Roman artefact studies tend to explore the tension between specific finds at a given site and their comparison to materials found at different sites and regions. This tension between similarity and difference is not uniquely linked to object types or material categories, functionality of deposits or nature of sites, and is not even specific to the Roman imperial period an sich. Moreover, such tensions surface in a variety of thematic debates, on form vs. function, globalisation, consumerism, class systems, technology, imperialism, cultural identity, gender, ethnicity, religion, nationalism and regionalism.
Figure 7.1 Coin minted at Sagalassos, during the reign of Valerian I. The reverse publicizes Sagalassos as friend and ally of the Romans as well as the first city of Pisidia. CNG electronic auction 169, 25 July 2007, 108.
This chapter addresses one variety of this tension: how regional differences in artefact assemblages are balanced against the integrative forces of the Roman commonwealth. For instance, the votive and funerary relief stelai of Phrygia dated to the later second to early fourth centuries AD are often considered parochial in style and content, reflecting rural Anatolia’s society. Their codified and repetitive corpus is considered to result from particular, regional predilections, which are unattested in contemporaneous urban contexts in western Asia Minor, to name but the most obvious contrasting tradition. Nevertheless, Jane Masséglia (2013) pointed out how the typical Phrygian focus on self-display in portrait and text, with an emphasis on personal industry, is actually a translation of social flexibility, successfully mixing Anatolian, Hellenistic and Roman cultural traits. These stelai exemplify how the local, while meaningful on its own, simultaneously evolved in reciprocal dialogue with more universal traditions.
At times, such mixing of cultural elements did not work. In discussing the late Hellenistic to early Roman imperial traditions of mould-made ceramic oil lamp making at Pergamon, Ephesos and Knidos, Anita Giuliani (2007) demonstrated how the early first century AD phase of lamp types mixing Hellenistic and Roman morphological traits was short-lived. Notwithstanding pre-existing distribution patterns and exchange mechanisms of these lamps in late Hellenistic times and the general demand for this type of objects serving all sectors of ancient daily life and death, in this case, mixed messages were no commercial success. The Hellenistic stylistic traditions remained separate from Roman Republican lamp repertoires, which hardly circulated in these regions. But even the early Roman imperial imported lamps would not affect local traditions. Only from the second quarter of the first century AD did local potters adopt the designs of Italian oil lamps and abandon the traditional Hellenistic models. This case study shows that the production of mundane objects could evolve in apparent disconnection from contemporary geo-political proceedings.
To be sure, examples of material culture translating active policies of mutual dependency are also available. The imperial Roman mints did not supply the cities and markets in the East with small coinage. Yet, local administrations wishing to fuel their economies with coin were generally granted the privilege by the imperial authorities to do so (Stroobants in press). Both local and imperial authorities had an interest in continuing these practices. The associated iconography makes this clear: messages of imperially-induced ideological unity are combined with local self-propagation in competition with peers (Fig. 7.1). Julie Dalaison (2014) recently presented an overview of how Pontic cities propagated their independence even in a context of firm dependency on Roman central authorities, in a region that had shown formidable resistance to Rome. In this way, all parties involved built on the complexities of inter-related realities, expounding different traditions, histories, claims and aspirations.
Many initiatives were also taken by private parties, which sometimes tried to boost local economic success by tapping into wider phenomena. On the basis of survey evidence from the Cide Archaeological Project, for instance, Philip Bes (2015a) proposed the presence of estates involved in probable vine cultivation in late Roman times. The local provenance of the related amphorae can be presumed based on the particularities of the clay fabric of the transport vessels. Their morphology, on the other hand, was very similar to the amphorae produced on estates related to Herakleia Pontike and Sinope, with documented distribution patterns around Black Sea, the Sea of Marmara and the Aegean Sea. The shape of the late Roman Cide region amphorae resulted from conscious decisions on behalf of estate owners or managers to try and blend into existing exchange patterns, based on regional produce from the coasts of northwestern Asia Minor.
Whether the artefacts under study are stelai, oil lamps, coins, or amphorae, meaning emerges from the tension between patterns of similarity and difference. These patterns seem to work in complementary ways and therefore form an essential part of how the Roman world worked, or at least materialised itself. Rather than enumerating more examples of this same phenomenon, this chapter engages in exploratory modelling. Based on the disciplines of archaeology and geography, we will consider whether the linguistic and socio-cultural concept of koine has pertinence to the economic concepts of opportunity cost and oligopoly, and finally whether comprehending the workings of these concepts affects our understanding of regionalism – the level at which many of the discussed tensions seem to play out.
In a 2011 paper, we argued for the re-introduction of an old term into the study of late Roman pottery tablewares (Poblome and Fırat 2011): that of ‘Late Roman D’ ware (henceforth LRD), as originally introduced by Frederick O. Waagé in 1948 in relation to material from Antioch-on-the-Orontes. LRD represents a recognisably similar way of designing and producing tableware shapes and fabrics shared by a range of documented production centres in southwestern Asia Minor and western Cyprus. The common language or koine of LRD highlights the existence of ‘a range of regional production centres involved in the making of a cohesive and consistent range of tableware types and forms’ (Poblome and Fırat 2011, 49). Recognising LRD as a koine places the ware on the same footing as the other contemporary types of tableware with wide circulation in the late Roman East – African red slip ware (henceforth ARSW) or Late Roman C ware (henceforth LRC), which is important for the general positioning of the ware. Acknowledging the internal cohesion of LRD as a stylistic product range, however, does not resolve issues related to attested fabric differences, their coupling to specific types linked to particular production centres, their archaeometrical provenancing to specific sites and the attribution of specific origins in LRD distribution patterns. We remain convinced that the re-introduction of LRD as a term and its recognition as a koine represents an opportunity for research, potentially making typologies more meaningful in their ancient contexts (Poblome and Fırat 2011, 54).
According to the Oxford English Dictionary (oed.com, accessed on 3 February 2016) the ancient Greek word κοινή is the feminine singular derived from κοινός for ‘common, ordinary’. Koine is defined as:
•Originally the common literary dialect of the Greeks (ἡ κοινὴ διάλεκτος) from the close of classical Attic to the Byzantine era. Now extended to include any language or dialect in regular use over a wide area in which different languages or dialects are, or were, in use locally.
•A set of cultural or other attributes common to various groups.
Building on this definition, the concept of koine implies that the commonness of the language or cultural attributes would be impossible to grasp if it were not for the existence of variations or differences at other operational levels. In this way, scale, context and tension between similarity and difference circumscribe each koine. Looking into the linguistic context of koine helps to understand how similarities and differences co-constitute each other.
As a linguistic phenomenon, koine does not exist but comes into being. The Hellenistic koine of Greek is traditionally cited as the original language koine. The First Athenian Sea League (477–404 BC) constituted the framework in which larger speech communities could emerge: providing the matrix for the mixing of dialects within the league’s territories, leading to phonological and morphological compromises and levelling between the dialects allowing the koine to settle. In its stable form, a koine is the ‘result of mixing of linguistic sub-systems such as regional or literary dialects. It usually serves as a lingua franca among speakers of the different contributing varieties and is characterised by a mixture of features of these varieties and most often by reduction or simplification in comparison’ (Siegel 1985, 363). It was the stabilised form of Greek, the Attic-Ionic koine, which was exported throughout the Hellenistic kingdoms leading to its adoption for literary purposes by the likes of Polybios, Plutarch, Flavius Josephus and Lucian (Bubenik 2010).
The problem with LRD is that it never existed an sich, in contrast to the lingua franca of Hellenistic koine Greek. LRD is useful as a meaningful level of classification, in the same way as ranges of pottery sherds are identified as ARSW or LRC tableware. In this sense LRD is a construct, albeit one that works, as it only takes little training to differentiate LRD from ARSW and LRC. As with ARSW and LRC, however, the common denominator of LRD koine represents a conglomerate of specific production centres, including Sagalassos; a site near Pednelissos (Jackson et al. 2012); and elsewhere in south-western Asia Minor and western Cyprus. Koine is not only an artefact of archaeological classification; it was also meaningful in antiquity, as the potters and customers wanted to make and use LRD-type of wares and not ARSW or LRC bowls and dishes. LRD represents a fashion, a taste, a design of particular ranges of products, which is consciously different from other wares and could bind people together, feeding the notion of regionalism.
The concept of koine helps explain why a classification that did not exist as such in the past, such as LRD, could nonetheless have been meaningful in antiquity. This paradox can be resolved by further considering the linguistic nature of koine, in particular through semiotic phenomenology as developed by Charles Sanders Peirce (1839–1914). According to the latter’s theory of signs, the process of semiosis which includes the production of meaning is realised through icons, indices and symbols. An icon is a sign inoculated on the ‘thing’ it refers to, signifying by resemblance. A Roman floor mosaic with a banquet scene retains some of the visual qualities of actual banquets. An index refers to things by participation, as in smoke being an index of fire, and signifies through cause and effect. Language, on the other hand, does not build on links to things nor stimuli, but on symbols, whose meaning must be learned, with language mediating the construction of meaning (Christidis 2010). That is why different words in different languages can mean the same thing, as these refer to a generalised abstraction of that thing. LRD as a koine of tableware does not refer to actual tableware, but to a generalised abstraction of it, which was meaningful to its users. LRD conformed to an underlying set of real objects, dishes and bowls projected to a more general idea of tableware, which was, in the minds of its users, sufficiently cohesive in design of forms, attributes and material qualities to be recognisable as LRD, and as different from ARSW or LRC.
If LRD was meaningful as a koine in antiquity, the question arises how to approach the actual tablewares constituting LRD. We take our cue from Sagalassos red slip ware (henceforth SRSW), one of the tablewares under the LRD umbrella. SRSW exists with its typological and chronological specificities (Poblome 1999), own path dependency (van der Enden, Poblome and Bes 2014), local chaînes opératoires (Murphy and Poblome 2012; Poblome 2016), embedded in local urban society (Poblome et al. 2013) and networked into its own economic framework (Willet and Poblome 2015). Taken together, these elements are what it takes to make SRSW into a ‘ware’ – the common denominator of ceramic analysis. In this sense, an individual potter’s workshop does not represent SRSW, but all SRSW workshops active in a given period do. At the same time, SRSW forms part of a larger commonality, LRD. Individual wares constituting a tableware koine can be considered as linguistic sub-systems or dialects, creating ‘a recognisable mix of vocabulary, pronunciation and grammar used by a particular group of speakers, who are regionally or socially connected’ (Poblome and Fırat 2011, 54). In ceramological terms, Michel Bonifay (2004) proposed the term faciès géographique to denote the real wares, within their regional zones of production and, when possible, the actual workshops or potters’ quarters.
As a matter of fact, there can be different faciès géographiques without there necessarily being a koine and their existence also does not entail an inevitable evolution towards a koine. Once a koine can be identified, however, it does presuppose the existence of different faciès géographiques, as well as dialectic relationships between these wares. Faciès géographiques and koine should be easy to define, based on tangible archaeological criteria and should combine etic and emic qualities. Etic criteria should best be defined based on workable practices of classification suited to the specifics of the artefacts under study; emic criteria ideally result from the analysis of the individual chaîne opératoire of each of the constituting wares (Read 2007). In that way, faciès géographiques and koine can be operationalised methodologically. As a result, both concepts are not synonymous with terms as culture, style, social identity, micro-regional interaction or macro-regional networking (Galanakis 2009), which remain problematic in definition and difficult to operationalise in archaeological terms, especially when trying to reach higher-level synthesis. Faciès géographiques and koine, in contrast, are not only within reach of archaeological methodologies, these concepts also derive their strength from the potential overlap of etic aspects of classification with emic ones related to past practices of production and use.
The LRD koine could in turn form part of other, larger stylistic frameworks of cultural pertinence. This could range from skeuomorphic inspiration from contemporary silver plate (Leader-Newby 2004; Willet 2012) to a globalised sharing of cultural symbols in the late Roman world. Miguel John Versluys (2015) recently demonstrated how the Roman globalised oikumene transformed the semantic system of the Hellenistic koine through processes of universalisation, in which styles and cultural elements lost part of their original meaning in order to play a role in a larger system; as well as particularisation, reconfiguring meaning in local contexts. When these larger systems share etic and emic analytical potential, these could also be labelled koine, but terminology preferably avoids redundancy.
A distinguishing feature of koine is that, whichever scale or context it represents, it is never static. Koine comes into being, develops and dissipates. Apart from the demonstrable cohesion in the development of form and design of LRD between the fourth and the seventh centuries AD, there might well be a shared origin (Özden 2015). A fill deposit containing various thousands of sherds of mostly tablewares was found sealed by a mortar layer during the 2006–10 excavations of Lot 159 within the western necropolis of the ancient town of Perge, extending from its West Gate (Fig. 7.2). Of the 1845 sherds studied, the majority conformed typologically to SRSW, whilst Cypriot sigillata, Eastern sigillata D (henceforth ESD) and Cypriot red slip ware (henceforth CRSW) were also represented albeit in markedly smaller quantities, as well as token presence of some other tablewares. Chronologically, most material was datable to the second half of the second and the first half of the third centuries AD, while the stratigraphy ran from the first to the seventh century AD. Macroscopically, the fabric and slip characteristics of the majority of material were comparable to SRSW, albeit containing qualitative variations which proved difficult to systemise macroscopically. This large group of material contained mostly SRSW types, but also some Cypriot and other types. A small group of sherds had different slip and clay fabric characteristics applied to both SRSW and Cypriot shapes.
XRF-analysis and thin sectioning applied to 22 samples of second/third centuries AD standard SRSW types of both macroscopic groups discriminated between three different compositional groups. Eight samples conformed entirely to all compositional characteristics of SRSW (Degryse and Poblome 2008). Eleven samples, macroscopically classified with the largest group of the material and representing traditional SRSW types, could not be traced to the Sagalassos clay beds, but were made from CaO rich clays of unknown provenance, albeit not from the region of Sagalassos (Figs. 7.2–3). Finally, three standard SRSW types were made from yet other clays, poor in MgO and containing medium-sized metamorphic rocks and minerals. These sherds conformed to the small group, distinguishable macroscopically by a more reddish hue of the fabric and a matt slip (Figs. 7.4–5). The archaeometrical composition of this last group was comparable to the fabric of an as yet unprovenanced pottery production centre in south-west Anatolia (Poblome et al. 2001), implying that it produced standard forms known from the SRSW, ESD and CRSW type series between the second and the seventh centuries AD. In its late antique phase the products from this centre have already been classified under the LRD umbrella (Poblome and Fırat 2011).
This evidence suggests at least mid-imperial origins for the phenomenon of the LRD koine. As LRD originates only in the course of the fourth century AD, the existence of an ESD koine is here suggested. As with LRD, the ESD koine highlights the existence of a range of regional production centres involved in the making of a consistent range of tableware types and forms located in the wider regions of southwestern Asia Minor and western Cyprus (Poblome and Fırat 2011, 49). Clearly, further research is required to elucidate the roles of the different production centres in these respective regions.
John Lund (2015) studied pottery circulation to throw light on the people and history of Cyprus – an important region in the koine discussed in this chapter – from the third century BC to the third century AD. Both Cyprus’ physical geography, with the Troodos mountain range structuring connectivity, and material culture bring the theme of archaeological regionality to the fore. Archaeology is rich in regional studies, be this driven by geographical aspects such as river valleys (Thonemann 2011), settlement patterns (Winther-Jacobsen and Summerer 2015), perceived historical cultural identities (Dörtlük et al. 2006), ancient regionality (Thonemann 2013), urban development (Raja 2012), or, as in the recent case of John Lund, the specifics of producing and consuming material culture, and the permeable boundaries this seems to indicate.
Lund concludes that the ‘spectrum of ceramic finewares and transport amphorae in Western Rough Cilicia and Western Cyprus was similar in many respects – but not identical’ (Lund 2015, 182). This cohesion between western Cyprus and south-western Asia Minor – our analysis reaches further west and northwest into Asia Minor – is linked to the importance of Nea Paphos as Cypriot metropolis and main port of entry for many goods into Cyprus (Lund 2015, 210), as well as the role of regional connections ‘in structuring maritime interaction and market formation’ (Leidwanger 2014, 33). Importantly, however, ‘the figures from Cyprus indicate … that pottery was primarily manufactured for local and regional consumption’ (Lund 2015, 217), implying that Cyprus participated in general socio-cultural terms in the globalised Roman world, but not in the strict economic sense of fully integrated markets. The importance of the big players on the tableware market, such as ESD, however, was to grow: ‘from the 1st century BC onwards, pottery consumers in the Eastern Mediterranean turned increasingly towards imported ceramic finewares with the consequence that local fineware producers were apparently forced out of business’ (Lund 2015, 220). In other words, momentum was building for the establishment of koine, such as ESD. The picture is again one of similarities and differences at different scales and times, which Lund tries to capture in the concept of archaeological regionality, defined by similarities in the production and exchange of different types of material culture. Of crucial importance is Lund’s argumentation against the traditional view that ascribes the apparent uniformity of Hellenistic and Roman culture on Cyprus as resulting from koine trends emerging in the Classical period. Instead, he reconstructs different regional compositions of material culture, relating pottery circulation patterning to burial practices and belief systems (Lund 2015, 230–6). The next question is whether we can move beyond archaeological and socio-cultural description in approaching these multi-scalar patterns of similarity and difference.
Figure 7.2 Sample 9. SRSW type 1B191, CaO rich group of non-Sagalassos provenance
Figure 7.3 Sample 15. SRSW type 1A150, CaO rich group of non-Sagalassos provenance.
Figure 7.4 Sample 20. SRSW type 1B190, MgO poor group of non-Sagalassos provenance.
Figure 7.5 Sample 28. ESD, Hayes form P40, MgO poor group.
What was the force behind the idea of LRD and its constituting wares? Why did potters not all opt to make ARSW dishes and bowls? Why did customers not all want only LRC? Why was a ware merging LRD, ARSW and LRC never made? How come that in the supposedly globalised Roman world most studied sites are characterised by discrete combinations of types of artefacts, or, the other way around, that distribution patterns of artefacts appear as multi-morphous sets expanding and contracting in time and space with little or no overlap? Social, political, cultural, religious, and even military factors can contribute to answering these questions, but when the distribution of goods is concerned, it makes sense to consider how the documented patterns could result from an economic circumscription of possibilities.
To be sure, eventually human beings and their intentions, relationships, interests, practices and decisions in changing time-space circumstances are behind the similarities and differences discussed in this chapter, but the nature of the archaeological record, especially in the Roman East, rarely allows us to come close to these. The available evidence induces abstraction. In his overview study of the chronological and geographical distribution of tableware in the Roman East based on the ICRATES-database, Philip Bes (2015b, 142–51) elucidated a couple of such mutually dependent abstractions, such as the notion of supply and demand, the nexus urban hub/productive countryside, pulling factors such as big cities, and connectivity.
The ARCHGLASS Project, coordinated by Patrick Degryse (2014), reminds of methodological differences between artefact types. Pottery is ultimately provenancable, if not analytically at least in descriptive terms. Few suppliers of raw glass in Syro-Palestina and Italy would deliver their semi-manufactured goods to the many secondary workshops throughout the Roman commonwealth who finished the objects. Sometimes recycled glass could be added to the mix, making inoculating morphological variety of finished products over compositional recipes a near impossible task. Considering the time-space frameworks involved, it would be hard to presume that each primary production centre would have been in direct contact with every one of its customers running a secondary glass workshop. The indirect nature of the attested exchange implies a range of networks in which information and goods were exchanged, as it were organically tapping into one another. Typical for pre-industrial economies is how these networks did not function as a free market economy nor as a highly centralised and institutionalised, hierarchical economy, but as something in between. Indeed, these networks were socially embedded relying on ‘relational trust, reciprocity, extra-legal sanctions, high commitment among parties and interdependence’ (Broekaert 2015, 147). In this sense, time, place, commodities, people and information became interconnected and interdependent, raising the possibility of degrees of connectivity/globalisation/market integration characteristic of the Roman economy (Pitts and Versluys 2015b; Morley 2015).
We tried to explore such basic economic market forces in two recent papers (Brughmans and Poblome 2016a; 2016b), based on the ICRATES-database (Bes and Poblome 2008). The papers respond to a need for more formal computational modelling in order to compare existing conceptual models of the Roman economy, and for evaluating their ability to explain patterns observed in archaeological data. Based on the distribution patterns of the four main types of sigillata tablewares in the Roman East (Eastern sigillata A, B, C and D), we combined exploratory analysis with computational modelling of hypothetical distribution mechanisms, applying agent-based modelling to examine the effects of different degrees of market integration in the Roman empire. The computational model simulated the structure of social networks between traders and thereby the flow of commercial information and goods. The results suggest that a high degree of market integration leads to generally widely distributed wares, while strong differences in the potential for large-scale production of tablewares resulted in variable distribution patterns among wares, similar to those observed in the archaeological record as collected in the ICRATES-database.
As a result, to a certain degree the economic rationale of market-functioning can also help explain the attested similarities and differences in Roman material culture. Can we actually describe that economic rationale? An earlier paper (Poblome 2013), concluded that pottery was both available and affordable in Roman markets, and that in those markets there was money to be earned for artisans. The concept of opportunity costs was introduced in order to translate each of these artisanal efforts as the value of the best alternative – but not chosen – employment of those means. In antiquity, the obvious alternative employment of means was agriculture in its widest sense. The opportunity costs of artisanal production to subsistence production ranged from very low to high depending on the scale of the combined output of the craftsmen. To be clear, we do not envisage opportunity costs to have determined or steered the socio-cultural aspects of material culture, discussed above. On the contrary, economic, social and cultural trends (along with other aspects not explored in this chapter) worked together to co-constitute material culture. But dependent on historical circumstances, the influence of these respective trends on the constitution of material culture need not always have been equal.
Production lines with mostly local distribution presumably represented very low opportunity costs to subsistence. At this level of production, typologies of artefacts could be idiosyncratic and only very loosely forming part of a koine. The spectrum of forms was fairly basic, with types possibly re-occurring in different fabrics. Connectivity was presumably low and the action radius town-countryside limited. In linguistic terms, sub-systems of regional dialects were only mixed to a limited degree. The locally produced wares from Boeotia can serve as a case in point. Local production can now be presumed or demonstrated at Thespiai, Koroneia, Tanagra and Askra. Each city’s ceramic repertoire presented both individual forms and traits, as well as a number of shared morphological and decorative characteristics (Fig. 7.6). Understanding the shapes and decorative styles is not straightforward, and the fact that we are dealing with surface survey material without much chronological granulation does not help. The nature and direction of associations with other wares, local or imported, cannot be elucidated, but as a whole the material does indicate that such processes played a role in the constitution of the local products. In any case such relationships, when they existed, did not necessarily follow linear paths, and can be traced geographically only in general morphological terms to south Italian and Greek/Aegean sources (Bes and Poblome in press).
If, however, artisanal production output proportionally increased, for instance as represented by SRSW (Willet and Poblome 2015), the opportunity costs rose through the greater loss of output of subsistence goods. To be clear, growth of the artisanal sector was not endless, but mediated by the lowest level of need for subsistence goods of the associated community and the connectivity resulting from the town-countryside nexus. Really big wares such as Eastern sigillata A were therefore logically dependent on large urban centres, as suggested by the agent-based modelling exercise, while the scale of output of each ware resulted from rational economic choices and behaviour. The higher the opportunity costs, the more successful distribution of the ware depended on the opportunities inherent in integrated markets, and the more typologies would cohere towards koine in order to appeal to more customers on more markets. By following the koine style, producers reduced the risk of their wares lacking appeal and not selling. The range of each koine is therefore dependent on its sustaining economic factors, such as the distribution of its inherent opportunity costs, connectivity of the associated production communities and the carrying capacity of the urban/rural framework within which opportunity costs were balanced out. In the minds of its producers and users, LRD referred to their notion of tableware, and this was simultaneously socially and culturally meaningful as explained above, as well as economically co-constituted.
In addition to the socio-cultural processes described above, the notion of oligopoly (Poblome 2013) can help explain, from an economic point of view, why typologies of individual wares constituting a koine cohere morphologically. Such typologies are mostly associated with production centres that aim at markets beyond their own locality. This is, for instance, the first century BC evolution John Lund sees in the field of tableware production in the eastern Mediterranean. The archaeological record of the Roman empire indicates that for most material categories typically only a few production centres developed an economic policy to integrate these higher opportunity costs, with concomitant large distribution reach. This is not only the case for the limited amount of widely distributed ceramic tablewares produced in the Roman East, but, as we saw, also for glass. Markets of this nature are considered to function within the framework of oligopoly.
Figure 7.6 Dish with off-set rim typical for Boeotian fabrics with attested production at Koroneia and Thespiai, and morphological parallels in ARSW, Eastern Sigillata B and Athenian products.
Within an oligopoly, the customers are many, preventing them from influencing pricing or the market individually or as a group. The suppliers are few, however, making strategic market behaviour possible, with suppliers needing to take each other’s strategic decisions into account. In this case, the policies and products of each supplier are influenced by those of the other suppliers in the same market … In this respect, the degree of competitiveness of oligopolitical markets can be situated between so-called perfect global competitive markets and monopolies. By definition, the ancient oligopolitical market strove towards stable conditions...with guaranteed sales for the few suppliers in the immediate known environment and satisfied customers with a product of a constant quality level (Poblome 2013, 92).
In this way, culturally defined koine and economically circumscribed oligopoly are really flip sides of the same coin. The dialectics between these fields hold the key to the attribution of meaning to patterns of similarity and difference, both in past practices and in present analysis.
This particular coin has three sides, however. The time-space framework in which meaning is constituted for Roman material culture needs to be considered as well. After all, an instituted economy such as the Roman imperial one is always also a political economy and hence involves spatio-temporal strategies of control and power aimed at establishing, maintaining or overturning oligopolistic situations. This discussion revolves not only around the political governance of oligopolistic production and distribution but alsp around the various spatial forms regionalism can take and its relation to socio-cultural expressions. In this final section of the chapter, we discuss political and competitive strategies of regionalisation which could explain the development of a koine in an oligopolistic situation, and reflect on the spatial form these could be expressed in.
To begin with, the spatial form of the relation between socio-cultural faciès géographique and koine will reflect the political organisation which, consciously or not, underpins oligopoly. Space is produced, contested, and transformed through a range of socio-political processes, strategies, and struggles, resulting in an institutionalisation of economic and cultural relations (Elden 2007; Harris and Alatout 2010; Allen 2016). Broadly speaking, the faciès géographique-koine nexus could be instituted as either nested hierarchical scales, relating to and fitting into each other like Russian dolls, or as interlocking networks co-determining each other in various directions.
The former interpretation of scale as nested hierarchies of territories, with regions occupying a particular level, implies that those further up the scalar ladder hold power over those further down the hierarchy by being able to constrain the latters’ activities (Brenner 2001; Marston et al. 2005; Collinge 2006). This is possibly the way in which most Roman emperors approached res publica, and imperialism by definition encapsulates such attitudes, but it is difficult to explain the similarity/difference issues of material culture within this framework. The other conceptualisation of power as operating in interlocking networks implies more complex scalarities in which parameters of scope and level are more spatially contingent (Allen 2003; Taylor et al. 2010; Agnew 2013). In this interpretation regions are overlapping and porous clusters of variable scope in a spatially discontinuous network (Derudder et al. 2003; Derudder and Taylor 2005; Allen and Cochrane 2007; Taylor et al. 2013). Clearly, this framework resonates more with how the discussed socio-cultural and economic concepts co-constituted meaning for material culture. In this respect, it is too often taken for granted ‘what centralized institutions are capable of bringing about at a distance’ (Allen 2004, 22). Territorial scales always have a network dimension to them in the sense that organisation and control are built upon infrastructures, allowing and directing the circulation of humans, things and ideas across space (Collinge 2006; Loopmans 2007; Harris and Alatout 2010). As a result, inequalities exist in the coverage of territories by more or less central powers, or in their particular policies on the ground, which, in the case of the Roman empire, can hardly be described as uniform in time/space/agency.
Importantly, a networked organisation of space is qualitatively different in the sense that it is relational and horizontal, instead of hierarchical and vertical. Relational spaces are spaces of mutuality, of sharing and trading, and hence are capable of stimulating agglomeration economies (Rosenthal and Strange 2004; Combes and Gobillon 2014), which could support connectivity and oligopolistic tendencies even without political control. In economic geography, the clustering of industries has been demonstrated to strengthen the competitive position of the actors involved in a variety of ways, apart from benefiting from local natural advantages such as the availability of natural resources. These advantages deriving from spatial proximity have been classified by Duranton and Puga (2004) as a) sharing of local infrastructure and facilities, input suppliers or workers with similar skills; b) matching between employers and employees (e.g. with specialised skills for specific industries), or buyers and suppliers (e.g. suppliers of specific product types catering for buyers with a specific cultural preference) and c) learning (e.g. about innovative or complex production technologies or business practices). The limits of the archaeological and historical record of the Roman empire do not allow to immediately list how these advantages would have worked in the case of LRD or any other koine of Roman material culture for that matter, but a wider comparative exercise amongst and between crafts should represent a promising avenue of research.
Indeed, whereas Classical Economic Geography studies have focused on territorial agglomeration (e.g. in certain cities or regions), more recent theorising about space and distance suggests that such agglomeration advantages do not necessarily rest on direct physical proximity, but rather on network proximity relying on infrastructural accessibility (Fujita and Thisse 2002; Ottaviano 2008; Yu et al. 2016). Infrastructures permit the circulation of goods, symbols and ideas in a spatially fragmented way (Graham and Marvin 2001). These infrastructures, which can be material, but also social or institutional (e.g. Simone 2010; Silver 2014), are increasingly regarded as vital mechanisms of clustering, but also of selection and exclusion (Graham and McFarlane 2015), and hence could function as tools to strengthen koine and oligopolies on the basis of agglomeration economies and political economic regulation.
In this way, debates in geography on the spatial form of regionalism can be instructive in an archaeological context to map the connection between political economic structure and the culture of everyday life. On the one hand, depending upon the political and economic processes underpinning oligopolistic production and distribution, the spatial distribution of material artefacts expressing a koine or faciès géographique can take different albeit mostly networked spatial forms. On the other hand, koine and oligopoly are potentially connected to spatial agglomeration and its benefits and drawbacks. The spatial distribution of artefacts is hypothesised to be crucially related to the material and social infrastructures through which these artefacts have circulated. Hence, the correlation (or not) of artefact distribution and the material relics of infrastructural networks can point to the specific spatio-temporal institutionalisation of the relation between society, culture, economy, and politics, and to how these co-constituted patterns of similarity and difference in material culture.
This research was supported by the Belgian Programme on Interuniversity Poles of Attraction, the Research Fund of the University of Leuven and the Research Foundation Flanders (FWO). We should like to thank the editors for providing structure and inspiration in setting up the workshop and bringing its results to publication. John Lund, Peter Talloen, Tom Brughmans, Rinse Willet, anonymous reviewers and especially the editors provided comments to drafts of this chapter, which very much helped in making the argumentation more inspiring.
* Departments of Archaeology and Geography, University of Leuven.
