Bernicia was the northern part of the old kingdom of Northumbria. It extended from Tees to Forth, including the later counties of Durham and Northumberland and south-east Scotland. Everyone agrees that (as with the southern kingdom of Deira) its name is Celtic. Yet its meaning has never had a satisfactory explanation, despite discussion by Professor Kenneth Jackson (1909–91), whose account remains standard. This paper, after summarizing Jackson’s comments and comparing them with opinion before and after, quotes evidence from Irish and other Celtic sources for a derivation based on the one he proposed.

Bede refers to the inhabitants as Bernicii, Bernicians. The ninth-century Welsh Latin Historia Brittonum gives the Old Welsh equivalents Berneich, Birneich, Bernech and Birnech. In Middle Welsh poetry the form is usually Bryneich, but Byrneich also occurs. Jackson rejected the derivation of these from Brigantes, ‘exalted ones; privileged ones, freemen’, used of a people of north Britain whose territory lay south of Bernicia, and we need not repeat what he said. Having established what it did not mean, he tentatively proposed a link with Old Irish bern, ‘gap, mountain pass’, and bernach, ‘full of gaps’. He reconstructed as their British cognates the forms Bernacci or Birnacci (with short a), ‘people of the land of the mountain passes’. These would have given unattested Primitive Cumbric Bernecc or Birnecc, the former giving Old English Bernice (in Northumbrian dialect) or Beornice (in West Saxon); and Bede’s Latinized Bernicii. Middle Welsh Bryneich would show simple consonant-shifting, the change that linguists call metathesis (as in Modern English bird from earlier brid).Footnote 1

Jackson’s rejection of a link between Bernicia and Brigantes meant taking issue with Sir John Rhŷs, whose status as an Oxford professor meant that his suggestion is influential even now, since it is accepted in Charles Plummer’s edition of Bede and Sir John Lloyd’s history of Wales (as was Rhŷs’s association of Deira with Welsh dwfr, ‘water’).Footnote 2 Jackson noted that Rhŷs’s etymology was also thought probable by Joseph Loth in France and Max Förster in Germany. It was taken seriously as well by Myres, who commented on how the ‘moorland and mountain fastnesses’ of Bernicia, a countryside ‘poor by nature and poorer still after the destructive raids of Picts and Scots had repeatedly passed over it on their way to the richer lowlands beyond’, hardly attracted English settlement.Footnote 3 Yet one person who did not embrace derivation of Bernicia from Brigant- with enthusiasm was Sir Ifor Williams (1881–1965), greatest of all Welsh scholars. He contented himself with noting early forms, but without comment on their meaning.Footnote 4 Professor Henry Lewis, also from the golden age of Welsh scholarship, noted that a thirteenth-century translator of Geoffrey of Monmouth translated his Deira as Deiuyr a Brennych, ‘Deira and Bernicia’. But Lewis likewise refrained from any etymology.Footnote 5 The silence of these two fine scholars on the point is telling and deserves respect.

Jackson’s rejection of Brigantes thus gave the coup de grâce to a theory that professional Celticists were already regarding coldly; and it is instructive to see how later writers received his arguments. Almost every book ever published on the Anglo-Saxons mentions Bernicia’s Celtic origins, with their authors often making a nod at the meaning of the word. Comment may begin with Jackson himself. In an outline of the phonological changes that marked the development of British into Welsh, Cornish and Breton, he stated as a fact that Bernicia could not ‘be derived from a stem *brigant-’.Footnote 6 Henry Loyn referred to Bernicia as an ‘Anglo-British kingdom’, where supposed British customs of rendering of services and dues survived for centuries; these were summarized (with a neat map) by Professor William Rees.Footnote 7 Charles Thomas looked to Scotland, relating the name to a British ruling dynasty of the Tweed basin, known solely from a sixth-century inscription at Yarrowkirk (in Selkirkshire/Borders) to Nudus and Dumnogenus, sons of Liberalis. He thought that these princes, who had thoroughly Celtic names despite the Latin one of their father, were perhaps of the Christian royal house of the ‘lost British *Berneich or *Bernaccia which, in the seventh century, continued in a fresh guise as the Bernicia’ of the Northumbrians.Footnote 8 In the late 1960s Jackson commented again on Bernicia, revealing a shift of ground. He still discounted the link with Brigantes, which he thought best to give up altogether. But on what the form did mean he was silent, since he described it as almost as difficult as that of Deira. So nothing now on bern or mountain passes.Footnote 9 By 1969 he must have felt less sure of the suggestion than he did in 1953.

Three books of 1971 comment on Bernicia. Leslie Alcock, from an archaeologist’s viewpoint, stressed Roman and British features of the Anglo-Saxon court at Yeavering and the lack of pagan burials in the region. He thought that the conclusion was ‘inescapable that in Bernicia a very small, and largely aristocratic, Anglian element ruled over a predominantly British population’.Footnote 10 Sir Frank Stenton wrote with sober eloquence of the Bernice as a people who, by the year 600, had survived fifty years of precarious existence on the coast, with their main stronghold on the rock of Bamburgh, to become formidable enemies of the Britons, so that within a generation they were to gain control of the whole country between Forth, Solway and Tees. But, despite explaining Dere, ‘Deirans’, as from British deifr, ‘waters’, he said nothing of their name.Footnote 11 Thomas repeats his own identification of the lost British polity of Tweeddale as Bernaccia, ‘which, like Deira and Elmet further to the south, became submerged (though not without trace) in the later Anglian kingdom of Northumbria’.Footnote 12 His remark raises a problem. Though Thomas did not mention it, his hypothesis weakens Jackson’s etymology, because Tweeddale contains the Merse, an extensive plain which is the very opposite of a region of ‘mountain passes’.

Other writers make their own points. John Morris emphasized the poverty of early Anglo-Saxon graves in Bernicia, their dead interred without goods, ‘save an occasional Saxon knife to prove their nationality’.Footnote 13 The English element in the local population was meagre. In lyric mode, Peter Hunter Blair (himself a native Bernician) portrayed Tweeddale as possessing ‘softer, richer lands, plentiful in rivers falling to the sheltered Tweed valley, the crimson lands of the Merse that bring wealth to those who own them, whether monastic orders or feudal earls’.Footnote 14 While some writers stress the poverty of Bernicia, Hunter Blair was more positive, indicating resources beyond those of its infertile coastline. That helps explain the growth of the English kingdom from puny beginnings to quasi-imperial power. Hunter Blair elsewhere described the etymology of Dere and Bernice as uncertain but as apparently based on ‘Welsh’ names, and cited Jackson’s etymology for the latter.Footnote 15 Peter Sawyer called Bernicia an ‘English kingdom with a British name’ and referred to research by Margaret Faull on Celtic aspects of it and Deira.Footnote 16 In discussing the Brigantes, a people of the Yorkshire region with their capital at Isurium, or Aldborough (near Ripon), Rivet and Smith followed Jackson and discounted any link with Bernicia.Footnote 17

In the 1980s, Charles Thomas spoke of native kingdoms that emerged in post-Roman Britain, including those of Deira, Elmet, Ergyng (south of Hereford) and Bernicia or *Bernaccia, which lay between Tyne and Tweed, and which (citing Jackson) he still took as perhaps ‘Land of the Mountain-passes’.Footnote 18 His enthusiasm for this derivation exceeded that of its originator, who had by then stopped mentioning it. James Campbell uses similar if brisker language. Bernicia’s name ‘is British (as is that of Deira to the south). It is likely that Bernicia was essentially British, but had fallen under the control of an Anglian dynasty’.Footnote 19 David Dumville comments on the absence from the older or B text of the Gododdin of any mention of Bernicia, the only threat to the Gododdin being Deira. This he uses as evidence for a sixth-century dating of these poems, earlier than that usually proposed.Footnote 20 Wallace-Hadrill emphasized the uneasy union of Bernicia and Deira, and cited Jackson on the etymology. He also gave a phrase (pre-dating Bede, and not mentioned by Ifor Williams or Jackson) from the Latin life of St Wilfrid formerly attributed to Eddius Stephanus, which styles Ecgfrith as rex Deyrorum et Berniciorum.Footnote 21 This is significant. It implies we should think of the forms as referring to a people (Bernicians) rather than a territory (Bernicia).

Patterns set out in previous decades recur in the 1990s and our own time. Wendy Davies observes how western and northern parts of England ‘always remained rather thinly sprinkled with people of Germanic stock’, whose language yet prevailed.Footnote 22 Ken Dark remarks that ‘Bernaccia, as its British name suggests, was an area of more intense British/Anglo-Saxon continuity, and probably had a principally British population in the fifth to seventh centuries’.Footnote 23 Coates and Breeze cite Jackson’s discussion, mention a link with Irish bern, ‘pass’, and discount one with Brigantes.Footnote 24 In a later study, Dark proposes that the ‘larger-scale’ style of Anglo-Saxon kingship in Bernicia and other realms is due to Celtic antecedents.Footnote 25 Maddicott, in similar vein, stresses the economic basis for such power, in terms of profits from land, cattle and slaves, so that Bede often refers to possession of gold.Footnote 26 The new Cambridge dictionary says that the name ‘is usually associated with the Roman British tribal name Brigantes’, but that there are ‘serious phonological difficulties with this derivation’ (do its editors know that Queen Anne’s health is giving cause for alarm?); it cites Jackson’s unattested Primitive Cumbric Bernecc ‘land of the mountain passes’ from unrecorded British Bernaccia, where the first element is cognate with Old Irish bern and Scottish Gaelic bearn, ‘mountain pass’.Footnote 27 Helena Hamerow assures us that the names of some Anglo-Saxon kingdoms derive from those of Romano-British tribes or districts, ‘such as Cantwarena, Bernicia, Deira, and Lindsey. This gave rise to the hypothesis that the earliest kingdoms were based on pre-existing territorial units that were taken over by Anglo-Saxon landlords’.Footnote 28 Bernicia does not, however, figure in a recent study relevant to the problem.Footnote 29 Finally, John Koch’s new atlas shows Bernech and Bernicia between Tyne and Tweed, and comments on the survival of Celtic toponyms in Anglo-Saxon England, though without considering that some Celtic names (like Din Guayroy for Bamburgh) are surely late scholastic inventions and not ‘survivals’ at all.Footnote 30

And so, nearly six decades after Jackson examined the question, we find no progress on the meaning of Bernicia. One can in part see why. There are three objections to the translation ‘dwellers of the land of mountain passes’. Although Jackson thought it a good description of the northern Pennines, it is not such a good description of Tweeddale, which lay nearer Bamburgh (Bernicia’s earliest capital) than does the Tyne Gap. Nor has anyone cited parallels to make it compelling. It is also the sort of description to appeal to a modern scholar, adept at map-reading, but hardly sounds like the name that any Celtic people (especially one famed for prowess in war) would desire. Nevertheless, one can vindicate Jackson’s 1950s proposal by locating his suggestion on Irish bern within Celtic idiom.

Here we turn to the Royal Irish Academy’s dictionary, which shows that early Irish bern meant more than ‘mountain pass’. It was used of other important gaps, in an enemy’s battle-line, for example, or in other spots where an opponent might seize an advantage. So it was a word of great significance for communities in which fighting men had ascendancy. The dictionary’s many quotations make that clear. In the early saga Scéla Mucce Meic Dathó, ‘Tidings of Mac Dathó’s Pig’, the Connacht hero Cet confronts the Ulsterman Celtchair, who ‘went into a gap (i mbernai)’ before him to hurl a spear (which misses, unlike Cet’s, which slices through Celtchair’s genitals and removes him from action).

Another tale, ‘The Battle of Rosnaree’ in the Book of Leinster, mentions near its violent close the bern or ‘breach’ in a battle-position crucial for victory. So powerful was the dangerous space in a line of warriors for victory or defeat, honour or disgrace, that bern churad, ‘champion’s breach’, and bern míled, ‘soldier’s gap’, were set phrases in early narrative. The ‘War of the Gaedhill with the Gaill’, on Brian Bóruma’s defeat of the Vikings at Clontarf in 1014, mentions a bern curad in the Danish ranks. Togail Troi, ‘The Sack of Troy’, the Irish translation of Darius of Phrygia, tells how a bern míled was made amongst the Greeks. Bern báegail, ‘gap of danger’, is another phrase meaning an undefended or weak position, whether in attack or defence. Bern is also used more generally of any situation of risk or combat, or even of Hell itself, which a medieval poet called a bern: a marshy abyss, with stones sharp as razors.Footnote 31 To these Irish instances may be added one from Scotland. Torquil MacLeod succeeded shortly before 1500 as chief of Lewis, in the Western Isles, and he was praised by an unknown poet as a hero equal to Cú Chulainn himself, of valiant and expert hand, fear briste gacha bearna, ‘a man who breaks every breach of battle’.Footnote 32 Since Torquil made a vigorous but unsuccessful attempt to support his nephew’s claim to the Lordship of the Isles, the boast of his rupturing an enemy’s line can have been no empty one.

These many references are telling. What was said of Torquil might be said of many Celtic fighters, that they burst through arrays of foemen and never flinched. Since bern and bearn were words with such resonance in heroic society, it seems they may well explain Bede’s Bernicii, Bernicians, Old Welsh Berneich, Birneich, Bernech and Birnech, and Middle Welsh Bryneich and Byrneich. Jackson’s reconstructed Bernacci or Birnacci could be explained thus. To original bern- would be added a suffix to give Bernaco-, meaning not ‘full of gaps’ but ‘associated with gaps (in battle)’, just as Welsh arfog (where -og is from -aco-) does not mean ‘full of weapons’ but ‘to do with weapons; armed’, and marchog does not mean ‘full of horses’ but ‘associated with horses; rider; knight’.Footnote 33 This unattested Bernaco- (with long a) developed a hypocoristic variant Bernacci or Birnacci (with short a). These would have given unattested sixth-century Primitive Cumbric Bernecc or Birnecc, and so, ultimately, Old English Bernice or Beornice and Bede’s Latinized Bernicii, the Bernicians, with a sense ‘men wont to be found in the breach, those whose place is a gap in the battle-line’.

When, in Henry IV, Falstaff the recruiter asks Feeble the woman’s tailor, ‘Wilt thou make as many holes in an enemy’s battle as thou hast done in a woman’s petticoat?’, he is concerned with Feeble’s boldness in assault. To explain Bernicii as referring to heroes forcing a position thus accords with a perennial concern for those in attack (and defence). It also goes with other warlike terms for ancient leaders and peoples, where Celtic nomenclature tends to the ferocious. This can be shown from surveys of ancient Gaul, Britain and Ireland. With the first may be mentioned Orgetorix, ‘king of murderers’, Vercingetorix, ‘great king of warriors’, Namantobogius, ‘breaker of enemies’, Boudica, ‘victorious one’, or Nettas, ‘champion’.Footnote 34 For the second are the Catuvellauni, ‘battle experts’ of Hertfordshire, and Ordovices, ‘hammer fighters’ of North Wales.Footnote 35 For the third are Irish names in Cath, ‘battle’, Dún, ‘fortress’, Guss, ‘strength’ (including Fergus, ‘manly strength’), or Sluaig, ‘host, army’, some of which must be very old, because we have the form NETTA SLOGI (= Nad Slúaig, ‘champion of the host’) in an ogam inscription. Bern, ‘pass, gap’ itself figures in Irish personal names, supporting the case that it does likewise in a British ethnic one.Footnote 36 In the light of these warlike and even bloodthirsty appellations, there seems no difficulty in seeing the Bernicians as brave combatants, who did not cower, but who were resolute when it came to piercing an enemy battle-line.

Two last points. Even if no exact Brittonic cognate of Middle Irish bern, ‘pass, gap’, is otherwise known, we do have Welsh bâr, ‘anger, fury, passion’, containing the root *bher-, ‘break, split, rend’, found in Latin ferio, ‘I strike, I kill’, and Irish bern itself. It apparently also existed in Ambibarii, ‘furious ones, enraged ones’, a people mentioned by Caesar in the region of what is now Ambrières, Brittany.Footnote 37 The existence of bâr thus strengthens present arguments. Second, if the conclusions proposed here are cogent, they raise the possibility that many other names from early Britain can be explained on lines resembling those followed here. This would have unexpected results. Deira might be related, not to dwfr, ‘water’ (as Sir John Lloyd and Sir Frank Stenton believed), but to Welsh dewr, ‘brave’, and thus meaning ‘land of valiant ones, territory of heroes’. The Cornovii of Cornwall would be, not dwellers on a *Corn- or horn-shaped peninsula, as Dr Padel imagines, but ‘worshippers of a horned deity’ (like the stag deity of the Gundestrup Cauldron in Copenhagen), a point on which Rivet and Smith agreed with Dr Anne Ross.Footnote 38 The Durotrages (not ‘Durotriges’) of Somerset and north Dorset would be ‘warriors of the stronghold’, making forays from a *duro-, ‘fort, walled town’, like Durobrivae, ‘fort by a bridge’ or Rochester, or Duroliponte, ‘fort by a marshy place’, or Cambridge.Footnote 39 In short, analysis of the ethnic names of early Britain promises rapid advances, provided one can address the mentality of ancient fighting men, and avoid that of modern professional scholars (for the most part a docile bunch, attached to their books and studies). On this point one might regret that the ranks of professional Celticists are not recruited solely from those with military training and experience (and preferably having heard shots fired in anger). Were that the case, philology could not fail to benefit.