Re: status of ka (was Re: [lojban] x3 of du'

[Rob Speer <rob@twcny.rr.com>]

On Thu, Aug 23, 2001 at 02:21:57PM +0100, And Rosta wrote:
> The arguments have mostly been about whether empty places are
> interpreted as containing ce'u or zo'e or either. There are pros to
> the ce'u option and pros to the zo'e option.
> 
> I therefore proposed that empty places in du'u contain zo'e and
> empty places in ka contain ce'u. There is now no ambiguity and
> everybody has the best of all worlds. 
> 
> It would invalidate some prior usage -- many former kas would need
> to be read as du'us. But language evolution always invalidates earlier usage.

I hardly see it as evolution to leave ka entirely unusable. It's already been
pointed out that nobody uses more than 2 ce'us - under this proposal, you'd
often have to use a lot of zo'e just to get the number of ce'us under 3.

In fact, this seems like it would invalidate _all_ prior usage, given that
either people use {ka} with no ce'us to mean {nu} or {du'u}, or they include
one {ce'u} and assume that the rest of the places will behave. Now you're
taking both of those usages and making them into functions of up to 5
variables, which I assume nobody using {ka} intended when they wrote it.

I support the approach that if there is no {ce'u} in a {ka}, then it is
implied. Relative clauses get by with implied {ke'a} just fine, although it is
of course clearer to include the {ke'a}. What makes {ce'u} different?
-- 
Rob Speer