Due to its rich variety of forms and decorations, African Red Slip Ware (ARS) is an important type of find, especially with regard to resolving chronological problems. It is also ideally suited for analyzing questions concerning trade and distribution patterns, since it was very widespread, particularly in the western Mediterranean region. In this context, determining the origin of a certain vessel from one of the numerous, only partly known, and to an even lesser extent comprehensively analyzed potteries plays a central role. Identifying the place of origin of this pottery type was the focus of the master's thesis completed by C. Hasenzagl (H.) at the University of Vienna in 2015 on which this 2019 volume is based. In it, H. analyzes ARS from surveys conducted by the Dutch archaeologist J. W. Salomonson in the 1960s and 1970s.

In the introduction, H. describes the aims and methodology of her work, which centers on standardized fabric description (1). Ultimately, by identifying fabrics that are characteristic of certain workshops, she expects it will be possible to draw conclusions about their production and production centers, as well as the distribution structures. However, while these questions are formulated in the introduction (and conclusion), they are not further addressed in the text. There, the typo-chronological classification of the range of forms from the surveys and the definition of the fabrics are the focus.Footnote ¹

Subsequently, H. presents in detail the history and state of research on ARS (3–14), starting with the earliest attention given to this ware and investigations into production sites in the 19th c., through the first more detailed analyses by F. O. Waagé and N. Lamboglia, the investigations of J. W. Salomonson, and the work of A. Carandini et al. at Ostia, to the seminal monograph Late Roman Pottery by J. W. Hayes, published in 1972. In particular, the Italian investigations at Ostia and the international excavations at Carthage led to a refinement and increased specificity in Hayes's typological and chronological considerations in the 1980s. During the same period, large-scale surveys were carried out in both central and northern Tunisia; these are only partially analyzed, yet they led to the discovery of a large number of both smaller and larger ARS workshops. Excavations were also carried out in some of the northern Tunisian potteries.

Hayes and others subdivided ARS into different vessel types on the basis of not only the nature of the clay and the slip, but also a combination of form and decoration; it was then further subdivided into different fabrics (or “productions”, as H. calls them), some of which indicate a regional origin. The initial rough classification into the fabrics A, C, and D became increasingly differentiated over the course of time. In her work, which includes finds from surveys in northern Tunisia, H. presents and discusses only the northern Tunisian fabrics A, D, and F in detail (7–13). She points out that the initial aim of subdividing ARS was to arrange the vessel types chronologically, but not to determine their origins. She then briefly discusses previous thoughts on the organization and structure of the individual workshops (13–14).

In the next chapter (“Methodology of fabric classification,” 15–16), H. presents the methodology behind the standardized fabric description. This is limited to the evaluation of the clay, its color, hardness, and inclusions; particular attention is paid to a uniform description of each fragment examined. In contrast, the surface treatment – in the case of the ARS, the slip – plays no role. H. also points out the difficulties of this method; for example, that the clay of the various parts of a single vessel (e.g., rim, body, base) can be different (16).

Between 1960 and 1972, Salomonson conducted surveys on 15 Tunisian and three Algerian sites, some of which (e.g., Sidi AïchFootnote ²) have already been published, as H. explains (17–22). Primarily, the sites described are places where pottery, and ARS in particular, was produced, but some of these locations are within towns. As far as can be seen from the available documents, the surveys took place over only a short period of time, and their exact locations, as well as the way they were carried out, usually remain unclear. H. emphasizes the importance of the finds collected during the surveys, since most are fragments from workshops. In her opinion, this means that “the prospected sites are not only the find spots but also the places of origin” (20).

The subsequent chapter (“The North Tunisian tableware from the Salomonson survey,” 23–65) represents the core of H.'s work, presenting the sites under discussion: Bordj el Djerbi, Henchir el Biar, Oudna, and Pheradi Maius. She begins with an overview of the history of research into each production center, followed by a presentation of the ARS fragments collected by Salomonson at the individual sites. The number of sherds varies from site to site, ranging from 65 fragments at Henchir el Biar, to 70 at Oudna, 118 at Bordj el Djerbi, and 222 at Pheradi Maius. A generous selection of vessels is illustrated in drawings and photographs; these also include stamped floor fragments and lamps. Differences between the photographed and the redrawn decoration is sometimes noticeable (see, e.g., 45, fig. 23.3, 5, 6, 8, 9; 113, pl. 6, Bd.083, 086, 087). The documented forms are discussed with reference to the production period of the respective workshops, but no advance in knowledge can be found here compared to earlier work by Mackensen (Bordj el Djerbi, Henchir el Biar), Bonifay (Oudna), and Ben Moussa (Pheradi Maius).Footnote ³

For Bordj el Djerbi, Mackensen identified a production period ranging from 320/330 CE (for H., first third of the 4th c. CE [25]) to the first half or the middle of the 7th c. CE; this information can also be found in the current volume (27). The spectrum of forms presented by Mackensen largely corresponds to that of H. As a typical form he named plates Hayes 61 B nos. 29/30, which were also produced in larger quantities. In the material published by H., however, this form is rarely found: among the 94 classifiable fragments, plates Hayes 67 B and C dominate. With regard to the stamp decoration, Mackensen was able to identify pieces decorated primarily in style A(II)/A(III), whereby the crescent-shaped stamp type Hayes 73 can be regarded as pottery-specific. There is also a base decorated with this stamp among the fragments presented by H. (113, pl. 6, Bd.090). Vessels decorated in style E(I) were also produced, as a sherd with a small dove shows (112, pl. 5, Bd.080), but apparently only in small numbers.Footnote ⁴

According to Mackensen, the workshops in Henchir el Biar – like the potteries of Bordj el Djerbi and El Mahrine – began to produce ARS around 320/330 CE (for H., first third of the 4th c. CE [34]), yet production here ended earlier, in the (first half of?) the 6th c. CE (also H., 35).Footnote ⁵ In contrast to Mackensen's statements, the range of forms presented by H. is very reduced and essentially limited to plates of Hayes Forms 67 and 76 and bowls of Hayes Form 73. The decoration of the vessels is dominated by pieces decorated in stamp style A(II)/A(III); the styles A(III)/E(I) and E(I), on the other hand, are hardly attested. H. presents a fragment in this context (116, pl. 9, Hb.055).

Several production areas have been identified in Oudna, in the center of the city or in the suburban area. Where the finds from Salomonson's surveys originated from is not clear, due to the sparseness of relevant documents. In Oudna, ARS was produced from the last quarter of the 5th to the early or middle 7th c. CE (so, 44). Primarily, this included various bowls such as Hayes Forms 91 B–C, 96, 97, and 99; these seem to have been decorated with the stamp style E(I).Footnote ⁶ The material presented by H. also includes some plates such as Hayes 61 B, 67 C, and 104 A (42, table 5). None of these pieces are misfired, however, and their production in Oudna is not secured by the fabric analysis (96–99, nos. Ou.001–003, Ou.005, and Ou.050–051). In contrast to the premise expressed by H. that the find spots of the ARS are also the places of origin (20), this is probably not the case for these pieces, which should be regarded as domestic items. In the case of an urban site like Oudna, it cannot be ruled out that this also applies to other fragments, only a few of which were identified as misfired (98–100, nos. Ou.039, Ou.045, Ou.058–059, Ou.061, Ou.065, Ou.067, and Ou.070).

This problem also exists with the numerous ARS fragments presented by H. from Pheradi Maius. According to the fabric analysis, however, the vast majority of the vessels (exceptions: nos. Ph.001, Ph.003–004 [100]) come from local potteries. This is also indicated by the spectrum of forms from the survey's finds, which are dominated by plates and platters Hayes 61 B, 87 A, 87 A/88, 88, and 103, shapes that can be regarded as characteristic of Pheradi Maius.Footnote ⁷ Ben Moussa suggested a date of production of ARS there ranging from ca. 320 CE to the 6th c. CE.Footnote ⁸

After the typo-chronological presentation of the finds, H. presents the fabrics that have been identified for each site, each illustrated with four photographs at varying magnifications. In Bordj el Djerbi and Henchir el Biar, three fabrics can be distinguished, in Oudna, two, and in Pheradi Maius, four. These fabrics differ among themselves mainly with regard to the texture of the clay, and less so in their inclusions. In addition, there are three fabrics designated IG(nota), which is used for vessels that did not appear to have been made on site.

Finally, another site studied by Salomonson, called Sidi Rherib, is presented. Its location remains unclear, but ARS was apparently produced there. However, the number of fragments (12) documented is small.

In her conclusion, H. summarizes the results of her study (67–68). She again points out the differences and similarities between the various fabrics, and she addresses the questions and goals formulated at the beginning of the book. These, however, remain largely unanswered, with the exception of the origin of the ARS. Apparently, these questions are addressed in her PhD.

Two appendices follow the text, with a brief description of the forms treated in the work (77–81) and a clear overview of the individual fabrics and their characteristics (82–86). The book ends with a tabular catalog (92–107) and 33 tables.

Comparable fabric analyses of ARS were carried out almost 40 years ago by D. P. S. Peacock, who came to the conclusion that such investigations are not very useful, especially for a fine ware pottery like ARS: “It is clear that the petrological method has been taken to its limits in this geologically uniform area. What is now needed is a chemical study to add refinement and precision.”Footnote ⁹ Peacock's results and also difficulties in assigning individual fragments to the “old” Italian fabrics A–E (or F and G), and in particular to individual workshops, led M. Mackensen and G. Schneider to chemically examine selected ARS sherds, which generally came from the workshops themselves, by means of wavelength dispersive X-ray fluorescence (WD-XRF). This resulted in characteristic reference groups for the individual production centers of El Mahrine, Sidi Marzouk Tounsi, and Henchir el Guellal near Djilma, and for the places examined by H. (Bordj el Djerbi, Henchir el Biar, and Oudna).Footnote ¹⁰ It would have been useful also to chemically analyze at least a small selection of the products identified by means of the standardized fabric description, despite the costs involved. This would have made it possible to verify or confirm H.'s own results, and such analyses might also have provided clues to the interpretation of the fabrics: are these due to different raw materials or different clay preparation and processing techniques (68)? However, the fabrics differ mainly in the size of the inclusions and their coarseness or fineness, which are probably related to each other, and not in the type of inclusions. Since there are no indications of the use of different clay deposits in the chemical analyses of samples from the individual sites so far, with the possible exception of Oudna (44–46), it seems obvious to attribute the fabrics analyzed by H. to differences in the preparation of the clay. Whether these differences can be traced back to individual potters or workshops of a pottery center (at least this is what H. assumes [68]) may be doubtful, especially since there is no obvious connection between fabrics and forms (31, 48, 63, 68). Rather, in the case of a non-industrial, artisanal production such as that of ancient pottery, a certain variance in processing and manufacture must be assumed, which, in the case of the fabrics, probably had no relevance for the consumers. For them, the form, the decoration and the quality of a vessel, visible from the outside (gloss and smoothness of the slip, uniform color, etc.), were probably of greatest importance.

In general, H. seems to overestimate the possibilities offered by fabric analysis for the study of ARS. That conclusions can be drawn about the trade patterns of a production center by determining the exact origin of a vessel (“where ARS was made” [1; 68]; cf. thus 15) is beyond question. However, it is not clear to me how information on the production process (“how … ARS was made” [1; 68]; cf. thus 15) and possibly evidence for different workshops in a production center could be obtained. This would essentially require geophysical surveys of the workshops and excavations (so also 13). In addition, the question of “for whom it was produced” (1; 68) will probably not be solved by means of fabric analyses. In any case, one may be curious to see how the methodology of standardized fabric description presented by H. proves itself in archaeological practice in the field, and how the potentials and questions of fabric analyses announced in her master's thesis are treated and exploited in her PhD thesis.