Hostname: page-component-77c89778f8-swr86 Total loading time: 0 Render date: 2024-07-22T05:06:34.409Z Has data issue: false hasContentIssue false

Statistics and the Conversion of the Roman Aristocracy*

Published online by Cambridge University Press:  14 March 2012

T. D. Barnes
University of Toronto


In a justly famous paper published in 1961, Peter Brown set out a model for understanding the historical process whereby the formerly pagan aristocracy of imperial Rome became overwhelmingly Christian during the course of the fourth and fifth centuries. Brown's paper has deeply influenced all who have subsequently studied this historical phenomenon, at least in the English-speaking world. Since this article argues that the Roman aristocracy became Christian significantly earlier than Brown and most recent writers have assumed, it must begin by drawing an important distinction. Brown's paper marked a major advance in modern understanding because it redirected the focus of scholarly research away from conflict and confrontation, away from the political manifestations of paganism culminating in the ‘last great pagan revival in the West’ between 392 and 394, away from episodes which pitted pagan aristocrats of Rome against Christian emperors, away from ‘the public crises in relations between Roman paganism and a Christian court’, towards the less sensational but more fundamental processes of cultural and religious change which gradually transformed the landowning aristocracy of Italy after the conversion of Constantine. This change of emphasis was extremely salutary in 1961, it has permanently changed our perception of the period, and it entails a method of approaching the subject which remains completely valid. Unfortunately, however, Brown also adopted prevailing assumptions about the chronology of these changes which are mistaken, on the basis of which he asserted that the ‘drift into a respectable Christianity’ began no earlier than the reign of Constantius. The evidence and arguments set out here indicate that the process began much earlier and proceeded more rapidly than Brown assumed, but they in no way challenge the validity of his approach to understanding the nature of the process.

Copyright © T. D. Barnes 1995. Exclusive Licence to Publish: The Society for the Promotion of Roman Studies

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)


1 Brown, P. R. L., ‘Aspects of the Christianization of the Roman Aristocracy’, JRS 51 (1961), 111Google Scholar, reprinted in Religion and Society in the Age of Saint Augustine (1972), 161–82.

2 See, for example, the sensitive recent discussion of ‘conversion and uncertainty’ by Markus, R., The End of Ancient Christianity (1990), 2743Google Scholar.

3 Eck, W., ‘Das Eindringen des Christentums in den Senatorenstand bis zu Konstantin d. Gr.’, Chiron I (1971), 381406Google Scholar, cf. A. Harnack, Die Mission und Ausbreitung des Christentums in den ersten drei Jahrhunderten 4 (1924), 946–58, esp. 950 (Rome).

4 viz., the African martyr Crispina: the Acta Crispinae (BHL 1989a/b) imply that she lacked the senatorial status which Augustine attributes to her (Enarr. in Ps. 120.13 (CCL 40.1799)).

5 Eusebius leaves her anonymous, while Rufinus names her Sophronia, apparently misunderstanding Eusebius' description of her as σωϕϱονεστάτη γυνή(HE VIII.14.16). She was plausibly identified as the wife of Junius Flavianus, praefectus urbi from 28 October 311 to 9 February 312, by A. Chastagnol, Les fastes de la préfecture de Rome au Bas-Empire (1962), 58–9.

6 Rapsaet-Charlier, M.-Th., ‘Les femmes sénatoriales du IIIe siècle. Étude préliminaire’, Prosopographie und Sozialgeschichte. Studien zur Methodik und Erkenntnismöglichkeit der kaiserzeitlichen Prosopographie. Kolloquium Köln 24.-26. November 1991 (1993), 147–63Google Scholar, at 162, cf. PIR 2 A 319; C 573; H 236; J 6.

7 I have deliberately chosen the lowest possible estimate and ignored adlecti: on the probable size of the Senate during the third century, see briefly Talbert, R. J. A., The Senate of Imperial Rome (1984), 2938Google Scholar, 131–4; Jacques, F., ‘Le nombre des sénateurs aux IIe et IIIe siècles’, Epigrafia e ordine senatorio I (Tituli IV, 1982, pub. 1984), 137–42Google Scholar.

8 von Haehling, R., Die Religionszugehörigkeit der hohen Amtsträger des römischen Reiches seit Constantins I. Alleinherrschaft bis zum Ende der Theodosianischen Dynastie (1978) (hereafter von Haehling, Religionszuhörigkeit).Google Scholar

9 Von Haehling, Religionszugehörigkeit, 507 (‘Tabelle VI: Der Anteil von Heiden und Christen bei den ermittelten Amtsinhabern unter den einzelnen Kaisern’), 510 (‘Graphische Darstellung zu Tabelle VII’).

10 Von Haehling, Religionszugehörigkeit, 614–18.

11 In addition to the reviews cited in nn. 12–15 and 27–8, I am aware only of those by Gaudemet, J., RHDFE 58 (1980), 648–50Google Scholar, and von Ungern-Sternberg, J., MH 37 (1980), 265Google Scholar.

12 Nellen, D., Gymnasium 89 (1982), 378–9Google Scholar.

13 Chastagnol, A., Latomus 39 (1980), 487–8Google Scholar.

14 Eck, W., HZ 231 (1980), 139–41Google Scholar.

15 Martindale, J. R., JRS 69 (1979), 194–6Google Scholar.

16 Salzman, M. R., ‘Aristocratic women: conductors of Christianity in the fourth century’, Helios 16 (1989), 207–20, at 208Google Scholar; On Roman Time. The Codex Calendar of 354 and the Rhythms of Urban Life in Late Antiquity (1990), esp. 195 n. 8, 223; Cooper, K., ‘Insinuations of womanly influence: an aspect of the Christianization of the Roman aristocracy’, JRS 82 (1992), 150–64, at 150 n.4Google Scholar.

17 Salzman, M. R., ‘How the West Was Won: The Christianization of the Roman Aristocracy in the Years after Constantine’, in Deroux, C. (ed.), Studies in Latin Literature and Roman History, Collection Latomus CCXVII (1992), 451–79Google Scholar. Her Table 3: ‘Religious Identification by Emperor at Highest Appointed Office’ gives figures of eleven pagans and six Christians for the period 312–337, eight pagans and three Christians for the period 337–350, and eleven pagans and ten Christians for the period 351–360. Unfortunately, she does not make clear exactly what these figures represent or how they have been obtained, except to state that they are ‘senatorial aristocrats’ culled from PLRE I and that she has ‘included only those people for whom there was explicit evidence for religious preference’ (456–8).

18 Cameron, Alan, ‘The date and owners of the Esquiline Treasure,’ AJA 89 (1985), 135–45, at 144CrossRefGoogle Scholar. The inscription, which reads ‘ Secunde et Proiecta vivatis in Chri[sto]’, is reproduced in Shelton, K. J., The Esquiline Treasure (1981), 31–3Google Scholar, cf. 72–5 no. 1.

Recently, with appeal to a forthcoming study entitled The Last Pagans of Rome, Cameron, Alan and Long, Jacqueline have reiterated that ‘the aristocracies of Athens and Rome continued to be substantially pagan into the late fourth century’ (Barbarians and Politics at the Court of Arcadius (1993), 14Google Scholar).

19 As argued in PLRE I.817, adding that the postulated son ‘was possibly the first of the Turcii to become a Christian’.

20 Shelton, K. J., ‘The Esquiline Treasure: the nature of the evidence’, AJA 89 (1985), 147–55CrossRefGoogle Scholar. As in her book of 1981, Shelton steadfastly refused to date the casket any more precisely than between the broad termini of 330 and 370.

21 Cameron, Averil, Christianity and the Rhetoric of Empire. The Development of Christian Discourse (1991), 41, 121Google Scholar.

22 Lane Fox, R., Pagans and Christians (1986), 667.Google Scholar

23 Cameron, Averil, Christianity (1991), 191Google Scholar.

24 MacMullen, R., Christianizing the Roman Empire (1984). 83Google Scholar.

25 Cameron, Averil, The Later Roman Empire AD 284–430 (1993), 78Google Scholar. Cameron also turns Petronius Probus, cos. 371, into a ‘pagan aristocrat’ (73).

26 e.g., recently, Thrams, P., Christianisierung des Römerreiches und heidnischer Widerstand (1992), 170Google Scholar: ‘Sehr aufschlussreich ist die Tabelle zu den heidnischen und christlichen Inhabern hoher Ämter bei R. von Haehling’.

27 Chantraine, H., BZ 73 (1980), 362–4Google Scholar; Demougeot, É., RHE 74 (1979), 389–95Google Scholar; Noethlichs, K. L., JAC 21 (1978), 193–8Google Scholar. The fallacy emerges clearly from a comparison of von Haehling's Tabelle II and Tabelle III, which arrive at the overall total of 757 when considering both the ‘Zahl der ermittelten Religionsangehörigen in den einzelnen Ämtern’ and ‘Feststellbarer Anteil von Heiden und Christen bezogen auf die Amtsinhaber’ (Religionszugehörigkeit, 492, 495). The review by Klein, R. dutifully follows von Haehling and erroneously speaks of ‘757 Amtsinhaber’ (ZKG 91 (1980), 401–6)Google Scholar.

28 Phoenix 33 (1979), 364–5Google Scholar.

29 From The New Empire of Diocletian and Constantine (1982), 131–9, 158, 160, 171; ‘Praetorian Prefects 337–361’, ZPE 94 (1992), 249–60Google Scholar, unless otherwise stated.

30 As argued in ‘Publilius Optatianus Porfyrius’, AJP 96 (1975), 173–86Google Scholar.

31 So, rightly, PLRE I (1971), 100Google Scholar. The only evidence is the horribly confused account of the Council of Tyre in Rufinus, HE X.16–18, which explicitly dates the council, which is irrefragably dated to 335, after the death of Constantinus in 340 and alleges that Constantius ordered Athanasius to be condemned by bishops assembled at Tyre under the supervision of (1) a comes sent from court, (2) Archelaus, who was comes Orientis at the time, and (3) the governor of Phoenice.

32 Two senators under Constantine’, JRS 65 (1975), 40–9Google Scholar.

33 Schumacher, W. N., ‘Zum Sarkophag eines christlichen Konsuls’, Röm. Mitt. 65 (1958), 100–20Google Scholar, cf. Fuhrmann, H., ‘Studien zu den Consulardiptychen verwandten Denkmälern I. Eine Glasschale von der Vicennalienfeier Constantins des Grossen zu Rom im Jahre 326 nach Chr.’, Röm. Mitt. 54 (1939), 161–75Google Scholar.

34 Against the earlier assumption that he was a pagan, see A. Chastagnol, Fastes (1962), 91. In the present context, I refrain from challenging von Haehling's classification of Domitius Zenophilus as a pagan (Religionszugehörigkeit, 420, adducing AE 1915.30 (Lambaesis)), lest I appear to be constructing a circular argument.

35 On the motives which may have led Constantius to make the unusual appointment of Placidus and Vulcacius Rufinus to this eastern post in the early 340s, see von Haehling, Religionszugehörigkeit, 179.

36 As argued in PLRE 1 (1971), 814Google Scholar.

37 Malcus, B., Opuscula Atheniensia 7 (1969), 106–8Google Scholar.

38 See Himerius and the fourth century’, CP 82 (1987), 206–25, at 218–20Google Scholar.

39 ZPE 94 (1992), 258, 259Google Scholar, cf. Norman, A. F., ‘The Illyrian prefecture of Anatolius’, Rh. Mus., N.F. 100 (1957), 253–9Google Scholar; R. Penella, Greek Philosophers and Sophists in the Fourth Century A.D. Studies in Eunapius of Sardis (1990), 90–1, 96–8, 130–2.

40 As von Haehling, Religionszugehörigkeit, 200. O. Seeck, Briefe des Libanius (1906), 188, deduced that Italicianus was a pagan from Libanius, Ep. 8, which von Haehling disallows.

41 As does von Haehling, Religionszugehörigkeit, 195–6, asserting that in its context ‘ist παϱαβάτης kein Schimpfname fur die Anhänger der Athanasianischen Gegenpartei’. Similarly, he classifies the general Sebastianus as a Manichee on the strength of Athanasius, Fug. 6.5; Hist. Ar. 59.1, 61.3 (Religionszugehörigkeit, 260): for the fallacy in this case, see Tardieu, M., ‘Sebastianus étiqueté comme manichéen’, Klio 70 (1988), 494500CrossRefGoogle Scholar.

42 See Athanasius and Constantius. Theology and Politics in the Constantinian Empire (1993), 72, 83, 85, 167Google Scholar.

43 Von Haehling, Religionszugehörigkeit, 61, uses precisely the same argument to classify Strategius Musonianus as an Arian.

44 For the attestations, see PLRE I.611/2; 438/9.

45 Augustine, , De sermone domini in monte I.50 (PL 34.1254)Google Scholar; Athanasius, , Hist. Ar. 22.1Google Scholar, cf. O. Seeck, Die Briefe des Libanius (1906), 156.

46 Von Haehling, Religionszugehörigkeit, 522, mistakenly includes as appointees of Constantinus three men who held office under Constans between 337 and 340, viz., Apronianus, L. Turcius, PUR 339Google Scholar, Fabius Titianus PUR 339–341, and Aurelius Celsinus, proc. Africae 338–339.

47 CP 82 (1987), 216Google Scholar.

48 Haehling, Von, Religionszugehörigkeit, 417Google Scholar, includes Clodius Celsinus Adelphius in his list of proconsuls of Africa under Constans. However, the only evidence for a proconsulate (province unspecified) is Isidore of Seville, who identifies the centonist Proba as uxor Adelphii proconsulis’ (De viris illustrious 18 (22) (PL 83.1093))Google Scholar: that is diagnosed as a mistake for praefecti by J. F.Matthews, ‘The poetess Proba and fourth-century Rome: questions of interpretation’, Institutions, société et vie politique au IVeme siècle ap. J. C. (284–423). Autour de l'oeuvre d'André Chastagnol (1992), 277–304, at 284 n. 1.

49 Rosen, K., ‘Ein Wanderer zwischen zwei Welten: Carmen ad quendam senatorem ex Christiana religione ad idolorum servitutem conversum’, in Dietz, K., Hennig, D. and Kaletsch, H. (eds), Klassisches Altertum, Spätantike und frühes Christentum. Adolf Lippold zum 65. Geburtstag gewidmet (1993), 393408Google Scholar, at 393: ‘Um 350 n. Chr. war fast noch die gesamte Senatsaristokratie in Rom heidnisch’.

50 JRS 51 (1961), 9Google Scholar = Religion and Society (1972), 177.

51 Barnes, T. D. and Westall, R. W., ‘The conversion of the Roman aristocracy in Prudentius' Contra Symmachum’, Phoenix 45 (1991), 5061CrossRefGoogle Scholar.

52 Ambrose's claims are treated with extreme scepticism, and in effect denied by McLynn, N., Ambrose. Church and Court in a Christian Capital (1994), 33–5Google Scholar. Their validity or otherwise does not affect the point at issue here.

53 Novak, D. M., ‘Constantine and the Senate: an early phase in the Christianization of the Roman aristocracy’, Ancient Society 10 (1979), 271310Google Scholar.

54 Champlin, E. J., ‘Saint Gallicanus’, Phoenix 36 (1982), 70–6CrossRefGoogle Scholar.

55 Table III A–C supersedes the less systematic analysis which I prepared in 1987 and which was published in Pagans and Christians in the reign of Constantius’, Entretiens sur l'Antiquité Classique 34 (1989), 301–37, at 315–20Google Scholar.

56 Who were the nobility of the Roman Empire?Phoenix 28 (1974), 444–9CrossRefGoogle Scholar.

57 Von Haehling, , Religionszugehörigkeit, 505Google Scholar. His discussion of the ‘Zahl der erfassten Amtsinhaber in Relation zur wahrscheinlichen Gesamtzahl’ calculates that 70 per cent of the proconsuls of Asia and 56 per cent of the proconsuls of Achaea are known between 324 and 450 (487–489, with Tabelle I). However, the figures which he gives for ‘ermittelte Anzahl’ and ‘geschätzte Gesamtzahl’ (including anonymi, respectively, forty two out of sixty for Asia and thirty five out of sixty three for Achaea) indicate that his ‘geschätzte Gesamtzahl’ in these cases represents not the total of all proconsuls who held office between 325 and 450, but an estimate of the number of proconsuls who are unattested. By my count, for the period 325–450, the names are known of about twenty five proconsuls of Asia and of about thirty five proconsuls of Achaea: admittedly, some proconsuls served several years, especially in Asia, but the number of proconsular years in which the name of the proconsul of Asia and Achaea are unknown exceeds seventy five and ninety respectively (in each case out of 125), so that we probably know the names of less than 30 per cent of the proconsuls of Achaea and less than 40 per cent of the proconsuls of Asia between 325 and 450. (Table II Column (5) gives my estimate of the percentage of office-holders in von Haehling's categories who are attested for the period 324–361.)

58 New Empire (1982), 171; ‘Proconsuls of Africa, 337–392’, Phoenix 39 (1985), 144–53, at 145CrossRefGoogle Scholar.

59 New Empire (1982), 158, 160.

60 PLRE I.834; Matthews, J. F., Western Aristocracies and the Imperial Court AD 364–425 (1975), 147Google Scholar.

61 Novak, D. M., Ancient Society 10 (1979), 293Google Scholar; Champlin, E. J., Phoenix 36 (1982), 76CrossRefGoogle Scholar.

62 See the edition of G. Polara (1973), 1. 12, 32, 41, 57, 69, 72.

63 Morris, J., ‘Prosopography of the Later Roman Empire’, Klio 46 (1965), 361–5, at 363–4CrossRefGoogle Scholar, cf. PLRE I.951.

64 More missing names (A.D. 260–395)’, Phoenix 27 (1973). 135–55, at 149CrossRefGoogle Scholar.

65 On the plausibility of the charge, see Athanasius (1993), 52.

66 As by A. Alföldi, Die Kontorniaten. Ein verkanntes Propagandamittel der stadtrömischen heidnischen Aristokratie in ihrem Kampfe gegen das christliche Kaisertum (1943), 8–85, largely reprinted in Alföldi, A. and E., Die Kontorniat-Medaillons II (1990), 1263Google Scholar.

67 Alan Cameron, ‘Orfitus and Constantius: a note on Roman gold-glasses’, (forthcoming), cf. ‘Religious Affiliation’ (Table III), 10 n. 1.

68 Marcone, A., ‘Costantino e l'aristocrazia pagana di Roma’, in Bonamente, G. and Fusco, F. (eds), Costantino il Grande II (1993), 645–58, at 653Google Scholar.

69 For a converging argument based on the results of excavations at La Magliana, see Broise, H. and Scheid, J., Recherches archèologiques à la Magliana. Le balneum des frères arvales, Roma antica I (1987), 275–7Google Scholar. If the Arval Brethren ceased to use the sacred grove of Dea Dia and the attached private balneum shortly after 334/5, that implies that Christianity had already made serious inroads into the social strata from which the confraternity was recruited.