Re: THEORY: Laxness?

>Why is [I] lax and [i] tense, [E] lax and [e] tense? I understand the
>terms "tense" and "lax" are relatively poorly defined, but if a
>language has spots in which vowels go that differ in length (e.g.
>because of a length distinction, or open vs closed syllables), it is
>common to find the "lax" one in the short position, and the "tense"
>one in the long position. This is often described in terms of
>undershoot: in short spots, in order to reach the tense target, the
>tongue would have to move too fast (or something similar), and so it
>doesn't quite reach the target; this failure may become phonologised
>so that these accidents become standard allophones or even even
>distinct phonemes with loss of the earlier length.
>
>This description implicitly takes as its starting point a neutral
>resting position of the tongue, something in the vicinity of [@] (or
>perhaps even lower; [E] and [O] are lower than [@]). But in connected
>speech, vowels don't alternate with a neutral resting tongue position,
>nor (for the most part) with other vowels (which might draw them to an
>average position likely around [@]). They alternate with consonants,
>which mostly involve closures towards the top of the mouth, and draw
>the tongue upwards (the main exceptions to this are pharyngeals, which
>draw the tongue down, bilabials, which do not involve the tongue, and
>possibly uvulars---I'm not quite sure). So wouldn't undershoot
>encourage [i] and [e] in comparison to [I] and [E], which are further
>away from where the tongue is, and where it's going to?
>
>Am I missing something comparatively obvious here?
>--
>Tristan.
I'm not shur' if it should be obvious, but there's the fact that the tongue
is not a single entity. Coronals (=dentals, (post/)alveolars & retroflexes)
mostly use the front part; vowels and dorsals (=palatals, velars & uvulars)
use the middle part. So just like gutturals and labials, a coronal such as
/t/, in itself, does not really en- nor discourage /i/ nor /I/. Laxing,
AIUI, is more of a Sprachbund-like feature. A seemingly mutually exclusiv
alternate would be palatalization; so if a language tends to lax the tongue
during consonant articulation, you'd be more likely to get [ti] > [tI]; if
it doesn't, you already have [t_ji].

A follow-up question might be why even velars, however, seem to just about
never raise vowels...? A partial explanation could ibe that an allophonic
[q] or [k_-] before low back vowels is more common that we realize, while
low front vowels are equally far from all POAs to default to [k]... But I'm
just speculating by now.

Ooh, wait! One more thing. I've understood that with lax vowels, the
rounding (or lack of it) tends to be much less exagerrated than with tense
vowels. I remember when I introduced sum'one to the concept of vowel
rounding & he proceeded to classify (English) /U/ as "unrounded" and /i/ as
"wide" ("anti-rounded"?)...

John Vertical