Cultural fu'ivla: summary and list of the ISO generated ones
Leo Molas
countries (ISO 3166), languages (ISO 639-3) and currencies (ISO 4217).
The script also generates a jbovlaste-like definitions, separated with
';', so I created a table (like .csc), and upload it to google docs. (If
you are not confortable using it, you can download the file in several
different formats.) I did this because the idea is to mass upload them;
I asked Robin if he can do that, and he only needed that the files had
that format.
My idea is to discuss now if there is something wrong with the
definitions. You can change little things if you feel it, and if you
want to change something in all the definitions, tell me and I'll change
it in the script. Those files are going to be the ones I give Robin to
upload to jbovlaste.
The columns are:
- valsi
- definition
- Prefix (I don't know if this should go there.)
- gloss word (this can be changed)
- in the sense of (this too)
The last points are where I feel insecure. I can also make a column with
"comments", telling that this are ISO generated or whatever... and can
also have a link to other word meaning the same (like mergu'e in gugde'usu).
The files are:
http://spreadsheets.google.com/ccc?key=0Ai4mbIPr2PUwdENVckVaOTJLbVdGVE5SeGJ0MTBTbGc&hl=en
: countries (247 fu'ivla) and currencies (172)
http://spreadsheets.google.com/ccc?key=0Ai4mbIPr2PUwdGNYUFhrdmV1cmpLeVVQZkNSU3dBYUE&hl=en
: languages part 1 (from banga'a'a to banlukusu, 3687 fu'ivla)
http://spreadsheets.google.com/ccc?key=0Ai4mbIPr2PUwdC1ENjlQTmZ1ZjlidFAwWDdvZDBUUWc&hl=en
: languages part 2 (from banlukutu to banzuzuju, 4232 fu'ivla)
That gives us a total of 8338 fu'ivla .uesai
I know it's a really huge number, but this may be really useful .ui
mu'o mi'e .leos.
--
My lojban journal: http://learninglojban.wordpress.com
My personal blog: http://leomolas.tumblr.com
Christopher Doty
Jorge Llambías
> I am still absolutely, one hundred percent opposed to the idea of having
> 7,000 words with an identical first syllable for related concepts. This is
> not how language works, and it is not how the human brain works.
> Jorge brought up an example previously of things (something) like "Iceland,
> Swaziland, etc." to point out that words having the same morpheme in the
> same place wasn't an issue. But, again, this is demonstrably NOT how
> languages work. Having the same suffix on a word is not the same as having
> the same prefix. Having the same prefix will make the words VERY hard to
> learn.
Doesn't (Ki)Swahili name all languages with a ki- prefix? (This may be
yet another myth, like that all country names end in "guó" in Chinese,
so I'm just asking, but I tend to be skeptical about claims that this
is "not how the human brain works".)
But if this is demonstrably not how languages work, can you give some
pointers to the demonstration?
mu'o mi'e xorxes
Jonathan Jones
1) Having gismu for some cultures and not others is biased, and therefore violates the "culturally neutral" stance of Lojban. The only acceptable solution to this is to either have a word for *all* cultures, or to not have a word for *any* of them. I support the latter: lujvo and fu'ivla are open space, let jbopre create the word they want to use *when they need it*.
2) The cultural gismu basically have too many meanings. For example, {merko} means "x1 pertains to USA/American culture/nationality/dialect in aspect x2". The solution to this is to separate the various bits into individual words, without reference to a specific culture. I offered my suggestion (the lujvo I posted earlier) as to what words to use instead of the gismu.
As far as I'm concerned, *it doesn't matter* what words to use for the languages/countries/cultures/what-have-you, as long as the people can understand what you're talking about.
--
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.
--
mu'o mi'e .aionys.
.i.a'o.e'e ko klama le bende pe denpa bu
Jonathan Jones
Considering that this discussion started out as "the cultural gismu are are bad idea and should be replaced", I honestly don't care what people choose to call any of those things. The reasons given as to why the aforesaid gismu are a bad idea are twofold:
1) Having gismu for some cultures and not others is biased, and therefore violates the "culturally neutral" stance of Lojban. The only acceptable solution to this is to either have a word for *all* cultures, or to not have a word for *any* of them. I support the latter: lujvo and fu'ivla are open space, let jbopre create the word they want to use *when they need it*.
2) The cultural gismu basically have too many meanings. For example, {merko} means "x1 pertains to USA/American culture/nationality/dialect in aspect x2". The solution to this is to separate the various bits into individual words, without reference to a specific culture. I offered my suggestion (the lujvo I posted earlier) as to what words to use instead of the gismu.
geography - matytutra "m1/t1 corresponds to territory of m2/t2 in aspect m3"
Jorge Llambías
>
> My idea is to discuss now if there is something wrong with the
> definitions. You can change little things if you feel it, and if you
> want to change something in all the definitions, tell me and I'll change
> it in the script.
I would definitely change the place structure for currencies from:
"x1 is the curency with ISO 4217 code "AED" (United Arab Emirates, Dirhams)."
to:
"x1 is measured in curency with code ISO-4217 "AED" (United Arab
Emirates, Dirhams) as x2 (number)."
This is the form that all measurement units take in Lojban.
Leo Molas
http://spreadsheets.google.com/ccc?key=0Ai4mbIPr2PUwdGVWcWpsSTVBQ0htZS1pb3pTYUNSNVE&hl=en
I understand that the list seems awful, but they are not terrible. You
don't always talk about 7000 countries and their languages and
currencies. You don't know all their names. And, finally, no one forces
anyone to know all of them; if someone doesn't like a word, just creates
another one and use it. You may link that new word in the comments of
the word in jbovlaste.
The seem difficult to learn, but again, you've only to know 2 or 3
letters, and most of the times they resemble what you're talking about.
The prefix are useful as they are in lujvo, or fu'ivla type-3; you may
not know which, for example, country is the other person talking about,
but you'll know is a country (and you can ask later which country is it).
mu'o mi'e leos
> --
> 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.
Christopher Doty
Doesn't (Ki)Swahili name all languages with a ki- prefix? (This may be
yet another myth, like that all country names end in "guó" in Chinese,
so I'm just asking, but I tend to be skeptical about claims that this
is "not how the human brain works".)
But if this is demonstrably not how languages work, can you give some
pointers to the demonstration?
Christopher Doty
I understand that the list seems awful, but they are not terrible. You
don't always talk about 7000 countries and their languages and
currencies. You don't know all their names. And, finally, no one forces
anyone to know all of them;
if someone doesn't like a word, just creates
another one and use it.
Yoav Nir
Bob LeChevalier
> 1) Having gismu for some cultures and not others is biased, and
> therefore violates the "culturally neutral" stance of Lojban.
Only in the same sense that basing the gismu-making process on only 6 of
the 12 most-spoken languages is culturally-biased, because it excludes
all the rest.
If one thinks only about gismu, then perhaps it is. If one considers
fu'ivla made from the name of the language/culture IN the language, then
we've more than neutralized that bias.
Gismu are shorter than fu'ivla, reflecting a presumption under Zipf's
law of higher usage frequency, and references by speakers to their own
language is a high frequency concept, so giving shorter words to
more-spoken languages is one form of cultural-neutrality.
> The only
> acceptable solution to this is to either have a word for *all* cultures,
> or to not have a word for *any* of them.
We decided otherwise.
> 2) The cultural gismu basically have too many meanings. For example,
> {merko} means "x1 pertains to USA/American culture/nationality/dialect
> in aspect x2". The solution to this is to
... make lujvo based on them. And in fu'ivla space, make culture words
using the experimental space that allows fu'ivla to be made into rafsi
and combined with the gismu used with the cultural gismu to make lujvo.
lojbab
--
Bob LeChevalier loj...@lojban.org www.lojban.org
President and Founder, The Logical Language Group, Inc.
Jonathan Jones
Jonathan Jones wrote:Only in the same sense that basing the gismu-making process on only 6 of the 12 most-spoken languages is culturally-biased, because it excludes all the rest.
1) Having gismu for some cultures and not others is biased, and therefore violates the "culturally neutral" stance of Lojban.
If one thinks only about gismu, then perhaps it is. If one considers fu'ivla made from the name of the language/culture IN the language, then we've more than neutralized that bias.
Gismu are shorter than fu'ivla, reflecting a presumption under Zipf's law of higher usage frequency, and references by speakers to their own language is a high frequency concept, so giving shorter words to more-spoken languages is one form of cultural-neutrality.
That's a very good argument, and one I wish had been presented at least a week ago.
We decided otherwise.
> The only
acceptable solution to this is to either have a word for *all* cultures, or to not have a word for *any* of them.
... make lujvo based on them. And in fu'ivla space, make culture words using the experimental space that allows fu'ivla to be made into rafsi and combined with the gismu used with the cultural gismu to make lujvo.
2) The cultural gismu basically have too many meanings. For example, {merko} means "x1 pertains to USA/American culture/nationality/dialect in aspect x2". The solution to this is to
lojbab
--
Bob LeChevalier loj...@lojban.org www.lojban.org
President and Founder, The Logical Language Group, Inc.
--
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.
Jorge Llambías
>
> The situation with Swahili (and other Bantu languages) is different for any
> number of reasons. First off, the language prefix exists as part of a
> system in which EVERY NOUN starts with a prefix that indicates its noun
> class. The prefixes thus exist as part of an entire system; it's not just
> something that language names do, it is something that all nouns everywhere
> in the language do.
Are you saying that prefixes are acceptable to the human brain as long
as all words in the language use them, but
unacceptable/hard-to-deal-with if only a restricted class of words use
them? That doesn't sound plausible to me.
> Plus, the prefixes on the nouns (or, more
> appropriately, the class of nouns to which a word belongs, which is
> indicated by the prefix) are used for a variety of syntactic functions--you
> can tell who is doing what to whom by virtue of prefixes that go on the
> verbs, which correspond to the noun classes that are marked on the nouns.
So prefixes are ok as long as they do enough work, but not if they
don't do enough work? (In the present case, they serve two purposes:
morphologically, they help to identify the word as belonging to a
certain syntactic class, a brivla, semantically, they indicate what
class of things the predicate refers to. Isn't that enough load?
> There are any number of experiment verifications of this:
> your brain processes language as it goes along, and if that information is
> not helpful or is not what it expected, it creates problems, because your
> brain now has to backtrack or, in the case of having language names start
> with the same thing, has to suddenly sort through the whole list, instead of
> having paired down in steps.
But the ban-/baur- prefix is quite helpful! It tells you to expect a
brivla, and furthemore, one related to a language.
> My point with starting everything with some form of "bang" means that your
> brain will not be able to do anything at all to get any closer to what you
> were trying to say. Plus, it won't help me if we're in a loud room and I
> don't quite here what you said. If the first element was meaningful with a
> system, then it can help in both of those ways.
But the prefix is meaningful! Perhaps a bit less meaningful than a
language-family prefix might be, but more meaningful than a random
string of letters.
tijlan
>
> In terms of "how the brain works," what I mean is that the brain processes speech linearly. But, as with everything in nature, the brain is also lazy, and tries to use context to determine what is coming up. So, imagine that I say, "The longest word in the English language is a..."
> At this point, your brain has already thrown out a bunch of possibilities. It's thrown out anything that doesn't start with /a/. It's also likely already thrown out short words, because I'm looking for something long. "Ant" is likely not part of the set of words that my brain is looking at.
> "...nt..."
> I still know it isn't ant, but now I also know what it starts with, and what it doesn't. By the time I get to:
> "...idises..."
In French, adjectives are basically backtrack:
langue construite (constructed language)
So, if the speaker starts with "langue", the listener waits for the
following element which specifies the kind of "langue", which is
"construite". And this is how my brain would process the fu'ivla
"banjubu'o" as well. "ban...", language, "...jubu'o", Lojban: a
language which is Lojban.
In Japanese (and Turkish, Tamil, Basque etc.), the modifier-nucleus
order can be almost reverse to English (and Spanish, Arabic, etc.):
[1]What [-]do [2]you [3]think [4]about [5]this [6]language [-]which
[-]has [8]no [9]syntactical ambiguity?
[9]koubunteki-aimaisei-ga [8]nai [5]kono [6]gengo-[4]o [2]anata-wa
[1]dou [3]omou?
Literally:
syntactical ambiguity no this language about you what think?
As it goes along, "-o" (about) is followed by "anata" (you); do you
consider this flow of information helpful? If it's not helpful, it
would create problems, according to you. In fact, the word with the
most important sentential information, the predicate, "omou" (think),
comes at the end of the expression. But proficient Japanese speakers
are not troubled by that, since their brain works such that it
captures the sentence non-linearly, in its entirety, after all of its
components have been laid out. It's processing 'as it has gone' rather
than 'as it goes along'.
> or, in the case of having language names start with the same thing, has to suddenly sort through the whole list, instead of having paired down in steps.
Still for language names that don't start with the same thing, like
"jbobau", the brain has to sort through the whole list of items evoked
by that leading element. "jbo...", jbota'a? jboce'u? jboge'a? jbopre?
and so on, until it gets decided with "...bau".
mu'o mi'e tijlan
Christopher Doty
> or, in the case of having language names start with the same thing, has to suddenly sort through the whole list, instead of having paired down in steps.Still for language names that don't start with the same thing, like
"jbobau", the brain has to sort through the whole list of items evoked
by that leading element. "jbo...", jbota'a? jboce'u? jboge'a? jbopre?
and so on, until it gets decided with "...bau".
Bob LeChevalier
> I mean, I think I kind of understood this last part. I thought the idea
> was that Lojban would perhaps, at some point, serve a rather
> international function. If that's really not the goal, that's
> fine--I'll shut up and remove myself from the lists.
I hope you don't.
But, no, that is not "the" goal, and was in fact explicitly disavowed as
such because we offend Esperantists who think we are competing with
them, and we offend linguists, who as a class tend to disparage the
"international language" movement (and this disavowal indeed helped me
in discussing Lojban with the linguistic community).
But it is a goal for some Lojbanists, and it is one that we do not
disparage, even while not making it a formal goal that might drive
decision-making.
There is certainly an international aspect to Lojban in the sense that
any linguistics research using Lojban would need to use speakers of
multiple natural languages, especially of unrelated families, to give
credible results.
> Real languages do things that are WAY more mind-bending than Lojban; I have no reason to
> be interested in or study the language if it's only purpose is to
> mind-bend, since it fails at that, as far as I'm concerned.
I'm sure there are a few things that would qualify %^)
As an example for a linguist: All languages have raising phenomena, but
Lojban's sumti-raising is both explicit and extremely powerful (and I
personally find lambda calculus to be more mind-bending than I can
handle %^).
Most languages have several words corresponding to our attitudinals, but
again I think Lojban goes far beyond any of them, and allows productive
compounding, which I don't think exists in natural language attitudinals
(though I won't be offended if you know an example).
The tense stuff all comes from natural languages - several different
ones. No *one* language puts so many tools for expressing tense in one
place: ZAhO, TAhE, etc. And then there are the time-travel tenses.
Will speakers use all these things? That's what we'll hopefully find out.
But if we put too much "real languages do things this way" into Lojban,
we are basically limiting the possibilities. We don't know all the ways
that real languages do things, and we cannot know the full range of how
they MIGHT do them. Lojban is a test bed that allow people to try
things that they cannot in their native language, and to use expressive
tools in combination that no extent natural language has.
tijlan
> At this point, your brain has already thrown out a bunch of possibilities. It's thrown out anything that doesn't start with /a/. It's also likely already thrown out short words, because I'm looking for something long. "Ant" is likely not part of the set of words that my brain is looking at.
> "...nt..."
> I still know it isn't ant, but now I also know what it starts with, and what it doesn't. By the time I get to:
> "...idises..."
In French, adjectives are basically backtrack:
langue construite (constructed language)
So, if the speaker starts with "langue", the listener waits for the following element which specifies the kind of "langue", which is "construite". And this is how my brain would process the fu'ivla "banjubu'o" as well. "ban...", language, "...jubu'o", Lojban: a language which is Lojban.
In Japanese (and Turkish, Tamil, Basque etc.), the modifier-nucleus order can be almost reverse to English (and Spanish, Arabic, etc.):
[1]What [-]do [2]you [3]think [4]about [5]this [6]language [-]which [-]has [8]no [9]syntactical ambiguity?
[9]koubunteki-aimaisei-ga [8]nai [5]kono [6]gengo-[4]o [2]anata-wa [1]dou [3]omou?
Literally:
syntactical ambiguity no this language about you what think?
As it goes along, "-o" (about) is followed by "anata" (you); do you consider this flow of information helpful? If it's not helpful, it would create problems, according to you. In fact, the word with the most important sentential information, the predicate, "omou" (think), comes at the end of the expression. But proficient Japanese speakers are not troubled by that, since their brain works such that it captures the sentence non-linearly, in its entirety, after all of its components have been laid out. It's processing 'as it has gone' rather than 'as it goes along'.
Bob LeChevalier
> My point with starting everything with some form of "bang" means that
> your brain will not be able to do anything at all to get any closer to
> what you were trying to say.
???
Just the opposite, I think.
In Lojban with lots of type 3 fu'ivla, they know that in all likelihood
that you are talking about some kind of language. In many instances,
the most critical information is that it is some kind of language, in
others perhaps, context makes it clear that a language is being referred
to and the bang- part is perhaps superfluous, but serves the cause of
linguistic redundancy, something Lojban is often criticized as lacking,
but is especially applicable to your noisy room problem.
lojbab
tijlan
slightly broken.
tijlan
> Nothing I said had anything to do with order of words in a phrase, it had to
> do with words starting with the same set of sounds. Your brain does a
> lexical search for words, not whole sentences.
But when I see/hear "banjubu'o", I don't stop at "ban..."; immediately
"...jubu'o" comes to break the searching process. And why should the
brain start the analysis upon perceiving the first component rather
than its combination with the following component(s)?
> > Still for language names that don't start with the same thing, like
> > "jbobau", the brain has to sort through the whole list of items evoked
> > by that leading element. "jbo...", jbota'a? jboce'u? jboge'a? jbopre?
> > and so on, until it gets decided with "...bau".
>
> I'm not sure what you're getting at here. Unless there are 7,000 words with
> related meanings that all start with jbo, this isn't an issue.
And unless 1) I recognise every one of those 7,000 words, 2) I fail to
grasp the context, and 3) I fail to catch the latter part of the word
in question, "banjubu'o" is not so much a problem for me.
Pierre Abbat
> Still for language names that don't start with the same thing, like
> "jbobau", the brain has to sort through the whole list of items evoked by
> that leading element. "jbo...", jbota'a? jboce'u? jboge'a? jbopre? and so
> on, until it gets decided with "...bau".
Not to mention "jbonobo" and "jboia" ;)
Pierre
--
lo ponse be lo mruli po'o cu ga'ezga roda lo ka dinko
Michael Turniansky
> My point with starting everything with some form of "bang" means that your
> brain will not be able to do anything at all to get any closer to what you
> were trying to say. Plus, it won't help me if we're in a loud room and I
> don't quite here what you said. If the first element was meaningful with a
> system, then it can help in both of those ways.
>
Speaking as a professional Scrabble® player, there are 4012 words
legal in North American/Israeli/Malaysian English Scrabble that all
start with UN-. (5259 in other countries). The vast majority mean
"the opposite of (the rest of this word)". So the same argument can
be brought to bear about English. Seems to me that languages *do*
work that way, in broad strokes.
--gejyspa, still trying valiantly to plough through
the lojban email backlog. Still haven't gotten to the thread with
Robin's essay link, yet.
Leo Molas
> Jonathan Jones wrote:
>> 1) Having gismu for some cultures and not others is biased, and
>> therefore violates the "culturally neutral" stance of Lojban.
>
> Only in the same sense that basing the gismu-making process on only 6 of
> the 12 most-spoken languages is culturally-biased, because it excludes
> all the rest.
This is not that accurate, since lojban doesn't sound like any of those.
However, I don't think a guaraní speaker would be hurt by that; but not
having a gismu for Paraguay (and having one for argentina).
>
> If one thinks only about gismu, then perhaps it is. If one considers
> fu'ivla made from the name of the language/culture IN the language, then
> we've more than neutralized that bias.
Making fu'ivla like that would be the best; I've searched for a list
with the languages names as said by natives, but I failed. That's why I
thought of "ISO generated" as the best possible solution.
>
> Gismu are shorter than fu'ivla, reflecting a presumption under Zipf's
> law of higher usage frequency, and references by speakers to their own
> language is a high frequency concept, so giving shorter words to
> more-spoken languages is one form of cultural-neutrality.
>
This is a good argument. However, it doesn't work for many many people
from the other countries.
>> The only
>> acceptable solution to this is to either have a word for *all*
>> cultures, or to not have a word for *any* of them.
>
> We decided otherwise.
This is not a good argument...
>
>> 2) The cultural gismu basically have too many meanings. For example,
>> {merko} means "x1 pertains to USA/American culture/nationality/dialect
>> in aspect x2". The solution to this is to
>
> ... make lujvo based on them. And in fu'ivla space, make culture words
> using the experimental space that allows fu'ivla to be made into rafsi
> and combined with the gismu used with the cultural gismu to make lujvo.
>
> lojbab
In this discussion, I've seen people who are against the cultural gismu,
people who don't care about it, and conservatives that don't want to
change a thing. But, besides what I marked IMO as a good argument, I
haven't read much of those.
I understand if there are people that won't like this change, but I
think others that, like myself, don't like the cultural gismu, won't use
them; we'll create fu'ivla, and encourage peers to use it.
MorphemeAddict
tijlan
Jorge Llambías
> Why does "san" (Sanskrit) become "bansu'anu" rather than less arbitrary
> "bansa'ana"?
s -> su
a -> 'a
n -> nu
If you are going to choose the vowel that goes with each consonant by
some other method on a case by case basis, the rule becomes more
arbitrary, not less. And you have to decide what to do with:
h -> xe
q -> ke
w -> ve
y -> je
mu'o mi'e xorxes