This volume is clearly the outcome of a massive piece of research, drawing together between the covers of a single book all that we have learned of stamped decoration on African Red Slip Ware (ARS) since the publication of the seminal work of John Hayes (Reference Hayes1972) 50 years ago. It originates in a doctoral thesis completed in 2018 for La Sapienza University, Rome; but since it makes extensive reference to visits by the author to known production sites in Tunisia in the spring of 2019 and incorporates data from the publication of the K. Wilhelm Collection by Michael Mackensen (Reference Mackensen2019; reviewed in Libyan Studies 51 [2020], 174–76), it must have been extensively revised and extended thereafter.
The extent to which the observations and opinions of John Hayes have held up over the last 50 years is quite remarkable and is frequently paid tribute to in this work. But the list of stamps which he published – and that were compiled for the Atlante I volume (Tortorella Reference Tortorella1981) – has been extended in a number of ways since then, not least with the progressive identification and study of the production sites in Tunisia. The core of the entire work, upon which various essays are based, is therefore composed of a catalogue (149–371, with pls 1–117) of slightly less than 900 types or sub-types of stamped decorative motifs found on ARS between the fourth and sixth centuries AD. Each entry consists of a description of the type; an estimate of the area of production (based on various criteria); the number ascribed to the type in any other typologies in which it has previously been described (typically Hayes Reference Hayes1972, Tortorella Reference Tortorella1981, Mackensen Reference Mackensen1993); the style of stamping with which it is associated (the styles A–E defined by Hayes in 1972 still being considered useful, with additions F and G); other types in association with which it has been recorded; the vessel-shapes upon which it has been found; and then finally a list of all the findspots from which it has been recorded. The distribution lists have been constrained by the very sensible decision to cover the diocese of Italy as fully as possible (including Corsica, Sardinia and Malta), but to exclude other parts of the Mediterranean except where finds in the East were previously reported by Hayes (21 n. 15).
In Chapter 1, ‘I luoghi e i modi della produzione’ (25–74), the list of known production centres is rehearsed. Here, two names inevitably stand out as pre-eminent in the study of this ware: Michael Mackensen and Michel Bonifay. Between them they are responsible for 31 entries in the bibliography (as against a maximum of five for any other author) and their contributions to the field are referred to repeatedly throughout. There is relatively little to add to Mackensen's recent review of production sites (Reference Mackensen2019, 35–77), but in Cardarelli's account each entry is terminated by lists of the styles, stamp-types and vessel forms associated with that source. Cardarelli also notes that three sites (Oudna, Henchir es Srira and Sidi Aïch) have been treated in unpublished PhD theses at the universities of Bordeaux, Tunis and Provence (20 with n. 10) which she has not seen: these would surely have something to add to the picture. The inclusion of Sidi Aïch alongside the other sites – originally treated separately by Hayes (Reference Hayes1972, 300) as being of only local significance – has led her to define the particular style of decoration practised there as style G (52, style F being an addition made previously by Bonifay Reference Bonifay2004, 193). The map of Tunisia in fig. 1 (26) furnishes a useful summary of the styles associated with each of the known production centres.
A brief section on the organisation of the production centres (55–56) emphasises our lack of knowledge, given the fact that most of these are known only from surface finds and not from excavation. The discussion under the same heading by Mackensen (Reference Mackensen2019, 71–77) is more extensive, and considers the use of saggars in the kilns and the pugilla with marks of ownership on them from Sidi Marzouk Tounsi. However, the next section (56–60, ‘Considerazioni tecnico-produttive e strumenti di lavoro’) offers a stimulating new perspective. Cardarelli considers how the punches used for the stamped decoration were made and what they might have been made of, illustrated by practical experiments. These ideas were first aired by the author a couple of years earlier (Cardarelli Reference Cardarelli2020) and certainly merited repetition here. The point she makes is that the various dies used to stamp the leather-hard vessels must necessarily have been produced in two different ways. The first involves a blank die, into which the details of, say, a palm-branch, are incised directly by hand. An impression made from such a (primary) die will show the areas incised or cut away in the die level with the surface of the vessel, while those parts of the die which are proud will appear impressed/intaglio on the vessel. Typically, though not always, these depressions are reproduced in drawings with depressed areas in black, just like the marks made on paper by a rubber stamp today. In these circumstances, the die might be made of any hard material which can be cut or carved (see 78 for suggested dies of metal). Those examples which have been found (55 of them, catalogued on 373–75 and fully illustrated in pls 118–26) are all in clay, and such dies might have been inscribed before firing, or chiselled after firing. The alternative (‘secondary’) technique is more sophisticated and involves a two-stage process. In this case, the craftsman carves or incises into a soft material (here, surely always clay) the precise design which he wishes to appear on the finished product, as a combination of recessed lines or dots. Once this archetype has been hardened (e.g., by firing), a die – which must be of clay – is pressed into it. The die itself needs to become hard, one might assume by firing as typified by the examples which have been found. Cardarelli is concerned at the very small number of dies of either type that has been found and suggests (60) that an unfired clay die could still become hard enough for repeated use, while it would ultimately leave no trace of its decay. This is a valid theoretical point, but virtually incapable of proof. In the context of name-stamps on early Roman terra sigillata, finds of dies are also extremely rare, but are known in clay and in bone and some of the impressions imply that on occasion dies were made from wood or of metal (Reference ComfortOCK, 12–13).
What particularly tantalised me about this discussion, but which is not pursued by the author, is the question of which workshops used either of these techniques and when, for they do indicate different levels of sophistication. In many cases, it is not difficult to judge which kind of die has been used. All of those on fig. 41 (68) are surely examples of impressions from primary dies. To my eye, this is almost essentially true of anything with concentric circles, since these are generally very neat, and have probably been turned in wood or bone on a lathe (the whirligig fringes can easily be cut in the outermost ring afterwards). The diamond composition in fig. 42.2 (69) also shows unequivocal evidence of the primary method, since the white lines on the drawing (those incised in the die) intersect one another at the top. On fig. 44 (72), note the contrast between the two crosses nos 4 and 5. The jewelled crosses are also instructive: compare type CR.3.2 (534, fig. 96.5) with type CR.3.8 (535, fig. 97.2; see also the contrasting dies in pls 125.1 [primary – but the black and white in the drawing of the impression are reversed!] and 122.2 [secondary]). In the first case, the jewels appear raised in the impression, therefore as hollows in the die; in the second, the reverse is true (for photographs exemplifying the same contrast, see 105, fig. 59). Now it is very much easier to drill or punch a small indentation than to create a raised dot by cutting all around it. Therefore, in the first case, hollows have been drilled or impressed into a primary die, while in the second, hollows have been drilled or impressed into an archetype from which a (secondary) die has subsequently been made.
The introduction of an archetype reverses the relationship between design and finished product but facilitates much more subtle detail; as a technological evolution, it may be compared with the development of the red-figure technique as successor to the black-figure technique in Greek painted pottery. All of the big human and animal compositions must have been designed in this way, and many of the small animal figures too (as in 70, fig. 43). Who first thought of using this technique for decorating ARS dishes, and who then copied it or learned of it later?
In ‘Sperimentazione e networks tra centri di produzione’ (63–74), Cardarelli traces potential relationships implied by the use of similar motifs in different production centres. For very simple motifs, any possible significance is treated with due caution, and there is certainly no evidence yet of an individual die being used in more than one place. The weakness of dating evidence mostly also prevents one from deducing in which direction any idea was communicated. But there are several close parallels between motifs used at El Mahrine and at Sidi Khalifa in the north (67 and fig. 41). Cardarelli also discusses the potential migration, not just of ideas but of actual craftsmen, between centres (72–74). This is equally difficult to demonstrate, but she does draw attention to the sixth-century incursions of Moors into central Tunisia as a potential reason for the ending of production at Sidi Marzouk Tounsi and a migration of personnel from there to workshops further north.
Chapter 2, ‘La decorazione a stampo: stili e motivi’ (75–112) lists the various styles defined by Hayes and added to by Bonifay (F) and Cardarelli (G), describing the character of each and then listing the production centres with which it is associated, the stamps attributed to it and the vessel forms on which they are found. The associations of styles with production centres is very usefully summarised in fig. 1 (26) and the styles are well illustrated in figs 45–66 (91–112). This fills in a lacuna in Hayes (Reference Hayes1972, 217–22), where the styles, as opposed to the individual motifs, were quite difficult to visualise for paucity of illustration. In northern Tunisia (the workshops employing styles A and E) the use of secondary dies seems first to appear in style A (iii), around the middle of the fifth century. Styles B and C (Henchir el Guellal and Sidi Marzouk Tounsi in central Tunisia) display particularly neat motifs, perhaps derived from metal (and therefore primary) dies. Style D at Sidi Marzouk Tounsi certainly sees the introduction there of secondary dies around AD 430 (Mackensen Reference Mackensen2019, 179), so perhaps marginally earlier than in the north.
Chapter 3, ‘La circolazione della sigillata africana decorata a stampo nella diocesi italiciana’ (113–45) discusses the recorded distribution of this material, divided into two periods: ca. AD 325–480 and ca. AD 480–600. The second period is distinguished from the first by the appearance of new vessel forms and new varieties of stamped motif (115). There is a very large imbalance in quantities, since 87 percent of assignable items are attributable to Period I and only 13 percent to Period II; and of those assigned to Period I, 97 percent were made in northern Tunisia (116). In Period II the domination of north Tunisian products is maintained, but at what precise level is not stated; naturally, far fewer sites were supplied. These changing patterns are demonstrated graphically in a number of maps (figs 70–74, 76–79). It is argued (129) that the importation of stamp-decorated ARS to Italy increased massively just at the time when Egyptian grain began to be redirected on a large scale to Constantinople; but is not this simply the time when the stamped decoration begins? For my satisfaction, I would like to see these figures compared with the importation of earlier ARS forms that did not bear stamped decoration. On the other hand, the Vandal conquest of Africa (AD 429+) is not reflected in any obvious drop in exports to Italy. At the end of the sequence, fig. 81 (132) shows a map of Italy in which the respective areas of Lombard and Byzantine control at the end of the seventh century are indicated. Here the finds of the very latest products (style E [ii], ca. AD 530–80) are plotted – and they fall exclusively within the areas of Byzantine domination. The chapter concludes (134–38) with a list of the findspots/publications upon which the detailed geographical survey is based.
More than half of the book (149–583, referred to by Tortorella in the ‘Presentazione’ as the second volume) is taken up by the detailed catalogue of the individual motifs and known dies and the supporting illustrations and indices. These are the serious data upon which everything else is based. There are also scattered throughout the book a number of important tables. These are not separately listed anywhere, which would have been useful. They are as follows:
The book is massive, and therefore occasional slips are only to be expected, but they can be disconcerting. On p. 87 Hayes is cited as describing the single central motif often found on dishes in style E (i) as ‘impressed in the manner of a potter's products’ which was extremely puzzling, until I checked that the original text was ‘… in the manner of a potter's stamp’. On p. 119, the caption to fig. 72 identifies circular dots as findspots of style A (ii) but does not identify the symbol used for style B – presumably the asterisk, of which there are just eight on the map. On p. 137, information on Vagnari (site 110) is attributed to ‘Faia, Giuliani 2011’ and a corresponding entry appears in the bibliography; the name of the first author is actually Favia. On pp. 400–409 there is surely an error in the fourth column-heading of table 4, which ought to read simply ‘Pars orientalis’. Entries in this column, when traced back to the catalogue of motifs, appear to represent items from the Eastern Mediterranean, included because they were in Hayes Reference Hayes1972.
The volume, published by Quasar, is presented as the first member of a new publication series, ‘Material Culture through Ancient and Modern Mediterranean’ (MACAM), under the direction of Antonio Ferrandes of La Sapienza University, Rome. It is a handsome and worthy contribution to the field of late Roman pottery studies. It gave pleasure to John Hayes to see it.
