BPFK: Old Modal Aspects

From Lojban
Jump to: navigation, search

 forum_name: Modal Aspects
topic_title: shepherd volunteer
    subject: shepherd volunteer
   username: And
  post_time: 2003-06-03 15:24:16
  post_text:

I volunteer to shepherd this.


 forum_name: Modal Aspects
topic_title: shepherd volunteer
    subject:
   username: nitcion
  post_time: 2003-06-03 23:30:12
  post_text:

Woah.

OK, you got it. I think you'll do a good job of it...


 forum_name: Modal Aspects
topic_title: shepherd volunteer
    subject:
   username: And
  post_time: 2003-06-04 16:00:19
  post_text:
Woah. OK, you got it. I think you'll do a good job of it...

Better "woah" than "whoa". Feel free, people, to nag me to start off discussion if you get impatient. I've got a lot of wiki-poring to do before I can summarize the issues.


 forum_name: Modal Aspects
topic_title: Meaning of {ka'e}
    subject: Meaning of {ka'e}
   username: And
  post_time: 2003-06-05 08:47:42
  post_text:

The gloss of {ka'e} as "innately capable of" cannot be wholly correct, because "innately capable of" is a relationship between X and a property. But CAhA does not work thus, as can be seen most easily when applied to a selbri with more than one sumti place. In translating {ka'e} as "can", we mean it in the sense used in (1) rather than (2).

(1) The mud can get 18 inches deep in this field.

   = "It can happen that the mud gets 18 inches deep..."

(2) I can run a mile in just under 5 minutes.

   = "I have the ability to run a mile...."

If {ka'e} does not (quite) mean "innately capable of", what does it mean? Some sort of possibility, obviously, but why then the non-obvious & somewhat misleading glosses? I suggest the purpose of the glosses is to allow things like (3) while excluding things like (4).

(3) The mud could have been as much as 18 inches deep yesterday.

   [On a reading different from (4): "It could have happened that...", "Nothing intrinsic to the way the world is prevented the mud from being deep".]

(4) The mud may have been as much as 18 inches deep yesterday.

   [On a Standard English reading: "I don't know that the mud wasn't deep".]

In technical terminology, {ka'e} expresses 'root' rather than 'epistemic' modality. That is, it's not about how (im)perfect the speaker's knowledge is; it's about dividing up alternative states of the world into those that are possible and those that are impossible, where 'possible' means something like 'consistent with the way the world actually is'.

I suggest that the above statement is as consistent with CLL and usage as possible.


 forum_name: Modal Aspects
topic_title: Meaning of {ca'a} and {pu'i}
    subject: Meaning of {ca'a} and {pu'i}
   username: And
  post_time: 2003-06-05 08:50:23
  post_text:

The following pattern is both necessary/desirable in a modal paradigm & partly evident in CAhA.

+possible +/-actual = ka'e +possible -actual = nu'o +possible +actual = ca'a

But what then does {pu'i} mean? (Or, what is the difference between {pu'i} and {ca'a}?) {ca'a} is glossed as "actually is" and {pu'i} as "can and has". This seems to be conflating modality with tense/aspect. What is the difference between "John ca'a attended the last logfest" and "John pu'i attended the last logfest"? I can't see any. I conclude that there is no valid difference between {ca'a} and {pu'i}.

{ca'a} has seen a lot of usage. {pu'i} hasn't (I say impressionistically). I would propose that {pu'i} be abolished on the grounds of redundancy.


 forum_name: Modal Aspects
topic_title: Implicit CAhA
    subject: Implicit CAhA
   username: And
  post_time: 2003-06-05 08:51:36
  post_text:

In Lojban, (1) is equivalent to (2) rather than being neutral between (2) and (3) and requiring glorking to determine whether it is (2) or (3) that is being claimed.

(1) ko'a broda (2) ko'a ja'a broda (3) ko'a na broda

In contrast, (4) is not equivalent to (5), but is neutral between (5) and (6) (and (7)?), and requires glorking to determine which of them is being claimed.

(4) ko'a broda (5) ko'a ca'a broda (6) ko'a ka'e broda (7) ko'a nu'o broda

This means that from (4) we can infer (6) without glorking, but have to glork in order to decide whether the stronger claim of (5) or (7) is being made. In a sense, then, the encoded meaning of (4) is equivalent to (6). We would use (6) to make it clear only that we are not trying to imply (5) or (7).

To the best of my knowledge, (4) *in usage* is always used to mean (5). (With the possible exception of when {nu} is involved -- but that's an issue for NU.) Thus there is a conflict between the prescription and usage.

I suggest that the prescription should be changed to make (4) equivalent to (5). This reflects usage and the fact that we more often wish to talk about things that actually broda than things that could broda. That would make Lojban more like natural languages, where 'actually' is the unmarked and 'possibly' the marked. (OTOH, avid whorfians might appreciate the whorfian mind-f*ck of reversing that.) Under this proposed change, {ca'a} becomes analogous to {ja'a} not only phonetically but also functionally, in that it explicitly affirms the default.


 forum_name: Modal Aspects
topic_title: Meaning of {ka'e}
    subject:
   username: nitcion
  post_time: 2003-06-05 08:56:06
  post_text:

A shameless, biassed, self-serving, yet I think fair comment from me: you both buy your possible-world ka'e and stick with baseline. Cool.

My concern with the extant wording was actually how "innately capable" could transfer to inanimates and abstractions. But while I'd obviously like ka'e to be "possible", let's ensure whether there is anything in "possible" not covered by "innately capable", and therefore whether this constitutes a reinterpretation or a change to the baseline.


 forum_name: Modal Aspects
topic_title: Meaning of {ca'a} and {pu'i}
    subject: Re: Meaning of {ca'a} and {pu'i}
   username: nitcion
  post_time: 2003-06-05 09:02:22
  post_text:
{ca'a} has seen a lot of usage. {pu'i} hasn't (I say impressionistically). I would propose that {pu'i} be abolished on the grounds of redundancy.

Or at the least, defended explicitly: this is a Supplication issue (as in, "what on earth were you thinking when you created this?") You, And, will need to do searches for all pu'i that have ever been used, and see if they are saying something that ca'a wouldn't be. I agree that conflating tense and modality is undesirable: pu'i should merely not be equivalent to caca'a or ba'oca'a. The prescription is, I believe, "actualised potential"; we would need some clarification as to what makes ca'a other than "actualised potential", that is not already conveyed by ba'oca'a.

If no convincing distinction can be made, and none has emerged in usage, then we can consider turfing or redefining pu'i; but of course, not before a usage survey, and a chance for Lojban Central to make an explicit case.

(If we were able to reallocate it, there is all sorts of good modal use we could put it to; but as with all reallocation, the case would need to be strong, and pu'i would need to have seen next to no usage.)


 forum_name: Modal Aspects
topic_title: Implicit CAhA
    subject:
   username: nitcion
  post_time: 2003-06-05 09:08:22
  post_text:

I suspect that supplication will indeed say something whorfian, and to anticipate the argument likely to be made, one piece of research necessary is: are there natlangs where irrealis is morphologically unmarked? Is that possible with Chinese, or is that merely hype? If and where it is possible, is realisood at least a default implicature, with irrealishood bound to particular expressions and contexts?

Is p. 243 wrong about {ro datka cu flulimna} because of a confusion of individuals and generics? This issue may well end up bound to gadri.

Hm. Not sure about this one.


 forum_name: Modal Aspects
topic_title: Meaning of {ka'e}
    subject:
   username: And
  post_time: 2003-06-05 09:13:16
  post_text:

As I see it, the main problem with "innately capable" is that if "ka'e broda fa ko'a fe ko'e", what is it that has the innate capability?

I think my proposal involves a change to the documentation of the prescription, but not to the underlying intent.


 forum_name: Modal Aspects
topic_title: Implicit CAhA
    subject: Chinese: semantics without syntax
   username: JohnCowan
  post_time: 2003-06-05 12:15:34
  post_text:

The notorious Chinese sentence in Woldy ("If you saw my sister you'd know she's pregnant"/"If you were to see my sister you would know she was pregnant") is straight out of Li & Thompson's Mandarin Chinese: A Functional Reference Grammar, and as such may be presumed sound. Randy LaPolla says that even subject/objecthood is only pragmatic in Ch., so why not realis/irrealis?

"People talk about what they wish to talk about, and they do not think about what they see no need to think about."


 forum_name: Modal Aspects
topic_title: Meaning of {ca'a} and {pu'i}
    subject: A more complete paradigm of CAhAs
   username: JohnCowan
  post_time: 2003-06-05 15:19:32
  post_text:

 +possible    +actual:    ca'a
 +possible  +/-actual:    ka'e
 +possible   -actual:     nu'o

+/-possible +actual: reduces to +possible +actual +/-possible +/-actual: no'eka'e +/-possible -actual: no'eka'ejenu'o?

 -possible    +actual:    self-contradictory
 -possible  -/+actual:    reduces to -possible -actual
 -possible   -actual:     to'eka'e

This pattern suggests to me that pu'i is indeed bogus, but that ka'e needs scalar and not contradictory negation.


 forum_name: Modal Aspects
topic_title: Meaning of {ca'a} and {pu'i}
    subject: Re: A more complete paradigm of CAhAs
   username: rlpowell
  post_time: 2003-06-05 15:45:19
  post_text:

+possible +actual: ca'a +possible +/-actual: ka'e +possible -actual: nu'o +/-possible +actual: reduces to +possible +actual +/-possible +/-actual: no'eka'e +/-possible -actual: no'eka'ejenu'o? -possible +actual: self-contradictory -possible -/+actual: reduces to -possible -actual -possible -actual: to'eka'e

This pattern suggests to me that pu'i is indeed bogus, but that ka'e needs scalar and not contradictory negation.

Seems to me that +/-possible +/-actual should merely be the default. Or if not that, a default should be assigned explicitely regardless (I'm a big fan of ca'a as the default, for clarity of communication, but that would contradict the red book).

As for +/-possible -actual, what about no'enu'o ?

-Robin


 forum_name: Modal Aspects
topic_title: Meaning of {ca'a} and {pu'i}
    subject: By the way...
   username: rlpowell
  post_time: 2003-06-05 15:56:36
  post_text:

That chart (probably with more verbosity) is *exactly* the kind of cmavo clarification that Lojban needs. It gives me the feeling that the BPFK is doing a Good Thing.

-Robin


 forum_name: Modal Aspects
topic_title: Meaning of {ca'a} and {pu'i}
    subject: Re: A more complete paradigm of CAhAs
   username: xorxes
  post_time: 2003-06-05 16:42:30
  post_text:

+possible +actual: ca'a +possible +/-actual: ka'e +possible -actual: nu'o +/-possible +actual: reduces to +possible +actual +/-possible +/-actual: no'eka'e +/-possible -actual: no'eka'ejenu'o? -possible +actual: self-contradictory -possible -/+actual: reduces to -possible -actual -possible -actual: to'eka'e

{no'eka'ejenu'o} reduces to {nu'o}.

If {no'e}/{to'e} touch only the 'possible' and not the 'actual', then we have:

 +possible    +actual:    ca'a
 +possible  +/-actual:    ka'e
 +possible    -actual:    nu'o

+/-possible +actual: no'eca'a [reduces to ca'a] +/-possible +/-actual: no'eka'e [tautological] +/-possible -actual: no'enu'o

 -possible    +actual:    to'eca'a [contradictory]
 -possible  -/+actual:    to'eka'e [reduces to to'enu'o]
 -possible    -actual:    to'enu'o

But I don't know how justified that would be.


 forum_name: Modal Aspects
topic_title: Implicit CAhA
    subject: Re: Implicit CAhA
   username: rlpowell
  post_time: 2003-06-05 19:21:53
  post_text:

In contrast, (4) is not equivalent to (5), but is neutral between (5) and (6) (and (7)?), and requires glorking to determine which of them is being claimed.

(4) ko'a broda (5) ko'a ca'a broda (6) ko'a ka'e broda (7) ko'a nu'o broda

...snip...

I suggest that the prescription should be changed to make (4) equivalent to (5). This reflects usage and the fact that we more often wish to talk about things that actually broda than things that could broda.

.iesai

-Robin


 forum_name: Modal Aspects
topic_title: IRC POV
    subject: IRC POV
   username: rlpowell
  post_time: 2003-06-05 19:25:37
  post_text:

The following conversation occured when this was discussed on IRC. I am "#lojban>".

  1. lojban> You know, http://www.lojban.org/phpbb/viewtopic.php?p=365#365 really has a point. pu'i does look amazingly useless.

<xod> has, but is not right now? <xod> do pu'i gletu <xod> I think it's useful <xod> an answer to xu ka'e <tsali> xod: pujecanai? <xod> it's the same difference betwee ca and ca'a

  1. lojban> xod, that's pu ca'a. You're conflating tense with aspect.

<xod> what's pu ca'a? <xod> pu'i == pu ca'a?

  1. lojban> No, "has, but is not right now" == pu ca'a

<tsali> xod: that *seems* logically equivalent, yes. <tsali> rlpowell: bzzzt, wrong. Lojban is aorist.

  1. lojban> aorist?

<tsali> Yes, pu broda does not entail canai broda. <tsali> And vice versa and for all tenses.

  1. lojban> Oh, sorry.
  2. lojban> canai ca'a then.

<xod> canai ca'a?? <xod> do canai ca'a mo <tsali> mi canai ca'a klama fu le relcukmarce <xod> pu'i go'i <tsali> go'i <xod> je'e <xod> .ua <xod> but then what is the difference between ca'a and pu'i?

  1. lojban> None, that's the point.

<tsali> rlpowell: maybe that "ca'a" is unmarked as to the innate capability? <xod> not capable of, yet does?

  1. lojban> Heh heh heh.

...snip... <tsali> If CAhA is scalar, doesn't that mean that CAhA+NAI falls? <xod> probably

  1. lojban> Not necessarily; it would just be == to'e ca'a

<tsali> You mean the +/-foo? That's very... linguistic.

  1. lojban> Yes, but it's *important*.

<tsali> It would be very cool with a dictionary that had en entire chart like that in the definition of a word.

  • rlpowell nods.

<tsali> No natlang dictionary has that sort of thing.


 forum_name: Modal Aspects
topic_title: How much more to discuss?
    subject: How much more to discuss?
   username: And2
  post_time: 2003-06-09 08:53:03
  post_text:

There has been lots of discussion on the wiki & Jboske on issues pertaining to the same semantic area covered by CAhA. As far as they pertain to the BF, I think these issues mostly boil down to whether the BF wants to even consider adopting any new cmavo in order to express stuff that can't currently be expressed by means of cmavo.

Is there any point in my writing a summary of the candidate functions/meanings for such new cmavo? It clearly wouldn't count as mending the current prescription, and nor would it count as providing ways to say the unsayable (for anything can be said with brivla).

({mu'ei}, with ROI syntax & CAhA semantics, is a different case, because it has seen a fair deal of usage, so needs to be considered because of that.)

If we don't want to discuss new cmavo, then all I have to add to the CAhA discussion is an uncontroversial note on CAhA as sumtcita.

Please let me know your views, especially people who will kick up a stink if things aren't done the way they want them to be done.


 forum_name: Modal Aspects
topic_title: How much more to discuss?
    subject: Re: How much more to discuss?
   username: rlpowell
  post_time: 2003-06-09 13:28:20
  post_text:
There has been lots of discussion on the wiki & Jboske on issues pertaining to the same semantic area covered by CAhA. As far as they pertain to the BF, I think these issues mostly boil down to whether the BF wants to even consider adopting any new cmavo in order to express stuff that can't currently be expressed by means of cmavo. Is there any point in my writing a summary of the candidate functions/meanings for such new cmavo? It clearly wouldn't count as mending the current prescription, and nor would it count as providing ways to say the unsayable (for anything can be said with brivla).

Unfortunately, I can't answer the question without having at least some idea of what it is you'd write up. 8(

Sorry.

-Robin


 forum_name: Modal Aspects
topic_title: How much more to discuss?
    subject:
   username: nitcion
  post_time: 2003-06-09 21:30:16
  post_text:

new cmavo are within the ambit of the bpfk, since they count as "how do you say this", and the alternative is humpty-dumptying existing cmavo. We may or may not decide that we don't want Lojban to encode X, but it needs to be an informed decision. And additive change does not intrinsically damage backward compatibility, the way subtractive does.

After all, a Unique gadri would definitely be additive.

So you have my blessing. Is And2 the evil twin, btw? :)


 forum_name: Modal Aspects
topic_title: How much more to discuss?
    subject:
   username: And
  post_time: 2003-06-09 22:01:32
  post_text:

Okay. I'll post the list of How-do-you-say-its in due course when time permits.

So you have my blessing. Is And2 the evil twin, btw? :)

I couldn't log on as And, for some reason, so had to create new membership. I'm sure it was something I did wrong & not a glitch in anything Robin set up.


 forum_name: Modal Aspects
topic_title: How much more to discuss?
    subject:
   username: noras
  post_time: 2003-06-10 19:38:38
  post_text:
new cmavo are within the ambit of the bpfk, since they count as "how do you say this", and the alternative is humpty-dumptying existing cmavo. We may or may not decide that we don't want Lojban to encode X, but it needs to be an informed decision. And additive change does not intrinsically damage backward compatibility, the way subtractive does.

...

I think it wise, however, not to make any real decision on the proposed new cmavo until all the existing cmavo are better defined. The maneuverings of meaning for existing cmavo may obviate the need for the proposed one(s).


 forum_name: Modal Aspects
topic_title: CAhA-related meanings
    subject: CAhA-related meanings
   username: And
  post_time: 2003-06-11 13:36:56
  post_text:

This is a very terse & schematic attempt to summarize key meanings & ideas that have arisen in previous discussion.

A. Real versus imaginary

1. x1 is the world in which x2 occurs 2. x1 is true of world x2; x2 is a world of which x1 is true

(1-2) allow "Sherlock is a detective" to be clearly true (if claimed to be true of the world in which Sherlock occurs) or clearly false (if claimed of the world in which dei ['this utterance'] occurs). Likewise it allows us to distinguish between "every detective" including and excluding Sherlock. It is reasonable that the default is that this information about which world the predication is true of should be glorked, but it is needful to have an explicit glorkless alternative.


B. mu'ei "PA mu'ei zo'e broda" means: "the proposition 'broda' is true of PA (relevant/potential) worlds of which zo'e is true". I.e. it expresses the predicate "x1 is true of PA worlds of which x2 is true". Thus, mu'ei is a shorthand for a logical formula based on predicate (2) (with the added element of relevance/potentiality).


C. ba'oi Like mu'ei, but retricts worlds to those that share an identical past with the local world. Useful for future conditionals ("If it rains tomorrow, then..."). However, judicious use of {ba} (in combination with {mu'ei}) can obviate the need for {ba'oi}, I think.


D. CAhA ka'e = su'o mu'ei = "is true of some (potential) worlds of which zo'e is true"

ca'a and nu'o make additional claims about actuality:

ca'a = "is true of the local world and of some (potential) worlds of which zo'e is true" nu'o = "is false of the local world and true of some (potential) worlds of which zo'e is true"

The zo'e in these formulas corresponds to the sumti tagged by the CAhA.


E. Added actuality claims If it is desirable to conflate claims of actuality and potentiality (as with ca'a/nu'o), then mu'ei leads to the following expanded paradigm: 3a "x1 is true of PA worlds of which x2 is true" 3b "x1 is true of the local world and x1 is true of PA worlds of which x2 is true" 3c "x1 is false of the local world and x1 is true of PA worlds of which x2 is true" 4a "x1 is true of PA worlds of which x2 is true, and x2 is true of the local world" 4b "x1 is true of the local world and x1 is true of PA worlds of which x2 is true, and x2 is true of the local world" 4c "x1 is false of the local world and x1 is true of PA worlds of which x2 is true, and x2 is true of the local world" 5a "x1 is true of PA worlds of which x2 is true, and x2 is false of the local world" 5b "x1 is true of the local world and x1 is true of PA worlds of which x2 is true, and x2 is false of the local world" 5c "x1 is false of the local world and x1 is true of PA worlds of which x2 is true, and x2 is false of the local world"


F. bi'ai ka'e broda = su'o mu'ei broda = "if zo'e then it could be true that broda"

Symmetry calls for:

bi'ai broda = ro mu'ei broda = "if zo'e then it would be true that broda"


G. Summary If we were (re)creating modals from scratch, it seems certain that we would not end up with just ka'e/ca'a/nu'o. If su'omu'ei gets ka'e as a single cmavo equivalent, then why should there not also be bi'ai? If ka'e has ca'a/nu'o as (non)actuality counterparts, then why not also have 6 more counterparts analogous to 4a-c and 5a-c, and why not have these counterparts also for bi'ai and mu'ei? However, this is the BF and we're not recreating modals from scratch. The fact that ka'e/ca'a/nu'o are abitrarily selected for lexicalization from a much larger unlexicalized paradigm is not reason for the BF to get rid of them or to lexicalize the rest of the paradigm.


 forum_name: Modal Aspects
topic_title: Logical scope of CAhA
    subject: Logical scope of CAhA
   username: xorxes
  post_time: 2003-06-11 15:37:29
  post_text:

Which of the following is {roda ka'e na broda} equivalent to?

1. naku ka'eku roda broda 2. naku roda ka'eku broda 3. ka'eku naku roda broda 4. ka'eku roda naku broda 5. roda naku ka'eku broda 6. roda ka'eku naku broda

I think cases can be made for 1, 2, 3 and 6, based on different aspects of the official prescription.


 forum_name: Modal Aspects
topic_title: Logical scope of CAhA
    subject: Re: Logical scope of CAhA
   username: And
  post_time: 2003-06-11 18:12:52
  post_text:
Which of the following is {roda ka'e na broda} equivalent to?

1. naku ka'eku roda broda 2. naku roda ka'eku broda 3. ka'eku naku roda broda 4. ka'eku roda naku broda 5. roda naku ka'eku broda 6. roda ka'eku naku broda

I think cases can be made for 1, 2, 3 and 6, based on different

aspects of the official prescription.

By my knowledge of the prescription, the official meaning is 2, but that is an abomination that I hope the BF will rectify (i.e. overriding the prescription for the sake of our sanity). The meaning should be 6.

I'm curious how you get 1 & especially 3 from the prescription.


 forum_name: Modal Aspects
topic_title: Logical scope of CAhA
    subject: Re: Logical scope of CAhA
   username: xorxes
  post_time: 2003-06-11 18:55:26
  post_text:
Which of the following is {roda ka'e na broda} equivalent to?

1. naku ka'eku roda broda 2. naku roda ka'eku broda 3. ka'eku naku roda broda 4. ka'eku roda naku broda 5. roda naku ka'eku broda 6. roda ka'eku naku broda

I think cases can be made for 1, 2, 3 and 6, based on different

aspects of the official prescription.

By my knowledge of the prescription, the official meaning is 2, but that is an abomination that I hope the BF will rectify (i.e. overriding the prescription for the sake of our sanity). The meaning should be 6.

I'm curious how you get 1 & especially 3 from the prescription.

Well, I always assumed that the precedence of na was a special case of the general precedence of the selbri tcita (of which na is a part) over all terms. I thought that was the official prescription until the recent discussion in jboske. I'm not sure whether I got that from some official source, but that's how I understood it for years, so that my answer until recently would have been 3 for the official prescription: selbri tcita first, followed by terms.

1 would be obtained by a rule that takes na first, then the rest of the selbri tcita, then the terms.

The rule that plucks na out of its position within the selbri tcita is so crazy that I still can't believe it is the official one. I don't find much in CLL that deals with this issue, but there is an interesting comment on page 384 (Ch.15 Sect.8):

8.5) la djan. [cu] na ba klama la paris. .e la rom. John [false] later-will-go-to [both] Paris and Rome. It is false that John will go to Paris and Rome.

8.7) la djan. [cu] ba na klama la paris. .e la rom. John later-will [false] go-to [both] Paris and Rome.

We stated in Section 3 that sentences like Example 8.5 and Example 8.7 appear to be semantically identical, but that subtle semantic distinctions may eventually be found.

Now, if a distinction can be found between {ba na} and {na ba}, it will indeed be subtle, but if the door is open for such distinction then they are not equivalent and na can't be just plucked out of its position within the tag. The distinction between {ka'e na} and {na ka'e}, on the other hand, would not at all be subtle, but CLL doesn't say anywhere, as far as I can tell, that they are equivalent.

In any case, I agree that 6 is the sensible meaning, but if not 6, at least one that preserves the {ka'eku naku} (= possibly not) order.


 forum_name: Modal Aspects
topic_title: Implicit CAhA
    subject: Re: Chinese: semantics without syntax
   username: nitcion
  post_time: 2003-06-21 02:57:19
  post_text:
The notorious Chinese sentence in Woldy ("If you saw my sister you'd know she's pregnant"/"If you were to see my sister you would know she was pregnant") is straight out of Li & Thompson's Mandarin Chinese: A Functional Reference Grammar, and as such may be presumed sound. Randy LaPolla says that even subject/objecthood is only pragmatic in Ch., so why not realis/irrealis?

Unfair counterexamples, John. if there's an "if" there, the possibilty of irrealis is there; the same if there is a generic referent (as I'd contend "ducks" is.) But is it at all possible that, for example, the standalone sentence

.i la fred. ca jatna le merko

can *ever* be interpreted as

.i la fred nu'o ca jatna le merko

?

If not --- and I believe not --- then we have to say when an irrealis interpretation is licit. My guess is, only with generic referents, conditionals, and other overt linguistic signals. The current unmarkedness of irrealis is not what people are going to do in reality --- they will assume a realis default where they can. I doubt even Mandarin does in all sentences, but my copy of Li & Thompson is at work...


 forum_name: Modal Aspects
topic_title: Implicit CAhA
    subject: Re: Chinese: semantics without syntax
   username: And
  post_time: 2003-06-21 08:53:34
  post_text:
The notorious Chinese sentence in Woldy ("If you saw my sister you'd know she's pregnant"/"If you were to see my sister you would know she was pregnant") is straight out of Li & Thompson's Mandarin Chinese: A Functional Reference Grammar, and as such may be presumed sound. Randy LaPolla says that even subject/objecthood is only pragmatic in Ch., so why not realis/irrealis?

Unfair counterexamples, John. if there's an "if" there, the possibilty of irrealis is there; the same if there is a generic referent (as I'd contend "ducks" is.) But is it at all possible that, for example, the standalone sentence

.i la fred. ca jatna le merko

can *ever* be interpreted as

.i la fred nu'o ca jatna le merko

?

If not --- and I believe not --- then we have to say when an irrealis interpretation is licit. My guess is, only with generic referents, conditionals, and other overt linguistic signals. The current unmarkedness of irrealis is not what people are going to do in reality --- they will assume a realis default where they can. I doubt even Mandarin does in all sentences, but my copy of Li & Thompson is at work...

I'd meant to post a similar reply, but forgot. That said, one can get, in English, things like:

"You see my sister and you know she's pregnant." "He disses me one more time and I'll have his guts for garters."

But anyway, I'm not yet convinced that CAhA does neatly map to realis/irrealis. E.g. irrealis doesn't have the restriction to potentiality.


 forum_name: Modal Aspects
topic_title: Implicit CAhA
    subject: Re: Chinese: semantics without syntax
   username: nitcion
  post_time: 2003-06-21 12:06:21
  post_text:
I'd meant to post a similar reply, but forgot. That said, one can get, in English, things like:

"You see my sister and you know she's pregnant."

"He disses me one more time and I'll have his guts for garters."

Both the intonation and the particular construction (2nd/3rd person present coordinated with 1st person future) are unambiguous for the second sentence as a conditional marker. You can't do that with "I dissed you and you had my guys for garters". Note too that the constructions use the simple present --- which is gnomic and generic in effect; not the progressive. So these are still irrealishoods licensed by the linguistic context. Lojban on paper allows irrealis for every single possible unmarked utterance, and I just don't buy it.

But anyway, I'm not yet convinced that CAhA does neatly map to realis/irrealis. E.g. irrealis doesn't have the restriction to potentiality.

Because of possible vs. impossible worlds? But talk of impossible worlds is pretty forced in natural language...


 forum_name: Modal Aspects
topic_title: Implicit CAhA
    subject: Re: Chinese: semantics without syntax
   username: And
  post_time: 2003-06-21 18:05:57
  post_text:
I'd meant to post a similar reply, but forgot. That said, one can get, in English, things like:

"You see my sister and you know she's pregnant."

"He disses me one more time and I'll have his guts for garters."

Both the intonation and the particular construction (2nd/3rd person present coordinated with 1st person future) are unambiguous for the second sentence as a conditional marker. You can't do that with "I dissed you and you had my guys for garters". Note too that the constructions use the simple present --- which is gnomic and generic in effect; not the progressive. So these are still irrealishoods licensed by the linguistic context. Lojban on paper allows irrealis for every single possible unmarked utterance, and I just don't buy it.

But anyway, I'm not yet convinced that CAhA does neatly map to realis/irrealis. E.g. irrealis doesn't have the restriction to potentiality.
Because of possible vs. impossible worlds? But talk of impossible worlds is pretty forced in natural language...

In a sentence like "If X then Y", X will be irrealis (e.g. subjunctive), but not paraphrasable by {ka'e}. Rather, {ka'e broda} is most closely translated as "could be broda", which is much more specific than general irrealis. I can live with the idea that the general realis/irrealis distinction is glorked, but not that the distinction is glorked between, on the one hand, "that ko'a is/be/were broda" and, on the other hand, "that ko'a could be broda".


 forum_name: Modal Aspects
topic_title: Implicit CAhA
    subject: If = negation!
   username: JohnCowan
  post_time: 2003-06-23 12:55:50
  post_text:

The reason that things under "if" are irrealis is because (as Lojban tells us) there is a negation involved, and nothing can be more irrealis than negation. "If you saw my sister, you would know she was pregnant" = "Either you do not see my sister, or you know that she is pregnant."


 forum_name: Modal Aspects
topic_title: Implicit CAhA
    subject: Re: If = negation!
   username: And
  post_time: 2003-06-23 21:07:13
  post_text:
The reason that things under "if" are irrealis is because (as Lojban tells us) there is a negation involved, and nothing can be more irrealis than negation. "If you saw my sister, you would know she was pregnant" = "Either you do not see my sister, or you know that she is pregnant."

(Is this an elegant jest? Or a real argument?)

McCawley has convincing reasons why "if X then Y" does not mean "either not X or Y". "Whether" also creates an irrealis context without "not".


 forum_name: Modal Aspects
topic_title: Implicit CAhA
    subject: Re: If = negation!
   username: xod
  post_time: 2003-06-23 23:32:49
  post_text:
McCawley has convincing reasons why "if X then Y" does not mean "either not X or Y".

This isn't the forum, but as a vociferous defender of ganai, gi for if, then, and having deemed every opposing argument to be misguided though reasonable, I'm interested in this.


 forum_name: Modal Aspects
topic_title: Implicit CAhA
    subject: Re: If = negation!
   username: And
  post_time: 2003-06-24 08:33:49
  post_text:
McCawley has convincing reasons why "if X then Y" does not mean "either not X or Y".
This isn't the forum, but as a vociferous defender of ganai, gi for if, then, and having deemed every opposing argument to be misguided though reasonable, I'm interested in this.

I'll hunt through McCawley, unless someone beats me to it. (The relevant info is on the wiki.) This debate exists not just within Lojban: the reason McCawley discusses it is that he was disputing the mainstream view that English IF = 'logical IF' = either not P or Q.

Note that in comparing ways to do IF, we are comparing non-brivla methods, because brivla methods can do anything. As for how to do IF without brivla, "mu'ei" does "if P then it could/would be that Q", and ka'e/bi'ai respectively do could/would versions.


 forum_name: Modal Aspects
topic_title: Implicit CAhA
    subject:
   username: JohnCowan
  post_time: 2003-06-24 11:24:45
  post_text:

When I talk of if-then, I of course mean the Lojban version, which is equivalent to "either not ... or ...".


 forum_name: Modal Aspects
topic_title: Implicit CAhA
    subject:
   username: And
  post_time: 2003-06-24 14:55:30
  post_text:
When I talk of if-then, I of course mean the Lojban version, which is equivalent to "either not ... or ...".

Ah. Okay. Then I'm not wholly sure I get your point. But a negated clause is not necessarily irrealis. However, it could be that OR makes things irrealis: "Either it be safe or it be too hazardous to venture". In which case, I take your point, if we make it about OR rather than NOT!


 forum_name: Modal Aspects
topic_title: "CAhA le du" versus "CAhA zi'o"
    subject: "CAhA le du" versus "CAhA zi'o"
   username: And2
  post_time: 2003-06-25 10:45:33
  post_text:

There is an important difference in meaning between "CAhA le du" (or CAhA + mostly anything but zi'o) and "CAhA zi'o".

1a. ka'e le du mi cliva "If a certain state of affairs obtained, then I could leave." "In some possible worlds in which a certain state of affairs obtains, I leave." "I could leave."

1b. ka'e zi'o mi cliva "In some possible worlds, I leave." "It is possible that I leave." "I can leave."

To see the difference, contrast:

2a. If I were a woman, I could fondle my breasts. [true] 2b. It is possible for me to fondle my breasts. [false] 3a. If I were a woman, I could father a child. [false] 3b. It is possible for me to father a child. [true]

The difference comes out more clearly if we shift from ka'e (= su'o mu'ei) to ro mu'ei:

4a. ro mu'ei le du mi cliva "If a certain state of affairs obtained, then I would leave." "In all possible worlds in which a certain state of affairs obtains, I leave." "I would leave."

4b. ro mu'ei zi'o mi cliva "In all possible worlds, I leave." "It is necessarily the case that I leave. "I will leave."

Here are some possible positions one could take wrt the above.

A. implicit zo'e includes zi'o. CAhA broda = CAhA zo'e broda. Therefore CAhA broda is unspecified wrt the a/b distinction (can vs could), and is resolved through glorking. One can disambiguate by using "CAhA zi'o" and "CAhA le du".

B. implicit zo'e does not include zi'o. CAhA broda = CAhA zo'e broda. Therefore CAhA broda has meaning (a) ('could', not 'can'). One can get (b) ('can') by using "CAhA zi'o".

C. CAhA broda != CAhA zo'e broda. Therefore CAhA broda is equivalent to CAhA zi'o broda, and has meaning (b) ('can', not 'could'). One can get meaning (a) ('could') by using "CAhA le du".

Discussion please...


 forum_name: Modal Aspects
topic_title: "CAhA le du" versus "CAhA zi'o"
    subject: Re: "CAhA le du" versus "CAhA zi'o"
   username: xorxes
  post_time: 2003-06-25 11:12:33
  post_text:

Here are some possible positions one could take wrt the above.

A. implicit zo'e includes zi'o. CAhA broda = CAhA zo'e broda. Therefore CAhA broda is unspecified wrt the a/b distinction (can vs could), and is resolved through glorking. One can disambiguate by using "CAhA zi'o" and "CAhA le du".

B. implicit zo'e does not include zi'o. CAhA broda = CAhA zo'e broda. Therefore CAhA broda has meaning (a) ('could', not 'can'). One can get (b) ('can') by using "CAhA zi'o".

C. CAhA broda != CAhA zo'e broda. Therefore CAhA broda is equivalent to CAhA zi'o broda, and has meaning (b) ('can', not 'could'). One can get meaning (a) ('could') by using "CAhA le du".

Discussion please...

Couldn't "CAhA zi'o" be characterized also as "CAhA le cabna", or "CAhA some situation that actually obtains"? If so, then B is not really an option. We could say that the default is C ("CAhA le cabna") in the same sense that "PU le cabna" is default (i.e. it can be defeated by certain contexts). I think the conservative stand requires "can" rather than "could" as default.


 forum_name: Modal Aspects
topic_title: "CAhA le du" versus "CAhA zi'o"
    subject: Re: "CAhA le du" versus "CAhA zi'o"
   username: And
  post_time: 2003-07-06 07:04:37
  post_text:
Couldn't "CAhA zi'o" be characterized also as "CAhA le cabna", or

"CAhA some situation that actually obtains"? If so, then B is not really an option. We could say that the default is C ("CAhA le cabna") in the same sense that "PU le cabna" is default (i.e. it can be defeated by certain contexts). I think the conservative stand requires "can" rather

than "could" as default.

Not hearing any views to the contrary, I will (as shepherd) take this to be the BF position on the matter.


 forum_name: Modal Aspects
topic_title: SE CAhA: grammar change?
    subject: SE CAhA: grammar change?
   username: And
  post_time: 2003-07-11 14:19:32
  post_text:

Consider 1a-b:

1a. ki'u gi broda gi brode

    "Because broda, brode."

1b. se ki'u gi brode gi broda

    "Brode, because broda."

And 2a-b:

2a. ka'e gi broda gi brode

    "If broda then perhaps brode."

2b. *se ka'e gi brode gi broda

    "Perhaps brode, if broda."

2b is ungrammatical, but symmetry makes that ungrammaticality an anomaly.

Should SE CAhA be made grammatical? (I would say yes.) If so, then how? I would suggest simply abolishing CAhA and moving its contents into BAI. (Incidentally, that would make CAhA NAI grammatical... And it would have the trivial & irrelevant virtue of making less egregious the current practise of describing BAI as 'modals'.)


 forum_name: Modal Aspects
topic_title: SE CAhA: grammar change?
    subject:
   username: kd
  post_time: 2003-07-11 14:45:31
  post_text:

I would support this if you can tell me what the following would mean:

3. ca'a le broda ku brode

I would like to point out that And's comments about this rendering CAhA NAI grammatical ought not scare off the anti-ka'enai camp - it would remove the ambiguity of what the nai meant in that context. It should absolutely not be supported simply because of this, but I support anything that simplifies the grammar once it has been shown that it doesn't break anything that isn't already broken.


 forum_name: Modal Aspects
topic_title: SE CAhA: grammar change?
    subject:
   username: And
  post_time: 2003-07-11 14:56:12
  post_text:
I would support this if you can tell me what the following would mean: 3. ca'a le broda ku brode

"If le broda is the case, then it could be the case (and in fact is the case) that brode".

Note that (3) is *already* grammatical (I cannot locate the BF page where its meaning is discussed, but I'm sure I did one).

I would like to point out that And's comments about this rendering CAhA NAI grammatical ought not scare off the anti-ka'enai camp - it would remove the ambiguity of what the nai meant in that context. It should absolutely not be supported simply because of this, but I support anything that simplifies the grammar once it has been shown that it doesn't break anything that isn't already broken.

I agree with all these points.


 forum_name: Modal Aspects
topic_title: SE CAhA: grammar change?
    subject:
   username: xorxes
  post_time: 2003-07-11 15:03:19
  post_text:

I am definitely in favour of this kind of simplification of the grammar.

One possible problem I can see is that KAhE can be combined with tense while BAI can't, so for example {pu ka'e broda} would no longer parse in the way it parses now. (It would still parse as {puku ka'e broda}. Of course, the current rule that {pu bai broda} doesn't parse like {pu ka'e broda} makes no sense to me.


 forum_name: Modal Aspects
topic_title: SE CAhA: grammar change?
    subject:
   username: And
  post_time: 2003-07-11 17:39:21
  post_text:

I have realized that my argument in favour of SE CAhA generalizes to SE + tag. I.e. wherever we have "tag gi broda gi brode", we would like also to have "se tag gi brode gi broda".

Hence just moving CAhA into BAI would be an insufficiently general solution. A more general solution, but a more substantial change to the grammar, would be to redefine stag as "[SE] simple-tense-modal ... (etc.)" and in the rule for simple-tense-modal change "[SE] BAI" to plain "BAI". (I am referring to EBNF rules, here.)

All the same, there do seem to be unnecessarily many types of tag, and much scope for simplification (such as moving CAhA and other selmaho into BAI). I do think the BF should weigh simplification (for the benefit of future learners) against conservatism (especially, not invalidating contents of CLL). Indeed, this is such an important point that I urge Nick to consider polling the larger community (beyond the BF commissioners) on it. That is, given a change that would (a) invalidate no usage, (b) making learning the grammar substantially simpler, and (c) invalidate the relevant bit of CLL, are current lojbanists for or against it.


 forum_name: Modal Aspects
topic_title: SE CAhA: grammar change?
    subject:
   username: nitcion
  post_time: 2003-07-12 11:05:02
  post_text:
I am definitely in favour of this kind of simplification of the grammar.

Predictably, I'm opposed, and regard it as tinkering. (As you know, And has the "if Jorge supports it, it must be right" rule; I've come to believe the opposite :-) .)

We can have the poll And calls for, but (a) you won't like the outcome; and anything but a 90% majority leads to an unacceptable risk of schism, which I shall not allow; (b) I don't feel like redoing the bpfk charter, and I'm making it purposefully difficult to do so --- after all, the bpfk does have a mandate now, and the vote was cast on the present charter; (c) we have repeatedly seen that such polls have failed on this forum, even among us commissioners. There are no broad policy decisions, only specific issues: the broad issues are the same for the grammatical status of Y and CAhA --- and you already know how Jay feels about the former, and how it is highly unlikely he feels the same about the latter.

And, I may have been pissed off with what Bob said to me about the BPFK needing to prove to conservatives that it isn't just a reformist factory, but I am obliged to say the same to you now. Per my edict on baseline violation, you are free to set up a proposal, and put it to side for final vote. But there is a rather more pressing issue here which you *must* address if ka'e is to be meaningfully used as a sumtcita at all --- namely, is the CLL's "innately capable" the same as "possible world", and if so, what is the status of pu'i, and what does it mean that ca'a doesn't. That's a simple, elemental point that needs clarification yesterday. (I know you've started on it, and John seems amenable to it, but we need a decision, so this is where we need the poll.) se ka'e, in my book, is a luxury, and too late for us to implement.

Now, that said: additive rules have a lesser threshold than subtractive rules: we *can* add SE CAhA, as a separate rule. But I'm reluctant to do so by simplifying the grammar and effacing the distinctions between tags. Yes, the distinctions are pointless, noone will learn them anyway, and in practice people will use their analogy whether the grammar allows it or not. But no, we do not have a mandate to revise the grammar where it is not demonstrably broken. I keep saying this; and Jay was quite assertive about it the last time it came up. Do you really want to go through this again? And given what was said here, do you think you can get this mandate outside the BPFK? Do you think the Silent Majority is *likelier* to be reformist than the commissioners?

Anyway. The rest goes to meta. We can have this poll; but I see no reason to think it will succeed; and a plurality just ain't going to cut it with this kind of issue.


 forum_name: Modal Aspects
topic_title: SE CAhA: grammar change?
    subject:
   username: And
  post_time: 2003-07-12 19:11:42
  post_text:

fuck. I've accidentally edited nick's message instead of replying to it. Can't retrieve the original via back button.

OK -- have now pasted in the email version of Nick's original.


 forum_name: Modal Aspects
topic_title: SE CAhA: grammar change?
    subject:
   username: And
  post_time: 2003-07-12 19:14:25
  post_text:
I am definitely in favour of this kind of simplification of the grammar.

Predictably, I'm opposed, and regard it as tinkering. (As you know, And has the "if Jorge supports it, it must be right" rule; I've come to believe the opposite :-) .)

But is it benign tinkering? -- I can't predict what opinion on that would be. Is making Lojban easier to learn worth invalidating bits of CLL -- when usage & prior learning would be unaffected?

We can have the poll And calls for, but (a) you won't like the outcome; and anything but a 90% majority leads to an unacceptable risk of schism, which I shall not allow; (b) I don't feel like redoing the bpfk charter, and I'm making it purposefully difficult to do so --- after all, the bpfk does have a mandate now, and the vote was cast on the present charter; (c) we have repeatedly seen that such polls have failed on this forum, even among us commissioners.

There are no broad policy decisions, only specific issues: the broad issues are the same for the grammatical status of Y and CAhA --- and you already know how Jay feels about the former, and how it is highly unlikely he feels the same about the latter.

Tinkerings that only xorxes & I & a few others support are different from ones that might command far more widespread support. Part of the antipathy to tinkering is based on the belief that a small handful of maniacs are interfering with the wishes of the great majority.

And, I may have been pissed off with what Bob said to me about the BPFK needing to prove to conservatives that it isn't just a reformist factory, but I am obliged to say the same to you now. Per my edict on baseline violation, you are free to set up a proposal, and put it to side for final vote. But there is a rather more pressing issue here which you *must* address if ka'e is to be meaningfully used as a sumtcita at all --- namely, is the CLL's "innately capable" the same as "possible world", and if so, what is the status of pu'i, and what does it mean that ca'a doesn't. That's a simple, elemental point that needs clarification yesterday. (I know you've started on it, and John seems amenable to it, but we need a decision, so this is where we need the poll.) se ka'e, in my book, is a luxury, and too late for us to implement.

I haven't put anything to a poll, because there have been no disagreements so far. I have stated my conclusions about "innately capable" and "pu'i". I am not going to put these or others to a poll unless I am either specifically asked to or it is apparent to me that there is a difference of opinion that can't be resolved by further discussion.

As new issues pertaining to CAhA occur to me, I am adding them to the phpbb. There is no backlog of issues I haven't got round to dealing with,

Now, that said: additive rules have a lesser threshold than subtractive rules: we *can* add SE CAhA, as a separate rule. But I'm reluctant to do so by simplifying the grammar and effacing the distinctions between tags. Yes, the distinctions are pointless, noone will learn them anyway, and in practice people will use their analogy whether the grammar allows it or not. But no, we do not have a mandate to revise the grammar where it is not demonstrably broken. I keep saying this; and Jay was quite assertive about it the last time it came up. Do you really want to go through this again? And given what was said here, do you think you can get this mandate outside the BPFK? Do you think the Silent Majority is *likelier* to be reformist than the commissioners?

I do, actually, if the commissioners accepted it as a benign reform.


 forum_name: Modal Aspects
topic_title: "CAhA le du" versus "CAhA zi'o"
    subject: Re: "CAhA le du" versus "CAhA zi'o"
   username: nitcion
  post_time: 2003-07-14 11:39:45
  post_text:
Not hearing any views to the contrary, I will (as shepherd) take this to be the BF position on the matter.

The use you're putting zi'o to in the above really worries me; after all, this modal use of ka'e is quite new, and there's no real precedent for pu zi'o either. I agree with xorxes' conclusion, though; anything but this will lead to utterly unacceptable breaking of backward compatibility. It just seems to me you're really talking about ka'e zu'i, and that ka'e zi'o and pu zi'o are meaningless: the possibility and tense are always relative to something, and the default something is the here-and-now. (So in fact, pu = pu le nau tcini, ka'e = ka'e le nau tcini).


 forum_name: Modal Aspects
topic_title: CAhA-related meanings
    subject:
   username: nitcion
  post_time: 2003-07-14 12:05:02
  post_text:

Might I kibbitz some issues for polling then? (Both baseline-maintaining and -violating.)

(A) The CLL statement of ka'e as innate possbility should be shifted to the more global notion of possible world, as this is understood in 20th century philosophy.

(B) pu'i should be deprecated. (Like I said, we must have a usage survey before we can decide on this.)

(C) We need a bridi for worlds in which things are true, so that we can do prenex statements of worlds without having to resort to mu'ei. (jetnyvanbi, say.)

(D) We should have a Necessarily True modal (bi'ai), analogous to ka'e.

The necessity of Necessarily True is blindingly obvious to anyone who has ever done modal logic, and an embarrassing absence in a purportedly logical language.

(E) We should make mu'ei or some equivalent official.

Now, if we have (C) and (D), I'm not convinced we actually need (E). As we know, possible worlds can differ infinitesimally from each other. so you can never have a discrete number of them: the only cardinalities of possible worlds that ever get used are: (none), (some), and (all). If we have CAhA corresponding to all three (naka'e, ka'e, bi'ai), then frankly I don't see the point of (nomu'ei, su'omu'ei, romu'ei). They have the advantage of analogy with {noda su'ode rodi zo'u); but then using ROI for these always struck me as kludgy, and logicians straight out quantify possible worlds in prenexes the old fashioned way: {su'o jetnyvanbi zo'u}.

(F) We need an equivalent of {ba'oi}

Not convinced of this either. It is not terribly painful to say {baka'e lenu carvi kei} --- with the underlying assumption that the pu and ca are in the realm of ca'a, so I guess spelled out as {pujecaca'a le nau jetnyvanbi baka'e lenu carvi kei}. In fact, this serves a good paedagogical purpose: it makes explicit, in a way that ba'oi does not, that ifs are all about a world whose past is what we know, but whose future is openended. If we can avoid ba'oi, I think we should.

(G) We should have lexicalisations of 4a-c, 5a-c

I, uh, don't really get these, and I'd need compelling examples before I approved them. Six more modals is wasteful anyway, we should have more economic solutions.

My minimalist counterproposal is just bi'ai, I think, though fully accepting the possible world interpretation of CAhA. The Founders must now answer whether they regard "innate possibility" as distinct from possible worlds in any way, and whether this is a distinction worth preserving. And they should avoid talking about ducks; I want to know what innate possibility means when applied to rocks.

I think And and John and I :-) are right that pu'i is a dangerous conflation of aspect and modality that doesn't belong in the paradigm, and should be deprecated. In an ideal world, it would be reused for bi'ai, but this is not an ideal world, so at most it will be freed, and retained as a synonym of ba'oca'a for backward compatibility.

In casting out mu'ei, I am casting out something that many Lojbanists have grown attached to; I will not object much to it being adopted as well. But if we admit the su'omu'ei = ka'e equivalence and introduce bi'ai, then there isn't much left for mu'ei to do: it's not harmful, but it is pretty much redundant. If (by some attack of fundamentalism I happen to be opposed to this time) we do not end up accepting that su'omu'ei = ka'e, then the introduction of mu'ei or a similar expression in the language is a matter of urgency.


 forum_name: Modal Aspects
topic_title: Meaning of {ca'a} and {pu'i}
    subject: Re: A more complete paradigm of CAhAs
   username: nitcion
  post_time: 2003-07-14 12:19:13
  post_text:
{no'eka'ejenu'o} reduces to {nu'o}.

no'eka'ejanu'o then. But yes, we should do better.

If {no'e}/{to'e} touch only the 'possible' and not the 'actual', then we have:

+possible +actual: ca'a +possible +/-actual: ka'e +possible -actual: nu'o +/-possible +actual: no'eca'a [reduces to ca'a] +/-possible +/-actual: no'eka'e [tautological] +/-possible -actual: no'enu'o -possible +actual: to'eca'a [contradictory] -possible -/+actual: to'eka'e [reduces to to'enu'o] -possible -actual: to'enu'o

But I don't know how justified that would be.

Yeah. It's stupendously elegant, and it just doesn't feel right; it's certainly an idiosyncratic way of negating. But we can define this into being anyway; we can say that this shall be the Lojbanic scale of modality.

The real elements are:

 +possible    +actual:    ca'a
 +possible  +/-actual:    ka'e
 +possible    -actual:    nu'o

+/-possible +/-actual: no'eka'e [tautological] +/-possible -actual: no'enu'o

 -possible    -actual:    to'enu'o

to'eka'e feels more right than to'enu'o, but in your scheme the two are equivalent anyway. The distinction between no'eka'e and no'enu'o is the one that takes getting used to; I say for expediency's sake we get used to it anyway...


 forum_name: Modal Aspects
topic_title: Meaning of {ca'a} and {pu'i}
    subject: Re: A more complete paradigm of CAhAs
   username: xorxes
  post_time: 2003-07-16 17:27:08
  post_text:
{no'eka'ejenu'o} reduces to {nu'o}.

no'eka'ejanu'o then. But yes, we should do better.

{no'eka'ejanu'o} reduces to {no'eka'e}. :)

If {no'e}/{to'e} touch only the 'possible' and not the 'actual', then we have:

+possible +actual: ca'a +possible +/-actual: ka'e +possible -actual: nu'o +/-possible +actual: no'eca'a [reduces to ca'a] +/-possible +/-actual: no'eka'e [tautological] +/-possible -actual: no'enu'o -possible +actual: to'eca'a [contradictory] -possible -/+actual: to'eka'e [reduces to to'enu'o] -possible -actual: to'enu'o

But I don't know how justified that would be.

Yeah. It's stupendously elegant, and it just doesn't feel right; it's certainly an idiosyncratic way of negating. But we can define this into being anyway; we can say that this shall be the Lojbanic scale of modality.

There might be a problem, though. NAhE doesn't always apply to CAhA directly. The rule is:

simple-tense-modal = [NAhE] (time [space] | space [time]) & CAhA [KI]

so who knows on what scale NAhE is playing there.


 forum_name: Modal Aspects
topic_title: "CAhA le du" versus "CAhA zi'o"
    subject: Re: "CAhA le du" versus "CAhA zi'o"
   username: And
  post_time: 2003-07-17 05:15:44
  post_text:
Not hearing any views to the contrary, I will (as shepherd) take this to be the BF position on the matter.
The use you're putting zi'o to in the above really worries me; after all, this modal use of ka'e is quite new, and there's no real precedent for pu zi'o either. I agree with xorxes' conclusion, though; anything but this will lead to utterly unacceptable breaking of backward compatibility. It just seems to me you're really talking about ka'e zu'i, and that ka'e zi'o and pu zi'o are meaningless: the possibility and tense are always relative to something, and the default something is the here-and-now. (So in fact, pu = pu le nau tcini, ka'e = ka'e le nau tcini).

"le nau tcini" may not quite capture it, because the meaning "can" is not deictic; it is not tied to the here and now of the discourse. But that's a quibble, & we can agree that the intended meaning is "that what is the case is the case".

However, the reason I formalized this in terms of zi'o is as follows. If the elided sumti is a zo'e, then by definition the value of zo'e is glorked from context. In the present case that would mean that selbri tcita {ka'e} is neutral between "can" and "could". To force selbri tcita {ka'e} to have only the "can" meaning, we must get rid of the zo'e. (The same goes for {pu}.)

Now, I have no strong feelings about whether selbri tcita {ka'e} should mean "can" or be neutral between "can"/"could". But I do maintain that the way to get the former is by the zi'o method, while the zo'e method results in the latter.


 forum_name: Modal Aspects
topic_title: CAhA-related meanings
    subject:
   username: And
  post_time: 2003-07-17 05:52:04
  post_text:
Might I kibbitz some issues for polling then? (Both baseline-maintaining and -violating.)

(A) The CLL statement of ka'e as innate possbility should be shifted to the more global notion of possible world, as this is understood in 20th century philosophy.

Oughtn't we to have some idea of what anybody opposed to (A) would actually propose? The current alternative to (A) is nothing.

(B) pu'i should be deprecated. (Like I said, we must have a usage survey before we can decide on this.)

I had forgotten the survey. I will try to get round to posting it. All the exx I found seem to use it as per the ma'oste, i.e. "can and has".

(C) We need a bridi for worlds in which things are true, so that we can do prenex statements of worlds without having to resort to mu'ei. (jetnyvanbi, say.)

A brivla, you mean? We need *some* way of saying it, but if it's not done by cmavo then I think it's outside the scope of the BF.

(D) We should have a Necessarily True modal (bi'ai), analogous to ka'e.

The necessity of Necessarily True is blindingly obvious to anyone who has ever done modal logic, and an embarrassing absence in a purportedly logical language.

The problem then is that we get an ugly gappy paradigm:

ka'e   bi'ai
ca'a   ??
nu'o   ??

(E) We should make mu'ei or some equivalent official.

Now, if we have (C) and (D), I'm not convinced we actually need (E). As we know, possible worlds can differ infinitesimally from each other. so you can never have a discrete number of them: the only cardinalities of possible worlds that ever get used are: (none), (some), and (all).

Also, 'most' and '(proportionally) many/few', and 'almost all', 'all but a few', and so forth. I have seen {so'e mu'ei} used (on IRC).

If we have CAhA corresponding to all three (naka'e, ka'e, bi'ai), then frankly I don't see the point of (nomu'ei, su'omu'ei, romu'ei). They have the advantage of analogy with {noda su'ode rodi zo'u); but then using ROI for these always struck me as kludgy, and logicians straight out quantify possible worlds in prenexes the old fashioned way: {su'o jetnyvanbi zo'u}.

{ka'e} and {bi'ai} aren't enough.

Of course we can do virtually anything using an extremely minimal syntactic apparatus. It is only the exceptional cmavo that can't be paraphrased by brivla. In creating a logical language, everything comes down to issues of convenience: a minimalist grammar, which is trivial to create, makes utterances unfeasibly cumbersome, and the nontrivial challenge is to find ways to expand the grammar that reduce the cumbersomeness.

(F) We need an equivalent of {ba'oi}

Not convinced of this either. It is not terribly painful to say {baka'e lenu carvi kei} --- with the underlying assumption that the pu and ca are in the realm of ca'a, so I guess spelled out as {pujecaca'a le nau jetnyvanbi baka'e lenu carvi kei}.

I'd go for "ka'e lenu ba carvi" as paraphrase. I haven't thought about this enough, but I don't think we should adopt ba'oi without better understood reasons.

(G) We should have lexicalisations of 4a-c, 5a-c

I, uh, don't really get these, and I'd need compelling examples before I approved them. Six more modals is wasteful anyway, we should have more economic solutions.

It's just an argument about paradigm symmetry. Modals express a relation between two states of affairs, p and q. {ca'a} and {nu'o} single out one particular relation ('possibility') and one particular argument (the apodosis) to be marked for its actuality.

This is completely arbitrary. But as I said in the message you're responding to, the BF's mission is to repair brokenness, not mere ugliness and arbitrariness.

In casting out mu'ei, I am casting out something that many Lojbanists have grown attached to; I will not object much to it being adopted as well. But if we admit the su'omu'ei = ka'e equivalence and introduce bi'ai, then there isn't much left for mu'ei to do:

Except for the several examples I gave above. If we adopt mu'ei, then there's no point in adopting bi'ai (other than brevity and partial symmetry), and we could instead preserve ka'e as merely a redundant relic.


 forum_name: Modal Aspects
topic_title: CAhA-related meanings
    subject:
   username: xorxes
  post_time: 2003-07-17 09:40:33
  post_text:

(D) We should have a Necessarily True modal (bi'ai), analogous to ka'e.

The necessity of Necessarily True is blindingly obvious to anyone who has ever done modal logic, and an embarrassing absence in a purportedly logical language.

The problem then is that we get an ugly gappy paradigm:

ka'e   bi'ai
ca'a   ??
nu'o   ??

bi'ai doesn't have ca'a and nu'o equivalents because bi'ai entails ca'a and is incompatible with nu'o, so I don't think these gaps are particularly bad.

(G) We should have lexicalisations of 4a-c, 5a-c

I, uh, don't really get these, and I'd need compelling examples before I approved them. Six more modals is wasteful anyway, we should have more economic solutions.

It's just an argument about paradigm symmetry. Modals express a relation between two states of affairs, p and q. {ca'a} and {nu'o} single out one particular relation ('possibility') and one particular argument (the apodosis) to be marked for its actuality.

This is completely arbitrary. But as I said in the message you're responding to, the BF's mission is to repair brokenness, not mere ugliness and arbitrariness.

There should probably be a UI that marks for actuality, maybe there is one already. Then mu'ei and that UI would cover all cases, I think.

In casting out mu'ei, I am casting out something that many Lojbanists have grown attached to; I will not object much to it being adopted as well. But if we admit the su'omu'ei = ka'e equivalence and introduce bi'ai, then there isn't much left for mu'ei to do:

Except for the several examples I gave above. If we adopt mu'ei, then there's no point in adopting bi'ai (other than brevity and partial symmetry), and we could instead preserve ka'e as merely a redundant relic.

Indeed I have used {romu'ei} many times, and {bi'ai} none, though it is nice that it exists for the sake of symmetry.


 forum_name: Modal Aspects
topic_title: CAhA-related meanings
    subject:
   username: And
  post_time: 2003-07-17 10:14:42
  post_text:

(D) We should have a Necessarily True modal (bi'ai), analogous to ka'e.

The necessity of Necessarily True is blindingly obvious to anyone who has ever done modal logic, and an embarrassing absence in a purportedly logical language.

The problem then is that we get an ugly gappy paradigm:

ka'e   bi'ai
ca'a   ??
nu'o   ??

bi'ai doesn't have ca'a and nu'o equivalents because bi'ai entails ca'a and is incompatible with nu'o, so I don't think these gaps are particularly bad.

Not so. Given {bi'ai gi broda gi brode}, "ca'a brode" is entailed if broda is true. But if broda is not true, then "nu'o broda" is also possible.

(G) We should have lexicalisations of 4a-c, 5a-c

I, uh, don't really get these, and I'd need compelling examples before I approved them. Six more modals is wasteful anyway, we should have more economic solutions.

It's just an argument about paradigm symmetry. Modals express a relation between two states of affairs, p and q. {ca'a} and {nu'o} single out one particular relation ('possibility') and one particular argument (the apodosis) to be marked for its actuality.

This is completely arbitrary. But as I said in the message you're responding to, the BF's mission is to repair brokenness, not mere ugliness and arbitrariness.

There should probably be a UI that marks for actuality, maybe there is one already. Then mu'ei and that UI would cover all cases, I think.

I agree, except I'm not certain UI does the trick. Maybe it does; I don't know. At any rate, its scope matters, and in lojban a word normally has scope over what it precedes. That's not how UI works, normally.


 forum_name: Modal Aspects
topic_title: CAhA-related meanings
    subject:
   username: xorxes
  post_time: 2003-07-17 10:32:58
  post_text:

bi'ai doesn't have ca'a and nu'o equivalents because bi'ai entails ca'a and is incompatible with nu'o, so I don't think these gaps are particularly bad.

Not so. Given {bi'ai gi broda gi brode}, "ca'a brode" is entailed if broda is true. But if broda is not true, then "nu'o broda" is also possible.

I meant that {bi'ai gi broda gi brode} entails {ca'a gi broda gi brode} and is incompatible with {nu'o gi broda gi brode}. Or, {bi'ai broda} entails {ca'a broda} and is incompatible with {nu'o broda}.


There should probably be a UI that marks for actuality, maybe there is one already. Then mu'ei and that UI would cover all cases, I think.

I agree, except I'm not certain UI does the trick. Maybe it does; I don't know. At any rate, its scope matters, and in lojban a word normally has scope over what it precedes. That's not how UI works, normally.

UI has scope over the single word that precedes it, but if this word is the head or closes some structure, then UI has scope over that whole structure. So it would seem that in something like {ka'e gi UI ... gi UI ...}, the UIs would have the desired scope.


 forum_name: Modal Aspects
topic_title: CAhA-related meanings
    subject:
   username: xorxes
  post_time: 2003-07-17 14:55:12
  post_text:

I meant that {bi'ai gi broda gi brode} entails {ca'a gi broda gi brode} and is incompatible with {nu'o gi broda gi brode}. Or, {bi'ai broda} entails {ca'a broda} and is incompatible with {nu'o broda}.

Ok, perhaps I was too hasty with that. The first part holds only if brode is true of the actual world, and the second part if the default complement is something true of the actual world.


 forum_name: Modal Aspects
topic_title: CAhA-related meanings
    subject:
   username: And
  post_time: 2003-07-17 18:25:48
  post_text:

bi'ai doesn't have ca'a and nu'o equivalents because bi'ai entails ca'a and is incompatible with nu'o, so I don't think these gaps are particularly bad.

Not so. Given {bi'ai gi broda gi brode}, "ca'a brode" is entailed if broda is true. But if broda is not true, then "nu'o broda" is also possible.

I meant that {bi'ai gi broda gi brode} entails {ca'a gi broda gi brode} and is incompatible with {nu'o gi broda gi brode}.

But this is incorrect. The ca'a version says brode is the case. The nu'o version says brode isn't the case. The bi'ai version makes no claim about whether brode is the case, and can be true even if brode is not the case.

I'm not sure where we're misunderstanding one another.

Or, {bi'ai broda} entails {ca'a broda} and is incompatible with {nu'o broda}.

Maybe, depending on what we decide selbri tcita CAhA means.


 forum_name: Modal Aspects
topic_title: Meaning of {ca'a} and {pu'i}
    subject: Re: A more complete paradigm of CAhAs
   username: nitcion
  post_time: 2003-07-18 05:47:39
  post_text:
There might be a problem, though. NAhE doesn't always apply to CAhA directly. The rule is:

simple-tense-modal = [NAhE] (time [space] | space [time]) & CAhA [KI]

so who knows on what scale NAhE is playing there.

This time, I'm not going to defend a misfeature: a NAhE with ambiguous scope? Unconscionable. We want a scalar CAhA, and if the grammar doesn't give us an unambiguous scalar, then the grammar shall change: end of story. In fact, since we want a scalar CAhA anyway...

... I'm now switching to supporting CAhA NAI.

Those who attack CAhA NAI are now obligated to convince me either that we don't want a scalar CAhA, or to defend the ambiguity in this grammar rule. Remember, just because it's established, doesn't mean it still doesn't have to be defended. Over to you, Bob.


 forum_name: Modal Aspects
topic_title: Meaning of {ca'a} and {pu'i}
    subject: Re: A more complete paradigm of CAhAs
   username: And
  post_time: 2003-07-18 18:12:16
  post_text:
This time, I'm not going to defend a misfeature: a NAhE with ambiguous scope? Unconscionable. We want a scalar CAhA, and if the grammar doesn't give us an unambiguous scalar, then the grammar shall change: end of story. In fact, since we want a scalar CAhA anyway...

... I'm now switching to supporting CAhA NAI.

Those who attack CAhA NAI are now obligated to convince me either that we don't want a scalar CAhA, or to defend the ambiguity in this grammar rule. Remember, just because it's established, doesn't mean it still doesn't have to be defended. Over to you, Bob.

I'm happy to agree that the BF must determine how the scope of NAhE is determined, but I don't see clearly what the case for scalar CAhA is. I'm not disputing that there is such a case; I'm just asking for you to spell it out a bit more explicitly.


 forum_name: Modal Aspects
topic_title: Meaning of {ca'a} and {pu'i}
    subject: Re: A more complete paradigm of CAhAs
   username: nitcion
  post_time: 2003-07-19 10:39:16
  post_text:
I'm happy to agree that the BF must determine how the scope of NAhE is determined, but I don't see clearly what the case for scalar CAhA is. I'm not disputing that there is such a case; I'm just asking for you to spell it out a bit more explicitly.

What am I missing? +possible = ka'e, +/- possible (i.e. unmarked) is no'eka'e, -possible is to'eka'e. The minute we use to'e and no'e, we have switched to scalar negation.


 forum_name: Modal Aspects
topic_title: Meaning of {ca'a} and {pu'i}
    subject: Re: A more complete paradigm of CAhAs
   username: And
  post_time: 2003-07-19 12:38:16
  post_text:
I'm happy to agree that the BF must determine how the scope of NAhE is determined, but I don't see clearly what the case for scalar CAhA is. I'm not disputing that there is such a case; I'm just asking for you to spell it out a bit more explicitly.
What am I missing? +possible = ka'e, +/- possible (i.e. unmarked) is no'eka'e, -possible is to'eka'e. The minute we use to'e and no'e, we have switched to scalar negation.

But what exactly do you want CAhA+NAI for, that can't be expressed by other means? (I do see of course that possibility is scalar, though NAhE is a little crude as a device for handling scales.)


 forum_name: Modal Aspects
topic_title: pu'i corpse
    subject: pu'i corpse
   username: And
  post_time: 2003-07-20 06:56:45
  post_text:

These are all the uses of pu'i that I've found by googling. On casual inspection, I don't see examples that can't be replaced by ca'a.

... ri lo xadbraxau je melbi gi'eji'a bacru ledu'u ge ra nu'o turni .ei lei cipni gi leda'inu ra cu se voksa lo xamgu cu nibli lenu pu'i turni . www.lojban.org/files/texts/aesop

... ca le nu ko'a fo'a co'a viska cu cusku fi le ko'a xacysantyse'u fe lu be'ucu'i le su'u le dimri'a ca'o gidva le mi'o cuntu cu zabna zmadu le mi'o pu'i se pacna ... www.lojban.org/wiki/index.php/donkixotes

fe lesu'u lo ve pemci pu'i ja nu'o zasti .ijonaibabo ciskycusku lo za'e valsi morna ni'o mi'o se pemci fu'o www.lojban.org/files/texts/cukta1.html

i cibybacru lu .e'u doranji li'u ni'o fu'eza'anai voksa gi'e voksa vi la sancypurdi .i pale voksa pu'i cusku lu ... www.lojban.org/twiki/pub/Files/Documents/guz.html

... ) --jay. .i mi'e xod .i ru'a le fetspe cu natfe vlipa .iki'ubo za'a fy. pu'i fanta vu'enai le zu'o le nakspe cu penmi. www.lojban.org/wiki/index.php/Lojban%20Central

... xagji pemci. leka gusni kei pu'i te zmadu loi gusni loi xagji. www.lojban.org/wiki/index.php/xagji%20pemci

.i .u'i.ieju'o frica .i mi pu'i selxa'e lenu lo genrubju'o ninmu cu nerkla lo velmicy le mi spti

www.lojban.org/files/texts/ckafybarja/C2N.TXT .i le laurxajmi pu'i vlipa .au toi)

mi ba'o mojytadni le gi'uste .i mi pu'i fanva ro ... www.lojban.org/files/texts/1994m2.txt

piso'oda ca'ojamo'u lifri le jdakra gizu'unai piso'oda jiktra po'o .i ji'a de poi se narjda zo'u go'i .i .uuru'e ga'iru'e .i'isai pu'i tcaci le zamnarkri li'a li'u .i denpa .i fo'a ledi'u ca'o pensi .i le bitmu cu xunre bunre mudri .i bacru lu zo'onai di poi ze'a ve pruce le lizyjdakra zo'u ca di ku le tolpei cu simsa lo cilce poi ka'enai vasru


>55 You are in a maze of twisty little passages, all alike. >56 Dead end >57 You are on the brink of a thirty foot pit with a massive orange column >57 \down one wall. You could climb down here but you could not get back >57 \up. The maze continues at this level.

do diklo lo korbi be lo srake'a poi mitre li pano leni condi .i ne'i ra lo tilju ke narju kamju cu lamji pale bitmu .i do vika'e mo'ini'a cpare .iku'i leda'inu pu'i go'i cu nibli lenu do banaka'e mo'iga'u cpare .i le lujypludi'u cu ranji vi levi pinta


 forum_name: Modal Aspects
topic_title: "CAhA le du" versus "CAhA zi'o"
    subject: Re: "CAhA le du" versus "CAhA zi'o"
   username: phma
  post_time: 2003-07-20 09:25:22
  post_text:
There is an important difference in meaning between "CAhA le du" (or CAhA + mostly anything but zi'o) and "CAhA zi'o".

1a. ka'e le du mi cliva "If a certain state of affairs obtained, then I could leave." "In some possible worlds in which a certain state of affairs obtains, I leave." "I could leave."

1b. ka'e zi'o mi cliva "In some possible worlds, I leave." "It is possible that I leave." "I can leave."

Putting {zi'o} in a place deletes the place. {ka'e zi'o mi cliva} is therefore equivalent to {mi cliva}.


 forum_name: Modal Aspects
topic_title: "CAhA le du" versus "CAhA zi'o"
    subject: Re: "CAhA le du" versus "CAhA zi'o"
   username: And
  post_time: 2003-07-21 06:54:15
  post_text:
There is an important difference in meaning between "CAhA le du" (or CAhA + mostly anything but zi'o) and "CAhA zi'o".

1a. ka'e le du mi cliva "If a certain state of affairs obtained, then I could leave." "In some possible worlds in which a certain state of affairs obtains, I leave." "I could leave."

1b. ka'e zi'o mi cliva "In some possible worlds, I leave." "It is possible that I leave." "I can leave."

Putting {zi'o} in a place deletes the place. {ka'e zi'o mi cliva} is therefore equivalent to {mi cliva}.

No, it changes ka'e from a 2-place predicate to a one-place predicate. Imagine if the grammar allowed, say, NA to tag a sumti. Since that would have no meaning, {na broda} would be equivalent to {na zi'o broda}, but not to {broda}. In other words, you're right that putting zi'o in a place deletes the place, but wrong that {ka'e zi'o mi cliva} is therefore equivalent to {mi cliva}.

Well, when I say you're wrong, I mean that the dictates of logic don't necessitate your conclusion. But that doesn't mean we couldn't adopt your conclusion as a rule of Lojban grammar, if we felt it desirable.


 forum_name: Modal Aspects
topic_title: "CAhA le du" versus "CAhA zi'o"
    subject: Re: "CAhA le du" versus "CAhA zi'o"
   username: clsn
  post_time: 2003-07-23 09:59:52
  post_text:
Not hearing any views to the contrary, I will (as shepherd) take this to be the BF position on the matter.
The use you're putting zi'o to in the above really worries me; after all, this modal use of ka'e is quite new, and there's no real precedent for pu zi'o either. I agree with xorxes' conclusion, though; anything but this will lead to utterly unacceptable breaking of backward compatibility. It just seems to me you're really talking about ka'e zu'i, and that ka'e zi'o and pu zi'o are meaningless: the possibility and tense are always relative to something, and the default something is the here-and-now. (So in fact, pu = pu le nau tcini, ka'e = ka'e le nau tcini).


However, the reason I formalized this in terms of zi'o is as follows. If the elided sumti is a zo'e, then by definition the value of zo'e is glorked from context. In the present case that would mean that selbri tcita {ka'e} is neutral between "can" and "could". To force selbri tcita {ka'e} to have only the "can" meaning, we must get rid of the zo'e. (The same goes for {pu}.)

I agree with Nick. {zi'o} should not mean some particular subset of the possible worlds (i.e. the ones consistent with the "real world" of discourse). That's not what {zi'o} is all about. If anything it's {zu'i}, by convention at least. You're not "getting rid" of the {zo'e}, you're restricting it to a particular kind of possible world.


 forum_name: Modal Aspects
topic_title: Meaning of {ca'a} and {pu'i}
    subject: Re: A more complete paradigm of CAhAs
   username: clsn
  post_time: 2003-07-23 10:10:22
  post_text:
There might be a problem, though. NAhE doesn't always apply to CAhA directly. The rule is:

simple-tense-modal = [NAhE] (time [space] | space [time]) & CAhA [KI]

so who knows on what scale NAhE is playing there.

This time, I'm not going to defend a misfeature: a NAhE with ambiguous scope? Unconscionable. We want a scalar CAhA, and if the grammar doesn't give us an unambiguous scalar, then the grammar shall change: end of story. In fact, since we want a scalar CAhA anyway...

... I'm now switching to supporting CAhA NAI.

I'm a little slow on this one. I can see why you're after a scalar CAhA, but how does going to CAhA NAI fix it? NAI only has one cmavo in it, a negator. If we add {ja'ai}, it'll have an affirmator. But we'd still need to add another one ({no'ai?}) for middle-of-the-road, like {no'e}. Or am I missing something?

I think CAhA NAI is likely a good idea anyway, but I want to be sure I'm understanding what we're up to.


 forum_name: Modal Aspects
topic_title: "CAhA le du" versus "CAhA zi'o"
    subject: Re: "CAhA le du" versus "CAhA zi'o"
   username: And
  post_time: 2003-07-23 10:12:08
  post_text:
Not hearing any views to the contrary, I will (as shepherd) take this to be the BF position on the matter.
The use you're putting zi'o to in the above really worries me; after all, this modal use of ka'e is quite new, and there's no real precedent for pu zi'o either. I agree with xorxes' conclusion, though; anything but this will lead to utterly unacceptable breaking of backward compatibility. It just seems to me you're really talking about ka'e zu'i, and that ka'e zi'o and pu zi'o are meaningless: the possibility and tense are always relative to something, and the default something is the here-and-now. (So in fact, pu = pu le nau tcini, ka'e = ka'e le nau tcini).


However, the reason I formalized this in terms of zi'o is as follows. If the elided sumti is a zo'e, then by definition the value of zo'e is glorked from context. In the present case that would mean that selbri tcita {ka'e} is neutral between "can" and "could". To force selbri tcita {ka'e} to have only the "can" meaning, we must get rid of the zo'e. (The same goes for {pu}.)

I agree with Nick. {zi'o} should not mean some particular subset of the possible worlds (i.e. the ones consistent with the "real world" of discourse). That's not what {zi'o} is all about. If anything it's {zu'i}, by convention at least. You're not "getting rid" of the {zo'e}, you're restricting it to a particular kind of possible world.

I still think you and Nick haven't grasped the issue. {zi'o} would not mean some particular subset of possible worlds. {zi'o} kills a sumti place; it turns a 2 place predicate CAhA into a 1-place predicate CAhA. It seems pretty natural to associate the nonconditional "can" meaning with a 1-place predicate and the conditional "could" meaning with a 2-place predicate (the extra place being for the condition, of course).

As I said at the outset, syntactically, it would be more consistent with the rest of the language if selbri tcita CAhA were equivalent to {CAhA zo'e}. But that is a little less consistent with the prescription meaning ("can") & requires overt use of zi'o/zu'i/whatever to unambiguously get the prescription meaning.


 forum_name: Modal Aspects
topic_title: Meaning of {ca'a} and {pu'i}
    subject: Re: A more complete paradigm of CAhAs
   username: xorxes
  post_time: 2003-07-23 10:24:47
  post_text:
We want a scalar CAhA, and if the grammar doesn't give us an unambiguous scalar, then the grammar shall change: end of story. In fact, since we want a scalar CAhA anyway...

... I'm now switching to supporting CAhA NAI.

I'm a little slow on this one. I can see why you're after a scalar CAhA, but how does going to CAhA NAI fix it? NAI only has one cmavo in it, a negator. If we add {ja'ai}, it'll have an affirmator. But we'd still need to add another one ({no'ai?}) for middle-of-the-road, like {no'e}. Or am I missing something?

I think CAhA NAI is likely a good idea anyway, but I want to be sure I'm understanding what we're up to.

Nick might be thinking of CAIs, which is where {nai} belongs. In that case, {cu'i} would be the middle-of-the-roader.


 forum_name: Modal Aspects
topic_title: "CAhA le du" versus "CAhA zi'o"
    subject: Re: "CAhA le du" versus "CAhA zi'o"
   username: nitcion
  post_time: 2003-07-30 10:55:00
  post_text:
I still think you and Nick haven't grasped the issue. {zi'o} would not mean some particular subset of possible worlds. {zi'o} kills a sumti place; it turns a 2 place predicate CAhA into a 1-place predicate CAhA. It seems pretty natural to associate the nonconditional "can" meaning with a 1-place predicate and the conditional "could" meaning with a 2-place predicate (the extra place being for the condition, of course). As I said at the outset, syntactically, it would be more consistent with the rest of the language if selbri tcita CAhA were equivalent to {CAhA zo'e}. But that is a little less consistent with the prescription meaning ("can") & requires overt use of zi'o/zu'i/whatever to unambiguously get the prescription meaning.

OK, I now rule that this is going off-topic, and into a discussion of zi'o rather than ca'a.

But the point of the thread was to establish that tense ca'a (which can be paraphrased as ca'a and some prosumti, be it zi'o or zu'i or le nau co'e or whatever) invokes a subset of what ca'a da does --- that ca'a as tense is here-and-now, while ca'a as sumtcita goes off into possible worlds. If we all agree on this (whatever the details of the paraphrase of tense ca'a), then we have here a prime candidate for a baseline-compliant decision that the bpfk can take. And I think we do agree. Let's concentrate on that for now, and leave the particular paraphrase issue to the KOhA paradigm...


 forum_name: Modal Aspects
topic_title: Meaning of {ca'a} and {pu'i}
    subject: Re: A more complete paradigm of CAhAs
   username: nitcion
  post_time: 2003-07-30 10:56:04
  post_text:
Nick might be thinking of CAIs, which is where {nai} belongs. In that case, {cu'i} would be the middle-of-the-roader.

Yes, that's what I was up to.

In truth, I also said this to stir people up. :-)


 forum_name: Modal Aspects
topic_title: "CAhA le du" versus "CAhA zi'o"
    subject: Re: "CAhA le du" versus "CAhA zi'o"
   username: And
  post_time: 2003-07-30 18:09:12
  post_text:
But the point of the thread was to establish that tense ca'a (which can be paraphrased as ca'a and some prosumti, be it zi'o or zu'i or le nau co'e or whatever) invokes a subset of what ca'a da does --- that ca'a as tense is here-and-now, while ca'a as sumtcita goes off into possible worlds. If we all agree on this (whatever the details of the paraphrase of tense ca'a), then we have here a prime candidate for a baseline-compliant decision that the bpfk can take. And I think we do agree. Let's concentrate on that for now, and leave the particular paraphrase issue to the KOhA paradigm...

I don't think we can agree on the meaning of selbri tcita CAhA without having agreed on what its sumti tcita paraphrase is, because certain paraphrases could be consistent with the rest of the grammar while other paraphrases could be inconsistent. Hence if we insist on consistency, then certain paraphrases would be excluded.

This is a more general matter than a CAhA one, but it is not just a question about zi'o: it is about whether there is any consistent relationship between selbri tcita function and sumti tcita function.


 forum_name: Modal Aspects
topic_title: "CAhA le du" versus "CAhA zi'o"
    subject: Re: "CAhA le du" versus "CAhA zi'o"
   username: xorxes
  post_time: 2003-07-31 17:14:30
  post_text:

I don't think we can agree on the meaning of selbri tcita CAhA without having agreed on what its sumti tcita paraphrase is, because certain paraphrases could be consistent with the rest of the grammar while other paraphrases could be inconsistent. Hence if we insist on consistency, then certain paraphrases would be excluded.

This is a more general matter than a CAhA one, but it is not just a question about zi'o: it is about whether there is any consistent relationship between selbri tcita function and sumti tcita function.

Yes, this is relevant to ZAhO too (if there is a consistent relationship, the current prescription for ba'o and pu'o will almost certainly violate it). There are many questions, such as matters of scope, that have to do with all tags in general, so perhaps it would be worthwhile to have a specific forum for that.