Lowland Ok

Timothy Usher, Santa Fe Institute

The Lowland Ok famil­y is comprised of perhaps six closely-related languages spoken in the lowlands south of New Guinea's central cordillera and east of the Kao river, along and around the courses of the Iwur, Muyu and Tedi rivers, on both sides of the border between Indonesia's Papua province and Papua New Guinea's Western province. Its nearest relatives are other subfamilies of Ok to the northwest and northeast.

The internal classification of Lowland Ok is as follows:

Lowland Ok

Ninggerum

Iwur-Koma

Iwur

Koma (Komanarepket)

Muyu

North Muyu

South Muyu

Yongkom

The number of languages represented by the very numerous Muyu vocabularies isn't entirely clear. All are very similar to one another, whereas Ninggerum and its dialects are sharply distinct.

Austen (1920-1921: 132) vocabulary of (North) Muiu of Kandam village

Austen (1921-1922: 159-162) 132 comparative terms for Marapka (Yongkom,) Ort (Yongkom,) Kandam (North Muyu) and Anu (South Muyu)

Austen (1924-1925: 75) 37 comparative terms for Ionggom (Yongkom) and Upper Tedi River (Ninggerum)

Austen (1925-1926: 75) vocabulary of Daupka (Ninggerum) (unobtained)

Geurtjens (1933: 398-433) 457 comparative terns for Digoeleesch (South Muyu)

Boelaars (1950: 68-76) English language recension of Drabbe's then-unpublished grammar of Niinati (North Muyu)

Drabbe (1954: 150-211) grammar and (pp. 214-229) 419 comparative terms and (pp. 231-255) 100 comparative terms for Kati Niinatì (North Muyu) and Katì Metòmka (South Muyu)

Drabbe (1959: 162-181) 431 comparative terms for Kati I (Metomka) and Kati II (Niinati)

Healey (1964: 78-97) 120 comparative terms for Lowland Ok languages and (pp. 170-180) 156 Lowland Ok reconstructions

Voorhoeve (1975: 381-384) general descriptions of Southern Kati, Northern Kati, Yonggom, Iwur and Ninggirum

Voorhoeve (1975: 97) 39 comparative terms for Ninggirum, 29 for Yonggom of Lake Murray and (p. 98) 40 for North Kati and South Kati drawn from Drabbe

Stokhof ed. (1982) Holle comparative vocabularies of (pp. 123-124) Central & Lower Uwimerah (South Muyu) (unattributed) , (pp. 125-127) Proa bivouac (Koma) (unattributed) , (pp. 131-132) Uwimerah (source) (Iwur or Koma) (unattributed), (pp. 151-161) Kaôh (South Muyu) after van Doorn, (pp. 163-173) Muyu (North Muyu) after van Doorn and (pp. 175-185) Tanah Merah area (mix of Muyu dialects) (unattribued)

Christensen (1995) grammar of Yongkom (unobtained)

Christensen (1999) orthography of Yongkom (unobtained)

Christensen (2000) phonology of Yongkom (unobtained)

Christensen (2000) grammar of Yongkom (unobtained)

Christensen (2002) sketch phonology of Yongkom

Christensen (2010) Yongkom discourse

Christensen (2013) grammar of Yongkom (unobtained)

Donohue (1996: 18-50) 151 comparative terms for Muyu

Jang (2003: 39-47) 239 comparative terms for Dintere of Iwur village and (Komanarepket) Muyu of Ikcan

Kroneman unpublished survey vocabulary of Niggerum obtained in digitalized form from Paul Whitehouse via the Summer Institute of Linguistics.

Additionally, unpublished survey vocabularies of Wogham (Ninggerum,) Kataut, Ok Bari, Moian, Upper Yong River, Kabon, Akamba, Kakap and Lower Yong River were obtained in digitalized form from Paul Whitehouse via the Summer Institute of Linguistics; however these are unattributed.

[under construction]

The presence of a number of loans from Lowland Ok into the basic vocabulary of Aekyom (Awin) (q.v. Austen 1924-1925: 75) led Greenberg (1971: 825) to wrongly include Aekyom in his Ok subgroup of Southwest New Guinea.

[under construction]

Proto-Lowland Ok had perhaps 14 consonants and 6 vowels as follows:

It's possible that there was another consonant, represented as /*j₂/ below.

Laminal /*s/ is uncommon, and is not known to occur root-medially.

Apical non-stop /*ɾ/ does not occur initially.

Consonant clusters do not occur except across morpheme boundaries in compounds.

Stress falls invariably upon the final syllable of the root (Voorhoeve 1975: 381, 384,) motivating vowel agreement in Muyu. Healey (1964: 61) states that Yonggom has contrastive tone, all nine given examples of which are monosyllables. These do not consistently correspond to tones in their Mountain Ok cognates.

Only a restricted set of consonants occurs word-finally:

Initial consonants correspond as follows, with Healey's (1964: 169) values presented for comparison:

Medial consonants correspond as follows:

Final consonants correspond as follows:

Vowels correspond as follows:

These correspondences are exemplified as follows, with all Iwur and Ikcan (Koma) attestations drawn from Jang's (2003: 39-47) Dintere of Iwur village and Muyu of Ikcan, and Ninggerum from Healey (1964: 78-97, 170-180) unless otherwise indicated. Ninggerum forms drawn from Voorhoeve, Kroneman (n.d.) and SIL's unattributed Wogham vocabulary are designated [cv], [rk] and [sil] respectively. For clarity of presentation, the general dispositions of initial consonants will presented first, then those of medial consonants, then those of final cononants, followed by those of vowels. Some details of the forms which follow are necessarily tentative.

Any consonant except apical non-stop /*ɾ/ can occur initially.

Initial bilabial nasal /*m/ is retained as such in all descendants:

Initial apical /*n/ is retained as such in all descendants. It is substantially less common than /*m/, an assymetry shared with other subfamilies of Ok and Digul River:

One example shows an irregular alternation between bilabial nasal /m/ in Iwur and apical /n/ elsewhere. This root is found with /*m/ in other subfamilies of both Ok and Digul River, leaving no doubt that /*m/ is original; however Jang's (2003: 45) Dintere attestation may be immediately a loan from Nakai to the west or in error.:

Initial bilabial voiceless stop /*p/ is deoccluded to /h/ in Ninggerum and dropped elsewhere:

Two examples shows initial /*p/ retained in Iwur and Ikcan as voiced [b] in a compound, in keeping with the medial reflex of /*p/ (below):

Initial apical voiceless stop /*t/ is retained as such in all descendants:

Initial laminal voiceless /*s/ is preserved as such in Ikcan and merged with apical voiceless stop /t/ in Ninggerum and Muyu. The Iwur reflex isn't known:

Initial velar voiceless stop /*k/ followed by non-front vowels /*a *o *u/ is retained as such in all descendants:

When followed by front vowels /*i *e/, initial /*k/ is palatalized to /*kʲ/ in Muyu; this becomes contrastive due to neutralization of the first vowel of disyllables (below):

…:

Initial rounded velar voiceless stop /*kʷ/… :

… /*kʷ/… :

When followed by high front vowel /*i/, initial /*kʷ/… :

Initial bilabial voiced prenasalized stop /*mb/ is realized as plain voiced[b] in all decendants:

Initial apical voiced prenasalized stop /*nd/ is known in only two examples, in which it is realized as plain voiced [d ɖ] in all decendants:

Initial velar voiced prenasalized stop /*ŋg/ is distinguishable only in Koma, having been merged with plain stop /*k/ in all other Ok languages:

A number of examples show a different correspondence, with initial velar voiced prenasalized /*ŋg₂/ yielding nasal [ŋ] in Iwur and in Ikcan, as do medial reflexes of /*ŋg/ (below,) and /g/ in Ninggerum, which appears to be synchronically distinct from /k/ (above.) It's possible that these are loans from Ngalum to the north, in which the realization of initial /*ŋg/ as /ŋ/ rather than [g ɢ] is regular, or that they were elicited in non-initial environments:

Initial bilabial non-stop /*w/ is generally retained as such in all descendants:

Apical non-stop /*ɾ/ does not occur initially because proto-Ok initial /*ɾ/ has merged with palatal /*j/ (below.)

Initial palatal non-stop /*j/ is occluded to [d] in Ninggerum when followed by a non-nasal consonant in the onset of the next syllable:

When the onset of the following syllable is bilabial nasal /*m/, initial /*j/ is generally occluded to apical nasal /n/ in Ninggerum; however this is not exceptionless:

The quality attested by the following palatal correspondence is unknown. There's no reason to think that it was /*kʲ/ as in Healey (1964); these have been indicated as /*j₂/ pending resolution of the problem:

Since /*j₂/ is distinguished from /*j/ only in Ninggerum, it is possible that some examples shown with /*j/ which are supported only by Muyu and outcomparison really had /*j₂/:

Any consonant can occur medially, excepting perhaps rounded laminals /*s/ and rounded velar stops /*kʷ *ŋgʷ/ which have not been found in medial position.

Medial bilabial nasal /*m/ is retained as such in all descendants:

In two examples, medial /*m/ is dropped in Muyu due to syllable reduction:

Medial apical nasal /*n/ is generally retained as such in all descendants, though a few Ninggerum examples are given with [d]:

Root-medial bilabial voiceless stop /*p/ was rare in Proto-Ok, and has been found in only one Lowland Ok example. While the Iwur and Ikcan reflexes are indistinguishable from those of voiced prenasalized /*mb/ (below,) the medial of the Mountain Ok cognate is/*p/, not /*mb/ (q.v. Healey 1964: 158):

Otherwise, medial /*p/ is found only as the final component of verb roots, which are manditorily suffixed

Medial apical voiceless stop /*t/ is usually lenited to [d r] in Ninggerum:

Laminal voiceless /*s/ has not been found to occur medially.

Medial velar voiceless stop /*k/ is lost in Ninggerum and Ikcan, and generally also in Muyu, typically resulting in long vowels. Iwur attestations are inconsistent in this respect:

The behavior of erstwhile root-initial /*k/ in compounds …:

…:

Medial bilabial voiced prenasalized stop /*mb/ is denasalized to plain voiced/b/ in Iwur and Ikcan as is initial /*mb/ (above) and is usually deoccluded to non-stop [w] in Ninggerum:

In one example the Ninggerum reflex of bilabial voiced prenasalized stop /*mb/ is given as nasal [m]:

Medial apical voiced prenasalized stop /*nd/ is denasalized and merged with non-stop /r/ in Iwur and Ikcan, and like medial /*r/ (below) is dropped in Ninggerum:

In one example phonetic [nd] resulting from cluster /*n/+/*j/ is likewise dropped in Niggerum. Here Iwur and Ikcan [l] appear to reflect root-initial /*j/ (above) rather than medial [*nd]:

Medial velar voiced prenasalized stop /*ŋg/ is reflected as nasal /ŋ/ in Ninggerum, Iwur and Ikcan:

Clusters /*nk *ŋgk/ created by compounding are also reflected as [ŋg]:

(Muyu /*mun/ “child” is a loan from Wambon.)

Medial bilabial non-stop /*w/ …:

One example of medial /*w/ is dropped in Muyu when followed by /*i/ in the next syllable; this is likely a consequence of vowel leveling (below) following an intermediate stage where these words had assumed the form /*Ciwi(C)/:

Medial apical non-stop /*ɾ/ is dropped in Ninggerum:

In one example, medial /*ɾ/ is lost in Muyu due to syllable reduction:

Apparent medial palatal non-stop /*j/ is known in only one example, in which it is occluded in Ninggerum and in Iwur. The failure of this word to undergo vowel leveling in Muyu (below) suggests it to be a compound:

Only a restricted set of consonants occurs finally, including nasals /*m n/, voiceless stops /*p *t *k/, velar voiced prenasalized stop /*ŋg/, which as in other Ok subgroups is invariably realized as nasal [ŋ], and apical non-stop /*ɾ/.

Final bilabial nasal /*m/ is retained as such in all descendants:

Final apical nasal /*n/ is retained as such in all descendants:

Final bilabial voiceless stop /*p/ is retained as such in all descendants. It is typically realized as voiced [b] in Ninggerum and in the Muyu dialects. As with finals /*t *k/ (below,) Iwur and Ikcan reflexes are unreleased:

Final apical voiceless stop /*t/ is retained as such in Ninggerum and in Muyu. It's not clear if the Iwur and Ikcan reflexes are distinct from those of final non-stop /*ɾ/ (below):

Final velar voiceless stop /*k/ is retained as such in all descendants:

Final velar voiced prenasalized stop /*ŋg/ is realized as nasal [ŋ] in all descendants, as it is throughout the Ok family:

Final apical non-stop /*ɾ/ is dropped in Ninggerum with compensatory length in the preceding vowel. It's not clear if the Iwur and Ikcan reflexes are distinct from those of final plain stop /*t/ (above;) these are provisionally given here as /t/:

One example has final /*t/ in Muyu, but with the curiosity of a nasalized vowel in Ninggerum, which is usually associated with the loss of medial /*nd/:

This is likely the reflex of Ok final laminal /*s/ or /*ndz/; cf. Mountain Ok /*às/ rather than /*àɾ/. Muyu final /*t/ does not appear in compounds such as /*a-moɾ/ “embers” and /*a-wop/ “fire.”

Vowels in monosyllables are generally retained as such in all descendants, as are vowels of identical quality in disyllables.

High front vowel /*i/ in simple environments is retained as such in all descendants:

High central vowel /*ʉ/ is uncommon, but its existence is supported by comparison to cognates in other subfamilies of Ok:

The vowel in one example corresponds to /*ʉ/ in other subfamilies of Ok, but is given only as /i/ in the dialects of Muyu:

The base of the first person plural shows /u/ rather than /i/ in Ikcan:

High back vowel /*u/ in simple environments is retained as such in all descendants:

Mid front vowel /*e/ in simple environments is retained as such in all descendants:

Mid back vowel /*o/ in simple environments is retained as such in all descendants:

Low central vowel /*a/ in simple environments is retained as such in all descendants:

Where the vowels of the first and second syllable differ, Muyu has undergone changes which bring them into agreement.

Roots of the form /*CaCV₂(C)/ are reflected in Muyu as /*CV₂CV₂(C)/:

…:

Examples which violate these rules can only be compounds or loans:

Initial /*a/ in roots of the form /*aCV(C)/ is unaffected by vowel levelling. By far the most common of these patterns is /*aCo(C)/:

Words of the form /*CiCu(C) *CuCi(C)/ are reflected as /*CuCu(C) *CiCi(C)/ in Muyu:

One word of the form /*Cowi(C)/ is reduced to /*Ci(C)/ in Muyu, probably following an intermediate stage /*Ciwi(C)/:

…

…

…

…

…:

Similar processes affect Ninggerum and Muyu alike relative to proto-Ok:

It's possible that these changes postdate proto-Lowland Ok and affect both descendants due to areal influence. For example, Aekyom /kendoke/ “ear” appears to be a loan from a form of Lowland Ok in which the root was /*kendoŋg/ rather than /*kende/.

Lowland Ok free pronouns are reconstructed as follows, with Ninggerum attestations drawn from Voorhoeve (1975: 384) and Iwur and Ikcan (Koma) from Jang (2003: 40):

(note: Voorhoeve gives Ninggerum /ni/ 1 pl. incl., /nip/ 1 pl. excl., but comparison to Ngalum and Mian shows this to be in error.)

An explanation of the unusual gloss “3 pl. incl.” is in order. It may be seen that the suffix /*-p/ indicates the inclusion of the second person. There are two forms of the second person plural in both Lowland Ok and Mountain Ok, one with initial /*k/ and one without it. The first of these is the regular reflex of Digul River-Ok /*ŋgip/; the second is found only in Ok . These cannot be regularly reduced to a single form. The solution adopted here assumes that /*-p/ could be added to the third person plural with the meaning, “they and you (sg.),” as opposed to “you (sg.) and [your companions],” and that this form /*[j]ip/ has replaced /*kip/ in many descendants. South Muyu, for example, has both forms (q.v. Drabbe 1954: 162-163;) however, the semantic distinction proposed here is not directly attested.

As in other Ok subgroups, the free pronouns function as possessives when preposed to nominals (q.v. Drabbe 1954: 163-164.)

… (q.v. Healey 1964: 71):

Aekyom appends low mid front vowel /ɛ/ to Lowland Ok words with final non-nasal consonants, with Aekyom forms drawn from Stewart (n.d.):

It's conceivable that some of these comparisons are legitimate cognates, for example the term for “shoulder” is found also in Kamula and Doso far to the southeast as /make/; however these languages are by no means closely related to Digul River-Ok, and any genuine cognates must necessarily be of great antiquity. Aekyom /mVgate/ “mouth” is found also in Kamula as well as a number of other subgroups of Trans-New Guinea, and may be one such example.