Re: Here we go loup-garou

Jeff Rollin wrote:
> In the last episode, (On Tuesday 10 July 2007 15:05:54), T. A. McLeay wrote:
>> Jeff Rollin wrote:
>>> I know there are languages which don't have voiced consonants, like
>>> Finnish (except approximants, which appear to be always voiced; from
>>> loans; and /d/, which is regarded as somewhat artificial and replaced by
>>> various phonemes in non-standard dialects), but are there any languages
>>> which have voiced consonants but no voiced ones at all?
>> I assume you mean "but are there any languages which have voiced
>> consonants but no voiceLESS ones at all?".
>>
>> No. If a language has no voicing distinction, then you will almost
>> always find a situation in which initial and final obstruents are
>> voiceless, and intervocallic ones may be allophonically voiced. This is
>> because (a) the pressure in your lungs needs to be a certain amount
>> greater than that of your oral cavity for voicing to occur, but blocking
>> the air from escaping while forcing air from your lungs into your mouth
>> causes the pressure to equalise --- this makes it hard to voice stops
>> and (b) because to pronounce a fricative you need to force a lot of air
>> through a small space (otherwise you either get no air coming through
>> and a stop, or you get air passing through cleanly and you get an
>> approximant), but vibrating your vocal chords makes it harder to obtain
>> this critical level --- this makes it hard to fricate while voicing.
>>
>> Note that these two process work in opposite ways: It is hard to vibrate
>> your vocal chords while making a stop; but it is hard to make a
>> fricative while vibrating your vocal chords. This explains why in many
>> languages, voiced fricatives are often pronounced as approximants.
>>
>> So in order to have a language in which only voiced consonants occurred,
>> either the speakers would need to be doing something harder than to
>> include voiceless consonants (and would therefore quickly include
>> voiceless consonants into their repertoire either allophonically or
>> phonemically); or else the language will lack obstruents entirely and
>> have only vowels and sonorants.
>>
>> A language with no obstruents is exceedingly unlikely because it makes
>> the hearer's job a lot harder. Nasals and laterals at different points
>> of articulation are distinguished by what frequencies are *missing*
>> rather than which ones are *present*, and so they sound a lot more
>> similar (notice how hard it is to clearly say "*em*, not *en*"). Central
>> approximants are much harder to keep apart from vowels, being
>> essentially vowels in a part of a syllable designated for consonants.
>> I'm not sure what, if anything, is wrong with taps, as they are
>> essentially voiced stops pronounced so quickly that the difficulty of
>> maintaining voicing doesn't come up, but precisely because of this I
>> suppose you'd be better able to hear it if you took longer to say it and
>> turned it into a proper voiced stop.
>>
>> (Australian Aboriginal languages often use d, rd, dj, g for /t t` c k/;
>> this is as much due to the (Australian) English pronunciation of
>> /t _ tS k/ (which are aspirated) and /d _ dZ g/ (which are unvoiced or
>> voiced only lightly before stressed vowels) as the Aboriginal
>> pronunciations.)
>>
>> If you meant "but are there any language which have voiceLESS consonants
>> but no voiced ones at all", this would entail the absence of sonorant
>> consonants and allophonic intervocalic vocing. I'm not sure whether or
>> not any language does the former, and in the absence of a voicing
>> distinction I'd be surprised if the latter occurred. I wouldn't rule it
>> out (like I would the other way), but I don't know of any.
>>
>> HTH,
>
> Thanks, that was a really clear explanation. And yes I did mean the first of
> the two possibilities you posited.
>
> FWIW I think there are Aust Abor languages that lack /s/, but I don't know if
> any lack all sonorants - if /s/ is an example of what you mean by sonorant.
Oh no --- Perhaps I should have defined my terms.

Obstruents are consonants which have obstruct the air flow creating
either a blockage (as with oral stops) or heavy turbulance (as with
fricatives). They generally will not be syllable nuclei altho I gather
there are languages in which phonemically just about anything can be...

Sonorants are the rest, i.e. those in which there is a non-turbulant
escape of air: nasals, (lateral and central) approximants,
trills/taps/flaps. I was obviouly using this term differently from how
Mark defines it (and I must say I've never heard a definition of
sonorant than includes fricatives, nor one that excludes approximants).
(Nasals, also called nasal stops, have a blockage in the oral cavity,
but air escapes quite happily through the nose. Nasal fricatives are
impossible; you can't get the build-up of pressure necessary to have
frication/turbulance if you're letting air escape through the nose.)

[s] is a sibilant fricative --- and most definitely an obstruent. Most
Australian Aboriginal languages lack fricatives entirely, but they do
include sonorants: lateral approximants at most of dental, alveolar,
post-alveolar and palatal, and nasals at most of the above as well as
bilabial and velar, and central approximants at most of the lateral POAs
as well as labiovelar. (Personally, I'm glad I was brought up speaking
English --- learning to make all of those distinctions is a hard job! I
gather many languages prestop their nasals to help ease the distinction,
but some also have prenasalised stops! It sounds like an evil conlanger
had his hand on the Australian continent...)

--
Tristan.