Dunbar, Castle Park

A third season of excavation on the headland took place between December 1990 and March 1991.

The earliest activity on the site consisted of two construction slots one of which appeared to have stone settings for upright posts, and a group of four post holes. These features are thought to date to the pre-Roman or Roman Iron Age being similar to other such structures found during previous phases of excavation. This was followed by Dark Age occupation represented by boundary or enclosure ditch containing large amounts of animal bone.

Early post-medieval activity consisted of a large ditch cut which is thought to be associated with the French fortification constructed about 1550. Later post-medieval activity included a stone building of possibly Cromwellian date.

Sponsors: SUAT, HS.

R S Sermon 1991; SUAT 1992.

Excavations 1987-91 in advance of the development for the swimming pool overlooking the harbour at Dunbar showed that the medieval castle stands on a headland that had not only been fortified in the early medieval period, but also earlier in the Roman or pre-Roman Iron Age. The earliest defences identified in the excavations comprise three ditches, which were probably drawn across the headland subsequently occupied by the artillery blockhouse constructed in the early 16th century on the very tip of the promontory, the inner ditch perhaps enclosing 0.17ha, and the outer possibly as much as 0.3 ha. Apart from the blockhouse and the massive ditch isolating it from the rest of the promontory, and the medieval castle standing on the seaward stack to the NE, no trace of these defences are visible on the surface, so reconstructing the full extent of this early fort, and indeed the early medieval fortifications that followed, is fraught and, while the excavators felt that they probably did not include the castle stack, this possibility cannot be excluded. Furthermore, there is no stratigraphic evidence to show whether the three early ditches are contemporary, or indeed whether they represent several periods of construction. In each case the ditches cut earlier features, and those preceding the middle ditch were sealed by a deposit from which a single radiocarbon date of AD 85-380 was obtained. Three dates spanning AD 125-495 were also returned from burnt deposits in an overlying yard, demonstrating that the middle ditch at least, and the demolished rampart that must have accompanied it, were built in the Roman Iron Age, though the inner and outer ditch may conceivably belong to an earlier period. Other evidence of enclosure during the Roman Iron Age is provided by the foundation trenches for several successive timber fences cut into the fills of both the inner and middle ditches before the area was buried beneath the yard, while yet another foundation trench for a fence or palisade on much the same line as the earlier ditches is cut through the deposits in that yard.

Subsequently these successive defences and enclosures were buried beneath a deposit of soil on which a series of early medieval buildings were erected, possibly in the mid-late 6th century AD, with occupation continuing in the early 7th century. No evidence of a defensive line across the promontory at this time were found within the area excavated, but in the succeeding phase a rampart was constructed, probably replacing a palisade on the same line to cut off an area of some 0.47ha. The rampart is thought to have been built of turf on a stone foundation, and though in itself undated, the excavators suggested that the mid to late 7th century was a likely date for its construction; it may have been a relatively short-lived fortification, however. A probable entrance through the palisade facing S may have been repeated in the rampart. In subsequent periods major stone buildings stood on the promontory prior to the sack of Dunbar by Kenneth MacAlpin in the mid 9th century. During these periods occupation seems to have been bounded by a ditch, though again the precise date at which it was dug is unknown and there was no trace of a rampart on its inner lip. The ditch had been recut after it had silted up, and a Norse comb was found in the top of the later fill, which itself contained debris from mortared buildings. Again, the precise extent of this enclosure is unknown, but may have extended to some 0.56ha.

Information from An Atlas of Hillforts of Great Britain and Ireland – 18 November 2016. Atlas of Hillforts SC3911