Comp. by: PG2846 Stage : Proof ChapterID: 0001306139 Time:16:21:09 Filepath:d:/womat-filecopy/0001306139.3D Date:17/6/11 OUP UNCORRECTED PROOF – FIRST PROOF, 17/6/2011, SPi 8 Where to Park? Carts, Stables, and the Economics of Transport in Pompeii Eric E. Poehler A critical factor in the growth of preindustrial economies has been the capacity of those economies to develop effective methods of transportation.1 Roman methods of transport are normally thought to have been so backward that they severely hindered economic development . . . It is primarily the high cost of land transport that makes historians pessimistic about the capabilities of the Roman transport system.2 How can scholars reconcile the complex commercial landscape of a city like Pompeii with the commonly held belief in the inefficiency of land-based transportation? Are we too pessimistic about the costs of land transport or do we misunderstand the economic sophistication of the urban landscape? On the one hand, it is hard to ignore the hundreds of (apparently successful) commercial properties at Pompeii and the complex economic relationships they reveal in their physical associations with other buildings, especially houses. Similarly, the enormous reconstruction efforts following the earthquake(s) of 62 ce demonstrate that vast quantities of materials were transported in and out of the city in a short amount of time.3 On the other hand, the idea of the inefficiency of land transport has been undermined by Laurence, who has shown the comparison with waterborne transportation to be misleading.4 Recent research by the author has also demonstrated that wheeled traffic in Pompeii was organized into an efficient system governed by the rule of right-side driving, the imposition of congestion-relieving detours, and by 1 3 2 Martin (2002: 151). Harris (1993: 27). 4 Dobbins (2007: 173–5; 1994). Laurence (1998b: 126–32). Comp. by: PG2846 Stage : Proof ChapterID: 0001306139 Time:16:21:09 Filepath:d:/womat-filecopy/0001306139.3D Date:17/6/11 OUP UNCORRECTED PROOF – FIRST PROOF, 17/6/2011, SPi The Economics of Transport in Pompeii 195 alternating series of one-way streets.5 The confluence of these facts—that a large amount of goods were continually transported into and out of the city and that the municipal authorities acted to facilitate that supply—argues in general for transport to have been more efficient economically than is generally accepted. Without stepping too strongly into the modernist/primitivist debates, it now seems more certain that the answer to this apparent paradox is the need for new commonly held beliefs. What remains to be understood, however, is how this efficiency was accomplished and how it was enacted within the urban environment and beyond. This chapter takes the first step towards addressing this question by considering one aspect of Pompeii’s transportation economy: the storage of wheeled vehicles.6 Rather than attempting to define the transport economy outright, it is better to ask, as Glenn Storey has done for the Roman economy more broadly, what are the ‘kinds of [economic] institutions and how did they work?’7 The traffic system can be seen as one such institution, but less formal mechanisms—the economic reasons that carts were travelling in the first place—were certainly more important underlying motivators. This kind of segmented approach is necessitated by an absence of evidence. While mules and human porters made up the greatest proportion of the means of transport, these methods of conveyance simply don’t make the same physical impact (e.g., ruts and worn street features).8 Additionally, mules and porters are far less archaeologically recognizable in their secondary contexts. Thus, how does one determine if a manger in a bakery represents the storage of animals for transport, traction, or both?9 Carts, however, are singular in purpose, traceable through their impacts across the urban environment, and identifiable through secure architectural proxies. Although carts made up only a portion of the total means of transport at Pompeii, the evidence for their circulation and storage make them an excellent sample from which to draw conclusions about the economic rationalities governing transportation choices in and around the ancient city. The following discussion therefore first focuses on those locations where carts can be observed to leave the space of the street—ramps—and the spaces to which 5 Poehler (2006: 59–74; 2009: 38–121). See also the chapters by Hartnett and Kaiser, in this volume. 6 The term ‘transport economy’ is used here to mean all economic relations of the act of transporting goods and people as well as those operations that support such transportation. As there is no body of water within Pompeii, all transport is necessarily land-based even if significant quantities of materials were carried by boat prior to their arrival in Campania. 7 Storey (2004: 107). 8 These groups are represented in epigraphic (mules, CIL IV.113, 97, 134; porters, CIL IV.497) and artistic (Nappo 1989: 80–1, figs. 1–2, 4; Ling 1990b: 56, fig. 4.14; Guasti 2007: 140, fig. 3) evidence at Pompeii. 9 e.g., draft animal skeletons were found in a stable (no ramp is known) at the bakery at IX.12.6–7. See Varone (2008) for the most recent work and bibliography. Comp. by: PG2846 Stage : Proof ChapterID: 0001306139 Time:16:21:09 Filepath:d:/womat-filecopy/0001306139.3D Date:17/6/11 OUP UNCORRECTED PROOF – FIRST PROOF, 17/6/2011, SPi 196 Eric E. Poehler they led—stables. An examination of the spatial and social distributions of these features then reveals important clues about who owned carts and/or the means to serve those travelling and how these distributions reflect different economic strategies within the transport economy. The specificity of our conclusions about the transport economy is constrained by the specificity of the evidence that can be brought to bear upon the question. Thus, even if the evidence for the traffic system’s organization can demonstrate the direction that a cart could travel on each excavated street in the city, without knowing where a vehicle began its journey and its intended destination, the restrictions on direction offer an abstraction of how traffic did circulate generally rather than any historical description of the particular and purposeful paths across the city. To assess the reason for travel one must establish a destination. The most obvious destinations for travel in Pompeii are monumental buildings and the city gates. For cart drivers and pedestrians alike, monumental buildings were important landmarks that required a large volume of supply. For example, the macellum, the baths, and the amphitheatre all would have required vast quantities of animals, fuel, and equipment appropriate to their activities.10 The monumental building as a destination, however, is only different from the humble shop in scale; although we know what must have been delivered, we cannot observe it archaeologically. Conversely, it is known that all traffic that entered and exited the city passed through the seven city gates. Compared to monumental buildings, gates offer a specific point at which the passage of vehicles can, necessarily, be established. Because every entering cart passed under the gates, however, there is no differentiation in purpose; gates suffer from the opposite problem, a lack of specificity. RAMPS AND STABLES The identification of a ramp leading out of the street combines the means to definitively establish the presence of vehicles with the ability to limit the purposes of their travel by the types of properties they entered. Moreover, as markers of destination, ramps are important because their different forms offer evidence for the economics of cart driving generally as well as the investment choices of specific properties. When a ramp is compared to other treatments of sidewalk space—such as its elaboration through decorative pavements or its truncation by a building to a line of curbstones or even less— 10 On evidence for animal processing at the macellum, cf. Maiuri (1947: 166); also McKinnon (2004: 184–94). Comp. by: PG2846 Stage : Proof ChapterID: 0001306139 Time:16:21:09 Filepath:d:/womat-filecopy/0001306139.3D Date:17/6/11 OUP UNCORRECTED PROOF – FIRST PROOF, 17/6/2011, SPi The Economics of Transport in Pompeii 197 it becomes obvious that the choice to construct a ramp indicates the high importance of wheeled vehicles and the material supply they carried to that property.11 Just as an oven or a millstone attests to the kinds of economic activities that took place within a property, a ramp signals the specific need for a large volume of material to be brought into and/or out of the property or for a high number of vehicles to be staged within the property. Of the thirty-six identified ramps, three ramp types can be identified at Pompeii: Paved Ramps, Rut Ramps, and Side Ramps. Paved Ramps are openings in the curbstones where an inclined section of paving permits access for wheeled vehicles between the street and a stable area (Fig. 8.1). Rut Ramps are identified by rutting worn into an uninterrupted line of curbstones, demonstrating the movement of wheeled vehicles between the street and a stable area (Fig. 8.2). In some cases, the curbstones have been cut or worn down completely to create a smooth, but very short incline.12 Side Ramps are sections of a street where the curbstones have been removed and replaced by an area of paving stones, usually inclined, in order to provide a place for wheeled vehicles to park out of the space of the street (Fig. 8.3). Some ramps are identified only by the existence of guard stones at the corner of the ramp— either at the street or at the door to the stable—or by an opening in the curbstones. These ramps are identified, but categorized as Unknown type. Except for Side Ramps, all ramps lead to a stable area. As part of the property’s economic infrastructure, stables were even more costly than ramps because stables were far larger in area and competed for space with other activities that might have been more productive. Economically, the equation is simple: one could profitably use the space to operate an inn with stabling services or one could reduce their own transport costs by owning the means of transportation. Properties with ramps and stables (hereafter transport properties) have two important distributions: across the types of properties that make up the urban fabric and across the space of the city.13 The first distribution shows that while the variety of transport properties was relatively restricted, the kind of ramp employed was strongly correlated with a type of property to which it led. Table 8.1 expresses these associations. Thus, nearly half of all Paved Ramps led to inns, which is almost four times more than any other identified transport 11 On the elaboration and truncation of sidewalks see e.g. that on the west side the Casa dei Vettii, which was already only the width of the curbstones, was covered by decorative masonry. 12 These are I.4.28, VII.13.17, and IX.1.28. 13 The identification of these transport properties (listed in Table 8.3) was made through onsite survey conducted between 2007 and 2009 combined with research into the publication record of each property. The main sources consulted were Eschebach (1993), Fiorelli (1875), Jashemski (1973, 1977, 1979), Mau (1899), Mayeske (1972); Overbeck (1884); Packer (1978); van der Poel (1986), and the publications Pompeianarum Antiquitatum Historia; Giornale degli Scavi di Pompei; Notizie degli Scavi di Antichità. Comp. by: PG2846 Stage : Proof ChapterID: 0001306139 Time:16:21:09 Filepath:d:/womat-filecopy/0001306139.3D Date:17/6/11 OUP UNCORRECTED PROOF – FIRST PROOF, 17/6/2011, SPi 198 Eric E. Poehler Fig. 8.1. Paved Ramp at I.1.4 (photo by Eric E. Poehler with permission of the Soprintendenza Archeologica di Pompei). Fig. 8.2. Rut Ramp at II.5.4 (photo by Eric E. Poehler with permission of the Soprintendenza Archeologica di Pompei). Comp. by: PG2846 Stage : Proof ChapterID: 0001306139 Time:16:21:10 Filepath:d:/womat-filecopy/0001306139.3D Date:17/6/11 OUP UNCORRECTED PROOF – FIRST PROOF, 17/6/2011, SPi The Economics of Transport in Pompeii 199 Fig. 8.3. Side Ramp at VIII.6.10 (photo by Eric E. Poehler with permission of the Soprintendenza Archeologica di Pompei). Table 8.1 Distribution of ramps across property types Property Commercial - Inn Commercial - Bakery Commerical Agricultural Commercial Theatre Residential Workshop Residential Unknown Totals Paved % of Ramp Group % of Rut Ramp Group % of Side Ramp Group Unidentified % of Group 8 1 1 47.1 5.9 5.9 2 0 2 22.2 0.0 22.2 0 1 0 0.0 20.0 0.0 1 0 0 20.0 0.0 0.0 2 0 0 11.8 0.0 0.0 1 0 0 11.1 0.0 0.0 0 1 3 0.0 20.0 60.0 1 0 1 20.0 0.0 20.0 2 3 17 11.8 17.6 47.2 3 1 9 33.3 11.1 25.0 0 0 5 0.0 0.0 13.9 2 0 5 40.0 0.0 13.9 property. Rut Ramps are much more evenly distributed across all transport properties with large houses having a slightly larger proportion (33 per cent). Side Ramps are more variable in their forms and therefore their potential costs. The 36-m-long Side Ramp on Via Stabiana at the teatrum tectum is the type form for these ramps, though the only example on this scale. Other Side Comp. by: PG2846 Stage : Proof ChapterID: 0001306139 Time:16:21:12 Filepath:d:/womat-filecopy/0001306139.3D Date:17/6/11 OUP UNCORRECTED PROOF – FIRST PROOF, 17/6/2011, SPi 200 Eric E. Poehler Table 8.2 Spatial Distribution of transport properties across property types Entrance Ramp Type % of Property Area used for Stable Property Type Identification At Gate? 1.2.22 20.2 Commercial Yes VI.16.23 VI.17.1 VIII.7.1 I.1.3 I.1.8 III.11.a IV.5.c VI.1.1 VI.15.18 VI.1.4 I.20.5 Unknown Guard Stones Paved Ramp Paved Ramp Paved Ramp Paved Ramp Paved Ramp Paved Ramp Paved Ramp Rut Ramp Rut Ramp Unknown Paved Ramp 13.9 17.0 33.l 34.7 39.7 Unknown (unexcavated) Unknown (unexcavated) 36.3 38.2 49.7 6.1 Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes II.5.4 Rut Ramp Unknown (unexcavated) III.7.6 Rut Ramp Unknown (unexcavated) VI.1.24 Vl.1.22 IV.3.h IX.2.24 1.8.12 IX.1.28 1.3.27 Vl.6.17 II.4.4 IX.9.12 1.6 27.7 Unknown (unexcavated) 27.0 65.2 31.3 32.2 n/a 0.9 14.0 VII.1.43 I.4.28 VI.6.13 Vll.13.17 I.10.14 Vl.3.28 Paved Ramp Paved Ramp Paved Ramp Paved Ramp Paved Ramp Rut Ramp Paved Ramp Side Ramp Paved Ramp Unknown Guard Stones Paved Ramp Rut Ramp Rut Ramp Rut Ramp Unknown Side Ramp Commercial - Inn Commercial - Inn Commercial - Inn Commercial - Inn Commercial - Inn Commercial - Inn Commercial - Inn Commercial - Inn Commercial - Inn Commercial - Inn Commerical Agricultural Commercial Agricultural Commercial Agricultural Residential (A/P) Unknown Unknown (unexcavated) Commercial Commercial Commercial Commercial - Bakery Commercial - Bakery Commercial - Inn Residential (A) No No No No No No VII.2.22 Side Ramp n/a VIII.6.10 Side Ramp n/a 1.13.14 Unknown 4.9 Residential (A/P) Residentail (A/P) Residential (A/P) Residential (A/P) Residential (A/P) Residential-Workshop (Bakery) Residential-Workshop (Bakery) Residential-Workshop (Bakery) Residential-Workshop (Wine) Theatres Unknown (unexcavated) Unknown (unexcavated) VIII.7.17–21 Side Ramp III.7.2 Paved Ramp III.6.2 Rut amp 2.9 1.6 2.0 2.7 8.3 n/a n/a Unknown (unexcavated) Unknown (unexcavated) Yes Yes Yes Yes Yes No No No No No No No No No No No No No Comp. by: PG2846 Stage : Proof ChapterID: 0001306139 Time:16:21:12 Filepath:d:/womat-filecopy/0001306139.3D Date:17/6/11 OUP UNCORRECTED PROOF – FIRST PROOF, 17/6/2011, SPi The Economics of Transport in Pompeii 201 Ramps are identified by large sections of sidewalk that have been replaced by street paving,14 while still others demonstrate parking occurred on the sidewalk area itself.15 There is very little variation, however, in the property association of Side Ramps: 80 per cent are found at bakeries, either within a residence or as a stand-alone workshop. The typology of ramps and their associated properties (Table 8.2) indicates that there were several choices available to those who were interested to invest in vehicular access to their property. Cost, durability, and visibility were the most important factors influencing this choice. The paved Side Ramps were the most costly in terms of materials, but this may have been offset by their size, which allowed multiple deliveries to occur simultaneously. Moreover, the absence of a stable meant that more of the internal area could be used for productive activities. Thus, the association of Side Ramps with bakeries fits well with an expected investment strategy in which the use of space and animal power would have been primarily for baking and/or milling. Similarly, the association of Paved Ramps with the hospitality industry demonstrates that Paved Ramps not only provided a durable, even surface for vehicles to traverse, they also advertised stabling services to those entering the city.16 Conversely, Rut Ramps were not intended to be as obtrusive, an observation which helps to explain their strongest correlations with the secondary entrances of houses and with the large vineyards near the Porta di Sarno.17 The second distribution of transport properties is spatial; they are either clustered very near the city gates or spread out more widely across the rest of the city (Fig. 8.4). Drawing a 100-m arc around each of the city gates circumscribes seventeen of the thirty-six transport properties in the city.18 While the number of transport properties found in proximity and away from gates is equivalent, the area of the city near the gates is only 19.9 per cent of the total area.19 Of the ramps near the gates, fourteen open directly onto the main, twoway streets leading to a gate with the remaining three found on one-lane streets that intersect or parallel that main street.20 Four of the seven gates (Ercolano, Stabia, Sarno, and Nola) have transport properties located on the street immediately inside the gate. In fact, at each of these gates the very first doorway inside the city is accessed by a ramp.21 As a whole, the concentration 14 These are VII.6.10 and VI.3.28. These are VII.2.22 and VI.6.17. Parking on the sidewalk area would have inconvenienced pedestrians, on which more broadly see Hartnett, in this volume. 16 Bar counters also were oriented to advertise to potential customers. Ellis (2004: 381–4). 17 These are II.5.4 and III.7.6. 18 100 m was chosen arbitrarily. Expanding or contracting this arc might alter the percentages, but the clustering of transport properties very near the gates would still show as significant. 19 The total area of the city is approximated to be 628,048.22 m2 and the total area within 100 m of a city gate is 125,187.43 m2. 20 These are I.2.22, VI.1.22, and VI.1.24. 21 These are II.5.4, III.11.a, VI.1.1, VI.17.1, and VIII.7.1. 15 Comp. by: PG2846 Stage : Proof ChapterID: 0001306139 Time:16:21:12 Filepath:d:/womat-filecopy/0001306139.3D Date:17/6/11 OUP UNCORRECTED PROOF – FIRST PROOF, 17/6/2011, SPi 202 Eric E. Poehler Fig. 8.4. Distribution Map of Transport Properties (image by Eric E. Poehler). of so many ramps and stables near the gates elides with the expected high volume of traffic here and demonstrates the economic importance of traffic for the operation and development of the city. In fact, this traffic establishes the gates as important peripheral generators of urban space.22 More specifically, the kinds of traffic-related space generated at the gates as well as away from them both reveal particular economic rationales. Table 8.3 shows the spatial distribution of transport properties in relation to their distribution across property types. Property types are listed along the left side of the table while the number and percentages of these types are shown in two groups of three columns. The first column lists the number of each property type near to or away from the gates, while the second column (% of Category) expresses this difference in number between the two groups as a percentage. The third column (% of Group) shows the percentage that a particular property type makes up within each group, that is, either at or not at the gates. Two trends are remarkably clear from this tabulation. The first is the considerable concentration of inns at the city gates: over 90 per cent (90.9) of all inns with a ramp are located here. Of these, 70 per cent have Paved Ramps. Such a concentration of inns shows this service to have been a very profitable use of space, especially when compared to residential properties 22 On gates as generators of space, see Malmberg and Bjur in this volume, for examples in Rome. Comp. by: PG2846 Stage : Proof ChapterID: 0001306139 Time:16:21:13 Filepath:d:/womat-filecopy/0001306139.3D Date:17/6/11 OUP UNCORRECTED PROOF – FIRST PROOF, 17/6/2011, SPi The Economics of Transport in Pompeii 203 Table 8.3 List of all transport properties Property Type Commercial - Inn Commercial - Bakery Commercial Agricultural Commercial Residential Workshop Residential Theatre Unknown Totals At % of Gates Category % of Group Not At Gates % of Category % of Group Total 10 0 3 90.9 0.0 100.0 58.8 0.0 17.6 1 2 0 9.1 100.0 0.0 5.3 10.5 0.0 11 2 3 1 0 25.0 0.0 5.9 0.0 3 4 75.0 100.0 15.8 21.1 4 4 1 0 2 17 14.3 0.0 50.0 5.9 0.0 11.8 100.0 6 1 2 19 85.7 100.0 50.0 31.6 5.3 10.5 100.0 7 1 4 36 with a ramp (14.3 per cent). In fact, the one house near the gates is the only non-commercial property to have wheeled access.23 The second trend revealed by Table 8.3 is the disproportionate number of residential properties with ramps spread throughout the rest of the city. Fully 85.7 per cent of all houses with stables are found away from the gates. These six houses make up 31.6 per cent of the group away from the gates, a percentage that is approached only by houses that have workshops within them (21.1 per cent). The increase in the number of houses with stables in this group equals the combined total of commercial properties (31.6 per cent), with the dual nature of Residential— Workshops nicely balanced in between. The strictly residential properties are remarkable not only for their having vehicular access, but also for the presence of status architecture—all have an atrium and six of seven also have a peristyle—and their very large size. The average size of these houses is 1344.7 m2, more than three times the average area (404.7 m2) of all houses in the city. Such a large average area suggests that there was a threshold in residential size, likely related to a larger need for supply, where it became both economically viable and socially desirable to dedicate an area to the storage of a vehicle and draft animals. Averages, however, have outliers and the equation is not that simple. The 1344.7 m2 average is made up of two houses over 2000 m2, two houses below 600 m2, and three houses spaced evenly between these.24 The decision to invest in transport infrastructures is not simply proportional to property size. If it were, one would expect more than four of the twenty-five houses larger than 1000 m2 to be equipped with a ramp and stable. Instead, the 23 This excludes the two unidentified properties, one of which (IV.3.h) was probably an inn. These are: I.4.28 at 2,308.11 m2, VI.6.13 at 2,263.39 m2, I.10.14 at 1,835.95 m2, VI.1.24 at 1,124.69 m2, VII.1.43 at 789.17 m2, I.10.17 at 590.71 m2, and IX.9.12 at 500.86 m2. 24 Comp. by: PG2846 Stage : Proof ChapterID: 0001306139 Time:16:21:13 Filepath:d:/womat-filecopy/0001306139.3D Date:17/6/11 OUP UNCORRECTED PROOF – FIRST PROOF, 17/6/2011, SPi 204 Eric E. Poehler purpose of a ramp and stable is intertwined in the specifics of a particular property. Well-studied houses, such as the Casa del Menandro (see below), offer strong evidence for the urban house being one node in a larger economic unit. Houses without a ramp and stable, however, rarely reveal the reason for a choice not made. Although we should not underestimate the volume of their supply, without a ramp and stable it is safe to assume that most of that supply, by mass and/or weight, did not require a cart. At the same time, those few residential properties with transport infrastructure therefore represent the highest level of investment in the transport economy. This division, based on ramps and stables, leads to two conclusions about the supply of houses in Pompeii. First, the majority of houses were supplied by mules and porters, a fact that gives still greater importance to these methods of conveyance. Second, as important as they were, these mules and porters can only be represented by their archaeologically observable proxies, ramps and stables. In combination, these two intersecting trends—the property distribution of ramps and the spatial distribution of transport properties—reveal two distinct economic rationales for constructing an access to and specific area for the storage of wheeled vehicles. The clustering of inns with ramps near the gates demonstrates that property owners saw the compression of traffic at the gates as an economic opportunity and invested in these particular commercial infrastructures to offer a specific service. The number of inns and the competition between them to provide ramps at the very first doorways inside the gate indicate the success of these enterprises. Conversely, away from the gates, ramps most often led to large private residences, illustrating a very different economic rationale. Rather than advertising themselves with well-paved ramps on main thoroughfares, these ramps and stables were constructed because the supply to the house, for whatever purpose, was great enough to outstrip the abilities of donkeys and porters. Moreover, this supply was both so large and so constant that it was economical for the property owner to also own (a significant portion of) the means of supply. Of course, the total volume of traffic at a particular property or for the city as a whole is far beyond what the evidence from stables can demonstrate. Still, these two divisions of transport properties operate as proxies for what is archaeologically invisible (i.e., donkeys and porters) and represent two modes of the transport economy, the Household Mode and Commercial Mode, which at times complemented and competed with each other. THE HOUSEHOLD MODE OF TRANSPORT The Household Mode of transport was driven by the constant demand for supplies required to maintain the people and activities of an elite house as well Comp. by: PG2846 Stage : Proof ChapterID: 0001306139 Time:16:21:13 Filepath:d:/womat-filecopy/0001306139.3D Date:17/6/11 OUP UNCORRECTED PROOF – FIRST PROOF, 17/6/2011, SPi The Economics of Transport in Pompeii 205 as its associated, if not always architecturally interconnected, dependencies. Although this claim about the elite house seems obvious, packed within it are three assumptions that need to be outlined. The first is that the elite house was often filled with large numbers of people, including many beyond the family and slaves of the paterfamilias.25 The second assumption is that Roman elites did not abhor involvement in trade, even if they chose to distance themselves from it physically through architecture and administratively through middlemen.26 Finally, the elite urban townhouse might be best understood as one part of a larger economic portfolio that often included urban rental and/or production properties as well as rural villae whose relationship to the townhouse is more reciprocal than the Consumer City model admits.27 How the Household Mode actually worked, that is, how it gave economic advantage is revealed by an overlooked aspect of the well-known episode in Cato’s De Agricultura (22.3). Factoring into his decision about the economics of purchasing a mill from Suessa or Pompeii were not only price differences, but also the transport costs.28 At 25 miles distant from his villa, Cato reports the transportation cost for bringing the mill from Suessa was 72 sesterces using six men for six days, while the cost for delivery from Pompeii, 75 miles away, was nearly four times that amount, 280 sesterces. These figures seem to show transportation costs growing dramatically with distance. There is however, an important difference. While the trip to Suessa was a round trip as Cato’s own carts were used, making the trip a total of 50 miles, the mill from Pompeii was being delivered.29 Comparing these costs per mile, the trip to and from Suessa was 1.44 sesterces per mile while the trip from Pompeii was 3.73 sesterces per mile.30 The difference in price is striking; transportation costs are two and a half times less when one owns the means of transport. Such variance in cost is the economic advantage behind the Household Mode of transport. 25 Wallace-Hadrill (1994: 95). Aubert (1993: 172–8); Wallace-Hadrill (1991). 27 Or, to put it another way, as Purcell (1995a: 172) has done, the townhouse is more like a villa; ‘the topography and institutions are those of a town but the social and cultural fabric is that of a cluster of villae and their dependencies’. 28 Laurence’s (1998) conclusions and methodological approach are adopted in the following calculations. 29 Laurence (1998b: 131–6); Yeo (1946: 221–4). 30 These figures assume the costs of the return trip from Pompeii are not included in the Commercial Mode, since these drivers could contract to carry other materials for the return trip. The rationale for this assumption is that in the competitive market of the Commercial Mode, businesses that routinely ‘deadheaded’ (i.e., returned empty) would quickly lose out to those who did not because the later businesses could therefore offer prices that were lower by 50% or more. There is no reason to assume that ‘deadheading’ was any more common in the Household Mode. Since the vehicle necessarily had to go and return, however, there is no change in the cost of the travel per mile even if, by percentage, the profitability of the trip could be increased. In the Household Mode, costs per mile are relatively fixed; in the Commercial Mode, costs per mile are price-dependent. 26 Comp. by: PG2846 Stage : Proof ChapterID: 0001306139 Time:16:21:13 Filepath:d:/womat-filecopy/0001306139.3D Date:17/6/11 OUP UNCORRECTED PROOF – FIRST PROOF, 17/6/2011, SPi 206 Eric E. Poehler On the other hand, this same example also demonstrates that the Household Mode was restricted by distance. If Cato had sent his own vehicles to Pompeii, the return trip of 150 miles would have cost 216 sesterces by his figures, or 77 per cent of the cost to have it delivered. More importantly, the time necessary for the trip would have tripled as well, taking away six men and six oxen from other work for eighteen days and delaying the arrival of the mill by nine days. The value of lost production from men and machinery, though not discussed by Cato, would have made the Household Mode of transport less efficient than the Commercial Mode at this distance.31 The Household Mode was therefore economically limited to a relatively local circulation, within which the urban residence was one of several nodes repeatedly visited. Any rural properties associated with the house would certainly have been common destinations, as the transport needs of the villa were greater than those of the urban residence.32 Varro is explicit about the link between availability of the infrastructure of transport—especially roads and rivers—and the success of a villa.33 Moreover, if Strabo’s description of the area of the Bay of Naples as having been given over completely to villas and agricultural production by Augustan times is accurate (and archaeological evidence seems to support it), then the success of that production is in part related to the strong integration of the economic networks of urban and rural production and consumption.34 The means of that integration was the efficiency of the local transportation of raw materials and finished products, an efficiency made possible by effectiveness of the Household Mode of transport. The Casa del Menandro (Fig. 8.5) exemplifies the complex economic interconnections of an urban household in the service of its owner’s social, political, and commercial ambitions and the extra-urban sources of its supply and wealth. At 1,835 m2, the Casa del Menandro is the ninth largest house in the city and includes an atrium, peristyle, suite of dining rooms, private bath complex, and large service quarter. Thus, it is easy to imagine that the day-today operation of the house would have required an enormous amount of food and drink appropriate to master, visitor, and slave.35 The existence of a large (169.53 m2) stable area in the south-east corner of the house, the remains of a 31 If the cost of travelling to Suessa was 72 sesterces or 2 sesterces per day/per person (including animals), it is not unreasonable to assume that the same figure is close to the minimum value of the labour that these same persons and animals would produce should they have not gone on the trip. Therefore, at least this amount should be calculated as loss in productive value for the trip to Pompeii. 32 Whittaker (1985) argues from epigraphic and archaeological data on amphorae that most transportation occurred between private properties whether owned by one or multiple individuals. More recently, Pena (2007: 61–118) has argued for commercial resale and reuse of amphorae. 33 Varro Rust. 1.2.23; Laurence (1998b: 134–5). 34 Strabo 5.4.8; Moorman (2007: 436, fig. 28.1). 35 Artefactual information, including nineteen human skeletons, from Allison (n.d.). Comp. by: PG2846 Stage : Proof ChapterID: 0001306139 Time:16:21:13 Filepath:d:/womat-filecopy/0001306139.3D Date:17/6/11 OUP UNCORRECTED PROOF – FIRST PROOF, 17/6/2011, SPi The Economics of Transport in Pompeii 207 Fig. 8.5. Plan of the Casa del Menandro (image by Eric E. Poehler). cart and the other artefacts found there physically demonstrate the means to transport the volume of materials suggested by the size, architecture, and amenities of the house.36 The finds in the stable area further illustrate the Household Mode of transport by associating real objects with the means of their conveyance. First, over forty amphorae were found against the south wall of the stable area, indicating either the delivery of (likely) wine for consumption within the house, sale in the associated bar property (I.10.13), or the transportation of empty vessels for filling elsewhere.37 A handmill for grinding grain was also found here, one of five discovered in the house, along with two piles of lime and various ceramic and metal finds. Throughout the rest of the house, the excavators found at least fifty-nine lamps, lanterns, lamp bases or other 36 On the cart, stable, and finds cf. Allison (2004: 106–7, 110–12, fig. 5.18); Ling (1997: 108–14, 139); Maiuri (1933: fig. 179). The large stable yard and its manger area was divided in a manner suggesting four animals were kept here, although no skeletons were found. This absence led Maiuri (1933: 193–4) to believe the animals were in the fields at the time of the eruption, while Allison (n.d.) follows Ling (1997: 114) in noting a number of possible reasons for their absence. 37 Cf. Ellis (2005: 268–70, figs. 48b–48c) on this property. The later interpretation is preferred by Ward-Perkins and Claridge (1980: n. 88). Comp. by: PG2846 Stage : Proof ChapterID: 0001306139 Time:16:21:14 Filepath:d:/womat-filecopy/0001306139.3D Date:17/6/11 OUP UNCORRECTED PROOF – FIRST PROOF, 17/6/2011, SPi 208 Eric E. Poehler lighting fixtures as well as five masonry ovens for cooking and heating, all of which needed a constant supply of oil and wood for fuel. In the service quarter of the house a great supply of agricultural tools were recovered that exceeded the needs of the peristyle garden.38 A plough was even reportedly discovered within the house.39 These objects, in association with the structural arrangement of service quarters in the house’s final phase help to identify the Casa del Menandro’s role in supporting ‘an agricultural operation outside the city’.40 The combined evidence of architecture and artefacts as well as literature (though not discussed here) describing the rich material culture of Roman households shows the variety of objects that were constantly moving into and out of large urban houses. Moreover, the kinds of objects—the agricultural implements, building materials, and large quantities of storage amphorae— demonstrate the interconnection of many houses and rural production. The house was not an economic dead-end for production, the parasite of the consumer city model, but rather was an integral part of the movement of goods between the urban and rural environments. Indeed, Cato’s mill was transported from the city into the countryside. The Household Mode of transport was the most cost-effective means by which this integration occurred, as the owners of urban and rural properties moved materials to, from, and between these locations for their own internal supply, the supply of their dependants, or to deliver raw materials or finished goods to urban, local, or foreign markets. THE COMMERCIAL MODE OF TRANSPORT The Commercial Mode of transport operated along side the Household Mode and complemented it by offering transportation services to those who could not afford or did not need to own their own means of transport. Three main factors informed the decision to rent transportation or to own it (i.e., the Household Mode), not least of which was the high cost of an animal and its maintenance.41 Beyond the cost of owning the means of transport, the frequency of the need to travel and distance to be covered greatly impacted the profitability of the endeavour and therefore the choice of the mode of transport. As the discussion of delivering Cato’s mill demonstrated, in a short trip to a local destination (perhaps a frequent trip from Cato’s villa) owning the means of transport was far more cost-effective. On the other hand, a single trip 38 These tools were: eighteen pruning knives, four hand mills, ten hoes, three picks, a mattock, a shovel, a rake and a set of shears (Allison n.d.). 39 The plow is recorded by Kolendo (1985) and cited in Purcell (1995a: 165 n. 56). 40 Ling (1997: 105). 41 Martin (2002: 164 n. 59). Comp. by: PG2846 Stage : Proof ChapterID: 0001306139 Time:16:21:14 Filepath:d:/womat-filecopy/0001306139.3D Date:17/6/11 OUP UNCORRECTED PROOF – FIRST PROOF, 17/6/2011, SPi The Economics of Transport in Pompeii 209 to a distant town offered no advantage and further excursions would have even operated at a loss. In the Commercial Mode, the means of transport were hired from a range of contractors who might have been an individual with a single cart and donkey for hire to a business with a large number and variety of vehicles, draft animals, and manpower. Thus, the proprietor of a taberna might hire a cart and draft animal to pick up a load of stones and the dolia to build a masonry counter in order to offer and advertise the sale of prepared foods. For the owner of a large private property who is charged with the responsibility of repaving the adjacent street, a number of carts, drivers, draft animals, and slaves would need to be contracted at once so that the large amount of lava needed for paving stones could be brought to the street in quantity and with efficiency. In these examples, individuals enter into two different kinds of contracts. The first is through locatio conductio rei, a contract whereby property is rented, but no services are part of the agreement. Conversely, in the second example, the contract falls under locatio operis in which the contractor is liable for the entire project, including the objects being delivered.42 An example of the latter, preserved in a second-century ce papyrus, is instructive: ‘And I will weigh and give to your cameleer another twenty talents for loading up for the road inland to Coptus, and I will convey [sc. the goods] inland through the desert under guard and under security to the public warehouse for receiving revenues at Coptus, and I will place [them] under your ownership and seal, or of your agents or whoever of them is present, until loading [them] aboard at the river, and I will load [them] aboard at the required time on the river on a boat that is sound, and I will convey [them] downstream to the warehouse that receives the duty of one-fourth—the charges for conveyance through the desert and the charges of the boatmen and for my part of the other expenses’.43 The language of this contract gives us a glimpse of the complexities of the Commercial Mode of transport and the detailed manner in which those complexities—the costs including duties, the route, timing, security, storage arrangements, and communications with agents—were set out and agreed upon. Pompeii was typical of Roman towns concerning the spatial organization of the Commercial Mode of transport. If one wished to hire transportation services he went to the city gates to obtain it.44 The clustering of transport properties near the city gates demonstrates the density of wheeled vehicles expected to be and (since they were the only way in or out of the city) necessarily found at the gates.45 Literary and epigraphic evidence for the 42 Martin (2002: 156–7). Translation by Casson (1990: 195). Cf. also Meyer and van Nijf (1992: 127–8). 44 Casson (1994: 179–80); MacMullen (1974: 70, 72). 45 Kleberg (1957: 50) noticed this concentration but with inns generally, not ramps specifically. Cf. also DeFelice (2007: 478). 43 Comp. by: PG2846 Stage : Proof ChapterID: 0001306139 Time:16:21:14 Filepath:d:/womat-filecopy/0001306139.3D Date:17/6/11 OUP UNCORRECTED PROOF – FIRST PROOF, 17/6/2011, SPi 210 Eric E. Poehler concentration of transport services at the city gates support the archaeological data. Characters in Apuleius and Plautus, for example, find lodging very near to the gates amongst a number of other commercial properties.46 Inscriptions dedicated by members of the mule-drivers’ association (collegium iumentaio) also place cart drivers near the gates of Roman cities.47 These properties, however, operated mainly as part of an economy geared towards providing for those who were already travelling. Therefore, we must distinguish between the services of transportation and the services for transportation even if these were undoubtedly intertwined in the business of running an inn. Paradoxically, it is in this distinction and also in this place (and at bakeries) that the Household and Commercial Modes of transport can be seen to intersect. For example, had Cato’s men come to Pompeii to pick up the desired mill (Household Mode), they might have taken lodgings in one of the inns near the city gate, not only for their comfort but also for the security of their animals, cart, and cargo.48 Or, if Cato had chosen to have the mill delivered from Pompeii (Commercial Mode), the vehicles, animals, and slaves might have been contracted though the manager of the same inn. Finally, when not in use for the Household Mode of transport, the cart and mules from the Casa del Menandro (for example) might have been posted near the Porta Stabia for their profitable use in the Commercial Mode of transport.49 The economic opportunity to cater to travellers as they entered and left the city not only determined where inns with ramps would be located at Pompeii, but also influenced the form of the building and kinds of services offered. The inn at I.1.4–6 provides a typical example (Fig. 8.6).50 A broad, Paved Ramp (Fig. 8.1) leads from the street into a wide corridor (A) before opening into a courtyard for vehicle storage (B). As the carts moved through the corridor, their hubs hit the masonry piers, leaving a characteristic rounding as evidence of their passage (Fig. 8.7).51 These brick piers supported an upper storey accessed by an internal staircase, but in several examples an external stairway is found leading from the street, a fact that suggests a separate apartment might have been operated by the inn.52 Such an interpretation is bolstered by the location of the external stairway. It is often placed between the ramp entrance of the inn and the entrance to a bar (L) and/or shop (K) connected to 46 47 Plaut. Pseud. 658; Apul. Met. 1.21. CIL V.5872 (Mediolanum). On the liabilities in transportation, cf. Martin (2002: 159 nn. 39 and 40). 49 Martin (2002: 154–8). This is an alternate version of Maiuri’s explanation for the absence of the draft animals in the Casa del Menandro’s stable area. 50 DeFelice (2007: 478) describes the inn at VI.1.2–4 for its similar typicality, while Packer (1978: 6–9) does the same for I.1.6–9. 51 The rounding begins at approximately 50 cm, which matches the axle height of the carts found in the Casa del Menandro (see above) and the Villa di Arianna at Stabiae. On the later cart see Miniero (1987). For another example of the rounding from wheel hubs as proxy evidence for particular types of movement, see Newsome (2009a) on the Casa del Marinaio (VII.15.1–2). 52 These are I.1.7, VI.1.3, and VIII.1.2. 48 Comp. by: PG2846 Stage : Proof ChapterID: 0001306139 Time:16:21:14 Filepath:d:/womat-filecopy/0001306139.3D Date:17/6/11 OUP UNCORRECTED PROOF – FIRST PROOF, 17/6/2011, SPi The Economics of Transport in Pompeii 211 Fig. 8.6. Plan of the inn at I.1.3–5 (image by Eric E. Poehler). the inn. These businesses not only communicated internally with the inn, but also opened wide doorways on the street to attract customers not accommodated by the inn.53 The additional commercial enterprises associated with inns demonstrate a sophisticated and diversified business strategy to attract customers whether on wheels or on foot. CONCLUSIONS The Household and Commercial Modes of transport identified in the archaeological remains at Pompeii add an intention to the circulation of traffic that is absent in the articulation of the traffic system itself. By adding and defining destinations (established through the presence of ramps and stable areas) it becomes possible to distinguish the internal supply of large households from the flow of traffic on commercial errands. Conversely, the complexity of individual properties (e.g., the Casa del Menandro) and their (discernable) 53 All inns with ramps except one (VI.1.1) have associated retail properties. Comp. by: PG2846 Stage : Proof ChapterID: 0001306139 Time:16:21:15 Filepath:d:/womat-filecopy/0001306139.3D Date:17/6/11 OUP UNCORRECTED PROOF – FIRST PROOF, 17/6/2011, SPi 212 Eric E. Poehler Fig. 8.7. Rounding on brick pier in I.1.4 (photo by Eric E. Poehler with permission of the Soprintendenza Archeologica di Pompei). urban investment strategies demonstrate that these were overlapping rather than opposing sectors of the transport economy. Indeed, the large inn at VI.1.2–4, immediately inside the Porta Ercolano, was once connected to the Casa delle Vestali and previously served as its stable area. Once it became more Comp. by: PG2846 Stage : Proof ChapterID: 0001306139 Time:16:21:17 Filepath:d:/womat-filecopy/0001306139.3D Date:17/6/11 OUP UNCORRECTED PROOF – FIRST PROOF, 17/6/2011, SPi The Economics of Transport in Pompeii 213 profitable to commercialize that space, the area was closed off and a new ramp was built at the rear of the house.54 It is even tempting to link the assumed profits of this change of operation to the Commercial Mode with the lavish decor of the following phase of the house. Regardless, the wheeled access at the rear of the Casa delle Vestali maintained its connection to the Household Mode, the efficiency of which was crucial in local transportation to the internal integration of the regional market around Pompeii. Similarly, the cost-effectiveness of Commercial Mode over relatively longer distances permitted time and price difference to matter, a fact that served to tie different regional economies together. Thus, Cato’s mill would have cost only 15 per cent more by having it delivered from Pompeii than fetching it himself from Suessa,55 a margin more than made up for in the production of his men and animals. The economics of transportation at Pompeii is perhaps best summed up in an example. The fifth doorway on Via dell’Abbondanza west of the Porta di Sarno is accessed by the broad Paved Ramp leading into the Praedia Iuliae Felicis. The ramp announced its stabling services while the rest of the frontage invited travellers with its high curbs, projecting stairway, and monumentalizing architecture. An advertisement on the façade detailed the amenities to investors and visitors alike: To rent for the period of five years from the thirteenth day of next August to the thirteenth day of the sixth August thereafter, the Venus Bath fitted for the well-todo, shops with living quarters over the shops, apartments on the second floor located in the building of Julia Felix, daughter of Spurius.56 The commercial investment in this building was enormous in scale and complexity: bigger than the Central Baths in area, the property joined two formerly separate insulae and completely suppressed a pre-existing street.57 The speed of the property’s construction following the earthquake(s) of 62 ce and its advertisement for rent before the eruption seventeen years later attests to the expectation of substantial profit from the traffic that would pass by this location due to its proximity to the amphitheatre and two city gates.58 The size of the building, its well-appointed rooms, and the quality of the services it 54 Jones and Robinson (2007: 398–9). Laurence (1998b: 127). CIL IV.1136, ‘In praedis iuliae SP F felicis locantur balneum venerium et nongentum, tabernae, pergulae, cenacula ex idibus aug primis in idus aug sextas, annos continuos quinque. S Q D L E N C.’ Translation by Bernstein (2007: 529). 57 Evidence for the previous flow of traffic on this street still exists on the former corner curbstones remaining in the curb. The street has also been detected in excavations within the praedia (Parslow 1998: 203–6). 58 Parslow (1998: 206). 55 56 Comp. by: PG2846 Stage : Proof ChapterID: 0001306139 Time:16:21:17 Filepath:d:/womat-filecopy/0001306139.3D Date:17/6/11 OUP UNCORRECTED PROOF – FIRST PROOF, 17/6/2011, SPi 214 Eric E. Poehler offered, however, belies the similarities in location and function that the praedia shares with humbler hospitality properties at the other gates. At the same time, the shop (II.4.1), baths (II.4.6), bars and restaurant (II.4.5, 7), the inn and its garden (II.4.2, 3, 8–12) all required a substantial flow of materials themselves. In fact, the stable area where visitors might have parked their vehicles was directly connected to the furnace for the baths, which was in need of fuel and ash removal. The design of the Praedia Iuliae Felicis reveals these combined concerns for the property’s internal supply and for the attraction of both local (baths) and non-local (inn) traffic. The advertisement shows a similar connection, even conflation in the progression of commercial and residential activities described: from commercial baths to ‘shops with living quarters’ to second-floor apartments. There is also a commensurate progression in the amount of supply each would need and the amount of traffic it would receive. Thus, the bath area was equipped with a ramp and stable, the shops were given wide doorways, and private apartments were relegated to the upper floors. Within the singular design of the Praedia Iuliae Felicis, the Household and Commercial Modes of Transport intersected to the point of ambiguity. Such uncertainty, however, underscores the interdependence of these modes and the ease with which Pompeians could move between them. Moreover, the decision to invest in transport related services generally (e.g., inns and taverns) and infrastructures specifically (e.g., ramps and stables) shows the importance of this sector of the overall economy and establishes it as a particular driving force in the development of the urban landscape.