Bleek 1956: 342. Plural form: ɳǀãː ~ ɳǀãŋ. Transcribed by W. Bleek as ɳǀaː, emphatic and plural forms ɳǀaŋ ~ ɳǀaːŋ ~ ɳǀáŋŋan. Quoted as ɳǀaː in [Bleek 1929: 46].
Bleek 1937: 211; Bleek 1929: 46; Bleek 1956: 342. The situation is complex, since another word for 'head' is xːuu [Bleek 1937: 208], quoted as xú ~ xúu in [Bleek 1956: 261]. Although it continues the common !Wi root *xu 'face' (see the comparative evidence in [Bleek 1956: 261]), in several attested ǀʼAuni examples the meaning is clearly 'head', not 'face', cf.: ti tani ke xú "carry on the head" (hardly "on the face"!). D. Bleek, therefore, suggests that ɳǀaː is, in fact, a Nǀuu/ǂKhomani word in ǀʼAuni, along the same lines as in her argument for 'mouth' q.v. The situation here is, however, different from 'mouth', because: (a) only ɳǀaː and not xú is found in the meaning 'head' in Bleek's early records, published in [Bleek 1929]; (b) unlike the words for 'mouth', the word ɳǀaː 'head' is pan-Southern Khoisan, and its re-introduction into ǀʼAuni from ǂKhomani is a rather complex scenario. It would probably be easier to simply think of the situation in terms of "transit synonymy": xu 'face' gradually replacing the original ɳǀaː 'head'. As is common in such cases, we list both synonyms.
Story 1999: 22. The morpheme ŋ= is probably the 1st p. possessive prefix.
Proto-!Wi:*ɳǀa
Distribution: Preserved throughout "Narrow !Wi"; in Lower Nǂossob, possibly replaced either already at the proto-level or in individual dialects. Replacements: In either Proto-Lower Nǂossob or a sub-section of ǀHaasi-ǀʼAuni dialects, replaced with *xu, originally 'face' (cf. ǀXam xú 'face' [Bleek 1956: 261], etc.); the semantic shift {'face' > 'head'} looks quite reasonable. Reconstruction shape: Correspondences are trivial. Semantics and structure: The original paradigm is reconstructible as sg. *ɳǀa, pl. *ɳǀa-ŋ.
Bleek 1956: 206, 239, 240. Transcribed as tːú ~ tːúi ~ tuːi ~ tːóa by W. Bleek. Quoted as tuː ~ tuːi ~ tum in [Bleek 1929: 46]. All forms traceable back to the stem *tu; vocalic variation is likely due to merger with various class suffixes, as in most simple verbal stems.
Ziervogel 1955: 36, 40, 52. The short stem, according to ZIervogel, is only used in the future tense. Past tense stem: tu-wa; present tense stem is tu-bi. Imperative forms glossed as to (sg.), to-u (pl.) [Ziervogel 1955: 48]. Cf.: ʔa ʔa tuʔbi ʔa-ʔe "do you not hear" [Lanham & Hallowes 1956: 115]. Quoted as tuːi ~ taːã in [Bleek 1929: 46], but only as tuːi in [Bleek 1956: 240].
Bleek 1937: 207; Bleek 1956: 239, 240. The earlier source lists this word as taːã ~ taːa [Bleek 1929: 46]; this either reflects a contraction with a class marker or particle or, more likely, represents a different root (cf. also tiãn 'to feel' [Bleek 1937: 206]).
ǀHaasi:
Not attested.
Proto-!Wi:*tu
Distribution: Preserved in all daughter languages. Additionally, cf. ǀǀKxau tú, Seroa tu [Bleek 1956: 239]. Reconstruction shape: Correspondences are regular and unequivocally point to *tu (with predictable palatalization of the initial dental consonant in most dialectal varieties of Nǀuu).
Bleek 1937: 208; Bleek 1956: 271. Quoted as ǀʼɛː in [Bleek 1929: 46], but that source has an additional synonym in the same meaning: ǂkʼa, not confirmed in later sources (possibly = ǂan 'chest' q.v.?).
Story 1999: 22. Initial n= is probably the 1st person possessive prefix; the second click-containing morpheme -ɡǀe is unclear.
Proto-!Wi:*ǀe ~ *ǀʼe
Distribution: Preserved in all daughter languages except for ǀǀXegwi. Additionally, cf. ǀǀKxau ǀae ~ ǀai-si, ǀǀKuǀǀe ǀʼɛ̃ː id. [Bleek 1956: 271, 296]. Replacements: In ǀǀXegwi, the old word seems to have been replaced with a Bantu borrowing. Reconstruction shape: The stem shows a somewhat chaotic distribution of zero (velar) click efflux vs. glottalized efflux in daughter languages (usually glottalized in ǀXam and ǀʼAuni, non-glottalized in Nǀuu and ǀHaasi). Ignoring this glottalization is impossible, since it shows up in too many sources; the situation possibly reflects glottalized vocal articulation (i. e. *ǀeʔe), as in several other such cases, but this solution is not conclusive.
Bleek 1956: 567, 574. Plural form: ǀǀʰe-ǀǀʰẽi ~ ǀǀʰẽ́i-ǀǀʰẽ́i ~ ǀǀé-ǀǀéi ~ ǀǀẽ́i-ǀǀẽ́i (with reduplication). Transcribed as ǀǀɛ̃:, plural (emphatic?) ǀǀẽi-ǀǀẽ́i-yaŋ by W. Bleek. Quoted as ǀǀẽ́ĩ́, pl. ǀǀẽ́ĩ́-ǀǀẽ́ĩ́ in [Bleek 1929: 47]. Of note is the transcription's fluctuation between the aspirated and the zero efflux, possibly indicative of a "non-trivial" type of articulation not detected properly by W. Bleek and L. Lloyd. None of the materials distinguish this word from 'tooth' q.v., although external data very clearly speak in favour of their separate origin.
Bleek 2000: 18. Plural form: ǀǀãĩŋ ~ ǀǀãĩŋ-sa. In [Bleek 1956: 568] the word is confused with 'tooth' q.v., and most textual examples are on 'tooth', although the plural variant ǀǀẽĩŋsa ~ ǀǀãĩŋ for 'horn' is also mentioned. Quoted as ǀǀẽĩ, pl. ǀǀẽĩn-sa in [Bleek 1929: 47].
Bleek 1937: 216; Bleek 1956: 568. Meaning glossed as plural: 'horns'.
ǀHaasi:
Not attested.
Proto-!Wi:*ǀǀẽĩ
Distribution: Preserved in all daughter languages (although see further on the somewhat divergent form in modern Nǀuu). Reconstruction shape: Most of the sources agree on the lateral click influx, zero (velar) accompaniment, and a front root vowel for the original protoform. The only uncomfortable exception is the form attested in Modern Nǀuu (ǀǀqʰoe-), which suggests a uvular aspirated efflux instead, as well as a labial component in the root vocalism. These correlations are quite irregular and find no explanation; however, due to lack of alternate etymologies, some degree of phonetic similarity, and the fact that uvular accompaniments had not been generally recognized by researchers prior to the modern era, we tentatively mark the form as an etymological cognate, pending further research on the issue. Additionally, the root is almost never encountered without nasalization, which may be part of the root or represent fusion with an old nasal class marker (< *ǀǀe-iŋ ?).
Bleek 1956: 140. Emphatic form: ŋ-ŋ. Also attested in a rare phonetic variant n, as well as m before the following labials (by assimilation). Possessive forms, according to W. Bleek, include the variants ŋ ~ ŋ-ka ~ ŋ-ta ~ ŋ-ga. Both W. Bleek and L. Lloyd also indicate the existence of a velar stem ka ~ ke ~ ki [Bleek 1956: 74, 85, 91], rarely, if ever, found in as the subject of the main clause and mostly limited to various bound usage in subordinate clauses. Available data are too ambiguous, however, to allow us to treat it as a regular "indirect stem-type" synonym.
Bleek 1956: 140; Bleek 1929: 49; Bleek 2000: 21. D. Bleek notes that both variants are freely interchangeable. The rare positional variant m (before words starting with a labial consonant) is mentioned in [Bleek 1956: 132] and [Bleek 1929: 49]. Note also a special dative form ke, e. g. ha saː ke !xe e ɳǀa "she gives me a cloth for the head" [Bleek 2000: 22].
Maingard 1937: 244. All four variants are mentioned as representing the subject form; in the object position only ŋ is encountered. Transcribed as na ~ ɲa in [Doke 1936: 63]; both variants, as subject pronouns, are also said to be in free variation.
Ziervogel 1955: 46. Cf. also the emphatic (absolute) form: ʔn-ʔe ~ ʔin-ʔe; the object form ʔne ~ ʔŋe; the possessive form ŋe ~ ŋ [Ziervogel 1955: 45-47]. The absolute form is quoted as ʔiŋ-ʔe ~ iŋ-ʔe in [Lanham & Hallowes 1956: 108]. Quoted as ŋ ~ aŋ ~ am in [Bleek 1929: 49]; as am ~ aŋ ~ n ~ ŋ in [Bleek 1956: 9, 10, 141] (the variant am is an assimilated form, used before words beginning with labials).
Bleek 1937: 197; Bleek 1956: 132, 141, 142. The variant m is the result of samdhi, encountered only before words beginning with labial consonants. Quoted as ŋ ~ ŋ-ŋ in [Bleek 1929: 49]. Possessive forms, as per [Bleek 1937: 197], include: n ~ n-ka ~ n-ga ~ m; there may also be a special dative form ki 'to me'.
Story 1999: 31. Cf. also the emphatic forms: gʸà=ŋa ~ gʸa=ŋ ~ ka=ŋ (the last form is not marked as "emphatic" by Story, but also represents a contraction with a prefix).
Proto-!Wi:*ŋ
Distribution: Preserved in all modern languages. Additionally, cf. ǀǀKxau ŋ ~ n, ǀǀKuǀǀe ŋ, Seroa ʔn, G!ãn!e n ~ ŋ [Bleek 1956: 141]. Reconstruction shape: The primary and most common form of the root is that of a syllabic velar nasal. Everything else is the result of contextual assimilations or combinations with various emphatic particles.
Bleek 1956: 293, 315. Transcribed by W. Bleek as ǀá ~ ǀíː. Vocalic variation is typical of most simple verbal stems and is explained through merger with various class markers. Not to be confused with ǀʼa 'to die' q.v.: despite superficial similarity, the two stems are very consistently distinguished both in W. Bleek's and L. Lloyd's records through their effluxes (glottal stop for 'die', zero or aspiration for 'kill'). Quoted as ǀiː ~ ǀaː ~ ǀʼaː in [Bleek 1929: 50] (where this exact confusion has actually taken place).
Bleek 1956: 293, 313, 315. Forms quoted as ǀʰiː 'to kill' : ǀʰaː 'killed' in [Bleek 2000: 24], although text examples in [Bleek 1956] show that the vocalic gradation is dependent not on tense or voice, but, most likely, on class characteristics of the accompanying nouns. Quoted as ǀiː ~ ǀaː ~ ǀʼaː in [Bleek 1929: 50] (where, as in ǀXam, the word has been confused with the entirely different stem ǀʼa 'to die' q.v.).
Ziervogel 1955: 51. The simple stem is said to function as the future tense. Present tense stem is ƛiŋ-we; past tense stem is ƛiŋ-wa. Somewhat dubious. The meaning on p. 51 is glossed as 'hit, strike', but cf. on p. 62: ʔi ƛeŋ-we ǀʰwi "we kill a bird", ʔi ƛeŋ ǀʰuriŋ "we kill a spur-winged goose". The same word is quoted as ƛʼeuŋ 'to hit' in [[Lanham & Hallowes 1956: 100].
Distribution: Found only in the ǀXam-Nǀuu cluster. Replacements: (a) ǀǀXegwi ƛiŋ could reflect something like Proto-!Wi *ǂiŋ or *ǂeŋ. However, the word has no external parallels, and its original meaning, as can be seen from attested examples, is clearly 'to hit' rather than 'to kill'; (b) ǀHaasi !au 'to kill' = ǀʼAuni !au 'to beat; to knock down' [Bleek 1956: 411]; in this case, the local replacement {'to beat' > 'to kill'} is more than evident. It is important to note that ǀXam-Nǀuu *ǀʰa is the only etymon in the whole group whose semantics is strictly confined to 'to kill', which is an additional argument for regarding it as a better candidate for Proto-!Wi status. Reconstruction shape: Modern Nǀuu shows that the root has to be reconstructed with an aspirated efflux. Original vocalism fluctuates between *a and *i, reflecting an obscure morphological pattern.
Bleek 1956: 349. Plural form: ɳǀu-ɳǀua-dːe ~ ɳǀũ-ɳǀũa-dːe. Transcribed as ɳǀo̯á-ɳǀo̯aː-déyakǝn (pl. form) by W. Bleek. Quoted as ɳǀɔaŋ in [Bleek 1929: 50].
Bleek 1937: 217. Quoted as ǀǀoiǀǀoi in [Bleek 1929: 50]. In [Bleek 1956: 584], this form, quoted as ǀǀweː-ǀǀwe, is marked as the (reduplicated) plural correlate to the singular ǀǀoe. This would seem logical, but, for some reason, only the reduplicated form is attested in [Bleek 1937], glossed as singular 'knee'.
ǀHaasi:
Not attested.
Proto-!Wi:*ɳǀu-aŋ ~ *ɳǀo-aŋ
Distribution: Attested in most varieties of "Narrow !Wi". Cf. also Seroa gno-ma-teŋ 'knee' [Bleek 1956: 47], where g- transcribes a click. Replacements: "Narrow !Wi" *ɳǀu-aŋ is opposed to ǀʼAuni (Lower Nǂossob) *ǀǀoe ~ *ǀǀõẽ. There are two reasons why the !Wi equivalent is preferable: (a) purely technical (the word is reconstructible to a much higher chronological level); (b) more importantly, ǀʼAuni ǀǀoe coincides segmentally with the basic Central Khoisan equivalent for 'knee', and, although this does not per se prove borrowing, makes it quite probable. Reconstruction shape: Dental click with nasal efflux is regularly reflected in all daughter languages, as is the labial vowel (although its exact quality is hard to determine). The segment -aŋ in ǀXam and the corresponding nasalization of the vowel in Nǀuu is most likely of suffixal origin, since in ǀǀXegwi the same root is encountered with a different suffixal extension: *ɳǀu-ma ~ *ɳǀo-ma.
Bleek 1956: 643. Transcribed as ǂʼeńn ~ ǂʼenńa by W. Bleek. Quoted as ǂʼen ~ ǂʼen-a in [Bleek 1929: 51]. The connection between these forms and ǂʼẽː ~ ǂʼĩː 'to think, remember' [Bleek 1956: 642, 652] is unclear, as is their connection to phonetically similar Central Khoisan forms with the same meaning (cf., for example, Nama ǂʼan 'to know', ǂʼãĩ 'to think'); nevertheless, due to external parallels within !Kwi itself, we do not count this as a borrowing (at least, not a recent one). In [Bleek 1929: 51] another stem, ǀǀwaka, is listed as synonymous, but in [Bleek 1956: 596] it is glossed as 'to understand, be wise, clever, cunning', and textual examples confirm that the word is unlikely to have simply meant 'to know'.
Bleek 1937: 217, 218; Bleek 1956: 601, 631. It is not quite clear if these two forms are really phonetic/morphological variants or represent different roots.
Distribution: An isogloss between ǀʼAuni and Nǀuu. Replacements: (a) In ǀXam, there is some evidence for a secondary merger between 'to think' and 'to know': ǀXam ǂʼeńńː 'to know' = ǀǀNg!ke ǂẽ 'to think' [Bleek 1956: 642], ǂKhomani ǂʼĩ 'to think, thoughts' [Maingard 1937: 257]. It is not clear if, on the next chronological level, this word had been borrowed from Central Khoisan *ǂʼan 'to know; to think' or not, but, in any case, of all the languages that share this form, it is only ǀXam for which the semantics of 'to know' has been attested; (b) ǀǀXegwi ci- has no etymology whatsoever; (c) ǀHaasi ǀüma has a strong parallel outside !Wi in !Xóõ ɡǀûmã 'to know' (Taa branch of !Wi-Taa). Under different circumstances, this fact would have made the ǀHaasi form an optimal candidate for Proto-!Wi 'to know' (external comparison argument). However, this would mean that the Nǀuu-ǀʼAuni isogloss has to be definitively judged as either a homoplasy or a result of borrowing (presumably, from Nǀuu into ǀʼAuni). In order to make that judgement, one has also to be sure that the ǀHaasi-!Xóõ connection might not be due to historical contact (given the geographical location of !Xóõ and Lower Nǂossob languages, this may not be excluded). Consequently, for the present we prefer to formally regard the ǀHaasi form as an innovation, not an archaism, since its internal distribution is weak compared to Nǀuu-ǀʼAuni *ǀǀxae. Reconstruction shape: We tentatively project the Nǀuu shape of the form onto the proto-level, although vocalic correspondences between ǀʼAuni and Nǀuu require further study.
Number:46
Word:leaf
ǀXam:
Not attested. The form ɡǀuḿmː in [Bleek 1956: 283] is tentatively glossed as 'leaf (?), stick (?)', based on a context from L. Lloyd's records in which the exact meaning is impossible to determine properly. Words with the meaning 'leaf', as such, are extremely met in South Khoisan languages as a whole.
ǀǀNg!ke:xerroː1
Bleek 1956: 259. Quoted as xero in [Bleek 1929: 52]. In the former source the meaning is glossed as 'leaves, foliage' (ɡʘo kan xerroː "the tree's leaves").
ǂKhomani:kānɾū2
Doke 1936: 72. Not attested in Maingard's data. An alternate, phonetically similar, but probably etymologically incompatible form, also glossed as 'leaf', is kāŋkāˤmū-sí ~ kã̄kã̄ˤmū-sí [Doke 1936: 69, 75]. Judging by the structure of the latter word, it is almost certainly a borrowing, but the source is unclear. The form kānɾū is also phonetically questionable (the cluster -nɾ- is unique for this entry and should indicate a non-!Kwi origin).
Nǀuu:blaːr-si-1
Sands et al. 2007: 60. A transparent borrowing from Afrikaans blaar.
ǀǀXegwi:li=kʰasi-zi-1
Lanham & Hallowes 1956: 112. Borrowed from Swazi li=kʰasi id. Not attested in [Ziervogel 1955].
ǀ'Auni:
Not attested.
ǀHaasi:
Not attested.
Proto-!Wi:
Not reconstructible. Due to areal conditions, the word 'leaf' consistently shows highly restricted usage and is in most cases /re/introduced as a borrowing.
Bleek 1956: 185, 196, 198. Transcribed as tá ~ taː ~ tːɛ̃ː ~ tẽː ~ teːŋ ~ teːn by W. Bleek. Quoted as taː ~ teːn ~ tiŋ in [Bleek 1929: 53]. Polysemy: 'to lie / to lie down'; also attested in causative meanings ('to lay (down)'). Vocalic variability is typical of verbal roots with short stems and probably reflects mergers with various class markers, although it is not clear which of the variants should be considered closer to the "pure" root (probably *ta, but it is also possible to denote the root as a monoconsonantal *t-). On the possible synonym ǀʼũːŋ see under 'sleep'.
Bleek 1956: 91, 202. Quoted as tu ~ tiä ~ kiä ~ keːŋ in [Bleek 1929: 53]. The form tieŋ is said to represent "past tense" in [Bleek 1956: 202], but its phonetic variant kieŋ [Bleek 1956: 91] ~ keːŋ [Bleek 1956: 87] is, however, said to be employed "after verbal particles". In [Bleek 2000: 18] it is explicitly acknowledged that tiä and kiä represent alternative pronunciations, which should be interpreted as one more example of a typologically frequent (for Khoisan languages) palatal articulation of t- before front vowels. The verb is frequently translated as 'sleep' in Bleek's examples, but 'lie' is almost certainly the primary meaning; for discussion, see 'sleep'.
Sands et al. 2007: 61. Meaning glossed as 'lie (down)'.
ǀǀXegwi:ǀaː #2
Bleek 1929: 53; Bleek 1956: 294. The latter source gives a textual example: haŋ ǀaː, ha ʘueŋ "she lies, she sleeps". The former source quotes the word !xwaː as a synonym, but its existence is not confirmed in [Bleek 1956]. Unfortunately, the word "to lie" is not attested properly in more reliable sources. In [Lanham & Hallowes 1956: 99], the phrase iŋ-ʘiɲe is translated as "I lie down", but the usual meaning of the underlying verb, both in ǀǀXegwi and in related languages, is 'to sleep' q.v.; without additional confirmation, we would rather regard this as a potential mistranslation and leave the slot empty. In [Lanham & Hallowes 1956a: 47], the form kala is adduced with the meaning 'lie down' (and explained as a potential borrowing from giTonga kʰala 'stay, sit'), but it is not clear whether it also has the required static semantics in ǀǀXegwi.
Bleek 1937: 207. Meaning glossed as 'down, to lie down, come down'; cf. also tũa 'to lie curled up' and the derived stem tõã-a 'to lay down, to bury' [ibid.]. In [Bleek 1929: 53], the meaning 'to lie down' is glossed as tũ ~ toa. Textual examples show both a dynamic and a static meaning for the verb: cf. kʰai tòa "water comes down" vs. o toa ki ɡ!ã "(he) lies on the ground" [Bleek 1956: 206]. Cf. also: ǂɔ̃ː di se tũa "the dog lies curled up" (maybe just "lies" as such?) [Bleek 1956: 240]. There is also a different stem, ɡǀò ([Bleek 1937: 209], [Bleek 1956: 280]), glossed only as 'to lie down', and attested only in examples with "dynamic" semantics.
ǀHaasi:
Not attested.
Proto-!Wi:*ta
Distribution: Preserved everywhere, with the possible exception of ǀǀXegwi. Additionally, cf. ǀǀKxau da ~ ta 'to lie' [Bleek 1956: 19, 185]. Replacements: The situation with ǀǀXegwi remains unclear; in any case, the hypothetical ǀaː 'to lie' has no reliable external etymology. Reconstruction shape: The most frequent variant is *ta, with the first consonant regularly palatalized in Nǀuu. The variants *teŋ (< *ta-iŋ?) in ǀXam, *tu in ǂKhomani, and *toa in ǀʼAuni should be regarded as morphological variants.
Sands et al. 2006. The two latter forms are said to represent the Eastern dialect.
ǀǀXegwi:
Not attested.
ǀ'Auni:
Not attested.
ǀHaasi:
Not attested.
Proto-!Wi:*ɳǀǀaN
Distribution: Preserved in all languages where attested. Additionally, cf. ǀǀKxau ŋaŋa [Bleek 1956: 143] with regular elision of the click influx. Reconstruction shape: Labialization in ǀXam seems to be secondary, as in multiple similar cases. A nasal is always present in the stem coda, but the variants fluctuate rather chaotically between *-aŋ, *-an, and *-ain; the optimal scenario for working out this problem has not yet been found, but in any case, the nasal seems to be an inherent part of the root rather than a suffixal addition.
Bleek 1956: 501. Polysemy: 'tall / long / high'. Plural form: !xó-!xó-ka (with reduplication). Transcribed as !xoːwa, pl. !xo-!xó-ka ~ !xó-!xó-kǝn by W. Bleek. Quoted as !xoːwa in [Bleek 1929: 55]. This adjectival stem is clearly derived from !xoː 'to grow up, climb up; make upright, make tall' [Bleek 1956: 500].
Bleek 1956: 267; Bleek 1929: 55; Bleek 2000: 23. Applied to objects (ha !u ǀʼaː "his (ostrich's) neck is long") as well as time periods (ɡǀǀã ǀʼaː "the night is long").
Bleek 1937: 208; Bleek 1956: 270. Meaning glossed as 'big, long, tall', although 'big' is an incorrect addition (see notes on 'big' for more details). Entirely different equivalent for 'long' found in [Bleek 1929: 55]: xaras (not confirmed by later research or external comparison).
ǀHaasi:
Not attested.
Proto-!Wi:*ǀʼã
Distribution: Preserved everywhere except for ǀXam. Replacements: In ǀXam, the only word glossed with the meaning 'long' is !xóː-wa, related to ǀǀNg!ke !xoːwa 'tall, big' [Bleek 1956: 501] = Nǀuː !xoː id. (see under 'big'). This allows to tentatively suggest a semantic shift {'big' > 'long'}, although much depends on the degree of accuracy in the glossing of ǀXam data. Reconstruction shape: Correspondences are relatively straightforward. Vowel nasalization seems to be innate in the root.
Lanham & Hallowes 1956: 105. Not attested in [Ziervogel 1955].
ǀ'Auni:
Not attested.
ǀHaasi:
Not attested.
Proto-!Wi:*ɳʘu-
Distribution: Preserved in all languages where attested. Reconstruction shape: Initial labial click with nasal accompaniment is attested in all languages and automatically projected onto the proto-level. Vocalism is harder to reconstruct; most dialects agree on a labial vowel, but ǀǀXegwi does not, and it is not excluded that an assimilative process was at work in at least some of these dialects. Still, we provisionally follow the "majority rule". Semantics and structure: For the ǀXam-Nǀuu cluster, the word is reconstructible as the complex stem *ɳʘu-iŋ, with a productive nominal suffix (of class?).
Number:51
Word:man
ǀXam:!wi=gwaiː1
Bleek 1956: 447, 466. Transcribed as !ü go̯áːi by W. Bleek (for more on the shorter variant !u ~ !ü see under 'person'). Quoted as !wi=gwai in [Bleek 1929: 56]. The noun !wi by itself normally means 'person' without specification of gender (q.v.). The compound form !wi=gwaiː literally translates as 'person-male', where gwaiː = 'male' (of human beings as well as animals) [Bleek 1956: 52]. It is unclear how frequent this compound was in actual speech, but there are some neutral contexts in L. Lloyd's records suggesting that it could have indeed been the default designation for 'man' as opposed to 'woman' in the singular number. In the plural number, the usual designation for 'men' is suppletive: tːú-kǝn (Lloyd), tú-kǝn (W. Bleek) [Bleek 1956: 239]. The singular stem *tu is, however, not attested in ǀXam at all, and the lexeme is therefore ineligible for inclusion in the list. The lexicostatistically relevant morpheme is gwaiː.
Bleek 1956: 240; Bleek 2000: 19. Plural form: tu-kǝn ~ tũ-nyǝn ~ tu-ŋǝn. Quoted as tũ, pl. tũŋǝn, túːkǝn in [Bleek 1929: 56]. This is the regular equivalent for 'male human being' as opposed to 'woman' (cf.: ǀake, hŋ ǀõːˤ, tukǝn ɳǀǀa ke, hŋ ǀõːˤ "women, they dance, those men, they dance" [Bleek 1956: 240]). The stem ɡ!oː [Bleek 1956: 383] more properly refers to 'male' being in general (cf.: kue ɡ!oː e "male ostrich it is" [ibid.]), although is occasionally used to designate male people as well (cf.: a ɡǀaiki ha e, a ɡ!oː ki e "this is a woman, that is a man (= male)" [ibid.]).
Maingard 1937: 239. Functions both as the independent noun 'man' and the semi-suffix 'male', attached to names of animals (e. g. !ai ǂoː 'male gemsbok', etc.). Transcribed as ɡǂō (with a voiced efflux) in [Doke 1936: 85].
Miller et al. 2009: 155. Meaning glossed as 'man' (no text examples, so it is not clear if the exact semantics is that of 'male human being' or 'person'; external data clearly speak in favor of the former). Suppletive plural: ɕuː-ke [Miller et al. 2009: 157]. Quoted as ǂó, pl. ɕú-kwe in [Westphal 1965: 139].
Ziervogel 1955: 38. Meaning glossed as 'male'. In [Lanham & Hallowes 1956: 112], the meaning 'man' is rendered by the compound form kwi-ƛʼoː, literally 'person' + 'male' (cf. 'woman' q.v.). Quoted as ǀǀɔː in [Bleek 1929: 56]; as ǀǀo in [Bleek 1956: 582].
ǀ'Auni:bɛ4
Bleek 1937: 201; Bleek 1929: 56. Quoted as be ~ bɛ in [Bleek 1956: 15]. Glossed as 'man, male' in [Bleek 1937: 201].
ǀHaasi:biː4
The word n=ǀʰã̀ː is glossed as 'man' in [Story 1999: 22], but textual examples rather suggest the meaning 'husband', cf.: n=ǀhã̀ː a "it is my husband" [Story 1999: 24] (the concatenation with the 1st p. possessive prefix n= further confirms this). The word biː is encountered in the phrase biː a 'it is a man' [Story 1999: 25], and the meaning 'man' is confirmed by the same situation in the closely related ǀʼAuni.
Proto-!Wi:*ǂo ~ *ǂʼo #
Distribution: Preserved in most (but not all) dialects of Nǀuu and in ǀǀXegwi. Additionally, cf. ǀǀKxau ǂoː 'man, male, person' [Bleek 1956: 662]; ǀǀKuǀǀe tʼo 'male' [Bleek 1956: 206]. Replacements: (a) In some dialects of Nǀuu, including Bleek's ǀǀNg!ke, the singular form *ǂo 'man, male' seems to have been replaced by the formerly suppletive plural stem *tu 'people; men'; (b) the origins of ǀXam =gwai 'male' are obscure; it is not even completely excluded that ǀXam !u=gwai < *!u=ɡ!wai < *!u=ɡǂo-ai with extra suffixation and irregular click loss (through dissimilation of two clicks in one compound formation?), in which case there would be no need to postulate a replacement. However, this scenario, resting on several unprovable assumptions, is somewhat far-fetched, and we currently prefer to regard the ǀXam situation as a lexical replacement of unknown origin; (c) Lower Nǂossob *be ~ *bi 'man, male' finds no obvious parallels in the other !Wi languages, and its archaicity is quite dubious, since the phoneme *b- is not well reconstructible for Proto-!Wi. Reconstruction shape: Click influx is reliably reconstructible as palatal (with a regular shift to lateral affricate in ǀǀXegwi), but click accompaniment unpredictably fluctuates between zero (velar) and glottal stop, sometimes even within the same language (cf. different transcriptions in different sources on ǀǀXegwi). This could suggest an original *ǂoʔo (glottalization on the vowel, occasionally transferred to the consonant), but more research has to be carried out on the issue. Semantics and structure: The stem *ǂo ~ *ǂʼo is sometimes employed on its own, but just as often functions as part of the compound *!ui-*ǂo 'person-male', i. e. the original semantics is probably 'male' (of any species) rather than specifically 'man' = 'male person'.
Bleek 1937: 201. Meaning glossed as 'man, person'; technically, this word could be either completely synonymous with bɛ or with ǂʼi 'person' q.v. However, the first option is preferable for the following reason: according to [Bleek 1937: 196], the suppletive plural for both bɛ and de 'man' is tutos ~ tutus. In [Bleek 1956: 240], the list of freely interchangeable plural variants is given as tu-ke ~ tu-ku ~ tu-tu-s ~ tu-tu-se, and the accompanying examples clearly support the semantics of 'men = male human beings', since the word is opposed to 'women', cf.: ɡǀɛki a !oeke, tuke na ǀõː ǂʼui "women are clapping, men are dancing" etc. Additionally, in [Bleek 1929: 56], only the word ǂʼi is given as the equivalent for 'person', although, admittedly, this is not a very strong argument due to the poor quality of the source. The fact that da ~ de is not mentioned in [Bleek 1929] at all makes plausible the hypothesis that da ~ de is the original ǀʼAuni term for 'man = male person', and that in the early 20th century it was being replaced by the newer equivalent bɛ, whereas the original suppletive plural was still being retained. Regardless of whether the scenario is true, we have to treat bɛ and da as synonyms.
Ziervogel 1955: 39. Quoted as qʼiŋ in [Lanham & Hallowes 1956: 117] (in the phrase ɲa šaː ʔe qʼiŋ "it is much water"). Quoted as ǀǀxain in [Bleek 1929: 57] and [Bleek 1956: 632].
Bleek 1956: 5, 9, 37, 39. Transcribed as ãː ~ ãːŋ ~ ãŋ ~ eŋ ~ eŋ-eŋ ~ eín-ya ~ ẽĩŋ-ẽĩŋ-ya by W. Bleek. Reduplicated forms are emphatic in nature. Quoted as ãː ~ eŋ in [Bleek 1929: 57]. The word is most likely a ǀXam-exclusive nominal derivate from hã 'to eat' q.v. (although initial aspiration is completely lacking in the nominal forms, its status in the verbal root is probably secondary as well), and, as such, has replaced the older form ʘu̯aiː, preserved only in the specific meaning 'game' ('meat that moves') [Bleek 1956: 685].
Maingard 1937: 246. Transcribed as ʘoe ~ ʘoi in [Doke 1936: 66]. Secondary synonym: hẽ̄-ǂʼīˤ 'meat' [Doke 1936: 77] (a strange compound form with no etymology). Finally, cf. also the word gāɾú [Doke 1936: 69], with the meaning glossed as 'flesh'.
Ziervogel 1955: 52. Quoted as ʘaː in [Lanham & Hallowes 1956: 102]. Quoted as ʘʌaː in [Bleek 1929: 57]; as ʘwaː ~ ʘwaː-gǝn in [Bleek 1956: 687] (recorded with a "special" variety of the labial click, said to be "released absolutely without the sound of a kiss, more like a plosive p" [Bleek 1956: 682]).
Distribution: Preserved everywhere except for ǀXam. Replacements: In ǀXam, the word is preserved only in the narrow meaning 'game'; otherwise, replaced by ãː, a nominal derivative (through conversion) from the verb 'to eat'. The semantic development {'to eat' > 'food' > 'meat'} is typologically common. Reconstruction shape: Reconstruction of the labial click with zero (velar) accompaniment is completely reliable; vocalic reconstruction remains less clear due to possible contractions of the root vocalism with class marker suffixes in different languages.
Bleek 1956: 417, 419. Transcribed as !au!áuru ~ !a!áuru ~ !a!árro by W. Bleek. Quoted as !a!aro ~ !au!auru in [Bleek 1929: 59]. Technically, the word looks like a reduplication, but it could just as easily be a compound, consisting of two parts that are not etymologizable internally (*!au + *!aro).
Bleek 1956: 242, 443, 454. Quoted as !ɔro in [Bleek 1929: 59]. The form turro is mentioned as an "unusual form of !oro"; it is not highly likely that it goes back to a different root, but it does present a curious dialectal enigma.
Distribution: Preserved in all Narrow !Wi languages. Additionally, cf. ǀǀKxau ǂʼɔrɔ, ǀǀKuǀǀe tʼɔlo [Bleek 1956: 207, 675]. Replacements: The Swadesh item 'moon' reflects the binary split between Narrow !Wi and Lower Nǂossob languages, with no etymological way of determining which particular term, Narrow !Wi *ǂoro or Lower Nǂossob *!(ʰ)oi, is more archaic. Naturally, the former is reconstructible at a deeper time level, making it a slightly more probable choice for Proto-!Wi status. Reconstruction shape: The "fluctuating" glottalization of the click efflux (as seen in some Nǀuu dialects, ǀǀKxau and ǀǀKuǀǀe) crops up too frequently to be brushed off as a transcriptional inaccuracy; most likely, the original form was *ǂoʔoro (i. e. contained a glottalized vowel, with glottalization occasionally transferred to or perceived as part of the click efflux).
Number:54
Word:moon
ǀXam:
ǀǀNg!ke:
ǂKhomani:
Nǀuu:
ǀǀXegwi:
ǀ'Auni:
ǀHaasi:
Proto-!Wi:*!ʰoi(-ŋ) #
The Lower Nǂossob equivalent for 'moon'. See notes on *ǂoro ~ *ǂʼoro.