Chris Ehret's Nilo-Saharan language tree with Proto-Northern Sudanic dated at 10,500 BP
(Expand by clicking on the diagram)
[From Tishkoff et al Supplementary Material link, but also available in the above reference]
"The earliest point in the differentiation of the Nilo-Saharan languages to which some practice of food production can be traced is Proto-Northern Sudanic. For the next successive periods of Proto-Saharo-Sahelian and Proto-Sahelian, the volume of reconstructible vocabulary which is unambiguously indicative of food production continues to increase. It might be thought at first glance that the increase reflects simply greater retention of evidence because the latter two periods are more recent in time. But one key repeated feature of the evidence makes this explanation untenable: at each of the three successive periods, the newly identifiable evidence includes one or more coherent bodies of new vocabulary expressing a suite of practices or knowledge not found at all in the vocabulary reconstructible for the immediately preceding era.
"To the Proto-Northern Sudanic language can be attributed the following subsistence vocabulary (with known distributions of each in brackets):
1 *ndw 'to milk' [Kunama; Eastern Sudanic (Tama; Gama; Kuliak)]
2 *su:k 'to drive' (domestic animals)' [Kunama; Saharan; Eastern Sudanic (Nubian)]
3 *a:yr 'cow' [Kunama; Eastern Sudanic (Nara; Southern Nilotic of Kir group)]
4 *Way 'grain' [Kunama; For; Eastern Sudanic]
5 *ke:n 'ear of grain' [Kunama; Songay]
6 *p'εl 'grindstone' [Kunama; Eastern Sudanic (Western Nilotic or Kir group)]
"Of these six roots, only the first two are actually diagnostic of food production. Together with the third item, they indicate that proto-Northern Sudanians raised at least some cattle. The root for 'cow' is not diagnostic by itself of food production since cattle would have been known in wild form to early Nilo-Saharans if they lived far enough north in the modern Saharan zones; and the three grain terms would have been as necessary to the vocabulary of wild grain collectors as to that of the cultivators of domestic grains.
"Two other Proto-Northern Sudanic roots deserve notice for their archaeological implications:
7 *sa:p or *sa:B 'temporary shelter' [Kunama; Songay; Eastern Sudanic (Southern Nilotic)]
8 *ted 'to make a pot' [Kunama; For; Maban; Eastern Sudanic (Western Nilotic)]
"No word for any kind of more permanent structure, such as a house, can yet be reconstructed for the Proto-Northern Sudanic language or for the two earlier periods of Nilo-Saharan history. One additional root word applying to an earthenware container may be traceable back to the proto-Sudanic era, however. Its suggested reflexes are Kunama dosa 'earthenware bowl' and [a similar word in] Proto-Central Sudanic [meaning] 'water pot'. Its validity remains uncertain because its postulation requires some sound shifts which, though probably regular, are not yet fully substantiated. If it is a validly reconstructable root, it would show the invention of pottery among Nilo-Saharans to predate the Northen Sudanic period and the appearance of cattle-raising. An interesting possibility, in view of the meanings of the Kunama and Central Sudanic terms, is that the earliest ceramic vessels may not have been used for cooking."