9
In 721 Anstruda of Piacenza in northern Italy made an unusual charter. She sold her own legal independence to the brothers Sigirad and Arochis, because she had married their unfree dependant (servus). She and they agreed that her future sons should remain the brothers’ dependants in perpetuity, but her daughters could buy their independence at marriage for the same money, 3 solidi, that Anstruda herself had received. Although Lombard Italy was a relatively legally aware country (and Piacenza is not far from the capital), this charter breaks at least three laws: the law forbidding free-slave marriages; the law, or at least assumption, that the unfree were not legal persons, so Anstruda’s daughters could not be assigned future rights; and the law prohibiting female legal autonomy. Anstruda’s father Authari, a vir honestus or small landowner, consented to the document, but the money for Anstruda’s legal rights went to her directly, and she is the actor throughout. There is an ironic sense in which this account of a young peasant woman, even though she was selling her own freedom, shows how she could make her own rules, create her own social context, even in as restrictive a society for female autonomy as Lombard Italy. This may say something about Anstruda as a person; it also says something about the fluidity of peasant society in Italy.
So also do Sigirad and Arochis, who were some way from home. They were medium landowners and small-scale village leaders in Campione near Lugano in the Alpine foothills, 140 kilometres to the north of Piacenza. They kept charters about their servile dependants; a parallel text for 735 shows them buying control over a second free woman who married one of their dependants, in Campione itself, this time (in more orthodox fashion) from her brother. Their kinsman Toto successfully claimed the ownership of another dependant, Lucius of Campione, in a court case of the 720s, against Lucius’ firm opposition; Toto is also found buying a slave from Gaul called Satrelanus from a woman, Ermedruda, in Milan in 725. The members of this family got about, that is to say, and were interested in obtaining or keeping hold of dependants in a variety of different contexts. They were tough to deal with, as Lucius found; perhaps Anstruda’s daughters would have found it hard to get out of their control in the future. But this dealing in itself marks a certain fluidity; social relationships in and around Campione seem to have been quite complex.
I begin here with Anstruda and Campione as a way into understanding the complexity of early medieval peasant societies. But it has to be said at once that we do not know much about most of them; peasant social practices were too far from the aristocratic and ecclesiastical interests of the great bulk of our written sources. For the most part, our evidence for peasants in the pre-Carolingian West is archaeological; the relatively small number of western villages which give us enough documents to allow us to discuss real peasant actions tend, with only a few exceptions, to be ninth-century rather than earlier, and this chapter will indeed stray into the ninth century as a result. Otherwise, peasants are seen resolutely from the outside, by legislators and hagiographers, who have very moralistic reasons for mentioning them, and little sympathy for their values. But these hostile external observers were also in all our societies from social groups who were rather more powerful than the peasantry, and who were entirely prepared to coerce them if it was in their interests to do so. If we want to understand peasant society in the round in our period, we have to see it in the framework of an understanding of how much wealth and thus power other social groups had as well. This is why this chapter links general problems of economic structure with peasant society. We have to understand the issue of the distribution of wealth before we can understand how much peasant social action really was constrained, in all the different local realities of the West. But the distribution of wealth also has implications for every sector of the economy, which we shall look at in the second half of the chapter.
We saw in Chapters 5-7 that aristocrats varied substantially in their wealth across western societies. In Merovingian Francia, there were some really rich landowners, with dozens of landed estates each, and a highly militarized factional politics. Bavaria was like Francia, although probably on a smaller scale; only a handful of families (apart from the ruling dukes) seem to have been important owners. In Lombard Italy, however, the wealth of the aristocratic strata was much more modest, and the political dominance of kings was overwhelming. Visigothic Spain was more like Italy in that respect, as it seems from thinner data. And the wealth of aristocrats in Britain and Ireland was, as far as can be seen, markedly less; societies there were on a much smaller scale, and the economic difference between the aristocracy and the peasantry was much less marked. In all these cases, too, except for northern Francia (and Ireland, which the Romans never ruled), levels of aristocratic wealth were far lower in the early Middle Ages than they had been under the Roman empire.
These are important contrasts, and they have several implications. The implications for differences in political practice have already been discussed, and we need not return to them here. There are also implications for peasant societies, as just indicated: the less land an aristocracy owned, the more land was in the hands of the peasantry, and therefore the more space there was for peasant autonomy; if an aristocracy was richer, the opposite was true. So the fluidity of action of some of our Italian village societies was made more feasible by the relatively contained wealth of Italian aristocracies; we might not expect Frankish village communities to be as autonomous. This point is reinforced by the fact that in Italy landowning was usually very fragmented; even an aristocratic estate could be divided into dozens, sometimes hundreds, of separate land plots. Aristocratic-owned lands, and the free or unfree tenants who worked them, were thus not all in a single block, and could well be next door to the lands, and houses, of small peasant landowners, who are quite well documented in Italy. There was space for fairly complicated local social relationships in the interstices of estates as a result, even when Italian aristocrats were locally dominant, which they usually were not.
In some parts of Francia, we find the same degree of fragmentation; the Rhineland is one example. Here, aristocrats were very powerful, and we can indeed identify at least two levels of a Rhineland aristocracy, a lesser level with a few estates each, generally in several different villages, and a greater aristocracy with a vast wealth in land spread over a wide region (this aristocracy by the end of the eighth century included major local monasteries like Lorsch and Wissembourg). Inside that framework, peasants had to be careful, for aristocrats were everywhere, and could do them harm. Peasant landowners attached themselves to aristocratic clienteles in a routine way, to obtain protection. But, as we saw in Chapter 5, in the Merovingian period aristocrats were in general more interested in obtaining wealth and status in royal courts than they were in achieving local domination over peasantries. Peasant society could remain largely autonomous even in Francia at the level of the village, and we can see active groups of small owners running some of the best-documented villages of the Rhineland, such as Dienheim near Mainz and Gœrsdorf in Alsace, in the eighth century.
The major exception to this seems to have been Neustria, particularly the well-documented Paris region, where estates tended to be large blocks of land. Here, peasants less often owned their own land, and village autonomy would have been quite difficult. Most of the villages we know about around Paris are indeed documented as a result of monastic estate surveys, polyptychs, which are a feature of the Carolingian period. The estates of the monastery of Saint-Germain-des-Prés in the Paris suburbs often contained whole villages, such as Palaiseau south of the city, which were thus entirely dependent on their landlord. We know the names of nearly every peasant, including children, who held land from Saint-Germain in the 820s, and what rent they owed, thanks to the monastic polyptych; they are among the completest records of village society we have. The peasants listed in them would have lived their lives largely by landlordly rules, and even the markers for local status would largely have depended on the different relationships each peasant family had with its landlord: the amount of land it held, the amount of rent and services it paid, and the free or unfree status of each of its members.
These Parisian villages were regarded as typical of the whole of western Europe by historians of two generations ago. Now that other sorts of document collection have been looked at in more detail, however, they seem the opposite: they were highly unusual in the early Middle Ages in the degree to which peasants in them were dependent on landlords. In other parts of the Continent, the fragmented landowning of aristocrats meant that very few villages had a single landlord, and most such settlements had a mixture of inhabitants: unfree and free tenants; tenants who owned a little land as well; small peasant proprietors who owned all the land they cultivated, medium owners like Sigirad and Arochis of Campione, who did not cultivate their own land (and were thus not peasants) but who were not rich enough to operate politically very far outside their own village; and only in a minority of cases anyone richer - only in the villages where aristocrats themselves happened to live, in fact. These mixed villages were dominated by their richest inhabitants, who were not necessarily peasants, but village collectivities could have a considerable practical authority, and peasants could have a voice in that.
Let us look at a couple of examples of villages which have a substantial documentation in the eighth and ninth centuries, to show how this worked in practice. Gœrsdorf in Alsace is one example, documented as it is in nineteen documents from the period 693-797. These texts survive in the charter-collection of the nearby monastery of Wissembourg, which shows in itself that the monastery gained a large amount of land there across the eighth century; nearly all the texts are gifts and sales to Wissembourg, in fact. The dukes of Alsace owned land there too, and so did the Sigibald family, significant aristocratic dealers in the eighth-century Rhineland. But between the lands of these three large owners, other people lived too. Medium owners lived in Gœrsdorf, like Adalgis-Allo, who with his wife and son sold land to Wissembourg in 695 (two tenant houses) and in 712 (four areas of arable land and woodland), and who stood witness for other donors and vendors in 693, 696 and 713. So did small-owning peasants, like Asulf, who stood witness along with Adalgis-Allo in 693 and who sold all his property to the monastery in the 696 document. What he did after that is unclear, though he could well have rented it back and become a free monastic tenant; such processes are documented elsewhere. There were certainly free tenants of the duke of Alsace in Gœrsdorf, for in the 730s they witnessed concerning the rent they had owed him on land now ceded to the monastery; probably the tenants were contesting the level of that rent, but the fact that they could do so in public shows that they had free status. Most tenants in the village were probably unfree, all the same; they are called mancipia in the charters, which means ‘unfree dependants’. Gœrsdorf was probably most sharply divided between unfree and free. The unfree were all tenants; the free were partly tenants, partly peasant cultivators, partly medium owners. It was the free who stood witness in front of the duke as ‘the men who live in Gœrsdorf’, as the text says. They also probably ran village affairs: perhaps a small-scale law court (called a mallus in Frankish law codes and dispute documents), and almost certainly any collective decisions that had to be made about the economic activities of the villa (village) of Gœrsdorf. The village seems to have been a compact settlement surrounded by its marca, fields, meadows and woods, all of which would have been exploited for grain-and wine-growing, stock-raising and wood-cutting. Gœrsdorf was near the edge of the great forest of the Vosges, but it was already by 700 in a fully settled landscape, with several other villages close by, and its own woodland would already have been restricted in size and quite fully used for its products. There were expanses of wild land in early medieval western Europe, especially in the woodland zones of central and southern Germany, but mostly people lived in territories that had been created and developed by humans for centuries, even millennia, and Gœrsdorf was certainly one of these.
Gœrsdorf was not directly dependent on Wissembourg (or the duke of Alsace), but it had to exist in a political framework dominated by such figures, and the monastery would have been more powerful than any rival there by the end of the eighth century, leaving less space for autonomous peasant action. Most of the villages we have documents for are like this, but sometimes we can find evidence for more independent communities. One example is the group of villages in eastern Brittany around the monastery of Redon which are documented in Redon’s cartulary. These villages, Carentoir, Ruffiac, Bains and others, certainly had tenants, both free and unfree, but it seems that here the majority of local inhabitants were landowning peasants when the Redon charters begin in the 830s (the monastery was founded in 832). Only a minority of these had more than a single peasant holding, or land in more than one village; these were often priests, or else local notables with an official position, called machtierns. Every village had amachtiern (we know the names of most of the ninth-century machtierns of Ruffiac, for example), and they were always among the richest people in the village, sometimes owning well outside it; they had their own special house, often called a lis (cf. modern Welsh llys, a princely court: the Breton language is closely related to Welsh). One might call them aristocratic, but, by the standards of aristocracies elsewhere in Europe, machtierns were not at all rich and powerful; they were no more than medium owners, on the level of Sigirad and Arochis of Campione, and it is not even clear that they were very militarized. In no sense did they dominate their villages, in fact. Only a small minority of landowners in Brittany were large-scale landowners with a military lifestyle: they made up the entourage of the princes of Brittany (who called themselves kings in the late ninth century, at least briefly). As in the Rhineland, if such people lived or had a lot of land in any given village, then that village would be effectively subject to them. But most villages were not; for them, machtierns and priests were the most powerful people around.
The east Breton villages were called plebes in the Redon documents: literally, ‘peoples’ (cf. Chapter 7 for Ireland). They were unusually organized and coherent communities by the standards of the earliest Middle Ages. They ran their own village-level law courts, presided over by machtierns or other village officials, where disputes were settled; other public village business was done at such law courts, too. When disputes were dealt with, it was the villagers who reached judgement; they also acted as oath-swearers for the disputing parties, and as sureties to ensure that losers accepted defeat. In one notable case of 858 in the plebs of Tréal, Anau had tried to kill Anauhoiarn, a priest of the monas tery of Redon, and had to give a vineyard to Redon in compensation, as an alternative to losing his right hand; here, six sureties were named, who could kill him if he tried such a thing again. In that case two of the six were machtierns, perhaps because the case was so serious, but most judgement-finders and sureties were peasants; the villages around Redon policed themselves.
Once again, we know about these Breton villages because Redon obtained lands (and associated documents) there, steadily from its foundation, reaching a peak in the 860s. The monastery was also given political rights in the villages around it, over the head of the peasants, by Carolingian kings and Breton princes; by the 860s at the latest, it was at least as locally dominant as Wissembourg was in Gœrsdorf, and perhaps more so. Here as elsewhere, peasant societies are only clearly visible in the early Middle Ages when they are just about to be taken over by powerful outsiders, the people who were likely to have archives that would survive into later periods. But the plebes which Redon’s land expanded into had, strikingly and unusually, begun as autonomous of landlordly power, and in the 830s their flat social hierarchy still seemed relatively stable. If Redon had not been founded, we would not know anything about them, but, conversely, there is no particular reason to think that their local autonomy could not have continued for a long time.
Document collections in the early Middle Ages generally tell us about the alienation of land, and little else; as noted at the start of this book, these were the kinds of documents which were most normally preserved. They dealt, that is to say, with land which was given or sold (usually to churches and monasteries), or pledged in return for loans, or else leased in return for rent. Reading texts of this kind is sometimes frustrating: surely they give us a very external view of village-level society, documenting as they do the most formal actions villagers could undertake, and, often, the dullest? Court cases, when they survive, generally do so because land was involved too (Anau’s vineyard, for example), but at least they can contain detail of more ‘human’ interaction - hatred, violence, bad faith. They are all the more illuminating because of that. But land transactions are not to be underestimated: they were of crucial importance, for they had to do with the resources available to each peasant family for their very survival. One chooses whom to cede land to; one will alienate or lease to friends or patrons or clients, not to enemies (unless one is forced to by an extreme situation, such as debt, penury or climatic disaster). As a result, if we have a large number of documents for any given village, we can build up pictures of social relationships which are only attested through land deals, but which had wider resonances too. So, for example, it is interesting that the 860s, when Redon got the highest number of gifts from its neighbours, is also a period when we have more court cases between the monastery and its neighbours. In the 830s Redon was a local, still relatively small house, to which one might give land for one’s soul without there being any political implications. In the 860s, however, it was the largest local landowner around, and, if one gave it land, one was seeking a patron. Such gifts by then usually involve the cession of the same land back to the donor in precaria, for rent; if one feared Redon’s power, however, rather than seeking its patronage, one might well oppose it instead, by taking its land, stealing its produce, contesting its property boundaries, or claiming that one’s kin had no right to sell to the monastery, hence the court cases. The document collections of the early Middle Ages are still fairly thinly spread, and we seldom have a critical mass sufficient for a dense description of local realities, but when we do, as around Redon, we can get closer to peasant society.
Palaiseau, Gœrsdorf and Ruffiac can stand for three early medieval peasant realities: the village all owned and dominated by a single lord; the village with powerful external owners but also fragmented property-holding and a significant presence of peasant landowners; and the village where small owners predominated and ran their own lives much more autonomously. How common were each? As already stated, Palaiseau was probably the least typical of the three, at least up to 800; there were village-sized estates in every part of western Europe, but they were only common in a small number of areas, such as the Paris basin. (Royal estates, too, tended to be of the Palaiseau type.) Gœrsdorf was probably a very widespread type indeed; there were, after all, major aristocrats all over Europe, and they had to own land somewhere - indeed, the more scattered their land, the more places they owned it. The Gœrsdorf model can perhaps be seen as typical of most of southern and eastern Francia, Italy (as in Campione), and - though here the evidence is less good - the non-mountainous sections of Spain. Ruffiac can stand for parts of Europe where aristocrats were weaker: Brittany, obviously, but Britain too; other parts of Europe north of the Frankish world; and also more marginal parts of southern Europe, such as the Pyrenees and the Appennines. But there were probably examples of autonomous villages scattered quite widely across Europe, at least in the sixth and seventh centuries. By the ninth and tenth centuries there would be far fewer, as we shall see in Chapter 22. In England, in particular, village-level societies with a relatively high degree of autonomy in 700 or so, at most paying recognitive dues to a king or, increasingly, a church, would have become by 900 or so much more subjected, paying higher rents to a single landlord. England moved, as a whole, from a collection of local societies on a Ruffiac model directly to a collection of societies on a Palaiseau model - a considerable social change, even though a poorly documented one. We shall look at how that process can be characterized in Chapter 19.
Villages were various in many other ways, too: far more ways than can be described in detail here. They varied in their size and spatial coherence, from big nucleated settlements (Palaiseau had 117 holdings, perhaps representing nearly 700 inhabitants), through small hamlets, to sets of isolated farms, and mixtures of all these forms. They varied in the strength of their internal organization; some had structured patterns of decision-making (although this was rare before 1100 in the West, it was not unknown - the Redon villages seem to be examples); some had collectively run pastoral economies (by contrast, collective decisions about agriculture were rare before three-field systems expanded across northern Europe in the central Middle Ages, and before the Arabs expanded irrigation agriculture in Spain and Sicily in the ninth to eleventh centuries). Before 800, overall, villages tended to be smaller and less structured than they would be later, and some historians indeed prefer not to call them ‘villages’ in this period at all. But the idea of all the people living in a given geographical territory, landowners or tenants, being seen as inhabitants of the same place, the villa of Palaiseau or Gœrsdorf or the plebs of Ruffiac (vicus, locus and many other Latin words were also used), is in itself an important element that can be said to be the basis of ‘village-ness’, and I am happy to use the word here. Some villages were fairly weak or small, some coherent or large, and village coherence would slowly increase between around 600 and around 1000, but in all centuries villages and their territories were important as the basic stage on which the peasant majority, 90 per cent of the population of Europe and maybe more, lived their lives throughout our period.
Villages were not egalitarian communities in any period, even if they did not have lords, and large landowners were marginal or absent. Peasants were divided between owners and tenants, and between richer and poorer owners, in a complex pecking-order. The free-unfree dividing-line was also of crucial importance in most villages, separating people who had legal rights, in public law courts and local decision-making (and also duties such as army-service), from people who had none. This line was violently policed by kings, and marrying across it was illegal everywhere, although we have seen, with Anstruda of Piacenza, that people frequently did so in practice. The practical importance of the free-unfree line was probably very variable regionally, too. It mattered more when all tenants were legally unfree, for example, than when unfree tenure was just one version of dependence beside others (as at Palaiseau, where free and unfree tenants lived side by side, and indeed intermarried on a regular basis). But everywhere it marked an important status difference inside the village, and thus a break in local solidarity: village collectivities would not often be powerful and coherent until unfreedom became less common, which was, once again, a feature of the tenth and eleventh centuries more than the sixth to eighth.
Peasant families were not egalitarian either. Many peasants had unfree servants and farm workers; and gender relations were unequal as well. Certain tasks were highly gendered: weaving was called ‘womanly work’; ploughing was ‘manly work’. And the legal subjection of women (already characterized for the aristocratic world in Chapter 8) was at least as complete in the peasant environment, or indeed more so: hence the interest of a woman like Anstruda, who could at least control the terms of her subjection. Few women appear as independent actors in any of our documents, and even fewer of them are peasants. Normally men acted for them, as alienators of land or as plaintiffs in court, or else they appeared alongside brothers or husbands. Their space was in general terms not the public world of law (they appear in it, in fact, hardly more often as independent actors than do the unfree, who at least sometimes appear in law courts to contest their status), but more the world of the household and the house. We do not have any sort of account of internal family relationships at the peasant level, but it is likely that women ran the peasant household commissariat, as we can show that peasant women did in later centuries and that women did at the aristocratic level already in the early Middle Ages. One indicator of this is that in the furnished burials of the sixth and seventh centuries, women are often buried with keys, which seem to represent their control over household money and supplies. We saw in Chapter 3 that the cliché of public roles for men, private roles for women did not fully describe the late Roman world; even in the early Middle Ages it is misleading unless it is properly understood, for many important economic roles were taken by women inside the household, weaving, certainly, and probably other artisanal activities as well (it is a case that has been put for household-level, unprofessional, pottery production, the kind that was normal in early Anglo-Saxon England, for example). But, that said, the public world was not for the most part very accessible to peasant women anywhere in the early Middle Ages. This marked a real change from late Rome.
The kings of early medieval Europe all saw themselves as drawing an element of their legitimacy from their links with the entire free (male) people of their kingdoms, seen in ethnic terms: free Lombards, Franks, West Saxons, men of Dál Ríata and so on. One result is that law codes deal with the whole free population, and often pay a good deal of attention to village-level, peasant society, as we see in the laws of Liutprand in Italy (dating from the years 713-35) or of Ine in Wessex (dating to c. 690), or the FrankishPactus Legis Salicae (c. 510). These are not descriptions of those societies; as was argued in Chapter 1, royal legislation tells us almost nothing in this period except what was in the mind of the legislator, for, in relatively simple political systems like these, written law was seldom enforced in detail or even known about at the village level. Liutprand, at least, often responded to real cases that were presented to him for judgement, but the Pactus might be an entirely imaginative recreation of a peasant society in reality unlived by anybody, a Frankish ideal, as the mythical nature of its legislators indeed hints may be the case. All the same, that ideal reconstruction does at least tell us about expectations of peasant activity. An important law in the Pactus Legis Salicae is law 45, ‘Concerning migrants’, which envisages that any newcomer to a Frankish village (villa) could be vetoed by any current (free, male) inhabitant, as long as the latter could obtain the sworn support of ten other free males for an oath-swearing ceremony held three times in as many weeks: a substantial proportion of the village, then, not just a single person, but still a right of veto. Even though we have no account of such a procedure actually being carried out, and no idea how many Frankish villagers even thought that vetoes existed, we can at least conclude that the political culture of the Frankish kingdom assumed that local-level solidarity was coherent enough for such a process to be conceivable. This links back to the identity which is visible in 700 or so for villages like Gœrsdorf; but it also shows that at least some peasant-level protagonism was recognized as legitimate by legislators.
This recognized protagonism was also associated with the duty of free peasant males to attend law courts and to bear arms. The Lombards, Franks, etc. were armed peoples; the royal link to the free, associated with the public assemblies referred to in Chapter 4, was above all expressed in terms of justice and military service, always the key elements in any medieval political system. Law courts could be local, as we have seen for the Redon villages; it is unlikely that many peasants went to larger-scale, county-level, hearings, which were more the preserve of élite political communities. Whether many peasants really participated in warfare can also be doubted; armies in this period were generally small, up to 5,000 for the Merovingians and far less than that for the Anglo-Saxons, and could usually have been made up of aristocrats and their entourages, who were also, unlike peasants, trained to fight. We saw in Chapter 8 that the Northumbrian aristocrat Imma in the Mercia of 678 saved his life by claiming to be a peasant, who could therefore be presumed not to have taken part in the battle his army had lost. But it is nonetheless striking how consistently legislation assumes that everybody, including free peasants, was liable for military service - in Visigothic Spain in the 670s-680s, indeed, even some of the unfree were. This was partly a royal image: if you were free, you could and should fight, even if in practice you did not. It was also partly shorthand for wider public obligations. Under Charlemagne, men who had fewer than four tenant houses had to club together and send one of their number to fight, meaning that peasant cultivators would rarely go; conversely, those who did not fight did other public service, building roads or bridges or fortresses. But the existence of these assumptions also meant that if a peasant really did want to serve in the army, and had the money to buy a horse and a sword in order to do so, then such service was possible. Medium owners, in particular, could well have been able to fight as often as there was a war, which in some places (eighth-century Francia is the classic example) was every year.
This network of assumptions about public obligation also presumed that there were no sharp dividing-lines between the various social strata of free society. There was a legal break between unfree and free, but there was as yet no division between a free peasant and an aristocrat. The leaders of village society, if they joined the army, might end up the retainers of a powerful lord; consistent patronage or lucky marriages (or both) over a generation or two might allow them to be lords themselves, for there was no formalized boundary to cross. This must have been rare, but it was possible, and we find low-born bishops and even (but more rarely) counts in our sources on occasion, like Leudast of Tours (d. 583), about whom Gregory of Tours was so disdainful (see Chapter 5). ‘Being aristocratic’ was as yet a fairly informal affair; being close to kings (Königsnähe), holding office, controlling substantial lands, living a military lifestyle, were all necessary to a greater or a lesser degree, according to the time and the place, but people who satisfied local assumptions about aristocratic practice seem to have been more or less acceptable to other aristocrats (except to their enemies, at least, as Gregory was to Leudast) whatever their origin. This would not change until 1000 or so, and, when it did, society itself would change too, as we shall see in Chapter 21.
The early medieval peasantry, even if they were landowners, were circumscribed by their aristocratic neighbours, who were so much more wealthy and powerful than they, but in the last two millennia the period 500-800 was probably when aristocratic power in the West was least totalizing, and local autonomies were greatest - taking into account regional differences, as we have seen. This is one of the main markers of the specificity of the earliest Middle Ages. Another that has often been invoked by historians is a relatively low population, and a relative lack of control of the natural world. An image of the early Middle Ages as one of small groups, huddled together in tiny settlements, surrounded (menaced) by uncultivated woodland and waste, is still widely shared, even among professional historians and archaeologists. This wildness is certainly an exaggeration, however. Woods and pastures were not limitless; the Vosges forest did not extend to Gœrsdorf, and in England there was relatively little woodland at all. Even in what is now Germany, where there were great forests well into the modern period, these were for the most part exploited at least for timber and rough grazing (as well as hunting), already in our period, although it is certainly true to say that using - and clearing - woodland would be more systematic in later centuries, as we shall see in Chapter 22. Early medieval landscapes were less fully controlled than they would be after 1000 or so, but they were by no means wildernesses. Archaeology, too, shows that villages could be ordered. Regular sets of wooden buildings and outbuildings in courtyards are common in north European archaeology from Northumbria and Denmark to Bavaria, particularly from the seventh century onwards, and often before. Vorbasse in Jutland, Kootwijk in the Netherlands, Cowdery’s Down in Hampshire, Lauchheim in Alemannia and Kirchheim in Bavaria are particularly well-studied examples (see below, Chapter 10). In southern Europe, village organization was regionally more variable and could be more fragmented, but there are even fewer signs that any part of the land was empty. Indeed, on the rare occasions when we can estimate the size of the population of individual villages in written sources, as in the polyptych of Saint-Germain, or when collective groups of villagers are listed in legal documents, squaring up to expansionist lords (there are examples from the ninth century or early tenth in both the Appennines and the Pyrenees), we can see that in some places early medieval settlement levels could match those of later centuries.
All the same, it would be wrong to leap from a catastrophist reading of the early medieval economy to too much of a continuitist one. It is likely that there was a population drop between the Roman empire and the earliest Middle Ages, not reversed until the tenth century or in some places even later. The density of archaeological sites falls in most places after the Roman period; in both northern France and eastern England, low plateau areas may have been left to pasture, with settlement and fields tending to concentrate in river valleys. Field surveys in other areas have often suffered because identifiable early medieval pottery (the standard marker of settlement in field archaeology) was less widely available or is less well known, but even the most generous interpretation of our Italian or Spanish evidence could not argue for settlement densities matching those of the Roman empire. Any quantification of this would be dangerous, but, overall, it is likely that the landscape was less intensely used in the earliest Middle Ages than either before or after, even if few zones saw any significant land abandonment. Why this population drop took place remains obscure. The early medieval bubonic plague epidemic, which began in the eastern Mediterranean in 541 and is attested in the West on several occasions in the late sixth and seventh centuries, is often invoked as a deus ex machina to explain it, along the lines of the Black Death of 1347-9. This argument relies, however, on some very literal readings of narrative texts, which tend to describe plague in apocalyptic terms. The plague existed, certainly, and killed people too, but neither the archaeology of Syria nor the documents of Egypt support a population collapse in the mid-sixth-century East. As for the West, if there was a population decline in northern France and England, it had already begun in the fifth century, far too early for the plague. Demographic drops do seem to coincide with periods of political crisis and a lessening of aristocratic power, however, and it is possible that a decreasing intensity of peasant subjection, together with a lessening concern for systematic estate management (something we shall return to later in this chapter), allowed for slow reductions in local populations. The slow demographic growth of the Carolingian period, conversely, went hand in hand with an increase in aristocratic landowning and in the intensity of exploitation of a tenant population. Rather than being a guide to the very early Middle Ages, in fact, the polyptych of Saint-Germain tells us most about that period of growth. We shall return to the economic system of the polyptychs in Chapter 22.
The early medieval period was also one in which exchange became much more localized. We have already observed that the fifth century saw the weakening of the great Mediterranean routes when the Vandals broke the Carthage-Rome tax spine in 439. These routes by no means vanished overnight, however. African olive oil and Red Slip fine pottery, both of which are easily identifiable archaeologically (the former because it was transported in amphorae), continued to be exported to Italy, southern Gaul and Spain; less went to Italy, but more is attested in Spain, at least at the outset. Nonetheless, across the sixth and seventh centuries African goods are less and less visible in the northern Mediterranean: they vanish first from inland sites, and then from minor coastal centres. By the late seventh century they are found only in major sites, Rome, Naples, Marseille; nor was this compensated for by the late sixth-century revival of exchange with the East, after the east Roman reconquest of Africa in 534. When, around 700, African productions stopped altogether, nothing replaced them in the western Mediterranean on that scale. What we find instead on the steadily increasing number of Mediterranean excavations are local products, of very variable quality and range of distribution. This variability is even more marked if we add the productions of northern Francia and Britain, which had been part of a separate exchange network focused on the Rhine army in the late empire. Let us look at this variability briefly, the simplest productions first, the most complex and wide-ranging last.
Early Anglo-Saxon England is the best-documented example of a really simple exchange system. Its archaeology shows us that all English pottery before around 720 was handmade, and mostly very locally produced, not necessarily by professional potters, and not even in kilns. Nor did the Anglo-Saxons import much wheel-turned pottery from the Continent (most of it is found in Kent). The frequent presence of weaving tools in house-compounds and female graves shows that cloth was made inside individual households, as well. Metalwork was perhaps not so localized - the brooches found in burials could have been transported over wider areas - but this, too, could have been the work of single travelling craftsmen, working to order in local communities. Little else seems to have been exchanged on more than the local level: a little amber, glass beads, the small-scale (and relatively inexpensive) luxuries of a peasant society. Only the relatively rich and powerful had access to more expensive luxuries, in worked and enamelled metal (including gold and silver) for instance, often bought from Francia, but often also made by dependent craftsmen in royal courts; slaves were part of this luxury network too, largely locally produced in the context of the inter-kingdom wars of the early Anglo-Saxon period (Imma was sold as a slave in London to a Frisian). It would be difficult, however, to say that England had much of a market economy before the eighth century; the huge bulk of production of artisanal goods was at the level of the single village. England can here stand for Wales, Scotland and Ireland, where much the same was true. These lands rarely made any pottery at all; they used wood, leather and iron instead, with equally localized production patterns. They imported some pottery from Francia, and, in the years around 500, occasionally even from the Mediterranean, but this was a high-status luxury, and there is, overall, even less evidence of such imports in western Britain and Ireland than there is in eastern England. Outside Britain and Ireland, we can find an equal simplicity in artisan production in northern Germany and Scandinavia, beyond the Roman frontier. Inside the former western empire, only parts of the Spanish Mediterranean coast show similar patterns so far; but more small areas with simple productive patterns are likely to appear, particularly in Spain, as archaeological work becomes denser.
More complex patterns of production and exchange are visible in the western Mediterranean. Here we find more professional types of artisan, almost always made on a wheel, in both fine and coarse (kitchenware) types; these ceramic types were often available across a city territory, and sometimes further afield, in distribution networks which must have been market-driven. We can see patterns of this kind in parts of southern Gaul, Lombard and Byzantine Italy, and at least some of Visigothic Spain. Gaul, Italy and Spain had in fact already in the late empire had productive systems of this type, alongside African imports. In the post-Roman period, these systems became rather more localized, but they survived when African imports dried up. After 700, Africa itself seems to have retained local productive systems of this kind as well. Ceramics are our best guide to the scale of these systems; but there are some signs that iron- and bronze-work was produced professionally at this local level as well - western Andalucía and Rome with its hinterland are two well-studied examples - and metal products of this kind seem to have been available across a wider geographical range than pottery. Italy and Spain contained networks of relatively localized economies in the late sixth to eighth centuries; every zone had a slightly different history, and clearly differentiated products. Some of these economic areas were larger-scale than others, too; Rome, in particular, seems in the eighth century to have been the focus for a much wider region than was normal by now in the western Mediterranean, covering much of the Tyrrhenian Sea: it imported wine from Calabria and Naples, oil-lamps from Sicily, and in the later eighth century developed a new glazed fine ware, now called Forum ware (it was first found by archaeologists in the Roman Forum), which would in the ninth century be available (in small quantities, at least) from Sicily to Provence. Rome was a big city by eighth-century standards, however, and had for long been a transport focus. The eighth century in the rest of the western Mediterranean, except the Adriatic, was pretty quiet, with almost no sign of any interregional trade except for luxuries. Marseille, the traditional entrepôt at the mouth of the Rhone for all traffic going from the Mediter ranean north into what was by now the Frankish heartland, went into an eclipse at the beginning of the eighth century, and not even the luxury trade had much effect on it after that for some time. Localized production systems do not need such entrepôts, and it is this localization, even if at a decent quality of product, which marks the seventh and, even more, the eighth century in the western Mediterranean as a whole.
The largest-scale economy in the early medieval West was the Frankish heartland. Here the networks of late Roman ceramic productions, based on supplying the Rhine army but extending across the whole of northern Gaul, in the Argonne forest above Verdun for terra sigillata tableware, in the Mayen industrial kiln complex near Trier for coarse-ware containers and tableware, continued after the army vanished, a little reduced in scale but still available over wide areas. Argonne ware had gone by 600, and Merovingian carinated fine wares were generally made on a rather smaller scale, but the Badorf ware of the kiln sites near Cologne, which replaced them after 700, was a new centralized production which could be found throughout the middle and lower Rhine valley, and further afield, and Mayen ware continued to be available over similar areas without a break. We can add to this archaeological material a range of anecdotal documentation of what seems to be fairly large-scale exchange in letters, saints’ lives and narratives. Among others, we have a bishop of Reims who writes to the bishop of Verdun in the 540s to ask about the price of pigs; Gregory of Tours who tells us that the merchants of Verdun set themselves up again after a period of trouble in the 530s with a loan of 7,000 aurei with interest from King Theudebert - he did not ask for it back, and in the 580s, Gregory says, the merchants were doing pretty well; a king (probably Sigibert III) who tries to stop the citizens of Cahors in the 630s or 640s from going to the fair at Rodez, 110 kilometres away, for fear of plague; the annual fair of Saint-Denis, for wine and other products, set up in the 630s and transferred to Paris as a going concern in the years before 709. Cologne, whose centre has been excavated, was a major metal manufacturing centre throughout the early Middle Ages; Paris was not only a fair but also had shops selling jewellery opposite Notre Dame in the 580s and quite a number of resident merchants who appear in documentary sources of various kinds. Northern Francia even had new towns, such as Maastricht, a seventh-century development with pottery-making, metalwork, bonework, and glass-making. An interlinked network of production extended all across the Seine-Rhine region, some of it very widely available, throughout the pre-Carolingian period. This network was destined to expand even further after 800, but it had active roots.
The core of the evidence presented here is the production and distribution of pottery, always the best evidenced product in archaeological excavations. Metal and also glass seem to have had similar patterns, generally showing distribution networks a little wider than those of ceramics, though they are less clearly visible (one can often tell from petrological analysis of potsherds where they came from; metal and glass are too often melted down for this to be possible, and we are reliant on stylistic analysis, which can be misleading, as there was much local copying of successful styles in our period). Cloth, the most important of all, is the great unknown out of such artisanal productions, for it so seldom survives on sites, but it would be reasonable to argue that the scale of its production often matched that of ceramics, and this seems relatively clear in England at least. These were the major artisanal products of the early Middle Ages, and they are the essential markers of economic complexity, along with more occasional agricultural specializations for sale, like the vineyards of northern Francia and also of parts of the south Italian coast. It is reasonably clear from this evidence that northern Francia had a much more complex and active exchange system than anywhere else in the West before 800, that the Mediterranean lands were more fragmented, with pockets of greater complexity and greater simplicity; and that Britain and the rest of the North was as a whole far simpler in exchange terms than almost anywhere further south. The difference between the two sides of the English Channel was particularly acute, and certainly not overcome by imports into England, which were anyway not so very numerous.
So far, no assumptions have been made here about what sort of exchange these patterns represented. As we saw in Chapter 2, the movement of goods in the Roman period was often the work of the state, taking taxes in food and artisanal products from one province to another, to feed the capitals and to feed and clothe the army. But even in the Roman period this was only one part of exchange, and commerce took other goods further, to cities and rural settlements whose supply was in no sense a fiscal concern. The state was much weaker in the post-Roman world, and one would not expect much of a tax-based movement of goods; an equivalent might be the movement of rents from one estate-centre to another, to feed landowners and kings who were located elsewhere, but the evidence we have for exchange, even in the relatively localized early Middle Ages, seems more capillary than that for the most part. With the exception of the luxuries on high-status sites, which were in some (not all) cases produced by subjected craftsmen, dependent on aristocrats and kings, most of the non-local goods found on archaeological sites were probably bought, and produced for sale. This does not mean, however, that aristocrats and kings were irrelevant to the networks that have been sketched out. Far from it: they were the most reliable buyers, for élites had large entourages who needed to be fed and clothed. The threefold division of the West just sketched out has an exact correlate in the differences in levels of aristocratic (and ecclesiastical, and royal) wealth described in earlier chapters and summarized at the start of this one: for Francia had the richest ruling class by far, and the societies of Britain and Ireland by far the least rich, with the different regions of Spain and Italy somewhere in the middle. A rich aristocracy went with an elaborate exchange system, and vice versa. When looking at the factors which underpinned the geographical range and complexity of exchange, the extent of aristocratic demand was the most important. Globally, we have also seen that aristocracies were less rich in the earliest Middle Ages than they had been under the empire (and, as we shall see in later chapters, than they would be later); globally, too, early medieval exchange was simpler than either before or after. But the contrasts between the regions of the West were as significant as those global differences.
This account of the trends of early medieval exchange is different from that found in many books of the last seventy years. These took their cue from Henri Pirenne’s Mohammed and Charlemagne, which first appeared in French in 1937. Pirenne argued for the survival of an essentially late Roman economy, focused on Mediterranean trade, even in Merovingian Francia, until the seventh-century Arab invasions, which broke the unity of the Mediterranean and forced the economies of Europe in on themselves until a commercial revival, this time centred on the North Sea, in the eleventh century. His theory was pre-archaeological, and so the evidence discussed here was simply not available to him; but, beyond that, his model had at least two serious flaws. The first was that it laid far too much stress on long-distance exchange, between the East (sometimes the Far East) and the West, which was always marginal to the main lines of trade; these latter operate above all inside regions or between neighbouring regions, and only very exceptionally extend beyond them (as with the African hegemony over the late Roman Mediterranean, which was, precisely, a product of the needs of an exceptionally powerful state). The second was that most of Pirenne’s arguments concerned luxuries: the availability of gold, spices, silk and papyrus in the West (the last of these was certainly not a luxury in Egypt - it was an industrial product - but arguably had become so in the West by the seventh century). This was perhaps forgivable, as luxuries are almost all the examples of traded goods that are mentioned in early medieval written sources. But luxuries, too, are marginal to economic systems; they are defined by their high price and restricted availability, so that only the rich can possess them, and they therefore represent wealth, power and status. (The jewellery shops of Paris presumably sold exclusively to the rich; they certainly sold to Count Leudast, who was arrested and taken for execution while shopping there in 583.) The reason why they tend to be the products which most often appear in written sources is that these tell us about the rich; but they are the surface gloss on economic systems taken as a whole, which depend for their complexity on much more mundane products: clothes, knives, plates. Luxuries also exist in every economy, whether simple or complex - they were present in Ireland and Francia alike - so they are not much use as discriminators. Now, Pirenne was actually wrong to say that the Arabs closed the Mediterranean; well before the Arabs arrived, the western part of the sea already had dramatically less shipping, as we have seen, and on the luxury level ships continued to link East and West even after the Arab conquests (spices were always accessible in the West, contrary to Pirenne’s opinion). But even had he been right, the luxury level he was discussing was still marginal; the real economic changes were inside regions.
It is not easy to say who made profits from large-scale production in the early Middle Ages. The pottery industry of Mayen might have had a single owner (this is not very likely, unless it was the king, but it is not unimaginable); it might also have been a collection of autonomous potters and kiln-owners, turning out similar wares quasi-competitively. This latter pattern seems to be how it worked in contemporary Egypt, judging by sixth-century papyri, which show the rentals of individual workshops to potters and contracts between individual potters and landowners to supply wine amphorae; it seems to me the most plausible hypothesis in the West as well. But we cannot really be sure, for there are no documentary sources for places like Mayen. It is easier to see who made profits out of distribution, for we have quite a lot of references in narrative sources to merchants. They were often quite small operators, like the debt-ridden merchant Cosmas the Syrian whom Gregory the Great helped out in 594, but they could be both important and influential, like Priscus of Paris (d. 582), a Jewish confidant of King Chilperic, or Eusebius the Syrian, who bought the bishopric of Paris with his profits in 591. The most successful merchant of the period by far was Samo, a Frank who actually became king of the Wends in the 620s, and united the neighbouring tribes against King Dagobert I; he apparently reached this status by helping the Wends in war, so even when still a merchant he must have had a certain political visibility (there is no evidence, unfortunately, about what he traded).
These were independent operators, but merchants could also operate in groups. Examples include the eastern merchants who came to Mérida in the mid-sixth century bearing gifts for Bishop Paul, or indeed the mercantile consortium Samo began with before he struck out on his own. They were also often the employees of aristocrats, trading for the latter, presumably with goods from the latter’s estates, like Jacob the Jew who sold cloth in Carthage in the 630s on behalf of a Constantinople notable, and who had the option of going on to Gaul; or the traders acting for the monastery of Saint-Denis, who got a royal privilege from Carloman II in 769 not to have to pay tolls on the rivers of Francia. But it is unlikely that most merchants were regularly employees; any of them could have been sometimes, but the markets and fairs of northern Francia, in particular, seem to have been the focus of interest of too wide a range of people for landowners’ representatives to have been more than a small part of their number. Some were ‘Syrians’, that is, from the eastern Mediterranean, particularly in the sixth century; some were Jews (though by no means all Jews were merchants); increasingly after 600, many, particularly in Francia, were Frisians, from the Rhine delta and the islands of the modern Netherlands; but merchants could in reality come from anywhere. Unfortunately, we cannot link either the origin or the economic scale of merchants to what kind of goods they carried. Our documentary sources tell us most about luxuries, as already noted; but it cannot be that most merchants concentrated on luxury exchange - there was not enough of it for them, and anyway the bulk goods discussed in earlier pages must have been bought and sold by someone. A miracle-book by Wandalbert of Prüm, dating to 839, describes one ship on the Rhine filled with pottery, and another with wine sent for sale from the monastery of St Gereon in Cologne - the former was wrecked, the latter saved from wreck, by the miraculous power of St Goar. Historians have seized on these as examples of a more normal pattern of trade than most sources give us, and rightly so. But they remain anecdotal (as well as late, by the standards of this chapter); our best source for what goods moved around is still archaeology.
We have seen that exchange across the Mediterranean slowly became less complex in the sixth and seventh centuries, and that African exports stopped by 700. In the eighth century only one important long-distance Mediterranean route is documented at all, as Michael McCormick’s work makes clear, the route from Rome, around the south of Italy and across into the Aegean, up to Constantinople. North-westwards from Rome, a link still existed to Genoa and Marseille, but it is not well documented either historically or archaeologically by now; the same is true of the eastern extension, from the Aegean to Syria and Palestine. The Anglo-Saxon pilgrim Willibald did get from England to Rome, and to Jerusalem and back to Rome, in 721-9, but it was a major enterprise, particularly once he got past the Aegean, and it occupies a large space in Hugeburc’s life of the eventual saint. Other routes did not appear at all until after 750. Inland in Europe, the main routes were certainly rivers: the Rhine, important all through; the Seine and the Meuse, increasingly; the Rhône decreasingly. In the South, the Spanish rivers are less attested, and even the Po in northern Italy had as yet relatively little documented traffic; a trade treaty between King Liutprand and the men of Comacchio, an active port under Byzantine control in the Po delta, from 715 or 730, stresses salt more often than anything else, from the delta salt-pans. This would change, slowly, from the ninth century onwards. But this restriction of long-distance trade routes was only a marginal aspect of the history of exchange, which was overwhelmingly focused on buying and selling inside regions. The Rhine and the Meuse were important because they linked different zones of northern Francia together, not because they were the start of longer-distance routes out into the North Sea. These did exist, all the same, as we shall see at the end of the chapter.
Two other points need to be made about exchange. The first relates to money. All documented early medieval societies had standards of value, and these were almost all in coins (the exception was Ireland, where valuations were in slave women and cows). The Romans had minted a range of coins, in bronze, silver and gold, to aid tax-collection above all. Given the simpler fiscal systems of the post-Roman world, a complex set of coins might have been no longer seen as necessary, and the successor states certainly minted fewer types and on a smaller scale, after the Vandals and Ostrogoths at least, who followed Roman patterns. The Franks after 550, and the Visigoths and Lombards, minted gold coins above all (with silver coins alongside these in Provence and Lombard Italy). In Francia, where minting was especially decentralized (there were up to a thousand mints in Francia), the percentage of gold in these coins began to drop in the 630s-640s, and by about 675 coins had become entirely silver. Around 760 the Carolingians reformed the coinage, formally establishing the silver denarius as their currency, and they extended this single coinage to Lombard Italy in 781, after Charlemagne’s conquest. The denarius dominated the next several centuries of western European coins. In England, debased gold coins had been minted since the early seventh century, and silver since the 670s; in the 760s these, too, were replaced by silver pennies that were parallel to those of the Carolingian monetary reform. These changes show, first, a narrowing of the range of coins minted; and secondly, a switch from gold to silver, which was complete in Latin Europe (apart from in the independent principality of Benevento, which remained closer to Byzantine traditions) by 800.
These changes are good guides to the simplification of state structures in the West, and also to the gradual lessening of the availability of gold, which was barely mined in Europe in this period. They do not tell us much about exchange, however. Historians traditionally put a great deal of weight on monetary issues, for it seemed to them that commerce was impossible without coins. This is not actually true; any merchant in a traditional society can cope with barter perfectly well as part of a bargaining process, as long as a common standard of value exists, and only an unsuccessful merchant will come away from a market with money rather than with goods to sell at the next market: coins themselves do not have to be involved in the process at all. It must also be noted that once the bronze or copper small change of the Roman empire was unavailable, almost all early medieval issues were fairly high-value: a Carolingian denarius was worth around £12 in the money of 2007 judging by the bread prices listed in the acts of the synod of Frankfurt in 794, and a Merovingian or Lombard gold triens or tremissis was nominally worth four times that, around £50. Only some Northumbrian and Italian issues seem to have been worth substantially less. Coins were thus in this period somewhat clumsy aids to exchange; they were standards of value for bargainers, and they were convenient ways of hoarding wealth, but they were not as yet the metonyms for commercial activity that they would later become. Coins do, on the other hand, if they are found in archaeological excavations, give us reliable guidance as to the geographical scale of economic networks, because where they were minted is normally made clear on the coin, and they can be fairly closely dated. These networks have not been studied as rigorously as one might have expected (the best distribution maps are currently for England), but broadly they seem at present to support the patterns, based on ceramics, already described. There is more work to do in this field, all the same.
The second point relates to gift exchange. Gift exchange is an alternative way of exchanging goods to commerce: it passes goods from person to person, but the purpose of this is to cement social relationships, not simply to allow each party to get what they really need, which they can do from a stranger as easily as they can from a friend. Indeed, gifts do not have to be essential items at all, as Christmas-present buying clearly shows. Exchanges of gifts (whether objects or services) were very common in the early Middle Ages. Embassies regularly took gifts with them, and kings could be quite competitive in their generosity to each other, sometimes taking pains to make points to the recipients. A letter of Cassiodorus concerning a water-clock that Theoderic the Ostrogoth gave to the Burgundian king Gundobad around 506 makes it clear that the gift was intended to show the superiority of Italian/Roman technology; so too, we can assume, was the mechanical organ given by the Byzantine emperor Constantine V to Pippin III of Francia in 757, which the Franks wrote up in chronicles. Kings gave gifts to their dependants, too, on a far richer scale than the dependants gave them in return, and part of the quid pro quo was personal loyalty; gifts of land, indeed, had the same assumption underlying them. Donors of land to the church, similarly, expected at least clerical or monastic prayers in return, and often made explicit that they hoped to be rewarded by going to heaven after their death. In England and Wales, giving a lavish feast might mean that the invited guests were expected to fight for their host, as we saw in Chapter 7. All personal relationships were sealed by gifts. They could also be ambiguous, just as personal relationships were, as when Bishop Praetextatus of Rouen, at his trial for treason to King Chilperic in 577, said that he had not bribed men to oppose Chilperic, but had simply given them gifts because they had already given him horses - the gifts (according to Praetextatus, at least) had a different meaning from what outside observers thought.
It was argued by Philip Grierson in 1959 and Georges Duby in 1973 that, in an early medieval economy relatively weak in commerce, much of the movement of goods visible in narrative sources and particularly archaeology could best be described in terms of gift exchange. The large Byzantine silver dish found in the Sutton Hoo burial of around 625, for example, was far more likely to have reached Suffolk as a result of diplomatic gifts, or of a chain of such gifts, than any sort of long-distance commerce. More generally, much luxury exchange could well have been in the form of gifts. But not all of it was - or else the West would not have needed merchants, or the Paris jewel shops; and, above all, none or almost none of the bulk exchange described here could have been restricted to the ‘gift economy’. Some of the village-level exchange in places like England could well have been on the level of gift-giving, between people who, inevitably, knew each other very well. (By contrast, merchants were the object of suspicion, and laws survive from both England and Italy which aim to safeguard buyers from the accusation of buying stolen goods from merchants, as long as they bought in public.) But gifts, like luxuries, however central to social relationships, were marginal to economic systems taken as a whole, even in the early Middle Ages.
Production of artisanal goods simplified considerably almost everywhere in the post-Roman West, because large-scale demand dropped, as aristocrats became less rich and as states no longer bought goods on a huge scale for armies (or else took them in tax). It would follow that this was likely to be the case for agricultural production, too. The fragmentary signs that we have for the organization of estates in the earliest Middle Ages support that statement. Roman estate-management was very complex and variable, and at least some of it was visibly for profit, like the slave plantations of first-century Italy or the demesnes worked by wage labour in third- to seventh-century Egypt. Post-Roman estates seem, in all the documents we have, to have been worked essentially by tenants who, whether free or unfree, owed stable, customary, rents: the simplest and least flexible way of extracting surplus from cultivators, and the one which left most autonomy to the peasants themselves. This sort of management, which can be found in Francia and Italy, and also in central Spain (in the fragmentary accounts written on slate found in the provinces of Salamanca and Ávila), does not show any particular focus on profit, or sale. The only specializations we see are along the northern edge of wine production, from Paris to the middle Rhine, where in the seventh century there are casual documentary references to vineyards, sometimes run directly by the landowner with an unfree vinedresser: these could well have been for sale, to merchants from further north coming to fairs such as Saint-Denis. The rapid expansion of a more complicated - and exploitative - ‘manorial’ estate management would come later, in the Carolingian period essentially, in a period in which exchange became more generalized and intense, whether in regions like northern Francia where it was already relatively large-scale, or in northern Italy where it was more localized. We shall look at those forms of management in more detail in Chapter 22.
The earliest sign of that change in the North, at least, was however a little earlier, around 700, and I will end this chapter with it. In the seventh century, at least two Frankish channel ports appeared, Quentovic south of Boulogne and Dorestad in the Rhine delta. Both, particularly Dorestad (which has been excavated), expanded considerably in the eighth century, and they began in the decades around 700 to have equivalents on the other side of the channel, at Hamwic (now Southampton) in Wessex, London in Mercia, Ipswich in East Anglia, York in Northumbria - as well as Ribe in Denmark and Birka in Sweden. These emporia, as archaeologists call them (the word is sometimes used in early medieval sources too), were interconnected, and buying and selling across the English Channel and North Sea developed consistently in the eighth century and early ninth, when other such ports came onstream as well, such as Domburg in the Rhine delta and Hedeby on the Baltic coast of Denmark. Actually, in England at least, the greater part of the economic activity of such ports was the work of local artisans, the metalwork and glass of Hamwic or the pottery of Ipswich (the first kiln-fired and wheel-turned pottery of the Anglo-Saxon period); regional and local exchange mattered more than the traffic across the sea even here. But it is nonetheless significant that these emporia were on the coast, or on rivers with easy coastal access; whatever their origins (which were diverse), they were developed, almost certainly by kings, in order to funnel whatever maritime exchange there was. We have a letter from Charlemagne to Offa in 796 which makes reference to the size of the cloaks that the Anglo-Saxons were exporting to Francia; there are almost no other diplomatic letters mentioning commerce in this period, and it must have been significant (at least as a political initiative; we cannot say on what scale it was operating). Kings valued maritime trade, and helped it on. And as the Carolingians took power in the eighth century and recentralized Frankish politics, they could give a powerful impulse to trading emporia.
The North Sea in the eighth century almost certainly had more shipping than the Mediterranean. Comacchio in the Po delta was a focus of Adriatic-wide exchange in this period, as well as some exchange up the Po, as we have seen; but there are no equivalents to the nodal ports of the North in the Mediterranean between the decline of Marseille around 700 and the rise of Venice after around 780. As we shall see in Chapter 22, Venice was a centre for the slave trade, channelling slaves, created by the Carolingian wars, to the Arabs for domestic service, and getting spices and other eastern luxuries in return. Venice was, that is to say, a gateway port which based its wealth on luxuries directed to Frankish and other buyers, and was probably as yet even more marginal to the economy of northern Italy than Dorestad was for northern Francia and Hamwic was for Wessex. But things were changing here; more ports would appear in Italy in the ninth century, and Venice would eventually, after 950 or so, develop more of a relationship with its hinterland, too. There was, in the end, more scope for the development of commerce in the Mediterranean than in the North Sea after 800 (see Chapter 15). The Mediterranean connected several complex economies, which after the pause of the eighth century would rediscover the advantages of at least limited levels of exchange. The problem of the North Sea was that, even though the Frankish economy was so active, those of its neighbours were not. It was important for the Anglo-Saxons or Danes to get Frankish goods, as luxuries for the most part, but their élites were not yet rich enough to be able to buy all that many of them. Nor were the economies of the North very diversified; Hamwic’s artisanal products resemble those of Maastricht and Dorestad in their range, and could not easily have been intended for sale outside Wessex. Economic specialization and diversification would slowly develop in later centuries; but the North Sea trade of the eighth century was more a spin-off of Carolingian wealth and political influence than a sign of the future economic dominance of north-west Europe.