Settlement on the Western Seaboard c. AD300-800: Dalriada and North Uist

Ewan J. Innes, MA(Hons Scot. Hist.) FSA Scot

© 1993

Synopsis:  This essay describes the settlement patterns on the western seaboard of Scotland from AD300-800.

Please see my copyright policy if you wish to cite any part of this essay.

Introduction

The lack of knowledge in many areas of the history of the western seaboard of Scotland is astounding, yet little effort is made to try and answer some of the fundamental questions that arise in this area. Some periods are slightly better known than others as detailed comparisons with other places have shown similarities in material culture and artifactual style.

One of the key periods in Scottish history however is little known, documented, or studied. This is a period which marks the beginnings of modern Scotland and the introduction of a new people into the geographical area of Scotland. A new people who were ultimately to give their name to that geographical area and who were to found the modern nation of Scotland. It is this period from c. AD 300-800 that I will be looking at.

Who were these people, why did they settle here and, more importantly, how can we identify and evaluate their settlements? It is these questions and others which I will be trying to answer in this dissertation. I will be using information from a site of acute archaeological importance in the Western Isles and trying to link it with the little archaeological information known from the 'heartland' of Dalriada to try to see what conclusions we can make -if any- about the Scots as a nation their culture and settlement patterns.

The date of AD 300 is not as arbitrary a starting point as it sounds. It was chosen because in 297 complaints were made of attacks on the fortified Roman frontier by two peoples the "Picti" and the "Scotti". Two peoples who were to be heavily involved both with each other and ultimately in forming the kingdom of Scotland.

The links between the two peoples were significant. Gildas wrote c. 450 of "foul hordes of Scots and Picts" when detailing the Roman evacuation of Britain the century before. The Scotti seem to have become allies of the Picts at an early stage and the seaborne attacks mounted on the Roman province to the south caused havoc and led to changes in Roman strategy. By 367 the Scots and the Picts were part of the Barbarian conspiracy which led to the overrunning of Hadrian's wall.

It is more than likely that this involvement of the Scotti in affairs outside their territory led to the settling of some Scots in Scotland after all, they could see Kintyre from Antrim and would no doubt have some idea what the geography across the water was.

Across Europe in this period there was a great wave of migrations of tribes and peoples, it seems only logical therefore, to expect evidence of this in relation to the Scots. It is possible that the evidence from North Uist to be outlined here may indeed represent an outpost of such a wave early in the 4th century.

The Scots of Dalriada were originally from Ireland, from an area along the Antrim coast and part of the province of Ulster (now counties Antrim and Down). The originator of the political territory of the Dál Riata in Scotland was Fergus Mór mac Eirc who arrived in Kintyre c. 500.

When Fergus Mór removed from Ireland to Scotland, there was no sundering of ties or relinquishing of authority between the two sections; and this continued to be the case under Fergus Mór's successors. Evidence for the continued rule of Dál Riata in Ireland by the Scottish branch is found at the Convention of Druim Cett. This was convened c. 575 to discuss the future relations and status of the Irish Dál Riata between Aed, son of Ainmire (d. 598) the leader of the Northern Uí Néill -the most powerful people in the north of Ireland at the time- and Aedán mac Gabráin king of Dál Riata in Scotland (d. c. 608).

The accounts of the convention, show that the status of the Scottish dynasty was more or less confirmed in its existing form. The right to levy taxes and tribute went to Aedán, while the right to raise the armed forces of Dál Riata in Ireland went to Aed, as overlord of Ireland. Essentially this meant that while Aedán and his successors could maintain their authority over the Dál Riata in Ireland, they had control of the government of the territory.

Involvement of the Scottish dynasty in Ireland was bound up with conflict between themselves, and the other two groups who shared Ulster with them. The Dál Fiatach (often known as the Ulaid)¹ on the Down coast, and the Dál nAraide or Cruithne in the interior.² The province of Ulster was ruled in turn by the Dál Fiatach and the Dál nAraide, and conflict between the two groups appears to have been common. The Dál nAraide and the Dál Riata became close and possibly went into an alliance with each other at least from the end of the sixth century.

Involvement of the Scottish Dál Riata in Northern Ireland came to a shattering end at the battle of Magh Rath. Domhnall Brecc brought the kings of Scottish Dál Riata into conflict with the kin of the abbots of Iona by backing his ally and possibly nephew, Congal king of the Dál nAraide and Ulster against the Uí Néill highking, Domhnall son of Aed mac Ainmerech.

The annals do not mention the Irish Dál Riata acting independently at all until after Magh Rath in 637. From this time, Scotland and Ireland began to go their separate ways, and it is possible that the Scottish dynasty forsook their claims to territory in Ireland.

The period after Domhnall Brecc's death in 642 at the hands of Owen king of Strathclyde, marked the beginnings of a decline of Dalriada and also Iona. Iona suffered a setback of huge proportions at the Synod of Whitby over the date of Easter and the tonsure. It was the political setback over control of the fledgling English church that was to cause the most harm to Iona. About 661 the seventh Abbot of Iona, wrote that Dalriada was being held down by "strangers", strangers who belonged to one or other of the four groups struggling for political hegemony at the time.

Kingship in Dalriada followed the Irish system with a rí ruirech, 'king of overkings', with two further grades below this, the ruirí or 'great king' and the basic level of the king of a tribe or petty kingdom, the rí. When the Scots arrived, they were divided into three kindred groups each with its own rí and territory. The Cenél Óengusa ('kindred of Óengus') occupied Islay. The Cenél Loairn ('kindred of Loarn') held Colonsay as well as present day Lorn and looked on the northern march with the Picts. And the Cenél nGabráin ('kindred of Gabran'), held Kintyre, Cowal, Bute and Arran and also the overlordship of the Scots.

The organisation of these groups for fiscal and military purposes detailed in the Senchus fer nAlban gives us a very important view of how their society was organised. The Cenél nGabráin enjoyed the overlordship of Dalriada for the sixth and much of the seventh century. and there is little doubt their status was enhanced by the ordination of Aedán mac Gabráin as overking of Dalriada by Columba in 573.

By the end of the seventh century the military failures that had befallen the Cenél nGabráin allowed a fourth kindred to emerge: the Cenél Comgaill, while they themselves were replaced as kings of Dalriada by the Cenél Loairn. This sudden rise to power of the Cenél Loairn was eclipsed less than a hundred years later by the rise of Pictish power in the east and the 'smiting of Dalriada' in 741. Yet, a century after this a king of the Cenél Loairn would become king of the Picts and unite the two nations into one, Alba.

Having laid out a brief outline of the history of Dalriada, what information can be gathered for settlement within that region and to what extent can it be juxtaposed in other areas? Let us first look at three sites from Mid-Argyll, which had, seemingly, by the late seventh century become part of the territory of the Cenél Loairn. We will then turn to look at a site of premier importance in Scottish settlement history, to see what we can gain in terms of useful comparison. We will then turn to look at a few settlement sites outside the Mid-Argyll area but important for giving a true picture of the settlement patterns of the Dalriadic Scots.

Chapter I

The Mid-Argyll Evidence³

Let us look now at the Scots of Dalriada, their social customs and organisation, and see if it can lead us to any conclusions regarding settlement. There have been very few documented sites that can be firmly and completely associated with the Scots of Dalriada and I shall look at some of these, situated in Mid Argyll within the territory of the Cenél Loairn.

The first of these sites is a crannog in Loch Glashan (map 3 no.60). This site yielded important information about the Scots and their culture, for many artifacts survived that are normally missing from dry land sites. Not only was there evidence of metal working in the form of crucibles and slag, but also leather and wooden objects point to other manufacturing processes. There were also examples of imported class E pottery vessels (E-ware) from the Atlantic coast of France whose date would be between the 6th and 8th centuries.

The crannog was investigated when low water levels associated with the building of a dam revealed part of it. It was constructed using layers of brushwood on top of mud on the landward side, while seaward layers of logs and revetted stones overlay brushwood. There was a thin scatter of stones across the site with indications of a hut on the north-west side. A rectangular structure 25ft x 15ft was defined by a series of massive oak timbers with flattened tops suggestive of the floor of a house. The hearth seems to have been constructed from stone slabs and sherds of Rhineland pottery were associated with it. As only part of the site could be excavated at the time there was no attempt to investigate the indications of an older structure underlying the other two buildings.

The large numbers of wooden objects included troughs, bowls, a paddle, scoop, worked timber pegs and pins, together with scrap leather. Also found were sheaths, shoes a jerkin and a dug-out canoe. A bronze brooch with a possible amber setting gives an approximate date for the abandonment of the site at around 850. In addition a 'bearded' iron axehead and large numbers of quern stones, both whole and broken, were found.

When we link this with evidence from Dunadd and the somewhat enigmatic site of Brouch an Druimein (nos 18 and 67), we can get a picture of a fairly dynamic and sufficient society with trade links to the Continent. A fact which is borne out by Adomnan in his life of Columba.⁴

Brouch an Druimein is a contemporary to Loch Glashan although the evidence here is slight due to destruction by gravel workings. Discovered here were two timber buildings overlying one another which had at one time been destroyed by fire, it was at least partially enclosed by a ditch and evidence of metal-working and again imported pottery was produced. It is fairly safe to presume from the evidence here that several periods of reconstruction over a considerable time span is involved.

The next and possibly most important site is Dunadd. Several excavations here have produced a wealth of material and together with its source references and carvings give it a possibly unique place in the history of the region. Dunadd was a centre for the production of jewellery and the working of precious metals, as the quantity of moulds and trial pieces testifies. Imported pottery of class D (D-ware) comprising wheel made bowls and E-ware jars, beakers, bowls and pitchers were discovered, indeed, the assemblage of E-ware is the single largest known from any site in Britain.

The structure of Dunadd consists⁵ of some 5 plateaus each of which is defended by dry stone walling. At the top is the upper fortress that was extended at a later date to the original building giving it a curious kink in the north-west wall. Access was on the north-east where an uneven path had been cut into steps allowed access to the summit. Only a short stretch of the walling to this earlier fort is revealed today (see plate 1) some 4m thick and overlies earlier occupation. Possibly about the same time as this alteration was carried out, the shelf that lies some 5m below the fortress was surrounded by a wall some 2m thick linking rocky outcrops. It is well preserved on the north-east (plate 2) where four irregular courses can be picked up and standing to a height of 0.6m. This wall encloses the important group of carvings and inscriptions of the fort (plates 9,10 &11). Deposits in this area in the 1904 excavations were up to 5ft deep in places and produced many fine finds -although no record of stratification exists and there is no clear indication available where exactly finds were made. On the north-east side of B there exists a rectangular stonework structure, originally described as a buttress. The outer facing stones are still visible and it overlooks the east side of the interior of the fort. (plate 4).

The largest area of the fort occupies the irregular terrace outlined by the walls that run from the natural rock in the north in a series of straight sections with right angled turns round to the base of the rock below the summit fort. Parts of this wall are still well preserved, although much of it exists as tumbled ruin (plate 6). Excavation of the wall showed two phases of construction on it. Originally the wall was constructed between 2.6 and 3.1m thick with battering on both faces. A later phase increased this thickness to some 4m and in places nearly 5m.

The entrance to the fort is through a rather dramatic natural cleft in the rock on the south-east. The entrance is about 3m wide and has sheer sides of about 2.3m and presumably had a very substantial timber gateway at the junction of the cleft with the defensive walling.

Many walls were traced during the successive excavations in the interior of the fort. Most of these are now no longer visible on the surface. To the west there is a small cell or chamber (e) although quite what relation this had with the walling is unclear. In the area of wall D, there were discovered copious quantities of charcoal-rich black soil with large quantities of clay mould and crucible fragments. Indications are that wall D was demolished about the eighth or ninth century after this period of manufacturing was over.

Dunadd, it is clear, was not constructed in one period but over several periods although the precise relation one to the other will probably never be known satisfactorily.

Artifactual evidence abounds from Dunadd, although the lack of stratigraphy does not help us to date accurately objects from elsewhere. One of the most interesting aspects of these artifacts is the numbers of moulds and templates for the making of jewellery. Pieces of slate have been used as drawing boards for the working out of designs. One slate has animals, a bird and rosette motifs engraved onto it with one design drawn using compasses. Although known to be from the 'Upper Ridge-fort', it lacks stratigraphy and remains a datable problem.

There were large numbers of elaborate metal finds including a high quality 6th-7th century Anglo-Saxon example. A garnet in gold setting with filigree. Ironwork was also produced in large quantity, including weapons, knives, and tools. The pottery found here included large numbers of imported ware from the late 6th century to the 8th. Other pottery of later date was also found. Indications of a literate, probably ecclesiastic population at Dunadd were provided by an incised stone with the inscription I(N)NOMINE ('in the name of the Lord') of possible 8th century date and a piece of yellow arsenic sulphide used in the illumination of manuscripts.

A look at Map III shows to what extent Dunadd was one of many forts within a small area. The chain of forts and monuments up Loch Craignish, along the valley of the Add and up towards Loch Awe via Kilmartin and Ford, point to a highly able warrior aristocracy during the Iron Age, and it would be unlikely if these fortifications were not utilised by the incoming Scots. The lack of surviving settlement sites in Dalriada and wider afield in Pictland may indicate a penchant for building in wood, a material that will not last except on crannog sites, which is why the Loch Glashan site is so important. To look at other sites within the Mid-Argyll area but not on the map we can see a few cases of surviving indications of post-holed wooden buildings. At Auchategan (NS000643) we have a site with a possible long term occupational history.⁶ The interesting point to be made here is that a C14 date was taken from the hearth of a structure relating to period 4 (on-site terminology) gave a date of AD 660 ± 100 with involvement in iron working indicated. Two sites at Ardnadam (NS163791 & NS163789) again showed signs of industrial activity during this period. These seem to prove a scenario where recoverable and datable settlement is related to industrial and not necessarily undefended activity.

Two interesting sites to keep in mind regarding 'figure of eight' style buildings are at Brenfield Hill (NR822827, NR823829) and at Allt na Sac (NM989048)⁷ in Mid-Argyll. Both are at present on Forestry Commission land and access was impossible when I went to look at them, however Campbell and Sandeman in their survey⁸ noted that they bore a 'figure of eight' plan, and were alongside several others. Of particular interest here is the Brenfield complex which Campbell and Sandeman described as consisting of:

A large group along the hillface, with some of types E [small ovals, ave. 18ft x 13ft] and G [sub-rectangular, ave. 23ft x 15ft], and one rectangular foundation overlying. A typical pair is 25ft x 15ft 6in max., and consists of an oval 10ft x 8ft (int.) with walls 3ft to 4ft thick, touching at one end an oval 6ft x 4ft (int) with walls 2ft thick.

The Allt na Sac site is described thus

"A group of about six pairs, near a burn, with a large spread mound containing charcoal."

From the descriptions and the geographical position of them these are vitally important sites and cry out for detailed investigation. They are also important for another reason. If archaeological evidence can be produced dating these to the same period as the Udal settlement we will have proof of Dalriadic settlement on North Uist.

It is more than likely that many other examples of this type of building or settlement complexes await discovery, and investigations of the machairs of Islay, Tiree and others may well turn up examples. Further detailed archaeological and field investigation is desperately needed to solve this problem.⁹

Chapter II

The Udal Evidence

The site of Coileagan an Udail situated on a peninsular on the north side of North Uist, marks a watershed in the settlement history of the western seaboard of Scotland. The sites there show an unbroken settlement pattern from the Neolithic until desertion in 1697, and as such are invaluable for dating evidence of other sites on the west coast.

The period around 300 marks a watershed concerning this site as with mainland Scotland. In 1972, underneath early Viking levels and overlying late Iron Age field systems and 'cremation platforms' a settlement of immense importance was discovered. It lay within the stratigraphic sequence of an hitherto unknown period in settlement terms, that period from AD 300 to 800. The only other excavated site existing in Scotland is at Buckquoy in Orkney, where there are problems due to a lack of stratigraphy. The evidence from the Udal site points to a possible tribal or political centre only some 75km north of the main Dalriadic heartlands.

The settlements at Coileagan an Udail are set into a shell sand machair landscape. In antiquity, a ridge ran from the tell Udal South (US), some 200m North to the tell Udal North (UN). In the Iron Age, the cultivation fields ran down this ridge, at the bottom of which were situated burning ghats of a ritual funerary purpose. The US sites hold the late Iron Age and Bronze Age foci with the UN tell consisting of the post Iron Age through until 1697 when the site was abandoned due to severe sand blow.

There is evidence to show that while there was settlement continuation there may well not have been an Iron Age population continuation, as the structures and artifacts change character abruptly and drastically. In all there were excavated some 8 'figure of eight' buildings with associated 'four poster' minor outbuildings, surrounded by palisading (see plan I).

The buildings themselves are almost unique in Scotland, and are constructed in an entirely new manner and form. They were sunk much deeper into the ground than the techniques for sand building normally require, suggesting the builders were unfamiliar with the matrix and building techniques required for it.

There appears to have been several phases of style and building custom at the Udal. The earliest buildings were constructed during the period XIV-XI.3¹⁰ and consisted of simple oval structures (see plan I: buildings h, j, and c and the primary form of ST) with single cell 'satellites'. Slab lined hearths lay along the long axis together with a single 'sleeping' platform. A second phase of design saw the symmetrical placing of a satellite on the end of a large oval chamber 6m long. There a slab lined hearth in a long narrow floor of trampled peat ash was framed by two revetted platforms -probably 'sleeping platforms'- (n3.63 in its primary form and later phases of ST). The third and most complex phase was XI.1. This phase saw the basic 'figure of eight' heavily embellished with minor satellites¹¹ (n3.1 and z). This phase must represent the growth of a fairly successful and growing settlement around the late eighth century before it is violently and finally wiped out by a Norse invasion around the year 800.

These later buildings were flanked by extensive palisading, shown up by the remains of post holes. This palisading was very important for N3.1 as it was replaced at least 10 times and on one occasion with 30cm squared timbers -in an area likely to have been without major timber stands.

Throughout these periods, small '4 poster' structures existed 2.4m square. These show slight slab lined hearths, trampled floors and post hole settings for roof support.The perimeters of these buildings however, are rarely well defined and no walling evidence exists. The possibility of outhousing or servant quarters is the conclusion which possibly has to be made here.

The drastic change in material culture around the fourth century associated with these buildings does tend to compel an invasion theory. A key indicator of this, is the drastic change in the style and quality of the pottery associated with them. Whereas the pottery associated with the late Iron Age occupation horizon is well fired, decorated and of fine quality, that associated with the new inhabitation is of very poor quality with simple undecorated shapes. It does exist however, in enormous quantities -probably due to its quality! An indicator of new practices comes from level XII where there was a heavily grass-marked and impressed pottery horizon that can be associated with similar Irish and Cornish examples and dated to the 6th century AD.¹²

Apart from pottery, other artifacts indicative of manufacture were uncovered including crucibles and metal casting moulds. Produced from level XI.11 was a gold gilt penannular brooch pin head. From level XII iron double-pronged and socketed forks and a decorated penannular silver ring were discovered. Worked bone was produced in copious quantity with indications of a bone pin industry. There were also bone gaming pieces, buckles and composite combs -another indicator of material culture change as during the late Iron Age here, the combs used were invariably of bone only.

What does all of this evidence suggest then? Can we say, as has been argued,¹³ that what we have is an example of the baile bíataigh of Irish documentation? Or do we have to look for other possible scenarios?

It is possible to argue the case for a baile bíataigh and we may indeed have here the settlement of a noble (aire déso) of the Dalriadic Scots with his client dóer-chéli. The number of buildings at the Udal corresponds to the division for rent and tribute laying (tech) and the division for naval recruitment (ceathramh) detailed in the Senchus fer nAlban.¹⁴ In both cases the lowest number is five and it is conceivable therefore, that we may indeed have a baile bíataigh here.

This may be of course, (and is likely to be) simply coincidence. The likely and logical answer seems to suggest a settlement of a small farming community on the outer arm of the kingdom of Dalriada, although no doubt participating in its affairs to an extent. It must always be kept in the forefront of the mind that the Udal is not actually all that isolated. In terms of sea transport it is very well located. Indeed there are good anchorages close to the site itself, with future excavation possibly likely to prove that this situation had existed back into Bronze and possibly earlier periods.

The importance of the Udal is not just in terms of this particular period. The ramifications of the knowledge and artifactual evidence from here in all periods from the Neolithic until the late Middle Ages are huge and incalculable. The scope for further investigation and insight into the history of the western seaboard of Scotland is greatly enhanced by this site and it is to be hoped that full publication will soon be forthcoming.¹⁵

Chapter III

Conclusion

What conclusions can be made, from the evidence available, about the settlement on the western seaboard of Scotland between the years 300-800 AD? Logically, we should be looking for some form of clear material cultural differences between the incoming Scots and the indigenous people of the country that they invade. There is a distinct possibility however, that in many respects there were no initial differences in the material culture between the Celtic peoples Scot, Pict or Briton in the Argyll area. Distinct material cultural differences exist at the Udal, and point to the survival of a pre-Celtic population there.

It is this fact which may partly explain why it has been extremely difficult to identify and categorise early settlement sites. At present, there is no difficulty, archaeologically, in identifying whether a settlement is Bronze, early or late Iron Age. The problem we have, is that while we can identify subsequent occupation, we can't put a population group to it.

It seems that one of the principal reasons for this inability to identify, date and categorise settlement sites in this and other areas is the materials used to build the domestic and dwelling structures. It would appear that most, if not all, of the dwelling structures were constructed of stone foundations with wooden walls, uprights and supports. Few of these foundations now survive as they were robbed in antiquity to build other houses, or dykes or enclosures. Consequently, we cannot get that first 'fix' of where to look for settlement and therefore we cannot pin down settlement to any great extent.

One means of trying to pin settlement down is though place-name study. Nicolaisen has shown that the key forms that we should be looking for on the west coast are the generics, sliabh, cill, baile and achadh. The problem here is the separation of these terms from their later 9th and 10th century forms, and also the fact that in many parts of Scotland the original Gaelic or P-Celtic place-names were replaced by Norse names. We will never be able to completely say where all of the settlement sites were in Scotland, although the discovery of a few more would certainly help to shed some light on the societies of the west coast of Scotland during the early historic period.

To look at the evidence which I have produced here, it appears clear to me that there is a sequence of events which could be logically and I feel successfully put together.

Taking Argyll first, around the year 300, in the face of pressure from the the northerly migrating Uí Néill, some Scots leave Antrim and move across the north channel to Kintyre and its surrounding area. There they mix happily with the indigenous people of Celtic origin. Over the next few centuries the Scots population gradually increases their territorial area and appear to thrive through utilisation of the indigenous strongholds: the forts and fortified settlements of previous periods. The Scots gradually expand and explore along the sea routes both north and south of Kintyre. They settle in Galloway and investigate the Western Isles.

The people of Loch Glashan, situated not far from Dunadd, capital of the Dalriadic kingdom, enjoy both the protection of it and their crannog dwelling. There were few disputes to disturb them and life is prosperous. Occasionally their menfolk march off to battle in Ireland or against Pict, Briton or Angle. Success is generally with them until disaster strikes in 736 and 741 when the Picts attack and overrun Dalriada.

For the people who live around Dunadd, life is extremely prosperous. Not only do they take advantage of the wealth of the capital, through its production of jewellery, they also used the land routes to Loch Awe and the sea routes south through Loch Craignish to take their wealth to other parts of the Dalriadic empire and elsewhere. Foreign traders follow in their tracks bringing goods and pottery from Gaul and further afield, adding to the richness and diversity of life in and around Dunadd.

Religion plays an important part in the lives of the Scots of Dalriada. With the coming of Christianity to mainland Scotland through the energies of Ninian and later Columba, power, both religious and secular, was exercised and increased over the pagan peoples to the east.

The foundation of monastic communities in Pictland was a major achievement for the church in its missionary mode. Iona in particular was the base for the setting up of these monasteries, Maelrubai the abbot of Bangor used Iona as a base when he set up the monastery of Applecross in definitive Pictland while others set up communities on Rum and Eigg from early in the seventh century if not earlier. This ecclesiastical growth and the huge prestige gained from it was perhaps one of the major successes of the Columban church. One which reflected much glory on the kingdom of Dalriada.

While investigating the Western Isles, a group of Scottish settlers comes across a settlement on a peninsular of North Uist. The people of this settlement are descended from the original Neolithic settlers of the area and have formed a small thriving complex. The Scots see this and move in, either enslaving or driving off the indigenous people before settling down.

When they settle down they move away from the insert buildings of the wheelhouse and build themselves a new settlement at the end of the existing field system. These early settlers don't have a particularly good idea of how to build in sand, and so they dig their buildings in to a good depth and construct their homes in a figure of eight style brought with them from Antrim.

Over the years, this settlement thrives and, no doubt, is involved with the political situation to the south. The arrival of Fergus mac Erc in Dalriada around the year 500, sees the population become part of a new political state. The lord of the Udal settlement is no doubt subject to overlords in Loairn and they undoubtably provide military service in the form of men for seven bench warships.

The period until the late eighth century appears to have been quite good for the inhabitants of the Udal, there is slight evidence of some adverse climate change,¹⁶ but things appear to have been generally prosperous for them. Then, around the year 800 Norse longships draw through the narrows of the sound of Harris and descend upon and destroy the settlement.

These new arrivals also see the advantage of staying on the fertile machair lands. They build over the buildings of the previous inhabitants, placing walls along the long axis. To protect themselves they also construct a polygonal fort at the site and begin their own manufacturing in metals and bone. A new epoch in the history of the western seaboard begins.

A factor of settlement along the west coast which was as true in the Neolithic as it was during the Norse period was maritime communication. The ease of navigation along the chains of islands that make up the west coast of Scotland and the ease by which access to the interior of the country can be made by boat facilitated settlement.

It is quite clear that the use of the Great Glen throughout the prehistoric and into the early historic period enabled communication and trade to be carried out. The Great Glen was, before the construction of the Caledonian canal, navigable by boat for most of its way with only a small number of portages required. It has been described as the "main communication route between Scottish Dál Riada and northern Pictland"¹⁷ a route which was to be used by the Norse with even greater proficiency in later times.

Looking at the evidence which has been presented here, we can see that it is possible, to a certain degree, to detail and discover a lot about the settlement patterns and society of the western seaboard from the archaeological evidence. Yet problems present themselves. The antiquarian interest of the late eighteenth and more especially nineteenth centuries led to the discovery of many fine and beautiful artifacts. It was not however, interest in the knowledge which could be gleaned from meticulous and careful excavation of the sites that was at the forefront of many antiquarians minds, rather, fortune hunting was the order of the day.

Many of the most important sites which could have given us so much rich knowledge were, unfortunately, disturbed and in many cases 'vandalised' by these antiquarians. Had they remained undiscovered, we could, with our improved techniques and knowledge on dating, greatly increase our meagre knowledge. Instead, we are left with collections of artifacts with little real information about where and when these artifacts came from or how they came to be where they came to be.

One example of this type of situation can be looked at by studying the great antiquarian Erskine Beveridge. Much of the collection of Dark Age finds in the National Museum of Antiquities in Queen Street was bequeathed by him during the late nineteenth and early twentieth century. He bought the island of North Uist around the turn of the century, and lived on the tidal North Uist island of Vallay not far from the Udal. In addition to much excellent work on the Island he explored and documented many other parts of Scotland.

His excavation technique was quite simple; he hired labourers to dig in likely, easy to dig spots until they found something! Occasionally he would visit the site to check on their progress and draw or photograph any interesting structures or finds. If they didn't produce anything exciting after a few weeks, he would move them on to another site to try his luck there.

An example here is his work at the Udal, on US. Beveridge knew that there was a great possibility of finding a site of importance as there were many artifacts here which could be picked up as they eroded out of the sand hills, and so he set his labourers to work. In due time they picked up the outer wall of a wheelhouse and souterainn. They didn't however, produce much in the way of artifacts and consequently Beveridge abandoned the site. He left a double wheelhouse complex and bronze smithy untouched, plus several other buildings and structures with their artifacts. The knowledge now gained from the Udal would have been considerably lessened had he continued to dig here.

We must not however, place too much blame on early archaeologists. Many closer to our own time have contributed to the disastrous direction which studies of settlement, and other sites underwent. The successive digs at Dunadd were conspicuous not only for the finds produced but also for the complete lack of any scientific or orderly procedures for execution of the dig. Reference to the excavation notes show that little attempt was made to stratify the site and so gain more knowledge from it.

So, where can we take settlement study to in the future? I feel that an extensive multi-disciplinary campaign of investigation involving archaeologists, historians, onomasticians and government bodies is seriously required. It is a disgrace that we continue to accept the paucity of present evidence and blithely repeat it without any concerted attempt to investigate even known or suspected sites for further evidence. If we continue to go along the lines where sites are only investigated or excavated when development or damage is threatened we will ultimately lose much of the evidence which must be, and is, out there for us to find.

In particular, I feel that the machair and westerly sites of the inner and outer Hebrides need to have extensive field work done on them -if nothing it will indicate the state of erosion on these coasts. The Udal sites are being seriously eroded each year by the weather, and it is only the huge mass of deposits that have helped to keep the sites intact. There must be several similar sites out there which are now being eroded with little attempt to find and excavate them.

It has to be said that, despite my criticism here, much good work is being done to increase our knowledge. In particular the publication of Anna Ritchie's book on the Picts and Scots should open new avenues of investigative possibility. The full publication of the Udal evidence will again open up and show the way forward for investigation and dating of west coast settlements.¹⁸

So in conclusion, the history of settlement along the western seaboard of Scotland can be seen in terms of a maritime and climate orientated culture. the political influences on the settlement here were at first only marginal- yet severe enough to force groups of Scots to leave Antrim and set sail for Kintyre some twelve miles away.

Once there, these people -as indeed all migrating peoples did- did not settle down at once but continued to explore and settle further afield. The political upheavals of the later eighth and ninth centuries subsequently caused a new wave of migration and settlement towards the east and the rich lands of Fortriu. In their wake, the Norse changed the linguistic and political map irrevocably and dramatically.

Ewan Innes, October 28 1993

Footnotes

1. O'Rahilly T.F. Early Irish History and Mythology, pp346
2. Ibid pp344
3. This area has been chosen as it was the subject of a detailed examination both in PSAS Vol. XCV and in Vol. 6 of the Royal Commission for Ancient and Historic Monuments in Scotland series on Argyll.
4. Anderson A.O. and M.O.(ed.) Adomnan's Life of St. Columba:, pp263
5. The plan at the end of this dissertation is reproduced from vol. 6 of Royal Commission on the Ancient and Historical Monuments of Scotland series on Argyll pp153, as it provides a handy sketch guide to the layout of the various structures of Dunadd. The letters used correspond therefore to those of the plan.
6. PSAS 109 pp36
7. Allt na Sac is noted in Discovery and Excavation Scotland 1958 pp15
8. Ibid pp100
9. Anna Ritchie refers in her book on Viking Scotland to the possibility that the Udal and Buckquoy may be linked to Dalriadic settlement . She draws comparison to similar buildings in Co. Antrim and hopefully she will develop this in her forthcoming book on "Pictish and Scottish Scotland" with reference to these sites and others which have undoubtably escaped the notice of myself and other recent authors on the subject.
10. Crawford I.A. Udal Interim Reports (1972-76) The Roman numerals represent phases, decimal and alphabetic suffixes indicate levels. The phases are given the same number as their most significant level (either largest or most intact level).
11. Professor Martin O'Mhurchu has suggested that the major satellite cells of phases two and three may represent the Iomaidh of early Gaelic literature (the inner sanctum where elders sat and observed the household at a discreet distance.) Ibid 1973 pp7
12. Thomas A.C. Britain and Ireland in Early Christian Times. (London 1971)
13. Crawford I.A. 50 Centuries pp13     Interim Report (1975) pp1 & 7
14. Bannerman J. Senchus pp132-141
15. See below fn pp4
16. Sand blow and erosion evidence not really noticeable pre AD 300. Crawford I.A. Interim Reports (1973) pp4
17. fn Crawford B.E. Scandinavian Scotland pp228. Alcock L. et al., Excavations at Urquhart and Dunotter Castles: Interim Report (Dept. of Archaeology, University of Glasgow 1985) pp2
18. When this will occur is a moot point. The volume of the material is extremely vast and is posing problems of finance and time to ensure it is published within the next few years. There is the possibility that it will never be published and will fall into an archaeological limbo. We must hold out in hope that this never happens. I must however, thank Ian for his generosity in offering me the opportunity to study the artifactual evidence from the Udal. Unfortunately due to severe financial constraints I was unable to take him up on this offer. The opportunity to review and study this evidence must therefore, wait for either another time or another person.

Forts, Duns and Settlement Sites for Map II

Bibliography