3.1 Pre-AD 1000

With the single exception of the anomalous later Mesolithic oak log from Cumminess, Orkney (Timpany et al 2017), all the pre-AD 1000 dendro-dated sites are found in SW Scotland. There are two main reasons for this pattern. Firstly, the area contains a concentration of wetland settlements which have been the focus of research interests since the 19th century, and which have produced large assemblages of waterlogged timbers. Secondly, the area has benefitted from its geographical proximity and climatic similarity to Ireland, where there is plentiful dendrochronological data covering both the 1st millennia BC and AD, thus making it easier to date Scottish samples. This has resulted in the development of regional tree-ring chronologies covering the period 797–153 BC (Crone forthcoming) and AD 250–752 (Crone 1998), one obvious outcome of which is that it should now be relatively straightforward to date any new material from this area which falls within those spans. The benefits of dendro-dating clusters of proximate sites are apparent in the chronological relationships between the three Iron Age wetland settlements at Dorman’s Island, Cults Loch and Black Loch of Myrton (Cavers and Crone 2019).

A 3D digital image of an excavated roundhouse and trackway
Black Loch of Myrton; 3D photogrammetric model showing roundhouse ST3 in the foreground and the trackway in the background. Oak used in the construction of the trackway was felled over a few years, in the winter/spring of 438/437 BC and the winter/spring of 436/435 BC

Instead of a smear of settlement building activity in the wetlands across the latter half of the 1st millennium BC, potential spikes in such activity are being identified, in the late 5th century BC and in the late 3rd and early 2nd century BC. Black Loch of Myrton and Cults Loch are near-contemporary; they could have been constructed within a year or two of each other and thus, it becomes possible to speculate about the real-time relationships between the two sites. The detailed chronologies from some of the dated sites, particularly Black Loch of Myrton and Buiston, are highlighting the intermittent nature and short duration of occupation at both the scale of individual houses and settlement (see Case Study: Black Loch of Myrton). 

Elsewhere in Scotland other archaeological sites have produced oak timbers which have been analysed but remain undated. These include the Neolithic sites of Blackshouse Burn, Lanarkshire (Crone 1998), Parks of Garden, Stirlingshire (Crone 2002) and Upper Largie, Argyll (Cook et al 2010), the Iron Age sites of Cullykhan, Aberdeenshire, Birnie, Morayshire and Oakbank crannog, Perthshire (see Case Study: Oakbank Crannog), and Early Historic sites including Iona, Dundurn and Loch Glashan crannog (Crone 1998). These sites remain undated largely because they are geographically remote from all the available dated reference chronologies but also because most have not produced sufficiently large assemblages of timber to construct robust site chronologies with strong common signals. 

Progress on some of these sites has been made by combining dendrochronological analysis with radiocarbon wiggle-match dating (WM). This uses statistical techniques to compare or match a time-series of radiocarbon dates – i.e. a series of radiocarbon dates taken from sequential tree rings – with the ‘wiggles’ on a radiocarbon calibration curve. At Cullykhan this has resulted in a floating tree-ring chronology, three centuries long spanning the third quarter of the 1st millennium BC and anchored between circa 290–235 BC, which may in time form the building block for a more extensive, absolutely dated chronology for northern Scotland (Crone and Hamilton forthcoming). 

Carbonised timbers are also potentially datable using dendrochronology. Analysis of carbonised oaks from roundhouses at Birnie, Moray has produced two mean chronologies, 81 and 111 years in length. As with Cullykhan, the lack of any local master chronologies has hindered their dendro-dating so WM dating has also been employed to tie down the date of the outermost years, placing them in the 2nd and 3rd centuries AD. 

Wood from non-oak species has been used extensively in the construction of wetland settlements and their analysis has also proved useful in providing relative chronological evidence. More often than not alder is the predominant timber on Scottish wetland settlements (Crone 2014) and it has proved possible to build alder site chronologies for Buiston, Cults Loch and Black Loch of Myrton, all of which have contributed to refining the dating of those sites (Crone 2000a; Cavers and Crone 2018; Crone forthcoming). Combined dendro- and WM dating of alder sequences has also been undertaken as part of the Living on Water project in which the chronologies of seven Iron Age crannogs on Loch Tay have been investigated, demonstrating that crannog-building in the loch probably commenced in the 5th century BC (Cook et al forthcoming).  

At Black Loch of Myrton ash and hazel have also been used successfully to provide information about the chronological relationships between different structures (Crone forthcoming). The key to success with all the non-oak species is the size of the available assemblage; there needs to be significant quantities of the same species used within a single context. 

Bog oaks, although not archaeological, could provide valuable building blocks for the development of a prehistoric chronology (Baillie 1982). To this end collections of bog oaks from the shores of the Solway Firth (Cressey et al 2001), the Sunart peninsula and Denny bog, Stirlingshire have been analysed. Despite very long sequences of up to 400+ years very few of these bog oaks have been successfully dated. Newbie Cottage 1, from the Solway Firth collection, spans 31122849 BC and Sunart6 spans 25412196 BC but there is little that can be usefully surmised with so few datapoints. Sunart2 has also been dated, producing a medieval date, indicating felling sometime between AD 1325 and AD 1361 (span AD 1115–1315) (Crone 2011a); this wide variation in dates within the Sunart assemblage illustrates the difficulties of analysing an eclectic and random collection of timbers.

A more coherent assemblage was collected from Denny Bog; 14 oak logs were retrieved from a small basin of peat and two chronologies were constructed. However, the two chronologies did not correlate with each other, nor could they be dendro-dated. Radiocarbon dates from the outermost rings of the grouped sequences produced a Mesolithic date of 4519–4366 cal BC for one and a Neolithic date of 3786–3660 cal BC for the other, explaining why there was no overlap and therefore no correlation between the two site chronologies. Consequently, there are now floating tree-ring chronologies 379 years long for the Mesolithic and 323 years for the Neolithic in the Forth Valley. Given that these periods cover many millennia the scale of the task of building prehistoric chronologies becomes very apparent!


Leave a Reply