Heine Bernd, Nurse Derek. A Linguistic Geography of Africa

Подождите немного. Документ загружается.

c. Wolane

ksa

slı

-llba

lna

-n

afternoon.snack PURP-eat.IPF.1PL-COP.3MSG

‘We intend to eat our afternoon snack’

Remarkable is the situation in Oromo and Libido. As mentioned earlier, both

languages do not distinguish morphologically between intentional and pro-

spective. Oromo expresses both categories by a construction consisting of a

verbal noun in the dative case followed by the copula.

(12) Oromo

a. prospective

nuyi barumsakeen

at’umuruu-f-i

1PL.NOM study.POSS.1PL ﬁnish.VN-DAT-COP

‘We are about to ﬁnish our studies’

b. intentional

nuyi laak’ana n

aaccuu-f-i

1PL.NOM lunch eat.VN-DAT-COP

‘We intend to eat lunch’

Furthermore, Oromo has an additional construction to express the prospective.

This construction contains a noun with possessive sufﬁxes.

(13) Oromo

ani barumsa t’umuruu koo-ti

1SG.NOM study ﬁnish.VN POSS.1SG-COP

‘I am about to ﬁnish my studies’

This construction is identical to the prospective construction in the Ethio-

Semitic languages. Therefore, it may have developed in Oromo due to contact

with these languages. Libido expresses intentional and prospective by a copula

construction with a converb in the predicate position. This strategy is identical

to the intentional const ruction in K’abeena.

(14) Libido

a. prospective

losano guullena-tte

study.ACC ﬁnish.CNV.1SG-COP

‘I am about to ﬁnish my studies’

b. intentional

hurbaata ’intena-tte

dinner.ACC eat.CNV.1PL-COP

‘We intend to eat lunch’

In conclusion, the languages which have a morphologic al distinction between

intentional and prospective use a subordinate verb to express the intentional,

and a modiﬁed noun to express the prospective. Oromo and Libido, however,

choose one out of the two strategies – either a subordinate verb (Libido) or a

case-marked verbal noun (Oromo) – to express both categories.

Joachim Crass and Ronny Meyer242

7.3.2.2 Benefactive focus All languages have a construction consisting

of a predicate involving a converb of the r espective verb ‘to say’ followed

by a copula. The verb ‘to say’ has two internal arguments, namely a noun

phraseinthedativecaseandacomplementclause.Thedativenoun

phrase refers to a n emphasized beneﬁciary of the event described by the

complement clause.

(15) a. Amharic

ne ya

mmı

tagga

-w l-anta

biyye

1SG REL.struggle.IPF.1SG-DEF.M DAT-2MSG say.CNV.1SG

COP.3MSG

b. K’abeena

’ani ’app’amaammiihu kesaa-ni-ma

1SG.NOM struggle.IPF.1SG .NOM 2SG.DAT-too-after.all

yiyee-ti

say.CNV.1SG-COP

c. Libido

’ani gubamaammooki keesa

1SG.NOM struggle.IPF.1SG .NOM 2SG.DAT

yee-tte

say.CNV.1SG-COP

‘That I am struggling is for the sake of you’

7.3.2.3 Different copulas in main and subordinate clauses All languages

of our sample have different copula constructi ons in main and subor dinate

clauses. A copula with agreement markers for the subject occurs in the main

predication of all investigated Ethio-Semitic languages except Zay. In the

Cushitic languages, however, agreement is more restricted.

(16) a. Amharic

ssu asta

mari na

3MSG teacher COP.3MSG

b. Guma

hut asta

mari-w

3MSG teacher-COP.3MSG

c. K’abeena

’isu rosisaanco-ha

3MSG.NOM teacher.ACC-COP.M

‘He is a teacher’

In subordinate clauses, a fully inﬂected verb with the meaning ‘to live, to

become’ is used in the function of a copula. This holds true for all languages,

Ethiopia 243

i.e. also for Zay. The respective verb occurs in the perfective aspect but refers

to a present or future event.

(17) a. Amharic

t’ı

ru dan

aya

-hona

w ...

good judge REL-be.PFV.3MSG person

b. Guma

he dan

aya

-ha

b ...

good judge REL-be.PFV.3MSG man

c. K’abeena

maat’aaree ‘ikkoo mannu ...

wise.ACC be.PFV.3MSG.REL person.NOM

‘A person who is wise / a good judge ...’

7.3.2.4 Experiential perfect According to Comrie (1976: 58), the

‘‘experiential perfect indicates that a given situation has held at least once

during some time in the past leading up to the present.’’ In all investigated

languages this category is expressed by a construc tion which consists of the

verb ‘to know’ in the main clause and its com plement expressed by a converb

clause. The event, which was experienced by the subject, is encoded in the

converb clause.

(18) a. K’abeena

’ameerikaani ’orooteeni kasseenta-’i?

America.LOC go.CNV.2PL know.PRV.2PL-Q

‘Have you ever been to America?’

’ee, ’ameerikaani ’oroo’ni kansoommi

yes America.LOC go.CNV.1PL know.PFV.1PL

‘Yes, we have been in America’

b. Oromo

ameerikaa deemtan-i beektuu?

America go.PFV.2PL-CNV know.IPF.2PL.Q

‘Have you ever been in America?’

eeyee, ameerikaa deemnee beekna

yes America go.PFV.1PL.CNV know.IPF.1PL

‘Yes, we have been in America’

c. Wolane

’amarikan hedkum-ani tı

cluw-ahum?

America go.PFV.2PL-CNV know.IPF.2PL-AUX.2PL

‘Have you ever been in America?’

’aw, ’amarikan hedı

-na

yı

clı

nan

yes America go.PFV-1PL.CNV know.IPF.1PL.AUX.3MSG

‘Yes, we have been in America’

Joachim Crass and Ronny Meyer244

7.3.2.5 Exist > ‘have’ possessive > obligation The grammaticalization of

the verb ‘to exist’ to a marker of possessive ‘have’ constructions (Heine &

Kuteva 2002: 127f.) can be linked with the grammaticalization of possessive

‘have’ constructions to a marker of obligation (Heine & Kuteva 2002: 243f.) to

the chain exist > ‘have’ possessive > obligation. This grammaticalization

chain is attested in all languages except Libido and Oromo.

‘Have’ possession is expressed by sufﬁxing object agreement markers to

the respective verb ‘to exist’ as exempliﬁed in the sentences (19b) to

(21b). In order to express obligation, an additional morpheme is sufﬁxed

to the verb in Ethio-Semitic languages. In K’abeena, however, there is no

morphological difference between possession and obligation. The two

categories are distinguished only by the word class of the subject. If it is a

noun, possession is expressed, if it is a verbal noun, obligation is

expressed.

(19) Amharic

a. Existence

ahun wuha alla

now water exist.PFV.3MSG

‘Now there is water ’

b. Possession

yanta

yahı

llı

jj alla

-n

GEN.2SG size child exist.PFV.3MSG-1SG.O

‘I have a child of the same age/size as yours’

c. Obligation

hed alla

-bbı

-n

go-VN exist.PFV.3MSG-OBL-1SG.O

‘I have to go’

(20) Zay

a. Existence

ahu ma

y ala

-n-u

now water exist.PFV.3MSG.-FC-DC

‘Now there is water’

b. Possession

haata

-naah ba

wu ala

-n

GEN.2MSG-FC.size child exist.PFV.3MSG-1SG.O

‘I have a child of the sam e age/size as yours’

c. Obligation

sha

gga

r woheedaat-ı

n ala

-bi-n

Addis Ababa go.VN-FC exist.PFV.3MSG-OBL-1SG.DC

‘I have to go to Addis Ababa’

Ethiopia 245

(21) K’abeena

a. Existence

teesu wuu yoo

now water.NOM exist.PFV.3SG/PL

‘Now there is water’

b. Possession

kii bikku c’uulu yoo-’e

2S.GEN size.NOM child.NOM exist.PFV.3SG/PL-1SG.O

‘I have a child of the same age/size as yours’

c. Obligation

’oro’-u yoo-’e

go-VN.NOM exist.PFV.3SG/PL-1SG.O

‘I have to go’

In Libido, only the grammaticalization from the verb ‘to exist’ to ‘have’ pos-

session is attested. Obligation is expressed by the verb ‘be necessary.’ Further-

more, Libido does not possess object agreement sufﬁxes on the verb. Therefore,

the possessor is marked by a personal pronoun in the dative case (see (22b)).

(22) Libido

a. Existence

kaaru wa’i yooko

now water.NOM exist.PFV.3SG

‘Now there is water’

b. Possession

kaa k’at’i beeti ’eessa yooko

2SG.GEN size.NOM child.NOM 1SG.DAT exist.PFV.3MSG

‘I have a child of the same age/size as yours’

c. Obligation

’ani ba’immi hasisaako

1SG.NOM go-VN.NOM be.necessary.PFV.3MSG

‘I have to go’

In Oromo the situation is completely different. Neither of the grammaticali-

zations is attested. ‘Have’ possession is expressed by a verb with the meaning

‘to have’:

(23) Oromo

a. Existence

bishaan-ni hamma jir-a

water-NOM now exist-IPF.3MSG

‘Now there is water’

b. Possession

ijjoollee amma-kee-n qab-a

child as.much.a s-POSS.2SG-FC.1S have-IPF.1SG

‘I have a child of the same age/size as yours’

Joachim Crass and Ronny Meyer246

Obligation is either expressed by the verb ‘to have’ or by a construction

involving the verb ‘to exist’ and a personal pronoun in the ablative case.

Neither of the construc tions is commonly accepted.

(24) Oromo

Obligation

a. ? deemuu-n qab-a.

go.VN-FC.1SG have-IPF.1SG

b. ? deemuu-n na-irra jir-a

go.VN-NOM 1SG.O-ABL exist-IPF.3MSG

‘I have to go’

More often the simple imperfective is used which may have the semantic

implication of obligation.

(25) Oromo

Obligation

n-an deem-a

FC-FC.1SG go-IPV.1SG

‘I will go’

7.3.2.6 Past > apodosis of an irrealis conditional clause Afurther

grammaticalization process involves the past marker. Beside its function to

express tense, it is used to mark irrealis conditional clauses. In all investigated

language the past marker occurs in the apodosis of an irrealis conditional clause.

(26) Guma

a. Past

b-abba-na

bet k’e

LOC-father-POSS.1SG house wait.IMP.2SG.1SG.O

bahu-m ambwa

r-hu bana

say.PFV.1SG-CNV NEG.go.PFV-1SG AUX.PAST.3MSG

‘Didn’t I leave, saying to you: ‘‘Wait in my father’s house!’’?’

b. Irrealis condition

trama

na zı

rab tanza

yesterday rain SUB.NEG.rain.PFV.3MSG

hı

nnı

dı

rgna

bana

corn thresh.JUS.1PL AUX.PAST.3MSG

‘If it had not rained yesterday we would have threshed corn’

(27) Oromo

a. Past

mana abbaa kooti-tti na

house father.GEN POSS.1SG.GEN-LOC 1SG.O

Ethiopia 247

dheegi siin jedhee-n

wait.IMP.2SG 2SG.O.INST say.PFV.1SG.CNV-FC.1SG

turee?

AUX.PAST.1SG.Q

‘Didn’t I tell you to wait for me in my father’s house?’

b. Irrealis condition

osso kaleessa bokkaa-n hin-roobne

if yesterday rain-NOM NEG-rain.PFV

sila t’aaﬁi tuma-na turre

IRR tef thresh-IPF.1PL AUX.PAST.1PL

‘If it had not rained yesterday we would have threshed tef’

(28) K’abeena

a. Past

’anni»i bokkooni ’agar’e

father.GEN house.LOC 1SG.O.wait.IMP.2SG

yiyehe ’oro’yoommi-kk’ i-ba-indo?

say.CNV.1SG.2SG.OBJ go.PFV.1SG-AUX.PAST-NEG-Q

‘Didn’t I leave, saying to you: ‘‘Wait in my father’s house!’’?’

b. Irrealis condition

bereta t’eenoo ’ubboba’ikkaani

yesterday rain.NOM rain.PFV.3MSG.NEG.SUB

t’aafaa ’udunnaammi-kk’i

tef.ACC thresh.IPF.1PL-AUX.PAST

‘If it had not rained yesterday we would have threshed tef’

When no adverb indicates tense, a past or a non-past interpretation of the

sentence is possible. The sentences in (i) are irrealis conditional clauses, the

sentences in (ii) hypothetical conditional clauses.

(29) a. Amharic

zı

nab bayza

nb t’ı

ru na

bba

rain CND.NEG.rain.IPF.3MSG good AUX.PAST.3MSG

(i) ‘It would have been good if it had not rained’

(ii) ‘It would be good if it did not rain’

b. K’abeena

t’eenoo ’ubbo-ba’i-kkaani t’uma-ha-kk ’i

rain.NOM fall.PFV.3MSG-NEG-CND good-COP.M-IRR

(i) ‘ It would have been good if it had not rained’

(ii) ‘It would be good if it did not rain’

The usage of past markers in the apodosis of an irrealis conditional clause is

a rare grammaticalization. Very often past markers occur in the protasis of a

conditional clause, e.g. in English. According to Fleischman (1989:4f.)the

Joachim Crass and Ronny Meyer248

Table 7.3 Possible new features of the ELA

K’abeena Libido Oromo Amharic Zay Wolane Guma

r Muher

Morphological features

Ablative > comparative þþþþþþþ

Ablative > ‘since’ temporal þþþþþþþ

‘Since’ temporal > real conditional þþþþþ

Simile > complementizer þþþþþþþþ

Complementizer > purpose clause þþþþþþþþ

Syntactic features

Prospective and intentional as separate

categories

þþþþþþ

Experiential perfect þþþþþþþþ

Benefactive focus þþþþþþþþ

Different copulas in main and subordinate

clauses

þþþþþþþþ

Exist > ‘have’ possession þþþþþþþ

‘Have’ possession > obligation þþþþþþ

Past > marker of irreal conditional clauses

in the apodosis

þþþþþþþþ

‘‘relationship between past tense and non-actuality has been widely acknowl-

edged in the linguistic literature.’’ With regard to conditional clauses the ‘‘scalar

view of time and probability ... may be inferred ... The greater the likelihood

that a situation will be realized, i.e. the closer to ‘reality’ the speaker perceives it

as being, the closer to ‘now’ ( ¼ present) will be the tense used to represent it;

similarly, the lesser the likelihood ascribed by the speaker to the situation, the

further in the direction of past will be the tense used to represent it.’’ However,

the ‘‘typical’’ place for a past marker in bi-clausal conditional sentences is the

protasis of hypothetical or contrary to fact conditions. The use of the past tense

for the apodosis seems to be quite rare (Fleischman 1989:6f.).Apossible

explanation for the occurrence of past markers in the apodosis in certain lan-

guages is given by James, cited from Fleischman (1989:6):‘‘James (1982)

suggests that the apodosis is the more hypothetical, whence more irrealis, of the

two clauses. The protasis ... sets up an imaginary world where X is the case.

The fact that it is an imaginary world means that we are already one step away

from reality ... Under the condition ‘given X, Y,’ a further logical step is

required for the realization of Y, which, being contingent on X, is therefore more

hypothetical and further removed from reality than X.’’

The relative rareness of the occurrence of past markers in the apodosis in the

languages of the world leads to the assumption that in the case of Amharic and

K’abeena language contact is one possible explanation.

7.4 Summary

Most of the features discussed occur in all investigated languages, as illustrated

in table 7.3. In Amharic, K’abeena, and Zay all the features are attested.

Wolane, Guma

r, and Muher lack one feature, Libido lacks two; and Oromo

lacks ﬁve features.

The new features that are proposed in this chapter support the assumption

that Ethiopian languages indeed form a linguistic area, and they enlarge Fer-

guson’s(1976) number of features considerably. Since Amharic and Oromo

are lingua francas in most parts of Ethiopia, we expect contact-induced spread

of at least some of these features into languages spoken in other parts of the

country to have played some role. However, the current number of investigated

languages and features is too small to propose a deﬁnite conclusion. Further-

more, most of the investigated languages are spoken in the Highland East

Cushitic/Gurage subarea. Therefore, further research on the occurrence of the

features in languages of other parts Ethiopia is necessary.

Joachim Crass and Ronny Meyer250

The marked-nominative languages

of eastern Africa

Christa Ko

nig

Africa is a continent where grammaticalized case systems are a rare phenomenon.

Of the roughly 2,000 languages there are only a few with grammaticalized case,

probably less than one-tenth of all African languages. Of the two basic case

systems, (nominative/)accusative and ergative(/absolutive) distinguished world-

wide, the latter hardly occurs in Africa, and the former accounts for less than one-

third of all African case languages. The majority of African case languages,

that is, roughly two-thirds, belong to the so-called marked-nominative type.

Elsewhere in the world, marked-nominative systems are virtually non-existent.

African case languages show an areal and genetical distribution: Afroasia tic

and Nilo-Saharan are primarily the phyla with case languages. East ern Africa

is a region with a high concentration of case languages in gener al and of

marked-nominative languages in particular. In the border region of Uganda,

Kenya, Ethiopia, and Sudan, marked-nominative languages are nearly the only

type of case-mark ing languages to be found. Neighboring marked-nominative

languages may belong to different phyla, namely Afroasiatic or Nilo-Saharan.

In this chapter it is argued that marked-nominative systems are at least to some

extent an areal phenom enon.

8.1 Introduction

Marked-nominative case systems stand out against other types of case systems,

their deﬁning property being that in such systems the nominative case is func-

tionally marked vis-a

-vis the accusative case (see section 8.2.1 for more details).

In the relevant literature, the phenomenon of marked-nominative languages

has been recognized mostly from the perspective of each regional philology:

Cushitists, Berberologists, and Niloticists all have looked at the phenomenon

within their own language subgroup. One resu lt of this more narrow inspection

is terminological confusion: one and the same phenomenon receives two or

more different labels, or one label is used to refer to different phenomena. So

far there is no general treatment of this language type.

The concern of this chapter is this particular type of case marking, which

is found mainly in eastern and northeastern Africa. There are also a few

251