An immensely detailed and thorough study of seven centuries of British kingdoms and their neighbours

Not the easiest book to score, this is the first part of a history of Wales and, for many people, it will be a specialist reference text. I got a huge amount from it (hence the 5*) but that does not mean to say there are not difficult, highly academic sections. It is also costly but will take you at least twice the time it would to get through your average read. Furthermore, you don’t have to be Welsh; there is everything here: the development of England, the kingdoms before Scotland, Ireland’s influence on the western shores, the Isle of Man, even the Hiberno-Norse.

Early chapters present a survey of ‘post-Roman’ Britain: from the Manaw border between Gododdin and the Picts through Rheged, North Wales – where the Ordovices are replaced by Gwynedd, ultimately of Irish origin, Powys which is developing as the ‘paganes’ opposition to urban-centred Cernyw / Cornouui, South Wales where the Demetae had become Dyfed, through to the territories of the Southwest: the relatively ‘untouched’ lands of the Durotriges (Dorset) and the Dumnonii (Devon / Cornwall).

The section on Brittany was difficult thanks to unfamiliarity and its repetition of British names – Kernev, Dumnonia, Gwenod. In retrospect, I see how important it was in the context of the geospatial linguistic analysis which follows. The first attestation of a sizeable British presence in Armorica belongs to the 460s, the time of Riothamus, and settlement would have been with Imperial authority. The lands of the Veneti were probably gradually taken by force over the sixth century. Some settlement between the Bretons and the Franks was likely but it cannot be dated to 497, although relations were peaceful until Childebert I’s death in 558. Intriguingly, whilst Latin appears to have been a normal means of communication in Britain, in the lands west of the Vilaine in Brittany, the dominant language appears to have been British.

One of the central arguments of the early chapters is that British led some linguistic changes. For much of the island, monolingualism was a complete myth – at least in the fifth and sixth centuries. Many areas which had been most Romanised were already dominated by Germanic tribes, but Latin had been the language of the occupation and, under threat, western and northern Britain clung to it. Latin was far from static though. Just as it set out to become French, Sard or Catalan with less complex grammar, something similar was underway in Britain. There was a second affinity here: Irish.

I got through the next section relatively unscathed by complexities but the linguistics here are technical, the explanations somewhat limited. The road from Brythonic to Welsh saw several major shifts including partial assimilation of single, intervocalic consonants to vowels. In Insular Celtic this happened not just within words but across word boundaries. Welsh also experienced a ‘North Sea group’ development: the loss of the final syllable or apocope. This is common to the development of the Germanic languages – the word for ‘man’ lost its final syllable (‘mannaz’) just as Celtic, ‘wiros’ did. Lenition (soft mutation) came to replace the grammatical function that case inflection had in Brythonic. This is unskippable because loss of the genitive case inflection is central for the inscriptions section. But case inflection and final syllable losses were independent of one another: Irish retained a rich declension system but underwent apocope. The most important change in Welsh, the shift in syllable stress, occurred in Cornish and Breton at much the same time and as late as the ninth century. The establishment of Hwicce circa 600 separated land access between the Southwest and Wales – but contact continued by sea, usually the fastest means of communication.

A significant section of the book is about inscriptions, which although usually brief, offer crucial evidence. In the early period, inscriptions were too few in number to support specialisation. One issue not discussed is that, if the ogham alphabet was ‘specifically designed’ for Primitive Irish, why should it have contained letters which were never used? For the Isle of Man, there was a change in ethnic attribution between Orosius and Bede – even though the latter was familiar with the former. Between 400 and Viking times there were evidently both British and Irish with dominance determined by Irish Sea balances of power. There is also a clear distribution of monolingual and bilingual inscriptions on the island. Epigraphically, Gwynedd is similar to Scotland whilst Dyfed is similar to the Isle of Man. Only in Brycheiniog with its Irish influence do we find a region well-stocked with inscriptions close to England’s border. Across a broad area there was a general shift from an epigraphy derived from late Roman Britain to a more ecclesiastical one, tied to book scripts and using a Latin learned in schools. In Northern Britain this development was cut short around 700 by the expansion of Northumbria.

The collapse of a treaty with the Irish in 360 led to a series of raids culminating in the ‘barbarian conspiracy’ (367). Later documentation suggests that there were three waves: from Leinster, Ulster and the Féni, who acquired an outlet to the sea around 500. Accepting the Féni hypothesis means rejecting Gwynedd’s establishment through Cunedda’s migration from Manaw Gododdin. We know that Maelgwyn’s reign overlapped with Ida of Bermicia’s in Northumbria, which commenced around 547, implying that Maelgwyn’s great-grandfather might have arrived circa 400. That would imply Irish settlement stopping shortly after. But it did not. The Penbryn inscription (in Demetaen territory) implies that the Ordovices lasted through much of the fifth century and that their number included an Irish component.

Challenges to received wisdom continue with the St Patrick story pitted against the forgotten legacy of Palladius and the overthrowing of the myth of an integrated Celtic Christianity. British and Irish Churches probably maintained contact through personal relationships. Southern Britain circa 500 was much like Eastern Europe in 1900 with intermingled ethnicities. By 700 that had changed. Amongst earlier assimilations (circa 600) was the integration of the Irish south of the Clyde which heralded a change in attitudes towards language: the British language became a defining mark of identity.

When Rome left, material culture in Britain deteriorated sharply even though the west had been less than thoroughly Romanised. Britons were worse off than they had been four centuries earlier. Nevertheless, at a limited number of concentrations of political power – Tintagel, Dinas Powys and Garranes (County Cork and possibly in Éoganacht Raithlinn) – there was a rich accumulation of pottery from the Eastern Mediterranean over 475 to 550. This aligns with Gildas’ characterisation of that period as one of recovery and subsequent peace. But shipments post-550 were from the Atlantic seaboard and, increasingly, the bulk of the population was excluded from an elite material culture, most Britons ceasing to use ceramics.

There then follows a section on laws and charters, focussing on the Book of Llandaff, which I found heavy going alongside the next section on kin relationships. Part of the section on the laws of Hywel Dda might have been easier after reading about Hywel himself. But the kin relationships section is important, remembering that kinship is next to kingship in the English language – to which Charles-Edwards progresses.

North of Hadrian’s Wall a kingly tradition had survived the Roman period even though some between there and the Antonine Wall had sporadically come under Roman rule. Further south, in the fifth century, Britain acquired new kings. Some British kingdoms were on a small scale – Dumnonia, Demetia, Ventia (Gwent). Vortigern presumably ruled a wider area but we have no documentary evidence and his name survives in Welsh form in a tiny kingdom: Gwerthrynion. By the ninth century in Wales some kingdoms were even smaller. By then Brycheiniog was then quite separate from any Silures successor state, Ventia was divided into the precursors of Glamorgan & Gwent, a Romano-British vicus, ‘Ariconium’, had become Ergyng and there were assorted other small territories such as Buellt (Builth), whilst Rhôs in Gwynedd’s sphere appears to have had its own dynasty.

British kingship harked back to a recent Roman past and a deeper, Irish inheritance. But, following the advent of Christianity, there was a further influence from the Old Testament. Different words were used for the concept of ‘king’: arbennig (ruler of a significant kingdom); rhi (poetic or minor), gwlaedig (‘lord’, related to the modern Welsh for ‘country’) – and the new arrival, brenin, perhaps originally connected with the concept of ‘freeman’ or even with the sovereignty deity, Brigantia, more often tied to the Brigantes but also present in the river name, Brent.

Æþelberht of Kent’s role in the Christianisation of Anglo-Saxon England changed perceptions, Rome switching its efforts to Germanicised parts of the island, whilst the British became characterised as heretics. In the eighth century Wales’ relationship with the Angles becomes clouded by the inner and outer zones surrounding ‘Mercia Proper’ (not the original heartland of Mercia, it should be added). Two fantastic material survivals from the period are brought in to explore this: the Pillar of Eliseg (Elisedd / Elise) at Llantisilio-yn-Ial, and Offa’s Dyke.

Patron of the ‘pillar’ was Cyngen ap Cadwell, King of Powys from around 808 until his death in 854. It celebrated the achievements of his great-grandfather, Elisedd ap Gwylog. Although Elisedd is never mentioned in the annals, it is claimed that he recovered Powys from Mercia in 757, the year Æþelbald was murdered by Beornred (prior to his ousting by Offa). The text – no longer so legible – is reliant on Edward Lhuyd’s translation in the late seventeenth century and may be compromised by two separate recoveries – one under Elisedd. However, it is clear that in 750 there was a rebellion of southern English against Æþelbald and Powys may have exploited this state of affairs.

Asser, writing from the Court of Wessex but originally from southwest Wales, claims that Offa built a defensive dyke from ‘sea to sea’, echoing earlier descriptions of Hadrian’s Wall. Modern archaeology suggests that this is a little distant from the truth: there was probably already Wat’s Dyke (which does not neatly link-up and which has been the subject of recent dating controversies) whilst, in the south, the new construction did not reach the Severn Estuary. But it was still some kind of frontier, if only one delimiting lands from which a Mercian king might not raise taxes. Right up against Wales, Shropshire has less British names than Staffordshire and this may reflect the importance of fortifying the frontier zone. Shropshire seems to have experienced the greatest intensity of Anglicisation, declining both to the north and – more rapidly – to the south.

There is a tenth century document concerning the small Dunsætan area, bordering on Gwent (which had, apparently, ‘at one time’ belonged to the Dunsæte) and which seems to have included lands on both sides the Wye. North of the Monnow lay Ergyng / Archenfield – which even Domesday indicates had some Welsh population. In Dunsætan, the Welsh on the other bank of the Wye appear to have been in a tributary relationship. The boundary on the Wye had not rendered Dunsætan obsolete and it is clear that interactions continued across the river.

There are two (credible but incompatible) interpretations. One denies the existence of settled Mercian overlordship in Wales between 679 and 796 (i.e. Trent to Offa’s death). By the ninth century, there was far more extensive subjugation, the frontier having become sharply defined in cultural & linguistic terms. This would attribute much to the military conquests of Cenwulf in Wales from 816 to 822.

The alternative views Mercian hegemony in Wales as lineal descendent of Penda’s control of southern England. In this scenario, the Pillar of Eliseg refers specifically to Powys after a period of direct Mercian rule. Cenwulf’s aggression would have commenced immediately after succession, slaying Caradog ap Meirion of Gwynedd shortly after the beginning of his reign, later campaigns being merely an attempt to defend an authority under threat.

In the late seventh century, Wessex still faced Dumnonia across its western frontier. In 722 the Annales Cambriae records the death of Beli of Alt Clud and 3 battles against the ‘Cornish’: Hehil. Garth Maelog and Pencon (locations all now difficult to identify). Independent confirmation regarding Beli exists in Irish annals. 722 is only 12 years after the ASC records Ine (with the help of his South Saxon ally) against Geraint, King of Dumnonia. So, by 722 ‘Defna scir’ had fallen and British victory secured the survival of Cornish independence.

Around 830 an anonymous Welsh scholar wrote ‘Historia Brittonum’. The history of the Roman and post-Roman periods is dealt with strangely, Ambrosius appearing a Christ-like figure, raised in Glamorgan from a virgin birth. It also conflicts with the Pillar of Eliseg regarding the genealogy of Powys’ kings, implying interchange with Gwerthrynion.

Following disputes between Rhôs-based Hywel and Cynan of Gwynedd, the ninth century saw a new dynasty in North Wales: the Merfynion. Relationships you might find through a simple Google search are different to those presented here. Hywel’s reign had been a time of substantial territorial loss to Mercia’s benefit. Over the course of a century the Merfynion would come to dominate much of Wales. Both Merfyn Frych (‘the Freckled’) and his son, Rhodri Mawr, seem to have ruled both Man and North Wales. But Merfyn took Gwynedd under obscure circumstances. It is possible that he was part of a great move southwards driven by the Vikings who attacked Ulster in 825, Brega & eastern Leinster in 827. But this is not Charles-Edwards’ default.

More likely is that North Wales factions associated with diverse Mercian parties. Offa seems to have found Caradog acceptable but the latter was slaughtered under Cenwulf. After Cenwulf’s death, attacks resumed – including Deganwy’s destruction in 822. With Ceolwulf deposed in 823, internal Mercian disputes handed Wessex’s Ecgberht an historic opportunity.

Over the period from Æþelwulf’s death in 858 to Northumbria’s incorporation in 958, England shifted from division to single, unified administration. In 850 there were still four separate kingdoms and there continued to be instability across the Central Belt of Scotland and the Thames Valley. In 870 two Viking kings based in Ireland, Olaf and Ivar, besieged Alt Clud and captured Dunbarton. When Arthal died in 872 at the instigation of Constantine, he had a new title: ‘King of Strathclyde’, implying that the British had been unable to rebuild Dunbarton. Whereas Alt Clud had looked north to Lomond, the new capital, Govan, had Renfrewshire as hinterland. Strathclyde enjoyed significant territorial expansion in the late ninth century and by 927 it extended as far as Penrith. There is no positive evidence to suggest an alliance between Strathclyde and the Vikings before 920. However, for Ragnall in York, it may have seemed politic to permit expansion.

As early as 841 the Vikings had transformed Dublin into a major fortress and port, winning important battles by 848 whilst their other major target, Frisia, experienced increased stability under Charles the Bald. The new target was to be Britain. Between 865 and 878 the Great Heathen Army conquered Northumbria, East Anglia, much of Mercia – and very nearly Wessex. What remained of Mercia, no longer Wessex’s equal partner, became a satellite of its southern neighbour. But how exactly did this happen and what role did Wales play?

Ceolwulf is seen as Viking client king. In the 877 partition the Vikings left him with the territories bordering Wales. That same year on Anglesey, Rhodri was defeated by the ‘Dark Foreigners’ and fled to Ireland. Edington (Ethandun) took place in 878 and that same year Rhodri was killed by the ‘English’. It is inconceivable that Alfred was responsible but Ceolwulf may have had a ‘free hand’. Relations between Mercia and Gwynedd are betrayed as nothing less than a blood-feud. However, over 881 to 893 Mercia went into collapse mode. Rhodri’s sons defeated it at the Battle of Conwy. Æþelred submitted to Alfred and the latter sought to replace him as overlord in southeast Wales as Æþelred’s dominance slipped away even amongst the Hwicce (which had been Mercian since Penda) and the Magonsæte.

Charles-Edwards argues Glywysing and Gwent’s submission preceded Mercia’s, citing a chronology: Edington (878), Conwy (881), Alfred encouraging Hywel ap Rhys, Brochfael ap Meurig and Ffernfael to transfer allegiance (881/882), Æþelred’s acceptance of Wessex’s overlordship (882 or 883) and the Watling Street boundary treaty with Guthrum no earlier than 883. Dyfed and Brycheiniog were threatened by expansionist Gwynedd and therefore submitted whilst Anarawd himself most likely submitted between 888 and 892. Powys had probably already been incorporated into Gwynedd by 886.

Threatened by Viking forces from the late 880s to 896, Brycheiniog built a unique monument in Wales: a crannog on Llangorse Lake. ‘Brecanmere’ in English, it would be burned down by the Mercians within thirty years. In 893 Cadell ap Rhodri replaced the sons of Hyfaidd as ruler of Dyfed. Southeast Wales remained under West Saxon overlordship, whilst most of Wales returned to Mercia’s sphere.

In times of strength Dyfed expanded into the Tywi Valley but it may have lost this to the petty kingdom of Ceredigion in the late seventh century. From the eighth to eleventh centuries it was heavily subjected to Viking raiding compared to most of Wales. In 904 Llywarch ap Hyfaidd died, leaving his daughter as heiress. Core Dyfed then became subject to a takeover by neighbouring Seisyllwg (Ceredigion & the Tywi) under Hywel the Good, grandson of Rhodri. It is not discussed but Hywel may not actually have controlled all Seisyllwg. It is also possible he inherited Dyfed from his father, Cadell ap Rhodri, implying acquisition sometime between 894 and 909. In 894 Anarawd and the ‘Anglii’ (presumably Mercia) laid waste to Seisyllwg, although not ‘core Dyfed’. Meanwhile, Alfred’s attentions were on the Vikings.

Parker ASC implies that the Welsh kingdoms which formed an alliance with Edward the Elder were those formerly subject to Mercia, having transferred their allegiances after Edward’s coup against Ælfwynn in December 918. Whilst Parker presents a unified group, the truth may have been rather different. By 927 Chronicle D names ‘Uwen’ (Owain?) king of Gwent and there is no mention of the kings of Glamorgan. The Mercian register indicates that Brycheiniog possessed its own king under Lady Æþelflaed, but in retribution for the murder of Mercian abbot, Ecgbryht, she had Llangorse crannog burned. So, at least two kingdoms in the southeast were not Merfynion-controlled. There was probably an alliance between Æþelred and the Merfynion in exchange for overlordship – at the expense of both Dyfed and Wessex (in 893 all of Wales had been under West Saxon control). Asser seemed unaware of Hyfaidd’s death (despite originally being from Dyfed), perhaps because it did not take place until fractionally later than traditionally chronicled.

Merfynion aspirations in Dyfed were not fulfilled until Edward the Elder’s reign. When Llywarch ap Hyfaidd died in 903 and Rhodri was decapitated in 904, Edward had his own concerns with his cousin, Æþelwald. In 924 Edward died at Farndon-on-Dee. He may have planned that Ælfweard was to take care of Wessex whilst his other son, Æþelstan, was for Mercia, perhaps having been fostered by the royal court there. This may even have been the logic behind Edward’s coup against Ælfwynn. Regardless, it was not to be. Ælfweard died in Oxford almost immediately.

In 927 Æþelstan acquired the Kingdom of York from Sihtric / Sigtryggr despite an attempt by Guthfrith / Gofraid ua Ímair to succeed. Hywel and Morgan ap Owen (Gwent’s king) probably went north with Æþelstan. But different sources suggest a varying story for Eamont Bridge – or, as it might actually have been, Dacre – if William of Malmesbury is to be believed. Had the Cumbrians given Gofraid some support?

It was only at this point with Gofraid removed that Æþelstan was in a position to get Owain of Cumbria (just beyond the border) and Constantine, the Scottish king, to come together in his presence. There is much said about ‘idolatry’ but both Owain and Constantine were Christians. The presumption must be that Gofraid was not and that offering support should be considered idolatrous. Gofraid died of a painful illness in 934 and his son, Olaf, became the leading figure amongst the Uí Ímair. Æþelstan led an army as far as Dunnottar, even ravaging Caithness. The Scottish king was forced to give his son as hostage and attend celebrations of his defeat. It may be that Constantine had previously switched support back to the Uí Ímair, thereby violating Eamont.

Olaf’s campaign in 937 led to defeat at Brunanburh (not Bamburgh; probably in the Wirral). On both occasions when Olaf had led combined armies against Wessex, Hywel had remained faithful to the English. In the last round of this battle between the Uí Ímair and Wessex, Edmund conquered the Kingdom of York and ravaged Cumbria, giving the latter to Malcolm of Scotland in exchange for overlordship. In Wales only Idwal Foel had allied with Olaf. Anarawd, Idwal’s father, had been allied to the Viking ruler of York until around 890. ‘Armes Prydein’ was written in this context, foretelling a time when the Welsh would lead an alliance against the English and drive them from the island.

The poem is awkward to date. Olaf’s allies and the Armes alliance are only a partial match. Olaf’s were Constantine (Alba), Owain (Cumbria) and Gebeachan (presumably, the Hebrides). In Armes, there are two groups. Firstly, the ‘cymod’ (people to whom the Welsh have been reconciled): Vikings and Gaels. However, some Gaels are described as ‘of Anglesey’ – problematic because it was British. That might imply a later composition date but the fact that the Irish Vikings wouldn’t follow St David’s flag might imply that they were still heathen at the time. The wording might just imply Gaelic-speaking?

The second group is the ‘cludwys’ (‘included amongst us’) with an inheritance right to Britain: Cornish, Cumbrians and (via the revenant leader, Cynan), Brittany. Note the etymological connection with the River Clyde (Clud). The mismatch comes in this group as neither Bretons nor Cornish came out against Æþelstan, only Cumbrians from Strathclyde. In Brittany’s case it would have been an unlikely scenario as Alan was a former guest at Æþelstan’s court. In 939/40 the allies might well not have included Scotland either. Constantine was approaching abdication and not participating may help elucidate why Cumbria was handed over to the new king, Malcolm (Máel Coluim in Gaelic), by Æþelstan.

Talk of ‘Men of Dublin’ suggests that Armes is post-841 and the establishment of the first longphort (and before the second half of the tenth century) whilst the fact that the tribute reeves of Cirencester fled to Winchester suggests that this is not Burgred’s Mercia in 853. Furthermore, the tribute regime does not smack of Alfred and single English overlordship. Charles-Edwards suggests that the composition date is between 927 and 942. Idwal’s death in the latter meant there was no prospect of a co-ordinated uprising and the poem is serious political propaganda to that end. As to where it was written, the propaganda target and writer’s domicile residence should not be confused. It is possible the target was Dyfed and Glywysing whilst the poet wrote from Gwynedd – and another poem with the same opening lines was definitely from the north.

Hywel’s death (as leading king since 916) in 950 represents a turning point, broadly coinciding with weakening West Saxon cohesion. I grew up with the notion that Edgar was ‘first king of all England’. In fact, in 957 he became simply king of Mercia, not taking on Wessex until Eadwig’s death (959). During the gap, most of Wales probably came under Edgar but the southeast was probably under Eadwig. Appointed by Eadwig in 956, Ælfhere was probably the most active Englishman in Wales. Although initially closely connected with Eadwig, his authority remained unharmed under Edgar.

Hywel’s son, Owain, succeeded him without opposition in Dyfed but he did not retain power in Gwynedd. There seems to have been a major battle at Carno (in the cantref of Arwstli) between Hywel’s sons and Idwal’s (Iago & Ieuaf). There was also conflict along a fault line extending from Gŵyr east as far as Cowbridge as Gower shifted back and forth between Dyfed and Gwent.

In 961 the ‘sons of Olaf’ sacked Caergybi / Holyhead, ravaging Lleyn and there were further Anglesey attacks from other Uí Imair branches. It is a confused picture, particularly with regard to Olaf’s identity: Olaf Guthfrithsson or Olaf Cuarán, king of Dublin until 980?

Whilst Cornwall’s religious landscape continued to be quite different from Devon’s, it was now an English county. Brittany had been conquered by the Franks and yet, during the decline of the Carolingian Empire, there was a large measure of independence – even expansion, meaning that alongside Breton-speaking, there were considerable Romance-speaking areas.

During the late ninth and tenth centuries Cumbria expanded too, taking over former Northumbrian territories. But two developments weakened Cumbria in the mid-eleventh century. The first was the establishment of a separate entity: the Rhinns in the southwest of modern Galloway. Whether the Gallgaedil were the dominant element in Galloway society in the early eleventh century we cannot tell, but the name means ‘foreign Irish’ and is the precursor of the modern name. Gaelic survived here until the early modern period and they were considered separate from the Gaelic Scots (perhaps also in their laws), holding strong links with Man, Dublin and the Hebrides. Dumfriesshire and coastal Cumbrian names have stronger Viking elements than the Rhinns. But King Suibne’s ‘Galloway’ may not have been in the Rhinns. Echmarcach, who was to die as ‘King of the Rhinns’ retired to Galloway and the people became attached to a region. However, topographically the name was first associated with Bute and Cowal – i.e. further north than modern Galloway, in the Firth of Clyde. Only later were the Rhinns added to an expanding Galloway (perhaps in the twelfth century). The second change was that Siward, Earl of Northumbria, gained control of all Cumbria.

Moving beyond the Millennium, the Welsh picture becomes more confused. When Maredudd ab Owain died in 999, Cynan ap Hywel of the northern Merfynion took the kingship of Dyfed but he died in 1003 and there is no record of his successor. Edwin ab Einion had ravaged Maredudd’s Deheubarth (literally, the ‘right-hand part’) in 992 and he was still alive at the beginning of Cnut’s reign. Edwin had hoped to succeed to Dyfed but there is no evidence that he did.

Gwynedd is only marginally less obscure. One possible ruler is Aeddan who had his power base west of the Conwy whilst there is a hint Llywelyn ap Seisyll was in power east of the river. One ‘Eilaf’ (probably Eilifr Thorgilsson) came into Wales and laid waste to Dyfed and St David’s. He was a major figure in Gloucestershire under Cnut, hinting that there may have been some alliance between Cnut and the Irish. Wales remained a concern to the English but beyond their capacities.

Rhydderch ab Iestin heralded another new dynasty in the south whose origin is unclear from the chronicles. In 1027 Cynan ap Seisyll was killed but the chronicles do not indicate by whom. The Book of Llandaff says that Rhydderch ab Iestin succeeded Cynan as king of Gwynedd (although Iago ab Idwal of the northern Merfynion held Anglesey). This does not mean that Rhydderch was the killer since in 1030 the English and Dubliners attacked Wales.

After Rhydderch ab Iestin’s slaughter by the Irish in 1033, Iago gained Gwynedd (presumably expanding from Anglesey) and Edwin ab Einion’s sons, Hywel and Maredudd, gained the southern kingdom. But there was further fragmentation in the Southeast with different kings (both grandchildren of Rhydderch ab Iestin) in Gwent uwch Coed / Ystrad Yw (Caradog ap Gruffudd) and Gwent is Coed / Ewyas (Rhydderch ap Caradog). This Rhydderch is likely to have taken Brycheiniog and thus established a position barring the two doorways into Wales from Herefordshire.

Cumbria and Brittany were – in territorial terms – British successes of the ninth to eleventh centuries. But those successes encouraged cultural dilution. In contrast, Wales was more distinctly ‘British’: less Welsh territory had been taken by the Vikings and there had been no expansion into English territories. In Northern Britain the ninth to twelfth centuries would prove the graveyard of two British languages: Pictish and Cumbric.

Initially, two of the final three chapters about the organisation of the Church and Latin education really feel as though they belong somewhere else. But the importance of Latin learning emerges in the final chapter on poetry when there is evidence of rivalry. Two bardic praise poems from the Book of Taliesin are examined: Edmyg Dinbych (Dinbych being Tenby) and Echrys Ynys (‘Desolate the island’). Charles-Edwards dates the first to be from between 814 and 870 when Longbury Bank’s role was at an end but Penally and Caldey Island were still active – but before the 870 dynastic switch to the Merfynion. Triffun’s cousin, Tangwystl, married Bleddri, father of Hyfaidd of Dyfed. The names Bleddri and Bleiddudd (in the poem) have the same first element, ‘wolf’ followed by a different name for ‘ruler’.

Echrys Ynys may also be datable if Aeddan is the one killed in 1017 by Llywelyn. Finally, in total contrast, there is the dialogue between Llywarch and Gwên. This brings the whole work to a moving and universal end, revealing truths about humanity alongside their equivalents regarding the natural world.