Pivotal sentences in Pandunia. How to express them in Lojban ?

50 views
Skip to first unread message

gleki

unread,
Mar 15, 2012, 12:10:45 PM3/15/12
to lojban-b...@googlegroups.com
Here is an example if a sentence in Pandunia language.
mi peti ti tshau ta. = I ask you to say hello to him. 
parsing: [I ask you greet he/she/it]

So the second "sumti" ti in mi peti ti becomes the first "sumti" of the next construction "ti tshau ta".

I wonder whether it's possible to do so in Lojban. May {be} or {.i ri} can do the trick

selpa'i

unread,
Mar 15, 2012, 1:38:40 PM3/15/12
to lojban-b...@googlegroups.com
mi cpedu fi do fe lo nu vo'i rinsa [ko'a]/zo'e
The vo'i is easable inferrable from context though. (In this case, repeating do would be shorter, too. But with longer sumti, it's a different story)
--
You received this message because you are subscribed to the Google Groups "Lojban Beginners" group.
To view this discussion on the web visit https://groups.google.com/d/msg/lojban-beginners/-/u__MPlGXSw4J.
To post to this group, send email to lojban-b...@googlegroups.com.
To unsubscribe from this group, send email to lojban-beginne...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/lojban-beginners?hl=en.


-- 
.i da xamgu ganse fi no na'ebo lo risna 
.i lo vajrai cu nonselji'u lo kanla

Jonathan Jones

unread,
Mar 15, 2012, 3:50:12 PM3/15/12
to lojban-b...@googlegroups.com
Not bothering with keeping the exact order the same, and allowing the person the request is being made of to be implicit, {mi cpedu lonu do rinsa [ko'a,zo'e,lo prenu, etc.]}
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 )

Oleksii Melnyk

unread,
Mar 16, 2012, 2:31:11 AM3/16/12
to lojban-b...@googlegroups.com
On 15 March 2012 18:10, gleki <gleki.is...@gmail.com> wrote:
So the second "sumti" ti in mi peti ti becomes the first "sumti" of the next construction "ti tshau ta".
I wonder whether it's possible to do so in Lojban. May {be} or {.i ri} can do the trick

>Pandunia ... lacks other complex grammatical structures ...

while lojban mandates them. So, in that example, it's hard to ?factor? the sumti out of the abstraction. In some specific cases you can get close enough with:

1. common set of sumti with 2 «bridi tails» - with reversed order, common comes first
2. NOI, with the hardcoded semantic

--
mi'e lex mu'o

gleki

unread,
Mar 16, 2012, 7:15:32 AM3/16/12
to lojban-b...@googlegroups.com
Of course I know that Pandunia lacks other nice features.
I want Lojban to be the best in every feature, not most features. It's not a competition. 
The victory over other logical projects must be total.

The question is similar to recent discussion about {kakne} - x1 is capable of {x2} which accomplishes the requirement of pivotal sentences whereas many other gismu do not.
{mi cpedu lonu do rinsa [ko'a,zo'e,lo prenu, etc.]} is not the answer. It's too ambiguous. Where is x3 ? Who are you asking this for ?
It's like {mi kakne lo nu do rinsa mi} - an impractical solution that could be logically possible.
Compare with {djica} - {mi djica lo nu do rinsa mi} - that *is* within the baseline.

The major flaw in Lojban is that it cannot compactly express "I ask you to greet me". You have to say
{mi cpedu [fe] lo nu do rinsa [ku] [fi] do}
And here is the lack of compactness (two valsi {do} instead of one).
Even English (not Pandunia !) is better. "I ask you to greet me" - absolutely no need to repeat the word "you".


SO THE MAJOR FLAW OF LOJBAN IS THE LACK OF CONNECTION BETWEEN SUMTI in many gismu.
I can see only one solution.
{mi cpedu fi do poi rinsa} therefore completely ignoring x2 (the act of ignoring at least won't break the baseline).
But I'm not sure whether it preserves the needed meaning.

.arpis.

unread,
Mar 16, 2012, 8:30:42 AM3/16/12
to lojban-b...@googlegroups.com
I disagree that this is a major flaw; this is simply one grammatical
feature lojban lacks (I dont know of any language besides Pandunia
that has it generally; I don't believe that your English example
generalizes).

Also, Pandunia omits unnecessary pronouns, so "I ask you to greet me"
is frequently said "peti tshau mi", which is exactly analogous to
lojban's {cpedu lo nu rinsa mi}.

I remember another grammatical feature I miss in lojban: the inability
to put a selbri at the end of a bridi as an adverb: e.g. "I go from my
house to the store prouldly"
{mi jgira klama lo zdani lo zaisle}; {mi klama lo zdani lo zaisle fi'o
se jgira} is the closest equivalent, but it feels awkward and
unnatural.

Neither of these is a big cost to expressivity pe'i.

> --
> You received this message because you are subscribed to the Google Groups
> "Lojban Beginners" group.
> To view this discussion on the web visit

> https://groups.google.com/d/msg/lojban-beginners/-/uwkF_b01VvIJ.


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

--
mu'o mi'e .arpis.

MorphemeAddict

unread,
Mar 16, 2012, 8:32:38 AM3/16/12
to lojban-b...@googlegroups.com
Indeed, it's not a competition. Therefore there are neither victors nor losers. And Lojban is *so* imperfect, in *so* many ways. But it's still fun to learn and use. 

stevo

On Fri, Mar 16, 2012 at 7:15 AM, gleki <gleki.is...@gmail.com> wrote:

gleki

unread,
Mar 16, 2012, 10:56:17 AM3/16/12
to lojban-b...@googlegroups.com
On Friday, March 16, 2012 4:32:38 PM UTC+4, stevo wrote:
Indeed, it's not a competition. Therefore there are neither victors nor losers. And Lojban is *so* imperfect, in *so* many ways.

Oh please don't tell me that ! Tell me that I'm right in my suggeston for {mi cpedu fi do poi rinsa mi}.
I hope there are no traditions in lojban for expressing this or that. And therefore may be we can invent something that will sound nice and laconic
but still within the baseline.

Or tell me that may be we should try creating another bunch of gismu, may be a hundred or so (zo'o) that will work better.

stevo

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

.arpis.

unread,
Mar 16, 2012, 11:42:29 AM3/16/12
to lojban-b...@googlegroups.com
I would interpret {mi cpedu fi do poi rinsa mi} as "I ask [something]
of you, who greets me." This is not the only interpretation (depending
on context), but the {poi} functions as a description, not the
request.

I respect you too much to lie to you zo'o ru'e, so I can't tell you
that your suggestion is correct.

What I can tell you is that I would translate "I ask you to greet me"
(If I wanted to mention the pronouns) as {mi cpedu lo nu do rinsa mi},
which leaves the person being asked implicit, but taken in the present
tense, I can't easily think of a {te cpedu} that isn't {do}.

>>> To post to this group, send email to lojban-b...@googlegroups.com.


>>> To unsubscribe from this group, send email to

>>> lojban-beginne...@googlegroups.com.


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

> --
> You received this message because you are subscribed to the Google Groups
> "Lojban Beginners" group.
> To view this discussion on the web visit

> https://groups.google.com/d/msg/lojban-beginners/-/7ottFYPBuU0J.
>
> To post to this group, send email to lojban-b...@googlegroups.com.


> To unsubscribe from this group, send email to

> lojban-beginne...@googlegroups.com.


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

--

gleki

unread,
Mar 16, 2012, 1:07:48 PM3/16/12
to lojban-b...@googlegroups.com
OK. At least I can express "I want to sleep" as "mi djica lo nu ri sipna" and I have to memorise just {lonuri}, a lengthy but practical string.
As for {cpedu} it's a more tricky example.
{mi se te se cpedu do lonuri rinsa} would do the trick. As for me, I have to memorise {setesecpedu} instead of {cpedu}.

Why not create a new short cmavo for {lo su'u ri} ?

>>> To post to this group, send email to lojban-beginners@googlegroups.com.


>>> To unsubscribe from this group, send email to


>>> For more options, visit this group at
>>> http://groups.google.com/group/lojban-beginners?hl=en.
>>
>>
> --
> You received this message because you are subscribed to the Google Groups
> "Lojban Beginners" group.
> To view this discussion on the web visit
> https://groups.google.com/d/msg/lojban-beginners/-/7ottFYPBuU0J.
>

> To post to this group, send email to lojban-beginners@googlegroups.com.


> To unsubscribe from this group, send email to


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

Lindar

unread,
Mar 16, 2012, 1:52:14 PM3/16/12
to lojban-b...@googlegroups.com
OK. At least I can express "I want to sleep" as "mi djica lo nu ri sipna" and I have to memorise just {lonuri}, a lengthy but practical string.

There's an implicit {ce'u} there. You don't need {ri} because it's already assumed.
{mi djica lo nu sipna} is {mi djica lo nu ce'u sipna} is {mi djica lo nu mi sipna} is {mi djica lo nu mi sipna vau kei ku vau}.
 

What I can tell you is that I would translate "I ask you to greet me"
(If I wanted to mention the pronouns) as {mi cpedu lo nu do rinsa mi},
which leaves the person being asked implicit, but taken in the present
tense, I can't easily think of a {te cpedu} that isn't {do}.

...or {mi cpedu lo nu rinsa kei do}.
{ko rinsa se va'u mi}
{ko rinsa fi'o se pluka mi}

The problem is that you're being Francocentric about your phrasing. Also, whatever language that is in the first post just looks like coded English. That's why you can get an English sentence to fit so nicely. If all you want is to be able to easily transliterate natlang phrases, just speak Esperanto. I'll stick with a few extra words instead of having to rely on tone and timing.

Also, {mi cpedu fi do poi rinsa mi} is "I ask of You Who Greets Me" kinda like it's a title. {poi} is probably not what you want and that's probably not what you want to say, even if you used {noi}, because it's not even implicit what you're asking.

gleki

unread,
Mar 16, 2012, 2:15:00 PM3/16/12
to lojban-b...@googlegroups.com


On Friday, March 16, 2012 9:52:14 PM UTC+4, Lindar wrote:
OK. At least I can express "I want to sleep" as "mi djica lo nu ri sipna" and I have to memorise just {lonuri}, a lengthy but practical string.

There's an implicit {ce'u} there. You don't need {ri} because it's already assumed.
{mi djica lo nu sipna} is {mi djica lo nu ce'u sipna} is {mi djica lo nu mi sipna} is {mi djica lo nu mi sipna vau kei ku vau}.
 

Don't you think that {mi djica lo nu sipna} is too ambiguous ? Don't you think that it can include such cases as {mi djica lo nu ro sipna} {mi djica lo nu do sipna} {mi djica lo nu mi sipna} {mi djica lo nu mi'o sipna} ?

Oleksii Melnyk

unread,
Mar 16, 2012, 2:35:49 PM3/16/12
to lojban-b...@googlegroups.com


On 16 March 2012 19:07, gleki <gleki.is...@gmail.com> wrote:
"mi djica lo nu ri sipna" {mi se te se cpedu do lonuri rinsa}

For that we have «au mi sipna» and «e'o ko rinsa mi».
 
Why not create a new short cmavo for {lo su'u ri} ?

Considering the worst case «ko'a cpedu lo nu fo'a rinsa do kei fo'a» vs. «fo'a te cpedu {LOSUURI} rinsa do kei ko'a» is only 1 cmavo shorter.
 

Jonathan Jones

unread,
Mar 16, 2012, 2:47:48 PM3/16/12
to lojban-b...@googlegroups.com
On Fri, Mar 16, 2012 at 5:15 AM, gleki <gleki.is...@gmail.com> wrote:
Of course I know that Pandunia lacks other nice features.
I want Lojban to be the best in every feature, not most features. It's not a competition. 
The victory over other logical projects must be total.

The question is similar to recent discussion about {kakne} - x1 is capable of {x2} which accomplishes the requirement of pivotal sentences whereas many other gismu do not.
{mi cpedu lonu do rinsa [ko'a,zo'e,lo prenu, etc.]} is not the answer. It's too ambiguous. Where is x3 ? Who are you asking this for ?

I don't agree that that is ambiguous. If I'm talking to a specific person, and I say that TO that specific person, I think context alone makes it quite clear that the elided cpedu3 is {do}.

After all, who else is capable of causing the event of {do rinsa ko'a}, besides {do}? Nobody. Context alone therefore dictates that {lo te cpedu} must be {do}.

It is true that being explicit in Lojban has the effect of making utterances longer, and that redundancies will occur, however, you don't HAVE to be explicit about everything, and leaving things to context when it is possible to do so tends to make Lojban utterances shorter than the natlang counterparts.

A major advantage of Lojban - at least IMO - is that it is possible to leave quite a lot up to context- something which Asians languages (and possibly others, I don't know) have, but languages like English do not.

--

Oleksii Melnyk

unread,
Mar 16, 2012, 2:57:28 PM3/16/12
to lojban-b...@googlegroups.com
On 16 March 2012 20:15, gleki <gleki.is...@gmail.com> wrote:
Don't you think that {mi djica lo nu sipna} is too ambiguous ? Don't you think that it can include such cases as {mi djica lo nu ro sipna} {mi djica lo nu do sipna} {mi djica lo nu mi sipna} {mi djica lo nu mi'o sipna} ?

«mi djica lo nu MI sipna» is also too ambiguous. What I really want is to sleep for 9 hours in a quiet and dark room, with a thin blanket and pillow,  22°C warm  20%(volume) oxygen 2% water air, on a soft bed, horizontal in a gravity field, slightly less then 10m/s², less is better...  But who cares?

iesk

unread,
Mar 16, 2012, 3:13:54 PM3/16/12
to Lojban Beginners
On Mar 16, 6:07 pm, gleki <gleki.is.my.n...@gmail.com> wrote:
> OK. At least I can express "I want to sleep" as "mi djica lo nu ri sipna"
> and I have to memorise just {lonuri}, a lengthy but practical string.

Actually, you can’t, can you? See CLL, chapter 7, section 6 (I’d link
but the web version seems to be down while I’m writing this). AFAIK
{ri}’s backcounting ignores {mi} (which I always thought made sense,
though).

mu'o
mi'e ·iesk·

Felipe Gonçalves Assis

unread,
Mar 16, 2012, 4:03:49 PM3/16/12
to lojban-b...@googlegroups.com
On 16 March 2012 14:52, Lindar <lindar...@gmail.com> wrote:
>> OK. At least I can express "I want to sleep" as "mi djica lo nu ri sipna"
>> and I have to memorise just {lonuri}, a lengthy but practical string.
>
>
> There's an implicit {ce'u} there. You don't need {ri} because it's already
> assumed.
> {mi djica lo nu sipna} is {mi djica lo nu ce'u sipna} is {mi djica lo nu mi
> sipna} is {mi djica lo nu mi sipna vau kei ku vau}.
>

There is no implicit {ce'u}. Implicit {ce'u} only happen in ka-abstractions,
where a {ce'u} must occur. {mi djica lo nu sipna} could very well be
used to mean {mi djica lo nu lo cifnu cu sipna} or even
{mi djica lo nu da sipna}.

By the way, I never understood why nu-abstractions are used in djica2.
There is no specific event you desire for. You just want an event with certain
properties to happen. Maybe the more abstract {du'u} would express that?

Felipe Gonçalves Assis

unread,
Mar 16, 2012, 4:21:40 PM3/16/12
to lojban-b...@googlegroups.com
On 16 March 2012 17:03, Felipe Gonçalves Assis <felipe...@gmail.com> wrote:
> By the way, I never understood why nu-abstractions are used in djica2.
> There is no specific event you desire for. You just want an event with certain
> properties to happen. Maybe the more abstract {du'u} would express that?

On a more logical level, there is a problem when you try to express desire
for something impossible. That is
1. Let broda be a predication that is impossible to hold, even in hypothetical
worlds, such as a logical contradiction.
2. Let ko'a be a subject with so little grasp of logic that ko'a can not
understand that broda is impossible, and naively wants broda.
How would the speaker express that ko'a wants broda? He can't say
{lo nu broda}, because there is no event of broda. There is no problem
with the existence of the predication broda, on the other hand.

mu'o
mi'e .asiz.

ianek

unread,
Mar 16, 2012, 4:39:47 PM3/16/12
to Lojban Beginners
I thought that under xorlo {lo broda} doesn't imply the existence of
broda. (And likewise, {lo nu broda} doesn't imply the existence of an
event of broda).

mu'o mi'e ianek

On 16 Mar, 21:21, Felipe Gonçalves Assis <felipeg.as...@gmail.com>
wrote:

Lindar

unread,
Mar 16, 2012, 7:57:58 PM3/16/12
to lojban-b...@googlegroups.com
UI is not a replacement for whole phrases. 《.au mi sipna》means 《mi sipna》 and you feel desire about it. 《.e'o rinsa》 means 《rinsa》 and you're suggesting something by it.

Robin Lee Powell

unread,
Mar 16, 2012, 8:00:29 PM3/16/12
to lojban-b...@googlegroups.com

That is not currently the case in the language, as you are well
aware, given your note at the end of
http://www.lojban.org/tiki/BPFK%20Section:%20Irrealis%20Attitudinals

{.au mi sipna} most definitely does *not* currently assert {mi
sipna}.

-Robin

--
http://singinst.org/ : Our last, best hope for a fantastic future.
.i ko na cpedu lo nu stidi vau loi jbopre .i danfu lu na go'i li'u .e
lu go'i li'u .i ji'a go'i lu na'e go'i li'u .e lu go'i na'i li'u .e
lu no'e go'i li'u .e lu to'e go'i li'u .e lu lo mamta be do cu sofybakni li'u

Felipe Gonçalves Assis

unread,
Mar 16, 2012, 8:33:25 PM3/16/12
to lojban-b...@googlegroups.com
On 16 March 2012 17:39, ianek <jan...@gmail.com> wrote:
> I thought that under xorlo {lo broda} doesn't imply the existence of
> broda. (And likewise, {lo nu broda} doesn't imply the existence of an
> event of broda).
>

Quite the opposite, actually.

mu'o
mi'e .asiz.

> mu'o mi'e ianek
>
> On 16 Mar, 21:21, Felipe Gonçalves Assis <felipeg.as...@gmail.com>
> wrote:
>> On 16 March 2012 17:03, Felipe Gonçalves Assis <felipeg.as...@gmail.com> wrote:
>>
>> > By the way, I never understood why nu-abstractions are used in djica2.
>> > There is no specific event you desire for. You just want an event with certain
>> > properties to happen. Maybe the more abstract {du'u} would express that?
>>
>> On a more logical level, there is a problem when you try to express desire
>> for something impossible. That is
>>  1. Let broda be a predication that is impossible to hold, even in hypothetical
>>      worlds, such as a logical contradiction.
>>  2. Let ko'a be a subject with so little grasp of logic that ko'a can not
>>      understand that broda is impossible, and naively wants broda.
>> How would the speaker express that ko'a wants broda? He can't say
>> {lo nu broda}, because there is no event of broda. There is no problem
>> with the existence of the predication broda, on the other hand.
>>
>> mu'o
>> mi'e .asiz.
>

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

> To post to this group, send email to lojban-b...@googlegroups.com.
> To unsubscribe from this group, send email to lojban-beginne...@googlegroups.com.

Jonathan Jones

unread,
Mar 16, 2012, 9:05:32 PM3/16/12
to lojban-b...@googlegroups.com
On Fri, Mar 16, 2012 at 6:00 PM, Robin Lee Powell <rlpo...@digitalkingdom.org> wrote:
On Fri, Mar 16, 2012 at 04:57:58PM -0700, Lindar wrote:
> UI is not a replacement for whole phrases. 《.au mi sipna》means
> 《mi sipna》 and you feel desire about it. 《.e'o rinsa》 means
> 《rinsa》 and you're suggesting something by it.

That is not currently the case in the language, as you are well
aware, given your note at the end of
http://www.lojban.org/tiki/BPFK%20Section:%20Irrealis%20Attitudinals

{.au mi sipna} most definitely does *not* currently assert {mi
sipna}.

-Robin

My understanding of that note means that it could mean any combination of "I desire that I sleep." and "I'm sleeping, about which I feel desire."

Personally, I don't see much use for the second interpretation, but based on the note, it's a valid one.

--

Robin Lee Powell

unread,
Mar 17, 2012, 12:12:03 AM3/17/12
to lojban-b...@googlegroups.com
On Fri, Mar 16, 2012 at 07:05:32PM -0600, Jonathan Jones wrote:
> On Fri, Mar 16, 2012 at 6:00 PM, Robin Lee Powell <
> rlpo...@digitalkingdom.org> wrote:
>
> > On Fri, Mar 16, 2012 at 04:57:58PM -0700, Lindar wrote:
> > > UI is not a replacement for whole phrases. 《.au mi
> > > sipna》means 《mi sipna》 and you feel desire about it. 《.e'o
> > > rinsa》 means 《rinsa》 and you're suggesting something by it.
> >
> > That is not currently the case in the language, as you are well
> > aware, given your note at the end of
> > http://www.lojban.org/tiki/BPFK%20Section:%20Irrealis%20Attitudinals
> >
> > {.au mi sipna} most definitely does *not* currently assert {mi
> > sipna}.
> >
> > -Robin
> >
>
> My understanding of that note means that it could mean any
> combination of "I desire that I sleep." and "I'm sleeping, about
> which I feel desire."
>
> Personally, I don't see much use for the second interpretation,
> but based on the note, it's a valid one.

Sure, if you read *the note*. The note is Lindar's opinion, it is
not the current state of the language. The current state is
represented by the rest of the page.

-Robin

Jonathan Jones

unread,
Mar 17, 2012, 12:17:34 AM3/17/12
to lojban-b...@googlegroups.com
Well, as long as the current state is that the latter is wrong, I'm happy, then.

Robin Lee Powell

unread,
Mar 17, 2012, 12:19:17 AM3/17/12
to lojban-b...@googlegroups.com

Yes, currently {.au mi sipna} means only "I want to sleep", or
something very much like it.

gleki

unread,
Mar 17, 2012, 9:12:35 AM3/17/12
to lojban-b...@googlegroups.com
Oh, thank you for pointing it out. Indeed, {ri} doesn't count {mi}.
Can we fix the problem by inventing a new cmavo ?

Please, give me place for a new experimental cmavo that will mean {lo su'u (any last sumti of the current selbri)}.
What is the current situation with Free Cmavo Space ?
AFAI remember, there are still free space for
ja'u ne'e bi'a ci'a zi'a zi'i zi'u te'i po'a 
bo'V
and many cmavo that are inthe form of xV'V
and many more CV'Vi

Could you please kindly provide me with some cmavo so that I (and at least one more jbopre) can start using them in order to implement complex verbs in Lojban ?
I guess most experimental cmavo have been used by not more than 2-3 jbopre each.

MorphemeAddict

unread,
Mar 17, 2012, 1:01:15 PM3/17/12
to lojban-b...@googlegroups.com
This list is for beginners. I recommend learning Lojban as it is before suggesting changes, and making those suggestions on the main Lojban list. 

stevo

--
You received this message because you are subscribed to the Google Groups "Lojban Beginners" group.
To view this discussion on the web visit https://groups.google.com/d/msg/lojban-beginners/-/KFZbXvCz9roJ.
To post to this group, send email to lojban-b...@googlegroups.com.
To unsubscribe from this group, send email to lojban-beginne...@googlegroups.com.

Michael Turniansky

unread,
Mar 23, 2012, 3:32:33 PM3/23/12
to lojban-b...@googlegroups.com
  Incidentally, you do know you don't have to change the order of the arguments of the main bridi from the default, right? You can use a vo'i that hasn't been defined yet -- "mi cpedu lo nu vo'i rinsa ko'a kei do" (I know that wasn't the point -- you were trying to mimic the structure of the original.  Just pointing it out in case you didn't know)  But to the main question, you could leave the x1 out of the abstraction entirely.  It doesn't necessarily default to the x1 of the outer bridi.  It's still "zo'e" and in most cases, it would be more reasonable to assume I'm asking you do take an action then asking if I can greet him.
         --gejyspa


On Thu, Mar 15, 2012 at 1:38 PM, selpa'i <m...@plasmatix.com> wrote:
mi cpedu fi do fe lo nu vo'i rinsa [ko'a]/zo'e
The vo'i is easable inferrable from context though. (In this case, repeating do would be shorter, too. But with longer sumti, it's a different story)

Am 15.03.2012 17:10, schrieb gleki:
Here is an example if a sentence in Pandunia language.
mi peti ti tshau ta. = I ask you to say hello to him. 
parsing: [I ask you greet he/she/it]

So the second "sumti" ti in mi peti ti becomes the first "sumti" of the next construction "ti tshau ta".

I wonder whether it's possible to do so in Lojban. May {be} or {.i ri} can do the trick

--
You received this message because you are subscribed to the Google Groups "Lojban Beginners" group.
To view this discussion on the web visit https://groups.google.com/d/msg/lojban-beginners/-/u__MPlGXSw4J.

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


-- 
.i da xamgu ganse fi no na'ebo lo risna 
.i lo vajrai cu nonselji'u lo kanla

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

gleki

unread,
Mar 24, 2012, 4:41:48 AM3/24/12
to lojban-b...@googlegroups.com
The problem with vo'i that it refers to a sumti with fixed position (No.3 in this case) whereas Pandunia chain sentences can theoretically be very long and (last sumti) marker would work much better. In Lojban it's {ri} except when last sumti is {mi/do} or similar.


On Friday, March 23, 2012 11:32:33 PM UTC+4, lincro wrote:
  Incidentally, you do know you don't have to change the order of the arguments of the main bridi from the default, right? You can use a vo'i that hasn't been defined yet -- "mi cpedu lo nu vo'i rinsa ko'a kei do" (I know that wasn't the point -- you were trying to mimic the structure of the original.  Just pointing it out in case you didn't know)  But to the main question, you could leave the x1 out of the abstraction entirely.  It doesn't necessarily default to the x1 of the outer bridi.  It's still "zo'e" and in most cases, it would be more reasonable to assume I'm asking you do take an action then asking if I can greet him.
         --gejyspa
On Thu, Mar 15, 2012 at 1:38 PM, selpa'i <m...@plasmatix.com> wrote:
mi cpedu fi do fe lo nu vo'i rinsa [ko'a]/zo'e
The vo'i is easable inferrable from context though. (In this case, repeating do would be shorter, too. But with longer sumti, it's a different story)

Am 15.03.2012 17:10, schrieb gleki:
Here is an example if a sentence in Pandunia language.
mi peti ti tshau ta. = I ask you to say hello to him. 
parsing: [I ask you greet he/she/it]

So the second "sumti" ti in mi peti ti becomes the first "sumti" of the next construction "ti tshau ta".

I wonder whether it's possible to do so in Lojban. May {be} or {.i ri} can do the trick

--
You received this message because you are subscribed to the Google Groups "Lojban Beginners" group.
To view this discussion on the web visit https://groups.google.com/d/​msg/lojban-beginners/-/u__​MPlGXSw4J.
To post to this group, send email to lojban-beginners@googlegroups.​com.
To unsubscribe from this group, send email to lojban-beginners+unsubscribe@​googlegroups.com.

For more options, visit this group at http://groups.google.com/​group/lojban-beginners?hl=en.
-- 
.i da xamgu ganse fi no na'ebo lo risna 
.i lo vajrai cu nonselji'u lo kanla

--
You received this message because you are subscribed to the Google Groups "Lojban Beginners" group.
To post to this group, send email to lojban-beginners@googlegroups.​com.
To unsubscribe from this group, send email to lojban-beginners+unsubscribe@​googlegroups.com.
Reply all
Reply to author
Forward
0 new messages