[YG Conlang Archives] > [jboske group] > messages [Date Index] [Thread Index] >
John: > And Rosta scripsit: > > > > > Why then would the antecedent be "la djan" in "le la djan mamta ri"? > > > > Isn't "le la djan mamta" the first complete sumti? > > > > > > No. Look at the grammar and you'll see why. "la djan." is a > > > complete sumti > > > > Anyway, is "le la djan mamta" not a complete sumti? If it is, > > how come it is "la djan" that is the *first* complete sumti? > > The idea is that ri is coreferential with the *rightmost* complete sumti > which precedes it, where sumti are ordered by their *leftmost* words > Thus "le la djan mamta" is left of "la djan", and so "la djan" is the > rightmost complete sumti > > Arguably this definition violates a universal (it ignores nesting) but > it is the definition The rule is clear and doesn't seem particularly unnatural to me (relative to the average degree of unnaturalness in Lojban and in Lojban anaphora in particular). --And.