Irish Gaelic: Pronunciation of /ɾˠ/ (broad r)


Senior Member
Dear all,

I've just come across this very short thread, where Tadhg an Mhargaidh convincingly argues that both slender r /ɾʲ/ and broad r /ɾˠ/ have traditionally been (and therefore should be) pronounced "as an alveolar tap (with or without a fricative quality)." That's also what the notation itself suggests as the default, since the IPA symbol ‹ɾ› alone stands specifically for an alveolar tap. The current version of "Irish phonology" at Wikipedia makes no mention what-so-ever of the possibility that Irish /ɾʲ/ or /ɾˠ/ be realized as an alveolar approximant (be it retroflex [ɻ] or not [ɹ]), which could give the impression that realizing either of them as such is not (good) Irish at all.

What I gather from what I've heard and read so far is that slender (or palatalized) r /ɾʲ/ never is a [ɹ] (or [ɻ]) in "traditional" native accents (where it has a continuum of realizations, ranging from pure [ɾʲ] to more fricativized or spirantized versions, quite reminiscent of what happened to Proto-Slavic /ɾʲ/ in Polish, where it is now realized as /ʒ/, sometimes transcribed as /ʐ/, cf. rzeka), which is why I chose to ask about broad (or velarized) r /ɾˠ/ only. If I'm wrong in this assumption, please correct me.

Now, the question is, are there varieties of Irish where broad r has traditionally (let's say, for generations) been realized as an approximant? If there are, is the distribution complementary, i.e. does it depend on factors such as position (syllable onset, syllable coda, intervocalic position, syllable final position before another consonant, absolute final position…) or stress (stressed syllable, unstressed syllable)?

For those who like more elaborated initial posts :p, here comes a "discussion trigger." :) Let's start with an excerpt from a thread on

seano said:
(…) they don't roll the R, though in some Connaught dialects you might find the R has a "tap", like a Spanish single R (in other words, the tongue flicks against the hard palate behind the front teeth. In Munster, many speakers do an R which resembles a French R - it's back in the throat, almost like someone gargling cough medicine. And in Ulster, we usually have a slightly retroflex R - that means the tongue is bent upwards in the palate. The most extreme English dialect where there is a retroflex R is Indian English (think Apu in the Simpsons). The Ulster Irish retroflex is not as marked as this. If you have ever heard a West Country English speaker, that is something like the sound of an Ulster speaker doing a broad R. (As I said before, the slender R is somewhere between a Y and a Z!) (…)
  1. the "tap" is an alveolar tap [ɾˠ]
  2. the "gargling" sound must be either a uvular trill [ʀ] or a voiced uvular fricative [ʁ] (the description matches the trill better, if you ask me :D )
  3. "the slightly retroflex R" must be a [ɻ] (or a [ɹ] leaning towards a [ɻ])
(See here for an overview of r-sounds)

It looks like, from seano's perspective, the unmarked pronunciation of /ɾˠ/ is [ɹ]. Let's call it pronunciation 0. This pronunciation appears to have (gained), in the eyes of many, the status of thé "default" (unmarked) realization of /ɾˠ/. At least, judging by what we get to hear in numerous language courses and what we come across in some pronunciation guides (including this one found at But… isn't it diachronically just a widespread transfer from Hiberno English? How widespread and established is it?

I'm particularly interested in the history of Irish /ɾˠ/ and I'd really appreciate it if someone could provide academic sources that could shed some light on the topic. I haven't had much luck finding any, other than Hickey (2011: 376), where he explains how "[i]n Irish English during the 1990s (…) the traditional velarised [ɹ] came to be replaced by a retroflex [ɻ], first in the Dublin area among young females, then spreading out to the rest of the country (…). As all young speakers of Irish are bilingual, features of recent Irish English are being transferred to their Irish, both in grammar and pronunciation (…). One obvious feature of the speech of young female Irish speakers is the use of a retroflex [ɻ], especially in wordfinal syllable-codas after a vowel, a position where it is prominent in Irish English, e.g. north [no:ɻt̪], sore [so:ɻ]. In instances where the R of Irish is palatal, the retroflex [ɻ] is not acoustically obvious, but it is where R is nonpalatal (…) where leor /lʲo:r/ 'enough' is realised as [lʲo:ɻ]."

Hickey represents "the traditional velarised [ɹ]" using an ‹ɹ› with stroke, i.e. with a bar through the letter (to represent velarization). The sound is, naturally, referred to as traditional in Irish English, the traditional Irish Gaelic pronunciation is not mentioned.

In A sound atlas of Irish English, Hickey (2004: 78-79) sees the presence of uvular realizations of English /r/ (as a voiced uvular fricative [ʁ]) along the east coast of Ireland as a "relic of the first form of English in the east coast." The Wikipedia article "Munster Irish" never mentions the existince of a uvular allophone (2) of /ɾˠ/ in Munster Irish. Likewise, the Wikipedia article "Connacht Irish" never mentions the existince of an approximant allophone (0) of /ɾˠ/ in Connacht Irish. All that, coupled with the fact that the retroflex (3) [ɻ] is characteristic of northern Irish English (Hickey 2004: 78) and the Wiki article "Ulster Irish" never mentions its existence in Ulster Irish either, leads me to (preliminarily) assume that pronunciation 2 and 3 mentioned by seano, as well as pronunciation 0, are all likely to be just transfers from local Englishes. Are there any good reasons to think otherwise?
  • djwebb1969

    English - England
    In Munster Irish, an initial broad r is an alveolar approximant (like the English r), with the flapped r occurring later in the word.

    rí: [ɹi:]
    ort: [ort]
    < Previous | Next >