As with biological diversity, linguistic diversity is unevenly distributed across the world. Despite many attempts from a variety of perspectives, we still do not know why certain regions are particularly suited for the emergence and maintenance of linguistic diversity. Nor do we fully understand the role played by context-specific environmental, social and cultural factors in any given ecoregion. In Africa, where modern humans and their languages once started to diverge, the Congo-Ubangi watershed is a major hotbed of linguistic, cultural, and genetic diversity spanning multiple ecozones within the Congo rainforest. Languages of two major African phyla, i.e. Bantu belonging to Niger-Congo and Central Sudanic of putative but unproven Nilo-Saharan affiliation, are interspersed with the geographically fragmented Ubangi groups, whose genealogy and putative Niger-Congo affiliation is highly debated. Featuring a deep history of human occupation, this critical region stands out in several respects. Shaped by major forest crises since the end of the African Humid Period (around 5000-4000 Before Present – henceforth BP) it is definitely not a refugial area where linguistic and biological diversity are expected to accrue. Further, it is characterized by deeply rooted exogamy, marriage alliances and long-lasting riverain trade networks. These features are virtually the opposite of those characterizing other African hotbeds of linguistic diversity such as the Nuba Mountains in Sudan. Despite the unique insights that could be gained from this region about language evolution and the deep human past, it remains severely understudied. Chief reasons include difficulty of access, lack of a local support network for expertise and infrastructure, and the astonishing intricacy of its linguistic and sociocultural configuration.
CongUbangi will realize a breakthrough in our understanding of how linguistic diversity correlates with material culture and genetic diversity and why this ecoregion was particularly suitable for the genesis and preservation of linguistic diversity. It will create a holistic analytical framework transferrable to other accretion zones in Africa and the wider world to investigate the coexistence of types of diversity.
The fact that the southern part of the Congo-Ubangi watershed is linguistically more uniformly Bantu than its northern counterpart where Bantu, Ubangi, and Central Sudanic are interspersed (see Map A) probably led to a first, deepseated
hypothesis devoid of archaeological and linguistic evidence according to which Ubangi and Central Sudanic speech communities gradually encroached into pre-existing Bantu speech communities, who reached the area from further north. The opposite scenario, whereby Bantu speakers represent a much later demographic stratum in the Congo-Ubangi watershed is equally if not more plausible based on the highly fragmented distribution of some Ubangi groups.
Linguistic research
At present, Ubangi is best understood as an areal group comprising at least the seven lineages shown in Map A. Of these, some scholars no longer consider Gbayaic and Zandic as Ubangi but rather as independent branches of likely Niger-Congo affiliation (Güldemann 2018b). Gbayaic would be lexically more similar to Central Gur Niger-Congo languages spoken thousands of kilometers away in West Africa than to any other Ubangi group (Moñino 2010). The geographically compact Bandaic, whose greatest diversity is found in the border area along the Ubangi River, is believed to be the result of Central Sudanic populations quickly adopting Ubangi varieties. Some exceptions notwithstanding, too little historical comparative work exists to make any claims at present about the internal relationships and genealogical affiliation of lineages characterized by considerable geographic fragmentation such as Mundu-Baka, Mbaic, Ngbandic and Ndogoic.
Numerous sources report Bantu speakers shifting to Ubangi and Central Sudanic languages, Ubangi speakers shifting to a Bantu language or to other allegedly related Ubangi varieties, as well as Central Sudanic speakers shifting to Ubangi. Nothing is known at present about the mechanisms regulating these multiple directions of shift, but available evidence suggests that linguistic and cultural assimilation can happen independently of one another. Nor do we have empirical evidence to prove that language shift truly happens in an unconstrained multidirectional fashion in this ecoregion. The first Research Objective (RO1) within linguistics is to understand the processes that regulate contact-induced change and language shift in the Congo-Ubangi watershed in order to determine whether these contributed, shaped or constrained the genesis of linguistic diversity (and its stratigraphy) in this ecoregion.
While language-shifting communities are a hallmark of the Congo-Ubangi watershed, there are also Ubangi linguistic enclaves such as Mbaic (see map A) which are geographically far removed from one another and surrounded by Bantu and/or Central Sudanic speakers. These isolated shift-resistant linguistic pockets are often considered as remnants of the earliest migration wave into the region. The Mbaic lineage hosts languages where pronominal gender systems coexist with Niger-
Congo-like nominal classification systems, and in one language also with a system of possessive classifiers. This combination of features is unique to this Ubangi lineage in the entire African continent. RO2 is to identify distinctive linguistic features of enclave varieties to determine whether these are also found in surrounding language groups. This will allow to establish whether refraction or continued exposure to language contact/shift (or a combination thereof) impacted their development in the enclaves.
(Ethno)archaeological research
The heterogeneous Congo-Ubangi ecoregion has favored highly diversified group specializations and material cultures
which appear to outrank ethnic and linguistic identities in this complex ecoregion. This configuration is likely to have considerable time depth, but there is currently little to no data on the spatial distribution of similarities and divergences in material cultures across this ecologically heterogenous region. Nor have there been any archaeological inquiries into
whether and how present-day material cultures relate to the past. Because subsistence strategies and/or material cultures appear to supersede linguistic and/or ethnic groups in the Congo-Ubangi watershed, RO3 is to document the spatial variability of cultural (dis)similarities in crafting and specialization strategies of groups targeted by linguistic research. This ethnoarchaeological approach enables our team to develop hypotheses about local-scale group mobility and historical connections among populations and connect the ethnographic present with the archaeological past through the identification of possible diagnostic material indicators such as pottery. If a group shifted to a different language in the past, the original linguistic origins of the group might still be visible in the technical vocabulary for conservative pottery shaping techniques which likely differ from those of the group to whose language they shifted but are identical to those of groups still speaking the ancestral language of the shifters.
The Congo-Ubangi watershed has a deep history of occupation as evidenced by several Early Stone Age sites with Acheulean and pre-Acheulean industries in the Central African Republic (CAR), particularly along the Sangha River, as well as Middle Stone Age sites featuring Sangoan and Lupemban industries between the Ubangi and Sangha Rivers. Despite the clear antiquity of human habitation throughout the Ubangi-Congo watershed, the earliest sites identified within the current extent of the forest zone belong to the Batalimo-Maluba pottery group and date approximately to 2200 BP. Nevertheless, it is a millennium after ceramics appear in the sequence from the not-too-distant site of Nangara-Komba in CAR. This incongruence is likely due to the dearth of archaeological fieldwork in the region. Only eight archaeological sequences from southern CAR, northern DRC, and western Cameroon – undergirding a significant portion of the Bantu expansion into the Congo Basin – have been radiocarbon-dated within the entirety of the northwest Congo Basin, an area of approximately 5,000 km2. This paucity of archaeological underpinnings has severely hampered prehistoric reconstructions of the region. Moreover, no attempt has ever been made to connect these past material cultures to the complex present-day mosaic, so that it is impossible to know whether, and to what degree, some of these past traditions have continuity into the present. Archaeological explorations and the projection of the ethnoarchaeological present into the archaeological past (RO4) will dramatically advance our limited understanding of this region’s settlement history, and shed new light on related conflicting hypotheses.
Genetic research
Available genetic analyses suggest significant genetic differentiation among populations also having distinct cultural and/or linguistic backgrounds in the Congo-Ubangi watershed. However, genetic sampling in this region is insufficient, especially amongst Ubangi and Central Sudanic speakers not considered to be foragers.
Interdisciplinary studies in other regions of Africa suggest that there is not a one-to-one correspondence between linguistic and genetic admixture. While in some cases visible linguistic traces of language contact or shift correlate with a substantial degree of genetic admixture genetic admixture does not leave any visible traces in the linguistic record, and instances where linguistic features attributable to language shift/contact do not appear to correlate with genetic admixture. Collecting new genetic data on groups which are undergoing or have undergone language contact, shift and/or a shift in material culture will allow us to determine for the first time in this ecoregion whether/how a given language-contact or language-shift scenario correlates with genetic admixture (RO5). Along with admixture patterns, the dating of genetic admixture will reveal the dynamics of early contacts between migrant groups in this region which might be otherwise undetectable in language, material cultures and/or subsistence strategies.
As is the case with language contact and shift, linguistic enclaves do not always correlate with genetic isolation. Genetic studies on linguistic enclaves offer important insights on the population history of enclaves and on whether languages spoken in a present-day enclave arrived there via demic expansion or cultural diffusion. The investigation of genetic isolation and endogamy patterns will also allow to establish whether linguistic enclaves are more endogamous compared to surrounding groups. New genetic data on linguistic enclaves will generate insights on whether and/or how the genetic makeup of speakers of enclave varieties might differ from those of speakers undergoing language contact and shift and on the correlations between the genetic structure and linguistic features of enclaves (RO6).