Conservative dilemma

[Larry Trask]

On Wed, 25 Aug 1999 ECOLING at aol.com wrote:

> referring to Algonquian - Wiyot - Yurok (Algonquian-Ritwan):

>>> The question arises then, why is it that the best historical
>>> linguists of American Indian languages were quite thoroughly
>>> convinced AT SOME POINT that these languages were related,
>>> yet the standard kind of evidence was somehow lacking?
>>> What was it that they found so thoroughly convincing?

> Larry Trask's message today in which he quotes the above paragraph
> completely fails to understand the point,
> because he does not know the Amerindian field,
> and does not relize that the hypothesis I mentioned is now considered
> by nearly everyone to be solidly demonstrated scientific hypothesis.

If you are referring to Algic (Algonquian-Wiyot-Yurok), then *of course*
I know that it is almost universally accepted.  I say so in my textbook,
and I say so again in my forthcoming dictionary.

But the original passage I was replying to did not appear to be about
Algic, but rather about the assorted super-families set up mainly by
Sapir, though often extended (and occasionally originated) by others.
Lloyd appeared to be asking why early Americanists were so enthusiastic
about these groupings, and my posting was a response to this query.

> This was not some iffy hypothesis believed by uncareful researchers,
> it is a hypothesis believed by the most conservative researchers.
> The comment Trask quotes from Campbell simply is not relevant
> to this case.

Agreed, but it is highly relevant to a lot of other cases.

> The "best historical linguistics" I was referring to are the CONSERVATIVES
> whom Larry Trask I think most admires.
> It was THOSE CONSERVATIVES who were completely convinced
> that the two "outlier" languages are related to Algonquian,
> yet some of THOSE SAME CONSERVATIVES also believed that
> their standard comparative method was not adequate to prove this
> relationship.

So far as I know, the case for Algic rests solidly on shared
inflectional morphology, which is normally considered adequate for
proof.  Comparative reconstruction is of limited applicability because
there is so little shared vocabulary between Algonquian and the
California languages.

> The POINT is that the most CONSERVATIVE linguists
> were able to base their conclusions on something beyond the
> standard comparative method.  THEREFORE,
> this is a prime field in which to explore WHAT these
> conservatives found so convincing,
> precisely in order to slightly extend our tools,
> our self-conscious understanding of how we reason.

But shared morphology is not a new idea.  Proof of relationship by
shared morphology is an idea that's been around since the beginning.
It played a major role in establishing IE, and it is the evidence
appealed to in order to establish Afro-Asiatic and Niger-Congo, neither
of which has yet been reconstructed to general satisfaction.

> (Not all conservatives consider this a dilemma,
> some consider that the standard conservative method
> CAN demonstrate the relationship.  But some consider
> it cannot, and for a larger number, there was an earlier
> period when they considered it had not done so,
> yet they were convinced the relationship was real and
> was genetic.  For THOSE conservatives,
> and at THAT time period, the dilemma exists and
> provides us a great opportunity.)

Sorry, but I don't see a dilemma here.  What dilemma?

> I think Trask (along with many others)
> considers that the standard traditional comparative method
> is practically a definition of valid evidentiary reasoning.

I have never said any such thing.  You are putting words into my mouth.

> They do not consider it merely a tool, which can be
> extended, or which might have weaknesses (such as
> sometimes, when properly applied to certain data sets,
> giving the wrong result, a case I mentioned earlier).

Oh, for heaven's sake, Lloyd.  Read my textbook of historical
linguistics.  There you will find that I say that the comparative method
is only a tool, that it has its limitations, that it is potentially
subject to pitfalls, that there exist other tools which also have their
virtues, and that there is hope that mathematical methods might be
developed to take us further back in time than the comparative method
can be applied.  Where are you *getting* this stuff from?  You seem to
be setting up a fantasy Trask of your own creation and going after that.
Why not just read what I write?

> They simply are unable to treat its validity and range
> of applicability as an EMPIRICAL question.

Read my textbook, and then see if you can still post this assertion
without turning crimson.

> (Many of them would not say this as a theoretical claim,
> but in practice they act just as if they believed it.)

I don't know just who `they' is supposed to include, but I can assure
you it doesn't apply to me.

Larry Trask
COGS
University of Sussex
Brighton BN1 9QH
UK

larryt at cogs.susx.ac.uk