S might be a vowel

[Daniel]

I just came across this paper and found it to be very interesting:
http://journals.linguisticsociety.org/proceedings/index.php/amphonology/article/view/50

[Jeff L.]

It's interesting stuff! I was going to send it to my prof because she has a lot of related work...then saw the first author IS my prof! She's also got some work on acquisition and some postulations about acoustic factors that are quite interesting as well.

I find it's also a cool was to describe the surface forms in Laurentian French where you have vowel deletion or "devoicing", which, when it happens between [s ]/[z]s (common given that there's affrication of /t/, /d/ before high front vowels/glides), gives the impression that the sibilants become nuclei...and acoustically, there's often no break in frication and not always formants accompanying that frication! If you ask a speaker, they'll often perceive the syllable as being present where [s ] or [z] is present (for a long string of deleted/devoiced vowels, often perceive one syllable fewer), though that could simply be reconstruction knowing the forms or underlying phonotactic constraints.

Edit: ...Forgot I'd end up with a strikeout for having [s ] without a space!

[freknu]

Oh? A voiceless sibilant is suddenly a vowel? /rolls eyes

Or do they simply mean it's syllabic?

[Daniel]

What's a vowel?

[mallu]

Is any sound that is syllabic be called a vowel?

[Daniel]

If there's no differentiation in the phonological distribution of that sound and "other vowels" then why/how could we separate them?

[freknu]

What's a vowel?

A sonorant open sound. Are sibilants sonorants?

Or are you confusing vowel with syllabic nucleus?

[Daniel]

What's a sonorant? Terminology is either circular going back to phonological features which here include S or just ostensively pointing to "English vowels".

(Of course you could argue the author is simply wrong, but if not...)

[freknu]

What's a sonorant? Terminology is either circular going back to phonological features which here include S or just ostensively pointing to "English vowels".

Have it your way. I'll just refer to "consonants" and "vowels" as poop and piss sounds, since the curret terminology is too whimsical and insubstantial, belonging on the same heap of trash as "god".

[Daniel]

This is a serious issue, though. We use these terms as if they are primitives, and it's not clear that they are. Their essence ends up either circular or going back to some arbitrary description from a hundred years ago or so.
We are mostly likely "on to" something, with a real phenomenon lurking under the inadequate terminology (or inadequately defined terminology).

Of course this is a particularly extreme example (and possibly only applicable in one language), but it happens all the time. The terms we rely on fall apart once we interrogate them.

In my case, I'm currently going through about 300 languages to attempt to identify "serial verb constructions" and I'm having a really difficult time because the definition (which I have extensively researched and polished from a number of sources) seems completely arbitrary. In fact, my point in the end will be to argue against any deeper meaning of "serial verb construction" beyond just what certain languages seem to have on the surface, but to get there I'm trying to work within the definition and it's an interesting challenge.


Anyway, as for S as a vowel, why not?

If speakers of the (so-called) Khoisan languages can produce clicks reliably and contrastively at a volume effective for speech, I can certainly imagine someone producing /s/ as a nucleus, which is what vowels are. Vowels are the sounds that don't need other sounds (vowel) to support them in a syllable. In fact, we have a sort-of-word in English that has no vowels and just S as a nucleus: psst (as in "hey, psssst, look at me!"). And it would helpfully explain why S can anomalously go before stops. Of course this would contrast with a language like Spanish where s- must be preceded by a vowel for support, as in estudiar 'study', especial 'special', escuela 'school'.

I'm intrigued by this argument, but (not the fault of the argument itself) a little dizzy from all of the circles the terminology seems to drag us in.

[MalFet]

Oh? A voiceless sibilant is suddenly a vowel? /rolls eyes

Or do they simply mean it's syllabic?

What they authors are saying is actually a bit more interesting than that. In fact, they're not even trying to suggest that the terminology that exists is circular or insufficiently defined. Instead, they're claiming that what is salient and consequential about [s ] within the phonotactics of Blackfoot are the things that are salient and consequential about vowels (in contrast to what is salient and consequential about consonants). It's a cool idea. I'm skeptical on a few points, but they're operating within totally conventional definitions.

[Daniel]

Thanks for phrasing it that way. You're right

[I didn't mean that these authors are questioning it. I mean that our expectations of what a definition "should" be are problematic-- an S isn't a vowel... why? Because it isn't. Unless we look at it this way, which, assuming that vowel is defined coherently in the first place, is a natural progression (again, assuming the analysis). The result, though, is that "vowel" means something quite different from the traditional sense of it that we have of "A, E, I, O, U". In some of the other cases, I don't know how much of a real phenomenon can actually be discovered, because often the definition is ostensibly arbitrary-- these are vowels, and this isn't, so now what? The authors are doing the right thing, assuming the argument and data hold up.]