[YG Conlang Archives] > [jboske group] > messages [Date Index] [Thread Index] >
On 3/26/07, Arnt Richard Johansen <arj@hidden.email> wrote:
da'i is a UI, so it should follow the same left-influencing behaviour as any other free modifier, right?
Yes (with the caveat for "left-influencing" that if the word to the left heads a construct, then the UI influences that construct, which will be to the right of the UI).
Now, consider what is going on here, in the CLL's sole example of da'i: ganai da'i do viska le mi citno mensi gi ju'o do djuno le du'u ri pazvau | | GA: ga | | NAI: nai | UI: da'i |- sentence | | GI: gi | UI: ju'o |- sentence See where the da'i ends up in the tree?
It should modify the first "sentence", and ju'o the second one.
Now, even assuming that the semantic effect of da'i can reach back through both the NAIClause and the gek (which is quite a stretch), what is it supposed to be doing to the gekSentence? I thought it was the _antecedent_ that was supposed to be marked as counterfactual, not the implication in its entirety!
I think da'i is indeed marking the antecedent, since GANAI and GI are
at the same level, so GANAI couldn't really be taken as heading the whole
gek-sentence.
But I don't really understand what difference da'i and da'inai are meant
to signal in the CLL examples. They seem to be a ba/pu distinction rather
than a da'i/da'inai one:
If you were to see my younger sister,
you would certainly know she is pregnant.
If you saw my younger sister,
you would certainly know she is pregnant.
mu'o mi'e xorxes