Marriage and Family

1

26

Marriage and Family

Susan Treggiari

1. Introduction. This chapter attempts to explain what Romans took for granted in thinking of and writing about family life, and what was especially praised and valued. To look at values in any society, we examine ideas and practice. In looking at Rome, we switch between theory (attested especially in philosophical writing, derived from Greece, but also in literary sources, e.g. Vergil’s Aeneid or Horace’s Odes); prescription/prohibition (e.g. customary and statute law); convention and fashion (reflected in visual arts and ‘the epigraphic habit’ as well as in literature); representation (what people say about what they or others do and think, including tendentious/moralising sources, such as Cicero in his court-room speeches, Valerius Maximus or Juvenal); practice (what people do, which might be documented by historiography, private letters, jurists’ description of cases or problems, tomb-inscriptions). We would like to see how all these change over time.

But development is often difficult to document because of the disparity of sources for different periods. Is it true that idealization of family life increased when the Principate limited aristocratic power (Veyne [1978] demolished by Saller and Shaw [1984]: 134-6, Dixon [1991])? Or is it just that it shows up more in our sources? Did the lower classes have different values? Can we get at them through what they put on their tombstones or their appeals to the emperor ? How can we distinguish the Roman part of a man’s thinking when he might be a Greek aristocrat like Plutarch or a Parthian freedman (CIL 11.137= ILS 1980)? The discussion which follows focuses on Roman citizens, the period from the late Republic to the early third century AD, husband/wife and parent/child relationships.

Where I have not given a reference to bibliography, the reader may find it in OCD. Some of the material is further explored in Treggiari (1991).

1. Prefatory. Fewer people marry and rear children than in the old days; more married people divorce. You can read it now or in writings of the Roman upper classes:

If we could manage without wives, ...we would all do without the annoyance they cause, but since nature has taught us we cannot live comfortably with them, nor live at all without them, we must take thought for our eternal welfare rather than our temporary pleasure. (Q. Caecilius Metellus Macedonicus, censor 131 BC, or Q. Caecilius Metellus Numidicus, censor 102-1 BC, quoted by Gell. NA 1.6.)

Augustus read the whole of Metellus' speech to the Senate when he in his turn was trying to pressure people into marriage.

Or take this, Cicero’s remark about his brother to the ex-wife’s brother after a divorce which ended about a quarter-century of marriage:

He's far from having any idea of marrying again. He says he finds it much pleasanter to have his bed to himself.

(Cic. Att. 14.13.5, 26 April 44 BC)

Or a husband commemorating his wife and pleased with their record of a marriage which lasted over forty years:

Such long unions, ended by death, not interrupted by divorce, are uncommon. (Laudatio ‘Turiae’ 1.27)

In dealing with such sources, in various genres, we should ask: Are such perceptions accurate? What are people's motives in saying this? How did they expect their hearers to react? Cicero, who had just divorced twice in two years, may have been sympathetic to his brother's disillusionment. But Atticus, who married late, seems to have been happy with his wife and daughter and had loyally supported his own sister, may have been less inclined to smile. (We cannot tell if the Cicerones would have remained unmarried from choice, for both died in December 43.) Similarly, let us ask what statistics the happy husband had to back his generalisation. He will have known divorced people, but surely his statement is impressionistic. The censor, and, later, Augustus, will have had census evidence for thinking Roman men did not marry (early enough and often enough).

3. Theory and Prescription. In his treatise on the Republic, which was circulated to acclaim during his absence as a provincial governor (51-50 BC), Cicero shows a predictable liking for the pairing of the individual and the community to which he belongs and of the family and the city. While Ennius (ap. Cic. Rep. 5.1) makes Rome depend on traditions and viri (males), Cicero, though he pays deference to the series of Roman heroes, has an eye to the rôle of women such as Lucretia (Rep. 2.46, cf. Leg. 2.10). The comparison of family and state is a continuing theme (e.g. Rep. 1.8, 61, 1 fragment 2; 3.45, 4.7, 5.4). Cicero presents the standard Greek view that the family developed before the wider community (Rep. 1.38, cf. Arist. Eth. Nic. 1162a16-22). This view is forcefully restated at the very end of his life in the book on duty addressed to his son:

Because the urge to reproduce is an instinct common to all animals, society originally consists of the pair, next of the pair with their children, then one house and all things in common. This is the beginning of the city and the seed-bed of the state. (Cic. Off. 1.54)

The doctrine that marriage is based on natural law was juristic orthodoxy (e.g. Dig. 1.1.1.3, Ulp. i inst.). Secondly, Cicero believes that the health of the state depends on the family and that the education of children and self-control of grown men, are vital (e.g. Legg. 1.57, 3.29-30). Senators set an example (Legg. 3.10, 28). Good states depend on the family:

For the sake of life and the practice of living, a prescription has been made for recognised marriages, legitimate children and the sacred homes of the household gods and family Lares, so that everyone should enjoy common and individual blessings. For living well is impossible without a good community and there is nothing happier than a well set up polity. (Cic. Rep. 5.7).

The family, religion and the wider community are inextricable. The state, however large, for instance all Roman citizens, is constructed on couples who obey their animal instinct to mate and reproduce. Justice and law emerge from Nature. Justice is to give each individual what belongs to him and this principle is as true for peoples as for individuals. In a good community, all its members have a share in the common good: a commonwealth or res publica, is one in which everything belongs to the people, res populi. A Roman theorist would not make a state and its citizens into opposites: the state is the Roman People. Anyone reviewing the rise and fall of cities inevitably sought the cause in the nature of their organisation and life-style (Polyb.1.1; cf. Sall. Cat. 5.9, 6.3, Livy Praef. 9). Imperial expansion brought worries about a change in Roman morality, which included unease with the idea and reality of the well-dowered wife who might control her husband (e.g. Malcovati, ORF³Cato158; Plaut. Asin. 87). `All men rule their wives; we rule all men, our wives rule us'. (Plut. Apophth. 198D3, is attributed, aptly, to Cato the Censor, who opposed the repeal of the Oppian law and later sponsored a law limiting inheritance-rights which Cicero characterises as full of injustice to women (Rep. 3.17; Dixon [1985]).

4. Roman peculiarities. What, in the ancient view, made the Romans ‘top nation’? Piety, a mixed constitution or features of their family life? Dionysius of Halicarnassus, who chose the third answer, idealised the ancient sacramental marriage which put the wife under her husband's control (Ant. Rom. 2.24-26.1). Cicero considers it right and Roman that no public official should keep women in order, but

There should be a censor to teach husbands to guide their wives (Rep. 4.6).

Philosophers conceived the household as a monarchy. The father might consult his opposite-numbers from other households. He answered to the law, custom and opinion of his community, but, as Cicero's Scipio puts it to Laelius, `Isn't it true that no-one but you rules your entire house?' (Rep. 1.61) This concept had long been enshrined in the practically unique (Gai. Inst. 1.55) paternal power (patria potestas) held by a head of household (paterfamilias), any male who was not himself under the power of a male ascendant. A grandfather might have in his power his sons (filiifamilias), daughters (filiaefamilias), and grandchildren through sons. If the grandfather died, his sons automatically each became paterfamilias. If a son predeceased his father and himself had a son, that grandson would similarly become an independent household-head on his grandfather's death. Women who on marriage passed into the husband's family because he obtained control over them, manus, came under something like a paternal power of the husband or his paterfamilias. Originally, the power of the paterfamilias was absolute. He could put those in his powerto death (very rare: Harris [1986]), and any property they acquired was his, but it was, as it were, held in trust and if the father died without a will then all dependants took equal shares. Since filiifamilias could not act independently, the father’s will was necessary for their marriage or divorce. The household (familia) included slaves. What we would call the family might be the group of liberi or free persons (a word which comes to mean descendants in the male line) under one paterfamilias, plus his wife. Uxor liberique (‘wife and children’) was the usual expression for a man’s nuclear family.

The legal structure dictated emphasis on agnates, kin through males (who shared the gentile name). In the extended family, agnates had a claim on inheritance and male agnates would be guardians to fatherless children. But emotional ties to the female-line asserted themselves (e.g. Saller [1997]). Sisters might maintain strong ties, so the word for cousins who are the children of sisters became the word for cousins in general (consobrini [Gai. Inst. 3.10]). By Augustus’s time, descent in the female line (maternum genus) was valued in sentiment, appraisal of status and inheritance- practices (Gardner [2001]). Members of the upper classes, at least, would try to make a will and to take all close kin into account (Champlin [1991] 103-30). Demography affected the number who might qualify. For instance, Cicero had been close to a patrilineal cousin, L. Cicero, but he died young (68 BC; Cic. Fin. 5.1; Verr. 2.3.170, 4.25, 137, 145; Att. 1.5.1). His own daughter died in 45 BC. The close kin who were left in 43 BC were a son, brother and nephew, but the latter two perished, like Cicero, in the proscriptions of that year. Being divorced from two women, Cicero had no moral obligation to name a wife in his will.

We do not know whether marriage in which the wife did not enter the control of her husband was as old as manus, but we know that it was well established by the fifth century (Gai. Inst. 1.111 citing XII Tables). The two continued to exist as options, but by Cicero's time it was relatively rare for a wife to be in her husband's control, in manu mariti (Treggiari [1991] 20-1). So her legal ties were to her family of birth, with the curious consequence that her legal connection with her own children was played down and only slowly asserted by the emperors. Moral duties and emotional ties cut across this (Dixon [1988]).

Although patria potestas shaped relationships, it is hard to find historic fathers using it to coerce: neither of the Cicerones seems to have appealed to it in difficulties with their sons.

Like moderns, Romans could think of who constituted ‘family’ in ways which changed according to the individual’s circumstances (Dixon 1992: 1-11). Although pietas mandated especial attention tochildren/parents (reflected in epigraphic practice), and husband/wife commemoration is the commonest type (Saller and Shaw), which confirms that the nuclear family was the usual household unit, siblings might be important (Bradley [1991a] 177-204, Bannon [1997]) and cousins, uncles etc had claims on loyalty and affection. Nor must we forget that the nuclear or extended family provided scope for conflict as well as affection (Dixon 1997b).

5. Marriage. A man took a wife in order to produce legitimate children, liberorum quaerendorum causa.Matrimonium, the usual word for marriage, means an institution for making mothers. If the couple were married according to Roman law -- that is, if they were qualified to marry each other and the necessary consents were given -- then the children took the status and family-name of their father. The wife took the social status of her husband.

The general rules on who could marry whom were not unduly restrictive. A Roman could marry any other Roman of the opposite sex who was not a close relation or under age (12 for a girl, perhaps 14 for a boy). Incest taboos ruled out ascendants and descendants, siblings etc. Augustus added further regulations based on social class and also (probably) ruled that soldiers could not marry (Lex Julia de maritandisordinibus and Lex Papia-Poppaea; Phang [2001]).

The Romans expressed the right of A to marry B by conubium. It was fundamental that this capacity existed between Romans and citizens of other Latin cities, but not (unless specifically granted) with foreigners (peregrini) and not between a free person and a slave, or between two slaves. Peregrini formed marriages according to their own customs; Rome did not legislate for them. Valid Roman marriages involved transmission of citizenship to the children, an increasingly valuable privilege vis à vis other citizenships. Slaves might, with owners’ permission or encouragement, form comparatively lasting relationships, contubernia. It is possible to trace such unions at various stages: when both partners were slave and any child born a slave; when the man had been freed but the woman not and any child was born a slave; when the woman had been freed and the man not and any child was born free but illegitimate; when both had been freed and become legally married citizens, so that a child was freeborn and legitimate.

It was normal to assess the success of matches by socio-economic compatibility: birth, wealth, male talent, female beauty. Philosophers added virtue, medical men fertility. Yet, when friends suggested to Cicero a lady whom they thought suitable for an aging divorcee, he rejected her outright: `I've never seen anything uglier' (Cic. Att. 12.11, 46 BC).

Older kin, particularly the paterfamilias,might claim to decide on behalf of young people. But the father’s power was modified by the expectation that he would take no major decisions without consulting kinsmen. By the late Republic, it was apparently normal for him to consult his wife and other women. Also by the late Republic the father could not engage or marry off a child without his or her technical consent. Girls might be allowed a veto on individual candidates. More than that, we can find Cicero's nephew Quintus negotiating a marriage for himself without obtaining his father’s approval (Cic. Att. 15.29.2, 16.2.5).When it came to a woman's second marriage, when she had already participated in social events as a wife, she could claim more independence. Cicero mentions one candidate whom he thinks Tullia could not be persuaded to accept (Att. 5.4.1), and left the decision on her third husband to her and her mother.

Since I was going to be so far away, I instructed them not to consult me but to do what they thought proper. (Fam. 3.12.2)

The man she picked turned out disastrously.

So individual taste could cut across the utilitarian motives for choosing a husband or wife which society approved, and the future husband or wife, particularly if the paterfamilias and other kin were dead, might act for himself or herself. What of love? Critics sometimes say people marry because of passion. This may not be true of the individual instance, but it suggests that the motive was seen as possible. The best confirmation is Augustus' law, Lex Aelia-Sentia, of AD 4 (Gai. Inst. 1.18-19), which laid down that a male slave-owner, under the age of 20, who needed to apply for permission if he wanted to manumit, could free a woman in order to marry her. There was no social advantage for anyone in marrying his own ex-slave, so the motive must be personal attraction. Augustus recognises it as a valid reason. (There was a strong socio-legal prejudice, however, against matches between a free woman and her ex-slave, unless they had previously been fellow-slaves: Evans Grubbs [1993]).

Consent was necessary in classical law for engagement (Dig.23.1.7.1, Paulus xxxv ad edictum) and marriage.

A marriage cannot exist unless everyone consents, that is, those who come together and those in whose power they are. (Dig. 23.2.2, Paulus, xxxv ad edictum)

Engagements could be made simply, by letters or verbal statements (Dig. 23.1.4, 5, 7 pr., 18). In the second century AD a great lawyer held, and others agreed, that if a filiafamilias acted by herself, her father's consent was to be taken for granted unless there was contrary evidence (Dig. 23.1.7.1, Paulus xxxv ad edictum, citing Iulianus).