointly with the precise date of delivery of the answer. The delays themselves correspond reasonably well to the five months in our document: if only one month and a half was needed for the constitution of Gratian, Valentinian, and Theodosius (but the place of reception is not known), three months were necessary in 318–19 from Sirmium (at this time the imperial residence: see Barnes, T. D., The New Empire of Diocletian and Constantine (1982), 74CrossRefGoogle Scholar) to Carthage for CJ 9. 17. 1, and six months between Trier and Hadrumetum in CJ 9. 47. 16.

One must note that, when mentioned, these towns where the constitutions were acceptae were all provincial capitals: Antioch (CJ 5. 35. 5 and 9. 18. 2), Carthage (CJ 9. 17. 1) and Hadrumetum (CJ 9. 47. 16 under Constantine: from the time of Diocletian capital of the provincia Valeria Byzacena). It must therefore be seriously doubted whether the recitatio was conducted only in the presence of the city magistrates who had sent the embassy; the evidence obliges us to believe rather that it took place in a provincial capital, that is, wherever the governor was and could preside at the reading. There is a second point to take into account: as the different inscriptions of the Severan period show (n. 12 above), the dating formula seems one of the most indispensable elements of a subscript (maybe as an authentication, cf. n. 32 below); and one cannot imagine local magistrates modifying the text which came into their hands to adapt it to the conditions of reading in their small municipium. It must therefore be assumed that the recitatio was a solemn form of reading under gubernatorial supervision. For some mentions of it, see the edict of L. Aemilius Rectus immediately preceding the letter of Claudius to the Alexandrians (P. Lond. 1912)Google Scholar, where a public reading is indicated; possibly, as suggested by W. Williams, ZPE 17 (1975), 42 n. 11, PUG 10. 8; and the edict of Marcus Aurelius to the Achaeans (Oliver, op. cit. (n. 6), 5–6, 11. 37–8), where the date of public reading is assimilated to a true date of promulgation.

¹⁸ Cf., on the basis of its diplomatic form, Williams, art. cit. (n. 6), 197–8; on the basis of its style, Honoré, op. cit. (n. 7), 102; and, from the vagueness itself of the term litterae, Coriat, op. cit. (n. 11), 94–5.

¹⁹ González (1986), 237.

²⁰ For such vagueness as characteristic of subscripts, cf., beside Williams, art. cit. (n. 6), 197–8, Honoré, op. cit. (n. 7), 50–1, and in the Apokrimata dossier the famous P. Col. vi. 123. iii, perhaps the shortest imperial pronouncement, with its ‘Obey what has been judged’.

²¹ Cf. Honoré, 58–9: a subscript is the most direct type of written answer which may be conceived from an emperor, and does not usually explain the legal notions involved, as in CJ ₄. 6. 1,2. 23. 1, etc. One of the basic characteristics is the absence of any introduction to the operative part: cf. Williams, 197–8, ‘even the most curt of Trajan's replies to Pliny differs … by including some account of the problem raised by Pliny … Such silence is entirely characteristic of subscripts’.

²² Honoré, op. cit. (n. 7), to be read in conjunction with Millar, F., ‘A New Approach to the Roman Jurists’, JRS 76 (1986), 272–80.Google Scholar On all these points, cf. also Williams, art. cit. (n. 6), 197.

²³ For his moderate and sensible conclusions in this field, Honoré, 1–23, besides his major book on Tribonian, the director of Justinian's legal commission which compiled the Corpus Iuris Civilis, Tribonian (1978)Google Scholar. See now also Coriat, op. cit. (n. 11), 756–82.

²⁴ As no such list of these documents is available, it is worth giving one here: (1) the subscript to the Samians from 38 b.c. (Reynolds, J., Aphrodisias and Rome (1978), no. 13Google Scholar); (2) possibly another to the same from 19 b.c. (RDGE 62); (3) the three subscripts from the second half of the first century collected on the bronze tablet from Vardacate (AE 1949. 24); (4) the subscript of Trajan to the Smyrnaeans (see our n. 15); (5) certainly the rescript of Trajan or Hadrian to the ordo of Italica (CIL 11. 5368); (6) a subscript from the reign of Commodus (AE 1894. 61 and 1903. 202); (7) the sacrae litterae of a.d. 204, certainly addressed to a magistrate (see our n. 12); (8) the document of Julia Domna addressed to the Ephesians (I. Ephesos 212. 9–14) which with Nörr, art. cit. (n. 11), 24 n. 66 we must consider as a subscript; (9) and the subscript from a.d. 244–7 addressed to the villagers of Aragua (OGIS 519). But this list of subscripts addressed to villages or cities of the Roman Empire must be supplemented by a list of subscripts addressed to the communities of a non-civic nature: (1) P. Berol. inv. 16456 (temple of Soxis); (2) ILS 7784. 1–17 (school of the Epicureans in Athens); (3)AE 1958. 9+ SEG 15. 108 (identity of the community unknown); (4) CIL viii. 10570 (the coloni from the saltus Burunitanus); (5) IGUR i. 35 (two subscripts to the Peanists); (6) P. Lond. inv. 2565. 105–6 (on the exemption of peasants from the urban liturgies); (7) P. Oxy. Li. 3611 (to the hieroneikai from Antinoopolis); and IGLS 4028 (temple of Baetocaece). As observed by Coriat, op. cit. (n. 11), 389–97, the problem is not only the nature of the addressees, but also the point of law (private or public) involved, and, more precisely, the nature of the whole procedure, contentious or honorific, the latter developing before the other: the first preserved subscript in private law dates only from a.d. 121 (P. Tebt. ii. 286). It may be owed to the fact that, as assumed by Honoré, op. cit. (n 7), 4–11, the developments of the libellus procedure in this field are caused primarily by the crisis of the practice of the responsum prudentium.

²⁵ Besides the evidence already presented, see Pliny, Ep. 10. 47. 1–2 (the colony of Apamea) and 10. 83 (the request of the Nicaeans); and in addition CJ. 8. 37. 1 (Severus and Caracalla), with its curious introductory sentence ‘licet epistulae, quam libello inseruisti’, discussed by Nörr, art. cit. (n. 11), 17. As demonstrated by Coriat, op. cit. (n. 11), 390–6, since nearly all their subjects addressed them by way of a petition, the emperors usually had full liberty of choice in the form of their answer. It is well known, however, that such a universal use of the petition altered its form under the Roman Empire—on this point, see still Ziemann, F., De epistularum Graecarum formulis solemnibus quaestiones selectae (1910), 261–6Google Scholar. For a vigorous enforcement of these informal rules of protocol, see the inscription from the saltus Burunitanus (CIL viii. 10570. iii. 15–19), where the soldiers are sent by the procurators to mistreat the coloni ‘only for the reason that we intended to beseech Your Majesty by way of an epistle’.

²⁶ It may be significant that many subscripts addressed to civic communities (see n. 23), when their content can be discovered, report refusals by the imperial power, and as such, were very often inscribed by the opposite party (see the subscript to the Smyrnaeans from 38 b.c., that of Trajan to the same, and the sacrae litterae of a.d. 204, quoted n. 24). Though not a general rule, the more honorific form of the epistula was more appropriate to a formal display of imperial benevolence.

²⁷ Cf. ERW, 426: ‘Though it is clear that emperors assented to requests whenever they could, the reason for this rarity of negative replies is not so much the universality of their benevolence as the fact that in the case of refusal the city had no motive for going to the expense of inscription’, and 431–2.

²⁸ Cf. 11. 2 (‘sollicitudo vestra’) and 4–5 (‘me/mineritis legis’). Further proof may be found in the fact that quite often, when the petition is presented by an intermediary, there is an attraction in the imperial pronouncement, which thus becomes an answer to the intermediary, from the second person plural to the second singular: see the petition of the Skaptopareni presented ‘per Pyrrum militem conpossessorem’ (Syll. ³ 888). Prof. P. Le Roux informs me that he is preparing an article on these preparatory committees for ZPE.

²⁹ To our knowledge, no subscript quoted and used by the addressee shows a similar absence either of the imperial heading or of the libellus which had been sent to the emperor.

³⁰ See Williams, art. cit. (n. 6), 197; Jones (n. 12), 99; Drew-Bear (n. 12), 363.

³¹ This is a very important parallel, for, as convincingly demonstrated by Harris, W. V., ‘The Imperial Rescript from Vardacate’, Athenaeum 59 (1981), 338–52Google Scholar, this document dates approximately from the same period, indeed the same decade as our document, if we agree with his identification of the author of the rescript as Nerva: addressed to Clodius Secundus (certainly a Roman official and friend of the emperor), who was not their original addressee, for his personal use, the rescript quotes the three imperial subscripts collected in this letter, omitting in the same way as in our document the name of the addressee. For other examples of such omissions, see P. Oxy. IX. 1202. 5–12 and XLIII. 3105. 1–10,SB 5294. 12–15, W. Chr. 41. iii. 20, P. Flor. 382. 27–35, P- Ryl- 117. 27, P. Mich. xiv. 675. 14.

³² This may be an important point to add to the discussion on the hypothetical existence of a special file for subscripts. See Wilcken, art. cit. (n. 7), Martin and d'Ors (n. 11), Williams (n. 6), Nörr (n. 11), and Williams, ibid., 198–204: the problem is that the entire discussion has been conducted on the basis of the difficult Skaptopara inscription. For a thorough and sensible conclusion, see F. Millar, ‘L'empereur romain comme décideur’ (to be published in Typologie des Etats antiques, ed. Cl. Nicolet): ‘Malheureusement, nous n'avons aucune raison de croire que de telles archives provinciales aient jamais existé, pas meme seulement des archives centrales, sises à Rome ou destinées à accompagner l'empereur dans ses voyages … L'énigme de la répartition et de la connaissance des souscriptions reste à résoudre’.

³³ The dating does not depend only upon palaeographical arguments: the subscript (with its mention of ‘idus Domitianae’) is undoubtedly a pronouncement of Domitian, even if it was not sent to the city of Irni. Further, it must not be forgotten that after his murder Domitian underwent damnatio memoriae; if a copy of this law had been engraved after his reign, his name would certainly not have appeared. As he is the living emperor in all the oath formulas, which could in fact be very easily adapted or modified, the main part of the law—up to and including the sanctio—must have been engraved under his reign. For the contemporaneity of the engraving of the ‘Addendum’, see n. 35.

³⁴ This is another argument to add to those presented by Prof. Millar (to whom I am grateful for having provided me with a draft of his paper) at the Lex Irnitana Colloquium, held in London on 10 Nov. 1986, to support a separation, in Spain, of the processes of the application of the municipal law from the sending of petitions with embassies: it appears that, at least in Spain (but even elsewhere it is poorly attested), the emperor had not presided personally at the giving of the law to each town, that very likely being the governor's task.

³⁵ For these palaeographical remarks, cf. González (1986), 238: the point is relevant, even if palaeographical differences in Latin epigraphy within a short period of time are not very easily perceived.

³⁶ For this common meaning of the word lex, close to the English use of the word ‘law’, see Thesaurus Linguae Latinae, s.v. ‘Lex’, coll. 1238–56.

³⁷ For another possibility, saving another unattested hypothesis, see below n. 48.

³⁸ In fact, there are two possible contexts for the sending of the libellus which dictated this imperial subscript: (a) an embassy, anxious (‘sollicitudo vestra’: there may be here some scornful irony of Domitian) to see the status of its citizens clearly established without any possible contest, asked some indulgentia of the emperor, that is (cf. n. 39 below) some derogation from the strict dispositions of the law. This libellus should then be understood in the context of ‘petition-and-response’ between city and emperor so well studied in ERW, esp. 375–8s; (b) during a trial possibly conducted by the governor himself, a legal ambiguity was raised, and it was necessary to resort to the ‘free legal advice’ of the emperor—see on this point Honoré, op. cit. (n. 7), esp. 24–33, and on the formal means of attracting the attention of the imperial power, Coriat (n. 11), 346–403.

³⁹ On this notion of indulgentia, see Gaudemet, J., Indulgentia principis (1st.d.st.d.dir.Univ. Trieste, Publ. no. 3, 1962)Google Scholar, mostly interested in its developments during the later empire, and Waldstein, W., Untersuchungen zum römischen Begnadigungsrecht. Abolitio—Indulgentia—Venia (1964), esp. 108–44Google Scholar. Very recently, Cotton, H., ‘The Concept of Indulgentia under Trajan’, Chiron 14 (1984), 245–66Google Scholar, more interested in the unity of the concept than in its particular legal applications: in fact, in a monarchical structure directed by imperial patronage, the functioning of many normal legal mechanisms would naturally be described as a result of the imperial will. In our passage, the solemn venia granted by the emperor Domitian may mean little more than a respect for the almost universal principle of non-retrospective application of a law.

⁴⁰ For an example of this impatience and irritability, in particular towards the noisy reactions of the Roman people in the amphitheatre, see Suet., Dom. 10. 1. The harsh order σιωπήσετε was known as the ‘word of Domitian’: Dio 79. 6. 1. But, without showing any ‘indulgence’ for Domitian, one must remember that listening to wordy petitions on obscure points of law certainly did not predispose emperors to refuse a favour with courtesy (cf. Williams, art. cit. (n. 6), 197–8 and ERW, passim).

⁴¹ For the official character of the recitatio, see n. 17. On this semi-public diffusion, see, beside the reflections of Prof. Millar quoted n. 34 and Reynolds, op. cit. (n. 15), 97, Honoré (n. 7), 1–23, and Coriat (n. 11), 756–82—‘la diffusion des constitutions impériales par le biais des jurisconsultes’. For a concrete example, cf. Westermann, W. L., Schiller, A. A., Apokrimata. Decisions of Septimius Severus on Legal Matters (1954), 99–101Google Scholar, where all the constitutions preserved in P. Col. vi. 123 are assumed to have been collected to serve as a memorandum either for a lawyer or a judge, or (opinion then abandoned) for a notary.

⁴² It is possible that this legal ambiguity arose from some difficulties created by the imposition of a general Roman mould with its formalist categories on a situation regulated by peregrine law. For a mention of these iuris periti, certainly regional practicians, see the Severan constitution of a.d. 204 quoted above, and for a provincial example, Katzoff, R., ‘Responsa prudentium in Roman Egypt’, Studi in onore di A. Biscardi 11 (1982), 524–35.Google Scholar However, there is a difficulty in this strictly ‘municipal’ hypothesis: as has been noted (n. 26 above), the reasons for the engraving at public expense of an imperial refusal are always obvious, most of the time because the community which had the inscription put up benefited from it—quite the opposite of what is suggested here.

⁴³ There is a problem with this hypothesis: pace Martin, op. cit. (n. 8), 282–6, we have simply no evidence. Furthermore, the correspondence between Pliny and Trajan (esp. Ep. 10. 58. 3) seems to demonstrate that subscripts granted in the province under his predecessors were not available to the present governor. We could, of course, theoretically suppose that he knew this subscript because he had presided at its reading, but that would be yet another unattested hypothesis.

⁴⁴ For it must be noted that rubric LXXXXVII, from the point of view of the themes treated, and rubrics xxii and xxiii, can neither have been conceived, nor originally composed, separately from each other; their common object is to allow the preservation of patronage relations despite the attribution of Roman citizenship, and all the possibilities are explored: those who acquire citizenship when their freedmen do too (rubric xxii), those who acquire citizenship when their freedmen do not (rubric xxiii), and those who do not acquire citizenship when their freedmen do (our rubric LXXXXVII). Such a unity of object is proof that rubric LXXXXVII has been displaced, and such a displacement may be the reason why, contrary to the use current at the time of this inscription (cf. the exemplars of the Flavian municipal law from Malaca and Salpensa— Lois des Romains IV. 4 and 5), no numeration has been given to the rubrics, to avoid any confusion with the corresponding paragraphs in the laws of other municipia. Prof. Crawford assures me that the positioning of a rubric outside the space delimited by the sanctio is attested in no other law.

⁴⁵ For the exclusion of freedmen from municipal magistracies see Treggiari, S., Freedmen in Late Roman Republic (1969), 63–4Google Scholar, and Duff, A. M., Freedmen in the Early Roman Empire (1928), 137.Google Scholar On the limits of accession of the sons of freedmen to municipal magistracies, see Boulvert, G., Domestique et fonctionnaire sous le Haut-Empire Rotnain: la condition de l'affranchi et de l'esclave du prince (1974), 323–5CrossRefGoogle Scholar.

⁴⁶ One should not forget that Domitian had experienced in his own chancery the pressure of freedmen for intermarriages, with the scandalous wedding of the father of Claudius Etruscus with a senatorial woman, probably Tettia Etrusca—cf. Evans, J. K., ‘The Role of Suffragium in Imperial Political Decision-Making: a Flavian Example’, Historia 27 (1978), 102–28. But, if one excepts the prohibition of any nuptiae with a member of the ordo senatorius, the marriage laws of Augustus seem to have favoured unions between freeborn and freedmen. Among such marriages, those between citizens and freedwomen seem to have been the most frequent—see Duff, 60–3 discussing Dig. 23. 2. 44, CJ 5. 4. 28. pr., and Dio 54. 16.Google Scholar

⁴⁶ One should not forget that Domitian had experienced in his own chancery the pressure of freedmen for intermarriages, with the scandalous wedding of the father of Claudius Etruscus with a senatorial woman, probably Tettia Etrusca—cf. Evans, J. K., ‘The Role of Suffragium in Imperial Political Decision-Making: a Flavian Example’, Historia 27 (1978), 102–28. But, if one excepts the prohibition of any nuptiae with a member of the ordo senatorius, the marriage laws of Augustus seem to have favoured unions between freeborn and freedmen. Among such marriages, those between citizens and freedwomen seem to have been the most frequent—see Duff, 60–3 discussing Dig. 23. 2. 44, CJ 5. 4. 28. pr., and Dio 54. 16.Google Scholar

⁴⁷ For this very vague sense of coisse, to designate any form of union, see for example Quint., Inst. 5. 11. 32, Gaius, Inst. 3. 59, and Paul., Dig. 23. 1. 2. 2. It is possible that the form of marriage recognized by Spanish peregrine law, under which the commoners in a municipium were perhaps living (see ERW, 485 and Appendix 4, ‘Freeborn cives Latini in the Roman Empire?’, 630–5), was in fact assimilated to concubinage, not proper conubium. Further, one should not dismiss the possibility, as in modern societies, of fake marriages undertaken only for the acquisition of Roman citizenship and possibly followed by a quick divorce. Such an evasion could explain the irate tone of Domitian. On the notion of venia, see n. 37.

⁴⁸ For this sense of considerate, see Cic, Scaur. 37, Plane. 72, Phil. 4. 6. For the diminished capacity of freedwomen under legacy laws, see now Fabre, G., Libertus. Recherches sur les rapports patron-affranchi à la fin de la République Romaine (1981), 304Google Scholar. The reason why this hypothesis is the most probable lies in the difference of the clauses preserving patronage obligations in rubrics XXIII and LXXXXVII: the clause ‘idem ius eademque condicio esto, quae esset, si civitate mutati mutatae non essent’ of rubric xxiii becomes in rubric LXXXXVII ‘idem ius … esto, quod esset, si a civibus Romanis manumissi manumissae essent’. The explanation for this Domitianic pronouncement must be found here, and in the differences in patronage obligations between Roman and peregrine law. Either the provincials have petitioned the emperor to remain in the old regime of obligations, and all that they have obtained was a temporary exemption (but then, why would others have reused this subscript?), or more probably the provincials have found the regime granted by Roman law more advantageous (in particular perhaps the independence conferred on a freedwoman by the ius liberorum) and have petitioned the emperor for a modification of the clause, which would have been originally similar to that of rubric xxiii. Having obtained it, they had, as did everyone reusing this text, to engrave this rubric with the subscript of Domitian specifying its conditions of use. From this point of view, the solemn refusal of any further indulgentia should be interpreted as a ‘clause of protection’, aimed at the preservation of all the other points of the law, as, for example, in the letter of Marcus Aurelius and Lucius Verus to Coiedius Maximus from the Tabula Banasitana (ILM 94, 1. 5), or in the Edict of Caracalla from Banasa (ILM 100, 11. 16–18).

⁴⁹ See for this point of view Fabre, op. cit. (n. 48), 313–14 and Duff (n. 45), 188–90.

⁵⁰ For some thorough remarks on the transcription of a municipal law from a papyrus roll, see Mallon, op. cit. (n. 3), 53 on the presentation of the Osuna bronze tablet. For the necessity of seeing in the office of the governor the real dispatcher of the various municipal laws, see our n. 34. For the role of the governor in dealing with the imperial subscript, see our n. 17.

⁵¹ Also relevant is the so-called Fragmentum Italicense (Girard, p. 124 = FIRA no. 35 =Lois des Romains IV. 7 = J. González, ‘Italica, municipium iuris Latini’, Mel. Casa Velazquez 20 (1984), 17–32); one does not see why this city, a colony at the time of the engraving at the beginning of the third century, would then have inscribed its old municipal law. It bears fragments of two columns, one of which seems to be a piece of our rubric LXXXX, while the other bears what is also the last line of our sanctio rubric LXXXXVI, ‘cuiq(ue) per h(anc) l(egem) actio petitio pe[rsecutio esto]’, alas a very common formula in the municipal laws, with a wide blank space after it. It might therefore be possible that in this example from Cortegana there was no ‘Addendum’ (displaced rubric and subscript) after the sanctio rubric. However, such a remark does not prove anything, as we are not even sure that this fragment is the lower right corner of the tablet.

⁵² For this notion of different versions of the same law, supervised by Roman authority, and the use of one version instead of another according to local needs, see A. Lintott, ‘Notes on the Roman Law inscribed at Delphi and Cnidos’, ZPE 20 (1976), 65–82, more precisely 79–81.

⁵³ There is room for surprise at such an opinion, when both Gaius, Inst. 1. 3–5 and Ulpian, Dig. 1. 4. 1 explicitly say the contrary, declaring that the legislative activity of Roman emperors has always been considered as a source of law. For this wide-spread opinion, see for example Girard, P. F., Manuel élémentaire de Droit Romain (1911), 60Google Scholar; Jolowicz, H. F., Historical Introduction to the Study of Roman Law (1932), 371Google Scholar; Gaudemet, J., Institutions de l'antiquité (1972), 349Google Scholar; and recently Green, E., in The Roman World (1987), 447Google Scholar.