Kapauri-Sause

Timothy Usher, Santa Fe Institute

The Kapauri-Sause family consists of two languages, Kapauri (Kapaori, Kapori) and Sause, spoken along and north of the middle Taritatu (Idenburg) river in the Airu and Nimboran districts of Jayapura regency in Indonesia's Papua province (q.v. Menanti and Rumaropen 2009: 3.)

The internal classification of Kapauri-Sause is as follows:

Kapauri-Sause

Kapauri

Sause

Voorhoeve (1975: 115) 23 comparative terms for Sause after Anceaux and 38 comparative terms for Kapori after Bromley (n.d.)

Smits and Voorhoeve (1994: 18-266) 121 comparative terms for Sause of Guai village after Anceaux (n.d.)

de Vries (1976) 175 comparative terms for Sause of Sause village

Walker (1978) 110 comparative terms for Sause of Sause and Ures villages

Barr and Walker (1978: 11-14) 110 comparative terms for Sause of Sause and Ures villages after Walker (1978)

Dommel (1985) 175 comparative terms for of Sause of Sebum/Lereh village

Wambaliau (2008) 260 comparative terms for Sause of Sebum and Ures villages

Menanti and Rumaropen (2009: 23-30 ) 260 comparative terms for Kapauri of Pagai village

Voorhoeve (1975a: 45) placed Kapori (Kapauri) and Sause alongside the Kaure Family, consisting at that time of Kaure and Narau, in a Kaure Stock. The Kaure Stock was further asserted to constitute a Subphylum-level Stock within the much larger Trans-New Guinea Phylum (1975b: 422-423,) a classification followed by Wurm (1982: 197) who added the recently-documented Kosare language to the Kaure Family. No arguments were presented in support of this proposal save for what is visible in the accompanying 40-term wordlists (1975a: 114-115.) Wurm characterizes the degree of relationship between Kapori, Sause and the Kaure Family as “on the low stock level” in contrast to the “lower medium to high family level” relationship uniting Kaure, Narau and Kosare.

Voorhoeve (1975b: 423) states, “Outside the stock, the closest relative of the Kaure language seems to be Kwerba of the Dani-Kwerba Stock.” As formulated, this is certainly untrue, however was likely meant to read as something closer to Wurm's assertion that the Kaure Stock's (as opposed to the Kaure language's) “strongest [lexical links] are with Kwerba of the Dani-Kwerba Stock.” The Dani-Kwerba Stock was Voorhoeve's (1975a: 35-38, 1975b: 400-406, Wurm 1982: 149-151) proposal to unite Dani (Balim Valley) on the one hand with Kwerba, Samarokena and Saberi (Isirawa) constituting a northern division on the other.

In recent years, Voorhoeve's classification has been modified or abandoned by most Papuan taxonomists. Silzer and Heikkinen (1984: 31) retain Kapori (Kapauri) in Voorhoeve's Kaure stock but (p. 41) move Sause into Voorhoeve's Tor-Lakes Plains stock, a conflation of Orya-Tor River with the unrelated Lakes Plains, most likely based upon to the vocabularies of Barr and Walker (1978: 11-14) as referenced which show a number of loans between Sause and Orya; this classification is reiterated in Silzer and Heikkinen-Clouse (1991: 28-29.)

Hammarström (2010: 188, 191-192) considers both Kapauri and Sause to be isolates. While Foley (2017: 469- 472) tentatively lists Sause as a member of the Tor (Orya-Tor River) family within the largr Kwerba-Tor family, following a brief lexical comparison of Oyra-Tor languages he writes (p. 471-472,) “On the basis of these data it seems unlikely that Sause is a Tor language at all…it also has no obvious cognates with the languages of the Kwerba family…I conclude that Sause is probably yet another isolate, in accord with the skepticism expressed in Hammarström (2010b.) This is not too surprising as Sause is spoken away from the main area of the Tor languages, sandwiched between Foau,a member of the Lakes Plain family, and Kapauri, an aberrant member of the Kaure family that is also probably an isolate.”

Of Voorhoeve's (1975) Kaure Stock, we find that Kapauri and Sause form one group and Kaure (and its Narau dialect) and Kosare another we designate as Nawa River, with Kapauri-Sause being related most closely to Masep, Isirawa and Apauwar River (i.e. Kwerba and Samarokena) and having no special relationship to either Nawa River or to Balim Valley (Dani.) The fundamental division between Kapauri-Sause and Nawa River is supported by Dommel and Dommel (1999: 1-2,) who counted only 1% lexical similarity between Sause of Badrun village and Kaure, and by Wambaliau (2006: 12-13) and Menanti and Rumaropen (2009: 13-14,) who found only 5-6% lexical similarity between Kapauri and Kosare, most of which can be attributed to loans (below.) Unfortunately, none of the SIL researchers are known to have compared Kapauri with Sause or Kaure with Kosare, but a fair accounting of the resemblances within each of these two pairs would yield a much higher figure, as demonstrated in the sections which follow.

[under construction]

Proto-Kapauri-Sause had at least 10 consonants and 7 or 8 vowels as follows:

In addition to the simple vowels given above, at least three or four diphthongs are found which are tentatively identified as follows:

Kapaur-Sause's consonant inventory is conspiuous for its gaps. Some of these might be filled upon further reconstrucrion from more extensive vocabularies.

It is not clear that there is either a voiceless biilabial stop /*p/ or any corresponding fricative. Bilabial voicless fricative [ɸ f] exists in both languages, but in some instances reflects initial /*hu *huw/ and in others results from a loan as well as being an allophone of Kapauri's regular reflex [p] of medial medial /*b/.

There is no apical voiced stop /*d/, suggesting non-stop /*ɽ/ to be its systemic equivalent; this is the case also in the neighboring Nawa River family.

It is not clear that there is a velar voiced stop /*g/, although words with /g/ exist in both languages.

Sause medial glottal stop [ʔ] is interpreted as a feature of synchronic vowel length /ː/ as it is found primarily in conjunction with /*aː/ or with other long vowels or diphthongs derived from sequences /*VwV/. Glottal stops on other positions are considered to be non-phonemic.

Root-finally, two and perhaps three archiphonemes are found; at least the first two are dropped in Kapauri

The same archiphonemes are found in Sause when a root-final segment /*-CV/ has been reduced to /C/ due to loos of a final vowel.

Phonotactically, Kapauri and Sause have undergone developments analagous to those which differentiate Kamula-Elevala River on the one hand from Awyu-Dumut and Awbono-Bayono on the other within the Digul River family (q.v. Suter and Usher 2017: 127-129.) In both instances, antecedent roots fall into one of three phonotactic classes: 1) monosyllabes /*(C)V/ which continue as such in both descendants 2) roots with final consonants which are dropped in Kapauri and Kamula-Elevala River and are reduced to archiphonemes in Sause, Awyyu-Dumut and Awbono-Bayono, 3) multisyllabic roots with final vowels which lose final vowels in Sause as well as Awyu-Dumut and Awbono-Bayono with concomitent reduction of now-final consonants to archiphonemes. This is a purely historical typological parallel, as Kapauri-Sause is not geographically adjacent to any of these families, but areally it is evident that Sause has been influenced by the Nawa River languages which have likewise reduced root-final consonants to archiphonemes.

In addition to these three classes, there is an uncommon fourth class in which apparent final vowels are retained in both Kapauri and Sause; howver the factors conditioning this retention, supposing that the vowels really were final as opposed to being followed by a consonant which has been dropped in both languages, are unkmown:

Thus Kapauri has no final consonants at all, while Sause roots are not exclusively but predominantly consonant-final.

There are no unambiguous consonant clusters except those which have apical non-stop /*ɽ/ as the second member. Only /*mɽ/ ,/*bɽ/ and /*tɽ/ have thusfar been reconstructed for Kapauri-Sauise but other such clusters exist in the daughter languages were likely allowed.

Sequences of multiple vowels do not occur unless some of our diphthongs or medial /*w *j/ are to be reinterpreted as their reflexes.

Initial consonants correspond as follows:

Medial consonants correspond as follows:

Subfinal consonants correspond as follows:

Final consonants correspond as follows:

Vowels and diphthongs correspond as follows:

These correspondences are exemplified as follows, with Kapauri attestations drawn from … and Sause from ….

Probably any consonant can occur initially.

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

Initial apical nasal /*n/ is generally retained as such in both languages, though Barr and Walker's (1978: 11) Ures [ⁿdaʔod] “path” show occlusion as is found regularly in the Nawa River languges to the east, presumably reflecting the influence of Kaure:

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

Initial laminal voiceless /*s/ is retained as such in Kapauri and deoccluded to /h/ in Sause, sometimes heard as zero in Barr and Walker's attestations:

Initial velar voiceless stop /*k/ is retained as such in Kapauri and seemingly lost entirely in Sause:

Intiial biabial voiced stop /*b/ is retained as such in both languages:

IHere we assume that a segment /*buw/ has been reduced to [b] in Kapauri because a similar phenomenon of coalescence affects segments /*hu *huw/ (below); this would require that Kapauri be the source of Kaure /mbaɺe/ (Dommel, Dommel, Auri and Pokoko 1991) (below) rather than vice-versa:

Intiial voiceless unoccluded /*h/ is generally retained as such in both subgroups, though sometimes heard as zero in Sause:

When initial /*h/ is followed by high back rounded vowel /*u/, the segment [hu hw] can be realized as bilabial fricative [ɸ] in both Kapauri and in Sause. Other Kapauri examples of this alternation are Menanti and Rumaropen's [ɸɽinɪ] vs. Barr and Walker's [hur̃ini] “worm” and Menanti and Rumaropen's [fʊkani] vs. Barr and Walker's [hogwane] “eye”. Voorhoeve (1975: 15) follwing Bromley gives the root below as [hwini] and cognates are found in other Foja Range languages with initial /*s/:

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

One example shows initial /*w/ occluded to [g] in Sause, suggesting a doublet loaned into Sause from Apauwar River or Orya-Tor where the reflex /ʷgʷ/ is regular (although this Foja Range root is not attested in Orya-Tor.) The vowel alternation is likewise unexpeted:

The existence of initial apical non-stop /*ɽ/ is based upon one comparison in which Sause /ɽ/ is initial but in Kapauri is preceded by a vowel, a phenomenon found also in Telefol of the Mountain Ok family (q.v. Healey 1964: 133-167.) Since vowels have not been observed to be dropped in Sause when followed by any other consonant, it is tentatively assumed that /*ɽ/ in this instance was initial:

Intiial palatal non-stop /*j/ is occluded to voiced fricative /z/ in Sause and sometimes to africate [dʒ] in Kapauri. Sause /z/ appears to contrast with the medial allophone of /*j/ (below):

Occlusion of initial /j/ occurs also in this loan from Orya (Uria) into Sause:

Probably any consonant can appear root-medially except for unoccluded fricative /*h/, a restriction found also in the neighboring Nawa River family.

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

Medial apical nasal /*n/ is generally retained as such in both languages:

In one example, medial /*n/ is palatalized presumably due to preceing high front vowel /*i/ (alternately this word is a coompund; cf. Masep [wine] (Erickson.) etc.):

…/*n₂/ …

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

…:

Medial laminal voiceless /*s/ …:

Two examples in which apparent medial /*s/ is deoccluded to /h/ in Sause are reduplications::

Another example is a compound with the base “two”::

Medial velar voiceless stop /*k/ is lenited to voiced [g] in Bar rand Walker's Kapauri and is probably deoccluded to glottal stop /ʔ/ in Sause, though it's not clear that Sause [ʔ] is a reflex of /*k/ rather than the subphonemic concomitant of vowel length in monosyllables (below) and the only unambiguous example of glottal stop is a reduplication:

Medial bilabial voiced stop /*b/ is devoiced to /p/ [p f] in Kapauri:

… /*b₂/ …:

Unoccluded fricative /*h/ has not been observed to occur medially, which is true also of the Nawa River family to the east.

Medial bilabial non-stop /*w/ when following high mid central vowel /*ɵ̝/ is retained as such in both languages:

In several examples where medial /*w/ is followed by low central vowel /*a/, the sequence /*wa/ is backed to [o u] in Sause with spradic lengthening and interpolation of glottal stop[ [ʔ]

There is however one exception, possibly indicating a comound (cf. Maspe [ˈkɔnɔ] (Donohue,) Isirawa [ono] (Oguri)):

Medial apical non-stop /*ɽ/ …:

Medial palatal non-stop /*j/ …

… /*j/ …:

Subfinal consonants are medials which appear in final segments /CV/ in Kapauri multisyllabic roots but are reduced to final consonants or archiphonemes in Sause.

Subfinal /*mV/ is reduced to final nasal archiphoneme /N/ in Sause:

Subfinal /*nV/ is reduced to final nasal archiphoneme /N/ in Sause:

Subfinal /*n₂V/ is reduced to final apical non-stop /ɽ/ in Sause:

No examples of subfinal /*tV/ have thusfar been identified.

Subfinal /*sV/ is reduced to final laminal voiceless /s/ in Sause:

One probable example of /subfinal *kV/ has been found which is reduced to final /k/ in Sause. It is tentatively assumed here that a cluster /sk/ has been reduced to /k/ [k g] in Kapauri:

No examples of subfinal /*bV/ have thusfar been identified.

No examples of subfinal /*hV/ have thusfar been identified. and probably will not be found as there is no indication that unoccluded fricative /*h/ occured medially.

No examples of subfinal /*wV/ have thusfar been identified.

Subfinal /*ɽV/ is reduced to final /ɽ/ in Sause:

No examples of subfinal /*jV/ have thusfar been identified.

A minority of multisyllabic words retain their final vowels in Sause under conditions which are not currently understood:

As is the case in the Nawa River languages to the east, there are two final archiphonemic consonants, /*N/ and /C/, and perhaps also apical non-stop /*ɽ/.

Final nasal archiphoneme /*N/ is realized as a velar nasal [ŋ] in Sause and lost in Kapauri:

Final stop archiphoneme /*C/ is realized as velar voiceless stop [k kʰ] in Sause and lost in Kapauri. This is reminiscent of the situation in the neighboring Kaure language, where final <k> is a grapheme for glottal stop /ʔ/ (Dommel and Dommel 1991: 51.)

One Sause word with outcomparisons in other Foja Range families suggests tha there was also a final apical non-stop /*ɽ/ by the hypothesis that long vowel /*aː/ is found only in monosyllables:

Kapauri-Sause had at least seven and possibly eight vowels as well as at least three diphthongs. Kapauri looks to have an unusually simple inventory of only four vowels /i u æ a/ with realizations of /u/ being especailly variable. Sause is more conservative in this respectm preserving all of Kapauri-Sause's original distinctions.

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

In one example, /*i/ becomes back rounded /u/ in Kapauri, presumably under the influence of neighboring /*u/:

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

High mid front vowel /*e̝/ …:

High mid central rounded vowel /*ɵ̝/ …

… /*ɵ̝/ …

High mid back rounded vowel /*o̝/ in simple environments or in combination with /*o̝/ in adjacent syllabes is reflected as high back [u] in Kapauri:

A number of roots have a final vowel which is usually given as [u ʊ ʉ y] in Menanti and Rumaropen's Kapauri but as [o] in Barr and Walker's and is lost in Sause; these are somewhat arbitarily reconstructed here with /*o̝/:

Low mid front vowel /*ɛ/ …:

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

One seeming exception may be an error in Wambaliau's Sause as Dommel gives [anaŋ] while Anceaux gives [aneŋ]:

… /*a/ …

… /*a/ …

Long low central vowel /*aː/ is found in at least two Sause words with clear outcomprisons to other Foja Range families. If both of these roots were monosyllables [*CaːC] in Kapauri-Sause, then [aː] may not need to be phonemic; however there is one possibly counterexample in “eat” (above.) The subphonemic interpolation of glottal stop [ʔ] in long [aː] is found also in some languages of the Gondu River and Ndu families as well as in Waran (Banaro) and is also found in Sause words where long [oː] reflects /*VwV/ (above):

At least three or four diphthongs have been found; however it is difficult to be certain of their qualities as each is supported only by a single example.

Mid front rising diphthong /*eⁱ/ …

Low front rising diphthong /*aⁱ/ …

Low back rounded rising diphthong /*aᵘ/ … Another Kapuari example of this alternation is Menanti and Rumaropen's [uvuny ~ ufunʉ] vs. Barr and Walker's [aufunu] “skin/bark”, where Dommel, Dommel, Auri and Pokoko;s(1991) Kaure /áhóˈɺîʔ/ confirms the presence of [a]:

Long low back rising diphthong /*aːᵘ/ … It is conceivable that this developed from an earlier sequence /*VwV/; cf. Isirawa [awa] (Oguri):

Kapauri-Sause free pronouns are as follows, with Kapauri attestations drawn from Menanti and Rumaropen (2009: 24) and Barr and Walker (1978: 13.) and Sause from Wambaliau (2008: 2) and Barr and Walker (1978: 13.) Only Kapauri-Sause singular pronouns are recontructable from internal evidence alone. The originality of Sause's first person plural is based upon outcomparison to other Foja Range languages. We assume that Kapauri-Sause did not distinguish inclusive from exclusive first person plurals:

No information about Kapauri-Sause verbal morphology is currently available to us.

[under construction]

Kapauri and Sause have borrowed words from several neighboring languages, notably Kaure and Kosare of the unrelated Nawa River family to the east and the distantly related Orya language o9f the Orya-Tor River family to the northeast, which has greatly influenced Sause in particular. In both instances these loans motivted incorrect classifications, first Voorhoeve's (1975 : 45, Wurm 1982: 197) placement with Kaure and Kosare in a Kaure Stock and later Silzer and Heikkinen's (1984): 31 placement of Sause. with Orya-Tor River.

Of the following Nawa River roots, only “knowledge” can be independently reconstructed for Kapauri-Sause; the others we assume to be loans from Nawa River on distributional grounds:

Other words were probably loaned into Kaure or Kosare from Kapauri-Sause, with Kaure attestations drawn from Dommel, Dommel, Auri and Pokoko (1991) and Kosare from Wambaliau (2006: 24-30:)

…Kapauri and Kosare …:

… Kapauri and Kaure …:

… Sause and Kaure … According to Barr and Walker (1978: 6, q.v. p. 10) both Sause and Kaure are spoken in the village of Bogogo. A few of the Sause words which follow include segments which do not exist in native roots (e.g. initial /k/, medial /h/.) The word for “black cockatoo” is found also in Orya (below):

No loans involving only Sause and Kosare have been found.

… Sause and Orya … According to Barr and Walker (1978: 6, q.v. p. 10,) both Sause and Warpu (Orya) are spoken in the village of Witi… According to Fields (1991: 30,) Sause and Orya share 10% cognacy in a list of 200 common words. None of the words in the chart which follows are known to occur in Kapauri, while several of them contain segments which do not occur in native words (e.g. initial [nd d], medial [h]):

That some of these were borrowed quite recently may be eeen in the word for “two” for which Anceaux (Smits and Voorhoeve 1994: 211-213) gives [nembe ~ nembre], cognate to the Kapauri form, where all of the more recent vocabularies give [dan ~ daŋ].

Still more loans between can be found in Anceaux's vocabularies of Uria (Orya) and Sause, which are said to have been elicited in the late 1950s (Smits and Voorhoeve 1994: 18-266)::

Several examples are loans into Orya rather than vice-versa. For example, Orya's reflex of Foja Range initial /*mb/ in “moon” is irregular for Orya, where /h/ is expected (cf. Tor River /*fɛⁱn/.) but regular for Kapauri-Sause, while “ear” and “tail” show developments peculiar to Sause as opposed to Kapauri and are otherwise not found in Orya-Tor River:

…:

One word which is phonologically abberant in both languages may have been borrowed from some third souce:

…: