Hittite <hurkis>/wheel

[Larry Trask]

Steve Long writes:

>  (When I asked Sean Crist to identify the "telltale signs of borrowing" that
>  he offered that would tell him if "the wheel word" was borrowed in at least
>  some IE languages, he never replied.)

OK.  I'm no IEist, and I can't evaluate the IE evidence here.  But I can
illustrate the general point Sean was making with a Basque example.

All but the westernmost dialects of Basque have a word <(h)anka> 'haunch', and
also 'buttocks', 'leg', 'paw', 'foot' in places.  This word is regarded by all
Vasconists as a loan from Romance.

A certain long-ranger has recently been interested in finding evidence of a
genetic link between Basque and the two North Caucasian families.  He has noted
that Basque <(h)anka> looks quite a bit like something in Caucasian, and he
therefore denies the loan status of <(h)anka>, insisting that the word must be
native and ancient in Basque, and therefore cognate with the Caucasian item.
How can we reply to him?

Well, the problem is that cluster /nk/.  This was indeed perfectly normal in
Pre-Basque.  But, in the early medieval period, Basque underwent a categorical
phonological change, in all but the easternmost dialects, by which plosives
were uniformly voiced after /n/.

For example, the Latin word <incude(m)> 'anvil', which was borrowed early into
Basque, appears today as <ingude> in all but the easternmost dialects.

Likewise, the native adverb-forming suffix <-ki> appears today as <-gi> after
/n/.  For example, <eder> 'beautiful' forms <ederki> 'beautifully', but <on>
'good' forms <ongi> 'well' in all but the easternmost dialects, which alone
preserve <onki>.

Now, the word in question is *everywhere* <(h)anka> in Basque, and no such form
as *<(h)anga> is recorded anywhere.  Therefore, the very form of the word is
enough to *prove* that it was not in the language at the time of the change,
and must have entered the language later -- from whatever source.

As it happens, we know the source: it is the very widespread Romance <anca>
'haunch', with regular regional developments like French <hanche>, all
ultimately from a Frankish *<hanka>.

But, even if we didn't know the source, the form of the word would tell us at
once that this is a late entry into the Basque lexicon, and therefore probably
a borrowing.

Larry Trask
COGS
University of Sussex
Brighton BN1 9QH
UK

larryt at cogs.susx.ac.uk