Renfrew and IE Overlords

[X99Lynx at aol.com]

Dear Rohan,
In a message dated 1/21/00 5:10:32 AM, you wrote:

<<I personally find Renfrew's analysis of the literature unconvincing,
but that may be just me.  To prove that historical linguistics is
built upon faulty assumptions about archaeology requires going through
the latest literature to demonstrate that linguists are not regularly
revisiting their assumptions, as they should.  How does quoting
Friedrich 1970 or Gimbutas 1960 prove that?>>

First, let me emphasize that nothing I've written is meant to be
disrespectful of historical linguistics.  I certainly should not give any
impression that Colin Renfrew disrespects linguistics, which I have every
reason to believe is the farthest thing from the truth.  And I'm not
insisting that "linguists are not regularly revisiting their assumptions..."

In fact, quoting "Friedrich 1970 or Gimbutas 1960" was meant to show that
dates given by linguists for PIE dispersal HAVE changed.  If you recall my
original point was that "the earliest possible date" given by some members of
this list have backed up - and I did check back in the archives to verify
that.  The quote from the UT Austin web page in my last post should assure
you that the 3000BC-2200BC dates are still around and quite clearly, even
with the benefit of the doubt, they were intended to reflect the earliest and
latest possible dates - and are simply repeating old and dated information.
This is not an indictment of historical linguistics.  It started as and is
nothing more than an observation that old ideas die hard.

So my purpose was not "to demonstrate that linguists are not regularly
revisiting their assumptions, as they should."  But rather to figure out how
those assumptions are affected by changes in archaeological information.
Renfrew's neolithic hypothesis stretches the information but how does it
stretch the linguistic assumptions?  Is it enough to simply say that the
earliest possible date of PIE dispersal is now 3300BC?  Or is there something
that now goes on with the linguistic analysis that might alter other
assumptions or conclusions?  Does the fact that PIE speakers now appear to be
exposed to the horse a thousand years before this date and to chariots (once
considered the equivalent of PIE) a thousand years after, change anything in
the use of those items with regard to the dating PIE unity?

Also let me make it clear again that my purpose is and was not "to prove that
historical linguistics is built upon faulty assumptions about archaeology."
Faultiness is not what I'm pointing to - although on this list there have
been a number of statements about the archaeological evidence that have been
quite faulty (e.g., with regard to the evidence in early Greece.)  The
connection between archaeology and linguistics is a complex matter.  It is
not always clear to me what a piece of new archaeological evidence "means" in
terms of linguistics.  A lot of what I have been doing is just trying to
figure out what "the assumptions" are.

You use the phrase assumptions that historical linguistics "is built upon."
And that is perhaps an issue I am running into.  Is historical linguistics
finished being built?  I'm sure you would say it is not, but it raises the
question:  just how built is it?  Consider that archaeology is designed to
operate in a state of almost constant dissatisfaction.  Especially among
American archaeologists, there is a constant race to find the next oldest
date, etc.  (This parallels the mentality one sees in physics and biology,
where there is a persistent dissatisfaction with the limits of both existing
data and theory.)

There are some on this list who have privately suggested to me that they
think that the only mechanism that allows for historical linguists "to
revist" their assumptions is archaeology.  I want to believe that is not
true.  I think that a consciousness of uncertainty  allows one to explore
accepted assumptions again even without new data.  Archaeology can of course
remind one of that uncertainty.  The kind of uncertainty expressed in the
fact that if the earliest possible date of PIE dispersal might now be say
4000BC in conservative terms, then what would that extra 1500 years over the
Childe-Kossina dates mean to historical linguistics?  Does that extra time
mean anything in terms of the linguistic journey traveled by the language
between PIE and the attested daughters?  Or is everything reconstructed and
are we now finished with all that and should we move on to the Nostratic
list? :)

And finally I should say that Colin Renfrew has been working with real
historical linguists and that my speculations on how his work affects
linguistics are the efforts of an amateur and may not be of much use in that
direction.

Regards,
Steve Long