Marked-nominative languages of Afroasiatic are the only ones in which defi-niteness played a crucial role. he impact of defidefi-niteness can be seen in the fact that only marked-nominative languages of Afroasiatic have split systems which work with definite nouns only, opposed to a non-case distinction with indefi-nite nouns. Some languages, such as Maale and Wolaitta, show a different case-marking system, one for definite nouns and one for indefinite nouns; again, definiteness is the factor which triggers these different encodings. No marked-nominative language of the Nilo-Saharan phylum has a case system which is sensitive to definiteness. Within Afroasiatic, all branches which have marked-nominative languages are involved, namely Berber, Cushitic, and Omotic.
Afroasiatic is the only phylum which has marked-nominative languages of type 2, in Nilo-Saharan no such type is found for marked-nominative lan-guages. As already suggested, this might be the result of a further feature which is unique to marked-nominative languages of the Afroasiatic phylum: Only here do languages show an amalgamated encoding of gender and case in the form of suffixes, vowel loss, or stress change.
has been extended to cover A and S. In this way, the new case pattern is born:
he old absolutive has changed into a new accusative and the old ergative into the new nominative. he accusative came into existence by the reduction of the old absolutive and the nominative by the extension of the old ergative (cf.
stage III in Table 16). herefore Plank uses the terms extended ergative versus restricted absolutive. Between stage I and stage III there is usually an inter-mediate stage where the language shows a split-S system (called active pattern by Plank 1985:273f.). In the intermediate stage (cf. stage II in Table 12), the behavior of S is not homogeneous. Some S behave still in accordance with the old pattern, S = O, other S behave already in accordance with the new pattern:
S = A. he result is a split-S system with SO = O and SA = A. he split-S system is semantically motivated. he extension of the ergative starts with intransitive agents (SA) and not with intransitive patients (SO). At stage III, all S behave homogeneously again, the change has been completed when SA & SO = A. he short notes about the languages which according to Plank (1985:274) are ex-amples for this development do not allow a final assessment. Not all of these languages show the full range of changes listed in Table 16. It seems that the development illustrated in Table 16 is a more idealized view. Some of them only show a development of ergative to split S, such as Kartvelian, Udi (Lez-gian, Northeast Caucasian), Sherpa (Tibeto-Burman), Akkadian (Afroasiatic), Wappo, Burushaski, and Japanese.
Among the languages listed by Plank with regard to this development, there are two which are of major concern here, as they both are at least to some ex-tent marked nominative: First Afroasiatic, Akkadian in particular, and second Wappo. Plank refers to Sasse (1982) and Diakonoff (1965) when claiming:
In Akkadian and perhaps further Afroasiatic languages a similar extension of ergative, or active, case marking, purportedly reconstructible for Proto-Afroasiatic, seems to have taken place (cf. Diakonoff 1965, Sasse 1982). (Plank 1985:274)
Plank’s hypothesis is to some degree in line with Aikhenvald’s conclusions about Berber (see Section 8c) in so far as the split-S system (called active by Plank) found in Proto-Berber could have been Proto-Afroasiatic. Aikhenvald however gives no hint about the possibility of having an ergative case system as the origin of all case traces within Proto-Afroasiatic.
(ii) Accusative to ergative. It is also possible that an accusative system devel-ops into an ergative system according to Plank (1985:279ff.). he rise of an ergative starts in transitive clauses with agents expressed in a more peripheral
relation, such as in passives. At stage I, the nominative encodes A and S, and the accusative O. At stage III the change is completed, when the nominative has been restricted to A and the accusative has been extended to O and S. he old nominative has become the new restricted nominative, which is the ergative, the old accusative has become the new extended accusative, which is the abso-lutive. Again, it is likely that there is a split-S system as an intermediate stage.
his comes into existence when some patient S’s already are aligned to transi-tive patients, O, while others still behave in the old way, being aligned with A.
Between stage I and stage III there is an intermediate stage where the change starts and that is the intransitive clause, such as passive clauses (other possi-bilities would be perfective-aspect or stative nominalized constructions). he new basic transitive construction emerges via the reinterpretation of an earlier existent intransitive pattern in which patients in a core relation and agents be-ing expressed by a peripheral relation. he original O of stage I is reinterpreted due to its semantics as S, and the original agent, expressed as peripheral, is reinterpreted as A. According to Li and Lang (1979,§4), Enga and other Pap-uan languages have developed an ergative system via the reinterpretation of the passive clause. Prior to this change there was a significant increase in the use of passive constructions. In the general case literature, a development from ergative to accusative, or from ergative to split S by way of intransitive clauses, however, is viewed to be rare or even not to exist at all (cf. Anderson 1977 or Dixon 1979:78, 101). hey doubt that such developments ever occurred. Dixon (1979:78 and 1980:338) mentions that three Australian languages with a split accusative-ergative system in the beginning have become ergative throughout.
he split was determined by the semantics of the core participants. he accu-sative, originally used primarily for O, with first and second person and some Table 6. he development from ergative to accusative case system according to Plank (1985).
Case system
Stage I A O ERG
S
ERG ABS
Intermediate stage II S1old = O
S2 new = A
Split S
Stage III S ACC
A O
extended ERG = new NOM
restricted ABS
= new ACC
human nouns, such as proper nouns, has been extended to all kinds of nouns, expressing S in intransitive clauses in Dhalandji, and to proper nouns express-ing S in Warluwarra, and in the Western Desert languages. he extension of the accusative, according to Dixon, was not gradually via intransitive clauses, but the accusative spread to agent-like S, such as proper names or human nouns.
Plank’s hypothesis with regard to the development from accusative and er-gative is in accordance with what Reh has claimed for Anywa (1996), where the ergative constituent order, not ergative case though, has emerged via the reinterpretation of a passive clause (see Reh 1996).
Table 7. he development from accusative to ergative case system according to Plank (1985).
Case system
Stage I A O ACC
S
NOM ACC
Intermediate stage II S1old = A
S2 new = O
Split S
Stage III A S ERG
O restricted NOM = new ERG
extended ABS
= new ABS
Both directions are possible: Nominative/accusative systems can develop into ergative languages and the other way round. Split S in both developments may occur as an intermediate stage. Plank assumes that change starts most likely in intransitive clauses. he development of case systems in two opposite direc-tions is not controversial in grammaticalization theory, as within grammatical-ization a concrete form in a concrete construction is studied and its grammati-calization path presented. Plank’s hypotheses are theoretically possible options which might occur but not with one and the same form in one and the same construction.
For African case languages in general and marked-nominative languages in particular, there is not enough historical data to make any statements about the historical origins further than the ones presented in Section 8. Plank’s sce-narios I and II allow for speculations about possible origins of marked-nomi-native languages: First, marked-nomimarked-nomi-native could be an intermediate stage of scenario I: he ergative is functionally extended to the new nominative, as it covers A and S. he markedness of the new nominative reflects stage I of the
ergative, as it is still the marked form within the case opposition. Second, the origin of a marked nominative within a marked-nominative system could be similar to the origin of the ergative as in scenario II. As has been shown in Sec-tion 8, within the Nilotic languages it is possible that the marked nominative goes back to a former agent participant of a passive clause. he same holds for the development of the ergative in scenario II above. Plank’s scenarios are therefore of relevance for marked-nominative languages as well, even if lack of appropriate data does not allow for any more concrete generalizations.