The Age of Marriage in Ancient Rome

Review by James Jope published in Italian Quarterly winter/spring 2006

Arnold A. Lelis, William A. Percy, and Beert C. Verstraete, The Age of Marriage in Ancient Rome. Studies in Classics vol. 26, Edwin Mellen Press, 2003

Richard P. Saller and Brent Shaw have studied the age of marriage as indicated by tombstones from the Roman Empire using modern demographic methods, and concluded that men and women married later than traditionally believed. They believe that men, previously thought to marry around 20, wed for the first time in their late twenties; and women, previously thought to marry around the time of menarche, married only late in their teens. The traditional view had been that men wed in their late teens or early twenties and women from 12 to 15.

Lelis, Percy and Verstraete (henceforth ‘LPV’) set out to refute the demographers’ view in this book, and defend the traditional belief.

This is actually the latest round in a dispute dating over a century. Harkness (1896) already argued from inscriptions for late marriage, but Hopkins (1964), for example, albeit himself using mainly epigraphic evidence, rejected Harkness’ interpretation of the data, remarking that Harkness was motivated by Victorian revulsion at the supposed abuse of young girls.

Indeed, more important than the narrow question of marital age are the broader conclusions drawn by authors concerning marriage and the family. Saller classes Rome, under a demographers’ typology, as the ‘Mediterranean type’, and infers that in the later periods and lower classes more reflected by the tombstones, the nuclear family replaced the extended family as the primary structure. LPV insist that the extended family prevailed throughout the Roman period, rejecting the demographic typology as inapplicable to a premodern society.

LPV’s monograph originated in a senior’s honours thesis, and it still reads like a thesis, albeit an outstanding one. The only unfortunate aspect is that the authors’ expressed aim to make it accessible for general readers is neglected. The reader must cope with untranslated and often unexplained legal terminology…

Suffice it here to say that, of the two main types of dowry, a dos profectitia could be recovered by the donor in case of the wife’s death as long as the donor was an agnatic ascendant of the bride. In the case of a dos adventitia, the wife could recover the dowry upon divorce or widowhood by the actio rei uxoriae, which might, however, be reduced in case of her misbehavior or justified expenses on the part of the husband (p. 42)

as well as the mysteries of demographic mathematics:

The relative chances of survival of parents and offspring can be calculated from the same Coale-Demeny Level 3 West-female life-tables that Saller used for his computer model. Realizing that it is not (e0 = 25) but (eaverage fatherhood = x) that is the issue, the following percentages can be calculated… (p. 85)

But the reader who perseveres is doubly rewarded. First, he will find in the appendices a detailed database drawn from written sources, listing far more marriages than were previously known, with biographical circumstances which are important for assessing the data. This admirable compilation, which will be useful for future researchers, could easily have been a separate publication. Secondly, in the body of the text, he will read a diachronic survey drawing on a wide range of ancient sources, which not only backs early marital age, but explains its logic in the overall structure of Roman demography and culture.

Hopkins had already briefly stated the essential points: The high mortality rates of a pre-modern society necessitated early marriage of females and high fertility; and the economic conditions of an agricultural society encouraged an extended family structure. Marriage was not expected to culminate a romantic relationship, but to perpetuate the survival and property of the clan.

LPV explain in detail how these factors unfolded in Rome. Perhaps the strongest aspect of their survey is the thorough presentation of the logic and consequences of Roman law, in particular the patria potestas which was the keystone of the extended family structure-and which, they stress, remained in force throughout the imperial period-and Augustus’ marital reforms. The chapters on the Republic, however, draw on Plautus and Livy as well, and there are several references in subsequent chapters to Pliny’s letters and the imperial historians. The authors allow important concessions signifying a drift toward greater importance for the nuclear family: Marriage sine manu and easy divorce were introduced in the Late Republic, and the lower classes under the Empire were subject to different economic constraints. But the nuclear family could not replace the extended family in a pre-industrial age. And the only group that did indulge extensively in delayed marriage-the old senatorial families during the Empire-became extinct.

The argumentation, as in any fine thesis, is generally reasonable and convincing. Not surprisingly, in view of the controversy, there are a few tendentious exceptions. For example, in spite of Augustus’ traditionalism, it is hardly convincing to present the tyrannical political manipulations of emperors as a typical exercise of patria potestas. Surely any monarchs, especially when their regime is lacking in an established principle of succession, would readily find whatever legal pretexts they required.

As might be expected in a thesis, the authors are always respectful and sometimes too generous towards their opponents. They justify their use of literary as well as epigraphic material, and its speculative assessment, by pointing out that the available material evidence cannot satisfy the requirements of scientific demography, and that the assessment of the tombstones too involves problems and relies on speculation (e.g., how to deal with inscriptions showing the duration of marriage without specifying whether it was the first marriage). Yet still they apologetically accept the designation of their own assessment as ‘impressionistic’.

A glance at one of the major arguments shows little difference between LPV and their opponents in this regard. Saller (1987) compared the frequency of men’s tombstone dedications by parents and by wives. Wives were more frequent for older men, but rare for men between 15 and 19. He concluded that these young men were not married, that wives replaced parents as commemmorators for married men. LPV, however, argue that older men were commemmorated by their wives only because the men’s fathers predeceased them. Analogously, if women dying in their early teens were commemmorated by their parents, Saller inferred that they were not married, but LPV postulate that they had not yet produced surviving offspring and so were still regarded more by their parents than by their husband’s family.

Incidentally, it is surprising that LPV do not object to the inclusion of freedmen’s inscriptions in the sample on which this argumentation is based. A freedman had no paterfamilias. Although his independence was restricted by certain obligations to his former master, he represented the start of a new citizen lineage. His natural father, who, if not dead, might still be enslaved, would not be a likely commemmorator.

If Saller’s explanation seems more ‘natural’ to us, it is because our own society has traditionally been structured on the nuclear family. Both explanations are logically possible; both are unproven; in short, both are ‘impressionistic’. The difference is precisely that Saller’s impression seems to be based on unexamined prejudice- e.g.,

Since in older age brackets…wives…were preferred as commemorators…the complete absence of wife commemmorators for this group surely means that teenage marriages were generally rare for men. (1987, p. 25; my emphasis)

whereas LPV’s conjectures conform with the broader patterns elaborated in their diachronic survey.

As alreadly mentioned, LPV do concede a drift toward a partially enhanced role for the nuclear family precisely in the periods and classes where Saller and others claim to observe it. Some future researcher might be able to construct a reconciliation here. But it would have to be rooted in the kind of broadly based understanding of Roman culture exemplified by LPV’s monograph, and not in the modern presuppositions which seem to have motivated their opponents.

LPV’s use of legal and literary sources to fill out a picture not attested by epigraphy alone does not need any apology. It is rather one of this book’s major strengths. Indeed it should have been taken further. Only the early Republican chapter, with its extensive consultation of Plautus and Livy, really exploits the literary sources satisfactorily. Pliny could be mined more deeply. For example, Letter I, 14, even while attesting the normalcy of arranged marriages, shows an intriguing tinge of hesitancy at placing financial considerations alongside of the character and good looks of the proposed groom. An informed and well reasoned interpretation (a scholarly ‘impression’) of an incident in Petronius which describes the mock wedding and deflowering of a young girl could tell us more about Roman attitudes on this prickly issue. The recurrence in late ancient novels of those far-fetched comedy plots designed to reconcile romantic love with extended-family requirements could be studied to determine to what extent they reflect similar social conditions. Probably also the satiric and the elegiac poets could provide relevant material.

The nuclear family was a Victorian value. It cannot offer the economic security afforded by an extended family, by equal access to careers, or by a welfare state. The persistent role of extended family in immigrant groups throughout the industrial age and in the many households today burdened by the return of underemployed adult offspring shows that the nuclear ideal did not serve the needs of the working poor. It rather served the need of employers for an amenable labour market. Today women’s access to career opportunities, the collapse of real wage levels for menial jobs, and the excessive working hours now required of professional or managerial jobs are already making the ideal of the nuclear family history. It is time to let the Romans too occupy their own place in history.


Harkness, A.G. “Age at Marriage and at Death in the Roman Empire”, Transactions and Proceedings of the American Philological Association (1896) 35-72

Hopkins, M.K. “The Age of Roman Girls at Marriage”, Population Studies 18 (1964-65) 309-327

Saller, Richard P. “Men’s Age at Marriage and its Consequences in the Roman Family”, Classical Philology 82 (1987) 21-34

No comments: