It is quite common knowledge that the history of English has been significantly influenced by the settlement of North Germanic speakers in the north-eastern part of what is now England. From relatively trivial words like skirt and window to the less commonly borrowed items like they and all the way to the general simplification of the Old English system of inflections (e.g. case in nouns) in the Middle English period: you name it, the Vikings have been blamed for it.
Here, I’d like to explore another part of the Norse influence on the language landscape of the British Isles: their impact on Scotland. Again, it is commonly acknowledged that the Norsemen are part of the cultural and linguistic mix that produced the Scottish nation as we know it. This is probably most obvious in the Northern Isles, where the Norn language, whose closest relatives are found in the west of Norway, on the Faroe Islands, and in Iceland, perisisted for about the same length of time as Cornish did in the south-west of Britain; but the Nordic linguistic heritage in Gaelic Scotland is also considerable.
History, archaeology, and literary sources all bear out a significant Norse presence in the Hebrides and around most of the Irish Sea. For some parts of this region the historical record is (relatively) abundant – this is, of course, not surprising in view of the strength of the chronicle tradition in Ireland; for one recent overview of the region, see David Griffiths’ Vikings of the Irish Sea. For others, this historical record is much sparser – and the Hebrides, today’s Gaelic heartland, are a case in point.
The use of linguistic evidence to tease out unrecorded contacts between populations has a long and venerable history. Very often this takes the form of the search for a substrate – a language whose users shift to a different one but which ends up influencing their version of the ‘new’ variety. Notable examples of substrate theories include the suggestion, going back at least to John Morris-Jones, that the Celtic languages have a ‘Hamito-Semitic’ substrate, Sigmund Feist’s proposed pre-Germanic language, or Theo Vennemann’s Vasconic effort. In all of these cases, certain features of the modern languages are surmised to be borrowings, but the problem is that while the linguistic conclusions may or may not be sound, depending on one’s standard of evidence, the very existence of population contact that would lead to the ‘correct’ sort of language contact remains largely speculative.
A slightly different facet of the same problem is presented by phenomena such as putative Celtic influence on English. Here, the fact of contact is not usually in doubt: there is widespread agreement that incoming Germanic speakers would have encountered a Celtic-speaking population.1 Whether this population contact would have led to significant linguistic influence has been the source of much debate, a lot of it helpfully summarized in a relatively recent book.
There are two main issues around the ‘Celtic hypothesis’. One is more of a chronological nature: the putative contact features do not really show up until the Middle English period, although people have proposed various ways of dealing with the issue. A second problem is our lack of knowledge about the precise circumstances of the contact, which, as Peter Trudgill in particular has shown in detail, are crucial to determining the existence and nature of any linguistic contact effects. (For what it’s worth, though, Trudgill endorses the Celtic hypothesis in Investigations in sociohistorical linguistics – a short and really accessible book that I cannot recommend highly enough.)
What about the Norse linguistic influence in Scotland, and more specifically in Gaelic Scotland? In many ways the questions are similar to those posed of the Celtic hypothesis for English: we know there was population contact, but we don’t really know what sort of contact it was, and the chronology doesn’t square up very nicely. Yet there are also interesting, if maybe subtle, differences.
A lot of early descriptive work on the living Celtic varieties happened to be carried out by Nordic, and particularly Norwegian scholars. The ‘grand old man’ of Celtic studies in Norway was Carl Marstrander. He trained as an Indo-Europeanist under Sophus Bugge and Alf Torp. Although some of his early work concerned the historical and linguistic aspects of Norse settlement in the British Isles seen from the Nordic perspective (as in Bidrag til det norske Sprogs Historie i Irland [Contributions to the History of the Norwegian Language in Ireland], published 1915), he was also a pioneer in the documentation of the living Celtic languages, having worked on the Blasket Islands with Tomás Ó Criomhthain of An t-Oileánach fame and on the Isle of Man. Among his disciples were Carl Borgstrøm (author of several descriptions of Scottish Gaelic dialects), Alf Sommerfelt, who worked not only on several varieties of (mostly Ulster) Irish but also on Welsh and Breton, and Magne Oftedal (author of what is probably the most comprehensive published description of a Scottish Gaelic dialect, The Gaelic of Leurbost, Isle of Lewis).
All of these scholars had a background not just in Celtic studies but also in the Nordic philological and dialectological traditions (for instance, Oftedal’s dissertation, Ordstrukturen i Gjestalmålet was a phonological description of the dialect of Gjesdal in south-west Norway). With this sort of history, it is not really surprising that they would get onto the case of some striking linguistic similarities between Nordic and Gaelic.
Marstrander was probably the first to draw attention to these, in Okklusiver og substrater [Stops and substrates], published in the 1932 issue of Norsk tidsskrift for sprogvidenskap (Norwegian Journal of Linguistics); his theories received support from Oftedal (Jærske okklusivar [Jæren stops], also in NTS, 1946) and were elaborated and extended by Borgstrøm (The influence of Norse on Scottish Gaelic, 1974).
There are three key areas where Norse influence on Gaelic has been claimed to be most evident: the lexicon (i.e. loanwords), the pronunciation of voiceless stops (p t k), especially in the middle of words, and the use of pitch to distinguish between otherwise identical words. The similarities between Norse and Gaelic in all these domains are indeed quite striking. We will consider especially the latter two in the next posts of this series.
1 Dissenting opinions do exist: the Celticist Peter Schrijver has proposed that the locals in question would be Latin rather than Celtic speakers. It also seems to me that if Guy Halsall’s proposals in the excellently thought-provoking (to this non-medievalist, non-historian) Worlds of Arthur are on the right track, then this picture may need to be revised as well.