1 The Latin text is Rufinus' translation of Commentariorum in epistolam S. Pauli ad Romanos libri tres 10.41 (PG 14.1289C); the translation is mine.

2 Lagrange, M. J., Saint Paul: Épître aux Romains (Études Bibliques; Paris: J. Gabalda, 1916) 376Google Scholar; Weiss, J., Der erste Korintherbrief (Kritisch-exegetischer Kommentar über das Neue Testament; Göttingen: Vandenhoeck & Ruprecht, 1910 ⁹) 21Google Scholar; Zahn, T., Der Brief des Paulus an die Römer (Leipzig: A. Deichert, 1910) 614Google Scholar; Dodd, C. H., The Epistle of Paul to the Romans (Moffatt New Testament Commentary; London: Hodder and Stoughton, 1932) 244Google Scholar; Lietzmann, H., An die Römer (HNT 8; Tübingen: J. C. B. Mohr [Paul Siebeck], 1933 ⁴) 129Google Scholar; Barrett, C. K., A Commentary on the Epistle to the Romans (London: A & C Black, 1962) 286Google Scholar; Schlier, H., Der Römerbrief (HTKNT 6; Freiburg/Basel/Vienna: Herder, 1977) 451Google Scholar; Käsemann, E., Commentary on Romans (Grand Rapids: Eerdmans, 1980) 421Google Scholar; Dunn, J. D. G., Romans (WBC 38A–B; Dallas, TX: Word Books, 1988) 910Google Scholar; Fitzmyer, J. A., Romans: A New Translation with Introduction and Commentary (AB 33; Garden City, NY: Doubleday, 1993) 749Google Scholar; Jewett, R., Romans (Hermeneia; Minneapolis: Fortress, 2007) 980–1Google Scholar.

3 Theissen, G., The Social Setting of Pauline Christianity: Essays on Corinth (trans. J. H. Schütz; Philadelphia: Fortress, 1982) 89, 94Google Scholar; idem, ‘The Social Structure of Pauline Communities: Some Critical Remarks on Meggitt, J. J., Paul, Poverty and Survival’, JSNT 84 (2001) 65–84, at 83Google Scholar: ‘We can say that the house of Gaius must have been larger than average houses’ in order to accommodate ca. 30–50 members; Meeks, W. A., The First Urban Christians: The Social World of the Apostle Paul (New Haven: Yale University Press, 1983) 57Google Scholar: ‘evidently a man of some wealth’; Murphy-O'Connor, J., St. Paul's Corinth: Texts and Archaeology (Good News Studies 6; Wilmington: Michael Glazier, 1983) 156Google Scholar; Stegemann, E. W. and Stegemann, W., The Jesus Movement: A Social History of its First Century (trans. O. C. Dean; Minneapolis: Fortress, 1999) 295Google Scholar: ‘relatively prosperous’ because he ‘owns a house that was apparently large enough to serve as a place of assembly for all the confessors of Christ in Corinth’; Friesen, S. J., ‘Poverty in Pauline Studies: Beyond the So-Called New Consensus’, JSNT 26 (2004) 323–61, at 356Google Scholar; ‘perhaps the wealthiest person we know of from Paul's assemblies'; Jewett, Romans, 980; Welborn, L. L., An End to Enmity: Paul and the ‘Wrongdoer’ of Second Corinthians (BZNW 185; Berlin and New York: de Gruyter, 2011) 247CrossRefGoogle Scholar: ‘among the wealthiest individuals in the Corinthian church’.

4 Stegemann and Stegemann, Jesus Movement, 295: ‘if he was a citizen of the Roman colony of Corinth … then membership in the local Corinthian upper stratum (without official function) is certainly a possibility’; Judge, E. A., ‘The Early Christians as a Scholastic Community’, JRH 1 (1960–1) 4–15CrossRefGoogle Scholar, 125–37, at 130; idem, ‘The Roman Base of Paul's Mission’, TynBul 56.1 (2005) 103–17Google Scholar, at 112.

5 Gillman, J., ‘Gaius’, ABD 2 (1992) 869Google Scholar; Welborn, An End to Enmity, 299: Gaius ‘may have been a descendant of the Italian settlers; a Greek immigrant to the city from the time when Greek immigration became more frequent; or a freedman who gained wealth and a name following his manumission; or a freedman now enjoying Roman citizenship or a freeborn citizen of higher rank’. Earlier, Ramsay, W. M. (‘A Historical Commentary on the Epistles to the Corinthians’, Expositor, 6th series, 1 (1900) 19–31Google Scholar, 91–111, 203–17, 273–89, 380–87, at 101) argues that the use of the praenomen rather than a cognomen was a mark of a freedman, proud of his manumission: ‘Gaius of Corinth … was probably a rich freedman, to whom the honourable duty of entertaining the guests of the Church was assigned (Rom. xvi. 23). In his Pagan days he would have aimed at the honourable position of a Sevir Augustalis.’

6 Theissen, ‘Social Structure of Pauline Communities', 79: ‘Gaius must also have been a Christian with a central position in the congregation’.

7 Deissmann, A., Paul: A Study in Social and Religious History (trans. L. R. Strachan; New York: Hodder and Stoughton, 1912) 216Google Scholar. Deissmann (217) suggested that the elite women mentioned in Acts 17.4 soon abandoned the group.

8 Friesen, ‘Poverty in Pauline Studies’, 356–7 (not ‘higher than category 4 on the poverty scale’ – i.e. those with a ‘moderate surplus’ and positioned immediately below ‘municipal elites'); Longenecker, B. W. (Remember the Poor: Paul, Poverty, and the Greco-Roman World (Grand Rapids: Eerdmans, 2011) 239)Google Scholar places Gaius along with Phoebe and Erastos ‘conservatively’ in the middle of ‘ES4’ even though he argues that if Gaius’ house was ‘suitable and welcoming to a gathering of at least 45 people’ he could be put at the top of ES4.

9 Chow, J. K., Patronage and Power: A Study of Social Networks in Corinth (JSNTSup 75; Sheffield: JSOT Press, 1992) 90Google Scholar.

10 Goodspeed, E. J., ‘Gaius Titius Justus’, JBL 69 (1950) 382–3Google Scholar.

11 Bruce, F. F., The Letter to the Romans (Grand Rapids: Eerdmans, 1985) 265–6Google Scholar; Dunn, Romans, 910 (fits with the little information); Gillman, ‘Gaius’; Jewett, Romans, 980; hesitantly, Judge, ‘Scholastic Community’, 130; Longenecker, Remember the Poor, 239 n. 66 (‘cannot be confirmed’).

12 Τιτου א E 453 945 1175 1739 1891 2818 sy^p co; omit A B² D* L *Ψ 33 323614 1241 1505 M; Τιτιου B* D¹ sy^h.

13 Iustus is one of the most frequent Roman cognomina and commonly attested of both elite and ingenui (Kajanto, I., The Latin Cognomina (Societas Scientiarum Fennica. Commentationes Humanarum Litterarum 36/2; Helsinki: Societas Scientarium Fennica, 1965) 133, 252Google Scholar). By itself it is not an indication of a distinguished family.

14 Welborn, An End to Enmity, 299–300; see also Malherbe, A. J., Social Aspects of Early Christianity (Baton Rouge: Louisiana State University Press, 1977) 74 n. 30Google Scholar.

15 Welborn, An End to Enmity, 299.

16 Kajanto (The Latin Cognomina, 133) points out that cognomina such as Honoratus, Iustus, Magnus, Maximus and Verus that were common among the senatorial nobility (and which might underscore an aristocratic identity) were very rare among freedmen.

17 Judge, ‘Roman Base’, 111.

18 Welborn, An End to Enmity, 309–20 and the sources cited there. Earlier, Chow (Patronage and Power, 48–51) had commented on the careers of Spartiaticus and his grandfather Eurycles.

19 Welborn, An End to Enmity, 316.

20 Although it is common to import Acts' description of Crispus as an archisynagogos into discussions of the Crispus in 1 Cor 1.14, it is noteworthy that 1 Cor 1.14 gives no hint of either his role in a Judaean assembly or even that he was Judaean. For this reason Pervo, R. (Dating Acts: Between the Evangelist and the Apologists (Santa Rosa, CA: Polebridge, 2006), 103)Google Scholar believes that in composing Acts 18 Luke was dependent upon 1 Cor 1.14–16, which has Paul baptising Crispus (v. 14) and the household of Stephanas (v. 16). In the process of conflating these two portions of 1 Corinthians, Luke has Crispus' entire household being baptised, Stephanas disappears, and Crispus is made into a Judaean and ‘promoted … to a prominent place in the synagogue’. Puzzlingly, Crispus' conversion in Acts 18.8 comes only after Paul left the synagogue, when he declared that he was now turning to the Gentiles (Acts 18.6) and moved his preaching venue next door (18.7). It could be added that Crispus is not commonly attested as a Judaean name. It appears only in a funerary inscription from Cyrenaica: CJZC 12 (imperial period): Ἰωσης Κρίσπου (ἐτῶν) δ’, ‘Yoses son of Crispus, aged 4 years’. The feminine form Crisp(e)ina is found in JIWE 281, 282 (Rome, 3rd–4th cent. ce). There are no instances of the name in CIIP i–iii or IJO i–iii. Josephus three times refers to a Crispus (Vita 33, 388, 393), evidently a Lüdemann, Judaean. G. (Early Christianity according to the Traditions of Acts (Minneapolis: Fortress, 1989) 203–4)Google Scholar takes the material regarding Crispus in Acts as probably historical, as do most interpreters of 1 Corinthians, who read the Acts account into 1 Cor 1.14.

21 Welborn, An End to Enmity, 317.

22 Welborn, An End to Enmity, 241–50.

23 Welborn, An End to Enmity, 317–18.

24 Lagrange, Saint Paul: Épître aux Romains, 376–7 (my translation).

25 Chrysostom, Commentarius in Epistolam at Romanos 32 (PG 60.677B), alluding to Matt 10.11 and the admonition to receive only those who are ‘worthy’: ἀσπάζεται ὑμᾶς Γάϊος ὁ ξένος μου καὶ τῆς ἐκκλησίας ὅλης. εἶδες οἷον αὐτῷ στέφανον ἔπλεξε, τοσαύτην φιλοξενίαν μαρτυρήσας, καὶ ὁλόκληρον τὴν ἐκκλησίαν εἰς τὴν οἰκίαν συναγαγὼν τὴν ἐκείνου; τὸν γὰρ ξένον ἐνταῦθα τὸν ξενοδόχον φησίν. ὅταν δὲ ἀκούσης, ὅτι Παύλου ξενοδόχος ἦν, μὴ τῆς φιλοτιμίας αὐτὸν θαύμαζε μόνον, ἀλλὰ καὶ τῆς κατὰ τὸν βίον ἀκριβείας· εἰ μὴ γὰρ ἦν ἄξιος τῆς ἀρετῆς τῆς έκείνου, οὐδ’ ἂν ἐκεῖνο; ἐκεῖ κατήχθη. ὁ γὰρ πολλὰ τῶν ἐπιταγμάτων τοῦ Χριστοῦ σπουδάζων ὑπερβαίνειν, οὐκ ἂν τούτον παρέβη τὸν νόμον τὸν κελεύοντα περιεργάζεσθαι τοὺς ὑποδεχομένους, καὶ παρὰ ἀξίοις κατάγεσθαι.

26 Zahn, An die Römer, 614 n. 78: ‘Der Ausdruck verbietet die Deutung des Orig[en], daß Gajus der Korinth. Gemeinde sein Haus als Versammlungslokal zur Verfügung stellte. Dafür gebraucht Pl andere Worte … Dagegen weist ξένος (Chrys deutet es durch ξενοδόχος) auf Übung der φιλοξενία, gastliche Aufnahme zureisender Fremder …’ Similarly, Lietzmann, An die Römer, 129: ‘für alle durchreisenden Christen’; Schlier, Der Römerbrief, 451 n. 2; Käsemann, Romans, 421; Jewett, Romans, 980–1; and others.

27 Jewett, Romans, 980.

28 Adams, E., The Earliest Christian Meeting Places: Almost Exclusively Houses? (Library of New Testament Studies; London and New York: Bloomsbury T&T Clark, 2013) 28Google Scholar.

29 Belsheim, J. E., Epistulæ paulinae ante Hieronymum latine translatae ex codice sangermanensi graeco-latino (Kristiania: Cammermeier, 1885) 16Google Scholar.

30 Dunn, Romans, 910.

31 Dunn, Romans, 911. Likewise, Theissen, ‘Social Structure of Pauline Communities’, 83: ‘Other clubs in antiquity provide valid comparative figures. They rarely have more than 100 or less than 10 members; most of them comprise 20 to 50 members. These figures coincide roughly with the figures reached by archaeological research: on the basis of an analysis of archaeological ground plans we can assume that c. 30 to 40 could meet in a private house. The Corinthian congregation was rather large (cf. Acts 18.10) and probably met in different house circles (or house churches). Gaius, however, was able to assemble the “whole congregation” in his house. We can say that the house of Gaius must have been larger than average houses. In the flat of an insula there would not have been enough space for the “whole congregation”.’

32 Welborn, An End to Enmity, 324. Similarly, de Vos, C. S., Church and Community Conflicts: The Relationships of the Thesssalonian [sic], Corinthian, and Philippian Churches with their Wider Civic Communities (SBLDS 168; Atlanta: Scholars, 1999) 204 n. 98Google Scholar.

33 Welborn, An End to Enmity, 328–79, at 378.

34 Zahn, An die Römer, 614.

35 Dunn, Romans, 910; Jewett, Romans, 980.

36 Dio Chrysostom 7.5, 10, 37, 39, 71, 78, 82, 88, 91, 92, 93, 94, 97, 141.

37 Pollux, Onomasticon 1.74. BDAG 684.2.c cites Il. 15.532, Melito of Sardis and Xenophon as instances of ξένος meaning ‘host’. But this is hardly obvious in the case of Melito, Peri Pascha 51 (375): καὶ γὰρ πατὴρ ἐπὶ υἱὸν ξίφος ἐπηνέγκατο, καὶ υἱὸς πατρὶ χεῖρας προσήνεγκεν, καὶ μαστοὺς τιθηνοὺς ἀσεβὴς ἐτύπτησεν, καὶ ἀδελφὸς ἀδελφὸν ἀπέκτεινεν, καὶ ξένος ξένον ἠδίκησεν, καὶ φίλος φίλον ἐφόνευσεν, ‘for even a father lifted a dagger against his son; and a son used his hands against his father; and an impious man struck the breasts that nourished him; and brother killed brother; and xenos wronged a xenos; and friend murdered friend’. It is far from clear that the first ξένος means ‘host’ and the second ‘guest’. Likewise in Xenophon, Anab. 2.4.15: Μένωνα δὲ οὐκ ἐζήτει, καὶ ταῦτα παρ’ Ἀριαίου ὢν τοῦ Μένωνος ξένου, ξένος appears to mean ‘friend’: ‘he did not ask for Menon, despite the fact that he came from Ariaeus, Menon's friend’.

38 Last, R., The Pauline Church and the Corinthian Ekklēsia: Greco-Roman Associations in Comparative Context (SNTSMS 164; Cambridge and New York: Cambridge University Press, 2015) 62–71Google Scholar.

39 ἑστιάτωρ: Agora xvi.161.12, 24 = GRA i.14 (early 3rd cent. bce); πατήρ: IJO i Mac 1.4 = GRA i.73 (2nd–3rd cent. ce).

40 ἑστιάω: IG ii².1343.26 = GRA i.48 (Athens; 37/36 or 36/35 bce); ὑποδέχομαι: IEph 3080 (Ionia, Asia Minor; 167 ce); IEph 951 (Ephesos, Asia Minor; unknown date).

41 Stählin, G., ‘ξένος, ξενία, ξενίζω, ξενοδοχέω, φιλοξενία, φιλίξενος’, TDNT v (1968) 1–36, at 20Google Scholar.

42 Stock, S. G., ‘Hospitality (Greek and Roman)’, Encyclopedia of Religion and Ethics vi (1914) 808–12Google Scholar, at 808.

43 Last, Pauline Church and the Corinthian Ekklēsia, 66.

44 Last, Pauline Church and the Corinthian Ekklēsia, 67–8.

45 Last's lists, supplemented with additional examples: guests simpliciter: SB iii.7182, fr.1.2.12–26; fr. 2.39; PTebt. i.118.4, 12; i.224 recto.3; iii/2.894, fr. 2 recto.1.5, 12; fr. 2 recto.2.37; fr. 5 verso.2.16; guests of members: PTebt iii/2.894, fr. 4 recto.l.8: ξένοι Ἡρακλείδ[ου]; and fr. 5 verso.2.16, which indicates ξένοι θ’ (9 guests) and then itemises those guests by the member who invited each one. Only one name is still legible (Καγῶς ἀ(νὰ) ξ’).

46 Last, Pauline Church and the Corinthian Ekklēsia, 71.

47 The question naturally arises as to whether ‘Gaius’ is a praenomen or a cognomen, especially in Greek-speaking areas and in the early Imperial period. Judge (reported by Welborn, An End to Enmity, 291 n. 18) suggested that Gaius might be a cognomen, since Lucius and Marcus also occur as cognomina (or as single Greek names, see below, n. 48). Welborn (ibid.) rightly regards ‘Gaius’ in Rom 16.23 as a praenomen. According to Salomies, O. (Die römischen Vornamen: Studien zur römischen Namengebung (Commentationes Humanarum Litterarum 82; Helsinki: Societas Scientiarum Fennica, 1987) 164–5)Google Scholar, using praenomina as cognomina ‘ist jedoch nie besonders üblich gewesen und scheint vor der Mitte des 2. Jh. kaum belegt zu sein’ (164). These occur mainly in the lightly romanised areas: ‘man wenigstens in Italien und in den romanisierten Westprovinzen bis in diese Zeit [3 Jh.] zumindest eine Ahnung davon gehabt haben muss, was ein Pränomen war, und wie es sich von einem Cognomen unterschied’ (165). Since Corinth was a thoroughly romanised city, it is likely that at least in the first century ce the distinction between a praenomen and a cognomen was still observed. Solin, H. (‘Latin Cognomina in the Greek East’, The Greek East in the Roman Context: Proceedings of a Colloquium Organised by the Finnish Institute at Athens, May 21 and 22, 1999 (ed. Salomies, O.; Helsinki: Suomen Ateenan-instituutin säätiö, 2001) 189–202, at 195–6Google Scholar) notes that Greeks adopted Latin praenomina as personal names beginning in the second century bce, with the bulk of occurrences in the first and second centuries ce. Solin's data, however, all comes from Athens. For Corinthia, however, there is only one possible example, which is too late: Rizakis, A. D. and Zoumbaki, S. (Roman Peloponnese, vol. i (Athens: Kentron Hellēnikēs kai Rōmaikēs Archaiotētos, 2001) 275, no. 116Google Scholar) adduce Γ(αϊος) Αλέξανδρος ΓΟ[3–4]ΟΔΟ[– – –] (= ICorinth.Meritt 15.58 = SEG xi.62), dated to the ‘latter part of the second century ad’. Rizakis and Zoumbaki i.401, no. 661: C(AIUS) [– – –] is probably not an instance of a praenomen as a name since it is a dedicatory inscription to an agonthete who undoubtedly bore a tria nomina. Rizakis and Zoumbaki i.405, no. 685: ΓΑΪΟ[Σ] dates from the second century ce.

48 Ramsay (‘A Historical Commentary’, 101 n. 2) observes that in Asia Minor name like Gaius or Lucius were assumed by provincials as a single name (like a Greek name), which in those cases does not imply Roman citizenship. He notes that this was not common in Greece at this time, ‘but belonged rather to the less educated cities'. See also n. 47.

49 One of the referees rightly points out that if the Gaius of 1 Cor 1.14 is not necessarily the same Gaius as that in Rom 16.23, and if there is no reason to connect Erastos of Rom 16.23 with the aedile Erastus of I.Corinth.Kent 232, then the link between the latter Gaius and Corinth is severed, and Romans might have been penned in some other location. Friesen, S. (‘The Wrong Erastus: Ideology, Archaeology, and Exegesis’, Corinth in Context: Comparative Studies on Religion and Society (ed. Friesen, S. J., Schowalter, D. N. and Walters, J. C.; NovTSup 134; Leiden and Boston: Brill, 2010) 231–56)CrossRefGoogle Scholar has indeed shown that Erastos of Rom 16.23 is not the Erastus of the Corinthian inscription, notwithstanding the special pleading of Brookins, T. A., ‘The (In)Frequency of the Name “Erastus” in Antiquity: A Literary, Papyrological, and Epigraphical Catalog’, NTS 59 (2013) 496–516CrossRefGoogle Scholar. Dunn's claim (Romans, xliv) that there is ‘scarcely any dispute’ about a Corinthian provenance for Romans is too strong. Nevertheless, the fact that Phoebe of Cenchreae is named at the head of Romans 16 as the bearer of a letter of introduction and greetings, and that Rom 15.25 indicates that the Achaian collection is now complete and that Paul is on his way to Jerusalem makes Corinth or Cenchreae the most likely location for the writing of Romans. Yet even if Romans were penned in some other location, it must be assumed that, wherever it was, the Christ group there had welcomed Gaius, whom the Romans must know, as its guest.

50 Lucius (Λούκιος) is obviously another Latin praenomen, but is sometimes treated as a single Greek name: Judge (‘Roman Base’, 112) notes that whether it is treated as a Latin or a Greek name in Rom 16.21 hangs on the meaning of ‘relative’. From Corinthia, the clearest examples of Lucius as a Greek name (not a Latin praenomen) are from the second century ce: Rizakis and Zoumbaki, Roman Peloponnese, i.345, nos. 379, 382 = ICorinth.Meritt 95 (2nd/3rd cent. ce): Λούκ[ιος – – –] Λουκίου [υἱός]; i.345, no. 380 (2nd/3rd cent. ce): Luciu|s [et] Cratinus; i.281, no. 140 = ICorinth.Kent 353 (early Imperial period): ΛΟΥΚΙΟΣ ΚΑΝ[Ι]ΟΣ, which might also be read as ΛΟΥΚΙΟΣ ΛΑΝ[Ι]ΟΣ, ‘Lucius lanius’, i.e. the butcher, but Solin, H. and Salomies, O. (Repertorium nominum gentilium et cognominum Latinorum (Hildesheim: Olms/Weidmann, 1988))Google Scholar report Lanius as a gentilicum, in which case Lucius is a praenomen; i.406, no. 692 (1st cent. ce): LUCI(US), on the base ring of a small lamp; i.406–7, no. 693.1–7 (2nd cent. ce): ΛΟΥΚΙΟΥ, on the reverse of seven oil lamps, probably potter's marks. As a praenomen: ICorinth.Kent 276.5 (250–300 ce): [– –]ωι Λευκίου ἀπελ[ευθέρωι Δηλμ – – ]; uncertain: SEG xi.61.24 (Corinth, 3rd cent. ce) … c. 6… Λεύκιο[ς ..c. 3.]ρα̣τ̣ε – – – –.

51 Salway, B., ‘What's in a Name? A Survey of Roman Onomastic Practice from c. 700 bc to ad 700’, JRS 84 (1994) 124–45Google Scholar, at 125. These are Aulus, Appius, Gaius, Gnaeus, Decimus, Lucius, Manius, Marcus, Numerius, Publius, Quintus, Servius, Sextus, Spurius, Titus, Tiberius and Vibius.

52 The PHI database of Greek inscriptions attested ‘Gaius’ in the Peloponnese 211 times in 154 inscriptions. The Corinthia section in Rizakis and Zoumbaki, Roman Peloponnese lists 70 Gaii, almost all part of tria nomina.

53 Salomies, Die römischen Vornamen, 29.

54 Assuming a population of Corinth at 87,000 (Engels, D. W., Roman Corinth: An Alternative Model for the Classical City (Chicago: University of Chicago Press, 1990) 82Google Scholar), two thirds of the population adult, with half of the population represented by slaves and freeborn Greeks (neither of whom had a praenomen), one might expect that of the approximately 14,350 male ingenui and liberti, about 2,900 would have the name ‘Gaius’.

55 I assume for the sake of argument that one third of the adult Roman population were slaves (and hence with no praenomina), and the remaining two thirds (with men being somewhat more numerous than women) included ingenui and liberti (who would bear a tria nomina).

56 R. S. Ascough (‘Implications of Association Meeting Places for Imagining the Size of Pauline Christ Groups’, unpublished paper presented at the 2014 Annual meeting of the Studiorum Novi Testamenti Societas, Szeged Hungary (2014), 7) has suggested that the ‘Roman group’ was comprised of five separate groups who may not ever have met together as a whole. This led him to ask (per litt.) whether we could assume that all of the Roman Christ-followers would know which Gaius was meant by Rom 16.23. Yet however many sub-groups in Rome there might be, Paul evidently assumes that his greetings will be conveyed to each. Whether the Gaius in question is known to each sub-group, it will be obvious to the group of which he was a member which Gaius had gone to Corinth, and it will be obvious to each of the others that Paul is conveying the greetings of a Roman Gaius now in Corinth, whether they know this Gaius personally or not. Recently, Stowers, S. (‘The Social Formations of Paul and his Romans: Synagogues, Churches, and Ockham's Razor’, A Most Reliable Witness: Essays in Honor of Ross Shepard Kraemer (ed. Harvey, S. A. et al. ; Brown Judaic Studies 358; Providence, RI: Brown Judaic Studies, 2015) 77–87)CrossRefGoogle Scholar has challenged the entire idea of Paul sending a letter to the ‘Roman church’, suggesting instead that Romans, because of its specialised and learned nature, cannot have been addressed to a ‘general Christian or Jewish population’, but is directed to a highly educated, specialised and apparently very small audience. It is unclear from Stowers' account what he thinks the relationship is between those greeted in Romans 16 and the actual addressees. He cites Mullins, T. (‘Greeting as a New Testament Form’, JBL 87 (1968) 418–26, at 420)Google Scholar: ‘In this way [viz., by using a second-person greeting: ‘Greet NN’], the writer of the letter becomes the principal and the addressee becomes his agent in establishing a communication with a third party who is not intended to be among the immediate readership of the letter.’ But for Mullins, these greeting formulae, as well as the third-person formula (‘NN greets you’), ‘informs us chiefly of relationships which exist beyond the writer-reader dialogue and beyond the specific occasion of the letter’ (422). While this might imply that those greeted are not intended as the primary recipients of the letter, it does at least mean that they belong to the network of the letter recipient. And it implies that those offering the third-person greetings do belong to the immediate network of the recipient. Hence we should expect the immediate addressees to know the identity of Gaius. (I thank Richard Last for directing me to Stowers' essay.)

57 The research for this paper was supported by the Social Sciences and Humanities Research Council of Canada and a working group on Greco-Roman associations, including (then) doctoral students, Callie Callon, Richard Last and Sarah Rollens, now all having completed and defended their respective dissertations. I would like to thank Richard Ascough, Phil Harland, Edwin Judge, Richard Last and Mariana Mastagar for reading versions of this paper and sparing me from various errors and omissions. Finally, I thank the anonymous referees for NTS for suggestions for improvement.