English questions

[vlad]

I've noticed that in my English there's three different types of postalveolars, but I don't know how to describe their articulation or what symbols to use to represent them. The terminology used for sibilants is pretty complicated (what is "domed"? "palato-alveolar" and "alveolo-palatal" are apparently different things?), and there are a lot of phonetic issues that even phoneticians don't seem to agree on.

The first type is the default (e.g. chain, Jane, shed, Asia). These sounds are not labialized, despite most descriptions of English phonetics saying they are (Wikipedia even says "strongly labialized").

The second type is found before /r/ in the same syllable (e.g. train, drain, shred). These sounds are labialized, and the /r/ is also labialized.

There is also an intermediate type, which is found before /r/ across a syllable/word boundary (e.g. switchroom, hedgerow, mushroom, beige room). These have the same tongue position as the second type, but no labialization (and the /r/ is labiodentalized rather than labialized).

Questions: Does anyone else have a similar distribution of allophones? Or can anyone confirm that their /tʃ dʒ ʃ ʒ/ are consistently labialized, as claimed?

What would be the best way to figure out what these sounds actually are? Is there an acoustic way to distinguish between different types of sounds, or would it require something esoteric like x-rays or electropalatograms?

[Estav]

I've noticed that in my English there's three different types of postalveolars, but I don't know how to describe their articulation or what symbols to use to represent them. The terminology used for sibilants is pretty complicated (what is "domed"? "palato-alveolar" and "alveolo-palatal" are apparently different things?), and there are a lot of phonetic issues that even phoneticians don't seem to agree on.

The first type is the default (e.g. chain, Jane, shed, Asia). These sounds are not labialized, despite most descriptions of English phonetics saying they are (Wikipedia even says "strongly labialized").

The second type is found before /r/ in the same syllable (e.g. train, drain, shred). These sounds are labialized, and the /r/ is also labialized.

There is also an intermediate type, which is found before /r/ across a syllable/word boundary (e.g. switchroom, hedgerow, mushroom, beige room). These have the same tongue position as the second type, but no labialization (and the /r/ is labiodentalized rather than labialized).

Questions: Does anyone else have a similar distribution of allophones? Or can anyone confirm that their /tʃ dʒ ʃ ʒ/ are consistently labialized, as claimed?

What would be the best way to figure out what these sounds actually are? Is there an acoustic way to distinguish between different types of sounds, or would it require something esoteric like x-rays or electropalatograms?

Labialization is one of the easiest articulatory gestures to observe, you can do it with a mirror! But I don't know exactly what different types of lip positions are called or what the correspond to acoustically. I think my lips are positioned differently for all of /s/, /ʃ/, /w/ and /r/. For /s/, my lips are spread and not brought together at all—e.g. in "said", "seat" there seems to be no/minimal movement of the lips between the /s/ and the following vowel. For /w/ (as in "wed", "wheat"), my lips are rounded and brought very close together, nearly touching. For /ʃ/, my lips are brought closer together than for /s/, but not as close as for /w/. Maybe it is what is called "compressed" labialization? I would not describe it as strongly labialized, but I don't think it's totally unlabialized like /s/. For /r/ ("red", "read"), I have a labiodental coarticulation.

[vlad]

Labialization is one of the easiest articulatory gestures to observe, you can do it with a mirror! But I don't know exactly what different types of lip positions are called or what the correspond to acoustically.

The tongue position is the part I was asking about.

[bradrn]

The first type is the default (e.g. chain, Jane, shed, Asia). These sounds are not labialized, despite most descriptions of English phonetics saying they are (Wikipedia even says "strongly labialized").

The second type is found before /r/ in the same syllable (e.g. train, drain, shred). These sounds are labialized, and the /r/ is also labialized.

There is also an intermediate type, which is found before /r/ across a syllable/word boundary (e.g. switchroom, hedgerow, mushroom, beige room). These have the same tongue position as the second type, but no labialization (and the /r/ is labiodentalized rather than labialized).

Questions: Does anyone else have a similar distribution of allophones? Or can anyone confirm that their /tʃ dʒ ʃ ʒ/ are consistently labialized, as claimed?

Saying the same words, I think I have something similar. My chain etc. do seem to be slightly labialised, but not very much: [tʃʷ]ain to me sounds more like train. However, I’m pretty sure the relevant sounds are all the same in terms of tongue position.

What would be the best way to figure out what these sounds actually are? Is there an acoustic way to distinguish between different types of sounds, or would it require something esoteric like x-rays or electropalatograms?

Acoustically, I guess it might be possible to record yourself saying [tʃ tʃʷ tʃʷʷ] with various different degrees of labialisation, and compare the resulting spectrograms to spectrograms of the words in question. But really this is an articulatory question, and it requires articulatory tools — though as Estav mentioned, something as simple as taking a video of your lips could suffice here. From what I’ve seen, tongue position measurements do require either X-ray or palatographic imaging.

[Travis B.]

I have noticed that there are certain words in the English I am familiar with that are normally pronounced with non-morphological geminates, particularly raccoon [ˌʁʷˤɛʔk̚ˈkʰũ(ː)n]. What might account for this, since Modern English is not supposed to have geminates that cannot be accounted for by either morphology or cluster assimilation?

[bradrn]

I have noticed that there are certain words in the English I am familiar with that are normally pronounced with non-morphological geminates, particularly raccoon [ˌʁʷˤɛʔk̚ˈkʰũ(ː)n]. What might account for this, since Modern English is not supposed to have geminates that cannot be accounted for by either morphology or cluster assimilation?

What other words have this? For me raccoon is non-geminate (not that I ever talk about raccoons anyway).

[Travis B.]

I have noticed that there are certain words in the English I am familiar with that are normally pronounced with non-morphological geminates, particularly raccoon [ˌʁʷˤɛʔk̚ˈkʰũ(ː)n]. What might account for this, since Modern English is not supposed to have geminates that cannot be accounted for by either morphology or cluster assimilation?

What other words have this? For me raccoon is non-geminate (not that I ever talk about raccoons anyway).

As I have mentioned here before, foreign names with intervocalic /t/ are often pronounced with geminates, such as Kyoto [ˈkʰjɵʔtːɵ(ː)], but that is to avoid flapping or elision, which does not apply to raccoon, since intervocalic /k/ is never flapped or elided (and even if it were, it would not here since flapping/elision does not apply to the onsets of stressed syllables). Then there are words with non-morphological geminates in them derived from cluster assimilation, such as vodka [ˈvaːkːə(ː)], problem [ˈpʰʁˤaːmː], and federal [ˈfɜːʁˤːɯ(ː)]. However, raccoon is not one of these, since it never had a cluster to assimilate.

[Travis B.]

One note - I'd analyze vodka as /ˈvɑɡkə/, problem as /ˈprɑbm/, and federal /ˈfɛdrəl/ because these make sense phonological sense - vodka cannot be */ˈvɑkkə/ due to vowel length and lack of preglottalization, problem cannot be */ˈprɑmm/ due to a lack of vowel nasalization, and /dr/ makes sense in federal because here [d] readily surfaces in alternation with the geminate pronunciation (but it is never pronounced with [tʃɻʁ] (there should be a tie bar above the [ɻʁ]), which implies that /dʒr/ and /dr/ are phonemically distinct).

[Darren]

I have noticed that there are certain words in the English I am familiar with that are normally pronounced with non-morphological geminates, particularly raccoon [ˌʁʷˤɛʔk̚ˈkʰũ(ː)n]. What might account for this, since Modern English is not supposed to have geminates that cannot be accounted for by either morphology or cluster assimilation?

It might be something to do with lax vowels, since the TRAP vowel usually likes closed syllables. I'm trying to think of any other words with unstressed /æ/ in an open syllable to compare, but right now I can't think of any good ones (perhaps ballerina?). Can you think of any other words with spontaneous gemination that can't be explained by borrowing or cluster reduction?

[Travis B.]

I have noticed that there are certain words in the English I am familiar with that are normally pronounced with non-morphological geminates, particularly raccoon [ˌʁʷˤɛʔk̚ˈkʰũ(ː)n]. What might account for this, since Modern English is not supposed to have geminates that cannot be accounted for by either morphology or cluster assimilation?

It might be something to do with lax vowels, since the TRAP vowel usually likes closed syllables. I'm trying to think of any other words with unstressed /æ/ in an open syllable to compare, but right now I can't think of any good ones (perhaps ballerina?). Can you think of any other words with spontaneous gemination that can't be explained by borrowing or cluster reduction?

I cannot think of any other words off the top of my head which have geminates that cannot be accounted for by borrowing (e.g. pizza) or cluster reduction; in this regard, raccoon is a very odd word.

For me both syllables in raccoon are stressed, just the second syllable has primary stress. This is not surprising at one level, since unstressed /æ/ is simply verboten in my dialect as far as I can tell. At the same time, it is very odd, because monomorphemic disyllables typically do not have both syllables stressed for me. Still, stressed /æ/ is normally permitted in open syllables, or more properly before ambisyllabic consonants.

My guess is pretty much the same as yours, i.e. is that it is somehow due to the oddity of having two stressed syllables - the stressed lax vowel, as you mention, of the first syllable makes it want to have either a coda or a following ambisyllabic consonant, but the fact that the second syllable has primary stress means that its onset cannot be ambisyllabic, so an extra /k/ is added just to satisfy the need for either a coda or ambisyllabicy.

[zompist]

The first type is the default (e.g. chain, Jane, shed, Asia). These sounds are not labialized, despite most descriptions of English phonetics saying they are (Wikipedia even says "strongly labialized").

The second type is found before /r/ in the same syllable (e.g. train, drain, shred). These sounds are labialized, and the /r/ is also labialized.

I just checked this, using a mirror, to the point where I'm not sure what happens anymore. I'm think my [tʃ dʒ ʃ ʒ] are labialized or not depending on the following vowel: strongly for /u/, very lightly for [æ], not at all for [ɪ].

My [ʃ] is apical and maybe a bit retroflexed, and post-alveolar.

[Travis B.]

I must be weird (as if that were not already obvious) because not only my /tʃ dʒ ʃ ʒ/ but also my /s z/ are labialized before /u/ [y] (they are not labialized before my /æ ɪ/). (As I have mentioned before, my /tʃ dʒ ʃ ʒ s z/ are all also palatalized before /u/, as [tɕʷ dʑʷ ɕʷ ʑʷ sʲʷ zʲʷ].) My /tʃr dʒr ʃr/ are all not labialized, but my /r/ is labialized initially (but in no other positions).

[anteallach]

I've noticed that in my English there's three different types of postalveolars, but I don't know how to describe their articulation or what symbols to use to represent them. The terminology used for sibilants is pretty complicated (what is "domed"? "palato-alveolar" and "alveolo-palatal" are apparently different things?), and there are a lot of phonetic issues that even phoneticians don't seem to agree on.

I tend to go with Ladefoged and Maddieson (The Sounds of the World's Languages) here, partly because of their reputation and partly because what they describe tends to fit well with what I feel my own mouth doing.

My understanding of what they say is that for postalveolars there are two main variables, whether they are apical or laminal and their degree of palatalisation, and that they distinguish three degrees of palatalisation (not palatalised, or "flat", weakly palatalised, or "domed", and fully palatalised). Of course there are also secondary articulations other than palatalisation.

They take "palato-alveolar" as corresponding to their type 6, an apical or laminal domed postalveolar like English sh, and "alveo-palatal" as corresponding to their type 8, a laminal palatalised postalveolar like Polish ś) Other postalveolar types in their table are type 4, "laminal flat", like Polish sz, type 5, apical postalveolar, type 7, a definitely laminal domed postalveolar, and type 9, a laminal postalveolar with the tongue tip position of a laminal alveolar found in North West Caucasian. Complicated indeed, especially if you want to try to learn Ubykh.

The first type is the default (e.g. chain, Jane, shed, Asia). These sounds are not labialized, despite most descriptions of English phonetics saying they are (Wikipedia even says "strongly labialized").

The second type is found before /r/ in the same syllable (e.g. train, drain, shred). These sounds are labialized, and the /r/ is also labialized.

There is also an intermediate type, which is found before /r/ across a syllable/word boundary (e.g. switchroom, hedgerow, mushroom, beige room). These have the same tongue position as the second type, but no labialization (and the /r/ is labiodentalized rather than labialized).

Questions: Does anyone else have a similar distribution of allophones? Or can anyone confirm that their /tʃ dʒ ʃ ʒ/ are consistently labialized, as claimed?

Mine have rounded lips, though I don't think I'd go for "strongly labialised".

With the caveat of the dangers of self-analysis, my impression is that my "default" type (not before /r/) are basically as Ladefoged and Maddieson describe them: weakly rounded domed postalveolars. For me the fricatives are usually apical and the affricates laminal; except I think in words with historic /tj/ and /dj/ where the merger with the original affricates isn't quite complete and I seem to use apical affricates (at least sometimes).

The ones before /r/, and I think this applies, at least within a word, regardless of whether they come from original /tr/ and /dr/ clusters as opposed to where historic affricates have ended up being before /r/ (dangerous, which is bisyllabic, hedgerow) seem to me to be less palatalised, like the description of laminal flat postalveolars, L & M's Type 4. Indeed I feel the /n/ is noticeably less palatalised in dangerous than in danger. But note that caveat about self-analysis, and I'm not that confident about hedgerow.

What would be the best way to figure out what these sounds actually are? Is there an acoustic way to distinguish between different types of sounds, or would it require something esoteric like x-rays or electropalatograms?

Section 5.4 of L & M talks about acoustic differences. I'm not sure, though that there's enough there to be confident of how to spot certain subtle differences like those above, and the x-ray tracings earlier in chapter 5 look more useful. I know that some time ago I tried to see if I could spot the difference between my dune and June and struggled, even though I knew my tongue did different things in the two recordings.

[Travis B.]

I've noticed that in my English there's three different types of postalveolars, but I don't know how to describe their articulation or what symbols to use to represent them. The terminology used for sibilants is pretty complicated (what is "domed"? "palato-alveolar" and "alveolo-palatal" are apparently different things?), and there are a lot of phonetic issues that even phoneticians don't seem to agree on.

I tend to go with Ladefoged and Maddieson (The Sounds of the World's Languages) here, partly because of their reputation and partly because what they describe tends to fit well with what I feel my own mouth doing.

My understanding of what they say is that for postalveolars there are two main variables, whether they are apical or laminal and their degree of palatalisation, and that they distinguish three degrees of palatalisation (not palatalised, or "flat", weakly palatalised, or "domed", and fully palatalised). Of course there are also secondary articulations other than palatalisation.

They take "palato-alveolar" as corresponding to their type 6, an apical or laminal domed postalveolar like English sh, and "alveo-palatal" as corresponding to their type 8, a laminal palatalised postalveolar like Polish ś) Other postalveolar types in their table are type 4, "laminal flat", like Polish sz, type 5, apical postalveolar, type 7, a definitely laminal domed postalveolar, and type 9, a laminal postalveolar with the tongue tip position of a laminal alveolar found in North West Caucasian. Complicated indeed, especially if you want to try to learn Ubykh.

I have three parallel sets of alveolars and postalveolars, a non-palatalized series, a palatalized series, and (as I had not realized previously) a labialized palatalized series. The palatalized series arises from coronals preceding /ɜr~ər/ and many consonant clusters such as /st/ and /rs/. The labialized palatalized series arises from coronals preceding /w ʊ u/. However, some of these consonants are not stable IMD, particularly palatalized and labialized palatalized /t/ tends to affricate outside of the cluster /st/ (but does often affricate in some words with historical /st/ such as sister and yesterday). (It tends to become [tɕʰ tɕʷʰ] even though I myself am odd for favoring [tsʰ tsʷʰ] in some words.). Furthermore /s/ readily becomes postalveolar before [tʃ tɕ] and palatalization spreads to adjacent coronals, and /t d/ usually become postalveolar but not palatalized before /r/ (but in some words this fails to occur, and also if /s/ precedes these both the /s/ and the /t/ or /d/ become palatalized), to make things more complex.

[Travis B.]

I just realized I pronounce /ɔɪ/ as [ɔʏ̯] in English - my rounding does not change throughout the whole diphthong. I for the longest time wondered how StG /ɔʏ/ differed from English /ɔɪ/ because I could not perceive a difference...

[anteallach]

I just realized I pronounce /ɔɪ/ as [ɔʏ̯] in English - my rounding does not change throughout the whole diphthong. I for the longest time wondered how StG /ɔʏ/ differed from English /ɔɪ/ because I could not perceive a difference...

Whereas I, though my brain interprets them as essentially the same sound, definitely have more rounding at the end of the German diphthong than the English one. (I find this interesting: it suggests my German phonetics are better than might be expected from four years of a few classes a week at secondary school when I didn't know much about phonetics, and I don't think the difference between the two sounds was mentioned.)

[Travis B.]

I just realized I pronounce /ɔɪ/ as [ɔʏ̯] in English - my rounding does not change throughout the whole diphthong. I for the longest time wondered how StG /ɔʏ/ differed from English /ɔɪ/ because I could not perceive a difference...

Whereas I, though my brain interprets them as essentially the same sound, definitely have more rounding at the end of the German diphthong than the English one. (I find this interesting: it suggests my German phonetics are better than might be expected from four years of a few classes a week at secondary school when I didn't know much about phonetics, and I don't think the difference between the two sounds was mentioned.)

In more recent times I have tried to realize StG /ɔʏ/ as more rounded than my native /ɔɪ/, but I now wonder whether that is just hypercorrection or not.

[Travis B.]

I was listening to the song Proud Mary by CCR and noticed that Mary in the song was repeatedly pronounced as [ˈmɛːɻə], such that I did not recognize it as being Mary when I first heard the song. The thing, though, is that John Fogerty did not sing the song in his native variety; he was from California, but he adopted affectations of New Orleans English (aka "Yat"), as shown by his pronunciations of turning and burning with [ɜɪ̯] (which many people identify with traditional New York English, even though John Fogerty was not trying to emulate New York English). So are there any NAE varieties which actually regularly merge HAPPY and COMMA as John Fogerty does here, or is this just a weirdness that John Fogerty added of his own accord?

Edit: in other parts of the song I notice he pronounces city as both [ˈsɪɾeɪ̯]~[ˈsɪɾi].

[bradrn]

I've noticed that in my English there's three different types of postalveolars, but I don't know how to describe their articulation or what symbols to use to represent them. The terminology used for sibilants is pretty complicated (what is "domed"? "palato-alveolar" and "alveolo-palatal" are apparently different things?), and there are a lot of phonetic issues that even phoneticians don't seem to agree on.

I tend to go with Ladefoged and Maddieson (The Sounds of the World's Languages) here, partly because of their reputation and partly because what they describe tends to fit well with what I feel my own mouth doing.

I can second this recommendation! Their analysis of postalveolars in particular is excellent.

[Travis B.]

I notice that the voice in Google Maps on my phone uses [tɕʰ] in turn...