¹ Eschewing the prejudicial ‘mistress’, ‘entirely wrong’ according to O'Donnell, J. J., Augustine: Confessions, vol. 2: Commentary on Books 1–7 (1992), 384Google Scholar n. 10 and incorrect ‘au sens strict’ per A. Solignac in Tréhorel, E., Bouissou, G. and Solignac, A., Les Confessions BA 13 (1962), 679,Google Scholar and the technically correct, but unacceptable-in-English ’concubine’. She appears as ‘La mère d'Adeodat’ in the Augustinus-Lexikon, ed. Mayer, C., vol. 1 (1986–1994), 87–9Google Scholar.

² Their total time together could have been thirteen to fifteen years: A.D. 370/72–385.

³ Shaw, B. D., ‘The family in Late Antiquity’, Past and Present 115 (1987), 3–51,CrossRefGoogle Scholar at 45, distorts what Augustine says, ‘he therefore rid himself of the concubine by rudely dismissing her back to Africa’.

⁴ See also Conf. 6.13.23.

⁵ Conf. 6.13.23 ‘maxime matre dante operam’.

⁶ Brown, P., Augustine of Hippo (1969), 89Google Scholar.

⁷ See TLL s.v. ‘impedimentum’ 528.74–84.

⁸ i.e. not ‘because she was a hindrance’, (Chadwick, H., Saint Augustine: Confessions (1991), 108)Google Scholar but ‘on the grounds that she was a hindrance’. See below for a self-distanced virtual oratio obliqua used of Augustine's own self-deceptive pretext, namely that he would have to wait two years.

⁹ The same passives adorn Augustine's account of the search for a new wife. See Conf. 6.13.23 ‘et instabatur impigre … instabatur … petebatur … expectabatur’. Note also the impersonal ‘quia ea placebat’, and Conf. 2.2.2 ‘non tenebatur modus’ for Augustine's adolescent sexuality.

¹⁰ Compare the death of Patricius at Conf. 4.4.7 ‘defuncto patre ante biennium’. Note especially the ellipsis of the subject of the ablative absolute, the implied antecedent of qua.

¹¹ See Old Latin Judges 19.24, as preserved at, for example, Ambrose, De Off. 3.19.114, ‘Tune senior filiam suam virginem et coaequalem ejus cum qua cubitare solitus esset, offerebat viris iniquitatis, tantum ne vis irrogaretur hospiti’ compared to the Vulgate ‘et hie homo habet concubinam’.

¹² TLL s.v. ‘concubina’. Also Treggiari, S., Roman Marriage: Iusti Coniuges from the Time of Cicero to the Time of Ulpian (1991), 52Google Scholar.

¹³ Brown, op. cit. (n. 6), 89.

¹⁴ Civ. Dei 16.25 and 16.34; Quaest in Hept. 1.70, 90, and 124; De bono coniugali 14.16.

¹⁵ cf. the Ars Bernensis in Gramm. Lat. Suppl. p. 74.1 Keil ‘ut cubo concubina’.

¹⁶ Brown, P., The Body and Society (1988), 389Google Scholar.

¹⁷ See DGAL 9.5.9 ‘Quanto enim congruentius ad convivendum et conloquendum duo amici pariter quam vir et mulier habitarent!’

¹⁸ See Conf. 4.5.10 ‘lenitum est vulnus meum’.

¹⁹ See O'Meara, J., The Young Augustine (1954), 129Google Scholar, for derision of sentimental scholars, ‘the best years of her life’, and practicality, ‘Augustine had no duty to marry his mistress’, followed by a relentment, ‘she was well loved by him for many years’. For a modern novel about Anonyma 1, see Gaarder, J., That Same Flower: Floria Aemilia's Letter to Saint Augustine, trans. Born, A. (1998)Google Scholar.

²⁰ O'Donnell, op. cit. (n. 1), 385.

²¹ e.g. Shaw, op. cit. (n. 3), 45, ‘Augustine kept a concubine for the purposes of sexual enjoyment for a period of at least fourteen years’.

²² See O'Donnell, op. cit. (n. 1), 384 n. 10.

²³ I disagree with Zumkeller, A., ‘Die geplante Eheschlieβung Augustins und die Entlassung seiner Konkubine’, in Signum Pietatis: Festgabe für Cornelius Petrus Mayer, OSA (1989), 21–35, at 24Google Scholar, who sees here a sign of respect for the ‘Unbekannte’.

²⁴ A latere can naturally also mean ’on the side’ in some contexts, e.g. a latere occidentis, or ‘at the side’, as in Ps. 90.7 ‘cadent a latere tuo’. Ex latere can indicate ‘made out of’, as in ‘mulier ex latere iam facta erat’ or ‘eique formata uxor ex latere’. De covers some of the same range, i.e. ‘material’ as in DGAL 6.46 ‘mulier illi de latere’; Enarr. in Ps., PL 47.1324 ‘de latere coniunx fieret ecclesia’; but ex also can indicate separation and source, e.g. Contra Faustum, PL 42.274 ‘sacramenta ecclesiae manentia ex latere hominis illius’; as can de, cf. Civ. Dei. 22.17 ‘de latere viri dormientis costa detracta femina fieret’.

²⁵ See Conf. 8.11.27 ‘Alypius affixus lateri meo’.

²⁶ It appears in DGAL combined with a number of different prepositions, e.g. 9.17.31 ‘ex viri latere feminam fieri’; 9.18.34 ‘quod ita mulier facta est de latere viri’; 10.1.1 ‘at ilia de illius latere’. DGAL 9.16.30 ‘virile autem latus unde femina fieret non habebat’.

²⁷ e.g. Civ. Dei. 22.17 ‘de latere viri dormientis costa detracta femina fieret’.

²⁸ De Quant. An. 31.62–3, PL 32.1070 ‘cur nonnullum animal concisum in omnibus partibus vivat’; In Joh. Evang. Tract. 11.5, PL 35.1477 ‘Jesum, quod eum possent concisum sicut agnum coquere’.

²⁹ See DGAL 9.1.1.

³⁰ Both in the Vulgate and in most Itala versions. See Jülicher, A., Itala 1. Matthäus-Evangelium (1972), 133Google Scholar. Coniungetur and adiungetur are two occasional variants, the latter being found in DGCM 2.1 PL 34.196. The passage is also quoted at Eph. 5.31–2.

³¹ Mt. in the Greek New Testament reads κολληθήσεται.

³² Mt. 19.6 ‘Quod ergo Deus coniunxit, homo non separet’.

³³ i.e. he may have had ‘marital intent’. See Clark, G., Women in Late Antiquity (1993), 31–3.Google Scholar

³⁴ Chadwick, op. cit. (n. 8), 109.

³⁵ F. Troncarelli, Il Ricordo della sofferenza. Le Confessioni di Sant'Agostino e la psicoanalisi (1993), 167–8.

³⁶ J. N. Adams, The Latin Sexual Vocabulary (1982), 190.

³⁷ Also, naturally, Conf. 4.2.2 ‘non quod legitimum vocatur coniugio mihi cognitam’.

³⁸ See Zumkeller, op. cit. (n. 23), 33.

³⁹ Given the loaded nature of God's question, and the chain of exculpation that follows, Adam's reply was almost certainly intended as a plea of innocence and an attempt to ‘pass the buck’, not as a bald statement of fact. See DGCM 2.17.25 ‘Deinde iam more superbiae in se non accusat quod consensit mulieri, sed in mulierem refundit culpam suam … voluit ad ipsum Deum pertinere quod peccavit’. Also DGAL 11.35.47.

⁴⁰ Augustine however in DGAL 11.42.59 also read Adam's imitation of the woman as due not to concupiscence of the flesh, but to amicalis benevolentia, the desire not to make her unhappy.

⁴¹ DGAL 9.18.34 ‘Quae per ipsum firma facta est, tamquam eius osse firmata, ille autem propter ipsam infirmus, quia in locum costae non costa sed caro suppleta est’. Already present in De Genesi contra Manichaeos 2.13.18 ‘Hoc nunc os … os de ossibus: fortasse propter fortitudinem’.

⁴² Mt. 26.41; Me. 14.38.

⁴³ p. 163.

⁴⁴ Brown, op. cit. (n. 16), 393.

⁴⁵ For fuller documentation of Augustine's use of the word, see the Chadwyck-Healey Patrologia Latina Database. All literal uses in Augustine apply to the devil's allies, Circumcelliones, those associated with Gildo, and Manichees. For an interesting figurative use in a similar context to that of the Confessions, see In Joh. Evang. Tract. 41.12 ‘quia non poterat facere ut non concupisceret: faciebat tantum ut concupiscentiam refrenaret, ut concupiscentiae non consentiret, et concupiscentiae membra ad satellitium non praeberet’.

⁴⁶ Brown, P., Augustine and Sexuality, The Center for Hermenuetical Studies (1983), 3 translates consuetudo throughout this passage as ‘habit’.Google Scholar

⁴⁷ TLL s.v. ‘consuetudo’ 561.46–75. i.q. amor, concubitus, matrimonium.

⁴⁸ For further passages in the Confessions where it may have the same meaning, see Conf. 6.12.21 ‘delectationes consuetudinis meae’; Conf. 8.5.12 ‘lex enim peccati est violentia consuetudinis, qua trahitur et tenetur etiam invitus animus eo merito, quo in eam volens inlabitur’; Conf. 8.5.13 ‘et de vinculo quidem desiderii concubitus quo artissimo tenebar’; Conf. 8.7.18 ‘remanserat muta trepidatio et quasi mortem reformidabat restringi a fluxu consuetudinis, quo tabescebat in mortem’; Conf. 8.11.26 ‘cum diceret mihi consuetudo violenta “putasne sine istis poteris?”’

⁴⁹ De Doctr. Christ. 3.18.26–7 ‘Item cavendum est ne forte, quod in Scripturis veteribus pro illorum temporum condicione, etiamsi non figurate, sed proprie intellegatur, non est flagitium neque facinus, ad ista etiam tempora quis putet in usum vitae posse transferri. Quod nisi dominante cupiditate, et ipsarum quoque Scripturarum, quibus evertenda est, satellitium quaerente, non faciet; nec intelligit miser ad hanc utilitatem ilia sic esse posita, ut spei bonae homines salubriter videant et consuetudinem quam aspernantur posse habere usum bonum, et earn quam amplexantur esse posse damnabilem, si et ibi caritas utentium, et hic cupiditas attendatur. Nam si multis uxoribus caste uti quisquam pro tempore potuit, potest alius una libidinose. Magis enim probo multarum fecunditate utentem propter aliud, quam unius carne fruentem propter ipsam. Ibi enim quaeritur utilitas temporum opportunitatibus congrua, hic satiatur cupiditas temporalibus voluptatibus implicata inferiorisque gradus ad Deum sunt, quibus secundum veniam concedit Apostolus carnalem cum singulis conjugibus consuetudinem propter intemperantiam eorum (I Cor. VII, 2), quam illi qui plures singuli cum haberent, sicut sapiens in cibo et potu nonnisi salutem corporis, sic inconcubitu nonnisi procreationem filiorum intuebantur’.

⁵⁰ I owe to the late Harry Jocelyn references to such passages as Caecilius' Plocium in Gellius, NA 2.23; Mostellaria 692, 699, and 703; and Aulularia 158 and 167–9. Modern ideas on the topic continue in the quip that the husband of a rich wife has ‘two sets of cheeks to kiss’. Jerome as always faced the issue squarely. See Contra Jov. 1. 47 ‘pauperem (sc. uxorem) alere difficile est, divitem ferre tormentum’.

⁵¹ Conf. 6.19. The fullest statement is in Solil. 1.17.4 cited above p. 157. The male professor's dream graduate student, the ‘litterata, vel quae abs te facile possit erudiri’, as admiring, unpaid research assistant!

⁵² Mt 17.10 ‘adveniat regnum tuum’.

⁵³ See Conf. 6.11.19 ‘Si feminae privarer amplexibus et medicinam misericordiae tuae ad eandem infirmitatem sanandam non cogitabam’.

⁵⁴ DGAL 9.15.26 ‘aliter ergo quaeritur quemadmodum sit soporatus Adam costaque eius sine ullo doloris sensu a corporis compage detracta sit’.

⁵⁵ See O'Donnell, op. cit. (n. 1), 384–5; Brown, op. cit. (n. 16), 393.

⁵⁶ Zumkeller, op. cit. (n. 23), 34, despite his distinction between two types of concubinatus, clearly identifies Augustine's relationship with Anonyma 1 as ‘Konkubinat auf Zeit’. See also O'Donnell, op. cit. (n. 1), 384: ‘In a nearly contemporaneous passage that must refer to this relationship’ (Italics mine).

⁵⁷ Zumkeller, op. cit. (n. 23), 32, who nonetheless identifies the second type of concubinage with Augustine's first relationship (ibid., 34). Augustine recognizes the concept of temporary concubinage at De bono coniugali 14.16.

⁵⁸ Pace O'Donnell, op. cit. (n. 1), 386: ‘No mention is ever made again of the second concubine’.

⁵⁹ Zumkeller, op. cit. (n. 23), 34, assumed that he knew from the start that he was going to leave her: ‘Die Entlassung war … war also von Anfang an eingeplant’.

⁶⁰ Conf. 6.13.23 ‘Et instabatur inpigre ut ducerem uxorem …’ For Augustine's following the path of least resistance, one might compare his decision to stay with the Manichees after being disillusioned by meeting Faustus. See Conf. 5.7.13 ‘ceterum conatus omnis meus, quo proficere in ilia secta statueram, illo homino cognito, prorsus intercidit, non ut ab eis ominino separarer, sed quasi melius quicquam non inveniens eo, quo iam quoquo modo inrueram contentus interim esse decreveram, nisi aliquid forte, quod magis eligendum esset, eluceret’.

⁶¹ cf. DGAL 9.7.12 ‘Hoc autem tripertitum est; fides, proles, sacramentum. In fide attenditur ne praeter vinculum coniugale cum altera vel altero concumbatur: in prole, ut amanter suscipiatur, benigne nutriatur, religiose educetur: in Sacramento autem, ut coniugium non separetur, et dimissus aut dimissa nec causa prolis alteri coniungatur’.

⁶² This clause permits his first concubinage to qualify.

⁶³ Inveniat throws dust in prying eyes and serves to disguise his own case somewhat. He had technically ‘found’, i.e. ‘located’, or become engaged to his future wife, but he had not yet gained possession of her.

⁶⁴ Adulter ‘avant la lettre’ so to speak: he has committed adultery in his heart with his promised wife.

⁶⁵ The obfuscation in inveniat necessitates this embarrassing explanation of why he is committing adultery in his heart when involved with the temporary concubine: in truth he had already ‘found’, i.e. become engaged to, his future wife.

⁶⁶ This should probably be taken as equivalent to Mt.5.28 ‘iam moechatus est eam in corde suo’.

⁶⁷ The obvious alternative paired opposition was the less biblical vir and femina. Gal. 3.28 has ‘non est masculus et’, echoed by Justinian, Novel 5.2.

⁶⁸ For example it has been assumed that Augustine dismissed his concubine in A.D. 385 because higher officials (such as professors of rhetoric) could not have concubines. See Zumkeller, op. cit. (n. 23), 27, who states that a marriage to her in A.D. 385 would have been not just ‘undenkbar’, but also ‘gesetzlich unmöglich’. But this false conclusion is derived from interpretation of Augustine's actions to the exclusion of the case of Libanius. Zumkeller, 35, contrasted the higher social class and greater security of Libanius who stayed with a concubine throughout his life, and claimed that Augustine's status was too low to get away with living in professorial concubinage. Thereby the saint could be exonerated: at that stage in his life he could not have married her — even had he wanted to.

⁶⁹ For more fine material along these lines see Zumkeller, op. cit. (n. 23), 21–2. Troncarelli, op. cit. (n. 35), 153, rightly speaks of our embarrassment about Augustine's conduct.

⁷⁰ Zumkeller, op. cit. (n. 23), 22, calls it research into the social and legal background of the relationship, but notes that it is not his intention to excuse Augustine and Monica ‘at any cost’. Troncarelli, op. cit. (n. 35), 157, characterizes such views as ‘il santo pur sofferendo, non si cura molto della sua donna’.

⁷¹ See the word plays, above p. 158.

⁷² Troncarelli, op. cit. (n. 35), 152–3, attributes the authority of this view to Solignac's notes in the Bibliothèque Augustinienne edition of the Confessions (op. cit. (n. 1), 677 ff.).

⁷³ Brown, op. cit. (n. 6), 62; Chadwick, H., Augustine (1986), 10Google Scholar, ‘a girl-friend of servile or low social class’; Evans-Grubbs, J., Law and Family in Late Antiquity: the Emperor Constantine's Marriage Legislation (1995), 295 and 300Google Scholar.

⁷⁴ Gillian Clark's secret explanation: that she was ‘an actress, from one of those sexy shows in Carthage’, because it would help to explain Augustine's obsession with theatre!

⁷⁵ Troncarelli, op. cit. (n. 35), 154.

⁷⁶ See below pp. 173–4.

⁷⁷ Troncarelli, op. cit. (n. 35), 154; at 157 he uses Libanius as a viable model for what Augustine could have done.

⁷⁸ See the epigraph above p. 157.

⁷⁹ Plassard, J. B., Le concubinat romain sous le haut empire (1921), 59Google Scholar, notes the fallacy: ‘Cette allégation suppose en effet que toutes les fois qu’un texte mentionne une concubine affranchie, l'homme avec qui elle vit appartient à la classe sénatoriale; elle suppose aussi que toutes les concubines ingénues sont, au moins au premier siecle, des mulieres famosae. Or la plupart de ces textes ne contiennent aucune preuve qui l'établisse, aucun indice qui permette de le supposer.’

⁸⁰ Rousselle, A., Porneia (1993), 100Google Scholar.

⁸¹ One could call this the ‘Lay the blame on Monica school’.

⁸² Conf. 2.2.4 covering November 369 to November 370.

⁸³ Lit. ‘turning points’.

⁸⁴ Or ‘with the result that’.

⁸⁵ Conf. 2.2.6 ‘quasi iam ex hoc in nepotes gestiret’.

⁸⁶ For the (telling and deliberate) biblical euphemism missed by commentators on the Confessions, see Adams, op. cit. (n. 36), 92–3. Jerome (Ep. 22.11) in the course of convincing Eustochium that the Devil invades both men and women through their private parts, makes an unpromising start with Job 40.16 — a description of Behemoth: ‘Virtus eius in lumbis eius et potestas eius in umbilico’, ‘See the strength in his loins, the power in his massive belly’. He cites various biblical passages where lumbus may refer to the male genitalia. The principle of euphemism is invoked: ‘it is only fit that men's and women's genitalia should be referred to by other names’, ‘Honeste viri mulierisque genitalia immutatis sunt appellata nominibus’. He then adduces an allusion to women's genitalia in in umbilico. This is not pure fantasy. The Septuagint has ἐπ ὀμΦαλοῦ γαστρός, (both sexes have navels), and the Hebrew word is apparently the same as that used in Ct. 7.3 to mean ‘vulva’. See Marvin H. Pope, Song of Songs (1977), 617: The word used means umbilical cord at Ezech. 14.4.

⁸⁷ Conf. 2.3.8.

⁸⁸ Conf. 2.3.8 ‘sed ibat in ceteris eius tardior’.

⁸⁹ The sicut sets up the following ‘ita curavit’, and has the force of a concessive qui-clause.

⁹⁰ Representing the ‘sicut… non ita’.

⁹¹ Meanings such as ‘to feel’, ‘to perceive’, etc. are impossible if cohercere (as is most natural) is construed with sentiebat rather than with curavit, as below n. 92.

⁹² Tréhorel and Bouissou, op. cit. (n. 1), 342, miss the ‘sicut’, take ‘cohercere termino’ etc. as the complement of ‘ita curavit’ (which at least is in line with Monica's actions), but run into trouble with the ‘quod’ and the ‘iamque … sentiebat’. They end up taking the same relative pronoun first as what Monica heard and then as (effectively) Augustine (depersonified as a pestilential thing). O'Donnell, likewise, op. cit. (n. 1), 125, believes that ‘ita curavit’ governs ‘cohercere’ and that the latter takes as its object the ‘compound relative clause “quod … sentiebat”’. Presumably this would yield something along the lines of ‘Thus she did not take care to confine within the limit of conjugal affection what she had heard from her husband and already sensed was diseased and henceforth a danger, etc’. But the parallel construction with the subsequent ‘non curavit hoc’ renders such a translation improbable and awkward. In addition, pestilentiosum and periculosum work better as attributes of a person, rather than of some abstract news heard. O'Donnell, 125, takes ‘non curavit hoc’ as ‘Finally the explicit answer to the question posed at the beginning of 2.2.3’, despite a separation of approximately four pages. ‘Non curavit hoc’ must rhetorically pick up ‘ita curavit quod’.

⁹³ Translators regularly resort to fancy footwork to take it that way, e.g. O'Donnell, op. cit. (n. I), 125; Chadwick, op. cit. (n. 8), 28, ‘and which she felt to hold a clanger for the future’, and R. S. Pine-Coffin, Saint Augustine: Confessions (10.61), 46, ‘She saw that I was already infected with a disease that would become dangerous later on … she did not think it right to restrain them to the bounds of married love’.

⁹⁴ Or, leaving the text as it is, ‘and [he] was voting …’

⁹⁵ The assessment (pestilentiosum and periculosum) may be Augustine's own retroactive one, not his father's opinion at the time.

⁹⁶ Augustine's affectus. This protasis may represent a concession of Patricius to Monica's opinion.

⁹⁷ ‘Ut caederer virgis ferreis ardentibus zeli et suspicionum et timorum et irarum atque rixarum.’ The symptoms are close to those that poison marriage. See De Sancta virginitate 16 ‘ipsam carnis tribulationem, quam praenuntiavit eis qui eligunt nuptias, in suspicionibus zeli conjugalis, in procreandis filiis atque nutriendis, in timoribus et moeroribus orbitatis. Quotus enim quisque, cum se connubii vinculis alligaverit, non istis trahatur atque agitetur affectibus?’

⁹⁸ See Brown, op. cit. (n. 6), 39; Chadwick, H., Augustine (1986), 10Google Scholar; Clark, G., Augustine: The Confessions (1993), 24Google Scholar.

⁹⁹ O'Donnell, op. cit. (n. 1), 203.

¹⁰⁰ Conf. 9.6.14.

¹⁰¹ O'Donnell, op. cit. (n. 1), 207, acknowledges the possibility: ‘“in illis annis” … suggest a liaison that must have begun in 371 or perhaps even 370, thus apparently probably in the first year of studies at Carthage, but conceivably during the year of indolence recorded at 2.3.5–6 (the philoprogenitive optimism of Patricius did not have long to wait).’

¹⁰² Conf. 2.3.8.

¹⁰³ As Brown, op. cit. (n. 6), 62 n. 4 takes it.

¹⁰⁴ Conf. 2.3.4 ‘non fuit cura meorum ruentem excipere matrimonio’.

¹⁰⁵ Particularly fornication with a woman who, if she was Anonyma 1, might have been marriageable. See Dig. 48.5.35 ‘Modestinus libro primo regularum Stuprum committit, qui liberam mulierem consuetudinis causa, non matrimonii continet, excepta videlicet concubina. Adulterium in nupta admittitur: stuprum in vidua vel virgine vel puero committitur’. There might always be the fear of the sort of charge alluded to by Ulpian and Aticilinus in Dig. 25.7.1.2 ‘Cum Aticilino sentio et puto solas eas in concubinatu haberi posse sine metu criminis, in quas stuprum non commititur’. Paulinus of Pella, Eucharisticon 159–86 has interesting comparative material on opportunities available to respectable young men (with resources). He prudently avoided liaisons not only with unwilling women, married women, and others' slaves, but also with women who merited special watchfulness, ingenuis oblatis sponte, free women offering either sexual relations or concubinages. He confined his attentions to slaves in his own household. Paulinus' parents nonetheless compelled him to marry at approximately eighteen.

¹⁰⁶ Patricius died in Augustine's seventeenth year (Conf. 3.4.7).

¹⁰⁷ Brown, op. cit. (n. 16), 390.

¹⁰⁸ For a concise statement of the issue, see H. Chadwick, ‘The attractions of Mani’, in E. Romero-Pose (ed.), Pléroma: Salus carnis. Homenaje a Antonio Orbe, S.J. (1990), 219. ‘A partner in bed and board from the lower classes of Carthage would hardly be acceptable at the governor's residence as hostess.’

¹⁰⁹ Conf. 6.6.9.

¹¹⁰ Necessity? Or obligation? Theophrastus, Cicero, and Jerome disagreed. See Jerome, Contra Jov. 1. 47 ‘Theophrastus de nuptiis … Non est ergo uxor ducenda sapienti. Primum enim impediri studia philosphiae, nec posse quemquam libris et uxori pariter inservire’ and 1. 48 ‘Cicero dicens post non posse se uxori et philosophiae pariter operam dare’.

¹¹¹ Or ‘relationship’, see Modestinus, Dig. 23.24 ‘In liberae mulieris consuetudine non concubinatus, sed nuptiae intellegendae sunt, si non corpore quaestum fecerit’.

¹¹² The protasis of the condition represents an original contrafactual condition in past time in oratio obliqua. The oratio recta would have been: ‘Multum interest inter illud quod tu expertus es, quod paene iam ne meministi quidem atque ideo nulla molestia facile contemnes et delectationes consuetudinis meae ad quas si accidisset honestum nomen matrimonii, non te mirari oportet cur ego illam nequeam spernere.’

¹¹³ It is important to note that the protasis of the condition is accidisset, implying simply that it had not happened, not that it could not have happened.

¹¹⁴ While a monogamous concubinage was acceptable, both a wife and a concubine were not. See CJ 5.26.1 and A. Arjava, Women and Law in Late Antiquity (1996), 208.

¹¹⁵ Troncarelli, op. cit. (n. 35), 162.

¹¹⁶ Conf. 6.14.24 ‘sed posteaquam coepit cogitari, utrum hoc mulierculae sinerent, quas et alii nostrum iam habebant et nos habere volebamus, totum illud placitum, quod bene formabamus, dissiluit in manibus atque confractum et abiectum est’.

¹¹⁷ Augustine's usage conforms to the definition at TLL s.v. ‘muliercula’ 1575.32–1756.33 mulier parva, miseranda, contempta, nee non familiaris. See, for example, Civ. Dei 10.16 ‘una muliercula’, ‘one weak woman’, Ep. 137.3.12 ‘quae abiecta muliercula’, or Contra Gaudentium 1.31.39 ‘blandienti mulierculae’ (of Delilah).

¹¹⁸ The tamen in 6.13.23 ‘instabatur tamen’ suggests that the dream, whatever its nature was, did not give a positive augury on Augustine's matrimony.

¹¹⁹ Unless the delectationes were those enjoyed with Anonyma 2. This seems unlikely given the respectful tone Augustine uses and the fact that he contemplates the addition of the name of matrimony to the relationship. See Conf. 6.15.25 ‘non utique coniugem’.

¹²⁰ See the persuasive discussion of Troncarelli, op. cit. (n. 35), 116–19, especially 119. To this one might add his own statements about ‘suppressio veri’ in the Contra mendacium, cited above p. 166.

¹²¹ See, for example, De bono coniugali 8 on divorce: ‘Ceterum aliter se habere iura gentilium quis ignorat?’

¹²² In De bono coniugali I. I he called the union of man and wife the ‘prima naturalis copula’ of human society.

¹²³ Augustine calls it ‘tamquam regula’ in which the fertility of nature is made honourable and the wickedness of incontinence is controlled.

¹²⁴ See TLL s.v. ‘dimitto’ 1210.69 ff. It means to repudiate a spouse, almost invariably a wife. Jerome, Ep. 55.4.5 ‘sive ipsa dimiserit virum sive a viro dimissa sit’ is a broad-minded exception.

¹²⁵ Compare De bono coniugali 7.7 ‘Quae si ita sunt, tantum valet illud sociale vinculum coniugum ut, cum causa procreandi colligetur, nee ipsa causa procreandi solvatur’. Also 15.17 ‘Manet enim vinculum nuptiarum, etiamsi proles cuius causa initum est, manifesta sterilitate subsequatur, ita ut iam scientibus coniugibus non se filios habituros separare se tamen vel ipsa causa filiorum atque aliis copulare non liceat’. Male infertility is addressed at 17.20: ‘ita uni feminae plures viros nee prolis ipsius causa, si forte ilia parere posset, ille generare non posset’.

¹²⁶ De bono coniugali 18.21.

¹²⁷ Even though the Church frowned upon divorce among Christians, it was permissible under Roman law and regularly occurred. See Evans Grubbs, op. cit. (n. 73), 242–53.

¹²⁸ Augustine acknowledged such an unlikely possibility on the part even of a temporary concubine in De bono coniugali 14.16.

¹²⁹ See Conf. 4.2.2. He would have passed the standard of the DGAL 9.7.12.

¹³⁰ See Plassard, op. cit. (n. 79), 85–6.

¹³¹ Contrast Augustine's verdict on Anonyma 2, above p. 164. For stuprum, see Evans Grubbs, op. cit. (n. 73), 217–18.

¹³² See Treggiari, op. cit. (n. 12), 51. Presumably analogous (aside from the case-history of impediment) to modern common-law marriage.

¹³³ Treggiari, op. cit. (n. 12), 52; Clark, op. cit. (n. 33), 32; Evans Grubbs, op. cit. (n. 73), 294.

¹³⁴ See Arjava, op. cit. (n. 114), 205–6.

¹³⁵ See Beaucamp, J., Le Statut de la femme à Byzance (4e–7e siècle) (1990–1992), vol. 1, 304;Google Scholar Clark, op. cit. (n. 33), 31; Arjava, op. cit. (n. 114), 205.

¹³⁶ See Plassard, op. cit. (n. 79), 35: ‘Cette intention qui est un fait d'ordre psychologique.’

¹³⁷ Treggiari, op. cit. (n. 12), 52. Rousselle, op. cit. (n. 80), 80–1, makes the same point: all respectable concubines could attain the status of matron while they stayed with their partners.

¹³⁸ Treggiari, op. cit. (n. 12), 280 cites Ulpian D. 48.5.14pr for the view that a freedwoman concubina could be considered a matrona and be prosecuted for adultery by her patronus, if she were in a sexual liaison with him.

¹³⁹ See Treggiari, op. cit. (n. 12), 52; Evans-Grubbs, 294; Arjava, op. cit. (n. 114), 205.

¹⁴⁰ Arjava, op. cit. (n. 114), 206.

¹⁴¹ Troncarelli, op. cit. (n. 35), 153–62 is the notable exception.

¹⁴² The debate concerns the words ‘excepta videlicet concubina’ (interpolation or gloss? or authentic?) in Digest 48.5.35pr. See Treggiari, S., ‘Concubinae’, PBSR 49 (1981), 73–4Google Scholar. For a recent discussion, see L. A. Olsen, La femme et l'enfant dans les unions illégitimes à Rome (1999), 166–9. The latter is concerned exclusively with the Republic and the early Empire.

¹⁴³ See Marcianus cited in Justinian, Digest. 25.7.4 ‘alioquin si honestae vitae et ingenuam mulierem in concubinatum habere maluerit, sine testatione hoc manifestum faciente non conceditur. Sed necesse est ei vel uxorem eam habere vel hoc recusantem stuprum cum ea committere’. This suggests that ingenuae were taken as concubines, but that a testatio was required, presumably to protect against the charge of stuprum.

¹⁴⁴ ‘Donationes in concubinam collatas non posse revocari convenit, nee, si matrimonium inter eosdem postea fuerit contractum, ad irritum recidere quod ante iure valuit’, discussed by Plassard, op. cit. (n. 79), 37. See Arjava, op. cit. (n. 114), 212–13, for legitimation by subsequent marriage.

¹⁴⁵ Beaucamp, op. cit. (n. 135), vol. 1, 297. There is a certain amount of inscriptional evidence in Plassard, op. cit. (n. 79), no, 112, 115 n. 2 (numerous examples), 135, 140–1, and 155–6. The difficulty with inscriptional evidence for concubinage between free individuals, as Plassard himself notes (161), is that it is impossible to determine the profession of the woman, and any, or all, could be women without conubium, actresses or prostitutes. Here the mute stones are silent. Treggiari, op. cit. (n. 142), 65 and 67, notes that one often cannot tell whether a free woman was free born or freed. Her study of Roman and Italian inscriptions (79 and 80–1) yields very few examples of ingenuae in concubinages.

¹⁴⁶ Beaucamp, op. cit. (n. 135), 298; Arjava, op. cit. (n. 114), 207.

¹⁴⁷ Historia Augusta, Marcus Aurelius 29.

¹⁴⁸ Treggiari, op. cit. (n. 142), 77, ‘Known freeborn concubines are few and not displayed’. An exception is the case discussed in Dig. 34.9.16.1 ‘Papinianus libro octavo responsorum Quoniam stuprum in ea contrahi non placuit, quae se non patroni concubinam esse patitur, eius, qui concubinam habuit, quod testamento relictum est, actio non denegabitur. idque in testamento Coccei Cassiani clarissimi viri, qui Rufinam ingenuam honore pleno dilexerat, optimi maximique principes nostn iudicaverunt: cuius filiam, quam alumnam testamento Cassianus nepti coheredem datam appellaverat, vulgo quaesitam apparuit’. See Plassard, op. cit. (n. 79), 73–84 and Olsen, op. cit. (n. 142), 168–9.

¹⁴⁹ Contrast the legal nightmare attested in Cicero, De Oratore 1.183 and 238.

¹⁵⁰ See Plassard, op. cit. (n. 79), 100–3, for some of the problems involved in assessing the epigraphic evidence to determine whether the relationship was a concubinage.

¹⁵¹ Unfortunately the poetry of blameless concubinage fails to rival in quantity even the slim volume of the poetry of married love, and the institution failed to come under the lash of satire.

¹⁵² Concubinages suffer from numerous disabilities in the competition for air time in the sources: they involved women, usually ones of lower social status, they were extra-legal, and were sometimes concealed.

¹⁵³ Arjava, op. cit. (n. 114), 208, notes the apparent disappearance of young wastrels' concubinages during the early Principate until the time when Church Fathers began to attack such arrangements.

¹⁵⁴ If one took a modern analogy, there are probably quite a few concubinages passing as marriages in life (and in statistics) and considerable numbers of real concubinages that simply cannot be recovered from written records.

¹⁵⁵ See Conf. 9.6.14 ‘Quod enim et nutriebatur a nobis in disciplina tua, tu inspiraveras nobis’.

¹⁵⁶ See Chadwick, op. cit. (n. 108), 218–19.

¹⁵⁷ Schmitt, E., Le Manage chrétien dans l'oeuvre de Saint Augustin (1983), 27Google Scholar, suggests that a Manichee Akousmatic would have preferred co-habitation to marriage (Chadwick, op. cit. (n. 108), 219 disagrees) and that marriage would have made a promotion to one of the Elect impossible.

¹⁵⁸ For quasi-marital relationships among the religious see Jerome, Ep. 22.14, who denounces subintroductae in terms that condemn the deliberate blurring of the borders between the wife, the concubine, and the whore: ‘Unde sine nuptiis aliud nomen uxorum? Immo unde novum concubinarum genus? Plus inferam: unde meretrices univirae?’

¹⁵⁹ Clark, op. cit. (n. 33), 33. Arjava, op. cit. (n. 114), 209, suggests that they normally were given custody of the children.

¹⁶⁰ For restrictions on donations and bequests to illegitimate children in the Constantinian marriage legislation, and the subsequent relaxation of prohibitions under Valentinian I in A.D. 371, see Evans Grubbs, op. cit. (n. 73), 213, and Arjava, op. cit. (n. 114), 212–14. The law at this date was presumably that of Cod. Theod. 4.6.4 (A.D. 371, apparently abrogated in A.D. 397 by Cod. Theod. 4.6.5), whereby a father with legitimate descendants might leave one-twelfth of his estate to his natural children and/or their mother. If he had no legitimate descendants, then one-quarter.

¹⁶¹ This may possibly be the implication of ‘vovens tibi’ in Conf. 6.15.25.

¹⁶² For the baptism, see Conf. 9.6.14 ‘sociavimus eum coaevum nobis in gratia tua, educandum in disciplina tua: et baptizati sumus …’

¹⁶³ In Augustine of Hippo Peter Brown, op. cit. (n. 6), 39, suggested that Augustine's concubinage was effectively the marriage that he wanted, but not one he enjoyed. ‘He had, in this way, got what he wanted: he had at last been “washed up on the shores of matrimony”. Whether he particularly enjoyed the experience is another matter.’ This view is favourably modified in Body and Society, op. cit. (n. 16), 390: ‘Augustine chose his companion because he loved her; and he slept with her because he loved to do so, and not so as to produce grandchildren for his mother or citizens for his home town.’ He thus laid the ground for the equation of concubinage with de facto marriage and equated the coniugale litus of Book 2 with the relationship with Anonyma 1, first mentioned in Confessions 4.

¹⁶⁴ It almost certainly could have illustrated the Rashomon syndrome too and seemed a marriage to Anonyma 1 and to Augustine and a temporary concubinage to Monica. Treggiari, op. cit. (n. 142), 61, allows for such situations: ‘Whether a given relationship is a marriage or not may therefore be privileged information. The will of both partners is needed to make a marriage; the lack of intention of one partner suffices to reduce the union to concubinatus, conceivably unbeknownst to the other.’

¹⁶⁵ See Shanzer, D. R., ‘Latent narrative patterns, allegorical choices, and literary unity in Augustine's Confessions,’ Vigiliae Christianae 46 (1992), 40–56Google Scholar.

¹⁶⁶ Lancel, S., Saint Augustin (1999)Google Scholar, no: ‘Ce qu'on sait de la banalité, alors, d'une telle pratique ne la rend pas moins choquante, quand son bénéficiaire s'appelle Augustin.’

¹⁶⁷ Solignac, op. cit. (n. 1), 679: ‘toute la netteté désiderable.’

¹⁶⁸ Aeneid 4. 172 ‘coniugium vocat, hoc praetexit nomine culpam’.

¹⁶⁹ Lancel, op. cit. (n. 166), 112 agrees. For a fine later example of the wound caused by the separation of man and wife's ‘one flesh’, see Neyt, F. and de Angelis-Noah, P. (eds), Barsanuphe et Jean de Gaza, Correspondance (1998)Google Scholar, Ep. 129, p. 487. I am grateful to Peter Brown for the reference.