Message-Id: <199512212152.XAA03285@xiron.pc.helsinki.fi>
Date:         Thu, 21 Dec 1995 18:37:55 +0000
Reply-To: ucleaar <ucleaar@UCL.AC.UK>
Sender: Lojban list <LOJBAN@CUVMB.BITNET>
From: ucleaar <ucleaar@UCL.AC.UK>
Subject:      Re: TECH: {loi} & {loe}
To: Veijo Vilva <veion@XIRON.PC.HELSINKI.FI>
Content-Length: 13685
Lines: 294

John:
> la .and. joi la xorxes. joi la pycyn. cusku be di'e casnu
> > 1. Xorxe, pc & me agree that past discussion of "masses" [i.e.
> > the meaning of {loi}/{lei}] has involved several distinguishable
> > factors:
> I grant that they are distinguishable, but are they necessarily to be
> distinguished?

Well if they are to be distinguished, we at least agree on which has
best claim to {loi}. As for whether they should be distinguished, we
can begin a debate on that. But I hope this will feed into a refgrammar
paper.

> For that matter, I believe the distinction between "collective"
> and "porridge" to be a mirror of the (hazy) English distinction
> between mass nouns and count nouns (unfortunate in this context,
> but firmly fixed in our terminology).

Maybe it would help if we changed our terminology, e.g. by lojbanizing
it.

> When the selbri of a "loi" represents an English mass noun, we speak of
> "porridge", whereas when the selbri represents an English count noun,
> we speak of "collective", it seems to me.

Maybe; I'd have to think about it. But at any rate, if you're right
it confirms my point, which is that there are genuine semantic
distinctions here. You will be wanting to argue that there are
important commonalities.

If you take any common noun, you can make it plural and it will then
be interpretable collectively or distributively, and you can make it
singular mass or singular countable. [Incidentally, I notice now that
"cattle" is not a mass noun, since it takes plural agreement.] So
for any lexeme whose instances can be common nouns you get a contrast
between (among other things) collective and mass interpretations.

> BTW, I take "Napoleon" and the like to be mass nouns; I don't know
> if this is standard, but they fit the profile otherwise.

My impression is that this is very non-standard, but I agree with you.
If there is a distinction between proper nouns and common nouns it is
principally phono-morphological rather than semantic or syntactic:
proper nouns are unconstrainedly homonymous; the dictionary ought to
have a zillion entries for JOHN/1 ... JOHN/99999999.

But Lojban is different.

> > 2. Xorxe, pc & me agree that {loi}/{lei} (&, I presume, {joi} & other
> > stuff labelled "mass") should be a COLLECTIVIZER.
> > pc:
> > > The collective sense seems to be the one we get the most use out of,
> > > so we should probably tie it to _loi_ and its analogs.
> > Xorxes to me:
> > > > As for masses, I don't want to debate them all over again until there is
> > > > an official refgrammar treatment of them. Until that exists, I will
> > > > continue with what is my current belief - that we either don't know
> > > > or disagree about what "masses" are.
> > > I think I've already formed an idea of what they are: the collective
> > > plural. If they are something else, then I would like to know how to
> > > do the collective plural, which is something extremely necessary given
> > >  that with le/lo you can only get the distributive one.
> I grant that one of the functions of a mass is to represent the
> collective plural (of count nouns, that is). But what is the function
> of "loi" applied to a mass noun?

There are no "mass nouns" in Lojban - every brivla is countable (or
at least every gismu place is, or at least every gismu x1 is).

> Lojban selbri represent mass nouns under the cover of "x1 is a quantity
> of ...".

In other words, nouns that are usually mass in English are made countable
in Lojban. I noted this in the message you're replying to.

> Therefore, "lo djacu" represents one or more water-quantities
> (of indeterminate but definite size);

Yes...

> "loi djacu" represents a collective plural of these water-quantities,

yes...

> which in fact is a porridgification of water.

No.
Or rather, not necessarily. Suppose it is the argument of
"is sufficient to make the pitch unusable". This is (in the context
I conjure up) true of the collectivity of le djacu, but not of
each djacu distributively. None of this means that I have to
conceptually eradicate the boundaries that distinguish one djacu
from another. As proof, notice that in English we can say
"here are cheeses weighing 10 kilos together" - which is distinct from
"here is cheese weighing 10 kilos"; {loi} gives the former, not the latter.

> > 3. I think we remain uncertain as to whether the {pi} fractionators
> > before {loi} will now make any sense. In my view they don't.
> > Xorxes:
> > > There was and maybe still is disagreement as to the default quantifier
> > > for {loi}. Is {loi broda} "all the broda there are, collectively", or
> > > is it "some broda, collectively". I think that the second one is the
> > > more useful and the more consistent with the other defaults. In any
> > > case both can be explicited: {piro loi broda} and {pisu'o loi broda}.
> It has always been true that "loi broda" means "a-portion-
> of(massification-of (all-of(brodas)))", hence "some broda taken
> collectively" is one interpretation, but not the only one.

I never liked that. Other singulars don't get that "a portion of"
interpretation. I'd have preferred "a porridgification of some/all/n
brodas". But now I prefer "a collectivity of some/all/n brodas".

> "mi joi do" normally means "you and I (collectively)", but given the
> right interpretive context might mean "my hand plus yours, exclusive of
> the rest of our bodies".  This was the example at which {la xorxes.}
> originally jibbed, as I recall.

I don't know what point this is making. {mi e do} can mean "my hand
and yours, exclusive of the rest of our bodies" just as much as with
{joi}. This is a matter of when my hand can and can't "count" for me,
and doesn't seem to be bound up with mass/gadri issues.

> > 5. The other two erstwhile now-rejected candidate meanings for {loi}
> > have previously been labelled respectively Myopic Singularizer, which
> > is pc's "species"/Mr. Rabbit, and Porridgifier, which is [I think] pc's
> > "shiftingly bounded continuities" [which soulds like a cowanism],
> Jeeg and Talen and Zipf and R forbid!  I would never coin a phrase like
> that, at least not intentionally.
> > i.e. true masses.
> >   To understand these, we must be aware of the distinction between
> > (i) multiply-instantiated/many-membered categories, such as the
> > category of cats, and (ii) what can be seen either as individuals or
> > as single-membered categories, e.g. Napoleon, London, (the mass of all)
> > wine. Lojban is fine on (i), but not fine on (ii). I will try to make
> > sense of the current situation.
> This looks like count nouns and mass nouns again.

Yes.

> >   (a) All selbri must be defined so as to conform to (i).
> Agreed.  We do not talk of "x1 is wine", but "x1 is a quantity of wine".

Good. (a) might be a bit sweeping, but as I said above, it is true
of all x1s of gismu at least.

> But with "x1 is Napoleon", {ko'e cu me la napoleon.}, we don't have a
> problem, because we can take the one-membered (or however-many-membered,
> in cases like "x1 is one of the Three Kings") in the categorical
> interpretation.

I still don't understand new {me} well enough to agree or disagree.
But at any rate, if we adopted a fuivla for Napoleon, e.g. prenrnapoleone,
it would have to be defined as, e.g. "x1 is a quantity of Napoleon
Bonaparte" or somesuch, and the ridiculouser that sounds, the less fitting
it is to have a brivla for Napoleon. [I wish we could have fuivla with
the grammar of sumti, so we don't have to use cmene.]

> >   (b) By processes of myopic singularization and of porridgification,
> >       (ii)-type meanings derive from (i)-type meanings.
> Just so, at least with regard to porridgification. I never understood
> myopic singulars: when the mythical Trobriander says "Ah, Mr. Rabbit
> again", I understand him to be saying "Ah, another outlier of
> Rabbit-Porridge ({loi ractu})."  Kind of like wandering the world and
> occasionally spotting another outcrop of the Midgard Serpent.

For me, that is the porridgification of all ractu.
At any rate, I understand exactly what you mean and it's not what
I mean by myopic singularization. Suppose on Monday you see Flopsy
and on Tuesday you see Mopsy. How many rabbits did you see? Two.
Now suppose you saw Cottontail on Monday and Cottontail again on
Tuesday. How many rabbits did you see? One. Now suppose that
on Monday you see a rabbit and on Tuesday you see a rabbit. How
many rabbits did you see? Well, to answer you have to find out
whether it was the same rabbit. Myopic singularization just assumes
it was the same rabbit - it says "as far as I can (be bothered to)
tell, there is just the one rabbit".

> >   (c) If a (ii)-type meaning can't be derived from a (i)-type, we
> >       must use a cmene to refer to it.
> Such as?

La napoleon.

> >   (d) Meanings that are naturally (ii)-type, but which we wish to be
> >       denoted by a selbri, must be altered to give them a (i)-type
> >       meaning.
> Yes, except that Lojban (as I read it) denies that there are any
> "natural" (ii)-type meanings: all that is the result of Whorfian
> Mind-Lock (TM).

(a) Some natural (ii)-types: la Napoleon, la djan kau,n.
(b) Lojban would be wrong to deny that there are natural (ii)-type
meanings, but it may well be making life easier for itself by insisting
on brivla sumti having (i)-type, natural or not.

> The natural-to-Lojban mode allows for both individual and
> collective/porridge reference, but makes the distinction uniformly
> across all selbri, without calling some selbri "naturally mass" and
> others "naturally count".

Yes. This is what I was saying.

> (This rule is not always and everywhere satisfied: some selbri
> subcategorize for masses.)

I have gone through the giuste & cannot really see this. Certainly
nothing subcats for only loi/lei. Of things that are said to require
a mass, I see that the sumti is a collectivity, but not that it
is a porridge.

> >  Porridgification involves taking a number of individuals and erasing
> > their boundaries. Thus, e.g., a heap of cheeses can be seen as cheeses
> > or as cheese. A load of cows can be seen as cows or as cattle.
> Just so.  This is porridgification, but it is also collectivity.
> Indeed, I would say that collectivity is a special case of
> porridgification.

In a collectivity the whole has various properties independent of its
parts; it has a certain autonomy from them. And crucially, it has
discernible parts. A porridge has no discernible parts (though it
can have ingredients); and it is not autonomous from whatever
constitutes it. (Most individuals are porridges.)

I find it hard to see one as a special case of the other. But, on
the other hand, I have yet to build up a load of examples that make
sense with one but not the other. (Maybe casnu has to be done by
a collectivity rather than by a porridge.)

> >  Myopic singularization involves identifying every member of the
> > category with every other member, i.e. failing to recognize the
> > differences between them.
> I don't see this as corresponding to mass behavior at all, vide supra.

You said you don't understand it.

> Certainly it doesn't match my concept of a species (jutsi), which is
> an individual, not a class or set or collective.  >Homo sapiens< is
> an individual, ontologically on a par with other individuals; its
> components are various (biological) individuals alive or dead.

Classes, sets and collectivities are individuals. (As far as I can see.)
I'd have thought a species is a class. I can't imagine what's your
concept.

> > the definition of {djacu} must contain criteria for distinguishing
> > between {pa djacu} and {re djacu}. I would propose that {pa djacu} is
> > continuous and entirely surrounded by non-djacu. The important point
> > is that there must be such criteria built into the definition of
> > {djacu}.
> Why so? In particular cases, we may have criteria for distinguishing
> them. But I have no trouble with saying that there are {recigeixa djacu}
> in this bucket (1 mole, or 10 g if I remember the rules correctly).

Part of the definition of the selbri is the criteria for distinguishing
one broda from another when counting them. We use these criteria when
deciding whether {mi viska re djacu} is true.

> > .. pc's loe cikagoan, which is the average chicagoan. So pc-wise,
> > your examples mean "the average apple is wanted by
> > me" and "the average cow-killing lion will be hunted by me".
> Yes, except that "average" refers here to an abstraction, not the
> concrete object which most nearly resembeles that abstraction (aka
> "the most typical Yale man").

That's what I meant - the abstraction.

> >  X is right that "he is a house builder" does not entail that
> > there is a house he has built.
> First of all, I have trouble with this statement if extended tenselessly.
> I "He is a house builder" really compatible with "He never has and never
> will build any house whatsoever?"

Yes. But, for pragmatic reasons, not usually.

> Second, note that tenseless Lojban bridi (unlike their English
> translations) are potentially any of "caa", "ka'e", "nu'i", or "pu'e".

I am so appalled by this rule that I haven't yet been able to bring
myself to face up to it.

Few concepts are as ill-understood as these. (If you think logic or
linguistics has a ready account of them, please point me to it.) Thus
they shouldn't even have cmavo status, and to have the default
unspecified is just a nightmare. (Usage goes against it, though.)

> Is "He is a house builder" really compatible with "There does not
> [tenselessly] exist any house that he is capable of building?"

I take the 5th on that.

You tell me what capability is.

---
And