Linguistic Change Observed: Three Types of Phonological Change in the Ulster-Scots Dialects*

Robert J. Gregg

Recent developments in generative phonological theory have led to an increased interest in the historical as well as the diatopic aspect of dialectology. The somewhat static systems evolved by the strict structuralists have given way to a more fluid concept, based on the discovering of rules which apply to sets and subsets of linguistic forms in given dialects, over specific areas, through limited periods of time, and in some specifiable order. This new approach enables us not only to formalize coherently the relationships of divergent forms within a dialect group but to approach a solution of the Saussurean dilemma by drawing clear lines between historical rules for which the underlying forms may be no longer retrievable and synchronic rules involving variations such as morphophonological alternations.

The material under discussion in this paper was collected during a detailed survey of the Ulster-Scots (US) dialects in Ulster, the northern province of Ireland, between 1960 and 1963. Because this was an extremely fine-mesh survey — it has been labelled 'microdialectology' — the data enable us to see at close range early evidence of new rules as well as the last vestiges of old ones.

The first rules to be dealt with concern the various reflexes of ME ọ̄ and probably go back to the later part of the 17th century. For the majority of the Ulster-Scots dialects the rule nowadays is that ME ọ̄ → /ɪ̈/. It was found however that in the Laggan district of East Donegal and in the Mid Ards Peninsula the rule is ME ọ̄ → /i/. The settlement history as well as the phonetics would suggest that this second rule represents the older situation, while the first is probably an innovation stemming ultimately from the prestigious dialect of the Edinburgh area. The vowel /ɪ̈/ thus comes to replace /i/ in certain marked lexical items, namely those that had ME ọ̄. In this way the statement I lost my good shoes above the school would be in the Laggan and the Mid Ards dialects:

[a ˈlo:st ma ˈgid ˈs’in ə ˈbin ðə ˈskil]

and I'll have your boots done by the afternoon would be:

[al ˈhe: jəɹ ˈbits ˈdin be ðə ˈɛ:ftəɹˈnin]

whereas with the newer rule applied elsewhere we find:

[a ˈlo:st ma ˈgɪ̈d ˈs’ɪ̈n ə ˈbɪ̈n ðə ˈskɪ̈l]

[al ˈhe: jəɹ ˈbɪ̈ts ˈdɪ̈n be ðə ˈɛ:ftəɹˈnɪ̈n]

The formula /i/ → /ɪ̈/ is therefore NOT a normal phonological change with gain or loss of certain features, but rather a replacive type of rule. The synchronic evidence from Southwest Scotland confirms this view.

Along the western fringe of the Laggan district in Donegal it was found by chance that [ba:k] (which would mean 'back' in the Antrim and Down dialects) means 'beam' i.e., balk. Further enquiry elicited that back was [ba:ḱ] and that similar contrasts existed between

[ta:k] talk [ta:ḱ] tack
[sta:k] stalk [sta:ḱ] stack
[ha:k] hawk [ha:ḱ] hack

Palatal vs. non-palatal opposition is, of course, alien to English, but is basic not only to the phonology but to the morphophonology of Irish Gaelic, the language that forms the boundary with Ulster-Scots in the district concerned. This very much restricted innovation arises obviously by interference from the substratum and/or adstratum Gaelic and suggests a rule by which some sub-phonemic feature of frontness and backness (the latter associated with the deleted /l/ or /w/-glide in earlier Ulster-Scots or Jacobean English) became formally as well as meaningfully opposed to each other in the above-mentioned pairs. This is therefore an innovative type of rule whose source is essentially outside the language, and which has arisen from the interaction of languages in contact.

A third type of phonological change in the Ulster-Scots dialects affects the suprasegmental feature of length. The vowel /o/, which is intrinsically long in all environments, has begun to undergo shortening in a very limited number of cases in the East Antrim dialect, so that contrasting pairs have arisen such as:

[po:k] 'poke' (v.) vs. [pok] 'poke' (n.)
[spo:k] 'spoke' (v.) vs. [spok] 'spoke' (n.)

This shortening rule seems to be confined to the area mentioned, and it is strictly incipient as it affects only the vowel /o/, only in the environment of a following /k/, and only in a very limited subset of lexical items, the two cited above and also /fok/ folk /smok/ smoke and /tśok/ choke. In all other cases the /o/ preserves its traditional long quantity.

On the other hand a long-established rule for the shortening of the Ulster-Scots vowel /i/ in specific contexts is to be found all over Ulster. By this rule /i/ remains long in absolute Auslaut and in the environment of following voiced fricatives (including /ɹ/), but otherwise is shortened, hence /lif/ leaf as against /li:v/ leave, etc. However, a few dialect speakers in the Mid Ards have failed to apply this shortening rule in two or three lexical items only. Instead of the regular Ulster-Scots /kik/ 'peep' /kɹip/ 'creep' and /ˈkɹipe/ 'three-legged stool' these Ards speakers have maintained the older long quantity:

/ki:k/ /kɹi:p/ /ˈkɹi:pe/

This area was settled from SW Scotland in the early 17th century and has retained an old-fashioned model of speech in general. It has been discovered that the same and similar items are also still be found in relic areas on the Scottish side. The general conclusion is that the long /i/ remains (in spite of the rule) in certain marked lexical forms — marked +hearth and home, or +affection towards children, or the like.

These changes in quantity thus show on the one hand functional motivation in the recent rule for the shortening of /o/ which yields at least two meaning contrasts, but in the other case they show how a very small sub-set of lexical items may resist the otherwise universal application of a shortening rule because of some special marking.



I wonder whether you have any sociolinguistic data (such as age of speakers, groups of speakers who prefer certain types) to go together with the linguistic data.


Yes. The speakers who have not adopted the shortening rule for the three items [ki:k], [kɹi:p] and [ˈkɹi:pe] all belong to the oldest generation.

* Originally published in Rigault, A., and R. Charbonneau (eds.). Proceedings of the VIIth International Congress of Phonetic Sciences (The Hague: Mouton, 1972), 722-724.