Are the h and ng sounds allophones?

Discussion of natural languages, or language in general.
Fooge
Sanci
Sanci
Posts: 55
Joined: Sat Jan 21, 2006 8:29 pm

Are the h and ng sounds allophones?

Post by Fooge »

I've heard some say that the sound at the beginning and end of the word "hang" might be allophones. I don't think that's the case. Because people don't really perceive them as being the same sound.

User avatar
KathTheDragon
Smeric
Smeric
Posts: 2139
Joined: Thu Apr 25, 2013 4:48 am
Location: Brittania

Re: Are the h and ng sounds allophones?

Post by KathTheDragon »

The only "evidence" for h and ŋ being the same phoneme is their complementary distribution. But obviously they aren't the same phoneme, so usually people insist on "phonetic similarity". This probably isn't the right approach either, so I agree with your idea, that native speaker intuition is a better judge of which phones belong together.

User avatar
mèþru
Smeric
Smeric
Posts: 1984
Joined: Thu Oct 29, 2015 6:44 am
Location: suburbs of Mrin
Contact:

Re: Are the h and ng sounds allophones?

Post by mèþru »

People usually don't perceive the latter to be a sound at all due to an extremely orthography-centric approach to consonants in English language education. It is usually thought of as n + g. I've actually thought of what you're saying before. I think that it isn't thought of as an allophonic pair by linguists because phones don't have any relationship - [h] can never be modified by its surroundings into becoming [ŋ] or vice versa.
ìtsanso, God In The Mountain, may our names inspire the deepest feelings of fear in urkos and all his ilk, for we have saved another man from his lies! I welcome back to the feast hall kal, who will never gamble again! May the eleven gods bless him!
kårroť

Fooge
Sanci
Sanci
Posts: 55
Joined: Sat Jan 21, 2006 8:29 pm

Re: Are the h and ng sounds allophones?

Post by Fooge »

Yeah. I don't think any linguists seriously consider the sounds as being allophones. They use them as an example of why phonemes can't be defined only by complementary distribution.

User avatar
Hydroeccentricity
Avisaru
Avisaru
Posts: 257
Joined: Mon Aug 05, 2013 10:01 pm

Re: Are the h and ng sounds allophones?

Post by Hydroeccentricity »

The distribution isn't even totally complementary:
"There's a singer ahead."
"I'm sorry, when you have all As in every class in every semester, it's not easy to treat the idea that your views are fundamentally incoherent as a serious proposition."

User avatar
KathTheDragon
Smeric
Smeric
Posts: 2139
Joined: Thu Apr 25, 2013 4:48 am
Location: Brittania

Re: Are the h and ng sounds allophones?

Post by KathTheDragon »

It still is, as the syllabification is different: sɪŋ.ər against ə.hɛd

User avatar
linguoboy
Sanno
Sanno
Posts: 3681
Joined: Tue Sep 17, 2002 9:00 am
Location: Rogers Park/Evanston

Re: Are the h and ng sounds allophones?

Post by linguoboy »

KathTheDragon wrote:It still is, as the syllabification is different: sɪŋ.ər against ə.hɛd
Are there other phonemic contrasts in English where this needs to be taken into account?

User avatar
Salmoneus
Sanno
Sanno
Posts: 3197
Joined: Thu Jan 15, 2004 5:00 pm
Location: One of the dark places of the world

Re: Are the h and ng sounds allophones?

Post by Salmoneus »

linguoboy wrote:
KathTheDragon wrote:It still is, as the syllabification is different: sɪŋ.ər against ə.hɛd
Are there other phonemic contrasts in English where this needs to be taken into account?
Possibly? For instance, I have two different forms of /S/, but these are not different phonemes because their distribution is determined by syllabification: one occurs in /tS/, and one occurs in /t.S/. Similarly, I have different forms of /r/ in /tr/ and /t.r/. So, for example, "rat-rap" and "rat-trap" are audibly different, but we don't have to assume two rhotic phonemes (or, say, a phonemic coda glottal stop or preglottalised stop occuring in the latter, or an unreleased or unaspirated stop in the former) because we instead ascribe this to syllabification differences.
Blog: [url]http://vacuouswastrel.wordpress.com/[/url]

But the river tripped on her by and by, lapping
as though her heart was brook: Why, why, why! Weh, O weh
I'se so silly to be flowing but I no canna stay!

Sumelic
Avisaru
Avisaru
Posts: 385
Joined: Sat Mar 28, 2015 7:05 pm

Re: Are the h and ng sounds allophones?

Post by Sumelic »

mèþru wrote:People usually don't perceive the latter to be a sound at all due to an extremely orthography-centric approach to consonants in English language education. It is usually thought of as n + g.
The common perception of [ŋ] as n + g (or at least, the perception that it is similar to this) may not be just based on the orthography. The orthography is the way it is because of the diachronic development of [ŋ] in English, and synchronically, the distribution of [ŋ] is still similar in some respects to the distribution of consonant clusters of a nasal + a non-coronal consonant like [ŋk] and [mp]: none of them can occur at the start of a syllable, and none of them comes after diphthongs except for in marginal words like "boing" and "oink". There is alternation between [ŋ] and [ŋg] in a few words: strong (stronger strongest), long (longer, longest), young (younger, youngest).
I've actually thought of what you're saying before. I think that it isn't thought of as an allophonic pair by linguists because phones don't have any relationship - [h] can never be modified by its surroundings into becoming [ŋ] or vice versa.
Yes, I think this is an important point. Productive paradigmatic alternations are stronger evidence for allophony than mere complementary distribution. I think this can make it difficult to analyze phonemes in certain languages like Mandarin Chinese where there are few processes like suffixation or ablaut/vowel harmony.
Hydroeccentricity wrote:The distribution isn't even totally complementary:
"There's a singer ahead."
linguoboy wrote:
KathTheDragon wrote:It still is, as the syllabification is different: sɪŋ.ər against ə.hɛd
Are there other phonemic contrasts in English where this needs to be taken into account?
Aside from Salmoneus's examples of the need for syllabification to explain the constrast between different kinds of consonant sequences, there are also ways to explain the distribution of [ŋ] and [h] intervocalically to some extent that don't rely on syllabification. For example, in many accents, /h/ cannot occur after a stressed lax vowel (at least not in ordinary English words), so there could not be any minimal pair like /ˈsɪŋər/ vs. */ˈsɪhər/.

This is complicated a bit by the existence of accents where /ɪ/ is tensed to /i/ before /ŋ/. Even in General American English, /ŋ/ does occur after the tense vowel THOUGHT, so you do have to use syllabification to explain the contrast between "rawhide" and "wrong-eyed". But there are also expected to be (smaller) differences based on syllabification between (imaginary) pairs like "moss-eyed" and "maw-side", or "hawk-eyed" and "haw-kide", and more saliently, between stuff like "thought-eyed" and "thaw-tide".

User avatar
ˈd̪ʲɛ.gɔ kɾuˑl̪
Avisaru
Avisaru
Posts: 255
Joined: Wed May 18, 2016 11:11 pm
Location: Łódź

Re: Are the h and ng sounds allophones?

Post by ˈd̪ʲɛ.gɔ kɾuˑl̪ »

Sumelic wrote:"moss-eyed" and "maw-side"
How would anyone make a difference between these two?

I think the other pairs make sense as the stop consonants have different realisations in the onset and the coda, but this example and thinking singer and ahead differ in the place of the syllable boundary for me have no basis in the phonetic details. I personally don't like phonologists insisting that the stressed lax vowels appear only in closed syllables and thus playing with the placement of the dot and making other conclusions about English phonology out of that statement.
In Budapest:
- Hey mate, are you hung-a-ry?

User avatar
KathTheDragon
Smeric
Smeric
Posts: 2139
Joined: Thu Apr 25, 2013 4:48 am
Location: Brittania

Re: Are the h and ng sounds allophones?

Post by KathTheDragon »

ˈd̪ʲɛ.gɔ kɾuˑl̪ wrote:
Sumelic wrote:"moss-eyed" and "maw-side"
How would anyone make a difference between these two?
I distinguish them by vowel quality, /ɒ/ against /ɔː/, though I doubt that's what you were asking.
but this example and thinking singer and ahead differ in the place of the syllable boundary for me have no basis in the phonetic details.
How else do you propose the (tmk) universal understanding of what syllables are and where boundaries are located, especially in languages with frequent complex clusters?
I personally don't like phonologists insisting that the stressed lax vowels appear only in closed syllables and thus playing with the placement of the dot and making other conclusions about English phonology out of that statement.
Fun fact, I have bitten, bittern /ˈbɪ.tn̩/ with an open first syllable.

User avatar
Salmoneus
Sanno
Sanno
Posts: 3197
Joined: Thu Jan 15, 2004 5:00 pm
Location: One of the dark places of the world

Re: Are the h and ng sounds allophones?

Post by Salmoneus »

ˈd̪ʲɛ.gɔ kɾuˑl̪ wrote:
Sumelic wrote:"moss-eyed" and "maw-side"
How would anyone make a difference between these two?
Well, most people of course have totally different vowels in these words.

But between, say, "peace-eyed" and "pea-side"? Well, there's something going on with the /s/ - maybe it's just lengthening, maybe it's glottalisation or something? I think the closed vowel is also shorter. And there may be something with tongue-backing that's not actually glottalisation... it's marginal with /s/, but when I consider "tine-eat" vs "tie-neat", there's a definate "darkening" of the nasal and the preceding vowel, though what that constitutes exactly I'm not sure.
I think the other pairs make sense as the stop consonants have different realisations in the onset and the coda, but this example and thinking singer and ahead differ in the place of the syllable boundary for me have no basis in the phonetic details.
Are you a native speaker?
Blog: [url]http://vacuouswastrel.wordpress.com/[/url]

But the river tripped on her by and by, lapping
as though her heart was brook: Why, why, why! Weh, O weh
I'se so silly to be flowing but I no canna stay!

User avatar
ˈd̪ʲɛ.gɔ kɾuˑl̪
Avisaru
Avisaru
Posts: 255
Joined: Wed May 18, 2016 11:11 pm
Location: Łódź

Re: Are the h and ng sounds allophones?

Post by ˈd̪ʲɛ.gɔ kɾuˑl̪ »

Salmoneus wrote:
ˈd̪ʲɛ.gɔ kɾuˑl̪ wrote: I think the other pairs make sense as the stop consonants have different realisations in the onset and the coda, but this example and thinking singer and ahead differ in the place of the syllable boundary for me have no basis in the phonetic details.
Are you a native speaker?
I admit I'm not, but I don't think it has anything to do with my aural ability to hear that English speaker don't really pronounce words like singer with an audible gap between the velar nasal and the schwa.
KathTheDragon wrote:
but this example and thinking singer and ahead differ in the place of the syllable boundary for me have no basis in the phonetic details.
How else do you propose the (tmk) universal understanding of what syllables are and where boundaries are located, especially in languages with frequent complex clusters?
Well, as I'm a native speaker of such a language, I think there may be multiple ways to divide a word into syllables (e.g. I can think of three different divisions of the Polish word for "die", kostka (in a simple phonemic transcription): /'ko.stka/, /'kos.tka/, /'kost.ka/), but I don't link the pronunciation to the morphological structure of the words, except when the phonemes have different allophones based on their position in the syllable, which doesn't happen in the case of the velar nasal.
In Budapest:
- Hey mate, are you hung-a-ry?

Fooge
Sanci
Sanci
Posts: 55
Joined: Sat Jan 21, 2006 8:29 pm

Re: Are the h and ng sounds allophones?

Post by Fooge »

There's also the fact that the sounds [h] and [N] are represented differently orthographically. Allophones of a single phoneme tend to be represented the same way. The unaspirated p in "spin" and the aspirated p in "pin" are both represented with "p".

User avatar
Salmoneus
Sanno
Sanno
Posts: 3197
Joined: Thu Jan 15, 2004 5:00 pm
Location: One of the dark places of the world

Re: Are the h and ng sounds allophones?

Post by Salmoneus »

ˈd̪ʲɛ.gɔ kɾuˑl̪ wrote:
Salmoneus wrote:
ˈd̪ʲɛ.gɔ kɾuˑl̪ wrote: I think the other pairs make sense as the stop consonants have different realisations in the onset and the coda, but this example and thinking singer and ahead differ in the place of the syllable boundary for me have no basis in the phonetic details.
Are you a native speaker?
I admit I'm not, but I don't think it has anything to do with my aural ability to hear that English speaker don't really pronounce words like singer with an audible gap between the velar nasal and the schwa.
Errr... no. But... "followed by an audible gap" isn't the definition of a syllable. That's just... not what people mean.

It's absolutely audibly clear that the /N/ in 'singer' is in the first syllable, and that the /h/ in 'ahead' is in the second syllable. I'm sorry if you can't hear that, but it is.
Blog: [url]http://vacuouswastrel.wordpress.com/[/url]

But the river tripped on her by and by, lapping
as though her heart was brook: Why, why, why! Weh, O weh
I'se so silly to be flowing but I no canna stay!

Fooge
Sanci
Sanci
Posts: 55
Joined: Sat Jan 21, 2006 8:29 pm

Re: Are the h and ng sounds allophones?

Post by Fooge »

Salmoneus wrote:
linguoboy wrote:
KathTheDragon wrote:It still is, as the syllabification is different: sɪŋ.ər against ə.hɛd
Are there other phonemic contrasts in English where this needs to be taken into account?
Possibly? For instance, I have two different forms of /S/, but these are not different phonemes because their distribution is determined by syllabification: one occurs in /tS/, and one occurs in /t.S/. Similarly, I have different forms of /r/ in /tr/ and /t.r/. So, for example, "rat-rap" and "rat-trap" are audibly different, but we don't have to assume two rhotic phonemes (or, say, a phonemic coda glottal stop or preglottalised stop occuring in the latter, or an unreleased or unaspirated stop in the former) because we instead ascribe this to syllabification differences.
An example in my speech is "holy" vs. "slowly". In my speech those don't rhyme. "Holy" has a dark l and "slowly" has a light l. However it doesn't mean I have a phonemic light l dark l distinction. The difference is a syllable difference.

User avatar
Salmoneus
Sanno
Sanno
Posts: 3197
Joined: Thu Jan 15, 2004 5:00 pm
Location: One of the dark places of the world

Re: Are the h and ng sounds allophones?

Post by Salmoneus »

Fooge wrote:
Salmoneus wrote:
linguoboy wrote:
KathTheDragon wrote:It still is, as the syllabification is different: sɪŋ.ər against ə.hɛd
Are there other phonemic contrasts in English where this needs to be taken into account?
Possibly? For instance, I have two different forms of /S/, but these are not different phonemes because their distribution is determined by syllabification: one occurs in /tS/, and one occurs in /t.S/. Similarly, I have different forms of /r/ in /tr/ and /t.r/. So, for example, "rat-rap" and "rat-trap" are audibly different, but we don't have to assume two rhotic phonemes (or, say, a phonemic coda glottal stop or preglottalised stop occuring in the latter, or an unreleased or unaspirated stop in the former) because we instead ascribe this to syllabification differences.
An example in my speech is "holy" vs. "slowly". In my speech those don't rhyme. "Holy" has a dark l and "slowly" has a light l. However it doesn't mean I have a phonemic light l dark l distinction. The difference is a syllable difference.
That's... weird? Normally the distinction is the other way: "wholly" with dark l, and "holy" with light l. But yes, syllabification.
Blog: [url]http://vacuouswastrel.wordpress.com/[/url]

But the river tripped on her by and by, lapping
as though her heart was brook: Why, why, why! Weh, O weh
I'se so silly to be flowing but I no canna stay!

User avatar
Ryusenshi
Lebom
Lebom
Posts: 189
Joined: Sun Mar 12, 2017 3:31 am
Location: Montrouge, France

Re: Are the h and ng sounds allophones?

Post by Ryusenshi »

The goat/goal split may also play a part: some people have the goal allophone before intervocalic /l/, as in holy, but not in slowly because of a morpheme boundary.

I don't think anyone ever suggested seriously that [h] and [ŋ] are allophones; it seems to me a reductio ad absurdum to show that complementary distribution is a necessary but not sufficient condition.

Sumelic
Avisaru
Avisaru
Posts: 385
Joined: Sat Mar 28, 2015 7:05 pm

Re: Are the h and ng sounds allophones?

Post by Sumelic »

KathTheDragon wrote:
ˈd̪ʲɛ.gɔ kɾuˑl̪ wrote:
Sumelic wrote:"moss-eyed" and "maw-side"
How would anyone make a difference between these two?
I distinguish them by vowel quality, /ɒ/ against /ɔː/, though I doubt that's what you were asking.
Salmoneus wrote: Well, most people of course have totally different vowels in these words.
Earlier in the paragraph, I wrote "in General American English", and the first example I gave was "rawhide" vs. "wrong-eyed". I thought it was sufficiently obvious that I was talking about an accent where CLOTH = THOUGHT; obviously words that are mininal pairs in one accent might not be in another accent that hasn't undergone the same vowel changes. Replace it with Salmoneus's example of '"peace-eyed" vs. "pea-side"' if that makes it easier to follow the argument.
ˈd̪ʲɛ.gɔ kɾuˑl̪ wrote:
Sumelic wrote:"moss-eyed" and "maw-side"
How would anyone make a difference between these two?
Salmoneus wrote: But between, say, "peace-eyed" and "pea-side"? Well, there's something going on with the /s/ - maybe it's just lengthening, maybe it's glottalisation or something? I think the closed vowel is also shorter. And there may be something with tongue-backing that's not actually glottalisation... it's marginal with /s/, but when I consider "tine-eat" vs "tie-neat", there's a definate "darkening" of the nasal and the preceding vowel, though what that constitutes exactly I'm not sure.
I think the originally word-final /s/ would tend to be preceded by a shorter vowel than the originally word-initial /s/, but there definitely isn't a huge difference, and it's hard to know whether it would be at all easy to detect when listening to naturally produced speech. I would certainly say that the examples I gave would sound similar enough to be confused (and the history of English shows that rebracketing of word-initial or word-final consonants has occurred: e.g. a nadder > an adder), and in contrast "rawhide" and "wrong-eyed" are not as easy to confuse.
ˈd̪ʲɛ.gɔ kɾuˑl̪ wrote: I think the other pairs make sense as the stop consonants have different realisations in the onset and the coda, but this example and thinking singer and ahead differ in the place of the syllable boundary for me have no basis in the phonetic details. I personally don't like phonologists insisting that the stressed lax vowels appear only in closed syllables and thus playing with the placement of the dot and making other conclusions about English phonology out of that statement.
Well, we've had discussions on this site about syllabification before (e.g. the thread I made "English syllabification discussion") and I don't think it's necessary to reproduce all of the arguments here. But I'll just say that I don't think it's really possible to syllabify words without using some criteria based on phonotactics and phonology in addition to phonetics.
KathTheDragon wrote: Fun fact, I have bitten, bittern /ˈbɪ.tn̩/ with an open first syllable.
Salmoneus wrote: Errr... no. But... "followed by an audible gap" isn't the definition of a syllable. That's just... not what people mean.

It's absolutely audibly clear that the /N/ in 'singer' is in the first syllable, and that the /h/ in 'ahead' is in the second syllable. I'm sorry if you can't hear that, but it is.
I don't agree with the idea that it's intuitively obvious to native English speakers which syllable an intervocalic consonant falls in. Or rather, I think it does seem obvious to some people, but they don't all reach the same conclusions, which is a problem for anyone trying to base a theory of syllabification just on personal intuitions abut syllabification (unless you're willing to bite the bullet and argue that different English speakers actually have significantly different underlying systems of syllabification, whatever that means). Some people feel it's obvious that the first syllable of a word like "bitten" is /bɪ/, some people feel it's obvious that it's /bɪt/, and some people feel it's obvious that the middle consonant is "ambisyllabic" and shared between both syllables. Bruce Hayes presents an ambisyllabicity analysis in Chapter 13A of "Introductory Phonology" that says that /ŋ/ is ambisyllabic in "Singapore" and "gingham"; this roughly corresponds with my own intuition that if the /t/ in the middle of a word like "knitter" counts in some way as the onset of the second syllable (my personal intuitions don't give me a clear answer about whether this is true), the /ŋ/ in the middle of a word like "singer" seems like it could count in some way as the onset of the second syllable.

User avatar
mèþru
Smeric
Smeric
Posts: 1984
Joined: Thu Oct 29, 2015 6:44 am
Location: suburbs of Mrin
Contact:

Re: Are the h and ng sounds allophones?

Post by mèþru »

Sumelic wrote:unless you're willing to bite the bullet and argue that different English speakers actually have significantly different underlying systems of syllabification, whatever that means
I'm not saying the answer is yes, but why should we immediately discount that line of thinking?
ìtsanso, God In The Mountain, may our names inspire the deepest feelings of fear in urkos and all his ilk, for we have saved another man from his lies! I welcome back to the feast hall kal, who will never gamble again! May the eleven gods bless him!
kårroť

Richard W
Avisaru
Avisaru
Posts: 363
Joined: Sat Oct 16, 2010 8:28 pm

Re: Are the h and ng sounds allophones?

Post by Richard W »

mèþru wrote:It is usually thought of as n + g. I've actually thought of what you're saying before. I think that it isn't thought of as an allophonic pair by linguists because phones don't have any relationship - [h] can never be modified by its surroundings into becoming [ŋ] or vice versa.
h > ŋ is a possible outcome of rhinoglottophilia, and can be phonetically conditioned. Syllable-initial Proto-Tai *ŋ > h in a Thai dialect spoken near Songkhla contrasts with the retention of final /ŋ/. Proto-Tai *hŋ shows a fair amount of seemingly inconsistent development to modern /h/ or /ŋ/, though it may be phonetically conditioned.

User avatar
mèþru
Smeric
Smeric
Posts: 1984
Joined: Thu Oct 29, 2015 6:44 am
Location: suburbs of Mrin
Contact:

Re: Are the h and ng sounds allophones?

Post by mèþru »

I have heard of rhinoglottophilia before, but not in English.
ìtsanso, God In The Mountain, may our names inspire the deepest feelings of fear in urkos and all his ilk, for we have saved another man from his lies! I welcome back to the feast hall kal, who will never gamble again! May the eleven gods bless him!
kårroť

Richard W
Avisaru
Avisaru
Posts: 363
Joined: Sat Oct 16, 2010 8:28 pm

Re: Are the h and ng sounds allophones?

Post by Richard W »

mèþru wrote:
Sumelic wrote:unless you're willing to bite the bullet and argue that different English speakers actually have significantly different underlying systems of syllabification, whatever that means
I'm not saying the answer is yes, but why should we immediately discount that line of thinking?
Because it's an inconvenient truth? It runs counter to the lie that native speakers do not have any problems in acquiring their native tongue.

I've seen personal native speaker testimony to syllabifying <singing> as si-nging, corroborated by the fact that the person concerned had no difficulty with syllable-initial /ŋ/ in foreign languages.

Travis B.
Sumerul
Sumerul
Posts: 3570
Joined: Mon Jun 20, 2005 12:47 pm
Location: Milwaukee, US

Re: Are the h and ng sounds allophones?

Post by Travis B. »

One thing to consider is that in almost all these cases of supposed phonemic syllabification (e.g. singing versus ahead, rawhide versus wrong-eyed, peace-eyed versus pea-side, tine-eat versus tie-neat, wholly versus holy) what we really have are phonemic morpheme boundaries or phonemic stress (e.g. Wis-consin versus Wi-sconsin). Therefore I am highly skeptical of the existence of phonemic syllabification, because practically always when it is proposed there is a better explanation.
Dibotahamdn duthma jallni agaynni ra hgitn lakrhmi.
Amuhawr jalla vowa vta hlakrhi hdm duthmi xaja.
Irdro. Irdro. Irdro. Irdro. Irdro. Irdro. Irdro.

User avatar
mèþru
Smeric
Smeric
Posts: 1984
Joined: Thu Oct 29, 2015 6:44 am
Location: suburbs of Mrin
Contact:

Re: Are the h and ng sounds allophones?

Post by mèþru »

I agree that phonemic syllabification does not exist in most English varieties. That does not mean it does not exist in other languages.
ìtsanso, God In The Mountain, may our names inspire the deepest feelings of fear in urkos and all his ilk, for we have saved another man from his lies! I welcome back to the feast hall kal, who will never gamble again! May the eleven gods bless him!
kårroť

Post Reply