St Abb's Head, Kirk Hill

A trench 19-0 m long was laid out across the north-western sector of the Kirk Hill rampart to examine the structure and chronology of the Kirk Hill rampart, and to pick up possible evidence for internal buildings. The cutting was originally laid out to a width of 3-0 m; but because the size of the

rampart had been underestimated, our resources proved to be insufficient to excavate the whole trench down to bedrock at that width. For much of its length, only a 2-0 m width was excavated, while the deepest parts were only 1-0 m wide. Only special features, such as the front revetment and palisade slot, were explored to the full intended width. Inevitably, therefore, both observation and interpretation are more tenuous than is normal when data are derived from a single cross-rampart cutting. A further problem in understanding the stratification was presented by the very intensive rabbit burrowing destroyed much of the stratigraphy on the south-west face of the trench.Dating is provided by three radiaocarbon dates: GU-1387 Quercus (oak) 1395±60 bp, GU-13881 Mixed charcoal, including Betula 1285±60 bp and GU-1389/ (birch), Corylus (hazel), Salix (willow) 1265±60 bp. These may be calibrated to calendar dates AD using the calibrations of Klein et al (1982) GU-1387 570-760 AD

GU-1388 610-880 AD

GU-1389 615-885 AD

The earlier period had two successive phases of timber palisade, or less probably, a single-phase double palisade. In the general context of palisaded sites in north-east England and south-east Scotland, such palisades might have been expected to fall into the Pre-Roman Iron Age, or even the later Bronze Age. The radiocarbon dates, however, rule this out, placing palisade A firmly in the second half of the first millennium AD. It is thus broadly contemporary with the single palisade at Doon Hill (NCR NT 6875), with the considerably more massive double palisade of the fort or Great Enclosure at Yeavering

(NGR NT 9331) and perhaps with the double palisade of the second phase at Harehope (NGR NT 2044), which is linked to Yeavering by the peculiarities of its entrance arrangement.

The latter period comprised a turf and clay bank, with a sub-phase represented by the front revetment. The stone wall on the crest probably marks an altogether later work. The dating of Kirk Hill 2 is inferential rather than proven. The radiocarbon dates provide no more than a probability estimate for its earliest possible date: there is a 19:20 chance that it was built after AD 615 on Klein's calibration. It could even be the defence of the burh, adopted

as the vallum of St Aebbe's monastery. The only argument against it being a pre-monastic work is provided by the toe of quarry-dressed stone most likely available on site during the building of the monastery itself. On these grounds, the Kirk Hill 2 rampart may indeed be the monastic vallum enclosing the exposed hilltop with a bank about 8 m wide, and over 3 m high, which backed onto the vertical sea cliffs. There were entrances at the northern and

southern tips, that on the north approached by a terraced trackway. Between the bank and the cliff was a fairly level area of about 3 ha, containing both wooden and masonry buildings, communal and private, as Bede described them.

Alcock et al 1986

A trench 19-0 m long was laid out across the north-western sector of the Kirk Hill rampart to examine the structure and chronology of the Kirk Hill rampart, and to pick up possible evidence for internal buildings. The cutting was originally laid out to a width of 3-0 m; but because the size of the

rampart had been underestimated, our resources proved to be insufficient to excavate the whole trench down to bedrock at that width. For much of its length, only a 2-0 m width was excavated, while the deepest parts were only 1-0 m wide. Only special features, such as the front revetment and palisade slot, were explored to the full intended width. Inevitably, therefore, both observation and interpretation are more tenuous than is normal when data are derived from a single cross-rampart cutting. A further problem in understanding the stratification was presented by the very intensive rabbit burrowing destroyed much of the stratigraphy on the south-west face of the trench.Dating is provided by three radiaocarbon dates: GU-1387 Quercus (oak) 1395±60 bp, GU-13881 Mixed charcoal, including Betula 1285±60 bp and GU-1389/ (birch), Corylus (hazel), Salix (willow) 1265±60 bp. These may be calibrated to calendar dates AD using the calibrations of Klein et al (1982) GU-1387 570-760 AD

GU-1388 610-880 AD

GU-1389 615-885 AD

The earlier period had two successive phases of timber palisade, or less probably, a single-phase double palisade. In the general context of palisaded sites in north-east England and south-east Scotland, such palisades might have been expected to fall into the Pre-Roman Iron Age, or even the later Bronze Age. The radiocarbon dates, however, rule this out, placing palisade A firmly in the second half of the first millennium AD. It is thus broadly contemporary with the single palisade at Doon Hill (NCR NT 6875), with the considerably more massive double palisade of the fort or Great Enclosure at Yeavering

(NGR NT 9331) and perhaps with the double palisade of the second phase at Harehope (NGR NT 2044), which is linked to Yeavering by the peculiarities of its entrance arrangement.

The latter period comprised a turf and clay bank, with a sub-phase represented by the front revetment. The stone wall on the crest probably marks an altogether later work. The dating of Kirk Hill 2 is inferential rather than proven. The radiocarbon dates provide no more than a probability estimate for its earliest possible date: there is a 19:20 chance that it was built after AD 615 on Klein's calibration. It could even be the defence of the burh, adopted

as the vallum of St Aebbe's monastery. The only argument against it being a pre-monastic work is provided by the toe of quarry-dressed stone most likely available on site during the building of the monastery itself. On these grounds, the Kirk Hill 2 rampart may indeed be the monastic vallum enclosing the exposed hilltop with a bank about 8 m wide, and over 3 m high, which backed onto the vertical sea cliffs. There were entrances at the northern and

southern tips, that on the north approached by a terraced trackway. Between the bank and the cliff was a fairly level area of about 3 ha, containing both wooden and masonry buildings, communal and private, as Bede described them.

Alcock et al