The consistency with which these Latin labels have been applied to Pompeian buildings is the agent of perpetual misinterpretation. Latin labels carry with them modern definitions of functions and activities, which, if not critically assessed, allow erroneous, or at least weak, secondary 'evidence' to be transmitted as primary data. Kleberg's list of terms and their meanings are now accepted as if certain, and their untested use and conversion into actual data has promoted his interpretations to primary evidence. Other large indices of Pompeian building types, such as the Consorzio Neapolis database (see De Simone 1988), or that by Hans Eschebach (1970), which also present text-based labels as if they were part of the original Pompeian evidence, are also continually drawn on by Pompeian scholars as though primary data. (For retailing: Ruddell 1964; Della Corte 1965; Eschebach 1970; Packer 1978; Raper 1977, 189-221; Gassner 1986; La Torre 1988; Laurence 1994 [revised 2007]; DeFelice 2001. On recent alternate approaches to textual nomenclature, though mostly for domestic spaces: see Leach 1997; Riggsby 1997; Allison 2001a.) This insistent, and yet inconsistent, reliance on the literary evidence to label the archaeological remains has left us with an unreliable definition of what constitutes a food and drink outlet, their various types, and where they were distributed throughout the city. Few of the properties included in such registers share any common characteristics: some have counters, others do not; some are equipped with hearths and ovens, others are without; and some are large complexes with many rooms, while others are small one-room establishments. There is no pattern or inherent reason for this approach to these labels. Why was the property at I.2.29 with its counter and attached hearth labelled by Liselotte Eschebach a thermopolium, but not the property with similar fixtures at I.10.13 (Eschebach and Müller-Trollius 1993, 19, 56-7)? Why does the Consorzio Neapolis database list I.3.2 and VII.7.3 as thermopolia when they were clearly without cooking arrangements (De Simone 1988, 106, 156)? Examples of such inconsistency result in a lack of concordance between Eschebach's count of 69 thermopolia and 146 cauponae at Pompeii with the figure of 89 thermopolia and 120 cauponae derived by the Consorzio Neapolis database (Eschebach and Müller-Trollius 1993, 453-64; La Torre 1988, 78). In spite of this disorder, such inconsistent lists of building types and activities become the legacy data that are plotted onto maps to examine the distribution of urban activity. This is evident in a collection of papers that drew their evidence for the function of each building from the haphazard list of building names found in the Consorzio Neapolis database of Pompeian properties (De Simone 1988). G.F. La Torre, for example, assessed this 'data' on the function of buildings for what it could reveal about commercial space across the city (La Torre 1988). He adopted the classifications of the Consorzio Neapolis database, namely cauponae (used interchangeably with tabernae), thermopolia (partnered with popinae), and hospitia (used in conjunction with stabula). Several properties even each have multiple labels such as thermopolium, taberna and caupona. La Torre offers no explanation for how he believes the properties once functioned, yet the purpose of his study is built on that interpretation of the use of space that is provided by the activity with which these labels are associated in the texts.
That interpretations can transform into archaeological data in Pompeian scholarship is perhaps most emphatically demonstrated in the work of John DeFelice, who wrote a PhD thesis on the hospitality and retail industry of Pompeii, later published as a book (DeFelice 2001; see also DeFelice 2007 for a condensed account). Rather than visit the site itself to collect data, DeFelice drew his primary evidence from secondary works: first, his 'Master List' of the retail and hospitality businesses was 'defined' by the work of Kleberg (1957) and Eschebach (1970), while his descriptions for each, in order to 'incorporate archaeological evidence from Pompeii into each definition', were a series of paraphrases lifted directly from Kleberg, Ruddell (1964), Van der Poel (1977), and Packer (1978); second, his identifications for each building were confirmed by his 'eyes in Pompeii' (DeFelice 2001, 3), Roy Bowen, who took photographs of some properties (only four were included in the published version) 'which were generously lent for my research, as well as his extensive field notes' (DeFelice 2001, 176). The resulting 'exhaustive appendix' omits properties that should have been included, and includes those that should have been omitted. Also his 'results' are ambiguous: for example, the property at VI.1.2 is regarded 'at once a hospitium, caupona, stabulum and taberna/popina (DeFelice 2001, 31).
This lack of engagement with the actual archaeological data at Pompeii, or at the very least to attempt some verification of these property listings, constitutes a certain laziness in scholarship, and is perhaps borne out of an intimidation by the gigantic body of primary and legacy data that exists for the site. The recent surge in spatial and statistical studies at Pompeii, particularly since the 1990s, have only perpetuated this problem of using interpretative information as if it were second-generation 'legacy data' (for this term, see Diacopoulos 2004). In almost impatient efforts to manage huge bodies of Pompeian data so that they may be run through databases for statistical analyses, rather than to generate one's own data from the primary archaeological data, scholars are persisting with the unverified and interpretative 'legacy data' as though these represent something more than unsophisticated and inaccurate compilations of information about how certain types of buildings might have functioned in antiquity (e.g. De Felice 2001; Laurence 1994 and 2007). What is needed is a clearer understanding of the differences between primary data and secondary interpretations.
With this in mind I now outline my own approach to the study of huge bodies of data available to Pompeian archaeologists. This approach is one that gives priority to the archaeological site for the generation of data, but also relies on a more considered use of the original excavation records - the creation of the legacy data - to develop criteria that can identify and provide a typology of the retail food and drink outlets at Pompeii.
© Internet Archaeology/Author(s)
URL: http://intarch.ac.uk/journal/issue24/4/2.1.2.html
Last updated: Mon Jun 30 2008