Re: [tied] numeral ( it was Romanian Swadesh list -> 10% substratua

09-12-03 18:41, alex wrote:

> I sustain:
> PIE *oinos > Rom. "unã" via substrate and not Latin.

How foolish of everybody else not to have noticed such an obvious thing!

> My argument: For the change oi > u we have attested the change oi > oe >
> in Thracian:
> Oiscus > Oescus > Uscu

Non sequitur. Even if the Thracian change were genuine, that wouldn't
disprove the straightforward Latin etymology.

> I sustain:
>
> PIE *duu > doo > douã in Rom, the masculine form "doi" being probable
> influenced by trei, but keeping its "do-" there. The fact the there is
> an "o" , exclude any Latin speculation.

Sorry, but here you slide into sheer idiotcy. First, there is no "PIE
*duu", and if there were such a thing, the rest of the "derivation" is
completely arbitrary. The /o/ is Latin, of course. In Vulgar Latin,
*d(u)o came to be treated as a root from which new gender forms were
derived, including, among other variants, masculine *d(u)o-i and
feminine *d(u)o-e (with regular plural endings). The latter form
(perhaps with some analogical influence of 'nine') gave ORom. (attested)
doauã (with an epenthesised glide and regular diphthongisation of *o),
simplified in Mod.Rom. douã.

> My argument:
>
> -Latin "dues" cannot give "do-" in Rom.

Your ignorance is abysmal indeed, and matches your phenomenal hubris.
There was no Latin "dues". Latin had masculine/neuter <duo> and feminine
<duae>.

> -PIE short "u" appears to be very open in Thracian (see Reichenkron
> arguments for it. If you don't have them, let me know and I will show
> them here).

No, thanks. This is irrelevant, no matter if Reichenkron is right or wrong.

> I sustain:
> PIE *tri > trei in Rom.
>
> My arguments:
>
> a) -Latin tres cannot give trei in Rom. The "i" is not explainable
> trough Latin tres

Of course it is. The simples derivation is that shown by Miguel: VLat.
tres > Eastern Romance *trei > Romanian trei. There are lots of parallel
cases, so the correspondence between <tre:s> and <trei> is well
supported and systematic.

> b) -Thracian glosses: Treibitos ( too Traibitos), Treicentos, Treisucu,
> Tricornii, Triballi.
>
> The Thracian forms have all the requested "i" and will explain the Rom.
> "trei".

Russian also has <tri> for '3'. Does it mean that Romanians are Russians?

> Agains Thracian glosses one can counter-argue that "we don't know what
> "trei-", "trai-", "tri-" in Thracian meant". I agree but a such argument
> is to put in co-relation with all PIE roots we know and to see if there
> fits anything as the PIE *tri-

As I have already told you, Latin was also IE, so <trei> no doubt comes
from PIE *trejes. Poor Thracians are dead and can't protest against your
handling of the sparse evidence they left to posterity. You could at
least have the decency to check the Latin words and PIE reconstructions
you quote (or rather misquote -- nay, invent). A fragmentarily
documented language is easy prey to somebody like you -- it is easy to
write irresponsible fiction about it, since there is no concrete
evidence to constrain your imagination. But you treat attested and
well-known languages with the same cavalier disrespect for the facts.

> Since I am sure you won't agree with this demonstration, or maybe you
> will call it a laugh-number but not a demonstration,

Of course you haven't demonstrated a single thing. You don't even seem
to know what counts as a demonstration in linguistics. Which is odd,
given the time you've spent on this list.

> I will beg you to
> show me why and where are the weak points in my argumentation for
> considering it as having no value.

See above. And now you've had your fun, so please stop it.

Piotr