Bleek 1956: 408, 444. Transcribed as !áo-gǝn ~ !áo-ka ~ !aːo-kǝn by W. Bleek. Quoted as !au ~ !ou in [Bleek 1929: 59]. Same word as 'stone' q.v.; it is possible that in the meaning 'mountain' the root is more frequently used in conjunction with suffixal extensions, but no strict morphological opposition between 'stone' and 'mountain' can really be determined from available materials.
Ziervogel 1955: 60. Encountered once in the texts, within the noun phrase thaŋ ʔe čwa "black mountains"; not very certain. In [Lanham & Hallowes 1956: 106], a completely different compound form is quoted in the meaning 'hill, mountain', with two phonetic variants: ɡǀǀu-ɳǀa ~ gu-ɳǀa (the second component here may be ɳǀa 'head' q.v.).
ǀ'Auni:ǀǀʼwa #3
Bleek 1937: 218; Bleek 1956: 629. Meaning glossed as 'hill' (no special word for 'mountain' is known).
ǀHaasi:
Not attested.
Proto-!Wi:*!ao #
Distribution: Well attested in ǀXam and Nǀuu; situation in all other languages is dubious. Replacements: Supposed semantic equivalents for 'hill, mountain' in ǀǀXegwi and ǀʼAuni do not look particularly reliable and find no suitable etymological support. Consequently, ǀXam-Nǀuu *!ao 'stone / mountain' emerges as the only candidate for this meaning with sufficient distribution and confirmation in textual sources. See 'stone' for further discussion.
Bleek 1956: 239. Transcribed as tːú, emphatic form tːú-kǝn, pl. tːú-tːu ~ túwa-kǝn by W. Bleek. Quoted as tuː in [Bleek 1929: 59]. Polysemy: 'mouth / hole'.
Lanham & Hallowes 1956: 108, 109. Plural form: tu-ŋ. Not attested in [Ziervogel 1955]. Quoted as tu in [Bleek 1929: 59] and [Bleek 1956: 239].
ǀ'Auni:ǂuː2
Bleek 1937: 219; Bleek 1956: 664. A somewhat complicated case. The same source lists, as synonymous, the form tu ~ tʰu 'mouth' [Bleek 1937: 207], adding that the form "may be ǂkhomani", i. e. borrowed from Nǀuu-ǂKhomani with whom the ǀʼAuni, as described by D. Bleek, had been in close contacts. Although *tu 'mouth' is a relatively stable !Wi stem, and there is nothing per se that would make its appearance in ǀʼAuni unusual, Bleek's idea is corroborated by the fact that her own earlier data on ǀʼAuni, collected in 1911, although generally inferior in quality to later research, only lists ǂuː and nothing else in the meaning 'mouth' [Bleek 1929: 59]. We go along with her suggestion and include ǂuː as the basic equivalent for this meaning.
Story 1999: 22. The element n= is probably the 1st p. possessive prefix.
Proto-!Wi:*ǂu
Distribution: Preserved only in ǀʼAuni. Replacements: This is a case where our decision rests exclusively on external evidence. ǀʼAuni ǂuː precisely corresponds to !Xóõ (Taa) ǂû-e 'mouth', and, unlike the isogloss between ǀʼAuni tu and early Nǀuu tu 'mouth', the ǀʼAuni-!Xóõ parallel is not easily interpretable as the result of areal contacts. This is significant evidence for regarding the ǀʼAuni form as an archaism, and interpreting Common Narrow !Wi *tu 'mouth' as an innovation (possibly 'mouth' < 'hole'?) that took place before the subgroup's primary split into ǀǀXegwi and ǀXam-Nǀuu. Another problem is the form attested in ǀHaasi, which corresponds to ǀʼAuni ǂuː in its consonantal structure, but shows an entirely different root vowel. Unless it can be shown that ǀHaasi =ǂa is contracted from *ǂu-a, where -a is a fossilized class suffix, we prefer to treat it as another lexical replacement of an obscure nature.
Story 1999: 22. Etymologically probably = "your name" (a= is the 2nd p. possessive prefix), but cf. aǀaŋa kʼas ǀǀkabbo "my name is ǀǀKabbo" [Story 1999: 29].
Proto-!Wi:*ǀe ~ *ǀẽ
Distribution: Preserved in all languages. Reconstruction shape: Correspondences are fairly straightforward. Nasalization is consistently marked everywhere except for ǀǀXegwi, but variation in ǀXam suggests that nasal articulation of the vowel may be due to contraction with a suffixal marker. The extended form in ǀHaasi may be emphatic in origin.
Bleek 1956: 412, 428. Transcribed as !au ~ !ʰou, emphatic form !ʰóu-gǝn by W. Bleek. Quoted as !ʰou in [Bleek 1929: 60]. Fluctuation of efflux articulation between aspiration and zero (!au ~ !ʰou) should be interpreted as a possible reflection of improperly identified non-trivial articulation rather than reflexation of two different roots.
Bleek 1956: 103, 448. The clickless variant is mentioned as an "occasional form", reflecting a particular lect that tends to drop the alveolar click. Quoted as !úː in [Bleek 1929: 60].
Ziervogel 1955: 41. Quoted as !eleŋ in [Lanham & Hallowes 1956: 104]. This form looks quite similar to Ziervogel's ǀʼele, but could represent the same word only if the transcription of the initial click in one of the sources is due to a mistake or typographic error.
Bleek 1937: 216. This word is reproduced in [Bleek 1956: 591], with a supporting textual example: sa ko ǀxʼẽsi, ho ha ǀǀú ǀǀo "bring beads, on my neck put them". There is also a synonymous word for 'neck' attested in all the sources: ǂõĩ ([Bleek 1929: 60]; [Bleek 1937: 219]; [Bleek 1956: 663]); however, no textual confirmations for it are present.
ǀHaasi:
Not attested.
Proto-!Wi:*ǂqu
Distribution: Preserved everywhere except for ǀǀXegwi. Additionally, cf. ǀǀKxau ǂʼu [Bleek 1956: 676]. Replacements: Replaced in ǀǀXegwi with ǀʼele, a word of unclear provenance. Reconstruction shape: Attestation in different varieties of Nǀuu suggests *ǂqu as the original form of this word. Potential cognates in ǀXam and ǀʼAuni are somewhat problematic: (a) in ǀXam, alveolar ! is an expected correlate for the palatal click in Nǀuu, but the "epenthetic" vowel -a- (or -o-) is not very well understood; (b) ǀʼAuni ǀǀú is only acceptable as a cognate if the lateral click in D. Bleek's records is a mistake for a palatal click (there is also an exotic possibility that *ǂq- > ǀǀ- in ǀʼAuni could have been a regular development; cf. a similar case in the etymon 'short' q.v.).
Bleek 1956: 556, 565, 573. Plural form: ǀǀaˤǀǀaˤr̃r̃a. Transcribed as ǀǀaːŋ ~ ǀǀʰaːŋ by W. Bleek. Polysemy: 'new / fresh / raw'. Attested contexts clearly show that the word may be used both with the semantics of 'freshly grown / raw' and 'replacing a previous object' (e. g. ǀǀʰoː ǀǀaːˤŋ "new bag", ǀǀʰoǀǀʰo ǀǀaˤǀǀaˤr̃r̃a "new bags"). Quoted as ǀǀaːŋ in [Bleek 1929: 61]. The latter source adds a potential synonym: ǀǀwe = ǀǀwèː ~ ǀǀwẽː ~ ǀǀwɛ̃̀ː [Bleek 1956: 599] (only in W. Bleek's records), but rather scarcely attested examples only show this word in conjunction with 'moon', making its "basic" character in ǀXam rather dubious (the adverbial form ǀǀwẽː 'strongly' in the same vocabulary entry is supposed to be derived from this stem, but this is semantically questionable).
ǀǀNg!ke:!xeː-kʸa ~ !xeː-tʸa2
Bleek 1956: 499. Polysemy: 'new / young'. Quoted as simply !xeː in [Bleek 1929: 61]; the forms in [Bleek 1956] look like bimorphemic compounds, but the second component is unclear. The earlier source also lists ǂʼẽ as a synonymous form, but it is not confirmed in [Bleek 1956], whereas !xeː- is even propped up by contextual examples (e. g. !xe !xeː-tʸa "new dresses"; ǀǀxeː ǀʼa ha !xeː-kʸa "the girl is young").
ǂKhomani:
Not attested.
Nǀuu:
Not attested.
ǀǀXegwi:ǀǀʼi3
Lanham & Hallowes 1956: 105. Attested in the phrase haː debe ʔela ǀǀʼi-wa "it is a new knife". Not attested in [Ziervogel 1955].
ǀ'Auni:
Not attested.
ǀHaasi:
Not attested.
Proto-!Wi:
Not properly reconstructible based on available data. External data in Taa show that the "marginal" form ǀǀwe in ǀXam, not eligible for basic item position in the attested 19th century varieties of the language, could actually be the best candidate for the meaning 'new' in Proto-!Wi.
Lanham & Hallowes 1956: 105; Bleek 1929: 61. Attested in the expression e ɡǀǀaː "at night". Not attested in [Ziervogel 1955]. Quoted as ɡǀǀa in [Bleek 1956: 522].
Story 1999: 22. Reduplicated stem (reason for reduplication is, however, unknown).
Proto-!Wi:*ɡǀǀa
Distribution: Preserved in all languages. Additionally, cf. ǀǀKxau ǀǀaː [Bleek 1956: 546], ǀǀKuǀǀe ɡǀǀà [Bleek 1956: 522]. Reconstruction shape: The most frequently encountered variant is *ɡǀǀa. The diphthong -au in ǀʼAuni finds no confirmation in other sources and must be the result of contraction with a nominal suffix. C. Doke's marking of glottalized articulation in this root (either as a click efflux or as glottalization on the vowel) is also not confirmed elsewhere and possibly reflects an extended variant like *ɡǀǀa-ʔa.
Bleek 1956: 352. Plural form: ɳǀũ-ɳǀṹː-tu (with reduplication). Transcribed as ɳǀur̃u, pl. ɳǀũɳǀũ-tu "nostrils" by W. Bleek. Quoted as ɳǀur̃u in [Bleek 1929: 62].
Bleek 1956: 353; Bleek 1929: 62. Plural form: ɳǀu-tu-yǝn ~ ɳǀu-tu-ke (plural form meaning is given as 'nostrils' in [Bleek 1956]). The suffix -tu is a standard element in anatomical (and some other) terms.
Distribution: Preserved in all languages. Additionally, cf. ǀǀKxau ɳǀú-tú, ǀǀKuǀǀe ɳǀu-tu [Bleek 1956: 353]. Reconstruction shape: Correspondences are trivial and straightforward. Semantics and structure: In ǀXam and Nǀuu, the root is most commonly encountered in conjunction with the suffix *-tu (< *tu 'hole; mouth'); judging by the situation in ǀʼAuni, this complex formation may have originally referred specifically to 'nostrils'.
Bleek 1956: 121. Transcribed as xʼau ~ xʼaú-ki by W. Bleek. Quoted as kʼau ~ kʼau-ki in [Bleek 1929: 62] (only the complex variant kʼau-ki also quoted as the predicative negation 'no'). There also used to exist a dialectal variant ouki [Bleek 1956: 155], with areal dropping of the initial velar affricate. The negative root morpheme is clearly *xʼau; -ki is a verbalizing suffix ("one of the connectives of double verbs", according to [Bleek 1956: 121]).
Bleek 1956: 342, 348. Quoted as ɳǀa ~ ɳǀɔ in [Bleek 1929: 62]; ɳǀa ~ ɳǀe ~ ɳǀi ~ ɳǀo in [Bleek 2000: 21]. The negation seems to behave like a verbal stem, judging by unpredictable vocalic variation that may have something to do with the usual verbal mergers with class markers in ǀǀNg!ke. Examples: ŋ ɳǀa xŋ tia ha "I did not understand him"; ŋ ɳǀa ǀí ʘwaiki "I do not have any meat" [Bleek 1956: 342]; ŋ ɳǀo ǀǀai "I do not know"; ŋ ɳǀi kieŋ "I do not sleep" [Bleek 1956: 348].
Ziervogel 1955: 43, 53; Lanham & Hallowes 1956: 114. Said by Ziervogel to be pronounced with high tone, as opposed to ʔa 'thou' q.v. with low tone. Translations of 'not' as ǀeːwa and ka in [Bleek 1929: 62] are erroneous; the latter form is not confirmed at all in [Bleek 1956], and ǀeːwa is glossed in [Bleek 1956: 310] as the negative predicate 'not to be', with the example han ǀeːwa ka "he is not here" (there is a possibility of incorrect morphemic segmentation).
Bleek 1937: 203, 206; Bleek 1956: 91, 185, 202. All the three particles are probably variants of one and the same negation *ta (with palatal realization of the initial consonant among some speakers). In [Bleek 1929: 62], 'not' is translated as either taku (probably a contraction of ta and the verbal particle ku) or ǀǀa (probably erroneous, since its existence not confirmed in [Bleek 1937]).
Distribution: On a strictly distributional basis, the item is not properly reconstructible; however, external comparison suggests that the lateral click stem, attested in various forms of Nǀuu, is the most archaic of all alternatives. Additionally, cf. Seroa ǀǀau 'not' [Bleek 1956: 562]. Replacements: Essentially, each language or dialect cluster within !Wi is represented by its own basic negation: ǀXam xʼau, Nǀuu *ǀǀV, ǀǀXegwi ʔa, Lower Nǂossob *ta. It does not seem possible at present to establish a reliable scenario for the historic development of this Swadesh item in !Wi. Semantics and structure: Most of the variants of 'not' in !Wi, including *ǀǀ-, behave like verbal stems, with vocalic gradation in the root typical of most non-derived verbal stems.
Bleek 1956: 566, 582, 590. Other than ɳǀV, no other forms for the basic negation are mentioned in either [Bleek 1929] or [Bleek 2000]. [Bleek 1956], however, offers ample evidence for an alternate negative marker, ǀǀV, which also behaves like a monoconsonantal root with vowel gradation. Examples: !aːˤ ǀǀu ǂóː "rain does not fall" [Bleek 1956: 590]; n ǀǀɔ́ ɳǀǀi ǀǀõẽ "I do not see the sun" [Bleek 1956: 582]; ha ǀǀãũː ǀǀe a "her brother he is not" [Bleek 1956: 566]. Based on available evidence, no differentiation whatsoever can be established between these two forms; we have to treat them as synonymous.
Bleek 1956: 459. Polysemy: 'one / alone / once'. Transcribed as !waːi by W. Bleek; also as !ʼoáːi ~ !ʼo̯áːi only in the meaning 'alone' [Bleek 1956: 490] (but this is clearly the same word). Transcribed as !waːi in [Bleek 1929: 63]. The variety of transcribed variants suggests a "non-trivial" original articulation for the word; may either reflect a glottal stop breaking up a vocalic sequence (*!oʔai) or an unrecognized uvular glottalized efflux (*!qʼoai).
Bleek 1956: 599, 630. Cf. also a variant with a suffixal extension: ǀǀoeː-nso [Bleek 1956: 584]. Quoted as ǀǀwe in [Bleek 1929: 63] and [Bleek 2000: 20]. The glottalized variant ǀǀʼweː is said to be quite rare.
Story 1999: 22. The "suffix" -kʼa may actually be the same as the verbal copula kʼa (i. e. 'one-is'); cf. ǀǀuaː-kʼa 'three'.
Proto-!Wi:*ǂ1oʔ-
Distribution: Preserved in all daughter languages. Additionally, cf. ǀǀKuǀǀe ǀǀxʼoa 'one' [Bleek 1956: 606], Seroa ǀǀoai 'one' [Bleek 1956: 625]. Reconstruction shape: The click influx in this item shows more or less the same reflexes as in the word 'bone' q.v. and a few others; we tentatively reconstruct it as *ǂ1-, indicating that the reflexation is similar to the "normal" palatal click (for ǀXam and Lower Nǂossob languages at least; the Nǀuu reflexation is lateral, and the ǀǀXegwi reflexation is alveolar). Additionally, the efflux fluctuates between velar (zero) and glottalized, which we typically ascribe to glottalic articulation of the vowel, sometimes transferred to the consonantal onset of the syllable. Root vowel is -o- everywhere except for Lower Nǂossob, but the second (suffixal?) vowel varies across languages: *ǂ1oʔ-e ~ *ǂ1oʔ-a, possibly also *ǂ1oʔ-/a/ŋ in Lower Nǂossob.
Bleek 1956: 466. Transcribed as !üi, emphatic form !üi-ten ~ !üi-ya by W. Bleek. The word is sometimes translated as 'man', but generally in ambiguous contexts; the proper equivalent for 'male human being' is the compound form !wi=gwaiː q.v. There also exists a shorter variant of the same root, namely, !u, but it is almost exclusively encountered in compound forms, such as !ü go̯áːi 'man', !ü ǀʼa 'girl', !ü ǀʼaːiti 'woman' (W. Bleek), !u-de 'someone' (L. Lloyd), etc. [Bleek 1956: 447]. This would, however, indicate that the original root is simply !u, whereas -i is an additional nominal suffix. Quoted as !wi ~ !u in [Bleek 1929: 65]. The plural form is suppletive: !é ~ !kʼé, emph. !é-tǝn ~ !kʼé-tǝn (L. Lloyd), !ɛ́ ~ !ɛː ~ !ʼɛ ~ !ei, emph.: !ʼei-tǝn (W. Bleek) 'people' [Bleek 1956: 373, 419, 420] (possibly reflecting an actual form like *!qe or *!qʼe).
Bleek 1956: 458, 466; Bleek 2000: 18. The latter source also notes the existence of a clickless variant kwa for the same word, although it is not mentioned again in any of the dictionaries. The variant !wa is by far the most frequent, but a couple of examples are also given on the use of !wi, which suggests a morphological segmentation into *!u-a, *!u-i. Quoted as !wa in [Bleek 1929: 65]; this source also adds ɡ!e as a possible synonym, but the only known context in [Bleek 1956: 380] gives only plural usage: a ɡ!e nke "they are people". Clearly, this is the same suppletive plural as in ǀXam: cf. also the more common variants, such as !ʼe ~ !ʼe-gǝn ~ !e ~ !kʼe [Bleek 1956: 373, 420; Bleek 2000: 19], which also participate in the self-designating compound form ǀǀŋ-!e, literally 'home people'.
Sands et al. 2006. See notes on ǂKhomani for analysis. It is not clear if the stem !ui is encountered in isolation, but, in any case, it is !ui that carries the central meaning of 'person', with ɳǀǀŋ ('home'?) serving as a modifier for self-designation. Suppletive plural: ɳǀǀŋ ǂeː 'people'. Quoted as ɳǀǀ-!wé, pl. ɳǀǀ-ǂe in [Westphal 1965: 139].
Bleek 1937: 218; Bleek 1956: 643, 652. Although both forms are glossed as 'men, people' in [Bleek 1937] (e. g. as plural forms), this is contradicted by the fact that the same source also lists the specifically plural form ǂʼi-te; textual examples confirming both variants as singular forms are also found in [Bleek 1956: 652], e. g. ǂʼi ti ǂʼú-u "one person". Quoted as ǂʼi, pl. ǂʼiː-te in [Bleek 1929: 65]. On the possibility of the word da to be translated as 'person', see under 'man'.
Distribution: Preserved throughout Narrow !Wi, but seemingly dropped in the Lower Nǂossob branch. Replacements: Most languages of the Narrow !Wi branch present firm evidence that the paradigm of this noun was suppletive in the protolanguage: *!u- (sg.) vs. *ǂe ~ *ǂʼe (possibly = *ǂeʔe) (pl.). In comparison, Lower Nǂossob languages only show reflexes of the second root both in sg. and pl. forms; this is most logically interpreted as an innovative generalization of the plural form, with a "new" plural formed wherever necessary (e. g. sg. ǂʼi vs. pl. ǂʼi-te in ǀʼAuni). The semantic shift {'people' > 'person'} is quite trivial. Reconstruction shape: The basic root shape of the sg. 'person' is straightforwardly reconstructible as *!u-; *!u-i and *!u-a are probably morphological variants, although only the former is widely distributed and clearly traceable back to proto-status.
Bleek 1956: 431, 457. Transcribed as !ʰwaː by W. Bleek. Quoted as !ʰwaː in [Bleek 1929: 68] (the latter source also adds the compound !ʰwaː ǀǀi, literally 'water's liquid', in the same meaning, but this is hardly the basic denotation). External data show that the Proto-!Wi roots for 'rain' and 'water' must have been different, but phonetically similar, and it is not excluded that the two words had been confused in transcription, both by W. Bleek and L. Lloyd. There is, however, no direct evidence for such a confusion in any of the records, and it must be assumed that in ǀXam the two words simply merged into one.
Bleek 1956: 423; Bleek 2000: 28. Quoted as !ʰàː in [Bleek 1929: 68]. It is unclear if this word is exactly the same as 'water' q.v., or a phonetically similar stem obscured by inadequate transcription. See notes on ǀXam for an identical situation in that language.
Lanham & Hallowes 1956: 100. Not attested in [Ziervogel 1955]. Attested mostly as a verbal root (e. g. ha gaʔa haɲ ɕʰeuŋ-we "the sky is raining" [Lanham & Hallowes 1956: 118]), thus, slightly dubious. In [Bleek 1929: 68] an entirely different word is quoted: gaːa (gaa in [Bleek 1956: 41]).
Bleek 1937: 216. Quoted as ǀǀʰàː in [Bleek 1929: 68]. See notes on 'water' for further details.
ǀHaasi:
Not attested. Cf., however, ǂĩ́ 'to rain' (verb) [Story 1999: 22].
Proto-!Wi:*ǂqa-
Distribution: Preserved everywhere where attested. Reconstruction shape: Correspondences for the click influx between ǀXam, Nǀuu, and ǀʼAuni are the same as in 'neck' q.v., allowing to trace all the attested forms to a common prototype; lateral click in ǀʼAuni may have been a mistake on D. Bleek's part, or the result of regular phonetic change from an original *ǂq-. For the affricate in ǀǀXegwi, cf. 'short' with an identic development. Click efflux varies between zero, aspirated, and uvular; we tentatively assume uvular articulation as primary because of the more recent and allegedly accurate Nǀuu data, but *ǂqʰ- is almost equally probable as the proto-phoneme. The vocalic coda is realized as -a(a) or -au, probably reflecting different nominal suffixation, but this situation is actually quite unique, and bears further investigation.
Bleek 1956: 427, 434. Verbal root: 'to be red'. Transcribed as !iː by W. Bleek. Quoted as !íː in [Bleek 1929: 69].
ǀǀNg!ke:xreː-kʸa2
Bleek 1956: 260. Meaning is glossed as 'red, brown', with a supporting text example: ǀǀʼõĩ e xreːkʸa "the sun is red" (?). In [Bleek 1929: 69], the word is quoted simply as xreː, without the second auxiliary morpheme. Root structure here is atypical of ǀǀNg!ke and Khoisan in general and may represent a secondary reduction from *xVre.
Sands et al. 2006. Meaning glossed as 'red; reddish brown'. The form is a transparent borrowing from Khoekhoe (Nama ǀʼawa, !Ora ǀxʼaba 'red'). There is also a (presumably less used) synonym ku=ǂqĩ, which must be more archaic (cf. the same equivalent in C. Doke's recordings of ǂKhomani).
Lanham & Hallowes 1956: 115. Meaning glossed as 'be red'. The entry 'red' = ǀamse in [Bleek 1929: 69] has to be amended in the light of the same word in [Bleek 1956: 300]: the earlier publication accidentally misprints 'reed' as 'red'.
ǀ'Auni:
Not attested.
ǀHaasi:cxwe-kʼa4
Story 1999: 22.
Proto-!Wi:*ǂ1e ~ *ǂ1i
Distribution: An isogloss between ǀXam and ǀǀXegwi, making the word the optimal candidate for "Proto-Narrow !Wi". Reconstruction on a deeper level is impossible. Replacements: Except for Nǀuu ǀxʼaba, which is most likely a recent introduction of Khoekhoe origin, most of the other equivalents for the meaning 'red' in !Wi languages do not easily find internal correspondences. They also tend to be phonetically peculiar (e.g. xreː- in ǀǀNg!ke), suggesting further possibilities of borrowing that have to be investigated. Reconstruction shape: The consonantal correspondence between ǀXam !iː- and ǀǀXegwi !e is the same as in the etyma 'bone' and 'one' q.v., which we currently mark as *ǂ1-. The vowel is unquestionably a front one, but the exact quality is not easily determinable. In Nǀuu, the expected click influx correspondence would have been lateral ǀǀ-; consequently, ǂKhomani ǂʼī in Doke's transcription cannot be compared.
Bleek 1956: 497. Meaning glossed as 'path'. Transcribed as !xárra by W. Bleek. Quoted as !xaːra in [Bleek 1929: 64]. The latter source also mentions a second synonym: !auː-ö = !au ~ !ao in [Bleek 1956: 408, 412], where this word is understood as a figurative meaning of 'stone, mountain' q.v. Regardless of whether this is just homonymy or, indeed, a result of semantic shift, the accompanying text examples do not refer to man-made paths, cf.: he ha siŋ ǂakka ka !auː "of which (ostrich) he told me its path" [Bleek 1956: 412]; ŋ kwaŋ taŋ kaŋ ǀǀaː !kʼãũː ǀʰiŋ ɡ!uru-ɳǀǀaː-ka !áó "I therefore intend to go passing through !Guru-ǀǀNa's pass" [Bleek 1956: 408].
Bleek 1937: 213 (quoted as kʌ́n, with the click symbol omitted through a typographic error); Bleek 1956: 456 (with the click symbol correctly restored). Quoted as !ʼane in [Bleek 1929: 65]: probably the same word in a misheard (or dialectal) variant. Meaning is glossed as 'path' in all sources. Cf. also additional synonyms: (a) ǀǀùru 'path' ([Bleek 1937: 216]; [Bleek 1956: 593]); (b) ǂxʼei 'road, path' ([Bleek 1937: 219]; [Bleek 1956: 668]); semantic difference unknown, but only !án has external cognates.
ǀHaasi:
Not attested.
Proto-!Wi:*!an
Distribution: Preserved in Nǀuu, ǀǀXegwi, and ǀʼAuni (unless the ǀʼAuni word is a borrowing from Nǀuu, in which case the reconstruction is only reliable for the "Narrow !Wi" level). Replacements: Forms in ǀXam and "old Nǀuu" (ǀǀNg!ke) have no known etymology. Reconstruction shape: Alveolar click *!- is reliably reconstructed based on the correspondence between Nǀuu !- and ǀǀXegwi k- (click loss, like in the word for 'person' q.v.).
Number:68
Word:root
ǀXam:
Unclear. The only unambiguous candidate would be !aui, given as the equivalent of 'root' in [Bleek 1929: 71]; however, the same word is translated more specifically as 'wild onion' in [Bleek 1956: 414]. Other possible candidates from the same source include ɳǀũnu "roots, fibres" [Bleek 1956: 352]; ǂʼʰáˤna, pl. ǂʼʰaˤǂʼʰáˤnu "fibrous root" [Bleek 1956: 650] (both from L. Lloyd's records); neither explicitly qualifies as the default word for 'root' (gen.), and it is not even clear if such a generic term existed in ǀXam in the first place.
ǀǀNg!ke:
Not attested.
ǂKhomani:
Not attested.
Nǀuu:!ʰabe-si #1
Sands et al. 2006. According to B. Sands (p.c.), "only one of the Eastern Nǀuu speakers knows this word", and a more suitable candidate may be ɡǂao-si ~ ɡǂãũ-si 'root of shepherd's tree (Boscia albitrunca)', also used in the meaning 'root (general)'. The issue needs further elaboration.
ǀǀXegwi:
Not attested.
ǀ'Auni:ǂau-si #2
Bleek 1937: 219; Bleek 1956: 658. Meaning glossed as 'small roots', therefore, the lexicostatistical entry is dubious (although the word itself is not, since it is clearly the same as Nǀuu ɡǂao-si ~ ɡǂãũ-si 'root of shepherd's tree'. The earlier, less reliable source lists the form ɳ!uma 'root' [Bleek 1929: 71], but in [Bleek 1937: 214] the exact same form is listed with the meaning 'tobacco' (?).
ǀHaasi:
Not attested.
Proto-!Wi:
Not reconstructible due to lack of proper attestation.
Bleek 1956: 113, 116. Reduplicated verbal stem, applicable to round objects (e. g. 'sun', 'egg'). Transcribed as kwórre-kwórre ~ ku̯órrɛː-ku̯órrɛː by W. Bleek. Quoted as kɔ̀re-kɔre ~ kwǝre-kwǝre in [Bleek 1929: 71]. Cf. also ǀǀérritǝn-ǀǀérritǝn [Bleek 1956: 570], also glossed as 'round' in one example (applied, however, to "a small troop of springbok", so the exact meaning may be 'to surround, encircle').
Bleek 1929: 71. Dubious; the word is not attested at all in the large dictionary [Bleek 1956].
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. Items in ǀXam and ǀǀNg!ke are quite likely cognate with each other, but their authenticity is not confirmed in any newer sources, and the forms belong to the expressive lexicon layer, not to mention the easy possibility of their areal diffusion.
Bleek 1956: 372, 412. Same word as 'earth' q.v. Quoted as !ãũ in [Bleek 1929: 71]. For all the phonetic/transcriptional variants with discussion, see under 'earth'.
ǀǀNg!ke:
Not attested. Should probably be the same word as 'earth' q.v., but there is no explicit confirmation of this in either [Bleek 1929] or [Bleek 1956].
Sands et al. 2006. Same word as 'earth' q.v. Quoted as !ʼãũ in [Westphal 1965: 144].
ǀǀXegwi:
Not attested.
ǀ'Auni:
Not attested.
ǀHaasi:
Not attested.
Proto-!Wi:*!ʼãũ ~ *!qʼãũ
In both of the languages where the meaning 'sand' is attested, it is expressed by the same word as 'earth' q.v.; there is no factual reason to suggest that this polysemy was not present in Proto-!Wi as well.
Bleek 1956: 654, 655. Transcribed as ǂá-kkǝn ~ ǂá-kka ~ ǂaˤ-ǂáˤ-kkǝn by W. Bleek. Quoted as ǂa-kǝn in [Bleek 1929: 71]. This is rather transparently the primary equivalent for 'say' in ǀXam, being the main means of introducing direct speech (cf. ha oakǝn ǂakka hĩ... "their father says to them...", etc.). The form is morphologically complex, with one of ǀXam's most common derivational suffixes that usually serves as a nominalizer. Other potential synonyms, listed in [Bleek 1929: 71], include: (a) ka, glossed as 'to wish, intend, think, say' in [Bleek 1956: 73] and, judging from the contexts, generally applied to mental rather than verbal activity; (b) kuːi = kúːi ~ kúi-tǝn "speaking" [Bleek 1956: 104], also glossed in the meaning 'think' and featuring in a very limited set of contexts/examples.
Bleek 1929: 71. Dubious, since the word is not found in the large dictionary [Bleek 1956]. However, ka as the basic word for 'say' is strongly supported by external data (ǂKhomani, Nǀuu), and the other quasi-synonym listed in [Bleek 1929: 71], ǀǀãla, is clearly ineligible, since its meaning is confirmed as 'to speak' rather than 'say' in [Bleek 1956: 554], where it is quoted as ǀǀalaʔã ~ ǀǀàˤlaˤ, cf.: a ǀǀàˤlaˤ ǀǀŋ-!wá !winkʸa "thou speakest the Bushman language".
Bleek 1937: 203; Bleek 1956: 95. Meaning glossed as 'says, cries' (in [Bleek 1937]) or as 'to say, scream' (in [Bleek 1956]). This is one of the two main verbs that introduce direct speech in ǀʼAuni. The other is ǀu ([Bleek 1937: 210]; [Bleek 1956: 322]); the difference is unclear - for instance, within one text dictated by a single woman ([Bleek 1937: 198]) the narrative begins with ǀu used several times, after which the narrator suddenly switches to ko for no apparent reason. We have to count both forms as synonyms.
Distribution: Preserved everywhere except for ǀXam (where it seems to have undergone a semantic shift) and ǀHaasi (if Story's source is to be trusted). Replacements: (a) the source of ǀXam ǂa- remains unclear, but its textual attestation suggests that it had managed to more or less permanently replace the older ka as the basic equivalent for 'say'; (b) Common Lower Nǂossob *ǀu- 'to say', functioning as a complete synonym of ko in ǀʼAuni, finds no reliable parallels in "Narrow !Wi". Reconstruction shape: The verb is commonly represented by either the variant *ka or the variant with a labial vowel (usually *ku), probably reflecting the usual (poorly understood) scenario of vocalic fluctuation in basic (CV-type) verbal roots.
Bleek 1956: 341, 345, 347. Transcribed as ɳǀaː ~ ɳǀãː ~ ɳǀɛː ~ ɳǀĩː ~ ɳǀiː ~ ɳǀi by W. Bleek. Quoted as ɳǀiː ~ ɳǀa in [Bleek 1929: 72]. Vocalic fluctuation is a typical ǀXam feature of short verbal stems and may reflect mergers with various class suffixes; the original root is arguably best denoted as ɳǀV-. Secondary synonym: ǀǀxʼóen ~ ǀǀxʼóeŋ [Bleek 1956: 607], glossed as "to see, look" (transcribed as ǀǀxʼoen ~ ǀǀxʼoein by W. Bleek; quoted as ǀǀkʼoen in [Bleek 1929: 72]); some attested contexts suggest 'see' as a better semantic equivalent than 'look', but, overall, it is statistically evident that ɳǀV is the more basic item in this meaning.
Bleek 1956: 341, 345, 347; Bleek 1929: 72; Bleek 2000: 24. Vocalic fluctuation, same as in ǀXam, is typical of short verbal stems. Secondary synonym: ǀeːn(-ya) ([Bleek 1929: 72]; [Bleek 1956: 309]), attested in one dubious example (n ǀeːnya ǀǀwe "I have seen the animals"); clearly a different root, but statistically quite unlikely to represent the basic equivalent for the meaning 'see'.
Maingard 1937: 245, 246, 252. Secondary synonym: ǂai [ibid.]. The semantic difference is unclear and suggests incorrect translation (cf., among Maingard's examples, such a strange opposition as a ǂai ŋ "you see me", but a ɳǀaʔ a "you see yourself"): most likely, ǂai is actually 'look' rather than 'see'. This suggestion is further corroborated by the occasional translation 'look' even in [Maingard 1937] itself, as well as the more modern data from Nǀuu.
Ziervogel 1955: 36, 49. The past tense form is transcribed as ɳǀaː [ibid.]. Infinitive: ɳǀa-ziŋ, imperative forms: ɳǀaː (sg.), ɳǀa-u (pl.) [Ziervogel 1955: 48]. Quoted as ɳǀi ~ ɳǀa in [Lanham & Hallowes 1956: 105]; as ɳǀeː ~ ɳǀaː in [Bleek 1929: 72]; as ɳǀaː ~ ɳǀe in [Bleek 1956: 341, 345].
Bleek 1937: 211; Bleek 1956: 341. Cf. also the phonetically similar form ǀʌn ([Bleek 1937: 210]; [Bleek 1956: 327]; quoted as ǀɯn in [Bleek 1929: 72]). It is not clear if this is a separate root (and if yes, what is the difference between the two), or if it is yet another morphological variant of the same stem ɳǀV-.
Distribution: Preserved in all daughter languages. Reconstruction shape: Nasalized dental click is preserved in all daughter languages and reconstructed without hesitation. Vocalic fluctuations are quite typical for all CV-type verbal roots.