From the latter part of the reign of Alfred (871–99) to the beginning of that of Æthelstan (924–39), the West Saxon king's style in charters was usually some variant of Angul Saxonum rex, reflecting his domination over both Saxon Wessex and Anglian Mercia.Footnote ² After the seizure of York and the meeting at Eamont in 927, Æthelstan was frequently styled Rex Anglorum and also presented as the ruler of Britain: a clear majority of his extant charters articulate such a claim. Following Æthelstan's death and the succession of Edmund (939–46), Hiberno-Scandinavian rule was reestablished at York until 944: this probably explains why Edmund was (so far as we know) not described as a ruler of Britain except in the last two years of his reign.Footnote ³ During the reigns of Eadred (946–55) and Eadwig (955–9), an assertion of rulership of Britain appears in a fair proportion of royal diplomas, and it is very common in the charters of Edgar (959–75). The idea was not confined to charters: many coins in Æthelstan's name call him Rex totius Britanniae, a poem probably written just after the Eamont meeting alludes to him readying for battle per totum Bryttanium, Æthelweard styled Edgar monarchus Brittannum and Byrhtferth of Ramsey referred to the same king as totius Albionis imperator, Albion being a synonym for Britannia. Similar examples could be multiplied.Footnote ⁴

Claims to rulership of Britain became less pervasive after the late tenth century. They by no means disappeared, but the proportion of eleventh-century diplomas in which they appear was lower than for Æthelstan or Edgar. Moreover, the number of diplomas being issued declined sharply during the eleventh century; writs, which simultaneously proliferated, did not articulate such claims. Nor was the idea prominent in eleventh-century narrative sources. The Life of Edward the Confessor occasionally hints that Edward's sphere of authority extended across Britain, but more often characterises him as ruler of the English alone.Footnote ⁵ The Encomium Emmae Reginae states that Cnut was imperator of England, Wales and Scotland (as well as Denmark and Norway), rather than depicting Britain as a single political unit.Footnote ⁶ After the Norman Conquest, as Rees Davies pointed out, claims to rule Britain became rarer still, despite increasingly frequent English interventions in Wales and Scotland.Footnote ⁷

It is often noted that tenth-century English kings were presented as rulers of Britain, but the question of why this was so has received little extended comment, probably because the answer appears obvious. The island is a definite physical unit and had long been recognised as such: its geography had been described by Pliny, Orosius, Isidore and Bede, to name but four particularly influential writers.Footnote ⁸ There was, moreover, venerable precedent for the idea that one man might be presented as the preeminent figure on the island. Bede had described the power of Edwin and Oswald, seventh-century Northumbrian kings, in such terms.Footnote ⁹ Æthelbald and Offa, eighth-century Mercian kings, had appeared in charters as rex Britanniae and rex et decus Britanniae.Footnote ¹⁰ Asser had portrayed Alfred as the rector of all the Christians of the island of Britain.Footnote ¹¹ Given that the Chronicle states (in the extract quoted at the outset) that Æthelstan had subordinated all other kings on the island, there is, on the face of it, little need to explain why he too was presented as the ruler of Britain. The matter is not, however, quite so simple, since few historians would now claim that any tenth-century (or earlier) English king had much more than a loose hegemony over the other kings in Britain: one could say that English power in Wales and northern Britain was ‘extensive’, rather than ‘intensive’.Footnote ¹² Why then were several tenth-century English kings nonetheless presented as rulers of Britain? And why, moreover, did such pretensions become less ubiquitous after Edgar's reign?

Before addressing these questions, we should note two long-running historiographical controversies. The first revolves around whether documents in the names of tenth-century kings were commonly written by permanent members of the royal household. The king might alternatively have commissioned people to write documents ad hoc, or indeed have expected the beneficiaries of royal grants to arrange the drawing up of diplomas themselves.Footnote ¹³ I cannot resolve this debate here (or, for that matter, elsewhere) but, for the purposes of my argument, it can be sidestepped. Whoever produced a diploma, it is unlikely that the draftsman would use a style of which the king disapproved, as royal backing was fundamental to the future security of the grant. But given that a wide variety of titles was used in tenth-century charters, it does not appear that kings routinely stipulated (at least in detail) how they were to be described: charter styles thus probably reveal most about how the king's position was perceived by literate churchmen who, whether working in a royal writing office or not, were almost certainly those who drafted diplomas.

The second debate concerns the significance of ‘imperial’ terminology. There are a couple of dozen cases of kings from Eadred to Cnut being styled imperator. Edmund Stengel and Carl Erdmann interpreted these as manifestations of a ‘nichtrömische Kaiseridee’, and similar connotations have sometimes been imputed to other titles, most notably basileus, which crops up frequently from the reign of Æthelstan to that of Edward the Confessor. The title imperator could be conferred by the pope (e.g. to Charlemagne in 800), but Stengel and Erdmann demonstrated that papal coronation was not a prerequisite for a man to be styled imperator or described as exercising imperium in the early medieval west. Even when a ruler had no particular connection to Rome, such terms were sometimes used if his power extended over several subordinate kingdoms or peoples.Footnote ¹⁴ We should not, however, assume that any particular political idea lay behind each use of a word like imperator or basileus. Tenth-century Anglo-Latin was marked by attempts to exhibit erudition through obscure vocabulary, and calling the king basileus may have been more about displaying knowledge of a grecism than likening him to a Byzantine emperor: that basileus was glossed rex suggests that it did not necessarily evoke ‘imperial’ associations.Footnote ¹⁵ Even imperator might have been used for the sake of lexical variety and a general air of grandeur, rather than because the writer associated the term with a specific ‘imperial’ ideology: one could translate it ‘commander’, as well as ‘emperor’.Footnote ¹⁶

While the contemporary import of words like imperator and basileus cannot be ascertained with confidence, it is not unlikely that some people were thinking along the lines sketched by Stengel and Erdmann, since tenth-century English kings were frequently presented as ruling a collection of peoples. The ordo that was probably first used for Æthelstan's consecration referred to the king ruling two peoples (i.e. the Saxons and the Mercians), and in later versions this was increased to three, who were named as the Saxons, the Mercians and the Northumbrians.Footnote ¹⁷ Edgar's obit in the D and E texts of the Chronicle calls him ‘ruler of the Angles, friend of the West Saxons and protector of the Mercians’.Footnote ¹⁸ Especially in the mid-tenth century, charter styles quite frequently assert that kings ruled both the English and unnamed neighbouring peoples, as in rex Anglorum ceterarumque gentium in circuitu persistentium gubernator et rector (‘king of the English and governor and ruler of other peoples dwelling round about’).Footnote ¹⁹ It is particularly notable that the largest cluster of occurrences of the word imperator is in a group of mid-tenth-century charters that refer to an assemblage of peoples, using styles like rex Ængulsæxna ond Norðhymbra imperator paganorum gubernator Brittonumque propugnator (‘king of the Anglo-Saxons and imperator of the Northumbrians, governor of the pagans and defender of the Britons’).Footnote ²⁰ The notion of rulership of Britain is often mentioned in the context of analyses of possible ‘imperial’ claims, but these two pretensions should be distinguished. The examples just cited show that a king did not need to claim authority throughout Britain to be understood as the ruler of several peoples and therefore a potential imperator. Equally, a claim to rulership of Britain would not require an ‘imperial’ ideology. Since my argument is not predicated upon the presence or absence of any ‘imperial’ conception of kingship, I leave that slippery and probably irresolvable issue open, and focus on the question of why it was claimed that certain kings ruled the whole island.

There is reason to think that charter styles did not merely reflect drafting fashions, but also bore some relation to political reality: the presentation of Æthelstan as ruler of Britain began quite soon after the Eamont meeting, and Edmund was not thus described while York was under Hiberno-Scandinavian rule. To understand why English kings were sometimes styled rulers of Britain, we therefore need to assess the nature of their relationships with neighbouring rulers in the late Anglo-Saxon period, focusing initially on the century or so between Alfred and Edgar. The extant contemporary evidence is sparse and almost all English in origin: as is often pointed out, there is a risk that this may tempt us to overstate English domination. Each reference to contact between English and neighbouring kings has been extensively picked over, and there are many hypotheses about what was demanded or agreed on particular occasions. My intention here is neither to assemble a complete corpus of the evidence nor to dwell on the detail of what may have happened in any individual incident. Rather, I aim to assess the general character of English kings’ dealings with their neighbours, within the severe constraints of the extant sources.

Charter attestations constitute the clearest evidence that English kings were at least sometimes dominant over other rulers. We may begin with the only attendee of the Eamont meeting to whom the Chronicle does not accord the title cyning (‘king’), Ealdred of Bamburgh. Between 930 and 932, an Ealdred dux, very probably the man who was at Eamont, attested several of Æthelstan's charters, most of which were issued south of the Thames: that Ealdred travelled far from his base on the Northumbrian coast to attend assemblies shows his acknowledgement of Æthelstan's superiority.Footnote ²¹ Ealdred probably died in 934, but his family retained power in north-eastern Northumbria for over a century and continued to witness diplomas from time to time, which indicates that they recognised the dominance of Æthelstan and his successors, at least intermittently.Footnote ²² Men whom English sources at least sometimes styled ‘king’ also appear in witness lists. Five Welsh kings, the most famous of whom is the Hywel who was present at Eamont, witnessed several charters from between 928 and 935, the majority of which were issued well within Æthelstan's heartlands. Constantine II, the Scottish king and another Eamont attendee, is known to have attested at Buckingham and Cirencester in 934 and 935, immediately after an English campaign in Scotland; Owain, the king of Cumbria or Strathclyde (the terms are interchangeable), appears at Worthy, Cirencester and Dorchester in 931 and 935. These men were styled subregulus (‘subordinate ruler’), which was at least one notch below the titles sported by the English king, such as rex or basileus.Footnote ²³ Like Ealdred of Bamburgh, the subreguli journeyed far into the core territories of the English king, thereby recognising his superiority. No genuine extant charter is witnessed by a Scottish king after 935, and the only possible subsequent attestation by a Cumbrian ruler is the Malcolm dux who appears in a dubious diploma dated 970, but there is another clutch of Welsh attestations between 946 and 956.Footnote ²⁴ Save for Malcolm's possible appearance in 970, all the Welsh, Cumbrian and Scottish attestations occur in two distinctive groups of diplomas, the so-called ‘Æthelstan A’ and ‘alliterative’ charters.Footnote ²⁵ That these rulers’ presence was recorded may be a result of the foibles of the draftsmen responsible for these particular series: Welsh, Cumbrian and Scottish rulers could have attended assemblies at other times without being noted in extant witness lists, and the lack of attestations after 956 need not indicate that English domination of Wales and northern Britain had loosened.Footnote ²⁶ In this regard, it is interesting to note that no extant witness list records the presence of any Welshman at Edmund's court in 945, but a Welsh king apparently aided his ravaging of Cumbria that year.Footnote ²⁷

The best-known occasion on which a tenth-century English king met with neighbouring rulers was in 973, when the D and E texts of the Chronicle state that Edgar went with a naval force to Chester, where six kings met him and ‘with him they all pledged (getreowsodon) that they would be cooperators (efenwyrhtan) on sea and on land (on sæ 7 on lande)’.Footnote ²⁸ The twelfth-century re-workings of this passage by John of Worcester and William of Malmesbury supply the names of eight men, who can plausibly be identified with Scottish, Cumbrian, Hiberno-Scandinavian and Welsh (or Breton) potentates active in the 970s. The discrepancy in the number of kings need not be suspicious, since the compiler of the Chronicle could have regarded two of those present as sub-regal. John and William may well have drawn on some early account of those who gathered at Chester, although their lists could alternatively reflect an eleventh- or twelfth-century guess as to who was likely to have attended a meeting in 973.Footnote ²⁹ We can, however, be reasonably confident that the king of Scots was one of those present, since Ælfric, writing no more than about twenty-five years later, stated that ‘all the kings who were on this island, of Cumeras and of Scottas, came to Edgar – once, on one day, eight kings – and they all bowed (gebugon) to Edgar's rule (wissunge)’.Footnote ³⁰ The meaning of Cumeras is doubtful: it could be a reference to the Welsh, the Cumbrians or perhaps both. The precise identities of those present are not critical here: the key point is simply that the king of Scots and a number of other insular rulers met Edgar at Chester. It has been claimed that Chester, which was frequented by merchants from around the Irish Sea, constituted a neutral venue.Footnote ³¹ It was not. It was a town with which Edgar had a longstanding connection: one of his earliest extant charters granted six estates in western Cheshire to a minster in Chester.Footnote ³² That Chester was firmly under Edgar's control in the last years of his reign is demonstrated by the minting there of coins of his reformed type. The reform was associated with a marked reduction in the number of moneyers active in Chester; the suddenness of this change is suggestive of an administrative imposition, rather than a shift in popular demand for coins.Footnote ³³ As in Æthelstan's reign, neighbouring kings were in 973 acknowledging the superiority of an English king by meeting him on his own ground, clearly at his instigation.

In guessing what Edgar wanted to achieve at Chester, we should bear in mind two events that happened earlier the same year. The first is well known: shortly before the Chester meeting, Edgar was consecrated at Bath at Pentecost. Janet Nelson argues persuasively that he had already been consecrated near the beginning of his reign, and that the Bath consecration was intended to stress his rule of Britain, rather than merely of the English: the ordo most likely followed in 973 was very similar to that which was probably used for Edgar's first consecration, but a prayer that the king be honoured pre cunctis regibus terrae (‘above all kings of the land’) was changed to pre cunctis regibus brittaniae.Footnote ³⁴ If Nelson is correct that the Bath consecration was concerned with articulating pan-British pretensions, the Chester gathering would serve to prove that these were far from empty pomp. The second event is less familiar to most English historians. It occurred seven weeks before the Bath consecration, at Quedlinburg in eastern Saxony: Otto I, who had recently conquered much of Italy and whom the pope had crowned emperor in 962, celebrated Easter in the company of the dukes of Poland and Bohemia, along with legates from the Greeks, Beneventans, Hungarians, Bulgars, Danes and Slavs. Shortly before, Edgar had sent an embassy to Otto and surely learned of plans for the Quedlinburg gathering, to which the Chester meeting presents clear parallels.Footnote ³⁵ Whatever specific promises of cooperation Edgar wanted to secure at Chester, it is hard to avoid the inference that a large part of the meeting's purpose was to demonstrate Edgar's strength and to ape the most powerful ruler in Europe. If displaying power was indeed a central objective, Edgar may have selected Chester as the venue because its visible Roman remains, like those of Bath, would call to mind an earlier Britain-wide hegemony.

The superiority of English kings was at certain times demonstrated in various other ways, besides visits to their courts. We have already seen that a Welsh king assisted Edmund in his ravaging of Cumbria in 945, and charter attestations suggest that Æthelstan may have received military aid from three Welsh kings on his Scottish expedition of 934.Footnote ³⁶ Referring to this campaign, the Chronicle merely records that Æthelstan ravaged much of Scotland with both land and naval forces, but John of Worcester, quite possibly reproducing an earlier source, adds that Constantine then handed over his son as a hostage, along with ‘worthy gifts’, presumably tribute in some form.Footnote ³⁷ There is firmer evidence of the Welsh paying tribute: Armes Prydein, a prophetic poem that could plausibly date from almost any time between the mid-ninth and eleventh centuries, bemoans the rapacity of the stewards of an unnamed English king.Footnote ³⁸ Spiritual patronage could be used to express superiority and to create or reinforce reciprocal bonds of obligation: Asser states that Alfred stood sponsor to Anarawd ap Rhodri at his confirmation and, less reliably, William of Malmesbury reports that Æthelstan ordered the baptism of Constantine's son and became his (the son's) godfather in 927.Footnote ³⁹ English kings also used spiritual patronage in their dealings with Scandinavians: Alfred became the godfather of both Guthrum and a son of Hæsten, and Edmund sponsored Olaf Sihtricson at his baptism and Ragnald at his confirmation.Footnote ⁴⁰

References to encounters between English and neighbouring kings often allude to some kind of agreement of common purpose. Thus, the Chronicle states that those who assembled at Eamont in 927 renounced idolatry, that the Scottish king agreed in 945 to be Edmund's cooperator (his midwyrhta – literally ‘his together-worker’) and that the kings who met at Chester in 973 likewise promised to be cooperators (efenwyrhtan).Footnote ⁴¹ It is important to remember that cooperation is not the same as equality. Aside from the references in the D and E texts of the Chronicle for 973, there are four extant occurrences of efenwyrhta, efenwyrcend and emnwyrhta. All appear in Old English translations of Latin texts and all translate Latin cooperator or operator. In the two occurrences in the Old English Bede, the cooperators are an abbot and a monk (Benedict Biscop and Ceolfrith) and an archbishop and an abbot (Theodore and Hadrian): in neither case were the men who worked together equals in power or status.Footnote ⁴² The same point can be made about midwyrhta: the Old English Soliloquies states that God is a mydwyrhta to each well-working person.Footnote ⁴³ Cooperation need not be voluntary and parties can cooperate in pursuit of an objective that is for one party's benefit: it is quite conceivable that English, Welsh, Cumbrian and Scottish kings agreed to work together to further a largely English agenda. It is likely that the key demand that English kings made was that their counterparts should not support Hiberno-Scandinavians seeking to maintain or establish a position in Northumbria. The exclusion of Hiberno-Scandinavians was not necessarily congruent with Welsh, Scottish or Cumbrian interests: Hiberno-Scandinavian rule at York might reduce the number of footloose raiders in the Irish Sea and would provide a buffer against northward West Saxon expansion. Indeed, Welsh, Cumbrian and Scottish rulers sometimes allied with Scandinavians: Anarawd of Gwynedd appears to have cooperated with Guthfrith of Northumbria in the late ninth century, and in 937 Æthelstan defeated an alliance of Olaf Guthfrithson, Constantine and (probably) a Cumbrian king.Footnote ⁴⁴ The Armes Prydein poet dreamed of an even grander anti-English coalition, encompassing the Welsh, Dubliners (i.e. Hiberno-Scandinavians), Bretons, Cornish, Irish, Cumbrians and Scots.Footnote ⁴⁵

Why did Welsh, Cumbrian and Scottish kings recognise English superiority, particularly if this involved handing over hostages and tribute, or promising to forsake potentially attractive Hiberno-Scandinavian alliances? The obvious answer is that English kings were in all probability the wealthiest and most militarily powerful men on the island and could therefore coerce their neighbours. The clearest instance of this is Æthelstan's ravaging of Constantine's kingdom in 934: immediately after, Constantine attested charters at Buckingham and Cirencester, and quite possibly paid tribute and gave his son as a hostage. Having regained York in 944, Edmund demonstrated that he too could inflict destruction in northern Britain: the next year, he ravaged Cumbria and blinded two sons of its king.Footnote ⁴⁶ It is clear that Edgar likewise had formidable coercive capabilities: in the decades after his death, several writers remarked upon his naval strength and there is good circumstantial evidence that he employed a powerful mercenary fleet.Footnote ⁴⁷ If Edgar's neighbours wished to avoid their lands being ravaged, the invitation to Chester was probably not one that they could decline. It is, moreover, interesting that the Chronicle notes that Edgar was accompanied to Chester by his scipfyrd or sciphere (naval force), which presumably served to focus the mind of any visitor who had contemplated refusing to cooperate.Footnote ⁴⁸

Certain historians of Wales and especially Scotland have sought to explain away evidence that ‘their’ kings recognised English superiority.Footnote ⁴⁹ Some others accept that submissions were made, but protest that they were forcibly extracted, thus exonerating Welsh and Scottish kings from the charge that they cravenly sucked up to the English.Footnote ⁵⁰ While the threat of coercion was important, we should not assume that kings were reluctant to acknowledge English domination. There is no reason why they ‘ought’ to have seen themselves as on a par with militarily stronger English kings: indeed, Dauvit Broun argues that it was only in the first half of the thirteenth century that Scottish kings began to assert that they were of equal status to their English counterparts.Footnote ⁵¹ Submission need not have been humiliating: when Welsh, Cumbrian and Scottish rulers visited the English king, they appear to have been treated with honour, attesting above the non-royal English lay witnesses and, in the ‘Æthelstan A’ charters, also above the bishops and sometimes the archbishops.Footnote ⁵² Nor need the burdens associated with submission have been particularly onerous. The English king might demand non-alliance with Hiberno-Scandinavians, tribute, hostages and occasional military aid, but there is no evidence that a Welsh, Cumbrian or Scottish king's authority within his own kingdom was significantly compromised.

The ‘costs’ of accepting English domination are thus unlikely to have seriously undermined a king. The benefits, on the other hand, were not trivial and extended beyond being spared from attack. Asser spelled out some of these benefits in his account of Alfred's dealings with the Welsh. The passage in question is the most detailed extant description of a submission to a pre-Conquest English king, but its value does not lie merely in its detail. Since Asser was a Welshman in an English king's service, and may well have envisaged both Welsh and English readers for his work, his account is probably the nearest thing we have to a ‘balanced’ report of any relationship between English and neighbouring kings.Footnote ⁵³ Asser emphasises the dignified nature of the submissions: the Welsh kings initiated contact, Alfred received Anarawd honorifice (‘honourably’) and their relationship was one of amicitia (‘friendship’).Footnote ⁵⁴ It is, however, plain that the Welsh kings recognised Alfred's superiority: Hyfaidd of Dyfed ‘submitted himself to [Alfred's] royal overlordship’ (regali se subdiderat imperio), the kings of Glywysing, Gwent and Brycheiniog sought Alfred's dominium (‘lordship’) and Anarawd of Gwynedd agreed to be obedient to Alfred's will in all things (in omnibus regiae voluntati sic oboediens). Asser stresses the benefits that Alfred could offer the Welsh kings in return. One was familiaritas (‘friendship’ or ‘intimacy’), the attraction of which should not be dismissed: there are plenty of examples from better-documented periods of dignitaries being keen to hobnob with foreign royalty. Submission to Alfred could also apparently lead to an increase in worldly power, perhaps because familiaritas with a powerful external figure boosted a king's prestige in the eyes of his own subjects. Furthermore, Asser states that the Welsh kings gained wealth and mentions the extravagant gifts that Alfred showered on Anarawd. The bestowal of presents would indicate Alfred's high esteem for his Welsh clients and give them resources with which to reward their followers, but would also underline Alfred's superiority, unless the Welsh kings gave him equivalent gifts. Asser implies, however, that the most important attraction of Alfred's lordship, at least for the southern Welsh rulers, was the possibility of protection from Ealdorman Æthelred of Mercia (Alfred's son-in-law) and from Anarawd and his brothers. Alfred had some capacity to restrain Æthelred, who had (according to Asser) promised obedience to Alfred, and Alfred could threaten to attack (or to have Æthelred attack) Gwynedd if Anarawd troubled the southern Welsh kings. The effectiveness of this protection is doubtful, since the Annales Cambriae records that in or around 893 Anarawd ravaged south-west Wales cum Anglis. The Angli involved could have been Mercians rather than West Saxons, but Alfred had failed to shield some of his Welsh clients from assault; his domination in most of Wales may thereafter have been severely attenuated.Footnote ⁵⁵ A promise of protection was not without value, however: even though a protégé could not rely on support, the mere possibility that his lord might intervene would likely make potential aggressors hesitate before attacking.

It is probable that tenth-century English kings likewise gave familiaritas, gifts and promises of protection in return for cooperation, and there are a couple of clear examples of Scottish kings deriving benefit from being on friendly terms with their English counterparts. The Chronicle reports that in 945 Edmund ravaged Cumbria and granted (let) it to Malcolm I on condition that Malcolm be ‘his cooperator (his midwyrhta) both on sea and on land’. Contrary to what has sometimes been asserted, Cumbria (or Strathclyde) was at this time a distinct kingdom, not an apanage for the heir to the Scottish kingship: Edmund was helping one northern king to assert overlordship over another.Footnote ⁵⁶ The second case is Edgar's alleged grant of Lothian (the area between the Forth and the Tweed or some part thereof) to Kenneth II, reported in an early twelfth-century Durham text. If the account is reliable, the ‘grant’ must have been made between Kenneth's accession in 971 and Edgar's death in 975, most probably in connection with the Chester meeting in 973, when Kenneth and Edgar likely promised to be efenwyrhtan ‘on sea and on land’. Lothian had probably been under the de facto domination of the kings of Scots for some time and had never been subject to more than very loose West Saxon overlordship.Footnote ⁵⁷ Edgar is thus unlikely to have regarded his ‘grant’ as a substantive concession, but the security afforded by English acceptance that Lothian was within the Scottish sphere will have been of some value to Kenneth. In both 945 and the 970s, English kings recognised and aided the extension of Scottish royal power in return for promises of cooperation on land and sea.

It thus appears that for much of the century or so between Alfred and Edgar, Welsh kings recognised the superiority of the West Saxon king, who was able to constrain and mould their actions through threats and promises. From at least about 927, and perhaps (as we shall see) slightly earlier, the same was so vis-à-vis the rulers of Bamburgh, and the Cumbrian and Scottish kings. But English overlordship was not continuous and did not create a troop of pliant acolytes. The clearest demonstration of this is the Scottish and Cumbrian kings’ alliance with Olaf Guthfrithson against Æthelstan in 937, and it is likely that English kings’ dominance in northern Britain was minimal at times when there was a Hiberno-Scandinavian kingdom in Northumbria. There are also references in the so-called Chronicle of the Kings of Alba to Scottish raids on Northumbria in c. 949, c. 971 and c. 972, and the possibility of similar forays by Welsh kings cannot be discounted.Footnote ⁵⁸ But unless Welsh, Cumbrian and Scottish kings allied with each other or with a Hiberno-Scandinavian king, there is no sign that any of them could threaten the English king in any fundamental way. By contrast, it is implicit in Asser's account that already in the 880s Alfred was sufficiently powerful that the southern Welsh kings thought that he could constrain the king of Gwynedd in north-west Wales. By 934, moreover, the English king could cause serious problems to any ruler on the island: that year, Æthelstan's forces ravaged as far as Caithness, going further north than even Edward I is known to have ventured.Footnote ⁵⁹ The campaign of 934 was not a campaign of conquest, intended to leave Æthelstan with any enduring capability to impinge directly upon the lives of the mass of the population of northern Britain. Rather, it enforced the English king's personal domination over another ruler, but only temporarily. Tenth-century English kings’ power in Wales and (even more so) northern Britain was a loose and somewhat intermittent hegemony, ‘extensive’ rather than ‘intensive’.Footnote ⁶⁰ English kings were nonetheless quite capable of coercing neighbouring rulers: the terms upon which they dealt with their Welsh, Cumbrian and Scottish counterparts did not, pace some recent commentators, approach anything like equality.Footnote ⁶¹

This assessment of the tenth-century evidence resembles Rees Davies's characterisation of how English kings kept Welsh and Scottish kings in line in the century after the Norman Conquest. From the late twelfth century, English domination became increasingly precise, onerous, threatening and sometimes humiliating: tacit understandings were replaced with written codifications, and English kings asserted tenurial and jurisdictional claims over subordinate rulers’ territories. Before the latter part of the twelfth century, however, English kings' hegemony was rather less oppressive: their loose superiority was honourably acknowledged through intermittent visits to their courts, the handing over of hostages and tribute and the provision of occasional military aid.Footnote ⁶² It is quite likely that most post-Conquest English kings had somewhat more capacity than their tenth-century predecessors to intervene in Scottish and Welsh affairs, but, as in the tenth century, English domination of Wales and northern Britain was not continuous in the century after the Norman Conquest, collapsing during Stephen's reign. The tenor of English kings’ relationships with Scottish and Welsh rulers does not seem to have been radically different between the Conquest and the late twelfth century from how it had been between c. 927 and c. 973.

The theme of English domination of Britain is less prominent in modern discussions of the years between c. 973 and c. 1066. For this period, we have a scattering of references to English magnates raiding in Wales, and to incursions by both Welsh and Scottish kings, but contemporary references to contacts between English kings and the Welsh, Cumbrians and Scots are even scarcer than in the period between Alfred and Edgar.Footnote ⁶³ One might hypothesise that this near-silence reflects that English domination in Wales and northern Britain crumbled in the ninety years between Edgar and the Norman Conquest. Such a conjecture may have some substance, particularly with regard to the second half of the reign of Æthelred II (978–1016), who must have been preoccupied with attempting to resist Scandinavian attack. Even then, however, Æthelred was able to ravage Cumbria in 1000 and it is unlikely that the offensive capabilities of English kings were significantly weakened during the eleventh century.Footnote ⁶⁴ Nor is it likely that the power of eleventh-century Welsh and Scottish rulers (with the possible exception of Gruffudd ap Llywelyn) was in a different league from that of their predecessors.Footnote ⁶⁵ If English kings’ domination did loosen in the ninety years after Edgar's death, this was probably at least as much a consequence of a change in the balance of power within Ireland as of any change in the balance of power within Britain. In 980, Máel Sechnaill, king of Mide, defeated Olaf Sihtricson, king of Dublin and sometime king of York. The Hiberno-Scandinavian kingdom of Dublin survived, but was thereafter a more minor political player, with no realistic chance of obtaining a substantial foothold in Britain.Footnote ⁶⁶ The principal reason why Edgar and his predecessors sought domination over Welsh, Cumbrian and Scottish rulers was probably to deter them from giving succour to anyone seeking to establish or maintain a Hiberno-Scandinavian kingdom in Northumbria, a prospect that receded after Olaf's defeat in 980.The Scandinavians who afflicted Æthelred primarily came across the North Sea and did not (so far as we know) attempt to ally with any king in Britain: the need for English kings to expend effort and resources on keeping neighbouring rulers in line thus became less pressing.

Even so, we should be wary of the conclusion that eleventh-century English kings were substantially less dominant over their Welsh, Cumbrian and Scottish counterparts than tenth-century kings had been. In or around 1031, Cnut went to Scotland and, according to the Chronicle, received some sort of recognition of superiority from Malcolm II, the king of Scots, and two other northern kings, Macbeth and Echmarcach.Footnote ⁶⁷ In 1054, Earl Siward of Northumbria led an expedition into Scotland and defeated Macbeth, by now king of Scots. The D text of the Chronicle states that some of Edward the Confessor's housecarls were killed during this campaign, which indicates that it was undertaken with Edward's approval, and perhaps at his instigation. Unlike in Shakespeare's play, Macbeth did not die in battle against Siward, but was killed in 1058 by Malcolm III. What happened during the intervening four years is obscure, but it appears that Macbeth was replaced as king before his death by his stepson, whom Malcolm also killed in 1058. Siward was not necessarily seeking to set Malcolm in Macbeth's place, but defeat probably undermined Macbeth and played a major part in his downfall.Footnote ⁶⁸ Something similar happened in Wales in 1063/4, when Earls Harold and Tostig attacked Gruffudd ap Llywelyn's territory. According to the C text of the Chronicle, Gruffudd had sworn in 1056 to be ‘a loyal and faithful underking (underkingc) to King Edward’, but two years later he aided the exiled Earl Ælfgar, a rival of Harold and Tostig. Following Harold and Tostig's campaign, Gruffudd ‘was killed by his own men because of the fight he fought against Earl Harold’. Gruffudd's head was brought to Harold, who presented it to Edward. Edward and Harold then set up two of Gruffudd's half-brothers in his place; they gave hostages and swore oaths that they would pay tribute and ‘be everywhere ready on water and on land’, presumably a promise of military assistance.Footnote ⁶⁹ The English king, sometimes acting through his earls, could still represent a serious threat to any other ruler on the island.Footnote ⁷⁰ Cnut and Edward perhaps gave practical expression to their superiority even more intermittently than their tenth-century predecessors or their twelfth-century successors, but looking across a long time span, the broad continuities between Æthelstan and Henry II are striking: while the extent of English kings’ domination over their neighbours fluctuated, a loose but unmistakable overlordship over the other kings in Britain was a general feature of these two-and-a-half centuries.

With this in mind, we can now tackle the specific problem posed by my title, approaching it via the question of why claims to rule Britain became less ubiquitous after the tenth century. One could explain their reduced prominence in the eleventh century by pointing to the possible loosening of English overlordship, but this explanation would not account for the rarity with which Norman, Angevin and Plantagenet kings were described as rulers of Britain. Rees Davies's central concern in his Ford Lectures was to explain why twelfth- and thirteenth-century kings were by and large not presented as rulers of Britain, in spite of both their undoubted power and of post-Conquest writers’ celebration of the pan-British pretensions of Æthelstan and Edgar. In proposing an explanation for this problem, Davies acutely observed that to explain how the English understood their relationship with Britain one ‘must look towards England itself’.Footnote ⁷¹ Above all, he stressed the importance of a strong sense of English identity, which was often coupled with strident contempt for the allegedly barbarous Welsh and Scots: this, Davies argued, precluded the presentation of post-Conquest kings as rulers of Britain.Footnote ⁷² Furthermore, Davies drew attention to how the uniformity and ubiquity of royal administration within the English kingdom contrasted with the looser structures of central rule in other parts of Britain: since English kings’ power was fundamentally different within the English kingdom from how it was elsewhere, it was hard to think of them as kings of Britain.Footnote ⁷³

Davies regarded analysis of the English kingdom as central to understanding the dealings of the English with other peoples in the British Isles, but also highlighted that understanding of the English kingdom itself can be sharpened through analysis of how the English related to their neighbours.Footnote ⁷⁴ My object here is to use Davies's arguments about identity and administration in the twelfth and thirteenth centuries to ask questions about the English kingdom in the period when it was coming into being, the tenth century. Davies declared, following Patrick Wormald and Sarah Foot, that ‘the medieval English state . . . had already by the tenth century forged its identity, as the polity of the gens Anglorum, the Angelcynn’, and that this identity was associated with a belief that the English were ‘the new Israel’.Footnote ⁷⁵ The logic of Davies's argument about the post-Conquest period does not seem to have led him to question these axioms that Anglo-Saxonists presented to him. But if Davies was correct that the chauvinistic rigidity of English identity was a prime reason why post-Conquest kings were not called rulers of Britain, and if this English identity was already deeply entrenched and coupled with an exclusivist ideology of divine election by the tenth century, would it not be odd that Æthelstan and Edgar were so often lauded as kings of Britain?

There is plenty of evidence from throughout the Anglo-Saxon period that the Germanic inhabitants of Britain could collectively be regarded as the gens Anglorum. From at least the second half of the ninth century, this expression was rendered in the vernacular as Angelcynn. But, as I have argued elsewhere, there are no good grounds to think that this English identity was tied to any particular political vision in the late ninth century: the conclusions of Wormald and Foot are predicated upon dubious assumptions about how Bede was received by later generations. There is no sound basis for believing that the English conceived of themselves as ‘the new Israel’, God's specially chosen people, at any time during the Anglo-Saxon period. There is scarcely more reason to think that Alfred and his successors set out to unify the English. The greatest cluster of occurrences of the word Angelcynn is in the Old English Bede. Once one recognises that this text is not concerned with promoting any notion of the English as a single special people, the argument that other instances of the word Angelcynn indicate a programme of nation-building and unification dissolves.Footnote ⁷⁶ That Edgar was content to recognise Scottish authority over Lothian implies that English unification was not a priority, since Lothian had been part of the old Northumbrian kingdom and long continued to be seen as in some sense ‘English’: Adam of Dryburgh, writing in what is now the Scottish Borders in 1179/80, regarded himself as living ‘in the land of the English (terra Anglorum) and in the kingdom of the Scots (regno Scotorum)’.Footnote ⁷⁷ The aim of Alfred and his tenth-century successors was not to create a kingdom of all the English, but to contain, subdue and ultimately expel the Scandinavian potentates who had gravely threatened Wessex. Welsh, Cumbrian and Scottish rulers did not present anything like the same level of menace: if English kings could secure their cooperation, the potential benefits of seizing their territory probably did not in most cases (including that of Lothian) appear commensurate with the effort that would have been required.

If one sets aside the idea that the West Saxon kings were seeking to unite the English, it is unsurprising that they were accorded a range of titles. As we have seen, Æthelstan and his successors could be the king of the English, or the king of Britain or Albion, or the king of a collection of peoples.Footnote ⁷⁸ This variety reflects the fact that the geographical extent of West Saxon power was increasing greatly in the tenth century, and contemporaries did not know what form the expanding kingdom would take: draftsmen therefore experimented with different ways to conceptualise royal authority. In the tenth and eleventh centuries, the strident sense of English superiority that Davies saw as so inimical to any idea of a kingship of Britain remained inchoate at most. A couple of tenth-century charter styles alluded vaguely to the barbari who neighboured the English, and the Life of Edward the Confessor (probably written 1065–7) made a few disparaging remarks about the Scots and Welsh, but, as John Gillingham has shown, it was in the first half of the twelfth century that sustained contempt for the Scots and Welsh was first seriously articulated by English writers.Footnote ⁷⁹ That Æthelstan and Edgar were frequently called rulers of Britain reflects that their kingship was not tied to a specifically English ideology, let alone to a clearly formed belief in English moral superiority.

While Davies particularly emphasised the significance of English chauvinism, he also argued that it was difficult to think of post-Conquest kings as rulers of Britain because their loose and intermittent domination of Wales and Scotland was so unlike their intensive and arguably ‘bureaucratic’ rule within the English kingdom. Davies neither wholeheartedly endorsed nor seriously challenged the claims of James Campbell and Patrick Wormald about the administrative sophistication of ‘the Anglo-Saxon state’, but his argument again implicitly raises questions about the tenth century: if the intensity of royal administration within the English kingdom made it hard to conceive of post-Conquest kings as rulers of Britain, was Æthelstan called rex totius Britanniae in part because his administrative apparatus in the nascent English kingdom was rudimentary?Footnote ⁸⁰ Was his power there not so very different in nature from in Wales and northern Britain?

It has been established beyond reasonable doubt that in the last century or so of the Anglo-Saxon period kings were capable of impinging as a matter of routine upon the lives of quite ordinary people living across most of what is now England.Footnote ⁸¹ Æthelred II raised vast sums to buy off Scandinavian attackers and to hire mercenaries, and an annual land tax appears to have been levied between 1012 and 1051. The assessment and collection of taxation must have required formidable coercive and administrative power: the division of the kingdom into shires and hundreds appears to have been crucial.Footnote ⁸² These same administrative units, and the network of local agents (especially sheriffs) associated with them, formed the basis of a judicial system through which the king could regulate society at a local level.Footnote ⁸³ Similar conclusions about royal capabilities can be drawn from the coinage. From sometime towards the end of Edgar's reign onwards, a single design of coin was almost always struck throughout the kingdom at any one time, with the design being changed frequently. Since many hoards are composed of a single type, or a small number of chronologically close types, it appears that old coins were to a significant extent withdrawn from circulation, probably under some form of compulsion. That people could seemingly be induced to exchange their coins for the new type, very likely at disadvantageous rates, is again indicative of administrative sophistication and coercive strength.Footnote ⁸⁴ Some of the more extravagant claims that have been made for Anglo-Saxon administrative genius are dubious: it is, for example, surely permissible to doubt that all variations in the weight and fineness of coins reflect a meticulous royal plan. But it is clear that late tenth- and eleventh-century kings, like their post-Conquest successors, ruled the English kingdom in an ‘intensive’ manner, quite different from their loose, intermittent, ‘extensive’ domination of Wales and northern Britain.

Much work on how kings ruled the early English kingdom suffers from a degree of chronological imprecision. Surveys of the means of royal rule typically treat the period between Alfred and the Norman Conquest as a block. While those who have written such accounts often allude to developments over the course of the period, they are usually vague about what changes are likely to have occurred when: in consequence, the tenth century seems broadly similar to the eleventh and Alfred's reign appears (at least implicitly) as the most decisive stage in the formation of the administrative basis of late Anglo-Saxon kingship.Footnote ⁸⁵ Given the scarcity of sources, it is tempting to take a lengthy timespan and to construct a single picture from the scattered fragments of evidence. Such an approach, however, risks fudging change over time.

A detailed analysis of how tenth-century kings ruled what was becoming the English kingdom is beyond this article's scope. Here we need only focus on two salient problems. The first is whether the northern extension of West Saxon power was accompanied by mass expropriation of existing landholders, with their estates being retained by the king or granted to his West Saxon associates. Since we have no tenth-century Domesday Book, and only a few diplomas concerning lands north of an imaginary line stretching from Worcester to the Wash, we can only make educated guesses. It is almost certain that a considerable quantity of land entered royal hands during the territorial expansion of the first half of the tenth century: this is the most obvious way to explain how kings had by 1066 acquired a fair amount of land in most shires north of the Thames. Ælfwynn of Mercia and the Hiberno-Scandinavian kings of Northumbria were deprived of power and it is likely that many or all of their lands were appropriated by the king.Footnote ⁸⁶ But the evidence for widespread expropriation of persons below the highest social ranks is slight. The Chronicle refers time and again to people in the East Midlands, East Anglia and Essex ‘turning’ to Edward the Elder or ‘choosing him as lord’ after he showed up with his army. This may suggest that those who accepted Edward's lordship, albeit under threat of force, were allowed to retain at least some land.Footnote ⁸⁷ Such a conclusion would accord with an account of a dispute about land in Huntingdonshire, which occurred soon after Edgar's death. The matter hinged on whether the claimants’ uncle's grandmother had made a timely submission to Edward. Since she had not, she forfeited her land but, had she submitted, she would apparently have retained it: the case reveals expropriation, while simultaneously suggesting that it was far from universal.Footnote ⁸⁸

Even scarcer than evidence of expropriation is evidence of the intrusion of people of West Saxon origin into newly acquired territory. It can hardly be doubted that some West Saxons gained lands north of the Thames, but the supposition that the upper aristocracy was overwhelmingly West Saxon is not founded on any secure evidence. It is sometimes stated that Æthelstan Half-King and Ælfhere, ealdormen north of the Thames and the two most powerful lay magnates of the mid-tenth century, hailed from Wessex.Footnote ⁸⁹ The basis for this belief is that they are known to have held lands there, but one could just as well postulate that they were Mercians who had been granted estates south of the Thames. Such a possibility is suggested by Asser's account of how Wærferth and Plegmund, two Mercian churchmen who entered Alfred's service, were showered with honours and powers in Wessex, in addition to those that they had in Mercia.Footnote ⁹⁰ On a balance of probabilities, it appears that West Saxon expansion was achieved primarily through a series of local agreements in which incumbent landholders were cowed into submission, rather than through widespread forcible expropriation.

The second fundamental issue that needs to be considered is the character of tenth-century royal rule in what was becoming the English kingdom. Much of the secondary literature implies (and sometimes states) that Alfred had a sophisticated administrative apparatus south of the Thames and that this was briskly replicated elsewhere, with shires and hundreds being rolled out as Edward the Elder and Æthelstan pressed north.Footnote ⁹¹ Even on grounds of general probability, one might doubt this: it would be surprising if kings could effect immediate and fundamental administrative reorganisation of newly acquired territories while simultaneously expending significant time and resources on campaigns to secure further submissions. Such a scenario would be even more unlikely if incumbent landholders for the most part remained in place. To overcome the argument from basic probability, one would want firm evidence, which is altogether lacking. Except in the case of Essex, the first evidence that any part of the area north of the Thames was divided into shires is from the 970s.Footnote ⁹² Essex and the shires south of the Thames are attested before the end of the ninth century, but their only identifiable function at that time was as military units: there is no sign of the routine holding of shire meetings in any part of the kingdom before Edgar's reign.Footnote ⁹³ It is also around this time that the sheriff, the king's key local agent in the eleventh century, made his earliest appearance.Footnote ⁹⁴ Hundreds are first mentioned in the legislation of Edmund (939–46), but it is from Edgar's reign onwards that references to them become ubiquitous: whatever previous existence hundreds possibly had, it appears that it was in the second half of the tenth century that they became fundamental units of royal administration.Footnote ⁹⁵

Earlier in the tenth century, fortifications (burhs) built south of the Thames by Alfred, and in the Midlands by Edward the Elder and Æthelflæd, served as outposts of royal authority, but only a small proportion of the population lived in them. Edward and Æthelstan sought to ban or limit trade outside a port, and it appears that a port was much the same sort of place as a burh: this suggests that early tenth-century kings had administrative structures (most likely a network of reeves) through which to monitor transactions in burhs, but that kings lacked a comparable apparatus elsewhere.Footnote ⁹⁶ By contrast, Edgar's later legislation presupposed that the whole kingdom was divided into hundreds (or wapentakes), and ordered that transactions be witnessed either in a burh or in a hundred (or wapentake).Footnote ⁹⁷ A similar shift from Æthelstan to Edgar can be detected with regard to royally ordained public meetings. Æthelstan's legislation implied that such meetings would be held in burhs, since the leading men of the burh were to seize the property of non-attenders.Footnote ⁹⁸ Edgar's legislation contains a similar provision, but ‘the leading men of the burh’ are replaced with ‘men from the meeting’, the location of which is unspecified. That this is probably an allusion to hundred meetings can be inferred from the statement, not found in Æthelstan's legislation, that half of the non-attender's property was to go to the hundred.Footnote ⁹⁹ With regard to both trade and meetings, there appears to have been a significant change between Æthelstan and Edgar: administrative structures that had formerly been restricted to burhs were extended into the countryside through hundredal organisation. Kings thus developed mechanisms for social regulation throughout the kingdom, rather than merely in particular fortified centres.

That ‘absence of evidence is not evidence of absence’ is a well-known mantra. But when we search (in vain) for evidence of a sophisticated and kingdom-wide royal administrative apparatus in the first half of the tenth century, we are to some extent dealing with silent sources, not with a straightforward absence of sources: it is particularly striking that the substantial corpus of legislation from Alfred to Æthelstan says nothing about hundreds, and indeed ascribes to burhs functions that burhs would later share with hundreds.Footnote ¹⁰⁰ It is, moreover, notable that the Midland shires, routine shire assemblies and sheriffs appear almost simultaneously, around the time of Edgar's reign, and that the first descriptions of actual shire meetings occur very soon after the first reference to a scirgemot in royal legislation.Footnote ¹⁰¹ That different types of sources break their silence at the same time suggests that earlier silence indeed reflects absence.

The argument from silence is reinforced by numismatics. The ability of Edgar and his successors to enforce uniformity in coin design and frequent changes of type is one of the surest signs of the sophistication of royal administration in the last ninety years of the Anglo-Saxon period. Between Alfred's death and Edgar's reform, the situation was quite different: there were no kingdom-wide changes of type and different designs were struck simultaneously in different regions. There were also significant variations in weight and fineness, especially in the third quarter of the tenth century. Alfred appears to have had reasonably close control over minting in his kingdom south of the Thames and, from the late 870s, in western Mercia: his coinage was fairly uniform and he carried out changes of type.Footnote ¹⁰² His successors may well have maintained such control in the south, which displays a significant degree of coherence in terms of the type being struck. They were, however, either unable to enforce uniformity of minting in their newly acquired territories, or uninterested in doing so.Footnote ¹⁰³ Edgar's reign saw major change in the coinage: this supports the conclusion that the breaking of the silence of the written sources around the same time reflects similarly substantial changes in other aspects of royal administration. The reasons for these changes cannot be explored here: for my present argument, it is enough to establish that the latter part of the tenth century was critical in the development of a fairly uniform and sophisticated administrative apparatus across the kingdom.

Without such an apparatus, earlier kings are unlikely to have been able to impinge as a matter of routine upon the lives of more than a small proportion of their subjects. They could, however, use their wealth to secure the loyalty of the majority of the greater aristocracy, the only politically significant social stratum. Through gifts of land, treasure, office and assistance in disputes, kings could buy magnates’ support; in case of disloyalty, they could withdraw their favour. Even the most powerful men were vulnerable to forfeiture: a charter of Edward the Elder reveals that an ealdorman had been deprived of his land for breaching his oath to Alfred.Footnote ¹⁰⁴ Kings’ extensive landholdings almost certainly enabled them to maintain bigger and better military retinues than any of their subordinates.Footnote ¹⁰⁵ They could thus visit destruction upon those who displeased them: Eadred ordered a great slaughter in Thetford in 952 to avenge the killing of an abbot and Edgar had all Thanet ravaged in 969 as a reprisal for the mistreatment of some merchants.Footnote ¹⁰⁶ Tenth-century royal ‘justice’ could be crude and arbitrary, but the mere threat of a visit from the king's goons was probably sufficient to induce compliance with the royal will. Kings did not cease to manipulate magnates or ravage the lands of the recalcitrant after the late tenth century, but it was then that these means of rule appear to have been supplemented by administrative structures through which royal power could be routinely projected at a local level throughout the kingdom.Footnote ¹⁰⁷

Until the latter part of the tenth century, royal power in the nascent English kingdom was based squarely upon using rewards and threats to secure the loyalty of the greater magnates, not upon regulating the lives of the mass of the population. Something similar could be said of English kings’ domination in Wales and northern Britain: they coerced and cajoled neighbouring rulers into some measure of obedience, but did not impinge in any sustained way upon these rulers’ subjects. It will have been easier for English kings to punish disobedience in Thanet or Thetford than in Wales or northern Britain, and they did not have burhs in the latter areas, but the means by which Æthelstan and his immediate successors enforced their will on Scottish, Cumbrian and Welsh rulers were not fundamentally dissimilar to the methods they used within most of what we think of as England: before the late tenth century, the variations in English royal power in different parts of Britain were more quantitative than qualitative. Contemporary writers implied as much. Asser stated explicitly that Anarawd submitted to Alfred on the same terms as Æthelred of Mercia.Footnote ¹⁰⁸ Asser referred to Æthelred as comes while almost certainly thinking of Anarawd as rex (the title he accorded to weaker Welsh kings), but the practical power of a Welsh king and a leading English ealdorman may well have been fairly similar. A charter of Alfred from 889 styled Æthelred subregulus et patricius Merciorum, an obvious parallel with the presentation of Scottish and Welsh rulers as subreguli in tenth-century diplomas.Footnote ¹⁰⁹ A couple of writers are even known to have treated Æthelred as rex: Æthelweard twice styles him thus, and he appears as the final name in a regnal list that begins with Penda and is headed de regibus Merciorum.Footnote ¹¹⁰ Æthelred was not the only tenth-century ealdorman whose power was understood in quasi-regal terms: the most powerful English magnate of the second quarter of the tenth century was known as Æthelstan semi-rex.Footnote ¹¹¹

The words with which the Chronicle narrates English kings’ dealings with Wales and northern Britain often echo those used with regard to the area that was becoming the English kingdom. The Chronicle uses the same verb, oferhergian, to refer to English kings ravaging Thanet (under English rule), ‘between the dykes and the Ouse’ (under Danish rule), a part of Northumbria that was under Hiberno-Scandinavian rule, Cumbria and eastern Scotland.Footnote ¹¹² The Chronicle's statement that peace was established by Æthelstan, Constantine, Hywel and others at Eamont ‘with pledge and oaths’ (mid wedde 7 mid aþum) is reminiscent of the legislation of Alfred and Edward the Elder: Alfred ordered that every man abide by his að 7 his wed and Edward prescribed the payment of compensation by any man who breached the að and wæd that the whole people (ðeod) had given.Footnote ¹¹³ The oaths of cooperation on sea and on land made in 945 and 973 echo the promise of the Danish army of East Anglia in 917 to keep the peace ‘both at sea and on land’, and the same couplet appeared in Edward the Elder's legislation.Footnote ¹¹⁴ The specific terms upon which cooperation was agreed surely varied, but they were not so divergent as to require different conceptual frameworks.

The Chronicle refers to people from Bedford, Northampton, Cambridge and Stamford seeking Edward the Elder's lordship in 914, 917 and 918, then goes on to say that the Welsh kings ‘sought to have [Edward] as lord’ in 918.Footnote ¹¹⁵ Two years later, the Scottish and Cumbrian kings, the rulers of Bamburgh and a Hiberno-Scandinavian potentate who had recently established himself at York ‘chose [Edward] as father and lord’.Footnote ¹¹⁶ These relationships can hardly have been identical, but lordship was a sufficiently elastic concept to encompass them all: it entailed one party recognising the other's superiority and both agreeing obligations to each other, but there was plenty of scope for variation in what these obligations might be.Footnote ¹¹⁷ By the latter years of Edward's reign, it was probably clear that the English king was the most powerful man on the island: it is by no means implausible that Welsh and northern rulers acknowledged Edward's superiority and agreed to further his interests, perhaps in return for promises of non-aggression and protection.Footnote ¹¹⁸ Æthelstan may therefore not have been the first tenth-century king to be recognised by other rulers as the preeminent man in Britain.Footnote ¹¹⁹ More importantly, a contemporary writer did not need to draw a conceptual distinction between Edward's relationship with the people of Bedford and his relationships with the men ruling at York and Bamburgh, or indeed with the Welsh, Cumbrian and Scottish kings: all these relationships could be understood in terms of lordship.

English royal power will never have been uniform across the island, but in the first half of the tenth century the variations were much less marked than they later became. Through a mix of fear and favour, Edward the Elder and Æthelstan secured some measure of obedience from Welsh, Cumbrian and Scottish rulers, as they had previously done in the Midlands, East Anglia and Northumbria: they had power across Britain, but this power was more ‘extensive’ than ‘intensive’. The administrative reforms that appear to have been implemented in or around the reign of Edgar did not produce homogeneous administration within the new English kingdom: East Anglia may well not have been divided into shires until well into the eleventh century, and even at the end of the Anglo-Saxon period royal control was loose north of the Humber, and even looser north of the Tees.Footnote ¹²⁰ But the reforms that took place around Edgar's time did make the king's rule within much of the English kingdom substantially more pervasive, routine and ‘intensive’, while English domination of Wales and northern Britain remained loose, intermittent and ‘extensive’. The intensification of royal rule within the English kingdom fostered an increasingly strong sense that it was a single and discrete political unit, which was to be ruled by one rex Anglorum: Bishop Æthelwold (writing 964×84) condemned the partition between Eadwig and Edgar (957–9), and the Encomium Emmae Reginae (written 1041×2) presented the death of Edmund Ironside (1016) as an act of divine mercy since it ended the division of the kingdom between him and Cnut.Footnote ¹²¹ By the early eleventh century, the English kingdom had gained clearer definition: fundamental to this was the development of a marked difference in the nature of royal power inside and outside the kingdom.

While Æthelstan and Edgar were often described as plain rex Anglorum, or as rulers of the English and surrounding peoples, the great majority of their diplomas incorporated some explicit claim to rulership of Britain. These titles could be combined. Æthelstan was often rex Anglorum per omnipatrantis dexteram totius Brytanie regni solio sublimatus (‘king of the English elevated by the right hand of the Almighty to the throne of the whole kingdom of Britain’), and many charters of Edgar style him totius Brittannie basileus in the dispositive section and rex Anglorum in the witness list.Footnote ¹²² After Edgar's reign, claims to rulership of Britain by no means disappeared, but the proportion of diplomas that mentioned Britain decreased. Changes within the nascent English kingdom itself are the key to understanding this shift in titles. The notion that Æthelstan was rex totius Britanniae was not a bombastic fantasy. He did have power throughout Britain, but that power seems almost everywhere to have been ‘extensive’, founded upon personal bonds with potentates who recognised his superiority, not upon an administrative network that would enable routine engagement with the mass of the population. Contemporaries could see him as rex Anglorum, since he was a member of the gens Anglorum and could most easily cajole and coerce those living in the part of Britain that this gens inhabited. But as yet it often did not seem necessary to think of the king's authority as varying in nature in different parts of Britain. By the end of the tenth century, this was changing: from around the time of Edgar, a major difference opened up between ‘intensive’ rule and ‘extensive’ overlordship, such that it became increasingly problematic to conceive of Britain as a unitary realm. As with the pan-British domination attributed to Edwin, Oswald, Æthelbald and Offa, the titles accorded to Æthelstan and Edgar reflect strength, relative to contemporary neighbours. But such titles also reflect weakness, relative to subsequent English kings. Æthelstan and Edgar's successors, were, like them, in some sense overlords of Britain. These later kings, however, were also direct rulers of a kingdom that was more intensively governed than any other in the Latin west. This was not the case with Æthelstan's reign, and was only just becoming so in Edgar's time: until the late tenth century, intensive royal administration inside the English kingdom was not so entrenched as to impede the presentation of its kings as rulers of Britain.