Unstressed aspirants in Texas English

[TzirTzi]

I've known a couple of Texans and heard a couple of speakers on the radio who shared this feature. It sounds like voiceless stops being aspirated in odd contexts - like coda position in unstressed syllables. For example: <...not a very...> [nAt_h@vEr\i]. What I'm actually hearing is just an unexpectedly long period of voicelessness, so [t_h] is not necessarily the right transcription ([th]? [t@_0]? though I'm pretty sure the vowel is still there and voiced really). Equally I could be wrong about associating this with Texas - I'm English so I don't at all expect to be able to distinguish between accents of the US, I just happened to identify this in two people who I knew to be from Texas.

So is this a well recognised phenomenon? What is it?

[Whimemsz]

I live in Texas and I have no idea what you're talking about...I can't ever remember hearing this as a common feature in someone's speech. Do you have any recordings or whatever that you can link to as an example?

[TzirTzi]

I live in Texas and I have no idea what you're talking about...I can't ever remember hearing this as a common feature in someone's speech. Do you have any recordings or whatever that you can link to as an example?

I don't off hand - I'll have a look around online and see if I can find any examples. If you - as someone living in Texas - have no idea what I'm talking about then it seems quite probable that I'm either drastically mishearing this or have misassociated it with Texas.

In case it helps, one word that I found particularly salient was <water> as [wAt_hr\=] and it occurs to me than in US varieties one would normally expect a voiced tap/flap here? So is it possible that what I'm hearing is actually a careful, "correct" or spelling-based pronunciation? Something like hypercorrection which would be inconsistent and associated with high formality situations? Or would you not expect a speaker to be especially aware of this particular variable?

[finlay]

Perhaps they're conscious of t-flapping because it's not done so much in england, and try to take it out their accent.

[Melteor]

You could look for King of the Hill clips online with it, on Youtube.

[Whimemsz]

Perhaps they're conscious of t-flapping because it's not done so much in england, and try to take it out their accent.

Oh. Yeah, it's certainly relevant whether these speakers are living in Texas now, or if they're in England with you. The normal way to say "water" in Texan English would be essentially the same as GA: something like ['waɾɚ]

And actually...do these people actually speak with any sort of Texas or Southern accent, or are they just GA speakers from Texas? (some form or another of GA is by far the most common accent among younger people that I come into contact with, at least in non-rural areas)

[Travis B.]

I live in Texas and I have no idea what you're talking about...I can't ever remember hearing this as a common feature in someone's speech. Do you have any recordings or whatever that you can link to as an example?

I don't off hand - I'll have a look around online and see if I can find any examples. If you - as someone living in Texas - have no idea what I'm talking about then it seems quite probable that I'm either drastically mishearing this or have misassociated it with Texas.

In case it helps, one word that I found particularly salient was <water> as [wAt_hr\=] and it occurs to me than in US varieties one would normally expect a voiced tap/flap here? So is it possible that what I'm hearing is actually a careful, "correct" or spelling-based pronunciation? Something like hypercorrection which would be inconsistent and associated with high formality situations? Or would you not expect a speaker to be especially aware of this particular variable?

I am not a Texan myself either, but I should note that there are North American English dialects, such as my own, that do maintain a consistent phonemic distinction between /t/ and /d/ in unstressed intervocalic positions. For instance, in my own dialect the usual distinction then is [Vɾ̥] versus [Vːɾ], where [V] is any preceding vowel's quality.

Hence, in deliberately careful speech in my own dialect, one may actually see that distinction turn into [Vt] versus [Vːɾ] or even, when emphasized, [Vtʰ] versus [Vːɾ]. This is not actually a case of hypercorrection, as it reflects an already extant underlying phonemic distinction, but rather is a case of phonological changes brought about by speaking more carefully than normal.

[Radius Solis]

I've known only one native American English speaker who, I think, consistently did not have t-flapping and instead had full aspirated stops for /t/ almost everywhere, aside from maybe after /s/. Everyone noticed his unusual pronunciation of it - it was a Thing, other kids would mimic him to be funny - but it was before I knew anything about phonology/phonetics, so I can't be certain of specifics or whether it was truly universal in his speech. FWIW he had lived in Hawaii his whole life until moving to the mainland at 13, which is when I met him.

[Travis B.]

I've known only one native American English speaker who, I think, consistently did not have t-flapping and instead had full aspirated stops for /t/ almost everywhere, aside from maybe after /s/. Everyone noticed his unusual pronunciation of it - it was a Thing, other kids would mimic him to be funny - but it was before I knew anything about phonology/phonetics, so I can't be certain of specifics or whether it was truly universal in his speech. FWIW he had lived in Hawaii his whole life until moving to the mainland at 13, which is when I met him.

Yeah, the people that I hear that do actually often leave unstressed intervocalic /t/ unflapped usually leave it unaspirated unless they are specifically strongly stressing it.

In this case, I wonder whether his having lived in Hawaii might be a factor here... (i.e. adstratum influence).

In general here, I wonder how common it is for NAE dialects to actually maintain a clear unstressed intervocalic /t/ versus /d/ distinction that is not merely learned or an affectation...

[Whimemsz]

In general here, I wonder how common it is for NAE dialects to actually maintain a clear unstressed intervocalic /t/ versus /d/ distinction that is not merely learned or an affectation...

Or, if it's only in very slow, careful, emphatic speech (the case for me), a spelling pronunciation

[Travis B.]

Perhaps they're conscious of t-flapping because it's not done so much in england, and try to take it out their accent.

Oh. Yeah, it's certainly relevant whether these speakers are living in Texas now, or if they're in England with you. The normal way to say "water" in Texan English would be essentially the same as GA: something like ['waɾɚ]

And actually...do these people actually speak with any sort of Texas or Southern accent, or are they just GA speakers from Texas? (some form or another of GA is by far the most common accent among younger people that I come into contact with, at least in non-rural areas)

Aren't varieties of Texas English properly Midland dialects, not Southern ones?

Of course, the distinction between Midland dialects and General American is very, very fuzzy, as GA is in essence a somewhat standardized northern Midland dialect...

[Whimemsz]

Aren't varieties of Texas English properly Midland dialects, not Southern ones?

No, not really. Maybe in the northern part of the state to some extent, but not really in the central or southern areas (certainly not the Gulf Coast region). For those speakers in Texas who actually have a non-GA accent, the most salient phonological trait is monophthongization of /aɪ/ to [aː]~[æː], which is also the defining phonological trait of Southern English generally. Speakers of Texan English also generally have, to some degree, the Southern Shift (IME it's much more pronounced/further advanced among older rural speakers than among middle-aged people and younger).

Labov's team agrees with me here

[Travis B.]

Aren't varieties of Texas English properly Midland dialects, not Southern ones?

No, not really. Maybe in the northern part of the state to some extent, but not really in the central or southern areas (certainly not the Gulf Coast region). For those speakers in Texas who actually have a non-GA accent, the most salient phonological trait is monophthongization of /aɪ/ to [aː]~[æː], which is also the defining phonological trait of Southern English generally. Speakers of Texan English also generally have, to some degree, the Southern Shift (IME it's much more pronounced/further advanced among older rural speakers than among middle-aged people and younger).

Labov's team agrees with me here

Interesting; I was used to seeing maps where the line between southern Midland and Southern proper arcs precipitously southward such that Texas as a whole falls under southern Midland rather than Southern.

(To be entirely honest, I would not take that map as is necessarily, but for other reasons; today the features associated with Inland North extend far further west and north than they do on the map, and by all accounts Inland North essentially fades into North Central rather than there being any clear line between the two anymore.)

[TzirTzi]

Interesting stuff..

@Travis - I had indeed vaguelly imagined that most US varieties eliminated the distinction between intervocalic /t d/, so it's interesting to hear that that isn't universal. I suppose that makes it a little more likely that I'm actually hearing what I think I'm hearing (easier to recover a distinction that's still there at some level than one only available from spelling).

Sadly I haven't been able to find any examples (though I haven't had time to do an exhaustive search by any means). The two speakers I knew I've lost contact with - hearing someone with the same feature on the radio reminded me of the whole thing, but the programme isn't on listen again.

On the issue of geographical place - the two I knew lived in the UK at the time; the two on the radio, I don't know (they were interviees and only given the briefest of introductions). So it's quite possible it was some sort of correction for foreign linguistic context.

What accents they had generally, I'm a pretty bad judge, never having studied accents of North America. The one guy whom I knew better I'm pretty confident didn't have monophthongisation of /aI/ or the breaking of short front vowels, and though his short front vowels were probably all higher/tenser than mine, they didn't seem strikingly so - so if those are reasonable diagnostics of a Southern US variety then that wasn't what he was speaking. Another striking (to me) feature he exhibited was full voicing of word initial /b d g/, but again I don't know whether that's particularly diagnostic of anywhere or is a widespread AmEng feature.

Anyway, it sounds likely that it was something sociolinguistic - by which I mean a spelling pronunciation or some odd pronunciation resulting from attempts at clarity in a foreign linguistic context - which is a bit less interesting than if it was a known feature of a specific variety

[Travis B.]

In my own dialect at least, though, all this is complicated by that the history of English in this area is not entirely natural, so to speak. That is, within about the same window of time there was a massive shift from languages other than English, and especially non-Anglic West Germanic varieties and North Germanic varieties, to English.

The significance of this is that the natural history of unstressed intervocalic /t/ and /d/ in North American English could very well have been disrupted by there being a stage at which NAE was a second language to a large portion of the population, particularly as this is a feature that would be very likely to be learned wrong by speakers of other Germanic languages. This would compare with how /eɪ̯/ and /oʊ̯/ became [e] and either [o] or, intervocalically and in some varieties sometimes finally, [ow] is widespread in the Upper Midwest and how my own dialect and apparently Chicago dialect have final fortition of lenis obstruents except /b/ and /g/ (and still the usual final devoicing of those).

The reason I mention this is that I have practically never heard of a systematic preservation of an unstressed intervocalic /t/ versus /d/ contrast existing in "normal" NAE varieties from anyone - it has almost always been learned or affected at some level - yet my dialect somehow magically preserves this distinction naturally and consistently in all registers.

So hence it would only make sense that this actually unnatural after all, especially as there is already evidence of such unnatural features in the phonology (and also the syntax and usage) of my own dialect.

[Travis B.]

What accents they had generally, I'm a pretty bad judge, never having studied accents of North America. The one guy whom I knew better I'm pretty confident didn't have monophthongisation of /aI/ or the breaking of short front vowels, and though his short front vowels were probably all higher/tenser than mine, they didn't seem strikingly so - so if those are reasonable diagnostics of a Southern US variety then that wasn't what he was speaking. Another striking (to me) feature he exhibited was full voicing of word initial /b d g/, but again I don't know whether that's particularly diagnostic of anywhere or is a widespread AmEng feature.

Anyway, it sounds likely that it was something sociolinguistic - by which I mean a spelling pronunciation or some odd pronunciation resulting from attempts at clarity in a foreign linguistic context - which is a bit less interesting than if it was a known feature of a specific variety

The full-voicing of word-initial /b d dʒ ɡ/ is something that I do tend to be noticeable about more southerly North American English varieties, both Southern and southern Midland, including varieties of General American under the influence thereof, that seems to be largely overlooked as as a feature that is variable in English overall. It sticks out strongly to my ears, even though I really have never heard of it as a feature anywhere.

Of course, this is not very surprising considering my own dialect is essentially at the opposite extreme as English dialects go, among obstruents only reliably voicing lenis obstruents followed by vowels and not preceded by consonants and unstressed intervocalic /d/ and only intermittently voicing unstressed intervocalic lenis obstruents other than /d/ other than /dʒ/, which is always voiceless. And my own dialect does not seem to be out of the ordinary here; other dialects from the northern US seem to tend more towards the voiceless end of things here here, and at least dialects from the Upper Midwest particularly tend towards this extreme.

If anything, the balance between voicedness and voicelessness for lenis obstruents in English dialects seems to be a major feature of different English varieties, and one that which seems by all accounts to be very under-studied. I have really heard very little about any of this in writing, even though it seems to actually be a major set of variables from the variation in it that is apparent from individuals' actual speech combined with what English varieties they themselves speak.

[bulbaquil]

I am from Northern Texas and have never heard intervocalic /t/ pronounced as just about anything other than [ɾ], except possibly mothers enunciating very clearly to toddler children. This smacks of a dialectical hypercorrection to me.

I also pronounce /l/ as /ɫ/ in all positions, including word-initially and intervocalicly. I know a person from Louisiana who does this as well and someone from the Milwaukee area who I think does it but I'm not sure if I'm overlaying my accent onto his.

[Ducane]

A lot of university professors I have encountered in the States tend to shift "t" [ɾ] back to [tʰ] as a part of the linguistic-cultural atmosphere. This has become a habit for me when speaking in classes, and early on if I was not careful I might have pronounced a "d" as a [tʰ]. There would perhaps be nothing more disgraceful and humiliating than that.

[Chagen]

I live in the deep south of Texas, in San Antonio, and I think I make a distinction between intervocalic /t/ and /d/.

Specifically, wheh intervocalic, /t/ is /4_h/ while /d/ is just /4/.

In addition, when doing the /4_h/, it's more laminal. /4/ is just apical.

[Wattmann]

Seriously, dude, you cannot aspirate voiced consonants.
And I doubt that Texan English has breathy voice.

[Whimemsz]

Seriously, dude, you cannot aspirate voiced consonants.

My assumption is he means he devoices them (so, [ɾ] vs [ɾ̥]). But if so, I'd like a recording. I suppose it's possible, but I find it hard to believe anyone from Texas would say, for example, [bəɾ̥ɚ] and not [bəɾɚ] for "butter". It sounds very unnatural (or alternately, very British) to me.

Chagen, are you sure you're not over-analyzing your own speech here, like, hearing things that aren't there because you're being influenced by spelling? Or perhaps that's influencing you into mishearing a tonal change or vowel length distinction as a distinction in the following consonant?

[Pole, the]

A lot of university professors I have encountered in the States tend to shift "t" [ɾ] back to [tʰ] as a part of the linguistic-cultural atmosphere.

Amazing, how do people fight with their tongues.

[Chagen]

I may be over-analyzing.

Also, voiced obstuents can be aspirated. It's called breathy voice.

Another test seems to say that intervocalic /t/ for me is actually just /t_h/.

My accent is wierd. I also lower coda /k g/ to fricatives.

[Wattmann]

I may be over-analyzing.

Also, voiced obstuents can be aspirated. It's called breathy voice.

Another test seems to say that intervocalic /t/ for me is actually just /t_h/.

My accent is wierd. I also lower coda /k g/ to fricatives.

Chagen, please do read before responding:

Seriously, dude, you cannot aspirate voiced consonants.
And I doubt that Texan English has breathy voice.

Anyway, that's not aspiration, that's a phonation type

[Whimemsz]

I may be over-analyzing.

My accent is wierd. I also lower coda /k g/ to fricatives.

So what, you say [bæx] for "back"? Anyway, now I definitely want a recording. Do a recording of yourself saying the "Please Call Stella" text or something.

Dooo eeet!

Another test seems to say that intervocalic /t/ for me is actually just /t_h/.

Huh? What test are you talking about?