Talk:Checked and free vowels

Why is ʌ listed as a checked vowel, or even as a vowel at all? It's identical to schwa so far as I can tell.
 * In American English, is phonetically very close to or the same as schwa, but in other accents it's quite different. Phonologically, their distribution is also completely different, as  is used only in stressed syllables and must be followed by a consonant, while schwa is used only in unstressed syllables and need not be followed by a consonant. Angr (talk • contribs) 07:46, 6 May 2006 (UTC)
 * But if it's not present in American dialects, why is it listed as a checked vowel for General American?
 * Another question: Is there a multi-syllable word where is used? The example given is a single syllable word, and stress in such words can be somewhat ambiguous. (While there is only one syllable for the stress to fall on, most one syllable function words are left completely unstressed when pronounced in context.) I'm also confused about the claim that rhotacized schwa can't stand in unstressed syllables: What about "murderer"? /"m@`=d@`=.r@`=/

(actually that last syllable is more like /M\`@`=/, but I'm not quite sure of the X-SAMPA for a rhotacized velar approximant, which is basically the semi-vowel equivalent of /@`=/). Linguofreak 15:50, 7 May 2006 (UTC)
 * is present as a phoneme in American English, it just happens to be acoustically very similar to . There are plenty of polysyllabic words with : mother, cupboard, puppy, putter, scuppernong, buttocks. The stressed syllable of murderer is considered to have rather than, although there probably really is no phonetic difference between these two. Both symbols are used primarily to facilitate comparison between rhotic and nonrhotic accents: murderer is rhotic  and nonrhotic . Not everyone follows these conventions, though; I have certainly seen transcriptions of American English where bud and bird are written  and  rather than  and . Neither system is wrong, but I think the system that uses different symbols for stressed syllables than unstressed syllables is more popular because there's less potential for confusion. Angr (talk • contribs) 16:36, 7 May 2006 (UTC)
 * Can you give a minimal pair between and ? I hear no difference. Aren't two phones considered allophones of the same phoneme if they sound identical to a native speaker?

Linguofreak 18:38, 7 May 2006 (UTC)
 * I can't think of a minimal pair off the top of my head, and if there were one, the two words still have to have different stress patterns since appears only in stressed syllables and  only in unstressed ones. A near-minimal pair, differing otherwise only in the initial consonant, is pickup  vs. hiccup . There are certainly people who argue that  and  are the same phoneme for precisely the reason you gave. Other people say  isn't a phoneme at all, the distinction between all lax vowels (except  in some accents) is lost in unstressed syllables. Following this reasoning, the phonemic form of hiccup doesn't have to be  because it could just as easily be, , ,  or  anyway, so why posit an extra phoneme? Other people would say that stress is predictable from vowel quality, so  and  have to be separate phonemes to distinguish  from . There's not one easy answer. Angr (talk • contribs) 19:07, 7 May 2006 (UTC)

Problem being, I don't have any stress contrast between pickup and hiccup.

Just as a side note, do you know why certain schwa's in certain dialects become /I/? It's quite prevalent in my dialect, but I can't find any rhyme or reason to it. "Banana" is /b@n{n@/, but allophone can be either /{l@fon/ or /{lIfon/, and "reason" is always /rizIn/. I can't figure out under what conditions /@/ stays the same, under which ones it becomes /I/, and under which ones it alternates. Linguofreak 23:28, 7 May 2006 (UTC)


 * According to the IPA chart (see International Phonetic Alphabet), is a back vowel (the unrounded version of ), and in classic RP and other accents it is spoken as such. But in a large proportion of accents (certainly my own, which is approximately Estuary English, and I think General American) its actual realisation is, a near-open central vowel (i.e., the same as  but more open). It's probably debatable whether there is a phonemic contrast since  only appears in unstressed vowels.


 * Having said that, I would urge people not to write anything in the article based solely on discussions here - we should be citing sources at every step. Hairy Dude 16:11, 16 November 2006 (UTC)
 * Just to throw my belated 2¢ in here... in Wisconsin English, ʌ and ə are definitely phonemic. If someone were to pronounce "bud" as mentioned above (i.e., as [bəd]), at least all alone (w/o context), it would be assumed that they were saying "bird" w/o proper rhoticizing.
 * On a different subject, the statement that [ə] is only found in non-stressed syllables strikes me as not exactly true, since, as far as I'm aware, [ə:] is the non-rhotic realization of [ɝ]. I could launch into a whole discussion about [ɝ ɚ] as opposed to [ɹ̩ː ɹ̩] (and how ridiculous I think the former is), but I'll spare y'all, at least for today... Tom e rtalk  23:55, 25 March 2007 (UTC)

"Missing" phoneme
I notice that this article doesn't list the RP vowel as a checked phoneme. Obviously it doesn't exist in General American (which has for the most part undergone the father-bother merger), but I think it is worth listing all the same. Thoughts? Hairy Dude 15:58, 16 November 2006 (UTC)


 * It may be desirable to include it, but as it is now, the examples are explicitly only from General American, so this'd need to be rewritten. -- j. 'mach' wust ☞ ☏ 17:33, 16 November 2006 (UTC)

Problematic statement
The article presently asserts: "Only a few interjections like nah /næ/, yeah /jæ/~/jɛ/, uh /ʌ/, duh /dʌ/ have a checked vowel at the end." Well, just a little bit of OR here, but "nah" is [na:] or [nã:], "yeah" is never [jɛ], always [jæ:] (although it really is the only word that ends with [æ]), "uh" is not [ʌ], but rather [ə:] and "duh" is either [də:] or [dʊ:], the difference correlating w/ intention (and breathiness as well). That said, "a" and "the" are [ʌ] and [ðʌ], respectively, and neither of them are interjections. Furthermore, every multisyllable word I can think of that ends in "a", ends with [ʌ] ("Minnesota", for example...for people from Missouri, "Missouri", for example, although for me Missouri ends with an [i] (not [ɪ] like the experts claim)).

Anyways, enough on that rant...if this classification of vowels is so sturdy, why do phoneticists say that diphthongs end with [ɪ] and [ʊ] instead of with [ʲ] and [ʷ]? Tom e rtalk 23:55, 25 March 2007 (UTC)
 * Because and  aren't sounds; they're diacritics indicating palatalization and labialization (respectively) of a preceding consonant. —Angr 04:26, 26 March 2007 (UTC)
 * Time to expand their use, methinks... :-) Tom e rtalk  06:38, 26 March 2007 (UTC)
 * That's silly. Some people, however, do analyze diphthongs as [aj] and [aw] rather than [aɪ] and [aʊ] for example. There is no reason to use the superscript there. -MP
 * I've never heard "duh" as [dʊ:], although I have heard "dur". I have heard [jɛ] for "yeah" - I think it would be used for a shorter, curter "yeah" than [jæ]. I don't really understand how you're getting that "uh" is [ə:] but "Minnesota" as ending with [ʌ]. The vowel in Minnesota sounds more reduced to me, so I feel like you have them reversed. As a random note, the interjection "meh" (of more recent coinage) is pronounced [mɛ] -MP 25-9-2009

More on exceptions for checked vowels
Presently the only exception given in the article, for a normally checked vowel being unchecked, is yeah as /jæ/. Apparently the alternative pronunciaton /jɛ/ was once there but was deleted. But I agree with MP&mdash;in my General American English, both are equally common in my (OR) experience. My dictionary, Random House Webster's College Dictionary, gives the only pronunciation as the vowel in the word "dare" without the "r", which I believe is halfway between the checked vowel /ɛ/ and the free vowel /e/ without a glide.

According to this dictionary, "the" and "a" cannot have [ʌ], only a schwa or a free vowel, contrary to Tomer above (although my OR introspection suggests that [ʌ] does in fact occur frequently in both of these when there is an in-between degree of emphasis).

Also according to this dictionary, other examples of normally checked vowels being unchecked are:

Interjections with [ʌ]: huh, uh, uh-uh (I believe the first syllable ends in a glottal stop, which this dictionary doesn't recognize, but the second syllable is still an example), uh-huh, and uh-oh (but again maybe the first syllable ends in a glottal stop).

Nouns with [æ]: taboo (alternatively with a schwa), Hanoi (alternatively with a schwa), caffeine (two options: the "f" can be in either syllable), cafe (alternatively with a schwa), plateau (American English; it says British English can (must?) put the "t" in the first syllable), tattoo, raccoon.

Noun with []: Tibet (surprisingly to me, no alternative pronunciation with a schwa).

Given that the rule for the five checked vowels almost always holds, would it be worthwhile to put these exceptions into the article, especially since this dictionary is a secondary source? 75.183.96.242 (talk) 17:12, 12 July 2010 (UTC)

Also:

Words with unchecked [ɛ]: eh (alternatively [e]), Yahweh

75.183.96.242 (talk) 15:07, 14 July 2010 (UTC)


 * The rule is that checked vowels must be followed by a consonant in a stressed syllable. Most of your exceptions have checked vowels in unstressed syllables; most of the others are onomatopoeias, which can be mentioned, although it's not unusual for onomatopoeias to behave phonologically differently from "normal" words. +Angr 17:18, 28 July 2010 (UTC)


 * Sorry I missed the part about "in a stressed syllable". I left the examples on the talk page for two weeks before putting them into the article, but didn't get any comments.  It's interesting that, except for my unstressed examples that you deleted, the lax vowels in English are followed by a same-syllable consonant even in unstressed syllables.
 * I noticed that you referred in the article to the other examples as mostly onomatopoeias, which the Wikipedia page defines as "a word that imitates or suggests the source of the sound that it describes". But do any of them (yeah; eh; duh, huh, uh, uh-uh, and uh-huh) imitate something else? I would have called them interjections.  75.183.96.242 (talk) 20:01, 28 July 2010 (UTC)
 * Sorry, I had overlooked your comments on the talk page. I don't think yeah is an onomatopoeia; it's the reason I wrote "mostly" rather than "all". As for the others, hmm. I see what you mean about them not exactly fitting the definition of onomatopoeia the way things like "bang" and "moo" do; but they're also not simply ordinary interjections like "Gosh!" and "Hey!". Are "yes" and "no" interjections? Surely their synonyms "yeah"/"uh-huh" and "uh-uh" are the same part of speech as they are. Likewise "eh" and "huh", which are synonyms of "What?" should be the same part of speech. Of course, interjection and onomatopoeia are not mutually exclusive terms: interjection refers to a part of speech, and onomatopoeia refers to a type of etymology. If I say "The door slammed shut - bang!" I'm using an onomatopoeia as an interjection, but if I say "The door banged shut" I'm using it as a verb, and if I say "The door closed with a bang" I'm using it as a noun. "Duh" could be a true onomatopoeia if it's imitative of the nonlinguistic grunts popularly assumed to be made by the profoundly mentally retarded. +Angr 08:57, 29 July 2010 (UTC)


 * My immediate reaction was that "yes" and "no" are indeed interjections, since they can be a complete sentence by themselves; but my dictionary calls them adverbs, and it says an interjection must be associated with emotion. So they're not interjections, but with the possible exception of "duh" they're also not onomatopoeia.  So could you replace "There are a few exceptions, mostly in onomatopoeias: ...." with "There are a few exceptions, mostly words that are commonly used in one-word sentences." ?


 * Getting back to the question of stressed versus unstressed syllables: The article tenseness says "Since in Germanic languages, lax vowels generally only occur in closed syllables, they are also called checked vowels, whereas the tense vowels are called free vowels as they can occur at the end of a syllable." There's nothing here limiting it to stressed syllables.  Do you have a source that defines it?  75.183.96.242 (talk) 17:29, 29 July 2010 (UTC)
 * I'll see what I can find. +Angr 21:22, 29 July 2010 (UTC)


 * I looked in the PE1133 section of the library, and I found one definition: Kurath, Hans, A Phonology and Prosody of Modern English, Univ. of Michigan Press, 1964: ch. 2   says on p. 17:


 * All checked and free vowels occur under full-stress and half-stress, some of them also under weak stress and unstress. Checked vowels do not occur at the end of morphemes; they are always followed by one or two consonants.  Free vowels, on the other hand, appear both finally and before consonants.


 * This says that we're both wrong: what's relevant is the end of a morpheme, not the end of a stressed or unstressed syllable. On the other hand, the following quote backs up your position if you interpret short and long vowels as checked and free vowels.  Trnka, Bohumil, A Phonological Analysis of Present-Day Standard English, Univ. of Alabama Press, 1968   says on p. 22:


 * Stressed short vowels in English are operative solely before consonants and never occur finally. At the end of stressed words or syllables only "long" vowels and "diphthongs" are admissible.


 * 75.183.96.242 (talk) 17:00, 30 July 2010 (UTC)
 * The morpheme-based definition doesn't seem to allow for accents in which happy (for example) is pronounced . +Angr 08:46, 31 July 2010 (UTC)


 * Interesting example, especially since there must be a huge number of equivalent examples. I've seen the pronunciation    in old dictionaries.  But are there really accents that pronounce it this way, as opposed to    without a lengthening : after the [i]?    75.183.96.242 (talk) 16:57, 31 July 2010 (UTC)
 * I'm pretty sure there are still people who pronounce words like that with a lax vowel. It's recessive, but I don't think it's extinct yet. +Angr 18:46, 31 July 2010 (UTC)

Onomatopoeias?
The article refers to the following as "onomatopoeias":


 * yeah /jæ/; eh /ɛ/; duh, huh, uh, uh-uh, and uh-huh with /ʌ/

If an onomatopoeia is properly defined as a word that imitates a sound which exists independent of that word, such as crash, burp, whir, or gulp, then none of the listed words is an onomatopoeia. Yeah (which is not pronounced /jæ/ in General American but /ˈjeə/) is simply a particle equivalent to yes or yea (whatever part of speech one considers those to be). Eh (which likewise is not pronounced /ɛ/ in General American but /ˈeɪ/) may be an interrogative (c.f. Ger. nicht wahr?, and oder?, Fr. non?), or it may serve merely as a meaningless place-holder like duh, uh, um, etc. (Cf. Fr. hein.)  Huh is usually an interjection expressing puzzlement, surprise, or doubt. It can also take the place of interrogative eh. Uh-uh and uh-huh mean, respectively, no and yes, and are therefore in the same category as yeah. None of them imitates a sound that exists independent of the word itself, so none is an onomatopoeia. (Arguably, huh is in some cases an onomatopoeia for a kind of snort: cf. the onomatopoeia humph!; but if so, it is the only one on the list.)

The standard of Wikipedia's linguistic articles is usually higher than this. An editor who understands the concept of checked and free vowels (I don't) should revise the passage. J. D. Crutchfield &#124; Talk 17:08, 29 January 2014 (UTC)


 * In the absence of correction by a better-qualified editor, I have substituted "particles" for "onomatopoeias" and eliminated "yeah", which is neither.

J. D. Crutchfield &#124; Talk 15:03, 22 March 2014 (UTC)

Baa
offers baa as an onomatopoeia in free /æ/ for American English. My impression is that General American pronounces baa as /baː/, not as /bæː/. I grew up pronouncing it /bæː/, but that's because my parents are from South Carolina, where ma and pa were traditionally pronounced /mæː/ and /pæː/. When I was small, I had a book of riddles that included the following:


 * Q: Where does a sheep go for a haircut?


 * A: To the baa-baa shop.

That seems to presume /baː/ for baa in General American. The riddle made no sense to me, because in my dialect baa-baa sounds nothing like barber, and /baː-baː/ isn't the sound a sheep makes. Only when I started school in Virginia, and heard my classmates singing "/baː-baː blæk ʃiːp/" did I get the joke (such as it is). I seem to remember, also, that the TV series "Baa-Baa Black Sheep" used /baː-baː/.

I would therefore suggest that /æ/ is not a free vowel for General American, even in baa, though it is for many American dialects.

Of course, /æ/ is a free vowel (if I understand the concept correctly) in Southern American and African-American English, not only in /bæː/, /mæː/, and /pæː/, but also wherever RP /ai/ is monophthonguized, as in eye, rise, wine, etc. Unfortunately, IPA can't distinguish between the vowel in South-Carolinian pa and that in SAE pie. My pa had pie, in traditional South-Carolinian contains two distinct vowels, but, as far as I'm aware, IPA would give it as /mæː pæː hæd pæː/. That's a problem for another discussion, though.

I note, incidentally, that Merriam-Webster offers both pronunciations for baa (as "\ˈba, ˈbä\"), but prefers /baː/. See baa.

J. D. Crutchfield &#124; Talk 15:23, 4 April 2014 (UTC)


 * Why do you say it prefers /baː/? It lists /bæ/ first, and the sound file has /bæ/. The monophthongized PRICE vowel in Southern U.S. English, though, isn't /æ/, it's /aː/, which is why tap and type aren't homophones. As for the other pronunciation of baa, it's /bɑː/ with a back vowel, homophonous with bah. Aɴɢʀ (talk) 18:21, 4 April 2014 (UTC)


 * I see that I was confusing /a/ and /ɐ/. (I'm relying on the audio IPA examples at the University of Victoria.)  What I meant was that M-W prefers /bɐː/ because I was taking its "\ˈba, ˈbä\" to translate into IPA /bɐː/, /bæː/, in that order.  I had not listened to the audio example, and it does indeed give /bæ/, in what sounds to me like a Mid-Western accent, verging on a diphthong, /bæə/. Having looked up M-W's pronunciations for father and fatter I see I was also mistaken about M-W's phonetic symbols.  On the other hand, M-W gives the following as rhymes for baa:


 * aah, ah, bah, blah, Blois, bra, dah, droit, fa, Fra, ha, Jah, Kwa, la, ma, nah, pa, pas, Ra, rah, schwa, shah, ska, spa, ta


 * Not an /æ/ in the bunch! I take all of these to be in /ɐ/ (or /ɐː/) for General American.  The /ɑː/ Angr proposes occurs in those words for some regional accents, including that of the Great Northwest; but I think the American broadcast standard has /ɐː/.


 * As for the Southern American English monophthong, for most Southerners I believe price is pronounced /pɹais/. (It becomes /pɹæːs/ only from the mountains westward.) Prize, on the other hand, is almost universally /pɹæːz/;  /pɹaːz/ occurring only in some regional accents, such as southern Louisiana and parts of southeastern Virginia.  Tap and type aren't homophones for most Southerners (including speakers of African-American Vernacular English wherever found) because the former contains the monophthong /æ/, while the latter employs the diphthong /ai/.  The same is true for trap and tripe.  On the other hand, for many (and I suspect most) Southern speakers, the difference in vowels between trap and tribe is purely one of length.  J. D. Crutchfield &#124; Talk 18:21, 7 April 2014 (UTC)
 * I've never seen any vowel of American English transcribed as /ɐː/, and only very rarely have I seen /ɐ/ used to transcribe the vowel of . /ɑ/ is the usual transcription for the vowel of father and bother in American accents where the two words rhyme. It's true that coastal Southern accents don't monophthongize the vowel before voiceless sounds, but considering Texas is the second-largest state in the union, it's probably fair to say the majority of people with Southern accents do say [hwaːt raːs] rather than [hwəɪt rəɪs] for white rice. But I still don't think anyone merges  with  even before voiced sounds; rad and ride are distinct in both quality and duration, as are span and spine. (However, the Southern breaking of the  vowel plays a role in avoiding homophony; bad and bide are distinct at least in part because bad is /bæɪ(ə)d/.) The Southern monophthongized  vowel is very much like the Bostonian  vowel; I knew a Bostonian named Mark who had to make phone calls to people all around the country in his job, and he complained how many people got his name wrong, because his pronunciation /maːk/ sounded just like a (non-coastal) Southerner's pronunciation of Mike. Angr (talk) 18:59, 7 April 2014 (UTC)


 * I'm no expert in IPA, but, again, I'm going by the audio IPA examples at the University of Victoria. The example for /ɐ/ given there sounds to me like the vowel in Gen. Am.  and .  I can see that it might be used to represent the vowel in RP, though for Gen. Am.  I think I'd use /ʌ/.


 * This discussion of the Southern vowel is interesting but I think it takes us too far afield for this talk page, so I'll respond on Angr's talk page when I get some time, hopefully later today.  J. D. Crutchfield &#124; Talk 16:36, 8 April 2014 (UTC)

Rhotacized schwa is neither?
The article says: "The schwa /ə/ and rhotacized schwa /ər/ are usually considered neither free nor checked because they cannot stand in stressed syllables."

What about the word "fur" /fər/ ? Or, since the article includes a bunch of diphthongs, "bear" /bɛər/? — Preceding unsigned comment added by Mechanic1c (talk • contribs) 01:04, 8 May 2016

"milieu" and "pot-au-feu"
(regarding my removal that was reverted)

@Sol505000, how are "milieu" and "pot-au-feu" examples of /ʌ/? In the source, they're given as [-u, -ɚ] and [-ɚ], respectively. — W.andrea (talk) 17:02, 2 August 2023 (UTC)


 * Do you not see they also have \-ˈyə\ and \-ˈfə\? Read the entire transcriptions, not just the ones accompanied by audio. Nardog (talk) 22:17, 2 August 2023 (UTC)


 * Oh! I don't know how I missed those. But, those are alternative pronunciations, right? Like, I don't think I've ever heard an American pronounce "milieu" as [mil'jə]. So should that be indicated, something like this? the loanword pho for when pronounced in American English, as well as sometimes milieu and pot-au-feu. — W.andrea (talk) 23:09, 2 August 2023 (UTC)
 * Sure. Nardog (talk) 00:38, 3 August 2023 (UTC)