There are numerous indicators which suggest that Welsh PRIA society shared traits common to the Celtic world at large, with power articulated through social hierarchies based upon tribute networks, whose basis will have been aristocratic lineage, status or prowess in war. Both economy and society, then, will have been centred upon the aristocratic household where status-building activities - feasting and the patronage of the arts - were focused. The recognition of such in the archaeological record, against a background of gradations in absolute ranking and clear differences in social organization over time as reflected in settlement archaeology, is a major challenge.
Recent research has suggested the existence of two basic socio-economic systems: an eastern redistributive economy in those zones dominated by large hill-forts, where agricultural surpluses were stored and exchanged, and a western clientage economy in those where hill-forts are few and where a dichotomy exists between producer farmsteads and consumer settlements of higher status (Cunliffe 1991: 394-8). Certainly the large, complex hill-forts of the Marches imply a much higher population density and one more amenable to manipulation and organization. A multifocal system may be envisaged here, with the biggest as the strongholds of the most powerful clans; and, though absolute power was fragmented, common cultural traits and lively exchange networks imply a strongly knit alliance of clans. Hidden social distinctions may lie within the settlement range of powerful hill-forts and producer farms, and within the structural range of individual settlements as exemplified at the Breiddin and Collfryn (Musson 1991). Defensive complexity and house dimensions at the latter (Figure 35.6) suggest an elite residence, and whilst the Breiddin apparently has smaller houses it has proportionately four times as many four-posters as Collfryn, suggesting a key role at least as a central repository of agricultural produce.
In south-west Wales Williams (1988) perceives two contrasting social systems: a redistributive system based upon the larger hill-forts of the north and east, with a clientage system in the inland south and west - certainly from the third century BC - based upon producer farms, and small promontory forts and ‘ring-forts’ which were high-status consumers of agricultural produce. If the latter were indeed residences of elite family groups, then their sheer numbers imply no regional political focus in the MPRIA and LPRIA and we may envisage a system of small clans who rarely acted in concert. However, though no central authority is visible in the settlement record or any other tangible indicator of PRIA political allegiance, it did exist by the time that the civitas Demetarum was in being by the later second century AD.
If conditions in Wales were seemingly inimical to the creation of centralized tribal communities on the southern or north-eastern English model, by the later second century AD Roman sources show that it was divided into a minimum of four tribal areas - Ordovices, Deceangli, Demetae and Silures (Jarrett and Mann 1969; Rivet and Smith 1979) - roughly approximating to the broad geographical divisions. Several writers have mused on the possibility that this political division was already reflected in the four main ‘cultural provinces’ of the LBA (Burgess 1980; Savory
1980) though Jones (1984: 33) warns against an uncritical acceptance of such views and writes, ‘The same geographical elements as helped to form the earlier cultural provinces by influencing the distribution of LBA metalwork are also likely to have influenced the later pattern of trade and settlement, and of social and political cohesion.’ The tribal distribution of the second century AD is likely to hold good at least for the immediate pre-Roman period, but although there is a broad consensus as to their geographical placement - the Ordovices excepted - recognizing these entities is difficult. They were most probably composed of clans who only came together at times of stress; instanced in the first century AD by the Roman threat. The distribution of ceramics, a traditional means of recognizing ethnicity or ‘identity-conscious interest groups’, is a less reliable guide to clan or tribal distribution in Wales for reasons stated above, though the chevron/'eyebrow’-decorated Lydney-Llanmelin pots (Figure 35.9) have a tight distribution which may conceivably define a ‘Silurian’ heartland (Figure 35.10b). Where tribal coinage impinges upon Wales we can attempt to be more specific. Both Manning (1981) and Sellwood (1984) define a region west of the lower Severn which did not share the monetary system of the Dobunni c.35 BC-AD 43 (Van Arsdell 1989) and must have been excluded from it (Figure 35.10a). Whether this formed part of Siluria or represented a border zone is speculative.
The recognition of the Silures and Demetae as civitates may either reflect their precocity or a higher level of political consciousness, and by implication the continuation of an essentially PRIA form of power structure. Furthermore, the Silurian and Demetian elites were sufficiently wealthy to create an urban focus, which in turn presupposes a substantial agricultural surplus and the possibility that such was also the case in the LPRIA. The absence of civitas development elsewhere in the principality may be a silent commentary upon the inability of the agrarian base in those areas to produce the necessary surplus, which together with geographical constraints which inhibited political co-operation fostered the continuance of highly segmented societies.
World History
