to the
Derivation of Scottish Surnames

William L. Kirk, Jr., Ph.D.
1 July 1992

The whole subject of Scottish surnames, and their connection with kinship, is surrounded by complications. Professional etymologists have attempted to classify how Scottish names came about. And, whereas the work is largely successful, the many exceptions and the metamorphosis of names make the study of Scottish names an unfinished one.


The use of surnames seems to have commenced in France about the year 1000, and surnames were introduced into Scotland through the Normans a little over one hundred years later, although the custom of using them was hardly common. The first official reference to the practice is from a general council held at Forfar in 1061, during the reign of Malcolm Ceannmor (1057- 1093). Malcolm directed his chief subjects to create surnames from the names of their territorial possessions.

Thus, the first people in Scotland to acquire fixed surnames were the nobles and great landowners, who called themselves, or were called by others, after the lands they possessed. The form of the names was Norman "de"; for instance, Robert de Brus (Brus in Normandy), John de Balliol (Balleul-en-Vimeu in Picard), William de Buchan (Buchan in Scotland), Christopher de Seton (Sayton in Scotland), William de Kirkhaugh (Kirkhaugh in Northumberland, my family), etc. One interesting example comes from the surname of Maxwell. Sometimes confused with the Norman, Maccusville, the name actually came from Maccus, the son of Unwin, a Saxon lord, who obtained a salmon pool on the river Tweed near Kelso Bridge. The pool was then called, Maccus's wiel (pool). The adjacent lands got the name, and the descendants of Maccus became known as, "de" Maccuswel, and, subsequently, became the powerful Maxwell family of Dumfriesshire and Galloway.

But, since not too many persons held significant lands, place-names quickly began to refer to the region or district from where a family originated. For example, Andrew de Moravia (of Moray), William de Douglas (of Douglas - dubh glas), Adam de Haddyngton (of Haddington), etc. As the need for a surname became more pressing, residents of the burghs often adopted street names, such as, Henry de Fishergate, Henry de Cunigstrete, etc.

So, the first surnames were place-names and originated with a man who lived in or came from a place, sometimes a big district like Moray (Murray) or Lothian, often a small rural community. A proprietor was particularly likely to take his name from his estate, but tenants also often took their names from the estate where they lived. Clearly many individuals, and ultimately of families, could originate in the same place, and take their names from it, without being related to each other. Besides, the same or similar names were given to different places, and so individuals or families who came from different parts of the country, and shared neither blood nor territorial affinity, could nevertheless have the same surname. Thus, anyone called Calder (or its variant, Caddell) may derive from an ancestor resident in Calder in West Lothian, Calder (or Cadder) in Lanarkshire, Calder (or Cawdor) in Nairnshire, or Calder in Caithness. Similarly, there is no necessary relationship among the many families called Blair, a place-name which occurs in at least a dozen different areas.

Official and Trade Names

There are surnames that derive from a craft, occupation, or official station. Smith, which is the most common name in Scotland, is an outstanding example. Wright, Baxter or Baker, Tailor, Carpenter, Mason, Shepherd, Slater, are among many others. It would clearly be an unparalleled absurdity to think that one smith was the ancestor of all the people now bearing the name Smith. The same is true when a name of this type arose in the Highlands, where a designation Coinneach Gobha (Kenneth the smith) produced the surname Gow. The Norman form of these names was "le"; so for example, from the Ragman Roll: Symon le Glover, Robert le Taillor, Walter le Goldsmith, Aleyn le Barbur, William le Barker, etc. As with the "de" in the place-names, the "le" was eventually dropped, giving us the modern form of the names.

Many offices were hereditary in feudal as well as in later times. The Stewarts, for example, were the first to be named after their office alone. Although, the first Stewart, Alan, had a son who called himself Walter Fitz Alan, and his son called himself Alan Fitz-Walter.

Offices associated with hunting and the king's lands yielded many names, for example: Forest, Warren, Hunt, Park, Woodward, etc.


The great prevalence of certain surnames, in some small towns and villages, led to the use of to-names, "other names," from the Old English, t -nama. To illustrate this, Black cites the following story.

A stranger had occasion to call on a fisherman, named Alexander White, living in a Buchan fishing village. But the stranger was ignorant both of the fisherman's to-name and his house. Unfortunately there were many persons of that name in the village. Meeting a young woman, he asked:

"Cou'd you tell me fa'r Sanny Fite lives?"
"Filk Sanny Fite?"
"Muckle Sanny Fite."
"Filk muckle Sanny Fite?"
"Muckle lang Sanny Fite."
"Filk muckle lang Sanny Fite?"
"Muckle lang gleyed Sanny Fite," shouted the stranger.
"Oh! It's 'Goup-the-lift' ye're seeking," cried the young woman, "and fat the deevil for dinna ye speer for the man by his richt name at ance?"


"Could you tell me where Alexander White lives?"
"Which Alexander White?"
"Big Alexander White."
"Which big Alexander White?"
"Big tall Alexander White."
"Which big tall Alexander White?"
"Big tall cross-eyed Alexander White," shouted the stranger.
"Oh! It's '(Stupidly gaping person)-the-thief' you're seeking," cried the young woman, "and why the devil didn't you ask for the man by his right name in the first place?"

Another example from the Borders shows how difficult it can be for a researcher to identify persons if the researcher doesn't know that: John Bell was called Quhitheid, Edward Bell was called the Dansair, John Bell was called Ranyis Johnne, and Andrew Irvin was called Tailyeour curst Geordie.

So accustomed were many Scots to being recognized only by their to-name that it became their official name. Familiar examples are: Black, Little, White, Campbell (caimbeul or crooked- mouthed), Cameron (camshron or crooked-nosed), and Meikle (big). Clearly, almost any surname of this type could arise quite independently at different times and in different places. So, again, caution should be taken in arguing a relationship between persons simply because of the similarity in their names.

Patronymic Names

These are the names usually represented in Lowland Scotland by the suffix, son. But with them must be taken the forenames, which became surnames, and are really truncated patronymics, such as, Henry, Mitchell (for Michael) and Arthur. The development of names of this type was more subtle than it was in the first three categories. Designations changed generation by generation. Robert's son might be John Robertson, his son Andrew Johnson, his son Peter Anderson, etc. This system was general in all the northern lands, and it extended to women, with forms that would translate as, for example, Elspeth Johnsdaughter. In Shetland, patronymic names persisted in many families until the nineteenth century, so that there are examples such as, Arthur Anderson, son of Andrew Robertson, or James Manson, son of Magnus Oalson.

Strangely, women sometimes used the suffix, son, as in Isabella Johnson, daughter of John Williamson.

Throughout most of Lowland Scotland, genuine patronymic practice ended after the fifteenth century. This happened when an individual decided, or some authority decided for him, that he would adopt his father's patronymic as his own surname. Thus, the son of John Robertson called himself not Andrew Johnson but Andrew Robertson. And from that point, Robertson became the surname of his descendants. It was clearly a matter of chance in which generation the patronymic became the final surname. So that in the example just given, had the decision been taken a generation later, the surname of the family would have been Johnson, not Robertson.

The above example shows the dangers of attempting to use surnames of patronymic form as guides to remote ancestry. Half-a-dozen Robertsons are probably descended from half-a-dozen different Roberts who lived in different parts of the country at different times, and have no kinship with each other. It is only by chance that they are called "Robertson" and not Johnson or Anderson.

In the Highlands and Galloway, where "son of" was denoted by the prefix, Mac, rather than by the suffix, son, patronymic names were more common than they were in the Lowlands. The Mac could be prefixed to craft names as well as to forenames, giving, for example, Mac an t-saoir, son of the joiner, which became Macintyre, also, for example, the group of names denoting descent from an ecclesiastic: Macnab, Mactaggart, Macpherson and Macvicar, meaning son of the abbot, the priest, the parson, and the vicar, respectively.

Names were sometimes carried into two or three stages by using Vic (Mhic, the genitive of Mac). Sixteenth-century examples are Angus MacDonald Vic Angus, son of Donald MacAngus, and Alastair MacAllane Vic Ane Vic Coull. Here is an interesting example from 1617: Hector MacGorrie Vic Achan Vic Allester Vic Ean duff, son of Gorrie MacAchan Vic Allester Vic Ean duff. In women's names, Nean (nighean), meaning "daughter of" could replace Mac, giving patronymics like Margaret nean Ean glas Vic Ilespig.

Names of this type, recorded in official registers, were not surnames, and, while individuals so recorded may have had surnames, their surnames were not used in the record. Identification may consequently be difficult for the researcher. For instance, but for their territorial designation of Lochiel, how would anyone know that the men recorded in the mid-sixteenth century as Ewan Allanson, John Dow, his son, and Ewan, his grandson, were in fact all Camerons?

The use of genuine patronymics in records continued well into the eighteenth century. For example, in South Uist, in 1721, there were names like Jonn MacEwan Vic Ean Vic Charles, and Murdo MacNeil Vic Ean Vic Duill. In some Highland families the prefix Mac meant not only son of but also descendant of. Such a patronymic, persisting generation by generation, became a surname. An example is MacDonald. Angus of the Isles, in the later thirteenth century, was the son of Donald, and his successors retained MacDonald, not so much as a surname in the modern sense, but as a mark of their descent. However, most of the many MacDonalds of later times had no kinship with the descendants of Angus, or necessarily descended from anyone called Donald.

The process by which the genuine surname replaced personal designations that changed from generation to generation, took a long time to complete. There are instances throughout the sixteenth century, in almost any part of the country, that show some people with more than one designation, and it is difficult to say which, if any, of the designations was a real surname. If a man had a name of patronymic form and a craft name, like Robertson or Pottar, it is difficult to tell which is a genuine patronymic and which is no more than the name of his actual occupation. William Davidson or Litstar, and Matthew Paterson or Litstar, were both priests, but the Litstar (i.e. dyer) is the surname. In the case of Andrew Wilson or Tailor, the son of Andrew Wilson, the name Wilson was the surname, not Tailor. Also there are surnames originating from a place-name combined with a patronymic, such as Alexander Murray or Anguson.

Miscellaneous Naming

Under the Clan system, prevailing on the Borders and in the Highlands, men often assumed the names of their chieftains or feudal superiors. There was a compelling reason for this sometimes, as when the earl of Menteith -- for some reason -- declared war against all men, except the king and those of the name of Graham. This attracted considerable popularity to that surname in the district of the Lennox and Menteith. But, because of the adoption of Clan names, when a man moved from one estate to another, he might change his name. In the 1750's, John MackDonell was really a Campbell, having changed his name to that of MacDonnel upon coming to live in the bounds and under the protection of the MacDonnel family of Glengary. The use of the landlord's name explains why, in the 1580s, a servant of the Earl of Huntly was called, Gordon or Page -- Gordon because his master, Huntly, was a Gordon and Page, he (or an ancestor) being a page.

I think that there is a preoccupation in this country with trying to be identified as a descendant of a Highland family or Clan. The fact that, at any given time, the Highlands represented a small minority of the population of Scotland, should make it obvious that most persons of Scottish extraction came from lowland families. When discussing this with people I meet, I usually am told that some family ancestor had a Highland name. From what has been written here so far, it should be clear that this use of name to establish a relationship is suspect.

Further, many Highland names are also lowland names. Gordon is a particularly good example of a name usually regarded as a Highland Clan name without there being any grounds for considering it so. The name originated with Richer de Gordun, lord of the barony of Gordon in the Merse. But when a branch of the family later settled in Strathbogie, the whole country around soon became full of men calling themselves Gordon. This name is now widespread throughout Scotland, and there is even a Polish family of Gordon. Generally speaking, if you think you have descended from a Highland family, or Clan, you are probably wrong.

Occasionally there is a switch from one kind of designation to another. In the 1470's, the three sons of Thomas Soutar were David, John and Thomas Thomson, and whether their descendants were Soutars or Thomsons is unknown. There was, also, an inclination for people to give up the more outlandish names to adopt names that were familiar or distinguished. It seems, to take a curious example, that the Scandinavian, Sigurdsson, which became Shuardson in Shetland, was Scotticised as Stewartson and finished as Stewart. So, not all Stewarts were related to a Stewart king. Further some Stewarts probably descended from the stewards of this or that estate, not from royal stewards.

Name changes sometimes occurred on inheriting or otherwise acquiring landed property. Sometimes charters even specified that the proprietor must bear a certain name. For similar reasons, husbands sometimes took their wives' names. In any of such cases, the surname ceases to be a guide to more remote ancestry.

In ancestral research, variations in the spelling of a name can be confusing, particularly if there is no significance in variant spellings of the same name. For example, prior to sixteen hundred, my ancestors' family name was spelled, Kirkhaugh, Kirkhaucht, Kirkalch, Kirkhalche, Kirkhaulch, Kirkhauch, Kyrkhauch, and Kyrkhalch; all of which would have sounded alike when spoken. Also, in Highland names, there is no significance in the variation between Mac and Mc, and between the use of a capital or a small letter in the second part of the name, such as, MacLean and Maclean. The variation in spelling is easy to understand when one realizes that most people in the middle ages could not read or write. If a person could not spell their name, someone recording the name did so phonetically. Different scribes used different spellings, and the same scribe might use different spellings within the same document. Even an individual might spell his own name in different ways on different occasions. In fact, until about two and a half centuries ago, the spelling of proper names, and many other words, was quite arbitrary.

So no significance should be attached to different spellings as indicative of ancestry or relationship. It sometimes was simply a matter of chance that a family adopted a particular spelling, while other families, possibly closely related to them, adopted different spellings.

On the other hand, similar spellings of different names may lead a researcher astray. Livingston is a Lowland name, of West Lothian origin, but Livingstone is a Highland name, and there is no relation between the two. Similarly, Johnson is a patronymic name, but Johnston derives from John's toun or settlement, while Johnstone might originate in the name of some landmark. Some Camerons -- perhaps most -- are Highland Camerons from Lochiel, but others take their names from the places called Cameron in Lothian and Fife. Dewar and Shaw are other examples of names with distinct Highland and Lowland origins, and Dunn, while it may derive from the Gaelic, donn, may equally well derive from the place Dun, in Angus. The distinction between a Highland and Lowland origin has often been effaced when a Gaelic name has been translated into English, so that MacNeacail becomes Nicolson, and MacGille-mhoire becomes Morison -- which means that they are added to the host of unrelated patronymics spanning the whole country, with no affinity among them.

Compilers of official records did not always have a consistent preference for a surname, and when there was consistency, it was often based on utilitarian considerations, by using a designation that most clearly identified the individual. On the other hand, it may be that the official recording of names had a certain influence in stabilizing surnames, and in some areas the establishment of the Register of Sasines in 1617 clearly had some effect. Variation of names further declined because ministers, in their registers of baptisms, marriages and burials, preferred names which they did not feel was outlandish. In the Highlands, many names indicative of remote ancestry were lost because ministers had difficulty in recording Gaelic names unfamiliar to them. They substituted names which had well established Anglicized forms.

Variations that survived into the nineteenth century were further curbed by the compulsory registration of births, marriages and deaths in 1855. Registrars began to insist that individuals use the same surname as his father.


Naming conventions of the forename (Christian name) can sometimes help in the tracing of families. A very widespread custom was for the eldest legitimate son to be named after his paternal grandfather, the second son after his maternal grandfather, and the third son after his father. The eldest daughter took the name of her maternal grandmother, the second, that of her paternal Grandmother, and the third was named after her mother. Mothers' or grandmothers' surnames were sometimes used as forenames for boys: Graham, Murray, etc. But even with these conventions difficulties can arise. For example, my grandfather named my father, William. My father named me, William, I named my eldest son William, and he must now name his eldest son William, and so on forever. Within a few generations, with early marriages and long lives, it may become difficult to sort out the individuals.

Also, it was a not an uncommon practice in the late Middle Ages for two brothers to have the same forename; a custom which has caused confusion for genealogists. There were two Davids in the family of the father of Cardinal Beaton. One of my own ancestral families had two Williams. But William de Viteri Ponte had three sons named William, distinguished as, Willelmus primogenitus, Willelmus medius, and Willelmus junior. King James V had three illegitimate sons named James, and on 26 February 1532 he wrote to Pope Clement VII asking him to declare them eligible to hold ecclesiastical dignities.

I think that there are two principal reasons why these duplicates came about, besides caprice. Sometimes a child died young, and the duplicate named brother was his replacement in the family. But perhaps most commonly, was the necessity of naming a child after each of its grandparents, and, as both happened to have the same forename, the only way of bestowing the honour and getting out of the difficulty was by having the double set.


Black, George F. The Surnames of Scotland: Their Origins, Meanings, and History New York: Publishing Center for Cultural Resources, 1946.

Jamieson, John. A dissertation on the origin of the Scottish language. In Jamieson's dictionary of the Scottish language Edinburgh: William P. Nimmo, 1867.

Kirk, William L., Jr. From Kirkhaugh to Kirk In The Augustan Society Omnibus Book 13, ed. Sir Rodney Hartwell. Torrance: Directors of the Society, 1991. 134-136.

Maxwell, Sir Herbert. A history of Dumfries and Galloway Edinburgh: William Blackwood and Sons, 1900.

Scottish Naming Customs Dumfries & Galloway Family History Society Newsletter. November 1988, p. 10.

William L. Kirk, Jr., Ph.D. [1 July 1992]

Clan Macrae Online