WikiDiscuss

WikiDiscuss


BPFK Section: lerfu Shifts

posts: 14214
Use this thread to discuss the BPFK Section: lerfu Shifts page.
posts: 14214

Please note that there is no point seperating out a "proposals" section, or numbering the proposals, or anything like that: your section stands or falls as a whole.

The goal of a comissioner is to continue refining their proposal until consensus-1 is reached.

-Robin

posts: 953

Rob, please take this in good faith. I am very happy with the thoroughness on your research, but I have to disagree with several on your conclusion, viz. the deprecation(sp?) of unused cmavo.

1. There is no tradition in the Loglan project for declaring parts of the language as "deprecated". A construct is either permissible, or ungrammatical. The LLG should decide what is and what is not part of the language, but should not decide how people should use it.

2. The non-usage of a construct is not a sufficient reason to change the status quo. I believe we have been through a similar discussion, either on the PHPBB or on Jboske.

posts: 152

I was somewhat afraid of that.

I suppose I will go ahead and define the alphabet shifts. But I believe Lojban will eventually need a concept of deprecation.

Human languages have words that become obsolete due to lack of usage. Computer languages have deprecation so that clunky features of the language don't have to be maintained forever. As halfway in between, I think Lojban needs a similar concept. When an advanced-level Lojban textbook appears, it shouldn't have to describe things like lerfu shifts, and when a Lojban-using AI appears, it shouldn't have to understand them.

So what I mean by "deprecated" is this: the construct is permissible, but future instructional materials in Lojban should feel free to not teach it. Its meaning is left up to the interpretation of whatever listener might understand it. I don't see this as prescribing how people should use the language; I see this as getting rid of baggage so we can get on with accurately specifying the parts of the language that are useful.

posts: 208

> arj:
> Rob, please take this in good faith. I am very happy with the thoroughness on your research, but I have to disagree with several on your conclusion, viz. the deprecation(sp?) of unused cmavo.

And I disagree with your disagreement.

> 1. There is no tradition in the Loglan project for declaring parts of the language as "deprecated". A construct is either permissible, or ungrammatical. The LLG should decide what is and what is not part of the language, but should not decide how people should use it.

Deprecation *IS* deciding "what is and is not part of the language". Furthermore, it is making that decision based, in this case, on total and complete lack of usage. I don't see a problem as long as (and this is key) there are other ways to acheive the same effect.

> 2. The non-usage of a construct is not a sufficient reason to change the status quo. I believe we have been through a similar discussion, either on the PHPBB or on Jboske.

Then please point me to it, because it seems perfectly sufficient to me, if there's another way to accomplish the same effect (as there is in this case, AFAIK).

-Robin

posts: 14214

On Thu, Nov 13, 2003 at 11:41:08AM -0800, wikidiscuss@lojban.org wrote:
>
> Re:On declaring a set of cmavo "deprecated"
>
> > arj:
> > Rob, please take this in good faith. I am very happy with the
> > thoroughness on your research, but I have to disagree with
> > several on your conclusion, viz. the deprecation(sp?) of unused
> > cmavo.
>
> And I disagree with your disagreement.

I should note, however, that my disagreement was made in total
absence of reading the proposal; it was disagreement in principle.

-Robin


posts: 953

> So what I mean by "deprecated" is this: the construct is permissible, but future instructional materials in Lojban should feel free to not teach it. Its meaning is left up to the interpretation of whatever listener might understand it. I don't see this as prescribing how people should use the language; I see this as getting rid of baggage so we can get on with accurately specifying the parts of the language that are useful.

Here is a proposal that I hope will be acceptable to both of us. The intent is to provide guidance to future (and present) writers of instructional materials, while not actually changing the language per se.

Given that Lojban does not seem to be intended for holding multilingual spelling bees, and learning material containing many unused cmavo with bizarre functions could confuse learners of the language, the BPFK does not recommend to include alphabet shifts (including zai as well) in learning materials intended even for advanced learners.


(I actually edited this into the page, being unaware of the proper procedures, but it is rolled back now. Sorry.)

posts: 1912


> ga'e is only used correctly in
> algebra, to mark variables named
> with capital letters. It occurs one lerfu from the end of a lerfu string, so
> that it seems to have the same effect as tau.

Notice that it is only used twice in the whole text, whereas the capital
letter variables appear a lot more times, so it is not quite the
same as tau.

> !! na'a (cancel shifts)
>
> This word is not used in the corpus.

In the above text, {nau} is used. Presumably that {nau} is {na'a}'s
ancestor.

> A letteral shift lasts until another shift of the same type replaces it, it
> is cancelled by na'a, or the end of the lerfu string is reached.
>
> (The sole usage of ga'e assumed that it would end at the end of a lerfu
> string, which seems reasonable.)

It may be reasonable, but the usage in question assumes the effect
lasts indefinitely, or at least for much longer than the lerfu string.
Otherwise, almost every variable there should be preceded by {ce'a fy ga'e}.

mu'o mi'e xorxes


__
Do you Yahoo!?
Protect your identity with Yahoo! Mail AddressGuard
http://antispam.yahoo.com/whatsnewfree


posts: 1912



> The page BPFK Section: lerfu Shifts was changed by rab.spir at
> Thu 13 of Nov, 2003 22:14 UTC

Unfortunately, it seems Arnt was editing it at the same time
and your changes were lost. I suggest only the shepherd touch
their page, and the others make their comments on the discuss
forum.

> !! na'a (cancel shifts)
>
> This word is not used in the corpus, though nau was used in the algebra
> text where na'a was probably intended.

That nau was changed to na'a so that nau could be given
a more useful meaning, even though it had seen at least one use
(which is more than can be said for some other cmavo) sets a
good precedent for changing lau, tau, tei, foi to
something less wasteful.

mu'o mi'e xorxes


__
Do you Yahoo!?
Protect your identity with Yahoo! Mail AddressGuard
http://antispam.yahoo.com/whatsnewfree


posts: 14214

> > !! na'a (cancel shifts)
> >
> > This word is not used in the corpus, though nau was used in
> > the algebra text where na'a was probably intended.
>
> That nau was changed to na'a so that nau could be
> given a more useful meaning, even though it had seen at least one
> use (which is more than can be said for some other cmavo) sets a
> good precedent for changing lau, tau, tei, foi to
> something less wasteful.

Do we have proof this occured? If so, did it occur after the
publication of the CLL?

-Robin


posts: 1912


> > > !! na'a (cancel shifts)
> > >
> > > This word is not used in the corpus, though nau was used in
> > > the algebra text where na'a was probably intended.
> >
> > That nau was changed to na'a so that nau could be
> > given a more useful meaning, even though it had seen at least one
> > use (which is more than can be said for some other cmavo) sets a
> > good precedent for changing lau, tau, tei, foi to
> > something less wasteful.
>
> Do we have proof this occured?

Shouldn't be hard to find.

> If so, did it occur after the
> publication of the CLL?

Obviously not. CLL {nau} has its current meaning, and
{na'a} appears in CLL.

mu'o mi'e xorxes


__
Do you Yahoo!?
Protect your identity with Yahoo! Mail AddressGuard
http://antispam.yahoo.com/whatsnewfree


posts: 953

Rob,

Our BPFK sections rely on each other, since we both have alphabet shifts. In Section: lerfu Forming Cmavo, I say that "letteral shifts last indefinitely or until superseded by another shift in the same category." Even if you insist on deprecating the alphabet cmavo &c. (which I can give you a number of good reasons not to), you should make a statement on the scope of these cmavo, since the CLL doesn't say so explicitly. You can use my result, or you can find out some of your own, in which case I'm asking you to send it, along with justificiations, to me.

posts: 152

On Wed, Jan 28, 2004 at 01:30:09PM -0800, wikidiscuss@lojban.org wrote:
>
> Re: BPFK Section: lerfu Shifts
>
> Rob,
>
> Our BPFK sections rely on each other, since we both have alphabet shifts. In Section: lerfu Forming Cmavo, I say that "letteral shifts last indefinitely or until superseded by another shift in the same category." Even if you insist on deprecating the alphabet cmavo &c. (which I can give you a number of good reasons not to), you should make a statement on the scope of these cmavo, since the CLL doesn't say so explicitly. You can use my result, or you can find out some of your own, in which case I'm asking you to send it, along with justificiations, to me.

I thought I accepted that on my page (which I'm going to get around to
finishing Real Soon Now). I used it for case and font shifts, for
example.
--
Rob Speer



posts: 152

Okay, I've updated my recommendation to include the new sections.

I've avoided the word "deprecated", instead using arj's wording that the words should not be included in future learning materials.

The scope of lerfu shifts should be clear, and I believe it is consistent with the proposal at BPFK Section: lerfu Forming cmavo.

People should vote on this now.

posts: 14214

On Thu, Jan 29, 2004 at 08:52:06PM -0800, wikidiscuss@lojban.org
wrote:
>
> Re: BPFK Section: lerfu Shifts
>
> Okay, I've updated my recommendation to include the new sections.
>
> I've avoided the word "deprecated", instead using arj's wording
> that the words should not be included in future learning
> materials.

Please make sure that it is absolutely clear whether your proposal
is that these cmavo be freed for future, different uses. You are
welcome to even state a proposed timeline for making them available
for re-use, if you wish.

-Robin

--
Me: http://www.digitalkingdom.org/~rlpowell/ *** I'm a *male* Robin.
"Constant neocortex override is the only thing that stops us all
from running out and eating all the cookies." — Eliezer Yudkowsky
http://www.lojban.org/ *** .i cimo'o prali .ui


posts: 953

> rlpowell:
> On Thu, Jan 29, 2004 at 08:52:06PM -0800, wikidiscuss@lojban.org
> wrote:
> >
> > Re: BPFK Section: lerfu Shifts
> >
> > Okay, I've updated my recommendation to include the new sections.
> >
> > I've avoided the word "deprecated", instead using arj's wording
> > that the words should not be included in future learning
> > materials.
>
> Please make sure that it is absolutely clear whether your proposal
> is that these cmavo be freed for future, different uses. You are
> welcome to even state a proposed timeline for making them available
> for re-use, if you wish.

Am I missing something here? Isn't the purpose of the BPFK project to make the last changes to Lojban, ever? If cmavo forms are to be relinquished sometime in the future, someone has to have the authority to make that change happen.

For the record, I can live with keeping the cmavo, and I can live with scrapping or reassigning the cmavo, but at the moment I'm not able to see any benefits of a solution in between.

posts: 14214

On Fri, Jan 30, 2004 at 03:30:56AM -0800, wikidiscuss@lojban.org wrote:
>
> Re: BPFK Section: lerfu Shifts
>
> > rlpowell:
> > On Thu, Jan 29, 2004 at 08:52:06PM -0800, wikidiscuss@lojban.org
> > wrote:
> > >
> > > Re: BPFK Section: lerfu Shifts
> > >
> > > Okay, I've updated my recommendation to include the new sections.
> > >
> > > I've avoided the word "deprecated", instead using arj's wording
> > > that the words should not be included in future learning
> > > materials.
> >
> > Please make sure that it is absolutely clear whether your proposal
> > is that these cmavo be freed for future, different uses. You are
> > welcome to even state a proposed timeline for making them available
> > for re-use, if you wish.
>
> Am I missing something here? Isn't the purpose of the BPFK project
> to make the last changes to Lojban, ever?

Not quite. It's a bit more complicated then that. See
http://lojban.org/llg/baseline.html

For example:

During the baseline freeze, LLG will not consider any change
proposals to the language, and will encourage anyone seeking to
have a proposal considered after the freeze ends, to formally
write their proposal up in Lojban. Any such proposals will be
collected in a controlled "wiki" format and accessible on the
lojban.org website, but will not be formally considered. If
errors in the baseline documents are detected during the
baseline period, they shall be similarly documented in Lojban,
and may be formally considered by the appropriate language
standard group of the time (which by default will be the rump
form of the commission established below).


> If cmavo forms are to be relinquished sometime in the future,
> someone has to have the authority to make that change happen.

Not so; the BPFK has the authority to make such a decision now.
There is nothing that says that all BPFK decisions must take effect
immediately.

> For the record, I can live with keeping the cmavo, and I can live
> with scrapping or reassigning the cmavo, but at the moment I'm not
> able to see any benefits of a solution in between.

Neither am I, particularily. I didn't say it was a *good* idea,
just that it is a possibility.

-Robin


posts: 14214

Question about:

"the BPFK does not recommend to include the alphabet shifts (to'a, ge'o, je'o, jo'o, lo'a, ru'o, zai) in learning materials intended even for advanced learners."

Are you sure you meant to have to'a be in that list? It's certainly not an alphabet shift.

-Robin

posts: 14214

Unfortunately, I can't let your proposal include words from another shepherd's section; that just opens up too large a can of worms.

You are welcome to convince Arnt to take a similar stance with repsect to zai as you have, but in the mean time please remove zai from your list of pseudo-deprecated cmavo.

-Robin, speaking as BPFK jatna.

posts: 152

to'a being in that list was a mistake. Thanks.

It's my opinion that 'zai' was assigned to the wrong section, but I'll just leave it alone.

posts: 14214

On Fri, Jan 30, 2004 at 11:00:45AM -0800, wikidiscuss@lojban.org
wrote:
>
> Re: BPFK Section: lerfu Shifts
>
> to'a being in that list was a mistake. Thanks.
>
> It's my opinion that 'zai' was assigned to the wrong section, but
> I'll just leave it alone.

It's done by selma'o. If Arnt wants to give you the whole of LAU,
I care very little.

-Robin


posts: 2388

I assum that {zai BY} defines a character set much as {ge'o} et al did, with the advantage of being openended about possibilities (Tengwar, for example, or Devanagari}, provided we agree on what letter stands for what alphabet. I suppose that, outside of specialists in languages or culture, no more than a dozen alphabets are actually used (math rarely gets beyond Greek, German, and Hebrew and it is about as character greedy as they come).

Does something deal with non-alphabetic symbols (implication arrows are my favorite) or are they dealt just with using {bu} and not being specific about what the resulting character looks like?
pc




posts: 1912


> Does something deal with non-alphabetic symbols (implication arrows are my
> favorite) or are they dealt just with using {bu} and not being specific about
> what the resulting character looks like?

{zai zoi gy non-alphabetic symbol gy bu zoi gy implication arrow gy bu}?

(I don't remember whether {zoi} acts before {bu}, I suppose it must.)

But I'd rather use the implication arrow itself for an implication arrow,
and something like {lo nibli sinxa} to talk about them.

mu'o mi'e xorxes


__
Do you Yahoo!?
Yahoo! SiteBuilder - Free web site building tool. Try it!
http://webhosting.yahoo.com/ps/sb/


posts: 953

> rlpowell:
> Unfortunately, I can't let your proposal include words from another shepherd's section; that just opens up too large a can of worms.
>
> You are welcome to convince Arnt to take a similar stance with repsect to zai as you have, but in the mean time please remove zai from your list of pseudo-deprecated cmavo.

I will add the note on zai to BPFK section: lerfu Forming cmavo.

-arj

posts: 2388

I hope not something with {zai zoi gy...}, at least something in lojban. I suppose I had something like {na.a bu} in mind (or {ganai bu} but placeable between sentences, not prefixed), but that is going to make problems for the symbols for OR, AND, IFF, and whatever you call the U function, since e.g. {abu} is already taken. Or is that where {zai} come in (but what selector gets dingbats)? {zai dy na.a bu} for => (and the like)?
All of this is for reading off wffs, the analog of spelling, I suppose. For talking about the character {lo nibli sinxa} (well, {nibli} is sorta wrong, but I get the idea) is also possible, and I suppose could be used for reading off well.
pc
xorxes wrote:


> Does something deal with non-alphabetic symbols (implication arrows are my
> favorite) or are they dealt just with using {bu} and not being specific about
> what the resulting character looks like?

{zai zoi gy non-alphabetic symbol gy bu zoi gy implication arrow gy bu}?

(I don't remember whether {zoi} acts before {bu}, I suppose it must.)

But I'd rather use the implication arrow itself for an implication arrow,
and something like {lo nibli sinxa} to talk about them.

mu'o mi'e xorxes


__
Do you Yahoo!?
Yahoo! SiteBuilder - Free web site building tool. Try it!
http://webhosting.yahoo.com/ps/sb/




posts: 1912


> I hope not something with {zai zoi gy...}, at least something in lojban. I
> suppose I had something like {na.a bu} in mind (or {ganai bu} but placeable
> between sentences, not prefixed),

You could do {na bu .a bu} or even {tei na bu a bu foi}.
Another possibility is {na zei .a bu}, using the lujvo {na zei .a}.

> but that is going to make problems for the
> symbols for OR, AND, IFF, and whatever you call the U function, since e.g.
> {abu} is already taken.

You could always base them on se.a, se.e, se.o (or na.onai) and .unai
So: {se bu .a bu}, {tei se bu .a bu foi}, or {se zei .a bu}.

> Or is that where {zai} come in (but what selector
> gets dingbats)? {zai dy na.a bu} for => (and the like)?

{dy} is for "dingbats"? Perhaps {zai norle'u bu}?

> All of this is for reading off wffs, the analog of spelling, I suppose. For
> talking about the character {lo nibli sinxa} (well, {nibli} is sorta wrong,
> but I get the idea) is also possible, and I suppose could be used for reading
> off well.
> pc

Presumably if the wff was part of a Lojban text it would use
the MEX module, in which case the implication would be some
operator and you would not have to mention the symbol by name.
But I don't know if MEX is really up to the task.

mu'o mi'e xorxes


__
Do you Yahoo!?
Yahoo! SiteBuilder - Free web site building tool. Try it!
http://webhosting.yahoo.com/ps/sb/


posts: 2388


Jorge LlambĂ­as <jjllambias2000@yahoo.com.ar> wrote:


> I hope not something with {zai zoi gy...}, at least something in lojban. I
> suppose I had something like {na.a bu} in mind (or {ganai bu} but placeable
> between sentences, not prefixed),

You could do {na bu .a bu} or even {tei na bu a bu foi}.


That would be (at best) "~v" — a compound of “literals,” not a literal for a compound concept.


Another possibility is {na zei .a bu}, using the lujvo {na zei .a}.
Can this really make a lujvo? (Yuck!)


> but that is going to make problems for the
> symbols for OR, AND, IFF, and whatever you call the U function, since e.g.
> {abu} is already taken.

You could always base them on se.a, se.e, se.o (or na.onai) and .unai
So: {se bu .a bu}, {tei se bu .a bu foi}, or {se zei .a bu}.

Somehow this misses the goal. Of course there are the internal forms: {najabu} and the like.

> Or is that where {zai} come in (but what selector
> gets dingbats)? {zai dy na.a bu} for => (and the like)?

{dy} is for "dingbats"? Perhaps {zai norle'u bu}?

OK. How exactly are alphabet selectors to be expressed?
> All of this is for reading off wffs, the analog of spelling, I suppose. For
> talking about the character {lo nibli sinxa} (well, {nibli} is sorta wrong,
> but I get the idea) is also possible, and I suppose could be used for reading
> off well.
> pc

Presumably if the wff was part of a Lojban text it would use
the MEX module, in which case the implication would be some
operator and you would not have to mention the symbol by name.
But I don't know if MEX is really up to the task.

No more do I, which is why I am looking at this (which, I think, actually is a part of MEX).







posts: 1912


> Another possibility is {na zei .a bu}, using the lujvo {na zei .a}.
> Can this really make a lujvo? (Yuck!)

Yes, any pair of words joined by {zei} constitute a lujvo.
(Maybe there are a couple of exceptions (like zoi?) but {na}
and {.a} are ordinary words for this.)

> You could always base them on se.a, se.e, se.o (or na.onai) and .unai
> So: {se bu .a bu}, {tei se bu .a bu foi}, or {se zei .a bu}.
>
> Somehow this misses the goal. Of course there are the internal forms:
> {najabu} and the like.

{najabu} won't work, as {na ja} is two words and {bu} takes just one.

> > Or is that where {zai} come in (but what selector
> > gets dingbats)? {zai dy na.a bu} for => (and the like)?
>
> {dy} is for "dingbats"? Perhaps {zai norle'u bu}?
>
> OK. How exactly are alphabet selectors to be expressed?

I think it works like other shifts: any lerfu that comes after the
selector will be taken to be in that alphabet until a new alphabet
is specified. So if you want just one symbol you have to return to
the normal alphabet after it.

> Presumably if the wff was part of a Lojban text it would use
> the MEX module, in which case the implication would be some
> operator and you would not have to mention the symbol by name.
> But I don't know if MEX is really up to the task.
>
> No more do I, which is why I am looking at this (which, I think, actually is
> a part of MEX).

{ni'e} turns a selbri into an operator, so {ni'e nibli},
or {ni'e brodynibli} (whatever brivla means "implication")
could be used for the implication in an expression.

mu'o mi'e xorxes


__
Do you Yahoo!?
Yahoo! SiteBuilder - Free web site building tool. Try it!
http://webhosting.yahoo.com/ps/sb/


posts: 2388


Jorge LlambĂ­as <jjllambias2000@yahoo.com.ar> wrote:


>--- John E Clifford <clifford-j@sbcglobal.net> wrote:

>> You could always base them on se.a, se.e, se.o (or na.onai) and .unai

>> So: {se bu .a bu}, {tei se bu .a bu foi}, or {se zei .a bu}.

>>

>> Somehow this misses the goal. Of course there are the internal

forms:

>> {najabu} and the like.



>{najabu} won't work, as {na ja} is two words and {bu} takes just one.



Well, one can’t help being suspcious of the definition of “word” invovled there, but the parser does seem to insist on it (in its usual deterministic but wrong-headed way). So we need – and {sei} seems to be it – a way to make one word out of the lawfully two.



>> OK. How exactly are alphabet selectors to be expressed?



>I think it works like other shifts: any lerfu that comes after the

>selector will be taken to be in that alphabet until a new alphabet

>is specified. So if you want just one symbol you have to return to

>the normal alphabet after it.



Yeah, but how are the selectors expressed? That is, what goes after {zai} to indicate what alphabet is selected? I was operating on the idea that this was coded into letters; xorxes seems to thing that a descroptive phrase is called for. Which is right or are both permitted?



>> Presumably if the wff was part of a Lojban text it would use

>> the MEX module, in which case the implication would be some

>> operator and you would not have to mention the symbol by name.

>> But I don't know if MEX is really up to the task.

>>

>> No more do I, which is why I am looking at this (which, I think,

actually is

>. a part of MEX).



>{ni'e} turns a selbri into an operator, so {ni'e nibli},

>or {ni'e brodynibli} (whatever brivla means "implication")

>could be used for the implication in an expression.



I think there are two different actions involved here, the analog of spelling – what does the expression look like – and the analog of reading – what does the expression mean. The {ni’e} trick is fine for the latter (the usual way we would use a formula), but the spelling form is also valuable – for teaching and for metatheory and that is what I am trying to figure out how to do. It’s too bad we don’t have a short, clear word for “arrow” (the lujvo look as much like spear as anything).






posts: 1912


> Jorge LlambĂ­as <jjllambias2000@yahoo.com.ar> wrote:
> >{najabu} won't work, as {na ja} is two words and {bu} takes just one.
>
> Well, one can’t help being suspcious of the definition of “word” invovled
> there, but the parser does seem to insist on it (in its usual deterministic
> but wrong-headed way). So we need – and {sei} seems to be it – a way to make
> one word out of the lawfully two.

Yes, "word" is probably not the best word here, as {zoi gy this sentence
has several words gy bu} is fine as a lerfu, as is {dei zei so'o zei mei
zei lo zei valsi bu}. We might call it a "bu-able string", and we just
have to know that {naja} is not bu-able.

> Yeah, but how are the selectors expressed? That is, what goes after {zai} to
> indicate what alphabet is selected? I was operating on the idea that this
> was coded into letters; xorxes seems to thing that a descroptive phrase is
> called for. Which is right or are both permitted?

Any lerfu is permitted after zai. The examples in CLL are:

5.2) zai .devanagar. bu
Devanagari (Hindi) alphabet
5.3) zai .katakan. bu
Japanese katakana syllabary
5.4) zai .xiragan. bu
Japanese hiragana syllabary

> I think there are two different actions involved here, the analog of spelling
> – what does the expression look like – and the analog of reading – what does
> the expression mean. The {ni’e} trick is fine for the latter (the usual way
> we would use a formula), but the spelling form is also valuable – for
> teaching and for metatheory and that is what I am trying to figure out how to
> do. It’s too bad we don’t have a short, clear word for “arrow” (the lujvo
> look as much like spear as anything).

Maybe something like farsni (farna sinxa)?

mu'o mi'e xorxes


__
Do you Yahoo!?
Yahoo! SiteBuilder - Free web site building tool. Try it!
http://webhosting.yahoo.com/ps/sb/


posts: 152

On Fri, Jan 30, 2004 at 12:08:52PM -0800, Jorge Llamb?as wrote:
> But I'd rather use the implication arrow itself for an implication arrow,
> and something like {lo nibli sinxa} to talk about them.

I'd like to draw more attention to this useful comment.

While the rest of the thread goes off with Stupid Zai Tricks, I want to
step back and ask: why in holy heck would you want to say an
_implication arrow_ out loud?

If you're expressing some logical statement, you don't say the arrow,
you say "implies" or "nibli". The arrow is a shorthand.

Compare the following two mathematical statements, and observe which one
would actually be spoken.

"For all x and all y, x equals y implies y equals x."

"Upside down capital A, x, upside down capital A, y, x, equals sign, y,
double arrow pointing to the right, y, equals sign, x."
--
mu'o mi'e rab.spir



posts: 2388

Well, when you put it that way... . There have been, however, quite a few occasions in the 43 years I taught logic (off and on) when spelling things out literally was important. How, for example, does one learn that your second sentence means your first on unless one finds out what the symbols mean, which requires symbol and meaning brought together somewhere (I rarely used a book). How, as the Polish logicians asked (this is not my experience, obviously), do you do logic in the dark for fear the Nazis would break in on you (one reason for Polish notation is that it is easier aurally than infix, lacking all those confusing parentheses)? It is not a major point (nothing about literals is a major point, since we rarely use them as such), but it does need a solution nonetheless. We don't only express formulae, we also describe them. (And {nibli} and its compounds are still the wrong words here for material implication. Invert A is also a moderately lousy version of the universal
quantifier sign, especially if combined with invert E for particular.)
pc
Rob Speer <rspeer@MIT.EDU> wrote:On Fri, Jan 30, 2004 at 12:08:52PM -0800, Jorge Llamb?as wrote:
> But I'd rather use the implication arrow itself for an implication arrow,
> and something like {lo nibli sinxa} to talk about them.
I'd like to draw more attention to this useful comment.

While the rest of the thread goes off with Stupid Zai Tricks, I want to
step back and ask: why in holy heck would you want to say an
_implication arrow_ out loud?

If you're expressing some logical statement, you don't say the arrow,
you say "implies" or "nibli". The arrow is a shorthand.

Compare the following two mathematical statements, and observe which one
would actually be spoken.

"For all x and all y, x equals y implies y equals x."

"Upside down capital A, x, upside down capital A, y, x, equals sign, y,
double arrow pointing to the right, y, equals sign, x."
--
mu'o mi'e rab.spir






posts: 1912


> (And {nibli} and its compounds are still the wrong words here
> for material implication.

What would be a good word for material implication?

Given {kanxe} and {vlina}, the place structure should be
"x1 is a material implication stating that x1 (du'u) is not
true or x2 (du'u) is true".

Peerhaps something based on {vlina}? {narvlina} would seem to
be very misleading, though perhaps no more misleading than
{na.a} is for someone who has not been told how it works.

BTW, why is it called "material" implication?

mu'o mi'e xorxes


__
Do you Yahoo!?
Yahoo! SiteBuilder - Free web site building tool. Try it!
http://webhosting.yahoo.com/ps/sb/


posts: 2388

Maybe {nafrvlina} (corrected to fit the rules, of course), following the pattern of swinging the 2nd place around to the front in tanru. Of course, that is not quite right either, but it does get toward the idea better than anything so far.
"material" because not essential — just a fact, not a necessity, as would be strct implication (and entailment and half a hundred other things in this area, all existing in countless versions.
Jorge LlambĂ­as <jjllambias2000@yahoo.com.ar> wrote:


> (And {nibli} and its compounds are still the wrong words here
> for material implication.

What would be a good word for material implication?

Given {kanxe} and {vlina}, the place structure should be
"x1 is a material implication stating that x1 (du'u) is not
true or x2 (du'u) is true".

Peerhaps something based on {vlina}? {narvlina} would seem to
be very misleading, though perhaps no more misleading than
{na.a} is for someone who has not been told how it works.

BTW, why is it called "material" implication?

mu'o mi'e xorxes


__
Do you Yahoo!?
Yahoo! SiteBuilder - Free web site building tool. Try it!
http://webhosting.yahoo.com/ps/sb/





posts: 152

On Wed, Feb 04, 2004 at 12:25:34PM -0800, John E Clifford wrote:
> Well, when you put it that way... . There have been, however, quite a few occasions in the 43 years I taught logic (off and on) when spelling things out literally was important. How, for example, does one learn that your second sentence means your first on unless one finds out what the symbols mean, which requires symbol and meaning brought together somewhere (I rarely used a book).

If you're speaking out loud, saying "(thing that means implication) bu"
won't tell you what the symbol for implication is. It will tell you how
to pronounce the symbol for implication, which I've already stated is a
useless endeavor.

If you're writing, then obviously instead of using some complicated
phrase to describe the symbol, you should draw the symbol.

And, if you're in a context where you can't draw the symbol at all:
(1) What will you gain from describing the symbol?
(2) Wouldn't you be better off describing how it's drawn, instead of its
lerfu name?

--
mu'o mi'e rab=>spir



posts: 1912


> Maybe {nafrvlina} (corrected to fit the rules, of course),

That would be {nafyvlina} or {na'ervlina}.

> following the
> pattern of swinging the 2nd place around to the front in tanru. Of course,
> that is not quite right either, but it does get toward the idea better than
> anything so far.

Yes, it does seem better than {narvlina}.

In jbovlaste I used {rolnonvlina} for {.o}. "All-nothing alternation",
"x1 states that both x2 and x3 are true or none of them is true".

mu'o mi'e xorxes


__
Do you Yahoo!?
Yahoo! SiteBuilder - Free web site building tool. Try it!
http://webhosting.yahoo.com/ps/sb/


posts: 2388

What we want is a transform of {pybu ?bu kuybu cu vlina le natfe be py (?) kuy (first ? for the unknown symbol, second for whether there need be a break to kee the p and q from running together in a literal string)
Jorge LlambĂ­as <jjllambias2000@yahoo.com.ar> wrote:

> Maybe {nafrvlina} (corrected to fit the rules, of course),

That would be {nafyvlina} or {na'ervlina}.

> following the
> pattern of swinging the 2nd place around to the front in tanru. Of
course,
> that is not quite right either, but it does get toward the idea
better than
> anything so far.






rob speer:

<<If you're speaking out loud, saying "(thing that means implication) bu"
won't tell you what the symbol for implication is. It will tell you how
to pronounce the symbol for implication, which I've already stated is a
useless endeavor.

If you're writing, then obviously instead of using some complicated
phrase to describe the symbol, you should draw the symbol.

And, if you're in a context where you can't draw the symbol at all:
(1) What will you gain from describing the symbol?
(2) Wouldn't you be better off describing how it's drawn, instead of
its
lerfu name?
If you're speaking out loud, saying "(thing that means implication) bu"
won't tell you what the symbol for implication is. It will tell you how
to pronounce the symbol for implication, which I've already stated is a
useless endeavor.

If you're writing, then obviously instead of using some complicated
phrase to describe the symbol, you should draw the symbol.

And, if you're in a context where you can't draw the symbol at all:
(1) What will you gain from describing the symbol?
(2) Wouldn't you be better off describing how it's drawn, instead of
its
lerfu name?>>


The same objections apply, of course, to all lerfu. I would assume that the various implication marks had different names (single arrow, double arrow, horseshoe, hook, to mention the most common). And why, exactly (though you claim to have explained it before) is pronouncing the symbol useless?

In the good old days (about last week it seems) the fact that somone thought he had a use for something in lojban was enough for half-a-dozen people to start with suggestions about how to do it; it was rarely the case that the response was to try to convince him he really had no such use (and, maybe, didn't even rally think he did).

pc



posts: 1912


> What we want is a transform of {pybu ?bu kuybu cu vlina le natfe be py (?)
> kuy (first ? for the unknown symbol, second for whether there need be a
> break to kee the p and q from running together in a literal string)

"P" is to "py" as "Q" is to "kybu"
"py" is to "pybu" as "kybu" is to "kybubu"

So I think what you mean, taking say {implik. bu} for the symbol
in question, is:

pybu implik.bu kybubu vlina le natfe be py boi kybu

But, shouldn't it be {me'o pybu implik.bu kybubu}?
Otherwise we seem to have just some odd pronoun.
An I'm not quite sure why {pybu} and {kybubu} instead
of {py} and {kybu}. After all, {py} and {pybu} are simply
two different lerfu. It is not clear that one is more of
a symbol than the other.

In answer to Rob:
> The same objections apply, of course, to all lerfu.

I too think all lerfu, except in their use as pronouns,
are objectionable. I think mixing lerfu with names
of symbols and characters is very confusing.

> I would assume that the
> various implication marks had different names (single arrow, double arrow,
> horseshoe, hook, to mention the most common).

And presumably different authors might use different symbols
for the same concept, and one symbol might be used by different
authors for different concepts. So "single arrow" may or may
not correspond with "implication symbol", depending on context.
I've no idea whether a lerfu could represent "single arrow",
"implication symbol", neither, or both, or some combination.

> And why, exactly (though you
> claim to have explained it before) is pronouncing the symbol useless?

I'm not even sure at this point whether the symbol is to be
pronounced according to its form or to its function.

mu'o mi'e xorxes


__
Do you Yahoo!?
Yahoo! Finance: Get your refund fast by filing online.
http://taxes.yahoo.com/filing.html


posts: 152

On Thu, Feb 05, 2004 at 07:23:52AM -0800, John E Clifford wrote:
> And, if you're in a context where you can't draw the symbol at all:
> (1) What will you gain from describing the symbol?
> (2) Wouldn't you be better off describing how it's drawn, instead of
> its
> lerfu name?>>
>
>
> The same objections apply, of course, to all lerfu. I would assume that the various implication marks had different names (single arrow, double arrow, horseshoe, hook, to mention the most common). And why, exactly (though you claim to have explained it before) is pronouncing the symbol useless?
>
> In the good old days (about last week it seems) the fact that somone thought he had a use for something in lojban was enough for half-a-dozen people to start with suggestions about how to do it; it was rarely the case that the response was to try to convince him he really had no such use (and, maybe, didn't even rally think he did).

The problem is not that the name of a symbol isn't expressible in
Lojban. It's that it's as expressible as it is in other languages. If
you have a symbol you want to call a horseshoe, you can call it "lo
xircutci tarmi sinxa".

But there's something that makes people want to express this, not using
a meaningful phrase like "lo xircutci tarmi sinxa", but using some
clever, contrived, and generally incomprehensible cmavo trick.

I remember hearing that when Lojban was being originally developed,
people wanted to be able to express every conceivable typographical
symbol with its own cmavo, so the _entire_ c??? cmavo space was
suggested to be set aside for lerfu. (pc, you could probably get the
details more accurate than I'm getting them, of course).

I assume that somebody suddenly remembered that Zipf guy, because
thankfully that didn't happen.

Was that the quashing of a useful feature that someone wanted? No, that
was keeping sanity in the language. (Still, the lerfu fetish managed to
leave its mark in the form of 'lau'.) The trend that I try to fight is
that someone says they don't know how to express a certain thing in
Lojban, and several people propose _new language constructs_ (sure, they
can be hacked out of old cmavo, but they're a new interpretation) to say
that thing, even when existing constructs in the language already work.
Coming up with new language constructs is fun, but unless they're
actually beneficial they simply make Lojban more complicated.

--
mu'o mi'e rab.spir



posts: 2388


Rab spir:

<<On Thu, Feb 05, 2004 at 07:23:52AM -0800, John E Clifford wrote:

> And, if you're in a context where you can't draw the symbol at all:

> (1) What will you gain from describing the symbol?

> (2) Wouldn't you be better off describing how it's drawn, instead of

> its

> lerfu name?>>

>

>

> The same objections apply, of course, to all lerfu. I would assume

that the various implication marks had different names (single arrow,

double arrow, horseshoe, hook, to mention the most common). And why,

exactly (though you claim to have explained it before) is pronouncing the

symbol useless?

>

> In the good old days (about last week it seems) the fact that somone

thought he had a use for something in lojban was enough for

half-a-dozen people to start with suggestions about how to do it; it was rarely the case that the response was to try to convince him he really had no

such use (and, maybe, didn't even rally think he did).



The problem is not that the name of a symbol isn't expressible in

Lojban. It's that it's as expressible as it is in other languages. If

you have a symbol you want to call a horseshoe, you can call it "lo

xircutci tarmi sinxa".



But there's something that makes people want to express this, not using

a meaningful phrase like "lo xircutci tarmi sinxa", but using some

clever, contrived, and generally incomprehensible cmavo trick.>>



I am not clear why the trick is incomprehensible – once the rule is worked out. It may not be inferable from other things in the language, but then neither are most things. And, of course, I did not require (even suggest, I think) that the device was a cmavo in the strict sense of that term – though it would seem to have to fit into some cmavo grammatical category, probably BY.



<<I remember hearing that when Lojban was being originally developed,

people wanted to be able to express every conceivable typographical

symbol with its own cmavo, so the _entire_ c??? cmavo space was

suggested to be set aside for lerfu. (pc, you could probably get the

details more accurate than I'm getting them, of course).



I assume that somebody suddenly remembered that Zipf guy, because

thankfully that didn't happen.>>



Not quite what happened but to that effect. The original Loglan, the 1960 Scientific American form, had enough not just puctuations but typesetting marks to make LaTeX look simpleminded. But it was not (directly) Zipf that killed them; it was JCB needing space to do a mess of other things and just taking what he thought he could get along without (but, as you note, never losing them entirely). That is ultimately Zipfy, to be sure, but the move was not conscious. I seem to recall that some similar proposal was made in among the thousands tht flew around in the late 80s-early 90s for Lojban, but the fate of the Loglan cases was enough to quell that particular swell (if it was that big) and we fell back on essentially the Loglan solution (in spirit, if not always in details). We did keep the lerfu properly so called and the whole muck of MEX.



<<Was that the quashing of a useful feature that someone wanted? No, that

was keeping sanity in the language. (Still, the lerfu fetish managed to

leave its mark in the form of 'lau'.) The trend that I try to fight is

that someone says they don't know how to express a certain thing in

Lojban, and several people propose _new language constructs_ (sure,

They can be hacked out of old cmavo, but they're a new interpretation) to

say that thing, even when existing constructs in the language already work.

Coming up with new language constructs is fun, but unless they're

actually beneficial they simply make Lojban more complicated.>>



I don’t follow this. Where is the new language construct here? All the suggestions so far have been well within the existing constructs, being just new instances of those constructs. And what different form does the job in mind, being pronounced ins spelling out a physical string? Would you want to spell by describing each letter as in the jazz spelling of Mississippi (em eye crooked letter crooked letter eye/ crooked letter crooked letter eye humpback humback eye) only completely?



xorxes:




<<> What we want is a transform of {pybu ?bu kuybu cu vlina le natfe be

py (?)

> kuy (first ? for the unknown symbol, second for whether there need

be a

> break to kee the p and q from running together in a literal string)



"P" is to "py" as "Q" is to "kybu"

"py" is to "pybu" as "kybu" is to "kybubu">>



Thanks. I can never remember how to handle odd letters in Lojban’s deficient alphabet. I suppose that {kuy} is illegit in some way. Pity.



<<So I think what you mean, taking say {implik. bu} for the symbol

in question, is:



pybu implik.bu kybubu vlina le natfe be py boi kybu



But, shouldn't it be {me'o pybu implik.bu kybubu}?>>

Otherwise we seem to have just some odd pronoun.

An I'm not quite sure why {pybu} and {kybubu} instead

of {py} and {kybu}. After all, {py} and {pybu} are simply

two different lerfu. It is not clear that one is more of

a symbol than the other.>>

Thanks again. The {me’o}, however otherwise obnoxious, does allow me to come down a {bu} and still avoid the pronoun problem (which I did not avoid anyhow, as you note).





<<In answer to Rob:

> The same objections apply, of course, to all lerfu.



I too think all lerfu, except in their use as pronouns,

are objectionable. I think mixing lerfu with names

of symbols and characters is very confusing.



> I would assume that the

> various implication marks had different names (single arrow, double

arrow,

> horseshoe, hook, to mention the most common).



And presumably different authors might use different symbols

for the same concept, and one symbol might be used by different

authors for different concepts. So "single arrow" may or may

not correspond with "implication symbol", depending on context.

I've no idea whether a lerfu could represent "single arrow",

"implication symbol", neither, or both, or some combination.



> And why, exactly (though you

> claim to have explained it before) is pronouncing the symbol useless?



I'm not even sure at this point whether the symbol is to be

pronounced according to its form or to its function.>>



This discussion had begun to feel hauntingly familiar until xorxes’ last couple of remarks brought the “original” to mind. In the Middle Ages, writing was free of most punctuation marks (except abbreviations and indications of those). This led to major ambiguities in closely reasoned debates – philosophical/theological arguments. So a whole theory developed – supposition – to account for and remove the ambiguities. Though the names varied, there were usually at least the following types of “meanings” of expressions:

The physical expression itself: “Man has three letters”

The expression as a bit of language: “Man is a noun”

The expression as referring to a concept: “Man is subsumed under animal.”

The expression as presenting a concept: “Man is a rational animal”

The exprssion as referring to objects “Man is any human or any male human”



What is happening here is a failure to distinguish among some of the earlier of these critters. I think I was looking for a way to present the physical expression orally rather than in writing. In that case, different forms of the implication sign would probably have different representations. But I may have meant bits of language (and I a pretty certainly shifted back and forth between at least these two), in which case the shape of the sign would be unimportant and its function (as a contrast to say alternation sign or quantifier) would be more significant. In the sentence I aimed at and missed, however, I was pretty clearly dealing with meaning and so the logical function comes to the fore, the shape being even less important than before.



Now, quite frankly, I want to do at least all three of these things – on different occasions – and to keep them clearly separated when they come close to one another (and it would be nice if this could be by devices that could be extended throughout MEX, since it seems to be pervaded by this kind of imprecision – leaving aside all its other problems). And it does not seem to me that we can always rely on context to sort these critters out, so we “need” (no one but an occasional philosopher is going to get caught in this except occasionally – Korzybski to the contrary notwithstanding) either different symbols or a disambiguating device to be called up when needed. I have no problem with descriptive names, suitably modified to indicate they are symbols ({lo xircutri tanxi sinxa} is just the wrong grammatical type for use in the form of an expression as such): {zai xircutri bu} or {lau…} or whatever does for that. Similarly, {nafyvlina bu} does find once the conceptual level is
reached; the linguistic one could go either way I guess.

And maybe part of the solution lies in {la’e} and {lu’e}, though that may require something new.




posts: 1912


> xorxes:
> "P" is to "py" as "Q" is to "kybu"

("kybu" should have been "ky.bu")

> Thanks. I can never remember how to handle odd letters in Lojban’s deficient
> alphabet. I suppose that {kuy} is illegit in some way. Pity.

h, q and w are {.y'y.bu}, {ky.bu} and {vy.bu}.

CLL lists iy and uy among the Lojban diphthongs, but they can only
appear in cmene. I avoid them completely, as I find them very
hard to pronounce.

> <<
> pybu implik.bu kybubu vlina le natfe be py boi kybu
>
> But, shouldn't it be {me'o pybu implik.bu kybubu}?>>
>
> Thanks again. The {me’o}, however otherwise obnoxious, does allow me to come
> down a {bu} and still avoid the pronoun problem (which I did not avoid
> anyhow, as you note).

That still leaves the question as to the difference if any between
{me'o py. implik.bu ky.bu} and {me'o vei py. ni'e nafyvlina ky.bu} and
{me'o vei py. na.a ky.bu}. They are all grammatical, but are they all
sensical?, and if so, do they mean the same thing?

...
> Though the names
> varied, there were usually at least the following types of “meanings” of
> expressions:
>
> The physical expression itself: “Man has three letters”
> The expression as a bit of language: “Man is a noun”

Those two would appear to be {zo nanmu} in Lojban:

zo nanmu cu lerfu mumoi
zo nanmu cu gismu

> The expression as referring to a concept: “Man is subsumed under animal.”
> The expression as presenting a concept: “Man is a rational animal”
> The exprssion as referring to objects “Man is any human or any male human”

Those three would be {lo nanmu} with my understanding of {lo}:

lo nanmu cu klesi lo danlu
lo nanmu cu racli danlu
lo nanmu cu du ro nakni remna

> What is happening here is a failure to distinguish among some of the earlier
> of these critters.

Lojban at least makes a two-way distinction, where English makes none.

> I think I was looking for a way to present the physical
> expression orally rather than in writing. In that case, different forms of
> the implication sign would probably have different representations.

What would be the analogous of the first two "man" for symbols?
"1+1" is pronounced in Lojban "(li/me'o) pa su'i pa"
So we could say:

zo su'i kruca
"+" is a cross

zo su'i mekfancu valsi
"su'i" is a MEX operator word

> But I
> may have meant bits of language (and I a pretty certainly shifted back and
> forth between at least these two), in which case the shape of the sign would
> be unimportant and its function (as a contrast to say alternation sign or
> quantifier) would be more significant.

But so far there appears to be no reason to leave "zo" or eventually
"lu-li'u" to talk about these things, which is the normal Lojban way
to deal both with physical expressions and with bits of language.

{me'o} for example can't deal with {su'i} because it is not a complete
expression. Similarly it can't deal with the physical expression nor with
the bit of language {na.a}.

> In the sentence I aimed at and
> missed, however, I was pretty clearly dealing with meaning and so the logical
> function comes to the fore, the shape being even less important than before.

I can't quite grasp what {me'o} is about. Is it about bits of language,
like gismu/cmavo/mekso?

> Now, quite frankly, I want to do at least all three of these things – on
> different occasions – and to keep them clearly separated when they come close
> to one another (and it would be nice if this could be by devices that could
> be extended throughout MEX, since it seems to be pervaded by this kind of
> imprecision – leaving aside all its other problems).

It seems so to me too.

> And it does not seem to
> me that we can always rely on context to sort these critters out, so we
> “need” (no one but an occasional philosopher is going to get caught in this
> except occasionally – Korzybski to the contrary notwithstanding) either
> different symbols or a disambiguating device to be called up when needed.

It would be useful to first have some examples of the kinds of things
we would like to say about these things.

> I
> have no problem with descriptive names, suitably modified to indicate they
> are symbols ({lo xircutri tanxi sinxa} is just the wrong grammatical type for
> use in the form of an expression as such): {zai xircutri bu}

{zai xircutci bu} would seem to select the Horse-shoe alphabet...

>or {lau…} or
> whatever does for that. Similarly, {nafyvlina bu} does find once the
> conceptual level is
> reached; the linguistic one could go either way I guess.

ko ciska lo xircutci tarmi sinxa le tanbo
Write a horse-shoe symbol on the board.

?ko ciska me'o xircutci bu le tanbo
?Write the expression consisting of the horse-shoe (operand) on the board.

?ko ciska me'o nafyvlina bu le tanbo
?Write the expression consisting of the implication (operand) on the board.

ko ciska me'o vei py na.a ky.bu le tanbo
Write the expression "if P then Q" on the board.
(Would this be satisfied by writing "P->Q"?)

?ko ciska me'o py nafyvlina bu ky.bu le tanbo

mu'o mi'e xorxes


__
Do you Yahoo!?
Yahoo! Finance: Get your refund fast by filing online.
http://taxes.yahoo.com/filing.html


posts: 2388


Jorge LlambĂ­as <jjllambias2000@yahoo.com.ar> wrote:

xorxes:

<<--- John E Clifford <clifford-j@sbcglobal.net> wrote:

> xorxes:

> "P" is to "py" as "Q" is to "kybu"



("kybu" should have been "ky.bu")



> Thanks. I can never remember how to handle odd letters in Lojban’s

deficient

> alphabet. I suppose that {kuy} is illegit in some way. Pity.



h, q and w are {.y'y.bu}, {ky.bu} and {vy.bu}.



CLL lists iy and uy among the Lojban diphthongs, but they can only

appear in cmene. I avoid them completely, as I find them very

hard to pronounce.>>



Both those diphthongs are in my native idiolect, so I find them natural. And kuy, uy and iy as more natural forms for q, w, and y (in the English sense).



<<> <<

> pybu implik.bu kybubu vlina le natfe be py boi kybu

>

> But, shouldn't it be {me'o pybu implik.bu kybubu}?>>

>

> Thanks again. The {me’o}, however otherwise obnoxious, does allow me

to come

> down a {bu} and still avoid the pronoun problem (which I did not

avoid

> anyhow, as you note).



That still leaves the question as to the difference if any between

{me'o py. implik.bu ky.bu} and {me'o vei py. ni'e nafyvlina ky.bu} and

{me'o vei py. na.a ky.bu}. They are all grammatical, but are they all

sensical?, and if so, do they mean the same thing? >>



I don’t get the imprssion from the wordlists that {ni’e} would work in this way, but then it is remarkably unclear just how it does work. In terms of the supposition sets, they seem to be different: first about the formula as object, second about the meaning of the formula and third simply stating the situation. But then again maybe not.



...

<<> Though the names

> varied, there were usually at least the following types of “meanings”

of

> expressions:

>

> The physical expression itself: “Man has three letters”

> The expression as a bit of language: “Man is a noun”



Those two would appear to be {zo nanmu} in Lojban:



zo nanmu cu lerfu mumoi

zo nanmu cu gismu>>



Which suggests that {zo} (and the rest of the quotation apparatus, covering longer expressions in and out of lojban) is still ambiguous at a certain (extremely rare) level, not distinguishing a bit of language (its usual use) from the bare physical object as such.



<<> The expression as referring to a concept: “Man is subsumed under

animal.”

> The expression as presenting a concept: “Man is a rational animal”

> The exprssion as referring to objects “Man is any human or any male

human”



Those three would be {lo nanmu} with my understanding of {lo}:



lo nanmu cu klesi lo danlu

lo nanmu cu racli danlu

lo nanmu cu du ro nakni remna>>



I happily don’t remember what your {lo} is – something incoherent like Mr. nanmu comes to mind but I don’t believe that of you – so I can’t speak to what that catches of the distinctions intended. For standard {lo}, none of these work for the intended sense (that is they are all false sentences). I think that applies as well for Mr. nanmu, but incoherent concepts are notoriously hard to pin down in real cases.



<<> What is happening here is a failure to distinguish among some of the

earlier

> of these critters.



Lojban at least makes a two-way distinction, where English makes none.>>



Well, written English has, like lojban, quotes, but they work no better there than in lojban. Written English also uses capital letters to do a bit of the work (though that helps not at all at the beginning of sentences, as here). I suppose that lojban also can make some use of {ka} and {du’u} and, indeed, {lo’i}, and maybe other items in their neighborhoods, just as English can say explicitly “the set of” or “the concept of” and the like.



<<> I think I was looking for a way to present the physical

> expression orally rather than in writing. In that case, different

forms of

> the implication sign would probably have different representations.



What would be the analogous of the first two "man" for symbols?

"1+1" is pronounced in Lojban "(li/me'o) pa su'i pa"

So we could say:



zo su'i kruca

"+" is a cross



zo su'i mekfancu valsi

"su'i" is a MEX operator word>>



That looks about right.



<<> But I

> may have meant bits of language (and I a pretty certainly shifted

back and

> forth between at least these two), in which case the shape of the

sign would

> be unimportant and its function (as a contrast to say alternation

sign or

> quantifier) would be more significant.



But so far there appears to be no reason to leave "zo" or eventually

"lu-li'u" to talk about these things, which is the normal Lojban way

to deal both with physical expressions and with bits of language.



{me'o} for example can't deal with {su'i} because it is not a complete

expression. Similarly it can't deal with the physical expression nor

with

the bit of language {na.a}.>>



And the result is an ambiguity: the zo su’I of which the first is true is not one of which the second is true and conversely. The difference is not exactly use-mention but is clearly another sort of map-teritory mixup.



<<> In the sentence I aimed at and

> missed, however, I was pretty clearly dealing with meaning and so the

logical

> function comes to the fore, the shape being even less important than

before.



I can't quite grasp what {me'o} is about. Is it about bits of language,

like gismu/cmavo/mekso?>>



I fear that it may be indsicriminately about all three and leave us to sort it out by context, which we can usually do, but for which we can pretty easily construct unsorted cases: {zo su’i sampu} for example.



<<> Now, quite frankly, I want to do at least all three of these things –

on

> different occasions – and to keep them clearly separated when they

come close

> to one another (and it would be nice if this could be by devices that

could

> be extended throughout MEX, since it seems to be pervaded by this

kind of

> imprecision – leaving aside all its other problems).



It seems so to me too.



> And it does not seem to

> me that we can always rely on context to sort these critters out, so

we

> “need” (no one but an occasional philosopher is going to get caught

in this

> except occasionally – Korzybski to the contrary notwithstanding)

either

> different symbols or a disambiguating device to be called up when

needed.



It would be useful to first have some examples of the kinds of things

we would like to say about these things.>>



Well, you have given some above: the shape of the physical object, the grammatical category of a bit of language, the specific meaning of a word, and so on. And, of course (see my example above) there are cases that could be meant for any of these: a cross is simple as a design, being at most a second order pattern, {su’i} is simple as a basic operator symbol, and addition is simple, being one of the basic arithmetical operations (after successor).



<<> I

> have no problem with descriptive names, suitably modified to indicate

they

> are symbols ({lo xircutri tanxi sinxa} is just the wrong grammatical

type for

> use in the form of an expression as such): {zai xircutri bu}



{zai xircutci bu} would seem to select the Horse-shoe alphabet...>>



Yes, what seems to be needed is {lau}, though what the “two-word” means in its definition is unclear. Nor is it clear that we then need {bu}.



>or {lau…} or

> whatever does for that. Similarly, {nafyvlina bu} does find once the

> conceptual level is

> reached; the linguistic one could go either way I guess.



ko ciska lo xircutci tarmi sinxa le tanbo

Write a horse-shoe symbol on the board.



?ko ciska me'o xircutci bu le tanbo

?Write the expression consisting of the horse-shoe (operand) on the

board.



?ko ciska me'o nafyvlina bu le tanbo

?Write the expression consisting of the implication (operand) on the

board.



ko ciska me'o vei py na.a ky.bu le tanbo

Write the expression "if P then Q" on the board.

(Would this be satisfied by writing "P->Q"?)



?ko ciska me'o py nafyvlina bu ky.bu le tanbo>>

“?” indeed. I don’t know about any of these, which is why I am

asking.

pc








posts: 1912

> <<
> That still leaves the question as to the difference if any between
> {me'o py. implik.bu ky.bu} and {me'o vei py. ni'e nafyvlina ky.bu} and
> {me'o vei py. na.a ky.bu}. They are all grammatical, but are they all
> sensical?, and if so, do they mean the same thing? >>
>
> I don’t get the imprssion from the wordlists that {ni’e} would work in this
> way, but then it is remarkably unclear just how it does work.

ni'e makes an operator, something like su'i, out of a selbri.
{ni'e nafyvlina} works like {su'i}. (In fact {su'i} can be taken
as a short form of {ni'e sumji}.) {vei py. su'i ky.bu} is the
sum P + Q, {vei py. ni'e nafyvlina ky.bu} is the implication
if P then Q. How would you expect it to work?

...
> <<{me'o} for example can't deal with {su'i} because it is not a complete
> expression. Similarly it can't deal with the physical expression nor
> with
> the bit of language {na.a}.>>

I had forgotten about {me'o su'ibu}, the MEX expression of
(operand?) su'ibu.


...
> {zai xircutci bu} would seem to select the Horse-shoe alphabet...>>
>
> Yes, what seems to be needed is {lau}, though what the “two-word” means in
> its definition is unclear. Nor is it clear that we then need {bu}.

{lau} has to be followed by a lerfu, such as {by}, {ky.bu}, {valsi bu},
etc. {bu} is needed for all lerfu except members of BY.


> ko ciska lo xircutci tarmi sinxa le tanbo
> ?ko ciska me'o xircutci bu le tanbo
> ?ko ciska me'o nafyvlina bu le tanbo
> ko ciska me'o vei py na.a ky.bu le tanbo
> ?ko ciska me'o py nafyvlina bu ky.bu le tanbo>>
>
> “?” indeed. I don’t know about any of these, which is why I am
> asking.

We can only speculate. The clearest path is usually to avoid
mekso altogether.

mu'o mi'e xorxes


__
Do you Yahoo!?
Yahoo! Finance: Get your refund fast by filing online.
http://taxes.yahoo.com/filing.html


posts: 324
{se'e} should be terminated by the next word that isn't a digit. If you use {na'a} to terminate it, that also cancels uppercase and other alphabet shifts. For instance, {ga'e dy .ebu se'e pa mu rei na'a .ibu} spells "DEĂ…Ĺľi", not "DEĂ…ĹľI". What the proper way to terminate it when it's followed by a digit is, I'm not sure.
posts: 1912


>
> Re: BPFK Section: lerfu Shifts
>
> {se'e} should be terminated by the next word that isn't a digit. If you use
> {na'a} to terminate it, that also cancels uppercase and other alphabet
> shifts. For instance, {ga'e dy .ebu se'e pa mu rei na'a .ibu} spells "DEĂ…Ĺľi",
> not "DEĂ…ĹľI". What the proper way to terminate it when it's followed by a
> digit is, I'm not sure.

Probably you have to do {tei se'e pa mu rei foi pa}.

Couldn't there be character codes that use letters as the code instead
of numbers?

mu'o mi'e xorxes



__
Do you Yahoo!?
Yahoo! Finance: Get your refund fast by filing online.
http://taxes.yahoo.com/filing.html


posts: 23

I will vote "yes" so as not to stand in the way of progress.

However:

The definitions of ga'e and that of the other equivalent words, includint to'a, are not parallel. There is a little more detail about scope in the ga'e definition. In a dictionary, these entries will not be grouped together; so, really, the extra detail should also be included in the other definitions.

As to the non-usage: A likely interpretive convention (a usage Bob says was intended to be allowed for), and one that does not contradict any of the grammar, is that a parenthetical shift or font-and-face change that is not followed by lerfu would be taken as applying to the words - sort of like a mark-up language. For example: "to'i ga'e toi mi to'i to'a toi klama" would be "MI klama". This convention might mean that these words would get used more in the future.

posts: 14214

On Thu, Mar 04, 2004 at 08:03:22PM -0800, noras wrote:
> I will vote "yes" so as not to stand in the way of progress.

While I appreciate that, I *do* want the results of the BPFK to be the
best they can be. Your *timely* feedback is more important to me than
your capitulation. :-)

> However:
>
> The definitions of ga'e and that of the other equivalent words,
> including to'a, are not parallel. There is a little more detail about
> scope in the ga'e definition. In a dictionary, these entries will not
> be grouped together; so, really, the extra detail should also be
> included in the other definitions.

That makes excellent sense. treed, please make it so.

> As to the non-usage: A likely interpretive convention (a usage Bob
> says was intended to be allowed for), and one that does not contradict
> any of the grammar, is that a parenthetical shift or font-and-face
> change that is not followed by lerfu would be taken as applying to the
> words - sort of like a mark-up language. For example: "to'i ga'e toi
> mi to'i to'a toi klama" would be "MI klama". This convention might
> mean that these words would get used more in the future.

Oh. That's actually rather interesting. Probably worth mentioning, at
least briefly, on the page itself.

-Robin


posts: 14214

> rlpowell:
> On Thu, Mar 04, 2004 at 08:03:22PM -0800, noras wrote:
> > However:
> >
> > The definitions of ga'e and that of the other equivalent words,
> > including to'a, are not parallel. There is a little more detail about
> > scope in the ga'e definition. In a dictionary, these entries will not
> > be grouped together; so, really, the extra detail should also be
> > included in the other definitions.
>
> That makes excellent sense. treed, please make it so.

Errr. "Rob", not "treed". Wrong section.

-Robin

posts: 1912


> For example: "to'i ga'e toi mi to'i to'a toi klama" would
> be "MI klama". This convention might mean that these words would get used
> more in the future.

Would that mean that we would write it down as "MI klama" and read
it aloud as "to'i ga'e toi mi to'i to'a toi klama"?

What would its meaning be? Something similar to "ba'e mi klama"?

Upper case in English is sometimes used to represent in writing
a certain way of speaking. In lojban we would be doing a second order
representation: we would use some words to represent upper case, which
in turn would represent something else (emphasis?).

mu'o mi'e xorxes



__
Do you Yahoo!?
Yahoo! Search - Find what you’re looking for faster
http://search.yahoo.com


posts: 14214

> xorxes:
>
> > For example: "to'i ga'e toi mi to'i to'a toi klama" would
> > be "MI klama". This convention might mean that these words would get used
> > more in the future.
>
> Would that mean that we would write it down as "MI klama" and read
> it aloud as "to'i ga'e toi mi to'i to'a toi klama"?

Speaking only for myself, I would both write and speak it in Lojban with the to'i ... toi phrase. But in other circumstances, for example when using it as input to something that needed capitilization, or in an overall English text where I felt some words needed special emphasis or whatever, I would write it MI klama.

> What would its meaning be? Something similar to "ba'e mi klama"?

Probably. Something that is often forgotten in these kinds of discussions is that it is perfectly reasonable for an author to say "lo valsi poi galfi sepi'o zo ga'e cu sinxa lo broda" where broda might be "glico valsi poi mi fanva fi la .lojban." or "basna" or something else entirely.

Not every interpretive convention needs to be defined in advance; leave a bit of wiggle room for future authors to engage in nonce practices without mis-using the language, please!

-Robin