Annotated Swadesh wordlists for the Jebel group (East Sudanic family).

Languages included: Aka [jeb-aka]; Molo [jeb-mol]; Kelo [jeb-kel]; Beni Sheko [jeb-bsh]; Gaam [jeb-gaa].


I. Aka, Molo, Kelo, Beni Sheko.

Main sources

Bender 1997 = Bender, Lionel M. The Eastern Jebel Languages of Sudan I: Phonology. In: Afrika und Übersee, 80, pp. 189-215. // General notes on the phonologies of four East Jebel lects (Aka, Kelo, Molo, Beni Sheko), including comparative 300-item wordlists collected by the author himself.

Evans-Pritchard 1932 = Evans-Pritchard, Edward E. Ethnological Observations in Dar Fung. In: Sudan Notes and Records, 15, 1, pp. 1-61. // This primarily ethnographical description of several populations in the Dar Fung region contains brief wordlists on a variety of minor languages, rife with phonetic inaccuracies but sometimes quite valuable from a historical point of view.

Additional sources

Bender 1989 = Bender, Lionel M. The Eastern Jebel Languages. In: Topics in Nilo-Saharan Linguistics. Ed. by M. Lionel Bender. Hamburg: Helmut Buske Verlag, pp. 151-179. // A brief comparative grammatical description of Gaam (Ingassana), Aka, Kelo, and Molo.

II. Gaam (= Gaahmg, Ingassana).

Main source

Bender & Malik 1980 = Lionel M. Bender, Malik Agaar Ayre. Prelimi\-nary Gaam-English-Gaam Dictionary. Carbondale: Southern Illinois Prin\-ting Service. // Comprehensive dictionary of the Gaam language, produced in collaboration with a native speaker, although the accuracy of phonetic notation and semantic glossing is sometimes questionable.

Additional sources

Stirtz 2012 = Stirtz, Timothy. A grammar of Gaahmg, a Nilo-Saharan language of Sudan. Utrecht: LOT. // Detailed grammatical description of Gaam, well illustrated with lexical and syntactic examples.

Verri 1955 = Sisto Verri, Di. Il Linguaggio degli Ingessana nell'Af\-rica Occidentale. In: Anthropos, 50, pp. 282-318. // Mostly of historical interest, this is nevertheless a very important source that collects most of the lexical and grammatical data made available on Gaam by different researchers over a period of about fifty years. Some of this data reveal important clues as to the phonological history of Gaam and other Jebel languages.


1. General.

Aka, Molo, Kelo, Beni Sheko.

These four idioms are spoken by small groups of people in Eastern Sudan, south of the Ingessana Hills, scattered among the local "jebels" (hills). The only more or less modern and generally reliable (or, at least, "usable") source on the lexical inventory of these languages are the wordlists in [Bender 1997], sufficient for performing lexicostatistical calculations and etymological research, though, as is often the case with Bender's data, some phonetic and semantic inaccuracies may be suspected. Some additional grammatical data are pulled from the earlier sketch [Bender 1989].

For a broader historical perspective and at least partial verification, we also make use of a much earlier source: the very brief wordlists, collected by the ethnologist E. Evans-Pritchard and published in the comparative survey [Evans-Pritchard 1932]. The idioms in this source are named after the geographical landmarks ("jebels") rather than the people themselves: Bender's Aka = Sillok, Molo = Malkan, and Kelo = Tornasi (data on the language of the small clan of Beni Sheko, which is most close to Kelo, was only collected by Bender). Evans-Pritchard's forms show surprisingly few discrepancies with Bender's data, collected much later, and are included in the notes section whenever they are available.


Gaam (= Gaahmg, Ingassana) is the largest and also the most far removed member of the small Jebel group. Since its speakers occupy a much larger territory than those of the small East Jebel languages, there is much more data available on Gaam, much of which dates back to the beginning of the 20th century and even before; most of the early sources have been diligently collected in [Verri 1955], and this data is also adduced by us in the comments section, where (Marno) = data from Ernst Marno's collection of 179 words, originally collected in 1869-73 and published in 1874; (De Pruyssenaere) = data from E. De Pruyssenaere's collection of 81 words, published as an appendix to his ethnographical report, originally published in 1877; (Seligman) = data from Brenda Z. Seligman's collection of about a hundred words, originally published in 1912; (Robertson) = data from J. W. Robertson's "Further Notes on the Ingessana Tribe", originally published in 1934. (For more detailed bibliographical information on all these sources, please refer to Verri 1955; in the database, all references are made exclusively to Verri 1955).

The main source of lexicostatistical data on Gaam should unquestionably be identified as the large and comprehensive dictionary by M. L. Bender and Malik Ayre [Bender & Malik 1980]. It is not very well illustrated with textual examples, and some of the semantic notation may be questioned, so that occasionally several equivalents have to be included as "technical synonyms" for lack of required information. As a reliable control source for both semantics and phonetics, we choose T. Stirtz's recent grammatical description of Gaam [Stirtz 2012], where many (though far from all) of Bender & Malik's words may be found supported with textual examples. It should be kept in mind, however, that the dialect described by Stirtz is less phonetically archaic than previously described dialects (for instance, it demonstrates the total deletion of intervocalic -ɬ-, a consonant that is essential for external comparison), so even if the data there were sufficient for a lexicostatistical database, it would still make sense to choose an older source as primary.

2. Transcription.

Aka, Molo, Kelo, Beni Sheko.

Bender's transcription mostly follows standard IPA conventions employed in Nilo-Saharan studies and requires only minor conversion to UTS (palatal affricates c > UTS ɕ, j > UTS ʓ). Long vowels are rendered by Bender as doubled symbols aa, ee, uu, etc., and are converted to UTS aː, eː, uː, etc. Additionally, for Kelo Bender sometimes uses a separate symbol ǝ̈ (distinct from simple ǝ) without explaining its articulation. If this is indeed a "centralized ǝ", it seems reasonable to render it with UTS ɜ (see Kelo 'dog', etc.). However, in most cases it seems to be a conditioned phonetic variant of simple e.


The same conventions as for Aka-Molo-Kelo also apply to re-transcribing Bender & Malik's Gaam material. The only addition is the standard UTS re-transcription of retroflex consonants , as ʈ, ɖ respectively. Also, the dictionary does not mark tones on a consistent basis: although three basic level tones are recognized (high, mid, and low), they are only marked for about half of the entries, presumably those in which they play a particularly important contrastive role. We preserve all tonal markers where they are present in the original dictionary, but there is nothing to be done about entries where tones were never marked in the first place.

The phonological analysis in [Stirtz 2012] differs from Bender & Malik's in several respects. Most importantly, B&M's "dental" t, d are marked as t̪, d̪, whereas B&M's "alveo-palatal retroflex" , are simply called "alveolar" and are marked as t, d. We did not introduce this unification, since all of Stirtz's data are only quoted in the comments section anyway. Other discrepancies involve phonology (e. g. Stirtz's Gaam informants all drop the lateral fricative -ɬ-; many consonants find themselves geminated in specific contexts where no gemination is indicated by Bender & Malik, etc.) and have even less reason to be unified.

The earlier sources, collected in [Verri 1955], use a variety of transcriptional conventions that have partially been unified by the author. The following re-transcriptions into UTS have been performed, based on the phonetic description in [Verri 1955: 288]: (a) interdental th > UTS θ; (b) palatal nasal ny > UTS ɲ; (c) velar nasal > UTS ŋ; (d) long vowels ā, ū, etc. > UTS aː, uː, etc. Digraphs such as gy or kh have been left unchanged, since their phonetic nature is unclear (gy may actually be palatal ʓ, and kh may be velar fricative x or uvular fricative χ, but nothing is certain). It should be noted that most of the transcriptions in the old sources are highly approximate and do not reflect an adequate understanding of Gaam phonology in general (e. g. none of them distinguish properly between dental and alveolar consonants, or identify the lateral fricative ɬ as a distinct phoneme).

Database compiled and annotated by: G. Starostin (last update: August 2015).