In the classification of African languages, Volta–Congo is a hypothetical major branch of the Niger–Congo family. Comparative linguistic research by John M. Stewart in the sixties and seventies helped establish the genetic unity of Volta–Congo and shed light on its internal structure, but the results remain tentative. Williamson and Blench (2000) note that in many cases it is difficult to draw clear lines between the branches of Volta–Congo and suggest that this might indicate the diversification of a dialect continuum rather than a clear separation of families. This had been suggested before by Bennet (1983 as cited in Williamson and Blench 2000:17) in the case of the Gur and Adamawa–Ubangi languages, which apart from Ubangian are now linked together as Savannas. Other branches are Kru, S