III
The immediate answer to the question posed in the title is given with characteristic dry clarity by James Murray in that great work of English history the Oxford English Dictionary. Murray’s first definition is “English Saxon, Saxon of England: orig. a collective name for the Saxons of Britain as distinct from the ‘Old Saxons’ of the continent. Hence, properly applied to the Saxons (or Wessex, Essex, Middlesex, Sussex, and perhaps Kent), as distinct from the Angles.” After explaining that, “in this Dictionary, the language of England before 1100 is called, as a whole, ‘Old English,’ ” Murray then goes on to say that the adjective “Anglo-Saxon” is “extended to the entire Old English people and language before the Norman Conquest.” Neither he nor the Supplement mentions explicitly the almost purely chronological use of “Anglo-Saxon” to describe the whole period of English history between 400 and 1066 that is now current, but it is easy to see how this has derived from the usage they expound.1
What the original edition goes on to do, moreover, is to give an account of a wider use of the word that beautifully encapsulates the beliefs about culture and descent that lie behind it. The expression “Anglo-Saxon,” according to the Oxford English Dictionary, was then—that is, in the late nineteenth century—used “rhetorically for English in its wider or ethnological sense, in order to avoid the later historical restriction of ‘English’ as distinct from Scotch, or the modern political restriction of ‘English’ as opposed to American of the United States; thus applied to (1) all persons of Teutonic descent (or who reckon themselves such) in Britain, whether of English, Scotch, or Irish birth; (2) all of this descent in the world, whether subjects of Great Britain or of the United States.” Part of this is now out of date, thanks to the increased use of the words “Britain” and “British,” rather than “England” and “English,” for the United Kingdom and its inhabitants; to the growth of Irish, Scottish, and Welsh nationalism; and to the fact that large parts of the “Anglo-Saxon” world are now formally subject neither to Great Britain nor to the United States. Nonetheless, the passage exemplifies very well the link that was simply assumed to exist between language and biological descent all through history until it was made explicit in the eighteenth and nineteenth centuries as the fundamental premise of the philological nationalism that was then developing in Europe. The philologists themselves, of course, started from the universal assumption that language and descent went together, so that when they divided languages into Germanic, Latin, Celtic, and so forth, they thought that they were also dividing mankind into “races” of common descent and common inherited characteristics.2 The advance of genetics, linguistics, history, and archaeology ought to have revealed the fallacy in this, but unexplored assumptions are slow to change, especially if it suits politicians to use them for good or ill, to cement the political unity of those they rule or to divert collective energies into hating those who are labeled “outsiders.”
The concern of this article is not with the problems of the modern world that are created or aggravated by the continuing confusion of culture with genetics but with the impediments such confusion puts in the way of understanding one particular part of the past. This historical confusion comes not primarily from the inexact way that historians themselves sometimes use words—which could be cured by making more precise definitions, agreeing them, and maintaining them among ourselves, which would be hard enough—but from our failure, as historians, to look at the way that words are and always have been used in real life, that is, loosely, with overlapping and inconsistent connotations, and above all changing their significance through time. Words shed or accumulate meanings as the world they represent changes and as people’s ideas change. Words, concepts, and things all change through time, but they change out of kilter. Historians cannot deal with the problems that changing terminology raises for them either by imposing definitions or by using the words we find in the sources without comparing them with our uses: both these procedures simply bypass the problems.3 We must study the words of the past in such a way as to discover the concepts of the past, and that means acknowledging the modern concepts that are likely to influence the way that even the most learned and devoted medievalist uses words that he or she thinks of as medieval.
Names are particularly interesting. The names of social and political collectivities suggest ways in which members of a group perceived it and in which that group was perceived by others. Human beings tend to be very realist (in the sense of the word used by medieval philosophers) about their social relations. We tend unconsciously to deduce entities from names and an unchanging entity from an unchanging name. Can there be a people without a name? Some of the debates about the unity or disunity of tenth- and eleventh-century Germany presuppose that the lack of a single and new name for a new entity must imply a lack of solidarity. It is, however, typical of the intuitive character of much historical reasoning that the hypothesis should simply be assumed, that the arguments started from nineteenth-century ideas about the nineteenth-century state of Germany, and that no proper comparative work seems to have been done—for instance on tenth- and eleventh-century France or England—to test either the assumed hypothesis or its wider application to that period of history.4
The object of this article is to look both at the modern use of the word “Anglo-Saxon” as applied to the inhabitants of the south and east of Britain between about 500 and 1100 and at what they seem to have called themselves, in order to try to disentangle what we think about them, about what sorts of groups they formed, and about the ways in which they were divided from other groups from what they themselves thought about these matters. First, something more must be said about the history of the word. Murray, in the first definition quoted above, suggested that it probably originated on the Continent around the ninth century, as a name designed to distinguish the Saxons of Britain from those east of the Rhine. Further study, while adding a few references that push the first recorded instance back to the late eighth century, has more or less confirmed this, though it is worth noting that the exact form of the compound name with an o in the middle occurs comparatively rarely in sources surviving from the early Middle Ages. Forms like Angli Saxones (for the people), Engelsaxo (for one man), or Anglorum Saxonia (for the country) are more common. From Italy come also references to the gens Anglorum et Saxonum, in which the singular gens preserves the compound implications that the et would otherwise deny. Meanwhile, the inhabitants of what would come to be called England were, by the early eighth century if not before, using the simple word “English” (Angli, Anglici) to refer to themselves. As far as they were concerned, the name “Saxon” seems to have been reserved for the people of the southern kingdoms, while even southerners sometimes applied the description “English” to themselves as individuals and acquiesced in being subsumed under it as a group. Though Bede occasionally used expressions like Anglorum sive Saxonum gens, the genuinely compound name seems to appear in surviving native sources only from the late ninth century on, when West Saxon kings and their successors sometimes referred to themselves as kings of the Angli Saxones, Angolsaxones, Anglosaxones, or Angulsaxones, perhaps using such forms to help them build bridges toward parts of the country that were, or had once been thought of as, quintessentially Angle.5 This usage had become rare by the later tenth century. Though it crops up occasionally in Aethelred’s and Cnut’s charters, its use in an opening clause is even then sometimes balanced by the king’s subscription as rex Anglorum.6 Except when they were making vague but grandiloquent claims to supremacy over the whole island of Britain, the normal title of all kings from the later tenth century on seems to have been rex Anglorum. They were reges Anglorum, and their subjects, correspondingly, were Angli—English. The compound name did not reappear until it was resurrected in the sixteenth century in order to distinguish the language and history of the inhabitants of England before the Norman Conquest from those of later periods. In summary, therefore, the expression “Anglo-Saxon” originated abroad, possibly less as a compound of two names than as a single noun (“Saxon”) with a geographical adjective (“English, of England”) attached so as to distinguish the Saxons of England from those elsewhere. That possibility—though it is no more than that—suggests that the native use of the blanket description “English” was already giving rise to the geographical concept of Angleland or England. On the other hand, when the natives—or their kings—picked up the compound, it seems to have been as a genuine compound of two names. It did not catch on, however, presumably because the unitary name Angli was by then too well established.
Against this background, our use of the word raises a lot of questions. When others called them Saxons and they called themselves English, did the inhabitants of England think of themselves as one people or even as the sort of hyphenated compound—hyphenated but nonetheless distinct from all others—that might be inferred from our name for them? When did they become distinct from others? In what did they think that their unity, such as it was, consisted? In language, descent, or politics? What sort of unity did they in fact enjoy, and, insofar as it changed, did their perception of it change too? Which came first, political unity or a sense of unity? Finally, can we justify the degree of the break in continuity that is liable to be inferred from references to “the Anglo-Saxon period,” especially when it is contrasted, as it often is by British archaeologists, with a “medieval period” that began only in 1066? Did the identity of the people and their culture change as much in 1066 as the change of name implies?
The consideration of these questions here will start from the premise that medieval people seem to have envisaged their world as divided into “peoples” (gentes, nationes, populi) of common biological descent and culture who normally and naturally formed separate political units. By the twelfth century, though maybe not before the tenth, these political units were quintessentially thought of as kingdoms.8 It is important to remember that these assumptions about peoples and kingdoms are not quite the same as the assumptions about nations and states that many people hold today and that, unless we recognize them and allow them, will color our view of the past. In the Middle Ages, at least after the tenth century and perhaps earlier, it was assumed that a people who, as some kind of political unit, shared common customs and law—and quite often a common language, though the texts mention that less often—were therefore of common descent. Nowadays, on the other hand, it is often thought that people of common descent share a common culture and therefore ought to form a separate political unit. The medieval assumption was an unreasoned justification for the status quo; the modern belief is an often controversial justification for change. The modern ideas have therefore been sharpened and altered by controversy and abuse in a way that the medieval assumptions were not. Though medieval people perceived differences between “peoples” and felt collective loyalties to their own, we cannot assume without argument that their perceptions of identity and difference correspond in all respects with modern perceptions of national and “ethnic” divisions, based as they are on modern ideas of popular government, nationalism, and race.
* * *
The most difficult problems of all in what we think of as Anglo- Saxon history come at its very beginning, with what is traditionally called the Anglo-Saxon settlement and with the consequent break in continuity—whether great or small—with Roman Britain. Increasingly close and careful study of the evidence does not so far seem to have made those problems less difficult; but, as archaeologists, philologists, and historians have all given increasingly close attention to the reasoning they apply to the evidence, it has at least become possible to identify and define the most important issues.9 One important part of the reasoning has, however, not yet received the attention that it deserves. That is the very idea of peoples or nations as discrete and objective entities, whether these are classified as Celts and Germans, as Britons and Anglo-Saxons, or as Angles, Saxons, and Jutes, possibly with the odd Frank or Alaman thrown in. Many of those who criticize the evidence with great rigor nevertheless seem to take these entities, or at least this sort of entity, on trust. They may accept that different “tribes” (as they are often called) became mixed together during the Age of Migration and Settlement, but they still imply that each tribe or people that had (or has) a separate name formed some kind of cultural entity, which was altered in proportion to its mixture with other tribes. The culture of the resulting society was therefore determined as much or more by the proportion of individuals of one or another descent within it as by any change of political or economic circumstances. It is, however, salutary to remember two things. One is that, difficult as it is to draw the boundary between what is transmitted biologically and what is transmitted culturally, we now know, as earlier ages did not, that there is a difference between the two processes. Most of the human social behavior in which we are interested at this juncture—such as religion, language, and the making of artifacts—was not genetically programmed. It was cultural. The other is that a great deal of our picture of the Age of Migration or Völkerwanderung is based on myths of origin and descent of a very conventionally mythical kind that have not gained in credibility by being adapted to fit the later ideas of philological nationalism.
One consequence of this line of thought is that one may go on to resolve some of the problems of the period by deducing that the so- called Germanic tribes of the Dark Ages probably formed much less tidy entities than is traditionally supposed and that there may have been comparatively few wholesale movements of groups that we can describe as tribes without grave risk of being misunderstood. Having suggested this, however, one has to concede that the migrations into England look like a special case. The change of language and religion was greater than in other Roman territories, except for a few border areas, so that it is reasonable to suppose that immigration may have been on a proportionately greater scale. That, however, is only a supposition: the language evidence, for instance, seems to be discussed without the benefit of much comparative evidence about the normal operation of this sort of linguistic change. Perhaps no comparative work has been done or is possible, but if that is the case, then the lack of it weakens the argument.10
In fact we do not know how consistently the Germanic-speaking invaders of Britain behaved like a group or felt themselves to be a group during the fifth and sixth centuries. We do not know what they called themselves or what others called them, if indeed they had any collective name. To the Roman authorities, those whom they called Saxons may have been particularly noticeable. Saxons, Scots, and Picts were the only barbarian invaders whom Gildas thought worth naming.11 That does not necessarily mean that all the Germanic- speaking invaders thought of themselves as Saxons or automatically allied themselves with Saxons against Britons, Picts, or Scots. By the eighth century, however, a sense of unity had somehow developed that enabled Bede to write in Latin of the gens Anglorum and, perhaps, to speak in the vernacular of the Angelcynn or Angelðeode—or both.12 Those people in other countries, meanwhile, who by then spoke of the inhabitants of eastern Britain as “Saxons” (perhaps as a continuance of the earlier Roman usage) seem, as the qualifications “English Saxons” or “Saxons of England” imply, to have meant by this much the same people as Bede meant by gens Anglorum. Given the idea of peoples that prevailed at the time, both usages may imply a certain sense of common descent. That such a sense was felt is confirmed by the way that, although the kings of the various kingdoms between which the gens was divided cherished their own separate genealogies, there was some sort of idea that the English as a whole were kinsmen of the Continental Saxons. That must imply that they were thought to be of common descent themselves.
When and how the sense of unity that is implied by the common name and sense of common descent had originated needs more thought, however, than it did when everyone took it for granted that the English nation was a natural unit of blood and culture, including language, that was foreordained to come into existence as soon as it arrived in its promised land. The culture of the Germanic invaders, judging from their material remains, varied somewhat, and their common language may not have been very uniform. They may perhaps have been sufficiently like each other, and sufficiently different from both the native British and the simultaneously invading Picts and Scots, to be drawn together against them, but we cannot be sure. They were certainly very different in culture from the upper ranks of Roman Britons, but less so, surely, than from the lower ranks or from the other invaders—or indeed from the ruling Britons after the pax romana had become a hazy memory. Given the fluid and confused conditions that are likely to have prevailed through most of Britain in the fifth and sixth centuries, military and political conflicts may not always have followed “ethnic” lines: by the seventh century, when a sense of English unity is easier to postulate and explain, they did not. It may be a mistake to try to make sense of the fifth and sixth centuries by picturing the countryside as divided into British settlements, English settlements, and empty areas, which can be mapped by dating either archaeological finds or place-names and attributing them neatly to particular ethnic groups. The reasoning depends on assumptions about ethnic continuity and ethnic separation that are difficult to justify. Apartheid is hard enough to maintain even when physical differences are obvious, political control is firm, and records of births, deaths, and marriages are kept. After a generation or two of post-Roman Britain not everyone, perhaps comparatively few people, can have been of pure native or invading descent. Who can have known who was descended from whom?
The implication of this argument—which is, of course, neither original nor provable—is that those whom we call Anglo-Saxons were not consistently distinguishable from everyone else. They were definable primarily by their military allegiance. Kings who called themselves (or whom Bede and others would later call) English or Saxon— although, incidentally, some of them bore British-sounding names—ruled areas that were therefore thought of, or later came to be thought of, as English or Saxon. How soon all the subjects of these kings came to speak English as their first language, and why it was that their English, when it came to be written, was so little affected by Celtic forms, seems very mysterious. It is not made less so by the way that some philologists and place-name scholars seem to be working with models of language transmission through “race” or “blood” that look a little old-fashioned.13 This is not just an interdisciplinary grouch: many historians use a similar terminology, and it is worth drawing particular attention to philological arguments only because the linguistic evidence is so important. Roman Christianity, like Roman political and material culture, could have been submerged merely by war and disorder. It is the linguistic change that most strongly suggests a large-scale immigration of a separate and possibly homogeneous people. If we try to explain it from unjustified premises about a link between biology and culture, genes and language, then we cannot hope to arrive at reliable conclusions.
It seems pointless to try to guess what proportion of the population of eighth-century England was descended from invaders of the fifth or sixth, let alone how, when, and why the descendants of the British among them changed their language. These questions, interesting and important as they are, need not, however, be directly relevant to the creation of a sense of English unity. That was a matter not of physical descent but of changes in political and social solidarity. Anyone who lived in an area dominated by an English king and who therefore owed allegiance to him was likely to come to consider himself and to be considered English. In time everyone in these areas came to speak English. There were probably exceptions everywhere at first and in newly conquered areas for longer. As late as the seventh century the Briton who rode past the site of the battle of Maserfelth (if that was in Mercian rather than British-ruled territory) may have been one, while those of King Penda’s subjects who were Christians before a formal mission arrived may have been people who had inherited or picked up British Christianity.14 A little later Ine’s laws referred to Welshmen and their wergelds in terms that suggest that some of his subjects kept a sense of British separateness. Nevertheless, he legislated for them. Royal authority in Britain seems to have been territorial, which would tend to work against the survival of this kind of separate group identity within a kingdom.
Presumably the common language of the gens Anglorum, on which Bede commented, was a help in creating its sense of unity, but history does not suggest that a common language is either a necessary or a sufficient cause of a sense of national unity. Fighting together, perhaps under King Aelle, King Ceawlin, or any of the rest of the “great kings,” may have helped, but Patrick Wormald has made that look more doubtful than it used to seem. The real founders of the sense of English unity may well have been, as he argues, Gregory the Great and Theodore of Tarsus.15 Once they had set the English church on its way, owing as it did devoted allegiance to their memories, divided as it was from the British church, and flourishing as it flourished in the late seventh and eighth centuries, the scene was set for that combination of unity and conflict that we find in the sources of what is sometimes oddly called the “middle Saxon” period. The English were united in their loyalty to Rome and Canterbury, in the division from the British that was symbolized by the separate names, languages, and ecclesiastical traditions, and in their sense of common descent. But they were divided by their separate kingdoms, the often fierce hostility between them, and the correspondingly fierce loyalties their kings could evoke. Unity and diversity coexisted on different layers of social consciousness and political allegiance.
The layer of unity was as yet a matter of feeling, not one of either political authority or genuinely common descent. Most subjects of English (or Saxon) kings who thought about it may have thought that their ancestors came from much the same areas as Camden, J. R. Green, Chadwick, or Stenton later believed they did, but in the European Middle Ages, as in other places and periods, the idea of common descent was more often the result than the cause of social and political solidarity. It was not based on good history, let alone on recognizable inherited characteristics or knowledge of the relationships between languages. Those who are commonly called the early and middle Anglo-Saxons (or Saxons) can best be identified as those inhabitants of Britain who, wherever their ancestors came from or whatever the mixture of genes in each of them, were the subjects of kings who claimed descent from gods or kings of the mainland of northern Europe and who apparently thought of themselves or their ancestors as Angles, Saxons, and maybe Jutes. The majority of these people were probably not pure-blooded descendants of invaders, nor were any pockets of British speakers among them probably pure-blooded descendants of the earlier native population. The subjects of English kings came to form a single linguistic, religious, and cultural community that seems to have called itself the English people. Some called themselves Saxons, and outsiders sometimes called all of them English Saxons, but they are not recorded as having called themselves Anglo-Saxons. So far as they were disunited it was not because some were Angles and some Saxons but because they formed separate kingdoms that cherished internal loyalties and mutual hostilities that did not derive from ethnic divisions.
* * *
The next stage of the story conventionally finds its plot in the evolution of the “bretwaldaship” toward the unification of England, but Wormald’s criticisms seem to have robbed that plot of much of its coherence.16 One of the reasons why “overlords” or “bretwaldas” have loomed so large since the time of H. M. Chadwick may be that, according to the unexamined assumption about nations that was current in Chadwick’s day and still is today, each nation or people (and that is normally assumed in Europe to mean a people of common language and “common stock”) moves through the attainment of national consciousness toward political unity, ultimately, if it is lucky, finding its rightful boundaries as a nation-state.17 The stage of the “bretwaldas” represents the transition from cultural to political unity and represents it in a nicely constitutional way: the English, being English—being a nation—wanted to be one “national kingdom” (as it were, a proto-nation-state), and any top king’s ambitions were therefore met by a measure of voluntary submission. Rival kings might vie for the top job, but there was a job—a vacancy—to be filled. We should not assume, however, that there was any such vacancy: the Franks, after all, felt some sense of solidarity but did not therefore remain a single kingdom. When the English became united in one kingdom in the tenth century, their unification was made easier by the preexisting sense of solidarity and by all the ideas of kingdoms and peoples that were then current, but we cannot assume that it was the culmination of a steady trend toward political unity. It was the creation of new political circumstances.
The most important of those circumstances were the Scandinavian invasion and conquest of all but one of the English kingdoms in the late ninth century and the ultimately successful wars waged against the invaders and their successors by the surviving monarchy of Wessex that followed. Here we face another set of questions about the identity of the “Anglo-Saxons.” What was the relationship between them and the Danes and Norwegians who settled permanently in the country? How did contemporaries perceive these categories and the difference between them, and how long did they continue to see them as separate? According to modern assumptions about “ethnic groups,” it may seem reasonable that “the Scandinavians formed an important recognizable element” in the population, “a distinct community living under separate laws.”18 It is, however, a little difficult to envisage how this worked in practice: most historians seem to assume that people of Danish descent all over the country were recognizable as such—and presumably enjoyed some kind of “personal law”—yet it also seems to be often assumed that everyone living in the “Danelaw” was thought of as Danish. In Stenton’s magisterial words, “The whole of eastern England, from the Thames to the Tees [was] regarded as the sphere of a distinct form of customary law.”19 It may also seem reasonable to suppose that the descendants of the Danish and Norwegian settlers of the ninth and early tenth centuries would have felt a sense of kinship with the new invaders of the early eleventh, so that in the end they sided with them.20 A division based on “ethnic” loyalties has even been thought to underlie the obscure conflicts of Edward the Confessor’s reign and to explain the allegedly incomplete unity of England in 1066.21 There seems, however, to be very little hard evidence for all this. It is an interpretation imposed on the evidence under the influence of the same ideas of philological nationalism—the same belief in the fundamental importance and connection of common descent and common language—as lie behind the expression “Anglo-Saxon” itself.
The arguments that the descendants of Danes did not continue to be thought of as Danes and that the Danelaw was not a consistently separate unit of government do not depend on decisions about the scale of Scandinavian immigration or the proportion of the population in northern and eastern England in, say, the eleventh century that was descended from them: the evidence of Danish influence is not dependent on that, and it is easier to avoid circular arguments if one ignores these problems. The difference between Danes and Norwegians is also irrelevant here, though it might be noted that arguments that they were consistently distinguished run into difficulties and inconsistencies.22 Nor is it necessary for present purposes to assess—let alone to depreciate—the impact of Scandinavian influences on the language or institutions of the country. The argument is simply that, whatever their number and whatever their influence either on the whole kingdom or on part of it, once the immigrants or their descendants had become part of the kingdom they do not seem to have been perceived as a separate and identifiable group within it. It is not easy to see how they could have been. Personal names are evidence of fashion rather than of descent, and given the probable size of the preexisting population of eastern and northern England, the similarity of the English and Norse languages, and the apparently quick disappearance of active paganism among the Danes, social mixing and intermarriage must have started soon. It is even harder to see how most people in the north and east could have been of exclusively English or Danish descent by 1000, or could have known if they were, than to know about the relationship of English and British in earlier centuries.
One explanation might be that anyone who mattered in the areas that were thought of as following Danish law therefore assumed himself to be a Dane, but it is not obvious that this was the case. The geographical definition of any such area is far from clear. The threefold division of the kingdom and its law into West Saxon, Mercian, and Danish became traditional, and law codes went on occasionally referring to different penalties under English and Danish law, but some of these references are not tied to geographical areas, and even where they seem to be, the areas are seldom defined. All the records of various sorts that survive from the tenth century on refer to the Danelaw as a geographical area far less often than most modern accounts imply.23 The more frequent use of other geographical descriptions like East Anglia, Norðleoda, or Northumbria or indeed of county names (none of which, however new, explicitly recognized the difference between English and Danes) suggests that the Danishness of such areas was not their most obvious characteristic. Altogether the mutual influence of English and Scandinavian terminology and custom—so far as customs differed significantly—cuts across boundaries in a way that makes it doubtful whether the following of English or Danish law in practice divided the population in such a way as to preserve the sense of two separate communities, whether living together according to “personal law” or in two separate regions.24 By the eleventh century, moreover, if not before, few people probably realized that earls, bovates, and wapentakes came from Scandinavia.
The crucial point seems to be that, whatever the local variations in the law practiced in the kingdom, and however much contemporaries may have recognized some of the varieties as Danish, subjects of the kings of the English were normally assumed to be English themselves. It is true that kings sometimes referred to Danes among their subjects, but we should not assume that each time they did so they were thinking of all the people who—or whose ancestors—had ever been considered Danes. Probably those thought of as Danes at any date were recent Danes, people who had themselves come from Denmark and might be going back there, not the descendants of earlier and now absorbed waves of immigrants.25 These occasional references to Danes as subjects of the kingdom must, moreover, be set against the implications of the title rex Anglorum and the way that names were used in the Old English Chronicle. To the chroniclers, Danes were invaders and enemies, not subjects of the kingdom, and it is noteworthy that they hardly mentioned Danes at all between 920 and 990.26 Even in the early years of the tenth century, not all the armies against which West Saxon kings fought were described as Danish, and not all may have been perceived as such. In the war-torn power vacuum north of the Thames troops of local vigilantes, free companies, and supporters of all kinds of local leaders or interests may have posed as much of a problem to Edward the Elder and his successors as did passport-carrying Danes.27 Many a band described as a here must have been ethnically mixed, including some bands that a chronicler loyal to a West Saxon king might well think of as Danish (or Norse) because the leaders were invaders from abroad.28 Individual Vikings, after all, from what we know of a few famous careers, operated all over the place, and a good many belonged, if anywhere, only to areas of already mixed settlement like Iceland or Ireland. The ultimate ethnic origin of their followers was even more irrelevant and unknown than their own and became still more so after war was ended and feelings of military solidarity were overlaid by those of political solidarity.
During the reign of Aethelred, a new series of Danish invasions met with dogged resistance for many years. When things began to go wrong, the chronicler, despite his eager search for scapegoats, did not blame treachery on Danes or suggest that the unhappy divisions of the kingdom followed ethnic lines: to him Danes were foreigners. The traitors, so far as he was concerned, were English. There are two possible exceptions, which bear a closer look. In 993 a raiding fleet came into the Humber, where a very large army was collected for defense but failed to give battle because its leaders fled. The twelfth- century Worcester chronicle may have been drawing on its now lost version of the Old English Chronicle when it added that this happened because the three generals of the defending army force were Danes on the father’s side.29 This does not suggest that the northeast as a whole was regarded as Danish or partly Danish: rather the reverse. The half- Danishness of the generals seems exceptional. The other case came twenty years later, when King Svein’s invading army also landed in the Humber, and Northumbria, Lindsey, the Five Boroughs, and (in Whitelock’s translation) “all the Danish settlers north of Watling Street” quickly submitted to him.30 Here, however, Whitelock is not so much translating as interpreting. The original does not mention Danish settlers. It says eall here. One reason for Whitelock’s choice of words was presumably the belief that here was used only of Danish forces. That is not quite true, though it was more often used of an invading, hostile, or rebellious force than of an honorable defending one.31 Apart from that, however, the translation could only make sense on the premise that the descendants of Danish settlers were Danish settlers forever (“settler” being, like “immigrant” in Britain today, a hereditary status) and that all the arms bearers of the region thought of themselves as Danes—which makes nonsense of the events of 993, when three generals were singled out as half Danish. That is not to say that no one in the northeast ever felt Danish, but the evidence that they did so is less good than apparently authoritative modern translations and interpretations suggest. Too many of such interpretations depend on unstated, intuitive premises, such as the idea that loyalty to ethnic roots outweighs loyalty to the king and community of the kingdom in which one lives. There is also too much intuitive reasoning, like that which deduces the original intentions of a general from the course his campaign takes. While Kappelle has argued that the earlier raids of Aethelred’s reign spared the northeast because it was Danish, Stenton thought that Svein started there in 1013 because he expected that the “men of Danish England would be prepared to welcome a Danish king, and that a base established in their country would be secure.”32 Either deduction might be correct, but neither need be, and they seem to proceed from different premises. In the event, the impeccably English south (except for London) submitted quite quickly to Svein too, and Cnut started his invasion of 1015 there, while Edmund Ironside afterward got an apparently quick submission at one stage from the allegedly Danish Five Boroughs. Perhaps ethnic origins were not the only factors at work.
Postulating that permanent inhabitants of the kingdom were normally considered to be English may also make sense of a third episode in Aethelred’s reign: the Saint Brice’s day massacre of 1002. Stenton found the king’s order to kill all the Danish men in England futile as well as repugnant: “Within more than a third of England no order of this kind could ever have been carried out. York and Lincoln, for example, were Danish rather than English towns.”33 But the author of the Life of Saint Oswald did not think that York was Danish. To him Danes were either the wicked invaders of the kingdom or the most numerous group among the foreign merchants who visited the city.34 The intended victims of the massacre were surely not those who might have had a Danish ancestor but visiting aliens or, at most, recent immigrants. Those among the Danes, thus defined, at whom the royal directive may have been primarily aimed were people like Pallig, who, just a year before, had deserted King Aethelred in spite of all the gifts of gold and silver he had received from the king and had joined up with an invading force to terrorize the southwest.35 Fury with Pallig, and fear of other fifth columnists like him, might explain the massacre.
Cnut’s reign, like William I’s later, brought in foreigners who probably continued to be thought of as foreign in so far as they retained homes and interests elsewhere. It is not surprising, therefore, that Cnut referred to some of his subjects as Danes. Whether any of the children of Danes who settled permanently at that time went on being thought of as Danes is doubtful, as it is for earlier generations. In Edward the Confessor’s reign the reasons why any earl favored a Danish, Norman, or any other alliance must have shifted over the years as political circumstances at home and abroad changed. Personal links with Denmark may have influenced some, but we should be wary of assuming, for instance, that Scandinavian names in Godwin’s family are evidence that the family “represented the Anglo-Danish interest.” There does not seem to be any reason either, except perhaps the apparently Norse origin of the words butsecarl and lithsman, to suppose that Edward the Confessor’s fleet was “foreign, probably Viking.”36 Housecarl is also a Norse word by origin, but that does not mean that members of royal bodyguards from Cnut’s time on were themselves all Scandinavians any more than, in later history, all dragoons were French. Some of Cnut’s followers no doubt were, but then he was said, as one might expect, to have drawn his soldiers from all his realms.37 Some at least, therefore, are likely to have been, or become, or been thought of as English. The origins of words are a poor guide to the birthplace or loyalties of those they describe.
* * *
Ten sixty-six marks the end of what is called “the Anglo-Saxon period.” It would be impossible, at the end of an article that has already skimmed over some 600 years, to embark on the controversies that have long raged around the changes that resulted from the Norman Conquest. One suggestion may, however, be made very briefly. Some of the problems about the Conquest seem to be made worse by the tendency to opt for package deals for continuity or change, without close enough examination of the ideas that lie behind old interpretations and that continue, so long as they are unexamined, to influence new ones. It is surely impossible to believe any longer, as pre-Marxists believed, that a military conquest by a relatively small aristocracy can change class structures and relationships significantly while the basis and nature of the economy remains the same. Economy, society, and government were changing all over western Europe in the eleventh and twelfth centuries. With some variations in timing and character, many of the changes traditionally associated with the Norman Conquest took place in other countries too. Post hoc may not mean propter hoc. In some spheres, of course, it may. In the arts, for instance, the importance of patronage combined with the regional diversity of styles suggests that conquest by new patrons could help to explain some of the changes that undoubtedly occurred. Yet it was possible to come away from the recent exhibition of English Romanesque in London with nagging doubts about the supreme significance of 1066 and about the reigns of kings thereafter as useful divisions of time in tracing the evolution of artistic styles. Even in architecture, where regional styles are particularly clear and Norman looks so different from “Anglo- Saxon,” may not the contrast be exaggerated by the destruction of all the grander English churches built in the century before 1066? In the purely political sphere, on the other hand, where some changes that followed in the long run can be connected more or less plausibly to the Norman—and Angevin—links of future kings, the long run involved many contingent events that were not foreordained by the conquest of 1066.
However all this may be, so far as the sense of Englishness and English solidarity are concerned, the chief change that followed from the Conquest was that relative peace and unity were succeeded by war, destruction, and division. It was the best part of a century before the inhabitants of the kingdom of England could once more be assumed to be a single people—the English. When the solidarity of the kingdom was rebuilt, it was based on the same assumptions and ideas as had underlain the solidarity of the years before 1066: the English people were a people because they formed a kingdom. If theirs was more united than some other kingdoms, that was because, before 1066 as well as after, the government traditionally commanded and received a good deal of obedience. William I did not conquer a hyphenated or compound kingdom of Anglo-Saxons. He conquered the kingdom of the English, and although during most if not all of his reign those who thought of themselves as English remained dangerously hostile to his army of occupation, it was the kingdom of the English that he and his successors continued to rule.
* * *
The conclusion may be brief. The “Anglo-Saxon period” is too long and contained too much change to form a significant unit for study—a coherent “period.” If it is held to begin with the end of the pax romana in Britain, then it certainly begins with a real break in civil life, but the sense in which one could say that the “Anglo-Saxon” period begins then is not immediately apparent. The people we call the Anglo-Saxons can barely be identified at that stage. Not until almost two centuries later do we know that the English, or English Saxons, were beginning to perceive themselves and be perceived as a unit. Until the tenth century their sense of unity was frequently undermined by political conflict, and we cannot assume that they had a manifest destiny of political unity. When a single kingdom was formed in the tenth century within boundaries that more or less approximate those of what we think of as England, it was called the kingdom of the English—not of the Anglo-Saxons, nor yet of the Anglo-Danes. Right through the “Anglo-Saxon period,” therefore, the term “Anglo- Saxon” invites us to beg questions and confuse our own ideas with those of the period we study. It would be overpresumptuous to attempt to stop the terminological world of historians—let alone of the general public—and try to get off; but even if we must continue to use a name that has become well established in tradition, we might do well to remember that the early medieval English did not call themselves Anglo-Saxons. If we want to call them that, we ought to think hard about what we mean, and what others may think we mean, by the name that we have chosen to use.
AFTERTHOUGHTS
p. 397: On the German kingdom’s lack of a name see afterthought to II.
p. 384. pp. 398, 401-2: M. Richter, ‘Bede’s Angli: Angles or English’, Peritia, 3 (1984), 99-114, produces more arguments for the early use of the name Saxons, including by the inhabitants of south-eastern Britain themselves, than I suggested.
p. 401 (comparative evidence on language change) and 403-4 (scale of immigration): H.R. Loyn, Anglo-Saxon England and the Norman Conquest (London 1962, 1991 edn. substantially unchanged), 5-14, sets out the problem posed by the language change. Since 1962 much archaeological evidence of continuity has appeared and has generally been interpreted to mean a large survival of the British population: see e.g. B. Yorke, Kings and kingdoms of early Anglo-Saxon England (London, 1990), 7 and, less strongly, P. Kirby, The earliest English kings (London, 1991), 12-14. Neither discusses the change of language. Conclusions drawn from archaeological evidence alone (or even from archaeological and place-name evidence) are less convincing than they would be if they took the language change into account. D. Dumville, Britons and Anglo-Saxons (Aldershot, 1993), III. 3, 6, 10-11, alludes to it (cf. IX. 7 on the problem of what he calls ‘the biological, or racial, relationship between the Anglo-Saxon rulers and their peoples’). On this last point, cf. T. Charles-Edwards, ‘Early medieval kingships in the British Isles’, in The origins of Anglo-Saxon kingdoms, ed. S. Bassett (Leicester, 1989), 28-39, at 34.
p. 405, 406: on ‘great kings’ or ‘bretwaldas’: S. Fanning, ‘Bede, imperium and the Bretwaldas’, Speculum, 66 (1991), 1-26.
p. 406-8: See now P. Stafford, ‘Danes and the Danelaw’, History Today (October 1986), 17-23. Gneuss, ‘Anglicae linguae interpretatio’, Proc. Brit. Academy, 82 (1992), 139, discusses ‘mutual intelligibility’ between English and Norse invaders.
p. 408, n. 23: I should also have cited R.H.C. Davis, ‘East Anglia and the Danelaw’, Trans. R. Hist. Soc. ser. 5,5 (1955), 23-39.
p. 410, n. 28 (on fyrd and here): N. Hooper, ‘The housecarls in England in the eleventh century’, Anglo-Norman Studies, 7 (1984), 165.
p. 413, n. 37: See Hooper, as in afterthought on n. 28, above.
1 Though it is noted, e.g., by D. Bullough in Artemis Lexikon des Mittelalters (Munich, 1980-), vol. 1, cols. 619–20.
2 See, e.g., H. A. MacDougall, Racial Myth in English History (Hanover, N.H., 1982).
3 See M. Bloch, The Historian’s Craft (Manchester, 1967), p. 159.
4 I discuss this briefly in Kingdoms and Communities (Oxford, 1984), pp. 289–92.
5 Though note Eadwig’s reference to (apparently) the Northumbrians as aquilones Saxones (W. de G. Birch, ed., Cartularium Saxonicum [London, 1885–99], no. 926).
6 For example, J. M. Kemble, ed., Codex Diplomaticus Aevi Saxonici (London, 1839–48), nos. 705, 714, 736, 770, 787, 793, 1308.
7 P. Wormald, “Bede, the Bretwaldas, and the Origins of the Gens Anglorum,” in Ideal and Reality in Frankish and Anglo-Saxon Society, ed. P. Wormald with D. Bullough and R. Collins (Oxford, 1983), pp. 99–129.
8 This is argued by S. Reynolds “Medieval Origines Gentium and the Community of the Realm,” History 68 (1983): 375–90, and (for the period after 900) in Kingdoms and Communities, pp. 256–331.
9 For example, C. Hills, “The Archaelogy of Anglo-Saxon England in the Pagan Period: A Review,” Anglo-Saxon England 8 (1979): 297–329; D. N. Dumville, “Sub-Roman Britain: History and Legend,” History 62 (1977): 173–92; P. Sims-Williams, “The Settlement of England in Bede and the Chronicle,” Anglo-Saxon England 12 (1983): 1–41.
10 The only work I have found that comes near to discussing it is U. Weinreich, Languages in Contact (The Hague, 1966). More may well have been published since, but the fact that such works do not seem to be cited in discussions of the change from British to Old English may explain my concern.
11 Gildas, The Ruin of Britain, ed. M. Winterbottom (Chichester, 1978), s.v. “Saxon,” “Scot,” and “Pict.”
12 Oxford English Dictionary 3:179; T. Miller, ed., Old English Version of Bede’s Ecclesiastical History, Early English Text Society, O.S., 95 (London, 1890), e.g., pp. 40, 70, 108, 110.
13 T. H. Parry-Williams, “English-Welsh Loan-Words,” in J. R. R. Tolkien et al., Angles and Saxons: O’Donnell Lectures (Cardiff, 1963), pp. 71, 73; W. Rees, “Survival of Ancient Celtic Custom in Medieval England,” in ibid., p. 148. J. R. R. Tolkien, “English and Welsh,” in ibid., pp. 9–12, however, expresses a different view.,
14 Bede, Historia Ecclesiastica 3.10, 21, ed. C. Plummer (Oxford, 1896), 1:146.
15 Wormald. It would be tedious and ungrateful to argue about small points, but it is not quite correct to say that Bede described the common vernacular as Saxon. He calls the language of Wessex or Essex Saxon, but when he is talking of the common language—one of the four languages between which Britain was divided—he normally refers to it as that of the English (Anglorum) (see Plummer, ed., s.v. Anglorum lingua, Saxonum lingua).
16 Wormald.
17 H. M. Chadwick, Origin of the English Nation (Cambridge, 1907), pp. 88–89; see also, e.g., F. M. Stenton, Anglo-Saxon England, 3d ed. (Oxford, 1971), pp. 33–35, 202, 236–37; P. H. Blair, Introduction to Anglo-Saxon England (Cambridge, 1962), pp. 49–54, 201–4; E. John, Orbis Britanniae (Leicester, 1966), pp. 1–63; and P. H. Sawyer, From Roman Britain to Norman England (London, 1978), pp. 48, 99–114. I have discussed the influence of modern nationalist ideas on the writing of medieval history in Kingdoms and Communities (n. 4 above), pp. 7–8, 250–56.
18 H. R. Loyn, The Vikings in Britain (London, 1977), p. 113. D. Ó Corráin, Ireland before the Normans (Dublin, 1972), p. 105, accepts this as the standard view in order to make comparisons with Ireland.
19 Stenton, p. 506.
20 For example, D. Whitelock, introduction to English Historical Documents, ed. D. C. Douglas, 2d ed. (London, 1979), p. 48.
21 On Edward’s reign, see, e.g., F. Barlow, Edward the Confessor (London, 1970), pp. 89, 92–93, 102, 191–92; and W. E. Kapelle, The Norman Conquest of the North (London, 1979), pp. 28–29, 47; on incomplete unity, see Kapelle, pp. 12–15; and R. A. Brown, 4The Norman Conquest,” Transactions of the Royal Historical Society, 5th ser., 17 (1967): 109–10, 116–20.
22 A Mawer, “Redemption of the Five Boroughs,” English Historical Review 38 (1923): 551–57, argued the case with particular reference to 942 but applied his argument more or less explicitly to a longer period. Stenton was commenting on 942 when he referred to “the antagonism between Danes and Norsemen, which is often ignored by modern writers, but underlies the whole history of England in this period” (p. 359). He also mentioned “an aristocracy of Norse extraction” around York as a survival of the Norse kingdom there but elsewhere referred to York in 993 as a Danish town (pp. 358, 380). It seems to be generally agreed that the Danelaw included Yorkshire.
23 See the references listed by F. Liebermann, ed., Die Gesetze der Angelsachsen (Halle, 1916), 2:51, 347–48. The translations in A. J. Robertson, ed., Laws of the Kings of England from Edmund to Henry I (Cambridge, 1922), sometimes impose a geographical meaning where it is not always clear in the text: e.g., pp. 118 (8 Aethelred 5.1) and 156 (1 Cnut 3.2). The documents printed by Stenton in Documents Illustrative of the Social and Economic History of the Danelaw (London, 1920) do not, to judge from the index, refer to the Danelaw at all. For the boundary of the geographical Danelaw, see R. H. C. Davis, “Alfred and Guthrum’s Frontier,” English Historical Review 97 (1982): 803–10.
24 An obvious example of a boundary between supposedly Danish and English institutions that fails to follow the boundary of the geographical Danelaw as traditionally defined by historians is that between hundreds and wapentakes. Other anomalies are implied by the concept of “English Northumbria,” which may even have been devised to accommodate them. The Five Boroughs also raise difficulties. How far and for how long they were perceived as Danish is doubtful. The “confederation of the Five Boroughs” may have been a consistently defined entity from around 942 to 1015 or later, but the only contemporary references are three in the Anglo-Saxon Chronicle (942, 1013, 1015 [J. Earle and C. Plummer, eds., The Saxon Chronicles Parallel (Oxford, 1892), 1:110, 143, 146]) and one in a law (3 Aethelred 1.1 [Robertson, ed., p. 64]). These give less information than one might expect from modern histories, while the reference to Seven Boroughs in 1015 suggests some elasticity of constitution. The language of 3 Aethelred is apparently heavily Scandinavianized, in contrast, e.g., to that of 1 Aethelred (see P. Wormald, “Aethelred the Lawmaker,” in Ethelred the Unready, ed. D. Hill, British Archaeological Reports, British series, 59 [Oxford, 1978], p. 61); but arguments that, though issued at Wantage and self-evidently assuming royal authority over Danes, it nonetheless represents the essentially Danish custom of the “territory of the Five Boroughs” (see Stenton, pp. 508–12) involve a good deal of circular reasoning about the Danishness of the area, the difference between Danish and English custom, the boundaries of royal authority, and the continuing political importance of ethnic divisions in general.
25 For example, Robertson, ed., pp. 32, 36 (4 Edgar 2.1.2; 12); D. Whitelock, ed., Anglo-Saxon Wills (Cambridge, 1930), pp. 44–46; cf. Anglo-Saxon Chronicle, 918A, 920A (in Earle and Plummer, eds. [1:104], as 922, 924).
26 See Earle and Plummer, eds., s.v. Dene, Denisc (the scip here of 980 was presumably Scandinavian, though not identified by nationality; and see also the wicinga of 982).
27 P. Stafford, “Reign of Aethelred II,” in Hill, ed., 17–21, suggests causes of provincial resentment against Wessex in the later tenth century quite irrespective of Danishness.
28 A here was not invariably Danish: see scip here and land here in the glossary of Earle and Plummer, eds.; and in annals for 684E, 910E (ibid., 1:95: this entry may be a slip), 1052CD (ibid., 1:175, 178, 179), and 1054CD. In some cases where a here is not described as Danish (though it is often translated by Whitelock as “a Danish army”), the context shows that it was or was perceived as such (e.g., annals 896A, 903, 992). In 917 a micel here included both þæt land here and þara wicinga they enticed to join them, though in this case the land here may perhaps have been more Danish than the folc of the country they dominated. Here may sometimes have had the particular sense of an army in the field, especially a rebellious or invading army: cf. Ine 13.1, 15.1 (F. L. Atten-borough, ed., Laws of the Earliest English Kings [Cambridge, 1922], p. 40), and annal for 917 (East Anglian and Mercian forces), though this would not apply, e.g., to 4 Edgar 15 (Robertson, ed., p. 38), or to all the examples given above. Note also ut here and unfriðhere in annals 1009C and a possible link with the verb hergian. There does not, however, seem to be any reason why fyrd (or folc, another word that could be used for an army) and here should have consistently been used with mutually exclusive meanings (a fyrde could have heretogan: annal 993CDE) or that, if they were, ethnic connotations were uppermost in the user’s mind.
29 Florence of Worcester, Chronicon, ed. B. Thorpe, English Historical Society (London, 1848–49), 1:150–51.
30 Anglo-Saxon Chronicle, ed. and trans. D. Whitelock with D. C. Douglas and S. I. Tucker (London, 1961), p. 92.
31 See n. 28 above.
32 Kapelle, The Norman Conquest of the North (n. 21 above), pp. 14–15; Stenton, pp. 384–85.
33 Stenton, p. 380.
34 J. Raine, ed., Historians of the Church of York, Rolls Series, 71 (London, 187–994), 1:454, 455.
35 Earle and Plummer, eds., 1:132 (1001A); cf. William of Malmesbury, Gesta Regum, ed. W. Stubbs, Rolls Series, 90 (London, 1887–89), 1:207.
36 Barlow (n. 21 above), pp. 89, 92–93, 170. I do not understand what is meant by “Englishmen in disguise” (p. 191).
37 E. A. Freeman, History of the Norman Conquest of England, 3d ed. (Oxford, 1877), app. KK, 1:756.