Earliest Roman Divorces: Divergent Memories or Hidden Agendas?

Gary Martin

[This paper was originally written in Autumn, 2000 for the Latin 457 course at the University of Washington in Seattle, in partial fulfillment for the B.A. in Classics (received in 2001). The paper was modified slightly in December, 2001 for presentation on this web page. The study itself remains incomplete, and text citations need checking and updating.]



Unlike ancient Mesopotamian laws[1] that codify in detail matters of marriage, divorce, the handling of betrothal gifts and dowries at a period that pre-dates the founding of Rome by over a millennium, evidence for early Roman legislation concerning divorce is limited to few and scattered references distanced by centuries from the time the laws were said to be in force. It should therefore not be surprising if we find a divergence of opinion about the precise content and nature of those laws among even the earliest Roman and Greek jurists, comics, and historians who treat the subject. It is also certainly possible that where there is lack of original source material, depictions of the past might be flavored by agendas for the present and future.

Part I of the paper assembles some of the recollections of earliest divorce rights and the first reported cases of divorce in Rome. Part II briefly discusses possible reasons for the divergence among the records, and attempts to account for an unresolved historical question in regard to the long period of time when no divorce is said to have occurred.

Part I: Divergent Views of Divorce Rights Granted under Romulus 

A. The Laws of Romulus as Given by Dionysius of Halicarnassus

For the period of the kings (ca. 753–510 B.C.E.) Dionysius of Halicarnassus in Antiquitates Romanae, 2.25 (written ca. 7 B.C.E.) praises the marriage laws put into effect by Romulus. He first deals with various “flawed” attempts at regulating marriage and sexual intercourse in other (unnamed) societies (2.24), stating that in some societies the model is taken from the beasts (i.e. there are no regulations); in some monogamy is the rule, but there are no formal regulations governing the marriages; in some there are many rules and provisions, especially those designed to “look after the good conduct of women,” which turn out to be impractical, since they are “incapable of bringing the women of unvirtuous nature to the necessity of a modest behavior.” The implication of 2.26.1 is that the “excellent laws which Romulus enacted concerning women” were like the next series of laws discussed by Dionysius concerning children and parents—“more august and of greater dignity and vastly superior to our [Greek] laws.”

It is important to note that Dionysius views disruptions in marriage as arising primarily from the side of women, particularly in their propensity to act “immodestly.” Romulus’ wisdom is highlighted in the fact that he was able to provide in a social structure a reward for women’s modest behavior that surpassed their bent toward immodesty. As a result of this “single law” (reward), Romulus was able to provide for a long period of stability in Roman marriage practice without creating an entire corpus of laws:

But Romulus, without giving cause either to the husband an action against his wife for adultery or for leaving his home without cause, or to the wife an action against her husband on the ground of ill-usage or for leaving her without reason, and without making any laws for the returning or recovery of the dowry, or regulating anything of this nature, by a single law which effectually provides for all these things, as the results themselves have shown, led the women to behave themselves with modesty and great decorum. The law was to this effect, that a woman joined to her husband by a holy marriage should share in all his possessions and sacred rites[2].

Essentially Romulus created a community property state. Dionysius next discusses the etymology of the term by which marriages are called to set the background for the act that brought about a long period of stability in early Roman marriages:

The ancient Romans designated holy and lawful marriages by the term “farreate,” from the sharing of far, which we call zea [spelt]; for this was the ancient and, for a long time, the ordinary food of all the Romans, and their country produces an abundance of excellent spelt. And as we Greeks regard barley as the most ancient grain, and for this reason begin our sacrifices with barley-corns which we call oulai, so the Romans, in the belief that spelt is both the most valuable and the most ancient of grains, in all burnt offerings begin the sacrifice with that. For this custom still remains, not having deteriorated into first-offerings of greater expense[3].

On the basis of a type of “first communion” in which husband and wife shared in a joint offering of spelt, and the sharing of property that this offering signified and celebrated, the marriage bond was firmly established:

The participation of the wives with their husbands in this holiest and first food and their union with them founded on the sharing of all their fortunes took its name from this sharing of the spelt and forged the compelling bond of an indissoluble union, and there was nothing that could annul these marriages[4].

The ideal was made reality—that is, as long as both parties behaved themselves, or especially, as long as the women behaved themselves. “Obligations” are placed on both, but the obligation of the wife is obedience, whereas that of the husband is the ruling of the wife. The only tangible benefits for honoring these obligations are those granted to the obedient wife:

This law obliged both the married women, as having no other refuge, to conform themselves entirely to the temper of their husbands, and the husbands to rule their wives as necessary and inseparable possessions. Accordingly, if a wife was virtuous and in all things obedient to her husband, she was mistress of the house to the same degree as her husband was master of it, and after the death of her husband she was heir to his property in the same manner as a daughter was to that of her father; that is, if he died without children and intestate, she was mistress of all that he left, and if he had children, she shared equally with them[5].

This sets up a corollary: If the only tangible benefits are for the wife’s obedience, then there are tangible punishments for the wife’s disobedience. Conversely, since there are no tangible benefits to a husband’s honoring of his obligations, neither are there tangible punishments for infractions on his side. This pattern is based on the premise of an already existing structure in which the husband (or more generally, the oldest male in direct ascent) was called the “father of the household” paterfamilias, and as such held the power over his wife and children, or paternal power, patria potestas. That power gives him (or a council of his relatives along with him) the right to punish his wife for major offenses, the most serious of which, punishable by death, were adultery and drinking wine, the latter of which was assumed to lead to drunkenness, and that ultimately to adultery[6]:

But if she did any wrong, the injured party was her judge and determined the degree of her punishment. Other offenses, however, were judged by her relations together with her husband; among them was adultery, or where it was found she had drunk wine—a thing which the Greeks would look upon as the least of all faults. For Romulus permitted them to punish both these acts with death, as being the gravest offences women could be guilty of, since he looked upon adultery as the source of reckless folly, and drunkenness as the source of adultery[7].

Dionysius began the discussion by describing a reward system that led wives to want to be obedient to their husbands, referring to it as a “single law.” It turns out, however, that there is also a severe punishment attached to disobedient acts of the wife, and one has to wonder then just how different Romulus’ laws were from those of other societies when in fact he brought nothing other than what is commonly found in law codes: reward the good; punish the evil. Perhaps Dionysius wanted to emphasize the presence of a reward and its value as something new, as opposed to systems that only punish the bad without tangible incentives for proper behavior. At any rate, Dionysius did include the law of punishment in Romulus’ display of wisdom. In fact, it is not really the “single law” of joint ownership, but “this law” of punishment that brought about a long period of “indissoluble” marriages:

And both these offences continued for a long time to be punished by the Romans with merciless severity. The wisdom of this law concerning wives is attested by the length of time it was in force; for it is agreed that during the space of five hundred and twenty years no marriage was ever dissolved at Rome. But it is said that in the one hundred and thirty-seventh Olympiad, in the consulship of Marcus Pomponius and Gairus Papirius, Spurius Carvilius, a man of distinction, was the first to divorce his wife, and that he was obliged by the censors to swear that he had married for the purpose of having children (his wife, it seems, was barren); yet because of his action, though it was based on necessity, he was ever afterwards hated by the people[8].


B. The Dates of the First Roman Divorce

Dionysius referred earlier to marriages as “indissoluble.” Of course it is assumed that when a husband or wife dies, the marriage naturally ceases to exist. By granting the right of the husband to put a wife to death, marriages can therefore still be ended by the will of at least one of the parties apart from “natural causes.” Dionysius’ mention of offenses being punished with merciless severity proves that Romulus’ laws didn’t actually result in a long period of exclusively harmonious marriage unions. And even when there was no cause for the death penalty, one cannot infer the existence of happy households. Therefore an interesting historical question raised by Dionysius’ treatment of early Roman marriage remains to be resolved: Is it not inconceivable that a society could exist without a single case of divorce for over half a millennium on the basis of two simple laws: community property and capital punishment for a limited number of offenses? Are Dionysius’ sources in error? Or is Dionysius fabricating his account for some personal or political agenda? Before these questions can be handled, other sources need to be compared to the account of Dionysius.

The tradition related by Dionysius is repeated, with a few variations and additions, nearly two centuries later by Aulus Gellius in the Noctes Atticae (ca. 150 C.E.), 4.3.1, who cites an author other than Dionysius:

It is on record that for nearly five hundred years after the founding of Rome there were no lawsuits and no warranties in connection with a wife’s dowry in the city of Rome or in Latium, since of course nothing of that kind was called for, inasmuch as no marriages were annulled during that period. Servius Sulpicius too, in the book which he compiled On Dowries, wrote that security for a wife’s dower seemed to have become necessary for the first time when Spurius Carvilius, who was surnamed Ruga, a man of rank, put away his wife because, owing to some physical defect, no children were born from her; and that this happened in the five hundred and twenty-third year after the founding of the city, in the consulship of Marcus Atilius and Publius Valerius. And it is reported that this Carvilius dearly loved the wife whom he divorced, and held her in strong affection because of her character, but that above his devotion and his love he set his regard for the oath which the censors had compelled him to take, that he would marry a wife for the purpose of begetting children[9].

Gellius repeats the story later in Book 17, drawing on a different source. In introducing the chapter (17.21.1), he at first highlights his bent for accuracy in order to avoid the embarrassment of erring in public:

I wished to have a kind of survey of ancient times, and also of the famous men who were born in those days, lest I might in conversation chance to make some careless remark about the date and life of celebrated men, as that ignorant sophist did who lately, in a public lecture… In order, I say, to guard against such errors in dates and periods of time, I made notes from the books known as Chronicles of the times when those Greeks and Romans flourished who were famous and conspicuous either for talent or for political power, between the founding of Rome and the second Punic War[10].

In the following sentences, however, Gellius presents a disclaimer:

And these excerpts of mine, made in various and sundry places, I have now put hastily together. For it was not my endeavor with keen and subtle care to compile a catalogue…but merely to strew these Nights of mine lightly here and there with a few of these flowers of history[11].

Now to his second account of the first divorce in Rome (17.21.44):

Five hundred and nineteen years after the founding or Rome, Spurius Carvilius Ruga, at the advice of his friends, was the first Roman to divorce his wife, on the ground that she was barren and that he had taken oath before the censors that he married for the purpose of having children[12].

Below is a summary comparing the accounts of Dionysius and Gellius:

Year of Divorce (after founding of Rome, or from the time of Romulus)



Alleged Source

D 2.25.7

523 years

Not given specifically. In general, Dionysius claims his sources came “orally from men of greatest learning…and…histories written by approved Roman authors…” D 1.7.3

G 4.3.1

Nearly 500 years

(See next.)

G 4.3.2

523 years

Servius Sulpicius, On Dowries

G 17.21.44 519 years Chronicles




D 2.25.7

Marcus Pomponius, Gaius Papirius

G 4.3.2

Marcus Atilius, Publius Valerius

G 17.21.44

(not given)

Reasons for the Divorce



D 2.25.7

Obliged by the censors that purpose of marriage was to bear children

His wife had no children

G 4.3.2

Oath which the censors had compelled him to take, married for children

No children were born to her

G 17.21.44

Oath before the censors, married for children

Wife was barren

Instigators — only G 17.21.44

At the advice of his friends

Spurius Carvilius’ Relationship to Wife — only G 4.3.2

Loved her dearly; held her in strong affection because of her character

Held his oath above his devotion and love to his wife

Reactions of the People — only D 2.25.7

Ever afterwards hated by the people

If these were the only accounts of this first divorce, the differences would hardly suggest an entirely confused state of affairs. Whether it was 523 or 519 ab urbe condita; exactly who were the consuls, and whether all of them were intended to be given by both writers; whether it was really the friends, the censors, or Carvilius himself who initiated the divorce; whether he really loved his wife, or it was merely an alleged love to make him appear noble; whether people really hated him ever afterwards—these issues don’t detract from the consensus that: (1) Spurius Carvilius, (2) divorced his wife, (3) about 230 B.C.E., and that (4) this was regarded as the first recorded divorce in Roman history.

As far as these basic points are concerned, the situation would be clear (though still not entirely to be explained) if it weren’t for the fact that other authors have recorded different memories about the case of Spurius Carvilius. But even in these cases the differences involve only the date of the divorce, not the person, nor the circumstances, nor the assertion that this is the “first” recorded divorce.

Plutarch (1st–2nd cent. C.E.) records the event as having occurred in the year 230 A.C.U., or 524 B.C.E., in his Lives in two passages, Lyrcugus–Numa [25(3).12.13] and Theseus–Romulus [35(6).3.4]:

…the Roman writers report it as the first example, that Spurius Carvilius divorced his wife, being a case that never before happened, in the space of two hundred and thirty years from the foundation of the city.

And to the reverence and love and constancy he [Romulus] established in matrimony, time can witness; for in two hundred and thirty years, neither any husband deserted his wife, nor any wife her husband; but, as the curious among the Greeks can name the first case of parricide or matricide, so the Romans all well know that Spurius Carvilius was the first who put away his wife, accusing her of barrenness[13].

Now the difference in dates is significant: Between the date assigned by Dionysius and Gellius on the one hand, and that assigned by Plutarch on the other, lies a difference of over 290 years[14]. And there is yet another date assigned to this event, this time by Valerius Maximus (1st cent. C.E.) in his Facta et Dicta Memorabilia (2.1.4), but again the name, cause, and mention as the first recorded divorce agree with Dionysius, Gellius, and Plutarch:

Divorce between a wife and a husband did not take place from the foundation of the city until the one hundred and fiftieth year. Now Sp. Carvilius was the first to divorce his wife, for the cause of barrenness[15].

On the basis of another passage by Valerius Maximus, modern authors[16] point out the possibility that it was not Spurius Carvilius, but a certain L. Annius who was the first to divorce. The passage relates why the censors removed Annius from the Senate at a date of 307/06 B.C.E.:

They removed L. Annius from the senate, because he had divorced the virgin he had married without taking calling in the counsel of his friends[17].

The suggestion that L. Annius was really the first person on record to divorce assumes that Dionysius and Gellius are correct in regard to the dates of Carvilius’ divorce, and that Valerius Maximus is wrong that the date, but correct in the account of L. Annius. But for our purposes, none of these conjectures or reconstructions really help settle the fundamental question: Can it be that whether for 150 years, 230 years, 519 years, or 523 years, there was not a single case of divorce that occurred in Rome?

C. The Nature of Divorce Laws and Practice

Now apart from the divergences of dates in regard to the first divorce, we are also faced with divergences concerning the earliest causes for divorce, or whether or not in principle divorce was allowed to occur. Dionysius says marriage was indissoluble, but eventually did happen. Plutarch agrees that at some time divorce did happen for the first time, but disagrees with Dionysius not only over when it happened, but over the fundamental question of the nature of the marriage and divorce laws instituted by Romulus. In Romulus, 22:3 Plutarch records:

He [Romulus] instituted also certain laws, one of which is somewhat severe, which suffers not a wife to leave her husband, but grants a husband power to turn off his wife, either upon poisoning her children; or counterfeiting his keys, or for adultery; but if the husband upon any other occasion put her away, he ordered one moiety of his estate to be given to the wife, the other to fall to the goddess Ceres; and whoever cast off his wife, to make an atonement by sacrifice to the gods of the dead[18]. 

Accordingly when the right to divorce surfaced, it was at first unilateral in favor of a husband divorcing his wife, but only for a limited number of cases. Other cases would be theoretically possible, but they would practically be ruled out by the consequent  bankruptcy of the husband. This is all quite different, however, from Dionysius’ account of Romulus’ laws.

As a final piece of early testimony, Cicero (1st cent. B.C.E.) refers to the earliest written Code of Roman Law—the Twelve Tables (ca. 450 B.C.E.). Unfortunately, they have come down to us only in fragmentary form. Cicero remembers the following about the Twelve Tables in general in De Oratore, I.44:

Though all the world exclaim against me, I will say what I think: that single little book of the Twelve Tables, if anyone look to the fountains and sources of laws, seems to me, assuredly, to surpass the libraries of all the philosophers, both in weight of authority, and in plenitude of utility[19].

In Philippics, 2.69 Cicero discusses the case of Antony’s divorce of his mistress, and tells us something about what the Twelve Tables recorded as to divorce procedure:

Now in his house every bedchamber is a brothel, and every dining room a cookshop. Although he denies this, do not, do not make inquiries. He is become economic. He desired that mistress of his to take possession of whatever belonged to her, according to the laws of the Twelve Tables. He has taken his keys from her, and turned her out of doors. What a well-tried citizen! of what proved virtue is he! the most honorable passage in whose life is the one when he divorced himself from this actress[20].

The pattern for divorce according to Cicero’s recollection of the Twelve Tables is thus: (1) pronouncement of injunction to take personal belongings, (2) turning in of keys, (3) sending her out of the house.

Now, if the first historical divorce in Rome didn’t take place until 230 B.C.E. because Romulus never provided for divorce to happen, as Dionysius asserts, it seems strange that the Twelve Tables should make provision for divorce in 450 B.C.E., when there was no practical need for such a law for another 220 years! On the other hand, if either the dates 604 or 574 B.C.E. are correct, then divorce that happened contrary to Romulus’ laws needed regulation, which may have given rise to the inclusion of divorce laws in the Twelve Tables. In any event, the historical facts are not at all clear. Various reconstructions are possible, and such reconstructions are likely to be based on what agendas are attributed to the authors who wrote on the subject of divorce in early Rome.


Part II: Discussion of the Accounts

In analyzing the divergent accounts of various ancient authors on early Roman divorce, we will present two areas that are deserving of further study, one involving the similarities of the account, the other involving their differences.

A. Issues Involving the Similarities in the Accounts

The striking similarities in so many of our authors in regard to the first recorded divorce, namely that of Spurius Carvilius at some time quite distant from the age of Romulus needs explanation. Common sense, as well as the recorded activity of societies elsewhere in the world in all times, ancient and modern, make it difficult to believe that divorce in early Rome did not occur for any period at all of its existence, let alone for a period over a hundred or even hundreds of years. If in fact the writers are correct, we have an anomaly of history that has yet to be adequately explained from a sociological context. If on the other hand our authors are in error, the question remains: Were their sources in error, or were they attempting to promote some agenda by skewing the past in their own directions? If the sources were in error, the same question will repeat itself, but we will not be in any position at all to address it since we do not know the sources, and therefore cannot infer anything in terms of agendas that may have been relevant during their own times.

Now if we intimate hidden agendas for our authors, we can begin by noting that they all lived either shortly before, during, or shortly after the Augustan reforms in family law[21]. Their writings may be a response to—mostly in favor of—those laws. Such an approach is taken by Treggiari in Roman Marriage. In discussing Dionysius, who wrote during the time of Augustus, Treggiari sees an attempt by Dionysius to draw a parallel between Romulus and Augustus in support of the Augustan laws.

It is, however, worth noting that he [Dionysius] picks out the indissolubility of Roman marriage as a praiseworthy characteristic and that he agrees with the philosophers and Cicero in idealizing community of property and the consequent absorption of the wife into her new family. He published his work at a time when the new Romulus [emphasis mine, GM] had imposed state controls on marriage, attempting both to regulate and encourage marriage and to check adultery. Romulus, theoretically, was starting from nothing and could create institutions good in themselves. Augustus, finding a corrupt system, in which, for instance, the indissolubility of marriage and community of property had ceased to be the norm, could only tinker with it and hedge the family about with regulations and penalties. [22]

Treggiari mentions other authors with similar views as Dionysius, and it may well be that, although the authors we have investigated here differ in certain details, they all tend toward a view that marriage needs a boost in the times in which they live. Thus, their “memories” of the past are not only selective, but in some cases clearly fabricated to support their views.

B. Issues Involving the Differences in the Accounts

It is less clear as to why the accounts differ in their details. Why, for example, should the dates of Spurius Carvilius’ divorce so widely differ? What would account for Dinoysius to postulate a “no divorce” period under Romulus, while Plutarch gives regulations? Perhaps in these details there is a combination of real “divergent memories”—source materials that simply differ, and conscious, but variable, modifications for the purpose of either softening or intensifying the agenda for reform held by each author. This would mean that the hypothesis suggested by the subtitle: “Divergent Memories or Hidden Agendas” is too simplistic. If both are present and interwoven in our materials, it will require a much deeper analysis to identify and separate the two.




Dixon, Susan. The Roman Family. Baltimore and London: The Johns Hopkins University Press, 1992.

Rawson, Beryl. Marriage, Divorce, and Children in Ancient Rome. Oxford: Clarendon Press, 1991.

Rawson, Beryl and Weaver, Paul, eds. The Roman Family in Italy: Status, Sentiment, Space. Oxford: Clarendon Press, 1997

Scafuro, Adele C. The Forensic Stage: Settling Disputes in Graeco-Roman New Comedy. Cambridge: Cambridge University Press, 1997 (pp. 309–313).

Treggiari, Susan. Roman Marriage: Iusti Coniuges from the Time of Cicero to the Time of Ulpan. Oxford: Clarendon Press, 1991.

Augustan Laws

Mette-Dittmann, Angelika. Die Ehegesetze des Augustus: Eine Untersuchung im Rahmen der Gesellschaftspolitik des Princeps. Stuttgart: Franz Steiner Verlag, 1991.

Csillag, Pal. The Augustan Laws on Family Relations. Budapest: Akadémiai Kiadó, 1976.

[1] 21st–20th cent. B.C.E, Laws of Ur-Nammu, 8 of the extant portions of 25 laws deal with marriage-related issues; 19th cent. B.C.E., Lipit-Ishtar Code for Sumer and Akkad, 13 of 38; 19th–18th cent. B.C.E., Laws of Eshnunna, 10 of 59; 18th–17th cent., Code of Hammurabi, 60 of 282. For the laws see Pritchard, Ancient Near Eastern Texts Relating to the Old Testament, (ANET) 3rd edition (with supplement) (Princeton University Press: 1969). Of course, these laws are extant primarily due to the nature of the material upon which they were written — baked clay tablets which didn’t disintegrate through the ages.

[2] Cary in Loeb, I, p. 381.

[3] Ibid. pp. 381–382.

[4] Ibid. p. 383.

[5] Ibid. p. 383.

[6] Treggiari, Roman Marriage, pg. 461 footnote 120: “Moderns usually connect this prohibition [drinking wine] with the risk of sexual misconduct…” Not only moderns, but Dionysius himself is explicitly of this opinion.

[7] Cary in Loeb, I, pp. 383–384.

[8] Ibid. p. 384.

[9] Rolfe in Loeb, I, 323.

[10] Rolfe in Loeb, III, 273.

[11] Ibid. pp. 273–275.

[12] Ibid. pp. 285–287.

[13] Texts from [to complete].

[14] One is tempted here to suspect a transposition of dates A.U.C. and B.C.E. by later copyists. In Dionysius and Gellius, it’s 523 A.U.C, or ca. 231/0 B.C.E. In Plutarch it’s 230 A.U.C, or ca. 524/3 B.C.E. But the dates are clearly ab urbe condita, and furthermore dating by B.C./A.D. (now B.C.E./C.E.) wasn’t begun until the 6th century C.E. (Dionysius Exiguus, 525), and the new system and wasn’t widely used for B.C.E. dating until many centuries later. Also, there remains another alternative given below by Valerius Maximus of 150 A.U.C., or 604 B.C.E.

[15] Repudium inter uxorem et uirum a condita urbe usque ad centesimum et quinquagesimum annum nullum intercessit. Primus autem Sp. Caruilius uxorem sterilitatis causa dimisit. [Text from Internet site: http://www.gmu.edu/departments/fld/CLASSICS/valmax2.html]

[16] Treggiari, Roman Marriage, 442; Scafuro, The Forensic Stage, 310.

[17] L. enim Annium senatu mouerunt, quod quam uirginem in matrimonium duxerat repudiasset nullo amicorum [in] consilio adhibito. [Text from Internet site: http://www.gmu.edu/departments/fld/CLASSICS/valmax2.html]

[18] Text from [to complete].

[19] Text from Perseus Internet site.

[20] Text from Perseus Internet site.

[21] The laws on marriage are: Lex Iulia de maritandis ordinibus (18 B.C.E.) and Lex Papia Poppaea (9 B.C.E.); laws on divorce are: Lex Iulia de adulteries coercendis (18 B.C.E.), and their primary sources are the commentaries of later jurists.

[22] Treggiari, Roman Marriage, p. 212.