Asmat-Kamrau Bay

Timothy Usher, Santa Fe Institute

The Asmat-Kamrau Bay family consists of nine languages spoken along nearly the entirety of the southwestern coastal lowlands of New Guinea, from Kamrau Bay in the northwest to just north of Kolopom Island in the southeast.

The internal classification of Asmat-Kamrau Bay is as follows:

Asmat-Kamrau Bay

Kamrau Bay

Buruwai

Kamrau

South Kamrau

North Kamrau

Asmat-Kamoro

Kamoro-Sempan

Kamoro

Sempan

Asmat

Casuarina Coast Asmat

Citak Asmat

North Asmat

Central Asmat

Modera (1830) variety of Kamoro (unobtained)

Miklucho-Maclay (1876) variety of Kamoro (unobtained)

Robidé van der Aa (1879) Lakahia variety of Kamoro (unobtained)

van der Sande (1907) vocabularies of Nagramadu, Angadi and Goreda varieties of Kamoro (unobtained)

Dumas (1911) large vocabulary of Mimika variety of Kamoro (unobtained)

Ray (1912: 326-330) comparative study of Angadi-Mimika (Kamoro) dialects, including Robidé van der Aa's Lakahia and van der Sande's Angadi, Nagramadu and Goreda as well as Utanata River, Mimika, Kupera Pukwa and Kiruru varieties of Kamoro

Feuilletau de Bruijn (1913) Kaja-kaja variety of Asmat (unobtained)

Feuilletau de Bruijn (1915) Kaja-kaja variety of Asmat (unobtained)

Drabbe (1950) unpublished comparative vocabularies of Kamoro, Sempan and Asmat

Boelaars (1950: 90-102) English-language recension of Drabbe's then-unpublished grammar of Tarjà Kàmoro

Drabbe (1953) grammar of Tarjà Kàmoro, comparison of Kamoro dialects and (pp. 96-104) 378 comparative terms for Kàmoro, Sémpan and Asmat

Drabbe (1954: 232-255) 100 comparative terms for Kàmoro, Sémpan and Kawenak (Central Asmat<)br>     Drabbe (1958) grammar of Keenok (Central Asmat)t

Drabbe (n.d.) dictionary of Keenok

Drabbe (1959) grammar of Kawenak (unobtained)

Drabbe (1959) dictionary of Kawenak (unobtained)

Drabbe (1963) study of Asmat dialects and (pp. 212-233) 377 comparative terms for Kawenak, Keenok and Keenakap varieties of Central Asmat, Kaünak (Citak) and Kaweinag (Casuarina Coast)

Drabbe (n.d.) dictionary of Kàmoro (unobtained)

Drabbe (n.d.) dictionary of Sémpan (unobtained)

Galis (1955) 16 comparative terms and numbers drawn from the preceding literature

Anceaux (n.d.) Asienara and Iria vocabularies as excerpted in Voorhoeve (1975) and in Greenberg's notebooks (n.d.)

Anceaux (1956) (unobtained,) reprinted in English as (1958)

Anceaux (1958: 119-120) 10 comparative terms for Asienara and Iria

Voorhoeve (1965) dictionary and grammar of Flamingo Bay variety of Central Asmat

Voorhoeve (1975: 369-374) general description of member languages based primarily upon Drabbe (1953, 1963) and Anceaux (n.d.)

Voorhoeve (1975: 100) 40 comparative terms for Iria (North Kamrau) and Asienara (Buruwai) after Anceaux (n.d.)

Voorhoeve (1980) survey of Asmat dialects and (pp. 61-121) 455 reconstructed terms for Proto-Asmat, including supporting forms from Kamoro and Sempan drawn from Drabbe (1953) and Iria from Anceaux (n.d.)

Voorhoeve (2005: 148-149, 152-154, 158-164) Proto-Asmat-Kamoro correspondences and (unexemplified) reconstructions, including Sabakor (Kamrau Bay)

Roesler (1972) phonology of Asmat of Ajam (unobtained)

Roesler and Roesler (2011) dictionary of Asmat (Kawenak) of Ayam village

Bromley (1973) orthography of Citak (unobtained)

Bromley (1975) orthography of Citak (unobtained)

van Arsdale (1974: 31-32) 66 comparative terms for Awok (Keenakap,) Wooi-Karmbis, Jinak (Tamnim,) Brazza and Upper Eilanden (Citak)

Peckham (1991:160-166) 208 comparative terms for Nanesa (Etna Bay) Kamoro

Lebold, Kriens and Susanto (2013) 239 comaprative terms for Central Asmat (Kawenak) of Isyaman village

Drabbe (1950: 545-546, 1953: 3, ibid.) recognized the close relationship between Kamoro, Sempan and Asmat. Voorhoeve would later include Drabbe's Central South Coast as Asmat-Kamoro in his proposed Central and South New Guinea Phylum (1968: 1, 3-4, McElhanon and Voorhoeve 1970: 10.)

Anceaux' (1958: 116-117) survey of the Bomberai Peninsula recognized the close relationship between Asienara and Iria, sharply distinguishing them from the other Papuan languages of the Bomberai. Anceaux' divisions are recapitulated in Voegelin and Voegelin (1965: 30-31.)

Greenberg (1971: 840) joined Kamrau Bay with Asmat-Kamoro, writing: “It has not been previously noted that these two languages [Asienara and Iria] are to be connected with the Kamoro group to the east from which they are separated both by the Etna Bay subgroup of the Western New Guinea subfamily and by Irahutu, an AN language.”

Voorhoeve (1975: 369-374) includes Iria and Asienara as a branch of Asmat-Kamoro alongside Kamoro, Sempan and Asmat, though without explicit citation to Greenberg. He suggests that Iria-Asienara may be closer to Asmat than to Kamoro. Voorhoeve's classification is recapitulated defectively in Wurm (1982: 137-138,) where both Iria and Asienara are given as first-order members coequal to Kamoro, Sempan and Central, Casuarina Coast, Citak and North varieties of Asmat; this defect was to propagate through derivative works such as Ruhlen (1987: 355) and versions of the Ethnologue (q.v. Grimes ed 2000: 469-486.) This is remedied (q.v. Lewis ed. 2009) following Voorhoeve (2005: 148-149) which reiterates (1975''s) four subgroups and designates Iria-Asienara as Sabakor.

[under construction]

Proto-Asmat-Kamrau Bay had 12 to 14 consonants and at least 5 vowels as follows:

Bilabial voiceless /*p *ɸ/ are in nearly complimentary distributionm with /*p/ occuring only word medially between unlike vowels. If the distinction between them at the Asmat-Kamrau Bay level is valid, it is only barely so and probably of recent origin.

Apical stops /*t *d/ are marginal at best, appearing in only a few roots.

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

Consonant clusters do not occur.

Ours differs from the 11-consonant inventory presented in Voorhoeve (2005: 148-149) in that Voorhoeve does not recognize the existence of voiced stops /*b *d *g/ and treats /*ɟ/ as voiceless /*c/:

…

Besides the correspondences themselves (below,) the originality of Kamrau Bay voiced stops is evidenced indirectly. In all Asmat-Kamoro languages, the lack of phonemic voiced stops leaves nasals /m n/ free to vary with phonetic [mb nd b d] (Drabbe 1953: 6, 88, 1963: 10-11, Voorhoeve 1980: 14-42.) This occurs also in the Cook River languages to the south, which likewise lack phonemic /b d/ (Voorhoeve 1971: 84.) This phenomenon is very conspicuous in materials which have not been adjusted to account for this allophony (q.v. Ray 1912: 327, Drabbe 1963: 212-233, Voorhoeve 1980 ibid.) No trace of it is found in Anceaux' (n.d.) phonetic transcriptions of Asienara or Iria, because this phonetic space is already occupied by voiced stops /*b *d/. It is therefore a southwest coast regional phenomenon, probably originating in Proto-Asmat-Kamoro, which must postdate the breakup of Asmat-Kamrau Bay and the subsequent loss of contrastive voice.

Only a restricted set of consonants occurs finally:

Voorhoeve (2005: 148-149) gives consonant correspondences as follows, with forward-slashed alternates asserted to represent variation between Sabakor (Kamrau Bay) dialects:

(n.b. Voorhoeve gives /j/ as <y>)

…

Initial consonants correspond as follows, with Voorhoeve's (2005) values presented for comparison:

Medial consonants correspond as follows:

Subfinal consonants correspond as follows:

The reconstruction of Proto Asmat-Kamoro adopted in the charts below is largely based upon that presented in Voorhoeve (1980,) but differs from it and the system presumed in Voorhoeve (2005) in three significant respects: First, we reconstruct Asmat-Kamoro /*f/ where Voorhoeve (1980) has zero, or zero alternating with /*f/. Voorhoeve (pp. 62-63, 79-80) was aware of but uncertain about this solution, and didn't consistently implement it across his protoforms; this is addressed in Voorhoeve (2005: 148-149.) Second, Voorhoeve's (1980) palatal stop <*T>. (2005 /*c/) when followed by /*i/ is treated as as an allophone of Asmat-Kamoro/*t/, as it does not follow the same correspondence in Kamrau Bay as does Asmat-Kamoro /*c/ : Kamrau Bay /*j./ < /*ɟ/.

All consonants besides apical non-stop /*r/ and the rare apical voiced stop /*d/ are found initially..

Initial bilabial nasal /*m// is retained as such in both subgroups:

Initial apical nasal /*n/ is retained as such in both subgroups:

Voiceless bilabial stop /*p/ has not been found to occur initially.

Initial apical voiceless fricative /*s/ is retained as such in both subgroups:

Initial velar voiceless stop /*k/ is lost in Kamrau Bay as it is in Kamoro and Sempan:

As initial /*k/ is retained only sporadically in Asmat varieties (Voorhoeve 1980: 90, 117, 118,) it's probable that some forms which have been reconstructed here with an initial vowel in fact began with /*k/. Some of these may be recoverable through outcomparison; for example Asmat-Kamoro /*ew/ “crocodile” was probably really /*kew/ as evidenced by Cook River /*kVw/, Marind /*kiu/, even though there is no longer any direct evidence for this in Asmat (p. 76.) Similarly, /*apimi/ “sago pulp” was probably really /*kapimi/ as witnessed by Gondu River /*kaipm/ “sago tree” We conclude that Asmat likely is (or was) in the process of eliminating initial /*k/ altogether.

Initial bilabial voiced stop /*b/ is devoiced to /*p/ in Asmat-Kamoro:

Apical voiced stop /*d/ has not been found to occur initially.

Initial palatal voiced stop /*ɟ/ is deoccluded to non-stop /*j/ in Kamrau Bay, filling the gap left by loss of Asmat-Kamrau Bay initial /*j/ (below,) and is devoiced to /*c/ in Asmat-Kamoro:

When root-initial /*ɟ/ has become word-medial in Kamrau Bay due to compounding, it is not deoccluded to /*j/ but remains as /*ɟ/:

The second person plural shows a secondary and possibly irregular change of Kamrau Bay /*ja/ to /*e/:

One example appears to be intractably irregular, with Voorhoeve's (1980: 94) Kawenak forms clearly indicating apical /*t/ rather than /*c/:

Initial velar voiced stop /*g/ becomes /*t/ in Asmat-Kamoro, presumably having merged with /*d/ prior to its devoicing (above) :

Initial bilabial voiceless fricative /*ɸ/ is not distinct from voicless stop /*p/ which has not been found to occur initially. It is retained as such in both subgroups:

Initial bilabial non-stop/*w/ …:

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

Palatal non-stop /*j/ is lost in Kamrau Bay when followed by vowels other than /*u/ (below):

…/*j/ …:

When followed by high back rounded vowel /*u/, initial /*j/ is lost in Kamrau Bay but not before fronting and derounding /*u/ to /*i/:

Any consonant can occur medially.

Medial bilabial nasal /*m// is retained as such in both subgroups except when in subfinal position (below):

Medial apical nasal /*n/ is retained as such in both subgroups except when in subfinal position (below):

Medial bilabial voiceless stop /*p/ is retained as such in both subgroups except when in subfinal position (below):

Medial bilabial voiceless fricative /*ɸ/ is genarally retained as such in both subgroups, although is typically lost in most Asmat-Kamoro dialects so in a few instances its reconstruction is supported only by outcomparison to Kamrau Bay:

…/*ɸ/ …:

In one puzzling example, Kamrau Bay /*ɸ/ looks to correspon d to Asmat /*p/ (cf. Muli Strait /*p[ie̝]/ where /*p/ corresponds to either):

Medial apical voiceless stop /*t/ is merged with velar /*k/ in Asmat-Kamoro:

Medial apical voiceless fricative /*s/ is generally retained as such in both subgroups:

There is one puzling exception in which Kamrau Bay /*s/ corresponds to Asmat-Kamoro /*t/ (or /*c/):

Medial velar voiceless stop /*k/ is retained as such in both subgroups except when in subfinal position (below):

The behavior of one word suggests that it must have been a compound , as medial /*k/ is not normally dropped in Kamrau Bay or in Asmat (q.v. Voorhoeve 1980: 99):

Medial bilabial voiced stop /*b/, like initial /*b/ (above,) is devoiced to /*p/ in Asmat-Kamoro. Voorhoeve (2005:148-149) derives both /*b *p/ from original /*p/, but medially these are plainly contrastive in Kamrau Bay:

Medial apical voiced stop /*d/ has been found in only one example, where it is devoiced to /*t/ in Asmat-Kamoro:

Medial palatal voiced stop /*ɟ/ is, unlike initial /*ɟ/ (above,) retained as such in Kamrau Bay and is devoiced to /*c/ in Asmat-Kamoro:

Medial velar voiced stop /*g/, like initial /*g/ (above,) becomes /*t/ in Asmat-Kamoro, presumably having merged with /*d/ prior to its devoicing (above) :

Medial bilabial non-stop /*w/ is retained as such in both subgroups:

… /*w/ …:

Medial apical non-stop /*r/ is generally retained as such in both subgroups:

Medial palatal non -stop /*j/ …:

Consonats which aremedial in Asmat-Kamoro but becomefinalin Kamrau Bay due to loss of final vowels are here called subfinal consonants. These develop very differently from medials which remain medial in Kamrau Bay. Verbs are mandatorily followed by suffixes and are not affected by subfinality.

Voorhoeve (1980: 41-42) misinterprets the resulting patterns as morphological rather than phonological in origin: “It is possible that we have here the remnants of an old class marking system. Some evidence that -t, -to, -ne, -ke and perhaps -kot are petrified suffixes can be found when comparing the Asmat languages with other members of the family, especially Iria-Asienara.” Analogous errors are found in Bee's (1965) Eastern Highlands and in Lloyd's (1969) Baruya, where the morphophonemic variations of a postposed article are likewise misanalyzed as variations in the underlying form of the article itself, leading to the impression of noun classes.

Subfinal bilabial nasal /*m/ yields Kamrau Bay /*m/:

Subfinal apical nasal /*n/ following front vowels /*i *e/ yields Kamrau Bay final /*n/:

When subfinal /*-nV/ follows low central /*a/ or mid back rounded *o/, it is reflected as bilabial /*m/ in Kamrau Bay:

Subfinal bilabial voiceless stop /*p/ …:

…:

…:

Subfinal bilabial voiceless fricative /*ɸ/ yields Kamrau Bay final /*k/:

Subfinal laminal voiceless fricative /*s/ becomes final /*t/ in Kamrau Bay. There is only one known example:

However, there are counterexamples, suggesting along with the rarity of Kamrau Bay final /*t/ that the paradigm was in the process of being eliminated in favor of zero:

Subfinal palatal voiced stop /*ɟ/ …:

Subfinal apical non-stop /*ɾ/ yields final /*r/ in Kamrau Bay. …:

Kamrau Bay final /*ɟ/ …:

…:

…:

…:

…:

…:

…:

… vowels …

/*i/ …:

/*e/…:

High back rounded vowel /*u/ is retained as such in both families, although is usually merged with front /*i/ in Sempand and in Asmat:

…:

…:

… /*ae/ …:

…:

…:

…:

Asmat-Kamrau Bay free pronouns are reconstructed as follows:

[under construction]