Inrect questions are a problem for English (and familiar language) speakers because (since at least I-E days) one set of pronouns get used in a variety of ways. So "He asked who the murderer was" involves a clear indirect question, a transform from "He asked, 'Who is the murderer?'", but "He knows who the murderer is" seems only remotely related to a question, perhaps as "He know the correct answer to the question 'Who is the murderer?'". But then *"He believes who the murder is" (OK in Lojban, of course: ko'a krinu lo du'u makau morgau) could only mean "He believes some answer to the question "Who is the murderer?'" ("He believes he knows ..."). And when we get to
something like "They differ in what they wear", the connection to questions has virtually disappeared ('They give different answers to the
question 'What do you wear?'"?!) What they wear is in Lojban just lo se dasni (or, in the example, maybe te), similar to "what he said" in "He said who the murder was, but I didn't hear what he said". In this last case, even the first is pretty clearly not about a question but rather an assertion with the crucial part left out ("He said '... is the murderer'") All of this got shifted to Lojban (culturally neutral, ma derriere) as an indirect question, rather than the several things it might reasonably be -- relatives of {ce'u} with matter of focus thrown in mainly. All that being said, {makau} seems to work in place, despite the incongruities in the semantics. Of course, it would be somewhat easier if the use of {ce'u} were not limited to abstraction, so that we could say {lo se dasni be ce'u} without throwing in the irrelevant property marker and fictional indirect question. On the other hand, the separation is
sometimes needed, since (in English) "He knows who the murderer is" is quite different from "He know the murderer" (something someone always says in closed suspect list mysteries).
On a different topic slightly, the raising of sumti in many cognitive and phatic predicates is a terrible mistake, since it opens violations of opacity right and left, since things mentioned in propositions or properties or what have you need not be in the world in which the propositions, etc. are referred to, but sumti raising puts them there: from "John mistakenly believes there are unicorns" one gets in two steps to "There are some things that John mistakenly believes there to be".
On Sat, Feb 2, 2013 at 12:46 PM, Jacob Errington <
nictytan@gmail.com> wrote:
>
> {djuno2} is actually a ka, but we pretend that it isn't because the gimste
> made its definition clumsy to use that way. Indeed, the djuno2 is a property
> of the djuno3.
We could say that "djuno" is a sumti raising predicate, in the
linguistic sense of "raising"
http://en.wikipedia.org/wiki/Raising_(linguistics) since its x3 is the
semantic argument of an embedded predicate. The same can be said of
almost all predicates that take a ka-argument. (There are a couple of
oddball predicates that the gi'uste says should take ka but are not in
this category.)
mu'o mi'e xorxes
--
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?hl=en.
For more options, visit
https://groups.google.com/groups/opt_out.