Quote needed. Format Footnotes.
We have defined the clan as a self-sufficient group of people, permanently settled within a defined area, who lived by hunting and herding. Depending on topography this territory might be a shore, a glen, a promontory, a small island, or part of a large island, and it was defined both by geography and by a central beacon located at the clachan or administrative centre. A few clans were small and isolated, but most had a number of related branches living in proximity and many belonged to a wider grouping based on mutual interest. Native clans claimed to own their lands allodially, by right of first occupancy, and defended them by the sword. In the case of the Fletchers, who continued in Glenorchy up to 1587, the traditional formula survives: ‘Se Clann-an-leisdeir a thog a chiad smuid a thug goil air uisge ‘an Urcha – ‘The Fletchers were the first to raise smoke and to boil the water of the Orchy’.1
The most basic and persistent territorial division in Highland Scotland was a self-contained hunting forest, typically a glen and its hinterland to the surrounding watershed, which provided a defined area capable of being exploited by several hundred people and containing enough resources to support them in a sustainable fashion. At some point, no doubt as populations rose, the better pasture was allocated to ten or twelve townships or pastoral farms. One factor that no doubt contributed to this sub-division was the need to accommodate refugees from low-lying land in the west as sea levels rose. A township supported perhaps fifty people and their domestic animals. Transhumance was the universal practice. The low-lying pasture, which was often in dispersed units or ‘dals’ along the valley, was occupied only in winter and was always complemented by extensive summer grazings or shealings in the upland, which were also formally defined and allocated. G. dail ‘field, dale, meadow’, dàil ‘portion, share; tribe’, dail ‘distribute, deal’. cf. E dale, dole, W. dol. The laying out of defined townships and shealings almost certainly predates the Late Bronze Age and may be earlier. This assumes that the native wild cattle or aurochs was exploited before improved domestic cattle were introduced, a point explored in the chapter on milking. This arrangement was formalised by the application of the soum, which defined the allowable number of cattle and was evidently designed to prevent overgrazing. Erosion is found already in the Mesolithic and reached catastrophic proportions in the Neolithic. It appears that souming predated the arrival of sheep in the Highlands, for cattle and horses are universally soumed in the native units of 2, 4 and 8 while sheep are soumed in imported decimal units of 5, 10 and 20, at least in Argyll.
Tribal beacons
The physical and spiritual focus of every clan settlement was its beacon. To be a member of a clan meant, in practical terms, that you were obliged to muster whenever the beacon was lit. Conversely, the settled area was limited to those places which had sight of the beacon, since it was necessary for every household to have a clear view of this central point and to be able to muster there rapidly.
Duncan Campbell gives a remarkable example of a prehistoric beacon system still in use in 1855 in Fortingall, a parish in the heart of Perthshire. When news of the fall of Sevastopol, marking the end of the Crimean War, finally arrived there, it was late at night. The good news was read out to those people who were still waiting to hear it, then ‘a rush was made to the church, the bell was rung furiously, and from different places, at some distance from each other, bonfires blazed up to show that the message of the bell was understood and welcomed.’2
Settlement round a central beacon was not only prevalent in Highland Scotland but may be traced in Western Europe generally. For it has left its mark, notably in the division of land into parishes and in the visibility of medieval churches, a feature which is evident in France where parishes were defined and many churches, complete with belfries and bells, built in the ninth or tenth centuries. But it is not uncommon for a Highland parish to consist of a single glen defined by the surrounding watersheds and still centred on its old beacon site or clachan, now occupied by the parish church. Settlement is typically dispersed to make best use of the better land but the old townships are all sited so as to have a view of the central beacon.
An excellent example of a self-contained tribal enclave organised on these lines is Balquhidder, Perthshire. At or near the church we find a considerable number of communal features: Puderach, a standing stone, Tom Anghais, a beacon site (G. ain ‘fire), Tom nan Aingeal, another beacon site (G. aingeal ‘bright fire’) and Creag an Tuirc, a third beacon site or outlook site on a rock above the church. There was also once a small stone setting said to be the oratory of St Angus, but as G. ain means ‘fire’, and3 means ‘to unite, join’, perhaps Angus was originally an aonach'‘fair, great assembly’ and that the oratory was a muster point. Angus' fair was held in Balquhidder in August and, like all fairs, no doubt originated as a hunting event.
Many beacon sites are still associated with clans. A little field work would show whether they are local gathering beacons or cover a wider area. According to one source, when Macdonnell of Glengarry wished to summon his clan, he first lit a beacon on the highest tower of Invergarry Castle, which was built on Craigan an Fhithich ‘raven rock’ and immediately a second beacon was lit on Craig an Gour, whence it could be seen the length of Glengarry and up and down the Great Glen.4
Buchanan: Clar Innis, an island in Loch Lomond.
Colquhoun: Cnoc Ealachain (G. ealachain ‘fire place’).
Farquharson: Càrn na Cuimhne (G. coimhe ‘outlook’).
Grant: Craigellachie (G. ealachain ‘fire place’).
Logan: Druim nan deur (G. doire ‘deer forest’).
Macdonnell of Glengarry: Creag an Fhithich, Invergarry (a pun on G. fiadh ‘deer’).
Macewen: Caisteal mhic Eoghuin, Lochfyne (G. aodhan ‘fire’).
MacGregor of Glenorchy: Tigh Mor, Dalmally, covers the north half of Loch Awe.
MacGregor of Rannoch: Eilean Faolag, Ardlarich, visible the length of Loch Rannoch.
Mackenzie: Tulach Ard (G. tul ‘fire’).
Macnaughton: Fraoch Eilean, an island in Loch Awe.
Munro: Castle Foulis, Ferindonald, Easter Ross.
Weem, a singularly dissevered and fragmentary parish …
That the parish was once equated with the territory of a single clan led by a single chief is shown in a very different way by the chaotic boundaries of certain old parishes in Perthshire where the eight parts of Weem were ‘strangely intermixed with wings and detachments of Logierait, Dull, Fortingal, Kenmore, Killin, Comrie, and Little Dunkeld.’5 A detached portion of the parish of Logierait was entirely surrounded by the lands of Fortingall and half of Glenlochay and a farm on Lochtayside adjacent to the parish church of Killin belonged to Weem whose church was 22 miles away. Other parts of Weem were as much as 30 miles from its church. Before the parishes were reformed in 1893, a person travelling the short distance from Kenmore to Aberfeldy passed through portions of the parishes of Weem, Dull, Fortingall, Logierait and Dull again.6 This eccentric arrangement reflected feudal possession at the time when the parishes were set out. The local landlords claimed jurisdiction over their own scattered properties, like native chiefs, without consideration for the fact that their lands no longer lay compactly around the clachan. Thus all the scattered properties of Menzies of Weem were included in the parish of Weem, all the bits of Logierait belonged to the Earls of Atholl, Kenmore and Killin were Campbell parishes, and so on.
The regional clan grouping
It is not always possible to reach the pre-feudal situation but the distribution and organisation of native clans was very dependent on local geography, and that has not changed since sea levels stabilised. The concept of the regional clan grouping is a useful way of creating a possible reconstruction of the prehistoric situation. It can be defined as a voluntary association of several local clans, linked by a single communications system. The size of such a grouping depends on the range of visibility or field of the master beacon. Beacons were more effective over a much wider range among the islands of the Hebrides than they were among the mountains of the mainland. We saw in the previous chapter that the origins of extended clan groupings such as Clan Donald and Clan Gillean can be explained by the range of beacons on Skye and Mull.
At its simplest, controlled settlement in certain areas which share access and facilities, such as the lands around an inland loch or the two sides of a sea-loch, implies the willing collaboration of all local groups. From a need for co-ordination, a local community of autonomous groups or clans subject to a hierarchy of control would naturally evolve. The hierarchy was based not on conquest but on the geographical fact of intervisibility which gave some places a natural advantage, and on the acceptance of necessary leadership. Highland Perthshire provides examples. The possibility of controlling Loch Tay, which is 30 km (20 miles) long, using beacon islands at both ends and a number of intermediate relays has already been discussed. All these points were under the control of elite families who constituted an effective heirarchy depending on the importance of their long-distance beacon. A similar system links the east end of Loch Rannoch with the elite centre at Ardlarich at the west end using the artificial island known as Eilean Faolag which is built very far out from the shore to provide a view of a distant signal. In addition, higher sites such as Ben Lawers and Ben More are widely visible both within Breadalbane and at a greater distance and added to the importance of the area since they could be part of a regional or national network. It is also worth noting that some local beacons seem to be positioned where its visibility was narrowly restricted. It is logical to suppose that a local beacon had to be placed where there could be no possibility of confusing their signal with beacons at a higher level in the hierarchy.
Signals could serve many purposes but the main importance of this communications structure was to organise and control hunting. Since deer and other wild animals are a mobile resource, it was important that the local food supply was not disturbed prematurely, or driven into another area, whether by purpose or design. From various surviving conventions it is clear there were fixed boundaries to clan areas, limited access to deer forests, and a calendar which co-ordinated the dates of hunts. Like agreeing to drive on the right or left side of the road, these rules, particularly the last, did not need any compulsion to enforce. All that was needed was a reliable way of identifying and broadcasting the dates of local hunts. The full moon or the new moon was reliable within a day or two, and both appear to have been used. Midwinter sunrise was another popular date which was easy to establish. At various times much effort has gone into making very exact local calculations, using standing stones and posts, wooden and stone circles, and alignments of various kinds, but these systems became obsolete. A much simpler and more powerful system was achieved by a single signal sent by a regional or national network from a central point. So long as everyone received the same signal - at Hallowe'en for example - it was not important if the date was exactly right or not. A lunar date might be out by three or four weeks but did not need to be more exact. Long-distance beacons were also needed to exploit the more remote areas. This brings us back to the regional clan groupings. The regional differences which are a such a notable feature of Scottish prehistory probably reflect the existence of a local RCG with a centre and a periphery, which was once devoted to making hunting easier and more productive. Moving forward to the Dark Ages, the Picts, who were great hunters, evolved a control system which covered the whole country from the Forth-Clyde as far as Shetland.
This very briefly outlines how the Highlands and Islands might have been organised in prehistory. Very similar tribes were also typical of the Scottish Borders, a hilly area noted for stock-rearing and raiding and less commonly recognised as an outpost of Gaelic culture into the Middle Ages.
Leadership
If the physical focus of the clan was its beacon, its human focus was its chief. The members of a clan were much intermarried – ‘in this insulated state, with a very limited admission of strangers, intermarriages and consanguinity were the natural consequence’7 – and blood ties played their part in creating a feeling of community, but the essential link between the members of a native clan was not blood but their formal sworn allegiance to a common chief who controlled the gathering beacon. The bonds of manrent recorded in the Black Book of Taymouth were used in a later context but the wording suggests that even unrelated individuals could join a clan by swearing an oath of loyalty to its chief. In prehistory we may rather envisage a coming-of-age ritual which recognised that a boy had passed certain tests and was now strong enough and wise enough to join the men on their hunts.
The existence of an elite did not manifest itself in material terms, as there are disincentives to accumulating material possessions in a mobile hunting society, but social stratification is implicit in the organisation of communal hunts involving several hundred men. The elite responsible for organising these hunts, on which the survival of the clan as a whole depended, consisted of a chief and his deputy or tanist, G. tànaiste ‘anything parallel or second to another; lord; thane; the third name of dignity among the ancient Caledonians.’ and, more obscure but certainly present, a council of elders who were responsible for appointing, advising, and on occasion replacing the chief. Every clan also had its families of poets, lawyers, doctors, historians and musicians who acted as repositories for the various aspects of oral learning. The elite was responsible for creating and applying tribal law, managing the deer forest, calculating the seasons and keeping the calendar, planning the hunts, allocating manpower, summoning hunters, distributing food, and teaching the rising generation. It had to negotiate with neighbouring clans to control outlaws, maintain boundaries and organise joint hunts. This degree of social organisation has no evident starting point. It is found all over Scotland and is implicit in its territorial divisions. Life may well have been equally complex in the Upper Palaeolithic.
As an individual the chief was expendable, to the extent that he was shadowed in life by a competent replacement. The election of a tanist was probably not, as is often said, to eliminate bloodshed on the death of the chief, since there seems to have been nothing to prevent any number of candidates presenting their claims at that point, but to ensure that if the leader were to be killed or disabled at a critical moment during a raid or a hunt, he would be replaced immediately by an agreed deputy. The rank and file were highly valued, since the success of the hunt largely depended on their skill, stamina and numbers. In the traditional view, a chief with lands but no following was chief of nothing, and in almost as bad a position as a leaderless or ‘broken’ man who had lost his legitimate means of support. This interdependence explains a chief’s pride in a numerous following and the clansman’s reciprocal pride in his chief. An old woman is said to have hailed Cameron of Lochiel as ‘our own great god of the Camerons!’8 ‘There is a connection between the proprietor in the Highlands and his tenants that cannot be easily described. He regards them as part of his family; and though he may occasionally use severity against individuals, yet he appears to have an interest in the welfare of the whole’.9‘My uncle says that if ever the Gairloch family had a devotee it was Banosdair Ceann-Loch-Iubh [the inn-keeper at Kenlochewe], and he believed she would cheerfully have gone to the gallows if she were quite sure that would please the laird’.10. (His uncle was born in 1803.) ‘I have lived to woeful days,’ said an Argyllshire chieftain in 1788. ‘When I was young, the only question asked concerning a man’s rank, was how many men lived on his estate – then it came to be how many black cattle it could keep – but now they only ask how many sheep the lands will carry’.11 He was right to be depressed.
As far back as it can be traced, leadership of a native clan was elective. The earliest description of an elective kingship is the account of the king of the Hebrides given by Iulius Solinus in his Polyhistor which was compiled in the third century AD. This monarch was not a clan chief but a high-king or judge responsible for administering the territories of several clans spread over several islands - in other words, a regional clan grouping. His neutrality was scrupulously promoted.12
"The isles of the Hebudes are separated from one another by narrow sounds, and by reason of their contiguity are governed by one King. This Monarch has no property but is supported at the expense of the public. He is bound by established laws to rule according to the principles of equity. Lest he should be tempted by avarice to commit any acts of oppression, his poverty keeps him within the rules of justice. He has no personal interest to promote. He has no wife that can with any propriety be called his own but may enjoy any woman for whom he conceives a passion. Hence he has neither hopes nor desires with regard to children, to whom he cannot claim any personal right."
We are not told how this king was selected but he must have been trained up in traditional law from a young age. This suggests a hereditary system. The Hebrides were part of the Pictish state and the Picts are said to have ruled by mother-right. This still allowed a son to be trained by his maternal uncle. It appears that Pictish kings, also, were often succeeded by their brothers or nephews but never by their sons.
In more recent times there have been several well-documented cases where Highland clans did not observe feudal succession or where they deposed a chief whom they found inadequate. The candidates in every case came from a small number of elite families, no doubt reflecting the value of accumulated management experience. However the esteem of the clan was not given gratuitously to the feudal heir and there was keen competition among young contenders to prove themselves. ‘The first specimen of manhood expected in a young chieftain was dexterity in hunting: the next was, to make an incursion attended with extreme hazard on some neighbour, with whom he was at open variance, and to carry off by force of arms whatever cattle he and his followers fell in with. In this manner conflicts and feuds were nourished and kept constantly in existence among our Scottish Highlanders. But these conflicts ceased almost entirely about the middle of the seventeenth century’.13
Dugald, the 6th chief of Clanranald, was so vicious that his clansmen killed him and put his uncle Alasdair in his place. When Alasdair died in 1530, his natural son, the great John of Moydart, succeeded him. John was still chief of Clanranald when he died in 1584 I. Grimble 1973, 144-5. but in the interim he had been imprisoned for several years by James V. During this enforced absence he was replaced as chief by Ranald Galda, an individual who had a much better claim in law and who appears to have been a competent leader. But, as soon as John of Moydart was freed, the whole clan, including the sons of the murdered Dugald, again acknowledged him as their chief and Ranald Galda was deposed.14
When the male line failed among the Macleods of Harris in the middle of the sixteenth century, the right to the feudal title passed to an heiress, but the clan, needing a leader, reverted to traditional ways and chose first the brother of the late chief, then a cousin who murdered him, and finally Tormod Macleod, another uncle of the heiress. Meanwhile Argyll, her feudal superior, had married her off to Duncan Campbell of Auchinbreck who thereby acquired a claim to the title. To avoid a dispute, Tormod gave Argyll a bond of manrent, resigned certain lands, and paid five hundred merks towards his niece’s dowry, thus confirming to the rule when negotiating with a Campbell: ‘Heads he wins, tails I lose.’
The story of the Hen Chief is also associated with Clanranald. As told by Walter Scott, who is not the most reliable source of Gaelic tradition, the heir had been fostered by the Frasers and had acquired parsimonious Lowland attitudes. His first remark on arriving home was to complain that the number of cattle which had been slaughtered to celebrate this great occasion was excessive, when a few hens might have been enough. The clan contemptuously named him the hen-chief and appointed one of his brother’s sons in his place.15 This story does at least assume that a clan could choose its own chief but it is likely to be an attempt to explain an obscure Gaelic phrase translated into English as ‘hen-chief’ – a ‘hen’ which is not a hen appears in many old stories. The story-teller has also got in a dig at the Frasers.
The most consistent picture of elective leadership is provided by that archaic and anarchic body, Clan Gregor. The memorial drawn up by Lord President Forbes of Culloden in 1745 makes it plain that they did things differently: ‘They have no present Chief, that being elective, and continuing no longer than the current expedition; he is chosen on the principle of detur digniori’ – ‘Let it be given to the most worthy.’ D. Stewart 1822, II, vii. The earliest election of which there is mention is said to have been held in the old church of Strathfillan shortly after the death of Iain Dubh in 1603 and the execution of his brother Alastair of Glenstrae in 1604 left the clan without a leader, though Iain Dubh left three young sons. As reported, a candidate was proposed by the southern branch, but the head of the collateral branch preferred Gregor, ‘a natural son of the last laird, a man of martial fire’, and, dragging the first candidate from the inaugural chair in the Kirk of Strathfillan, placed Gregor in his stead.16 This story does not make much sense and none of the details can be verified, but it is certainly true that in 1604 Clan Gregor urgently needed a competent leader and it correctly assumes that this clan would resort to election in such circumstances.
They did this again in 1714. To allow them to claim the pension which Queen Anne proposed paying to Highland chiefs, fourteen ‘Heads of the Families of the Clan of Macgregor’ met and elected Alexander MacGregor of Balhaldie as their hereditary chief and confirmed the right of his son to succeed him, to fulfil the requirements of the claim but acting according to their own traditions.17 This son succeeded in due course and was an active Jacobite but in 1745 it was Gregor Glun Dubh of Glengyle who was appointed colonel of the clan regiment, a position invariably occupied by the chief of a clan. When the Balhadie line faded out, the heirs of Glun Dubh did not bestir themselves to claim the chiefship but were generally regarded as chiefs of Clan Gregor if there was such a thing. Then in 1787 the chiefship was claimed by Sir John MacGregor Murray, an extremely wealthy man of obscure ancestry who had endeared himself by buying property in the old clan lands. He promoted an election with himself as sole candidate and his claim was accepted by the Lord Lyon, though not before 1795. This was surprising, because he was not even certainly head of his own family, there were others with valid claims, and the genealogy presented has been described by a clan historian as ‘a piece of sustained fiction marred only by the occasional intrusion of fact.’18 Sir John subsequently disqualified himself by persistently using the surname Murray. He did this 'out of loyalty to the Duke of Atholl', as he put it, but one cannot be chief of one clan and owe allegiance to the chief of another. When he died in 1822, his son, Sir Evan, rapidly changed his name to Murray MacGregor, which is still the family surname, and resorted to another single-candidate election. His descendants are still chiefs of Clan Gregor but there is more than a hint of tribal law in the hope expressed in the election proposal of 1822 that Sir Evan’s heirs may prove to be ‘worthy of the respect and attachment’ of this ancient body.
‘The chiefs being now deprived of their jurisdiction, have already lost much of their influence; and as they gradually degenerate from patriarchal rulers to rapacious landlords, they will divest themselves of the little that remains. That dignity which they derived from an opinion of their military importance, the law which disarmed them has abated. An old gentleman that delighted himself with the recollection of better days, related that forty years ago a chieftain walked out attended by ten or twelve followers with their arms rattling. That animating rabble has now ceased. The chief has lost his formidable retinue, and the Highland walks his heath unarmed and defenceless, with the peaceable submission of a French peasant or an English cottager.’19
