Personal Pronouns / Ergativity

[Larry Trask]

On Wed, 2 Jun 1999, Eduard Selleslagh wrote:

> In Basque (an agglutinating ergative language), both the ablative
> and the ergative case contain the -k ending, which also occurs in
> the nominative plural, in my view derived from a construction
> implying a kind of 'genitivus/ablativus partitivus' (cf. French 'des
> gens'; Lat. 'de' also had an 'ablative meaning'!).

With respect, I don't think it's possible to relate any two of these
three Basque inflectional endings.

A native and ancient Basque lexical item never ends in a plosive, except
in a few cases in which the final plosive is secondary.  In Basque of
the historical period, however, several inflectional suffixes end in, or
consist of, /t/ or /k/.  Some of these endings are clearly secondary,
like first-singular /-t/ from earlier */-da/. while others are of
unknown origin.  Possibly all are secondary.

Ergative /-k/ is universal and attested in the 9th century.  There is no
way of knowing if it derives from the reduction of something longer,
since the ergative can never be followed by any other suffix.

The ending /-tik/ is the most widespread ablative marker today, but it
is clearly recent.  In addition to /-tik/, we also find /-ti/ in earlier
Basque (still preserved today, I think, in some regions), /-ik/ in early
texts, and /-(r)ean/ in early Bizkaian.  In fact, all the local case
suffixes exhibit significant variation in time and space, and there
seems to me little doubt that the local case-endings are generally of
recent origin in Basque.

The common ablative /-tik/ may well result from a combination of /-ti/
with /-ik/, though there are problems with this.  In modern Basque,
/-ik/ is strictly the partitive ending, but its early attestation as an
ablative suggests that it may have originated as an ablative and then
become specialized as a partitive after the rise of other ablative
endings.

Complications: old Bizkaian, which has ablative /-(r)ean/, consistently
uses /-ti/ to mean `by way of', `via', which may therefore have been the
earlier function of /-ti/.  And both ablative /-tik/ and partitive /-ik/
have extended forms /-tika/ and /-ika/, respectively; these longer forms
may well be conservative.

The absolutive (not nominative) plural /-k/ is the most interesting of
all.  We find plural /-k/ only in the absolutive, which generally has
case-marker zero, while all the oblique cases exhibit an apparent plural
marker /-e-/.  One possibility is that the plural marker was originally
*/-g/, with devoicing to /-k/ in final position in the absolutive.  In
this view, the addition of a further overt case-suffix of any kind
triggered the automatic Basque insertion of /-e-/ to separate this
*/-g-/ from a following consonant, and then the medial */-g-/, being
invariably intervocalic, simply dropped (as is common in Basque),
leaving only the originally non-morphological /-e-/ as the apparent
marker of plurality in the oblique cases.

But note something curious: it is trivial to reconstruct an earlier
stage of Basque in which the plural marker /-k/ (or whatever it was)
occurs *nowhere* but in the three demonstratives.  Even in modern
Basque, this /-k/ occurs only in the three demonstratives, in the
ordinary (`definite') article, which itself plainly derives from a
reduction of the distal demonstrative, and (in some varieties)
pleonastically after the indefinite plural suffix /-zu/.  An example of
the last is provided by /bat/ `one, a' (< */bade/), plural /batzu/
`some, several', extended pleonastically to /batzuk/ in some varieties.
Plural /-k/ occurs nowhere else at all.

Not sure what all this means, but ablative /-tik/ and ablative/
partitive /-ik/ cannot possibly be identified with either ergative /-k/
or absolutive plural /-k/, and it seems most unlikely that these last
two can be identified with each other.

Larry Trask
COGS
University of Sussex
Brighton BN1 9QH
UK

larryt at cogs.susx.ac.uk