John Lorne Campbell’s firmly held belief that Canna had been the sacred island Hinba, where Columba landed first in 563, before arriving on Iona most probably in the early 570s where he established his monastery, still remains to be proved or disproved, but there is no question of the fact that there had been a historical association with the saint. His researches uncovered the entry in the Ravenna Cosmography of the Seventh century, indicating that the name given to the island preceded the Norse invasions. The earliest recorded descriptions of Canna in the sixteenth century make mention of a parish church, and Martin Martin informs us (1695) that it had been dedicated to St Columbus (Colum Cille). However, the artistic importance of the two carved stone crosses, and of the large number of cross-marked slabs that have survived in the graveyard and have been taken for safekeeping to Canna House, point strongly to the fact that Keill had also been the site of a monastery, of which nothing now survives. It would have been part of a chain of Celtic paruchia which extended over the vast area of north-west Pictland which had been colonised within a century of the saint’s death, and it provides an explanation as to why Canna was ceded to the Abbot of Iona in whose ownership, and that of the abbot’s successor, the Bishop of the Isles, it remained until the early seventeenth century. Excavations in the area of the carved cross undertaken in 1994 led to the suggestion that there could have been a pre-Christian burial chamber here.
The natural communication channels for the Early Christian missionaries were by sea, connecting the Scottish islands and the Western Seaboard, with Iona at the centre, with Ireland. It was to Ireland that the monks would withdraw at times of Viking raiding. As a target for portable wealth Iona was attacked on no less than three occasions in the space of nine years, and on the last of these occasions, in 806, as many as 68 monks were massacred. To Columba’s missionaries, who withstood the constant raiding and ransacking of the monasteries, a barrier to extending their influence more widely on the mainland became the mountain ranges which formed the backbone of the Highlands.
The evidence is all the more remarkable for the existence of a further religious site on Canna, of which local tradition and the archaeology of the site point to it having been a cashel, or a nunnery. The inaccessible location of Rubha nam Ban-naomha, its proximity to the sea together with the extensive archaeological remains, all indicate that it had been a self-sufficient retreat and that it could not have been reliant upon the field systems of the island. Provisions, were they ever needed, could only have been brought to the site by boat. After the monastery was abandoned this remote site was venerated by those who lived at the nearest settlement at Tarbert. The inhabitants cared for it, and used it for occasional worship.
How the religious sites on Canna fared during the frequent Viking raids of the late eighth and ninth centuries can only be guessed at. The strongest evidence of their interest in the islands, and that they might have settled and became integrated with the community comes with the survival of numerous place-names which are of Norse origin. These names include Sanday, for the south island. The raids were partially a consequence of over-population on the west coast of Norway, and to the harassed communities of the islands the raiders were known as either the Black or White Gentiles, depending on whether their country of origin was Denmark or Norway.
With the emergence of the dominant clans throughout the isles in the early medieval period, fealty to the feudal superior was laid down in specified numbers of ships and oars that should be capable of being raised in the battles to control the seaways, adopting the Viking principle of servitude. Thus a galley may have as many as 34 oars, and a birlinn as few as 12. In an archaic document drawn up in 1672 to acknowledge a lease granted to Clanranald for the continuing occupation of Canna, there was a throwback to earlier charters with the request that the feudal superior should be served, when needed, with ‘a galley of sixteen oars, sufficiently appointed with men and necessaries for thirty days yearly between the isle of Canna and Icolmkill….’.
One of the strongest physical associations with the Viking longboats is derived from the survival of a number of long, oblong stone enclosures, of which the best known is that on the grassy plateau on the north of Canna at Rubha Langan-innis, known as the ‘King of Norway’s Grave’. It has long been held that these represent evidence of ship-burials, although the conclusions from the fieldwork carried out by the RCAHMS team suggest less certainty as to what their purpose might have been on other parts of the islands where similar remains were identified.
About life on Canna during the early medieval period, when the monastic settlements were well established, very little is known. Some of the boundary walls of stone boulders, and of stone and turf, defining the landscape of the upper plateaux may have been laid down at this time. Transhumance, and the use of shielings for summer grazing of animals, would have been commonplace, and is likely to have died out by the eighteenth century. Consistent with the patterns associated with monastic land use, it is thought that tidal fish-traps (caraidh, or cairidh) were to be found on both Canna and Sanday, consisting of stone walls erected below the high tide line to provide a basin in which to trap the fish. The chapel at Keill must have been a prominent feature within the landscape until all traces of it had disappeared by the mid-nineteenth century; Thomas Pennant noted that it had been ruinous in 1772.
The claim of Iona over Canna is ratified in a papal bull of 1203, which lists Canna as one of the endowments of the monastery. However, just as had been the case at the time of the Viking raids, church protection was insufficient in itself to spare the island communities from disturbance from raiding. The problems on Canna appear to have escalated during the 1420s, and may have been occasioned by feuding within the Lordship of the Isles, which must have led to the reforming abbot of Iona, Dominicus, taking the unusual step of securing a papal mandate which disallowed anyone of noble blood from entering the monastery there. The consequence of this extended bout of lawlessness was that the Canna had to be abandoned. In 1428 the abbot pleaded with Rome for the offenders to be excommunicated, and for the community to be left in peace. He wrote:
Since in the island of Canna …… by reason of wars and other calamities in the past divers homicides, depredations and other ills were perpetrated, so that some strong men of the familiars of the Abbot and convent were slain by pirates and sea rovers, and divers farmers and inhabitants of the island were afraid to reside there and cultivate the land, and transferred themselves elsewhere, deserting the island to the no little loss of the said monastery….
The ‘problem’ of the Isles and how to keep it under control troubled a whole dynasty of Scottish kings. James IV, having assumed the title of the Lordship of the Isles for himself after it had been declared forfeit in 1493. He was often in attendance during the sea raids on the area during his reign. Insurrection was rife: there were no less than six major risings in the Isles between 1494 and 1545, for which James V pursued a policy of ‘daunting of the isles’ and was prepared to resort to deception and treachery which he used against the clan chiefs in pursuit of his cause. James VI had still not resolved the problem by the end of the sixteenth century. His efforts to call the clan chiefs to order and require them to display their title to the land they claimed to possess in 1597, and to find sureties for good behaviour led only to a further prolonged spell of violence and disorder. In his private thoughts he could only express despair, regarding the chiefs ‘that dwelleth in the Isles’ as being ‘alluterlie barbares’.
The Canna Local History Group would like to thank ANDREW PK WRIGHT; OBE, BArch, RIBA, PPRIAS, FRSA, FSAScot.
Chartered Architect & Heritage Consultant
for his contribution to this text.