Re: ciska bai tu'a zo bai

[Logical Language Group <lojbab>]

la lojbab. cusku di'e

> =In this case, however, Zipf may argue for a reversal of meaning for ga'i,
> =since almost all examples of actual usage are of "ga'inai", and a lot more
> =people have reason to be obsequious than blatantly pompous.

la nitcion. cusku di'e

> Ouch; major cultural presumption there.

Agreed, but the mere choice to use polar scales and mark only one end
(.ui vs. uinai) is a cultural bias.  Reversing a scale or two doesn't make
much difference.

> The story isn't that simple.

[...]

> relational:
> 	speaker & referent (referent honorific)
> 	speaker & addressee (addressee honorific)
> 	speaker & bystander (bystander honorific)
> 	speaker & setting (formality levels)
> absolute: (eg. emperor pronouns; gender pronouns)

Is this really absolute?  How does the emperor refer to his deceased
ancestor: with a god pronoun?  What about other emperors?

[Korean example deleted]

> So what of our ga'i? Oh. I've just realised I've raised the wrong point here;
> Lojbab and the others are talking about what the deictic centre of {ga'i}
> should be --- that is, what is the honorific relative to: does it mean
> "I am honourable relative to the referent" or "the referent is honourable
> relative to me". Lojbab's answer is consistent with other UI cmavo; for
> the alternative, we might allow {do'a} (or whatever the empathy cmavo is
> these days) to shift the deictic centre --- and, since the referents will 
> often be inanimate, we should take the opportunity to change its definition
> to "deictic centre shift". 

It's "dai".  But I think we leave the definition alone: "empathy" is confusing
enough without dragging in "deictic reference shift".  I will add a note to the
paper saying that in some circumstances we attribute emotion (or its analogs)
to inanimate objects:

	le bloti .uudai klama le xasloi
	The boat [Pity!] [empathy] goes-to the ocean:floor
	The boat, poor thing, sank.

> The reason Lojbab and John wouldn't change it to this when I first talked 
> of it is that attitudinals are always self-expressions, and so are always 
> centred on the "self"; "empathy" is something that can still be attributed 
> to the self, whereas "deictic centre shift" would mean that .o'onaido'a
> would mean "He's angry!", not "Oh, I think he's angry" or "He must be angry".

I think in the realm of attitudinals proper, this distinction is unreal.
You do not >know< that someone else is angry; you infer it from your
observation (za'a.o'onaidai) or your intuition (se'o.o'onaidai).

> Well, that'll generate some response from Lojban Central. The other thing
> we should beware of is not to conflate addressee and referent honorifics.
> What does {le patfu cu klama vauga'inai} mean? Does the father outrank me,
> or does the person I'm talking to?

Neither: see below.

> What about {le ga'inai patfu cu klama}?

That means that the father outranks you.

> Do we in fact have a way of doing addressee honorifics as distinct from
> referent honorifics?
> 
> I suspect that, due to the nature of UI cmavo, {ga'i} is a referent honorific
> (so father is the outranker in the *second* example), and that the only
> way to do addressee honorification is a la tu/vous, by making the addressee
> grammatically part of the utterance: {doido'uga'inai le patfu cu klama}.

I agree completely.  Another, rather less absolute, version is

	do ga'inai zo'u le patfu cu klama
	You [who outrank me] : the father goes

> What the interpretation of {le patfu cu klama vauga'inai} should be is
> a mystery to me. It should *not* be construed as an addressee honorific;
> that's playing havoc with UI semantics (such as they are). It can only
> mean... that you're honouring the sentence.

Not quite: you are honoring the >referent< of the sentence: as you say,
we have >referent< honorifics.  So you are honoring an event.  This probably
makes more sense with "ga'i":

	le xarju cu citka vauga'i
	The pig ate [which is an event beneath my notice].

> The problem is that attitudinals proper can hang off sentences because they
> aren't really relational between two entities: they are centered on the
> self. Honorifics don't work like that; they're not about "Oh, I feel humble"
> in the same ways as "Oh, I feel happy". They honour someone specific.

Yes: this is the difference between "ga'i" and ".o'a".

-- 
John Cowan		sharing account <lojbab@access.digex.net> for now
		e'osai ko sarji la lojban.