The Battle of Gwen Ystrat

Six years ago, in a post at this blog, I discussed the battles of Catraeth and Gwen Ystrat. The former is mentioned in a medieval Welsh poem (or group of poems) known as Y Gododdin (‘The Gododdin’) where it is presented as a defeat for the Britons of Edinburgh at the hands of Anglo-Saxon foes. Gwen Ystrat was the scene of a victory for the North British king Urien Rheged, apparently over Pictish enemies, in a poem attributed to the bard Taliesin. The poem has the title Gueith Gwen Ystrat (Welsh gueith = ‘strife, battle’) and its historical setting is the sixth century AD.

In my blogpost from 2009, I referred to a suggestion by the Celtic scholar John Koch that the battles of Catraeth and Gwen Ystrat were one and the same. I also mentioned an objection to this idea by Graham Isaac, who proposed that Gueith Gwen Ystrat should be seen not as an authentic poem from sixth-century North Britain but as an ‘antiquarian’ composition from eleventh-century Wales. I ended the blogpost on a rather sceptical note, sharing Isaac’s doubts as to whether the battle of Gwen Ystrat was a historical event rather than a literary creation.

Fast forward to 2015 and to the latest issue of the journal Northern History. This contains an article by Andrew Breeze in which the case for seeing Gueith Gwen Ystrat as a genuine North British poem of the sixth century is re-stated. Professor Breeze also discusses the possibility that the battle may have been fought in the valley of the River Winster in southern Cumbria. This idea was first proposed by the nineteenth-century Welsh scholar Thomas Stephens who observed that another poem attributed to Taliesin mentions a place called Gwensteri as the scene of a battle fought by a northern king called Gwallawg. Stephens wondered if Gwensteri might be an error for Gwen Ystrat and suggested that both names refer to Winsterdale. Breeze thinks Stephens was right to identify Gwensteri as the River Winster but sees the name Gwen Ystrad as an error for Gwensteri rather than the other way around. He proposes Gwensteri as a native Celtic (Brittonic) name for the Winster.

Other river-names in Cumbria certainly have names of Celtic origin but Winster has generally been regarded as Norse. An explanation of this name based on Norse vinstri (‘left’) was suggested by the Swedish philologist Eilert Ekwall nearly 100 years ago but Andrew Breeze thinks a more likely base is Brittonic gwen (‘white’). A name with the simple meaning ‘white river’ does indeed seem more likely than the harder-to-explain ‘left-hand river’. Taliesin’s poem also mentions a place called Llech Wen (‘White Stone’?) which Breeze identifies as Whitbarrow, a high fell on the east side of Winsterdale. Whitbarrow has a name of Anglo-Saxon origin meaning ‘White Hill’ and a prominent crag called White Scar.

Placing the battle beside the River Winster fits with Breeze’s broader vision of the geography of Urien’s kingdom (Rheged) which he envisages as straddling the Pennine hills to encompass parts of Yorkshire and Cumbria. This is in keeping with a consensus view on sixth-century political geography in which Urien is envisaged as ruling an extensive realm centred on the Solway Firth. I am not with the majority on this matter, nor do I think the name ‘Rheged’ can be written on a map with any measure of confidence – or without an accompanying question mark. The reasoning behind my scepticism (or heresy?) is set out in my book The Men of the North.

Andrew Breeze makes a good case for amending Gwen Ystrad to Gwensteri and for identifying the latter as a Celtic name for the Cumbrian river Winster. His article builds on his other recent work on the Old North and adds further weight to conventional opinion on the location of Urien’s realm. Heretical voices, such as mine, can only respond with a rather pessimistic view on the authenticity of Taliesin’s poems, or with the bleak suggestion that Rheged’s geography might be irretrievably lost.

River Winster

* * * * *


Andrew Breeze, ‘Urien Rheged and Battle at Gwen Ystrad’ Northern History 52 (March 2015), 9-19. [I am grateful to Andrew Breeze for sending me a copy of this article]

Other Rheged-related articles by Andrew Breeze….
‘The Names of Rheged’ Transactions of the Dumfriesshire and Galloway Natural History and Antiquarian Society 86 (2012), 51-62.
‘Early Welsh Poetry and Rossett, Cumbria’ Northern History 49 (2012), 129-34.
‘Northumbria and the Family of Rhun’ Northern History 50 (2013), 170-9.
‘Yrechwydd and the River Ribble’ Northern History 47 (2010), 319-28.

Graham Isaac, ‘Gweith Gwen Ystrat and the Northern Heroic Age of the Sixth Century’ Cambrian Medieval Celtic Studies 36 (1998), 61-70.

Tim Clarkson, The Men of the North: the Britons of Southern Scotland (Edinburgh, 2010) [see pp.68-78 for a discussion of Rheged]

My blogpost from 2009: Catraeth and Gwen Ystrat.

* * * * * * *


29 comments on “The Battle of Gwen Ystrat

  1. Peter says:

    If Gwen Ystrat is Winster, perhaps we can localise the battle site a bit more precisely. The Winster river rises about 2 kms from Lake Windermere -(” Win” place-name element again) near the hamlet of Winster. It is more likely that the river is named from the hamlet, rather than the other way around. Also the sources from what you say do not appear to mention a river, so this is surplus to the speculation. Is it possible that the battle can be located at the hamlet?
    As regards “Windermere” as a placename, it would appear from Wkipedia that the meaning of the “Winder” element is very uncertain, and certainly a derivation would be possible from “gwen” or indeed “vin” Early Celtic, both meaning “white” or “bright” e.g. the sun reflecting off the lake or whatever.

    • Tim says:

      Yes, I suppose the hamlet could be original the source of the place-name, and therefore the site of the battle. However, a line in the poem mentions ‘yn drws ryt’, usually translated ‘at the approach to the ford’, which does suggest a riverside setting.

      Regarding the name of Windermere, the early form ‘Winendermer’ (c.1170) seems to support the usual assumption of Scandinavian origin, i.e. ‘Vinandr’s lake’. But you’re right about the uncertainty over this name, which means a Celtic origin is also possible.

      • dearieme says:

        ‘at the approach to the ford’: the Big Deal ford in those parts was presumably the ford across the Solway. Was there perhaps a similar ford across Morecambe Bay? Is there any reason to suppose that the ford was a big deal that people would recognise by allusion, or will any old common-or-garden ford do? Could mentioning a ford but not a river lead you to consider a ford across an estuary?

        • Tim says:

          I suppose the low-tide route across Morecambe Bay counts as a ford, like the ‘wath’ over the Solway Firth. Presumably the term ryt/rhyd covered all types of ford – including estuary crossings – but my knowledge of Welsh is too small to be certain of that. If the River Winster had a particularly significant and ancient fording-point, a glance at a large scale OS map might show it. The trouble with fords is that most rivers probably have a couple of them, so we’re dealing with a common feature that might be difficult to pin down even if we knew the approximate location of the battle.

          • Dan Elsworth says:

            Morecambe Bay was traditionally forded at a couple of points on what is known as the ‘cross or over sands route’, one linking the shore at Hest Bank near Lancaster with the south side of the Cartmel Peninsula, and the other linking the north side of the Cartmel Peninsula with the south side of the Furness Peninsula, at one of a variety of places near Ulverston, initially apparently Conishead. There is evidence to suggest that at Conishead it was met by an early road known as the Streetgate, perhaps of Roman origin (as I argued in 2007). There is then another ford at the Duddon estuary on the north side of the Furness peninsula. These fords were a ‘big deal’ in that area because they cut an enormous amount of time off any journey into these peninsulas, but the route was very dangerous and even now not to be attempted without a guide. Any attacker coming inland would otherwise be forced along difficult land routes, potentially including coming through the low flat and marshy valley that contains the Winster and Levens, the confluence of the Rivers Winster and Kent. You might ask, ‘why would anyone want to attack this area?’ Two good reasons spring to mind – firstly the very good access to the Irish Sea from this point, secondly the extremely rich iron (and copper) reserves in the area. Control of Furness and Cartmel would also then provide easier access from the south to the whole of the West Coast of Cumbria, again full of iron and with open access to the Irish Sea.

            • Tim says:

              Thanks for this useful info on routes and crossings, Dan. I’m thinking the presence of the great abbey of Furness also adds considerable weight to the idea that this part of (post-1974) Cumbria was regarded as valuable territory in medieval times.

              • Dan Elsworth says:

                It always seems to me that Cumbria is somewhat sidelined from discussions, not a large enough modern population to give it much voice (no local universities carry out extensive research for example), a border area with control often swapping and a poor recorded history, and not the national interest that Wales and Scotland can bring to bear on the poetry being fairly peripheral to matters of English history. Don’t get me started on Furness Abbey – I’m sure there was something significant going on there in the early medieval and possibly Roman period, but again pinning it down is very hard. The area also suffers from an obsession with the Vikings, which is perfectly understandable as they are fascinating, but they have come to dominate discussion of early medieval Cumbria (the modern county that is).

  2. dearieme says:

    One thing I’ve never understood about “Picts” is how St Ninian could preach to them in Galloway. It makes me wonder whether “Picts” is a word with more than one meaning, perhaps simultaneously, perhaps as times change. Once you stop assuming that “Picts” must be from eastern Scotland north of the Forth, how are you to pin them down at all?

    • Tim says:

      The label ‘Picts’ is certainly more fluid than it used to be. Current thinking seems to be moving towards the idea that the Romans regarded everyone north of Hadrian’s Wall as Pictish, and that the label lingered in Southern Scotland for a long time afterwards, which might explain why Pictish symbols are found in Galloway and Lothian. Distinguishing the Picts from the Scots and Britons in the 5th/6th centuries is becoming more and more difficult..

      • dearieme says:

        I’ve seen gallant efforts by Irishmen to justify translating “Cruithne” as Picts, but I suspect it’s essentially a bogus effort to deny that parts of Ireland were inhabited by Britons in ancient times. They don’t seem to grasp that just because many Ulster Unionists believe something about history doesn’t mean it must be wrong.

        • Norval Smith says:

          I am reminded of the Galloway nicknames used by inhabitants of the Rhinns for each other.
          The Heehh-enders (Heich-enders) were called Fin-gauls (formerly tall fair-haired people, with light blue eyes, supposedly descendants of Norsemen). They were found in the southern part of Kirkmaiden parish.
          The Laihhenders (smaller, dark-haired, with dark brown eyes) were supposedly the descendants of savages they called Kreenies, and lived elsewhere on the Rhinns. The Fin-Gauls called them Gossoks.

          (from “Galloway Gossip” by Mrs Maria T*****r, 1877. Choppington, Northumberland: Robert Trotter. (pp. 17-18) [available from].

          W.J. Watson discusses these names in CPNS pp. 178-9. His explanations are:
          Fin-gaul (stressed on the last syllable) is (Gaelic) Fine Gall ‘tribe of strangers.’ Gall is often used for Norsemen, as in the name of Galloway, itself.
          Kreenie is (Gaelic) Cruithnigh (plural of Cruithneach) (a people from Ulster) with the same Gaelic name as the Picts.
          Gossok is (Welsh) Gwasog ‘servant’

          Tim mentions the context for this mixture of races and languages on p.148 of “Strathclyde.” Here we supposedly have people of Norse descent, but known by a Gaelic name, referring to other people of Irish origin, and also known by a Gaelic name, by a derogatory term that is Strathclyde Welsh!

          • Tim says:

            Thanks for this, Norval. I wasn’t aware of the Rhinns nicknames but they’re an interesting aspect of the folklore. I’ve made a note to look up the ‘Galloway Gossip’ book at

  3. Dan Elsworth says:

    I’m particularly glad Andrew has published this as I was clumsily discussing the Winster in my article about hillforts, because of the possible relevance of the nearby Castle Head, from which a collection of artefacts very similar in description (although none now survive to check) to finds from early historic forts in Sctland (Dunnad and the Mote of Mark for example) were found in the 18th century. Given that it has been argued that it is possible to place the battle of Argoed Llywfain at nearby Levens (earliest version Lefuenes), and I would suggest that Rhossed (the subject of another article by Andrew in 2012 and apparent site of another battle) can be equated with Roose, again a short distance away, and you have three battles in a small area of coast. I am in danger of appearing obsessed with my own local area (but then, why not?) but but at both Levens and Roose random ‘Christian’ burials have been found, which need re-examining. Could it be that here we have Catraeth – the ‘battle shore’? (I’ve always been intrigued by its similarity to Cartmel – the second half of which of course Norse for ‘sand dunes’, something that is present on the shore in the area) Not only that but doesn’t Arfynydd, which is one of a number of places mentioned in Argoed Llywfain equates to ‘at the mountain(s)’, which is a remarkably similar meaning to the political area containing all of the places mentioned above in the Domesday Survey, Hougun, from the Norse ‘at or amongst the hills’. I would gladly welcome some critical assessment of what is at best an amateur approach!

    • Tim says:

      Having had a look at your hillforts article, Dan, I’m inclined to think there was a lot more happening on the north side of Morecambe Bay in the Dark Ages than I had previously thought. I’ll email you on that, after I give the article a more detailed reading. Whether this area contains any Taliesin place-names is another matter, but a valid philological case can certainly be put forward (as both you and Andrew Breeze have shown). It’s interesting to note the similarity in meaning of Hougun & Arfynydd. Hougun must surely be an ancient unit of lordship, preceding the Scandinavian settlements, and no doubt with a Celtic name beneath the Norse one.

      • Dan Elsworth says:

        I’m relieved no-one said I was talking complete nonsense at least! I think I mentioned to you, Tim, by email the place-name ‘Ragged Gill’ at Pennington in Furness, the earliest version of which is Reget Gill, which seems potentially of relevance. There’s also Carrkettle and Carley, both of which, I think, indicate a potentially late revival of British interest in the area, perhaps because of the resurgence of Strathclyde, but this needs more work. Fiona Edmonds has recently covered the British survival in Furness and Cartmel, and the real meaning of Ecgfrith’s grant of Cartmel (King Ecgfrith and all the Britons with him, i.e. in collusion with him, not all the Britons given with Cartmel, as slaves or whatever), but it is interesting to note that this also mentions Cuthbert reviving a boy from the dead at the villa at Exanforda – is this the ford on the River Esk on the Cumbrian West Coast, which is essentially the northern boundary of Hougun. Is the vill therefore the nearby Roman fort at Ravenglass? This too potentially connects to the poetry – Gwallawg. He is described as the defender of Llan Lleenawg, which has been equated with a site near Millom, not very far from Ravenglass at all. At Ravenglass are the very well preserved ruins of the bath house (well-preserved enough, I would argue, to have potentially been functioning buildings in the 5th-6th century), which were known as ‘Walls Castle’ – is this Gwallawg’s Castle? Also interesting is the legendary King Aveling of Ravenglass. Is this actually a British name? Perhaps ‘Elfin’? One thing that did dawn on me recently, when considering the early poetry and its possible connections with Cumbria is that there are apparent references connected to two areas – Penrith/Upper Eden Valley and the south-west coast. These two areas are almost directly physically connected by one thing – the High Street Roman road, known as the bretstrett or ‘road of the Britons’!

  4. Helen McKay says:

    Hi Tim, sorry about the bleak, but the story isn’t over yet. I still think that you are right to focus on the general historical context and the logisitics, military and civilian, of possible battle sites. As for Gwen Istrat, the problem is that ‘white’ is such a ubiquitous element in placenames from one end of the country to another, that it really doesn’t provide much of a clue to location. If Urien’s Reged was basically in the middle of southern Scotland, as I think you have laid out a strong argument for, then he is possibly fighting PIcts to the north, or Picts to his west, or Britains to the NW. Now, why should I say Picts to the west? well I suspect dearieme hit the nail on the head. Today we’re used to the idea that in historical times the ‘Cruithnigh’ of Galloway and the neighbouring area of Ulster were confused with ‘Picts’ (mainly because the Picts were also called Cruighnigh.) But the question is, how far back does that confusihon go? Symeon writing in the 1100s spoke of Ninian converting the ‘southern Picts’ and in his day there was a great fad of interest in the Picts of Galloway, so there’s very little doubt this was who he was referring to. But he basically lifted that story from Bede, so was Bede also under the same confusion? If we assume that Ninian who converted the ‘southern Picts’ was St Finnian (as is generally thought) then we can trace his dedications in a great arc from Whithorn around the coast of Galloway and Ayrshire up to Glascow area – as well as directly opposite the channel in Ireland. So its highly likely that Bede too thought of the ‘southern Picts’ as the Cruighnigh of Galloway. Especially as he should have had good advice In this matter given his connection with Whithorn and Iona. So how far back does this confusion actually go? and how would the political ramifications of this confusion be likely to play out for the two groups involved, the ‘real’ Picts and the Galloway Picts? Hard to say. But these group identification myths can be very strongly emotional and powerful politically. I cant say if this confusion is reflected in the symbols at Trusty’s Hill, or whether its reflected in the presence of the Whitecleugh silver chain with its symbols. But I suspect there is a connection.
    So coming back to the battle of White Valley, Gwen Istrat. Now I’m cracking up a bit here. I started off saying that ‘white’ is useless as an indicator of the battle site, but I’ve just mentioned the silver chain deposited at Whitecleugh – White Valley, the same name as Gwen Istrat. Now that’s hilarious… Whitecleugh is interesting as its the valley that holds the headwaters of the river Clyde – always a place of symbolic importance to a tribe and often marks out its boundaries – at least symbolically. In this case, it could be the boundary of Strathclyde, Reged (assuming its in the middle) and perhaps Galloway to the south. Still sounds a bit of an odd place for a battle, but these places are ritually and symbolically critical, and there must be a reason for the silver chain’s presence. Perhaps it was deposited to mark the battle site … Ok didn’t see that coming …

    • Tim says:

      Thanks for these useful thoughts and questions, Helen. It’s good to have some positive feedback on my Rheged=Peeblesshire theory (a topic for a long overdue blogpost, btw).

      Interesting to bring Whitecleuch into the mix. It’s certainly unexpected, but I reckon it has just as much claim as anywhere else to be Gwen Ystrat – or, at least, nobody can say it definitely wasn’t the place Taliesin had in mind.

      Your musings on the Galloway Picts show how much leeway we’ve got these days, especially when questions of Pictishness (Who? Where? Why?) crop up – and how much uncertainty still exists about Ninian, Trusty’s Hill, etc.

  5. dearieme says:

    Here’s a speculation. Perhaps “Picts’ meant Britons who had been subject to very little Romanising influence. So, suppose Roman domination [in modern Scotland] was pretty thin west of, say, Annandale or more likely Nithsdale, and was only fleeting roughly north of the Forth-Clyde line. Then the Picts would be the Britons who lived in Scotland north of the Forth, and those who lived in Galloway. I dare say that the German immigrants could pick up hints about who had been Roman-influenced (as in “Welsh”, “Walloons”, “Vlachs”) and those who had not, so perhaps use of “Picts” would seem natural to them too.

    • Tim says:

      I think this could turn out to be pretty close to the mark. One recent theory on the use of wealh/walh terminology in areas of Germanic settlement in various parts of Europe suggests that it did indeed mean ‘Roman’ or ‘Romanised’.

      • Dan Elsworth says:

        I’d be interested to know if there is a specific reference relating to research on this.

        • Tim says:

          Here’s the reference, Dan:
          Alex Woolf – ‘Reporting Scotland in the Anglo-Saxon Chronicle’ in Reading the Anglo-Saxon Chronicle: Language, Literature, History, edited by A. Jorgensen (Turnhout: Brepols, 2010), pp. 221-239.
          …and here’s a quote from the paper, at page 232:
          ‘The popular explanation that wealh simply denoted ‘foreigner’ holds no water since it is always used of peoples who had at one time been within the Roman Empire and never of Gaels, Picts, Finns, Slavs, or other more exotic peoples.’
          [This comment was edited on 9 June 2015, to correct the Woolf reference]

    • Helen McKay says:

      Apart from the outlying and fairly late symbols on Trusty’s Hill (with the Pictish name Drustan as well), plus the string of silver chains that seem to mark a mid line across Scotland between the walls, is there any other evidence that ‘Picts’ (as the Romans knew them) were ever thought to be as far south as Hadrians Wall? The Romans did have a fair amount of knowledge and interaction with all tribes above the wall and especially with the Picts north of the Forth, so I cant really see them making an error like that. Unless there is something else?
      As for the symbols, they appear fully blown on small items and stones around the same time that the Romans start referring to the Picts around 300 AD. And there are no symbol stones below the line of the Forth (Trusty’s Hill being the exception rather than the rule), so we can reasonably safely say that by the time the Romans start referring to the Picts, this meant the area beyond the Forth which was united by a strong pagan bureaucracy who knew the nature of the symbols and could reach over the country to decide where to raise a stone and what and how it should be carved. This doesn’t necessarily mean that Pictland was fully functioning as a united political or military region (Scotland never has been perhaps even to this day one could argue) but that it was an area and a people that was united in its idea of ‘self’ and which still had a strong thriving culture which allowed and supported its bureaucratic elite to act without ties to the tribe and move across the land, not far distant in concept to how the Celtic monasteries would also act in coming years. In fact, if we assume that the Picts has been receiving druidic refugees from Gaul and southern Britain, and they were then watching Christianity move across the world including Ireland, then they would have been fully aware of being the last of their kind and that their fate was coming too. This sort of situation we know pushes people to become highly conservative and idealistic in their defence of their beliefs, and possibly actively and aggressively so. Its quite possible that the symbol stones are marking out the land and their belief system with symbols, for the very purpose that we their descendants will see and understand and, hopefully, finally, remember.

      • dearieme says:

        “by the time the Romans start referring to the Picts, this meant the area beyond the Forth”: but did it continue to mean only that over the next few centuries?

        • Helen McKay says:

          All the evidence that I can find says yes, it continued to mean the main area of Pictland beyond the Forth. There is no evidence that any tribe below the Forth was part of the Pictish political entity, and the Romans had all sort of ‘spies’ and fortlets and communications with all of Scotland above Hadrians wall. Unless we can come up with some other evidence?

          • Tim says:

            Helen’s mention of Pictish artefacts (silver chains, etc) reminded me of a useful and relevant discussion of Pictish identity in Fraser Hunter’s 2003 Groam House Lecture Beyond the Edge of the Empire: Caledonians, Picts and Romans (published in 2007). The Caledonians, a people of the Highlands who fought against Rome, are usually seen as one of the major components of what later became the Pictish ‘nation’. However, things might not be quite so straightforward….

            “…although ‘Picts’ are not mentioned until the late third century, there is no simple sequence of Caledonians being succeeded in the sources by Picts. The problem with both ‘Caledonian’ and ‘Pict’ is that Roman authors were very variable in their usage….. both were used very imprecisely and it is arguably more realistic to see them as general shorthand for ‘north British barbarians’ rather than seeking any greater detail in their meaning.” [Hunter 2007, p.6]

            • Helen McKay says:

              I’m not sure its fair to blame the Romans for their lack of precise documentation, its more how textual transmission from the time worked. We know so little of how the Caledonians became ‘Picts’ (apart from their origin story), but then we also don’t really know the process of how Pictland became Fortriu became Alba became Scotland either, despite considerably better documentation. These changes of identity and naming are complex cultural events at best. But what I’m reading from HF’s quote is that he is wondering if the Romans may have been vague at times in what they called people, not that the Caledonians and Picts actually lived in the area between the two walls – is this correct?

              • Tim says:

                Yes, your reading is correct. He’s making an observation on the vagueness of ethnic/cultural labels in the Roman texts, which seems quite relevant to the question of who/where/what were the Picts. He isn’t referring directly to the area between the walls, but I think his point about North British natives in general being casually lumped together under a vague ‘Pictish’ label can be applied south as well as north of the Forth-Clyde line.

  6. Helen McKay says:

    > … I think his point about North British natives in general being casually lumped together under a vague ‘Pictish’ label can be applied south as well as north of the Forth-Clyde line.

    I’m a bit confused, is there some historical need for (real) Picts to be below the Forth-Clyde line? and where is there any evidence of this situation? its just that these notions can be insidious if we let them pass unquestioned …

    • Tim says:

      Some Roman writers, and even Britons such as Gildas, seem to include areas south of the Forth-Clyde isthmus in their vision of ‘Pictland’. This doesn’t mean we ourselves should describe the inhabitants of these areas as ‘Picts’. It’s simply a reminder that our modern understanding of these terms was not necessarily valid 1500 years ago.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s