The database consists of several hundred entries from Hadza, tentatively linked to the "Macro-Khoisan" hypothesis. Since the genetic status of Hadza is at the present time debatable (it cannot be excluded that Hadza is, in fact, not Khoisan at all, but rather a language of entirely different affiliation with a substrate Khoisan layer), all the comparisons with Khoisan material from the other branches are highly hypothetical. The forms should be approached with caution, since a large percentage of them may probably constitute unidentified borrowings from neighbouring languages (Nilotic, Cushitic, and Bantu).
Data sources: as of now, there are no published dictionaries of Hadza. Most of the material has been collected from various articles and fieldnotes by Derek Elderkin (E.), Bonnie Sands (S.), and A. Tucker (Tc., TBW). A large number of Hadza entries can be found in Dorothea Bleek's "Bushman Dictonary" (B.), but the quality is not reliable.
The database consists of the following fields:
1. Proto-Hadza: a unified phonological transcription of the various transcription styles in available data sources. Hadza does not have much dialect differentiation, so the term "Proto-Hadza" is provisional.
2. Meaning: the general meaning of the Hadza stem.
3. Hadza: the data itself. In most cases, the data have been transliterated exactly the way they are found in original sources.
4. Notes: various notes and observations.
Notes on transcription: the "Proto-Hadza" entries follow some of the usual conventions for Khoisanistics, i. e.:
Clicks: | = dental click, ! = alveolar click (Hadza does not phonologically distinguish between alveolar ! and palatal ǂ), || = lateral click. Click effluxes include zero (no special marking), nasalisation (ɳ|, etc.), and the glottal stop (|ʔ, etc.).
Non-click consonants are transcribed as follows: ƛ = tl = voiceless lateral affricate; ʎ = ɫ = hl = voiceless lateral fricative; c, ʒ = hissing voiceless and voiced affricates; š, č (tš) = hushing fricative and affricate; the dot under the consonant marks glottalisation (ḳ, ƛ̣, etc.).
In the data field, conventions employed by researchers in their own transcription are usually respected. However, it should be noted that many of them use phonetic rather than phonological transcription (e. g., the numerous "prenasalized glottalized clicks", transcribed as ŋ|ʔ, etc., are phonologically just glottalized clicks, with a secondary nasalization feature).