5 2 Marriage and divorce

(Inst.Gai.1.108.-15b., Inst.1.10., D.23.2.)

Marriage, or matrimony, is a joining together of a man and a woman, implying a united life­

style. (Inst. 1.9.1.)

Modestinus, Rules, book 1: Marriage is the union of a man and a woman, a partnership for the whole life involving divine as well as human law.

The word miitrimoninm—the norma! word for marriage—is partly derived from mater (mother). The emphasis on motherhood suggests that the Romans regarded marriage (at least on some level) as an institution that resulted in the birth of legiti­mate children. See Treggiari, S., Roman Marriage (1991), 5 ff. Marriage in Roman law was not a sacrament (a medieval canonist notion), but purely a social fact with certain legal consequences.

Although there were times, notably the late Republic, when the evidence points to a certain moral laxity and a casual approach to personal relationships, marriage was, nevertheless, regarded for much of Rome's history as a solemn union with important consequences. It was the prime way by which potestas could be acquired. The seriousness with which marriage was treated is demonstrated by the profusion of rules laying down various bars and impediments. These rules suggest that the Romans had a highly developed sense of what constituted a desirable and law­ful marriage. See Evans Grubbs, Women, 81-2. For a good recent survey, see now Urbanik, J., 'Husband and Wife', in OHRLS, 473-86.

5.2.1 Betrothal

(D.23.1.)

Betrothals were common in Roman society at all periods, especially among the wealthier classes. The betrothal ceremony would often be conducted by the respec­tive fathers. The presence of the happy couple was not necessary:

Ulpian, Eitiit, book 6: When betrothals are being contracted, it does not matter much whether this is done by the parties themselves (in person, by sending a messenger, or by letter) or by someone else.

The betrothal ceremony was usually a solemn affair, especially in early times, involving the exchange by the fathers of promises of marriage on behalf of their children. But solemnity was not obligatory: senators are said to have sometimes betrothed their children during the conviviality of a dinner party. In early Rome, a paterfamilias could insist on betrothing his child to whom he wished, but by the late Republic the consent of the parties was required. However, one cannot imagine that the wishes of a paterfamilias would often be thwarted (cf. D.23.1.12.).

At what age could children be betrothed?

Modestinus, Distinct ions, book 4: There is no fixed age for the parties in betrothal as there is in the case of marriage. So betrothal can take place at a very early age, provided that what is being done is understood by both parties, that is, as long as they are not under seven years of age. (D.23.1.14.)

The betrothal of young children to one another in their absence was not peculiar to Roman society. It occurred frequently in the past—e.g. in the dynastic marriages of European sovereigns—and still occurs in some societies of today.

In early law, it seems that betrothal promises were actionable if broken (because the ceremony was similar to the undertaking of a stipulation but by the late Republic they could no longer be sued on, even where a penalty had been agreed for breach. After the conversion to Christianity, betrothal again came to be regarded as of great importance, the law attaching to it some of the consequences of marriage (e.g. infidelity by a betrothed woman was regarded as adultery). Moreover, the prac­tice developed of exchanging tokens of good faith on betrothal. These tokens could be quite valuable. They were forfeited on breach of promise, Justinian allowing double forfeiture (i.e. the token was forfeited plus its equivalent value).

It must be stressed that despite its social and legal importance, betrothal was never a formal requirement of a valid marriage.

5.2.2 Formal requirements of marriage

A Roman civil law marriage (matrimonitim itistum) had to satisfy various require­ments: the parties had to be of the required age; they had to consent; they had to have comibium; and the marriage had to be free of bars and impediments. See Frier and McGinn, Casebook, 27-40.

5.2.2.1 Age

1'he basic rule was that marriage was not legally possible until the attainment of puberty. In early law, a physical examination of the parties was involved, but even­tually the ages were fixed as twelve for females, fourteen for males:

Roman citizens can marry each other when they are legally united, males having reached puberty and females having become marriagable... (Inst.l.lOpr.)

The ancients judged puberty in males not by age, but also by the development of their bodies. But we, from a desire to conform to the purity of the present times, have thought it proper that what seemed indecent even to the ancients, that is, the inspection of the body, should be thought equally so in the case of males... (Inst.l.22pr.)

The schools had disputed the issue, the Sabinians favouring a physical examina­tion, the Proculians typically preferring a fixed rule. The latter view prevailed.

Women tended to marry at a relatively young age. Men tended to marry at a higher age; that fact, coupled with their greater susceptibility to dying in war or civil dis­turbance, meant that widowhood was a frequent occurrence. Cicero's daughter was apparently betrothed at twelve, married at sixteen, and widowed at twenty-two, not an unusual scenario. The sensational Agrippina, Claudius' last wife, was first married at twelve. The general tendency to marry at a young age must be viewed in the context of the average life span of Roman citizens—probably below thirty years of age in the early Empire, see Gardner, Women in Roman Law, 38 ff.

The view of marriage as a means of increasing the birth rate was one particu­larly emphasized by Augustus. He imposed a duty to marry on men aged between twenty-five and sixty, and on women between twenty and fifty. Sanctions were enacted against those in breach of their duty (see 8.4.1.4), see also Evans Grubbs, Women, 88.

5.2.2.2 Consent

Consent to marriage has to be distinguished from consent to betrothal, e.g. a party could consent to a betrothal, but might then refuse to marry. The consent of the parties to marry was normally necessary because they had to have affectio maritalis, the intention to enter marriage. Consent could be proved in a variety of ways, par­ticularly by evidence that the couple had undergone the ceremonial customs tradi­tional in Roman marriages. For example, it was normal for the couple to exchange vows in the bride's home, normally in the presence of an auspex, a formal witness to a wedding (originally someone adept at reading omens from studying the flight of birds). There then followed a procession to the bridegroom's home, during which the guests often indulged in pranks and risque humour—not unlike the case at some weddings of today. The deductio in domum ('leading into the home') was a legally significant moment:

I’omponius, Submits, book 4: It is settled that a woman can be married by a man in his absence, either by letter or by messenger, if she is led to his house. But where she is absent, she cannot be married by letter or by messenger because she must be led to her husband's house, not her own, since the former is, as it were, the domicile of the marriage. (1) 23.2.5.)

The parties had to have the capacity to consent at the time of the marriage— supervening insanity did not invalidate it. If either party was in potestns, the con­sent of the paterfamilias was required (see 5.1.2.1).

5 2.2.3 Conubium

A Roman civil law marriage required that both parties should have conubium, the ins civile right to enter such a union. Roman citizens had conubium, but non­citizens did not (unless it was specially granted). So, a marriage where one or both of the parties lacked conubium could not be an ins civile marriage, but was, neverthe­less, recognized as a valid marriage, governed by the ins gentium. Since both were valid, did it matter whether a marriage was an ins civile or ins gentium union? Most certainly—the legal consequences were quite different. In a ins gentium marriage, the children took the status of their mother rather than their father; they were suiiuris since potestas did not arise in such a marriage; there was no agnatic relation­ship; and the wife could not be under the legal control (mantis) of her husband (see 5.2.3.1, cf. Inst.Gai.1.78.).

What if the parties were mistaken as to their status? If one of the parties, mistak­enly thought to be a citizen, lacked citizenship (and therefore conubium), the mar­riage could not be regarded as a Roman civil law marriage. However, legislation in the early Empire allowed erroris causae probatio (see 4.4.3.1) in such cases, whereby the non-citizen was entitled to citizenship on proof that a child had been born to the marriage. This converted the marriage from an ius gentium union to an ius civile one. The same consequence resulted if a citizen, mistakenly thought not to be a citizen, married a spouse without conubium·. proof that a child had been born to the marriage converted the union into a civil law marriage, see Frier and McGinn, Casebook, 31-3.

5.2.2.4 Prohibited marriages

The parties to a civil law marriage had to be free of the many bars and prohibitions imposed by the law in its attempt to prevent undesirable marriages.

As a general rule, a marriage in breach of these rules was void, and was often pun­ishable, e.g. through the imposition of infamia. For a good survey of the texts, see Evans Grubbs, Women, ch. 3.

(a) Status and occupation Under the Twelve Tables a marriage between a patri­cian and plebeian was banned, a rule repealed within a few years of its enactment. Marriage between the freeborn and freed persons was originally not possible, although this ban fell into disuse during the Republic. Augustus allowed such mar­riages except for members of the senatorial order. The marriage of a patron and his freedwoman was not prohibited (in later law) but such unions could incur social disapproval (see McGinn, Prostitution, ch. 3). Consequently, the relationship between them would often be one of concubinage rather than marriage. A mar­riage between a patroness and her freedman was thought to be very bad form. Such marriages were banned in the late classical period except in the following case:

Ulpian, Sabinus, book 34: If a patroness is so degraded that marriage at least with her own freedman is honorable, it should not be prohibited by the judge who is investigating the matter. (D.23.2.13.)

In an attempt to tighten military discipline, Augustus banned soldiers from marry­ing while on military service. And the marriage of a soldier was considered ended if he entered the army as a married man, any children becoming illegitimate. This resulted in recruits divorcing their wives, since the children of divorced parents remained legitimate. These rules proved very unpopular and contributed to a pro­nounced increase in concubinage during the early Empire. The ban was lifted in AD 197 by Septimius Severus in an effort to retain the support of the army.

A person could not legally be a party to two marriages at the same time—the Romans regarded marriage as a monogamous union. Bigamy incurred infamia, and occasionally could have extreme consequences, as Messalina discovered when she attempted to marry her lover while still married to Claudius. A widow could not marry within ten months (extended in the Empire to one year) of the death of her husband. A guardian could not marry his ward, nor could the guardian's son. A conflict of interests could well arise in such a case, if the rule were otherwise, as also where a high-ranking officer in provincial government attempted to marry a woman from the province—the marriage was void.

(b) Religious factors Vestal virgins obviou sly could not marry—indeed, if they lost their virginity, they were subject to the death penalty. However, once they retired from office, they became eligible for marriage. Rome's conversion to Christianity led to a number of restrictions on religious grounds—those who became monks or entered the priesthood could not marry; and marriage between Christians and Jews was prohibited.

(c) Moral grounds Augustus prohibited the senatorial order from marrying not only freed persons, as we have seen, but also a variety of people engaged in activi­ties considered to be morally dubious, such as prostitutes, procurers, actors, and actresses:

Modestinus, Formation of Marriage, sole book: As far as marriages are concerned, it is always necessary to consider not just what is lawful but also what is decent. [1] If the daughter, grand­daughter, or great-granddaughter of a senator marries a freedman, or someone who was an actor, or whose father or mother were actors, the marriage will be void. (D.23.2.42pr.-1.)

Justinian eventually abolished this rule, having been granted dispensation by his predecessor to marry that grande dame of the stage, Theodora. Augustus banned an adulterous wife from marrying her lover after divorce. Justinian limited the ban to the lifetime of the innocent party.

(d) Consanguinity Marriages were prohibited between persons related by blood. The basic rule was that lineal ascendants and descendants could not marry, e.g. parent and child; grandparent and grandchild. As regards collaterals, the rule emerged by the late Republic that collaterals could not marry if at least one of them was only one degree removed from the common ancestor. Thus, brother and sister could not marry; nor could uncle and niece, nor aunt and nephew. Cousins could marry, even first cousins (both are two degrees removed from the common ances­tor). Claudius altered the rules to enable him to marry his niece, Agrippina: he allowed a man to marry his brother's daughter, but not that of his sister. Claudius's decree was repealed in AD 342:

If anyone should be so abominable as to believe that the daughter of his brother or sister should be made his wife or should fly to her embrace not as her paternal or maternal uncle, he will be liable to a penalty of capital punishment. (C.Th.3.12.1.)

It is not clear how strictly the rules on consanguinity were applied throughout the Empire. It seems that regional variations did occur, with confusing consequences after the general grant of citizenship in AD 212—old practices may have continued in ignorance of the general rules. In AD 295, a decree condemned those who 'rushed into illicit unions in the promiscuous manner of farm animals or wild beasts, driven by execrable lust, with no regard for decency or righteousness' (quoted in Gardner, Women in Roman Law, 36).

(e) Affinity One could not marry an ascendant or descendant of one's former spouse, e.g. a man could not marry his former mother-in-law. Nor could one marry the former spouse of an ascendant or descendant, e.g. a former stepmother or step­daughter. Marriage to a brother/sister-in-law was possible until prohibited after Rome's conversion to Christianity.

(f) Adoption When persons were adopted, they became for all legal purposes members of the family of the adoptive parents. Hence, there could be no mar­riage between an adopted child and the adoptive parent, even if the child was later emancipated. As regards persons related collaterally through adoption, Gaius states:

Gaius, Provincial Edict, book 11: When the relationship of brother and sister arises because of adoption it is an impediment to marriage while the adoption lasts. So I will be able to marry a girl whom my father adopted and then emancipated. Similarly, if she is kept in power and I am emancipated, we can be married. (D.23.2.17pr.)

Compare to Frier and McGinn, Casebook, 34-8.

5.2.3 Types of marriage

Roman civil law marriage was of two types: mantis marriage, in which the wife was in 'the hands' of her husband, i.e. in his legal control; or 'free' marriage, where she was not subject to that control.

5.23.1 Manus marriage

(a) Form Manus marriage was very common in early law but gradually dimin­ished in importance and frequency during the Republic (see Looper-Friedman, S. E., 'The Decline of Manus-Marriage in Rome', (1987) 55 TR, 281-96, on the reasons for the decline). See Frier and McGinn, Casebook, 89-95 as well as Lewis, 'Slavery, Family, and Status', 166-7 and now Cantarella, 'Women and Patriarchy in Roman Law', in OHRLS, 419-31.

A manus marriage could be created in the following ways (see Inst.Gai.1.110.):

(i) Coemptio (‘purchase') A form of bride purchase, effected through a formal conveyance per aes et libram in the presence of witnesses. The husband 'bought' the bride in a manner similar to that used for the purchase of certain types of property, the bride 'selling' herself with the permission of her paterfamilias or if she was sui iuris with the permission of her guardian.

(ii) Conf arreatio (‘sharing ofbread') A religious ceremony before witnesses and the chief pontiff, requiring, inter alia, the eating of special bread by the couple. It was reserved for marriages of patricians and members of the college of priests. Its significance lay in the fact that it provided a means of entry to the highest echelons of the priesthood: only children of a marriage by confarreatio were eligible for such elevated office.

(iii)Usus If a man and woman cohabited for a year with affectio maritalis, i.e. regarding themselves as man and wife, a manus marriage was created even though the partieshad not undergone any form of wedding ceremony. The required marital intention would normally be presumed from the fact of cohabitation. However, manus would not arise (according to the Twelve Tables) if the woman absented herself for three nights in each year—probably three consecutive nights. There was a marriage in these circumstances, but it was a free marriage, not with manus.

(b) Legal consequences In manus marriage, the husband (or his paterfamilias if alive) exercised a power and control over the wife comparable to the potestas of a father over his child. Indeed, the wife was regarded as her husband's daughter in certain respects, e.g. under the intestacy law (see 8.3.1.1). She could not own prop­erty except for very personal belongings. Property owned by her before marriage automatically passed to the husband or paterfamilias, subject to the rules on dowry (see 5.2.5.1), as did anything that she acquired during marriage. She could not make binding contracts without her husband's permission, and generally was not able to participate in litigation and the processes of law. Moreover, entry into manus mar­riage severed her former agnatic ties: she exchanged one family for another and left the power of her paterfamilias. When the paterfamilias of her husband died, the wife came under the potestas of her husband (if he became paterfamilias) and remained under his manus by virtue of their marriage. A wife could be released from manus by a similar procedure to that used for the emancipation of a daughter (Inst.Gai.1.137.). There were, however, important differences between the position of the wife in a mantis marriage and a child in potestas. The husband did not have the power of life and death over his wife; nor could he sell her into slavery, or surrender her noxally, or give her in adoption. She attained the position of wife in manu in the household.

5.2.3.2 Free marriage

This type of marriage, the norm from the late Republic onwards, was 'free' mainly in the sense that the wife was legally independent—the husband did not hold legal power over her, see Frier and McGinn, Casebook, 96-103.

(a) Form A free marriage was created by the cohabitation of the parties, provided that they regarded themselves as man and wife. As soon as such cohabitation began, i.e. with the necessary intent, the marriage came into existence. For Ulpian, the intention of the parties seems to have been more important than the fact of cohabitation:

Ulpian, Sabinus, book 35:... for it is consent, not sleeping together, which makes a marriage. (D.35.1.15.)

This is somewhat misleading if interpreted to mean that everything depended on the intention of the parties. Most probably, it was cohabitation, which was the cru­cial element, and from which the necessary marital intention could be presumed. Besides, the parties would often have undergone the ceremonies customary in Roman weddings, which served as evidence of the required intent. The betrothal or the giving of a dowry—both customary, if not strictly necessary, would provide fur­ther evidence. It could be, however, that there was no betrothal, no dowry, no wed­ding ceremony, but simply cohabitation with the necessary intent. Such a union was a perfectly valid marriage. The ease with which free marriage could be created may suggest a rather casual approach by the Romans to marriage as a legal institu­tion, but this is belied, as we have seen, by their great concern to prevent undesir­able marriages.

(b) Legal consequences The wife did not exchange families—she remained an agnate of her original family. If she had a paterfamilias, she remained in his power. If she was sui iuris, she continued under the guardianship of her tutor (see generally 5.4.2) but otherwise was legally independent: her property remained her own and she could keep whatever she acquired during the marriage. The wife could make contracts and could generally participate in the legal process. Indeed, one sena­tor's wife, Gaia Afrania, earned some notoriety as a vexatious litigant. She prob­ably was the 'shameless woman' who, according to Ulpian (D.31.1.5.), annoyed a magistrate, thus provoking a praetorian edict prohibiting women from acting as representatives in litigation, see Gardner, Women in Roman Law, 262-3; as well as Evans Grubbs, Women, 60-70.

Since the husband in a free marriage did not acquire his wife’s property, he might be tempted to persuade her to make gifts to him. The possibility that wives could be tricked out of their property resulted in a ban on inter vivos gifts between husbands and wives:

Ulpian, Sabinus, book 32: As a matter of custom, we hold that gifts between husband and wife are not valid. This rule is upheld to prevent people from impoverishing themselves through mutual affection by means of gifts which are not reasonable, but beyond their means. (D.24.1.1.)

It is not clear when the ban on gifts was introduced—Ulpian's reference to the ban as 'a matter of custom' is not very illuminating—but most probably it was imposed in the late Republic, see Cherry, D., 'Gifts Between Husband and Wife: The Social Origins of Roman Law', in Speculum Juris, 34-45. By then free marriage was preva­lent, and the problem of the tricked wife had surfaced. There were some exceptions to the general rule, e.g. inexpensive anniversary presents were allowed, as were gifts that did not enrich the donee:

Ulpian, Sabinus, book 31: If a husband gives his wife money to buy perfumes and she pays the money to his creditor and then buys perfumes with her own money—she will not be held to have been enriched by this. (D.24.1.7.1.)

Important financial protection for wives was provided in the early Empire when women were prohibited from standing as guarantors for their husbands or taking over their debts. On the property of the spouses, see the texts collected in Frier and McGinn, Casebook, 122-39; as well as Evans Grubbs, Women, 98-101.

5.2.4 Divorce

Divorce was normally affected by the act of one or both of the parties to the mar­riage. It was, like marriage, concluded through a private process and it was not judicially controlled. A marriage could also end in divorce at the insistence of the paterfamilias of the husband or wife, even where the marriage was a happy one (but the paterfamilias ran the risk of infamia if he behaved unreasonably). In the Empire the power was restricted, but never abolished—it could be exercised if grave cause was shown.

Divorce by the parties differed according to whether the marriage was free or with mantis. For a good survey of the texts, see Evans Grubbs, Women, ch. 4.

5.2.4.1 Divorce in manus marriage

In early Rome, when manus marriage was prevalent, divorce appears to have been a rare phenomenon. It seems that if a husband wished to divorce his wife, he would normally consult their respective families, who might try to reconcile the parties or at least to supervise the divorce arrangements. The first attested divorce—c. 306 BC—concerned a senator said to be expelled from the Senate for divorcing his wife without proper consultation. See Treggiari, S., Roman Marriage (1991), 442. If the husband divorced his wife without good cause, he ran the risk of incurring infamia. Evidence of what constituted good grounds is scanty, but they apparently included poisoning of children, theft of keys, and the imbibing of wine—conduct thought to be unbecoming in the responsible wife of the early Republic. It is not clear whether a wife could divorce her husband in a manus marriage. In early law she probably could not initiate a divorce; but in the early Empire, it seems that she could end the marriage unilaterally by repudlum (see 5.2.4.2).

The procedure required to terminate a manus marriage was largely a variation of the ceremonies required to create the marriage, since the wife had to be transferred back to the authority of her (remaining) family. For example, a ceremonial resale of the wife terminated marriage by coemptio (and probably by tisus, too). By the end of the Republic, the forms of divorce in manus marriage had become virtually obsolete, so that such marriages came to be terminated in the same way as free mar­riages. See Frier and McGinn, Casebook, 94-5.

5.2.4.2 Divorce in free marriage

No grounds were necessary to end a free marriage, but a causeless divorce could incur the wrath of the censor, as in the case of a manus marriage. As regards form, terminating a free marriage was basically as informal as the manner in which it was created:

Paul, Edict, book 35: A true divorce does not take place unless an intention to remain apart permanently is present. So things said or done in anger are not effective until the parties show by their persistence that they are an indication of their considered opinion. So where repudiation takes place in anger and the wife returns shortly afterward, she is not held to have divorced her husband. (D.24.2.3.)

In essence, all that was required was that at least one of the parties should have ceased to have the intention to be married (see Frier and McGinn, Casebook, 156-69). The parties could divorce by mutual consent (divortium) or unilaterally (repudium). The former method required that the parties should intentionally cease living together, whereupon it could be presumed that they had both lost the neces­sary marital intent. Repudium occurred where only one spouse wished to end the marriage—it was normal to notify the other party by letter or messenger:

A woman in the manus of her husband can, if she has sent him a notice of repudiation, compel him to release her [by emoncipatio] just as if she had never been his wife. (Inst.Gai l.137a.)

Augustus required the presence of seven witnesses for the sending of a notifica­tion of repudium (see D.24.2.9.), but it is not clear whether this rule applied gener­ally or only where a husband repudiated his wife on the grounds of her adultery. Although the sending of the notification terminated a marriage, the consequences could be unexpected and unwelcome, especially for the third party. Consider what happened to this playboy:

Papinian, Adultery, book 2: The deified Hadrian relegated a man for three years where he had led someone else's wife to his house while she was on a journey, and she sent a notice of repu­diation to her husband from there. (D.24.2.8.)

Relegation involved banishment and possible loss of citizenship and civil rights—a fate best avoided.

In AD 48 Messalina, Claudius' third wife, went through a ceremony of mar­riage with her lover, Silius. She was considered to have committed bigamy and was executed. Mere cessation of affectio maritalis was insufficient: in practice, there had to be some visible evidence of it, such as sending a note. But then, was not going through a very public wedding with Silius sufficient evidence? It seems not: the divorce logically must precede the subsequent marriage—thus, it is difficult to maintain that the same act, the wedding, can simultaneously create a new marriage and extinguish the previous union, see Treggiari, S., Roman Marriage (1991), 457-8.

5.2.4.3 The lex Julia de adulteriis

(D.48.5., C.9.9.)

The alleged decline in moral standards (at least among the Roman upper classes), coupled with the prevalence of free marriage and the ease with which divorce could be obtained, apparently caused a sharp increase in the divorce rate in the late Republic. Augustus tackled the problem in typically direct fashion. His lex Julia de adulteriis c. 18 BC was a remarkable piece of 'social engineering', designed to improve moral standards by making adultery by a wife a criminal offence— triable by special adultery courts—with serious consequences. The wife was banned from remarrying her lover after divorce; she suffered a diminution of her rights in property—half her dowry entitlement and confiscation of a third of her sepa­rate property—and she could lose her citizenship and be relegated to an island. A husband who knew of his wife's adultery was placed under a duty to prosecute his wife. A conviction led to the termination of the marriage. If he failed to act within the required time, he was guilty of the crime of lenocinium, i.e. acting like a pimp, which was punished in a similar way to adultery. Any act by the husband, which could be regarded as encouraging or condoning his wife's adultery, amounted to lenocinium·.

Ulpian, Adulteries, book 4: The statute has punished the lenocinium of a husband who after catching his wife in adultery has kept her and let the adulterer go; for he ought to have avenged himself on the man and also vented his rage on his wife, who has violated their mar­riage. (D.48.5.30pr.)

The wife had to be prosecuted by the husband within sixty days of the event. Her paterfamilias could prosecute her within that period, but the husband had prec­edence. If neither took action, anyone had the right to prosecute within the next four months, i.e. up to six months from the event.

What of the lover? He was subject to prosecution and penal consequences— forfeiture of half of his property and possible relegation; or sentencing to hard labour, (e.g. in the mines) if he was of lowly status. He could be killed in certain circumstances (see 5.1.2.1) by the paterfamilias of the woman. The husband, too, had a right to kill the lover if he caught him committing adultery with the wife in the husband's home, provided that the lover was of inferior status, e.g. an actor or dancer. Or the husband could detain the lover (of whatever status) for up to twenty hours 'for the purpose of testifying to the matter' (D.48.5.26pr.). The wife's paterfa­milias probably had the same right. But the husband had no right to kill an adul­terous wife (although in practice husbands who did so were often not convicted). Persons outside the triangle of husband, wife, and lover could be punished under the lex Julia de adulteriis. It would be lenocinium for anyone to aid the commission of adultery—e.g. by acting as a go-between or allowing premises to be used by the lovers.

The enormous scope of Augustus's law can be further appreciated from the fact that it applied to adultery in all marriages, not just Roman civil law marriages. As can be imagined, the lex Julia de adulteriis had a sensational impact on Roman soci­ety. The satirists had a field day. The element of compulsion, whereby a husband was forced to divorce his wife, was ridiculed as not being conducive to the stability of family life that Augustus was apparently seeking. The Emperor was castigated for having apparently conflicting aims. See further Norr, D., 'The Matrimonial Legislation of Augustus: An Early Instance of Social Engineering' (1981) 16 IJ, 350-64. Nevertheless, Augustus's legislation remained the cornerstone of Rome's matrimonial law for centuries—if somewhat erratically enforced—although 'there is little evidence that Augustus changed opinions or standards of behaviour in his own time’ (Treggiari, S., Roman Marriage (1991), 292). On this law, see extensively McGinn, Prostitution, chs 5 and 6.; see also Evans Grubbs, Women, 83-7.

5.2.4 4 Later developments

The Christian Emperors frowned on divorce but did not abolish it. The tendency was to penalize causeless divorce harshly through bans on remarriage, loss of prop­erty rights, and deportation. For example, Constantine punished repudium by a husband unless his wife was proved to have committed undesirable acts such as adultery or homicide (see C.Th.3.16.1.). Moreover, it seems that Constantine made adultery by a wife a capital offence. Justinian abolished divorce by mutual con­sent, apart from exceptional cases, but the ban was lifted after his death. And some attempt was made in the late Empire to make divorce procedure less informal: in repudium it became compulsory to send a formal divorce petition as notification of the ending of the marriage. For a survey of cases and examples, see Evans Grubbs, Women, ch. 4.

It is important to appreciate that throughout the whole history of Roman divorce law it was never essential to have a ground for divorce. Causeless divorces were valid but attracted penal consequences. For example, under Justinian the penalties were: for a wife—loss of dowry and compulsory entry into a nunnery; for a husband— loss of a third of his estate and compulsory entry into a monastery. Justifiable grounds for divorce under Justinian included adultery by the wife, cohabitation by the husband with a concubine, entering the religious life, and plotting against the Emperor. On this topic, see extensively Evans Grubbs, J., Law and Family in Late Antiquity—The Emperor Constantine's Marriage Legislation (1995).

5.2.5 Dowry (dos)

(D.23.3-5, D.24.3, D.25.1.)

5.2.5.1 The expectation of a dowry

Dowry was a crucial element in property relations between spouses throughout Roman legal history. The dowry consisted of property (or some other contribution) given to the husband by the wife or her paterfamilias or others (like the clients of her paterfamilias) on her behalf (see Sailer, R. R, 'Roman Dowry and the Devolution of Property in the Principate' (1984) 34 CQ, 195-205 for an interesting discussion of the use of the dowry to ensure the continuation of senatorial family lines dur­ing the Principate). It was normally given on marriage but could be given before or after it. Although a husband was not strictly under a legal duty to maintain his wife, or vice versa, in practice he would feel a moral and social obligation to do so. The dowry would help him—it was regarded as the wife's contribution to the running of the matrimonial home. The giving of dowry by a wife constituted the major excep­tion to the rule invalidating gifts between spouses. See Frier and McGinn, Casebook, 72-86. There is a natural affinity between any discussion of the dowry and discus­sions of broader elements of the law of property. These should, therefore, be taken into account when studying this area of law. Since the dowry was intimately con­nected to marriage and divorce in Roman law, we prefer to treat it here.

On a manus marriage, the husband acquired the wife's property if she was sui iuris (see 5.2.3.1). Originally, it seems that such property was not regarded as dowry, but by the late Republic it was. If the wife was not sui iuris, then her paterfamilias would be expected to provide a dowry. The position was similar in a free marriage: either the wife or her paterfamilias would be expected to provide dowry. The hus­band did not have a right to a dowry unless it had been specifically promised; but he certainly had an expectation of a dowry since it became an invariable practice to give one, at least among the Roman propertied classes. The search for wives with handsome dowries by ambitious young men was a familiar scenario for Roman satirists. It seems that a decree of the late classical period imposed a legal duty on the bride's paterfamilias to provide a dowry: he could be forced by a magistrate to do so (D.23.2.19.).

5.2.5 2 Rights in the dowry

Who owned the dowry? The general position in early law was that the husband became the absolute owner of the dowry. He could do what he liked with it, and did not have to return it, or any part of it, on the termination of the marriage. So it became the practice for the donor to insist on a formal promise by the husband to return the dowry when the marriage ended, or to dispose of it in the agreed manner (see Gardner, J. E, 'The Recovery of Dowry in Roman Law' (1985) 35 CQ, 449-53 whose article builds on the article by Sailer (cited earlier)). See Frier and McGinn, Casebook, 140-53 as well as Evans Grubbs, Women, 91-101.

Such a dowry was known as dos receptitia ('a returnable dowry'), the husband’s promise enforceable by the actio ex stipulatu, one of the standard contractual rem­edies. From this practice emerged the concept that the husband was not entitled to the capital, but was entitled to the fruits (i.e. the profits or income) of the dowry:

Ulpian, Sabinas, book 31: Equity demands that the profits on a dowry shall belong to the husband; since he bears the burdens of marriage, it is only fair that he receives the profits. (D.2.3.3.7pr.)

An alternative practice was for the dowry to be valued (dos aestimata), the husband promising to pay the agreed valuation on the termination of the marriage. The advantage to the husband, compared with dos receptitia, was that he could do as he wished with the property. He did not have to return the capital; his only duty was to pay the valuation. However, he did take the risk of deterioration or damage to the property.

Eventually, the praetors introduced the actio rei uxoriae (‘the action for the wife's property') in an attempt to protect wives, particularly as regards dowries that were not valued or returnable. It is likely that the introduction of this remedy was a response to a notorious divorce in c. 230 BC when Spurius Carvilius Maximus Ruga divorced his wife because she was allegedly sterile. Under the action the divorced wife could recover whatever the judge thought was a fair share of the dowry, subject to allowing the husband to make reasonable deductions. The action was extended to cases where the husband predeceased the wife. The effect of these developments was that the husband's position by the end of the Republic was in many respects little better than a form of temporary stewardship. Indeed, the husband came to be regarded as owing a duty of care in respect of the dowry—consider the following text concerning one of the most famous of all Romans, Gaius Gracchus (see 1.2.3.1):

Javolenus, From the Posthumous Works ofLabeo, book 6: According to Servius, a husband is responsible for fraud and negligence in connection with all the property in the dowry apart from money. This is also the view of Publius Mucius; for he stated it in the case of Licinnia, the wife of Gracchus, whose dotal property had perished during the insurrection in which Gracchus was killed, saying that the property should be restored to Licinnia since Gracchus was to blame for the insurrection. (D.24.3.66pr.)

What degree of negligence was required to make the husband liable for losses? Paul states:

Paul, Sabinus, book 7: In matters relating to the dowry, the husband... must exercise the same diligence as he shows in his own affairs. (D.23.3.17pr.)

This was a subjective standard of care—how had that husband behaved in the past? The 'diligence' expected of a husband may have been different in the early classical period, when the duty of care that was required was probably that of the bonus pater­familias (the prudent head of the household)—normally this was a higher standard of care (but not necessarily), and an objective one.

5.2.5.3 Reforms of Augustus

Augustus introduced important reforms, the general trend of which was to improve the position of the wife. The husband was banned from selling dotal immovables (dowry consisting of land) without the consent of the wife if they comprised Italic land, i.e. land south of the river Po (in the Lombard Plain). In addition, mortgages created on such land (even with the wife's consent) were invalidated. Further, Augustus established detailed rules (some originally introduced previously) con­cerning the return of the dowry on termination of marriage (see Frier and McGinn, Casebook, 170-87). The rules applied particularly to dowry other than dos receptitia or aestimata (which continued to be governed by the original agreement between the parties) and depended on whether the marriage ended through divorce or death:

(a) Divorce The wife (or her paterfamilias, if alive) could recover dowry by the actio rei uxoriae as before, but Augustus formalized the position regarding deduc­tions and the recovery of expenses. The husband generally had to care for the dowry at his own expense; but he could claim for 'necessary' expenses, i.e. those that arose out of necessity:

Ulpian, Sabinas, book 36:... if a husband rebuilds a house which was falling into ruin but was useful to his wife, or if he replants an olive orchard where trees have blown down... [Paul, Sabinas, book 7 [2]j or spends money on curing slaves, [Ulpian, Sabinas, book 36 [3JJ: or plants vines or looks after trees or nurseries for the good of the land, he will be held to have incurred necessary expenses. (D.25.1.1.3.; D.25.1.2.; D.25.1.3.)

The husband could deduct the value of gifts made by him to the wife de facto (such gifts being legally invalid under the general rule prohibiting gifts between spouses). And the husband could deduct the value of property wrongfully appro­priated by the wife; alternatively, he had the actio rerum amotarum ('the action for things removed') for the recovery of such property (D.25.2.). Further, the husband could retain a sixth of the dowry if the wife was guilty of serious misconduct, or an eighth for lesser behaviour. If the wife (or her paterfamilias) was responsible for the ending of the marriage in divorce, the husband could retain, in addition to the other deductions, one-sixth of the dowry per child to a maximum of half the dowry. The person 'responsible' for ending a marriage was the one who ended it without good reason or caused its breakdown, but not necessarily the one who initiated the divorce:

Papinian, Adulterers, sole book·. I have married a woman charged with adultery; I have divorced her as soon as she was convicted. Am I regarded as having provided the grounds for the divorce? The reply was: Since, in accordance with the lex Julia, you are forbidden to keep a wife of this kind, it is clear that you are not to be regarded as having provided the grounds for the divorce. For this reason then, the question of right shall be treated as though the divorce had been brought about because of the fault of the woman. (D.48.5.12.13.)

How soon after the divorce did the husband have to return the dowry? The basic rule was that dotal property or its value had to be returned immediately except for fungibles, i.e. property consumed through use (such as money or corn), which could be returned in no more than three annual instalments. But if the husband was to blame for the divorce, the whole dowry had to be returned immediately, and it seems that he could claim no deduction.

(b) Death When the wife predeceased the husband, the basic rule was that he retained the dowry but had to return any part that had been donated by the wife's paterfamilias or other donor (if alive). The duty to restore was subject to the hus­band's rights to deduct for expenses, gifts, and misappropriated property, as on divorce. And the husband could retain one-fifth of the returnable dowry in respect of every child of the family. If the husband predeceased the wife, she (or her pater­familias) could recover the dowry from the husband’s heirs, who could make the usual deductions for expenses, gifts, and misappropriated property.

Following Augustus' reforms, the position of the wife as regards dowry was rea­sonably secure. The husband's rights were similar to those of a usufructuary (see generally 6.3.4.1), i.e. he was entitled to use the property and take its fruits, but with a duty to restore the capital intact. And if he was insolvent, the wife was given priority over all unsecured creditors.

52.5.4 Reforms of Justinian

Further improvements to the wife's position were enacted later, especially by Justinian. He prohibited the alienation of dotal immovables even with the wife's consent. Moreover, the wife was given the right to bring proceedings during the marriage: she no longer had to wait for its termination. For example, she was allowed to recover the dowry if her interests in it were threatened. The actio rei uxo- riae was abolished—the standard remedy for recovery of the dowry became the actio ex stipuiatu, which could now be brought even though the parties had not made an agreement about the disposal of the dowry. Justinian held that there was an implied agreement in such a case that disposal would be determined by the standard provi­sions of the law. The husband's rights to automatic deductions were abolished, but he was allowed to bring a claim for expenses. If the wife predeceased the husband, the dowry went to her heirs, not to the husband. Dowry had to be returned within one year unless it consisted of land, in which case return had to be immediate. Further, the wife was allowed a tacit hypothec (implied mortgage) over the hus­band's property to secure her interests, thus giving her priority over all creditors of the husband, whether secured or unsecured.

The improving position of the wife as regards dowry, from the late Republic onwards, helped to secure some measure of financial protection for the wife, espe­cially on divorce. This was a major development in the gradual improvement of the legal position of women in Roman law.

5.2.5.5 Donatio propter nuptias ('Gift on account of marriage')

(D.24.1.)

This was the converse of dowry—a gift by the bridegroom to his wife before the marriage. The property belonged to the wife but would normally be managed by the husband during marriage. The usual purpose of such gifts was to ensure provi­sion for the wife in the event of the husband predeceasing her. In the East, there was considerable legislation in the late Empire that elevated donatio from a customary practice into an important legal institution subject to many of the rules applicable to dowry. A transfer of property was originally necessary, but later a promise of a donatio became actionable. The gift had to be made before the commencement of the marriage so as to evade the rule banning gifts between spouses. But Justinian enacted that a gift made during the marriage could be a valid donatio. Moreover, he specified that the value of the donatio had to be equal to the dowry; and he applied rules similar to those of dowry to the matter of the alienability of the property and its ultimate disposition.

5.2.6 Relationships outside marriage

Some relationships outside marriage could have legal consequences. The most important cases were stuprum and concubinage. For a good survey of the texts, see Evans Grubbs, Women, ch. 3.

5.2.6.1 Stuprum

A sexual union outside marriage between free persons was termed stuprum. The term comprised relations between unmarried persons, and adulterous unions. In the Republic there were occasional trials (but no regularized procedure) in which stuprum was alleged—usually against adulteresses. Condemnation normally involved exile or fines. However, there were instances where a father put his daugh­ter to death for her stuprum. Her 'misbehaviour' was seen as ruining the chances of a respectable marriage. A child born in stuprum was regarded as illegitimate.

Although Augustus was apparently not averse to stuprum in his own personal life, he legislated against such behaviour and established special courts to deal with alleged cases. The legislation (the Lex Julia on adultery) was aimed at stamping out immorality (as it was perceived) and ensuring that Roman women remained eligi­ble for marriage. This may explain why intercourse with a prostitute was not con­sidered to be stuprum (she was not a suitable prospective wife), whereas intercourse between betrothed couples did constitute the offence.

5.2.6.2 Concubinage

(D.25.7.)

In some ways, concubinage resembled marriage. It was regarded (according to cer­tain modern scholars) as a monogamous union—a man could not have two concu­bines (at the same time) or a wife and a concubine (though see D.25.7.3.1.). It seems that a partner could prosecute an unfaithful concubine, but such proceedings were infrequent. Concubinage usually involved a union between a man and woman infe­rior to him in status, e.g. a patron and his freedwoman. Emperors were occasionally known to take concubines, e.g. Vespasian and Marcus Aurelius. The children of con­cubines were regarded as illegitimate until the Christian Empire, when legitimation by subsequent marriage was introduced (see 5.1.2.6). Gifts between a concubine and her partner were valid since the ban on gifts applied only to married couples.

Concubinage strictly amounted to stuprum, and thus could be prosecuted. This could occur especially where the partners were of equal legal status—the view was taken (at least by Augustus) that they should be married and not living in concubi­nage. To avoid prosecution for stuprum it was usually desirable that the concubine should be of lowly status:

Marcian, Institutes, book 12: Another person's freedwoman can be kept as a concubine as well as a freeborn woman, especially where she is of low birth or has been a prostitute. But if a man would rather have a freeborn woman with respectable background as his concubine, he will not be allowed to do this unless he clearly states the position in front of witnesses. (D.25.7.3pr.)

The spread of concubinage, particularly following Augustus's ban on marriage by soldiers, helped to make the institution more acceptable. However, the Christian Emperors made various attempts to discourage concubinage but failed to eradicate it.

5.3