² In using the words ‘ideals of marriage’ in this paper, I do not wish to suggest that people over long periods seriously set these ideals before them as objects of conduct and endeavour, but only that at an early stage the ideals had social and religious origins and were taken to be characteristic elements in the marriage-union. Then gradually (from a time, on the whole, later than Plautus and Terence) the ideals became conventional and approximated more to pious expressions which might suitably be inscribed on tombstones: that is, they were largely emptied of their practical implications for everyday conduct. This whole process of development will have been complex and uneven, and subject to all the vagaries of human nature.

³ On this poem, see Fraenkel, Eduard JRS XLV, 1955, 1 ffGoogle Scholar.

⁴ See, in particular, the detailed monograph of K. Meister, Die Hausschwelle in Sprache und Religion der Römer, Sitzber. d. Heidelb. Ak. d. Wiss., Phil.-Hist. Kl., 1924–5, 3 Abh.

⁵ This advice bears no relation to that of LXII, 59 ff., for there pugnare means ‘refuse to marry’: the young men are persuading the girls that marriage is desirable. This is not advice on married behaviour.

⁶ See n. 12, below.

⁷ The translation is that of F. C. Babbitt in the Loeb Moralia II.

⁸ The best account of this theory is to be found in Fraenkel, Plautinisches im Plautus, 292–313, who thought, however, that Plautus transferred the ending from another Greek play.

⁹ Plautinisches und Attisches, chap. II.

¹⁰ On what follows, see, for instance, Hermann-Blümner, Lehrb. d. griechischen Privataltertümer ³, 268 ff. On Roman marriage-customs, see, for example, Marquardt-Mau Privatleben ² 39 ff.; Blümner, Die röm. Privataltertümer, 345 ff.

¹¹ On Greek wedding-hymns, see Maas in P-W IX, s.v. ‘hymenaios’.

¹² Maas (l.c., col. 131) tries to claim this advice as Greek: ‘Die komischen Mahnungen sind vielleicht Parodien von Paraenesen in echten H. (zu tua vox superet vgl. Plut. Mor., 139 c.d.)’. But this is not legitimate: see above, pp. 16 f.

¹³ The words ‘integrae atque imperitae huic impercito’ (832 f.), may reflect another ritual phrase, behind the very obvious joke on the male bride. Themes depending on a contrast between the violence of the husband and the virgin bride are frequent in marriage contexts: cf. e.g. Catullus LXI, 56 ff., ‘tu fero iuveni in manus/floridam ipse puellulam/dedis a gremio suae/matris’ (also LXII, 23); Macrob. 1, 15, 21, ‘nuptiae in quibus vis fieri virgini videtur’. Plutarch several times in the quaest. Rom. touches on this theme of violence: e.g. quaest. 87. It was symbolized partly in the taking of the bride from her mother's arms. A plea for gentleness of treatment, especially if spoken by the mother, would be appropriate. The idea of the virgin bride is important, but is so taken for granted that it is seldom mentioned specifically; but its importance is, for instance, clear from the provision of special regulations for the marriage of widows (cf. e.g. Plut., quaest. Rom. 105; Macrob., 1, 15, 21). The theme is fairly frequent on epitaphs: cf. e.g. CIL 10, 3720; 12, 2244; 14, 1641, 5210; 13, 2189 and see Lattimore (see n. 29 below), p. 278 no. 98. Add statements like those of Val. Max., 2, 1, 3; 6, 1, 3–4; Servius on Ecl., 8, 29; also Juvenal, 6, 203 (the pretium virginitatis). To this circle of ideas belongs Virgil's specific statement about Dido, : Aen. I, 344 ff.Google Scholar, ‘cui pater intactam dederat primisque iugarat/ominibus’; this is one more aspect of Dido's single, complete and ideal marriage to Sychaeus (see below, pp. 23–5).

¹⁴ See further, p. 22 below. There is little satisfaction in giving this as a duet to two ancillae; for then Plautus has simply replaced mechanically Diphilos' good luck-song (continuing the wedding-hymn begun by the old man and Olympio), and the tone of the song has no significance in the mouths of these remote characters. Of course, the matrona and her slave might have sung it as a duet, but there is no sign of this and it seems unlikely. In fact, Cleustrata is the only character certainly on the stage in IV, 4—cf. (835) iamne abscessit uxor?

¹⁵ On this and related words, see Appendix below.

¹⁶ cf. the much later Christian epitaph in which a husband praises his wife for the same virtue, in more modern terms: Carm. Epigr. 761, 11 moribus apta meis. The wife in the Men. specifically rejected this virtue when she said (726), ‘vidua vivam quam tuos mores perferam’.

¹⁷ This scene will be discussed in detail below, pp. 22 ff.

¹⁸ The points are explained by Donatus on 11. 582, 592 and 593.

¹⁹ The Roman marriage reference in solum (293) also and its connection with Roman religion and marriage ideals should be noted: the point is discussed in detail below, p. 23 f.

²⁰ It is only to be expected, particularly in a people at once so conscious of history and tradition and so uniquely sensible of the formal elements in their customary observances, that ritual practices, which themselves often have an origin in social and economic facts, should have been a frequent source of ideals and symbols, even after the ritual had faded or was forgotten. For instance, praise of a wife's devotion to wool-making is a frequent theme in literature and on the epitaphs: cf. e.g. Carm. Epigr. 52, 8; 63, 14; 237; 492, 16; 1123; 1988, 14; 1996 (collected by Lattimore—see below) p. 23, n. 29; also e.g. Laud. Tur. 1, 30. I take it that such praise is not just to be understood literally, despite the fact that Augustus tried to instil the virtuous practice into his family (Suet. Aug. 64, 2Google Scholar). The wool-making for which busy women of affairs like Turia are praised symbolizes devotion to the home, and this duty of devotion was itself represented in the marriage-ceremony by the spindle and wool carried by the bride (see e.g. Blümner, o.c., 359 n. 1).

²¹ On the scanty evidence for what was said in Roman marriage-ceremonies, see Marquardt-Mau, o.c., 34 n. I, Blümner, o.c., 355 n. 8 and P-W, IV, cols. 198 f., 863.

²² On the personal status of a wife in a manus-marriage, see Schulz, Principles of Roman Law, 193 f., with the references quoted by him.

²³ Philologus XVII, 1861, 466.

²⁴ Hermes LXXVI, 1941, 128 ff.

²⁵ Museum Helveticum, VI 1949, 105 ff.

²⁶ Most editors since Ritschl, including Leo and Lindsay, have reversed the order of these lines (largely because the MSS begin 204 with solam). It is a modern way of thinking to feel that the order of expression should be (i) grounds or evidence, (ii) conclusion. In spoken language the conclusion is very often stated first and the grounds given in the second place, frequently in asyndeton but often in Latin with such a connective as ita: see e.g. Schmalz-Hofmann, 654 and Hofmann, Umgangsspr. 112 (the same phenomenon can be seen too in Greek, see Kühner-Gerth II, 344 ff.). Here, this consideration is reinforced by the fact that the repeated solus follows more naturally on the single occurrence in the previous line. The confusion of the MSS over the beginning of 204 should not outweigh these points.

²⁷ Very exceptionally the husband claims credit for faithfulness to one wife: cf. e.g. CIL I, 2, 1221. But clearly the normal view was that such faithfulness was a becoming and necessary virtue in woman, not in men. So too the eternal nature of the marriage-bond applies normally only to the wife. The essential onesidedness of the whole conception is neatly illustrated in Prop, IV, II, where Cornelia boasts of being univira but envisages Paullus' marrying again (see further below, p. 24).

²⁸ Roman in tone are passages like Plautus, Cist. 76 ff.; Merc. 824 ff.; and especially Ter., Heaut. 392 ff.

²⁹ cf. also the variations in Carm. Epigr. 558, 4, unicuba uniiuga. Lattimore, Themes in Greek and Latin Epitaphs, Illinois, 1942, 296 n. 251, adds further examples from CIL.

³⁰ See p. 18, n. 13 above.

³¹ The frequent occurrence of sine crimine on epitaphs should be noted: see Thes. LL. IV, 1192, 8 ff. and Pease, ad. loc.; also Lattimore, o.c., 279 n. 108 and 290 n. 202. That the phrase may already have found a place on epitaphs is perhaps suggested by Hor., Epist. I, 7, ss fGoogle Scholar., ‘Volteium nomine Menam,/praeconem, tenui censu, sine crimine …’, which, though it corresponds to the brief questions, also recalls the succinct asyndetic summaries of careers and mores, so characteristic of Roman epitaphs: cf. e.g., also on a praeco, CIL 1, 2, 1210, ‘prudentis hominis frugi cum magna fide praeconis Oli Grani sunt ossa heic sita’ (cf. also, e.g., 1, 2, 1211; 1221 (b); 1259; 1378; VI, 26192).

³² For the interpretation of more ferae, see R. G. Austin, Aeneid IV, ad. loc. The interpretation which he adopts, and which is demanded by the context, is completely vindicated by a glance at the numerous quotations adduced from ancient writers by Lovejoy, and Boas, , Primitivism and Related Ideas in Antiquity, Baltimore, 1935, chap, XIIIGoogle Scholar ‘The Superiority of the Animals’.

³³ Mr. Shackleton Bailey (Propertiana p. 265) is not justified in saying of this line: ‘Not necessarily an injunction against remarriage’. This is an injunction against remarriage and should be seen in the general context of the ideal univira. But, of course, few people would like in their lives to be held to what they hope will be written over their graves.

³⁴ In considering the themes of Greek epitaphs, I have derived much help from Lattimore's book quoted above, n. 29.

³⁵ In recommending virginity, on the usual grounds that it represents a marriage with Christ, Tertullian says (de monog. II) unius Christi unica sponsa, clearly recalling the type of thought recorded on the epitaphs quoted above.

³⁶ cf. also Felix, Minuc., Oct. 31, 5Google Scholar, ‘unius matrimonii vinculo libenter inhaeremus, cupiditate procreandi aut unam scimus aut nullam’.

³⁷ Of course not every occurrence of unus and solus in approximately marital contexts is to be taken as a reference to marriage-ideals; sometimes, as in Ter., Eun. 119–23, there may be a hint of such a reference (esp. in 122), but the main point is the joke with unus. In Asin. 163 ff.; 208, there is no hint of marriage.

³⁸ cf. e.g. Alexis frg. 146 Kock. It is significant that a piece of advice like Eur., Troad. 700 (φίλον διδοῦσα δέλεαρ ἀνδρὶ σῶν τρόπων) is given to a woman who is the slave of her captor, and that the advice is to play the part of a slave. The phrase may be compared with Amphis frg. 1, Kock 4 (τοίς τρόποις ὠνητέος ἄνθρωπός ἐστιν), describing a meretrix. Cf. the lena in Prop. 4, 5, 45, in mores te verte viri.

³⁹ Many inscriptions could be added: cf. e.g. CIL I, 2, 1570, huic autem obsequens; VIII, 23808, feminae marito obsequentissimae.

⁴⁰ To this aspect of the ideal marriage, Carm. Epigr. 959 (time of Sulla's death) gives excellent and moving expression. Many instances from literature could be mentioned: e.g. Publilius Syrus 108, ‘casta ad virum matrona parendo imperat’; 492, ‘obsequio nuptae cito fit odium paelicis’.

⁴¹ It was this aspect of marriage which Atalanta (apparently) felt to be characteristic of it and to which she so much objected: Accius, trag. 442 Klotz (442 R.) ‘quam invita ancillans dicto oboediens viri’ (on the use of dicto oboedire etc. to express obedience to the word of magistrates cum imperio, see Hermes 85 (1958), P. 97 n. 1).

⁴² The MSS of Festus read paupertatem; but perpetuitatem is suggested by Augustine and the excerpt of Paulus reads ut singulare perseveret matrimonium.

⁴³ So Propertius in II, 7 (see below) speaks of marriage, but constantly uses idealizing marriageterms of his relationship with Cynthia: unus etc.—I, 2, 26; II, 1, 47 f.; 13, 35 f.; 30, 23. solus etc.—II, 7, 19; 32, 55. Also, generally, 1, 12, 20. The eternal bond: II, 6, 42; 15, 31 ff.; 20, 15 ff.

⁴⁴ On this distinctive hair-style, see e.g. Marquardt-Mau, o.c., 45, n. 6.

⁴⁵ An additional argument to the same end is that the lines between the two speeches introduce, abruptly and incoherently, a topic which has already been settled in 168 ff. (see further below), and that the lines are largely pilferings from the surrounding scene (on this, see Fuchs, o.c., 114 ff.).

⁴⁶ A place might possibly be found for the lines (on the hypothesis that they are simply misplaced in the MSS) after 181 or between 185 and 186. But there are three decisive objections to any such procedure: (i) the immediate relevance of (208) quae istuc cures to 167 is lost; (ii) the obvious pilferings from the rest of the section 186–246 are not accounted for; (iii) the repetition of 224 after 246 is left unexplained.

⁴⁷ It should be noted that whoever wrote these lines was adept at imitating and reproducing the true Plautine style: see Fraenkel, Plaut. im Pl. 37, n. 3.

⁴⁸ The lines 168–85 are simply a series of Plautine jokes; but, since one cannot be quite sure that they had no basis in the Greek play, it is safer to leave them out of account.

⁴⁹ It seems hard to take them (with Mr. Shackleton Bailey, Propertiana, 73) as ‘lose my passion for you’, in the absence of a genitive (for this reason Ovid, Rem. Am. 453, is not a parallel), faces (normally the object which inflames passion) might be taken in a more general way to mean true passion in contrast to the feeling aroused by a wife: so Ovid, , Heroid. 13, 160Google Scholar, makes Laodamia by a pun stress the conjunction of both— ‘perque pares animi coniugiique faces’. (It is true, however, that the juxtaposition of nuptae and faces is misleading and suggests the meaning ‘marriage torches’ for faces.)

⁵⁰ This hypothesis is unaffected by the question to what extent the free form of marriage (which existed at least from the time of the XII Tables side by side with the forms of manus-marriage) was, in fact, used: see e.g. Schulz, Principles 192 ff.; Jolowicz, Historical Introd. to the Study of Roman Law ² 112 ff.

⁵¹ I can find no indication whatever that morem gerere etc. have Greek models. Such Greek phrases as come nearest in meaning to them have a fundamental difference of viewpoint: e.g. Aesch. ap. Aristoph. Ran. 1479 τοῖς τρόποις ὑπηρετεῖν, or perhaps τροποφορείν (Cic, ad. Att. 13, 29, 2 and Schol. ad Aristoph., Ran. 1433; the word may be a Hellenistic colloquial invention. For τῡφος with τρόπος, cf. e.g. Antiphanes 195, 1 f. Kock). In Greek the point of reference is always the τρόποι of the other person (i.e. of the object or implied object); in Latin it is the mos of the person primarily concerned (i.e. of the subject or implied subject). Passages such as those quoted in note 38 above are different, since they are formally very remote and since there τρόποι has a narrowly specific (usually sexual) significance.

⁵² In Lucr. 5, 80, ‘morigera ad fruges augendas atque animantis’, where the meaning of morigera is slightly widened by poetic adaptation, the gerundive construction with ad has been substituted for the personal dative.

⁵³ Similar to this, but surprisingly bold (if the text is sound) is Naevius trag. I Klotz (18 R.) ‘ne mihi gerere morem videar lingua, verum lingula’.

⁵⁴ This is clearly indicated by the pattern of its distribution among authors: e.g. rhet. ad Her., Varro, Cicero's letters, Propertius, Apuleius etc.

⁵⁵ ad Fam. 2, 17, 7; 6, 8, 3; ad Att. 3, 20, 3; 8, 10; 13, 42, 1; 15, 26, 1.

⁵⁶ See, for instance, Lindsay, Latin Language 485.

⁵⁷ Cf. e.g. Cicero, Orator 159, ‘voluptati autem aurium morigerari debet oratio’.