From: tgpedersen
Message: 54908
Date: 2008-03-09
>http://tech.groups.yahoo.com/group/cybalist/message/6678
> On Sun, 9 Mar 2008 10:21:18 +0100, "fournet.arnaud"
> <fournet.arnaud@...> wrote:
> >Having no particular dogma to sell,
> >I try to look at data in order to understand
> >what we can get from it.
> >
> >Chinese Mandarin san1 "three"
> >and Tibetan gsum have a clear relationship
> >with Uralic Hungarian ha:rom.
>
> Of course those stupid Uralicists derive the Hungarian form
> from PFU *kolmi (Finn. kolme, Saami gol'bmâ, Mordvin kolmo,
> Mari këm ~ kuum, Udmurt kwin'(m-), Komi kujim; Mansi kuurem,
> Xanty käälem, Hung három).
>
> >and Basque hiru with loss of -m
>
> The Basque form is hirur, as shown by the dialects that
> preserve the final flap -r. There is no room for *m there.
>
> Speculatively, the Basque form can derive from any of the
> following pre-stages: {ptk0}i{ltd}u{td}. If we unite t/d as
> T, that gives the following possibilities:
>
> piluT piTuT
> tiluT tiTuT
> kiluT kiTuT
> iluT iTuT
>
> It is obviously tempting to choose *tilud, which would
> connect the Basque form with PSem. *t_ala:t_ (c^ala:c^),
> PBerber *s^rad. ~ *k.rad. and PIE *trej-/tr.ti- (<
> **tiláti).
>