On 2006-07-04 09:30, Abdullah Konushevci wrote:
> First, I found that the verb is <erepho> 'to cover, to build',
Actually, both <erépHo:> and <erépto:> are attested in Ancient Greek.
The first is a simple thematic present, the second a *-je/o- stem.
> future <erepso> and <erepsomai>, aor. <erepsa>; erepsimos 'belonging
> to roof'. So, for sure, we have *H1rebh-o in verbal root.
The root is *h1rebH-, right? So far, so good.
> Initial /o/ is only found in "Rasmussen derivatives" with the
>> O-fix, <óropHos, oropHé:> 'cover, roof' and must be due to Gk.
> vowel
>> assimilation, as in <odoús> and <ónoma>. It's far from obvious
> that the
>> Albanian word belongs to the same etymon at all.
>
> [AK]
> But, that is to be waiting, due to e-o Ablaut in verb and noun stem,
> noun stem to be *H3robh-o > Greek he orophe and ho orophos 'roof',
> especially ho orophos 'rod, cane' that cover the roofs.
There is no such thing as *h1/*h3 ablaut. *h1 remains *h1 whatever
happens to the root vowel. To be honest, I'm not quite sure what should
be expected in a root like *h1rebH- with an O-fix as the regular
development. Jens claims that before a laryngeal or *r the O-fix
remained a prefix and metathesis was blocked. But what about a complex
onset with _both_ a laryngeal and a rhotic? Perhaps Jens will find the
time to comment on this. However, what seems to me to be at least a
possibility is metathesis with loss of the initial laryngeal, i.e.
*O-h1r[&]bHo- > *robHo-, not unlike *O-h2w[&]lh1no- > *wolno- (Jens's
own example). Gk. <óropHos> would then contain a genuine prothetic
vowel, with the same quality as the root vowel, added to avoid a
word-initial /r/. The analogy of erépto: ~ erépHo: (where the first
vowel seems to echo the root vocalism) may have played a role.
As an alternative explanation, we might simply assume *h1robHos,
analogical after *tómh1os etc., becoming *eropHos > oropHos by
assimilation. This, indeed, seems to be the mundane standard
explanation, though it's less clever than the one above :)
Piotr