Hittite has some of the Neolithic technological terms shared by the
other IE groups, e.g. <hars-> 'to plough, dig, cut the soil', which
is relatable to *h2ar-h-, and <jukan> 'yoke' < *jugom. However, it
has no words based on *weg^H- (the most widespread non-Anatolian root
meaning "convey, transport in a vehicle"); the Hittite "wheel" word
is mysterious <hurki>, which may or may not have a Tocharian cognate
(<wärkänt> 'circle, wheel', a possible reconstruction being derived
from _very_ hypothetical *h2werg- 'turn'; Puhvel adduces Germanic
*wurgjan- 'strangle', but the semantics is so-so at best), but has
also been suspected of Hattic origin. The Anatolian words
for "horse" -- Hieroglyphic Luwian <a-su-wa-> and Lycian <esb-> (for
Hittite we only have the Sumerogram ANU.KUR.RA) -- look more like
loans from (Mitanni) Indo-Aryan *as'wa- than native developments of *
(h1)ek^wo-.
The percentage of inherited vocabulary is not a very good measure of
relatedness (quite apart from the question what should be taken into
account -- types or tokens, basic vocabulary or the entire known
lexicon). In contact situations borrowing from a prestigious language
may happen quickly on a massive scale. Still, the complete absence
of "wagon terminology" from Anatolian (especially of the most popular
terms like *weg^H- or any of its numerous derivatives, *kWekWlo-
'wheel', *h2ak^-s-i- 'axle', etc.) is curious. In all the other
branches there may be one or two accidental gaps in this lexical
field, but not so much blank space.
Exactly how much in Anatolian was borrowed is hard to estimate
because of the general philological problems complicating Anatolian
etymologies. Moreover, since Anatolian seems to be a primary branch
(or subfamily) of IE, the chances that many words survived in
Anatolian but were lost elsewhere are relatively high. We can be
reasonably sure that an Anatolian word is a loan only if a likely
source form is documented in Hattic, Hurrian, Akkadian or another
known language and/or if the form of the word clearly gives away its
non-IE origin.
Piotr
--- In cybalist@..., george knysh <gknysh@...> wrote:
> *****GK: May I ask a question here (which is actually
> related to my previous question concerning non-IE
> elements in IE languages)? If Anatolian IE languages
> do not share words re "wheel" "vehicle" and "horse"
> with other IE languages, what is the source for their
> words as to these (for I assume that Hittite etc. do
> have words for "wheel" "vehicle" and "horse" do they
> not?). And a follow up: if Mallory (citing someone
> whose name escapes me for the moment==I left his book
> at the dacha (:=))====) claims that perhaps as much as
> 50% if not more of Hittite vocabulary is non-IE then
> are far-flung conclusions here not a bit chancy? What
> guarantee do we have that the "Hittites" when they
> moved into "the land of Hatti" were not subjected to
> enormous cultural pressure from the locals, which led
> to the abandonment of many IE words in the eventual
> hybrid language developing from this interplay? I
> suppose, as a logician rather than a linguist, that
> the archaic nature of the Hittite IE word baggage can
> otherwise be most useful for cautious genetic
> reconstructions, but is one entitled to assume that
> non-presence of an expected word in their language is
> proof that it was never there? Mind you this argument
> has significance to the extent that the % cited above
> is in effect accurate or close.******