Re: replies re. ka & mamta be ma

[ucleaar <ucleaar@ucl.ac.uk>]

Chris:
> >"Go away" is imperative, and the grammar says it is a command. "I
> >command you to go away" is declarative, and the grammar says it is
> >an assertion. The same goes for Lojban. The exact nature of the
> >difference is still a matter of debate, but [...]
> Perhaps adding "ca'e" would help: does "ko klama" mean "ca'e mi minde lenu
> do klama"?

{cahe} = "I define"? I don't understand. I suspect you might be seeking
a rendition of "hereby", though I don't see how this might relate to
{cahe}. Anyway, if it is "hereby" you're after, then I agree it's almost
a solution, as follows:

   dei minde le nu do klama

or

  dei nu mi minde le nu do klama

I say "almost a solution", because technically the utterances are
assertions, subject to truth-conditions rather than satisfaction-
conditions.

Jorge:
> > The exact nature of the
> > difference is still a matter of debate, but it is clear that there
> > is a difference (I have my own analysis, but shan't inflict it upon
> > my fellow list-members). The difference is more than a matter of brevity.
> > It is similar to the difference between "me" and "the person speaking".
> Well, why not inflict it upon us? Maybe it's relevant to Lojban.

In brief, the meaning of all word-tokens with a deictic element of
meaning is defined with respect to the word token itself. E.g. the
referent of a token of {mi} is whoever has said the {mi}-token. A
token of an imperative verb is grammatically specified as being a
command by the speaker of the token, to the addressee of the token,
to perform an action that is an instance of the sense of the token.
In contrast, a token of a declarative verb is grammatically specified
merely as an assertion: the truth value of "I command you to go" is
not necessarily contingent on whether the utterance itself is a
command by the speaker, to the addressee.

> I just don't think there needs to be a simple
> way to say "for a certain card, pick that card". If you want to say
> it with imperatives instead of using {minde}, it can be done:
> {ko'a goi da poi karda  i ko cuxna ko'a}. It is such an unusual
> request, that the xo'u form is not really needed.

Okay. I accept that the case for xohu is reduced to a matter of
convenience. Roughly speaking, it can be used to override default
scope in rule-governed ways, which is useful either for brevity
or as an afterthought (in preference to deleting what one has already
said and starting over again); all methods of avoiding xohu involve
forethought (e.g. & mainly, sticking the variable it would attach to
into the prenex of a higher bridi than the bridi the variable is a
sumti in).

> > I happen to work in a framework where the exceptionality
> > of a rule is costless (i.e. the rule itself adds complexity to the
> > grammar, but in most cases its status as an exception imposes no cost).
> > This view seems to me to suit natural language, which is riddled with
> > exceptions to virtually everything.
> That would seem to allow for a lot of rules. I would think the less rules,
> the simpler, but as you say, who can say for sure what simplicity is?

I agree that simplicity corresponds more to the number of rules than
to the exceptionlessness of rules.

> > I wasn't claiming that {xohu da} suffices for {makau} as indirect
> > interrogatives in a duhu-clause. I have previously offered "She knows
> > the identity of the set of colours of the house" as a way of
> > rendering "She knows what colour the house is".
> Yes, and I disagreed with you, because an identity is not a fact, and
> one knows (djuno) facts.

I'm not persuaded I agree with you, but anyway, here's a remedy:

  Ex, she knows that x is the identity of the set of the colours of the
    house.

> I would agree that you can say "she knows something (some fact) about
> the identity of...", or that "she knows what is the identity of...",
> but the first is ambiguous,

Yes it's hopelessly vague.

> and for the second you fall into makau again.

See my remedy above.

> > >        ko'a se pluka le nu le zdani cu skari makau
> > > is more general, because it includes the possibility for example that
> > > {le zdani cu skari noda}, or {le zdani cu skari reda}. The speaker
> > > doesn't know the answer to {le zdani cu skari ma}, but does know
> > > that whatever is the answer, ko'a likes that fact.
> > "She likes the identity of the set of colours of the house" would
> > allow for these possibilities.
> "She likes what the identity of the set of... is", not the identity in
> itself. What's likable about an identity?

Again, see my remedy.
But anyway, the identity of the set is the sum of those properties
that differentiate the set from all others. Why can one not like
this sum of properties?

> (And what was your Lojban rendering of "the identity"? I don't
> think {le ka du} works, but I don't remember if it was that.)

How to render "the identity" is a separate problem. I did suggest
{lo ka du}. Since Lojbab's pronouncement on {ka} I'm now rather
short of intuitions about what it means, so I can't judge what
{lo ka du} means.

> > I think I'm clear now about {makau},
> > and don't think it is necessary (i.e. we could manage without it),
> > but as you're so fond of it I wouldn't want to argue for abandoning
> > it.
> I suppose we could do without it, but at the cost of a lot of verborrhea.
> How would you say {ko'a cusku le sedu'u xokau prenu ba klama} = "She said
> how many people are coming."

  Ex, she said that x is the cardinality of the set of people that are
   coming.

> > > > However, let's go back to makau in the frica example, since
> > > > I have at long last understood adequately (I hope) what you intend.
> > > >   koha kohe frica le ka keha se skari makau
> > > >   Koha and kohe differ in terms of the colour they are.
> > > >   Koha and kohe differ by virtue of the fact that they are the colour
> > > >     they are.
> > > >   ro da, da is member of {koha, kohe}, ro de, da skari de:
> > > >      koha kohe frica le nu da skari de
> > > > There's an ugly repetition of {skari}, but it is sayable.
> > > There you are essentially saying two sentences:
> > >        da zo'u ko'a ko'e frica le nu ko'a skari da
> > >        ije de zo'u ko'a ko'e frica le nu ko'e skari de
> > Don't you think that your makau version is in a sense saying two
> > sentences as well?
> No, that is exactly what {ke'a} (or whatever is the lambda variable)
> saves you from. It leaves the property as a function with its
> argument place empty, so that the difference is not that function
> evaluated at any particular argument, but the function itself.
> The reason that that function is the difference is that it evaluates
> to different things, but those values are not the difference, the
> difference is the function.

Okay: that's an objection to my avoidance of keha. But do we need
makau? Say koha is red and kohe is blue. Then

  ro da, da is member of {koha, kohe}, ro de, da skari de:
     koha kohe frica le ka keha skari de

says that being red and being blue are differences between koha and
kohe. I don't see what your objections to this are going to be, but
I'm sure there will be some. Look at this as practice for explaining
to the community at large.

> But ke'a is not equal to either ko'a or ko'e. ke'a is a place keeper,
> for a place that remains empty in this sentence.

Can we say {lo ka keha mamta keha} to mean "the mother relation",
"the function from mothers to offspring"? If we can, I start to
see a strong case for it.

----
And