Chapter 3: Introduction to the data
3.1. Problems of the sources
The study of Scottish burials is beset with methodological problems. Drawing this research together has highlighted the most obvious of these: the fact that the majority of sites are actually unconfirmed by excavation (compare Figure 3.1 and Figure 3.3). This is because they are either from old reports, like the vague notices of ‘stone coffins found here’ on 19th century Ordnance Survey maps, or they are only known as cropmarks, as with the majority of barrows. Antiquarian finds, even when excavated, were often poorly recorded and can be hard to relocate; quite often, these notices consist of nothing more than the testimony of local informants. Barrows and cairns bring their own problems as well, since mounds are known to have been used for burial since the Neolithic period, and vague reports of
‘tumuli’ could refer to any period unless diagnostic material culture is found. Of course, only those mounds that turned up ‘relics’ were likely to have been reported in the first place, and so unfurnished graves of the first millennium AD may be under-reported.
Despite these issues, unconfirmed burials often make their way onto distribution maps, and so the database had made this distinction clear (2.4.1). This research will thus focus on confirmed sites, while acknowledging the backdrop of such unconfirmed burial evidence.
Figure 3.1: Simplified distribution of burial evidence from the first millennium AD, presented as a binary opposition; compare to Figure 3.2.
Figure 3.2: Complex distribution of all burial evidence from the first millennium AD;
compare to Figure 3.1.
Figure 3.3: The distribution of burial sites confirmed by modern excavation highlights how much of the evidence is based on antiquarian reports and unsubstantiated cropmark data;
compare to Figure 3.1.
Another limitation that needs to be confronted is the poor preservation of bone in much of Scotland, in upland areas where acidic soils predominate. Combined with the general preference for unfurnished burial, unlined or ‘dug graves’ are likely to be under-represented in the archaeological record. Cropmark data is also problematic. The advent of
aerial archaeology in the mid-20th century led to the discovery of a previously-unrecognised grave type: the square barrow (Ashmore 1980). This distinctive burial monument is easily spotted in cropmarks, and a large number are recorded in the National Monuments Record of Scotland (NMRS). The problem with using this data is that it is confined to areas where cereal crops are grown and where aerial reconnaissance is undertaken. Furthermore, at least some of these cropmarks may turn out to be other forms of enclosed settlement or structure once excavated, so sites known only from cropmarks should be treated only with care (Cachart 2008; Halliday 2006: 12-13).
Round barrows are also attested at early medieval sites like Redcastle ANG (Alexander 2005), but an isolated ring ditch seen as a cropmark will rarely be interpreted as a Late Iron Age barrow – without excavation or a telltale central grave pit, ring ditches are often assumed to be prehistoric burials or settlements (Cowley 2009). A number of the ring ditches in the NMRS may be contemporary with the square barrows but will not be interpreted as such unless they are in close association or in a linear arrangement, typical of barrow cemeteries (6.4.3). There is also the question of scale: while round barrows are typically 5-10m across, there are some larger ones, for instance at Back Park, Kettlebridge FIF where the cropmark cemetery includes what seem to be round barrows 25-35m across (DES 1997: 39); such huge ring ditches would be interpreted as prehistoric barrows or roundhouses if found in isolation. Thus the already skewed distribution of barrows, found largely by aerial photography, will be biased toward the more diagnostic square barrows.
Cairns, on the other hand, are unlikely to create distinct cropmarks. Instead, these are most often found either by ploughing or coastal erosion. The distribution of cairns is markedly coastal (Figure 3.2), bringing to mind sand dune sites like Lundin Links FIF (Greig 2000) or Ackergill CAI (Edwards 1926). The kerbed cairn may well have been an adaptation for coastal areas, where sandy soils do not lend themselves to mound-building, but the small number of inland cairns shows it is not restricted to beaches. However, these inland cairns are much harder to spot; very few have been found as a result of archaeological survey, the rest being reported by farmers who have come across a Pictish stone or a long cist. If a cairn covered only a dug grave, the odds it would be reported or even noticed during field clearance are quite low. This may explain the relative scarcity of cairns in lowland sites.
Figure 3.4: The square and round barrows at Forteviot focus attention away from the more numerous dug graves, in green; polygons in red indicate areas excavated in 2007 and 2009 (after Campbell and Gondek 2009).
Finally, our neat dichotomy of flat grave vs. mound cemetery does not always stand up to scrutiny (Williams 2007a: 149-150). It should be noted, first of all, that the vast majority of the barrows and cairns under discussion here contain the same kinds of extended, supine, orientated (west-east) inhumation burials as flat grave sites. Further, excavations of barrow and cairn cemeteries regularly turn up a number of adjacent flat graves alongside the mounds. For instance, at Forteviot PER, the two conjoined barrows each cover a single dug grave, but directly north of these barrows is a large inhumation cemetery of which ten dug graves were excavated in 2007 (Figure 3.4; Poller 2008).