Let's move {soi} to JOI. And why can't places be interconnected in lojban predicates?

[la gleki]

At first some examples.

1.{mi ce do simxu lo nu viska} - presumably it's "We see each other" although we don't know what places of {viska} must be filled with {mi,do}.

2."The clown and the grocer despise one another."

So in the chat it has been suggested that we could move {soi} to JOI with the following result:

3.{mi soi do ze'e prami}, quite a laconic way to say "We loved each other since the beginning of time".

Even more examples:

4."France and Britain had been at war [with each other] for 100 years".

So don't you think that those multiple meaningful examples that are supposed to be frequently realised in our speech are enough to settle the issue with {soi}?

More reading: http://www.tlg.uci.edu/~opoudjis/lojbanbrochure/lessons/lessappendix.html

One more global note.

Predicate arguments seems never to be interacting with each other in gismu. This proposal is one of such attempts to overcome the issue.

[selpa'i]

Am 18.08.2012 17:20, schrieb la gleki:
> At first some examples.
>
> 1.{mi ce do simxu lo nu viska} - presumably it's "We see each other"
> although we don't know what places of {viska} must be filled with {mi,do}.

mi jo'u do simxu lo ka ce'u ce'u viska
(also works with joi and ce)

> So in the chat it has been suggested that we could move {soi} to JOI
> with the following result:
>
> 3.{mi soi do ze'e prami}, quite a laconic way to say "We loved each
> other since the beginning of time".

But what is prami2?

> Even more examples:
>
> 4."France and Britain had been at war [with each other] for 100 years".

lo fraso joi lo brito pu jamsi'u ze'a lo nanca be li panono

> So don't you think that those multiple meaningful examples that are
> supposed to be frequently realised in our speech are enough to settle
> the issue with {soi}?

We have simxu, which is much cleaner, and you can always make -si'u-
lujvo to shorten it, so I don't really see the point of this proposal.
Some time ago, I proposed to allow "empty" soi, that is, allow it to be
used without anything inside, and also to allow inifinitely many
arguments so that both these become grammatical:

mi do soi prami
"I love you and you love me."
"We love each other."

se slabu soi mi do lo pendo be mi [se'u]
"Me, you and my friend all know each other."

This is a sort of afterthought simxu, which I sometimes wanted to have,
but which doesn't exist (I don't like the current soi).
All of the arguments in the soi-phrase reciprocally do [bridi], in all
[meaningful] ways possible.
The proposal didn't get much attention back then, but I guess I can
bring it up again.

mu'o mi'e la selpa'i

--
pilno zo le xu .i lo dei bangu cu se cmene zo lojbo .e nai zo lejbo

[djandus]

Whenever I first learned {soi}, I had a similar thought, but slightly different. It's not to take SOI and make it JOI, because that's basically completely undesired. Instead, it would say:

Use SOI like JOI between sumti, but only between sumti. (SOI thus allows {ku} elision) SOI, like JOI, is allowed to make a SOI series. ({ko'a soi ko'e soi ko'i} is allowed.) SOI asserts that all sumti in the series can be swapped in any order without changing the validity of the overall bridi. Otherwise, you can remove all SOI (and insert the possibly elided {ku} where necessary) and parse the bridi exactly. (So every sumti still falls into the place structure as if SOI wasn't there, except that the components of the series can be swapped and replaced in any order.)

The benefit is that

mi prami do soi vo'a vo'e

is shortened to

mi soi do prami

And even more beneficial are longer statements, such as

mi bevri lo tanxe ku ti soi tu ca'o bi cacra ca lo cabdei

"I carried boxes there and back for eight hours today."

(In the current system, the placement of {soi} is rather annoying as well, as it kind of... sucks, no matter where you put it.)

The main reason I had for the idea was that Lojban goes through so much effort to allow any place structure order that it is always the case that {soi} could be between the necessary sumti, with proper rearranging. Secondarily, it allows the extension of "vice-versa" to include multiple items. (The current SOI uses the restriction of requiring exactly two items to help with VOhA elision.)

However, this type of change sorely disagrees with current usage, so a new word would be more proper than a replacement.

On a complete side note, I also dislike the current SOI process as it feels like it was created with lame reasons behind it. I get the feeling that somebody came up with VOhA for reflexives and somebody didn't like that a whole series of words was being created for one purpose. That, or maybe somebody liked that {vo'a vo'e} parallels "vice-versa" phonetically. Both of which are pretty lame reasons to not think of a better way of doing things.

And on a final side note, I would also like to mention that the reciprocity should be able to be expressed without filling the places, as in the current:

{mi prami soi vo'a vo'e} for "I love something (and it loves me)"

The way I would do that with my proposal would be:

{mi soi vo'e prami}, where vo'e fills the place it represents.

(Once again, if anyone actually likes this proposal, we would need a word other than {soi} to use.)

mu'o mi'e djos

[la gleki]

On Saturday, August 18, 2012 10:14:29 PM UTC+4, djandus wrote:

Whenever I first learned {soi}, I had a similar thought, but slightly different. It's not to take SOI and make it JOI, because that's basically completely undesired. Instead, it would say:

Use SOI like JOI between sumti, but only between sumti. (SOI thus allows {ku} elision) SOI, like JOI, is allowed to make a SOI series. ({ko'a soi ko'e soi ko'i} is allowed.) SOI asserts that all sumti in the series can be swapped in any order without changing the validity of the overall bridi. Otherwise, you can remove all SOI (and insert the possibly elided {ku} where necessary) and parse the bridi exactly. (So every sumti still falls into the place structure as if SOI wasn't there, except that the components of the series can be swapped and replaced in any order.)

The benefit is that

mi prami do soi vo'a vo'e

is shortened to

mi soi do prami

And even more beneficial are longer statements, such as

mi bevri lo tanxe ku ti soi tu ca'o bi cacra ca lo cabdei

"I carried boxes there and back for eight hours today."

(In the current system, the placement of {soi} is rather annoying as well, as it kind of... sucks, no matter where you put it.)

The main reason I had for the idea was that Lojban goes through so much effort to allow any place structure order that it is always the case that {soi} could be between the necessary sumti, with proper rearranging. Secondarily, it allows the extension of "vice-versa" to include multiple items. (The current SOI uses the restriction of requiring exactly two items to help with VOhA elision.)

However, this type of change sorely disagrees with current usage

Pardon, current usage? Does anybody use it? How many times in the corpus it is met? Even L4B mentions it as unsettled and the CLL has two different examples without proper explanations. It sounds more like a waste of precious cmavo space.

[MorphemeAddict]

On Sat, Aug 18, 2012 at 11:43 AM, selpa'i <sel...@gmx.de> wrote:

Am 18.08.2012 17:20, schrieb la gleki:

At first some examples.

1.{mi ce do simxu lo nu viska} - presumably it's "We see each other" although we don't know what places of {viska} must be filled with {mi,do}.

mi jo'u do simxu lo ka ce'u ce'u viska
(also works with joi and ce)

So in the chat it has been suggested that we could move {soi} to JOI with the following result:

3.{mi soi do ze'e prami}, quite a laconic way to say "We loved each other since the beginning of time".

But what is prami2?

Even more examples:

4."France and Britain had been at war [with each other] for 100 years".

lo fraso joi lo brito pu jamsi'u ze'a lo nanca be li panono

Are these specific enough? I expected something like "lo frasygu'e" and "lo britygu'e" (I don't remember their short rafsi). 

stevo 

So don't you think that those multiple meaningful examples that are supposed to be frequently realised in our speech are enough to settle the issue with {soi}?

We have simxu, which is much cleaner, and you can always make -si'u- lujvo to shorten it, so I don't really see the point of this proposal. Some time ago, I proposed to allow "empty" soi, that is, allow it to be used without anything inside, and also to allow inifinitely many arguments so that both these become grammatical:

    mi do soi prami
    "I love you and you love me."
    "We love each other."

    se slabu soi mi do lo pendo be mi [se'u]
    "Me, you and my friend all know each other."

This is a sort of afterthought simxu, which I sometimes wanted to have, but which doesn't exist (I don't like the current soi).
All of the arguments in the soi-phrase reciprocally do [bridi], in all [meaningful] ways possible.
The proposal didn't get much attention back then, but I guess I can bring it up again.

mu'o mi'e la selpa'i

--
pilno zo le xu .i lo dei bangu cu se cmene zo lojbo .e nai zo lejbo

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

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

[selpa'i]

Am 18.08.2012 20:14, schrieb djandus:

Whenever I first learned {soi}, I had a similar thought, but slightly different. It's not to take SOI and make it JOI, because that's basically completely undesired. Instead, it would say:

Use SOI like JOI between sumti, but only between sumti. (SOI thus allows {ku} elision)

What do you mean? ku can always be elided, even with JOI: lo mlatu joi lo gerku
There was a time when this didn't work due to parser limitations, but we're long past that.

SOI, like JOI, is allowed to make a SOI series. ({ko'a soi ko'e soi ko'i} is allowed.) SOI asserts that all sumti in the series can be swapped in any order without changing the validity of the overall bridi. Otherwise, you can remove all SOI (and insert the possibly elided {ku} where necessary) and parse the bridi exactly. (So every sumti still falls into the place structure as if SOI wasn't there, except that the components of the series can be swapped and replaced in any order.)

A similar thing happens in my proposal too, but this is a bit weird, because you're filling one sumti place while leaving all the others empy.

The benefit is that

mi prami do soi vo'a vo'e

is shortened to

mi soi do prami

And even more beneficial are longer statements, such as

mi bevri lo tanxe ku ti soi tu ca'o bi cacra ca lo cabdei

"I carried boxes there and back for eight hours today."

mi bevri lo tanxe ti fa'u tu ze'a lo cacra be li bi ca lo cabdei

[...]

However, this type of change sorely disagrees with current usage, so a new word would be more proper than a replacement.

soi sees extremely little usage; almost any change is an improvement, but moving soi to JOI is a bit odd, it still feels hackish. My suggestion doesn't break usage though, if that's really of importance.

[Jacob Errington]

I created a selma'o once for an idea I had (i.e. multiple-sumti relative clauses). It turns out that it does what you want {soi} to do. That is to say, "connect" two sumti, while preserving the fact that they fill different places in the selbri's structure. In particular, {ce'e} does this, but moving {soi} to CEhE would cause the awful side effect of the resulting joint sumti being connectable with pe'e+connective. 

Anyway, the selma'o I had made up for this purpose is JOhOI and it has essentially the same grammar as CEhE, minus the pe'e stuff. The fact that the grammar is the same as CEhE means that it is implementable. "Move" soi into JOhOI (call it SOI still) and change se'u such that it doesn't terminate SOI anymore (only SEI). SOI still only connects sumti (or terms, due to the nature of CEhE) so there's no issue as there would have been with JOI.

Formally define the interpretation of SOI such that any of the sumti in the SOI-connected series can be arbitrarily swapped and the bridi still holds.

Problem solved? :)

mu'o mi'e la tsani

--

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

To post to this group, send email to loj...@googlegroups.com.

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