User talk:Mr KEBAB/Archive 13

Archive 10

Archive 11

Archive 12

Archive 13

Archive 14

Archive 15

→

Archive 19

Vowel patterns in GenAm and RP

I like your recent edits at the General American article. Illuminates the tense-lax 5-vowel systems of /i, u, eɪ, oʊ, ɑ/ and /ɪ, ʊ, ɛ, ʌ, æ/ in the vowel space. The problem was that /eɪ, oʊ/ are frequently regarded as diphthongs like in RP which obfuscates the relation in GenAm (you have moved them to the main table which is better as they are mostly monophthongal). Any way to make the table for RP in the English phonology article reflect the phonological vowel space? I think a table for RP, like the one you created for GenAm, would have to treat /iː, uː/ on par with /eɪ, əʊ/.--Officer781 (talk) 15:24, 14 November 2017 (UTC)

@Officer781: Thanks. I'll think about something, but for now I'll just say that GA /eɪ, oʊ/ are not mostly monophthongal (see the article). To call them such was one of the biggest mistakes in the 8th edition of Gimson's Pronunciation of English. They can be monophthongized from time to time, but probably never to the quality of Scottish /e, o/ (especially in the case of /oʊ/), with which they are equated in that book. At least in regions in which the normal pronunciations are monophthongal diphthongal.

Another mistake Cruttenden makes is to show /ɑː/ as central. I strongly disagree with this, as the vowel is noticeably more back than this - it's between central and back or even slightly backer than that (but only slightly, Geordie and some South Africans have a truly back vowel.) This is evident when you hear an RP-speaking person pronouncing BATH words in a conversation with an American. A central vowel wouldn't have such a shockingly different sound (yes, hearing RP-speaking Britons pronounce BATH words still takes me aback after listening to American English for too long, as my ears have to readjust.) Mr KEBAB (talk) 21:03, 14 November 2017 (UTC)

@Officer781: I'm actually thinking about removing the Modern RP vowel chart from Commons. It has at least two noticeable mistakes. We need something better. Mr KEBAB (talk) 21:05, 14 November 2017 (UTC)

@Officer781: What do you think? Mr KEBAB (talk) 03:51, 23 November 2017 (UTC)

Sorry, I was on holiday in London and it was difficult for me to reply using my phone. Yes the vowel placements don't seem very accurate, although it certainly looks progressive. The vowel values look to me more like popular London or how RP will turn out in the near future. We can always say that it's modern British English rather than RP? The only real issue I have with it is that it shows STRUT as a range but the rest as vowels. I checked the source and it appears to give STRUT as discrete values as well depending on how "modern" the RP is?--Officer781 (talk) 17:03, 23 November 2017 (UTC)

@Officer781: No problem! Did you by any chance listen to the way Londoners pronounce their /ɑː/? To your ears, was the central realization in any way common?

I'm not sure about that vowel chart being progressive, especially in the case of /ɔː/. It's shown as lower than on Roach's chart, which falsely suggests /ɔː/-lowering as an actual change in RP, a change that is not happening.

The lower left corner of the /ʌ/ range is a more [ɐ]-like realization, the upper left right corner is a more [ʌ]-like realization. Both are shown in the source. Certainly, the [ʌ]-type can be thought of as an innovation triggered by the TRAP lowering (to the open front position). It wasn't a mainstream RP realization until very recently, and even now its mainstream status is rather questionable. Mr KEBAB (talk) 18:19, 26 November 2017 (UTC)

Ah, okay that makes sense. Thanks for pointing the /ɔː/ and /ʌ/ points out. That makes me wonder where /ʌ/ "theoretically" (like all those [+back] stuff) belongs in the RP vowel space: central like AusE or back like GenAm? I noticed that the back variant of /ɑː/ still seems to be the common realization (I heard it more often than the central variant), although curiously the recorded train announcement (if I remember correctly. Was it the bus or the train?) pronounces /ɑː/ as central despite pronouncing /ɪə/ as a pronounced diphthong (suggesting conservative RP).--Officer781 (talk) 20:17, 26 November 2017 (UTC)

@Officer781: There isn't one correct way of presenting the vowel space of RP. /ʌ/ can be analyzed as mid central, mid back, open central or open back, depending on the analysis. We also have to remember that the central-back contrast exists only in the case of mid vowels. Elsewhere, the contrast is between front and non-front or back and non-back vowels. I'm not sure which set of names is more correct in phonology.

The vowel space of modern RP is roughly this:

Close, front /ɪ, ɪː/

Close, non-front /ø, øː/ (which Lindsey writes /ɵ, ɵː/)

Mid, front /ɛ, ɛː/

Mid, non-front:

Central /ə, əː/

Back /ɔ, oː/

Open, front /a/

Open, non-front /ɐ, ɑː/

Diphthongal, with front offset /ɪi, ɛi, oi, ɑi/

Diphthongal, with non-front offset /øʉ, əʉ, ao/ (which Lindsey writes /ɵu, əu, au/, which is completely incorrect in the case of the GOOSE vowel, which isn't a backing but closing diphthong)

The vowel space of conservative RP could be said to be roughly like this:

Close, front /ɪ, ɪi/

Close, non-front /ʊ, ʊu/

Mid, front /e/

Mid, non-front:

Central /ə, əː/

Back /oː/

Open, front /æ/

Open, non-front:

Central /ɐ/

Back /ɒ, ɑː/

Diphthongal, with front offset /eɪ, ɔɪ, aɪ/

Diphthongal, with non-front offset /ɪə, ʊə, ɛə, əʊ, aʊ/

Doesn't it look considerably less natural to you? Because to me, it does! And apparently, conservative RP also contrasts short open central and back vowels.

Again, I'm not sure whether /ɑː/-fronting is an actual change in RP. To me, a central [äː] sounds non-RP and non-Estuary and is a sign of a near-RP accent of someone from Wales/Northern England/Norfolk/some other place. But I could be wrong.

[ɪə ~ iə] for /ɪə/ isn't really a conservative realization, as the change to [ɪː] is far from complete. However, if you consistently use a centering diphthong (especially one of the [ɪ̯ə ~ i̯ə] type) instead of [ɪː], chances are that you're a conservative speaker indeed. Mr KEBAB (talk) 07:18, 27 November 2017 (UTC)

Hello

Hello, long time no see. I have a few things I’ve been meaning to say/ask.

1. Interestingly, Zuch—a German pronunciation spelling of my surname—is actually Luxembourgish for Zug.

2. Would you happen to know roughly when -owa surname suffixes went out of fashion/ceased to be used much?

3. Is Danusia usually a diminutive of Danuta?

4. I’ve been doing some research into family history. What do you think would be the best way to continue? I could perhaps pay for a Polish genealogist to have a look at some records. There’s also the possibility of going there one day and maybe accessing some birth certificates, etc.

Rovingrobert (talk) 06:43, 5 December 2017 (UTC)

@Rovingrobert: Hi. Luxembourgish is just a standardized dialect of German, so that's not surprising. But actually, Zuch is just a High German loanword, but loaned with a local (Western German) pronunciation. The native counterpart is Zoch and it means a current of air. Interestingly, because Luxembourgish /o/ is typically quite close (about close-mid [o], Zoch can sound just like Northern German Zug [tsʊχ]. IMO it's hugely unfair to consider Luxembourgish a separate language while Alemannic and Bavarian dialects are just lumped together with High German as German. I know that standardizing a language isn't very easy, but for goodness sake...

I'd say quite a long time ago, at least about 30-40 years ago. See also [1] and [2], if you can find someone to translate those for you.

Yes.

To be honest, I'm not sure. This information is readily available for me whenever I choose to read about it. I have at least two family members that are into this stuff. With that being said, what you've written makes sense, and I'd go for it. Mr KEBAB (talk) 08:22, 5 December 2017 (UTC)

Found out anything interesting about your family history you wouldn't mind sharing? By the way, do your family use the old declensions when referring to relatives who lived before the surname reforms?

You said that "⟨C⟩ as /s/ before a back vowel (/ʊ, uː, ʌ/ are phonologically back even in Australia) is counter-intuitive for natives." In America, would this be more common, owing perhaps to the Spanish influence?

Which name is Janek most likely to be a diminutive of?

Also, if you don't mind me asking, what is the secret to your command of the English language despite living in Poland?

Many thanks, Rovingrobert (talk) 08:03, 16 December 2017 (UTC)

Carl Sagan pronunciation

Mousing over your proposed change to the pronunciation of Carl Sagan shows the last syllable pronounced as "on" as in button. I have never heard Carl's name pronounced that way. The previous pronunciation seems correct to me. Thank you. JeanLucMargot (talk) 07:07, 20 January 2018 (UTC)

@JeanLucMargot: You must have because it's the only correct pronunciation. The spelling is different, but remember that we're talking about English. Its spelling is hopelessly complicated in comparison with Dutch or Polish. The last two phonemes of both of the words are /ə/ and /n/. In button, they can be turned into a syllabic [n̩] as in Dutch laten [ˈlaːtn̩] as pronounced by people from West Flanders and Eastern Netherlands (your accent, if it's the one of Leuven, probably just leaves out the /n/ as many other varieties, yielding [ˈlaːtə]). When button is pronounced with a syllabic 'n', then the preceding /t/ can (and often is) turned into a simple glottal stop, yielding bu'n [ˈbʌʔn̩].

Ok, but what about Sagan? It's the same as far as the last two sounds are concerned (so [ˈseɪɡn̩ ~ ˈseɪɡən]), except /ɡ/ is never glottalized in any of the standard accents. Maybe it can be glottalized in Cockney (which is dying out), but only rarely.

Here's more on syllabic consonants: [3]

By the way, maybe my edit summary was too harsh. Sorry for that. Mr KEBAB (talk) 11:31, 20 January 2018 (UTC)

Thank you for the explanation. It's the "o" in "button" that really confused me. Indeed, the phonetic pronunciations of "button" given in various dictionaries do not include /ən/. Would it be reasonable to propose that the example given when mousing over /ən/ be improved? It seems awfully confusing, at least to this wikipedia user. Thank you. JeanLucMargot (talk) 17:29, 20 January 2018 (UTC)

@JeanLucMargot: Can you give me an example of that? Differentiating the final sounds of Sagan and button strikes me as strongly non-native. Mr KEBAB (talk) 02:54, 21 January 2018 (UTC)

Sure. The final sounds of "pagan" and "button" are shown as distinct in at least three dictionaries: Merriam-Webster, dictionary.com, and Apple's. Are they all wrong? Seems unlikely. JeanLucMargot (talk) 18:22, 21 January 2018 (UTC)

I think the principle laid out at Wikipedia:Manual of Style/Pronunciation and Help:IPA/English is that we regard any sequence of an obstruent + /n/ or /ən/ as underlyingly /ən/, defined as "'on' in 'button'", because they do not contrast (English phonology § Consonants), even if one was more popular than the other in a given word. Speakers are not consistent about this: e.g. There were signs of a fast increasing minority tendency for General British speakers to favour the use of schwa plus (unsyllabic) consonant where previously syllabic consonants were the norm, eg in cotton, garden, bottle and struggle and even increasingly in such items as assembly, doubly, gambling, cackling etc for which it is doubtful that they ever previously contained a syllabic consonant.[4]

This is also the very reason {{IPAc-en|ən}} and {{IPAc-en|ə|n}} are designed to yield ostensibly the same result except in the mouseovers. At Help talk:IPA/English there have been suggestions that we use superscript "ᵊ" or parenthetical "(ə)" like some dictionaries do, but in general editors seem to have agreed that such practices would bring too much clutter without much gain. That is, the distinction between [ən] and [n̩] is considered so insignificant or uncertain that they do not deserve to be readily differentiated.

Cambridge and Oxford are some of the dictionaries which transcribe button and pagan the same way, by the way: [5][6][7][8]. Nardog (talk) 18:11, 22 January 2018 (UTC)

@JeanLucMargot: Yep, pretty much what he said. Mr KEBAB (talk) 08:52, 24 January 2018 (UTC)

Berlin stress explanation

What does it mean in a two-syllable word for one of the syllables to have "secondary stress"? I'd always assumed secondary stress appeared in words of three syllables or more. (Perhaps, if this is easier to explain through audio files, e.g. on Forvo, do you know of any that help clarify this idea?) Thank you. Wolfdog (talk) 09:43, 26 January 2018 (UTC)

@Wolfdog: In that particular case it probably means that /ɜːr/ has a slightly greater length ([ˌbɜˑ]) and more prominence than the unstressed /ɜːr/ ([bɜ]). I guess this distinction isn't relevant in GA, in which it could be shown as /bərˈlɪn/ vs. /bɜːrˈlɪn/, but remember that transcriptions enclosed within IPAc-en are diaphonemic. Mr KEBAB (talk) 09:56, 26 January 2018 (UTC)

Hm, it seems a bit vague to me. Yes, I do understand the diaphonemic nature of the transcriptions, but I also see that Oxford Dictionaries shows /ər/ as well, so perhaps I can add it back in as a third option.

Also, on an only semi-related note, can you think of a minimal pair that distinguishes /ɜːr/ versus /ər/ (obviously, in something like a UK accent)?— Preceding unsigned comment added by Wolfdog (talk • contribs) 10:17, 26 January 2018 (UTC)

Secondary stress (and stress in general) can imply many things depending on author. Mr KEBAB's is the canonical definition, but I don't think /bɜːrˈlɪn/ and /ˌbɜːrˈlɪn/ would result in any appreciable difference when said in isolation. Rather, what dictionaries mean by e.g. /(ˌ)bɜːrˈlɪn/ is that the prominence on syllables may or may not shift in connected speech. In general, British dictionaries only recognize secondary stress in this sense. American ones attribute it to unstressed full vowels as well, and also use it to distinguish [ʌ] vs. [ə], [t] or [d] vs. [ɾ], etc. (But even dictionaries in the same school disagree when it comes to secondary stress – LPD attributes it to the penultimate syllable of -ism words while CEPD permits it to occur after primary stress only in compounds.) Nardog (talk) 11:16, 26 January 2018 (UTC)

@Wolfdog: Oxford Dictionaries gives /bɜːˈlɪn/ for RP and /bərˈlɪn, -ˈliːn/ for GA. Are you saying that there's a relevant difference between unstressed /ɜːr/ and /ər/ in GA? If so, I disagree (and so does our article on General American - or doesn't it?). But we should mention the /-ˈliːn/ variant.

Even ignoring length, there isn't a single BrE speaker for whom /ə/ and /ɜː/ have exactly the same allophonic range, as the unstressed, extra-short nature of the former causes it to move around the mid central vowel space depending on the consonants in its vicinity. Non-final /ə/ varies from close-mid to mid and central to near-back and it can be rounded to [ʊ] before the velarized allophone of /l/. /ɜː/ is typically mid to open-mid (but can also be close-mid) and is always central and unrounded, for many speakers also sulcalized. The difference is even more pronounced in Brummie, New Zealand, South Africa and some Cockney, in which a fronted, rounded [œː] type of vowel is used. So it doesn't really matter that there aren't many minimal pairs (one is forward /ˈfɔːwəd/ vs. foreword /ˈfɔːwɜːd/, I don't remember more than this one). Native speakers can readily distinguish /ə/ from /ɜː/ and to them, /bəˈlɪn/ is diffferent from /bɜːˈlɪn/. But I don't know whether they can consistently differentiate /bɜːˈlɪn/ from /ˌbɜːˈlɪn/.

Also, in the word-final position, there's almost always a difference in height. The final schwa is very often too ah-like to be classified as belonging to the /ɜː/ phoneme, but I'm not convinced that [ɐ] is always the best transcription. To me, it sometimes sounds like a centralized cardinal [ʌ]. Either way, it's a very consistent error that English speakers make when trying to speak German (it's also a great example of pronunciation errors causing grammatical errors, as -e [-ə] is a different ending from -er [-ɐ]).

Another thing to consider is that the sequences of /ə/ + sonorant often result in a syllabic sonorant. This is never true in the case of the sequence of /ɜː/ + sonorant, so that e.g. U-turn is always [ˈjuːtɜːn] (or [ˌjuːˈtɜːn]), never *[ˈjuːtn̩]. The same is true of /ʌ/ + sonorant, so airgun and megaton are [ˈɛːɡʌn, ˈmɛɡətʌn], not *[ˈɛːɡn̩ ~ ˈɛːɡŋ̍, ˈmɛɡətn̩]. I'd be really interested if [ˈɛːɡn̩ ~ ˈɛːɡŋ̍, ˈmɛɡətn̩] are possible pronunciations in any accent with the STRUT-COMMA merger by the way. Mr KEBAB (talk) 06:09, 27 January 2018 (UTC)

@Nardog: Yes, I added the /ˌbɜːr-/ variant per LPD. I have no problem with removing it. Mr KEBAB (talk) 06:09, 27 January 2018 (UTC)

Oxford Dictionaries actually uses /əː/ (though, now that I notice it, the indicator of length here probably suggests that this is simply their way of transcribing what Wells would transcribe as /ɜː/). No, of course, there's certainly no "relevant difference between unstressed /ɜːr/ and /ər/ in GA". I'd admit that before anyone else; after all, I'm the one with all the questions and having a hard time imagining how Brits distinguish the two.

Honestly, I can't say I've ever heard an American use "the /-ˈliːn/ variant" for Berlin, though it certainly does better approximate the German pronunciation.

You say that "there isn't a single BrE speaker for whom /ə/ and /ɜː/ have exactly the same allophonic range" and while this may be true (I have no way to test it), this doesn't close off the possibility that for BrE speakers the allophonic range of /ə/ and /ɜː/ may still sometimes overlap. If I can be a bit non-technical, I assume by "sulcalized" in this context you mean something like "tensed up" or "not relaxed", as if forming the mouth-shape for /ɹ/ without pronouncing it. Anyway, the fact that you're saying it "doesn't really matter" suggests that you find me argumentative or feel that I'm skeptical or disagreeable. That's not true at all. I was just asking for minimal pairs for my own personal amusement, knowledge, and so on, so no it doesn't "matter" in the way of any argument, just for own personal curiosity. I think the distinction fascinating.

I agree that English speakers tend to mistakenly neutralize the difference between the German endings ⟨-e⟩ and ⟨-er⟩.

Your final paragraph makes me think of the rare few Americans I sometimes hear saying /ˈpinət/ for peanut or /ˈdoʊnət/ for doughnut, which always comes across as strange to me, whereas I certainly say /ˈpinʌt/ and /ˈdoʊnʌt/. Wolfdog (talk) 04:43, 28 January 2018 (UTC)

@Wolfdog: Yes, of course they're the same phoneme.

The source for that is the latest edition of Gimson's Pronunciation of English. It says that the range of the non-final /ə/ is [ɘ ~ ɘ̠ ~ ə]. The range of /ɜː/ is [ɘː ~ əː ~ ɜː ~ ɐː], with the last one being extremely old-fashioned.

I can't explain it better than our article. All I can add is that to me, sulcal /ɜː/ at times sounds almost creaky-voiced, but that realization seems to be more widespread among older speakers.

Oh, sorry. Apparently I missed on an only semi-related note. The reason I said that it didn't matter was that the distinction is obvious for Anglo-Welsh speakers, but that's pretty much stating the obvious anyway. Yes, the distinction is quite interesting (it doesn't appear word-initialy by the way, at least when you discount /ə/ that historically wasn't followed by /r/), as is the distinction between strong /ɪ/ (phonetically [ɪ ~ ë]) and weak /ɪ/ (phonetically [ɘ ~ ë ~ ɪ]). They both belong to the /ɪ/ phoneme. Generally, that [ɘ] of weak /ɪ/ isn't quite the same as the [ɘ] that belongs to the /ə/ phoneme. It's less central and perhaps slightly higher than the latter. To an outsider, this distinction might be very, very hard to hear.

I vaguely remember a restaurant joke in which an Englishman ended a sentence with bitter ('bitter') instead of bitte ('please'), but my German is too bad to say it properly. But yeah, it's really common.

In RP, intraalveolar /ə/ tends to be raised to [ɘ], which isn't a normal realization of /ʌ/ in most accents. It's more like /ɪ/ than /ʌ/, that's why when RP speakers talk about lizards they pronounce that word almost like lizids ([ˈlɪzɘdz]), though the sounds are still distinct and the phonemic form of the word is /ˈlɪzədz/ - see above. What I'm saying is that the same might apply to some forms of American English, which is why /ˈpinət, ˈdoʊnət/ sound so weird to you. Mr KEBAB (talk) 12:51, 28 January 2018 (UTC)

Understanding broad variants of English

Cześć, I wonder how easy it is for you to understand broad, non-standard accents of English? For example, in this comedy video two people from Glasgow speak in a very unique accent. Also another video a speaker mentions about their apparently very strong accent from Arkansas. If you can understand them, I wonder if you can analyse them for me as well? Even I myself, as an Australian, could pretty much understand the whole thing. I think these should be interesting for you! — ^{they call me} AWESOMEmeeos ... [ˈɔɪ̯]! 22:54, 28 January 2018 (UTC)

@Awesomemeeos: I haven't done any serious research on those. Have you checked the literature? Mr KEBAB (talk) 09:53, 30 January 2018 (UTC)

I did do some research on both of them (more on the second accent), but the main question was, how much did you understood them? — ^{they call me} AWESOMEmeeos ... [ˈɔɪ̯]! 10:07, 30 January 2018 (UTC)

@Awesomemeeos: Everything, but I tend to read the comments as I watch the video, which helped me a bit. Mr KEBAB (talk) 10:10, 30 January 2018 (UTC)

Very interesting accent

I think you'll like this speech which shows a city councillor from Medford, Massachusetts legitimately speaking in a very broad accent from over there. I think it's a really unique accent that you wouldn't hear anywhere else in the US, probably which you haven't heard before, but correct me if I'm wrong. — ^{they call me} AWESOMEmeeos ... [ˈɔɪ̯]! 09:00, 8 February 2018 (UTC)

@Awesomemeeos: Sounds like a mix of New York English (because it's non-rhotic), General American (because THOUGHT is rather low) and Australian English (because PALM/START is central or even front of central [aː]). It's pretty good. Mr KEBAB (talk) 10:18, 8 February 2018 (UTC)

Although I must say my Australian PALM/START vowel is ever-so-slightly back, something like [ɐː]. His just vowel sounded extremely fronted, even as an Aussie I must say. But I liked 0:43-0:44 "[...] go through every park in the community [...]" and 0:47-0:48 "[...] start doing the cleanups [...]", where it's emphasised the most — ^{they call me} AWESOMEmeeos ... [ˈɔɪ̯]! 10:56, 8 February 2018 (UTC)

@Awesomemeeos: Yeah, a truly front [aː] is rather old-fashioned in Australia. You'd be more familiar with it if you were from [ˈkæːdɪf] or Liverpool. Mr KEBAB (talk)

Weird pronunciations

I just happened to be browsing Wikipedia when I bumped into these unusual (non-standard) pronunciations from Great Lakes Region of the words accent, coupon and naturally. Whaddya reckon? That tense [æ] in those pronunciations tho! — ^{they call me} AWESOMEmeeos ... [ˈɔɪ̯]! 10:09, 15 February 2018 (UTC)

@Awesomemeeos: Non-pre-nasal [eə] for /æ/ doesn't sound terribly strange to me, but [ˈkupan] sounds like something an EFL friend would say when mocking the General American accent. It is a bit funny. Mr KEBAB (talk) 12:37, 15 February 2018 (UTC)

About this accent, I managed to find an informative short documentary about it, featuring William Labov. Please watch it, as it has vowel charts explaining the Northern Cities Vowel Shift. — ^{they call me} AWESOMEmeeos ... [ˈɔɪ̯]! 07:21, 16 February 2018 (UTC)

@Awesomemeeos: Thanks, but I've already watched it a long time ago. Our article on the NCVS is also pretty good. Mr KEBAB (talk) 11:49, 16 February 2018 (UTC)

Apical [ɹ̺] vs. retroflex [ɻ]

Do you know the difference between these two? Even after reading stuff like this Wells post and the relevant sections in Ladefoged's A Course in Phonetics ("Rhotic Vowels", "Rhotacized Vowels"), I never understood what the difference between a postalveolar approximant and a retroflex one people sometimes claim there is was, apart from the one between apical and molar. But if people claim a need for a distinction between apical and molar [ɹ] in notation, as seen in the Wells post and in extIPA, that means they do feel the need to distinguish something that cannot be accomplished by simply using ⟨ɹ⟩ and ⟨ɻ⟩, right? In fact I have found scholars refer to the apical /r/ as "retroflex".[9][10] Perhaps this is the distinction between the true, subapical retroflex and the apical postalveolar "retroflex", and some people would like a symbol other than ⟨ɻ⟩ for the apical /r/ because they are opposed to referring to the latter as "retroflex"? This has puzzled me for so long, and I wonder if you have some knowledge about it. Thank you for your contribution and keep up the good work, Nardog (talk) 17:31, 15 February 2018 (UTC)

@Nardog: I'll check my sources. Sorry for the delay, I'm a bit busy (also on WP). Mr KEBAB (talk) 13:29, 17 February 2018 (UTC)

@Nardog: The molar [ɹ] as described by Wells is not only not retroflex but also not even coronal. The way he describes it, it's a kind of a dorso-uvular (dorso-preuvular?) approximant with pharyngealization. So it's the dorsum that is the active articulator, not the tip/blade of the tongue.

I'll try to do a better research on it. Mr KEBAB (talk) 13:44, 17 February 2018 (UTC)

It's okay, I just wondered if you already had the answer. So there's no rush, but I appreciate your looking into it too.

You mean by Catford? No, he's just describing the pharyngealization and sulcalization that may accompany the molar articulation of [ɹ] by telling the reader to produce [ʀ] and then lower the tongue. The primary place of articulation is indeed postalveolar (or perhaps lamino-palatal?); see e.g. [11]. Nardog (talk) 17:59, 17 February 2018 (UTC)

@Nardog: Haha, I knew there was something wrong with that description. You're right, it's a laminal postalveolar approximant with a secondary articulation of pharyngealization and sulcalization. And yes, it's Catford, not Wells. Sorry.

The more I read about phonetics the more I just want to ditch the term retroflex. Subapical postalveolar is good enough and it correctly tells the reader that retroflexes are extremely similar to postalveolars both auditorily and in being coronal rather than dorsal. Mr KEBAB (talk) 19:25, 17 February 2018 (UTC)

I've posted this at Talk:Pronunciation of English /r/, so you can add your input there if you find anything in the future. Nardog (talk) 15:20, 18 February 2018 (UTC)

@Nardog: Ok, so I'll just archive this discussion. Mr KEBAB (talk) 15:40, 18 February 2018 (UTC)