Context and precision (was:Re: [lojban-beginners] Special reference, underspecified)

[Jonathan Jones]

(In the event that my comments below spark a discussion, I've moved this to the main list and altered the subject to reflect the topic at hand.)

On Thu, Jan 3, 2013 at 8:34 AM, Ian Johnson <blindb...@gmail.com> wrote:

This drastically hinders the precision (in fact it basically makes it exactly the same as PA lo GISMU), but yes, PA GISMU under this model is an extremely compact form for a rather uncommon usage.

But fair enough; at the very least the {pa se pelkre} version is sketchy, and we'll leave it at that.

I'm a firm believer in determinism via context and precision via verbosity. That is, I believe that anything about an expression that can be figured out from the context that expression is in, should be left to context, and that a more precise expression should always, without exception, require more words to express than a similar, less precise one.

My reasoning behind both is simply that following those doctrines means we can say things in Lojban more succintly, a definite boon considering the wordiness the language has, IMO.

And it's not at all difficult to add {roroi fe'eroroi} to make the context be "always and everywhere". It does, obviously, require more words, but that's in line with the above.
 

mi'e la latro'a mu'o

On Thu, Jan 3, 2013 at 10:29 AM, Jonathan Jones <eye...@gmail.com> wrote:

On Thu, Jan 3, 2013 at 8:21 AM, Ian Johnson <blindb...@gmail.com> wrote:

The first version is wrong; that says "there exists exactly one yellow haired thing which ..." (Rather, there is a very slightly less controversial version that I tend to subscribe to: "there is exactly one yellow-haired thing that is at all possible to come up in discussion which ..." The point is that the default universe of discourse should be the universe of discourse, there shouldn't be anything that be reasonably anticipated to come up in conversation that isn't actually in it.)

I subscribe to the idea that the "universe of discourse" is bound by the context that discourse is in, mainly because forcing PA GISMU to mean that there is exactly PA things that GISMU makes the construct pretty much useless, seeing as statements like that are pretty much always incorrect. But this isn't the list to talk about thee issues, being the beginner's thread and all.
 

If you use su'o instead it's not terrible, but really {da} constructs when you don't want nice sharp precision tend to be undesirable, in my experience.

At any rate, here I'd probably say {lo se ke pelxu kerfa}. {le} might also be appropriate in this context, but I doubt I'd use it. That said, I think more likely the structure of the discussion would be different among fluent in-real-life Lojban speakers. (Few of these exist; I, for example, am pretty quick on IRC, but am almost completely helpless when it comes to spatial things.) I'd expect a lot of up-front establishment of pronouns, for example.

mi'e la latro'a mu'o

On Wed, Jan 2, 2013 at 1:30 PM, Jonathan Jones <eye...@gmail.com> wrote:

On Mon, Dec 31, 2012 at 10:01 AM, neizyn. <sja...@gmail.com> wrote:

coi

.i mi'e neizyn.

 

Sorry, I'm just a beginnger! but I was wondering about a special situation.

Suppose there's a group of girls with one guy; I'm talking to one of the girls, and I make a statement referring to the guy without unambiguously referring to him by name or other reference. This is because he's the only person in the group like that. Say he's blonde. Is there a way I can say that "someone is blond" oooooor 'blonde exists/obtains such that x is blonde' ? Better yet, could I use that with some kind of reference? 

{pa pelxre} == {pa da poi pelkre} => "one X which-is golden-haired"
{pa lo prenu poi pelxre} => "A singular person which-is blonde"
{lo pelxre prenu} => "One or more blonde type-of person"
{ko'a goi (any of the above)} => "It-1, which-is what we will call {any of the above)"

I'm sure there's other ways to do it as well, but those are the most obvious, to me at any rate.

--
mu'o mi'e .aionys.

.i.e'ucai ko cmima lo pilno be denpa bu .i doi.luk. mi patfu do zo'o
(Come to the Dot Side! Luke, I am your father. :D )

--
You received this message because you are subscribed to the Google Groups "Lojban Beginners" group.

To post to this group, send email to lojban-b...@googlegroups.com.
To unsubscribe from this group, send email to lojban-beginne...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/lojban-beginners?hl=en.

--
You received this message because you are subscribed to the Google Groups "Lojban Beginners" group.

To post to this group, send email to lojban-b...@googlegroups.com.
To unsubscribe from this group, send email to lojban-beginne...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/lojban-beginners?hl=en.

--
mu'o mi'e .aionys.

.i.e'ucai ko cmima lo pilno be denpa bu .i doi.luk. mi patfu do zo'o
(Come to the Dot Side! Luke, I am your father. :D )

--
You received this message because you are subscribed to the Google Groups "Lojban Beginners" group.
To post to this group, send email to lojban-b...@googlegroups.com.
To unsubscribe from this group, send email to lojban-beginne...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/lojban-beginners?hl=en.

--
You received this message because you are subscribed to the Google Groups "Lojban Beginners" group.
To post to this group, send email to lojban-b...@googlegroups.com.
To unsubscribe from this group, send email to lojban-beginne...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/lojban-beginners?hl=en.

--
mu'o mi'e .aionys.

.i.e'ucai ko cmima lo pilno be denpa bu .i doi.luk. mi patfu do zo'o
(Come to the Dot Side! Luke, I am your father. :D )

[la gleki]

On Thursday, January 3, 2013 8:01:16 PM UTC+4, aionys wrote:

(In the event that my comments below spark a discussion, I've moved this to the main list and altered the subject to reflect the topic at hand.)

On Thu, Jan 3, 2013 at 8:34 AM, Ian Johnson <blindb...@gmail.com> wrote:

This drastically hinders the precision (in fact it basically makes it exactly the same as PA lo GISMU), but yes, PA GISMU under this model is an extremely compact form for a rather uncommon usage.

But fair enough; at the very least the {pa se pelkre} version is sketchy, and we'll leave it at that.

I'm a firm believer in determinism via context and precision via verbosity. That is, I believe that anything about an expression that can be figured out from the context that expression is in, should be left to context, and that a more precise expression should always, without exception, require more words to express than a similar, less precise one.

i'e

 

My reasoning behind both is simply that following those doctrines means we can say things in Lojban more succintly, a definite boon considering the wordiness the language has, IMO.

And it's not at all difficult to add {roroi fe'eroroi} to make the context be "always and everywhere". It does, obviously, require more words, but that's in line with the above.

sorry for asking questions not necessarily relevant to this discussion but can't we just say 

{ca ro da bu'u ro de} instead of {roroi fe'eroroi} or even {cajebu'u ro da} ?

[Jonathan Jones]

On Thu, Jan 3, 2013 at 10:01 AM, la gleki <gleki.is...@gmail.com> wrote:

On Thursday, January 3, 2013 8:01:16 PM UTC+4, aionys wrote:

(In the event that my comments below spark a discussion, I've moved this to the main list and altered the subject to reflect the topic at hand.)

On Thu, Jan 3, 2013 at 8:34 AM, Ian Johnson <blindb...@gmail.com> wrote:

This drastically hinders the precision (in fact it basically makes it exactly the same as PA lo GISMU), but yes, PA GISMU under this model is an extremely compact form for a rather uncommon usage.

But fair enough; at the very least the {pa se pelkre} version is sketchy, and we'll leave it at that.

I'm a firm believer in determinism via context and precision via verbosity. That is, I believe that anything about an expression that can be figured out from the context that expression is in, should be left to context, and that a more precise expression should always, without exception, require more words to express than a similar, less precise one.

i'e

 

My reasoning behind both is simply that following those doctrines means we can say things in Lojban more succintly, a definite boon considering the wordiness the language has, IMO.

And it's not at all difficult to add {roroi fe'eroroi} to make the context be "always and everywhere". It does, obviously, require more words, but that's in line with the above.

sorry for asking questions not necessarily relevant to this discussion but can't we just say 

{ca ro da bu'u ro de} instead of {roroi fe'eroroi} or even {cajebu'u ro da} ?

I don't see how either of those could mean "always and everywhere: the whole of time and space"

--
You received this message because you are subscribed to the Google Groups "lojban" group.
To view this discussion on the web visit https://groups.google.com/d/msg/lojban/-/j3NErWcKXg8J.
To post to this group, send email to loj...@googlegroups.com.
To unsubscribe from this group, send email to lojban+un...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/lojban?hl=en.

[Jonathan Jones]

On Thu, Jan 3, 2013 at 10:59 AM, Jonathan Jones <eye...@gmail.com> wrote:

On Thu, Jan 3, 2013 at 10:01 AM, la gleki <gleki.is...@gmail.com> wrote:

On Thursday, January 3, 2013 8:01:16 PM UTC+4, aionys wrote:

(In the event that my comments below spark a discussion, I've moved this to the main list and altered the subject to reflect the topic at hand.)

On Thu, Jan 3, 2013 at 8:34 AM, Ian Johnson <blindb...@gmail.com> wrote:

This drastically hinders the precision (in fact it basically makes it exactly the same as PA lo GISMU), but yes, PA GISMU under this model is an extremely compact form for a rather uncommon usage.

But fair enough; at the very least the {pa se pelkre} version is sketchy, and we'll leave it at that.

I'm a firm believer in determinism via context and precision via verbosity. That is, I believe that anything about an expression that can be figured out from the context that expression is in, should be left to context, and that a more precise expression should always, without exception, require more words to express than a similar, less precise one.

i'e

 

My reasoning behind both is simply that following those doctrines means we can say things in Lojban more succintly, a definite boon considering the wordiness the language has, IMO.

And it's not at all difficult to add {roroi fe'eroroi} to make the context be "always and everywhere". It does, obviously, require more words, but that's in line with the above.

sorry for asking questions not necessarily relevant to this discussion but can't we just say 

{ca ro da bu'u ro de} instead of {roroi fe'eroroi} or even {cajebu'u ro da} ?

I don't see how either of those could mean "always and everywhere: the whole of time and space"

Also, keep in mind that I'm specifically talking about, for example: {roroi fe'eroroi pa da zo'u da prenu} : "In the whole of time and space there is exactly one X, such that X is a person.", which is what {pa prenu} means if you /don't/ limit it to context. (If you do limit it to context, {pa prenu} means pretty much the same thing {pa lo prenu} does, as latro'a pointed out.)
 

[Ian Johnson]

The problem with this is that we don't have mechanisms for explicitly handling the universe of discourse. Anything you talk about is automatically bound within the hidden variable "the universe of discourse", and you can only indirectly influence what is in this domain. "Everything that has ever been at any location in the universe at any point in time" is implicitly "Everything that ... and is also in the universe of discourse" by default.

Another example of this is the approach to what outer quantifiers should mean. Assuming we've come to some agreement on what the universe of discourse is for the moment, should {ci da} mean "exactly three things" as the CLL proclaims? The way I understand what you're saying, you would think that it shouldn't, and instead there should be another PA for "exactly", and without that addition {ci da} should be something like "at least three, and probably not tremendously larger than three". At least as an outer quantifier; in {lo cacra be li ci} it would be more like "close to three, possibly with some error on either side". Am I right here?

My problem with making the "verbosity and precision correlate tightly" doctrine into law is that precise statements all vanish for being too verbose. It's like pedantic English, you can be very careful in English if you try really hard, but we don't speak that way, and those that try are criticized for sounding awkward.

mi'e la latro'a mu'o

You received this message because you are subscribed to the Google Groups "lojban" group.
To post to this group, send email to loj...@googlegroups.com.
To unsubscribe from this group, send email to lojban+un...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/lojban?hl=en.

[Ian Johnson]

An addition: the way I see it, there's a few levels of "universes":
The full linguistic universe. In the case of Lojban this universe is uselessly enormous, since we work with events as objects and so on. It contains parallel universes, events that never happen, people that will never be born, etc. I agree that putting {da} here is useless; in fact it probably makes {da broda} vacuous for any non-contradictory {broda}.
The "anything that might come up" universe. I think {da} should be somewhere around here, albeit probably slightly smaller than this. Basically, if you say {ro da broda}, and something comes up in conversation that doesn't broda, you spoke falsely. (I wouldn't say you "lied", because you might not have thought of that thing as being able to come up.) But the fact that I can make up a universe where something doesn't broda doesn't automatically mean you spoke falsely, in this model.
The "anything relevant" universe. This is tremendously smaller than both of the preceding. It means pretty much what you'd think, though it is worthy of comparison with the following.
The "some specific-ish relevant thing" "universe". This is not actually a universe, but rather my way of distinguishing between {da} and {zo'e} in the previous universe. While {ro da} there is everything relevant, {ro zo'e} could be just one thing, because it is allowed to be more concrete (in particular, it can have referents) than just "something in the universe" (which can't).

mi'e la latro'a mu'o

[Jonathan Jones]

On Thu, Jan 3, 2013 at 12:19 PM, Ian Johnson <blindb...@gmail.com> wrote:

The problem with this is that we don't have mechanisms for explicitly handling the universe of discourse. Anything you talk about is automatically bound within the hidden variable "the universe of discourse", and you can only indirectly influence what is in this domain. "Everything that has ever been at any location in the universe at any point in time" is implicitly "Everything that ... and is also in the universe of discourse" by default.

Another example of this is the approach to what outer quantifiers should mean. Assuming we've come to some agreement on what the universe of discourse is for the moment, should {ci da} mean "exactly three things" as the CLL proclaims? The way I understand what you're saying, you would think that it shouldn't, and instead there should be another PA for "exactly", and without that addition {ci da} should be something like "at least three, and probably not tremendously larger than three". At least as an outer quantifier; in {lo cacra be li ci} it would be more like "close to three, possibly with some error on either side". Am I right here?

No. I think it should mean exactly three. The difference is that I think "exactly three" should only be considered within context. If we're talking about the house next door, and I say {ci prenu} = {ci da poi prenu}, it should mean that there are exactly three people in the context of the house next door, not that there are exactly three people, in the whole of time and space, real and imagined, etc..., whatever the scope of "universe of discourse" is.
 

My problem with making the "verbosity and precision correlate tightly" doctrine into law is that precise statements all vanish for being too verbose. It's like pedantic English, you can be very careful in English if you try really hard, but we don't speak that way, and those that try are criticized for sounding awkward.

I'm not sure that's true, and I'm not certain that's really much of a problem if it is. I highly doubt that requiring more words to be more precise- which, by the way, is already true in the vast majority of Lojban as it is- automatically makes higher precision statements cease to be made. I also don't think your comparison with English is very on the ball, but that's neither here nor there.
 

[Jonathan Jones]

On Thu, Jan 3, 2013 at 12:34 PM, Ian Johnson <blindb...@gmail.com> wrote:

An addition: the way I see it, there's a few levels of "universes":
The full linguistic universe. In the case of Lojban this universe is uselessly enormous, since we work with events as objects and so on. It contains parallel universes, events that never happen, people that will never be born, etc. I agree that putting {da} here is useless; in fact it probably makes {da broda} vacuous for any non-contradictory {broda}.

^^ This is what seems to me be the "universe of discourse", and why I think anything that currently operates within that field should instead operate within the of context, where it would be much less useless.
 

The "anything that might come up" universe. I think {da} should be somewhere around here, albeit probably slightly smaller than this. Basically, if you say {ro da broda}, and something comes up in conversation that doesn't broda, you spoke falsely. (I wouldn't say you "lied", because you might not have thought of that thing as being able to come up.) But the fact that I can make up a universe where something doesn't broda doesn't automatically mean you spoke falsely, in this model.
The "anything relevant" universe. This is tremendously smaller than both of the preceding. It means pretty much what you'd think, though it is worthy of comparison with the following.

^^ My "universe" is determined by context, which I would consider as both of the above two, sort of.
 

The "some specific-ish relevant thing" "universe". This is not actually a universe, but rather my way of distinguishing between {da} and {zo'e} in the previous universe. While {ro da} there is everything relevant, {ro zo'e} could be just one thing, because it is allowed to be more concrete (in particular, it can have referents) than just "something in the universe" (which can't).

Honestly, I don't understand any of that, so I'll just ignore it.
 

[selpa'i]

la .aionys. cu cusku di'e

I'm not sure you're saying this, but "ci prenu cu broda" does not mean
that "there are exactly three people in the universe period", it means
"there are exactly three people in the universe that broda", which is a
big difference. Even without much context (though a certain tense and
place is almost always automatically assumed), such da-based statements
are not as extreme as some some seem to have claimed in the past and
also in this thread, since there is always an the additional restriction
that is the bridi they are contained in.

mu'o mi'e la selpa'i

[Jonathan Jones]

Your pretty much jives with what I mean about contextual determinism.

BUT, there's apparently disagreement as to the scope, since, to give a recent example, {ci prenu cu nanmu}, according to at least one person, means "there are exactly three people that are male in the entirety of the universe of discourse". The problem here is obviously what the universe of discourse is, which is where the disagreement seems to stem from.

I firmly assert that this "universe" should be determined by context, which has the effect of making {PA GISMU} roughly equivalent to {PA lo GISMU} in all cases.

Others seem to be of the view that the "universe" should be "the whole of everything, real and imagined, in any possible world, .....", which renders DA practically useless IMO.

And there are still other who think the universe should be somewhere between the two.

Tangentially relevant to this discussion, how would one say "in all possible worlds" anyway? I seem to remember there's a cmavo for this, but I can't recall what it is.
 

mu'o mi'e la selpa'i

--
You received this message because you are subscribed to the Google Groups "lojban" group.
To post to this group, send email to loj...@googlegroups.com.

To unsubscribe from this group, send email to lojban+unsubscribe@googlegroups.com.

For more options, visit this group at http://groups.google.com/group/lojban?hl=en.

[.arpis.]

Are you thinking about {ro mu'ei} (http://vlasisku.lojban.org/vlasisku/mu%27ei)?

To unsubscribe from this group, send email to lojban+un...@googlegroups.com.

For more options, visit this group at http://groups.google.com/group/lojban?hl=en.

--
mu'o mi'e .arpis.

[Jonathan Jones]

On Thu, Jan 3, 2013 at 4:29 PM, .arpis. <rpglover...@gmail.com> wrote:

Are you thinking about {ro mu'ei} (http://vlasisku.lojban.org/vlasisku/mu%27ei)?

Yes, thank you. I don't keep up very well with the experimental cmavo.
 

[Ian Johnson]

No. I think it should mean exactly three. The difference is that I think "exactly three" should only be considered within context. If we're talking about the house next door, and I say {ci prenu} = {ci da poi prenu}, it should mean that there are exactly three people in the context of the house next door, not that there are exactly three people, in the whole of time and space, real and imagined, etc..., whatever the scope of "universe of discourse" is.

This answer isn't really {na go'i} so much as {na'i}. I was saying to choose a fixed universe of discourse, then decide what an outer quantifier means. In other words I wanted to simplify the issue of outer quantifiers by decoupling it from the issue of determining the universe of discourse. You're saying that the universe of discourse shrinks and expands rapidly from sentence to sentence, and that the idea of holding one constant isn't reasonable. I think this is troubling in general, partly because everything done in logic depends on a background universe of discourse. In other words the first thing you do is say what you're talking about precisely (for example, the real number system), and then you start saying things about it. Rejecting this idea is sa'u illogical, but not necessarily irrational nor necessarily damaging.

 

My problem with making the "verbosity and precision correlate tightly" doctrine into law is that precise statements all vanish for being too verbose. It's like pedantic English, you can be very careful in English if you try really hard, but we don't speak that way, and those that try are criticized for sounding awkward.

I'm not sure that's true, and I'm not certain that's really much of a problem if it is. I highly doubt that requiring more words to be more precise- which, by the way, is already true in the vast majority of Lojban as it is- automatically makes higher precision statements cease to be made.

I phrased that the way I phrased it for a reason. Insisting that verbosity and precision correlate tightly makes it so every precise statement is necessarily verbose. I think part of the point of Lojban is that logic cleanly and succinctly expresses certain ideas (quantifiers, for example, and in particular ro, su'o, and to a lesser extent pa), and by reworking ordinary language carefully we can bring everyday ideas into that structure.

I also don't think your comparison with English is very on the ball, but that's neither here nor there.

It's an exaggeration, but I think the basic point is valid. That is, once you get to a certain level of verbosity, your listener tunes out. The threshold is probably higher in Lojban because the grammar is both simpler and less obtrusive, but it's still there.

[Jonathan Jones]

On Thu, Jan 3, 2013 at 4:47 PM, Ian Johnson <blindb...@gmail.com> wrote:

No. I think it should mean exactly three. The difference is that I think "exactly three" should only be considered within context. If we're talking about the house next door, and I say {ci prenu} = {ci da poi prenu}, it should mean that there are exactly three people in the context of the house next door, not that there are exactly three people, in the whole of time and space, real and imagined, etc..., whatever the scope of "universe of discourse" is.

This answer isn't really {na go'i} so much as {na'i}. I was saying to choose a fixed universe of discourse, then decide what an outer quantifier means. In other words I wanted to simplify the issue of outer quantifiers by decoupling it from the issue of determining the universe of discourse. You're saying that the universe of discourse shrinks and expands rapidly from sentence to sentence, and that the idea of holding one constant isn't reasonable. I think this is troubling in general, partly because everything done in logic depends on a background universe of discourse. In other words the first thing you do is say what you're talking about precisely (for example, the real number system), and then you start saying things about it. Rejecting this idea is sa'u illogical, but not necessarily irrational nor necessarily damaging.

I would say situation to situation, not sentence to sentence, because context isn't determined by what is said, but by the environment it is said in. Strictly speaking "environment" isn't really the right word, but you can't define context as "context".

And as far as logic wrt Lojban is concerned, Lojban is logical only in the sense that its grammar is built from predicate logic. Beyond that, it is no more nor less logical than any other language.
 

 

My problem with making the "verbosity and precision correlate tightly" doctrine into law is that precise statements all vanish for being too verbose. It's like pedantic English, you can be very careful in English if you try really hard, but we don't speak that way, and those that try are criticized for sounding awkward.

I'm not sure that's true, and I'm not certain that's really much of a problem if it is. I highly doubt that requiring more words to be more precise- which, by the way, is already true in the vast majority of Lojban as it is- automatically makes higher precision statements cease to be made.

I phrased that the way I phrased it for a reason. Insisting that verbosity and precision correlate tightly makes it so every precise statement is necessarily verbose. I think part of the point of Lojban is that logic cleanly and succinctly expresses certain ideas (quantifiers, for example, and in particular ro, su'o, and to a lesser extent pa), and by reworking ordinary language carefully we can bring everyday ideas into that structure.

I don't like the phrase "every precise statement is necessarily verbose". I would agree with "more precise statements are necessarily more verbose", however. A tight correlation means that the desired amount of precision would require a higher degree of verbosity, but not that every precise statement would be too verbose to actually be used, which is what it seems you are asserting.
 

I also don't think your comparison with English is very on the ball, but that's neither here nor there.

It's an exaggeration, but I think the basic point is valid. That is, once you get to a certain level of verbosity, your listener tunes out. The threshold is probably higher in Lojban because the grammar is both simpler and less obtrusive, but it's still there.

Well, certainly. I just don't think it's as big a problem as you seem to. Also, what I'm suggesting isn't a radical change from the way things are: the vast majority of Lojban does have a strong correlation between verbosity and precision, I'm just saying this correlation should be "all", not "most".
 

mi'e la latro'a mu'o

--

You received this message because you are subscribed to the Google Groups "lojban" group.
To post to this group, send email to loj...@googlegroups.com.
To unsubscribe from this group, send email to lojban+un...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/lojban?hl=en.

[Ian Johnson]

On Thu, Jan 3, 2013 at 7:08 PM, Jonathan Jones <eye...@gmail.com> wrote:

On Thu, Jan 3, 2013 at 4:47 PM, Ian Johnson <blindb...@gmail.com> wrote:

No. I think it should mean exactly three. The difference is that I think "exactly three" should only be considered within context. If we're talking about the house next door, and I say {ci prenu} = {ci da poi prenu}, it should mean that there are exactly three people in the context of the house next door, not that there are exactly three people, in the whole of time and space, real and imagined, etc..., whatever the scope of "universe of discourse" is.

This answer isn't really {na go'i} so much as {na'i}. I was saying to choose a fixed universe of discourse, then decide what an outer quantifier means. In other words I wanted to simplify the issue of outer quantifiers by decoupling it from the issue of determining the universe of discourse. You're saying that the universe of discourse shrinks and expands rapidly from sentence to sentence, and that the idea of holding one constant isn't reasonable. I think this is troubling in general, partly because everything done in logic depends on a background universe of discourse. In other words the first thing you do is say what you're talking about precisely (for example, the real number system), and then you start saying things about it. Rejecting this idea is sa'u illogical, but not necessarily irrational nor necessarily damaging.

I would say situation to situation, not sentence to sentence, because context isn't determined by what is said, but by the environment it is said in. Strictly speaking "environment" isn't really the right word, but you can't define context as "context".

The way you described it, it sounded like it would be perfectly

And as far as logic wrt Lojban is concerned, Lojban is logical only in the sense that its grammar is built from predicate logic. Beyond that, it is no more nor less logical than any other language.

No, it's not true that just its grammar is built from predicate logic. If that were literally true, its semantics could be not at all tied to predicate logic, whereas the definitions of things like {da} clearly reflect semantic ties to predicate logic. People use the term "grammar" in a terrible way, grammar alone is just structure, it doesn't mean anything. Instead they use "grammar" to refer to grammar in combination with "structural semantics", which are essentially invariants, aspects of semantics that persist within a given grammatical structure no matter what components you put in place. {lo broda cu broda} is one.

At any rate, I reject the premise that the grammar of Lojban is built from predicate logic only for it to completely throw out the semantic ideas of predicate logic at random. I think it's built this way so that it will be good at expressing the things that predicate logic is good at expressing. Quantifiers are one such thing, but only if you are reasonably careful about the universe of discourse.

 

 

My problem with making the "verbosity and precision correlate tightly" doctrine into law is that precise statements all vanish for being too verbose. It's like pedantic English, you can be very careful in English if you try really hard, but we don't speak that way, and those that try are criticized for sounding awkward.

I'm not sure that's true, and I'm not certain that's really much of a problem if it is. I highly doubt that requiring more words to be more precise- which, by the way, is already true in the vast majority of Lojban as it is- automatically makes higher precision statements cease to be made.

I phrased that the way I phrased it for a reason. Insisting that verbosity and precision correlate tightly makes it so every precise statement is necessarily verbose. I think part of the point of Lojban is that logic cleanly and succinctly expresses certain ideas (quantifiers, for example, and in particular ro, su'o, and to a lesser extent pa), and by reworking ordinary language carefully we can bring everyday ideas into that structure.

I don't like the phrase "every precise statement is necessarily verbose". I would agree with "more precise statements are necessarily more verbose", however. A tight correlation means that the desired amount of precision would require a higher degree of verbosity, but not that every precise statement would be too verbose to actually be used, which is what it seems you are asserting.
 

I also don't think your comparison with English is very on the ball, but that's neither here nor there.

It's an exaggeration, but I think the basic point is valid. That is, once you get to a certain level of verbosity, your listener tunes out. The threshold is probably higher in Lojban because the grammar is both simpler and less obtrusive, but it's still there.

Well, certainly. I just don't think it's as big a problem as you seem to. Also, what I'm suggesting isn't a radical change from the way things are: the vast majority of Lojban does have a strong correlation between verbosity and precision, I'm just saying this correlation should be "all", not "most".

The previous two paragraphs contradict one another.

[Ian Johnson]

The way you described it, it sounded like it would be perfectly

Didn't finish this part, sorry. I meant to say that it sounded like you thought it would be perfectly fine, if there were 5 people in a house all of whom have already been mentioned, to say {ci prenu cu zvati lo zdani}, since they had become the topic of discussion as of this sentence.

[Jonathan Jones]

On Thu, Jan 3, 2013 at 5:33 PM, Ian Johnson <blindb...@gmail.com> wrote:

On Thu, Jan 3, 2013 at 7:08 PM, Jonathan Jones <eye...@gmail.com> wrote:

On Thu, Jan 3, 2013 at 4:47 PM, Ian Johnson <blindb...@gmail.com> wrote:

No. I think it should mean exactly three. The difference is that I think "exactly three" should only be considered within context. If we're talking about the house next door, and I say {ci prenu} = {ci da poi prenu}, it should mean that there are exactly three people in the context of the house next door, not that there are exactly three people, in the whole of time and space, real and imagined, etc..., whatever the scope of "universe of discourse" is.

This answer isn't really {na go'i} so much as {na'i}. I was saying to choose a fixed universe of discourse, then decide what an outer quantifier means. In other words I wanted to simplify the issue of outer quantifiers by decoupling it from the issue of determining the universe of discourse. You're saying that the universe of discourse shrinks and expands rapidly from sentence to sentence, and that the idea of holding one constant isn't reasonable. I think this is troubling in general, partly because everything done in logic depends on a background universe of discourse. In other words the first thing you do is say what you're talking about precisely (for example, the real number system), and then you start saying things about it. Rejecting this idea is sa'u illogical, but not necessarily irrational nor necessarily damaging.

I would say situation to situation, not sentence to sentence, because context isn't determined by what is said, but by the environment it is said in. Strictly speaking "environment" isn't really the right word, but you can't define context as "context".

The way you described it, it sounded like it would be perfectly [...] fine, if there were 5 people in a house all of whom have already been mentioned, to say {ci prenu cu zvati lo zdani}, since they had become the topic of discussion as of this sentence.

Yeah, no. Sorry if that's the impression I gave. I suppose I should've emphasized the "context" bit more than I did.
 

And as far as logic wrt Lojban is concerned, Lojban is logical only in the sense that its grammar is built from predicate logic. Beyond that, it is no more nor less logical than any other language.

No, it's not true that just its grammar is built from predicate logic. If that were literally true, its semantics could be not at all tied to predicate logic, whereas the definitions of things like {da} clearly reflect semantic ties to predicate logic. People use the term "grammar" in a terrible way, grammar alone is just structure, it doesn't mean anything. Instead they use "grammar" to refer to grammar in combination with "structural semantics", which are essentially invariants, aspects of semantics that persist within a given grammatical structure no matter what components you put in place. {lo broda cu broda} is one.

At any rate, I reject the premise that the grammar of Lojban is built from predicate logic only for it to completely throw out the semantic ideas of predicate logic at random. I think it's built this way so that it will be good at expressing the things that predicate logic is good at expressing. Quantifiers are one such thing, but only if you are reasonably careful about the universe of discourse.

I'm not going to quibble about this. I'm apparently being much less capable of expressing myself than I'd like. I am in agreement with you wrt this specific issue.
 

My problem with making the "verbosity and precision correlate tightly" doctrine into law is that precise statements all vanish for being too verbose. It's like pedantic English, you can be very careful in English if you try really hard, but we don't speak that way, and those that try are criticized for sounding awkward.

I'm not sure that's true, and I'm not certain that's really much of a problem if it is. I highly doubt that requiring more words to be more precise- which, by the way, is already true in the vast majority of Lojban as it is- automatically makes higher precision statements cease to be made.

I phrased that the way I phrased it for a reason. Insisting that verbosity and precision correlate tightly makes it so every precise statement is necessarily verbose. I think part of the point of Lojban is that logic cleanly and succinctly expresses certain ideas (quantifiers, for example, and in particular ro, su'o, and to a lesser extent pa), and by reworking ordinary language carefully we can bring everyday ideas into that structure.

I don't like the phrase "every precise statement is necessarily verbose". I would agree with "more precise statements are necessarily more verbose", however. A tight correlation means that the desired amount of precision would require a higher degree of verbosity, but not that every precise statement would be too verbose to actually be used, which is what it seems you are asserting.
 

I also don't think your comparison with English is very on the ball, but that's neither here nor there.

It's an exaggeration, but I think the basic point is valid. That is, once you get to a certain level of verbosity, your listener tunes out. The threshold is probably higher in Lojban because the grammar is both simpler and less obtrusive, but it's still there.

Well, certainly. I just don't think it's as big a problem as you seem to. Also, what I'm suggesting isn't a radical change from the way things are: the vast majority of Lojban does have a strong correlation between verbosity and precision, I'm just saying this correlation should be "all", not "most".

The previous two paragraphs contradict one another.

Huh? Which two? The immediately preceding two, or the previous two of mine? Either way, I don't see where the contradiction is.
 

mi'e la latro'a mu'o

--

You received this message because you are subscribed to the Google Groups "lojban" group.
To post to this group, send email to loj...@googlegroups.com.
To unsubscribe from this group, send email to lojban+un...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/lojban?hl=en.