Annotated Swadesh wordlists for the Sandawe group (Sandawe family).

Languages included: Sandawe [snd-snd].


Eaton & Hunziker 2007 = Eaton, Helen; Hunziker, Daniel; Hunziker, Elizabeth. A Sandawe Dialect Survey. SIL Electronic Survey Report, August 2007. SIL International. // A large paper presenting the results of a dialect survey of seven Sandawe villages, conducted by the authors in 2003. Contains a large wordlist elicited from a speaker in the Magambua village (complete wordlists for all the villages of the survey have not been published).

Eaton & Hunziker 2008 = Eaton, Helen; Hunziker, Daniel; Hunziker, Elizabeth. A Description of the Phonology of the Sandawe Language. SIL International. // Detailed description of Sandawe phonology; contains some lexical data unavailable in [Eaton & Hunziker 2007].

Eaton 2010 = Eaton, Helen. A Sandawe Grammar. SIL International. // A detailed grammar of the Sandawe language.

Kagaya 1993 = Kagaya, Ryohei. A Classified Vocabulary of the Sandawe Language. Tokyo: Institute for the Study of Languages. // Large classified vocabulary of Sandawe, based on the author's personal fieldwork conducted with a speaker from the Kurio village. Transcription quality is fairly good, but no tonal notation is available.

Dempwolff 1916 = Dempwolff, Otto. Die Sandawe. Hamburg: L. Friederichsen & Co. // One of the oldest works on the Sandawe, with a detailed ethnographic description, grammar, vocabulary, and text examples. Transcription quality is not thoroughly reliable, but the work remains an important historical source.

Tucker, Bryan & Woodburn 1977 = Tucker, Archibald; Bryan, Margaret; Woodburn, James. The East African Click Languages: a phonetic comparison. In: Zur Sprach\-ge\-schichte und Ethnohistorie in Afrika. Ed. by Wilhelm J. G. Möhlig, Franz Rottland & Bernd Heine, pp. 300-323. // A brief description of the phonetic structures of Sandawe, Hadza, and Dahalo, richly illustrated by lexical material.


1. General.

Despite the relatively "low profile" of Tanzania's Sandawe language, there are actually plenty of data sources today that allow for the construction of a comprehensive wordlist. Our primary source for the GLD wordlist is [Eaton & Hunziker 2007], since the source already contains a prepared wordlist and represents the results of recently performed, highly accurate, team fieldwork. However, for additional comparison we also employ the important dictionary [Kagaya 1993]; the fieldnotes from [Tucker, Bryan & Woodburn 1977]; and the historically important early source [Dempwolff 1916], since in between all of them, they contain certain minor, but important, variations.

Dialectal variety in Sandawe, according to most sources, is relatively low, and it does not seem to make sense to construct different wordlists in order to conduct internal Sandawe lexicostatistics (differences would at best revolve around 2-3%, and most of them would be questionable in terms of accuracy of semantic glossing anyway). However, it is significant to notate transcriptional - sometimes, perhaps, reflecting real phonetic - variability between the sources, which we tend to preserve in the UTS transliteration.

2. Transliteration.

The UTS system, being essentially based on the IPA with minor changes, is quite close to the completely IPA-based system of Helen Eaton et al.; transcription systems for Sandawe employed by other researchers generally require more transliterational efforts. It should also be noted that serious discrepancies are attested in various researchers' interpretation of the phonological structure of Sandawe (see notes below).

UTS Eaton/Hunziker Kagaya Tucker & Bryan Dempwolff
p p p p p
ph ph ph
b b b b b ~ b̬
m m m m m
w w w w w
t t t t t
th th th
d d d d d ~ d̬
n n n n n
r r r r r
č (= c) c ts ts
tsʼ tsʼ tsʼ tsʼ
čʰ (= cʰ) tʃʰ ch tsh ts
ǯ (= ʒ) j dz ~ z dz
ƛ tl tl
ƛʼ tɬʼ tlʼ tlʼ t̰ʼ
Ł dl dl
ɬ ɬ hl hl
y j y y y
k k k k k
kh kh kh
g g g g g ~ g̬
x x x x χ
ŋ ŋ ng ŋ
ʔ ʔ ʔ ʔ ʼ
ǀ ǀ ǀ ǀ ~ gǀ ǀ` ~ ǀk
ǀʼ ǀʼ ǀʔ ǀʼ ǀʼ
ǀʰ ǀʰ ǀh ǀh ǀh
ɳǀ nǀ ǀn ǀṅ
! ! ǂ ! ~ g! !` ~ !k
ǂh !h !h
ɳ! n! ǂn n! !ṅ
ǀǀ ǀǀ ǀǀ ǀǀ ~ gǀǀ ǀǀ` ~ ǀǀk
ǀǀʼ ǀǀʼ ǀǀʔ ǀǀʼ ǀǀʼ
ǀǀʰ ǀǀʰ ǀǀh ǀǀh ǀǀh
ɳǀǀ nǀǀ ǀǀn nǀǀ ǀǀṅ
a a a a a
e (= ɛ) e e e ~ ɛ e ~ e_
i (= ı) i i i ~ ı i ~ i_
o (= ɔ) o o o ~ ɔ o ~ o_
u (= ʋ) u u u ~ ʋ u ~ u_
Ṽ ~ Vŋ Vṅ


1. Aspirated stops and affricates are an integral part of Sandawe inventory, but are not systematically marked in [Dempwolff 1916], and other sources allow for some variation as well; we mark everything as originally transcribed in the sources.

2. The "voiced" and "semi-voiced" (marked with an additional subscript diacritic: , , etc.) consonants as marked by Dempwolff do not represent a valid phonological opposition (no other source confirms this), and the "semi-voiced" feature is ignored in transliteration (everything is simply marked as voiced). The same goes for Dempwolff's obviously fictitious opposition between "hard" and "pressed" articulation of the ejectives (the latter marked with an additional diacritic, e. g. vs. k̄ʼ), also not supported as a valid phonological or even phonetic opposition in later sources; it is safe to omit this transcriptional detail in order to avoid superfluous complexity.

3. There is no phonological opposition between alveolar and post-alveolar affricates in Sandawe, and all the sources fall in two distinct categories: those that consistently mark all the affricates as alveolar (c, , , ʒ; Dempwolff, Tucker & Bryan) or those that ascribe plain alveolar articulation only to the ejective affricate while at the same time defining the rest as post-alveolar (Eaton & Hunziker) or palatal (Kagaya). We do not unify this discrepancy in our transliteration, since it seems to reflect real dialectal peculiarities rather than individual tastes of the field data collectors.

4. According to most sources, Sandawe only has four distinct click effluxes: (a) velar = zero; (b) glottal stop; (c) aspiration; (d) nasalization. Additionally, Tucker & Bryan mention the possibility of a voiced efflux, but indicate that it is usually encountered in free variation with the zero efflux; we do not eliminate this notation in the transliterated forms, but it is never encountered in the main field (since most of the entries there are from Eaton & Hunziker). Likewise, Dempwolff tries to distinguish between the "explosive velar efflux" (ǀk, !k, etc.) and the "weak efflux" (ǀ`, !`, etc.); most likely, this is not a valid opposition either, but just in case, we retain this distinction in the transliteration (Dempwolff's ǀ` = UTS ǀ; Dempwolff's ǀk = UTS ǀk).

5. It seems that Sandawe does not have a +/-ATR phonological distinction, but some sources (Tucker & Bryan; Dempwolff) still try to mark phonetic variants; we do not omit them in our transliteration, but it should be noted that Dempwolff's transcriptions, in particular, are not highly reliable in this respect.

6. Additional vocalic features of Sandawe include vowel length (more or less consistently marked by everybody except for Dempwolff) and nasalization (which Kagaya and Dempwolff often interpret in consonantal terms - as combinations of vowels with velar ŋ). Kagaya also postulates a separate set of "creaky" vowels for the language that he marks as underlined a_, i_, etc.; we transliterate them as pharyngealized , , etc., but it must be noted that this extra feature has not been marked by any other researcher and may be fictitious.

7. The (semi-)vocalic labial element preceding the full root vowel is interpreted differently by different specialists: Eaton & Hunziker formally mark it as "click labialization" (e. g. ǀʷV, etc.); Tucker & Bryan, as well as Dempwolff, mark it as a separate glide phoneme w (e. g. ǀwV, etc.); and Kagaya often marks variation between glide and purely vocalic articulation (e. g. ǀwV ~ ǀoV, etc.). We preserve the individual styles, since they may actually represent phonetic peculiarities of the respective dialects.

8. Sandawe also has a small subset of word-medial / word-final reduced vowels, defined as "voiceless" by Eaton & Hunziker; these authors are the only ones to mark this feature on a more or less consistent basis, and we preserve their transcription of and in our transliterations.

9. The tonal structure of Sandawe has been studied in detail by Eaton & Hunziker [2008], and all the tonal notation of these authors has been preserved. Tone in Sandawe is also regularly marked by Tucker & Bryan and by Dempwolff, but not by Kagaya. There seems to be a general consensus about Sandawe possessing three basic register tones (high, mid, low) and a variety of contour tones, but individual notations frequently do not correlate in between different researchers.

Database compiled and annotated by: G. Starostin (last update: December 2012).