BMCR 2006.05.29

The Age of Marriage in Ancient Rome

, , , The age of marriage in ancient Rome. Studies in classics ; v. 26. Lewiston, Queenston, Lampeter: Edwin Mellen Press, 2003. x, 146 pages ; 24 cm.. ISBN 0773466657. $99.95.

The Age of Marriage in Ancient Rome (AOMIAR) by Arnold Lelis, William Percy, and Beert Verstraete reconsiders the common beliefs regarding the age at first marriage (AAFM) for both males and females in the Roman world. It questions the current methodology and seeks to establish a new means of considering a topic traditionally based on a data set that can only be described as sparse and inadequate when compared to that of other investigations which utilize modern statistical and demographic analysis techniques. Furthermore, the authors note that historians of the modern era are too often influenced by their own cultural norms and biases and will not allow the possibility of unusually early marriages within industrialised nations even during earlier, pre-modern periods when such situations might have been warranted.

In the opening chapter, the authors inform us that, traditionally in pre-modern societies, AAFM for girls corresponded with the visible onset of puberty. In these early societies, high mortality rates demanded an offsetting high-birth rate. As a result, younger marriages were required to facilitate population growth. For the Classical world, some inscriptional evidence supports such a claim, but the data is scanty. Historical evidence which informs us of such associated issues as pro-natalist legislation and infanticide is also relevant, as are works of men such as Plato and Aristotle, who discuss matters of ideal population and the checks required to keep a balanced community. The majority of this evidence, however, is Greek. Nevertheless, the authors insist that this model is suitable for Rome as well, particularly for girls.

Chapter two, which focuses on the nature of the evidence, outlines the authors’ hypotheses and establishes their method of development. They begin by presenting a model that they consider to be the most representative for pre-modern societies, that of an AAFM for females between 12 and 16 and for males on average between 15 and 21 or 22. This model is very different from that of Saller.1 In fact, both Saller and Shaw suggest an age of 28 for males and 18 for females as the “Mediterranean model.”2 The authors admit that such a pattern does emerge around 600 in Greece, but not in Rome. Furthermore, the authors stress that regional variations undermine the possibility of applying a single, all-encompassing model to the Mediterranean or the ancient world as a whole. As evidence, they look at alternate cultural systems found in areas such as Egypt and Israel and note the variations within each with regards to marriage. If indeed there is a constant in the ancient world, it is a young AAFM for females, which is necessary to maintain a 0.5 % growth rate, or a doubling of the population every 140 years.

A primary factor behind disagreement among scholars is the nature of the evidence. Literary evidence promotes a young AAFM for both sexes, while epigraphy implies one much older. The authors have compiled and included in appendices a large database of biographical data drawn from all available sources, including both literary and epigraphic evidence. These appendices, which supplement the text, grew out of an honours thesis by Thomas Permatteo under the supervision of Mr. Percy and are a vast improvement upon past efforts. In all, the authors provide the pertinent biographical information for 83 Roman males and 31 females in two appendices. These numbers represent a substantial increase over the figures offered earlier by Hopkins.3 Hopkins, like Saller and Shaw, tends to shy away from purely literary evidence and instead focuses more on epigraphic evidence, particularly funerary inscriptions. As justification, Hopkins cites what he refers to as the ‘Everest fallacy’: that literary evidence does not offer a typical view of society but instead highlights extraordinary circumstances.4 The authors concur with this sentiment, stating that recorded history is not the story of those leading usual lives (p. 58). They also note, however, that “the careers of the people in [their] appendices … can tell us more about the actual life histories of the elite … than any model based on statistics otherwise devoid of context” (pp. 19-20). In other words, if nothing else, the appendices do allow for a more authoritative albeit limited statement regarding AAFM for upper classes.

For Roman culture as a whole, the authors stress the need for an alternate, non-statistically based approach, particularly for the Republic. In response, they develop what they call a “demographic theory” based on a survey of the significant periods of Roman history and the development of a subsequent demographic model for each. These models, when created independently from the purely statistical data offered in the appendices, may then be compared to the visible trends that emerge within the limited biographical data. The authors admit that this approach is “impressionistic” and offers “plausibility” in place of hard, statistically-based conclusions (p. 19). Nevertheless, the strength of this approach lies in its recognition that statistics are as open to interpretation as other evidence derived from material remains and that numbers alone, devoid of their social and historical context, have limited value.

The one true danger in applying a demographic model to the data is that of modifying the data to fit a presumed conclusion. The authors walk a fine line in this regard, qualifying their conclusions with statements such as “if interpreted according to our hypothesis … “. Nevertheless, the authors everywhere take a neutral view of the data, one that is independent of the constructed models, and thus are able to illustrate important similarities and differences between the two models outside of any foregone conclusion.

Chapter three considers the earliest evidence from Rome down to the second century BC. The epigraphic data from this period supports the idea of a young AAFM but is not representative of all classes. Nevertheless, the authors note that, at a very early time, all classes must have married young in order to account for the great population increase during the Neolithic revolution. Such an AAFM would have been particularly appropriate for women and the lower classes. As the primary component of an increasing agrarian society, labour was at a premium. Thus the idea of early marriages, especially for women, fits in well with an overall increase in births, a substantial shift in gender roles, and the need for employable children, all of which are hallmarks of Neolithic revolution. Elite classes, meanwhile, were exploiting the ager publicus in an attempt to acquire prestige goods. As a result, there was no need to delay marriage. In fact, such behaviour would only have promoted a pro-natalist mentality, as upper classes sought to maximize the production of all available land.

The authors consider also the effects of ongoing seasonal warfare and the payment of taxes through military service. These two factors resulted in astrong need for both labourers and soldiers. Thus larger families and a younger AAFM would have been necessary to ensure that these requirements were met. Success in war meanwhile prompted an increase in colonial enterprise. Livy (27.9.11) says this enterprise was initiated in an attempt to increase the size of families beyond the limits allowed by the available farmland in Italy and not necessarily in response to an already ever-increasing population.

Perhaps the most interesting element of this demographic model for early Rome is the role of the paterfamilias in arranging the marriages of his immediate and extended family. For elites, these marriages helped control such aristocratic interests as the preservation of territory. As a result, earlier betrothals were more convenient and effective than later ones. Furthermore, there was no need to wait for financial situations to improve, as evidence for inflated dowries would seem to suggest. In fact, as the authors state, “most writers concede that senatorial female AAFM were very early” (p. 37, n.13). The authors extend early marriages to males as well, allowing a range of 17 to 22 for AAFM, since boys at this age would be tractable and more easily exploited by their fathers.

In chapter four, the authors observe that this pattern of early AAFM does not seem to vary significantly in the later Republic and into the Empire. In fact, economic exploitation by elites is even more prevalent between 100 and 31 BC and becomes archaeologically visible with the spread of elite-style villas. Marriages do, however, vary slightly with the increased prevalence of sine manu weddings and no-fault divorce. As a result, this period shows a higher number of men with older wives than would be expected. The authors stress, however, that these disparate data do not alter the overall model, but rather underscore the dangers of trying to identify common tends in a history fraught with frequent macrohistorical changes. This period in particular is known for its excessive political violence and features various disruptions of normal family life and marriage patterns. As a result, this era should not be seen as the norm, but rather, like all others, should be considered individually.

Throughout the early and mature Empire, the demographic data demonstrates a wide spectrum of AAFM for both sexes. Within the royal family, Augustus commonly arranged early marriages for his extended family while Tiberius did just the opposite. Certainly early AAFM would be important with respect to dynastic continuity. For Augustus, however, early marriages may also have represented an attempt to re-establish traditional values. Thus, as the authors rightly suggest, the actions of Augustus may help substantiate previous conclusions with regards to early AAFMs for earlier periods.

The delayed marriages promoted by Tiberius, meanwhile, may reflect a growing anti-natalist movement within the common citizenry of Rome. Contributing to this trend was the practice of delaying marriage until after military service. This behaviour would have had a drastic effect on the AAFM for men, but a slight one for girls. According to the authors, the result was an increase in older grooms with younger brides. Often this arrangement would include even prepubescent girls. To combat this shift in marital practices, the Lex Julia demanded that penalties be placed on those who did not marry and procreate by a certain age. To guarantee inheritance, therefore, men were forced to marry young, by 18 to 21 at the latest, while, according to the authors, precedence allows for even younger AAFM.

In chapter five, the authors pause from their survey of the main periods of Roman history to consider the epigraphic record and its value as evidence for AAFM. As they mention on several occasions, prior investigators tended to focus almost exclusively on epigraphic evidence, particularly funerary inscriptions, believing it to be more representative of the whole of society than is literary evidence. As the authors clearly demonstrate, however, such inscriptions, while not exclusively elite, also tend to focus more on respectable classes and not on the plebs or slaves. In addition, they note that, throughout the long history of scholarship, there has been no real consensus on what exactly such evidence can tell us about the demographic make-up of society. In support, they cite Parkin, who states that “using funerary inscriptions to reach conclusions about demographic behaviour of ancient populations is an endeavour fraught with special problems.”5

The authors do not deny the value of epigraphic evidence in the study of Roman demography. Issues such as dowries, inheritance, and life expectancy—all dependent on the study of inscriptions—are also relevant to the study of AAFM. Instead, the point of this chapter would seem to be that epigraphic evidence is as limited as literary or historical and that by no means should it serve as the sole basis for the study of Roman demography in any form. Instead, it should be an important component of the full spectrum of evidence just as it is in the investigation at hand.

In chapter six the authors conclude their examination of ancient Rome with a brief consideration of marriage in the late Empire. Here the authors demonstrate that similar trends visible throughout Roman history are generally present in later epochs. The only noticeable shift occurs at the onset of Christianity: epitaphs from this period seem to demonstrate a higher AAFM than ever before. No longer were elites marrying off their children to the same extent. Also, new aversions had arisen towards sexuality, particularly involving youths that were deemed to be underage.

In the end, the authors admit that their monograph does little to provide empirically significant conclusions nor do they in any way suggest that the evidence proves an early AAFM for most Romans. Instead, they state that the “overall shape of the available evidence, even though it does not allow us to fix the ages of first marriage with extreme precision, strongly suggest” an early AAFM for most Romans. They are also aware that inconsistencies within the data set stand in opposition to the demographic model they have created. Given that the authors have not relied on purely statistical analyses, however, such discrepancies by no means undermine the integrity of the conclusions.

The real success of this work lies in the two innovations that the authors bring to the subject of Roman marriage. First, they provide a complete corpus of the available demographic data concerning ancient marriage. This data is drawn from a variety of sources including epigraphic, literary, historical, and archaeological evidence. The result of these efforts is a very impressive set of appendices of AAFM for both men and women from a variety of historical periods. Although the elites are best attested, almost all classes are at least represented; slave marriages are not considered. To supplement this core data, the authors also consider evidence from the Greek world, and draw on a number of comparable sources when necessary. The result is the most comprehensive collection of data on the topic to date, an ideal source for anyone approaching the subject for the first time. In fact, the authors stress that this work was targeted for a wider audience than mere Classics scholars. This statement reveals the importance and relevance of both the subject matter and the approach and likewise reiterates the need for this work.

Second, the authors develop a strong theoretical model that may be applied effectively to the topic of AAFM from any period. They use this model in tandem with their statistical analysis, not in place of it. One of the greatest weaknesses in our field is the idea that statistical data do not require the same degree of interpretation as other kinds of evidence. In essence, raw data, no matter how complete or dense, still requires interpretation and integration into the ever-evolving understanding of the ancient culture it illuminates.


1. R. P. Saller, “Men’s Age at Marriage and its Consequence in the Roman Family,” CP 82 (1987), 22-30; id., Patriarchy, Property and Death in the Roman Family (Cambridge, 1994), 26-41.

2. R. P. Saller, “Men’s Age at Marriage and its Consequence in the Roman Family.” CP 82 (1987), 21; and id., Patriarchy, Property and Death in the Roman Family (Cambridge, 1994) 25-26; B. Shaw, “Age of Roman Girls at Marriage: Some Reconsiderations,” JRS 87 (1987), 30-1.

3. K. Hopkins, “The Age of Roman Girls at Marriage,” Population Studies 18 (1964-5), 309-27.

4. K. Hopkins, Death and Renewal, Sociological Studies in Roman History, Vol. 2 (Cambridge, 1983), 41.

5. T. G. Parkin, Demography and Roman Society, (Baltimore, 1992), 18.