The Central Sudanic languages in the context of Nilo-Saharan: a new overview 13th Nilo-Saharan ɓonference University of Addis Ababa, 6th May, 2017 Roger Blench McDonald Institute for Archaeological Research, Cambridge Department of History, University of Jos Kay Williamson Educational Foundation Visiting Fellow, University of New England, Armidale Academic Visitor, Museu Emilio Goeldi, Belem, Brazil Background to Nilo-Saharan  Nilo-Saharan remains a controversial grouping and various claims have been made for excluding different branches, with Songhay a favourite.  Usually these claims are made without any examination of the evidence, simply by assertion  For what it is worth, I have gradually been compiling the evidence these last two decades with a view towards an eventual monograph  The broad conclusion is that Nilo-Saharan seems to be a coherent phylum, and the thirteen proposed branches are all included.  The map shows the distribution of Nilo-Saharan according to this model  The figure shows my current model of the internal structure of Nilo-Saharan  ? Excluding Shabo The Nilo-Saharan languages A proposal for the internal structure of Nilo-Saharan Proto-Nilo-Saharan Bertha Kunama Koman ɗumuz Shabo *ɓentral African Songhay Saharan Kuliak Maban ɖur ɓentral Sudanic Kadu Eastern Sudanic Where and when did Nilo-Saharan originate?  Nilo-Saharan is extremely fragmented and scattered across the African map  A proposal to account for this was made in 2010, suggesting that it is linked to the ‘green Sahara’  Prior to 12000 BP, the Sahara was no desert, but a region of lakes and rivers, rich in aquatic resources  A typical artefact of the period was the bone harpoon, used to hunt hippos and croccodiles  If we superimpose the map of finds of bone harpoons and the palaeo-rivers, the match with Nilo-Saharan is fairly good  So the proposal is that hunter-gatherers on the Ethio-Sudan borderlands took advange of this sudden increase in fish and huntable resources to spread westwards across Africa  And once the desert became arid, they became fragmented and isolated from one another Nilo-Saharan languages superimposed on the ‘green Sahara’ Background to Central Sudanic I The Central Sudanic languages are a large, complex subgroup of Nilo-Saharan stretching between Nigeria, western Chad and Northeast DRC and Uganda They were first identified by Joseph Greenberg (1963) and they have been broadly accepted since then Even sceptics about the remoter branches of Nilo-Saharan usually include Central Sudanic. It was formerly part of a larger subgroup, ‘Chari-Nile’, but this is no longer considered valid. Overviews of Nilo-Saharan have generally treated Central Sudanic as a unity (e.g. Bender 1997). The unity of Central Sudanic is usually accepted but the published evidence for this is thin. Pascale Boyeldieu & Pierre Nougayrol (2008) leave the question open, pointing out that a lexicostatistical evaluation falls to values as low as 10%, which is only just above chance. Background to Central Sudanic II The paper is intended to support the argument for the unity of Central Sudanic and to provide evidence for a proposed internal classification. The evidence is mainly lexical; grammatical descriptions of Central Sudanic languages are scattered and not every branch is covered by even a sketch, so identifying common grammatical structures at more than a basic level is a task for the future. Central Sudanic is notorious for some extreme phonology, especially in the Mangbetu and Lendu-Ngiti subgroups The family is usually divided into two major branches, East and West. The documentation of Birri is too weak to be sure of its position and it is provisionally given a branch of its own, pending further research. Formona-Sinyar is also sparsely documented and its position as a part of the Western branch must remain at best aguess. Branches of Central Sudanic The independent branches are; Sara-Bongo-Bagirmi (SBB) Kresh-Aja Birri ɖormona-Sinyar The independent branches are; Mangbutu-Efe Mangbetu-Asua Lendu-Ngiti Moru-Madi Of these, SBB is by far the most complex and ramified Central Sudanic languages Source: Updated from Boyeldieu (2004) Classification of Central Sudanic languages Proto-ɓentral Sudanic West Sara-Bongo -Bagirmi Sinyarɖormona East KreshAja Mangbetu-Asua Birri Mangbutu-Efe LenduNgiti Moru-Madi In the year 9040 Boyeldieu recalls a science fiction story, published in 1911, which envisaged Bagirmi, the language of a small state in Central Africa, becoming the lingua franca of Africa in the year 9040, when Europe has sunk beneath the waves (Van Gennep 1911). En l'année 2211, comme on sait, toute l'Europe, pivotant autour des Monts Oural comme autour d'une gigantesque charnière, s'affaissa de 880 mètres et disparut sous les flots […]. A la tête d'une de ces missions archéologiques fut placé le célèbre T.ɔ.B. Abdallah Sénoufo, professeur d'épigraphie comparée à l'université des Etats-Unis du Tchad […] Enfin Abdallah Sénoufo avait encore sur ses confrères l’avantage de connaître quelque peu de latin, et on lui devait une excellente traduction en néo-baguirmien des passages des auteurs romains qui traitaient de la géographie et de l'ethnographie des Alpes. This is still in the future, but it is pleasant to see an African language being given such high status in fiction. Acronyms for Central Sudanic branches Short form Expansion SBB Sara-Bongo-Bagirmi KA Kresh-Aja Birri ɖS ɖormona-Sinyar ME Mangbutu-Efe MA Mangbetu-Asua LN Lendu-Ngiti MM Moru-Madi Subgroups: Lendu-Ngiti Lendu and Ngiti are two related languages spoken in the Northeast DRC. Ngiti is well known from a comprehensive grammar (Kutsch-Lojenga 1993) and an unpublished dictionary. Material on Lendu (Bhadha ~ Bálɛ́dhá) was first published by Struck (1913) and a first grammar was published by Deleu (1934). The striking phonology of Lendu was first marked in the literature by Hertsens (1940) and then the subject of studies by Dimmendaal (1986) and Kutsch-Lojenga (1989, 1991). Trifkovic (1973) completed a thesis on Lendu grammar which has remained unpublished. Bokula & Irumu (1994) give data on Ndru, a southern Lendu dialect which may be a distinct language. Subgroups: Moru-Maɗi • Moru-Maɗi consists of very large number of lects spoken between northeast DRC, Uganda and Sudan. There are considered to be ten language clusters, but many of them are subdivided into marked dialects. The figure shows the internal structure of the Moru-Maɗi group. Moru-Madi South (Ma iOlubo) ɓentral (Avokaya, Lugbara, Logo etc.) North (Moru) Subgroups: Mangbetu-Asua Mangbetu-Asua (MA) consists of a group of four languages, Asua, Lombi, Mangbetu, Bendi, spoken in the northeast of the DRC. One of these, Bendi, is erroneously listed as Lendu-Ngiti in the Ethnologue. The literature on the group as a whole is very sparse, but Demolin (1992) includes a comparative wordlist of several Mangbetu dialects plus Lombi and Asua, though not Bendi, and also posits reconstructed forms. Almost all the published literature concerns Mangbetu, beginning with Autrique (1912), Vekens (1928), Larochette (1958), Demolin (1992) and Bokula & Irumu (1994). Subgroups: Mangbutu-Efe Mangbutu-Efe (ME) is a group of six languages spoken in northeast DRC and adjacent parts of Uganda. The first identification of this group appears to be in Van Geluwe (1957). Mamvu-Lese is the main language which has been studied in depth, particularly by Vorbichler (1965, 1971, 1974, 1979). A complete bibliography to 1993 is in Bokula & Irumu (1994). Subgroups: Birri The Birri [also Biri] language is only known from publications by Santandrea (1950, 1965-66). According to him, the language is correctly known as Vìrì. There is also a Bviri language, aligned with Sere, with which this Biri has sometimes been confused (e.g. in Ethnologue 2009). Given the chronic insecurity of the region and the declining status of the language during the 1960s, Biri may well be extinct today. Although there has been a tendency to class Birri with Sara-Bongo-Bagirmi, but the evidence for this is weak. Biri has cognates both with roots which spread across the whole of the family, but also cognates with both the Eastern and Western branches. For this reason it is given its own branch in the family tree. Subgroups: Kresh-Aja The Kresh-Aja (KA) languages are one of the smallest and least-known subgroups of Central Sudanic. They are spoken in two pockets south of Nyala in the extreme west of Sudan. Kresh is often known as Gbaya. The main source (and in some cases the only source) for this group is Santandrea (1976). Boyeldieu (2000a) had access to wordlists from Richard Brown and J-P. Caprile, cited in his appendix on the affiliation of Kresh. These indicate that KA is the nearest relative of SBB, but should not be included within it. Subgroups: Formona-Sinyar Formona-Sinyar [FS] is another poorly documented Central Sudanic subgroup, formerly spoken in Chad and Sudan. According to the Ethnologue (2016) the Sudanese population may well have crossed into Chad. Sinyar [=Shemya] has 5-10,000 speakers but the number of speakers of Formona is unknown. The main source on these two languages is Haaland (1978) which consists of comparative orthographic wordlists I understand new information on Sinyar willl become available Subgroups: Sara-Bongo-Bagirmi The Sara-Bongo-Bagirmi (SBB) languages constitute the single largest and most coherent subgroup of Central Sudanic. The first outline of the languages in the group appears in Tucker & Bryan (1956:10-19) SBB languages were the subject of a general monograph and reconstruction (Boyeldieu 2000), including substantial wordlists. Further papers and overviews can be found on Pascale Boyeldieu’s website The study of the Sara languages themselves has been revolutionised by the publication of material from the surveys of John Keegan. Apart from lexicons of many individual languages, Keegan has published comparative dictionaries and an overview of the family as a whole (Keegan 2016a,b,c) Classification of SBB languages Proto-SBB Western Barma West ModoBaka Bongo Kaba Na ɗulaɖongoro ɓentral East Source: Updated from Boyeldieu (2004) ɖer Yulu Location of SBB languages Lau Laka Source: Updated from Boyeldieu (2004) Bagirmi and Fongoro • The position of Bagirmi [=Barma] remains difficult to determine, but Keegan (2016d) has now published a dictionary and grammar sketch in modern transcription. The early material of Gaden (1909) and the grammar of Stevenson (1969). Keegan notes that many of Stevenson’s examples are not confirmed by his informants. • A language for which published data is very limited is Fongoro, a language of hunter-gatherers in Dar Fongoro, straddling the Chad-Sudan border (Doornbos & Bender 1983: 74). • The language is that of an outcast group and there were only a few speakers in the 1970s when the sparse data we have was collected. However, Fongoro is a close relative of the Gula languages described in Nougayrol (1999). Central Sudanic in Nigeria  A curious addendum to the map of SBB languages emerged in September 2016, when it was discovered by a student of Mark van der Velde that the ‘Laka Lau’ living on the Benue River in Nigeria also spoke a language of this group.  Their name previously misled classifiers to assume it was the same as the (Adamawa) Laka of Northern Cameroun, but a preliminary wordlist shows that it is related to the Kaba of Chad.  Surprisingly, although the basic lexicon is very similar, the speakers no longer have a clear idea of when and why they migrated.  They live in symbiosis with a group of Jukunoid speakers, formerly the Lau Habe, but now renamed Win Lau. Reconstructing Central Sudanic phonology • Despite some of the highly unusual consonants in some branches, notably the bilabial trills in MA and ME languages, it seems unlikely these are PCS. PCS probably had a quite small underlying inventory Proposed consonant inventory for PCS Labial Plosive p b Alveolar t d s z Retroflex Palatal Velar ʈ k g Labiovelar kp gb Implosive ɖricative Affricate Nasal Trills ʧ ʤ m n ŋ r Approximant Laterals ɲ y l w Reconstructing Central Sudanic phonology • • Proto-Central Sudanic certainly had ATR vowel harmony, like many other branches of Nilo-Saharan. Many languages are reported with nine vowels and so probably either nine or ten should be reconstructed. Inadequate transcription of branches such as Formona-Sinyar and Kresh-Aja makes the number of vowels uncertain for their particular subgroup. Proto-Central Sudanic vowels Front Close Mid Central i u ι υ e (ә) ε Open Back a o Reconstructing Central Sudanic tones  For those which have descriptions, Central Sudanic languages have predominantly three level tones and limited inventories of glide tones.  However, four tone heights have been reported for some SBB languages such as Yulu and Gula Koto (Boyeldieu & Nougayrol 2004) and two levels for Mangbetu and eastern SBB languages.  Boyeldieu (1989, 1998, 2000b) is a much more detailed discussion of tone for SBB languages. Tone-heights in Central Sudanic languages Group Languag e Level tones Comment Lendu-Ngiti Ngiti 3 Kutsch Lojenga (1994) Moru-Madi Madi 3 Andersen (1986) MangbetuAsua Mangbetu 2 ɔemolin (1992) Mamvu-Efe Mamvu 3 Vorbichler (1971) Kresh-Aja Kresh 3 Boyeldieu (2000) ɖormonaSinyar Sinyar 2 SBB Bongo 2 Nougayrol (2006) SBB Yulu 4 Boyeldieu (2000) SBB ɗula 3 Nougayrol (1999) SBB Sara 3 Boyeldieu (2000), Keegan (2016) plus downstep Reference Boyeldieu (2013) Reconstructing Central Sudanic morphology  The reconstruction of the sounds of Central Sudanic depends strongly on the model developed to understand its morphology.  Boyeldieu (2000) has used a consonantal template. In other words, the reconstructed form is notated as consonants with variable vowels.  Sometimes these are retained across Central Sudanic common lexemes, but in many cases the vowels seem to vary remarkably, from front to central and back and at different heights.  In other words, e.g. ‘breast’ the low central vowel is retained across the entire family.  Similarly, some languages have strong copying and harmonisation rules, so that the vowels in both root syllables are identical. Central Sudanic ‘breast’ #mbà breast Subgroup LN LN MM MA KA ɖS MM SBB SBB SBB SBB Language Lendu Ngiti PMM Mangbetu Kresh Sinyar Mödö Bongo ɗula Bagirmi proto-Sara Attestation ba ɨba, -bà *bà nɛbà pl. ɛbà mbāmbā mbàár mbà òmb(ә) mvà mbà *mbà Gloss breast breast breast sein sein breast breast breast sein sein sein Source RɓS ɓKL B&W96 ɔe92 Bo00 Bo13 PP Bo13 Bo00 Ke16 Ke12 Central Sudanic ‘ear’ #mbile ear Family Subgroup Language Attestation Gloss Source ɓS LN Lendu bɪ oreille ɔɔ ɓS LN Ngiti bɨ ear ɓKL ɓS MM *MM *bí, * í ear B&W96 ɓS MM Lugbara bílɛ ear B&W96 ɓS MA Mangbetu nɛbɪ pl. ɛbɪ oreille ɔe92 ɓS ME Mamvu úbí oreille ɔɔ ɓS KA Kresh mbímbí oreille Bo00 ɓS ɖS ɖormona amb to hear Ha78 ɓS SBB Lutos mvi ear KO ɓS SBB Modo mbɪlɪ ear PP ɓS SBB ɖongoro mvi ear ɔB83 ɓS SBB Bagirmi mbī oreille Ke16 ɓS SBB Proto-Sar *mbī oreille Ke16 Is there a special relationship with Niger-Congo?  A puzzling aspect of Central Sudanic is the number of shared features with Niger-Congo languages.  I previously claimed this was part of a genetic relationship, i.e. that Niger-Congo and Nilo-Saharan were related, i.e. ‘NigerSaharan’ and the Central Sudanic was co-ordinate with NigerCongo  I now think this cannot be true and that borrowing is more credible. But even so, this must have been at some level close to the proto-language.  Take the labial-velars, /kp/ and /gb/. These are not typical of Nilo-Saharan but are clearly reconstructible to Proto-Central Sudanic.  And some lexical items, but not systematic enough to be a genetic relationship Is there a special relationship with Niger-Congo? II  For example; #tarV to say Ph Family Subgroup Langua ge NS ɓS MM NS ɓS NS ɓS NS Attestati on Gloss Source Logo tà say B&W96 Birri ala speak Sa66 KA Kresh àdә dire Bo00 ɓS ɖS ɖormona adama say Ha78 NS ɓS SBB Bagirmi láà dire Ke16 NS ɓS SBB Ngamba y tàr word PWS tá- Nɓ W Reconstructing Central Sudanic history I  Historically speaking, the puzzle is to explain how Central Sudanic has become so territorially dispersed.  The core of its diversity appears to be west of Lake Victoria on the Uganda-Zaire border, although today most languages are far to the northwest, in the Sara area.  Fragmented population islands connect them, including some relatives of Sara-Bagirmi now spoken in the extreme southwest of Sudan.  Separating them today is the great eastward salient of Ubangian languages which extend from North Cameroun to Southern Sudan.  Bouquiaux & Thomas (1980) point out that the Ubangian expansion must have taken place north of the forest prior to the Bantu expansion and they assign it a tentative date of 35-4000 BP. Reconstructing Central Sudanic history II  The clearer picture now available of the structure of Central Sudanic shows that in what is now Central African Republic, the Ubangian expansion must have broken up a continuous chain of Central Sudanic languages from NE Zaire to the borders of Cameroun. The Sara languages subsequently underwent a secondary expansion north of Ubangian.  Boyeldieu (2000) uses the rich ethnohistorical data to accompany his linguistic reconstructions of Sara-BongoBagirmi, spoken between Chad, Sudan and Central African Republic.  He shows through a detailed reconstruction of the historical phonology of the group that their origin must lie on the border of Sudan, where only remnant languages are now spoken and that the much more extensive populations of Sara speakers in western Chad are secondary expansions. Where next?  As this overview should make clear, the density of documentation for individual branches varies greatly.  Many languages, especially small ones, exist in areas of civil insecurity where linguists dare not tread  And even if they did, populations may well have been displaced  Birri may be lost forever and Kresh-Aja is uncertain  Sinyar is live but Formona?  The SBB group is far better documented in Chad than CAR T H AN K S To numerous scholars who have provided unpublished or hard to get data over the years, including Don Killian, Harald Hammarstrom, Mark van der Velde, Pascal Boyeldieu, Andreas Joswig