Ulrich Kleinewillinghöfer Institut für Ethnologie Afrikastudien Genealogical classification

Ulrich Kleinewillinghöfer Institut für Ethnologie &Afrikastudien Genealogical classification of African languages beyond Greenberg The Gur - “ADAMAWA” Interface Or The “Adamawa” Dilemma

Greenberg 1963: Adamawa-Ubangi (A. -Eastern) • A. Adamawa 14 Groups 1: Tula, Dadiya, Waja, Cham, Kamu. 2: Chamba, Donga, Lekon, Wom, Mumbake. 3: Daka, Taram. 4: Vere, Namshi, Kolbila, Pape, Sari, Sewe, Woko, Kotopo, Kutin, Durru. 5: Mumuye, Kumba, Gemgle, Teme, Waka, Yendang, Zinna. 6: Dama, Mono, Mbere, Mundang, Yasing, Mangbei, Mbum, Kpere, Lakka, Dek. 7: Yungur, Mboi, Libo, Roba. 8: Kam. 9: Jen, Munga. 10. Longuda. • 11. Fali. 12. Nimbari • 13. Bua, Nielim, Koke. • 14. Masa = Kim • • – Later additions: Day ; Bikwin (Burak)

“Adamawa“ • 1. Inventory • • • Samarin 1971 Köhler 1975 Boyd 1989 Blench (Homepage, updated 2004) Ethnologue 2009

“Adamawa“ • 2. Classifications • • Köhler 1975 Bennett & Sterk 1977 Bennett 1983 Schadeberg 1986 Boyd 1989 Williamson & Blench 2000 Ethnologue … 2009

• Samarin : … “relationship [of Ubangi ] to Adamawa … not at all clear yet. (1971: 224) • Köhler : prefers to list 'Adamawa‘ and 'Ubangi- languages‘ seperately because • Ubangi shows ". . . die geringere Zahl der lexikalischen Beziehungen zum Westnigritischen. . . “ • while Adamawa shows ". . . allgemeine Beziehungen zum Westnigritischen. . . darunter anscheinend einige besondere Beziehungen zum Gur" (1975: 197).

• Bennett & Sterk 1977: 249 "Statistically, there is some evidence for a link between Gur and Adamawa-Eastern. Though the percentages are not large enough to be impressive, the two groups seem to form a continuum. Longuda and Tula of Adamawa-Eastern, for example, show significantly higher percentages of cognacy with More (Gur) than with some other Adamawa-Eastern languages. ”

• Schadeberg: 1986: 74 • "Adamawa-Ubangi has been marked as doubtful because it is only supported by the FN [Furthest Neighbour] classification; in the BA [Branch Average] classification, Tula clusters with the Gur languages and creates a link between Gur and Adamawa-Ubangi".

Bennett 1983: 44 • Finally, the third question: is there an Adamawa-Eastern? If we mean by this, is there a clear border between an approximation to Greenberg’s A. -E. and all other parts of Niger-Congo, the answer must be no. There is such a border between A-E and the neighbouring branches of South Niger. Congo, … no such border can be found between Gur and Adamawa. • … it is possible that a set of features defining a border may be found when adequate data are at hand… It is also not impossible that new data may indicate unity for most of A-E, but force the reassignment of the Trans-Benue languages to Gur

Bennett 1983:

Boyd 1989 • 3 core groups within the Adamawa branch to which most of the languages could be assigned to. 1 southwestern groups (Samba, Duru, Mumuye) 2 Mbum and the northern groups (Boa-Kula, Day, Kim) 3 a “northwestern unit might be posited" BENNETT's "Trans-Benue".

Williamson & Blench 2000

ETNOLOGUE 2009 ØLeko - Nimbari ØMbum -Day ØWaja - Jen Ø… ?

Gur – “Adamawa“ Interface • Relation were based mainly on the lexicon of Tula (and Longuda ) • Which then were extended to Trans – Benue • = 'Waja-Jen‘ • = Northwestern “Adamawa” (groups)

“Trans Benue“ ? • Apart from Bennett assumption that these groups "constitute a unit" (1983: 41), no one ever presented any evidence showing that : Bikwin-Jen, Tula-Waja, Longuda, and Ɓəna-Mboi (Yungur) • Form a valid genetic unit as against other groups like • Ba (= Kwa), Yandang, Mumuye and/or Gur. • (Boyd 1989, Kleinewillinghöfer 1996 a & 1996 b. )

NW – “ADAMAWA“ a genetic unit ? All the “Adamawa“ groups in this area report migrations. The autochthonous people (of the Muri Mountains) most probably spoke languages belonging to an unrelated language group (family? ). At least 2 or 3 different languages of this group were used to borrow vocabulary to replace tabooed lexems. Jalaa Extinct (Kleinewillinghöfer 2001, 1995)

NW – “ADAMAWA“ ? Lexicon : number of common possible cognates is not impressive. No clear cases of shared innovations known yet. Nominal morphology : Very heterogenous. Bikwin – Jen: almost no marking of plurals if at all then with a particular prefix For Tula – Waja, Yungur and Longuda see the handout.

Relationship between “Adamawa” and Gur (i) The case of Waja and Tula. Lexicon Phonology Nominal Morphology (Kleinewillinghöfer 1996)

Relationship between “Adamawa” and Gur (i) The case of Waja and Tula. (Kleinewillinghöfer 1996) Revised Conclusion : Tula – Waja ought to be included into Central Gur, its position remains to be defined. The morphological evidence points to OV But: How to define the border to “Adamawa“?

‘person‘ • • • Waja Dagara Tobote Nateni Fali nÄ-r-Ä / n. U-wa ni-r / ni-b. E nè-lé, nì-lé / nè-bé ni-ro, ni-lo / ni-ba ní-rù (Gorimbari) • -mba class 2 a • Pron. : 3 p. sg-

Gur - “Adamawa” interface (ii) The “cultural” vocabulary. (Kleinewillinghöfer, in print) Conclusion : North Volta Congo groups (Tula-Waja, Central Adamawa groups, Gbaya and Gur groups) formed a continuum in the Nigerian Savannah. On this basis more Adamawa languages have to be included.

Morphological Relationship between “Adamawa”groups and Gur (iii) Noun class systems in Central Adamawa languages. (Kleinewillinghöfer (Ms) LONto (Voko) Gimm. E – Vere group

Lɔŋtó -yɔ, -ø YO BO -bó, ‑mó -ø YO BO LE ‑dɛ, ‑rɛ RE -gó, ‑ŋó KO ‑bó, ‑mó BO ‑nɛ Ne -mɛ ‑lɛ -vɛ QE -yɛ, -Qɛ TO ‑dó, ‑ró, ‑tó TO -ntó

fish animal bark Lɔŋto du ‑rɛ / ‑do nam‑dɛ / nan‑to kɔ ɔ ‑rɛ / ‑ro Beiya du k‑sə / ‑tə Pana kɔ b‑sə / kɔ ŋ‑nə na m‑ze / na a m‑e Riiti du k‑se / ‑te Momi dúg-à / duuk-t na m‑ze / na a m‑e kɔ p‑se / ‑ti re rim‑z / rim‑i ‑k

load sinew kəə‑rɛ Lɔŋto dust ní‑rɛ / ní‑dó Beiya gə m-zə Riiti gə m-zə ku -se / ku -re Momi gəm-z kə -k (ak) ní‑s / ní‑r

• Rhotacism • intervocalic *s > z > r [d] • Class sɩ > -r. V Gurunsi : Deg, Phwie, Lyele. Oti Volta: Dagara, … • Spirantisation ?

Lɔŋto YO 1, 3 1 a 13 (? ) 14 10 23 2 2 a LE RE KO BO NE QE 6 TO 21 -nto 10 A -ø 5 15 BO -vɛ -mɛ

Nominalklasssen des Proto Gur Miehe, Kleinewillinghöfer, von Roncador, Winkelmann. ( 2007 ). *U, *a *ba *NU *ni *