1) Alexander Lubotsky's suggestion that *indra does not conform to
the expected Indo-Iranian vocalization, and may, therefore, be a non-
IE word;
==========
It's absurd to start with the wrong root
and then declare the word must be a loanword
because the wrong root does not provide the expected result....
A.
==============
3) John Colarusso's speculation that Indra may represent an early
North-West Caucasian loan into IE -- cf. Circassian /y@.../ 'huge +
present participle', Abkhaz /รก-yna-r/ 'the huge + present
participle', the name of the god of the forge Aynar ('the Huge One')
> *inra > *indra (with intrusive -d-), in this case originally
meaning 'the Great One'.
=============
Considering the fact that Indra belongs to
the Second class of tripartite ideology : warriors.
I suggest *?int?ara is PIE :
?i "the one who"
n- "is not"
t?ar "true, righteous, orderly"
PIE *d_r- true has #t?- as initial.
Burns temples, kills cattle, etc
Arnaud.
=============
Here is the link to the page from a book by Colarusso where this
etymology is discussed (N.B. Colarusso theorizes that North-West
Caucasian languages may be genetically related to the IE family in a
larger "Pontic" family):
http://tinyurl.com/357ype
==========
This pontic = PIE family looks really nice !!
I agree with Colarusso.
A.
=============