The importance of the artefact 'fingerprint' is that it gives an impression of the nature of the overall assemblage produced by the PAS dataset. This reflects the circulation of certain artefact types in early medieval England, and provides a benchmark against which individual sites can be compared in Section 4. One of the objectives of the VASLE project was to use these 'fingerprints' to develop comparisons of site assemblages as undertaken by Richards (1999a), Leahy (2000) and Naylor (2004) and to investigate if differences between sites might relate to function.
The 'fingerprint' should illustrate major trends in the dataset, but also allow for comparison between sites or regions without confusion. It was therefore decided that nothing above the overall artefact type (e.g. strap-end, pin) would be used, as this would simply cause too many categories to be present. Coins would also be grouped by broad categories only: sceattas, broad flan pennies, stycas and 'other'.
Figure 63: The artefact 'fingerprint' for the VASLE National dataset (for artefact categories with greater than 20 records)
Figure 64: The artefact 'fingerprint' for the VASLE National dataset (for artefact categories with greater than 50 records)
It was decided to produce two initial charts: one showing all artefact groups for which there were more than 20 records (Fig. 63), and another for groups with more than 50 records (Fig. 64). Owing to the difference in scale, the EMC and PAS assemblages were kept separate as far as calculations were concerned, as counting both as one assemblage would inevitably mean that coinage predominates. Given the different collection histories of the two sources it would also be misleading to make direct comparisons between the proportions of coinage and metalwork in any given region. Therefore the PAS and EMC assemblages are normalised with each group totalling 100%. On Figure 64 there are 25 groups (21 PAS, 4 EMC), accounting for 84.4% of all PAS finds. Many of the smaller groups accounted for <1% of the PAS records. On Figure 63 there are 16 groups (12 PAS, 4 EMC), accounting for 73.8% of all PAS finds. It was therefore decided that Figure 64 would be used as the main comparative chart as the smaller number of groups makes for easy comparison, while on Figure 64 too many groups account for very little of the overall database (<1%). For the purposes of comparing individual site samples in Section 4 these categories were grouped even further.
Figure 63 shows the predominance of personal-related items (the first eight categories comprising strap-ends to strap fittings) and of horse-related items. The former cover the whole period, although pins are the only substantial group which are mostly dated pre-850. Strap-ends virtually all date to after AD 800, and the type continues into the 10th century. The high level of horse-related items is a surprise, accounting for 24.5% of the total. We have suggested elsewhere that this represents large-scale horse ownership and the development of an equestrian class from the 10th-century onwards (Richards and Naylor forthcoming). Hill (2004, 40-1) has also argued for large numbers of mounted cavalry in England from at least the late 9th century, with Athelstan (924-39) demanding that two mounted men be supplied for every plough. The general pattern of coinage reflected in the 'National dataset' illustrates the high numbers of sceattas lost (c. 680-760) compared with pennies (c. 765-1066). Finds of stycas (c. 790-870) are virtually all made in Northern England.
There is variation in these charts by region, especially for the south-west where there are relatively few finds, especially of an earlier date, and in northern England where there is a generally narrower range of material culture, and virtually no 11th-century horse-related material has been found. The latter is very interesting and must indicate that region's identity was expressed somewhat differently to other regions of the country. As noted above, stycas are a peculiarly Northumbrian coinage (unique in Northern Europe), and there are also a number of insular artefacts - some Trewhiddle-style strap-ends form a northern group (Bailey 1992; Thomas 2006), as does one group of buckles (Class A1ib). Overall, this is a useful method for comparing assemblages from different regions and sites across the country, which allows for the major elements of most assemblages to be assessed.
© Internet Archaeology/Author(s)
URL: http://intarch.ac.uk/journal/issue25/2/3.2.1.html
Last updated: Tues Apr 21 2009