Respect goes both ways!

[Larry Trask]

On Tue, 5 Oct 1999 ECOLING at aol.com wrote:

> Steve Long this morning sent me a reference to p.177 of Larry
> Trask's "Historical Linguistics" where he claims to be quoting Trask
> to the effect that regional varieties of a language "would
> eventually become so different from one another that they would
> cease to be mutually comprehensible at all, and we would be forced
> to speak, not of different dialects, but of different languages."

> Here it appears that Trask is indeed using just the criterion he says
> he does not use, at least as a part of his definition which in this
> case is a determining part.

> Even if the quotation Long sent me is within a context of a
> criticism of the technical "mutually intelligible" definition, it
> still demonstrates that Trask CAN think logically using this
> definition, and come to the exact same conclusion as I believe
> anyone else.  Simply common sense. If so, why does he refuse to do
> so in the present discussion on our email list?

> Is the reference Steve Long sent me correct, anyone?

Yes; it's correct.  I wrote this, and I see nothing wrong with it.

But let's consider the context.  I wrote this in an elementary textbook
of historical linguistics for beginning students.  Now, beginning
students often do not understand what we mean by the term `genetic
relationship', and they often do not understand that a language can give
rise to a family of diverse daughters.  Since this point is fundamental,
I try hard in my textbook to get it scross.  In the process, I silently
overlook any number of complicating factors.  This is necessary.
If I simply present all known complications from page one, my readers
will be lost: they won't be able to see the wood for the trees.
Students have to learn the basic concepts before they can get to grips
with the complications.

Does a clarinet teacher try to teach all known techniques on day one?
No?

But what I say to beginning students is one thing.  And what I say to
professional colleagues is another.

> Even before receiving this reference, I wrote the following
> paragraph: If the correspondent really believes there is NO
> acceptable technical definition of "same language" vs, "different
> language", then perhaps the correspondent should simply have said
> that and declined to participate in this discussion.

I think I *have* said this.

To repeat: there is no principled and watertight definition of what
constitutes a single language.  That is a fact known to all linguists,
I think -- though sometimes I wonder about the Chomskyans. ;-)

Nevertheless, it is often convenient, and even necessary, to proceed as
though individual languages exist and can be identified -- so long as we
never forget that we are making simplifying assumptions.  However, as I
have pointed out, if we forget that we are making these simplifying
assumptions, and start reifying languages into discrete and individual
entities, comparable to the individual seagulls that wander about
outside my window, then we're in deep trouble.  And it's this last point
that I was emphasizing.

Larry Trask
COGS
University of Sussex
Brighton BN1 9QH
UK

larryt at cogs.susx.ac.uk