“Are dead babies good evidence for a Roman brothel?”
That’s the question raised by a BBC story about analyses of materials from an almost century-old excavation at a Roman villa in the Thames Valley.
The data: remains of 97 infants, all of whom died close to birth. To the researchers, the coincidence suggests deliberate killing of newborn babies:
Archaeologist Dr Jill Eyers said: “The only explanation you keep coming back to is that it’s got to be a brothel”.
Really? Let’s think about the set of assumptions necessary to get from infants dying close to birth, to the place that they were found being a brothel. While the BBC story (promoting a new BBC TV series) doesn’t sketch these arguments out, the implicit chain of thinking must go something like this:
- having lots of babies is a consequence of having lots of sex
- having babies that die at birth is unnatural so they must have been killed deliberately
- the most likely person to kill them would have been their mothers
- the mothers must have had lots of sex, which means they were sex workers
And of course, there is the link in the chain of interpretation that initially drew my attention, this one made explicit in the BBC article:
 With little or no effective contraception, unwanted pregnancies could have been common at Roman brothels.
As a scholar and teacher, I recognize this entire set of assumptions as what historian of anthropology George Stocking (following in a long line of scholars) called “presentist” history: interpreting the past in terms that are rooted in the present. Most remarkably, the key assumptions are contradicted by what we actually know about Roman family planning, contraception, how to recognize brothels (Roman and otherwise) archaeologically, and the actual features of the specific site at Hambledon in the UK.
Let’s start with the claim that there would have been “little or no effective contraception” in Roman Britain. Since the publication of John Riddle’s books Contraception and Abortion From the Ancient World to the Renaissance (1994) and Eve’s Herbs (1997), there is no excuse for making such an argument. Riddle identifies a wide range of plants used as contraception by the ancient Greeks and Romans.
Riddle also points to a demographic argument suggesting that Romans were successfully practicing birth control: during the first five centuries of the Christian Era, Roman family size apparently was stable (at two children), and life expectancies increased. This pattern extended to the provinces: as historian Malcolm Todd says in his 2004 Companion to Roman Britain, “family planning was widely practiced in Roman Britain”.
The mistaken claim that contraception would have been unavailable and ineffective in antiquity fits a pair of common assumptions about the present and the past: first, that medicine has seen constant progress over time; second, that people in pre-modern societies experienced sex as a less mediated, more natural part of life. Putting these two together, Roman society should have had less medical knowledge, and fewer efforts (or less effective efforts) to control sexuality and fertility. But there is actually quite a lot of evidence to argue that the recent past– the past two centuries in particular– has been a period when medical practices concerned with women’s reproductive health have narrowed and were controlled more tightly than before, eliminating previous options to control reproduction.
So, it isn’t a natural assumption at all that people in Roman Britain in general would have been unable to control fertility. But perhaps women working as prostitutes would have had fewer options. Which brings us to the question: what other evidence is there for the identification of this site as a brothel?
There is actually quite a lot of archaeological and documentary evidence about prostitution in the Roman empire. Thomas McGinn, in his 2004 book The Economy of Prostitution in the Roman World, provides a useful model, surveying the Roman world (Chapter Eight) and providing an appendix attempting to identify more and less formal sex work establishments at Pompeii. These include specific kinds of graffiti (including prices for specific sex acts), specific kinds of architectural layouts, and patterns of quantities and kinds of food and drink that would have been served to clients. McGinn concludes that brothels “were located with an eye to maximizing profit” (p.4).
So is there anything that makes the British Roman villa obviously a brothel site?
Well, actually, no. The record for the site at English Heritage describes it as
A Romano-British homestead built before the mid-1st century and occupied until the end of the 4th, comprising four buildings with an enclosure wall. The principal dwelling house, 92 x 82 ft, was of the double corridor type; the large number of furnaces found suggest that the establishment was engaged in corn production on a large scale.
“Corn production on a large scale”. More than a dozen corn ovens were part of the complex.
Not a brothel: a commercial farm. And, presumably, one that drew on a lot of labor, including a lot of women’s labor.
Hambledon apparently is not (just) a villa, but more of a village; the buildings that yielded the data being interpreted as a brothel
are part of an extensive complex of buildings and fields arranged alongside a paved road. It seems likely that this was more than a villa complex: traces of at least 21 buildings have been recorded, all with stone foundations.
Malcolm Todd, in his Companion, actually comments on the Hambledon site, writing that it is “now interpreted as the discovery of an official infant cemetery in a rural community, where burial in family groups was not the tradition”.
There are ways that an interpretation as a brothel might be supported. But the fact that there were a large number of newborns who might have been deliberately killed would only be support for this interpretation if we knew, or thought we knew, that sex workers in Roman Britain were especially apt to experience unwanted pregnancy, and to have no option but infanticide.
This is a presentist assumption here, and it tells us more about contemporary stereotypes than about the reality of sex work in Rome or in Roman Britain.
There have been discoveries of newborns informally disposed of near historically known brothels from the nineteenth century in the US. The 2005 article “Babies in the Privy: Prostitution, Infanticide, and Abortion in New York City’s Five Points District” by Thomas A. Crist, in the special issue of Historical Archaeology “Sin City” (vol. 39, no. 1) is a classic instance.
But here’s the difference: Crist is exceptionally careful to think about the circumstances that would have made this happen. (And the number of cases is quite small, by the way– probably two events, one a miscarriage).
Crist writes, “the discovery of the skeletal remains provides an opportunity to trace changes in American social and legal attitudes regarding infanticide, abortion, and prostitution and explore the difficult choices faced by workingwomen in New York City from the colonial period to the middle of the 19th century.”
Difficult choices. And choices, specifically, taking place during a period when contraceptive options were being redefined and limited in the United States.
While it is pretty clear from historical texts that early infanticide was not condemned in Roman society, to connect this practice to prostitution without considering the wider context is not just presentist history: it is bad history.
This blog entry follows up on and reprints parts of my original post at Ancient Bodies, Ancient Lives, including issues raised in the comments responding to it.