PRECISIATION OF MEANING

Lotfi A. Zadeh
Computer Science Division
Department of EECS
UC Berkeley

NTL Seminar, ICSI

April 30, 2010
URL: http://www-bisc.cs.berkeley.edu
URL: http://www.cs.berkeley.edu/~zadeh/
Email: Zadeh@eecs.berkeley.edu

Research supported in part by ONR N00014-02-1-0294, BT Grant CT1080028046,

Omron Grant, Tekes Grant, Azerbaijan Ministry of Communications and
Information Technology Grant, Azerbaijan University of Azerbaijan Republic and
the BISC Program of UC Berkeley.
1 /168 LAZ 6/9/2010
2 /168 LAZ 6/9/2010
 PRECISIATION OF MEANING—PREAMBLE
z The coming decade is likely to be a
decade of automation of everyday
reasoning and decision-making. In the
world of automated reasoning and
decision-making, computation with
information described in a natural
language is certain to play a pivotal
role.

3 /168 LAZ 6/9/2010

 CONTINUED

z Precisiation of meaning is a
prerequisite to computation with
information described in a natural
language. In turn, understanding of
meaning is a prerequisite to
precisiation of meaning.

4 /168 LAZ 6/9/2010

 CONTINUED

z Precisiation of meaning is of particular

importance in the realm of human-
machine communication (HMC).
Humans can understand imprecise
language. Machines cannot.
z As an operation, precisiation
presupposes that its operand is
imprecise. In the case of natural
languages, this is indeed the case.

5 /168
 IMPRECISION OF NATURAL LANGUAGES

z Natural languages are intrinsically

imprecise. Basically, a natural language
is a system for describing perceptions.
Perceptions are imprecise, reflecting
the bounded ability of human sensory
organs and ultimately the brain, to
resolve detail and store information.
Imprecision of perceptions is passed
on to natural languages.

6 /168 LAZ 6/9/2010

 NATURAL LANGUAGE AND PERCEPTIONS

description evocation
perception NL perceptions

z p: perception
z NL(p): description of p; semantic entity
z p+: perceptions evoked by NL(p)
z p+: meaning of p; denotation of p

7 /168 LAZ 6/9/2010

 IMPRECISION OF NATURAL LANGUAGES

zThere are many different forms of

imprecision in natural languages. A
principal source of imprecision is
unsharpness of class boundaries.
Everyday examples:
Words(phrases, predicates)
z tall z young
z near z high fever
z not very tall z beautiful
z mountain z recession
8 /168 LAZ 6/9/2010
 CONTINUED

Propositions
z Most Swedes are tall
z Icy roads are slippery
z Speed limit is 65 mph
z Check out time is 1pm

Commands
z Keep under refrigeration
z Handle with care
9 /168 LAZ 6/9/2010
 UNSHARPNESS OF CLASS
BOUNDARIES=FUZZINESS

z Fuzzy set= a class with unsharp

boundaries.

z Fuzzy word (phrase)= label of a fuzzy

set.

z Fuzzy proposition= a proposition

containing fuzzy words.

z Most words and propositions in a

natural language are fuzzy.
10 /168 LAZ 6/9/2010
 BASIC STRUCTURE OF PRECISIATION
precisiation language
p: object of precisiation p*: result of precisiation

precisiend precisiation precisiand

cointension
z precisiand = model of meaning
z extension= name-based meaning
z intension = attribute-based meaning
z cointension = qualitative measure of proximity of
meanings
= qualitative measure of proximity of
the model (precisiand) and the object
of modeling (precisiend)
11 /168 LAZ 6/9/2010
 CONTINUED
z A desideratum of precisiation is high
cointension.
z Informally, a precisiand is cointensive if
it is associated with a high cointension
in relation to the precisiend.
z Cointension principle: Achievement of
cointensive precisiation generally
requires that if the precisiend is fuzzy
so must be the precisiand.

12 /168 LAZ 6/9/2010

 UNSHARPNESS OF CLASS BOUNDARIES
z Imprecision, in the form of
unsharpness of class boundaries, is a
basic characteristic of natural
languages. What is remarkable is that
in the enormous literature of linguistics
and computational linguistics the issue
of imprecision is largely ignored. There
is a reason. In large measure, theories
of natural language are based on
bivalent logic—a logic which is
intolerant of imprecision and partiality
of truth.
13 /168 LAZ 6/9/2010
 CONTINUED

z What is widely unrecognized is that

bivalent logic is not the right logic for
dealing with imprecision of natural
languages. What is needed for this
purpose is fuzzy logic. Basically, fuzzy
logic is the logic of classes with
unsharp boundaries. In fuzzy logic, as
in natural languages, everything is or is
allowed to be a matter of degree, with
the understanding that degrees are
allowed to be fuzzy.
14 /168 LAZ 6/9/2010
 BINARIZATION

zHow is the unsharpness of class

boundaries dealt with in bivalent logic?
In bivalent logic, the standard tool is
binarization—creation of a sharp
boundary between classes.
Examples:
Tall: Height is over 6 ft
Middle age: Age between 40 and 50
Recession: Decline in GDP in two
successive quarters
15 /168 LAZ 6/9/2010
 CONTINUED

z Bear Market: 30 percent decline after 50

days, or a 13 percent decline after 145
days. (Robert Shuster)

16 /168 LAZ 6/9/2010

 A PROBLEM WITH BINARIZATION

z There is a serious problem with

binarization. When binarization is
applied to a fuzzy concept, e.g.,
recession, the precisiand is not
cointensive, that is, the meaning of the
precisiand does not fit closely the
meaning of the precisiend. In effect,
binarization violates the cointension
principle.

17 /168 LAZ 6/9/2010

 GRADUATION IN FUZZY LOGIC
zHow is the unsharpness of class
boundaries dealt with in fuzzy logic? In
fuzzy logic, the basic tool is graduation—
making class membership a matter of
degree (membership function). The
degrees are allowed to be fuzzy.
Example: μ
middle-age
1
0.8
core of middle-age
0 40 45 55 60
43
definitely not middle-age definitely definitely not middle-age
middle-age

Note: Parameters are context-dependent

18 /168 LAZ 6/9/2010
 EXAMPLE OF GRADUATION
HONDA FUZZY LOGIC TRANSMISSION
fuzzy sets Not Very Low
High
1 1 Close 1
High Low High
Low
Not Low

0 30 130 180 0 54 0 5
Speed Throttle Shift

Control Rules:
1. If (speed is low) and (shift is high) then (-3)
2. If (speed is high) and (shift is low) then (+3)
3. If (throt is low) and (speed is high) then (+3)
4. If (throt is low) and (speed is low) then (+1)

19 /168 LAZ 6/9/2010

 GRADUATION OF PERCEPTIONS

z Humans have a remarkable capability

to graduate perceptions without any
measurements or any computations.
More specifically, assume that I am
given an object, a, and a class, A, and
am asked to put a mark on a scale from
0 to 1 indicating my perception of the
degree to which a fits A. Generally, I
would have no difficulty in doing this.

20 /168 LAZ 6/9/2010

 CONTINUED

z This is what I do when I am asked to

rate a restaurant on the scale from 0 to
10.

21 /168 LAZ 6/9/2010

 CONTINUED

z If I am not sure what the degree is, I am

allowed to put a fuzzy mark, using a
spray pen with an adjustable spray
pattern.
z A fuzzy mark reflects imprecision of my
perception.
z In most cases, a crisp mark should be
interpreted as the centroid of a fuzzy
mark.
22 /168 LAZ 6/9/2010
 CONTINUED

z For example, if I am asked to estimate

the probability that Obama will be able
to solve the financial crisis, and I put a
crisp mark at .7, the crisp mark should
be interpreted as the centroid of my
fuzzy perception of the probability that
Obama will be able to solve the
financial crisis. What this points to is
that more often than not fuzzy real-
world probabilities are treated as if they
were precise. LAZ 6/9/2010
23 /168
 FUZZY DATA ENTRY & RETRIEVAL VIA Z-
MOUSE
z A Z-mouse is an electronic
implementation of a spray pen. The Z-
mouse is a convenient means of fuzzy
data entry and retrieval. In a Z-mouse,
the curser is a round fuzzy mark called
a globule. The color of the mark is a
matter of choice. A dot identifies the
centroid of the mark. The cross-section
of a fuzzy mark is a trapezoidal fuzzy
set with adjustable parameters.
24 /168 LAZ 6/9/2010
 Z-MOUSE—AN EXAMPLE OF APPLICATION

z Consider the following problem.

z Question: What is the probability that

Robert is home at 6:15pm?
z Information set—information in which
the answer is to be inferred:
p1: Usually Robert leaves office at
about 5pm.
p2: Usually it takes Robert about an
hour to get home from work.
25 /168 LAZ 6/9/2010
 USE OF Z-MOUSE

Time of Travel Usually Probability

departure time 1 1
*.8
*1 *.7

*5 *.4
globule
0 0 0 0

Information set answer

*a means approximately a
26 /168 LAZ 6/9/2010
 CONTINUED

z Note. A Z-mouse serves primarily as a

means of fuzzy data entry and retrieval.
Computation of an answer to a
question is carried out through the use
of the machinery of Computing with
Words (CW or CWW).

27 /168 LAZ 6/9/2010

 GRADUATION –EXTENSIONAL VS. INTENSIONAL

z Imprecisiation of meaning, it is important to

differentiate between extensional graduation
and intensional graduation.

graduation

extensional intensional

Name-based Attribute-based

28 /168 LAZ 6/9/2010

 CONTINUED

z The concepts of extension and

intension are drawn from logic. (R.
Belohlavek and V. Vychodil, Attribute
Implications in a Fuzzy Setting, Lecture
Notes in Artificial Intelligence 3874,
Springer-Verlag, 2006.) In extensional
graduation, grades of membership are
name-based.

29 /168 LAZ 6/9/2010

 CONTINUED
z In intensional graduation, grades of
membership are attribute-based. As an
illustration, consider a population of
Swedes represented as a relation:
POPULATION.SWEDES [Name]
z Let T be the fuzzy set of tall Swedes in
POPULATION. The extension of T in
POPULATION is a fuzzy set described
as the relation TALL[Name;µ], where µ
is the grade of membership of Name in
the fuzzy set TALL.
30 /168 LAZ 6/9/2010
 CONTINUED
z The intension of T in POPULATION is a
fuzzy relation described as
TALL[Name; Height; µ], where µ is the
grade of membership of Height in
TALL.

z Note that in the example used earlier, p:

Carol lives in a small city near San
Francisco, the relations SMALL.CITY
and NEAR.SF.CITY are extensional.
31 /168 LAZ 6/9/2010
 CONTINUED

z The extensional and intensional

graduations of T lead to extensional
and intensional meanings of T,
abbreviated as e-meaning and i-
meaning, respectively. Use of an e-
meaning of a concept is a necessity
when its attributes are not defined.
Example: Honesty. Fundamentally, the
meaning of p is its intension rather
than its extension.
32 /168 LAZ 6/9/2010
 GRADUATION OF PROPOSITIONS?

z What is meant by graduation of

propositions? If I were asked to
graduate the proposition, p: Most
Swedes are tall, what would I do? What
is the connection between graduation
of p and precisiation of p?

33 /168 LAZ 6/9/2010

 CONTINUED

z In general, a proposition, p, may be

associated with a variety of attributes.
A basic attribute is the truth-value of p,
t(p). In this perspective, graduation of p
may be related to graduation of truth-
value of p. As will be seen later,
graduation of truth-value of p is a
byproduct of precisiation of p.

34 /168 LAZ 6/9/2010

 CONTINUED

z The truth-value of p cannot be

assessed in isolation. If I were asked
what is the truth-value of p: Most
Swedes are tall, I would have to know
how most and tall are defined, and be
given the distribution of heights of
Swedes. Let us call the needed
knowledge the Information Base, IB(p).

35 /168 LAZ 6/9/2010

 CONTINUED

z The question is: How can the truth-

value be computed given the
information base, IB(p)? What is
needed for this purpose is generalized-
constraint-based semantics, GCS.
Generalized-constraint-based
semantics is rooted in test-score
semantics (Zadeh 1982, 1986.)

36 /168 LAZ 6/9/2010

 CONTINUED

z Generalized-constraint-based
semantics is a generalization of truth-
conditional and possible-world
semantics. In the following, precisiation
of propositions through the use of
generalized-constraint-based
semantics is discussed in greater
detail.

37 /168 LAZ 6/9/2010

38 /168 LAZ 6/9/2010
 INFORMAL EXPOSITION OF GCS
CLARIFICATION DIALOGUE

z The basic ideas which underlie

precisiation of meaning and, more
particularly, generalized-constraint-
based semantics, are actually quite
simple. To bring this out, it is expedient
to precede a formal exposition of GCS
with an informal narrative in the form of
a dialogue between Robert and Lotfi.

39 /168 LAZ 6/9/2010

 DIALOGUE

Robert: Lotfi, generalized-constraint-

based semantics looks complicated to
me. Can you explain in simple terms
the basic ideas which underlie GCS?

40 /168 LAZ 6/9/2010

 CONTINUED

Lotfi: I will be pleased to do so. Let us

start with an example, p: Most Swedes
are tall. p is a proposition. As a
proposition, p is a carrier of
information. Without loss of generality,
we can assume that p is a carrier of
information about a variable, X, which
is implicit in p. If I asked you what is
this variable, what would you say?

41 /168 LAZ 6/9/2010

 CONTINUED

Robert: As I see it, p tells me something

about the proportion of tall Swedes
among Swedes.
Lotfi: Right. What does p tell you about
the value of the variable?
Robert: To me, the value is not sharply
defined. I would say it is fuzzy.
Lotfi: So what is it?
Robert: It is the word “most.”
42 /168 LAZ 6/9/2010
 CONTINUED

Lotfi: You are right. So what we see is

that p may be interpreted as the
assignment of a value “most” to the
variable, X: Proportion of tall Swedes
among Swedes.

43 /168 LAZ 6/9/2010

 CONTINUED

As you can see, a basic difference

between a proposition drawn from a
natural language and a proposition
drawn from a mathematical language is
that in the latter the variable and the
value assigned to it are explicit,
whereas in the former the variable and
the assigned value are implicit.

44 /168 LAZ 6/9/2010

 CONTINUED

There is an additional difference. When

p is drawn from a natural language the
assigned value is not sharply defined—
typically it is fuzzy, as “most” is. When
p is drawn from a mathematical
language the assigned value is sharply
defined.
Robert: I get the idea. So what comes
next?

45 /168 LAZ 6/9/2010

 CONTINUED

Lotfi: There is another important point.

When p is drawn from a natural
language the value assigned to X is not
really a value of X—it is a constraint
(restriction) on the values which X is
allowed to take. This suggests an
unconventional definition of a
proposition p, drawn from a natural
language. Specifically, a proposition is
an implicit constraint on an implicit
46 /168 variable. LAZ 6/9/2010
 CONTINUED

Robert: I notice that you are talking

about generalized constraints. What is
a generalized constraint? Why do we
need generalized constraints?
Lotfi: A generalized constraint is
expressed as:
X isr R

47 /168 LAZ 6/9/2010

 CONTINUED

where X is the constrained variable, R

is the constraining relation—typically a
fuzzy set—and r is an indexical variable
which defines how R constrains X. Let
me explain why the concept of a
generalized constraint is needed in
precisiation of meaning of a
proposition drawn from a natural
language.

48 /168 LAZ 6/9/2010

 CONTINUED

Standard constraints are hard in the

sense that they have no elasticity. In a
natural language, meaning can be
stretched. What this implies is that to
represent meaning a constraint must
have elasticity. To deal with richness of
meaning, elasticity is necessary but not
sufficient. Consider the proposition:
Usually most flights leave on time.

49 /168 LAZ 6/9/2010

 CONTINUED

What is the constrained variable and

what is the constraining relation?
Actually, for most propositions drawn
from a natural language a large
repertoire of constraints is not
necessary. What is sufficient are three
so-called primary constraints and their
combinations. The primary constraints
are: possibilistic, probabilistic and
veristic.
50 /168 LAZ 6/9/2010
 CONTINUED

The role of primary constraints is

analogous to the role of primary colors:
red, green and blue. In most cases,
constraints are possibilistic.
Possibilistic constraints are much
easier to manipulate than probabilistic
constraints.

51 /168 LAZ 6/9/2010

 CONTINUED

Robert: OK, suppose I figured out what is

the constrained variable, X, and the
constraining relation, R. Is there
something else that has to be done?
Lotfi: Yes, there is. You see, X and R are
described in a natural language. What
this means is that we are not through
with precisiation of meaning of p. What
remains to be done is precisiation
(definition) of X and R.
52 /168 LAZ 6/9/2010
 CONTINUED

For this purpose, we construct a so-

called explanatory database, ED, which
consists of a collection of relations in
terms of which X and R can be defined.
The entries in relations in ED are
referred to as database variables.
Database variables are assumed to be
uninstantiated.

53 /168 LAZ 6/9/2010

 CONTINUED
Robert: Can you be more specific?
Lotfi: To construct ED you ask yourself
the question: What information—in the
form of a collection or relations—is
needed to precisiate (define) X and R?
Looking at p, we see that to precisiate
X we need two relations:
POPULATION.SWEDES[Name; Height]
and TALL[Height; µ].

54 /168 LAZ 6/9/2010

 CONTINUED

In the relation TALL[Height; µ], µtall(h) is

the grade of membership of a value of
Height, h, in the fuzzy set tall. So far as
R is concerned, the needed relation is
MOST[Proportion; µ], where µ is the
grade of membership of a value of
Proportion in the fuzzy set most.

55 /168 LAZ 6/9/2010

 CONTINUED

Equivalently, it is frequently helpful to

ask the question: What is the
information which is needed to assess
the degree to which p is true?

56 /168 LAZ 6/9/2010

 CONTINUED

At this point, we can express ED as the

collection:

ED= POPULATION.SWEDES[Name;
Height]+
TALL[Height; µ]+
MOST[Proportion; µ]

in which for convenience plus is used

in place of comma.
57 /168 LAZ 6/9/2010
 CONTINUED
Robert: So, we have constructed ED for
the proposition, p: Most Swedes are tall.
More generally, given a proposition, p,
how difficult is it to construct ED for p?
Lotfi: For humans it is easy. A few
examples suffice to learn how to
construct ED. Construction of ED is easy
for humans because humans have world
knowledge. At this juncture, we do not
have an algorithm for constructing ED.
58 /168 LAZ 6/9/2010
 CONTINUED
Robert: Now that we have ED, what
comes next?
Lotfi: We can use ED to precisiate
(define) X and R. Let us start with X. In
words, X is described as the proportion
of tall Swedes among Swedes. Let us
assume that in the relation
POPULATION.SWEDES there are N
names. Then the proportion of tall
Swedes among Swedes would be the
number of tall Swedes divided by N. LAZ 6/9/2010
59 /168
 CONTINUED
Here we come to a problem. Tall
Swedes is a fuzzy subset of Swedes.
The question is: What is the number of
elements in a fuzzy set? In fuzzy logic,
there are different ways of answering
this question. The simplest is referred
to as the ΣCount. More concretely, if A
is a fuzzy set with a membership
function µA, then the ΣCount of A is
defined as the sum of grades of
membership in A.
60 /168 LAZ 6/9/2010
 CONTINUED
In application to the number of tall
Swedes, the ΣCount of tall Swedes may
be expressed as:
N
ΣCount(tall.Swedes)= ∑ μ tall ( hi )
i =1

where hi is the height of Namei.

Consequently, the proportion of tall
Swedes among Swedes may be written
as:
1 N
X = ( ∑ μ tall ( hi ))
N i =1 LAZ 6/9/2010
61 /168
 CONTINUED
This expression may be viewed as a
precisiation (definition) of X in terms of
ED.

Precisiation (definition) of R is simpler.

Specifically, R=most, where most is a
fuzzy set. At this point, we have
precisiated (defined) X and R in terms
of ED.

62 /168 LAZ 6/9/2010

 CONTINUED
Robert: So what have we accomplished?
Lotfi: We started with a proposition, p:
Most Swedes are tall. We interpreted p
as a generalized (possibilistic)
constraint. We identified the
constrained variable, X, as the
proportion of tall Swedes among
Swedes. We identified the constraining
relation, R, as a fuzzy set, most. Next,
we constructed an explanatory
database, ED. LAZ 6/9/2010
63 /168
 CONTINUED
Finally, we precisiated (defined) X and
R in terms of ED. In this way, we
precisiated the meaning of p, which
was our objective. The precisiated
meaning may be expressed as the
constraint:
1 N
( ∑ μ tall ( hi )) is most
N i =1
Robert: So, you precisiated the meaning
of p. What purpose does it serve?
64 /168 LAZ 6/9/2010
 CONTINUED
Lotfi: The principal purpose is the
following. Unprecisiated (raw)
propositions drawn from a natural
language cannot be computed with.
Precisiation is a prerequisite to
computation. What is important is that
precisiation of meaning opens the door
to computation with natural language.

65 /168 LAZ 6/9/2010

 CONTINUED
Robert: Sounds great. I am impressed.
However, it is not completely clear to
me what you have in mind when you
say “opens the door to computation
with natural language.” Can you clarify
it?
Lotfi: With pleasure. Computation with
natural language or, more or less
equivalently, Computing with Words
(CW or CWW), is largely unrelated to
natural language processing. LAZ 6/9/2010
66 /168
 CONTINUED
More specifically, computation with
natural language is focused on
computation with information
described in a natural language.
Typically, what is involved is solution
of a problem which is stated in a
natural language. Let me go back to our
example, p: Most Swedes are tall. Given
this information, how can you compute
the average height of Swedes.
67 /168 LAZ 6/9/2010
 CONTINUED
Robert: Frankly, your question makes no
sense to me. Are you serious? How can
you expect me to compute the average
height of Swedes from the information
that most Swedes are tall?
Lotfi: That is conventional wisdom. A
mathematician would say that the
problem is ill-posed. It appears to be ill-
posed for two reasons.

68 /168 LAZ 6/9/2010

 CONTINUED
First, because the given information:
Most Swedes are tall, is fuzzy, and
second because you assume that I am
expecting you to come up with a crisp
answer, like the average height of
Swedes is 5’10”. Actually, what I expect
is a fuzzy answer—it would be
unreasonable to expect a crisp answer.
Robert: Thanks for clarification. I am
beginning to understand your question.
69 /168 LAZ 6/9/2010
 CONTINUED
Lotfi: I should like to add a key point. The
problem becomes well-posed if p is
precisiated. This is the essence of
Computing with Words.

70 /168 LAZ 6/9/2010

 CONTINUED
Robert: I am beginning to understand the
need for precisiation, but my
understanding is not complete as yet.
Can you explain how the average
height of Swedes can be computed
from precisiated p?
Lotfi: Recall that precisiated p is a
possibilistic constraint expressed as:
1 N
( ∑ μ tall ( hi )) is most
N i =1
71 /168 LAZ 6/9/2010
 CONTINUED
From the definition of a possibilistic
constraint it follows that the constraint
on X may be rewritten as:

1 N
t = μ most ( ∑ μ tall ( hi ))
N i =1

What this expression means is that

given the hi, µtall and µmost, we can
compute the degree, t, to which the
constraint is satisfied.
72 /168 LAZ 6/9/2010
 CONTINUED
It is this degree, t, that is the truth-value
of p. Now, here is a key idea. The
precisiated p constrains X. X is a
function of database variables. It
follows that indirectly p constrains
database variables. This has important
implications. Let me elaborate.

73 /168 LAZ 6/9/2010

 CONTINUED
What we see is that the constraint
induced by p on the hi is of the general
form
f(h1, …, hN) is most
What we are interested in is the
induced constraint on the average
height of Swedes. The average height
of Swedes may be expressed as:
1 N
have = ( ∑ hi )
N i =1
74 /168 LAZ 6/9/2010
 CONTINUED
This expression is of the general form
g(h1, …, hN) is ?have
where ?have is a fuzzy set that we want
to compute.

75 /168 LAZ 6/9/2010

 CONTINUED
At this stage, we can employ the
Extension Principle of fuzzy logic to
compute have. (Zadeh 1975 I, II & III) In
general terms, this principle tells us
that from a given constraint of the form
f(x1, …, xN) is A
in which A is a fuzzy set, we can derive
an induced constraint on g(x1, …, xN),
g(x1, …, xN) is B,
76 /168 LAZ 6/9/2010
 CONTINUED
in which B is a fuzzy set defined by the
solution of the mathematical program
µB(v)=supx1, …, xN µA(f(x1, …, xN))
subject to
v=g(x1, …, xN)
In application to our example, what we
see is that we have reduced computation
of the average height of Swedes to the
solution of the mathematical program
77 /168 LAZ 6/9/2010
 CONTINUED
µB(v)=suph1, …, hN µmost(f(h1, …, hN))
subject to
1 N
v = ( ∑ hi )
N i =1
In effect, this is the solution to the
problem which I posed to you. As you
can see, reduction of the original
problem to the solution of a
mathematical program is not so simple.
78 /168 LAZ 6/9/2010
 CONTINUED
However, solution of the mathematical
program to which the original problem
is reduced is well within the capabilities
of desktop computers.

79 /168 LAZ 6/9/2010

 CONTINUED
Robert: I am beginning to see the basic
idea. Through precisiation, you have
reduced the problem of computation
with information described in a natural
language—a seemingly ill-posed
problem—to a well-posed tractable
problem in mathematical programming.
I am impressed by what you have
accomplished, though I must say that
the reduction is nontrivial.
80 /168 LAZ 6/9/2010
 CONTINUED
Without your explanation, it would be
hard to see the basic ideas. I can also
see why computation with natural
language is a move into a new and
largely unexplored territory. Thank you
for clarifying the import of your
statement: precisiation of meaning
opens the door to computation with
natural language.

81 /168 LAZ 6/9/2010

 CONTINUED
Lotfi: I appreciate your comment. May I
add that I believe—but have not verified
it as yet—that in closed form the
solution to the mathematical program
may be expressed as:
have is ≥ most × tall
where most × tall is the product of
fuzzy numbers most and tall.
Robert: This is a very interesting result, if
true. It agrees with my intuition. LAZ 6/9/2010
82 /168
 CONTINUED
Lotfi: I appreciate your comment. I would
like to conclude our dialogue with a
prediction. As we move further into the
age of machine intelligence and
automated reasoning, the complex of
problems related to computation with
information described in a natural
language, is certain to grow in visibility
and importance.

83 /168 LAZ 6/9/2010

 CONTINUED
The informal dialogue between Robert
and Lotfi has come to an end. What
follows is a formal exposition of
generalized-constraint-based
semantics.

84 /168 LAZ 6/9/2010

 THE BASICS OF GENERALIZED-
CONSTRAINT-BASED SEMANTICS
z The point of departure in generalized-
constraint-based semantics, GCS, is an
unconventional definition of the concept
of a proposition. More specifically, in
GCS there are two equivalent definitions
of a proposition.

85 /168 LAZ 6/9/2010

 CONTINUED
z Definition 1. A proposition, p, in a
natural language may be interpreted as
an implicit assignment statement. p
assigns an implicit value, R, to an
implicit variable, X. Explicitness of X
and/or R is allowed as an exception.

86 /168 LAZ 6/9/2010

 CONTINUED
z Strictly speaking, R is not a value of X; R
is a restriction on the values which X is
allowed to take. To place this in
evidence, in Definition 2 the concept of
an assignment statement is replaced by
the concept of a constraint.

87 /168 LAZ 6/9/2010

 CONTINUED

z Definition 2. A proposition, p, in a
natural language may be interpreted as
an implicit constraint. In p, an implicit
variable, X, is constrained by an implicit
relation, R. R plays the role of a
restriction on the values which X is
allowed to take. Generally, R is a
predicate.

88 /168 LAZ 6/9/2010

 CONTINUED

Usually, R is a fuzzy set. Explicitness of

X and/or R is allowed as an exception. X
and R are described in a natural
language. X and R are identified by
inspection.

89 /168 LAZ 6/9/2010

 CONTINUED
z Definitions 1 and 2 are designed to open
the door to computation with natural
language and, more specifically, to
Computing with Words. In Computing
with Words, precisiation of meaning
plays a pivotal role.

90 /168 LAZ 6/9/2010

 EXAMPLES
p: Vera is middle-aged.
p Age(Vera) is middle-age

X R (fuzzy set)
p: Most of Robert’s friends are rich.
p Proportion (rich.friends.Robert/
friends. Robert)
is Most X

91 /168
R (fuzzy set) LAZ 6/9/2010
 CONTINUED
z In the following, Definition 2 will be
employed in preference to Definition 1.

z Standard constraints are of the form

XεC
where the constraining set, C, restricts
the values which X can take.

92 /168 LAZ 6/9/2010

 CONTINUED
z Standard constraints are not flexible
enough to be of use in precisiation of
meaning of propositions drawn from a
natural language. What is needed for
this purpose is the concept of a
generalized constraint, GC(X). A
generalized constraint is expressed as:

GC(X): X isr R

93 /168 LAZ 6/9/2010

 CONTINUED
z It is important to note that in contrast to
standard constraints, generalized
constraints have elasticity. Elasticity of
generalized constraints matches the
elasticity—stretchiness of meaning—of
natural languages.

94 /168 LAZ 6/9/2010

 CONTINUED
In the expression X isr R, X is the
constrained variable, R is the
constraining relation and r is an
indexical variable which defines the
modality of the constraint, that is, the
way in which R constrains X. To place in
evidence that X is the focus of
precisiation, X is referred to as the focal
variable, and R is referred to as the focal
relation.
95 /168 LAZ 6/9/2010
 CANONICAL FORM of p: CF(p)

z When X is represented as a generalized

constraint, the expression X isr R is
referred to as the canonical form of p,
CF(p). Thus,
CF(p): X isr R
z The concept of a canonical form of p
has a position of centrality in
precisiation of meaning of p.

96 /168 LAZ 6/9/2010

 PRIMARY CONSTRAINTS

z There are many types of generalized

constraints but only a few are needed
for precisiation of natural languages.
The primary constraints are:
possibilistic (r=blank); probabilistic
(r=p); and veristic (r=v). Natural
language constraints are
preponderantly possibilistic.

97 /168 LAZ 6/9/2010

 CONTINUED
z In the case of a possabilistic
constraint, X is R, assume that R is a
fuzzy set in a universe of discourse, U.
As a fuzzy set, R is precisiated through
graduation, that is, through association
with a membership function, µR, which
associates with each u in U its grade of
membership, µR(u), in R. A possibilistic
constraint on X is defined as:

X is R Poss(X=u)= µR(u)
98 /168 LAZ 6/9/2010
 CONTINUED

with R playing the role of the possibility

distribution of X.

99 /168 LAZ 6/9/2010

 CONTINUED
z A probabilistic constraint is expressed
as:
X isp R

where R plays the role of the

probability distribution of X.
z A veristic constraint is represented as:
X isv R,

with the truth-value (verity) of the

proposition, X=u, equated to µR(u).
100 /168 LAZ 6/9/2010
 CONTINUED

z Note. The importance of the concept of

primary constraints derives from the
fact that most constraints of interest
are derivable from primary constraints
through combination.

101 /168 LAZ 6/9/2010

 CONTINUED

z In representation of p as a generalized
constraint, p: X isr R, there are two
important points that have to be noted.
First, X need not be a scalar variable. X
may be vector-valued or, more
generally, have the structure of a
semantic network.

102 /168 LAZ 6/9/2010

 CONTINUED

z For example, in the case of the

proposition, p: Robert gave a ring to
Anne, X may be represented as the 3-
tuple (Giver, Recipient, Object), with the
corresponding value of R being
(Robert, Anne, Ring).

103 /168 LAZ 6/9/2010

 CONTINUED

z Second, in general, X is not unique.

However, it is usually the case that
among possible choices either there is
one that has higher plausibility than
others, or there a few that are closely
related. For example, if
p: Leslie is much taller than Ixel
then a plausible choice of X is
X: Height(Leslie)

104 /168 LAZ 6/9/2010

 CONTINUED

in which case the corresponding

constraining relation is
R: Much taller than Ixel.
Another plausible choice is
X: (Height(Leslie); Height(Ixel))
Correspondingly,
R: Much taller

105 /168 LAZ 6/9/2010

 CONTINUED

z Note that R is determined by X and p.

An important condition is that p should
be recoverable from X and R. X and R
are analogous to the subject and
predicate of a sentence.
z In CW, the choice of X is dictated by
external considerations. More
specifically, in CW the point of
departure is a question of the form:
What is the value of a variable, Y?
106 /168 LAZ 6/9/2010
 CONTINUED

In this setting, X should be a variable in

p which is question-relevant in relation
to the value of Y.
z As was noted earlier, in general X and
R are identified by inspection.

107 /168 LAZ 6/9/2010

 CONTINUED
z It is important to note that the use of
generalized constraints in precisiation
of propositions drawn from a natural
language is greatly facilitated by the
fact that, as noted earlier, natural
language constraints are for the most
part possibilistic—and hence are easy
to manipulate.

108 /168 LAZ 6/9/2010

 CONTINUED
z Less common are propositions whose
canonical forms involve a combination
of possibilistic and probabilistic
constraints. An important example is
the concept of a usuality constraint, X
isu R, meaning that usually (X is R.)

109 /168 LAZ 6/9/2010

 CONTINUED
z A usuality constraint may be expressed as a
combination of a possibilistic constraint and
a probabilistic constraint (Zadeh, Syllogistic
reasoning in fuzzy logic and its application
to usuality and reasoning with dispositions,
IEEE Transactions on Systems, Man and
Cybernetics 15, 754-763, 1985; Outline of a
theory of usuality based on fuzzy logic,
Fuzzy Sets Theory and Applications, A.
Jones, A. Kaufmann and H.J. Zimmermann
(eds.), 79-97. Dordrecht: Reidel, 1986.)
110 /168 LAZ 6/9/2010
 CONTINUED
z In conclusion, the concept of a
generalized constraint is the
centerpiece of generalized-constraint-
based semantics. The importance of
the concept of a generalized constraint
derives from the fact that it makes it
possible to standardize precisiation by
expressing a precisiated form of p as a
generalized constraint.

111 /168 LAZ 6/9/2010

 THE CONCEPT OF EXPLANATORY
DATABASE (ED)

z In generalized-constraint-based
semantics, the information base, IB(p),
is represented as a collection of
relations, with the names of relations
drawn, but not exclusively, from the
constituents of p.

112 /168 LAZ 6/9/2010

 CONTINUED
z The collection of relations is referred to
as the explanatory database, ED (Zadeh
1984). Basically, ED may be viewed as
the information which is needed to
define X and R. For example, for the
proposition, p: Most Swedes are tall,
ED may be represented as:
ED=POPULATION.SWEDES[Name;
Height]+TALL[Height;µ]+
MOST[Proportion;µ],
where + plays the role of comma.
113 /168 LAZ 6/9/2010
 CONTINUED

z It is important to note that definition of

X and R may be viewed as precisiation
of X and R. Precisiation of X and R is
needed because X and R are described
in a natural language.
z In relation to possible world semantics,
ED may be viewed as the description of
a possible world.

114 /168 LAZ 6/9/2010

 ADDITIONAL EXAMPLES OF ED
z p: Brian is much taller than most of his
friends.
X: Height of Brian.
R: Much taller than most of his friends.
ED=HEIGHT[Name; Height]+
FRIENDS.BRIAN[Name; µ]+
MUCH.TALLER [Height1; Height2; µ]+
MOST[Proportion; µ]

115 /168 LAZ 6/9/2010

 CONTINUED

In FRIENDS.BRIAN, µ is the degree to

which Name is a friend of Brian. The
relation is extensional.

116 /168 LAZ 6/9/2010

 CONTINUED

z p: The incidence of lung cancer among

heavy smokers is much higher than
among those who are non-smokers.
X: Incidence of lung cancer among
heavy smokers.
R: Much higher than among non-
smokers.

117 /168 LAZ 6/9/2010

 CONTINUED

ED=POPULATION[Name;
µHeavy.Smoker; µNon-smoker; µLung.Cancer] +
MUCH.HIGHER[Proportion 1;
Proportion 2; µ]

Note: In µNon-smoker, µ takes the values 0 or

1.

118 /168 LAZ 6/9/2010

 CONTINUED
z p: Over the past few years Sam made a
lot of money.
X: Sam’s income over the past few years.
R: Lot of money.
ED=INCOME.SAM[Name; Amount; Year]+
FEW[Number; µ]+
LOT.OF.MONEY[Amount; µ]

119 /168 LAZ 6/9/2010

 CONTINUED
z It is important to note that relations in ED
are uninstantiated, that is, the values of
database variables—entries in relations—
are not specified.

120 /168 LAZ 6/9/2010

 THE CONCEPT OF A PRECISIATED
CANONICAL FORM, CF*(p)

z After X and R have been identified and

the explanatory database, ED, has been
constructed, X and R may be defined as
functions of ED. As was noted earlier,
definition of X and R may be viewed as
precisiation of X and R. Precisiated X
and R are denoted as X* and R*,
respectively.

121 /168 LAZ 6/9/2010

 CONTINUED
z A canonical form, CF*(p), with
precisiated values of X and R, X* and R*,
will be referred to as a precisiated
canonical form.
z In the following, construction of the
precisiated canonical form of p is
discussed in greater detail.

122 /168 LAZ 6/9/2010

123 /168 LAZ 6/9/2010
 z The concepts discussed so far provide
a basis for a relatively straightforward
procedure for constructing the
precisiated canonical form of a given
proposition, p. The precisiated
canonical form may be viewed as a
precisiated form of p. Effectively, the
precisiated canonical form may be
interpreted as a representation of
precisiated meaning of p.
124 /168 LAZ 6/9/2010
 z A summary of the procedure for
computing the precisiated canonical
form of p is presented in the following.

125 /168 LAZ 6/9/2010

 PROCEDURE
z Step 1. Clarification
The first step is clarification, if needed,
of the meaning of p. This step requires
world knowledge.
Examples:
clarification
z Overeating causes obesity
Most of those who overeat are obese.

z Obesity is caused by overeating clarification

Most of those who are obese, overeat.
126 /168 LAZ 6/9/2010
 CONTINUED
z Young men like young women clarification
Most young men like mostly young women.
z Swedes are much taller than Italians
clarification
Most Swedes are much taller than
most Italians.
z Step 2. Identification (explicitation) of X and
R.
Identify the constrained variable, X, and the
corresponding constraining relation, R.

127 /168 LAZ 6/9/2010

 CONTINUED

z Step 3. Construction of ED.

What information is needed—but not
necessarily minimally—to precisiate
(define) X and R? An answer to this
question identifies the explanatory
database, ED. Equivalently, ED may be
viewed as an answer to the question:
What information is needed—but not
necessarily minimally—to compute the
truth-value of p?
128 /168 LAZ 6/9/2010
 CONTINUED
z Step 4. Precisiation of X and R.
How can the information in ED be used
to precisiate the values of X and R? This
step leads to precisiated values of X and
R, X* and R*, and thus results in the
precisiated canonical form, CF*(p).
z It is important to note that CF*(p)
induces a constraint on the decision
variables in ED. This constraint may be
interpreted as the possibility distribution
of decision variables in ED.
129 /168 LAZ 6/9/2010
 CONTINUED

z Note. It should be underscored that

definition of X and R serves the function
of precisiating the meaning of X and R,
but not that of instantiating the values of
X* and R*. Instantiation of the values of
X* and R* requires instantiation of
relations in ED.
End of procedure.

130 /168 LAZ 6/9/2010

 CONTINUED
z Step 5. (Optional) Computation of truth-
value of p. The truth-value of p depends
on ED. The truth-value of p, t(p, ED), may
be computed by assessing the degree to
which the generalized constraint, X* isr
R*, is satisfied.
z It is important to note that humans have
no difficulty in learning how to use the
procedure. The reason is: Humans have
world knowledge. It is hard to build world
knowledge into machines.
131 /168 LAZ 6/9/2010
 EXAMPLES
z Note. In the following examples
r=blank, that is, the generalized
constraints are possibilistic.
1. p: Most Swedes are tall
Step 1. Clarification. Clarification not
needed
Step 2. Identification (explicitation) of X
and R.
X is identified as the proportion of tall
Swedes among Swedes.
132 /168 LAZ 6/9/2010
 CONTINUED

Correspondingly, R is identified as
most.
Digression.
In fuzzy logic, proportion is defined as
a relative ΣCount. (Zadeh 1983) More
specifically, if A and B are fuzzy sets in
U, U={u1, …, un}, the
ΣCount(cardinality) of A is defined as:
ΣCount ( A ) = Σ i μ A ( ui )
133 /168 LAZ 6/9/2010
 CONTINUED

The relative ΣCount of B in A is defined

as:

ΣCount ( A I B )
ΣCount ( B / A ) =
ΣCount ( A )
Σ i μ A ( ui ) ∧ μ B ( ui )
=
Σ i μ A ( ui )
where I =intersection and ∧ =min
134 /168 LAZ 6/9/2010
 CONTINUED
In application to the example under
consideration, assume that the height of
ith Swede, Namei, is hi and that the grade
of membership of hi in tall is µtall(hi), i=1,
…, N. X may be expressed as:
1
X = ( Σi μ tall ( hi ))
N
Step 3. Construction of ED.
The needed information is contained in
the explanatory database, ED, where LAZ 6/9/2010
135 /168
 CONTINUED
ED= POPULATION.SWEDES[Name;
Height]+
TALL[Height; µ]+
MOST[Proportion; µ]

Step 4. Precisiation of X and R.

In relation to ED, precisiated X and R may
be expressed as:
1
X * = ( Σi μ tall ( hi ))
N
R* = MOST[Proportion; µ] LAZ 6/9/2010
136 /168
 CONTINUED
z The precisiated canonical form is
expressed as:

CF*p=X* is R*

where
1
X * = ( Σi μ tall ( hi ))
N

R* = MOST[Proportion; µ]
137 /168 LAZ 6/9/2010
 CONTINUED
Step 5. The truth-value of p, t(p, ED), is
the degree to which the constraint in
Step 4 is satisfied. More concretely,

1
t ( p , ED ) = μ most ( Σi μ tall ( hi ))
N

Note. The right-hand side of this equation

may be viewed as a constraint on
database variables h1, …, hN.
138 /168 LAZ 6/9/2010
 CONTINUED

2. p: Overeating causes obesity

Step 1. Clarification.

p is restated as: Most of those who

overeat are obese.

139 /168 LAZ 6/9/2010

 CONTINUED

Step 2. Identification of X and R.

X is identified as the proportion of
those who are obese among those who
overeat. Correspondingly, R is
identified as Most.

X= Proportion of those who are obese

among those who overeat.
R= Most

140 /168 LAZ 6/9/2010

 CONTINUED
Step 3. Construction of ED

ED=POPULATION[Name; µObese;µOvereat]
+ MOST[Proportion; µ]

Step 4. Precisiation of X and R

X*= Σi(µObese(Namei) ^ µOvereat(Namei))

ΣiµOvereat(Namei)
R*= MOST[Proportion; µ]
141 /168 LAZ 6/9/2010
 CONTINUED
The precisiated canonical form is: CF*(p)=X is
R*, where
Σi(µObese(Namei) ^ µOvereat(Namei))
X*= ΣiµOvereat(Namei)
R*= MOST[Proportion; µ]

Step 5. The truth-value of p, t(p, ED), is

t(p, ED)=µmost (Σi(µObese(Namei)^µOvereat(Namei))/
ΣiµOvereat(Namei))
142 /168 LAZ 6/9/2010
 EXAMPLE

3. p: Carol lives in a small city near San

Francisco.
Step 1. Clarification. Not needed.
Step 2. Identification of X and R.
X= Location(Residence (Carol))
R= Small city near San Francisco
Step 3. Construction of ED
ED= LOCATION.RESIDENCE[Name;
City.Name]+
143 /168 LAZ 6/9/2010
 CONTINUED

SMALL.CITY[City.Name; µ] +
NEAR.SF.CITY[City.Name; µ]

z In the relation SMALL.CITY[City.Name;

µ] µ is the degree to which City.Name is
a small city.
z In the relation
NEAR.SF.CITY[City.Name; µ] µ is the
degree to which City.Name is a city
near San Francisco. LAZ 6/9/2010
144 /168
 CONTINUED

Step 4. Precisiation of X and R

X*=City.NameLOCATION.RESIDENCE[Name
= Carol; City.Name]

R*=SMALL.CITY[City.Name; µ] ∩
NEAR.SF.CITY[City.Name; µ]

145 /168 LAZ 6/9/2010

 CONTINUED

The precisiated canonical form, is

expressed as CF*(p)= X* is R*, where

City.NameLOCATION.RESIDENCE[Name=
X*=
Carol; City.Name] is

SMALL.CITY[City.Name; µ] ∩
R*=
NEAR.SF.CITY[City.Name; µ]

146 /168 LAZ 6/9/2010

 CONTINUED

z Note. R* is a fuzzy set of cities. The

grade of membership of a city,
City.Name, in R* may be interpreted in
two ways. First, as the possibility that
Carol lives in City.Name, given the
proposition p: Carol lives in a small city
near San Francisco; and second, as the
truth-value of the proposition p: Carol
lives in a small city near San Francisco,
given that she lives in City.Name.
147 /168 LAZ 6/9/2010
 EXAMPLE
4. p: Robert’s income is significantly
higher than the combined income of
John and Peter.
Step 1. Clarification. Not needed.
Step 2. Identification of X and R
X= Robert’s income(IR)
R= Significantly higher than the
combined income of John and Peter

148 /168 LAZ 6/9/2010

 EXAMPLE
Step 3. Construction of ED
ED= INCOME[Name; Income] +
SIGNIFICANTLY.HIGHER[Income1;
Income2; µ]

Step 4. Precisiation of X and R

Precisiation of X

149 /168 LAZ 6/9/2010

 CONTINUED
X*=IR=IncomeINCOME[Name=Robert;
Income]
Note. The notation AR means read A in R.
Precisiation of R:
Find the incomes of John, IJ, and Peter,
IP
IJ= IncomeINCOME[Name=John;
Income]
IP= IncomeINCOME[Name=Peter;
Income]
150 /168 LAZ 6/9/2010
 CONTINUED
Compute the combined income, IJP
IJP=IJ + IP
In the relation
SIGNIFICANTLY.HIGHER[Income1;
Income2; µ]
instantiate Income2 to IJP and read
Income1 and µ. The result is precisiated R
R*=Income1;µSIGNIFICANTLY.HIGHER[Inco
me1; Income2=IJP; µ]
151 /168 LAZ 6/9/2010
 CONTINUED
The precisiated canonical form may be
expressed as CF*(p)=X* is R*, where
X*=IncomeINCOME[Name=Robert; Income]

R*=Income1;µSIGNIFICANTLY.HIGHER[Inco
me1; Income2=IJP; µ]

152 /168 LAZ 6/9/2010

 EXAMPLE

5. p: Swedes are much taller than

Italians.
Step 1. Clarification.
p clarification
Most Swedes are much taller
than most Italians.
Step 2. Identification of X and R.
X= Proportion of Swedes who are
much taller than most Italians.
R= Most
153 /168 LAZ 6/9/2010
 CONTINUED

Step 3. Construction of ED
POPULATION.SWEDES[Name;
Height] +
POPULATION.ITALIANS[Name;
Height] +
MUCH.TALLER[Height1; Height2; µ]
+ MOST[Proportion; µ]

154 /168 LAZ 6/9/2010

 CONTINUED

Step 4. Precisiation of X and R

Precisiation of X
Find the height of ith Swede, Namei, i=1,
…, m
hi=HeightPOPULATION.SWEDES[Name=
Namei; Height]
Find the height of jth Italian, Namej, j=1,
…, n
kj=HeightPOPULATION.ITALIANS[Name=
155 /168
Namej; Height] LAZ 6/9/2010
 EXAMPLE

Find the degree, aij, to which Namei is

much taller than Namej
aij=µMUCH.TALLER[Height1=hi;
Height2=kj; µ]

Compute the proportion, pi, of Italians

in the relation to whom Namei is much
taller
1
pi = Σ j aij
n
156 /168 LAZ 6/9/2010
 CONTINUED

Compute the degree, qi, to which pi

satisfies most

1
qi = μ MOST [Pr oportion = Σ j aij ; μ ]
n
Compute the proportion, X*, of Swedes
who are much taller than most Italians
1
X * = Σi qi
m
157 /168 LAZ 6/9/2010
 CONTINUED

Precisiation of R
R*= MOST[Proportion; µ]
The precisiated canonical form may be
expressed as CF*(p)= X is R, where
1
X * = Σi q i
m
R*= MOST[Proportion; µ]
Additional examples may be found in
cited papers.
158 /168 LAZ 6/9/2010
 SUMMATION

z Natural languages are pervasively

imprecise, especially in the realm of
meaning. The primary source of
imprecision is unsharpness of class
boundaries. In this sense, words,
propositions and commands in natural
languages are preponderantly
imprecise.

159 /168 LAZ 6/9/2010

 CONTINUED

z Precisiation of meaning is a
prerequisite to achievement of higher
levels of mechanization of natural
language understanding. Precisiation
of meaning plays a particularly
important role in communication
between humans and machines.

160 /168 LAZ 6/9/2010

 CONTINUED

z Despite its intrinsic importance,

precisiation of meaning has drawn
little, if any, attention within linguistics
and computational linguistics. There is
a reason. In large measure, theories of
natural languages are based on
bivalent logic.

161 /168 LAZ 6/9/2010

 CONTINUED

z Bivalent logic is not the right logic for

dealing with imprecision of natural
languages. What is needed for this
purpose is fuzzy logic. Basically, fuzzy
logic is the logic of classes with
unsharp boundaries.

162 /168 LAZ 6/9/2010

 CONTINUED

z In the approach presented in this

lecture, the point of departure is an
unconventional definition of the
concept of a proposition. There are two
equivalent versions. In the first version,
a proposition, p, is defined as an
implicit assignment statement. p
assigns to an implicit variable, X, an
implicit value, R.

163 /168 LAZ 6/9/2010

 CONTINUED

z In the second version, p is defined as

an implicit constraint. In this case, X is
an implicit constraint variable and R is
an implicit constraining relation.
Basically, R restricts the values which
X is allowed to take. Usually, R is a
fuzzy set. Since there are many ways in
which R can constrain X, the constraint
on X is referred to as a generalized
constraint.
164 /168 LAZ 6/9/2010
 CONTINUED

z A generalized constraint is an
expression of the form, X isr R, in
which X is the constrained variable, R
is the constraining relation and r is an
indexical variable which defines the
way in which R constrains X. The
primary constraints are possibilistic
(r=blank); probabilistic (r=p); and
veristic (r=v).

165 /168 LAZ 6/9/2010

 CONTINUED

z The meaning of p is precisiated by

representing p as a generalized
constraint, X isr R. Generally, the
constraint is possibilistic or a
combination of a possibilistic
constraint and a probabilistic
constraint. Usually, X and R can be
identified by inspection.

166 /168 LAZ 6/9/2010

 CONTINUED

z Representation of p as a generalized
constraint is the centerpiece of
generalized-constraint-based
semantics, GCS. Through precisiation
of meaning, GCS opens the door to a
wide-ranging enlargement of the role of
natural languages in scientific theories.

167 /168 LAZ 6/9/2010

 CONTINUED

z Precisiation of meaning plays a pivotal

role in CW. What should be
underscored, however, is that in CW
precisiation of meaning does not
require the elaborate machinery which
is described in this lecture.

168 /168 LAZ 6/9/2010