Excluding data

[Larry Trask]

On Wed, 15 Sep 1999, Jon Patrick wrote:

[LT, on Basque <taup> `heartbeat']

>     Of course.  But this word is not general in Basque.  It is more or less
>     confined to the center of the country, being restricted, as far as I can
>     determine, to the Gipuzkoan dialect and to adjoining parts of the
>     Bizkaian dialect.  It appears to be unknown in the French Basque
>     Country, unrecorded in the Pyrenean dialects and in High Navarrese, and
>     not general in the Bizkaian dialect.  Furthermore, the word is only
>     first recorded, in the form of its derivative <taupada>, in 1888.

>     On top of this, the word violates at least four of the
>     morpheme-structure constraints which are generally obeyed by words
>     meeting my criteria:

>     (1) No initial voiceless plosive;
>     (2) No initial coronal plosive;
>     (3) No final plosive;
>     (4) No final labial.

> This is a key point that I opened up in my previous mail item. I
> speculate that the initial 6 criteria you defined, create a close
> coincidence with an expectation about what the phonological profile
> of early euskara should be (as for example what is defined above) ,
> hence the possibility of revealing more structure to euskara is
> limited by this unspoken correspondence.

Sorry, but I don't see how my criteria do anything of the sort.  The
word <taup> is excluded because of its very late first attestation and
because of its limited distribution in the language, not because of its
form.  If the word had proved to be found throughout the language and
first recorded in the 16th century, then it would be in my list,
regardless of its bizarre form.  The four observations listed above are
just that: observations about what appears to be generally true about
words that *do* satisfy my criteria.  The word <taup> is not excluded
because it fails to meet these four criteria: it is excluded for other,
and highly principled reasons, and I merely note afterward that it
furthermore has a bizarre form.

Now, in a study of Pre-Basque, as opposed to a study of modern Basque,
what would be the point of including <taup> in a list of lexical items?
This word is *most* unlikely to be ancient in Basque, and it is a fine
example of the kind of thing I want to eliminate from consideration at
once.

> I seek to test my speculation by using a less restrictive criteria
> and by studying classes of words not previously given individual
> scrutiny. You could say why bother to do this when talented scholars
> have already worked over all the material available in euskara. I
> have two responses to that. Computers enable us to do a more
> systematic job on a larger volume of data, more quickly, hence we
> are likely to pick up omissions and oversights of earlier workers.

Well, I have nothing against computers, but I don't regard them as a
form of magic.  If you put together a list of words 95% of which are not
ancient in Basque, what exactly can your computer program do that will
be informative about Pre-Basque?

> A bit like using modern technology to reprocess the tailings of 19th
> century gold mines.

A colorful analogy, but I'm afraid I don't follow it.

> Secondly, my experience has taught me that linguists don't know
> their material as well as they think they do, so to me statments of
> generalisations I take a little more scpetically than most others.

But the generalizations can come only after the list has been compiled
in the first place.  We are talking about how the list should be
compiled, not about the generalizations that will emerge from it --
though, as I have pointed out often, I *think* I have a pretty good idea
what those generalizations will look like -- though I'm prepared to be
surprised on occasion.  But, once more: I *never* exclude a word from my
list because it doesn't match any generalizations about form which I may
have in mind.

Larry Trask
COGS
University of Sussex
Brighton BN1 9QH
UK

larryt at cogs.susx.ac.uk