Talk:Scottish Gaelic phonology and orthography

Tones section needs example tone contour recordings

The section on Tones needs example tone contour recordings. E.g., in the minimal pair example:

fitheach ('raven') vs. fiach ('debt')

it's hard for a non-Gaelic reader to know whether the examples have one or two (spoken) syllables. This makes it a poor example, and it's an even worse one, because I've no idea what either of them sounds like. (Despite years of using the IPA to make necessary distinctions in other linguistic pursuits, I've never consciously heard Scottish Gaelic spoken.) Would it be possible for one of our excellent editors to record a brief sound file showing the difference between fitheach and fiach? And another showing the difference between typical tonal contours for one- and two-syllable words?

Also, could we please "eschew obfuscation" and not conclude the section with the pretentious-seeming:

... in the Scottish Gàidhealtachd.

? - I simply don't understand it, and should not be left guessing, e.g., that it might mean "Gaelicdom" (sic) or "Gaellaw" (sic). Thanks! yoyo (talk) 18:17, 8 May 2010 (UTC)

I can't make the recording, but I can try to explain the difference between fitheach and fiach as I understand it. They're both roughly [fiəx], but fitheach is two syllables [fi.əx] and fiach is one syllable [fiə̯x], and the primary way of hearing the difference between them is that fitheach has the falling pitch contour of uncontroversial dissyllables and fiach has the steady (perhaps very slightly rising) pitch contour of uncontroversial monosyllables. As for Gàidhealtachd, I've added a link so that those who don't know what it means can quickly find out. It may sound pretentious to throw in a Gaelic word as if everyone knew what it meant, but there just isn't any convenient English term for the Gaelic-speaking part of Scotland. +Angr 18:42, 8 May 2010 (UTC)

Allophones?

I am somewhat interested in Gaelic and what I can hear when I'm listening to it is voiceless alveolar retracted sibilant as the allophone of broad "s" after "r", voiced alveolar approximant as the allophone of "r" before all dentals & alveolars, and (very pronounced in Karen Matheson's speech) all velar fricatives going back in, the broad 1s sounding almost uvular and the slender 1s almost velar. Could anyone elaborate on this? 46.186.37.98 (talk) 01:28, 19 June 2011 (UTC)

In rC groups, the C often heads towards retroflex and in some cases, all you end up with is a retroflex C without any r, for example àrd /aːʈ/ Akerbeltz (talk) 10:14, 19 June 2011 (UTC)

You mean [aːʈ]? Peter238 (v̥ɪˑzɪʔ mɑˑɪ̯ tˢʰoˑk̚ pʰɛˑɪ̯d̥ʒ̊) 22:46, 7 November 2014 (UTC)

Why is spelling the prominent focus of this article?

This is a page about phonology. The phonemes and phones of the language should come first, followed by spelling information. As it is, the page starts with spellings, and then provides pronunciations. This is not how linguistics work is done. 112.169.28.146 (talk) 23:07, 25 September 2011 (UTC)

I agree, though it's primarily the vowels section that is the problem. Most of the orthographic information should be moved to Scottish Gaelic orthography if it's not already there. — Ƶ§œš¹ _{[ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ]} 19:07, 27 December 2011 (UTC)

Massively agree—the existence of both an orthography and a phonology+orthography page adds needless confusion. It's well past time for someone to restructure this article so that it more exclusively describes phonology, if anyone is inclined to do so. Acdavie-wiki (talk) 00:24, 5 December 2023 (UTC)

Vowel quadrangle

While vowel quadrangles are nice, I think the recent addition is not very good. The template/format used is really best for illustrating the cardinal vowel points, not the language-specific ones. Using it it this way misleadingly implies that the vowels are at or close to their cardinal values. I would rather we use an image of a trapezium that more precisely shows the values. In the meantime, the type of table that this recent addition is meant to replace is nice in its simplicity and would be good to augment a trapezium image, as is done at Received Pronunciation#Vowels. — Ƶ§œš¹ _{[ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ]} 03:45, 8 January 2012 (UTC)

The table was not entirely accurate. The main problem is, a very small study aside, we don't know where exactly the vowels lie. On the other hand, it was done by Ladefoged, so even though the sample is small, it should be kosher within that. My main problem is that I don't know how to convert that into Wikiformatting. I'll look into it, shall we leave it as it is for now and if I can't figure it, we'll use something else? Akerbeltz (talk) 12:13, 8 January 2012 (UTC)

Ok, run into the problem that for a quadrangle, we need a perception study and all we have are F1/2 measurements. Grrr... anyway, I redid it in table format, a bit clearer than the other table though I think. Akerbeltz (talk) 13:00, 8 January 2012 (UTC)

I have a similar problem with Russian phonology. I have the formant values for the vowels but I don't know how to convert them into a quadrilateral. There's a way to do it, but I'm not sure how exactly. — Ƶ§œš¹ _{[ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ]} 15:48, 8 January 2012 (UTC)

I made File:Welsh vowel chart.svg with Excel and Inkscape. I used Excel to plot the formants in a chart, and Inkscape to flip it around to put the high front vowels in the upper left corner, to stretch and compress the sides till they fit into a reasonable approximation of a vowel trapezoid, and draw the chart. Angr (talk) 16:32, 8 January 2012 (UTC)

Is that accurate though? I was swatting up a bit and it would appear that F1/2 values and the perceived distance, which is what the quadrangle measures, are not the same. Akerbeltz (talk) 17:18, 8 January 2012 (UTC)

Well, it's accurate for what it is, which is a graph of F1 plotted against F2. It approximates a vowel trapezoid, but that doesn't mean you could just put the points on a trapezoid in exactly the same position as they are on the graph (quite apart from the problem of ɑ, which seems to be in completely the wrong place). Angr (talk) 17:30, 8 January 2012 (UTC)

Maybe I should do that with the Russian vowel information I've got. Do the the points in the Welsh vowel grid reflect an average of the different contextual allophones or do they depict the phones in a context deemed most neutral? — Ƶ§œš¹ _{[ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ]} 20:31, 9 January 2012 (UTC)

AFAICT they depict the phones as uttered several times by a single speaker in the same context. For example, /i/ is from several recordings of the same person saying Mae tîm yma [mɑɨ tim əma] "There's a team here"; /ɪ/ is from several recordings of Mae cimwch yma [mɑɨ kɪmʊx əma], and so forth. I don't know criteria were used to decide on these contexts rather than others. Angr (talk) 21:03, 9 January 2012 (UTC)

BTW, you may like to read the comments at John Well's blog that includes a discussion of my Welsh vowel chart, in particular the antepenultimate comment, posted by Paul on 15 December 2009 at 10:24. If you understand what he's talking about (I don't) maybe you can avoid some of the problems that my chart apparently has. Angr (talk) 21:15, 9 January 2012 (UTC)

I actually do know what he's talking about. I brought up the issue of logarithmic vs. linear scales a couple of years ago in the reference but forgot that someone gave me a helpful link to convert to a logarithmic scale. I ought to investigate this before school starts. — Ƶ§œš¹ _{[ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ]} 23:49, 9 January 2012 (UTC)

I don't have your original data but looking at this study [1] the F1 you used seems to be odd when compared with the values for southern ɑ (page 9), you're bang on 800, the others cluster around 650 ish. Akerbeltz (talk) 22:40, 9 January 2012 (UTC)

Well, changing it on the basis of that other study is beyond my ability now. Maybe once Aeusoes learns how to draw logarithmic charts, he'll make a new one for Welsh based on the data from that study. Angr (talk) 00:18, 10 January 2012 (UTC)

nasalisation

with your kind permission, i would like to completely redo the current short nasalisation para which is devolid of citations. i am beyond unhsppy eith it. sorrry :-) i see that people have put in a serious amount of fine wirh in the article as a while and fir this kuch respect indeeed.

i would like ti doma new start based on a fully split, bruef dialectal verdion based kn the tabke un the aolendjx kf o maolalaigh and mac aonghais 1996 ScG in 300 Years. which accords extremely well with my exoerience and with rhe oerceotions of nstive soeakers regards dialect groupngs.

would that be ok?

it will takr me a while to finish this off, as my health is currently extremely poor im afraid. CecilWard (talk)

Sorry but you're not making sense. At all. There is no 'blocked lenition' in an taigh, it's not a leniting environment, taigh being masculine. Secondly, even if this was a leniting environment, the rules of synchronic nasalisation would still apply i.e. irrespective of whether it's an taigh or air an taigh, the phonetic outcome is /d/. So overall, this is NOT a reason to delete that section - which incidentally has TWO references.

I have, however, removed the ˠ marks from the IPA, Irish has a tendency to write these in the IPA but this is Gaelic where we do not have that. Akerbeltz (talk) 01:09, 14 June 2014 (UTC)

PS you even posted this on the wrong page I'm thinking, you actually deleted the section on the Lenition page but commented on this page. Akerbeltz (talk) 01:15, 14 June 2014 (UTC)

dear Akerbeltz, a charaid, thank you for sorting me out. i respect your work and am aware that you know what you're on about. :) indeed, thats what you get for having multiple windows open and getting confused between them. id like you to understand that i live on an hourly dose of very strong pain killers and they have a very debilitating effect on me, so i beg wikipedians' understanding.

however, unless yiu object, i would like to present at some point a simple three-way dialectal split example taken straight from ó maolalaigh 1996 appendix dialectal table fir phonetic surface realisations of nasalisation. to appear in the talk page for your approval and your yea or nay. if yiu should hate the proposal, then it wont happen. is mise le meas CecilWard (talk) 02:21, 14 June 2014 (UTC)

Hey. Sorry to hear of your health issues and yes, I understand how that can affect stuff.

I have no issue with expanding the details on nasalization, no reason why I should. Ó M is a good starting point, you'll have to tweak his IPA slightly. Other than that, go for it. Perhaps if you write up the section in your sandbox first, we can play around with it there and avoid toing and froing in the mainspace article? Akerbeltz (talk) 11:13, 14 June 2014 (UTC)

new references

they seem pretty out of date, or at least, no reference to new instrumental phonetic work - and why not cite http://www.jstor.org/discover/10.2307/40732061?uid=3738032&uid=2&uid=4&sid=21104711563543 ref the pre-aspiration map? — Preceding unsigned comment added by 81.86.104.165 (talk) 14:20, 19 September 2014 (UTC)

Table of vowel phonemes

@Akerbeltz: You're mistaking phonemic classification for narrow phonetic description. I'm not saying that /ɪ/ is a fully front vowel, I'm saying that it's a front vowel phonologically. There isn't such a thing as a phonological near-front vowel as no language contrasts front, near-front and central vowels. Phonetically, it probably varies from centralized front to central.

I'm reverting.

Is /ɪ/ even a phoneme? Or is it just a short counterpart of /iː/? If it's the latter, we shouldn't even list it in the table. Mr KEBAB (talk) 13:46, 28 March 2018 (UTC)

You clearly know a lot about Gaelic... not. The Short counterpart of /iː/ is /i/. Both of which may occur in stressed position, neither in unstressed. You can indeed argue that /ɪ/ is an unstressed allophone of /i/ but not /i:/. Akerbeltz (talk) 15:08, 28 March 2018 (UTC)

@Akerbeltz: You're completely confusing allophones with phonemes. Only the latter can be enclosed within slashes. You've also misunderstood my post - I said short, not unstressed.

We now have a reason to remove ɪ from the table, as it doesn't seem to be a phoneme. Mr KEBAB (talk) 15:12, 28 March 2018 (UTC)

I'd just like point out that if you are going to make changes you have to cite a scholarly source as per Wikipedia policy. Since Akerbeltz wrote the book that the table currently has as a reference, we can assume that any changes are not reflected in the source. Catrìona (talk) 01:13, 29 March 2018 (UTC)

@Catrìona: That explains WP:OWNERSHIP vibes I got from his response.

Judging by the constant confusion of allophones and phonemes (also see other topics on this talk page), his book might not even be WP:RELIABLE enough to cite, at least in this particular case. I haven't read it so I wouldn't know for sure, I'm just sharing my observations and will leave that for other editors to decide. I'll just repeat: "near-front" isn't a phonological label for a vowel but a narrow phonetic description thereof. No language contrasts front, near-front and central vowels and they're always classified as either front or central, depending on the language. See The Sounds of the World's Languages (1996).

MacAulay (1992), who's cited directly above the table lists [ɪ] as one of the allophones of /i/ which can be either [i], [ɪ] or [ɨ]. Now, the question is - what does the other book say and does it get the allophone-phoneme distinction right? Mr KEBAB (talk) 01:29, 29 March 2018 (UTC)

Hah, the vibe you're getting is that I don't really care that much about the English Wiki any more. As for WP:OWNERSHIP, don't make me laugh. Check the page history, I have had virtually no input to this page apart from somy tiny tweaks to the IPA as Wikipedia uses a closer transcription that Celticist publications often use. As for the reference to my book, that was written as a practical textbook and not a linguistic treatise. For purely practical didatic reasons, I do not distinguish phonemes and allophones in the same rigid way a linguist would in a scientific publication. That's presumably the reason why /i:/ /i/ and /ɪ/ ended up in the vowel table (again, check the history, not my doing). Akerbeltz (talk) 19:59, 29 March 2018 (UTC)

Acute accents

This article states é and ó are still used for /e:/ and /o:/, and è and ò are for the open-mid versions. However, as per here on Wiktionary's Tea Room, it would seem the acute accent has been done away with. MGorrone (talk) 19:47, 19 May 2021 (UTC)

It's complicated. They were abolished in the education system in Scotland and many official bodies choose to adhere to the GOC reforms too but it's not universally accepted. In addition, Nova Scotia, which is a sizable and historic emigrant community, did not carry any of the GOC reforms. We should probably stick in a note saying that é > è and ó > ò under GOC but leave it up to editors to chose their orthography of choice, it would turn into a mess trying to enforce one over the other. Akerbeltz (talk) 09:35, 20 May 2021 (UTC)

More links and references reqired

I feel that much of this article is promoting the work of a particular author from the Akerbeltz website, for example "a near-exhaustive list". I have checked this website and it is very detailed but, 'near-exhaustive' seems like self-promotion, especially given that the website has very little information about variations in phonology, is there any chance that we could add some other links and references? — Preceding unsigned comment added by 2.31.240.51 (talk) 10:22, 24 January 2023 (UTC)

The wording of that external link was not by me but it was, as you said, a bit OTT, so I changed it to "detailed" (which it is). In terms of variation, there is quite a bit on the page and while there aren't that many refs, that's down probably for a lack of time/editors but on the other hand, none of the things stated are controversial and *could* easily be ref'd if someone had the time and inclination. Akerbeltz (talk) 10:55, 25 January 2023 (UTC)