The etymological database for the Taa subgroup of the South Khoisan, or Taa-!Wi (!Wi-Taa), language family. In order to simplify things and also due to specific problems concerning the quality of data on !Wi, the Taa database is directly linked to the Peripheral Khoisan (Juu-Taa) database rather than to its intermediate ancestor (South Khoisan).

Data sources: the main and, in fact, only major source of data here is the excellent !Xóõ dictionary by Anthony Traill (Traill 1986). Considering the extreme complexity of !Xóõ phonology, it is not surprising that "Proto-Taa" essentially coincides with the corresponding !Xóõ forms (with but a few exceptions that are specifically noted in the database). Additional data on two Taa idioms - that of the Masarwa of Kakia and the |Nu//en - are extracted from Dorothea Bleek's materials (Bleek 1929 and Bleek 1956). They do not so much add to our knowledge of comparative Taa phonology as provide extra insight into the lexical history of this subgroup. The relatively poor quality of the transcription makes it hard to talk in terms of phonetic correspondences; for the most part, it seems that all the three analyzed idioms are quite close and the correspondences are mostly trivial, but this triviality is frequently made obscure by erroneous transcription of the forms (e. g. with a major confusion of both click influxes and effluxes).

The database consists of the following fields:
1. Proto-Taa: the hypothetical protoform, usually stripped of easily detachable grammatical morphemes and unified in terms of transcription; otherwise practically the same as the corresponding !Xóõ entry.
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 meaning of the protoform (most often, directly copied from Traill 1986).
4. !Xoong: the main entry in Traill 1986. All the forms are given together with the relevant grammatical information, following the system of notation employed by Traill in his dictionary. [For nouns, the Arabic numbers indicate the corresponding noun classes (from 1 to 4); the Roman numbers indicate the tonal class (from I to II). For additional details, please consult Traill 1986.]
5. Masarwa: the default sources are Bleek 1929/Bleek 1956.
6. |Nu//en: the default sources are Bleek 1929/Bleek 1956.
7. !Xoong derivates: a listing of additional !Xóõ forms derived from the main stem as presented in Traill 1986.
8. Notes: additional comments and considerations.
9. References: bibliographical links.

Notes on transcription:
For old data (Bleek 1956 and older), the original transliteration systems are followed. For data taken from Traill's dictionary, some of the usual transliterating conventions of ToB are followed (such as transcribing prevoiced clicks as ɡ|, etc., and nasalized ones as ɳ|, etc.); in most cases, though, Traill's transcriptional idiosyncrasies are respected, especially in the area of rendering complex vowels, since they represent a "semi-official" transcription system for !Xóõ. For Proto-Taa, however, all of the standard ToB conventions are in order.
Clicks: ʘ = labial click, | = dental click, ǂ = palatal click, ! = alveolar click, || = lateral click.
The Taa click efflux system is the most complex among all attested click systems and requires special attention. It includes the following:
- zero efflux (Proto-Taa and Traill: no special marking; Bleek: |k, etc.);
- voiced efflux (Proto-Taa/Traill: ɡ|, etc.; Bleek: |g, etc.);
- velar fricative efflux (Proto-Taa/Traill/Bleek: |x, etc.);
- prevoiced velar fricative efflux (Proto-Taa/Traill ɡ|x, etc.; Bleek ?);
- velar ejective affricate (Proto-Taa |kx, etc.; Traill |kxʔ, etc.; Bleek |kx, |k", etc.);
- prevoiced velar ejective affricate (Proto-Taa ɡ|kx, etc.; Traill ɡ|kxʔ, etc.; Bleek ?);
- uvular stop (Proto-Taa/Traill |q, etc.; Bleek ?);
- prevoiced uvular stop (Proto-Taa/Traill ɡ|q, etc.; Bleek ?);
- aspirated uvular stop (Proto-Taa/Traill |qh, etc.; Bleek ?);
- prevoiced aspirated uvular stop (Proto-Taa/Traill ɡ|qh, etc.; Bleek ?);
- prenasalized voiced aspirated uvular stop (Proto-Taa/Traill G|qh, etc.; Bleek ?);
- ejective uvular stop (Proto-Taa/Traill |qʔ, etc.; Bleek |k', ?);
- aspirated efflux (Proto-Taa/Traill |h, etc.; Bleek |kh, |h, etc.);
- voiceless nasalized efflux (Proto-Taa/Traill |n̥, etc.; Bleek ?);
- voiced nasalized efflux (Proto-Taa/Traill ɳ|, etc.; Bleek |n, etc.);
- preglottalized nasalized efflux (Proto-Taa/Traill ʔ|n, etc.; Bleek ?);
- glottal stop (Proto-Taa/Traill |ʔ, etc.; Bleek: no special marking).
[Note: question marks merely denote that Bleek does not recognize the existence of this particular type in the idioms she is describing, not the fact that there are no cognates in Masarwa and |Nu//en that correspond to items with these effluxes in !Xóõ and not even the fact that these effluxes did not exist in these idioms; they may have simply been unnoticed. Thus, words with uvular effluxes are usually treated by Bleek as if containing corresponding velar ones; preglottalized nasals are transcribed the same as regular nasals, etc.].
Non-click consonants: c, ch (= Traill's ts, tsh) = voiceless hissing unaspirated and aspirated affricates; ʒ, ʒh (= Traill's dz, dtsh) = voiced hissing unaspirated and aspirated affricates; q, G, qh, Gh (= Traill's q, G, qh, Gqh) = uvular stops; cʔ, kʔ, qʔ, etc. (= Traill's tsʔ, kʔ, qʔ, etc.) = ejective consonants; š = hushing fricative (only in Bleek's materials).
Vowels: ɛ, ɔ = open correlates to closed e, o. Pharyngealized vowels are marked as a̰, o̰, etc.; breathy vowels are marked with a superscript ^h (= Traill's regular h); nasalized vowels are marked with a tilde (ã, õ, etc.); á, à, ǎ, â = tonal markings.