Cosa has played a fundamental role in the American archaeology of Roman Italy and the list of participants in its excavation and study over the years includes some of that subject’s most distinguished names. Its publication has proceeded on what might appear a cyclical basis, beginning with Frank Brown’s study of the history and topography of the site in 1951. We now seem to be in the earlier part of a new cycle, and one therefore gives a warm welcome to this study of the lamps.
This is also the first of the MAAR series (and the first of the Cosa publications) to be published by the University of Michigan Press, and the result is a handsome product.
The work begins with an introduction (pp. 3-8) on the history of lamps in the ancient world in general and at Cosa in particular. The unreferenced statement (p. 5) that lamps were reintroduced to Greece in the seventh century will eventually need to be modified: for example, one has been found in the settlement at Zagora, Andros, in an eighth-century context—even if it still lacks proper publication.
Pages 9-17 provide a very useful chronology of the site, outlining the major phases of the town’s existence. The authors then fit the lamp-types into this scheme. This section has the great asset of providing a good basic summary for those who do not already know the site. It also includes a handy double-page chart giving the usage of the various lamp-types through time. There is a brief survey of the dated contexts in which lamps were found, though for the most part there is no immediate cross-reference to the actual lamps involved but instead to the original publications of the deposits. Indeed there is little involvement of stratigraphic evidence in the evaluation of the lamps either here or in the catalogue. There is a brief summary (pp. 14-15) of the lamps found in a deposit apparently sealed by the collapse of the Basilica, ca. ad 40-45. This is a potentially important context, situated as it is a full generation before the destruction of the Vesuvian cities. One notices, however, that on p. 196 (under the discussion of the chronology of Factory Lamps), it is admitted that the context is not as secure as one might have hoped, whether through later intrusion or (that bane of so many excavations) mis-recording. For the most part, contexts regarded as significant are mentioned in an introductory section to each lamp-type.
The Catalogue takes up the remainder of the volume. The recording of the objects, their dimensions and state of preservation is seemingly meticulous. Inevitably perhaps, given the apparent lack of contextual evidence for most of the lamps, a great deal of the classificatory work rests on the catalogues of other major collections. Thus this may be considered a secondary rather than a primary source—as if it were a publication of another museum collection rather than an excavation report.
A more serious worry in an excavation report is the relatively small number of drawings and photographs. So far as one can tell without the original objects to hand, the drawings seem good and finely done—although, especially in the earlier part of the book, they tend to normalise the fragments or to be composites, pulling together a number of fragments into an ideal’ form. The validity of such a procedure is open to question. There is some emphasis on rim types, drawings of which are often collected at the beginning of a chapter. Many of the drawings are presented at full size, yet one would rather have had two or three times as many at 1:2 or 1:3.
Only 79 of the 1094 items catalogued are reproduced in photographs. That is we have to rely to an inordinate degree on the authors’ verbal descriptions and classifications. (Drawings, too, inevitably involve interpretation.) For example, among the wheel-made lamps and in particular what is called the Truncated-Cone Type Lamp, we are told that a “large number” were identified (whatever that means); of them 45 are catalogued and given verbal descriptions, and the whole lot are represented by a single drawing, Fig. 7, which is of cat. no. 32. They are given a date-range of mid-third century to later first century. Of what are called Doughnut Type Lamps, we are told that “more than 100” were identified, and of these 86 are catalogued and one illustrated, Fig. 8 (no. 83). Decorated lamps do rather better: of the 51 ‘Warzenlampen’ (called rather awkwardly Raised-Dots Lamps), there are eight drawings.
The sense of a museum catalogue also affects the discussion. Most of the lamps at Cosa seem to have been imported to the site, with the exception of a number belonging to the so-called Esquiline Type which were probably made in the workshop that produced the Sestius amphorae. From the first century of the city’s life, it would have been worth comparing the earlier wheel-made lamps with those of South Italy, to which they have a generic similarity. Interconnections at this period are especially fascinating and deserve more comprehensive study. Taranto fell to the Romans in 272 bc, the year following Cosa’s foundation. Already, in what we suppose on normal chronology to be the later part of the fourth century, there are demonstrable influences from Apulian vase-painting on Etruscan red-figure. In the early part of the third, a Tarentine painter of Gnathia vases, the Volcani Painter, settled in Rome, setting up a shop which not only produced the famous pocola with their inscriptions to Roman divinities, but a style of pottery which had major influence in the southern regions of Etruria, not least Cerveteri. Somewhat later, in the later second and earlier first centuries bc, there are hints of interconnections with Delphiniform lamps. While present at Cosa, they seem to have had their base in Sicily, and the type is also found in Delos. These sorts of connections are typical of that period and are seen most obviously, perhaps, with the so-called Magenta series of plastic vases, but one thinks too of the important work on the distribution of black-glaze pottery carried out by J.-P. Morel. It is difficult not to feel that this was an opportunity missed.
The definition of the chronology of Birds’-Heads Lamps (better known as Vogelkopflampen) on p. 79 is important: they appear to predate the destruction of 70 bc, though they continued in use throughout first century ad and possibly on into second. There is useful information too (p. 137) on the period of popularity of lamps with so-called heat-shields. There is a full discussion of Factory Lamps (194 ff) though in fact fewer are catalogued (thirty-six) than one might have expected, and seemingly none from the FORTIS factory (although this last is given a considerable amount of space in the discussion). There is, however, a FORTIS pinecone lamp (no. 1046).
One senses that this publication represents a labour of love, and as such it brings advantages and disadvantages, moving as it does from the specialist to the curiously generalist. The initial introduction on the history of lamps in the ancient world (not all of it well referenced) is very broadly based. The introduction to techniques of manufacture is also in this vein and it adds little or nothing to other such introductions to be found elsewhere. Equally, there is a rather banal Glossary at pp. 252-5, yet an Index seems not to have been thought necessary. Inasmuch as this is a book about material from a single site, it is a pity that there was no opportunity to pursue a physical analysis of the clays since it could have proved useful on questions of copies and imitations. At the same time one would readily agree that physical analysis provides no magic answers, and it would in fact have added a major and perhaps time-consuming dimension to the publication.
It is curious to make reference to unpublished limestone moulds from Caesarea Maritima in the context of lamps from Cosa. Why not refer to published examples? There are eccentricities in the terms applied to some of the lamp-types, especially among the wheel-made versions, as if excavation nicknames had been translated into the publication. These can slowly be coped with. What is slightly less easy is the use of terms such as one finds on p. 79: Gravisca Type 2, or Sartorio Type 1. They are not explained, even in the Introduction, and not easily found through the Bibliography. For the former the reader needs to be experienced enough to look under H (R. Hannoun, “Lampes de Graviscae”, MEFRA 82, 1970, 237-262), and for the latter under P (G. Pisani Sartorio, “‘Vogelkopflampen’ e lucerne ‘da spedizione'”, RendPontAcc 42, 1969-70, 81-93). As a detail, is no. 986 (shown only in a drawing) really a barbarian, as stated, or is he in fact a papposilenos?
There are no Conclusions, and, although it may be argued that others can make them from the evidence provided, it is regrettable. No one else in our time is likely to acquire a similar knowledge of this body of material, and, given the manner of publication, we are so very reliant on the authors’ verbal descriptions and classifications. Among many reasons the material is important is the fact that it comes from a settlement site, not from tombs like most of the material in museum collections. The excavation therefore provided an important opportunity to develop discussion of the use of lamps in domestic and public contexts at the various periods during the life of an Italian site from foundation (273 bc) to its final extinction in the early fifth century ad. (See now E. Fentress, “Cosa in the Empire: The Unmaking of a Roman Town,”JRA 7, 1994, 208-222.) One would also have liked to see more on issues of trade in a colony and its patterns through time, or on the selection of decorative motifs in a settlement context. As to this last, themes of popular entertainment seem relatively common, with motifs drawn from the circus and the arena. But on the sample presented in this report, there is hardly anything from the theatre. It is legitimate to ask why. The answer can hardly be because there is no evidence of a theatre-building at Cosa—there is no circus or amphitheatre either. Could it be that lamps decorated with masks were thought primarily appropriate for the grave? And if so, what does that tell us about the function of images of masks in Roman Italy?