by Tim Pestell
This broad picture is, again, likely to mask a much greater variety and complexity of choices in the location of markets and trading sites, which will provide ample scope for future research. For example, the site of Melton Ross in Lincolnshire lay inland at a junction of regional pathways, while the settlement at Hollingbourne was situated on the cross-roads of routeways linking the territories of East and West Kent (Leahy; Brookes, this volume). Similarly, analysis of the local geography and geology of an inter-regional and perhaps international market site ‘near Carisbrooke’ on the Isle of Wight, has suggested that easy access to a navigable river seems to have been of secondary importance in the choice of its location, the site instead being situated three kilometres inland, close to a fording point at the junction of three very different ecological zones (Ulmschneider, this volume).
It will only be through more detailed topographical work and a full programme of archaeological excavations that some of the most stubborn elements in the interpretation of many ‘productive’ sites – their precise nature and the wide variety of places that this term invariably encompasses – will eventually be teased out. One of the main difficulties in assessing these sites at present remains our inability to ascertain accurately the nature and intensity of occupation on any site, and its development over time. It is perfectly conceivable that many ‘productive’ sites were only temporary fairs. However, it is becoming increasingly clear that some of the recovered assemblages closely reflect those found on more permanent settlements. The excellent communication links which most ‘productive’ sites had make it unsurprising that they might develop as central places, acting as foci not simply for trade and exchange, but production, consumption and, like Melton Ross in Lincolnshire, administration and justice. The very notion of which elements came first or should take the lead in their interpretation – trading site, administrative centre (such as the caput to local landholdings or as a multiple-estate centre) or religious estate – is therefore difficult. In a radical reassessment, based on his fieldwork at Cottam, Julian Richards (1999a and b, and this volume) has suggested that they need not be any or all of these but may simply represent ‘normal’ settlements which have seen above-average metaldetecting activity. While Richards’ argument may be contradicted by the number of ‘productive’ sites known from elsewhere to have had a visible importance by Domesday, his approach is important in questioning our terms of reference. Just how much material must a site yield to be called ‘productive’? And upon what types of artefact must such a definition be made?
There remains a desperate shortage of the sorts of detailed data needed to allow better interpretations of the likely multivariate functions of such economically important sites, and their role within the hierarchy of Early Medieval markets and other settlement types. These problems should not be seen as insurmountable. Important new approaches in the study of settlement and economic hierarchies in England are already beginning to emerge, from large-scale fieldwalking such as John Newman’s work for the East Anglian Kingdom Survey and Paul Blinkhorn’s detailed study of Ipswich ware (Newman 1992; Blinkhorn 1999). Similarly, Ben Palmer’s and Stuart Brookes’ studies in this volume act as important signposts as to how detailed temporal and spatial analysis, aided by Geographical Information Systems (GIS) models, can help us to understand the wide variety of, and interactions between, sites involved in the economy and trade, and provide crucial information on their origins and development against the background of the landscape.
With so much of the study of inland markets and ‘productive’ sites still in its infancy, what might we expect in the future? As this book shows, the vast body of new archaeological and numismatic evidence derived through metal-detecting in countries such as England and Denmark has allowed the identification of an hitherto unsuspected network of inland markets and other sites involved in trade. This is providing archaeologists with an unprecedented opportunity not only to reconsider the complexities of the Early Medieval economy, but also to study its influence on settlement hierarchy and on Early Medieval society at large. Our prime agenda in Britain must now be both to harness successfully the information from these dramatic new datasets, and even more importantly, to lobby for a programme of large-scale excavations of such metal-detected sites, to shed light on key aspects such as their possibly seasonal nature, their varying functions and status, and their development over time. On the present evidence, these questions simply cannot be answered. The need is all the more pressing given the immediacy and scale of destruction occurring on most rural archaeological sites in the United Kingdom through agricultural activity. There is a real danger now that without the excavation of a representative selection of those endangered archaeological sites our chance of fully understanding and assessing the ‘productive’ site phenomenon will be irretrievably lost.
Equally as important – if not more so – is the need now to view and compare the English findings against the background of developments elsewhere in Northern Europe. As this book shows, the identification of new market and trading sites by metal-detector is not just rife in the United Kingdom, but is also a prominent feature of Danish, Dutch, and, more recently, Swedish archaeology (Hårdh 2000). The systematic excavations on such sites, together with ongoing large-scale research projects on known trading sites and the much greater abundance of Continental written sources, provide a wealth of data from which archaeologists must develop a more integrated and European-wide research agenda for the study of smaller markets and fairs. It is only through such an integrated and multi-disciplinary approach, that we will really begin to appreciate the complex and sophisticated pattern of marketing, trade and exchange in seventh-to ninth-century Europe. Each contributor makes this case in their own fashion, but the broader message of this volume is unequivocal. The study of inland markets and ‘productive’ sites has begun fundamentally to challenge and rewrite the history of the economy of Early Medieval Europe.
* According to the most recent annual report of the Portable Antiquities Scheme,39,346 objects were recorded in the twelve months from October 1999 to September 2000. However, the 1,788 finders apparently responsible for this material are well short of the estimated 13–30,000 metal-detector enthusiasts in Britain (Hobbs 2001, 25 and n. 1).
I
History, Numismatics and the Early Medieval Economy
CHAPTER 2
Production and Distribution in Early and Middle Anglo-Saxon England
James Campbell
Difficillima ars nesciendi is the best motto for a Dark Age historian. What can be more necessary than the most difficult art of not knowing, granted that sources are very thin, heavily skewed and, for many important areas of activity, completely absent? Paradoxically, students of ‘productive’ sites, in adding importantly to the archaeological record, have demonstrated how deficient it is. The Faustian pact which they have (rightly) made with the detectorists has led to the discovery of much, the existence of which had been hardly suspected, let alone known, a commanding reminder of how much there is we cannot know. A guide – some guide – to the scale of the unknown, is to follow the advice of Marc Bloch: ‘Let the figures speak, they have a brutal eloquence’.
Suppose that the population of the area under English control in the time of Bede was half a million *. What quantity of materials would have been needed to keep them reasonably warm and decent? On a very rough, but restrained, calculation, an Anglo-Saxon man or women could hardly have been provided with clothing made of less than three square yards of cloth (I do not set store by the detail of any of the guesstimates here; except to say that, on any guess, we are dealing with very large quantities). Let us allow two square yards each for children under fourteen and assume them to compose a third of the population. The total amount of cloth being worn at any one time would have been of the order of 1.3 million square yards (1.08 million square metres), no allowance being made for persons owning more than one suit of clothes. * Surely one may safely assume that all the population over todd
ler age wore footwear; call it at least three quarters of a million shoes in use at one time, most of a kind which would wear out quite easily and would be hard to repair (possibly some kinds of patten were also in use).† How much metalwork did the population as a whole wear? It is possible that in the lower reaches of society the functions of brooches and buckles were performed by thorns, frogs and pieces of rope (buttons were not introduced to western Europe until the thirteenth century). But we can reasonably assume that a large proportion of the population depended for warmth and decency on small metal goods.
The question of how large a proportion is important because the answer affects judgement on the social significance of the commonest grave-goods. If it is assumed that brooches, strap-ends and the like were normal kit then (on the half-million population assumption) something like two million pieces of metal could have been in wear at one time. It will not escape attention that questions implied here are relevant to Martin Biddle’s thesis on the availability of metal goods (Biddle et al. 1990, 42–74). His argument is that there was a major increase in the supply of these in the tenth century, perhaps not paralleled until the sixteenth. This can be but indirectly relevant to our immediate concerns, yet it leads one to wonder how far there may have been a comparable increase in supply in the Early and/or Middle Anglo-Saxon periods. Biddle and his colleagues demonstrate the importance of the assiduous archaeological recording of what could appear to be inconsiderable trifles. They almost show that the study of numerous fragments can be more revealing than that of objects larger, more complete, and more attractive. Detectorist exploration can hardly lead to minute investigation and recording of Biddle’s kind. Nevertheless, his arguments lend importance to minor metal items and fragments found on ‘productive’ sites; ‘odds and ends’ can help provide a major means to understanding the economic state and development of society. *
Other figures, somewhat more secure and somewhat more surprising, are those for the physical extent of emporia. To take three of the best known: Hamwic extended over 100 acres, Ipswich over some 125 acres, while Lundonwic seems to have been even more extensive, perhaps 148 acres (Ulmschneider 2000a, 83–5; Russo 1998, 142 and 151). Naturally there are problems in relation to density and continuity of occupation. Nevertheless, these are very large areas and clearly indicative that emporia had wide functions, involving many people and/or quite widely distributed articles, imported, exported or manufactured. We also have to bear in mind the strong likelihood that there are other emporium sites still to be discovered. A major question here is that of how late the palatalised -wich element which, as Ekwall showed (1964, 14–29), seems very specially associated with major commercial sites, was newly applied to appropriate places: for example Greenwich, Harwich, Dunwich, Swanage, Norwich and Woolwich have to fall under consideration. The last case has some special importance. In his place-name dictionary (Ekwall 1960, 533) Ekwall interpreted Woolwich as meaning either ‘farm where wool was produced’ or ‘town where wool was exported’; but in his later work (Ekwall 1964, 17) he chose the second option because of the palatalised form of the suffix. If his case is accepted it is important because it takes evidence for the export of English wool back to at least the tenth century. There is no direct evidence for such export until the early twelfth century, though a powerful case has been made for the eleventh (Sawyer 1965, 162–3).
A warning against assuming the absence of an eighth-century emporium site because one has not been found so far is Professor Hodges’ distinguished book Dark Age Economics in which, arguing from the absence of archaeological evidence, he expressed considerable scepticism about the existence of emporia at either London or York, notwithstanding the plain words of the literary evidence (Hodges 1989a, 70 and 73). The great extent of emporia, implying involvement with large quantities of relatively low-value goods, has to imply a hierarchy of lesser places involved in distribution and collection.
Not the least interesting of the palatalised ‘wich’ names is Droitwich. Mawer and Stenton interpreted this place-name as follows: -wic, they say, ‘is simply descriptive of a settlement’ (1927, 285–6). Ekwall has shown that this can hardly be true of the palatalised form. The ‘droit’ element they take to mean ‘muddy’ or ‘dirty’. The possibility of derivation with Old English dryhten with the archaic meaning of ruler or lord is touched upon but dismissed. However, Droitwich (Salinae) had been a major source of salt in Roman times and appears in a comparable role as soon as we have charters to tell us about it. The early references or alleged references to Droitwich are revealing as forms corresponding to ‘Droitwich’ do not appear until the fourteenth century. In earlier sources the ‘droit’ does not appear. Thus in a charter (judged spurious) of 716 it is called Wiccium emptorium; in another of 716X717, better regarded, it is Wico emtorio salis quem nos saltwich vocamus; in an imperfectly recorded grant of 716X757 we have in vico emptorio salis (Sawyer 1968, Nos 83, 97 and 1824). These texts are not such as to inspire general confidence, but it is notable how each combines wiccium, wico or vico (it being unclear whether the word is being used as a proper noun or not) with emptorium (which I take to be the same word as emporium).
Interest attaches also to the name of a nearly place, Wychbold (Mawer and Stenton 1927, 285). The suffix can have a meaning varying from an ordinary house to a manor house (Mawer 1924, 7–8). Of note in this context it has been argued, with reference to continental evidence, that it can indicate a fortified place of importance (Morris 1973, 192 and 472–3). That Wychbold was significant is shown by its having been a meeting place for ninth-century Mercian councils (Cubitt 1995, 219 and 222). Thus the emptorium of Droitwich resembles others with a different economic basis, in being near to a centre of royal power; notwithstanding the authority of Stenton and Mawer, one must incline to suppose that the first element in the name had to do not with mud, but with early power.
Consideration of Droitwich brings one naturally to salt. Those of us who generally only see salt as a tiny mound in the corner of a plate can easily forget the vast quantities of salt earlier societies needed to preserve their food. Consider how much salt is needed to preserve one herring. Sources from the eighteenth century and later give a highest estimate of i.7oz (48g) per fish and a lowest of 0.6oz (i7g) (Bridbury 1955, 3; Young 1813, 230). Suppose the hypothetical halfmillion inhabitants of the Anglo-Saxon lands in early England ate the salted equivalent of ten herrings annually and that it took an ounce of salt (28g) to preserve one herring, then the total of salt required would have been nearly i4 tons (i4,225kg). Yet ten ‘herring units’ per head is likely to have been vastly less than the true Early Anglo-Saxon consumption of salted food. One can see the great economic importance of emptorium salis and reflect on the significance of the possible continuity of Droitwich salt production from Roman times (Hurst 1998). Salt was boiled out in lead vessels, so one may wonder also about the continuity of lead production.
Consideration of ‘productive’ sites inevitably forces us to peer, if in relative ignorance, then in hopeful ignorance, at the whole economic system of Anglo-Saxon England. All I can offer is a handful of guesses, easily open to correction or amplification.
A starting point for one such guess is the recent reinterpretation by Steven Plunkett of discoveries made so long ago as 1953 at Pakenham, Suffolk (Plunkett 1999). He shows that at this site there was a weaving establishment, probably of the sixth century, containing two or more looms. A reconstructed version of one of these has been made for the Ipswich Museum. The original loom was certainly 8ft (2.4m) wide and was probably the same high. This machine and its companion (there may have been more than one such) are argued to ‘indicate organised production workshops making large textiles of standardised quality and character probably for commercial as well as domestic purposes’ (ibid., 295).
Let us put texts beside this archaeological evidence. It is well known, from a letter of Charlemagne (796), that in negotiation with Offa he objected that cloaks being shipped from England were too short. This passag
e is complemented, most interestingly, by another in the account of Charlemagne by Notker, a monk of St Gall writing 884–7. Notker describes the ruler’s testy concern with the excessive brevity of a new style of riding cloak:
Notker writes of these cloaks as supplied by Frisians, but the link between his account and what appears in Charlemagne’s letter is strong enough to support the possibility that these cloaks were woven in England but exported by Frisians. Furthermore, the importance of cloaks from England is emphasised by references in the eighth-century letters of Boniface and Lul. One letter refers to a gift of two woollen cloaks from thelbert II, king of Kent, to St Boniface; their value is suggested by the fact that the other part of the gift was a silver cup, lined with gold, and weighing three and a half pounds (Tangl 1916, No. 105). The second letter is from Alhred, king of Northumbria, sending twelve cloaks and a gold ring to Bishop Lul (ibid., No. 121; Whitelock 1955, No. 187). Important store was set by the cloaks of great men: thus Theodulf mentions the ‘double cloak’ of Charlemagne’s son Charles (Raby 1934, 190).
Let us set beside these references another, from Bede’s Ecclesiastical History. Here is described how King Æthelwealh of Sussex endowed Wilfrid with a vast estate at Selsey. Bede tells us what the great man did upon receiving this gift of 87 hides; he freed 250 servi et ancillae: male and female slaves (Historia Ecclesiastica, iv, 13). Query: what were these people supposed to be doing? Was it a matter of an agricultural workforce, or could we be looking at the domestic staff of something like a major Roman, or indeed Carolingian, villa? The reference to ancillae somewhat favours the latter supposition.