Re: misc responses to And #2

[ucleaar <ucleaar@UCL.AC.UK>]

Lojbab:
> > >> Presumably most recursion and repetition in natlangs has limits,
> >Not in any interesting sense.
> >> for every such grammatical structure you can probably generate some
> >> unacceptable nonsense.
> >I've not usually found this to happen. And if it does happen, then it
> >means there's something wrong with the grammar.
> I again cite center embedding as a well-known example both of the limits
> of recursion and of formal grammatical structures.

That's no good. I've already told you that's a special case. Plus,
centre embedding does not result in nonsense, though it does result
in unacceptability.

> >Most of the problems I'm aware of are due to dissimilar lexemes
> >belonging to the same selmao.  For example, {zio} is in KOhA, but not
> >every KOhA can meaningfully be replaced by a {zio}.
> I'm not sure why not. zi'o represents elimination of a place.  I assume
> that if you eliminated ALL the places you would have nothing meaningful
> left in a predicate, but that is a philosophical issue, not a linguistic
> one.

Not every sumti is a sumti of an explicit selbri/bridi. What does
{le zio cukta} mean? Or {le cukta nou zio} or {le cukta goi zio}?

> Besides, zi'o is in the language only because certain other people
> insisted on it.  Hummpf.

Yes, well I was one of them, but I certainly didn't insist on it being
in KOhA.

> >> 1. Jabberwocky
> >Jabberwocky is either ungrammatical or has defined semantics.
> I guess it is ungrammatical then, assuming that "defined semantics"
> means anything.  The moment you put a nonsense word, even of firmly
> defined syntax, into a language, and do not define its semantics, you
> cannot determine the semantics of the whole.

That's one way of looking at it. But what tends to happen in the case
of Jabberwocky at least is that the "nonsense" word (= novel, unfamiliar
word) is given a conjectural defined-semantics.

> >> 2. Curious green ideas sleep furiously.
> >Is perfectly grammatical, and has defined and agreed on semantics.
> It has hypothetical semantics based on some metaphorical interpretations
> of the individual terms.

No. Metaphor only enters pragmatically. The semantics is, very very very
roughly, "Ex ideas(x), green(x), curious/colourless(x), sleep(x),
furious(x)". For the meanings of the predicates consult a good dictionary
(okay, there aren't any, but imagine there were).

> I can probably come up with similar metaphorical interpretations for
> the most nonsensical Lojban, but that does not make the semantics
> "defined and agreed upon", because you might come up with a different
> metaphorical interpretation than I do.

{li ci re boi} means "thirty-two" and not "twenty-three". We know that
because of a rule of semantics. There are no such rules that tell you
how to interpret {le zio cukta}. Indeed you may be able to creatively
succeed in communicating something useful with {le zio cukta}, but you
wouldn't be using rules of lojban to do so.

> >> 3. The US tax code
> >I haven't read it, but I expect it is perfectly grammatical, and has
> >defined and relatively agreed on semantics.
> Been a lot of court cases to prove you wrong.

Not at all. First, the majority of cases will actually be about pragmatic
rather than semantic issues. Second, I said the semantics is relatively
agreed on, not utterly agreed on. Third, just because the semantics is
defined (and even agreed on) doesn't mean it is always sufficiently
defined for legal purposes.

> >I should have downplayed your "agreed-on", because this is actually
> >irrelevant to the definition of grammaticality, although it is of
> >course vital to successful communication.
> If it cannot be agreed upon, then you have to treat the individual
> interpretations as idiolects and not as a language.  Every idiolect has
> some variations from the standard grammar of the language.  S presumably
> the standard grammar MUST be that portion of the grammar that is agreed
> upon by most or all speakers.  Or you can define it to include the
> combined grammar of ALL idiolects, in which case it is hard to say what
> is ungrammatical without checking with all speakers to make sure none
> has an interpretation.

The relationship between an idiolect and a language is like the
relationship between a creature and a species. Languages and species
are cluster patterns of similarities among idiolects and creatures.
If you're going to do linguistic or biological dissection, you must
operate on idiolects and creatures.

> And I guess I left out one other case "Time flies like an arrow" and
> other such demonstrably ambiguous sentences do not have a defined
> SYNTAX, much less semantics, because you need to kow the context to
> determine which semantics holds.

All we have here is several different sentences that are very similar
in certain respects, chiefly in their sound. The syntax and semantics
of each sentence is perfectly clearly defined.

> >Singularities could be selbri, but that would be pointless.
> I cite munje, and the monotheistic version of jegvo.

If munje really is a singularity, which I doubt, as, given its place
structure, should you, then it should be a sumti. {jegvo}, I presume
you have mistakenly thought to denote only a deity. If it did denote
only Jehovah, then it should be a sumti.

> And at one time you argued that all names should be selbri if I recall.

I argued that many non-lojban names should be rendered in lojban by
fuivla. I still have the same view. This is partly because in Lojban
cmevla are allowed to be ambiguous, and aren't standardized, so if you
want an unambiguous standardized designation you have to turn elsewhere.
Further, I would argue that the denotation of, say, "London" is not that
qualitatively different from that of "water"/"djacu", by the criterion
of whether they warrant selbrihood or sumtihood.

> >> >My objection to {nu} is that really the event is an argument of the
> >> >bridi, so {jai fau broda} is truer to the meaning.
> >> If the event is an argument of the bridi that is reflected in the event
> >> I see infinite recursion.
> >I don't understand.
> If the "1-place" bridi "da broda" has an implied x2 event place which we
> night represent as "lenu da broda" for convenience, then that implied
> form of the bridi is "da broda lenu [da broda]", and that in turn has
> the implied form "da broda lenu (da broda lenu [da broda])" ad nauseum.

True. So although you might represent it as "lenu da broda" for
convenience, you would be unwise to assume that logical consequences
of your choice of notation necessarily tell you about logical
consequences of what your notation is intended to represent.

> >> >> (As an aside, I don't think that something like {mi nelci le ka
> >> >> do melbi} makes any sense. It should be {mi nelci le nu do melbi}.
> >> >I agree.
> >> Then "nu" is useful?
> >No, I meant the x2 of {nelci} shd be a situation rather than a property.
> >I didn't mean that to refer to a situation we shd use {nu}.
> So you are defining a new abstraction which might or might not be
> equivalent to what we call "nu" and refusing to label it "nu".

I'm not defining any abstraction. I am saying that just as the x1
of nelci refers to a sentient consciousness, so the x2 can sensically
refer to a situation, but not to a property (at least with the intended
meaning in Jorge's example). Situations are indeed what we refer to
by {nu}, but I reject the use of {nu} because its existence leads to
a to me obnoxious mismatch between syntax and semantics.

> >> The fact that the number of places in gismu that are abstractions
> >> alone shows how useful it is to recognize multiple levels of
> >> abstraction in analyzing semantics.
> >I don't understand. Can you rephrase, please?
> We use lots of lenu clauses in actual Lojban, therefore they are useful.
> Argument by pragmatics (as are most Lojban decisions).

.oicai!!! I have at no point disputed that it is useful to be able to
refer to situations. I just contend that {nu} is an ugly way to do
so.

> >> I like "nu" itself simply for the the capability of bundling up an
> >> occurance of a bridi relationship as a semantic unit,
> >Yes, it is commonly misused that way.
> Since fasnu - the defining predicate of "nu", refers to "occurance",
> better to say that it is DEFINED that way.

Sorry: I interpreted you as having said "bundling up a bridi relationship
as a semantic unit". That's what I meant is a misuse.

But how to understand "occurrence of a bridi relationship"? That makes
no sense to me. Bridi relationships "obtain" (i.e. "are true") rather
than "occur". "Occur" means "to be a situation". I have no problems
at with the semantics of {fasnu}, or with the idea that the denotation
of {fasnu} and {nu} are the same. I merely dispute that {lo bridi
cu fasnu} is ever true.

> >> and the subcategories allow me even to define a substructural way of
> >> looking at that occurance.
> >{lo pruce jai fau broda} would do just as well, and would have a
> >proper syntax-semantics match.
> The use of jai IMPLIES the existence of an abstraction by
> transformation.  Provide the explicit transform, please, without using
> any abstractors.

I don't understand. Please try again.

> >There is a fair bit that is mind-expanding in Lojban, but those 4
> >somewhat outdated aktionsart distinctions don't expand my mind. The
> >more basic properties (dynamicity, extent/duration and telicity, which
> >itself is a temporal manifestation of bounding in general) that underlie
> >this taxonomy are most certainly reflected extensively in natlangs.
> You would have to clarify - maybe we need words for them %^)

As and when they're needed. I don't really want to launch into a
disquisition on these matters unless I have to.

> Seriously, if they are reflected clearly in natual language, you should
> be able to give me, say for some common English sentences, examples of
> varying each of the 3 parameters and how they are reflected in changes
> to the sentence.

I don't think any construction necessarily expresses something nondynamic.
Progressives must denote things that are dynamic. Duration can be tested
by acceptability of durative _for_. Telicity in verbs is harder to test
- if you can say things like "in 3 hours" then it's telic.

> >> >> {za'i}, {zu'a}, {mu'e} and {pu'u} are all kinds of {nu}.
> >> >Someone had been reading linguistics textbooks when these were
> >> >added, and gave them too much weight.
> >> Actually, I hadn't read ANY linguistics textbooks when I added them.
> >I'm skeptical that you dreamt them up purely by thinking about situation
> >types.
> pc was explaining "nu" on the phone one day, and got into these types of
> "events", and I came back a couple days later saying I thought it might
> be useful to embed the distinctions in the language, "nu" already having
> proven so useful, leaving "nu" as the non-specific general case.  So pc
> read the textbooks, and I dreamed up the Lojban application %^)

It is the application that I've been objecting to.

> [Actually, I just got inspired to look in the archives, and it was in a
> letter to me from pc, not a phone call, which I annotated.  So I have
> documentation of EXACTLY when and how those came into the language.  See
> a separate post forthcoming.]

My comments are in a separate post.

> >> But the event/achievement/state and processes of running a race are
> >> all to my mind quite different concepts.
> >Absolutely they are. Running a race, as I understand the concept, is
> >a process. So to describe it as an activity, achievement or state is
> >to create a new category, different from normal race running.
> No, it is just a different way of looking at the same race.

Well of course in principle that is possible. You can construe a race
running as a herd of purple elephants too, if you so wish.

Less expostulatorily, I agree that a particular instance of race
running may be conceptualized as something other than a process,
but, to repeat, you are thereby creating a new category, of
atypical race runnings.

Consider a tennis ball. You can conceptualize it as a mere aggregation
of matter, its shape being an accidental irrelevance, or you can
conceptualize it as a sphere, with its shape being an integral property
that makes it eligible for spherehood. But the definition of the
category of balls includes the requirement of approximate sphericality,
so when you're conceptualizing some ball as a mere splodge of matter,
you're not conceptualizing it as a ball.

>> I'd like
>> to see some explicit examples of meaningless statements using PA.
>The other day John offered {suo ro suo}. Let that do for starters.
> "At least all of the at least 1". Makes sense to me; just sounds either
> useless or redundantly lawyerish (which may mean the same thing).
> Now whether John agrees with me about these semantics is irrelevant
> (though he should have recognized that I would come up with that),
> just as you claim I don't have to agree with your semantics for
> "...green ideas..."

You don't have to agree with my semantics as long as you have different
rules. If you have no rules, though, then this is what I'm saying
is bad.
I guess there's a rule saying suo+PA means "at least PA", so {suo ro}
is "at least all". I'm not familiar with the rule that says ro+PA
means "all of the PA", but I'll believe there is one if you say so.

How about {ro pa fiu ro sohi fiu mahu mahu mehi mehi daa rae}?

Whichever answer you give, I'll be satisifed. If you can tell me
the interpretation rules then my objection is answered, and if
you can't then at least I've got you to see that the phenomenon
of rulelessness exists.

---
And