Cambridge and Greenberg's methods

[Larry Trask]

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

> I am very grateful to Larry Trask for his discussion of
> Joseph Greenberg's methods and Cambridge methods today.
> I hope we can have more discussion of the Cambridge methods.

> But I will insist (and provide evidence below)
> that Trask still does not understand that Greenberg is doing
> something COMPLETELY DIFFERENT FROM
> claiming to prove particular language families are related.

No.  Here is a direct quote from page 38 of Greenberg's Amerind book:

"The thesis of this book is that all the indigenous languages of the
Americas, except those of the Na-Dene and Eskimo-Aleut groups, fall into
a single vast assemblage.  The bulk of this volume will be devoted to a
demonstration of the genetic unity of what I will call the Amerind
family."

Nothing ambiguous about this.  Note in particular that word
`demonstration'.  Greenberg is claiming not only that Amerind is a
single genetic family but that his book *proves* it.

I'm afraid that Lloyd Anderson is trying to see Greenberg as doing
something very different from what Greenberg plainly *is* doing.
Greenberg is not putting forward hypotheses for possible consideration
by others.  He's telling us loud and clear "It's over.  Amerind is
proved."

Elsewhere in this book, and in other publications, Greenberg makes it
clear that he considers Amerind to be established beyond discussion, and
that the only interesting question is to find its closest relative (his
own candidate is his Eurasiatic construct).

> Barring the difference of whether one expresses one's conclusions as
> rooted family trees (Greenberg, I don't remember whether he used
> tree diagrams for this)

G does indeed propose rooted family trees of the familiar kind, though
he considers his subgrouping of Amerind to be less secure than the
reality of the family.

> or unrooted family trees (the Cambridge Group that Trask refers to),

I am not aware that G has ever made any use of unrooted trees, or even
expressed any interest in them.

> there may be little difference between them (little difference in
> matters noted in Trask's discussion today; of course there may be
> other differences that we consider important).

There are huge differences.  The Cambridge work merely tries to quantify
linguistic distance; the results are (or may be) expressed as unrooted
trees, and *no claims whatever* are made about genetic relations.

> [LT]

>> The assumption that all languages are related is out of order.

>> The assumption that all languages *might* be related is hardly an
>> assumption at all, and in any case such an idea is excluded by no one.

> The assumption is of course quite in order, to see what it might
> lead to, as long as it is merely that, an assumption. So there is
> really no difference between the above two statements just above.

Statement 1: All languages are related.

Statement 2: All languages might be related.

You really can't see any differences between these?  Gad.

I know of no one who queries (2), but I know of very few people prepared
to defend (1).

> Except some emotional flavoring which antagonizes some people.

Sorry, but the difference between (1) and (2) looks to me like far more
than "emotional flavoring".

> All one needs to understand Greenberg's reasoning is to assume
> that all languages *might* be related.

No.  I am happy to agree that all languages might be related, a position
which commits me to nothing at all, apart from a rejection of the
contrary proposition that polygenesis is certainly false.  But it does
not follow that I have to accept *any* of Greenberg's conclusions.

> As far as I can tell, the only difference noted by Trask
> is in whether one expresses one's tentative conclusions as

> (if all languages are related, then here are some
> conceivable family trees for the deeper connections),
> or as
> (here are some conceivable deeper connections,
> though we do not express them as family trees).

Well, I won't be drawn into this.  I'll just point out, yet again, that
neither of these has anything to do with Greenberg's work.  Greenberg's
work, on Amerind and elsewhere, takes the following form and no other:

	"All the languages listed below are *certainly* related."

> In terms of politics, it would perhaps have been more useful in the
> long run if Greenberg had not expressed his conclusions AS IF they
> were family trees.

A fundamental misconception.  Family trees are *intrinsic* to what
Greenberg does.  Indeed, he's not even interested in anything other than
family trees.  And politics has nothing to do with it: it's the
*linguistics* that's at issue.

If Greenberg had wanted to be politic, he might have said something like
this: "Here is some interesting evidence suggestive of a possible
hypothesis.  Maybe it's worth looking into this."

But he does no such thing, and moreover he plainly has no interest in
such a softly-softly approach.  He thinks he's proving family trees.
End of story.

> That is clearly a red flag to some people.  They seem to treat such
> expressions as if they were final conclusions proven by sufficient
> evidence,

But this is exactly *Greenberg's* position!  Read what he writes.

> for two languages or language families taken in isolation from all
> others, rather than one investigator's claim that the evidence he
> evaluated as he evaluates it suggests those trees rather than some
> others (merely that, the preferable conclusion to others rather than
> a proven conclusion).

Again, nothing to do with Greenberg, unless we impute to him the idea
that monogenesis is true and provable.  In fact, he probably does
believe this, since he's hinted it heavily on several occasions, but so
far he has stopped short of asserting it in print.

The point is not whether tree X is better or worse than tree Y: the
point is whether there is any justification for drawing a tree at all.

> Personally, I have no problem at all treating that as the expression
> of a tentative hypothesis only (how could it have ever been anything
> else?),

But that's not how Greenberg presents it.

> and I have been simply amazed at the antagonisms.

Here I have some sympathy.  The stridency of a few of G's critics is
greater than I consider proper.  But we shouldn't let that unfortunate
stridency blind us to the very substantial *linguistic* critiques of G's
work.

> Shouldn't we always try to make the best use of the contributions of
> each one of us? Didn't Greenberg attempt to systematize a large
> amount of data which others can then correct and improve on?

G's own view is that such corrections and improvements can extend only
to the details, to the subgrouping and to the individual comparisons.
He does not allow for the possibility that his vast mega-agglomerations
like Amerind are merely phantasms -- as many of his critics maintain.

[LT on the Cambridge work]

>> If the same is true of Greenberg's highly informal approach, then G
>> cannot distinguish relatedness from unrelatedness,

> Exactly.  I have always, from the very beginning, understood that
> was exactly what Greenberg's approach did.

Nope.  Read what he writes.  He damn well *does* believe that he can
distinguish relatedness from unrelatedness -- insofar as he recognizes a
concept of unrelatedness at all, which I frankly doubt that he does,
given his constant asides about "Proto-Sapiens".

> He would classify Basque and Chinese as the most divergent members
> of such a set.

You have a point here, though not the one you intend.

For G, *no* language ever gets left out of his agglomerations.  The only
issue, for him, is which agglomeration to put it into.  And even that
seems to be only a temporary measure, pending the revelation of the
single Great Tree to which all languages belong.

Johanna Nichols has described G's refusal to accept isolates or
unaffiliated tiny families as "pathological", a view which I endorse.

The Cambridge group don't distinguish relatedness from unrelatedness
because, with their methods, they can't.  Greenberg doesn't because he
won't.

> He would conclude neither "related" nor "unrelated" if he did not
> assume all languages were related.

Can you cite one single passage from any of G's publications in which he
concludes that any two languages or families are "unrelated"?  Or
"neither related nor unrelated"?

> Given the assumption of all languages
> potentially related, which since unprovable amounts merely to a way
> of expressing one's hypotheses,

Well, starting out by assuming the truth of something you consider to be
unprovable strikes me as a daft way of proceeding.

Here's something I consider unprovable (and unfalsifiable):

	The universe was created last Tuesday morning, and all of us
	were created complete with false memories of our earlier
	existence.

Now, would it make any sense for me to assume the truth of this, and try
from there to establish anything interesting?

> he concludes "most divergent", just as he in fact did for the
> families Athabaskan and Eskimo-Aleut.  The difference in these modes
> of expression is utterly trivial as long as we are concerned with
> tentative hypotheses, not with ultimate truth.

But G firmly believes that he has established "ultimate truth".

> [LT]

>> and he has no
>> business setting up imaginary "families".

> Given that this is merely a way of expressing degree of divergence,
> under Greenberg's assumptions, I find it unobjectionable.
> I don't draw any more conclusions from it than are warranted,
> one investigator's tentative judgements of stronger vs. weaker
> resemblances.  That is all.

I am speechless.  I guess it's time for me to withdraw from this
discussion, dubtless to the great relief of the moderator.

I close with a reaffirmation of my central point: Lloyd Anderson's
attempt at characterizing Greenberg's work bears no relation to
Greenberg's work.

Larry Trask
COGS
University of Sussex
Brighton BN1 9QH
UK

larryt at cogs.susx.ac.uk