Astrid Van Oyen directs our attention to the significance of storage in the Roman economy. This is perhaps an unexpected, but all the more welcome turn of perspective — and one not without polemical sting. Inspired by the New Institutional Economics, Roman economic history has recently been preoccupied by questions of economic performance. Institutions have been examined for their potential for lowering transaction costs and facilitating the integration of markets. This agenda has probably been most successfully advanced by the efforts of the Oxford Roman Economy project to update the quantitative data-sets available for ancient economic history. Objects have been counted and, each time a new find could be added to the accumulating lists, the claim seemed to have been substantiated that the Roman economy had generated greater activity than previously thought (177). Now, V.O. argues, it is time to move beyond such mechanistic interpretations of the material record. Quantification must be complemented by more qualitative approaches.

To do so, she returns to a key theme of Finley's Ancient Economy: the household and its management as the central form of organisation. Among anthropologists and development economists, interest in the household has often focused on risk-buffering strategies of subsistence peasants. But the economic strategies of households involve much more than storing to insure against a failed harvest. Storage serves a range of purposes, not least to underwrite claims to status and power, as Finley would have said. What is stored reflects cultural values and political preferences. In short, the storage practices of the Roman household occupy a historically inflected middle-ground between the abstract macro taxonomies of quantification and what we might call minimal subsistence. It is this historicity that the book then moves on to examine across a range of contexts in chs 2–5 before a general model is attempted in ch. 6 and some final reflections added in ch. 7.

Chs 2–5 move determinedly beyond the peasant household, to explore contexts where aristocratic estates developed in the wake of Roman imperial conquests. Ch. 2 explores the growth of cash-crop-producing estates in central and south Tyrrhenian Italy during the first century b.c. The wine produced by many of these estates was exported around the western Mediterranean and frequently crops up as Dressel 1 amphorae in excavations. But this market-oriented production was also reflected in a growth of what V.O., with a felicitous phrase, terms conspicuous production. Large separate structures for storage, imposing barns, copious over-ground cisterns and facilities for several vintages of wine became part of the regular equipment of aristocratic villae in Italy. These structures were not merely functional, but also served to manifest the status and values of the household. Ch. 3 leaves Italy, Dressel 1 and the first century b.c. behind, to turn to the era normally marked out by Dressel 2–4. Accordingly, the context now moves to south-western France and north-eastern Spain of the first century a.d. to explore how communal forms of small underground silos for food grains, often in the highlands, gave way to large, centralised barns on the plains. Local elites seized the opportunities of empire to embark on new market-oriented forms of agriculture (and taxation). Ch. 4 returns to first-century Italy, to explore the top end of urban households in Pompeii and Herculaneum and the micro-dynamics of urban storage practices, and not least the agency left to the slave domestics. Ch. 5 takes several steps up in scale to analyse the warehouses of Ostia and Portus and their role in the grain supply of Rome during the high empire. V.O. rightly rejects as anachronistic attempts to classify this system as based on either state redistribution or private market. The emperor and magistrates, including their freedmen, were active both through their state office and as ‘private’ estate owners and investors. Instead, V.O. identifies a two-pronged pattern. At Ostia, large warehouses were organised around a courtyard and storage space could be allocated and rented very flexibly according to need. Portus, on the other hand, offered a series of standardarised rectangular storage units arranged round the hexagonal harbour. The entire ensemble was closed on itself and could be seen much as a giant version of the courtyard warehouses of Ostia (149), but with one significant difference. The greater regularity and standardisation of Portus made the structure and stored quantities more legible: a stable and easily manageable base for the food supply of Rome. By contrast, the greater flexibility of Ostia emphasised local experience, embodied knowledge and personal networks — much as in a bazaar — to allow traders to act in response to opportunity.

Ch. 6 then ties the individual case studies together to offer a basic model of the Roman economy. Instead of seeing different types of institutions operating at different scales, V.O. posits household and family (including enslaved domestics) as the key organising principle. In kaleidoscopic fashion, the household can be seen as the crucial unit reproduced and active at all levels. Above and absorbing everything else was the imperial household. In possession of ‘extraordinary metaphorical and logistical reach’, the imperial household, unlike some early state societies, did not merely keep its large storage facilities together with the palace. Its warehouses were scattered, and located where necessary, across a much bigger area. Even so, V.O. insists, ‘the imperial family was not constituted of different stuff and relations than any other family’ (166).

This is the major wager of V.O. and one which is used to underwrite a theoretically ambitious, if at times a little self-indulgent, agenda (summarised in ch. 7). Occasionally, the reader stumbles across theoretical flourishes such as seeing the topic of storage as ‘schizophrenic’ (158) or a claim that the Roman Empire, of all places, was ‘slowly eroding the taken-for-grantedness of stuff's equation with power and wealth’ (173). Few people who have visited Rome could walk away with such an impression. Rarely in pre-industrial history did power amass so much physical matter. If the theoretical signalling occasionally strains credulity, the overall project is right on target. The Roman Empire never became an administratively standardised space. Roman bureaucracy was minuscule, governmental power a composite and authority dispersed across a jumble of overlapping networks of grand households. The result, as V.O. writes in this important and thought-provoking study, was a state edifice with plenty of cracks and a fragmented knowledge-scape.