Study

This article considers the way in which academic linguistics misrepresents the Shetland tongue in order to mine it for linguistic minerals. Particularly noteworthy is the way in which David Crystal uses a deliberately Anglicised version of Shetland speech to support his ideological stance that 'of course' many Shetlanders speak a 'much less distinctive dialect' than an example taken from J.J. Graham, the essential phonological distinctiveness of which he dismisses as 'choice of spelling.'

The Study of Shetlandic.

This article addresses the following issues:

1. Local and Academic Study. The dichotomy which exists between local and academic study of Shetlandic.

2. Representation and Presupposition. Presuppositions behind statements made about Shetlandic.

3. Methodological Questions in the study of Spoken Shetlandic. Problems arising in the modern study of spoken Shetlandic.

1. Local and Academic Study.

The study of Shetlandic could be roughly divided into two categories:

1. Local studies, such as the glossaries of Edmonston (1866) and Angus (1924?), The Shetland Dictionary of John J. Graham, and Grammar and Usage of the Shetland Dialect (GUSD). Generally speaking, local studies - with the important exception of GUSD - have been mainly concerned with vocabulary, often with a particular emphasis on words of Norse etymology.

A further division could be made between local studies which are primarily antiquarian in intent, and those which seek to enhance the use of modern Shetlandic. The Shetland Dictionary and GUSD are primarily in the latter category, in that they concentrate on relatively modern rather than older usage; give practical advice about issues such as spelling; and are not concerned with issues such as etymology which do not affect the actual use of the language.

2. Study by individuals and organisations outwith Shetland. Examples are the lexicographic studies of Jakobsen and the Scottish National Dictionary; the lexical and phonological studies of the Linguistic Survey of Scotland; and the ongoing studies of Scandinavian scholars such as Gunnel Melchers and Bengt Oreström.

Studies from outwith Shetland have typically dealt with Shetlandic as part of wider fields of study. Jakobsen, for example, was explicitly concerned to find survivals of Norn, and was not at all interested in how modern Shetlandic functions as a linguistic system. Consequently, his phonetic script is often quite difficult to interpret, because he was interested in identifying the reflexes of Old Norse phonemes, not in understanding how they function in modern Shetlandic phonology. As Barnes says:

In a rare outburst of criticism, Stewart (1964: 172) describes the listing in the dictionary of twenty-five variant pronunciations of the word gopn, ‘the hollow of the hand, a handful’ (1928-32: 253) as ‘phonetics run riot’, and it is hard to disagree with this verdict. The question we would wish Jakobsen had addressed is: how did speakers distinguish gopn from other words - that is, what were the distinctive units of which his twenty-five variant pronunciations were the realisations?

Shetland’s Northern Links, p. 6

Barnes is here pointing out that Jakobsen’s investigations of Shetlandic pronunciations were phonetic, not phonological. However, Barnes, too, is interested in the phonology of modern Shetlandic only as a means to an end. As he goes on to say:

Had he been able to establish, however tentatively, a system (or systems in the case of significant dialectal variation) of distinctive speech sounds, we would not only have known more about Norn in its last years as a living language, but also have been able to trace more easily the lines of its development from Old Norse. In the case of Faroese, for example, we can for the most part predict with certainty what the present-day reflexes of an Old Norse word will be. That is in no way true of Norn.

The point I have made by reference to sounds and sound systems applies equally to other levels of linguistic analysis. Few answers are provided by Jakobsen for the person with questions about the morphological, syntactic or semantic systems of Norn.

Viewed from the modern perspective, then, Jakobsen’s Norn studies leave many gaps...

ibid, p. 9-7.

It is clear here that, although Barnes deplores the gaps in Jakobsen’s material caused by his failure to use modern methods (which he scarcely could have been expected to do, not being modern) his motivation is essentially similar to Jakobsen’s, in that it is the extinct Norn, rather than the living Shetlandic, that is his focus of interest. In both cases, Shetlandic is a field to be excavated in the service of Nordic studies, rather than a linguistic system to be studied in its own right.

Barnes’s implication that the opportunity to establish the reflexes of Old Norse phonemes in Norn is past, and correlatively that this is due to a failure in Jakobsen’s methodology, could be questioned. Firstly, it assumes that the survivals of Norn vocabulary in modern Shetlandic - that is, the Shetlandic of the late 19th Century studied by Jakobsen - would have been retained in an accurate enough form to facilitate such analysis. Secondly, it begs the question of whether any results which Jakobsen could have obtained could not still be arrived at by a holistic study of Shetlandic phonology, which would place Jakobsen’s evidence in the context of how Shetlandic phonology works as a whole, rather than the attempt to extrapolate evidence for Norse without properly understanding the system in which these Norse words have now been embedded.

Conversely, the Linguistic Survey of Scotland is interested in Shetlandic primarily for the light it can cast on the historical development of Scots. Thus J.C. Catford opens his article Shetland Dialect (Shetland Folk Book 3, 1957) with a defence of the relevance of the Orkney and Shetland dialects for extracting evidence for older Scots pronunciations, and finishes by saying that ‘Shetland dialect, then, on account of its somewhat archaic character, is of the greatest interest to Scots as well as Scandinavian scholars’ and that the work of the Linguistic Survey should ‘make it possible on the one hand...to relate the Norn features of Shetland dialect fairly precisely to particular Scandinavian dialect areas, and on the other, to extract as much as possible of the evidence for earlier Scots pronunciation, in which Shetland dialects are so rich.’ Again, the motivation for the academic study of Shetlandic is external, related to the academic interests of bodies outside of Shetland, and unrelated to the holistic study of Shetlandic as used by Shetlanders.

Thus the internal study of Shetlandic is usually undertaken by local enthusiasts whose main interest is in vocabulary, often with an etymological bent; and academic study by non-Shetlanders whose interests are in using Shetlandic evidence within wider areas of study. The academic study of Shetlandic by speakers of Shetlandic, with the accruing two-way interaction between that study and the spoken and written use of Shetlandic, such as exists with e.g. English, Gaelic and Faroese, is almost entirely lacking. As a result, there is, for example, no dictionary which gives an adequate guide to Shetlandic pronunciation.

2. Representation and Presupposition.

Misleading representations of Shetlandic in works of reference can sometimes be traced back to the presuppositions of the studies on which they are based. Whereas the article on Orkney and Shetland Dialect in The Oxford Companion to the English Language, by Scots scholar A.J. Aitken, is excellent within its space limitations, the treatment of Shetlandic by David Crystal, in The Cambridge Encyclopedia of the English Language (p. 332) is another matter. First he gives an example of Shetlandic writing by John J. Graham:

Shetland’s ‘ain aald language’ has its röts awa back ida Norn tongue at wis spokken in Shetland fae aboot da nint tae da seeventeent century. Da Scots fock at cam among wis fae da sixteent century an on brocht der ain leid, an at da lang an da lent da twa languages melled tagidder to mak da tongue we caa Shetlandic. While dis wis gjaan on, anidder wye o spaekin an writin wis shapin da local speech. Dis wis English - ösed by da Kirk, da laa-coorts an ida sköls.

From J.J. Graham, 1981.

Then, remarking that ‘much of the visual identity of this style is due to the choice of spelling (e.g. da for the)’ he comments: ‘Many Shetlanders, of course, speak a much less distinctive dialect, as seen in this extract from a 64-year-old man’s account of the fire festival, Up-Helly-Aa. The language is essentially Standard English, with just the occasional dialect form.’

well - hit’s a procession of maybe nine hundred guizers and if you do not ken what a guizer is that is somebody that dresses up to pretend to be something else...in the old days part of the reason for the festival was to celebrate the end of the dark days of winter and the return of the sun well whatever else has changed the weather has not changed very muckle - hit’s still the dark days of winter..and there’s nobody now more than in the old days blither to see the return of the sun even though it’s still not very high in the sky than the Shetlander is.

After B. Oreström, 1985.

This is certainly little different from standard English. It is, however, very unlikely to be representative of the speech of a 64 year old man in Lerwick in 1985. The first impression one gets is of someone doing his best to speak standard English in order to be understood by a foreigner. However, on examining the source of the above extract (A Corpus of Shetland English, edited by Bengt Oreström) it becomes evident that there was a deliberate policy of Anglicising the texts in order to make them more readable by non-Shetlanders. Oreström says:

As the purpose of this corpus is to supply a material chiefly for studies in grammar and intonation, the local writing convention with all its problems would be an unnecessary complication for a reader not familiar with this form. The written representation of the corpus material should provide for easy readability also to non-Shetlandic users. Therefore, the standard English orthographical convention has been adhered to as far as possible. Only in cases where a word is not easily recognisable as an ordinary English one and/or is exclusively local, has it been written in a specific way and given in the Glossary.

In order to illustrate the effect of this policy of Anglicisation, the same text is given, by Oreström, in both a typical Shetlandic spelling and the Anglicised form.

(a) Noo dan does du see dey wir dis peerie Jockie boy at bade up aboot Hurdibeck wi his grandmidder an he hed a lovely peerie dug for his very owen. An so dey wir wan time at he happened to be oot a daeks lookin among da twartree sheep at dey hed at Aisha. An him an his dug wir comin hame ower be da sooth side a Turrieshun on dy wye a da Dutch Loch, him whistling an his dug keepin closs be his fit laaching away with his tongue hingin oot.

(b) Now then does du see there were this peerie Jockie boy that bade up about Hurdibeck with his grandmother and he had a lovely peerie dog for his very own. And so there were one time that he happened to be out of dykes looking among the twartree sheep that they had at Aisha. And him and his dog were coming home over by the south side of Turrieshun on their way of the Dutch Loch, him whistling and his dog keeping close by his foot laughing away with his tongue hanging out.

It can be seen here that the Anglicisation of the text involves (a) the replacement of all English cognate words by their English cognates (e.g. hame by home; da by the; dug by dog; daek by dyke; noo by now; laach by laugh; closs by close; owen by own; be by by; fit by foot) thus obscuring all the phonological differences between Shetlandic and standard English, including (b) ‘replacing’ consonants which are present in English but not in Shetlandic (e.g. at by that; a by of; and dey by there - curiously the first version also has with rather than wi.) The last of these - the replacement of dey by there - is a particularly bad representation of the Shetlandic, because the grammatical use of dae (as I spell it) is different from that of English there, in that it is always followed by a plural verb; and its replacement by there gives the effect of an in-between language which is neither Shetlandic nor English. In fact, the passage as a whole reads like a Shetlander trying hard to speak English and ‘faain trowe it’ (falling through it) - exactly the impression given by the example which Crystal cites to prove that many Shetlanders speak a much less distinctive dialect.

It can be seen that, when this Anglicisation of the second text is allowed for, there is little difference between the linguistic nature of the Graham and the Lerwick extracts to justify the claim that the Lerwick extract is ‘much less distinctive’. Neither passage contains very many words without close cognates in English. To a Shetlander, the only unusual word in the first extract is leid, which Graham seems to have borrowed from Lallans or literary Scots. (In Graham’s own Shetland Dictionary, the only meaning given for leid is ‘diligence’, this being a completely different word from Scots leid meaning language). Of the words glossed by Crystal, röts (roots), ida (in the), fock (folk), at wis (that was), brocht (brought), and ösed (used) are all close cognates of English, of the type which Oreström’s policy would have Anglicised in the second text. The only peculiarly Shetland idiom is at da lang an da lent, meaning at long last; and the only word which does not have a particularly close English cognate is the familiar Scots kirk. In the Lerwick extract, although fewer words are glossed, two of these - ken and muckle - are not close cognates of standard English words, though again they are common Scots; and the implication is that the type of words which are glossed in the first text have simply been replaced by their English cognates in the second. It can be seen, then, that it is really only the stated process of Anglicisation followed by Oreström which makes the second extract seem significantly less distinctive from standard English than the first. Whatever may be said about the validity of this practice for the specialised purposes of Oreström, it is certainly remarkable that Crystal should have taken, as an example to illustrate the non-distinctiveness of Shetlandic, a passage which has explicitly had most of the distinctive features removed.

When we investigate the type of differences from standard English which occur in the Graham text - differences which Crystal attributes partly to ‘the choice of spelling’ - we can see that most of these reflect actual differences in Shetlandic pronunciation as opposed to that of standard English; and moreover, that most of these differences are phonological rather than merely phonetic. For example, there is a phonological difference between the ‘d’ sound and the ‘th’ sound (both voiced and unvoiced) in Shetlandic, and therefore it would not be practical to represent the ‘d’ sound by a written <th>. Similarly, the <ö> spelling used by Graham represents a different phoneme from that represented by the <oo> in e.g. coorts; <ae> in spaekin is different from both <ea> and <ai>; and <aa> is phonologically distinct from <a>. The choice of spellings is therefore a fairly minimal representation of the essential features of Shetlandic phonology. There has been no attempt to make the spelling more distinctive from that of standard English in other ways - for example, by spelling language as *langwidj, tongue as *tung or back as *bakk. The only obvious concessions to non-English spelling conventions, apart from necessary indications of phonological distinctions, are the use of <j> rather than <y> in gjaan and the double <k> - which is uncommon in English spelling - in spokken. Graham does not even go out of his way to avoid English spellings where the Shetlandic sound is different - using e.g. has rather than the more typical Shetlandic haes; Shetland rather than Shaetlan; and the English form speech rather than the more colloquial spaek. Far from choosing non-English spellings deliberately to emphasise, far less exaggerate, the differences between Shetlandic and English, Graham’s spelling is conservative, using only the minimum changes from standard English convention in order to represent the Shetlandic pronunciation. Apart from the borrowed word leid, the passage as a whole is written in natural everyday Shetlandic.

This is an illustration of the effect which presuppositions can have on the representation of language in writing. Writers of literary Scots are often accused of exaggerating the difference between Scots and English. Here, however, we seem to have the opposite tendency. It would seem that Crystal - as his comment that ‘of course’ many Shetlanders speak a much less distinctive dialect would imply - is presupposed to regard the Graham text as artificially distinct from English, and thus has chosen a text in which the distinctive features have been removed in order to represent it as closer to standard English. It is true that many younger Shetlanders now speak a much less distinctive language than that represented in the Graham text - a language which is little different from Scottish Standard English - but that would have been extremely atypical of the speech of a 64 year old in the early 1980s.

3. Methodological Questions in the study of Spoken Shetlandic.

Apart from its inappropriate use by Crystal, the treatment of Shetlandic by Oreström raises questions about methodology in studying Shetlandic.

Oreström’s stated aims are as follows:

The aim of the field work that started in 1982 was to record a cross-section of Shetland English in the form of informal, spontaneous speech, to transcribe the recordings, and to supply the transcriptions with a prosodic analysis. The linguistic bias of the corpus was decided to be primarily for studies on grammar and intonation, to some extent on lexis. It is not intended for detailed phonological studies, which, in order to capture subtle differences of pronunciation, will require specific methods, and tape-recorders of a high quality.

page 14

Oreström recognises that this aim - the emphasis of which is on ‘informal, spontaneous speech’ - is, in fact, unattainable. He quotes Labov:

the aim of linguistic research in the community must be to find out how people talk when they are not being systematically observed; yet we can only obtain this data by systematic observation.

The logical conclusion of these two statements would be that the scientific observation of language is impossible, unless done by ethically unacceptable means such as secret recordings. Oreström presents a compromise approach; quoting Duncan & Fiske:

the goal is to minimise, within the constraints of research requirements and ethics, intrusion upon the interaction in question.

Oreström recognises the effects on an informant of being recorded as (a) increase of tension, resulting in less spontaneity, and (b) orientation ‘away from his normal degree of broadness in the direction of more standard English in a misguided attempt to "do his best" by trying to come out as clearly and intelligibly as possible on the tape’; and he comments that ‘many of us...when asked to produce speech in front of a recorder...would change our language to some extent, accommodating to a listener whom we are not certain will otherwise fully understand what we say.’ Recognising that ‘Even if these undesired effects cannot be wholly eliminated, they have to be reduced’ Oreström describes measures taken to minimise these problems, including making recordings in the speaker’s own home with natural seating arrangements in order to decrease tension. To address the problem of linguistic accommodation to the interviewer:

They were informed that their speech, however broad, did not cause me any problems of interpretation and they were asked to speak as they would do to local people. Presumably, one of the best ways of obtaining natural broadness is to avoid mentioning (or at least to tone down the fact) that it is their speech you are concerned with and instead show great interest in the story that is to be told. This approach was therefore generally taken here.

The second of these points is immediately open to question. It seems equally likely that an informant, under the impression that the story is the main interest, will try all the more to make the content intelligible to non-Shetlanders. I remember as a child hearing a story about a shipwreck, recorded in Foula, and being surprised that so much of it was in English. Although this recording had been made by a Shetlander, the informant had nevertheless altered his speech for recording purposes.

As regards the first point, it could be suggested that Shetlanders would be unlikely to believe this. The actual difficulty which Oreström evidently had in understanding the Shetlandic speech is illustrated by several comments, such as ‘notes were taken down concerning unclear words in order to facilitate the later transcription work’ and:

As could be expected, the transcription work proved to be time-consuming, involving endless reruns on the recorder with 3-5 words transcribed at a time. The video turned out to be a surprisingly valuable help here because many doubtful words could be solved by reading the movements of the informants’ lips. When all this work was finished, a batch of new queries concerning unclear words had accumulated and it was necessary to visit some of the informants again.

The picture presented here of the difficulty involved in transcribing the text scarcely inspires confidence in the assurance that even the ‘broadest’ speech would be easily understood. Indeed, it is not clear that all the difficulties were ironed out even after this process of repeated consultation and lip-reading. In the text in two orthographies given above, the following sentence occurs:

Text (a): An him an his dug wir comin hame ower be da sooth side a Turrieshun on dy wye a da Dutch Loch...

Text (b): And him and his dog were coming home over by the south side of Turrieshun on their way of the Dutch Loch...

The phrase (a) on dy wye a; (b) on their way of makes sense in neither version. To a native speaker of Shetlandic, the immediate impression is that the first phrase should be on da wye o, meaning in the direction of, and that the definite article da has somehow been transcribed as dy, perhaps because the idiom has not been understood. How this came to be interpreted as their in the second version is not clear, however, because in Shetlandic dy does not mean their, but thy (which would obviously be meaningless in context) unless the idiom has been conflated with on dir wye tae = on their way to. Without the first - Shetlandic - version to refer to, the second text would have been incomprehensible. It is not, of course, possible to say categorically that the phrase on dir wye o does not exist in any part of Shetland; and if the first text had not been available - as texts in Shetlandic orthography are not available for the other texts in the corpus - it would have been impossible to say whether this was an unusual local idiom; a misinterpretation by the researcher (which is what it seems to be); or a slip of the tongue on the part of the informant.

Even if the informants had accepted in theory that an English-speaking researcher could understand their Shetlandic speech (and it is clear from the explanations offered in the texts - e.g. ‘the lighter ground, or what we call taatie möld’ (Text C1, p. 32) - that they, perhaps wisely, did not) it is questionable whether they would have been able to speak naturally in Shetlandic under these circumstances. Most Shetlanders automatically accommodate their speech, to one extent or another, to English speakers, unless the English speakers have been resident in Shetland long enough for the Shetlandic speaker to be confident that they will understand everything. It is likely that much better results could be obtained if the researchers themselves were speakers of Shetlandic.

Oreström’s glossary and description of phonology are likewise surprising. In the glossary, there are several strange translations, e.g.

atidda prep. into (This is actually itae da meaning in the. In addition to ignoring the definite article, the researcher has failed to appreciate that Shetlandic itae or ita means in, not into which would be intae.)

bannock n. a flat home-made loaf made of oat meal. (I have seen, and eaten, beremeal bannocks and white flour - flouermael - bannocks, but never oatmeal ones.)

geng vb. to dwell (geng means to walk or go)

knave adj. fist-sized (nev, Scots knieve, normally a noun meaning fist)

swa adj. small (presumably a misprint for sma)

put vb. [pat] (Usually [pɪt] - the pronunciation [pat] is preterite, or past tense).

In the phonological description, there are some strange assignations, e.g:

StE (RP) ShE

[ɪ] [e] dinner, sit, fish

[aɪ] [əi] time, climb

[əʊ] [o:] home, go

The headings StE (RP) and ShE stand for Standard English (Received Pronunciation) and Shetland English respectively.

In the first example, dinner does not normally have the same vowel in Shetlandic as do sit and fish. Common pronunciations would be [dɛnər], [sɜt] and [fəɪʃ], representing three different phonemes: /ɛ/, /ɪ/ (of which the [ɜ] pronunciation is here an allophone) and the diphthong /əɪ/. It is possible, of course, that some Shetlanders do pronounce all three words with the same vowel even when speaking Shetlandic; but in general this would be more characteristic of a Shetlander speaking English. It is therefore unclear whether the choice of [e] is intended to represent the phonetic difference between an RP pronunciation and that of a Shetlander speaking English, or is intended to represent a phonological characteristic of Shetlandic speech per se.

In the second example, time and climb do not normally rhyme in Shetlandic - common pronunciations would be [təɪm] and [klɪm]. Again, although some Shetlandic-speaking Shetlanders may pronounce them with the same vowel, [kləɪm] would be more characteristic of ‘knappin’ (speaking English).

In the third example, the characteristic Shetlandic pronunciations of home and go are haem - variously [hem]/[hɛɪm]/[him] - and gae [ge:]. The pronunciation [go:] is used in Shetlandic, both as a regional variant and as a free variant (perhaps particularly for emphasis) in other regions. However, although there is a variant of haem with a rounded vowel, this does not rhyme with go, but has a more open vowel, [hɔm], as the Shetlandic spelling hom suggests.

The reason for these strange assignations appears to be a failure to appreciate that, although there is a continuum (or, it might be more accurate to say, there are different kinds of intermediate stages and accommodations) of speech in Shetland between Shetlandic and standard English, there is nevertheless a difference, both as perceived by Shetlanders and in phonological terms, between the two; and a description of Shetlandic phonology which does not make this distinction (as, for example, the Linguistic Survey of Scotland does) is of dubious value. Secondly, the type of English spoken by Shetlanders when they are speaking English, as opposed to Shetlandic, is not Received Pronunciation English but Scottish Standard English (more or less); and thus the attempt to contrast ‘Shetland English’ with ‘Standard English’ fails to distinguish those features which are characteristic of Shetlandic as opposed to Scottish Standard English from those features which are characteristic of Scottish Standard English as opposed to Received Pronunciation English.

Taken as a whole, these anomalies underline the dangers of studying Shetlandic without the complementary advantages of a thorough understanding of the phonology and the intuition of Shetlandic speakers (of which phonology is essentially a description). More particularly, they underline the dangers of regarding speech in Shetland as a whole as Shetland English, without recognising the differences between Shetlandic and SSE.

Methodologically, it is difficult to avoid the conclusion that many of the anomalies could have been avoided had there been a Shetlandic speaker involved throughout the research process. The informants would have been much more likely to reply in authentic Shetlandic to a Shetlandic than to an English speaker, particularly one with a foreign accent. A Shetlandic speaker would probably have had no difficulty understanding the ‘unclear words’ which apparently occasioned repeated consultations and lip-reading, and would be unlikely to make transcription errors such as appears to have been made in the translation of on dy wye o as on their way of.

Overall, like the approach of Jakobsen, Oreström’s approach raises the question of how far Shetlandic can be used as a source for particular kinds of information - whether etymological or prosodic - without an understanding of how Shetlandic functions as a whole.

A more useful treatment of Shetlandic phonology is offered by the Linguistic Survey of Scotland (in The Linguistic Atlas of Scotland, vol 3.) which was published around the same time as Oreström’s study. Although undertaken within the overall framework of Scots phonology, this study recognises most of the peculiar features of Shetlandic phonology, such as contrastive short and long vowels in words such as need [nid] and meid [mi:d] (landmark), or guid [gød] and roed [rø:d] (speak nonsense). There remain, however, some strange allocations which seem to be owing to a failure to appreciate the extent to which vowels in Shetlandic undergo a systemic mutation before certain types of consonant. Thus the LAS data requires considerable analysis in order to be useful for practical purposes such as the representation of Shetland pronunciation in word lists. Overall, however, the LAS illustrates that it is necessary for Shetlandic to be approached from a thorough understanding of (a) Scottish Standard English, and (b) Scots, before its own phonology can be holistically described.

Questions raised:

1. The possibility of identifying the reflexes of Norse phonemes in modern Shetlandic without an understanding of how Shetlandic phonology, in which these reflexes are now embedded, works as a whole.

2. The possibility of giving an informed representation of Shetlandic without adequate consideration of the presuppositions, scope and methodology of sources.

3. The possibility of giving a single description of the phonology of Shetland English, without a further division into Shetlandic (or The Shetland Dialect), and Scottish (or Shetland) Standard English.

4. The difficulties caused by conducting an investigation into spoken Shetlandic through the medium of spoken English.

5. The difficulties caused by studying Shetlandic phonology from a starting-point of Received Pronunciation English.

6. The desirability of Shetlandic being studied, and consequently represented to the academic community, by researchers who do not fully understand it but must approach it through the medium of standard English.

Further Notes.

1. Jakobsen’s Treatment of gopn.

In Jakobsen’s defence, it should be said here that his transcriptions of gopn are not as confusing as it might seem. Although he records multiple variants - e.g. [gɔpən], [gjɔpən], [gɔpəm], [gjɔpəm] - these can be analysed into categories of variation:

1. Whether or not the final [n] is assimilated to the penultimate [p] to become [m]

2. Whether or not the initial [g] is followed by [j]

3. Which vowel is in the first syllable.

The first of these is a straightforward case of an original /n/ being assimilated to the foregoing labial /p/ to give [m], and is not phonologically significant. Similar conditioned variations occur in words like lippen [lɜpən] or [lɜpm] (expect). The second may have some bearing on etymology - Jakobsen postulates Shetlandic [jo] as a reflex of O.N. <au> - but the insertion of [j] after certain initial consonants, particularly velars, is a common feature of regional variants in Shetlandic and so the variation may be explicable by analogy. This leaves the vowel of the first syllable as the principle variation. Conveniently (and unusually) Jakobsen offers a regional breakdown of pronunciations of this word which gives the regional values for this syllable (with more than one pronunciation given for some regions):

[ɔ] Whalsay, Sandsting, Aithsting, Sandness, Delting, Northmavine, North Isles, North Roe

[ɒ] Foula, Sandsting, Aithsting, Sandness, Delting, Northmavine, North Isles

[o] Foula

[o:^ə] ([ə] superscript) Unst, Lunnasting

[u] Dunrossness, North Roe

(The above lists are extrapolated from Jakobsen’s different arrangement, with conventional symbols corresponding to his idiosyncratic phonetics.)

Of the above, [ɔ] and [ɒ] are fairly obviously phonetic variations of the same phoneme; and the single occurrence of [o] - which would correspond to written <u> in conventional Shetlandic spelling, - seems to be a rogue variant of the same. This leaves three essential variants:

/ɔ/ - the most prevalent

/o:/ - in Unst and Lunnasting

/u/ - in Dunrossness and North Roe

Of these, /u/ is a regular development from Scots gowpen, a double handful (the usual meaning of the word in Shetlandic) - cp. loop, usually [lup], Scots lowp, usually [loup]; coop [kup], Scots cowp [koup]. The other two pronunciations are, according to Jakobsen, proof that the word is descended from O.N. gaupn, the hollow of the hand. What can be seen here, then, is a word with regional variants which reflect both Norn and Scots phonology. The usual sense of the word is that of Scots gowpen, a double handful.

However, Jakobsen rarely provides regional references for his recorded pronunciations which can be analysed in this way; and to that extent the criticisms of both Barnes and Stewart are justified.

2. John Stewart’s Phonetic Transcription.

John Stewart, whom Barnes quotes as a critic of Jakobsen’s over-indulgence in phonetic detail, takes an opposite approach to Jakobsen’s in his Shetland Place Names. Although he includes a phonetic transcription of the place names which he records, he offers no guide to his use of phonetic symbols which would relate them to characteristics of Shetlandic (or Scots, or Norn) phonology. Also, although he shows length distinctions for some vowels he does not do so for others, and even where he does it is not clear whether he does so consistently. For example, he sometimes uses a diacritic to indicate the long [e:] sound which occurs in words such as Mainland [me:nlənd] as opposed to gaet [get] (path). When, then, he fails to indicate a long [e:] sound in his transcription of Northmaven [nɔrtmevən] it is impossible to tell whether he is recording a local variant with a short vowel (not inconceivable before [v] in Shetlandic); or relying on the SVLR to determine the length (unlikely in Shetlandic); or simply being inconsistent. Moreover, he gives no length indication for other vowels such as [i], [u] or [ø] - two of which are recorded as having contrastive length in the LAS - though he records the difference between [ø] and [y], which is phonetic - characteristic of the West side of Shetland - rather than phonological. As a result his transcriptions are not adequate as descriptions of the pronunciation of Shetland words - at most, they give a rough idea - and although his approach is at an opposite extreme to the hyperphonetic approach of Jakobsen, the practical outcome is the same: the actual phonology of the words listed is in many cases a matter of guesswork. Both the Jakobsen and the Stewart approaches to representation of Shetlandic pronunciation are inadequate for any practical, or indeed academic, purpose, because neither of them is based on a holistic understanding of Shetlandic phonology.

References

Oreström, Bengt (ed.) A Corpus of Shetland English, Almqvist & Wiksell International, Stockholm 1985.

Barnes, Michael P. Jakob Jakobsen and the Norn Language of Shetland in Doreen J. Waugh (ed.) Shetland’s Northern Links - Language and History.

Crystal, David Shetland Speech in The Cambridge Encyclopedia of the English Language, CUP 1995.

Stewart, John, Shetland Place Names, Shetland Library and Museum, Lerwick 1987.

Mather, J.Y and Speitel, H.H. (eds), The Linguistic Atlas of Scotland, vol 3, Croom Helm, 1986.

John M. Tait.

March 2000