[lojban] Re: emotions
Jorge Llamb�as
--- Robin Lee Powell <rlpo...@digitalkingdom.org> wrote:
> > > lo gerku cu zdani -- A dog is a house.
> >
> > lo gerku cu zdani lo risna curnu - A dog houses heartworms.
>
> Fine fine. lo gerku cu dilnu, then.
>
> > (I would rather say {selparji}, of course.)
>
> Which is what?
What's a dictionary for? :)
http://www.lojban.org/jbovlaste/dict/parji
mu'o mi'e xorxes
__________________________________
Do you Yahoo!?
The New Yahoo! Search - Faster. Easier. Bingo.
http://search.yahoo.com
And Rosta
is bad per se. I say that deviation is bad in proportion to the
adverse consequences it has (in terms of causing us to not speak
the same language). There's not much more to be said about that.
(I agree with xorxes that ironically the discussedness of
experimentals means that they stick in the memory better than
many official words. For instance, I have no idea what the
official word for 'fungus' is, but I do know 'gumri'. And
among the cultural gismu, I no longer have any notion of
which are and aren't official, except for obvious cases.)
--And.
------------------------ Yahoo! Groups Sponsor ---------------------~-->
Get A Free Psychic Reading! Your Online Answer To Life's Important Questions.
http://us.click.yahoo.com/Lj3uPC/Me7FAA/CNxFAA/GSaulB/TM
---------------------------------------------------------------------~->
To unsubscribe, send mail to lojban-un...@onelist.com
Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
Robert LeChevalier
>la djorden cusku di'e
> > [...]
> > > > Gismu, lujvo and fu'ivla
> > > > are just different types of brivla; none are more privledged than
> > > > the others.
> > >
> > > If that were so, then why all the fuss when a new gismu is proposed,
> > > but no fuss when a new lujvo or fu'ivla is proposed?
> >
> > Because the gismu list is frozen, and there's no reason to prefer
> > a gismu rather than another brivla.
>
>And yet people do seem to prefer gismu. They are used much more frequently
>than other brivla.
That's because the gismu list is handy, and there is no equivalently handy
dictionary of the other words. They might also have used LogFlash or some
other tool to memorize them. You use what you know.
> > If you ask why it is frozen (as you no doubt will do); it would not
> > be smart to allow open season on gismu: the knee-jerk creation of
> > gismu for things which can be other brivla damage our ability to
> > increase rafsi for lujvo in the future (and in the process betrays
> > a lack of understanding about the purpose of the different types
> > of brivla).
>
>Hmm... {parji} doesn't seem to limit any increase in future rafsi
>availability, because par, paj, pai, pa'i are all already taken.
In which case, there is even less reason to give it a gismu rather than a
rafsi fu'ivla form.
>In any case, the best way to oppose these words is to provide
>good canonical alternatives, as Nick said.
Nora suggested that no noncanonical word should be allowed to be added to
the jbovlaste UNLESS a canonical (perhaps longer) alternative has already
been added.
lojbab
--
lojbab loj...@lojban.org
Bob LeChevalier, President, The Logical Language Group, Inc.
2904 Beau Lane, Fairfax VA 22031-1303 USA 703-385-0273
Artificial language Loglan/Lojban: http://www.lojban.org
And Rosta
> At 03:32 AM 5/25/03 -0400, Rob Speer wrote:
> >On Sat, May 24, 2003 at 01:16:29PM -0400, Robert LeChevalier wrote:
> > > Why is it important that infrequently used jargon words have very short
> > > forms?
> >
> >(glances at lau, tei, and foi)
>
> Content words. Obviously in Lojban, cmavo will be shorter than any content
> word,
Not in principle. There are CVVVVVVVVVVVVVVV cmavo, in principle.
> (I'm
> surprised that there has seldom been criticism of re'a, which I myself
> thought was a questionable addition, deferring to people with more
> mathematical orientations.)
It's disyllabic. There is no shortage of bisyllabic cmavo space.
> it is important that there be short ways to say acronyms even when they
> don't use Lojban-alphabet lerfu,
Not so important that the need has manifested itself in usage yet...
--And.
Robert LeChevalier
>JCB had the habit in the 70s of coining a new gismu that seemed useful at
>the moment, but which really had no justification as a root word in the
>language. Thus TLI Loglan has gismu for "olive", "football" (ambiguously
>never defined as to whether he meant this to be soccer or American rugby),
>sodium, beefsteak, and billiards. All of these should have been lujvo, or
>fu'ivla if there was no obvious lujvo. But in the 70s, there was no
>distinct fu'ivla form, so borrowings were in the morphological form of
>gismu and lujvo.
>
>I'm fighting the recurrence of this bad habit.
First of all, Nora notes that sodium is still a gismu (sheepish grin),
though our reason was different.
Nora adds several arguments against ad hoc expansion of the gismu list by
simple addition to jbovlaste, which I summarize.
1. every added gismu makes the goal of "learning the gismu list", a worthy
goal for new Lojbanists, that much harder
2. every added gismu makes the goal of learning rafsi (or deducing their
meaning) that much harder. Assume that parji is added even with no rafsi
assigned. Because it is there, then when you see rafsi paj, par, pai, or
pa'i, or even pra, then this is one more gismu that they MIGHT be, and
hence a little harder to learn.
3. all of the gismu added, whether people agree they should be or not, went
through a certain amount of debate before we even made a gismu for
them. The sheer necessity of looking up a word in 6 languages means that
we had to consider the meaning carefully, so we'd know what to look up, and
there were at least three of us involved in looking up words, so we
therefore always debated (and Tommy and I had MANY long debates, since he
was a gismu minimalist - as few as possible).
4. Once we got past the basic start of analyzing, weeding, and redoing the
TLI Loglan list words, words were added only with a careful consideration
of a)semantic completeness (e.g. of sets of food-grains), b) usability in
lujvo to cover semantic space. New words should have to be justified in
terms of necessity AS GISMU.
5. Words made from one language, as parji was, should be fu'ivla. Whether
people think there is a lot of meaning to the 6-language word-making, it
offers a couple of things: an objective way to decide the "best form",
dissociation of the word from the keyword in any single source language, so
that it is less likely to be encoded English (or whatever language). This
is also why fu'ivla should be dispreferred when one can make a lujvo: a
lujvo has its own lojbanic meaning, whereas a fu'ivla starts with the
meaning in some other language and is not really lojbanic. lujvo-making
forces you to think about meaning, and jvajvo force you to think about
place structures (whether you choose to follow jvojva or not, considering
them is a good idea).
Nora looked up other experimental gismu in jbovlaste, and points
out that even more than parji, "mango" has no business as a gismu, and
benzo is almost as questionable.
6. (hard to explain) the list of existing gismu slants the choice of how
one makes and interprets lujvo. The semantics of the language is based on
what has gone before. Adding a new gismu to the coverage of semantic space
changes the semantic map, and thus could change the color of meaning of
other words in unexpected ways.
7. Without disparaging the contributions of new people to the language,
there is a tendency of many new people to, early in their Lojbanic career,
say "it would have been better to do it 'this way'" without fully
understanding the reasons why it was done 'the other way', so they advocate
for change without learning the language as it is. Without baseline
controls, the momentum of LOTS of usage, and a dictionary with words of all
varieties so that people can find most of the words they want without
inventing them, coining new gismu for every concept they want to say, is
natural. I myself am guilty of this, with my favorite "pitsa", but I would
never argue for adding it to the gismu list because I know better (and I
don't really care to make more gismu for pepperoni, sausage, peppers, ham,
and pineapple %^)
If it is "easy" to add words without thinking about meanings,
place structures, people will do so. I contend that, for gismu, this is
NOT a good thing.
8. Finally, before there was a byfy, adding gismu to the original baseline
list was consider fundamental enough that each one was put to a membership
vote (at LogFest). People were expected to make a case for their word and
submit it for consideration by the members, and to abide by the
result. Hence I abided by the elimination of gumri. The current method of
putting words out there, and having them see usage without the debate,
without the research, without the discussion, and without abiding by what
was decided in the past, is disparaging of stability, tradition, and the
opinions of members who put time and effort into the language in the past.
lojbab
--
lojbab loj...@lojban.org
Bob LeChevalier, President, The Logical Language Group, Inc.
2904 Beau Lane, Fairfax VA 22031-1303 USA 703-385-0273
Artificial language Loglan/Lojban: http://www.lojban.org
------------------------ Yahoo! Groups Sponsor ---------------------~-->
Robert LeChevalier
>On Monday 26 May 2003 21:38, Jorge "Llamb�as" wrote:
>li'o
> > Even so, there were some gaps left. For example, one that came up
> > recently on the list:
> >
> > tirna sance
> > viska jvinu
> > sumne panci
> > pencu tengu
> > ????? vrusi
>
>Another one: bratu, snime, *sicpi. carvi doesn't mean "is rain"; it refers to
>the falling of any kind of precipitation.
no idea what sicpi means, but liquid rain is simply djacu carvi (which
we've also used for "shower" around here since the earliest days.
>Also, the set of bird gismu is decidedly lopsided: two Anseriformes, two
>Galliformes, and none of any other order including the most speciose by far,
>Passeriformes.
The birds that were added were primarily because they were major food birds
or they were especially symbolic in many cultures.
I've said it before, Pierre, that Lojban was NOT trying to design to the
Linnean system, and indeed we explicitly chose not to - a good thing since
there were only 2 kingdoms when we started instead of the present 5 or 6;
we never had a gismu for algae, blue-green or otherwise. JCB always
intended Linnean classification to use fu'ivla, and indeed he had a formal
borrowing scheme to use in making Linnean names.
jjllambias2000
la lojbab cusku di'e
> > A few words like
> >mango, pitsa or taksi have a special status in that they are
international
> >_and_ are already gismu-form without any need of adaptation.
[...]
> > It is hard
> >to resist those, since they don't even need a dictionary
definition in order
> >to be understood.
>
> Which is a good reason to resist them. You understand the word
without a
> place structure, and you have a bunch on 1-place predicates - the
language
> isn't much of a predicate language if most relations are unary.
But none of those would be 1-placers. (And in any case, why would
exptal gismu differ from fu'ivla in this respect?) The place
structure of mango is trivially: "x1 is a mango of species x2".
Anyone with a minimum familiarity with the gi'uste can guess
that. The place structure for pitsa is not so immediately
self-evident, but I would bet everyone would come up with "x1
is a pizza with ingredients x2". Again some familiarity with
the gi'uste almost imposes that place structure. The place
structure for taksi is perhaps the least obvious of the three,
but almost certainly I would bet it has the passengers/cargo
in x2.
mu'o mi'e xorxes
------------------------ Yahoo! Groups Sponsor ---------------------~-->
Robin Lee Powell
> How do you express emotions with a language that is logical?
In the case of Lojban, very effectively. In fact, the amazing
facility of the emotive words is one of the reasons I stuck with the
language.
> How can we speak a logical language if we are not logical?
Well, I can't speak for you, but some of the people on the main list
are so logical it makes my head hurt. 8)
But, in fact, you shouldn't have to understand formal logic to speak
Lojban, at least at a concious level. Or at least, we think so.
That's a question that won't be answered until there are native
speakers.
> Emotions can not be expressed unambiguiously.
Sure they can!
"I love you" and "I hate you" are both *very* unambiguous.
But just because Lojban's underlying *grammer* is unambiguous
doesn't mean that you can't express complicated or strange or
meaningless things with it.
lo gerku cu zdani -- A dog is a house.
-Robin
--
Me: http://www.digitalkingdom.org/~rlpowell/ *** I'm a *male* Robin.
"but I'm not stupid and people are not stupid who think samely with me"
-- from an actual, real, non-spam mail sent to webm...@lojban.org
http://www.lojban.org/ *** .i cimo'o prali .ui
Robin
> On Sat, May 24, 2003 at 07:51:29AM -0400, Pierre Abbat wrote:
>
>>On Saturday 24 May 2003 06:50, Nick Nicholas wrote:
>>
>>>Your issue (on those particular words) is with Abbat, not with Kominek
>>>or Powell. Your solution is to propose a canonical word for parasite,
>>>not to vent revulsion. (And I remind you of our recent discussion on
>>>big and small tents.)
>>
>>And that word I did not invent, I only entered into the dictionary. It was
>>already on the Lojban wiki, along with {didni}, {nusna}, {sicpi}, {gumri},
>>and several others. The ones I invented are mostly chemical terms, such as
>>{benzo} and {zmase}, which are either common enough concepts or used in
>>enough compounds that I think there ought to be gismu for them.
>
> [...]
>
> Why?
>
> What is with this thought that gismu are somehow privledged brivla?
> This is the same thing that makes people assert that all cultures
> should have gismu, instead of some with gismu and some with lujvo.
>
> If you're talking about rafsi, go use zei. If you're talking about
> word length, many lujvo have only 2 syllables, and 3 is totally
> fine (hell "parasite" is 3 in english). What advantage could you
> possibly see for it being a gismu?
>
> I think this all rests on a fundamental misunderstanding of lojban
> word classes. People like to think about gismu, cmavo and lujvo.
> But it's actually brivla, cmavo and cmene. Gismu, lujvo and fu'ivla
> are just different types of brivla; none are more privledged than
> the others.
>
here.
The issue with creating gismu is less how useful that gismu would be
itself than how generative it is. "Parasite" might be a good candidate,
not because we use the word "parasite" a lot, but because it might help
create a lot of lujvo (for things like fleas, lice, tapeworms, ticks etc.).
--
"A Perl script is "correct" if it gets the job done before your boss
fires you."
- Larry Wall
Robin Turner
IDMYO
Bilkent Univeritesi
Ankara 06533
Turkey
Craig
>Lojban, at least at a concious level. Or at least, we think so.
>That's a question that won't be answered until there are native
>speakers.
Right. As in, I don't know jack about formal logic, but I *think* I can
speak Lojban anyway.
------------------------ Yahoo! Groups Sponsor ---------------------~-->
Get A Free Psychic Reading! Your Online Answer To Life's Important Questions.
http://us.click.yahoo.com/Lj3uPC/Me7FAA/uetFAA/GSaulB/TM
Robert LeChevalier
>And of course, the obligatory flame:
> > Message: 18
> > Date: Sat, 24 May 2003 00:32:13 -0400
> > From: Robert LeChevalier <loj...@lojban.org>
> > Subject: Re: emotions
> >
> > At 03:28 PM 5/23/03 -0700, Robin Lee Powell wrote:
> >>>> What's a dictionary for? :)
> >>>>
> >>>> http://www.lojban.org/jbovlaste/dict/parji
> >>>
> >>>
> >>> BTW, the inclusion of such things in jvovlaste means that I for
> >>> one have no particular interest in using the thing. It cannot be
> >>> a standard for the language until the byfy decides to weed out the
> >>> garbage.
> >>
> >> You've just rejected something on the order of 80 man-hours on the
> >> basis of a *single* *word*.
> >
> > It did unfortunately happen to be the first word I've ever looked up.
> > I
> > can't help it that I feel extreme revulsion when it seems like my 15
> > years
> > of fighting for a solid baseline, with clear delineation between valid
> > and
> > invalid usage according to the language prescription, is being
> > undermined.
>
>Your issue (on those particular words) is with Abbat, not with Kominek
>or Powell. Your solution is to propose a canonical word for parasite,
>not to vent revulsion. (And I remind you of our recent discussion on
>big and small tents.)
Please review the thread.
Pierre casually mentioned "selparji" as an alternative to a peculiarly
Lojbanic application of zdani.
Robin, who if anyone should know how to use jbovlaste asked "what is selparji".
[At this point I note that "selparji" is NOT defined in jbovlaste.]
Jorge then gave the URL for the dictionary record for parji.
I looked up that URL, knowing that parji is not on the gismu list, and did
not notice any bright red flags.
Clicking on the links on that record I find a page
"Record For Natural Language Word "parasite" In Language English"
which gives parji as the translation with no flags whatsoever.
I fired off a quick, late-night flame.
I went hunting, eventually found the search page, and it responds to parji
and parasite both with no indication that it is non-standard, (nor,
checking now, can I find the obvious way to get from the search listing to
the dictionary record).
I don't view these as things that Robin or Jay did wrong. I think it is
premature to dispense with the flat text dictionary file as the
quasi-official standard when things are obviously at such a preliminary
stage. And if "use the dictionary" means use jbovlaste rather than the
flat file, then that jump has already happened.
> >> Go stick your head in a pig.
> >
> > As a dictionary it is useless to me until the non-standard words are
> > excluded. That presumably will be done by the byfy.
>
>Since it's the only dictionary we really have outside the disparate
>wordlists,
The official raw dictionary file is what I have used for years, and I guess
I will continue to use.
>and it is not yet standardised anyway, your rejection is
>premature. It's certainly more useful than the vapourware dictionary
>that LLG never produced.
I find the raw dictionary file more useful, and given what has now come up,
more trustworthy.
>And any publication under bpfk auspices of a
>dictionary, generated by jbovlaste or not, will be reviewed by the
>bpfk. jbovlaste itself (which will continue to exist in the long term)
>has its own democratic process of review; this may or may not be a good
>or canonical thing, this may or may not fall under bpfk or LLG
>auspices, but I trust that the project administrators will issue all
>caveats and disclaimers necessary about non-canonical words (as long as
>the baseline holds, at least). Jay's involved in it, for heaven's sake.
And by raising the issue, presumably Jay can and will take action if he
agrees that disclaimers are needed.
> > Because the policy of
> > its use allows standard and non-standard Lojban to be entered as if
> > the two
> > were equal in value, I strongly question that policy.
>
>Exptals are clearly delimited.
Not from what I could tell.
> > It means that
> > someone looking up the word for keyword X may get an invalid answer,
> > and
> > non discriminating users (probably most people) will take that answer
> > as
> > gospel.
>
>When jbovlaste is closer to being a standard (and of course, jbovlaste
>is very much evolving right now in its content), it will be trivial for
>the programmers to add a little flag next to any expt'al word that
>comes up on lookup. If you think it should right now as a matter of
>priority, log a feature request.
I thought that Robin already said once that feature requests can go hang.
>Of course, I believe that anyone non-discriminating (that isn't going
>to bother to look up the Lojban gloss that comes up, e.g. to check its
>place structure) has no business speaking the language. (This will be
>viewed as exclusionary, I'm sure.)
Not sure how this is relevant.
lojbab
--
lojbab loj...@lojban.org
Bob LeChevalier, President, The Logical Language Group, Inc.
2904 Beau Lane, Fairfax VA 22031-1303 USA 703-385-0273
Artificial language Loglan/Lojban: http://www.lojban.org
------------------------ Yahoo! Groups Sponsor ---------------------~-->
Get A Free Psychic Reading! Your Online Answer To Life's Important Questions.
http://us.click.yahoo.com/Lj3uPC/Me7FAA/CNxFAA/GSaulB/TM
Robert LeChevalier
>la lojbab cusku di'e
> > >good canonical alternatives, as Nick said.
> >
> > Nora suggested that no noncanonical word should be allowed to be
>added to
> > the jbovlaste UNLESS a canonical (perhaps longer) alternative has
>already
> > been added.
>
>{parji} to be adopted then I'd be tempted to enter an unappealing
>canonical alternative so that it won't really compete with my
>proposal.
One would hope that people would abide by the spirit of the rule, and not
play such games.
lojbab
--
lojbab loj...@lojban.org
Bob LeChevalier, President, The Logical Language Group, Inc.
2904 Beau Lane, Fairfax VA 22031-1303 USA 703-385-0273
Artificial language Loglan/Lojban: http://www.lojban.org
------------------------ Yahoo! Groups Sponsor ---------------------~-->
Rob Speer
> At 12:48 PM 5/26/03 +0100, And Rosta wrote:
> >It's disyllabic. There is no shortage of bisyllabic cmavo space.
>
> "la'edi'u" being longer than one syllable, and I can imagine few things
> that need a shorter one.
Does it fit within the BPFK guidelines to make "lau" mean "la'edi'u"?
Sure, it's an arbitrary change that has no usage so far, but for one
thing it would be very useful, and for another thing "lau" has never had
any usage anyway, and I can't even imagine what it would be used for.
--
mu'o mi'e rab.spir
And Rosta
> At 12:48 PM 5/26/03 +0100, And Rosta wrote:
> >It's disyllabic. There is no shortage of bisyllabic cmavo space
>
> frequent and fundamental words of the language like the logical
> connectives?
Well, right or wrong, that's why no fuss is made about uselessish
disyllabic cmavo.
As for the big need for monosyllabic cmavo, many many people do care
about the syllable count of their utterances. Reasons would include
syllable count being a measure of effort, and being a measure of
comparative verbosity compared to other languages.
> No one has been hurt, so far as I know, by "le du'u"/"le se du'u" and
> "la'edi'u" being longer than one syllable, and I can imagine few things
> that need a shorter one
I am an atypical lojbanist, but I have been hugely bugged by "lo'edu'u"
and xorxes by "la'edi'u". They don't actively harm me, but it pains me
to use them, to the extent that it poisons my pleasure in using Lojban.
It wouldn't surprise me if future newcomers to Lojban who had the sort
of refined linguistic sensibilities one observes in the likes of
xorxes would be similarly pained.
Not that I don't think it is too late to change Lojban. Lojban is far
into the "take it ot leave it" stage, and can no longer be adapted to
suit the tastes or even the needs of its users.
> The argument for the lerfu-builders being monosyllables was that they never
> would stand alone, but would always be in a string of several cmavo, and we
> wanted to keep the string short. This is turn was motivated by the
> importance (and frequency) of acronyms in many languages, and the already
> difficult rendering of Chinese and other non-Roman characters in
> Lojban. Unicode has solved the latter, and we seem to be satisfied with
> other ways to use lerfu. (Don't know if gamma ray has a lujvo yet,
> though). So I could let those go (out of the language or longer cmavo -
> either way)
I know the argument. We've discussed it before and it is not really worth
rehashing, because the effort of arguing about lau/toi wouldn't really
justify the proportionally small benefits of reassigning them. But, to
remind you, the counterargument to your point is that the shortest
cmavo should be assigned to maximize the saving of syllables in actual
usage. The price of making foreign-alphabet acronyms easier is not
worth the cost of making much higher frequency strings longer.
--And.
Jorge Llamb�as
> [...]
> > > I think this all rests on a fundamental misunderstanding of lojban
> > > word classes. People like to think about gismu, cmavo and lujvo.
> > > But it's actually brivla, cmavo and cmene.
> >
> > Morphologically, the classes are gismu, cmavo, lujvo, fu'ivla and cmene.
>
> Right.
>
> > Syntactically, the classes are BRIVLA, CMENE, KOhA, A, BAI, TAhE, VAU,
> etc...
>
> I.e. brivla, cmavo and cmene, as I mentioned.
cmavo is not a syntactic class, it is a morphological class.
GOhA for example is syntactically much closer to BRIVLA than
to any other selma'o. gismu are priviledged morphologically,
not syntactically, over other brivla. In a similar way CV cmavo
are morphologically priviledged over CVV, and CV'V cmavo.
> [...]
> > > Gismu, lujvo and fu'ivla
> > > are just different types of brivla; none are more privledged than
> > > the others.
> >
> > If that were so, then why all the fuss when a new gismu is proposed,
> > but no fuss when a new lujvo or fu'ivla is proposed?
>
> Because the gismu list is frozen, and there's no reason to prefer
> a gismu rather than another brivla.
And yet people do seem to prefer gismu. They are used much more frequently
than other brivla.
> Having a good number of gismu
> is desirable, to get rafsi for forming lujvo, however if we decide
> we need more it needs to be because we need more for forming lujvo,
> not because someone thinks that gismu are privledged and that some
> concept (e.g., 'parasite') seems to be 'common' enough or 'important'
> enough to deserve it, or that all cultures should be given the
> supposedly privledged status.
Of course it won't be because of one person's decision. Only if people
use them will they be used.
> If you ask why it is frozen (as you no doubt will do); it would not
> be smart to allow open season on gismu: the knee-jerk creation of
> gismu for things which can be other brivla damage our ability to
> increase rafsi for lujvo in the future (and in the process betrays
> a lack of understanding about the purpose of the different types
> of brivla).
Hmm... {parji} doesn't seem to limit any increase in future rafsi
availability, because par, paj, pai, pa'i are all already taken.
In any case, the best way to oppose these words is to provide
good canonical alternatives, as Nick said. Then people will choose
what they prefer in their usage. Talking so much about {parji} only
reinforces it in people's minds, so it is not an efficient way to
oppose it. (A similar campaign against {gumri} in the past helped
a lot to fix that word in my mind, so that it is now effectively a
part of my active vocabulary, which I can't say is the case for all
official gismu.)
mu'o mi'e xorxes
Robin Lee Powell
> At 01:04 PM 5/23/03 -0700, Jorge "Llamb?as" wrote:
>
> >--- Robin Lee Powell <rlpo...@digitalkingdom.org> wrote:
> >> >
> >>
> >> Fine fine. lo gerku cu dilnu, then.
> >>
> >> > (I would rather say {selparji}, of course.)
> >>
> >> Which is what?
> >
> >
> >http://www.lojban.org/jbovlaste/dict/parji
>
> Maybe he, like me, refuses to recognize words that are not Lojban.
>
> BTW, the inclusion of such things in jvovlaste means that I for
> one have no particular interest in using the thing. It cannot be
> a standard for the language until the byfy decides to weed out the
> garbage.
You've just rejected something on the order of 80 man-hours on the
basis of a *single* *word*.
Go stick your head in a pig.
-Robin
Nick Nicholas
those who wants that), it is not for la'edi'u that I would destine
it, but for the Unique/Kind gadri (jboske, Oct 2002 - Jan 2003).
Which is what Bob is hinting at.
Discussion on that is a long time off, though.
--
**** **** **** **** **** **** **** **** **** **** **** **** **** **** ****
* Dr Nick Nicholas, French & Italian Studies ni...@unimelb.edu.au *
Rm 637 Arts Centre, Melbourne University, Australia www.opoudjis.net
* "Eschewing obfuscatory verbosity of locutional rendering, the *
circumscriptional appelations are excised." --- W. Mann & S. Thompson,
* _Rhetorical Structure Theory: A Theory of Text Organisation_, 1987. *
**** **** **** **** **** **** **** **** **** **** **** **** **** **** ****
Philip Newton
> The place structure for taksi is perhaps the least obvious of the
> three, but almost certainly I would bet it has the passengers/cargo in
> x2.
Probably also a place for area of operation (most taxis I know operate
only in one city, or in a defined rural area).
Not so sure about driver or fare.
I've amended the Wiki page accordingly. (
http://www.lojban.org/wiki/index.php/experimental%20gismu )
mu'o mi'e .filip.
--
filip.niutyn. <pne...@gmx.de>
Robert LeChevalier
>As for usability in lujvo, one that I've often missed is something
>correponding to Esperanto -inda, "deserving of".
We considered "deserve" as a gismu, and decided that it was polysemous. It
either is the perfective of earned, or it is innately-obliged to
> > 5. Words made from one language, as parji was, should be fu'ivla.
>
>The "ji" part seems like it could be from Chinese, but I wouldn't
>really know. English-only would have given 'parsi', no?
I don't know.
> > Whether
> > people think there is a lot of meaning to the 6-language word-making, it
> > offers a couple of things: an objective way to decide the "best form",
> > dissociation of the word from the keyword in any single source
> language, so
> > that it is less likely to be encoded English (or whatever language).
>
>This goal was sort of defeated by the English keyword list. People
>learn the keywords to the point that they sometimes use the wrong
>place structure because of a misleading keyword.
I think that depends on whether they memorize the keywords before they
start using the language.. Certainly that error happens, but I think
people move beyond it quickly. The words I have the most problem with, are
the words that most resemble English.
> > This
> > is also why fu'ivla should be dispreferred when one can make a lujvo: a
> > lujvo has its own lojbanic meaning, whereas a fu'ivla starts with the
> > meaning in some other language and is not really lojbanic. lujvo-making
> > forces you to think about meaning, and jvajvo force you to think about
> > place structures (whether you choose to follow jvojva or not, considering
> > them is a good idea).
> > Nora looked up other experimental gismu in jbovlaste, and points
> > out that even more than parji, "mango" has no business as a gismu, and
> > benzo is almost as questionable.
>
>And that's about the whole list, isn't it? It seems like you are making
>the issue seem far bigger than it really is. Even if all the experimental
>gismu from the wiki were transferred to jbovlaste, I don't think they are
>more than 50, and almost all of them are cultural words. A few words like
>mango, pitsa or taksi have a special status in that they are international
>_and_ are already gismu-form without any need of adaptation.
But we HAVE resisted those up till now, and equally important, it was
decided in strong terms (and not just be lojbab) that we wanted to resist
these. My own reason for resisting the obvious gismu is because we
ridiculed JCB for allowing them willy-nilly, after having lots of people
complain. (The number of complaints about "billiards" and "olive" getting
gismu were highly prejudicial when we thought of "pitsa". The reaction was
more or less "ya gotta be kidding!"
> It is hard
>to resist those, since they don't even need a dictionary definition in order
>to be understood.
Which is a good reason to resist them. You understand the word without a
place structure, and you have a bunch on 1-place predicates - the language
isn't much of a predicate language if most relations are unary.
> > 6. (hard to explain) the list of existing gismu slants the choice of how
> > one makes and interprets lujvo. The semantics of the language is based on
> > what has gone before. Adding a new gismu to the coverage of semantic
> space
> > changes the semantic map, and thus could change the color of meaning of
> > other words in unexpected ways.
>
>Unexpected = bad ?
Yes.
Changes the semantics after usage has been established with the earlier
semantics. Forces relearning.
> > If it is "easy" to add words without thinking about meanings,
> > place structures, people will do so. I contend that, for gismu, this is
> > NOT a good thing.
>
>I agree. Not only for gismu, but also for lujvo and fu'ivla. They should
>not be added willy-nilly and without due consideration. Especially so in
>the case of gismu forms.
We agree for once %^) That is why I pushed for so long with the old
noralujv file that people keyword them, then define them with place
structures, then decide if they were worth keeping, all before adding them
to a dictionary.
All this takes time, which I've been short of while tackling other fires
(Nick keyworded most of the file on one of his visits here).
Those other fires are why I resent people saying I should be adding
bonafide words to compete with the illegitimate ones. People also want
book orders kept up, the accounting transitioned and turned over, and when
I have time for technical work, I'm 3 weeks behind in looking at byfy.
> > 8. Finally, before there was a byfy, adding gismu to the original baseline
> > list was consider fundamental enough that each one was put to a membership
> > vote (at LogFest). People were expected to make a case for their word and
> > submit it for consideration by the members, and to abide by the
> > result. Hence I abided by the elimination of gumri. The current
> method of
> > putting words out there, and having them see usage without the debate,
> > without the research, without the discussion, and without abiding by what
> > was decided in the past, is disparaging of stability, tradition, and the
> > opinions of members who put time and effort into the language in the past.
>
>The members will have to realize at some point that the language will
>belong more and more to the users than to the members.
Certainly. What was once done by the members should now be passed to byfy,
which for the most part consists of users. But the prescription shouldn't
go uncontrolled before it is fully documented (and we wanted the 5 year
period on top of that, though lujvo could be added).
Jorge Llamb�as
la pier cusku di'e
> As far as I can see, voting can't be used to weed out definitions - once
> someone has voted that {parji} means "parasite", the word is in the
> dictionary and the only way to get it out is to edit the TeX file.
The keyword "parasite" could be voted for another word, in which case
I think {parji} won't appear in the English to Lojban part. Perhaps
every new word could have an "exclude" alternative so that they
can be voted against as well. Then people could choose whether they
want to print the dictionary with or without those words that have been
voted out. Or, there could be an option to include or exclude certain
kind of words (exptal gismu, or compound cmavo, or whatever) when
generating a dictionary.
Robin Lee Powell
> At 03:28 PM 5/23/03 -0700, Robin Lee Powell wrote:
> >> >
> >> >http://www.lojban.org/jbovlaste/dict/parji
> >>
> >> Maybe he, like me, refuses to recognize words that are not
> >> Lojban.
> >>
> >> BTW, the inclusion of such things in jvovlaste means that I for
> >> one have no particular interest in using the thing. It cannot
> >> be a standard for the language until the byfy decides to weed
> >> out the garbage.
> >
> >You've just rejected something on the order of 80 man-hours on
> >the basis of a *single* *word*.
>
> up.
As xod once said, "And you admit this publically?"
> I can't help it that I feel extreme revulsion when it seems like
> my 15 years of fighting for a solid baseline, with clear
> delineation between valid and invalid usage according to the
> language prescription, is being undermined.
Did you even go to the link? Right at the top it says "experimental
gismu".
> >Go stick your head in a pig.
>
> I am not rejecting the man-hours. I'm sure you and Jay have done
> a Very Good Thing.
>
> As a dictionary it is useless to me until the non-standard words
> are excluded.
See above.
Jay and I have discussed the experimental issue, btw, and we've
agreed that the Right Way to handle experimental words is to allow
people to enter them all they want, but disallow them from appearing
in the best guesses listing (which, by extension, stops them from
appearing in the dictionary).
This change will be implemented the next time I get a chance to work
on jbovlaste.
Nick Nicholas
> Message: 18
> Date: Sat, 24 May 2003 00:32:13 -0400
> From: Robert LeChevalier <loj...@lojban.org>
> Subject: Re: emotions
>
> At 03:28 PM 5/23/03 -0700, Robin Lee Powell wrote:
>>>> What's a dictionary for? :)
>>>>
>>>> http://www.lojban.org/jbovlaste/dict/parji
>>>
>>> Maybe he, like me, refuses to recognize words that are not Lojban.
>>>
>>> BTW, the inclusion of such things in jvovlaste means that I for
>>> one have no particular interest in using the thing. It cannot be
>>> a standard for the language until the byfy decides to weed out the
>>> garbage.
>>
>> You've just rejected something on the order of 80 man-hours on the
>> basis of a *single* *word*.
>
> It did unfortunately happen to be the first word I've ever looked up.
> I
> can't help it that I feel extreme revulsion when it seems like my 15
> years
> of fighting for a solid baseline, with clear delineation between valid
> and
> invalid usage according to the language prescription, is being
> undermined.
Your issue (on those particular words) is with Abbat, not with Kominek
or Powell. Your solution is to propose a canonical word for parasite,
not to vent revulsion. (And I remind you of our recent discussion on
big and small tents.)
>> Go stick your head in a pig.
>
> As a dictionary it is useless to me until the non-standard words are
> excluded. That presumably will be done by the byfy.
Since it's the only dictionary we really have outside the disparate
wordlists, and it is not yet standardised anyway, your rejection is
premature. It's certainly more useful than the vapourware dictionary
that LLG never produced. And any publication under bpfk auspices of a
dictionary, generated by jbovlaste or not, will be reviewed by the
bpfk. jbovlaste itself (which will continue to exist in the long term)
has its own democratic process of review; this may or may not be a good
or canonical thing, this may or may not fall under bpfk or LLG
auspices, but I trust that the project administrators will issue all
caveats and disclaimers necessary about non-canonical words (as long as
the baseline holds, at least). Jay's involved in it, for heaven's sake.
> Because the policy of
> its use allows standard and non-standard Lojban to be entered as if
> the two
> were equal in value, I strongly question that policy.
Exptals are clearly delimited.
> It means that
> someone looking up the word for keyword X may get an invalid answer,
> and
> non discriminating users (probably most people) will take that answer
> as
> gospel.
When jbovlaste is closer to being a standard (and of course, jbovlaste
is very much evolving right now in its content), it will be trivial for
the programmers to add a little flag next to any expt'al word that
comes up on lookup. If you think it should right now as a matter of
priority, log a feature request.
Of course, I believe that anyone non-discriminating (that isn't going
to bother to look up the Lojban gloss that comes up, e.g. to check its
place structure) has no business speaking the language. (This will be
viewed as exclusionary, I'm sure.)
> I can't support a policy of "usage will decide" along with a
> policy that promotes non-baseline solutions as being equal to
> baseline-compliant solutions.
Well, you know what I think of your "usage will decide", and how it
inherently undermines any baseline...
**** **** **** **** **** **** **** **** **** **** **** **** **** ****
****
* Dr Nick Nicholas, French & Italian, University of Melbourne,
Australia *
ni...@unimelb.edu.au http://www.opoudjis.net
Robert LeChevalier
>--- Robin Lee Powell <rlpo...@digitalkingdom.org> wrote:
> > > > lo gerku cu zdani -- A dog is a house.
> > >
> > > lo gerku cu zdani lo risna curnu - A dog houses heartworms.
> >
> > Fine fine. lo gerku cu dilnu, then.
> >
> > > (I would rather say {selparji}, of course.)
> >
> > Which is what?
>
>What's a dictionary for? :)
>
>http://www.lojban.org/jbovlaste/dict/parji
Maybe he, like me, refuses to recognize words that are not Lojban.
BTW, the inclusion of such things in jvovlaste means that I for one have no
particular interest in using the thing. It cannot be a standard for the
language until the byfy decides to weed out the garbage.
lojbab
Jordan DeLong
> la djorden cusku di'e
> > > > word classes. People like to think about gismu, cmavo and lujvo.
> > > > But it's actually brivla, cmavo and cmene.
> > >
> > > Morphologically, the classes are gismu, cmavo, lujvo, fu'ivla and cmene.
> >
> > Right.
> >
> > > Syntactically, the classes are BRIVLA, CMENE, KOhA, A, BAI, TAhE, VAU,
> > etc...
> >
> > I.e. brivla, cmavo and cmene, as I mentioned.
>
> cmavo is not a syntactic class, it is a morphological class.
It is also a class of syntactic classes. (And as you hopefully can
tell, that is to what I was refering).
> GOhA for example is syntactically much closer to BRIVLA than
> to any other selma'o. gismu are priviledged morphologically,
> not syntactically, over other brivla. In a similar way CV cmavo
> are morphologically priviledged over CVV, and CV'V cmavo.
Sure, but that's not relevant to the issue of this brivla privledge
concept.
> > > > Gismu, lujvo and fu'ivla
> > > > are just different types of brivla; none are more privledged than
> > > > the others.
> > >
> > > If that were so, then why all the fuss when a new gismu is proposed,
> > > but no fuss when a new lujvo or fu'ivla is proposed?
> >
> > Because the gismu list is frozen, and there's no reason to prefer
> > a gismu rather than another brivla.
>
> And yet people do seem to prefer gismu. They are used much more frequently
> than other brivla.
Frequency of use is only because they are the only words with set
standards about use.
Furthremore, whether people prefer gismu does not prove anything
about whether they are privledged. People unknowingly (and in your
case sometimes knowingly) misuse lojban features all the time
(especially gadri (in all cases; but of course this is mostly the
fault of the lack of clarity in the language prescription on the
subject)).
[...]
> > If you ask why it is frozen (as you no doubt will do); it would not
> > be smart to allow open season on gismu: the knee-jerk creation of
> > gismu for things which can be other brivla damage our ability to
> > increase rafsi for lujvo in the future (and in the process betrays
> > a lack of understanding about the purpose of the different types
> > of brivla).
>
> Hmm... {parji} doesn't seem to limit any increase in future rafsi
> availability, because par, paj, pai, pa'i are all already taken.
[...]
It takes the "parj" rafsi.
--
Jordan DeLong - frac...@allusion.net
lu zo'o loi censa bakni cu terzba le zaltapla poi xagrai li'u
sei la mark. tuen. cusku
Pierre Abbat
> or Powell. Your solution is to propose a canonical word for parasite,
> not to vent revulsion. (And I remind you of our recent discussion on
> big and small tents.)
And that word I did not invent, I only entered into the dictionary. It was
already on the Lojban wiki, along with {didni}, {nusna}, {sicpi}, {gumri},
and several others. The ones I invented are mostly chemical terms, such as
{benzo} and {zmase}, which are either common enough concepts or used in
enough compounds that I think there ought to be gismu for them.
> Since it's the only dictionary we really have outside the disparate
> wordlists, and it is not yet standardised anyway, your rejection is
> premature. It's certainly more useful than the vapourware dictionary
> that LLG never produced. And any publication under bpfk auspices of a
> dictionary, generated by jbovlaste or not, will be reviewed by the
> bpfk. jbovlaste itself (which will continue to exist in the long term)
> has its own democratic process of review; this may or may not be a good
> or canonical thing, this may or may not fall under bpfk or LLG
> auspices, but I trust that the project administrators will issue all
> caveats and disclaimers necessary about non-canonical words (as long as
> the baseline holds, at least). Jay's involved in it, for heaven's sake.
This includes adding a note to lujvo and type-3 fu'ivla made from experimental
gismu (which jbovlaste does not note as such) saying that they are
experimental. I'm going to enter some now as examples.
phma
--
.i toljundi do .ibabo mi'afra tu'a do
.ibabo damba do .ibabo do jinga
.icu'u la ma'atman.
Pierre Abbat
> The issue with creating gismu is less how useful that gismu would be
> itself than how generative it is. "Parasite" might be a good candidate,
> not because we use the word "parasite" a lot, but because it might help
> create a lot of lujvo (for things like fleas, lice, tapeworms, ticks etc.).
Or different kinds of parasitizing: nerparji, barparji, mrori'aparji,
cinvi'uparji, and even cinga'iparji (Wolbachia).
Robin Lee Powell
> Please review the thread.
>
> Pierre casually mentioned "selparji" as an alternative to a
> peculiarly Lojbanic application of zdani.
>
> Robin, who if anyone should know how to use jbovlaste asked "what
> is selparji".
I did that specifically to point out that expecting people to know
experimental gismu is dumb. I could have used jbovlaste, I suppose,
but i didn't actually care.
> I looked up that URL, knowing that parji is not on the gismu list,
> and did not notice any bright red flags.
Dictionary record
valsi parji
type experimental gismu
creator phma
time entered Sat Mar 22 01:47:46 2003
How much more clear would you like us to be?
If you have a specific way in which you would like experimental
words to be 'red flagged', submit a request.
Robert LeChevalier
>On Saturday 24 May 2003 11:22, Jordan DeLong wrote:
> > I think this all rests on a fundamental misunderstanding of lojban
> > word classes. People like to think about gismu, cmavo and lujvo.
> > are just different types of brivla; none are more privledged than
> > the others.
>
>have rafsi; some gismu have short rafsi. So {malgaci zei smani} cannot be
>shortened, but {glauka zei cnebo} can be shortened to {glaukyne'o}, and
>{xamgu zei zmadu} can be shortened to {xagmau}.
Why is it important that infrequently used jargon words have very short forms?
It was part of the Loglan design strategy that infrequently used words
would tend to be longer than frequently used words, based on Zipf's Law.
Pierre Abbat
>
> What is with this thought that gismu are somehow privledged brivla?
> This is the same thing that makes people assert that all cultures
> should have gismu, instead of some with gismu and some with lujvo.
>
> If you're talking about rafsi, go use zei. If you're talking about
> word length, many lujvo have only 2 syllables, and 3 is totally
> fine (hell "parasite" is 3 in english). What advantage could you
> possibly see for it being a gismu?
I repeat, I did not invent {parji}. Go ask whoever did. I did invent {zmase},
because "-ase" is a common suffix.
> I think this all rests on a fundamental misunderstanding of lojban
> word classes. People like to think about gismu, cmavo and lujvo.
> But it's actually brivla, cmavo and cmene. Gismu, lujvo and fu'ivla
> are just different types of brivla; none are more privledged than
> the others.
Not true. Some fu'ivla have rafsi (proposed); all gismu except {brod(i,o,u)}
have rafsi; some gismu have short rafsi. So {malgaci zei smani} cannot be
shortened, but {glauka zei cnebo} can be shortened to {glaukyne'o}, and
{xamgu zei zmadu} can be shortened to {xagmau}.
phma
And Rosta
> Nora adds several arguments against ad hoc expansion of the gismu list by
> simple addition to jbovlaste, which I summarize
I'll add some counterarguments, for the record. My own view is that
experimental gismu are relatively trivial, with no strong arguments
pro or con. It is striking that nobody has proposed any really basic
experimental gismu (such as "deserve/worthy" (not "earn") or "intend").
All experimentals fall very much among the semantic marginalia. I
suppose the reason for that is that they are less intrusive.
> 1. every added gismu makes the goal of "learning the gismu list", a worthy
> goal for new Lojbanists, that much harder
Hard to see why it is such a worthy goal. Many gismu don't see usage,
because the need for them so seldom arises. A learner is better off
learning the most basic and frequent gismu, & then moving on to
other things.
Furthermore, I imagine that a learner is going to encounter gumri
more often than stero. So the grounds for learning the latter before
the former are decidedly ideological rather than utilitarian.
> 2. every added gismu makes the goal of learning rafsi (or deducing their
> meaning) that much harder. Assume that parji is added even with no rafsi
> assigned. Because it is there, then when you see rafsi paj, par, pai, or
> pa'i, or even pra, then this is one more gismu that they MIGHT be, and
> hence a little harder to learn
Maybe true. I have never exerted myself to learn rafsi other than useful
CCVs, so I wouldn't know.
> 3. all of the gismu added, whether people agree they should be or not, went
> through a certain amount of debate before we even made a gismu for
> them. The sheer necessity of looking up a word in 6 languages means that
> we had to consider the meaning carefully, so we'd know what to look up, and
> there were at least three of us involved in looking up words, so we
> therefore always debated (and Tommy and I had MANY long debates, since he
> was a gismu minimalist - as few as possible)
But the debates are closed, and resulted in an inventory of gismu that
many people find uncompelling in many respects. Hence those debates
do not necessarily command much respect -- if they'd produced excellent
results, or if they were ongoing, it would be a different story.
> 4. Once we got past the basic start of analyzing, weeding, and redoing the
> TLI Loglan list words, words were added only with a careful consideration
> of a)semantic completeness (e.g. of sets of food-grains), b) usability in
> lujvo to cover semantic space. New words should have to be justified in
> terms of necessity AS GISMU
1. It is hard to believe that many gismu could not satisfactorily have
been rendered by lujvo or fu'ivla. So the specialness of gismuhood is
not apparent from inspection of the gimste.
2. For some people the biunique association between 5-letter words and
a particular closed class of words is not especially compelling, &
certainly not fundamental to the functioning of the language.
> 7. Without disparaging the contributions of new people to the language,
> there is a tendency of many new people to, early in their Lojbanic career,
> say "it would have been better to do it 'this way'" without fully
> understanding the reasons why it was done 'the other way', so they advocate
> for change without learning the language as it is. Without baseline
> controls, the momentum of LOTS of usage, and a dictionary with words of all
> varieties so that people can find most of the words they want without
> inventing them, coining new gismu for every concept they want to say, is
> natural. I myself am guilty of this, with my favorite "pitsa", but I would
> never argue for adding it to the gismu list because I know better (and I
> don't really care to make more gismu for pepperoni, sausage, peppers, ham,
> and pineapple %^)
I agree that to achieve stability, Lojban has to take a very firm line
against innovation, to avoid opening the floodgates, the slippery slope,
etc. I also agree that the sooner Lojban moves from a design to an
actively used medium of communication, the sooner newbies will stop
saying "it would be better to do it this way".
> If it is "easy" to add words without thinking about meanings,
> place structures, people will do so. I contend that, for gismu, this is
> NOT a good thing
It's not a good thing for any brivla.
> 8. Finally, before there was a byfy, adding gismu to the original baseline
> list was consider fundamental enough that each one was put to a membership
> vote (at LogFest). People were expected to make a case for their word and
> submit it for consideration by the members, and to abide by the
> result. Hence I abided by the elimination of gumri. The current method of
> putting words out there, and having them see usage without the debate,
> without the research, without the discussion, and without abiding by what
> was decided in the past, is disparaging of stability, tradition, and the
> opinions of members who put time and effort into the language in the past
Quite possibly. But see response to (3).
Nick Nicholas
class. Per definitionem. If it's bullshit to introduce "kir" and "git"
as gender-neutral pronouns in English (and it is), then it is just as
bullshit (per definitionem) to introduce new gismu into Lojban.
Innovating gismu is not held to the same standard as innovating lujvo,
and if they are, then we abandon the distinction between the two, which
is foundational to Lojban morphology. So your parallel with lexical
innovation of nouns or verbs in English is false. Any additions to the
gismu set must be gradual and well-considered: *obviously* gismu are
privileged, purely in terms of morphology and productivity, and I don't
get why anyone would argue otherwise.
And with all respect to jbovlaste, I don't see anyone expecting that
the entirety of its contents will ever constitute a baseline
(especially if it is to be issued in the next three years). Gismu and
camvo are baselined. I really don't see the point in a large set of
lujvo or fu'ivla being in a baseline. In a reference dictionary, sure.
But I don't set on such a dictionary the canonical value I place on the
fundamental building blocks of the language.
(This counters Bob's vision of the dictionary. But then, Bob has his
vision, and I have mine. And I continue to think his vision naive.)
**** **** **** **** **** **** **** **** **** **** **** **** **** ****
****
* Dr Nick Nicholas, French & Italian, University of Melbourne,
Australia *
ni...@unimelb.edu.au http://www.opoudjis.net
* "Eschewing obfuscatory verbosity of locutional rendering, the
*
circumscriptional appelations are excised." --- W. Mann & S. Thompson,
* _Rhetorical Structure Theory: A Theory of Text Organisation_, 1987.
*
**** **** **** **** **** **** **** **** **** **** **** **** **** ****
****
Robin Lee Powell
> At 12:14 PM 5/27/03 -0700, Robin Lee Powell wrote:
> >On Sat, May 24, 2003 at 12:32:13AM -0400, Robert LeChevalier wrote:
> >> I can't help it that I feel extreme revulsion when it seems
> >> like my 15 years of fighting for a solid baseline, with clear
> >> delineation between valid and invalid usage according to the
> >> language prescription, is being undermined.
> >
> >"experimental gismu".
>
> "invalid according to the prescription", since other experiments
> are condoned in the baseline.
>
> And that line is only visible if you call up the word directly.
> It doesn't show up when you use search of click on a reference to
> the word elsewhere.
That's been fixed. Rather drastically, in fact.
Jorge Llamb�as
la djordan cusku di'e
> What is with this thought that gismu are somehow privledged brivla?
They are somehow privilidged. It is rather obvious that they are,
from the point of view of the morphology.
> This is the same thing that makes people assert that all cultures
> should have gismu, instead of some with gismu and some with lujvo.
Yes.
> If you're talking about rafsi, go use zei.
But zei-words don't quite feel like words. I'm not even sure they
are defined as a single word in terms of lojban. Can they be used as
the word delimiter for zoi, for example? Can they be quoted with
zo? (Maybe the answer is yes, but it is not at all intuitive.)
> If you're talking about
> word length, many lujvo have only 2 syllables, and 3 is totally
> fine (hell "parasite" is 3 in english). What advantage could you
> possibly see for it being a gismu?
Well, if it's a gismu you don't have to worry about possible
misinterpretations of grouping, for example. If you form a lujvo
with a gismu by adding another gismu, there is no possible
grouping ambiguity. If you do it with a two-part lujvo, then
you have to use brackets to make sure. Not that this is the
main advantage of gismu, but it is one.
> I think this all rests on a fundamental misunderstanding of lojban
> word classes. People like to think about gismu, cmavo and lujvo.
> But it's actually brivla, cmavo and cmene.
Morphologically, the classes are gismu, cmavo, lujvo, fu'ivla and cmene.
Syntactically, the classes are BRIVLA, CMENE, KOhA, A, BAI, TAhE, VAU, etc...
(The Book weirdly compares Lojban morphological classes with English
syntactic classes.)
> Gismu, lujvo and fu'ivla
> are just different types of brivla; none are more privledged than
> the others.
If that were so, then why all the fuss when a new gismu is proposed,
but no fuss when a new lujvo or fu'ivla is proposed?
mu'o mi'e xorxes
Robert LeChevalier
>Okay. Nick & Lojban say that any deviation from the prescription
>is bad per se. I say that deviation is bad in proportion to the
>adverse consequences it has (in terms of causing us to not speak
>the same language). There's not much more to be said about that.
>
>(I agree with xorxes that ironically the discussedness of
>experimentals means that they stick in the memory better than
>many official words. For instance, I have no idea what the
>official word for 'fungus' is,
mledi, which is memorable for the initial, and quite mnemonic with "mold".
--
lojbab loj...@lojban.org
Bob LeChevalier, President, The Logical Language Group, Inc.
2904 Beau Lane, Fairfax VA 22031-1303 USA 703-385-0273
Artificial language Loglan/Lojban: http://www.lojban.org
------------------------ Yahoo! Groups Sponsor ---------------------~-->
Robert LeChevalier
> > >
> > >http://www.lojban.org/jbovlaste/dict/parji
> >
> > Maybe he, like me, refuses to recognize words that are not Lojban.
> >
> > BTW, the inclusion of such things in jvovlaste means that I for
> > one have no particular interest in using the thing. It cannot be
> > a standard for the language until the byfy decides to weed out the
> > garbage.
>
>You've just rejected something on the order of 80 man-hours on the
>basis of a *single* *word*.
It did unfortunately happen to be the first word I've ever looked up. I
can't help it that I feel extreme revulsion when it seems like my 15 years
of fighting for a solid baseline, with clear delineation between valid and
invalid usage according to the language prescription, is being undermined.
>Go stick your head in a pig.
I am not rejecting the man-hours. I'm sure you and Jay have done a Very
Good Thing.
As a dictionary it is useless to me until the non-standard words are
excluded. That presumably will be done by the byfy. Because the policy of
its use allows standard and non-standard Lojban to be entered as if the two
were equal in value, I strongly question that policy. It means that
someone looking up the word for keyword X may get an invalid answer, and
non discriminating users (probably most people) will take that answer as
gospel. I can't support a policy of "usage will decide" along with a
policy that promotes non-baseline solutions as being equal to
baseline-compliant solutions.
lojbab
Jorge Llamb�as
la lojbab cusku di'e
> > > >In any case, the best way to oppose these words is to provide
> > > >good canonical alternatives, as Nick said.
> > >
> > > Nora suggested that no noncanonical word should be allowed to be
> >added to
> > > the jbovlaste UNLESS a canonical (perhaps longer) alternative has
> >already
> > > been added.
> >
> >That seems to put the onus on the wrong party. If I wanted
> >{parji} to be adopted then I'd be tempted to enter an unappealing
> >canonical alternative so that it won't really compete with my
> >proposal.
>
> One would hope that people would abide by the spirit of the rule, and not
> play such games.
Even if they disagree with the spirit of the rule? Why would
you expect people to act against their convictions?
Robert LeChevalier
>On Sat, May 24, 2003 at 12:32:13AM -0400, Robert LeChevalier wrote:
> > At 03:28 PM 5/23/03 -0700, Robin Lee Powell wrote:
> > >> >What's a dictionary for? :)
> > >> >
> > >> >http://www.lojban.org/jbovlaste/dict/parji
> > >>
> > >> Maybe he, like me, refuses to recognize words that are not
> > >> Lojban.
> > >>
> > >> BTW, the inclusion of such things in jvovlaste means that I for
> > >> one have no particular interest in using the thing. It cannot
> > >> be a standard for the language until the byfy decides to weed
> > >> out the garbage.
> > >
> > >You've just rejected something on the order of 80 man-hours on
> > >the basis of a *single* *word*.
> >
> > It did unfortunately happen to be the first word I've ever looked
> > up.
>
>As xod once said, "And you admit this publically?"
Yes. And I should feel no embarrassment about it when everyone is so
insistent that it isn't "official". After all, we know that I have more
than enough to do on official stuff.
> > I can't help it that I feel extreme revulsion when it seems like
> > my 15 years of fighting for a solid baseline, with clear
> > delineation between valid and invalid usage according to the
> > language prescription, is being undermined.
>
>Did you even go to the link? Right at the top it says "experimental
>gismu".
Yes. I did not see it at first, and "experimental" does not mean "invalid
according to the prescription", since other experiments are condoned in the
baseline.
And that line is only visible if you call up the word directly. It doesn't
show up when you use search of click on a reference to the word elsewhere.
>Jay and I have discussed the experimental issue, btw, and we've
>agreed that the Right Way to handle experimental words is to allow
>people to enter them all they want, but disallow them from appearing
>in the best guesses listing (which, by extension, stops them from
>appearing in the dictionary).
>
>This change will be implemented the next time I get a chance to work
>on jbovlaste.
If you can have the "experimental gismu" show up on the search and
cross-reference per above that would also help.
robin
>On Sat, May 24, 2003 at 01:40:47PM -0700, Jorge Llambías wrote:
>
>
>>la djordan cusku di'e
>>
>>
>>>What is with this thought that gismu are somehow privledged brivla?
>>>
>>>
>>They are somehow privilidged. It is rather obvious that they are,
>>from the point of view of the morphology.
>>
>>
>
>They are not. They have different behaviors (or they wouldn't be
>different classes). But none is privledged.
>
>If I need a brivla that expresses some predicate relation, it does
>not matter what type of word it is.
>
>[...]
>
>
>>>If you're talking about rafsi, go use zei.
>>>
>>>
>>But zei-words don't quite feel like words.
>>
>>
>[...]
>
>Well, that's probably because they aren't words. But so what?
>
>
>
>>I'm not even sure they
>>are defined as a single word in terms of lojban. Can they be used as
>>the word delimiter for zoi, for example? Can they be quoted with
>>zo? (Maybe the answer is yes, but it is not at all intuitive.)
>>
>>
>
>ZEI is parsed in the tanru rules, so no.
>
>[...]
>
>
>>>I think this all rests on a fundamental misunderstanding of lojban
>>>word classes. People like to think about gismu, cmavo and lujvo.
>>>But it's actually brivla, cmavo and cmene.
>>>
>>>
>>Morphologically, the classes are gismu, cmavo, lujvo, fu'ivla and cmene.
>>
>>
>
>Right.
>
>
>
>>Syntactically, the classes are BRIVLA, CMENE, KOhA, A, BAI, TAhE, VAU, etc...
>>
>>
>
>I.e. brivla, cmavo and cmene, as I mentioned.
>
>[...]
>
>
>>>Gismu, lujvo and fu'ivla
>>>are just different types of brivla; none are more privledged than
>>>the others.
>>>
>>>
>>If that were so, then why all the fuss when a new gismu is proposed,
>>but no fuss when a new lujvo or fu'ivla is proposed?
>>
>>
>
>Because the gismu list is frozen, and there's no reason to prefer
>a gismu rather than another brivla. Having a good number of gismu
>is desirable, to get rafsi for forming lujvo, however if we decide
>we need more it needs to be because we need more for forming lujvo,
>not because someone thinks that gismu are privledged and that some
>concept (e.g., 'parasite') seems to be 'common' enough or 'important'
>enough to deserve it, or that all cultures should be given the
>supposedly privledged status.
>
>If you ask why it is frozen (as you no doubt will do); it would not
>be smart to allow open season on gismu: the knee-jerk creation of
>gismu for things which can be other brivla damage our ability to
>increase rafsi for lujvo in the future (and in the process betrays
>a lack of understanding about the purpose of the different types
>of brivla).
>
>
>
Exactly. As for the old cultural gismu debate, the whole way cultural
gismu were created was wrong, IMO, but we won't solve it by making
more. I'd like to scrap all cultural gismu and use fu'ivla, but it's
too late for that.
--
"Hackers appear to wear black more because it goes with everything and hides dirt than because they want to look like goths."
- Eric Raymond
Robin Turner
IDMYO
Bilkent Univeritesi
Ankara 06533
Turkey
Jorge Llamb�as
la lojbab cusku di'e
> 1. every added gismu makes the goal of "learning the gismu list", a worthy
> goal for new Lojbanists, that much harder
One can always settle for "learning the official gismu list". I would
not recomend that as a goal, though. It is much easier, I think, to learn
them as needed through usage.
> 2. every added gismu makes the goal of learning rafsi (or deducing their
> meaning) that much harder. Assume that parji is added even with no rafsi
> assigned. Because it is there, then when you see rafsi paj, par, pai, or
> pa'i, or even pra, then this is one more gismu that they MIGHT be, and
> hence a little harder to learn.
This is probably true, if you use that method of learning rafsi, but you
are the only person I have heard saying they learn them that way. For me
rafsi are the hardest things to learn in the language. Except for the
few that are used very often, I have not learned very many yet, even after
several years of not insignificant usage. For most gismu, I can't tell
what their rafsi is, or even whether they have short rafsi.
> 3. all of the gismu added, whether people agree they should be or not, went
> through a certain amount of debate before we even made a gismu for
> them. The sheer necessity of looking up a word in 6 languages means that
> we had to consider the meaning carefully, so we'd know what to look up, and
> there were at least three of us involved in looking up words, so we
> therefore always debated (and Tommy and I had MANY long debates, since he
> was a gismu minimalist - as few as possible).
Well, it seems that lack of debate won't be a problem in this case.
> 4. Once we got past the basic start of analyzing, weeding, and redoing the
> TLI Loglan list words, words were added only with a careful consideration
> of a)semantic completeness (e.g. of sets of food-grains), b) usability in
> lujvo to cover semantic space. New words should have to be justified in
> terms of necessity AS GISMU.
Even so, there were some gaps left. For example, one that came up
recently on the list:
tirna sance
viska jvinu
sumne panci
pencu tengu
????? vrusi
As for usability in lujvo, one that I've often missed is something
correponding to Esperanto -inda, "deserving of".
> 5. Words made from one language, as parji was, should be fu'ivla.
The "ji" part seems like it could be from Chinese, but I wouldn't
really know. English-only would have given 'parsi', no?
> Whether
> people think there is a lot of meaning to the 6-language word-making, it
> offers a couple of things: an objective way to decide the "best form",
> dissociation of the word from the keyword in any single source language, so
> that it is less likely to be encoded English (or whatever language).
This goal was sort of defeated by the English keyword list. People
learn the keywords to the point that they sometimes use the wrong
place structure because of a misleading keyword.
> This
> is also why fu'ivla should be dispreferred when one can make a lujvo: a
> lujvo has its own lojbanic meaning, whereas a fu'ivla starts with the
> meaning in some other language and is not really lojbanic. lujvo-making
> forces you to think about meaning, and jvajvo force you to think about
> place structures (whether you choose to follow jvojva or not, considering
> them is a good idea).
> Nora looked up other experimental gismu in jbovlaste, and points
> out that even more than parji, "mango" has no business as a gismu, and
> benzo is almost as questionable.
And that's about the whole list, isn't it? It seems like you are making
the issue seem far bigger than it really is. Even if all the experimental
gismu from the wiki were transferred to jbovlaste, I don't think they are
more than 50, and almost all of them are cultural words. A few words like
mango, pitsa or taksi have a special status in that they are international
_and_ are already gismu-form without any need of adaptation. It is hard
to resist those, since they don't even need a dictionary definition in order
to be understood. I'm sure those will end up as part of the language in
any case.
> 6. (hard to explain) the list of existing gismu slants the choice of how
> one makes and interprets lujvo. The semantics of the language is based on
> what has gone before. Adding a new gismu to the coverage of semantic space
> changes the semantic map, and thus could change the color of meaning of
> other words in unexpected ways.
Unexpected = bad ?
> 7. Without disparaging the contributions of new people to the language,
> there is a tendency of many new people to, early in their Lojbanic career,
> say "it would have been better to do it 'this way'" without fully
> understanding the reasons why it was done 'the other way', so they advocate
> for change without learning the language as it is. Without baseline
> controls, the momentum of LOTS of usage, and a dictionary with words of all
> varieties so that people can find most of the words they want without
> inventing them, coining new gismu for every concept they want to say, is
> natural. I myself am guilty of this, with my favorite "pitsa", but I would
> never argue for adding it to the gismu list because I know better (and I
> don't really care to make more gismu for pepperoni, sausage, peppers, ham,
> and pineapple %^)
> If it is "easy" to add words without thinking about meanings,
> place structures, people will do so. I contend that, for gismu, this is
> NOT a good thing.
I agree. Not only for gismu, but also for lujvo and fu'ivla. They should
not be added willy-nilly and without due consideration. Especially so in
the case of gismu forms.
> 8. Finally, before there was a byfy, adding gismu to the original baseline
> list was consider fundamental enough that each one was put to a membership
> vote (at LogFest). People were expected to make a case for their word and
> submit it for consideration by the members, and to abide by the
> result. Hence I abided by the elimination of gumri. The current method of
> putting words out there, and having them see usage without the debate,
> without the research, without the discussion, and without abiding by what
> was decided in the past, is disparaging of stability, tradition, and the
> opinions of members who put time and effort into the language in the past.
The members will have to realize at some point that the language will
belong more and more to the users than to the members.
mu'o mi'e xorxes
Robert LeChevalier
>On Mon, May 26, 2003 at 10:46:12AM -0400, Robert LeChevalier wrote:
> > At 12:48 PM 5/26/03 +0100, And Rosta wrote:
> > >It's disyllabic. There is no shortage of bisyllabic cmavo space.
> >
> > No one has been hurt, so far as I know, by "le du'u"/"le se du'u" and
> > "la'edi'u" being longer than one syllable, and I can imagine few things
> > that need a shorter one.
>
>Does it fit within the BPFK guidelines to make "lau" mean "la'edi'u"?
Not directly.
It fits within the BPFK guidelines to define lau. Then it fits to decide
that lau is unnecessary and eliminate it. Then it fits to decide to assign
it to something else (and I would not support la'edi'u as a high priority
choice, since the ONLY justification I can imagine for that assignment is
Zipf, to which is counterbalanced the structural clarity of la'eXXXX (which
need not be "di'u", so it also loses the parallelism with la'ede'u, which I
doubt would warrant a parallel monosyllable). More likely, lau would be
used somehow in resolving the gadri semantics problem which seems to bother
more people on purely technical grounds.)
Each of those steps must be justified - you can't eliminate something as
non-useful if you don't know what it means and why it was put in the
language in the first place. And the byfy's job isn't to improve the
language, but to complete the prescription, fixing any bugs in that
prescription that have arisen through usage (which means places where usage
has shown a problem, or where usage has been consistent but at odds with
the intended prescription and thus needs adjudication).
>Sure, it's an arbitrary change that has no usage so far,
And arbitrary change is definitely NOT within the byfy charter.
> but for one
>thing it would be very useful, and for another thing "lau" has never had
>any usage anyway, and I can't even imagine what it would be used for.
It was the original solution to the problem that there are more lerfu
symbols and alphabetic characters in the languages of the world, than there
are lojban lerfu, in effect anticipating the Unicode approach of two byte
lerfu, but without nearly as much thought put into it. There ARE better
solutions now that Unicode exists, and usage has chosen other methods for
expressing common lerfu, so the byfy has reason to consider its elimination
as being un-useful (and maybe un-usable BECAUSE it isn't consistent with
the new international standard).
Robert LeChevalier
>To And's puzzlement, all I have to respond with is: gismu are a closed
[Agreed with Nick's response]
>And with all respect to jbovlaste, I don't see anyone expecting that
>the entirety of its contents will ever constitute a baseline
>(especially if it is to be issued in the next three years).
It is not that it will constitute a baseline, but rather that it will be
USED as the reference standard for the language "because it is there". (As
you noted, the official dictionary is not there - even though it isn't
"vaporware", but merely draft - but then so is jbovlaste.)
Thus I note that jbofi'e seems to have displaced the official parser, and
the E-BNF which was the unofficial grammar standard displaced the YACC in
usage to the point that it was added to CLL and hence made part of the
baseline, and now people find "errors" in the language based on jbofi'e and
the E-BNF.
Because users can freely add to it WITHOUT any vetting on their additions,
people will add to jbovlaste. And because it is readily available for word
lookup, those who live on the net will USE jbovlaste. Thus it stands to
overwhelm any and all baseline standards by its mere existence, if it
permits non-baseline-standard entry/display on equal terms with
baseline-standard work.
I don't see any easy solution - I realize the amount of work that went into
jbovlaste, and the apparent fact that people like it a lot better than they
liked editing "noralujv" and the draft dictionary files (which I admit we
never provided a good editing/commenting mechanism for).
(And I realize your criticism that I should be adding standard words to
compete with the nonstandard ones. But people want me to serve my last
time as President productively, and you also want me to respond to
supplication on byfy and maybe do more besides, so official obligations
have taken precedent over word building, even though I admit the latter is
more fun for me than the other stuff).
>Gismu and
>camvo are baselined. I really don't see the point in a large set of
>lujvo or fu'ivla being in a baseline. In a reference dictionary, sure.
Which jbovlaste, unedited, will be.
>But I don't set on such a dictionary the canonical value I place on the
>fundamental building blocks of the language.
As canon, I agree. But I look at what the community USES, and they are
USING jbovlaste.
>(This counters Bob's vision of the dictionary.
Not sure what you think my vision is. But your vision as presented is not
that unlike mine. I just recognize what the user-community and the world
think about dictionaries supersedes what the linguists think about them.
> But then, Bob has his vision, and I have mine. And I continue to think
> his vision naive.)
My argument on this issue is based on the fact that I KNOW my ideal for the
dictionary is naive when jbovlaste is a reality. And I don't have time (or
web-programming knowhow) to come up with my own alternative, so I am stuck
with complaining, and hoping I can stimulate others into acting as needed.
lojbab
--
lojbab loj...@lojban.org
Bob LeChevalier, President, The Logical Language Group, Inc.
2904 Beau Lane, Fairfax VA 22031-1303 USA 703-385-0273
Artificial language Loglan/Lojban: http://www.lojban.org
------------------------ Yahoo! Groups Sponsor ---------------------~-->
Robert LeChevalier
>On Sat, May 24, 2003 at 01:16:29PM -0400, Robert LeChevalier wrote:
> > Why is it important that infrequently used jargon words have very short
> > forms?
>
Content words. Obviously in Lojban, cmavo will be shorter than any content
word, and people don't normally consider cmavo to be jargon. (I'm
surprised that there has seldom been criticism of re'a, which I myself
thought was a questionable addition, deferring to people with more
mathematical orientations.)
I will note in passing that lau, tei, and foi were added to the language
before Unicode, and before Gary Burgess and I came up with word+bu as
alternate approaches to letteral words. The problem remains the same, and
it is important that there be short ways to say acronyms even when they
don't use Lojban-alphabet lerfu, but we now have more solutions than we
need. I have no doubt that the byfy can come up with the appropriate ones
to delete, and this is one area where I will likely support a baseline
change, since we have a combination of non-usage, good alternate solutions,
and an important problem that key people feel needs the cmavo (the gadri
issue).
Jorge Llamb�as
la lojbab cusku di'e
> >But none of those would be 1-placers. (And in any case, why would
> >exptal gismu differ from fu'ivla in this respect?)
>
> Precisely the question. fu'ivla are second rate not-quite-lojban words,
> usually concocted on the fly without a lot of thought because they were
> needed to talk about things in an area of specialty. gismu are the
> fundamental roots of the language, supposedly the most carefully reviewed
> area of the language. I want to keep the distinction quite clear. lujvo
> are greatly preferred to fu'ivla even at the expense of length because they
> are internal to the language, and are capable of being analyzed for a
> predicted place structure, even if usage does not necessarily abide by
> jvojva.
The fu'ivla that go into the dictionary should not be concocted
on the fly, they should be given as much consideration as the
other words in the dictionary. And they are fully lojban words,
despite their origin as borrowings. Other than that, I agree with
the above, which does not require gismu to be a definitively closed
class.
Robert LeChevalier
>It's disyllabic. There is no shortage of bisyllabic cmavo space.
What is this big need for monosyllabic cmavo, except perhaps for the REALLY
frequent and fundamental words of the language like the logical connectives?
No one has been hurt, so far as I know, by "le du'u"/"le se du'u" and
"la'edi'u" being longer than one syllable, and I can imagine few things
that need a shorter one.
The argument for the lerfu-builders being monosyllables was that they never
would stand alone, but would always be in a string of several cmavo, and we
wanted to keep the string short. This is turn was motivated by the
importance (and frequency) of acronyms in many languages, and the already
difficult rendering of Chinese and other non-Roman characters in
Lojban. Unicode has solved the latter, and we seem to be satisfied with
other ways to use lerfu. (Don't know if gamma ray has a lujvo yet,
though). So I could let those go (out of the language or longer cmavo -
either way).
I am also favoring that the byfy start using the xVV cmavo, as the reason
for reserving them seems to be fulfilled by the current effort.
Jordan DeLong
> On Saturday 24 May 2003 11:22, Jordan DeLong wrote:
> > Why?
> >
> > What is with this thought that gismu are somehow privledged brivla?
> > This is the same thing that makes people assert that all cultures
> > should have gismu, instead of some with gismu and some with lujvo.
> >
> > If you're talking about rafsi, go use zei. If you're talking about
> > word length, many lujvo have only 2 syllables, and 3 is totally
> > fine (hell "parasite" is 3 in english). What advantage could you
> > possibly see for it being a gismu?
>
> I repeat, I did not invent {parji}. Go ask whoever did. I did invent {zmase},
> because "-ase" is a common suffix.
But you endorsed the concept that some ideas are "common enough"
that there "ought to be gismu for them".
> > I think this all rests on a fundamental misunderstanding of lojban
> > word classes. People like to think about gismu, cmavo and lujvo.
> > But it's actually brivla, cmavo and cmene. Gismu, lujvo and fu'ivla
> > are just different types of brivla; none are more privledged than
> > the others.
>
> Not true. Some fu'ivla have rafsi (proposed); all gismu except {brod(i,o,u)}
> have rafsi; some gismu have short rafsi. So {malgaci zei smani} cannot be
> shortened, but {glauka zei cnebo} can be shortened to {glaukyne'o}, and
> {xamgu zei zmadu} can be shortened to {xagmau}.
What's not true? I agree with everything you just said regarding
rafsi.
I already (preemptively) addressed the rafsi = privledge argument;
the existence of zei destroys it.
Jordan DeLong
> > What is with this thought that gismu are somehow privledged brivla?
>
> They are somehow privilidged. It is rather obvious that they are,
> from the point of view of the morphology.
They are not. They have different behaviors (or they wouldn't be
different classes). But none is privledged.
If I need a brivla that expresses some predicate relation, it does
not matter what type of word it is.
[...]
> > If you're talking about rafsi, go use zei.
>
> But zei-words don't quite feel like words.
[...]
Well, that's probably because they aren't words. But so what?
> I'm not even sure they
> are defined as a single word in terms of lojban. Can they be used as
> the word delimiter for zoi, for example? Can they be quoted with
> zo? (Maybe the answer is yes, but it is not at all intuitive.)
ZEI is parsed in the tanru rules, so no.
[...]
> > I think this all rests on a fundamental misunderstanding of lojban
> > word classes. People like to think about gismu, cmavo and lujvo.
> > But it's actually brivla, cmavo and cmene.
>
> Morphologically, the classes are gismu, cmavo, lujvo, fu'ivla and cmene.
Right.
> Syntactically, the classes are BRIVLA, CMENE, KOhA, A, BAI, TAhE, VAU, etc...
I.e. brivla, cmavo and cmene, as I mentioned.
[...]
> > Gismu, lujvo and fu'ivla
> > are just different types of brivla; none are more privledged than
> > the others.
>
> If that were so, then why all the fuss when a new gismu is proposed,
> but no fuss when a new lujvo or fu'ivla is proposed?
Because the gismu list is frozen, and there's no reason to prefer
a gismu rather than another brivla. Having a good number of gismu
is desirable, to get rafsi for forming lujvo, however if we decide
we need more it needs to be because we need more for forming lujvo,
not because someone thinks that gismu are privledged and that some
concept (e.g., 'parasite') seems to be 'common' enough or 'important'
enough to deserve it, or that all cultures should be given the
supposedly privledged status.
If you ask why it is frozen (as you no doubt will do); it would not
be smart to allow open season on gismu: the knee-jerk creation of
gismu for things which can be other brivla damage our ability to
increase rafsi for lujvo in the future (and in the process betrays
a lack of understanding about the purpose of the different types
of brivla).
--
jjllambias2000
la lojbab cusku di'e
> >In any case, the best way to oppose these words is to provide
> >good canonical alternatives, as Nick said.
>
> Nora suggested that no noncanonical word should be allowed to be
added to
> the jbovlaste UNLESS a canonical (perhaps longer) alternative has
already
> been added.
That seems to put the onus on the wrong party. If I wanted
{parji} to be adopted then I'd be tempted to enter an unappealing
canonical alternative so that it won't really compete with my
proposal.
Adding {brodr-} to a gismu form will always result in a valid
fu'ivla. So an automatic way of canonizing experimental gismu
would be to prefix them with brodr-, as in {brodrparji}.
mu'o mi'e xorxes
------------------------ Yahoo! Groups Sponsor ---------------------~-->
Robin Lee Powell
> On Friday 23 May 2003 14:22, Robin Lee Powell wrote:
> li'o
> > But just because Lojban's underlying *grammer* is unambiguous
> > doesn't mean that you can't express complicated or strange or
> > meaningless things with it.
> >
> > lo gerku cu zdani -- A dog is a house.
>
> lo gerku cu zdani lo risna curnu - A dog houses heartworms.
Fine fine. lo gerku cu dilnu, then.
> (I would rather say {selparji}, of course.)
Which is what?
-Robin
Pierre Abbat
li'o
> There is a voting procedure already in place to do the weeding out.
> We just need more people to actually use it, write some definitions
> and/or do some voting.
As far as I can see, voting can't be used to weed out definitions - once
someone has voted that {parji} means "parasite", the word is in the
dictionary and the only way to get it out is to edit the TeX file. What
voting does is pick which of several alternative definitions to put in the
dictionary. E.g., both {samcrkasava} and {mandioka} refer to the same plant;
one can vote whether "cassava" or "manioc" is to be translated by one or the
other.
phma
And Rosta
> > cmavo is not a syntactic class, it is a morphological class
>
> It is also a class of syntactic classes.
It doesn't strike me as a natural class of syntactic classes, since
the only thing that those syntactic classes have in common is that
they all have the morphological property of being expressed by
cmavo.
--And.
Jorge Llamb�as
la lojbab cusku di'e
>(I'm
> surprised that there has seldom been criticism of re'a, which I myself
> thought was a questionable addition, deferring to people with more
> mathematical orientations.)
Probably most people, even among those who use the language often, have
no idea what re'a means. I had to look it up to see what it was. I'm
sure it would have received lots of criticism if it had occupied
valuable monosyllabic space. As a CV'V word it is not so harmful.
mu'o mi'e xorxes
Pierre Abbat
li'o
> But just because Lojban's underlying *grammer* is unambiguous
> doesn't mean that you can't express complicated or strange or
> meaningless things with it.
>
> lo gerku cu zdani -- A dog is a house.
lo gerku cu zdani lo risna curnu - A dog houses heartworms. (I would rather
say {selparji}, of course.)
mu'omi'e pier.
Pierre Abbat
li'o
> recently on the list:
>
> tirna sance
> viska jvinu
> sumne panci
> pencu tengu
> ????? vrusi
Another one: bratu, snime, *sicpi. carvi doesn't mean "is rain"; it refers to
the falling of any kind of precipitation.
Also, the set of bird gismu is decidedly lopsided: two Anseriformes, two
Galliformes, and none of any other order including the most speciose by far,
Passeriformes.
"mango", as I've pointed out before, is almost the same in five of the six
languages (it's called variants of "ampah" in Indic) and several of the next
six. It is also one of the most popular fruits in the world - the most
popular, I've read.
phma
Craig
>> >mango, pitsa or taksi have a special status in that they are
>international
>> >_and_ are already gismu-form without any need of adaptation.
>> > It is hard
>> >to resist those, since they don't even need a dictionary
>definition in order
>> >to be understood.
>>
>without a
>> place structure, and you have a bunch on 1-place predicates - the
>language
>> isn't much of a predicate language if most relations are unary.
>But none of those would be 1-placers. (And in any case, why would
>exptal gismu differ from fu'ivla in this respect?) The place
>structure of mango is trivially: "x1 is a mango of species x2".
>Anyone with a minimum familiarity with the gi'uste can guess
>that. The place structure for pitsa is not so immediately
>self-evident, but I would bet everyone would come up with "x1
>is a pizza with ingredients x2". Again some familiarity with
>the gi'uste almost imposes that place structure. The place
>structure for taksi is perhaps the least obvious of the three,
>but almost certainly I would bet it has the passengers/cargo
>in x2.
Laying aside my somewhat contorted views on exptal gismu, I wouldn't blink
at ko'a ko'e ko'i taksi. The x3 is clearly the destination, IMO.
Robin Lee Powell
> I thought that Robin already said once that feature requests can
> go hang.
No, Robin said that he is very busy and isn't likely to get around
to them any time soon.
You are more than welcome to make the request. E-mail to
rt-...@lojban.org
Rob Speer
> Why is it important that infrequently used jargon words have very short
> forms?
(glances at lau, tei, and foi)
--
mu'o mi'e rab.spir
Robert LeChevalier
>la lojbab cusku di'e
>without a
> > place structure, and you have a bunch on 1-place predicates - the
>language
> > isn't much of a predicate language if most relations are unary.
>
>But none of those would be 1-placers. (And in any case, why would
>exptal gismu differ from fu'ivla in this respect?)
Precisely the question. fu'ivla are second rate not-quite-lojban words,
usually concocted on the fly without a lot of thought because they were
needed to talk about things in an area of specialty. gismu are the
fundamental roots of the language, supposedly the most carefully reviewed
area of the language. I want to keep the distinction quite clear. lujvo
are greatly preferred to fu'ivla even at the expense of length because they
are internal to the language, and are capable of being analyzed for a
predicted place structure, even if usage does not necessarily abide by jvojva.
lojbab
--
lojbab loj...@lojban.org
Bob LeChevalier, President, The Logical Language Group, Inc.
2904 Beau Lane, Fairfax VA 22031-1303 USA 703-385-0273
Artificial language Loglan/Lojban: http://www.lojban.org
------------------------ Yahoo! Groups Sponsor ---------------------~-->
Jorge Llamb�as
la djorden cusku di'e
> > > > > I think this all rests on a fundamental misunderstanding of lojban
> > > > > word classes. People like to think about gismu, cmavo and lujvo.
> > > > > But it's actually brivla, cmavo and cmene.
> > cmavo is not a syntactic class, it is a morphological class.
>
> It is also a class of syntactic classes. (And as you hopefully can
> tell, that is to what I was refering).
Then you were wrong. The preference for gismu is for morphological, not
for syntactical reasons.
> > Hmm... {parji} doesn't seem to limit any increase in future rafsi
> > availability, because par, paj, pai, pa'i are all already taken.
> [...]
>
> It takes the "parj" rafsi.
But there is no shortage of those.
sshiskom
> 5. Words made from one language, as parji was, should be fu'ivla.
xorxes wrote:
> The "ji" part seems like it could be from Chinese, but I wouldn't
> really know. English-only would have given 'parsi', no?
You are right. As I said, "parasitism" is "gisaeng" in Sino-Korean,
and "ji4 sheng1" in Chinese. You note the similarity. It is obvious
to me that "ji" in "parji" comes from Chinese.
mi'e sanxiyn.
Robert LeChevalier
>Nick to Lojbab:
> > > I can't support a policy of "usage will decide" along with a
> > > policy that promotes non-baseline solutions as being equal to
> > > baseline-compliant solutions
> >
> > inherently undermines any baseline..
I first note in answer to Nick that the concept of "usage will decide" was
intended to apply to selection among multiple options built into the
language. This concept has unfortunately been extended to inventing
non-baseline solutions and hoping that they catch on enough to force a
change in the baseline.
>Wholly without heat, I observe that I don't understand the
>either of you's rationale. We need a baseline (frozen or not) so
>that we are all (a) speaking the same language & (b) truly speaking
>a language (& not making it up as we go along, pidginwise). But the
>evidence of natural language is that vocabulary innovation is
>the most innocuous and useful variety of usage deciding.
Yes. And I WANT vocabulary innovation. But I'd like it to proceed along
the lines of the design-intent, with a priority on tanru/lujvo for common
words, Type III fu'ivla for uncommon words, with frequent usage of the
latter justifying the coining of a Type IV shorter version, and rafsi
fu'ivla for those concepts that people think deserve a short form from the
start because they want to sue them in lujvo.
I also think that, while nonce usage will "invent" lots of words, serious
vocabulary development of a non-nonce type should involve checking how the
concept is represented in multiple languages including some non-European ones.
>It's
>innocuous because there is scant room for misunderstanding (because
>if a word is novel, there is no competing prior definition for
>it) -- the risk of misunderstanding is much more pernicious
>than the risk of not understanding at all, and useful for obvious
>reasons.
For me, the risk is not one of misunderstanding, but of the establishment
of bad habits early in the vocabulary development that persist, and in fact
create problems later.
JCB had the habit in the 70s of coining a new gismu that seemed useful at
the moment, but which really had no justification as a root word in the
language. Thus TLI Loglan has gismu for "olive", "football" (ambiguously
never defined as to whether he meant this to be soccer or American rugby),
sodium, beefsteak, and billiards. All of these should have been lujvo, or
fu'ivla if there was no obvious lujvo. But in the 70s, there was no
distinct fu'ivla form, so borrowings were in the morphological form of
gismu and lujvo.
I'm fighting the recurrence of this bad habit.
>Furthermore, the value of a baseline (as opposed to
>frozenness) is that it provides some sort of shared explicit
>reference standard. If jbovlaste were used as such, then an
>experimental gismu listed in jbovlaste would be more part of
>the baseline than a lujvo that is not listed.
Indeed. And it is obvious that jbovlaste will become such a reference
standard whether it is approved as a "baseline" or not.
>And I don't see any problem with that. Anybody encountering {parji} will
>either
>know it or not understand it and look it up in jbovlaste, just
>as with any other gismu.
Except that there aren't supposed to be any experimental gismu. Indeed
some of them, including the infamous gumri, were explicitly voted out of
the language by the membership (and I lost on that vote, so that in
fighting gumri, I fight for a principle over my own preferences for the
language).
lojbab
--
lojbab loj...@lojban.org
Bob LeChevalier, President, The Logical Language Group, Inc.
2904 Beau Lane, Fairfax VA 22031-1303 USA 703-385-0273
Artificial language Loglan/Lojban: http://www.lojban.org
------------------------ Yahoo! Groups Sponsor ---------------------~-->
Jordan DeLong
> On Saturday 24 May 2003 06:50, Nick Nicholas wrote:
> > Your issue (on those particular words) is with Abbat, not with Kominek
> > or Powell. Your solution is to propose a canonical word for parasite,
> > not to vent revulsion. (And I remind you of our recent discussion on
> > big and small tents.)
>
> And that word I did not invent, I only entered into the dictionary. It was
> already on the Lojban wiki, along with {didni}, {nusna}, {sicpi}, {gumri},
> and several others. The ones I invented are mostly chemical terms, such as
> {benzo} and {zmase}, which are either common enough concepts or used in
> enough compounds that I think there ought to be gismu for them.
[...]
Why?
What is with this thought that gismu are somehow privledged brivla?
This is the same thing that makes people assert that all cultures
should have gismu, instead of some with gismu and some with lujvo.
If you're talking about rafsi, go use zei. If you're talking about
word length, many lujvo have only 2 syllables, and 3 is totally
fine (hell "parasite" is 3 in english). What advantage could you
possibly see for it being a gismu?
I think this all rests on a fundamental misunderstanding of lojban
word classes. People like to think about gismu, cmavo and lujvo.
But it's actually brivla, cmavo and cmene. Gismu, lujvo and fu'ivla
are just different types of brivla; none are more privledged than
the others.
--
Jorge Llamb�as
la lojbab cusku di'e
> BTW, the inclusion of such things in jvovlaste means that I for one have no
> particular interest in using the thing.
That would be your loss, as that is where we are concentrating all
definitions of lujvo. Also, notice that experimental gismu and cmavo
have their own separate sections, so you can use it without ever
seeing the parts you don't like.
> It cannot be a standard for the
> language until the byfy decides to weed out the garbage.
There is a voting procedure already in place to do the weeding out.
We just need more people to actually use it, write some definitions
and/or do some voting.
mu'o mi'e xorxes
And Rosta
> > I can't support a policy of "usage will decide" along with a
> > policy that promotes non-baseline solutions as being equal to
> > baseline-compliant solutions
>
> Well, you know what I think of your "usage will decide", and how it
> inherently undermines any baseline..
Wholly without heat, I observe that I don't understand the
either of you's rationale. We need a baseline (frozen or not) so
that we are all (a) speaking the same language & (b) truly speaking
a language (& not making it up as we go along, pidginwise). But the
evidence of natural language is that vocabulary innovation is
the most innocuous and useful variety of usage deciding. It's
innocuous because there is scant room for misunderstanding (because
if a word is novel, there is no competing prior definition for
it) -- the risk of misunderstanding is much more pernicious
than the risk of not understanding at all, and useful for obvious
reasons. Furthermore, the value of a baseline (as opposed to
frozenness) is that it provides some sort of shared explicit
reference standard. If jbovlaste were used as such, then an
experimental gismu listed in jbovlaste would be more part of
the baseline than a lujvo that is not listed. And I don't see
any problem with that. Anybody encountering {parji} will either
know it or not understand it and look it up in jbovlaste, just
as with any other gismu.
--And.