Diet and Ethnicity during the Viking Colonization of Northern Scotland: Evidence from Fish Bones and Stable Carbon Isotopes
Barrett, James H., Beukens, Roelf P., Nicholson, Rebecca A., Antiquity
The legitimacy of migration in archaeological explanation waned in the mid to late 20th century, at least in the Anglophone scholarly tradition (Harke 1998: 19). Revisionist interpretations of Viking colonization in northern Scotland emerged in step with this trend. In the 1950s, F.T. Wainwright (1962: 125-6) was confident that `Scandinavian settlement amounted to a mass-migration. The Scandinavians arrived in numbers sufficient to overwhelm the earlier inhabitants politically, socially, culturally and linguistically'. By the mid 1970s, however, Anna Ritchie (1974; see also Ritchie 1993: 25-6) proposed what can be described as the integration or acculturation model (see Morris 1996: 77) -- which assumes considerable continuity of native culture -- originally based on the discovery of Pictish artefacts in presumed Viking Age houses. This model of cultural continuity -- with implicit or explicit acceptance of a Norse speaking elite stratum -- has since become quite widely accepted for at least the earliest phase of Scandinavian immigration (e.g. Buteux 1997: 261-4; Hunter 1997: 249-53; Sharples & Parker Pearson 1999).
Given a renewed appreciation of migration in the wider archaeological community (e.g. Anthony 1990; Chapman & Hamerow 1997; Burmeister 2000), however, it is appropriate to reassess the acculturation model. This paper will do so from a single perspective -- that of subsistence and diet. It revisits a recent hypothesis (built on observations first made by Wheeler 1977) that a more maritime-oriented subsistence strategy was introduced to Viking Age northern Scotland by Norse farmer-fishermen (Barrett et al. 1999: 369-70). The evidence includes zooarchaeological assemblages and new stable carbon isotope data for human bone of `Pictish' (c. 300-800 AD) and `Viking Age' (c. 800-1050 AD) date (see FIGURE 1).(1)
[Figure 1 ILLUSTRATION OMITTED]
Methods and potential biases
A shift in the role of marine resources at the Pictish--Viking Age transition was discussed previously in a study of the northern Scottish fishbone record from the Neolithic to the Middle Ages (Barrett et al. 1999). To address this question in detail, however, it is useful to focus exclusively on fish assemblages of Pictish and Viking Age date.
TABLE 1 includes data regarding a series of variables chosen to isolate trends in both the intensity and kind of fishing conducted:
* the total number of identified fish specimens (NISP) recovered,
* the ratio of identified fish to mammal bone,
* the ratio of cod family (Gadidae) to other fishes,
* the ratio of cod (Gadus morhua) to saithe (Pollachius virens) and
* the ratio of ling (Molva molva) and torsk (Brosme brosme) to rocklings (Ciliata or Gaidropsarus species), wrasse (Labridae) and cottids (Cottidae).
[TABULAR DATA 1 NOT REPRODUCIBLE IN ASCII]
The total NISP and the ratio of fish to mammal are both intended to measure the absolute intensity of fishing activity and, by implication, the role of fish in the diet. Conversely, the ratio of cod family to other fishes is included to detect any shift in prey selection towards the gadid species known to have been of paramount importance in Iron Age, Viking Age and medieval Norway (Enghoff 1999:55; Perdikaris 1999). The last two ratios, which are intended to measure the relative importance of `offshore' versus littoral fishing, require some explanation. During recent centuries in northern Scotland saithe of a size common in the assemblages under consideration (the majority of bones represent individuals between 100 and 400 mm total length, see Barrett et al. 1999: 361-2) were caught from the shore and from boats in shallow water using a net or simple rod and line (Fenton 1978: 527-30; Baldwin 1982). Conversely, cod fishing, often for larger individuals (the majority of specimens from the Viking Age assemblages considered represent fish over 800 mm total length, see Barrett et al. …