2005:043 - OLD FREEHOLD, Antrim

The site is set within a field of c. 2.3ha situated on the eastern edge of Agnew’s Hill, near Kilwaughter, at a height of c. 290m OD. The field’s potential archaeological significance came to light when it was ploughed, apparently for the first time, in early 2005. The landowner’s son noticed prehistoric pottery and worked flint disturbed by the plough. Following prompt reporting of the discoveries to the Environment and Heritage Service: Built Heritage, fieldwalking was carried out during March and April 2005. Fieldwalking identified four potential prehistoric sites, which were represented by scatters of worked flint and prehistoric pottery. The largest of these extended over an area 20m by 40m and formed the focus for Queen’s University’s undergraduate training excavation in June 2005. The fieldwalked finds from this site included a number of hollow scrapers, which are probably indicative of activity in the Middle Neolithic period, as well as a pottery assemblage dominated by Middle Neolithic types.

It was proposed that the fieldwalking exercise would be supplemented by an evaluative excavation intended to inform a management strategy for the site. Because of restrictions placed upon the excavation by both the size of the excavation team and the farmer, it was only possible to open up an area of 20m by 15m. In comparison to Early Neolithic houses, Middle Neolithic settlements are under-represented in the Irish archaeological record. This presumably reflects a normative architectural tradition in the Middle Neolithic, which does not result in significant disturbance of the ground and the concomitant production of easily recognisable archaeological features. Consequently, the excavation of the potential prehistoric site at Old Freehold, with its concentration of Middle Neolithic artefacts, provided a rare opportunity to excavate the site of a possible Middle Neolithic settlement that had apparently been ploughed only once. Given the possibility that the only archaeological evidence for the Middle Neolithic structures might exist in the ploughsoil, rather than the preserved stratigraphy of subsurface features, an excavation methodology was developed to maximise the recovery of information from the ploughsoil. That the site had only been ploughed once suggested that the spatial coherence of any evidence for prehistoric structures within the ploughsoil (such as artefact distributions and variations in mineralised phosphate and magnetic susceptibility levels) would be significant. Soil samples for phosphate analysis, magnetic susceptibility and micro-debitage analysis were taken every square metre. To provide a ‘control’ for assessing the results of these analyses it was decided to locate the excavation trench partly over the fieldwalking ‘hotspot’ and partly over an adjacent area from which relatively few finds had been recovered. Consequently, a rectangular area (20m by 15m) was excavated. The excavation was intended to be as much an evaluation of the recording methodology, as it was an evaluation of the site.

The stratigraphically latest deposit, which extended throughout the trench, was a thin layer of harrowed soil that supported a sparse crop of recently seeded grass. This overlay the truncated remains of the ploughed furrows which also extended throughout the trench and varied in depth between 0.15m and 0.25m. Most of the furrows consisted of a layer of ‘bleached’ grass, a band of brown loamy soil and a thin band of redeposited boulder clay, which respectively represented the sod, soil profile and truncated upper part of the natural boulder clay which had been cut and turned by the plough. In the eastern part and southern edge of the trench, the excavation of the ploughed furrows revealed a number of exposures of the basalt bedrock, whose presence had caused the plough to rise. Consequently, in those parts of the trench the ploughing had not truncated the surface of the boulder clay, resulting in the survival of patches of the impoverished base of the pre-ploughing soil profile, several mixed deposits of trampled boulder clay that contained cultural material and an artificial deposit of redeposited boulder clay, which possibly formed the remains of a house platform, and which sealed a charcoal-rich silty loam. A small number of probable and possible negative features cut into the natural boulder clay were sealed by the ploughed furrows. In addition, one possible negative feature was sealed by the charcoal-rich silty loam. All of these features were heavily truncated, hence the uncertainty in ascertaining whether they are artificial or natural.

Large assemblages of prehistoric pottery and worked flint were recovered from the harrowed soil and ploughed furrows. Provisional analysis suggests that Middle Neolithic types dominate both the pottery and worked flint assemblages. In a distribution pattern that mirrored the results of the fieldwalking exercise, the bulk of this material was recovered from the northern part of the trench. In comparison, the possible house platform and the majority of the probable, truncated, negative features identified during the course of the excavation were uncovered in the southern third of the site, where relatively few finds were recovered. Three stones with flat sides containing crudely incised chequerboard patterns, consistent with the known forms of the early medieval game of hnefatafl, were recovered from the charcoal-rich silty loam. Although the gaming boards were not recovered from parts of the deposit physically sealed by the surviving portion of the redeposited boulder clay, if their provisional identification as an early medieval type is correct then they provide a terminus post quem for the possible house platform. Apart from the gaming boards, no other obviously early medieval material culture was recovered.

There is no meaningful stratigraphic relationship between the artefactual assemblages and the truncated, negative features. If the negative features are contemporary with the prehistoric flint and pottery assemblages, it is notable that their distribution does not coincide. Such a pattern would be consistent with flint being worked, and pottery being dumped on middens, adjacent to buildings within a settlement site. If the house platform is early medieval or later in date, then this indicates that a significantly later site has been superimposed upon the earlier prehistoric site. Such an occurrence would have considerable implications for the interpretation of the results of the ongoing analyses of the mineralised phosphate and magnetic susceptibility samples taken from the harrowed soil and ploughed furrows.

School of Geography, Archaeology and Palaeoecology, Queen’s University, Belfast