Not attested. [Bleek 1929: 15] gives the form ʒã̀ in this meaning; it is then later quoted as ʒã̀ˤ in [Bleek 1956: 30]. However, not only does this morpheme have no external parallels whatsoever in the meaning 'all', but it also looks very much like a variant of the numeral 'two' q.v., and the sole text example quoting it (i ʒã̀ˤ ǀu ǀxoa 'we are all alive') most likely means just 'both of us are alive', cf. Juǀ'hoan è-cá 'us two' etc.
König & Heine 2008: 71. Quoted as wèēsè in [Heikkinen 1986: 26].
Proto North Khoisan:*wòè-še
Distribution: Preserved in all daughter dialects. No alternate stems. Reconstruction shape: The reconstruction follows the Ekoka variant as phonetically more archaic in its vocalism; the variant *wè-še is also possible. Extra low tone in Ekoka is not, however, confirmed by the rest of the data. Semantics and structure: The form is morphologically complex: the derivation is transparently seen in Ekoka, cf. wȍhà 'forever', wȍhè 'some time ago, already, just' [König & Heine 2008: 71]. The meaning of the suffix *-še, however, remains unknown.
Bleek 1956: 206. Quoted as tɔ̀ː in [Bleek 1929: 17]. Secondary synonym: ɡǂwá [Bleek 1956: 649] ~ ɡǂwa [Bleek 1929: 17]. The latter root is less common in North Khoisan and may easily represent a recent borrowing from Central Khoisan.
Bleek 1956: 649. Quoted as ɡǂwàː in [Bleek 1929: 17]; ɡǂòā in [Snyman 1980: 33]. The latter source adds a secondary synonym: tȁoˤ-tȁoˤ [ibid.]. This may be a more archaic (and "genuine" root) than ɡǂoa, a highly probable borrowing from Central Khoisan, but it is completely absent in Bleek's primary sources for the wordlist.
König & Heine 2008: 34. Translated as 'ash'. The dictionary also lists two more compound forms glossed as 'ashes' (sic): dàʔà šŋ̄ŋ́ and dàʔà !̯ȍʰà, where dàʔà = 'fire'. The first form literally means 'fire-excrement'; the main morpheme in the second, !̯ȍʰà, is suspiciously similar to ɡǀǀȍʰà, which indicates either a case of mistranscription or areal transmission (the latter is quite possible, since !̯ȍʰà is a direct parallel to Juǀ'hoan ɡǂȍȁʰ 'soap', a cultural word most likely transmitted from Central Khoisan *ɡǂóà 'ashes').
Proto North Khoisan:*tȍˤ
Distribution: The original root is well preserved in the Southern and Central clusters. The Northern cluster seems to have displaced it with a borrowing of Central Khoisan origin (*ɡǂóà 'ashes'; the initial lateral click in Ekoka indicates a different root, but is possibly mistranscribed). Reconstruction shape: The reconstruction follows the Juǀ'hoan variant, although it is not in perfect agreement with Angolan !O!Kung tȁoˤ-tȁoˤ in terms of vocalism. (A secondary development *oˤ > aoˤ is not out of the question).
Bleek 1956: 348. Quoted as ɳǀɔː in [Bleek 1929: 19]. Same word as 'skin' q. v. More specifically, may be expressed by the compound !au-ɳǀɔ, lit. 'tree-skin' [Bleek 1956: 348].
Bleek 1956: 622. Quoted as ɳǀǀʌ̀li in [Bleek 1929: 19]. The same source also lists a secondary synonym: ɳǀàni, confirmed in [Bleek 1956: 344]. Semantic difference is unclear, but the second word has no external parallels.
König & Heine 2008: 61. Quoted as ɳǀǀùrì 'peel or bark' in [Heikkinen 1986: 26].
Proto North Khoisan:*ɳǀǀoˤʔrV
Distribution: The original root is fully preserved in the Northern cluster and partially in the Southern one. Known alternates are limited to a metaphoric semantic transfer {'skin' > 'bark'} in the Central cluster. Reconstruction shape: The reconstruction is problematic in the vocalism department; however, the development *o > u is attested for Ekoka several times, particularly before the front vocalism of the second syllable. We tentatively follow the more complex shape of the Juǀ'hoan variant.
König & Heine 2008: 30; Heikkinen 1986: 24. The former source lists the meaning 'stomach', the latter source gives 'belly'; most likely, a case of polysemy as in all the other dialects in this family.
Proto North Khoisan:*ɡ!ú
Distribution: Preserved in all daughter dialects. Reconstruction shape: Correspondences are generally regular and trivial. Semantics and structure: Polysemy {'belly' & 'stomach'} was present on the proto-level.
Bleek 1956: 471; Bleek 1929: 22. Secondary synonym: ɳ!wiː [Bleek 1956: 488; Bleek 1929: 22]. It is unclear if this form, with no parallels in other NK dialects, is related to ɳ!a or represents an entirely different root.
Bleek 1956: 612. The transcription ɳǀe-ǀǀá in [Bleek 1929: 22] is most likely a misprint for ɳǀǀe-ǀǀá; the latter form, along with several variants, is quoted in [Bleek 1956: 618] with the meaning '(to be) large, increase' (a compound of 'big' with another stem).
Bleek 1956: 612, 613. Quoted as ɳǀǀáːa in [Bleek 1929: 22]. Singular subject action verb/adjective; the plural subject form is ɳǀǀéɳǀǀe [Bleek 1929: 22]. Secondary synonym: ǀǀaǀǀa [Bleek 1956: 565]; [Snyman 1980: 33]. The latter is clearly a different root (with reduplication and without the nasal click efflux), but the semantic difference is unclear, and external parallels for ǀǀaǀǀa are lacking.
König & Heine 2008: 57. Quoted as ɳǀǀàʔà in [Heikkinen 1986: 26]. Singular subject action verb/adjective; the plural subject form is ǀǀȁhã̀ ibid.
Proto North Khoisan:*ɳ!̯àʔà ~ *ɳ!̯ã̀ʔã̀
Distribution: Preserved in all daughter dialects. Reconstruction shape: Correspondences are generally regular, indicating an original retroflex nasalized click and a glottal stop between the vowels. Tonal reconstruction is highly approximate. Various forms show fluctuations between nasalized and plain vowels; this may be due to the influence of the nasalized click.
König & Heine 2008: 67. Quoted as cʼāmà in [Heikkinen 1986: 22]. Plural form is čʼá-m̏hè, indicating that -mà, pl. -m̏hè is detachable as the standard diminutive suffix. The original form is still found as čʼám 'poultry, bird (life form), aeroplane', i. e. čʼámà < *čʼám-mà.
Proto North Khoisan:*cʼā(m)-mà
Distribution: Preserved in all daughter dialects. Reconstruction shape: Correspondences are more or less straightforward. The alveolar affricate regularly develops into postalveolar in Ekoka. Tonal reconstruction is approximate. Semantics and structure: the word is morphologically complex, the second component clearly identifiable as Proto-North Khoisan *-ma 'small; diminutive suffix'. The root morpheme is therefore *cʼā or, perhaps, *cʼam (cf. the notes to Ekoka !Xung).
Bleek 1956: 477. Quoted as ɳ!éː in [Bleek 1929: 22]. The latter source also quotes taː in the same meaning, but the word is Central Khoisan in origin (in [Bleek 1956] it is only attested for Naro, a close contact of !Kung).
Grootfontein !Kung:
Not attested.
!O!Kung:ɳǀà2
Bleek 1956: 341; Bleek 1929: 22. [Snyman 1980: 34] quotes the forms ɳ!aì-ǀǀòá (sg.), ɳ!ài-kxóm (pl.) in the meaning 'to bite off'. The first root in these compounds is clearly of Proto-North Khoisan descent, unlike the isolated form ɳǀà in D. Bleek's transcription. Bleek, however, does not mention the existence of a separate ɳ!ai 'bite' in !O!Kung, and it is not highly likely that ɳǀà could be a mistranscription of the former.
König & Heine 2008: 50. Compound form (second component is ḿ 'to eat' q.v.). A synonymous form [König & Heine 2008: 50] is !ʼāè; this may be simply a phonetic variant of the original form with the complex preglottalized nasalized click. Quoted as ʔɳ!āē (Western dialect) ~ ɳ!é (Eastern dialect) in [Heikkinen 1986: 25].
Proto North Khoisan:*ʔɳ!āē
Distribution: Preserved in most daughter dialects, with the possible exception of Bleek's !O!Kung, with a recent replacement of unknown status. Reconstruction shape: Correspondences are generally regular. The special preglottalized nasalized click is preserved in Ekoka, but merges with the simple nasal click in the other dialects. The vocalic correspondence "Juǀ'hoan ai : Ekoka ae" indicates an older *ae.
Bleek 1956: 292, 621; Bleek 1929: 22. L. Lloyd's texts contain examples of both words used in the meaning 'blood', with no obvious differentiation between the two. External data indicate that this may be a gradual replacement of the old root ǀʼĩ with the innovation ɳǀǀoru.
Bleek 1956: 72. Quoted as yalu ~ yula in [Bleek 1929: 22]; yàlò ~ yűlà in [Snyman 1980: 34]. The word contains a near-unique case of initial y- (not only in !O!Kung, but in North Khoisan as a whole), shows no external parallels and is consequently, in all likelihood, a borrowing from an unknown source. J. Snyman adds a secondary synonym, ǀàḿ, whose external connections are much stronger, but D. Bleek does not mention this word in her recordings of "N3".
König & Heine 2008: 60. Quoted as ɳǀǀóˤɽú in [Heikkinen 1986: 26].
Proto North Khoisan:*ǀʼaŋ
Distribution: The original root is well preserved everywhere except for the Northern dialect cluster, but is encountered even there, seemingly as an archaism. Replacements: (a) !O!Kung yalo of unknown origin, probably non-native because of initial y-; (b) Northern *ɳǀǀoˤru, probably an authentic root but without a good etymology; (c) Northern *ǀam, quoted here for Angolan !Xung and also reflected in several other dialects of this cluster; the similarity with *ǀʼāŋ̀ is notable, but the origin is probably different (click influxes as well as finals are incompatible). Reconstruction shape: Reconstruction of the vocalic part of the syllable is highly questionable and problematic (the correspondence series is unique). Only the presence of a velar nasal is certain; neither the quality of the first vocalic element nor the original tonal scheme can be fully ascertained.
Bleek 1956: 447, 492; Bleek 1929: 23. D. Bleek seems to occasionally confuse !ʼu 'bone' with !ʰu 'horn' (her recorded polysemy 'bone / horn' is most likely fictitious), which explains the unexpected appearance of variants with simple velar efflux (instead of the expected glottal stop).
Bleek 1956: 447. Quoted as !úː in [Bleek 1929: 23]. The lack of glottal stop efflux is strange and may imply that L. Lloyd, just like D. Bleek after her, was confusing the word 'bone' with the word 'horn'.
Distribution: Preserved in all daughter dialects. Reconstruction shape: Correspondences are generally regular and trivial. The word is normally assigned the regular high tone, but P. Dickens indicates an extra high tone for Juǀ'hoan. This is reflected in the reconstruction, but there is no certainty about the presence of a separate extra high tone in Proto-North Khoisan.
Bleek 1956: 390; Bleek 1929: 28. Different from cʼa 'female breast' [Bleek 1929: 24] (curiously, the form is absent in [Bleek 1956]).
!Kung:ǂé2
Bleek 1956: 659; Bleek 1929: 28.
Grootfontein !Kung:
Not attested.
!O!Kung:toa ǀʼõ3
Bleek 1956: 355. Quoted as tʼoa ǀʼõ in [Bleek 1929: 28]. Obviously a compound; neither of the two parts are, however, attested individually, or have transparent equivalents in closely related dialects. Obscure.
König & Heine 2008: 66. Quoted as cō̃ā̃ in [Heikkinen 1986: 22]. Different from kūú 'female breast' [König & Heine 2008: 42] and šám 'female breast, nipple' [König & Heine 2008: 17].
Proto North Khoisan:*ɡ!òʔá #
Distribution: Found in the Southern and Central clusters, but conspicuously absent in the Northern dialects. Replacements: (a) !O!Kung toaǀʼõ, of unknown origin; (b) Ekoka čō̃ā̃, possibly = various dialectal forms, recorded by J. Snyman, with the meaning 'lung', e. g. Tsintsabis soʔã, Cuando čõʔã, etc., eventually = Juǀ'hoan čȍˤʔò; if so, clearly an innovative development {'lungs' > 'chest'}. On the other hand, the word čàʔnà 'lung' is attested in Ekoka all by itself [König & Heine 2008: 65], so this may be some sort of semantic contamination. Reconstruction shape: the only reliably transcribed form is in Juǀ'hoan; reconstruction variants may include *ɡ!!òʔá, with a retroflex click.
Bleek 1956: 83, 102. Both forms are phonetic variants, to which it is perhaps possible to add a third one, gu [Bleek 1956: 49], although in this form the quoted example for the word reflects intransitive use. [Bleek 1929: 25] quotes two more roots, duː and če, in the meaning 'burn (tr.)', but neither is verified through [Bleek 1956], where the form duː is translated as 'to heat slowly', and the form če is not found at all.
Bleek 1956: 102, 108. Quoted as kuúː in [Bleek 1929: 25].
Grootfontein !Kung:
Not attested.
!O!Kung:
Not attested. In [Bleek 1929: 25] the possible equivalent is listed as ɡǀǀ ɑ̀lasǝ, but the more accurate entry in [Bleek 1956: 525] explains this as ɡǀǀàla-se 'to burn in (smth.)' = ɡǀǀàla 'to mark, tattoo, write'. [Snyman 1980: 35] has ɳǂȁeˤ 'to burn down'.
König & Heine 2008: 42; Heikkinen 1986: 21. Polysemy: 'to burn / to roast / to shine' (used as both transitive and intransitive stems). Secondary synonym: !̯ʼʰȕbȉ [König & Heine 2008: 95; Heikkinen 1986: 24] (exact difference in meaning stays unclear, but this form has no external parallels).
Proto North Khoisan:*kūʔú
Distribution: Preserved in all daughter dialects. Reconstruction shape: Correspondences are regular and trivial, except for the tonal reconstruction, which is approximate (the pattern is probably mid-rising, as in Ekoka).
König & Heine 2008: 84. Quoted as !ūʔúrú in [Heikkinen 1986: 24]. Also encountered as a compound: ɡǀǀāō !ūlú, lit. 'hand-nail' [König & Heine 2008: 34].
Proto North Khoisan:*!̯ūʔrú
Distribution: Preserved in all daughter dialects. Reconstruction shape: The rare retroflex click *!̯ is clearly indicated by the variation between alveolar and lateral articulation in particular dialect clusters, and is explicitly attested in the Grootfontein dialect. Ekoka !Xung, however, irregularly yields alveolar reflexation instead of lateral, possibly due to semantic contamination with *!uru 'quiver' ('nail' as 'hand sheath'?). The correspondence "Juǀ'hoan -VʔV- : Ekoka -V-" is currently interpreted as reflecting an original sequence *-Vʔ-[C]-.
Bleek 1956: 505, 522; Bleek 1929: 29. Same word as 'rain'. Judging by available data, the !Kung dialect described by L. Lloyd is the only one where the meaning 'cloud' itself can be expressed by the same word as 'rain', without an additional second component ('hair', etc.): cf. examples such as ɡǀǀa ti ǂʰauwa ɳǀeá 'clouds travel in the sky'. It cannot be verified whether they may be interpreted as a metonymical 'rain travels in the sky'; the dialect may have had a special word for 'white cloud', but it is not attested in L. Lloyd's data.
Bleek 1929: 29. Dubious entry, since it is not confirmed in [Bleek 1956], nor anywhere else. Also, D. Bleek indicates that the plural form is !ɔli, possibly implying that the form is morphologically segmentable, but no such mechanism of segmentation exists in living North Khoisan dialects.
Ekoka !Xung:ǀǀòbō-xà4
König & Heine 2008: 100. Glossed as 'white cloud', distinguished from !ʰȍwà 'cloud' (= 'raincloud') [König & Heine 2008: 82].
Proto North Khoisan:*ɡ!!à=!kxúí #
Distribution: The compound form, lit. 'sky-hair', is encountered in the Southern and Central clusters, but not in the Northern one, where several different roots are present, none of them with a solid common North Khoisan etymology (none, however, can be seen as obvious borrowings). Since even different dialects of the Northern cluster disagree in between themselves, we present no alternate etyma. Reconstruction shape: See under 'rain' and 'hair', resp.
Bleek 1929: 29; Bleek 1956: 680. Alternately transcribed as ǂxẽ in [Bleek 1956: 679]. A possible synonym is ǀau 'to be cold, bare' [Bleek 1956: 303]; however, in the English-ǀǀKxauǀǀen vocabulary of [Bleek 1929] only the first root is adduced.
Bleek 1956: 656. [Bleek 1929: 29] gives the form ǂarau instead, but it is unconfirmed in [Bleek 1956] and unsupported by external data; possibly erroneous.
König & Heine 2008: 89. Quoted as ǂàʔō in [Heikkinen 1986: 23]. Polysemy: 'cold / cool / good, well'.
Proto North Khoisan:*ǂàʔū
Distribution: Preserved (mostly) in the Northern and Central clusters. Replacements: (a) Southern cluster: *ǂxãĩ, possibly reflecting a rare semantic development {'to tremble' > 'to be cold'}; (b) !O!Kung ǀkxáú of unknown origin (not likely to be a mistranscription of *ǂaʔu). Reconstruction shape: Correspondences are regular and trivial.
Dickens 1994: 275. Secondary synonym: ɡǀàè [Dickens 1994: 202], glossed as 'come, arrive (by day)'. The two words are also encountered as a compound: ci̋-ɡǀàè 'come'. Contexts, idiomatic use, and statistical frequency all suggest that ɡǀàè means a more temporally marked 'arrive' rather than simply 'come'.
Bleek 1956: 277. Quoted as ɡǀe in [Bleek 1929: 30]. The latter source also adds a secondary synonym še ~ ša; analysis of contexts in [Bleek 1956: 178] as well as external data, however, suggests that the basic meaning of this word is 'to return, come back', whereas ɡǀe is simply 'come' (cf. the following example: sa ti u, sa ti ɡǀeː 'they two go out, they two come in', where it forms an antonymous pair with u 'to go').
Doke 1925: 153. Somewhat dubious; corresponding forms in related dialects (e. g. Juǀ'hoan hȍȅ) usually have an imperative meaning ('come!'). Since Doke gives no syntactic contexts, it is possible that in the Grootfontein dialect it is also used primarily as an imperative. Alternately, the word ɡǀȅː [Doke 1925: 157] is also translated as 'to travel, to come', which is, however, unlikely to represent the Swadesh meaning either.
Bleek 1956: 216, 228. Quoted as siː ~ ciː ~ čiː in [Bleek 1929: 30]. The latter source adds ɡǀeː as a potential synonym, only scantily confirmed through examples in [Bleek 1956] (statistically, ci is much more frequent).
König & Heine 2008: 28; Heikkinen 1986: 23. Quasi-synonyms include ǀǀxāì [König & Heine 2008: 101] (glossed as 'come, cross'; meaning given as 'to come (Western dialect), to come out (Eastern dialect)' in [Heikkinen 1986: 25]) and čí 'come!' (imperative) [König & Heine 2008: 65].
Proto North Khoisan:*cí
Distribution: The original root is strongly preserved in the Central and Southern clusters, but in the Northern cluster it seems to have been relegated to the imperative function ('come!'), replaced in the original meaning by *ɡǀàè. The latter root is also of Proto-Northern Khoisan origin, and co-exists in most dialects with *cí with subtle differences in semantics (judging by certain examples and translations, *cí = 'to come (proceed in the listener's direction)', *ɡǀàè = 'to arrive (reach the final destination)'. Replacements: Northern cluster: *ɡǀàè {'to come' > 'to arrive'}. Reconstruction shape: Correspondences are regular and trivial, including reconstruction of a high tonal pattern.
Bleek 1956: 566. Quoted as ǀǀáí in [Snyman 1980: 36]; ǀǀé: in [Bleek 1929: 33]. The latter source additionally lists ǀǀʼòː as a synonym; this is a transparent borrowing from Central Khoisan. It is not confirmed in [Bleek 1956]. Singular subject action verb; the plural subject form is ǀǀau [Bleek 1956: 561].
König & Heine 2008: 96. Singular subject action verb; the plural subject form is ǀǀàō [ibid.]. Quoted as ǀǀāē (Eastern dialect), ǀǀē (Western dialect) in [Heikkinen 1986: 25] (plural subject form is ǀǀāō).
Proto North Khoisan:*!̯e
Distribution: Preserved in all daughter dialects. Reconstruction shape: The click is to be reconstructed as retroflex, based on perfect correspondences all around (except for the strange transcription with palatal ǂ- in //Kxau//en). The vowel was most likely *e, with regular diphthongization in Juǀ'hoan and certain Ekoka lects. Tonal information is contradictory and so far inconclusive. The word had a lexical variant *!̯ao for the plural action verb; the two may have been morphologically related on a Pre-Proto-North Khoisan level, but the model is not seen in any other examples.
Dickens 1994: 207. Secondary synonyms: (a) ɡǂʰòà (with the exact same rare click as the initial consonant, this must be a dialectal variant of the same root) [Dickens 1994: 207]; (b) !kxúí (= 'hair, fur'; probably a tabooistic equivalent) [Dickens 1994: 315].
König & Heine 2008: 32. Quoted as ɡǂʰōē in [Heikkinen 1986: 23].
Proto North Khoisan:*ɡǂʰo-ĩ ~ *ɡǂʰo-e
Distribution: Preserved in all daughter dialects, although in different morphological shapes. Reconstruction shape: All reliably transcribed sources agree upon *ɡǂʰ- as the original click articulation, and labial articulation of the following vowels. Juǀ'hoan, however, shows at least two different morphological variants: ɡǂʰo-a and *ɡǂʰo-ĩ > ɡǂʰũ-ĩ, and Ekoka adds a possible third (ɡǂʰo-e). This can be interpreted as reflecting three original nominal suffixes (archaic class markers?) and is analogous to similar variations in many other cases. The bare root would then be reconstructible as *ɡǂʰo-.
Bleek 1956: 180, 230; Bleek 1929: 34. Secondary synonym: kxaː [Bleek 1956: 117; Bleek 1929: 34]. The latter, however, is unquestionably a Central Khoisan word, not met in actual texts.
König & Heine 2008: 18. Quoted as cʰȁŋ (Eastern dialect), ṣːàŋ ~ ṣʰàŋ (Western dialect) in [Heikkinen 1986: 22].
Proto North Khoisan:*čʰȉŋ
Distribution: Preserved in all daughter dialects. Reconstruction shape: Initial *č- clearly indicated by Juǀ'hoan data. Reconstruction of the vocalic part is problematic, mainly because of the lack of any traces of a nasal consonant or nasalization in Juǀ'hoan. Nevertheless, loss of the nasal in this particular context may still have been conditioned phonologically. Aspiration in Juǀ'hoan is confirmed in several other dialects; it may be connected with the ultra-low tone attested for Ekoka.
Bleek 1956: 470, 558. [Bleek 1929: 35] does not contain these items (only ǀǀaːo in the verbal meaning 'to dry (intr.)'), but gives a strange, unparalleled entry ča in the adjectival meaning. It is, perhaps = ča 'to pour, spill' (i. e. 'spilt' mistaken for 'dry').
König & Heine 2008: 103. Adjectival form; the verbal stem is formed through tonal change (ǀǀkxāò 'be dry, hard'). Quoted as ǀǀkxāō 'become dry or hard' in [Heikkinen 1986: 26].
Proto North Khoisan:*!̯kxau
Distribution: Preserved in all daughter dialects (at least, where the form is attested). Reconstruction shape: Correspondences are mostly regular and trivial. The correspondence between Ekoka ǀǀ- and Juǀ'hoan !- indicates an original retroflex click. The tonal pattern, however, is unclear (level tone in Ekoka vs. rare "ultra-high" tone in Juǀ'hoan); the original pattern was probably composite.
Dickens 1994: 296. The synonym !õ̀ã̀-ǂáé is. like many other synonyms for bodyparts, most likely an "educated" word of Khoekhoe provenance (Nama ɡǂae- 'ear'), although the first component remains unclear.
Distribution: Preserved in all daughter dialects. Reconstruction shape: Juǀ'hoan and Ekoka forms (the most reliable ones) completely coincide and justify the current reconstruction.
Bleek 1956: 118; Bleek 1929: 35. The latter source also quotes čà as a synonym, but the form has not reappeared in [Bleek 1956] and has not been confirmed by any other data, so it must have been erroneous.
Bleek 1956: 118. Quoted as kxa in [Bleek 1929: 35]. The latter source also quotes ɡǀǀwí as a synonym; the word is seemingly confirmed in the meaning 'earth' in [Bleek 1956: 538], but lacks any sort of external parallels. It is, however, nearly homophonous with Kavango (Ekoka) !Kung ɡǀǀùʔi̋ 'termite hill'; comparing this fact with the lone textual example supporting it (ǯú tanki ti !um ɡǀǀwí 'other people fill in the earth' /over a dead man's body/) would suggest that the real meaning is a more specific '(earth) mound'.
Distribution: Preserved in all daughter dialects. Reconstruction shape: Juǀ'hoan and Ekoka forms (the most reliable ones) completely coincide and justify the current reconstruction. Semantics and structure: The polysemy 'earth / sand' must have been inherited from the proto-level (no perceived semantic difference between the two substances due to areal conditions).
Bleek 1956: 131; Bleek 1929: 35. Both sources quote kxɔː ~ kxɔ as a synonym in the meaning 'to eat meat (hard food)'; this is undoubtedly a recent borrowing from a Khoekhoe source (not to mention ineligible for exclusion due to semantic reasons).
König & Heine 2008: 44. Quoted as ʔḿ in [Heikkinen 1986: 27].
Proto North Khoisan:*ʔḿ
Distribution: Preserved in all daughter dialects. Reconstruction shape: The root, just as it is attested in most dialects, should be reconstructed with a high-toned syllabic *m preceded by a glottal stop.
Dickens 1994: 254. Used independently only in the meaning 'glans of the penis'; otherwise, in compounds like cʼàmà-ɳ!ù 'bird's egg', kʰúkʰú-ɳ!ù 'chicken egg' etc.
König & Heine 2008: 54. Quoted as ɳ!ūú in [Heikkinen 1986: 25].
Proto North Khoisan:*ɳ!u
Distribution: Preserved in all daughter dialects. Reconstruction shape: Correspondences are mostly regular and trivial with the exception of the tonal pattern where even Ekoka has several close variants.
Distribution: Preserved in all daughter dialects. Reconstruction shape: Correspondences are regular and trivial, including a clearly indicated glottal stop "break"; the only possible minor problem is with the tonal pattern (probably low-mid, but low-high is not excluded either).
Bleek 1956: 347. [Bleek 1929: 37] quotes ɳǀǀ wiː instead, but the latter is obviously a Central Khoisan root by descent, and its existence in !Kung (even as a borrowing) is not further confirmed in [Bleek 1956].
König & Heine 2008: 48. Quoted as ɳǀí 'oil' in [Heikkinen 1986: 23].
Proto North Khoisan:*ɳǀí
Distribution: Preserved in all daughter dialects. Reconstruction shape: Correspondences are regular; original vowel *-i undergoes diphthongization in Juǀ'hoan and a few other dialects.
Bleek 1956: 128, 466. Same as 'hair' (the meaning 'feather' is illustrated by the example čaː !wi 'feather on arrow'). Quoted as !kxwi in [Bleek 1929: 38].
Ekoka !Xung:
Not attested.
Proto North Khoisan:*!kxúí
Distribution: Although precise data on many dialects are missing, it is fair to assume that the basic words for 'hair' and 'feather' were not distinguished in Proto-North Khoisan (at best, as in Juǀ'hoan, the meaning 'feather' could be specified as 'bird-hair'). For more details, see under 'hair'.
König & Heine 2008: 20. Quoted as dàʔā in [Heikkinen 1986: 21].
Proto North Khoisan:*dàʔá
Distribution: Preserved in all daughter dialects. Reconstruction shape: Correspondences are regular and trivial, except for the questionable tonal pattern (probably low-high, but Ekoka indicates low-low or low-mid).
König & Heine 2008: 30. Quoted as ɡ!òrō in [Heikkinen 1986: 24]. Very dubious. König & Heine actually list three synonyms, with no attempts at differentiation; the other two are ǀǀʼáú (p. 104) and !ʰȁnnù (p. 82). The primary choice on this list is motivated only by the fact that ɡ!òlō is the only word out of the three that finds confirmation in [Heikkinen 1986]. External connections would rather speak in favor of ǀǀʼáú, but since this word is suspicious as a recent loan from Khoekhoe, such connections should not qualify as a serious argument in this particular case.
Proto North Khoisan:
Distribution: Technically, the form *ǀǀʼau is reconstructible for Proto-North Khoisan, being found in at least two out of three main dialect clusters. Nevertheless, the following reasons prevent us from adopting such a reconstruction: (1) the overall distribution of the root is quite poor; it is very marginally attested or not attested at all in old sources; (2) the form closely matches Proto-Central Khoisan *ǀǀʼau, from descendants of which the several attested variants could be borrowed individually; (3) the existence of an authentic word for 'fish' in North Khoisan is questionable considering their desert homeland; for comparison, there is not a single South Khoisan language for which the word 'fish' has even been attested, let alone considered "authentic". Ekoka ɡ!òlō, without any clear etymology or transparent source of borrowing, is in this respect a better candidate for Proto-North Khoisan than *ǀǀʼau; still, in the light of its even more limited distribution as well as factor (3), we prefer to leave the slot empty.
Bleek 1956: 196. Dubious (translated as "to fly up" in the one available context). [Bleek 1929: 40] gives ǀʼwíː in the meaning 'fly', but [Bleek 1956: 362] gives the meanings 'to fly up' and 'to come out' (e. g. of ants). Still another form, ɳ!oːa [Bleek 1956: 479], may be etymologically related to Juǀ'hoan ɳ!ȍm̏, etc., but is translated as 'to fly about' (of birds).
Bleek 1956: 525 (only in the meaning 'to fly up'). Quoted as ɡǀǀama in [Bleek 1929: 40] (in the meaning 'to fly' as such). Dubious entry, but no better one can be suggested.
Grootfontein !Kung:
Not attested. Cf., however, kā-ɡǀíː 'to fly up' [Doke 1925: 155], where the second part = ɡǀíː 'to come out, to rise'.
Bleek 1956: 196. The confirming textual example is given as soŋgu tẽː "the arrow flies". Quoted as tẽ in [Bleek 1929: 40]. The same source also adds !aú as a secondary synonym, but this is not confirmed in [Bleek 1956] (according to that source, such forms in the meaning 'to fly' only exist in South Khoisan languages). Of additional note is ɳ!oa [Bleek 1956: 479], translated as 'to mount up' (in the sky) and compared with //Kxau//en ɳ!oːa 'to fly about'. Everything is highly dubious due to ambiguity of the contexts.
Ekoka !Xung:
Not attested. Cf., perhaps, čāō 'to fly up' [König & Heine 2008: 64], but the meaning is fully glossed as 'wake up, rise, stand up, fly up, jump up', indicating the idea of ascension rather than flight.
Proto North Khoisan:*ɳ!om #
Distribution: The etymon is highly unstable and, furthermore, few of the sources are trustworthy (see notes on particular items). Juǀ'hoan ɳ!ȍm̏ is the only form from a reliable source glossed directly as 'fly', and can be very tentatively projected onto a higher level, especially if it is in some way related to //Kxau//en ɳ!oːa 'to fly about' (maybe < *ɳ!om-a?, cf. the Juǀ'hoan derived form ɳ!ȍm̏-ȁ 'to fly over (a village)').
König & Heine 2008: 76; Heikkinen 1986: 23. The same sources also list !ʼō in the meaning 'foot', but this is dubious: the complete meaning is given as 'foot, base, back' [König & Heine 2008: 88; Heikkinen 1986: 24], and all external parallels point to 'back' as the default semantics.
Proto North Khoisan:*ǀkxáí
Distribution: Preserved in all daughter dialects. Reconstruction shape: Correspondences are regular and trivial, including the tonal pattern (high-high).
Bleek 1956: 381; Bleek 1929: 41. Erroneously contaminated in [Bleek 1956] with ɡǂẽ 'long' q. v. Secondary synonym: !au [Bleek 1956: 411; Bleek 1929: 41].
!Kung:
Not attested. Cf., however, the word for 'to fill': ɡ!ʼàŋ ~ ɡ!aä́ŋ [Bleek 1956: 376], etymologically the same as the adjectival form 'full' in other languages.
Grootfontein !Kung:
Not attested.
!O!Kung:
Not attested. Cf., however, ǂʼàè-ɡ!àŋ̀ʔŋ́ 'to fill' in [Snyman 1980: 38].
König & Heine 2008: 30. Quoted as ɡ!àʔáŋ in [Heikkinen 1986: 24]. Secondary synonym: bálí [König & Heine 2008: 15] ~ bárí [Heikkinen 1986: 20] 'to be full'. The latter form is probably not of Khoisan origin, judging by the initial b-, and has no parallels outside Ekoka. It is unclear from the existing data what is the exact difference in semantics between the two items.
Proto North Khoisan:*ɡ!àʔŋ
Distribution: Preserved in all daughter dialects in which it is attested. Reconstruction shape: correspondences are mostly trivial, with the exception of the last mora after the glottal stop; *ɡ!aʔŋ, following the Ekoka structure, seems to be the least controversial variant for the protoform.
Bleek 1929: 42; Bleek 1956: 2, 267. The forms are given as synonymous in [Bleek 1929] (a third synonym included there is na, disqualified because its use is restricted to the imperative 'give!'), and analysis of the contexts in [Bleek 1956] does not allow to establish any serious semantic difference. Considering external data, it is likely that au is a recent replacement for ǀʼaː, still competing with the old form.
Bleek 1956: 267, 293. Quoted as ǀʼã́ ~ ǀá in [Bleek 1929: 42]. The latter source also quotes aː as a synonym; however, the form is not confirmed in [Bleek 1956] (it is typical rather of Southern Khoisan languages and could have penetrated the North Khoisan material accidentally).
Doke 1925: 152. Secondary synonyms: ǀʼã́ː 'to give, gift' [Doke 1925: 156] (the way the meaning is quoted and the lack of contexts, in comparison to the far better represented āū, implies that, in this dialect, this verb may have a more specialized function); nāː 'to give, to bring' [Doke 1925: 154] (judging by external data, most likely an imperative form).
Bleek 1956: 293; Bleek 1929: 42. Quoted as ǀa in [Snyman 1980: 39]. [Bleek 1929] also quotes na as a synonym, confirmed in [Bleek 1956: 141], which includes both imperative ('give!') and indicative textual examples. It is possible that in some subdialects the older imperative verb has indeed replaced the old indicative form.
König & Heine 2008: 74; Heikkinen 1986: 22. Secondary synonym: nèʔè [König & Heine 2008: 46]. This is probably the imperative form, as in other languages of the subgroup (textual confirmation needed).
Proto North Khoisan:*ǀãʔã
Distribution: The original root *ǀʼa is present in all three major dialect clusters and clearly reconstructible. It should be distinguished from *na ~ *neʔe, whose functions in at least several distinct dialects and, most likely, in Proto-North Khoisan as well are restricted to the imperative function. Replacements: *au, of unknown origin, in the Grootfontein and ǀǀKxauǀǀen dialects. Reconstruction shape: Attested forms fluctuate between nasalized and non-nasalized variants, without much evidence for the primacy of either one. Also, the Northern cluster shows glottalized articulation of the vowels (ǀaʔa), whose archaicity is possible, but uncertain (the development *ǀaʔa > ǀʼa in Juǀ'hoan is quite irregular). The proposed reconstruction is somewhat speculative (it includes both glottalization and nasalization, presupposing that different dialects simplified the original structure in different ways).
Bleek 1956: 405; Bleek 1929: 43. Secondary synonym: ǀḿ [Bleek 1929: 43], translated, however, as 'nice; pretty' in [Bleek 1956]. The form !ãĩ must be the principal word, since it functions in texts as the antonym of ču 'bad'.
Bleek 1956: 32. Within actual text examples translated as 'nice', 'sweet', 'glad'. [Bleek 1929: 43] quotes an entirely different root: kaiä, but it is not confirmed in [Bleek 1956].
König & Heine 2008: 38. Quoted as kã̀ĩ́ˤ in [Heikkinen 1986: 21]. In the English-Ekoka section of their dictionary, König & Heine list several other synonyms, but all of them are probably secondary according to the main section: (a) n̏ǀʰúm is glossed as 'to be beautiful' (p. 48); (b) tú 'good' is a word said to be 'known but not used' (p. 70); (c) !̯ʰánnù is glossed as 'be good, harmless' (p. 90).
Proto North Khoisan:*žã
Distribution: The widest distribution (South and Central clusters) by a clearly non-borrowed (at least, not recently) root is shown by *žã. Its replacements in various dialects, for the most part, seem to represent recent borrowings of various variants of the Proto-Central Khoisan root *!ã́ĩ̀ 'good' (some of them still preserve the original click articulation and some, as seen in the Ekoka variant, have already dropped the click influx). Reconstruction shape: The form of the original root is established based on the Juǀ'hoan form.
Bleek 1956: 303; Bleek 1929: 44. Quoted as 'blue, light green' in [Bleek 1956]. [Bleek 1929: 44] also gives the form ɳǀoussi in the meaning 'dark green'; in [Bleek 1956: 345] it is then quoted as ɳǀausi 'green'. Considering the lack of any special parallels in other languages, it is highly likely that this is really the same word as ǀãũ (-si is typically the plural suffix), misheard or dialectal. See also notes to 'yellow'.
König & Heine 2008: 74. Polysemy: 'green / blue / yellow' (basic meaning glossed as 'coloured' in the source).
Proto North Khoisan:*ǀȁŋu
Distribution: The original root is preserved almost everywhere, except for a possible replacement of unclear origin in the Grootfontein dialect (provided Doke's glossing is correct). Reconstruction shape: Correspondences indicate the presence of a final sequence *-aŋu, with the subsequent development > -ãũ in Juǀ'hoan (see 'neck' for more details on intervocal *-ŋ- in Proto-North Khoisan). Tonal characteristics are not entirely clear, but at least one of the two morae must have been characterized by an ultra-low tone, connected with the aspiration/breathiness elements in Juǀ'hoan and Ekoka. Semantics and structure: The root must have been the main denotation for the entire 'blue / green / yellow' spectrum in Proto-North Khoisan.
König & Heine 2008: 86; Heikkinen 1986: 25. The meaning 'head hair' may be more specifically expressed by the compound form ʔɳǀē !kxúí, where ʔɳǀē = 'head' q.v.
Proto North Khoisan:*!kxúí
Distribution: Preserved in all daughter dialects. Reconstruction shape: Correspondences are regular and trivial.
Distribution: Preserved in all daughter dialects. Reconstruction shape: Correspondences are regular, indicating an original voiced retroflex click (still preserved in Grootfontein). The only problem is with the tonal pattern (ultrahigh-high in Juǀ'hoan vs. low-low or mid-mid in Ekoka, probably indicating a rare and unstable combination in the protolanguage).
Bleek 1956: 346, 347; Bleek 1929: 46. All variants quoted in [Bleek 1956] with the second meaning 'berry', but this is likely to be the result of confusion with a different root = Juǀ'hoan ɳǀàŋ 'wild currant', !Kung (Doke) ǀŋ 'berry' rather than true polysemy.
König & Heine 2008: 47. Polysemy: 'head / top / peak'; also, same word as ʔɳǀē 'reflexive marker' [König & Heine 2008: 48].
Proto North Khoisan:*ʔɳǀē
Distribution: Preserved in all daughter dialects. Reconstruction shape: Correspondences are generally regular. The preglottalized nasal click is reconstructed based on Ekoka data. Juǀ'hoan regularly undergoes diphthongization of *-e > -ai. Tonal pattern is mid-level.
König & Heine 2008: 64. Polysemy: 'hear / feel understand'. Quoted as càʔā in [Heikkinen 1986: 22].
Proto North Khoisan:*sàʔā
Distribution: Preserved in all daughter dialects. Reconstruction shape: The fluctuation between affricate (c-) and fricative (s-) articulation is resolved in favour of the fricative articulation as original; affricativization probably occurs under the influence of the glottal stop, especially considering that the first half of the complex vowel sequence is frequently reduced or even completely deleted in the actual articulation. Everything else is fairly regular, including the tonal correspondences between Juǀ'hoan and Ekoka.
Distribution: Preserved in all daughter dialects. Reconstruction shape: Correspondences are mostly regular. The development of the velar affricate efflux into a glottal stop in Grootfontein !Kung is not 100% regular, but very frequent (see 'dry', 'foot', etc.). In !O!Kung, Bleek records dialectal articulation that drops the click influx, possibly under local Khoe influence. Semantics and structure: The polysemy {'heart' & 'inside'} is reflected in at least several dialects and may easily go back to Proto-North Khoisan stage.
Distribution: Preserved in all daughter dialects. Reconstruction shape: Correspondences are mostly regular and trivial. The retroflex click articulation in Grootfontein must represent erroneous transcription on the part of C. Doke (correspondences do not confirm original retroflex articulation).
Bleek 1956: 132, 136, 137; Bleek 1929: 49. The latter source adds ŋ and na as additional variants; the forms are confirmed in [Bleek 1956: 141, 142], but do not seem to be primary choices (they function more frequently as possessive forms rather than independent pronouns, and may be positional phonetic variants in at least some contexts).
Doke 1925: 153. Form said to be used "in conjunction". The same source also quotes the "isolated" (most likely, emphatic) form mı́-hı. Of note is also the morpheme nà [Doke 1925: 154] that Doke interprets as "1st pers. singular tense formative" (e. g. na=tàm 'I don't know').
Bleek 1956: 132, 136, 137; Bleek 1929: 49. The variant mi-hi, quoted in both sources, is most likely emphatic. Quoted as ma, mi in [Snyman 1980: 41]; the latter source also adds another stem, na, as synonymous, without explaining the difference.
Distribution: Preserved in all daughter dialects. Reconstruction shape: One of the very few solid cases of initial *m- in Proto-North Khoisan. The stem is often found in multiple variants, of which *mí is the most frequent; the others may perhaps be explained as lenited, assimilated, or, in a few cases, suffixal variants. The emphatic stem *mi-hi is also of Proto-North Khoisan origin (cf. the corresponding form *à-hi for the 2nd p. pronoun).
Bleek 1956: 429, 453, 502. Quoted as !ũ ~ !uŋ in [Bleek 1929: 50]. The same source also lists ǀǀõã as a synonym; this must be the plural object action equivalent = Juǀ'hoan !ʼṍã́.
König & Heine 2008: 82; Heikkinen 1986: 24. Singular object action verb; the plural object form is ǀǀʼŋ́ [ibid.].
Proto North Khoisan:*!ʰṹ
Distribution: Preserved in all daughter dialects. The plural object form is more difficult to reconstruct, since several different stems are attested in that function (Juǀ'hoan !ʼṍã́ is not equal to Ekoka ǀǀʼŋ́). Reconstruction shape: Correspondences are generally trivial, except for the final position where a velar nasal consonant may or may not have been present in Proto-Northern Khoisan (variation in Ekoka seems to favour this reconstruction, but it is unclear if the variant !ʰúŋ is truly archaic; at the very least, correspondences indicate a coda that is different from the one in 'nose' q.v.).
Bleek 1956: 458; Bleek 1929: 50. Translated as 'knee' in [Bleek 1929], but as 'knee-joint' in [Bleek 1956] and transparently equivalent to Ekoka ɡ!xȍȁ ʔɳǀē 'knee-cap', where the first component is 'knee' proper and the second = 'head' q.v.
Bleek 1956: 48, 281. Quoted as ɡǀǀɔm ~ gɔm in [Bleek 1929: 50]; ǀòˤm̀ʔḿ in [Snyman 1980: 42]. Transcription with the lateral click in [Bleek 1929] is probably in error (a misprint), although it is also reproduced in the cross-reference in [Bleek 1956: 48]; Snyman's data clearly corroborate the variant with the dental click.
Distribution: Preserved in most dialects, with the possible exception of some Northern lects (provided the meanings have been attested correctly). Replacements: !O!Kung ɡǀɔ̀m, probably representing a metonymic meaning shift {'knee-cap' > 'knee'}, cf. the distinction in Ekoka between ɡ!xȍȁ 'knee' and ɡǀȍˤḿ 'knee-cap' [König & Heine 2008: 28]. Reconstruction shape: Correspondences are regular and trivial, except for the tonal pattern (the correspondence between low-low in Juǀ'hoan and ultralow-ultralow in Ekoka is not well understood). Semantics and structure: The stem is frequently found as a compound with *ʔɳǀē 'head', either explicitly or, perhaps, implicitly in the meaning 'knee-cap', lit. 'knee-head'.
Bleek 1956: 640. Quoted as ǂʼãː ~ ǂã́ in [Bleek 1929: 51]. It should be noted that in [Bleek 1956: 393] we also find !ʼʰãː 'to know'; the semantic difference between the two words in the language is hard to establish.
Bleek 1956: 672; Bleek 1929: 51. Quoted as ɳǀǀʼʰáí in [Snyman 1980: 42] (provided that this is really the same root; the click correspondences are irregular, although this may simply represent a transcription mistake on Snyman's part). Snyman also adds a secondary synonym: ɡǀǀàŋ (similar to Grootfontein !Kung !̯ʼʰã, but phonologically incompatible).
König & Heine 2008: 55. Quoted as ʔɳǂìí (Western dialect) ~ ɳǂìí (Eastern dialect) in [Heikkinen 1986: 24].
Proto North Khoisan:*!ʼʰã
Distribution: The original root *!ʼʰã is well preserved in the Southern dialect cluster, but tends to be displaced elsewhere. Replacements: (a) ǀǀKxauǀǀen ǂʼãː < Proto-Northern Khoisan *ǂʼaŋ {'to think' > 'to know'}; (b) !O!Kung ɳǂi, etc. < Proto-Northern Khoisan *ʔɳǂai, reconstructible with the semantics 'to be able to; to understand' {'to be able' > 'to know'}, cf. the meaning 'to be able' in several sources listing the same root [Bleek 1956: 672]. An alternate scenario is that *ʔɳǂai could be the original root and all of the forms listed under *!ʼʰã merely borrowings from a Central Khoisan source (cf. Proto-Central Khoisan *!ʼã́ 'to know'). However, the distribution of reflexes of *!ʼʰã is so wide that, even if such a borrowing did occur, it should be postulated at the proto-stage, which renders the scenario irrelevant to the current issue. Reconstruction shape: Correspondences are mostly regular (in Grootfontein !Kung, Doke incorrectly records a retroflex click, just like he does with 'horn', q.v.); the tonal pattern is, however, hard to establish without Ekoka data.
Bleek 1956: 27. Quoted as dɔ̀raː in [Bleek 1929: 52], with a second synonym: ɳǀɔbu. The latter is, however, not found in [Bleek 1956] and may be an erroneous entry in the earlier vocabulary.
Bleek 1956: 390. Quoted as ɡ!wáː in [Bleek 1929: 52]; likely to be the most neutral word in this dialect, as illustrated by textual examples (e. g. ǂã ti ma !ãũ ɡ!wa-siŋ 'the wind blows the tree's leaves'). The latter source also adds ɳ!ɔ̀bu as a synonym; existence of the word is confirmed in [Bleek 1956: 480], where it is quoted as ɳ!ʼóˤbbu, but no text examples are given.
Bleek 1956: 47, 532. Quoted as ɡǀǀɔ̀ba ~ ǀǀɔ̀a ~ gɔaː in [Bleek 1929: 52]. Different root quoted in [Snyman 1980: 43]: ɡ!oaˤ. Theoretically, Bleek's gòaː may fit in better with Snyman's variant, reflecting a secondary dialectal post-click-drop variant of ɡ!oaˤ.
König & Heine 2008: 30. Quoted as ɡ!ȍāˤ in [Heikkinen 1986: 24]. Secondary synonym: gè-ǀǀkxáé [König & Heine 2008: 25] (internal structure is not quite clear; gè probably = 'be, stay, have', second component has not been identified).
Proto North Khoisan:*ɡ!oa #
Distribution: A complex case; the etymon is highly unstable, usually with several quasi-synonyms for each dialect, frequently with dubious semantics. The best isogloss so far is *ɡ!oa, with the meaning 'leaf' (as such) attested for L. Lloyd's !Kung and Ekoka, but in Juǀ'hoan yielding ɡ!òà 'wet leaf' [Dickens 1994: 212]. Better data from future descriptions may, however, change the situation. Replacements: (a) Juǀ'hoan dòàˤrà, etc., of unclear origin (an areal isogloss with Naro tòàˤrà, although the latter also lacks a reliable Central Khoisan etymology; both forms may be possibly inherited from an unknown substrate); (b) Grootfontein ɳ!ʼúbú, etc. = Juǀ'hoan ɳ!ùʔúbú 'to come into leaf (vb.); leaf, cabbage (n.)' [Dickens 1994: 254] {'to come into leaf' > 'leaf'}. Reconstruction shape: Correspondences among the few surviving reflexations are mostly regular, except for some difficulties with the tonal pattern.
Bleek 1956: 182; Bleek 1929: 53. The latter source also quotes tam in the same meaning, but this is erroneous; in [Bleek 1956: 190] the correct meaning is stated as 'to fall down, to throw', confirmed by external data.
Bleek 1956: 182, 236; Bleek 1929: 53. The latter source adds ɳ!waː as a synonym, but textual examples in [Bleek 1956], compared with external data, show that the more correct meaning is 'to stand'.
König & Heine 2008: 19. Quoted as sú in [Heikkinen 1986: 26]. Polysemy: 'lie / lie down / lay down / fall'. Singular subject action verb; the plural subject form is ɡ!̯à.
Proto North Khoisan:*šú
Distribution: Preserved in all daughter dialects. Reconstruction shape: Correspondences are regular and trivial. Semantics and structure: The meanings 'lie' (state) and 'lie down' (action) seem to be generally expressed by the same stem or lexeme, as opposed to 'sleep' q.v.
König & Heine 2008: 18. Quoted as sáŋ ~ čáŋ in [Heikkinen 1986: 26].
Proto North Khoisan:*čiŋ́
Distribution: Preserved in all daughter dialects. Reconstruction shape: The correspondence between Juǀ'hoan č(ʰ)- and Ekoka š- is irregular, possibly reflecting a specific development *č- > š- before a syllabic nasal. The coda is provisionally reconstructed as *-iŋ, although alternate variants are possible.
Bleek 1956: 277 (translated as 'long; tall; big'). Quoted as ɡ!ẽ in [Bleek 1929: 55], i. e. with a different click influx; considering that old Khoisan sources consistently misidentify the palatal click ǂ, the variation between the earlier and the later source is an indirect argument in favor of a palatal click in this word, i. e. the same as in related languages.
Bleek 1956: 275, 644. Translated as '(to be) long, tall, deep, powerful' (most text examples are in the meaning 'tall', but the two meanings normally merge in North Khoisan).
König & Heine 2008: 31. Adjectival variant; the verbal stem ('to be long, tall, big') is attested with tonal variation as ɡ!̯āʔŋ̀. Plural form for both stems is ɡ!̯àáŋ.
Proto North Khoisan:*ɡǂaʔŋ
Distribution: Preserved in most daughter dialects, with the possible exception of !O!Kung. Replacements: !O!Kung ǂxana is probably related to Juǀ'hoan ǂxã́ 'far' [Dickens 1994: 303], !O!Kung ǂxa id. with additional suffixation {'far' > 'long'}. Reconstruction shape: Correspondences are generally regular, although the coda is relatively rare and not entirely certain; however, the presence of a segmental nasal *-ŋ is almost certain.
König & Heine 2008: 21. Quoted as ʒʼáŋ in [Heikkinen 1986: 22].
Proto North Khoisan:*cʼíŋ́
Distribution: Preserved in most of the daughter dialects in which it is attested. Reconstruction shape: Juǀ'hoan and Ekoka forms clearly belong together; the correspondence "-ĩ : -ŋ" is recurrent (cf. 'liver', etc.). The voicing in Ekoka is, however, irregular (all of the dialectal forms collected by J. Snyman show that voiceless *cʼ- is primary).
Bleek 1956: 458. Quoted as !wã̀ in [Bleek 1929: 56]. Alternately quoted as !wã̀ˤ in [Bleek 1956: 458]; attested contexts suggest this may be actually a plural form, although in [Bleek 1929] the singular and plural forms of this word are put down as homonymous. On the other hand, formation of the plural number through pharyngealization of the vowel is never attested in any reliable source on North Khoisan languages, so there may be some mistake involved.
Doke 1925: 160. Polysemy: 'man / husband' (from which it can be ascertained that this is indeed the basic form for 'man (male)' rather than 'human being'). Cf. also ɡ!̯ō-mà 'small (of males); son' [ibid.]. Suppletive plural: ɳǀǀȁȉ [Doke 1925: 161].
Bleek 1956: 447; Bleek 1929: 56. Of note is also ɡǀǀɔ̀ 'man, male' [Bleek 1956: 531]. The word is only listed in the meaning 'male' in [Bleek 1929], but such textual examples as ɡǀǀɔ̀ ǀwi !ei dàà 'a man does not bring firewood' may indicate that the word is at least partially synonymous with !ũ in its usage in the meaning 'human being of male sex'. [Snyman 1980: 44] lists the form !xṹ-ɡǀǀòˤ, pl. !xṹ-ɳǀǀàé, lit. '!Xung-male', as well as an entirely different root, ɳ!áú, = Juǀ'hoan ɳ!àù 'old man' (most probably, this is the exact meaning in Angolan !Xung as well).
Distribution: The main stem denoting a 'male human being', as opposed to both 'female human being' and 'human being, person (in general)', seems to have been generally preserved in all three dialectal clusters, despite occasional instability as well as strange phonetic changes. Replacements: Grootfontein ɡ!̯oː, etc., reflecting Proto-North Khoisan *ɡ!̯oˤ 'male' (of any species; cf. the use of -ɡ!oˤ as a suffix denoting male sex in Juǀ'hoan [Dickens 1994: 213]) {'male' > 'man'}. Reconstruction shape: The root is highly problematic; attested variants presuppose at least four irreconcilable variants (Juǀ'hoan !ʼʰõã, ǀǀKxauǀǀen *!õã, !Kung *!ũ, !O!Kung /Snyman/ !xũ = Ekoka !xũ̄ṹ 'person' [König & Heine 2008: 86]). Mechanistically suggesting four different entries is out of the question. It is more likely that the Juǀ'hoan variant is the most archaic, whereas the old sources fail to properly transcribe aspiration; as for the velar fricative efflux -x- in the Northern cluster, it may have appeared secondarily through contamination with *!xuni 'to live, reside' (cf. Ekoka !xúnní [König & Heine 2008: 86], etc.). Semantics and structure: In Proto-North Khoisan, the root must have denoted a 'male person' without specific ethnic connotations (as it is attested in Juǀ'hoan and ǀǀKxauǀǀen); in many dialects, however, there is a clear tendency to restrict its usage to the denotation of native speakers (hence the appearance of the ethnonym !Kung, !Xung).
Doke 1925: 158. The word !āū ~ !̯āū is also translated as 'three, many' in [Doke 1925: 158, 159], but the numeric meaning is likely to be the principal one.
Bleek 1956: 313, 661. Quoted as ǂʰíː in [Bleek 1929: 57]; ǂʰáí in [Snyman 1980: 44]. The latter source adds a secondary synonym, hò (no etymology).
Distribution: Preserved in all daughter dialects. Reconstruction shape: Correspondences are regular. The click efflux is reconstructed as "regular aspirated", despite the (probably erroneous) transcription with a glottal stop in C. Doke's Grootfontein materials. The original vowel is *-i, undergoing regular diphthongization in Juǀ'hoan.
Bleek 1956: 402, 423, 546, 654. Quoted as !ʰá in [Bleek 1929: 57]. Fluctuating articulation of the click influx reflects former retroflex (!̯) articulation.
Distribution: Preserved in all daughter dialects. Reconstruction shape: Correspondences are regular and indicate an original retroflex click, still preserved in the Grootfontein dialect. Tonal pattern is most probably level even, based on Ekoka data.
Distribution: Preserved in all daughter dialects. Reconstruction shape: Correspondences are regular and trivial, indicating retroflex click articulation in Proto-North Khoisan.
Bleek 1956: 485; Bleek 1929: 59. Polysemy: 'stone / mountain'. The latter source also lists ɳ!ɔ́ː as a synonym, but the form is unconfirmed in [Bleek 1956].
Doke 1925: 159. Polysemy: 'stone / mountain', although Doke postulates tonal differences between the two: ɳ!ȕːm 'mountain' vs. ɳ!ūːm 'stone'. No other dialect confirms any such differentiation.
Bleek 1956: 485. Not quite certain, since the word is not found under 'mountain' in [Bleek 1929]; no better candidates, however, are recorded in [Bleek 1956].
König & Heine 2008: 53. Quoted as ʔɳ!òḿ (Western dialect), ɳ!om (Eastern dialect) in [Heikkinen 1986: 25]. Both sources gloss the word as 'hill' and explicitly distinguish this root from the very similar-sounding 'stone' q.v.
Proto North Khoisan:*ʔɳ!òm
Distribution: Preserved in all daughter dialects. Reconstruction shape: Both major sources on Ekoka agree that the word is pronounced with a glottalized nasal click rather than simple nasal in the case of 'stone' q.v. Considering the usual polysemy 'stone / mountain' in African languages, this phenomenon looks very strange. On the other hand, if the two roots were indeed different in the protolanguage, it would have been natural for them to contaminate. Provisionally, following the data, we reconstruct 'mountain' with a preglottalized nasal click and 'stone' q.v. with a simple nasal one. C. Doke does not mark any differences in click articulation, but mentions that the two words have different tonal characteristics ; this is not confirmed by Ekoka or Juǀ'hoan data at all. The basic tonal pattern is low-low or low-high (as in Ekoka).
Bleek 1956: 216. Quoted as cíː in [Bleek 1929: 59]. The latter source adds ǀǀkʼwe as an alternative synonym, but the word is not confirmed either in [Bleek 1956] or by external data.
Distribution: Preserved in all daughter dialects. Reconstruction shape: Correspondences are regular and trivial, except for an unexpected dental (rather than the expected alveolar) click transcription by C. Doke in the Grootfontein dialect; possibly just a misprint.
Bleek 1956: 556. Quoted as ǀǀãũ in [Bleek 1929: 60]; in [Bleek 1956], the given meaning is 'back of the neck', but no separate root for 'front of the neck' is indicated.
Distribution: Preserved in all daughter dialects. Reconstruction shape: Correspondences are generally regular, indicating an old retroflex click in the initial position. Coda correspondences fall under the recurrent pattern "Juǀ'hoan -ãĩ ~ -ãũ : Ekoka -aŋ" that we provisionally mark as reflecting Proto-North Khoisan *-aŋi and *-aŋu respectively (alternately, one could reconstruct *-aiŋ and *-auŋ, but, with three vocalic morae in a row, that would contradict the general laws of Bushman phonotactics); see 'green' and 'tree' for more examples.
Bleek 1956: 264. Quoted as zɛː in [Bleek 1929: 61]. The same source adduces one more synonym: gàù, later quoted as gàùˤ in [Bleek 1956: 45]. This word is cognate to Juǀ'hoan gàˤʔú 'to be clean', and it is not clear whether a semantic shift has actually taken place or if the whole thing is merely a matter of inadequate translation.
Distribution: Preserved in all daughter dialects. Reconstruction shape: Correspondences are mostly regular and trivial, with the exception of the tonal pattern that ranges from simple rising in Juǀ'hoan to ultra-low in Ekoka (with additional aspiration, frequently accompanying low tones). It is also unclear whether the Juǀ'hoan form zàíˤ, with a dipthong and pharyngealization, is merely a dialectal variant or a different root. The reconstruction of the coda as simple *-e is therefore provisional.
König & Heine 2008: 67, 68. Quoted as ʒgxáŋ 'nostrils' in [Heikkinen 1986: 22].
Proto North Khoisan:*ckxúŋ́
Distribution: Preserved in all daughter dialects. Reconstruction shape: Correspondences are complex and relatively rare. In the initial position, Ekoka as well as some other dialectal data indicate an old affricate cluster *čkx- (phonologically opposed to the simple post-alveolar glottalized affricate *čʼ-). The coda contains a velar nasal, presumably with a preceding labial vowel (reflecting the correspondence "Juǀ'hoan -ũ : Ekoka -(a)ŋ"), although this particular part of the reconstruction is provisional.
Distribution: Preserved in all daughter dialects. Reconstruction shape: Correspondences are generally regular, although there are occasional fluctuations of vocalism in this negative marker, possibly due to sandhi-type change. !O!Kung dialects as recorded by D. Bleek also tend to drop click articulation.
König & Heine 2008: 48. Quoted as ɳǀeʔe in [Heikkinen 1986: 23] (no tone marks).
Proto North Khoisan:*ɳǀèʔe
Distribution: Preserved in all daughter dialects. Reconstruction shape: Correspondences are regular and trivial except for the tonal pattern, where the best-transcribed dialects contradict each other. Most likely, the double marking of the vowel in ǀǀKxauǀǀen and !Kung orthographically indicates a sequence of two vowels divided by a glottal stop.
Doke 1925: 152. Although the immediate source of borrowing is unknown, the word cannot possibly be of native origin. The plural form, however, is the same as in other North Khoisan dialects: žùː ~ ǯùː 'people' [Doke 1925: 155].
König & Heine 2008: 86; Heikkinen 1986: 24. Secondary synonym: ǯù-ǀxòà [König & Heine 2008: 23]. The latter term is not as well confirmed by textual examples and may be more stylistically marked (it should be noted that the simple root ǯù 'person' has shifted in Ekoka to the meaning 'we excl.', although the form zù 'man, person' is still quoted in [Heikkinen 1986: 26] next to the pronominal meaning).
Proto North Khoisan:*ǯù
Distribution: The original root *ǯù is found in all the major dialect clusters. Replacements: (a) atȁ in C. Doke's Grootfontein records, of unclear origin (if the word is transcribed correctly, it cannot be a native item); (b) Ekoka !xũ̄ṹ < '!Xung, (Bushman) male human being'. Reconstruction shape: Correspondences are generally regular and trivial. It is remarkable that in Juǀ'hoan, the word forms its plural through tonal change, the only recorded (and so far unexplained) instance of this phenomenon in North Khoisan. Semantics and structure: Of note is the compound form *žù-ǀʰõã, lit. 'true person', denoting North Khoisan-speaking people; given its presence in both Juǀ'hoan and Ekoka, it has to be projected onto the proto-level as a very archaic ethnic self-designation.
Distribution: Preserved in all daughter dialects. Reconstruction shape: Correspondences are regular and trivial, indicating an original voiced retroflex click and a low tonal pattern.
König & Heine 2008: 29. Adjectival form; the corresponding verbal stem 'to be red' is glossed as ɡ!àè (with tonal variation). Quoted as ɡ!àē (Western dialect), ɡ!èē (Eastern dialect) in [Heikkinen 1986: 24].
Proto North Khoisan:*ɡ!ã̀ ~ *ɡ!ae
Distribution: Preserved in all daughter dialects. Reconstruction shape: The Northern dialect cluster generally agrees on the reconstruction *ɡ!ae, whereas the rest of the dialects agree on *ɡ!ã. This may either reflect some rare, non-trivial combination of features (a nasalized diphthong?), or, more likely, two different morphological variants, making it likely that the original root was simply *ɡ!a-.
Dickens 1994: 248. Synonym: ǂʰà 'path, road'; external comparison shows this to be the older root, but it is unclear which of the two words is currently used as the default word for 'road'.
Bleek 1956: 295; Bleek 1929: 64. The latter source quotes dàù-sa, a transparent borrowing from Central Khoisan, as a synonym, confirmed also in [Bleek 1956: 22]. It is not quite clear which root is the default one in the transcribed dialect.
Bleek 1956: 660; Bleek 1929: 64. The latter source quotes kú as a synonym. In [Bleek 1956: 103] this word is translated as 'footprints, road', with text examples illustrating both meanings. Since external data are clearly in favor of ǂʰa as the more archaic designation of 'path, road', it may be assumed that the older root ku 'footprints' has not yet firmly replaced the old 'road' in the dialect(s) described by L. Lloyd, as it has, for instance, in Ekoka !Kung.
König & Heine 2008: 42. Quoted as kūú in [Heikkinen 1986: 21]. Meaning given as 'path, road, way'. Derived form: kȕʰù-mà 'path' (with the regular diminutive suffix). Secondary synonym: !̯ʰà 'path' [König & Heine 2008: 90] (arguably not eligible, since the meaning 'road' is not glossed; the word is likely to be restricted for designating explicitly small paths, rather than the "default" type of walkway).
Proto North Khoisan:*ǂʰà
Distribution: The original root *ǂʰà 'path, road' is relatively stable according to old data sources, but its usage seems to have diminished in recently transcribed dialects, possibly due to its being replaced by lexemes denoting different kinds of 'roads'. However, this may be a dictionary-based illusion. Replacements: (a) Juǀ'hoan ɳ!àmà, with no known parallels or sources; (b) Ekoka kȕhù {'footprint' > 'footpath' > 'road'} (see notes on !Kung). Reconstruction shape: Correspondences are regular and trivial. D. Bleek seems to have mistranscribed a dental click for both ǀǀKxauǀǀen and !O!Kung instead of palatal articulation (a rather common error in her records).
Dickens 1994: 209. Secondary synonym: ǀǀàrì [Dickens 1994: 329]. The form is slightly less reliable, since (a) it is only quoted by P. Dickens after J. Snyman's earlier dictionary (not attested by himself); (b) related forms in other dialects of the Central and Southern clusters usually have it in the meaning 'root fibre'.
Bleek 1956: 375. Erroneously quoted as ɡǀãĩ in [Bleek 1929: 71]; later, confused by D. Bleek with the phonetically similar, but ultimately different forms for 'tree' (the difference is clearly seen in Juǀ'hoan: ɡ!ȁnì 'root' vs. !ã̏ĩ̀ʰ 'tree'). There are also several forms with the meaning 'root fibre' ([Bleek 1929: 71]: ǀǀari, ǂabe) that are ineligible.
Bleek 1956: 448. Translated as 'main root of plant'; quoted as !ube 'root' in [Bleek 1929: 71]. The latter source also lists, in the meaning 'root fibre', the forms ǀǀeri and daŋ; external data show that ǀǀeri may indeed mean 'fibre', whereas daŋ is more likely to be a metaphor for 'root' proper, since its usual meaning is 'bottom'.
Bleek 1956: 554. Dubious; the meaning is given as 'branch; root fibre'. In the exact meaning 'root' the word is quoted as ǀǀari in [Bleek 1929: 71]. Only a complete lack of synonyms and the fact that the same word functions as 'root' in the closely related Ekoka !Xung keeps us from leaving the slot unfilled.
König & Heine 2008: 97. Quoted as ǀǀàrí in [Heikkinen 1986: 25]. Polysemy: 'root / handle'. Secondary synonym: ɳǀǀàì 'root, carrot' [König & Heine 2008: 58].
Proto North Khoisan:*ɡ!ani #
Distribution: Proto-North Khoisan *ɡ!ani is well attested in Juǀ'hoan and ǀǀKxauǀǀen, restricting its total absence only to the Northern dialect cluster, and has no known semantics other than simply 'root'. Still, it is frequently competing with *ǀǀari, also quoted with the meaning 'root' for Juǀ'hoan as well as Ekoka !Xung (Northern cluster). On the other hand, the latter in some old sources is glossed as 'root fibre' rather than 'root' (D. Bleek very explicitly states the difference and lists minimal pairs in [Bleek 1929]); the semantic transition {'root fibre' > 'root'} is easily understandable and could have occurred in the common ancestor of all Northern cluster dialects. Replacements: (a) !Kung !ubbe, of unclear origin, perhaps = Juǀ'hoan !úbè 'species of shrub' [Dickens 1994: 317]; (b) !O!Kung ǀǀale, etc.,see above. Reconstruction shape: The protoform *ɡ!ani is tentatively modeled after the Juǀ'hoan form, but, since there are no correlates in the Northern dialect cluster, could have easily been *ɡ!̯ani as well.
Distribution: Preserved in all daughter dialects. Reconstruction shape: See 'earth'. Semantics and structure: The meanings 'earth' and 'sand' were not differentiated in Proto-North Khoisan any more than they are differentiated in modern day descendants.
Bleek 1956: 154. The form is a composite one, consisting of o 'to make, do' ([Bleek 1956: 152]) + kxwi 'to speak' ([Bleek 1956: 128]) = Juǀ'hoan òː-kxúí 'speak, talk'. Although existing textual examples do not show the meaning 'say' for kxwi as such, this may be accidental; [Bleek 1929: 72] quotes both kxwí and o=kxwí in the meaning 'to say', but only o=kxwí in the meaning 'to speak'. Another synonym is ǂá [Bleek 1956: 653] ~ ǂáː [Bleek 1929: 72] 'to say, talk', with no external parallels.
Bleek 1956: 154. Quoted as kxwí ~ o=kxwi in [Bleek 1929: 72]. See notes on the ǀǀKxauǀǀen entry (most of them apply to the !Kung item as well). The latter source also quotes kue as a synonym; the form is not confirmed in [Bleek 1956].
Bleek 1929: 72. Neither of the forms is confirmed in [Bleek 1956]. Additionally, [Bleek 1929: 72] lists kʼwi = kxwí in [Bleek 1956: 128], but with textual examples confirming the semantics of 'speak, sound' rather than 'say' proper. A third synonym is ɳǀǀaŋ, also unconfirmed in [Bleek 1956]. Quoted as ka in [Snyman 1980: 48].
König & Heine 2008: 41, 42. Quoted as kòè (Western dialec), kè (Eastern dialect) in [Heikkinen 1986: 21]. All the variants look like idiolectal variants of the same word, although the fluctuations between -oe and -uya cannot be easily explained.
Proto North Khoisan:*kò
Distribution: The original root *kò still functions as the main term for 'say' in Juǀ'hoan and Ekoka, the two best recorded North Khoisan dialects that represent two distant clusters; this is clearly the best candidate for proto-status. Replacements: (a) ǀǀKxauǀǀen o=kxwi, etc.; a compound from Proto-North Khoisan *o 'to do, make' + Proto-North Khoisan *kxúí. Both in Juǀ'hoan and Ekoka this compound is translated as 'speak, talk (about smth.)' rather than 'say (smth. specific)'. Furthermore, P. Dickens even assigns the morpheme kxúí a pronominal rather than verbal meaning: 'be thus, be so (e. g. of the sound, sight or way of doing something)' [Dickens 1994: 232], although a more detailed analysis of contexts is needed to clarify the situation. Still, some of the examples given by D. Bleek in [Bleek 1956] suggest that this compound form may have actually replaced the original *kò in at least some of the described dialects, i. e. the original *'speak' and *'say' have merged in one word. Reconstruction shape: The basic form of the root is *kò, as seen in Juǀ'hoan; Ekoka kòè (and the contracted variant kè) is likely to represent a suffixal extension (possibly old).
Dickens 1994: 219. The form sé is listed as a synonym in the English-Juǀ'hoan section of the dictionary (p. 144), but is only translated as 'look (at), look after, investigate' in the main section.
Bleek 1956: 62, 169, 176; Bleek 1929: 72. Quoted as sĩ́ŋ in [Snyman 1980: 48]. The latter source also quotes ɳǀe as a synonym, but the root is not explicitly confirmed in this meaning in [Bleek 1956: 345], with 'perceive, understand, know' more likely to be its general semantics.
König & Heine 2008: 36. Quoted as háŋ (Western dialect), sáŋ (Eastern dialect) [Heikkinen 1986: 22].
Proto North Khoisan:*séŋ́ ~ *héŋ́
Distribution: Preserved in all daughter dialects except Juǀ'hoan. Replacements: Juǀ'hoan hȍ < Proto-North Khoisan *hȍ(ò) 'to find' (cf. Ekoka hȍhò id. [König & Heine 2008: 37], etc.; a basic semantic shift {'to find' > 'to see'}. Reconstruction shape: Juǀ'hoan and ǀǀKxauǀǀen drop the nasal part of the coda in this root just as they do for 'drink' q.v., for not quite clear reasons; nevertheless, most of the dialects, including those from J. Snyman's survey, confirm -ŋ. The vowel is tentatively reconstructed as -e- based on Juǀ'hoan data (in most dialects the two codas, -iŋ and -eŋ, seem to have merged). Initial *s- is occasionally lenited to h-; this is either a unique positional development before a syllabic nasal, or reflects an original aspirated *sʰ- (not enough data to reach a definite conclusion).
Dickens 1994: 329. Polysemy: 'seed / kernel / marrow'; more precise denotation of the former is ǀǀxàrà-ǀǀàʔá, lit. 'plant-seed'.
ǀǀKxauǀǀen:ǂʼaː2
Bleek 1956: 642 (quoted in the plural form: ǂʼaː-si 'seeds'). [Bleek 1929: 72] has an entirely different entry: ɳǀoː-si (also a plural form), but this is almost certainly a mistake, since ɳǀoː actually means 'skin, bark, shell' q.v.; in the generally more reliable source [Bleek 1956] the meaning 'seed' for this root is unattested.
!Kung:
Not attested (seemingly). [Bleek 1929: 72] quotes the plural form ɳǀeː-siŋ in the meaning 'seeds'; the basic meaning of the root is 'head' (q.v.) and, although it is commonly used in the metaphorical meaning 'bulb' or 'berry' (see text examples in [Bleek 1956: 346]), the meaning 'seed' (of fruit or berry) is not really confirmed by any other source, be it this particular dialect or any other North Khoisan language.
König & Heine 2008: 82. Polysemy: 'grain / seed'. Glossed as !ó 'pip, grain' in [Heikkinen 1986: 24]. Secondary synonym: kōbā [König & Heine 2008: 41] (also, as a verb, = 'to plant'; not confirmed in [Heikkinen 1986]).
Proto North Khoisan:*!ó
Distribution: Proto-North Khoisan *!ó is well preserved in the Northern dialect cluster; in Juǀ'hoan, !ó is still attested with the meaning 'pip' [Dickens 1994: 315]. Replacements: (a) Juǀ'hoan ǀǀàʔá = Ekoka ǀǀã̀ʔã̀ 'bone marrow', implying a shift like {'pith' > 'seed'}; (b) ǀǀKxauǀǀen ǂʼaː of unclear origin (the form itself is somewhat dubious). Reconstruction shape: Correspondences are regular and trivial.
König & Heine 2008: 48. Quoted as ʔɳǀíŋ (Western dialect), ɳǀíŋ (Eastern dialect) in [Heikkinen 1986: 23]. Singular subject action verb; the plural subject form is ɡ!ʰō [ibid.].
Proto North Khoisan:*ʔɳǀáŋ́
Distribution: Preserved in all daughter dialects. Reconstruction shape: Correspondences are regular and trivial. Preglottalized nasal click is reconstructed based on T. Heikkinen's transcription of the Ekoka entry. The coda is probably the same as in 'blood' q.v. and, alternately, may be reconstructed as syllabic *-ŋ. Semantics and structure: Like many basic verbal stems, this one is also suppletive: the plural subject form is also preserved everywhere and is safely reconstructible as *ɡ!ʰo.
Bleek 1956: 348; Snyman 1980: 49. Quoted as ɳǀɔ́ in [Bleek 1929: 76]. The latter source adds da as a synonym, but the word rather means 'cloth, kaross', as can be seen from both the examples in [Bleek 1956: 19] and external parallels.
Distribution: Preserved in all daughter dialects. Reconstruction shape: Correspondences are regular and generally trivial. Some old sources seem to reflect an additional suffixal variant *ɳǀo-a (or ɳǀo-ba?) that is not, however, encountered in recently transcribed material.
Bleek 1956: 210. Quoted as cá: in [Bleek 1929: 76]. The stem ɡǂa, quoted in the latter source as a synonym, is actually the plural subject form of the verb 'to lie (down)'.
König & Heine 2008: 67. Quoted as cʼā in [Heikkinen 1986: 22].
Proto North Khoisan:*cʼā
Distribution: Preserved in all daughter dialects. Reconstruction shape: Correspondences are regular and trivial (including the development *cʼ- > čʼ- in Ekoka).
Dickens 1994: 279. The simple form cʼè is used after nouns with diminutive suffixes (čʼu-macʼè 'small house'); the compound form cʼè-mà (where -mà itself is a diminutive suffix) is used more frequently.
Bleek 1956: 215; Bleek 1929: 76. The latter source also quotes such ineligible forms as -ma (actually a diminutive suffix) and ɡ!ɔ̀ (actually 'short', as clarified by examples in [Bleek 1956]).
Bleek 1956: 215. Quoted as ce-ma in [Bleek 1929: 76]. The latter source also quotes -ma separately in the same meaning (the morpheme is actually a diminutive suffix).
Doke 1925: 153. The same source also quotes māː per se [ibid.] in the meaning 'small', interchangeable with cʼēː-mā (e. g. dàː māː and dàː cʼēːmā both mean 'small blanket'). It is perhaps reasonable, as in several other North Khoisan dialects, to think of -māː as a diminutive suffix and cʼēː-mā as the proper adjective 'small'.
Bleek 1956: 215. Quoted as ce-ma in [Bleek 1929: 76]; čʼèmà ~ cʼèmà in [Snyman 1980: 50]. [Bleek 1929] also quotes the compound form -ma !oː as a synonym, where -ma is the standard diminutive suffix and !oː = Juǀ'hoan !ò- in !ò-mà 'short'. Textual examples in [Bleek 1956: 436] confirm the usage of both -ma-!oː and simply !o in the meaning 'small' (e. g. !ü !ó 'a small star'), but the semantic difference between these cases and ce-ma remains unclear. We mark the latter form as primary, since its general distribution in the meaning 'small' indicates its archaic nature.
König & Heine 2008: 68. Quoted as cʼema in [Heikkinen 1986: 22] (without tonal marking).
Proto North Khoisan:*cʼe ~ *cʼe-mà
Distribution: Preserved in all daughter dialects, although, for the most part, only in the form of the derived stem *cʼe-mà. Reconstruction shape: Correspondences are regular and trivial, with the exception of the tonal pattern (the Ekoka mid-level pattern may be more archaic, with assimilation to the low tonal level of the suffix *-mà in Juǀ'hoan). Semantics and structure: The root *cʼe, already on Proto-North Khoisan level, must have been most frequently used with the diminutive suffix *-mà, usually (but not always) adjacent to nominal roots. There is, however, no solid evidence for *mà having ever functioned as an independent adjective 'small' on the same chronological level.
König & Heine 2008: 19. Polysemy: 'smoke / tobacco'. Quoted as súrē in [Heikkinen 1986: 26]. Secondary synonyms: (a) !̯ ȍhà [König & Heine 2008: 91]; (b) ɳǀǀòʔò [König & Heine 2008: 60]. Semantic differences between all these stems are unclear; the only transparent external connections with the meaning 'smoke' are displayed by šúlè.
Proto North Khoisan:*šórè ~ *šórà
Distribution: Preserved in most daughter dialects. Replacements: !O!Kung ǀɔnu, of unknown origin (although the original root is still preserved as čule ~ čuli 'tobacco, snuff' [Bleek 1956: 237]; the semantic shift {'smoke' > 'tobacco'} is trivial). There is some phonetic similarity between ǀɔnu and Nama ǀʼanni 'smoke', but not enough to postulate borrowing with any degree of certainty. Reconstruction shape: Correspondences are regular and trivial except for the unclear vocalic variation in the second syllable. Semantics and structure: Polysemy {'smoke' & 'tobacco'} is widespread for this root both between and within dialects, but there is no clear evidence for 'tobacco' having been the original semantics.
Bleek 1956: 471, 486. Quoted as ɳ!ṹ in [Bleek 1929: 79] (both variants found in [Bleek 1956] represent suffixal extensions); ɳ!ṹ in [Snyman 1980: 50]. Singular subject action verb; the plural subject form is ɡǀǀa ~ ɡǀǀà [Bleek 1956: 522].
König & Heine 2008: 84. Quoted as ʔɳ!ṹ (Western dialect), ɳ!ṹ (Eastern dialect) in [Heikkinen 1986: 25]. Singular subject action verb; the plural subject form is ɡǀǀà [ibid.].
Proto North Khoisan:*ʔɳ!ṹ
Distribution: Preserved in all daughter dialects. Reconstruction shape: Correspondences are generally trivial, although T. Heikkinen's data on Ekoka conflict with König & Heine as to the exact articulation of the click efflux. Perhaps the rare case of variation between !- and ɳ!- in König & Heine's data actually reflects an unrecognized case of preglottalized nasalisation. Semantics and structure: The etymon was almost certainly suppletive in Proto-North Khoisan: sg. action *ʔɳ!ṹ vs. pl. action *ɡǀǀà. There was also a derived verbal stem *ʔɳ!ṹ-a, whose protolanguage meaning is hard to establish (P. Dickens gives a causative meaning for Juǀ'hoan, but for Ekoka, König & Heine adduce a continuative meaning instead).
König & Heine 2008: 92. Quoted as ǂũ̏ in [Heikkinen 1986: 23].
Proto North Khoisan:*ǂũ̏
Distribution: Preserved in all daughter dialects. Reconstruction shape: Correspondences are generally regular and trivial. The tonal pattern is ultra-low, frequently accompanied with breathy articulation in Juǀ'hoan.
König & Heine 2008: 52. Quoted as ɳ!òm̀ in [Heikkinen 1986: 25]. Probably the same word as 'mountain' q.v., but both sources on the Ekoka dialect mark different types of nasalized clicks for the two words. Secondary synonyms: (a) ǀúnní 'stone; money' [König & Heine 2008: 75]; (b) ǀxùm̄ 'stone; flint' [König & Heine 2008: 76].
Proto North Khoisan:*ɳ!òm̀ ~ *ɳ!ùm̀
Distribution: Preserved in all daughter dialects. Reconstruction shape: Correspondences are regular. There seems to have been a neutralization of the u / o opposition before syllable-final *-m in Proto-North Khoisan, hence all the vocalic fluctuations. Semantics and structure: Further discussion on the polysemy {'stone' & 'mountain'} see under 'mountain'.
Doke 1925: 156. Also encountered in the same source as ɡǀam 'sun, day' [Doke 1925: 157]; this may be an idiolectal variant or a most recent re-borrowing from a Central Khoisan source.
König & Heine 2008: 24; Heikkinen 1986: 22. Polysemy: 'sun / day / time / thirst'.
Proto North Khoisan:*ǀáḿ
Distribution: Preserved in all daughter dialects with the sole exception of Ekoka. Replacements: Ekoka gàò is attested in this dialect alongside ǀám, which is, however, only assigned the secondary meanings 'day, hour' [König & Heine 2008: 73]. In theory this raises the question of whether this could not have been the original North Khoisan word for 'sun', whereas all the instances of *ǀáḿ could then count as more recent borrowings from Central Khoisan (cf. Proto-Central Khoisan *ǀáḿ id.). However, given our current information, the answer must be negative, since: (a) in Khoekhoe, the subgroup that includes Nama and serves as the most common source for recent Central Khoisanisms, the word for 'sun' is *sore-; (b) the general distribution very clearly speaks for the Proto-North Khoisan status of *ǀáḿ, regardless of whether its further connections with Central Khoisan are horizontal or vertical; (c) Ekoka gàò is clearly related to Juǀ'hoan gàʔáró, glossed by Dickens as 'to drink too little to quench one's thirst' [Dickens 1994: 200], i. e. basically '(still) be thirsty'. This implies that 'thirst' may, after all, be the original meaning for this root. The semantic shift {'thirst' > 'sun'} is certainly unusual, but, considering that the polysemy {'sun' & 'thirst'} is quite typical of all Bushman languages, not at all impossible. Reconstruction shape: Correspondences are regular and trivial throughout.
Bleek 1956: 29. Quoted as dum in [Bleek 1929: 82]. The latter source also adds ǂʰau as a synonym; this word is, however, translated as 'to walk, travel, crawl, stalk game, swim' in [Bleek 1956: 651] and is the regular equivalent of the item 'to walk, move' in other North Khoisan dialects.
Bleek 1956: 480, 482; Bleek 1929: 82. Somewhat dubious; the translation equivalent in [Bleek 1956] is given as 'to row across, swim across' for ɳ!ɔ̀va and simply 'row' for ɳ!òba. No better suitable equivalents have been found. [Snyman 1980: 52] lists this stem, transcribed as ɳ!òaˤ, as synonymous with dòm; the latter is clearly cognate with Ekoka dhȍm̀, but cannot be listed here as the main item since [Snyman 1980] is not the main source.
Distribution: The root *dȍm̀ in the meaning 'to swim' has the widest distribution, being attested in the Northern cluster (Ekoka) as well as L. Lloyd's !Kung data. Replacements: (a) Juǀ'hoan ǯxà, of unknown origin; (b) !O!Kung ɳ!ɔ̀va, with a dubious meaning (see notes on !O!Kung). Reconstruction shape: The protoform is essentially based on the Ekoka form (the only more or less reliably transcribed item). Semantics and structure: Apart from the meaning 'swim', the root is also found in some sources in the meanings 'wash', 'bathe', and even 'shelter from rain' (Heikkinen). This is consistent with the general (and understandable) scarceness of water basin-related terminology in native North Khoisan lexicon.
Bleek 1956: 433, 667; Bleek 1929: 82. Variation in click articulation (! ~ ǂ) may have something to do with the original retroflex articulation of the click (!̯), but not necessarily.
Distribution: Preserved in all daughter dialects. Reconstruction shape: Correspondences are regular; click correspondences clearly indicate original retroflex articulation, and the diphthong is to be reconstructed as *-oe based on Ekoka (and !Kung) data.
Bleek 1956: 227, 229; Bleek 1929: 83. An uncertain entry; the word is ascribed both the meanings 'this' and 'that', indicating a general rather than proximity-specific deictic stem. [Bleek 1929: 83] also lists úhe as a potential synonym, but the word is not confirmed in [Bleek 1956].
Bleek 1956: 27. Uncertain (there is a semantic difference indicated for two near-homonymous forms dóä 'this' and doä́ 'that', with not enough textual evidence to fully justify this). In [Bleek 1929: 83] the item 'that' is rendered by two different morphemes, ši and ka, neither of which finds confirmation in [Bleek 1956]; the former most likely refers to some specific auxiliary use of the noun ši 'thing', and the latter may be a particle mistaken for a deictic pronominal stem (or, perhaps, the third person pronoun ka 'it').
Bleek 1956: 27. Attested textual examples do not provide the required contexts, but external comparison with Ekoka material indirectly confirms this as the basic stem for far deixis. Quoted as n doa in [Bleek 1929: 83].
König & Heine 2008: 46, 70. The derived nominal stem ('that thing') is čá-à-ndòʔà [König & Heine 2008: 64]. All forms seem to be in free variation.
Proto North Khoisan:*toʔà ~ *ndoʔà
Distribution: Preserved in most daughter dialects, with the possible exception of ǀǀKxauǀǀen. Replacements: (?) ǀǀKxauǀǀen či, of unclear provenance (perhaps = Proto-North Khoisan *či 'thing'?). Reconstruction shape: No single certain variant; the dialectal fluctuation as reflected in four distinct Ekoka variants, including structurally rare forms with initial nasal clusters, may go as far back as Proto-North Khoisan. Semantics and structure: The form is analyzable as a composite stem, from *to ~ *ndo, indicating far deixis, and *ʔa 'general relative morpheme' (= Juǀ'hoan -à [Dickens 1994: 188]). Initial *nd-, a cluster normally prohibited in North Khoisan, may only reflect traces of further compounding, perhaps from some earlier unclear structure like *Vn-to.
Bleek 1956: 227, 229; Bleek 1929: 84. An uncertain entry; the word is ascribed both the meanings 'this' and 'that', indicating a general rather than proximity-specific deictic stem. [Bleek 1929: 84] also lists ɳǀi and ǀe as potential synonyms, but the latter is assigned only the meaning 'here' in [Bleek 1956], and textual examples for the former are rather confusing.
Bleek 1956: 36, 38. Quoted as e ~ eːa ~ eːya in [Bleek 1929: 84]. In [Bleek 1956: 27] the stem dóä is also given in the meaning 'this', with no clear differentiation between the two. Considering, however, that the latter is translated both as 'this' and 'that', and comparing this situation to the distinction in Juǀ'hoan, it is likely that e is still the most neutral morpheme in this dialect to designate near deixis.
Grootfontein !Kung:
Not attested.
!O!Kung:ŋ3
Bleek 1956: 141; Bleek 1929: 84. The latter source adds či as a synonymous form. Its existence is corroborated in [Bleek 1956: 229], however, examples for ŋ are more numerous. [Snyman 1980: 52] under 'this' lists the form àŋ, which is clearly the same item.
Ekoka !Xung:ŋ̄ŋ̀3
König & Heine 2008: 61. According to König & Heine's grammar sketch, this stem is the "neutral" version of the close-deixis pronoun. The word ē [König & Heine 2008: 24], on the other hand, has a contrastive function: 'this (particular), rather than any other one'. It is also said to be used less frequently than ŋ̄ŋ̀, and is therefore ineligible for the wordlist.
Proto North Khoisan:*ŋ
Distribution: Two main morphemes glossed in English as 'this' are attested in North Khoisan: *e and *ŋ. The presence of both at the same time is noted only for Ekoka. However, since there is no evidence, nor any general reason, to imply an external provenance for *ŋ in the Northern dialect cluster, the simplest solution is to project the opposition onto the proto-level, stating that Ekoka (and, possibly, other dialects of the Northern cluster) preserves the original distinction between *ŋ 'this (one)' and *e 'this (one) in particular', whereas the other clusters have generalized *e as the only morpheme for both cases. This hypothesis will remain the likelier one until an external etymology has been found for *ŋ that would explain its apparition in the Northern cluster.
Bleek 1956: 6; Bleek 1929: 85. The suffixless form a is also quoted in the meaning 'thou' in [Bleek 1929], but mentioned only as the possessive form 'thine' corresponding to the independent form a-hi 'thou' in [Bleek 1956].
Bleek 1956: 3. Cf. also the "isolated" form á-hn [Bleek 1956: 6]. [Bleek 1929: 85] does not mention the existence of these forms, but quotes instead a different form: ma. [Bleek 1956: 132] shows that the more correct transcription is mʔa, and that the form is regularly used in subject position, whereas a is more common as the object and the possessive form. This is similar to the difference between à and bà in Ekoka q. v. The form mʔ-a can thus be analyzed as -a (root morpheme) with a "subject" particle prefixed.
Bleek 1956: 4, 6; Bleek 1929: 85; Snyman 1980: 56. Both the simple and complex form are attested in textual examples in subject position as well as in the possessive function.
König & Heine 2008: 15. According to König & Heine's grammar sketch, à is the pronominal stem used in all syntactic contexts. In the subject position, however, especially sentence-internally, it tends to be replaced with bà [ibid.]. The latter is probably best viewed as the result of merger between à and some other morpheme (emphatic particle, etc.).
Proto North Khoisan:*à
Distribution: Preserved in all daughter languages. Reconstruction shape: Correspondences for this monovocalic morpheme are perfectly regular and obvious. The special emphatic variant fluctuates between *à-hi and *à-hŋ; it is possible that *à-hŋ is the primary variant, restructured in some dialects by analogy with the 1st p. emphatic form *mi-hi q.v. Odd forms with initial labial consonants should probably be treated as compounds, although the meaning and function of these initial labial elements are not yet understood.
König & Heine 2008: 24. Quoted as dʰàrrì in [Heikkinen 1986: 24].
Proto North Khoisan:*ntʰari
Distribution: Preserved in all daughter dialects. Reconstruction shape: Correspondences are generally trivial, with the exception of the initial consonant, with irregular fluctuation of laryngeal features and even completely unexpected prenasalization in Grootfontein !Kung. Likewise, König & Heine actually transcribe the initial consonant in Ekoka as a very rare example of a "fortis" initial stop (spelling the word as {d̀thàlì}, with the initial voiced onset bearing a low tone) [König & Heine 2008: 24], which shows that Doke's prenasalization was certainly not an accidental transcriptional error. This is consistent with the typologically aberrant (both for Africa and other world areas) phonetic behavior of the word 'tongue' and indicates some sort of hard-to-reconstruct phonosemantic phenomenon. The situation is provisionally denoted in the guise of a unique initial cluster *ntʰ- (= the variant attested in [Doke 1925]), although, obviously, this is far from the only way to interpret the data. Tonal correspondences are also complex, although P. Dickens and J. Snyman in [Snyman 1980] agree on the tonal pattern "extra low-low" for two dialects from different clusters.
König & Heine 2008: 68. Quoted as cʼāō in [Heikkinen 1986: 22].
Proto North Khoisan:*cʼau
Distribution: Preserved in all daughter dialects. Reconstruction shape: Correspondences are regular and trivial except for the tonal pattern (low in Juǀ'hoan vs. mid-level in Ekoka, not allowing for an adequate reconstruction).
Distribution: Preserved in all daughter dialects. Reconstruction shape: Correspondences are complex and, in some aspects, unique. The basic structure of the word is more or less the same as in 'neck' q.v., which is reflected in the reconstruction of the coda *-aŋi for both items. Seemingly random fluctuations are, however, observed in secondary click articulation (ranging from simple velar release to prevoicing to aspiration), additional vowel properties (breathy articulation in Juǀ'hoan) and the tonal scheme (usually low tones, but cf. also the ultra-high notation in Snyman's !O!Kung data). The provisional reconstruction with *ɡ!ʰ- and ultra-low tone on the first mora merely reflects the fact that some particularly complex bag of features must have been present on the proto-level in order to yield such a large variety of reflexations.
König & Heine 2008: 64. Quoted as cā in [Heikkinen 1986: 22].
Proto North Khoisan:*cā ~ *cã̄ ~ *cã̄ˤ
Distribution: Preserved in all daughter dialects. Reconstruction shape: The initial consonant is unambiguously reconstructed as *c-, but the coda demonstrates so far unexplainable variations of secondary features (nasalization and pharyngealization that seem to "come and go"). This is perhaps a reflection of former morphological variability, but there is no evidence to prove it.
Bleek 1956: 246; Bleek 1929: 42. The latter source also quotes such dubious synonyms as !ũ and ǀu, but their existence is not confirmed in [Bleek 1956].
Bleek 1956: 246. Quoted as ú in [Bleek 1929: 42]. The latter source adds !ú as a synonym, but it is only confirmed in the meaning 'to run' in [Bleek 1956: 446].
Bleek 1956: 246; Bleek 1929: 42. The latter source adds two other stems: ó, which may be an error or just a phonetic variant of ú, and toa, not confirmed in [Bleek 1956].
König & Heine 2008: 71. Also met in a compound with an etymologically unclear first part: ɳǀũ̄ṹ ú [König & Heine 2008: 50].
Proto North Khoisan:*ú
Distribution: Preserved in all daughter dialects. Reconstruction shape: This monovocalic stem is relatively trivial (Doke's transcription of it as a diphthong is not confirmed by any other source) and should be reconstructed with a high or, possibly, ultra-high tone. Semantics and structure: All sources agree on this item as representing the basic meaning of moving from one place to another, rather than anything more specific.
Distribution: Preserved in all daughter dialects. Reconstruction shape: Correspondences are regular and trivial. Semantics and structure: The word is antonymous to *ǂàʔū 'cold' q.v., and is strictly distinguished from *ǀǀʼṹ 'warm' (Juǀ'hoan, Ekoka ǀǀʼṹ, etc.).
Bleek 1956: 387, 535. Quoted as ɡ!úː in [Bleek 1929: 90]. Dialectal variation between alveolar and lateral articulation of the click indicates original retroflex (ɡ!̯) articulation.
Distribution: Preserved in all daughter languages. Reconstruction shape: Correspondences are regular, reflecting an original retroflex click and a high tonal pattern.
Bleek 1956: 36. D. Bleek notes a special variant for the masculine form (e-!ka) and one for the feminine (e-he) (ibid.; also [Bleek 1929: 90]). According to her notes, there is no difference between inclusive and exclusive.
Bleek 1956: 36; Bleek 1929: 90. Exclusive form. Both sources also list the synonymous form šiši; this must probably be an occasional Khoekhoeism, cf. Nama sí-ke 'we (excl. m.)', sí-se 'we (excl. f.)'; its statistical frequency is unclear.
Bleek 1956: 36, 37; Bleek 1929: 90. [Bleek 1929] also quotes i, not confirmed in [Bleek 1956] (probably a phonetic variant of e). D. Bleek does not record an inclusive/exclusive opposition for this dialect; J. Snyman, however, does, and quotes e as only the exclusive meaning [Snyman 1980: 54].
König & Heine 2008: 23. Exclusive form. Quoted as zù (Western dialect) in [Heikkinen 1986: 26]. The more archaic synonymous form è [König & Heine 2008: 24] is said by the authors to be much less used, meaning that it is in the final stages of being displaced by ǯù.
Proto North Khoisan:*è
Distribution: Preserved in all daughter languages except for Ekoka. Replacements: Ekoka ǯù, reflecting an interesting semantic shift {'people' > 'we'}; see 'person' for more details. Reconstruction shape: Correspondences for this monovocalic stem are regular and trivial. Semantics and structure: Proto-North Khoisan *è is unambiguously reconstructed as the exclusive version of the 1st pronoun pl.
Bleek 1956: 62; Bleek 1929: 90. Inclusive form. Said to be rarely used. Also attested in the dual variant: mː-sa ~ m-sa 'we two' [Bleek 1956: 138]. An alternate synonym is i [Bleek 1956: 67], "rarely used before numbers"; unclear form, lacking external parallels in other North Khoisan dialects.
Grootfontein !Kung:
!O!Kung:
D. Bleek does not record any special inclusive forms for this dialect. J. Snyman [Snyman 1980: 54], along with e, records the synonymous forms m and m-!a. Judging by the external parallels, the distinction must be between the exclusive e and the inclusive m, although Snyman does not mention any differences in meaning.
König & Heine 2008: 45. Inclusive form. The morpheme -hm̀ is probably assimilated from -hŋ̀, the standard emphasis marker [König & Heine 2008: 36].
Proto North Khoisan:*m̀
Distribution: The original form is not attested in some of the dialects, which could indicate occasional loss of the clusivity opposition. On the other hand, most of these non-attestations are evident in old sources, whereas most modern descriptions clearly state the presence of this opposition. Reconstruction shape: Correspondences are generally regular and trivial. The ultra-low tone in Ekoka may be connected with the aspiration of the emphatic particle. Semantics and structure: Proto-North Khoisan *m̀ is unambiguously reconstructed as the inclusive pronoun. The compound form *m̀-!a id. must also be of Proto-North Khoisan origin, although the meaning and function of the second morpheme are obscure. Of note is the phonetic similarity with the 1st p. sg. *mí, possibly, although not necessarily, indicating a common origin for the two.
[Bleek 1929: 91]. Simple -ba is also quoted in the same meaning, but [Bleek 1956: 13] gives ba as a general interrogative particle = Juǀ'hoan bȁ:ʰ id. The first component če probably = čí ~ či 'thing' [Bleek 1956: 229]. The bimorphemic stem, however, is not attested in [Bleek 1956]; instead, one finds čo (p. 232), perhaps, a dialectal contraction from *či-ba.
!Kung:
Unclear. [Bleek 1929: 91] lists the compound form ca-de, but its existence is not confirmed in [Bleek 1956] except for the second part dé, indicated as a general interrogative word on p. 23. Rather than trust the dubious evidence of [Bleek 1929], we prefer to leave the slot unfilled.
Bleek 1956: 138; Bleek 1929: 91. J. Snyman lists a different compound form: m-čí [Snyman 1980: 55], easily analyzable as the general interrogative morpheme m plus čí 'thing'. Bleek's -pai component is less easy to understand, but it is clear that the main interrogative meaning is placed on the m- element as well.
König & Heine 2008: 45. The form m̄-čí literally means 'what thing?'; -čá is a pronominal stem. The main interrogative meaning is expressed by the morpheme m̄.
Proto North Khoisan:*hà-čí
Distribution: Each dialect cluster has its own preferable interrogative marker, which makes reconstruction problematic. Nevertheless, analysis of comparative data in J. Snyman's dialect survey shows that occasional forms of interrogative pronouns with initial a- (< *ha-) can be found even in the most distinct Northern cluster. We choose *hà as the most representative equivalent, and reconstruct the compound form 'what?' as *hà-čí. Replacements: (a) ǀǀKxauǀǀen če-ba; the transformation mechanism is not quite clear, but it looks like the noun 'thing' has dropped the preceding interrogative marker in this language and then linked itself to a more general interrogative particle, transposing the interrogative meaning on itself; (b) Ekoka m̄-čá, etc. The provenance of the interrogative morpheme *m̄ is unclear. It is not excluded that it appeared under the influence of similar morphemes in Central Khoisan languages, but direct borrowing is out of the question; there must have been some internally driven reason for replacement as well. Reconstruction shape: The reconstruction is essentially based on the Juǀ'hoan form. The second component must have been the equivalent of Proto-Northern Khoisan *čí 'thing' = Juǀ'hoan čí, etc., although some dialects show an irregular development to če; this is the result of further contraction with the deictic stem *-e, i. e. *ha-či-e > *hače (cf. the same situation with 'who' q.v.).
König & Heine 2008: 81. Adjectival stem; the corresponding verbal stem is !àʔō [ibid.], with tonal gradation.
Proto North Khoisan:*!àʔū
Distribution: Preserved in all daughter dialects. Reconstruction shape: Correspondences are regular and trivial, except for the tonal pattern (Ekoka data shows traces of tonally marked morphological gradation, which could account for the irregularities).
Bleek 1956: 4. Lit. 'what person?' = Juǀ'hoan hȁžòè. [Bleek 1929: 91] lists an entirely different form: !ũ-de, where !ũ = 'man' q.v., and -de is a general interrogative morpheme. However, that form is not confirmed in any way in [Bleek 1956], whereas textual confirmation for a-ǯu 'who?' is quite plentiful.
König & Heine 2008: 45. Lit. 'what person?', with the same interrogative morpheme as in m̄-čá 'what?' q.v.
Proto North Khoisan:*hà-ǯù
Distribution: The situation here is exactly the same as with 'what' q.v. Replacements: See under 'what'. Reconstruction shape: Based on the Juǀ'hoan form. The exact lexeme is a compound with *ǯù 'person', and in some dialects is further expanded with the deictic stem *-e: *hà-ǯù-e > Juǀ'hoan ha-žoe, etc.
Bleek 1956: 31, 164, 263. Quoted as zau ~ ʒau in [Bleek 1929: 93]. The latter source also adds de as a synonym, but most of the textual examples as well as external comparison rather suggest the meaning 'female' for this word.
Bleek 1956: 213; Bleek 1929: 93. Secondary synonym: dama [Bleek 1956: 21; Bleek 1929: 93]. Textual examples in [Bleek 1956] show that dama and cau are either completely synonymous (indicating that dama is in the process of replacing the older word cau), or that cau may already be restricted to indicating the more specialized 'wife'. Additional research on existing texts is needed to resolve the issue. However, J. Snyman, for the meaning 'woman', only quotes čʰàò ~ cʰàò ~ šàò ~ sao [Snyman 1980: 55], not mentioning the existence of dama at all. A third synonym is !ũ-de [Bleek 1929: 93] = !xṹ-dẽ̀ [Snyman 1980: 55], but this form is easily understood as lit. '!Xung female', i. e. specifically 'a !Xung woman'.
König & Heine 2008: 20. Quoted as ʒʰāō (Western dialect) ~ sʰāō (Eastern dialect) in [Heikkinen 1986: 22]. Attested also as ǯʰāō-mà id., with a standard diminutive suffix. Secondary synonym: dȁh-mà 'woman, wife' [ibid.]. The semantic distinction between the two words is not quite clear, but text examples suggest that the latter is more closely associated with the specified meaning of 'wife' or 'married woman', whereas the former is still the more broad term.
Proto North Khoisan:*ʒʰau
Distribution: Preserved in all daughter dialects. Reconstruction shape: Correspondences are generally regular and trivial; Proto-North Khoisan *ʒʰ- automatically becomes preglottalized in Juǀ'hoan, so there is no need to carry it over as a phonological feature onto the proto-level. Tonal correspondences are not clear enough to allow the reconstruction of a distinct pattern. Semantics and structure: This is unambiguously the main North Khoisan word denoting a female human being, and should be distinguished as such from *de ~ *di 'female (in general)'. The occasionally encountered synonym da(h)ma is much more localized (could this be simply a "naturalized" reanalysis of German/Afrikaans Dame?).
Bleek 1956: 276; Bleek 1929: 94. Translated as 'yellow, pale blue, green' in [Bleek 1956] and most likely the same word as 'green' q.v., unless there is some confusion of similar-sounding roots involved (not highly likely given lack of confirming external evidence).
Distribution: Two different terms denoting the color of 'yellow' in Juǀ'hoan and Ekoka are transparent recent innovations, which, in the light of the rest of the data, means that in Proto-North Khoisan 'yellow' was not likely to have been different from 'green' q.v. Replacements: (a) Juǀ'hoan ɡǀǀȍˤnì-ɡ!ú {'jewel beetle belly' > 'yellow'); (b) Ekoka ɳ!ùū {'egg' > 'yellow'}.
Bleek 1956: 295, 362, 401, 423, 653, 678. Quoted as !xãː ~ ǀã́ in [Bleek 1929: 37]. The multiple variants are, in all likelihood, dialectal or misheard forms of the same root (the palatal click ǂ is very unstable in D. Bleek's transcription).
König & Heine 2008: 92. Quoted as ǂxã̄ in [Heikkinen 1986: 23]. Used in verbal ('to be far') and adverbial ('far away') functions. Cf. also !ʼà 'far' [König & Heine 2008: 86], only as a directive adverb.
Proto North Khoisan:*ǂxã̄
Distribution: Preserved in all daughter dialects. Reconstruction shape: Correspondences are regular and trivial. Fluctuation between various click types in old records is fairly typical of the palatal click, usually the hardest one to properly identify (at least for researchers in the first half of the XXth century).
König & Heine 2008: 69. Quoted as tȉ in [Heikkinen 1986: 21].
Proto North Khoisan:*tȉʰ
Distribution: Preserved in all daughter dialects. Reconstruction shape: Correspondences are regular. Both in Juǀ'hoan and Ekoka the word displays an ultra-low tonal pattern that frequently accompanies breathy vowel articulation; in this case, it seems that breathy articulation has to be reconstructed on protolanguage level as a distinctive phonological feature.
Bleek 1956: 27, 207, 241. Quoted as tɔmm in [Bleek 1929: 60]. The latter source also quotes ǀuː in the meaning 'to be near', confirmed in [Bleek 1956: 322], but this is a recent borrowing from Central Khoisan that evidently has not succeeded in replacing the original word.
König & Heine 2008: 70; Heikkinen 1986: 21. Verbal form ('be near'), also quoted as gè tōʔm-ā id.
Proto North Khoisan:*toʔm
Distribution: Preserved in all daughter dialects. Reconstruction shape: Correspondences are regular and trivial, except for the tonal pattern which requires further exploration. Frequently attested in the "junctive" form *toʔm-a (with following object).
Bleek 1929: 71. For some reason, not attested in [Bleek 1956]. In any case, clearly a borrowing from the same Central Khoisan source as the Juǀ'hoan item.
Bleek 1956: 53; Bleek 1929: 71. J. Snyman [Snyman 1980: 48] incorrectly translates 'salt' as ǂʼʰùì, even though the latter is a verb/adjective meaning '(to be) salty'; this is made clear in a textual example on the same page: gui ǂʼʰùì 'the salt is salty'.
Distribution: Preserved in most daughter dialects. Replacements: (possibly) ǀǀKxauǀǀen dabe, borrowed from Central Khoisan. J. Snyman records the variants dabi and dibi for a few more North Khoisan dialects, but the distribution is still scarce enough to be explainable as the result of recent areal diffusion. Reconstruction shape: Correspondences are regular and trivial.
Dickens 1994: 316. Singular subject action verb; the plural subject form is !òà-mʰí (second part of the compound stem in both cases is =mà / =mʰí 'small').
Bleek 1956: 446. In the meaning 'short' only the reduplicated stem is attested in [Bleek 1956], but in [Bleek 1929: 74] both !o!o and !oː are quoted with the meaning 'short'.
König & Heine 2008: 83. Functions both as adjectival and verbal stem; another derived verbal stem is !ò-mà 'be short' [ibid.] (quoted as !ō-mā in [Heikkinen 1986: 24]).
Proto North Khoisan:*!ò ~ *!ò-mà
Distribution: Preserved in all daughter dialects. Reconstruction shape: Correspondences are regular and trivial. The proper monosyllabic root is *!ò, but the word must have existed in the expanded form *!ò-mà, with a further diminutive suffix, already on the protolevel.
Dickens 1994: 304, 305. The second morpheme is the diminutive suffix =mà (cf. the plural form ǂʼã̀gà-mʰí). Of the two variants of the root morpheme, ǂʼàː is likely secondary (a dialectal contraction from ǂʼã̀gà-).
Bleek 1956: 380; Bleek 1929: 77. The latter source also quotes ɳǂìː as a synonym; in [Bleek 1956: 672] the word is, however, attested with the more specialized meaning 'great watersnake'.
!Kung:
Unclear. [Bleek 1929: 77] lists two items: (a) ɡǀǀéː, confirmed in [Bleek 1956: 530] in the specific meaning 'a short thick snake' (corresponds regularly to Juǀ'hoan ɡ!áí 'puff-adder'); (b) ɡǀǀaŋɡǀǀani-še, confirmed in [Bleek 1956: 527] as the plural form 'snakes', but no corresponding singular form is listed.
Distribution: Ekoka and Juǀ'hoan agree on the same item, which is therefore the best candidate for 'snake' in Proto-North Khoisan. Replacements: (a) ǀǀKxauǀǀen ɡ!eː, etc. = Juǀ'hoan ɡ!áí 'puff-adder' [Dickens 1994: 209], etc., implying a generalization {'puff-adder' > 'snake'}; (b) !O!Kung ǀã́ṹ probably is the same word as Juǀ'hoan ǀkxã̀ũ̀ 'blind snake (Typhlops sp.)' [Dickens 1994: 291], although the lack of velar affricate efflux is suspicious. It should also be noted that the exact semantics for these forms could have been indicated erroneously (i. e., 'snake' instead of 'a certain kind of snake'). Reconstruction shape: Correspondences for the first two segments are regular and trivial, but the stem on the whole is somewhat unique; its violation of the general phonotactic laws of North Khoisan (-g- or -w- in word-medial position) can be explained either as a result of compounding (i. e. *ǂʼã- + *-ga, although neither of the two morphemes is encountered separately) or borrowing (although the source is unknown). The Juǀ'hoan variant with -g-, even worse from a general phonotactic view than Ekoka's variant with -w-, must have been primary. Also, in Juǀ'hoan, the word is further expanded with the standard derivative suffix *-mà.
Bleek 1956: 265. Secondary synonym: ǀwi [Bleek 1929: 85], confirmed as ǀwí in [Bleek 1956: 334] = Juǀ'hoan ǀùʔí 'to be thin (of paper)', i. e. the semantics is probably somewhat more specialized than that of žaḿm.
König & Heine 2008: 103; Heikkinen 1986: 26. Meaning given as 'to be thin, lean'.
Proto North Khoisan:*ǯȁˤm
Distribution: The original form is attested in all major dialect clusters, replaced completely only in Ekoka. Replacements: Ekoka ǀǀkxàí = Juǀ'hoan ǀǀkxàì 'wrinkled' [Dickens 1994: 331], implying a metonymic shift {'wrinkled; contracted' > 'thin'}. Reconstruction shape: The reconstruction essentially just follows the form of the word in Juǀ'hoan.
Distribution: Proto-North Khoisan *ǂʼa has the widest dialectal distribution in this case, the other two roots being somewhat more localized. Replacements: (a) Juǀ'hoan màˤ, etc. probably reflects verb-to-noun conversion: {'to blow (of wind)' > 'wind'}, since in most other dialects this root, clearly reflecting Proto-North Khoisan *maˤ, is only attested in this verbal meaning (cf. in ǀǀKxauǀǀen: ǂʼa mà 'the wind blows' [Bleek 1956: 132]); (b) !O!Kung ǀǀuli, etc., is comparable with Juǀ'hoan ǀǀȍrò 'whirlwind' [Dickens 1994: 333], hence probably a semantic specialization: {'whirlwind' > 'wind'}. Reconstruction shape: Apart from old sources, the word is also attested in many dialects in J. Snyman's survey, always in the form ǂʼa; Bleek's transcription with a nasalized vowel in ǀǀKxauǀǀen must, therefore, be either an error or an occasional irregularity. Semantics and structure: The phonetic similarity with Proto-Central Khoisan *ǂʼã 'wind' is hardly coincidental, implying either cognation or borrowing. If this was a borrowing, however, it must have occurred already on the Proto-North Khoisan level.
Bleek 1956: 107; Bleek 1929: 94. The latter source also lists maːka, confirmed in [Bleek 1956: 133], as a synonym; this must be a very recent Bantu borrowing. J. Snyman lists an entirely different form: ǂʼʰáú [Snyman 1980: 56].
König & Heine 2008: 33. Same word as 'rain' q.v. Secondary synonym: ǂʰáú [Heikkinen 1986: 23].
Proto North Khoisan:*kuri #
Distribution: Although the form kuri is far more widespread in the meaning 'year' than any other one, it is also most likely of Central Khoisan origin (Proto-Central Khoisan *kuri id.); the only question is whether it had already penetrated into Proto-North Khoisan, or was borrowed independently into different dialects. There are, however, no serious alternate candidates for 'year' on the protolanguage level anyway. Replacements: (a) Grootfontein ɳ!̯ȁȕ, of unknown origin; (b) Ekoka ɡǀǀà = 'rain' q.v. Reconstruction shape: The reconstruction is based directly on Juǀ'hoan.