Bryn Mawr Classical Review

Bryn Mawr Classical Review 2000.10.27

David Soren, Noelle Soren, A Roman Villa and a Late-Roman Infant Cemetery: Excavation at Poggio Gramignano, Lugnano in Teverina (Bibliotheca Archaeologica 23).   Rome:  "L'Erma" di Bretschneider, 1999.  Pp. 687 + 310 figs., 269 pls..  ISBN 88-7062-989-9.  

Reviewed by W.V. Harris, History, Columbia University (
Word count: 2017 words

The Roman villa at Lugnano lay in the hills on the left bank of the Tiber, 7 kms. from ancient Ameria and about 70 kms. north of Rome and was therefore in Umbrian territory. It was probably built, according to the excavators, in the 10s B.C. It went into serious decline, they believe, in the years around 200 A.D., and thereafter enjoyed "limited or sporadic use" (p.43) until it was employed in the mid-fifth century as a cemetery for young children. The villa was first excavated, for two seasons, by the Soprintendenza Archeologica, represented by Daniela Monacchi, then for six more seasons by a joint team put together by David Soren (under the aegis of the University of Arizona) and by the Soprintendenza. The report published here is naturally the work of many hands, but chiefly those of Soren, Monacchi, William Aylward, Archer Martin, Carla Piraino and Michael McKinnon.

Ten or eleven rooms were dug, and there were at least four more. This was therefore a substantial property, and the most elaborate room was a quite imposing colonnaded chamber (7.2 m. x 7.1) with a vault and an expensive floor; but there are no signs of great luxury or of culture. The villa may have been built by one of the decurions of Ameria, but it might also have been one of the secondary properties of a man of greater wealth. The edifice is in any case of a fairly common Italian type. It is a pity that this report does not provide a systematic comparison with other similar sites, such as the not-far-distant villas at Alviano Scalo and Pennavecchia (though these are mentioned from time to time). What Monacchi has to say (37-38, 423-424) about the likely social status of the proprietors seems, in the absence of such comparisons, to be a little ambitious.

The basis of the excavators' chronology, and hence of their history of the site, is not as firm as one might like. It may be too much to ask whether the villa was built about 50 B.C. or in the 10s (though the editors must be aware that the exact date of the onset of the villa economy is a matter of the greatest interest). What is disappointing is that the authors do not explain why Soren gives the later date while Martin (p.229) says that the villa was built "about the middle of the first century B.C." Since the site yielded more black-gloss pottery than terra sigillata, Martin's view seems to be an eminently reasonable one.

It is also hard to see what justifies supposing that there was a big break in habitation about 200 A.D. The "terra sigillata chiara italica," which is later than that date, is relatively abundant. The site also yielded a noteworthy amount of African red-slip ware of the third and/or fourth centuries.(p.237) The brick stamps (of which there are 30 in all) mostly date from the years around 200, with the Figlinae Publilianae best represented. This implies a serious restructuring of the villa at that time, as Monacchi indeed remarks (p.382). The supposed breakdown around 200 A.D. thus seems to owe more to the tired old convention that everything went wrong in the Italian economy at that date than to any actual archaeological evidence. The date is probably at least fifty years off.

This is not a review of the excavation but of the report. It must be said, however, that there is every sign of meticulous care in the dig itself and in the work done on the finds. Most categories of excavated material certainly received expert treatment. The excavators must have been disappointed not to have dug the whole site, and the possibly quite extensive section SE of Room 6 arouses curiosity. It seems to be implied (46) that it was not because of an uncooperative landowner that this area was neglected, so one wonders what the reason can have been. Simple lack of funds? At all events, the mission was not fully accomplished.

One should never be ungrateful when the excavators of a worthwhile site get a full report out in a timely fashion. And this report provides a high proportion of what a reader may require (the most serious gap -- the lack of comparisons with other sites of the same type -- has already been pointed out). But there are some editorial issues, and some historical ones too.

No one would want such a report to be a terse summary, I think. But this one weighs 4 kilos, which makes it tiresome to consult, the number of plates is lavish -- and the book apparently costs 1.4 million lire per copy ($640 or so, even with the lire very cheap), and that price does not get you an index. This costliness was not inevitable, but on the contrary resulted from two somewhat questionable editorial decisions. One was to entrust the publication to "L'Erma" di Bretschneider. The publishing branch of this fine old firm honors the god of theft daily by practicing the kind of price-gouging which leaves even such companies as Cambridge University Press, Gieben and De Gruyter gasping in ineffective envy. The pricing formula seems to be: multiply publication costs plus overhead by three, then divide by the number of libraries (mostly American and German, I presume) that can be guaranteed to buy the book at virtually any price. American archaeologists ought to swear an oath to take no more manuscripts to these guys. Many others have complained of "L'Erma"'s greed.

The second editorial error was to print too much information and too many photographs. It is true that one may legitimately want to see exactly how the excavators proceeded, and some excavation reports offer too little. We are glad to know that the exact dimensions of each sounding are on record, but we do not need to be told what they were, certainly not on the printed page; many of these printed pages are full of perfectly useless information. As for the visual record, most of it is of excellent quality (the introductory maps, figs.1 and 2, and the map of page 304 are marked exceptions), but were all those ceramic profiles really necessary? For a few people, yes. And that leads to the obvious conclusion: fully half of the information in this book -- and half is a moderate estimate -- should have been made available on-line only. Electronic publication has many problems to resolve before it can supplant the academic book, but the volume under review is now anachronistic.

A short review of this book has unavoidably to neglect a number of intriguing questions about this site. Economic historians may be most interested by the 309 kg. of amphora fragments, expertly discussed by Archer Martin. They establish that the villa was, as we might have expected of a Tiber valley site, well integrated into the water-borne trade of the Tyrrhenian Sea. There is an interesting contrast, however, between the oil and garum amphorae, some of which come from overseas, and those that contained wine, which are all Italian.1 Anyone who recalls the pages about Umbria in Tchernia's Le Vin de l'Italie Romaine2 will realize that this must have something to do with the strength of the wine-producers in the Tiber basin itself; it may also be another indication of the less-than-plutocratic condition of the Poggio Gramignano proprietors -- no imported vintages for them (not of course that all imported wine was of the first quality). The amphorae also appear to show that in the fifth century A.D. Poggio Gramignano was still part of a long-distance trading economy.

But the most distinctive feature of this site is the infant cemetery of the fifth century A.D., and it receives extended treatment (pp.461-651). No fewer than nine other scholars besides Soren contributed to this part of the report, and it is clear that all or many of them have engaged in a productive debate. Unfortunately, many of the contributors go over the same ground again and again, and a much firmer editorial hand was needed.

The cemetery consists of 47 burials (there were originally at least a few more), which makes it the largest set of child burials yet discovered in Italy. With the exception of one two-to-three-year-old, all those buried here were premature fetuses or very young infants. The excavators argue that they were all buried within a short period, and suggest that they died as a result of a malaria epidemic. What leads them to think that the burials took place within a very short period of time, "perhaps a month," (487) is (1) the lack of compaction in the fill in which the skeletons were discovered and (2) a series of joins between sherds found at relatively widely separated places within the cemetery. It seems a strong case.

A Roman infant cemetery is not in itself a surprising phenomenon, for we know from Pliny and Juvenal (cf.478) that according to the Roman practice of their time infants were not cremated--a custom which can easily have remained in existence in the fifth century. The oddest features of the Poggio Gramignano cemetery are, I think, three (but only the third of them receives a detailed comment in the published report): (1) fully 22 out of the 47 infants were fetal in age, which cannot possibly be the result of "spontaneous" miscarriages; (2) if the burials all belong to a very short period, the population from which they were drawn must have been a substantial one by the standards of the late-antique Italian countryside (to include 50 late-term and new-born infants, a population would have had to number at least 2,000--and remember that Poggio Gramignano is not even in the outskirts of the small town of America); (3) then there are the puppies--the cemetery includes the remains of at least a dozen young canines, some of them partially dismembered before burial.

The age pattern of these dead children inevitably makes one think of abortion and exposure, for a cross-section of the victims of childhood diseases would have contained some slightly older children. The evidence for an epidemic of malaria is entirely indirect, and its presentation here is not helped by the evocation of "the fall of the Roman Empire." The event that killed these children will have had a dramatic effect on a village in Italy in the mid-fifth century; it cannot have debilitated the imperial armies of one or two generations earlier.

It is not my intention to pick apart what this report has to say about malaria, still less to exclude the possibility of demographically important outbreaks of malaria in late-antique Italy, but it should be registered that there is no direct evidence for malaria in Italy at this time. (However, I understand that Walter Scheidel will soon be putting forward the case for the late-antique importance of this disease.)

Finally, the puppies: Soren collects in a very thorough fashion all the evidence about canine burials and sacrifices (including some from the Iguvine Tables, perhaps suggesting Umbrian continuity?). He argues that the connection of dogs with the chthonic goddess Hecate is crucial in this case (there is no sign of Christianity in this cemetery). As it happens, however, there seems to be no Italian parallel for canine burials mixed with human burials. That may be because archaeologists have missed the evidence. It seems likely in any case that what we have here is the result of a religious reaction to an event of quite exceptional severity; that event may have been an outbreak of malaria, but it may have been simply a disastrous harvest--a frequent cause, in early-modern Italy as elsewhere, of abrupt increases in child-exposure.

Whether we should think of these canine sacrifices as magical and as signs of witchcraft is another matter. The question turns largely on the skeleton of a toad (pl. 235) and a single raven's claw (pl.234) found in very close association with the skeletons of infants. But these are isolated elements, whereas the puppies are numerous.

In short, this is an extraordinarily interesting report, notwithstanding some debatable chronologies and debatable editorial decisions.


1.   But what became of the graphs referred to on page 361?
2.   André Tchernia, Le Vin de l'Italie Romaine (Rome, 1986), 254-255.

Read Latest
Index for 2000
Change Greek Display
Books Available for Review

HTML generated at 13:27:03, Friday, 03 April 2009