Re: E and e (was: A break in the evils of English (or, Sturnan is beautiful))

Ray Brown wrote, inter alia:
>as [ij] or [Ij]), phonemically we have, as I see it, to choose between /ij/
>~ /i/, /i/ ~ /I/ or even, possibly, /i:/ ~ /i/.
Exactly.  Those are the three systems I'm aware of.
1. /ij/ vs. /i/ (more commonly in the US /iy/ : /i/)
2. /i/ vs /I/
3. /i:/ vs. /i/ (actually equiv. to #1, replacing |glide| with |length|)

Well make that four, if you include Distinctive Feature analysis: (there are
plus signs here, which may not reproduce)--
underlying /i/ [+voc, +hi, +tense]-- at the surface (phonetic) level,
[+tense} --> the appropriate homorganic glide [j], or [w] for back vowels;
the 3 phonetic lengths that we see in e.g. "beat, bead, bean" are also
assigned by rules at the surface level.

underlying /I/ [+voc, +hi, -tense] (with surface length, "bit, bid, bin")
and so on for /ey/ : /e/, as well as /uw ~ow/ : /u ~o/

In old phonemic analysis, there was no principled way to explain
alternations like divine~divinity, meet~met,
telegraph~telegraphy~telegraphic; one simply had to list the various stems
(allomorphs) and their occurrence:  /di'vayn/ in isolation; /di'vin-/ before
suffixes -ity, -ation; etc. etc.  Even Dist.Feat. analysis was not much of
an improvement, since it involved ad-hoc rules (with morphological
environments)  to change /ay/ to /I/ etc.  The problem with this school of
thought was that you were required to consider only the existing form of the
language; looking back at historical developments was forbidden.

Chomsky & Halle's "Sound Pattern of English" did manage to show the
relationships, but only at the cost of proposing underlying forms that in
many cases harked back to pre-vowel-shift English.  Thus 'divine' was
underlying // divi-n // (i- for i-macron), and the [ay ~i] alternation
arises in the course of applying the phonological rules and their cyclical
stress-rule.  (Their analysis certainly implied that at least portions of
the Great Vowel Shift were still present in a speaker's competence, an idea
which many many many people balked at).  Although they did (I think) explain
English vowel alternations & stress in a great tour-de-force, it was only by
introducing a great deal of history, and some very complex rules, into the
analysis;  and I suspect very few scholars have adopted the SPE system.

At about the same time, I recall seeing analyses of Spanish that (1) posited
a deletable underlying final /e/ on (modern) consonant-final forms (so e.g.
// animale //)-- which does simplify the Span. stress and pluralization
rules, and (2) an underlying  seven vowel system // i e E a u o O // --
which does simplify the radical-changing verbs (pEnsar~pienso, mOrir~muero
and other changes); again, this introduces a lot of history into one's
analysis of Spanish.  Underlying forms end up looking suspiciously like
Proto-Romance or even late "vulgar" Latin...(e.g. // lege // 'law' modern
ley~leg/al~leg/islativo) and one or two of the really irregular verbs looked
positively Indo-European.. ('poner' IIRC had the underlying stem // pos-n-
//)
>
>Of course, it may be that the phonemic theory is incorrect - but that's
>another story  ;)
Umm, yes. :-)

>>(why is French pardoned here?).
>
>It's not pardoned.  The plain fact is tho, e.g. French _peine_ and English
>_pen_ are likely to be both transcribed as /pEn/, the vowels in the two
>languages are not the same.  The English sound is both higher and more
>retracted than the French.
and _probably_ (for US speakers, at least) involves less muscular tension.