Tomui River

Timothy Usher, Santa Fe Institute

The Tomui River family consists of four or five languages spoken both north and south of the Tomui river, a tributary of the Guam and Ramu rivers to the west, in the Middle Ramu district of Papua New Guinea's Madang province. Its nearest relatives are the Sogeram River languages spoken immediately to the south and east.

The internal classificaiton of Tomui River is as follows:

Tomui River

Utarmbung (Osum)

Anam (Pondoma)

Anamuxra

Ikundun

Moresada

Wadaginam

Z'graggen (1971: 61) gives lexicostatistical relationships between the languages of the Josephstaal Stock (excluding Katiati i.e. Mum which belongs with the Sogeram River family, Daniels 2010, 2015) as follows:

The impression is that of a chaining relationship from Osum in the south to Wagaginam in the north. It is likely that the Pondoma figures have been depressed somewhat by a number of loans from Katiati (Mum) of the Sogeram River family (below.)

A primary division between Osum (Utarmbung) and the other Tomui River languages might be supported by a number of isoglosses in which the Osum word is found also in the Sogeram River family, thus presumably descending from proto-Sogeram and Tomui Rivers, while the other languages have their own term. It is however possible that most of these are loans from Katiati (Mum) (below):

Capell (1951-1952: 201-203) 75 comparative terms for Murusapa

Z'graggen (1971: 72) 17 cognate series for Moresada, Ikundun, Pondoma, Wadaginam and Osum

Z'graggen (1980: 1-80) 330 comparative terms and (pp. 81-82) pronimnal paradigms for Osum, Pondoma, Ikundun, Moresada and Wadaginam

Ingram (2001) grammar of Anamuxra

Ingram (2003) Anamuxra nominal classifiers

[under construction]

Capell (1951-1952: 206) placed Murusapa (Moresada) with the Ramu River languages

Wurm (1969) placed Murusapa with his Western Family (Lower Ramu River) while stating that it belongs typologically to his Central Family, i.e. the Adelbert Range languages Bunabun (Brem) and Wanambre.

Z'graggen (1971, 56-61) … Josephstaal Stock … Pomoikan Family … included Katiati (Mum) …

Z'graggen (1975) … added the newly-documented Sileibi (Sirva) language to the Josephstaal Stock, forming with Katiati the Sikan Family

Greenberg (1971: …) …

Daniels (2010: …, 2015: …) removed Mum and Sirva from Z'graggen's Josephstaal Stock, placing them immedaitely with Apalɨ (Emerum) in the Central Sogeram subgroup of his Sogeram River family, an expansion of Z'graggen's (1969: 61-64) Wanang Stock

[under construction]

Proto-Tomui River had 17 consonants and 4 vowels as follows:

This inventory is very similar to the system reconstructed for proto-Sogeram River in Daniels (2010, 2015,) but with a few key differences. Daniels does not reconstruct a velar non-stop /*ɣ/. Daniels reconstructs a palatal nasal /*ɲ/, but most of the examples are consistent with the interpretation of [ɲ] as a nasalized allophone of non-stop /*j/ as presented here (below.) Daniels reconstructs the prenassalized stops as voiceless because they are devoiced in Mand and Nend and he assumes voiceing rather than devoicing to be the more likely direction of change.

There are also very strong phonological similarities to the Guam and Moam subgroups of Ramu River immediately to the west, as well as a number of common roots which are presumed to be loans in one direciton or the other (below.).

…

Only a restricted subset of consonants occurs initially:

Any consonant cna occur medially.

Only a restricted subset of consonants occurs finally.

Initial consonants correspond as follows:

Medial consonants correspond as follows:

Final consonants correspond as follows:

Vowels correspond as follows:

These correspondences are exemplified as follows, with Osum, Pondoma, Wadaginam, Moresada and Ikundun attestations drawn from Z'graggen (1980: 1-180) and Anamuxra from Ingram (2001.)

… initial consonants …

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

Initial apical nasal /*n/ is found nearly only in monosyllabes, where it is retained as such in all descendants; the sole example of a disyllable with initial /*n/ is probably a reduplication:

Velar nasal /*ŋ/ has not been found to occur initially.

Initial peripheral voiceless stops /*p *k/ are lenited or dropped in all descendants, a latter-day phenomenon shared with most descendants of Ramu River to the west; however apical and laminal /*t *s/ are not affected by lenition.

Initial bilabial voiceless stop /*p/ is lenited to voiced fricative /β/ in Anamuxra and typically to non-stop [w] elsewhere, sometimes making it difficult to distinguish it from non-stop /*w/ (below):

Root-initial /*p/ in verbs however is retained as such when it is preceded by a prefix (compare “bathe” above):

Similarly, the initials of /*pʉsa/ “skin” and /*pʉriN/ “side” are not lenited when they are preceded by another word in a compound:

One example in which apparent initial /*p/ does not undergoe lenition is mandatorily a loan:

Initial apical voiceless stop /*t/ is usually lenited to non-stop /r/ in Osum, which does not otherwise occur initially, although this appears to be in free variation with [t]:

Initial laminal voiceless /*s/ is merged with apical /t/ in Ikundun and varies between [t s] in Wadaginam:

:

Initial velar voiceless stop /*k/ …:

… /*k/ …:

Root-initial /*k/ in verbs however is retained as such when it is preceded by a prefix (cf. “watch (intr.)” above):

One example in which apparent initial /*k/ does not undergo lenition is mandatorily a loan:

Initial rounded velar voiceless stop /*kʷ/ …:

As in the closely-related Sogeram River family, prenasalized voiced stops /*mb *nd *nz *ŋg/ do not occur word-initially.

Intiial bilabial non-stop /*w/ is sporadically dropped in Osum and Pondoma:

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

Intiial palatal non-stop /*j/ …:

When followed by a nasal consonant or a prenasalized voiced stop in the onset of the next syllable, initial /*j/ is retained as such only in Moresada and cocasionally in Osum. Elsewhere it is nasalized to [ɲ] in Osum and merged with apical /n/ in Pondoma, Wadaginam and Ikundun:

… initial /*j/ …:

Initial velar non-stop [*ɣ] exists phonetically, but in most if not all instances is the initial allophone of /*k/ (above.)

… medial consonants

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

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

Medial velar nasal /*ŋ/ …:

Medial bilabial voiceless stop /*p/ is retained as such in all descendants:

Medial apical voiceless stop /*t/ …:

… /*T/ …:

Medial laminal voiceless /*s/ is merged with apical /t/ in Ikundun and varies between [t ts s] in Wadaginam:

Medial velar voiceless stop /*k/ is retained as such in all descendants:

Medial bilabial prenasalized voiced stop /*mb/ is devoiced and denasalized to /p/ in Osum:

Onr example in which [*mb] remains as such in Osum results from a compound with /*-pʉsa/ “skin”:

Medial apical prenasalized voiced stop /*nd/ is generally devoiced and denasalized to /t/ in Osum:

There are everal unexplained exceotions in which /*nd/ is retained as such in Osum, two of which strangely look to have irregular reflexes /nz/ in other Tomui River languages::

Some examples of medial [nd] arise from the suffixation of auxilliary verb /*tɨ/ “do/male” to a root with a final nasal consonant:

Medial laminal prenasalized voiced /*nz/ is devoiced and denasalized to /s/ in Osum and merged with apical /nd/ in Ikundun and Wadaginam:

Medial velar prenasalized voiced stop /*ŋg/ is generally devoiced and denasalized to /k/ in Osum:

… medial /*ŋg/ …:

Medial rounded velar prenasalized voiced stop /*ŋgʷ/ …:

Medial bilabial non-stop /*w'/ …:

Medial apical non-stop /*r/ …:

… /*r/ …:

… /*r/ …:

Medial palatal non-stop /*j/ …:

When /*j/ …

Medial velar non-stop /*ɣ/ …:

… final consonants …

…:

Final apical nasal /*n/ …:

Final velar nasal /*ŋ/ …:

Final apical non-stop /*r/… :

… /*r/… :

… /*r/… :

… vowels …

…

…

Proto-Tomui River pronouns had both subject forms, which consist of only the uninflected base, and objects which are suffixed with /*-ɣʉ/. Osum, Wadaginam and Moresada have collapsed these two paradigms, with Osum and Wadaginam using the object forms of the singulars but the subejt form of the plurals and Moresada using object forms form both numbers; hence the gaps in the paradigms which follow.

Tomui River subject pronouns are reconstructed as follows, with Osum, Pondoma, Wadaginam, Moresada and Ikundun attestations drawn from Z'graggen (1980: 81-82) and Anamuxra from Ingram (2001: 194-203.) Ingram's Anamuxra third person singular looks to have been replaced with the plural, although their associated prefixed forms remain distinct (below):

Tomui River object pronouns are reconstructed as follows:

While Z'graggen does not give dual forms, the originality of the Ingram's Anamuxra forms is supported by outcomparison to Sogeram River (Daniels 2017: 86.) The separation of plural and dual suffixes is supported by not only the brute force of analogy but the fact that the plural suffix /*-ŋ/ is in Anamuxra used also for ordinary nominals (Ingram 2001: 186-187.)

In addition to subjects and objects, alienable possessve forms are reconstructed as follows (the second /*-ka/ on some forms is a nominal classifier):

An alternative alienable possessive series is found in Osum and in Ingram's Anamuxra:

In addtion to the free forms, prefixes are used to indicate inalienable possessors: and verbal objects.

…inalienable possessors, which Z'graggen glosses as genitive … kin terms and body parts (Ingram 2001: ):

… verbal objects …::

As the underlying forms are identical to those of the free pronouns (above) we can conclude that prefixation is in both instances of recent origin.