Kolopom

Timothy Usher, Santa Fe Institute

The Kolopom language family consists of four languages spoken in the Merauke regency of Indonesia's Papua Province, three on Kolopom (Yos Sudarso, Frederik-Hendrik) island and one, Moraori (Morari,) located east of Marind territory near the Papuan border. The Fly River language Marind is plainly intrusive to the region, and is presumably the reason for Moraori's geographical separation from the insular Kolopom languages.

The internal classification of Kolopom is as follows:

Kolopom

Moraori

Ndom

Central Kolopom

Riantana

Kimaghama

Voorhoeve (1968: 8) gives the following lexicostatistics for the Kolopom family, excluding Moraori which was not then known to belong:

The figure 13% is given for Kimaghama and Moraori (1968: 9.) As Moraori is riddled with loans from Kanum, Yei and especially Marind, any unadjusted figure will understate the genetic relationship. Since Voorhoeve did not specify here the asserrtions of cognacy underlying his figures, they're impossible to directly evaluate; however Voorhoeve (1968: 10) relied exclusively upon Drabbe (1949, 1954,) so the same data is available here for any of us to propose our own figures.

[under construction]

Geurtjens (1933: 398-433) 455 comparative terms for Kaladdar(sch) (west) and Teri Kalwa(sch) (northeast) varieties of Kimaghama.

Nevermann (1939: 60-69) comparative vocabulary of Moraori.

Drabbe (1949) brief grammars of and (pp. 14-24) … comparative terms for Kimaghama, Riantana and Ndom.

Drabbe (1954: 15-26, 53-68) detailed grammar and comparative vocabulary of Moraori.

Boelaars (1950: 33-36 44-48) English-language recension of Drabbe's Kimaghama and Moraori

Voorhoeve (1975: 364-366) grammar notes for Kimaghama, Riantana and Ndom following Drabbe (!949)

Voorhoeve (1975) …     Donohue (1996) notes on and comparative vocabulary of Moraori (Wasur.)

Gebze and Donohue (1998) dictionary of Moraori (unobtained)

Menanti and Susanto (2001) surtvey of Kimaam (unobtained)

Susanto (2001) 233 comparative terms for Kalilam (Ndom) of Wetau village

Choi (2015) grammar of Riantana and (pp. 55-57) 203 terms

Arka … Marori …

Arka … Marori …

Arka … Marori …

Arka … Marori …

Arka … Marori …

Arka … Marori …

Arka … Marori …

Arka … Marori …

In addition to these, Paul Whitehouse provided rekeyed versions of four unattributed survey vocabularies from the Summer Institute of Linguistics for Ndom, Riantana and two Kimaghama villages.

[under construction]

The taxonomic unity of the insular Kolopom languages has been evident since 1949, when Drabbe presented comparative vocabularies and brief notes on the grammars of Kimaghama, Riantana and Ndom. According to Boelaars (1950: 33,) Drabbe himself did not consider the languages genetically related, stating that “a comparison of the vocabularies of these three languages does not show resemblances and certainly no fixed soundshifting.”

Voorhoeve (1968: 8-9) places Kolopom, along with the unrelated Bulaka River and Morehead River families, with his proposed Central and South New Guinea Stock at the phylum level.

Moraori has been the object of taxonomic confusion.

Drabbe (1950: …) …

Voorhoeve (1968: …) …

Greenberg (1971: …) …

Wurm (1975: …) …

Wurm (1982: …) …

Ross (…) …

The Kolopom languages are too diverse to support a robust reconstruction without substantially more data. Accordingly, we're unable to present as complete and as accurate a system as we would have liked, for example those of Awyu-Dumut (Healey 1970,) Bulaka River or Marind, which are likewise based primarily on Drabbe, using more or less the same termlist. However, an outline of the most common correspondences can be presented, which can provide a basis for further investigation. Interpretations of least common correspondences are necessarily tentative.

[under construction]

Proto-Kolopom had between 10 and 14 [resolve *r *ɣ] consonants and [unaddressed] vowels:

There is as of yet no segmental evidence that /*j *w/ are distinct from sequential vowels /*i *u/, as they are in neighboring families.

All attestations in this section are from Drabbe (1949, 1954) unless otherwise specified. Where Moraori or Kimaghama attestations are separated by a semicolon, the first form is Drabbe's and the second is Donohue's or Geurtjens (1933) respectively.

We recognize that not all Central Kolopom roots date to Proto-Kolopom, but find it useful to present roots shared by only only Riantana and Kimaghama along with the others.

Initial bibabial nasal /*m/ …:

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

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

Medial apical nasal /*n/ …:

Initial bilabial voiceless stop /*p/ is fricated to /f/ in Ndom:

Medial bilabial voiceless stop /*p/ … [β] …. :

Moraori's apparent retention of /*p/ as [p b] in Donohue's and Arka's attestations of the following word probably result from excrescence of the cluster /mr/ :

Intitial apical voiceless stop /*t/ is retained as such an all four languages:

Medial apical voiceless stop /*t/ …

Initial velar voiceless stop /*k/ is lenited to x [x ɣ] in Ndom and to [w ø] in Moraori. …:

Initial velar voiceless stop /*k/ is lenited to x [x ɣ] in Ndom and to [w ø] in Moraori. …:

Medial velar voiceless stop /*k/ is lenited to x [x ɣ] in Ndom and to [w ø] in Moraori. It is sometimes realized as [ɣ] in Kimaghama. …:

Voiced prenasalized stops /*mb *nd/ …

…

…

…

The status of velars other than /*k/ is difficult to determine. The clearest example of /*ŋg/ is the second person plural:

The second person singular poses a problem in that the Ndom form suggests /*k/ or perhaps /*ɣ/ (see below):

…

Initial velar non-stop /*ɣ/ …:

Medial velar non-stop /*ɣ/ …:

Voiceless fricative or affricate /*s/ is realized by allophones [θ s ts] in Ndom and by voiceless laminal stop /tʲ/ in Central Kolopom:

[CHART]“knife”, “taro”, “bitter”, “break (stone)”, ?“sweet”, “tell lies”

Medial /*s/ has the same values, but in Kimaghama /tʲ/ is optionally lenited to [j]:

“Sneeze” is found also in (at least) Muli Strait, Marind, Lake Murray, Awyu, Bulaka River (where it is a loan from Kolopom or Marind) and as far away as Koiari-Managalas Plateau (q.v. Dutton 2010: 58) and Guhu-Oro, allowing us to assign the value /*s/ to this correspondence with some confidence. Ndom's allophones of /s/ [θ ts] are also found in Cook and Gondu Rivers (Voorhoeve 1971: 84) and in Asmat dialects (Voorhoeve 1980: CITE) as allophones of /s/. Riantana and Kimaghama's fortition of /s/ to a stop is shared with Yelmek immediately to the east and Yaqay to the northeast, where Bulaka River and Yaqay-Warkay's /*s/ are realized as /t/.

Prenasalized affricate /*ndz/ is merged with /s/ in Ndom and lenited to /r/ in Riantana. Its initial reflexes in Kimaghama appear to be indistinguishable from those of /*s/:

[CHART], “hard”, “shiver”, ?““see/hear”

Medial /*ndz/ …

[CHART]“intestines”

Strong support for the identification of this correspondence with /*ndz/ is found in Geurtjens' nasalized medial reflex [ndʲ] in “bathe” and “grandchild” as well as Kimaghama's nasalized [ĩ] in “man,” as there otherwise would seem to be no other place nasality could have come from. “Urine” is a compount /*kat ndzu/ “testicle water,” unambiguously equating Kimaghama's prenasalized medial allophone with the denasalized initial reflex.

Initial bilabial non-stop /*w/ is neither occluded nor aspirated, as it is in neighboring Muli Strait and Marind (and Maklew) respectively:

Medial bilabial non-stop /*w/ is neither occluded nor aspirated, as it is in neighboring Muli Strait and Marind (and Maklew) respectively:

Initial apical non-stop /*r/ …

Medial apical non-stop /*r/ …

Palatal non-stop /*j/ is occluded to /r/ [r d] in Riantana:

[are these medials distinct from /*-ndz-/?]

Medial palatal non-stop /*j/ …:

All final consonants are lost in Central Kolopom.

Final /*m/ is preserved as such in Moraori and Ndom:

One Marind word, in Riantana's case via Yelmek, indicates that Central Kolopom's final dropping is or was synchronically operative in recent times, postdating for example Yelmek's shift of /*ɣ/ to /ŋ/:

Ndom denasalizes final /*n/ to /r/ where a nasal /*m *n/ appears in a previous segment, a rule which also governs the reflexes of final /*ŋg/ (below):

As with initials, the final reflexes of peripheral stops /*p *k/ are fricated in Ndom and fricated or lost in Moraori:

[double-check separation of Riantana [-βo -bu -ko], possibly /*pV *kV/]

Final /*t/ is realized as /r/ in Moraori and Ndom [was there a distinction /*t *r/ that is lost here?]:

[outcomparison suggests /*dzisin/ “sand” – some of these may have been final /*n/]

Ndom has meged the reflexes of final /*ŋg/ with those of /*n/, with regular denasalization to /r/ where a nasal /*m *n/ appears in a previous segment. Moraori takes the opposite turn, denasalizing /*ŋg/ [ŋg ŋk ŋɣ] to /ɣ/ [ɣ w] when not preceded by a nasal:

[but Ndom “louse” does not follow this rule as stated]

…

Final prenasalized stops /*mb *nd/ have only been identified in a few examples:

Final /*s/ [?is this distinct from final /*ndz/] …

One example of Ndom final [t] occurs on a widespread regional term:

Final consonants are preserved in Riantana when suffixed with the particle /-aˈna/, which appears to signify a mass or collective:

Similarly, this loan from Marind:

An analogous phenomenon occurs in Kimaghama before [-aɣa] plural; Drabbe (1949: 4) interprets the surfaced consonant as a component of the suffix:

Several kin terms do not lose their finals in Kimaghama because they have been suffixed with a morpheme /-u/:

The nominative forms of the personal pronouns (q.v. Drabbe 1949: 5-6, 1954: 16) are as follow:

The oblique forms were derived from the nominatives by the suffixation of /*-i/:

Moraori has generalized the third person obliques to the nominatives, a grammatical change shared with Maklew, with the oblique coming to be signified by the addition of a prefix /ŋga/.

What we presume to be the original possessive pattern has been lost outside Moraori. The presumption of originality is based upon outcomparison to Cook and Gondu River's singular possessives (q.v. Voorhoeve 1975: 85-86,) with the shared idiosyncrasy of a zero base in the first person singular:

In the other Kolopom languages, the possessive base reflects either the oblique or the nominative:

Prefixed undergoers, too, have been lost outside of Moraori. The second and third person are original to Trans-New Guinea, and hence to Kolopom, but the first person form is unexpected:

[under construction]

…

Several roots are shared with the unrelated Bulaka River family to the north (q.v. Usher 2014):

Several of these roots are shared with other families. “Stone” is found also in Cook River, Kolopom's nearest relative, as well as in Muli Strait and thus might be a loan in either direction. “Cassowary” is found also in the Paho River and Oriomo Plateau families to the east. This must be of ancient origin as it is not found in either Marind or in the Morehead River family which now separates these enclaves. “Older sibling” … “Breast” …

…:

One word shared with Morehead River, most likely borrowed in the context of the senary counting system unique to this region of New Guinea (q.v. …, Choi 2015: 36,) has subequently been loaned from Kolopom into Marind and thence into Bulaka River. We can tell that Kolopom is the immiedate source for the Marind word because Marind /*h/ reflects previous /*ɸ/ (Usher and Suter 2015: 130) This loan must be ancient in origin because Morehead River medial /*n̪/ has since become /*θ/ in Bensbach River and /t/ in Yei:

… Muli Strait …

In more recent times, … Marind loans found only in Moraori are too numerous to list here …