Jane Webster, in her article ‘Less beloved. Roman archaeology, slavery and the failure to compare’, brings into the limelight the necessity for archaeologists who study slavery to engage with the use of comparative methods. A brief historiographical analysis shows that, so far, only a handful of archaeologists are interested in performing a study of comparative slavery and probably even fewer are aware of the methodological complexities of such a task. Webster's article could not have been more timely. Historians are increasingly interested in studying connections and in making comparisons between people, sectors, countries and chronological periods. The emphasis on the study of comparative slavery in the UK is, for instance, demonstrated by the recent foundation of the Institute for the Study of Slavery (ISOS) at the University of Nottingham and its teaching in several other universities (Leicester, Newcastle, Edinburgh and King's College London among them). The development of comparative methodology, though, is still in its infancy. This may be one of the reasons why researchers are discouraged from undertaking such a task. Similarly, as Webster rightly notes, scholars tend to complain that the material is scanty and inadequate for the comparative reconstruction of past civilizations. Prominent among these is Walter Scheidel, who altogether, despite his comparative studies on slavery, remains unconvinced about the usefulness of comparisons in archaeology (Scheidel 2003, 581).
On the other hand, students of ancient history have acknowledged the validity of diachronic comparative methods for decades. Jane Webster noted their contribution and acknowledged that this movement was not followed up by the majority of scholars. Specifically, in the 1970s and early 1980s we observed the attempts of Finley (1980) and Hopkins (1978) to find analogies between ancient and modern slavery. Similarly, Bradley (1994), Scheidel (2008), Harris (1999), Bodel (2005) and others preferred to make use of selected comparisons that illustrate the continuity or discontinuity of the ancient slave systems. Their efforts, though, remained restricted to ‘soft’ comparative points, laying special emphasis on the similarities rather than differences between the ancient world and early modern or modern societies. In addition, we should note the comparative studies of ancient and modern slavery by Orlando Patterson (1982; 1991) and Joe Miller (2002), both of whom attempted to demonstrate the continuity and change of slavery across the centuries and throughout the globe. While Patterson's emblematic study adopts comparative methodologies to analyse his global data, Miller prefers to employ cross-historical methods in order to analyse the ‘slaving’ process. Both undertakings are remarkable but they pose a series of problems for historians who do not have the institutional and other resources to undertake this task. Besides, no matter how impressive the breadth of knowledge of these books may be, the fact remains that they do not manage to cover each case in substantial depth.
The reluctance or inability of the majority of modern scholars to embark upon comparative slavery studies is based on the existence of a series of methodological problems. Webster mentions a series of the most important ones, such as, first, the racist element that is central in the modern world but does not exist in the classical Mediterranean; second, the insistence of scholars on comparing only ‘slave societies’ (in the strict Finleyan sense); third, the abundance of documents from the modern world and the lack of them from the ancient world; and fourth, the higher visibility of slaves in the American plantations. The ways that Webster proposes to overcome these difficulties are valid and should be considered seriously. However, in this comment I would like to note a few other problems and offer, in some cases, different solutions that might eventually persuade both archaeologists and historians to undertake the comparative reconstruction of antiquity.
Webster notes in her article that Enrico Dal Lago and I, who for the past few years have dedicated most of our time to the study of slavery in the Roman world and the antebellum American South, would not have approved of the use of ‘soft’ comparisons. Technically, this is not accurate, since we have already experimented with various different methodologies that could facilitate the comparative analysis of our diachronic data. The publication of our two edited volumes on ancient and modern slavery – Slave systems. Ancient and modern (Dal Lago and Katsari 2008c) and From captivity to freedom. Themes in ancient and modern slavery (Katsari and Dal Lago 2008) – exemplify the variety of the methods used by scholars in the Americas and in Europe and our indubitable approval. In our first book on Slave systems we decided to include articles that are explicitly comparative of the Atlantic world and the ancient Mediterranean, hoping that this decision would ensure the homogeneity of the adopted methodologies. The outcome, although admirable, persuaded us that there are as many methodological perspectives as scholars on the face of the earth. Some preferred to follow the methods employed in cross-history, others focused on the influence of the classical tradition on early modern societies, while yet others chose to implement the ‘contrast-of-context’ method. Despite the differences in approaching the material, the results remain remarkable, as each paper clarifies otherwise obscure aspects of history. In the second volume, From captivity to freedom, we opted for the obvious choice to juxtapose papers on several themes relevant to ancient and modern slavery. Two papers in each chapter, one on the ancient world and the other on the modern, dealt with the same or a similar topic. The point of such an exercise was to allow the readers to compare for themselves the two case studies, note similarities and differences and maybe be persuaded to embark in the future on the comparative analysis of the available data. Similar juxtaposed studies are common. The several edited volumes published in the past few years on specific topics on slavery include papers that range from the ancient Mediterranean to the end of slavery in the Americas (for example Kleijwegt 2006). The difference is that most of them are interested in the process of slavery, rather than in the direct comparison of two case studies. The experience of editing the above volumes and the mistakes we made in the process taught us a few value valuable lessons in addition to the ones presented by Webster that I would like to share with readers, especially if they are interested in the study of slavery in the Roman Mediterranean and the 19th-century USA.
First of all, even if I accept the value of employing different comparative methods – ‘rigorous’, ‘soft’, cross-history, classical tradition and juxtaposed studies – my preferred methodological approach remains the ‘contrast of contexts’ as it was described by Skocpol and Somers (1980). One of the modern historians who successfully employed this method in the field of slavery is Peter Kolchin (1987). His comparison of two unfree labour systems, slavery in the antebellum American South and serfdom in Russia, created new perspectives of both societies. Kolchin, in effect, gave modern historians the methodological tools for synchronic comparative studies, through the analysis of his examples. The emphasis on only two case studies allows the historian to study in depth (as far as this is possible), and give equal emphasis to, the two sides of the equation. Once the study of the individual cases is complete, the historian ought to indicate the existence of both similarities and differences. Such an exercise clarifies the reasons for the direction each society took in the solution of similar or dissimilar problems. In his latest book, A sphinx on the American land (2003), Kolchin explains in detail his methods for comparing the antebellum American South with other societies. Even if he does not discount the value of other comparative methods, his comparative study of two cases on an equal basis remains unsurpassed.
Diachronic comparisons, on the other hand, pose additional challenges. For example, according to Webster, the quantification of data in the study of modern slavery does not find its equal in the study of ancient slavery. The material extant in the plantations of the antebellum American South (population numbers and other similar data) is not matched by similar data from the ancient world. The evidence from the Roman period is usually patchy, unreliable and does not allow any meaningful statistical analysis. Even if, in the case of Roman slaves, such an approach seems to leave little room for other options, I cannot offer any alternative solutions. However, the amount of available data from inscriptions concerning freedmen is substantial and may be used in statistics. Funerary and benefactors’ inscriptions in Rome and the provinces are especially useful in the reconstruction of the social mobility of freedmen and their role in the Roman economy (Mouritsen 2005).
If the quantification of data is difficult or even impossible in some cases, then we should lay special emphasis on the qualification of our evidence. Qualification is a popular tool that historians use extensively in the assessment of the ancient material. Despite the acknowledged value of such a procedure, here too are pitfalls for the unwary. For example, ancient historians consistently use a variety of disciplines in order to understand their sparse and sometimes inadequate or biased textual material; hence they use linguistics, philology, archaeology, anthropology, sociology and the sciences in order to achieve the reconstruction of the ancient world. It would, therefore, be useful if modern historians – despite the large volume of their evidence – opened up to other disciplines. A sustained use of sociological theories, philological sources and anthropological perspectives could facilitate the purposes of the comparison.
Another problem relevant to the qualification of data is the attempt of many historians as well as archaeologists to locate analogies and assume that they should be used as comparative points. As Webster suggests, ‘analogy plays a central role in archaeological reasoning’ (p. 104), thus it may easily disorientate the researchers and lead them into false assumptions. Under no circumstance should scholars who study the ancient world presuppose that modern processes reflect ancient ones, if they do not have adequate evidence from antiquity to prove such a claim. Even modern commonsensical thoughts may have been different according to the social, religious or ideological background of peoples who lived centuries before or in different geographical areas. In order to overcome this obstacle a diachronic comparative study should always start with the analysis of the ancient data. This way, both archaeologists and historians will find out early in their analysis the themes on which they should focus. If they try to analyse the modern data first, they may not find comparable material from the ancient side, and hence lose valuable time. As Kolchin suggests, the comparison of the two case studies should be based on the fact that both of them can be analysed in depth and that they both offer comparable material.
One of the problems that Webster acknowledges is the visibility of slaves in the archaeological record. Specifically, she claims that ‘Graeco-Roman slaves are frequently assumed to be “archaeologically invisible”, leaving no clear material footprint, other than artefacts of restraint or ownership, for excavators to identify’ (p. 115). This assumption has influenced both archaeologists and historians who study ancient slavery and inhibited them from undertaking such studies. Only recently have archaeologists tried to distinguish a slave culture that is divergent from that of the masters and to recognize their material distinctiveness, Carandini and Morris being among them. Webster gives several examples of areas where slaves lived and worked: the mines in Laurion, the kitchen, the cellae and the ergastula of the villa rustica and the kitchens of urban households. I would agree that it is possible to locate spaces that slaves would have frequented in antiquity. However, how can we distinguish them from other unfree labourers? For example, even though mostly slaves worked in the Laurion mines of classical Athens, during the Roman Empire both free workers and slaves worked in the provincial mines (Domergue 1990). Also, in the villae rusticae the employment of other unfree labourers (e.g. tenants or seasonal workers) complemented the use of slaves in the fields (Pliny, Ep. 3.19). The labour system remained flexible throughout the Roman Principate and we may assume that both slaves and other unfree workers laboured side by side and frequented the same areas. The only exceptions would be the exclusive occupation of the ergastula by slaves and their predominance in the urban kitchens. Also, I am not entirely convinced (and neither seems Webster) that the graffiti we encounter on the walls of kitchen corridors and latrines were produced by slaves; freedmen are as likely candidates.
The methodological problems inherent in comparative studies should not inhibit the researcher. Instead, the undertaking is a challenge that gives a chance to both historians and archaeologists to cooperate and explore new issues, clarify otherwise incomprehensible aspects of the past and analyse their material from different angles. The lack of data from the ancient world may be a fact for some themes but simultaneously we may observe the abundance of sources in other instances. For example, in the case of the Roman world and the antebellum American South there is available evidence that could highlight the following historical aspects: slave revolts, masters’ ideology, social mobility of freedmen, sources of slaves, population studies, the classical tradition of slavery, slave management. There is still a lot to accomplish with regard to these topics but I hope that an interdisciplinary comparative approach will give the methodological and theoretical tools to move forward.
