Werner 2002: II, 31; Werner 1993: 81. Neuter gender. Plural form: ˈɔgdɛ-n {огдэн} ~ ˈɔgdɛ-niŋ. Quoted as ɔgde5, pl. ɔgden6 in [Werner 1977: 171]; as ogdi in [Castrén 1858: 163].
Werner 2011: 234. Neuter gender. Plural form: ˈɔχtiŋ-ɨn ~ ˈɔχtɨŋ-ɨn. Quoted as ɔχtɨŋ6, ɔχtɨŋ-ɨn5 in [Werner 1977: 171]; as oqteŋ in [Castrén 1858: 163].
Common Ket-Yugh notes:
Most likely, from proto-KY *ʔɔɢdɛ. Although both words are unquestionably related, reconstruction of the protoform is difficult, and the difficulty only increases if the Ket-Yugh forms for 'ear-ring' are taken into consideration: Ket ˈɔqtaŋ, pl. ˈɔqtanɨŋ ([Werner 2002: II, 45]; quoted as ɔqtǝn6, pl. ɔqtɨn-ǝŋ5 in [Werner 1977: 171]), but Yugh ɔgdɔn ~ ˈɔgdiŋey, pl. ˈɔgdɔn-ɨŋ ~ ˈɔgdiŋ-ɨn [Werner 2011: 234]. It looks as if the correspondences are reversed: Ket has voiced -gd- in the word for 'ear', but voiceless uvular -qt- in the word for 'ear-ring', whereas in Yugh it is vice versa. A fully satisfactory resolution of this problem seems unattainable, but it should be noted that voiced / voiceless alternations in Ket/Yugh are usually caused by consonantal assimilations; in this case, voicing / devoicing of the cluster could have been triggered by voiced / voiceless consonants of the suffix (or second root) that later disappeared due to contraction. If the original stem was *ʔɔɢdɛ, then the old word for 'ear-ring' could have been *ʔɔɢdɛ-Cǝn (where -C- is a certain voiceless consonant), with different scenarios of contraction in Ket and Yugh. As for Yugh ˈɔχt-ɨŋ, it is a rather strange form for a singular noun, also indicating a possible compound origin; perhaps < *ʔɔɢd-fɨʔŋ, where fɨʔŋ = 'swelling, tumor' [Werner 2002: I, 348] (i. e. 'ear' as 'ear-swelling', 'outer ear' as opposed to 'inner ear', 'hearing')?
Kott:kaloːx-1
Castrén 1858: 204. Plural form: kaloːg-ˈan. Most likely, borrowed from a Turkic source (cf. Yakut kulgaːk, Tatar qolaq, etc.), although the "vowel metathesis" remains unexplained. Cf. in older sources: kalˈogan (M., Dict., Pal., Kl.), kolog (Kh.) [Verner 1990: 385].
Dulzon 1961: 187 (M., Dict., Kl.); quoted as utkʸen-ˈoŋ in (Pal.). The form is marked as plural (Lat. aures), so -oŋ is most likely the plural marker. Cf. also utkuy 'ear' in (Kh.) [Werner 2002: II, 372].
S. Starostin 1995: 198 (*ʔɔgde ~ *ʔɔqtV). Distribution: Preserved in all daughter languages except for Kott. Replacements: Kott kaloːx is most likely borrowed from a Turkic source (see notes on the Kott entry). Reconstruction shape: This issue is very problematic. H. Werner places under doubt the common etymologization of Ket-Yugh items, on one hand, and the Arin / Pumpokol forms, on the other. However, the only serious problem preventing such a comparison is the metathesized order of consonants, and it may be easily circumvented. Pace S. Starostin's reconstruction, it seems more reasonable to view the Arin / Pumpokol consonantal sequence as original, since (a) Arin and Pumpokol belong to different primary branches of Yeniseian, with metathesis in Proto-Ket-Yugh a more economic solution than independent metatheses in Arin and Pumpokol; (b) all other instances of *-KT- ~ *-QT-type sequences in Proto-Yeniseian generally yield assimilated -T(T)-type sequences in Arin. The somewhat tentative reconstruction *ʔɔdɢe, despite the uniqueness of its medial cluster, accounts for all the resulting diversity, except for stem-final -n in Arin and Pumpokol, which is probably just the old plural marker, sometimes fused with the root (cf. the variation between the old singular utkuy and the plural-turned-singular utqʸöː-n- in Arin). Semantics and structure: The unusual stem structure *ʔɔdɢe almost certainly implies a compound origin, but whatever it was, the fusion probably pre-dated the Proto-Yeniseian stage, and it is doubtful that this particular mystery will ever be resolved to general satisfaction.
Werner 2002: I, 110; Werner 1993: 22. Neuter gender. Plural form: baŋ-ɨn {баӈэн}. Polysemy: 'earth / land / place'. Quoted as baʔŋ2 in [Werner 1977: 137]; as baŋ, pl. baːŋ-an ~ baːŋ-en in [Castrén 1858: 188].
S. Starostin 1995: 205; Werner 2002: I, 110. Distribution: Preserved in all daughter languages. Reconstruction shape: Correspondences are generally regular, although the front row vocalism in Arin (peŋ) and Pumpokol (biŋ) remains unexplained. Based on some additional evidence, it is quite possible that this situation reflects a former bisyllabic structure (i. e. Proto-Yeniseian *baʔŋi or *baʔŋe), with deletion of the final vowel after a glottal stop in Ket-Yugh and root vowel assimilation in Arin and Pumpokol; however, this hypothesis requires further verification.
Werner 2002: II, 191; Werner 1993: 87. This form phonetically functions as the 1st/2nd/3rd m. p. sg. member of the intransitive verbal paradigm ('I eat', etc., without explicit marking of the object within the verbal form), but in reality, as seen in comparison with the rest of the paradigm as well as data from phonetically conservative dialects, is to be analyzed as *d=sʸiy-a / *k=sʸiy-a etc., where the first morpheme is the subject marker, and sʸi-...-a is a composite verbal stem (usually assimilated to si-...-e in actual conjugation). Cf. the paradigmatic data in [Werner 2002: I, 359-360]: t=sʸîy ~ t=sʸîːy-e ~ t=sʸîːy 'I eat', past tense t=sʸî-lʸ ~ t=sʸîː-lʸ-e ~ t=sʸîː-lʸ 'I ate'.
Werner 2011: 127. The paradigm of this formally intransitive (without explicit object marking within the paradigm) verb is structured exactly the same way as in Ket: 1sg. di=siʰːy ~ t=siʰːy 'I eat', past tense di=siʰː-r ~ t=siʰː-r 'I ate' (reduced from earlier *di=siʰː-y-a, *di=siʰː-r-a, etc., where *-a is the same verbal root as in *diˑ-p < *di=b=a 'I eat it', see below).
Castrén 1858: 217. Past tense: toːp-o-l-oːk-ŋ, imperative: toːp-aː-l-če-k. Composite verb; the simple root toːp has either an infinitive meaning ('to eat') or a substantive one ('food'). The older sources quote a different, not easily identifiable form: bag (M., Dict., Pal., Kl.) [Verner 1990: 309], where b= may be the 3rd p. inanimate object prefix, but the root remains unclear.
Dulzon 1961: 166 (Dict., Pal., Kl.). The forms dɨ=si-an (Kl.), dɨ=siy-an (Pal.) are, in all likelihood, really Yugh in origin.
Proto-Yeniseian:*siː-
S. Starostin 1995: 274 (*siɢ-; later amended to *siː- in [YED # 632]). Alternately reconstructed as *siʔǝgǝ- in [Werner 2002: I, 360]. Distribution: Preserved as a verbal root with its original basic meaning in Ket-Yugh and possibly in Arin and Pumpokol, but only in derived stems in Kott. Replacements: In Kott, the old verbal root 'to eat' is still preserved in the nominal derivate ši-g 'food, meal', but in the verbal paradigm, it has been replaced with a complex formation on the basis of the verbal root toːp- = Ket-Yugh *=dɔp 'to drink'. The semantic shift {'to drink' > 'to eat'} seems strange, but the situation may have been more complex (see notes on 'to drink' for further details). Reconstruction shape: The reconstruction siː- is essentially based on Ket-Yugh forms, as well as on Kott ši-g if the velar element is detached as an old nominal suffix. Attested Arin and Pumpokol forms are more difficult to interpret as far as their morphemic constituency is concerned; in addition, initial s- in Pumpokol sogo is irregular (normally, the standard development *s- > t- should be expected). Unfortunately, we know so little of Arin and Pumpokol verbal morphology that no satisfactory explanations can be foreseen; we have to rely on the initial consonant as the only significant piece of evidence for not suggesting lexical replacement in any of these languages. Semantics and structure: There is no way of telling if the complex, tripartite system of lexical roots used to express the meaning 'to eat' in Ket-Yugh, was an innovation or should be traced back to the Proto-Yeniseian stage. Currently, Ket-Yugh *siː- is the only root out of the three (*siː-, *=a, *=dɔq) that finds more or less reliable external parallels, and for that reason, we choose it as the default equivalent for 'to eat' in Proto-Yeniseian.
Werner 2002: I, 359; Werner 1993: 36. This root is encountered in the transitive verbal paradigm ('I eat smth.', with the normally inanimate object of eating marked with the usual morpheme -b- ~ -v- ~ -p-), where it is phonetically preserved only in archaic dialects. Cf. diˑ=p= ~ diˑ=b=a ~ diˑ=v=a {дип ~ дипь} 'I eat it', past tense d=b=iˑ=lʸ= ~ d=b=iˑ=lʸ=e {дбиль} 'I ate it'. It must be exactly the same root as in the intransitive verb sʸi-...-a, and, since it is encountered in both paradigms, should be considered the main basic equivalent for the meaning 'eat' in Ket. However, we still include the morpheme sʸi- as a lexicostatistical synonym, since it is also encountered in basic usage, and neither internal nor external data show it to have ever possessed a meaning different from 'eat' or 'food'.
Other peculiarities of the meaning 'eat' in Ket include: (1) the infinitive for both -a and sʸi-...-a is ˈilʸiŋ ~ ˈilʸeŋ [Werner 2002: I, 359]. While -iŋ is a well-known productive suffix participating in the formation of infinitives, the morpheme ilʸ- looks suppletive, unless it is actually the result of a (rare, but attested) incorporation of the past tense marker within the infinitive. If so, ilʸiŋ historically = *i-lʸ-a-iŋ, with the original root once again lost due to complex morphophonology. Cf. also the perfectly regular imperative form i=lʸ= 'eat!' (< *i=lʸ=a), which could itself serve as the basis for this infinitive formation;
(2) An entirely different paradigm is available for situations that require multiple animated participants (e. g. 'I eat him', 'she eats me', etc.): in this case, the usual stem is -dɔq [Werner 2002: I, 201]: dˈi=ɣ=a=rɔq 'I eat him', d=ˈiː=rɔq (< *di=ɣ=i=rɔq) 'I eat her', etc. Since the meaning 'to eat smbd. (people, animate creatures etc.)' is less usual and basic than the meaning 'to eat (food)', we do not see this paradigm as eligible for inclusion. One old source (Dulzon 1968) also mentions the existence of the variant -dɔ in addition to -dɔq, but this has not been confirmed.
Werner 2011: 127. As in Ket, this root is primarily encountered in the transitive verbal paradigm: 1 sg. diˑ=p= 'I eat it' (< *di=b=a), past tense d=i=b=iˑ=r 'I ate it' (< *di=b=i=r=a). The original root is still seen in the plural form: 1 pl. diˑ=b=a-n 'we eat it', past tense di=b=iˑ=r=a-n 'we ate it'. The infinitive form is ˈiriŋ = Ket ˈilʸiŋ (see notes on Ket). Also, as in Ket, the meaning 'to eat (smbd. rather than smth.)' is expressed by a separate verbal root, =dɔχ [Werner 2011: 127].
Common Ket-Yugh notes:
Proto-KY *=a 'to eat'. The overall system was somewhat suppletive, depending on the object of the process: (a) inanimate object marker *b was used in conjunction with the simple verb stem *=a (*di=b=a 'I eat it'); (b) if the object marker was not present at all, an extra slot was occupied by incorporating the lexical root *siː- (*di=siː-y-a 'I eat'); (c) if the object was animate, the root *=dɔq was used instead (*di=ɣ=a=dɔq 'I eat him').
Werner 2002: I, 256; Werner 1993: 131. Neuter gender. Plural form: eˑŋ ~ ˈɛŋǝn {еӈ}. Quoted as ɛʔy2, pl. eˑŋ1 in [Werner 1977: 148]; as eŋ ~ eːŋ, pl. eäŋ-en in [Castrén 1858: 236].
Werner 2011: 112. Neuter gender. Plural form: ˈɛŋ-ǝn ~ ɛŋ-ɨn. Quoted as eŋ1, pl. ɛŋ-ǝn5 in [Werner 1977: 148].
Common Ket-Yugh notes:
The form *eŋ 'egg(s)' is definitively traceable to Proto-KY. Subsequent reconstruction scenarios are as follows: (a) H. Werner [Werner 2002: I, 256] regards the Ket situation as original, presuming that in Yugh the old plural form *eŋ was reinterpreted as singular, and a new productive plural, eŋ-ǝn, was formed - quoting the case of 'horn' q.v. as an analogy; (b) S. Starostin [S. Starostin 1995: 232] suggests the opposite, namely, that *eŋ is the original form and that it is actually Ket ɛʔy that is a back formation. While the latter scheme seems somewhat counter-intuitive, it is indirectly supported by external data (Arin and Pumpokol equivalents for 'egg' have only *ŋ-containing forms), and the back formation ɛʔy < eˑŋ is theoretically imaginable, e. g. as an analogy with another bird-related term: kɛʔy 'wing', pl. keˑŋ [Werner 2002: I, 420]. On the other hand, considering that the word 'egg' is more frequently used in the plural, Ket might indeed be preserving an archaic sg. form as opposed to all the other members of the Yeniseian family. No definite conclusions may be drawn.
Kott:šuley2
Castrén 1858: 215. Plural form: šuley-aŋ. It has been suggested by Ye. Khelimskiy that the word is a borrowing from Samoyed (cf. Mator šlʸoy 'egg'), but the word has no Common Samoyed, let alone Uralic, etymology, so for the present moment it is not recommended to score it as an obvious loanword. Cf. in older sources: šulˈey (M., Dict., Kl.), šulep (Kh.) [Verner 1990: 394].
Dulzon 1961: 189 (Dict., Kl.). Presumably a plural form, where -aŋ is the plural suffix. The alternate form eg (Kl.) is really Yugh (= eŋ q.v.).
Proto-Yeniseian:*yeʔy
S. Starostin 1995: 232 (*yeŋ ~ *yɔŋ). Alternately reconstructed as sg. *eʔy, pl. *eŋ in [Werner 2002: I, 256]. Distribution: Preserved (often in a morphologically modified form) in most Yeniseian records, with the exception of Kott (probably). Replacements: Kott šuley has no clear etymology. It is either (a) a borrowing from Samoyed (see notes on the Kott entry); or (b) a compound formation: šul- '?' + -ey 'egg'. If it were possible to prove the latter case, there would be no need to postulate lexical replacement; but since the origins and meaning of the mysterious component šul remain unclear, such a replacement has to be postulated, at least "technically". Reconstruction shape: A very tricky situation here. In Ket-Yugh, apart from the general paradigm 'egg' (sg. *ɛʔy, pl. *e-ŋ), there is also a form ɔŋ, attested in the words for 'roe': Ket sg. ɔŋ-disʸ (where the second component = 'eye' q.v.), pl. ɔŋ-nʸiŋ. H. Werner regards 'egg' and 'roe' as two different etyma; S. Starostin suggests that the two may be etymologically linked if the vowel gradation reflects some sort of old "Ablaut", as in certain other nominal stems (e. g. 'dog', etc.). The opposition between 'egg' and 'roe' is also not as well pronounced in the rest of the Yeniseian languages as it is in Ket-Yugh (cf. Arin aŋ 'egg', but Kott dʸanan 'roe', Assan dialect anaŋ id.). One possible scenario is therefore to suggest this early paradigm: sg. *yeʔy (> Ket ɛʔy), pl. *yɔ-ŋ (> Ket-Yugh ɔŋ- 'roe'; Arin aŋ 'egg', with pl. > sg.) with various subsequent shifts: (a) in Yugh, plural eŋ has replaced singular ɛʔy; (b) in Arin, *yɔ-ŋ 'eggs' > aŋ 'egg', *yɔŋ-nǝŋ 'roe' > uy-nun; (c) in Pumpokol, *yɔŋ > *taŋ (regularly, *d- should rather be expected, cf. 'leaf' q.v., but a transcriptional mistake is not out of the question); the old plural form becomes singular, and a new plural is formed: *taŋ-aŋ > *tanʸ-aŋ with dissimilation. The vowel gradation between *yeʔy and *yɔ-ŋ would then date back to pre-Proto-Yeniseian (e. g. early *yɔʔy > *yɛʔy under the influence of two palatal glides?).
Werner 2002: I, 187; Werner 1993: 35. Neuter gender. Plural form: dɛ̂sʸ {дэсь} (dual) / dˈɛsʸ-taŋ {дэсьтаӈ} (plural). Quoted as deˑsʸ1, pl. dɛʰːsʸi4 (N.-Imb.) / dɛsʸ4 (S.-Imb.) /dual/, dɛsʸtǝŋ6 /plural/ in [Werner 1977: 143]; as des, pl. deɛs in [Castrén 1858: 181].
Werner 2011: 77. Neuter gender. Plural form: dɛʰːs ('pair of eyes'), but dˈɛs-aŋ ~ dˈɛs-ɨŋ ('many eyes'). Quoted as des1, pl. dɛʰːs4 ('pair'), dɛs-eŋ5 ('many') in [Werner 1977: 143].
Common Ket-Yugh notes:
Proto-KY *des, pl. *dɛːs (possibly < *des-ŋ with contraction and compensatory vowel lengthening; the plural form with the reinstated ending -(t)aŋ is a recent formation).
Castrén 1858: 217. Plural form: teːč-ag-an (partial suppletion, formed from an unattested complex sg. stem *teːč-ag/a/). Cf. in older sources: tečagan (M., Dict., Kl.), tieŋ (Pal.) [Verner 1990: 298] (all of these forms are plural, but the one in (Pal.) must be more archaic than the rest).
Dulzon 1961: 162 (M., Dict., Kl., Pal.). Probably a plural form ('eyes'). Cf. also tenkt 'eye' in (Kh.) [Werner 2002: I, 187] (probably = te-ŋ 'eyes' + an unidentified suffix).
S. Starostin 1995: 220. Alternately reconstructed as *detʸ ~ *des in [Werner 2002: I, 187]. Distribution: Preserved in all daughter languages. Reconstruction shape: Correspondences are quite regular (Pumpokol -t is a regular reflexation of Proto-Yeniseian *-s). Semantics and structure: In S. Starostin's reconstruction, final *-s is interpreted as a fossilized singulative suffix, a fuller variant of which may also be seen in *xu-sa 'one' q.v. and several other archaic nominal stems (e. g. 'stone' q.v.). This argumentation is solidly supported by Arin tie-ŋ, which probably preserves a trace of the archaic paradigm: sg. *de-s, pl. *de-ŋ (the latter form shifted to *des-ŋ in Proto-Ket-Yugh by analogy).
Werner 2002: I, 481; Werner 1993: 61. Neuter gender. Quoted as kɨʔt2 in [Werner 1977: 158]; as kɨt ~ kɨːt ~ kɨet, pl. kɨt-ɜŋ ~ kɨed-eŋ in [Castrén 1858: 168].
Not attested. Cf., however, ki '(it is) fat' in (Kh.) [Werner 2002: I, 481], which probably contains the same root.
Pumpokol:
Not attested.
Proto-Yeniseian:*gɨʔd
S. Starostin 1995: 228. Alternately reconstructed as *kɨʔʎ ~ *kɨʔt in [Werner 2002: I, 481]. Distribution: Preserved in all daughter languages where attested, but not found in Arin and Pumpokol. Reconstruction shape: Correspondences are fully regular. Semantics and structure: As in the attested language, the Proto-Yeniseian word was probably applicable to both hard 'fat' and liquid 'oil'.
Werner 2002: I, 66; Werner 1993: 19. Neuter gender. Plural form: ˈasʸ-eŋ {асеӈ}. Quoted as asʸ4 (S.-Imb.) / aːsʸ4 ~ asʸ4 (Bak., Sur.), aːsʸi4 ~ asʸ4 (Kur.), pl. asʸeŋ5 in [Werner 1977: 135]; as aːs, pl. aːs-eŋ in [Castrén 1858: 159].
Werner 2002: I, 66. Attested only in (Kh.); the form is transparently plural ('feathers').
Pumpokol:
Not attested.
Proto-Yeniseian:*ʔaːsi
S. Starostin 1995: 205 (*ʔVːsi). Alternately reconstructed as *aʔǝsǝ in [Werner 2002: I, 66]. Distribution: Preserved in all daughter languages (but not attested in Pumpokol). Reconstruction shape: Consonantal correspondences are regular and transparent. Vocalic correspondences are unclear (S. Starostin does not reconstruct the root vowel), but the data suggest that, most likely, the stem-final *-i has influenced the root vocalism in the Kott-Arin branch (*ʔaːsi > *ʔiːsi).
Werner 2002: I, 145; Werner 1993: 28. Neuter or feminine gender. Quoted as bɔʔk2 in [Werner 1977: 140]; as bok, pl. bog-eːŋ ~ bog-aːŋ in [Castrén 1858: 190].
S. Starostin 1995: 212. Distribution: Preserved in Ket-Yugh and Pumpokol. Replacements: In Kott and Arin (probably, in Proto-Kott-Arin), replaced by a nominalization of Proto-Yeniseian *qɔʔt 'to burn' q.v.; the shift {'to burn' > 'fire'} is typologically trivial. Reconstruction shape: Correspondences are regular and trivial except for the palatalization *-k > -č in Pumpokol (there are very few examples on final *-k in Pumpokol altogether); it is possible that this palatalization reflects traces of an original second syllable vowel, i. e. the reconstruction could be amended to *boʔke or *boʔki.
Werner 2002: I, 396; Werner 1993: 49. Feminine gender. Singular and plural forms are identical. Same word as 'meat' q.v.; in the meaning 'fish', the idiomatic expression ulʸ-d îˑsʸ (literally 'water's meat') may be used if necessary. Quoted as iˑsʸ1 'fish / meat' in [Werner 1977: 151]; as isʸ 'fish' in [Castrén 1858: 162].
Werner 2011: 134. Feminine gender. Singular and plural forms are identical. Same word as 'meat' q.v. Quoted as iˑsʸ1 'fish / meat' in [Werner 1977: 151].
Common Ket-Yugh notes:
Proto-KY *ʔiˑs 'fish / meat' (no distinction between sg. or pl. numbers).
Dulzon 1961: 180 (M., Dict., Pal., Kl.). Initial il= is most likely a fossilized prefix, same as in 'dog' q.v. Quoted as il=ta in (Kh.) [Werner 2002: II, 267].
Dulzon 1961: 180 (Dict.). Quoted as gˈite in (Pal., Kl.).
Proto-Yeniseian:*ciːk
S. Starostin 1995: 214. Alternately reconstructed as *tʸiʔǝgǝ in [Werner 2002: II, 267]. Distribution: Preserved in the original meaning in Kott-Arin and in Pumpokol. Replacements: In Ket-Yugh, replaced in the meaning 'fish' with *ʔiˑs 'meat' and only preserved in the meaning 'snake': Ket tîɣ, Yugh čiːʰk (see under 'snake'). The shift chain {'meat' > 'fish / meat'}, {'snake' > 'fish'} is, overall, the most economic solution, given the distribution of cognates in daughter languages. Reconstruction shape: The correspondence "Kott-Arin *t- : Pumpokol h- ~ x-" is regular and reflects Proto-Yeniseian *c- (see 'hair'), although the proper phonetic interpretation of this phoneme is questionable. Likewise, Pumpokol -t- is known to at least occasionally reflect Proto-Yeniseian *k (cf. *ʔok 'sterlet' > Pumpokol ot), so this correspondence is also regular.
Werner 2002: I, 200. In Ket proper, this verbal root participates in several paradigms with very close meanings: (a) simple verb, cf.: dˈiˑ=rɔq 'I fly', past tense d=ˈi=n=dɔq; (b) with directional preverb =t=, cf.: da=t=ˈa=y=dɔq 'she flies', past tense da=t=ˈɔ=lʸ=dɔq (possibly in the meaning 'to fly (around)' rather than 'to fly (somewhere)').
Possible secondary synonym: kɨˑ {кы} 'to fly (forth)' [Werner 2002: I, 483; Werner 1993: 60], as in: da=kˈɨ-(ɣ)-a-vet 'she flies', past tense da=kˈɨ-ɣ-ɔ-lʸ-bet.
Still another option is a verb for 'flying' that usually appears in the form -ɔk or -ɜk in several paradigms, such as: diˑ=ɣ=ɜk ~ diˑ=ɣ=ɔk 'I (will) fly (forth)' [Werner 2002: I, 201, 482]; cf. also, with the preverb =k=, da=k=aˑ=y=ɔk 'she flies forth', past tense da=k=ˈɔ=y=ɔk. (It is this paradigm that is probably surmised in Castrén's present tense koːigaq, past tense kolʸaŋoq 'to fly' [Castrén 1858: 168], although his final uvular is quite out of place). Overall, the situation here is very complex, possibly with several contaminated stems. For the most basic entry, we choose the form that appears to be the least marked in terms of specifying the direction of flight.
Werner 2011: 137. Infinitive form; quoted as dɔʰːχ4 in [Werner 1977: 145]. As in Ket, this verbal root has two different paradigms: (a) simple, cf. di=y=dɔχ 'I fly', past tense d=iˑ=r=dɔχ ~ d=i=n=dɔχ; (b) with directional preverb =č=, cf.: di=č=ˈa=y=dɔˑχ 'I fly', past tense di=č=ˈɔʰː=r=dɔʰːχ.
Possible secondary synonyms are also the same as in Ket: (a) kɨʰːχ 'to fly' in da=kɨʰːχ-aʰː-getʸ 'she flies', etc. [Werner 2011: 138]; (b) =gɜk 'to fly (forth, away)' in dˈi=g=a=gɜk 'I fly away', etc. [ibid.].
Common Ket-Yugh notes:
Proto-KY *=dɔːq 'to fly'; the root was used either by itself or within the complex KY framework *=č=...=dɔːq. Various directional aspects of flying, whose semantics is not well established based on existing evidence, could also be expressed by (a) the stem *=kɨːq= (only as the first component of bimorphemic verb stems); (b) the complex framework *=k=...=kǝk. Despite the phonetic similarity of these two roots, they cannot be reconciled as allomorphs; nor is it possible to relate any of them to *=dɔːq by assuming fossilized prefixation, at least, not on the KY chronological level.
Castrén 1858: 225. An idiomatic expression, consisting of an auxiliary verb and the nominalized stem f=a=ta-ga, which is the same as the first half of the verb f=a=ta-g-aːk-ŋ 'to stand up, rise' (directional preverb f= + root =ta-).
Arin:
Not attested.
Pumpokol:
Not attested.
Proto-Yeniseian:*=doːq #
S. Starostin 1995: 223. Alternately reconstructed as *doʔǝqǝ in [Werner 2002: I, 200]. Distribution: Preserved in Ket-Yugh, but not in Kott; not attested in either Arin or Pumpokol. Replacements: In Kott, the word may have been replaced by an idiomatic expression, derived from the verbal root 'to rise', provided that the semantic notation of M. Castrén ('to fly' and not 'to fly up, to soar', etc., was correct. The root *=doːq, nevertheless, is still attested in Kott in the meaning 'to jump': a=š=toːk-ŋ, past tense a=l=toːk-ŋ. Reconstruction shape: Correspondences between Ket and Yugh are regular and trivial. Semantics and structure: Based on the Kott parallels, the verb may have been polysemous in Proto-Yeniseian: 'to jump / to fly' (although the situation could also reflect a semantic shift from 'fly' to 'jump' in Kott). S. Starostin suggests analyzing the verbal stem *=doːq as composite, with a fused directional prefix, based on the occurrence of semantically similar verbs with different initial consonants, e. g. Kott i=tʰak-ŋ 'jump' (? < *t=ɔq-) and Ket-Yugh *k=ɨːq- 'to fly (forth)' (see notes on Ket-Yugh). However, there is no firm Yeniseian-internal evidence to justify this conclusion: all of these verbs might just as well represent different roots (note also the significant differences in vocalism between all the three). More systematic research on this issue is necessary to clarify the situation; for now, it is premature to confidently segment out *=d= as a separate morphological element.
Werner 2002: I, 153. Neuter gender. Plural form: bˈulʸ-aŋ {буляӈ}. Quoted as buˑlʸ1, pl. bulʸ-ǝŋ5 in [Werner 1977: 142]; as bul, pl. buol-eŋ in [Castrén 1858: 190]. Quite distinct from kiʔsʸ 'leg' [Werner 2002: I, 434].
Werner 2011: 145. Neuter gender. Plural form: bˈul-ɨŋ. Quoted as bul1, pl. bul-ɨŋ5 in [Werner 1977: 142]. Quite distinct from kiʔs 'leg' [Werner 2011: 85].
Common Ket-Yugh notes:
Proto-KY *bul, pl. *bul-uŋ 'foot'; opposed to Proto-KY *kiʔs 'leg'.
Castrén 1858: 222. Plural form: pul-aŋ. There is no indication of any lexical opposition between 'foot' and 'leg' in Castrén's materials. Cf. in older sources: pul (Kh.) [Verner 1990: 338].
Dulzon 1961: 174 (M., Dict., Pal., Kl.). Quoted as a=pil in (Kh.) [Werner 2002: I, 153]; status of initial a= is unclear - it may be a fossilized possessive prefix.
Dulzon 1961: 174 (Dict., Pal., Kl.). The form is most likely plural in origin ('feet'). Actually, the meaning 'feet' is somewhat dubious in the light of external evidence - 'legs' seems to would have been a more accurate semantic equivalent, but, since the original language of the dictionary is Latin ('pedes'), the semantics remains ambiguous.
Proto-Yeniseian:*bul
S. Starostin 1995: 213; Werner 2002: I, 153. Distribution: Preserved everywhere except for Pumpokol, but even there the attested "replacement" is somewhat dubious. Replacements: In Pumpokol, Proto-Yeniseian *bul 'foot' may have been replaced with an- = Arin an 'thigh', Kott aːn-ar 'thigh' id. [S. Starostin 1995: 181]. The semantic development 'thigh' > 'foot' is, however, more dubious than the development 'thigh' > 'leg', so there is a probability that Pumpokol an-iŋ really means 'legs', whereas the proper word for 'feet' was not recorded. Reconstruction shape: Correspondences are regular except for the enigmatic i instead of u in all Arin sources.
Werner 2002: II, 381-382. Predicative form: uˑt-sʸ. Plural form: ˈut-iŋ ~ ˈut-iŋ-sʸ-in. In [Werner 1993: 115], listed only in the derived form ˈut-al {утал}; it may be so that in Ket proper it is this derived stem that has a more frequent usage, unlike in Yugh. Quoted as uːt ~ ut-aːl ~ uːt-al 'voll, ganz' in [Castrén 1858: 165]. Secondary synonym: qɔ ([Werner 2002: II, 93; Werner 1993: 66]), more frequently used in idiomatic expressions (such as sʸiˑlʸe-qɔ "all summer", qip-qɔ "full moon", etc.).
Werner 2011: 326. Plural form: ut-iŋ. Predicative form: ˈut-si 'it is full'. Quoted as uˑt1, pl. ut-iŋ5 in [Werner 1977: 189]. As in Ket, also attested as a derivative form with an additional suffix: ˈut-al 'full' [Werner 2011: 326]. Secondary synonyms also include (a) χo ~ χɔ 'whole, full, complete' (usually in idiomatic expressions such as χep-χo "full moon", etc.); (b) pˈɔlɔn- in pˈɔlɔn-aʰː-b-an 'it becomes full' (borrowed from Russian полный 'full').
Castrén 1858: 204. Cf. also uːt-am 'all, whole' [ibid.].
Arin:
Not attested.
Pumpokol:
Not attested.
Proto-Yeniseian:*ʔute
S. Starostin 1995: 201. Distribution: Preserved in all daughter languages where attested, but not found in Arin or Pumpokol. Reconstruction shape: Correspondences are regular and generally trivial (the vowel in the second syllable could probably have been *-i as well).
Number:33
Word:give
Ket:n=...=o1
Werner 2002: II, 23. A complicated verb, used only with the rare preverb n= and with the obligatory inanimate object marker =b=. Specific forms iclude d=ˈa=nʸ=b=o 'he gives (to) him', past tense d=ˈa=nʸ=b=i=lʸ=o, d=ˈi=nʸ=b=o 'he gives (to) her', past tense d=ˈi=nʸ=b=i=lʸ=o, etc.
Yugh:n=...=o ~ n=...=u1
Werner 2011: 148. See notes on Ket. Specific forms include d=ˈa=n=b=u 'he gives (to him)', past tense d=ˈa=n=b=ɨ=r=u, d=ˈi=n=b=u 'he gives (to) her', past tense d=ˈi=n=b=ɨ=r=u, etc.
Castrén 1858: 130, 210. 1st p. sg.; cf. also past tense hi=l=i=peːn-aŋ, imperative hi=lʸ=en. This complex verb consists of the root =peːn- (although the fact that p disappears in the imperative could hint at its original affixal status) and an unclear prefixal component hi= or h=. The strange form farak 'I give' (Kh.) in one of the older sources is unclear [Verner 1990: 302].
Werner 2002: I, 438. Attested only in (Kh.). The first two syllables are unclear (separate lexical stem in a compound verb? a string of adverbial prefixes and conjugation markers?), but =penʸ- is segmentable as a root and comparable with Kott =peːn- q.v.
Pumpokol:
Not attested.
Proto-Yeniseian:*n=...=o #
No previously postulated reconstructions. Distribution: Only attested in Ket-Yugh. Replacements: There are altogether three different roots / stems attested with the meaning 'to give' in various Yeniseian languages. Of these: (a) Kott-Arin *=pen- (always functions as the second root in a composite stem, with differing first elements) is compared by S. Starostin in [YED # 962] with Yugh =fɨn in the composite verbal stem χˈɜdʸiŋ=fɨn ~ χˈɜdʸiŋ=fan 'to give back; to give away' (also attested as χˈɜdʸiŋ=fɨtʸ with unclear consonantal mutation). However, external Ket evidence shows that it is Yugh χˈɜdʸiŋ-, not =fɨn, that carries the main lexical meaning of 'give back, give away' (= Ket qˈɜr-am ~ qˈɜr-aŋ id. [Werner 2002: II, 146]). Considering that in Kott and Arin, =pen- is also not found on its own, it is more likely that the verb was a general "directional" auxiliary in Proto-Yeniseian rather than an original 'to give'; (b) Ket-Yugh =aq, likewise, is a verbal root with much broader semantics than 'to give'. Consequently, the only verbal stem that is attested exclusively in the meaning 'to give' is Ket-Yugh *n=...=o. Furthermore, its highly unusual shape (monovocalic root + very rare directional prefix) is an additional indirect hint at archaicity. We very tentatively set it up as the optimal candidate for the Swadesh meaning 'to give' in Proto-Yeniseian. Reconstruction shape: Since the stem n=...=o is only attested in Ket-Yugh, the Proto-Yeniseian equivalent could be slightly different, but at least it is not highly likely that the monovocalic root could have contained additional consonants (in this position, most of them would have yielded traces in modern dialects).
Werner 2002: 1, 49; Werner 1993: 32. Cf. specific forms: d=ˈa=v=aq 'he gives (to) him', past tense d=ˈɔ=v=iy=aq; d=ˈi=v=aq 'he gives (to) her', past tense d=ˈi=v=iy=aq. The same forms are quoted as d=a=b=aːq, past tense d=oː=b=e=d=iːy=aq in [Castrén 1858: 180]. Like n=...=o, this verb is also always used with the inanimate object marker =b=, although, when the indirect object is expressed by the 1st or 2nd p., the object marker inexplicably migrates to the position preceding the indirect object (as in: d=b=ˈi=ɣ=oq 'he gives to you', past tense d=b=ˈu=ɣ=aq; d=b=ˈi=r=aq 'he gives to me', past tense d=b=ˈu=r=aq). This may only be explained through a suppletive paradigm, where forms with the 3rd p. obj. follow a different model from those with the 1st/2nd p. object. In these forms, =b(i)= should be judged as an independent root ("modifier"), derived from biʔ 'thing', rather than a grammatical marker. Hence, the complete paradigm would be aq / bi-...-aq.
The semantic difference between n=...=o and (bi-...)-aq is not entirely clear, although the latter verb is usually suspected of denoting momentary action, contrasting with the "prolonged" (habitual) semantics of the former. The issue is not entirely resolved, and in any case, even if this is correct, we have to include both verbs as synonyms.
Werner 2011: 147. Cf. specific forms: d=a=b=aʰːχ 'he gives (to) him', past tense d=o=u=m=nʸ=aʰːχ; d=i=b=aʰːχ 'he gives (to) her', past tense d=iː=m=nʸ=aʰːχ. The same irregular inversion of the object marker =b= as in Ket happens here as well: d=b=i=g=aʰːχ 'he gives (to) you', etc. The verb is also encountered in compound formations, e. g. with =us= 'away': d=ˈus=a=y=aʰːχ 'he gives him' (literally 'gives away to him'), etc.
Common Ket-Yugh notes:
Proto-KY *ʔaːq 'to give'. This root actually has a whole variety of meanings in KY, depending on the shape of the paradigm (presence or absence of verbal prefixes, modifying first stems, etc.), including 'to leave', 'to drive', etc.
Werner 2002: I, 54; Werner 1993: 14. Predicative form; cf. ˈaqta kɛʔt 'good man' without the suffix. Quoted as aːqta ~ aqta 'good, fresh' in [Castrén 1858: 157]. Cf. aqtǝ-m6 'good' (adverbial form) in [Werner 1977: 183].
Castrén 1858: 207. The root is hag-; -ši is an infrequent adjectival suffix. Synonymous form: hamaː 'good' [Castrén 1858: 209], cf. also haŋo 'better'. It is not clear if the two words are related through some morphological model, or belong to different etymologies. Cf. in older sources: xamagit 'good' (Kh.) [Verner 1990: 386] (probably xama-xit 'good person').
Werner 2002: I, 52. Attested only in (Kh.), as part of the compound bergar=ɨktu, probably = 'very good' (where berga-r- = 'big' q.v.).
Pumpokol:
Not attested properly. The recorded forms ekte-ket '(it is) good' [Dulzon 1961: 170], ekte-m 'good' (German gut) [Dulzon 1961: 187] are only present in (Pal.), where Yugh forms are frequently recorded as "Pumpokol", and these particular ones definitely look Yugh rather than "proper" Pumpokol.
Proto-Yeniseian:*haq-
S. Starostin 1995: 230. Alternately reconstructed as *aqtə in [Werner 2002: I, 54]. Distribution: Preserved in all daughter languages (but not attested in Pumpokol). Reconstruction shape: Proto-Ket-Yugh *aqta is segmentable into the original root morpheme *aq- + the same suffix as in im-da 'small' q.v. (with phonetic assimilation). This allows for an easy and direct comparison with Kott hag-, where all the correspondences are regular. If Arin =ɨk-tu belongs here as well, -tu is the predicative suffix, i. e. Arin would be preserving the original root without any further derivational suffixes, unlike Ket and Kott.
Werner 2002: II, 221. With the predicative suffix -sʸ; cf. also sʸˈɜnʸ-iy 'to become green' [ibid.]. Meaning glossed as 'blue / green / brown' (although 'brown' is very suspicious: it is only confirmed by North Ket sʸɜʔnʸ qɔːye 'brown bear', quoted by Werner without justification; note also the glottal stop, lacking in other examples - the expression should be verified). Quoted as sʸon 'blue' in [Castrén 1858: 188]; this source attributes the meaning 'green' to xaɢal-eŋ 'blue / green', derived from xaːɢal 'gall' [Castrén 1858: 171], but this is a Yugh form, not Ket proper, and the gloss is not confirmed in more recent sources.
Yugh:silʸˈɔna-1
Werner 2011: 161. Transparent borrowing from Russian зеленый 'green'. Werner also lists the form xaʁal-eŋ (= xaɢal-eŋ) from [Castrén 1858: 171], but it is not attested in his own Yugh records (see notes on Ket for etymology).
Common Ket-Yugh notes:
There must have been no direct equivalent for 'green' in Proto-KY: the entire spectrum from 'blue' to 'green' was probably covered by the root *sǝn- (not analyzable as a borrowing from Russian синий 'blue' because of differences in vocalism), lost in Yugh where this part of the spectrum got replaced by Russian borrowings (silʸˈɔna 'green' and also sinʸa 'blue' < синий [Werner 2011: 95]).
Kott:
Not attested. Cf., however, the form šapkan 'it is green' (Russ. зелено) (Kh.) in one of the older sources [Verner 1990: 314].
Dulzon 1961: 168, 182 (Dict.). Polysemy: 'green / blue'. Quoted as kˈomul-zi in (Pal.). Dubious (the form could actually be Ket-Yugh rather than Pumpokol; cf. Adelung's kɨalmas 'green' for Ket [Dulzon 1961: 168]). Final -si is a standard Yeniseian adjectival suffix.
Proto-Yeniseian:
Not reconstructible. The item is poorly attested in all extinct languages; not a single isogloss between two different languages can be detected; and there are reasons to assume that the meaning 'green' was not lexically distinct even in Proto-Ket-Yugh.
Werner 2002: II, 304; Werner 1993: 105. Plural (collective) form only. Quoted as tǝˑŋǝ1 (N.-Imb.) / tǝˑŋ1 (S.-Imb.) in [Werner 1977: 184]; as tʸɜːŋ ~ teɛŋ ~ tɨːŋ, pl. tʸɜːŋ-en in [Castrén 1858: 178].
Werner 2011: 161. Quoted as čǝˑŋ1 in [Werner 1977: 184].
Common Ket-Yugh notes:
Proto-KY *čǝˑŋ. H. Werner's attempt [Werner 2002: II, 304] to regard the word as a compound (< Ket tɨʔ, Yugh čɨʔ 'head' q.v. + Ket iʔŋ 'skin' q.v.) runs into significant phonetic problems and, furthermore, is quite unprecedented according to semantic typology ('skin' and 'head hair' do not develop into one another; the pair would at least have to look like 'hide, fur' and 'body hair' - see notes on 'skin' q.v.); the word has to be taken on its own, as an unsegmentable entity.
Castrén 1858: 209. Plural form: heŋay-aŋ. Synonym: eːk 'hair' [Castrén 1858: 200]. Although the semantic difference is not indicated in Castrén's dictionary, external comparanda make it very probable that heŋai meant 'head hair', whereas eːk would rather mean 'body hair', 'fur', etc. The suffix -ay is a frequent body part suffix in Kott (cf. 'head', etc.). Cf. in older sources: hˈiŋay-an (M., Kl., Pal., Dict.), inagay (Kh.) [Verner 1990: 294].
Dulzon 1961: 161 (Dict., Pal.). Quoted as qɨŋa in (Kl.). Distinct from tümen 'hair (on body)' [ibid.].
Proto-Yeniseian:*cǝŋe
S. Starostin 1995: 213. Alternately reconstructed as *tʸǝŋǝ in [Werner 2002: II, 304] and analyzed as a compound form ('head' q.v. + 'skin' q.v.). Distribution: Preserved in all daughter languages. Reconstruction shape: The correspondence "Ket-Yugh *č- : Kott h- : Arin q- or k- : Pumpokol x-" is recurrent and, in S. Starostin's system, is interpreted as reflecting Proto-Yeniseian *c-. Such an interpretation is open to debate - in particular, due to the typologically unlikely velarization of affricates in Arin and Pumpokol, and also because of the comparable rarity of Proto-Yeniseian *c-, which would rather imply some non-trivial cluster or contextual development (cf. especially the transcription qʸ- in Arin), but there is little doubt that all the listed items do indeed belong together.
Werner 2002: I, 339; Werner 1993: 124. Neuter gender. Plural form: hˈɜŋnʸ-eŋ {хъӈнеӈ}. Quoted as hɜŋn5, pl. hɜŋn-eŋ5 in [Werner 1977: 193]; as hɜɛŋnʸeŋ in [Castrén 1858: 174] (where this word, quite clearly plural in form, is still translated as 'Hand (von der Wurzel an)'). The stem also forms part of a large number of compound forms, e. g. hɜŋn.. teŋ 'to touch', hɜŋn-tɛt 'to move (with the hand)', hɜŋn-anʸ-sʸ 'handless', etc.
Werner 2011: 164. Neuter gender. Plural form: bˈiŋ-ɨn. Quoted as bieŋ, pl. bieŋ-en in [Castrén 1858: 189].
Common Ket-Yugh notes:
Proto-KY is not reconstructible. The meaning 'hand' is notoriously unstable in Yeniseian, particularly in Ket-Yugh: none of the three attested equivalents (two in Ket and one in Yugh) have reliable parallels in the opposite dialect. It might be noted that Yugh biʔŋ 'hand' is completely homonymous with biʔŋ 'things (pl.)' [Werner 2002: I, 134], but there is no plausible semantic scenario here that could make us regard this as anything other than mere coincidence.
Dulzon 1961: 180 (M., Dict., Pal., Kl.). Quoted as upega in (Kh.) [Werner 2002: I, 338]; u= may be an obscure prefix, as in u=pusɨr (Kh.) 'forehead', u=kuriy (Kh.) 'throat' etc.
Dulzon 1961: 180 (Dict., Pal., Kl.). Glossed as both 'hand' (manus) and 'arm' (brachium). This actually looks like a reasonable plural form from tok (Dict., Pal., Kl.) 'finger' [Dulzon 1961: 176], so the accuracy of the semantics 'hand' is somewhat questionable.
Proto-Yeniseian:
Not reconstructible. The meaning 'hand' is notoriously unstable in Yeniseian languages: almost every language has its own etymological equivalent (sometimes two!), and most of the etymological connections are problematic. In the order of increasing probability of representing the Proto-Yeniseian equivalent for 'hand', here is what may be said about the individual candidates:
(1) Pumpokol to-n is, in all likelihood, historically a plural from tok 'finger' = Ket tɜʔq, etc. < Proto-Yeniseian *tǝʔq 'finger' [S. Starostin 1995: 283], a concept that turns out to have been much more stable in Yeniseian than 'hand';
(2) Yugh biʔŋ 'hand', despite phonetic similarity with Ket hɜŋn, does not regularly correspond to this word, and has no other parallels whatsoever;
(3) The only etymological parallel for Ket lʸaʔŋ 'hand' is in Arin: lan-tʸuːŋ (M., Kl.), lʸan-puy (Kh.) 'wing' [YED # 936], and it is not very clear (e. g. both forms are clearly compounds, but their second halves remain unetymologized); nevertheless, the semantic shift 'hand' > 'wing' is theoretically possible;
(4) Kott keːgär is compared in [YED # 341] with Yugh kɜgdʸ-at 'arm', and also, possibly, with Arin koro-nun (Kh.) 'mittens'. The Arin parallel is dubious, but the Yugh form is a good match that allows to reconstruct Proto-Yeniseian *kǝgdi-ʔaʔd, where *ʔaʔd = 'bone' q.v., possibly with the meaning 'arm' or 'armbone', but less probably 'hand' (it must be noted that Kott, apparently, does not distinguish between 'hand' and 'arm');
(5) The best "chances" lie with the pairing of Ket hɜŋn and Arin pʸʰyaga (= pega), which allows S. Starostin to reconstruct the protoform as *pVg- [S. Starostin 1995: 254]. The semantic matching is exact, and the correspondences are generally reconcilable. However, there is some doubt as to whether the Ket word is indeed the primary equivalent for 'hand' (see notes on Ket), and, subsequently, this would influence Proto-Yeniseian semantics.
Given this very complicated situation, we currently prefer to leave the Proto-Yeniseian spot empty, even if the Ket and Arin forms may, at least technically, be counted as lexicostatistical matches (without any serious semantic or phonetic obstacles).
Werner 2002: II, 9. More frequently encountered as lʸˈaŋ-at (< lʸaʔŋ + aʔt 'bone' q.v.). In [Werner 1993: 78] only this compound form lʸaŋat {ляӈат} is translated as 'hand' ('рука'); lʸaʔŋ {ляʼӈ} is glossed only as 'part of hand close to the wrist' ('основание кисти'). In [Castrén 1858: 175], the form lʸaːŋ-at is translated as 'Arm mit der Hand'.
The exact difference between hɜŋn and lʸaʔŋ(-at) is hard to define. On one hand, dictionary glosses and derivatives speak in favor of a basic distinction between hɜŋn as 'hand (with fingers)' and lʸaʔŋ(-at) as 'hand + arm'. However, lʸaʔŋ is not 'arm' per se (in this meaning Ket uses the word ɨˑlʸ [Werner 2002: II, 434]). Also, published texts (such as collected by Kreynovich, Dulzon, and others) seem to regularly feature the word lʸaŋat in contexts like "take with the hand", etc., whereas hɜŋn is encountered quite rarely, if ever. For the time being, we propose treating the two words as synonymous; however, it seems that hɜŋn is the more "marked" member of the opposition, referring specifically to 'hand' as 'the fingers and the part of the hand closest to the fingers', used for 'touching', 'groping', etc., rather than 'taking'. If this explanation is proven to be correct, then hɜŋn has to be eliminated from the wordlist.
Build tree