5. Appropriate Indicators for Recognising Types of Food and Drink Outlets

5.1 Criteria previously used to generate Roman and Pompeian shop typologies

The difficulties involved in the collection of freshly derived and legacy data, and its integration and management, are just part of the problem when working with such a massive corpus of archaeological evidence. Through this process it became clear that various types of Pompeian properties had a masonry sales counter that could potentially be used to identify them as food and drink outlets. How to best sort these properties into a meaningful typology of Roman retail outlets has challenged a number of scholars over recent years. In the mid 1980s Verena Gassner made the first significant attempt to recognise a typology of shops from the available archaeological evidence at Pompeii (Gassner 1986). Gassner wanted to understand how the complex structural and social history of Pompeii has influenced various shop plans. The reading of the spatial arrangement of shop plans to explain their specific function has some tradition in Roman archaeological scholarship, particularly at Ostia. However, for the shops at Ostia, more effective structural typologies have been established because the evidence for these buildings is limited to the 2-dimensional layout of the ground floor, with only some structural evidence for upper levels. The number of rooms, the provision of a mezzanine level, and the type of buildings attached to the shops at Ostia are the common characteristics used to create the typologies of Roman commercial buildings. For example, Giancarla Girri recognised four types of shop at Ostia, the identifications of which were initially based on a wide shuttered threshold. Her four types consisted of: a single-room shop, of variable size; a single-room shop with a mezzanine; a shop with a back room (commonly referred to as the retrobottega); and a shop with a back room plus a mezzanine level (Girri 1956, 3, 6-7). Subsequent attempts to create typologies of Roman shops have involved variations on the same theme and have effectively sorted the hundreds of 2nd- and 3rd-century AD Ostian and Roman shop types, many of which had very standardised structural plans and spatial layouts (Boëthius 1934; 1960; Meiggs 1960, 242-3; Packer 1971, 18; Staccioli 1959, 58). While the Ostian shops were sorted first by their number of rooms, and then sub-divided according to the type of buildings to which they were attached, for the Pompeian material Gassner reordered this formula by first sorting the shops according to the kinds of buildings they were attached to. By this process she identified four broad types: a shop attached to an atrium house; a shop that had no direct access to any attached building; a shop attached to an 'irregular house'; and a shop with habitation accessible only through the shop itself (Gassner 1986, 45-9). Gassner then divided each of her four types according to the number of rooms in each shop.

In an unpublished doctoral thesis on the tabernae of regio 1, Roseanne Gulino established a typology of shops and bars that was defined broadly by the size of each property, and then sub-divided by the various archaeological features that registered the most change: such criteria included the dining and garden areas; cooking facilities; the number of counters; the availability of working space at the counter; the number of inset vessels in the counter; the provision of a latrine; the various sorts of epigraphic evidence; and the decoration of the walls and counters (Gulino 1987, 130-4). Although Gulino's work represents the first attempt to define a typology of Pompeian bars by their extant installations, not just their structural plans, a number of these sets of data were inappropriately used to differentiate one type of establishment from another. For example, by differentiating properties by the various messages found among the graffiti, or by the number of remaining vessels that were once inserted into the counter, Gulino produced a classification that is not only too precise and sensitive for the anomalies in the surviving archaeological record, but also for the lack of clarity in its publication. She also produced a typology that is seemingly unrelated to the broader divisions that may have been recognised in antiquity. Nicolas Monteix recently applied the same formula to produce a similar typology of bars that depended on both the survival of graffiti among Pompeii's shops and the number of vessels found in their sales counters (Monteix 2007). This is not to discount the value of a careful study of the epigraphic record of retail outlets. However, an effective study of what the messages found scribbled over the walls of bars and shops might tell us about the different services each property once offered has yet to be carried out. In many ways, Gulino's typology was one of bricolage, not only dependent on features that were unlikely to have affected activity in antiquity, but also on the conditions of those features' preservation. Unlike the methods used to establish typologies of Roman shops, too much evidence encumbered Gulino's taberna typology, and the results were much too complex to be effectively synthesised.


© Internet Archaeology/Author(s) URL:
Last updated: Mon Jun 30 2008