si and quotes

21 views
Skip to first unread message

Luke Bergen

unread,
Oct 13, 2010, 9:32:58 PM10/13/10
to lojba...@lojban.org

So, I understand what happens with zo si si si.  But what happens if I say {.i mi cusku lu mi nelci lo xunre li'u si gerk li'u}.  Does the si successfully "take back" the first li'u such that the "gerku" is still part of the quote?

Jorge Llambías

unread,
Oct 13, 2010, 10:02:51 PM10/13/10
to loj...@googlegroups.com
On Wed, Oct 13, 2010 at 10:32 PM, Luke Bergen <lukea...@gmail.com> wrote:
> So, I understand what happens with zo si si si.

With the PEG grammar you only use one "si" to delete the whole "zo
si", and there is no way to delete the quoted word alone. This is
because magic words work on a "first come first served" basis, and
once a word has been quoted with "zo" they form an inseparable unit.

"Magic words" are ZO, ZEI, BU, SI, SA, SU, LOhU-LEhU and ZOI. The
peculiarity of these words is that they see all other words as just
words, they don't care what selma'o they belong to, they treat them
all the same. And once they've grabbed them they won't let them go, so
essentially they form a compound that behaves like a single word.

> But what happens if I say
> {.i mi cusku lu mi nelci lo xunre li'u si gerk li'u}.  Does the si
> successfully "take back" the first li'u such that the "gerku" is still part
> of the quote?

Yes. "lu" and "li'u" behave like ordinary words. They do care what
selma'o the words they have dealings with belong to.

But in "lo'u si si si si si si si le'u si", all the "si" inside
lo'u-le'u are deactivated (because lo'u came first and treats anything
until "le'u" as just-a-word, it doesn't care what word it is. But then
the whole construction "lo'u si si si si si si si le'u" is now a
single entity and the final "si" can take revenge and delete it all in
one swoop.

Similarly a single "si" is enough to delete "broda zei brode" just as
it deletes "brodybrode", and it will delete ".abu" just as it deletes
"by".

mu'o mi'e xorxes

Luke Bergen

unread,
Oct 13, 2010, 10:18:14 PM10/13/10
to loj...@googlegroups.com

.oi .uu.  So I guess the song zo si si si doesn't really work then

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

Jorge Llambías

unread,
Oct 14, 2010, 9:05:04 AM10/14/10
to loj...@googlegroups.com
On Wed, Oct 13, 2010 at 11:18 PM, Luke Bergen <lukea...@gmail.com> wrote:
> .oi .uu.  So I guess the song zo si si si doesn't really work then

Well, assuming the PEG grammar becomes official (which it will).

The song may still work, with a slightly different meaning, or it
might be tweaked to work. What are the full lyrics?

Luke Bergen

unread,
Oct 14, 2010, 9:42:17 AM10/14/10
to loj...@googlegroups.com
Well... I guess it could work... it would just bite into whatever came before the song.  It goes:

zo si si si   zo si si si
lo'ai sa'ai le'ai
zo si si si   zo si si si
lo'ai sa'ai le'ai
re kanla sa
re kanla .e lo rebla co'e sa
re kanla sa
re kanla .e lo rebla co'e su

I forget who invented it but I first heard about it from selkik ages ago when I asked how zo interacted with si

2010/10/14 Jorge Llambías <jjlla...@gmail.com>

Jorge Llambías

unread,
Oct 14, 2010, 9:52:28 AM10/14/10
to loj...@googlegroups.com
On Thu, Oct 14, 2010 at 10:42 AM, Luke Bergen <lukea...@gmail.com> wrote:
> Well... I guess it could work... it would just bite into whatever came
> before the song.  It goes:
> zo si si si   zo si si si
> lo'ai sa'ai le'ai
> zo si si si   zo si si si
> lo'ai sa'ai le'ai
> re kanla sa
> re kanla .e lo rebla co'e sa
> re kanla sa
> re kanla .e lo rebla co'e su
> I forget who invented it but I first heard about it from selkik ages ago
> when I asked how zo interacted with si

Oh, so it uses all kinds of unofficial stuff. Not a problem then, it's
almost in a different language. :)

A "si" at the beginning of text is allowed, so the only question is
whether something from the first "lo'ai sa'ai le'ai" gets deleted or
not. has anybody ever worked out the formal grammar rules for "lo'ai
sa'ai le'ai"? I remember there being talk about it, but I don't
remember any conclusions.

Luke Bergen

unread,
Oct 14, 2010, 10:01:53 AM10/14/10
to loj...@googlegroups.com
hmm, yeah, that's a good question.  I don't know if anyone has formally worked out lo'ai sa'ai le'ai.  Maybe I should learn PEG and write up a proposed formal definition.  How does the grammar treat to ... toi?  I would think that lo'ai sa'ai le'ai would have a similar structure. I see it as a metalinguistic comment about previous statements.  I'm guessing that it would probably be too complex to have the grammar go back and make the actual substitutions.

And it could cause problems if one did something like ".ui mi nelci lo selsa'a beme'e lu zo si si si li'u" which, I'm not sure how that would behave.  zo si si gets negated and then the third si kills lu then another zo si si kills itself and another third si kills me'e.

2010/10/14 Jorge Llambías <jjlla...@gmail.com>

Jorge Llambías

unread,
Oct 14, 2010, 11:09:17 AM10/14/10
to loj...@googlegroups.com
On Thu, Oct 14, 2010 at 11:01 AM, Luke Bergen <lukea...@gmail.com> wrote:
> hmm, yeah, that's a good question.  I don't know if anyone has formally
> worked out lo'ai sa'ai le'ai.  Maybe I should learn PEG and write up a
> proposed formal definition.  How does the grammar treat to ... toi?

It's a simple free modifier: "to (text) toi". It can be inserted
(almost) anywhere in another text.

> I would
> think that lo'ai sa'ai le'ai would have a similar structure. I see it as a
> metalinguistic comment about previous statements.  I'm guessing that it
> would probably be too complex to have the grammar go back and make the
> actual substitutions.

Ah, yes, I now seem to recall something like that.

But internally it's more like "lo'u ... le'u" in the sense that it
surrounds any number of words without caring about their selma'o,
right? I fact that's probably where "lo'ai" and "le'ai" come from. If
so, one "si" would delete the whole structure.

> And it could cause problems if one did something like ".ui mi nelci lo
> selsa'a beme'e lu zo si si si li'u" which, I'm not sure how that would
> behave.  zo si si gets negated and then the third si kills lu then another
> zo si si kills itself and another third si kills me'e.

Right, you'd have to quote it with "lo'u ... le'u", or else use two
dummy words after "lu" if you wanted to preserve the grammatical
structure of the song.

Reply all
Reply to author
Forward
0 new messages