Annotated Swadesh wordlists for the Daju group (East Sudanic family).
Languages included: Logorik [daj-log]; Caning (= Shatt) [daj-can]; Sila (=Dar Sila) [daj-sil]; Eref (= Dar Daju) [daj-erf]; Nyala [daj-nyl]; Lagowa [daj-lag]; Nyalgulgule [daj-nyg].
Thelwall 1981a = Thelwall, Robin. The Daju Language Group. Systematic Phonetics, Lexicostatistics and Lexical Reconstruction. School of Humanities of the New University of Ulster. D. Phil. // The single most detailed historical-comparative analysis of the Daju languages so far, based almost exclusively upon the author's own field data. Contains a list of Proto-Daju lexical reconstructions together with the comparative data from attested languages, as well as lexicostatistical wordlists.
Thelwall 1981b = Thelwall, Robin. Lexicostatistical Subgrouping and Lexical Reconstruction of the Daju Group. In: Proceedings of the First Nilo-Saharan Linguistics Colloquium, Leiden, September 8-10, 1980. Ed. by Thilo C. Schadeberg and M. Lionel Bender. Dordrecht/Cinnaminson: Foris Publications, pp. 167-184. // A companion piece to Thelwall 1981a, presenting the main conclusions of that work in condensed form, and also including a large comparative lexical wordlist for Daju languages that is slightly different from the data in Thelwall 1981a.
Jungraithmayr 1978 = Jungraithmayr, Hermann. A Lexical Comparison of Darfur and Wadai Daju. In: Aspects of Language in the Sudan. Ed. by Robin Thelwall. New University of Ulster, pp. 145-154. // This paper contains wordlists for several varieties of Daju, including Sila, Nyala, and Mongo, collected by the author. Unfortunately, the wordlists overlap only partially with the standard Swadesh list, making them unfit for use as primary sources.
Alamin Mubarak 2006 = Alamin Mubarak, Suzan. An initial description of Laggori noun morphology and noun phrase. In: Insights into Nilo-Saharan Language, History and Culture. Ed. by Al-Amin Abu-Manga, Leoma Gilley, & Anne Storch. Köln: Rüdiger Köppe Verlag, pp. 9-25. // A short paper on Laggori nominal morphology, based upon the author's own field research and containing a significant amount of illustrative lexical data.
Alamin 2013 = Alamin, Suzan. Verbs in Laggori. In: Nuba Mountain Language Studies. Ed. by Thilo C. Schadeberg and Roger M. Blench. Köln: Rüdiger, Köppe Verlag, pp. 451-462. // A paper on verbal morphology in Laggori. Contains a rather small amount of illustrative material, but is of assistance in understanding the internal segmentation of verbal forms attested in other sources.
Alfira et al. 2013 = Alfira, David Abbi; Kafi, Timothy Kuku; Stirtz, Tim et al. Caning-English Dictionary. Juba: Sudan Workshop Program. // A brief educational vocabulary of the Caning language.
Palayer 2011 = Palayer, Pierre. Grammaire du dadjo d'Eref (Tchad). Louvain/Paris: Peeters. // A comprehensive grammar of the Eref dialect of Dar Daju, well illustrated by lexical material and containing some text examples, but no separate vocabulary.
Ismail 2000 = Ismail, Muhammad Abbaker. Linguistic analysis of the noun morphology of the Daju language. M.A. Degree Thesis. University of Khartoum: Department of Linguistics. // A detailed study of the nominal system of Lagowa Daju, well illustrated by material and very useful as a control source for much of Thelwall's material.
Our main source of comparative lexical information on the various Daju languages of Sudan are the etymologies and Swadesh-type comparative wordlists collected by Robin Thelwall and published in his works, most importantly the PhD thesis [Thelwall 1981a]. Since this is not only the largest, but also the most cohesive and consistent set of data, entries from [Thelwall 1981a] are consistently considered for inclusion in our "primary slots", whereas material from additional (usually later) sources is listed in the notes section for comparison.
It must, however, be stated that this approach has one serious drawback: Daju lexical items frequently appear in Thelwall's works in different phonetic versions, usually without any explanation for the discrepancies. As a rule, the differences are slight (for instance, the word may be listed with or without a suffix; with a "fully articulated" vowel such as a or in a reduced variant with ǝ instead of a, etc.), but sometimes they even involve the appearance of different lexical equivalents for the same meaning. The most significant discrepancies are encountered between the so-called "preliminary" lexicostatistical wordlist and the so-called "refined" wordlist, printed next to each other in [Thelwall 1981a]. Although "preliminary" and "refined", as explained by the author, are supposed to relate not to the data itself, but to the way in which cognation is scored (purely based on phonetic similarity in the "preliminary" version; taking into consideration phonetic correspondences in the "refined" version), the data in these wordlists are actually different as well.
Our temporary solution for this issue is as follows: (a) where discrepancies are minor and not particularly significant for lexicostatistical calculations or even for automated phonetic comparison (e. g. the same word listed as wunet or wunett, or as aga or agǝ, etc.), we usually just list the "refined" variant as primary and mention all the other ones in the comments section; (b) where the discrepancies are significant (major phonetic dissimilarities or even completely different lexical equivalents), we go by the "majority rule" and list the equivalent that is most frequently encountered in all of Thelwall's works, as well as seek confirmation in our auxiliary sources. The less common (and possibly mistaken) alternate equivalent(s) is/are listed in the notes section. This seems to be the best strategy to be followed until more detailed and accurate dictionaries on individual Daju languages begin to appear.
An additional general source that can sometimes be used for control purposes is [Jungraithmayr 1978], where a comparative wordlist for several varieties of Daju, based on the author's own fieldwork, is presented; accuracy of notation seems to be at least as reliable as in the case of Thelwall, if not more, but, unfortunately, the wordlists are too short and too different from the standard Swadesh list to be selectable as primary sources. (This is why, in particular, it is impossible to complete a proper GLD wordlist for the Mongo variety of Daju).
Accurate morphological segmentation of Daju items is a problem, since its nominal and verbal morphology are rather complex, and quite a few formerly productive morphemes become fossilized in individual languages. To ensure the best results for both manual and automated comparison, we try to segment out fossilized morphemes where their status of suffixes or prefixes is made clear through external comparison with related languages.
The notation system of Robin Thelwall is largely retained as it is presented in the original sources. The predictable exception are coronal affricates and fricatives, recoded into UTS as follows: c > UTS ɕ, j > UTS ʓ, jʔ > UTS ʄ, ʃ > UTS š. Long vowels are marked as doubled (aa, ii, etc.) in Thelwall's works and have been recoded to UTS aː, iː, etc. accordingly. Daju languages are not tonal, but stress is distinctive; however, Thelwall usually does not mark stress in his data.
Individual sources by other authors on various Daju languages have their own specific notation, especially those sources that rely on specially elaborated "official" orthographies for particular languages. In particular:
(1) For Caning, the data in [Alfira et al. 2013] uses a specially designed alphabet, where:
'd = UTS ɗ; ng = UTS ŋ; ny = UTS ɲ; ä = UTS ǝ;
(2) For Logorik, Suzan Alamin [2006, 2013] uses ɟ to denote the implosive palatal (recoded to UTS ʄ); the rest of the symbols is the same as in Thelwall's works.
(3) For Daju of Eref, Palayer  mostly employs IPA symbols, but uses n̰ for ɲ.
Database compiled and annotated by: G. Starostin (last update: February 2016).