Sio language

The low back vowel is pronounced [ɔ]. All vowels vary in length, but length is rarely contrastive. Monosyllabic nouns and adjectives tend to be lengthened more than monosyllabic verbs, adverbs, or prepositions. Word stress generally falls on the penultimate syllable.

When Stephen and Dawn Clark of SIL International began to work with Sio speakers in 1985, the latter expressed a desire to revise their orthography to make it more similar to what people had become familiar with in Tok Pisin and English. The eventual results are tabulated in the following chart. The community at first resisted writing the labialized consonants as digraphs, since they clearly regarded them as unit phonemes. They insisted on writing the labialization as superscripts rather than as separate segments. However, by 1992, after many materials were produced in the new orthography, Sio teachers and church circuit officers approved writing the indicator of labialization on the same line, thus accepting mw instead of mʷ (Clark 1993).

The first orthography of Sio was devised by the missionary Michael Stolz, based on that of the Kâte language, which the German Lutheran mission used as a church and school lingua franca among speakers of Papuan languages. (Sio appears to have been assigned to the wrong language circuit.) The linguist Otto Dempwolff served as mentor and adviser to all the German missionaries in New Guinea on language questions. After Stolz died, Dempwolff analyzed his language materials and compiled a short sketch (1936). His analysis differs in several key respects from that of Clark (1993), who has had firsthand experience with the language. The most striking difference pertains to the labiovelars, which Dempwolff analyzed as coarticulated [k͡p], [ɡ͡b], [mɡ͡b], [ŋ͡m], but which Clark finds to be labialized labials (rounded on release) [pʷ], [bʷ], [mbʷ], [mʷ]. (The boldfaced q in the table below here stands for a curly q with hooked serifs that cannot properly be rendered online.) But Clark also found that g- [ɡ] and -c- [ɣ] were positional variants of the same phoneme; that trilled [r] is just a conditioned variant of flapped [ɾ]; and that the approximants are conditioned variants of their corresponding vowels.