The longphort phenomenon in Early Christian and Viking Ireland

Published in Features, Issue 3 (Autumn 2004), Pre-Norman History, Volume 12

Recent discoveries of Scandinavian settlement remains at Woodstown, Co. Waterford, have produced great excitement, and not a little hyperbole. Settlements such as this are described as longphuirt or longphorts in the present literature. But what exactly was a longphort? Michael Gibbons investigates.


The early stages of Scandinavian settlement in Ireland and Britain remain difficult to locate in the archaeological record, which is why the recent discoveries at Woodstown, on the upper tidal reaches of the River Suir, are so important. Woodstown has produced ships’ nails, hack-silver, the remains of Viking weaponry and a high-status pagan burial, although early claims that this is the grave of the Viking leader Rodolph seem premature. If it can be precisely dated, Woodstown may give us a good look at the original Viking presence in Waterford harbour and hence a good analogy for the origins of Dublin.


The archaeological evidence

Reconstruction of a Viking winter camp at Repton, 873–4. (Vikings and the Danelaw, p.60)

Reconstruction of a Viking winter camp at Repton, 873–4. (Vikings and the Danelaw, p.60)

Much of the discussion of the likely significance of Woodstown has drawn on other supposed examples of longphuirt at Dunrally (Co. Laois), Athlunkard (Co. Clare) and Anagassan (Co. Louth). Scandinavian fortifications, such as at Hedeby (Germany) and Århus (Denmark), feature a D-shaped enclosure, but these are much more mature than the early longphuirt. The remains of a winter camp of AD 873–4 at Repton, on the Trent near Derby and one of the few such sites to be thoroughly excavated, do feature a D-shaped enclosure. A burial mound next to the enclosure contained a Viking war-grave and the enclosure itself incorporates an Anglo-Saxon church.
The three Irish sites all feature D-shaped enclosures on the edge of rivers but other than that have little in common. John O’Donovan, and later Eamonn P. Kelly, suggested that the name Dunrally is an Anglicisation of Dún-Rothlaibh (Rodolph’s fort), which they connect to Longphort-Rothlaibh of AD 862. The annals refer to


‘. . . the destruction of Longphort-Rothlaibh by Cinnedidh, son of Gaithin, Lord of Laighis, on the fifth of the ides of September; and the killing of Conall Ultach and Luirgnen, with many others along with them’.


The fleet, which was destroyed, had recently arrived from Lochlainn and a ‘dreadful slaughter’ was made of them. The dún itself is a ringfort on a small hill surrounded by a large D-shaped ditched enclosure. Kelly draws attention to the similarities of these arrangements to those described at a Viking fortification on the River Dyle in Belgium, which was destroyed in AD 891.


The site commonly known as Athlunkard (Ath-an-longphuirt) in County Clare is not connected to any specific event and is actually in Fairyhill townland (two townlands to the north), half a mile from Athlunkard bridge. It consists of a ditched bank with a raised platform in the centre, on a bend in a river next to its confluence with a tributary. Ninth-century, but not Scandinavian, artefacts have been discovered on the site and also from St Thomas’s Island nearby. A Viking silver weight was found directly opposite the site at Corbally. Kelly and O’Donovan theorise that St Thomas’s Island may form part of the complex.
The Athlunkard name is of uncertain date; it is not directly connected to the enclosure and it could be related to another site. This may well have been Irish and later in date. If the name does refer to a Scandinavian longphort then it seems more likely that the longphort in question is Limerick City, only a kilometre and a half from Athlunkard bridge.
Lisnaran/Linn Duachaill fort sits on a cliff face overlooking the River Glyde to the north-east. It contains the remains of a circular and a rectangular structure and may have featured more extensive defences outside the main enclosure. The only find from the site is a coin hoard dated to the fourteenth century from the base of the cliff. The origin of the belief that this is Linn Duachaill appears to be a note by Godard H. Orpen in an article, ‘Origins of Irish motes’, from the 1908 Journal of the County Louth Archaeological and Historical Society:


‘Linn Duachaill was a ship-shelter of the Northmen near the junction of the rivers Clyde and Dee in the County Louth. The longphort at Linn Duachaill (Ann. Ulst. 840) can I think be confidently identified with a headland fort at the mouth of the Glyde called Lis-na-Rann.’


Orpen seems to have derived this belief from Wright’s Louthiana, but this book tends to ascribe every fortification type to Danish influence. In any case, the combination of a ringfort with circular and rectangular structures is diagnostically Irish rather than Scandinavian.
The three proposed longphuirt vary radically in size. Dunrally measures 350m long by 150m wide, Athlunkard 75m by 30m, and Lisnaran 73m by 34m. Dunrally in particular dwarfs the enclosed area at Repton, c. 45m by 80m.
Until we know more about what went on inside these enclosures—and that at Repton appears to have been empty—we can say little about the significance of these variations. None of the Irish examples have any particular connection to the Vikings other than a Scandinavian presence in the area, their shape, and place-name evidence. Only the excavation of one of these sites will clear up the large areas of uncertainty about them. Until this has been done we are forced to rely on the historical evidence, in particular that of the annals, to understand the longphort phenomenon.


The meaning of longphort in the Viking Age

Reconstruction of a Viking knarr based on findings from Hedeby. (Oxford Illustrated History of the Vikings, p.190)

Reconstruction of a Viking knarr based on findings from Hedeby. (Oxford Illustrated History of the Vikings, p.190)

Writing in the 1998 Journal of Irish Archaeology on ‘Forts and fields: a study of monastic towns in seventh and eighth century Ireland’, Catherine Swift noted that:


‘The number of scholars working in the field of Old Irish is historically very small and the production of the Dictionary of the Irish Language has involved the energies of many of the key figures working in the field between 1913 and 1976. Such men and women were linguists, interested in the grammatical complexities of the Irish language and, for the most part, particularly concerned to elucidate the connections between Irish and its ancestor Common Celtic and, further back, Indo-European. As archaeologists we tend to imagine that the primary focus of a dictionary is translation, but at least as important to the Dictionary and its compilers was the identification of specific stem classes (for nouns) and the ancestral pre-verbs which made up the verbal complexes in Old Irish. English translations were often not their primary interest and they tended to be taken verbatim from editions extant at the time the particular section was compiled.’


Writing on ‘The origins of Dublin’ in Studies on early Ireland, Patrick F. Wallace observed that:


boat‘Ó Corráin talks of a defended ship harbour, Clarke of an emporium or trading station, and de Paor of a shore fortress. The RIA Dictionary defines longphort as a camp, encampment, temporary stronghold (or a) mansion, princely dwelling, stronghold, fortress.’


The precise meaning of longphort varied until it acquired its ‘final’ form. In this it was rather like the Viking settlements themselves. Kieran O’Conor has demonstrated that the word denoted a fortified Irish palace site by the fourteenth century. Translations, such as O’Donovan’s of the Annals of the Four Masters, seem to have projected this meaning into the Viking Age. Following the example of these earlier scholars, many archaeologists seem to have taken a preconceived view of a longphort—as some form of distinctive Viking castle—into the field with them.
The first hint of permanent Viking settlement in Ireland may be in the annals for AD 835, when ‘Cill-Dara was plundered by the foreigners of Inbher-Deaa and half the church was burned by them’. Inbher-Deaa was somewhere in Wicklow, possibly at the mouth of the River Vartry. The raiders were connected specifically with this area so there may have been a small camp there. From 837 the Viking leaders begin to acquire names in the Irish annals. These may have been learnt from freed slaves or perhaps from direct contacts with the invaders, who began to overwinter on the outlying islands in the late 830s. In 837 two fleets of 60 ships appeared on the Liffey and on the Boyne. After this the inland lakes became hunting grounds for the raiders.
In the first direct annalistic reference to the longphort phenomenon, the Annals of the Four Masters in 839 refer to


‘The plundering of Lughmhadh by the foreigners of Loch-Eathach; and they made prisoners of many bishops and other wise and learned men, and carried them to their fortress (longphort), after having, moreover, slain many others’.


It is unclear where exactly this longphort was, but in the next year the picture is more precise:


‘A fortress (longphort) was erected by the foreigners at Linn-Duachaill out of which the territories and churches of Teathba were plundered and preyed. Another fortress (longphort) was erected by them at Duibhlinn, out of which they plundered Leinster and the Uí Neill, both territories and churches as far as Sliabh-Bladhma. The plundering of Cluain-eidhneach, and the destruction of Cluain-Iraird and Cill-achaidh-Droma-Fota, by the foreigners.’


The Linn Duachaill longphort may have been on the site of a monastery. In 841 ‘Caemhan, Abbot of Linn-Duachaill, was killed and burned by the foreigners’.
The invaders may have simply moved into the abbot’s property after eliminating him. Events at Duibh-Linn may have been similar. An abbot is mentioned there in the annals of AD 650 and it has been suggested that the original Viking presence in Dublin may have consisted of raiders who had seized the monastery and its belongings for short-term gain. The settlement at Ath-Cliath, slightly further up the river, might have been a later development. The annals for 841 refer to


‘A fleet of Norsemen on the Boinn, at Linn-Rois. Another fleet of them at Linn-Saileach, in Ulster. Another fleet of them at Linn-Duachaill . . . Maelduin, son of Conall, lord of Calatruim, was taken prisoner by the foreigners.’


It is not clear what the Norse presence at Linn Saileach (Lough Swilly or Strangford Lough) or Linn Rois (opposite Rosnaree on the Boyne) consisted of. The locations, pools on rivers, seem similar to Linn Duachaill and Duibh Linn so there may have been some form of base at these locations. Another group, the ‘foreigners of Cael-Uisce’, raided Diseart Diarmada, and a dunadh is reported on Lough Ree in 845 but it is unclear how, if at all, this differed from a longphort.
In 853 a fleet, probably from Viking Scotland, arrived and took hostages from the ‘foreigners of Ireland’ as well as tribute from the local population. This suggests that the Scandinavians were in direct control of Irish groups rather than merely raiding the locality. The longphuirt from which this control was exercised must have been radically different from the original examples reported in 841–2.
Viking sites are referred to by a variety of terms, including longphort, dún, caisteoll and island. The different terms may simply describe their function (as appears to have been originally the case with the various longphuirt), their physical layout (as with the caisteol of Cork destroyed in 865), or their location (as with the Vikings of Loch Cuan, apparently expelled from their ‘island’ in 942).
Aed Fhinnliath, in his operations on the north coast in the 860s,


‘. . . plundered the fortresses (longphuirt) of the foreigners, wherever they were in the North, both in Cinel-Eoghain and Dal-Araidhe; and he carried off their cattle and accoutrements, their goods and chattels’.


The cattle in the longphuirt suggest that they controlled areas of pastureland. Then,


‘The victory was gained over the foreigners, and a slaughter was made of them. Their heads were collected to one place, in presence of the king; and twelve score heads were reckoned before him, which was the number slain by him in that battle, besides the numbers of them who were wounded and carried off by him in the agonies of death and who died of their wounds some time afterwards.’


The lack of the distinctive Scandinavian place-name elements, such as -by, aster or staed, which indicate settlement by Scandinavian farmers, suggests that these lands were farmed by Irish clients controlled by the quite small number of warriors described above.
The word longphort was adapted to circumstances by those who used it. It did not even necessarily have any Scandinavian connotations. In 858 Maelseachlainn made war on Aed Fhinnliath and there was


‘A hosting of Leinster, Munster and Connaught, and of the southern Uí Néill, into the North, by Maelseachlainn, son of Maelrunaidh; and he pitched a camp (longphort) at Magh-Dumha (plain of the mound) in the vicinity of Ard-Macha. Aedh Finnliath, son of Niall, and Flann, son of Conang, attacked the camp (dunaidh) that night against the king, and many persons and destroyed by them in the middle of the camp (longphort); but Aedh was afterwards defeated, and he lost many of his people; for Maelseachlainn and his people manfully defended the camp against the people of the North.’


Maelseachlainn had established his longphort on the night it was attacked and its defences cannot have been particularly well developed. Thus longphort could vary in meaning from a site of 60 years’ standing with Irish clients dependent on it and its own hinterland, such as Dublin at the time of the expulsion in 899/902, to a fleeting, if impressive, collection of soldiers and whatever structures they had established in the space of a few hours. Nor is this an anomaly; in 969 we read that


‘An army was afterwards led by Domhnall Ua Néill, with the soldiers of the North, i.e. the races of Conall and Eoghan, against the men of Meath and the foreigners, so that he plundered all their forts (dunadh) and fortresses (longphuirt), and spoiled Uí-Failghe and Fotharta; and he took revenge on them on that occasion for their opposition to him, for he erected a camp (longphort) in every cantred (tuatha) of Meath from the Sinainn to the Bealach-Duinn.’


Longphort, as used here, seems to refer to an inland garrison without any nautical/riverine element.
Edel Bhreathnach has discussed the development of the term longphort and has noted its use as an Irish military camp. The earliest appearance of the two constituent words in one phrase is in the TriPartite Life of St Patrick, in the sense of looking for a landing place for a ship, gab port a long. She discusses the later development of the idea into an Irish military camp and demonstrates that, by the eleventh century, this meaning was in general circulation and had no echoes of the Viking naval camps of the early 840s.


Scandinavian operations in the Anglo-Saxon chronicles


Viking raids, after Graham-Campbell. (The Viking World, p.23)

Viking raids, after Graham-Campbell. (The Viking World, p.23)

The Viking bases of the 830s and 840s were probably originally conceived as temporary bases of operations from which the invaders would subdue the natives and then divide up the land. Outside of the Dublin hinterland this policy failed, and only after its failure did the commercial possibilities of the longphuirt become important. This is why the evidence from England, where the invasion succeeded and where many of the camps and fortresses did not develop any further, might be useful in understanding the original longphuirt.
The large Scandinavian campaigns in England were launched immediately after the first wave of longphort-building in Ireland. In many cases those planning their descents on Mercia or Wessex had probably honed their skills in Ireland. In England ship-armies often cooperated with raiding-armies, which relied on horses for mobility. The Saxon ‘great armies’ found it extremely difficult to bring the Danes to battle in the field or to drive them from their fortifications once they had settled down behind Roman or Saxon walls. Operations in Ireland may have followed a similar pattern. Some of the Scandinavian material from Irish sites no doubt resulted from such temporary occupations.
A persistent pattern of Viking operations in England was a rapid advance, by horse or ship, from one year’s overwintering spot to secure an undefended strongpoint. From here the Danes would plunder, take slaves and defy any native forces that came to unseat them.


The raiding-army of the 860s/870s broke out of East Anglia in 866 and headed for York, which they then held. The divided Northumbrians were defeated when they tried to retake the old city and both claimants to the throne were killed. The survivors were forced to make peace with the invaders and the groundwork for a powerful Danish presence in Northumbria had been laid.


This pattern continued throughout the 870s. At the outset of a year’s campaign an important and defensible objective such as London, the Royal Church at Repton or Chippenham would be captured, often by surprise, and then defended while the land was ‘ridden over’ and plundered. The main objectives of the first Viking thrusts seem to have been Roman cities, with their surviving fortifications and Anglo-Saxon administrative centres. The approach was highly successful. The Danes in England seem to have started building their own fortifications in the mid-880s/890s. While none of these have been conclusively identified, one wonders whether Irish chroniclers might have described any or all of them as longphuirt.


Conclusion: the likely nature of Viking encampments and the difficulties of using the name longphort to find them
There is as yet no diagnostic site type for a Viking longphort, nor is it clear how one might differ from a Viking dúnadh, or indeed an Irish river-edge enclosure of any date. The locations of Viking sites might be expected to share certain characteristics. Islands were popular with the Scandinavians in Ireland, Britain and France, particularly those in strategically important harbours and lakes. Another common location mentioned in the annals is next to a pool on the upper tidal reaches of a river, as with the various –linn locations of the 840s. Locations on the edge of marshes are attested to in England and in Belgium. Viking encampments probably tended to be built on or around existing sites. These may have been the focus for the original attack and have later acted as bases of operations. This would fit with historical accounts of Viking operations in England and France.
The actual physical footprint of a Viking encampment, whether naval or otherwise, may have been extremely small in comparison to the size of the fleet or army involved. The encampment at Repton would have offered completely inadequate housing for a ‘great army’ of thousands with its ships and horses. The enclosure itself, which contains no settlement materials, was possibly the fortified centre-piece of the camp, with the soldiers either billeted in Saxon buildings or else in more temporary accommodation—perhaps aboard ship?
To the eyes of Irish observers, the tents, campfires, light structures etc. of an army settled around the captured monastic site of Linn-Duachaill might have appeared very similar to the same army camped elsewhere. After a time the word, which had come to mean simply ‘military camp’, came to refer to a fortress. This meaning was then applied anachronistically by the translators of a later era to describe the original, rather humble, Viking structures of 839–41. The greatest meaning that one can draw from the word would seem to be ‘military site of some nature, probably Viking Age or later, duration of use uncertain’. When the word longphort is encountered as a place-name or in the written evidence it has no necessary connection with Vikings or even with ships. Interpreting sites as though it does may lead to serious errors. Therefore it might be wise not to use longphort as an archaeological term. The Gaelic inland sites it described by the fourteenth century presumably bore little resemblance to whatever it was the Vikings built near the Black Pool in 840–1. We should instead discuss longphuirt, and indeed dunadh, using the English translation, i.e. naval camp, fortified island, castle, etc., that seems most appropriate.
The archaeological evidence for these sites, in both Ireland and Britain, is still thin on the ground. This is what makes the Woodstown site so important. If the Scandinavians used a similar modus operandi in Ireland and England (with greater success in England), then Woodstown will join Repton as one of the only Viking camps to be excavated in the British Isles. Whatever is discovered there will help to illuminate Viking operations on both sides of the Irish Sea.


Michael Gibbons is an independent archaeologist, former co-director of the Sites and Monuments Record Office and a member of the Archaeology Committee of the Heritage Council.


Further reading
E. Bhreathnach, ‘Saint Patrick, Vikings and Inber Dée—longphort in the early Irish literary tradition’, Wicklow Archaeology and History 1 (1998).
J. Graham-Campbell, R. Hall, J. Jesch and D. Parsons (eds), Vikings and the Danelaw: select papers from the Proceedings of the Thirteenth Viking Congress (Oxford, 2001).
E.P. Kelly and E. O’Donovan, ‘A Viking longphort near Athlunkard, Co. Clare’, Archaeology Ireland 12 (4) (Winter 1998). 
K. O’Conor, The archaeology of medieval rural settlement in Ireland, Discovery Programme Monographs No. 3 (Dublin, 1998).


Copyright © 2024 History Publications Ltd, Unit 9, 78 Furze Road, Sandyford, Dublin 18, Ireland | Tel. +353-1-293 3568