Skip to content Skip to navigation

More Caveats than Conclusions about Roman Singles

Monday, September 5, 2022 - 07:00

It's common for articles about demographic studies to focus heavily on the methodology and definitions used for interpreting the data. This is of vital importance, as all such interpretations are conditional on the accuracy of the premises. But this sort of approach can give the impression that nothing at all is known for certain. To some extent, that's an accurate impression if one focuses on the "for certain" part, though not with regard to the "nothing is known" part. Still, in exploring "how we know what we know" the authors necessarily lay out a complex understanding of the social structures and conditions that underlie their interpretations. It's all fascinating.

This is my last day at Worldcon in Chicago and I'm down to a book signing and a small group fan-chat. I've had a good time, though I've also learned some shifts in the boundaries of my energy. Most of the panel discussions I've participated in have not simply been entertaining, but have been exciting and energizing. Particularly memorable were a discussion of "proto-sci-fi" works, especially the long history of "hollow earth" stories; an examination of gendered magic in historic fantasy; and a look at how alternate history stories can give voice and visibility to historically marginalized characters.

It's too early to tell whether my Covid precautions have held, especially since I have another day of airports to get through tomorrow. The convention required proof of vaccination and universal masking, but of course it's still possible for those to fail if the luck goes against you. Keep your fingers crossed for me, and hold a thought for those who would like to return to attending in-person conventions but don't yet consider it safe.

Major category: 
Full citation: 

Goessens, Thomas. 2019. “Singles and Singleness in the Christian Epigraphic Evidence from Rome (c. 300-500 CE)” in Sabine R. Huebner & Christian Laes (eds), The Single Life in the Roman and Later Roman World. Cambridge University Press, Cambridge. ISBN 978-1-108-47017-9

Publication summary: 

A collection of papers addressing (and definine) the state of "singleness" in the Roman Empire, both in pre-Christian and early Christian times. There is a strong focus on Egypt as well as Rome proper, as well as wider Byzantine material. Comparative material is offered from Jewish sources, as well as a small selection of studies from specific cultures of more modern date.

Goessens, Thomas. “Singles and Singleness in the Christian Epigraphic Evidence from Rome (c. 300-500 CE)”

Late Antique Christianity: The Rise of the Ideal of Being Single

# # #

This article uses early Christian funerary inscriptions in the city of Rome as a data source for life-long singleness, allowing for a quantitative and statistical analysis. The corpus of relevant inscriptions includes over 40,000 items though many are fragmentary. As the vast majority of inscriptions from this period are funerary in nature, and due to the typical content of such inscriptions, we have perhaps 20,000 epitaphs that include not only the name, but also age at death, length of marriage (if any), and references to familial relationships. In date, they range from the early 3rd to mid-7th century CE, thugh most fall in the century between the death of Constantine the Great and the early 5th century.

Several caveats are necessary. Due to social differences in burial preferences, these inscriptions are less likely to belong to the social elite. Due to Christian avoidance of identifying people by social status, it is not possible to make group distinctions between enslaved people, freedmen, and citizens. In addition, the shift from use of the tria nomina name formula to the use of a single name can make family connections difficult to trace. The epigraphic information rarely touches on questions of divorce and remarriage.

Given these caveats, the criteria used to classify an inscription as referring to a “single” person are: of marriageable age and not in a spousal relationship at the time of death (regardless of reason). Three women are used as examples to introduce the analysis.

Maximilla died in 389 at age 51. She was the daughter of a diaconus (deacon), but the epitaph was commissioned by a friend, the daughter of a man of senatorial rank. Maximilla is described as virgo (virgin) and ancilla Dei (female-servant of God). There is no reference to a spouse. It is a reasonable conclusion that she never married and may have deliberately chosen a life of consecrated virginity.

Cassia Sophrosyne was commemorated in 402 by her niece Cassia Vindicia. Although Sophrosyne’s age is not given, the fact that she had an adult niece suggests she may have been past the age of expected marriage. Sophrosyne is described a a virgo sacra (sacred virgin) respected for her sexual abstinence, and Vindicia identifies herself as a virgo Deo dedicate (virgin dedicated to God), though her age is not known.

The contextual information for these three women is atypically detailed, and it’s possible that such details over-represent a religious motivation for singlehood (as other reasons would not be celebrated for posterity). In addition, all three belonged to the social and economic elite. Therefore caution must be observed in extrapolating from the more detailed inscriptional data.

Singlehood can either be mentioned explicitly, or implied by the lack of reference to a spouse, with the latter being more common. Many inscriptions contain only the name of the deceased (which indicates gender) accompanied by formulaic expressions. Most often, there is no mention of what the relationship is to the person doing the commemoration, and the author has chosen to exclude these from analysis as no conclusion about singlehood can be made. Another problematic group are epitaphs that indicate they were arranged for by the deceased themselves—a context that suggests but does not prove single status.

Searching for more certainty, the author considers what explicit language might indicate single status. “Caelebs” is a good candidate but in funerary inscriptions seems to be limited to military veterans. “Vidua” (widow) may also be used to indicate an unmarried woman (as opposed to a woman whose husband had died but who might have remarried). (There seems to be a suggestion that “vidua” might also be used to mean “never-married”?) Overall, the author concludes that the use of specific terminology is unhelpful in identifying those unmarried at death.

Use of the word “virgo” is also a problem for interpretation as it is associated primarily with young women (up to age 25), and doesn’t absolutely correlate with unmarried status. Overall both female and male virgins are highly likely to be never-married, but it may sometimes refer to a married person in a chaste relationship. To confuse the issue, the words virginia/virginius are used almost exclusively for married people and seem to refer to virginal status at the time of marriage. The phrases “virgo Dei” or “virgo sacra” seems much more likely to refer to a deliberate never-married state, on religious principles, but these phrases are quite rare and there is some indication the label would only be applied after reaching a certain age. And in contradiction to this interpretation is Ioviniana who was described as “virgo sanctae memoriae” but also as a wife and mother.

The use of the formulas “ancilla Dei” or “servus Dei” [Note: typically translated “servant of God” but with a nod to my classicist friend who emphasizes this point, a more contextual translation is “slave of God”] seem to refer purely to religious devotion with no implication regarding marital status, with some explicitly noted as being married.

References to the status of clergy carries no indication of singlehood as there was no requirement for clergy to be unmarried, though it was considered an ideal.

The author once again comes to a conclusion that identifying singles via specific keywords in the inscriptions is fraught with uncertainty. He moves on to an analysis of inscriptions in which it is married status that is explicitly indicated. Based on the demographic data in the inscriptions, a normative life course for married people can be identified, after which this can be used as a context for evaluation. Then a metric can be established to categorize people who had passed the normative age for marriage (at the time of death) but for which no explicit reference to marriage was present.

After much hedging about, he suggests that women married between the ages of 14 and 21, and men between 20 and 25, with an average for women of 20 and for men of 26. This estimate might be corroborated from inscriptions that include an age at death and length of marriage, but there are relatively few that contain this data. This approach suggests that women married between 12 and 27 (with 90% married by that age) and men between 18 and 34 (with 90% married by that age).

After yet more hemming and hawing about what weight can be put on the data, th author proposes that for lower estimates of expected age at marriage, perhaps 1 out of 7 women were unmarried at death and 1 out of 4 men; while for higher estimates of expected marriage age, perhaps 1 in 10 women and 1 in 7 men were unmarried at death. But then he notes only that this 1/10 and 1/7 no doubt included some single people, though not necessarily never-married ones.

All in all, the paper takes a great deal of time and analysis to conclude that we can’t really be certain about anything except in the few specific cases where the person’s singleness is explicitly noted in the inscription. But there are some interesting data tables for specific keywords in the inscriptions and graphs for some of the demographic patterns.

Time period: 
Place: 
Misc tags: 
historical