Finds of imported Mediterranean pottery of 5th- and 6th-century date have been seen as directly connected to the formation and maintenance of secular, hierarchical power structures in post-Roman Britain (Harris 2003, 147; Campbell and Bowles 2009, 301). The imported wares have been identified at a number of British and Irish sites thought to be centres of local political control such as South Cadbury and Cadbury Congresbury in Somerset, Dinas Powys in Glamorgan and Garranes in Co. Cork (Campbell 2007, 62,138). Tintagel, despite earlier interpretations as a religious site, is now also interpreted as a centre of high-status, political control (Barrowman et.al. 2007, 335). Other sites with imported pottery have been interpreted as seasonal 'beachmarket' or trading centres, although Bantham has recently been described as a 'port' (Reed et al. 2011, 132), while structures excavated at Mothecombe in south Devon provide evidence of 'long-standing settlement' (Agate et al. 2012, 390) (Figs 4 and 5). Sites further from the main focus of distribution in south-west Britain, or those with smaller quantities of pottery, may represent secondary redistribution systems within Britain and Ireland (Campbell 2007, 138) connected to political or ecclesiastical networks (Harris 2003, 147).
Many aspects of the relationship between this pottery and other imported material (including glass) and their significance to post-Roman economic and political systems remain to be fully established, particularly the exchange of commodities that these finds might represent. Demand for minerals, specifically tin, has been typically seen as the driving force of this exchange (Radford 1956, 59; Campbell 2007, 138), although some have proposed an underlying political or diplomatic function to this exchange system rather than a purely commercial basis (Harris 2003, 152). Olive oil and wine have been proposed as the potential contents of the LRA1 and LRA2 amphorae, although wine seems increasingly likely (University of Southampton 2005; Pieri 2005, 85, 93; Campbell 2007, 24). LRA4, specifically, has associations with fine wine from Gaza (Pieri 2007, 152), although this type is noted by Campbell as rare within British post-Roman imports (Campbell 2007, 22). However, the amphorae and finewares can potentially be considered as proxies for other commodities such as grain. Paul Reynolds has suggested that the presence of ARS without accompanying North African amphorae at sites on the Atlantic Seaboard and in Britain might indicate that these regions were receiving grain shipments (Reynolds 2010, 111).
In spite of a long period of data collection and extensive publications on the Mediterranean imported pottery, interpretations have remained relatively static, particularly in relation to the chronology and logistics of its arrival. Radford's 1956 publication laid the foundations for these interpretations, suggesting the mid-5th century or later for the beginning of this exchange. The distribution of these wares is described as 'western and exclusive of Roman Britain' and belonging 'to an age when trade was once more flowing along the Atlantic Seaways' (Radford 1956, 67). Thomas' 1959 article describes pottery 'being brought directly by sea from the Byzantine world through the Straits of Gibraltar' (Thomas 1959, 105). The mechanisms for the importation of this pottery are formalised in Michael Fulford's 1989 article; direct shipment from the east Mediterranean is proposed, which is used to demonstrate direct contact between parts of Britain and the Byzantine world in the period c.AD 475-550 (Fulford 1989). He suggests that any Tunisian material might have been collected en route (Fulford 1989, 4). Campbell's synthesis presented a refinement of this model into two non-exclusive phases of Aegean imports c.AD 475-525 and African imports c.AD 525-550, again based on direct, though not 'non-stop', shipments (Campbell 2007, 26, 138).
The model of direct shipment to Britain has a dual foundation – the relative scarcity of comparable pottery on the Atlantic Seaboard, and observations of the unique composition of the British assemblage. In particular, the higher proportion of east Mediterranean ceramics (amphorae and LRC) to North African products (amphorae and ARS) in British assemblages is seen as distinctively different from the pattern in the west Mediterranean (Fulford 1989, 3). This is taken to indicate that the shipments reaching Britain originated in the Byzantine east and were not redistributed from the west Mediterranean, while the smaller quantities of imported DSP argue against a model of redistribution through Gaul (Fulford 1989, 3). The specific and consistent character of the British assemblages is therefore used to argue for Britain being a 'deliberate objective' of east Mediterranean shipments as part of a wider expansion of east Mediterranean trade in the later 5th century; a model reinforced by reference to contemporary texts and epigraphic evidence (Fulford 1989, 4-5).
Variations on this argument have been proposed, principally Wooding's 'tramp-steamer' model, which, nevertheless, argued for shipments of east Mediterranean origin taking on additional cargo further west (Wooding 1996, 15). Although Wooding incorporated the 'scattered' finds of imports in Atlantic Portugal and Spain into his model, the apparent absence of comparable Mediterranean imports in western France (specifically LRA2 and LRC at urban sites such as Bordeaux) led him to conclude that that shipments did not land between 'Iberia and Cornwall' (Wooding 1996; 41-3) For Campbell, however, the 'coherence' of the 'Aegean package' of imports argued against a model of 'tramp-steaming' and instead for a model of direct transport from the two respective Mediterranean sources (Campbell 2007, 128).
It must be noted that Fulford raised the possibility of future discoveries on the Atlantic Seaboard, highlighting two isolated sites with late Mediterranean imports, specifically Phocaean Red Slip Ware/LRC at Conimbriga in Portugal, and a sherd of Late Roman 1 amphora from Brittany (Fulford 1989, 3). Likewise, Campbell's 2007 published database includes isolated finds of Mediterranean pottery at sites in western France, while the accompanying monograph presents a map of the distribution of Phocaean Red Slip Ware (LRC) across the western Mediterranean and Atlantic Seaboard (Campbell 2007, 16). In general, however, the limited analysis of comparative data from sites on the Atlantic Seaboard has left the British finds to be largely examined in isolation, which, in turn, has reinforced the apparent exceptional character of the British assemblage. Recent publications from the Continent, particularly on sites in north-west Spain and south-western France, have, however, offered new information on the supply of late Roman pottery to the Atlantic region. These emerging data provide a new opportunity to question these established models, and as a result, to examine patterns of trade or exchange to Britain through the 4th, 5th and 6th centuries.