implied co'e and Multiple-Question Questions

From Lojban
Revision as of 16:52, 4 November 2013 by Gleki (talk | contribs)
Jump to navigation Jump to search

{CODE(wrap=&quot;1&quot;)}16:34 <udjalus> coi ro do ro do ro do ro do

16:34 <vensa> xalbo: the desk at the office is the place one is at

16:35 == MayDaniel [email protected]/maydaniel has quit

16:36 == udjalus [email protected]/web/freenode/ip.138.100.208.216 has quit Client Quit

16:36 <vensa> short for {mi de'a zvati lo jibni be lo skami} :)

16:37 <@xalbo> je'e

16:37 <vensa> you think {mi de'a jibni} would also be understandable?

16:38 == Imami [email protected] has quit

16:39 <vensa> coi udjalus

16:39 <@xalbo> probably, yeah

16:39 <vensa> lo do jufra pu nalgenra

16:40 == Sxem [email protected] has joined #lojban

16:40 <vensa> si nalgendra

16:41 <vensa> .y .u'u li'a ja'a gendra

16:41 <vensa> gerna coi ro do ro do ro do ro do

16:41 <gerna> (0[[jbocre: {coi <(1ro BOI)1 do> DO'U} {<(1[jbocre: {ro BOI} do|{coi <(1ro BOI)1 do> DO'U} {<(1[jbocre: {ro BOI} do]] [[jbocre: {ro BOI} do|{ro BOI} do]])1 (1ro BOI do)1> VAU}])0

16:41 <vensa> cizra

16:42 <soto> gerna lo cizra le cizra la cizra

16:42 <gerna> (0[[jbocre: {<(1lo cizra KU)1 (1le cizra KU)1> <la cizra KU>} VAU|{<(1lo cizra KU)1 (1le cizra KU)1> <la cizra KU>} VAU]])0

16:42 <vensa> hmm, aparently a list of sumti with no selbri is also a gramatical utterance

16:42 == kribacr [email protected]/web/freenode/ip.66.192.126.3 has joined #lojban

16:42 <kribacr> coi

16:42 <vensa> probably in order to be able to answer questions like {ma zvati ma}

16:42 <@xalbo> Exactly.

16:42 <vensa> :)

16:42 <Volatile> heh. nice.

16:43 <kribacr> Sorry, my computer froze, so I missed anything you guys said.

16:43 <kribacr> What about ma zvati ma?

16:43 <vensa> we said: {coi ba'e la kribacr .i mi'a prami do}

16:44 <vensa> nm ma zvati ma

16:44 <Volatile> Hm, is no &quot;omitted selbri&quot; cmavu implictly involved somewhere?

16:45 <Volatile> co'e

16:45 <vensa> volatile:im not sure, but I can think of cases where it shouldnt be

16:45 <vensa> e.g.: {ma djica lonu ma cliva}

16:45 <vensa> the answer to that does not involve a single {co'e} relation between the two {ma}s

16:46 <vensa> xalbo: am I right?

16:46 == eternaleye [email protected]/developer/eternaleye has quit Ping timeout: 240 seconds

16:46 <Volatile> Hm. And the answer does not have to have the relationship pointed out?

16:47 <@xalbo> I think there's some debate on whether one can omit {co'e} or not.

16:47 <vensa> IMO: no. the order of the sumti defines which {ma} each one is answering...

16:47 <Volatile> Is { zo'e zo'e } a legit answer, or do you have to do { zo'e co'e lonu zo'e co'e } ? :)

16:47 <vensa> yay! I love debate :)

16:47 <@xalbo> {zo'e zo'e} is absolutely a legit answer.

16:48 <@xalbo> And there's also a {bu'a} such that {mi ti bu'a} is the same as {mi djica lo nu mi ti cliva}, it's just a complicated one :)

16:48 == kucli [email protected]/web/freenode/ip.82.232.145.133 has joined #lojban

16:48 <kucli> coi ke'u ro do

16:48 <vensa> xalbo: how do you define that bu'a?

16:48 <vensa> (using cei)

16:49 <vensa> coi la kucli

16:50 <@xalbo> I'm not sure, actually. With one place free, I can use ckaji, but there needs to be something for more variables.

16:51 <Volatile> Is it always possible to interpret the answers as full structures omitting a lot of relation wordsL

16:51 <Volatile> s/L/?/

16:51 <@xalbo> That is, {ckaji lo ka ce'u xi pa djica lo nu ce'u xi pa ce'u xi re cliva} is *almost* it, but not quite.

16:51 <vensa> xalbo: I dont follow. but I gather that's the argument &quot;for&quot; including {co'e}. i.e. saying that there IS &quot;some&quot; selbri that relates the two sumti, so that selbri can be {co'e} even if its unclear what {co'e} actually is...?

16:52 == LogicalDash [email protected] has joined #lojban

16:52 <vensa> wow! that was beautifuly complicated

16:52 <vensa> I think I understood the &quot;gist&quot; of it

16:52 <@xalbo> {djica co cliva} is pretty darn close, of course :)

16:52 <vensa> yes

16:53 == LogicalDash [email protected] has quit Client Quit

16:53 <vensa> but an *exact* selbri is possible? because {ckaji} isnt *exactly* the same.... even if it were gramaticaly standardised to use all those {xi}s

16:54 <@xalbo> So, absent a question we're answering, {mi lo mensi be do} is a grammatical utterance, and {mi co'e lo mensi be do} is a grammatical and sensible utterance. The question is whether the former has the same meaning/interpretation as the latter.

16:54 <vensa> also: I misunderstand {ce'uxipa ce'uxire}. y not just {ce'uxire}

16:54 <@xalbo> I wanted {mi ti cliva}, not {ti cliva}

16:55 <vensa> oh &quot;I leave here&quot;.. ok

16:56 <@xalbo> And I'm not even sure what your question about an exact selbri means.

16:56 <vensa> xalbo: but still: {mi ti ckaji lo ka ce'u xi pa djica lo nu ce'u xi pa ce'u xi re cliva} puts {ti} in the x2 of ckaji. not in the ce'uxire

16:58 <vensa> I think that {mi co'e lo mensi be do} means that there is an expressable relationship between me and your sister. wether we want to allow using {co'e} even in cases where that relationship cannot be exactly expressed (in the same form as it were expressed in the question) is what the debate is about (I reckon )

17:00 == kpreid [email protected] has quit Quit: Offline

17:01 <kribacr> .i li'a si'a ji'a mi co'e lo mamta be do

17:01 <vensa> doi kribacr xa'a'a

17:01 == lojysmanix [email protected] has quit Quit: Colloquy for iPhone - http://colloquy.mobi

17:02 <selpa`i> kribacr: Do you still teach like you did two years ago?

17:02 <@xalbo> I assert that it means there is a relationship that is relevant to the conversation, not that it can necessarily be expressed *simply*.

17:02 <kribacr> Wow... has it really been two years? O_O

17:02 <selpa`i> Yes, almost exactly-

17:02 <kribacr> Yes, I still teach. Well, in theory. I don't have as much t ime as I used to.

17:02 <vensa> xalbo: can it be expressed AT ALL?

17:02 <kribacr> But yes, I can still teach.

17:02 <@xalbo> (and I knew the {ckaji} was wrong, I was complaining about it at the time)

17:02 <selpa`i> Uh, sucks to hear that you are so busy.

17:02 <kribacr> Today seems like a slow day, so I may have time for a lesson in a few hours.

17:03 <kribacr> xalbo is a wonderful teacher as well.

17:03 <@xalbo> If there is a relationship that cannot be expressed, then lojban is utterly broken. I do not believe that is the case, though.

17:03 == kpreid [email protected] has joined #lojban

17:03 == mode/#lojban +o kpreid by ChanServ

17:03 <selpa`i> Well, it was more of a general question, and I wouldnt be your pupil this time.

17:03 == sanotehu_ [email protected] has joined #lojban

17:03 <kribacr> On the subject of implied co'e, if that's what you're talking about, I'm sort of a proponent for it.

17:03 <selpa`i> I know, xalbo is a nice teacher as well

17:03 <kribacr> Ah, gotcha.

17:03 <kribacr> Who? And how new?

17:03 <selpa`i> It was mostly a hypothetical question.

17:04 <selpa`i> I know someone who I would like to learn lojban or atleast get into it.

17:04 == sanotehu_ [email protected] has quit Client Quit

17:04 <kribacr> Ah, gotcha.

17:04 == LogicalDash [email protected] has joined #lojban

17:04 <vensa> xalbo: expressed with a single selbri. how? you cant even express a simple {ma broda lonu brode ma} with a single selbri, so what will you do with a huge number of {ma}s?

17:04 <@xalbo> Anyway, I'd love to have a word that means &quot;x1 (relation with an arbitrary number of empty spots marked by ce'u) is true with x2 filling ce'u1, x3 filling ce'u2, etc&quot;

17:04 <selpa`i> But it's unclear whether that person will actually do it.

17:05 <vensa> xalbo: that would probably solve it

17:05 <vensa> but as of currently, it seems that lojban is broken

17:05 <vensa> IF you add the implied co'e

17:05 <selpa`i> Does anyone happen to have the old chatlogs

17:05 <selpa`i> ?

17:05 <selpa`i> From 2 years ago?

17:06 <vensa> I think Hugglesworth has a bunch of logs on his machine. maybe they go back 2 yrs

17:06 <vensa> He once looked for something for me

17:06 <soto> http://www.lojban.org/resources/irclog/lojban/ ?

17:06 <@xalbo> vensa: The point is that that's a content word (it's just a selbri), and that the class of selbri is wide open. I could coin a fu'ivla that means just that, and there you go.

17:07 <vensa> hmmmmm

17:07 <vensa> but the original question did not use the fu'ivla. so is it still the same thing?

17:07 <@xalbo> There's nothing *fundamentally* unexpressible about that.

17:08 <vensa> yeah. I suppose I could/should be accepted (an implied co'e)

17:08 <vensa> so, whats the main argument AGAINST?



17:08 <vensa> soto: that's stuff expressed IN LOJBAN only

17:08 <vensa> it filters out the english

17:09 <soto> oh right

17:09 <@xalbo> I'm not sure. And I was previously somewhat against it. But I really don't have a good argument against it.

17:09 <vensa> hehehe

17:09 <vensa> this is exactly what my discussion topics log is for

17:09 <@xalbo> I tend to include explicit {co'e}, but I'm not sure there's a reason to do so.

17:09 * vensa is archiving

17:09 <kribacr> I think the problem with implied {co'e} is when people just speak vocatives.

17:10 <vensa> kribacr: what does that mean?

17:10 <vensa> example?

17:11 <kribacr> Well...

17:11 <kribacr> coi la vensa

17:11 <vensa> coi .u'i

17:11 <kribacr> Is there an implied {co'e} there?

17:11 <vensa> ahhhh

17:11 <vensa> dunno. and if there is. what harm does it do?

17:11 <kribacr> If there is, is it harmless?

17:11 <kribacr> Right.

17:11 <vensa> I think its harmless

17:11 <kribacr> I'm just playing devil's advocate here.

17:12 <vensa> obviously, if I am addressing you, I am telling you something

17:12 <vensa> perhasp {coi la kribacr zo'e co'e zo'e} is {coi la kribacr mi rinsa do}

17:13 <vensa> btw: you could say the same thing about bare UI

17:13 <@xalbo> Answering questions is weird anyway. {.i ma fanta lo nu do mo}, for instance, naïvely produces an aswer that doesn't seem right at all.

17:13 <vensa> but there too I believe there is an implied {co'e}

17:14 == LogicalDash [email protected] has quit Quit: Leaving

17:14 <vensa> xalbo: I think it's like the difference between &quot;a complete answer&quot; and fragments

17:14 <@xalbo> Yeah.

17:14 <vensa> I would reply {do fanta lonu mi surla} to be clear

17:14 <vensa> but is {do surla} a valid answer???

17:15 <vensa> that seems wrong

17:15 <UukGoblin> I'd just reply {go'i} ;-]

17:15 == LogicallyDashing [email protected] has joined #lojban

17:15 <vensa> because e.g. {broda pa} is ungramatical

17:15 <vensa> but I could have asked: {do mo xo gerku}?

17:16 <@xalbo> ke'u Answering questions is weird anyway.

17:16 == Amie [email protected] has quit Read error: Connection reset by peer

17:16 <vensa> I think full answers should be mandatory for questions with more than one question word

17:17 <vensa> or perhaps: {.i}-seperated answers

17:17 <UukGoblin> does {go'i} re-ask a {ma}/{mo} question?

17:17 <vensa> {do .i. surla}

17:17 <vensa> {broda .i pa}

17:17 <vensa> uuk: yes IMO

17:18 <vensa> you are repeating the question, leaving it in question form.

17:18 <vensa> you could be asking yourself aloud

17:18 <selpa`i> What?

17:18 <vensa> or asking the listener to answer instead of you

17:18 <@xalbo> There may be a case to be made for a I to separate answer words. It would also give an unambiguous way to answer a question instead of making a new, unrelated statement.

17:18 <vensa> .iesai

17:18 <UukGoblin> why 'unrelated'? ;-]

17:18 == kucli [email protected]/web/freenode/ip.82.232.145.133 has quit Quit: Page closed

17:19 * vensa is so happy he's archiving these new ideas

17:19 <UukGoblin> definitely related, although repeating a lot of what was said

17:19 <@xalbo> Maybe not &quot;unrelated&quot;, but &quot;dodging the question&quot;

17:19 == lojysmanix [email protected] has joined #lojban

17:19 <vensa> uuk: the statement {do surla} answers the question {ma fanta lonu do mo} but it makes an unrelated statement

17:20 <@xalbo> &quot;What were you doing with that woman I saw you with last night?&quot; &quot;The Phillies swept the NLCS.&quot;

17:20 <UukGoblin> oh, I see

17:20 <vensa> xalbo: &quot;dodging questions should still be allowed&quot;... just frowned upon :)

17:21 <UukGoblin> I thought replying with a full sentence made an unrelated statement

17:21 <vensa> no

17:21 <@xalbo> Yes. But the point is that if you ask a question with {mo}, I need a way to *not* answer it, and any bridi I saw *will* answer it.

17:22 <UukGoblin> mhm

17:22 <vensa> xalbo: example?

17:22 <tcatipax> mi na djuno?

17:22 <@xalbo> ke'u &quot;What were you doing with that woman I saw you with last night?&quot; &quot;The Phillies swept the NLCS.&quot;

17:22 <UukGoblin> like, &quot;What are you doing?&quot; &quot;Nice weather, isn't it?&quot;

17:22 <vensa> xalbo: how would that be in lojban? simpler version

17:23 <vensa> xalbo: whats wrong with answering {do mo} with {lo tcima cu pluka}

17:23 <vensa> ?

17:23 <@xalbo> .i do mo le mi mensi / .i .yy lo tcima ku melbi

17:23 <vensa> yeah. so whats the problem?

17:24 == donri [email protected]/dagodenhall has joined #lojban

17:24 <@xalbo> Well, it *probably* carries over the x2, at least.

17:24 <vensa> wha?!

17:24 <vensa> why does it carry stuff over?

17:24 <@xalbo> Think about it. Is not {cinba} a valid answer there?

17:24 <vensa> you said {melbi}. not {go'i}. not {co'e}

17:25 <vensa> yes. cinba is valid

17:25 <vensa> BUT

17:25 <vensa> the minute you replace the original x1 of the question with a diff x1, you are no longer answering the question. no?

17:25 <@xalbo> The answer to {mo} is some relationship such that its x1, x2, whatever other places were given to the {mo} make it true.

17:25 <vensa> but the minute you replace the original x1 of the question with a diff x1, you are no longer answering the question. no?

17:26 <@xalbo> I don't see where you get that from.

17:26 <vensa> common sense

17:26 <vensa> x1 or any other x

17:26 <kribacr> Tuesday's coming. Did you bring your coat?

17:27 <vensa> kribacr: is &quot;Tuesday's coming&quot; the answer?

17:27 <kribacr> I live in a giant bucket.

17:27 <vensa> kribacr: that is a y/n question.

17:27 <vensa> so unless I answered go'i or na go'i I dodged your question

17:28 <@xalbo> Well in {.i do mo / citka lo badna}, we're replacing in {lo badna} for the x2. Or is this new interpretive convention only for sumti that were previously explicitly filled?

17:28 <vensa> hmmm

17:28 <vensa> xalbo: in that example, you only ADDED X's. you didnt OVERRIDE any

17:28 <vensa> IMO when you OVERRIDE one of them, it becomes a &quot;dodging&quot; statement

17:29 == lojysmanix [email protected] has quit Quit: Colloquy for iPhone - http://colloquy.mobi

17:30 <UukGoblin> well

17:30 <vensa> xalbo: also, I dont understand how my proposition about {i} seperating answers to a multiple-question question &quot;solves&quot; this for you

17:31 <UukGoblin> let my put my question into the discussion, which is probably what xalbo already mentioned: say someone asks {do mo}, and you want to make an observative about a rain that's just started so you want to say {carvi}, but that'll make /you/ rain

17:31 <@xalbo> My idea was to make a new I that would do nothing but separate/precede answers. Then {.i} would always be dodging, and the new I would be for answering.

17:32 <kribacr> .i but for answers?

17:32 <kribacr> I like that.

17:32 <selpa`i> Hm..

17:32 <kribacr> .i ma gletu ma

17:32 <vensa> xalbo: ohhhh

17:32 == MayDaniel [email protected]/maydaniel has joined #lojban

17:32 == ksion [email protected] has joined #lojban

17:32 <kribacr> new-I la .kribacr. new-I lo mamta be do

17:32 <kribacr> .i'e

17:32 <vensa> .u'isai

17:33 <kribacr> Hmm.

17:33 <selpa`i> Seems unnecessary

17:33 <vensa> and {new-I la kribacr .i lo tcima} would be a partial answer

17:33 <kribacr> Are there any CVV or CV'V that could be derived from ... danfu is it?

17:34 <vensa> you could change {paunai} to mean &quot;answer follows&quot; :P

17:35 <kribacr> Eww... no.

17:35 <@xalbo> {.i do catra le nolraitru ki'u lo nu mo}.

17:36 <vensa> uuk: in the {carvi} case I would just say {ti carvi} thereby overriding hte x1 {do} and making it into a statement not a question

17:36 <vensa> valsi nolraitru

17:36 <valsi> nolraitru = t1=n1 is a regent/monarch of t2 by standard n2.

17:36 <@xalbo> Just *try* and change the topic on that one. Note that there are no places to override.

17:36 == bortzmeyer [email protected] has quit Quit: Leaving.

17:36 <vensa> xalbo: ooohhh

17:36 <vensa> you got me

17:37 <kribacr> D'oh, {dau} is taken.

17:37 <kribacr> Stupid hex.

17:37 <vensa> how about {mi na catra .i do bebna} :P

17:38 == Amie [email protected] has joined #lojban

17:38 == tama [email protected] has quit Ping timeout: 240 seconds

17:38 <UukGoblin> there was this meta-negator

17:38 <UukGoblin> metalinguistic negator

17:38 <UukGoblin> na'i

17:39 <vensa> yes! good point Uk

17:39 <vensa> it seems very handy here

17:39 <vensa> {.i do catra le nolraitru ki'u lo nu mo}. {na'i}

17:39 <vensa> I wanted to say also that the &quot;new i&quot; should be for cases where you DONT intend to answer becuz those are the less frequent cases

17:40 <vensa> so using {na'i} for that purpose exactly seems brilliant. (and the intended way)

17:41 <vensa> so the answer to the {carvi} problem would be. {.i na'i carvi}

17:41 <UukGoblin> hm.

17:41 <selpa`i> How do you say &quot;this&quot; as in &quot;this house&quot; ?

17:42 <UukGoblin> I kinda thought {na'i} would mean &quot;your question is wrong&quot; rather than &quot;I don't feel like answering it&quot;

17:42 <@xalbo> vensa: That says it's not raining.

17:42 <UukGoblin> selpa`i, {ti}?

17:42 <vensa> selpa'i {lo vi zdani}

17:42 <@xalbo> selpa`i: {ti poi zdani}, roughly.

17:42 * vensa is looking up na'i

17:42 <@xalbo> (could also be {noi} instead of {poi})

17:43 <selpa`i> Thx gusy.

17:43 <kribacr> lo bu'u zdani

17:43 <selpa`i> Guys.

17:43 <dbrock-> I too would like to have �a word that indicates that something is an answer

17:43 <dbrock-> the opposite of {pau}

17:44 <kribacr> Hmm. I wish there was more CVV and CV'V space available. ._.

17:44 <vensa> xalbo: so, {na'i .i carvi}

17:44 == Suprano [email protected] has quit Ping timeout: 245 seconds

17:44 <dbrock-> if I got to choose, {pau nai} would mean &quot;answer follows&quot;, {pau cu'i} would mean &quot;question does not follow&quot;, and {pau nai cu'i} would mean &quot;answer does not follow&quot;

17:45 <vensa> I still think maybe the word should be for &quot;this is NOT an answer&quot;. I would hate to be required to utter another syllable for 99% of the time

17:45 <@xalbo> dbrock-: Then I'm glad you don't get to choose.

17:45 <dbrock-> :)

17:45 <@xalbo> Sorry, had to go there, but I don't think that's a natural scale at all, and it changes way too much.

17:45 <kribacr> da'au

17:46 <vensa> dbrock: does {pau} currently have a {cu'i}?

17:46 <dbrock-> yeah, I think of {UI nai} as being a separate scale

17:46 <dbrock-> but that's not how most people think of it

17:46 == kpreid [email protected] has quit Quit: Offline

17:46 <@xalbo> I just don't like {.i} between multiple parts of the same answer. And I think even then I'm not sure all answers can be done without restating the whole sentence.

17:46 <vensa> xalbo: what was that an answer to?

17:47 <dbrock-> to me, the {pau} scale would be &quot;how much of a question is this&quot;, whereas the {pau nai} scale would be &quot;how much of an answer is this&quot;

17:47 <dbrock-> so you could have {pau pau nai} for &quot;answering with a question&quot;

17:47 <vensa> how about {pauna'i} for &quot;I dont intent to answer you&quot;? :P

17:47 <dbrock-> well, {pau nai pau} would be a more natural order, I guess

17:48 <@xalbo> vensa: What was what an answer to?

17:48 == urandom_ [email protected] has joined #lojban

17:49 <@xalbo> ({ge'i}, for instance, I don't think can be answered except with a whole sentence)

17:49 <dbrock-> {fu'e pau nai i broda i brode i brodi fu'o}?

17:49 <vensa> xalbo: the statement you said above the statement I said that asked tha

17:49 <vensa> *that

17:49 <dbrock-> for a three-part-answer

17:49 == tcatipax [email protected]/web/freenode/ip.217.171.129.74 has quit Quit: Page closed

17:50 <@xalbo> vensa: Just quote the mabla sentence.

17:50 <vensa> valsi ge'i

17:50 <valsi> ge'i = logical connective: forethought all but tanru-internal connective question (with gi).

17:50 <vensa> <@xalbo> I just don't like {.i} between multiple parts of the same answer. And I think even then I'm not sure all answers can be done without restating the whole sentence.

17:50 <vensa> gerna ge

17:50 <gerna> not grammatical: ge ⚠

17:50 <vensa> gerna e

17:50 <gerna> (0e)0

17:50 <vensa> hmmm.. geks alone seem to be ungramatical

17:50 <dbrock-> gerna ge co'e gi co'e

17:50 <gerna> (0[[jbocre: ge {co'e VAU} gi {co'e VAU} VAU|ge {co'e VAU} gi {co'e VAU} VAU]])0

17:51 <vensa> gerna ge gi

17:51 <gerna> not grammatical: ge _gi_ ⚠

17:51 <dbrock-> what's the problem with that?

17:51 <vensa> yeah, xalbo?

17:51 <@xalbo> Means that a question with {ge'i} is harder to answer.

17:51 == urandom [email protected] has quit Ping timeout: 265 seconds

17:51 <vensa> not if we add implied {gi} to the grammar parser

17:51 <dbrock-> true

17:51 <@xalbo> At least, the only way to answer it is to make an entire bridi, not just fill in the blank.

17:52 <dbrock-> I guess you could answer with an afterthought connective?

17:52 <vensa> de'a

17:52 <@xalbo> Um, *{ge gi} isn't legal either.

17:53 <dbrock-> see any problem with answering with afterthoughts, xalbo?

17:54 <@xalbo> Don't know. In general it's odd to answer with something other than the form of the question.

17:58 == Suprano [email protected] has joined #lojban

17:59 == syllogism [email protected] has quit Read error: Connection reset by peer

18:00 <vensa> dbrock: a question could contain both {ge'i} AND {ji} so that answering in a diff form would be confusing

18:00 <vensa> xalbo: add implied co'es too and you'll get {ge co'e gi co'e}

18:01 <@xalbo> If you try to answer out of order, though, you *really* screw things up, so I don't think that's a problem.

18:01 <vensa> xalbo: why do you think {.i} between multiple parts of a fragmented answer cant be a complete reply?

18:03 <@xalbo> {.i} separates bridi by the same speaker. That seems pretty different from separating fragments that are all used to fill parts of a single bridi.

18:03 <vensa> why?

18:03 == Suprano [email protected] has quit Ping timeout: 245 seconds

18:03 <vensa> ma tavla ma -> .i mi .i do

18:04 <vensa> means: {.i mi tavla .i do se tavla}

18:04 <vensa> (remeber the implied co'e)

18:04 <@xalbo> That seems very different from {mi tavla do}.

18:05 <vensa> why? context welds them together IMO

18:05 <vensa> how do you solve the {do surla} bug with something other than a seperating {i}?

18:06 <dbrock-> xalbo has already proposed the addition of new I

18:06 <vensa> oh.

18:06 <vensa> so {newI mi newI do} is acceptable xalbo?

18:06 <dbrock-> danfu ze'ei i mi danfu ze'ei i do

18:06 <@xalbo> Seems much more so, yes.

18:07 <vensa> i c

18:07 <vensa> fine we need the newI for other things too (specifying dodging answers)

18:07 <@xalbo> (I'd still probably just answer {mi do}, but for more complicated ones, yes)

18:07 <dbrock-> I don't really see why we need a new I

18:08 <@xalbo> The point is that if newI is for answering, then oldI (spelled {.i}) keeps its completely normal function, which just happens to work out to question dodging.

18:08 <dbrock-> well, it's not a matter of need, of course

18:08 <vensa> in that case I am &quot;for&quot; dbrock's {paunai} def

18:08 <dbrock-> but I mean other similar things are solved using UI

18:08 <@xalbo> (and in most cases, you start speaking without either, so there's no problem)

18:09 <vensa> xalbo: isnt there an implied {oldI} at the start?

18:09 <@xalbo> I strongly oppose changing {pau nai}. You can argue for a UI, but you can't have that one.

18:09 <dbrock-> I don't propose changing {pau nai}

18:09 <dbrock-> you can't do things like that

18:10 <vensa> {paucu'i} is currently undefined

18:10 <dbrock-> it's impossible, so debating it is a waste of time

18:10 <vensa> dbrock: 1. anything is posible

18:10 <vensa> 2. didt you suggest that earlier?

18:10 <vensa> http://www.lojban.org/tiki/tiki-index.php?page=BPFK+Section%3A+Discursives

18:10 <UukGoblin> hm.

18:11 <dbrock-> I have long been an advocate of thinking of {UI ja'ai} and {UI nai} as completely separate scales

18:12 <vensa> so you did suggest to change {paunai}

18:12 <UukGoblin> with stuff like 'paunai', what is there to distinguish between definitions like 'answer follows', 'no question follows' or 'unquestion follows'?

18:12 <dbrock-> that's why I said &quot;if I got to choose, {pau nai} would mean &quot;answer follows&quot; &quot;

18:12 <UukGoblin> :-]

18:12 <dbrock-> I didn't say &quot;I propose we change the meaning of {pau nai}&quot;

18:12 <vensa> oh

18:12 <vensa> I read that as that

18:13 <vensa> who cares about the old meaning of {paunai} its probably rarely been used

18:13 <dbrock-> yeah, I can see how you'd read it as a proposal

18:13 <@xalbo> (Note that I also didn't say &quot;we need a new I&quot; but &quot;a case could be made for a new I&quot;

18:13 <dbrock-> {pau nai} has seen significant enough use that people will just say &quot;NO&quot;

18:13 <vensa> uuk: what you mean by 'no question follows' or 'unquestion follows'

18:13 <dbrock-> and the only effect of trying to change its meaning will be to cement the old meaning even further

18:13 <vensa> xalbo :)

18:14 <vensa> &quot;cement&quot;?

18:14 <UukGoblin> vensa, &quot;the following is not meant to be intepreted as a question&quot; and &quot;the following is meant to cancel the question in question&quot;

18:15 <vensa> I am very much an advocate of changing the old for the benefit of the future. as an answer to the nay-saying conservatives I have proposed the &quot;version\scripting&quot; system

18:15 <UukGoblin> cementing is a popular technique of postponing trouble with blown up nuclear reactors for later

18:15 <vensa> uuk: regular {i} is the first. and you cant obliterate a question once it was asked. you can just choose to not answer it with regular {i}



18:17 <UukGoblin> vensa, nah, it's kinda not my question... my problem is {pau nai} is a cluster, but because {pau} can be negated in different ways, doesn't it make {pau nai} a bit ambiguous?

{CODE}

{CODE(wrap=&quot;1&quot;)}17:18 <vensa> xalbo: how does {i'au} sound to you as the &quot;newI&quot;? (from {i} + {danfu})

17:18 * Volatile citka

17:18 <vensa> Volatile: {zo'oi} is for one word quote only

17:19 <Volatile> vensa: well, that was one word quoted. Then, I kinda changed language.

17:19 <vensa> Volatile: you can translate word by word using valsi. it still wont help you understand the grammar

17:19 <vensa> Volatile: that doesnt parse

17:19 <@xalbo> vensa: Feels like an attitudinal to me.

17:19 <vensa> you need {zoi .gy. bla bla bla .gy}

17:20 <vensa> xalbo: does {i} feel like an attitudinal?

17:20 <vensa> are there experimental-cmavo attitudinals?

17:20 <vensa> I guess ur right tho

17:21 <@xalbo> I'm used to single vowels being connectives, and multiple vowels being UI1. It's not set in stone of course, but probably not good to mess with.

17:21 <Volatile> vensa: I meant to just quote one word, but then I realize that I don't really know the correct grammar (modals, no?) to express what I wanted anyhow...

17:21 <vensa> V: fine

17:21 * Volatile cu citka.i co'o

17:22 <vensa> xalbo: {da'au}?

17:22 <@xalbo> Better.

17:22 <vensa> or: {ni'au} (ni'o} + {danfu}

17:24 <kribacr> .u'i sai coi jungo

17:24 <vensa> yeah it does sound a bit chinese

17:24 <vensa> kribacr: did you hear about our idea?

17:24 <kribacr> Yes.

17:24 <kribacr> I for responses.

17:24 <kribacr> I was here yesterday.

17:25 <vensa> those are different I's in ur 2 sentences...

17:25 <vensa> english ambiguity :P

17:25 <kribacr> I lamented the fact that {dau} was unavailable.

17:25 <kribacr> Indeed.

17:25 <@xalbo> Right now I have my head in the *huge* bpfk thread from the weekend, about where &quot;texts&quot; begin and end with multiple speakers

17:26 <ksion> coi rodo .i ma lamji je fanza se stidi la vensa u'i

17:26 <vensa> thats also a big one

17:26 <vensa> .u'iru'e .oiro'a doi ksion

17:26 <@xalbo> Unfortunately, we don't have a convention for quoting selma'o names in running English text (since for all but I there's not much problem), nor even for talking about them in Lojban.

17:27 <vensa> doi ksion ni'au lo danfu valsi

17:27 <ksion> ue

17:27 <ksion> xu do stidi tu'a lo cmavo pe lo danfu pe fi'o simsa zo pau

17:28 <ksion> s/lo danfu pe/lo danfu zi'epe

17:28 <vensa> ksion: {ni'au} (sounds cooler) is the proposed cmavo which will act &quot;like&quot; an {i} but signal that the utterance is a &quot;partial reply&quot; to a question word, and not a full sentence

17:28 <@xalbo> .i cmavo lo selma'o be zo .i

17:28 <ksion> And the need for having this is ...?

17:29 <vensa> imagine you are asked a multiple-question question, e.g.:

17:29 <vensa> {ma djica lonu do mo}

17:29 <ksion> xalbo: I usually say {zo'oi FAhA}.

17:30 <vensa> if you want the full answer to be {do djica lonu mi surla} would you say {do surla}?

17:30 <ksion> No, {do .i surla}.

17:30 <vensa> becuz that seems to imply some other nonexistent bridi relationship

17:31 <vensa> yes, that may be enough. but

17:31 <vensa> then we got into decding how we are supposed to &quot;avoid&quot; a question

17:31 <vensa> I ask you {do mo} but you dont want to answer. you want to point out that it's raininng so you say {carvi}

17:31 <vensa> it &quot;seems&quot; as though you are claiming that {mi carvi}

17:32 <ksion> {.i co'e .i carvi}

17:32 <selpa`i> how bout ni'o

17:32 <ksion> {ni'o} or {ta'o} is fine too.

17:32 <vensa> xsion: what about if I ask {xo} and you dont want to answer?

17:33 <ksion> ji'i

17:33 <vensa> so basically the questioner forces the listener to respond to their question, even if its with a vague answer, get the question &quot;out of the way&quot; before he can say anything?

17:33 <ksion> (if you want a question type where I don't have a 'neutral' reply, try {cu'e} :) )

17:33 <vensa> IMO that is a little annoying

17:34 <vensa> valsi cu'e

17:34 <valsi> cu'e = tense/modal question.

17:34 <selpa`i> I dont think its a problem

17:34 <ksion> Then {.i .i <your stuff>} works.

17:34 <vensa> do'e

17:34 <selpa`i> When I answer by saying something unrelated, then context will show that I didnt care to answer.

17:34 <vensa> {do'e} is vague of {cu'e}

17:34 <vensa> but there is no vague for {fi'a}

17:34 <vensa> valsi fi'a

17:34 <selpa`i> That happens all the time in natural languages too

17:35 <valsi> fi'a = sumti place tag: place structure number/tag question.

17:35 <ksion> {faxiji'i}

17:35 <vensa> selpa'i: but there can be unclear cases where it's not clear if you are answering or not

17:35 <vensa> ksion: nice

17:35 <selpa`i> In such a case, the question asker will ask for clarification like normal

17:35 <vensa> still. y force the listener to &quot;get the question out of the way&quot;.

17:35 <ksion> vensa: Question is not enforced grammatically. You don't have to escape it by grammatical means, really.

17:36 == Dessous [email protected] has joined #lojban

17:36 <vensa> if I ask you {ma mo mo xo ma mo xu}

17:36 <selpa`i> That's a stupid question

17:36 <vensa> selpa'i: it's just an example

17:36 <ksion> Then the correct answer is {ko ko gletu} :P

17:36 <selpa`i> yes

17:36 <vensa> I can fill it with other &quot;meaning&quot; words and leave the same number of Q words

17:36 <selpa`i> u'i

17:36 <vensa> xa'a

17:37 <selpa`i> If you ask me such a dumb question, you cant expect me to answer it

17:37 <ksion> Or {ki'a}, if you still want to be polite (I'd not be).

17:37 <vensa> ksion: you say I dont have to escape the question. so why do you propose the {co'e .i broda} approach?

17:38 <selpa`i> co'e is a polite evasion move

17:38 == kucli [email protected]/web/freenode/ip.82.232.145.133 has quit Ping timeout: 265 seconds

17:38 <ksion> vensa: Because it is customary to expect an answer after a question. It's not by-grammar though, but only by-semantics.

17:38 <vensa> IMO if you ask me a question I should be able to say whatever I want. but only if I want to ONLY &quot;fill in the slots&quot; of the question words, I need something like {ni'au}

17:38 <selpa`i> I agree.

17:39 <timonator> ni'au?

17:39 <vensa> ksion: I think this should be decided by grammar. much like the {go'i} answer is a gramatical mechanism

17:39 <ksion> ta'a ni'o ta'o a'anai -- Possible solutions.

17:40 <ksion> {go'i} being answer is not grammatical mechanism. {go'i} being last bridi is.

17:40 <vensa> I think not defining this issue is leaving room for some sort of ambiguity

17:40 <timonator> right, go'i is by far not only for answers

17:40 <vensa> not only

17:40 <timonator> {i za'a ta muvdu} {i do go'i gasnu} {i mi na go'i}

17:40 <vensa> nm go'i

17:41 <kribacr> .i do ja'a go'i

17:41 <kribacr> coi spitaki

17:41 <ksion> 17:40:49 <vensa> I think not defining this issue is leaving room for some sort of ambiguity -- And?

17:41 <kribacr> ko cikna binxo

17:41 <vensa> And? do you like ambiguity?

17:41 <selpa`i> lojban is hella ambiguous anyway

17:41 <vensa> says you

17:41 <vensa> it's not supposed to be

17:41 <ksion> Semantically, I'm from neutral to positive.

17:42 <selpa`i> Semantically it is.

17:42 <ksion> Of course it is supposed to be.

17:42 <selpa`i> Grammatically, not so much.

17:42 <ksion> .i mi za'e firxance lo se cusku be la vensa

17:42 <vensa> valsi firxance

17:42 <valsi> no results. http://vlasisku.lojban.org/firxance

17:42 <vensa> oh

17:42 <vensa> is that an example of semantic ambiguity

17:42 <vensa> ?

17:43 <ksion> Nope ;)

17:43 <vensa> but it is

17:43 <ksion> Well, nonce words are an example of it.

17:43 <vensa> xalbo: help me out here

17:43 <ksion> ({firxance}, as it's not-so-hard to figure out, is meant to mean &quot;facepalm&quot; :) )

17:43 <vensa> why did we think it was a good idea yesterday?

17:45 <vensa> ksion: you want to go over the discussion and see if you agree with any of it?

17:45 <vensa> http://www.lojban.org/tiki/tiki-index.php?page=Implied+%7Bco%27e%7D+and+Multiple-Question+Questions

17:45 <ksion> Sure.

17:45 <vensa> gr8

17:45 * vensa is glad there is a use for his archiving

17:47 <@xalbo> .oi

17:48 <ksion> vensa: You can extract a place from any number of nested abstractions be using an appropriate number of {jai} and SE.

17:48 <@xalbo> It seems wrong that we use the same cmavo ({.i}) for starting a new complete bridi, and for filling in sequential fragments of someone else's bridi.

17:49 <Volatile> &quot;facepalm&quot; sounds like some kind of tree to me. Guess it's some malglico...

17:49 <vensa> I think we should decide between 3 options: 1. you need to say {co'e} and {ji'i} for every question to get it out of the way first (i dont like this option) 2. you say {ta'a} or {ni'o} or something to imply that you are NOT answering the question 3. you use {ni'au} for cases when you want to indicate that you ARE answering

17:49 <vensa> xalbo: &quot;seems wrong&quot; is a bit short of an argument IMO

17:50 <selpa`i> #3 seems terrible

17:50 <ksion> 4. You say what you want since question aren't grammatically binding.

17:50 <vensa> ksion: ki'e

17:50 <selpa`i> I like 4 the best.

17:50 <vensa> ksion: if 4, then who knows whether I'm answering you or not?

17:51 <selpa`i> Context.

17:51 <vensa> I can imagine cases where it's unclear from context

17:51 * Volatile does not like &quot;

17:51 <Volatile> argh

17:51 <Volatile> 4

17:51 <vensa> so what if &quot;natlangs get away with it&quot;

17:51 <selpa`i> Sure, but in those cases, you can clarify.

17:51 <ksion> &quot;What are you doing?&quot; &quot;Raining&quot;

17:51 <@xalbo> Well, 4 is always going to be the most common, and with good reason. But it still seems important (there's that &quot;seems&quot; again) to be able to specify one way or another.

17:53 <Volatile> I'd say that in this language more than others, specificity matters...

17:54 <vensa> wouldnt it be prettier if lojban had an elegent way to deal with it?

17:54 <ksion> 17:38 <UukGoblin> na'i -- uasai, how could I forget it!

17:54 <ksion> There you go, your miraculous &quot;escape-all-questions&quot; cmavo.

17:54 <vensa> selpa'i: calrifying with more sentences is exaclty the thing we want to avoid in lojban

17:54 <@xalbo> The example I use before, I think, was {.i do catra le nolraitru ki'u lo nu mo}. Answering {co'e} there is a bad idea.

17:55 <selpa`i> Is that so?

17:55 <@xalbo> ksion: Problem is that {na'i} isn't avoiding an answer, it's specifically saying that there *isn't* one.

17:55 <vensa> xalbo: yes! thank you! the {catra} example

17:56 <vensa> .ie on the {na'i} not working

17:56 <vensa> {na'i} is something else

17:56 <selpa`i> Why did you kill the monarch?

17:56 <@xalbo> If you ask the above catra question, I can't plead the fifth in Lojban. I can use {na'}, or I can give a reason, but I can't just say &quot;I want a lawyer!&quot; without that being my answer for why I killed him.

17:57 <ksion> je'e

17:58 <ksion> Then that's why we have {na'i}.

17:59 <@xalbo> {na'i} doesn't do that, though. It says that the question itself is wrong (in this case, because I didn't kill him), not that I'm not going to answer it.

17:59 <vensa> ksion {na'i catra .i mi djica tu'a la'oi lawyer}?

17:59 * Volatile klama .i co'o

17:59 <ksion> vensa: {na'i} is UI. Thus {.i go'i na'i}

18:00 <vensa> xalbo: so why not {i go'i na'i .i <what you want to say>}

18:00 <@xalbo> That means {mi catra le nolraitru ki'u lo nu no'a na'i}, which sure seems b0rken to me.

18:00 <ksion> ...What?

18:00 <vensa> huh?

18:00 <selpa`i> ..?

18:01 == Wolvenreign [email protected] has joined #lojban

18:01 <@xalbo> I asked for a {mo}. You gave me a bridi.

18:01 <vensa> it says {na'i mi catra le nolraitru ki'u lo nu ma}

18:01 <vensa> ohhhh

18:01 == Wolvenreign has changed nick to labnytru

18:01 <ksion> No, I gave you a selbri. Which is incidentally the same.

18:01 <vensa> wow - this raises another issue

18:02 <ksion> ...

18:02 <@xalbo> But even without that, the {na'i} doesn't not answer the question. it asserts that the question is wrong. That's different.

18:03 <vensa> A says {do djica lonu mo} B wants to repeat the question to A. does {go'i ra'o} work?

18:03 <ksion> xalbo: It doesn't answer it.

18:03 <labnytru> coi rodo

18:03 <vensa> xalbo: legal differences. &quot;I didnt say I didnt do it&quot; :P



18:05 <@xalbo> For whatever legal reason, I don't want to say that I *didn't* kill him, but I sure as Hell don't want to say I did. All I *want* to say is &quot;I want a lawyer&quot;.

18:05 <@xalbo> {na'i} does the first of those three, {co'e} the second.

18:06 <labnytru> So, folks.

18:06 <ksion> Congratulations. You made me use the biggest cannon.

18:06 <labnytru> How many of you know what SEO (Search Engine Optimization) is?

18:06 <ksion> Behold, {sei}! ... {.i sei na pinka}

18:08 <@xalbo> labnytru: I only know of SEO as &quot;Evil people trying to hijack Google to show me what they want instead of what I want.&quot;

18:08 <ksion> (Although I still think xalbo misunderstands {na'i} giving it less &quot;power&quot; than it really has)

18:08 <labnytru> Good enough answer, although it doesn't have to be that way.

18:09 <labnytru> Ultimately, you could have a website with valuable information related to the keyword and not have it show up on Google because of it's lack of optimization.

18:09 <ksion> {na'i} is metalinguistic. It invalidates EVERYTHING linguistically associated with statement it marks. It does not only negate the &quot;truth case&quot;, but also &quot;false case&quot;.

18:09 <@xalbo> Point.

18:09 <vensa> xalbo: why not go with option2? {do pu catra lo nulraitru ki'u lonu mo} -> {ni'o mi djica tu'a la'oi lawyer}

18:10 <labnytru> Well, with that in mind...

18:10 <vensa> xalbo: does &quot;point&quot; mean you agree about {na'i} with ksion?

18:10 <@xalbo> It means that ksion made a good point about {na'i}, and I'm stepping back to reconsider in light of that.

18:11 <labnytru> I've been working with an SEO forum to learn more...and I've been chosen to be the sole co-moderator of it.

18:11 <vensa> ok. so we're going with option2 and {na'i}?{CODE}

{CODE(wrap=&quot;1&quot;)}19:00 <ksion> vensa: So, did we come to any conclusions regarding your question issues?

19:00 <vensa> selpa'i: {i mi kakne lonu ca lonu} sounded like you were correcting yourself

19:00 <selpa`i> *fai

19:00 <selpa`i> I wasnt

19:00 <kribacr> Yes, exactly.

19:00 <vensa> ksion: I think we said to use {na'i} if you want to avoid answering a question

19:00 <selpa`i> or was I? It should be in the text

19:01 <kribacr> The x1 becomes the {fai}-tagged slot.

19:01 <selpa`i> Yes.

19:01 <vensa> you wasnt

19:01 <kribacr> Tagless {jai} basically implies a {tu'a} for the x1.

19:01 <vensa> but your voice sounded like you were

19:01 <vensa> it confused me

19:01 <selpa`i> okay

19:01 <kribacr> You understand {.i tu'a mi bandu do}?

19:01 <selpa`i> valsi bandu

19:01 <valsi> bandu = x1 (event) defends/protects x2 (object/state) from threat/peril/potential x3 (event).

19:01 <selpa`i> Yes.

19:01 <selpa`i> I do.

19:01 <ksion> vensa: Okay. Thing is, I don't like it. xalbo was wrong about {na'i} invalidationg only the &quot;truth variant&quot; of the question but was right about it invalidating the question and not only expressing the desire to avoid answering it.

19:01 == tcatypatxu [email protected]/web/freenode/ip.217.171.129.72 has joined #lojban

19:02 <kribacr> {.i mi jai bandu do} means pretty much the same thing.

19:02 <selpa`i> *head explodes*

19:02 <tcatypatxu> mi citka lo pitnanba be lo vo cilra

19:02 <tcatypatxu> Guess!

19:02 <ksion> cilra ki'a

19:02 <vensa> ksion: &quot;invalidating the question&quot; gives you the option to say something else. you dont need to &quot;express your desire to not answer it&quot; IMO. you could do that with an additional attitudianl

19:03 <selpa`i> So {jai bandu} is the selbri?

19:03 <kribacr> Yes.

19:03 <selpa`i> Which has x1 = tu'a something

19:03 == MayDaniel [email protected]/maydaniel has quit

19:03 <ksion> vensa: True. But {na'i} also states that question is metalinguistically wrong *regardless* of you wanting or not to answer it.

19:03 <selpa`i> Okay

19:04 <tcatypatxu> I assume my sentence is correct since no one is telling me 101 ways I'm horribly wrong :D

19:04 <kribacr> With a place structure of &quot;x1 defends/protects x2 (object/state) from threat/peril/potential x3 (event) with event of defending fai&quot;. Or something like that.

19:04 <ksion> vensa: And the other way around: you would have to invalidate a totally valid question if you didn't want to answer it and used {na'i} to express that unwillingness/

19:05 <selpa`i> uanai

19:05 <kribacr> Where am I losing you?

19:05 == smajis [email protected] has joined #lojban

19:05 <vensa> ksion: ok. so do you feel better about {ni'o} instead of {na'i}?

19:05 <selpa`i> It's just so weird and confusing

19:05 <selpa`i> Especially since it seems to double

19:06 <selpa`i> fai and x2 are the same there or not?

19:06 == vilfredo [email protected] has quit Quit: Ex-Chat

19:06 <ksion> vensa: Yes.

19:06 <selpa`i> And in {mi jai bandu do} fai is not even used.

19:06 <smajis> .i coi

19:07 <kribacr> {fai} is just another spot.

19:07 <vensa> ksion: good. then let that be the new conclusion. except we'll need the BPFK to include that in the definition of {ni'o}

19:07 <kribacr> You don't have to fill every place.

19:07 <kribacr> Same with SE.

19:07 <selpa`i> I know.

19:07 <kribacr> You can easily just say {.i mi te vecnu}.

19:07 <ksion> vensa: Maybe. Not sure if it's needed. &quot;New topic&quot; being the key part of {ni'o}'s definition is pretty clear.

19:08 <vensa> ksion: ok. then maybe in the second-layer guidelines

19:08 <selpa`i> It's confusing that {tu'a mi bandu du} = {mi jai bandu do}

19:08 == zyzazezuzizo [email protected] has joined #lojban

19:08 <ksion> vensa: u'i You like the layers! :)

19:08 <vensa> yes :)


19:09 <ksion> Better to like the layers than be liked by lawyers.

19:10 <vensa> .u'i{CODE}