August 4, 2014

(Part 1 is here)

We will now zoom in on the linguistic features that have been identified as showing evidence of Norse-Gaelic contacts: the lexicon, preaspiration, and pitch accents. As a phonologist, I won’t have very much to say about the lexicon, but the other two concern me quite a bit. So I’ll start with them.


We start some way off our destination: in English. (If you’ve ever taken an introductory linguistics course, feel free to skip the rest of this paragraph.) In most varieties of English, the sounds written as <p t k> before a vowel are aspirated: if you listen very closely to a word like pit, you might hear a small ‘puff of air’ just after the [p]. This is called aspiration, and in this respect English is different from a language like French or Russian, where this is not usual (and so the lack of aspiration is a good sign of a French or Russian accent in English).

Now when a [p t k] type of sound follows a vowel in English, a number of things may happen, depending on the type of English you speak. One thing that does not tend to happen in English, in this position is that there is a similar ‘puff’ before the [p], [t], or [k]. This phenomenon is called preaspiration, because it’s just like aspiration but comes before the consonant itself.

If you believe the literature, preaspiration is quite rare in the world’s languages (see in particular this paper by Daniel Silverman), but, interestingly, it seems to cluster strongly in Northern Europe. One language that is quite well-known for having it is Icelandic: go here to hear someone saying the word drekka ‘to drink’ – it sounds like [drehka], the [k] is pre-aspirated. And it’s not just Iceland that has it – preaspiration is also found in the closely related Faroese, and by Marstrander’s time it was well established as a feature of the dialects of Western (read: south-western) Norway. Its true extent in the North Germanic languages is actually pan-Scandinavian, as shown in Pétur Helgason’s dissertation.

Now Gaelic most definitely has quite strong preaspiration of the stops [p t k]: we have known this since the earliest descriptions (which, remember, were done by the ear), and a recent paper by Claire Nance and Jane Stuart-Smith has the gory details for those of a phonetic inclination. What’s more, Gaelic shows an amazing variety of ways in which preaspiration is realized, as we shall have the opportunity to discuss later; but see this paper by Anna Bosch for a good overview if you’re impatient, or even just this map. If you want to have a listen, here are some samples.

‘Old Irish as spoken by Western Norwegians’

Apart from Nordic and Gaelic, preaspiration is also a feature of the Sámi languages of central and northern Scandinavia. This clustering is not only noteworthy because the languages are spoken in geographical proximity to each other. The more specatcular fact is that Icelandic, Gaelic, and Sámi languages cannot have inherited it from a common ancestor: Gaelic and Icelandic are of course related, but not very closely; and the Sámi languages are related to both very distantly indeed, if at all. Given all this, it is not at all surprising that scholars have turned to contact in order to account for these similarities.

Why would a very rare feature pop up in three unrelated languages spoken in geographical proximity to each other, and indeed by peoples who have been in close contact? The borrowing explanation seems fairly obvious, especially if you’re coming from a Scandinavian background and have worked on the history (and linguistic consequences) of the Norse settlement in the the British Isles. So it shouldn’t be surprising that preaspiration would be seen as a loan from Norse into Gaelic. Indeed Marstrander goes so far as to say that Scottish Gaelic is, in effect, Old Irish overlaid on the phonetic habits of settlers from Western Norway.

This conclusion is, at face value, eminently reasonable, and has indeed been adopted by scholars working in the area – especially those with a Scandinavian background. Similarly, Lauri Posti argued, as early as in 1954,1 that a similar explanation was available for the Sámi preaspiration, based on the fact that Sámi’s closest relatives lack the feature, and also on the fact that it is less widespread in the eastern Sámi languages – those furthest from the supposed source, the North Germanic language area. So – it’s the Vikings whodunit.

A certain sing-song quality

Another feature of Gaelic that has often been claimed to be connected to the Viking settlement is the so-called ‘pitch accent’. Without going into too much detail, the basic idea of a ‘pitch accent’ system is one where you can have words that consist of exactly the same individual sounds but differ both in meaning and in the movement of pitch. You may have heard of ‘tone languages’ such as Chinese, where (simplifying horribly for now) basically every syllable has to have a particular pitch pattern. Languages with ‘pitch accent’ are not quite as drastic: pitch differences do exist, but are generally restricted to a single syllable – generally the stressed syllable.

For many linguists, languages like Swedish and Norwegian are the typical, best-known examples of ‘pitch accent’ systems. There is great diversity of potential realizations within Scandinavia – here are some examples (click on the transcriptions to hear the words), and these are only from Norway – but the feature is pervasive throughout the region.

The pitch accents are quite a ‘prominent’ feature of Norwegian and Swedish: people tend to know about them and they’re actually discussed at least to some extent when teaching them to foreigners (sometimes under the highly misleading rubric of ‘word tones’). One reason for this would seem to be the rather accidental fact that, for reasons we don’t really care about here, there is a large number of cases where the tones actually are the only thing that distinguishes two words, as in Norwegian bønder ‘farmers’ and bønner ‘beans’ (the nd is pronounced nn in the former), or Swedish anden ‘the duck’ and anden ‘the spirit’. Many of the cases where the pitch accents are crucial in this way involve plurals and definite articles, so it is a good thing to know when you’re learning the language.2

(The existence of these accents seems to have contributed to the folk-linguistic idea of North Germanic languages as ‘sing-song’ – Google ‘scandinavian languages sing-song’ if you don’t believe me. I should like to put this to rest: the description ‘sing-song’ is pretty meaningless linguistically. In fact, I’m told that Swedish speakers from Sweden refer to Finland Swedish as ‘sing-song’ too, even though Finland Swedish lacks the pitch accent system.)

Now Gaelic happens to have a somewhat similar feature, as certain words have to be associated to a certain pitch pattern. There are even a few pairs like the ones in Swedish or Norwegian that only differ in pitch, as in balg ‘belly’ (pronounced roughly as [balag] with a rising tone throughout) and ballag ‘skull’ ([balag] with a high or rising tone on the first syllable and a fall on the second). That certainly looks a lot like the North Germanic system, and has indeed been analysed as such (notably by Elmar Ternes, here). Now it’s important to note that the system is nowhere near as pervasive, or prominent for the learner, as the one in Swedish or Norwegian, largely because it’s not as involved in productive morphology (although you do get changes of the pitch pattern in forms of the same word: for instance, the first syllable in leabhar ‘book’ has a different pattern from the first syllable in leabhraichean ‘books’), but it still exists.

Settlers and geography

One striking fact about both of these features is that they are so clustered in the north-west of the Gaelic-speaking area. The pitch accents are certainly best described for places like Lewis and Applecross (Wester Ross). The type of preaspiration that’s most like the Icelandic one is also characteristic of Lewis and Wester Ross. Lewis in particular, of course, is known as a stronghold of Norse settlement. So it all fits together: Norse settlers occupy the place, but then shift to Gaelic and import their phonetic habits to produce both preaspiration and pitch accents.

This rather makes sense, on the face of it. It is also supported by Thomas Stewart’s study of lexical borrowings: they are not only numerous but also, Stewart argues, show evidence of imposition, i.e. speakers of the ‘source’ language shifting to the ‘new’ language and bringing their habits with them (exactly as Marstrander argued), rather than borrowing by already-extant Gaelic speakers from their Norse neighbours. A similar scenario is endorsed by Gunnar Ólafur Hansson in his paper on preaspiration.3

It all makes sense, right? Rare features jsut pop up in unrelated languages. And we know that the languages definitely were in contact (even though we don’t actually know much about some pretty crucial aspects of this contact). So it’s a reasonable explanation.

In the next part we’ll see that it is (of course) more complicated than that.


1 Posti, Lauri. 1954. On the origin of the voiceless vowel in Lapp. Svenska landsmål och svenskt folkliv 7677, 199–209.

2 The whole thing might be a bit overblown. OK, one is never going to sound completely like a native speaker until one gets a grip on the pitch accent, but how many contexts are there when one can mix up ducks and spirits (not the drinks)?

3 Gunnar Ólafur Hansson. 2001. Remains of a submerged continent: preaspiration in the languages of Northwest Europe. In Laurel J. Brinton (ed.) Historical Linguistics 1999. Amsterdam: John Benjamins, 157-173.

Blog archive  |
comments powered by Disqus

About me

I’m Pavel Iosad, and I'm a Lecturer in Theoretical Phonology in the department of Linguistics and English Language at the University of Edinburgh. ¶ You can always go to the start page to learn more.



Subscribe to the  RSS feed, or follow me on Twitter at  @anghyflawn.