Summary for query on negation and case marking

[Matti Miestamo]

Dear Colleagues,

a week ago I posted a query on this list asking for pointers to 
languages where case marking (or the marking of nominal participants 
more generally) is affected by negation. The original query can be found 
at 
<http://listserv.linguistlist.org/cgi-bin/wa?A2=ind0910e&L=lingtyp&T=0&F=&S=&P=88>.

Altogether 12 people replied, either on-list or off-list. Many thanks 
for your replies! The following is a short summary of the responses.

Bernhard Wälchli wrote that they addressed this question briefly in 
Koptjevskaja-Tamm & Wälchli (2001: 633ff, 729; Circum Baltic languages). 
He also mentioned the situation in Basque and noted that the phenomenon 
occurs in a related way in older stages of Indo-European languages, 
notably in Germanic.

Larisa Leisiö pointed to the discussion of the phenomenon in Russian in 
her dissertation <http://acta.uta.fi/pdf/951-44-5029-9.pdf>.

Greville Corbett gave a couple of references on the situation in 
Slavonic: “First a bibliography, now rather dated, but it gives a sense 
of just how much has been done:
Greville Corbett. 1986. The use of the genitive or accusative for the 
direct object of negated verbs in Russian: a bibliography. In: R. D. 
Brecht and J. S. Levine (eds) Case in Slavic, Columbus, Ohio: Slavica, 
pp. 361–372.
And second, a more recent article:
Alexander Krasovitsky, Matthew Baerman, Dunstan Brown and Greville G. 
Corbett. Changing semantic factors in case selection: Russian evidence 
from the last two centuries. To appear in: Morphology.”

Paul Hopper pointed to the discussion in the article "Transitivity in 
grammar and discourse" (Paul Hopper & Sandra Thompson, Language 1980).

Ljuba Veselinova wrote that in Oceanic languages the NP in negated 
constructions regularly changes over to the indefinite article or in 
some cases to the non-specific article. She further noted that such 
effects on NPs in negated clauses, either by case marking, or article 
change, or something else actually seem to be very common. She added 
that she has mainly looked at existentials and predicative possession. 
Her complete on-list reply can be read at 
<http://listserv.linguistlist.org/cgi-bin/wa?A2=ind0910e&L=lingtyp&T=0&F=&S=&P=871>.

Claire Moyse-Faurie clarified that what Ljuba had pointed out is indeed 
common in Polynesian languages,  but not in all Oceanic languages. She 
added the following reference:
Hovdhaugen Even and Ulrike Mosel (eds), 1999. Negation in Oceanic 
languages. Lincom Studies in Austronesian Linguistics 2.

Alex François mentioned the Oceanic language Araki (Vanuatu), where 
affirmative sentences in Realis mood use a bare NP for the object and 
then the verb must bear some transitive morphology.  But negative 
sentences have to use a Partitive marker /re/, which must be understood 
as a marker for Non-Referential (=non-specific indefinite). The verb 
form used in the affirmative would be compatible with the negative only 
with a referential/definite reading. The Indefinite Specific marker is 
incompatible with negation, and conversely, the Partitive /re/ is 
incompatible with an affirmative realis sentence. In Affirmative 
Irrealis the two constructions are possible, but  differ semantically as 
to the referentiality (= specificity) of the object. Alex’s reply 
included a large number of examples and a summarizing table which I’m 
not including in this short summary. He also mentioned the following 
references:
François, Alexandre (2002). Araki: A disappearing language of Vanuatu. 
Pacific Linguistics, 522. Canberra: Australian National University. 375 pp.
François, Alexandre (2002). ‘I need somebody, not just anybody’: The 
grammar of referentiality in Araki.  Paper presented at the 5th 
International Conference of Oceanic Languages (COOL5), Canberra,  Jan 2002.

John Stewart brought up the related phenomenon in Arabic, where negative 
existential sentences use a special negative verb /laysa/ that requires 
the nominal predicate to take accusative case (instead of the nominative 
used in the affirmative). He also noted that /laysa/ is not the only 
element that can have this effect on phrases in the predicate. His 
complete reply is found in 
<http://listserv.linguistlist.org/cgi-bin/wa?A2=ind0910e&L=lingtyp&T=0&F=&S=&P=1341>

Wolfgang Schultze followed up on this and added that the selection of 
the accusative case is not confined to /laysa/ but many other elements 
have a similar effect, see the complete reply  at 
<http://listserv.linguistlist.org/cgi-bin/wa?A2=ind0910e&L=lingtyp&T=0&F=&S=&P=2269>. 
He further noted that the selection of the accusative has to do with the 
original Semitic case system more generally, not with negation as such, 
and pointed to the following reference: Michael Waltisberg 2002. Zur 
Ergativitätshypothese im Semitischen. ZDMG 152,11-62.

Bernard Comrie pointed out that a number of Bantu languages have the 
shorter form of the noun class prefix within the scope of negation (and 
some other environments).

Larry Hyman mentioned the Bantu language Aghem, where the object of a 
negative sentence is treated like an oblique, and attached a forthcoming 
paper where the issue is treated in more detail. He added that “there 
are quite a few African languages where negation is treated as 
"inherently focused", thereby having either a morphological, syntactic 
or prosodic effect that is different from the corresponding affirmative.”

Eduardo Ribeiro wrote that in Northern Jê languages (Central Brazil), 
negation triggers an ergative pattern in transitive constructions, and 
the verb is nominalized. The ergative pattern is linked to the limited 
argument structure of deverbal nouns, and is triggered by any situation 
that requires a nominal form of the verb. His complete reply, which 
includes links to several downloadable grammatical descriptions of 
Northern Jê languages, is at 
<http://listserv.linguistlist.org/cgi-bin/wa?A2=ind0911a&L=lingtyp&T=0&F=&S=&P=735>. 


Once again, many thanks to all those who responded!

Best wishes,
Matti

--
Matti Miestamo
<http://www.ling.helsinki.fi/~matmies/>