This third volume of Oxford Studies on the Roman Economy contains a substantial introduction by the editors and nine papers on different aspects and areas of agricultural production in the Roman world in the period c. 100 b.c. to a.d. 350. Kehoe's brief survey of the state's impact on agricultural production focuses on tenancy arrangements, especially the Mancian regulations in North Africa and fourth-century rules about loans to private farmers. His argument for a policy of protecting small farmers is optimistic because, as he also demonstrates, the primary aim of these rules was to protect state revenues. The unexamined reverse of the coin is Rome's peculiarly strong ideology of private ownership, which favoured the growth of large private estates whose élite owners lobbied effectively for increasing exemptions from civic burdens.
The other eight papers all adopt a quantitative approach. Three analyse the results of regional field surveys. De Vos summarizes the findings of a survey project around Thugga in the Medjerda (Bagradas) valley of Tunisia, chosen for its epigraphy of imperial and private estates, including the Mancian and successor ‘laws’, to which the survey has added important new finds. Although the rich data invite other questions, she concentrates on the emergence of large tenant-based estates, which she assumes represent productive investment. Particularly striking is the now precise delineation of the estates of the Statilii and the Passieni, which were subsumed under Nero into a large imperial estate. I suspect, however, that they were imperial properties from the 30s b.c., assigned as gift-estates to Statilius Taurus and the younger Sallust until reclaimed from their descendants, and that the Mancian regulations were attempts to maintain land in production against a chronic tendency to abandonment. Goodchild argues the merits of GIS-modelling of ancient landscapes using the example of the middle Tiber valley. Potentially the ability to map material findspots (‘sites’) onto land-types systematically and in detail is exciting, even if Goodchild modestly calls it no more than enhanced site-catchment analysis. The problem remains our dependence on ancient data: apart from the huge issue of the categorization of ‘sites’, the agronomists are unreliable — e.g. Goodchild cites Varro 1.44.1 and Columella 2.9.1 on sowing ratios, although they are contradictory. Friedman's analysis of survey data from the Wadi Faynan (Jordan, third–fourth century) shares the unhappy reliance on written ‘facts’, but presents a fascinating, if speculative, micro-story of mining stimulating agriculture and then destroying it by pollution.
Another trio of papers investigates the diffusion of technology. Marzano notes that although only 44 of the 338 ‘élite villas’, and another 96 ‘farms’, known in the hinterland of Rome have produced secure evidence of wine- or oil-presses, this should not undermine De Sena's argument from amphora studies for extensive wine and oil production in that area. In her second paper she collates and discusses the data for multiple presses in Gaul, Spain and the Black Sea area, which she finds peaked in the later first to second centuries. Wilson and Malouta study the Roman use of water-lifting machines, based on an analysis of documentary references from first- to seventh-century Egypt set in the context of a wider selective survey of archaeological evidence. This conclusively shows that there was widespread use of various types of water-lifting machines long before the Islamic period (but equally dense or not?), with an apparent peak in the long second century and then again in Byzantine Egypt. The questions which all three papers raise are specificity and significance of the surviving data. Lever-presses, with their unmistakeable material footprint, and especially multiple presses, were arguably linked to a phase of mass production of medium-quality wine and oil, and show investment in that type of agriculture, but their absence may only reflect a switch to less detectable screw-presses, which could be used in series for mass production, and commonly were in Roman Egypt, but were also suited to smaller-scale but high-investment production of quality wines. Wilson and Malouta valiantly try to sort out the terminology of water-lifting machines in the papyri, but the suspicion remains that the ‘machines’ of the Byzantine period are shadufs (simple and small), appropriate to its specific tenant-based system of agriculture.
The remaining two papers treat Roman Egypt. Bowman provides a magisterial conspectus of what has been achieved by attempts to quantify agricultural production in terms of gross areas of cultivable land, size of village territories, distribution of land by crop-types and farm-sizes, and populations and consumption needs of towns and villages. Blouin argues for a tension in the Mendesian nome (Nile Delta) between a market-oriented focus on wheat by the rich and the traditional autarkic diversification of the poor, with points of comparison to the North African situation. Blouin's paper encapsulates the big question of the volume, as fully and frankly laid out in the editors’ Introduction. What can we quantify reliably, and how far does it take us? J. Andreau, L’Économie du monde romain (2010) has sharply criticized modelling the ancient economy by quantified estimates. While Bowman and Wilson doubt the validity of macro-estimates, they urge faith in more specific data-rich studies. This volume exemplifies how much potential ‘thick’ data we have, and how we are beginning to confront seriously and overcome its limitations. What we have yet to work out is how we might meld disparate data from Roman Egypt, let alone Spanish wine-press data, North African tenancy regulations and Nabataean industrial pollution, into a coherent economic picture.
