Time for the perenial other-centric-.ui conversation

21 views
Skip to first unread message

Luke Bergen

unread,
Nov 27, 2010, 1:45:47 PM11/27/10
to lojba...@lojban.org

That's actually not what I intend but it is what will inevitably happen.

I was talking to my wife who is perpetually freezing and at one point I remembered somehing and wanted to say "oh yeah!  Your dad said he was going to make a fire tonight _____".  Where "______" was that cool new cmavo of COI that takes a UI before it and somehow applies it to the object of the COI.  I've searched around but must be using the wrong search terms.  Anybody remember what I'm talking about?

What I'm trying to express is something like what you want to express when you play peekaboo with a baby.  It's like some kind of .uadai like "surprise!"  But it's not a command or an observation, but more of like an expectation.  Almost like a .uipeikau if that makes any sense?

Remo Dentato

unread,
Nov 27, 2010, 2:14:03 PM11/27/10
to loj...@googlegroups.com
jbovlaste has {da'oi} : http://vlasisku.lojban.org/da%27oi is it what you mean?

> --
> 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.
>

John E Clifford

unread,
Nov 27, 2010, 3:12:12 PM11/27/10
to loj...@googlegroups.com
Well, this raises the interesting OT issue of what exactly something like "peekaboo" is linguistically.  It seems to be related to "taDA" and magicians' exclamations when the the trick comes off, but it seems to have a bit more content than some attention getter in COI, since the seeing is vital.  And, of course, I am not sure what the linguistic status of the supposedly similar things is either.  Come to think of it, there is a double seeing here -- the child sees the parent again and the parent sees the child ("I see you" maybe left over from its origin in a hide-and-seek game).  "Surprise" seems to be a totally different sort of thing. Or maybe no so different: it too is used to announce the completion of a trick (even if the recipient is not surprised).  But it clearly is not, as this note implies, a case of the speaker empathetically experiencing the emotion of the recipient -- the speaker, a part of the party, is not surprised.  He might be joyful and also experience some empathetic joy with the recipient, but that isn't surprise. I'd be happy to hear more suggestions about this.
I would not be happy to see cmavo used to indicate emotions of a third party not empathized by the speaker.



From: Luke Bergen <lukea...@gmail.com>
To: "lojba...@lojban.org" <lojba...@lojban.org>
Sent: Sat, November 27, 2010 12:45:47 PM
Subject: [lojban] Time for the perenial other-centric-.ui conversation

That's actually not what I intend but it is what will inevitably happen.

I was talking to my wife who is perpetually freezing and at one point I remembered somehing and wanted to say "oh yeah!  Your dad said he was going to make a fire tonight _____".  Where "______" was that cool new cmavo of COI that takes a UI before it and somehow applies it to the object of the COI.  I've searched around but must be using the wrong search terms.  Anybody remember what I'm talking about?

What I'm trying to express is something like what you want to express when you play peekaboo with a baby.  It's like some kind of .uadai like "surprise!"  But it's not a command or an observation, but more of like an expectation.  Almost like a .uipeikau if that makes any sense?

--

Luke Bergen

unread,
Nov 27, 2010, 3:25:56 PM11/27/10
to loj...@googlegroups.com

Yep, da'oi is exactly what I was looking for.  ki'e la .remo.

See, john that's actually what I AM looking for.  I expect that's somewhat how da'oi to be used.

I would read a magician saying ".ue.u'e da'oi [do]" as something like english "tada" or as a shortcut for the lojban ".ue.u'epeikau" like "surprised and in awe, aintcha".  This seems like a really handy feature to me.

On Nov 27, 2010 3:12 PM, "John E Clifford" <kali9...@yahoo.com> wrote:

John E Clifford

unread,
Nov 27, 2010, 3:37:09 PM11/27/10
to loj...@googlegroups.com
But, of course, "surprised and in awe, aintcha" is merely a derived sentence fragment, not an expression of a feeling.  The magician is not surprised or awed, nor does he feel that of the watcher empathetically.  He may feel nothing at all or pride in duping the rubes or embarrassment for the same reason and so on.  I agree that 'da'oi' as described is what you are looking for but you are looking in the wrong place -- it ain't in UI and related things but in plain old bridi, describing the emotions of another.  'da'oi' is merely a conceptual mistake brought on by the ever-renewed habit of using UI instead of bridi.  Notice, "surprised and in awe" can be denied, declared false; a UI cannot (though it may be genuine or faked).


From: Luke Bergen <lukea...@gmail.com>
To: loj...@googlegroups.com
Sent: Sat, November 27, 2010 2:25:56 PM
Subject: Re: [lojban] Time for the perenial other-centric-.ui conversation

Luke Bergen

unread,
Nov 27, 2010, 4:01:50 PM11/27/10
to loj...@googlegroups.com

So what you're saying is, in english when I see someone faceplant and say "oooooo, that looked like it hurt", there simply isn't a way to say this in lojban.  I don't feel pain so {.oiro'odai ta simlu lo ka cortu} is wrong.  To translate english "ooooo" in this context, how would YOU say it john?  In lojban please.  For me, using lojban in everyday life, being able to assert that others UI (with the implicit pe'i) like in ".oiro'o da'oi la nik ta simlu lo ka cortu (where nick has just faceplanted)" is hella useful. 

So just for clarification from other users of da'oi, am I doing it right? Or should I just start using {.oipeikau} (dunno why I hadn't thought of that before).

On Nov 27, 2010 3:37 PM, "John E Clifford" <kali9...@yahoo.com> wrote:

Remo Dentato

unread,
Nov 27, 2010, 4:17:04 PM11/27/10
to loj...@googlegroups.com
On Sat, Nov 27, 2010 at 10:01 PM, Luke Bergen <lukea...@gmail.com> wrote:
> So just for clarification from other users of da'oi, am I doing it right? Or
> should I just start using {.oipeikau} (dunno why I hadn't thought of that
> before).

I've not used {da'oi} yet but being of selma'o DOI I expect I'll use it like:

{mi viska da'oi .liuk. .ue }

to indicate that you were surprise that I had seen you, and

{mi viska doi .liuk. .ue }

to indicate that I was surprised to see you.

John E Clifford

unread,
Nov 27, 2010, 4:23:41 PM11/27/10
to loj...@googlegroups.com
Before I translate it into Lojban, I want to know what you mean by it.  As far as I can see, this has nothing to do with the topic at hand (da'oi).  It appears to be a expression of an emotion, either a sympathetic one of pain ("Owie") or a primary one of some level of discomfort ("Eeew"), followed by a declarative sentence (the expression goes with but does not modify the sentence) about how something looks (presumably open wounds, perhaps contorted face, perhaps just knowledge of how that event has felt when you did what looked like the same thing).  But none of this involves  the use of 'da'oi': the sentence is just a sentence and the long o is either a UI or a UI dai depending.on which expression you meant.

Sent: Sat, November 27, 2010 3:01:50 PM

Subject: Re: [lojban] Time for the perenial other-centric-.ui conversation

So what you're saying is, in english when I see someone faceplant and say "oooooo, that looked like it hurt", there simply isn't a way to say this in lojban.  I don't feel pain so {.oiro'odai ta simlu lo ka cortu} is wrong.  To translate english "ooooo" in this context, how would YOU say it john?  In lojban please.  For me, using lojban in everyday life, being able to assert that others UI (with the implicit pe'i) like in ".oiro'o da'oi la nik ta simlu lo ka cortu (where nick has just faceplanted)" is hella useful. 

So just for clarification from other users of da'oi, am I doing it right? Or should I just start using {.oipeikau} (dunno why I hadn't thought of that before).

On Nov 27, 2010 3:37 PM, "John E Clifford" <kali9...@yahoo.com> wrote:

Luke Bergen

unread,
Nov 27, 2010, 4:24:25 PM11/27/10
to loj...@googlegroups.com

Oh, from vlasisku I thought that the /preceding/ UI was what was dai'ed to the DOIed person.  E.g. "mi viska do .ue da'oi la .remo." for "I see you aren't you surprised la remo" vs "mi viska do .ue doi la .remo."  -> "I see you (surprised that it's /you/), remo"

John E Clifford

unread,
Nov 27, 2010, 4:27:19 PM11/27/10
to loj...@googlegroups.com
And what emotion (etc.) did the speaker feel when the listener was surprised to
see him? Nothing relevant comes to mind and it very probably was not surprise,
even empathetically.


----- Original Message ----
From: Remo Dentato <rden...@gmail.com>
To: loj...@googlegroups.com
Sent: Sat, November 27, 2010 3:17:04 PM
Subject: Re: [lojban] Time for the perenial other-centric-.ui conversation

--

John E Clifford

unread,
Nov 27, 2010, 4:33:21 PM11/27/10
to loj...@googlegroups.com
But the first is just two sentences (maybe the second is a rhetorical question, but probably not) and the second just an ordinary UI attached to a vocative to indicate (pragmatically) that the surprise is at the identity of the object in view.  So, in the first 'da'oi' is misused (or rather used according to instruction even though the instructions make no sense) and the second doesn't even need the empathetic modifier, let alone the more remote connection that 'da'oi' suggests.

Sent: Sat, November 27, 2010 3:24:25 PM

Subject: Re: [lojban] Time for the perenial other-centric-.ui conversation

Remo Dentato

unread,
Nov 27, 2010, 4:40:38 PM11/27/10
to loj...@googlegroups.com
On Sat, Nov 27, 2010 at 10:24 PM, Luke Bergen <lukea...@gmail.com> wrote:
> Oh, from vlasisku I thought that the /preceding/ UI was what was dai'ed to
> the DOIed person.  E.g. "mi viska do .ue da'oi la .remo." for "I see you
> aren't you surprised la remo" vs "mi viska do .ue doi la .remo."  -> "I see
> you (surprised that it's /you/), remo"

Well, vlasisku just says that {da'oi} assigns UI to someone else than
the speaker and it is part of the DOI selma'o. So my understanding is
that like {doi} sets who the listener is, {da'oi} sets who the
"feeler" is. Then the UI of that sentence are tied to the preceding
sumti.

John, UI and brivla are not substitutes one of the other. If I use a
UI to express an emotion is to convey something different than just
"sayng" that the emotion is felt.

If in a natural language I would say something like "I love Alice" I
would translate with "mi prami la .alis." but if I would say "Oh! How
I love Alice!" I would render it with "mi la ,alis..au prami". I might
get it wrong but it seems to me that you consider UI as a bad idea.

.remod.

Luke Bergen

unread,
Nov 27, 2010, 4:42:33 PM11/27/10
to loj...@googlegroups.com

Here you go.  I leave for work and then 20 minutes later I come home for some reason to find remo putting the moves on my wife.  So I sneak into the room with a baseball bat and think "oh perfect opportunity to use my lojban IRL".  So I bust in bat a blazin' and say "do na kanpe lo nu mi cazi sevxruti .iicai.ue da'oi la .remo.  .i ko mrobi'o doi pe'a kalci".

Ignoring the fact that I don't know what remo looks like, this seems like a decent example of "da'oi"s usage

On Nov 27, 2010 4:33 PM, "John E Clifford" <kali9...@yahoo.com> wrote:

Remo Dentato

unread,
Nov 27, 2010, 4:42:50 PM11/27/10
to loj...@googlegroups.com
On Sat, Nov 27, 2010 at 10:27 PM, John E Clifford <kali9...@yahoo.com> wrote:
> And what emotion (etc.) did the speaker feel when the listener was surprised to
> see him?  Nothing relevant comes to mind and it very probably was not surprise,
> even empathetically.

Exactly. The speaker emotion in {mi viska da'oi .liuk. .ue } is not
specified. There's no need for it: either the speaker doesn't want to
say or it is clear from the context.

Remo Dentato

unread,
Nov 27, 2010, 4:44:37 PM11/27/10
to loj...@googlegroups.com
On Sat, Nov 27, 2010 at 10:42 PM, Luke Bergen <lukea...@gmail.com> wrote:
> Here you go.  I leave for work and then 20 minutes later I come home for
> some reason to find remo putting the moves on my wife.  So I sneak into the
> room with a baseball bat and think "oh perfect opportunity to use my lojban
> IRL".  So I bust in bat a blazin' and say "do na kanpe lo nu mi cazi
> sevxruti .iicai.ue da'oi la .remo.  .i ko mrobi'o doi pe'a kalci".

Let me say I'm really happy you though of such a long sentence. I
might have the opportunity to ran away before you finish saying it! :)

Luke Bergen

unread,
Nov 27, 2010, 4:49:12 PM11/27/10
to loj...@googlegroups.com

Well obviously this is what I'd be yelling WHILE I was running into the room "bat a blazin' "  ;)

And the bit I was talking about earlier is that vlasisku says that it takes the preceding attitudinal (note that it is singular attitudinal).  So I would read "mi viska do .ue da'oi la .remo. .ui" as "I see you" with you being surprised and me being happy

On Nov 27, 2010 4:44 PM, "Remo Dentato" <rden...@gmail.com> wrote:

Remo Dentato

unread,
Nov 27, 2010, 4:58:58 PM11/27/10
to loj...@googlegroups.com
On Sat, Nov 27, 2010 at 10:49 PM, Luke Bergen <lukea...@gmail.com> wrote:
> And the bit I was talking about earlier is that vlasisku says that it takes
> the preceding attitudinal (note that it is singular attitudinal).

You're right, I had missed that part. We should ask to the creator of
that cmavo, if he/she is still around here ....

John E Clifford

unread,
Nov 27, 2010, 5:05:12 PM11/27/10
to loj...@googlegroups.com
No, UI, properly used, was a great idea (I'm not too sure about some of the
Lojban elaborations, but the idea is still excellent). It allows us to make
distinctions that are impossible in English (well, very hard, any how). Take an
English sentence (I think this works for most naturla languages in the case of
at least some emotions) "I am sorry I spilled the drink". Assuming for the
moment that I did spill the drink (what happens if I didn't goes off in into
philosophical cloud-cuckoo-land), this sentences has two different uses. On is
a *report* about how I feel -- it may be true or false, depending upon whether I
really do feel sorry. The other is an *expression* of my regret. the latter is
neither true nor false (though it may be sincere or not); it has the same
logical force as "Oops."
As for the difference between saying that someone else has an emotion (that I do
not share empathetically) and .... what? It can't be expressing an emotion
since, by the assumption, I don't feel it, and have no reason to fake it (and
with love, for example, good reasons not to around her bruiser boyfriend).


----- Original Message ----
From: Remo Dentato <rden...@gmail.com>

To: loj...@googlegroups.com
Sent: Sat, November 27, 2010 3:40:38 PM
Subject: Re: [lojban] Time for the perenial other-centric-.ui conversation

.remod.

--

Luke Bergen

unread,
Nov 27, 2010, 5:06:15 PM11/27/10
to loj...@googlegroups.com

Looks like it was entered into jbovlaste by daniel.  I'm assuming that's dbrockman

John E Clifford

unread,
Nov 27, 2010, 5:11:27 PM11/27/10
to loj...@googlegroups.com
Then what is getting expressed by the 'ue'?. It isn't the speakers emotion, so
he can't express it, and the person who has the emotion can't express it, since
he is not the speaker. I can imagine making a kind of sentence that was built
on what other people would express if they were the speaker and even include the
UI in it somewhere (in quotes, preferably, but we have gotten a ways away from
accuracy of this sort). Indeed, we have such sentences, even in English "He
would say 'ue'" for example.

----- Original Message ----
From: Remo Dentato <rden...@gmail.com>
To: loj...@googlegroups.com
Sent: Sat, November 27, 2010 3:42:50 PM
Subject: Re: [lojban] Time for the perenial other-centric-.ui conversation

--

John E Clifford

unread,
Nov 27, 2010, 5:14:50 PM11/27/10
to loj...@googlegroups.com
Same old question: do you empathize with his surprise.  If yes, then you only need 'da'; if no,then you need a whole sentence, and the 'a'oi' snippet is not a recognized sentence, I think.

Sent: Sat, November 27, 2010 3:49:12 PM

Subject: Re: [lojban] Time for the perenial other-centric-.ui conversation

Well obviously this is what I'd be yelling WHILE I was running into the room "bat a blazin' "  ;)

And the bit I was talking about earlier is that vlasisku says that it takes the preceding attitudinal (note that it is singular attitudinal).  So I would read "mi viska do .ue da'oi la .remo. .ui" as "I see you" with you being surprised and me being happy

On Nov 27, 2010 4:44 PM, "Remo Dentato" <rden...@gmail.com> wrote:

John E Clifford

unread,
Nov 27, 2010, 5:17:17 PM11/27/10
to loj...@googlegroups.com
You might also ask him what the Hell he thought he was doing.  Maybe he has an explanation that is not flat out contradictory..

Sent: Sat, November 27, 2010 4:06:15 PM

Subject: Re: [lojban] Time for the perenial other-centric-.ui conversation

Oren

unread,
Nov 27, 2010, 6:39:17 PM11/27/10
to loj...@googlegroups.com
I see the immediate utility and expressive freedom granted by { da'oi
} and no rock-solid argument that non-experimental cmavo can do the
same.

John E Clifford

unread,
Nov 27, 2010, 7:00:09 PM11/27/10
to loj...@googlegroups.com
It's not doing a cmavo's job, but a bridi's. As I say, if you want to work up a
sentence type that does this, go ahead. But, in keeping with the logical
language idea, do make it look like a sentence and not a UI or a cmavo smashup.
For clarity, the UI should appear either in quotes or not at all.

----- Original Message ----
From: Oren <get....@gmail.com>
To: loj...@googlegroups.com
Sent: Sat, November 27, 2010 5:39:17 PM
Subject: Re: [lojban] Time for the perenial other-centric-.ui conversation

--

Luke Bergen

unread,
Nov 27, 2010, 7:02:01 PM11/27/10
to loj...@googlegroups.com

Now that I've remembered the absurdly useful "kau", my understanding of da'oi is that it is basically a shortcut for ... peikau doi...  e.g.  "do flira farlu .oiro'o da'oi la .nik."  =  "do flira farlu .oiro'opeikau doi la .nik."

On Nov 27, 2010 6:39 PM, "Oren" <get....@gmail.com> wrote:

Luke Bergen

unread,
Nov 27, 2010, 7:05:49 PM11/27/10
to loj...@googlegroups.com

Good luck using lojban outside of a computer program, john.

"I see that you have injured yourself.  I imagine that you must feel extreme pain.  You have my sympothy selrirni"

On Nov 27, 2010 7:00 PM, "John E Clifford" <kali9...@yahoo.com> wrote:

John E Clifford

unread,
Nov 27, 2010, 7:29:57 PM11/27/10
to loj...@googlegroups.com
Well, perhaps good luck to you using it inside a program, since violation of type tend to lead to chaos (if you're very lucky).  If I only wanted to describe my emotions, what you write would be quite alright.  But, chances are, I have some emotions in this situation that I want to express and then  -- in English or Lojban -- I would probably put in some words that express them.  Your last sentence is, in fact, one such expression (it can also be used as a report and this is what makes English so messy and is the source of the muck-up in Lojban, like Loglan before it).



From: Luke Bergen <lukea...@gmail.com>
To: loj...@googlegroups.com
Sent: Sat, November 27, 2010 6:05:49 PM

Subject: Re: [lojban] Time for the perenial other-centric-.ui conversation

Good luck using lojban outside of a computer program, john.

"I see that you have injured yourself.  I imagine that you must feel extreme pain.  You have my sympothy selrirni"

On Nov 27, 2010 7:00 PM, "John E Clifford" <kali9...@yahoo.com> wrote:

Pierre Abbat

unread,
Nov 27, 2010, 8:13:16 PM11/27/10
to loj...@googlegroups.com
On Saturday 27 November 2010 16:17:04 Remo Dentato wrote:
> I've not used {da'oi} yet but being of selma'o DOI I expect I'll use it
> like:
>
> {mi viska da'oi .liuk. .ue }
>
> to indicate that you were surprise that I had seen you, and
>
> {mi viska doi .liuk. .ue }
>
> to indicate that I was surprised to see you.

If "da'oi" is in selma'o DOI, does that make cmevla like "mada'oitik" invalid?

Pierre

--
Jews use a lunisolar calendar; Muslims use a solely lunar calendar.

Luke Bergen

unread,
Nov 27, 2010, 8:17:12 PM11/27/10
to loj...@googlegroups.com

I would think so.  I never though about that.  I guess that'd mean that maki'etik would be problamatic as well.  Never thought of that.  COI is a fairly large selma'o too.

Jorge Llambías

unread,
Nov 27, 2010, 8:20:11 PM11/27/10
to loj...@googlegroups.com
On Sat, Nov 27, 2010 at 9:02 PM, Luke Bergen <lukea...@gmail.com> wrote:
> Now that I've remembered the absurdly useful "kau", my understanding of
> da'oi is that it is basically a shortcut for ... peikau doi...  e.g.  "do
> flira farlu .oiro'o da'oi la .nik."  =  "do flira farlu .oiro'opeikau doi la
> .nik."

Not quite. "da'oi" can be used to express empathy with a third party,
while "doi" only idenifies your interlocutor. "da'oi" is an expanded
"dai", such that "dai"="da'oi zo'e".

Also, you may be confusing "kau" with "paunai". "kau" outside of a
subordinate clause will tell you "whatever the answer to this question
is", it is not "this is a rhetorical question". So "oipeikau" is more
like "whether you like it or not, it doesn't really matter what your
answer is".

mu'o mi'e xorxes

Jorge Llambías

unread,
Nov 27, 2010, 8:24:11 PM11/27/10
to loj...@googlegroups.com
On Sat, Nov 27, 2010 at 10:13 PM, Pierre Abbat <ph...@phma.optus.nu> wrote:
>
> If "da'oi" is in selma'o DOI, does that make cmevla like "mada'oitik" invalid?

It should be in COI. In any case, with dot-side it's really a moot point.

Luke Bergen

unread,
Nov 27, 2010, 8:29:24 PM11/27/10
to loj...@googlegroups.com

Yeah, my understanding of kau has never been very solid though it seems really handy.  It's a shame that I don't /really/ get it. 

Yeah, I guess paunai makes more sense.  Though I though that it marked that a question was to follow (or not with nai).  Is it ok to say paunai to say that the /preceding/ question isn't really a question at all?

Jorge Llambías

unread,
Nov 27, 2010, 8:43:04 PM11/27/10
to loj...@googlegroups.com
On Sat, Nov 27, 2010 at 10:29 PM, Luke Bergen <lukea...@gmail.com> wrote:
>   Is it ok to say paunai to say
> that the /preceding/ question isn't really a question at all?

Yes, you can use "paunai" at the beginning of the bridi, to announce
that the question that follows is just rhetorical, so not really
something you expect an answer to, or you can use it directly after
the question word itself.

John E. Clifford

unread,
Nov 27, 2010, 10:56:47 PM11/27/10
to loj...@googlegroups.com
Thanks! That clarifies matters nicely. So 'da'oi' just marks the speaker's empathetic sharing of the emotion of someone other than the you in the situation. So no problem with that. There is a problem, however, if the move is made from this expression of my second-hand emotion to either a claim that the third party is feeling that emotion or an expression of that third party's emotion. It is not clear which of these -- or something else -- the various participants in this discussion are proposing but what they say seems to be one or the other or the two mixed in some not very useful way. In any case, I take it that, in fact, both are quite correctly not supported in the actual system.

Sent from my iPad

On Nov 27, 2010, at 19:20, Jorge Llambías <jjlla...@gmail.com> wrote:

On Sat, Nov 27, 2010 at 9:02 PM, Luke Bergen <lukea...@gmail.com> wrote:
Now that I've remembered the absurdly useful "kau", my understanding of
da'oi is that it is basically a shortcut for ... peikau doi... e.g. "do
flira farlu .oiro'o da'oi la .nik." = "do flira farlu .oiro'opeikau doi la.nik."

Not quite. "da'oi" can be used to express empathy with a third party,
while "doi" only idenifies your interlocutor. "da'oi" is an expanded
"dai", such that "dai"="da'oi zo'e".

Also, you may be confusing "kau" with "paunai". "kau" outside of a
subordinate clause will tell you "whatever the answer to this question
is", it is not "this is a rhetorical question". So "oipeikau" is more
like "whether you like it or not, it doesn't really matter what your
answer is".

mu'o mi'e xorxes

--

Luke Bergen

unread,
Nov 27, 2010, 11:58:07 PM11/27/10
to loj...@googlegroups.com

So long as empathy doesn't require that I feel the actual emotion myself, I'm fine with that.  I don't want to say .oidai and accidentally imply that I .oi

Pierre Abbat

unread,
Nov 28, 2010, 12:13:11 AM11/28/10
to loj...@googlegroups.com
On Saturday 27 November 2010 20:17:12 Luke Bergen wrote:
> I would think so. I never though about that. I guess that'd mean that
> maki'etik would be problamatic as well. Never thought of that. COI is a
> fairly large selma'o too.

The rule about substrings of cmevla applies only to selma'o LA and DOI, not
COI. One should therefore write me'o denpa bu between COI and a cmevla. I
often forget though.

mu'omi'e .pier.

Jorge Llambías

unread,
Nov 28, 2010, 7:25:43 AM11/28/10
to loj...@googlegroups.com
On Sun, Nov 28, 2010 at 12:56 AM, John E. Clifford <kali9...@yahoo.com> wrote:
> Thanks!  That clarifies matters nicely. So 'da'oi' just marks the speaker's empathetic sharing of the emotion of someone other than the you in the situation.

It could also be with the you (and "dai" could also be with someone
other than the you.) "da'oi" just allows you to be explicit about who
it is with.

> So no problem with that.  There is a problem, however, if the move is made from this expression of my second-hand emotion to either a claim that the third party is feeling that emotion or an expression of that third party's emotion.

Right, "da'oi" is not used to make any claims. But if I hear you
expressing empathy with X on attitude Y, I can legitimetely conclude
that you are attributing attitude Y to X. It is of course impossible
to conclude from that that X actually does have that attitude. There
are a million reasons why you may be attributing that attitude to
them. I cannot even conclude that you really think they have that
attitude, since in some contexts it is perfectly sensible to attribute
attitudes that you know they don't have (tongue-in-cheek, humor,
deceit, etc.)

>It is not clear which of these -- or something else -- the various participants in this discussion are proposing but what they say seems to be one or the other or the two mixed in some not very useful way.  In any case, I take it that, in fact, both are quite correctly not supported in the actual system.

I really don't see what all the brouhaha about this is.

In the case of "peekaboo!", I would just say "ua". Not because I'm
actually discovering anything, but because I'm expressing "discovery"
for the benefit of the baby.

In the case of the magician's "surprise!", "uedai" is perfectly fine,
not because the magician is actually surprised, but because acting
surprised is part of their act. Of course it depends on the magician,
some may prefer acting cool and detached, others pretend to be as
surprised as their audience, or even more surprised, in which case a
first hand "ue" might be even more appropriate. Magicians are
performers, and what they say is part of their performance.

Jorge Llambías

unread,
Nov 28, 2010, 7:33:10 AM11/28/10
to loj...@googlegroups.com
On Sun, Nov 28, 2010 at 1:58 AM, Luke Bergen <lukea...@gmail.com> wrote:
> So long as empathy doesn't require that I feel the actual emotion myself,
> I'm fine with that.  I don't want to say .oidai and accidentally imply that
> I .oi

When you use ".oi" as a verb like that is when you send pc into a fit.

Neither saying ".oi" nor saying ".oidai" require that you feel
anything. People seem terribly confused sometimes about the difference
between feeling X, expressing X and claiming that they feel X. They
are three different things.

John E Clifford

unread,
Nov 28, 2010, 7:45:11 AM11/28/10
to loj...@googlegroups.com
Well, empathy does require exactly that you feel "the same" emotion.  Incidentally, your last remark, using 'oi' as a verb, is exactly the problem that that I am worried about and your use suggest that my worry is not unfounded.

Sent: Sat, November 27, 2010 10:58:07 PM

Subject: Re: [lojban] Time for the perenial other-centric-.ui conversation

John E Clifford

unread,
Nov 28, 2010, 7:49:56 AM11/28/10
to loj...@googlegroups.com
Exactly, and the latter two are linguistic but of very different linguistic
types.

----- Original Message ----
From: Jorge Llambías <jjlla...@gmail.com>
To: loj...@googlegroups.com
Sent: Sun, November 28, 2010 6:33:10 AM
Subject: Re: [lojban] Time for the perenial other-centric-.ui conversation

--

Jorge Llambías

unread,
Nov 28, 2010, 8:34:59 AM11/28/10
to loj...@googlegroups.com
On Sun, Nov 28, 2010 at 9:49 AM, John E Clifford <kali9...@yahoo.com> wrote:
> Exactly, and the latter two are linguistic but of very different linguistic
> types.

I would say that expressing may or may not be linguistic, and that it
is not a linguistic type that contrasts with claiming.

Suppose that John is surprised. That is obviously not a linguistic
matter, it's just something that happens. But John may express his
surprise in many different ways:

(1) By opening his mouth very widely.
(2) By saying "Holy shit!"
(3) By saying "I am truly surprised."
(4) By doing all of the above.

All of those are perfectly good ways to express surprise. (1) is
non-linguistic, (2) and (3) are linguistic. (3) also makes a claim,
but all can be used to express surprise. (2) is linguistic but does
not make any claim.

Saying "ue" is of course like (2). It is a linguistic way of
expressing surprise, not by making a claim but by making an
exclamation.

Also, not all UIs are used to make exclamations, which is one of the
unwarranted overgeneralizations that are sometimes made. UIs have
several other uses.

John E Clifford

unread,
Nov 28, 2010, 9:43:00 AM11/28/10
to loj...@googlegroups.com
Right as usual; I apologize for generalizing beyond the data. My only
disagreement would be over whether "I am truly surprised" must make a claim any
more than "I hope he comes" does.As for UI, the interesting question is whether
all are expressions of some emotion (in a veeerry broad sense). Some of them
typically are combined with sentences and affect the status of that sentence,
but the whole might still reasonably be called an expression. There are a few
that are harder to place.

----- Original Message ----
From: Jorge Llambías <jjlla...@gmail.com>
To: loj...@googlegroups.com
Sent: Sun, November 28, 2010 7:34:59 AM
Subject: Re: [lojban] Time for the perenial other-centric-.ui conversation

--

Craig Daniel

unread,
Nov 28, 2010, 10:01:38 AM11/28/10
to loj...@googlegroups.com
On Sat, Nov 27, 2010 at 11:58 PM, Luke Bergen <lukea...@gmail.com> wrote:
> So long as empathy doesn't require that I feel the actual emotion myself,
> I'm fine with that.  I don't want to say .oidai and accidentally imply that
> I .oi

I always understood it as expressing empathy with the perceived oi,
which can't possibly mean you feel oinai. There is absolutely a
difference between recognizing pain in somebody else and empathizing
with it!

I'm sorta with JEC on this one, in that UI should be expressing your
emotion, but if da'oi is really just about expressing your empathy
with a specified person then it makes total sense to me. Some
da'oi-advocates seem to indicate that this is what it is - something
semantically equivalent to a way to specify the referent of dai
(although syntactically quite distinct); that seems useful. (Although
if it's in COI, doesn't it have the side effect of resetting the
referent of "do"?) Some seem to want it to mean "I believe so-and-so
feels the emotion indicated by saying whatever attitudinal (or,
apparently from some example sentences, string of attitudinals -
something dai cannot modify, because I can uedai after oiing or after
oidaiing*) and am not saying anything at all about my own emotional
state." In this case, you are stating apparent facts about the world,
not expressing your own feelings; statements of fact or belief like
that are what bridi are *for.* I'm against any experimental cmavo
whose advocates can't agree on what it means, because that kind of
imprecision is incompatible with what the non-experimental parts of
the language strive to be (although they have sometimes been every bit
as murky in their own way), so you can put me in the anti-da'oi bin
until you guys make up your mind.

The notion that saying "no, da'oi shouldn't work like that even though
nothing else does" is telling you that there's no good way to say
"ooh, that must have hurt" in Lojban is just silly, because nobody but
you seems resistant to using the vast majority of the grammar in the
way it was intended - the "ooh" is an English UIesque interjection
about the *speaker's* emotion, and the rest of the sentence is a
declarative sentence and really ought to be translated as one. The
emotional gismu were created for a reason.

That said (tangent warning!), I think there's quite a difference
between zo'o and u'idai. The "surprise!" of an unexpected party is
much more akin to the former, and is not empathizing with anything at
all. It is not a perceived emotion, but an intended one. If it is to
be expressed with a UI at all, and I'm not sure it needs to be, it's
definitely not one modified with dai (or da'oi, if that's a
specified-referent dai relative).

Now, I can see the value of a possible experimental dai-alike for
intended emotions, such that u'iblah and zo'o are synonymous, and
ueblah conveys something like "this is said/done with the intent that
it will be surprising!" But such a hypothetical cmavo is not and
should not be confused with dai. If da'oi is a semantically dai-like
cmavo, then this hypothetical would probably quickly get a
corresponding experimental COI. And I'm not sure the dai-for-intent
cmavo is even remotely necessary - one could just as easily say "spaji
.ai" in the three syllables needed for any experimental cmavo not
starting with x, and use the observative "spaji" instead of "spaji
da'oi."

- mi'e .kreig.

* John: by "oiing" in this context I mean "expressing pain through
the use of zo oi" rather than "feeling pain"; it's an English
shorthand for "cusku lu .oi li'u" rather than for "cortu."

John E Clifford

unread,
Nov 28, 2010, 10:25:17 AM11/28/10
to loj...@googlegroups.com
Yup! Except, again, that the English sentence "That must hurt" may well be part
of the expression as well (the "must" is a clue that this is not merely
descriptive). I think that something like 'da'oi' is useful (for expressing
empathetic emotions, etc. with third parties) but can't see why 'dai' (possibly
with a change of selma'o) couldn't fill the bill, defaulting to 'do'.

----- Original Message ----
From: Craig Daniel <craigb...@gmail.com>
To: loj...@googlegroups.com
Sent: Sun, November 28, 2010 9:01:38 AM
Subject: Re: [lojban] Time for the perenial other-centric-.ui conversation

- mi'e .kreig.

--

Luke Bergen

unread,
Nov 28, 2010, 10:34:39 AM11/28/10
to loj...@googlegroups.com

Ok, I see where you're going.  So "oooo, that looked like it hurt" might become something like ".uu ta simlu lo ka cortu".  I suppose.  It's just unfortunate that there's this rich exclamation system that I can only use to indicate my own emotional state.  But I guess it makes sense and I should stop trying to shoehorn .ui and friends into shortcuts for bridi that involve do.... or just say .uipeipaunai =p

John E Clifford

unread,
Nov 28, 2010, 10:36:48 AM11/28/10
to loj...@googlegroups.com
I do worry about the notion that 'zo'o' is predictive. I suspect that it is
more often a warning: I don't expect that you will be amused but I do mean this
in a humorous way. Nor do I think that "Surprise" is predictive; it is closer
to a performative: the utterance constitutes (or is a significant part of) the
surprise, which is intended to engender a feeling of surprise in the recipient.

----- Original Message ----
From: Craig Daniel <craigb...@gmail.com>
To: loj...@googlegroups.com
Sent: Sun, November 28, 2010 9:01:38 AM
Subject: Re: [lojban] Time for the perenial other-centric-.ui conversation

- mi'e .kreig.

--

John E Clifford

unread,
Nov 28, 2010, 10:41:14 AM11/28/10
to loj...@googlegroups.com
Still not clear what the point of 'uipeipaunai' is in all this.


From: Luke Bergen <lukea...@gmail.com>
To: loj...@googlegroups.com
Sent: Sun, November 28, 2010 9:34:39 AM

Subject: Re: [lojban] Time for the perenial other-centric-.ui conversation

Pierre Abbat

unread,
Nov 28, 2010, 10:56:15 AM11/28/10
to loj...@googlegroups.com
On Sunday 28 November 2010 10:25:17 John E Clifford wrote:
> Yup! Except, again, that the English sentence "That must hurt" may well be
> part of the expression as well (the "must" is a clue that this is not
> merely descriptive). I think that something like 'da'oi' is useful (for
> expressing empathetic emotions, etc. with third parties) but can't see why
> 'dai' (possibly with a change of selma'o) couldn't fill the bill,
> defaulting to 'do'.

What does "dai" after "pei" default to?

Pierre
--
La sal en el mar es más que en la sangre.
Le sel dans la mer est plus que dans le sang.

Luke Bergen

unread,
Nov 28, 2010, 10:58:56 AM11/28/10
to loj...@googlegroups.com

It means exactly what it looks like.  Pei asks the receiver how much or if at all they are feeling the .ui and then the paunai says "but that wasn't a question".  In other words, I would read it as an exclamation of "I know to what extent or whether or not you are feeling .ui".  In other words, a cheap way of expressing .ui for them, or rather expressing that I know the extent to which they could accuratly express .ui (be it cai, cu'i or nai)

John E Clifford

unread,
Nov 28, 2010, 1:34:44 PM11/28/10
to loj...@googlegroups.com
Ah! So 'uipei' is a version of 'uicai' to 'uinai', not a separate speech act.  I was reading your phrase as three, not two.  But whose 'ui' is it?  The fact that it is an expression seems to mean it is the speaker's.  The fact that the speakers asks about its intensity seems to require that it is not , even if the question is rhetorical.  If it were an expression of the listener's emotions (which it cannot be, by definition), then the complete complex speech act of asking a rhetorical question would indeed imply something what your sentence says (reading it generously) or perhaps that I don't really
care about your feelings at all.  But I see no reason to think it actually works that way nor that, even if it did, it would imply your putative sentence.  The problem with the sentence is just the root problem of this whole issue, the use of 'ui' as a noun (in this case; it was a verb elsewhere).  It is an exclamation, so "feeling ui" is like "feeling huzzah".  If it makes any sense at all, it means something like "feels like wanting to say 'ui'" or so.  But it is a totally opaque way of saying that and Logjam (the logical language, before all) is meant to be transparent (at least to the extent of having its opacities clarifiable by following a few, mainly logical, rules).  So, this expression is not and should not be a part of Lojban.

Sent: Sun, November 28, 2010 9:58:56 AM

Jonathan Jones

unread,
Nov 28, 2010, 1:40:13 PM11/28/10
to loj...@googlegroups.com
It is my understanding that {.uipei} is "How happy are you?", in other words, it is asking the listener where on the .ui scale the listener is.
mu'o mi'e .aionys.

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

John E Clifford

unread,
Nov 28, 2010, 1:52:14 PM11/28/10
to loj...@googlegroups.com
But it "says" (not really) "How much happiness do I express?"  You are saying that this is an idiom, an expression totally detached from its base meanin in an illogical way and tucked into the grammar (and semantics and pragmatics) of the logical language.  Nice to have a short expression for that, I suppose, since we have others for other conventional greetings (and in that case, the 'paunai' is definitely called for).  But wouldn't just 'pei' do as well -- or better -- and be shorter, too?


From: Jonathan Jones <eye...@gmail.com>
To: loj...@googlegroups.com
Sent: Sun, November 28, 2010 12:40:13 PM

Jonathan Jones

unread,
Nov 28, 2010, 2:26:22 PM11/28/10
to loj...@googlegroups.com
No it doesn't.

"pei - CAI - emotion ? - attitudinal: attitudinal question; how do you feel about it? with what intensity?"

It "says" "How much happiness do YOU express?"

{pei} by itself would be asking about emotion in general, and can be answered by, for examples, {.iinai}, {.aucu'i}, {.o'icai}, etc., whereas {.uipei} is asking about happiness specifically, anything on the .ui scale is a sensical answer, but only that which lies on the .ui scale.

Jorge Llambías

unread,
Nov 28, 2010, 4:08:06 PM11/28/10
to loj...@googlegroups.com
On Sun, Nov 28, 2010 at 12:01 PM, Craig Daniel <craigb...@gmail.com> wrote:
> (Although
> if it's in COI, doesn't it have the side effect of resetting the
> referent of "do"?)

"mi'e" is in COI and does not have that effect. The reason to put
"da'oi" in COI/DOI is because of its syntactic behaviour, not because
of it's meaning. It clearly does not create a vocative like almost all
the other COIs do.

> That said (tangent warning!), I think there's quite a difference
> between zo'o and u'idai. The "surprise!" of an unexpected party is
> much more akin to the former, and is not empathizing with anything at
> all. It is not a perceived emotion, but an intended one. If it is to
> be expressed with a UI at all, and I'm not sure it needs to be, it's
> definitely not one modified with dai (or da'oi, if that's a
> specified-referent dai relative).

That's true. But human beings are imitative creatures, so a common way
of inducing (or trying to induce) an emotion in someone is by
expressing that emotion yourself. So while "u'i" and "zo'o" do have
different definitions, their use is not that far appart, because you
can't very credibly say that something is meant as humor but you are
not amused, or express amusement and deny that you mean it to be
humorous.

Another similar case is (the way I use) ".o'i", which is not so much
to express a feeling of caution as to induce that feeling in someone
else, by the same mechanism of contagion.

> Now, I can see the value of a possible experimental dai-alike for
> intended emotions, such that u'iblah and zo'o are synonymous, and
> ueblah conveys something like "this is said/done with the intent that
> it will be surprising!" But such a hypothetical cmavo is not and
> should not be confused with dai.

I don't see a need, because the distinction between ".u'i" and "zo'o",
while understandable, has always seemed somewhat artificial. What does
it mean when someone adds a smiley to something they write? That they
find it amusing or that they want others to find it amusing? What does
it mean when you say something with a smile? Is it really worth making
such a subtle distinction?

> If da'oi is a semantically dai-like
> cmavo, then this hypothetical would probably quickly get a
> corresponding experimental COI. And I'm not sure the dai-for-intent
> cmavo is even remotely necessary - one could just as easily say "spaji
> .ai" in the three syllables needed for any experimental cmavo not
> starting with x, and use the observative "spaji" instead of "spaji
> da'oi."

I will leave that to TV show scriptwriters (those surprise parties
where the lights are suddenly turned on and everyone says "surprise!"
only happen in TV shows, don't they?) For more natural scenarios, I
think empathetic surprise works well.

John E Clifford

unread,
Nov 28, 2010, 4:10:52 PM11/28/10
to loj...@googlegroups.com
So, the question is how a word that expresses the speaker's emotions comes to have a propositional function: How are you feeling on the happiness scale?.I assume one of the reasons for using this idiom is that this question is hard to formulate as a full sentence.  The word for "happy" probably doesn't have a place for quantitative measure, nor are such measures easy to formulate and turning that all into a question raises extra problems.-- all soluble, of course, but taking a moment's thought. So, here is a cute dodge; never mind it literally makes no sense (indeed, a mark of a good idiom).  The other UICAI are expressions of MY emotion; this is now suddenly of YOURS which, of course makes no sense.  The problems seems to lie ultimately with 'pei' itself: it calls upon a person to express his emotion, but not as an expression of an emotion, rather as an factual answer to a factual question -- which just isn't an expression's job (again, an expression can be sincere or not, but it can't be true or false). Of course, once we start looking at cmavo for sensibility, the task could be a very long one, so I'll leave 'pei' be as totally useless in the real Logjam, but perhaps fitting into some illogical argot on the periphery.



From: Jonathan Jones <eye...@gmail.com>
To: loj...@googlegroups.com
Sent: Sun, November 28, 2010 1:26:22 PM

John E Clifford

unread,
Nov 28, 2010, 4:26:27 PM11/28/10
to loj...@googlegroups.com
Impeccable as your Lojban is, your descriptions of human behavior leave
something to be desired, mainly accuracy. Saying something is meant as humor
but I am not amused is one of the most common moves in watching and critiquing
tv shows, for one example, and being ammused by something that was not meant to
be funny is a painful memory from both sides for most people. The encouraging
imitation use of expressions is an interesting theory, but doesn't seem to fit
the facts very well, at least for the "surprise" case, where the person will be
surprised (or not) even if nothing is said (the emotion is inherent in the
situation). The 'o'i' case is better, but then, the emotion of caution seems
an odd concept from the get-go.
In the US at least surprise parties of the dark room-lights up-shout "surprise"
sort are common enough to be a recognized listing on incident reports: murders,
beating, heart attacks, etc.


----- Original Message ----
From: Jorge Llambías <jjlla...@gmail.com>
To: loj...@googlegroups.com
Sent: Sun, November 28, 2010 3:08:06 PM
Subject: Re: [lojban] Time for the perenial other-centric-.ui conversation

--

Jorge Llambías

unread,
Nov 28, 2010, 4:31:24 PM11/28/10
to loj...@googlegroups.com
On Sun, Nov 28, 2010 at 11:43 AM, John E Clifford <kali9...@yahoo.com> wrote:
> Right as usual; I apologize for generalizing beyond the data.  My only
> disagreement would be over whether "I am truly surprised" must make a claim any
> more than "I hope he comes" does.

They both make claims in English, as well as expressing how the
speaker feels. In those cases the speaker expresses how they feel by
making a claim.

In English you don't really have much of an option for expressing hope
other than making a claim, unless you go with something ridiculous for
modern English like "would that he come!". In Spanish you have both
options, like in Lojban:

¡Ojalá que venga!
.a'o (ko'a) klama

¡Espero que venga!
mi pacna lo nu (ko'a) klama

(In fact the Spanish exclamation "ojalá" comes from Arabic meaning
something like "God willing", but in Spanish it is just an
interjection and speakers are not normally aware of the etymology.)

> As for UI, the interesting question is whether
> all are expressions of some emotion (in a veeerry broad sense).

They are not, unless you distort the meaning of "emotion" so much that
it just means "anything expressed by selma'o UI". ".e'a" for example
is used to grant permission, and has nothing to do with any emotion in
the usual sense of "emotion".

> Some of them
> typically are combined with sentences and affect the status of that sentence,
> but the whole might still reasonably be called an expression.  There are a few
> that are harder to place.

Most utterances can reasonably be called an expression. Normally when
you say something, you mean to express something. Not all expressions
are made through claims, but that doesn't mean that claims can't be
made in order to express something.

Jorge Llambías

unread,
Nov 28, 2010, 4:52:49 PM11/28/10
to loj...@googlegroups.com
On Sun, Nov 28, 2010 at 6:10 PM, John E Clifford <kali9...@yahoo.com> wrote:
> So, the question is how a word that expresses the speaker's emotions comes
> to have a propositional function: How are you feeling on the happiness
> scale?

No, "ui pei" just asks the interlocutor to express themself using the
"ui ___" format. The paraphrase "How are you feeling on the happiness
sacle?" is just that, a paraphrase.

> So, here
> is a cute dodge; never mind it literally makes no sense (indeed, a mark of a
> good idiom).

It makes as much sense as any other question.

> The other UICAI are expressions of MY emotion; this is now
> suddenly of YOURS which, of course makes no sense.

When I make I claim, say "la .djan. klama lo zarci", the claim is MY
claim. When I ask a question, the bridi is sudenly no longer MY claim,
but a pattern for you to make a claim. The relationship between "ui
pei" and "ui sai" is no different from that between "la .djan. klama
ma" and "la .djan. klama lo zarci" in that respect.

> The problems seems to
> lie ultimately with 'pei' itself: it calls upon a person to express his
> emotion, but not as an expression of an emotion, rather as an factual answer
> to a factual question

No, "ui pei" asks the person to respond "ui ja'ai" or "ui nai" (or any
of the other possibilities).

John E Clifford

unread,
Nov 28, 2010, 5:02:24 PM11/28/10
to loj...@googlegroups.com
Well, I was raised on speech-act theory and so I would maintain that "I hope he
comes" has two different uses, can be used in two different speech acts,
corresponding pretty much to the two in Spanish and Lojban. It is this
possibility that leads to so many problems with UI and some other classes. The
difference of course is just whether the sentence can be true or false or not:
the expressive ('a'o klama') cannot, since it makes no claim and is not even a
reliable guide to whether I hope this or not. The other ('mi pacna lo nu
klama') is false if I don't really have that hope (both are, of course, totally
independent of whether he comes or not). People jump from the fact that the
'a'o' form is in English "I hope' to using 'a'o' also for the 'pacna' form and
then generalizing to all UI and beyond. I agree that "emotion" is a bad word
for what UI does, but there isn't any good word, since it covers a wide range of
speech acts: performative, expressive of real emotions, directive, and so on
through most of the list, possibly excluding declarative. Maybe "expression" is
a bad word, too, since it is after all a common word for any linguistic
productionm but again. there are not a lot of good choices here if we want to
cover the range (doing which may be the original bad idea - trying to cover a
syntactic class with a pragmatic classification).


----- Original Message ----
From: Jorge Llambías <jjlla...@gmail.com>
To: loj...@googlegroups.com

--

Jorge Llambías

unread,
Nov 28, 2010, 5:05:00 PM11/28/10
to loj...@googlegroups.com
On Sun, Nov 28, 2010 at 6:26 PM, John E Clifford <kali9...@yahoo.com> wrote:
> Saying something is meant as humor
> but I am not amused is one of the most common moves in watching and critiquing
> tv shows, for one example,

Saying *that something I'm saying* is meant as humor, not that
something someone else has said was meant as humor. "zo'o" marks the
speaker's speech, not someone else's speech.

> and being ammused by something that was not meant to
> be funny is a painful memory from both sides for most people.

We are talking about being amused *by something you are saying*, not
by something someone else has said. "zo'o" marks your own speech, not
someone else's.

> In the US at least surprise parties of the dark room-lights up-shout "surprise"
> sort are common enough to be a recognized listing on incident reports: murders,
> beating, heart attacks, etc.

(Ils sont fous ces américains.) I can see a possible connection with
heart attacks, but I'm not sure I see how it relates to murders and
beatings. Is the person being surprised the murderer or the murdered
one, and why?

John E Clifford

unread,
Nov 28, 2010, 5:34:21 PM11/28/10
to loj...@googlegroups.com
But there is something odd about asking someone to express on a scale.
Presumably, he has already expressed his degree of happiness in some way, since
that is what we are asking about -- refine your expression. But under these
conditions what we are asking for in not actually an expression but information,

we want to know a fact. And that just isn't what UI (and UICAI) are about. If
I stop and consider whether to say 'uicai' or 'uicu'i' or 'uinai' or decide to
stick with just 'ui' I am no longer expressing my happiness in any natural sense

of the word, but rather describing it. If I burst in the room and say "uicai, I

passed", then I am probably expressing my extreme happiness. If, on the other
hand, I stop to analyze my feelings and then say 'uicai' I am more likely
seeking to give information -- especially if I do it in answer to a question.
And UICAI is not about giving information.


----- Original Message ----
From: Jorge Llambías <jjlla...@gmail.com>

To: loj...@googlegroups.com
Sent: Sun, November 28, 2010 3:52:49 PM
Subject: Re: [lojban] Time for the perenial other-centric-.ui conversation

--

Robin Lee Powell

unread,
Nov 28, 2010, 5:48:10 PM11/28/10
to loj...@googlegroups.com
On Sat, Nov 27, 2010 at 01:45:47PM -0500, Luke Bergen wrote:
> That's actually not what I intend but it is what will inevitably
> happen.
>
> I was talking to my wife who is perpetually freezing and at one
> point I remembered somehing and wanted to say "oh yeah! Your dad
> said he was going to make a fire tonight _____". Where "______"
> was that cool new cmavo of COI that takes a UI before it and
> somehow applies it to the object of the COI. I've searched around
> but must be using the wrong search terms. Anybody remember what
> I'm talking about?
>
> What I'm trying to express is something like what you want to
> express when you play peekaboo with a baby. It's like some kind
> of .uadai like "surprise!" But it's not a command or an
> observation, but more of like an expectation. Almost like a
> .uipeikau if that makes any sense?

{ko se spaji} ; why does this need to be a UI?

-Robin

--
http://singinst.org/ : Our last, best hope for a fantastic future.
Lojban (http://www.lojban.org/): The language in which "this parrot
is dead" is "ti poi spitaki cu morsi", but "this sentence is false"
is "na nei". My personal page: http://www.digitalkingdom.org/rlp/

John E Clifford

unread,
Nov 28, 2010, 5:53:47 PM11/28/10
to loj...@googlegroups.com
Well, in that case, 'zo'o' is back in the "fair warning" category: "Don't take
this seriously" but it does not say that I think it is funny. And it is
certainly not a laugh, which I take 'u'i' to be. You keep insisting that 'zo'o'
marks my speech not someone else, contrary to at least a few cases to be found
in the annals, but then you have no problem with 'ui' being someone else
emotion. I suppose you are right about the proper use of 'zo'o' though your
original remarks were not perfectly clear (obviously). But that does not make
your point, merely restricts it to a more limited set of cases, satirists, say.
On the other side, I have cracked up giving a funeral eulogy and have seen
several others do similar things (indeed, one identical one -- with a different
urn).
Yeah we are cuckoo, not that any other country is in a position to judge. The
most common murder/mayhem scenario is that the surprisee reacts instantly to the
presents of a lot of people in a dark room and just blasts away (I admit that
particular response is almost uniquely American and almost only possible in
America). Most of the others involve someone thinking that the person coming in
is not the surprisee (this typically involves a long pre-party party). The
point is that this kind of surprise party is not uncommon (it is also a frequent
request at significant birthdays, say, "No surprise party, please!")

----- Original Message ----
From: Jorge Llambías <jjlla...@gmail.com>
To: loj...@googlegroups.com
Sent: Sun, November 28, 2010 4:05:00 PM
Subject: Re: [lojban] Time for the perenial other-centric-.ui conversation

--

John E Clifford

unread,
Nov 28, 2010, 5:55:03 PM11/28/10
to loj...@googlegroups.com
Or, indeed, how can it be?


----- Original Message ----
From: Robin Lee Powell <rlpo...@digitalkingdom.org>
To: loj...@googlegroups.com
Sent: Sun, November 28, 2010 4:48:10 PM
Subject: Re: [lojban] Time for the perenial other-centric-.ui conversation

-Robin

--

Jorge Llambías

unread,
Nov 28, 2010, 6:14:19 PM11/28/10
to loj...@googlegroups.com
On Sun, Nov 28, 2010 at 7:34 PM, John E Clifford <kali9...@yahoo.com> wrote:
> But there is something odd about asking someone to express on a scale.

Yes/no is the main gist of the question. How nuanced to make the
answer is up to the one responding. For example: "iepei", "do you
agree?". I don't really care for a nuanced answer, but you may feel
compelled to make it nuanced anyway.

> Presumably, he has already expressed his degree of happiness in some way, since
> that is what we are asking about -- refine your expression.

I don't think that would be the usual case.

> But under these
> conditions what we are asking for in not actually an expression but information,
> we want to know a fact.  And that just isn't what UI (and UICAI) are about.

We are asking about someone's attitude: "iepei", "do you agree?",
"e'apei", "may I?"(Do you give your permission?"), ".u'upei", "any
regrets?", "je'epei", "OK?" and so on.

> If
> I stop and consider whether to say 'uicai' or 'uicu'i' or 'uinai' or decide to
> stick with just 'ui' I am no longer expressing my happiness in any natural sense
> of the word, but rather describing it.

The idea that UIs must somehow come directly from the gut and not pass
through the brain is one of those Lojbanic myths that have no reason
of being.

> If I burst in the room and say "uicai, I
> passed", then I am probably expressing my extreme happiness.  If, on the other
> hand, I stop to analyze my feelings and then say 'uicai' I am more likely
> seeking to give information -- especially if I do it in answer to a question.
>  And UICAI is not about giving information.

"la'acai" means that I consider something extremely likely. It is
certainly informative. I don't see why UIs can't contain any
information.

Luke Bergen

unread,
Nov 28, 2010, 6:55:02 PM11/28/10
to loj...@googlegroups.com
What do you mean?  Of course UI contain information.  They contain information regarding my emotional state.  If I burst into a room and yell ".ui" then people will have been informed as to my emotional state.  ".uipei" asks the listener to inform [me|us] about their location in the ".ui" dimension.

Is it any more ridiculous to ask for a persons location on an x, y, or z plane than it is to ask where they are on an emotional scale?

2010/11/28 Jorge Llambías <jjlla...@gmail.com>

John E Clifford

unread,
Nov 28, 2010, 10:27:06 PM11/28/10
to loj...@googlegroups.com
The point keeps being that what you are doing is seeking information not an
expression of some sentiment, emotion, or what have you. The answer is, as you
say, yes or no, and those are the answers to factual questions, nothing to do
with 'ui' or even 'a'o', so why bring them into the question at all. The way to
ask whether you agree or not is 'xu do tugni' not 'iepei', which is something
like "You damned betcha , innit?" You are asking about sommone's attitude and
you want a factual answer; therefore, you are asking a factual quest, a bridi
with 'xu' attached -- or with a question word at some point in it. That's how
you perform that speech act in a logical language. This is not Neanderthal,
after all, where the conversation is entirely in grunts.
Expressing a whatever need not come from the gut and may go through the brai,
but it is still a different act from stating a fact or asking a factual question
(any kind of question as far as I can see). The reason for the myth is to drive
home tis fundamental point, which obviously needs some more driving.
'la'a cai' expresses your confidence in the following statement being true, less
than 'ju'o' more than 'la'a' alone, but it is a discursive, not a modal and is
not false if the event is unlikely, as the modal case would be. It is grounds
for thinking that you believe the event likely (though not definitive grounds);
it is not grounds for thinking the event is likely, nor does it claim to be (or
anything else for that matter).


----- Original Message ----
From: Jorge Llambías <jjlla...@gmail.com>
To: loj...@googlegroups.com
Sent: Sun, November 28, 2010 5:14:19 PM
Subject: Re: [lojban] Time for the perenial other-centric-.ui conversation

--

John E Clifford

unread,
Nov 28, 2010, 10:33:57 PM11/28/10
to loj...@googlegroups.com
Well, it depends on what you mean by information.  Someone bursting in saying 'ui' is evidence (maybe even decisive evidence) for his being in a certain emotional state.  But it is not a claim that he is in that state; it is neither true nor false.  But evidence is information, just not (in this case) stated information.  It does, of course, make sense to ask about someone emotional state, and the way to do that is ask "How is your emotional state?" or something like that.  There is, after all, no grunts and wheezes that ask where one is in three (or four or more) space, one asks "Where the fugawee?"  Why should the matter be different for emotional space?


From: Luke Bergen <lukea...@gmail.com>
To: loj...@googlegroups.com
Sent: Sun, November 28, 2010 5:55:02 PM

Subject: Re: [lojban] Time for the perenial other-centric-.ui conversation

Luke Bergen

unread,
Nov 28, 2010, 11:28:19 PM11/28/10
to loj...@googlegroups.com
well.. if for no other reason... because it's in the language definition and in active use (at least I seem to recall seeing pei used around IRC and such).  Maybe in loCCam3 you can take pei out.  I have a feeling like when (if) you eventually start to actually speak it though, you'll find that you miss good ol' pei.

Jorge Llambías

unread,
Nov 29, 2010, 8:04:46 AM11/29/10
to loj...@googlegroups.com
On Mon, Nov 29, 2010 at 12:27 AM, John E Clifford <kali9...@yahoo.com> wrote:
> The point keeps being that what you are doing is seeking information not an
> expression of some sentiment, emotion, or what have you.

You keep thinking of UI as things used to express a sentiment, emotion
or what have you, but UIs are not *just* that.

True observation: some UIs are most often used to express an emotion.

Invalid inference #1: Every UI is always used to express an emotion
and for nothing else.

Invalid inference #2: Only UIs can be used to express an emotion.

It is strange that in the Lojban community, this jumping from Some X
is Y to Every X is Y and Only X is Y is so frequent.

Invalid inference #3: Since UIs are expressive, they cannot be
communicative. If you want to communicate something, you need to use a
proposition.

> The answer is, as you say, yes or no, and those are the answers
> to factual questions, nothing to do with 'ui' or even 'a'o', so why
> bring them into the question at all.

The main answer to "pei" is "ja'ai" or "nai". For more nuanced
answers, there are other members of CAI. And you can throw in other
specifiers like the ro'V series for even more nuance. And others.

> The way to
> ask whether you agree or not is 'xu do tugni' not 'iepei', which is something
> like "You damned betcha , innit?"  You are asking about sommone's attitude and
> you want a factual answer; therefore, you are asking a factual quest, a bridi
> with 'xu' attached -- or with a question word at some point in it.

"iepei" is a perfectly good way to ask whether someone agrees with
something you are saying or not.

> That's how
> you perform that speech act in a logical language.  This is not Neanderthal,
> after all, where the conversation is entirely in grunts.

UIs are not grunts. They are words with meanings, like all other words
of the language. Some of them are most often used purely to express an
emotion. Please don't jump from there to "each one of them can only be
used to express an emotion".

> Expressing a whatever need not come from the gut and may go through the brai,
> but it is still a different act from stating a fact or asking a factual question
> (any kind of question as far as I can see).

Of course they are different speech acts. Indeed the function of some
UIs is precisely to specify the kind of speech act you are performing.
They are not all and always used for the same type of speech act.

> The reason for the myth is to drive
> home tis fundamental point, which obviously needs some more driving.

Unfortunately, that "fundamental point" is wrong, and driving it home
only creates more confusion in an area where we already have too much
of it.

> 'la'a cai' expresses your confidence in the following statement being true, less
> than 'ju'o' more than 'la'a' alone, but it is a discursive, not a modal and is
> not false if the event is unlikely, as the modal case would be.

Which has nothing to do with the fact that "la'apei" is a perfectly
good way of asking for someone's confidence on something being the
case.

> It is grounds
> for thinking that you believe the event likely (though not definitive grounds);
> it is not grounds for thinking the event is likely, nor does it claim to be (or
> anything else for that matter).

And "la'apei" is a perfectly reasonable question, with a potentially
informative answer. It is perfectly reasonable to ask someone to be
explicit about how certain they are of something.

Jorge Llambías

unread,
Nov 29, 2010, 8:27:18 AM11/29/10
to loj...@googlegroups.com
On Mon, Nov 29, 2010 at 12:33 AM, John E Clifford <kali9...@yahoo.com> wrote:
>  There is, after all, no grunts
> and wheezes that ask where one is in three (or four or more) space, one asks
> "Where the fugawee?"  Why should the matter be different for emotional
> space?

If there was a UI that was used to express you were at some particular
point in space, then tagging it with "pei" would be a perfectly
reasonable way to ask whether you were there.

There aren't any spatial points that are special enough across points
and general enough across people to have a UI assigned to express our
being there, and also it does not seem to be something humans feel
compelled to express.

Maybe a UI for "I'm here!" wouldn't be unreasonable. If there was such
a UI, we could use it to ask, although it would probably make more
sense as a COI. Let's say "bu'au" was such a COI. Then "bu'au pei
<name>" would be a good way to do a roll call.

mu'o mi'e xorxes.

John E Clifford

unread,
Nov 29, 2010, 10:06:19 AM11/29/10
to loj...@googlegroups.com
Well, since I made none of the claims or inferences mentioned here, I don't
entirely see the relevance of these remarks (although I agree that the moves are
made more often than they should be -- and not just in Lojban). Yes, UI is used
for several other purposes than expressing emotions and mark a variety of speech
acts. I have not yet found a useful way of expressing that explicitly but have
several times warned that I use the "express emotion" line as a shorthand for a
longer description. If you want to insist that I list all the things out, I
will beut it gets tedious for me and boring for you (and you will probably dig
up some cases I missed, anyhow). On the flip[ side, of course other things can
be used to express emotions: I suppose that 'mi gleki', for example, can be used
as expressively as "I'm happy", though with the same potential for
misunderstanding -- which 'ui' doesn't have. As for inference 3, I don't quite
know what you mean by "communicate". In something like a normal meaning, just
about every language act communicates something or other, as do most
non-linguistic acts. Presumably, 'ui' communicates, among other things, that
the speaker is happy. But it does not state that. It is neither true nor
false. It is evidence but not a claim. And so on. Do you mean something more
by "communicate"?
While I have a certain amount of difficulty with 'pei', I can see it usefulness,
both as another greeting and in more intense examination. I can't fathom
'uipei', however, except as an idiom of a particularly illogical sort. Come to
that, I have some trouble with your standard responses, which seem not to be
responsive at all -- well, 'nai' is OK, but 'ja'ai', aside from being an
innovations whose rationale is obscure, seems to be simply incoherent and have
nothing to do with what 'pei' is presumably asking (of course, the incoherence
may conceal a useful kernel). For the rest, is "ha ha" a legitimate answer to
"How are you feeling?" Maybe so, though in a rather extended sense. And of
course 'ie' is a perfectly good answer to 'xu do tugni' since that is in fact
its main purpose, as a "Yes" for a particular sort of question. But whether
'iepei' is a legitimate question is not thereby decided; it means "Yes, innit"
and it might be possible to extract from that something like a real question,
but the path is tortuous -- unless you just say it's an idiom, so don't expect
it to be logical.
The "grunts and wheezes" line was hyperbole, as you well know, But the point
there was that if it takes a real question (declarative statement with an
interrogative particle of some sort) to locate3 one in physical space, why
shouldn't it take one to locate one in emotional space. I don't think much of
the analogy, but it was Bergen's line not mine. But, along that same paragraph,
just what other purposes does 'ui' have than expressing happiness?
I have to admit that 'la'apei' makes more sense than 'uipei' and then suspect
that that sense carries over to other cases where it is relatively absurd. On
the other hand, the absurdity of 'uipei' makes me skeptical about 'la'apei' as
well. But. as I have said. we have an idiom here and though they are illogical,
we seem to be content to allow them, so let them ride (but they are another mark
against the "logical language" claim, even in the official restricted version).

From: Jorge Llambías <jjlla...@gmail.com>
To: loj...@googlegroups.com
Sent: Mon, November 29, 2010 7:04:46 AM

Subject: Re: [lojban] Time for the perenial other-centric-.ui conversation

On Mon, Nov 29, 2010 at 12:27 AM, John E Clifford <kali9...@yahoo.com> wrote:

--

John E Clifford

unread,
Nov 29, 2010, 10:16:02 AM11/29/10
to loj...@googlegroups.com
Well, yes, given we have the idiom, extending it to new cases seems perfectly
reasonable (once you shoot a hole in balloon, you might just as well tear the
whole thing up as you plunge downward). 'bu'au pei' would still be a dumb
question, since it can't meaningfully have the answer 'nai' and any positive
answer would be not specifically informative, i.e., 'cai' would be no different
from 'cu'i' or 'ro'e' -- contrary to the purported purpose of the idiom, Again,
this is not my amalogy and I fon't give much weight to it either way.


----- Original Message ----
From: Jorge Llambías <jjlla...@gmail.com>
To: loj...@googlegroups.com

Sent: Mon, November 29, 2010 7:27:18 AM
Subject: Re: [lojban] Time for the perenial other-centric-.ui conversation

--

Jonathan Jones

unread,
Nov 29, 2010, 2:10:39 PM11/29/10
to loj...@googlegroups.com
On Mon, Nov 29, 2010 at 8:06 AM, John E Clifford <kali9...@yahoo.com> wrote:
But whether 'iepei' is a legitimate question is not thereby decided; it means "Yes, innit"

 
No, it means, "Do you agree, and with what intensity do you (dis)agree?"

John E Clifford

unread,
Nov 29, 2010, 2:36:56 PM11/29/10
to loj...@googlegroups.com
Idiomatically, yes; literally, more like what I said.  And, as I have noted, idioms are anti-logical in the appropriate sense, since there is no mechanical rule to get them back to there logical form.


From: Jonathan Jones <eye...@gmail.com>
To: loj...@googlegroups.com
Sent: Mon, November 29, 2010 1:10:39 PM

Subject: Re: [lojban] Time for the perenial other-centric-.ui conversation

Jonathan Jones

unread,
Nov 29, 2010, 2:41:44 PM11/29/10
to loj...@googlegroups.com
No, by definition. No idiom needed. Must I quote the definition of {pei} again?


"pei - CAI - emotion ? - attitudinal: attitudinal question; how do you feel about it? with what intensity?"
 
In the case of {.iepei}, the "it" is "agreement": {.iepei} = "how do you feel about agreement? with what intensity?" = "Do you agree, and with what intensity do you (dis)agree?"

Luke Bergen

unread,
Nov 29, 2010, 3:00:45 PM11/29/10
to loj...@googlegroups.com

Lojban is a language "built on predicate logic", but does that mean that it is exclusively predicate logic?  The way that you talk about lojban and logic, John, it sounds like you would also be upset by joi or any of the other things that aren't explicit logical connectors

John E Clifford

unread,
Nov 29, 2010, 3:58:21 PM11/29/10
to loj...@googlegroups.com
Well, I didn't say where the idiom strarted but, as I suspected. apparently it is with 'pei'.  'ie' is not an it, but a mark of agreement. the "yes" of my version. 'pei' is Lord knows what,. but apparently a question about some of the things that fall into UI and the like.  But what kind of a question i9s it? Not a "yes/no" clearly, so an identity question.  But it doesn't look like an identity question (or a yes/no one, for that matter) and something that performs a linguistic function that it is not designed to do is an idiom. 'pei' is in CAI (I assume), a rough marker of intensity of emotion (or whatever in the flotsam and jetsam of discursives).  The only discursive here is one about MY agreement -- it can't be anyone else's.  So, the whole means "I agree, how much?" or, presumably, "How much do I agree?"  You don't come into it at all.  Now, I also think that using a CAT to ask a factual question is a mistake, but that doesn't affect the confusion that is involved in this idiom (propagated, it seems, from the core).


From: Jonathan Jones <eye...@gmail.com>
To: loj...@googlegroups.com
Sent: Mon, November 29, 2010 1:41:44 PM

Jonathan Jones

unread,
Nov 29, 2010, 4:11:08 PM11/29/10
to loj...@googlegroups.com
Okay, no. Just no. It's "You agree how much?" or "How much do you agree?" It is asking a question of the listener. Every other question word in Lojban is asking the LISTENER, not the speaker. {pei} is no different in that respect.

John E Clifford

unread,
Nov 29, 2010, 4:12:04 PM11/29/10
to loj...@googlegroups.com
Having spent thirty five years adapting Logjam and logic to more or less coincide  -- working on extensions of logic as well as of Logjam, I think it fair to say that you are talking through a nether orifice.  I remember the fights I had to do to get 'joi' in (well, pre-'joi' since this was back in Loglan).  Won that one eventually, lost a lot of others -- like trying to organize UI and kin into something resembling logical structures (I won't even talk about the semantic mishmash there).  The "built on" line used to mean that logical transformation were transparent (or, at least, rule governed starting with the surface structure).  That typically involved scraping away a lot of stuff, but, again, the scraping was rule governed: a question had certain marks to identify it, similarly an exhortation or a promise or an expression of hope.  And eventually we got down to the declarative core and the logical rues applied.  There was never the intention of sticking with that core (you can't talk very long in it without boring the socks off everyone) but it was the heart.  What 'iepei' (and I would bet a number of other things as well) does is take away the rule-like behavior of the scraping process.  This may not be a catastrophe, but it is a serious knock on the claim to be, in any sense, a logical language.


From: Luke Bergen <lukea...@gmail.com>
To: loj...@googlegroups.com
Sent: Mon, November 29, 2010 2:00:45 PM

Jonathan Jones

unread,
Nov 29, 2010, 4:14:02 PM11/29/10
to loj...@googlegroups.com
Why do you keep reffering to Lojban as Logjam?

Craig Daniel

unread,
Nov 29, 2010, 4:15:42 PM11/29/10
to loj...@googlegroups.com
On Mon, Nov 29, 2010 at 4:14 PM, Jonathan Jones <eye...@gmail.com> wrote:
> Why do you keep reffering to Lojban as Logjam?

He's not - he's referring to Lojban and its precursor, Loglan, in one
breath, and making fun of the cat-herd of a mess that has been
Loglan/Lojban history and culture from day one. And he was there, so
he would know.

John E Clifford

unread,
Nov 29, 2010, 4:18:26 PM11/29/10
to loj...@googlegroups.com
Right, it asking the listener how much the speaker agrees.  And that is a proper question, not some fiddle-faddle off in CAI.  But a remarkably dumb question in most contexts.  It would be better to get a nice snappy way of saying "how much in proper bridi form, rather than the runaround not required, so that one could simply say 'do togni ---?'  Maybe that is 'pei' but then it is probably in the wrong selma'o.


From: Jonathan Jones <eye...@gmail.com>
To: loj...@googlegroups.com
Sent: Mon, November 29, 2010 3:11:08 PM

Jonathan Jones

unread,
Nov 29, 2010, 4:19:33 PM11/29/10
to loj...@googlegroups.com
No, it's asking the listener how much the listener agrees.

John E Clifford

unread,
Nov 29, 2010, 4:24:23 PM11/29/10
to loj...@googlegroups.com
1.  Because I get tired of talking about "Loglan and/or Lojban (and/or LoCCan3)"  2.Because I think that Lojban ut nunc has some serious problems that are pretty regularly avoided (usually, by running off on tangents about what a word means or whether it fits in this slot or...) and "Logjam" is a mildly amusing (to me, anyhow) way or indicating that.  The problems often go back to Loglan, so the extended reference is appropriate. 3. Oh, I've already said that I think it is funny. 

From: Jonathan Jones <eye...@gmail.com>
To: loj...@googlegroups.com
Sent: Mon, November 29, 2010 3:14:02 PM

John E Clifford

unread,
Nov 29, 2010, 4:25:24 PM11/29/10
to loj...@googlegroups.com
Full marks (though I would like to think you found it funny as well).

----- Original Message ----
From: Craig Daniel <craigb...@gmail.com>
To: loj...@googlegroups.com
Sent: Mon, November 29, 2010 3:15:42 PM
Subject: Re: [lojban] Time for the perenial other-centric-.ui conversation

--

John E Clifford

unread,
Nov 29, 2010, 4:26:55 PM11/29/10
to loj...@googlegroups.com
That may be what you (think you) are doing, but it is not what you said. 


From: Jonathan Jones <eye...@gmail.com>
To: loj...@googlegroups.com
Sent: Mon, November 29, 2010 3:19:33 PM

Craig Daniel

unread,
Nov 29, 2010, 4:27:25 PM11/29/10
to loj...@googlegroups.com
On Mon, Nov 29, 2010 at 4:25 PM, John E Clifford <kali9...@yahoo.com> wrote:
> Full marks (though I would like to think you found it funny as well).

Less so after many repetitions, but still at least a little bit so
when I'm in the right mood.

Jonathan Jones

unread,
Nov 29, 2010, 4:28:44 PM11/29/10
to loj...@googlegroups.com
That is what the definition of {pei} is.

John E Clifford

unread,
Nov 29, 2010, 4:33:33 PM11/29/10
to loj...@googlegroups.com
First time is the most I would expect.

----- Original Message ----
From: Craig Daniel <craigb...@gmail.com>
To: loj...@googlegroups.com
Sent: Mon, November 29, 2010 3:27:25 PM
Subject: Re: [lojban] Time for the perenial other-centric-.ui conversation

--

John E Clifford

unread,
Nov 29, 2010, 4:36:45 PM11/29/10
to loj...@googlegroups.com
Okay, the "I agree.  How do you feel about that?  With what intensity?"  A better question, but surely the answer will not be 'iecai' (nor should it be -- but then that shouldn't have been the question either).


From: Jonathan Jones <eye...@gmail.com>
To: loj...@googlegroups.com
Sent: Mon, November 29, 2010 3:28:44 PM

Jorge Llambías

unread,
Nov 29, 2010, 5:55:50 PM11/29/10
to loj...@googlegroups.com
On Mon, Nov 29, 2010 at 12:06 PM, John E Clifford <kali9...@yahoo.com> wrote:
> Well, since I made none of the claims or inferences mentioned here, I don't
> entirely see the relevance of these remarks (although I agree that the moves are
> made more often than they should be -- and not just in Lojban).

Sometimes you seem to pick a property of some particular UI and
generalize it indiscriminately to all UIs. Like calling them "grunts".
Yes, some UIs can be likened in some respects to grunts, but others
have nothing in common with them.

> Yes, UI is used
> for several other purposes than expressing emotions and mark a variety of speech
> acts.

Right.

> I have not yet found a useful way of expressing that explicitly but have
> several times warned that I use the "express emotion" line as a shorthand for a
> longer description.  If you want to insist that I list all the things out, I
> will beut it gets tedious for me and boring for you (and you will probably dig
> up some cases I missed, anyhow).

I don't insist that you do anything, but if you ever feel like making
that list, it might be quite interesting. I don't think I would find
it boring.

>  On the flip[ side, of course other things can
> be used to express emotions: I suppose that 'mi gleki', for example, can be used
> as expressively as "I'm happy",

Yes.

>though with the same potential for
> misunderstanding -- which 'ui' doesn't have.

I'm not sure what kind of misunderstanding you have in mind.

> As for inference 3, I don't quite
> know what you mean by "communicate".  In something like a normal meaning, just
> about every language act communicates something or other, as do most
> non-linguistic acts.

I was contrasting the merely expressive function of language, where an
interlocutor is irrelevant, with its communicative function, where an
interlocutor or intended audience is crucial.

> Presumably, 'ui' communicates, among other things, that
> the speaker is happy.

In its core use, the speaker just expresses happiness when saying
"ui". If that happens to communicate their happiness, it is as a side
effect. But that's "ui", not UI, just in case someone wants to make
the usual jump of "whatever we say of 'ui' goes for all of UI". For
many UIs, an audience is crucial or at least highly relevant.

> But it does not state that.  It is neither true nor
> false.  It is evidence but not a claim.  And so on.  Do you mean something more
> by "communicate"?

I hope I made myself a little more clear now.

> While I have a certain amount of difficulty with 'pei', I can see it usefulness,
> both as another greeting and in more intense examination.  I can't fathom
> 'uipei', however, except as an idiom of a particularly illogical sort.

"uipei" is not the best choice to examine "pei", just because "ui" is
mainly purely expressive, and asking someone to provide a purely
expressive locution is slightly silly. (But still meaningful.) But
don't go and conlude from that that "UIpei" in general is silly.
"UIpei" does not mean, as you seem to think, that the speaker says UI
and then, independently, asks the listener to provide some kind of
comment on that. "pei" modifies the meaning of the preceding word, in
such a way that "UIpei" is a question. "Uipei" has a perfectly
compositional meaning, but it is not the meaning of UI and then
separately the meaning of "pei". (The same can be said of "UInai" for
example. When you say "UInai" you are not expressing something with
UI, and then somehow reversing what you just expressed. You are
expressing something with "UInai".) UIpei asks the listener to answer
with "UI" or "UInai". It's really quite simple, and it seems to me you
are just trying hard to not understand.

> Come to
> that, I have some trouble with your standard responses, which seem not to be
> responsive at all -- well, 'nai' is OK, but 'ja'ai', aside from being an
> innovations whose rationale is obscure, seems to be simply incoherent and have
> nothing to do with what 'pei' is presumably asking (of course, the incoherence
> may conceal a useful kernel).

"UI ja'ai" is equivalent to UI by itself, I'm quite sure you
understood that perfectly. It is the identity modifier in the CAI/NAI
series, if you prefer to put it that way.

> For the rest, is "ha ha" a legitimate answer to
> "How are you feeling?" Maybe so, though in a rather extended sense. And of
> course 'ie' is a perfectly good answer to 'xu do tugni' since that is in fact
> its main purpose, as a "Yes" for a particular sort of question.

"ie" makes more sense in response to a statement than to a question,
because a question makes no claim with which to agree or disagree.

Pragmatically, since the question as posed is about agreement,
answering "ie" would probably be understood, but strictly speaking
there is no claim advanced with which to agree or disagree.

> But whether
> 'iepei' is a legitimate question is not thereby decided; it means "Yes, innit"
> and it might be possible to extract from that something like a real question,
> but the path is tortuous -- unless you just say it's an idiom, so don't expect
> it to be logical.

No, you are simply wrong about how "pei" works. It is perfectly
compositional, and it forms a compound with the preceding UI that asks
the listener to answer with "ie[ja'ai]" or "ienai" (or any of the
other variants).

> The "grunts and wheezes" line was hyperbole, as you well know,  But the point
> there was that if it takes a real question (declarative statement with an
> interrogative particle of some sort) to locate3 one in physical space, why
> shouldn't it take one to locate one in emotional space.

Because we happen to have words to express one's location in emotional
space but we just don't happen to have any to express one's location
in physical space (other than propositionally).

> I don't think much of
> the analogy, but it was Bergen's line not mine.  But, along that same paragraph,
> just what other purposes does 'ui' have than expressing happiness?

"ui" is for expressing happiness. UI has lots and lots of different
purposes besides expressing emotions. UI is not only, maybe not even
mainly, for expressing emotions.

> I have to admit that 'la'apei' makes more sense than 'uipei' and then suspect
> that that sense carries over to other cases where it is relatively absurd.  On
> the other hand, the absurdity of 'uipei' makes me skeptical about 'la'apei' as
> well.

See? You think that some property of one UI somehow must generalize to
all UIs. But there's no reason to expect that, because other than
their common syntax, there's not that much that can be said of all of
them in general.

> But. as I have said. we have an idiom here and though they are illogical,

No we don't. "UIpei" is no more idiomatic than "UInai" or "UIcu'i", or
many other compounds with compositional meaning. Even less so, since
UInai is sometimes hard to figure, but UIpei is always transparent.

> we seem to be content to allow them, so let them ride (but they are another mark
> against the "logical language" claim, even in the official restricted version).

Logic doesn't really enter into it, but "pei" is certainly nice and regular.

Oren

unread,
Nov 29, 2010, 10:09:05 PM11/29/10
to loj...@googlegroups.com
This lengthy conversation seems like a stark gap in interpretation of a moderately used word. Doesn't LLC have some like, official example sentence that explains the usage of a word? Isn't this what those cmavo definitions are for? It seems really silly to watch this go back and forth without any examples. 

2010/11/29 Jorge Llambías <jjlla...@gmail.com>
--
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.




--
Oren Robinson
102 Morrison Ave
Somerville, MA 02144

John E. Clifford

unread,
Nov 29, 2010, 10:40:27 PM11/29/10
to loj...@googlegroups.com
Well, the  CLL examples (and the discussions) are mixed bag, often dwelling on cute but peripheral issues --the section on 'sei' is a mildly typical case.  Also, of course, sometimes CLL just gets it wrong.  I didn't realize we were short on examples, but then they would hardly solve anything since we disagree about what the examples meant and whether they were legitimate examples even.
As for this being lengthy, you ain't seen nuthin yet.  Wait till the next go round of the genuine perennial conversation, the one about 'lo' which I came into the middle of 35 years ago and hasn't stopped yet.  Even the official line has changed half a dozen times just in that period (and at least once before).

Sent from my iPad

Robin Lee Powell

unread,
Nov 29, 2010, 11:58:35 PM11/29/10
to loj...@googlegroups.com
On Mon, Nov 29, 2010 at 10:09:05PM -0500, Oren wrote:
> This lengthy conversation seems like a stark gap in interpretation
> of a moderately used word. Doesn't LLC have some like, official
> example sentence that explains the usage of a word? Isn't this
> what those cmavo definitions are for? It seems really silly to
> watch this go back and forth without any examples.

You mean LLG.

I haven't been paying attention, because PC (John Clifford) was
involved, and that's just how I roll. I had been under the
impression they were arguing over the semantics of UI in general,
and potentially new UI in particular. If they're arguing over the
meaning of particular word, which one?

-Robin

--
http://singinst.org/ : Our last, best hope for a fantastic future.
Lojban (http://www.lojban.org/): The language in which "this parrot
is dead" is "ti poi spitaki cu morsi", but "this sentence is false"
is "na nei". My personal page: http://www.digitalkingdom.org/rlp/

Craig Daniel

unread,
Nov 30, 2010, 12:11:44 AM11/30/10
to loj...@googlegroups.com
On Mon, Nov 29, 2010 at 11:58 PM, Robin Lee Powell
<rlpo...@digitalkingdom.org> wrote:
> On Mon, Nov 29, 2010 at 10:09:05PM -0500, Oren wrote:
>> This lengthy conversation seems like a stark gap in interpretation
>> of a moderately used word. Doesn't LLC have some like, official
>> example sentence that explains the usage of a word? Isn't this
>> what those cmavo definitions are for? It seems really silly to
>> watch this go back and forth without any examples.
>
> You mean LLG.
>
> I haven't been paying attention, because PC (John Clifford) was
> involved, and that's just how I roll.  I had been under the
> impression they were arguing over the semantics of UI in general,
> and potentially new UI in particular.  If they're arguing over the
> meaning of particular word, which one?

Originally, da'oi (an experimental COI some people have proposed).
Then it was dai for a while, and now it's mostly pei.

Robin Lee Powell

unread,
Nov 30, 2010, 12:15:51 AM11/30/10
to loj...@googlegroups.com

*nod*

If someone wants to summarize the confusion over dai and pei and put
it in the appropriate BPFK section, or email it to me, I'd
appreciate the help.

Craig Daniel

unread,
Nov 30, 2010, 12:26:59 AM11/30/10
to loj...@googlegroups.com
On Tue, Nov 30, 2010 at 12:15 AM, Robin Lee Powell

<rlpo...@digitalkingdom.org> wrote:
> On Tue, Nov 30, 2010 at 12:11:44AM -0500, Craig Daniel wrote:
>> On Mon, Nov 29, 2010 at 11:58 PM, Robin Lee Powell
>> <rlpo...@digitalkingdom.org> wrote:
>> > On Mon, Nov 29, 2010 at 10:09:05PM -0500, Oren wrote:
>> >> This lengthy conversation seems like a stark gap in
>> >> interpretation of a moderately used word. Doesn't LLC have some
>> >> like, official example sentence that explains the usage of a
>> >> word? Isn't this what those cmavo definitions are for? It seems
>> >> really silly to watch this go back and forth without any
>> >> examples.
>> >
>> > You mean LLG.
>> >
>> > I haven't been paying attention, because PC (John Clifford) was
>> > involved, and that's just how I roll.  I had been under the
>> > impression they were arguing over the semantics of UI in
>> > general, and potentially new UI in particular.  If they're
>> > arguing over the meaning of particular word, which one?
>>
>> Originally, da'oi (an experimental COI some people have proposed).
>> Then it was dai for a while, and now it's mostly pei.
>
> *nod*
>
> If someone wants to summarize the confusion over dai and pei and put
> it in the appropriate BPFK section, or email it to me, I'd
> appreciate the help.

On it. But not tonight; I should go to bed.

- mi'e .kreig.

John E Clifford

unread,
Nov 30, 2010, 2:17:13 PM11/30/10
to loj...@googlegroups.com
This is going to be long and boring, if you are not interested in the topic
(and probably even if you are) so feel even freer than usual to skip it.
e xorxe

On Mon, Nov 29, 2010 at 12:06 PM, John E Clifford <kali9...@yahoo.com> wrote:
> Well, since I made none of the claims or inferences mentioned here, I don't
> entirely see the relevance of these remarks (although I agree that the moves
>are
> made more often than they should be -- and not just in Lojban).

Sometimes you seem to pick a property of some particular UI and
generalize it indiscriminately to all UIs. Like calling them "grunts".
Yes, some UIs can be likened in some respects to grunts, but others
have nothing in common with them.

[I refuse to be attacked for my jokes. Indeed, no UI sounds much like a grunt
since they don’t have final consonants, nor like a wheeze either (no initial
h). But, if that is the worst generalization I made (though I didn’t make that
one), I don’t feel too bad anyhow. I have indeed generalized about classes of
UI (and related areas) and perhaps not been clear about the limits of the
generality – just as I have used “express” for a number of different activities,

with diminishing similarities as we get farther from the core. Sorry ‘bout
that. Perhaps if the list were a bit better organized, it would be possible to
deal with problems more precisely. As it is, it is hard to remember, just
looking at it, whether a word is a pure expression of a feeling or a world
changer, so some of my particular remarks may have been out of line for
particular example. But I do think we have not gotten very far from expressions

proper in this discussion, so the fault may not be very great.]

> Yes, UI is used
> for several other purposes than expressing emotions and mark a variety of
>speech
> acts.

Right.

> I have not yet found a useful way of expressing that explicitly but have
> several times warned that I use the "express emotion" line as a shorthand for
a
> longer description. If you want to insist that I list all the things out, I
> will beut it gets tedious for me and boring for you (and you will probably dig
> up some cases I missed, anyhow).

I don't insist that you do anything, but if you ever feel like making
that list, it might be quite interesting. I don't think I would find
it boring.

> On the flip[ side, of course other things can
> be used to express emotions: I suppose that 'mi gleki', for example, can be
>used
> as expressively as "I'm happy",

Yes.

>though with the same potential for
> misunderstanding -- which 'ui' doesn't have.

I'm not sure what kind of misunderstanding you have in mind.

[Of course you are. The one between a slightly more complex expression of joy
than ‘ui’ and a claim about your emotional state. The one is sincere or not,
the other is true or false.]

> As for inference 3, I don't quite
> know what you mean by "communicate". In something like a normal meaning, just
> about every language act communicates something or other, as do most
> non-linguistic acts.

I was contrasting the merely expressive function of language, where an
interlocutor is irrelevant, with its communicative function, where an
interlocutor or intended audience is crucial.

> Presumably, 'ui' communicates, among other things, that
> the speaker is happy.

In its core use, the speaker just expresses happiness when saying
"ui". If that happens to communicate their happiness, it is as a side
effect. But that's "ui", not UI, just in case someone wants to make
the usual jump of "whatever we say of 'ui' goes for all of UI". For
many UIs, an audience is crucial or at least highly relevant.

[Is that jump really usual? If it is, maybe we should look to see if there may
not be something in it. I don’t think I made that jump, by the way.]

> But it does not state that. It is neither true nor
> false. It is evidence but not a claim. And so on. Do you mean something
more
> by "communicate"?

I hope I made myself a little more clear now.

[Yup. But I am not clear about the relevance. To be sure, a person can say
‘ui’ when there is no one about, and be performing the same act she does when
someone is about. But that is true of most speech acts (though occasionally we
have to split the speaker into speaker and hearer). Now, it is the case that
the rational use of ‘pei’ does require another person, but the split self
routine still works.]

> While I have a certain amount of difficulty with 'pei', I can see it
>usefulness,
> both as another greeting and in more intense examination. I can't fathom
> 'uipei', however, except as an idiom of a particularly illogical sort.

"uipei" is not the best choice to examine "pei", just because "ui" is
mainly purely expressive, and asking someone to provide a purely
expressive locution is slightly silly. (But still meaningful.) But
don't go and conlude from that that "UIpei" in general is silly.
"UIpei" does not mean, as you seem to think, that the speaker says UI
and then, independently, asks the listener to provide some kind of
comment on that. "pei" modifies the meaning of the preceding word, in
such a way that "UIpei" is a question. "Uipei" has a perfectly
compositional meaning, but it is not the meaning of UI and then
separately the meaning of "pei". (The same can be said of "UInai" for
example. When you say "UInai" you are not expressing something with
UI, and then somehow reversing what you just expressed. You are
expressing something with "UInai".) UIpei asks the listener to answer
with "UI" or "UInai". It's really quite simple, and it seems to me you
are just trying hard to not understand.


[ But of course ‘uinai’ is a simple blend of ‘ui’ and ‘nai’: “Whee – not!”,
totally natural (well, only lately) English, as is “Whee – sorta” and the like.

What is a case where this sort of thing is not true? So, ‘uipei’ comes out to
mean “Whee – but how much?” or something like that, possibly meaningfull but
basically dumb -- nothing like the use you claim for it. Something like it has
a perfectly meaningful use, of course: A: ui B: pei. No change of meaning of
either ‘ui’ or ‘pei’ and a sensible question (if a bit rude in this case). But
you would have attaching ‘pei’ change ‘ui’ from a first person expression to a
prompting for a second person expression, with a tag yet. None of the others
work anything like that.]

Ø Come to


> that, I have some trouble with your standard responses, which seem not to be
> responsive at all -- well, 'nai' is OK, but 'ja'ai', aside from being an
> innovations whose rationale is obscure, seems to be simply incoherent and have
> nothing to do with what 'pei' is presumably asking (of course, the incoherence
> may conceal a useful kernel).

"UI ja'ai" is equivalent to UI by itself, I'm quite sure you
understood that perfectly. It is the identity modifier in the CAI/NAI
series, if you prefer to put it that way.

Ø
Ø [Well, as I said, this is an innovation whose purpose is obscure. I
suppose it is meant to reassure that I really meant thisUI rather than some
other: “Whee – yes indeed”, probably in response to a ‘pei’. It seems like
there are other places where something like this would be more useful, but I
most of them can be covered by the placement of ‘ja’a’]

> For the rest, is "ha ha" a legitimate answer to
> "How are you feeling?" Maybe so, though in a rather extended sense. And of
> course 'ie' is a perfectly good answer to 'xu do tugni' since that is in fact
> its main purpose, as a "Yes" for a particular sort of question.

"ie" makes more sense in response to a statement than to a question,
because a question makes no claim with which to agree or disagree.

Pragmatically, since the question as posed is about agreement,
answering "ie" would probably be understood, but strictly speaking
there is no claim advanced with which to agree or disagree.

Ø
Ø [I suppose this is a contextual matter. One doesn’t ordinarily ask for
agreement unless there has been a position set out already, the x2 and x3 of
‘tugni’. I couldn’t think of a case to lay out, so I skipped it, figuring tou
could fill in the blanks.]

> But whether
> 'iepei' is a legitimate question is not thereby decided; it means "Yes, innit"
> and it might be possible to extract from that something like a real question,
> but the path is tortuous -- unless you just say it's an idiom, so don't expect
> it to be logical.

No, you are simply wrong about how "pei" works. It is perfectly
compositional, and it forms a compound with the preceding UI that asks
the listener to answer with "ie[ja'ai]" or "ienai" (or any of the
other variants).

Ø
Ø [See above. In what sense is it compositional? It more like Montague
grammar than item and arrangement (or process); the end result has practically
nothing to do with its parts: the first person ‘ui’ has been put in someone
else’s mouth. The question, which made sense of after someones exclamation, is
asking for such an exclamation, whether or not the second person was so
inclined. And so on.]
Ø

> The "grunts and wheezes" line was hyperbole, as you well know, But the point
> there was that if it takes a real question (declarative statement with an

> interrogative particle of some sort) to locate one in physical space, why


> shouldn't it take one to locate one in emotional space.

Because we happen to have words to express one's location in emotional
space but we just don't happen to have any to express one's location
in physical space (other than propositionally).

Ø
Ø [I doubt that it is happenstance, but still, even if we had expressions
of some sort to express our perceived location, it would still make sense to as
for a GPS reading in a straightforward question. And to get an answer in a
straightforward declarative sentence. And no amount of expressive content is
going to come up to that standard.]
Ø
Ø > I don't think much of


> the analogy, but it was Bergen's line not mine. But, along that same
>paragraph,
> just what other purposes does 'ui' have than expressing happiness?

"ui" is for expressing happiness. UI has lots and lots of different
purposes besides expressing emotions. UI is not only, maybe not even
mainly, for expressing emotions.

Ø
Ø [You here (and occasionally earlier) seem to be making the confusion
between ‘ui’ and “UI” of which you unjustly accuseme. I asked about the word
‘ui’ not about the whole class UI. Though, the question applies equally to each

member of that class: what else does it do than such and such?]
Ø

> I have to admit that 'la'apei' makes more sense than 'uipei' and then suspect
> that that sense carries over to other cases where it is relatively absurd. On
> the other hand, the absurdity of 'uipei' makes me skeptical about 'la'apei' as
> well.

See? You think that some property of one UI somehow must generalize to
all UIs. But there's no reason to expect that, because other than
their common syntax, there's not that much that can be said of all of
them in general.

Ø
Ø [ No, I am quite aware that ‘la’a’ is a disclaimer not the expression of
an emotion and personal only insofar as it is one persons assessment of an
objective reality. This connection with objective reality makes the question
possible without contradiction. But the fact that it is personal to any extent
does make me wonder if it is also as logically absurd as ‘uipei’. However, on
the other side, ‘la’apei’ makes sense even before the second says ‘la’a’, again
thanks to the objective component. I keep going back and forth on this one. I
have no problem coming down on “absurd” with ‘uipei’.]
Ø

> But. as I have said. we have an idiom here and though they are illogical,

No we don't. "UIpei" is no more idiomatic than "UInai" or "UIcu'i", or
many other compounds with compositional meaning. Even less so, since
UInai is sometimes hard to figure, but UIpei is always transparent.

> we seem to be content to allow them, so let them ride (but they are another
>mark
> against the "logical language" claim, even in the official restricted
version).

Logic doesn't really enter into it, but "pei" is certainly nice and regular.

[Well, no. ‘nai’, say, takes a first person expression and then modifies it in
this case rejecting it “Whee – not”, as we say, and similarly for “Whee – sorta”

and so on. But ‘pei’ does not start out with a first person expression and add
something to it. It somehow changes the first person expression into a second
person and then asks about it. There is a perfectly legitimate (is so far as
‘pei’ is legitimate at all) use that looks like this: Speaker says ‘ui’,‘pei’say

the hearer. No person shifting and a reasonable sort of thing to ask.]



Jorge Llambías

unread,
Nov 30, 2010, 6:07:39 PM11/30/10
to loj...@googlegroups.com
On Tue, Nov 30, 2010 at 4:17 PM, John E Clifford <kali9...@yahoo.com> wrote:
>
> [ But of course ‘uinai’ is a simple blend of ‘ui’ and ‘nai’: “Whee – not!”,
> totally natural (well, only lately) English, as is “Whee – sorta” and the like.

No, "uinai" is an expression of saddness, it is not an expression of
happiness followed by the negation (or reversal) of the preceding
expression.

> What is a case where this sort of thing is not true? So, ‘uipei’ comes out to
> mean “Whee – but how much?” or something like that, possibly meaningfull but
> basically dumb  -- nothing like the use you claim for it.

You misanalyze both "uinai" and "uipei".

> Something like it has
> a perfectly meaningful  use, of course: A: ui   B: pei. No change of meaning of
> either ‘ui’ or ‘pei’ and a sensible question (if a bit rude in this case).  But
> you would have attaching ‘pei’ change ‘ui’ from a first person expression to a
> prompting for a second person expression, with a tag yet.  None of the others
> work anything like that.]

You seem to be confusing words with expressions (i.e. the use of
words). "nai" does not change an expression, it changes a word. If you
really think that expressing sadness is akin to a blend of expressing
happiness and then doing something else, (expressing reversal? or
what?) then I'm afraid we won't get anywhere with this.

> "UI ja'ai" is equivalent to UI by itself, I'm quite sure you
> understood that perfectly. It is the identity modifier in the CAI/NAI
> series, if you prefer to put it that way.
> Ø
> Ø      [Well, as I said, this is an innovation whose purpose is obscure.

It is often necessary to have the syntactic support even when the
semantic effect is neutral.

If you understand the purpose of "ja'a" and "ja'e", you will also
understand the purpose of "ja'ai".

> I
> suppose it is meant to reassure that I really meant thisUI rather than some
> other: “Whee – yes indeed”, probably in response to a ‘pei’.  It seems like
> there are other places where something like this would be more useful, but I
> most of them can be covered by the placement of ‘ja’a’]

I'm surprised that you think "ja'a" can replace "ja'ai", since "ja'a"
is a strictly propositional operator.

>> And of
>> course 'ie' is a perfectly good answer to 'xu do tugni' since that is in fact
>> its main purpose, as a "Yes" for a particular sort of question.
>
> "ie" makes more sense in response to a statement than to a question,
> because a question makes no claim with which to agree or disagree.
>
> Pragmatically, since the question as posed is about agreement,
> answering "ie" would probably be understood, but strictly speaking
> there is no claim advanced with which to agree or disagree.
> Ø
> Ø      [I suppose this is a contextual matter.  One doesn’t ordinarily ask for
> agreement unless there has been a position set out already, the x2 and x3  of
> ‘tugni’.  I couldn’t think of a case to lay out, so I skipped it, figuring tou
> could fill in the blanks.]

A: xu do mi tugni lo du'u lo snime cu blabi
B: ie

The expected answers to a xu-question are "go'i" or "na go'i", meaning:

go'i= mi do tugni lo du'u lo snime cu blabi
na go'i= mi do na tugni lo du'u lo snime cu blabi

The answer "ie" is pragmatically acceptable, but strictly speaking
doesn't make much sense. It doesn't mean: "ie [go'i]", B agreeing that
they agree, because there was no claim put forth that they agree. It
means "ie [lo snime cu blabi]", but that claim was not put forth
directly either.

The natural use of ïe is in:

A: lo snime cu blabi
B: ie

As a response to a claim, not to a xu-question, or also:

A: lo snime cu blabi iepei
B: ie

as a response to a "iepei" question.

> Logic doesn't really enter into it, but "pei" is certainly nice and regular.
>
> [Well, no.  ‘nai’, say, takes a first person expression and then modifies it in
> this case rejecting it “Whee – not”, as we say, and similarly for “Whee – sorta”
>
> and so on.

No, that's wrong. When someone says "uinai", they don't start by
expressing happiness. The only thing they express is sadness.

> But ‘pei’ does not start out with a first person expression and add
> something to it.

The word "pei" modifies the preceding word. The compound "UI pei" is
used to ask a question. The meaning of "UI pei" is easily and
regularly derived from the meanings of "UI" and "pei".

> It somehow changes the first person expression into a second
> person and then asks about it.

You keep confusing words with expressions (=the use of words).

> There is a perfectly legitimate (is so far as
> ‘pei’ is legitimate at all) use that looks like this: Speaker says ‘ui’,‘pei’say
> the hearer.  No person shifting and a reasonable sort of thing to ask.]

That's not how "pei" is meant to work, and not how it works in practice either.

It is loading more messages.
0 new messages