Yakhai

Timothy Usher, Santa Fe Institute

[under construction]

[under construction]

Nevermann (1940: 191-195) vocabulary of Enemur Sohur (North Yakhai)

Drabbe (1954: 73-98) grammar and (pp. 128-142) 428 comparative terms for Jaqàj (North Yakhai)

Kainakainu, Paidi, Rinantanti and Morin (1998) phonology and (pp. 97-184) large comparative vocabulary of Yakhai (South Yakhai)

Kanakaimu (2000) Yakhai verbs (unobtained)

Kriens and Lebold (2010: 40-53) 239 comparative terms for Yaqay (North Yakhai) of Kogo, Kotup, Rayam, Wairu and Wanggate villages

Olsson (n.d.) vocabulary of Nambioman-Bapai (South Yakhai) (unobtained)

[under construction]

Drabbe (1954: 73-75) gives 15 consonants and 7 vowels for Jaqàj (North Yakhai) of the Mapi and Qobàmarao rivers region as follows:

In addition to the simple vowels, at least one diphthong is found:

The most conspicuous phonological differences between North Yakhai and South Yakhai as exemplified by the attestations of Kainakainu, Paidi, Rinantanti and Morin (1998: 97-138) are South Yakhai's frication of bilabial voiceless stop /p/ to labiodental /f/ and the frication and deocclusion of /*s/, which is defricated to /t/ in North Yakhai, to /h/ (Drabbe 1954: 73, 74.)

North Yakhai's defrication of /*s/ to /t/ follows the lenition of apical voiceless stop /*t/ to /r/ in both dialects, although both are sporadically attested with their original qualities (q.v. Nevermann (1940: 191-195.)

While the overwhelming realization of Yakhai-Warkay initial apical voiceless stop /*t/ is non-stop [r] in both dialects, there are sporadic reflexes in which it is retained as [t] … /*-tak/ “walk” … …Kainakainu, Paidi, Rinantanti and Morin give only ten examples of words with initial /t/, suggesting it to exist only in loans. … Besides the regular loss of Yakhai-Warkay final /*ɾ/, which is exceptionless in Yakhai,, the only known example of outcomparison involving initial /*ɾ/ is Drabbe's Jaqàj [rod] “fence” (“omheining”, p. 135,) which may be a loan from Marind /*ɾod(-a)/. …However, /*t *ɾ/ remain contrastive medially, where either can occur between any vowels.

A small number of roots are found with initial [h] in South Yakhai, suggesting /*s/, but initial [r] in North Yakhai, suggesting a phoneme which might be reconstructed as /*ts/; e.g. /*tsep[a/o]/ “kunai grass”, /*ts[a/o]p/ “fence.” Another example, Drabbe's [muku-rap]“skull, corresponds to /haf/ in South Yakhai, but is probably the same root as [tap] “coconut shell” < /*sap/. No outcomparisons for these roots are known and the origin of this phoneme, if valid, is a mystery.

Drabbe's (1954: 2) Jaqàj <q>, given here as velar non-stop /ɣ/, is characterized as a fricative with the quality like a trill, and is found only when followed by low central and low mid back vowels /a o/, where it contrasts with unvoiced stop /k/.

Despite Drabbe's inclusion of velar voiced stop /g/ as a phoneme, his attestations of [g] are allophones of /ŋg/ ([gem] “zand” p. 136) or in error for <q>, /ɣ/ ([kajagaˈmaerɛ] “twee” p. 142.)

Kainakainu, Paidi, Rinantanti and Morin (1998) give only one example of Yakhai initial voiced [ɣ] ([ɣadin] 'memintal tali', p. 112,) possibly in error. Generally, only voiceless [x] is found initially while only [ɣ] is found medially.

Kainakainu, Paidi, Rinantanti and Morin (1998) give only one example of Yakhai initial [x] followed by /i/ ([xirdiʔ] “putus”, p. 129) and three examples of initial [k] followed by /ɤ/ (pp. 110, 112, 117. Otherwise [x] is found only before /a ɤ o/ and never before /i e u/, leaving initials /k x/ contrastive primarily when followed by /a/.

Root-medially, [ɤ] occurs both following and followed by any vowel, with [k] attested only marginally in words that may be loans ([akamin wiri] “kuil (rumah pemujaan)” p. 107, [ika] 'pinang' p. 111, [deːka ~ deka] “kasbi (ubi pohon)” p. 113, [kakef] “bubu” p. 117) or possibly in error ([mbakɤ] 'tua' p. 106.)

Both velars /k ɣ/ are neutralized to glottal stop [ʔ] word-finally and syllable-finally in clusters.

There is no voiced velar stop [g] except as a component of prenasalized /ŋg/, which is found in all positions.

The situation as it is described above leads to the question of whether what is given here as velar non-stop /*ɣ/ shouldn't be reconstructed as fricative /*x/ with voiced [ɣ] as its medial allophone alongside initial [x]. The reason …

…

It is worth noting that the synchronic situation in the Awyu languages is quite similar in these respects, with velar fricative /*x/ appearing and thus contrasting with voiceless stop /*k/ only when followed by vowels /*a *ɔ *o̝/, although the history behind this arrangment is quite different: besides the neutralization of Central Digul River rounded velar /*kʷ/ to /*k/, the contrast between /*k *x/ is found only in suspected loans especially from Marind-Yakhai, with all examples of Central Digul River /*k/ before the aforementioned vowels having been fricated. Nevertheless, it is difficult to avoid the impression that these two systems have converged through mutual influence towards a common system.

It's not clear that bilabial non-stop [w] should be reconstructed as a phoneme rather than the first member of a vowel sequence /uV/,Marind-Yakhai /*w/ having been dropped altogether in Yakhai and Warkay alike.

Palatal non-stop /j/ is needed because it is retained from proto-Marind-Yakhai and contrasts with sequential vowels /iV/ as well as with final /i/ (below.)

…

Based upon these correspondences and those to proto-Marind-Yakhai values, 14 or 15 consonants are reconstructed for Proto-Yakhai as follows:

Most likely under the influence of the western dialects of Marind, the dialects of Yakhai have been developing a new series of occluded rounded velar and palatal non-stops. These do not arise where original palatal non-stop /*j/, which is never occluded, is present, while Marind-Yakhai /*w/ has been outright dropped in all environments: the new occlusives arise where high vowels /*i *u/ combine with another vowel in a sequence or where they are word-final. It is this emergent contrast between segements /*iV *i#/ on the one hand and /*jV *j#/ on the other which mandate the reconstruction of /*j/ in proto-Yakhai.

…:

…:

…:

…:

…:

…:

…:

…

…:

…:

… /*ae#/ …:

… /*aeC/ …:

… /*aɣuC/ … [aɣaᵘC] … :

…

[under construction]

Yakhai free pronouns are reconstructed as follows, with Jaqàj (North Yakhai) attestations drawn from Drabbe (1954: 80) and Yakhai (South Yakhai) from Kainakainu, Paidi, Rinantanti and Morin (1998: 132):

The originality of third person /*anVp/ is demonstrated by comparison to Marind. Drabbe's Jaqàj medial [r] in these forms is irregular and unexplained; the root is found with expected [n] in the compound [anVp-ˈadə] “veel” (p. 142.) Kainakainu's [anif] masculine is given only in the possessive formation [anif-iːn] where it is homophonouns with the plural and likely in error.

… demonstratives … /*kV-de/ and /*kVp(-de)/ … Kainakainu, Paidi, Rinantanti and Morin (1998: 133) gloss /xɤ-de/ as “this/here” (“ini,” “sini”) and /xɤp-de/ as “that/there” (“itu”, “situ”) …:

Possessives are indicated by the suffix [-ndiːn ~ -iːn], the latter allophone occurring following root-final /*p/ [f]. …

…

[under construction]

… preposed possessors on kin terms (Drabbe 1954: 80-82) Here first persons are not distinguished by number. In a number of first person forms, an invariable root is found which we assume to originate as address forms; these are indicated in brackets below. Unlike gender-inflected free third person pronouns (above) and nominals (below,) third person possessors are not differentiated to indicate masculine, feminine or neuter gender but only by number. The forms of the third person singular and plural are often confused, suggesting this distinction to be in the process of collapsing:

Two roots of the form /*-uC(u)/ induce the backing and rounding of preceding /a/ in the possessor to [o]:

[under construction]

… (Drabbe 1954: 77, 77,-79 also some kin terms pp. 80-82.) In some instances an underlying vowel may be discerned in some forms which is leveled with the gender vowel in others; these are indicated by below brackets in the form of the root:

Two unexplained exceptions to the use of /e/ to indicate the masculine is found in the terms “older sibling” and “younger sibling”; here we assume /a/ and [œ], found as [a] in the attestations of Kainakainu, Paidi, Rinantanti and Morin (1998: 105) and Kriens and Lebold (2010: 41,) to be the underlying vowels of the root. No neuter or plural forms are given for these roots:

[under construction]

… Drabbe (1954: 85-94) …

… tense …

… subjects …

… objects … Unlike third person subjects (above,) third person objects are not distinguished by gender but only by number, with the neuter vowel /o/ standing for both masculine and feminine singular referents:

…

… /-ar-/ [-ar- -r- -ø-] … in the feminine singular as [-ɛr- -ø-] … present … (pp. 86-88) … third person singular object unexpectedly given as [-e-] rather than [-o-] …:

… /k-/ past … (pp. 87-88):

… /b-/ past … (pp. 87-88):

… /na-/ potential … (p. 90) …

… /na-/ … prohibitive … (p. 92) ….

… /nana-/ [nana- nan-] future … (p. 89) The genders of third person singular subjects are not distinguished…

The morphologically simplest construction, with no prefix /ø-/ but only a subject, is found in a marginal semantic environment of the present interrogative (p. 91.) Only a small sample of possible object forms is attested:

… /t-/ past interrogative … (p. 91) …

… /ka-/ imperative (pp. 91-92) …

…

…

…

…

[under construction]

…

…from Awyu into Yakhai … none have known cognates in Marind or other subgroups of Fly River:

… from Yakhai or Marind to Awyu … neither has known cognates in Wambon or in other subgroups of Digul River-Ok:

One root common to Yakhai and Awyu, that for “canoe,” was loaned into Awyu from Marind, with “paddle” presented for comparison; the Awyu sound [*xʷ] is found only in loans from Marind and presumably continues aspirated reflexes [hʷ] of bilabial non-stop /*w/ as characterize Marind's western dialects:

… wihtout known cognates in either Wambon or Marind or in other subgroups of Digul River-Ok or Fly River. Some are tentatively assumed to be loans from Yakhai into Awyu as unfricated voiceless stops /*p *k/ have no known native origin in Awyu, while prenasalized stops such as velar /*ŋg/ are found in native Awyu terms only in compounds:

Where no outcomparison is known and …, the direction of borrowing is unclear:

… regional …

…

…

… Kriens and Lebold's (2010) Kotup and Wairu …despite … Nohon and Taim of Awyu's central division immediately to the east along the Pasue River and … the borrowed terms are more similar to forms found in West Awyu…:

…

…

…

… Marianne Strait … with Mombum attestations drawn from Drabbe (Drabbe 1950: 566-574) and Komelomsch (Mombum) and Konerawsch from Geurtjens (Geurtjens 1933: 398-433) … none have known cognates in Marind:

As excepting only Warkay, Marind is Yakhai's nearest relative, loanss from Marind into Yakhai can be difficult to distinguish from native words. At least one example is identifiable as such a loan because the suffixed formative /*-hei/ is characteristic of Marind where the base /*do/ means “blood”, while /*h/ would become zero not /*ɣ/ in a native term (the native Yakhai term /*kab-kob/ is likewise based upon /*kab/ “blood”):

…