experimental cmavo {ko'oi} of UI6 as imperativity

41 views
Skip to first unread message

guskant

unread,
Nov 16, 2012, 12:39:03 PM11/16/12
to lojban
I would like to propose an experimental cmavo {ko'oi} of UI6 as imperativity.

I know an experimental cmavo {xa'e} of LAhE, but it does not satisfy my demand.
When {xa'e} qualifies sumti with a logical connective, a problem occurs.

1) xa'e do .onai lo mi speni co'a morsi
2) xa'e do lu'u .onai xa'e lo mi speni co'a morsi

1) and 2) have different meanings.
{xa'e} of 1) qualifies {do .onai lo mi speni}. The speaker wants one
of them to die.
On the other hand, two {xa'e}s of 2) qualify {do} and {lo mi speni}
respectively. The speaker's imperativity is separated into two parts
with {.onai}, and each of them is therefore indefinite.

A problem occurs when I try to transform them.

The imperativity of 2) can be expressed in two sentences connected with ijek:
2-1) xa'e do co'a morsi .ijonai xa'e lo mi speni co'a morsi
while that of 1) cannot.

Why the imperativity of Lojban is so inflexible, while the
interrogativity of it is quite flexible? I want an imperative cmavo be
of UI6 like {xu} instead of LAhE.

I propose therefore an experimental cmavo {ko'oi} of UI6 as imperativity.
With {ko'oi}, the above mentioned sentences 1) and 2) can be expressed
respectively as follows:

3) do .onai ko'oi lo mi speni co'a morsi
4) do ko'oi .onai lo ko'oi mi speni co'a morsi

and both can be transformed to sentences connected with ijek:

3-1) do co'a morsi .ijonai ko'oi lo mi speni co'a morsi
4-1) do ko'oi co'a morsi .ijonai lo ko'oi mi speni co'a morsi


As a basis of this proposition, I was inspired by an opinion of xorxes
in the following thread:
https://groups.google.com/group/lojban/browse_thread/thread/f3bc98e058baa736
Jorge wrote, Wed, 11 Oct 1995 19:45:21 EDT
> What's more inconsistent is that mi, mi'o, mi'a, ma'a and do'o don't
> have an imperative version. But they are not needed, just as {ko} is
> not really needed. In my opinion, the "imperativity" does not really
> belong in a sumti.
> Consider these:
> au do lo plise mi dunda
> <wish> You give me an apple.
> a'o do lo plise mi dunda
> <hope> You give me an apple.
> e'o do lo plise mi dunda
> <request> You give me an apple.
> e'u do lo plise mi dunda
> <suggestion> You give me an apple.
> e'a do lo plise mi dunda
> <permission> You give me an apple.
> ei do lo plise mi dunda
> <obligation> You give me an apple.
> They have different forms in normal English: "I wish you would give
> me an apple", "I hope you give me an apple", "Please, give me an apple",
> "How about giving me an apple?", "You may give me an apple", "You must
> give me an apple".
> Why should "<command> You give me an apple" be special? Just because
> there is a special tense in English and other languages for that?
> Of course, {ko} may be useful because it's nice and short, but there
> wouldn't really be any loss in expressive power without it.

I consent to this opinion. In fact, in my translation work "lo nenri
be spati denmi", I finally decided to use {e'o} instead of any
"imperative" word:
do .onai .e'o lo mi speni co'a morsi
http://guskant.github.com/yabu/yabu5.html

However, I think one sometimes cannot easily define one's attitudinal
of imperativity, {au}, {a'o}, {e'o}, {e'u}, {e'a} or {ei}. I would
like to express imperativity more easily without wavering between the
attitudinals. For this purpose, the experimental cmavo {ko'oi} is
defined as a broader term than {au}, {a'o}, {e'o}, {e'u}, {e'a} and
{ei}.

Some examples of usage:

ex.1) http://fr.wikisource.org/wiki/Les_Confessions_(Rousseau)/Livre_VI
Enfin je me rappelai le pis-aller d'une grande princesse à qui l'on
disait que les paysans n'avaient pas de pain, et qui répondit: Qu'ils
mangent de la brioche.
ra ko'oi citka lo brioco
It may be expressed rather with a narrower term:
ra .e'u citka lo brioco

ex.2) Puissent vos projets réussir !
lo ko'oi se platu be do ku te snada
With a narrower term:
lo .a'o se platu be do ku te snada
(or possibly {e'o})

ex.3) Que je sois pendu si je mens.
mi ko'oi dandu janai jifsku
With a narrower term:
mi .au dandu janai jifsku
(or possibly {e'u}, {ei})


How do you think about {ko'oi}? Any comments will be appreciated.

la gleki

unread,
Nov 16, 2012, 12:47:48 PM11/16/12
to loj...@googlegroups.com
There was a thread on {xa'e} sev. months ago (actually I started it) with someone replying that {.e'i} is the solution.

guskant

unread,
Nov 16, 2012, 12:54:26 PM11/16/12
to loj...@googlegroups.com
2012/11/17 la gleki <gleki.is...@gmail.com>:
> There was a thread on {xa'e} sev. months ago (actually I started it) with
> someone replying that {.e'i} is the solution.
>

I don't think {.e'i} implies {.e'a}. I need a broader term than the
attitudinals {au}, {a'o}, {e'o}, {e'u}, {e'a} and {ei}.

la gleki

unread,
Nov 16, 2012, 1:22:42 PM11/16/12
to loj...@googlegroups.com
je'e

by the way, here is the link.

selpa'i

unread,
Nov 16, 2012, 3:33:18 PM11/16/12
to loj...@googlegroups.com
Am 16.11.2012 18:39, schrieb guskant:
> How do you think about {ko'oi}? Any comments will be appreciated.

Since xa'e as LAhE is quite awkward, I have proposed to move it to UI.
UI are easier to handle and more flexible. I added an alternate
definition for xa'e on jbovlaste which only needs more votes to become
the preferred one. I'm not sure we need two different cmavo.

mu'o mi'e la selpa'i
--
pilno zo le xu .i lo dei bangu cu se cmene zo lojbo .e nai zo lejbo

doị mèlbi mlenì'u
.i do càtlu ki'u
ma fe la xàmpre ŭu
.i do tìnsa càrmi
gi'e sìrji se tàrmi
.i taị bo pu cìtka lo gràna ku


.


.

Jacob Errington

unread,
Nov 16, 2012, 5:32:40 PM11/16/12
to loj...@googlegroups.com
I created LAhE, and to be perfectly honest, I am discontent with the actual choice of word for it. It should have been something like ko'oi or ko'ei. I also agree that its status as a LAhE is quite silly. Rather than recycle a word that was poorly chosen to begin with, let's use something that will actually look like an attitudinal. How about {.e'ei} ? It looks like {.ei} for "obligation" but it will mean "command". Seeing as usage of xa'e is very limited, I simply propose that we remove it entirely. LAhE-based imperatives is silly.

.i mi'e la tsani mu'o





.

--
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

unread,
Nov 16, 2012, 6:03:42 PM11/16/12
to loj...@googlegroups.com
Am 16.11.2012 23:32, schrieb Jacob Errington:
> I created LAhE, and to be perfectly honest, I am discontent with the
> actual choice of word for it. It should have been something like ko'oi
> or ko'ei. I also agree that its status as a LAhE is quite silly. Rather
> than recycle a word that was poorly chosen to begin with,

Yeah I agree. xa'e might not be the best choice.

> let's use
> something that will actually look like an attitudinal. How about {.e'ei}
> ? It looks like {.ei} for "obligation" but it will mean "command".
> Seeing as usage of xa'e is very limited, I simply propose that we remove
> it entirely. LAhE-based imperatives is silly.

Or ko'oi. At least it resembles ko. I'd say we shouldn't use up .V'VV
cmavo yet. I think the current wave of new cmavo should use only CV'VV
and CVV'V. This will make it easier in the future to see during which
era of Lojban history a cmavo was created.

guskant

unread,
Nov 16, 2012, 6:13:49 PM11/16/12
to loj...@googlegroups.com
.u'u the discussion developped during I was creating a Lojbanic
definition of {xa'e} as UI6. So do you agree with {ko'oi}?

Jacob Errington

unread,
Nov 16, 2012, 7:35:13 PM11/16/12
to loj...@googlegroups.com
{ko'oi} seems fine then. With that in mind, I'll completely erase my old definition of xa'e and downvote it, such that it no longer be returned from searches.


On 16 November 2012 15:13, guskant <gusni...@gmail.com> wrote:
.u'u the discussion developped during I was creating a Lojbanic
definition of {xa'e} as UI6. So do you agree with {ko'oi}?
--
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.

selpa'i

unread,
Nov 16, 2012, 7:56:07 PM11/16/12
to loj...@googlegroups.com
Am 17.11.2012 01:35, schrieb Jacob Errington:
> {ko'oi} seems fine then. With that in mind, I'll completely erase my old
> definition of xa'e and downvote it, such that it no longer be returned
> from searches.

I'm fine with ko'oi too, so I guess I'll remove my xa'e definition as
well to free the word up for a new use.

guskant

unread,
Nov 16, 2012, 8:09:07 PM11/16/12
to loj...@googlegroups.com
2012/11/17 selpa'i <sel...@gmx.de>:
> Am 17.11.2012 01:35, schrieb Jacob Errington:
>
>> {ko'oi} seems fine then. With that in mind, I'll completely erase my old
>> definition of xa'e and downvote it, such that it no longer be returned
>> from searches.
>
>
> I'm fine with ko'oi too, so I guess I'll remove my xa'e definition as well
> to free the word up for a new use.
>
>

OK, so I added {ko'oi} to jbovlaste:
http://jbovlaste.lojban.org/dict/ko'oi
Reply all
Reply to author
Forward
0 new messages