Peter:
> cardinal *[d]k^Nt-iH1
> cardinal *[d]k^Nt-es or *[d]k^onts; gen. *[d]k^Nt-óm in *k^Ntóm
> "100" < *dek^Nt dk^Ntóm "decad of decads"
> cardinal *[d]k^ónt-H2
I know that it's a very popular view to add *d's where they are never
ever attested, but I like to deal with the facts... The facts are that
*d- is NEVER found in the plural decad ending *-komt-/*-kmt-. We've
discussed this before and all that people can come up with is the
preceding lengthening seen before the *k in some later languages which
many presume to be compensatory for a theorized loss of former *-d-.
In reality, this lengthening is not standardized amongst the cognates
that show this length and further we all can see how lengthening can be
caused by a lot of things, more frequently by a loss of _laryngeal_.
While compensatory lengthening caused by a bona fide loss of *d is
completely rare if not absent altogether from the extensive records we
have, it's hard not to study IE for any length of time without tripping
upon compensatory length caused by loss of any of *h1, *h2 or *h3. That
I believe is the origin of the lengthening that has then spread to other
decads out of textbook assimilation seen in most of the world's number
systems. Namely, the preceding lengthening is either caused by *-x (the
collective *-h2) or *-h (the dual *-h1) which was suffixed to the
preceding numeral before the decad ending, thereby causing the length.
Not good enough for you? Need more facts? Another fact is that while
we not only NEVER see **-dkmt-/**-dkomt-, we also coincidentally NEVER
see **dkmtom. Why? We can either conclude based on absolutely nothing
that *d assuredly exists, which amounts to nothing more than the trappings
of a linguistic cult, or we can accept that *d isn't there in any of
the non-singular forms and never was. The actual attested form *kmtom
is *komt- plus the genitive plural *-om. Like other words like *yugom
and *pedom, *kmtom signified a collection of "tens" (note PLURAL "tens",
not singular *dekm). Some think that it is a shortened form of a
phrase "(tens) of tens" but I don't this is necessary. Regardless, both
of these etymologies suggest the development "a group of many tens" =>
"a hundred".
What we in fact see if we open our eyes to the facts, rather than
assumptions, is that there are _two_ forms of "ten", one singular (*dekm)
and the other non-singular or dual-plural (*komt-). As I alluded to in
a previous post, one is from a fossilized phrase meaning "(one) ten"
and the other from the plural form with former plural *-it (> IE *-es).
The *t is preserved in *-kmt- because the *t is never in final position
to be sibilantized in any of these "tens" words.
= gLeN