As defined in 1.1, in type-2 languages all case forms used to encode the core participants A, S, and O are derived forms. here is no morphologically un-marked, that is zero, form. In marked-nominative languages both the accusa-tive and nominaaccusa-tive are derived forms in a type-2 language as opposed to a
type-1 language, where only the nominative is derived whereas the accusative is the basic form.
Only a minority of the marked-nominative languages belong to type 2, namely 16 of the 54 languages studied. Out of these 16, two are only partly type 2, in that some nouns in the language follow type 1 and others type 2.
here is a concentration of type-2 languages within Highland East Cushitic and North Omotic.
Sidamo (East-Highland, Cushitic, Afroasiatic)
Tucker and Bryan (1966) list the following case endings in Sidamo: Accusa-tive is zero, nominaAccusa-tive either suffix -i or -u, geniAccusa-tive also either suffix -i or -u.
Among the determiners, a suffix distinguishes two cases following type 2, cf.
Table 6.
Table 6. Case inflexion on determiners in Sidamo (Tucker and Bryan 1966:514 &
517).
ACC M -ha NOM M -hu
F -ta (also -te) F -tu (also -te)
One could argue that the determiners follow type 2.
K’abeena (East-Highland, Cushitic, Afroasiatic)
According to Crass (2003) seven different cases are distinguished in K’abeena.
Case marking follows complex rules. Nouns fall into thirteen different classes.
Case encoding is sensitive to gender, number, and the number of morphemes of the word stem. Two genders are distinguished, feminine and masculine.
Gender and case are shown by the final vowel. he accusative is encoded by different suffixes, depending on gender. he nominative is expressed either by vowel shit, pitch accent shit, or zero. Table 7 gives an overview of a few case endings (raised symbols stand for voiceless vowels). Morphologically, the cases are divided into two subclasses: Case marking in the accusative, nominative and genitive are all derived directly from the wordstem. he dative, instru-ment, locative, and ablative are derived from either the accusative or the geni-tive form; the latter two are case doubling.
Table 7. Case marking in K’abeena (Crass 2003).
Noun Accusative Nominative Genitive Dative Instrument Comitative Locative
Ablative
246M a, aa, ee, i, o, oo, u
i, u, oo ee, i, o VVha VVni VVcci
2 F e, o i,u e, o VVta VVni VVcci
4 F ata, aata, eeta, ita, uta
ati, aati, eeti, iti, uti
a, aa, ee, e, o VVta VVni VVcci
Functionally, the nominative covers S & A (as in 58) and the accusative is used in citation, covering O (as in 58), nominal predicates (cf. 59), and S in imper-sonal constructions (cf. 60). Note that case and gender endings in the following examples are not separated from the stem because of their interwovenness.
(58) faangoo lalu ÁaaÁitoÁ.
thief.NOM cow.ACC take.PFV.3SG.F/3.PL
hiefs have stolen the cow/the cows. (Crass 2003) (59) Áisa haakime-ha.
3SM doctor.ACC-COP.M He is a doctor. (Crass 2003)
(60) Áeedisi zabbooni hureennu-baÁ.
aids.ACC medicine.INST heal.IPV.IMP-NEG Aids is not curable with medicine. (Crass 2003)
Nominal modifiers show a defective case inflexion. As Table 8 illustrates, demonstratives occur in two different forms only, one being morphologi-cally marked and the other morphologimorphologi-cally unmarked. he morphologimorphologi-cally marked form is used when the head noun occurs in the nominative. In all other environments the morphologically unmarked form is used. One could argue that the morphologically marked form corresponds to the nominative (cf. 61c) and the morphologically unmarked form to the accusative (cf. 61a). he latter has the status of a default form. In 61b the accusative form is used with its head being in the genitive.
Table 8. Determiners in K’abeena (Crass 2003).
Morphological
status Demonstrative
determiner Head noun
derived, e.g. ti NOM NOM
zero marked, e.g. ta ACC ACC, DAT, GEN, ABL, LOC, INST
= all other environments
(61) a. ta kÁankÁuta hiÁriyoon-se. DEM.F.ACC medicine.for.tapeworm.F.ACC buy.PFV.1S-3SF his medicine for tapeworm, I have bought (it). (Crass 2003) b. ta cÁaa eÁiÁliti hulbuhulbuyitaaÁ.
DEM.F.ACC girl.GEN bottom.NOM constantly.waggle.with.one’s.
bottom.IPV.3SF.
he bottom of this girl is constantly waggling. (Crass 2003) c. ti Áadancuti zumaa Áadafaffi Áaffi
DEM.F.NOM cat.NOM rat.ACC lie.in.ambush.CON.3SF catch.CON.3SF ittaÁ.
eat.IPV.3SF
his cat lies in wait for rats, catches and eats them.
As Table 9 illustrates, adjectives occur in up to three different forms: One is morphologically unmarked and two are morphologically marked. he mor-phologically unmarked form modifies head nouns in the accusative only.
One of the morphologically marked forms (called ‘marked form 2’ by Crass) modifies masculine nominative heads only, while in all other environments the marked form 1 is used.
Table 9. Adjectives in K’abeena (according to Crass 2003).
Morphological status
Adjectives Head noun Example
roora ‘big, much’
derived (called marked form 2)
NOM.M NOM.M rooru
zero marked ACC ACC roora
derived (called
marked form 1) NOM.F, DAT, GEN,
ABL, LOC, INST NOM.F, DAT, GEN, ABL, LOC, INST
= all other environments roora
he case system suggests the following generalizations: If K’abeena has a mor-phologically unmarked form then it is the accusative (see determiners of Ta-ble 8). here are traces where the accusative is functionally the most unmarked:
First, the accusative of the determiner is used as the default case. Second, the accusative is the basis for the ablative case. hird, the accusative has the widest range of applications: It encodes O, nominal predicates, and is used as the cita-tion form. But the accusative is not the only case which is a candidate for being used in a wider range of functions. As mentioned earlier, the genitive is also used as the basis for case doubling.
he fact that gender and case marking are interwoven might have led to a situation where no case form in K’abeena is morphologically unmarked.
Alaaba (Highland-East, Cushitic, Afroasiatic)
Alaaba, another Highland East Cushitic language, shows a similar behavior as K’abeena with regard to case. Eight cases are distinguished. It also is a marked-nominative language of type 2, meaning that there is no morphologically un-marked case form in Alaaba. Both cases, the nominative and the accusative, are morphologically derived forms. In Alaaba as well, case marking is inverwo-ven with gender marking. Again there are two genders, namely masculine and feminine. Case is encoded by accent shit, vowel devoicing, and suffixes. he accusative is used as the citation form (cf. 62), it encodes O (cf. 63), nominal predicates (cf. 65), and, in contrast to K’abeena, also time participants (cf. 66).
he nominative encodes S (cf. 64) and A (cf. 63). As in K’abeena, cases are divided into two subclasses: he first class consists of the accusative, the nomi-native and the genitive, which are all derived from the word stem. All other cases belong to the second class, such as dative, ablative, locative, instrument, and similative. hey are derived from the accusative. he latter are instances of case doubling.
(62) Alaaba miní house.ACC
‘house’ (Schneider-Blum 2003)
(63) mánc(u) ka elóo albaaÁl-isée A O V
man.S.M.NOM DEM1ACC hole.ACC be.wide-CAUS-3S.M.PFV2 he man enlarged this hole. (Schneider Blum 2003)
(64) hiku mánc(u) orróoÁ-y(o) S V
DEM.2.NOM man.S.M.NOM leave-3SG.M.PFV hat man let (Schneider-Blum 2003)
(65) tiin(i) mancootáan(i) S N.PRED
DEM.1F.S.NOM woman.S.F.ACC.CL.S.F.L/I his is a woman. (Schneider-Blum 2003)
(66) sasíic(i) shoolú iill(a) ameet-áam(i) Time S three.ABL four.ACC until come-1S.IPV
I’ll come until three to four (o’clock). (Schneider-Blum 2003)
Nominal modifiers always show a reduced pattern of case inflexions. One set of dependent demonstratives occurs in three case forms only, namely accusa-tive, nominaaccusa-tive, and one further form, which is used for all remaining cases.
he first person feminine singular demonstrative meaning ‘near’ belongs to this set. It occurs with the forms ta for accusative, ti for nominative, and tan elsewhere.
Even more than in K’abeena it is the accusative the case which is function-ally the least marked. his can be seen in the functional range covered by the accusative: In addition to the typical functions O, nominal predicate and the citation form, it also covers a peripheral participant, namely time.
he genetically closed related language Kambaata, also Highland East Cushitic, shows a similar behavior with regard to case as Alaaba and K’abeena.
Again, case marking is interwoven with gender marking in Kambaata. Here again, nominative and accusative are both derived forms (Yvonne Treis, p.c.).
In a similar fashion, Libido47 (East Cushitic, East Highland) belongs to type 2 as well, also encoding case by means of suffixes (Joachim Crass, p.c.).
Haro (East Ometo, Omotic, Afroasiatic)
Haro, an Omotic language, is a marked-nominative language according to Woldemariam (2003). Case is encoded by suffixes. In total, ten cases are dis-tinguished. he core cases nominative, accusative, and the genitive are only distinguished with definite nouns. Indefinite nouns show no case distinction with these three cases. All remaining cases are distinguished with definite and indefinite nouns.
he accusative covers the direct object, O (cf. 67a). As indefinite forms show no case distinction, the citation form is not the accusative but the in-definite form. Nominal predicates, when in-definite, are encoded in the accusa-tive (cf. 67c). In this respect, Haro behaves like a typical marked-nominaaccusa-tive language.
hree cases are only distinguished with definite nouns by suffixes. hese are the nominative, marked by the suffix -i, the accusative, marked either by the suffix -a (masculine, plural), or -o (feminine, singular) (called absolutive by Woldemariam) and the genitive suffix -i. Seven further cases, which all en-code peripheral participants, are always used irrespective of whether the noun is definite or indefinite. he latter are dative, instrumental, locative, ablative, directive, comitative, and vocative.
(67) a. Haro48
‘ís-í garmá-z-a ‘í-wo1-ín-e
she-NOM lion-M.DEF-M.ACC 3FS-kill-PAST-A.DEC She killed the lion.
b. ‘ís-í garmá-t-o ‘í-wo1-ín-e she-NOM lion-F.DEF-F.ACC 3FS-kill-PAST-A.DEC She killed the lioness. (Woldemariam 2003:64) c. yé‘-í garma-z-á-kko
that-NOM lion-M.DEF-M.ACC-FOC hat is the lion. (Woldemariam 2003:64)
d. gaarmá-z-í deyššá-z-a-kko ‘é-m-ín-e
lion-M.DEF-NOM goat-M.DEF-M.ACC-FOC 3MS-eat-PAST-A.DEC he lion ate the goat. (Woldemariam 2003:65)
e. šaató-z-i maačč-i1é-z-a ‘é čÁaš-ín-e
boy-M.DEF-NOM woman-PL-M.DEF-M.ACC 3MS insult-PAST-A.DEC he boy has insulted the women.
he nominative shows no sensitivity to gender. It is always marked by the suf-fix -i (cf. 67b & d). All plural nouns take the masculine accusative suffix -a, even if feminine (cf. 67e). he nominative and the genitive are encoded by the same suffix, namely -i (cf. Woldemariam 2003:67). Woldemariam considers this similarity to be accidental:
he Nominative and Genitive cases involve the same morphological marking as against the Absolutive. he formal similarity between the Nominative and Genitive is considered as a mere historical coincidence or homonymy since functionally there may not be much that can be said about the closeness be-tween the Nominative and the Genitive. (Woldemariam 2003:61)
here is case doubling in Haro. All seven peripheral cases (listed above) are added to the noun bearing a case suffix -i. Considering the homophony of gen-itive and nominative, the question arises whether this suffux -i is the nomina-tive or the geninomina-tive. According to Woldemariam it has the status of the geninomina-tive and not of the nominative (cf. Woldemariam 2003:61). Her main argument is based on genetic comparison: In genetically related languages, such as K’abeena (Crass 2003) and Wolaitta (Amha forthc.), the genitive is also used as the basis for case doubling; e.g. the noun ‘lion’ appears as garmá when indefinite for S, A, O and as the citation form. he accusative masculine definite form of ‘lion’
is garm-z-a (cf. 67a), with -z- encoding masculine definite, and -a masculine accusative. he corresponding feminine form is garmá-t-o with -t being femi-nine definite, and -o femifemi-nine accusative (cf. 67a & 67b). Nevertheless, as also
mentioned by Woldemariam (loc. cit.), there are other Ometo languages which take the accusative as the basis for case doubling, such as Maale (Amha 2001, see above) and Zayse (Hayward 1990).
Unlike type-1 languages, the morphologically unmarked form in Haro presents the indefinite form of the noun, which is ‘caseless’. he latter form is used as the equivalent of the definite case forms nominative, accusative, and genitive. Definiteness and case are both marked by separate suffixes which fol-low the noun stem. Nominative, accusative and genitive case forms are built on the following pattern:
Noun – DEF – CASE
All remaining cases involve case doubling: Case is suffixed to the noun in the genitive case form. All nouns bear case irrespective of whether definite or in-definite. Case doubling takes the following form:
Noun – (DEF) – GEN – CASE
he encoding pattern type 2 came into existence in Haro in a slightly different way than in the languages of this type discussed so far. Like in the other type-2 languages, case and gender are interwoven as well in Haro. With definite nouns gender is marked twice on each noun: he definite suffixes and the case suffixes are both gender-sensitive. Nevertheless, unlike in many other type-2 languag-es, this interwovenness does not lead to an encoding pattern following type 2 in Haro. Type 2 in Haro has been triggered by a different source: Haro has a split marked-nominative system, which is neutralized with indefinite nouns.
he fact that with indefinite nouns case is neutralized results in the situation that the morphologically unmarked form is already employed for indefinite nouns.
Beja (North Cushitic, Afroasiatic)
he data for Beja are not entirely clear. It seems that Beja has a defective marked-nominative system, following type 2 with regard to nouns and type 1 with regard to determiners. Case is hardly seen on nouns, case and gender are interwoven (see Table 10). Instead, case is seen with a preceding qualifier, which itself is case-inflected (see Table 11). As Table 10 illustrates, case is de-fective in the feminine gender: here is no distinction between nominative and accusative.
Table 0. Case inflexion in Beja49 on nouns (Tucker and Bryan 1966:514).
NOM M zero
F -t
ACC M -b ater a vowel, otherwise zero
F -t
GEN M -i ater a consonant, -y ater a vowel
F -ti
With a preceding qualifier, the noun has to appear in the morphologically un-marked form, showing no gender or case ending. As Table 11 illustrates, the accusative is the morphologically unmarked form. It also is functionally the unmarked form as it covers accusative and genitive.
Table . Case inflexion in Beja with determiners (Tucker and Bryan 1966:516).
Determinative Noun
NOM. (ABS) M (w)uu- Áu:- PL (y)aa- Áa:-
zero suffix (absolute form)
F tuu- tu:- taa-
ta:-ACC M (w)oo- Áo:- (y)ee
Áe:-F too- to:- tee-
te:-GEN M (w)oo- Áo:- (y)ee Áe:- -i or -y PL -e
Genitive form
F too- to:- tee- te:- -ti PL -te
Functionally, the case system follows a marked-nominative pattern. According to Reinisch the accusative is used as the citation form (see Reinisch 1893:4450), it encodes nominal predicates (see Reinisch 1893:8351), O and IO (Reinisch 1893:6752). Beja is a split marked-nominative language of type S2. It either fol-lows a marked-nominative pattern or shows no distinction at all. On nouns, the split is triggered by gender and phonology, only some masculine nouns show a case distinction between nominative and accusative.
Split-type 2 languages
In addition, languages belonging partly to type 2, namely Arbore, Maale and Wolaitta. As mentioned above, in Arbore only few nouns have a basic form which differs tonally from the accusative. With these nouns, the accusative and the nominative are morphologically derived forms. As mentioned above for Maale, all definite nouns belong to type 2, in that accusative and nominative are both encoded by suffixes. Indefinite nouns however all behave according to type 1.
In Wolaitta (Ometo, North Omotic), indefinite nouns follow type 1 with a morphologically unmarked accusative and a morphologically marked nomi-native, either encoded by vowel length, suffix or high tone. here are a few exceptional nouns ending on a vowel -e, where the nominative is also zero marked. Definite nouns however follow type 2 in that both case forms nomi-native and accusative are morphologically marked forms. Case is marked by portmanteau morphemes which encode gender, definiteness and case. Gender is only encoded with definite nouns (Azeb Amha, p.c.).