The etymological database for the !Wi subgroup of the South Khoisan, or Taa-!Wi (!Wi-Taa), language family. In order to simplify things and also due to specific problems concerning data quality, the !Wi database is directly linked to the Peripheral Khoisan (Juu-Taa) database rather than to its intermediate ancestor (South Khoisan).
This is by far the most problematic of all the low-level databases on Khoisan. Despite the fact that as late as about a hundred years ago islands of !Wi-speaking tribes were scattered all over South Africa, at the present time only the N|huki language has avoided complete extinction, and the absolute majority of our data stems from late XIXth - early to mid XXth century sources that are consistently unreliable in terms of data transcription quality. Before the fortuitous "rediscovery" of N|huki a few years ago, not a single language in this branch could be used as an "anchor" to tie the rest of the data to (the way that, e. g., Ju|'hoan functions for North Khoisan or Naro and Kxoe function for West Central Khoisan). As a result, reconstruction of proto-forms for !Wi is a very hard task, and most of the asterisked forms here - as well as some of the more questionable comparisons - have to be taken with a large grain of salt.
Data sources: the main source is Dorothea Bleek's dictionaries (Bleek 1929 and Bleek 1956), containing both her own and other researchers' flawed, but still priceless data on a dozen different !Wi idioms; some of the primary sources for Bleek 1956 have been considered as well. In addition, for //Xegwi some valuable data have been pulled from two articles by L. W. Lanham and D. P. Hallowes (mid-50s), and, finally, for N|huki we now have data of fairly good quality, recorded by N. Crawhall, B. Sands, and A. Miller-Ockhuizen, drawn from mostly unpublished or Internet-available sources (a dictionary by Sands and Miller-Ockhuizen is currently in preparation).
The database consists of the following fields:
1. Proto-!Wi: the hypothetical protoform, where possible, stripped of easily detachable grammatical morphemes. In many cases the protoform will look similar to the attested N|huki form, but this is really an approximation, since not all the correspondences have been worked out yet. For some information on the current state of research on Proto-!Wi, please consult G. Starostin's article "From Modern Khoisan Languages to Proto-Khoisan" in the "Articles and Books" section.
2. Stems: some of the forms are frequently met in conjunction with specific (class?) suffixes, sometimes two or more for one single root. In a few cases it was deemed useful to specify these combinations in a separate entry.
3. Meaning: the approximate meaning of the protoform.
4. |Xam: default sources are Bleek 1929/Bleek 1956 ("SI" in Bleek 1956). The forms marked (B.) denote Wilhelm Bleek's transcriptions, the ones marked (Ll.) refer to Lucy Lloyd's, exactly the way they are marked in Bleek 1956. Forms marked (Lich.) are from M. H. C. Lichtenstein's notes (1803-1806), also included in Bleek 1956.
5. //Ng: default sources are Bleek 1929/Bleek 1956 ("SII" in Bleek 1956), reflecting Dorothea Bleek's own transcriptions of the idiom.
6. N|huki: taken from Internet-published materials by N. Crawhall, B. Sands, and A. Miller-Ockhuizen (variants delimited by commas represent the pronunciation of different informants). N|huki is essentially the same language as Bleek's //Ng, but there are obviously dialectal distinctions between the two, which is why they are treated separately.
7. #Khomani ("SIIa" in Bleek 1956): this idiom, with data stemming from two separate descriptions by C. Doke (Dk.) and L. Meingard (Mg.), is also a very close variant of N|huki, yet treated separately here for historical and technical reasons.
8. //Kxau ("SIIb" in Bleek 1956): all the data are taken from a brief description by C. Meinhof (Meinhof 1929).
9. //Ku//e ("SIIc" in Bleek 1956): the only source is D. Bleek's own fieldnotes, included in Bleek 1956.
10. Seroa ("SIId" in Bleek 1956): represented exclusively by a very brief article by C. Wuras (Wuras 1920).
11. !Gã!ne ("SIIe" in Bleek 1956): all data are from Bleek 1956 (ultimately from an article by H. Anders).
12. //Xegwi (Batwa) ("SIII" in Bleek 1956): the default source is Bleek 1956 (D. Bleek's own notes), but most of the data from two articles by Lanham & Hallowes (LH) have been included as well. The latter are generally more credible, but, unfortunately, less numerous.
13. |Auni ("SIV" in Bleek 1956): the only source is Bleek 1956 (D. Bleek's own notes).
14. |Haasi ("SIVb" in Bleek 1956; the numeration implies close relationship between |Haasi and |Auni but, in fact, lexicostatistical analysis shows that |Auni has much more in common with //Xegwi, while |Haasi is the most distant member of the family): the default source is R. Story's manuscript, most (but not all) of which has been included in Bleek 1956 and which has recently (1999) been published with helpful notes by A. Traill.
15. Khatia ("SIVa" in Bleek 1956): the only source is Bleek 1956 (D. Bleek's own notes). Data are extremely scarce.
16. |Nusan ("SVIa" in Bleek 1956): the only source is J. G. Krönlein's manuscript dictionary included in Bleek 1956. For some reason, Bleek files this under SVI, implying closeness to |Nu//en; however, |Nu//en is an obvious member of the Taa subgroup, whereas data on |Nusan unambiguously shows it to be a rather close dialect of |Xam (SI).
17. Notes: additional comments and considerations.
18. References: bibliographical links.
Notes on transcription:
For old data (Bleek 1956 and older), for the most part, the original transliteration systems are followed (especially since in many cases, no particular references are given as to the exact meaning of the symbols). The data on N|huki are given with some of the usual transliterating conventions of ToB (such as transcribing prevoiced clicks as ɡ|, etc., and nasalized ones as ɳ|, etc.), but also reflect most of the peculiarities of the original phonetic transcription (especially where vocalism is concerned).
Clicks: ʘ = labial click, | = dental click, ǂ = palatal click, ! = alveolar click, || = lateral click. Click transcriptions with subscript indexes in the reconstructed Proto-!Wi forms (ǂ1, ǂ2) reflect specific rows of correspondences for which no phonetic interpretation is so far available.
The Proto-!Wi click efflux system is not yet fully established due to inadequate data (e. g., not a single "old" source on !Wi languages marks the presence of uvular click effluxes in those languages, even though they are quite prominent in N|huki and must have been present in Proto-!Wi as well). Effluxes that are more or less regularly - though often incorrectly - marked throughout the data include the following ones:
- zero efflux (Proto-!Wi and N|huki: no special marking; Bleek: |k, etc.);
- voiced efflux (Proto-!Wi and N|huki: ɡ|, etc.; Bleek: |g, etc.);
- velar fricative efflux (Proto-!Wi, N|huki, Bleek: |x, etc.);
- velar ejective affricate (Proto-!Wi |kx, etc.; N|huki |xʔ, etc.; Bleek |kx, |k", etc.);
- aspirated efflux (Proto-!Wi and N|huki |h, etc.; Bleek |kh, |h, etc.);
- nasalized efflux (Proto-!Wi and N|huki ɳ|, etc.; Bleek |n, etc.);
- glottal stop (Proto-!Wi |ʔ, etc.; N|huki |ʔ or ŋ̊|ʔ [denoting facultative prenasalization]; Bleek: no special marking or, in some manuscripts, |ʔ, etc.).
Uvular effluxes recorded for N|huki and superimposed on the Proto-!Wi level include:
- simple uvular efflux (|q, etc.; in "older" records usually not distinguished from the zero efflux);
- aspirated uvular stop (|qh, etc.; in "older" records usually not distinguished from the aspirated efflux);
- ejective uvular stop (|qʔ, etc.; in "older" records usually confused with the zero efflux, velar ejective affricate efflux or, sometimes, represented as an ejective velar stop - |k', etc.).
Non-click consonants: c, ch (= ts, tsh) = voiceless hissing unaspirated and aspirated affricates; ʒ, ʒh (= dz, dzh) = voiced hissing unaspirated and aspirated affricates; q, qh = uvular stops; ƛ, Ł = voiced and voiceless lateral affricates (in //Xegwi); cʔ, kʔ, qʔ, ƛʔ, etc. = ejective consonants; š = hushing fricative.
Vowels: ɛ, ɔ = open correlates to closed e, o. Pharyngealized vowels are marked as a̰, o̰, etc.; nasalized vowels are marked with a tilde (ã, õ, etc.; in some cases where diacritics are too abundant a superscript ŋ after the vowel is used instead); á, à, ǎ, â = tonal markings, used inconsistently in "older" materials and highly unreliable.