Re: more sources of opacity-like phenomena

[Jorge Llambias <jorge@PHYAST.PITT.EDU>]

la djan
> The current grammar allows both "NA <tense> <selbri>" and
> "<selbri> NA tense";

I suppose that was meant to be <tense> NA <selbri>

> the negation paper says that there are no known
> differences between these two forms, but that is remote from the
> discussion of "-roi", which is very brief.  The selbri paper says there
> is no difference, period; the negation paper allows that there might be,
> but its nature is not known.  Perhaps the relative scope of negation and
> tense should be, in fact, determined by order in this case (or indeed in
> every case, but indistinguishably most of the time).

That sounds reasonable.

>  However, the doubtful case is:
>
> 4)      ci lo cukta cu reroi se tcidu mi
>         some-three of-those-that-are books two-times are read by-me
>
> On the view that selbri-attached tenses have bridi scope, this means the
> same as Example 1; on the view that they have scope only from where they
> are, then this means the same as Example 2.  Making selbri tags have
> bridi scope has a certain appeal, but then Example 4 is different from:
>
> 5)      ci lo cukta reroi ku se tcidu mi
>
> where the "reroi" isn't part of the selbri but is free-floating (and happens
> to be just in front of the selbri). This result is unpleasing.

Why? It's exactly the same thing that happens with negation.

        ci lo cukta na se tcidu mi

is different from:

        ci lo cukta naku se tcidu mi

> I think
> I have to hold that selbri-attached tenses don't have bridi scope after all.

No, no, please reconsider!  :)

Jorge