Script on excavated terracotta seals from Harappa deciphered ????

[Jacob Baltuch]

Luis Gonzalez-Reimann wrote:

>Probably the main difference between Semitic and Indo-European verbal roots
>is that Semitic roots don't include vowels, whereas Indo-European roots not
>only do include them, but they also incorporate an elaborate system of
>vocalic transformations (guNa, vRddhi).

I think you are using "IE" for Sanskrit. As far as I know IE roots
do not have an intrinsic vowel which then undergoes two degrees of
lengthening, guNa and vRddhi, but are roots with no intrinsic vowel
in which the vocalization can take two quantitative degrees (long/short)
combined with two qualitative degrees (e/o) or of course the reduced
degree with no vowel, for a total of 5 shapes the root can come in.

The loss of the qualitative distinction in Sanskrit comes I think from
the collapse of IE *e and *o to Sanskrit a. As to the intrinsic vowel
of Sanskrit roots I've never quite understood where it came from in IE
terms (but that's my fault). I mean, it is obvious it is obtained by
saMprasaaraNa from a semi-vowel. But I didn't know all IE roots _had_ to
have a semi-vowel. I probably should look into this more carefully some day :)

Anyway, to get back to Semitic and IE:

I'd say the main difference between Semitic and IE roots is that the
great majority of Semitic roots have three consonants where the great
majority of IE roots have only 2 consonants (observing however that
the "consonants" of IE can sometimes be clusters e.g. a root can start
with cluster /sw-/ whereas in Semitic they are always single consonants.

Otherwise I don't know if there is such a big difference in how
the roots are treated.

Just like an IE roots has 5 basic forms in which it can be used
(CeC, Ce:C, CoC, Co:C, CC), Semitic has vocalic patterns which
are mapped into the root. These vocalic patterns are made of
a, i, u, a:, i:, u: and (according to the prevailing description)
gemination of the 2nd root consonant (which gives the so called
heavy forms) For example from a root CCC you could have forms
such as Ca:CC, CaCiC, etc. The only difference with IE is that
in the first case a choice is made from a set of 5 elements, and
in the second case 2 choices are made from a set of 7 elements.
Note though that not all possible patterns are used in Semitic.
In a Hebrew e.g. there's only about 20 to 30 vocalic patterns of
the root which combined with the use of suffixes and prefixes
gives the roughly 100 patterns or so according to which all
adjectives, nouns and verbal forms are formed. (The restriction
to a, i, u, a:, i:, u: is only valid for proto-Semitic, Arabic
and some other but not all Semitic languages; for example
Akkadian, Hebrew, Aramaic, Ethiopian, ... have more than three
vowel colors which in principle can increase the number of patterns)

One minor difference is that in IE suffixes too can undergo
vocalic alternance whereas in Semitic they can't.

In any case, Semitic roots do _not_ get used as clusters of
three consonants with no vowels, but use an intricate system
of vowel patterns to give rise to the actually occurring forms.