South Cadbury Environs Project
From the outset of its fieldwork it was intended that the substantial comparative database available in the form of Leslie Alcock’s ceramic archive ought to be supplemented by new local assemblages integrated with it. This has increased the range and depth of the assemblage. The series is currently in preparation, although work on the pre first millennium material is nearing completion. The latter first millennium and 1st century AD will be supplemented greatly by the integration of the pottery from extensive excavations of an enclosure and pits at Sigwells and from a small trench at The Moor, South Cadbury.
The Pre 1st millennium BC pottery is presented in chronological order below. This is followed by the 1st Millennium BC to 1st Century AD assemblage according to vessel form and not chronological. Pottery illustrations are by Colette Maxfield, Joleen O'Neill and Amanda Tabor.
Pre 1st millennium BC
Leslie Alcock recovered a small but significant assemblage of Early and Late Neolithic material. The former was from pits on the plateau and was of Windmill Hill type, the latter from under the inner rampart bank. The Windmill Hill affinity contrasts with the Hembury/Carn Brea style pottery found in pits at Milsoms Corner (see below). The Late Neolithic material (see below) fills a gap in the SCEP material record, which comprises only a very few isolated sherds.
Early Neolithic pottery from Milsoms Corner and Cadbury Castle
Late Neolithic sherds from Cadbury Castle; Early Bronze Age sherds from Milsoms Corner and Crissells Green; Middle Bronze Age sherds from Milsoms Corner and various test pits.
Middle Bronze Age pottery from Sigwells Trenches 8, 9, 10 and 19.
The most substantial and diagnostically valuable assemblage was recovered from a later Middle Bronze Age enclosure at Sigwells (see above), the majority of the sherds from a single pit. The vessels fit into the Deverel-Rimbury Globular Urn tradition, the closest published comparison being the group from Shearplace, Sydling St. Nicholas. The Sigwells material was found in association with casting mould fragments for Wilburton metal work.
1st millennium BC
Writing in 1972, Alcock was reluctant to discard invasionist theories which largely had been abandoned for a decade. His perspective was informed by the belief that he could identify a very distinctly varied pottery sequence extending from the Late Bronze Age and through five Iron Age phases. He considered that the decorative motifs of the Early Iron Age represented a stylistic rupture with indigenous potting traditions, whilst stressing that the Bronze Age/Iron Age transition itself was gradual. He went on to discern stylistic similarities to material from northern France, southern Germany and the Netherlands, while pointing out the similar geographical affinities for some of Cadbury's bronzes.
The authors of the English Heritage report published in 2000 sought to 'avoid the many pitfalls of a spurious chronological accuracy' (Barrett et al. 2000) but Alcock's more chronologically resolved classification (Alcock 1980) still has much to offer. Ann Woodward's contribution to the assemblage has been within a scheme developed at Hengistbury Head, Dorset and refined at Danebury but without clarification by local absolute dating she preferred to blur temporal boundaries.
Although the respective hillfort assemblages are distinct they share enough traits to allow chronological alignment. Comparisons of a particular jar type found at Danebury and Houghton Down, JB1.3, bears a close family resemblance to similarly named vessels assigned to Alcock's Cadbury 6 and to a lesser degree Cadbury 5. It is a class which Barry Cunliffe and his colleague Cynthia Poole have given a date range of 7th to 5th centuries BC (Danebury ceramic phase or cp1-3), close to Alcock's scheme but up to two centuries earlier than that preferred by Ann Woodward, who has moved the whole series forward in varying degrees. The Danebury Environs Programme chronology is supported by a substantial battery of carbon dates so, where correlation seems appropriate, SCEP has leaned towards it.
The most significant disagreement between the systems of Woodward and Alcock applies to the first centuries BC/AD. Alcock had observed a clear break between Cadbury 8, characterised by the highly decorated Glastonbury style and saucepan pots, and the Cadbury 9 Durotrigan wares, amongst which bead rim bowls and jars predominated. In retrospect Woodward noted that in the bank and associated contexts Cadbury 9 material always occurred with diagnostically Romano-British pottery or metalwork and, citing work in other parts of Somerset, she has asserted that Durotrigan vessels did not reach the Cadbury area until around the time of, or soon after, the Roman invasion. Drawing on the local archaeological literature, she observed that the Glastonbury style persisted until, but not beyond, the invasion. In contrast Alcock envisaged the arrival of 'quite alien'; Cadbury 9A Durotrigan ware as early as the first half of the first century BC and that there were two further discernible subdivisions up to a massacre episode in around AD 60. Both authors noted that whilst Cadbury 8 material dominated the pit assemblages on the plateau it was comparatively under-represented in the banks.
Early Bronze Age material has been sparse and, where diagnostic, has been in the shape of very small decorated Beaker sherds or soft, thin-walled sherds with oxidised exteriors and reduced interiors. The best examples were from a burial at Milsoms Corner comprising displaced sherds in the fills of a ditch cutting it. Ann Woodward has attributed more rustic sherds from a ring ditch at Crissells Green to the Early Neolithic but after careful consideration it has been considered appropriate to include them in the Early Bronze Age assemblage.
Some of the areas investigated by SCEP have provided a much sharper focus on a particular time precisely because they have been used less intensively or for a shorter period than Cadbury itself. Later phases have not degraded and destroyed earlier ones and particular ceramic ranges can be defined clearly. Even here there are qualifications: some of the material from test pits has been dated mainly by fabric and finish rather than the more reliable form. The problem is more acute during periods, or at sites, where there is little decoration on the pottery as quite simply there are fewer fragments with diagnostic potential. During the first millennium there are two phases where this is true. Alcock’s Cadbury 4 assemblage broadly equates with the so-called Post Deverel Rimbury or Late Bronze Age Plain Ware range.
JB type jars
In crude terms it is the filling of a chronological sandwich between the Globular Urn style pottery from Sigwells Middle Bronze Age enclosure and the Earlier Iron Age Cadbury 5 and 6 vessels, as well as some highly decorated material from Milsoms Corner and especially Sheep Slait, described in the sections below. The specialised character of these sites ought to sound a warning bell: to what extent is the chronology distorted by function? If we look at the material from Sigwells it is easy to see stylistic relationships with some vessel classes from Sheep Slait; the Plain Ware assemblage must overlap both.
JC to JF type jars and DA dish types.
There is also important evidence regarding the Late Iron Age pottery from an enclosure sequence and pits in the north west of Sigwells. Here the later pits and ditches contain a great deal of Dorset-sourced Durotrigan pottery but not a single diagnostically Romano-British sherd. Immediately south of the enclosure and pits, Romano-British and Durotrigan material were found in later banks bounding a trackway. The pottery in the pits and ditches is unlikely to have accumulated over a very short span of time. After they had been filled, several pits containing Dorset material, some accompanied by shell tempered Cadbury 8 sherds, were cut by later pits also containing it. The same was true of the two latest phases of the enclosure ditches, although a brooch from the uppermost fill of the latest ditch phase had currency in the late second to third quarters of the first century AD. It is hard to avoid the conclusion that a considerable amount of Durotrigan pottery was circulating well before the invasion. Alcock’s early first century BC date seems unlikely but its arrival in the late part of that century is quite conceivable. This is the chronological framework adopted by SCEP, see figures below.