zompist bboard

What is the difference between [s\] and [S]? The descriptions on Wikipedia don't really help, and I can't really hear a difference.

Also, are there any languages which contrast the two?

What is the difference between [s\] and [S]? The descriptions on Wikipedia don't really help, and I can't really hear a difference.

[S] has the tongue just behind the alveolar ridge while [s\] lifts the body of the tongue toward the palate at the same time.

Also, are there any languages which contrast the two?

Possibly some Slavic languages. I thought Polish did, but apparently it contrasts alveolar / alveolo-palatal / retroflex instead.

Possibly some varieties of Swedish and Norwegian, too.

Well, presumably you've got no problem with [S], so you just need to get a handle on [s\]. The quick and dirty method that will get you most of the way there is to just pretend it's a "palatized" [S]. Say [S_j] or [s_j] many times really fast and you'll get pretty close. The best way, though, is to listen to sound samples of languages with that sound. Youtube videos in Japanese, Mandarin, Korean, or Vietnamese will put you up to your eyeballs in [s\].

Montenegrin apparently has [s\ z\] as possible realisations of /sj zj/ which would therefore contrast with [S Z].

brandrinn wrote:The quick and dirty method that will get you most of the way there is to just pretend it's a "palatized" [S].

Isn't it exactly the same thing (by definition) as [S_j]? What would the difference be?

Some forms of Scottish Gaelic have [s`] for /rs/ and, as usual, [S] for slender /s/.

Thanks for the replies, I think I'm getting there.
Are the diagrams at the bottom of this article accurate? The one for [S] seems more 'palatalised' than the one for [s\].

Aid'os will probably correct me on this, but I think this is the right canIPA:


For some reason, the IPA characters for the alveolopalatal fricatives are used in canIPA for the bilabialized prepalatals. Weird.

jwt wrote:Thanks for the replies, I think I'm getting there.
Are the diagrams at the bottom of this article accurate? The one for [S] seems more 'palatalised' than the one for [s\].

It looks like the back of the tongue is higher in [ʃ] but the site of greatest constriction in farther forward than [ɕ].

I confess: I have no idea about the S-family consonants.

For instance, the 'palatalised S' description and the above picture seem very contradictory. Well, no, to start with, the idea of palatalising S is weird, since, for me, /S/ is basically laminal-palatal to start with. So I assume you're talking about an apical consonant - but that picture is clearly laminal.

For me, I can make: a 'back S' (probably palatal), a 'front S' (closer to the ridge), a 'very front S' (like that picture, I think - tongue behind the lower teeth, blade pressed all the way forward), a 'double S' (like that, but the tip moved up behind the front teeth to add an extra level of frication), a 'front apical S' (like the double S but the blade lowered so that it's just the tip making the sound), a 'medium apical S' (tongue moved back halfway to the ridge), and a 'back apical S' (tip on the cusp, maybe a bit behind it). And, unreliably, a 'moving r-S' (apical S, tip flicks past the cusp, sometimes causes tongue to roll a bit).

But I don't know what any of these are, and readin descriptions of the 'real' sounds, I don't know how to make any of them.

Adding to the complexity: I'm aware most people manage to fit an /s/ in there as well, but I don't know where or how (my /s/ is apical-interdental-bilabial).

Salmoneus wrote:For instance, the 'palatalised S' description and the above picture seem very contradictory. Well, no, to start with, the idea of palatalising S is weird, since, for me, /S/ is basically laminal-palatal to start with. So I assume you're talking about an apical consonant - but that picture is clearly laminal.

[ɕ] and [ʑ] are necessarily laminal, IIRC, just as [ʂ] and [ʐ] are necessarily apical. That is also a major point of contrast between the two of them in a lot of language.

I recall reading that in a large majority of languages with two postalveolar consonants, one will be apical and the other will be laminal.

ɕ sounds like a cross between ʃ and ç. or it does to me. that doesn't necessarily help in making it though. but kinda halfway between, sorta

finlay wrote:ɕ sounds like a cross between ʃ and ç. or it does to me. that doesn't necessarily help in making it though. but kinda halfway between, sorta

Yeah, that's about how I consider it. The difference between /ɕ/ and /ç/ is not completely clear to me, and UPSID has a different way to divide those sounds than IPA, or can vary from language to language. Personally I consider /ɕ/ as a /ʃ/ but with a "lighter sound" (the frequency of the noise you hear is higher), where /ʂ/ has an even darker sound.

I consider both /ɕ/ and /ç/ to be pretty much devoiced /j/, but /ɕ/ is more like /ʃ/ (and tends to pattern with sibilants, see Mandarin or Ubykh), and /ç/ is more like /x/ (and tends to pattern with dorsals, see Greek). But they're both very similar sounds.

In practice the /ɕ/ symbol is usually used in languages that distinguish a darker post-alveolar sibilant with a lighter one, and the details of pronunciation vary from language to language.

In terms of articulation, /s,ʃ,ɕ,ʂ/ are all considered coronal (same place of articulation as /t,d,n,l/...), but they can be articulated by different parts of the tongue: the tip (apical), or a small hump right after the tip (laminal - ie using the blade of the tongue), however:
* /ɕ/ has to be laminal (you can't make an apical /ɕ/ afaik).
* /ʂ/ has to be apical (retroreflexes have to be apical).
* /s/ and /ʃ/ can be either apical (see English) or laminal (see French).

MadBrain wrote:Personally I consider /ɕ/ as a /ʃ/ but with a "lighter sound" (the frequency of the noise you hear is higher), where /ʂ/ has an even darker sound.

I think this sums it up quite well.

About [ʂ], isn't Mandarin [ʂ] laminal? Or is it not actually [ʂ]?

jwt wrote:

MadBrain wrote:Personally I consider /ɕ/ as a /ʃ/ but with a "lighter sound" (the frequency of the noise you hear is higher), where /ʂ/ has an even darker sound.

I think this sums it up quite well.

About [ʂ], isn't Mandarin [ʂ] laminal? Or is it not actually [ʂ]?

According to Wikipedia it's laminal rather than apical. It also states though that /ʂ/ is inherently vague, representing a post-alveolar/retroflex sibilant but underspecified for laminal or apical tongue shapes. It goes on to state that the difference between laminal /ʂ/, /ʃ/ (which is also, in general, underspecified for laminal/apical distinction in IPA and can be used for both) and /ɕ/ (always laminal) involves degree of palatalisation, AFAICT. That is laminal /ʂ/ is unpalatalised, /ʃ/ is partially palatalised and /ɕ/ is highly palatalised.

bricka wrote:Possibly some varieties of Swedish and Norwegian, too.

I think almost all Norwegians merge /S/ and /s`/ now. Even if there are some who maintain the distinction as [S~s\] vs. [s`], the phones are practically in complementary distribution anyway. But many Norwegians contrast their /S~s`/ (probably underspecified for laminal/apical) with their /C~s\/ (underspecified for... something else).

jwt wrote:

MadBrain wrote:Personally I consider /ɕ/ as a /ʃ/ but with a "lighter sound" (the frequency of the noise you hear is higher), where /ʂ/ has an even darker sound.

I think this sums it up quite well.

About [ʂ], isn't Mandarin [ʂ] laminal? Or is it not actually [ʂ]?

Hmm, I have no idea actually... I use an apical articulation for that Mandarin sound but I'm not very good at Mandarin, so this would need more info to confirm.

Nortaneous wrote:Aid'os will probably correct me on this, but I think this is the right canIPA:


For some reason, the IPA characters for the alveolopalatal fricatives are used in canIPA for the bilabialized prepalatals. Weird.

How did you extract the image?

Anyway, that is the correct canIPA. Simple old [s' z'].

Aid'os wrote:

Nortaneous wrote:Aid'os will probably correct me on this, but I think this is the right canIPA:


For some reason, the IPA characters for the alveolopalatal fricatives are used in canIPA for the bilabialized prepalatals. Weird.

How did you extract the image?

I didn't. I took a screenshot and cropped it.

Ah, I have remembered what the distinction is – the IPA doesn't have the best way of denoting it, but the difference between [θ] and [s] is mainly one of tongue shape – the [s] is more bunched up in the middle to make a "groove", which funnels air towards the back of the teeth, making what we might call sibilance. The [θ] can be made at the same POA, say alveolar instead of dental, by making a "slit", ie a flatter tongue shape with a horizontal slit between the tongue and the roof of the mouth – this doesn't funnel air towards the back of the teeth and we don't get sibilance. The difference between [ç] and [ɕ] is (or can be) exactly the same: [ç] is a slit fricative and [ɕ] is a grooved fricative.

bricka wrote:Some forms of Scottish Gaelic have [s`] for /rs/ and, as usual, [S] for slender /s/.

Kind of similar to Central and maybe half of Southern Swedish and I guess most of Norwegian too. Northern Swedish usually has [s`] for both.
I think they're more or less allophones for me (/rs rS/ are always [s`], but /S/ is some times [s`] and some times [S], but I guess this is just sloppy speech, as I can have [z`] and [Z] some times too).