Etna Bay

Timothy Usher, Santa Fe Institute

The Etna Bay family, also known as the Mairasi family, is comprised of three closely related languages spoken in the Bird's Neck region of Indonesia's West Papua province.

The internal classification of Etna Bay is as follows:

Etna Bay

Mairasi

Mer

Semimi

Peckham (1991: 152) calculates 62% lexical similarity between Mairasi and Semimi, or 128 cognates out of 205 comparative terms. No figures involving Mer are available.

Galis (1955-1956) 30 comparative terms for Kaniran (Mairasi)

Anceaux (1956, 1958: 119-120) 10 comparative terms for Mairasi and Etna Bay

Anceaux (n.d.) comparative vocabularies of Mairasi, Faranjao (Mairasi) and Etna Bay

Greenberg (n.d.) comparative vocabularies of Mairasi, Faranjao and Etna Bay after Anceaux (n.d.)

Voorhoeve (1975: 101) 40 comparative terms for Mairasi and Semimi after Anceaux (n.d.)

Peckham (1982) Mairasi verbal morphology

Peckham (1991) phonology of Mairasi of Lobo village

Peckham (1991: 173-180) 209 comparative terms for Mairasi, Mer and Semimi

Peckham unpublished survey vocabulary of Mer of Javor obtained in digitalized form from Paul Whitehouse via the Summer Institute of Linguistics

Peckham unpublished survey vocabulary of Semimi of Kiruru obtained in digitalized form from Paul Whitehouse via the Summer Institute of Linguistics

Walker and Peckham unpublished survey vocabulary of Mairasi obtained in digitalized form from Paul Whitehouse via the Summer Institute of Linguistics

Walker and Peckham unpublished survey vocabulary of Semimi obtained in digitalized form from Paul Whitehouse via the Summer Institute of Linguistics

Walker unpublished survey vocabulary of Mairasi of Sisir obtained in digitalized form from Paul Whitehouse via the Summer Institute of Linguistics

The close relationship between Mairasi and Semimi (Etna Bay) was first recognized by Anceaux (1956, 1958: 118,) and has not been controversial. The Mer language was included upon elicitation (Peckham 1991: 147.)

[under construction]

The external relationships of Etna Bay are unclear. … Anceaux (1956, 1958) … Greenberg (1971) … Voorhoeve (1975) … Wurm (1982) … Ross …

Proto-Etna Bay had 13 consonants and 5 vowels as follows:

Prenasalized laminal /*ns/ is uncommon.

Neither prenasalized stops /*mb *nd *ns *ŋg/ nor apical non-stop /*ɾ/ occur word-initially.

Neither final consonants nor consonant clusters occur.

Vowels may occur in any sequence except geminates. Vowel length is not phonemic. Observed sequences include:

All attestations below are drawn from Peckham (1991: 173-180,) with Mairasi forms also drawn from (1982 ibid., 1991 ibid.,) unless otherwise indicated. Forms drawn from Anceaux (1958, n.d.,) Greenberg (n.d.) after Anceaux or Voorhoeve (1975: 101) after Anceaux are indicated with [ja].

[under construction]

Bilabial nasal /*m/ is preserved as such in all three languages:

Apical nasal /*n/ is generally preserved as such in all three languages:

In segments of the form /*VnVk/, /*n/ is coalesced with following /*k/ to yield /ŋg/ in Mairasi (below.)

Some examples of initial segment /*na/ are interpreted as representing the third person singular possessor /*na-/ (below.) This might not be correct in all instances:

As geminate vowels do not occur, it's possible that some of these roots have an underlying intial vowel /*a/ which is not discernible in these constructions.

Bilabial fricative /*ɸ/ is preserved as such in all three languages. Word-medial /ɸ/ is lenited to voiced [β] in Mairasi (Peckham 1982: 92, 1991: 127) and Mer, and sometimes in Semimi:

One example is irregular, with Semimi initial /*ɸ/ answered by /mb/ [b] in both Mairasi and Mer in the manner of a reduplication:

Apical plain stop /*t/ is preserved as such in all three languages. Word-medial /t/ is optionally lenited to plain voiced [d] in Mer:

Laminal /*s/ is preserved as such in all three languages. Word-medial /s/ is optionally lenited to voiced [z] in Mer:

Velar plain stop /*k/ is reflected as glottal stop /ʔ/ [ʔ ø] in Mairasi (Peckham 1991: 152-153.) Word-medial /k/ is typically lenited to plain voiced [g] in Mer:

Where medial /*k/ is followed by high back rounded vowel /*u/, the segment /ʔu/ is realized as [ʔgu gu] in Mairasi:

In segments of the form /*VnVk/, /*k/ is coalesced with previous /*n/ to yield Mairasi /Vŋg/ (Peckham 1991: 152-153.) Both known examples are compounds:

Some examples of initial segment /*ka/ appear to be morphological in origin, with /*ka-/ designating an impersonal possessor, as in part-of-whole contructions, or the referent of an adjective. This /*ka-/ is dropped when a personal possessor (e.g. /*na-/ 3 sg. above) is present, or when the root appears as the second member of a compound:

This phenomenon is undocumented in the works of Peckham (1982, 1991a, b,) and there are a number of examples in which it's unclear whether /*ka/ is part of the root. Provisionally, we assume them to be separable, but this might not be correct in all instances:

As geminate vowels do not occur, it's possible that some of these roots have an underlying intial vowel /*a/ which is not discernible in constructions with /*ka-/ or third person singular /*na-/.

Bilabial prenasalized stop /*mb/ …:

Apical prenasalized stop /*nd/ …:

Laminal prenasalized /*ns/ …:

The term for “dog” is a loan from Indonesian (Peckham CITE.)

Velar prenasalized stop /*ŋg/ …:

… /*w/ …:

Apical non-stop /*ɾ/ is generally retained as such in all three languages. It is very common medially, but has not been observed to occur word-initially:

In a few instances, postthematic /*ɾ/ is dropped in Mer under unknown conditions:

… /*j/ …:

…

…

Etna Bay free pronouns are reconstructed as follows, with forms drawn from Peckham (1982: 75-76, 80-82, 1991: 178, 1991 ibid., Walker and Peckham n.d.) and Mairasi and Semimi inclusives from Anceaux:

(Peckham's Mer third persons are demonstrative phrases and are not included here.)

…

[under construction]

… possessive …:

In addition to these, there is an impersonal possessor /*ka-/ which appears on adjectives and in part-of-whole constructions (above.)

[under construction]

…