Pseudo-Austronesian Conlang

This conlang is inspired by the phonology and grammar of the Austronesian languages. I decided to create this to show how an Austronesian-like language could develop, from something that is less 'unusual', in terms of phonotactic restrictions, and unusual affixation patterns, such as the infixes that are common in Austronesian languages.

Consonant Inventory
	labial	coronal	palatal	velar	glottal
nasals	*m	*n		*ŋ
stops	*p	*t		*k	*ʔ
stops	*b	*d		*g
fricatives		*s
approximants		*ɾ	*j	*w

Monophthong Inventory
	front	central	back
close	*i		*u
mid	*e		*o
open		*a

Onset	Nucleus	Coda
Obligatory consonant	Vowel	Optional voiceless occlusive or voiced continuant

Weak vowel shift: unaccented high vowels are weakened to /ə/, while unaccented mid vowels /e o/ are raised to /i u/.
Non-head syllable reduction: all syllables except the last one (which is stressed) drop their codas.
Accent retraction: accent shifts leftward to the penult in polysyllabic words, unless that penultimate vowel is schwa, in which case final accentuation is retained.
Infixation: VC pre-head dependent morphemes are infixed to follow the rule of every syllable having an onset, and every unstressed syllable being open. However, before vowel initial stems, these morphemes remain prefixes.
Initial /ʔ/ epenthesis: Fulfilling the onset constraint, a glottal stop is epenthesised to ensure onsets; you can probably analyse this as occuring before infixation, as all prefixed infixes are preceded by a glottal stop. However, underlyingly, these glottal stops cannot be considered to be contrastive, and do not contribute to syllable weight or the no-coda requirement i.e. ʔakal~ʔakal > ʔakalakal vs. kal~kal > kakal.