‘THE Highlands” are a given in describing Scotland and its history. They are variously seen as the location of a kind of landscape, a distinct language or a semi-autonomous collection of distinct cultural practices. Sometimes they are one of these, sometimes both, sometimes all three.

But there are good reasons to challenge all these categories and to insist on being more clearly aware of the history of the concept of the “Highlands” ­itself, which has changed over time and is a historic ­category like any other. My new book, Scotland: The Global History, published by Yale University Press on July 26, was written to extend and challenge, as well as to describe the role that Scotland has played in the world, and in doing so it examines the way we describe Scotland.

The Highlands are sometimes aligned with the terms “Gaeldom” or the “Gàidhealtachd”, ­usually without defining – certainly in the case of ­“Gaeldom” – whether or not exactly the same area is being ­described. This is at once problematic. In 1600, Gaelic was spoken north of Forth-Clyde in western and central Scotland, but not so on the east. ­However, it was still spoken in Galloway, which is ­almost never included as “Gaeldom”. By 1800 the ­language had ­retreated north by some 40 miles and had ­disappeared from Galloway and Carrick, ­although there appear still to have been speakers there at the time of the Jacobite Rising in 1746, and Gaelic ­survived in ­Arran into the 20th century, where parts of the island were still more than 50% Gaelic speaking in the 1901 ­census, by which time many areas far to the north that are still identified as “Gaeldom” had all but ceased to speak the language.

The next distinction is one of geography: the ­“Highlands” are associated with the line of the ­Highland Boundary Fault, which runs conveniently close to demarcating the area of northern Scotland which primarily spoke Gaelic in 1746, thus enabling a synergy of the linguistic and geographical ­boundaries of “Highland” space. Unfortunately, the Highland Boundary Fault does not define the Highlands but was defined by the place they already held in the ­British imagination when the “Fault” was identified, for we owe the identification of this geological ­phenomenon to the English geologist George Barrow (1853-1932). The “Highland” boundary fault is itself a term of art framed by the expectations of history.

What about the clear distinctions of cultural ­practice between the “Highlands” and “Lowlands”? The historian Allan Macinnes describes the clans as “Anglian, Anglo-Norman and Flemish as well as Celtic and Norse-Gaelic in origin”, which just about covers everyone living in Scotland. Earlier commentators such as John of Fordoun (c1384) are often quoted in support of the concept of Highlanders as “wild Scots”’ but recent scholarship suggests that Fordoun was simply using a rhetorical formula rather than making an ethno-cultural point.

“Highlanders” were again identified as “Wild Scots” by John Mair (1467-1550), and a narrative of Highland savagery and Scots civility took root some time during or after the end of the career of ­Alexander Stuart, the Wolf of Badenoch (1343-1405) and the Battle of Harlaw (1411): in other words, it seems to have arisen in response to particular ­political circumstances. Yet Alasdair Mór mac an Rígh – the Wolf – was the son of Robert II, King of Scots, while at Harlaw, Domhnall, Lord of the Isles, was in competition for the Earldom of Ross, also claimed by the House of Stewart, and was beaten off by one of its members, Alexander, the Earl of Mar.

The politics of this classic “Highland v Lowland” conflict were in fact over a great earldom under the Crown.

Undoubtedly the balance between ­kinship and association and ­feudal ­practice varied between Scottish ­magnates and their followers, but this was a matter of emphasis rather than a disjunction.

Remote lords were also Scotsmen ­“under the Kingis of Scotland”, and this was normative. The great territorial ­lordships may have been better at ­raising men in the north than in the south of Scotland, but Erskine, Forbes or Ogilvy, Drummond, Oliphant or ­Murray, are hardly families fit for Gaelicising, and indeed attempts to define great ­magnates like the Gordon earls of Huntly as ­having “superiorities within Gaeldom” while not being Highland (though knowing ­Gaelic) are signs of the awkward ­definitions still commonplace with respect to ­“Highland/Lowland” terminology.

Lord Lovat, Chief of the Name of Fraser, was ­educated at Aberdeen Grammar School, as were some of the academic Gregory dynasty, cousins to Rob Roy, that ­archetypical “Highlander”, who was thus kin to ­several distinguished men among the urban Scottish professoriate.

One of the major issues that any ­concept of the Highlands has to deal with – but ­often does not – is the nature of “Highland” social behaviour.

Chiefs of the Name in the West ­Highlands, such as Cameron of ­Lochiel, mixed as easily in the ranks of European nobility as did those from the Scots-speaking areas of Scotland such as the Marquess of Montrose, Chief of the Name of Graham. There is much more to be said, but any clear dividing line is difficult to draw. And this was what the British Army and government found in the Jacobite era.

For General Wade in the 1720s, the Highlands had no precise county ­boundaries: various “lines” had been drawn to suggest where they might start, but even these differed.

In the “Memoriall Anent the True State of the Highlands as to their Chieftenries” (1745), the Duke of Perth was at the same time described as of “no Claned familie” while also being “the head of a ­Considerable Number … of the Name of Drummond”.

Forbes of Culloden (though he ­included other Dukes in a list of ‘the clans’) ­described the Murrays of Atholl as “no Highland Family” while others describe them and their lands as “Highland”, while a 1715 explanation of the risk posed by Jacobitism appears to render the whole of Scotland as tribally “Highland”.

In William Roy’s 1747-52 military ­survey, the Highlands appear to include all of Scotland north of the Forth-Clyde line, and indeed Roy (unlike ­Lemprière, whose map was “one of Wade’s ­surveillance maps”) tends to exaggerate mountain contours even south of Forth.

In the wake of the 1745 rising, James Campbell wrote to the Duke of Argyll that “Highlands is no certain description. The disarmed countys would do”.

Campbell’s problem was that which I have mentioned above, the ­disparity ­between the “Highlands” and the fact that the counties of Scotland’s civil ­administration had been formed without this “Line” in mind. As indeed they had been.

For British officers in Scotland ­after Culloden such as James Wolfe ­matters could be even more stark: for him ­“Highlander” was a term extended to the Earl of Kilmarnock and the Irish brigade officer Brigadier Walter Stapleton. It was thus not just the “disarmed countys” but opponents of the British government as a whole who could receive the epithet of “Highlander”.

The National: The Battle of CullodenThe Battle of Culloden

It may be objected that Highland elites were one thing, rank and file ­another. Yet in the Jacobite Rising of 1745, many “Lowlanders” were incorporated in “Highland” regiments. The army as a whole was called “Highland” because of the signification of true Scots patriotism attaching to the north, a cultural ­typology that can be traced back as far as the courts of James V and Mary ­expressing patriotism through the “aboriginal attire” associated with the north; but the army received its orders in English.

The so-called “Highland line” as a more or less defined conflation of language, culture and geography seems to have its origins in the legislation of the post-Culloden era. Notably, this legislation not only bisected existing counties, but also was not consistent in either its definition or its implementation.

The bisection of counties was ­ultimately the legislative decree of the British ­Parliament, legislation which clearly ­suggested an incompatibility ­between Scottish civil administration and the divided “Highland and Lowland” culture of what it was administering.

The initial Act of Proscription in 1746 laid the foundations for the concept of the “Highland Line” through its ­banning of the carrying of arms from those “within the shire of Dunbartain, on the north side of the water of Leven, Stirling on the north side of the river of Forth, Perth, Kincardin, Aberdeen, Invernefs, Nairn, Cromarty, Argyle, Forfar, Bamff, Sutherland, Caithnefs, Elgine, and Ross”. Notably Perthshire, Aberdeenshire and Forfarshire (Angus) were included in the entirety.

This is the source of the maximalist understanding of the “Highlands” as all Scotland north of the Forth-Clyde line. In 1784, the formal concept of the ­“Highland Line” as a Parliamentary fact was ­introduced by the Wash Act of ­Parliament “as a convenient demarcation for the whisky-still taxation”. It was held to be: “A certain line or boundary beginning at the east point of Loch Crinan, and ­proceeds from thence to Loch Gilpin …along the west coast of Loch Fyne to ­Inveraray and to the head of Loch Fyne from thence … to Arrochar … to ­Tarbet; from Tarbet … straight eastward on the north side of Loch Lomond, to ­Callander … north eastward to Crieff … and to ­Ambleree [Amulrie] and Inver to ­Dunkeld; from thence along the foot or side of the Grampian Hills to Fettercairn … northward … to Kincardine O’Neil, Clatt, Huntly and Keith to Fochabers …westward by Elgin and Forres, to the coast on the river Findhorn, and any place in or part of the county of Elgin which lies southward of the said line…”

This precise definition is of course rather less precise than it seems. Its very county-crossing complexity and the tiny settlements held to be its border towns should give one pause for thought. Was Crieff in or out? Half in or half out? Did the inhabitants of Dunkeld speak Gaelic on alternate days of the week and did anyone south of the Findhorn in the county of Elgin need an interpreter to ­visit the county town?

The National: DunkeldDunkeld

These are ridiculous questions, but they reveal how the definitions of the “Highlands” on which we still rely were largely arbitrary and factitious. What was “Highland” was rendered even more problematic by the introduction of an “Intermediate” zone in 1797, neither Lowland nor Highland, including areas such as southern Argyll and much of the counties of Aberdeen and Banff. British legislation also found the “Highlands” problematic as a concept.

As it is most popularly used today, the “Highlands” tends to include all Argyll, Ross and Cromarty, ­Sutherland and ­Inverness, and – to a less clear and ­unstable extent – the upland parts of Perthshire, Aberdeenshire and ­sometimes Angus and Stirlingshire.

Immediately there are problems here: Caithness is explicitly excluded, but is seldom thought of as “Lowland”. Several counties are bisected on the basis that some parts of them are hilly, though this does not earn Dumfries and Galloway or parts of the eastern Borders the term “Highland”, even though Gaelic was ­spoken in the former longer than in some of the areas still popularly categorised in this way.

One distinguished Scottish historian also adds Banffshire, Dunbartonshire, Kincardine and Nairnshire to the list of counties bisected by the Highlands. Even those describing where they often cannot agree on the geography.

If one examines early maps of ­Scotland, it is its counties and/or their associated earldoms which are most frequently marked, rather than any internal division of landscape, culture of language: Paolo Forlani’s Regno Di Scotia, Abraham Ortelius’s Scotiae tabula, John Leslie’s Scotiae Regni Antiquissimi, Gerhard Mercator’s Scotia Regnum, and many others all display a Scotland divided up only by feudal authority or county governance.

As Charles Withers, until recently ­Scotland’s Geographer Royal, put it many years ago, “the Highlands have been ­created” as “a sort of natural or ­anthropological curiosity’, which later became “a recreational commodity”.

The definition we inherit is one we should question more closely.

Murray Pittock is a Scottish historian and holds the Bradley Chair of English Literature at the University of Glasgow