fIJIAN lANGUAGE
fIJIAN lANGUAGE
Fijian has a very rich system of reduplicative morphology which interacts in interesting and subtle ways with the stress system. The focus here is on the phonological
aspects of reduplication, not the semantic aspects, so don’t be too concerned about the glosses.
For the data in (1), please provide a formal statement of the reduplication process. In particular, argue both for and against the proposals discussed in class
regarding whether reduplication is best viewed as a copy of segments, copy of syllables, or is prosodically constrained by templates, also, whether reduplication is
sensitive to the prosodic structure of the base or not. Point out particular forms that are crucial to your argument. What is the nature of the reduplication process,
and how might we formalize it? Morpheme boundaries are indicated by dashes. (Some forms have been simplified)
(1)
bo}se-a confer about it }bose-bo}se-a confer about it at length
bo}so-k-a mix it }boso-bo}so-k-a mix it intensely
bo}ta-n-a patch it }bota-bo}ta-n-a patch it often
bu}ki-a fasten it }buki-bu}ki-a fasten it often
bu}ta: remove it (food) }buta-bu}ta: take it (food) out often
}vau-c-a tie it up }vau-}vau-c-a tie it up often
}ba: deny }ba:-}ba: deny often
}dre: it is pulled }dre:-}dre: it is difficult
}va: four }va:-}va: all four
}droka uncooked }droka-}droka green
}tsaka doing }tsaka-}tsaka working
ja}qona kava }jaqo-ja}qona a plant related to kava
vi}naka good }vina-vi}naka thanking
bu}tao steal }buta-bu}tao steal several times
The data in (2) show a variation on the pattern in (1). What is the difference? First provide a formal account of the reduplication process, again defending your
analysis with crucial data. Look in particular at the stress pattern in evidence here. Every phonological word has one and only one primary stress. What does this tell
you about the nature of the reduplication in (2) versus (1)? The notion of “minimal word” might prove helpful here. Recall that many languages possess a requirement
that content words must be two moras in length. Assuming that Fijian has such a requirement, how can we combine the stress facts and the reduplication facts into a
coherent analysis?
(2)
}la:i small la-}la:i a collection of small items
}leka short le-}leka a collection of short items
ba}lavu long ]ba-ba}lavu a collection of long items
bo}lomu steeped (food) ]bo-bo}lomo steep continually
tso}butu covered (with hand) ]tso-tso}butu cover continually
The data in (3) exemplify yet another pattern. If you solved (1) and (2), these data should not be problematical. Please provide full discussion.
(3)
ta-}basu broken (by itself) ]ta-ta-}basu-}basu (freq.)
ta-}bili go in (great no.) ]ta-ta-}bili-}bili (freq.)
tsa-}druti broken (rope+) ]tsa-tsa-}druti-}druti (freq.)
tsa-}lidi crackling noise ]tsa-tsa-}lidi-}lidi (freq)
tsa-}tsuvu explode ]tsa-tsa-}tsuvu-}tsuvu (freq)
ta-}buki knotted ]ta-ta-}buki-}buki (freq)
ka-}dala opened ]ka-ka-}dala-}dala (freq.)
tsa-}dri: (no gloss) ]tsa-tsa-}dri:-}dri: (freq.)
tsa-}qou (no gloss) ]tsa-tsa-}qou-}qou (freq.)
Fijian has a glide formation process that applies under certain conditions, but is blocked under others. although only a few forms are provided in (4). But given your
results so far, you should be able to establish the proper account of its behavior.
(4)
tu}ia hammer }tui-tu}ia hammer it a lot
}ilo (no gloss) }ilo-}ilo (*ilojlo) (no gloss)
}tawru (*ta.uru) become slack
(<ta-}uru)