O.Ir. cin, ‘guilt’, M.Ir. cintach, ‘guilty’ and I am inclined to think that the ‑i꞉ in gα꞉ri꞉, ‘garden, small enclosure’, M.Ir. garda and o̤mwi꞉, ‘many a’, O.Ir. immda, is also due to analogy; for the latter form cp. Derry People 24 x ’03 p. 3—siomaidh sgéul atá innisiste fa daobh dó. Henebry (p. 65) states that “gh broad or slender after l, n, r contracts w or y with the svar. thrown out by the liquid and becomes ū or ī” and quotes as instances feadghaile, murrghach. Pedersen (p. 15) says of Mod.Ir. gardha, “the word is now pronounced garī on Arran with a regular change of dh > j, development of svarabhakti vowel and change of əjə > ī; Scotch garradh”. Pedersen unfortunately fails to give us any further instances of the change of non-palatal dh > j and this sound-law has certainly not operated in Donegal, where Mod.Ir. dh after r, m disappears, as far as can be seen[A 1]. Thus M.Ir. gruamda appears as gruəmə; α NʹiəLəs, M.Ir. Mac Niallguis (Fergus and Oengus unfortunately appear as fʹαrəgəs and N⅄̃꞉s); *mórdhachas, cp. Di. mórdhacht, gives mo̤Rαχəs with assimilation of rd > R and shortening of the preceding vowel; fαurə, ‘eclipse’ is obscure, but it may be mentioned here as it represents urdhubhadh. Further in a number of trisyllabic adjectives in ‑rdha the result is ‑rə, e.g. kʹαχərə, ‘miserly’, Meyer cecharda; dαnərə, ‘cruel’, Di. danardha; similarly kɔrpərə (Meyer corporda), mαsərə, mwiNʹtʹərə, ʃαskərə.
§ 122. However in the case of O.Ir. palatal g after lʹ, rʹ the svarabhakti i + j + vowel invariably gives i꞉, e.g. dælʹi꞉s, ‘difficulty’, Keating doilgheas; Nʹiən, ‘daughter’ < inʹijən (the loss of the initial i is due to the word being frequently used proclitically) O.Ir. ingen. In a few cases a post-consonantic palatal ch is treated as if it were g (Mod.Ir. gh), only the preceding consonant must be voiceless, e.g. fwætʹi꞉s, ‘timidity’, M.Ir. faitches; tɔ꞉r̥i꞉s, ‘number at birth, parturition’ (ro̤g ʃi꞉ tʹrʹu꞉r ə jɛəN tɔ꞉r̥i꞉s, ‘she had triplets’), Di. toircheas, M.Ir. torrchius.
§ 123. By shifting of stress fuirʹ, ‘got’, O.Ir. fúair, becomes fwi꞉rʹ. smwi꞉tʹuw, ‘to think’, represents M.Ir. smúainim, cp. § 443. For bwi꞉rʹuw, ‘trouble’, see § 66. For i꞉, iə < O.Ir. ua see forms with y §§ 66, 67.
§ 124. But one of the most frequent sources of i꞉ is O.Ir. ái, ói, now written aoi, e.g. fwi꞉ʃuw, ‘improvement’, Di. faoiseamh
- ↑ αli꞉nʹ, ‘art’, M.Ir. elathain, eladain (dat.), has arisen by way of αləꬶinʹ > αləinʹ.
