zompist bboard

How does it develop?

Primarily, vowel length is seen in Ancient Greek, Latin and Sanskrit. As should be clear, all three are used as liturgical languages. It is clear that vowel length in these languages is a result of the sonorous chanting of clergy and religious professionals, who need to draw out the vowel sounds in order to make them fit the cadence of the chant.

*"Compensatory lengthening"--that is, a consonant is lost, and the preceding (or, occasionally, the following) vowel lengthens to "make up" for the lost segment. For example, in the history of Cayuga (Iroquoian): *íhraks --> iha:s "he eats". Also seen in non-rhotic English dialects (those that drop coda [syllable-final] "r"), e.g. cart = [kʰɑ:t] (I think that's the vowel?)

*Coalescence of multiple vowels (or vowels separated by glides or [h] or something similar) into a single long vowel. For example, in Lakhota Sioux, sequences of vowel+(w,y,h)+vowel often coalesce into a single long vowel whose form is dependent on which the two vowels in the original sequence were; an example is iyáye "he left for there" --> [iyæ:].

*Influence from neighboring consonants; English exhibits this. Vowels before voiced consonants are allophonically lengthened, so bad = [bæ:d] while bat = [bæt]. A loss of the conditioning environment (such as, say, the devoicing of all word-final stops) could lead to the distinction becoming phonemic. For instance, you'd have /bæt/ = "bat" and /bæ:t/ = "bad".

Suprasegmentals can give rise to vowel length as well. The role of stress should be readily apparent to an English speaker. In modern Korean, the loss of pitch-accent results in distinctive vowel length in some dialects, e.g. MK /mal_H/ "measure", /mal_R/ "speech" > ModK. /mal/, /ma:l/.

Whimemsz wrote:*"Compensatory lengthening"--that is, a consonant is lost, and the preceding (or, occasionally, the following) vowel lengthens to "make up" for the lost segment. For example, in the history of Cayuga (Iroquoian): *íhraks --> iha:s "he eats". Also seen in non-rhotic English dialects (those that drop coda [syllable-final] "r"), e.g. cart = [kʰɑ:t] (I think that's the vowel?)

*Coalescence of multiple vowels (or vowels separated by glides or [h] or something similar) into a single long vowel. For example, in Lakhota Sioux, sequences of vowel+(w,y,h)+vowel often coalesce into a single long vowel whose form is dependent on which the two vowels in the original sequence were; an example is iyáye "he left for there" --> [iyæ:].

*Influence from neighboring consonants; English exhibits this. Vowels before voiced consonants are allophonically lengthened, so bad = [bæ:d] while bat = [bæt]. A loss of the conditioning environment (such as, say, the devoicing of all word-final stops) could lead to the distinction becoming phonemic. For instance, you'd have /bæt/ = "bat" and /bæ:t/ = "bad".

AFAIK, influence from neighbouring consonants is a major cause of "compensatory" lengthening which develops in a similar manner to phonemic tone. That is to say that the vowel is already phonemically lengthened before certain consonants with that consonant merging with another or being lost all together rather than the vowel becoming long as a result of this loss of the following consonant.

As linguoboy hinted at, stress can also play a part in the development of phonemic length. For examples, stress may be reinforced by allophonic vowel lengthening which might become phonemic if the stress shifts to another syllable.

Long vowels may also develop allophonically in open syllables but not in closed ones. If geminate consonants occur then reduction of phonemic length in consonants can lead to phonemic vowel length, e.g. /makkar/>[makkar]>/makar/ vs. /makar/>[ma:kar]>/ma:kar/.

Long vowels can arise as the result of collapsing diphthongs, e.g. /uo/>/o:/, /ou/>/u:/, /ei/>/e:/, /ie/>/i:/, etc. where a language with only short vowels and diphthongs shifts to one with length distinction in vowels but no diphthongs (although these may arise from other sources, e.g. /ovu/>/owu/>/ou/ with the last stage occurring after original /ou/ shifts to /u:/ so as to not have /ovu/>/owu/>/ou/>/u:/).

sangi39 wrote:Long vowels may also develop allophonically in open syllables but not in closed ones. If geminate consonants occur then reduction of phonemic length in consonants can lead to phonemic vowel length, e.g. /makkar/>[makkar]>/makar/ vs. /makar/>[ma:kar]>/ma:kar/.

Or, alternatively, the exact opposite. Proto-Semitic seemingly allowed a degree of variation in V:C and VC: (ie, length could transfer between a vowel and the preceding consonant), so you could easily have /makkar/ > /ma:kar/ while /makar/ > /makar/. One prominent example is in the D verbal stem (involving gemination of the medial root consonant) – if said consonant is gutteral and therefore cannot be geminated (as was the case in many Semitic languages), the preceding vowel would be lengthened instead. So a root *K-T-B "write" would give kattaba in the D-stem, while P-ʕ-L "do, perform, act" would give *pāʕala.

Whimemsz wrote:For example, in Lakhota Sioux, sequences of vowel+(w,y,h)+vowel often coalesce into a single long vowel whose form is dependent on which the two vowels in the original sequence were; an example is iyáye "he left for there" --> [iyæ:].

These long vowels also have falling/rising tone depending on which vowel is stressed. If a word iyáye is stressed on the syllable preceding the deleted glide, the new vowel [æ:] has falling tone; if it is stressed on the syllable following, it will have rising tone.

If the two vowels are identical, the long vowel which is generated from deletion of the glide will not change. The second new vowel is [ɔ:], resulting from such back-vowel sequences as awa, etc. Also, if either of the vowels is nasal, the resulting long vowel will also be nasalized.

Mecislau wrote:

sangi39 wrote:Long vowels may also develop allophonically in open syllables but not in closed ones. If geminate consonants occur then reduction of phonemic length in consonants can lead to phonemic vowel length, e.g. /makkar/>[makkar]>/makar/ vs. /makar/>[ma:kar]>/ma:kar/.

Or, alternatively, the exact opposite. Proto-Semitic seemingly allowed a degree of variation in V:C and VC: (ie, length could transfer between a vowel and the preceding consonant), so you could easily have /makkar/ > /ma:kar/ while /makar/ > /makar/. One prominent example is in the D verbal stem (involving gemination of the medial root consonant) – if said consonant is gutteral and therefore cannot be geminated (as was the case in many Semitic languages), the preceding vowel would be lengthened instead. So a root *K-T-B "write" would give kattaba in the D-stem, while P-ʕ-L "do, perform, act" would give *pāʕala.

There's also the theory that the Proto-Semitic long vowels were originally diphthongs which were simplified to long vowels, with only /aw/ and /aj/ retained longer into development, which themselves simplify in many languages into /o/ and /e/ - this was the case of Sanskrit as well.
In Arabic, many of the long /a:/'s are actually [supposedly] the result of massive vowel+sonorant leveling; i.e. PS */k'awama/ "he stood" > Arabic /qa:ma/, */s'ajara/ "he became" > Arabic /s_?\a:ra/, etc.

Quebec French went the way of merging vowels that ended up together. Loosing many /l/ phonemes, especially with the definite article, and /r/ in /syr/, many vowels ended up in hiatus. These, in turn, merged into diphtongs that were ultimately monophtongized (with the last vowel being the result, and nasalization being kept if any vowel had it).

(1) Je m'en vais à l'école :
(a) /Z9ma~vAalekOl/
(b) /Zma~vAaEkOl/
(c) /Zma~vE::kOl/

(2) Il est dans les pommes :
(a) /ileda~lepOm/
(b) /jeda~epOm/*
(c) /jede~:pOm/

* "Il" always turned into /j/ before a vowel, preventing long monophtongization.

(3) Elle est sur la table :
(a) /ElesyRlatabl/
(b) /Eesyatab/
(c) /Etab/*

* "Elle est" is always /E:/ for some reason, probably because of an older rule I don't know. There are also theories that QcFr never had "sur" /syr/ and always had "su" /sy/, attested in Middle French and Classical non-Modern French.

In turn, today, some of them might again diphtongize.

(So yes, QcFr is full of long vowels today, and part of morphosyntax, so a lot harder to teach when teaching Norm French.)

Yiuel wrote:(1) Je m'en vais à l'école :
(a) /Z9ma~vAalekOl/
(b) /Zma~vAaEkOl/
(c) /Zma~vE::kOl/

...how do you even pull words out of that?

Nortaneous wrote:

Yiuel wrote:(1) Je m'en vais à l'école :
(a) /Z9ma~vAalekOl/
(b) /Zma~vAaEkOl/
(c) /Zma~vE::kOl/

...how do you even pull words out of that?

I was wondering that. I doubt I'll ever be able to even begin understanding the most basic Quebec French...

Nortaneous wrote:

Yiuel wrote:(1) Je m'en vais à l'école :
(a) /Z9ma~vAalekOl/
(b) /Zma~vAaEkOl/
(c) /Zma~vE::kOl/

...how do you even pull words out of that?

Have you seen close transcriptions of casual English? It's just as complex, if not worse. Words are messy in natural speech, without clear boundaries and with lots of featural spreading. This looks perfectly natural to me.

Yeah, [aU~n@U@42fVgiz tAg@mbaU?].

That could get even worse, too.

"I dunno what (other?) (?) talkin' 'bout"?

Colzie wrote:

Nortaneous wrote:

Yiuel wrote:(1) Je m'en vais à l'école :
(a) /Z9ma~vAalekOl/
(b) /Zma~vAaEkOl/
(c) /Zma~vE::kOl/

...how do you even pull words out of that?

Have you seen close transcriptions of casual English? It's just as complex, if not worse. Words are messy in natural speech, without clear boundaries and with lots of featural spreading. This looks perfectly natural to me.

Agreed completely, I must say.

Mbwa wrote:Yeah, [aU~n@U@42fVgiz tAg@mbaU?].

That could get even worse, too.

"I don't know what the fuck he's talking about"?

Whimemsz wrote:"I dunno what (other?) (?) talkin' 'bout"?

I don't know what the fuck he's talking about

at worst, though, I think I'd say something like [ãːɯ̯ɽ̃əʊ̯wɐð̠əfɐkʰistʰɔːkm̩bæə̯ʔ], which isn't too bad (edit: the /r`/ should be nasalized, not the /@U_^/... I don't know why it's doing that)

Yeah, what they said. But the 2 wasn't supposed to be there in the first one, I just missed the shift key, it should've been @.

I suppose the final [?] could drop out and I bet it would still be comprehensible if [tAg@mbaU] went to [tA(@)mbaU], although I probably wouldn't say that unless I was really tired or something.

But I admit the French example looked crazy to me too, but that's probably because I don't know French.

Mbwa wrote:Yeah, what they said. But the 2 wasn't supposed to be there in the first one, I just missed the shift key, it should've been @.

I suppose the final [?] could drop out and I bet it would still be comprehensible if [tAg@mbaU] went to [tA(@)mbaU], although I probably wouldn't say that unless I was really tired or something.

But I admit the French example looked crazy to me too, but that's probably because I don't know French.

now that I think of it, there's probably a hell of a lot of prosodic information that these transcriptions are missing (in the case of english, distinguishing long vowels arising from coalescence of vowel sequences from regular long vowels)

The Unseen wrote:

Mbwa wrote:Yeah, [aU~n@U@42fVgiz tAg@mbaU?].

That could get even worse, too.

"I don't know what the fuck he's talking about"?

ˈaɛ̯ə̃ˌnːoː wʌ d̪əˈfʌʔk çiːz ˈtʰakʰn̩ ʔbæoʔ

your diphthongs are weird

Oddly enough, the odd triply lengthed vowel /E::/ does not seem to show any prosodic difference, except for its length. And quite frankly, I don't know how we can actually take out any meaning from such sentences, even though I can. That can be part of the whole polysynthetic hypothesis of QcFr, where we should analyse all that as a messy phrase incorporating many things.

Indeed, the triple length marks "present-tense to preceeding word WITH dative+definiteness+initial vowel to the following word" That's worse than Spanish's "-í".

TaylorS wrote:

The Unseen wrote:

Mbwa wrote:Yeah, [aU~n@U@42fVgiz tAg@mbaU?].

That could get even worse, too.

"I don't know what the fuck he's talking about"?

ˈaɛ̯ə̃ˌnːoː wʌ d̪əˈfʌʔk çiːz ˈtʰakʰn̩ ʔbæoʔ

[ˈaː õn ˈnoː ˈwʌʔt̚ d̥ə ˈfʌʔk iːsʲ ˈtʰɒkn̩ː ˈb̥ɑ̟̆ŏ̯ʔ]

I think that this sort of thing is really atypical for English as a whole, and especially North American English, by any means.

(Actually, I do not think that this case is particularly dramatic when it comes to elision and assimilation in English overall by any means.)

Yiuel wrote:Oddly enough, the odd triply lengthed vowel /E::/ does not seem to show any prosodic difference, except for its length. And quite frankly, I don't know how we can actually take out any meaning from such sentences, even though I can. That can be part of the whole polysynthetic hypothesis of QcFr, where we should analyse all that as a messy phrase incorporating many things.

Indeed, the triple length marks "present-tense to preceeding word WITH dative+definiteness+initial vowel to the following word" That's worse than Spanish's "-í".

Dunno, /Zma~vE::kOl/ is maybe pushing it a bit (wouldn't that be /Zma~vE E:kol/?), but yeah something is up with that particular combination ("va à" and "à l'école), probably because of the high frequency of those words.

As for the polysynthetic French hypothesis, I think there are some arguments against that - mostly adverbs and adjectives landing in the middle of some of those clusters ("j'ai mangé"->"j'ai poliment mangé" for instance), or prepositions being matched with a too large variety of things to be affixes rather than words ("j'ai une télé pour toi", "je suis pour", "pour que tu viennes", "je suis dessus", "je suis au dessus", "Il parle de pourquoi tu veux pas", etc...)

Mbwa wrote:Yeah, [aU~n@U@42fVgiz tAg@mbaU?].

That could get even worse, too.

I've more-or-less got got [æ dənə wɔʔ ðə fʊkiz tɔkɪn əbaʊʔ] in quick speech and [æ dəno: wɔʔ ðə fʊk i:z tɔ:kɪn əbaʊʔ] in slower speech (couldn't be a***d with marking aspriation )