Re: PLI: evidentials in reported speech

["Mark E. Shoulson" <shoulson@CS.COLUMBIA.EDU>]

>Date:         Wed, 18 Dec 1996 13:41:03 GMT
>From: Don Wiggins <dwiggins@BFSEC.BT.CO.UK>
>
>I am not trying to bring fo'a's opinion in, but the jufra:
>
>    .i ko'a cusku lesedu'u   pe'i     fo'a   ca'o   klama   lo  zarci
>        He  said   that<-evidential-my-opinion...
>
>I take to mean that it is my opinion that the particular thing he was saying
>was...  But, it is not my opinion, but his.  Hence, the "dai".  Indeed, if
>further we want fo'a's opinion is that "pe'idaidai"?

This is different from lu/li'u quotes.  I would understand from the above
sentence that it's your opinion, not ko'a's, but if you said

ko'a cusku lu pe'i fo'a ca'o klama li'u

would to me mean that it's ko'a's opinion, not necessarily yours.  lu/li'u
take in everything that was said, it's all part of the quote, it's all part
of the reported speech (except sa'a-marked things).  At least, that's what
I thought.

As to pe'idaidai, it won't do what you want, I think.  dai is not so
mechanically defined.  It just means that the emotion in question is felt
more as "empathy" with someone else, and not the speaker.  Who is that
someone else?  There's no clear answer.  I often wish there were a simple
way to say a UI really belongs to someone else (but then again, where do I
get off expressing someone else's feelings, except when quoting?  That's
why "dai" is really "empathy": I'm feeling the emotion of empathy in which
I fancy I feel that X is experiencing this or that emotion.  The trouble is
there's no way to say who X is).

>>For "mi" to express mi's feelings, the attitudinal needs to be
>>outside of the quotes either in actuality or in some editorial parens
>>(to'i/toi?) with sa'a to unquote those (as I recall, lu/li'u quotes
>>contain, semantically, ALL their contents except sa'a, which can be used to
>>unquote other stuff).
>
>"to'i" and "sa'a" are interesting suggestion.  Let me ponder awhile.

I thought that was what they were for.

~mark