Heine Bernd, Nurse Derek. A Linguistic Geography of Africa

Подождите немного. Документ загружается.

The Sudanic belt is well deﬁned by the next group of features: labial ﬂaps,

labial-velar stops, implosives, nasal vowels, 2H vowel systems, multiple tone

levels and ‘‘lax’’ question prosodies, among others. None of these features are

as common in other zones. Nor, as we have seen, are they equally distributed

within it; however, their overlap deﬁnes the Sudanic belt quite well, with the

exception of the extreme northwest (northern Atlantic languages) and the

northeast (the eastern Sudan, as discussed in section 3.2.10).

A third zone, the South, is sharply delineated by the remaining features in

table 3.11: ejective and aspi rated stops, clicks, and sla ck voic ed stops. To these

features we could add their characteristic series of lateral affricates and fri-

catives. All these features are widely shared by Khoisan and Bantu languages

in the region.

Less well demarcated is the East zone, whose languages share many features

with those of the North due to their common Afroasiatic heritage. Neverthe-

less, the two non-Semitic families in this zone, Omotic and Cushitic, display

several features different from those of the North, notably the widespread

presence of tone or tonal accent systems, and the common occurrence of

implosives and ejectives (sometimes in the same language). We have seen that

only in this zone does



occur as the unique implosive. Ejectives also occur in

Ethio-Semitic languages (see Crass 2002 for a fuller account).

The Center is well deﬁned by the inherited features of the Bantu languages

spoken within it, and does not as a whole display the characteristic features of

the Sudanic languages spoken to the north nor the Khoisan and Bantu lan-

guages spoken to the south. We have seen that it is well characterized by a

unique system of vowel harmony.

As far as the more diverse Rift zone is concerned, this survey has not

succeeded in identifying large-scale diffusion of phonological features across

major genetic boundaries, the hallmark of a genuine phonological area. This

fact might well call the independence of this zone into question. It remains to

be seen, however, whether further study will reveal cases of such diffusion, at

least in micro-areas.

Appendix

The African phoneme database used for this study is composed of most of the

African languages contained in the UPSID database (Maddieson & Precoda

1989) together with others that we have added, mainly in the Sudanic belt, in

order to improve the geographical coverage. It is divided into six parts,

according to zone. It contains 88 Niger-Congo languages, 30 Nilo-Saharan

languages, 27 Afroasiatic languages, and 5 Khoisan languages, for a total of

150 languages spoken indigenously on the African continent. We have cor-

rected and updated information on certain UPSID languages based on more

G. N. Clements and Annie Rialland82

Table 3A Composition of the Sudanic database, by genetic groups

NIGER-CONGO (66)

Dogon: Dogon

Atlantic: Wolof, Pulaar, Diola, Konyagi, Ndut, Temne, Bidyogo,

Kisi

Mande: Kpelle, Bambara, Bobo-Fing, Dan, Bisa

Kru: Aizi, Klao, Bete

Gur: Dagbani, Mo

ore

, Bwamu, Tampulma, Senadi, Bariba

Kwa: Alladian, Adioukrou, Attie

, Akan, Ga

, Lelemi,

Siya, Ewe-Gen

Ijoid: Ijo (Izon)

West Benue-Congo: Yoruba, Isoko, Igbo, Gwari, Igede

East Benue-Congo:

non-Bantoid: Amo, Birom, Tarok, Kpan, Eﬁk, Ogbia, Kohumono

Bantoid, non-Bantu: Mambila, Ejaghem, Noni, Aghem, Fe’fe’

Bantu: Kpa/Baﬁa, Ewondo, Makaa, Basaa, Yaka/Aka, Egbuta,

Bila

Adamawa: Doayo, Mumuye, Mbum, Lua

Ubangi: Gbeya, Azande, Mba (Mba-Ne), Sango

Kordofanian: Moro, Jomang

NILO-SAHARAN (23)

Central Sudanic: Yulu, Sar, Furu/Bagiro, Kresh, Lugbara, Ngiti,

Mangbetu

Eastern Sudanic: Nera, Nyimang, Tama, Mursi, Tabi, Temein, Daju,

Dinka

Other: Zarma, Central Kanuri, Maba, Fur, Berta, Kunama,

Koma, Kadugli

AFROASIATIC, CHADIC (11)

West: Hausa, Kanakuru, Angas, Ngizim

Biu-Mandara: Tera, Margi, Kotoko, Higi

East: Kera, Dangale

Masa: Lame

Table 3B Composition of the North, East, Rift, Center, and South databases

NORTH (7)

Afroasiatic, Berber: Shilha, Tamasheq

Afroasiatic, Semitic: Egyptian Arabic

Afroasiatic, Cushitic: Beja

Nilo-Saharan: Tedaga, Nobiin, Koyraboro Senni Songay

EAST (12)

Afroasiatic, Semitic: Amharic, Tigre, Chaha

Afroasiatic, Cushitic: Awiya/Awngi, Oromo, Somali, Dahalo

Afroasiatic, Omotic: Dizi, Hamer, Kefa/Kafa, Kullo

Niger-Congo, Bantu: Swahili

Africa as a phonological area 83

RIFT (9)

Afroasiatic, Cushitic: Iraqw

Nilo-Saharan, Nilotic: Luo, Maasai, Sebei

Nilo-Saharan, Kuliak: Ik

Niger-Congo, Bantu: Kikuyu, Ganda

Khoisan: Hadza, Sandawe

CENTER (13)

Niger-Congo, Bantu: Tsogo, Teke, Beembe, Mongo-Nkundu, Lega, Rwanda,

Bemba, Mwera, Makhuwa, Luvale, Umbundu, Herero,

Zezuru Shona

SOUTH (9)

Central Khoisan (Khoe): Nama

Northern Khoisan: !Xu

(!Kung)

Southern Khoisan: !Xo

Niger-Congo, Bantu: Gciriku/Diriku, Tsonga, Yeyi, Copi, Tswana, Zulu

Table 3C Composition of the tone level database

NIGER-CONGO (55)

Mande: Samo (3), Guro (3), Santa Dan (5), Tura (4), Bobo-Fing (3)

Kru: Wobe

(4), Be

(4), Vata (4), Godie

(3), Neyo (3), Dewoin/

De (3), Nyabwa (4), Krahn (3)

Gur: Togolese Moba (4), Bariba (4), Ncam (3), Akaselem (3),

Nateni (3), Gulmancema (3), Kasem (3), Nuni (3), Biali (3)

Kwa: Attie

(4), Abbe

(3), Alladian (3)

West Benue-Congo: Nupe (3), Igede (4), Kana (4), Gwari/Gbari (3), Yoruba (3),

Yala (3), Igala (3)

East Benue-Congo:

non-Bantoid: Mbembe (5?), Jukun (3), Birom (3), Kpan (3)

Bantoid, non-Bantu: Tikar (3), Mambila (4), Ndoola (4), Bafut (3), Babanki (3)

Bantu: Ewondo A70 (3), Mbati C13 (4?), Nyali D23 (3), Bira D21

(3), Bila D311 (3)

Adamawa-Ubangi: Banda-Linda (3), Ngbaka (3), Zande (3), Sango (3), Tupuri

(4), Yendang (4), Munzombo (4), Doayo (4), Mumuye (3)

NILO-SAHARAN (8)

Central Sudanic: Moru-Madi (4), Lugbara (4), Mangbetu (4), Bedionde (3),

Yulu (4)

Nilotic: Dinka (3), Shilluk (3), Nuer (3)

AFROASIATIC (9)

Chadic: Tera (3), Ga’anda (3), Angas (3), Kera (3), Lame (3)

Omotic: Dizi (3), Sheko (3), Yem (3), Bench Gimira (5)

KHOISAN (4) Nama (3), Kxoe (3), Tsoa (3), !Xo

(4)

Table 3B (Cont.)

G. N. Clements and Annie Rialland84

recent or more accurate information. Non-African phoneme systems have been

drawn from the unmodiﬁed UPSID database.

Table 3A show s the genetic composi tion of the Sudanic database. A few

units, notably Gur, Kainji-Platoid, Cross River, Adamawa, Ubangi, and

Chadic, are underrepresented in proportion to their numbers, but as these units

are centrall y located in the Su danic belt this should not lead to a severe

underestimation of shared Sudanic properties.

Table 3B gives the com position of the North, East , Rift, Center, and South

databases.

Table 3C g ives the com position of the tone level database , consi sting of

76 languages with three or more cont rastive tone levels. The number of

contrastive levels in each language is shown in parentheses.

Table 3D gives the composi tion of the yes/no questio n intonat ion database ,

containing 75 languages for which relevant information was found.

Table 3D Composition of the question intonation database

NIGER-CONGO (60)

Atlantic: Wolof, Fulfulde

Mande: Mende, Baule, Samo, Guro

Kru: Godie

, Neyo, Bassa, Klao, Wobe

, Vata

Gur: Togolese Moba, Mo

ore

, Ncam, Akaselem, Kusaal,

Nateni, Moyobe, Farefare, Dagaare

, Gulmancema,

Kasem, Kabiye, Tem, Nawdem, Lobiri

Kwa: Adioukrou, Akan, Ga, Ewe, Gun, Fon

Adamawa/Ubangi Banda-Linda, Munzombo

Ijoid: Ijo (Izon), Nembe

West Benue-Congo: Isoko, Igbo, Yekhee, Degema, Engenni, Nupe

East Benue-Congo

non-Bantoid: Eﬁk

Bantoid, non-Bantu: Tikar, Ejagham

Bantu: Bafut, Bajele, Mongo-Nkundu, Holoholo, Dzamba,

Nyanga, Rundi, Shi, Saghala, Jita, Chewa, Southern

Sotho, Zulu, Swahili

NILO-SAHARAN (6)

Songay: Zarma

Central Sudanic: Bagiro, Ngiti

Nilotic: Dholuo, Turkana, Nandi

AFROASIATIC (8)

Chadic: Hausa, Angas, Sayanci, Pero, Miya, Tera

Cushitic: Dahalo, Arbore

KHOISAN (1) Nama

Africa as a phonological area 85

Africa as a morphosyntactic area

Denis Creissels, Gerrit J. Dimmendaal, Zygmunt

Frajzyngier, and Christa Ko

nig

4.1 Introduction

The aim of the present chapter is to review a list of morphosyntactic features,

involving categories and their syntactic functions, which are likely to provide a

typological characterization of languages in general . The properties in ques-

tion are investigated from the following two points of view:

(i) Does the proportion of African languages possessing a given feature (or

being devoid of it) seem to be roughly similar to the proportion observed

at world level, or to be signiﬁcantly different?

(ii) Are the features whose frequency or rarity characterizes Africa as a

whole evenly distributed, or are there geographical clusterings, i.e. areas

with a particular concentration of languages possessing or lacking the

features in question, and if so, to what degre e do such clusterings

coincide with or diverge from genetic groupings?

We do not proceed by systematically testing the features we consider on the

basis of a language sample pre-established on the basis of statist ical methods

that would ensure its representativity. Apart from a few reference works whose

quality can be taken for granted, this chapter mainly relies on the authors’ ﬁrst-

hand knowledge of African data. This undoubtedl y limits the bearing of our

conclusions (or rather suggestions), but the set of African languages docu-

mented in a sufﬁcient way to be systematically used in such a study is so

limited that it is simply imposs ible to extract from it a sample representative of

the diversity of African languages.

4.2 Core grammatical relations

4.2.1 The recognition of subjects and objects

The relevance of a syntactic function “subject” grouping together the only

argument of semantically monovalent verbs (hereafter S) and the more agent-like

argument of semantically bivalent verbs (her eafter A) is obvious only for

languages in which, regardless of the precise semantic nature of the verb, both

have the same coding characteristics (in terms of case marking, indexation,

and/or constituent order). In other lang uages, the possibility of recognizing a

syntactic function “subject” is less obvious, since it cannot be justiﬁed on the

basis of immediately visible coding properties, but only on the basis of

behavioral properties in mechanism s such as reﬂexivization, relativization,

questioning, focalization, or clause chaining; in some languages, it may hap-

pen that (some of) these mechanisms function in a way that does not justify

grouping together S and A, which in turn may raise doubts about the relevance

of the notion of subject for the description of such languages.

From this point of view, African languages can be characterized as lan-

guages in which the notions of subject and object are not problematic, and most

often, the clearest manifestation of these notions can be found in systems of

pronominal afﬁxes of a type particularly widespread among Bantu languages

(see section 4.2.3), in which the same set of pronominal afﬁxes is used to

represent S and A, and another set is used to represent the patient of proto-

typical action verbs, and more generally the less agent-like argument of

semantically bivalent verbs.

Outside Africa, some language groups (e.g. Caucas ian languages) show a

marked tendency towards a closer correspondence between semantic and

syntactic roles, whereby verbs tend to divide into classes characterized by

different constructions according to the type of event they represent and the

types of semantic roles they assign to their arguments, and the assimilation of

the construction of semantically bivalent verbs to that of prototypical transitive

verbs is systematically limited by semantic conditions. The available docu-

mentation on African languages suggests a strong predominance of the

opposite tendency: to the best of our knowledge, syntactic descriptions of

African languages do not mention the existence of systematic semantic lim-

itations to the assimilation of the construction of semant ically bivalent verbs to

the prototypical transitive construction, and variations in the morphosyntactic

treatment of the single argument of semantically monovalent verbs are

exceptional in African languages.

4.2.2 Subject/object case marking typology

In this section, case marking is taken in a wide sense, including not only

morphological case, but also the use of adpositions to code the distinction

between subjects and objects.

In the majority of African languages, both subjects and objects are

unmarked for case, i.e. they do not exhibit any marking (afﬁx, adposition, or

prosodic contour) distinguishing noun phrases in subject or object function

Africa as a morphosyntactic area 87

from noun phrases quoted in isolation, or used in an extra-syntactic function of

pure designation. However, the situation is not exactly the same in all phyla

and in all geographical areas of Africa.

In Khoisan, as a rule, there is no case marking of the distinction between

subject and object. However, in some languages, the status of certain mor-

phemes as marking the discourse roles topic/focus or the syntactic roles sub-

ject/object is not entirely clear.

Case-marked subjects or objects are exceptional in Niger-Congo languages;

in the few Niger-Congo languages in which a case distinction between subjects

and objects may be recognized, it resu lts from recent historical developments:

either the reanalysis of a verb ‘take’ in a serial verb construction as a prep-

osition introducing objects, in some Kwa languages (Lord 1993), or the

reanalysis of a distinction between a deﬁnite and an indeﬁnite form of nouns as

a case distinction, in some western Bantu languages – see Blanchon (1999),

Schadeberg ( 1999). The western Bantu languages that have “tone cases”

constitute an compact geographical area from Gabon to Angola. The following

example is from Ngangela (Maniacky 2002):

(1) a. nouns in quotation, or in extra-syntactic function of pure designation:

nike ‘child’, kahu

mbi ‘hen’

b. kanike na

mono

kahu

mbi

12:child:SUBJ TAM:see 12:hen:ABS

‘The child has seen the hen’

Case distinctions between subjects and objects are less rare in the other two

phyla (Afroasiatic and Nilo-Saharan). Within Afroasiatic, the majority of

Chadic languages have no case system, but some Central Chadic languages use

prepositions to code objects, and many Berber, Omotic, Cushitic and Ethio-

Semitic languages distinguish between subject and object by way of mor-

phological case on nouns. Thus, in the Omotic language Maale, case sufﬁxes

(whose form depends on gender and deﬁniteness) as well as tonal inﬂection are

used to distinguish these two core functions (Amha 2001: 56–8). Compare:

(2) a. na-att-a

bayi yenk’-a

-ne

child-PL-SUBJ cattle:ABS herd-IPF-POS:DECL

‘Children herd cattle’

b. ?z na-att-o

k-a

-ne

3MSG:SUBJ child-PL-ABS like-IPF-POS:DECL

‘He likes children’

Within Nilo-Saharan, case marking is common in a range of language

groups stretching from Nigeria in the west to Ethiopia and Eritrea in the east,

more speciﬁcally in primary branches such as Saharan, Maban, For, various

Denis Creissels et al.88

Eastern Sudanic groups, e.g. the Tama group, Nara, Nubian, Surmic, and

Nilotic. Within this phylum, case-marking systems appear to be absent in a

number of Eastern Sudanic groups in the Nuba Mountains and west of this area

(Temein and Keiga Jirru, the Daju group) and in Central Sudanic.

Since these

groups are spoken in areas where Niger-Congo languages are found (which

typically lack case marking), areal contact may have played a role in the

disappearance of case.

An inventory of African languages having case contrasts between subjects

and objects can be found in Ko

nig (2004a, 2004b, and this volume).

In African languages that have case-marking systems distinguishing sub-

jects from objects, this distinction is often marked by tonal differences. This is

the case not only in all western Bantu languages that have a case distinction

(see above), but also, for example, in the Cushitic language Somali, or in the

Nilotic languages Dinka and Maasai. As shown by Amha (2001) for Maale,

and by Andersen (1988) for the Nilotic language Pa

ri, case marking through

tonal inﬂection may be used in tandem with sufﬁxation. In Pa

ri, the two

strategies appear to be entirely phonologically conditioned (Andersen 1988:

294–5). With nouns ending in a consonant in the absolutive, the erga tive is

formed by doubling the ﬁnal consonant if the latter is a sonorant, and by adding

a low-toned sufﬁx -i or -I (depending on vowel harmony). With vowel-ﬁnal

nouns, the ﬁnal vowel is replaced by the same sufﬁx if the ﬁnal root vowel is

high, and by -e or  (depending on vowel harmony) with non-high vowels. If

both vowel replacement and tone addition apply vacuously, the ergative form

is phonologically indistinguishable from the absolutive form:

(3) Absolutive Ergative



O˛ 



O˛˛-



I ‘spear’

lplp-



I ‘tongue’



d-i



 ‘stone’



IN-



O w



IN-

 ‘bird’

ob-i



ob-i



‘buffalo’



 pa



 ‘knives’

The drift towards loss of segment al layers and the maintenance of tonal

marking is part of a more general areal tendency to use tone as an exponent of

some inﬂectional or derivational process in African languages (similar

observations can be made, for example, regarding the expression of TAM

distinctions, or of deﬁniteness distinctions see section 4.2).

Among the languages that have a case distinction between c ore syntactic

roles with the same case form for S (intransitive subject) and A (transitive

subject), and a different case form for O, the most common type w orldwide is

that i n which the ca se form for S and A (nom inative) coinci des with the form

Africa as a morphosyntactic area 89

of nouns used for quotation or in the extra-syntactic function of pure des-

ignation, whereas the form for O, traditionally called Accusative, is a form

occurringonlyincertainsyntacticconditions. By contrast, the type of case

marking in which the quotation or designation form of nouns is used for O, a

case form distin ct from the quota tion or design atio n form of nouns being used

for both S and A, is very rare at world level. But African languages show a

very different ratio of these two variants of the accusative type of case

marking (Ko

nig 2004a, 2004b). The ﬁrst variant, which very strongly pre-

dominates at world level, is not particularly common on the African con-

tinent; in particular, it seems to never occur in case-marking s ystems

involving tonal distinctions. By contrast, so-ca lled “marked-nominative”

systems ( i.e. systems i n w hich t he quotation or designation form o f nouns is

used for O, a nd a case form distinct from the quotation or designation form is

used for both S and A)

are very common among African languages that have

case. Such systems are found in three different areas of the African continent.

The case systems found in Berber languages and western Bantu languages

(see the Ngangela example (1) above) all belong to the “marked-nominative”

type, and this t ype is also relatively common among the northeast African

languages that have distinct case forms for subjects and objects: most case

systems found in Nilotic, Surmic, Omotic, and Cushitic languages belong to

the “marked-nominative” type; for a detailed inventory, see Ko

nig (2004a,

2004b, and this volume).

A fairly common type at world level is the ergative type of case marking, in

which S and O share the same case form (absolutive), and A takes a special

case form (ergative). In general, ergativity is not a common phenomenon

among African languages; most manifestations of ergativity observed in

African languages can be characterized as “ubiquitous” or “pervasive” erga-

tivity, i.e. manifestations of ergativity that tend to occur even in the most robust

accusative languages. In particular, the ergative type of case marking is

exceptional on the African continent. It has been recognized only in some

Western Niloti c languages, i.e. in a language family in which case systems of

the “marked-nominative” type are predominant. According to Andersen

(1988), in basic sentences in Pa

ri, S and O precede the verb and are in a form

that coincides with the quotation or designation form, whereas A follows the

verb and shows a different form, identiﬁable as erga tive case:

(4) a. dha



O a



th



woman:ABS TAM:laugh

‘The woman laughed’

b. dha



O a

aN u

rr-i



woman:ABS TAM:insult Ubur-ERG

‘Ubur insulted the woman’

Denis Creissels et al.90

A similar system has been reported for the closely related Western Nilotic

languages Anywa (Reh 1996), Jur-Luwo (Buth 1981), and Shilluk (Miller &

Gilley 2001). With respect to the situation commonly found in non-African

languages that have a case marking of core syntactic roles of the ergative type,

the most striking characteristic of these Western Nilotic languages is a strong

correlation between word order and case marking. This is a general property of

Nilotic and Surmic case-marking systems: in Nilotic and Surmic languages

with case marking of the “marked-nominative” type, it is common that only

postverbal subjects are inﬂected for case; the same applies to the Western

Nilotic languages having a case-marking system of the ergative type, with the

additional constraint that in ergative systems, intransitive subjects (but not

transitive subjects) are restricted to occur in preverbal position.

In the languages of the wor ld, deﬁniteness and case marking often correlate,

and some northeast African languages that have case-marked core constituents

conﬁrm this tendency: the use of an accusative marker correlates with the

deﬁniteness distinction in Semitic languages like Amharic or Tigre,

as well

as some Cushiti c and Omotic languages. In the Omotic language Haro, sub-

jects as well as objects are inﬂected for case only when they are deﬁnite

(Woldemariam 2004).

4.2.3 Subject/object indexation typology

4.2.3.1 Types of pronominal markers In this section, the term

“pronominal marker” is applied to bound morphemes possessing the following

characteristics:

 A pronominal marker refers to an entity that is represente d elsewhere in the

same clause by a noun phrase, or could be represented by a noun phrase in

a clause identical in all other respects.

 A pronominal marker is attached to a word in a certain syntactic relation

with the noun phrase that represents or could represent the entit y to which

it refers.

Most commonly, pronominal markers attac h to verbs and

represent arguments of the verb they are attached to, but this is not the only

possible situation: so-called “possessive afﬁxes” are pronominal markers

attached to nouns (see section 4.4.6), and pronominal marker s may also

attach to adpositions (see section 4.5).

 A pronominal marker shows variations reﬂecting, either semantic character-

istics of the entity it refers to, or grammatical features of a noun phrase

referring to the same entity.

Pronominal markers typically show variations expressing distinctions that

parallel those expre ssed by free pronouns, in particular distinctions in person

and number.

Africa as a morphosyntactic area 91