The x3 of {srana}

12 views
Skip to first unread message

Jacob Errington

unread,
Jun 29, 2012, 12:57:37 PM6/29/12
to loj...@googlegroups.com
coi rodo

Twice now, I've had to use {sekai} in order to get a sort of "complete
srana" such that "srana = x1 pertains to x2 in aspect x3".
I feel like this should be a part of the place structure of srana
proper. Adding a place to a gismu breaks no usage, as we all know, so
there is no fear of corrupting existing Lojban.

Does anyone agree?

mu'o mi'e la tsani

p.s. I'm not really *proposing* that we effect this change *now*. But,
I'm just saying that it would be really neat, if maybe when a bigger
reworking of the language comes around, that we add this place, which
seems really srana to srana, if you know what I mean.

I'll also be adding this to the gismu issues page.

la .lindar.

unread,
Jun 29, 2012, 1:48:42 PM6/29/12
to loj...@googlegroups.com
Makes sense to me.
I had to go check to see that it didn't have that already.
Weird that it doesn't.

Jonathan Jones

unread,
Jun 29, 2012, 1:49:24 PM6/29/12
to loj...@googlegroups.com
I agree. There's actually a lot of gismu that I feel are missing a place, most of which are specifically for reasons of regularity, i.e. all the other gismu of the same class have place structure "x1 is broda of x2 in way x3", and this one is only "x1 is broda of x2".


--
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.




--
mu'o mi'e .aionys.

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

la .lindar.

unread,
Jun 30, 2012, 12:13:48 AM6/30/12
to loj...@googlegroups.com
Meanwhile there are a bunch of gismu with superfluous places. =P
Maybe they should trade?

Jonathan Jones

unread,
Jun 30, 2012, 12:41:05 AM6/30/12
to loj...@googlegroups.com
I'm a bit less agreeable to removing possibly extraneous places than to adding apparently needed ones.

After all, you can always just /not/ use those places you find extraneous.

On Fri, Jun 29, 2012 at 10:13 PM, la .lindar. <lindar...@gmail.com> wrote:
Meanwhile there are a bunch of gismu with superfluous places. =P
Maybe they should trade?

--
You received this message because you are subscribed to the Google Groups "lojban" group.
To view this discussion on the web visit https://groups.google.com/d/msg/lojban/-/oIKuSSogC9kJ.

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

Ian Johnson

unread,
Jun 30, 2012, 2:55:26 AM6/30/12
to loj...@googlegroups.com
Sometimes these are like that because "this gismu is that gismu but without this being inherent to it". I agree in many cases, though.

mu'o mi'e la latro'a noi ge lo bebna cu cmene ke'a gi zo latros pu cmene ke'a

Jonathan Jones

unread,
Jun 30, 2012, 3:03:42 AM6/30/12
to loj...@googlegroups.com
Absolutely none of the gismu I am referring to are like that. I'd have to actually look to find out, but there are for example, the familial gismu, of which all are "x1 is [famillial relationship] to x2 by bond/tie x3" except those which are missing the x2, and IIRC some of the measurement gismu do not have a subunits place, some of the metric measurement gismu do not have a by standard place, etc.

I'm not referring to, for example, the difference between {klama} and {litru}, which are similar relationships, but to those gismu which are in the same category of relationship. ("Brother" and "Cousin" are not similar relationships but are of the same category (for my current chosen meaning of similar), for example.)

gleki

unread,
Jun 30, 2012, 3:05:51 AM6/30/12
to loj...@googlegroups.com
sorry for a question from a nintadni. Can't {sekai} be used as an alternative to x3-srana and similar brivla?


On Friday, June 29, 2012 9:49:24 PM UTC+4, aionys wrote:
I agree. There's actually a lot of gismu that I feel are missing a place, most of which are specifically for reasons of regularity, i.e. all the other gismu of the same class have place structure "x1 is broda of x2 in way x3", and this one is only "x1 is broda of x2".

On Fri, Jun 29, 2012 at 10:57 AM, Jacob Errington <nict...@gmail.com> wrote:
coi rodo

Twice now, I've had to use {sekai} in order to get a sort of "complete
srana" such that "srana = x1 pertains to x2 in aspect x3".
I feel like this should be a part of the place structure of srana
proper. Adding a place to a gismu breaks no usage, as we all know, so
there is no fear of corrupting existing Lojban.

Does anyone agree?

mu'o mi'e la tsani

p.s. I'm not really *proposing* that we effect this change *now*. But,
I'm just saying that it would be really neat, if maybe when a bigger
reworking of the language comes around, that we add this place, which
seems really srana to srana, if you know what I mean.

I'll also be adding this to the gismu issues page.

--
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.

Jonathan Jones

unread,
Jun 30, 2012, 3:11:39 AM6/30/12
to loj...@googlegroups.com
No, but only because srana has no x3. IFF srana had the x3 "in aspect", then yes.

The point of the discussion is that we think srana /should/ have that x3, thus negating the need to use sekai.

To view this discussion on the web visit https://groups.google.com/d/msg/lojban/-/mR_hQRaAV6QJ.

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

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

Jonathan Jones

unread,
Jun 30, 2012, 3:13:05 AM6/30/12
to loj...@googlegroups.com
On Sat, Jun 30, 2012 at 1:03 AM, Jonathan Jones <eye...@gmail.com> wrote:
... except those which are missing the x2....

I meant "missing the x3".

Ian Johnson

unread,
Jun 30, 2012, 2:43:39 PM6/30/12
to loj...@googlegroups.com
It's a matter of perspective, basically. You can say that the ones which delete places for reason of non-necessity are irregular, but if so they are still useful, or you can say they are different entirely, in which case it's a moot point. In either case there's no real disagreement here.

mu'o mi'e la latro'a

Jonathan Jones

unread,
Jun 30, 2012, 2:54:21 PM6/30/12
to loj...@googlegroups.com
I'm not saying either, though. I think those gismu which lack the places of their compatriots /do/ need them, and not merely because their lack makes them irregular. For example, if you wanted to say that ko'a is the adopted sister of ko'e, you could easily do so, as mensi is "x1 is the sister of x2 by bond/tie x3". However, if you wanted to say that ko'a is the adopted daughter of ko'i, you have a more difficult time, because tixnu lacks the by bond x3: "x1 is the daughter of parent(s) x2 (no necessarily biological)".

Jonathan Jones

unread,
Jun 30, 2012, 2:56:54 PM6/30/12
to loj...@googlegroups.com
I mean, can you even name the BAI needed to replace the missing x3? I can't. I might be able to if I scanned through them all, but certainly not off the top of my head.
Reply all
Reply to author
Forward
0 new messages