The Feature 'Dentality' in Ulster-Scots Dialects and its Role as a Sociolinguistic Marker*

Robert J. Gregg

It is part of our normal basis of articulation that, in spoken versions of Standard English, we have a large set of coronal consonants — phonetically specified as alveolars, viz., /t, d, n, s, z, l/ — which corresponds to the set of dentals found in many other languages (See Table I, Set A).(1) We have in fact only two coronal consonants — /θ/ and /ð/ — that can accurately be labelled dental or interdental (See Table I, Set B). These two dental consonants are phonologically contrastive with the corresponding alveolar continuants, /s/ and /z/, and it should be observed that, when either of the latter occurs with one of the dentals, assimilation does not take place. On the other hand, this small dental set seems to have some kind of hierarchical priority over the other alveolars in the matter of assimilation. Members of the alveolar set — apart from /s/ and /z/ — adjust their articulatory position to a following /θ/ or /ð/, in other words, by a kind of anticipatory coarticulation they acquire a feature that might be called 'dentality', marked in Table I, Set B, by the conventional I.P.A. subscript symbol /  ̪  /.


The assimilation works regressively across morpheme boundaries, within a word form, e.g.

eιt eight eιt̪θ eighth
waιd wide wιd̪θ width
ten ten ten̪θ tenth
wel well wel̪θ wealth

This assimilatory rule also operates regressively (as far back as two segments, if the first is /n/ or /l/) across word boundaries as in:

kʌt̪ ðιs cut this haιd̪ ðæt hide that
bιˈgιn̪ ðιs begin this fιl̪ ðæt fill that
ɹen̪t̪ ðιs rent this bιl̪d̪ ðæt build that

It should further be noted that, for many speakers, the flapped r [ɾ] — surviving from an earlier phase of English — belongs in Set B, as it is maintained in the environment of a preceding /θ/, whereas preceding alveolars call for the now generalized approximant r [ɹ] thus:

tɹʌst trust θɾʌst thrust

In addition to these dental assimilations in contact with /θ/ and /ð/ which are general among native speakers of English, there are several types of dialectal English that have a further domain in which the dentality feature crops up, namely in the environment of a following /r/. The latter dialects include those spoken in

(1) northwestern England

(2) southwestern Scotland (2)

(3) the whole of Ireland

Since the dialects of English spoken in Ireland derived mainly from (1) and (2), the latter regions may be the source of this extension of the dentality phenomenon. The actual phonetic realization of the dentals (or more accurately the interdentals), however, is identical with the corresponding set of substratum or adstratum Irish Gaelic 'broad' (i.e., non-palatalized) dental consonants, frequently transcribed by Celticists as /T, D, N, L/. These Irish sounds, of course, do not arise from the assimilatory process. They are independent functional units, in phonological contrast with corresponding 'slender' (i.e., palatalized) consonants: / tˊ, dˊ, etc./. It is noteworthy, however, that in loan words from English, Irish Gaelic speakers (e.g., in Donegal) use an alveolar /t/ initially in words like /tïḱedˊ / (from English ticket), although alveolar /t/ is foreign to the Irish consonant system. On the other hand they use their 'broad' /T/ (i.e., t̪) in contact with /r/ as in the loanword /ˈt̪ɾəisïḱəlˊ /, which is thus identical with the neighbouring Hiberno-English and Ulster-Scots forms.(3)

Apart from links with the northern English dialects which may have provided the dentality pattern, it thus looks as if there is a very strong connection with the indigenous Irish Gaelic, whose preference for the 'broad' dentals in the proximity of /r/ seems to have influenced all the incoming English and Scottish dialects.

The occurrence of the dentality feature in the northwest of England was noted almost a century ago by Alexander J. Ellis.(4) It was observed also by many other writers in their descriptions of individual dialects in the northwest region,(5) as well as by Joseph Wright in his English Dialect Grammar (1905). Eduard Kolb's Phonological Atlas of the Northern Region (1966), based on Harold Orton's dialect survey material, shows that this dentality still persists in NW England down to the present time.(6)

In the earlier dialects of this region that spread to Ireland in the 16th and 17th centuries the occurrence of dentality was probably much more widespread if not universal, for it crops up still in all Hiberno-English dialects, including the Ulster-Scots in the province of Ulster.(7) Thus we find everywhere in Ireland — in the environment of a following /r/ — forms such as:

t̪ɾi: tree ˈlad̪ər ladder

For many speakers in the southern Irish provinces, however, the situation is complicated by the fact that the English continuant θ → t̪ (i.e., loses the continuant feature) and ð → d̪, which means that the two words quoted above are phonologically identical with three /t̪ɾi:/ and lather /ˈlad̪ər/. In contexts other than that of a following /r/ the dental stops, /t/ & /d/, however, as in /t̪ιn/ thin and /d̪ɛn/ then, are in general phonologically opposed to the simple alveolars as in /tιn/ tin and /dɛn/ den.

In the dialects of Ulster, whether they originated in England or in Scotland, the dentalized consonants have the same distribution as those of NW England, while /θ/ and /ð/ are generally given their normal articulation as fricatives.

The main purpose of this present study is, first, to explicate linguistically the phonological rules that govern dentalization in two Ulster dialects, viz., Ulster-Scots Rural (U-S R) and Ulster-Scots Urban Non-standard (U-S U N), and to discuss the sociolinguistic implications of its use in these dialects vis-à-vis Ulster-Scots Urban Standard (U-S U S) speech, in which this type of dentality never occurs.

The most general rule can be simply stated as follows:

i.e., immediately before /r/ (or /ə + r/), the alveolars /t, d, n/ become /t̪, d̪, n̪/ respectively.

At this point a corollary needs to be stated thus:

and an 'elsewhere' rule:

i.e., the unspecified underlying form of /r/ is realized at the surface level as a flap — [ɾ] — when preceded by a dentalized consonant and followed by a vowel, whether or not a schwa intervenes between the dentalized consonant and the /r/. In other environments, namely when final or before a consonant, the /r/ is realized as a continuant — [ɹ], for example:

Exx 1
t̪ri: tree d̪raι dry hɛn̪re Henry
Exx 2
batəɾe battery ʃʌd̪əɾən shuddering skʌn̪əɾən sickening
Exx 3
bʌt̪əɹ butter ʃʌd̪əɹd shuddered skʌn̪əɹ sicken

A special rule is needed to cover the case of the lateral:

i.e., alveolar l becomes l̪ immediately before another dentalized consonant (t̪ or d̪) as in:

Exx 4 ˈʃɛl̪t̪əɹ shelter ˈbo:l̪d̪əɹ boulder

Note in relation to Rule 3 that there is an optional rule as follows:

e.g., ˈʃʌd̪əd shuddered is in free variation with ˈʃʌd̪əɹd.

Unlike the general rules for assimilation in standard English outlined at the beginning of this discussion, the special Ulster dialect rules for dentalization do not operate across a perceived morpheme boundary. This difference leads to internal and external dialectal contrasts such as the following:

U-S Urban
U-S Urban
U-S Rural
flatter (< flat) t t t
flatter (vb.) t
better (< bet) t t t
better (< good) t(9)
assist 'er t t t
a sister t
starter (< start) t t t
barter t
quilter (< quilt) lt lt lt
filter lt l̪t̪ l̪t̪
sprinter (< sprint) nt nt nt
splinter nt n̪t̪ n̪t̪
redder (< red) d d d
rudder d
rider (< ride) d d d
spider d
recorder (< record, vb.) d d d
order d
builder (< build) ld ld ld
bewilder ld l̪d̪ l̪d̪
spender (< spend) nd nd nd
fender nd n̪d̪ nd̪

In conclusion, an attempt will be made to summarize the social distribution within the Ulster dialect system, and the sociolinguistic implications of the dentality feature described and analysed above. All forms of the standard language and all dialects share the general assimilatory processes in contact with /θ/ and /ð/ outlined at the beginning of this study. No contrasts emerge in this area, and thus speakers are in general quite unaware of the dentalization involved. On the other hand, only the rural and non-standard urban dialects share the dentality feature Rules 1-4 above, which consequently form a contrast with the standard urban speech.

As a result of this distribution, speakers of all these Ulster dialects tend to be — consciously or subconsciously — aware of the dentality involved with these rules. For educated urban speakers it is a clear-cut marker of non-standard or rural dialect and hence stigmatized. It belongs to what they would call 'talking broad'. With non-standard urban speakers, conflict may arise, for, as part of the normal education process, they are under pressure — among other things — to give up their dentality. If they are interested in upward social mobility it is mandatory. For girls this seems easier to accept than for boys, who may feel that the non-dental forms sound affected or snobbish.

The country dweller who speaks a broad rural dialect has a different problem. As part of his education he has to become bidialectal by learning the standard language which, of course, has the non-dental forms that he will be expected to acquire. To him it is perhaps clearest that each style of speech has its proper place. What he scorns is the pretentious rural speaker who mincingly tries to imitate standard urban speech but keeps mixing it up with his broad dialect. This is called speaking 'silk and drugget'.(10) Equally to be scorned — he thinks — is the urban speaker who mixes styles when he attempts to use dialect words, pronouncing [ˈskʌnəɹ] instead of [ˈskʌn̪əɹ] 'sicken', [ˈgʌldəɹ] instead of [ˈgʌl̪d̪əɹ] 'an incoherent shout' or [tɹɔ:x] instead of [t̪ɾɔ:x] 'trough'. Only the skilled and sensitive bidialectal speaker will know how to switch the dentality rules so that his speech is suitable to all social environments, and he avoids the opprobrious possibility that every time he opens his mouth some other Ulsterman will despise him.


* Paper delivered at the Linguistic Association of Great Britain Spring Meeting at The Queen's University of Belfast, 10-12 April 1989.
(1)The term basis of articulation is a translation of the German term Artikulationsbasis, used in the late 19th century to refer to the concept (elaborated by phoneticians such as Eduard Sievers, Otto Jespersen, Henry Sweet, etc.) that for each language variety there is a characteristic articulatory configuration of the speech organs:

For English, styled as cacuminal, i.e., with the tip of the tongue raised towards the alveoli and with the upper tongue surface concave

For French and German, dorsal, with the tongue tip in contact with the lower front teeth and the upper surface of the tongue convex

For Irish Gaelic, rim articulation, with the tongue flat and spread laterally.

Nowadays, with the improved facilities provided by the modern electronic equipment of our phonetics laboratories, this concept has been refined and extended, and the currently preferred term is articulatory setting. [See Beatrice Honikman (1964) and John Laver (1984)].
(2)The advancement of alveolars to a dental or interdental articulation in the environment of a following /(ə)r/ is not so well documented for Scotland as it is for Ireland and NW England. It is, however, said to be characteristic of some southwestern Scottish dialects (the main source of Ulster-Scots speech). See T. Wright (1929). The dentality phenomenon is also described by William Grant (1914), who attributes it to Gaelic speakers (Note that Ayrshire and Galloway still had remnants of Gaelic speech in the 17th century or later). The main purpose of Grant's book is to 'correct the faults' of Scottish teachers in training who were expected to acquire a 'standard' pronunciation.
(3)The Donegal Irish examples were given to the writer by Donall Ó Baoill in a personal communication. See also D. Ó Baoill (1975), 19 ff., for contrastive examples in Gaelic. A detailed description of the articulation and distribution of the Irish dentals in an Inishowen (NE Donegal) dialect may be found in Emrys Evans (1970).
(4)Ellis, Alexander J., On Early English Pronunciation (London, 1889), 542-3.
(5)CUMBERLAND: Penrith (P. H. Reaney, 1927); Lorton (B. Brilioth, 1913)
WESTMORLAND: Kendal (T. O. Hirst, 1906)
LANCASHIRE: Adlington (A. Hargreaves, 1904)
YORKSHIRE: Hackness (G. C. Cowling, 1915); Nidderdale (M. L. Annakin, 1922)
Note that the county boundaries have recently been re-drawn to create the new entity named Cumbria.
(6)Kolb, Eduard, Phonological Atlas of the Northern Region (Bern, 1966). A review article on Kolb's Atlas (G. B. Adams, 1967) draws attention to the dentality feature common to Ulster dialects and those, particularly of the western part of the region, mapped by Kolb.
(7)See Henry, P. L. (1957), 54 ff; Gregg, R. J. (1958), 401 & 405; Gregg, R. J. (1964), 184 ff.; Todd, L. (1970), 22, 30 & 31; Nally, E. V. (1971), 36-38; Sullivan J. P. (1976), 61-72.
(8)The feature 'continuant' is questionable for the lateral [see Chomsky, Noam, & Morris Halle, Sound Pattern of English (1968), 318]. See also Wells, J. C., 'A Scots Diphthong and the Feature "Continuant"' in JIPA 1:1 (1971).
(9)In this case the form better, being irregularly derived from good, is perceived by rural and non-standard urban speakers as a solid morpheme.
(10)The word drugget (pronounced /d̪ɾʌgət/, of course, by the rural US speaker) refers to a coarse, woollen material.