Talk:General American English

Canadian raising of äɪ to ʌɪ

I have a doubt whether Canadian raising actually would raise the diphthong [äɪ] to [ʌɪ]. The vowel [ʌ] is pronounced too far back in the mouth. Could anybody verify this claim? I don't need a source, only somebody's word. Thank you.LakeKayak (talk) 02:30, 27 December 2016 (UTC)

Wikipedia's audio recording for [ʌ] sounds right to me, as does the audio recording for [ə], and possibly [ɐ]. [ɜ] sounds a little wrong to me. This is just my own perception of the Wikipedia audio recordings. Wolfdog (talk) 14:47, 27 December 2016 (UTC)

I beg your pardon, sir. I am little confused. I fail to see the relevancy of the audios. Could you please elaborate? Thank you.LakeKayak (talk) 00:43, 28 December 2016 (UTC)

The vowels I mentioned all sound right to me as the nucleus for Canadian-raised /äɪ/ in the U.S. Wolfdog (talk) 02:31, 28 December 2016 (UTC)

Oh, you lost me for a second there. Thank you, sir.LakeKayak (talk) 02:33, 28 December 2016 (UTC)

@Wolfdog: When referring to the audio for [ʌ], which audio are you talking about? I'm a little confused. Do you mean the audio on this page for [ʌ~ɐ] or the audio [ʌ] on IPA vowel chart with audio. I can see the former as the nucleus for the raised [äɪ] more than the latter.LakeKayak (talk) 03:15, 28 December 2016 (UTC)

It's worth noting that the page you just shared has audio from a Canadian user uploading sounds. There are numerous linguists that put Canadian English in a different umbrella than any american english. Some of his sounds can still be found on this page for general american pronunciations. This page does say that canadian english "arguably" falls under general american. I'm on the disagreement side of this, personally. Accent coach Claudette Roche has said Canadians and Americans have different vowels. https://www.facebook.com/permalink.php?story_fbid=10151861837052908&id=124536002907 Obviously, there are going to be contradictions here and there on any wikipedia page (which is why I never though of removing a couple of that users sound files from this page). I don't agree with the part in the GA article text that says two thirds of americans speak with a general american accent. But since it's sourced, it should stay. That said, I do know a Canadian actor that had to modify his speech to sound more american when he tried to enter the US market. And he had to change the way many of the vowels sound. This, along with the accent coach that has a profession in this, along with numerous linguists (Charles Boberg is an enthusiastic supporter of the uniqueness of Canadian English) I've talked to in regards to this, makes me very weary that Canadian English is the same as any american English. Bisnic95 (talk) 06:25, 28 December 2016 (UTC)

@Bisnic95: I beg your pardon, sir/ma'am. I don't recall sharing any page. Furthermore, I don't think I even know how to share a page.LakeKayak (talk) 15:52, 28 December 2016 (UTC)

@Bisnic95: As another note, I would prefer to say on topic. Whether or not Canadian English is a branch of American English is of little to no relevancy to my question. Thank you.LakeKayak (talk) 15:57, 28 December 2016 (UTC)

I'm not sure of Bisnic95's point either. Anyway, LakeKayak, I was referring to the vowels of the vowel chart. Wolfdog (talk) 16:47, 29 December 2016 (UTC)

Thank you.LakeKayak (talk) 02:51, 7 January 2017 (UTC)

Replacing /ʌ/ with /ʌ̈/

The latter symbol is used to describe the result of fronting /ʌ/ in the section "Fronting of short u." Therefore, I think possibly this vowel should be the one used to describe the General American short u. I am going to make the edits. However, I do respect that anybody else has the right to revert them.LakeKayak (talk) 21:14, 29 December 2016 (UTC)

Done. In the end only three cases of [ʌ] were altered to [ʌ̈]. Over and out.LakeKayak (talk) 00:26, 30 December 2016 (UTC)

How is this different from [ɜ]? If it's somewhat fronted but not completely centralized, then it feels like it should be [ʌ̟] instead. Nardog (talk) 04:01, 6 June 2017 (UTC)

@Nardog: It was simply the notation already used on the page, and it didn't seem appropriate at the time to change everything. If you strongly feel otherwise, go ahead. I don't object.LakeKayak (talk) 18:30, 7 June 2017 (UTC)

[ɜ] is typically the representation of the Southern STRUT vowel. I'd be wary of using that. Listen to it:

That certainly doesn't sound like a General American "uh" to me. Wolfdog (talk) 00:00, 8 June 2017 (UTC)

Well according to Relative articulation#Centralized vowels a dieresis doesn't necessarily mean completely centralized, so I'm not implementing a change at the moment. Whatever the case, any transcription on Wikipedia must adhere to what the relevant reliable source describes rather than one's empirical and/or anecdotal evidence. Nardog (talk) 05:19, 8 June 2017 (UTC)

Velarized l

To my knowledge, in American English, the initial "l" still remains un-velarized. Regardless of whether or not this is correct, I am confused either way. Can anybody help me out? Thank you.LakeKayak (talk) 00:00, 7 January 2017 (UTC)

• What it says now seems right to me. If it's unvelarized, it often comes across as sounding foreign or ethnically Latino. Wolfdog (talk) 02:02, 7 January 2017 (UTC)

There is still a distinction made between the "l" in "little" and the "l" in "all." The latter sound fainter than the former. If the former is still velarized, then how is pronounced?LakeKayak (talk) 02:47, 7 January 2017 (UTC)

I think I might know my real issue. The page Dental, alveolar and postalveolar lateral approximants says that the "clear l" appears in New York City English in the word "let", with a citation from Wells provided. I live within the New York metropolitan area, and I may have a slight influence from the accent.LakeKayak (talk) 03:16, 7 January 2017 (UTC)

Yes, that might be the case with NYCE, which has been highly influenced by foreign languages like Italian, Yiddish, and, more recently, Spanish. Also, in NYCE, a syllable-final /l/ may even be dropped. In my own American accent and that of most others I hear, little is [ˈɫɪɾɫ̩]. Also, in my particular accent, all is [ɒ(ə)ɫ]. Wolfdog (talk) 21:39, 7 January 2017 (UTC)

@Wolfdog: Do you happen to have an audio recording of a speaker with the velarized initial "l"? I want to see if I can hear the difference.

Yes I can point you in the right direction and, again, you shouldn't be hearing any major difference. Here is a whole list of English speakers saying "little". All the American (and Canadian) speakers show some degree of velarization in both "l" sounds with perhaps the most velarized initial "l" sounds represented by speakers like ijarritos, Matt3799, rbedsole, and SpanishKyle. For user Wasch, you can in fact hear some difference between the two, but the initial "l" is still not strictly "clear-sounding" and the difference is still nowhere near as extreme as Australian speaker faye2 or UK speaker gcarter. Hope that helps. Wolfdog (talk) 13:15, 8 January 2017 (UTC)

I wouldn't even go as far to say all of the speakers have some degree of velarization of the initial "l." SeanMauch seems to show little to no velarization. Either way, I think I can hear the difference. Thank you.LakeKayak (talk) 20:08, 8 January 2017 (UTC)

I strongly disagree - you can clearly hear velarization of the initial "l" on his recording. Mr KEBAB (talk) 23:57, 9 January 2017 (UTC)

I think there is a difference between the initial and final l in SeanMauch's recording. Perhaps the first l is less velarized, or perhaps the last l is vocalized (i.e., the tip of the tongue is not touching the alveolar ridge). I find when I analyze my own speech, (syllable- or word-) final ls often don't have the tongue touching the alveolar ridge. I like to think of myself as understanding the IPA, but I'm not totally sure how the sound would be transcribed. If it's [ɰ] (a velar approximant) after a vowel, then perhaps it's [ɯ], the vocalic equivalent of that sound, in little. But there may be lip-rounding, and it may be opener. — Eru·tuon 00:16, 10 January 2017 (UTC)

The very initial portion of the initial /l/ to me sounds pretty strongly velarized. Perhaps it gets gradually more 'clear' because of the relative frontness of the vowel it precedes. Mr KEBAB (talk) 00:19, 10 January 2017 (UTC)

I also definitely disagree. The SeanMauch recording is certainly one where both l sounds are noticeably velarized, though the second one perhaps more so. Wolfdog (talk) 00:36, 10 January 2017 (UTC)

It can be noted that I am not the best analyzer. However, I do thank you all for your help.LakeKayak (talk) 16:59, 10 January 2017 (UTC)

Mythical California accent

This article attempts to perpetuate the myth of the California accent as the source of the "General American" accent. It is a myth Hollywood enjoys contributing to but it is untrue. Californians have historically had Western accents (Howdy y'all). Hollywood and NYC promoted the central Midwestern accent as a standard accent and this rapidly became the standard in broadcast and film (indeed media outlets deliberately hired a lot of people out of the Iowa, Nebraska, etc. specifically for their accents). Over the years regional accents have been fading around the country, moreso in the South and West and less so in the Northeast. California in particular, because of its rapid immigration during the 20th century has seen more of a dulling of the native accent over the years than some other regions. But you still hear the native accent out in the rural areas (and, yes, people in California still use y'all out in the country and in some of the suburbs, though not as much as in the past).

This myth should not continue to be promoted.

--MC — Preceding unsigned comment added by 141.131.2.3 (talk) 19:15, 6 February 2017 (UTC)

@141.131.2.3: Do you happen to have a source to support your claim? If so, it will be possible to alter the article in order to tell the story with the most accuracy. If not, you only would need to find one. Either way, this seems to be an easy fix.LakeKayak (talk) 01:00, 7 February 2017 (UTC)

Hmmm, I've read about this before but I'd have to go hunting to find it. I largely speak from having lived in California and know something about the culture. Either way, though, no sources are provided backing up this myth either.

Thanks. --MC

I completely agree with LakeKayak. If your claim is true, I've never heard of it. I'm aware that rural Californians have more Midland and Southern features, but the claim that across the board "Californians have historically" used "howdy y'all" (which, by the way, is most associated with Southern not Western accents) is new to me. The claim as it stands on the page right now is that Californians inherited Midwestern speech patterns, which also definitely needs verifying. But your idea that Hollywood alone promoted Midwestern speech also requires verification; why was this accent then "chosen" as the media standard over others? Wolfdog (talk) 18:27, 7 February 2017 (UTC)

Vowel charts

I'm about to replace the current vowel charts with [1] and [2]. The problems with the current charts I have are:

• /ɪ/ is somewhat too front;
• /ɛ/ is somewhat too close;
• /ɔ/ is definitely too close for a normal /ɔ/ (even the source says so!);
• /ɑ/ is too back and perhaps too close for an average speaker with the cot-caught distinction;
• The ending point of /oʊ/ may be too back;
• The starting points of /aɪ/ and /aʊ/ are definitely too close, they are more like pre-fortis /aɪ, aʊ/ found in speakers with Canadian raising.

The vowel charts I'd like to use are from Wells's Accents of English. They're not perfect either, but they're way better than the current charts. Revert me if there are any objections, but I think my reasoning is pretty good. Mr KEBAB (talk) 05:24, 15 July 2017 (UTC)

Per a recent discussion on my talk page, I feel a need to further explain the post above. What I wrote about /ɪ/ and /oʊ/ is based on my subjective auditory impressions of the typical phonetic values of these vowels in GA. However, what I wrote about /ɛ, ɔ, ɑ, aɪ, aʊ/ is discussed by Wells and some other sources:

/ɛ/:

• Wells (1982:485) says that GA /ɛ/ is somewhat opener than the corresponding RP vowel. In the previous volumes, he says that the RP vowel is typically mid [e̞], so that must mean that the GA vowel is open-mid [ɛ], otherwise he'd be talking about too small a difference to mention it. This is confirmed by his vowel chart.

/ɔ/:

• Wells (1982:476) says that it is lightly rounded open-mid back [ɔ], which is opener and less rounded than the corresponding RP vowel, which in the previous volumes he describes as mid [ɔ̝]. This is confirmed by his vowel chart and by Gimson's Pronunciation of English (8th ed., page 129).

/ɑ/:

• Wells (1982:476, 483) says that for most Americans /ɑ/ is central or perhaps somewhat backer than that. On the vowel chart, he puts it in the open central position.

/aɪ/:

• Wells (1982:487−488) says that /aɪ/ generally has an open front starting point, but it can be central instead. On the vowel chart, he puts it between the front and the central position.

/aʊ/:

• Wells (1982:488) says that the starting point of /aʊ/ is in practically the same phonetic range as the starting point of /aɪ/ and that [æʊ] is becoming increasingly common. The last statement is confirmed by Kretzschmar, Jr. (2004:266, in A Handbook of Varieties of English).

Last but not least, let's not forget that vowel charts don't have just one purpose, which is to tell you how exactly (well, up to a point) vowels of a certain accent are pronounced. The other purpose is just showing you vowel systems of accents, regardless of whether the positioning of the vowels is phonetically accurate or not.

Considering all that and the fact that General American is in fact an American-Canadian accent (rather than just American), I feel that the closeness of the starting points of /aɪ, aʊ/ is one of the most important reasons for which we should opt for the 1982 charts rather than the 1993 (not 2009, that's not the original year they were made) ones. When people unfamiliar with the IPA come here from Canadian raising and see the starting points of /aɪ, aʊ/ in a place where [ɐ] resides on the official IPA chart, it can surely confuse them. After all, we and sources such as Boberg (2004) say that one type of the raised allophones of /aɪ, aʊ/ are [ɐɪ, ɐʊ]. I'm not saying that speakers that don't use Canadian raising in their speech can't realize /aɪ, aʊ/ as [ɐɪ, ɐʊ]. That may or may not be true, but even if it is true, that's definitely not the only possible realization.

Let's also not forget that you can only show so much on vowel charts. You can't account for all the phonetic and regional variation, but rather just illustrate typical, general phonetic values of vowels.

I'm really curious whether Geoffreybmx can read cardinal vowel charts or interpret formant values in Labov's paper with any accuracy. I'm weak with formants, so I don't bring them up. Mr KEBAB (talk) 13:38, 1 August 2017 (UTC)

/ɔː/ in General American

Hi all. Mr KEBAB and I have been discussing /ɔː/, or the THOUGHT vowel, in General American. Wells indeed represents it with the phoneme /ɔ/, though I don't think I've ever heard any American use that as the actual realization, myself a lifelong American. I've heard [ɑ] (for speakers with the cot-caught merger), [ɒ~ɒə] (my own personal realization and common to many speakers who have the merger in perception but not production, or who have no merger at all, like myself), and a definitely diphthongal [ɔə] (as heard by speakers in the Northeast -- New York City, Philly, Connecticut, Rhode Island, etc. by otherwise GA speakers -- and in older US speakers elsewhere). The sound [ɔ] for THOUGHT, though, comes across to me as strictly British-sounding. Obviously, these are all only my impressions (strong ones), so the important question is: Are there any sources to confirm any of these as GA realizations? Wolfdog (talk) 21:23, 8 August 2017 (UTC)

@Wolfdog: I don't think Open-mid back rounded vowel.ogg is correct, the vowel on it is too rounded and perhaps too high to be a correct audio for the cardinal [ɔ]. Indeed, it's quite close or perhaps even identical to the RP /ɔː/, which is a strongly rounded mid back vowel, auditorily closer to the cardinal [o] than [ɔ]. Wells says that the GA vowel is lower and less rounded than the corresponding RP vowel.

Wells does discuss the [ɒ] realizations, but I don't think they fall under what is described as General American:

Turning to the question of the realization of /ɑ/ and /ɔ/, we can say that where they contrast in the GenAm area /ɑ/ is usually open (retracted) central unrounded (much as RP /ɑː/), while /ɔ/ is a lightly rounded half-open back [ɔ] (and therefore opener and less rounded than the usual RP /ɔː/). In the north-central area, however, /ɑ/ tends to be fronted to [a̠] or even [a]. The THOUGHT vowel, in turn, is often as open as [ɒ], particularly away from the Atlantic coast. (In eastern New England, though, it is [ɒ], with loss of the LOT-THOUGHT opposition.) But in Philadelphia and Baltimore, as also in New York City, it is no opener than [ɔ], and is well rounded.

The quote (slightly changed/shortened) is from AoE, p. 476. The second (Bostonian) [ɒ] is obviously non-GA, but the first one is more controversial. I thought that Wells describes a part of the Northern Cities Vowel Shift (because /ɑ/ as [a] is a part of it), but perhaps I was wrong. So what is it?

I don't deny that /ɔ/ as [ɒ] is a part of GA. Most probably, it is. The problem is that the source may not be the best one to determine that. If anyone can interpret formant values, please check Labov and tell us what he says about this issue. Mr KEBAB (talk) 21:45, 8 August 2017 (UTC)

I don't know whether we should preoccupy ourselves with the monophthong-centering diphthong distinction. When you analyze the GOAT vowel as a tense mid monophthong, /ɔ/ can be conveniently analyzed as a lax mid monophthong, and lax monophthongs in GA have centering allophones in some positions (see Wells), not only in GA but also in some speakers in England (see Gimson 2014). IMO it's just a rather predictable phonetic variant. Mr KEBAB (talk) 23:55, 8 August 2017 (UTC)

Wells specifically seems to describe the American (unmerged) THOUGHT vowel in a way that could be represented as [ɔ̞]. Your quoted text shows that Wells indeed does mention the [ɒ] variant; do you feel he's only referring to the north-central U.S. region? I'd disagree, since Wells again brings up and is consenting of [ɒ] or [ɒː] as a GenAm variant explicitly in a blog post of his here; he admits to not using that variant in the Longman Dictionary merely due to [ɔː] being a convenient way to represent both AmE and BrE. Wells in the blog further states of the GenAm vowel that "we could decide to write it ɒ", though he fears "that would create confusion" in respect to traditional RP phonemes. He only seems to fear using that symbol in situations of GenAm/RP comparisons; otherwise, however, it's phonetically accurate. Wolfdog (talk) 12:34, 9 August 2017 (UTC)

@Wolfdog: Woah, let's not conflate what different sources say on this topic. The fact that they were written by the same author is only somewhat relevant, they're still different sources and we must treat them as such.

The problem I have with [ɒ] as described by the AoE is that I have a strong impression that Wells describes features of the NCVS and the Boston accent, nothing more. Look at how it's worded - [a] is a strongly non-GA realization of /ɑ/.

We'd be walking on shaky grounds if we used Wells's blog to source anything here. Someone could come, remove it and we wouldn't really be able to say a word about it since it'd be in a perfect alignment with the sourcing rules of WP. Using it to back up what AoE is saying is also unacceptable. However, if you or someone else could check Sounds Interesting (2014) or Sounds Fascinating (2016), which are books that are basically collections of posts from Wells's blogs, that'd be great. If you find that post in one of those books, that'll solve our problem.

I checked Labov (2006:108), and his map (which doesn't differentiate between cot-caught and non-cot-caught merged speakers) states that the typical height of /ɔ/ in the US has the F1 values between 937 and 674 and so it's most likely in the [ä ~ ɑ ~ ɒ (~ ɔ)] range (rounding can't be reliably determined from formant values alone). The raised THOUGHT, typical of New York, has the F1 value between 674 and 520 and it's most likely in the [ɔ ~ o] range. This may be what we're looking for. I'm pinging @Aeusoes1: to help us, since I suck with formants.

Notice that Wells says in his blogpost that he transcribed the GA vowel with ⟨ɒː⟩ in the first edition of LPD. @Nardog: has it, maybe he could check whether Wells elaborates on his choice in the book. Then again, a quote from the third edition would be much better, as I know of at least one instance of an incomplete/outdated explanation from the first LPD - see this discussion, search for 'There's something wrong with that quote'.

Longman Dictionary of Contemporary English also uses ⟨ɒː⟩ for the GA vowel. But that's not good enough, the fact that a certain vowel is transcribed with a certain symbol in phonemic transcription gives us only a very rough idea on its phonetic realization. Mr KEBAB (talk) 03:03, 10 August 2017 (UTC)

The fact that two sources were written by the same author is absolutely relevant. What if an author published a text that said "Some information in an older text of mine is now outdated or discredited"? I'm baffled that you can think the author is not relevant to a source, even though I understand that blogs are frowned upon as sources. I don't actually expect us to quote and cite the blog; I'm just showing you that an obvious and respected expert in the field approves of the vowel. If you just wanted better sources, that would be one thing: we'd be on the same page in our search. Instead, though, we seem to continue arguing about the vowel itself. I did look at the AoE wording by the way. Why would he say "particularly away from the Atlantic coast", when the north-central area is already defined as being away from any coast? I believe he's saying that the unmerged vowel everywhere is "often as open as [ɒ]", while along the (North) Atlantic coast, it is higher, starting somewhere in the vicinity of [ɔ]. I know Sounds Interesting or Sounds Fascinating, but they are basically lists of the pronunciation quirks of certain words, not entire lexical sets, so I can't see their helpfulness for our situation (which is about GenAm allophones of /ɔː/ of course). Thanks for involving Aeusoes1, since I too can't really work in numbers. I appreciate your trying to find other sources to get to the bottom of this. In my opinion, [ɒ] is a completely common GenAm option (even if not the most common, though it well could be -- of course I couldn't possibly prove either concept) and so should be included in the possible realizations, whereas [ʊə] or [ɔo] or the British-sounding [ɔː~oː] variants, for example, would mark someone as a non-GA speaker. (By the way, do you also find to be incorrect? To my ears, it sounds the exact way I personally pronounce the THOUGHT vowel. Also, interestingly, the page Open back rounded vowel presents [ɔ̞] as an alternative representation for [ɒ].) Wolfdog (talk) 22:29, 10 August 2017 (UTC)

I'm watching this thread but I'm not sure what contributions I can make to the conversation. I don't have personal experience with that vowel (I'm from California) and I don't have additional sources to bring to the table. Wells seems like a trustworthy source to rely on, though it wouldn't hurt to find a broad base of phonetics sources. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 21:14, 11 August 2017 (UTC)

@Aeusoes1: I was hoping you could help me decipher Labovian formants. I have no idea whether my interpretation is correct. Mr KEBAB (talk) 16:19, 12 August 2017 (UTC)

@Mr KEBAB: To answer your request, here's what LPD1 says:

There is considerable variability in GenAm vowels in the open back area. LPD follows tradition in distinguishing the vowel of lot lɑːt from that of thought θɒːt. (Note, though, that other books generally use the symbols ɑ, ɔ, or ah, ɔh respectively.) However, some Americans do not distinguish these two vowel sounds, using the same vowel sound in both sets of words; so a secondary AmE pronunciation with ɑː is given for all words having ɒː. LPD also makes provision of a difference between the vowel of thought θɒːt and that of north nɔːrθ (the symbol ɔː, implying a closer quality, like RP ɔː, being shown for AmE only after r); but many speakers do not make an appreciable difference between these qualities. (p. xiv)

I think "appreciable" is the key. In fact a GA back vowel chart on p. xvi shows /ɒː/ (unmerged THOUGHT) exactly halfway between the cardinal [ɒ] and [ɔ], while /ɔː/ (NORTH) is the cardinal [ɔ].

I don't understand how AoE's description of the THOUGHT vowel being [ɒ] could be interpreted NOT as of the NCVS. He says "in turn", so clearly the two sentences are describing the same accent(s) (the map on Northern Cities Vowel Shift covers Connecticut and upstate New York, so "particularly away from the Atlantic coast" is not strange at all).

As for the blog post, I think he, in response to the question, is talking about how the GA THOUGHT is opener than the RP counterpart (which is closer than the cardinal [ɔ]), not in comparison to the cardinal [ɔ] itself.

So it seems to me both [ɔ] and [ɔ̞] are totally acceptable descriptions of the GA THOUGHT, but definitely not [ɒ]. [ɒ̝] is acceptable as well, but THOUGHT is most often transcribed with ⟨ɔ⟩, so [ɔ̞] is more reasonable. Nardog (talk) 13:52, 12 August 2017 (UTC)

I'm on board with everything you say until you somehow conclude that what's acceptable includes [ɔ̞], but "definitely not [ɒ]." Huh?? According to your own investigation into this, [ɒ] (or more precisely, [ɒː], it seems) is perfectly acceptable... or, at the very least, very much within the "acceptable" range. Once again, for example, you show how Wells is entirely approving of [ɒː] in the LPD. If you prefer the transcription [ɒ̝], I'd be content with using that as one of the presented realizations on the article page. Can this we reach that agreement, at least? If not, I'll yield the argument, since the consensus seems against me. Wolfdog (talk) 15:13, 12 August 2017 (UTC)

LPD1 uses ⟨ɒː⟩ only in phonemic transcription, and we're deciding what symbol to use to represent the phonetic realization of the GA unmerged THOUGHT vowel. Nardog (talk) 15:39, 12 August 2017 (UTC)'

This is exactly the argument I've been making, but about /ɔː/, which is Wells's typical phonemic transcription for THOUGHT in AoE and later LPDs. He at least once openly admits that "In the first edition of LPD I actually represented the AmE THOUGHT vowel as ɒː, differently from BrE ɔː" but "I decided to use the symbol ɔː for both varieties. Apart from anything else, this makes for simpler entries, since the same transcription of words like thought θɔːt and law lɔː will do for both BrE and AmE." As you would agree, his words here are all about phonemic rather than phonetic transcriptions. At the same time, his words about avoiding [ɒ] as a phoneme in themselves suggest, however, that it is certainly one possible phonetic realization. He never says anything about avoiding [ɒ] because it is imprecise or illogical, just that it is a less convenient phonemic representation. Anyway, I'm fine with dropping the conversation on the grounds that we've found no positive/citable phonetic evidence of [ɒ].Wolfdog (talk) 16:08, 12 August 2017 (UTC)

@Wolfdog: You're quite wrong in thinking that what I'm arguing about is the vowel. No, that's not it - it's the sources and the kind of sources we can and should use here. Actually, per WP:SPS we can safely cite Wells's blog. Maybe it's a new policy, I'm not sure, but I could've sworn it said something else. Mr KEBAB (talk) 16:47, 15 August 2017 (UTC)

@Mr KEBAB: I'm not sure what to make of your realization. So do you think we can now present [ɒ] and cite the blog as its source? Wolfdog (talk) 16:36, 19 August 2017 (UTC)

@Wolfdog: Yes. Not only that, we can also pretty safely use blogs of Geoff Lindsey, Jane Setter, Peter Roach, Alex Rotatori and others as sources. This is good news. Mr KEBAB (talk) 16:43, 19 August 2017 (UTC)

@Mr KEBAB: OK, I've added the vowels and blog source. Please make any other changes you feel are appropriate. Wolfdog (talk) 23:10, 19 August 2017 (UTC)

@Wolfdog: Thanks, but I'm not sure if we want both [ɒ] and [ɔ̞] in the table. I'd go for the latter because the common phonemic symbol is /ɔ/. Soon I'll write a lengthy (or not that lengthy, we'll see :P) joint reply to you and Nardog. For now, I'm just curious what is the reason for including [ɔə] as a GA pronunciation? Mr KEBAB (talk) 23:57, 19 August 2017 (UTC)

@Mr KEBAB: Yes, of course, I forget that [ɒ̝] and [ɔ̞] are really the same. I think the diacritics are important, so long as the audio on Wikipedia remains potentially misleading. As for [ɔə], I thought that was included in the AoE, but I see it really was a regionalism of just Philly/NYC. I'll make adjustments. Wolfdog (talk) 00:17, 20 August 2017 (UTC)

@Nardog: Thanks again for the quote. Sorry for not replying earlier, I'm finishing my research and will soon expand upon my answer (which is incomplete). Mr KEBAB (talk) 16:47, 15 August 2017 (UTC)

Any vowel symbol with no diacritic, when describing the exact phonetic quality of a vowel, is by default taken to be that of the cardinal ones, correct? So just [ɒ] with no diacritic would be too open. This is what I implied the last I replied, and I just assumed Mr KEBAB would make this point so I didn't bother.

I also would like to reiterate that Wells' post is simply describing the AmE THOUGHT to be opener than the RP THOUGHT, but not necessarily than the cardinal [ɔ]. And as we know, the RP THOUGHT is closer than the cardinal [ɔ], sometimes rather closer to [o], so I just don't understand how the post could be used as a source to support the transcription [ɔ̞], regardless of whether a blog could be used as an RS or not (if anything, LPD1 p. xvi is such a source we could use to say it's [ɔ̞]). Nardog (talk) 06:31, 20 August 2017 (UTC)

@Nardog: Sorry, I tried to do too much in too short a period of time. Apparently I shouldn't try to multitask and give empty promises. This is going to be a rather long post, but that's what I said I'd write. Anyway:

Thanks for changing the source to AoE. Then again - isn't it at odds with what Wells writes in the third volume and our vowel chart? Just saying...

Any vowel symbol with no diacritic, when describing the exact phonetic quality of a vowel, is by default taken to be that of the cardinal ones, correct? Yes, that or a vowel similar in quality to the respective cardinal vowel. We need to remember that the cardinal vowel system is something that is obviously imperfect and to a certain extent purely subjective (unless we're talking about Luciano Canepari who obviously has a godlike hearing, no? ;))

So just [ɒ] with no diacritic would be too open. See, this is controversial. Acoustically, the cardinal [ɒ] itself is a near-open vowel between [ɑ] and [ɔ]. By that logic [ɒ] is a narrow transcription of GA /ɔ/. But that clashes with the cardinal vowel chart which gives the same space for rounded and unrounded vowels, instead of near-front to back and close to near-open for rounded vowels and front to near-back (front to back in the case of the fully open cardinals [a, ɑ]) and close to open for unrounded vowels. That is defendable because the cardinal vowel chart itself (as we know it today) isn't an accurate representation of where exactly vowels are made in the mouth, just an abstraction. But we should remember that the acoustic difference between the cardinals [ɒ] and [ɔ] is considerably smaller than the difference between the cardinals [ɔ] and [o]. This is partially because [ɒ] uses the same or very similar type of rounding as [ɔ], whereas [o] does not.

Also, there are varieties of English in which /ɒ/ is realized as an unrounded sulcalized vowel akin to RP /ɜː/. In that case, there's hardly any justification to write anything but [ɑ] in narrow transcription. After all, what difference is there between a sulcalized [ɑ] and a truly back [ɑ] you can hear in broad Cockney and some accents of Netherlandic Dutch? None. Exactly. The symbol [ɒ] is probably as useless as [ɶ, ɘ, ɜ, ɞ], you can pretty much always write [ɔ] or [ɑ] instead of [ɒ].

I took a look at 22 vowel charts on Commons which contain a (near-)open back rounded vowel or a (near-)open back vowel with variable rounding (some of the following are actually near-back, I obviously omit that for simplicity). Here's a list of them, along with the symbols they're usually transcribed with:

• Chart 1: File:Abercrave English monophthongs chart.svg, symbols: [ɒ, ɒː], type: near-open rounded
• Chart 2: File:Closing diphthongs of Cockney on a vowel chart.svg, symbol: [ɒʊ], type:near-open rounded (a diphthong onset)
• Chart 3: File:Dangme vowel chart.svg, symbol: [ɔ], type: near-open rounded
• Chart 4: File:Danish monophthongs chart.svg, symbol: [ʌ], type: near-open rounded
• Chart 5: File:General South African English vowel chart.svg, symbol: [ɒ], type:near-open rounded
• Chart 6: File:Geordie monophthongs chart.svg, symbol: [ɒ], type: open rounded ([ɒː], the PALM vowel has a variable rounding, the LOT vowel [ɒ] is consistently rounded)
• Chart 7: File:Hungarian vowel chart.svg, symbol: [ɑ], type: near-open rounded
• Chart 8: File:Ibibio monophthongs chart.svg, symbol: [ɔ], type: near-open rounded
• Chart 9: File:Igbo vowel chart.svg, symbol: [o̙], type: near-open rounded
• Chart 10: File:Khmer vowel chart.svg, symbol: [ɔ], type: near-open rounded (the chart is unsourced)
• Chart 11: File:Maastrichtian Limburgish monophthong chart.svg, symbol: [ɔː], type: near-open rounded
• Chart 12: File:Orsmaal-Gussenhoven Dutch monophthongs chart.svg, symbols: [ɒ, ɒː], type: near-open rounded
• Chart 13: File:Plautdietsch vowel chart (Canadian Old Colony dialect).svg, symbol: [ɔ], type: near-open rounded (nowadays this often merges with /a/ to [ɐ ~ ɑ])
• Chart 14: File:RP English monophthongs chart.svg, symbol: [ɒ], type: near-open rounded
• Chart 15: File:Scouse monophthongs chart.svg, symbol: [ɒ], type: open rounded
• Chart 16: File:Swedish monophthongs chart.svg, symbol: [ɑː], type: near-open weakly rounded
• Chart 17: File:Thai vowel chart (monophthongs).svg, symbol: [ɔ], type: near-open rounded
• Chart 18: File:Ulster Irish vowel chart.svg, symbol: [ɔː], type: near-open rounded
• Chart 19: File:Urban East Norwegian vowel chart.svg, symbol: [ɔ], type: near-open rounded
• Chart 20: File:Urban East Norwegian diphthong chart.svg, symbol: [ɔy], type:near-open rounded (a diphthong onset)
• Chart 21: File:Valencian vowel chart.svg, symbol: [ɔ], type: near-open rounded
• Chart 22: File:Yoruba vowel chart.svg, symbol: [ɔ], type: near-open rounded

Here are the results:

• The symbol [ɔ] was used 11 times;
• The symbol [ɒ] was used 7 times;
  • It was only once used for a non-English dialect/language (the Orsmaal-Gussenhoven one). The rest 6 cases are all English dialects (Abercrave, Cockney, General South African, Geordie, RP, Scouse).
  • Only in two cases (Geordie, Orsmaal-Gussenhoven) was there a need to use [ɒ] instead of [ɔ]. On the other charts (General South African, RP, Scouse) we have [oː] mistranscribed as [ɔː] (or at least an in-between 'true-mid' vowel that is better transcribed [oː] for simplicity), and in other cases (Abercrave, Cockney) we can simply write [ɔ, ɔː, ɔʊ] without any problems.
• The symbol [ɑ] was used 2 times;
• The symbol [ʌ] was used 1 time;
• The symbol [o̙] was used 1 time.
• The vast majority of these vowels (in fact almost all of them) are near-open, not fully open. Some of them are slightly more on the open side and some others are slightly on the close side, but according to the principles of the IPA the symbols ⟨ɔ⟩ and perhaps ⟨ɑ⟩ should be preferred to ⟨ɒ⟩ because they are simpler.
• Vowels that are transcribed with [ɒ] never contrast with an open-mid vowel but always a mid or a close-mid one (in the case of Geordie and the Orsmaal-Gussenhoven dialect, both mid and close-mid vowels are present). This is a further argument to write [ɔ].

Placing English vowels closer to the cardinal [ɒ] than the cardinal [ɔ] may be not only a result of the cardinal vowel system being subjective but also of an influence of British English accents themselves on phoneticians' judgement. Read this interesting article: [3]. Search for 'The chief disadvantage of impressionistic phonetics' if you don't want to read the whole thing.

Here's a Geoff Lindsey's article on which I based much of what I wrote: [4].

And as we know, the RP THOUGHT is closer than the cardinal [ɔ], sometimes rather closer to [o], so I just don't understand how the post could be used as a source to support the transcription [ɔ̞] To me, that post isn't ambiguous at all. To write [ɔ̞] to mean [ɔ] would be considerably redundant and simply wrong in narrow transcription (which Wells used in that particular case). To me, the post unambiguously states that RP /ɔː/ is above open-mid and more rounded than the corresponding cardinal vowel, whereas GA /ɔ/ is below open-mid and less rounded than the cardinal [ɔ]. Don't forget that he also wrote In the first edition of LPD I actually represented the AmE THOUGHT vowel as ɒː, differently from BrE ɔː, which would have pleased Bao Zhi-kun. Why would that please his reader if the GA vowel were open-mid?

You could bring up now how Danish is usually transcribed in the IPA, but that's a totally different thing and it's controversial whether we should even call it IPA (though it obviously uses IPA symbols, just in a strange and inconsistent manner and even Basbøll himself doesn't give a very good justification for that).

I'll try to respond to your previous message (and the Wolfdog one), but for now this is already long. I hope I'm making some sense... Mr KEBAB (talk) 14:38, 7 September 2017 (UTC)

Hmm, apparently I didn't make complete sense in the message above. I mean this: That is defendable because the cardinal vowel chart itself (as we know it today) isn't an accurate representation of where exactly vowels are made in the mouth, just an abstraction. doesn't make a lot of sense because the cardinal [ɒ] is made exactly where the cardinal [ɑ] is. The fact that it's higher than [ɑ] on formant charts is the direct result of it being rounded. But still - I've proved that the near-open variant is far more common in world's languages and I still think that it would make at least some sense if we placed the cardinal [ɒ] above the cardinal [ɑ]. Mr KEBAB (talk) 03:13, 8 September 2017 (UTC)

I also wanted to say that maybe parts of my rant would be better suited for Talk:Open back rounded vowel or Talk:Cardinal vowels. Maybe I posted it here prematurely, I'm not sure. Mr KEBAB (talk) 13:21, 8 September 2017 (UTC)

IPAc-en template

Why would we not use the IPAc-en template when it is the standard for all of Wikipedia? Wolfdog (talk) 16:00, 20 August 2017 (UTC)

@Wolfdog: It's quite a stretch to say that. We use it mainly in non-linguistics-related articles, in which it indeed is standard, but if someone needs the IPAc-en tooltips it's a sign that they won't be able to read this article with any fluency anyway. I'm for being noob-friendly but only when it has no consequences. I'm against using this template because it introduces inconsistencies in phonemic transcription. GA doesn't have phonemic vowel length, but if you use IPAc-en you're forced to write e.g. /iː, uː/ instead of /i, u/. Plus, marking length in phonemic but not phonetic transcription has hardly any justification as far as this accent is concerned. Mr KEBAB (talk) 16:15, 20 August 2017 (UTC)

@Mr KEBAB: Weird. Well, OK, whatever you think. Wolfdog (talk) 19:12, 20 August 2017 (UTC)

@Wolfdog: I'm not sure what's weird in my reasoning, but I won't press the issue. Mr KEBAB (talk) 19:37, 20 August 2017 (UTC)

@Mr KEBAB: No, sorry. I meant, it's weird how we use the IPAc-en template. Weird that we have this whole template and then don't use it consistently. But if that's how it's used in your experience, then I'll trust your judgment. Wolfdog (talk) 21:07, 20 August 2017 (UTC)

@Wolfdog: My bad then, but I don't see why we should prefer it over the simple IPA template in linguistic articles, if that's what you're saying. Again, users that need the tooltip feature will have other problems with this article anyway, it's possible that they won't even be able to understand a large part of it.

I see that you mainly edit articles on American English, and that may be where your idea that IPAc-en is standard came from. I on the other hand don't really edit anything AmE-related besides this article and American English, simply because I find other varieties of English more interesting. There may be an unofficial consensus among editors to use IPAc-en on AmE-related articles, that's probable. But if it introduces inconsistencies in transcription (like in this article), then it's questionable whether we should use it in other articles as well. Maybe you could check some of them when you have some spare time.

You don't even have to believe me, here are articles in which IPAc-en is used either sparingly or not at all: Received Pronunciation, Australian English phonology, New Zealand English phonology, Cockney and Scottish English. Mr KEBAB (talk) 22:51, 20 August 2017 (UTC)

You generally wouldn't want to use IPAc-en in linguistics articles unless maybe you are talking about English diaphonemes. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 03:17, 21 August 2017 (UTC)

@Aeusoes1: That's how we use it in the articles I mentioned. Mr KEBAB (talk) 08:32, 21 August 2017 (UTC)

I always assumed it was more for comparing dialects or generic English articles in which precise dialect was unimportant, rather than American English-specific articles. There are certainly elements of IPAc-en that make little sense to American English dialects, and I'm sure that's exactly your point. Thanks. Wolfdog (talk) 14:22, 21 August 2017 (UTC)

General American primarily spread via Midwesterners, via Californians, or via some other means?

I was wondering whether anyone could stir up any sources to help either lend credibility to the following sentence or else unearth the reality of how General American "spread":

The fact that a broad "Midwestern" accent became the basis of what is General American is often attributed to the mass migration of Midwestern farmers to California from where the accent system spread,[citation needed][disputed – discuss] since California speech itself became prevalent in nationally syndicated films and media via the Hollywood film industry.[disputed – discuss]

The sentence has remained without much-needed citations for a while (as other frustrated users have pointed out), and, though its claim seems reasonable to me, it could just as likely be completely inaccurate. Does anyone know of any sources that could verify or counter? Thanks! Wolfdog (talk) 15:10, 23 August 2017 (UTC)

Take it out. It actually sounds a lot like guesswork to me. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 15:57, 23 August 2017 (UTC)

Yes, that's one option. Do we all agree it's total guesswork and unverifiable/unsourceable? And if it is... Is there some source we can still find to discuss how a General American sound became prevalent? Wolfdog (talk) 02:49, 25 August 2017 (UTC)

Kretzschmar (2008:41–3) cites as potential major sources of the emergence of the standardized American speech suburban housing and advanced mobility in the 20th century and educated speakers' tendency to suppress regionally marked features in formal settings, and says "The typical speech of national news broadcasters is symptomatic – not a cause of the change, as many suppose" (p. 42) and "StAmE can best be characterized as what is left over after speakers suppress the regional and social features that have risen to salience and become noticeable" (p. 43). So the sentence does sound like total guesswork and Kretzschmar is probably a pretty good starting point to revise it. (And it also confirms descriptions like "spread" and "became prevalent" aren't accurate when it comes to General American.) Nardog (talk) 04:57, 25 August 2017 (UTC)

Yes, I was thinking about asking if we were even approaching the history of GenAm appropriately by talking about its "spread", since of course GenAm is a collection of similar dialects rather than a single uniform variety that "swept the country". It seems to be a very fine line. I agree that the idea of education and formality are completely relevant and even necessary in discussing the usage of a (real or perceived) GenAm umbrella. — Preceding unsigned comment added by Wolfdog (talk • contribs) 11:32, 25 August 2017 (UTC)

I've made some changes according to what we've discussed. Nardog, thanks for the source. Does anyone know of any reason why the highly educated of the twentieth century began "choosing" a non-coastal Northeastern rhotic sound as their supra-regional accent? Was it simply random that this location's accent was the accent with the least stigma/markedness? Wolfdog (talk) 13:56, 26 August 2017 (UTC)

@Wolfdog: Maybe this post will answer that: [5]. Mr KEBAB (talk) 23:44, 26 August 2017 (UTC)

@Mr KEBAB: A cool article. Thanks for finding it. But it seems to focus more on Hollywood than on the whole country. In any case, its main relevant claim is "As the 20th century progressed, America’s economic and demographic centre of gravity shifted westward. The heartland was largely white, gentile and rhotic, including many with Scots-Irish origins. They may have played a proportionally modest role in the creation of popular culture, but it was in the democratic nature of that culture to assimilate and reflect their speech. The result was that hypo-rhoticity became increasingly marked, in both lower and higher status speech." The reason given for why the "heartland" accent became GenAm, then, is basically due to the American "democratic nature." Hmm, seems a bit vague.

One other maybe more convincing argument is that TV encouraged that "heartland" sound in the media more than film had done: "The new medium of TV certainly reinforced General American speech. Where Hollywood had been a dream factory, relatively cosmopolitan and international, TV was more national and more reflective of real people’s lives, from news coverage to game shows with ‘ordinary’ contestants." Anyway, I think TV and film are the results of accent-spread more so than they are the causes of accent-spread.

Finally, a central argument for the blog post is that "popular culture as it spread from early 20th century America was to a large extent not rhotic" substantially due to the non-rhotic accents of "two ethnic minorities": African Americans and Jewish Northeasterners. This may be true enough, but it's very unconvincing to then make the leap that the white Hollywood elite developed their non-rhotic accent by trying to imitate black and Jewish accents. And of course, this still doesn't explain my original question of why GenAm was rhotic, except to say that most American had always been rhotic speakers. Wolfdog (talk) 14:58, 27 August 2017 (UTC)

Second-sentence change?

Is there any reason to change the second sentence of the article, "Due to General American accents being widespread throughout the United States, they are sometimes, though controversially, classified as Standard American English" to something more like "Due to the perception of General American being widespread throughout the United States, it is sometimes, though controversially, classified/known as Standard American English"? I italicized the word "perception" to show that this is the major addition here. This seems like a logical edit to me, but I was wondering what others thought first. The change from plural ("General American accents") to singular ("General American") also seems cleaner to me, though honestly I also kind of like the plural, because it reinforces the idea that GenAm is not just "one thing". Wolfdog (talk) 14:00, 26 August 2017 (UTC)

I feel like we should just take out the first clause. Either way I bet this is such a minute point we can safely apply WP:BRD. Nardog (talk) 14:22, 26 August 2017 (UTC)

Near-rhyming or rhyming?

@Wolfdog:, what are your arguments/proof for this edit? My argument is that /k/ unambiguously belongs to the first syllable as it clips the preceding sequence of sonorants /ɜr/, just as in RP. The source is LPD. Mr KEBAB (talk) 02:50, 2 October 2017 (UTC)

Hmm, maybe it's variable. The CEPD transcription is /ˈwɜː.kə, ˈwɜ˞ː.kɚ/. But the corresponding RP recording clearly says /ˈwɜːk.ə/ ([ˈwɜkə], with a short vowel). If anything, it's the LPD recording that has a longer vowel, yet it transcribes that word /ˈwɜːk.ə/... Mr KEBAB (talk) 07:35, 2 October 2017 (UTC)

Why don't we use murder or further or something, then? Nardog (talk) 08:54, 2 October 2017 (UTC)

@Nardog: I think stirrer is the best word. Murder and further have exactly the same problem as worker - the LPD places the syllable break before the schwa, whereas the CEPD places it after the NURSE vowel. Mr KEBAB (talk) 09:03, 2 October 2017 (UTC)

@Mr KEBAB: Syllabification of English words is something scholars never seem to agree on, so perhaps we're better off avoiding the word "rhyming" altogether and saying something like "the two vowels in ___ are pronounced with the same quality." Nardog (talk) 09:46, 2 October 2017 (UTC)

@Nardog: Maybe that's a better idea, but IMO stirrer has only one possible syllabification in GA, which is /ˈstɜr.ər/. This is because /ɜ/ doesn't occur outside of /r/ contexts in GA and because it's mandatorily rhotacized in e.g. New York English which preserves the hurry-furry distinction (and even /ˈstʌrər/ would have to be syllabified /ˈstʌr.ər/ because /ʌ/ is a checked vowel). Mr KEBAB (talk) 09:55, 2 October 2017 (UTC)

Ah! Sorry, I didn't realize the heart of the contention here was syllabification. If we want to use a word like "assonance" or just say "the same r-colored vowel" rather the "rhyming" (which depends on the organization of syllables and which, I admit, didn't even occur to me since in my head worker is pronounced "wər-kər" despite being syllabified as "wərk-ər"), that is fine with me. Or if we want to keep the word "rhyming," then I'm happy to use an alternative example word than worker, such as your aforementioned stirrer. "Near-rhyming" was confusing, though, since, it makes it seem as if there is some slight difference with the vowel quality. Wolfdog (talk) 10:54, 2 October 2017 (UTC)

@Wolfdog: Oh, that's what you meant. I'm fine with either option. Mr KEBAB (talk) 16:31, 2 October 2017 (UTC)

Can we use [ɹ̩] instead of [ɚ] by the way? It'd be consistent with other sonorants and we'd avoid using the schwa symbol for a stressed vowel (which may be a bit confusing for readers that are accustomed to the way e.g. Wells transcribes GA). Mr KEBAB (talk) 16:44, 2 October 2017 (UTC)

In narrow transcription, [ɹ̩] might be a good idea. Nardog (talk) 17:43, 2 October 2017 (UTC)

@Nardog: I meant phonetic transcription in general, broad or narrow. Mr KEBAB (talk) 16:32, 3 October 2017 (UTC)

@Mr KEBAB: You mean even phonemic? If so that would allow a stressed syllabic consonant, which is quite unusual. I thought /ɜr, ər/ were the de facto standard for (morpho)phonological reasons. Nardog (talk) 16:50, 4 October 2017 (UTC)

@Nardog: Phonetic. Phonetic transcriptions can be narrow and broad, phonemic transcriptions are broad by definition. Phonemically, these are undoubtedly /ɜr, ər/. Mr KEBAB (talk) 17:35, 4 October 2017 (UTC)

@Mr KEBAB: This is slightly off-topic but I'm curious, who else is using Wells' analysis? The syllabification I've seen in other authors' literature almost universally follows the maximal onset principle like CEPD, though with variation in the treatment of the consonants after checked vowels. If I understand correctly, Wells' analysis is based more on phonetic evidence with emphasis on speakers' production. So just by looking at the phonemic representation /ˈwɝkɚ/ without the knowledge of how the word is actually pronounced, one would not be able to syllabify it with certainty as /ˈwɝk.ɚ/. Which obviously is not to say that such syllabification is wrong, but it is impossible to draw from the (unsyllabified) underlying representation alone or otherwise it would be circular reasoning. And /ɝ/ is not a checked vowel, so I bet the statement "worker is realized with two rhyming syllables" wouldn't be any confusing to most people (and if you're concerned about clipping, we could use murder etc.).

But again, just because Wells is a minority here doesn't mean he's wrong (if anything they're just two different approaches—one from phonetics and the other from phonemics, to put it crudely), so I stand by my previous point that in the article such language is better avoided altogether. But I'd love to know what other scholars think of Wells' approach. Nardog (talk) 17:43, 2 October 2017 (UTC)

External links modified

Hello fellow Wikipedians,

I have just modified one external link on General American. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:

• Added archive https://archive.is/20130202223336/http://www.soundcomparisons.com/Eng/Direct/Englishes/SglLgAmericanStandard.htm to http://www.soundcomparisons.com/Eng/Direct/Englishes/SglLgAmericanStandard.htm

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

This message was posted before February 2018. After February 2018, "External links modified" talk page sections are no longer generated or monitored by InternetArchiveBot. No special action is required regarding these talk page notices, other than regular verification using the archive tool instructions below. Editors have permission to delete these "External links modified" talk page sections if they want to de-clutter talk pages, but see the RfC before doing mass systematic removals. This message is updated dynamically through the template {{source check}} (last update: 5 June 2024).

• If you have discovered URLs which were erroneously considered dead by the bot, you can report them with this tool.
• If you found an error with any archives or the URLs themselves, you can fix them with this tool.

Cheers.—InternetArchiveBot (Report bug) 12:46, 12 October 2017 (UTC)

John Samuel Kenyon

Numerous sources identify John Samuel Kenyon's influence regarding General American. However, I have removed the following paragraph (that I mostly or entirely wrote):

Influential to the "standardization" of General American pronunciation in writing was John Samuel Kenyon, author of American Pronunciation (1924) and pronunciation editor for the second edition of Webster's New International Dictionary (1934), who used as a basis his native Midwestern (specifically, northern Ohio) pronunciation. Ironically, Kenyon's home state of Ohio, far from being an area of "non-regional" English, has emerged now as a crossroads for at least four distinct regional accents, according to late twentieth-century research, and Kenyon himself was vocally opposed to the notion of any supreme standard of American speech.

Does anyone see a better place to put this and/or a better way to abbreviate the information? Or do other editors feel, as I'm beginning to, that maybe this information about him is too much minutiae for the scope of the article and should not be reinstated after all? Wolfdog (talk) 01:23, 6 November 2017 (UTC)

I think it should all stay. John Kenyon's General American accent deserves to be in the regions origins. It was the regional origin of "General American." I don't see how you couldn't mention his name when talking about that. So it makes sense for it to stay. I'd also like to keep the part that John was "vocally opposed to any supreme standard of speech." This goes to show at least he thought it was biased to have any sort of supreme speech over others. If were going to charge racism for General American, then it makes since to include Kenyon was opposed to any supreme standard of american speech, since he did. In the research I've done for this, John Kenyon's General American went out of style back in the middle of the 20th century. No-one in the Inland North region talks like that anymore due to the Northern City Vowel Shift. The thing about adding Bonfiglio's sources is that he seems to think the Inland North accent is still "General American", when others like William Labov, John Wells, Dennis Preston and Van Riper disagree with this. I'm not a source obviously, but I personally agree with those guys over Bonfiglo. William Labov thinks that general american was chosen because it does the best job of representing all of General America. Which mean't outside of the really well known and unique accents of Boston area, New York and the South. I don't think race had anything to do with it. Pre 1945, the New York accent and English accent was the standard. It was known as the mid-Atlantic accent. It adopted some New York features, but mostly English features. After World War 2, Europe was destroyed and America became the big super power. America to this day is the global super power. Before that, it was the British empire that was the big super power. Hence the British standard. I believe that in the 19th century up until 1945 most of the elite in both countries (both the US and UK) spoke with a mid-Atlantic accent. Since the UK wasn't the super power anymore, I think they wanted a generalized american english accent after world war 2. America has always looked down at kings and queens. We were born out of revolting against that. So I think that's why we never got a RP type of accent for america. Instead we just went with sort of a generalized american accent. If people think it's biased to have a "general" accent for things, like Kenyon did at least in the supreme sense, I can see it from that angle. Western New England, Midwest, Canadian English and West are considered by quite a few to be in the General American spectrum. If we do decide to keep Bonfiglio's sourced information, I think we should also add a source that includes Western New England and Canadian English. There are some sources that claim they're part of the General American spectrum, along with the Midwest and West, including on this very page by Van Riper. Usernamebradly (talk) 03:12, 6 November 2017 (UTC)

@Usernamebradly: Please name your actual sources. It's not the first time someone comes here and claims to know things without proving it. Mr KEBAB (talk) 12:24, 6 November 2017 (UTC)

The sources on John Kenyon? Seabrook (2005).

Jump up ^ Hunt, Spencer (2012). "Dissecting Ohio's Dialects". The Columbus Dispatch. GateHouse Media, Inc.

^ Jump up to: a b Hampton, Marian E. & Barbara Acker (eds.) (1997). The Vocal Vision: Views on Voice. Hal Leonard Corporation. p. 163.There were three of them. They were all deleted. it's against wikipedia rules to delete sources without taking it to the talk page. Usernamebradly (talk) 12:36, 6 November 2017 (UTC)

@Usernamebradly: Thanks. Wolfdog has already said that he deleted one of them by accident, so I'm not sure why you keep pushing the issue. Mr KEBAB (talk) 12:41, 6 November 2017 (UTC)

Because they are still removed from the main page. Deleting sourced content is considered vandalism. Usernamebradly (talk) 12:44, 6 November 2017 (UTC)

@Usernamebradly: Maybe instead of calling him a vandal invite him to a discussion? He said so in the edit summary. @Wolfdog: Mr KEBAB (talk) 12:55, 6 November 2017 (UTC)

But that is what it is, Kebab. I have nothing against Wolfdog. But deleting sourced content is vandalism. That's why you have many moderators on wikiepdia that guard pages where people vandalize and delete sourced content. Many even state that it's vandalism in their edit summaries. People can't just go around deleting sourced content because they feel like it doesn't belong. If this was allowed, people would abuse this and edit out whatever they want.. Usernamebradly (talk) 13:01, 6 November 2017 (UTC)

@Usernamebradly: So when someone says they made a mistake they're lying by definition. Ok, I get it. Mr KEBAB (talk) 13:08, 6 November 2017 (UTC)

He stated at the very top of this very talk that he removed it. It wasn't any mistake. I didn't think it should be removed until it was talked about. Usernamebradly (talk) 13:12, 6 November 2017 (UTC)

@Usernamebradly: Read the most recent edit summary: [6]. Mr KEBAB (talk) 13:14, 6 November 2017 (UTC)

He deleted the entire John Kenyon section from the Geographic origins part. This part of the write-up had three sources. Again, it was sourced content that shouldn't of been removed until a discussion was made to remove it. He quotes above in this talk section about the part that he removed. It had three sources. Usernamebradly (talk) 13:18, 6 November 2017 (UTC)

@Usernamebradly: You clearly didn't read the first message thoroughly. I'm starting to have a déjà vu... Mr KEBAB (talk) 13:25, 6 November 2017 (UTC)

I've decided to contact an adim to help with this. You are showing hostility and favoritism against me, which is against wikiepdia rules. I will not be replying to anymore of your messages. Any future messages will be about the removing of the John Kenyon content, which is what this talk topic was created for. Usernamebradly (talk) 13:30, 6 November 2017 (UTC)

@Usernamebradly: :DD I'm eagerly waiting for them. Mr KEBAB (talk) 13:34, 6 November 2017 (UTC)

What in the world is the problem here? I only "deleted" information in the form of consciously and carefully transferring it here (with all the sources intact) to the talk page so that we could have a reasoned conversation. How could this possibly be regarded as "vandalism"? (It's made even further bizarre because I'm being accused of this when I was the one who personally added that exact info and the sources to begin with. Now I'm bringing it to the talk page because I'm doubting whether my own info is relevant. This is honestly the first time I've heard of a call for discussion and collaboration being denounced as "vandalism".) Wolfdog (talk) 14:08, 6 November 2017 (UTC)

@Wolfdog: He's continuously assuming bad faith and trying to get you (and me too) in trouble. I can't see another explanation. The problem is, Usernamebradly, that what we write here is very different from what you're reading. That's why I found it so funny when you took this issue to the admins. Mr KEBAB (talk) 14:20, 6 November 2017 (UTC)

I stand by the fact that the sourced material should not of been taken off the page without it being talked about first. This si the edit where it was removed. https://en.wikipedia.org/w/index.php?title=General_American&diff=808920897&oldid=808918708 The John Kenyon information was sourced material and shouldn't of been removed without it being discussed first. Usernamebradly (talk) 14:47, 6 November 2017 (UTC)

@Usernamebradly: Classic WP:IDIDNTHEARTHAT ;) Mr KEBAB (talk) 14:49, 6 November 2017 (UTC)

When people want to discuss the topic I'm ready. Usernamebradly (talk) 14:55, 6 November 2017 (UTC)

Wolfdog why are you for deleting the John Kenyon section when it is sourced that the Geographical origins of GA come from John kenyon's northeastern ohio? Bonfiglo, on the other hand, still thinks that the Inland North region, along with Western new England are part of common day General American/Inland North region. Some sources disagree with this due to that part of the country undergoing the Northern City Vowel Shift. I actually had a recent email conversation with Bonfiglo on this, as can be seen here https://imgur.com/ZEObIzx Let's all try to not turn this into anymore drama at this point. I'm new at wikipedia. I will probably make mistakes. I havn't read all the rule webpages on wikipedia. i'd like to start this on a better footing. Thanks. I look forward to chatting with you. Usernamebradly (talk) 15:36, 6 November 2017 (UTC)

@Usernamebradly: We're happy to discuss amicably. I've talked and collaborated with Mr KEBAB a few times in the past and decorum has always been maintained even when we've disagreed.

In light of trying to make the GenAm page as non-technical and reader-friendly as possible, I wondered whether the Kenyon information was too much detail and overloading for the scope of this article. After all, why go into such detail about one man who has had some influence, though probably overall marginal, on the topic? Anyway, my whole point here was to have a discussion and get other editors' opinions, so I'm open to all suggestions.

Where do you see that Bonfiglio thinks the Inland North is still a General American region? My take is that he's talking about the history of the topic. Regardless, he is a credible source, so I don't see why we can't include his sociological argument. (The "racism" argument is already a fairly established theory of how Southern American English phonology developed (away from features associated with "black English"), so this kind of sociological thinking is certainly plausible. As I've worded it in the article, it's just one possible factor. We can make that even clear if you'd like.) Wolfdog (talk) 17:22, 6 November 2017 (UTC)

Well, I don't view John Kenyon's influence as marginal. He was the nation’s leading pronunciation expert at the time, as discussed by Edward Mccleland. "Perhaps most importantly, the nation’s leading pronunciation expert was John S. Kenyon, a philologist at Ohio’s Hiram College, which is just east of Cleveland. Kenyon was the author of two books, American Pronunciation (1924) and A Pronouncing Dictionary of American English (1944), and was pronunciation editor of Webster’s New International Dictionary of the English Language. In these roles, Kenyon championed rhoticity, the pronunciation of r’s wherever they appear in words: he preferred “war” to the Transatlantic “waugh.” Kenyon also favored pronouncing “not” like “naht,” instead of “nawt.” These were both features of the Inland North speech he heard in northeastern Ohio." http://beltmag.com/introduction-speak-midwestern/ It is discussed here again by both Edward Mccleland and William Labov https://m.clevescene.com/cleveland/the-origins-and-evolution-of-the-cleveland-accent-yes-you-have-an-accent-cleveland/Content?oid=4949852&storyPage=2 "Indeed, in a 2005 PBS documentary called Do You Speak American, Labov, the linguist who first reported the Northern Cities Shift, pointed out that the region from Rochester to Chicago was the closest thing to television news network standard pronunciation that existed in the U.S.

"It was what the NBC standard was based on," he said.

And it was a standard that, perhaps not surprisingly, had strong ties to the Cleveland area. The man who is credited with creating it, a linguist named John Kenyon, was a professor at Hiram University, just southeast of here. In 1924, Kenyon published the first version of a guide called American Pronunciation, later was a consulting pronunciation editor of the second edition of Webster's New International Dictionary and, in 1944, co-authored A Pronouncing Dictionary of American English.

One constant across all of Kenyon's works was that he was an unabashed proponent of the version of English spoken in the CLE.

"[T]he author has based his observations on the cultivated pronunciation of his own locality — the Western Reserve of Ohio," he wrote in American Pronunciation, referring to himself. "It is his belief, however, that this is fairly representative of what will here be called the speech of the North." He would later simply call this, "General American."" I don't know, it seems like it was John Kenyon that came up with the standards for what General American would be. He thought the speech in his area was most correct. I don't see him promoting any racial component for any of this. More than anything, it seemed like John Kenyon, who Edward McClellan pointed out, was the nations leading pronunciation expert in regards to this stuff back in the middle of the 20th century. Since he held this position, he most certainly was someone with big influence. He was the one that thought up of these standards. He favored pronouncing r's and using naht over nawt. As Labov says, his region was what the NBC Standard was based on. That's not to say there's no racism towards the way other people speak. Well before John kenyon was born there has racism towards certain speech. There probably will be decades from even now. But I'm not seeing any of this in the guy who started "General American" speech after World War 2. Kenyon vocally opposed of any supreme standard of american english. Wolfdog, thanks for the nice reply. I appreciate that freindlyness. Thanks a lot. I apologize for calling you a vandalizer earlier. That was uncalled for.Usernamebradly (talk) 19:26, 6 November 2017 (UTC)

@Usernamebradly: Again, I don't see any reason to remove the Bonfiglio source (unless you can positively identify how it isn't credible), but the rest of your argument about reinserting the Kenyon info seems convincing to me. By the way, when you said that Kenyon "thought the speech in his area was most correct", did you mean to say that Kenyon "didn't think the speech in his area was most correct"? After all, we seem to agree that he didn't believe in any supreme standard Wolfdog (talk) 02:22, 7 November 2017 (UTC)

I agree with your statement your entire statement. I think we should reinstate the John Kenyon stuff and keep the stuff that Bongfilio said. I'm glad we could reach agreement. Usernamebradly (talk) 03:02, 7 November 2017 (UTC)

@Wolfdog: Sorry for the late reply to this, but do you think this part "The shift in American media from a non-rhotic standard to a rhotic one cemented only in the late 1940s, after the triumph of the Second World War, with the patriotic incentive for a more inclusively all-American "heartland variety" in television and radio." is outdated in todays world? It seems to me that the spectrum of GA has expanded since the mid 20th century, to include other parts of the US other than the midwest. I really think this part should be removed. Usernamebradly (talk) 03:29, 23 November 2017 (UTC)

@Usernamebradly: I'm confused by why exactly you want to remove it. How is a statement about history "outdated"? Wolfdog (talk) 11:09, 24 November 2017 (UTC)

@Wolfdog: It's part of "In the Media" section, which today, according to numerous sources, expands beyond the midwestern heartland part of the country. Many of todays big radio talk host are from the Northeastern Coast. Mark Levin, Shawn Hannity, Bill O'reilly, John Stewart, Michael Savage... the list goes on and on. And those are some big names, not randoms. Things change over time, and I really feel like that's something that hasn't been the case in over a half century or more. Long enough that I don't see the point in putting it there. Basically, what I'm getting at, is that it's not the standard anymore. So why put it there? Usernamebradly (talk) 11:35, 24 November 2017 (UTC)

@Usernamebradly: We're talking about the shift in history that made rhotic GenAm the standard in the media. That shift happened in the 1940s, based on the 1940s perception of a heartland variety. The sentence I included is saying that that's where it started, not where it may or may not be used today. Wolfdog (talk) 11:54, 24 November 2017 (UTC)

@Wolfdog: Well, the source says "cemented" (settle or establish firmly), which is not the case in todays world. It's not "cemented" in todays world. I still feel that sentence should be removed. We can agree to disagree. Thanks for the timely responses. Usernamebradly (talk) 12:11, 24 November 2017 (UTC)

@Usernamebradly: Rhoticity is not cemented in American media? Where did you get that from? Mr KEBAB (talk) 16:01, 25 November 2017 (UTC)

@Mr KEBAB: My main problem with that sentence is that the GA of the 1940s actually started from the non-coastal Northeast. And then spread into the upper midwest, Western Pennsylvania and the Mid-Atlantic. GA has never been just a "heartland variety" (which in America means the Mid-west) of English. There are various sources that say this throughout the GA page. From William Labov, Van Riper, Thomas Bonfiglio. Usernamebradly (talk) 19:44, 26 November 2017 (UTC)

@Usernamebradly: Maybe I misunderstood you. My bad. Mr KEBAB (talk) 19:48, 26 November 2017 (UTC)

@Usernamebradly: Actually, now I'm more confused. 1) Would it make you feel better if we changed "with the patriotic incentive for a more inclusively all-American 'heartland variety' " to "with the patriotic incentive for a more inclusively all-American, notional [or supposed or alleged] "heartland variety' "? 2) Are you saying that the GenAm of the 1940s began in the non-coastal Northeast and then literally "spread into the upper midwest, Western Pennsylvania and the Mid-Atlantic" after the 1940s (until the present day, I guess)? If so, I think the Atlas of North American English debunks that; the timeline would have to be quite a bit earlier. Wolfdog (talk) 03:21, 28 November 2017 (UTC)

@Wolfdog: Sorry for the late reply. I recently had a discussion with Bonfiglio, a source you added to this page. GA of the 1940s was originally from the NorthEast. In our discussion he confirmed this. It did spread to certain parts of the mid-west, but it's origins are from the northeast. I've never heard of the NorthEast considered part of the heartland. Heartland is a vague term because there are numerous accents in middle america. The GA of the 1940s is a accent of Northeastern origin. It was changed due to the NCVS. A sterotype is that middle america has one accent, which it doesn't. Also, Labov states that settlers from the Northeast settled the upper midwest. Western New England and Upstate New York still fit in the "Northern" family that Labov defines. They are all part of a Northern Super Region. Usernamebradly (talk) 07:14, 8 December 2017 (UTC)

@Usernamebradly: I agree with, and already have heard of, everything you just said. So is there still an issue? Again, I'm happy to change the language of the article to the option "1)" I mentioned above. Does that appeal to you? Wolfdog (talk) 14:39, 8 December 2017 (UTC)

Also, did Bonfiglio specify where exactly in the Northeast? I'd be willing to bet it was a non-coastal inland region of the Northeast, as the article already says: somewhere in the vicinity of Western New England, inland Pennsylvania, and (northern?) Ohio. Wolfdog (talk) 14:44, 8 December 2017 (UTC)

@Wolfdog: Yes, he mentioned all the way to West New England. I personally think the whole sentence should be removed. Heartland is a vague term, and not entirely accurate. Also. I was originally the person that removed the Jewish American example, as Thomas Bonfiglio mentions various southern (Italian) and Eastern European examples (Slavic) In fact, I think it's pretty safe to say that outside of pockets of ethnics here and there, most immigrants that brought there cultures here, were frowned upon. Regardless of where at the time an european immigrant came from, they were encouraged to americanize. Most German, Polish, or any other ethnic group, has largely Americanized. I say we just leave southern european and eastern European, and not single out any groups like jewish and russian people? Usernamebradly (talk) 06:20, 9 December 2017 (UTC)

@Usernamebradly: I'm OK with your idea of not singling out any groups and just leaving it as "Southern and Eastern European". However, as for removing the "heartland variety" sentence, the problem is that it is still backed by a reputable source (John McWhorter); unless you can give another reputable source that contradicts it, I think leaving it makes sense. Otherwise, we're arguing about what, for the moment, seems to be WP:OR. Wolfdog (talk) 14:05, 9 December 2017 (UTC)

@Wolfdog: GA is not a exclusive heartland accent. Not then and not now. Numerous parts of the US (and even Canada) fall under the umbrella of GA. Labov (sourced on this very page) said the GA (post NCVS) could fall under Western US, Midland, Canada and even WNE. It's not a set or fixed/monolithic accent. Who puts the West Coast or WNE under heartland? The old school one you hear in old 1950s movies actually came from non-coastal northeast and spread to what one could arguably say a small part of the heartland, but far from all of it. And it wasn't a variety of the heartland. It was heard in parts of New England, Pennsyvania, and New YOrk (Not the city area.) Immigration spread westward. So it makes sense that certain parts of the Upper Midwest would have this accent. It came from the northeast as settlers settled westward. And the english speakers brought there accent with them. My argument is that the accent itself did not originate in the heartland. It's a northeastern accent that came from northeastern settlers who happen to settle is certain parts of what one could call the heartland. Labov talks about the settlement of Northeasterns to the upper midwest, where the standard used to be, and Thomas Bonfiglio talks about the non-coastal northeastern origins of GA in his book, which you sourced on this page. Thanks for the timely responses. Sorry I can't always be that way in my responses. Usernamebradly (talk) 20:30, 9 December 2017 (UTC)

@Usernamebradly: Didn't you yourself argue that Bonfiglio seems confused about the Inland North dialect? In any case, find me an exact source and page number that contradicts McWhorter and we can try to make sense of it. Wolfdog (talk) 23:23, 9 December 2017 (UTC)

@Wolfdog:He stated that the origins of the GA accent was from the non-coastal northeast. Isn't that pretty much what it says in the GA origins section? Didn't you include that part on this page? I think the whole sentence we talked about should be removed. Or we should put that the 40s accent was a non-coastal north eastern rhotoric accent. People like Labov and Ripler (sourced content on the GA wiki page) say todays GA standard is the West, Midland, Canada, Western New England. I wouldn't put most of those in the heartland category. There are some things about what Bonfiglio has said that contradict research that I've done into this. Both Labov and Edward McClelland seem to think that John Kenyon made his hometown the standard because he was the leading phonetician in the US at the time. This means that he had a lot of power and those with power listened to him. I really don't know if we should charge certain parts of the country as seeing themselves as WASP. And I don't know if I agree with the whole racism classism stuff in the GA popularity section. To me that whole section comes off as having pretty big charges against people. Some people that read that section will most certainly label those parts of the country as being classist and racist. I'm not going to say there is none of that. But reading a couple sources like I mentioned earlier shows that racism/classism is not how GA got started. Your thoughts? Usernamebradly (talk) 05:32, 12 December 2017 (UTC)

@Usernamebradly: My thoughts are that, at this far in the discussion, exact sources and page numbers are needed to change any already-sourced material on the page. Wolfdog (talk) 21:24, 12 December 2017 (UTC)

@Wolfdog:In terms of what I was talking about in regards to Bonfiglio, isn't it sourced in the origins sections of this page? Bonfiglio (2002), p. 43 Here's another source in the origins section "Though General American accents are not commonly perceived as associated with any region, their sound system does, in fact, have traceable regional origins: specifically, the English of the non-coastal Northeastern United States in the very early twentieth century.[23]." Usernamebradly (talk) 04:09, 14 December 2017 (UTC)

@Usernamebradly: I guess your problem is that "heartland" is vague or inaccurate, correct? I take "heartland" to mean simply "non-coastal" (as in "west of the Atlantic coast, where the US was originally settled") or to mean "Midwestern". What about the "Western Reserve of Ohio" and the "Great Lakes states" (Bonfiglio's words) is not Midwestern? And what about those areas, plus Upstate NY and Western New England is not "non-coastal"? If we're really just debating about the one word "heartland", we can certainly change that to a synonym that you think would be more clarifying. Wolfdog (talk) 16:19, 16 December 2017 (UTC)

/t/-glottalization

The person who wrote about it seems to have lumped together glottal stops, pre-glottalized alveolar stops and glottaly masked alveolar stops, which are different sounds. For now, I've removed what I could. Mr KEBAB (talk) 15:50, 10 November 2017 (UTC)

@Mr KEBAB: If you ever have any time and would like to outline the distinctions between the items you listed above, I'd love to read about it! Wolfdog (talk) 15:19, 25 November 2017 (UTC)

@Wolfdog: Sure:

- Glottal stop: read glottal stop and tell me if it makes sense to you.

- Pre-glottalized voiceless bilabial/alveolar/postalveolar/velar stops: [p], [t], [tʃ] or [k] that are preceded by a glottal stop. They're simple sequences [ʔp, ʔt, ʔtʃ, ʔk]. I'm not sure how widespread they are in General American, but they're extremely common in General British, so common that ESL speakers risk sounding non-native (and aggressive, consistent lack of pre-glottalization is perceived as rather aggressive-sounding in GB!) if they don't imitate that feature. AFAIK, Peter Roach is (or was) considering introducing pre-glottalization to transcriptions in his Cambridge English Pronouncing Dictionary.

On the other hand, Australian English almost never pre-glottalizes the fortis plosives, which is one of the most obvious differences between the General Australian accent and the General British one.

- Glottaly masked voiceless bilabial/alveolar/postalveolar/velar stops: read [7] and tell me if it makes sense to you. Mr KEBAB (talk) 15:44, 25 November 2017 (UTC)

Right. I guess the only one I didn't already know about was glottal "masking". Just didn't know that terminology. Thanks. Wolfdog (talk) 21:38, 25 November 2017 (UTC)

@Wolfdog: A reliable way of distinguishing glottaly masked plosives from proper glottal stops in the word-final position is that the latter often have a slight [h] offglide, so that they're actually affricates [ʔ͡h] or at least aspirated stops [ʔʰ]. Compare [nɑt͡ʔ] with [nɑʔ͡h ~ nɑʔʰ]. Do the latter sound American to you?

Unless I'm mistaken and this affrication occurs only in the UK, not the US. Then forget about what I've said. Mr KEBAB (talk) 21:52, 25 November 2017 (UTC)

Aspiration

I'm removing aspiration from the article. Most sources don't transcribe it, and it'd be incorrect to transcribe words such as play, crack, pure, quota [ˈpʰleɪ, ˈkʰɹæk, ˈpʰjʊɹ, ˈkʰwoʊɾə] (as it's the following sonorant that's devoiced, not the initial plosive that is aspirated [in the narrow sense "causing the following vowel to be partially devoiced"]) and inconsistent to transcribe them [ˈpleɪ, ˈkɹæk, ˈpjʊɹ, ˈkwoʊɾə], as /l, ɹ, j, w/ don't block aspiration in the wider sense. If we're transcribing e.g. pay as [pʰeɪ], then the only correct/consistent transcription of play is [pl̥eɪ], and this might be an overkill. Also, the devoiced allophone of /ɹ/ is neither alveolar (well, /ɹ/ is hardly ever alveolar in GA) nor an approximant, but a postalveolar non-sibilant fricative. There could be further arguments to transcribe that sound with [ɹ̠̊˔] rather than [ɹ̥]. We don't want that.

We don't transcribe devoicing of the lenis stops /b, d, ɡ/, which is just as striking to non-native ears, especially in the word-final position. This is another inconsistency. And if we start transcribing that, why stop there and not transcribe the allophonic vowel length too? After all, beat and bead differ mainly in the length of the vowel, don't they?

See this discussion, especially the last post. Mr KEBAB (talk) 16:53, 10 November 2017 (UTC)

I'm fine with less phonetically narrow transcriptions (unless we're talking about a feature that requires a greater degree of narrowness, of course), but I should note that it's not strictly inaccurate to transcribe e.g. play as [ˈpʰleɪ]. It's understood from the aspiration symbol that the delayed voice onset time carries into the /l/. After all, if the onset of [ˈpʰleɪ] doesn't manifest as [p] with a (partially) devoiced [l], how does it manifest? — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 20:53, 10 November 2017 (UTC)

I don't know why [ˈpʰleɪ] is inaccurate. After all, the transcription of pay [ˈpʰeɪ] that parallels [pl̥eɪ] is [ˈpĕ̥eɪ], with an extra-short devoiced vowel that corresponds to the aspiration symbol. And for languages with phonemic aspiration, it would be quite normal to put the aspiration symbol before a liquid. (For instance, Attic Greek ἆθλον /âːtʰlon/, not /âːtl̥on/. That's from a dead language, but there must be living examples.) Which version to use seems a question of convention or what you want to emphasize in the transcription: the fact that the aspiration is a quality of the consonant, or the fact that the aspiration devoices the following sound. — Eru·tuon 21:25, 10 November 2017 (UTC)

@Aeusoes1 and Erutuon: It could be misread as a devoiced vocalic release [pə̥ˈleɪ, kə̥ˈɹæk, pə̥ˈjʊɹ, kə̥ˈwoʊɾə], especially in the case of the /pl, kl/ clusters, in which the plosives are pronounced with a lateral (rather than central) release. That's why I find it a bad transcription.

The Attic Greek example is not entirely appropriate here, as you've enclosed your transcriptions within phonemic slashes. Plus, English plosives contrast not by aspiration but by articulatory strength, as fricatives. You don't have to transcribe the aspiration at all. In Attic Greek you do, because it's phonemic. Mr KEBAB (talk) 22:10, 10 November 2017 (UTC)

Could it be misread that way, though? I suspect you would be hard pressed to find a phonetician who makes that same argument. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 22:21, 10 November 2017 (UTC)

@Aeusoes1: Ok, perhaps [ˈpʰleɪ] etc. could be (or is) unambiguous to experienced users/readers of IPA, but perhaps not for laymen. I was hoping that Hans Basbøll brought up the issue of [l̥] etc. in his Phonology of Danish, but there's nothing of significance there.

The thing is that it's an established practice to transcribe partially devoiced vowels with [ʰ] but partially devoiced approximants with the devoicing diacritic attached to them, at least in English. See e.g. Gimson's Pronunciation of English or Australian English Pronunciation and Transcription. Mr KEBAB (talk) 22:56, 10 November 2017 (UTC)

(edit conflict) Yeah, I didn't indicate the example as phonetic because it's from a dead language. But if it were a living language and it had an aspirated stop before a liquid, I think the transcription would still be fine. (There are sequences like /kʰl/ in Thai and Icelandic, for instance ขลุ่ย and klukka. Not sure if the same symbols would also be used in a phonetic transcription.) As to articulatory strength being the distinguishing factor for English obstruents, that is true, but I don't know what bearing phonological factors like that have on phonetic transcriptions. The benefit of indicating aspiration before a sonorant consonant is that it suggests unity between the allophones of p in pay and play: they are both aspirated because they are at the beginning of a word or a stressed syllable. Transcribing them differently suggests that there is some difference between the two, but the only difference (in my view) is whether the sonorant that follows them is a consonant. — Eru·tuon 23:27, 10 November 2017 (UTC)

@Erutuon: I guess, yes. I'm not sure about Thai, but the Icelandic sequence is probably the same as in English clock, so [kl̥]. But it's probably never transcribed like that because what is typically transcribed as [l̥] in Icelandic is a different segment, a fully voiceless fricative [ɬ] as opposed to a devoiced approximant (I'm guessing the realization of the latter, I don't know that for sure).

The only reason I mentioned articulatory strength was because I was saying that you don't have to transcribe aspiration in English.

The benefit of indicating aspiration before a sonorant consonant is that it suggests unity between the allophones of p in pay and play - I'm not sure what you mean. Phonetically, aspiration is purely a property of the following sonorant, so that it's partially devoiced. It doesn't matter how you indicate it, and it'd be a rather newbie mistake to somehow think that there's no 'unity' between them. The most important thing is that they both belong to the /p/ phoneme. Mr KEBAB (talk) 00:26, 12 November 2017 (UTC)

@Mr KEBAB: It might be a newbie mistake not to see a connection, but many people are newbies. That is, they don't have the knowledge that aspiration ([pʰleɪ]) means devoicing of the following sonorant ([pl̥eɪ]). It is true that phenomena do not change depending on how we represent them, but still some representations are better for particular purposes than others. In this case, using [pʰ] in transcriptions of pay and play is a good way to indicate that their allophones of /p/ are similar. On the other hand, I suppose [pl̥eɪ] is a good way to indicate that the pronunciation of /l/ in play is different from the pronunciation in lay [leɪ], if that were your goal. — Eru·tuon 02:37, 12 November 2017 (UTC)

@Erutuon: I think using [CʰV] with vowels and [CC̥] with consonantal sonorants is a good and necessary compromise. First, because it's an established transcription (see above). Second, because transcriptions such as [pĕ̥eːɪ̯] may be rather confusing for laymen, also because hardly any source transcribes aspiration in English in that manner. Some dialects of English have phonemic vowel length, and IPA allows long vowels to be transcribed with either ⟨Vː⟩ or ⟨VV⟩, i.e. as a doubled vowel. It could introduce massive confusion. Also, I think that transcriptions such as [pʰleɪ] aren't common in English. Mr KEBAB (talk) 02:54, 12 November 2017 (UTC)

I agree with Aeusoes1 and Erutuon's initial concerns and ultimately feel that aspiration should be left in the transcriptions. The representation [pleɪ] as a narrow transcription is simply inaccurate for GenAm, and is even capable of being misconstrued by American ears as [bleɪ]. As long as we're keeping the transcription narrow, we should keep it accurate. Even your "overkill" transcription seems better to me than a wannabe-narrow transcription that ends up missing an important feature of the accent. Furthermore, such a distinction can become salient in terms of accent perception, since there are accents of English (e.g. I think first of the accents of India) in which a total lack of aspiration is typical and indeed even noticeable to speakers of other English dialects; GenAm, however, is not such an accent. Wolfdog (talk) 01:44, 11 November 2017 (UTC)

@Wolfdog: You're making a mistake that I've already noticed among a few WP editors. The fact that a transcription is enclosed within phonetic brackets doesn't mean that it's narrow, it means that it's phonetic. Narrow transcription shouldn't be used for general transcriptions but only when you need to transcribe a phonetic feature you're describing. There's too much variability to allow us to account for it in our transcriptions. If we don't, we're left with what you're calling wannabe-narrow transcriptions. There's no point in doing this.

The aim of this article isn't to compare GA with Indian accents (which sometimes are described as non-native), it's just to describe GA. Mr KEBAB (talk) 02:09, 11 November 2017 (UTC)

@Mr KEBAB: I'm confused then what you consider the point of phonemic transcriptions to be, which I've always considered the "broadest" possible transcription. (If there's already been a prior discussion somewhere making this distinction clearer, as I'm sure there has been, feel free to redirect me there.) And I would argue the aim of this article is, indeed, in a sense to compare and contrast it with other English accents (certain Scottish rather than Indian ones, then, if you prefer, though an accent's "nativeness" really has nothing to do with my point) in order to define what makes GenAm unique to a greater or lesser extent from other English accent groups. And what about my point that [peɪ] for example could be misconstrued for [beɪ] by Americans? Wolfdog (talk) 16:42, 11 November 2017 (UTC)

@Wolfdog: The broad-narrow dichotomy applies to phonetic transcription alone. Phonemic transcription, as its name suggests, can only contain phonemes, i.e. contrastive segments of a language. Even though it's purely abstract, most authors' aim is to use phonetically correct symbols to transcribe phonemes so that the most common allophone is displayed, though most commonly, they stop when they need to use diacritics (this is one of the reasons we don't write /pʰ, tʰ, kʰ/ in English phonemic transcription, but not the only one - English plosives are incredibly complex phonetically in comparison with e.g. Polish).

Sometimes you'll read that phonemic transcription is by definition broad. I disagree with this. While that may be true in the majority of cases, sometimes it's not. Consider the Polish word jaj (form of jajo 'egg'). How would I transcribe it phonemically, in broad phonetic transcription and in narrow phonetic transcription? /jaj/, [jaj] and [jaj] (yes, this is pretty much the narrowest transcription possible). In this case, it's the brackets alone that change the transcription to phonetic (and vice versa)! So sometimes (or often, depending on the language), phonemic transcriptions can overlap with the phonetic ones, sometimes completely so, as you can see.

I think that we should use phonemic transcription whenever possible. When phonetic transcription is called for, we should use an established set of phonemic symbols for all of the features that are not discussed in a given sentence, with few possible exceptions. If we don't, there's no telling when we should stop. English obstruents are a whole class of phonemes, and you focus only on the aspiration of /p, t, k/. Why aren't you for transcribing the allophonic devoicing of /b, d, ɡ, dʒ, v, ð, z, ʒ/ or transcribing allophonic vowel length?

I'll tell you why I am against it: first, because IPA is too imperfect to convey this. The lenis obstruents in English are only partially devoiced at the beginning of a syllable and strongly devoiced at the end of an utterance. You can't capture this with IPA without either using too many diacritics and thus preventing most browsers from displaying the correct symbols or using [p, t, k, tʃ, f, θ, s, ʃ] for the more strongly devoiced allophones, and in that case you're blurring the distinction between word-final /b, d, ɡ, dʒ, v, ð, z, ʒ/ and /p, t, k, tʃ, f, θ, s, ʃ/, especially when you don't transcribe allophonic vowel length. Not a good idea either, especially given the fact that the word-final /p, t, k/ can be pre-glottalized [ʔp, ʔt, ʔk] or glottaly masked [p͡ʔ, t͡ʔ, k͡ʔ] (or, in the case of /t/, also flapped, like /d/), which is impossible in the case of /b, d, ɡ/. Because of this allophonic variability, word-final /p, t, k/ are most commonly written [p, t, k] in phonetic transcription. Also, it may not be very correct to transcribe the final /ð/ as [θ] since it's usually closer to an approximant.

Second, because, again, this is variable - for example, word-final lenes are voiced if the next word starts with a vowel.

The allophonic vowel length is also problematic, because even though it applies to pretty much all syllables, unstressed syllables are also shorter than the stressed ones, and the schwa is shorter than all other vowels, but not when it's phonemically /ɜ/ or /ʌ/. Plus, in words such as shell, /l/ is also longer than in shelf because it doesn't precede a fortis consonant within the same syllable. So /l, m, n/ must also be taken into account when describing allophonic vowel length, because they're sonorants just like vowels. This may cause some readers to confuse /l, m, n/ of full length with the phonemic geminates /l.l, m.m, n.n/ (or not, depending on how we transcribe that).

If the aim of this article is to compare GA with other accents of English, why stop there and not make its aim to compare GA with all languages other than English? It's quite easy to take the narrowness of phonetic transcription to utterly absurd levels.

[peɪ] and [beɪ] are equally wrong narrow transcriptions of pay and bay (I know that you're talking about pay alone). Readers of this article are expected to know that the correct narrow transcriptions are [pʰeːɪ̯, b̥eːɪ̯]. Again, you're too rigidly focused on aspiration. It's not the only difference. With that being said, we could improve our coverage of obstruents. Mr KEBAB (talk) 18:41, 11 November 2017 (UTC)

[A]n accent's "nativeness" really has nothing to do with my point - Of course it does. Non-native accents are expected to have phonetic features that aren't commonly found in standard Englishes. As in Danish and Swedish, all major standard English accents aspirate stressed syllable-initial /p, t, k/. It's nothing unique. Mr KEBAB (talk) 19:04, 11 November 2017 (UTC)

@Mr KEBAB: First off, I appreciate you taking the time to explain your perspective, provide some more background, and give examples. This was very helpful. Now here are a few comments in response to some of yours:

• "If the aim of this article is to compare GA with other accents of English, why stop there and not make its aim to compare GA with all languages other than English?" You stop where the language more or less stops, but in reality you compare typical American English to typical British English and possibly others of certain countries. That's pretty obvious. I feel you are implying I'm taking things to a ridiculously extreme level. But to me, a page like this about a particular "accent" is the right place for details.
• "It's quite easy to take the narrowness of phonetic transcription to utterly absurd levels." Again, I get your slippery-slope argument, but I don't intend to take transcription to "utterly absurd" levels and I think you see that. Absence or presence of aspiration is a quite noticeable feature that any GA listener could comment on, and you're right to say that an absence of that feature is typically "foreign" to English. Yes, it is perceived as a foreign sound by Americans and therefore is not some throwaway feature of the accent. Aspiration is very real and consistent in a GA accent, not some tiny arguable option, and there is even a distinction for example between the type of /p/ in "pay" versus in "caper". The rules for aspiration have a systematic pattern, so my discussing their importance is far from utterly absurd.
• I now understand your point about non-native accents. Again, I appreciate your explanation. One thought occurs to me though: Aspiration may still be used differently based upon particular native accent, which I'll address in the next bullet.
• "you're too rigidly focused on aspiration. It's not the only difference." No, actually I'm not. I'd be happy to note all "narrow" but consistent features of accents. We just happened to be focusing on aspiration in this particular conversation. And I'm sure I could be better enlightened to other features too that I'm not yet aware of. As another example, Britons tend to pronounce the /t/ in "sixteen" as [t], while Americans pronounce it as [tʰ]. I'm quite sure the underlying pattern here is noteworthy and would be regarded as completely appropriate to discuss on a page like this, no? So, in my view, I'm focusing on aspiration meaningfully, not rigidly.
• I understand your devoicing comment and you're quite right. But, also, isn't it true that allophonic vowel length and certain types of devoicing apply much more to UK English than GenAm? (Some Britons practically pronounce bad in a way that Americans could hear as bat.)

Wolfdog (talk) 20:04, 11 November 2017 (UTC)

@Wolfdog: I apologize for the unfortunate tone of my message. I was trying to express a general concern, not to overly criticize your approach to the issue. I'll reply to your message soon. Mr KEBAB (talk) 20:12, 11 November 2017 (UTC)

I agree with Nardog that we're all going to be making arbitrary "lines in the sand" of some kind, whether more broad or less broad. I'd be comfortable drawing this line right where Nardog does in the example sentence Nardog gives starting "So I would transcribe...." And Mr KEBAB : no harm done. You've educated me quite a lot with your explanations, so thank you. Wolfdog (talk) 20:14, 11 November 2017 (UTC)

@Wolfdog: No problem and thanks for a very quick response.

But GA is exactly like RP in this regard, at least as far as stressed syllables are concerned and if you count out affrication that is common for /t, k/ in RP/Estuary. As I've already said, aspiration of stressed syllable-initial /p, t, k/ is mandatory in all major standard Englishes. Further evidence that we shouldn't transcribe it comes from papers on other Germanic languages such as German, Norwegian, Swedish and Luxembourgish. The [ʰ] diacritic is universally left out.

I think you can call it a throwaway feature if you look at other major standard Englishes, as every single variant thereof aspirates /p, t, k/.

I find it a bit surprising that you're bringing up pay and caper as well as sixteen. The difference is purely allophonic in the first case. In General American, syllable-initial /p/ is aspirated only when it's stressed. RP is variable in that regard, with younger speakers using [pʰ] in both pay and caper and older speakers having a GA-like distribution. The difference in the pronunciation of sixteen is in the treatment of the second /s/. If you don't aspirate the /t/, you're treating this word as /sɪkˈstin/ (sick-STEEN). If you do, you use more of a spelling pronunciation /sɪksˈtin/ (six-TEEN). In stressed syllables at least, the lack of aspiration of /p, t, k/ is caused by them being preceded by /s/. This is just one of the 'weird' pronunciations you can hear in GB, another example is /əˈtɔːl/ (narrow: [əˈtˢʰoː(ɫ)]). Guess what that means? At all. I don't see how these pronunciations are relevant to this article. You can't hear them in America, no?

Sorry, that doesn't sound accurate to me at all. There's a noticeable difference between the vowel length in bat (shorter) and bad (longer) in GA, just as in RP. And just as in RP, the final /t/ has a variety of allophones that just aren't used for /d/. Try voicing your /d/ at the end of bad. Doesn't it sound just as foreign as the failure to aspirate /p/ in pay? Maybe not, but I bet it's a bit weird. Try the same with vowel length. Pronounce bat with a lengthened vowel and bad with a shortened vowel. Does that sound native to you? Mr KEBAB (talk) 14:58, 12 November 2017 (UTC)

I'm not sure where we've gotten (a)round to in our discussion. I still like Nardog's middle-ground approach that gives us a basic place to pin down our arbitrariness. Mr KEBAB, I see what you mean about the "unnecessariness" of showing aspiration when it is almost universal to English dialects (though with, I think, the interesting exceptions of Cape Flats, possibly Singaporean, Cajun, some native White and Indian South African, and perhaps other dialects [yes, certainly contact varieties that in some cases though have gone native], and as an interesting side-note, "shorter duration of aspiration" reported in working-class rather than middle-class Scottish speakers). While recognizing your point, I don't really see why this should prevent us from transcribing aspiration either. Moreover, I'm happy to transcribe syllable-coda devoicing; I recognize that that's a fairly consistent feature in GenAm and admit that fully voicing the /d/ in "bad" does indeed sound odd to me. And how about the "dark L" commonly transcribed only in the syllable codas of British words; any reason not to show that is certainly dominant in all positions for GenAm? I still find the vowel-length feature (in just my own experience), to be less consistent than aspiration which, again, follows a systematic pattern. I suppose, yes, saying it myself and listening to others online, "bad" may often be longer than "bat" for Americans, but, actually, here's a great place to start drawing our line in the sand. Whereas the bad/bat vowel distinction is optional, the aspiration distinction is pretty much mandatory. Wolfdog (talk) 18:17, 12 November 2017 (UTC)

@Wolfdog: Broad South African English or Singaporean English don't count as Standard English in the same sense as Received Pronunciation, General American and General Australian do. The South African varieties you list don't even count as standard or at least socially neutral in South Africa itself (with a possible exception for the Indian variety, I'm not sure). Indeed, lack of aspiration itself isn't standard/neutral in South African English, because Cultivated and General varieties thereof mandatorily aspirate /p, t, k/ as GA does.

I understand that it's a side note, but English varieties do vary in the amount of aspiration /p, t, k/ receive. Absolutely. But the [ʰ] diacritic doesn't tell you anything about that (at least according to the official IPA usage), it just indicates that the aspiration is present.

English obstruents are devoiced in coda, but only before a pause and voiceless consonants. When immediately before vowels (including word-initial vowels of the following word), they stay voiced. They're also partially devoiced in the syllable onset. Again, this feature is fully shared with Received Pronunciation and General Australian, and indeed also with some other Germanic languages. It's nothing unique.

I'll create a separate thread for sonorants, since you're another editor that wants to discuss that topic.

What makes you say that the bad/bat distinction is optional? The greater length of /æ/ in the former is the reason you're able to perceive the final /d/ as opposed to /t/, especially in isolation. What you're thinking about is the emphatic lenghthening of /æ/ to [æː] in bat, which is possible but not very usual. Again, this allophonic feature is shared with RP and GenAus and, if we count out emphatic lenghtening before fortis consonants, it's mandatory in all of these varieties. Try saying bad, cheese and good with shortened vowels. Make them as short as the vowels of bat, fleece and put. Do they sound normal to you? Mr KEBAB (talk) 20:52, 12 November 2017 (UTC)

@Mr KEBAB: To be honest, yes, I think they would actually sound normal to me (especially in an environment of little stress), but then again maybe I've never actively "heard" this distinction and I've certainly never been taught to. I'm happy to admit there's a clear vowel difference between "bad" and "ban" (/æ/ tensing of course), but what you're talking about, so hard to perceive by an American even saying the two words ("bad" and "bat") side-by-side, strikes me as a feature that is "usual" rather than "necessary". Of course, I'm speaking with no sources. Looking this up online, I see it's well studied. I wonder how I've never come across it. Interesting. Anyway, I'd be happy to include this feature in a phonetic transcription if we agree it's another good place to draw a line. Seeing as I've heard of the idea just now, though, I doubt I should be allowed to voice my opinion either way. Wolfdog (talk) 22:04, 12 November 2017 (UTC)

@Mr KEBAB: Certainly in context, the more I really think about it, if someone said to me "That's really bat" (with the short duration of "bat"), I would simply interpret it as "bad" without batting an eye. Now, if these words were in isolation, perhaps the lengthening applies, but I feel that context certainly overrides length, and the length factor doesn't determine whether a word sounds foreign or not, unlike the aspiration factor, which does. Again, this is all OR on my part, so I don't expect to win any practicable arguments with it. Wolfdog (talk) 22:22, 12 November 2017 (UTC)

@Wolfdog: We're not talking about an environment of little stress but saying those words in isolation and in sentences spoken at a normal pace. When weakly stressed (or completely unstressed), all vowels lose some of their length in English. This may be more noticeable in GA, which obviously doesn't have phonemic vowel length, but it still applies in other dialects. English schwa is so short because it's an unstressed-only vowel.

[B]ut what you're talking about, so hard to perceive by an American even saying the two words ("bad" and "bat") side-by-side - Let me put it this way: the shorter length of the vowel in "bat" is a phonological property of the final /t/ (yes, I imagine that it's a bit hard to believe, but it really is so). Vowel length in stressed final syllables before non-fortis consonants and in an absolute word-final position is "full", before fortes it's "clipped", hence the term "pre-fortis clipping". You're not supposed to (so to say) hear this distinction, you're supposed to hear a final /t/ (or whatever the syllable-final consonant is) and you do.

This is a good example of Native speakers of a given language usually perceive one phoneme in that language as a single distinctive sound, and are "both unaware of and even shocked by" the allophone variations used to pronounce single phonemes. See the lead of allophone.

I don't think that there's a convincing reason to include it in phonemic phonetic transcription, just as there's no completely convincing reason to include aspiration.

We're not talking about context either. You can figure out from context what speakers with strong accents are saying far more often than it is the case with words in isolation. So that's not a good argument.

What you need is a blind test. Ask someone to record for you minimal pairs such as bid vs. bit, bed vs. bet, bud vs. but etc. Or download Forvo recordings, make sure they have neutral names, wait some time to make sure you don't guess the word from the hour you saved the file and listen to them.

Even better: ask someone to record full sentences in which the only phonemic difference is in the articulatory strength of the final consonant of one of the words. Something like that's really bad vs. that's really bat, but in proper English (i.e. there must be a real change of meaning). Ask them not to pronounce the sentences too quickly or too slowly, the pronunciation must be as natural as possible. I can't come up with anything right now (my English isn't that good), so you'll have to figure that out on your own. Mr KEBAB (talk) 09:09, 13 November 2017 (UTC)

@Wolfdog: Sorry if the bit about Forvo recordings sounded a bit patronizing, which wasn't my intent. I was just looking for a way to make that blind test reliable. Mr KEBAB (talk) 01:57, 17 November 2017 (UTC)

@Mr KEBAB: No, no problem! Sorry I didn't respond; I didn't notice the Nov. 13 message because my notification about this probably got lost among a whole slew of recent notifications. The more I try the words out and listen to others say them, the more I notice you're right. It's just something I've never come across or been previously conscious of. I had learned about GenAm not having phonemic vowel length a while back and in my head I think I mistakenly generalized this into the idea that GenAm also doesn't have any allophonic vowel length. So are you also saying we should edit the chart here in order to better represent vowel length? (It's also hard for me to believe that your spoken English isn't that good when your casual written English is excellent and when you understand such a depth of English phonetics.) Wolfdog (talk) 03:36, 18 November 2017 (UTC)

@Wolfdog: There seems to be a major issue with the replyto and ping templates. In my experience at least, they seem not to work in about 30% of cases, which is a lot. You need to put this page on your watchlist if you want to be sure that you'll be notified of new messages.

Yeah, it's a mistake, but it's only really relevant in stressed word-final syllables, especially when a pause or a fortis word-initial consonant follows. In other environments it's very different from the vowel length of e.g. German, in which [aː] is much longer than [a]. The length of the former would typically signal word boundary to native speakers of English, that's why it's important not to make English vowels too long.

When I hear lid or lead with an overshort vowel I immediately think of Scottish English. This is one of the reasons many people find Scottish people hard to understand.

I've already edited it, you can check it.

Thanks, but speaking about something you find interesting is a different thing than making up very specific sentences on the spot. The latter is much closer to amateur poetry than a casual discussion. With that being said, try these pairs:

• Where's my cab / cap? (utterance-final /Vb/ vs. /Vp/)
• Cathy lost her bed / bet. (utterance-final /Vd/ vs. /Vt/)
• These are my kids / kits. (utterance-final /Vdz/ vs. /Vts/)

I can't think of any sentence-internal minimal pairs in which word-final lenis consonant(s) are followed by a word-initial fortis one, as in my kids thought it was funny. Mr KEBAB (talk) 23:51, 18 November 2017 (UTC)

@Wolfdog: Thanks. Interesting points. I wonder if, for me, the distinction "kids / kits" has a few other differentiating features, since the unreleased element though perhaps somewhat present is also not as strong with final consonant clusters: [kʰɪ(ˑ)dz] vs. [kʰɪts]. Wolfdog (talk) 16:14, 19 November 2017 (UTC)

@Wolfdog: Thanks. Now that I think of it, was funny is actually not a very good example as /ə/ is an inherently short vowel which makes pre-fortis clipping less noticeable.

I'm afraid that's not in agreement with what I've read. It's the amount of unstressed syllables that follow the stressed one that reduce the length of the stressed vowel (it's called rhytmic clipping), not the amount of consonants that end the stressed syllable. With that being said, the final consonants of kids will be revoiced before word-initial vowels, though not necessarily before word-initial lenis consonants (but that may be dialect- or speaker-dependent, I'd have to do a better research on that). Mr KEBAB (talk) 12:19, 22 November 2017 (UTC)

I'm ambivalent. As much as I agree phonetic transcriptions need not be completely impressionistic (IPA Handbook, p. 29), I think it's natural for a reader to expect the same level of narrowness throughout a transcription upon encountering one that describes an allophone. So [ˈpɑɹɾi] strikes me as a weirdly semi-narrow-semi-broad transcription because it transcribes one consonant with a non-phonemic symbol but not the other. (The choice of ⟨ɹ⟩ rather than ⟨ɹ̠⟩ is totally valid though, because IPA doesn't define the exact place of articulation for ⟨ɹ⟩ and there's no other coronal approximant in English.)

Would you say the transcriptions illustrating T-glottalization and flapping with audio in the article should be [ˈmaʊnʔn̩], [ˈpɑɹʔnəɹ], [ˈliɾəɹ] and [ˈkæɾəl], rather than the current [ˈmæʊnʔn̩], [ˈpɑɹʔnɚ], [ˈɫiɾɚ] and [ˈkæɾɫ̩]? It's curious you removed the aspiration but left [æʊ, ɫ, ɚ]. (Perhaps because these are covered in the article? But if so, would you restore the aspiration if it was incorporated in the article?)

I venture to say this is a case where WP:LEAST and WP:UCS apply and where a line in the sand should be drawn somewhat arbitrarily, not rigorously, just like editors are allowed some leeway in prose even by MoS. So I would transcribe [pʰeɪ, pɫ̥eɪ, pɹ̥eɪ] rather than [peɪ, pleɪ, pɹeɪ] or [pʰeːɪ̯, pɫ̥eːɪ̯, pɹ̠̊˔eːɪ̯], even though I know the latter are technically more "consistent" than the first. Nardog (talk) 20:02, 11 November 2017 (UTC)

@Nardog: I can see that you linked to WP:THREAD in your edit summary. What was the problem with the indentation? I think we all used it properly.

I think it's natural for a reader to expect the same level of narrowness throughout a transcription upon encountering one that describes an allophone. That's not an established practice and I disagree. Also, what is your opinion on allophones that are in free variation? See the note on the pre-/l/ allophones of /i, u, eɪ, oʊ/ that I added a few days ago. That's what I mean when I talk about absurdity.

But flapping of the alveolar stops is not nearly as predictable as aspiration of the fortis ones. It's important to transcribe it, also because the consistency with which /t/ is flapped in GA is one of the most recognizable features of this accent. Aspiration, on the other hand, is not. It's much more noticeable in Estuary English, Northern Welsh English or Scouse. In those accents, /p, t, k/ are very strongly aspirated (much more so than in the US, especially in less stressed positions), not rarely with affrication. In that regard, they're more like certain Northern German dialects than General American.

Mountain should be [ˈmaʊnʔn̩], yes. As far as other transcriptions you've listed are concerned, maybe you should make a separate thread for them. To me, it's important to transcribe syllabic sonorants in English because what is two segments phonemically becomes one segment phonetically. Plus, it's also less predictable than aspiration. More so than flapping, but less so than aspiration.

Ok, but /p, t, k/ aren't the only obstruents we have in English. What's your opinion on what I've said regarding the lenis stops and fricatives? It's equally important. Mr KEBAB (talk) 20:38, 11 November 2017 (UTC)

I was referring to the last paragraph of that section. An empty line between indented paragraphs breaks the nesting in HTML and therefore the semantics of the code.

That's not an established practice – What makes you say that? Every author makes a decision on how much detail to be included every time they write a phonetic transcription, and it is not at all uncommon to see transcriptions with greater detail than needed to illustrate the relevant phenomenon, but not so much as the narrowest possible (e.g. [8]).

To me, it's important to transcribe syllabic sonorants – I know, right? So you're making arbitrary—not to be confused with baseless—judgments as to what is important information worth including and what is not, as do we all. Yet you're making this—as Wolfdog eloquently puts it—slippery-slope argument that to include information not crucially relevant would bring "absurdity". We do and must make judgments as to what's notable enough to include and what's not case by case, and there cannot possibly be a rule that universally applies, as we can clearly see. Nardog (talk) 22:28, 11 November 2017 (UTC)

@Nardog: Ok.

Fair enough - the reason I said that was because most sources I've read normally use phonemic transcription, which is probably what we should do (but then there's the /t/-flapping which is not very predictable, so it may not be the best idea). But can you prove that fairly narrow transcription is in any way common as far as English is concerned? I somehow doubt that.

Of course, I don't claim to be the final authority on phonetic transcription (far from it, actually). To call my opinions arbitrary is stating the obvious. This is a discussion. Also, if you're for transcribing the aspiration but not other phonetic properties of English obstruents (as one could conclude from your lack of response to that, which I find a bit confusing), it's just as arbitrary. If you want a further discussion on sonorants, you should create a separate thread, as I asked. Let's not derail this conversation.

Eventually it would. I'm not saying that aspiration alone would bring absurdity, but it's entirely unnecessary to include it in our transcriptions, especially if that causes discrepancies as far as transcribing other obstruents is concerned. As John Wells would put it, such details are best stated once in the article so that readers can 'convert' transcriptions from broad to narrow on their own in their minds.

I'd still like to know your opinion on allophones in free variation. Mr KEBAB (talk) 00:26, 12 November 2017 (UTC)

most sources I've read normally use phonemic transcription – Even when, like I said, describing an allophone? How does one describe an allophone in phonemic transcription? I'm afraid I don't quite know how to parse your second paragraph.

If you took my characterization of your comment as "arbitrary" as some kind of criticism, I'm sorry, I have failed to communicate my opinion clearly. My criticism is that you're making a classic slippery-slope argument of "Where does it stop?" when the answer is clearly "Somewhere." Arbitrariness (to some degree) is what I'm arguing for, not against.

I agree that it's inconsistent to transcribe aspiration but not devoicing, especially in codas; we should do the other if we were to do either. I'm actually being kind of swayed toward your side as far as whether to include aspiration in phonetic transcriptions because of this reason. However, to uniformly remove "extraneous" phonetic detail in allophonic transcriptions simply because it doesn't constitute a phonemic contrast or concern the phenomenon discussed in the context is not something I would be behind.

I'd still like to know your opinion on allophones in free variation. – Use the same symbols as the phonemes, of course. Nardog (talk) 03:56, 12 November 2017 (UTC)

@Nardog: Haha, no, obviously not. I thought that the argument you were making was that, let's say, because the correct narrow transcription of /eɪ/ is [eːɪ̯], we must use this symbol whenever we phonetically transcribe words with an unclipped /eɪ/ to be consistent, and it's just not something that most scholars would do. They typically decide on a rather broad set of symbols. Plus, aspiration is very commonly left out in transcriptions of most other Germanic languages. I can list a few papers if you want.

No problem. I also misunderstood your message.

I don't want to do it completely uniformly. I said that there's room for exceptions such as the alveolar flap or syllabic sonorants. Aspiration, on the other hand, is completely predictable and is a feature that is shared will all other major standard Englishes. Mr KEBAB (talk) 14:33, 12 November 2017 (UTC)

So where did we leave off? I'm sorry to harp on this, but where are we drawing a line of consensus, to help better protect the page from future edit wars? If aspiration and devoicing are two "completely predictable" features, isn't there some value in representing them (especially when representing individual audio files)? And what we can avoid representing, I'd think, is the narrowest of features that are variable or most likely to be open to debate. Wolfdog (talk) 14:49, 18 November 2017 (UTC)

@Wolfdog: Nardog is too busy to respond right now. And who's gonna edit war? None of us I suppose.

I'm for covering them once under the consonant table. Mr KEBAB (talk) 23:51, 18 November 2017 (UTC)

@Mr KEBAB: Ok, you got it. Wolfdog (talk) 16:14, 19 November 2017 (UTC)

@Mr KEBAB: Sorry it took so long to respond. After giving some thought, I'm still of the opinion that aspiration is better included in most contexts. I would also point to the fact that Wells did include it in transcription in AoE where it wasn't necessarily relevant to the point being illustrated (e.g. p. 31), but he didn't do so in a consistent manner either, so for now I would go no further than to reiterate that we shouldn't be adding or removing certain pieces of information from phonetic notations uniformly, which you seem to agree.

What was added while I was gone that I have more of a problem with, rather, is the notation /ʌr/. I'll discuss this in a new section. Nardog (talk) 20:41, 26 November 2017 (UTC)

Edit war

An anon keeps changing /oʊr/ to /or/ in one particular place in the article, creating inconsistencies in phonemic transcription. He quite clearly mistakes allophones for phonemes. It's understood that the FORCE vowel isn't a phonetic closing diphthong + [ɹ] but a close-mid monophthong + [ɹ]. If we were to change /oʊr/ to /or/ where the anon changes it, we must change /oʊ/ to /o/ in the whole article, and that will create discrepancies with /eɪ/, which we shouldn't change to /e/ since that's the most common representation of /ɛ/ in the British tradition. Plus, /oʊ/ and /eɪ/ are most commonly diphthongal, not monophthongal. There's no reason for that change. Mr KEBAB (talk) 22:17, 12 November 2017 (UTC)

Notation for NURSE

Mr KEBAB has changed the phonemic notation for the NURSE vowel from /ɜr/ to /ʌr/. This leads to hurry being transcribed /ˈhʌri/ in both GA and RP, making it difficult to illustrate the different pronunciations of the word in the two accents. Yes, Wells does call it "logically unassailable", but who's actually adopted this notation? Even Wells himself opts for /ɜr/ throughout his book, because "it makes for easier comparisons with other accents".

I know that from a purely phonological standpoint, it makes sense to regard NURSE in GA as STRUT + /r/. But then it makes at least more sense to transcribe STRUT as /ɜ/ than NURSE as /ʌr/, if you ask me―I bet central STRUT is even more common now than it was when AoE was written. But at which point I would say, again, who's actually using that notation?

The GA phonemes serve as reference points for comparison with other varieties (this is even truer for GA than for RP; see AoE p. 118), so we should never hasten to make notations completely phonemic, let alone apply that to the English phonology article. What we should do is to stick to the notation most common and highly regarded, in honor of WP:DUE.

If anything, /ər/ is the notation for GA NURSE that's widely adopted next to /ɜr/. But then I wonder if the STRUT–COMMA merger in GA is something widely agreed upon and can be backed up by reliable sources. If so, and only if so, I would opt for transcribing all STRUT/COMMA/NURSE/LETTER with ⟨ə⟩. Other than that, however, I see no scenario in which I would support transcribing NURSE in GA as anything other than /ɜr/. Nardog (talk) 20:51, 26 November 2017 (UTC)

@Nardog: The main problem I have with /ɜ/ is that it doesn't contrast with either /ʌ/ or /ə/, which makes forward homophonous with foreword, which isn't the case in RP. Using it isn't correct as far as phonemic transcription is concerned. In phonetic transcription, NURSE and LETTER should be transcribed identically as they have exactly the same quality ([ɚ]). Mr KEBAB (talk) 21:03, 26 November 2017 (UTC)

How is that a problem? Wouldn't forward be transcribed /ˈfɔrwərd/ while foreword /ˈfɔrwɜrd/, corresponding to RP /ˈfɔːwəd/–/ˈfɔːwɜːd/?

And this is Wikipedia, it doesn't matter whether something is correct or wrong in someone's opinion. It has to be verifiable by RSes. Nardog (talk) 21:13, 26 November 2017 (UTC)

@Nardog: It is a problem because this distinction doesn't exist in GA. The source is, of course, AoE and any dictionary that transcribes NURSE with /ər/. It's really surprising that you're accusing me of OR when the non-phonemic status of /ɜ/ is so obvious. It's the same as postulating a separate /ɒ/ phoneme in GA. It doesn't exist. Mr KEBAB (talk) 21:18, 26 November 2017 (UTC)

What I'm calling OR is the notation /ʌr/, not the lack of the opposition. Can you provide any reputable source that transcribes NURSE as /ʌr/ other than that one page of AoE? As we know, phonemic transcriptions don't always have to be completely phonemic. The "redundancy" of transcribing NURSE with /ɜr/ helps explain and illustrate the phonological history of the vowel manifested in pairs such as hurry/furry and forward/foreword―or, whatever the reason, that's the notation many reputable sources adopt and thus so should we per WP:DUE/WP:LEAST. Nardog (talk) 21:29, 26 November 2017 (UTC)

We might also consider how representative this analysis/transcription is. If it's unusual, we might want to avoid adopting it when discussing GA. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 16:50, 27 November 2017 (UTC)

@Nardog: Sorry for the wait.

It may a bit on the WP:FRINGE side, but it's not OR because it's sourced.

You can restore /ɜ/, but let's not change the phonetic transcription. Using /ɜ/ seems to be against the principles of the IPA, because it falsifies the number of actual phonemes in GA just to make comparison with other accents easier. That's not good at all. But if that's how other sources write it, I suppose we have no choice but to follow them.

By the way, I didn't change the transcription on English phonology. Officer781 did. Mr KEBAB (talk) 13:15, 6 December 2017 (UTC)

Sidetrack, but I did that because the two pages originally had identical copies of that paragraph. Im just trying to make the information consistent.--Officer781 (talk) 07:01, 7 December 2017 (UTC)

I'm aware of that, both of you. Nardog (talk) 07:17, 7 December 2017 (UTC)

@Officer781: I know. I didn't mean to sound accusatory. If I did, sorry. Mr KEBAB (talk) 09:31, 7 December 2017 (UTC)

@Mr KEBAB: It's fine thanks. I just gave Nardog the info for the edit.--Officer781 (talk) 09:40, 7 December 2017 (UTC)

@Nardog: Thanks for fixing the IPA on English phonology. Mr KEBAB (talk) 09:31, 7 December 2017 (UTC)

The "name" section doesnt make any sense

Mid-Atlantic English, Inland Northern English, and Western Pennsylvania English are obviously General American.