Journal of Roman Pottery Studies: Volume 17

The import and distribution of eastern amphorae within the Rhine provinces

Tyler V. Franconi

Abstract

This paper examines the evidence for amphorae of eastern Mediterranean origin present in the area of the Rhine basin. The nature of the available sample is considered. Issues arising from both the potential impacts of research agendas on the types of sites subject to excavation, and previous quantification methods used by those reporting assemblages, and, indeed, an absence of quantification, are noted. Whilst representing only a fraction of the amphorae borne consumables arriving in the Rhine provinces, eastern amphorae nonetheless occur in a wide variety of forms, though three types are more frequent than others, with wine featuring as the most prominent content. Figures suggest the supply to the region was ‘even’ between the military and the civilian populations, or at least they had equal access to amphorae from these sources. This raises questions regarding the supply mechanisms. The early Roman period saw the greatest level of import, with numbers declining through the first century AD. Transport costs and the rise of regional and local production of wine probably account in large part for the decline, though the trend to gradual fall-off of long distance supply is seen more widely in the Roman amphora trade. Examples from the east occur in low numbers in the Late Roman period, and are likely to have had a rarity value, and are probably to be associated with elite consumers.

Keywords: The Rhine basin; archaeology; Roman period; amphora; trade; economy

1. Introduction

Despite an increasingly large amount of published material, synthetic and comparative studies of Roman amphorae assemblages have remained uncommon. This lack of synthesis has meant that amphorae, a valuable dataset for studying the Roman economy, have only been able to be utilized in an anecdotal manner to discuss ancient trade, relying on individual, well-known sites that may or may not be representative of wider patterns of production, transportation, and consumption. These well-known sites have been crucial to make evident the value of studying amphorae to better understand the ancient economy; the large number of data now available both require and enable a new approach. In order to advance the discussion, new methods must be developed to incorporate quantified ceramic data into discussions and models of the ancient economy (Brughmans and Poblome 2016, 406), and this necessitates synthetic studies at the regional or supra-regional level.

The data suitable for this sort of study are unevenly published across the territory of the Roman Empire, with some regions being much better studied than others. The Rhine provinces of northeastern Gaul and Germany, corresponding to parts of modern Switzerland, France, Germany, and the Netherlands, provide a large amount of published ceramic assemblages. A full discussion of every aspect of these assemblages would require a much longer publication, comparable to work done elsewhere, as in northern France (Laubenheimer and Marliere 2010), but a specific discussion of one category of amphorae, those originating from the eastern Roman Empire (here meaning the provinces from Greece to Egypt around the eastern Mediterranean), is explored in this paper as an example of the contribution that regional studies are able to make to an understanding of ancient trade, focusing on the types of products imported to the Rhine region, their geographical distribution within this area, and the chronology of these trade connections.

This study first discusses the history of the collection of these data and the problems inherent to discussing large archaeological datasets. It goes on to examine aspects of the eastern amphorae present along the Rhine: the forms, contents and origins, the chronology of importation, and the places of consumption. We can see a significant decline in the consumption of eastern products over the course of the first century AD, and this can be directly related to local economic developments in Gaul and Germany.

2. The study region and its history of ceramic research

The Rhine frontier stretched from the Alps to the North Sea and included the Roman provinces of Germania Superior, Germania Inferior, and Gallia Belgica. Initially conquered under Caesar, intensive development from c 30 BC onwards led to a densely settled region in which a substantial civilian population lived alongside the permanent military garrison of the Limes Germanicus. The Rhine and its tributaries created an integrated transport network in which traffic and communication moved easily across a relatively large and varied landscape, effectively linking the settlements of this region along the riverine corridor.

With over a century of high quality excavation and documentation of archaeological sites, the material record of this region is one of the best in the Roman world. The study of amphorae from sites along the Rhine dates back to the early 20th century when excavations at Haltern (Loeschke 1909), Hofheim (Ritterling 1913), Niederbieber (Oelmann 1914), and Alzey (Unverzagt 1916) created important ceramic typologies and chronologies. Later publications from Oberaden (Loeschke 1942), Windisch (Ettlinger and Simonett 1952), and Nijmegen (Stuart 1963) paved the way for the more recent work at Augst (Martin-Kilcher 1987; 1994) and Mainz (Ehmig 2003) as well as regional studies of eastern France (Baudoux 1996), northern France (Laubenheimer and Marliere 2010) and the central Rhine (Ehmig 2007a). In addition, numerous excavation reports have included sections on recovered amphorae as a matter of course.

To date, at least 65 sites in the region have published quantified reports on their amphora assemblages. The abundant data from the Rhine allow for detailed examinations of the distribution and consumption of different amphora-borne products within the region. The documentation of trade connections and their changes plays an important role in discerning the economic history of both the Rhine basin and those regions with which it was connected.

Amphorae from the eastern Mediterranean are generally quite rare along the Rhine and in neighbouring regions of northern Gaul and the relatively small dataset has already been the subject of several studies. Over a decade ago, the main forms found across Germania Inferior, Gallia Belgica, and southeastern Britain were documented by Vilvorder, Symonds and Rekk (2000), and the 11 types that they outlined can be expanded by material from the Rhine. Van den Berg (2012) produced unquantified distribution maps of ‘exotic’ amphorae forms found at Nijmegen, that now augment the distribution maps of Cretan (Marangou-Lerat 1995) and Levantine (Vipard 1995) amphorae, as well as those of other forms in Peacock and Williams (1986) such as the Camulodunum (Cam) 184, Cam 189, and Kingsholm 117. The recent documentation and distribution of amphorae from western Asia Minor (Papaioannou 2011) also must be considered. It is the goal of this study, therefore, to discuss the quantified evidence for eastern imports into the Rhine region over time and to discuss the wider economic implications of their trade and consumption.

3. Data quality

Before discussing the amphorae directly, it is important to note that archaeological data are what has recently been termed ‘characterful’ (Cooper and Green 2016, 271): they have been generated and assembled by different people in different places at different times with different methods, goals, and agendas in mind. This ‘character’ means that all ceramic assemblage reports were not created equal, and this is a major obstacle to the study of regional patterns of consumption. Different sites have used different collection policies, different typological classifications, and different quantification methods. Not all amphorae sherds may have been kept and recorded (see, for instance van der Werff 2004, 290), they may have been identified as different forms, and they may well have been counted differently. We must also remember that we only have data from where they have been researched and published, and thus the geographical distribution of sites presented in Fig. 1 is not a distribution of ancient trade or consumption, but rather where it has been documented. The three clusters evident in the Netherlands, the Middle Rhine, and around the Upper/High Rhine are artefacts of publications, not (at least not definitely) evidence of ancient realities.

These problems of comparing data are significant but not insurmountable. The character of archaeological material needs to be appreciated and understood, and then needs to be assessed for its usefulness. For the purposes of this study, where amphorae generally are uniformly identified within a set number of typologies and published with a set number of quantification styles, there is much to be gained by taking a synthetic approach. While there is nothing to be done about sites where work did not fully record or publish all of their finds, different typologies can be equated and qualitative data extracted, especially where drawings of amphorae are published.

The issue of quantification is more complicated. Most sites in the region have been published with a count of minimum number of individuals (MNI), though some are published with counts of rims, bases, and handles (RBH), and two (Dangstetten and Waldgirmes) are published as a maximum number of vessels present (Max) (Table 1). It is often not possible to overcome these differences when synthesizing material, and so each table presented here explicitly states the quantification method and, when amalgamating data by form or site type, keeps the data separated by the method of quantification. This means that we cannot produce very neat numbers for the entire region, but we can compare like with like, as is done below in Tables 2 and 3. In some cases where the publication of pottery has been thorough, as at Xanten where amphorae were published by total sherds, weight, EVE, spikes, and handles (Carreras Monfort 2006), the data can be converted into an estimation of MNI despite not being originally published as such, and I have done this where possible.

Figure 1. Sites in the Rhine provinces with published and quantified eastern amphorae (see Table 1 for details).

Table 1. Sites in the Rhine provinces with published and quantified eastern amphorae (EM: total of eastern amphorae from site; EM%: the percentage of each assemblage made up by eastern amphorae)

Thus while plenty of issues exist, they should not deter attempts to discuss ceramic consumption at regional scales. There is much un-realized potential in comparative studies that critically assess and present the possibilities and limitations of data (see Laubenheimer and Marlière 2010 for a similar exploration in northeastern France). In what follows, I examine how the relatively small dataset of eastern amphorae along the Rhine frontier can be compared and what this consideration demonstrates in terms of trade and economic life.

Table 2. Eastern amphorae forms present in quantified assemblages from Rhine provinces, with totals arranged by quantification method

Table 3. Eastern amphora forms by site type and quantification method

4. The data

Published and quantified data from a total of 65 sites located within the Rhine basin were collected into a database, of which 43 had documented eastern types. Since only quantified assemblages are used, many other sites, though excavated, are left out of this discussion since they lack comparable data. For this reason, major Roman cities like Cologne (now under study with preliminary results in Höpken 2013, with previous work on tituli picti by Ehmig 2007b and 2009) and Trier (excepting the Kaiserthermen assemblage considered here) are unfortunately poorly represented.

The full collection of all amphorae assemblages contains data from 38 forts, 22 settlements, and five villas, accounting for 41%, 58%, and 2% of the total respectively, and offering a significant cross-section of sites in the region. The total assemblage for the region shows the dominance of products from Spain (54%) and Gaul (22%), while Italy (6%), the eastern Mediterranean (4%), and North Africa (1%) make up the remainder. The products once carried in these containers can be divided approximately between olive oil (38%), wine (including defrutum) (30%), and fish products (19%), with small amounts of other products such as fruit, alum, and beer (14%).

As noted the 43 sites with eastern amphorae are made up of 23 forts, 18 settlements, and two villas (Fig. 1), though for the majority of eastern amphorae are present in settlements, that produced 57% of all the eastern material; forts produced a further 42%, and villas only the final 1%. The distribution of finds by site type is discussed further below. Many of these sites, especially the military sites, date to the Augustan or early Julio-Claudian period, which is largely the outcome of recent excavation and publication agendas within Germany and the surrounding region. The chronology of the importation of eastern evidence is also discussed below, but we must remember that the focus on early military sites arising from targeted research priorities has had an influence on the prominence of these earliest data.

5. The form repertoire

Three main amphorae types account for the majority of the total eastern amphorae documented in the Rhine provinces, with 14 other forms known in small quantities (see Table 2).

The Rhodian Cam 184 form is the most common eastern form recorded, accounting for an average of 51% of all quantified eastern imports to the Rhine basin. These wine amphorae are recorded at 30 sites in the database. They appear in the earliest phases of Roman occupation, being documented at Neuss, Basel, and Windisch, and continue into the mid-second century at Augst (Martin-Kilcher 1994, 490–2). Peacock’s suggestion that these amphorae were linked to a Claudian tax imposed upon Rhodes (Peacock 1977, 270) is no longer tenable as the sole explanation for their presence (Carreras Monfort 2006, 35) considering the strong representation of Rhodian amphorae present in earlier, Augustan,