Interface

Orthography as Interface

Although the Sjætlan language is presented on this website as a conlang - constructed language - as it might exist in a parallel universe, it is in most respects the same as the real language traditionally spoken in Shetland, and as recorded in various dictionaries and Grammar and Usage of the Shetland Dialect by Graham and Robertson. The main difference - apart from the use of the deprecated word Shetlandic - is the spelling, or orthography.

Orthography, rather than an army and navy as is often claimed, is what distinguishes a language from a dialect in popular perception, and enables its speakers to become literate. The main turning point in the emancipation of the Faroese language is concisely expressed in the Faroese-English dialogue in W.B. Lockwood, An Introduction to Modern Faroese:

It was Hammershaimb, a native of Sandavágur, who in 1854 published the first grammar of Faroese and introduced the modern orthography. From that time on Faroese began to challenge the supremacy of Danish.

In other words, the creation of an orthography is a - perhaps the - essential step in the emancipation of a marginalised language. Orthography is an interface between the written and spoken language. In the modern world, unless a language has a written interface which is usable by its speakers, its speakers are functionally illiterate in their own language. And since they are constantly exposed to written as well as spoken influence from the dominant language, that language first influences the marginalised one, and then - often in the space of a generation or two - replaces it entirely. Naturally then, in a parallel universe in which Shetlandic is taken seriously, the first step would be to create a viable orthography.

In creating their orthography, the Faroese largely followed the example and conventions of the already established Icelandic. Since Shetlandic is predominently an Anglic rather than a Nordic tongue - albeit with a Norse substratum - and closely related to Scots and English, it would at first sight seem more sensible to follow English spelling conventions. However, since English spelling is notoriously inconsistent and Scots has no standard spelling, it is difficult to represent the Shetlandic sound system in this way without giving the impression of a series of random exceptions to English pronunciation. That representation suits the prevailing dialect narrative in Shetland, but militates against representing the sound system of Shetlandic as an entity in its own right, which is the approach in the parallel Sjætland.

Most languages (English is a notable exception) try to maintain a close correspondence between writing and pronunciation. In the case of Shetlandic, the first step is to identify the important phonemes, or sound distinctions which can carry meaning. In other words, the approach begins from Shetlandic itself, rather than from how it happens to differ from standard English.

The consonants of Shetlandic are fairly easily identified. The vowels, on the other hand, are regularly misunderstood and misrepresented owing to the inconsistencies of dialect spellings. For example, the following words all rhyme in Shetlandic, and logically should be spelt with the same vowel spelling, as they are in the Sjætlan orthography below.

læk - like

dæk - dyke

sæk - such

pæk - peck (v)

However, læk is often written in dialect spelling as lik - by simply missing out the final -e of English ‘like’ - and sæk often as sic, in imitation of Scots, which is presumably in turn an imitation of Latin spelling. But in Shetlandic pronunciation, I and Æ are different phonemes, which are pronounced differently and can distinguish the meanings of words, e.g. lik (lick) from læk (like), pik (knock) from pæk (peck) and pit (put) from pæt (peat.) But because English has no sound or spelling corresponding to the Æ phoneme in Shetlandic, it tends to be confused with I - from which it is distinct and with which it does not rhyme - and spelt inconsistently. In the Sjætlan orthography, it is consistently spelt with the Æ ligature, as in læk and pæt above.

Although they are regularly misrepresented by recent postmodernist linguists as well as in dialect spellings, the underlying vowels of Shetlandic were actually well understood by linguists of former generations, such as J.C. Catford, J.Y. Mather, and the Linguistic Survey of Scotland. They consist of the twelve phonemes and six diphthongs of traditional Scots, as identified by D. Abercrombie and defined by A.J. Aitken, plus a further three long vowels which occur mainly in words deriving from Old Norse - these were not recognised by Catford and Mather, although two of them were recorded by the Linguistic Survey of Scotland. (Note that this does not mean that only three of the phonemes of Shetlandic come from Old Norse and the rest from Scots. The majority of the Norn and Scots phonemes coincide, and so most of the Norn ones are already included in the Scots inventory by default.)

The approach of the Sjætlan orthography is to assign a practicable minimum of graphemes - letters - to each underlying Shetlandic sound, or phoneme. This could have been done in various ways, but because of the convenience of the notation and the historical relationship to Old Norse, an orthography based on Old Icelandic has been used.

There are certain aspects of the Shetlandic vowel system which need to be recognised at the outset.

1. Vowel Length.

In Shetlandic, some vowels are long and some short. The ‘ee’ sound in need (English need) is short, but in need meaning strain or sprain, it is long. The ‘oo’ sound is short in moorit, a shade of brown, but long in moorie, a blizzard.

Some distinctions can be expressed by approximations to English spelling, such as got as opposed to goat or baet (beat) as opposed to bait. But even these tend to throw up inconsistencies in dialect spelling, owing to the fact that only perceived differences from English are normally spelt. So the word rate rhymes with baet as above, not with hate, but it may still be spelt as rate, perhaps because there is no EA in the English word to reverse into AE, or because the difference from English is not felt to be sufficient to require a different spelling. This leads to rate and hate being spelt alike even though they do not rhyme, and baet and rate being spelt differently, even though they do rhyme. Also, there is no obvious way in English to distinguish e.g. long and short ‘ee’, ‘oo’ or ‘ø’ sounds, and this leads to essential features of Shetlandic phonology being obscured by the shortcomings of English spelling.

There are various reasons for the differences in vowel length, including differences inherited from the phonemes of Scots and/or English (e.g. baet - from Scottish English beat - as opposed to bait) differences owing to extant or historical following consonants (e.g. blyde - glad - from an original blythe, where the vowel sound is long before a voiced fricative according to the Scottish Vowel Length rule) and preservation of Old Norse vowel length inherited from Norn (e.g. need - sprain, strain - with a long vowel from Old Norse hníta, as opposed to the short vowel in English need.) There is no need for spelling to distinguish between these historical origins, however - only to show which vowels are short and which are long, wherever they came from.

In the Sjætland of a parallel universe, Shetlandic is treated as an entity in its own right, and each vowel phoneme is differentiated in spelling, as is the case with most languages. So using Y for the ‘ee’ sound - as in modern Icelandic - U for the ‘oo’ sound - as in most languages, including the Nordic ones - and the acute diacritic (accent) for long vowels, as in Old Icelandic, we have the following types of spellings:

bak - back

bák - balk, beam

hæt - heat

hǽt - hate

spret - unravel (E has no long form)

hip - hip, omit (I has no long form)

nyd - need

nýd - sprain, strain

rǫt - rot (Ǫ has no long form)

bot - living room

bót - boat

luk - look

rúg - heap

bøl - resting place

brǿl - bellow

In contrast to dialect spellings, which represent Shetlandic sounds as they are perceived from an Anglocentric viewpoint, these spellings represent the sound distinctions which are made by native speakers.

2. Vowel Mutation

One of the reasons for the confusion between I and Æ in Shetland dialect spelling is that before voiceless consonants such as P, T and K, the Shetlandic vowel Æ sounds like the English vowel I. So the vowel in sæt (seat) sounds like the vowel in English sit. But in Shetlandic, the vowel in sit sounds different from this - more like the final E vowel in the English word the. Sæt rhymes with words like pæt (peat), hæt (heat) and læt (late), whereas sit rhymes with words like bit (bit), pit (put), and fit (foot). This is owing to a little understood characteristic of Shetlandic - not evident in English, Scots, or the Nordic languages, and ignored in most recent linguistic descriptions (1) - whereby several vowels have different pronunciations before voiceless (e.g. P, T, K) and voiced (e.g. B, D, G) consonants. This characteristic is here termed Soft Mutation. (2)

Unlike vowel length, soft mutation is non-phonological. That is, it does not distinguish one word from another by sound, and therefore does not need to be represented in spelling. However, it can lead to confusion between Shetlandic phonemes - in dialect spelling, not in speech - and so needs to be taken into account in creating an orthography. It is also useful to have a way of representing it in word lists, and the diacritic commonly referred to as ‘Umlaut’ from its use in German is used.

Soft mutation can be illustrated by the difference in the sound of A in pairs of words such as bat and bad. In most if not all Shetland dialects, the vowel in bat is similar to that in Scots, ordinary Scottish English (as opposed to affected ‘Morningside’or ‘Kelvinside’ pronunciations) or Swedish, whereas the vowel in bad is similar to that in ‘posh’ RP English or Danish. The latter sound - which typically occurs before voiced consonants such as B, D and G as opposed to unvoiced ones such as P, T and K - can be shown at need by the ‘Umlaut’ diacritic - bäd. This is not necessary in spelling, because native speakers pronounce it this way automatically according to their regional accents, where it varies considerably (for example, A is not pronounced as Ä before G in West Side accents, as it is in most other regions) but it can be useful in word lists.

3. Dialect Variation

The reason dialect variation has to be taken account of in the creation of a Shetlandic orthography is that the intention of an orthography is not to create a standard spoken Shetlandic, but a spelling which is equally adaptable to all dialects. The variation in dialect pronunciations is often cited as an example of why a standard spelling is impossible. But this is almost the opposite of the case, and the constant reiteration of this mantra has nothing to do with the ontology of the language or its dialects, and everything to do with the determination of the Shetland and Scottish establishments - administrative, intellectual, literary and media - that Shetlandic and Scots should be treated as Cinderella tongues clad in rags and tatters which are - in the words of Shetland archivist Brian Smith - ‘unfitted’ to attend the ball in company with the unquestioned and sacrosanct inherited superiority of standard English.

Dialect variation in any language is a result of a family tree of sound changes, whereby an original sound comes to be pronounced differently in different regions. An obvious Shetland example is the phoneme Á, which is pronounced ‘ah’ in some dialects, and ‘aw’ in others, so that you get ‘taatie’ (potato) in most Mainland and many other dialects, and ‘tautie’ in the North Isles and the Sooth End (Dunrossness.) A rather less obvious example is the way in which the original Æ phoneme is pronounced differently before voiced consonants such as B, D, G, N, M and L. (This is one example of the effects of Soft Mutation.) So the word stæn is pronounced ‘stane’ in Burra, ‘sten/steyn’ in Central Mainland type dialects, and ‘steen’ in North Isles type dialects (where it hasn’t been completely replaced by the alternative form ‘ston,’ derived from English stone.) In dialect spelling, each region spells their own pronunciation as the nearest approximation to standard English, and the underlying identity of the vowel, and the language, is obscured.

The technical term for such underlying vowels - historical vowels which have come to be pronounced differently in different regions - is ‘diaphonemes.’ The approach of the Sjætlan orthography is to represent each underlying vowel by a single spelling - e.g. stæn, táti - which can be pronounced differently in different regions.

4. Consonants

Consonants present a different problem from vowels, in that they are mostly easily identified, but it is not always so easy to represent them accurately using an orthography based on Old Icelandic, and they are most satisfactorily dealt with on a case by case basis.

Double consonants are unnecessary in the Sjætlan orthography. In English, these distinguish the preceding vowels in words like hatter and hater. In the Scandinavian languages, they show that the preceding vowel is short. However, because the Sjætlan orthography shows both the identity and length of vowels - the A vowel is never pronounced as in hater - double consonants aren't needed for either purpose. The words hatter and hater are spelt hater and hǽter.

The Function of an Interface

The above descriptions may seem to be very complicated. But that is precisely why an orthography - an interface - is necessary. Cars and televisions are complicated internally, but with an interface - a steering wheel and pedals, or a remote control - you can use them without understanding their internal functions. An orthography is only complicated at the creation stage. Once it’s made, it can be used by anybody, and the complications pass unnoticed. A generation of children in the parallel Sjætland, who are taught the Sjætlan orthography, find no problem in pronouncing Á as ‘ah’ or ‘aw’ according to their own dialect, and becoming literate in their own language. By contrast, in the real world, the dialect narrative, by depriving Shetlandic of a systemic representation of its own characteristics and thus keeping the complications on the surface rather than under the bonnet, ensured that that could never happen.

(1) Apart, as far as I know, from those of Gunnel Melchers and Peter Sundkvist, who describes it as Shetland Vowel Mutation. More recent descriptions make no mention of it.

(2) Some Characteristics of the Shetlandic Vowel System.

References

W.B. Lockwood, An Introduction to Modern Faroese. Føroya Skúlabókagrunnur, Tórshavn 1977.

Catford, J.C. Shetland Dialect, in Shetland Folk Book vol. 3.

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

Mather, J.Y. 1964. 'Dialect Research in Orkney and Shetland after Jakobsen'. in Froðskaparrit 13 Bók 1, Mentunargrunnur Føroya Løgtings, Tórshavn.

Tait, J.M. 'Some Characteristics of the Shetlandic Vowel System,' Scottish Language 19, 2000.