¹ A refreshing, but little-heeded, exception was Morris, J., ‘Changing fashions in Roman nomenclature in the early Empire’, LF 86 (1963), 34–46Google Scholar.

² The classic account is E. Fraenkel, ‘Namenwesen’, RE 16 (1935), columns 1648–70, in which post-classical developments are discussed, under the emotive title ‘Die allmähliche Zerrüttung des alten römischen Namensystems’ (columns 1662–4). Greek influence is cited by Solin, H., ‘Die innere Chronologie des römischen Cognomens’, in Duval, N. (ed.), Actes du colloque Internationale du CNRS no. 564 sur l'onotnastique latine, Paris 13–15 octobre 1976 (1977)Google Scholar (hereafter Onom.Lat.), 429; Christianity by Kazhdan, A. P., ‘Names, Personal’, Oxford Dictionary of Byzantium (1991)Google Scholar (hereafter ODB), 1435; imperial nomina and weakening legal control by Kajanto, I., Onomastic Studies in the Early Christian Inscriptions of Rome and Carthage (1963)Google Scholar (hereafter Onomastic Studies), 16–18; a combination by Brown, T. S., Gentlemen and Officers: Imperial Administration and Aristocratic Power in Byzantine Italy A.D. 554–800 (1984), 20Google Scholar.

³ e.g. the papers of the international congress Onomastique latine (Paris, 1977)Google Scholar were dominated by discussion of epigraphic sources, taking little account of the important work recently done on Late Antique papyri.

⁴ C. Titius Probus, De praenominibus (preserved with Valerius Maximus in Iulius Paris' epitome) 4.1: ‘Varro simplicia in Italia fuisse nomina ait existimationisque suae argumentum refert, quod Romulus et Remus et Faustulus neque praenomen ullum neque cognomen habuerint.’

⁵ , Priscian, Institutiones grammaticae 2.23Google Scholar (H. Keil, Gramm.Lat. 2.57, 11. 13–17): ‘Praenomen est quod praeponitur nomini vel differentiae causa vel quod tempore, quo Sabinos Romani asciverunt civitati, ad confirmandam coniunctionem nomina illorum suis praeponebat nominibus et invicem Sabini Romanorum. Et ex illo consuetudo tenuit, ut nemo Romanus sit absque praenomine.’ Cf. Probus, De praenominibus 2: ‘Romanus autem est ab Albanis et Sabinis multiplicandorum nominum consuetudinem traxisse, quoniam ab illis orti sunt.’

⁶ On the ancient Greek naming system see the extensive discussion of E. Fraenkel, RE 16, columns 1611–48. The usage of the patronymic in -ides/-ades in a fossilized form among the Athenian aristocracy to denote members of a γένος — a group of families claiming descent from a single heroic or divine ancestor whom they worshipped in a collective ceremony (e.g. the Εὐμολπίδαι) — is highly atypical.

⁷ e.g. Lucius, Publius, and Servius are probably abbreviations of compound elements (cf. Λυσίας, Νιϰίας), while Gnaeus and Spurius are clearly examples of the descriptive, ‘Schimpfnamen’, type (cf. Σίμων, Δίδυμος). On their etymologies see Chase, G. D., ‘The origin of Roman praenomina’, HSCP 8 (1897), 153–8Google Scholar and 181.

⁸ Of a total fifty six known from the historians and the inscriptions of CIL 1. According to Probus, De praenominibus 3.1, Varro considered that there were only about thirty praenomina. G. D. Chase, HSCP 8 (1897), 135, counts sixty four, but I have excluded those cognomina latterly used as praenomina (see below p. 131).

⁹ I list here, with their ancient abbreviations, the seventeen praenomina in question: A. [Aulus]; Ap(p). [Appius]; C. [Gaius]; Cn. [Gnaeus]; D. [Decimus]; L. [Lucius]; M'. [Manius]; M. [Marcus]; N. [Numerius]; P. [Publius]; Q. [Quintus]; Ser. [Servius]; Sex. [Sextus]; Sp. [Spurius]; T. [Titus]; Ti(b). [Tiberius]; V. [Vibius].

¹⁰ Salomies, O., Die römischen Vomamen (1987), 215–25Google Scholar, notes some exceptions.

¹¹ cf. Syme, R., The Roman Revolution (1939)Google Scholar, Genealogical Table 1. The Metelli.

¹² Occasionally in lengthened form as -eius/-aeus; -a/-as and -o and -anus/-(i)enus forms belong to Etruria and Transappennine Italy respectively. See further Schulze, W., Zur Geschichte lateinischer Eigennamen (revised edn, 1991)Google Scholar.

¹³ Rix, H., ‘Zum Ursprung des römisch-mittelitalisch Gentilnamensystem’, ANRW 1. 2, 714–58Google Scholar; J. Morris, LF 86 (1963), 35: etymologies include reference to both animal (e.g. Porcius) and vegetable (e.g. Iulius), hair colour (e.g. Flavius), topographical provenance (e.g. Tarquinius) as well as fossilized patronymics akin to the -ides form in Greek (e.g. Marcius).

¹⁴ A phenomenon mocked by Cicero, Brutus 62; Tusculanae disputationes 1.38.

¹⁵ Momigliano, A. D., ‘The Origins of Rome’, CAH ² 7.2Google Scholar, saw the growth of the binominal system as concurrent with the urbanization of archaic Central Italy.

¹⁶ Heurgon, J., ‘Onomastique étrusque: la dénomination gentilice’, Onom. Lat., 27–31Google Scholar; ‘Vetusia’ (= ? Veturia) is on a silver bowl from the Bernadini tomb, Praeneste, of c. 700/650 B.C., cited by Cornell, T. J., ‘The tyranny of the evidence: a discussion of the possible uses of literacy in Etruria and Latium in the archaic age’, in Humphrey, J. H. (ed.), Literacy in the Roman World (1991), 18–21Google Scholar.

¹⁷ Such is Livy's usage; e.g. Appius Claudius is called simply Appius (II.24–27); and Kaeso Quinctius simply Kaeso (III.11–15).

¹⁸ Polybius, XVIII passim (except at XVIII.46.5 where, quoting a Roman declaration, he uses the correct Roman form ‘T. Quin<c>tius'); indeed, following Polybius’ usage, Plutarch's life of Flamininus is the only one to be entitled by praenomen alone.

¹⁹ Q. Oppius; see Reynolds, J. M., Aphrodisias and Rome (1982), 12, doc. 2b, 1. 13.Google Scholar

²⁰ G. D. Chase, HSCP 8 (1897), 159–74.

²¹ Kajanto, I., The Latin Cognomina (1965), 120–2Google Scholar; e.g. Ahala, Piso, Cicero, Scipio. For an example of the obscurity of their meanings, even to Romans sometimes, see the discussion of the various etymologies proposed for Galba in Suetonius, Galba 3.

²² Badian, E., ‘The clever and the wise: two Roman cognomina in context,’ in Horsfall, N. (ed.), Vir Bonus Discendi Peritus (1988), 8f.Google Scholar L. Cornelius Cn.f. Scipio (cos. 298), born c. 340 B.C., is the first case verifiable from a non-literary source according to I. Kajanto, ‘On the chronology of the cognomen in the Republican period’, Onom.Lat., 63; unless P. Cornelius P. f. Scapula of AE 1967.19 may be identified with, or related to, the magister equitum of 362 or consul of 328 (Solin, H., ‘Analecta epigrafica’, Arctos n.s. 6 (1969)Google Scholar, noff.; cf. T. R. S. Broughton, MRR III (1986), 70).

²³ Mau, A., ‘Cognomen’, RE 4 (1900)Google Scholar, column 225.

²⁴ I. Kajanto, Onom.Lat., 67.

²⁵ Adams, J. N., ‘Conventions of naming in Cicero’, CQ n.s. 28 (1978), 151–4CrossRefGoogle Scholar; Syme, R., ‘Imperator Caesar: a study in nomenclature’, Historia 7 (1958), 185f.Google Scholar = Roman Papers 1 (1979), No. 29, 374ff. On the retention of such ‘gentilicial’ cognomina by adoptees, see Salomies, O., Adoptive and Polyonymous Nomenclature in the Roman Empire (1992) (hereafter APNom), 83Google Scholar.

²⁶ This is less plausibly attributed to strict state control by C. Nicolet, ‘L'onomastique des groupes dirigeants sous la République’, Onom.Lat., 46ff.

²⁷ Although it has been argued (e.g. Doer, B., Die römische Namengebung (1937), 46–52, 68–71Google Scholar) that such cognomina ex virtute were the result of a formal grant by the Senate, Linderski, J., ‘The surname of M. Antonius Creticus and the cognomina ex victis gentibus’, ZPE 80 (1990), 159ffGoogle Scholar., demonstrates that it is better to understand them as being assumed unofficially; it was an added benefit if the Senate actually granted a corresponding triumph. Thus M. Antonius managed to snatch his ‘Creticus’ from the jaws of defeat (ibid., 163f.).

²⁸ However this was not the only measure used to reflect adoption during the Republic; on which see Bailey, D. R. Shackleton, Two Studies in Roman Nomenclature (1976), 83–7.Google Scholar

²⁹ e.g. the fourth-century Diomedes, Ars grammatica I (H. Keil, Gramm.Lat. 1.321, 11. 3–11): ‘Proprium nominum quattuor sunt species, praenomen, nomen, cognomen, agnomen. Praenomen est quod nominibus gentiliciis praeponitur, ut Marcus, Publius. Nomen proprium est gentilicium, id est quod originem familiae vel gentis declarat, ut Porcius, Cornelius. Cognomen est quod unius cuiusque proprium est et nominibus gentiliciis subiugitur, ut Cato, Scipio. Ordinantur enim sic, Marcus Porcius Cato, Publius Cornelius Scipio. Agnomen quoque est quod extrinsecus cognominibus adici solet ex aliqua ratione vel virtute quaesitum, ut est Africanus, Numantinus et similia.’ And Priscian, Inst.Gramm. 2.22 (ibid., 2.57, ll. 12–13): ‘Nam propria habent species separatim quattuor: praenomen, nomen, cognomen, agnomen.’

³⁰ Though for exceptions in both cases see Salomies, O., Die römischen Vornamen (1987), 233–8Google Scholar. In contrast while still servile the single name was placed in the position of praenomen to the master's gentilicium; e.g. Apollonius Laelius Q(uinti) s(ervus), Prepon Alleius M(arci) s(ervus) of ILLRP 1². 194 = ILS 9236 Delos, who contrast with the freedman M. Granius M(arci) l(ibertus) Heras from the same stone.

³¹ The following discussion is based on that of Syme, R., The Augustan Aristocracy (1986), 313–18Google Scholar and Genealogical Table 26.

³² Figures from H. Solin, Onom.Lat., 103; more strikingly still, if one subtracts from these 196 those which are Greek (which in CIL 1² more often belonged to freedmen rather than natives), only 66 Latin cognomina remain.

³³ As analysed by J. N. Adams, CQ n.s. 28 (1978), 145ff.

³⁴ Plutarch, , Marius 1Google Scholar.

³⁵ In the light of Quintilian's words, A. Mócsy's attempt (in discussion of I. Kajanto's, ‘The emergence of the late single name system’, Onom.Lat., 429) to identify Juvenal's tria nomina as praenomen, nomen, and patronym or tribe is highly implausible.

³⁶ On Junian Latins see Weaver, P. R. C., ‘Where have all the Junian Latins gone? Nomenclature and status in the early Empire’, Chiron 20 (1990), 275–305Google Scholar.

³⁷ CIL III. 399 Pergamum, c. A.D. 40. The cognomen's final position is precisely dictated by the regulations of the lex repetundarum, FIRA 1², no. 7, l. 14 and Tabula Heracleensis, FIRA 1², no. 18, l. 146.

³⁸ On which see Syme, R., Roman Papers I, 365–77Google Scholar.

³⁹ Instances are listed by Salomies, O., Die römischen Vornamen (1987), 313–18Google Scholar. The latest datable example is PIR ² E 82, Galeo Tettienus Severus M. Eppuleius Proculus L.f. Claud(ia) Ti. Caepio Hispo, suffect consul in A.D. 102 or 104.

⁴⁰ See the stemma of the T. Flavii, PIR ² III, 183.

⁴¹ Salomies, O., Die römischen Vornamen (1987), 378–89Google Scholar. An effect of spurious variation of praenomina was sometimes created by the inheritance of a fossilized praenomen through the maternal instead of the paternal line; which accounts for the examples of unexpected variation cited by Salomies, 387f., such as the curious instance of three or more sons sharing only two praenomina: C. Cassius Crispus, L. Cassius Secundus and C. Cassius Atilianus (CIL v.5991 Milan).

⁴² e.g. P.Oxy. 2565, A.D. 224; AE 1948.121, A.D. 240.

⁴³ Salomies, O., Die römischen Vornamen (1987), 411ff.Google Scholar

⁴⁴ For statistics on the use of nomina as cognomina at Rome, see Kajanto, I., Onomastic Studies, 22Google Scholar. The borrowing of famous surnames such as Stanley, Percy, Howard, Sydney, and Nelson as popular Christian names presents a modern British parallel.

⁴⁵ ILS 7591 Lugdunum. Despite the provenance, Constantinius Aequalis was originally from Germaniceia in Commagene.

⁴⁶ See principally Syme, R., ‘Clues to testamentary adoption’, in Panciera, S. (ed.), Titulus 4: Atti del colloquio interazionale dell'AIEGL su epigrafia e ordine senatorio, Roma 14–20 maggio 1981 1 (1982 [1984]), 397–410Google Scholar = Roman Papers IV (1988), No. 9, 159–73; idem, ‘The paternity of polyonymous consuls’, ZPE 41 (1985), 191–8 = Roman Papers v (1988), No. 37, 639–47; and, more generally, O. Salomies, APNom, passim.

⁴⁷ Suetonius, Tiberius 6.3: ‘Post reditum in urbem a M. Gallio senatore testamento adoptatus hereditate adita moxnomina abstinuit, quod Gallius adversarum Augusto partium fuerat’.

⁴⁸ This case was discussed by Syme, R., ‘Praesens the friend of Hadrian’, in Studia in Honorem Iiro Kajanto (1985), 274Google Scholar = Roman Papers v, No. 32, 563f. and again most recently by O. Salomies, APNom, 36f.

⁴⁹ Mommsen, Th., ‘Zur Lebensgeschichte des jüngeren Plinius’, Hermes 3 (1869), 31–139Google Scholar = Gesammelte Schriften IV. Historische Schriften I, No. 4, 366–468; in particular, ‘Plinius Adoption in lhrer rechtlichen Bedeutung’, 397–412. Maintenance of his adoptive cognomen would have resulted in the unaesthetic repetition of Secundus which had already been inherited from his mother. It is for similar aesthetic reasons that the second praenomen is omitted from the middle of his name.

⁵⁰ O. Salomies could find only one exception to the primacy of adoptive nomina, and that from the Greek East (APNom, 5, 42 and 83). He reckons that maternal nomenclature tends to be second, though adduces numerous counter examples (ibid., 63–7 and 75–8, cf. 67–9 and 78–80).

⁵¹ PIR P 492: Q. Pompeius Senecio Roscius Murena Coelius Sex. Iul(ius) Frontinus Silius Decianus C. Iul(ius) Eurycles Herculaneus L. Vibullius Pius Augustanus Alpinus Bellicius Sollers Iul(ius) Aper Ducenius Proculus Rutilianus Rufinus Silius Valens Valerius Niger Cl(audius) Fuscus Saxa Amyntianus Sosius Priscus (ILS 1104); at least six of these sets of names inherited from his father Q. Pompeius … Sosius Priscus, cos. 149 (PIR R 68). J. Morris, LF 86 (1963), 42–4 traces the origin of these names to a dozen persons of the Flavio-Trajanic period.

⁵² Salomies, O., APNom, 66 and 70f.Google Scholar

⁵³ Ausonius, Opuscula I.praefatiunculae I.9–12: ‘Hinc late fusa est cognatio; nomina multis / ex nostra, ut placitum, ducta domo veniant: / derivata aliis, nobis ab stemmate primo / et non cognati, sed genetiva, placent.’

⁵⁴ Ammianus XIV. 11.27; cf. PLRE I Cerealis 2.

⁵⁵ PLRE I Placidus 2, perhaps a descendant of C. Memmius M.f. Qui. Caecilianus Placidus a suffect consul of the mid-third century (PLRE I Placidus 3).

⁵⁶ On the purpose of military diplomas see Mann, J. C. and Roxan, M. M., ‘Discharge certificates of the Roman army’, Britannia 19 (1988), 344CrossRefGoogle Scholar. The Tabula Banasitana records imperial scrutiny of the claim of the chief of the Zegrenses, Aurelius Iulianus, for Roman citizenship for his wife and children; for the text see Sherwin-White, A. N., ‘The Tabula Banasitana and the Constitutio Antoniniana’, JRS 63 (1973), 86–7Google Scholar.

⁵⁷ BGC II.655 (Arsinoite nome, 16 August 215): Ἁὐρήλιος Ζώσιμος πρὸ μὲν τῆς θίας (sic) δωρεᾶς ϰαλούμενος Ζώσιμος Λεωνίδου…’.

⁵⁸ The only legal difference between Old and New Romans is that there is no evidence that the latter were ever enrolled in voting tribes. This is hardly surprising given the practical difficulty involved for the bureaucracy in digesting such an enormous number simultaneously.

⁵⁹ Bourazelis, K., Θεῖα Δωϱεά. Studies on the Policy of the Severans and the Constitutio Antoniniana (1989), 120–32Google Scholar; e.g. Aurelius accounts for c. 23 per cent of the nomina attested in the Christian epigraphy of Carthage and Rome, Iulius only c. 5 per cent (Kajanto, I., Onomastic Studies, 16Google Scholar).

⁶⁰ e.g. Aelia, Aurelia (CIL VI.2832 and 2833), Aelia, Antonia, Augusta, Flavia, Iulia, Septimia, and Ulpia (EE IV.891–5), all in laterculi praetorianorum from Rome, and the ex-praetorian ‘[M. Aur.] M.f. Ulp. Syrio’ posted to Carlisle (for whom see Hassall, M. W. C. and Tomlin, R. S. O., ‘Roman Britain in 1988. II Inscriptions’, Britannia 21 (1990), 331–3Google Scholar). On the history of the phenomenon before 212 see Forni, G., ‘Tribù e pseudo-tribù romane in epigrafi’, in Giuffrè, V. (ed.), Sodalitas I: Scritti in onore di A. Guarino (1984), 97–104Google Scholar.

⁶¹ In fact even their metronyms are included, according to local custom but entirely against traditional Roman practice: P.Oxy. 3906, ll. 2–5:

⁶² This was not a scribal habit confined to the army; see e.g. ‘Aurelius Ignatius Apollinaris’ of Nov./Dec. 213, from a family of L. Ignatii, in P.Diog. II, cf. 10 and 17.

⁶³ On the Aramaic/Arab genitives (as might be expected in a unit of Palmyrenes) represented by ‘Matthana’ and ‘Themarsa’ see Gilliam, J. F., P.Dura (1959) introduction, 60Google Scholar.

⁶⁴ For a statistical analysis see Gilliam, J. F., ‘The Dura rosters and the Constitutio Antoniniana’, Historia 14 (1965), 81–4Google Scholar.

⁶⁵ G. Daux, ‘L'onomastique romaine en Grèce. Appendice: Passage du nom grec au nom romain’, Onom.Lat., 413–16, discusses the development of this formula.

⁶⁶ Schubert, P., Les Archives de Marcus Lucretius Diogenes (1990), 16–19Google Scholar. Other proofs of citizenship, birth certificates and signet rings soon disappeared. The last known birth certificate (AE 1948.121) dates from 25 January 240 and finds of signet rings cease in mid-third-century contexts (oral communication of Dr M. Henig).

⁶⁷ e.g. the population of Altava and Pomarium in Mauretania, where the emperor Septimius Severus planted veteran colonies, habitually use gentilicia; the last datable instance is Iulia Rogatiana who died in 655 (IAM II.608, Volubilis). In Italy Melminius Cassianus, the last recorded member of a family which dominated the municipal offices of Ravenna since c. 500, was magistratus in 575 (P.Ital. 6).

⁶⁸ As noticed in the epigraphic record of Rome and Carthage by I. Kajanto, Onomastic Studies, 16–17. See also Kajanto, Onom.Lat., 421–30.

⁶⁹ On which see Syme, R. in Straub, J. (ed.), ‘The Ancestry of Constantine’, The Historia Augusta: a Call of Clarity (1971), 60Google Scholar = Historia Augusta Papers (1983), No. 5,76.

⁷⁰ CIL II.1228 Apulum, cited by Veyne, P., ‘Le «prénom» de Naucellius’, RPh 3^me série 38 (1964), 256Google Scholar. Veyne was evidently disturbed that the term ‘praenomen’ which in classical usage denoted the diacritical element was by the fourth century being used to denote the invariable nomen.

⁷¹ Ammianus XXVIII.4.7.

⁷² M. Aurelius Valerius Claudius (A.D. 268–270), M. Aurelius Claudius Quintillus (270), M. Aurelius Probus (276–282), M. Aurelius Carus (282–283), M. Aurelius Carinus (283–285), M. Aurelius Numerius Numerianus (283–284), and M. Aurelius Iulianus (284/5).

⁷³ Kajanto, I., Onomastic Studies, 49Google Scholar. Hobson, D. W., ‘Naming practices in Roman Egypt’, BASP 26 (1986), 189fGoogle Scholar.

⁷⁴ Of the 140 individuals with Greek names recorded on inscriptions culled from the index of Roueché, C. M., Aphrodisias in Late Antiquity (1988), 339–42Google Scholar, which covers A.D. 250–650, 61 (43 per cent) are of the -ios/-ia type.

⁷⁵ The earliest Latin -ius cognomen known to Kajanto, I. (Onomastic Studies, 29)Google Scholar is borne by the pantomimist M. Aurelius Augg. lib. Agilius Septentrio (CIL XIV.2113 and 2977) in 187 or 192, but they were not common amongst the freeborn until the late third century.

⁷⁶ Sophocles, E. A., Greek Lexicon of the Roman and Byzantine Period (memorial edn, 1887), 390Google Scholar Δόγμα 3.

⁷⁷ Kajanto, I., Onomastic Studies, 38–48Google Scholar.

⁷⁸ e.g. Maximinus in the heading of an imperial letter, recorded by Eusebius, Ιώβιος Μαξιμῖνος Σαβίνῳ (HE IX.9a.4) and Galerius as Iovius Maximianus on ILS 661 from Solva (Noricum).

⁷⁹ PLRE I, Proculus II (full name: CIL VI.1690, 1691) appears as Populonius Proculus in the dating formula of, e.g., P.Col. VII. 148, 149, etc.; for full references see Bagnall, R. S. et al. , Consuls of the Later Roman Empire (1987), 215Google Scholar (hereafter CLRE). The consistency of the formulae precludes the possibility of a simple misunderstanding of the syntax of their names. I will argue the case for Iulianus Ionius fully elsewhere.

⁸⁰ e.g. Preisigke, F., Namenbuch (1922)Google Scholar, Foraboschi, D., Onomasticon alterum papyrologicum (1967)Google Scholar, and the indices of ICVR.

⁸¹ Based on the figures of Kajanto, I., Onomastic Studies, 16Google Scholar.

⁸² Mócsy, A., ‘Der Name Flavius als Rangbezeichung in der Spätantike’, in Akte des int. Kongresses für griechischen und lateinischen Epigraphik, Wien 1962 (1964), 261Google Scholar; G. M. Browne, P.Mich. X (1970), introduction, 55.

⁸³ Keenan, J. G., ‘The names Flavius and Aurelius as status designations in later Roman Egypt’, ZPE II (1973), 33–63Google Scholar; idem, part 2, ZPE 13 (1974), 283–304 and ‘An afterthought on the names Flavius and Aurelius’, ZPE 53 (1983), 245–50.

⁸⁴ J. G. Keenan, ZPE 53 (1983), 250.

⁸⁵ He is plain Αὐρήλιος Διοσϰουρίδης Ἰουλιανός in P.Oxy. 2585 (Oct./Nov. 315), but Οὐαλέριος Διοσϰουρίδης Ἰουλιανός curator in P.Oxy. 42, 900, 2767, and 1509 (322–323).

⁸⁶ P. Aelius Vibullius Rufus at Athens, P. Aelius Flavius Apollonius at Miletus, and P. Aelius Otacilius Moschus at Pergamum, on whom see Salomies, O., APNom, 62 n.8Google Scholar.

⁸⁷ J. G. Keenan, ZPE II (1973), 63.

⁸⁸ The dynastic nomen is surely best seen as the natural accompaniment of rank conferred by certain appointments rather than the result of individual imperial codicils in each case as J. G. Keenan would have (ZPE II (1973), 40). For the transferability of honor to the spouse see CTh II.1.7 of 10 November 392.

⁸⁹ PLRE I Valentinus 12 (and cf. stemma 27, where a family connection with the Symmachi is proposed on the basis of the combination ‘Aur. Val.’, borne also by Symmachus 6); Vrbanus 4.

⁹⁰ As he is correctly called by T. E. Gregory, ODB, 5241. The form C. Flavius Valerius Iulius Constantius, given by PIR ² (F 390) is a chimaera produced by cross-breeding the evidence of the ill-informed (a few milestones) with the plain wrong (the Historia Augusta). The more cautious PLRE I (Constantius 12) does not give him a praenomen and sensibly rejects the Iulius. The ‘Caius’ of the milestones is hardly surprising given a public which had become unaccustomed to varying praenomina, the last imperial dynasty to differentiate them having been that of Septimius Severus.

⁹¹ For the ‘Iulius Aurelius’ adopted by the Palmyrenes in 212, perhaps commemorating a special favour of the empress Iulia Domna, as a possible precursor see Schlumberger, D., ‘Les gentilices romains des Palmyréniens’, BEO 9 (1942/1943), 53–82Google Scholar. A more certain example of the phenomenon may come as early as the 240s, i.e. the ‘Iulius’ adopted by Sentius Malchus βουλεύτης, σύνδιϰος ϰαὶ ἐπιμελητής of Philippopolis in Arabia in a dedication to the emperor M. Iulius Philippus (Segal, A., Town Planning and Architecture in the Province of Arabia (1988), 95–100Google Scholar No. 401a, cf. No. 395).

⁹² Licinius even celebrated the arrangement by the creation in 314 of the twin provinces Aegyptus Iovia and Herculia, on which see Barnes, T. D., The New Empire of Diocletian and Constantine (1982), 211CrossRefGoogle Scholar.

⁹³ The only exceptions are four Western emperors of the fifth century: Petronius Maximus (455), Iulius Valerius Maiorianus (457–461), Libius Severus (461–465), and Anicius Olybrius (472), all but Majorian from the Italian senatorial aristocracy.

⁹⁴ The adoption of Christian names was examined by Bagnall, R. S., ‘Religious conversion and onomastic change in early Byzantine Egypt’, BASP 19 (1982), 105–24Google Scholar to estimate the rate of conversion and was criticized by Wipszycka, E., ‘La valeur de l'onomastique pour l'histoire de la christianisation. A propos d'une étude de R. S. Bagnall’, ZPE 62 (1986), 173–81Google Scholar; cf. Bagnall, , ‘Conversion and onomastics: a reply’, ZPE 69 (1987), 243–50Google Scholar.

⁹⁵ A. P. Kazhdan, ODB, 1435. Chrysostom, John, Sur la vaine gloire et l'education des enfants (Malingrey, A.M., 146) 648–51Google Scholar:

⁹⁶ On the use of ‘Abba’ by clergy see J. G. Keenan, ZPE 13 (1974), 283–304.

⁹⁷ On the different employment of Flavius between the élites of East and West see CLRE, 26–40. They were similarly reluctant to replace the traditional vir clarissimus by new titles such as illustris or spectabilis.

⁹⁸ Ausonius, , Opuscula 16Google Scholar. Griphus ternarii numeri 80: ‘tria nomina nobiliorum’. Known only as Decimius (not Decimus cf. CLRE 292f.) Magnus Ausonius, we are, ironically, ignorant of the poet's own praenomen.

⁹⁹ Ammianus XXVIII.4.7: ‘Praenominum [i.e. nominum] claritudine conspicui quidam (ut putant) in immensum semet extollunt, cum Reburri et Flavonii et Pagonii Gereonesque appellentur, ac Dalii cum Taraciis et Ferasiis, aliisque ita decens sonantibus originum insignibus multis.’

¹⁰⁰ PLRE 2 Decius 2 and stemma 26, 1324.

¹⁰¹ PLRE 2 Basilius 11 and 12.

¹⁰² Considered as another son of Basilius (cos. 463) by PLRE 2 Basilius 13, and CLRE, 503.

¹⁰³ Cameron, A. D. E., ‘Polyonymy in the Late Roman aristocracy: the case of Petronius Probus’, JRS 75 (1985), 173Google Scholar.

¹⁰⁴ A. P. Kazhdan, ODB, 1453.

¹⁰⁵ IAM II.608 : ‘D.M.S. Memoria Iulia Rogatian(a) de Altava Ko(o)ptativa cui fili et nep(otes) fec(e)r(unt), vix(it) ann(os) pl(us) m(inus) LXXVI d(i)sc(essit) in p(ace) an(no) p(rovinciae) DCXVI’. The gentilicium was clearly still in use in Altava when she was born in A.D. 579. The last datable example of the full classical formula is C. Matrinius Aurelius C.f. Lem(onia tribu) Antoninus on ILS 6623, Hispellum (Umbria) from 333/337.

¹⁰⁶ SB 9460 (Arsinoë), l. 3 : Φλ. Τίτῳ εὐϰλεεστάτῳ δουϰὶ Ἀρϰαδίας ϰαὶ Θηβαίδος; P.Land. 1540 (Coptic document with Greek subscription): τὼν (sic) ἐνδοξοτάτον πάγαρχον.

¹⁰⁷ On Fl. Parsinus see Brown, T. S., Gentlemen and Officers, 272Google Scholar. Φλάυιοϛ Ιουστιανός in ll. 1 and 2–3 of the text provided by Vasiliev, A. A., ‘An edict of the emperor Justinian II, September 688’, Speculum 18 (1943), 5–6CrossRefGoogle Scholar.

¹⁰⁸ Paulus Diaconus, Historia Longobardorum 3.16.

¹⁰⁹ PLRE 2 Areobindus 1, Dagalaiphus 1, and Ariobindus 1. An analogous example is the Egyptian magnate Fl. Strategius Apion Strategius Apion (cos. 539), possibly ‘son of Strategius, grandson of Apion, great-grandson of Strategius’ (PLRE 3 Apion 3); for his ancestors see PLRE 2, 1325 stemma 27 and Strategius 8, Apion 2, and Strategius 9.

¹¹⁰ PLRE 3 Gregorius 3, Florentius 2, and Georgius 1.

¹¹¹ PLRE 2 Attalus 1; see also PLRE 3, 1545 stemma 12. On the family's wider eminent connections see Mommaerts, T. S. and Kelley, D.H., ‘The Anicii of Gaul and Rome’, in Drinkwater, J. F. and Elton, H. (eds), Fifth-century Gaul: a Crisis of Identity? (1992), 111–21Google Scholar.

¹¹² PLRE 2 Anastasius 17 and Magnus 5; they are simply (Flavius) Anastasius and (Flavius) Magnus in consular fasti, chronicle entries, and the consular formulae of contemporary papyri. Their full names are known only from their consular diptychs: CIL v. 8120.2, XIII.10032.5a + b (Anastasius); CIL XIII. 10032.6 (Magnus).

¹¹³ Needless to say ‘Pompeius’ here is simply a cognomen, no doubt inspired by the nomen of the great Republican general, but without any gentilicial force.

¹¹⁴ Cameron, A. D. E., ‘The House of Anastasius’, GRBS 19 (1978), 261Google Scholar; cf. PLRE 2 Anastasius 17 and Magnus 5, which considers them possibly brothers.

¹¹⁵ A near contemporary parallel example of the commemoration of the maternal grandfather is Flavius Licerius Firminus Lupicinus of Aries (PLRE 2 Lupicinus 3), who was grandson of a certain Firminus and nephew of bishop Magnus Felix Ennodius of Ticinum (PLRE 2, 1320 stemma 19) and who was sent to Milan for his education in 503. Licerius was the name of his paternal grandfather.

¹¹⁶ PLRE 1 Hypatius 6 and Pompeius 2; brothers according to Theophanes, A.M. 6024.

¹¹⁷ A. D. E. Cameron, GRBS 19 (1978), 229–62.

¹¹⁸ The wording of Marcellinus (under A. D. 532):‘Hypatius, Pompeius et Probus genere consobrini…’, would be peculiar if they were simply fratres, as Cameron argues. In its restricted sense consobrini designates sons of a mother's sister but was frequently employed to mean child of parent's brother or sister (cf. TLL 4, 473f., Lewis and Short, 434, OLD, 417), which allows a plausible identification of Probus as a son of Paulus.

¹¹⁹ The full nomenclature of the consul of 540, Fl. Marianus Petrus Theodorus Valentinus Rusticus Boraides Germanus Iustinus (PLRE 3 Iustinus 4), similarly acknowledged his uncle Boraides as well as his father Germanus; see PLRE 2, 1315 stemma 10.

¹²⁰ Note that the marriage of Paulus and Magna's daughter Irene to Olybrius (cos. 491) introduced the cognomen Proba (their daughter) into the Areobindi (PLRE 2, 1309 stemma 3) as well as Magnus, on which see A. D. E. Cameron, GRBS 19 (1978), 273ff.

¹²¹ e.g. Fl. Marianus Iaccobus Marcellus Aninus Addaeus, praefectus praetorio Orientis 551, praefectus urbis Constantinopoleos 565 (full name : SB 5.8939) is called simply Addaeus by Procopius (Anecdota 25.7).

¹²² PLRE 3 Aurelianus 1; in common with all the other known examples, he is probably an easterner since not only was he an appointee of Constantinople but moreover the orthography of his names (‘Michaelius Gabrielius’ instead of regular Latin ‘Michaelis Gabrielis’) suggests such an origin. The devotional nature of his nomenclature was first recognized by Gelzer, M., ‘Altes und Neues aus der byzantinisch-ägyptischen Verwaltungsmisere, vornehmlich im Zeitalter Justinians’, Archiv für Pap. 5 (1909–1913), 359f. n.5Google Scholar.

¹²³ Feissel, D. and Kaygusuz, I., ‘Un mandement impèrial du VI ^e siécle dans une inscription d'Hadrianopolis d'Honoriade’, Trav. et Mèm. 9 (1985), 397–419Google Scholar.

¹²⁴ Other bearers of this ‘Marian’ religious polyonymy from PLRE 3 are: Apion 3 (cos. 539), Athanasius 3 (dux Thebaidis ?566–568) and 8 (proconsul of Asia sixth/seventh century), Callinicus 4 (dux Thebaidis 568–569/70), Ioannes 21+22 (at Miletus c. 536) Iulianus 19 (dux Thebaidis 578), Iustinus 4 (cos. 540).

¹²⁵ PLRE 3, Hephaestus, praefectus praetorio Orientis A.D. 551–52 (cf. Dioscurus 8, Iacobus 1, and Thomas 10). On the cult of the martyr Menas, with its pilgrimage centre at Abū Mīnā where the emperor Arcadius had built a basilica, see Cabrol, F. and Leclercq, H., Dictionnaire d'archeologie chrètienne et de liturgie 12.1 (1933), 323–98Google Scholar.

¹²⁶ On the eclipse of the old aristocracy see Brown, T. S., Gentlemen and Officers, 21–37Google Scholar.

¹²⁷ Nikephoros, Brief History (C.A. Mango, 11.8). An earlier draft (the incomplete MS L) retains the original form Φάβιον, which, as Mango comments, is to be preferred since ‘Flavius is a mere gentilicium.’ Of course so is Fabius, but it is an ancient gentilicium used as a cognomen, whereas Flavius is a mere status-nomen.

¹²⁸ E. Fraenkel, RE 16, column 1659.