Ganfer 6

6. Orthography and Diaphonology.

One thing not mentioned in the video is the importance of recognising the underlying Shetlandic phonology for orthographic purposes. 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. It is also the reason why orthography is targeted as a weak point by anyone wishing to scupper any attempts to emancipate a language - Scots or Shetlandic - because speakers can easily be persuaded that a standard spelling would be detrimental to local pronunciations.

To attempt to solve this problem, it is necessary to have a systemic, diaphonemic, cross-dialect orthography to represent the underlying phonological system, which can then be pronounced regularly according to regional accents. The two main features which have to be taken account of in the case of Shetlandic are vowel length and soft mutation. Both are recognised by Melchers & Sundkvist (Lesser-Known Varieties of English, Sheier, Trudgill et al. 2010, p.24-25) with reference to my ‘Characteristics’ article.

‘There are three potential additional long items /i:, ø:, u:/. The most extensive account of these rather marginal items may be found in Tait (2000).’

‘Some of the most salient characteristics of the Shetland vowel system derive from a process of vowel mutation ... The most extensive treatment of Shetland vowel mutation is Tait (2000).’

One totally overlooked Norse characteristic of Shetland phonology is the long /i:/, /u:/ and /ø:/ vowels - recorded by Jakobsen as well as (apart from /u:/) the LAS - in words like nýb (nod off, ON hnípa) rúg (heap, ON hrúga) and rǿð (speak nonsense, ON hrjóða - the ð in my Sjætlan spelling is not purely etymological, but a diaphonemic grapheme to accomodate the regional pronunciations of ‘d’ and ‘ð’.) Final and intervocalic ON ‘p, t, k’ became ‘b, d, g’ in Norn, which is why these vowels don’t occur before voiceless consonants, and are missed in the default analyses of syllables closed by /t/.

A likely reason why this is overlooked in dialect perception is that there is no obvious way to represent it in English-derived spelling, and it is only marginally contrastive, affecting only a few words. However, as these long vowels also occur in non-Norse vocabulary for different historical reasons (usually extant or fossilised Scottish Vowel Length Rule) e.g. pǿr (poor, emaciated) rǿs (praise) prýv (taste) - all phonologically Scots in spite of obvious resemblances to Nordic cognates - stúr (dust) nýr (both Scots/English ‘near’ and Norn ‘kidney’) the entire phonology of Shetlandic can be represented by a single, coherent orthography which marks all long vowels irrespective of etymology.

I invented the term ‘soft mutation’ to describe a phenomenon which affects several vowels, as, without this, there is no way to refer to it as a whole without constant terminological complications.

Generally, the ‘soft’ allophones are further front and/or closer than the ‘hard’ allophones. The effect is not unlike that of i-mutation in Old Norse, and it is suggested - or hinted at - by Catford that it is owing to palatalisation, or former palatalisation, of the following consonants. It does not usually lead to phonemic mergers, as it does with the Æ phoneme in some dialect groups, and so does not usually need to be represented in spelling. But it has to be understood in order to avoid both conflation and division of phonemes.

I use the trema or ‘Umlaut’ sign to indicate the soft allophones at need - e.g. bäd (bad) hæ̈d (head) bëd (bed) sïd (corn husk) rǫ̈d (road) lög (ear) - and the Hungarian double acute for long soft vowels, e.g. la̋v (hover) mæ̋ð (maggot) kla̋g (cackle.) But since these mutations are determined by the following consonant, and native speakers pronounce them automatically according to their regional accent, this is not part of the general Sjætlan orthography. Again, the pronunciations in my accent are available in my ‘Short Guide’ PDF and video.

The word ‘kæn’ (know) demonstrates the effect of soft mutation on the historic Æ phoneme, and why this has to be understood for orthographic purposes. The ‘hard’ form of this phoneme in words like ‘læt’ (late) and ‘Sjætlan’ - mostly before voiceless consonants - has more or less the same pronunciation everywhere in Shetland. The ‘soft’ mutation - mostly before voiced consonants - merges with the E /ɛ/ phoneme in one group of dialects (e.g. Central Mainland) and with the Y /i/ ‘ee’ phoneme in another group (e.g. North Isles.) In other accents, such as my own, it retains a distinct pronunciation. So ‘kæn’ is pronounced variously as ‘kane’ [e], ‘keyn’ ([ɛɪ] being the soft pronunciation of the E phoneme with which it merges) and ‘keen.’ [i]. Again, recognition of this goes back to Catford’s article of 1957.

The same regional correspondences before voiced consonants occur in all (or almost all - I can only think of one or two exceptions) words which have this historical phoneme - hæd (head), æg (egg), gæng (go) næm (name), etc. In the parallel Sjætland, the Æ spelling in these is automatically read and pronounced according to regional accent.

Failure to recognise soft mutation can lead to a number of errors in phoneme identification, and consequently spelling, which feed into the narrative that Shetland dialect consists of irresolvable dialect variations. A case in point is the word ‘hæm’ - home. The Central Mainland pronunciation of this is ‘heym’ [hɛɪm] which is the soft allophone of the E phoneme with which it merges there, and it is sometimes spelt ‘hem’ in dialect writing. This could be related to Swedish ‘hem’ and the spelling justified or preferred on that basis. Alternatively, the diphthongal pronunciation (sometimes spelt as ‘heym’) could be taken as a justification for adopting the Icelandic spelling ‘heim’ - the pronunciation is quite similar.

But both of these are essentially regional spellings based on a Central Mainland pronunciation. In the case of ‘hem,’ the word would rhyme with ‘hem’ (of a garment) ‘dem’ (them) and ‘gem’ in the Central Mainland, but not in the North Isles, Burra or the Ness. In the case of ‘ei’ - if the intention was to represent the diphthongisation - the diphthongal pronunciation is, again, characteristic of the Central Mainland. In the Shetland Dictionary, the spelling ‘hame’ is used presumably because of passing familiarity with Scots, but some similarly spelt words such as ‘hate,’ ‘bake’, and ‘frame’ have a different phoneme in Shetlandic - a distinction which does not occur in most dialects of Scots.

This is a problem with the Shetland Dictionary, which reflects a Central Mainland pronunciation where soft Æ has merged with soft E as [ɛɪ]. Words with the soft allophone of the Æ phoneme are sometimes spelt E (e.g. ‘lem’, crockery) sometimes A-E (‘hame,’ home) and sometimes AE (‘fael’, turf) while the AE spelling, as well as being used for the hard Æ sound (‘gaet’, path) is also sporadically used for a different, long phoneme (‘maed’, maggot.) Sjætlan orthography spellings of ‘læm,’ ‘læn’ and ‘fæl’ put them in one group all with the same regional variants, and the acute on ‘mǽð’ specifies the long phoneme. This shows how, even though the soft mutation is not shown in the main Sjætlan orthography, it is necessary to understand it in order to allocate graphemes to phonemes so that they can be consistently pronounced according to regional accents.

Lack of recognition of the Æ phoneme also leads to mistakes in the pronunciation section. It states ‘In Whalsay... The short e sound as in English ‘men’ becomes a diphthongised sound ‘eh-e’ as in ‘meh-en’, ‘steh-en’ and ‘teh-en’ for ‘men’, ‘stane’ and ‘ten.’ But while this is true for ‘men’ and ‘ten,’ in Whalsay ‘stane’ is pronounced ‘steen,’ as in the North Isles. It is only the E phoneme, not the Æ one, that has the diphthongised pronunciation. Even using the different spelling ‘stane,’ John Graham - or his advisor - has apparently simply assumed that this word will rhyme with the others because it does so in the Central Mainland.

Consideration of the regional variants of ‘hame’, ‘heym’ and ‘heem’ show that this word does not have the E phoneme in Shetland as a whole, but the soft allophone of the Æ vowel, and fits into the large group of words which have the same regional variants as the examples given above, including kæn (know) næm (name) bæn (bone) hæl (whole) hæd (head) etc. Vagaland - T.A. Roberston, poet and one of the authors of ‘Grammar and Usage of the Shetland Dialect’ - spelt it ‘haem,’ as I did in my previous English style spelling. In the Sjætlan spelling, I spell it ‘hæm.’ Both spellings are diaphonemic, intended to identify the phoneme and accomodate the regional variants.

This sort of regional variation is often cited as proving that a standard spelling is impossible, illustrating how everything varies, dictionaries inevitably get it wrong, and any attempt to sort it out would be a ‘nightmare.’ But it is in fact an example of how a single orthography, with accurately allocated diaphonemes, could represent various regional pronunciations, whereas a spelling based on a particular dialect - such as ‘ken,’ ‘keyn’ or ‘keen’ for kæn - is not logically adaptable to others. It is precisely the lack of such a user interface for the language which perpetuates the ‘nightmare’ narrative. It is because this practical ‘user interface’ approach to spelling is unacceptable in the real Shetland under the dialect narrative - even though the basic information goes back to Catford’s article from 1957 - that I have decided to write up my mother tongue as if it existed in a parallel universe, where it is treated as a language in its own right.