History of Scots to 1700

History of Scots to 1700

2. The origins and spread of Scots (CM)

The distinctive linguistic mix of Scots was created by the language contacts of the late OE and Pre-Scots (PreSc) periods, and we must therefore consider the historical background to the language at this time in some detail. Unfortunately, the early history of Scots is obscure, to the extent that we are not certain whether the language descends primarily from the Anglian of Lothian or from the Anglo-Danish of Yorkshire four or five hundred years later, or from a mixture, in unknown proportions, of the two.

Because of the depredations of the Vikings, little OE survives from any part of Northumbria. For the same reason, and also because of the carrying off of the national records by Edward I of England (to be lost in subsequent centuries), the destruction of the great monasteries in Border warfare, and the vandalism of the Reformation, the documentary history of Scotland is thin, in any language, for the crucial centuries in which Scots emerged. And indeed the documentary record may not have been so rich to begin with as further south. Our understanding of this period therefore relies heavily on place-names and on archaeology. Since the evidence of archaeology is still coming in, it plays a particularly influential role in shaping new perceptions.

2.1 The Angles

The Angles were a Germanic people from what is now Schleswig-Holstein (the southern part of the Jutland peninsula), on the fringes of the Roman Empire. Together with neighbouring tribes of Saxons and Jutes they first came to Roman Britain as auxiliary troops. (There is however no specific evidence of Anglo-Saxons amongst the Roman legions in what is now Scotland.) When the Romans withdrew from Britain at the beginning of the 5th century, the Romanised tribes retained some links with Roman institutions through the Christian church, but in terms of secular power Roman Britain broke up into territories ruled by warlords, and was exposed in the west and north to the threat of the unconquered tribes, which in Scotland meant the Picts and Scots north of the Antonine Wall (see Map 1).[6] The Anglo-Saxons were invited into this volatile situation as mercenaries. There is now a substantial body of archaeological evidence for the presence of individual Anglo-Saxon strong-arm men or small bands of adventurers in Scotland, both north and south of the Antonine Wall, from the 5th century onwards, as well as Anglo-Saxon craftsmen whose influence is seen in Pictish sculpture (Proudfoot and Aliaga-Kelly, 1996).[7] Within a century and a half, the Anglo-Saxons established themselves as conquerors over most of what was to become England, with the northernmost kingdom of the Angles, Bernicia, extending into southern Scotland.[8]

The kingdom of Bernicia was founded in 547 and expanded after a victory at Degsastan in 603 (identified as Dawston in Liddesdale or Addinston in Berwickshire). The kingdom of Northumbria was created shortly afterwards by the incorporation of Deira to the south. Edinburgh was perhaps captured from the Britons during the reign of Oswald (633-41), allowing the occupation and settlement of Lothian (in the narrow sense of the lands between the Forth and the Lammermuirs).[9]

It had until recently been accepted that the Angles established themselves in the south-east of Scotland at a relatively late date – the fourth or fifth generation from the original invaders of Britain (Lowe, 1999: 10). This chronology was based on the apparent absence of pagan place-names,[10] and on the absence of names of the singular -ing type, which in England were thought to represent the earliest layer of Anglo-Saxon names, and the probable absence also of the nominative plural ‑ingas (Nicolaisen, 1976: 71). However, the chronology of the earliest Anglo-Saxon names has been questioned, with -hām names emerging as the earliest, and it has been pointed out also that pagan names are generally lacking north of the Humber, apparently because of later Christian renaming, so it may not be necessary to suppose that the first settlements in what is now Scotland were substantially later than those in the rest of Bernicia (i.e. perhaps three generations after the first settlement in England) (see Cameron, 1996: 66ff., 119; Hough, 1997).[11]

Nicolaisen (1976; also in McNeill and MacQueen (eds.), 1996 – hereafter ASH) has shown how the spread of settlements can be mapped by the distribution of selected name types (Maps 2 and 3: cf. the bōtl and wīc types with the earlier hām, ingahām and ingtūn types). Amongst the earliest are those with ingahām, as in Lyneryngham ‘the settlement of the people by the Linn’ (now East Linton) (Lowe, 1999: 33). This name is one of a number showing the interaction of Anglian and Cumbric: linn is a Cumbric word meaning ‘pool’. A mixed Anglian-Cumbric culture at the leading edge of Anglian settlement is also indicated by some of the archaeological evidence. As Proudfoot and Aliaga-Kelly point out, “Anglian Northumbrian settlement may have developed through relationships with the British landowners or farmers, independent of fluctuations in power in Northumbria and its neighbours” (1995: 23). Apart from some areas such as East Lothian that were already relatively treeless in the late Roman period (Patterson, 1999), the land had not reached its carrying capacity, even within the technological limits of the period. At this time there were still internal frontiers of cultivation, and land could be obtained without necessarily displacing the resident population. Nevertheless, as we shall see (§4.2.2.1), there seems to be little influence of p-Celtic on Older Scots.

Under Oswy (or Oswiu) (642-71), the Cumbric kingdom of Rheged (which may or may not have extended into Galloway: see Map 1), was acquired by marriage around 645. Brooke’s (1991) analysis of the Anglian names and archaeological evidence in Galloway shows that a succession of different powers – British, Roman, Anglian – were in control of the coastal defences and strategic inland passes. It is uncertain when the Angles took control of Galloway, but they may have held as much as half of the accessible land, and were present as free peasants as well as overlords. They were also well established in Cumberland, on the other side of the Solway Firth (Higham, 1985).

The relatively isolated names containing bo[ō]tl ‘a dwelling’ (Maybole) and wi[ī]c ‘farm’ (Prestwick and others) in the west (see Maps 2 and 3) “seem to point to some kind of Anglian overlordship or sporadic influence in the area at a fairly early date” (Nicolaisen, 1976: 79-80).

The northward expansion of Northumbria was halted by the Picts at the battle of Nechtansmere in 685 (possibly commemorated on the Pictish symbol stone at Aberlemno). The battle site is usually identified as Dunnichen Moss near Forfar. In an expansion to the west, Kyle was annexed from Strathclyde around 750.

What is now Scotland south of the Forth-Clyde isthmus remained part of Northumbria for about three hundred years until, with Northumbria weakened by the attacks of the Vikings (see below), it was ceded to the Scots. Exactly when the Scots acquired Lothian (in the broad sense of Scottish Northumbria from the Forth to the Tweed) is unclear – dates ranging from 973 (Lothian ceded by Edgar) to 1018 (Scottish victory at Carham) are given in different sources. However, Barrow considers that the supposed cession of Lothian simply accepted a fait accompli and that this territory had fallen under Scottish control already. Long before 973 “the Scots were exerting pressure upon, and indeed almost certainly appropriating for settlement, the territory north of Lammermuir” (1962: 12),[12] hence the large number of Gaelic place-names in Lothian. Nicolaisen contrasts baile ‘hamlet’ and achadh ‘field’ names – the absence of the latter in the south-east of Scotland suggests that the small numbers of Gaelic speakers in the east were “landowners rather than tillers of the soil” (1976: 128) (see Maps 4 and 5).[13]

The disruption of Bernicia by the incursions of Scots from the north and Danes from the south is illustrated by the fact that few of the large enclosed Anglian sites, such as Hoddom (near the site of the Ruthwell Cross) or St Abb’s Head (Berwickshire) developed into the commercial towns of the medieval period: only Whithorn and Dunbar in southern Scotland “made the vital step from villa regia or monastery to medieval urban centre” (Lowe, 1999: 55).

In the west in the 10th century, there was a resurgence of Cumbric power, with Strathclyde acquiring territory as far south as the North Riding of Yorkshire.[14] In 945 Edmund of Wessex overran Strathclyde and conferred it on Malcolm I,[15] but the Britons regained their independence, until their royal line died out c1018, and the Cumbric kingdom came under Scottish rule.[16]

The squeezing of Northumbria between the Danes and the Scots eventually created the Border as we now know it. Barrow writes:

In a fashion which seems awkward and unhistorical, the border cut both Bernicia and Cumbria in half … … The Solway-Tweed line brought under Scottish rule two tracts of non-Scottish territory, British on the west, English on the east … … their acquisition compelled Scottish kings and their subjects to find some fresh formula in which to express their relationship. It was found in the feudal concept of the regnum Scottorum or regnum Scotie, the kingdom of the Scots or of Scotland. Unlike Scotia, Scotland properly so called, which stopped short at the Forth, the kingdom of Scotland reached south to Tweed and Solway, and incorporated until well into the twelfth century land still regarded, racially or geographically, as Anglia, England. (1962: 20)

Nevertheless, Anglian, in the form of its descendant Older Scots, eventually became the dominant language of the most fertile and densely inhabited parts of Scotland, superseding Gaelic (which had itself displaced p-Celtic languages). A number of factors, all of them connected with the feudal system, aided this spread of Anglian beyond the areas that it had already reached by the start of the feudal period (Lothian and the South-West, together with coastal Fife and Angus). Before we examine the expansion of Anglian, however, we must first turn to the Scandinavian incursions.

2.2 The Scandinavians

A very thorough treatment of the Scandinavians in Scotland is provided by Crawford (1987). There are several distinct areas of Scandinavian settlement in Scotland (see Map 6). The Scandinavians did not become the dominant power or population group anywhere in Lowland Scotland south of Caithness, but they were present in sufficient numbers to leave a large legacy of place-names.[17] ON may also have influenced the Anglian speech of Scotland, but if so this would be largely masked by the later influence of Anglo-Danish (see below).

2.2.1 The Northern and Western Isles

In the last decade of the 8th century, the Western Isles and Ireland began to be raided by Vikings from what is now Norway. Part of the Viking strategy was to establish pirate lairs around the coasts and archipelagos, including Orkney, followed later by defended settlements based on trading and extortion. The Scandinavian presence in the Western Isles does not concern us here, as the linguistic contact there was with Gaelic. In the Northern Isles, there were farming and fishing settlements by the mid-9th century, backed up by the power of the kings of Norway. From this base, Caithness was also wrested from the Picts, but attempts to take Moray were unsuccessful. By the 11th century, the earls of Orkney (whose jurisdiction included Shetland) were in a dual political relationship, owing allegiance to the king of Scotland for their lands in Caithness and to the king of Norway for the Northern Isles.

The place-names of the Northern Isles are almost entirely ON in character, with little trace of the earlier p-Celtic language. The language spoken throughout the Older Scots period was a dialect of Norwegian, known as Norn (q.v., and McArthur ed., 1992, 1996 s.v. Norn). From 1379, the Earldom of Orkney was in Scottish hands, and the extant documentary record of Scots in Orkney begins in 1433. In 1468/9, Orkney and Shetland were pledged to Scotland for a dowry that was never paid, and thereafter the Northern Isles were dominated by Lowland rulers, administrators and clergy, with Scots as the sociolinguistically ‘high’ language of the islands from the 16th century or earlier.[18] In Caithness, Norn was apparently not entirely replaced by Gaelic by the time Scots became important there, probably the 15th century (Waugh, 1986).

2.2.2 Lothian

Little can be said about the linguistic implications of the scattering of place-names in the East of Scotland, except that some pioneering individuals may have formed part of the linguistic and cultural mix in the pre-Norman period. Taylor (1995, 2004) suggests that these names may be linked to the pro-Scandinavian policy pursued by the 10th century kings of Alba (apparently as a buffer against Wessex):

Within this general context it would come as no surprise to find the Scottish kings of the tenth century encouraging limited Scandinavian settlement within their kingdom, especially within those areas in which the Scots themselves were only beginning to establish real and lasting control. The details of the expansion of Alba into both Lothian and Strathclyde is a process which is still not fully understood, but the tenth century would appear to be the period when there was a major shift in Lothian from Northumbrian and towards Scottish control. For at least part of the century the border between Scottish and Northumbrian spheres of influence was formed by the Lammermuir Hills, and it may well be significant that the remarkable cluster of -names in Humbie parish E[ast] Lo[thian] … sits immediately below these hills’ north-eastern edge, some sixteen km from the coast. The question is justified as to whether this cluster is perhaps evidence of Scandinavian settlement countenanced or even positively encouraged by the kings of Alba on the very south-east frontier of their expanding kingdom (2004).

Map 6: The linguistic regions of Scandinavian Scotland (reproduced from Barbara Crawford Scandinavian Scotland, Leicester University Press, 1987: Figure 25, with permission from the Publishers). Area 4 should possibly be extended westwards to take in the points south of Clyde (see §2.2.2).

Fellows-Jensen (1989-90) is inclined to treat the -names to the south of Clyde as part of the same pattern as Crawford’s area 4 (see Map 6 here), [19] and Taylor agrees. The history of Strathclyde in this period is somewhat unclear, but “the overall picture is similar to that in Lothian: of increasing Scottish control … conditions would have been ideal for sporadic settlement of Scandinavians actively encouraged by the encroaching Scottish hegemony” (Taylor, 2004).

Most other Scandinavian names were probably given by Anglo-Danes moving northwards (see below). There is a brief period c1100 when men with Norse names like Thor, Cnut and Swein figure as witnesses to feudal charters in South-East Scotland (Murison, 1974). Such names appear also in occasional place-names in the South-East, but mostly with Old English or Gaelic generic elements, e.g. Dolphinston (containing the Scandinavian personal name Dólgfinnr), suggesting people of Scandinavian descent, though we cannot tell whether they were ON speakers (Nicolaisen, 1976: 114). The multicultural flavour of the time is summed up for us by one Liulf son of Elgi, a freeholder of Coldinghamshire in the late 12th century, who named his five sons Cospatric (Brittonic), Gamal (Scandinavian), Macbeth (Gaelic), Reginald (Anglo-Norman) and Eggard (Old English) (Barrow, 1980: 34).

2.2.3 The Danelaw

From c790 onwards, what is now England was also subject to Viking attack, mainly by Danes. They over-wintered for the first time, in Kent, in 850/1, and the escalating attacks culminated in the arrival of a micel here (great army) in 865. The Danes eventually conquered and settled the Anglo-Saxon kingdoms in the central part of the country, leaving only Wessex and the northern part of Northumbria (the former Bernicia) under Anglo-Saxon control. The boundary of the Danelaw with Wessex was settled in the 880’s by a treaty between King Alfred and King Guthrum. In 902, the Norse were expelled from Dublin, and a period of Norse settlement, mainly in the west of England, followed.

The nature of the Danish population movement has been much debated and seems insoluble. Keynes helpfully summarises the competing models:

We might suppose, quite simply, that the conquest and initial settlement of parts of eastern and northern England was conducted by members of large Viking armies, followed immediately (in the late ninth century) by a veritable peasant migration from Denmark … on a scale sufficient to swamp the indigenous population and to produce a distinctively ‘Danish’ society. Alternatively, we might suppose that the settlements were conducted by the remnants of relatively small Viking armies, whose political dominance enabled them to exert an influence (not least on language, and so on place-names) out of all proportion to their actual number; and that it was the descendants of these old soldiers … who expanded from the areas of initial settlement … Again, we might suppose that the initial settlements were on a small scale, made by members of the Viking armies who not unnaturally established themselves in the most advantageous positions; and that … further settlements took place on a much larger scale behind this protective screen, amounting to a secondary migration from Scandinavia … Or we might suppose that the settlements in the late ninth century were conducted by the remnants of relatively large Viking armies, … that from the outset these settlers mixed and intermarried with the indigenous English population .. creating an ‘Anglo-Danish’ society … (1997: 68).

What is clear is that the Danes had indeed an enormous influence on place-names[20] and language (see §4.2.2.3). The mixed dialect of English that resulted is sometimes known as Anglo-Danish or Anglo-Scandinavian.

A ‘Great Scandinavian Belt’, where ON-derived words and forms are particularly numerous, can be observed in the dialects of Modern English, with its focal area having its southern edge at the Humber in the east, and Samuels (1985) presents evidence for the existence of a similar linguistic and place-name division in the ME period. A number of Scandinavian sound-changes of the period, including the one that arguably produces scho/she from OE hēo (see §7.3.2), occur only within the focal area of the Belt. Scandinavian place-names, however, are just as numerous south of the Humber in Lincolnshire (see Map 7)[21], so this focal area appears to reflect relatively late survival of ON rather than density of the initial settlement.[22]

2.2.4 Cumbria, including Galloway

We use the term Cumbria in a wide sense here, from Strathclyde to the furthest extent of Cumbrian territory in the North Riding of Yorkshire. There are few indications of how long the Cumbric language survived. By the beginning of the feudal period, Galloway was a mashlum of different language groups, so the naming of ‘Galwegians’ in some charters (up to 1179 x 90) is difficult to interpret in linguistic terms. Two references to Walenses (Welsh, i.e. British or Cumbrians) in charters of Malcolm IV (1153-65) are more clearly references to a language group (Barrow, ed., 1960: nos. 240, 258), and there is also a mention of Cumbrenses in a charter of David I (Barrow ed., 1999: No.15).[23]

There appears to have been a peaceful settlement of Scandinavians in Cumberland and Westmorland between about 920 and 945 (when Edmund of Wessex handed the Cumbric kingdom over to the Scots). Higham points out that the distribution of place-names extends far outside the areas of artefacts, so “unless we are dealing with a grossly distorted pattern of artefact recovery” (1985: 45), this suggests Scandinavian settlers without a Scandinavian aristocracy. Around the head of the Solway (including the concentration of -names in eastern Dumfriesshire, see Map 7[24]), the place-name evidence suggests that Scandinavian speakers were on poorer sites. This region of Scandinavian influence appears to extend northwards into Lanarkshire, on the evidence of beck ‘stream’ names (see ASH: 67).[25] Galwegians were often mentioned in charters concerning Ayr and Lanark (Sharp, 1927: 108).

A characteristic feature of the place-names of South-West Scotland and Cumberland is the ‘inversion compound’, in which the generic element comes first, as in Gaelic, e.g. Crossraguel. The majority of these names in South-West Scotland contain the element kirk (e.g. Kirkcudbright), which is particularly problematic to interpret: although ON in origin,[26] it supplanted OE-derived chirch, and became a productive element in PreSc (Nicolaisen, 1976: 108ff., and see below): some of the names first appear as late as the 15th century. The earlier of the names may be the work of a mixed Gaelic/Norse speaking population from Ireland or the Hebrides. However, place-name scholars agree that the patterning of the clearly ON place-names (see Map 7 and ASH: 67) links the Scandinavian settlers around the head of the Solway with the Danes across the Pennines in the North Riding of Yorkshire,[27] rather than the Norse, or Gaelic/Norse, who were however present elsewhere in Cumberland and Westmorland, and perhaps in Galloway.[28]

Samuels did not explore whether the Great Scandinavian Belt stopped at what is now the Border,[29] but this question is taken up by Kries (1999), who argues the case for seeing the dialect of the South-West not only as a continuation of the Belt, but as part of its most densely Scandinavian band or focal area. This part of Scotland would accordingly fall within the area where Anglo-Danish was created, although it was undoubtedly the later influx from Yorkshire that was crucial in dispersing this influence throughout urban Scotland.

2.3.1 The Anglo-Normans and their followers

There was intermittent Norman influence on the English court for half a century before the Norman Conquest of England in 1066. The exiled English princess Margaret (Saint Margaret), the wife of Malcolm III (Canmore), introduced continental influences into the Scottish church. Like the Angles, the Normans had already been seen in Scotland, as mercenaries. The first were probably those who fought on both sides with the future Malcolm III and Macbeth (Barrow, 1973: 279; 1981: 26).

When Malcolm III was killed (in 1093), his brother Donald Bán expelled the Normans and the southern English brought in under Malcolm and Margaret, but he was driven out by Malcolm’s eldest son Duncan, trained as a knight by the Normans, who in turn held the throne for less than a year and was forced to dismiss his Anglo-Norman retainers. Edgar took the throne in 1097 with the help of William II (Rufus) of England.[30] The next king, Alexander I, built castles and created knight feus on royal estates. But it was with the reign of David I, beginning in 1124, that Normans were brought in in large numbers. David I, who had spent much of his early life at the English court, greatly accelerated the process of feudalisation, so that by the end of his reign (1153) most of Scotland south of the Forth, apart from Galloway and Carrick, had been allocated to tenants, almost all newcomers, holding by military service. North of the Forth there was settlement in Fife, Gowrie, Angus and the Mearns, and the Aberdeenshire districts of the Garioch and Formartine. The Earl of Moray rebelled in 1130 and after defeating him, David I annexed the whole province (at that time a very large territory) for the crown and set up foreign feudatories there.

Later described as “ane sair sanct to the croun” (see sanct B 2 (3)), David I continued the process begun by Saint Margaret of endowing Continental monastic orders in Scotland. As Sharp (1927: 123) points out, David I perhaps valued the Normans as much for their administrative as for their military talents, and it was from the ranks of churchmen, educated in schools run by churchmen, that administrators were drawn. As with secular migrants, the clerical migrants came from establishments in England as well as directly from the Continent. Crucially for the spread of Lowland Scots, David I also founded burghs (see below) in almost every part of his kingdom outside the Highlands (Barrow, 1981: 31-2).

The Clyde valley was systematically feudalised by Malcolm IV in the mid-12th century. A colony of Flemings was planted, hence the place-names Thankerton (after Tancard), Wiston (Wice), Lamington (Lambin), and Symington (Simon Loccard) (Barrow, 1973: 289). The penetration of Galloway was apparently difficult – there was “a violent anti-foreign reaction” that lasted from 1174 to 1185 (Barrow, 1981: 49).

The Anglo-Norman era lasted, Barrow tells us, from 1097 (the beginning of the reign of Edgar) to 1296 when war broke out between England and Scotland. For large parts of the Norman period, Scotland was a client kingdom of Norman England, and the kings of Scotland “kept the doors of their kingdom open for settlement, and in particular for settlement from England and northern France… … Scotland became a land of opportunity for sons whose fathers had not yet died, for younger sons with no patrimony to inherit” (Barrow, 1980: 7; see also ASH: 417). Barrow quotes the well-known comments of the Barnwell chronicler to the effect that “the more recent Scottish kings [Malcolm IV and William the Lion] count themselves Frenchmen by race, manners and speech, and retain only Frenchmen in their household” and of Jordan Fantosme that William the Lion “held only foreigners dear, and would never love his own people” (ibid.: 84).

Nevertheless, the Anglo-Norman period in Scotland differed from England in that Scotland had fewer external connections and retained its native dynasty and much of its native ruling class (ibid.: 153). Existing earldoms came to be held under feudal tenure, and the native magnates had to adapt by obtaining more land with which to reward new followers, who might be incomers, and by forming marriage alliances with the new order (ibid.: 87). There was nothing to prevent them obtaining lands in England, which they sometimes did.

Below the great lords, there was a stratum of tenants holding land by knight-service (the originals of the later laird class), many of whom were also migrants. “Although many of these families may have had continental ancestry, as far as Scotland was concerned they would to all intents and purposes have been English. Their speech was doubtless English, their experience was limited to England, and they would have regarded themselves as English by race” (Barrow, 1980: 82).

2.3.2 Feudalisation and the spread of OE/PreSc

Anglo-Norman French never acquired the importance in Scotland that it had in England (Murison, 1974: 77-8; Barrow, 1999). With rare exceptions, it was Latin that was employed in administration (which took on a new, bureaucratic form under the feudal system). French was a familial language amongst the Normans, and was of use for wider communication (with England and France), but OE (which we can now begin to call PreSc) was the shared language of feudal overlords (secular and clerical), their vassals, and the freemen of the burghs. The anglicising forces of the feudal system were:

  • the burghs;
  • the monasteries (see ASH: 340-1), and the parochial organisation of the church (see ibid.: 348ff.);
  • the local administration, by sheriffs, of feudalised territories (see ibid: 192-5).[31]

The numbers of people involved were small, but the social shift was radical, and eventually brought about a linguistic shift from Gaelic to Scots throughout the Lowlands, as the native population was assimilated into the new social system.[32]

The burghs, as foci of internal and external trade, were crucial in the spread of Lowland Scots, although the population even of the largest would have numbered hundreds rather than thousands (Barrow, 1981: 94). There was a great deal of internal migration as new burghs were created, and this would have increased the homogeneity of the dialect that spread as a result. The population of the early burghs would have included Lothian Angles, and a handful of Gaelic speakers, and in the west and south-west Cumbrians (conceivably still Cumbric speakers). “For the most part, however, they seem to have come into Scotland from Flanders, the Rhineland, northern France, and England especially eastern England” (Barrow, 1981: 92).

Right up to the 16th century, Flemish craftsmen were encouraged to immigrate, and they formed small enclaves, seen in such place-names as Flemington, of which there are four in Scotland, or settled in the burghs, where they played a prominent part in public life. Their linguistic influence is reflected in burghal terminology, e.g. guild, kirkmaister. They were allowed to have their own ‘Fleming law’ (s.v. Fleming n.).

The Atlas of Scottish History to 1707 (ASH) gives a series of maps (pp.196-8 and 212-4) showing the spread of burghs up to 1500. The second of these is reproduced here as Map 9. Taylor (1994) considers that the critical mass of Scots speakers in the St Andrews area was reached c1200: a man named Martin gave his name to Balemartin (coined in the mid-12th century), with the Gaelic generic baile, while his son Gillemuire, who was alive around 1200, gave his name to Gilmerton, with the OE generic tūn.[33] Nevertheless, Gaelic continued to be used in Fife in the first half of the 14th century, on the evidence of names and nicknames that occur in a Dunfermline document containing genealogies of neifs, indicating a Gaelic-speaking environment at least amongst the unfree peasantry. The outward spread of PreSc from the burghs into the countryside is examined for the North-East by Nicolaisen (1999), who finds evidence of Gaelic/English (i.e. PreSc) bilingualism from the early 13th century, and a dramatic increase in PreSc place-names (mostly additional to the existing Celtic nomenclature, but sometimes replacive) in the 14th century. Sharp (1927: 382ff) similarly found that whereas even the east coast of Angus was strongly Gaelic in 1219, by the end of the 14th century, the districts around Forfar were inhabited mainly by Scots (“English”) speakers. Scots must have begun quite early to differentiate into Northern, Central and Southern dialects (see §§5.2.5, 8.4 and Map 10).

Withers’ map of “Linguistic Changes” (ASH: 427, reproduced here as Map 11) attempts the difficult task of estimating the boundary between Gaelic and Scots (“English”) c1400 and c1500, on the basis of place-name and charter evidence. Of course, this boundary would not have been a thin line, but a transition zone with a mixed and often bilingual population (cf. the maps for later periods, where more detailed information is available, ibid.: 428-9).

2.3.3 Anglo-Danish population movement

In a chapter with important implications for the history of Scots, Barrow (1980) looks at the origins of the adventurers who came with the Norman lords into Scotland. By the time of the Anglo-Norman settlement in Scotland, it was fashionable to have surnames, mostly from the village or manor where the family held most of its property, so the Anglo-Normans in Scotland are mostly traceable. It was formerly thought that the Honour of Huntingdon (English lands held by the kings of Scotland as vassals, mainly in the shires of Huntingdon, Cambridge, Bedford and Northampton) was the single main source, because of its connections with the Scottish crown, but Barrow concludes that Somerset and Yorkshire were equally important. The remote English west country is a surprising source, and this suggests that the kings of Scotland (and their larger lords) were extremely eclectic in their recruitment of vassals (ibid.: 100).

The Yorkshire input is very interesting indeed from a linguistic point of view, as the English spoken by such migrants would have been Anglo-Danish.[34] “Although we cannot calculate the numbers of dependants involved, we may feel sure from the evidence available that the movement was large-scale in relation to the existing population … … I would argue strongly for the probability that Anglo-Norman settlement greatly reinforced the Middle English [i.e. Anglian] elements in Scots speech and culture, and had a decisive effect upon the texture of Scottish society as a whole” (ibid.: 117). Barrow also writes that “It is essential to grasp that whereas, for our purposes, the Picts, Britons, Irish, Scandinavians, and French appeared on the scene only once, the Anglo-Saxons came twice” (ibid.: 5-6). Following Barrow, Aitken accepts, in the Introduction to The Concise Scots Dictionary, that:

This Scandinavianized Northern English – or Anglo-Danish – was certainly the principal, though probably not the only, language of the early Scottish burghs and its contribution to the formation of the language later known as Scots is probably even greater than that of the original Old English of south-eastern and southern Scotland. (p.ix)

Barrow’s discoveries solve a number of puzzles: the relative lack of p-Celtic loanwords; the relative lack of Gaelic loanwords, given the absorption of Lothian into the Kingdom of the Scots; and the abundance of Scandinavian influence.

The lack of Scottish texts between early OE and the late 14th century, apart from occasional words in Latin documents, makes it difficult to trace the transition from OE to OSc in detail. When the written record of Scots really begins again, we find that it is heavily ON-influenced. An important early witness is ‘The Scone Gloss’ (Facsimiles of National Manuscripts of Scotland vol. II: no. 19), an interlinear gloss written before c1360 in a Latin lease. In this text, the present participle ending is already the ON-derived -and, not the native OE -ende; and we even find the ON-derived -and suffix of the ordinal numeral: the four and tuentiand fat ‘the four and twentieth vat’.

These puzzles are explained if the language that spread through the Lowlands in the feudal period owed more to an incoming population of Anglo-Danes than to the Lothian Angles, by then under Scottish domination for perhaps two centuries. To resolve the issue would require answers to such questions as: was there a Celtic influence in Lothian that later disappeared? How much Scandinavian influence was there already on the Anglian of Lothian? Was this distinguishable from Scandinavian influence on the Danelaw? Was there an abrupt or a gradual transition from Old English to borrowed Old Norse forms? Was the dialect of the Anglians in Scotland identifiably different from the dialect that spread to the early burghs?

To some extent, this period of the language’s history will always remain obscure. However, some light will perhaps be shed on these matters by ongoing work on place-names and on OSc dialectology. A detailed study of the language of the earliest records is also much needed, and work has begun on this in the Institute for Historical Dialectology, University of Edinburgh. At the moment, a few pointers exist (and cf. §6.1, Vowel 2).

The Scandinavian element in Scots, Kries (1999) demonstrates, is a mixture of West and East Norse (the ancestors of Norwegian and Danish respectively). Unfortunately, the distinction between dialects of ON is less conclusive than we might wish. At the time of the Scandinavian settlements, West and East Norse had only begun to diverge, and conservative forms still remained in West Norse that would later become peculiarities of East Norse. Scholars in this field tell us that the few criteria that can be relied upon suggest both Norwegian and Danish populations in Cumberland and Westmorland (Fellows-Jensen, 1985a, 1985b) and a generous admixture of Norwegians in the Danelaw (Thorson, 1936), confirmed also by the place-name evidence (Cameron, 1996: 77). Nevertheless, Kries is able to make a meaningful distinction between the two sources, with WScand loans being mainly rural in character (e.g. slak n.1, bow n.2, graip), and EScand both rural (e.g. intak n. 2, wikkir) and, as we would expect of vocabulary transmitted through the Anglo-Danish of the burghs, urban and commercial (e.g. osmond, keling and the interesting word kirseth, whose citations cast light on the relative attractiveness of different burghs).

The accepted view of ON loanwords in Scots follows Aitken (1954):

the Old Norse loans found in Scots (again leaving aside Caithness and the Northern Isles) are almost all found likewise in the dialects of the North of England. This is in contrast to direct borrowing from other languages such as Middle Dutch and Anglo-Norman, where the loans into Scots are independent of the influence of these languages in England. (Macafee, 1997: 201)

It is therefore very significant that Kries finds that there are at least a few loans and senses in OSc not found in England, which does indeed suggest an independent input, e.g. bathstoff, bing in the sense ‘a heap or pile’, hink, buller v.2 (and also kirseth, which is nevertheless associated with the burghs). A few items, as we might expect, are specific (within Scotland) to the South-West: hagarde, nolt price (s.v. nowt n. 2 c); and cf. the following, recorded in the modern dialect (see SND): dyke in the sense ‘hedge’ (s.v. dyke n. 2), ure n.1, choop n.1 (see §7.3.2 on the scho form of the personal pronoun).

A situation of dialect contact in the 12th century is suggested by the forms of the place-name element kirk. Within a century, the earlier native form chirche[35] gives way to the Scandinavianised form kirk. However, there are also transitional forms such as kirche and chirk, and these interdialect forms are very telling, as they point to an abrupt confrontation of the two forms at this point in time, rather than a gradual diffusion of the borrowed form into Lothian Anglian.[36] It would be a worthwhile study to examine the earliest names in the hope of establishing the nature and chronology of this transition. The effect of contact is also seen in the mistaken translation of scír-burna (‘shire burn’, if Brooke is correct) to Skyreburn, as if containing skire- (s.v. Skire-Thursday (cf. schire adj.)) ‘bright’. Unlike schire, shire has no ON cognate, and is therefore retained in the related name Shirmers (< scir-(ge)maere ‘shire boundary’) (Brooke, 1991: Appendix).[37]

It is tempting to speculate what Scots might have been like had it developed from the speech of the Lothian Angles without this Anglo-Danish reinforcement. Quite possibly it would not have survived, or would have lingered into modern times only in some isolated enclave, as an archaic form of English did in Ireland, in Forth and Bargy, after the ‘Old English’ colony was absorbed. Assuming that it did survive, and was not excessively isolated and archaic, it would necessarily have been most like the northern English dialect of the Northumbrian enclave seen in Map 12. This rump of Northumbria, in alliance with Wessex, held out against both the Danes and the Scots until England was unified in the mid-10th century. Its modern dialect often agrees with dialects south of the Scandinavian Belt in having words and forms of native origin, e.g. anvil rather than stithy, ladder rather than stee (Orton and Wright, 1974); hang rather than heng or hing, gosling rather than gesling, lie rather than lig v., ridge rather than rig (of a house) (Orton et al., 1978); churn rather than kirn or kern and birch rather than birk (Kolb, 1964). In many other cases, however, the ON item has diffused into the enclave and is found throughout northern English and Scots.[38] This diffusion would have taken place also in our speculative scenario, but unlike the sweeping effect of population movement, it would have been gradual and increasingly attenuated as it reached northward towards the limits of Anglian in Fife and Angus (if indeed, Anglian had survived north of the Forth).

In its lesser degree of ON influence, the modern descendant of Anglian would have been more like StE, but would have been distinguished from the latter by a much greater degree of Gaelic influence, and perhaps even a substantial p-Celtic element. It would not, of course, have been entirely lacking in ON influence even without diffusion from the Scandinavian Belt, and there might have been a marked dialectal difference between the more heavily Scandinavianised dialect of the South-West and the dialect of the South-East, but in any case these would merely be northward extensions of the corresponding dialects in England.

2.4 The later spread of Scots

The continuing geographical retreat of Gaelic, which for our purposes marks the advance of Lowland Scots monolingualism, has been researched in detail by Withers (1979, 1984). The stigmatising of Gaelic and its relegation to minority language status are discussed by Ó Baoill (1997; also Macafee and Ó Baoill, 1997). By the late 14th century, John of Fordun was making the distinction that was to become stereotypical, between “domesticated” Lowlanders and “wild” Highlanders, and he described the partition of the country between them:

The language and customs of the Scots vary with the diversity of their speech; for they use two languages, namely Scottish and Teutonic; those of the latter tongue possess the coastal and low-lying regions, whilst those of the Scottish tongue inhabit the mountains and outlying islands. (quoted in translation by Nicholson, 1978: 206)

Speitel and Mather (1968: 522ff.) offer a chronology of the spread of Scots into peripheral areas:

  • 15th century: Caithness (modest influx from the 13th century on); Orkney and Shetland (ceded to Scotland in 1468; modest influx previously, particularly in Orkney);
  • 16th century: presumed demise of Gaelic in the South-West;[39]
  • 1603 onwards: Ulster – Co. Antrim, Co. Down, Co. Donegal, later spreading into Co. Londonderry. Montgomery and Gregg give the following concise account:
    Lowlanders speaking Scots began to trickle over the channel in the second half of the sixteenth century (indeed it was (in part)[40] their presence in Ulster which first alarmed and provoked the Tudors to attempt early but largely unsuccessful plantations), but their first significant infusion occurred around the turn of the seventeenth century, in the very earliest years of the reign of James VI/I. Through private grants or other means, they arrived in east Ulster in numbers sizeable enough and were sufficiently successful in developing the land to exclude the counties of Antrim, Down and Monaghan from an official plantation … begun in 1610 …that initiated the recruitment of Scots and English to take up land in the province … … (London)Derry was included in the official plantation plans. Its settlement was the prerogative of the London companies, which had little luck in the enterprise. The Lowland Scots, because of their closer bases, were able to take over a good portion of the north-east corner of the county and penetrate loosely the rest of (London)Derry and Tyrone. It was also as part of the official plantation plans that Scots were brought over to Donegal from 1610 onwards … They were settled in the northern parts of the low-lying east Donegal region known as the Laggan. … … The plantation was only one phase of a wider process of Scottish migration that can be sketched only in outline, because much of the later to-ing (and fro-ing) between Lowland Scotland and Ulster was anonymous and untraceable. (1997: 572)
  • 1650 onwards: Kintyre, Arran, Bute (the last possibly as early as the 15th century);
  • 17th century onwards: inner Moray Firth, fishing villages on the east coast between the Moray Firth and Caithness;
  • 18th and 19th centuries: forestry and whisky-based settlements along the Highland Line.

The influence of Gaelic on Scots is that of a substratum (cf. §4.2.2.1), and is therefore likely to be expressed in subtle ways, mainly through phonological, syntactic and semantic features carried over inadvertently by Gaelic speakers shifting to Scots, rather than in copious lexical loans. Some influences are pervasive in Scots (see §4.2.2.2), and Macafee and Ó Baoill (1997) also found some evidence that the dialects of Scots still in contact with Gaelic in the MSc period were more heavily influenced than Central Scots in these ways: for instance, the NE /f/ in words like fa ‘who’ and fulp ‘whelp’ was probably substituted by Gaelic speakers for the consonant /ʍ/, not found in Gaelic. It is conceivable that the NE change of /w/ to /v/ in /wr/, e.g. vrang ‘wrong’, and post-vocalically, e.g. gnaave ‘gnaw’, was induced by the lack of /w/ in Gaelic (Macafee, 1989). In the SW long-term contact with Ireland makes it difficult to separate Gaelic and Irish influences. In areas where Gaelic was not lost until the ModSc period, the replacement is usually a form of English (Highland English, q.v. in McArthur ed. 1992, 1996) with varying degrees of Gaelic influence, and a considerable body of loanwords from Scots. There must formerly have been what might be called Highland Scots, a second language variety represented for instance by an early 18th century broadsheet (purportedly a letter home from Maryland) discussed by Millar (1996), and in highly stereotyped form in literature (see §8.5).

2.5 Anglicisation

2.5.1 Anglo-Scots

It may be useful, before discussing anglicisation, to mention the existence, before 1500, of a mixed language, known as Anglo-Scots, found, for instance, in Colkelbie Sow, Lancelot of the Laik and James I’s Kingis Quair. For the last of these, we can explain the odd language in terms of the personal history of the author, since James I spent the latter part of his youth in captivity in England. His language is not the kind of superficial mixture that arises through the scribal copying of texts. A ME feature integral to the language of the Quair is the use of final -e rather than final -is to give metrical flexibility (see §9.1.2.4).[41] The thoroughness of the mixture is also shown by the fact that there are rhymes in the Quair that are neither Scots nor ME, e.g. moon ‘moan’ : doon ‘do’ (ll.309-11) – Scots mane does not rhyme with do; nor does ME moan rhyme with the inflected infinitive done.

2.5.2 The changing relationship between Scots and English

Scots and English have never been isolated from each other, and have always formed a geographical continuum of dialects within which linguistic changes diffused and spread. At different periods their relationship was dominated by different processes:

1. In PreSc and ESc, we find a pattern that is normal, Joseph (1987) tells us, before the introduction of standardisation to a vernacular: changes flow freely within a group of language varieties. The main focus is nME, in a state of flux precipitated by language contact with ON, is the main focus;

2. the spread of StE (from about 1450 on) is at first irrelevant to Scots, which continues to diverge from its southern neighbour and to produce its own incipient standard in the early MSc period, while continuing to share in ongoing changes such as the Great Vowel Shift, and to borrow isolated features, especially in poetic diction (see §9.3.1);

3. in the transition to EModE (corresponding to late MSc), StE experiences a period of rapid elaboration, much of which is transmitted to Scots, including developments in the verb phrase and wh-/quh- relative pronouns. As Görlach (2002) describes it, Scots seems to lose any initiative in this period. Time and again, an innovation is shown to have appeared first in English. [42] This phenomenon might be termed ‘pre-emptive innovation’, on the analogy of ‘pre-emptive domestication’. This is a term used in anthropology to capture the observation that when a species has once been domesticated, there is no need to domesticate it again unless there is some geographical barrier to its spread. For instance, genetic analysis shows that in Eurasia the economically important plant and animal species have generally only been domesticated once, and have diffused across the width of the continent, whereas in the Americas, with their north-south orientation cutting across different climate zones, species such as cotton have been domesticated independently north and south of the Equator (Diamond, 1998: 178ff.).

England and Lowland Scotland occupy an easily traversed space on the same small island. The greater size of the population of the south-east of England, and the wealth and stability of the English economy, are sufficient to explain the creative vitality of English at this time (though much of its elaboration had already, of course, been pre-empted by French, itself following Latin),[43] and the overwhelmingly northward flow of innovation even before Scotland embraced StE.[44]

The predominant contemporary perception that Scots and English were the same language (see Inglis B 1, Scottis A 1 (e) and B 1 (1), Southern B 2)[45] allowed English elements to be “infiltrated into Scots writings and, later, speech, without appearing too incongruous” (Aitken, 1979: 89). Aitken (1997) traces the gradual adoption of anglicised forms in Scots prose from the early 16th century, and examines the writings of a small group of individuals whom he calls ‘Anglo-Scots’, whose personal life histories made them familiar with spoken English, and whose writing reflected this in varying degrees even before the Reformation of 1560.

Many clergy were brought into contact with English through the enforced exiles that affected first one party then another in the religious struggles of the 16th and 17th centuries, a notable instance, of course, being John Knox. By the turn of the 17th century, some individuals were able to modify their written language (Bald, 1927; Aitken, 1997), and perhaps also therefore their speech, according to the addressee. If the ‘1 GENT’ of Eastward Ho! is indeed James VI and I, as has sometimes been suggested, he is represented as bilingual:

Farewell, farewell, we will not know you for shaming of you. I ken the man weel; he’s one of my thirty-pound knights. (George Chapman, Ben Jonson and John Marston, ed. R. W. van Fossen, Manchester University Press, 1979:IV: 1)

Scottish familiarity with spoken English is shown by 17th century spellings such as <no> ‘know’ (Scots knaw, ken); <tu, tow, towe> ‘two’ (Scots twa); <how> ‘who’ (Scots quha) (Aitken 1979). Such misspellings could only be arrived at by ear. English troops were garrisoned in Scotland during the 1650s, and this may also have been influential.

MacQueen (1957: 197) and Meurman-Solin (1993a, 1997; see §8.3) both conclude that anglicisation was a pragmatic process, with no implication that southern forms were felt to be more ‘correct’:

anglicization … appears to be primarily motivated by the practical needs dictated by contact situations between the two varieties. Individual writers seem to have been tempted to adopt practices of the wider linguistic community, whereas institutions … tend to be resistant to abrupt overall change. (Meurman-Solin, 1993a: 49)

Indeed, missing Scots forms were sometimes introduced into the printed versions of parliamentary minutes in the last years of the Scottish Parliament (MacQueen, 1957: Appendix 8), and a Londoner who visited Scotland in 1689 recorded that:

They are great Criticks in Pronunciation, and often upbraid us for not giving every word its due sound … neglecting the gh as if not written (Rev. Thomas Morer, quoted by MacQueen, 1957: 263-4).

The southern forms were nevertheless felt to be more modern: cf. the 1678 quotation s.v. the def. art. 11, which recalls Winȝet’s well-known jibe about “curiositie of nouationis” (s.v. Scottis B 1 (1)).

The effect was what Joseph calls “involuntary language shift”, a gradual erosion of one language variety through mixing with another. The drift towards fashionable anglicisation, even against the writer’s own inclination and better judgment, is illustrated for instance by the A text of David Hume of Godscroft’s The History of the House of Douglas, of which the modern editor writes: “however conservative, and perhaps deliberately so, the language of A nevertheless shows how Scots had been penetrated by English forms and made uncertain in its usage” (Reid, 1996: I, xlix). Hume himself wrote of his linguistic preferences:

For the language, it is my Mother-tongue, that is, Scottish: and why not, to Scottish-men? Why should I contemne it? I never thought the difference so great, as that by seeking to speak English, I would hazard the imputation of affectation. … For my own part, I like our own, & he that writes well in it, writes well enough to me. Yet I have yeelded somewhat to the tyrannie of custome, and the times, not seeking curiously for words, but taking them as they came to hand. I acknowledge also my fault (if it be a fault) that I ever accounted it a mean study, and of no great commendation to learn to write, or to speak English, and have loved better to bestow my pains and time on forreigne Languages, esteeming it but a Dialect of our own, and that (perhaps) more corrupt. (From the preface to the 1644 edition, ed. Reid, 1996: 452-3)

Ironically, whoever prepared this edition for the press anglicised the text very thoroughly.

The period of mixing set up a continuum between Scots and English.[46] As it affects speech, the process continues to spread and is still ongoing;

4. the next phase is anglicisation proper, with Standard English adopted as a process of voluntary language shift, which involved the replacement of entire genres by Standard English (Görlach, 1997). In Murison’s pithy formulation:

Scots … lost spiritual status at the Reformation, social status at the Union of the Crowns, and political status with the Parliamentary Union. (1979: 9)

The lack of a Bible in the native language, which is considered to have been crucial in the maintenance of literacy and language use in other cases, such as Welsh, was one very important factor in the adoption of StE. The Geneva Bible of 1561 and an English Service Book were used in Scotland. In contrast to Latin, this was “spirituall foode to our soullis” in “our commoun toung” (1558, quoted by Robinson, 1983: 60).

The controversy surrounding the Protestant Reformation also gave an impetus to printing. Material produced in England was also widely read in Scotland. Although printing in Scotland was, of course, in Scots at first,[47] after the Reformation both Scotsmen and incomers printed works in both Scots and English (Bald, 1926; Watry, 1992). The Scotsmen had varying degrees of success in their attempts at English, but the effect of print was generally to homogenise the language of the text, whether in the direction of Scots or of English. After 1600, printed texts were considerably more anglicised than manuscripts overall (Meurman-Solin, 1997a: 15).

In retrospect, the choice between mither-tongue Scots and pitten-on English often seems to us to have been the decisive one for the future of the vernacular, and a few late MSc voices were evidently aware of this issue (see references above). At the time, however, the main debate and the most conscious choice for most writers of verse and literary prose was between Latin and the vernacular, and having chosen the vernacular, between a latinate and a ‘plain’ style.

Latin was often described as copious. As functions were transferred to the vernacular, writers strove to achieve the same facund eloquence, what Douglas called “fowth of langage” (see fouth (2)). In late 16th century and 17th century England, the means of elaborating the vernacular were explicitly discussed amongst three main factions: those in favour of extensive borrowing – their opponents objected to ‘inkhorn terms’, and in Scotland to minȝard terms; purists in favour of augmenting English from its own resources; and additionally in verse, archaisers following Spenser (see Barber, 1976: ch. 11). There were both adherents and opponents of the borrowing camp in Scotland, and Douglas in practice drew on archaisms (see e.g. gan p.t., to- prefix2 b, nocht-for-thy, himselvin, hir selvyn, and references passim in Bawcutt, 1976). A number of the makars are eloquent in their praises of the latinate style of Chaucer and Lydgate (see aureat adj. 2, laureat adj. 2, ornat 2 adj., rethorik n. 2). The author of The Complaynte of Scotlande, however, is scathing of “exquisite termis quhilkis ar nocht daly vsit . . ., dreuyn or rather to say mair formaly reuyn fra lating” (16/15, quoted by Aitken, 1971: 178), although finding it necessary sometimes to employ latinate vocabulary (see Latine B 2 (c)). The watchword of the anti-latinate camp is plain (q.v. adj.1 7), with the terms hamely (q.v. adj. 4) and rude (s.v. rude adj.1 10) used in self-deprecation.

Douglas is notable for his aspiration to elaborate Scots – and Scots specifically. His approach is one of judicious eclecticism (see Latine B 2 (a)). But for the most part, MSc writers in the borrowing or latinate camp seem to have seen the elaboration of the vernacular as a joint enterprise with English writers. Perhaps more than we now realise they felt an ownership of the StE that resulted, and which they simply read aloud as if it were Scots (Robinson, 1983), and occasionally described as Scottis (Bald, 1928), and frequently as Scottis or Inglis (s.v. Scottis B 1 (3)).

By the middle of the 17th century, as Corbett points out, the use of NE Scots to represent provincial speech in Urquhart’s translation of Rabelais is:

indicative of the way that educated Scots – even Scots who, like Urquhart, were educated at Aberdeen University – now perceived the way they spoke. In writing, it was reserved for contexts associated with provincialism and low comedy. (1999: 93-4)

– prefiguring one of its main literary roles thereafter.

In the 17th century, the Scottish nobility began, as a class, to acquire spoken English, and to intermarry with the English aristocracy. Following the Restoration of 1660:

every Scotsman of the nobility was likely to spend part of his time in southern England, at court or residing in the Home Counties, and nearly all other eminent Scots … visited London for shorter or longer periods. (Aitken, 1979: 91)

In the course of the 17th century:

The choice of word-forms and vocabulary in their private correspondence seems to suggest that their speech passed through a stage when there was rather inconsistent vacillation between native and imported southern options … to a variety which fairly consistently preferred southern English forms and words. (ibid.: 93)

(but see §8.3.3). As early as 1673, Mackenzie of Rosehaugh was writing, apropos of the question why the English undervalued the Scottish idiom:

that of our Gentry differs little from theirs; nor do our commons speak so rudely as those of Yorkshire: as to the words wherein the difference lies, ours are for the most part old French words … (quoted by MacQueen, 1957: 259)

though, writing in 1681, he also recommended the use of Scots (“firy, abrupt, sprightly and bold”) for pleading at the bar (quoted by Ouston, 1987: 20). Pitcairne’s play The Assembly, written in 1692 (though not published until 1722), represents the speech of the gentry as StE, apart from some of the older generation. StE speech was thus becoming widespread amongst educated Scots at the same time as it was becoming general at this social level in England.

By about 1760, it was distinctly quaint for a gentleman or lady to speak Scots in polite company, though some judges remained kenspeckle for their use of Scots in court, including Lord Auchinleck, the father of James Boswell (Bailey, 1987). However, in the absence of native StE-speaking role models, English speech was apparently achieved only after a period of false starts, in the form of spelling pronunciations, interdialectal forms and hypercorrections (see Macafee, 2004). By this time also, the ‘Augustan’ ideology was demanding that writing be stripped of every trace of impropriety, including any indication of a writer’s Scottish origins – a hurdle that few non-English or American writers could clear even now, as MacQueen (1957: 233) points out;

5. the final stage is the involuntary language shift (as above) of the remaining speakers, mostly at the lower end of the social scale. At the time of writing, this stage has been reached over wide areas of Central Scotland (see Máté, 1996; Macafee, 2000). At this point we begin to speak of language death.

There is a tendency to see the history of English as a single-minded march towards StE. Even in Scots scholarship we find this – the trajectory of the written language from the late 16th century on is so unremittingly towards StE that it seems sometimes to be overlooked that this is a movement of the written language away from any entity that we could call Scots. To some extent, scholars are led into this peat-hag by a refusal to treat Scots and English as distinct entities (and it is not necessary to regard them as separate languages to apply a contact linguistics approach). But some egregious mistakes could easily be avoided by consulting the Scots dictionaries and the data of The Linguistic Atlas of Scotland (LAS). The history of the development of Scots from the late 16th century to the 18th is largely still to be written.

[6] Until recently, historians had accepted the Scots’ own account of their origins, in the absence of evidence to the contrary. This stated that the Scots came from Ireland (at a time variously given as the third century AD or c500) and founded the kingdom of Dalriada in Argyll. However, this has now been challenged by archaeologists, who point out that there is only a partial overlap between the material culture of Ireland and the west of Scotland at this time, and, crucially, that there is no discontinuity in the archaeological record of Dalriada. Campbell (1999: 14-15) suggests that the origin myth was designed to explain the shared q-Celtic language of Dalriada and Ireland, which may in fact simply have been a shared conservatism, as the q-Celtic forms are older. The implications for Galloway, with its early slew (< sliabh) names, seen as evidence for a settlement contemporary with Dalriada (Nicolaisen, 1976: 39ff.; ASH: 58-9 ) is not yet clear.

[7] Early interaction with the Angles is indicated also by the poem Gododdin, a Cumbric composition eulogising the warriors who fought the battle of Catraeth (sometime between 540 and 600, probably at the Roman fortified town of Catterick Bridge in what is now Yorkshire). It is now thought that one of the leaders, Yrfai son of Golistan, was the son of an Angle, as Golistan has been identified with Anglo-Saxon Wulfstan (Lowe, 1999: 16).

[8] Bede distinguished amongst Angles, Saxons and Jutes and associated them with distinct territories in Britain, but Blair (1956: 10, 11) doubts whether the tribal distinctions were maintained in the crossing to Britain, and the dialect differentiae within OE emerged much later. The term Saxon was the usual one in the OSc period for the historical people. It was borrowed into Gaelic and thus back into 18th century Scots as Sassenach.

[9] The annals record only that there was an obsesio Etin in 638. Proudfoot and Aliaga-Kelly (1995) explore Anglian place-names in relation to pre-Norman boundaries in South-East Scotland, which may reflect phases of Anglian expansion.

[10] Nicolaisen (1976: 71). It has now been suggested that Harrow Law in Peebleshire may be OE hearg ‘a pagan shrine’ (Proudfoot and Aliaga-Kelly, 1995: 25, citing Wilson, 1985). What may be traces of pagan Anglo-Saxon cremations have also been found in Roxburghshire and East Lothian (Proudfoot and Aliaga-Kelly, 1996: 9).

[11] There is disagreement on whether the few wīc names north of Forth are early (Proudfoot and Aliaga-Kelly, 1996; Taylor, 2000).

[12] By the 940s the clergy of St Cuthbert seem to have lost their lands north of the Tweed (Barrow: 1962).

[13] Taylor has re-examined the place-names of Fife, and concludes that there are no achadh names there. He identifies the generic of names like Auchtermuchty as uachdar ‘upland’ (2000: 209).

[14] Thus defining the later Scottish Border with England up to 1157 (Barrow: 1962; see also Barrow’s maps in ASH: 76-7, 79). Higham (1985) suggests that the Angles in Cumberland and Westmorland may have connived at the expansion of Strathclyde as a way of strengthening themselves against the Dublin Norse on one side and the Yorkshire Danes on the other.

[15] Barrow (1962: 10) writes: “The Scoto-Pictish kings … played a part in north Britain which corresponded to the part played in south Britain by the kings of Wessex. Once the main impetus of Scandinavian invasion and settlement had slackened, it was inevitable that Saxons advancing northward should meet Scots advancing southward. It was simply a question of when, where, and upon what terms.”

[16] Gaelic, in a form influenced by close contacts with Ireland, was already one of the languages of the South-West. The strength of Gaelic speech in this area continued to increase in the Middle Ages (Brooke, 1983, 1991; O’Maolalaigh, 1998; and see note 1).

[17] We have to depend on place-names as the main evidence for Scandinavian settlement, but it should be borne in mind that the absence of names does not necessarily mean the absence of settlement. The opportunity and inclination to create new names, rather than take over existing ones, must have been affected by many factors. Ekwall (1925: 72-3) points out that there are parts of England known to have had a substantial Scandinavian population that nevertheless have few Scandinavian names.

[18] See Bald (1928) for some 16th– and 17th– century comments on the languages of Orkney and Shetland.

[19] She suggests a link with a portage route across the Forth-Clyde isthmus. See also Crawford (1987: 130) for a map of ‘hogback’ monuments, which, like the -names, point to a limited Scandinavian presence in the Central Lowlands, emanating from the Danelaw. The same map in ASH (p.72) does not continue over the Border, as Crawford’s version does.

[20] Higham (1993: 196-7) points out that the Viking conquest resulted in the transfer of extensive former Church lands into their ownership, with the consequent sub-division creating ample opportunities for new names.

[21] See also the map of Scandinavian-derived surnames in Crystal (1995: 26).

[22] It may be significant that names retaining the ON genitive singular inflection -ar, e.g. Bowderdale (Cumberland), are mainly found within the Belt (cf. Cameron, 1996: 79), and cf. Butterwhat and Butterdales in Dmf (Williamson, 1943: lvi). The retention of inflections is generally a sign of the living language being more recent. It is conceivable that a living Scandinavian tradition continued at the head of the Solway Firth when this area was settled by Anglo-Normans, and that this is indicated by the considerable number of continental names combined with -bý in Dmf and Cmb, e.g. Lockerbie (Williamson, 1943: 281). However, this is generally agreed to be unlikely. If still productive, the element -bý had perhaps been borrowed into Anglian (Williamson, 1943: 281) – or carried over into Anglo-Danish. But it is perhaps more likely that the Norman personal names were substituted for ON ones in existing -bý place-names (Barrow, 1980: 47; Fellows-Jensen 1984, 1991). A small number of runic inscriptions dating from the late 12th century seem to indicate the latest possible date for the survival of ON in England, but the language is often corrupt and the significance of the occurrence of the runes difficult to interpret. For a review of this question, see Page (1971). It is noticeable that Scandinavians are never addressed as a people in the surviving Scottish charters (Sharp, 1927: 108n.).

[23] I am grateful to Professor Barrow for these references.

[24] Not all of the Dmf -names are included in this map: cf. Nicolaisen’s map in ASH (p.67).

[25] Barrow (1980: Appendix C) comments on the use of Anglo-Danish vocabulary in Lesmahagow (Lanarkshire) in Latin documents before c1250, for example beck and holm.

[26] In fact, to complicate matters, ON originally borrowed it from OE, while the consonant was still /k/, prior to the OE sound-change of palatalisation (see §4.2.2.3).

[27] For instance, Fellows-Jensen writes, “The distribution pattern suggests very strongly that the býs mark the arrival of settlers from the Danelaw, who crossed the Pennines and made their way along the Eden valley to Carlisle and from there continued northwards into eastern Dumfriesshire and southwards along the coastal plain of Cumberland” (1985b: 288).

[28] Though this is unclear (Cowan, 1991). Brooke (1983) goes further, and interprets the kirk inversion-compound names as Anglian in origin, marking an extension of their influence beyond the strongly Anglian parishes of Galloway (where they are relatively lacking), subsequent to the 11th century collapse of Cumbrian power (see above), but this is not accepted by Fellows-Jensen (1987, 1991), who continues to see them as Scandinavian. The Gaelic element in the word-order is undisputed. The sudden ascendancy of Gaelic in a formerly p-Celtic area is as puzzling here as in Pictland. Brooke (1983, 1991) suggests that there may have been a Gaelic-speaking peasantry under Cumbrian and Anglian rule, with Gaelic gaining prestige after the territorial advances of the Kingdom of the Scots. As an added complication, many of the kirk- names have variant forms in Gaelic-derived kil-, and there has been much debate about whether these are the originals of the kirk- names or later translations.

[29] His data were drawn from the findings of the Linguistic Survey of England, which, unlike the corresponding Linguistic Survey of Scotland, stopped at the Border.

[30] In a charter of 1095, whose authenticity is disputed, Edgar acknowledged the overlordship of William Rufus (Donaldson, 1997: 17).

[31] On the functions of the sheriff, see Dickinson ed. (1928: lvi).

[32] On the effect of social shift on language shift, cf. Mallory (1999: 11).

[33] Taylor identifies a number of historical individuals who gave their names to tūn place-names in Fife in the late 12th and early 13th centuries. Their names were variously Anglian, Anglo-Danish and Gaelic (2000: 210-11).

[34] There may also have been a significant number of Anglo-Danish slaves, taken by Malcolm III in various expeditions into the north of England, according to English chroniclers. However, we would not expect that the speech of people at this level of society would have much influence.

[35] Apart from early spellings like chirche, Birchinside (1153-65, later Birkenside, Williamson, 1943: 144), etc., the existence of palatalised forms in the earlier Anglian dialect, in words where they were later replaced, is confirmed by various surviving pronunciations of place-names, e.g. Chesters near Dunbar, Chester near Kirkliston, Chalkielaw near Duns, and the local pronunciation /bɜ:rdʒəm/ for Birgham Berwickshire (Williamson, 1943: xiv).

[36] This is on the assumption that the spellings reflect pronunciation differences. There is a complication, however: early AN scribes sometimes wrote <ch> for /k/. Williamson (1943: xiii) considers this unlikely at least in the sources for the Borders, as “Norman influence is practically non-existent in place-name material in this area”.

[37] Samuels (1985: 279) also tentatively suggests that the fronting of Vowel 7 (see §6.10) may have originated as a hypercorrection amongst Anglo-Danish speakers who migrated to Scotland.

[38] To the extent that Thomason and Kaufman (1988: 71) have suggested that ONhb is not the principal ancestor of the English dialects north of the Great Scandinavian Belt, a suggestion parallel to that made for Scots (above).

[39] Lorimer (1949, 1951) discusses the persistent local tradition that Gaelic survived into the 18th century.

[40] Quotation amended at the request of the author (Michael Montomery).

[41] The mixed vocabulary is treated by McDiarmid (1973: 27-8). Jeffery (1978) has shown further that the vocabulary includes idioms drawn from the speech of both countries. See also Jeffery (1981) on Colkelbie Sow.

[42] But see §7.13.3.

[43] In Joseph’s analysis, standardisation is not an internal process that may happen spontaneously in any language when the right conditions prevail; rather it is a unitary historical process that has spread outwards from Latin, and affects languages as part of the larger process of modernisation/westernisation.

[44] There was a small counter-current flowing southwards: see Jacobsson (1962).

[45] See also Murray (1873: §§13, 14), Bald (1928), McClure (1981).

[46] Aitken (1983) uses the term ‘Anglo-Scots’ for the resulting mixed variety, or cline of mixtures, and this has the merit of making a distinction between anglicised Scots and Scots proper (see below). However, some distinction is also necessary between idiosyncratic texts like the Kingis Quair on the one hand, and texts that participate in a historical process of anglicisation. Here I have reserved the term ‘Anglo-Scots’ for the former. There are many fine degrees of anglicisation, which lead ultimately to a continuum between Scots and English, making the precise delimitation of the term in the latter sense problematic in any case.

[47] Printing arrived late in Scotland, in 1508, even after Scandinavia (see Mackay and Ditchburn, 1997, for a map of the spread of printing in Europe).

Macafee, Caroline and †Aitken, A. J. (2002) ‘A history of Scots to 1700’ in A Dictionary of the Older Scottish Tongue vol. XII, xxix-clvii. Online https://dsl.ac.uk/about-scots/history-of-scots/origins/