The position of Kam (Central-Eastern Nigeria) within Niger-Congo

And the overestimation of genealogical uniformity in African languages

Jakob Lesage (LLACAN, Paris)

jakob.lesage@gmail.com

+32479968126

ASIMIL, 03-Apr-2019, Leiden

1

African language

classification

• Since Greenberg (1963): four super-families (Afro-Asiatic, Niger-Congo, Nilo-Saharan, Khoisan)

• For decades, questions of uniformity and diversity preferred genealogical answers over areal patterns and contact scenarios (cf. Güldemann 2018b)

• This view neglects natural dynamicity and diversity of languages and cultures and gives an unrealistically static picture of the African past

• Contact scenarios are, however, everywhere in Africa (Güldemann 2018b: 454-472)

2

Reasons for this presentation

• Making comparative data available on an undescribed language

• To typologists: African languages are not adequately classified; be careful when sampling!

• Any feedback from traditions that have been more critical of Greenberg’s proposals and macro-comparative research?

• Case-study of language that is difficult to classify if the only model we want to use is genealogy

3

Niger-Congo classification

• Fairly well established as a unit (be it as a ‘domain’, ‘pool’, ‘phylum’ or ‘family’)

• Noun class morphology• Pronominal paradigms (Güldemann 2017)

• Numerals (Güldemann 2018a, Pozdniakov 2018)

• Some phylum-wide vocabulary (tongue, person, people)

• Sub-classification (41 low-level groups; 5 higher-level groups): different methods with different degrees of robustness

• B: Regular sound correspondences (+ morphology): 11 sub-subgroups, e.g. Bantu (which is actually a sub-sub-sub-subgroup)

• C: ‘Obvious’ relatedness (+ lexicostatistics, morphology): 22 sub-subgroups, e.g. Igboid

• D: ‘Scattered resemblances’ (mass comparison): 8 sub-subgroups and all 5 higher-level groups, e.g. Adamawa, Benue-Congo, Gur

Good (2017) 4

Kam (awɔm): an Adamawa (Niger Congo) language?

• 5,000-20,000 speakers, 20 villages

• Farmers-fishers-hunters

• Claim Jukun descent (kororofakingdom) – close contact (e.g. religion, Meek 1931b,c)

• Trade language: Hausa; Fulfulde major language since Fula jihad (early 19th century)

• Close contact with various other groups, e.g. Mumuye(‘Adamawa’), Jirim (‘Dakoid’), Chamba Leko (‘Adamawa’), recently Glavda (Chadic)

5

Kam classification

• Kam: claim Jukun origin

• Meek (1931c): not Jukun, but claimed resemblances with Cross River, Bendic, Kushi

Kam, Jirim

BendicCross River

Tula-Waja (Gur)

Kushi (Chadic)

Jukun

Hammarström et al (2018) 6

Kam classification

• Greenberg (1963): Adamawa (based on Meek’s 100 words)

• Adamawa-Gur continuum or genealogical/areal pool (Kleinewillinghöfer 2010)

• Little evidence for genealogical unity: high-level Niger-Congo isolate? (Güldemann 2018b; Hammarström et al. 2018)

GurAdamawa

7

Reevaluating Kam classification

• Sample of features:

• Unusual typology: phonology, logophoricity, serial verb constructions, clause-final negation, STAMP morphs (auxiliaries), plural words

• Noun structure: a- and N-prefixes

• Subject pronouns: 1SG, 2SG, 1PL, 2PL

• Lower numerals: 2, 3, 4, 5

• Basic lexicon: Leipzig-Jakarta list (Haspelmath & Tadmor 2009)

8

Reevaluating Kam classification

• Sample of languoids

• Proto-Niger-Congo: Güldemann (2017) for pronouns, Güldemann (2018a) for numerals and some basic vocabulary

• Potential genealogically distant contact languages? Chadic, Kanuri, Hausa, Adamawa Fulfulde → no obvious links at this point

• Geographically close Niger-Congo languages

• Jukunoid (Shimizu 1980, Storch 1999)

• Mumuyic (Shimizu 1979)

• ɲesam (Eveling Villa, p.c.) and Yendangic (Blench n.d. & Güldemann 2018a)

• Other Niger-Congo languages

• Edoid (Elugbe 1989)

• Potou-Akanic (Stewart 2002)

• Central Gur (Manessy 1979)

9

Comparative Kam 1 – Typology and the Macro-Sudan belt• Kam is areally and typologically part of a

linguistic Macro-area, i.e. the Macro-Sudan Belt (latest: Güldemann 2018b)• Labial-velar consonants (kp, gb)

• Nasal vowels (but rare)

• Logophoric pronouns

• STAMP morphs (TAM on pronouns)

• Clause-final negation

• Plural word

= chain of overlapping clusters with areal features; areal and not genealogical

Güldemann (2010)10

Comparative Kam 2: noun structure, nominal prefixes, and Central Jukunoid• 1/3 of noun stems in Kam have a prefix-like element

• a- (276), e.g. a-gbag ‘stool’; agbaŋ ‘chin’, abi ‘song’, atʃi ‘child’

• N- (108), e.g. mpuŋ ‘jug’; mfɔ ‘forest’; ntam ‘medicine’; ŋmkpɔŋ ‘agama lizard’

• No indications of noun class functions synchronically, but occurrence of a- is phonologically conditioned: non-postpausally → ɨ- or ø-

• cf. Central Jukunoid noun class prefixes, e.g. Jukun Takum (Shimizu 1980: 121-122); other Jukun (Storch 1999: 308-320)

• no other language in the sample has something similar.

11

Comparative Kam 3: subject pronouns and numerals → Mumuyic & Yendangic

1SG 2SG 1PL(.E) 2PL

Kam N-; ɲɨm a; ɲa juru ɲo

Proto-NC °mE °mO °TI NI

Mumuyic °mE; °N °mo °rO, wO °noO

Yendangic °m(E) °mO °tO °no

Jukunoid °-mi °-wu °-jE °-nE(n)

Potou-Akanic *-mE *wO/fO *jɛ *mO

Edoid *mhɛ n/a *-mhanhi *bha(dhi)

Central Gur °mE °mO °TI n/a

Note: N = homorganic nasal; I = close vowel; E = front vowel; O = back vowel; T = alveopalatal consonant (t, d, l, r); R = rhotic; C = unidentified consonant; ° = psuedo-reconstruction; * = cited reconstruction; cj = Central Jukunoid

two three four five

jirag tʃar/l/d n.nar/l/d ŋ.wun

°Ri °ta(C) °na(C) °nU

°zi.ti (?) °taa.ti °(d)nee.ti °maani

°iDNE(t) °taa.t °naa.t °nu.ŋ

°pan, cj °tar, cj °(-)NE.(n), cj °to.n, cj

*(-)ɲɔ *-ta *-na n/a

*-və *-cha.Gɪ *-niə *-chi.Nə.nhi

*le, °-nyo *ʈa, °sa *naː.(si), °nyi *nu, °-mU.wa

Note: analogical changes in numerals (Pozdniakov 2018): blurs regular sound correspondences, but at the same time provides other opportunities for comparison

12

Analogical changes in numerals (Pozdniakov 2018)

• Analogy = irregular phonetic changes of adjacent numerals which makes these numerals resemble each other

• Kam: 3-4: final consonant; 4-5: initial consonant

• 3-4 grouped together: also in other ‘Adamawa’ languages

• 4-5 grouped together: also in other ‘Adamawa’ languages

two three four five

jirag tʃar/l/d n.nar/l/d ŋ.wun

13

Comparative Kam 4: lexical comparison

• Collection of Leipzig-Jakarta list for sampled (proto-)languages

• Attempt to find comparative series → failed

• Superficial similarities (borrowing or inheritance with little change, or chance):

• Central Gur: 8/58 = 14%

• Mumuyic: 6/54 = 11%

• ɲesam (Yendangic): 10/90 = 11%

• Jukunoid: 7/79 = 9%

• Edoid: 4/59 = 7%

• (Potou-Akanic: 2/20 = 10%)

14

Comparative Kam 4: lexical comparison, shared across multiple languages (9)

bird bone breast dog mouth come grind water

Kam nuŋ a.kub a.vi wo we pɛ nam mɨŋ.kaŋa

Central Gur ? *ʔob, *ʔo ? *ɓa *nan, *no *ba *nam, *nem (*ni, na, ov)

Mumuyic *dʒaV-nuŋ ?*ka *min *za-V, *zaa *ɲa-V, *ɲaa *e ? *mi-V, *min

ɲesam ?bɛnuɛ kuɓi nwe bɛza ɲe inaŋ na mi

Jukunoid *ki-/li-nun *ku-/a-kup *ki-/a-bjan *i-bu *u-/i-ndut *ba *kwa *mbjed

Edoid *a-pɪ ? ? *a-bh ɞa *a-nuə ? *lɔ A-mɪN

Potou-Akanic

? ? ? ? *-nʊ ? ? *-dʒu

15

Comparative Kam 4: lexical comparison

• 12 words shared between Kam and one other language

• 76 words without any links so far (78%)

• Much more work has to be done on reconstruction and documentation (e.g. Yendangic), and probably more links – possibly with comparative series – will be found.

16

Comparative Kam 4: lexical comparison

One very interesting case of near-homonymy:

Detail that may only be explainable with common inheritance.

No other case of homonymy like this attested anywhere else in an African language (RefLex, Segerer & Flavier 2011; Pozdniakov, p.c.; Pozdniakov & Segerer under review)

bird eye

Kam nuŋ a.nuŋ

Mumuyic dʒaV-nuŋ nuŋ

17

Conclusions

• Introduction: genealogical explanations have traditionally been the default go-to models in Africa (and elsewhere?) for explaining convergence and divergence

• Both genealogy and contact are required to model language change and to paint a realistic picture of the (linguistic) past of any community

• The historical contexts for language contact are in fact much more natural than the contexts needed for ‘regular’ inheritance

• So there’s no problem for assuming contact scenarios, even if there is no historical evidence for it; at least no bigger problem than assuming inheritance scenarios

18

Conclusions• Contact is obvious in Kam on various layers

• Areal typological profile (Macro-Sudan Belt)• Nominal prefixes suggest a link with Jukun, but only in part of the vocabulary• Not mentioned: religious terminology and proper names often of Jukun origin• Not mentioned: imperfective aspect construction is very similar to Jukun• Vocabulary shared with different languages, sometimes extremely similar

• Genealogy?• Pronouns and numerals: Yendangic + Mumuyic; fits with Niger-Congo• Vocabulary: some shared vocabulary with geographically distant Gur, also

Yendangic & Mumuyic• But: no regular sound correspondences so far• Near-homonymy in Kam and Mumuyic: bird & eye

19

Some speculation

• Oral history: Kam encountered and conquered an unidentified ethnic group when they went to live in Kamajim

• Jukun elements and Mumuyic/Yendangic elements may suggest superstrate-substrate influence

• Possibly, some elements of Kam are the remnant of an absorbed earlier population

• Alternatively, Kam was the unidentified ethnic group (Mumuyic?), and the conquering involved ‘Jukunization’ of Kam in terms of religion, oral history, and part of the language

• Note that a specific group of Mumuye, Pugu, mysteriously claims Kam ancestry, and the Mumuye claim to come from a place they say was founded by the Kam, Yoro (Meek 1931a: 448-449)

20

Complex genealogy and contact in Central-Eastern Nigeria and West-Africa (Güldemann 2018b: 458)• Not just ‘meeting place of three major African phyla’ (Niger-Congo, Nilo-Saharan,

Afro-Asiatic)

• Majority of people belong to ‘minority groups’, incredible ethnic diversity and complex alliances that supersede simple categories such as ‘Christian vs. Muslim’, ‘Hausa vs. Jukun’, ‘Farmer vs. herdsman’, ‘tribe A vs. tribe B’. Ethnicity does not coincide with language affiliation.

• Language shift as a recurrent phenomenon

• Restructuring of many ‘colonizing’ Chadic languages as they came into contact with local languages.

• “Benue-Gongola-Chad Basin”: high language density of many related languages

• Identification of near-extinct language isolate Jalaa and its interaction with Waja-Jen languages (and these in turn with Chadic) (Kleinewillinghöfer 2001) points to high degree of genealogical diversity, as well as complex contact scenarios in the past.

21

kɔma, n tʃaŋaji

Jakob Lesage (LLACAN, Paris)

jakob.lesage@gmail.com

+32479968126

ASIMIL, 03-Apr-2019, Leiden

22

References (1)Blench, R. (no date). The Maya [Yendang] languages: Wordlists collected by Barau Kato and Zachariah Yoder.

Elugbe, B. O. (1989). Edoid. In J. T. Bendor-Samuel (Ed.), The Niger-Congo languages (pp. 291–304). Lanham: University Press of America.

Good, J. (2017). Niger-Congo Languages. In R. Hickey (Ed.), The Cambridge Handbook of Areal Linguistics (pp. 471–499). https://doi.org/10.1017/9781107279872.018

Greenberg, J. H. (1963). The Languages of Africa. Bloomington, Indiana: Indiana University Press.

Güldemann, T. (2010). Sprachraum and geography: Linguistic macro-areas in Africa. In Handbooks of Linguistics and Communication Science: Vol. 30. Language and space. An international handbook of linguistic variation, Vol. 2: Language mapping (Vol. 2, pp. 561–585). Berlin: Mouton de Gruyter.

Güldemann, T. (2017). A shared pronominal canon in the Macro-Sudan belt: Typological, areal and genealogical perspectives. In K. Raija & R. Kießling (Eds.), Mechthildian approaches to Afrikanistik: Advances in language based research in Africa, Festschrift fürMechthild Reh (pp. 101–146). Cologne: Rüdiger Köppe.

Güldemann, T. (2018a). 2. Historical linguistics and genealogical language classification in Africa. In T. Güldemann (Ed.), The Languages and Linguistics of Africa (pp. 58–444). https://doi.org/10.1515/9783110421668-002

Güldemann, T. (2018b). 3. Language contact and areal linguistics in Africa. In T. Güldemann (Ed.), The Languages and Linguistics of Africa (pp. 445–545). https://doi.org/10.1515/9783110421668-003

Hammarström, H., Forkel, R., & Haspelmath, M. (2018). Glottolog 3.3. Retrieved from http://glottolog.org23

References (2)Haspelmath, M., & Tadmor, U. (2009). Loanwords in the world’s languages: a comparative handbook. Berlin, Germany: De Gruyter Mouton.

Manessy, G. (1979). Contribution à la classification généalogique des langues voltaiques: Le proto-central. Paris: SELAF.

Meek, Charles Kingsley. (1931a). Tribal Studies in Northern Nigeria. London: Kegan Paul, Trench, Trübner.

Meek, Charles Kingsley. (1931b). A Sudanese Kingdom: An Ethnographical Study of the Jukun-speaking Peoples of Nigeria. London: Kegan Paul, Trench, Trubner & Co., Lts.

Meek, Charles Kingsley. (1931c). Tribal Studies in Northern Nigeria, Volume Two (Vols. 1–2). London: Kegan Paul, Trench, Trubner & Co., Lts.

Pozdniakov, K. (2018). The numeral system of Proto-Niger-Congo. Retrieved from http://www.oapen.org/download?type=document&docid=1001684

Segerer, G., & Flavier, S. (2011). RefLex: Reference Lexicon of Africa. Retrieved from http://reflex.cnrs.fr/

Shimizu, K. (1979). A comparative study of the Mumuye dialects (Nigeria). In Marburger Studien Zur Afrika- Und Asienkunde. Serie A, Afrika: Vol. 14. Berlin: Dietrich Reimer.

Shimizu, K. (1980). Comparative Jukunoid (Vols. 1–2). In Beiträge Zur Afrikanistik: Vol. 5, 6 (Vols. 1–2). Vienna: Afro-Pub.

Stewart, J. M. (2002). The potential of Proto-Potou-Akanic-Bantu as a pilot Proto-Niger-Congo, and the reconstructions updated. Journal of African Languages and Linguistics, 23(2), 197–224.

Storch, A. (1999). Das Hone und seine Stellung im Zentral-Jukunoid. In Westafrikanische Studien: Vol. 20. Köln: Rudiger Köppe. 24