A Simpler Connective System (blog article)

[selpahi]

"How does one say 'and' in Lojban?"
"Well... it depends, you know. For sumti, it's {e}. For tanru and
sentences, it's {je}. For bridi tails it's {gi'e}. And those were just
the afterthought connectives. Forethought 'and' is {ge} for everything
but tanru, which use {gu'e} instead. Relative clauses are connected
with {zi'e}."
"... oh."

With no less than six ways to say "and" as well as roughly the same
number for each of the other logical functions (OR, IFF, WHETHER-OR-NOT)
totaling 26, the Lojban connective system can be quite daunting. But
despair not, for there is a light at the end of the tunnel.

[...]

To read the full article follow this link:

https://solpahi.wordpress.com/2016/09/20/a-simpler-connective-system/

[MorphemeAddict]

How would adoption of this proposal affect the length of the current grammar? I assume it would shorten it, but how much? 

I favor the proposal. 

stevo

--
You received this message because you are subscribed to the Google Groups "lojban" group.
To unsubscribe from this group and stop receiving emails from it, send an email to lojban+unsubscribe@googlegroups.com.
To post to this group, send email to loj...@googlegroups.com.
Visit this group at https://groups.google.com/group/lojban.
For more options, visit https://groups.google.com/d/optout.

[Jonathan Jones]

I remember a similar proposal made years ago, I think by .xorxes., that does about the same thing for the same reason.

--

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 )

[Mark E. Shoulson]

There *used* to be zi'a/zi'e/zi'o/zi'u, completing the ZIhEk spectrum, but it was determined for some reason that non-conjunctive conjunctions of relative clauses didn't make sense, and only zi'e was needed (and the other cmavo got repurposed for other things).

~mark

To unsubscribe from this group and stop receiving emails from it, send an email to lojban+un...@googlegroups.com.

[And Rosta]

Having given the matter about .0001% of the thought you have, I wonder whether the gi'i terminator is optimal. Firstly it would not always be easy to work out on the fly when it is and isn't elidable, so the prudent strategy would be to leave it in except when certain it is elidable. Secondly, when it isn't elided it adds an extra word and two extra syllables. A better alternative would be to introduce medial conjuncts with _go_ rather than _gi_, and use _gi_ only for introducing final conjuncts: {ga JA A go B go C gi D}. (Or, one step neater, use _gu_ for medial conjuncts and _go_ for the tanru coordination introducer. Or _ge_.)

--And.

 

[selpahi]

On 26.09.2016 16:43, And Rosta wrote:
> Having given the matter about .0001% of the thought you have, I wonder
> whether the gi'i terminator is optimal. Firstly it would not always be
> easy to work out on the fly when it is and isn't elidable, so the
> prudent strategy would be to leave it in except when certain it is
> elidable. Secondly, when it isn't elided it adds an extra word and two
> extra syllables. A better alternative would be to introduce medial
> conjuncts with _go_ rather than _gi_, and use _gi_ only for introducing
> final conjuncts: {ga JA A go B go C gi D}. (Or, one step neater, use
> _gu_ for medial conjuncts and _go_ for the tanru coordination
> introducer. Or _ge_.)

The only times {gi'i} would not be elidible is if another connective
follows that is supposed to apply to the entire forethought connection
to its left. In all other cases {gi'i} is elidible, because each {gi}
can only devour exactly one sumti, after which the entire connection
ends automatically.

Your strategy with {go} would involve much more forethought than this,
because you would have to be absolutely certain that you only want to
add exactly one more item.

Do you have any situations in mind where working out whether {gi'i} is
elidible would not be easy?

Thank you for your comment.

~~~mi'e la solpa'i

[Jacob Errington]

I'm in favour of a simplification of the connective system. However, there is something that bothers me about this proposed simplification; in particular, it changes the mental model one has of {cu}, breaking a certain elegant symmetry between bridi heads and tails.

A particular type of sentence in Lojban is a bridi, of which the only mandatory component is the bridi-tail, which begins with a selbri, ends with {vau}, and encloses zero or more terms. If a bridi-tail is present, it may be preceded with a bridi-head, which begins after {zo'u}, ends with {cu}, and encloses one or more terms.

The symmetry that I'm speaking of is that {cu} parallels {vau}: the former terminates the bridi-head and the latter, the bridi-tail. Notice that in this proposal, it becomes possible to write {.i ko'a broda je cu brode}. Clearly, {cu} here is not operating as a terminator anymore, and the symmetry between {cu} and {vau} is lost.

In the official grammar, it can be observed that {cu} is a terminator by noticing that it is elided only when one or more terms precede the bridi-tail: contrast {ko'a [cu] broda [vau]} with {broda [vau]}. The proposed connective reform does more than reform connectives. It alters the operation of {cu}, so that rather act as a bridi-head elidible terminator, it acts as a bridi-tail elidible *initiator*.

Reforming the overly complex connective system is an excellent goal, but I am against the means to that end employed in this proposal, namely the alteration of {cu} to become an elidible initiator (which would also make it the first of its kind, I think).

.i mi'e la tsani mu'o

[selpahi]

On 26.09.2016 17:22, Jacob Errington wrote:
> In the official grammar, it can be observed that {cu} is a terminator
> by noticing that it is elided only when one or more terms precede the
> bridi-tail: contrast {ko'a [cu] broda [vau]} with {broda [vau]}. The
> proposed connective reform does more than reform connectives. It
> alters the operation of {cu}, so that rather act as a bridi-head
> elidible terminator, it acts as a bridi-tail elidible *initiator*.

{cu} is not really a terminator. The fact that you can't have {cu}
without a bridi-tail is evidence of that. If {cu} were really a
bridi-head terminator, {ko'a cu} should be a valid fragment. Note also
that people think that things like {sei broda cu brode} should be legal.
This shows that psychologically {cu} belongs to the tail, rather than
the head.

Also, {cu} and {vau} don't really form a pair. {vau} appears even in
phrases that contain zero {cu}, and the number of {vau} in sentences
containing bridi-tails exceeds the number of {cu}. If they were a real
pair, they should always appear equally often.

> Reforming the overly complex connective system is an excellent goal,
> but I am against the means to that end employed in this proposal,
> namely the alteration of {cu} to become an elidible initiator (which
> would also make it the first of its kind, I think).

I like to compare {cu} to FA. A FA is an optional marker for a
particular sumti, while {cu} is an optional marker for the selbri. Both
FA and CU mark particular slots in the bridi.

[Jacob Errington]

> On Sep 26, 2016, at 11:31 AM, selpahi <sel...@gmx.de> wrote:
>
> On 26.09.2016 17:22, Jacob Errington wrote:
>> In the official grammar, it can be observed that {cu} is a terminator
>> by noticing that it is elided only when one or more terms precede the
>> bridi-tail: contrast {ko'a [cu] broda [vau]} with {broda [vau]}. The
>> proposed connective reform does more than reform connectives. It
>> alters the operation of {cu}, so that rather act as a bridi-head
>> elidible terminator, it acts as a bridi-tail elidible *initiator*.
>
> {cu} is not really a terminator. The fact that you can't have {cu} without a bridi-tail is evidence of that. If {cu} were really a bridi-head terminator, {ko'a cu} should be a valid fragment. Note also that people think that things like {sei broda cu brode} should be legal. This shows that psychologically {cu} belongs to the tail, rather than the head.

{ko'a cu} is not a bridi-head, since a bridi-head can appear only before a bridi-tail. (I gave an informal characterization of bridi-heads and bridi-tails in my last message. Notice that my characterization rules out {ko'a cu} as a bridi-head, as it does not precede a bridi-tail.)

As for people wanting to say such things as {sei broda cu brode} as the basis of a psychological argument in favour of {cu} belonging to the tail, I argue that people think this way because the most common way of explaining {cu} to people is a lie: "{cu} is a cmavo that can be placed before the main selbri in order to force the insertion of all elidible terminators necessary to make the appearance of the main selbri grammatical". Sure, according to this explanation, {cu} belongs to the tail, but as I said, this explanation is a lie. According to this explanation, {cu broda} would be legal, since it forces the inclusion of all (zero) terminators necessary to make the appearance of the main selbri {broda} grammatical. However, {cu broda} is not allowed in the official grammar.

Therefore, as I understand it, your psychological argument boils down to this: we should change the operation of {cu} because of a convenient lie we've told to too many people. (I'm not at all claiming that I'm not guilty of perpetuating this lie too; I've explained {cu} this way countless times.)

As for people who want things like {sei broda cu brode} to be legal, according to my (rather straightforward) characterization of bridi as a bridi-tail plus an optional head, {sei broda cu brode} is ruled out, as {sei broda} is not a term. Hence, there is no bridi-head. ("If a bridi-tail is present, it may be preceded with a bridi-head, which begins after {zo'u}, ends with {cu}, and encloses *one or more* terms." Added emphasis.)

Perhaps we can just teach people this characterization instead of fundamentally changing the operation of {cu}. In fact, this is already how I explain {cu} to people now. Rather than change the grammar so that we can have been right all along, let's just admit that our explanation was wrong and move on.

>
> Also, {cu} and {vau} don't really form a pair. {vau} appears even in phrases that contain zero {cu}, and the number of {vau} in sentences containing bridi-tails exceeds the number of {cu}. If they were a real pair, they should always appear equally often.
>

There's no argument here. I agree with you entirely that {cu} and {vau} do not form a pair. This contradicts nothing of what I said though; I never said that they formed a pair. I merely said that there was a kind of elegant symmetry between bridi-heads and bridi-tails, since each has a terminator.

>> Reforming the overly complex connective system is an excellent goal,
>> but I am against the means to that end employed in this proposal,
>> namely the alteration of {cu} to become an elidible initiator (which
>> would also make it the first of its kind, I think).
>
> I like to compare {cu} to FA. A FA is an optional marker for a particular sumti, while {cu} is an optional marker for the selbri. Both FA and CU mark particular slots in the bridi.
>

And I like to compare {cu} to {vau}! {vau} terminates a bridi-tail whereas {cu} terminates a bridi-head. Both terminate important parts of the bridi.

Of course, {cu} cannot appear without {vau}, since a bridi-head cannot be present without an accompanying tail. But then again, this is no different than how CU differs from FA in your proposal, since {fa mi} is not a bridi whereas {cu broda} is.

[And Rosta]

On 26 September 2016 at 16:04, selpahi <sel...@gmx.de> wrote:

On 26.09.2016 16:43, And Rosta wrote:

Having given the matter about .0001% of the thought you have, I wonder
whether the gi'i terminator is optimal. Firstly it would not always be
easy to work out on the fly when it is and isn't elidable, so the
prudent strategy would be to leave it in except when certain it is
elidable. Secondly, when it isn't elided it adds an extra word and two
extra syllables. A better alternative would be to introduce medial
conjuncts with _go_ rather than _gi_, and use _gi_ only for introducing
final conjuncts: {ga JA A go B go C gi D}. (Or, one step neater, use
_gu_ for medial conjuncts and _go_ for the tanru coordination
introducer. Or _ge_.)

The only times {gi'i} would not be elidible is if another connective follows that is supposed to apply to the entire forethought connection to its left. In all other cases {gi'i} is elidible, because each {gi} can only devour exactly one sumti, after which the entire connection ends automatically.

Your strategy with {go} would involve much more forethought than this, because you would have to be absolutely certain that you only want to add exactly one more item.

Yes, but in designing an ergonomic loglang -- which okay, Lojban isn't and doesn't aspire to be, but we're kind of imagining if it were trying to be -- should minimizing speaker forethought be prioritized over, say, minimizing hearer backtracking, or minimizing verbosity? (I think no.)

 

Do you have any situations in mind where working out whether {gi'i} is elidible would not be easy?

I was thinking of {ga JA1 ga JA2 ga JA3 A gi B gi C gi D gi E gi F gi G gi H gi I gi J gi K}, which I presume would treat A--K as conjuncts of JA3, and would require two {gi'i}s (terminating JA3 and JA2) to mark the intended structure (whatever it is). Have I misunderstood? I suppose it's not hard to work out that gi'i is unelidable, so I will withdraw the first of my two objections. A reason for preferring the terminator over alternatives is that terminators are the Lojban way; but a reason for preferring terminatorless alternatives is that they can potentially involve incremental parsing without lookahead, which I think is psycholinguistically much less taxing, and that terminators are psycholinguistically alien.

--And.

 

Thank you for your comment.

~~~mi'e la solpa'i

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

To unsubscribe from this group and stop receiving emails from it, send an email to lojban+unsubscribe@googlegroups.com.

[John E Clifford]

Since JCB through out the basic structure of FOPL on day one of the development of Loglan and decided to graft a pseudo form onto a SAE base, things like trying to simplify the conjunction system have been a matter of ever increasing complexity, interrupted occasionally by attempts to get back to the basic underlying simplicity -- with scarcely visible success.  To be sure, Lojban has achieved the monoparsing with which it should have begun at the cost of Byzantine complexity (and questionable accuracy) But it seems unlikely that much reform can keep this result and cut through the mare's nest.  There are those that love the complexity and the documented structure (the best documentation in the language business, after all) and -- despite occasional complaints about not getting more new people -- glory in their isolated mastery, and  so they are not interested in "improvements".  Mere improvers are also too tied up in the status quo to consider scrapping the mess and starting over on the right foot this time.  So, changes, fueled merely be convenience or clarity, are not likely to occur.  Changes that add to complexity are always welcome, of course.  

To unsubscribe from this group and stop receiving emails from it, send an email to lojban+un...@googlegroups.com.

[Jonathan Jones]

Aha, I found .xorxes.'s proposal. Apparently it overloads {gi}, while this one does not?

https://groups.google.com/forum/#!topic/lojban/ExtEumbYoQg

To unsubscribe from this group and stop receiving emails from it, send an email to lojban+unsubscribe@googlegroups.com.

To post to this group, send email to loj...@googlegroups.com.
Visit this group at https://groups.google.com/group/lojban.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "lojban" group.
To unsubscribe from this group and stop receiving emails from it, send an email to lojban+unsubscribe@googlegroups.com.

To post to this group, send email to loj...@googlegroups.com.
Visit this group at https://groups.google.com/group/lojban.
For more options, visit https://groups.google.com/d/optout.

[Remo Dentato]

I've always found the connective system too complex. I never really managed to understand it in its entirety (and I suspect very few did).

To me this seems really a case where the Lojban community as all to gain from this change.  My only question would be if the peg grammar can be updated to reflect this change but I understand that an experimental parser including this change has already been made, so I'm fine with that.

P.S. I had missed the other "Simpler" article, which I favor as well. Again, the increase in simplicity would be huge and breaking compatibility with current lojban would be worthy.

[Ilmen]

You can test the JACU connective system here:

• http://lojban.github.io/ilmentufa/camxes.html

Select "Camxes: Beta" in the dropdown menu labeled "Parser:", under the text input field. (The parser selected by default is the BPFK standard grammar.)

This parser doesn't implement n-ary forethough connectives terminated with {gi'i} though, but the Zantufa parser does implement it:

• http://guskant.github.io/gerna_cipra/zantufa-1.html

(However I seem to understand that Solpahi proposes that each {gi} would introduce only one new term, with the benefit of requiring the connection terminator {gi'i} much less often, whereas in Zantufa, any number of bridi terms can appear after each {gi}, so they behave like termset connectives, and I suppose {gi'i} is required more often there. Please correct me if I misunderstood.)

—Ilmen.

To unsubscribe from this group and stop receiving emails from it, send an email to lojban+un...@googlegroups.com.

[selpahi]

On 27.09.2016 11:13, Ilmen wrote:
> This parser doesn't implement n-ary forethough connectives terminated
> with {gi'i} though, but the Zantufa parser does implement it:
>
> • http://guskant.github.io/gerna_cipra/zantufa-1.html
>
> (However I seem to understand that Solpahi proposes that each {gi} would
> introduce only one new term, with the benefit of requiring the
> connection terminator {gi'i} much less often, whereas in Zantufa, any
> number of bridi terms can appear after each {gi}, so they behave like
> termset connectives, and I suppose {gi'i} is required more often there.
> Please correct me if I misunderstood.)

{ge mi gi do ti gi ti ta} does not parse in zantufa, but {ge mi ti gi
ti ta tu gi ti ta} does. When you start with a gek-termset then each
connected unit allows an n-ary termset (of arbitrary length) and then
{gi'i} is not elidible if any other terms follow, but if you start with
a unary connectand, then each connected unit also must be unary. I think.

[Ilmen]

Hmm, under Zantufa, {ge ti gi ta gi tu mi} parses as {ge ti gi ta gi tu
GIhI mi VAU}, so there when there's only one term in the fist
connectand, it does seem to expect one term in each other connectand.
However things like {ge ti gi naku ta gi tu mi} parses as {ge ti gi naku
ta gi tu mi GIhI VAU}, and {ge ti gi ke ta tu (ke'e) gi do mi}
parsefails. Weird.
Maybe there's a bug in there.

—Ilmen.

[gleki.is...@gmail.com]

I have to agree with la tsani's arguments. Besides, feedback from users showed to me that variations of mad proposals lead to no gain. 

[gleki.is...@gmail.com]

Em segunda-feira, 26 de setembro de 2016 21:31:11 UTC+3, aionys escreveu:

Aha, I found .xorxes.'s proposal. Apparently it overloads {gi}, while this one does not?

https://groups.google.com/forum/#!topic/lojban/ExtEumbYoQg

You can also find a boiled down version of it.

 

To unsubscribe from this group and stop receiving emails from it, send an email to lojban+un...@googlegroup s.com.

To post to this group, send email to loj...@googlegroups.com.
Visit this group at https://groups.google.com/grou p/lojban.
For more options, visit https://groups.google.com/d/op tout.

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

To unsubscribe from this group and stop receiving emails from it, send an email to lojban+un...@googlegroups.com.

To post to this group, send email to loj...@googlegroups.com.
Visit this group at https://groups.google.com/group/lojban.
For more options, visit https://groups.google.com/d/optout.

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

To unsubscribe from this group and stop receiving emails from it, send an email to lojban+un...@googlegroups.com.

To post to this group, send email to loj...@googlegroups.com.
Visit this group at https://groups.google.com/group/lojban.
For more options, visit https://groups.google.com/d/optout.

[Susannah Doss]

As a nintadni whose opinion arguably doesn't matter that much*, I've avoided using connectives because I was horribly confused by what I've read about the existing system. .i .ie ji'a lo frica nintadni cu tugni lo du'u mi'a cinmo lo xrani .uanmonai There were so many words to remember for different situations! When I read the new proposal, I immediately understood the proposed system. It seems much more elegant than the existing system. I really like it. .i lo mibypre cu pa'itce lo melbi selti'i

* to .i a'o lo mrilu mibziljmina cu nalkansa lo donynabmi toi

.i ki'e la zabna donpre

.i mi'e la .suzanys.

To unsubscribe from this group and stop receiving emails from it, send an email to lojban+unsubscribe@googlegroups.com.

[Gleki Arxokuna]

2016-09-27 15:18 GMT+03:00 Susannah Doss <susann...@gmail.com>:

As a nintadni whose opinion arguably doesn't matter that much*, I've avoided using connectives because I was horribly confused by what I've read about the existing system. .i .ie ji'a lo frica nintadni cu tugni lo du'u mi'a cinmo lo xrani .uanmonai There were so many words to remember for different situations! When I read the new proposal, I immediately understood the proposed system. It seems much more elegant than the existing system. I really like it. .i lo mibypre cu pa'itce lo melbi selti'i

My experience with lo nintadni is that the problem is definitely not in the system itself but in how it is taught.

CLL by design is not a tutorial, and other textbooks by far only tried to copy it, successfully or unsuccessfully.

--
You received this message because you are subscribed to a topic in the Google Groups "lojban" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/lojban/ewQLBEaH52s/unsubscribe.
To unsubscribe from this group and all its topics, send an email to lojban+unsubscribe@googlegroups.com.

[Susannah Doss]

My experience with lo nintadni is that the problem is definitely not in the system itself but in how it is taught.
CLL by design is not a tutorial, and other textbooks by far only tried to copy it, successfully or unsuccessfully.

The new system is much more intuitive than the old. I don't think that is entirely due to how it is taught. Having 5 words to remember instead of 26 is nice assuming 5 can do the work of 26. I've been reading the CLL in various sections throughout the book and I haven't come across as much concept confusion as I have when attempting to make it through the section on the connective system. CLL is indeed not a tutorial, but I feel like I gain much more understanding out of it than reading the textbooks. It makes everything make sense and feel connected with the top-down view. I feel the theoretical basis for old connective system isn't great. It doesn't feel elegant. It doesn't seem to fit within the Lojban language as well as other concepts. I don't really notice others using the connectives too much either in the IRC. But errr, I may have not been paying the closest attention. I can at least speak for myself in that the old system confuses the hell out of me when reading about it in CLL compared to reading other concepts in CLL.

I want to use connectives but I don't know how!

[Gleki Arxokuna]

2016-09-27 15:56 GMT+03:00 Susannah Doss <susann...@gmail.com>:

My experience with lo nintadni is that the problem is definitely not in the system itself but in how it is taught.
CLL by design is not a tutorial, and other textbooks by far only tried to copy it, successfully or unsuccessfully.

The new system is much more intuitive than the old. I don't think that is entirely due to how it is taught. Having 5 words to remember instead of 26 is nice assuming 5 can do the work of 26.

I already tried this system and now hate it.

You can't judge by the number of words.

One important argument against this system is that people studying Lojban often incorrectly omit {lo} resulting in weird results under this system:

{lo nixli je nanla} means "someone who is a girl and a boy at the same time."

With the standard system an attempt to do that simply leads to an ungrammatical thing that can be easily detected without delving whether the text can be meaningful at all. The parser immediately warns of

{lo nixli .e nanla.}

The correct thing is {lo nixli .e lo nanla.}

The connective system is pretty simple:

.a, ja, gi'a, .i ja

.e, je, gi'e, .i je

etc.

One might say that

.ji, je'i, gi'i, .i je'i

makes complications but it doesn't really belong to the series of logical connectives. This is a series of question words.

Similarly, non-logical connectives like jo'u are used primarily to connect nouns (lo sumti). And when they connect other elements these are in majority of cases proposals to break their semantical/grammatical silliness.

Yes, you can't judge by the number of words.

The point is that there are indeed 5 distinct concepts and they simply cannot be replaced with one. {.e} connects nouns, {gi'e} connects tails of clauses etc.

They all are simple enough to remember since they all have the same "suffix" -e.

Variations of these mad proposals (initially proposed in year 1996) do not replace these 5 concepts but simply use the same word for some structures.

Please, note that this {je cu} is no difference from {gi'e}. It is still not {je}.

Other arguments include:

2. breaking the meaning of {cu} as pointe out by la tsani.

3. always keeping in mind where you can elide {cu} and where you cannot (as the paper with the proposal itself mentions)

4. always keeping in mind garden path sentences (what did I use {je} for? For a sumti or for something else?)

Interestingly, that the original proposals by la xorxes look even more elaborated. There you would use {gi je} instead of {gi'e}. In another proposal it was {vau je} instead of {gi'e}. Although of course it's no win. You still have to remember this construct {gi je} is a single entity just like you remember {lo nu} or {lo ka}.

I've been reading the CLL in various sections throughout the book and I haven't come across as much concept confusion as I have when attempting to make it through the section on the connective system. CLL is indeed not a tutorial, but I feel like I gain much more understanding out of it than reading the textbooks. It makes everything make sense and feel connected with the top-down view. I feel the theoretical basis for old connective system isn't great. It doesn't feel elegant.

What exactly doesnt feel elegant to you? there are only 5 endings. The rest is about syntax. How is {je cu} better than {gi'e}? I dont get it.

It doesn't seem to fit within the Lojban language as well as other concepts. I don't really notice others using the connectives too much either in the IRC. But errr, I may have not been paying the closest attention. I can at least speak for myself in that the old system confuses the hell out of me when reading about it in CLL compared to reading other concepts in CLL.

I want to use connectives but I don't know how!

But I suppose now after using this system you know perfectly what they are, do you?

[Susannah Doss]

I don't think I understand it, but whatever. The logical vs. non-logical distinction and other parts are still confusing me. Thank you for the explanation though. I probably just need to try and read more on it.

Anyway-- where can I find the mentioned previous, similar proposals? Would you include links, or are they buried in the mailing list? I'd like to learn more.

[Gleki Arxokuna]

2016-09-27 16:38 GMT+03:00 Susannah Doss <susann...@gmail.com>:

I don't think I understand it, but whatever. The logical vs. non-logical distinction and other parts are still confusing me. Thank you for the explanation though. I probably just need to try and read more on it.

Anyway-- where can I find the mentioned previous, similar proposals? Would you include links, or are they buried in the mailing list? I'd like to learn more.

la .aionys. posted earlier one link and there is "watered down" version. you may use google groups search.

{vau je} must be a part of http://lojban.github.io/ilmentufa/camxes.html

[Yon]

coi

On Tue, Sep 27, 2016 at 04:07:54PM +0300, Gleki Arxokuna wrote:
> One important argument against this system is that people studying Lojban
> often incorrectly omit {lo} resulting in weird results under this system:
> {lo nixli je nanla} means "someone who is a girl and a boy at the same
> time."

As a nintadni, I suppose I will probably sometimes make this error, and
I think this is a very interesting point, but, I wonder, does it really
generate that much confusion in practice? I mean, in natural languages
there are many sources of confusion resolved by context even when you do
things right: at least in this case it would occur only when you make
actual mistakes.

> One might say that
> .ji, je'i, gi'i, .i je'i
> makes complications but it doesn't really belong to the series of logical
> connectives. This is a series of question words.

One thing I'm not sure about is: the proposed system seems to allow
the use of all non-logical connectives in forethought connectives too.
Is there a way to do that for all such connectives with the current
system? This part of the proposal seems from my naive point of view
quite orthogonal, and could be used with little change with the current
system, am I right?

> What exactly doesnt feel elegant to you? there are only 5 endings. The rest
> is about syntax. How is {je cu} better than {gi'e}? I dont get it.

I suppose it's at least a little more mnemonic because it uses two known
words, and in the mind of a nintadni like me, «cu» is some sort of word
that says to elidable terminators «clear the way, I want to say the
selbri».

That said, as a nintadni, I haven't for now experienced real problems
with connectives. I have found others parts of Lojban more confusing
(for example the quantifier stuff), even though I did know a bit of
predicate logic. And learning vocabulary for me is perhaps the hardest
part, and yet it is simpler than with natural languages (apart perhaps
for the big number of cmavo, but I hope one can do quite well knowing
only a reduced number of them).

.i mi'e la .anasetos.

[And Rosta]

On 26 September 2016 at 16:58, Jacob Errington <ja...@mail.jerrington.me> wrote:

{ko'a cu} is not a bridi-head, since a bridi-head can appear only before a bridi-tail. (I gave an informal characterization of bridi-heads and bridi-tails in my last message. Notice that my characterization rules out {ko'a cu} as a bridi-head, as it does not precede a bridi-tail.)

Can cu-terminated bridi-heads be conjoined, e.g. {ge ti cu gi ta cu broda}? One would expect the answer to be Yes.

--And.

[Jacob Thomas Errington]

On 09/27/2016 01:57 PM, And Rosta wrote:
>
>
> On 26 September 2016 at 16:58, Jacob Errington <ja...@mail.jerrington.me

I wouldn't be against extending the connective system to allow it, but
in the official grammar it is not possible to connect bridi-heads.

[MorphemeAddict]

I disagree about the five concepts. There is only one: connecting like constructs. The constructs may have five different shapes or structures, but the connection is only one. 

stevo

[John E Clifford]

Right, there are only sentential connectives (how else could they be truth functional, after all?).  All {A x conj y B} come from {A x B conj A y B} in the same way.  It is an early and persistent flaw of Logjam that the local grammar of x and y is often lost in a maze of secondary matter, since, ultimately, there are only terms and predicates and they are (or ought to be) obviously distinct.  (That they are not at the crucial point is thanks to JCB's initial decisions, one of which was to make the connections between terms primary, against the logic he was supposedly following -- and even against SAE, which he was actually following, usually.)

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" group.

To unsubscribe from this group and stop receiving emails from it, send an email to lojban+unsubscribe@googlegroup s.com.

To post to this group, send email to loj...@googlegroups.com.
Visit this group at https://groups.google.com/grou p/lojban.
For more options, visit https://groups.google.com/d/op tout.

--

You received this message because you are subscribed to a topic in the Google Groups "lojban" group.
To unsubscribe from this topic, visit https://groups.google.com/d/to pic/lojban/ewQLBEaH52s/unsubsc ribe.

To unsubscribe from this group and all its topics, send an email to lojban+unsubscribe@googlegroup s.com.

To post to this group, send email to loj...@googlegroups.com.
Visit this group at https://groups.google.com/grou p/lojban.
For more options, visit https://groups.google.com/d/op tout.

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

To unsubscribe from this group and stop receiving emails from it, send an email to lojban+unsubscribe@googlegroup s.com.

To post to this group, send email to loj...@googlegroups.com.
Visit this group at https://groups.google.com/grou p/lojban.
For more options, visit https://groups.google.com/d/op tout.

--

You received this message because you are subscribed to a topic in the Google Groups "lojban" group.
To unsubscribe from this topic, visit https://groups.google.com/d/to pic/lojban/ewQLBEaH52s/unsubsc ribe.

To unsubscribe from this group and all its topics, send an email to lojban+unsubscribe@googlegroup s.com.

To post to this group, send email to loj...@googlegroups.com.
Visit this group at https://groups.google.com/grou p/lojban.
For more options, visit https://groups.google.com/d/op tout.

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

To unsubscribe from this group and stop receiving emails from it, send an email to lojban+unsubscribe@ googlegroups.com.

To post to this group, send email to loj...@googlegroups.com.
Visit this group at https://groups.google.com/ group/lojban.
For more options, visit https://groups.google.com/d/ optout.

[Curtis Franks]

> {lo nixli je nanla} means "someone who is a girl and a boy at the same time."

To me, this is a feature, not a bug. I dislike omitting {lo}.

[Gleki Arxokuna]

2016-10-24 9:52 GMT+03:00 Curtis Franks <curtis....@gmail.com>:

> {lo nixli je nanla} means "someone who is a girl and a boy at the same time."

To me, this is a feature, not a bug.

Who doubts that?

 

I dislike omitting {lo}. 

[Curtis Franks]

> Who doubts that?

Maybe no-one. I just thought that it would be good to point out that it is not a problematic property of the language to me.

[Gleki Arxokuna]

2016-11-06 7:20 GMT+03:00 Curtis Franks <curtis....@gmail.com>:

> Who doubts that?

Maybe no-one. I just thought that it would be good to point out that it is not a problematic property of the language to me.

I don't understand the reason for your comment. Earlier I simply noted that newcomers may forget {lo} (most likely because their native languages classify words into nouns and verbs). In this regard {.e} helps figure out such mistake easier.

[adamb...@gmail.com]

Is there a jo'au for the system?

[Ilmen]

I'd just say {jo'au (la) .jacus.}, from "JACU", a name sometimes used
for referring to the proposal allowing {broda je cu brode} as a
replacement for {broda gi'e brode}.