Dating PIE

[Miguel Carrasquer Vidal]

Larry Trask <larryt at cogs.susx.ac.uk> wrote:
 
>To begin with, the attested IE languages of the second and first
>millennia BC do give us a *terminus ante quem* for PIE, and the
>degree of divergence among the attested early languages is such that
>few people seem to be happy with putting PIE only a few centuries
>earlier than this.
>
>The linguistic evidence alone is not sufficient to give us a
>*terminus post quem*, but recall that there exists a standard
>argument for estimating this.  Many specialists are satisfed that we
>can reconstruct PIE words pertaining to technology, and most
>particularly to wheeled vehicles, including `wheel', `axle' and
>`nave'.  Now, if this is true, then it follows that the PIE-speakers
>must have known wheeled vehicles.  But the archeologists can find no
>evidence for wheeled vehicles before about 6000 BP.  Therefore, the
>arguments runs, 6000 BP is a *terminus post quem* for PIE.
 
Only approximately.  The oldest 4th millennium wheels are made of
wood, which is of course a perishable material.  Older wheels may yet
turn up.  It's also not clear whether the wheel was used first for
transport or for pottery (generally both turn up in the archaeological
record simultaneously).
 
Gamq'relidze and Ivanov list the following IE words for "wheel":
 
*kwel- (Celtic, Germanic, Balto-Slavic, [Latin])
  *kwe-kwlo- (Tocharian, Indo-Iranian, Greek, Germanic)
*Hwer-th- "to turn", but "wheel, wagon" in Iranian (Sogdian-Ossetic)
*Hwer-g^h- (Hittite [*], Tocharian)
*rotHo- (Indo-Iranian, Latin, Celtic, Germanic, Baltic)
 
Most of these words can also mean "wagon".  *rotHo- is derived from a
root meaning "to run", the others (*kwel-, *Hwer-) "to turn".  The
reduplicated form *kwe-kwl- has curious parallels in Kartvelian
(Georgian) gorgal, Hebrew galgal and Sumerian gi(r)gir.
 
Some others:
*H(o)i(e)s- "shaft" (Hittite, Sanskrit, Greek, Balto-Slavic)
*dhur(H)- "harness" (Hittite, Sanskrit)
*Hak^s- "axle" (Sanskrit, Greek, Latin, Celtic, Balto-Slavic,
Germanic)
*yugom "yoke" (general IE)
 
We can add "nave", from the root for "navel" *Hnebh- (etc.).
 
All of this makes a pretty reasonable case for not putting the
break-up of Indo-European too long before or after the invention of
the wheel (currently 4th millennium).  My own favourite date for the
split between Anatolian [in the Balkans] and non-Anatolian [Linear
Ware and related cultures], 5500 BC, is some two millennia before the
earliest known wheels.  I would obvioulsy welcome a slightly earlier
find, but I can't see any real objections against vocabulary
associated with a new technology spreading uniformly across a group of
still neighbouring and largely mutually intelligible languages
(comparable to, say, Western Romance), especially if much of the
vocabulary has transparent meaning ("turner", "spin-spin", "runner",
"armpit", "navel", "joiner").
 
 
[*] Because the Hittite word for wheel is <hurki-> (which, according
to G & I has a derivative <hurkel-> "crime" [the wheel as a torturing
device?]), I was slightly shocked to find <burkila> "spinning wheel"
while browsing through Azkue's Basque disctionary.  A Basque-Hittite
connection is ridiculous of course, but *wrg-/*wrgh-/*wrk- are good IE
roots for "to turn", and a Celtic connection would not be all that
far-fetched.  But as far as I can tell, the Celtic words for "spinning
wheel" are derived from the other variant *wer-t- (OIr. fertas, We.
gwerthyd).  Just to set my mind at ease, a question for Larry: what do
Agud and Tovar say?
 
 
=======================
Miguel Carrasquer Vidal
mcv at wxs.nl
Amsterdam