9.1 Introduction

1. The Early Medieval period is the term used to cover the centuries between AD500 and AD1100. Although archaeology is contributing more and more to the study of this period, we are still very much reliant upon literary sources when creating an overall framework. These suggest that it was a time of tremendous upheaval and unrest. The withdrawal of the Roman army from Britain left a power vacuum, and into this void came an influx of people from across Europe. Research, informed by contemporary literature and often supported to some degree archaeologically, has given us some insights into the character of these immigrant populations, and their likely places of origin. The literary sources refer to the movement of Danes and Norse peoples from the area now occupied by modern Scandinavia, Angles and Saxons from continental Europe and Gaelic-speaking Scots and Irish from Ireland.

2. Closer to home, there were movements of Picts southwards from northern Scotland. All this turmoil impacted upon an indigenous population of Brythonic-speaking ‘Celts’ whose society – in South West Scotland – had remained largely unchanged by repeated Roman incursions but whose numbers must also have ebbed and flowed as communities forced from their lands further afield may have sought safety elsewhere. In South West Scotland, amongst the earliest recorded arrivals is an Influx of Irish who reportedly arrived after 400AD. They are believed to have settled in the Mull of Galloway, to be followed by Irish missionaries who left their distinctive pecked crosses at Ruthwell, Nithsdale and Staplegorton, Eskdale.

3. It is believed that – following the Roman withdrawal – the local communities coalesced into kingdoms. South West Scotland incorporated part of the Kingdom of Rheged, the location of which has long been open to debate. Some authorities suggest that Rheged was centred upon Carlisle, while others argue that it could potentially have been located in Galloway, with its centre of power at Dunragit (which can be translated as the ‘Castle’ or ‘Stronghold’ of Rheged’’) (Watson 1923-4). To the north of Rheged lay the kingdom of Strathclyde, encompassing much of the lower Clyde valley with a political centre at Dumbarton.

4. The political picture is still unclear and open to debate amongst historians and archaeologists alike. The exact extent of Rheged and Strathclyde – both spatially and chronologically – remains uncertain, and it has been argued that the area now encompassed by modern Ayrshire was occupied by several small regional powers. Further supporting this interpretation is the fact that the modern subdivisions of Ayrshire – Carrick, Cunninghame, and Kyle (corresponding respectively to contemporary South, North and East Ayrshire) – can be traced back to documentary sources of the Early Medieval period, as does the placename ‘Ayr’ (Driscoll and Forsyth 2004).

5. It is entirely possible that Rheged extended north to encompass at least what is now modern Galloway. It is also possible that the northern part of the region, including North Ayrshire but also potentially some land further south, would at some point have fallen within the Kingdom of Strathclyde. At around AD 670, the fort of Alt Cluit (Dumbarton), which formed the centre of power for the Kingdom of Strathclyde, was sacked and ethnically cleansed by a combined force of Gaels and Norsemen. The native Brythonic speakers who escaped fled to Wales, where they settled. This coincided with a period when the Picts and Gaels moved southwards, potentially settling as far south as the Tyne-Solway line. By the 6th century AD, however, the Angles of Northumbria were moving northwards and westwards. During the 7th century AD, and throughout the eighth, Dumfries and Galloway would have been under Anglian rule. Norse incursions followed, which again appear to have resulted in settlement. Throughout this period, the level of complexity inherent in these interactions should not be underestimated. Scots religious teachers moved to settle in Northumbria, Irish scholars came to study at Whithorn, creating a pattern as complex as the interlacing designs which grace the sculptured crosses which are so characteristic of the time. The echoes of this complexity can still be found in modern place name evidence throughout the region.

6. Traditionally, archaeologists have used the literary sources as a means through which they can define specific ethnic groups through combinations of material culture forms. These might include carved stones, ornamental metalwork types and even architectural traditions. The resulting groups can then be mapped across time and space, often reinforcing (but occasionally challenging) the evidence gleaned from literary sources. This has the potential to create a situation where a nuanced and complex picture is lost beneath the broad-brush narrative which inevitably results from the patterns of differential survival encountered by archaeologists. It is not within the scope of this document to challenge the resulting terminology. Rather, it remains expedient to adopt it as a series of descriptive labels which recur within the literature. We will therefore utilise long-established terms like ‘Norse,’ ‘Anglian,’ ‘Irish,’ ‘Pictish’, etc without necessarily assuming that the people who built, acquired, used and lost or discarded particular material culture forms were associated with a specific ethnic group.

7. It was during the Early Medieval period that Christianity made its first appearance within the region. Early churches and religious houses were established: Whithorn appears to have been the first of these, with evidence of Christian religious practice dated to around the mid-5th century AD through inscribed grave markers at the site. The gospel was preached in Northumbria by migrant Scots, which meant that the influx of Northumbrians and Anglians who entered the region by the 7th century – and who ultimately formed its ruling class – were also Christian worshippers. Another wave of settlement has been postulated during the 9th or 10th centuries, largely through linguistic and place name evidence. This time it was an influx of Norse – or more accurately Hiberno-Norse – from Ireland. The name ‘Galloway’ is thought to have derived from a translated form of the term ‘Stranger Gaels,’ and while there is also ample evidence that these Hiberno-Norse incomers also worshipped Christianity, some pagan traditions appear to have continued in the form of the Norse boat burials identified at Kingscross and Millhill on Arran.

8. There is evidence of some degree of continuity from the preceding Native Iron Age, in particular through the continuing use (or perhaps reuse) of hilltop forts and some monumental roundhouse sites (in particular crannogs). While the crannog structures appear similar to their earlier Iron Age counterparts, the hilltop forts tend to be representative of the ‘nucleated fort’ type. Here, summit areas are typically enclosed in a complex arrangement of ramparts which subdivide the sites into smaller sections which often correspond to individual hilltop terraces. Important examples of nucleated forts within the region include Mote of Mark and Trusty’s Hill. These sites often show an earlier phase of native British occupation, followed by later layers which reveal material culture finds associated with the Anglians or Northumbrians. This may infer that the same site was reused as a centre of power by a later polity imposed on the indigenous communities by an invading group. Elsewhere, an Anglian presence has been inferred by the presence of a structural tradition quite alien to the roundhouses of the native Iron Age. This is the rectangular long house associated with the Anglian and Northumbrian traditions, an example of which has been excavated at Lockerbie Academy.

9. The presence of a political elite is attested by the often-impressive pieces of decorative metalwork recovered across the region. Sometimes these are modest items like the Whitby-type strap-ends – with their animal interlace decoration – which have been found on potential ‘beach-market’ sites like Stevenston Sands and Luce Sands. Occasionally this metalwork can take the form of impressive and potentially unique pieces like the Hunterston brooch. The presence of these items suggests that elite members of society were supporting skilled craftsmen who were capable of producing fine pieces of metalwork which could potentially be offered as gifts to cement alliances or offered to members of a retinue. Mould fragments derived from the production of these distinctive items – often manufactured in a distinctive regional style – have been found at nucleated fort sites like Mote of Mark. Elite status could also be inferred by the acquisition of exotic items sometimes sourced from distant places as far away as the Mediterranean. The ecclesiastical equivalents of the secular elites were the Bishops, who resided at sites such as Whithorn. Finds of Gaulish pottery and sherds of amphorae (inferring the consumption of wine) confirm that these individuals were linked in with wider trading networks, and again they could have used the acquisition of this exotic material to build alliances amongst the local political elites.

10. Certain items of decorative metalwork – in particular the fine penannular brooches, thistle-headed pins and intricately-decorated strap ends – appear to be associated with the secular realm. There is also ample evidence that the early Christian church acquired a wide range of exquisitely decorated items for religious use. In addition to the carved stone crosses, which are often items of considerable artistic merit which fall within distinct regional ‘schools’ of workmanship, a range of items have been found, composed of gold and sometimes incorporating semi-precious stones such as rock crystal. The Balmaghie hoard (also known as the Galloway Hoard) produced a large number of gold and metal objects, including some clearly linked with ecclesiastical use. These objects included a pectoral cross, worn as part of ecclesiastical vestments, and a pointer formed into the shape of a bird which would have been used as an aid when reading from the gospels.

11. Monks within monastic communities like Whithorn would have been literate and would have been capable of producing their own volumes of Christianity’s sacred texts. The resulting volumes may also have been lavishly illustrated using similar artistic styles to those seen in the Lindisfarne Gospels and the Book of Kells. The bird-shaped pointer would have been used in conjunction with such texts, but of the holy books themselves we now have no surviving examples from the region. The intervening centuries have no doubt robbed us of a large number of such treasures, and it is only very rarely that we encounter surviving examples. Sometimes items like reliquaries have been safeguarded within the church, but on other occasions these objects have only survived because they were stolen from the ecclesiastical establishment and hidden for later recovery by raiders who never returned to retrieve them. Such a scenario probably explains the massive and eclectic collection of gold and silver found near Balmaghie.

12. The earliest references to discoveries of Early Medieval sites and artefacts show that their origins and character were not always necessarily understood or viewed sympathetically. In 1684, reference was made by an authority named Sympson to a ship burial found in either Wigtownshire or Galloway which revealed copious numbers of ‘copper’ nails. Sympson’s account was subsequently quoted by Rev C H Dick, who in turn was referenced even later by William Collingwood in the early 1900s (Collingwood 1919-20, 114). Although not interpreted as such at the time, we can now infer that this discovery is likely to represent a Norse ship burial.

13. Unfortunately, records also show that early interactions with the region’s Early Medieval monuments were not altogether positive. In 1642, an Act of the General Assembly of that year was issued ordering the sculptured stone known now as the Ruthwell Cross to be cast down. Prior to the Reformation, it had stood within the parish church, but because it was inscribed with artwork and inscriptions, it was deemed idolatrous by the Post-Reformation church (Anon 1865-6, 12).

14. Although broken, the Ruthwell Cross survived in its entirety, and lay where it had fallen. It was noted by the antiquarian Thomas Pennant when he visited the church in 1772, and a later minister of the parish, the Rev. Henry Duncan, had the piece restored and erected at the churchyard in order to preserve it (Anon 1965-6).

15. In the autumn of 1826, an ornate metal brooch was found at the foot of a cliff near Hunterston. Described half a century later as ‘Scotland’s richest fibula’ by George Stephens (Stephens 1866-8; Stephens 1878, 76), it was made of silver, embellished on front and rear surfaces with gold filigree work in a late Celtic style, and inset on the front surface with amber. As well as being a beautiful and exquisitely crafted object in its own right, it attracted ongoing antiquarian interest on account of the Norse runes inscribed (in two different hands, according to Stephens) on its rear. Stephens transcribed and translated the inscription as follows: ‘Malbritha owns (or possesses) this brooch, Speaker (or Lawman) in Lar (unknown place)’ (Stephens 1878, 77).

16. The brooch has remained a focus of interest and fascination ever since. Not only does it represent the final flourishing of an art tradition that began much earlier in the Iron Age (see Chapter 7 Iron Age), but it also embodies the cultural melting pot that South West Scotland potentially became during the Early Medieval period. The piece itself showed Celtic and Scandinavian influences; the runic writing was Norse, but the name ‘Malbritha’ appeared to be Celtic.

17. Studies in the 1860s focussed largely on the literary tradition. W F Skene Identified Galloway – and in particular the area around Hoddom – as an area settled by the Frisians, who might be grouped around the early Saxon tribes (Skene 1860-2). Some fieldwork was also being undertaken, with a description of the surviving remnants of the linear earthwork known as the Deil’s Dyke included in one of the earliest volumes of the Transactions of the Dumfries and Galloway Natural History and Archaeological Society’s (the Transactions). This extensive monument, which is now thought to have potential associations with the Early Medieval period, was referred to in an annual address given by Vice-President James Starke (Starke 1866-7a), following fieldwork undertaken by Vere Irving (Starke 1866-7b). The same address contained a broad summary of the various types of field monuments and historic buildings encountered throughout Dumfries and Galloway. Some were already recognised as Early Medieval in date, in particular the sculptured and incised crosses. These included the Ruthwell Cross, with its runic inscription (deemed Anglo Saxon in origin), as well as fragments recovered from the churches of Hoddom and Dunrod (Starke 1866-7c).

18. The Proceedings of the Society of Antiquaries of Scotland’s (the Proceedings) of 1870-2 included a report by Arthur Mitchell, highlighting several cross-inscribed stones occurring at the burial ground of Kirkmadrine, Stoneykirk. One of these stones bore a Latin inscription which he ascribed to the period spanning the 5th-10th centuries AD (Mitchell 1870-2), while another example survived ex situ, built into the wall of a mill building at Drummore. In the following edition of the Proceedings, George Wilson highlighted a similar collection of sculptured stones in Wigtownshire, centred this time at Laggangarn (Wilson 1872-4). Seven remained in situ, with a further four removed for later re-use, either as lintels or as gateposts.

19. Initially, an important focus of study was early Christian sites, often flagged up through local traditions which potentially dated back centuries to the pre-Reformation period. Early church and chapel sites were important, but so too were holy wells and caves. The latter might be attributed to a saint or hermit who had allegedly used these natural features as a retreat during the early Christian period. Archaeological investigations of cave sites were recorded from the 1880s onwards, but it was not guaranteed that a link with early Christian worship or ritual practice could be confirmed.

20. Investigations would take place in several cave sites during the 1880s. Most revealed evidence of multi-period occupation, potentially dating back to at least the Iron Age. Excavations at Cleaves Cove, Dalry, proved inconclusive (Smith 1889), as did investigations at another cave in Galloway which had a reliable early Christian attribution and surviving structural elements: St Medan’s Cave and Chapel, Kirkmaiden (Trotter 1889). Reference to Gothic arches and vaulting suggested a possible Medieval attribution to the chapel structure, with nearby wells acknowledged as ‘holy wells,’ and used for votive deposits of coins and other items until relatively recent times. Robust evidence of Early Medieval use, in the form of inscribed crosses or sculptured crosses, was absent.

21. By contrast, excavations which took place at St Ninian’s Cave, Glasserton, near Whithorn produced plenty of evidence for early Christian worship during the Early Medieval period. Initial explorations carried out by a local gamekeeper revealed several incised crosses and cross-incised stones (Johnstone 1882-3). Later excavations on the site, carried out by Herbert Maxwell, also revealed evidence for episodes of occupation, punctuated by falls of debris from the roof. Occupation extended over a considerable period, commencing prior to the use of the cave during the Early Medieval period (Maxwell 1886-7, 138). Also revealed were the remains of a stone stair descending into the cave.

22. Maxwell presented a detailed record of finds recovered from the site, along with illustrations of the various incised crosses and cross-incised stones. His reports were published in the Proceedings and also in the regional journal, the Ayrshire and Galloway Collections (the Collections) (Maxwell 1884-5; Maxwell 1885). Further works on the site revealed a magnificent example of an incomplete sculptured cross with ‘Celtic’ interlaced decoration and a Runic inscription (Maxwell 1889). In a subsequent volume of the Collections, a useful inventory of sculptured stone finds from Wigtownshire was included. (Black 1894). Amongst their number were various Early Medieval cross-inscribed stones and sculptured crosses, all found from Galloway, and most from the vicinity of Whithorn. All formed part of the collections in the National Museum of Scotland.

23. Runic inscriptions were reported in St Molio’s Cave, Holy Isle, Arran (Rogers 1884-5), while the end of the same decade saw John Romily Allen compile an inventory of sculptured stones bearing ‘Celtic’ decoration then found in Scotland (Romily Allen, 1889-90). His list included 17 examples from the area now covered by modern Dumfries and Galloway, a single specimen from Ayrshire (Mansfield House, near Cumnock), and another single specimen from King’s Cave, Arran. All fell within his Class 3 group, which comprised ‘crosses, cross-slabs or recumbent coped stones with no symbols.’ His inventory included just one single outlier which did not fall within his Class 3 group. This was the rock surface at Anwoth, Gatehouse of Fleet, with its inscribed Pictish symbols, located at the site we know as Trusty’s Hill.

24. The 1883-6 volume of the Transactions included a paper by George Black which detailed the history of the Ruthwell cross and presented an equally detailed description of the object, including its imagery and inscriptions (Black 1883-6). Black also discussed the Runic inscriptions on the piece, arguing for an Anglo-Saxon origin.

25. The following year, James Starke discussed the Early Medieval crosses from Kirkmadrine, placing them no later than the 7th century AD. This meant they predated the creation of the Ruthwell Cross (Starke 1887-90). In 1892, another important piece of Early Medieval sculpture was found during grave-digging operations at St Bryde’s Chapel, Lamlash. This comprised a sculptured cross, with a representation of the Crucifixion on one face (Landsborough 1896-7). Another fragment bearing interlaced decoration was recovered in the same decade from a farm in Upper Nithsdale, near New Cumnock (Hewat 1899-1900). At around the same time, similarly early items were identified by George Irving in the grounds of the Old Church, Hoddom. Here, an object originally interpreted as a font was reinterpreted as the socketed base for a free-standing cross. Traces of decoration were also identified on a grave slab, suggesting the re-use of a much earlier object (Irving 1896-7). The decade – and indeed the century – ended with James Barbour arguing – countering theories put forward by various internationally renowned authorities – that the Ruthwell cross could have originated in its immediate locale. The distinctive hard red sandstone from which it had been wrought was, he argued, of similar character to rocks occurring nearby (Barbour 1899-1900).

26. It was during the 1880s that the first investigations took place at an ecclesiastical site which would come to dominate archaeological discourse for decades: Whithorn. This initial work was undertaken by the architect William Galloway under the patronage of the 3rd Marquis of Bute and continued into the 1890s. Galloway’s interest in the Early Medieval period had already been confirmed through a contribution to an earlier volume of the Collections. His paper focused on Early Christian remains in Ayrshire (Galloway 1882), presenting a detailed account of free-standing cross-bases recovered from Whitehill, Dailly and Colmonnell. A short final report on the Whithorn investigations – including a site plan – was subsequently prepared by the architects David MacGibbon and Thomas Ross, better known for the multiple-volume inventories devoted to the ecclesiastical and castellated buildings of Medieval and Post-Medieval Scotland. Galloway’s report was then published in Volume 10 of the Collections.

27. Over a century later, however, the site archive from Galloway’s excavations was subject to further research by Derek Hill, allowing a more detailed understanding of the findspots of some of the carved stones. A further summary of the surviving correspondence followed, compiled by Julia Muir Watt, in the 2001 volume of the Transactions (Watt 2001).

28. By the 1870s, antiquarian knowledge had embraced the Three Age system advocated by Thomsen in 1843, which advocated the subdivision of prehistory into the Stone, Bronze and Iron Ages (Thomsen, 1843). Against this backdrop, researchers still grappled with the complex array of field monuments that still survived within the landscape, but which were being increasingly threatened by development and improvements in both urban and rural environments.

29. One form of prehistoric settlement site that became a particular focus of interest in the 1870s and 1880s was the crannog. Crannog sites occupied either marine or riverine locations and comprised circular house structures built upon a raised mound. The mound could be entirely of artificial construction, or it might occupy a modified natural feature. These crannog sites appeared to originate in the Iron Age (see Chapter 7 Iron Age), but even at this early stage it was generally recognised that both the Scottish and Irish crannogs might also be occupied in the Early Medieval Period, and sometimes even later still, into the Medieval period (Munro 1880).

30. Robert Munro explored a number of crannogs in South West Scotland, identifying their distinctive mode of construction. This often took the form of brushwood mounds pinned and held in place by larger pieces of timber. Sometimes these timbers showed the use of skilled carpentry techniques. His reports, published variously in the Proceedings and the Collections, included detailed drawings of a diverse range of material culture, including iron tools and metalworking debris like moulds and crucibles, but only rarely do we see objects that can be specifically assigned to the Early Medieval period. Bone combs from Buiston (Munro 1882) and a jet pendant from Lochspouts represent good examples of such objects. This circular jet object is subdivided into quadrants, forming a rectangular cross in the interior. Commenting in Munro’s paper, Joseph Anderson suggested an Early Christian origin for the object (Munro 1884, 15). This would place this object within the Early Medieval period, an attribution supported by the use of repeated incised circles as a decorative motif.

31. Although crannogs dominated the literature throughout this period, there was increasing recognition that some of the region’s many defensive hilltop sites, classed variously as forts and duns, might also have been occupied during the Early Medieval period. At this stage, however, this association was often made through place name or traditional evidence. Authors who wrote widely on the topic of early fortified sites in the region included David Christison, Frederick Coles (Coles 1890-1; Coles 1892-3), and John Smith. Each compiled inventories dedicated to specific areas. Christison, for example, published an inventory of sites in Upper Annandale in the 1890-1 edition of the Proceedings (Christison 1890-1) and an inventory of their counterparts in Ayrshire in the following volume (Christison 1892-3).

32. It would be half a century or more before the intricacies of these sites’ time-depth could be better understood through widespread excavation and radiocarbon dating. It was, however, already being recognised that different classes of fort could be differentiated by variations in location, plan and layout, defensive form, as well as other attributes. For the time being, place name evidence often provided the starting point. Christison cited numerous sites which included either the word ‘mote,’ or a variant, which could infer Anglo-Saxon origin (Christison 1892-3). Another potential example was put forward by George Wilson: Dunwick, a Galloway fort attributed by local tradition to an unknown Viking who drew up on the shore nearby (Wilson 1885 69).

33. Another interesting but at the time misinterpreted discovery reported in the 1880s was that of six inhumations from Dounan, near Ballantrae (Moore 1880-1; Moore 1885). Found in association with a shell midden, the six individuals were originally thought to date to prehistory, perhaps even the Mesolithic (see Chapter 3 Palaeolithic and Mesolithic). It was only after the advent of radiocarbon dating that an accurate date could be ascribed to them. Their burial outside consecrated ground has led to the theory that they were victims of a shipwreck, buried close to where they were washed up on the shoreline.

34. Another site of ambiguity was the area of sand dunes around Luce Sands, which became a focus of interest for antiquarians for the numerous artefacts recovered there. These spanned a long period of time, dating from prehistory onwards. Along with the items catalogued and illustrated by George Wilson in 1881 were a bronze penannular brooch and an ornate bronze mount, both of recognisably Early Medieval form (Wilson 1880-1).

35. In 1903, an important summary of the then-current levels of knowledge was presented by Romilly Allen and Joseph Anderson, in a volume entitled the Early Christian Monuments of Scotland (Allen and Anderson 1903). Later that decade, in 1909, James Balfour was actively engaged in the excavation of two important Early Medieval sites on Arran. The first was a mound at Kingscross, which revealed an oval stone setting which contained calcined human bones. The presence of numerous iron rivets and a fragmentary sword in association with the deceased, inferred the burial place of a Norse or Viking seafarer. Following their death, they were cremated aboard a boat or ship, a practice known from the Viking sagas (Balfour 1908-9a). It was even possible to establish a rough date for the interment as a coin was recovered: a styca of Wigmund, Archbishop of York, minted between AD 834 and 851.

36. The other site which drew Balfour’s attention was Holy Isle, where he excavated the reported former site of a monastery. He also investigated St Molio’s Cave, Holy Isle, where the eponymous saint was believed to have had his cell (Balfour 1908-9b). The interior of the cave was paved and accessed via a short flight of steps. There was evidence for a fireplace, which led Balfour to suggest that the cave had been separated into two areas, one for habitation and the other for religious devotions. A dark midden deposit rich in marine shell and animal bone had accumulated over the paved floor.

37. Balfour then turned his attention to the excavation of a site at Kilpatrick, Arran, which he described as a ‘cashel’ and ascribed to the Early Christian (that is, Early Medieval) period. Its assignment as such was based on its typological similarity with Irish examples known to be associated with the ‘Celtic’ Early Christian church. The site comprised a thick-walled circular structure with intra-mural chambers which adjoined a polygonal enclosure. Several roundhouse footings were identified in the vicinity, outwith the enclosure (Balfour 1909-10). Interestingly, Balfour encountered a Bronze Age urn burial during his investigations; the builders of the ‘cashel’ wall had encountered the cist cover and built the later structure incorporating the prehistoric monument.

38. That same year, Balfour was given the opportunity to examine several items recovered during the removal of a gravel mound from Millhill, Lamlash over a decade before, in 1896. He identified an iron shield boss and fragments from an iron sword blade, inferring that the mound had – much like the recently-excavated example from Kingscross – housed a Norse or Viking burial, potentially another ship burial (Balfour 1909-10b).

39. It was during the same decade that the Rev Philip Robertson presented a talk to the Dumfries and Galloway Society which covered the topic of the Early Christian sculptured stones of Kirkmadrine. As well as discussing the stones’ iconography, inscriptions and original purpose, he reported how – in recent decades – the ruins of the former church building at Kirkmadrine had been restored as a monument. The two surviving stones were taken into the care of the Ministry of Works, while a third, unfortunately, remained lost (Robertson 1908-9). The talk was later published in the Transactions, in the pages of which – the following year – another paper devoted to an important piece of Early Medieval sculpture, the Ruthwell cross, was published by John Dinwiddie. As well as summarising the object’s turbulent history since the Reformation, Dinwiddie ascribed it a c.AD 680 date and talked at length about how it represented ‘a sermon in stone.’ (Dinwiddie 1909-10, 118).

40. The following decade began with a report in the Proceedings detailing the discovery – or more probably, the rediscovery – of a cross-shaft in the garden of Hastings Hall, Moniaive. The cross was known to have been relocated from an earlier location in Stroanfreggan, in the neighbouring parish of Dalry, but further information relating to its provenance was unfortunately lost (Corrie 1911-12). The cross-shaft may have corresponded with one first noted by Frederick Coles over a decade earlier. The author, J M Corrie. also tracked down another item initially recorded by Coles: the Auchenshinnoch Cross, which comprised a cross-incised slab re-used as a property marker in 1734.

41. Then, in May 1912, Herbert Maxwell was alerted to the discovery of a mixed hoard recovered during peat digging at Talnotrie, near Kirkcudbright. The hoard was small, and dominated by metalwork of mixed character, with a lead weight, an ornamental strap-end, two silver pins and a gold finger ring listed amongst the contents. A piece of glass, two claystone whorls and a piece of worked lignite were amongst the other items present, along with 12 coins of Anglo-Saxon or Northumbrian character (Maxwell 1912-3). In this particular instance, the original finder had the presence of mind to note that the objects had all been found at the interface between the peat and the underlying clay.

42. The same volume of the Proceedings included an account of excavations undertaken in 1907 on the site of St Cuthbert’s, Girvan (Morris 1912-13). The works were carried out by James Morris, in conjunction with a local man, Alexander Johnston, who was in the process of restoring elements of what had been an overgrown and neglected churchyard (Maxwell 1912-3). Amongst the items uncovered was an incomplete free-standing cross featuring a pattern of ‘Celtic’ interlace. Assuming the cross had not been relocated from elsewhere, its presence suggested an Early Medieval origin for the site.

43. The year 1913 saw an important milestone towards reaching an understanding of the region’s Early Medieval settlement sites when Alexander Curle undertook excavations on the Mote of Mark, a vitrified hillfort near Rockcliffe. The site had several phases of occupation, suggestive of occupation during prehistory (Chapter 7 The Iron Age). It was, however, its assemblage of Early Medieval finds that provided one of its most important contributions. In particular, it revealed evidence for metalworking in the form of crucibles and broken mould matrices. These had been used to create non-ferrous metalwork in a variety of styles, ranging from simple items like pins to ornate penannular brooches decorated with rich ‘Celtic’ style interlace-type decoration. There was, in addition, evidence for contemporary occupation in the form of fragments of glass, as well as iron objects, worked bone and pottery. Not only did this evidence suggest that the site had been inhabited during the Early Medieval period; it also inferred that the site was of high status, since it had housed a craftsman capable of turning out ornamental metalwork of the highest quality (Curle 1913-14a).

44. Another potentially important Early Medieval find highlighted by Alexander Curle in the same year was that of a double-edged iron sword from Torbeckhill, presented to the National Museum by the finder, John Bell of Torbeckhill. The sword had apparently been found in isolation, during quarrying. Curle noted that a similar sword had been found during the excavation of a burial in Norfolk, and had been dated, on the basis of its down-curving hilt plate, to the 10th century AD, or later, through comparison with examples recovered from Swedish graves. This led Curle to describe the object as a ‘Viking Sword’ (Curle 1913-14b).

45. Evidence for occupation at a similarly late date was found a few years later at Castlehill, Aitnock, by John Smith, who excavated the site. Although the finds recovered there did not appear to match those from the Mote of Mark in either quantity or levels of ornamentation, the presence of a small penannular brooch again confirmed that this site, too, had been occupied during the Early Medieval period. This inferred either continuity of occupation, or re-occupation at a point which post-dated the Roman occupation of the region (Smith 1918-9).

46. Meanwhile, the regional journal – the Transactions – was publishing a re-appraisal of the Ruthwell cross by art historian-turned-archaeologist William Collingwood. Though Collingwood emphasised the unique nature of the sculptured cross, he placed it nonetheless in a wider Anglian series. He proposed a later date than that previously suggested by Dinwiddie, ascribing it an origin in the 8th century AD (Collingwood 1916-18a). In the same volume, G P Robertson presented a short account of a cross-inscribed stone from Kirkmadrine, previously recorded as lost (Robertson 1908-9) but recently rediscovered (Robertson 1916-18). Robertson’s report was then followed by a short note from Collingwood, who talked in more detail about the character and iconography of the piece (Collingwood 1916-18b). Collingwood wrote a further contribution for the Transactions of 1919-20, this time compiling a summary of evidence for a Norse presence in Galloway, using literary sources and place names. The monumental evidence, represented mainly by carved stone crosses and cross-inscribed stones, was – Collingwood suggested – much more representative of Anglian workmanship. It was noted, however, how one stone cross-shaft from Whithorn bore a style of plait seen only on Norwegian metalwork (Collingwood 1919-20, 116), suggesting some degree of cultural cross-fertilisation.

47. More fragments from Early Medieval sculptured crosses were discovered and reported in the 1920s. The first was reported to the Proceedings of 1920-1 by Herbert Maxwell (Maxwell 1921-1). Maxwell had been the owner of a farm and steading named Kirkland, which occupied a former chapel site recorded at Longcastle. The chapel and burial ground had been ransacked for building material a century earlier, but the fragmentary cross-slab escaped notice, having been incorporated into a later grave. It was eventually unearthed when Maxwell’s tenant was hunting for a suitable stone to create a lintel for a pigsty, and the find was subsequently reported. A few years later, the Rev. R S G Anderson reported several new finds from the Machars. These comprised firstly two crossheads, the first found at Brighouse, three miles north of Whithorn, and the second from Elrig, Mochrum (Anderson 1922-3). R Anderson subsequently reported a previously unknown incised cross from within St Ninian’s Cave, Glasserton, forming an outlier to a group previously noted by Herbert Maxwell, another unrecorded cross-slab from Barscore, Kirkcowan (Anderson 1924-5), and a cross-incised slab built into the wall of a byre at Boghouse, Mochrum (Anderson 1926-7).

48. In the midst of this flurry of reporting, William Collingwood presented a lecture to the Dumfries and Galloway Natural History and Archaeological Society – subsequently published in their Transactions – which presented a synthesis of Early Christian sculptured and incised crosses. He discussed these objects in their context as creations post-dating the Roman occupation of Scotland and placed the known examples into a chronological scheme. For the Kirkmadrine crosses he ascribed a date of around AD 600, suggesting that Anglian clerics and craftsmen were active at Whithorn throughout the 8th and into the 9th century AD. From Whithorn, the influence of their work then passed to the Isle of Man.

49. Collingwood identified Northumbrian-style crosses from St Ninian’s Cave and Kirkmadrine, with a later transition to a Scottish style evidenced by a cross at Kilmorie, Corsewall House, an Anglo-Cumbrian style from crosses identified at Whithorn, and St Ninian’s Cave. Each could be differentiated by the detailed character of their plait-work. He identified, in addition, a ‘Whithorn School’ of carving, characterised by its disc-faced head and a limited, though distinctive, range of treatments used in the plait-work. Other examples consistent with this ’School’ came from Kirkmaiden, Glasserton and Elrig, Mochrum. Later 11th century crosses, he argued, showed more kinship with indigenous Scottish traditions originating in Argyll, with examples from Glenluce and Minnigaff. A selection of ‘Viking’ crosses provided the conclusion to a report which provided a thorough discussion of the material, as well as a comprehensive and copiously illustrated inventory of known finds to date.

50. In a slightly later volume of the Transactions, William Watson presented an account (again, reproducing a lecture he gave to the Society) which summarised the known evidence provided by various historical sources, commencing with Ptolemy in the 2nd century AD and ending with land grants of the later Medieval period. In this lecture, he drew together place name evidence which supported his conclusions. These included the argument that the place name ‘Dunragit’ suggested that the Brythonic Kingdom of Rheged extended as far north as Galloway, and that a Scottish element had entered Northumbria, through the spread of Celtic Christianity, prior to the Northumbrian expansion in the 7th and 8th centuries AD (Watson 1923-4). That same year, R J Reid published a paper in the Transactions discussing the parish of Whamphray, where a Norse Christian burial-slab had been built into the walls of a later parish church. Amongst the topics he considered in this paper was the Early Medieval history of the parish, and in particular the likely location of the early ecclesiastical site which might once have housed the burial-slab (Reid 1923-24a).

51. Reid also included an account of an ill-fated excavation at St Medan’s Cave (Reid 1923-4b). Following the removal of modern infill to a depth of six feet, a female statue – believed to have represented the figure of St Medan herself – was recovered. The works were left in abeyance for the weekend, but on the team’s return they found that, ‘A band of evil-disposed persons, believed to have come from Stranraer, augmented by others from Kirkmaiden, invaded the chapel, smashed up everything they could, and threw all the “finds,” including the laboriously reconstructed figure, into the sea’ (Reid 1823-4b, 182). It was thought that local rumours of a ‘treasure’ lurking in the cave may have been the reason behind this despicable act of vandalism. The excavations Reid described did not, however, confirm an Early Medieval occupation or use of the cave, although the tradition of Saint Medan certainly seems consistent with other Early Medieval cave sites throughout the region. Instead, although the ‘chapel’ remains he examined were definitely pre-Reformation in date, they seemed more likely to be of Medieval origin.

52. In 1925 another paper authored by William Collingwood was published in the Transactions. This account formed a counterpoint to earlier works which had focused primarily on Galloway, focussing instead upon the south-eastern part of the region: Dumfriesshire. Using literary sources, Collingwood examined evidence for the life of St Kentigern, exploring the nature of Northumbrian dominance within a social framework which was still essentially Brythonic. He also emphasised, once again, the important role of Whithorn in the regional context.

53. Collingwood highlighted the contrast between Galloway, which featured a large number of monuments, primarily cross-incised stones and sculptured crosses, and Dumfriesshire. Here, known physical evidence began with the Ruthwell Cross, which was crafted at a relatively late date (late 8th century AD) compared with its counterparts to the west. Crosses recovered elsewhere, at Hoddom, Whamphray, Closeburn, Durisdeer, Glencairn, Kirkconnel, Penpont and – reportedly – at Crawick, near Sanquhar, originated later, perhaps significantly so. However, given their small numbers, it seemed likely to him that they derived from much smaller establishments than that represented by Whithorn, or ‘Candida Casa’ as it was known in the historical sources. That further evidence might remain as yet undiscovered at Hoddom seemed to him entirely possible (Collingwood 1924-5).

54. In summary, Collingwood argued that between AD 800 and AD 1100, the early church in ‘Dumfriesshire had a continuous existence, and that it preserved its Northumbrian character very much like the contemporary church at Whithorn’ (Collingwood 1924-5, 60). However, he proposed that the native British church which preceded it had been lost because – unlike St Ninian’s foundation at Whithorn – it had opposed the Anglo-Roman reforms put forward by the Northumbrian church during the period the two churches first came into contact with one another.

55. The following year, Collingwood joined forces with R J Reid to report on the recent discovery of a cross-inscribed stone, dated to the late 11th or early 12th century AD, at Ardwall Island, near Knockbrex (Collingwood and Reid 1925-6). Judging by its size, they reckoned that it may have functioned as the headstone of a grave, which inferred the presence of an Early Medieval chapel and burial ground on the island. In a later report of a field meeting at Trusty’s Hill, Anwoth, Reid described the site as a vitrified fort likely to be contemporary with the recently excavated Mote of Mark, that is, likely to have been occupied during the Early Medieval period (Reid 1926-8). He also drew attention to the Pictish symbol carved on a rock outcrop on the site, the only example of its kind known in Galloway. The same decade saw a report on the early church site dedicated to St Conal at Kirkconnel, which had revealed a socketed cross-base and various sculptured stones ascribed by Collingwood to the 11th century AD (Forbes Charleson 1928-9).

56. The Early Medieval finds from Stevenston Sands near Irvine came under the spotlight in 1933 when Joseph Callendar compiled an inventory of the material culture items recovered from this enigmatic site (Callendar 1933). As well as discussing the Stevenston Sands finds, Callendar’s report presented a much wider-ranging discussion of several Early Medieval artefact types, including concave-sided shale rings and strap-ends with zoomorphic decoration, now known as the ‘Whitby Type.’

57. In the mid-1930s, Rev R S G Anderson reported on the discovery of a fragment of an early sculptured cross from Kilncroft, near the church of Old Luce, where similar examples had been recorded and had reportedly been standing prior to the founding of Glenluce Abbey in AD 1190 (Anderson 1935-6, 139-40). The same volume of the Proceedings saw the Irish archaeologist Robert MacAlister – at the invitation of Ian Richardson, then an Inspector of Monuments at the Ministry of Works – review the Early Medieval inscriptions found on stones recovered from Kirkmadrine and Whithorn (MacAlister 1935-6).

58. A more comprehensive overview of the Early Medieval period in Scotland was compiled by Cecil Curle and published in the 1939-40 volume of the Proceedings (Curle 1939-40). It was, Curle argued, the first such exercise to be undertaken since Allen and Anderson’s volume had been published more than two decades before (Allen and Anderson 1903).

59. There was a brief hiatus spanning the Second World War but work soon resumed in the immediate post-War period. An important theoretical advance was made when Robert Stevenston linked a particular type of hilltop fort – the ‘nuclear fort’- with ‘Dark Age’ capitals, thus placing their occupation in the Early Medieval Period. These distinctive sites are characterised by their enclosed summit, which sits above a series of terraces, each encircled and defended by ramparts that skirt the edge of the landform. His 1948-9 paper in the Proceedings cited several of these sites located across mainland Scotland, including Dunadd, Argyll and Bute, and Dalmahoy, Midlothian. At this stage, no potential examples were cited within South West Scotland (Stevenston 1948-9), although the Mote of Mark had already been identified as a potential Early Medieval regional capital by Alexander Curle. R C Reid suggested, just a few years later, that the Round Donnan, Dunragit, also had the potential for having Early Medieval origins (Reid 1950-1). Reid also argued, on the basis of placename evidence, that ‘Dunragit’ may have meant ‘the strong place of [the kingdom of] Rheged’, further supporting the site’s Early Medieval credentials.

60. The 1948-9 volume of the Proceedings also featured a short account by Angus Graham detailing the various surviving elements of the extensive linear feature known as the Deil’s Dyke. Graham’s paper made no attempt to assign any date of origin to this extensive and enigmatic structure (Graham 1948-9). Several years later, however, in a companion piece once again presented in the Proceedings, Graham argued for the possibility that if the date of the Deil’s Dyke was pushed back ‘to before the days of feudal organisation, we find a situation into which a work of this kind would fit a good deal more plausibly,’ that is, to within the period that we would call Early Medieval and which he termed the ‘Dark Ages.’ (Graham 1953-5, 154).

61. Meanwhile, several excavations on Early Medieval ecclesiastical sites were undertaken by C A Ralegh Radford. In July 1950, the site of Chapel Fininan, Mochrum was investigated, with trenches placed to explore the chapel building and its associated enclosure. Analogy with other early ‘Celtic’ church sites, particularly examples located in Ireland, led Radford to assign an Early Medieval date to the site, with documentary evidence gleaned from the literary sources used to bolster his argument (Ralegh Radford 1949-50a). In the same volume, Radford also presented a short article highlighting the much-earlier discoveries from St Ninian’s Cave, Glasserton. Though the cave was not being actively investigated, it was enjoying a significant change in status, having recently been brought into Guardianship by the Ministry of Works (Ralegh Radford, 1949-50b). Ralegh Radford’s investigations in the area also resulted in his publication of an Early Medieval cross-incised stone identified at Ruthwell (Ralegh Radford 1949-50c).

62. These investigations by Ralegh Radford proved to be supporting acts to the main event: the commencement of major archaeological works at the ecclesiastical site of Whithorn. The importance of these works was acknowledged from the outset: Professor John D Mackie later referred to the 1949-50 edition of the Transactions as the ‘Whithorn Volume’ (Mackie 1950-1, 37) in a later article titled ‘St Ninian and the Picts.’ This latter work explored evidence for the life of St Ninian, examined the accepted association between Whithorn and Candida Casa, and also considered wider issues such as evidence for the place of the Picts in South West Scotland (ibid.).

63. The main aims of the excavations at Whithorn were to investigate evidence for Early Medieval occupation of the site as well as the remains of the later Premonstratensian abbey which later occupied the site (Radford 1950-1). Ralegh Radford did not, however, concentrate on Whithorn to the exclusion of other Early Medieval sites. Two years later, he published another paper in the Transactions reiterating the importance of the evidence for early ecclesiastical activities at Hoddom, in the southeastern part of the region. Then in the 1953 edition of the Transactions, he published two important pieces of material culture from these sites: a fragment of staff reliquary believed to derive from Hoddom, and a circular reliquary – or phylactery – from Whithorn (Ralegh Radford 1953-4). An account of an Early Medieval incised cross of British type, discovered at Staplegorton, then followed in 1955 (Ralegh Radford 1954-5).

64. Ralegh Radford presented the final report of these first excavations at Whithorn in the 1955-6 volume of the Transactions (Ralegh Radford 1955-6). Four seasons’ excavation had taken place in 1949, 1950, 1951 and 1953 respectively. In 1949, the focus was on the ecclesiastical complex at Whithorn itself and St Ninian’s Chapel. Attention then turned to St Ninian’s Cave, Glasserton, in 1950, with the final two seasons devoted to what Ralegh Radford described as ‘various minor tasks’ (Ralegh Radford 1955-6, 131). The excavations did not, however, produce any direct evidence of early medieval occupation of the area, over and above the incised crosses already known from St Ninian’s Cave. Evidence instead appeared to derive from the Medieval period and onwards (see Chapter 10 Medieval). One item of potentially early material culture was, however, recovered. This comprised a mould fragment for a plain annular object, perhaps derived from the mould for a simple penannular brooch.

65. Despite the lack of new evidence from the excavations, Ralegh Radford took the opportunity to review the Early Medieval material from Whithorn, in particular the epigraphic evidence derived from the carved stones, and to consider the site’s importance afresh alongside similar important sites associated with the ‘Celtic’ church, including Tintagel, Cornwall and Glendalough, Eire.

66. In 1956, an early example of developer-led archaeology occurred when trenching for a water main near Terally took place, revealing a long cist cemetery which was excavated by Robin Livens. The archaeological intervention took place at a fairly late stage, but a total of 13 bodies were encountered, apparently laid in an extended position. They were placed within a stone slabbed chest or ‘cist,’ enclosed by another stone slab placed horizontally. At least two of the burials, and probably more, were aligned East-West. The character of the cists and the arrangement of the bodies, coupled with their alignment and the lack of grave goods, led the excavator to conclude that an Early Medieval date was likely for the cemetery (Livens 1956-7).

67. Several years later, Charles Thomas undertook excavations at Trusty’s Hill, Anwoth (Thomas 1959-60a), where he observed a situation similar to that observed by Alexander Curle almost half a century earlier (Curle 1913-14a). There were two phases of occupation at the defended hilltop site: an earlier phase which he dated to the Roman Iron Age, and later re-occupation which he dated to the 6th or 7th century AD (Thomas 1959-60a). In plan, the site shared many of the characteristics of the nuclear forts identified by Stevenson a decade earlier (Stevenson 1948-9). That same year, Thomas undertook measured plans of two Early Medieval ecclesiastical sites: St Ninan’s Chapel, Whithorn and Ardwell Island (Thomas 1959-60b). The same volume of the Transactions featured a revised summary of the Anglican sculptured crosses from the eastern part of the region by Rosemary Cramp. Cramp’s study focussed in particular on the Ruthwell Cross and the fragments from Hoddom (Cramp 1959-60).

68. Early Medieval re-use of an earlier, Iron Age, site was also attested at Camp Hill, Trohoughton, through excavations carried out by Douglas Simpson and Major General James Scott-Elliot on an Iron Age fort. The work was in response to a threat to the site by the landowner, a Mr McKeachie, who intended to flatten the ramparts in order to enable the land to be ploughed and cultivated. Work continued intermittently until 1962, with additional funding required from the Ministry of Works to cover the costs of excavation. What was of particular interest about the site was the fact that the earlier Iron Age monument had been re-used in the Early Medieval period as a cemetery. A total of 60 graves were identified, 12 of which were fully excavated. Orientations varied between east to west and north east to southwest, but no skeletal material had survived, with the exception of three teeth crowns (Simpson and Scott-Elliot 1962-3).

69. An important summary of recent excavations and discoveries in the southern portion of the region was then presented by Alfred Truckell in the Transactions of 1961-2. As well as including an important distribution map detailing all known sites and finds, Truckell’s work presented several advances in knowledge. Firstly, he recognised the likelihood that the Deil’s Dyke formed an Early Medieval land boundary over at least a portion of its length. He also noted the fact that many supposedly Iron Age sites that either pre-dated or were contemporaneous with the Roman occupation of the region also revealed evidence for a subsequent Early Medieval phase. Such sites included known examples of nucleated forts like the Mote of Mark and Trusty’s Hill and other probable examples, such as Tynron Doon and Stroanfreggan. Also showing evidence of this later phase of occupation were a range of artificial islands and crannogs, such as Dowalton Loch.

70. Truckell also discussed aspects of contemporary material culture. He noted the contrast between the two schools evident in the region’s sculptured stones, with the southern distribution of the Whithorn school contrasting with an ‘Iona’ School located further to the north (Truckell 1961-2,97). He also highlighted some important discoveries from the region, including a fragment of gold bracteate with filigree work of 7th century AD date from Tynron Doon (Truckell 1961-2, 94), two Northumbrian coins from the reign of Eanred (AD 806 – c. 841) recovered from Luce Sands (ibid. 90), and a Viking domestic axe of 9th-10th century AD date from Capel Rig.

71. The same volume of the Transactions also featured an important summary of the evidence for ecclesiastical occupation across Galloway and Dumfriesshire by Ralegh Radford. He presented a model of how sites may have evolved from earlier sacred enclosures set aside for services or sometimes burial, into the Medieval stone-built churches and chapels, in many cases associated with administrative units: ‘parishes.’ (Ralegh Radford 1961-2)

72. The Northumbrian coins from Luce Sands were discussed in more detail in a short note published in 1965 by William Cormack. Cormack highlighted the importance of this numismatic evidence in confirming what had previously been known only through the literary sources (Cormack 1965). He also highlighted the much-earlier discovery of a further four Northumbrian coins held within various museum collections across Scotland, and suggested that their appearance, which showed heavy erosion, as if caused by exposure to blown sand, was similar to that of the recent finds. It was possible, therefore, that they had originally derived from the same location, thus adding to its importance as an Early Medieval site.

73. The following year saw the publication of Charles Thomas’s excavation report detailing his work on Ardwall Island (Thomas 1966), which had been mentioned a few years previously by Alfred Truckell in his 1961-2 summary (Truckell 1961-2). Thomas noted two phases of Early Medieval occupation on the island. The first comprised a burial ground and a possible ‘slab shrine,’ similar to comparable examples on Irish early Christian sites (Thomas 1966, 90), within an enclosure. The second phase was represented by traces of a timber building which Thomas put forward as the remains of a timber chapel or oratory, predating the later stone chapel and coeval with another group of inhumations. Several inscribed stones were found supporting the site’s putative use as an early place of Christian worship.

74. Several interesting items relating to the region’s Early Medieval archaeology appeared in the 1967 of the Transactions. A short report by Jack Scott discussed a Viking war-axe, dug out of a peat moss prior to 1901 (Scott 1967). Another article, by James Williams, focussed on a group of decorated sandstone spindle whorls recovered from the southern part of the region. These consistently occurred either on early ecclesiastical sites, or on sites which appeared to have some kind of ecclesiastical connection inferred through place-name evidence (Williams 1967).

75. Finally, there was an interim report concerning work at Tynron Doon (Truckell 1967). Interest in this site had been sparked by the discovery of some fragments of goldwork revealed in a rabbit scrape in 1924. The fragments were subsequently identified as derived from a gold filigree panel from a bracteate pendant of 6th to 8th century AD date. Initially, work was carried out on the site over two seasons by Thornhill Extra-Mural Class in Archaeology, under the guidance of Alfred Truckell of Dumfries Museum. A final season was undertaken on the site in 1967 under the leadership of James Williams. The final report, authored by Williams, appeared in the Transactions of 1971 (Williams 1971). No traces of Early Medieval structures were identified on a site which saw its final phase of occupation during the Late Medieval period. However, a midden deposit with Early Medieval material incorporated within it was found on the southwestern slopes of the hill. Finds included a ribbed blue glass bead and worked bone pins, indicating later occupation of an earlier site. The site also produced evidence of vitrification and bloomery waste.

76. A further find of an Early Medieval ring brooch from Luce Sands was reported by Etienne Rynne in 1968. The object bore close parallels to mould fragments recovered during Alexander Curle’s excavations at the Mote of Mark in 1913 (Curle 1913-14), potentially placing it within the date range AD 475 to AD 675. Meanwhile, excavations on the site of a putative deserted Medieval village at Kirkconnel revealed evidence for a large rectilinear structure underlying a 17th or 18th century pillow mound. The earlier structure was interpreted by the excavators – T Clough and Lloyd Laing – as evidence for a ‘Dark Age’ timber hall (Clough and Laing 1968). Small finds which supported this interpretation included a glass eye bead dated typologically to the period AD 525-600 and a sherd of pottery consistent with a pre-12th century (possibly 6th or 7th century AD) date.

77. New investigations at the Mote of Mark followed in 1973. These were led by Lloyd Laing of the University of Liverpool. Laing argued against Curle’s earlier interpretation of the site, proposing instead that there had been no earlier phase of occupation during the Iron Age. He suggested instead that the earliest occupation of the site was represented by a defended settlement within a timber-laced rampart, established by native Britons during the late 5th century AD. The site was similar, Laing argued, to many other ‘Dark Age citadels’ which displayed the nucleated character, including Dunadd, Argyll and Bute (Laing 1973, 38).

78. The first phase of occupation at the Mote of Mark revealed evidence for metalworking, including mould debris for the manufacture of typologically distinctive pins, as well as for the acquisition of imported pottery and glass. At some point the timber-laced ramparts were burned and dismantled. Laing proposed that this episode of destruction could be linked with hostile activities by Anglian invaders, evidenced by finds of Runic inscriptions and a characteristic clay loom-weight of Anglo-Saxon type.

79. Laing finished his report with a summary of known finds of Anglo-Saxon material culture across mainland Scotland. These included the fragments of a gold bracteate pendant from Tynron Doon, a polygonal bead from Mouswald, a ceremonial whetstone from Dumfriesshire (the ‘Collin Whetstone’) and seven ‘Whitby’ strap ends recovered variously from Stevenston Sands in North Ayrshire and Luce Sands in Galloway. There was also a sword from Torbeckhill, Ecclefechan.

80. Laing also included the Talnotrie hoard, which, although in all likelihood buried by a Viking, was composed of objects entirely Anglo-Saxon in character. Several Anglo-Saxon coins and carved stone crosses of Anglo-Saxon character from the region were also noted. These were, he argued, suggestive of Anglian occupation centred in Bernicia but extending further westwards through Dumfriesshire and into Galloway. In the same volume, Laing wrote a short piece about the Luce Sands penannular brooch, questioning its accepted Irish associations and arguing instead for an indigenous Romano-British origin (Laing 1973b). The same article presented a comprehensive survey and discussion of Early Medieval pins from Scotland, as well as a more detailed discussion of the mould debris from the Mote of Mark. This material included moulds for the manufacture of distinctive pin types which could be dated to the 6th century AD (Laing 1973c).

81. Another important contribution to the analysis of the region’s Early Medieval material culture came in 1974, when Robert Stevenson published a description and analysis of the Hunterston brooch in the journal Medieval Archaeology (Stevenson 1974). He argued that the brooch combined a form characteristic of the Celtic north and west, with craft techniques and principles of design that had origins in Anglo-Saxon or Germanic forms (Stevenson 1983, 31). He would later return to the Hunterston brooch in a later paper which looked in more detail at certain aspects of the decorative elements, and discussed them alongside contemporary items of fine metalwork, including the Tara brooch from Eire and the Monymusk reliquary (Stevenson 1983).

82. In 1975, the published literature saw James Williams write brief notes on several sculptured stones. These comprised two fragments – including a fragmentary cross-shaft – from Ruthwell (Williams 1975a) and a cross slab from Kirkinner. (Williams 1975b). These appeared to be pieces first identified and recorded in the late 19th century, but which had in the intervening period gone astray and fallen out of collective specialist knowledge. The Proceedings would the following year include the Talnotrie hoard in James Graham-Campbell’s summary and discussion of Scottish hoards of Viking Age and Scandinavian character (Graham-Campbell 1975-6).

83. Williams then turned his attention to a re-evaluation of material from Luce Sands held within the Ludovic Mann Collection, in a short note which was published in the Transactions of 1976-7. Although perhaps better known for its prehistoric element, the collectors whom Mann had sponsored to comb the dunes for artefacts had also recovered some Early Medieval material. This was dominated by sherds of pottery of the type known as ‘E’ ware, but some metalwork was also present, including a gilt-bronze interlace-decorated mount of 9th or 10th century AD date (Williams 1976-7). The same volume of the Transactions contained a report of excavations undertaken a few years earlier, in August 1970, by Lloyd Laing and Eric Talbot on the deserted medieval village site of Kirkconnel, Waterbeck. Further evidence to confirm the presence of a ‘Dark Age’ timber hall at this location proved elusive, meaning that the interpretation of this site as an Early Medieval centre of power remained inconclusive (Laing and Talbot 1976-7).

84. Meanwhile, further excavations had been underway at Whithorn during 1972 and 1975. These revealed artefacts and some structural remains relating to occupation from the late Medieval period onwards. Evidence for land use and occupation co-eval with the Early Medieval settlement and ecclesiastical site was not, however, forthcoming (Tabraham 1979). However, sufficient evidence had been revealed to allow the presence of an 11th century Viking trading post and workshops of Scandinavian craftsmen at the gates of the Northumbrian monastery in Whithorn to be argued in a more theoretical paper by Daphne Brook on the administrative structure of the Early Medieval church in Dumfries and Galloway in 1983 (Brook 1983).

85. In 1983, a short note by Jack Scott was published in the pages of the Transactions, regarding the likely locations of Norse settlements in Galloway and in the vicinity of Kirkcudbright (Scott 1983). He acknowledged the universal acceptance of the presence of Norse settlement in the Machars, in the vicinity of Whithorn, but argued that the area only came under Norse control during or after the second half of the 10th century AD. This, he suggested, post-dated a Norse presence in and around what is now modern Kirkcudbright, evidenced by various finds of Norse material culture. These included a Viking glass linen-smoother, found during the laying of a gas pipeline within the burgh, and a Viking grave group comprising a sword, a ring-headed pin and a blue glass bead found in St Cuthbert’s churchyard (Scott 1983).

86. Aerial photographs taken by Barri Jones of the area around Ruthwell kirk in 1978 had revealed traces of a circular enclosure, in the field lying immediately to the northwest of the kirkyard. This was excavated by Chris Crowe, who sectioned the feature in 1978 and carried out more comprehensive excavations in 1980 (Crowe 1987). As well as traces of the enclosure ditch and bank, evidence for ironworking was recovered in the form of furnace lining and slags. The specialist report, compiled by James Cleland, suggested that the date of this industrial activity would have been from the period 150-250 AD onwards (Cleland 1987).

87. Work also took place at the medieval motte site at Cruggleton Castle. Excavations were undertaken by Gordon Ewart between 1978 and 1981 on the summit of the motte, which comprised a modified natural mound located on a coastal promontory. The site saw a long history of occupation, extending from the Iron Age to the Post Medieval period. Early Medieval use of the site was represented by the remains of a small timber hall with associated palisade (Ewart 1985).

88. An unusual gold mount of Hiberno-Saxon form was recovered from Rerrick, Dundrennan in 1992 Decorated with ‘stylised ribbon-beasts,’ it showed damage consistent with being forcibly removed from its parent object and a subsequent account of the object, published five years later in a short account jointly authored by Niamh Whitfield and James Graham-Campell, suggested that the damage may have resulted from ‘Viking vandalism and re-use,’ (Whitfield and Graham-Campbell 1992, 11).

89. The decade would end with a revised summary of the region during the Early Medieval period authored by Alfred Truckell of Dumfries Museum, creating what he described as a ‘proto-history’ (Truckell 1989). Truckell was able to draw upon the ongoing work at Whithorn, as well as wider discoveries made from across the region. These included a canoe recovered from the vicinity of a crannog in Loch Arthur which had provided a radiocarbon date confirming an Early Medieval origin.

90. Another work of synthesis which appeared at around the same time was by Daphne Brook. Her paper – published in the Proceedings – subdivided Galloway and Carrick into three administrative areas (‘shires’) during the ‘Northumbrian supremacy,’ the period spanning the mid-7th and the 10th century AD (Brook 1990). Brook used a variety of material to support her arguments, including place-name evidence and church dedications as well as historical, topographical and archaeological evidence. She argued for a scenario where there was co-existence between the indigenous British population and the Northumbrian elites who had seized political and administrative control of the region, with the local communities potentially having to provide tribute to their new Northumbrian overlords.

91. Excavations at Whithorn were still ongoing, undertaken now by the Whithorn Trust under the direction of Peter Hill. Traces of Hiberno-Norse settlement were identified in proximity to of the ruined medieval abbey, in an area once occupied by an earlier Northumbrian settlement of 8th century date. The latter could be dated through coin finds spanning a date range of AD 850 to AD 1000. The site of this settlement was later ploughed, indicating abandonment, but the church appeared to have been dismantled rather than violently destroyed. The second phase of settlement lasted 400 years (from AD 1000 to the mid 12th century AD) until the precinct was reorganised during the 13th century AD (see Chapter 10 Medieval). Amongst the new discoveries amassed during these long-running excavations were insights into Early Medieval cultivation techniques.

92. Further investigations were also taking place at Hoddom, which had long been put forward as a likely candidate for an 8th century Anglian monastic site. In 1991, excavations took place there in advance of gravel quarrying, with Chris Lowe, Derek Craig and Dougal Dixon co-authoring the initial report. Limited excavation took place around the perimeter of the enclosure which had been identified previously on the site, with radiocarbon dating of charcoal yielding a date of AD 920 +-50, consistent with Early Medieval occupation (Lowe et al. 1991, 15). Traces were also found of a rectangular timber post-built structure of likely 7th or 8th century AD date, as well as two corn-drying kilns. The dates of the latter fell late in the Early Medieval period and extended the site’s use onwards into the Medieval period, with 11th and 13th century AD dates recorded. In summary, this phase of excavations at the site was considered to indicate a rich potential for further detailed information relating to the ‘service-sector’ of an early Northumbrian monastery (Lowe et al. 1991. 25). One of the individuals involved in this work – Derek Craig – was also engaged in postgraduate research on an examination of the pre-Norman sculpture of South West Scotland. The results of his research were subsequently presented in a PhD thesis to the University of Durham in 1992.

93. The work at Whithorn and Hoddom focussed on two important religious sites located in the southern part of the region. Another religious site, albeit smaller in scale, was investigated at Barhobble, Mochrum, in a series of small-scale excavations undertaken by William Fleming Cormack between 1984 and 1994. The site was thought to have been the original location of the various fragments from stone crosses found throughout Mochrum parish, and while the earliest structural evidence recovered derived from a timber oratory and a possible church structure of 11th century date, evidence of occupation predating these structures was recovered in the form of Early Medieval coins (Cormack 1995).

94. Archaeological investigations were also underway at an important secular site located much further north, at Dundonald. Directed by Gordon Ewart, the works comprised three seasons’ excavations undertaken between 1986 and 1988, with a further season in 1993. The work confirmed a long history of occupation at the site, culminating in the Medieval royal castle. Later activity had removed much information from the earlier levels, but enough remained to enable the excavator to create a detailed narrative of how the site’s character had changed over time. Of particular interest was the fact that an earlier Iron Age hillfort had developed into a nucleated fort with a well-defined summit citadel and evidence for vitrified ramparts. The site’s substantial character, combined with evidence of imported pottery (sherds of ‘E’ ware from western Gaul, confirmed that it had been a high-status site. The authors of the final report – Gordon Ewart and Denys Pringle – argued that Dundonald would have played an important role in the kingdom of Strathclyde (Ewart and Pringle 2004, 127). Following the earlier argument put forward in the same volume by Stephen Driscoll and Katherine Forsyth, this would suggest that Dundonald functioned as a potentially autonomous regional capital rather than a nodal point in a vast and unified political entity (Driscoll and Forsyth 2004, 9).

95. Another medieval defensive site which was investigated at around the same time was Auldhill, Portencross. Excavated by David Caldwell, Gordon Ewart and Jonathan Triscott in 1987, it revealed similar evidence for longevity of occupation. The medieval layers were identified on a site which had seven distinct phases. The first phase was a timber-laced rampart, with some evidence for vitrification, which the excavator surmised might have had its origins as far back as the Bronze Age. This was followed by the construction of a much smaller feature: a dun. It was not possible, however, to ascertain whether this structure had been in use during the Early Medieval period, or the Iron Age, though it seemed likely that there had been some continuity of occupation from prehistory to the Medieval period (Caldwell et al. 1998).

96. The early 1990s also saw a number of discoveries during developer-led work undertaken by David Maynard, the Irish Gas Board Project Archaeologist, in association with a gas interconnector project. Archaeological evidence relating to Early Medieval activities in the region was identified at various points along the route. The first section was at Brighouse Bay (Maynard 1994). Although there was no evidence of structures, there were a number of artefacts, including an Early Medieval iron spearhead (Hunter 1994).

97. On a much larger scale were the discoveries at Chapelton, Haugh of Urr. The features identified were ambiguous, comprising two ditches of roughly contemporary date. These had been overlain by a stone spread, perhaps representing dumped stone or clearance, which in turn had been cut by several phases of pits (Alexander et al. 2004). Various artefacts were recovered, including a coin of the Northumbrian styca type, and two radiocarbon dates indicated occupation during the 1^st millennium AD. It was therefore postulated that the site represented an occupation site of Early Medieval date, and while no direct evidence of an ecclesiastical function was recovered, this could not – the excavators argued – be ruled out (ibid., 74).

98. Further east along the route, at Blairhall Burn, an Early Medieval metalworking site was identified (Strachan et al 1998). As well as smelting and smithing slags, the site revealed the remains of a bowl-shaped hearth. A radiocarbon date from oak charcoal yielded a date of AD 870-1030. Another example of small-scale activity was identified during excavations at Hayknowes Ring Ditch, Annan (Gregory 2001b). Once again, the only evidence of Early Medieval occupation was a hearth, which yielded a date range spanning the period AD 240 to 600, although in this case there was no indication that any industrial activity was being undertaken on the site. Further adding to the unusual nature of this site was the fact that the Early Medieval hearth was postdated by a circular enclosure 25m in diameter with unspecified internal features. It had originally been interpreted as a probable prehistoric monument.

99. A long-running series of excavations at Barhobble, Mochrum, were also reported by William Cormack. Here, a small 12th century chapel had been built on the site of an earlier burial ground. Fragments of worked stone, including cross fragments, derived from this earlier burial ground suggested an association with the Celto-Norse population who resided in this part of Galloway (Cormack 1995).

100. Meanwhile, Anne Crone was commencing a long-running series of research projects which would involve ground-breaking work in the study of crannogs. Her approach involved the application of various scientific dating techniques, including radiocarbon dating and dendrochronology. While the presence of Early Medieval material culture had long been recognised from crannog sites in the region, it was only when the Historic Scotland survey of crannogs in South West Scotland was undertaken that dating evidence to support this hypothesis was finally recovered (Crone 1993). Amongst the crannog sites where Early Medieval dates were recovered in the region were those from Buiston, Barean Loch and Milton Loch 3 (ibid., 248). Early Medieval occupation could still only be inferred, however, at Dowalton Loch and Lochlea through the presence of contemporary material culture.

101. The first of a series of regular roundups of Roman and Early Medieval/Medieval coin finds was included in the Proceedings of 1997. These were compiled over a ten-year period by researchers who specialised in numismatics, including Donal Bateson, then of the Hunterian Museum, and Nicholas Holmes. The majority of coin finds tended to be Medieval in origin, but some Early Medieval finds were noted. The 1997 edition (Bateson and Holmes 1997) included references to the find of a Northumbrian Phase II styca, dated to between 837 and 855 AD, during a watching brief led by Derek Alexander at Chapelton, Castle Douglas, and the discovery of two Northumbrian coins in a ploughed field near Hoddom. Mention was also made of Early Medieval coins recovered during the ongoing excavations at Whithorn (Hill 1997, referenced in Bateson and Holmes 1997).

102. During the 1990s, assessments and survey work were undertaken on the region’s large-scale linear monuments, in particular the Deil’s Dyke, as part of the Linear Earthworks Project. In 1998, a comprehensive review of these monuments was published, authored by John Barber and others. Part of this review comprised a classification of the various sub-types, including the definition of a ‘political class.’ Monuments in this group would have functioned as physical definitions of political and administrative boundaries. While it was acknowledged that these boundaries may have extended back into prehistory, it was argued that they developed their final form in the Medieval period, and in the Early Medieval period in particular (Barber 1998, 70).

103. Stephen Harrison published a reappraisal of the Millhill, Arran shield boss in 2000, arguing that it belonged to an ‘Irish Sea’ group which had Anglo-Saxon, as opposed to Norse, affinities (Harrison 2000). All were associated with Norse burials, in areas where an Anglo-Saxon tradition of warrior burials was unknown. Most also showed signs of wear consistent with battle damage. This led Harrison to argue that these particular items of material culture could be interpreted as evidence for cultural contact as much as symbols of cultural identity.

104. Two years later, in 2002, a paper by Mike McCarthy featured in the Proceedings arguing for a redefinition of the kingdom of Rheged. He suggested that its extent may have centred upon the Rhins of Galloway and also encompassed the Machars. The argument that ‘Dunragit’ was the site of its political centre had been proposed previously by Watson (Watson 1923-4), but Watson and others had argued that the kingdom of Rheged was an extensive polity encompassing Dumfries and Galloway and also, potentially, Cumbria. McCarthy argued instead for a much smaller entity, and one which had proved relatively short-lived (McCarthy 2002).

105. That same year, Ronan Toolis authored a report in the Transactions which presented the results of a series of site surveys targeting 16 promontory forts located on coastal sites around Galloway. These defensive structures typically occupy a defensive location, with the promontory protected on three sides by the topography (often sea cliffs) and the approach side defended by one or more banks or ramparts. Several examples had already been subject to limited excavations, but dating evidence often proved elusive. However, similar sites in Ireland and Northeast Scotland had provided evidence for Early Medieval occupation, so it was possible that – like the hillforts and the crannogs – these sites might have had origins in the pre-Roman Iron Age but were later subject to a secondary phase of occupation during the 1^st millennium AD (Toolis 2003).

106. Adding to the range of sites associated with Early Medieval occupation of the region was an interesting discovery made at Montgomerie Park in 2004. Here, excavations undertaken by Rathmell Archaeology for North Ayrshire Council revealed a circular palisaded enclosure measuring 27m in diameter, which revealed only three ephemeral features in the interior and no artefacts. Although a prehistoric date was anticipated, radiocarbon dates obtained from the palisade included an Early Medieval date of AD 770 – 980 from a burnt hazelnut shell. While this late date might once have been dismissed as intrusive, the site’s similarity to another circular palisaded enclosure at Titwood, Newton Mearns (which also produced an Early Medieval date) suggests that the Montgomerie Park example is comparable. During the same year, Thomas Rees and Rebecca Shaw identified more ephemeral traces of Early Medieval activity during developer-led excavations on the site of Arran High School. Here, a fire-pit rich in wood charcoal and burnt bone, but with no apparent associated structural evidence, yielded a date range spanning AD 770 to AD 980 (Rees and Shaw forthcoming).

107. Material culture studies continued to emerge during the late 2000s, with Timo Wegner undertaking doctoral research on the so-called ‘Dumfriesshire fragments.’ First reported in 1906, these comprised two fillets of bronze with Christian motifs, and five decorative bronze bosses. Wegner saw these objects as potentially originating from an ecclesiastical object such as a sheet metal book cover. Lloyd Laing had suggested previously that they may have been recovered from Tynron Doon, but Wegner, advised by James Williams, argued instead that they represented objects originally recovered from the site of Corrie Loch by a Dr Moffat, who presented them to the Society of Antiquaries of Scotland (Wegner 2006). A silver terminal fragment recovered prior to 1997 from Whithorn also achieved a mention in a revised inventory of Viking and Late Norse gold and silver finds from Scotland, authored by James Graham-Campbell (Graham-Campbell 2008).

108. Work continued at Whithorn in 2008 and 2009, undertaken this time by the University of Bradford. This involved extensive topographic and geophysical survey in the environs of the site. Their focus included St Ninian’s Chapel and the promontory fort at Isle Head as well as the balance of ground in between the two sites. They concluded that the fort had the potential to be a complex site occupied over an extensive period, potentially comparable with other sites like the Mote of Mark. The initial phase of use was likely to have been during the Roman Iron Age, with a later phase in the Early Medieval period (McCarthy et al. 2010). Further work here, they suggested, might prove extremely rewarding.

109. For many years, if not decades, the majority of sites subject to the most detailed investigation – particularly the more complex ones – had been located in the south of the region, in Galloway, or Eastern Dumfriesshire. The emphasis changed, however, in 2011, when a groundbreaking discovery at Mountfode in North Ayrshire reached final publication. Excavated five years previously during developer-led works led by Candy Hatherley and carried out by Headland Archaeology, the site comprised an Early Medieval cemetery which numbered 60 graves in total, of varying character.

110. Some of the graves at Mountfode were enclosed in long slab-built cists. Others had partial cists or were placed in simple earth-cut graves. The main part of the cemetery seemed to fall within the bounds of a semi-circular enclosure, defined by a ditch. The grave cuts respected the internal line of the ditch, which suggested to the excavator that there may originally have been an internal bank present (Hatherley 2009, 199). Radiocarbon dates recovered from the site placed its use between the mid-6th and mid-7th centuries AD. The orientation of the cists varied, with some aligned NW-SE and others WNW-ESE (ibid. 202). It seems likely, however, that the earliest burials within the cemetery were laid out in a well-ordered fashion, probably accompanied by grave markers, with later additions added in a much more haphazard fashion. The preservation levels of the interred individuals were unfortunately very poor.

111. At roughly the same time, further evidence of settlement was being uncovered at two sites across the region. The first comprised a series of postholes revealed during community-led excavations in the South Range of Kilwinning Abbey (Rees 2012). Radiocarbon dates recovered from wood charcoal returned dates spanning the period AD 385 to AD 583, potentially indicating that a settlement had been present on the site centuries before the foundation of the Abbey. It should be noted that Kilwinning is also the findspot of a fragmented free-standing cross, thought to date to the 9th century AD.

112. A more substantial Early Medieval structure was identified during excavations carried out at Lockerbie Academy by Magnus Kirby of Headland Archaeology (Kirby 2011a and b). Amongst the discoveries made on this complex, multi-period site was a rectangular structure interpreted as an Anglian timber hall on the basis of its similarities with examples from classic Northumbrian type-sites such as Yeavering. Radiocarbon dating confirmed its Early Medieval origins, giving a date range – obtained from charred barley – of 420-630 cal AD (2σ).

113. The timber hall had immediately superseded an earlier post-built structure of similar scale, which yielded radiocarbon dates of 50–660 cal AD and 600–675 cal AD (2σ) from two of the post-holes. This indicated that the structure – though predating the Anglian hall – was roughly contemporary, and so Kirby and his team concluded that it must have belonged to an earlier British tradition of rectangular timber structure (Kirby 2011a, 60).

114. A settlement site of a different type was investigated during 2012 when Ronan Toolis and Christopher Bowles undertook excavations at Trusty’s Hill, Anwoth, as part of the Galloway Picts Project (Toolis and Bowles 2013). The project had begun with the first detailed measured survey of the site, undertaken by RCAHMS. This was coupled with a topographic survey, allowing the four trenches cut as part of the excavations to be accurately located.

115. The 2012 excavations reached the occupation layer that Charles Thomas had identified in his 1960s investigations on the site. Work then proceeded to remove this layer, demonstrating that the occupation layer had built up over a collapsed rampart, and indicating that this phase of use had post-dated the slighting of the fort’s earlier defences.

116. Conveying a more ambiguous picture were the features identified in association with road-widening works at Derskelpin Farm, Glenluce in 2012. Phil Moore and Ian Suddaby reported on the discovery of a burnt mound – usually associated with prehistoric activity – and its subsequent excavation (Moore and Suddaby 2012). The mound material, which comprised burnt stones, was consistent with prehistoric burnt mounds, but some elements were absent. In particular, the hearth and the stone-lined trough, which are regularly encountered, were missing. Further adding to the unusual character of this site were its dates, which ascribed it to the Early Medieval period. The mound of burnt material apparently overlay an old fence line, although the excavators argued that both features were broadly contemporary. Its function must remain uncertain, with its presence a reminder that there is still some uncertainty to be found in the range and character of Early Medieval sites occurring within the region.

117. Material culture studies of the decade 2010-20 included the publication of a paper by Alice Blackwell in the Proceedings, revisiting the Hunterston Brooch. Blackwell explored the consistent use of zoomorphic decoration and put forward the theory that the use of the opposed animals at the terminals of these brooches might have a religious meaning, in a society which did not depict Christ in human form. Other works of synthesis included a useful and comprehensive roundup of the physical evidence for early Scottish churches authored by Sally Foster (Foster 2011). Included in the volume Buildings for Worship in Britain and Ireland 350-950 (Barnswell 2011), Foster examined a wide range of factors, including site plan, type of structure and associated material culture form. Amongst the sites comprehensively discussed in her work were Whithorn and Hoddom.

118. Excavations at Hunterston North in 2014, led by Douglas Gordon of Rathmell Archaeology, uncovered a keyhole-shaped kiln which yielded a date range spanning the 8th to 10th centuries AD. Although an unusual discovery within the region, the kiln represented a form encountered relatively frequently in Ireland. Its discovery further strengthened potential links with Ireland that had first been inferred by the much-earlier discovery of the Hunterston brooch, recovered close to the site of this roughly contemporary kiln.

119. The year 2014 also saw the re-evaluation of a collection of Viking-age metalwork from Carronbridge. Comprising a sword, a penannular brooch and a sickle, the group was found in 1989 and roughly dated on typological grounds to the 9th or 10th century AD. They had been ascribed ‘uncertain’ status and interpreted as the property of a ‘lone traveller’ (McLeod 2014, 13). Shane McLeod argued against this interpretation, suggesting instead that what was represented by the finds at Carronbridge was a bona fide Viking-age burial. It was in fact possible that the deceased had been placed in an earlier monument – a penannular enclosure – which was located close to the findspot of the metalwork. The re-use of earlier monuments was, McLeod proposed, entirely consistent with Viking practice encountered elsewhere, such as at the Broch of Gurness in Orkney (ibid. 15). Although located inland, McLeod also suggested that the findspot’s location, close to the line of a Roman road and overlooking a major river, was again entirely consistent with Viking funerary practice. It had originally been proposed by Olwyn Owen and Charles Welander of Historic Scotland that the metalwork marked the resting place of a lone individual who had died, unattended, while travelling. Instead, McLeod argued that the deceased had been formally placed in a deliberately chosen location by others travelling with them.

120. It was during this same year – 2014 – that the region produced what was undoubtedly its most impressive Early Medieval find since the discovery of the Hunterston Brooch almost two centuries earlier. This was the hoard from Balmaghie (also known as the Galloway hoard), which was uncovered by metal detectorists on Church of Scotland land. Thanks to the prompt actions of the detectorists, who alerted the Treasure Trove unit immediately, Andrew Nicholson of Dumfries and Galloway Council was able to excavate the hoard in its entirety and learn more about the context of its deposition (Nicholson 2016). The find consisted of two separate deposits of material, separated by a ‘false bottom’ of clean gravel. There was also evidence, within these two separate deposits, that various items had been packaged separately, in containers which included textiles, leather bags or packets, a wooden box and a Carolingian silver pot. The hoard contained a variety of items, including silver arm-rings, a large Anglian silver cross and a pointer in the shape of a bird. The objects dated to roughly AD 900.

121. Further excavation in the vicinity of the find revealed traces of a roughly contemporary settlement, occupied at around AD 900 – AD 1000 and likely to be linked with Norse settlement in Galloway during the mid-9th and 10th centuries AD (ibid.). The hoard seemed to fit in with a wider pattern of similar hoards occurring at around this time along the Irish coast, with other examples recovered from Ireland, Anglesey and inland locations along the main trade route to York. Further research into the hoard and its wider landscape setting and historical context was then undertaken by Orla Craig, whose PhD thesis was submitted to the University of Glasgow in 2024.

122. New discoveries have been reported right up to the present. In 2020, GUARD Archaeology excavated a multi-phase site at Curragh, Girvan. The latest structural evidence on the site comprised three corn-drying kilns, one of which was radiocarbon dated to the Early Medieval period (Barbour and McNicol 2025). This was followed in 2021 by an important iron-working site at Glenlee. Here, three large working hollows associated with furnace bases were found, associated with various structural elements which included pits, postholes and ditches. The industrial residues were dominated by iron smelting slags, but slags and residues derived from smithing and blacksmithing were also recovered. As well as producing the largest Early Medieval iron smelting assemblage yet recovered in Scotland, the excavations at Glenlee also represented the country’s first known example of an Early Medieval bloomery site.

---