IN THE WINTER OF 999 AD, the Holy Roman Emperor Otto III left the city of Rome. He was a Saxon, not a Roman, and not particularly holy, but such was the draw of the imperial city that he had made his way there both to make a statement about his own importance, and to use its religious prestige to hold a synod in which to slap down an archbishop who had been causing him the odd problem. This much was more or less standard – if you happened to be an emperor. What’s really striking, however, and the point at which the third Otto’s activities intersect with the focus of this book, is where he went next. Normal imperial activities over the winter might include a spot of hunting, or heading off somewhere pleasant to hold a synod or a council, and/or celebrating one of the major Christian festivals with his great men, ecclesiastical and secular. But Otto did none of these. The emperor had heard of the miracles being performed at the tomb of a recent Christian martyr, the bishop and missionary Adalbert, and had resolved to pay the shrine a visit. Nothing out of the ordinary in that, you might think. First-millennium emperors, Roman or not, all thought they were appointed by God and had a vested interest in manifestations of divine power. But this is where it gets interesting.
Before turning to the brief and ill-fated missionary drive that led to his death, Adalbert had been Bishop of Prague in Bohemia. Otto, however, was not setting off for Prague, nor in fact for Bohemia at all, but Poland. There the latest representative of its ruling Piast dynasty, Boleslaw Chrobry (‘the Brave’), had ransomed Adalbert’s body and built a magnificent tomb for it at Gniezno. We can pick up the story of what happened next in the words of a contemporary chronicler, Bishop Thietmar of Merseburg.
[Otto] was led into the church where, weeping profusely, he was moved to ask the grace of Christ’s martyr. Without delay, he established an archbishopric there . . . He committed the new foundation to Radim, the martyr’s brother, and made subject to him Bishop Reinbern of Kolberg, Bishop Poppo of Krakow, and Bishop John of Wroclaw . . . And with great solemnity, he also placed holy relics in an altar which had been established there. After all issues had been settled, the duke [of Poland] honoured Otto with rich presents and, what was even more pleasing, three hundred armoured warriors. When the emperor departed, Boleslav and an illustrious entourage conducted him to Magdeburg, where they celebrated Palm Sunday with great festivity.1
For our purposes, it’s the backdrop to Otto’s imperial progress that is so significant.
At the start of the first millennium, Poland and Bohemia had been dominated by Germanic-speakers, and the basic pattern of life involved clusters of wooden huts – some larger, some smaller – grouped together amidst the prevailing forests. There were still plenty of trees left at the end of the millennium, but the ruling Premyslid and Piast dynasties of Bohemia and Poland were all Slavic-speakers. The wooden huts had been superseded by castles, cathedrals and armoured knights, which, as we shall see in a moment, had become pretty much standard appurtenances of power right across central and eastern Europe. Not only that, but central Poland had become a destination fit for an emperor, and a suitable location for an independent province of the Christian Church, complete with its own archbishop. There could be no greater symbol, if an imperial visit was not itself symbol enough, that Poland had just been welcomed to the club of Europe’s Christian states.
Nor was Poland alone. Prague, as we have just seen, had a bishop too, and although Bohemia didn’t yet rate an archbishopric, it too had its fair share of castles, cathedrals and knights. Its Premyslid dynasty had definitively converted to Christianity in the person, no less, of Good King Wenceslas – or perhaps just Wenceslas, since we are dealing with his historical incarnation – in the 920s. Subsequent members of the dynasty slipped in and out of Ottonian imperial favour, but this would be true of the rulers of Poland too, and doesn’t alter the fact that both these Slavic ruling lines were firmly members of the club. The first Slavic entity to demand and be granted recognition on this higher plane, in fact, had been ‘Great’ Moravia, which emerged from the wreck of the old Avar Empire in the mid-ninth century. It was the first Slavic state to convert to Christianity, receiving in the 860s, amongst other missionaries, the famous Byzantine Saints Cyril and Methodius, who were responsible for the first written form of a Slavic language, produced to translate key Christian materials for their new converts.2
In Scandinavia too, in the aftermath of a chaotic Viking century, matters were moving in a similar direction. From the mid-tenth century, a powerful state structure began to emerge, based on Jutland and the Danish islands and dominated by successive members of the Jelling dynasty, named after the place from which they originated. Originally pagan, the dynasty converted to Christianity in the person of Harold Bluetooth, and while maintaining a larger naval capacity than its continental Slavic counterparts it too was soon putting up castles and cathedrals, and likewise alternately squabbling with or receiving favours from different Holy Roman emperors. Intermarriages between the Danish and Slavic, particularly the Polish, dynasties soon followed, and they were all part of the same broader diplomatic and cultural orbit.3
As we saw in the last chapter, moreover, Scandinavian expansion had flowed as much eastwards as westwards, and one of its chief outcomes in this sphere was the Rurikid-dominated Rus state, centred on Kiev. This dynasty held on to its ancestral paganism for a little longer than its western counterparts, and, reflecting the particularities of its origins, took a bit longer to get round to castles and cathedrals. Not, though, that much longer: Prince Vladimir converted his state definitively to Christianity in the late 980s, and shortly after the year 1000 constructed in Kiev the famous Tithe Church dedicated to the Mother of God. Built of brick and stone, it measured twenty-seven metres by eighteen, and could boast three aisles, three apses and a cupola: the greatest structure yet seen so far east and north in the European landscape.4
The last two hundred years of the first millennium thus saw new political powers of considerable stature spring up right across central and eastern Europe, in some of the areas that had belonged to the most underdeveloped parts of the western Eurasian landscape. With their emergence, Europe finally took on something of the shape that it has broadly retained down to the present: a network of not entirely dissimilar and culturally interconnected political societies clustering at the western end of the great Eurasian landmass. But what, precisely, was the nature of these new entities, and how had they come into being? What, too, was the nature of their relationship to the patterns of Slavic and Scandinavian expansion we have been examining in the last two chapters. Did they just mean that you ended up with Slavic dynasties in some parts of old barbarian Europe and Scandinavians in another, or was migration central to the whole process of state formation?
As is often the case with the first millennium, it is easier to ask questions than answer them, and for all the usual reasons, though by the end of the period the situation as regards sources is a huge improvement on the era of Slavic expansion. Literacy eventually came to the Slavic world, as we have just seen, with the conversion to Christianity of Moravia in the mid-ninth century. But written Slavic did not acquire any non-religious purposes in this era, and in the centuries either side of the year 1000 even Latin and Greek remained largely restricted to religious uses within the new states. It was not until the early twelfth century that chroniclers around the courts of the new dynasties started to generate homegrown accounts of the past: Cosmas of Prague in Bohemia, theGallus Anonymus in Poland and the Russian Primary Chronicle in Kievan Russia. These texts do contain some useful information, but all had at least partly celebratory purposes vis-à-vis their intended dynastic audience and patrons, and their memories of the ninth and tenth centuries tend to verge on the mythical.5
Of necessity, then, we are again often forced back on historical texts produced by outsiders: the western and southern European states with whom our new states quickly came into contact. This material poses the usual problems of reliability, although they are, in fact, less pressing than those we faced with Roman writers. For one thing, there is much more information. The Viking revolution in Baltic transportation brought Scandinavia into a much closer relationship with literate Europe, while Moravia, Poland and Bohemia were all its close neighbours. And much more was being written in literate Europe, in any case, thanks to the renewed emphasis on literacy that sprang out of the ninth-century Carolingian renaissance. The Emperor Charlemagne had made determined efforts to improve standards of literacy as part of his broader project of Church reform, and literacy continued to increase after Carolingian imperial collapse. When you also throw into the mix the fact that Islamic Arabic authors transmit some important information from another direction entirely, then you can immediately see why we are better endowed.6
A second point is nearly as important as the first. Within pretty short order, all our new entities converted to Christianity. This did not mean that their relations with more developed Europe, from whom they had acquired the religion, proceeded henceforth without conflict. Far from it; but the fact that they adopted Christianity did mean that they could not be viewed as barbarian ‘outsiders’ in the same unrelenting fashion that classical authors had adopted towards all non-Romans. In 1002, shortly after his Polish progress, Otto III went on another journey, this time to meet his maker. Sonless, he was succeeded by his cousin Henry II, whose arrival on the throne inaugurated over a decade of warfare between the Holy Roman Empire and the Polish state. Much of this is lovingly chronicled for us by Bishop Thietmar of Merseburg, but his narrative is striking for its lack of any real demonization of the Poles, despite the ferocity of much of the fighting. That this was at least in part due to the Poles’ Christianity shows up in Thietmar’s criticism of Henry for employing still-pagan Elbe Slavs as allies against the Poles.7
In addition, the new Scando-Slavic states constitute yet another subject area that has benefited hugely from the Soviet archaeological bonanza of the postwar years. Originally, of course, the usual distorting agendas were firmly in play, but so much information became available that they were fast losing their credibility even before the Berlin Wall came down. And, in overall terms, the Communist years brought into the scholarly domain a vast amount of information that would not otherwise have come to light. All in all, then, both texts and archaeology provide a great deal of information about our new dynasties and the political structures they erected. What does all this material allow us to say about how these new states worked?
State and Periphery
Like their earlier counterparts on the fringes of the Roman Empire in the fourth century, these new entities in some ways fell short of modern conceptions of the word ‘state’. The largely Germanic-dominated polities of the Roman frontier zone had been limited in their capacity to undertake centrally organized action. Politically, they were confederative, which meant that their overall rulers had to coexist with other ‘kings’, who retained real power, if usually perhaps within one locality rather than over the group as a whole. They were also limited in the quantity of resources – human and economic – that they could redirect for centrally designated purposes. Royal retinues numbered only in the few hundreds, even if the groups as a whole could field total military forces of ten thousand-plus. They have likewise left us few signs of any capacity to erect and maintain fortifications or other types of monument. Nor were these entities particularly large, although Roman counteraction was partly to blame for this, and the realm of the Gothic Tervingi, further east, covered a substantial area from the Danube to the Dniester. On all these counts, the new political entities of northern and eastern Europe in the late first millennium were much more impressive.
Geographically, the new states of the ninth and tenth centuries were huge. The Rus state ran from Kiev to Novgorod in a north–south direction, and east–west from the Dnieper to the Volga. All told, this amounted to a staggering million square kilometres, or near enough. The other states, too, were much bigger than their late Roman counterparts. Bohemia was the most contained, but this name is more than a little misleading since the kingdom usually encompassed most of what is now Slovakia as well (Moravia in ninth- and tenth-century parlance) – a much bigger area than that dominated by any Roman client state. The Piast dynasty of Poland, likewise, customarily governed lands all the way from the Oder to Volhynia and Galicia beyond the River Vistula, again an unimaginably large territory in mid-first-millennium terms. Even Denmark was bigger than modern preconceptions would lead you to think. The Jelling dynasty quickly put together Denmark and the largest of the adjacent islands (Öland, Skåne and Sjaelland), but also made their presence felt in nearly all of the most fertile lands of southern Norway, particularly around Oslo Fjord and in what is now western Sweden. True to the best Viking traditions and proper first-millennium logistics, water united these different components, giving Harold Bluetooth and his son and grandson, Svein and Cnut, a large enough power base from which to conquer the populous and prosperous Anglo-Saxon kingdom in twenty years of warfare from the mid-990s.8
The disparity in profile between these entities and standard client states of the Roman period becomes even more marked if you look at governmental capacity: the kinds of powers they had available and the institutions they used to activate them. Archaeologically, the most striking legacy of these later states consists of castles. These new political authorities were capable of erecting by the score. By the year 1000, the Piast dynasty had dotted its domains with no less than fifty. The Premyslids, likewise, used garrisoned forts to control their central areas. In this, the tenth-century dynasties were following firmly in the footsteps of their ninth-century Moravian predecessors. Piast and Premyslid fortifications were constructed largely in wood (in case ‘castle’ brings to mind something anachronistically grand, along the lines of Edward I’s constructions in Wales), but the Moravians had quickly learned to build in stone, and for good reason. One of our chroniclers notes the dismay felt by Carolingian forces in 869 when they suddenly found themselves faced with the ‘insurmountable’ – probably stone – fortifications of Rastiz (perhaps Stare Mesto now in the Czech Republic). On previous campaigns, they had been able to burn their way through Moravian obstacles, but not any more. The Moravians also used fortified centres to control landscapes. Their key political centre of Nitra was surrounded by a ring of forts: Devin, Novi Voj, Kolyka and Bratislava.
Individual population centres in Kievan Russia, likewise, were fully fortified, but here the archaeology provides us with a more striking echo of Rurikid power. Running south and east from Kiev for over a hundred kilometres are the ‘Snake Walls’: ramparts originally three and a half to four metres high, reinforced with a twelve-metre outside ditch. These were constructed in the very early eleventh century (so it’s not really cheating to include them in a book that notionally stops at the year 1000) to counter the threat posed by the Pechenegs, the latest nomads to crash into the adjacent steppe north of the Black Sea. And if the new Slavic dynasties were the past masters of castle construction, the habit at least partly rubbed off on to the Scandinavians. One of the most exciting finds of postwar Danish archaeology was a series of fortified power centres datable, thanks to dendrochronology, to the reign of Harold Bluetooth. Named ‘Trelleborg fortresses’ after the first of them to be excavated, they vary in size but are all beautifully circular monuments with a symmetrical arrangement of large halls within. Otherwise, being an entity whose constituent parts were linked much more by water than by land routes, its ruling Jelling dynasty was less obsessed with castle-building. Nonetheless, the list of late first-millennium monuments impressively underlines the capacity of these new states to engage in concerted construction. The most any Roman client state could manage was to put the odd wall round a king’s hillfort in the case of the Alamanni, or try to repair an existing line of Roman fortifications in the case of the Tervingi, and even this much stretched group loyalty to breaking point. It is also unclear whether what were basically single fortified dwellings, such as those put up among the Alamanni, reflect the public power of a state or state-like entity – as both the regularity and the mass of late first-millennium construction do – or the clout of a particularly important individual.9
The capacity of these new states to raise and maintain troops was equally impressive. It had to be, of course, since to build multiple castles and not garrison them would have been a charmingly pointless exercise. We have no detailed evidence for Moravia, although the concerted efforts of various Carolingian rulers to dominate the territory, together with their ultimate failure, are eloquent testimony to the overall military power of the first of the new states. In the case of Poland the evidence is more specific. First, one of the Arab geographers tells us that a Piast king was capable of maintaining an armed retinue of three thousand armoured knights, paid for out of personal funds. The figure may be questionable, but not the nature of the force, since Boleslaw Chrobry had promised to aid the Emperor, whenever needed, with three hundred ‘armoured men’ as part of the archbishopric deal in the year 1000.
The key word in the original Latin is loricati, lorica being Latin for ‘coat of mail’. The rise to military predominance of soldiers equipped in this expensive manner – the mailcoat being the single most costly item of contemporary military equipment – was a revolutionary development of the late first millennium. The fact that the Piast retinue was so equipped emphasizes that they were fully up to date. And the promise to send three hundred men when asked is compatible with a total retinue size in the thousands, as Ibn Fadlan reports, since no one would ever agree to send anything like their full force to a foreign war. Equally important, this retinue was only one part of the Piast war machine, which rested on a military obligation imposed more broadly on at least some categories of the wider population. Again, the early eleventh-century sources don’t give us the full rundown, but in the campaigns against Otto’s successor Henry I we see a Piast army of many thousands which was capable of operating in detached divisions towards a common aim – as in 1003, when a force of three thousand men represented only one out of four Polish divisions engaged in the defensive holding effort against Henry’s imperial might. What hits you about all this is the cost. Germanic retinues of the late Roman period numbered only in the few hundreds, and the indications we have suggest that coats of mail were at that point restricted only to a small elite. The Piasts reportedly maintained retinues on ten times this scale and were able to equip them fully with all the latest hardware. We have no contemporary documentation on how the money was raised, but later arrangements give some idea. Areas under Piast control were administered from the nearest castle, and, of the revenues gathered, one-third went to the castle commander, presumably in part to maintain his garrison forces, and two-thirds to the king. There were other, perhaps more important, sources of revenue too, but it is likely enough that the ancestor of this later system was already in place to help the Piasts maintain their forces.10
The observable patterns of military power in the other states are similar. They had to be. The Piasts, Premyslids and Rurikids regularly fought one another, the military balance swinging, depending on circumstances (usually which of the states was in the middle of a dynastic crisis), first one way and then the other. This cyclical pattern would not have been possible had not all three been able to deploy military forces that were roughly equal in size and nature. As part of their treaty obligations with the Byzantine Empire, the Kievan Rus agreed to send to the Emperor, when asked, a military force several thousand strong. This was large enough to play a crucial role in keeping the Emperor Basil II on the throne in the face of a major revolt, and again underlines the overall scale of Kievan Rus forces, since this expeditionary force would have represented only a part of the total available. Again, these forces were composed partly of specialist retinues, who figure at many points in our narrative sources, and partly of contingents drawn from the major settlements of the realm. We have no figure for the size of retinues, but the RPC gives some detail on two of the territorial contingents. One from Novgorod figures strongly in a civil war of 1015, a second from Chernigov in another of 1068. Both are said to have numbered three thousand men. Retinues likewise appear in our early Bohemian sources, a more general military obligation only in later documents. But, again, I’m confident that retinues alone would never have been sufficient for the rulers of Bohemia to compete so successfully on the international stage.11
The Jelling dynasty was rarely involved in these interdynastic struggles, but it did have to fend off the hostile attentions of successive emperors, and, as we have seen, was perfectly capable of sustained aggressive warfare against the Anglo-Saxon kingdom under Svein (986/7–1014) and Cnut. Exactly how they did so is controversial. Did they use retinue and mercenary soldiers and/or forces levied under a more general military obligation to the state? In thirteenth-century Danish documentation, the levied force is called theleding, and in that century could produce for the king a fleet of notionally one thousand ships, each manned by forty warriors. At issue is whether any kind of direct ancestor of the leding system was used by Svein and Cnut, in addition to the mercenaries (lithsmen), whom they certainly also employed. In my view, it is extremely likely that they did. Assessing and mobilizing a military obligation is one of the basic powers of any ruler, and it’s hard to see that the Jelling dynasty’s power could have amounted to much if they were not able to do this in at least some of the territories they controlled. It is also suggested by some of the more detailed evidence. The more or less contemporary Encomium of Queen Emma, wife successively to Aethelred the Unready and Svein’s son Cnut, records that, in gathering his expeditionary force, Svein ordered that it should contain ‘no slave, no freedman, no low born’. This sounds like a general mobilization order, and certainly overseas Scandinavian societies, such as those created in the Scottish islands during the Viking period, quickly organized with clearly defined military obligations.12
The power of these central political structures was not limited to the waging of war. We’ve already met Vladimir’s Tithe Church in Kiev. Not only was this the Empire State Building of its day – at least as far as the Dnieper region was concerned – but it was just one part of a larger palace complex built by Vladimir on the Starokievskaia Hill. Two-storeyed stone halls, each over forty metres long, were built to the south, west and possibly also the north-west of the church. All were floored with glazed ceramic tiles whose design included the eagle, one of the oldest symbols of empire, and decorated with mosaics and paintings. Nor was this much grander than the best the rest had to offer. The grandest of the Christian basilicas discovered in Great Moravia was constructed at 
and covered an area of four hundred square metres, making it very similar in scale to the Tithe Church, although little is known of its decoration. This was one of twenty-five stone-built churches known to have been constructed in ninth-century Moravia, and there were probably many others of wood. Denmark and Bohemia, similarly, quickly acquired a stock of more and less impressive churches, not least their chief cathedrals at, respectively, Roskilde and Prague. As befits so early an independent archbishopric, however, the Piasts trumped their rivals in the religious arms race. The cathedral at Poznan was a monstrous three-aisled basilica covering no less than one thousand square metres, while the tomb of Adalbert at Gniezno was adorned by Boleslaw Chrobry with a solid gold cross said to have been three times his own body weight. It has been estimated that there were, in addition, another thirty to forty churches constructed across the Piasts’ Great Polish heartland by the time of Boleslaw’s death in 1025.13
The capacity of these new states to make things happen also extended to communications. Odd bits and pieces of relevant evidence turn up in the narrative sources: the construction of bridges and roads, for instance, features in the RPC. More generally, in the earliest monastic documents from Poland and Bohemia, labour dues owed for bridges and roads feature as royal rights which were never given up when a piece of land was handed over to the Church. The land’s labour force, in other words, would periodically be turned out to work on the highways at the ruler’s command. Some of what this meant in practice has been elucidated by Danish archaeologists. Another of their great postwar treasures is the Ravning Edge Bridge, dated conclusively, again by dendrochronology, to the reign of Harold Bluetooth. This was a kilometre long, part causeway, part raised bridge over a particularly soggy bit of central Jutland. It required four hundred separate sections and the small matter of seventeen hundred posts to complete. Not exactly the Golden Gate, it was still a magnificent piece of determined construction, typical of the kinds of enterprise required to make the boggier parts of the North European Plain reasonably amenable to land transport.14
Looked at under these different headings – and I have chosen only a few examples – the new states of northern and eastern Europe appear potent indeed. They enjoyed considerable powers over their constituent population. More elite elements could be made to turn out and fight, the poorer to build roadways, palaces, churches and fortifications. Economic resources could also be mobilized to support rulers and their extensive retinues, not to mention an associated Christian priesthood, which was growing apace under princely sponsorship. There is not the slightest doubt that their achievements dwarf the political structures that emerged on the fringes of the Roman Empire. Nonetheless, there were still some important ways in which the new states remained profoundly limited.
For one thing, they operated with little in the way of administration or written record-keeping, even if writing played a slightly more important role among them than in Rome’s client states of the fourth century. International treaties were on occasion committed to paper. The RPC includes the texts of two trade treaties made in 911 and 944 between the rulers of Kiev and the Byzantine Empire. All the internal evidence indicates that these texts are authentic, but the Chronicle was put together two hundred years later. The Papal Archives, likewise, contain a short but fascinating text known as the Dagome Iudex. A summary of it was copied into a register of Pope Gregory VII in about 1080 AD. Examined closely – the author mistakenly thought that the original was talking about Sardinia! – it turned out to be the last mortal remains of a late first-millennium international manoeuvre whereby in 991 the Piast ruler Miesco I (father of Boleslaw Chrobry) granted some kind of highly notional overlordship over his kingdom to the Pope in return for persuasive lobbying with the Emperor. In this case, the Polish original disappeared at some point; but clearly some of the diplomatic backdrop to Otto’s progress of the year 1000 was conducted on paper.15
Literacy also played some role in the management of internal resources, but within the period covered by this study, only a marginal one. The oldest written records of land grants from Bohemia date to about the year 1000. They detail royal land grants to favoured monasteries, and provide some insight into how kings shared their existing rights over people and their labour with the new religious foundations. But even in Bohemia such texts are few and far between at this date, and in most of the other new states it is the later eleventh century before such grants took a written form, the twelfth in Kievan Russia. As the physical monuments surviving from these states imply, these early documents show that rulers had well-established rights to produce and services, and like early Anglo-Saxon England these states were capable of assessing the economic potential of populations and landscapes and of recording the fulfilment of the obligations thereby derived, but its scarcity suggests that not much of this was happening on paper.
This picture is confirmed by the other kind of written document to survive from the early years of these states: formal codes of law. From before the year 1000, evidence for the distribution of rules and regulations in written form come only from Church contexts. Among the materials translated into the first written form of Slavic by Cyril and Methodius in Moravia, for instance, were two Byzantine texts of Church law: the Nomokanon. Canon law texts in written form put in a similarly early appearance in Bohemia: surviving examples date to the second half of the tenth century. But despite convincing mentions of specific royal edicts in the chronicle texts, and the surviving physical manifestations of rulers’ capacities to enforce them, these states produced no codifications of written royal orders dating to this era. The first secular law books from Poland and Russia date from the thirteenth century, and even these look more like codifications of existing custom than monuments to royal power; and prevailing practice, even this late, seems to have left much real legal power in the hands of more local authorities. Again, comparisons with western Europe help put matters into perspective. Church legal texts came to Anglo-Saxon England with the missionaries at the start of the seventh century, but it was not until the tenth that royal law-making started to take a consistently written form, and the later twelfth and the thirteenth that the English monarchy instituted the complex legal bureaucracy and record-keeping required both to make, and to make it possible for, people to bring their cases to centrally organized law courts.16
Bureaucratic underdevelopment, however, is not the main reason for regarding these new entities as only a limited form of state organization. Looking at the overall narrative of their collective histories in the period 950 to 1050, what’s really striking is their capacity to trade vast tracts of land between themselves, seemingly at the drop of a hat. Take, for instance, Moravia – broadly what is now Slovakia. This fell under Bohemian Premyslid control in the time of Boleslav I (929/35–967/72), then under Polish control under Boleslaw Chrobry in 1003, back to Premyslid in 1013, Piast again in 1017 and Premyslid again two years later. Moravia saw the fastest-moving game of Pass the East European Parcel, but other territories had analogous histories. Silesia and Wroclaw were under Premyslid control in the mid-tenth century, passed to the Piast Miesco I in 989/90, back to the Premyslids in 1038, and were only definitively ceded to the Piasts in return for an annual payment of two hundred and thirty kilos of silver and fourteen of gold in 1054. Cracow in southern Poland suffered from a similar Piast/Premyslid identity crisis. What is now south-eastern Poland, from the Upper Bug to the Carpathians, was similarly swapped, but this time between the Piasts and the Rurikids. Under Rurikid control from the time of Vladimir in 981, it swapped back to the Piasts in 1018, then back again to the Rurikid Yaroslav the Wise in the 1030s.
Similar patterns are observable, if on a slightly different scale, in the outlying regions of Jelling territory. Southern Norway around the Oslo Fjord was always contested with rival lords established further west: first Olaf Tryggvasson in the 990s, then the dynasty of Olaf Haraldsson from whom medieval kings of Norway were destined to descend. The west coast of what is now Sweden, likewise, was eventually wrested from Jelling control by kings of Sweden based further east.17 What all this makes clear is that it is anachronistic to think of these states as possessing clearly defined territorial boundaries. Over much of central and eastern Europe, the lordship of any particular dynasty was a highly contingent phenomenon.
At the same time, each of our dynasties’ landed possessions comprised a much more intensively governed core, over which rulers were able to maintain a consistent authority and which rarely, if ever, passed into the hands of dynastic rivals. The heartland of the Piasts was Great Poland, the territory centred on Gniezno between the Rivers Oder, Warthe and Vistula that Otto III visited in the year 1000. Its extent is clearly marked out by the spread of tenth-century Piast castles (Map 20). Premyslid rule in Bohemia, likewise, had the region around Prague as its core, a zone again defined by the spread of early Premyslid strongholds. Kievan Russia had a double core, as we saw in the last chapter: Novgorod in the north, the Middle Dnieper around Kiev in the south. Even in the much smaller Denmark, the Jelling dynasty ruled Jutland and the major islands much more directly and with a firmer grip than the larger region that at different moments found itself incorporated into Cnut’s Baltic Empire. In the worst Premyslid dynastic crisis of all, in 1003/4, Piast Polish garrisons came as far as Prague, but this was only the briefest of phenomena, as was a parallel Bohemian annexation of Gniezno in 1038. Otherwise, these central areas were securely under the authority of their respective dynasties, and we clearly need to think of these states in terms of ‘core’ and ‘periphery’: core territories subject to permanent, more intensive control, and peripheral ones that were liable to fall under the control of others as the power of individual dynasts waxed and waned.
This is a common early medieval pattern, typical of entities that rely less on bureaucratic structures for their cohesion and more on the power and charisma of individual monarchs. The latter was classically expressed by regular patterns of itineration, with the ruler making the circuit of his kingdom, consuming food renders with his attendant military retinues and involving himself personally, as he went, in the needs and desires of his greater subjects. This kind of personal government worked perfectly well in small kingdoms, but characteristically generated patterns of core and periphery when geographical scale increased, to the extent that it’s a broad rule of thumb that an early medieval ruler really governed only where he regularly travelled. All our evidence suggests itineration was the key mechanism of government in the new entities of northern and eastern Europe. The main economic right of the ruler referred to in the earliest Bohemian and Polish texts, for instance, consisted of food renders – the basic means by which an itinerant ruler fed himself and his entourage. For logistic reasons, food renders were always consumed close to source rather than transported to one designated royal centre. The larger Piast and Premyslid castles presumably served as the local collection centres for food renders.18
Kievan Russia had different origins, a circuit of political itineration not being central to the original Scandinavian merchants’ gathering of furs, slaves and other trade goods, even if these did tend to be gathered on winter circuits. By the later tenth century, however, itineration and a more regular pattern of early medieval government were being established. When all of the necessary logistic structure was put in place is unclear. The RPC records, however, that, apart from avenging herself upon the Derevlians for killing her husband, Igor’s widow Olga (c.890–969, regent 945–c.963) did much to establish towns, trading posts and hunting grounds in their territories, both much further to the north around Novgorod and further south around the River Dnieper and its tributary the Desna. Hunting was the main royal entertainment, and the main occupation of rulers’ retinues on most afternoons. The will of Vladimir Monomachus tells us that he used to go hunting a hundred times a year. The establishment of royal hunting preserves, therefore, was – in a bizarre way – an important moment in instituting a regular cycle of government. I suspect that Olga’s actions extended over a much wider area the kinds of institutions of rule and support that were already in place closer to the main governmental centre in Kiev.19
Danish kings of the Jelling dynasty, likewise, eventually became itinerant, and some of their early constructions, such as the Ravning Edge Bridge, were clearly all about making land travel more efficient, quite possibly with royal itineration in mind. It’s hard to see what else the Trelleborg fortresses were for, if not for a monarch’s itinerations. When first discovered, they were identified as purpose-built bases for the military forces of Svein and Cnut, who undertook the conquest of England. Their real date is actually too early for this, however, since they were constructed under Harold Bluetooth, Svein’s father. Nor could their regular layout have served any straightforward military purpose, as has often been noted. On closer examination, the deceptively identical interior buildings actually served a wide variety of purposes: some were equipped with fireplaces as residences or for entertaining, others served as storage sheds and yet others for craftsmen such as blacksmiths and even goldsmiths.20 The likeliest answer to the puzzle, in my view, is that they were built to extend Harold Bluetooth’s capacity to express practical political power by itineration, an interesting moment in the territorialization of the Jelling dynasty’s control.
The new states of northern and eastern Europe present us, then, with something of a paradox. Capable of highly impressive acts of government and of building power structures over huge geographical areas, they were at the same time fragile. Bureaucratically underdeveloped, they could govern only relatively small areas with full intensity, and larger peripheral areas were always liable to be lost to rival powers in moments of dynastic crisis. The rule of the itinerant dynasts provides most of the explanation for their at first sight paradoxical nature, but still leaves unanswered questions. Where had these dynasties come from, and how did they build up their power bases in the first place?
Dynasty
The year is 995, the place eastern Bohemia at the confluence of the Libice and Elbe Rivers on the morning of St Wenceslas Day. But there’s nothing cool, crisp or even about it, since Wenceslas Day falls on 28 September. Nor are we anywhere near the forest fence or St Agnes’ fountain. We’re with a group of men standing quietly outside the wooden castle of Libice, headquarters of the powerful Slavnik family, currently led by Sobibor, son of Slavnik. Four of his seven brothers are inside the compound, though he himself is on a visit to the Emperor in Germany. The quiet is broken by shouts and violence, orchestrated by Boleslav II, current head of the other powerful Bohemian dynasty the Premyslids, and nephew of Good King Wenceslas himself. The action is swift and decisive. At its close, the compound and castle are burned out, the Slavnik males and their retainers destroyed. Slavnik power had been eliminated once and for all in what was arguably the most efficient hit of all time: certainly on a par with that February morning in 1929 when six members of Bugs Moran’s North Side German/Irish gang were lined up against a garage wall – along with an unfortunate mechanic who happened to be in the wrong place – and gunned down by footsoldiers of the South Side Italian gang. The only thing missing in 995 was any pretence of an alibi. Unlike Al Capone, Boleslav II didn’t bother to arrange a holiday; in any case, there were no packages to Florida currently available.21
Not only is it a great story, but the St. Wenceslas Day massacre represents the culmination of the political process behind the emergence of Premyslid-dominated Bohemia. Thanks to its position close to the Frankish imperial frontier and its own precocious literary tradition, which gives us two clusters of home-grown texts from the tenth century (one surrounding Wenceslas in the 930s, the other Adalbert at the back end of the century), Bohemia also provides the best-documented case study of dynastic emergence. It also gets us in the right frame of mind for thinking more generally about the emergence of all the new dynasties of central and eastern Europe. There were some important differences of detail in the political processes involved, clearly, but there was also enough in common for Bohemia to provide us with a general model of how the new game of dynasty was being played right across central and eastern Europe.
From the historical sources, one dimension of the story is easy enough to tell, and pretty well known. As it emerged from the Avar Empire following its destruction at the hands of Charlemagne in the years after 800, Bohemia was subdivided into a number of separate political units with their own leaders (called duces in Frankish sources, but with the general meaning of ‘leader’ rather than something as grand and hereditary as the modern English ‘duke’ implies). The ninth-and tenth-century sources give us a series of snapshots, which between them strongly imply that the Premyslids emerged from a Darwinian process whereby these different ducal lines eliminated each other, until only one remained.
The first snapshot comes from 845, when fourteen duces from Bohemia presented themselves for baptism at the Easter court of the Frankish King Louis the German. Fourteen ‘leaders’ strongly implies that each ruled only a relatively small area, but ducal numbers quickly declined. In 872 only five Bohemian princes turned up at the court of Louis the German, and by 895 there were only two. Part of this picture of ducal decline may be historiographical accident. I am not convinced that the sources are full enough, for instance, for us to be certain that there were only two pre-eminent leaders left in the game as early as 895. This would imply that Premyslids and Slavniks then managed a century of coexistence before their final showdown, and this seems unlikely. But the basic picture is clear enough. State formation in Bohemia was the result of a political process – played out over pretty much two hundred years from Charlemagne’s destruction of the Avar Empire – which saw one ducal line eliminating its peers to bring an ever larger core region under its control. As with similar processes affecting Germanic groups earlier in the millennium, each stage need not have been as violent as the St Wenceslas Day massacre. Some of the other, originally peer, families may have been willing to accept demotion rather than demolition. Nonetheless, there is every reason to suppose that violence regularly punctuated the process.22
As far as we can tell, similar dynastic games underlay the creation of the two of our other new states that emerged from the wreck of the Avar Empire: Great Moravia and Poland. Great Moravia was the first to appear, in the middle decades of the ninth century. Carolingian sources for c.800–20 mention in passing a whole series of regionally based small-scale political leaders at the head of their own groupings, as Avar dominion in central Europe unwound. One called Vojnomir supported the Franks against the Avars, a certain Manomir appears briefly, while a major revolt against Carolingian rule was led by Ljudevit. The sources are nothing like full enough for us to attempt a political narrative of how these different dynasties combined and eliminated one another to produce the much larger power that was Great Moravia. But that they did is clear enough and, again, we are given the odd snapshot. The first really prominent Moravian ruler, perhaps the real founder of dynastic pre-eminence, was called Mojmir, and Carolingian sources record what was clearly a highly significant moment in the 830s (the incident can be dated no more tightly than 833–6) when he drove a rival prince, Pribina, out of Nitra in Slovakia to bring a broader region under his direct control. Using this greater power base, the dynasty continued to extend its control, as and when it could. Carolingian power kept its westerly ambitions in check through most of the ninth century, but as that Empire waned in the early 890s the Moravians extracted the right to exercise hegemony over Bohemia. From then on, a still more exciting range of ambitions might have been open to this ruling line, had not its career been cut decisively short from 896 by the arrival of the nomadic Magyars as a major force in central Europe.23
If we had nothing but the available historical sources, the emergence of Piast Poland would be particularly mysterious. The Piast state suddenly jumps into Ottonian narrative sources in the 960s, already fully formed under the control of the Piast Miesco I. With a heartland west of the Vistula, beyond the immediate border region between the Elbe and the Oder, the new Polish state was simply too far away from imperial dominions for our chronicle sources to observe its growing pains. Not even the Anonymous Bavarian Geographer knew the political layout of lands so far to the east. Thanks to the wonders of dendro-chronolgy, however, archaeological evidence – which is usually so much better at observing long-term development than the immediately political – has in this case brilliantly illuminated at least the final stages of the rise of the Piasts. Because the dynasty built its castles of wood – as pretty much everyone else in Europe was still doing in the first half of the tenth century – it has become possible, within just the last decade, to date their construction precisely. The results are revolutionary.
The emergence of the first Polish state used to be construed as a long, slow process of political consolidation, which gradually brought ever larger areas together under the control of a single dynasty. Long-term developments, as we shall examine in a moment, were certainly of critical importance to create the necessary conditions, but the archaeology has shown with striking clarity that the last stage of the Piast rise to power was sudden and violent. Piast castle construction clusters in the second quarter of the tenth century, demonstrating that the dynasty expanded its control over broader areas of Great Poland very quickly, from an originally narrow base (Map 20). More than that, in many of these localities Piast castles replaced a much larger kind of fortified centre, often dating back to the eighth century, many of which seem to have been destroyed exactly at the moment of Piast construction. The conclusion seems inescapable. The creation of Piast Poland, the entity that suddenly bursts into our histories in the mid-tenth century, involved the destruction of long-standing local societies and the imposition of Piast military garrisons upon them. How many of these local societies were ‘tribes’, for want of a better word – the kind of unit listed in the Bavarian Geographer for more westerly regions of the Slavic world – is unclear, as is how they were distributed across the landscape.24 Just like Great Moravia and Bohemia, then, the new Polish state emerged by violent dynastic self-assertion, as the Piasts eliminated their rivals at the heads of those other, older political units.
Much is also obscure about the rise of the Riurikids. As we saw in the last chapter, the Russian Primary Chronicle (RPC) is both too Kiev-focused and too much composed from the hindsight of achieved Riurikid domination to provide a straightforward route into the complexities of early Rus history. Nor – at this stage at least – is the archaeological picture so arrestingly precise as that for Piast Poland. Nonetheless, the basic outlines of Scandinavian intrusion into Russia are clear enough, and hence too the key political developments that made possible the Rurikid state.
Amongst its other problems, as we saw, the RPC provides a thoroughly unconvincing account of both the date and circumstances by which political power came to be transferred to Kiev from Gorodishche in the north. The Chronicle both places the transfer a generation too early and seems to be hiding dynastic discontinuity or at least disruption in its odd and seemingly sanitized account of the relationship between Oleg, the first major political figure associated with Kiev, and Igor, Riurik’s son and heir. The story also presents Oleg as gathering an army in the north and taking control of the Middle Dnieper by force. Yet it closes by saying that, at the conclusion of these operations, an annual tribute of three hundred grivny was imposed by Oleg on Novgorod in return for peace. This, the Chronicle notes, was paid until the death of Prince Yaroslav in 1054, which is late enough to fall virtually within living memory of the compiler of the Chronicle in the early twelfth century. So the payment is presumably historical. But why would a ruler who came from the north to conquer in the south, as the story has it, end up imposing a tribute on the north?
There are two other big problems besides. First, the trade treaties with Byzantium confirm that, well into the tenth century, non-Riurikid Scandinavians ruled their own Russian settlements with a great deal of independent power, since they had to be represented individually in the negotiations. Second, before the end of the tenth century the Chronicle preserves only a very simplified version of Riurikid dynastic history. From that point on, the transmission of power from one generation to the next always involved many contenders and civil war, but before that date, even though we know that the early princes were multiply polygamous (as indeed were their successors), the Chronicle mentions only one son at each moment of succession and nothing but a smooth transition of power.
None of this is credible. The 944 trade treaty with Byzantium tells us that Igor had two nephews important enough to rate a separate mention. There is no record of them or their subsequent fate in the rest of the Chronicle, and it’s hard to resist the conclusion that history has been edited to give an impression of secure and smooth Rurikid dominance. Likewise, the Oleg story: was he a collateral relative of Riurik who first conquered Kiev and then imposed his rule on the north? Or was he a complete outsider who perhaps married into the dynasty so as in some way to legitimize his rule after the fact? And how, then, did power pass from him to Riurik’s son Igor? It is also hard to believe that Oleg didn’t have heirs of his own, so what happened to them? The politics of early tenth-century Russia were clearly much messier than the Chronicle would have us believe, with independent Viking leaders and a self-assertive dynasty all jockeying violently for position.
The full details of these internal political struggles will for ever escape us, but the kind of world we should be envisaging is clear. At this stage, as one commentator has evocatively called it, it was not so much a state as a ‘glorified Hudson Bay company’, composed of essentially independent trading operations located at various centres along the main river routes, loosely linked together by having to pay protection money to the most powerful among them. They acted in concert only in certain circumstances, such as when using their collective muscle to extract advantageous trade terms from the Byzantines, and no doubt also to engage in a little price-fixing. The Rus state began life, therefore, as a hierarchically organized umbrella organization for these merchants, no doubt established originally by force. Even so, the original merchant adventurers, or their descendants, were left with considerable powers and independence, and as late as 944 still ran their own localities.25
By the eleventh century, however, this stratum of independent non-Riurikid rulers in their own settlements had disappeared. By this date, the preferred solution to the dynastic mayhem which characteristically accompanied transfers of power between different Rurikid generations took the form of giving their own centre of power to each eligible contender. This was already happening by the year 1000, the Chronicle providing us with an exhaustive list of the twelve cities that were granted by Vladimir to his twelve sons, the products of five of his more official liaisons. How many other children he had generated from the 300 concubines he kept at Vyshgorod, the three hundred at Belgorod and the two hundred at Berestovoe is not recorded. At some point in the tenth century, then, the independent power of the descendants of the founding merchant princes had been curtailed, turning their formerly self-governing settlements into dynastic appanages. In fact, this was probably a steady process, which played itself out over a lengthy period. Oleg’s suppression of Askold and Dir, to the extent that this story might be taken as historical example, provides us with an early instance of this kind of action. The RPC also records some later instances of exactly the same thing. In the civil war between Sviatoslav’s two sons Yaropolk and Vladimir, new merchant settlements continued to be founded. Two Scandinavian leaders by the names of Rogvolod (Ragnvaldr) and Tury established their own trading centres at Polotsk and Turov. Their subsequent fate is not recorded, but both centres were among the twelve distributed in the next generation to the various sons of Vladimir, by which time their founders had clearly lost out. As part of the same civil war, another such locally dominant line, apparently a family established for a much longer period, that of Sveinald, also met its demise.26 The full story of the suppression of the independent merchant lines is hidden from us, but it clearly happened, and it represented the final stage in the evolution of mercantile settlements into a fully fledged political union. Although the unique origins of the Rus state meant that the Rurikids began as one set of merchant princes among several – rather than as the leaders of one regional tribal group among several, as was the case with the Piasts and Premyslids – nonetheless violent dynastic self-assertion was central to the process of state formation.
The same was true of the last of these new states, Denmark, although here, too, the process differed substantially from that unfolding in the Avar successor states. In the small settlement of Jelling in central Jutland stands a not very substantial church and two huge mounds: the northern one 65 metres in diameter and 8 metres high, the southerly 77 by 11. Within the northern mound there is a wood-lined chamber dated by dendrochronology to 958, which was nearly the last resting place of King Gorm. Gorm’s son and heir Harold Bluetooth originally buried him there, but transferred the body to the church when he himself converted to Christianity, probably around 965. Like the Mormons, Harold was taking no chances that his ancestors might be deprived of the joys of his new religion. Apart from shifting the corpse, he also erected a fabulous runestone whose inscription is still there to be read: ‘Harold had these monuments erected in memory of Gorm his father and Thyre his mother, that Harold who won for himself all Denmark and Norway, and Christianized the Danes.’
The case of Denmark differs substantially from the other states, by providing a timely warning against the assumption that political developments always move in a straight line. As we have seen, a powerful centralizing political structure had existed in southern Jutland before the Viking Age, from at least the mid-eighth century when the Danevirke was first constructed. But this monarchy was destroyed by flows of new Viking wealth into Scandinavia. Wealth translated pretty much directly into warriors, and warriors into power, so that new wealth in sufficient quantities could not but generate political revolution. The old monarchy fell because so many ‘kings’ could now buy in so much military muscle that political stability evaporated.27
By the mid-tenth century, there are further signs of substantial change. For one thing, there seem to have been fewer kings. Viking-period sources demonstrate that a multiplicity of royals had existed in ninth-century Scandinavia. Apart from the one extended, or possibly two, separate dynastic lines found competing for power in southern Jutland (Godfrid, Haraldr and their descendants), there were more independent kings in the Vestfold west of the Oslo Fjord in Norway in the ninth century, and on the island of Bornholm. Birka and Sweden, further east, likewise, also had kings. A large number of other kings also appeared in western waters in the Great Army period, from the 860s onwards, and these must all have had their origins in some particular corner of Scandinavia. By my reckoning about a dozen of them are named, at different points: not enough to suggest that ‘king’ was a status that just anyone might claim, especially as we also meet men of slightly lesser status – jarls – who held back from claiming to be royalty. From the time of Harold Bluetooth, by contrast, the historical narrative throws up other ‘kings’ consistently in Sweden only, and occasionally in Norway. It would appear, therefore, that the word had undergone a change of meaning (as it did in other cultural contexts, too) from something like ‘person from an extremely important family’ to ‘ruler of a substantial territory’, the normal meaning of the word today.28
That said, the Jelling dynasty did seemingly build up its power by bringing under its control disparate territories that had had their own leaderships in the chaos of the later ninth century. It may have been the dynasty’s success, of course, that brought about the substantive change of meaning in the word ‘king’. Gorm’s wife Thyre is called in another inscription ‘the pride of Denmark’. It has been convincingly argued, on the basis of contemporary usage, that in c.900 the ‘mark’ element in ‘Denmark’ meant ‘regions bordering the Danish kingdom’; in other words, somewhere other than the main centres of the Danish monarchy – perhaps northern Jutland or the southern Baltic islands. Like our other dynasties, therefore, despite the substantial differences in historical context, the political activities of the Jelling dynasty were fundamentally accumulative – putting together regions that had previously been independent. This process was begun by Gorm and carried on by subsequent members of the dynasty. Harold Bluetooth added control of southern Norway to the dynasty’s portfolio of assets after the battle of Limfjord, but ruled it indirectly through the Jarls of Lade. Svein and Cnut maintained this hegemony through most of their reigns, and at times dominated the west coast of what is now Sweden as well. Even so, the heritage of old independence did not disappear overnight. From the narratives of Danish history in the eleventh century, it emerges very clearly that Jutland and the islands of Fyn and Sjaelland were still functioning on occasion as separatist power centres.29
The political processes behind all these new states, therefore, were similar. In each case, one dynastic line was able to demote or eliminate a peer group of geographically proximate rivals to bring a larger region under its control. The vagaries of this process further explain the propensity of the states it created to swap intervening areas amongst themselves. Given that all these areas were originally independent, it is easy to see why some of them might maintain a capacity for autonomous political activity long after they first accepted a new dynasty’s domination, especially in a context where itineration and personal charisma rather than developed bureaucratic structures were being used to govern them. But while full of arresting stories and individuals of striking charisma, political narratives of achieved dynastic ambition do not remotely begin to tell the full story of state formation in the north and east at the end of the first millennium. History is littered with ambitious individuals trying to build their power and thereby eclipse every rival. In most cases, however, such ambition does not lead to new and impressively powerful state structures. Apart from looking at narratives of personal ambition, then, we also need to think about the broader structural transformations that made it possible for entirely ordinary ambitions to achieve such unusual outcomes.
State-building
Many of these changes were similar to those that had generated the larger political structures on the fringes of the Roman Empire in the first half of the millennium. Taking the long view, social and economic transformations of the most profound kind were structurally critical to the process of state formation in northern and eastern Europe. This is most obviously true of the Slavic-speaking world, but to a considerable extent applies to Scandinavia as well.
Up to the mid-first millennium, Slav or Slavic-dominated societies were characterized by little in the way of obvious social inequality. Whatever their exact geographical origins, the Slavic-speaking groups who burst on to the fringes of the Mediterranean in the sixth century had clearly emerged from the undeveloped, heavily wooded regions of eastern Europe, where settlements were small – no more than hamlets – and whose Iron Age farmers were operating at little above subsistence level and with few material markers of differing social status. This state of affairs had already begun to change radically in the sixth century, as a direct result of the migratory processes that brought some Slavic-speakers into a direct relationship with the more developed Mediterranean. From this, an unprecedented flow of wealth – the profits, more or less equally, of raiding, military service and diplomatic subsidy – quickly generated inequalities around which new social structures began to form. These showed themselves initially after c.575 in the rise of a new class of military leader, controlling quite substantial areas and groups several thousand strong – even if there is also some reason to think that other elements within Slavic society, represented by Korchak remains, retained older, more egalitarian social forms and were even using alternative kinds of migration, and in different directions away from the east Roman frontier, as a means to preserve them.30
The new Slavic states of the ninth and tenth centuries were constructed on a marked accentuation of these initial inequalities. This shows up most obviously in the existence of military retinues: that classic vehicle of social and political power, which had played such an important role in the transformation of the Germanic world. Presumably the new Slavic leaders of the sixth century had their henchmen, but large permanent retinues do not figure in any of the historical sources as a major force, military or social. The contrast with the ninth and tenth centuries is striking. The Arab geographers report that Miesco of Poland maintained a personal force of three thousand warriors – and this is just one account among many, stressing the importance of retinues at this time. In Bohemia, the fourteen dukes presenting themselves for baptism in 845 did so ‘with their men’, and the early Bohemian texts associated with Wenceslas refer both to his retinue and to that of his brother, Boleslav I. Frankish texts, similarly, mention the ‘men’ of both Mojmir and his nephew Zwentibald among the Moravians, and retinues were just as important in Russia. Again, Arab geographers pick out the four hundred men of the dominant Rus prince in the north in c.900, and retinues appear as important political pressure groups for several of the early kings in the narratives of the RPC. It was the need to satisfy the demands of his ‘men’, for instance, that led Igor to increase the tribute he customarily imposed upon the Derevlians. He may have regretted giving in, since, as we have seen, it led to his death at the hands of the aggrieved taxpayers. And as we saw among the Germani around the Roman Empire, the rise of permanent military retinues greatly increased the capacity of rulers both to bring rival dynasts into line and to enforce a range of obligations (such as army and labour services) upon the broader population. As such, it obviously played a critical role in the process of state formation, not least – again as among the Germani – in creating a much stronger dynastic component to power at the top. There is no sign among even the late sixth-century Slavs that power was in any sense hereditary, even if particular individuals could build up striking power bases.31 But by the ninth and tenth centuries, dynasties dominated politics, and hereditary power was the order of the day.
But retinues were only one aspect of a broader process of social change. Part of the problem in understanding this bigger picture as fully as one would like to stems from uncertainties about its starting point. The idea of a highly egalitarian Slavic world in c.500AD is strongly entrenched in both the scholarly literature and more popular mythology. It underlies the ‘happy hippy’ vision of Slavicization, and finds real support particularly from east Roman sources which note that sixth-century Slavic society was marked by a lack of structured social differentiation, as well as being unusually willing to take on prisoners as free and equal members. But such visions of Slavic equality need to be tempered with some caution. To echo a point made earlier in the case of the Germani, there are entirely non-material ways in which higher status can be all too real – if those enjoying it had to work less hard, enjoyed more food, and if their word counted for more when it came to settling disputes within the group.32
But even if we factor in a less egalitarian starting point for the evolution of Slavic society from c.500 (and as we have seen, any preexisting egalitarianism was being undermined rapidly at this point by the twin processes of migration and development), much had clearly changed by the tenth century. Not only was political leadership now hereditary and its clout more wide-reaching thanks to its permanent military retinues, but Slavic society as a whole was marked by the evolution of clearly differentiated, and therefore presumably also hereditary, hierarchically arranged social categories.
At the bottom of the social scale, unfree population groups now played a prominent role in all our late first-millennium Slavic and Scandinavian societies. The slave trade was a major phenomenon of central and eastern Europe from the eighth century onwards. Likewise, as the new state structures evolved, a major component of their economic makeup (as we shall explore in more detail in a moment) became the unfree ‘service village’. Given the available sources, it is not entirely clear whether the populations of these villages enjoyed a higher status than the slaves who were often exported – perhaps a status akin to those of the permanent ‘freed’ (better, ‘semi-free’) populations we encountered in the Germanic world. Either way, a large part of Slavic humanity had clearly been permanently relegated to hereditary lesser status (or statuses, if slaves and service villagers need to be distinguished) by the tenth century. However you model Slavic society in c.500 AD, the extent of change here can hardly be overemphasized.
Equally permanent at the other end of the social scale was a class of high-status individuals, often styled optimates in our sources. These men are recorded, for instance, attending assemblies in Bohemia to give their approval to their choice of Adalbert the Slavnik as Bishop of Prague in 982, and feature as the rulers of their own settlements under overall Rurikid rule in Russia (some of them independent enough to send their own ambassadors to Constantinople when the trade treaties were negotiated). Certain individuals with the same kind of high status also appear in the train of the King of Poland in the early eleventh century, and were presumably the kind of men Polish and other kings customarily offered hospitality to as they feasted their way round their kingdoms. Their existence a century earlier in Moravia is possibly also reflected archaeologically in the five so-called princely dwellings found in the hundred-hectare outer area at , although these could have belonged to junior members of the ruling dynasty. The evidence suggests that this group coalesced out of originally three component elements. First, there were the close supporters of the new dynasty from within their home group. These were reinforced, second, by the elites of originally independent units (whether of Slavo-Scandinavian trading enterprises in Russia or ‘tribal’ units in Bohemia, Moravia and Poland), who accepted the new dynasty’s domination; and, third, junior members of the ruling line. Before and even after they accepted Christianity, polygamy was usual – which made such junior royals a correspondingly numerous group, especially with a ruler like Vladimir who, as we know, numbered his concubines in the hundreds. Over time, the three became indistinguishable, between them eventually providing the nobility of the fully fledged kingdoms.
As in the earlier Germanic world, there was also an extensive free class operating in between the nobility and the unfree. They appear in some of the written legal sources from all the major kingdoms, except Moravia, which didn’t last long enough to have any. Based on parallels with the rest of late first-millennium Europe, this group probably provided the bulk of military forces deployed by these kingdoms, beyond the specialist military retinues of the rulers. Elsewhere, it was customary for unfree populations to perform lower-status functions, such as providing the labour with which many of the striking monuments of these kingdoms were presumably constructed. Military service, by contrast, was higher-status, despite its obvious dangers.33
Even if you don’t believe in an entirely egalitarian sixth century, Slavic society underwent a total restructuring between the years 500 and 1000. The sixth-century Slavic world evolved leaders who rose and fell in their own lifetimes, with no markedly hereditary element to their power. There is also no sign of a hereditary nobility, or of permanently unfree population groups. This might still have been true of at least some Slavic groups in the seventh century. The fact that a Frankish merchant, a complete outsider, like Samo, could still at this date be elected overall leader among a multiplicity of other Slavic duces would seem to indicate that such men were not sitting on top of a strongly hierarchical or hereditary social pyramid. But this had changed by the tenth century, and, equally important, the new states of the later period could never have appeared without these intervening transformations. Heritable dynastic power, the social and military clout of retainers and nobles, and a differentiated population who could be forced and/or persuaded to undertake necessary functions such as providing food and labour or military service: all of these were key elements of the new state structures, and none had existed in the sixth century.
The question of when the different elements of the restructuring had happened is very difficult to answer. The likeliest answer, as is so often the case, would appear to be mixed. Some of the change looks on balance to have had long roots. When it emerged from Avar control after Charlemagne’s campaigns of the 790s, central European Slavic society already had the capacity to throw up powerful princes. Within a couple of decades, the chronicle sources give us a cluster of leaders – Voinomir, Manomir and Liudevit – capable of mobilizing significant military power for a variety of ends. This degree of control seems unlikely to have sprung up overnight and had probably been generated during the Avar period. This is also suggested by the fact that in both Moravia and Bohemia we find leaders – duces – with well-entrenched hereditary power over particular localities as early as the mid-ninth century. On the other hand, this observation needs to be balanced by the fact that most of the hillforts built in the Slavic world up to the ninth century appear to have been communally generated places of refuge. On excavation they characteristically lack any sign of an elite dwelling (often any permanent dwellings at all) or any other sign that the guiding hand of some great man was behind the project.34 If a class of hereditary leader had emerged by c.800, then, it is important not to overestimate the extent of its social power.
Equally important, state-building had powerfully transformative social effects. Most obviously, the increasing wealth of particular dynasties led to the generation of retinues of increasing size and power. At the same time, much of the nascent nobility of the new states was a by-product of dynastic self-assertion, whether from the promotion of supporters and junior relatives, or the demotion of previously independent regional leaders. Castle-building in Bohemia, Moravia and Poland also involved the destruction of the old communal refuge-type hillforts and their replacement with new dynastic castles. And while the slave trade had certainly begun in the eighth century, it gathered pace dramatically in the ninth and tenth. Both of these latter developments probably played a major role in increasing the number of the unfree in the population (if not, perhaps, in initially generating them). My own best guess, therefore, would be that a longer, slower process of evolution had generated a body of hereditary group leaders by c.800 AD, but that the process of state formation after the collapse of the Avar Empire further revolutionized the situation.
How much of this broad model of social transformation is applicable to Denmark is a different question. State formation in Denmark differed substantially that in other cases, because there it was a question of state re-formation. A state-like structure comparable to the Slavic and Scando-Slavic ones already existed in southern Jutland and the islands from the eighth century, before being destroyed by the new wealth introduced into Scandinavia in the Viking period. As this would suggest and the sources confirm, Scandinavian society entered the last two centuries of the millennium with more entrenched social inequalities than was true of the Slavic world. Viking-period sources show us kings, jarls, freemen and unfree (thralls) already fully in existence. This makes good sense, not only because state-like structures already existed there but also because Scandinavia, or Denmark at the very least, had been part of the Germanic world (if belonging to its outer rather than inner periphery), interacting with the Roman Empire in the first half of the millennium, and had participated – as flows of Roman goods and bog deposits of weapons indicate – in some of the earlier processes of sociopolitical transformation. State formation in Denmark in the later ninth and the tenth century was probably much more a story of the promotion and demotion of existing power blocks and the dynasties at their heads, than of the fundamental social change that was central to the process among neighbouring Slavs.35
Social revolution, part cause and part effect, was absolutely necessary, therefore, to state formation at this time, at least in Slavic lands. But social change on this scale is never possible without parallel economic restructuring, and, again, there is plenty of evidence of this from contemporary central and eastern Europe. As with the social transformations, some of this preceded the formation of states and was a necessary precondition for their existence. Further change was then instituted by states themselves.
The hardest to document in a precise way is the development of the agricultural economy: food production. Not least, of course, is the fact that Slavic-speakers had come to dominate such a huge territory, with such a vast range of environments, that farming did not and could not have developed everywhere along a single trajectory. Nonetheless, the evidence indicates strongly, if at this point still rather generally, that farming outputs increased dramatically. At different speeds in different contexts, a revolution was under way that was bringing more land into production and instituting more productive farming practices, both in terms of the technologies employed and the management of land fertility. Most obviously, there was a substantial amount of forest clearance in central and eastern Europe in the second half of the millennium. In those parts of Poland with the right kind of lakes for taking pollen cores, the ratio of grass and tree to cereal pollen declined dramatically in these centuries from 3:1 to much more like 1:1, suggesting a doubling in the amount of cultivated land. This result can’t be simply applied to the whole of Slavic Europe. I would expect the degree of change to have been less, for instance, as you moved north and east. Nonetheless, even within Mother Russia, the spread of Slavic-type cultures in northerly and easterly directions was closely associated with the spread of full-scale agriculture in the wooded steppe and reasonably temperate forest zones of the East European Plain. The phenomenon of general agricultural expansion is clear enough, then, even if it is impossible to put figures and dates on its impact in particular localities.36
The spread of more efficient farming techniques, too, is easy enough to document in outline. Initial contacts between Slavic-speakers and the Mediterranean world led some Slavs to adopt more efficient ploughs, types that turned the soil over so that rotting weeds and crop residues released their nutrients back into it. This increased both the yields that could be expected and the length of time particular fields might be cultivated. Further improvements had not yet worked themselves out fully by the time these states came into existence in the ninth and tenth centuries. The height of medieval sophistication in arable production, for instance, was the manor. Its advantages lay in the fact that it was a self-contained, integrated production unit with a large labour force, where farming strategies could be centrally directed towards greatest efficiency, particularly when it came to crop rotation for maintaining fertility, and costs (particularly of ploughing equipment) could be pooled and minimized. It was also an instrument for brutal social control, but that’s another story. For present purposes, the point is that arable production in central and eastern Europe became fully manorialized only from the eleventh century onwards – after the new states had come into being. This finding came as a nasty shock for doctrinaire Marxists, since these were all supposed to be ‘feudal’ states, whose development was only made possible by mano-rializing agriculture, but the chronology is secure enough. Even if manorialization was still only nascent in the ninth and tenth centuries, however, we do have evidence that some key preparatory changes were under way. In particular, the amount of rye found in pollen cores in the second half of our period increases steadily. The use of rye, sown in the autumn rather than the spring, is associated with moves towards three-crop rather than two-crop rotation schemes. Three-crop schemes both increased the amount of land under cultivation at any one time (two-thirds rather than just a half) and preserved better fertility. This development perhaps also underlies the observation of the Arab geographers that Slavic populations gathered not one but two harvests each year.37 There were important further developments yet to come, but much more food was being produced in central and eastern Europe by the year 1000 than had been the case five hundred years earlier.
This was critical to state formation in a number of ways. Until food surpluses were generated in substantial quantities, it was quite impossible (as was also the case with the Germanic world in the first half of the millennium) for leaders to maintain large specialist military retinues and other functionaries not engaged in primary agricultural production. Without economic surpluses, likewise, it was impossible for social differentiation to become entrenched. More food also meant more people,38 and state formation could probably never have happened as it did without this increased population. It provided all the extra manpower required for ambitious construction projects. More important, but harder to measure in concrete terms, increasing the population density in central and eastern Europe increased substantially the competition for available resources. The need to belong to a group in order to flourish has always been a powerful reason why individuals are willing to accept the costs that usually accompany group membership. Put simply, one reason why peasants – or some of them – were willing to pay food renders and do labour service for rising dynasts was that they offered sufficient military organization to guarantee safe land retention in return.39
But much more changed in barbarian Europe between 500 and 1000 than the appearance of more people producing more food. Other economic developments were just as important to state-building, or nearly so. Military retinues, for instance, needed arms and armour, and, as we have seen, the new rulers of the tenth century controlled substantial reserves of precious metals (witness the gold cross erected by Boleslaw Chrobry over Adalbert’s tomb) as well as all the other resources, apart from the mere physical labour, required to erect, decorate and furnish prestige projects such as the cathedrals and palace complexes that are such a marked feature of the period. In part, the rulers’ capacity to do so arose from some general processes of economic development affecting the whole of central and eastern Europe at this time, processes beyond their control. In part, it was fostered by particular policies adopted by the dynasts themselves.
The biggest non-agricultural economic phenomenon in these years was the rise of an international trade network in furs and slaves. Some of its western axes began to operate in the seventh century, but it was in the eighth that it stretched into the Baltic, and the early ninth before it exploded more generally across eastern Europe. We’ve already encountered the role of waterborne Scandinavians in making this network’s longer-distance connections work, and its central role in the whole Viking phenomenon. Consequently, the period saw a huge outflow of raw materials – largely people and furs – from northern and eastern Europe, and a flood, in return, of due payment. Large quantities of Byzantine silks were presumably one part of this, but few traces of these materials survive in our written sources, and none archaeologically. Beautifully finished silks were the main item Byzantium had to offer in exchange for imported items. What the archaeological record has produced in vast quantities, however, is silver coinage, above all from the Muslim world but from western Europe as well. These coins survive in quite astonishing quantities: as we saw in Chapter 9, over 220,000 Muslim coins in hoards of five or more according to recent estimates. This is all the more impressive given that silver has never been without value. The coins that survive are probably no more than the tip of a silver iceberg that has been reworked many times in the intervening centuries.
Kings, of course, were hugely interested in such massive flows of wealth. Not least, they could milk it for tolls, offering merchants safe places for making their exchanges and charging fees in return. Right at the beginning of the ninth century, the then King of pre-collapse Denmark, Godfrid, forced merchants who’d previously operated on Slavic territory at the trading centre called Reric to move to his newly built trading station at Hedeby in southern Jutland. The move is recorded in a contemporary source, and has found archaeological confirmation. The dendrochronology dates of early timbers recovered from Hedeby show that, as the text reports, it was built in c. 810 when Godfrid was at the height of his powers. The king’s interests here can only have been financial. Prague, likewise, one of the key centres of the Premyslid dynasty of Bohemia, was also, as the Muslim geographers report, a major entrepôt in the slave trade. The tolls must greatly have swelled the coffers of the dynasty, and the trade’s importance was such that one of the reasons given for Wenceslas’ killing is that he was attempting to outlaw it.
Kiev, too, new home of the Rurikids in the tenth century, was a trading entrepôt of huge importance. Byzantine and Russian sources both confirm that it was the starting point for the Rus trade fleets that came to Constantinople every spring from the early tenth century onwards. And by the early eleventh, Thietmar of Merseburg tells us, the city boasted no fewer than eight markets. Only in the case of Poland do we lack explicit textual evidence of its participation in the new international trade networks, but this can only be an accident of (non-)survival. The territory of the Piasts has thrown up such a density of Muslim silver coin finds that there is no doubt that its population was in some way involved in the new trading networks.40 So all of our new dynasties effectively tapped in to the new wealth being generated.
Nor was their role confined merely to taking tolls. They also sometimes interfered actively to reshape, as it were, the networks and maximize their own profits. This is easiest to demonstrate in the case of the Rurikids. In the tenth century the dynasty mounted collective military action on two occasions, in 911 and 944, to force the Byzantine authorities in Constantinople to grant Rus traders increasingly favourable trade terms, including the stipulation that they would get free bed and board inside the city for a month while conducting business. Not surprisingly, given the origins of Scandinavian interest in Russia, the treaties show us that members of the Rurikid dynasty (and not just its current head) were themselves active traders, so they had every reason to want to increase activity and market share. But I would not jump to the conclusion that this was an enterprise limited to the Rurikids. Byzantine–Rus trade connections just happen to be comparatively well documented, and it seems to me entirely likely, although undocumented, that other dynasts took an equally active interest in developing international trade links along the particular lines that best suited them.41
Even though we don’t know as much about this as we would like to, a good case can be made that all these dynasties took a proactive role in organizing the economies of at least their dynastic heartlands. The picture has emerged from a mix of historical sources and archaeological investigation. Looked at archaeologically, the tenth-century dynastic heartlands of Poland and Bohemia are striking for the relative density of their populations. Once again, this observation is based upon detailed knowledge of pottery chronologies, whose spread gives you a reasonable guide to the existence of settlements. The evidence suggests that this population density was not an accident, but the outcome of deliberate interference. As the archaeology has shown, a key moment in the rise of both Premyslid and Piast dynasties came when they destroyed the defended centres associated with the old sociopolitical structures, or ‘tribes’, and replaced them with their own chain of castles, in the later ninth and the earlier tenth century respectively. This process, it seems, was accompanied by the deliberate transfer of at least some of the subdued population groups to the dynasty’s heartland. Some transfers are mentioned in our sources. The RPC, for instance, associates Prince Vladimir in the late tenth century with a mass transfer of various population groups – Slavs, Krivichi, Chud and Viatichi – to different places along the River Desna. Here, persuasion rather than force may have provided sufficient incentive for the move. In other cases – Poland and Bohemia – all we have is the archaeological reflection of the effects of such a process in the sudden creation of an unusually dense population cluster, but early texts (all gifts to monasteries) from both these kingdoms (and, indeed, from Russia too) make clear the purpose of these resettlements. The classic economic form to emerge from these early texts is the ‘service village’, already mentioned. These unfree villagers were required to fulfil particular economic functions for the king, such as bee-keeping or horse-breeding, in addition to providing standard food renders. The fact that they were unfree strongly suggests their origins lay in forced resettlements.42
Kings – or kings and their advisers – were economically alert enough, therefore, to maximize the exploitation of their dynastic heartlands. The way this was done suggests that they were operating in a world where there was little in the way of a functioning market economy in agricultural goods, since instead of simply being able to purchase the required items, specialist tasks had to be assigned to particular settlements. This is not surprising. The same was true of the ninth-century Carolingians who were still using service villages in certain areas, and is in line with both the coinage evidence and the fact that peasants owed the king food renders rather than cash taxes. North and eastern Europe was still at the time an economy lacking small change. Muslim silver coins were plentiful enough, but these were of much too high value to do your everyday shopping with. Use one of them at the baker’s and you’d come home with a few sackfuls of bread, which would have gone stale long before you ate it. Likewise, although kings generally preferred cash taxes because they were infinitely more flexible, they could demand them only when the possibility existed for peasants to sell on any surplus production to merchants.
All of this adds another dimension, of course, to the tendency of these late first-millennium states to operate with a distinct centre–periphery dichotomy. Not only was this an accidental offshoot of the logistic limitations of itineration, but it reflected something more fundamental about the states’ construction. Thanks to the policies of the triumphant dynasties, core and periphery were also distinct in population density and economic organization. In the case of the Kievan Rus state, because of the peculiarities of its origins, the process of core creation had an additional importance. Up to the mid-tenth century, the Riurikid dynasty displayed a distinct capacity to shift its centre of operations about. It first came to prominence in the north with an initial seat, it seems, at Gorodishche (old Novgorod) before transferring, as we have seen, south to Kiev on the River Dnieper in the late ninth century, when Oleg came south with his army.
The reasoning behind this transfer requires more than a little puzzling out. At first sight, it seems an odd move, since Gorodishche, as noted earlier, was better placed for controlling trade flows along the Volga to the Muslim world, which was a far richer trade route than its counterpart along the Dnieper to Constantinople. Kiev, however, had other advantages. Situated on the wooded steppe, Kiev was excellently placed to dominate the surrounding regions of what is now Ukraine, which had become home in the seventh and eighth centuries to a large Slavic agricultural population. This was organized into units with their own substantial political structures before the arrival of any Scandinavians. The Polian dominated the area immediately around Kiev, the other groups in the vicinity being the Derevlians, Severians, Radimichi and Dregovichi (Map 19). While less well placed in purely trade terms, Kiev offered Scandinavian dynastic wannabes far more in the way of exploitable human and economic resources. Already in the time of Oleg, the RPC tells us, the Grand Prince’s army consisted not just of Scandinavians, but of Slavic- and Finnic-speakers. For the Grand Prince, who wanted to be much more than a merchant prince, Ukraine had much more to offer than the north. Even so, the Riurikids were not quite yet ready to give up the gypsy life. Oleg’s son and heir Sviatoslav engaged in wide-ranging campaigns, as far east as the Volga and south all the way to the Caucasus. The RPC reports that, just before his death, he was contemplating relocating the capital of the dynasty a third time – to the Danube. The work of Sviatoslav’s son and eventual heir Vladimir, in generating a much stronger economic core in and around Kiev along the Desna, had the particular effect of rooting the Riurikid state once and for all in its Ukrainian heartlands.43
There is, of course, much else we’d like to know about how state-building intersected with the economic development unfolding in the later first millennium. One huge gap is the lack of detailed information about iron-mining and steelworking. The kinds of armies deployed by the new dynasts imposed a huge demand for these items, but there is no good information on how this was satisfied. Nonetheless, the big picture is clear enough in outline. State-building would have been impossible without a number of pre-existing social and economic transformations of fundamental importance: the generation of a much more productive agriculture, the substantial population increase that followed on from this, a huge increase in the amount of movable wealth, and the more developed social hierarchies that formed around its unequal distribution.
But if these huge changes provided the necessary backdrop, the dynasts were themselves responsible for further social and economic developments of profound importance. On the trading front, they not only milked the new international networks for all the tolls they could get, but actively extended them, as and when they could. At home, likewise, agricultural production in the core areas of their kingdoms was maximized. None of these processes was complete by the year 1000. Not only was agricultural production reorganized on a manorial basis from the eleventh century, but the grandsons and great-grandsons of the dynasts would also famously employ recruiting agents to bolster output still further by making hundreds of thousands of German peasants offers they could not refuse to move eastwards.44 Even if all this was still work in progress by 1000, we have arrived at a partial answer to our question. Normal dynastic ambition produced entirely abnormal results during this period because deeply rooted social and economic transformation meant that the dynasts were pushing at a door that was already opening.
But this, at best, is still only half an answer. Dynastic ambition provided the activating mechanism for state formation; being both cause and effect of the political revolution we have been observing, massive social and economic transformation was its necessary backdrop. But what underlay those initial social and economic transformations that gave such free range to dynastic wannabes to remake the map of central and eastern Europe?