Re: [lojban] Re: what's a du'u?

[Daniel Brockman <dbrockman@gmail.com>]

2009/12/3 Jorge Llambías <jjllambias@gmail.com>:
> On Thu, Dec 3, 2009 at 4:19 PM, Daniel Brockman <dbrockman@gmail.com> wrote:
>> 2009/12/3 Jorge Llambías <jjllambias@gmail.com>:
>>>
>>> See also:
>>> http://www.wiw.org/~jkominek/lojban/9511/msg00557.html
>>
>> That is very interesting.  Do you support the bridi1 = su'u1 position?
>
> Yes. Either that, or accept that "bridi" has two meanings, the second
> one being the purely syntactical one "x1 (text) is a bridi consisting
> of a selbri x2 (text) inserted among sequence of sumti x3 (each of
> which is a text)". Something like this is what "bridi" is used for in
> English when discussing Lojban syntax.

Interesting alternative.  I hadn't even thought of that possibility.

Assuming that we do want separate terms, though, and that we want
{bridi} to be about the predication rather than the text, --- do you
have any thoughts or opinions about how we could, would, or should
refer to the parts of speech?  I sketched up a kind of ad-hoc system
based on -pau --- any comments on that?

> The mixed version, where x1 is a text but has arguments x3 that are
> not text doesn't really make a lot of sense.

Agreed.

> Many of the other Lojban grammar words have similar problems too.

Hmm, {sumti} has a very analoguous problem.  The gimste says (even
more explicitly than in the case of {bridi}) that sumti1 and sumti2
are both text.  Would you prefer sumti1 to be lo cmima be bridi3 and
sumti2 to be bridi2?  I would, I think (for symmetry with {bridi}, if
nothing else).

On the other hand, the word words --- like {gismu}, {cmavo}, {lujvo},
etc. --- don't really have this problem.  At least not to this severe
degree.

What other grammar words do we have?

-- 
Daniel Brockman
daniel@brockman.se