Interpretation and results


This page is designed to give a broad overview of results from the project and our current interpretations. More detailed dissemination of VASLE's results can be found in our publications and presentations page. This is updated regularly as new work is published and presentations listed. After the completion of the project in late 2007, full publication of results will take place in Internet Archaeology.

Aim 11

The VASLE dataset

The data derived from the PAS and EMC forms the basis for the national mapping of artefact and coinage distributions in Aim 1. It has been cleaned and amended where necessary (see Naylor and Richards 2005). The resulting database contains 9,096 entries, 3,379 from the PAS and 5,717 from the EMC, over 94% of which have at least a four-figure national grid reference which allows for good quality mapping at a larger scale. The distribution of this data (Figure 1) shows a greater concentration of finds in the south and east, especially East Anglia, northern and eastern Kent, the Sussex coastline, North Lincolnshire, and the Vale of York and Yorkshire Wolds. This is not surprising for two reasons: firstly, the collection of metal-detected data has a longer history in eastern England, especially East Anglia and Lincolnshire and so the networks of contacts are well defined, and second, excavations across the country do indicate that in the Middle Anglo-Saxon period at least there was simply more coinage and metalwork in circulation in the east than the west. Defining to what degree each affects the final dataset is one of VASLE's over-riding aims. Converting the data to a plot showing the densities of finds made across England and Wales (Figure 2) gives a better indication of the relative concentrations of findspots. The darker the shading, the more finds are present. This shows quite clearly that the density of finds recovery generally increases towards the east, and only low densities visible west of the Rivers Severn and Trent.

The VASLE data also allows us to plot distributions of a whole range of artefact types. For some groups of finds the numbers known previously were simply too small to be able to discuss their distributions with any confidence- the fact we can plot them within VASLE illustrates the importance of recording schemes such as the PAS to furthering our understanding. For example, Middle Anglo-Saxon pins are well known from excavations at the urban emporia and from some rural settlements but, in reality, this amounted to a very small number of sites across the country. Through the PAS there are now several hundred examples of pins made by metal-detectorists which shows a broad spread of material suggesting their use was widespread (Figure 3). Compare this to the Viking Age weights and Islamic coins, known as dirhams, (Figure 4) made on sites dating from the ninth to late tenth/early eleventh century. These are very much constrained in the east and north-west of the country, which correlates to the area of highest Viking influence. Weights and dirhams are known from across the Viking world and were used as part of a bullion economy. These two examples illustrate the potential power of data we are using.

However, by only examining the early medieval records it would be difficult to safely interpret how much of this is due to the nature of finds recovery and how much to ancient metal use and settlement patterns. In order to do this we must use as much data as possible, preferably representing a longer timespan- in this case, we used the entire PAS dataset, providing finds data from Palaeolithic right through to modern times, over 120,000 records!

Identifying and assessing constraints on finds recovery

Using all of the PAS data gives a broad overview of metal-detected data across England and Wales into which period-specific datasets can be compared and conclusions drawn relating to settlement patterns for certain time periods. This is an important objective as it allows for the determination of the nature of constraints and bias in the data to be tested. A number of techniques have been developed by VASLE to assess this, built around GIS-based analysis comparing finds data against a range of controls and a number of base maps including the ‘constraints’ base map which is used here. This was specifically designed to illustrate where finds recovery may be problematic, limited or unduly affected my modern features, such as urban areas, forests, lakes, the limits of ploughzone farming, and ‘danger’ zones such as military practice areas2. We have also used basic topographic mapping, comparing data to the height of land and river systems, and have produced plots of finds density to determine areas of relatively higher concentrations of finds across the country. Like the VASLE dataset, the distribution of all PAS records (Figure 5) shows a greater concentration of finds in the south and east, especially East Anglia, Kent, North Lincolnshire and parts of Yorkshire with large numbers also visible in the central Midlands, especially Northamptonshire. In north-west England, Wales and the south-west (Devon and Cornwall) noticeably fewer finds are known. Comparing the distribution against the probable constraints on data collection indicates that urbanism and the limits of ploughzone farming are the major factors, but any area where access is restricted, including woodland and ‘danger’ zones also had a negative effect on recovery. Urban areas provide interesting results, with a two-fold effect apparent- there are virtually no finds made in built-up areas as little detecting is able to take place and few stray finds will be visible, but conversely, clusters of finds are often made in the immediate vicinity of many towns and cities, especially larger conurbations. This can be seen as a result of most detectorists working on land relatively local to them. Across the country, most finds have been made in low-lying areas, below 100m OD, and the vast majority below 300m OD although some can be seen at this height in the southern Pennines. Historical landscape elements have affected the patterns of finds produced with the majority of finds made dating to the period pre-AD1500. It is most obvious in areas such as the Weald and around the Wash where archaeological work has illustrated a low density of occupation for much of pre-Modern times.

Figure 6 illustrates the densities of finds made. As before, greater density is represented by darker areas, and these a mostly confined to eastern and southern counties. A further area of high density can be seen in the Northamptonshire/Warwickshire areas, and the few outliers elsewhere simply reflect a concentration of finds at a single spot or at least over a small area. It should be noted that density is lower to the north of the River Trent and west of the Pennines, in Wales, and along the Thames valley.

The main question to answer here must be to what extent the distribution patterns seen are a product of constraints on modern recovery of artefact and how much can be considered a real indicator of ancient settlement patterns. Clearly, this is a difficult question to answer, and there is much regional variation, producing the complex patterns summarised above. However, in general, we can be quite confident that the dearth of finds on higher ground is an indication of ancient settlement patterns and not just the limits of ploughzone, especially given that even on higher ground within the ploughzone the number of finds made is generally lower. In eastern England (excluding the north-east and south-east), constraints on data recovery are generally low, and levels of ploughzone very high, and so here it is likely that the distributions have a basis in ancient patterns of settlement. In the north-east there are known problems of access to land and so the sparse distribution there is constrained by modern recovery problems. South-east England faces very similar problems to the north-west and much of the Midlands with high levels of urbanism. The problems associated with these urban areas are most pertinent, and the density of finds immediately outside these areas would be expected across the country. Therefore, in the Midlands, north-west and south-east, outside of the highland areas, it should be expected that a more general spread of finds should be seen but urban areas distort this picture to produce a biased pattern.

The VASLE dataset: reality vs recovery

By comparing VASLE's data with the above it is possible to begin to divide patterns which relate to Anglo-Saxon settlement and metal/coin use from those derived from modern recovery (Figures 1 & 2 vs Figures 5 & 6). This technique would also work for any other period in history. The major variations between the two sets of data relate to areas in the west. There are fewer VASLE finds, and correspondingly lower densities in the north west, South Wales or the Midlands and virtually no early medieval finds have been made in south-west England. As for most pre-Conquest periods the Weald has produced few finds, and this corresponds well with much early medieval archaeology. Variations are also present in East Anglia, especially in north-east Suffolk where the relative finds density is very low. In all if these areas, however, the underlying reasons for the difference must be historic as good numbers of finds from other periods are known. Conversely, the effect of the EMC is visible on the Yorkshire Wolds areas of which exhibit higher density than for the overall PAS dataset. This can be explained by the fact that whilst the PAS only records new finds, the EMC contains all known single coin finds, including excavated finds. In most areas this has little effect on distribution patterns or relative density but eastern Yorkshire is somewhat different. The area is a major concentration of ninth-century styca coins, over 1000 single finds, which is so far not reflected to anywhere near the same extent on the PAS database.

1. This summary of Aim 1 is based on a report in the Medieval Settlement Research Group Annual Report for 2005. For full reference click here. (Back to text)

2. The background map detail is based on copyright digital map data owned and supplied by HarperCollins Cartographic and is used with permission. Contours are derived from LANDMAP datasets and are used with permission. Where shown, the network of Roman roads is based upon I. D. Margary, Roman Roads of Britain, (London, 1973). (Back to text)

This page was last updated by John Naylor on 20/12/2006

Back to top

VASLE is funded by

The Arts and Humanities Research Council