Irish Phonology - Phonological Processes - Lengthening Before Fortis Sonorants

Lengthening Before Fortis Sonorants

Where reflexes of the Old Irish fortis sonorants appear in syllable-final position (in some cases, only in word-final position), they trigger a lengthening or diphthongization of the preceding vowel in most dialects of Irish (O'Rahilly 1932:49–52, Ó Siadhail & Wigger 1975:89–94, Ó Siadhail 1989:49–50, Carnie 2002). The details vary from dialect to dialect.

In Donegal and Mayo, lengthening is found only before rd, rl, rn, before rr (except when a vowel follows), and in a few words also before word-final ll (de Búrca 1958:132–34, Mhac an Fhailigh 1968:163–64, Evans 1969:127, Ó Baoill 1996:16), for example, barr /bˠaːɾˠ/ "top", ard /aːɾˠd̪ˠ/ "tall", orlach /ˈoːɾˠl̪ˠax/ "inch", tuirne /ˈt̪uːɾˠn̠ʲə/ "spinning wheel", thall /haːl̪ˠ/ "yonder"

In Connemara (de Bhaldraithe 1966:109–12), the Aran Islands (Finck 1899), and Munster (Breatnach 1947:142–44, Ó Cuív 1944:121–23), lengthening is found generally not only in the environments listed above, but also before nn (unless a vowel follows) and before m and ng at the end of a word. For example, the word poll "hole" is pronounced /pˠəul̪ˠ/ in all of these regions, while greim "grip" is pronounced /ɟɾʲiːmʲ/ in Connemara and Aran and /ɟɾʲəimʲ/ in Munster.

Because vowels behave differently before broad sonorants than before slender ones in many cases, and because there is generally no lengthening (except by analogy) when the sonorants are followed by a vowel, there is a variety of vowel alternations between different related word-forms. For example, in Dingle (Ó Sé 2000:40–42) ceann "head" is pronounced /cəun̪ˠ/ with a diphthong, but cinn (the genitive singular of the same word) is pronounced /ciːnʲ/ with a long vowel, while ceanna (the plural, meaning "heads") is pronounced /ˈcan̪ˠə/ with a short vowel.

This lengthening has received a number of different explanations within the context of theoretical phonology. All accounts agree that some property of the fortis sonorant is being transferred to the preceding vowel, but the details about what property that is vary from researcher to researcher. Ó Siadhail & Wigger (1975:89–90) (repeated in Ó Siadhail 1989:48–50) argue that the fortis sonorant is tense (a term only vaguely defined phonetically) and that this tenseness is transferred to the vowel, where it is realized phonetically as vowel length and/or diphthongization. Ní Chiosáin (1991:188–95) argues that the triggering consonant is underlying associated with a unit of syllable weight called a mora; this mora then shifts to the vowel, creating a long vowel or a diphthong. Carnie (2002) expands on that analysis to argue that the fortis sonorants have an advanced tongue root (that is, the bottom of the tongue is pushed upward during articulation of the consonant) and that diphthongization is an articulatory effect of this tongue movement.