Transcription styles

[finlay]

And how old is he? The transcription is so dense or non-standard that I can't make sense of it, even though he's spent many paragraphs rationalizing it.

Oh, you haven't met Travis yet? His transcription is indeed dense, and non-standard, and in fact, totally unfalsifiable. I had an extended exchange with him in this thread. After a while, and as the discussion got more and more abstract, he stopped responding to me. I've come to think that his transcriptions are nothing more than a form of intellectual masturbation, rather than an actual form of scientific communication.

If that that is how you see how I transcribe things, I in kind see your position of being a mixture of sticking your fingers in your ears (you obviously had no willinginess to even listen in the first place), demanding orthodoxy with the idea that that is the only justifiable position (you never even considered the possibility that I might be right), and using the excuse if "falsifiability" when it is obvious that nothing is going to be falsifiable unless you get the person to hand over piles of transcriptions closely aligned with actual audio for anything they claim.

Uh, go on then. That is pretty much exactly what we're asking for. Even a recording now and then would be nice.

Falsifiability is a scientific principle – Rory and I have studied scientific phonetics (he's doing his phd now; I'm doing other stuff now) – and without it you could pretty much pluck anything out of the air and call it fact. You want us to believe that you "might be right"? We're more than open to the possibility – but the onus here is on you to prove it, in that case. Perhaps this means that our purposes in using transcriptions is at odds, but then I'm afraid that I don't understand what your purpose is. Is it simply to deliberately buck the trend and be a "special snowflake"? After all, you bandy around words like "orthodoxy" as if they're swearwords. This does confuse me, I have to admit. Tell us why the "orthodoxy" is objectionable to you, then!

[Travis B.]

And how old is he? The transcription is so dense or non-standard that I can't make sense of it, even though he's spent many paragraphs rationalizing it.

Oh, you haven't met Travis yet? His transcription is indeed dense, and non-standard, and in fact, totally unfalsifiable. I had an extended exchange with him in this thread. After a while, and as the discussion got more and more abstract, he stopped responding to me. I've come to think that his transcriptions are nothing more than a form of intellectual masturbation, rather than an actual form of scientific communication.

If that that is how you see how I transcribe things, I in kind see your position of being a mixture of sticking your fingers in your ears (you obviously had no willinginess to even listen in the first place), demanding orthodoxy with the idea that that is the only justifiable position (you never even considered the possibility that I might be right), and using the excuse if "falsifiability" when it is obvious that nothing is going to be falsifiable unless you get the person to hand over piles of transcriptions closely aligned with actual audio for anything they claim.

Uh, go on then. That is pretty much exactly what we're asking for. Even a recording now and then would be nice.

Falsifiability is a scientific principle – Rory and I have studied scientific phonetics (he's doing his phd now; I'm doing other stuff now) – and without it you could pretty much pluck anything out of the air and call it fact. You want us to believe that you "might be right"? We're more than open to the possibility – but the onus here is on you to prove it, in that case. Perhaps this means that our purposes in using transcriptions is at odds, but then I'm afraid that I don't understand what your purpose is. Is it simply to deliberately buck the trend and be a "special snowflake"? After all, you bandy around words like "orthodoxy" as if they're swearwords. This does confuse me, I have to admit. Tell us why the "orthodoxy" is objectionable to you, then!

It is partially it is that I have this nagging suspicion that if I did make recordings and did provide them it would not convince Rory, or you, of anything. You would listen to them and continue to say the same things you have been, or go and more overtly say "but that doesn't sound like your transcriptions", whether or not they do or not.

A big part of this, at this point, is that I do not really trust, I have no reason to trust, that those who seem to vehemently oppose anything I have to say are actually intellectually honest in it all. I am not going to put all this effort into providing audio and all that if you are going to say the same that you have been, if you are going to keep up with nagging me about my transcriptions and whatnot.

About "orthodoxy", that might not have been the best of words, but what I mean is that I strongly suspect that you are actually criticizing my transcriptions because they don't look like GA or RP, not because you actually honestly have problems with falsifiability here, i.e. all this business about what I say not being falsifiable is likely an excuse belying this real reason.

(And no, I do not seek to be a "special snowflake", it just happens that there is a big sum of a lot of little differences that happen to make what I actually speak quite different from GA, even though underlyingly it is almost a sister if not a child of GA. I just want to actually accurately transcribe how I actually say things, and I think I have been getting progressively closer to an accurate representation of it. So hence I find myself feeling like I am needlessly fighting you and Rory and like over this, simply because you have chosen to be contrarian, and only bring out things like "falsifiability" as excuses to support your contrarianness rather than as actual honest goals unto themselves.)

[Travis B.]

Hell, for that matter, I frankly haven't the faintest clue as to what about my transcriptions you even find objectionable, except that you find something about them so, and that you find them "unfalsifiable", largely because I haven't felt like giving in and providing audio providing evidence of... something. I am not sure what to even make audio samples of!

[Rory]

If that that is how you see how I transcribe things, I in kind see your position of being a mixture of sticking your fingers in your ears (you obviously had no willinginess to even listen in the first place), demanding orthodoxy with the idea that that is the only justifiable position (you never even considered the possibility that I might be right),

Travis, you have been on this forum for nearly 6 and a half years. I'd like to think in that time that I have had the chance to review your claims and consider them in light of the evidence that I have been provided with. I have listened, and asked, and listened some more, and asked some more. I've almost never received satisfactory answers, so I kept listening, hoping that maybe I'd understand more. In our last interaction, I asked you a lot of pointed questions about terminology, and you answered them well. Given the experience that I now have with your claims, my conclusion is that your transcriptions are of no scientific merit.*

Secondly, you've clearly misjudged me if you think that I am preaching orthodoxy. Many of my views are not mainstream views. Still, I can defend my views with reasoned arguments and evidence. Your views are also not mainstream, but I am yet to see any evidence in support of them. (This is not an issue of logic or rhetoric - areas that I know you are strong in - but a simple one of empiricism.)

and using the excuse if "falsifiability" when it is obvious that nothing is going to be falsifiable unless you get the person to hand over piles of audio for anything they claim.

I'm sorry, but maybe you don't understand how science works? Any claim or hypothesis should be falsifiable, and that is proved through evidence. In our case, in the absence of the ability to hook you up to an articulometer, we've asked (quite reasonably and politely, in my view) for audio. Other forumites don't get asked (as often) for audio, as their claims are not as wild as yours. Extraordinary claims require extraordinary evidence.

And I stopped responding to you because I did not want to abide by the terms you wished to impose, e.g. onerous values of falsifiable that I should not have to abide by in the first place.

As I've said, if you're not going to engage with phonetics in a serious and scientific way, I have no desire to talk about it with you. However, for the sake of the defense of the education of newbies on the board, I will continue to let everyone know not to listen to you, as you have more or less explicitly announced that you are working in an antiscientific framework. Apologies if I have misinterpreted your reluctance to provide evidence; please correct me if so.

Falsifiability is a scientific principle – Rory and I have studied scientific phonetics (he's doing his phd now; I'm doing other stuff now) – and without it you could pretty much pluck anything out of the air and call it fact. You want us to believe that you "might be right"? We're more than open to the possibility – but the onus here is on you to prove it, in that case. Perhaps this means that our purposes in using transcriptions is at odds, but then I'm afraid that I don't understand what your purpose is. Is it simply to deliberately buck the trend and be a "special snowflake"? After all, you bandy around words like "orthodoxy" as if they're swearwords. This does confuse me, I have to admit. Tell us why the "orthodoxy" is objectionable to you, then!

It is partially it is that I have this nagging suspicion that if I did make recordings and did provide them it would not convince Rory, or you, of anything. You would listen to them and continue to say the same things you have been, or go and more overtly say "but that doesn't sound like your transcriptions", whether or not they do or not.

Again, it looks like you don't understand science. If I make a claim, "I think DNA molecules are organized as a double-helix!", my peers in the scientific community will respond "go on then!" or "prove it!" or some other goading. If I turn around and say "oh, I would, but I have a suspicion that you'll just pooh-pooh my evidence. I know I'm right though!", then no-one will believe me. Indeed, I will become a laughing stock. One's peers (in this case, fellow forumites) are not necessarily friendly, and are not necessarily agreeable. That is the point. If one's theory can withstand a hundred sceptical eyes, then it is surely a strong theory. Even if it is met with mixed reviews - some agree, some disagree - that is surely many times better than making claims without providing evidence beyond your own suspicions.

Additionally, what do you have to lose? This is a forum on the internet. You're not going to lose money, or career options, or tenure. You're not even risking losing internet points™. Sure, you'll lose some time from actually making the recording - but if you're posting on a forum meant for people interested in constructed languages, you probably have a lot of time anyway. I think that this is the reason that many people have been upset at your obstinance with regards recordings.

A big part of this, at this point, is that I do not really trust, I have no reason to trust, that those who seem to vehemently oppose anything I have to say are actually intellectually honest in it all. I am not going to put all this effort into providing audio and all that if you are going to say the same that you have been, if you are going to keep up with nagging me about my transcriptions and whatnot.

I understand not trusting strangers on the internet. But we're not all strangers, and we're certainly not all anonymous. Both me and finlay use our real names here. Others do too, or their real identities are not obscured. I'm an academic, and I have a professional reputation to keep up. By being intellectually dishonest, I have nothing to gain and everything to lose.

About "orthodoxy", that might not have been the best of words, but what I mean is that I strongly suspect that you are actually criticizing my transcriptions because they don't look like GA or RP, not because you actually honestly have problems with falsifiability here, i.e. all this business about what I say not being falsifiable is likely an excuse belying this real reason.

(And no, I do not seek to be a "special snowflake", it just happens that there is a big sum of a lot of little differences that happen to make what I actually speak quite different from GA, even though underlyingly it is almost a sister if not a child of GA. I just want to actually accurately transcribe how I actually say things, and I think I have been getting progressively closer to an accurate representation of it. So hence I find myself feeling like I am needlessly fighting you and Rory and like over this, simply because you have chosen to be contrarian, and only bring out things like "falsifiability" as excuses to support your contrarianness rather than as actual honest goals unto themselves.)

Hell, for that matter, I frankly haven't the faintest clue as to what about my transcriptions you even find objectionable, except that you find something about them so, and that you find them "unfalsifiable", largely because I haven't felt like giving in and providing audio providing evidence of... something. I am not sure what to even make audio samples of!

There are two issues that I have with your transcriptions. One is empirical, the other is theoretical.

The empirical issue is similar to what most other people are saying, and prompted by your observation:

I strongly suspect that you are actually criticizing my transcriptions because they don't look like GA

The fact that your transcriptions are so alien-looking is indeed what makes them stick out and challenges us to investigate them. I have not made any claims about whether or not your transcriptions are accurate, whether or not they are right or wrong. I have simply asked for evidence. My though process goes like this:

• That doesn't look or sound like GA
• He claims to be a native speaker of an AmEng variety
• It doesn't sound like any AmEng I've ever heard before
• How interesting!

Again, extraordinary claims require extraordinary evidence.

My theoretical issue is the one I expounded at length in our most recent interaction. You write your transcriptions extremely narrowly, and this is not an appropriate use of the IPA, and it does not make sense to do so. I won't repeat my arguments here, instead, I refer the interested reader to here.

* Maybe they hold artistic merit, or something.

[Travis B.]

Hell, for that matter, I frankly haven't the faintest clue as to what about my transcriptions you even find objectionable, except that you find something about them so, and that you find them "unfalsifiable", largely because I haven't felt like giving in and providing audio providing evidence of... something. I am not sure what to even make audio samples of!

There are two issues that I have with your transcriptions. One is empirical, the other is theoretical.

The empirical issue is similar to what most other people are saying, and prompted by your observation:

I strongly suspect that you are actually criticizing my transcriptions because they don't look like GA

The fact that your transcriptions are so alien-looking is indeed what makes them stick out and challenges us to investigate them. I have not made any claims about whether or not your transcriptions are accurate, whether or not they are right or wrong. I have simply asked for evidence. My though process goes like this:

• That doesn't look or sound like GA
• He claims to be a native speaker of an AmEng variety
• It doesn't sound like any AmEng I've ever heard before
• How interesting!

Again, extraordinary claims require extraordinary evidence.

To be entirely honest, this would be a more credible position coming from an American who has been around a lot, rather than from a Brit.

I have heard other Americans not speaking AAVE yet clearly speaking some Anglic variety who I have not understood (as in like speaking-another-language not understood). (They were two European American guys my age from North Carolina, former coworkers particularly.) I have heard really odd dialects, and that is by my standards, in parts of northern Wisconsin, both ones spoken by European Americans and ones spoken by American Indians.

Then, on the other end of things in my own perceptual sphere, I have heard many Americans speak dialects only different by a small amount from my own, such as many people from Chicago, from southern Michigan, or from upstate New York.

I do not really consider it fair for a Brit to consider everything an American posts that does not look like General American questionable and to demand proof.

That said, my own accent does tend to raise eyebrows for many people who hear it for the first time, and that is amongst Americans. When I first moved out here to Maryland, I got more than a few comments made about it, and looks made by people who did not actually say anything.

(It is not me having a speech impediment or something, either, as my ex, who moved out here with me but moved back later, said she got quite a few comments when she was out by herself here as well.)

My theoretical issue is the one I expounded at length in our most recent interaction. You write your transcriptions extremely narrowly, and this is not an appropriate use of the IPA, and it does not make sense to do so. I won't repeat my arguments here, instead, I refer the interested reader to here.

* Maybe they hold artistic merit, or something.

This is partly why I started this particular thread in the first place, that I thought there might be some validity in the point that there might really be no difference between voiceless lenis and unaspirated, unpreglottalized voiceless fortis consonants in my own speech and that, in many positions, it might be essentially orthographic influence in my head (or me confusing phonemes with phones when making transcriptions a bit), and even where it is occasionally apparent, it might not be worth actually transcribing. (I do hear some who do maintain this distinction, but at least in the particular variety I speak and many in the suburb I grew up in speak, this distinction is seemingly questionable in everyday speech, even if it may be apparent in careful speech.)

[Bristel]

I'm an American who speaks Pacific Northwest English, a variety with the diphthongization of the LEG/EGG vowel and some creaky voice possibly.

I would transcribe this, but I don't trust my own bias.

Now, please. I'm one American who'd like to analyze some of your claims to your transcriptions.

I'm sure there are others here who are intrigued by your dialect and would like to hear a couple short sentences and a transcription.

[Travis B.]

Okay this is me talking on and on to myself for five full, albeit gap-filled, minutes.

[Ziz]

Okay this is me talking on and on to myself for five full, albeit gap-filled, minutes.

Your realization of /l/ is completely bizarre, and sounds sort of like a speech impediment, but nevertheless you had a good [l] at the very end in like. And elision has a short i, rather than a long one.

Sometimes you're pretty hard to understand, too...

[Travis B.]

Okay this is me talking on and on to myself for five full, albeit gap-filled, minutes.

Your realization of /l/ is completely bizarre, and sounds sort of like a speech impediment, but nevertheless you had a good [l] at the very end in like.

It is bizarre, but I assure you that it is a dialect feature and not a speech impediment.

That in that like, though, was no [l], however - it was a [ʟ̞] that happened to be fully lateral for once, all the way to the tip. Actual [l] is a foreign speech sound, albeit one I can easily enunciate, to me.

And elision has a short i, rather than a long one.

I was not talking about traditional English vowels, and yeah, I might have pronounced it wrong, but I was talking about actual realized vowel length.

Sometimes you're pretty hard to understand, too...

I usually assume it's all the elisions combined with my pronunciation of /l/ that tend to really cause people problems more than anything else.

[Bob Johnson]

Oh hey evidence. Most of the R and Ls just sound like [w] to me, i.e. rounded labiovelar approximants. The <like> at the end is different, but it's not a proper L; it's rounded at least, and something else is different; I'm tempted to say it's pharyngealized but I can't really tell that from uvular or velar anyway.

I've heard exactly one person who sounds anything like that, but I figured it was pathological.

also I think you mean "extemporaneous". Not sure what the first word you're stumbling over is. I can figure it out from context in the normal speech but the disordered speech just turns to mush.

[finlay]

I do not really consider it fair for a Brit to consider everything an American posts that does not look like General American questionable and to demand proof.

Why not? We speak the same language. And it's perhaps worth noting that Rory's lived in Ohio for the past two or three years, and just doesn't bother updating his location field on the board...

[Rory]

Travis, I believe the word you were looking for throughout most of the recording was extemporaneous. Thanks for providing a recording, I'll have a more detailed look at it when I have time. You're right, you certainly have a weird /l/ and a weird /r/. Contra Bob Johnson, they're not just [w], though.

[Travis B.]

I do not really consider it fair for a Brit to consider everything an American posts that does not look like General American questionable and to demand proof.

Why not? We speak the same language. And it's perhaps worth noting that Rory's lived in Ohio for the past two or three years, and just doesn't bother updating his location field on the board...

I did not know he's lived in Ohio recently, but a big part of what I was saying was that one needs to consider how familiar one personally is with the variations found in a given dialect group. This is why I tend to reserve judgement on English English, Scottish English, Irish English, and even Southern NAE and traditional Atlantic NAE dialects.

[Bristel]

I see that the /ɰ/ which Travis uses in prevalent in some of his transcriptions is accurate, for the most part. But I'm not very good at writing transcriptions.

He has /ilaɪʒən/ for /əlɪʒən/, which pops out, but isn't a weird pronunciation for a word which isn't commonly spoken aloud.

He had /ɛdɪtɔwiəl/ [ɛdɪtoɰiəl]? for <editorial>.

He has /l r/ vocalizations all over (I think), but not completely, such as in the word rime it is a more standard GA pronunciation.

<extrataneuously>?

/wændəm/?

/θri θəɹɾi/?

/ɬiɹɪŋ/ [ɬiəɹiŋ]? This is sometimes heard in some words with /h/ or /ʃ/.

[Travis B.]

<extrataneuously>?

That is me trying to say extemporaneously and failing miserably.

[Rory]

Okay, so Travis has graciously given us a recording, and noted that it's probably different from his transcriptions. Listening to it, it's clear that he's very aware of his own voice and of what he's saying, and that he's somewhat uncomfortable. So, naturally, any conclusions that we draw from this recording are not necessarily generalizable to the rest of his speech. Also, note that my issues with Travis's transcription are based on both empirical and theoretical grounds. A recording could satisfy my empirical concerns, but my theoretical concerns remain valid. In this post, I'll be discussing the two of these in tandem.

I've transcribed two portions of the recording, chosen randomly. I've highlighted particularly interesting aspects of the transcription to demonstrate various points along the way. Executive summary: Travis's /l/s are "strange" (i.e. different from GenAm), and some of his /r/s are also odd. He does a lot of phonetic reduction, which is not unusual for someone of his age, and exhibits some influence of the Northern Cities Vowel Shift. On the theoretical side of things, IPA cannot adequately display many of the finer-grained distinctions that both he and I are trying to talk about.

From 187s to 197s:
goʊɪŋə̆s ɑ̃n ə̃ɑ̃ə̃ɑ̃ (>weird stuff<) ɰaɪk ðɪs n̩ (>pause<) biŋ vɛɹḭ (>pause<) vɻʷɛ̃æ̃ndəm
going on and on and on (>weird stuff<) like this and ... being very ... v-random

If I'd transcribed the content in "weird stuff", it might look something like this: [ə̰̃ɑ̰̃ə̰̃ɑ̰̃ə̰̃ɑ̰̃]. Travis is saying on and on and on and on and so on. It's all creaky (in contrast with the early three ons, which were modally voiced. It's also all nasal. But the transcription that I could have given doesn't adequately describe what's going on. First of all, the vowel variation is not as strong as the transcription suggests. Travis is not jumping from central schwa to cardinal [ɑ] and back again. In fact, the vowel distinctions are so slight it's probably not even worth transcribing. The real variance, however, is in the nasality. Travis is increasing and decreasing the nasality, oscillating up and down, as we go through the words. I imagine an X-ray or ultrasound would have shown his velum to have been merrily flapping away.

This is interesting not because it's weird (it's not), but rather because it illustrates one of the drawbacks of the IPA. I want to label the whole stretch as nasalized, but I also want to point out that some of it is more nasalized than other bits; I want to note that the vowels, fairly rapidly, lose most of their distinguishing features; I want to highlight the interesting variation in degree of creakiness; but I cannot do any of that. With a categorical "creaky" vs "not creaky" and "nasalized" vs "not nasalized" distinction, I can't do this. The IPA is not designed for describing continuously variable phonetic properties. Using it to do so is useless.

The /l/ in like is almost definitely a velar approximant, with very little to no tongue contact with the roof of the mouth. (A palatogram would answer that question for us. If I ever meet Travis in person, I'd be happy to perform one on him - they're quite easy to do.) Cool!

The first very is notable because the /r/-sound here is not significantly different from a Standard American /r/. I transcribed it as such. The creaky voice at the end of the phrase also doesn't sound unusual to me.

The /r/ at the beginning of random, however, is quite different. There's significant [v] articulation to begin with, and I suspect this is a speech error - Travis had begun to say very, but stopped himself and transitioned smoothly into random. If he can recall his metacognition at the time, perhaps he can let us know if that's true. Anyway, the /r/ itself sounds labialized, and perhaps retroflex. I don't see any evidence for "epiglottalization"; although given that I'm not entirely sure what that means, it's hard to look for such evidence.

From 132s to 142s:
ɹʷiɫʷaes ae pʰɹãas ɪʔ vɔɪsʟəs (pause) tʰɤ dʒʌs naʊ (pause) bʌɾ ɪʔs stɪɤɰ ðə vaʊɰ ɪs ɰɑŋ
realized (that) I pronounce it voiceless (pause) til just now (pause) but it's, still, the vowel is long

I've decided to stop transcribing creaky voice here, as it's occurring fairly regularly at the end of a phrase.

Travis has really interesting l-vocalization. I haven't studied that much, so I can't really say how similar it is to other types, but it's certainly very interesting. In realize and voiceless, the /l/ is definitely lateral, but later productions there either is no /l/ pronounced (as we see in til) or it is not lateral at all (as in vowel). Again, we struggle to use IPA to adequately describe this. There is definitely a cline of lateralization, and I've used the symbols [ɫ], [ʟ], and [ɰ] to represent steps on that cline, but there are no discrete steps, and each production is incrementally at a stage that is unique from all others. It's like trying to describe real numbers using only integers.

Similarly, the first glottal stop (in it) is easy to hear, whereas the second, in it's is much weaker, and in fact I could be persuaded that it isn't even there at all. They're both a lot weaker than glottal stops that I have in my native dialect; but again, IPA has no way for us to encode this information. (Also, when does something stop being a glottal pulse and start being a glottal stop? There's no clear boundary.)

I'd love to transcribe more at this point, but it takes me a while to do this, and various real-world constraints are demanding my time. Travis, you've satisfied me that you have unusual liquids, but I still take issue with your hyper-precise transcription.

[Travis B.]

Some things I should note here that is not really obvious from my transcriptions:

My usual /r/ and /l/ transcriptions are rather idealized in that they are ones that are typically found in words sound clearly, albeit in register, in isolation, and they overlook some variation.

For instance, while I do always have some level of dorsal articulation of my /r/s, even if it is not always apparent, I do notice in practice I may at times have more postalveolar articulation than is always obvious from my transcriptions. In particular, I tend to "link" initial /r/ with final consonants in the preceding word, applying my allophonic postalveolar articulation when applicable there, but I seem to also sporadically do it elsewhere in syllable onsets despite tending to not transcribe it (as I rarely do so when I say individual words carefully).

(Though I did miss that I can also have intervocalically what I would transcribe as [ɹ̠͡ɰˤ]; I had figured that I only had what I would transcribe as [ʁˤ] intervocalically. Now what I wonder is whether this is purely free variation or is there something else conditioning it, which may have resulted in my missing it, such as a labial onset for the preceding syllable.)

Note that my /r/s can be quite open in their articulation, which might give the impression at times that I actually have r-vocalization, even to myself, but I do not as I still have actual POA even when I pronounce them this way.

It might be a phonetic trick "in my head" (literally) that I seem to hear the dorsal component of sounds much more than others do, even when there actually is at times a postalveolar component where I would not expect it. Likewise, I can in my head hear the epiglottalization very prominently when it is strong, and can deliberately make it stronger or lighter and hear the result to myself, but it could be more apparent to myself than it is to others.

It is correct that I only have a lateral /l/ in stressed syllable onsets ever, but I also change how I pronounce it dramatically, based both on stress and on register but also on a signficant degree of free variation. This is why in some onsets I have a very clearly lateral /l/ while in others I have no lateralness at all. In my transcriptions I try to be at least a bit consistent, which does overlook a lot of this variation to a good degree.

[Nortaneous]

The /r/ at the beginning of random, however, is quite different. There's significant [v] articulation to begin with, and I suspect this is a speech error - Travis had begun to say very, but stopped himself and transitioned smoothly into random.

That's possible, but not necessarily the only explanation. I know someone who, despite otherwise speaking fairly standard AmE, has some sort of labiodental frication on some /r/, although I've noticed it more before high front vowels than in any other environment. And it doesn't come off as a speech impediment, at least to me.

[AnTeallach]

Rory will be better able to say anything about this than I am, but the vowels in Travis's recording really don't sound as odd to me as Travis's transcriptions suggest they ought to. In particular his DRESS vowel sounds perfectly normal to me, and I'd have thought it wouldn't if the [ɜ] transcription were really justified. Maybe my ears are adjusting too easily; strange things can happen with how people perceive sounds in other people's speech .

Travis: have you actually looked at formants of your vowels in Praat or some similar package?

[Travis B.]

Rory will be better able to say anything about this than I am, but the vowels in Travis's recording really don't sound as odd to me as Travis's transcriptions suggest they ought to. In particular his DRESS vowel sounds perfectly normal to me, and I'd have thought it wouldn't if the [ɜ] transcription were really justified. Maybe my ears are adjusting too easily; strange things can happen with how people perceive sounds in other people's speech .

Travis: have you actually looked at formants of your vowels in Praat or some similar package?

No, I have not looked at my vowels with Praat at all.

However, I should note that my vowel transcriptions are generally idealized to some extent. Particularly, my DRESS vowel is often front-central rather than, well, central-central as [ɜ] taken at face value would indicate, particularly in more careful or stressed speech. Note that this is not consistent, as often in informal speech I do find it becoming more truly central where in more careful speech it is only partly centralized.

Regardless, I still use [ɜ] because I need to still demonstrate that it contrasts with my TRAP vowel, which I transcribe as [ɛ] rather than as [æ] due to it being just that high (excluding where it breaks into a diphthong [e̯ɛ] or even [i̯ɛ], which I generally do not transcribe), and that the nature of this contrast is usually a frontness contrast rather than a height contrast. This is not an oddity of my own speech, I should note, as I even remember reading articles, which I unfortunately cannot cite here, noting how the contrast between the DRESS and TRAP vowels in southern Wisconsin is a frontness contrast rather than a height contrast and that which is higher than the other is often highly inconsistent in practice, with the DRESS vowel not infrequently being lower than the TRAP vowel for many.