Re: Etymology of _insula_ (was Re: Thoughts on Word building)

On Tue, 13 Dec 2005 09:46:35 +0000, R A Brown <ray@...>
wrote:

>>>Of course it does. Yep, a Proto-Greek *na:sos would have give /na:os/
>>>(Doric), /nE:os/ (Ionic) and /neO:s/ (Attic).
>>
>> Interestingly enough, that is the Greek word for 'temple'.
>
>It is - as shown above in all those dialects. But there's more:
>Aiolian has _nauos_ and in Spartan inscriptions we have _nawos_, with
>'wau' or "digamma". The disappearance of original /w/ in most Greek
>dialects is well attested. We must assume, in fact, that the Spartan
>forms retain the earliest pronunciation /na:wos/
The Indo-European protoform, then, was *néxwos.  I wonder if there's any
way to relate this word to any IE root or other etymon.

[snip]

>>>However, he does connect the Celtic *srogn- with Greek _rhenkein_/
>>>_rhenkhein_ "to snore", which I suppose is possible if we have
>>>metathesis of -ghn- ~ ngh-
>>
>> Do the Greek variants have a dialectal distribution?  That is, are their
>> distributions mutually exclusive?
>
>There doesn't seem to be clear dialect distinction. It is surely to do
>with the Greek tendency to drop and aspirate if two occur in a word. One
>see this nicely demonstrated in the case forms of thrikh- "hair" -
>       sing.    plural
>Nom  thriks    trikhes
>Acc. trikha    trikhas
>Gen  trikhos   trikho:n
>Dat. trikhi    thriksi
>
>The Greek stem of 'snore' must have been rhenkh-
>
>As initial /r/ was always aspirated in earliest Greek, we might expect
>the from renkh- to become normal. But it may be that as there was no
>contrast between [r_h] and [r], the initial aspiration was not so
>strongly felt or, indeed, it may be that the forms where -nkh- are shown
>reflect a loss of aspiration in the initial /r/ - there would be no way
>of showing this in Greek.
I wonder if the initial aspirated /r/ arose due to the fact that there were
seemingly no native IE words that began with /r/.  So, all instances of
Greek initial /r/ come from *sr-, *wr-, and *yr- (if there are any of the
last one).

>The loss of initial /h/ before vowels was already well underway in the
>ancient language. It was standard in practically all Ionian dialects,
>for example; so it does not seem to me unlikely that the change from
>[r_h] --> [r] had already begun also in the ancient language.
>
>> I agree that metathesis of *-ghn- to *-ngh- is more likely than vice-
>> versa.  Plus IE */g_h/ becomes Celtic */g/.
>
>Yep.
>
>  However, one problem is the
>> Greek vocalism -- _rhenk(h)ein_ vs. _rhis_, _rhinos_.
>
>Absolutely - I don't think there is any connexion with rhi:n-, only that
>a proto-Celtic *srogn- may be connected with rhenkh-
Well, with the example you gave above, _thriks_, one wouldn't think that
there'd be a monophthong /i/ in the nominative singular of a root noun.
So, it seems that, if the word is indeed native to IE, the original
paradigm was e.g. nom. sg. *dhréighs, gen. sg. *dhrighós.  Greek then
regularized the vocalism of the word based on the oblique cases.  Such a
thing could also have happened with _rhi:s_.

>> Also, if the
>> presumed */g_h/ was part of the root, we should see Greek _rhinks_,
>> _rhinkos_ -- to my knowledge, we do not see that.
>
>We don't.
On the other hand, though, *-gh is a common "root extension" in IE verbal
morphology.

>> We *do* see, though, a
>> Greek word _rhunkhos_ (neuter s-stem), meaning 'snout' or 'muzzle'.
>> Semantically, this is similar to 'nose' in the sense of "jutting/sticking
>> out".  Interestingly, this might be related to _rhenk(h)ein_ 'to snore'
>
>Yes - it certainly looks like it. Tho I am not sure how one accounts the
>_u_.
I've been thinking about that as well.  IIRC, Greek raised /o/ to /u/
before medial coda nasals (and, apparently, any labial nasals), so it's
possible that _rhunkhos_ comes from earlier *rhonkhos.  If that is indeed
the case, I think that we can link it to the verb _rhenk(h)ein_ as an s-
stem neuter.

>> and
>> _-runkhs_ as in _pharunkhs_ 'pharynx'.  What do you think?
>
>..and in _larunkhs_ 'larynx'?  :-)
Exactly. :)

>The problem with the last two is that the words are _pharunks_ (at least
>the final letter is ksi) and _larunks_ with genitives _pharungos_ and
>_larungos_ - not an aspirate in sight.
Yes, you're right.  Sorry for the mistake.  The pair _pharunks_ and
_larunks_ seem to point to an element *runks that forms compounds, but such
a hypothesis cannot be strong unless there are other words containing that
element.  Do you know of any?

- Rob