Stressed /ə/ in English

[Travis B.]

Typically stressed /ə/ in English seems to be identified with /ʌ/ in the popular mind, as shown by the frequent writing of /ə/ in eye dialect as <u> or <uh>.

The thing is that stressed /ə/ seems to be pronounced either [a] or [ɘ] here, depending on position, not as [ʌ]. For instance, when trying to get my daughter Alyssa's attention I pronounce her name as [ˌaːˈʟ̞ɘˌsaː] instead of the usual [ˈɰɘsəː] not as *[ˌʌːˈʟ̞ɘˌsʌː].

Does anyone else have or hear "non-standard" pronunciations of stressed /ə/ in English?

[zompist]

Could this be a spelling pronunciation? If your daughter's name was Ursula, would you emphasize it as [ur sa la]?

If someone misspelled "photograph" as "photagraph", I can imagine telling them "It's [fo to græf]", not because there's an underlying /to/ but because of the spelling.

[Soap]

I thought schwa was /ʊ/ when I was 10 years old or so, and I still think it makes sense. The only true stressed schwa in American English is in words like "pull" and "full" which in a dictionary are listed as having /ʊ/. It cant be considered just an allophone of /ʌ/ after a labial because of the minimal pair "put"/"putt".

[Travis B.]

Could this be a spelling pronunciation? If your daughter's name was Ursula, would you emphasize it as [ur sa la]?

I would pronounce it as [ˈʁ̩ˌsʲaːˌʟ̞aː].

If someone misspelled "photograph" as "photagraph", I can imagine telling them "It's [fo to græf]", not because there's an underlying /to/ but because of the spelling.

The thing is that this is under the influence of the form photo.

[zompist]

If someone misspelled "photograph" as "photagraph", I can imagine telling them "It's [fo to græf]", not because there's an underlying /to/ but because of the spelling.

The thing is that this is under the influence of the form photo.

Yes, that's the point. Alyssa begins with A.

[Travis B.]

If someone misspelled "photograph" as "photagraph", I can imagine telling them "It's [fo to græf]", not because there's an underlying /to/ but because of the spelling.

The thing is that this is under the influence of the form photo.

Yes, that's the point. Alyssa begins with A.

Except I don't normally pronounce <a> as [a] (note not [ɑ]) except in father and foreign words such as pasta and potentially a few sporadic changes of /ɔ/ to /ɑ/ pronounced [a] despite no cot-caught merger (but all of these spelled <a> that I can think of, such as swan and wasp are pronounced [ɑ], but there are cases pronounced [a] but these are spelled <au> such as astronaut); its normal pronunciation here is [ɛ]~[eɛ]~[iɛ], [ɑ] (but I note my daughter has [ɒ] instead of [ɑ] here despite not being cot-caught merged), or in some words [e]~[eɪ].

Not that I was refering to influence from related lexemes and not mere spelling pronunciation above.

[Zaarin]

...I can't think of any instance of /ə/ being stressed in my dialect. The closest instances I can think of are stressed syllabic resonants like /n̩/ or /l̩/. To use your example, the second vowel in Alyssa for me is /ɪ/ (/əˈlɪsə/).

[Cognomen]

What about pronouncing "the" with a stress, without pronouncing it like "thee"?

[Zaarin]

What about pronouncing "the" with a stress, without pronouncing it like "thee"?

I'm pretty certain that the is either /ðə/ or /ˈði/ for me, but I suppose I could imagine [ˈðə(ː)] not being outside the realm of possibility.

[Travis B.]

...I can't think of any instance of /ə/ being stressed in my dialect. The closest instances I can think of are stressed syllabic resonants like /n̩/ or /l̩/. To use your example, the second vowel in Alyssa for me is /ɪ/ (/əˈlɪsə/).

I should note that the stressed vowel in Alyssa for me is /ɪ/, which I realize as [ɘ] except before /r/.

[Curlyjimsam]

/ə/ and /ʌ/ sound similar enough for me (native southern-ish British English speaker) that I'd quite happily analyse them as the same phoneme. If you'd asked me before I had any linguistic training, I'd have said London had the same vowel sound in both syllables.

[Travis B.]

/ə/ and /ʌ/ sound similar enough for me (native southern-ish British English speaker) that I'd quite happily analyse them as the same phoneme. If you'd asked me before I had any linguistic training, I'd have said London had the same vowel sound in both syllables.

One note is that /ʌ/ is typically more like [ɐ] than true [ʌ] in most English varieties, so this is not surprising. Inland North varieties, of which I speak one, are pretty atypical in having true [ʌ] for /ʌ/.

[gmalivuk]

The only true stressed schwa in American English is in words like "pull" and "full" which in a dictionary are listed as having /ʊ/.

Isn't this rather circular?

Of course you'll conclude that schwa is /ʊ/ if you start with the assumption that the sound in words with /ʊ/ is a stressed schwa.

Incidentally, I fully agree that "put" and "putt" are not homophones, but for me the schwa is definitely closer to the latter than the former.

[Sumelic]

The only true stressed schwa in American English is in words like "pull" and "full" which in a dictionary are listed as having /ʊ/.

Isn't this rather circular?

Of course you'll conclude that schwa is /ʊ/ if you start with the assumption that the sound in words with /ʊ/ is a stressed schwa.

Incidentally, I fully agree that "put" and "putt" are not homophones, but for me the schwa is definitely closer to the latter than the former.

@Soap: The thing is that "schwa" in English does not refer to a single phone. The symbol /ə/ is used to denote unstressed and reduced vowels that occur in various contexts. They don't contrast with one another, but the different contexts usually condition different phonetic realizations.

I can distinguish at least the following general allophones in my speech corresponding to the "schwa" phoneme (as it's usually transcribed):

• The vowel at the end of words like "comma" or "Rosa" is pretty close to [ɐ]. This sounds to me essentially the same as the vowel in "strut". I'm kind of suspicious about whether this should even be grouped in with "schwa" properly speaking, since I associate schwa with neutralization of vowel contrasts, but there are a number of other vowel phonemes that contrast with commA (at least happY and grotto), and the vowel of commA/RosA contrasts with the epenthetic vowel of rosEs.
• The vowel in the second syllable of words like "parrot" is possibly actually close to being phonetic [ə]. It's merged, or near-merged, with the vowel in words like "turret" or "rabbit", which is generally identified with KIT. I would say the nearest equivalent to the last vowel of "parrot" in stressed syllables is either KIT [ɪ] or STRUT [ɐ]; I can't determine which is closer. FOOT /ʊ/ (something like [ɵ] for me, I think) is somewhat close, but not the best fit.
• Words like "button" and "chasm" have syllabic [n̩] and [m̩], I think. The nearest equivalent in stressed syllables is, again, probably KIT or STRUT followed by /n/ or /m/ respectively.
• Words like "pencil" probably have syllabic [ɫ̩], I think. I would agree that the closest equivalent to this in stressed syllables for me is /ʊl/ as in "pull"; I'm not quite sure how this is realized phonetically (possibly [pɫ̩]?).
• Words like "letter" have syllabic [ɹ̩], I think. The equivalent in stressed syllables is the "nurse" vowel, which might also just be syllabic [ɹ̩]. I don't know if it's rounded or not. I wouldn't particularly object to analyzing this phonemically as /ʊr/, since I've merged CURE into NURSE (although POOR is merged into NORTH/FORCE).

[Travis B.]

I have the following realizations of the schwa (I do not have separate schwa and schwi phonemes):

I have a general-case low realization, as in comma or Rosa, which is [ə].
I have a general-case high realization, as in parrot, turret, or rabbit, which is [ɘ]. This is identical to my stressed KIT vowel. Note that the conditions that determine whether the general-case low or the general-case high realization are used are somewhat complex, but the general case is that morpheme-final schwas are low and other schwas are high except when adjacent to a labial or a liquid and not a coronal.
I have syllabic nasal, such as in button but chasm, i.e. [n̩] and [m̩]. Also note that if the nasal is followed by a vowel, it is realized again. Note that however low realizations are commonly found adjacent to /m/ and high realizations are commonly found adjacent to /n/ and these are frequently interchangeable except that /tn/ always is realized with a syllabic nasal.
I have two realizations which are combined with a following vocalized /l/, namely [ɯ] and [ʊ], which depend upon the rounding of that adjacent to them. Also note that if the /l/ is followed by a vowel, it is realized again. Note that pull is not a stressed counterpart of this, as I analyze it as /ʊl/ due to being realized as a [ʊ] which undergoes no centralization unlike actual /ʊ/.
I have a realization combined with a following /r/, namely [ʁ̩]. Note that the generalized low realization may be used instead in some environments, such as initially. Also note that if the /r/ is followed by a vowel, it is realized again. Note that this is identical to NURSE and, aside from the glide, some cases of CURE, and indeed I analyze NURSE and the non-glide component of some cases of CURE as involving a stressed schwa.

Note that these variations are truly allophonic, and despite the examples above do not match what is indicated by the orthography or the standard language, e.g. [ə]maginary or [ɘ]ttack or transmissi[ɘ]n.

[Soap]

The only true stressed schwa in American English is in words like "pull" and "full" which in a dictionary are listed as having /ʊ/.

Isn't this rather circular?

Of course you'll conclude that schwa is /ʊ/ if you start with the assumption that the sound in words with /ʊ/ is a stressed schwa.

Incidentally, I fully agree that "put" and "putt" are not homophones, but for me the schwa is definitely closer to the latter than the former.

Well, my answer to that is that I was 10 years old, and had never heard of the Internet, but I did have access to a dictionary that used IPA. Still, isnt it true that the vowel of PULL in General American is truly [ə], or at least closer to [ə] than to [ʊ]? I've never heard a rounded vowel in that word as far as I know, nor have I heard a syllabic [l]. I'm not saying my theory is the only correct one, just that it still bothers me when people seem to automatically associate the schwa with [ʌ].

I would say for my dialect, any vowel can be unstressed. Words like "above" and "Rosa" do not have a schwa, but rather an /ʌ/, as shown by the three-way minimal pair Rosa's/roses/Rose is. It's possible, though, I admit, that "roses" and similar words actually have /ɪ/, since I can't think of a minimal pair that doesn't cross word boundaries. Still, I think the vowel in words like "excite" is separate from /ɪ/, as I would pronounce a word like "ixcite" differently if it existed.

[Sumelic]

Isn't it true that the vowel of PULL in General American is truly [ə], or at least closer to [ə] than to [ʊ]? I've never heard a rounded vowel in that word as far as I know, nor have I heard a syllabic [l]. I'm not saying my theory is the only correct one, just that it still bothers me when people seem to automatically associate the schwa with [ʌ].

I would say for my dialect, any vowel can be unstressed. Words like "above" and "Rosa" do not have a schwa, but rather an /ʌ/, as shown by the three-way minimal pair Rosa's/roses/Rose is. It's possible, though, I admit, that "roses" and similar words actually have /ɪ/, since I can't think of a minimal pair that doesn't cross word boundaries. Still, I think the vowel in words like "excite" is separate from /ɪ/, as I would pronounce a word like "ixcite" differently if it existed.

Well, I tried analyzing some of my vowels in Praat. (I haven't done this before, so I may have made some mistakes). It seems to me like the vowel I use in PULL is nearer to the vowel I use in HOOD than the vowel I use in parrOt. parrOt is definitely the closest vowel: I measured an F1 of around 300-450. PULL was more open, with F1 around 480-540. I had a hard time measuring formants for FOOT because the vowel was so short, so I decided to use HOOD with a voiced consonant in the coda. HOOD had an F1 of around 480-490. So I think PULL was nearer to HOOD than to parrOt. I don't know about the realization of schwa in other words, though: as I said, it tends to be variable, and it probably would be more open before a labial or velar consonant. Also, maybe I should have compared to a schwa before a voiced consonant like /d/ (e.g. "naked") to be consistent, but I don't feel like taking the effort to do that now.

FOOT oddly seemed to have an even more open vowel, with F1 around 491-693, or maybe even up to 800, but as I said I think I might have had problems with identifying it correctly.

[Zaarin]

Still, isnt it true that the vowel of PULL in General American is truly [ə], or at least closer to [ə] than to [ʊ]? I've never heard a rounded vowel in that word as far as I know, nor have I heard a syllabic [l].

Speaking for my GenAm dialect, pull is definitely /pʰl̩/.

[Travis B.]

I would generally think it is a bad idea to believe that one speaks General American, e.g. I wouldn't be surprised if many people here would think they speak GA, had they any idea of what GA is, when in reality they speak anything but it.

[jal]

Speaking for my GenAm dialect, pull is definitely /pʰl̩/.

[pʰl̩]


JAL

[Travis B.]

It is hard for me to intuitively think of syllabic [l̩] in English, considering how overwhelmingly l-vocalizing the dialect I am used to is in all registers (and I don't mean the weird-ass variety I speak, I mean what everyone here speaks, even people who speak relatively close to standard).

[----]

Still, isnt it true that the vowel of PULL in General American is truly [ə], or at least closer to [ə] than to [ʊ]? I've never heard a rounded vowel in that word as far as I know, nor have I heard a syllabic [l].

Speaking for my GenAm dialect, pull is definitely /pʰl̩/.

I don't believe you.

[vokzhen]

I'm a mix of North Midlands and Inland North (a couple hour's drive north of where GA is/was naturally spoken), and it definitely feels in more careful pronunciation that I have a syllabic /l/ in <pull full>, except there's still optionally some pulling in of the lips on the /p f/. If it's artificially lengthened, the vowel is /ʊ/. In normal speech before a non-coronal consonant, I'm pretty sure it often loses the coronal contact though and becomes a pure vowel in the lower uvular/upper pharyngeal region, just behind where I'd normally pronounce /q/.

[mèþru]

Shudders at memory of a literature teacher who insisted that syllabic nasals do not exist.

[Sumelic]

Still, isnt it true that the vowel of PULL in General American is truly [ə], or at least closer to [ə] than to [ʊ]? I've never heard a rounded vowel in that word as far as I know, nor have I heard a syllabic [l].

Speaking for my GenAm dialect, pull is definitely /pʰl̩/.

I don't believe you.

Shudders at memory of a literature teacher who insisted that syllabic nasals do not exist.

I don't understand why anyone would care whether random people believe or don't believe stressed [l̩] exists in some accents of American English. If there is disagreement, people should find some type of convincing evidence, like audio files, spectrogram analysis, citations of linguistic literature.

Here is a citation I found:

Is there a stressed syllabic /l/? Answer: yes, in some dialects. This may come about from historical sequences of various short vowels plus /l/. Kantner & West (1938: 328) give the pronunciation [mɫ̩k] for "milk" (from /mɪlk/). Bailey (1985: 237) has "pull" (/pʊl/) as [pɫ̩ː], and has a spectrogram supporting the claim that there is really a pure syllabic consonant here, not a vowel followed by /l/. Wells (1982: 551) describes Southern American English dialects which have a stressed syllabic velar (not velarized alveolar) lateral [ʟː] out of /ʊl/ in words like "full", "bull", and "wolf", or even out of /ʌl/ in the words "bulge" and "bulk" (though it is not clear whether these latter two words had /ʌl/ directly before they had [ʟː] , or whether /ʊl/ was an intermediate stage). Hammond (1999: 143) mentions possible pronunciations of /ʊl/ and /ʌl/ in words such as "bull" or "mull" as [l̩].

—Guenter, Joshua. 2000. "The Vowels of California English Before /r/, /l/, and /ŋ/", pp. 104-105 http://escholarship.org/uc/item/602810m9

I would be interested in seeing any evidence that suggests that Bailey and Hammond were wrong, or that realizing /ʊl/ as [ɫ̩] is confined to non-standard dialects that should not be called General American.