la, lai, me

["And Rosta" <a.rosta@lycos.co.uk>]

It seems to me that there is a substantive difference between
la and lai only when the name is applied to a group, exactly
parallel to le/lei. For example, I can refer distributively
to a group of dogs as {le gerku}, or collectively as {lei
gerku}, or, if I name the group {gerk}, then distributively
as {la gerk} and collectively as {lai gerk}.

This has a couple of interesting implications.

The first is that la is sensitive to scope relative to le
in the way that le is -- i.e. a way that is nonvacuous only
when the sumti refers to a nonsingleton group.

The second is that 

{ko'a me la brod} 

must mean 

{ro da poi ke'a cmima lai/la'i brod zo'u ko'a me da}

(which is unlikely to be something one would want to say).

While

{ko'a na me la brod} 

must mean 

{na ku ro da poi ke'a cmima lai/la'i brod zo'u ko'a me da}

which is even less likely to be something one would want to
say.

I conclude from all of this that {lai} is far less likely 
than {la} to result in one inadvertently saying something one
didn't mean, and this is especially the case after {me}.

--And.