4.2. Utilitarismo y justicia
4.2.3. La justicia y su fundamento: la sensación de justicia
South Binandere is defined by changes involving the phonemes * wand *a.
The labial glide * w is reflected in the South Binandere languages as either w or a voiced velar obstruent, varying unpredictably across lexical items. The velar obstruent is reflected as g in the Orokaiva dialects and as y in the Coastal Binandere languages.3S The two obstruent reflexes are assumed to be the result of a sound change in Proto South Binandere as follows:
C41: * w > y l_ a ,o{Proto South Binandere)
20 pSBin *wauro 'bam boo’ > BBa, TBa, Kor yauro
366 pNucBin *wawa ‘sitting platform ’ > Baruga, Kor yaya
376 pBin *wa ‘rain’ > Not ya
417 pNucBin * worn ‘sheath’ > MBa yoru
60 pNucBin *sowa ‘breadfruit’ > Kor soya
177 pBin *atowo ‘father-in-law’ > Kar atoyo
295 pBin *wowora ‘loose’ > Kor yoyosa
See also:
49 pBin * saw- ‘boil’ 76 pBin * wamba ‘cane’ 77 pBin * wa(N) ‘canoe’ 79 pBin *orows ‘canoe pole’ 98 pBin *ow/a.o}w3 ‘chin, jaw ’
111 pBin *bu- ‘com e’
137 pNucBin *bunduwa 'disc-headed club' 147 pSBin *wasa 'd ry '
181 pBin *puwa ‘fence’ 187 pBin *pow /i,oj ‘fingernail’ 190 pBin *wo ‘fish, meat' 241 pBin *ji- ‘hold’ 253 pCstBin *iw oi ‘hungry’
258 pBin * warns ‘internal organ, sw eet/faf
303 pBin *ewa ‘m ango’ 346 pBin * wanda ‘arm ’ 349 pBin *jewa ‘pandanus’ 351 pCstBin * waito ‘pandanus mat' 355 pNucBin *ganewa ‘parrot’ 483 pSBin * iw oi ‘sun’
484 pBin * wacko ‘sun’ 513 pBin *di- ‘tie’ 524 pBin *iwo 'turtle' 554 pBin *wabu 'w ing' 559 pBin *siw3 ‘w orm ’ 560 pNucBin * aw- ‘w rap’ 562 pBin *ew- ‘w rap’
569 pNucBin *a-wa sedo ‘therefore (conjunction)’
Irregularly affected. 238 pBin *mbowu ‘heavy’
Correspondingly in Orokaiva (where the starred segment represents an intermediate stage):
38 The three sources for Yega show only < g > but this cannot be relied upon as phonetically faithful; it is therefore assumed that Yega shares the spirantised form y with the other Coastal Binandere languages. Cf. Yega in the discussion of Coastal Binandere (§2.11) below.
C42: * y > g l_ a ,o (Proto Orokaiva) 20 pSBin * wauro ‘bamboo' > Dob gauro
366 pNucBin *wawa ‘sitting platform" > Sos, Hun gaga
376 pBin * wa ‘rain’ > Aek ga
417 pNucBin * worn ‘sheath' > Sos goru
60 pNucBin *sowa ‘breadfruit’ > Sos oga
177 pBin *atowo ‘father-in-law’ > J-S atogo
295 pBin * wowora Moose' > Sos gogora
Items 79, 98, and 559 do not display the eligible environment at the Proto Binandere stage but would have displayed an eligible conditioning o at the Proto Orokaiva stage (following C48; see §2.15.6). The vocalic environments apply also when constituting the first element in a diphthong.
There is disagreement among the languages as to which eligible forms have undergone the sound change and which not. There is also a significant number of eligible forms which have not undergone the sound change in any represented South Binandere lect (e.g. 68 pBin *kowatu ‘brother-in-law of woman’, 224 pNucBin *kauwa
‘green/stale’). There are perhaps two possible explanations for this haphazardness.
The first is that in Proto South Binandere, allophony arose for a subset of lexical items in the specified phonological environment (the likely candidates for which are the forms which have an obstruent reflex in all represented lects), so that for some forms both [w] and [y] (the latter strengthened to [g] in Proto Orokaiva) were possible pronunciations of the segment. The change continued, drawing more lexical forms into its sphere, but this happened as speech communities were breaking up. Proto Coastal Binandere and Proto Orokaiva inherited the original subset as well as the change-in-progress and so could add to the original subset, but due to the splitting into separate speech communities, how newly-formed groups added to the subset differed from one group to another.
One would therefore assume a case of free variation between two allophones in a lexical subset which began at the Proto South Binandere node and flowed down, drawing more items along the way. In any language, it would be possible for one optional free allophone to be dropped in favour of the other, and the direction of that collapse could also be lexically conditioned.
The other explanation would be that the change was simply a change-in-progress during the break-up into speech communities and resulting lects mixed with and influenced one other, shifting or unshifting various words (cf. C20 and C33 in the discussion of the Baruga dialects in §2.12). This would account for the disagreement among lects as to whether the reflex for a particular form is w or gly, and also for the presence of a significant number of forms which appear not to have undergone the shift in
any represented lect even though they exhibit the eligible environment. (In relation to this, see §2.16 for discussion of Labov (2007) on transmission and diffusion.)
Several forms have the correct environment for changes C41 and C42 and are thus reconstructed with *w even though the phoneme is uncertain in them due to lack of representation from languages 01-06 - or variation among languages 07-15 - to prove the original segment was * w(20, 147, 351, 483).
Related to this sound change is a tendency for labial plosives to shift to velar articulation in the environment of adjacent rounded vowels. This occurs in the following items:
144 pBin *eupu 'drum ’ > 496 pBin *pumb- ‘take/carry’ > 261 pCstBin *kaubo 'knee' > 159 pBin *umbug3 'elbow ' >
Bin euku
Bin küban\ Amb, Not kumbari
Gen kauijgwo. Bar kwogo, BBa kougwo, MBa kougo
TBa ügiigo
In the case of the last item (159), the metathesis of the consonants which has occurred in both Yekora and Coastal Binandere reflexes may be related to this tendency as well, e.g.
*umbuga > umbuga (C44; see §2.15.6) > utjguba (metathesis placing velar consonant articulation between rounded vowels) in Yekora.
Ohala (1993:242) describes some confusions which occur due to the acoustic similarity of sounds that are very different in articulation, exemplifying the substitution of labial (or labialised) velars by labials which is well known to have occurred in the development of Classical Greek from Proto Indo European and many other unrelated languages (e.g. Indo European *ekwös > Classical Greek hippos ‘horse’, Proto Muskogean *uNk“i > Choctaw umbi ‘pawpaw’). He refers to a study by Winitz et al. (1972), which studied listeners’ identifications of CV syllables (formed from the set C = [p t k], V = [i a u]), heard under a variety of conditions. This reports confusions between the syllables [ku] and [pu] (where one assumes [k] was phonetically labialised before rounded [u]) as being among the highest of any of the CVs included.
The sound change that has occurred to the low vowel *a is problematic in that it cuts across subgrouping boundaries: it is exclusive to South Binandere languages but omits pockets of it, namely, Gaena, Karoto, and Korafe. The change is as follows:
C50: *a- > e- /_CM (Proto South Binandere?)
65 pBin * ami 'breast' > Aek, Orokaiva, Hun, Not, Yeg, Gen, Baruga, Dog em i
71 pBin *aw- 'bum ' > Hun, Not, Gen ev-
105 pBin * ami ‘club’ > Orokaiva, Hun Yeg emi
438 pBin * aw- 'sleep' > Aek, Orokaiva, Hun, Not, Yeg, Gen, TBa ev-lew-
533 pBin *asi(N) ‘vine, rope, string' > Aek, Dob, Sos, Hun, Not, Yeg, Gen esi\ Baruga, Dog eri
Irregularly affected. 142 pBin *aturo 'dream'
Irregularly unaffected. 454 pBin *aslsi ‘spirit’
This rule has applied to some verbs (71, 438) but not to others (133, 434, 560); in their infinitival form verbs are affixed in these languages with -art, which would supply the appropriate conditioning environment for the sound change.
The distribution of this sound change requires explanation in light of the Binanderean family tree, on which Gaena and Korafe (and Karoto) together form an exclusive branch (see Fig. 2.1). C50 (*a- > e- /_CM) calls into question the Gaena-Korafe node’s place within both South Binandere and its daughter group Coastal Binandere. Gaena and Korafe do, however, share in the other changes particular to South Binandere and Coastal Binandere (C7 *p > <p, C8 *g > y in certain environments, C41 *w > y /_a,o). A reversal of C50 is not plausible, as this would result in an absence of word-initial cases of
c in Gaena and Korafe, but such is not the case.
Gaena-Korafe could be excluded from the South Binandere node and recast directly beneath Proto Nuclear Binandere, with the effects of C7, C8, and C41 in the two languages accounted for by contact with Coastal Binandere languages. Alternatively, leaving Gaena-Korafe in its place on the tree, it could be hypothesised that both the old and the new forms of the lexical items affected by C50 were in free variation as late as the time of Proto Gaena-Korafe, at which node the older rather than the newer forms came to be exclusively used.
On the downside for South Binandere, only 20-odd lexical items are reconstructed to the level of pSBin, and no morpho-syntactic features have been identified which support the South Binandere languages as a subgroup (see §4.4). South Binandere may then be yet another instance of a dialect linkage.