Syllable Structure
More than 90% of the native lexicon is based on disyllabic root morphemes, with small percentages of monosyllabic and trisyllabic roots (Adelaar 1992). However, with widespread prefixing and suffixing, many words of five or more syllables are found.
In the native lexis, syllable structure is C1VC2, where both C1 and C2 are optional and C1 can be any consonant (though /w/ and /j/ occur word-initially only in one or two morphemes). In morpheme-final syllables, C2 can be any consonant except the laminals /tʃ, dʒ, ɲ/ or the voiced plosives. In non-final syllables in the native lexis, C2 is usually either a nasal (homorganic with a following obstruent, except that /ŋ/ precedes /s/, as in bangsa /baŋsa/ 'ethnic group'), or /r/ before any consonant except /h/, /w/ or /j/, for example in bersih /bərsih/ 'clean', terbang /tərbaŋ/ 'fly', and bernas /bərnas/ 'fertile'.
In loanwords, obstruents and other sonorants also appear in non-final C2 positions: akhbar /axbar ~ akbar/ 'newspaper' (from Arabic), saudara /sawdara/ 'brother' (Sanskrit), hairan /hajran/ 'amazed' (Arabic).
A wide range of consonant clusters occurs in the native lexis across morpheme boundaries before suffixes, as in kuatkan /kuatkan/ 'strengthen' and sampaikan /sampajkan/ 'deliver', where –kan is a verbal suffix. Initial clusters occur at the phonetic level only, as the result of either optional ellipsis of /ə/, as in bersetuju /bərsətuju/ [bɾ̩studʒʊ] ‘agree’, or the optional reduction of /u/ to [w] or /i/ to [j] before a following vowel, as in kuat /kuat/ [kwat] 'strong' of siapa /siapa/ [sjapa] 'who'.