Research on the Roman marble industry and trade has developed considerably over recent decades, largely thanks to new discoveries in epigraphy, archaeometry and extensive archaeological fieldwork. Since 1964, the German Archaeological Institute has been making a seminal contribution to this area of study through fieldwork at Simitthus (Chemtou) in Tunisia, where Numidian marble, giallo antico—probably the most widely traded colored marble of antiquity—was quarried. The excavations have resulted in the Simitthus publication series, the first volume of which appeared in 1993. The sixth volume by Dennis M. Beck, under review here, focuses on both mining and distribution, and offers an updated, holistic view of the trade in Numidian marble.
Beck’s study exemplifies the German Archaeological Institute’s continuation of research on Simitthus since 2009 and represents the Institute’s efforts to make publications and associated research data freely available online. In this case, 13 maps have been made available in PDF format and 1,221 entries added to the iDAI.objects/Arachne database. This review summarizes the chapters of Beck’s study, evaluates its data-driven approach, and highlights the most promising leads for future research.
In the introduction, Beck outlines earlier research[1] and develops his own research plan: to collate evidence from different disciplines, including epigraphy, archaeometry, and archaeological fieldwork across the western Mediterranean, with the aim of producing a diachronic and holistic survey of the production and trade of Numidian marble. This already makes it evident that an extensive body of data is presented and discussed in the bulk of the volume (Chapter Three, pp. 31–235), followed by thematic summaries in Chapters Four and Five. It is worth noting that the author visited around 40% of the 1,221 recorded sites and items, making first-hand observations of stone varieties that enrich his analyses.
Beck organizes his data into four major time periods: Late Republican, Julio-Claudian to Flavian, Trajanic to AD 161, and AD 161 to Severan. A chapter on a fifth, Late Antique period was part of the original dissertation manuscript but will be published separately. For each of the periods in this study, Beck presents the available evidence of the use of Numidian marble in accordance with a geographical progression, beginning in Simitthus, then expanding further afield into North Africa and the wider Mediterranean basin. Each section is followed by a concise but informative discussion of distribution patterns, quarrying sites, and evidence for how marble quarrying and trade was organized.
Numidian marble was first used in the second half of the 2nd century BC, starting with the victory monument on top of Djebel Bourfifa, the so-called Tempelberg of Chemtou, from which its building stone was quarried. The same variety of Numidian marble was being used in the wider North African region to make votive stelas and one Punic inscription from shortly after 118 BC. The earliest introduction of Numidian marble to the Italian peninsula, around 100 BC, was thanks to elite Roman families, who were probably connected to Numidian royalty or to North African trade. The Casa dei Grifi on the Palatine Hill in Rome or the Casa del Fauno in Pompeii are examples of the Romans’ early penchant for using colored stones as home decoration. The available archaeological evidence therefore generally corroborates Pliny’s statements in the Natural History (36.49) that note the early use of marble in the house of Marcus Lepidus (consul in 78 BC) and its growing popularity across Roman society. Beck also pays close attention to the use of Numidian marble in pavements, including the patterns in which it was arranged and slab size. In this context, Beck’s technical descriptions (p. 39, fig. 13) require familiarity with Federico Guidobaldi’s terminology[2]. From these analyses, Beck reveals that, contrary to common belief, Roman pavers used regular cuts and amorphous chips simultaneously, both randomly and in geometric patterns, thus rendering stylistic dating unreliable.
In the Julio-Claudian and Flavian periods, quarrying activities at Simitthus spread over a much larger area and the stone was distributed more widely. At this time the stone was used for columns and in columnar orders, for the scaenae frons of theaters (pp. 308–9), inscriptions, wall revetments, and opus sectile paving, including new styles such as the quadricromia Neroniana. Statistical analysis suggests that Numidian marble was typically the most widely used colored marble for pavements. The stone was introduced to Gallia Cisalpina, Narbonensis, and the Hispanic provinces. Numidian marble also saw application in sculpture, where it was used to carve a specific repertoire, such as small herms of Dionysian subjects (190 examples on pp. 310–19), small sculptures of kneeling or standing Barbarian prisoners in “Phrygian” attire (10 examples on p. 320), animal sculptures (6 examples, mostly lions, on p. 320), table supports, candelabras, and labras (25 examples on pp. 320–21). The uniformity of these examples at first suggests a centralized production at Simitthus. However, Beck also identifies similar sculptures carved from different imported or local stones, which he uses as evidence of the existence of several workshops active in different places and at different times (pp. 130–33). As he notes, these well-defined sculptural series offer rich potential for future research. Regarding the organization of quarrying and trade, it is usually assumed that Augustus established an imperial monopoly over quarries. However, Beck argues that, if such a monopoly was established, this was done much later. Until that time, marble quarrying and distribution was a multilateral system that involved private, civic, and imperial actors.
In the period AD 96–161, the exploitation of the Simitthus quarries intensified and it became common practice to mark blocks with the formula ex ratione. This formula hints toward the involvement of imperial slaves and freedmen in the procurement of Numidian marble. The establishment of the office of procurator marmorum Numidicorum by AD 138 also would seem to be an indicator of the increase of state influence in marble production[3]. This period also witnessed the construction in stone of a “workers’ camp,” presumably to house forced laborers and a military garrison. The details of earlier excavations of this camp were published in the second and third volumes of the Simitthus series, but Beck presents new geomagnetic survey data on the site (pp. 265–70). Between AD 96 and 161 in the western Mediterranean, the quantity of Numidian marble used and the size of these blocks reached their apogee. The stone finally reached Raetia, Upper and Lower Germania, and all Gallic provinces. Numidian marble nevertheless remained rare in the eastern part of the Empire[4], where many other varieties of colored marbles were available.
In the later 2nd and the 3rd centuries, Numidian marble was still transported over large distances but the trade decreased, with quarrying ceasing entirely after AD 235. The partly collapsed workers’ camp came to house a workshop where small marble objects, such as bowls and mortars, were crafted (pp. 226–30).
In Chapter Four, Beck addresses the questions of the chaîne opératoire from quarry to end user, of state involvement in marble quarrying, and of distribution in nine sections, arranged in a loose chronological order. Finally, Chapter Five covers the aesthetic judgments involved in marble use. Beck clearly states that Numidian marble was mainly valued for its vibrant golden hue, but not always or necessarily a symbol of imperial power, as has been recently assumed. A synthesis in English and German summarizes and concludes the main text. The study is accompanied by a series of appendices that includes Beck’s textual sources, a list of all recorded fragments of Numidian marble from the Forum of Augustus in Rome, a survey of the stone’s use in theaters and in sculpting, and tables on the epigraphical evidence for quarry administration, the military presence at Simitthus, and marble prices. Additional lists summarize the chronological development of the chaîne opératoire. Beck dedicates a final section to reports on the archaeometric analysis conducted at Simitthus by Philipp Hoelzmann and Daniel Steiniger. His extensive bibliography spans 64 pages. Maps, concordances, indexes, and color plates complete the volume, which weighs a hefty 2.16 kg.
The volume’s inclusion of complementary research data is itself a laudable contribution to digital scholarship. Readers can download the data, for example in CSV format, then manipulate them for their own purposes and use them as the basis of new research. For instance, anyone seeking the color variant of Numidian marble used in a specific votive stele, inscription, or column—an important detail in deducing quarry attribution and dating—can consult the digital catalogue.
However, some details of the presentation of research data invite criticism. These critiques relate more to platform design than to the author. First, the data are currently partly limited by the absence of a controlled vocabulary, for example in the description of stone varieties. Such features would enable greater data extrapolation and interoperability. Second, object references in the book rely solely on database IDs, formatted as ID-###. In the electronic version, the reader can use hyperlinks to access the database. In the print book, readers must consult the concordance (pp. 426–34) to find the corresponding Arachne ID and then manually enter a 7-digit code on the online catalog website (or scroll down a long list to find the ID). There, the researcher can alternatively “zoom in” on the desired location on a “heat map”. However, the study does not provide bibliographical references or CIL numbers, or links to identifiers in other databases (e.g., there is no indication that ID-68 = CIL 01, 1904 = EDR 114234). One example: Beck’s volume refers to pavements from Calvatone (Bedriacum) using their catalogue entries ID-279 and ID-280, and only a well-informed reader would be able to link them to Fabrizio Slavazzi’s 1998 publication, because Slavazzi is not referenced in the digital catalogue entries or in the text itself, although he is included in the volume’s bibliography alongside other previous literature on the catalogued objects.[5] Fortunately, the digital format permits the addition of such crucial references at a later stage. Third, the database would benefit from merging the study’s new catalogue entries with existing ones, which they duplicate (e.g., ID 7364441 = 1104933)—a frequent issue in Arachne. Fourth, future researchers wishing to update the distribution maps would be unable to do so in the print or PDF versions. The coordinates included in the digital catalogue do not always correspond to those used in the maps. For example, Map 5a on Numidian marble in Herculaneum cannot easily be redrawn because all of the database entries for Herculaneum (ID-335–419) share a generic, imprecise location. In short, the study aptly demonstrates the considerable effort involved in making complementary research data digitally available and the challenges in adapting it for the end user.
Arguably the most groundbreaking result of Beck’s study is based on his attention to marble varieties, as the Simitthus quarries can be categorized into smaller areas, based on the differently colored and brecciated varieties of Numidian marble found there. By making direct observations of the quarries and using dated usage contexts, Beck has been able to roughly establish when each variety was extracted (pp. 64, 133, 140, 182, 220, 292). Furthermore, archaeometric analysis proves that the presence of trace elements such as iron(III) oxide and strontium can also help to distinguish these varieties (see fig. C4 on p. 346). Beck sketches out this method at several points in the volume (pp. 4, 12, 26–29, 335–48), but it has not yet seen broader sampling. In this respect, Beck’s study is best considered a proof of concept. In future, high-resolution geological surveys of the quarries, combined with comprehensive sampling, will enable researchers to determine the quarry from which artefacts originated and hence their approximate dates.
Dennis Beck’s book is a major new contribution to scholarship on the Roman marble trade, offering a pioneering focus on one specific stone and how its distribution patterns evolved over time. The observations in the book are informed by extensive fieldwork and investigation, and another strength of the book lies in its engagement with fields of research that still lack evidence, such as trade organization. In contrast, the verdict on the distribution patterns of Numidian marble now appears to be set in stone.
Notes
[1] See also: S. Ardeleanu, Numidia Romana? Die Auswirkungen der römischen Präsenz in Numidien (2. Jh. v. Chr. – 1. Jh. n. Chr.) (Wiesbaden 2021) 335–357. Ardeleanu’s synthesis was partly based on an earlier database to which different researchers contributed and from which Beck’s database evolved.
[2] F. Guidobaldi, «Pavimenti in opus sectile di Roma e dell’area romana. Proposte per una classificazione e criteri di datazione,» in: P. Pensabene (ed.), Marmi antichi. Problemi d’impiego, di restauro e d’identificazione (Rome 1985) 171–233.
[3] For the procuratores see now also: D. M. Beck, «Die procuratores marmorum Numidicorum als kaiserliche Funktionäre und städtische Euergeten,» RM 129, 2023, 236–265
[4] Note that the columns in the ‘marble hall’ at Sardis have been erected in February 212 and not before AD 161 as stated on p. 175. For the correct date see F. K. Yegül, The Bath-Gymnasium Complex at Sardis (Cambridge 1986) 170.
[5] “Pavimenti di triclini a Bedriacum (Cremona),” AISCOM 5, 1998, 1–10.