Vote for the Future Global Language
najrut
------------------------
http://international-languages.webs.com/
MorphemeAddict
--
You received this message because you are subscribed to the Google Groups "lojban" group.
To post to this group, send email to loj...@googlegroups.com.
To unsubscribe from this group, send email to lojban+un...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/lojban?hl=en.
Lindar
isn't now a candidate auxlang. If it -does- get upvoted, people are
going to ask why it would make a good auxlang, and we can give no
legitimate response because it wouldn't be. Personally, I don't want
hundreds of douchebags every day asking why we're the best auxlang and
asking why we think we're better than $popular_auxlang.
Oren
--
You received this message because you are subscribed to the Google Groups "lojban" group.
To post to this group, send email to loj...@googlegroups.com.
To unsubscribe from this group, send email to lojban+un...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/lojban?hl=en.
Luke Bergen
.alyn.post.
incomplete, and completing it requires more effort, which
can be had by generating interest, which can be had by
generating attention.
Only a percentage of people in the community will work to
complete the language, and the community consists only of those
who spend time in it, which is a subset of those who are
interested, which is a subset of those who have heard about it.
There are ~500 subscribers to this mailing list, of which maybe 2
dozen regularly post. I don't have any reason to believe that
this is atypical, in that you don't see an increase in direct
participation without an increase in silent participants and
everything else leading up to that.
tl;dr: Lojban isn't going to finish itself. ;-)
-Alan
On Tue, Jan 04, 2011 at 04:43:43PM -0500, Luke Bergen wrote:
> I never understood that idea. If I were a serial killer I wouldn't want
> any press.
> And there is at least 1 good reason for lojban being an auxlang (at least
> at some point in the future after it's been completed). One of the design
> goals was cultural neutrality. Not many conlangs and certainly no natlangs
> have been designed with that in mind I think.
>
> On Tue, Jan 4, 2011 at 4:32 PM, Oren <[1]get....@gmail.com> wrote:
>
> But any press is good press >D
>
> On Tue, Jan 4, 2011 at 15:57, Lindar <[2]lindar...@yahoo.com> wrote:
>
> I'm -not- going to vote for lojban. It has not been and definitely
> isn't now a candidate auxlang. If it -does- get upvoted, people are
> going to ask why it would make a good auxlang, and we can give no
> legitimate response because it wouldn't be. Personally, I don't want
> hundreds of douchebags every day asking why we're the best auxlang and
> asking why we think we're better than $popular_auxlang.
> --
> You received this message because you are subscribed to the Google
> Groups "lojban" group.
> To post to this group, send email to [3]loj...@googlegroups.com.
> To unsubscribe from this group, send email to
> [4]lojban+un...@googlegroups.com.
> For more options, visit this group at
> [5]http://groups.google.com/group/lojban?hl=en.
>
> --
> Oren Robinson
> (315) 569-2888
> 102 Morrison Ave
> Somerville, MA 02144
>
> --
> You received this message because you are subscribed to the Google
> Groups "lojban" group.
> To post to this group, send email to [6]loj...@googlegroups.com.
> To unsubscribe from this group, send email to
> [7]lojban+un...@googlegroups.com.
> For more options, visit this group at
> [8]http://groups.google.com/group/lojban?hl=en.
>
> --
> You received this message because you are subscribed to the Google Groups
> "lojban" group.
> To post to this group, send email to loj...@googlegroups.com.
> To unsubscribe from this group, send email to
> lojban+un...@googlegroups.com.
> For more options, visit this group at
> http://groups.google.com/group/lojban?hl=en.
>
> References
>
> Visible links
> 1. mailto:get....@gmail.com
> 2. mailto:lindar...@yahoo.com
> 3. mailto:loj...@googlegroups.com
> 4. mailto:lojban%2Bunsu...@googlegroups.com
> 5. http://groups.google.com/group/lojban?hl=en
> 6. mailto:loj...@googlegroups.com
> 7. mailto:lojban%2Bunsu...@googlegroups.com
> 8. http://groups.google.com/group/lojban?hl=en
--
.i ko djuno fi le do sevzi
Remo Dentato
And there is at least 1 good reason for lojban being an auxlang (at least at some point in the future after it's been completed). One of the design goals was cultural neutrality. Not many conlangs and certainly no natlangs have been designed with that in mind I think
To me, what is more important is that IT WORKS!
Many conlangs are just more or less simplified "copies" of natural language or they are just toy languages and can't really match the expressiveness of natural languages.
Lojban seems to be the only conlang that is *completely" different from any natural language and still can be used for human communication.
The problem we have, I think, is that we have been given a too much powerful language that we are still no able to handle.
Ivo Doko
I'm -not- going to vote for lojban. It has not been and definitely
isn't now a candidate auxlang.
Robin Lee Powell
person I decided to respond to. Lots of people have said similar
things.
"very far"? *Really*??
Y'all have weird standards/requirements. Lojban is *FAR* more fully
defined than Esperanto.
No, really: it is. Esperanto doesn't have a formal grammar of any
kind, for starters.
We know far more about how Lojban grammatical structures work than
*any other actually spoken language on the planet*.
We have already won that prize: Lojban is the most precisely,
formally specified language that there is, for any language with its
number of speakers or higher. Period. I challenge anyone to find
anything even *remotely close* to the CLL in terms of covering every
*possible* grammatical combination. Even if you can find such a
thing, the formal grammar takes it so far ahead of everything else
they can't possibly hope to catch up.
That's "fully defined"; now for "complete and functioning".
The reason that Lojban *seems* flaky is:
1. When people have trouble saying something in Esperanto, they
simply import a word or phrase or grammatical structure from a
natlang, and everyone's OK with this. Current Lojban culture
refuses to do that.
As a historical note, this *was* acceptable, to some extent,
around the time the CLL was published, which I think was why they
thought the language was all-the-way-done. See noralujv if you
don't believe me; there are some natlang imports there that would
cause most current Lojbanists to scream.
2. We all are a bunch of picky, whiny, geeks (that is intended more
as a factual description rather than an insult), so our response to
"huh, I don't see how to do that" tends to be, since we can't just
import a natlang solution, "OMG LODGEBANS ARE
BROKEN!!!!!1!!one!!11!!cos(0)!1!".
3. We all have a tendency to do this *before* actually learning the
language all that well. The truth of the matter is that you really
*can* say anything you want in Lojban; LNC and alis prove that
pretty conclusively, I think. Most of the "problems" that people
freak out about are already well understood by oldbies.
4. Nobody shouts "Wow this is well specified!!!" at the top of
their lungs, but they certainly shout their complaints. Geeks have
a shared culture that compliments are private and insults are
public; it's deeply fucked up. See
http://lesswrong.com/lw/3h/why_our_kind_cant_cooperate/
I can say anything I need to say in Lojban, modulo my own vocabulary
knowledge. This puts it ahead of 99.999% of conlangs. Saying that
it is very far from being complete and functioning is ridiculous,
and pretty insulting to a lot of people's hard work.
-Robin
--
http://singinst.org/ : Our last, best hope for a fantastic future.
Lojban (http://www.lojban.org/): The language in which "this parrot
is dead" is "ti poi spitaki cu morsi", but "this sentence is false"
is "na nei". My personal page: http://www.digitalkingdom.org/rlp/
Robin Lee Powell
I think that Lojban would be decent for this, simply because it's
equally hard for everybody to learn :D (ignoring the lack of
learning materials in various languages).
... 0.o
Why do they have Loglan on their list? (and not Lojban, in
addition!)
It seems like a silly bit of easy publicity; I wrote Lojban in.
-_- They don't seem to be collating writeins. *sigh*
Matt Arnold
even when there is an infinity of work left to be done in its
universe. Lojban has been a finished language for years. I drifted
away and lost interest not because of the language itself, but the
culture of perfectionism.
-Eppcott
Remo Dentato
Thank you, Robin. You're absolutely right. A language can be finished
even when there is an infinity of work left to be done in its
universe.
Lojban has been a finished language for years. I drifted
away and lost interest not because of the language itself, but the
culture of perfectionism.
I agree with Robin too. As I said in that thread: LOJBAN WORKS!
If I can't see how to say something it's because there are things that I don't know, not because of flaw in the language.
Just by translating things into lojban I learned a lot and I'm firmly convinced that Lojban expressiveness is equivalent (if not superior) to natural languages.
I feel it's just that I don't know how to use all the feature of the language.
I'm not against exploring how much {si} will make the universe collapse in a black hole. It may be fun and entertaining, but I'm much more interested in understanding real cases of use of the language.
I would feel better, however, if more people would start *using* it rather than just exploring the sharp corners for the sake of doing it.
Alex Rozenshteyn
--
You received this message because you are subscribed to the Google Groups "lojban" group.
To post to this group, send email to loj...@googlegroups.com.
To unsubscribe from this group, send email to lojban+un...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/lojban?hl=en.
--
Luke Bergen
Matt Arnold
najrut
it's native numbering system is cumbersome) Hence, this language can't
be international.... Nonsense !
Lojban doesn't have a word for let's say some device. But it has such
magic words as la'o and la'oi. Does Esperanto have such words ?
> Why do they have Loglan on their list? (and not Lojban, in
addition!)
know about Lojban. Let them read Wikipedia ;-)
On Jan 5, 7:56 pm, Robin Lee Powell <rlpow...@digitalkingdom.org>
wrote:
> > 'The Future Global Language' site has a voting system for different
> > languages that can be international. Please vote for Lojban.
> > ------------------------
> >http://international-languages.webs.com/
>
> I think that Lojban would be decent for this, simply because it's
> equally hard for everybody to learn :D (ignoring the lack of
> learning materials in various languages).
>
> ... 0.o
>
> Why do they have Loglan on their list? (and not Lojban, in
> addition!)
>
> It seems like a silly bit of easy publicity; I wrote Lojban in.
>
> -_- They don't seem to be collating writeins. *sigh*
>
> -Robin
>
Michael Turniansky
> Chinese uses latin numbers in their writing system, doesn't it ? (as
> it's native numbering system is cumbersome) Hence, this language can't
> be international.... Nonsense !
>
No they use Chinese numbers, 一 二 三 四 五etc, or they use Arabic
numerals, 1, 2, 3, 4, 5... but they don't use Latin numbers (I II III
IV V), as far as I know... :-P
--gejyspa
Robin Lee Powell
> seconded. Totally agree. Awesome rant.
>
> But I'm really replying because... you read lesswrong too robin?!
> That's awesome! I just discovered that site like a month or two
> ago and have been steadily making my way through the sequences.
> Awesome site.
I'm not so much a lesswrong reader as an Eliezer Yudkowsky cultist;
see
http://teddyb.org/robin/tiki-index.php?page=My+Views+On+The+Future
Ivo Doko
Saying that
it is very far from being complete and functioning is ridiculous,
and pretty insulting to a lot of people's hard work.
Also, I've realised I made a typo: I meant to say that Esperanto is "far from being completely regular and logical", not that it's "far from being completely *ir*regular and logical", sorry if that confused anybody.
As for the title/subject, I never said that lojban is *broken*, pardon me, I only said that it isn't finished, which is not something I made up - I constantly read people both newbies and oldies ranting on about how there's so much work yet to be done on lojban and how every now and then someone will take upon him-/herself to finish whatever is left unfinished and it will look like he/she will accomplish it but then he/she gives up and yadda yadda. To me that seems like lojban is, to put it as simple as possible, not finished. Esperanto, on the other hand, is. Were you trying to say that Esperanto isn't finished either? It may not be as "fully defined" as lojban is, but that doesn't really say much. Esperanto never aimed to be what lojban aims to be - a completely logical and fully unambiguously defined language. Instead, Esperanto aimed to be a language which is as unambiguous and as regular as it can be while still operating like a naturally-evolved language as much as possible. In order for a language to be like that, it doesn't have to be as fully defined as lojban does in order to be finished, which is the reason why Esperanto is a finished language while lojban is not. Even though lojban is better defined than Esperanto, it's not as fully defined as it should be in order for it to be finished, because the current level of its well-definiteness is not good enough for what lojban aims to be.
Pierre Abbat
Esperanto has at least one word which proves that its words cannot be
unambiguously parsed: "avaro" is derived from the same root as "avarice", but
collides with "avaro", from "avo" (grandfather) and the suffix "ar". Also the
relation among the noun "brodo", the adjective "broda", and the verb "brodi"
depends on which was formed first, unlike the corresponding rule in Lojban,
which is that "le" or "lo" always takes something that fits or is described
as fitting in x1 of the selbri.
The main thing that Lojban lacks for being used as a global language is not
the precise definition of every corner case. It's vocabulary. And because its
morphology is defined so as to prevent collisions like "avaro", it takes
longer to invent vocabulary in Lojban. You can't take some Latinate term
that's commonly used in many languages, some of them unrelated to Latin, and
expect to make a brivla out of it just by changing "-us" to "-o". You have to
consider whether a lujvo would capture the meaning better, whether the second
consonant is in a cluster, and whether the same word could mean something
totally different (such as "malpigi" which could be either an acerola fruit
or an insect's kidney).
Pierre
--
I believe in Yellow when I'm in Sweden and in Black when I'm in Wales.
Robin Lee Powell
> On 5 January 2011 17:52, Robin Lee Powell <rlpo...@digitalkingdom.org>wrote:
>
> > Saying that
> > it is very far from being complete and functioning is ridiculous,
> > and pretty insulting to a lot of people's hard work.
>
> only said that it isn't finished,
You did, actually:
it's a fully defined, complete and functioning language, which
lojban is very far from being at the moment
"far from functioning" is very different from "incomplete".
> which is not something I made up - I constantly read people both
> newbies and oldies ranting on about how there's so much work yet
> to be done on lojban
This is why I started by saying that this isn't really about you.
There are a lot of things that can and should be done aronud here,
but very few of them have to do with making the language *complete*.
It's plenty functional as is, it's more about finishing the
codification of what is. Even things like xorlo were, overall,
extremely minor; it changed very little about how the language is
actually spoken. The only thing of any real significance was lo's
inner quentifier.
And yet people act like it's hugely full of holes and millions of
things need to be changed, and on and on. It's just not true.
I'm not saying *you* did anything wrong, because as you said people
say things like this all the time. It's a generic problem 'round
here.
> To me that seems like lojban is, to put it as simple as possible,
> not finished.
I think that that's fair; it's just not what I heard you say.
> Esperanto, on the other hand, is. Were you trying to say that
> Esperanto isn't finished either? It may not be as "fully defined"
> as lojban is, but that doesn't really say much.
That was all I was saying, yes. I had no intention of asserting
that Esperanto wasn't finished, which is why I divided my comments
into two sections. You said that Lojban was far from being fully
defined; I was responding to that. Then later I responded to you
saying that it is "very far" from complete or functioning. It is
not, in fact, very far from either of those.
"unfinished" yes, but there is a sense in which it will *never* be
finished; all languages change and grow.
"very far from finished", no.
Ivo Doko
Esperanto has at least one word which proves that its words cannot be
unambiguously parsed...
There are multiple, but that is irrelevant. Like I said, Esperanto never even aimed to be fully unambiguous and as thousands of languages worldwide (Esperanto included, because it has native speakers) prove, a language doesn't *need* to be fully unambiguous to be a usable and working language.
The main thing that Lojban lacks for being used as a global language is not
the precise definition of every corner case. It's vocabulary.
I.e. it's not finished, which is what I said.
...its morphology is defined so as to prevent collisions like "avaro", it takes
longer to invent vocabulary in Lojban. You can't take some Latinate term
that's commonly used in many languages, some of them unrelated to Latin, and
expect to make a brivla out of it just by changing "-us" to "-o". You have to
consider whether a lujvo would capture the meaning better, whether the second
consonant is in a cluster, and whether the same word could mean something
totally different (such as "malpigi" which could be either an acerola fruit
or an insect's kidney).
Speaking of which, I think that, unfortunately, is the main flaw of lojban. I understand that it can't possibly hope to be literally unambiguous if its vocabulary doesn't operate like that, but that ensures that if people ever do start to use lojban for everyday communication and if lojban ever gets native speakers, its so praised unambiguity will very soon melt away. Vocabulary assimilation is unavoidable and you can't possibly expect every native speaker of lojban to know which new brivla will create an ambiguity, so native lojban speakers would naturally start to incorporate words from other languages in their vocabulary, those words would inevitably create ambiguities, and after a couple of decades its precious ambiguity would be nowhere. (And that's without even mentioning other ways in which a language evolves when it's used by people as their main language for everyday communication.)
So... as far as I've understood it, this is how it goes:
1) Let's make lojban the world's official common language because it's completely logical and unambiguous.
2) lojban is made the world's official common language.
3) People use lojban every day to talk to each other.
4) As was the case with Esperanto, this eventually results in people having lojban as their native language, who proceed to use lojban as their main language for everyday communication.
5) This makes lojban evolve.
6) After a couple of decades, lojban is no longer unambiguous nor completely logical and as time goes by is more and more like languages which have naturally evolved among humans.
Wait, so what was the initial reason to use lojban as the world's official common language? After all, lojban's unambiguity and logicality seems to be one of the main arguments for that, and yet if it did get chosen for that role it will have stopped being unambiguous and logical not long after its use became widespread. So if we're going to have an "ordinary" language as the world's official common language in the end anyway, why not chose one which is not unfinished?
Ivo Doko
> As for the title/subject, I never said that lojban is *broken*, pardon me, IYou did, actually:
> only said that it isn't finished,
lojban is very far from being at the moment
it's a fully defined, complete and functioning language, which
"far from functioning" is very different from "incomplete".
"[Esperanto is] a fully defined, complete and functioning language, which
lojban is very far from being at the moment."
So what I *have* said about lojban, is that it's "far from being a fully defined, complete and functioning language", not just that it's "far from functioning".
I have to say that I'm disappointed and surprised at the same time at the irony of having my words so terribly misinterpreted and misunderstood by a lojban enthusiast.
Jim Carter
> ...
> Vocabulary assimilation is unavoidable and you can't possibly expect every
> native speaker of lojban to know which new brivla will create an ambiguity,
> so native lojban speakers would naturally start to incorporate words from
> other languages in their vocabulary, those words would inevitably create
> ambiguities, and after a couple of decades its precious ambiguity would be
> nowhere. (And that's without even mentioning other ways in which a language
> evolves when it's used by people as their main language for everyday
> communication.)
I haven't been doing too much lately with Lojban, but back in the Loglan
days I translated about 10,000 words of text into Loglan while creating
only four new brivla (torus, to use in "bagel"; noodle; bear (the animal,
which wasn't in old Loglan); and oar). All the rest could be represented
by lujvo, if I looked carefully in the word list. I expect that brivla
will be coined rarely once the language is "complete", but novel lujvo will
coruscate off the tongues of the lojbanistani. Particularly if we pay
attention to compatible definitions and reliable combining rules.
Your point is well taken that languages drift, particularly if used by the
workers, peasants and soldiers. Even so, the scope for drift in grammar
and syntax is limited because the grammar is so well defined. I would
expect to see more drift in phoneme sounds and in the usage of words, just
as we see in natural languages such as English.
As for making definitions unambiguous, words are defined circularly by
brief texts made of other words which are defined in terms of the first
one. I think we don't really have a good handle on making our definitions
provably unambiguous; instead, our goal has been to make them less
ambiguous than typical natlangs, relying on a human's nonlogical ability to
understand what a word means when given a few cues and examples.
I think it's less useful to ask whether Lojban or Esperanto is "finished",
and more useful to ask if it can do the job proposed for it: use as an
auxlang. I think both languages get a passing grade on that point.
James F. Carter Voice 310 825 2897 FAX 310 206 6673
UCLA-Mathnet; 6115 MSA; 520 Portola Plaza; Los Angeles, CA, USA 90095-1555
Email: ji...@math.ucla.edu http://www.math.ucla.edu/~jimc (q.v. for PGP key)
Jorge Llambías
>
> I haven't been doing too much lately with Lojban, but back in the Loglan
> days I translated about 10,000 words of text into Loglan while creating only
> four new brivla (torus, to use in "bagel"; noodle; bear (the animal,
> which wasn't in old Loglan); and oar). All the rest could be represented
> by lujvo, if I looked carefully in the word list.
In the "lo nu binxo" translation (26.565 words) the only brivla I used
were gismu, no lujvo and no fu'ivla. I did it sort of as an
experiment, so see if it could be done, and I was quite pleased with
the result.
(Actually, I did use one lujvo and one fu'ivla, anyone interested in
knowing which ones will have to read the translation :)
http://www.lojban.org/tiki/tiki-index.php?page=lo+nu+binxo
mu'o mi'e xorxes
Ivo Doko
I haven't been doing too much lately with Lojban, but back in the Loglan days I translated about 10,000 words of text into Loglan while creating only four new brivla (torus, to use in "bagel"; noodle; bear (the animal,
which wasn't in old Loglan); and oar). All the rest could be represented
by lujvo, if I looked carefully in the word list. I expect that brivla
will be coined rarely once the language is "complete", but novel lujvo will
coruscate off the tongues of the lojbanistani. Particularly if we pay
attention to compatible definitions and reliable combining rules.
I'm not talking about how lojban native speakers will come up with new words for things they don't have a word for. What I'm talking about is that speaker(s) of lojban will be introduced to a new invention/concept/thing which will have been named by people who don't speak lojban (but, for example, English) and lojban speakers will like the name those people have given it and will thus simply incorporate that word in their vocabularies and, thus, in lojban. It's just like French "écran" and English "software" got incorporated into Serbo-Croatian as "ekran" and "softver". Sure, purists didn't like that and invented replacement words, namely "zaslon" and "omekšje", respectively, but those words are simply not used and have failed to replace "ekran" and "softver" and these two have become a part of Serbo-Croatian vocabulary. Same thing would happen with lojban - purists would invent lujvo (or brivla) to replace the direct loanwords in order to leave the language's unambiguity intact, but people who don't care about whether the language is completely unambiguous or not (who would, mind you, make up a great majority of lojban speakers if it did become world's official common language) would not cease to use the loanwords in place of the new "proper" words and lojban would get screwed up pretty quick.
Of course, you could say that lojban is what a special committee of purists says it is and that people who don't use only the words which have been approved by the committee don't speak lojban, but no one would agree to make such a fascistic language the world's official common language and even if they did no one would give a crap what the committee says and lojban would still be what is spoken and not what is approved.
Luke Bergen
Jonathan Jones
--
mu'o mi'e .aionys.
.i.a'o.e'e ko cmima le bende pe lo pilno be denpa bu .i doi.luk. mi patfu do zo'o
(Come to the Dot Side! Luke, I am your father. :D )
Jonathan Jones
actually I read it the same way.Esperanto is X, Y, and Z. lojban is far from being [that] at the moment.=lojban is far from X, Y, AND Z.=lojban is far from X AND lojban is far from Y AND lojban is far from Z.If that's not what you meant to communicate maybe you should try saying it in lojban next time ;)
Yes, we Lojbanists do tend to think lojbanically about logic - thata's actually why I titled my recent poll-thread "taxonomic .a descriptive" instead of "taxonomic or descriptive", because English "or" is equivalent to {.onai}.
--On Wed, Jan 5, 2011 at 6:11 PM, Ivo Doko <ivo....@gmail.com> wrote:
You have misquoted me. My words were, as you quoted them correctly the first time:On 5 January 2011 23:05, Robin Lee Powell <rlpo...@digitalkingdom.org> wrote:
> As for the title/subject, I never said that lojban is *broken*, pardon me, IYou did, actually:
> only said that it isn't finished,
lojban is very far from being at the moment
it's a fully defined, complete and functioning language, which
"far from functioning" is very different from "incomplete".
"[Esperanto is] a fully defined, complete and functioning language, which
lojban is very far from being at the moment."
So what I *have* said about lojban, is that it's "far from being a fully defined, complete and functioning language", not just that it's "far from functioning".
I have to say that I'm disappointed and surprised at the same time at the irony of having my words so terribly misinterpreted and misunderstood by a lojban enthusiast.
--
You received this message because you are subscribed to the Google Groups "lojban" group.
To post to this group, send email to loj...@googlegroups.com.
To unsubscribe from this group, send email to lojban+un...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/lojban?hl=en.
You received this message because you are subscribed to the Google Groups "lojban" group.
To post to this group, send email to loj...@googlegroups.com.
To unsubscribe from this group, send email to lojban+un...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/lojban?hl=en.
Jonathan Jones
Of course, we don't even really need the BPFK for something like borrowed words - the grammar itself does most of the "what-is-and-is-not-lojban" all on it's own, and no one who fluently speaks Lojban (such as, for example, native speakers) would purposely invent a word that doesn't follow Lojban grammar rules.
Ivo Doko
actually I read it the same way.Esperanto is X, Y, and Z. lojban is far from being [that] at the moment.=lojban is far from X, Y, AND Z.=lojban is far from X AND lojban is far from Y AND lojban is far from Z.If that's not what you meant to communicate maybe you should try saying it in lojban next time ;)
Wow, you guys need to learn your logic. Let's do it properly:
A = "lojban is fully defined."
B = "lojban is complete."
C = "lojban is a functioning language."
"lojban is not a fully defined, complete and functioning language" can be written as:
¬(A ∧ B ∧ C)
which is equivalent to:
¬A ∨ ¬B ∨ ¬C
which is:
"lojban is not fully defined, or lojban is not complete, or lojban is not a functioning language."
Jonathan Jones
On 6 January 2011 01:40, Luke Bergen <lukea...@gmail.com> wrote:actually I read it the same way.Esperanto is X, Y, and Z. lojban is far from being [that] at the moment.=lojban is far from X, Y, AND Z.=lojban is far from X AND lojban is far from Y AND lojban is far from Z.If that's not what you meant to communicate maybe you should try saying it in lojban next time ;)
Wow, you guys need to learn your logic. Let's do it properly:
A = "lojban is fully defined."
B = "lojban is complete."
C = "lojban is a functioning language."
"lojban is not a fully defined, complete and functioning language" can be written as:
¬(A ∧ B ∧ C)
which is equivalent to:
¬A ∨ ¬B ∨ ¬C
which is:
"lojban is not fully defined, or lojban is not complete, or lojban is not a functioning language."
.i la.lojban. na mulno smugau gi'a na mulno gi'a na tolpo'u bangu
jboski parse
Like I said, we tend to think Lojbanically about logic. Notice how the two sentences, translated into Lojban, mean entirely different things.
Ivo Doko
Jonathan Jones
Wrong on both counts.
Where:
A = "lojban is fully defined."
B = "lojban is complete."
C = "lojban is a functioning language."
"lojban is not fully defined, or lojban is not complete, or lojban is not a functioning language." = "lojban is (not A) or (not B) or (not C)." = {.i la.lojban. na mulno smugau gi'a na mulno gi'a na tolpo'u bangu}
Craig Daniel
Or (as in .a/vel) *English* is not logical, and while a logician's OR
means one thing, the pragmatics of English are such that we often use
"or" in a very different way that does not logically mean the same,
making some sentences ambiguous. Their translations into Lojban are
unambiguous, and thus differ from one reading of the English to
another; in this context, there is nothing surprising about some
people reading the (non-logical) English one way and some reading it
the other.
- mi'e .kreig.
Ivo Doko
Wrong on both counts.
Where:"lojban is not a fully defined, complete and functioning language" = "lojban is not A, B, and C." = "lojban is not (A, B, and C)." = {.i la.lojban. na mulno smugau je mulno je tolpo'u bangu}
A = "lojban is fully defined."
B = "lojban is complete."
C = "lojban is a functioning language."
"lojban is not fully defined, or lojban is not complete, or lojban is not a functioning language." = "lojban is (not A) or (not B) or (not C)." = {.i la.lojban. na mulno smugau gi'a na mulno gi'a na tolpo'u bangu}
In any case, this is what I meant:
"It is not true that lojban is fully defined and that lojban is complete and that lojban is a functioning language."
That *must* be equivalent to:
Jonathan Jones
Regardless, what you meant is not what you said. Something that is easy to do in English, but rather difficult in Lojban.
Waiter: Do you want cream of sugar in your coffee?
Lojbanist: Yes.
Jonathan Jones
I think it does, but I am not proficient enough to translate it, so we're stuck on assuming.On Wed, Jan 5, 2011 at 7:11 PM, Ivo Doko <ivo....@gmail.com> wrote:If A = "lojban is fully defined." then "lojban is not A" means "lojban is not "lojban is fully defined."." Decide what A, B and C mean.On 6 January 2011 03:02, Jonathan Jones <eye...@gmail.com> wrote:Wrong on both counts.
Where:"lojban is not a fully defined, complete and functioning language" = "lojban is not A, B, and C." = "lojban is not (A, B, and C)." = {.i la.lojban. na mulno smugau je mulno je tolpo'u bangu}
A = "lojban is fully defined."
B = "lojban is complete."
C = "lojban is a functioning language."
"lojban is not fully defined, or lojban is not complete, or lojban is not a functioning language." = "lojban is (not A) or (not B) or (not C)." = {.i la.lojban. na mulno smugau gi'a na mulno gi'a na tolpo'u bangu}
In any case, this is what I meant:
"It is not true that lojban is fully defined and that lojban is complete and that lojban is a functioning language."
That *must* be equivalent to:
"lojban is not fully defined, or lojban is not complete, or lojban is not a functioning language."
Regardless, what you meant is not what you said. Something that is easy to do in English, but rather difficult in Lojban.
Waiter: Do you want cream of sugar in your coffee?
Lojbanist: Yes.
Craig Daniel
> In any case, this is what I meant:
> "It is not true that lojban is fully defined and that lojban is complete and
> that lojban is a functioning language."
>
> That *must* be equivalent to:
> "lojban is not fully defined, or lojban is not complete, or lojban is not a
> functioning language."
No, because "and" binds in two different ways in English when negators
get involved. ("Or" behaves similarly, and worse it also gets used for
both a logical OR and a logical XOR, with both meanings carrying the
same syntactic ambiguity. Natural languages misbehave like that.) It
*must* be equivalent to either that or (.onai/aut):
"lojban is not fully defined, and lojban is not complete, and lojban
is not a functioning language."
Which is not the meaning you intended, and not the one I personally
got from it, but in context I'm not surprised others read it that way.
Moral: Lojban is better at these things. Absent the clarifying
mediation of tone of voice (and sometimes even with it), it can be
useful to add the missing specificity in English - though it often
takes hindsight to realize when, pragmatically speaking, it wasn't
filled in by context. (Secondary moral: pragmatics is an important
area of study for Lojbanists, even though it's less precisely defined
- or definable - than syntax is. English pragmatics do an amazing job
of filling in some of the holes left by syntax.)
- mi'e .kreig.daniyl.
Ivo Doko
Waiter: Do you want cream or sugar in your coffee?
Lojbanist: Yes.
Proper response is "I do." :D Unless you drink your coffee with neither...
Also, that would depend on the waiter's emphasis of the sentence:
"Do you want CREAM or SUGAR in your coffee?"
means:
"Is this true: (You want cream in your coffee. ⊕ You want sugar in your coffee.)?"
"Do you WANT cream or sugar in your coffee?"
means:
"Is this true: (You want cream in your coffee. ∨ You want sugar in your coffee.)?"
("⊕" is XOR, "∨" is OR.)
So presuming the waiter emphasised "cream" and "sugar", not "want", "yes" or "I do" are not proper responses.
Robin Lee Powell
> On 5 January 2011 23:05, Robin Lee Powell
> <rlpo...@digitalkingdom.org>wrote:
>
> > > As for the title/subject, I never said that lojban is
> > > *broken*, pardon me, I only said that it isn't finished,
> >
> > You did, actually:
> >
> > it's a fully defined, complete and functioning language, which
> > lojban is very far from being at the moment
> >
> > "far from functioning" is very different from "incomplete".
>
>
> You have misquoted me.
That's a cut and paste, so I find that unlikely.
> My words were, as you quoted them correctly the first time:
>
> "[Esperanto is] a fully defined, complete and functioning
> language, which lojban is very far from being at the moment."
>
> So what I *have* said about lojban, is that it's "far from being a
> fully defined, complete and functioning language", not just that
> it's "far from functioning".
In the dialect of English I speak, "A is far from being X, Y, and Z"
means the same thing as "A is far from being X" and "A is far from
being Y" and "A is far from being Z".
If that's not what you meant, I apologize for responding to
something other than what you intended. That really is the *only*
meaning I can see from the English, though.
Feel free to use Lojban next time.
Just turned to the person next to me: "If I say, "that computer is
far from pretty and functional and cheap", what have I just accused
it of?". The response: "That it's ugly and broken and expensive."
This is even more extreme, to me, when it's "A is X, Y, and Z, which
B is far from being at the momment". I'm really surprised that a
native English speaker would take that to mean something other than
"B is not X" and "B is not Y" and "B is not Z".
This is why we're hear, because English does that.
> I have to say that I'm disappointed and surprised at the same time at the
> irony of having my words so terribly misinterpreted and misunderstood by a
> lojban enthusiast.
Again, I'm sorry for responding to something other than what you
said, but you have no basis for acting like your interpretation is
*clearly* correct, as so far more people have disagreed than agreed.
So, let's just move on, shall we.
I'm tempted to ask what you *were* trying to say, if it wasn't that,
but I don't actually *care*; I had a point to make about the
generally pessimistic and perfectionist attitude in this community,
and I've made it, and I'm pretty much done at this point.
xaujbes cindustus
tjgi...@gmail.com
I'm -not- going to vote for lojban. It has not been and definitely
isn't now a candidate auxlang. If it -does- get upvoted, people are
going to ask why it would make a good auxlang, and we can give no
legitimate response because it wouldn't be. Personally, I don't want
hundreds of douchebags every day asking why we're the best auxlang and
asking why we think we're better than $popular_auxlang.
John E Clifford
And, of course, finally, who the hell want to propose Lojban for an auxlang? That is way outside its design specs (not that it has done that good a job at meeting those specs, but that doesn't certify it for some other purpose instead).
From: Ivo Doko <ivo....@gmail.com>
To: loj...@googlegroups.com
Sent: Wed, January 5, 2011 4:55:38 PM
Subject: Re: [lojban] Re: Lojban is *NOT* broken! Stop saying that!
John E Clifford
Subject: Re: Lojban is *NOT* broken! Stop saying that! (was Re: [lojban] Re: Vote for the Future Global Language)
John E Clifford
From: Jonathan Jones <eye...@gmail.com>
To: loj...@googlegroups.com
Sent: Wed, January 5, 2011 7:14:49 PM
Subject: Re: Lojban is *NOT* broken! Stop saying that! (was Re: [lojban] Re: Vote for the Future Global Language)
Craig Daniel
> imho esperanto is a better aux lang
The only useful auxlang is the one people use.
Right now, that's English, for all its flaws. Two centuries ago it was
French, and a few centuries before that it was Sabir in many places. A
few centuries in the future, who knows?
Michael Turniansky
Personally, I neither want (cream) nor (sugar in my coffee). Cause
cream is kinda silly all by itself, and I don't drink coffee, so
putting sugar in something that I wouldn't drink anyway is also kind
of silly.
--gejyspa
Remo Dentato
Remo Dentato
Even of not so often used, in Italian, roman numerals are used to mean
ordinals so that XIV means "the 14th". I'm not sure about French but I
suspect they do the same
najrut
As to la'o, la'oi etc. the point is that there cannot be only one language on the Earth. We have numerous languages including languages of animals, computer internal languages (machine codes, not BASIC and C++), scientific notation, language of traffic signs...
Lojban as an auxiliary lang can and must interact with those languages but not replace them.
This is the power of Lojban. And lack of this is one of the biggest flaws of the other langs including Esperanto (which is actually an Indoeuropean one).
Krzysztof Sobolewski
> Lojban is a proscribed language. So, call it fascist if you want, but that's exactly how it works: it the BPFK doesn't approve it, it isn't Lojban.
And there's at least one high-profile example of a proscribed natlang: Sanskrit. It was proscribed more than 2000 years ago and AFAIK it's still the same language as then.
Of course the general populace evolved it into a myriad of different natlangs, just like Latin evolved into many Romance languages. But Latin is mostly dead, while Sanskrit is mostly still alive.
AFAIROW[1], of course :)
[1] As Far As I Read On Wikipedia
--
Ecce Jezuch
"Refuse to be denied refuse to compromise
Your ideology always a lie
It's jingoistic or nationalistic
Or whatever the hell you want to call it" - J. Bush
Krzysztof Sobolewski
> I don't think anyone uses Latin numerals anymore, with the exception of
> dating (as in movies.) Arabic's base-10 system is so much better. :D
As a side note, we call them "Arabic" but they were actually developed in India :)
--
Ecce Jezuch
"Children today get trophies for everything from sports to academics to
drive-by's. And what's more: they get awards when they don't win the
competition, or even when they clearly lost. Because there is understanding
in our society that our children need more emotional support for their
self-esteem. But emotional support is difficult, so we give them awards
instead." - Chet Haase
Robert LeChevalier
> The main thing that Lojban lacks for being used as a global language
> is not
> the precise definition of every corner case. It's vocabulary.
>
>
> I.e. it's not finished, which is what I said.
By that reasoning, no language is "finished", since no language has
vocabulary that is not continuously being added to, and no language has
definitions for all of its words (and even if it did, no speaker
actually KNOWS all the words and their definitions).
> Speaking of which, I think that, unfortunately, is the main flaw of
> lojban. I understand that it can't possibly hope to be literally
> unambiguous if its vocabulary doesn't operate like that, but that
> ensures that if people ever do start to use lojban for everyday
> communication and if lojban ever gets native speakers, its so praised
> unambiguity will very soon melt away.
That is entirely unclear.
>Vocabulary assimilation is unavoidable
Vocabulary assimilation into new languages can and does follow rules.
If the rules are clear and simple, people will tend to follow them.
This is one reason why I oppose making many Type IV fu'ivla until the
language is well-established. When people assimilate vocabulary, they
should habitually assimilate it as Type III, which rules are very easy
to follow
> and you can't possibly expect every native speaker of lojban
> to know which new brivla will create an ambiguity,
Well, very young kids might not, but adult speakers should be able to
manage Type III fu'ivla borrowing
> so native lojban
> speakers would naturally start to incorporate words from other languages
> in their vocabulary,
As long as they do so using Type I-III borrowing, problems are unlikely.
> those words would inevitably create ambiguities,
Obviously, I disagree.
There may be the occasional erroneous creation, but errors can be
corrected. "Refudiate" has not yet become a standard English word.
> So... as far as I've understood it, this is how it goes:
>
> 1) Let's make lojban the world's official common language because it's
> completely logical and unambiguous.
It isn't, so that isn't the reason.
It is MORE logical and MORE unambiguous in certain ways, and MORE
culturally neutral as well. It is also easier to learn than any natlang.
> 2) lojban is made the world's official common language.
Ain't gonna happen. I'd be happy with it having significant official
use, and UNofficial lingua franca status for informal communication
> 3) People use lojban every day to talk to each other.
That would be nice.
> 4) As was the case with Esperanto, this eventually results in people
> having lojban as their native language,
That part will happen
> who proceed to use lojban as their main language for everyday
communication.
My understanding is that the bulk of native Esperantists are
multilingual and do NOT use Esperanto as their "main language"
> 5) This makes lojban evolve.
Languages do evolve. The nature and rules for such evolution aren't
really known, and have not been much studied in the case of artificial
languages
> 6) After a couple of decades, lojban is no longer unambiguous
It already isn't. It has never been more than morphologically and
syntactically unambiguous.
> nor completely logical
Lojban has never been "completely logical" and never tried to be. It,
however, arguably ENABLES many forms of "completely logical" speech if
such is desired.
> and as time goes by is more and more like languages
> which have naturally evolved among humans.
This is a prediction which has no scientific basis. No one could
possibly know how Lojban would evolve as a quasi-native language.
> Wait, so what was the initial reason to use lojban as the world's
> official common language?
Not what you started with.
> After all, lojban's unambiguity and logicality
> seems to be one of the main arguments for that,
One of the arguments - but probably not the main one except in certain
domains. Cultural neutrality and simplicity and completeness of
specification all rate quite highly as arguments.
> and yet if it did get
> chosen for that role it will have stopped being unambiguous and logical
> not long after its use became widespread.
"not long"? is unsupported by any research. Languages do evolve, but
they evolve slowly. People do still understand Shakespeare after 400+
years.
> So if we're going to have an
> "ordinary" language as the world's official common language in the end
> anyway, why not chose one which is not unfinished?
Because there is no such language, and all of the other plausible
candidates have greater flaws than merely being "unfinished".
Overall, the argument is that if Lojban cannot be perfect, then it is
not only no better than other languages, it is worse than them.
lojbab
Oleksii Melnyk
word for. What I'm talking about is that speaker(s) of lojban will be introduced to a new invention/concept/thing which will have been named by people who don't speak lojban (but, for example, English) and lojban speakers will like the name those people have given it and will thus simply incorporate that word in their vocabularies
And lojban designers have already solved that problem. Most of such thing can just fit the place right to zo la , almost unchanged. So, everyone will be happy.
--
mu'o mi'e lex
Robert LeChevalier
> - I constantly read people both newbies and oldies ranting on about how
> there's so much work yet to be done on lojban
This is a relative thing.
Much of the "work yet to be done" consists of completing the originally
promised set of language documentation (something I promised back in
1989 or so, but was unable to deliver myself). The job is too big for
one or two people. It is NOT too big for the current community, but
will still take a lot of time because we don't have any paid, full-time,
workers (whereas most published dictionaries are made by a small army of
workers using specially designed software).
If we has 100 full time workers skilled in the language, and knew how to
use them effectively, the "so much work" would disappear rather quickly.
We don't, so it takes longer, and it seems that progress is extremely
slow.
And yet, how long does it take OED to put out a new edition, even with
and army of workers: decades. We should be able to do better, because
Lojban is not English, and we don't need an OED-equivalent to have a
useful language.
> and how every now and then
> someone will take upon him-/herself to finish whatever is left
> unfinished and it will look like he/she will accomplish it but then
> he/she gives up and yadda yadda. To me that seems like lojban is, to put
> it as simple as possible, not finished.
No. It means that it maintains higher standards than its current
all-volunteer work crew has been able to manage. The documentation lags
the reality.
> Esperanto, on the other hand, is.
No it isn't. No language is "finished" until it is a dead language.
> Were you trying to say that Esperanto isn't finished either? It may
> not be as "fully defined" as lojban is, but that doesn't really say
> much.
Actually, it says a lot.
> Esperanto never aimed to be what lojban aims to be - a completely
> logical and fully unambiguously defined language.
Lojban never aimed to be that either. It is a false ideal, indeed a
straw man.
> Instead, Esperanto
> aimed to be a language which is as unambiguous and as regular as it can
> be while still operating like a naturally-evolved language as much as
> possible.
It never achieved that either.
> In order for a language to be like that, it doesn't have to be
> as fully defined as lojban does in order to be finished,
Since no language can be finished, the argument is a waste.
The question is whether the language is well-enough defined that a
novice can come in, pick up the language materials and study them, and
come out speaking passable Lojban within a reasonable amount of time.
This has been demonstrated, but so far only with bright and highly
motivated people. Some who are less motivated, demand a more formal set
of documentation in order to be so-motivated, and they would demand that
of Esperanto as well. Esperanto doesn't have that either. It does have
dictionaries, but its primary advantage is its larger community and
larger body of already written materials. It also requires a relatively
low skill level of beginners for them to think of themselves as
"Esperantists".
We're not there yet, but we aren't far (though not far is still "a lot
of work" for part time volunteers), and we aren't trying to emulate
Esperanto's path in any event.
>which is the reason why Esperanto is a finished language
Esperanto is not a finished language. Nor is English.
> while lojban is not. Even
> though lojban is better defined than Esperanto, it's not as fully
> defined as it should be in order for it to be finished, because the
> current level of its well-definiteness is not good enough for what
> lojban aims to be.
"should be" is arbitrary, as are the "aims". So your argument merely
creates a straw man that isn't really the target, so it cannot possibly
be "finished".
lojbab
Robert LeChevalier
> You have misquoted me. My words were, as you quoted them correctly the
> first time:
>
> "[Esperanto is] a fully defined, complete and functioning language, which
> lojban is very far from being at the moment."
Esperanto is NOT fully defined. It's completeness is arbitrary. It is
a functioning language.
Lojban is NOT fully defined, but is more so than Esperanto. It's
completeness is arbitrary, but its vocabulary is less completely defined
than Esperanto. It is a functioning language.
> So what I *have* said about lojban, is that it's "far from being a fully
> defined, complete and functioning language", not just that it's "far
> from functioning".
But it is functioning.
The problem is there is a smaller cadre of people sufficiently skilled
in Lojban who can use it fully functioning, and most of the people who
post on the lists are NOT those. By contrast, Esperanto discussions are
dominated by those who ARE highly skilled.
That is a teaching problem that takes time, and a limitation of the
community until then. It is not a language problem. And moreover it is
a language problem that is self-remedying.
lojbab
Robert LeChevalier
> I'm not talking about how lojban native speakers will come up with new
> words for things they don't have a word for. What I'm talking about is
> invention/concept/thing which will have been named by people who don't
> speak lojban (but, for example, English) and lojban speakers will like
> the name those people have given it and will thus simply incorporate
> French "écran" and English "software" got incorporated into
> Serbo-Croatian as "ekran" and "softver".
They would not be incorporated into lojban that way. And the forms in
which they were incorporated would be valid Lojban words, and would
remain valid Lojban words even if better words were made to replace them.
> Sure, purists didn't like that
> and invented replacement words, namely "zaslon" and "omekšje",
> respectively, but those words are simply not used and have failed to
> replace "ekran" and "softver" and these two have become a part of
> Serbo-Croatian vocabulary.
If they are valid S-C words, that is no problem. There wold not be
valid Lojban words, and probably would therefore NOT become part of the
vocabulary.
>Same thing would happen with lojban
There is no evidence of this.
- purists
> would invent lujvo (or brivla) to replace the direct loanwords in order
> to leave the language's unambiguity intact,
Loanwords, in proper morphological form, leave the morphological
unambiguity intact, and it is not difficult to create such words.
>but people who don't care
> about whether the language is completely unambiguous or not
Probably won't learn the language in the first place.
>(who would,
> mind you, make up a great majority of lojban speakers if it did become
> world's official common language)
No language will "become world's official common language". That is
political impossibility. Any international language will only be used
by people who are motivated to use the language, and generally those who
are so motivated, want to use the language correctly.
> would not cease to use the loanwords
> in place of the new "proper" words and lojban would get screwed up
> pretty quick.
Lojban is not screwed up by well-formed loan words. It may be
aesthetically less pleasing, but it loses no functionality.
> Of course, you could say that lojban is what a special committee of
> purists says it is and that people who don't use only the words which
> have been approved by the committee don't speak lojban, but no one would
> agree to make such a fascistic language the world's official common
> language
NO one will agree to make any language "the world's official common
language", but if they did, there undoubtedly would indeed be some sort
of standards committee that would "fascistically" define the
international language, and denigrate anything else. That is the way
international standardization works.
> and even if they did no one would give a crap what the
> committee says and lojban would still be what is spoken and not what is
> approved.
In which case, the question of "completeness" is irrelevant. If people
are happy speaking the language or some approximation thereof, then the
"fascistic" standards of completeness and unambiguity will be
unimportant to them, just as the rules of every other language are.
But in fact, the evidence of the existing speaker base is that people DO
"give a crap" what the committee says in the case of Lojban, enough so,
that "xorlo" was "forceably" approved by the committee well ahead of the
standard procedure because the community wanted it to be officially
endorsed.
lojbab
Robert LeChevalier
> There's no "lojbanic" thinking about logic - there's logic and not
> logic.
False. There are multiple schools of logic, which conflict at times on
very fundamental issues. Though that probably isn't relevant for this
example.
> If the two sentences parse differently you've either
> mistranslated them or lojban is not logical. Yeah, I'll presume the former.
The better conclusion is that the original sentence, being written in an
ambiguous language, wasn't logically unambiguous in the first place.
> Wow, you guys need to learn your logic. Let's do it properly:
>
> A = "lojban is fully defined."
> B = "lojban is complete."
> C = "lojban is a functioning language."
>
> "lojban is not a fully defined, complete and functioning language" can be written as:
Not it can't. Your summary sentence is NOT
¬(A ∧ B ∧ C)
The closest English can come to that is
"It is not the case that lojban is fully defined, and that lojban is
complete, and that lojban is a functioning language." and even that is
potentially ambiguous in several ways, because the words themselves are
ambiguous given differing contexts. (for example, "Lojban is complete"
and "Lojban is a complete language" are not necessarily identical in
meaning.)
Your summary sentence uses "not" as a contrary rather than contradictory
negation, and combines the three independent logical terms into a single
complex modifier of the word "language". It thus is NOT the same as the
three separate sentences, logically ANDes and the whole negated.
Lojban makes the differences extremely clear. English obviously does not.
lojbab
Robert LeChevalier
> If A = "lojban is fully defined." then "lojban is not A" means "lojban
> is not "lojban is fully defined."." Decide what A, B and C mean.
>
> In any case, this is what I meant:
> "It is not true that lojban is fully defined and that lojban is complete
> and that lojban is a functioning language."
>
> That *must* be equivalent to:
> "lojban is not fully defined, or lojban is not complete, or lojban is
> not a functioning language."
That still ignores the difference between contradictory negation and
contrary negation, a difference that English pretty much cannot
distinguish except with a lot of context and convention, while the
distinction is fundamental in Lojban.
lojbab
MorphemeAddict
And Rosta
> Lojban is *FAR* more fully defined than Esperanto.
Natural languages are defined by what their speakers know (or do). An invented language may be defined either (A) explicitly, by means of formal grammars and suchlike, or (B), like a natural language, by what their speakers know (or do). Esperanto is defined only (B)-wise. There are some Lojbanists, such as Lojbab, who would prefer a (B)-wise definition for Lojban too, but I guess most folk attracted to the idea of a logical language would want an (A)-wise definition for it.
> No, really: it is. Esperanto doesn't have a formal grammar of any
> kind, for starters.
But nor does Lojban. Lojban's so-called formal grammar does nothing but define a set of structures of phonological strings. What a real grammar would do is define a set of correspondences between sentence forms and sentence meanings.
However, even though Lojban has no true formal grammar, I think it would be easier to write one for Lojban than for almost any other language that has a speech community, though one expects it would be hard for the community to accept it as definitional.
> We know far more about how Lojban grammatical structures work than
> *any other actually spoken language on the planet*.
This is one of the attractions of explicit,(A)-wise definitions. But of all actually spoken languages on the planet, Lojban is the only one that has an explicit, (A)-wise definition, so Lojban wins this competition by having no competitors.
> We have already won that prize: Lojban is the most precisely,
> formally specified language that there is, for any language with its
> number of speakers or higher. Period. I challenge anyone to find
> anything even *remotely close* to the CLL in terms of covering every
> *possible* grammatical combination. Even if you can find such a
> thing, the formal grammar takes it so far ahead of everything else
> they can't possibly hope to catch up.
The virtues of Lojban are indeed as you say they are, for any language with its number of speakers or higher. But this is pf course far more of a tribute to Lojban's success in acquiring a user community than to its formal specification.
> The truth of the matter is that you really
> *can* say anything you want in Lojban; LNC and alis prove that
> pretty conclusively, I think.
This is debatable in a number of ways. First, the formal specification doesn't explicitly cover everything ordinary language might require (cf. problems with "if", with alternatival questions, etc.). Second, the claim could be true in only the trivial sense that the basics of predicate structure are sufficient to express all needed meanings; i.e. you can ignore everything but predicate structure and define new predicates to express whatever meaning you need. Third, some of the conventions that have arisen in usage to express needed meanings are not compositional, so their status as licit Lojban is questionable.
> 4. Nobody shouts "Wow this is well specified!!!" at the top of
> their lungs, but they certainly shout their complaints. Geeks have
> a shared culture that compliments are private and insults are
> public; it's deeply fucked up. See
> http://lesswrong.com/lw/3h/why_our_kind_cant_cooperate/
The impressive thing is the vitality of the user community, and the amount of labour folk have invested in it, not the specification. It would be easy -- with the benefit of experience -- to improve on the specification enormously, i.e. easy to design a language better in every conceivable way. But it would be nigh-on impossible to achieve a lojban-scale user-community for it.
> I can say anything I need to say in Lojban, modulo my own vocabulary
> knowledge.
It may well be that for any meaning you want to express, you have a way of expressing it and find that others will understand you. This is not the same thing, though, as it being possible to take your sentences apart and show *how* they mean what you think they do. If you have cooperative interlocutors, you can speak a very broken mangled version of language X and still be understood. Indeed, when all interlocutors know the language only very imperfectly, they may simply be oblivious to all the mistakes. And it can happen that some mistakes are so frequent that in actual usage they override the formal specification (e.g. prexorlo gadri).
> This puts it ahead of 99.999% of conlangs.
But maybe not ahead of 99.999% of conlangs that somebody is at all likely to claim are adequate to all ordinary communicative requirements.
>Saying that
> it is very far from being complete and functioning is ridiculous,
> and pretty insulting to a lot of people's hard work.
Whoever is insulted by that misunderstands what people's hard work has achieved. The design of the language itself has little intrinsic excellence (when viewed ahistorically), and it is naive to deny that it is massively incomplete. The achievement has been in building and sustaining the user-community, so that of all languages with a user-community, Lojban is the one that comes closest to being an explicitly specified logical language. The language itself could not have been substantially improved without great detriment to the user-community.
--And.
Robin Lee Powell
> Robin Lee Powell, On 05/01/2011 16:52:
> >Lojban is *FAR* more fully defined than Esperanto.
>
> Natural languages are defined by what their speakers know (or do).
> An invented language may be defined either (A) explicitly, by
> means of formal grammars and suchlike, or (B), like a natural
> language, by what their speakers know (or do). Esperanto is
> defined only (B)-wise. There are some Lojbanists, such as Lojbab,
> who would prefer a (B)-wise definition for Lojban too, but I guess
> most folk attracted to the idea of a logical language would want
> an (A)-wise definition for it.
*nod*
> >No, really: it is. Esperanto doesn't have a formal grammar of
> >any kind, for starters.
>
> But nor does Lojban. Lojban's so-called formal grammar does
> nothing but define a set of structures of phonological strings.
That's what "formal grammar" *means*;
http://en.wikipedia.org/wiki/Formal_grammar
> What a real grammar would do is define a set of correspondences
> between sentence forms and sentence meanings.
I don't know what that is, but it's not a formal grammar. Ask
google if you don't believe me. :)
I have no idea how you could formalize such a thing (and I'm not
terribly sure I care, to be honest).
> >We know far more about how Lojban grammatical structures work
> >than *any other actually spoken language on the planet*.
>
> This is one of the attractions of explicit,(A)-wise definitions.
> But of all actually spoken languages on the planet, Lojban is the
> only one that has an explicit, (A)-wise definition, so Lojban wins
> this competition by having no competitors.
I *know*. :D Isn't it awesome!?!?
> >4. Nobody shouts "Wow this is well specified!!!" at the top of
> >their lungs, but they certainly shout their complaints. Geeks
> >have a shared culture that compliments are private and insults
> >are public; it's deeply fucked up. See
> >http://lesswrong.com/lw/3h/why_our_kind_cant_cooperate/
>
> The impressive thing is the vitality of the user community, and
> the amount of labour folk have invested in it, not the
> specification. It would be easy -- with the benefit of experience
> -- to improve on the specification enormously, i.e. easy to design
> a language better in every conceivable way. But it would be
> nigh-on impossible to achieve a lojban-scale user-community for
> it.
*nod*
> >I can say anything I need to say in Lojban, modulo my own
> >vocabulary knowledge.
>
> It may well be that for any meaning you want to express, you have
> a way of expressing it and find that others will understand you.
> This is not the same thing, though, as it being possible to take
> your sentences apart and show *how* they mean what you think they
> do.
Fair enough; we leave a lot to context in practice. I'm OK with
that.
> >This puts it ahead of 99.999% of conlangs.
>
> But maybe not ahead of 99.999% of conlangs that somebody is at all
> likely to claim are adequate to all ordinary communicative
> requirements.
Sssshhhh! Stop spoiling my rants with evidence! :P :)
> >Saying that it is very far from being complete and functioning is
> >ridiculous, and pretty insulting to a lot of people's hard work.
>
> Whoever is insulted by that misunderstands what people's hard work
> has achieved. The design of the language itself has little
> intrinsic excellence (when viewed ahistorically), and it is naive
> to deny that it is massively incomplete.
I completely disagree. I don't see anything even vaguely
approaching "massively incomplete" in any part of Lojban, except
maybe vocabulary. I'd ask you to point to specific examples, but
I'm honstly not sure that I'm terribly interested in debating the
issue.
And Rosta
> Ivo Doko wrote:
>> Wow, you guys need to learn your logic. Let's do it properly:
>>
>> A = "lojban is fully defined."
>> B = "lojban is complete."
>> C = "lojban is a functioning language."
>>
>> "lojban is not a fully defined, complete and functioning language" can be written as:
>
> Not it can't. Your summary sentence is NOT
> ¬(A ∧ B ∧ C)
>
> The closest English can come to that is
> "It is not the case that lojban is fully defined, and that lojban is complete, and that lojban is a functioning language." and even that is potentially ambiguous in several ways, because the words themselves are ambiguous given differing contexts. (for example, "Lojban is complete" and "Lojban is a complete language" are not necessarily identical in meaning.)
>
> Your summary sentence uses "not" as a contrary rather than contradictory negation, and combines the three independent logical terms into a single complex modifier of the word "language". It thus is NOT the same as the three separate sentences, logically ANDes and the whole negated.
>
> Lojban makes the differences extremely clear. English obviously does not.
In English, logical scope tends to be ambiguous, at least within the same clause. So English "not A, B and C" can mean "It is not the case that each of A,B,C is the case" or "For each x, where x is one of A,B,C, it is not the case that x is the case".
Unless it has been fixed by recent BPFK action, Lojban has *exactly the same ambiguity* with regard to logical scope between elements that are not explicitly prenexed. (At least Lojban has the option of prenexing to eradicate ambiguity, but it is an option almost never used and that if often used would be received with opprobrium as stylistically objectionable.)
--And.
Robin Lee Powell
>
> In English, logical scope tends to be ambiguous, at least within
> the same clause. So English "not A, B and C" can mean "It is not
> the case that each of A,B,C is the case" or "For each x, where x
> is one of A,B,C, it is not the case that x is the case".
>
> Unless it has been fixed by recent BPFK action, Lojban has
> *exactly the same ambiguity* with regard to logical scope between
> elements that are not explicitly prenexed.
Show me an example please.
> (At least Lojban has the option of prenexing to eradicate
> ambiguity, but it is an option almost never used and that if often
> used would be received with opprobrium as stylistically
> objectionable.)
That last bit isn't true for me, fwiw.
Krzysztof Sobolewski
> > What a real grammar would do is define a set of correspondences
> > between sentence forms and sentence meanings.
>
> I don't know what that is, but it's not a formal grammar. Ask
> google if you don't believe me. :)
>
> I have no idea how you could formalize such a thing (and I'm not
> terribly sure I care, to be honest).
Formal semantics? I'm pretty much sure such a thing exists, but I have no idea how would one define one...
--
Ecce Jezuch
"But I want to, I want to, oh I want to but my hands were always tied up
And I wish I had myself a dime for every time I cursed your Goddamn name"
- P. Keenan
And Rosta
> On Thu, Jan 06, 2011 at 04:51:59PM +0000, And Rosta wrote:
>> Lojban's so-called formal grammar does
>> nothing but define a set of structures of phonological strings.
>
> That's what "formal grammar" *means*;
> http://en.wikipedia.org/wiki/Formal_grammar
You're quite right, but you'll see that the article says that formal grammars are for formal languages, and (in the description of formal languages in the article on formal languages) that formal languages aren't human languages. A "formal grammar" in this sense is irrelevant to the specification of a human language.
"Formal grammar" has a further meaning in linguistics, which is "grammar formulated in an explicit way", and it's this meaning that is relevant to the specification of a human language.
>> What a real grammar would do is define a set of correspondences
>> between sentence forms and sentence meanings.
>
> I don't know what that is, but it's not a formal grammar. Ask
> google if you don't believe me. :)
>I have no idea how you could formalize such a thing (and I'm not
> terribly sure I care, to be honest).
If you think about it, I think you will find you do care. Obviously the essential function of a language is to define correspondences between forms and meanings. If your putative specification of a language describes only possible forms and says nothing of meanings, then it is simply not a specification of a language. (Rather, it would be a specification of a "formal language" in the sense referred to above.)
As for you having no idea how to formalize such a thing, surely you can imagine having and implementing the design goal of a speakable predicate logic (which was one of Loglan's original goals). Retrofitting such a thing onto existing Lojban would be difficult, but surely the principle of it is easy to grasp: rules that take the phonological forms of Lojban sentences and translate them into predicate logic.
>>> We know far more about how Lojban grammatical structures work
>>> than *any other actually spoken language on the planet*.
>>
>> This is one of the attractions of explicit,(A)-wise definitions.
>> But of all actually spoken languages on the planet, Lojban is the
>> only one that has an explicit, (A)-wise definition, so Lojban wins
>> this competition by having no competitors.
>
> I *know*. :D Isn't it awesome!?!?
Certainly awesome enough for it to have kept me interested in it for the last 20 years.
>> The design of the language itself has little
>> intrinsic excellence (when viewed ahistorically), and it is naive
>> to deny that it is massively incomplete.
>
> I completely disagree. I don't see anything even vaguely
> approaching "massively incomplete" in any part of Lojban, except
> maybe vocabulary. I'd ask you to point to specific examples, but
> I'm honstly not sure that I'm terribly interested in debating the
> issue.
The major incompleteness is in the specification of correspondences between forms and meanings (i.e. predicate logic). I don't mean the definitions of individual brivla, but rather the meanings of sentences containing nonbrivla stuff.
--And.
John E Clifford
----- Original Message ----
From: And Rosta <and....@gmail.com>
To: loj...@googlegroups.com
Sent: Thu, January 6, 2011 10:51:59 AM
Subject: Re: Lojban is *NOT* broken! Stop saying that! (was Re: [lojban] Re:
Vote for the Future Global Language)
Robin Lee Powell, On 05/01/2011 16:52:
> Lojban is *FAR* more fully defined than Esperanto.
Natural languages are defined by what their speakers know (or do). An invented
language may be defined either (A) explicitly, by means of formal grammars and
suchlike, or (B), like a natural language, by what their speakers know (or do).
Esperanto is defined only (B)-wise. There are some Lojbanists, such as Lojbab,
who would prefer a (B)-wise definition for Lojban too, but I guess most folk
attracted to the idea of a logical language would want an (A)-wise definition
for it.
**but, so far as we can tell and act upon, A and B are the same here, that is,
people do conform to the formal grammar. And will continue to do so for some
time.
> No, really: it is. Esperanto doesn't have a formal grammar of any
> kind, for starters.
But nor does Lojban. Lojban's so-called formal grammar does nothing but define a
set of structures of phonological strings. What a real grammar would do is
define a set of correspondences between sentence forms and sentence meanings.
However, even though Lojban has no true formal grammar, I think it would be
easier to write one for Lojban than for almost any other language that has a
speech community, though one expects it would be hard for the community to
accept it as definitional.
**Are we quibbling here about the difference between syntax and grammar? There
is a use of "grammar" that is all-encompassing, from phonology through
pragmatics, and Lojban certainly doesn't have that, though it lacks only the
last two chunks, But, since these have resisted formulation in Linguistics so
far, even at the theoretical level, it seems unfair to criticize Lojaban for
lacking what Logic and Linguistics have yet to provide good models -- or even
criteria -- for. Efforts along this line tend to involve and idealized
representational language, almost all of which end up looking a lot like first
order predicate logic, meaning that the crucial step in the process from Lojban
form to meaning would be -- with a few caveats -- a snap. The only interesting
question about Lojban's A syntax is whether all and only semantically
significant substructures are also syntactic substructures. This was certainly
not true in earlier versions, but I can't read modern syntax well enough to know
whether it is now or not (I seem to recall that bridi tail was a particular
problem in this respect).
> We know far more about how Lojban grammatical structures work than
> *any other actually spoken language on the planet*.
This is one of the attractions of explicit,(A)-wise definitions. But of all
actually spoken languages on the planet, Lojban is the only one that has an
explicit, (A)-wise definition, so Lojban wins this competition by having no
competitors.
** Being unique in this way can hardly be a flaw in the language, especially if
your aim (or your ultimate criterion) is the cionstruction of a complete
pragmatics).
> We have already won that prize: Lojban is the most precisely,
> formally specified language that there is, for any language with its
> number of speakers or higher. Period. I challenge anyone to find
> anything even *remotely close* to the CLL in terms of covering every
> *possible* grammatical combination. Even if you can find such a
> thing, the formal grammar takes it so far ahead of everything else
> they can't possibly hope to catch up.
The virtues of Lojban are indeed as you say they are, for any language with its
number of speakers or higher. But this is pf course far more of a tribute to
Lojban's success in acquiring a user community than to its formal specification.
> The truth of the matter is that you really
> *can* say anything you want in Lojban; LNC and alis prove that
> pretty conclusively, I think.
This is debatable in a number of ways. First, the formal specification doesn't
explicitly cover everything ordinary language might require (cf. problems with
"if", with alternatival questions, etc.). Second, the claim could be true in
only the trivial sense that the basics of predicate structure are sufficient to
express all needed meanings; i.e. you can ignore everything but predicate
structure and define new predicates to express whatever meaning you need. Third,
some of the conventions that have arisen in usage to express needed meanings are
not compositional, so their status as licit Lojban is questionable.
** I need to be reminded of what "compositional" means here and see some
examples of problem cases. The problems with "if" and the milk-or-cream joke
are real enough but clearly don't need solutions outside the existing syntax,
only a better use of what is already there (stiop thinking of them as
connectives being one useful approach).
> 4. Nobody shouts "Wow this is well specified!!!" at the top of
> their lungs, but they certainly shout their complaints. Geeks have
> a shared culture that compliments are private and insults are
> public; it's deeply fucked up. See
> http://lesswrong.com/lw/3h/why_our_kind_cant_cooperate/
The impressive thing is the vitality of the user community, and the amount of
labour folk have invested in it, not the specification. It would be easy -- with
the benefit of experience -- to improve on the specification enormously, i.e.
easy to design a language better in every conceivable way. But it would be
nigh-on impossible to achieve a lojban-scale user-community for it.
> I can say anything I need to say in Lojban, modulo my own vocabulary
> knowledge.
It may well be that for any meaning you want to express, you have a way of
expressing it and find that others will understand you. This is not the same
thing, though, as it being possible to take your sentences apart and show *how*
they mean what you think they do. If you have cooperative interlocutors, you can
speak a very broken mangled version of language X and still be understood.
Indeed, when all interlocutors know the language only very imperfectly, they may
simply be oblivious to all the mistakes. And it can happen that some mistakes
are so frequent that in actual usage they override the formal specification
(e.g. prexorlo gadri).
** Check. Do the semantic and syntactic substructures congrue?
> This puts it ahead of 99.999% of conlangs.
But maybe not ahead of 99.999% of conlangs that somebody is at all likely to
claim are adequate to all ordinary communicative requirements.
> Saying that
> it is very far from being complete and functioning is ridiculous,
> and pretty insulting to a lot of people's hard work.
Whoever is insulted by that misunderstands what people's hard work has achieved.
The design of the language itself has little intrinsic excellence (when viewed
ahistorically), and it is naive to deny that it is massively incomplete. The
achievement has been in building and sustaining the user-community, so that of
all languages with a user-community, Lojban is the one that comes closest to
being an explicitly specified logical language. The language itself could not
have been substantially improved without great detriment to the user-community.
**But, of course, a large portion of that community came to Lojban precisely
because of the claim to be unambiguous in one fairly major way. Without that
claim, the group would be significantly smaller, nearer, say, toki pona (maybe
50 with a little fudging and an awareness base pf a few hundred).
--And.
Robin Lee Powell
> "Formal grammar" has a further meaning in linguistics, which is
> "grammar formulated in an explicit way", and it's this meaning
> that is relevant to the specification of a human language.
Ah. I don't know that use. Can you point me to an example of such
a thing?
> >>What a real grammar would do is define a set of correspondences
> >>between sentence forms and sentence meanings.
> >
> >I don't know what that is, but it's not a formal grammar. Ask
> >google if you don't believe me. :) I have no idea how you could
> >formalize such a thing (and I'm not terribly sure I care, to be
> >honest).
>
> If you think about it, I think you will find you do care.
> Obviously the essential function of a language is to define
> correspondences between forms and meanings. If your putative
> specification of a language describes only possible forms and says
> nothing of meanings, then it is simply not a specification of a
> language. (Rather, it would be a specification of a "formal
> language" in the sense referred to above.)
Of course; the CLL does, in fact, cover semantics in quite a lot of
detail (and, I assert, more thoroughly than any such document
natural language; I have no way to measure this though).
> As for you having no idea how to formalize such a thing, surely
> you can imagine having and implementing the design goal of a
> speakable predicate logic (which was one of Loglan's original
> goals). Retrofitting such a thing onto existing Lojban would be
> difficult,
Wait what? How do we not have that?
> but surely the principle of it is easy to grasp: rules that take
> the phonological forms of Lojban sentences and translate them into
> predicate logic.
That doesn't do anything for general semantics, though. IsRed(x) as
a predicate is just a suggestively named lisp token ( see
http://singinst.org/ourresearch/publications/GISAI/meta/glossary.html#gloss_lisp_tokens
and http://lesswrong.com/lw/la/truly_part_of_you/ ); to formalize
actual semantics in the way I think you're talking about, you need
to formalize what it means for something to be Red. You can't do
that in bare predicate logic; you'd do samething like
HasWavelengthBetween(x,630nm,700nm), but that doesn't help, because
now you have to have predicates for nanometers, and what a
wavelength is, and on and on and on. Having a complete semantic
mapping of *anything* is a fool's errand, which is why the semantic
web is dead (and was dead before it started).
As far as I can tell, the semantic descriptions of Lojban in the CLL
are about as good as can reasonably be achieved without falling down
the rabbit hole of perfect semantic description, I don't see how it
differs from "spoken predicate logic" in that respect, and I'm very
curious as to whether you have evidence to the contrary.
> >>The design of the language itself has little intrinsic
> >>excellence (when viewed ahistorically), and it is naive to deny
> >>that it is massively incomplete.
> >
> >I completely disagree. I don't see anything even vaguely
> >approaching "massively incomplete" in any part of Lojban, except
> >maybe vocabulary. I'd ask you to point to specific examples, but
> >I'm honstly not sure that I'm terribly interested in debating the
> >issue.
>
> The major incompleteness is in the specification of
> correspondences between forms and meanings (i.e. predicate logic).
> I don't mean the definitions of individual brivla, but rather the
> meanings of sentences containing nonbrivla stuff.
I don't feel a significant lack there. If you do, please make
updates to the Notes sections of the various BPFK pages so I can try
to fix it.
And Rosta
> On Thu, Jan 06, 2011 at 05:08:02PM +0000, And Rosta wrote:
>>
>> In English, logical scope tends to be ambiguous, at least within
>> the same clause. So English "not A, B and C" can mean "It is not
>> the case that each of A,B,C is the case" or "For each x, where x
>> is one of A,B,C, it is not the case that x is the case".
>>
>> Unless it has been fixed by recent BPFK action, Lojban has
>> *exactly the same ambiguity* with regard to logical scope between
>> elements that are not explicitly prenexed.
>
> Show me an example please.
"su'o broda ro brode cu brodi"
"na ku a bu e by e cy cu broda"
Ten years ago these were ambiguous. Xorxes proposed a rule that items in higher clauses have scope over items in lower clauses (i.e. that items export to the prenex of the localmost clause) and that when two items are in the same clause, the leftward element has scope over the rightward. (It's a shame to have to 'pollute' the purely hierarchical structure of logical form with left-to-right order of forms, but it's by far the simplest way to rescue Lojban in its (then) current state. Perhaps the BPFK has made xorxes's rule official, in which case I wonder what happened to the rule about the scope of selbri tcita "na", and to the scope of selbri tcita in general.)
--And.
John E Clifford
we are beyond semantics into pragmatics -- about which we have even less of an
idea than semantics. As for return to formal logic, the untangling of Logjam's
involuting is fairly easy in theory but does occasionally break down: scopes, as
you note, interdigitating argument streams, and a few others. The residuum is
probably going to require at least intensional second order theories (whatever
the Hell they are) (feelings of deja vu -- did we have this discussion -- or
something like it -- over the last twenty years?)
----- Original Message ----
From: And Rosta <and....@gmail.com>
To: loj...@googlegroups.com
Sent: Thu, January 6, 2011 12:37:20 PM
Subject: Re: Lojban is *NOT* broken! Stop saying that! (was Re: [lojban] Re:
Vote for the Future Global Language)
--And.
--
Robin Lee Powell
> Well, the non-brivla stuff is, of course, also the non-predicate
> logic stuff, so we are beyond semantics into pragmatics
xorxes and I have done *some* work on "formal" definitions of cmavo;
see
http://www.lojban.org/tiki/tiki-index.php?page=BPFK+Section%3A+gadri
, and I have occassionally fantasized about picking a
few predicate-logic-related words as axioms and defining the entire
rest of the language in terms of them.
Perhaps that's what Andrew was aiming for.
Robin Lee Powell
> Robin Lee Powell, On 06/01/2011 17:19:
> >On Thu, Jan 06, 2011 at 05:08:02PM +0000, And Rosta wrote:
> >>
> >>In English, logical scope tends to be ambiguous, at least within
> >>the same clause. So English "not A, B and C" can mean "It is not
> >>the case that each of A,B,C is the case" or "For each x, where x
> >>is one of A,B,C, it is not the case that x is the case".
> >>
> >>Unless it has been fixed by recent BPFK action, Lojban has
> >>*exactly the same ambiguity* with regard to logical scope
> >>between elements that are not explicitly prenexed.
> >
> >Show me an example please.
>
> "su'o broda ro brode cu brodi"
>
> "na ku a bu e by e cy cu broda"
>
> Ten years ago these were ambiguous.
You're going to have to hold my hand more than that, I'm afraid.
Ambiguous how?
> Xorxes proposed a rule that items in higher clauses have scope
> over items in lower clauses (i.e. that items export to the prenex
> of the localmost clause) and that when two items are in the same
> clause, the leftward element has scope over the rightward. (It's a
> shame to have to 'pollute' the purely hierarchical structure of
> logical form with left-to-right order of forms, but it's by far
> the simplest way to rescue Lojban in its (then) current state.
> Perhaps the BPFK has made xorxes's rule official, in which case I
> wonder what happened to the rule about the scope of selbri tcita
> "na", and to the scope of selbri tcita in general.)
I didn't know that was xorxes' rule; I thought left-to-right
quantifier scope was in the CLL. Yes, indeed:
http://dag.github.com/cll/16/5/ "The rule for dropping the prenex is
simple: if the variables appear in the same order within the bridi
as they did in the prenex, then the prenex is superfluous.".
So, I'm probably failing to understand. Can you please explain it
like I'm very very stupid?
John E Clifford
that whole cult is still too much with us.
----- Original Message ----
From: Robin Lee Powell <rlpo...@digitalkingdom.org>
To: loj...@googlegroups.com
Sent: Thu, January 6, 2011 12:51:51 PM
Subject: Re: Lojban is *NOT* broken! Stop saying that! (was Re: [lojban] Re:
Vote for the Future Global Language)
On Thu, Jan 06, 2011 at 06:37:20PM +0000, And Rosta wrote:
> "Formal grammar" has a further meaning in linguistics, which is
> "grammar formulated in an explicit way", and it's this meaning
> that is relevant to the specification of a human language.
Ah. I don't know that use. Can you point me to an example of such
a thing?
**Complete for a language, no. Models for (small) parts of languages, look in
any old anthropological linguistics journal.
> >>What a real grammar would do is define a set of correspondences
> >>between sentence forms and sentence meanings.
> >
> >I don't know what that is, but it's not a formal grammar. Ask
> >google if you don't believe me. :) I have no idea how you could
> >formalize such a thing (and I'm not terribly sure I care, to be
> >honest).
>
> If you think about it, I think you will find you do care.
> Obviously the essential function of a language is to define
> correspondences between forms and meanings. If your putative
> specification of a language describes only possible forms and says
> nothing of meanings, then it is simply not a specification of a
> language. (Rather, it would be a specification of a "formal
> language" in the sense referred to above.)
Of course; the CLL does, in fact, cover semantics in quite a lot of
detail (and, I assert, more thoroughly than any such document
natural language; I have no way to measure this though).
> As for you having no idea how to formalize such a thing, surely
> you can imagine having and implementing the design goal of a
> speakable predicate logic (which was one of Loglan's original
> goals). Retrofitting such a thing onto existing Lojban would be
> difficult,
Wait what? How do we not have that?
**Well, scope limits are undefined for most quantifiers (which are
indeterminately defined themselves) and for negations and alternate world
functions (real modals), for starters. This just needs decisions by someone,
writing them up, and getting people to actually use them. It is not a flaw in
the syntax, except that it means some semantically distinctive substructures are
not (reliably, at least) syntactically distinct.
> but surely the principle of it is easy to grasp: rules that take
> the phonological forms of Lojban sentences and translate them into
> predicate logic.
That doesn't do anything for general semantics, though. IsRed(x) as
a predicate is just a suggestively named lisp token ( see
http://singinst.org/ourresearch/publications/GISAI/meta/glossary.html#gloss_lisp_tokens
and http://lesswrong.com/lw/la/truly_part_of_you/ ); to formalize
actual semantics in the way I think you're talking about, you need
to formalize what it means for something to be Red. You can't do
that in bare predicate logic; you'd do samething like
HasWavelengthBetween(x,630nm,700nm), but that doesn't help, because
now you have to have predicates for nanometers, and what a
wavelength is, and on and on and on. Having a complete semantic
mapping of *anything* is a fool's errand, which is why the semantic
web is dead (and was dead before it started).
**It is, of course, logically impossible to define all of the terms of a
language in that language without either circularity or contradiction. It is
possible, however, to define all the terms of a language in another language,
here the semantic metalanguage. It is incompletely specified, but can
completely specify Ebglish, say, or Lojban (of course, it doesn't exist and the
models for it have all so far been flawed).
-Robin
--
John E Clifford
----- Original Message ----
From: And Rosta <and....@gmail.com>
To: loj...@googlegroups.com
Sent: Thu, January 6, 2011 12:52:19 PM
Subject: Re: Lojban is *NOT* broken! Stop saying that! (was Re: [lojban] Re:
Vote for the Future Global Language)
** Were these really ambiguous as recently as 10 years ago? There are problems
about what the quantifiers mean exactly and there are always RHE problems, with
scope mainly. But seems to be pretty clear from the LHE (prenex) perspective
(issues about the second example were about the translation of an English case).
--And.
Ian Johnson
~(A(x) ^ B(x) ^ C(x))
and
~A(x) ^ ~B(x) ^ ~C(x)
Lojban has plenty of ways to do this. Here's one:
naku ko'a broda gi'e brode gi'e brodi
vs.
ko'a na broda gi'enai brode gi'enai brodi
These are completely different sentences. On the other hand, if you actually go through and try to say the same thing in English, you actually wind up with virtually identical sentences:
It is not the case that ko'a X and Y and Z
vs.
ko'a doesn't X and doesn't Y and doesn't Z.
"It is not the case that" is stylistically horrible, however, which is why classes on logic taught in English spend probably an entire week on the concept of a "useful negation" and thus introducing the De Morgan laws, what happens to quantifiers when a statement is negated, etc.
So the issue is really not with English, it's with idiomatic English. You can more or less remove ambiguity in "rigid" contexts like these, but when you try to also add in flavor and life to your English (such as in "X is far from being A, B, and C", which incidentally I read as ~(A(x) ^ B(x) ^ C(x)), as a counterexample to someone who was saying any native English speaker would read it the other way), it becomes extremely hard to preserve the lack of ambiguity. Lojban makes this a hell of a lot easier; UI alone is a tremendous help.
Also, for what it's worth, I've used prenexes quite a bit, and don't consider them especially hard to read. Hell, one of {me'ei}'s dominant uses is {ro me'ei bu'a zo'u}.
mu'o mi'e .latros.
And Rosta
> From: And Rosta<and....@gmail.com>
> Lojban's so-called formal grammar does nothing but define a
> set of structures of phonological strings. What a real grammar would do is
> define a set of correspondences between sentence forms and sentence meanings.
>
> However, even though Lojban has no true formal grammar, I think it would be
> easier to write one for Lojban than for almost any other language that has a
> speech community, though one expects it would be hard for the community to
> accept it as definitional.
>
> **Are we quibbling here about the difference between syntax and grammar? There
> is a use of "grammar" that is all-encompassing, from phonology through
> pragmatics, and Lojban certainly doesn't have that, though it lacks only the
> last two chunks,
I don't think there's quibbling going on here. For a language you need a level of form, the stuff that gets interpreted phonetically, and a level of meaning, the stuff that gets interpreted pragmatically, and correspondences between the two levels. What you call the levels and the correspondence rules is a separate matter.
> But, since these have resisted formulation in Linguistics so
> far, even at the theoretical level, it seems unfair to criticize Lojaban for
> lacking what Logic and Linguistics have yet to provide good models -- or even
> criteria -- for. Efforts along this line tend to involve and idealized
> representational language, almost all of which end up looking a lot like first
> order predicate logic, meaning that the crucial step in the process from Lojban
> form to meaning would be -- with a few caveats -- a snap.
Hopefully it would be a snap, but it's these rules that the formal definition/specification of the language requires, and not the formal grammar (save for whichever bits of the formal grammar are necessary for the form--meaning correspondence rules). Regarding the question of whether it would indeed be a snap, the requisite rules would in most cases need to be invented, so there'd be a political difficulty at least as much as a linguistic one.
>> The truth of the matter is that you really
>> *can* say anything you want in Lojban; LNC and alis prove that
>> pretty conclusively, I think.
>
> This is debatable in a number of ways. First, the formal specification doesn't
> explicitly cover everything ordinary language might require (cf. problems with
> "if", with alternatival questions, etc.). Second, the claim could be true in
> only the trivial sense that the basics of predicate structure are sufficient to
> express all needed meanings; i.e. you can ignore everything but predicate
> structure and define new predicates to express whatever meaning you need. Third,
> some of the conventions that have arisen in usage to express needed meanings are
> not compositional, so their status as licit Lojban is questionable.
>
> ** I need to be reminded of what "compositional" means here
The meaning of the whole is predictably composed from the meaning of the parts.
> and see some
> examples of problem cases. The problems with "if" and the milk-or-cream joke
> are real enough but clearly don't need solutions outside the existing syntax,
> only a better use of what is already there (stiop thinking of them as
> connectives being one useful approach).
I don't think anything needs solutions outside the existing syntax. But there is still stuff that needs solutions (within the existing syntax).
> The design of the language itself has little intrinsic excellence (when viewed
> ahistorically), and it is naive to deny that it is massively incomplete. The
> achievement has been in building and sustaining the user-community, so that of
> all languages with a user-community, Lojban is the one that comes closest to
> being an explicitly specified logical language. The language itself could not
> have been substantially improved without great detriment to the user-community.
>
> **But, of course, a large portion of that community came to Lojban precisely
> because of the claim to be unambiguous in one fairly major way. Without that
> claim, the group would be significantly smaller, nearer, say, toki pona (maybe
> 50 with a little fudging and an awareness base pf a few hundred).
It does seem that the great majority of Lojbanists are attracted by its aim or claim to be a logical language, but there are very few who are so dissatisfied with the design and/or specification that they would risk weakening the community by strengthening the language design. To put it another way, the great majority of Lojbanists are also attracted by its having a flourishing user-community, and rank the maintainance of the community higher than the quality of the language.
--And.
And Rosta
> On Thu, Jan 06, 2011 at 06:52:19PM +0000, And Rosta wrote:
>> Robin Lee Powell, On 06/01/2011 17:19:
>>> On Thu, Jan 06, 2011 at 05:08:02PM +0000, And Rosta wrote:
>>>>
>>>> In English, logical scope tends to be ambiguous, at least within
>>>> the same clause. So English "not A, B and C" can mean "It is not
>>>> the case that each of A,B,C is the case" or "For each x, where x
>>>> is one of A,B,C, it is not the case that x is the case".
>>>>
>>>> Unless it has been fixed by recent BPFK action, Lojban has
>>>> *exactly the same ambiguity* with regard to logical scope
>>>> between elements that are not explicitly prenexed.
>>>
>>> Show me an example please.
>>
>> "su'o broda ro brode cu brodi"
>>
>> "na ku a bu e by e cy cu broda"
>>
>> Ten years ago these were ambiguous.
>
> You're going to have to hold my hand more than that, I'm afraid.
> Ambiguous how?
In the case of the second one, ambiguous between the two meanings the English version has: "It is not the case that A, B and C broda" versus "For each of A,B,C it is not the case that it brodas".
>> Xorxes proposed a rule that items in higher clauses have scope
>> over items in lower clauses (i.e. that items export to the prenex
>> of the localmost clause) and that when two items are in the same
>> clause, the leftward element has scope over the rightward. (It's a
>> shame to have to 'pollute' the purely hierarchical structure of
>> logical form with left-to-right order of forms, but it's by far
>> the simplest way to rescue Lojban in its (then) current state.
>> Perhaps the BPFK has made xorxes's rule official, in which case I
>> wonder what happened to the rule about the scope of selbri tcita
>> "na", and to the scope of selbri tcita in general.)
>
> I didn't know that was xorxes' rule; I thought left-to-right
> quantifier scope was in the CLL. Yes, indeed:
> http://dag.github.com/cll/16/5/ "The rule for dropping the prenex is
> simple: if the variables appear in the same order within the bridi
> as they did in the prenex, then the prenex is superfluous.".
>
> So, I'm probably failing to understand. Can you please explain it
> like I'm very very stupid?
My mistake -- failure of memory. Presumably the then-unofficial rule was to generalize CLL's left-to-right rule for all elements in the bridi (with the possible exception of some or all selbri tcita) and to make explicit the rule that things export to the localmost rather than outermost prenex (when you have one bridi within another).
--And.
Arnt Richard Johansen
> On Thu, Jan 06, 2011 at 04:51:59PM +0000, And Rosta wrote:
>
> > What a real grammar would do is define a set of correspondences
> > between sentence forms and sentence meanings.
>
> terribly sure I care, to be honest).
For what it's worth, ERG is an example of such a thing:
http://www.delph-in.net/erg/
To summarize your objection from further downthread: “Formal semantics is just a system for transforming one string of meaningless symbols into another string of meaningless symbols, so what's the point?”
Well, to a certain extent you're right, but if you choose the right kind of semantic representation, you can do things like proving that two different strings of Lojban have the same meaning. Correct me if I'm wrong, but at the moment no machine grammar of Lojban represents the fact that “mi viska do” is equivalent to “do se viska mi”.
I don't think that we absolutely need to have such a thing, and I am certainly not volunteering to make it, but if we did have such a thing, I'm sure it would reveal one or two problems about Lojban grammar that no-one's thought about before.
--
Arnt Richard Johansen http://arj.nvg.org/
Information wants to be anthropomorphized!
Robin Lee Powell
wrote:
> On Thu, Jan 06, 2011 at 09:13:47AM -0800, Robin Lee Powell wrote:
> > On Thu, Jan 06, 2011 at 04:51:59PM +0000, And Rosta wrote:
> >
> > > What a real grammar would do is define a set of
> > > correspondences between sentence forms and sentence meanings.
> >
> > [...] I have no idea how you could formalize such a thing (and
> > I'm not terribly sure I care, to be honest).
>
> For what it's worth, ERG is an example of such a thing:
> http://www.delph-in.net/erg/
>
> To summarize your objection from further downthread: “Formal
> semantics is just a system for transforming one string of
> meaningless symbols into another string of meaningless symbols, so
> what's the point?”
>
> Well, to a certain extent you're right, but if you choose the
> right kind of semantic representation, you can do things like
> proving that two different strings of Lojban have the same
> meaning. Correct me if I'm wrong, but at the moment no machine
> grammar of Lojban represents the fact that “mi viska do” is
> equivalent to “do se viska mi”.
Right, very true. People have started playing with that. Also,
quoting from elsewhere in the thread:
xorxes and I have done *some* work on "formal" definitions of
cmavo; see
http://www.lojban.org/tiki/tiki-index.php?page=BPFK+Section%3A+gadri
, and I have occassionally fantasized about picking a few
predicate-logic-related words as axioms and defining the entire
rest of the language in terms of them.
> I don't think that we absolutely need to have such a thing, and I
> am certainly not volunteering to make it, but if we did have such
> a thing, I'm sure it would reveal one or two problems about Lojban
> grammar that no-one's thought about before.
*nodnod*
And Rosta
> ...What? You're trying to distinguish between:
> ~(A(x) ^ B(x) ^ C(x))
> and
> ~A(x) ^ ~B(x) ^ ~C(x)
>
> Lojban has plenty of ways to do this.
No, the point is firstly that the difference is merely one of scope, and secondly that Lojban too has (or perhaps merely had) the ambiguity that Lojbab was reproaching English for having. The point was not that Lojban (or English) has no way of disambiguating.
--And.
Here's one:
>
> naku ko'a broda gi'e brode gi'e brodi
> vs.
> ko'a na broda gi'enai brode gi'enai brodi
>
> These are completely different sentences. On the other hand, if you actually go through and try to say the same thing in English, you actually wind up with virtually identical sentences:
>
> It is not the case that ko'a X and Y and Z
> vs.
> ko'a doesn't X and doesn't Y and doesn't Z.
>
> "It is not the case that" is stylistically horrible, however, which is why classes on logic taught in English spend probably an entire week on the concept of a "useful negation" and thus introducing the De Morgan laws, what happens to quantifiers when a statement is negated, etc.
>
> So the issue is really not with English, it's with idiomatic English. You can more or less remove ambiguity in "rigid" contexts like these, but when you try to also add in flavor and life to your English (such as in "X is far from being A, B, and C", which incidentally I read as ~(A(x) ^ B(x) ^ C(x)), as a counterexample to someone who was saying any native English speaker would read it the other way), it becomes extremely hard to preserve the lack of ambiguity. Lojban makes this a hell of a lot easier; UI alone is a tremendous help.
>
> Also, for what it's worth, I've used prenexes quite a bit, and don't consider them especially hard to read. Hell, one of {me'ei}'s dominant uses is {ro me'ei bu'a zo'u}.
>
> mu'o mi'e .latros.
>
> On Thu, Jan 6, 2011 at 11:08 AM, And Rosta <and....@gmail.com <mailto:and....@gmail.com>> wrote:
>
> Robert LeChevalier, On 06/01/2011 15:00:
>
> Ivo Doko wrote:
>
> Wow, you guys need to learn your logic. Let's do it properly:
>
> A = "lojban is fully defined."
> B = "lojban is complete."
> C = "lojban is a functioning language."
>
> "lojban is not a fully defined, complete and functioning language" can be written as:
>
>
> Not it can't. Your summary sentence is NOT
> ¬(A ∧ B ∧ C)
>
> The closest English can come to that is
> "It is not the case that lojban is fully defined, and that lojban is complete, and that lojban is a functioning language." and even that is potentially ambiguous in several ways, because the words themselves are ambiguous given differing contexts. (for example, "Lojban is complete" and "Lojban is a complete language" are not necessarily identical in meaning.)
>
> Your summary sentence uses "not" as a contrary rather than contradictory negation, and combines the three independent logical terms into a single complex modifier of the word "language". It thus is NOT the same as the three separate sentences, logically ANDes and the whole negated.
>
> Lojban makes the differences extremely clear. English obviously does not.
>
>
> In English, logical scope tends to be ambiguous, at least within the same clause. So English "not A, B and C" can mean "It is not the case that each of A,B,C is the case" or "For each x, where x is one of A,B,C, it is not the case that x is the case".
>
> Unless it has been fixed by recent BPFK action, Lojban has *exactly the same ambiguity* with regard to logical scope between elements that are not explicitly prenexed. (At least Lojban has the option of prenexing to eradicate ambiguity, but it is an option almost never used and that if often used would be received with opprobrium as stylistically objectionable.)
>
> --And.
>
>
> --
> You received this message because you are subscribed to the Google Groups "lojban" group.
> To post to this group, send email to loj...@googlegroups.com <mailto:loj...@googlegroups.com>.
> To unsubscribe from this group, send email to lojban+un...@googlegroups.com <mailto:lojban%2Bunsu...@googlegroups.com>.
John E Clifford
----- Original Message ----
From: And Rosta <and....@gmail.com>
To: loj...@googlegroups.com
Sent: Thu, January 6, 2011 1:48:12 PM
Subject: Re: Lojban is *NOT* broken! Stop saying that! (was Re: [lojban] Re:
Vote for the Future Global Language)
**OK, but then don't complain if someone calls a syntax a grammar. When I say
only lacks two chunks, I should add that every other language also laks those
chunk and also the one before, which Lojban has (almost?)
> But, since these have resisted formulation in Linguistics so
> far, even at the theoretical level, it seems unfair to criticize Lojaban for
> lacking what Logic and Linguistics have yet to provide good models -- or even
> criteria -- for. Efforts along this line tend to involve and idealized
> representational language, almost all of which end up looking a lot like first
> order predicate logic, meaning that the crucial step in the process from
Lojban
> form to meaning would be -- with a few caveats -- a snap.
Hopefully it would be a snap, but it's these rules that the formal
definition/specification of the language requires, and not the formal grammar
(save for whichever bits of the formal grammar are necessary for the
form--meaning correspondence rules). Regarding the question of whether it would
indeed be a snap, the requisite rules would in most cases need to be invented,
so there'd be a political difficulty at least as much as a linguistic one.
**I'm not following here. What is the political difficulty in given obvious
rules for untangled conjoined terms or predicates or even blobs like briditail.
There are some less than obvious places, to be sure, but doing the first bit is
already more than we can do -- even by hand -- for any other language.
>> The truth of the matter is that you really
>> *can* say anything you want in Lojban; LNC and alis prove that
>> pretty conclusively, I think.
>
> This is debatable in a number of ways. First, the formal specification doesn't
> explicitly cover everything ordinary language might require (cf. problems with
> "if", with alternatival questions, etc.). Second, the claim could be true in
> only the trivial sense that the basics of predicate structure are sufficient
to
> express all needed meanings; i.e. you can ignore everything but predicate
> structure and define new predicates to express whatever meaning you need.
>Third,
> some of the conventions that have arisen in usage to express needed meanings
>are
> not compositional, so their status as licit Lojban is questionable.
>
> ** I need to be reminded of what "compositional" means here
The meaning of the whole is predictably composed from the meaning of the parts.
> and see some
> examples of problem cases. The problems with "if" and the milk-or-cream joke
> are real enough but clearly don't need solutions outside the existing syntax,
> only a better use of what is already there (stop thinking of them as
> connectives being one useful approach).
I don't think anything needs solutions outside the existing syntax. But there is
still stuff that needs solutions (within the existing syntax).
> The design of the language itself has little intrinsic excellence (when viewed
> ahistorically), and it is naive to deny that it is massively incomplete. The
> achievement has been in building and sustaining the user-community, so that of
> all languages with a user-community, Lojban is the one that comes closest to
> being an explicitly specified logical language. The language itself could not
> have been substantially improved without great detriment to the
user-community.
>
> **But, of course, a large portion of that community came to Lojban precisely
> because of the claim to be unambiguous in one fairly major way. Without that
> claim, the group would be significantly smaller, nearer, say, toki pona (maybe
> 50 with a little fudging and an awareness base pf a few hundred).
It does seem that the great majority of Lojbanists are attracted by its aim or
claim to be a logical language, but there are very few who are so dissatisfied
with the design and/or specification that they would risk weakening the
community by strengthening the language design. To put it another way, the great
majority of Lojbanists are also attracted by its having a flourishing
user-community, and rank the maintainance of the community higher than the
quality of the language.
Well, yes, success breeds success and all. I am inclined to think that at least
a significant portion of the present community would go over to an improved
language, if they were sure it was improved. But I fear that the kind of things
you suggest doing would not be obvious improvements (they may in fact be
improvements, but it would not be obvious that they are -- see the years of
xorlo discussions, for example).
--And.
--
Robin Lee Powell
> Robin Lee Powell, On 06/01/2011 19:01:
> >On Thu, Jan 06, 2011 at 06:52:19PM +0000, And Rosta wrote:
> >>Robin Lee Powell, On 06/01/2011 17:19:
> >>>On Thu, Jan 06, 2011 at 05:08:02PM +0000, And Rosta wrote:
> >>>>
> >>>>In English, logical scope tends to be ambiguous, at least
> >>>>within the same clause. So English "not A, B and C" can mean
> >>>>"It is not the case that each of A,B,C is the case" or "For
> >>>>each x, where x is one of A,B,C, it is not the case that x is
> >>>>the case".
> >>>>
> >>>>Unless it has been fixed by recent BPFK action, Lojban has
> >>>>*exactly the same ambiguity* with regard to logical scope
> >>>>between elements that are not explicitly prenexed.
> >>>
> >>>Show me an example please.
> >>
> >>"su'o broda ro brode cu brodi"
> >>
> >>"na ku a bu e by e cy cu broda"
> >>
> >>Ten years ago these were ambiguous.
> >
> >You're going to have to hold my hand more than that, I'm afraid.
> >Ambiguous how?
>
> In the case of the second one, ambiguous between the two meanings
> the English version has: "It is not the case that A, B and C
> broda" versus "For each of A,B,C it is not the case that it
> brodas".
I'm not going to comment on what I think it means because I have
problems with na ku and quantification :) and I'm pretty busy right
now, but if you could review the relevant CLL sections and
double-check whether there's remaining ambiguity, and if so show me
where, I'd *really* appreciate it.
John E Clifford
----- Original Message ----
From: And Rosta <and....@gmail.com>
To: loj...@googlegroups.com
Sent: Thu, January 6, 2011 2:13:49 PM
Subject: Re: Lojban is *NOT* broken! Stop saying that! (was Re: [lojban] Re:
Vote for the Future Global Language)
But the Lojban isn't ambiguous in this way and hasn't been even in Loglan back
as far as I can remember. The fact that I can't remember which way the rule
runs or what the override technique is doesn't mean they aren't in place.
>> Xorxes proposed a rule that items in higher clauses have scope
>> over items in lower clauses (i.e. that items export to the prenex
>> of the localmost clause) and that when two items are in the same
>> clause, the leftward element has scope over the rightward. (It's a
>> shame to have to 'pollute' the purely hierarchical structure of
>> logical form with left-to-right order of forms, but it's by far
>> the simplest way to rescue Lojban in its (then) current state.
>> Perhaps the BPFK has made xorxes's rule official, in which case I
>> wonder what happened to the rule about the scope of selbri tcita
>> "na", and to the scope of selbri tcita in general.)
>
> I didn't know that was xorxes' rule; I thought left-to-right
> quantifier scope was in the CLL. Yes, indeed:
> http://dag.github.com/cll/16/5/ "The rule for dropping the prenex is
> simple: if the variables appear in the same order within the bridi
> as they did in the prenex, then the prenex is superfluous.".
>
> So, I'm probably failing to understand. Can you please explain it
> like I'm very very stupid?
My mistake -- failure of memory. Presumably the then-unofficial rule was to
generalize CLL's left-to-right rule for all elements in the bridi (with the
possible exception of some or all selbri tcita) and to make explicit the rule
that things export to the localmost rather than outermost prenex (when you have
one bridi within another).
**I also don't see what your example has to do with the rule you miss: since
ancient Loglan times the rule is that the quantifiers come out in order and here
there is not problems of subordinate clauses.
--And.
--
You received this message because you are subscribed to the Google Groups
"lojban" group.
And Rosta
> Well, the non-brivla stuff is, of course, also the non-predicate logic stuff, so
> we are beyond semantics into pragmatics -- about which we have even less of an
> idea than semantics.
Some of the syntactically nonbrivla stuff is still predicate logic stuff. Indeed I think the vast majority of cmavo can be translated into predicate logic (or an extended version thereof).
--And.
John E Clifford
----- Original Message ----
From: And Rosta <and....@gmail.com>
To: loj...@googlegroups.com
Sent: Thu, January 6, 2011 2:40:00 PM
Subject: Re: Lojban is *NOT* broken! Stop saying that! (was Re: [lojban] Re:
Vote for the Future Global Language)
Ian Johnson, On 06/01/2011 19:35:
> ...What? You're trying to distinguish between:
> ~(A(x) ^ B(x) ^ C(x))
> and
> ~A(x) ^ ~B(x) ^ ~C(x)
>
> Lojban has plenty of ways to do this.
No, the point is firstly that the difference is merely one of scope, and
secondly that Lojban too has (or perhaps merely had) the ambiguity that Lojbab
was reproaching English for having. The point was not that Lojban (or English)
has no way of disambiguating.
**But there is no (significant, perhaps) scope ambiguity in the Lojban and never
has been, so far as I can recall (after some early mucking about with prenex
forms). The *English* is screwed up in the way you suggest,
John E Clifford
----- Original Message ----
From: And Rosta <and....@gmail.com>
To: loj...@googlegroups.com
Sent: Thu, January 6, 2011 2:52:32 PM
Subject: Re: Lojban is *NOT* broken! Stop saying that! (was Re: [lojban] Re:
Vote for the Future Global Language)
** Interesting. I suppose you mean the prepositions, which merely extend the
predicates, and maybe the actual modals which would fit into modal predicate
logic (second order intensional). Attitudinals and the like go directly to
pragmatics, not passing Go. The plethora of connectives and RHE markers are all
part of the basic transferal into predicate logic. So, yes, you are right that
there is some non-brivla stuff that is still predicate logic or an easy (well,
...) extension of that. Unfortunately, I've forgotten the original, so I can't
tell whether is for or against some position.
--And.
And Rosta
> On Thu, Jan 06, 2011 at 06:37:20PM +0000, And Rosta wrote:
>> "Formal grammar" has a further meaning in linguistics, which is
>> "grammar formulated in an explicit way", and it's this meaning
>> that is relevant to the specification of a human language.
>
> Ah. I don't know that use. Can you point me to an example of such
> a thing?
Google for "formal linguistics" and you get this use or variations on it.
>>>> What a real grammar would do is define a set of correspondences
>>>> between sentence forms and sentence meanings.
>>>
>>> I don't know what that is, but it's not a formal grammar. Ask
>>> google if you don't believe me. :) I have no idea how you could
>>> formalize such a thing (and I'm not terribly sure I care, to be
>>> honest).
>>
>> If you think about it, I think you will find you do care.
>> Obviously the essential function of a language is to define
>> correspondences between forms and meanings. If your putative
>> specification of a language describes only possible forms and says
>> nothing of meanings, then it is simply not a specification of a
>> language. (Rather, it would be a specification of a "formal
>> language" in the sense referred to above.)
>
> Of course; the CLL does, in fact, cover semantics in quite a lot of
> detail
Of course.
>> As for you having no idea how to formalize such a thing, surely
>> you can imagine having and implementing the design goal of a
>> speakable predicate logic (which was one of Loglan's original
>> goals). Retrofitting such a thing onto existing Lojban would be
>> difficult,
>
> Wait what? How do we not have that?
We have it partially but not completely.
>> but surely the principle of it is easy to grasp: rules that take
>> the phonological forms of Lojban sentences and translate them into
>> predicate logic.
>
> That doesn't do anything for general semantics, though. IsRed(x) as
> a predicate is just a suggestively named lisp token ( see
> http://singinst.org/ourresearch/publications/GISAI/meta/glossary.html#gloss_lisp_tokens
> and http://lesswrong.com/lw/la/truly_part_of_you/ ); to formalize
> actual semantics in the way I think you're talking about, you need
> to formalize what it means for something to be Red.You can't do
> that in bare predicate logic; you'd do samething like
> HasWavelengthBetween(x,630nm,700nm), but that doesn't help, because
> now you have to have predicates for nanometers, and what a
> wavelength is, and on and on and on.
No, it's the job of the language specification to link the form /red/ to the (notional) encyclopedia entry for Red, but not to specify the content of the encyclopedia entry. Similarly, it is the job of the language to say that phoneme /c/ is realized as [S], but not to then define the phonetics (aerodynamics, acoustics, etc.) of [S].
So the semantic task of the language specification is (i) to define the meanings of terms that don't simply point to an encyclopedia entry and (ii) to define how meanings combine to yield sentence meanings.
> As far as I can tell, the semantic descriptions of Lojban in the CLL
> are about as good as can reasonably be achieved without falling down
> the rabbit hole of perfect semantic description, I don't see how it
> differs from "spoken predicate logic" in that respect, and I'm very
> curious as to whether you have evidence to the contrary.
CLL is partial but incomplete (which is not to derogate CLL's achievement or excellence). Thinking back to ten years ago, there were two main sorts of problem. One was that even where CLL specifies what X means and what Y means, it doesn't specify what X and Y mean when they occur together, especially which has scope over which -- i.e. the 'syntax of semantics'. The other was that important stuff such as kau constructions didn't have translations into predicate logic (or anything similar).
>> The major incompleteness is in the specification of
>> correspondences between forms and meanings (i.e. predicate logic).
>> I don't mean the definitions of individual brivla, but rather the
>> meanings of sentences containing nonbrivla stuff.
>
> I don't feel a significant lack there. If you do, please make
> updates to the Notes sections of the various BPFK pages so I can try
> to fix it.
I appreciate the offer, and once upon a time I devoted a huge chunk of my spare time to pointing out stuff that needs fixing -- to what was at the time a community led by Lojbab that believed nothing should be fixed and everything left to usage. Nowadays I lack the time and to some extent the motivation.
--And.
Craig Daniel
> On Thu, Jan 06, 2011 at 09:13:47AM -0800, Robin Lee Powell wrote:
>> On Thu, Jan 06, 2011 at 04:51:59PM +0000, And Rosta wrote:
>>
>> > What a real grammar would do is define a set of correspondences
>> > between sentence forms and sentence meanings.
>>
>> [...] I have no idea how you could formalize such a thing (and I'm not
>> terribly sure I care, to be honest).
>
> For what it's worth, ERG is an example of such a thing:
> http://www.delph-in.net/erg/
>
> To summarize your objection from further downthread: “Formal semantics is just a system for transforming one string of meaningless
> symbols into another string of meaningless symbols, so what's the point?”
>
> Well, to a certain extent you're right, but if you choose the right kind of semantic representation, you can do things like proving that two
> different strings of Lojban have the same meaning. Correct me if I'm wrong, but at the moment no machine grammar of Lojban
> represents the fact that “mi viska do” is equivalent to “do se viska mi”.
It is not equivalent, because of different emphasis. The details are
not explicitly spelled out and may well vary a fair bit from one
Lojbanist to another, but no human speaker is going to deliberately
pick a longer phrasing of something if they don't intend the choice to
be meaningful. The only difference is pragmatics, but one of them
violates a Gricean maxim, resulting in the implication that the other
choice was less well-suited to what the speaker wished to convey.
Subtle, but basic - well, basic to a subfield of linguistics that we
have been largely ignoring since JCB started the whole project. (I'm
not saying we *should* try to formally spell out Lojban pragmatics,
either; I'm not sure that particular task is tractable even in theory.
I'm just disagreeing that "mi viska do" and "do se viska mi" will ever
be equivalent in actual usage generated by any human or any software
whose linguistic behavior obeys similar principles.)
Esperanto was defined with a lot of morphology and little syntax; in
practice, it is as syntactically rich as any natlang, because human
language use works like that. Lojban as a very precisely defined
syntax and absolutely no official pragmatics, but all human language
use is affected by considerations of such.
> I don't think that we absolutely need to have such a thing, and I am certainly not volunteering to make it, but if we did have such a
> thing, I'm sure it would reveal one or two problems about Lojban grammar that no-one's thought about before.
Almost definitely.
And Rosta
>> Efforts along this line tend to involve and idealized
>> representational language, almost all of which end up looking a lot like first
>> order predicate logic, meaning that the crucial step in the process from
> Lojban
>> form to meaning would be -- with a few caveats -- a snap.
>
> Hopefully it would be a snap, but it's these rules that the formal
> definition/specification of the language requires, and not the formal grammar
> (save for whichever bits of the formal grammar are necessary for the
> form--meaning correspondence rules). Regarding the question of whether it would
> indeed be a snap, the requisite rules would in most cases need to be invented,
> so there'd be a political difficulty at least as much as a linguistic one.
>
> **I'm not following here. What is the political difficulty in given obvious
> rules for untangled conjoined terms or predicates or even blobs like briditail.
> There are some less than obvious places, to be sure, but doing the first bit is
> already more than we can do -- even by hand -- for any other language.
Until Robin assumed his dictatorial powers recently, it was politically impossible to get any change implemented.
--And.
John E Clifford
Lojban surface to pred log deep are mostly fairly straightforward. You say that
getting the changes (mainly, it seems, specifications) needed to work with the
remaining cases would be politically difficult (or, at least, involve convincing
Robin). Of course, as in the days before Robin's imperatorship, changes can and
will be made without telling anyone, everyone picking them up on the fly (see
the last decade, at least).
----- Original Message ----
From: And Rosta <and....@gmail.com>
To: loj...@googlegroups.com
Sent: Thu, January 6, 2011 3:53:37 PM
Subject: Re: Lojban is *NOT* broken! Stop saying that! (was Re: [lojban] Re:
Vote for the Future Global Language)
--And.
--
Piermaria Maraziti
> Formal semantics? I'm pretty much sure such a thing exists, but I have
> no idea how would one define one...
/delurking
You can google for "ontologies" - systems that formally defines semantics and relations between concepts.
Weel, it's more complex than only this :-)
A whole field of study by itself.
/lurking again
Ciao!
---8<----------------------------------------------fnord------
Piermaria Maraziti pier...@maraziti.it ICQ744473 MSN:kall...@hotmail.it
www.eridia.it www.hovistocose.info www.wildboar.it www.labasebianca.it
-- Verrà il giorno in cui la mistica nascita di Gesù, ad opera dell'Essere Supremo come Padre, nel grembo di una vergine, sarà considerata come la favola della generazione di Minerva dalla testa di Giove. T.Jefferson
David
On Jan 6, 8:55 am, Ivo Doko <ivo.d...@gmail.com> wrote:
> On 5 January 2011 22:58, Pierre Abbat <p...@phma.optus.nu> wrote:
>
> > Esperanto has at least one word which proves that its words cannot be
> > unambiguously parsed...
>
> There are multiple, but that is irrelevant. Like I said, Esperanto never
> even aimed to be fully unambiguous and as thousands of languages worldwide
> (Esperanto included, because it has native speakers) prove, a language
> doesn't *need* to be fully unambiguous to be a usable and working language.
It's worth mentioning at this point that Lojban is not fully
unambiguous nor is it intended to be AFAICS -- Lojban is fully
grammatically-unambiguous while culturally somewhat preferring
semantic ambiguity (since that allows shorter utterances).
>
> The main thing that Lojban lacks for being used as a global language is not
>
> > the precise definition of every corner case. It's vocabulary.
>
> I.e. it's not finished, which is what I said.
Let me point out something:
If 'lacking vocabulary' == 'not finished', then no language in
existence is finished. There is no such thing as a universal ontology,
so that sense of finished cannot be a useful distinction.
>
> ...its morphology is defined so as to prevent collisions like "avaro", it
>
> > takes
> > longer to invent vocabulary in Lojban. You can't take some Latinate term
> > that's commonly used in many languages, some of them unrelated to Latin,
> > and
> > expect to make a brivla out of it just by changing "-us" to "-o". You have
> > to
> > consider whether a lujvo would capture the meaning better, whether the
> > second
> > consonant is in a cluster, and whether the same word could mean something
> > totally different (such as "malpigi" which could be either an acerola fruit
> > or an insect's kidney).
>
> Speaking of which, I think that, unfortunately, is the main flaw of lojban.
> I understand that it can't possibly hope to be literally unambiguous if its
> vocabulary doesn't operate like that, but that ensures that if people ever
> do start to use lojban for everyday communication and if lojban ever gets
> native speakers, its so praised unambiguity will very soon melt away.
> Vocabulary assimilation is unavoidable and you can't possibly expect every
> native speaker of lojban to know which new brivla will create an ambiguity,
> so native lojban speakers would naturally start to incorporate words from
> other languages in their vocabulary, those words would inevitably create
> ambiguities, and after a couple of decades its precious ambiguity would be
> nowhere. (And that's without even mentioning other ways in which a language
> evolves when it's used by people as their main language for everyday
> communication.)
Where you say 'brivla' above, do you mean generally brivla, or the
subset which is fu'ivla/zi'evla? Because only the latter could
generate substantial ambiguity IMO
>
> So... as far as I've understood it, this is how it goes:
>
> 1) Let's make lojban the world's official common language because it's
> completely logical and unambiguous.
> 2) lojban is made the world's official common language.
> 3) People use lojban every day to talk to each other.
> 4) As was the case with Esperanto, this eventually results in people having
> lojban as their native language, who proceed to use lojban as their main
> language for everyday communication.
> 5) This makes lojban evolve.
> 6) After a couple of decades, lojban is no longer unambiguous nor completely
> logical and as time goes by is more and more like languages which have
> naturally evolved among humans.
I agree with your predictions here, they are logical; I'll bet that's
one of the reasons why historically we have said 'lojban is NOT aiming
to become a universal auxlang at all'
>
> Wait, so what was the initial reason to use lojban as the world's official
> common language? After all, lojban's unambiguity and logicality seems to be
> one of the main arguments for that, and yet if it did get chosen for that
> role it will have stopped being unambiguous and logical not long after its
> use became widespread. So if we're going to have an "ordinary" language as
> the world's official common language in the end anyway, why not chose one
> which is not unfinished?
I would be fascinated to see any language you can point to that is
'finished' in that sense.
Jonathan Jones
>
> On 6 January 2011 03:17, Jonathan Jones <eye...@gmail.com> wrote:
>>
>> Waiter: Do you want cream or sugar in your coffee?
>>
>> Lojbanist: Yes.
>
> Proper response is "I do." :D Unless you drink your coffee with neither...
If you want your coffee with neither, the correct answer is "no". "I do" and "yes" are merely semantically different - the meaning is the same in both responses.
>
> Also, that would depend on the waiter's emphasis of the sentence:
>
> "Do you want CREAM or SUGAR in your coffee?"
>
> means:
>
> "Is this true: (You want cream in your coffee. ⊕ You want sugar in your coffee.)?"
>
> "Do you WANT cream or sugar in your coffee?"
>
> means:
>
> "Is this true: (You want cream in your coffee. ∨ You want sugar in your coffee.)?"
>
> ("⊕" is XOR, "∨" is OR.)
>
> So presuming the waiter emphasised "cream" and "sugar", not "want", "yes" or "I do" are not proper responses.
It doesn't matter what the waiter emphasized.
Let us assign the following:
p = wants cream
q = wants sugar
(Taken directly from http://en.wikipedia.org/wiki/Truth_table#Logical_disjunction):
The truth table for p OR q (also written as p ∨ q, p || q, or p + q) is as follows:
| p | q | p ∨ q |
|---|---|---|
| T | T | T |
| T | F | T |
| F | T | T |
| F | F | F |
L2 (wants sugar, but not cream): Yes.
L3 (wants both cream and sugar): Yes.
L4 (wants neither): No.
To translate the waiter's sentence into Lojban:
W: xu do djica lo nu lo kruji .a lo sakta cu nenri lo ckafi
Lojban {.a}, by the way, is Logical OR. XOR is {.onai}.
--
mu'o mi'e .aionys.
.i.a'o.e'e ko cmima le bende pe lo pilno be denpa bu .i doi.luk. mi patfu do zo'o
(Come to the Dot Side! Luke, I am your father. :D )
Jorge Llambías
> Robin Lee Powell, On 06/01/2011 19:01:
>> On Thu, Jan 06, 2011 at 06:52:19PM +0000, And Rosta wrote:
>>>
>>> Xorxes proposed a rule that items in higher clauses have scope
>>> over items in lower clauses (i.e. that items export to the prenex
>>> of the localmost clause) and that when two items are in the same
>>> clause, the leftward element has scope over the rightward. (It's a
>>> shame to have to 'pollute' the purely hierarchical structure of
>>> logical form with left-to-right order of forms, but it's by far
>>> the simplest way to rescue Lojban in its (then) current state.
>>> Perhaps the BPFK has made xorxes's rule official, in which case I
>>> wonder what happened to the rule about the scope of selbri tcita
>>> "na", and to the scope of selbri tcita in general.)
>>
>> I didn't know that was xorxes' rule; I thought left-to-right
>> quantifier scope was in the CLL. Yes, indeed:
>> http://dag.github.com/cll/16/5/ "The rule for dropping the prenex is
>> simple: if the variables appear in the same order within the bridi
>> as they did in the prenex, then the prenex is superfluous.".
I precede CLL in the Lojbanic timeline, so the fact that it is in CLL
doesn't automatically rule out I might have had something to do with
it, but in fact the left-to-right rule was already in place when I
started learning Lojban.
I have often argued against the related "na" rule however, and the
contradictions and/or convolutions that it causes, and I am confident
that that will eventually be fixed.
> My mistake -- failure of memory. Presumably the then-unofficial rule was to
> generalize CLL's left-to-right rule for all elements in the bridi (with the
> possible exception of some or all selbri tcita) and to make explicit the
> rule that things export to the localmost rather than outermost prenex (when
> you have one bridi within another).
Yes, I think I did take part in discussions about that part. I'm not
sure the BPFK has said anything on that matter though, so even though
I have a well formed opinion on what scopes over what for most cases,
it is doubtful that there is any official last word on the matter. In
fact the CLL has a pretty confusing section on termsets and
quantifiers with "equal scope", so I would say the matter is actually
still officially ill-defined.
mu'o mi'e xorxes
John E Clifford
----- Original Message ----
From: David <00a...@gmail.com>
To: lojban <loj...@googlegroups.com>
Sent: Thu, January 6, 2011 6:11:18 PM
Subject: [lojban] Re: Lojban is *NOT* broken! Stop saying that!
On Jan 6, 8:55 am, Ivo Doko <ivo.d...@gmail.com> wrote:
> On 5 January 2011 22:58, Pierre Abbat <p...@phma.optus.nu> wrote:
>
> > Esperanto has at least one word which proves that its words cannot be
> > unambiguously parsed...
>
> There are multiple, but that is irrelevant. Like I said, Esperanto never
> even aimed to be fully unambiguous and as thousands of languages worldwide
> (Esperanto included, because it has native speakers) prove, a language
> doesn't *need* to be fully unambiguous to be a usable and working language.
It's worth mentioning at this point that Lojban is not fully
unambiguous nor is it intended to be AFAICS -- Lojban is fully
grammatically-unambiguous while culturally somewhat preferring
semantic ambiguity (since that allows shorter utterances).
**It doesn't have a lot of choice, since it uses words in a culture of sorts.
>
> The main thing that Lojban lacks for being used as a global language is not
>
> > the precise definition of every corner case. It's vocabulary.
>
> I.e. it's not finished, which is what I said.
Let me point out something:
If 'lacking vocabulary' == 'not finished', then no language in
existence is finished. There is no such thing as a universal ontology,
so that sense of finished cannot be a useful distinction.
**universal ontology? A catalog of everything? a theory of what there is or
might be? I suppose philosophical languages do shoot for that, but Lojban does
not. It merely hopes that it can be extended to meet every new situation.
** Gismu don't generate ambiguity, they just are ambiguous. It comes with the
language game. And where two or theree are gathered together, it gets
exponentially worse (if ambiguity is a bad thing).
>
> So... as far as I've understood it, this is how it goes:
>
> 1) Let's make lojban the world's official common language because it's
> completely logical and unambiguous.
> 2) lojban is made the world's official common language.
> 3) People use lojban every day to talk to each other.
> 4) As was the case with Esperanto, this eventually results in people having
> lojban as their native language, who proceed to use lojban as their main
> language for everyday communication.
> 5) This makes lojban evolve.
> 6) After a couple of decades, lojban is no longer unambiguous nor completely
> logical and as time goes by is more and more like languages which have
> naturally evolved among humans.
I agree with your predictions here, they are logical; I'll bet that's
one of the reasons why historically we have said 'lojban is NOT aiming
to become a universal auxlang at all'
**Aside from objecting tho this use of "logical" (Spock has a lot to answer
for), I think the predictions are largely unlikely to happen, even if Lojban
were declared THE international auxiliary language. Lojban is not meant to be
an auxlang because 1) it takes too much thought for most people to bother with
and 2) the effing politics takes attention away from more interesting things.
>
> Wait, so what was the initial reason to use lojban as the world's official
> common language? After all, lojban's unambiguity and logicality seems to be
> one of the main arguments for that, and yet if it did get chosen for that
> role it will have stopped being unambiguous and logical not long after its
> use became widespread. So if we're going to have an "ordinary" language as
> the world's official common language in the end anyway, why not chose one
> which is not unfinished?
I would be fascinated to see any language you can point to that is
'finished' in that sense.
**I'm not sure yet what "unfinished" means when applied to a language --
especially if it means something that Lojban is but some other language is not.
John E Clifford
From: Jonathan Jones <eye...@gmail.com>
To: loj...@googlegroups.com
Sent: Thu, January 6, 2011 6:52:01 PM
Subject: Re: Lojban is *NOT* broken! Stop saying that! (was Re: [lojban] Re: Vote for the Future Global Language)
Jorge Llambías
>
> This is one of those little problems that needs to be solved, so the the answers: cream, sugar, both or neither are all correct.
Hasn't this been solved for ages? The answer is that question-or is
neither ".a" nor ".onai". Question-or is "ji", and the four helpful
answers are ".enai", "na.e", ".e" and "na.enai". (There are also
plenty of true but less than helpful answers, of course, but that's
the case for any question.)
Robin Lee Powell
> Robin Lee Powell, On 06/01/2011 18:51:
> >updates to the Notes sections of the various BPFK pages so I can
> >try to fix it.
>
> I appreciate the offer, and once upon a time I devoted a huge
> chunk of my spare time to pointing out stuff that needs fixing --
> to what was at the time a community led by Lojbab that believed
> nothing should be fixed and everything left to usage. Nowadays I
> lack the time and to some extent the motivation.
That's too bad. When CLLv2 comes out, see if you feel differently.
:)
Robin Lee Powell
> I think we are at cross purposes here. I say that the rules for
> shifting from Lojban surface to pred log deep are mostly fairly
> straightforward.
I agree.