RE: [lojban] li'i (was: Another stab at a Record on ce'u

["And Rosta" <a.rosta@dtn.ntl.com>]

pc:
> a.rosta@dtn.ntl.com writes:
>
>   When others want to say {X se li'i ce'u broda}, I want it to be {X se li'i
>   X broda}. In the most generalizable solution, the second X would be an
>   anaphor whose antecedent/binder is the first X, the experiencer. I couldn't
>   find any anaphor that would do the job, so proposed {no'au}, which works
>   like no'a but applies to all types of phrase, not just bridi.
>
> Now, can X have an experience of brodaing in general, not of something
> brodaing.  I guess I don't think so and so find {li'i ce'u broda} not to make
> sense. Must that something be X?  Clearly not, but that is an especially
> common case, I would think.  So the first temptation is surely to leave the
> first place of {broda} bare in that case -- and this is almost certainly what
> happened in the little bit of use {li'i} has gotten over the years.  Popping
> that up the {ce'u}, on the analogy of {ka}, or to {zo'e}, on the analogy of
> {du'u}, seem equally misguided.  Being explicit is, we now know from the
> toehr cases, the best policy, so we need "X" there or its anaphor.  Would
> {ri} worik in most cases?  Creating a new class of this situation (are there
> going to be others?) seems excessive.

{ri} would usually not work, since x2 of li'i normally follows the abstraction
and is often elided. I'm not arguing for li'i constituting a new class of
abstraction.

--And.