South Awyu

Timothy Usher, Santa Fe Institute

The South Awyu family is comprised of two closely related languages, Bamgi River Awyu and Ia River Awyu, spoken by approximately 3,000 people (1975) living on either side of the Digul River near its delta in Indonesia's Papua province (Voorhoeve 1975: 375.) These are better known in the literature by their former river names, Sjiagha (Siagha, Syaxa) and Jenimu (Yenimu) respectively, following Drabbe (1950, 1954, 1959) (Lebold, Kriens and Susanto 2013: 24.)

Nevermann (1940: 195-196) 45 terms for Oser (Bamgi River)

Drabbe (1950) grammar and (pp. 131-142) 408 comparative terms for Sjìagha (Ia River) and Jénimu (Bamgi River)

Drabbe (1954: 231-255) 100 comparative terms for Sjìagha and Jénimu

Drabbe (1959: 62-181) 431 comparative terms for Sjìagha and Jénimu

Voorhoeve (1975: 375-376) brief general description of Syiagha

Lebold, Kriens and de Vries (2010: 46-52) 97 comparative terms for South Awyu of Salamepe village (Bamgi River)

Lebold, Kriens and Susanto (2013: 36-47) 239 comparative terms for South Awyu of Yeloba and Konebi (Jas) villages and (pp. 48-50) 60 comparative terms for Banamepe and Konebi (Homeme) (Bamgi River)

Wester (2014) Awyu-Dumut comparative morphology including (pp. 195-210) 430 comparative terms for Shiaxa and Yenimu after Drabbe (1959: 162-181)

Proto South Awyu had 14 or 17 consonants and 7 vowels as follows:

Any consonant besides prenasalized stops can occur initially or medially. Prenasalized stops /*mb *nd *ŋg/ occur only medially. Their phonemic status is dependent upon whether they are to be analyzed as clusters or as unit phonemes. Their phonological behavior parallels that of final nasal archiphoneme /*N/ in combination with a following voiced stop /*b *d *g/, suggesting cluster /*NC/ to be the proper analysis. However, this would introduce root-internal consonant clusters, which otherwise occur only in compounds, while in most instances there is no evidence that words with prenasalized stops were ever compounds, and some may be loans from Yaqay.

High vowels /*i *u/ do not occur initially. Excepting diphthongs, sequential vowels do not occur. Otherwise, any vowel can occur in any position.

In addition to the seven cardinal vowels, five diphthongs are found:

Stress appears to be contrastive, but is not consistently indicated in Drabbe's vocabularies.

Only a restricted set of consonants occurs root-finally in nominals:

Final nasal /*N/ is an archiphoneme resulting from the merger of final nasals /*m *n/. With a few exceptions, it is discernible only in Ia River when a root is followed by a suffix. Since verbs are mandatorily suffixed, this is the most common way for /*N/ to be identified.

On verb stems, this set is futher reduced to two archiphonemes /*N *C/, the second of these reflecting Central Digul River final stops /*p *t *k/.

Palatal non-stop [j] isn't recognized as a final consonant are because diphthongs /*ɛⁱ *ɔⁱ aⁱ/ are needed anyway to avoid consonant clusters, which don't otherwise occur, and because /*j/ would be the only final consonant that is not dropped in Bamgi River.

Initial and medial consonants correspond as follows:

Final consonants correspond as follows:

Vowels and diphthongs correspond as follows:

It's conceivable that more extensive phonological descriptions of Ia River and Bamgi River might demonstrate that high mid vowels /e̝ o̝/ and low mid /ɛ ɔ/ are still distinguished from one another in one or both languages. Drabbe (1950: 94) asserts such a distinction between /e ɛ/ <é e> in both languages; however the correlation of these to /*e̝ *ɛ/ is statistical at best, and the arbitrary distinction <é e>, including for languages in which this contrast is definitely not phonemic, is a recurring weakness in Drabbe's documentations generally..

These correspondences are exemplified as follows. Attestations below labeled Sjìagha and Jénimu are drawn from Drabbe (1950: 131-142, ibid., 1959: 162-181.) Salamepe village forms are drawn from Lebold, Kriens and de Vries (2010: 46-52) and Konebi and Yeloba from Lebold, Kriens and Susanto (2013: 36-47.)

… initial consonants ….

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

Initial apical nasal /*n/ is significantly less common than initial /*m/, and is disproportionately represented by pronominal and imperative forms. It is retained as such in both languages:

Initial bilabial voiceless stop /*p/ is rrelatively uncommon. It is retained as such in both languages:

Initial apical voiceless stop /*t/ is retained as such in both languages:

Initial velar voiceless stop /*k/ is retained as such in both languages:

Initial bilabial voiced stop /*b/ is retained as such in both languages:

Initial apical voiced stop /*d/ is substantially less common than peripherals /*b *g/, an assymetry shared by Digul River-Ok languages as a whole. It is retained as such in both languages:

Initial velar voiced stop /*g/ is retained as such in both languages:

Initial bilabial fricative /*ɸ/ [ɸ f] is retained as such in both languages:

Initial fricative /*s/ is retained as such in both languages:

When followed by high front vowel /*i/, /*s/ is realized as [ʃ tʃ] in Ia River:

Initial velar fricative /*x/ is retained as such in both languages:

Initial bilabial non-stop /*w/ is retained as such in both languages:

Initial apical non-stop /*ɾ/ is retained as such in both languages:

Initial palatal non-stop /*j/ is generally retained as such in both languages:

Medial bilabial nasal /*m/ is retained as such in both languages:

Medial apical nasal /*n/ is not rare relative to medial /*m/ as found in initials, especially when it's considered that a number of examples of /*m/ above historically reflex the auxilliary verb /*mɔ/. It is retained as such in both languages:

Medial labial voiceless stop /*p/ is uncommon. It is retained as such in both languages:

Medial apical voiceless stop /*t/ is retained as such in both languages:

Medial velar voiceless stop /*k/ is retained as such in both languages:

Medial bilabial voiced stop /*b/ is retained as such in both languages:

Medial apical voiced stop /*d/ is retained as such in both languages:

Medial velar voiced stop /*g/ is retained as such in both languages:

Medial prenasalized stops /*mb *nd *ŋg/ are merged with plain voiced stops /b d g/ in Bamgi River:

Medial bilabial fricative /*ɸ/ [ɸ f] is retained as such in both languages:

In two examples, for unknown reasons, Bamgi River /ɸ/ is answered by Ia River /w/. We assume this sound to have been /*ɸ/ rather than /*w/ or a new consonant because “enemy” appears to be a loan from Yaqay /koˈpae/ id. (q.v. Drabbe 1954: 135); no explanation is offered for this discrepancy:

Medial /*s/ is retained as such in both languages:

In one example, Bamgi River reflects /*s/ as /t/ for unkown reasons; comparison to central Awyu suggests /*s/ to be original:

Another shows the opposite pattern:

Medial velar fricative /*x/ is realized as voiced [ɣ] in both languages:

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

Medial apical non-stop /*ɾ/ is retained as such in both languages:

Medial palatal non-stop /*j/ is retained as such in both languages:

Final archiphonemic nasal /*N/ is invariably lost in Bamgi River; this parallels the behavior of prenasalized stops (above.) There are three ways in which final /*N/ can be discerned in Ia River. First, it sporadically surfaces as [n ŋ] when a root is given in isolation:

Final /*N/ is realized as Ia River [m] when a root is suffixed with /*bɔge/ “bone”:

In a few examples, compounding shows /*N/ to have been present where it is not given in the isolated form:

The most common way for /*N/ to be identified is in verbs, which are mandatorily suffixed. Drabbe's citation forms with /*-d/ [d nd] indicate the first person singular, the second and third person being marked with /*g/ [g ŋg] according to the same prenasalzation rule (q.v. Drabbe 1950: 111-114):

Unfortunately, no verbs are given in Lebold, Kriens and de Vries (2010: 46-52.)

Final apical voiceless stop /*t/ is appended with a copy vowel in Ia River and dropped in Bamgi River:

Final velar fricative /*x/ is appended with a copy vowel in Ia River and dropped in Bamgi River:

One idiosyncratic example is tentatively reconstructed with final /*x/ because it appears to originate as a loan from Marind /*maɣuk/:

Final apical non-stop /*ɾ/ is dropped in Bamgi River:

Final /*ɾ/ is lost in both languages when followed by a consonant in a compound or phrase:

While regular within South Awyu, the retention of proto-Awyu final consonants other than /*N/ in verbs is not predictable. For example, the following doublet:

High front vowel /*i/ is retained as such in both languages:

High back vowel /*u/ is retained as such in both languages:

High mid front vowel /*e̝/ is reflected as /e/ in Ia River and most typically as [i] in Bamgi River:

In one example, /*e̝/ is reflected as /o/ in Ia River for unknown reasons:

High mid back vowel /*o̝/ is reflected as /o/ in Ia River and as /u/ in Bamgi River:

Low mid front vowel /*ɛ/ is generally reflected as /e/ [e ɛ] in both languages:

In roots of the form /*wɛCi/, Ia River backs and raises /*ɛ/ to /u/:

Roots of the form /*CɛCa/ are reflected as /CaCa/ in Bamgi River:

Low mid back vowel /*ɔ/ is generally reflected as /o/ in both languages:

Ia River raises /*ɔ/ to /u/ when followed by /u/ in the next syllable:

Ia River raises /*ɔ/ to /u/ when following /*i *u *o̝/ in the previous syllable:

One apparent exception is historically a compound with /*-ɾɔ/ “seed”:

Roots of the form /*CɔCa/ are reflected as /CaCa/ in Bamgi River:

An apparent exception is historically a compound of /*bɔn/ “mouth” and /*xa/ “skin”:

Low central vowel /*a/ is retained as such in both languages:

Diphthongs /*uⁱ *ɛⁱ *ɔⁱ *aⁱ *aᵉ/ are uncommon: