Enggano language

Phonology

The only major linguistic treatment of Enggano was conducted by Hans Kähler in 1937; he published a grammar (1940), texts, and a dictionary (1987). However, phonology is limited to a simple inventory and a short paragraph of basic features; the grammar and dictionary disagree with each other, the dictionary is not consistent, some words are not legible, and doubts have been raised about the accuracy of the transcriptions. Nothofer (1992) discusses loanwords and also lists phonemes. Yoder (2011) is a thesis on Enggano vowels, with some comments on consonants; it will be followed here.

Stress was once reported to be penultimate but is now on the final syllable. Alternating syllables preceding it have secondary stress.

Yoder and Nothofer report seven oral and seven nasal vowels:

Diphthongs are /ai, aɨ, au, ei, ɘi, oi/.

Vowels do not occur word-initially in Enggano apart from what Yoder analyzes as /i u/ before another vowel; these are then pronounced as semivowels [j w]. (Nothofer counts these as consonants /j, w/ restricted to initial position, which avoids the problem of not uncommon [ji] being analyzed as /ii/, when sequences of the same vowel are otherwise quite rare.) The vowels /i ɨ u e o/ are all pronounced as semivowels in vowel sequences after medial glottal consonants /ʔ h/, as in /kõʔĩã/ [kõʔjã] (a sp. tree) and /bohoe/ [boho̯e] 'wild'; otherwise, apart from diphthongs, vowel sequences are disyllabic, as in /ʔa-piah/ [ʔapi.ah] 'to graze'. /i/ optionally triggers a glide after a following glottal consonant, as in /ki-ʔu/ [kiʔu ~ kiʔju] 'to say'. Diphthongs lower to [aɪ, aʊ] etc. before a coda stop, as in /kipaʔãũp/ [kĩpãʔãʊ̃p] 'ten', and undergo metathesis when that stop is glottal, as in /kahaiʔ kak/ [kahaʔɪkak] 'twenty'. An intrusive vowel [ə̆] appears between glottal stop and another consonant (though not semivowels), as in /kaʔhɨɘ/ [kaʔ.ə̆.hɨ.ɘ] 'female leader'; this does not affect the pattern of stress.

In many words, a final vowel transcribed by Kähler is not found in Yoder.

The offglide of diphthongs lowers before glottal consonants, and a glottal stop may intrude when another word follows, as in /kahaiʔ mɘh/ [kahaʔɪmɘ̃h] 'another'.

Yoder notes that the voiced stops [b~m, d~n] are in complementary distribution, depending on whether the word has nasal vowels, but lists them separately. Voiced oral consonants, [b d l r], do not occur in words with nasal consonants or vowels. Nasal consonants nasalize all vowels in a word, and there is therefore no contrast between [m n] and [b d] apart from the contrast between nasal and oral vowels. For example, with the oral stem tax 'bag', the possessive forms are tahi’ 'my bag' and tahib 'your bag', but with the nasal stem 'age', the forms are ’umunu’ 'my age' and ’umunum 'your age'.

/l/ occurs in only a few native words. /s ~ x/ are infrequent and apparently a single phoneme; they only occur word finally, where they contrast with /h/: [x] occurs after the non-front vowels /ɨ ə u/, [ç] after the front vowels /i a ã/, and [s] after vowel sequences ending in /i/ (including /ii, ui/). The resulting [aç ãç] may actually be /aix ãĩx/, as most such words are attested with alternation like [kaç ~ kais] 'box'. When a suffix is added, so that this consonant is no longer word-final, it becomes /h/, as in tahi’ 'my bag' above.

Nothofer is similar, but does not list the uncommon consonants /l/ and /s ~ x/ and counts [j w] as consonants rather than allophones of vowels. Kähler's dictionary adds /ɲ/, as well as /f tʃ dʒ/ as marginal phonemes, and claims that /t r/ are only found in southern villages. However, Yoder states that at the time of his research in 2010 there were no differences among the six villages on Enggano Island, and that initial /t r/ and final /t d/ are rare in native words. Medial /d/ and /r/ are in free variation in a few words, with older people preferring /d/ and younger speakers /r/.

Grammar

Independent and possessive pronouns in Yoder (2010) are,

Most of these appear to be Austronesian: Compare Malay 1sg aku ~ ku, 1.EX kami, 1.IN kita, 2pl kalian, 3sg/pl dia, and suffixes 1sg -ku, 2sg -mu, 3sg -nya, with *k, *t (d), *l, *m, *n having shifted to ’, k, r, b, d in Enggano, and with final consonants and (where possible) vowels being lost.

The possessive suffixes appear on nouns, and they are often preceded by a vowel. Few forms are attested, but this vowel is i or ai after [ç] (as with 'bag' in the phonology section), an echo vowel after several other consonants, and with several words not predictable on current evidence: ’eam – ’ami’ '(my) fishing rod' (Blench notes that ’e- appears on many nouns in Kähler and may be a prefix, perhaps a determiner; cf. ’ẽ’ 'this'), dar – daru’ '(my) husband', pi – pia’ '(my) garden'. In a couple cases, an intact pronoun -’u or -ki is appended.

Adjectives commonly have prefixes ka-, ka’-, ki-; the first two are attested in derivation, and the last is assumed as it is very common and many such adjectives otherwise appear to be reduplicated, as in kinanap 'smooth' (Yoder 2010).

Verbs may have one or two prefixes and sometimes a suffix. Attested prefixes are ba-, ba’-, ia-, iah-, ka-, ka’-, kah-, ki-, kir-, ko-, pa-, pah-, ’a-. The functions of these are unknown. Ki- and pa- may occur together, as in pe, pape, kipe, kipape, all glossed as 'give'. The three attested verbal suffixes are -i, -ar, -a’ (Yoder 2010).

The counting system is, or at least once was, vigesimal: Kähler recorded kahai'i ekaka 'one man' = 20, ariba ekaka 'five man' = 100, kahai'i edudodoka 'one our-body' = 400. (The last may be based on two people counting together: each time I count all twenty of my digits, you count one of yours, so that when you have counted all of your digits, the number is 20×20 = 400.) However, most people now use Malay numerals when speaking Enggano, especially for higher numbers. Yoder (2010) recorded the following:

1–5 are Austronesian, assuming ka- is a prefix on 'one' and ’a- is a prefix on 2–5. Compare the remaining -hai’, -ru, -kər, -up, -rib with Lampung əsay, rua, təlu, əpat, lima; *s, *t, *l, *m have shifted to h, k, r, b in Enggano, and final consonants and (simple) vowels have been lost. ’aki’akin 6 may be reduplication of ’akər 3. ’arib he ’aru 7 is 'five and two'. The two forms for 8 mean 'hugging', from the verb pã’ĩõp 'to hug', and 9 appears to be 'eight, one coming'; it may be shortened to kaba kahai’ (no -i) in enumeration. Yoder believes 10 may also be a verb, based on an unelicited root ’ãũp, as ki- and pa- are verbal prefixes (as in ki-pa-pe 'to give'); indeed, the apparent prefixes on 1–5 are identical to verbal prefixes as well.

Numbers above 10 and 20 are formed with he ~ hi 'and': kĩpã’ãũp he ’aru 'ten and two' for 12, kahai’ kak he kĩpã’ãũp 'twenty and ten' for 30. kak is 'person', so twenty is 'one person'. Multiples of twenty are formed from kak, as in ’akər kak he kĩpã’ãũp 70, ’arib kak 100 (also kahai’ ratuh from Malay ratus).

Classification

Based on the low number of apparent Austronesian cognates, Capell (1982) concludes that Enggano is a language isolate rather than Austronesian as previously assumed. Blench (2009) picks up the question, though he leaves Enggano unclassified. Subsequent material collected by Yoder (2011), however, suggests the language is Austronesian after all, albeit lexically divergent. Bak 'eye', for example, corresponds regularly with Malay mata, but 'die' (Yoder ba’a, Kähler ka’a, presumably a prefix ba- or ka- with a root ’a) has no apparent connection to Austronesian mati ~ matay; indeed, of the most stable verb and noun roots, only a third have reasonable Austronesian cognates.