- Quiggin, E. C. 1909-1910. 'A Case of Stress Shifting in the Dialect of Tréguier', Zeitschrift für Celtische Philologie 7, 354-356.
Ce très court article relève des variations dans l'accentuation entre les dialectes du breton. Il suggère qu'en Tréguier, la négation ket est cliticisée sur le verbe tensé et compte comme syllabe finale pour l'accentuation.
"Perhaps the most interesting feature of the Breton dialects of the present day is the way in which the position of the stress varies in the different localities. Like Cornish and Welsh Breton shifted the accent during the middle period from the final (the primitive penult) to the penultimate syllable. The speech of Vannes has however retained the old stress down to the present day. In Haute-Cornouailles and Goelo on the other band the accent is placed on the ante-penult wherever possible (See F. Vallée, Leçons Elementaires de Grammaire Bretonne p. 10).
Traces of this are to be observed in the immediate neighbourhood of Treguier itself, any rate on the east side of the R. Jaudy. The dialects of Leon and Treguier have a strong expiratory accent on the penult, which reminds one not a little of Welsh. French as spoken locally is largely influenced by the stress system of the vernacular, thus one frequently hears mer'credi, petit [pœ́ti], donner [dó:ne].
Some time ago whilst copying two XVII Century canticles I was struck with the frequency with which the negative particle ket was joined on to a preceding verb form, e.g. comps guer er bet ne hallequet, querset na comps ne hallequet, car comps ne allequet. I do not know how frequent this orthographical peculiarity may be, as printed texts do not always follow the MSS. closely in such respects. Several instances however are to be found in the XVIII. Century Mystere de saint Crepin et de saint Crepinien published by V. Tourneur in Revue Celtique vol. XXV, pp. 299-343, e.g. l. 134 na gredan-quet', l. 141 n-en dint-quet, l. 226 na reon-quet, l. 250 na vet-quet, l. 334 ne allquet. From a Statement in the introduction the editor seems to consider these cases as mere orthographical mistakes. After spending some time however in the Treguier district I am inclined to regard these instances of seemingly careless copying as having a deeper significance.
Under certain well defined conditions the accent in the Treguier dialect is transferred from the penult to the final syllable. This occurs regularly in the case of a verb form followed by the enclitic negative particle ket, e.g. gwéled [gwə:lət], 'to see', but welan ket [we:lã́ kət], 'I do not see'; lar(et) ráfe an (d)en, 'One would say', but rafé ket, 'he would not do'. Similarly [ma mi:žé kət] 'if I had not', [ma nə ro:žen kət] 'if I had not given'. This holds good without exception and in consequence the complexion of some forms is liable to be completely changed.
Thus the present 1st sing, of rei, 'to give', is [róã] but with the negative one hears [rwã kət], similarly in the future [róĩ] but [rwĩ kət]. The pret. 3rd sing, is [róas] but [rwás kət]; ne roint ket appears as [rwánt'ət]. In all these cases the verb form and the negative particle kət are stressed as if they formed one word.
Precisely the same phenomenon is found when a numeral is followed by a frequently employed monosyllabic substantive such as bla, skoued, eur: e.g. náountek, 'nineteen', but naountég 'la, '19 years', pewár bla, '4 years'; ugent [ý:gen, y:gən] 'twenty', but [ygén la], '20 years'. In the latter case it may even happen that the indeterminate vowel ə receives the stress, for I have frequently heard [y:g'ən la].
Further examples hánter, 'half', but eun hanter eur, [oen anté:r ər], 'half an hour'; pemzek, 'fifteen', but [pemzé(k) te:], 'a fortnight'; da unnek eur, [dané:gər], 'at 11 o'clock'. Unnek alone is pronounced [ánek]. A Substantive thus standing enclitically after a numeral is liable to considerable reduction. For instance, skoued [skwet], 'a three franc piece', commonly becomes [skot] or [sko], e.g. [dáu skot], 'six francs', [pém sko war ný:gen], '75 francs'.
To my knowledge the only Breton texts in phonetic transcript are the song published by Loth in the Annales de Bretagne vol. XI, pp. 237 ff. and the fable published by J. Le Gall ib. vol. XII, pp. 423 ff., but neither of these throw any light on the point in question. Nor do the excellent grammars by Hingant, Vallee and Le Clerc notice the special rules relating to stress which obtain in Treguier. If I might judge from conversations with Bretons from districts other than the one with which I am acquainted, it would seem that the above mentioned cases of stress shifting are unknown in and round Morlaix, but that they occur south of Guingamp in parts of Cornouailles."