[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [lojban] Speaker specificity: {.i da'i na vajni}

la .and. cu cusku di'e

On 29 Sep 2014 01:40, "Pierre Abbat" <phma@bezitopo.org
<mailto:phma@bezitopo.org>> wrote:
 >
 > On Sunday, September 28, 2014 12:03:18 selpa'i wrote:
 > > As for old-{voi}, I agree it's utterly useless.
 >
 > I don't think so. It has its place, though I don't see much use for
it. {lo
 > rangutano cu me lo remsmismani voi se kerfa lo xunre bunre} allows an
ape to
 > still be an orangutan even if it's albino.

On further consideration, I agree it's not useless. But
You can find a possible use case for anything if you try hard enough. In practice, however, I have never needed {voi}, and nobody else uses it more than once per year. Not being a primitive, in the rare cases that it is needed one can always use a full form involving a predicate. (see below) 


(1) For every relative, noi, poi, ne, pe, no'u, po'u (that list from
memory -- hope it's right) a nonveridical counterpart would be at least
as useful as voi--poi is.
To me this just shows that "non-veridicality" should just be a predicate (from which you can define a UI or tag if necessary), so that you don't have to make non-veridical versions of everything.
If there is a predicate (currently {skicu} seems to come closest) then neither {voi}, nor any of the hypothetical non-veridical counterparts of the GOI you mentioned is necessary. 


(2) It's nonveridical noi that would figure in a logical expansion of le
and English definite descriptions, so is a candidate for usefulest.
I think the whole notion of veridicality and non-veridicality is overstated. Human speech tends to be metaphorical in general, so I don't think something like that should be a feature of a gadri, because that would limit it to sumti. A predicate, UI or tag on the other hand can be used anywhere.
For the definite description nature of {le}, which is its main purpose, non-veridicality is irrelevant, and it would more likely to be defined in terms of quantifiers in a formal logic. 


Selpa'i has suggested a UI for (non)veridicality. The snag with that is
that you'd not want the unmarked default to be "unspecified
veridicality", and usually you'd want the unmarked default to be
Veridical, but sometimes, specifically when the phrase has an
identificatory function, you want the unmarked default to be
Nonveridical. So this leads me to think that maybe better than a UI
would be a nonveridical poi'i, or maybe a nonveridical poi'i that
includes co'e in its meaning. You could even use /voi/ for that, tho not
at the cost of depriving poi'i of a shorter allomorph.

I'm kind of brainstorming here, not presenting a decidedly optimal change.
I think a predicate ({skicu} and related words like {simsa} and {simlu}) is all that is needed to capture non-veridicality.
Instead of {lo broda voi brode} you can always say {lo broda noi/poi mi skicu ke'a fo lo ka brode} and {noi simlu lo ka brode}. As this is something that is relatively rarely needed, it doesn't matter that it doesn't have a shortcut cmavo. But that's just my opinion. 

mi'e la selpa'i mu'o

--
You received this message because you are subscribed to the Google Groups "lojban" group.
To unsubscribe from this group and stop receiving emails from it, send an email to lojban+unsubscribe@googlegroups.com.
To post to this group, send email to lojban@googlegroups.com.
Visit this group at http://groups.google.com/group/lojban.
For more options, visit https://groups.google.com/d/optout.