Scribd
Upload
39 views

Lecture19 PDF

This document discusses models for quantified logic. It begins by explaining that models in quantified logic involve interpreting names, predicates, and quantifiers directly rather than just assigning truth values. Models specify a universe of discourse (set of objects) and interpret names as assignments of objects, predicates as extensions listing which objects satisfy them, and quantifiers by checking if every instance is true. The document provides examples of interpreting universal and existential quantifiers based on a model with a universe of bears and monadic/dyadic predicates.

Uploaded by

Bob0b
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
39 views

Lecture19 PDF

This document discusses models for quantified logic. It begins by explaining that models in quantified logic involve interpreting names, predicates, and quantifiers directly rather than just assigning truth values. Models specify a universe of discourse (set of objects) and interpret names as assignments of objects, predicates as extensions listing which objects satisfy them, and quantifiers by checking if every instance is true. The document provides examples of interpreting universal and existential quantifiers based on a model with a universe of bears and monadic/dyadic predicates.

Uploaded by

Bob0b
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 21

PHIL 220:

Symbolic Logic
Dr. Jonathan Jenkins Ichikawa

Lecture 19:
Models for Quantified Logic
1
Note on Translation

l For translation into English: make sure you translate into


reasonably natural English—the kind of thing an ordinary
English-speaker would understand without much trouble.
l ∀x((Px & Fx) ⊃Lax)
l “Every person who is is funny is something that Alice loves.”
l “Alice loves every funny person.” ß Much better

2
Goals for Today

l Understand models for quantified logic

3
An Argument

l Every Canadian is friendly l ∀x(Cx ⊃ Fx)


l Donald isn’t friendly l ~Fd
l Donald isn’t Canadian l ~Cd

l This is a valid argument; we want to give it a valid


argument form.
l We need to think about models in order to
understand validity in predicate logic
4
Models in Sentential Logic
l X ⊨ A:
l Every interpretation that satisfies X, satisfies A.
l Models in SL are assignments of atomic sentences to truth
conditions.

l Example:
l (P v Q) ⊃ R ⊨ (P ⊃ R) v (Q ⊃ R)
l The possible interpretations:
P=1 P=1 P=1 P=1 P=0 P=0 P=0 P=0
Q=1 Q=1 Q=0 Q=0 Q=1 Q=1 Q=0 Q=0
R=1 R=0 R=1 R=0 R=1 R=0 R=1 R=0

l Check that each model either satisfies A or falsifies X (or both).


5
Models in Quantified Logic

l Models in quantified logic can’t just be assignments of


truth values to propositions.
l Otherwise our argument form wouldn’t be valid.
l ∀x(Lx ⊃ Sx); ~Sd; therefore, ~Ld
l Let I(∀x(Lx ⊃ Sx))=1, I(Sd)=0, I(Ld)=1

l We need to interpret names, predicates, and quantifiers


directly.

l So we will be working with richer models as


interpretations
6
UDs

l We begin with the universe of discourse. A UD is a (non-


empty) set of objects; it identifies the objects our sentence
are talking about.

l For example, a domain might be:


l Bears: {Mama bear, Papa bear, Baby bear}
l MLB AL East: {Blue Jays, Red Sox, Yankees, Orioles, Rays}
l Integers: {0, 1, 2, 3, …}

l What a formula with quantifiers is talking about depends


on the UD.

7
Names
l Each name occurring in a formula must be associated with an
object in the domain.

l A model will include an assignment of (at least) one name for


every object.
l Consider the bears: {Mama bear, Papa bear, Baby bear}

l I(a) = Mama bear means that the interpretation assigns the


name a to the object Mama bear.

l Here are some possible assignments of the names a, b, and c:


l I(a) = Mama; I(b) = Papa; I(c) = Baby
l I(a) = Mama; I(b) = Baby; I(c) = Papa
l I(a) = Papa; I(b) = Mama; I(c) = Baby
8
Monadic Predicates

l Monadic predicates are one- I(P)


place predicates, like ___ is red
or ___ is in love with Freddy.
l Consider our previous UD of Mama 1
bears and the one-place
predicate P for ___ is a Parent. Papa 1
l The predicate P has this
extension: {Mama Bear, Papa
Bear} Baby 0
l Or we may represent it thus:
9
Monadic Predicates

I(S)

l Here’s another predicate Mama 1


l Let Sx stand for x has a very
soft bed Papa 0
l I(S)={Mama}

Baby 0

10
Monadic Predicates

I(S)
l Let Sx stand for x has a very
soft bed Mama 0
l Here’s a different extension:
l I(S)={Baby}
l Given this extension, Baby
Papa 0
has a very soft bed, but
Mama and Papa don’t. Baby 1

11
Monadic Predicates

l Some monadic predicates I(B)


may categorize every object
the same. Mama 1
l Consider the predicate B for
___ is a bear.
l All three objects in the Papa 1
domain satisfy that
predicate.
Baby 1
l I(B)={Mama, Papa, Baby}

12
Dyadic Predicates

l A two-place predicate has ordered pairs in its extension.


Consider Lxy, for x is larger than y.
l I(L)={<Papa, Mama>, <Papa, Baby>, <Mama, Baby>}
I(L) Mama Papa Baby
Mama 0 0 1
Papa 1 0 1
Baby 0 0 0
l Read these charts looking at the row first, then the
column. The red cell says that <Papa, Mama> is in L (i.e.,
Papa is larger than Mama)
13
Polyadic Predicates

l A three-place predicate will have ordered triples in its


extension.
l Pxyz: x likes their porridge at least as hot as y does, but
not as hot as z does
l I(P)={ <Baby, Mama, Papa>,
<Baby, Baby, Papa>,
<Mama, Mama, Baby>,
<Mama, Mama, Papa>}
l In general, the extension of an n-place predicate is a set of
n-tuples of objects in the UD
14
Rules for Evaluating Formulae
In a given model…

l Atomic sentences are assigned 1 or 0 directly.

l Fa1, … an is true iff <I(a1), … I(an)> is in the extension of F.

l Propositional connectives work as before:


l ~A=1 iff =0
l (A & B)=1 iff A=1 and B=1
l (A v B)=1 iff A=1 or B=1
l (A ⊃ B)=1 iff A=0 or B=1
l (A º B)=1 iff A=1 and B=1 or A=0 and B=0.

l (We’ll talk quantifiers soon.)


15
A Model (LC questions)
l UD: {Mama, Papa, Baby} I(B)
l I(a) = Mama
Mama 1
l I(b) = Papa
l I(c) = Baby Papa 1

Baby 1

I(L) Mama Papa Baby


I(H) I(S)
Mama 0 0 1
Papa 1 0 1 Mama 0 Mama 1
Baby 0 0 0
Papa 1 Papa 0

16 Baby 0 Baby 0
True or false: Lab ⊨ ~Lba?

l Tempting to say: true, because L means is larger than, and


so is irreflexive

l But L only means is larger than given this particular


evaluation; entailment is about all possible evaluations

l There are possible evaluations that satisfy Lab and ~~Lba,


so the QL argument form from Lab to ~Lba is invalid

17
Quantifiers and Models

l How do we understand ∀x Bx on our interpretation?

l A universal formula is true iff each instance is true.

l The instances of Bx in this model are Ba, Bb, and Bc.

l We replace the variable with each of these names.

18
Examples

l So on our interpretation about the bears:


l ∀x ~Sx is false
l one of its instances is false — ~Sa
l ∀x Bx is true
l Each of its instances is true – Ba, Bb, Bc

19
A Model (LC questions)
l UD: {Mama, Papa, Baby} I(B)
l I(a) = Mama
Mama 1
l I(b) = Papa
l I(c) = Baby Papa 1

Baby 1

I(L) Mama Papa Baby


I(H) I(S)
Mama 0 0 1
Papa 1 0 1 Mama 0 Mama 1
Baby 0 0 0
Papa 1 Papa 0

20 Baby 0 Baby 0
Quantifiers and Models

l More on interpreting quantified formulae—including


embedded quantifiers—next time

21

You might also like