55

7. CHADIC OVERVIEW
RUSSELL G. SCHUH

1. Background

The Chadic Language family comprises about 150 languages that lie in a band south of the
Sahara desert, stretching across southern Niger, northern Nigeria, northern Cameroon, and south-
central Chad Republic. A single Chadic language, Hausa, constitutes the entire western end of the
Chadic area. Hausa is the overwhelmingly largest Chadic language with well over 20 million
native speakers, whereas most other Chadic languages have fewer than 100,000 speakers. The
Hausa native-speaking area covers most of the northwest quadrant of Nigeria and most of southern
Niger. It is the lingua franca
for all of northern Nigeria
and most of Niger. Chadic
languages at one time
probably formed a block
from near the Niger River in
the west to their current
eastern extent in Chad Re-
public. Hausa has absorbed
many of its cousins in the
west, Kanuri has replaced
Chadic languages west and southwest of Lake Chad, and Kanembu and Chadian Arabic have cut
through the original Chadic area of Chad Republic.
In the 1930’s, Johannes Lukas presented the first comprehensive classification of languages of
the Lake Chad basin, proposing a “Chado-Hamitic” group and a “Mandara” or “Chadic” group
(Lukas 1936). A principal criterion for this categorization was the presence or absence of gram-
matical gender. Greenberg (1950) set the stage for the modern classification of Chadic. On the
basis of lexical resemblances and shared patterns in morphology, Greenberg proposed a unitary
group of “Chad” languages with nine subgroups, which included all Lukas’s “Chado-Hamitic” and
“Chadic” languages and was itself a branch of the larger Afrasian (= Afroasiatic) family.
In the first true comparative Chadic publication, Newman and Ma (1966) proposed a
subgrouping that unified Greenberg’s two largest groups of Chadic—group 1 at the western end
and group 9 at the eastern end—into a “Plateau-Sahel” branch, with the remainder of Greenberg’s
groups united in a “Biu-Mandara” branch. They identified a number of sound changes that united
or differentiated the various subgroups and reconstructed a list of 150 items for a proto-Chadic
lexicon.
Newman and Ma’s Plateau-Sahel group soon became untenable. Newman (1978) himself
noted that counts of lexical similarities did not justify a special linkage between the western branch
(Greenberg’s group 1) and the eastern branch (Greenberg’s group 9). Moreover, a major criterion
for Newman and Ma’s classification was a putative correspondence of Plateau-Sahel *s to Biu-
Mandara *Ò. Research in the late 1960’s revealed that many of the Plateau-Sahel languages of
Nigeria also have lateral fricatives, thus ruling this out as a criterion for subgrouping.
Drawing on a large amount of Chadic research and publication in the 1960’s and 1970’s, two
reclassifications of the Chadic languages emerged about the same time in the late 1970’s. Newman
(1977) updated what was known about sound changes among the Chadic languages and presented
56 RUSSELL G. SCHUH

a new list of 150 proto-Chadic reconstructions. Newman proposed a four-way division of the
Chadic languages into West (= Greenberg’s group 1, all spoken in northern Nigeria and including
Hausa), Biu-Mandara (= Greenberg’s groups 2-7, spoken in northeastern Nigeria and northern
Cameroon), Masa (= Greenberg’s group 8, spoken mainly in western Chad, extending into
eastern Cameroon), and East (= Greenberg‘s group 9, all spoken in Chad Republic). The map
above reflects this classification, lighter to darker shades of gray distinguishing the groups from
west to east.
The second reclassification is that of Jungraithmayr and Shimizu (1981), repeated in
Jungraithmayr and Ibriszimow (1994). This classification has West and East branches identical
to those of Newman (1977). Both reclassifications also recognize “Masa” as a unitary group.
However, unlike Newman’s distinct high-level groups of Biu-Mandara and Masa, the
Jungraithmayr et al. classification has a single Central Chadic branch. This branch, in turn, has
three subbranches, with the languages of Newman’s Masa branch included within one of those
subbranches. It may well be that Masa is linked to the other “Biu-Mandara”/“Central Chadic”
languages, but a low-level classification of Masa within Central Chadic is questionable. Shryock
(1989), using a variety of criteria from sound change, lexical change, and morphological change,
was unable to find evidence for a link between the Masa languages and the Biu-Mandara/Central
Chadic languages as opposed to any other branch of Chadic. The conservative Newman (1977)
classification, with four-high level branches, thus seems to be the most plausible among current
proposals. Branch affiliations of languages cited below will be indicated by [W], [B-M], [M], and
[E] following the language name.

2. Phonology

Consonants: Certain details in the reconstruction of a proto-Chadic consonant system remain

unresolved, e.g. the status of prenasalized consonants and velar fricatives (see Newman (1977)
and Jungraithmayr and Ibriszimow (1994) for proposed reconstructions and sound changes that
characterize particular languages and groups). All Chadicists agree that the proto-Chadic
consonant inventory included two noteworthy categories of consonants: a set of glottalized sounds
and at least one lateral fricative. All Chadic languages have both ∫ and fi, and there are
reconstructable proto-Chadic roots containing these sounds, e.g. *∫©r- ‘cut, break’ (Hausa [W]
∫a\ra¥ , Bade [W] ∫©æru ‘split into sections’, Bura [B-M] ∫©ælaæ ‘break (a stick)’), *f©fiu ‘four’
(Bade [W] f©fiu, Gude [B-M] næfwafi, Musey [M] fâfiì, Mokilko [E] pìfie). Most Chadic
languages have at least one glottalized sound in addition to these two, e.g. &y, ts&, ˚ , ƒ, but
convincing cognates providing evidence for the existence of, much less phonetic details of a third
glottalized consonant are harder to come by than those for ∫ and fi.
There are languages with lateral fricatives in all the major branches except East Chadic, and
lateral fricatives appear in several reconstructable lexical items, e.g. *Ò´w- ‘meat’ (Newman 1977)
as seen in Ngizim [W] Òu~wai, Mofu [B-M] aÒa~w, Musey [M] Òiw-na. The reconstructability of
lateral fricatives for proto-Chadic was not recognized until the 1970’s, in part because of
inadequate phonetic transcription in earlier documentation, in part because *Ò has shifted
independently in a number of languages to other sounds, usually sonorant [l] or a coronal
fricative, e.g. the root for ‘meat’ has become Bole [W] lo, Gude [B-M] lùwa, Kotoko [B-M] s' u ,
Mokilko [E] séy.
Vowels: A widespread Chadic vowel system is i (ê) © ªÈº u (u\)
a, a\
A macron over a vowel indicates a long vowel. We can reconstruct an *a vs. *a\ contrast, a
distinction that we find in languages in every branch of Chadic, as in Bade [W] kaædu ‘snap in
two’ vs. ka¥du ‘bite’. Many languages also distinguish [i] from [ê] and [u] from [u\ ] . Although
the phonologies of some languages now treat these pairings as a vowel-length distinction parallel to
/a/ vs. /a\/, the origin of the phonetic long high vowels is probably *V+ homorganic glide, an
origin still evident in morphological processes of many languages, such as Ngizim [W] zàyat
CHADIC OVERVIEW 57

‘slaughtering’ < [zÏdu] ‘to slaughter’, Hausa [W] duwa¥tsu\ ‘stones’ < [du\tse¥] ‘stone’. Though
many Chadic languages have mid vowels, these probably come historically from loanwords,
coalescence of diphthongs, or other processes that are active in modern languages. It is worth
pointing out that this rather sparse vowel system with a length distinction in the low vowel, which
is typologically similar to the vowel systems of Semitic and Berber, is widespread in Chadic
despite millennia of contact with Niger-Congo languages, which typically have five, seven, and
even nine vowels systems and no distinction in vocalic length.
Tone: All Chadic languages are tone languages. The family-wide presence of tone
distinguishes Chadic from other Afrasian families aside from Omotic, whose Afrasian affiliation is
questionable in this writer’s view. Since proto-Afrasian was probably not tonal, the most likely
source of tone in Chadic is early and continued contact with non-Afrasian tone languages.
By far the most common Chadic tone system has high (H) and low (L) level tones and a
downdrift intonation pattern, in which H following L does not rise to the pitch register of
preceding H. Compare the tone levels on Hausa [W] farê [ \  \ \  \] HH ‘white’, farÏ [ \  \ _] HL
‘drought’, faærê [_ –] LH ‘rolling the eyes’. Some Chadic languages have three tone levels, e.g.
Lele [E] kìrè LL ‘road’, kire MM ‘ruins’, kírwé HH ‘leopard’. Finally, some languages have
“terrace level” systems in which one may descend from H to L or from H to “downstepped H”
(!H), but from L one may only ascend to H (or, more accurately, to !H). !H sets the new register
level for H, which itself may descend to yet another !H or to L. For example, in Miya [W]
a;ts©;tsalìy [ \  \ – – _] H!HHL ‘cowry’, the second syllable has !H, which establishes the new H
register, maintained by the third syllable, and the fourth syllable is L.

3. Morphology

A number of features of morphology show both a connection among the Chadic languages and
a connection of Chadic to Afrasian. From proto-Afrasian, Chadic inherited a system which
distinguishes MASCULINE and FEMININE grammatical gender in the singular but has only a single
PLURAL category undifferentiated for gender (differentiation for gender in the plural in Semitic and
Berber is clearly a shared or convergent innovation). For example, Miya [W] has demonstratives
na;-ka ‘that one (m)’, ta;-ka ‘that one (f)’, nây-ka ‘those (common gender)’. These Miya
demonstratives illustrate a second aspect of Afrasian inheritance, viz. a pattern correlating *n-
masculine, *t- feminine, *n- plural (Greenberg 1960). This pattern emerges in languages of all
Chadic branches except East. For example, Musey [M] nouns in citation form have
gender/number marking suffixes /-na/ masculine and plural, /-ta/ feminine, e.g. Hu~-na~ (m)
‘billy-goat’, Hu~-ra~ (f) < *Hu~-ta ‘nanny-goat’, Huænâ-na ‘goats (any sex)’.
Rather than the n/t/n pattern, languages of the East branch have a k/t/k pattern, e.g. Lele [E]
bayndâ k-oloΩ ‘that man’, tama; t-oloΩ ‘that woman’, kara k-oloΩge; ‘those people’. The *k is
likewise of Afrasian origin, seen for example, in Cushitic. In Chadic outside the East branch, this
*k is a gender/number-neutral determiner base (cf. the Miya demonstrative forms above),
suggesting that its specific function as “non-feminine” may be an innovation within the East branch
of Chadic.
Another element of Afrasian morphology shared across Chadic is the pronominal system. I
limit discussion just to the striking pattern of second-person pronouns. The table here shows
cognate forms—languages have shifted the functions of these forms in various ways. “G” in the
Mokilko masculine form = “gemination of the next consonant”.
Egyptian Tuareg Ngizim Hausa Kotoko Mokilko
(Berber) [W] [W] [B-M] [E]
‘you’ (ms) c'w ka¨y c'ì ka\ -ku -kìG-
‘you’ (fs) c'm ka¨m k©æm kin -k©m -kìn-
‘you’ (pl) c'n ka¨wa¨na¨y kuæn kun -kun -uænn-
The striking pattern of specific correspondences in this small array of forms alone is enough to
demonstrate the genetic unity of Afrasian and the inclusion of Chadic within that unity: (1) *k- as
the base of all second-person pronouns; (2) a tripartite MASCULINE/FEMININE/PLURAL system
58 RUSSELL G. SCHUH

parallel to that for the nominal system; (3) *-n as a plural marker; (4) most striking, *-m as part
of the feminine pronoun, distinguishing it from the simple CV shape of all other singular pronouns
and from all other feminine gender-marked forms, where *t is the canonical sign of gender (see
Newman (1980) for a fuller discussion of Chadic and Afrasian pronouns).

4. Basic Sentence Structure and Function Marking

Basic sentence structure: No Chadic language uses nominal case marking. All Chadic
languages rely on word order and/or prepositions to show syntactic function. Bole [W] is typical.
Subjects are immediately preverbal, direct objects are immediately postverbal, and prepositions
mark other functions.

Bamoi kaæppu\ moærfio m; bo\-nì gaæ jaæfiaæ ‘Bamoi planted millet for his father with a hoe’
Bamoi planted millet for father-his with hoe

A widespread Chadic feature is to place NOMINAL indirect objects in prepositional phrases (as
above) but to realize PRONOMINAL indirect objects as verbal clitics preceding direct objects,
including pronominal direct objects, as in Bole

ita aæ fioæppaæ-no ishì ‘she will follow him for me’ not ‘she will follow me for him’
she imperf. follow-me(IO) him

The vast majority of Chadic languages have SVO as basic main clause order. The Bole
sentences above illustrate SVO order in a West Chadic language. Below are examples from other
branches:

Bura [B-M]: Anjukwi k© tl©ka yarì ‘Anjukwi planted millet’

anjukwi perf. plant millet
Zime [M]: Ekwa geæ zìw lo; sânara; ‘Ekwa threw a line into the river’
Ekwa threw rope in river
Mokilko [E]: guædre yiloæmme pu¥ni ‘the crocodile was hunting fish’
crocodile hunt fish

Along the Nigeria-Cameroon border are a number of languages with VSO order, e.g.

Gude [B-M]: k© &ush Ra\bi fia]fna ‘Rabi cooked mush’

perf. cook Rabi mush

All the Chadic VSO languages lie in a contiguous geographical area and all are members of the “A”
subgroup of Biu-Mandara (though most Biu-Mandara “A” have languages SVO typology). This
suggests that development of VSO order may be a local innovation in Biu-Mandara “A”. On the
other hand, there are arguments for considering this to be a remnant area preserving an earlier,
more widespread, if not proto-Chadic order. A shift of VSO to SVO is common in languages of
the world, whereas the opposite shift is rare. Moreover, the most ancient non-Chadic contacts
would have been with SVO languages, creating an environment for shifting AWAY from VSO
order. Finally, within Afrasian, VSO is the prevalent order in Berber and older Semitic languages,
making inheritance of this order from an ancestral language into Chadic plausible.
Questions: In forming questions containing question words, some languages place all question
words in sentence initial position. This is true of a minority of SVO languages, e.g. Hausa [W]
and Bura [B-M], and most VSO languages (compare the Gude example above with the
corresponding questons, wuæ &uæshi fiaf©\}n a\@ ‘who (wù) cooked this mush?’, mì &ush Ra\bi a\@
‘what (mì) did Rabi cook?’). However, a large majority of SVO Chadic languages and some VSO
CHADIC OVERVIEW 59

languages have in situ ordering for question words, i.e. the question word occupies the same
sentence position as a functionally corresponding noun.

Kirfi [W]: Ωæka\li je\-shì@ ‘who (Ωæka\li) put it down?’

kaæ je\wu ma¥mi@ ‘what (ma¥mi) did you put down?
Tera [B-M]: ki a! zurt´ Lua na? ‘who (ki) is frying the meat?’ (na, a are question particles)
Ali masa n©m a@ ‘what (n©m) did Ali buy?’
Zime [M]: sa;a gì kìr su;@ ‘who (sa;a) stole (it)?’ (sú is a question particle)
ti kìr mi su;@ ‘what (mi) did she steal?’
Kera [E]: mintâ l©;ta;Ω n©wri mo;@ ‘who (mintí) hit your sister? (mó is a question particle)
Ageæleæm l©;ta;Ω mintâ mo;@ ‘who (mintí) did Agelem hit?

Questioned subjects deserve special mention. Some SVO Chadic languages that use in situ
ordering for NON-subjects have the unusual feature of placing questioned subjects in postverbal
position. Compare the Bole questions below with the first example in this section:

Bole [W]: kaæppu\ moærfio loæ@ ‘who (lò) planted millet?’

Bamoi kaæppu\ leæ@ ‘what (lè) did Bamoi plant?’
The languages that question subjects in this way are all West Chadic and all lie along the edge of
the West Chadic area, extending from Bade in the north to Kanakuru in the south. This syntactic
property is surely not an accidental parallel innovation in these languages, and it seems unlikely to
be a result of areal drift—these language fall into a long, narrow north-south strip with Chadic and
non-Chadic neighbors to the east and west having preverbal subjects only. Postposing questioned
subjects must therefore have its source in inheritance. A scenario that brings in the VSO languages
would be as follows: the ancestral language had VSO order, with SVO as an alternative order in at
least some constructions, a typical feature of VSO languages around the world. Through parallel
internal developments and/or through areal drift, languages began to move toward SVO as their
basic order, ultimately leaving just the remnant VSO area that exists today. As languages drifted
toward SVO as the basic sentence structure, the postverbal position for subjects came to be a
“marked” position in at least part of the West Chadic area (note that the languages that postpose
questioned subjects also postpose focused subjects, as in Bole [W] kaæppuwo¥ Bamoi ‘BAMOI
planted (it)’). Most languages have abandoned postverbal placement of subjects in any function,
leaving just a strip of West Chadic languages that postpose subjects for questioning and focus and
a strip of Biu-Mandara languages that retain basic VSO sentence structure.

5. Lexicon

As in morphology, some of the strongest lexical indicators of the genetic unity of Chadic are
also indicators of the affiliation of Chadic with the rest of Afrasian, e.g. *m-t- ‘die’ (Bole [W]
motu, Musey [M] mít—cf. Egyptian mt, Arabic ma\t), *s-n- ‘know’ (Miya [W] s©n, Mokilko
[E] su¥ne—cf. Egyptian swn, Tuareg ©ss©n), *b-n- ‘build; house’ (Bole [W] bònò ‘house’,
Mokilko [E] bê  }no ‘grass hut’—cf. Arabic bana\), and *f©fiu ‘four’ (cf. Egyptian fdw) and *Ò´w-
‘meat’ (cf. Tuareg isan) mentioned in §2.
Of interest both for demonstrating the unity of Chadic as a distinct subfamily of Afrasian and
for saying something about the homeland of the speakers of proto-Chadic are a number
reconstructable words referring to flora and fauna. These items suggest a savannah homeland near
water, probably the great inland sea that Lake Chad once was. (Question marks indicate absence
of a meaning equivalent or cognate in available materials, hyphens indicate that the meaning
equivalent is not cognate.)
60 RUSSELL G. SCHUH

Hausa [W] Bade [W] Bura [B-M] Musey [M] Kera [E]
‘baobab’ ku\ka¥ kukwa;u kwaægu koΩgo;-na koΩkoΩ
‘Acacia nilotica’ gaæba¥r-uwa] ©ægva¥l-aækau @ @ @
‘fish’ kêfê --- k©ælfaæ kuluæf-naæ ---
‘crocodile’ kada¥ ©ægd©m ng©æl©æm huruæm-maæ kim
‘Nile monitor’ guza¥ ©ægzaæn gaædzaæ-gaædzaæ @ ---

REFERENCES

Greenberg, Joseph H. 1950. “Studies in African linguistic classification: IV Hamito-Semitic.

Southwestern Journal of Anthropology 6:47-63.
Greenberg, Joseph H. 1960. “An Afro-Asiatic pattern of gender and number agreement.”
Journal of the American Oriental Society 80:31-321.
Jungraithmayr, Herrmann and Kiyoshi Shimizu. 1981. Chadic Lexical Roots. Vol.2: Tentative
Reconstruction, Grading and Distribution.. Marburger-Studien zur Afrika- und Asienkunde,
Serie A, Afrika, 26. Berlin: Dietrich Reimer.
Jungraithmayr, Herrmann and Dymitr Ibriszimow 1994. Chadic Lexical Roots. Vol. 1:
Tentative Reconstruction, Grading, Distribution and Comments; Vol. 2: Documentation.
Sprache und Oralität in Afrika, 20. Berlin: Dietrich Reimer.
Lukas, Johannes. 1936. “The linguistic situation in the Lake Chad area in Central Africa.” Africa
9:332-349.
Newman, Paul. 1977. “Chadic classification and reconstructions.” Afroasiatic Linguistics
5(1):1-42.
Newman, Paul. 1978. “Chado-Hamitic ‘adieu’: new thoughts on Chadic language
classification.” In Pelio Fronzaroli (ed.), Atti del Secondo Congresso Internazionale di
Linguistica Camito-Semitica, pp. 389-397. Florence: Istituto de Linguistica e di Lingue
Orientali, Universita di Firenze.
Newman, Paul. 1980. The Classification of Chadic within Afroasiatic. Leiden: Universitaire
Pers Leiden.
Newman, Paul and Roxana Ma. 1966. “Comparative Chadic: phonology and lexicon.” Journal
of African Languages 5:218-251.
Shryock, Aaron. 1989. “Issues in Chadic classification: the Masa Group.” MA thesis, UCLA
Department of Linguistics.