Ruboni Range

Timothy Usher Santa Fe Institute

The Ruboni Range family consists of three languages spoken in the Ruboni hills, between the south coast of Hansa Bay in the east in the lower Ramu river to the west, inland of the northwest coast in the Bogia subdistrict of Papua New Guinea's Madang province (Z'graggen 1971: 76.) It is called the Misegian Family in the works of Z'graggen, after the names of its members, Mikarew, Sepen and Giri (Kire.)

The internal classification of Ruboni Range is as follows:

Ruboni Range

Mikarew

Akukem-Sepen

Kire (Giri)

Z'graggen (1971: 77) gives lexicostatistical relationships between the languages of the Misegian Family as follows:

The appearance is that of a chain, with Mikarew at one end and Kire at the other.

Schorr (n.d.) dictionary of Mikarew (unobtained)

Höltker (1961: 93-95) vocabulary and (pp. 96-97) counting system for Kire-Puir

Capell (1951-1952: 138-141) grammar sketch and (pp. 198-200) 73 comparative terms for Makarub (Mikarew)

Z'graggen (1969, 1971: 90-91) 13 comparative terms for Mikarew, Sepen (Akukem) and Giri (Kire) and (pp. 109-110) number-inflected nominals and (pp. 146-148, 162-164) 21 tense-inflected verb stems for Mikarew

Z'graggen (1972) 431 comparative terms for Mikarew, Sepen and Giri

Stanhope (1972) grammar of Kire including (pp. 53-57) 68 number-inflected nominals and (pp. 64-71) 396-term vocabulary and 119 tense-infected verb stems

Pryor (1981) phonology of Kire (unobtained)

Pryor and Clifton (1987) Kire nasalization

Foley (2005: 119-121) 61 comparative terms for Kire drawn from Stanhope (1972) and Mikarew from Capell (1951-1952) and Z'graggen (1971, 1972)

Daniels (2010) 376 terms for Akukem (digitalized by Russell Barlow)

In addition to these, Paul Whitehouse provided digitalized survey vocabularies of Mikarew and of Kumanang, Girian and Fuvura varieties of Kire via the Summer Institute of Linguistics Ukarumpa; however they are undated and unattributed.

[under construction]

Proto-Ruboni Range had 14 consonants and 6 vowels as follows:

In the first version of this reconstruction, the stop prosodies were based primarily upon Stanhope's (1972: 51-54) Kire, with aspirated voiceless stops <*pʰ *tʰ *kʰ> reflecting Ramu River voicelss stops /*p *t *k/ and plain (unaspirated) voiceless stops <*p *t *k> reflecting Ramu River prenasalized voiced /*mb *nd *ŋg/ in other Ramu River subfamilies.

Capell (1951-1952: 138-139) asserts that in Makarub there is no difference between voiceless and voiced stops: “The plosives are devoiced; i.e. intermediate in acoustic effect between voiced and voiceless. At first writing both series were used in the word-lists, but did not prove to have been written at all consistently, and a distinction was not phonemic. The same remarks apply to s for which z also was first used in some words.” However Z'graggen's representations of both Mikarew and Sepen (Akukem,) while not entirely consistent, correlate non-randomly with Stanhope's Kire values as do the Daniels' (2010) representations of Akukem. Accordingly, we posit that all three prosodies are synchronically present in each language, but are somewhat difficult for elicitors to distinguish.

Subsequent research however has shown that some traces of prenasalization remain discernible where a plain voiceless stop would usually be expected. Besides sporadic attestation in Daniels' Akukem, the first and second person duals consistently show it in the special environment where the prenasalized velar stop is the onset of a suffix upon the pronominal root. Thus, even though plain voiced stops are sychronically their overwhelmingly most common realization in all three languages, we now reconstruct prenasalized stops as such, resulting in a more archaic system.

While a number of roots have initial consonants in all three languages, these are all secondary developments due to the loss of a destressed initial vowel given below as /*ʉ/, since alone among vowels /*ʉ/ does not otherwise occur initially except in a very small number of originally monosyllabic roots (e.g. 1 sg..) In multisyllabic roots, there are only four genuinely initial consonants at the proto-Ruboni Range stage, each of which is lost in Kire:

Only a restricted set of consonants occurs finally:

A small number of apparent consonant clusters have been identified word-medially; these most probably indicate compounds.

Final vowels are not prohibited, but are nearly unattested following a postthematic consonant; that is, nominal roots of the form /*(C)VCVCV/ don't seem to exist, suggesting that vowels which hadd been in this position were lost

Initial consonants in monosyllables correspond as follows:

Initial consonants in multisyllbic words correspond as follows:

Subinitial consonants correspond as follows:

Medial consonants correspond as follows:

Final consonants correspond as follows, with Mikarew values assuming the presence of following article /-imu/ or attributive /-ir/:

Vowels correspond as follows:

These correspondences are exemplified as follows, with all attestions drawn from Z'graggen (1971, 1972,) Capell (1951-1952: 198-200,) Daniels (2010) and Stanhope (1972: 55-57, 64-71, ibid.) as indicated. A few of the comparisons presented below originate with Z'graggen (1971: 90-91.)

Only a handful of genuinelly monosyllabic roots have been identified as such.

Initial bilabial nasal /*m/ in a monosyllabe is retained as such in all three languages:

… /*ŋg/ …:

First and second person dual pronouns are suffixed with an monosyllabic element /*-ŋga/ leading to a correspondence which is distinct from that of initial /*ŋg/ (above) or medial /*ŋg/ (below):

Only four consonants occur initially in multisyllabic roots, bilabial nasal /*m/ and non-stops /*w *r *j/, all of which are dropped in Kire along with the following vowel.

Initial bilabial nasal /*m/ in multisyllabic roots is occluded to prenasalized voiced stop /mb/ in Akukem-Sepen and dropped in Kire along with the following vowel. Several examples where Z'graggen gives initial [m] as an alternative form suggest this process to be subphonemic/:

When initial /*m/ is followed by a nasal as the onset of the next syllable, occlusion in Akukem-Sepen is sporadic:

… /*m/ …:

Velar nasal /*ŋ/ has not been found to occur initially. Examples of initial /*ŋ/ in Ramu Coast have been dropped in Ruboni Range.

Initial bilabial non-stop /*w/ …:

Initial apical non-stop /*r/ is fortited to stop [t] in Mikarew, dropped in Kire and retained as such in Akukem-Sepen when followed by a vowel other than low central /*a/. When the following vowel is high front /*i/, Mikarew [t] is palatalized or affricated:

When followed by low central vowel /*a/, initial /*r/ is dropped in Akukem-Sepen as well as Kire:

Initial palatal non-stop /*j/ when followed by low central vowel /*a/ is occluded to apical voiced stop [d] in Mikarew and dropped in Akukem-Sepen and Kire:

Two examples are posited to have had initial /*j/ followed by high back rounded vowel /*u/ because while the initial vowels of all three languages are high front /i/ (which would usually be dropped in Kire) the following syllables have the onglide characteristic of initial /*u/ (below):

Subinitial consonants are those which which originally were medials of multisyllabic roots but have become initial in all three languages due to the mutual loss of destressed intiial high central vowel /*ʉ/. Their correspondences are distinct from those of genuinely initial consonants (above.)

Subinitial bilabial nasal /*m/ is retained as such in Mikarew and Akukem-Sepen and occluded to prenasalized stop /mb/ in Kire:

We assume that subinitial /*m/ followed by a nasal as the onset of the next syllable would not be occluded; however no examples of /*n in this environment have yet been found.

Subinitial apical nasal /*n/ is retained as such in Mikarew and Akukem-Sepen and occluded to prenasalized voiced stop /nd/ in Kire. This phenomenon is shared with the Guam River family to the south:

We assume that subinitial /*n/ followed by a nasal as the onset of the next syllable would not be occluded; however no examples of /*n in this environment have yet been found.

Subinitial bilabial voiceless stop /*p/ is fricated to /ɸ/ [f] in at least Mikarew and Akukem-Sepen and probably in all three languages:

Subinitial laminal voiceless /*s/ is maintained as such in all three languages:

No examples of subinitial velar voiceless stop /*k/ have yet been found.

Subinitial bilabial prenasalized voice stop /*mb/ …:

Subinitial apical prenasalized voiced stop /*nd/ …:

Subinitial velar prenasalized voiced stop /*ŋg/ …:

Subinitial bilabial non-stop /*w/ …:

Subinitial apical non-stop /*r/ ….:

No examples of subinitial palatal non-stop /*j/ have yet been found.

Subinitial velar non-stop /*ɣ/ is occluded to plain voiced stop [g] in all three languages when followed by a vowel other than high front /*i/ (below):

… medial consonants….

Medial bilabial nasal /*m/ is retained as such in all three languages, except where Kire has dropped an initial vowel (below):

When medial /*m/ in a disyllable becomes initial in Kire due to dropping of erstwhile initial vowel, /*m/ undergoes occlusion to /mb/ as does initial /*m/ in monosyllables (above,) unless it is followed by a nasal in the next syllable (below):

This occlusion does not occur if medial /*m/ is followed by a nasal in the next syllable:

One example in which /*m/ is expected to undergo occlusion in Kire …:

Medial apical nasal /*n/ is retained as such in all three languages, except where Kire has dropped an initial vowel (below):

When medial /*n/ in a disyllable becomes initial in Kire due to the loss of an initial vowel, /*n/ undergoes occlusion to prenasalized voiced stop /nd/ as does initial /*n/ in monosyllables (above) unless it is followed by a nasal in the next syllable (below):

This occlusion does not occur if /*n/ is followed by a nasal in the next syllable:

One exception to these rules remains unexplained and may be a compound:

Medial velar nasal /*ŋ/ in a multisyllabic root, like /*m *n/ in the same conditions (above,) is occluded to initial /ŋg/ in Kire unless followed by a nasal consonant in the onset of the next syllable (below):

When medial /*ŋ/ in a multisyllabic root is followed by a nasal consonant in the onset of the next syllable, it remains /ŋ/ in Kire:

… medial /*ŋ/ …:

Medial bilabial voiceless stop /*p/ … /ɸ/ [f] …:

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

Medial laminal voiceless /*s/ is maintained as such in all three languages:

Medial velar voiceless stop /*k/ …:

When medial /*kʰ is followed by high front vowel /*i/, it is merged with the reflexes of apical /*t/ in Mikarew and Akukem-Sepen:

Verb stems followed by present tense /*-i/ are likewise affected:

Medial bilabial prenssalized voiced stop /*mb/ …:

Medial apical prenasalized voiced stop /*nd/ …:

Medial velar prenasalized voiced stop /*ŋg/ … when followed by a vowel other than high front /*i/ (below):

When medial /*ŋg/ is followed by high front vowel /*i/, it is merged with the reflexes of apical /*nd/ in Mikarew and Akukem-Sepen:

Verb stems followed by present tense /*-i/ are likewise affected:

Medial bilabial non-stop /*w/ is typically occluded to [β v] in all three languages:

Medial apical non-stop /*r/ is retained as such in all three languages, except where initial vowel dropping has occured in Mikarew (below):

Where Mikarew has dropped an initial vowel,medial /*r/ is reflected as apical plain voiceless stop /t/, as is initial /*r/ (above):

Medial palatal non-stop /*j/ is most typically occluded to voiced fricative /z/ in all three descendants:

Several examples of apparent medial /*j/ both of which correspond to /*j/ in other subfamilies of Ramu River are occluded to /r/ in Kire under unkown conditions:

In one example where medial /*j/ is followed by bilabial non-stop /*w/ in the next syllable, Kire has dropped /*j/:

Where Mikarew has dropped an initial vowel, erstwhile medial /*j/ is reflected as /d/, as is initial /*j/ (above). The following example is one of several Kire nominals (q.v. Stanhope 1972: 55-57) which have /s/ in the singular and dual but /z/, reflecting medial /*j/, in the plural/ No explanation for this apparent irregularity is offered here:

Medial velar non-stop /*ɣ/ …:

… medial /*ɣ/ … Stanhope's (1972: 64-71) Kire vocabulary gives only six roots with /h/, which he characterizes as “roughly aspirated” (p. 53):

…medial /*ɣ/ … high front vowel /*i/ …:

When medial /*ɣ/ is followed by high back rounded vowel /*u/, it is merged with bilabial /β/ in Kire:

Only a restricted set of consonants occurs finally. This set includes nasals /*m *n *ŋ/, possibly archiphonemic stops /*p *t *k/, fricative /*s/ and non-stop /*r/.

Final bilabial nasal /*m/ is retained as such in Akukem-Sepen and Kire, but dropped in Mikarew before the nominal article /-imu/, which is then reflected as /-amu/. It is also dropped in Akukem-Sepen and Kire plurals:

Final apical nasal /*n/ is generally retained as such in all three languages. It is dropped in the Akukem-Sepen and Kire forms of the plural:

Several seeming exceptions to the retention of final /*n/ may be morphological in origin, perhaps indicating a singulative suffix /*-ʉn/ which appears on terms referring to items which are most typically found in masses (q.v. Stanhope 1972: 56):

Final velar nasal /*ŋ/ is sporadically dropped in Akukem-Sepen and Kire and invariably in Mikarew before the nominal article /-imu/, which is then reflected as /-amu/. In Akukem-Sepen and Kire its underlying presence is indicated by the lenghtening and nasalization of the preceding vowel (Pryor and Clifton 1987 ibid.):

Final bilabial voiceless stop /*p/ is retained as such in Akukem-Sepen and Kire and lenited to [β v w] in Mikarew before the nominal article /-imu/:

Final apical plain voiceless stop /*t/ when preceded by non-front vowels /*ʉ *a *ɔ *u/ is merged with the reflex of /*r/ in Mikarew andAkukem-Sepen:

When preceded by front vowels /*i *e/, final /*t/ is dropped in Mikarew before the nominal article /-imu/:

In Kire, finals /t r/ are distinct, but neither Z'graggen's nor Stanhope's attestations are consistent in this respect; for example there are a number of words given with final /r/ in Stanhope's general vocabulary (1972: 64-71) but with /t/ in his categorization of nominal plurals (pp. 55-57.) Since all examples with [t] are known to correspond to final /*t/ in other Ramu subfamilies, the variation between Kire [t r] is considered subphonemic. Both /*t *r/ correspond to the same sounds in Ramu Coast. There are a handful of roots which cannot be assigned to either /*t *r/ because for Kire is given only [r], and there are no known outcomparisons; this uncertainty is represented as /*[t/r]/.

Final laminal voiceless /*s/ is uncommon. It is retained as such in (at least) Mikarew :

Final velar voiceless stop /*k/ is merged with laminal /s/ in Mikarew when followed by the nominal article /-imu/. It is retained as such inAkukem-Sepen and Kire unless following low central vowel /*a/ (below):

When followed by attributive /-ir/ (q.v. Capell 1951-1952: 139, Z'graggen 1971: 109-110,) final /*k/ is merged with /s/ inAkukem-Sepen as well as in Mikarew:

When /*k/ is followed by a vowel-initial suffix, for example plural /-ɪ/ or singulative /-œn -ʊn/, it is lenited to voiced /g/ inAkukem-Sepen and Kire:

When final /*k/ follows low central vowel /*a/, it is reflected as glottal stop [ʔ] in Kire:

Final apical non-stop /*r/ is retained as such in all three languages when following non-front vowels /*ʉ *a * *u/:

When preceded by high front vowel /*i/, final /*r/, like final /*t/ (above,) is dropped in Mikarew before the nominal article /-imu/:

…vowels ….

High front vowel /*i/ …:

High back rounded vowel /*u/ …:

Most Ruboni Range roots begin with vowels, which are then very often lost, presumably due to thematic stress. This has progressed furthest in Kire, with Stanhope's Kire vocabulary showing seven instances of initial /ɪ/, but none of initials /ɛ œ ʊ/, while two examples of intial /a/ and one of /ɔ/ are interjections (q.v. 1972: 64-71):

…

…

…

…

…

…

If the vowel of the second syllable is high front /*i/ or low central /*a/, a rounded offlide [ʷ] is added to onset of the second syllable:

No rounded onset is added if the consonant of the onset of the second syllable is bilabial non-stop /*w/:

…

…

…

No examples of the pattern /*CeCV/ have been found.

…

No examples of the pattern /*CoCV/ have been found.

…

…

…