Locations, Peoples, Greek Pottery

This chapter focuses on the coastal area between two Greek colonies, Massalia (modern Marseille) in southeastern France and Emporion (modern Empúries or Ampurias) in northeastern Spain. The 64 sites in these regions accounted for 12,391 vases or fragments of vases (or almost 52% of the total number in the database). It is unknown when traders of Greek goods first started coming to this area; they had certainly arrived by the end of the sixth century BC, shortly before Massalia’s foundation. None of the Greek pottery cataloged for the province of Girona in Spain and the départements of Pyrenées-Orientales, Aude, Hérault, Gard, and Bouches-du-Rhône has a median production date earlier than 650, and only 96 pieces date to the half-century between 650 and 600 (figure 9.1).³ The presence of Etruscan and Phoenician goods indicates that there was probably competition between traders from different backgrounds in this area. With the foundation of Massalia around 600 and Emporion around 580–50, however, the consumption of Greek pottery grew dramatically, with 2,471 examples dating to between 600 and 550. Similarly high numbers of exports persisted throughout the sixth and fifth centuries, reaching a peak of 3,567 produced between 400 and 350, before the market for Greek imports seems to have collapsed. Only 265 examples in the catalog were produced between 350 and 300. It is not entirely clear what indigenous groups offered in compensation for the goods they consumed. Agricultural produce was probably an important component of the trade. Salt, readily available in the pans and lagoons of the Rhône delta region, was likely also a valuable commodity (Bouloumié 1984).

Figure 9.1. Number of imported Greek vases found in coastal northeastern Spain and southern France, by median production date, 650–300 BCE.

The identities of the non-Greek inhabitants of this coastal region also require clarification if we are to examine their relevance for understanding the distribution of vases. For our purposes, identity will be defined as claimed or real membership in a group. We do not have direct access to non-Greeks’ characterization of their own ethnic identity because we lack any firsthand accounts that would elucidate it. Only Greek and Roman descriptions survive, and they are problematic. Greek and Roman writers identified several different ethnic groups in the area covered by the study, particularly Celts and Iberians. The issues associated with interpreting the treatment of foreigners in ancient literature include three points: first, such accounts are often self-justifying and biased in other ways toward the authors’ own cultures and against those of the foreigners; second, they are frequently based on second- or third-hand reports of interactions with foreigners; third, the date of their composition is often centuries later than the events they purport to describe (Antonaccio 2003, 2005, 2009; Dietler 2010; Hodos 2006, 2009). It is not clear, for example, how accounts that define or describe ancient cultural groups reflect the actual ways in which people attributed to those groups would have identified themselves and their similarities to or differences from other groups. Did Celts—whose name, in any case, seems to have belonged only to a group in southeastern France but was applied by Greeks to people living from Central Europe to Provence and perhaps even farther west and south—think of themselves as belonging to a unified group, or were there distinctions (subtle or otherwise) made on a regional or local level?

With regard to the archaeological record, answers to this question have so far been mixed. Archaeologists have identified a cultural group spanning Central and Western Europe between roughly 800 and 450, referred to by the name of an Austrian type-site, Hallstatt. The Hallstatt culture was succeeded by a related one, named for the Swiss site La Téne, which persisted until the Roman conquests in the first century BCE. Hallstatt culture was characterized by a highly stratified society centered on hilltop settlements (the so-called princely sites, or Fürstensitzen), as well as prominent burials containing rich arrays of grave goods and marked by tumuli. The best-known and most extensively studied sites are Heuneburg in southwest Germany (Bittel and Rieth 1951; Kimmig 1983; Kimmig and Böhr 2000) and Vix (Mont Lassois) in eastern France (Chaume and Brun 1997; Chaume, Mordant, and Allag 2011; Chaume, Nieszery, and Reinhard 2012; Chaume, Olivier, and Reinhard 2000; Chaume, Reinhard, and Nieszery 2007). The La Téne period seems to have resulted from a collapse of Hallstatt sites, perhaps because of social strife. It was characterized by fortified towns still identified by the word Roman writers used to describe them: oppida. In general, scholars have equated these two cultures with the “Celts” first named by the Greek historians Hekataios (Jacoby 1923–98 1:F 54–55) and Herodotos (2.34, 4.48), although the precise range of their settlements is not clear. Hallstatt sites have been found in Galicia and northern Portugal, and several scholars have noted affinities between elements of Hallstatt culture such as jewelry and local products found as far away from Austria as El Carambolo, outside Seville in southwestern Spain, leading to the theorization of a hybrid group of “Celt-Iberians” (Almagro-Gorbea 1997; Lenerz-de Wilde 1995).⁴ This formulation draws attention to another problem in understanding ethnic groups from literary sources: discerning whether similarities in the material record reflect cultural influence or cultural connections. The same questions also exist for the other major ethnic group mentioned in this region: the Iberians (whose name comes from the one given by ancient authors for the Ebro River; Almagro-Gorbea 1988; Buxó and Pons i Brun 2000; Rouillard 1991; Sanmartí-Grego 1992; Vives-Ferrándiz 2010).

Similar to the Hallstatt and La Téne cultures, the material culture throughout most of the Iberian Peninsula, especially the eastern half, was socially stratified and controlled from a few major centers, such as Ullastret in Catalonia and Cástulo in Andalusia. These centers controlled territories of up to 2,000–3,000 km² (larger than most Greek city-states) through a network of smaller settlements ranging from individual farmsteads to small towns scattered across the landscape (Plana i Mallart and Martín Ortega 2001; Sanmartí and Santacana 2005). Just as the Hallstatt/La Téne cultures have been connected to the “Celts” described in literature, so, too, the culture of eastern Iberia has been linked with the literary “Iberians.” Connecting the material record to ethnicity has been a matter of some debate in studies of the Greek world, though, and so the opportunity to explore a possible relationship between the archaeological evidence and the concept of ethnicity is an important one (Antonaccio 2001, 2005, 2009; Hall 2000, 2005; Ruby 2006).

It is perhaps already clear that a distinction between Celts and Iberians is not easy to make. For example, where did the territory of one begin and the other end? In the late first century BCE, Strabo wrote that “historians of former times, it is said, gave the name of Iberia to all the land beyond the Rhodanos (Rhône) and that isthmus between the Galatic gulfs [i.e., the modern Iberian Peninsula]” (Jacoby 1923–98 3.4.19, trans. Jones). This statement seems to indicate that Iberians controlled a significant area north of the Pyrenees, including the modern regions of Rousillon and Languedoc. The mountains are less of a barrier than they might seem, however. Rousillon was part of the Principality of Catalonia until 1659, and even today, many inhabitants of that area speak Catalan in addition to French. Was Strabo therefore correct in suggesting that, in some period prior to his writing, a boundary did exist between non-Greek populations in the region between Massalia and Emporion? If such a boundary can be shown to have existed, where was it located? More specifically for the purposes of this chapter, to what extent can the consumption of imported pottery in large quantities (since the acquisition of these objects involved a clear act of choice by consumers) be used as a sign of ethnic boundaries? Some evidence collected over the last two decades already suggests that Iberian populations lived north and east of the Pyrenees (Dietler 1997, 2010:80). One sign of a difference may be the use of different alphabets in inscriptions (Untermann 1992). Material differences seem to have existed as well, at least early on. Local ceramic assemblages were different at the transition from the Bronze Age to the Early Iron Age in the seventh and sixth centuries, composed of two sets of material culture: to the east the so-called Suspendian type, similar to those of the Iberian Peninsula, and to the west the Mailhac II/Grand Bassin II type, which resembled Hallstatt styles (Dietler 1997:277; Garcia 1993:316; Py 1992:29–30). In the cases of both the writing systems and the cultural assemblages, however, the dividing line does not seem to have been found at the Rhône River but rather to the west, at the Hérault. The effort that follows tries to show how imported pottery might clarify the situation.

Pottery Distributions

A kriging analysis of the total number of imported vessels (or fragments of vessels) in the area between Massalia and Emporion between 650 and 300 BCE shows several patterns (figure 9.2). First, the zone with the greatest quantity of Greek imports is, perhaps unsurprisingly, around the colony of Massalia itself. This area also stretches about 50 km west to include a major site where thousands of fragments have been found, Saint-Blaise (Bouloumié 1992). This settlement, adjacent to the mouth of the Rhône, was largely inhabited by an indigenous population, to judge from its architecture, and it seems to have been a significant point of interaction and exchange between Greeks and non-Greeks.

Figure 9.2. Kriging results for the distribution of Greek vases, 650–300 BCE.

The territory around Emporion also featured high numbers of Greek imports but in a much smaller area (in fact, the intensity of the kriging prediction for this area is almost entirely a result of Emporion’s very large quantity of published material, while its neighboring sites contributed very little). Somewhat high levels of consumption spread along the coast to the north before dropping off precisely at the Hérault River. The kriging prediction does not take topographic features into account, so the fact that the river coincides with one of the contour edges is not only a striking marker of different patterns of consumption but also possibly an independent marker of a boundary between groups. Along the coast of eastern Languedoc, between the Hérault and Rhône Rivers, there is a noticeable decline in the level of consumption of Greek imports, a pattern similar to what is found in the hinterland of western Languedoc and Rousillon, 50 km or more inland. At the same time the quantity of vessels was lower in this area, though, on average the range of functional types (drinking, eating, household, storage, and transport) present at sites in eastern Languedoc was richer than at sites elsewhere, particularly to the west and south. Using a statistical measure known as Simpson’s Index of Diversity, which calculates how evenly distributed individual vessels are across categories (the index ranges from 0 to 1; a higher number means a greater evenness), it can be shown that sites between the Hérault and Rhône tended to have higher index scores of between 0.5 and 0.6.⁵ West of the Hérault, by contrast, sites showed less diversity, with scores ranging from 0.29 to 0.39. In other words, despite lower overall consumption of Greek imports in eastern Languedoc, people tended to buy a broader range of functional types of vessels when compared with their neighbors, either to the east in Saint-Blaise and Massalia or to the west.

A closer look at certain aspects of the pottery distribution can be equally illuminating—for example, individual functional types. The consumption patterns associated with vessels used for drinking—cups, jugs, mixing vessels—were essentially identical to the ones already noted for the complete set of vessels, since drinking vessels formed 74 percent of the corpus. The divergence between western Languedoc and Rousillon, on one hand, and eastern Languedoc, on the other, was quite strong, with the eastern sites showing only one-tenth the predicted values for drinking vessels as their counterparts on the other side of the Hérault. The range of drinking shapes was slightly higher in western Languedoc, but the diversity of drinking assemblages was again higher to the east. The spread of one specific type of drinking vessel, the krater, may be of special importance for understanding local interest in Greek pottery. This very large and distinctive shape, used for mixing wine and water prior to drinking, was emblematic of the Greek symposium, as it often appeared in depictions of symposia. It may even be seen as symbolic of Greek-ness itself, since Greek authors claimed the dilution of wine was the epitome of civilized behavior, while the drinking of unmixed wine, by contrast, was a behavior attributed to uncivilized foreigners (in some literary discussions, unmixed wine was seen as dangerous or even deadly). When kraters are found in non-Greek contexts, it is not possible to be certain that they were used for mixing, particularly when they are found outside of domestic spaces (e.g., in tombs). In keeping with all the other kinds of Greek pottery except transport amphorai, consumption of kraters at non-Greek sites between Massalia and Emporion was concentrated in western Languedoc (figure 9.3). Almost 250 were found at one site, Nissan-lez-Ensérune. Only the Greek colony Emporion, with 549 kraters, revealed more examples. But the vessels that appeared in western Languedoc were concentrated at only a few major centers rather than evenly distributed across all towns. East of the Hérault, by contrast, relatively few kraters appeared outside of Saint-Blaise and Massalia.

Figure 9.3. Kriging results for the distribution of kraters (mixing vessels for wine), 650–300 BCE.

Where vessels used for eating (primarily bowls and, to a lesser extent, plates) are concerned, a somewhat different pattern emerges. In the corpus as a whole, eating vessels comprised only 14 percent of the total, and they were most associated with Greek colonies (figure 9.4). Few sites in eastern Languedoc acquired much pottery used for eating—only one had more than 14 examples (Saint-Gilles-du-Gard, with 46). West of the Hérault, too, there was little apparent widespread interest in Greek eating vessels—large quantities were found only at a few major centers, especially in close proximity to the coast. Four of these sites ranked highest in the entire corpus of 233 sites for their range of eating vessels, and another 4 sites fell among the top 13. In this case, the important western sites not only showed a wider range of eating vessels, but there is also no difference in the diversity of eating assemblages between sites on either side of the Hérault. The eating function therefore seems to have been the subject of a particular lack of emphasis or interest for consumers of Greek pottery in the east, while it was more important west of the Hérault than anywhere else in Western Europe except for Massalia, Saint-Blaise, and Emporion.

Figure 9.4. Kriging results for the distribution for Greek vessels for eating, 650–300 BCE.

Transport amphorai might be considered a particularly important functional category for understanding differences between groups because their presence is assumed by scholars to imply the presence of wine as well. Although it is far from clear that non-Greeks consumed wine the way Greeks did at their symposia—that is, with water mixed in large quantities for a group of men to consume over several hours while reclining, often in a highly ritualized fashion—it has been argued that wine was first introduced to this region by Greek traders.⁶ The number of amphorai might suggest the interest of non-Greeks in a new beverage for banquets that had previously included beer or mead. It is clear that in the seventh century and early sixth century, Etruscan amphorai were dominant in this region, but within a few generations of the foundation of Massalia, that city’s products became the most popular (it was the source for 41% of all transport amphorai in this survey). As the kriging map for transport vessels shows (figure 9.5), Saint-Blaise and Massalia were again the largest consumers, but now they were joined by several other sites nearby that had many amphorai, especially Beaucaire and Saint-Gilles-du-Gard. The number of amphorai declined sharply—seemingly almost as a function of their distance from Massalia—so that in southwestern Languedoc and Rousillon, there seem to have been few to none present at any site. Proximity to Emporion seems to have been unrelated to wine consumption, suggesting that importation of wine was not caused merely by contact with Greeks.

Figure 9.5. Kriging results for the distribution of Greek transport amphorai, 650–300 BCE.

Some consideration of the find-spots of Greek imports within a site is also worthwhile for ascertaining cultural differences between various regions. As noted, a vase would likely have had a very different meaning if found in a tomb, a sanctuary, or a house. In the period discussed here, when Greeks buried their dead, they might have placed a few small vases (especially in the oil-flask shape known as the lekythos) and occasionally cups, bowls, or plates inside or near a tomb. They rarely placed larger shapes or shapes associated with commensal banqueting (such as table amphorai or kraters) with the dead. At non-Greek sites in the south of France, practices seem to have been variable in these regards.⁷ In western Languedoc and Rousillon, Greek vases were placed in tombs in significant numbers at a few sites (Nissan-lez-Ensérune, with 515 vessels, and Pézenas, with 339 vessels comprising all the Greek vases found at the site), while at most other sites (such as Sigean [Pech Maho] and Mailhac), they were found only within settlements. Only 3 sites east of the Hérault showed vases in tombs; of these, 2 were located along the river, and only 1 had more than 1 vase from a grave context (Agde, which had just 4 vessels dating to the seventh century, prior to its settlement by Greeks).

Interpreting the Distribution Patterns

Pottery assemblages clearly demonstrate the agency of consumers to make choices about what goods to acquire based on locally determined value systems and interests (Cook, Yamin, and McCarthy 1996; Dietler 1999; Marshall and Maas 1997). Generally speaking, there were notably different patterns on either side of the Hérault River. For the most part, there was higher consumption of Greek imports to the west and lower consumption to the east. The higher interest in Greek wares was pervasive across almost all categories in western Languedoc and, to a lesser extent, Rousillon. There were variations in practice not only between regions but also within them. To the west, major centers were the biggest consumers of Greek pottery, where much less was found at smaller sites; to the east, there was less differentiation among sites. Moreover, non-Greeks west of the Hérault did not seem to be concerned about using Greek vessels exactly the way Greeks did—for example, occasionally placing large vessels associated with commensal banqueting in tombs. In perhaps the most striking pattern, to the east of the Hérault, particularly in the area close to the mouth of the Rhône, consumers appear to have been more interested in Greek wine than in the vessels Greeks used to drink it.

Conclusion

The use of a geostatistical technique, kriging, allows the interpretation of a rich data set from a geographical perspective. Kriging also allows the analyst to plot the amount of error associated with the prediction. In the case of imports of Greek pottery to southern France and northeastern Spain, clear variations can be identified between regions, and the associated prediction error is very low because of the large number of sampled sites in a relatively small area. The identified variance seems to offer an independent confirmation of the distinction made by previous scholars between two non-Greek cultural groups in these regions. In contradiction of Strabo’s claim that the Rhône River formed the boundary between Iberians and Celts, the most significant geographical feature that can be associated with the pottery distribution patterns is instead the Hérault River. Even in analyses where the predicted distribution contours do not precisely align with the river, they typically fall within 10 km of it. It is therefore possible, based on the array of evidence now assembled from literary and material sources, to identify the two cultural groups with the Iberians and the Celts, although some caution should be exercised in using this terminology, since cultural mixing seems to have been a frequent occurrence; in addition, there are problems associated with taking ancient reports of ethnonyms at face value.

With regard to this question of cultural identity, it is perhaps of some interest to compare consumption patterns in western Languedoc with the rest of the Iberian Peninsula. In fact, a wide variety of patterns can be identified throughout Iberia, depending on the period and location. For example, in the earliest days of contact and trade, in the eighth through sixth centuries BCE, most consumption of Greek goods by indigenous groups happened in the southwest of Spain, in the area identified by many scholars as ancient Tartessos, around modern Huelva. This region was legendary for its wealth, purportedly derived from silver production (Jacoby 1923–98 3.2.8, 3.2.11, 4.155). Elsewhere on the Iberian Peninsula, however, there was hardly any interest in Greek vases during this period, either from indigenes or from Phoenicians who had founded their own emporia. Following the foundation of Emporion, there was strong interest in Greek pottery at the nearby regional center of Ullastret but not elsewhere, until new mines opened in eastern Andalusia in the fifth century around the town of Cástulo. The surrounding province of Jaén became the largest importer of Greek pottery in Iberia through the fourth century. These general trends show that Iberians generally were not enamored of Greek pottery but that in situations where increased wealth could spur status competitions, consumption of imports was one way distinctions might be made between individuals (Walsh 2014). Such an explanation also seems appropriate for the concentration of Greek vases at the major centers of Ibero-Languedoc. Likewise, the consumption of wine in eastern Languedoc may have been associated with status competitions in arenas of commensality (Dietler 1997, 2005, 2010).

Beyond these specific conclusions, this chapter has also demonstrated how the richness of the publications regarding pottery from one region relative to those on other wares can be used to identify important patterns of indigenous behavior following local principles. The modern colonialist bias toward objects associated with ancient Greeks, the desire of people in southern France and northeastern Spain for those products, and perhaps especially the abundance of pottery as a category of artifacts recovered through excavation (resulting from the utility of fired clay for creating containers and the durability of ceramics over very long periods of time) have combined with a new perspective on the agency of consumers to allow the development of significant new insights into ancient behaviors. The acquisition of imported pottery must have been more costly—however that cost was reckoned, whether in coined money, non-currency resources, or some other capacity to make choices—than locally produced options, and thus it represented a kind of luxury.⁸ It therefore seems probable that the varying patterns identified are reflective of status competitions operating within the rules of the local cultural context. Greater consumption, demonstrated by a greater abundance of imports, thus likely indicates more intense competitions.