Lehmann's Syllabicity

[petegray]

>... the unusual fact that most IE
>roots display a front-back vowel (also, potentially no vowel) contrast,
>which indicates morphosemantic differences only --- not lexicosemantic
>ones.

The word "vowel" needs defining here!   Woul you count a sequence "eu" as
the same as, or different from, the vowel "e"?  If (as I believe we must) we
count it as different, then your problem disappears.   Ignoring the e/o/zero
ablaut, PIE shows:
    e
    eu
    ei
    er
    el
    em
    en
    eH 1 -2 -3 - etc
All of these carry lexicosemantic differences.

Even if we don't like the sequences er, el, em, en as diphthongs, we must
still recognise the zero grades as vocalic elements - vowels - with
lexicosemantic significance.

Peter