For Horace (Epistles 1.14.21–25), the greasy cookshop (uncta popina) and the local tavern (taberna) were hallmarks of city living, which, he imagines, were hankered after by an urban slave unwillingly displaced to a country estate. Ellis’ new study of Roman retail rather confirms Horace's claim, demonstrating archaeologically the ubiquity of tabernae in Roman cities, which Ellis argues underwent three spikes in construction, termed here as ‘retail revolutions’. Tabernae are identified by their architectural typology, principally by a grooved threshold designed to hold shutters that could be removed to allow an entirely open frontage, ideal for commercial use. Most provide little in the way of additional archaeological evidence as to their specific function in antiquity, although Ellis classifies them throughout as retail units, relegating other probable uses such as administrative offices, barbers’ shops, doctors’ offices and the like to a single footnote (p. 5, n. 18). Food and drink outlets are by far the most readily recognizable shops, identified by the presence of masonry counters linked to hearths, and particularly visible at Pompeii, as Ellis has demonstrated extensively in his previous work. In this book he broadens his focus to include retail outlets more generally, while at the same time offering a wider study of retail landscapes across the Roman world. He draws on a remarkable range of examples from over 100 cities in a variety of regions, demonstrating a truly impressive and comprehensive knowledge of tabernae structures. Food and drink outlets do, however, still dominate here, as do Pompeian examples, leading at times to a tension between the specificity of the evidence and the broadness of some of the conclusions.

Once Ellis has established the archaeology of Roman shops, he goes on to consider the motivation behind capital investment in such permanent retail units (chapter 3). They are typically found in the busiest areas of urban centres, that is, along main roads, at intersections and near gates and public monuments. Many are attached to houses, particularly those in central locations, a trend which has often been seen as an attempt by home owners to maximize the profit from urban rents. However, Ellis modifies such market-driven explanations for investment in retail property, opting instead for an explanation rooted in social and political structures rather than economic rationality. He argues that the rent on a single shop cannot have been particularly high, and while the return from retail must have been higher, most shops were probably not particularly lucrative (pp. 104–5). Furthermore, rental on the open market may have been relatively rare; instead, shops were used by ‘elites … of varying financial clout’ to maintain their urban networks, setting up freedmen in shops in return for their social and political support (p. 108). Here, Ellis follows most notably Mouritsen (Reference Mouritsen2011). He is also offering a direct challenge (although he does not explicitly do so) to Mayer's recent picture of an independent freedmen middle class rooted in tabernae (Reference Mayer2014). A socio-political explanation for capital investment in retail units is, however, probably not geographically or chronologically universal, given that much of what we know about freedmen comes from Roman Italy in the early empire, nor does it account for any growing public investment in permanent retail space. Indeed, Ellis is cautious to note that this is not a universal explanation, and that profitability remained ‘an important part of the strategy for urban investment in retail’ (p. 107 and passim); in fact, throughout the book, rising levels of investment in retail space are linked with economic growth, indicating an underlying belief in the importance of the market as an explanatory factor. This is difficult to reconcile with Ellis’ simultaneous claim that retail was not a particularly profitable venture, nor is it clear in which circumstances economics was the chief motivating factor for investment in retail space, although socio-political and economic motivations are not in themselves mutually exclusive and could of course in many cases have coincided.

Ellis then turns to the first of his ‘retail revolutions’, linked to the urban building boom of the second century bc (chapter 4). Again drawing primarily on Italy, he shows that tabernae were an integral part of this period of urban development, driven by the demand for food and services created by a growing urban population. He argues that in this period tabernae served production and retail, acting as both workshops and shops. In his second ‘retail revolution’ (chapter 5), however, dated to the early empire, he contends that tabernae became more specialized, focusing primarily on retail rather than small-scale production, something indicated by the changing use of the Porta Stabia shops at Pompeii. In particular, he charts the rise of the specialized food and drink outlet at Pompeii, marked out in the archaeological record by distinctive masonry counters, which in themselves represent a marked growth in capital investment in retail. For Ellis, this is all part of a general Augustan economic upturn (p. 175). However, demographic expansion is also identified as key here (p. 177), and it is not always clear if the increased demand for food and drink outlets is down to a rise in urban living standards, or simply a consequence of increased levels of urbanization, with growing numbers of people reliant on the market to supply their needs. His ultimate conclusion here is that there were not just more shops than ever before in the early empire, but that they were much more specialized ones; it was, he claims, ‘a period that gave rise to the Roman retail city’ (p. 186). Ellis is surely correct to place retail at the centre of the Roman urban experience, but this does not in itself rely on the taberna, and beyond the development of bars at Pompeii, this specialization is not fully demonstrated, nor are the changes to production that drove a move away from the small-scale urban workshop explored in any detail. The vast majority of tabernae in the archaeological record remain generic in form, and could presumably have been used for a wide variety of functions. Ellis’ third and final ‘retail revolution’ marks the rise of retail homogenization in the early second century (chapter 7), with increasingly standardized shops with shuttered doorways and entrances on the right-hand side being constructed as part of much larger building complexes, incorporating shops, apartments and public spaces. I would argue that this standardization and ubiquity in fact reflect the flexibility of these units, rather than suggesting any standardization of retail practices; after all, it is often just the footprint of these units that survives, telling us very little about how they were used in antiquity.

Ellis’ concluding chapter considers the role of retail in the Roman city (chapter 8), although the focus is primarily on the place of bars in urban life. This discussion is important and interesting in its own right, but to a certain extent reflects the tension inherent throughout this book between the study of food and drink outlets specifically and the retail trade more broadly. There is little in this final chapter about the wider role of retail, about supply chains, about who staffed these shops, about the customers, about their profitability, and so on. There is some internal contradiction here, as earlier in the book Ellis claimed that shops were not particularly lucrative and that they were often staffed by freedmen, but the implications are not discussed further here. If this were indeed the case, then retail was not part of any economic boom per se, nor were shops openly available on the rental market to independent entrepreneurs. In this final analysis, however, retail is explicitly linked to economic growth. The development of food-and-drink outlets in particular reflects increased standards of living, fulfilling more of a want than a need; it is here that we see ‘cashed-up urbanites’ (p. 245) choosing to spend their money in bars, offering an economic opportunity for retailers.

Retail was for a long time neglected in studies of the ancient economy, rejected in favour of long-distance trade and a focus on production, rather than consumption. It is only recently that the retail trade has started to attract the interest that it deserves and Ellis makes an important contribution to this growing body of work. He offers a detailed study of the taberna in the archaeological record, confidently handling a vast amount of data and demonstrating the complexity not only of the world of the taberna, but also of the social, economic and political setting of the urban centres in which they operated.