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

Languages included: Ama (Nyimang) [nyi-ama]; Afitti (Dinik) [nyi-afi].


Main sources

Stevenson 1957 = Stevenson, Roland C. A survey of the phonetics and grammatical structure of the Nuba Mountains languages, pp. 3-5. In: Afrika und Übersee, 41, pp. 27-65, 117-152, 171-196. // This publication, among other things, provides a brief, but informative sketch of the grammar of Nyimang and Dinik, well illustrated by lexical material.

Stevenson 2000 = Roland Stevenson's Nyimang and Dinik Lexicon. Ed. by M. Lionel Bender. In: Afrikanistische Arbeitspapiere, 63, pp. 103-120. // A list of more than 500 lexical items for Nyimang and Dinik, edited by M. Lionel Bender after Roland Stevenson's notes with some transcriptional unifications.

Additional sources

MacDiarmid & MacDiarmid 1931 = MacDiarmid, P. A.; MacDiarmid, D. N. The languages of the Nuba Mountains. In: Sudan notes and records, 14, pp. 149-162. // This publication includes a very brief comparative wordlist of about 20 basic items for Nyimang and Dinik.

Rilly 2009 = Rilly, Claude. Le Méroïtique et sa famille linguistique. Louvain-Paris-Dudley, MA: Peeters. // Claude Rilly's monograph on Meroitic and its potential relationship with East Sudanic contains an etymological appendix for 200 items that includes material collected by the author himself, including wordlists for Nyimang (Ama) and Afitti (Dinik).

Stevenson et al. 1992 = Stevenson, Roland C.; Rottland, Franz; Jakobi, Angelika. The verb in Nyimang and Dinik. In: Afrikanistische Arbeitspapiere, 32, pp. 5-64. // An important publication that deals with the complex verbal systems of Nyimang and Dinik. Includes a large appendix with detailed paradigmatic information on most of the verbs.

De Voogt 2009 = De Voogt, Alex. A sketch of Afitti phonology. In: Studies in African Linguistics, 38:1, pp. 35-52. // A brief sketch of the phonology of Afitti, based on original field data collected by the author; contains some illustrative lexical material.


1. General.

Ama (Nyimang) and Afitti (Dinik), two closely related small languages of the Nuba mountains, have not yet received a comprehensive descriptive treatment, but enough data have been published on both of them to allow lexicostatistical and etymological research (which places them, rather reliably, within the Eastern Sudanic family). The largest available collection of data belongs to Roland C. Stevenson; unfortunately, different publications and manuscripts posthumously edited for publication sometimes give different phonetic variants and semantic glossings. As our primary source, we choose [Stevenson 2000], a collection of more than 500 words for both languages that was edited for publication by M. L. Bender. Although Bender took a few permissible liberties with transliteration (see "Transcription" notes below), on the whole this is probably the most accurate and thorough publication of a single body of lexical data. Additional data, including grammatical notes, are also available in the general sketch [Stevenson 1957] and in [Stevenson et al. 1992], a detailed paper on the rather complicated verbal system of Nyimang languages which includes a large list of suppletive and non-suppletive variants of verbal stems.

More recently, research on Afitti has been conducted by A. de Voogt, who has published a short sketch on Afitti phonology [De Voogt 2009] that we took into consideration; and, most importantly, Claude Rilly, who has published his own field data in [Rilly 2009], a monograph on the East Sudanic family and the hypothetical East Sudanic origins of Meroitic. Rilly's list consists of approximately 200 words and occasionally, but not frequently, differs from Stevenson's. Some forms are also listed from the separate (phonetically, somewhat more archaic) Mandal dialect of Nyimang, which we include in the Notes section.

2. Transcription.

The system of transcription for Ama and Afitti data in [Stevenson 2000] is generally simple and requires very few UTS transliteration efforts. Some notes are, however, necessary.

(a) Stevenson's original publications distinguish between +/-ATR variants ı / i and ʋ / u, respectively. Even in Stevenson's original works, they were already analyzed as positional variants, and Bender, in his edition of Stevenson's material, merges them as i and u, respectively. Since this has no bearing on cognacy judgements, accuracy of correspondences, or the results of automatic comparison, we follow Bender in just transcribing i and u. However, when quoting additional forms or alternate variants from [Stevenson 1957], we retain the discriminating transcription of that source. (It should be noted that Steveson uses the symbol ɨ to denote -ATR ı; this has been changed to avoid confusion).

(b) Both Ama and Afitti distinguish between two series of coronal obstruents: dental (occasionally realised as interdental) and alveolar/retroflex, where the articulation may seemingly vary between "closer to alveolar" and "closer to retroflex". This phonetic variation means that different authors may use different transcriptional signs, depending on either their preferences or the particular pronunciation aspects of particular speakers. In [Stevenson 2000], Bender uses the symbols t_, d_ (we retranscribe them more correctly as , ) to denote dental consonants and diacritically unmarked t, d to denote alveolar/retroflex consonants. following Stevenson's practice (Stevenson used t̩, d̩̩ instead of t̪, d̪). In [De Voogt 2009], it is stated that Afitti does not have a phonemic contrast between dental and alveolar consonants, contra Stevenson, and so only t and d are used in that source; however, Rilly does distinguish between the two, so it may be assumed that the merger has occurred in some subdialects of the language, but not in others. Additionally, it must be noted that the different sources sometimes contradict each other in their choice of dental or alveolar to transcribe a particular word, and it is not always easy to understand which of the variants is more accurate. We have not tampered with such conflicting transcriptions.

(c) Palatal affricates are usually marked as c, j in most of the sources; according to UTS standards, we retranscribe them as ɕ and ʓ, respectively. Palatal nasal ny is retranscribed as ɲ.

(d) Both languages are tonal, and most researchers agree in distinguishing three tonal levels (high, mid, low). In [Stevenson 2000], tones are indicated numerically: 1 for low tone, 2 for high tone, mid-tone is either unmarked or indicated with a hyphen sign -. We have retranscribed this based on UTS standards as V̀, V́, and V̄ respectively.

Morphological segmentation of Ama and Afitti forms is somewhat tentative. Nominal stems usually have the structure CVCV, where the last vowel may be of suffixal origin (always or in most cases), but is frequently not segmentable on the synchronic level; our solution is to separate it with a hyphen whenever transparent evidence exists either on the synchronic level (e. g. existence of same-root cognates with a different final vowel) or on the group level (e. g. Ama and Afitti show different coda vowels, possibly reflecting old morphological variation). Verbal stems frequently have suffixal and/or prefixal components, such as the prefix t̪/V/= in definite stems; these are often identifiable, for instance, through systematic comparison of various suppletive paradigms, but sometimes have to be assumed on very flimsy evidence. Most of the disputable cases are commented upon in the notes section.

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