Not attested. Possibly no generic term at all in the language.
ǀǀXegwi:
Not attested. There is a gloss ǀxʼuri 'seeds' in [Bleek 1929: 72]; however, it is not confirmed in [Bleek 1956] or any other, more reliable sources. Even if the entry is correct, it is in all certainty a borrowing from Central Khoisan (Proto-Non-Khoekhoe *ǀxuri 'seed'), and should be excluded from lexicostatistical calculations all the same.
Bleek 1937: 207; Bleek 1956: 220. Meaning glossed as 'pips, seeds'. Cf.: ǂwi cʼou "seeds of tsamma" [Bleek 1956: 220]. Also attested as a compound: ǂu-cʼo 'seeds, pips' [Bleek 1937: 219], although the first component is unclear. Another possible candidate is ǀǀwai [Bleek 1937: 216]; cf. n ǀʼʰõ ǀǀwai si-ǂʼa "I pick up seeds into the kaross" [Bleek 1956: 596].
ǀHaasi:
Not attested.
Proto-!Wi:
Not reconstructible due to lack of proper attestation. The form in ǀʼAuni may be etymologically connected with the Common !Kwi compound stem for 'eye' q.v.; if its internal etymology as 'seed+face' is correct, then cʼou is an archaism and may well represent the old equivalent for 'seed', but there is no direct comparative evidence to confirm that.
Bleek 1956: 171, 181. Transcribed as so ~ sːoː ~ sːʼo ~ šːoː by W. Bleek. Quoted as sʼoː ~ šʼoː in [Bleek 1929: 76]. This is the stative verb ('to sit, be sitting'); the dynamic action verb ('to sit down') is expressed by an apparently derivative stem: sːoéːŋ ~ sːo̯eŋ ~ sːoéːnya ~ sːwéŋ (W. Bleek), sːueːŋ ~ sːueːnya ~ sːueŋ-sːueŋ (L. Lloyd) [Bleek 1956: 172, 173, 175].
Bleek 1956: 171; Bleek 2000: 20. Quoted as sʼɔː in [Bleek 1929: 76]. This is the stative verb ('to sit, be sitting'); the dynamic action verb ('to sit down') is expressed by an apparently derivative stem: soeːŋ ~ soẽya ~ swünya [Bleek 1956: 172, 176].
Maingard 1937: 252. As in other cases, differently vocalized variants may reflect mergers with class markers or other types of suffixation. Transcribed as sôu in [Doke 1936: 63].
Ziervogel 1955: 40, 51. The simple form, according to Ziervogel, is used as the past or future tense; the present tense stem is šoge-ŋe [ibid.]. Possibly the same stem is quoted as šoɢaʔane 'sit and wait for me' in [Lanham & Hallowes 1956: 116], where it is analyzed as šo 'sit' + ɢaʔa 'wait' + ne 'me'. Quoted as šoː in [Bleek 1929: 76]; as šoː ~ šo ~ čo in [Bleek 1956: 181, 232].
Bleek 1937: 205, 206; Bleek 1956: 161, 164, 171. Meaning glossed as 'to sit down, put down, set, stay', but textual examples confirm that this is also the basic equivalent for the static verb 'to sit'. In [Bleek 1929: 76], the form sã is listed in the meaning 'to sit', and a strange form !ʼʰowa, not confirmed in later sources, in the meaning 'to sit down'. Should be distinguished from !ãũ 'to sit, to squat' [Bleek 1937: 213].
Story 1999: 23. Secondary synonym: !xi-kʼi [ibid.].
Proto-!Wi:*soʔ-
Distribution: Preserved in all daughter languages. Reconstruction shape: The first consonant is clearly *s- (compare with *cʼi 'bite' q.v. to make certain that reconstructing an affricate here is out of the question). However, in order to account for the glottal stop in ǀXam and ǀǀNg!ke, as well as the glottalized affricate in ǀHaasi, it has to be presumed that glottalic articulation was at least defined on the original vowel (i.e. ǀXam sːʼo < *soʔV, etc.). The basic root seems to frequently interact with verbal suffixes, e. g. *soʔ-ĩ (in Nǀuu > sõẽ or sũĩ with vocalic assimilation), *soʔ-ã (in ǀʼAuni), *soʔ-i (in ǀHaasi > cʼi).
Bleek 1956: 240. Transcribed as tːũ, emphatic form tːũŋ ~ tːũwaŋ by W. Bleek. Quoted as tũː in [Bleek 1929: 76]. The latter source also quotes a secondary synonym: ɡǀǀɔ, glossed, however, in [Bleek 1956: 531] as 'outer skin' and encountered in application to a snake's shed skin and the skin of a dead person. Clearly not the basic word for 'skin' in the light of both internal and external evidence.
Bleek 1956: 26, 240, 243. The form diõ is somewhat dubious and may represent a different lexeme. Quoted as tũː in [Bleek 1929: 76]. The same source lists ǀou as a synonym; in [Bleek 1956: 321] the word is glossed as 'skin, leather', with one textual example supporting only the second meaning (ŋ ǀǀʼayan ǀou !xí "I bring a skin dress").
Distribution: Preserved in Narrow !Kwi. Replacements: Narrow !Wi *tuŋ is opposed to Lower Nǂossob *ǀǀʼu (reconstruction relies on ǀʼAuni rather than ǀHaasi), but external comparanda in Taa languages (!Xóõ tùˤm, etc.) show that the Narrow !Kwi variant is more semantically archaic. The Lower Nǂossob variant finds an interesting etymological parallel in Narrow !Kwi; ǀXam ɡǀǀɔ 'outer skin', 'shed skin', allowing to suggest a possible replacement {'outer skin / shed skin' > 'skin (gen.')}, although the data are clearly insufficient to secure this conclusion. Reconstruction shape: The nasal coda seems to be retained in ǀǀXegwi, but as a rule, dissipates into vowel nasalization in the other languages (*tuŋ > tũ). Palatalisation in Nǀuu is regular, but the reasons for voicing of the initial consonant remain unclear - an issue to be investigated further, possibly affecting the phonological status of the reconstructed segment.
Bleek 1956: 684, 686. Transcribed as ʘoeŋ ~ ʘoeŋ-ya ~ ʘoe-ʘoeŋ by W. Bleek. Quoted as ʘoen in [Bleek 1929: 76]. There also exists a secondary synonym: ǀʼũːŋ (L. Lloyd, W. Bleek) [Bleek 1956: 359], quoted as ǀʼũŋ in [Bleek 1929: 76]; its exact meaning is hard to determine based on attested examples, but it is likely to have been the dynamic action verb 'to lie down (to sleep)' (cf. such a diagnostic example as heː si ǀʼũːŋ ĩː ... ŋ a xʼauki ʘwoiŋ "then we lay down... I did not sleep" [Bleek 1956: 359]).
Bleek 1956: 684, 686; Bleek 1929: 76. In D. Bleek's records, the verb kia 'to lie' q.v. is also very frequently glossed as 'to sleep' (e. g. within a complete transcribed "paradigm" ŋ kiaː, a kiaː, etc., translated as "I sleep, thou sleepest, etc." [Bleek 1956: 91]). However, it may be assumed that the primary meaning of kia is 'to lie', based on (a) external evidence and, more importantly, (b) the fact that only for kia we also have contexts such as ǀwi ǀʼi kia ha ɳǀǀai "the bird stays (= lies) in her nest" (certainly not *'sleeps') [ibid.]. Nevertheless, there must have been some actual overlapping between the two concepts, and it is possible that in this particular dialect the old etymon ʘoiŋ was being gradually replaced by kia.
Ziervogel 1955: 36, 39, 52. The simple stem, according to Ziervogel, is only used in the future tense. The present tense stem is ʘi-ɲe; the past tense stem is ʘi-ɲa [Ziervogel 1955: 52]. Quoted as ʘweːni ~ ʘweːŋ ~ ʘwen in [Bleek 1956: 685, 687] (with the same "special" type of labial click that D. Bleek perceives in the word 'meat' q.v.).
Bleek 1937: 220; Bleek 1956: 686. Quoted as ʘwoiŋ in [Bleek 1929: 76]. The latter source also adds kia as a synonym, but this word is rather the basic ǀʼAuni equivalent for 'to lie' q.v.
Story 1999: 23. Transcribed with a space (ʘwa ai), but must be a single word (perhaps, ʘwaʔai).
Proto-!Wi:*ʘu- ~ *ʘi-
Distribution: Preserved in all daughter languages. Reconstruction shape: The only certain thing about this reconstruction is the initial click (labial, with zero/velar release; glottal stop in old ǂKhomani records is most likely erroneous, since it is not confirmed anywhere else). Vocalism and presence/absence of nasalization (either as an additional vocalic feature or as a consonant) are heavily influenced by the verbal nature of the stem, which seems to be encountered with a variety of suffixes. Although nasality is a very common feature here (and is further confirmed by external data), it is interesting to note that in ǀǀXegwi as well as in ǀHaasi, variants without nasalization are attested, and it is not highly likely that nasalization could have been secondarily lost in these languages.
Bleek 1956: 643. Plural form: ǂʼénː (on a formal synchronic basis, the singular stem is derived from the plural one). Transcribed as ǂʼer̃r̃iː ~ ǂʼér̃r̃i, pl. ǂʼenn by W. Bleek. Quoted as ǂʼẽri in [Bleek 1929: 76]. This seems to be the least strongly marked equivalent for 'small' in ǀXam. Two more forms are listed in [Bleek 1929: 76]: (a) cʼǝre = sg. cːérre ~ cérre, pl. cːétten (L. Lloyd), cʼéttǝn ~ cʼɛ́ttǝn (W. Bleek) [Bleek 1956: 215]. This lexeme is said to be exclusively used in conjunction with the negative (xʼauki cerre 'not small'), and all textual examples confirm this; (b) -ʘwa = ʘu̯á [Bleek 1956: 684]; this is a diminutive morpheme, mostly found in conjunction with words for 'children' or 'young of birds/animals' (toi ʘwa 'little ostrich', etc.).
Bleek 1956: 652; Bleek 1929: 76. The secondary synonym ʘwain-ki [Bleek 1956: 685] ~ ʘwoin-ki [Bleek 1929: 76] is more rare and most probably applied only to young animals (cf. the given example mirriŋ ʘwain-ki "little goats").
Maingard 1937: 256. The efflux is marked with both a glottal stop and the velar obstruent k, whatever that might mean according to Maingard's notation. Cf. ʘónē 'small' in [Doke 1936: 73] (most probably the same root, but with a different click efflux and an extra suffix at the end).
Sands et al. 2006. Meaning glossed as 'small, few'. Secondary synonym: ʘũ (primarily used to form diminutives, as in ǀXam and other !Wi-Taa languages).
Ziervogel 1955: 58. Quoted as ƛʼini 'be small' in [Lanham & Hallowes 1956: 98]. The latter source also adduces a secondary synonym: ʘari [Lanham & Hallowes 1956: 104]. Quoted as ǀǀxʼeni in [Bleek 1929: 76]; as ǀǀxʼeːni in [Bleek 1956: 601].
Bleek 1929: 76. Only attested in this early, not very reliable, source, so the entry is dubious.
ǀHaasi:nʸái-si4
Story 1999: 23.
Proto-!Wi:*ǂʼeni
Distribution: Preserved in most of the Narrow !Wi dialects, but seems to have been largely lost in Lower Nǂossob. Replacements: (a) The situation with Lower Nǂossob languages is unclear, since the ǀʼAuni equivalent is only attested in an early dubious source, and the ǀHaasi entry, with its initial palatal nasal, is phonologically odd and looks like an "expressive" term; (b) ǂKhomani ʘʼkõ in Maingard's vocabulary is to be compared with a whole series of words meaning 'child, son, daughter; young; little, small' in !Wi languages [Bleek 1956: 682-687]; however, as a rule, this word is never used to denote objects that are small in size, so either the meaning in ǂKhomani was inaccurately glossed or there was a minor semantic shift in that dialect of Nǀuu ('young, little /of age/' > 'small /in size/'). Reconstruction shape: Initial palatal click is well preserved in ǀXam and Nǀuu and regularly shifts to a lateral affricate in ǀǀXegwi. Of particular note is the fact that in ǀXam, the palatal click is consistently marked as such, rather than as an alveolar click (cf. 'dog', etc.). The implications of this for the reconstruction are not yet clear (separate click phoneme? special positional development? result of recent borrowing?). Bisyllabic nature of the word is well preserved in ǀǀXegwi and is reflected as vocalic nasalization in Nǀuu.
Bleek 1956: 533. Quoted as ɡǀǀɔ̀ in [Bleek 1929: 77]. Somewhat dubious (only attested in the compound form ɡǀǀɔ́ː-ka !wi "smoke's man" (?)), but seems to be confirmed by external evidence.
Story 1999: 23. Said to denote both noun and verb, although the only textual example is within the phrase 'I smoke'.
Proto-!Wi:*ǀǀoˤ
Distribution: This is an isogloss between ǀXam, Nǀuu, and ǀHaasi, and, consequently, the optimal candidate for 'smoke' in Proto-!Wi. Replacements: ǀǀNg!ke ǀwiː, ǀǀXegwi kʰaʔa-zi, and ǀʼAuni ɳǀǀaːlu currently show no external parallels that would explain their origins. Reconstruction shape: We tentatively choose the Nǀuu form as the most conservative, simply because of the accuracy of transcription. It does not correlate well with such features as voiced articulation of the click in ǀXam, additional glottal stop in Doke's ǂKhomani, or the diphthong -au in ǀHaasi; however, we are not sure as to whether at least some of these features have not simply been inaccurately transcribed.
Bleek 1956: 425. Polysemy: 'stand / remain / stop'. Transcribed as !ʰeː ~ !ʰe by W. Bleek. Quoted as !ʰeː in [Bleek 1929: 79]. The verb is quite distinct from the dynamic 'to stand up': kːóaˤŋ (L. Lloyd), kóaŋ ~ kːóːaŋ (W. Bleek) [Bleek 1956: 97].
Bleek 1956: 611. Quoted as ɳǀǀá in [Bleek 1929: 79]. Slightly dubious; the main meaning of this verbal stem in ǀXam is 'to stay, dwell, be somewhere', and many particular contexts in ǀǀNg!ke display the same semantics, cf.: hn ɳǀǀa hn einki "they stay with their father", ha ɳǀǀá ɡ!au "he is at the door" [Bleek 1956: 611]. However, cf. also: ŋ !a ɳǀǀa "I stand" [ibid.]. Also, no serious competition for this stem is attested, with the potential weak exception of !uŋ [Bleek 1956: 453]. Overall, it must be stated that no truly diagnostic contexts, in which it would be transparently clear that the Swadesh meaning is present, are attested in Bleek's corpus of examples.
Ziervogel 1955: 48, 51. The simple stem, according to Ziervogel, is used as the past or future tense; the present tense stem is !ʔoʔo-ge-ne. The imperative forms glossed as: !ʔoʔo-geŋa (sg.), !ʔoʔo-u "stand ye!". The same word is possibly quoted as ǀǀɔːro in [Bleek 1929: 79], although it is not confirmed in [Bleek 1956].
Bleek 1937: 212. Meaning glossed as: 'up, to stand'. The same word is quoted in [Bleek 1929: 79] as !ã 'to stand' (with the zero efflux, possibly erroneous); and in [Bleek 1956: 368], in the textual example: misis kʼwi saː !ʼã́ "Missis is standing up".
ǀHaasi:ɳ!wa5
Story 1999: 23.
Proto-!Wi:
Reconstruction of this item is postponed due to too many difficulties: many languages feature phonetically similar forms that, however, do not manage to "come together" under any plausible historical scenario.
Bleek 1956: 596, 610. Plural form: ǀǀwaiˤ-nkʸe ~ ǀǀxʼweː-gǝn. Quoted as ǀǀwèi-sa in [Bleek 1929: 79]. The suffix -sa is a rarely encountered class marker.
Sands et al. 2007: 62. Polysemy: 'star / hedgehog' (apparently a common association for South Khoisan speakers, known also in !Xóõ). Quoted as ɡǀǀoiˤ-sí in [Westphal 1965: 143] (meaning glossed as pl. 'stars').
Ziervogel 1955: 38, 45. The form is plural ('stars'); the singulative form is ǀouni-zi. A completely different form, kale, is found in [Bleek 1929: 79] (quoted as kaleː 'stars' in [Bleek 1956: 78]).
Bleek 1937: 213; Bleek 1929: 79. Quoted as !ʼʰa in [Bleek 1956: 394].
ǀHaasi:
Not attested.
Proto-!Wi:*ǀǀxʼoeˤ- #
Distribution: Based on these data, the optimal candidate for Proto-!Wi (more precisely, Proto-Narrow !Wi) 'star' should be the isogloss between ǀXam and ǀǀXegwi. However, it must also be noted that the best external parallels (in Taa) are for the Nǀuu entry, and there is no sufficient evidence to think of any intense secondary Taa-Nǀuu contacts. Additionally, the item with the lateral click is also seen in several languages for which lexicostatistical calculations are impossible: cf. ǀǀKuǀǀe ǀǀan-te 'stars' (pl.), ǀǀKxau ǀǀʼɔan-si id. [Bleek 1956: 557]. We consider this evidence more substantial, which would mean that ǀXam and ǀǀXegwi may have independently replaced the original etymon (an easy possibility if the old root *ǀU- denoted a semantically close phenomenon, such as 'big star, planet', etc.). Replacements: No current etymologization for either the etymon seen in ǀXam and ǀǀXegwi or for ǀʼAuni !ʼʰaː. Reconstruction shape: All relevant data unequivocally agree on the reconstruction of the lateral click influx. Click efflux varies between several fluctuating reflexes. We tentatively accept the Nǀuu form as archaic, and ascribe the fluctuations to various assimilative/dissimilative processes, caused by its containing a glottalized velar affricate in the click efflux position and pharyngealization of the vowel at the same time. Naturally, this is a temporary decision, pending more detailed work on correspondences.
Bleek 1956: 412, 444. Plural form: !au-gǝn ~ !au-ukǝn. Quoted as !au ~ !ou in [Bleek 1929: 80]. Same word as 'mountain' q.v. (see 'mountain' for speculations on possible morphological differentiation between the two meanings).
Ziervogel 1955: 40. Quoted as ɕʼeo in [Lanham & Hallowes 1956: 100]. Cf. also the variant ǯwe 'stone' in [Lanham & Hallowes 1956a: 47]; the authors mention that "this is not the common Bushman word for 'stone', and it was obtained from one group of informants only". It is also proposed that this latter form may be a borrowing from Sotho liː=ǯwe id. In [Bleek 1929: 80] and [Bleek 1956: 265, 266], two forms for 'stone' are quoted: že and žuː. It is not clear if they represent the same word as Lanham & Hallowes' ɕʼeo or ǯwe (or both?). Overall, a rather complicated situation.
ǀ'Auni:ǀǀxʼɔː #3
Bleek 1929: 80. Only attested in this early, not very reliable, source, so the entry is dubious.
Distribution: Preserved almost everywhere, with the possible exception of some Nǀuu dialects and ǀʼAuni (the ǀʼAuni entry is dubious). Additionally, cf. ǀǀKxau !ao 'stone' [Bleek 1956: 408] and possibly ǀǀKuǀǀe dʔɔ 'rock' [Bleek 1956: 26]. Replacements: (a) ǂKhomani ǀǀuɾu = ǀXam ǀǀúːru ~ ǀǀurru 'stone knife, splinter of stone, quartz' [Bleek 1956: 593]; if Maingard's semantics is correct, this could reflect a semantic shift {'stone knife / stone splinter' > 'stone (gen.)'}; (b) ǀʼAuni ǀǀxʼɔː is a dubious form with no parallels. Reconstruction shape: Determined primarily by the shape of this word in ǀXam and Nǀuu. However, the actual vocalic structure may have been different, cf. ǀHaasi !òè and the palatalized reflexation in ǀǀXegwi žeu ~ ɕʼeo.
Bleek 1956: 607, 626, 627. Polysemy: 'sun / day / thirst'. Transcribed as ǀǀṍĩŋ ~ ǀǀʼõĩː ~ ǀǀʼõĩːŋ ~ ǀǀʼṹĩ́ŋ, emphatic form: ǀǀṍĩ́ŋ-yaŋ ~ ǀǀʼoiːŋ-ya by W. Bleek. In L. Lloyd's records, the form ǀǀʼũĩː is only recorded in the meaning 'day'. Quoted as ǀǀõĩŋ in [Bleek 1929: 81]. Irregular fluctuation of transcription between glottalic, velar affricate, and simple velar articulation of the click efflux suggests an original "non-trivial", undetected type of efflux, possibly uvular (*ǀǀqõĩ or *ǀǀqʼõĩ?).
Bleek 1956: 584, 585, 625, 626. Polysemy: 'sun / day / thirst'. Quoted as ǀǀõĩ in [Bleek 1929: 81] and [Bleek 2000: 20]. Irregular fluctuation of transcription between glottalic and simple velar articulation of the click efflux is the same as in ǀXam and suggests a "non-trivial" type of original articulation.
Lanham & Hallowes 1956: 117. Not attested in [Ziervogel 1955]. In [Bleek 1929: 81], two forms are quoted: ǀǀõĩ and ǀǀʼuːn; both of them are confirmed in [Bleek 1956: 584, 628]. It is not clear if these are two different words or dialectal variants. Likewise, neither of the forms can be equated with Lanham & Hallowes' ǀǀumi, although external comparison shows them to be more archaic in shape than ǀǀumi.
ǀ'Auni:ǀɛ̃́ ~ ǀɛn2
Bleek 1937: 210. Polysemy: 'sun / day'. Quoted as ǀɛ́ ~ ǀɛn in [Bleek 1956: 307]; as ǀẽː in [Bleek 1929: 81]. The latter source also adds a different synonym, ǀǀoi, not found in [Bleek 1937]. It corresponds to the general !Wi root for 'sun', but, since [Bleek 1929] is a generally unreliable source, we do not include ǀǀoi in our calculations - it could have been included by mistake, or represent a dialectal archaism, or, quite likely, a borrowing from Nǀuu.
ǀHaasi:ǀʼi2
Story 1999: 23.
Proto-!Wi:*ǀǀʼũĩ
Distribution: Preserved throughout Narrow !Wi. Additionally, cf. ǀǀKxau ǀǀʼoːeː 'sun' [Bleek 1958: 625]. Replacements: In Lower Nǂossob, the original root (whose archaicity is confirmed by external comparison with Taa languages) is replaced with *ǀ(ʼ)e ~ *ǀ(ʼ)i, a from that closely resembles the Common !Wi equivalent for 'fire' q.v.; however, we hesitate to postulate a lexical merger, since the actual recorded forms for 'fire' in ǀʼAuni and ǀHaasi are phonetically different. Reconstruction shape: The majority of the languages agree upon the phonetic shape *ǀǀʼũĩ ~ *ǀǀʼõẽ (it is not yet clear if the diphthongs ui and oe were opposed in Proto-!Wi or if they were mere phonetic variants). The only major dissenting form is ǀǀXegwi ǀǀumi, the way it is transcribed in [Lanham & Hallowes 1956]; assuming secondary labialization of the nasal under the influence of the preceding vowel, an earlier variant *ǀǀuni may not only indicate that *ǀǀʼũĩ is the result of consonantal lenition, but would also better agree with external parallels in Taa (cf. !Xóõ ǀǀʼân 'sun'). However, the lack of glottal efflux is rather confusing, as is the attestation of the variant ǀǀõĩ in ǀǀXegwi by D. Bleek. Such ambiguity of the data implies that one should probably adopt the "majority rule" here, and suspend amending the reconstruction from the widely supported *ǀǀʼũĩ to the ǀǀXegwi-supported *ǀǀʼuni until a detailed reconstruction of Proto-!Wi is produced.
Bleek 1956: 686. Quoted as ʘxu in [Bleek 1929: 82], along with the allegedly synonymous form gaːn = gːáːn [Bleek 1956: 43]. Both forms are only attested sporadically in W. Bleek's records. We choose the former as primary due to its being illustrated with an actual contextual example (ha-g ɳǀe kui ʘxuː u !ʌhí tiŋ !waː "he swam on top of the water"), but its status is questionable all the same.
ǀǀNg!ke:
Not attested.
ǂKhomani:
Not attested.
Nǀuu:
Not attested.
ǀǀXegwi:
Not attested.
ǀ'Auni:
Not attested.
ǀHaasi:
Not attested.
Proto-!Wi:
Not reconstructible due to lack of proper attestation.
Story 1999: 23. The prefix i= may be possessive ('his tail').
Proto-!Wi:*ǀa- #
Distribution: Preserved only in ǀHaasi, as may be established from external comparison. Replacements: The situation here is complicated. On the Narrow !Wi level (ǀXam + Nǀuu + ǀǀXegwi), the reconstruction *!ʰi 'tail' may be proposed beyond reasonable doubt. If this form could somehow be shown as related to ǀʼAuni ǂwi, it would have constituted the optimal candidate for the Proto-!Wi level. However, ǀʼAuni ǂ-never corresponds to Narrow !Wi !-, and a transcriptional error in this case is not highly likely (as a rule, it is old ǂ that gets mistranscribed as !, very rarely vice versa). If so, we have three distributionally permissible candidates for Proto-!Wi tail: Narrow !Wi *!ʰi, ǀʼAuni *ǂui, and ǀHaasi *ǀa-. On their own merits, Narrow !Wi *!ʰi would be preferable because of the largest chronological depth of reconstruction, but in the light of external comparison, it is ǀHaasi *ǀa- that has the best outside parallels in the Taa branch of South Khoisan - !Xóõ sg. ǀàũ, pl. ǀã̀ 'tail'. Because of this, the currently optimal scenario is to postulate preservation of archaic 'tail' in ǀHaasi, with independent replacements in ǀʼAuni and in Proto-Narrow !Wi. The nature of these replacements, however, remains unclear. Reconstruction shape: Since there are no !Wi parallels for the ǀHaasi form, we just tentatively project it onto the proto-level.
Number:85
Word:that
ǀXam:ǀeː ~ ǀe #1
Bleek 1956: 306. Transcribed as ǀeː ~ ǀèː ~ ǀɛː by W. Bleek. Correct definition of demonstrative pronouns in ǀXam is problematic, since reliable grammatical descriptions are lacking, most dictionary information is contradictory, and most of the textual examples inconclusive. ǀXam ǀe is almost certainly a stem that is used to indicate a far degree of deixis; however, it is glossed in [Bleek 1956] as "there, yonder, far, that, here", and the only textual example that is fully satisfactory for GLD standards is (from L. Lloyd's records) ɳǀa-kki ʘho ǀe "give me that piece of wood". The same stem is quoted in [Bleek 1929: 83] as ǀeː-á, most likely a compound of ǀeː with another deictic stem, a ~ e (see under 'this'). Unfortunately, no diagnostic contexts have been detected that would contrast 'this' and 'that' within one sentence. Another quasi-synonymous form in [Bleek 1929: 83] is ha, but it is glossed as "this, that" in [Bleek 1956: 55], with no possibility of evaluating the actual meaning.
Bleek 1956: 4. Listed as the first or default equivalent for 'that' both in [Bleek 2000: 20] and [Bleek 1929: 84], as opposed to a prosodically unmarked a 'this' q.v. The latter source lists three additional synonyms: (a) he, which is not confirmed in either [Bleek 1956] or [Bleek 2000]; (b) ɳǀeá, most likely a misprint for ɳǀǀe=á ([Bleek 2000: 20]; [Bleek 1956: 618]), a form which is best analyzed as a compound of a neutral-deictic stem ɳǀǀe with the far-deixis stem á (cf. the same situation for 'this' q.v.); (c) ǀǀŋ-á, most likely a contracted variant form of ɳǀǀe-a.
Ziervogel 1955: 55. The first morpheme is either a copula or a neutral deictic stem; the opposition between 'this' and 'that' is formed by the second component of the expression, cf. 'this'. In [Lanham & Hallowes 1956: 110], the form ʔeta is defined as the "rarest" of all demonstrative pronouns; no differences in degree of deixis are indicated between the three demonstrative pronouns ʔena, ʔela, ʔeta. D. Bleek gives ha as the main deictic stem ([Bleek 1929: 83]; [Bleek 1956: 55]); however, it does not really function as a demonstrative adjective, but rather as a general pronominal proclitic.
Bleek 1937: 202; Bleek 1956: 55. Only the variants he and hi are quoted in [Bleek 1929: 83]. According to [Bleek 1937: 197], the same word functions as the main 3rd p. sg. pronoun ('he', 'she', 'it'). The fluctuating vocalism most likely reflects contraction with nominal class markers.
ǀHaasi:cɔː-a5
Story 1999: 23.
Proto-!Wi:
Not properly reconstructible because of the highly unstable nature of the etyma and possible inaccuracies in available descriptions.
Bleek 1956: 4. Transcribed as a ~ aː by W. Bleek. There is also a different vocalic variant: eː, said to refer "to nouns in the pl. and to those in the sing. which take he, hi, instead of ha in the 3rd pers. sing." [Bleek 1956: 36]. This means that a(ː) and e(ː) both represent contractions of a basic deictic stem *V (vocalism not defined) with different class markers, a situation not atypical of other South Khoisan languages as well. Quoted as aː-á ~ aː ~ eː-á ~ eː in [Bleek 1929: 84]. For more details on the ǀXam system of demonstrative pronouns in general see under 'that'.
Bleek 1956: 4. Listed as the first or default equivalent for 'this' both in [Bleek 2000: 20] and [Bleek 1929: 84], as opposed to a prosodically different á 'that' q.v. The allegedly synonymous form ɳǀǀa ([Bleek 1929: 84]; [Bleek 2000: 20]; [Bleek 1956: 612]) is probably a contraction of the adverbial word ɳǀǀa 'here, there' + -a (for more details, see under 'that').
Ziervogel 1955: 55. The first morpheme is either a copula or a neutral deictic stem; the opposition between 'this' and 'that' is formed by the second component of the expression, cf. 'that'. There is also a phonetic variant (or, less probably, a separate synonymous stem) ʔe=na id. [Ziervogel 1955: 55]. Secondary synonym: cela [Ziervogel 1955: 56] (encountered relatively rarely; semantic difference from ʔe=la is unclear). Quoted as ʔila ~ ʔela in [Lanham & Hallowes 1956: 108], with the meaning is given as 'that'; however, Lanham & Hallowes actually fail to spot any semantic differences between ǀǀXegwi demonstrative pronouns (see notes on 'that').
Bleek 1937: 201; Bleek 1956: 4. In the earlier notes in [Bleek 1929: 84], the meaning 'this' is glossed as ti. However, in [Bleek 1937: 206] this word is already explained as part of the expression ti e 'that is, there', with no implications whatsoever about the semantics of "near deixis". Nevertheless, as in the case of nearly all !Wi languages, the data are somewhat controversial, and existing textual examples that never contrast 'this (near)' and 'that (far)' do not help matters much.
ǀHaasi:gʸa-ŋ3
Story 1999: 23.
Proto-!Wi:*a
Distribution: Seemingly preserved everywhere except for ǀǀXegwi and ǀHaasi. Replacements: The forms in ǀǀXegwi and ǀHaasi are not well understood from a historical perspective. Reconstruction shape: The monovocalic root *a generally stays the same in all descendant languages.
Bleek 1956: 3. Emphatic form: a-kǝn, possessive: á-ka. Transcribed as a ~ aː, emphatic form: a-kǝn, possessive: a-kka by W. Bleek. Quoted as a ~ a-a ~ aː-ken in [Bleek 1929: 85].
Ziervogel 1955: 46. Said to be pronounced with low tone, as opposed to ʔa 'not' q.v. with high tone [Ziervogel 1955: 43]. Cf. also the emphatic (absolute) form: ʔa-ʔe; the object form ʔaye ~ ʔai ~ ʔayi; the possessive form ʔaye [Ziervogel 1955: 45-47]. The absolute form is quoted as ʔa-ʔe ~ a-ʔe in [Lanham & Hallowes 1956: 108]. Quoted as a ~ aŋ in [Bleek 1929: 85]; as a, emphatic an, possessive a-ka in [Bleek 1956: 3].
Bleek 1956: 272, 310. Transcribed as ǀʼér̃r̃i ~ ǀér̃r̃iː by W. Bleek (the latter variant, not marking the glottal stop, is rare and probably erroneous). Quoted as ǀʼer̃i ~ ǀʼeni in [Bleek 1929: 86].
Sands et al. 2006. The first variant allegedly reflects the Western dialect, the second corresponds to the Eastern one. Quoted as ǀʼán in [Westphal 1965: 141].
Distribution: Preserved in all languages. Additionally, cf. ǀǀKxau ǀʼanan-si 'tongue' [Bleek 1958: 269]. Reconstruction shape: Initial click *ǀʼ- is preserved everywhere. Vocalic variation between -e- and -a- is most likely caused by dialectal assimilation (*CaCi > CeCi). The old syllabic structure is well preserved in ǀXam and ǀʼAuni, but tends to be simplified in descendant languages (either the final vowel gets dropped or the medial nasal gets "lenited", transforming into nasalized articulation of the surrounding vowels).
Bleek 1956: 574. Plural form: ǀǀʰeǀǀʰẽĩ. Quoted as ǀǀẽĩ, plural: ǀǀẽĩ-ǀǀẽĩ (with reduplication) in [Bleek 1929: 86]. None of the materials distinguish this word from 'horn' q.v., although external data very clearly speak in favour of their separate origin.
Bleek 1956: 551, 567, 568, 571. Plural form: ǀǀeŋǝn ~ ǀǀẽĩŋ ~ ǀǀẽǀǀẽ. Quoted as ǀǀẽĩ, plural ǀǀẽĩǀǀẽĩ in [Bleek 1929: 86]. None of the materials distinguish this word from 'horn' q.v., although external data very clearly speak in favour of their separate origin.
Bleek 1929: 86. Plural form: ǀǀẽĩwa-s. Not attested in later, more reliable, sources, but the entry is believable because of secure external cognates.
Story 1999: 23. Probably a plural form (kʼi= is one of the variants of the plural prefix).
Proto-!Wi:*ǀǀʰãĩ
Distribution: Preserved in all languages. Additionally, cf. ǀǀKuǀǀe kxe 'teeth' [Bleek 1958: 122]; !Gã!ne ǀǀʼĩ-ǀǀʼiŋ 'teeth' [Bleek 1958: 544]. Reconstruction shape: The lateral aspirated click is reconstructed based on ǀXam, Nǀuu, and ǀǀXegwi; in other languages, aspiration may have either disappeared or, more likely, not been noticed by researchers. The nasalized diphthong is preserved everywhere except for ǀǀXegwi (and even there it is still postulated by D. Bleek, if not by Lanham & Hallowes), so it is also reconstructed quite reliably.
Maingard 1937: 256. Meaning glossed as 'wood'; no special lexeme for 'tree' is attested, thus, the entry is somewhat dubious. Cf. also ɡʘōˤ-kē 'firewood' in [Doke 1936: 82].
Miller et al. 2007: 58. Meaning glossed as 'wood'. According to B. Sands, there is no special generic term for 'tree' in the language; the closest term is ǂʼʰiː 'shepherd's bush (Boscia albitrunca)', the plural form of which may possibly be used as a generic term. For the moment (until a published dictionary comes out), it is still reasonable to include the old word for 'wood / (growing) tree' on the list, keeping in mind that this may be an erroneous inclusion. Cf. also ǂʰí 'tree' in [Westphal 1965: 144].
Ziervogel 1955: 36, 42. Quoted as ɳʘòː-zì 'tree' (with specially marked low tone and the singulative suffix -zi) in [Lanham & Hallowes 1956: 106]. Cf. also, without this suffix, ɳʘoː 'wood', and with the plural suffix -ŋ, ɳʘoː-ŋ 'forest' [Lanham & Hallowes 1956: 98, 111]. Quoted as ʘʼʰoː ~ ʘʼʰoː-si, pl. ʘʼʰo-gǝn in [Bleek 1956: 682].
Distribution: Preserved in all languages, but the meaning is frequently restricted to 'wood' as material; this seems to be true in the case of at least ǀHaasi and, possibly, Modern Nǀuu, where the meaning 'tree' has not been explicitly attested for this item. Additionally, cf. ǀǀKxau ʘoː, pl. ʘoː-kn 'tree' [Bleek 1956: 684]. Replacements: External data (in Taa) clearly show that *ʘo used to be applied to 'wood' as material as well as 'tree' as a living organism. However, similarity of the attested situations in ǀHaasi and Nǀuu (representing both subbranches of !Wi) shows that already on the Proto-!Wi level, the word *ǂʼʰi 'Boscia albitrunca' may have been used as a generic term with specific reference to living trees. On the other hand, the related forms found in [Bleek 1956], e. g. ǀXam !ʰi 'umbrella-top tree, witgatboom' [Bleek 1956: 426], etc., always refer to the precise tree species, and there are also explicit textual cases where the old form *ʘo is found referring to living trees. Keeping this in mind, we have to treat the situations in Modern Nǀuu and ǀHaasi as lexical replacements: {'shepherd's tree' > 'tree (gen.)'}. Reconstruction shape: The only thing that is not easily reconstructible for this form is the click efflux. In old data collections, almost every possible variant is attested (zero efflux, aspirated efflux, aspirated glottalized efflux, voiced efflux, nasal efflux, etc.), which should probably be interpreted as the original presence of some rare type of efflux, simplifying in different ways in different languages and phonetically misinterpreted by inexperienced researchers in others. External data (e. g. !Xóõ ʼɳʘà-ye with a preglottalized nasalized efflux) indirectly validate that interpretation as well. However, until more examples become available, we prefer to refrain from making the reconstruction too complex and tentatively choose the simplest variant, attested in Modern Nǀuu (with the zero efflux). Semantics and structure: Should be reconstructed with the polysemy 'tree / wood'.
Bleek 1956: 448, 492. Transcribed as !úː ~ !ʼuː by W. Bleek. Quoted as !úː in [Bleek 1929: 88]. Fluctuation between simple and glottalized articulation for the click efflux in W. Bleek's records is hardly accidental or erroneous in the light of similar fluctuations in between other South Khoisan languages as well; for ǀXam, this may signify a "non-trivial" type of click efflux articulation.
Ziervogel 1955: 58. Quoted as kʼuː ~ ɕʼuː 'be two' in [Lanham & Hallowes 1956: 115]. Quoted as ǀǀuː in [Bleek 1929: 88]; as ǀǀu ~ ǀǀʼu in [Bleek 1956: 591, 627].
ǀ'Auni:ǀam-1
Bleek 1937: 209; Bleek 1956: 299. Correctly identified as a Nama borrowing. Quoted as ǀʌm in [Bleek 1929: 88].
ǀHaasi:s=ǀǀaː-maː2
Story 1999: 23. Composition of this numeral is unclear. The main root morpheme is probably ǀǀaː.
Proto-!Wi:*!uʔ-
Distribution: Preserved in all Narrow !Wi languages. Additionally, cf. ǀǀKxau !ʼuː, ǀǀKuǀǀe !ʼu [Bleek 1956: 492]. Replacements: (a) In ǀʼAuni, the original numeral was replaced by a borrowing from Nama; (b) internal composition of the form s=ǀǀaː-maː in ǀHaasi is unclear; the sequence s=ǀǀaː- can only reflect some sort of fused compound (*sV- + *ǀǀa-), but there are no definite ideas on what the separate parts might go back to. If this is indeed a complex idiomatic expression, it is clearly not archaic. Reconstruction shape: The word is commonly attested either in the variant !u or in the variant !ʼu; sometimes both variants seem to be in "free variation" within the same language. This indicates either some uncommon type of efflux or an original combination of a non-glottalized efflux with a medial glottal stop (thus, *!uʔ- could either be misheard as *!ʼu-, or the glottal stop could genuinely undergo "metathesis"). See similar cases with 'one', 'sit', etc.
Bleek 1956: 512, 519. Transcribed as ǀǀʼaː ~ ǀǀʼa ~ ǀǀʼa-kǝn ~ ǀǀʼaŋ ~ ǀǀʼe ~ ǀǀʼeː ~ ǀǀʼɛː by W. Bleek. Quoted as ǀǀʼe ~ ǀǀʼaː in [Bleek 1929: 42]. This is quite clearly the basic verb to designate beginning of movement towards an object, e. g.: ŋ ǀǀʼa ha toːi 'I go to that ostrich', etc. Vocalic fluctuation is typical of most basic monosyllabic verbal stems and may reflect either contextual phonetic variation or merger with class markers. Another quasi-synonym listed in [Bleek 1929: 42] is tàiˤ ~ tã̀ẽˤ, but textual examples in [Bleek 1956: 187] confirm rather strongly that the basic meaning of this verb is 'to walk' (i. e. with emphasis on the process of movement) rather than the required semantics of 'go'.
Bleek 1956: 513. Quoted as ǀǀʼa in [Bleek 1929: 42] and [Bleek 2000: 20]. Two more variants in [Bleek 1929: 42] are not confirmed in [Bleek 1956]: !eya is actually 'to bring' and ǀǀʼana is not found at all (it is probably just a different variant of ǀǀʼa with a postpositional particle).
Maingard 1937: 247, 257. Secondary synonym: ɕxan [Maingard 1937: 245], unsupported by data from other sources. Distinct from tãĩ 'to walk' [Maingard 1937: 257].
Ziervogel 1955: 36, 39. Past tense form is quoted as ǀǀaː-ya [Ziervogel 1955: 51]. There is a second quasi-synonymous root: tʼãʔã-ne [Ziervogel 1955: 50, 60, 61] = tʼaʔa ~ tʼaʔan [Lanham & Hallowes 1956: 118]. It is also listed as the main equivalent for 'go' in [Bleek 1929: 42], as teʌn ~ anteʌn (the second variant is contracted with a verbal or pronominal proclitic). However, comparative analysis of sources shows that the basic semantics of this second verb is more complex. Cf. the following examples with their (obviously, approximate) translations: !ʰoa antean "the cow goes away" [Bleek 1956: 10]; n tean "I walk", ha tean "she goes away", teaːni "run away" [Bleek 1956: 197]. Also, Ziervogel always gives the meaning 'walk' for this root; its external connections also confirm such meanings as 'go away', 'depart', 'travel'. Consequently, we prefer to exclude this lexeme from our calculations.
Bleek 1937: 214, 215; Bleek 1956: 513, 519, 545. The variety of variants indicates that the original form may have been *ǀǀaʔa. Not listed in [Bleek 1929: 42], with two other quasi-synonyms suggested instead: (1) taãĩ, glossed in [Bleek 1937: 206] as tãĩ ~ tai ~ taãĩ 'to walk, to go'; its general semantics seems to be more or less the same as the one of its cognate in ǀǀXegwi, q.v.; (2) !uŋ, probably a marginal borrowing from a Central Khoisan language (cf. Proto-Central Khoisan *!ũ 'to go').
Story 1999: 22. Secondary synonym: tʸá-ai id. [ibid.]. This second word etymologically coincides with ǀʼAuni taãĩ and must probably reflect the same semantics ('walk' rather than 'go').
Proto-!Wi:*ǀǀʼa-
Distribution: Preserved in all languages, except for ǀHaasi. Additionally, cf. ǀǀKxau ǀǀʼa ~ ǀǀʼaː ~ ǀǀʼaŋ ~ ǀǀʼa-i [Bleek 1956: 513]. Replacements: The provenance of ǀHaasi ǂa is unclear. There is a possibility that it is really an inaccurate transcription of *ǀǀʼa, but this is not highly likely (no other evident examples of misinterpreting ǀǀ- as ǂ- may be drawn from Story's vocabulary). Reconstruction shape: The majority of attested variants allows to reconstruct Proto-!Wi *ǀǀʼa- 'to go' without controversy. There are, however, some variants without the glottal efflux (e. g. in ǀǀXegwi) that are not so easily explainable. Semantics and structure: In Proto-!Wi, *ǀǀʼa- 'to go' was most likely opposed to *taˤ- ~ *taˤ-iŋ 'to walk' (without a specific direction).
Build tree