Date: Thu, 27 Nov 1997 12:09:40 -0500 (EST) Message-Id: <199711271709.MAA29865@locke.ccil.org> Reply-To: And Rosta Sender: Lojban list From: And Rosta Organization: University of Central Lancashire Subject: Re: Indirect questions X-To: LOJBAN@cuvmb.cc.columbia.edu To: John Cowan Status: OR X-Mozilla-Status: 0011 Content-Length: 6090 X-From-Space-Date: Thu Nov 27 12:09:43 1997 X-From-Space-Address: LOJBAN@CUVMB.CC.COLUMBIA.EDU Jorge: > >On reflection, I now think that it makes more sense for Zo, > >lu, zoi, etc. to denote text-types. This is because zo/lu/zoi > >are not selbri valsi. Since there may be many tokens of a > >given text-type, reference to a text-token should be by means > >of a selbri (e.g. "le nu cusku zo coi" - 'an utterance of the > >text-type _coi_'). > > Sounds reasonable. Then {le nu cusku re zo coi} would not > be meaningful. I think that if zo is a text-type, that should > really be {le nu reroi cusku zo coi} or {lei re nu cusku zo coi}. > The first one (re zo coi) would be wrong, right? Yes. > > le nu mi cusku lu go'i li'u cu danfu le nu cusku lu xu do badri > > li'u > > My saying "I am" is an answer to someone's saying "Are you sad?". > > Right. Or for short: > > tu'a lu go'i li'u cu danfu tu'a lu xu do badri li'u Yes. > >> >> {ko cusku le sedu'u xukau do badri} = "Say whether you're sad". > >> if {da de du'u xukau do badri}, then {da} is a proposition and {de} is > >> a text-type corresponding to that proposition. > > > >But I'd have thought that the text-type corresponding to > >the propositionoid "xu kau do badri" is {xu kau do badri}. > > No! The text type {lu xu kau do badri} is independent of context. > The propositionoid {du'u xu kau do badri} is context dependent. I see that, but have trouble grasping the rest of your point... > It seems to me that the only way you can associate a text-type > with a proposition is through an utterance, A text-type is a pairing of a sound-pattern (not the actual physical sounds themselves) and a meaning, which is a possibly incomplete proposition - a proposition fragment or (za`e) "propositionoid". > and different > utterances will link a given text-type with different propositions, > as well as different utterances will link a given proposition > with diffrent text-types. That's right. Different utterances of the same text-type will enrich the same propositionoid to yield a context-specific fully fledged complete proposition. > For example, the text type {lu mi badri} is linked with a different > propositions when you say it than when I say it. If I say it, it > is linked with the proposition {le du'u mi badri}, and if you say it, > with {le du'u do badri}. Of course. > The proposition {le du'u mi badri} is linked with different text-types, > depending on who wants to express it. I would have to use > the type {lu mi badri}, you would have to choose between the > types {lu do badri} or {lu la xorxes badri}, depending on who > you're talking to. Right. > So du'u is not a one-to-one relationship. Not sure what that means. But anyway, back to the original point: > >> >> {ko cusku le sedu'u xukau do badri} = "Say whether you're > >> >> sad". It means "ko cusku the text type that expresses the proposition(oid) (that is expressed in Lojban by) {xu kau do badri}". So, noting your corrections, but basically sticking to my original contention, I think {ko cusku le sedu'u xukau do badri} means not "Say whether you're sad" but "ko cusku lu xu kau mi badri li`u". Ah... I think the light is dawning. I reckon I sort of grasp your point now. Hmm. If {mi djuno le du`u xu kau ko`a badri} means "for every x, a jetlai of le du`u ko`a badri, I know that x is jetlai of le du`u ko`a badri". The crucial thing is that this only makes sense when {Q kau} occurs within a proposition that itself is an argument of an epistemic predicate. I no think {ko cusku le sedu'u xukau do badri} just doesn't make sense at all. > >I agree that {da} is a proposition and {de} is a text-type > >corresponding to that proposition, but what proposition and > >text-type do you think da & de are? > > [We were talking about {ko cusku le sedu'u xukau do badri}.] > > The propositionoid is "whether you're sad". That's not really a propositionoid. It is a linguistic expression that makes sense only when the clause is an argument of an epistemic predicate. > The text-type I don't > know, since there wasn't yet any utterance to link the prepositionoid > with a text-type. You'll have to choose an appropriate text-type and > say it in order to satisfy my ko-request. For example, it could be: > > le du'u xukau do badri cu du'u xukau do badri kei lu mi badri li'u > do'e le nu do cusku lu mi badri li'u > > The propositionoid whether you're sad is the prepositionoid > whether you're sad linked with the text-type {mi badri} by your > uttering the text type {mi badri}. > > >> How would you say "Say whether you're sad"? > > > > Ko ???? le du`u xu kau do badri > > > >-- I can't remember the appropriate word for "say". > >Possibly something like "selvlagau" would do, but there > >must be gismu for it. > > No, {cusku} should be the gismu for it, but for some reason it > got tangled with text-types rather than with propositions, so that > you have to use {sedu'u} which brings a text-type associated > with the proposition. and which seems not to work, moreover. > All very complicated. There is also {bacru} > for text-types. I thought {bacru} meant to make a vocal sound. A text-type is not a vocal sound. > I don't know about {selvlagau}, maybe {seljufrygau}. Both are suitable. Curious that one must use a lujvo for so common a concept as "say". > >> Would that work as an explication of the direct question? > >Yes. Direct questions would reduce to a subcase of indirect > >questions. > > Right. So now how do we explicate indirect questions? Obviously > my explication using {le danfu be la'e lu xu ...} would be going in > circles. So I think. > I would say that {mi djuno le du'u xukau do badri} means, > as a first approximation, {mi djuno le du'u do ja'a badri ija > mi jduno le du'u do na badri}. (It's only a first approximation > because I'm ignoring fuzzy answers. To include them I would > need an infinite number of alternations.) And likewise for {ma} questions. > Any buyers? Yes. > I forgot what was your explication using truth values. It's just a more elegant version of yours. (See above for {xu kau}. {ma kau} is only slightly more complicated.) --And