La céramique romaine d’Argos: Fin du IIe siécle avant J.-C.–fin du IVe siécle aprés J.-C.

This is a massive work, and the amount of dedicated scholarship is evident from the more than 500 items listed in the bibliography. It is a conventional pottery catalogue, but as such it puts in the public domain, in a uniform format, a huge amount of data covering the pottery in use at Argos between 146 B.C.E. (the Roman conquest) and 395 C.E. (the partition of the empire). For this reason alone, it will be cited for years to come and will be responsible for dots on innumerable distribution maps. I turn first to the issue of presentation.

Works of this kind are not for the fainthearted. The correct identification of pottery depends on endless minute descriptions and interpretations, which have to be set out before one can move on to wider matters of historical interest, such as the economic history of a region and its trade relations, or the rise and fall of different pottery production centers. It is therefore important to make the process of assimilating the data and following the arguments as easy as possible for the reader, and in this respect I found much in this volume that made life more difficult, rather than easier.

The overall structure of the work is composed of a typology, first of the fine wares, with separate sections for each ware (e.g., “Sigillée orientale A”) or category (e.g., “bols hellénistiques à reliefs”), and then of the coarse wares, divided by shape category only (e.g., “vases à cuire”), followed by an itemized list of 48 dated contexts or groups. There is an introduction (3–5), but it is more akin to a preface, describing briefly how the work came about and listing acknowledgements. Nowhere is there an explanation of the approach taken to the typology and its complicated numbering system or to the choice of dated contexts.

I turned to the contexts (275–322) at an early stage to verify my assumption that they had been chosen either because they had external historical dates or good dating evidence from other finds (coins, lamps) that might help refine the dating of the fine wares, or because they contained useful quantities of coherent material where a date established by the fine wares might help elucidate the chronology of the coarse wares. This proved to be true of some contexts, though there were few for which the dating was not directly dependent on the pottery. There were, however, several contexts for which I could divine no such rationale: in the Hellenistic period, in particular, the list includes a number of tomb groups (Groups 3–13) that comprise as few as two items each and for which a date is sometimes proposed with no explanation at all. Why are these here? They do not seem to contribute anything to the study of the pottery, and if they are included as part of some wider debate concerning the history of Argos, that is not explained.

The consultation and understanding of the dated contexts is hampered by an issue of presentation that bedevils the whole work: the numbering system (or systems). The typology makes use, in the text, of a three-tiered hierarchical system (e.g., “8.2.3”). The first of the three numbers defines a major category (e.g., “Hellenistic relief bowls,” “Eastern Sigillata A,” “coarse ware”). The second is used to subdivide these categories either morphologically (e.g., “plates and platters,” “cups and bowls,” “bases,” “cooking vessels”) or according to some other criterion (e.g., “Argive production,” “Asia Minor production”). The third number then defines (for the most part) individual forms: thus, section 8.2 comprises cups, bowls, and basins in Çandarli Ware, and form 8.2.3 corresponds to form 1 in the typology of Hayes (Late Roman Pottery [London 1972]). These weighty numbers are the primary basis for cross-references in the volume—except when it comes to the illustrations. These are numbered in a separate series from 1 to 466, which corresponds one-to-one with the tripartite form numbers (e.g., form 8.2.3 corresponds to cat. no. 190) but then has its own subset for multiple illustrated examples of the form (e.g., cat. nos. 190.1–3). This means that every time a form is referred to, either in discussion or in the list of contexts, to find the illustration (the basic currency of ceramic studies), one must look up the relevant text to discover the equivalent catalogue number for the illustration. This shortly becomes exasperating. If the tripartite numbers had been suppressed altogether, nothing would have been lost and understanding the material would have been made easier. (Oddly, this alternative approach is applied to citations in the context lists of previously published material and to references in the lists of stamps and applied motifs.)

Neglect of the reader (or failure to understand the reader’s perspective) extends also to the typology. Each entry is prefaced by some catalogue data (inventory number, context, illustration, brief description, previous publication reference—but again, only for the first example, where multiples are illustrated). This is followed by a list of published parallels from other sites, a discussion of the form, and then a list of further instances from Argos. So far, so good: this is just what one would expect. But the list of further instances is ordered by inventory number (an administrative convenience of no consequence for the reader), not by context; so if, for a prolific form (e.g., the 226 instances listed under cat. no. 185), one wishes to gain a coherent idea of its distribution among the dated contexts, it is necessary to copy out this list by hand and rearrange it in order to see the pattern.

The typology incorporates four additional catalogues: stamps (T1–98), applied motifs (A1–39), graffiti (G1–9), and dipinti (D1, 2). The last two are gathered together into “Annexe I” (273–74) and shown in the list of contents, but the others are not; they are broken up into individual sections and listed at the end of each relevant category. The stamps, when first referred to (16, cat. no. 4), are not explained, nor is one told where to look for any further information; the first applied motif (A1) is listed in its own catalogue entry, so the same bewilderment does not apply. It may be useful here to list the pages on which this material is given: A1 (19); A2–29 (47–51); A30 (not found); A31 (108); A32–9 (167); T1–6 (19); T7–47 (52–60); T48–51 (85); T52–80 (108–11); T81, T82 (128); T83, T84 (135); T85–98 (168–70).

The first stamp entries are decidedly shaky. For T1 and T2, the reader is told that both occur on inventory number 81.527.4. There is no other guidance that this is an additional instance of form 2.1.2/catalogue number 4, found in Group 22. Later on, under T29, for instance (56), one is told more helpfully, “Voir 80.12.86 (voir 87),” which shows that it is an additional instance of form 4.3.7/catalogue number 87 and that it was found in Group 23.

The preceding paragraphs have been taken up with matters of presentation, not of content, but for the reasons stated above, I think that it is perilous not to give this sufficient consideration, as poor presentation will discourage the reader from ever discovering the value or interest of the content. Turning now to the latter, the author does not state when the research was carried out or completed. The only clues are the statement in the introduction that only material excavated prior to 1984 was taken into consideration (3) and at the beginning of the bibliography that the citations do not, for the most part, extend beyond 1998 (323). One has the impression, however (e.g., when publications from the 1970s or 1980s are described as “recent”), that much was done some time ago. The ordering and the types defined for Eastern Sigillata A, B, and C (Çandarli Ware) follow the typology proposed by Hayes (“Sigillate orientali,” in Atlante delle forme ceramiche. Vol. 2, Ceramica fine romana nel bacino mediterraneo (tardo ellenismo e primo impero) [Rome 1985] 1–96). However, the lengthy and important section on Italian terra sigillata (24–66), while citing form numbers from the Conspectus (E. Ettlinger et al., Conspectus formarum terrae sigillatae Italico modo confectae [Bonn 1990]), uses the abbreviated title CFTS, contrary to the otherwise universal usage of Consp. recommended in the volume itself. The author makes no concession (in terms of ordering the forms) to that typology, nor does she ever refer to the chronology put forward in that publication. One has the impression that the Conspectus form numbers have simply been tacked onto a piece of work written previously and not accordingly revised.

There are a number of points of detail where I feel uneasy with identifications or lines of argument. That is perhaps inevitable, but one must always be careful of circularity (dating a group by its contents and then using the sequence of groups thus obtained to argue the chronology of the same products). This seemed to me a particular danger in regard to the Late Hellenistic fusiform unguentaria (cat. nos. 412–16). A great deal of weight is placed on these for dating the groups of the first century B.C.E. Catalogue number 413, in local fabric, “apparaît dans des contextes de la seconde moitié du IIe siècle et du début du Ier siècle av. J.-C.,” but the only dated context in the present work to which it is attributed is Group 7, ascribed to the first half of the first century on the basis of this and other similar unguentaria. Likewise, catalogue number 416 is said to occur at Argos in contexts of the second half of the second century—but the groups cited (Groups 5, 8, and 12) are attributed, largely but not exclusively on the basis of the unguentaria, to the first century B.C.E.

Group 18 is a rich burial containing gold, together with some glass unguentaria and a coarse ware flagon; the author states that “les unguentaria ainsi que la cruche datent de la seconde moitié du Ier siècle–début du IIe siècle ap. J.-C.” (288). However, in the catalogue entry for the flagon (cat. no. 386), she states that this is a local product that makes its first appearance at the beginning of the first century C.E. at the latest (in Group 14, a sound group) and is still current at the beginning of the second century. If that is the case, it is difficult to see how it can be used to argue a more precise date for Group 18.

The main text concludes (257–72), appropriately enough, with a series of conclusions. These seek to draw out and compare the evidence of the different classes of pottery described. There is an extraordinary paucity of amphoras (and not a single stamp), explained tentatively in terms of the agricultural self-sufficiency of the Argolid (259). The local coarse wares are identifiable as far afield as Mantinea, Nemea, and Corinth, while Argive coins are also common at Corinth. The author infers that the realities of geography always limited the basic economic zone of Argos. The site participated in wider Mediterranean trade (in pottery) to a more limited extent than Corinth but followed broadly the trends that are observable there, at Delos, and at Knossos in Crete. A certain conservatism with regard to changing trends (supposedly driven by Italian or Roman influence) is tentatively ascribed to a lack of Italian settlers who may have provided a stimulus for foreign imports at Corinth and Delos. African Red Slip Ware is barely present before the appearance of the Central Tunisian C fabric in the second quarter of the third century C.E. and does not become dominant (along with the North Tunisian D fabric) before the fourth century. It is suggested that this is due to the economic decline during this period in western Anatolia, which had previously dominated the market (with Eastern Sigillata B, Çandarli Ware, cooking pans, and amphoras). The story terminates in 395 C.E., which allows only for the introduction of Phocaean Red Slip Ware and not for the later importation of Cypriot and later African products or for the consideration of the later amphoras.

In summary, there is much information here that will add to the resources for studying the pottery trade in the Mediterranean during the Roman period, but it is not easy to extract, and each argument should be evaluated carefully before it is accepted.

Philip Kenrick
7 Abbey Close
Abingdon OX14 3JD
United Kingdom
philip.kenrick@arch.ox.ac.uk