Pre-Basque phonology (part 3)

[Larry Trask]

On Wed, 15 Sep 1999, Roslyn M. Frank wrote:

> Earlier I brought up a problem related to the reconstruction based
> on the following data set of attested items where in southern
> dialects the forms are monosyllabic, e.g., /be/ "low, below,
> beneath" while they are bisyllabic with aspiration in northern
> dialects, e.g., /behe/ "low, below, beneath." Northern dialects are
> characterized, in general, by aspirated consonants which are to my
> knowledge, generally speaking, absent from the southern dialects
> today

Correct, though only certain consonants, in certain positions, can be
aspirated in northern varieties today.  And recall that the aspiration
is abundantly recorded in the western dialects in the Middle Ages,
though only /h/.  Aspirated plosives are not recorded there, though
they may have existed anyway.  Other aspirated consonants *are* recorded
there: for example, medieval <Elhorriaga> for modern <Elorriaga>, a
place name.

> In an unrelated discussion Larry brings up the fact that he will
> accept the root-stem <ke> "smoke" for his database of reconstructed
> items:

Right.  I have to.

> However, my question is the following: since this root stem has
> several different attested representations, which one should you
> choose? You seem to have chosen a southern variant, namely, /ke/.
> I refer to the fact that this item is often pronounced /ke/ and
> /kee/ in southern dialects but frequently /khe/ in northern ones. I
> emphasize the fact that /khe/ is considered a common variant of this
> item in the northern dialects, but not */khehe/ to my knowledge. Is
> this evidence for anything?

The word is usually <khe> in the north, <ke> in the south.  I query the
reality of the suggested *<kee>, which I have never encountered
anywhere.  True, <k(h)e> plus the article <-a> is commonly pronounced
<k(h)eia> in many varieties, but this is easy to explain, and not
relevant.  Certainly *<khehe> does not exist, and it couldn't: the
aspirating varieties permit only one aspiration per word.

> And to make things more complicated there is ample evidence for a
> variant in /ekhe/ "smoke" in northern dialects whereas this appears
> as /eke/ in southern dialects.

I'd like to know what evidence this is.  There exists a localized High
Navarrese variant <eke>, which is quite mysterious.  But I know of no
such northern form as *<ekhe>.  (There *is* a northern word <ekhe> ~
<ekhei>, but it means `material for making something'.)

I must confess, though, that I would be pleased if the highly anomalous
<ke> could be shown to be a reduced form of an earlier *<eke>.  That
would remove the single most anomalous form from my list.  Don't see how
to do it, though, on present evidence.

> I assume that Hualde would list /kehe/ also.

No, I don't think so.  No such form as *<kehe> is recorded anywhere, and
nobody has proposed it as a reconstruction.

> So faced with these representations of the same word,
> how does one go about reconstructing the form?  Keeping in mind that
> the attested cases are /ekhe/, /khe/, /kehe/, /eke/ /ke/ and /kee/,

I can't agree, I'm afraid, unless you can cite some documentary evidence
for the reality of the ones I have queried.

> which one should be assigned the role of best representing the
> earlier form?

My choice is the virtually universal <ke>, since I ignore the aspiration
in citing forms (for my particular purposes here, I mean).  The variant
<eke> is too localized to be taken as more conservative, much as I might
like to do that.

> Or should none of them play that role? And was it
> originally monosyllabic or bisyllabic.

The evidence at present says monosyllabic.

> Finally, will the
> reconstruction of this form, i.e., the choices that are made, have
> any bearing on the way that we reconstruct /behe/ vs. /be/?

I don't think so.  An original *<behe> offers no difficulties at all.
The word for `smoke' is anomalous, that's all, with its initial /k/ and
its mysterious variant <eke>.  No other word in the language behaves
like this one, and I'm trying to work out the general rules, not to
account for isolated anomalies.

> Or stated differently, doesn't the set of choices we make about the
> reconstruction of the proto-form of /behe/ vs. /be/ bear on the way
> that we reconstruct the proto-form of the root-stem meaning "smoke"?

Not that I can see.

> Finally, aren't we caught in a dilemna when we admit at the onset
> that the mechanisms (formerly) governing the aspiration of b/t/k and
> /p/t/g/ in Euskera are poorly understood.

The *origin* of the aspiration is indeed poorly understood.  But that's
an entirely different issue from deciding whether forms with or without
the aspiration are more conservative.  We may safely reconstruct *<behe>
for `below' without bothering our heads about where that aspiration came
from originally.

Larry Trask
COGS
University of Sussex
Brighton BN1 9QH
UK

larryt at cogs.susx.ac.uk