Bʼaga languages

There are four to five Bʼaga languages. Grammatical forms are distinct between Northern Gumuz and Southern Gumuz.[5] Yaso is at least a divergent dialect, perhaps distinct enough to count as a separate language. Daatsʼiin, discovered in 2013, is closest to Southern Gumuz, while Kadallu in Sudan is attested by only two short word lists.

A comparative word list of Daatsʼiin, Northern Gumuz, and Southern Gumuz is available in Ahland & Kelly (2014).[6] The internal classification appears to be as follows:[7]

• Bʼaga (Gumuzic)
  • Dasʼin
  • Gumuz
    • Guba
    • Wenbera
    • Agelo Meti
    • Sira Abay
    • Eastern Gumuz
      • Yaso
      • North Gumuz
        • Metemma
        • Mandura
        • North Dibatʼe

Dimmendaal (2008) notes that mounting grammatical evidence has made the Nilo-Saharan proposal as a whole more sound since Greenberg proposed it in 1963, but that such evidence has not been forthcoming for Songhay, Koman, and Bʼaga/Gumuz: "very few of the more widespread nominal and verbal morphological markers of Nilo-Saharan are attested in the Coman languages plus Gumuz ... Their genetic status remains debatable, mainly due to lack of more extensive data." (2008:843) And later, "In summarizing the current state of knowledge, ... the following language families or phyla can be identified — ... Mande, Songhai, Ubangian, Kadu, and the Coman languages plus Gumuz." (2008:844)

This "Coman plus Gumuz" is what Greenberg (1963) had subsumed under Koman and what Bender (1989) had called Komuz, a broader family consisting of Gumuz and the Koman languages. However, Bender (2000) separated Gumuz as at least a distinct branch of Nilo-Saharan, and suggested that it might even be a language isolate. Dimmendaal (2000), who tentatively included Koman within Nilo-Saharan, excluded Gumuz as an isolate, as it did not share the tripartite singulative–collective–plurative number system characteristic of the rest of the Nilo-Saharan language families. Ahland (2010, 2012), however, reports that with better attestation, Gumuz does indeed appear to be Nilo-Saharan, and perhaps closest to Koman. It has grammatical forms that resemble what might be expected from an ancestral proto-Nilo-Saharan language. Gumuz may thus help elucidate the family, which is extremely diverse and has been difficult to substantiate.

Dimmendaal, Ahland & Jakobi (2019) summarize earlier work that the evidence "suggests that Gumuz and Koman may indeed form two subgroups within a broader 'Komuz' family" and that "there is some evidence that these two language families may indeed be part of a broader Nilo-Saharan phylum, albeit outliers in the family".[3]

• Gumuz word lists (Wiktionary)

• Ahland, Colleen Anne. "The Classification of Gumuz and Koman Languages", presented at the Language Isolates in Africa workshop, Lyons, December 4, 2010
• Lionel Bender, 2000. "Nilo-Saharan". In Bernd Heine and Derek Nurse (eds.), African Languages: An Introduction. Cambridge University Press.
• Dimmendaal, Gerrit J., 2000. "Number marking and noun categorization in Nilo- Saharan languages". Anthrolopological Linguistics 42:214–261.
• Gerrit Dimmendaal, 2008. "Language Ecology and Linguistic Diversity on the African Continent", Language and Linguistics Compass 2/5:842.

• video of Colleen Ahland speaking on the classification of Koman and Gumuz

• Gumuz basic lexicon at the Global Lexicostatistical Database