Message-Id: <m0sRyyA-0000YkC@xiron.pc.helsinki.fi>
Date:         Sat, 1 Jul 1995 05:18:41 -0400
Reply-To:     Logical Language Group <lojbab@ACCESS.DIGEX.NET>
Sender:       Lojban list <LOJBAN@CUVMB.BITNET>
From:         Logical Language Group <lojbab@ACCESS.DIGEX.NET>
Subject:      ci stedu, and a new dumb idea on quantifiers for people to tear up
To:         lojban@cuvmb.cc.columbia.edu
Content-Length: 5871
Lines: 140

>> Compare an English mass noun with the same
>> structure sentence:  "Water weighs 1 kg .or 1000 kg or 1x10e10 kg" seems
>> false because we don't know how much water is being talked about.  Only if we
>> put the English in a form that suggests the preselcted prenexed value for
>> loi djacu does it work: "There is some amount of water that weighs 1kg or
>> 1000 kg or 1x10e10 kg".
>
>We know that English is very ambiguous in these types of quantification.
>I thought Lojban was good at precisely this.  It doesn't seem right to
>change the rules at every step to fit this or that English expression.

I agree.  That is why I think I like using sumti number order for scope.

>> I don't think the English can be expanded into 3 sentences with different
>> amounts of water, so perhaps it is best to say that we have an ill-defined
>> situation about what happens to "loi remna", or "lo remna" for that matter,
>> when it has to be split over 3 sentences in expanding another sumti.
>
>What??!!  Are you saying that we should abandon any attempt to make
>Lojban logical just because English does not follow predicate calculus?

No.  I am saying that it wasn't intended that implicit quantifiers have
differing scopes based on position.  I always viewed use of "da" as
being an explicit attempt to be unambiguous as to scope.  I have not to
my knowledge ever used "lo", "loi" or "le" in ways where implicit
quantifiers caused logical trouble.  This suggests to me that when we
use implicit quantifiers, we are avoiding logical precision.  Because of
my old biases about "lo" in addition, I have presumed that its logical
scope was somewhat ambiguous.  "lo" has been for me a more natlang-like
usage than "da poi" and may thus lead to logical imprecision.

loi mass statements have always been to me so inherently logically weak
in their claim, that logical precision seemed superfluous.

The assumption that "lo broda" may contain some number of unspecified
restrictions (a more recent development in the language, but still
predating the last year's discussion) has led me to similarly consider
that "lo" usages are inherently logically imprecise.  Only "dapoi"
usages (for which I perhaps unlike others presume will be COMPLETE as to
restriction within some universe of discourse) have I considered to be
sufficiently logically precise that quantifier scope mattered.

Lojban allows, but does not necessarily require, logical precision.  Or
perhaps better, it has several levels of logical precision, with "lo"
and "loi" somewhat intermediate.

>> We would NOT say that it is true that
>> "lo remna cu se stedu ci stedu" intending this to expand into 3 sentences
>> that each is true: "lo remna cu se stedu pa stedu .ije lo remna cu se stedu
>> pa stedu ije lo remna cu se stedu pa stedu."
>
>No, in that sentence, one human is claimed to have three heads.  On the
>other hand:  {ci stedu cu stedu lo remna} claims that three heads are
>head of at least one human, not nec. the same human for each of the
>heads.  I thought all that was settled long ago.

lo remna != pa remna != ro remna

The first sentence does not necessarily say to me that 1 human has 3
heads, but that some number of humans (at least 1) has 3 heads.

pa remna cu se stedu ci stedu
There is (exactly?) one human with 3 heads

(or using sumti number order:  there are some particular 3 heads that
pertain to exactly one human - which I believe a weaker statement that
says nothing about any other heads that might pertain to some other
human).

These kinds of statements are not quite as constrained as the
traditional voda poi goat's legs.  "mi penmi ze simxu mensi" says that I
meet some 7 sisters in a particular (underspecified) event, and not that
in all my life there are only 7 sisters that I meet.

ro remna cu se stedu pa stedu

seems to mean either that every human has the same head, or that every
human has one head, depending on whether quantifier order is based on
sumti order or not.

ni'o
Let me toss a new stupid idea into the ring for people to tear up.  I'm
sure >I< am not thinking of the logical implications too clearly.

Another possibility might be that indefinite descriptions, and maybe
even quantified lo descriptions, need to be treated as singular
collectives for scope purpose, even when the number is plural.  This
makes indefinites more like masses ("ci broda" = "piro lei ci da poi
broda" XOR maybe "pa lo broda cimei" - not sure what the difference
would be)

Given this
ci remna cu citka mu zumri
and
mu zumri cu se citka ci remna

both say that there are 5 ears of maize in this event collectively
downed by 3 men.

To get the alternate quantification, you need to say
ro lo ci lo remna cu citka mu zumri
mu zumri cu se citka ro lo ci lo remna
Each of the 3 men separately downed 5 ears of corn.

with the possible abbreviation of "ro lo ci lo" as "roci" still an
undecided issue.


ro lo ci lo remna cu citka ro lo mu lo zumri
means that each of the 5 ears of corn are eaten by each of the 5 men (yecch!)
and the conversion would mean the same.

(With "ro lo mu lo zumri" and "ci remna" in either order, you get that
each of 5 ears of corn is separately downed by some 3 men.

"le" unlike "lo" already has that "ro" outer quantifier making the scope
issues more clear.

ci lo broda might be defined the same as ci broda, implying a collective.
Or it might have a different definition to be argued out.

Advantages of all this:

1) It reduces all explicit quantifiers on lo/indefinites to either
"pa/piro" or "ro" as the outermost quantifier, which I think removes the
scope/order question.

2) I think it follows the goat leg convention.

Still unanalyzed:

how su'o/non-specific outer quantifiers work under this convention

e.g. lo remna cu citka mu zumri
     lo remna cu se stedu ci stedu
     lo remna cu se stedu lo stedu

and their conversions.

lojbab