by Tim Pestell
If for the sake of argument one assumes these things were connected, could the Pakenham great looms be part of a luxurious estate-centre weaving establishment, staffed by slaves and providing very high quality goods to the king which might be given to an archbishop of Mainz or even sold to an emperor? This is, of course, no more than a crude hypothesis to illustrate an argument, or rather a dilemma. It is this: how far were manufactured goods produced in association with lordship? Was it the possession of skilled slaves, toiling away in sunken-featured buildings, which gave great men control of manufacture, establishments such as that at West Stow being, so to speak, their factories? The interreaction between lordship, production and what we might think of as market-directed production was doubtless a complex one and involved attitudes and procedures contrary to our normal ways of thinking.
The economy which concerns us is that which paid for Bede. No interpretation of the economy (nor, I would venture to say, of the society) of seventh-century England can make sense if it does not take account of Monkwearmouth-Jarrow. Bede’s monastery gives us an island of assured knowledge in a troubled sea, in which limited information is churned by cross-currents of speculation, not all of it happy. We have no doubt of the scale of the library of Monkwearmouth-Jarrow nor of the importance of the learning which was fostered there. We have no doubt about the Roman fineness of the buildings: not only are they described by Bede, their remains have been unearthed by Rosemary Cramp (for example Cramp 1976). Bede’s account, his own writings, and the physical remains prove that Monkwearmouth-Jarrow was an immensely, an almost extravagantly, expensive enterprise. Biscop’s repeated visits to Italy, his purchases of many manuscripts and other religious items there, the importation of craftsmen from Gaul to construct the fine buildings: all these things show him as a devout Maecenas, possibly at least as rich or richer than whoever was buried in Mound One at Sutton Hoo. By no means the least call on the funds for a founder was the maintenance of the community. Bede says that in 716 when Abbot Ceolfrith left on his fatal journey to Rome, there were nearly six hundred fratres in the twinned foundation of Monkwearmouth-Jarrow (Plummer 1896, i, 382). What did a frater cost? Writing on a rather later period and, of course, not in a position to give more than vague indication, Duby reckoned that it took thirty peasant families to maintain one monk of St Bertin (1978, 108). Maybe, one may say, English monks at an earlier date came cheaper. Bede, though possibly a distinguished exception in what he cost, can hardly have done so. Note that when, just before his death, he distributed his personal possessions, they included some pepper, which cannot have come from anywhere much nearer than southern India (Plummer i896, i, clxiii). Monkwearmouth-Jarrow was outstanding but it was not unique as a monastery: far from it. All the evidence indicates that monasteries became very numerous in the age of Bede and by 800 there were some thirty in the diocese of Worcester alone (Stubbs 1862). Of course, there were monasteries and monasteries but we know just enough about some of the most obscure to recognise even them as centres of learning.
In short, England was a rich area. It almost looks like a booming area and its economy was such that the emporia must have played a big part. It was such that the study of ‘productive’ sites, the theme of this volume, can begin to be seen as representing part of an economic hierarchy (Ulmschneider 2000a, chapters 2 and 3). In such a hierarchy, one can begin almost to glimpse the economic mechanisms which linked the emporia to the free peasants whom early law codes assume, however reasonably or unreasonably, to have been capable of paying large fines, denominated in gold (Campbell 2000, 69–70). In this connection it is just about worthwhile to notice the number of gold or pseudo-gold coins found on ‘productive’ sites (Ulmschneider 2000a; 20, 22, 29 and 47–8).
Perhaps something like a commercial revolution was in progress linked, as Dr Ulmschneider emphasises, in more ways than one with the progress of the Church (ibid., esp. 66–73). It may not have been for nothing that coining and Christianity seem to be introduced almost hand in hand. Bede tells how Bishop Wilfrid introduced new methods of sea-fishing to Sussex (Historia Ecclesiastica, iv, i3); this could be one instance among many of the introduction of new techniques by missionaries who, almost by definition, were men of wide experience. At the same time one should not lose sight of the possibility that economic progress depended largely on predation, and that behind the wealth which supported the minsters and the scholars lay the slave trade.
A familiar illustration of the role of the slave trade comes from Bede’s account of the battle between Northumbria and Mercia in 679, the battle of the Trent (ibid., iv, 22). After a Northumbrian nobleman, Imma, was taken prisoner his captor sold him to a Frisian at London. The incident illustrates the nexus, possibly the very important nexus, between (frequent) war and the slave trade (Pelteret 1995, 34). It is significant that the merchant was Frisian. The Anglo-Frisian connection, economic, monetary, linguistic and religious, is more than a mere background in our period.
The fruits of metal-detection and, in particular, the exploration of ‘productive’ sites are not only extending but transforming knowledge of the Anglo-Saxon economy in the Early and Middle periods. How can this knowledge be more fully extended and interpreted? I would suggest above all by surveying and studying similar material from later, better-documented, periods. Dr Ulmschneider rightly signals a problem when she indicates that the abundance of ‘detectorist’ finds in Lincolnshire may be connected with the outstandingly large number of deserted medieval villages in that county (Ulmschneider 2000a, 19). It is inevitable that such finds will, to a greater or lesser extent, be skewed towards sites at present uninhabited or thinly inhabited. The best way of estimating how far distribution patterns based on metal-detecting (and not from metal-detecting alone) truly reflect an overall economic position would be to make surveys of the later periods comparing what the ‘archaeological record’ and in particular its ‘detectorist’ element suggests with the comprehensive written evidence of such a source as Domesday Book.
* I do not know how to guess the population at the perid concerned with even spurious accuracy. The considerations advanced by Esmonde Cleary (1989, 174–5) suggest t at half a million could be a modest estimate. If, as Maddicott (1997) shows very possible, bubonic plague raged in the seventh century, then this may have greatly reduced the population. Not to be excluded is that such a reduction (resembling that after the Black Death) could have increased the prosperity of the survivors, whose resulting demand for consumer goods could have contributed to developments evidenced by ‘productive’ sites.
* An unknown portion of clothing would have been of leather; the earliest reference to such in England known to me occurs in the Life and Miracles of St William of Norwich ( Jessopp and James 1896, 15), and relates to the mid-twelfth century, but this must stand for much which went unrecorded before.
† For medieval shoes see Blair and Ramsay 1991, 308–11. The shoe problem is an instance of to how limited an extent important manufactured goods edge into the ‘archaeological record’. A relevant question is this. Off what did ‘ordinary people’ eat and out of what did they drink? If they drank from turned wooden cups, by what economic mechanism did they obtain these? On what did they sit? Not all, one may suppose, with their legs dangling into the pits of sunken-floored buildings.
* The significance of the metal finds on some ‘productive’ sites provokes speculation. Could such bits have been used as currency? For the notion that pins might have been used in this way see Ross (1991, 452).
CHAPTER 3
‘Productive’ Sites and the Pattern of Coin Loss in England, 600–1180
Mark Blackburn
The evidence for coin circulation in Early Medieval England has changed out of all recognition over the past twenty years. With the advent of the hobby of metal-detecting in the early 1970s, and its dramatic growth in the 1980s and 1990s, new data on coin single-finds was amassed on a scale never seen before. The timing was fortuitous, for Metcalf had already begun studying find
distributions as a means of identifying the origin and circulation patterns of the early eighthcentury pennies (the so-called ‘sceattas’) (Metcalf and Walker 1967; Metcalf 1974), and he took pleasure in publishing individual new finds at length as they were still something of a novelty (e.g. Metcalf 1976 and 1984b). As the numbers swelled and numismatists developed their contacts with the detecting community, the recording and publication of finds was put on a more systematic footing in the British Numismatic Journal, initially through a series of articles (Blackburn and Bonser 1984a; 1985; 1986) which in 1987 were subsumed into the annual ‘Coin Register’. Several surveys of the find evidence from a particular period or region appeared (e.g. Rigold 1975; Rigold and Metcalf 1977 and 1984; Pirie 1986b and 2000; Blackburn 1993b; Bonser 1998; Metcalf 1998a), but they quickly became out-of-date and one could not readily be compared with another. Since 1999 an on-line Corpus of Early Medieval Coin Finds from the British Isles, 410–1180 (EMC: via medievalcoins.org) has provided greater access to this material.
It soon emerged that the finds, far from being evenly spread across the country, displayed a regional bias towards the eastern and southeastern counties, and within those areas there were dramatic hot-spots – with certain fields yielding remarkable quantities of Anglo-Saxon coins and metalwork. The finds from two such sites were published in the 1980s – ‘Sancton’ (=Newbald, Yorks.) (Booth and Blowers 1983) and ‘Near Royston’, Herts. (Blackburn and Bonser 1986, 65–80) – and these could be compared with the equally prolific excavation finds from Hamwic (Southampton) (Metcalf 1988a). In 1989 a conference of numismatists and archaeologists organised in Oxford by Metcalf and myself considered the phenomenon of ‘productive’ sites, as these had by then become known, but regrettably the proceedings were not published. The term ‘productive’ site has always been problematic, for it was clear from the start that one could not make comparisons between sites simply on the basis of the number of coins found, as this depended so much on the area available for investigation and the opportunity and conditions of recovery. It is, however, helpful to distinguish fields that simply have little if any coinage or metalwork on them, from ones where there clearly was a considerable amount in the ground waiting to be recovered. It is true that the productive status of some sites only emerges very slowly as more finds are made, and that there is a graduation from sites that yield nothing to ones that are very prolific. Yet once recognised as ‘productive’ it highlights an aspect of the site that demands interpretation. That is not to say that a similar explanation will stand for all ‘productive’ sites, despite the temptation to regard them as the probable location of a market or occasional fair, for already we can see that they occur in a range of contexts: in urban and rural settlements, monastic sites, by rivers, roads and Roman forts.
The principal ‘productive’ sites are listed in the Appendix and their distribution mainly in the eastern counties of England is shown in Fig. 3.1.* Many of the finds have been published or are accessible from the web, but in only a few cases have they been discussed as site assemblages. Those reports should appear soon. The purpose of this paper is to consider the finds from a number of representative sites and set these in a context of the general pattern of finds across England. Before doing so, it is necessary to look at the theory of coin-find evidence and different approaches to its analysis.
FIGURE 3.1. Map of the principal ‘productive’ sites in Britain (see Appendix). 1. Hamwic; 2. ‘Near Carisbrooke’ 3. Hollingbourne; 4.‘Near Canterbury’ 5. Reculver; 6. Tilbury; 7. London; 8. Ipswich; 9. Coddenham; 10. Caistor by Norwich; 11. Thetford; 12. Barham; 13. ‘Near Cambridge’; 14. ‘Near Royston’; 15. Bedford; 16. Bidford-on-Avon; 17. Bawsey; 18. ‘South Lincolnshire’; 19. Lincoln; 20. Torksey; 21. Flixborough; 22. West Ravendale; 23. Riby; 24. South Newbold; 25. York; 26. Cottam; 27–8. ‘Near Malton’ 1 and 2; 29. Whitby; 30. Carlisle; 31. Whithorn.
The different nature of hoard and single-find evidence
There is a fundamental difference between the evidence provided by coin hoards and single-finds:
Hoards
Hoards are typically sums of money that have been put together and buried for safe keeping, and then for some reason not recovered by the owner. The coins may have been selected for the purpose and may not be representative of the money in circulation. Hoarding was a normal way of protecting money, but it may have been more intense in times of trouble or economic pressure, although these could also lead to the digging up of older hoards. What most affects the pattern of surviving coin hoards is the circumstances that led them not to be recovered by the owners or their families. The non-recovery of hoards may have increased, for example, when there was political turmoil, especially war, if people were killed or taken away from their homes. But that is by no means the only situation and it is wrong to seek a political context for every hoard discovered. There are also different categories of hoards – currency hoards, savings hoards, hoards with double peaks, grave deposits – each of which will influence the final composition of the hoard (Grierson 1975, 124–39). Hoards are the numismatist’s primary tool for determining the chronology of a coin series and how long particular coin types remained in circulation.
Single-finds
Single-finds are assumed for the most part to be individual accidental losses from the money in circulation. They will have been lost in many different ways, some while changing hands in a transaction, others slipping from people’s purses or pockets or having been dropped during a fight. Whatever the means, they should be a random sample of the coins in circulation, not specially selected because of their weight, fineness or coin type. Their chance of loss should not have been affected by political events as it was with hoards, so they ought to reflect genuine patterns of coin circulation. With some 6,500 single finds recorded from England 600–1180, they provide a much larger sample and a broader geographical and chronological coverage than the 250 or so English hoards of the same period (Blackburn and Pagan 1986).
The difference in the data presented by hoards and single-finds is vividly demonstrated in Fig. 3.2, which compares the incidence of hoards (above) with single-finds (below). Single-finds are plentiful during the first half of the eighth century, but become progressively fewer during the later eighth and ninth centuries. The hoards present a very different picture, there being very few in the eighth and early ninth centuries but swelling dramatically in the period 860–80, which coincides with the arrival of the great Viking army in 865 and its campaigns leading up to the conquest and settlement of the Danelaw in the later 870s. This is an unusually clear example of the impact a military campaign can have on the pattern of coin hoards.
In using the single-find data one should be aware of potential problems (Blackburn 1989a). Any coin that has been converted into jewellery by being pierced or mounted should be excluded from an analysis since it had probably ceased to have a monetary function. Some apparent single-finds may in fact come from disturbed graves, though this is really only a potential problem for the seventh century, as the practice of placing coins and other goods in graves died out in England at the beginning of the eighth century, while the few fifth-and sixth-century coins in graves had usually been worn as ornaments. Some coins may be strays from disturbed hoards, although as the number of new hoards found in England is tiny compared to the number of single-finds any distortion of the evidence through the inclusion of the odd hoard coin should be small. That this does indeed appear to be so is indicated by the very different chronological patterns shown by the hoards and the single-finds. Suchodolski (1996) has also warned about the influence of votive deposits, such as coins placed in hearths or foundation trenches, but while this may be a factor in Poland, we have no evidence that such practices were followed in Anglo-Saxon England. Our confidence in the general reliability of the single-find evidence is bolstered by the similar patterns that recur at different sites and in different regions.
FIGURE 3.2. Histogram comparing hoards and single-finds from England, 780–90
0.
The interpretation of single-finds: geographical and quantitative approaches
The single-find evidence can be used for interpretation in two fundamentally different ways. One is studying their geographical distribution, taking them as evidence that the coins were in use at the places where they were found. The other is a more quantitative and chronological approach, looking at the number of finds from different periods in a particular place or locality.
We can use the geographical distribution of particular coin types to help identify where they were produced and how they subsequently moved in circulation. The simplest way is to plot the findspots on a map or series of maps, but Metcalf has developed a number of sophisticated statistical techniques to present and interpret the data. Thus he has used a form of regression analysis to compare the average distances and directions coins from different mints travelled before being lost (Metcalf 1998a and b; this volume). This showed that in the ninth century, coins from Canterbury and London generally travelled further that those from East Anglia and Wessex, a feature also seen in the eighth century. For the Late Anglo-Saxon period, after c. 973, he has investigated the velocity of circulation of coinage seeing how far coins had travelled during the short period in which each type was valid (Metcalf 1980, 27–31). These methods look at the distribution of coins of a particular type or mint, but one could equally concentrate on the origins of coins found at a particular site, seeing in which direction it had contacts. Thus, the ‘productive’ site in ‘South Lincolnshire’ is exceptional for its high proportion of Continental coins of the seventh and eighth centuries, implying it had a significant role in North Sea trade. However, before drawing conclusions about the significance of a group of finds, one must take into account the normal composition of the currency in that region and the indirect routes by which particular coins might have arrived.