How far can you trust documents?

All history-writing, certainly all history-writing dealing with events before the twentieth century, is based on documents.   From the late nineteenth century onwards there are many more aids – sound-recordings, films and so on – but for events before these new inventions, historians have to depend on written documents.   But to say written documents are invaluable is not the same as to say that they are infallible.   Anyone reading a document from 200 or 300 years ago should have the same arm’s-length approach which (for example) anyone reading a political party’s manifesto now ought to have.   Here are some examples where eminent historians have failed to observe this (surely) elementary rule

Robert MacKay wrote in 1829, in his House and Clan of MacKay, that (in modern spelling) the MacKay country, “from Drumholstein, which divides it from Caithness on the north-east, to Kylesku, an arm of the sea dividing it from Assynt on the south-west, is about eighty miles in length”.1   Captain Ian Scobie, who was a Sutherland man, quoted this passage verbatim, without the slightest misgiving, in his 1914 book about the Reay Fencibles.2   Since there were now two documents to prove this “fact”, the author and expert on clan history Sir Iain Moncreiffe of that Ilk, eleventh baronet, Falkland Pursuivant, Albany Herald, Q.C., Ph.D. etc, repeated the information in 1969: the MacKay country was a tract “measuring eighty miles” in length.3   However, any inexpensive map, even in the hands of a less educated reader, shows that it is something under forty-nine miles from the Caithness border to Kylesku – not much more than half of the MacKay-Scobie-Moncreiffe figure.   The documents, unfortunately, are wrong, and anyone quoting from them (without correction) is simply confirming errors.

Again, the Highland ministers writing in the two Statistical Accounts of Scotland (the first one published in 1791-9, and the second one in 1845) often gave exaggerated accounts of the size of their parishes: no doubt the aim was to magnify the amount of work they had to do, and the esteem owed to them for doing it.   Some Highland parishes were certainly very big: several parishes in the Highlands were in fact bigger than Andorra, the independent country in the Pyrenees.   However, exaggeration made them bigger still.   Kilmallie,4 Gairloch,5 and Fortingall6 were claimed to have many more square miles then they had; the statistics of Kilmonivaig,7 Kilmorack,8 Ardchattan,9 Farr,10 Lochbroom,11 and Croy12 (for example) were also considerably overstated, either in the 1790s or the 1840s.   It was presumably these inflated claims in the O.S.A. and N.S.A., clearly wrong though they are (as the map shows), that led Professor Michael Lynch, F.R.Hist.S., F.R.S.E., F.S.A. Scot, and from 1993 to 2005 the Sir William Fraser Professor of Scottish History and Palaeography at Edinburgh University (“the oldest and most distinguished Scottish History Professorship in the world”) to state that the Highland parishes “on average were 400 square miles in area”.13   However, the fact is that there are only a handful (perhaps only four) as big as that: Kilmonivaig, Kilmallie (both of which have been described as the largest parish in Scotland, as has Kilmorack), Lochbroom, and (after 1891, when it took over some of Reay) Farr parish.   There were also a few other very extensive Highland parishes, but the average size of Highland parishes was much less than that.   All this topographical information is freely available in published sources and on the internet; but to some historians the words in a document, especially a venerable one, have a kind of sanctity, and cannot be gainsaid.   The Highlands (that is, four of the old Scottish counties, and ten part-counties) cover so far as I can calculate about 16,300 square miles,14 and contain 16215 parishes: so it cannot take long to work out that the average parish is just over 100 square miles.

The average size of the Highland parishes, then, is only a quarter as big as Professor Lynch said.   A defender of the “400 square miles average” theory might claim that my definition of the Highlands is at fault, and that if a different definition of the Highlands were adopted, perhaps with fewer parishes, a different average size might be found.   Let us then look at the four principal “Highland counties” – Argyllshire, Inverness-shire, Ross-shire, and Sutherland – by themselves.   However small an area a particular observer wished to demarcate as Highland, a definition of the Highlands which did not include these four counties would be less than convincing.   The four named counties contained about 113 parishes,16 covering something like 12,407 square miles.17   (These numbers are only tentative, because some parishes had land in two counties, which meant that not every parish, and not every part of every parish, is always listed under the same county.)   If, however, we take this reckoning as being approximately accurate, it is easy to work out that the average parish size in the four main Highland counties was about 110 square miles.   As for the Hebrides, Nelson’s Gazetteer and the Gazetteer for Scotland both said that the area of the Hebridean islands, from Lewis down to Islay and Gigha (all of them, of course, in the three counties of Ross-shire, Inverness-shire, or Argyllshire) was 2812 square miles;18 and since there are 25 Hebridean parishes, that would make an average size for those parishes of 112.5 square miles.   It appears, therefore, that the average Highland parish, while certainly much larger than the average Lowland parish, was not nearly as large as an incautious reading of some of the O.S.A. and N.S.A. reports might suggest.

In 1994 Professor T. M. Devine, the Research Professor and Director of the Research Institute of Irish and Scottish Studies at the University of Aberdeen, and recipient of many honours including two earned doctorates and three honorary ones (in his own words, “in 2001 I was presented by HM the Queen with the Royal Gold Medal, Scotland’s supreme academic accolade”),19 revealed a similar belief as to the size of Highland parishes; and he gave an example of what he meant.   After warning his audience that Highland history “has long been shrouded in romance and myth through the ingenious efforts of Victorian writers who virtually invented a Gaelic past which fitted in with the assumptions and expectations of their readers”, he said that now Highland history was “based on careful examination of contemporary documentary evidence”;20 and remarked that “there were few Highland parishes under 400 square miles in extent, a typical example being Harris which was forty-eight miles by twenty-four”.21   (Those asserted dimensions, if consistent, would of course mean that Harris contained over a thousand square miles.)   In fact, as we have seen, there were very few Highland parishes, probably only four, as large as 400 square miles in extent (so the fact is that all Highland parishes except four were smaller than 400 square miles): certainly Harris was nowhere near that size.   Professor Devine’s remark about Harris is clearly based on the Harris minister’s comment in the O.S.A. that his parish, “from the northern to the southern extremity, along the common track of travelling by land, and the course of navigation through the Sound, will be at least forty-eight miles long: its breadth varies much.   Near the northern extremity it is 24 miles; from thence to the Sound, it may be at an average from 6 to 7; and, of the Sound, navigators calculate the breadth as well as length at 3 leagues [or 9 miles].”22   The N.S.A. Harris reporter thought the length of the parish was greater still: “its extent from north to south is fifty miles”.23   The acceptance of these figures constitutes another case of over-reverence for a document, and indeed for an extreme interpretation of it: the width of “twenty-four miles” alleged for part of Harris by the O.S.A. minister only applies to a small section of the parish, while the “average of six to seven miles” is much nearer the truth.   The atlas shows that though the greatest width of Harris (both the island and the parish) may well be about twenty-four miles, the greatest length of the parish of Harris is at the most twenty-six miles, from the northern point of Loch Resort to the southern end of Berneray (rather than the forty-eight miles claimed in the O.S.A., or the fifty alleged in the N.S.A.).   Apart from that, the minister made it (and the map makes it) clear that most of the space inside these maximum measurements was occupied by extra-parochial water.   Two authoritative calculations of the area of Harris are 193 square miles (Gazetteer for Scotland 2002-13), and 195 square miles (Nelson’s Gazetteer 1941).   Whichever figure is correct, Harris clearly has fewer than half of the alleged 400 or more square miles.

Professor T. C. Smout, the present Historiographer Royal in Scotland (i.e. the leader of the profession), and formerly professor of history at Edinburgh University and at St Andrews University, did not suggest an average size, or a minimum size, but he did claim that the “Highland parishes were of immense size”, and gave four examples.24   The first was Kilmallie, where the O.S.A. minister had asserted that “the length . . . is about sixty miles in a straight line”,25 a statement endorsed by the N.S.A. minister: “the length . . . is about sixty miles”.26   There were thus two documents giving the length of Kilmallie: and Professor Smout was presumably relying on them when he wrote that “Kilmallie was sixty miles long”.27   It may well be as we speculated earlier that the O.S.A. minister was tempted to over-estimate distances, and therefore his own hard work, as he went about his parish business, and it may also be that the N.S.A. minister found it the easiest option to copy his predecessor; but anyone writing now can look at a map and see that Kilmallie’s length is about thirty miles, rather than sixty.28   So where an allegation in a historical document, and the easily discernible fact, are at odds with each other, it is the fact that is rejected by academic orthodoxy.   The document cannot be denied.

Dr Smout also claimed that Glenorchy was nearly as big: “Glenorchy was sixty miles long and twenty-four miles wide.”29   No doubt a document somewhere gave these “facts”, though I have not found it yet (Dr Smout supplied no reference; and even the O.S.A.30 and N.S.A.31 only claimed a length of twenty-four or twenty-five miles).   However, the map shows clearly that Glenorchy was at the most only about thirty-one miles long and no more than fourteen wide (half the length, and not much more than half the breadth, given by Dr Smout).32   Dr Smout’s third example was Buchanan parish, which he claimed was nine miles across.33   The map shows clearly that the width of Buchanan was at the most six miles; the Gazetteer for Scotland gives the same figure – six miles.   Even in the statistical accounts, the O.S.A. minister34 had claimed only six miles of breadth, and the N.S.A. report35 only five miles.

So the historian’s motto must be – look at all the documents you can find, by all means, but don’t assume that every document must tell the truth.


1 Robert MacKay, History of the House & Clan of MacKay, Edinburgh, 1829, 1.

2 Scobie, Captain I., An Old Highland Fencible Corps, Blackwood, Edinburgh, 1914, 8.

3 Moncreiffe, Sir Iain, The Highland Clans, Barrie & Rockliff, London, 1967, 175.

4 O.S.A. VIII 47, Kilmallie Inverness-shire.

5 N.S.A. XIV 90, Gairloch Ross-shire.

6 N.S.A. X 527, Fortingall Perthshire.

7 O.S.A. XVII 543, & N.S.A. XIV 503, both Kilmonivaig Inverness-shire.

8 O.S.A. XX 401, 402, Kilmorack Inverness-shire.

9 N.S.A. VII 469, Ardchattan Argyllshire.

10 N.S.A. XV 69, Farr Sutherland.

11 O.S.A. X 461, Lochbroom Ross-shire.

12 N.S.A. XIV 445, Croy Inverness-shire.

13 Lynch, Professor Michael, Scotland A New History, Pimlico, London, 1996, 364.

14 16,300 square miles is my best calculation: it must be approximately correct.

15 I have explained elsewhere how I reached the total of 162 Highland parishes: this figure too must be approximately correct.

16 I explain elsewhere that I have taken Argyllshire to have 35 parishes, Inverness-shire 32, Ross-shire 33, and Sutherland 13 (as in the O.S.A.) – a total of 113.

17 Sutherland 2028 sq. miles, Ross-shire 3078, Inverness-shire 4088, Argyllshire 3213 – total 12407.

18 The Hebrides – that is, Lewis, Harris, the Uists, Benbecula, Barra, Skye, the Small Isles, Coll &

Tiree, Mull, Colonsay & Oronsay, Jura, Islay, Gigha & Cara, plus the neighbouring islets:

Nelson’s Gazetteer and Gazetteer for Scotland.

19 website etc (the Edinburgh University website). At the event

“celebrating the career of eminent historian Professor Sir Tom Devine, Professor Devine was joined by Gordon Brown, former Prime Minister of the U.K.”

20 Devine, Sir T., Clanship to Crofters’ War, Manchester University Press, Manchester, 1998, preface.

21 Devine, Sir T., op. cit., 100.

22 O.S.A. X 343, Harris Inverness-shire.

23 N.S.A. XIV 155, Harris Inverness-shire.

24 Smout, Professor T. C., A History of the Scottish People 1560-1830, Collins, London, 1970, 461.

25 O.S.A. VIII 407, Kilmallie Inverness-shire.

26 N.S.A. XIV 117, Kilmallie Inverness-shire.

27 Smout, op. cit., 461.

28 The Gazetteer for Scotland, 2002-13, said that Kilmallie’s greatest length was 29 1/8 miles, and its greatest breadth 30¼ miles.

29 Smout, op. cit., 461.

30 O.S.A. VIII 336, Glenorchy Argyllshire.

31 N.S.A. VII 83, Glenorchy Argyllshire.

32 The Gazetteer for Scotland, 2002-13, said Glenorchy’s “utmost length . . . is 31½ miles”, and its greatest “breadth 13 5/8 miles”.

33 Smout, op. cit., 461.

34 O.S.A. IX 13, Buchanan Stirlingshire.

35 N.S.A. VIII 89, Buchanan Stirlingshire.

Leave a Reply

Your email address will not be published. Required fields are marked *