Query: Adverbs and the AEC

[Wolfgang Schulze]

Dear Typologists,

it is a well-known fact that there is a strong correlation between
accusative case marking and adverbial behavior in so-called accusative
structures (I remember that some times ago there was a discussion on
accusatively marked space/time concepts on this list). Obviously, this
has to do with at least a) the marked role that the 'objective' (O)
plays in an accusative strategy (S=A;O in terms of the
Accusative-Ergative Continuum (AEC)) and b) the fact that accusative
case marking is often co-paradigmatized with locatives (mainly
allatives) [there are additional explanations that come from Cognitive
Typology, but I'll skip this here for sake of brevity]. My question now
is, whether there is an analogic correlation in ergative behavior
(S=O;A). If we claim that the marked part of the S=O;A structure is 'A'
in an ergative behavior, we would expect to find (derived) adverbs that
show an ergative case marking (if present). In some cases this seems to
be true especially with adverbs of manor (< instrumental), cf. Udi

(1)    she-t'-in ich yoldash-ax zor-en-ne tap'-i
        he-SA-ERG REFL friend-DAT2 force-INSTR-3SG hit-AOR
        'He strongly hit his friend

But note that Udi (East Caucasian) already shows strong tendencies
towards an accusative behavior. In Burushaski, for instance, adverb-like
elements sometimes appear in the absolutive, cf.

(2)    tham-e ine hir waziir e-tim-i
        tham-ERG PROX man Wezir:ABS 3SGm:O-make:PAST-3SGm:A
        'The Tham appointed this man as a wezir.'

Secondary adverbs in Burushaski are often derived from locatives or
represent the grammaticalization of former constructions with gerunds,
but there is a certain class of local adverbs that use a morpheme -e,
itself a functionally complex morpheme that can best be described as an
genitivus-ergativus-locativus (cf. hol-e 'outside', kaat.-e 'together').
Naturally, this is not enough evidence to propose that the marked
portions of S=A;O and S=O;A, namely O (accusative) and A (ergative) are
a 'natural' source for secondary adverbs, also because Burushaski offers
both options. So my question is whether there is any systematic
correspondency between ergative case marking and secondary adverbs to be
observed, or whether there is a preference for absolutive 'marked'
adverbs as in the following Udi example:

(3)    she-n-o xib ghi tac-i-ne
        'he-SA-ABS three day go:PAST-PAST-3SG
        'He went three days...'

Can we claim that

S=A;O -> ADV ~ O(ACC)
S=O;A -> ADV ~ A(ERG) or O(ABS) ?

Note that here I disregard other adverbial forms stemming for instance
from locatives etc....

Wolfgang
*****************************
Prof. Dr. Wolfgang Schulze
Institut fuer Allgemeine und Indogermanische Sprachwissenschaft
Ludwig-Maximilians-Universitaet München
Geschwister-Scholl-Platz 1
D-80539 München
Tel.: +89-21805343 / Fax: +89-21805345
Email: W.Schulze at lrz.uni-muenchen.de
http://www.lrz-muenchen.de/~wschulze/
*****************************