The MethodThe Method

the caseAberffraw was the chief royal seat of the kings of Gwynedd from at least the ninth century until the thirteenth, and its identification with the tale's wedding site has never been questioned; excavation in 1957 and 1959 on the traditional llys field found only the remains of a later farmstead, so the medieval court's exact footprint is unconfirmed even though the place itself is certain.

the disputeThe 1957–59 excavation on the traditional llys field found only an eighteenth-century farmstead, so the precise footprint of the medieval court is unconfirmed.

the caseAn oval reef roughly 450 metres by 215 metres lies about 900 metres off the coast at Llandwrog, exposed only at the lowest spring tides, and has carried the name Caer Arianrhod in local tradition for as long as it has been recorded, which is why Coflein lists it under that name; RCAHMW aerial survey in 2010 found no convincing trace of walling on the reef itself beyond a possible later fish-weir, so while the toponymic identification is effectively unchallenged, the physical court the text describes is folklore rather than confirmed archaeology.

the disputeRCAHMW's 2010 aerial survey found no convincing structural remains on the reef beyond possible fish-weir lines closer to shore; the tale's palace is not archaeologically demonstrated, only the reef's name.

the caseLocal and antiquarian tradition, followed by Bollard and Griffiths' field survey, places the throne mound at Camp Hill, an Iron Age hillslope enclosure a short walk from Narberth Castle and still visible as an earthwork in a private field; the mound's proximity to the certain site of Arberth itself is the strongest argument for it, but the enclosure has not been excavated to confirm any continuity with the tale's gorsedd, so the precise earthwork intended remains an informed guess rather than a settled fact.

the disputeThe enclosure has never been excavated, so its identity with the tale's mound rests on proximity and tradition rather than confirmed archaeology.

the caseThe Welsh name Ynys Gwales survives directly as the island of Grassholm off the Pembrokeshire coast, and the text's own "Gwales in Penvro" ties it to Penfro, the historic name for Pembrokeshire, which together make Grassholm the dominant scholarly identification; a minority of commentators note that Grassholm has no fresh water and no view of the mainland, both awkward against a text that implies habitability and a door with something to look through, though no rival island has been seriously proposed in its place.

the disputeGrassholm has no fresh water and the mainland is not visible from it, both awkward for the tale's description, though no alternative island has been agreed on in its place.

the caseMedieval and later Welsh tradition fixes the White Mount beneath the Tower of London; the Triads' account of the Three Fortunate Concealments and Three Unfortunate Disclosures goes only so far as hiding Bran's head in the White Hill in London, and the association with the Tower's own mound has been repeated continuously enough, not least through the later folklore of the Tower ravens, to count as the dominant identification; but the Tower itself was built by the Normans centuries after the tale's own setting, and no excavation has confirmed a prehistoric mound beneath it, so the identification remains traditional rather than demonstrated.

the disputeNo excavation has confirmed a prehistoric mound beneath the Tower of London; the identification rests on continuous literary and folk tradition rather than physical evidence.

the caseGuest renders the name Linon; the modern normalised Llinon is generally read as the Liffey, Dublin's An Life, the obvious defensive line for a retreating Irish court once a Welsh fleet has made landfall on the coast facing Wales. Ifor Williams's edition of Pedeir Keinc y Mabinogi treats this as the most economical reading of the tale's own internal geography, and the identification is repeated by Bromwich and by later commentators including Breeze, though it rests on narrative logic and onomastic plausibility rather than a documented medieval place-name equation.

the disputeA minority reading takes Linon for the Shannon instead of the Liffey; the Shannon lies on Ireland's west coast, which sits awkwardly with a fleet crossing the narrow sea from Wales, and this is the chief argument scholars set against it.

rival candidateLlinon

the caseEvans, Nettleship and Perry's study of the Historia Brittonum's marvels identifies Llyn Lliwan with a lost tidal feature on the Caerwent shore of the Severn, the site locally remembered as the Whirlyholes, and reads the ninth-century Mirabilia's account of the pool's dramatic ebb and flow alongside the tale's own placing of it beside Aber Gwy; the location is well argued rather than certain, since the feature itself is no longer active and its precise historical footprint on the modern shoreline is reconstructed from documentary and place-name evidence.

the disputeBromwich and Evans thought the pool might equally have lain on the Gloucestershire shore; the Caerwent-side identification argued by Evans, Nettleship and Perry is not universally accepted.

the caseThe name Rhuddlan Teifi survives directly onto a small hamlet on the Teifi's north bank near Llanwenog, two miles west of Llanybydder, and the identification of this as the court the Fourth Branch means is the standing scholarly view; no excavation has confirmed an actual llys on the ground there, so the precise footprint of Pryderi's hall within the hamlet remains unproven even though the name is secure.

the disputeNo archaeological survey has confirmed a llys site within the hamlet; the identification rests on the onomastic survival rather than excavated remains.

the caseThe territorial name is real and well attested as Seisyllwg, but Ifor Williams and later commentators note that it belongs properly to the eighth-century political union under Seisyll ap Clydog, who flourished around 730, several centuries after Pryderi's own setting, so its use here reads as the redactor's own anachronistic label for the combined lands rather than a name Pryderi's contemporaries would have used.

the disputeIfor Williams treats the name as an anachronism, an eighth-century political term retrojected onto Pryderi's earlier realm.

the caseThe early Stanzas of the Graves, preserved in the Black Book of Carmarthen, locate Pryderi's grave at Aber Gwenoli, a spot within Coed Felenrhyd near the site of the later Maentwrog hydroelectric power station rather than at the parish church; this is an older and textually independent tradition beside the churchyard stone, but the two cannot both be right, and neither has been confirmed by excavation, so the grade for each candidate is possible.

the disputeRival to the St Twrog's churchyard stone tradition; the Stanzas of the Graves and the churchyard identification cannot both mark the same grave, and neither is confirmed archaeologically.

rival candidateMaen Tyriawc

the caseBedd Branwen is a real Bronze Age round cairn, opened when a cremation urn was recovered from a cist in 1813 and properly re-excavated by Frances Lynch in 1967; the cairn yielded eight further urns around a central standing stone, and radiocarbon determinations on material from the cairn ring and the cremations cluster between about 1400 and 1270 BC, many centuries before any plausible historical Branwen. The tale itself specifies only "the banks of the Alaw" and never names this cairn, so its identification with the literary grave is a piece of onomastic folklore attached to a genuine prehistoric monument rather than a location the text pins down itself, which is why the grade here is possible rather than certain or probable.

the disputeThe cairn's radiocarbon dates, roughly 1400 to 1270 BC, are many centuries earlier than any plausible historical setting for the tale, so the name records a folk association rather than a claim the text itself makes.

the caseThe name means simply the fort on the Tywi, and Bromwich and Evans leave it otherwise unfixed; Dinefwr, the Iron Age hillfort later refounded as the seat of the kings of Deheubarth at Llandeilo, is the obvious candidate on etymological grounds alone, since it is the one major fortified site of that description on the river's course, but the tale gives no further detail to confirm it and no editor claims the match as secure.

the disputeThe identification with Dinefwr is an inference from the name's meaning rather than a documented equation, and it remains unconfirmed in the scholarship.

the caseCulhwch names Esgair Oerfel three times as a fixed point of Irish geography, far enough off to anchor a boast about a shout carrying from Cornwall to the north of Britain and to Ireland alike. Sims-Williams reads the second element as reflecting Irish uarbél, a cold gap or cold mouth, and proposes an equation with the townland of Windgates near Delgany in Co. Wicklow, an English rendering of a lost Irish name of similar sense; the wider Irish place name Seiscenn Uarbeoil, a boggy tract in Leinster, has also been linked to the tale's Irish-derived boar-hunt material.

the disputeThe Windgates equation is Sims-Williams's own onomastic proposal rather than a settled consensus; other commentators leave Esgair Oerfel as an unlocated Leinster bog, and the name may work as a rhetorical far corner of Ireland as much as a real fixed point.

the caseA hill called Garth Grugyn survives in the Ceredigion countryside near Llanilar, and the coincidence with the piglet's own name Grugyn Gwallt Ereint has long been noted; but the district was also home to Eli, one of the two huntsmen the text names as pursuing the boar into Ceredigion, in the nearby place name Garth Eli, so the naming pattern may run in more than one direction and the precise hill meant by the tale is not established beyond doubt.

the disputeThe exact position of Garth Grugyn is, in Rhys's own words, still uncertain, and the coincidence of naming with the character Grugyn may be folk etymology rather than evidence.

the caseCornish antiquarians from the early nineteenth century onward identified Celli Wig with Callington, whose Cornish-language name Callywith was read as a corruption of the same root; the theory gave the modern town its Welsh-facing spelling but has been superseded in most recent scholarship by Bromwich's preference for Kelly Rounds near Egloshayle, recorded as the alternate to this entry.

the disputePadel and later commentators regard the Callington identification as etymologically weaker than Kelly Rounds.

rival candidateGelli Wic

the caseRachel Bromwich, editing the Welsh Triads in which Celli Wig appears as one of the Three Tribal Thrones of Britain, matched the name to Kelly Rounds, the Iron Age hillfort at Killibury in the parish of Egloshayle near Wadebridge, on the strength of the surviving field name; the identification is the scholarly favourite but rests on a single place name surviving into modern estate records rather than continuous attestation, and the alternate record below carries the older antiquarian candidate.

the disputeNo consensus exists; Padel himself treats the site's precise location as unrecoverable from the name alone.

rival candidateGelli Wic

the caseRhys read Glyn Ystu as Clyn Ystun, a farm name he placed between Carmarthen and the junction of the Amman with the Llychwr near Drefach, and found support in a 1609 survey of the lordship of Carnwyllion describing land of a similar name on the east side of the Gwendraeth Fawr; the identification rests on a chain of later documentary spellings rather than an unbroken name, so it stays a candidate rather than a certainty, fittingly for the one stretch of the story where the tale itself says the trail was lost.

the disputeThe 1609 survey evidence is suggestive rather than conclusive, and no modern place bears the name directly.

the caseA holed stone answering the text's description, roughly a metre by three-quarters of a metre with a hole about an inch across, was recorded on the Cynfal's bank by Frank Ward in 1934 and a stone believed to be the same, or a successor to it, was rediscovered around the turn of this century on a tributary now called Afon Bryn Saeth, where it stands on a farm of the same name; Patrick Ford's scholarly translation notes the tradition, but the stone's own findspot has shifted twice within living memory, so the identification is graded possible rather than probable.

the disputeThe stone identified as Llech Ronw has been reported at two different findspots within living memory, Ward's 1934 location on the Cynfal itself and a later rediscovery on the Bryn Saeth tributary, which weakens any claim to continuous local memory of a single stone.

the caseThe Shannon is Ireland's longest river and the obvious candidate for a storyteller's idea of Ireland's great water barrier, and this reading survives as an alternate in general and popular retellings of the tale. Against it stands the crossing itself: a fleet sailing the short sea from Aberffraw or Caer Seiont would make landfall on Ireland's east coast, nowhere near the Shannon's west-coast estuary, which is why specialist commentary on the Second Branch's Irish geography prefers the Liffey.

the disputeThe dominant scholarly reading favours the Liffey; the Shannon persists chiefly in general retellings rather than in specialist onomastics.

rival candidateLlinon

the caseRhys proposed that Llwch Ewin survived as a bog mere he found marked above a farmhouse called Llwch on the southern slope of the Amman valley near Bettws, though the map he consulted named the same feature Llwch is Awel, pool below the breeze, which weakens the onomastic link considerably; Bromwich and Evans call the evidence thin and prefer other candidates, recorded as the alternate to this entry.

the disputeRhys's own source names the pool differently from the reading he builds on, and Bromwich and Evans regard the identification as unconvincing.

rival candidateLlwch Ewin

the caseBromwich and Evans set out several candidates for Llwch Ewin without preferring one: Llangadog castle, previously recorded as Castell Luchewein; Llyn Llech Owain on the ridge above Gorslas; the Black Mountain generally; or a point west of the mouth of the Wye. Llyn Llech Owain is the strongest of these on onomastic grounds, since Ewin and Owain are close enough in form to have been confused in transmission, but the case remains circumstantial and the alternate record below carries the field's other leading guess.

the disputeRhys instead proposed a lost bog pool above a farm called Llwch in the parish of Bettws in the Amman valley, a reading Bromwich and Evans call thin; see the alternate record.

rival candidateLlwch Ewin

the caseMatholwch names the feature precisely in his own account to Bendigeidfran, the mound at the head of the lake which is called the Lake of the Cauldron, but gives no further Irish geography to fix it. Scholars have nonetheless hunted the name through medieval Irish sources, and Sims-Williams finds only two candidates: a Loch Cori in an eleventh-century poem on the lakes of Slieve Aughty in Co. Galway, possibly a scribal error for Loch nGaire, and a Loch nGualai at Devenish on Lower Lough Erne, which Whitley Stokes equated with Branwen's Lake of the Cauldron and which Sims-Williams calls the better of the two. The Book of Leinster sagas hide Conchobar's vast vat, the Iarnguala or Iron Abyss, in that lake and derive the lake's name from it, which supplies the Welsh tale with exactly what it needs: an eponymous vessel beneath an Irish water.

the disputeSims-Williams rejects Stokes's etymology, deriving Loch nGualai from Middle Irish guala 'pit, abyss' rather than any word for a cauldron, so the equation rests on the sunken vessel and not on the name itself; the rival Loch Cori of Slieve Aughty is the other candidate and the whole episode may be aetiology for the cauldron of rebirth rather than located geography.

the caseA sandstone pillar stands beside the south porch of St Twrog's church at Maentwrog, and local tradition names it Pryderi's grave-marker; but the manuscript reading is Maen Tyuyawc, which Lady Charlotte Guest herself emended to Maen Tyriawc specifically so that it would resemble Maentwrog, a piece of editorial harmonising rather than an independent scholarly identification, so the stone's link to the tale is graded possible and set against the rival tradition recorded at Aber Gwenoli.

the disputeGuest's own emendation of the manuscript's Maen Tyuyawc to Maen Tyriawc was made in order to produce a resemblance to Maentwrog; the identification is partly an artefact of her translation rather than a name independently attested.

rival candidateAber Gwenoli

the caseRCAHMW's own account of Pryderi's traditional itinerary suggests that Maenor Penardd is the township of that name adjoining Dinas Dinlle, near Caernarfon, which would place the battle and the swineherd's house on the coastal plain of Arfon; the suggestion is offered as plausible rather than proven, and no excavation ties any surviving earthwork there to the tale, so the grade given is possible.

the disputeRCAHMW's own phrasing is that the identification 'has been suggested', not confirmed; no excavated feature at Penardd township is tied to the battle by independent evidence.

the caseThe retreat the text describes runs from the coastal plain near Dinas Dinlle south-east toward Dolbenmaen, and RCAHMW's account of the episode places the fighting along an old trackway on that line; a stream called Nant Cyll north of Porthmadog has been proposed as the modern survival of the name, but the identification depends on a name-change too small to be certain and no independent record confirms it, so the grade is possible.

the disputeThe proposed match with Nant Cyll depends on a minor sound-change from Call and is not corroborated by an independent source; RCAHMW itself gives only the general line of retreat, not a named stream.

the casePelumyawc is read as Peuliniog, a lost medieval commote that Rhys places between Llanddewi Velfrey, Henllan Amgoed and Laugharne, roughly the parishes of Whitland and Llandysilio; the commote has left no surviving settlement of that name, so the location is Rhys's inference from later administrative boundaries rather than a fixed point, and Bromwich and Evans do not improve on it.

the disputeRhys's reconstruction of Peuliniog's bounds is itself provisional, built from later manorial and parish evidence rather than any securely dated medieval source.

the caseCaer Dathyl is named seven times across the Fourth Branch as Math's fixed court in Arfon, yet no single site commands scholarly agreement: nineteenth-century antiquarian tradition pointed to Pen y Gaer above Llanbedr-y-Cennin in the Conwy valley, well outside Arfon proper, while later writers have proposed Tre'r Ceiri, Caer Engan, Caernarfon itself, Pen Dinas, Craig y Dinas and Y Foel without any one candidate gaining consensus; Ifor Williams offers only the etymology of Dathyl, from Dathal or Tathal, an adaptation of the Irish Tuathal, and stops short of endorsing a site. The atlas records the name as the text's own true north for the Fourth Branch's geography without pretending to a grid reference no scholar can agree on.

the disputeProposed sites include Pen y Gaer (Llanbedr-y-Cennin), Tre'r Ceiri, Caer Engan, Caernarfon, Pen Dinas, Craig y Dinas and Y Foel; none commands scholarly consensus and RCAHMW's own survey record ties none of them to the tale by name.

the caseMatholwch's court is never named in the tale: Branwen simply travels to Ireland and is later degraded within an unnamed household. Tara, the traditional seat of the kings of Ireland, is the image a medieval Welsh audience would most readily summon for a king of Ireland's court and it is sometimes invoked loosely in that spirit, but no scholar identifies it with Matholwch's court, and Sims-Williams's chapter on the Irish geography of Branwen treats only the Lli and the Archan, the Liffey and Dublin and the origin of the five fifths of Ireland. With no name in the text and no candidate in the literature the court is recorded as lost rather than pinned to a real monument.

the disputeTara is a conjecture built on what a medieval Welsh audience would assume a king of Ireland's court to be, not on anything the text names; several commentators leave Matholwch's court unlocated altogether.

the caseThe tale recalls a time when only two rivers, the Lli and the Archan, kept Britain from Ireland, and has Bendigeidfran's fleet come to shoal water at their meeting. Discussion of the passage has placed the Lli on the Irish side and floated Loch Laoigh, the old Irish name surviving in Belfast Lough, as its referent, paired with the Archan on the Welsh side, itself disputed between the Arth and the Ystwyth; the whole passage reads as remembered cosmogony, a shrinking sea, as much as sailing directions, and no candidate commands enough agreement to plot with confidence.

the disputeSome read the Lli as a doublet of the Linon itself rather than a separate river, given the tale's habit of naming the same crossing twice; even readers who keep it separate disagree on whether Loch Laoigh is the right equation, and Belfast Lough in any case lies well north of the region this atlas otherwise maps.

the caseThe text names Maenor Coed Alun only once, paired with Maenor Penardd as the ground of the first battle; RCAHMW's own summary of the episode renders the name slightly differently, as Maenor Pen Alun, without proposing a location, and no coed, wood, of that name is confirmed on the ground near Caernarfon today, so the second of the two battle-maenors is recorded here without coordinates.

the disputeRCAHMW's summary of the episode gives the name as 'Maenor Pen Alun' rather than 'Maenor Coed Alun', and proposes no location for either form.