Three-Way Contrast of Secondary Articulations in PIE

[Stanley Friesen]

>> Trouble is, for the stage before Old Indian, CaC roots are very rare, and
>> CeC ~ CoC roots are morphologically conditioned alternatives.  Some writers
>> would only accept CeC roots , so for them your argument collapses.

>[PCR]
>My first comment would be that there are NO CaC roots in earliest IE, only
>Ca:C, the result of CaHC.

>Secondly, if, as I proposed in my answer to Nath, vowel-quality was
>neutralized with the semantic load carried by them transferred to glides, the
>natural neutral vowel is /a/ so my scenario would have all originally
>short-vowel roots as CaC, from which roots the e-o-Ablaut developed.

This stage, to the extent it existed, predates PIE proper, since even the
oldest directly reconstructible stage has e~o-Ablaut.  Any pre-ablaut stage
is arrived at by internal reconstruction.

>[PCR]
>I have a feeling that you might wish to retract this opinion upon further
>reflection. IE, as it has been reconstructed, is rife with purported homonyms.
>What originally distinguished these homonyms were glides, and before that, in
>Nostratic, different vowel-qualities.

Why need they always be distinguished?  Real languages today have much
homonymy.

But even then, what I see in the PIE vocabulary is many homonymous *roots*,
but relative fewer homonymous *words*.  The main distinguishing factors
were differences in suffixes and stem formants - and occasionally infixes.

>All roots with an unmodified vowel in IE are Ce/o/-C.

>Roots with Ca:C, Ce:C, Co:C are all the result of an internal 'laryngeal', or,
>in the very rare case, the effect of former aspiration. CaC roots are Ca:C
>roots that have lost length.

Long vowels in PIE seem to also come from other sorts of compensatory
lengthening, such as degemmination of a following double consonant.

--------------
May the peace of God be with you.         sarima at friesen.net