Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
Melvin Fitting
TRENDS IN LOGIC
Heinrich Wansing
The titles published in this series are listed at the end of this volume.
MELVIN FITTING
Lehman College and the Graduate Center,
City University of New York, U.S.A.
TYPES,TABLEAUS,
AND GO DEL'S GOD
~~~....
''
A C.I.P. Catalogue record for this book is available from the Library of Congress.
ISBN 1-4020-0604-7
Contents
PREFACE
Part I
Xl
CLASSICAL LOGIC
1. CLASSICAL LOGIC-SYNTAX
1
2
3
3
Substitutions
2. CLASSICAL LOGIC-SEMANTICS
Classical Models
1
Truth in a Model
2
Problems
3
Compactness
3.1
Strong Completeness
3.2
Weak Completeness
3.3
And Worse
3.4
Henkin Models
4
Generalized Henkin Models
5
A Few Technical Results
6
6.1
Terms and Formulas
Extensional Models
6.2
Language Extensions
6.3
3. CLASSICAL LOGIC-BASIC TABLEAUS
A Different Language
1
Basic Tableaus
2
Tableau Examples
3
v
11
11
12
15
15
16
16
17
19
24
29
29
29
30
33
33
35
37
vi
43
43
46
47
48
52
59
62
63
66
5. EQUALITY
Adding Equality
1
Derived Rules and Tableau Examples
2
Soundness and Completeness
3
69
69
69
73
6. EXTENSIONALITY
1
Adding Extensionality
A Derived Rule and an Example
2
Soundness and Completeness
3
77
Part II
77
77
79
MODAL LOGIC
83
83
86
89
90
92
94
95
101
102
8. MODAL TABLEAUS
1
The Rules
1.1
Prefixes
Propositional Rules
1.2
105
105
105
107
vii
Contents
1.3
Modal Rules
1.4
Quantifier Rules
1.5
Abstraction Rules
1.6
Atomic Rules
1.7
Proofs and Derivations
Tableau Examples
107
108
109
109
110
111
113
9. MISCELLANEOUS MATTERS
Equality
1
1.1
Equality Axioms
1.2
Extensionality
115
115
115
117
De Re and De Dicta
3
4
Rigidity
118
121
Stability Conditions
124
Definite Descriptions
Choice Functions
125
128
Part III
ONTOLOGICAL ARGUMENTS
133
133
134
134
137
138
139
145
General Plan
145
Positiveness
145
Objections
150
152
5
6
7
Essence
Necessarily God Exists
Going Further
Monotheism
7.1
156
160
162
162
viii
7.2
Positive Properties are Necessarily Instantiated
More Objections
162
9
10
A Solution
164
Anderson's Alternative
169
11
Conclusion
171
163
REFERENCES
173
INDEX
179
Preface
What's Here
This is a book about intensional logic. It also provides a thorough
look at higher-type classical logic, including tableaus and a completeness proof for them. It also provides a formal examination of the Godel
ontological argument. These are not disparate topics. Higher-type classical logic is intensional logic with the intensional features removed, so
this is a good place to start. Ontological arguments, Godel's in particular, are natural examples of intensional logic at work, so this is a good
place to finish.
The term formal logic covers a broad range of inventions. At one
end are small, special-purpose systems; at the other are rich, expressive
ones. Higher-type modal logic-intensional logic-is one of the rich ones.
Originating with Carnap and Montague, it has been applied to provide
a semantics for natural language, to model intensional notions, and to
treat long-standing philosophical problems. Recently it has also supplied
a semantic foundation for some complex database systems. But besides
being rich and expressive, it is also tremendously complex, and requires
patience and sympathy on the part of its students.
There are two quite different reasons to be interested in a logic. There
is its formal machinery for its own sake, and there is using the formal
machinery to address problems from the outside world. The mechanism
of higher-type modal logic is complex and requires serious mathematics
to develop properly. Models are not simple to define, and tableau systems are quite elaborate. A completeness argument, to connect the two,
is difficult. But, the machinery is of considerable interest, if this is the
sort of thing you have a considerable interest in. If you are such a reader,
applications concerning the existence of God can simply be skipped. On
the other hand, if philosophical applications are what you are after, the
xi
Xll
PREFACE
xiii
XlV
also Anselm) all have proofs that stick at the same point: showing that
the existence of God is possible.
If the Godel argument is what you are interested in, start with Part III,
and pick up earlier material as needed. Many of the uses of the formalism are relatively intuitive. Indeed, in Godel's notes on his ontological
argument, formal machinery is never discussed, yet it is possible to get
a sense of what it is about anyway.
PREFACE
XV
Acknowledgments
An earlier draft of this work was on my web page for some time, and I
was given several helpful suggestions as a result. In particular I want to
thank Peter Hajek, Oliver Kutz, Paul Gilmore, and especially Howard
Sobel.
CLASSICAL LOGIC
Chapter 1
CLASSICAL LOGIC-SYNTAX
1.
where cp(x1, ... , xn) is a formula with free variables as indicated. Such
axioms ensure that to each formula corresponds an 'object.' The other
approach is to elaborate the term-forming machinery, so that there is
an explicit name for the object specified by a formula cp. This involves
predicate abstraction, or >-.-abstraction:
3
M. Fitting, Types, Tableaus, and Gdel's God
Kluwer Academic Publishers 2002
The two approaches are equivalent in a direct way. I have chosen to use
explicit abstracts for several reasons. First, axioms are not as natural
when tableau systems are the proof machinery of choice. And second,
predicate abstraction has already played a major role in earlier investigations of modal logic [FM98], and makes discussion of major issues
considerably easier here.
Finally, one can characterize higher-order formulas more-or-less the
way it is done in the first-order setting, taking quantifiers and connectives as "logical constants." This is the approach of [Sch60]. Alternatively, following [Chu40], one can think of quantifiers and connectives as
constants of the language, which itself is formulated in lambda-calculus
style. In this book I take the first approach, though one can make
arguments for the second on grounds of elegance and economy. My justification is that doing things the way that has become standard for
first-order logic will be less confusing to the reader.
Recently one further alternative has become available. In [Gil99,
GilOl], Paul Gilmore has shown that by a relatively simple change, a
system of classical higher-order logic can be developed allowing a controlled degree of impredicativity-typing rules can be relaxed to permit
the formation of certain useful sentences that are not "legal" in the approach presented here. This, in turn, allows a more natural development
of arithmetic in the higher-order setting. I do not follow Gilmore's approach here, but I recommend it for study. Much of what I develop
carries over quite directly.
So these are my choices: no function symbols, explicit predicate abstraction, quantifiers and connectives as in the first-order setting, and
no impredicativity. With this out of the way I can begin presenting the
formal syntactical machinery.
In first-order logic, relation symbols have an arity--some are oneplace, some are two-place, and so on. In higher-order logic this simple
idea gets replaced by a typing mechanism, which is considerably more
complex. Terms, and certain other items, are assigned types, and rules
of formation make use of these types to ensure that things fit together
properly. I begin by saying what the types are.
DEFINITION 1.1 (TYPE) 0 is a type. If t1, ... , tn are types, (tl, ... , tn)
is a type. I generally use t, t1, t2, t', etc. to represent types.
An object of type 0 is intended to be a ground-level object-it corresponds to the designation of a constant symbol or variable in standard
first-order logic. An object of type (t 1 , ... , tn) is a predicate that takes
n arguments, of types t 1 , ... , tn respectively. Thus a constant symbol
of type (0, 0, 0), say, would be called a three-place relation symbol in
CLASSICAL LOGIC-SYNTAX
L( C). Its type is (t1, ... , tn), and its free variable occurrences are the
free variable occurrences in the formula~' except for occurrences of the
variables a1, . . . , an.
DEFINITION 1.5 (TERM OF L(C)) Terms of each type are characterized as follows.
If~
...,~
is a formula of L( C) so is
are those of<>.
...,~.
If~
If~
EXAMPLE 1. 7 Suppose a(o,o} is a variable of type (0, 0) (and so firstorder), (3 is a variable of type 0, and 'Y((O,O},O} is a variable of type
((0,0),0) (second-order). Both (3 and 'Y((O,O},o} are terms. Then the
expression 'Y( (o,o} ,o} (a (O,O}, (3) is an atomic formula. Generally I will
write the simpler looking 'Y( a, (3), and give the information contained
in the superscripts in a separate description. Since this atomic formula
contains a variable 'Y of order 2, it is referred to as a second-order atomic
formula.
DEFINITION 1.8 (SENTENCE) A formula with no free variables is a sentence.
CLASSICAL LOGIC-SYNTAX
Constants
Variables
a, b, c, ...
A,B,C, ...
x, y, z, ...
A,B,C, ...
X,Y,Z, ...
X, Y,Z, ...
ExAMPLE 1.9 For this example I give explicit type information (in superscripts), until the end of the example. After this I omit the superscripts, and say in English what is needed to restore them.
Suppose x 0 , X(O), and X((O)) are variables (the first is of order 0, the
second is of order 1, and the third is of order 2). Also suppose p((O)) and
g0 are constant symbols of L(C) (the first is of order 2 and the second
is of order 0).
1 Both X((O))(X( 0)) and X( 0)(x 0 ) are atomic formulas. All variables
present have free occurrences.
2 (>..X((o)).X((O))(X( 0 ))) is a predicate abstract, of type (((0))). Only
the occurrence of X(O) is free.
3 SinceP((O)) isoftype ((0)), (>..X((O)).X((O))(X( 0)))(P((O))) is a formula.
Only X(O) is free.
4 [(>..X((O)) .X((O))(X(0)))(P((O)))::) X( 0)(x 0 )] is a formula. The only free
variable occurrences are those of X( 0) and x 0 .
5 (VX( 0))[(>..X((O)).X((O))(X(0)))(P((O))) : => X( 0)(x0 )] is a formula. The
only free variable occurrence is that of x 0 .
6 (>..x 0 .(\fX(0))[(>..X((O)).X((O))(X(0)))(P((O))) : => X( 0)(x0 )]) is a predicate abstract. It has no free variable occurrences, and is of type (0).
The type machinery is needed to guarantee that what is written is wellformed. Now that the exercise above has been gone through, I will
display the predicate abstract without superscripts, as
(>..x.('v'X)[(>..X.X(X))(P) : => X(x)]),
DEFINITION
Note that since an atomic formula can involve terms containing predicate abstracts which, in turn, involve other formulas, the degree of an
atomic formula need not be 0, as in the first-order case.
2.
Substitutions
Formulas can contain free variables, and terms that are very complex
can be substituted for them. The notion of substitution is a fundamental
one, and this section is devoted to it. In a general way, I follow the
treatment in [Fit96].
1.11 (SUBSTITUTION) A substitution is a mapping from
the set of variables to the set of terms of L( C) such that variables of
type t map to terms of type t.
DEFINITION
DEFINITION
DEFINITION
1 Au
=A
CLASSICAL LOGIC-SYNTAX
4 [<P]u = [<Pu].
5 (<P 1\ w)u = (<Pu 1\ \llu).
6 [('v'a)<P]u
= ('v'a)[<Pua]
= 70"2.
2 If 0"1 and u2 agree on the free variables of the formula <P then <Pu1 =
<Pu2.
Not all substitutions are appropriate in all settings; some do not properly respect the role of bound variables, in the sense that they may replace a free occurrence of a variable in a formula with another variable
that is "captured" by a quantifier or predicate abstract of the formula.
The substitutions that are acceptable are called free substitutions. These
play a significant role throughout what follows.
1.16 (FREE SUBSTITUTION)
The following characterizes when a substitution u is free for a formula or term.
DEFINITION
is free for (>.a1, ... , an.<P) if O"a1, ... ,an is free for <P, and if ,6 is
any free variable of (>.a1, ... , an.<P) then ,6u does not contain any of
a1, ... , an free.
O"
10
4a
5
is free for (<P 1\ 'l1) if a is free for <P and a is free for 'l1.
a is free for (Ya)<P if CJ01 is free for <P, and if f3 is a free variable of
(Ya)<P then f3a does not contain a free.
DEFINITION
It is not generally the case that T(a1a2) = (Ta1)a2, for terms T, and
similarly for formulas. But it is when appropriate freeness conditions
are imposed.
1.18 Substitution is
cumstances.
THEOREM
<~compositional"
1 If a1 is free for the formula <P, and a2 is free for the formula <Pa1,
then (<Pa1)a2 = <P(aw2).
2 If a1 is free for the term
(Ta1)a2
T,
TCJI,
then
= T(a1a2).
Exercises
EXERCISE 2.1 Prove Proposition 1.15 by induction on degree. Conclude
that if <P is a sentence then <Pa = <P for every substitution a.
Chapter 2
CLASSICAL LOGIC-SEMANTICS
1.
Classical Models
Defining the semantics of any higher-order logic is relatively complicated. Since modalities add special complexities, it is fortunate I can
discuss underlying classical issues before bringing them into the picture.
In this Chapter the "real" notion of higher-order model is defined first,
and truth in them is characterized. Then Henkin's modification of these
models is considered-sometimes these are called general models-as
well as a non-extensional version of them.
I don't want just syntactic objects, terms, to have types. I want
sets and relations to have them too. After all, we think of terms as
designating sets and relations, and we want type information to move
back and forth between syntactic object and its designation.
DEFINITION
1 [0,8] =
For
s.
2 [ (t1, ... , tn), S] is the collection of all subsets of [t1, S] X x [tn, S].
12
tion in the model to each constant symbol of the language. But that
denotation must be consistent with type information.
DEFINITION 2.2 (CLASSICAL MODEL) A higher-order classical model
for L(C) is a structure M = (D,I), where D is a non-empty set called
the domain of the model, and I is a mapping, the interpretation, meeting
the following conditions.
1 If A is a constant symbol of L(C) of type t, I(A) E [t, D].
2 If = is the equality constant symbol of type (t, t) then I(=) is the
equality relation on [t, D].
2.
Truth in a Model
CLASSICAL LOGIC-SEMANTICS
13
DEFINITION
1 For terms T, T1, ... ,Tn, M if-v T(T1, ... ,Tn) provided
((v * I) (T1) , .. . , (v * I) (Tn)) E (v * I) (T) .
4M
1\
DEFINITION
4 M if-v (Va)q, if M if-v q,[a/0] for every object 0 of the same type as
a.
Defined symbols like~ and 3 have their expected behavior, which are
explicitly stated below. Alternately, this can be considered an extension
of the definition above.
5 M if-v q, V 'lt if M if-v q, or M if-v 'lt.
6 M if-v q,
14
As in first-order logic, if <I> has no free variables, M lf--v <I> holds for
some v if and only if it holds for every v. Thus for sentences (closed
formulas), truth in a model does not depend on a choice of valuation.
2.8 (VALIDITY, SATISFIABILITY, CONSEQUENCE)
be a formula and S be a set of formulas.
DEFINITION
Let <l>
1 <I> is valid if M lf--v <I> for every classical model M and valuation v.
2 S is satisfiable if there is some model M and some valuation v such
that M If-v <p for every <p E S.
3 <I> is a consequence of S provided, for every model M and every
valuation v, if M lf--v <p for all <pES, then M lf--v <I>.
The definitions above are of some complexity. Here is an example to
help clarify their workings.
2.9 This example shows a formula that is valid and involves
equality. In it, cis a constant symbol of type 0.
The expression (AX.(:lx)X(x)) is a predicate abstract of type ((0)),
where X is of type (0) and x is of type 0. Intuitively it is the "being
instantiated" predicate. Likewise the expression (>.x.x = c) is a predicate
abstract of type (0), where x and care of type 0. Intuitively this is the
"being c" predicate. Since this predicate is, in fact, instantiated (by
whatever c designates), the fir:;;t predicate abstract correctly applies to
it. That is, one should have the validity of the following.
EXAMPLE
(>.X.(3x)X(x))((>.x.x =c))
(2.1)
I now verify this validity. Suppose there is a model M = (V, I). I show
the formula is true in M with respect to an arbitrary valuation v. To do
this, I investigate the behavior, in M, of parts of the formula, building
up to the whole thing.
First, recalling that the interpretation of an equality symbol is by the
equality relation of the appropriate type, we have the following.
(v *I)((>.x.x =c))= {o I M lf--v (x = c)[x/o]}
= {o I o =I(c)}
= {I(c)}
15
CLASSICAL LOGIC-SEMANTICS
E {0
I 0 # 0}.
You might try verifying, in a similar way, the validity of the following .
(.AX.(3x)X(x))((.Ax.(x = x)))
3.
Problems
First-order classical logic has many nice features that do not carry
over to higher-order versions. This is well-known, and partly accounts
for the general emphasis on first-order. I sketch a few of the higher-order
problems here.
3.1
Compactness
The compactness theorem for first-order logic says a set of formulas is satisfiable if every finite subset is. This is a fundamental tool
for the construction of models of various kinds-non-standard models
of analysis, for instance. The higher-order analog does not hold, and
counter-examples are easy to come by. Here is one.
The Dedekind characterization of infinity is: a set is infinite if it can be
put into a 1-1 correspondence with a proper subset. Consequently, a set
is finite if any 1-1 mapping from it to itself can not be to a proper subset,
i.e. must be onto. This can be said easily, as a second-order formula.
Since function symbols are not available, I make do with relation symbols
in the usual way-the following formula is true in a model if and only if
the domain of the model is finite.
(\fX)[(function(X)
1\
one-one(X)) :J onto(X)]
for
for
for
(\fx)(3y)(\fz)[X(x, z) = (z = y)]
(\fx)(\fy)(\fz){[X(x, z) 1\ X(y, z)] :J (x = y)}
(\fy)(3x)X(x, y)
(2.2)
16
A2
A3
ol y
abbrevi-
= (:3xl)(:Jx2)[xl ol x2]
=
3.2
Strong Completeness
3.3
Weak Completeness
CLASSICAL LOGIC-SEMANTICS
17
1\
one-one(S)
1\
0-exclude(S)
1\
induction(S)
3.4
And Worse
18
CLASSICAL LOGIC-SEMANTICS
19
4.
Henkin Models
As we saw in the previous section, higher-order classical logic is difficult to work with. Indeed, difficulties already appear at the second-order
level. Not only does it lack a complete proof procedure, but the very
notion of validity touches on profound foundational issues. Nonetheless,
there are several sound proof procedures for the logic-any formula that
has a proof must be valid, though not every valid formula will have a
proof. So, there are certainly fragments of higher-order logic that we
can hope to make use of.
In a sense, too many formulas of higher-order classical logic are valid,
so no proof procedure can be adequate to prove them all. Henkin broadened the notion of higher-order model [Hen50] in a natural way, which
will be described shortly. With this broader notion there are more
models, hence fewer valid formulas, since there are more candidates for
counter-models. Henkin called his extension of the semantics general
models-! will call them Henkin models.
Henkin's idea seems straightforward, after years of getting used to
it. Given a domain V, a universal quantifier whose variable is of type
0, (Vx), ranges over the members of V. If we have a universal quantifier, (VX), whose variable is of type (0), it ranges over the collection of
properties of V, or equivalently, over the subsets of V. The problem of
just what subsets an infinite set has is actually a deep one. The independence of Cantor's continuum hypothesis is one manifestation of this
problem. Methods for establishing consistency and independence results
in set theory can be used to produce models with considerable variation
in the powerset of an infinite set. Henkin essentially said that, instead
of trying to work with all subsets of V, we should work with enough of
them, that is, we should take (VX) as ranging over some collection of
subsets of V, not necessarily all of them, but containing enough to satisfy
natural closure properties. Think of the collection as being intermediate
between all subsets and all definable subsets.
In a higher-order model as defined earlier, there is a domain, V, and
this determines the range of quantification for each type. Specifically,
20
4'. Let M = ('It, I) be a Henkin frame and let o:t be a variable of type t.
M lf-v (Vo:t)<I> if M lf-v <I>[o:t ;ot] for every ot E 1t(t), or equivalently,
if M lf-v' <P for every at-variant v' of v such that v'(o:t)
7t(t).
The revised version of item 4 above says that quantifiers of type t range
over just 1t(t) and not over all objects of type t.
CLASSICAL LOGIC-SEMANTICS
21
DEFINITION
(v*I*A)((.Xa1, ...
,an.~))
=A(v,(.Xa1, ...
,an.~)).
22
1 For terms T, T!, ... ,Tn, M lf-v,A T(TI, ... ,Tn) provided
((v *I* A)(TI), ... , (v *I* A)(Tn)) E (v *I* A)(T).
2 M lf-v,A --,q> if it is not the case that M lf-v,A <1>.
3M lf-v,A
4M
<1> 1\ \]i
if M lf-v,A
<1>
and M lf-v,A W.
For a given Henkin frame M it may be the case that no proper abstraction designation function exists. But, if one does exist it must be
unique.
2.18 Let M = (1t,I) be a Henkin frame and let both A
and A' be proper abstraction designation functions, with respect to L( C).
Then A= A'.
PROPOSITION
M lf-v,A
<1>
(v*I*A)(T)
M lf-v,A'
<1>
= (v*I*A')(T)
(2.3)
(2.4)
CLASSICAL LOGIC-SEMANTICS
23
Suppose (2.3) and (2.4) are known for formulas and terms whose degree is < k. It will be shown they hold for degree k too, beginning with
(2.4).
Suppose T is a term of degree k. Since k could be 0, T could be a
constant symbol or a variable. If it is a constant symbol, (v *I* A) (T) =
I( T) = (v *I * A') (T). Similarly if T is a variable. Finally, T could be a
predicate abstract, (>.a1, ... , an.<I>), in which case <I> must be a formula
of degree < k, so using the induction hypothesis with (2.3) we have
24
will say the Henkin frame M itself is a Henkin model, and write M lf-v <P
forM lf-v,A <P.
H(t)
DEFINITION
(H,I)
Since standard models are among the Henkin models, any formula
that is true in all Henkin models must be true in all standard models
as well. But there is the possibility (a fact, as it happens) that there
are formulas true in all standard models that are not true in all Henkin
models. That is, the set of Henkin-valid formulas (Definition 2.29) is a
subset of the set of valid formulas (Definition 2.8), and in fact turns out
to be a proper subset. By decreasing the set of validities, it opens up the
possibility (again a fact, as it happens) that there may be a complete
proof procedure with respect to this more restricted version of validity.
5.
CLASSICAL LOGIC-SEMANTICS
25
X1t(tn)
In addition, M is normal ifE(I(=(t,t))) is the equality relation on 1-l(t)
for each type t.
Much of this definition is similar to that of Henkin frame. The members of 1-l(t) are the objects of type t (which now need not be sets). The
new item is the mapping E. Think of E(O) as the extension of the object
0.
Generalized Henkin models are built out of generalized Henkin frames.
Much of the machinery is almost identical with that for Henkin models,
26
but there are curious twists, so things are presented in detail, rather
than just referring to earlier definitions. The definition of valuation is
the same as before.
DEFINITION 2.22 (VALUATION) The function vis a valuation in a generalized Henkin frame M = ('H, I,) if v maps each variable of type t
to some member of1i(t).
Next, just as with Henkin models, a function is needed that provides
designations for predicate abstracts, then later we can require that it
give us the "right" values. The wording is the same as before.
DEFINITION 2.23 (ABSTRACTION DESIGNATION FUNCTION) A is an
abstraction designation function in the generalized Henkin frame M =
('H, I,), with respect to the language L( C) provided, for each valuation
v in M, and for each predicate abstract (Aal, ... , an.<l?) of L(C) of type
t, A(v, (Aal, ... ,an.<I>)) is some member of1i(t).
Term denotation is like before-terms designate objects in the Henkin
domains.
DEFINITION 2.24 (DENOTATION OF A TERM) Let M = ('H,I,) be a
generalized Henkin frame, let v be a valuation, and let A be an abstraction designation function. A mapping, (v *I* A), is defined assigning
to each term of L( C) a denotation for that term.
1 If A is a constant symbol of L(C) then (v *I* A)(A) =I( A).
2 If a is a variable then (v *I* A)(a) = v(a).
3 If (Aal, ... , an. <I>) is a predicate abstract of L( C), then
27
CLASSICAL LOGIC-SEMANTICS
4M
ot E 7-l(t).
In item 1 above, T(Tl, ... , Tn) is true if the designation of (TI, ... , Tn)
is in the extension of the designation of T. For Henkin frames, we were
dealing with sets, and extensions were for free. Now we are dealing with
arbitrary objects, and we must explicitly invoke the extension function
.
I am about to impose a "correctness" requirement, analogous to Definition 2.16, but now there are three parts. The first part is similar to
that for Henkin models, except that the extension function is invoked.
The other parts need some comment. Suppose we have two predicate
abstracts (>.a1, . . . , an. <P) and (>.a1, . . . , an.\]!). In a Henkin model, if <P
and \]! are equivalent formulas, they will be true of the same objects and
so the two predicate abstracts will designate the same thing, since they
have the same extensions. But now we are explicitly allowing predicate
abstracts having the same extension to denote different objects. Still,
we don't want the designation of objects by predicate abstracts to be
entirely arbitrary-! will require equi-designation under circumstances
of "structural similarity."
DEFINITION 2.26 Let M be a generalized Henkin frame (or a Henkin
frame), and let A be an abstraction designation function in it. For each
valuation v and substitution a-, define a new valuation vu by:
Let M = (7-l, I,) be a generalized Henkin frame and let A be an abstraction designation function in it, with respect to L( C). A is proper
provided, for each predicate abstract (>.a1, ... , an.<P) we have
1 ((v*I*A)((>.ab ,an.<P))) =
{ (01, ... , On) I M lf-v,A <P[ai/01, ... , an/On]}
2 If v and w agree on the free variables of (>.a1, ... , an.<P) then
28
DEFINITION
DEFINITION
1 <I> is valid in generalized Henkin models if M lf-v,A <I> for every generalized Henkin model (M,A) for L(C) and valuation v.
2 S is satisfiable in a generalized Henkin model (M, A) for L( C) if
there is some valuation v such that M If-v,A <p for every <p E S.
3 <I> is a generalized Henkin consequence of S provided, for every generalized Henkin model (M,A) for L(C) and every valuation v, if
M lf-v,A <p for all <p E S, then M lf-v,A <I>.
Similar terminology is used when confining things to generalized Henkin
models that are normal, or to Henkin models themselves.
CLASSICAL LOGIC-SEMANTICS
6.
29
6.1
2.30
Let (M, A} be either a Henkin model or a generalized Henkin model, and let v and w be valuations.
PROPOSITION
(v*I*A)(T)
(w*I*A)(T).
Next I state a result that will be used in the next Chapter to establish
the soundness of the tableau system.
2.31
Let (M, A} be either a Henkin model or a generalized Henkin model. For any substitution a and valuation v:
PROPOSITION
Once again I omit the proof, and refer you to Proposition 4.16 for
a similar argument. (For generalized Henkin models, part 3 of Definition 2.27 is needed.)
6.2
Extensional Models
Among Henkin models the standard ones correspond to "true" higherorder models. A similar phenomenon occurs here-among the generalized Henkin models certain ones correspond to Henkin models.
30
is.
6.3
Language Extensions
Part of the definition of (generalized) Henkin model is that each predicate abstract must have an interpretation that is an object with the
"right" extension. But what predicate abstracts there are depends on
what the language is. Given a language L( C), one would expect models to depend on the collection of constants-members of C-which the
interpretation function, I, deals with. One would not expect the choice
of free variables of L(C) to matter, but this is not entirely clear, since
predicatr: abstracts can involve free variables. It is important to know
that the choice of free variables, in fact, does not matter, since the machinery of tableau proofs will require the addition of new free variables
to the language.
In what follows, L( C) is the basic language, and L + (C) is like L( C),
with new variables added, but with the understanding that these new
CLASSICAL LOGIC-SEMANTICS
31
PROPOSITION
32
Exercises
EXERCISE 6.1 Give a proof of Proposition 2.30.
EXERCISE 6.2 Give a proof of Proposition 2.31.
EXERCISE 6.3 Supply details for a proof that each generalized Henkin
frame that is extensional is isomorphic to a Henkin frame.
Chapter 3
Several varieties of proof procedures have been developed for firstorder classical logic. Among them the semantic tableau procedure has
a considerable attraction, [Smu68, Fit96]. It is intuitive, close to the
intended semantics, and is automatable. For higher-order classical logic,
semantic tableaus are not as often seen-most treatments in the literature are axiomatic. Among the notable exceptions are [Tol75, Smi93,
Koh95, GilOl]. In fact, semantic tableaus retain much of their first-order
ability to charm, and they are what I present here. Automatability becomes more problematic, however, for reasons that will become clear
as we proceed. Consequently the presentation should be thought of as
meant for human use, and intelligence in the construction of proofs is
expected.
This chapter examines what I call a basic tableau system; rules are
lifted from those of first-order classical logic, and two straightforward
rules for predicate abstracts are added. It is a higher-order version of the
second-order system given in [Tol75]. I will show it corresponds to the
generalized Henkin models from Section 5 of Chapter 2. In Chapters 5
and 6 I make additions to the system to expand its class of theorems
and narrow its semantics to Henkin models.
1.
A Different Language
34
35
2.
Basic Tableaus
I now present the basic tableau system. It does not contain machinery
for dealing with equality-that comes in Chapter 5. The rules come
from [Tol75], where they were given for second-order logic. These rules,
in turn, trace back to the sequent-style higher-order rules of [Pra68] and
[Tak67].
All tableau proofs are proofs of sentences-closed formulas-of L( C).
A tableau proof of q> is a tree that has --,q> at its root, grounded formulas
of L + (C) at all nodes, is constructed following certain branch extension
rules to be given below, and is closed, which means it embodies a contradiction. Such a tree intuitively says --,q> cannot happen, and so q> is
valid.
The branch extension rules for propositional connectives are quite
straightforward and well-known. Here they are, including rules for various defined connectives.
DEFINITION
XI\Y
(X V Y)
-.x
--,y
--,y
(X
=:J
Y)
For the conjunctive rules, if the formula above the line appears on a
branch of a tableau, the items below the line may be added to the end of
the branch. The rule for double negation is of the same nature, except
that only a single added item is involved.
DEFINITION
3.4
x
X
Next come the disjunctive rules. For these, if the formula above the
line appears on a tableau branch, the end node can have two children
added, labeled respectively with the two items shown below the line in
the rule. In this case one says there is tableau branching.
36
DEFINITION
3.5
(DISJUNCTIVE RULES)
XVY
XIY
(X A Y)
XiY
(X = Y)
(X ~ Y) I (Y ~ X)
DEFINITION
(:Jat)<P( at)
<P(pt)
(Vat)<P( at)
<I>(pt)
The rules above embody the familiar notion of existential instantiation. Since the convention is that parameters are never quantified or >.bound, we don't have to worry about accidental variable capture. More
precisely, in the rules above, the substitution {at jpt} is free for the formula <P(at).
The universal rules are somewhat more straightforward. Once again, .
note that in them the substitution {at /Tt} is free for the formula <I>( at).
DEFINITION
(3at)<I>( at)
<l>(Tt)
37
3.
Tableau Examples
Tableaus for first-order classical logic are well-known, but the abstraction rules of the previous section are not as widely familiar. I give
38
(3.1)
This formulation contains equality. I have not given rules for equality
yet, so I give an alternative formulation that does not involve it.
('v'R)(3X)('v'x)(3y){[R(x, y)
1\
X(y)] V [R(x, y)
1\
X(y)]}
(3.2)
O'l
M
?,p)
{Ax.~P(x,~7
os
-.P(p,p)
&S
C,)
-.-.P(p,p)
11.
-.-.(>.x.-.P(x, x))(p) 9.
(>.x.-.P(x, x))(p) 10.
8.
-.(>.x.-.P(x, x))(p)
-.-.P(p,p) 15.
12.
-.-.P(p,p) 13.
-.P(p,p) 16.
-.(>.x.-.P(x, x))(p)
14.
~
~
0
~
t3
Figure 3.1.
3.
40
('v'x)R(x, x)
(3.3)
Actually, the implication from left to right is straightforward-! supply a tableau proof from right to left.
(\ly)[7
7.
(Az.R(p, z))(p)
R(p,p) 13.
8.
41
(VP)(VQ)[(Vx)(P(x) :J Q(x)) :J
(Vx)(F(P, x) :J F(Q, x))] :J (3S)('v'x)(F(S, x)
= S(x))
(3.4)
I leave the construction of a tableau proof of this to you as an exercise, but I give the following hint. Let <P(P, x) abbreviate the formula
(Vy)(F(P,y) :J P(y)) :J P(x). An appropriate term to consider during
a universal rule application is: (>..x.(VP)<P(P, x)).
A comment on the hint above. Rewriting ('v'y) (F( P, y) :J P(y)) using
conventional function notation: it says F(P) ~ P. Then <P(P, x) says
that x belongs to a set P if P meets the condition F(P) ~ P. Then
further, (VP)<P(P, x) says that x is in n{P I F(P) ~ P}. So finally,
(>..x.(VP)<P(P, x)) represents the set n{P I F(P) ~ P} itself. In the
most common proof of the Knaster-Tarski theorem, one proceeds by
showing this set, in fact, is a fixed point of F.
Example 3.14 once again illustrates a fundamental point about higherorder tableaus. They mechanize routine steps, but do not substitute
for mathematical insight. The choice of which predicate abstract to use
during an application of a universal rule really contains, in distilled form,
the essence of a standard mathematical argument.
The problem of what choice to make when instantiating a universal
quantifier also arises in first-order logic, but there is a way around it--one
uses free variables when instantiating, then one determines later which
values to choose for them [Fit96]. This last step, picking values, involves
unification, the solving of equations involving first-order terms. There
are several unification algorithms to do this, all of which accomplish
42
the following: given two terms, if there is a choice of values for their
free variables that makes the terms identical, the algorithm finds the
most general such choice; and if the terms cannot be made identical, the
algorithm reports this fact. Unification is at the heart of every first-order
theorem prover.
If we attempt a similar strategy in automating higher-order logic, we
immediately run into an obstacle at this point. The problem of unification for higher-order terms is undecidable! This was shown for thirdorder terms in [Hue73], and improved to show unification for secondorder terms is already undecidable, in [Gol81]. This does not mean
the situation is completely hopeless. While first-order unification is decidable, and second-order is not, still there is a kind of semi-decision
procedure, [Hue75]. Two free-variable tableau systems for higher-order
classical logic, using unification, are presented in [Koh95]. The use of
higher-order unification in this way traces back to resolution work of
[And71] and [Hue72]. But finally, technical issues aside, we always come
back to the observation made above: the choice of predicate abstract to
use in instantiating a universally quantified formula often embodies the
mathematical "essence" of a proof. Too much should not be expected
from the purely mechanical.
Exercises
EXERCISE 3.1 Extending the ideas of Example 3.13, give tableau proofs
of the following.
1 (symmetry)
('v'x)('v'y)[R(x,y) :J R(y,x)] =
('v'P)('v'x){(3y)[R(x, y) 1\ ('v'z)(R(y, z) :J P(z))] :J P(x)}
2 (transitivity)
Chapter 4
This chapter contains a proof that the basic tableau rules are sound
and complete with respect to generalized Henkin models. Soundness is
by the "usual" argument, is straightforward, and is what I begin with.
Completeness is something else altogether. For that I use the ideas
developed simultaneously in [Tak67, Pra68], where they were applied to
give a non-constructive proof of a cut elimination theorem.
1.
Soundness
The second key fact takes more work to prove, but the work is spread
over several cases, each of which is rather simple.
43
M. Fitting, Types, Tableaus, and Gdel's God
Kluwer Academic Publishers 2002
44
LEMMA
45
(v*I*A)(f3) =v({3).
46
(v *I* A)(Tn) But, by items 1 and 2 above, vu itself is this a1, ... ,
an-variant of v. We thus have
that is,
Proof Suppose <P has a tableau proof, but is not true in every generalized Henkin model with respect to L( C)-I derive a contradiction.
Since <P is not true in every generalized Henkin model with respect to
L( C), {-.<P} is satisfiable, and by Proposition 2.33, is so in a generalized
Henkin model with respect to L + (C). A tableau proof of <P begins with
a tableau consisting of a single branch, containing the single formula
--.<P, so this must be a satisfiable tableau. As we apply branch extension
rules, we continue to get satisfiable tableaus, by Lemma 4.3. Since <P
is provable, we can get a closed tableau. Hence there must be a closed,
satisfiable tableau, which is impossible according to Lemma 4.2.
Essentially the same argument also establishes the following.
4.5 Let S be a set of sentences and <P be a single sentence
of L(C). If <P has a tableau derivation from S, then <P is a generalized
Henkin consequence of S.
THEOREM
2.
Completeness
The proof of completeness, for basic tableaus, with respect to generalized Henkin models, is of considerable intricacy. It is spread over
several subsections, each devoted to a single aspect of it. All the basic ideas go back to [Tak67, Pra68], where they were used to establish
non-constructively a cut-elimination theorem for higher-order Gentzen
systems. I also use aspects of the (second-order) presentation of [Tol75],
in particular the central goal, for us, is to prove that something called
47
2.1
Hintikka Sets
DEFINITION
48
5 Universal Cases.
(a) If (3at)<P( at) E H then <P(pt) E H for at least one parameter pt.
(b) If (Vat)<P(at) E H then <P(pt) E H for at least one parameter
pt.
7 Abstraction Cases.
2.2
Pseudo-Models
49
50
Roughly the idea is, any possible value for T should have in its extension
all those things the Hintikka set H requires, and should omit all the
things H forbids. Any entity that meets these conditions will serve as a
possible value. Clearly each possible value of a grounded term of type
t, relative to a Hintikka set H, is an entity of type t. Item 1 of the
definition of Hintikka set ensures that part 2 above is meaningful.
Now that we have the notion of possible value, Henkin domains for
our pseudo-models can be defined.
4.10 (RELATIVE HENKIN DOMAINS) Let H be a Hintikka
set. A mapping, 'HH is defined, from types to entities, as follows. For
each type t, 'HH(t) is the set of all entities of type t that are possible
values relative to H.
DEFINITION
The languages L( C) and L +(C) are allowed to contain constant symbols. How to interpret these is rather arbitrary, within broad limits.
4.8.
-v
-v
51
4M
1\
ot E 1-lH(t).
5 Let T =
The definition above has the usual complex recursive structure, with
truth at the atomic level needing (v * I * AH) and hence AH, and the
characterization of AH itself needing the notion of truth for formulas.
Of course what makes it work is the fact that, in every case, behavior
of some construct on a formula or term requires constructs involving
simpler formulas and terms.
The key point is, why is this called a pseudo-model, and not simply a
model? The answer is, we have the characterization of the abstraction
designation function backwards here. In Chapter 2 we assumed we had a
function A that mapped valuations and abstracts to members of Henkin
52
Exercises
EXERCISE 2.1 Show that if entity E is a possible value, then E must
be a possible value of T(E).
2.3
In this subsection valuations and substitutions are shown to be wellbehaved with respect to pseudo-models. It should be noted again that
valuations always mean valuations in a pseudo-model-they map variables to members of Henkin domains, to possible values. They do not
map to arbitrary entities. The proofs below are rather technical, so I
begin with the statements of the two Propositions to be established, after which their proofs are given, broken into a number of Lemmas. On a
first reading you might want to just read the Propositions and skip over
the proofs.
The first item should be compared with Proposition 2.30.
PROPOSITION 4.15 Let H be a Hintikka set, let M = ('HH,I,E) be
a generalized frame relative to H, and let (M, AH) be a pseudo-model
relative to H. Also let v and w be valuations.
1 If v and w agree on the free variables of the term T
(v*I*AH)(T) = (w*I*AH)(T).
2 If v and w agree on the free variables of the formula <I>
M II-v,AH <I> ::::::} M II-w,AH <I>
53
PROPOSITION
Now I turn to the proofs, which are given in considerable detail since
these results are critical to the completeness argument, and I want the
reasoning on record. On a first reading, skip the proofs and move on to
the next section.
Proof of 4.15 Suppose the result is known for terms and formulas
whose degree is < k. I show the result is also true for those of degree k
itself, beginning with terms.
Assume Tis a term of degree k, and v and w agree on the free variables
and parameters of T. If k happens to be 0, T is a constant symbol,
variable, or parameter. In these cases the result is immediate.
Now suppose k =/= 0, and soT= (.Xa1, ... , an.<P), where <Pis of degree
< k. Say (v *T*AH)((.Xal, ,an.<P)) = AH(v,(Aai, ,an.<P)) =
(a,S) and (w*T*AH)((Aal, ,an.<P)) =AH(w,(.Xal, ,an.<P)) =
(a', S'). We must show a= a' and S = S'.
Suppose a is a variable or parameter that occurs free in T. Then
air = T(v(a)) = T(w(a)) = a~, using the assumption that v and
w must agree on a. By definition, a = T'v and a' = T~, and the
substitutions 'v and ~ agree on the free variables of T, so a = a' by
Proposition 1.15.
Next we suppose (E1, ... , En) E S, and so M lf--v,AH <P[al/ E1, ... ,
an/ En] Since v and w agree on the free variables and parameters of
(.Xa1, ... , an.<P), they agree on the free variables and parameters of <P,
other than a1, ... , an. Then by the induction hypothesis, M lf--w,AH
<P[al/ E1, ... , an/ En], and it follows that (E1, ... , En) E S'. Thus S ~
S'. A similar argument shows S' ~ S.
This completes the induction step for terms, and I turn next to formulas. Suppose <P is of degree k and v and w agree on the free variables and
parameters of <P. By the induction hypothesis, we have the Proposition
for terms and formulas of degree< k, and by what was just shown, we
also have it for terms of degree k itself. Now we have several cases.
Suppose <Pis atomic, To(TI, ... , Tn), where each Ti must be of degree
~ k. Then, using the induction hypothesis and what was just proved,
54
::>
::>
::>
1 If w is an 0:1, ... , O:n variant of v then wu"'l ,a.n is an 0:1, ... , O:n
variant of vu.
2 Conversely, if u is an 0:1, ... , O:n variant of vu then u = wu"'l ,... "'n
for some 0:1, ... , O:n variant w of v.
3 vu<>1, ... ,<>n (o:i) = v(o:i), fori= 1, ... , n.
Proof
Part 1. Suppose w is an 0:1, ... , O:n variant of v. Let {3 be a variable or
parameter other than 0:1, ... , O:n. It must be shown that wu<>1>-- ,<>n ({3) =
va({3). Here are the steps; the reasons follow.
= (w * I * AH) ({3(})
(4.1)
(4.2)
= (v*I*AH)(f3(})
(4.3)
vu ({3)
(4.4)
Above, (4.1) is the definition of wu<>l, ... ,<>n, and (4.2) is because {3 is
different from 0:1, ... , O:n. Also (4.4) follows from (4.3) by the definition
of vu. The key item is the equality of (4.2) and (4.3), and for this it is
enough to show v and w agree on the free variables and parameters of
{3(}, and then appeal to Proposition 4.15. The argument for this follows.
If {3 does not occur free in (..Xo:1, ... , O:n.<P), {3(} = {3 by assumption.
Also we are assuming {3 is different from o: 1, ... , O:n, and v and w agree
on all variables except o: 1, . . . , O:n, so v and w must agree on {3, and
55
hence trivially they agree on the free variables and parameters of {JCJ in
this case.
Now suppose {3 does occur free in (Ao:1, ... , O:n-~). Since CJ is free for
(Ao:1, ... , O:n-~), {JCJ cannot contain any of 0:1, ... , O:n free. Once again
v and w must agree on the free variables and parameters of {JCJ, since v
and w can only differ on 0:1, . . . , O:n.
Part 2. Suppose u is an 0:1, ... , O:n variant of va. Define a valuation
w as follows.
(4.5)
(4.6)
(4.7)
(4.8)
In this, (4.5) is by definition of wa"'l .. "'n. Then (4.6) is because CJa 1 , ... ,an
is the identity on O:i Next, (4.7) is because o:i is a variable, and finally
(4.8) is by definition of w.
Now suppose f3 ::/= 0:1, ... , O:n.
(4.9)
(4.10)
(4.11)
(4.12)
(4.13)
LEMMA
56
Proof Suppose first that T is a predicate abstract. Then by Definition 4.14, (v *I* AH)(T) = AH(v,T) = (T*v,S) for a certain setS,
and so T((v *I* AH)(T)) = T*v. If T is a variable or parameter,
T*v = T(v(T)) by definition of tv", and v(T) = (v *I* AH)(T) by definition of (v * I * AH) again, for variables. If T is a constant symbol,
T*v = T, and also T((v *I* AH)(T)) = T(I(T)) = T because I is an
allowed interpretation.
The proof of Proposition 4.16 is by an induction on degree. Since the
steps are somewhat complex, I have separated the significant parts out,
in the following two Lemmas.
4.19 Let H be a Hintikka set, let M = ('HH, I,) be a generalized frame relative to H, and let (M, AH) be a pseudo-model relative to
H. Assume that for each formula <I> of degree < k, whenever substitution
C5 is free for <I> then
LEMMA
57
Similarly:
where
Part 1, a= a'.
First of all,
(4.14)
(4.15)
Part 2, S = S'.
Using Proposition 1.15 it can be assumed that a is the identity on
variables and parameters that do not occur free in (>.a1, ... , an. <I>).
58
(4.16)
(4.17)
(4.18)
<I>
(4.19)
(4.20)
(4.21)
Above, (4.17) is just (4.16) rewritten. Since a is free for (Ao:~, ... , O:n.<P),
by Definition 1.16, aa 1 , ... ,an must be free for <P, and since <P must be of
degree < k, (4.17) and (4.18) are equal by the hypothesis of the Lemma.
Then (4.18) and (4.19) are equal by part 3 of Lemma 4.17. The equality
of (4.19) and (4.20) follows by parts 1 and 2 of Lemma 4.17. Finally,
(4.21) is (4.20) rewritten.
4.20 Let H be a Hintikka set, let M = (HH,I,t:) be a generalized frame relative to H, and let (M, AH) be a pseudo-model relative
to H. Assume that
LEMMA
(4.22)
for each substitution a that is free for <I>, provided <I> is of degree < k.
Also assume that
= ( va
*I * AH) (T)
(4.23)
for each substitution that is free forT, provided T is of degree ::; k. Then
(4.22} also holds for each formula <P of degree k itself.
59
If <I> is of the form r(r1, ... , Tn), each of the terms r, r1, ... , Tn must
be of degree :=:; k. Then, using hypothesis (4.23) about terms,
M lf-v,AH [(Va)w]a
M lf-v,AH (Va)[Waa]
{:::> M lf-w,AH Waa for every valuation
w that is an a-variant of v
(4.24)
{:::>
{:::>
w that is an a-variant of v
{:::> M lf-u,AH W for every valuation
u that is an a-variant of va
{:::>
(4.25)
(4.26)
M lf-v",AH (Va)w
Proof of 4.16 The proof, of course, is by induction on degree. Suppose the result is known for formulas and terms of degree < k. Then
by Lemma 4.19 the result holds for terms of degree k, and then by
Lemma 4.20 it holds for formulas of degree k as well.
2.4
60
4.21 Let H be a Hintikka set, let M = (HH,I,) be a generalized frame relative to H, and let (M, AH) be a pseudo-model relative
to H. Also let v be a valuation in the pseudo-model.
THEOREM
1 For each term T of L +(C), (v *I* AH) (T) is a possible value for TlJ.
2 For each formula <J> of L+(C), if <J>"V E H then M lf-v,AH <J>.
X .. X
HH(tn)
61
fl. S.
<I>w E H.
<I> must be of lower degree than (>.a1, ... ,an.<I>), that is, k, so the induction hypothesis applies and
M 11-w,AH <I>.
Then (w(a1), ... , w(an)) E S, so (E1, ... , En) E S, which is what we
wanted.
This concludes the induction step for terms.
Part 2. Let <I> be a formula of degree k. By the induction hypothesis
the result holds for formulas and terms of degree < k, and by part 1 of
the proof it also holds for terms of degree k. Now we have several cases,
depending on the form of <I>. I only present a few of them.
Suppose <I> is 7o(71, ... , 7n) and [To(71, ... , 7n)]tv E H. That is,
62
and so
2.5
4. 23 Let H be a Hintikka set, let M = (1-lH, I,) be a generalized frame relative to H, and let (M, AH) be a pseudo-model relative
to H. Then (M, AH) is a generalized Henkin model.
THEOREM
Proof We need that AH is an abstraction designation function, Definition 2.23. Specifically, we need that it maps predicate abstracts to
63
Proof By Corollary 4.22 and the Theorem above. (Recall, it was shown
in Section 6 that a choice between L( C) and L + (C) was not significant
when considering models for formulas from the language L(C).)
2.6
Completeness At Last
64
Proof The set of grounded formulas of L + (C) is countable; let 'lT 1, 'lT 2,
W3, ... be an enumeration of all of them. Also, let PI, P2, P3, ... be an
enumeration of all parameters of L+(C) of all types. Now we construct
a sequence of sets of formulas. Each set in the sequence will meet two
conditions: it is consistent, and infinitely many parameters of each type
do not appear in it. Here is the construction.
Let 8 0 = S. This is consistent by hypothesis, and contains no parameters at all, so both of the conditions are met.
Suppose Sn has been defined, and the conditions are met.
1 If Sn U {'l'n+I} is not consistent, let Sn+l = Sn.
2 If Sn U {'l'n+I} is consistent, and Wn+l is not an existentially quantified formula or the negation of a universally quantified formula, let
Sn+l = Sn U {'l'n+I}
3 Finally, if Sn U {'l'n+I} is consistent, and Wn+l is (3o:)cll(o:), choose
the first parameter p in the enumeration of parameters, of the same
type as o:, that does not appear in Sn or in (3o:)cll(o:), and set Sn+l =
Sn U {(3o:)cll(o:), cll(p)}. And similarly if Wn+l is (Vo:)cll(o:).
Note that Sn+l meets the conditions again. In case 3, consistency needs
a small argument, which I leave to you.
Finally, let 8 00 be SoUS1 US2U .... I leave to you the easy verification
that 8 00 will be consistent, E-complete, and maximal.
4.29 If S is a set of grounded formulas of L+(C) that is
maximal consistent and E-complete, S is a Hintikka set.
PROPOSITION
65
xk
k.
k + 1.
1 If <I> is valid in generalized Henkin models, <I> has a basic tableau proof.
2 If <I> is a generalized Henkin consequence of S, <I> has a basic tableau
derivation from S.
66
3.
Two of the main results about first-order logic are the Compactness
and the Lowenheim-Skolem theorem. I already noted, in Section 3,
that compactness does not hold for "true" higher-order logic. It is also
easy to verify that the Lowenheim-Skolem theorem does not hold, since
one can write a formula asserting an uncountable object exists. But
things are very different if generalized Henkin models are used, instead
of standard models. Then both theorems hold, just as in the first-order
case. Compactness is easy to verify, now that completeness has been
shown. Lowenheim-Skolem takes more work.
THEOREM 4.31 (COMPACTNESS) LetS be a set of closed formulas of
L( C). If every finite subset of S is satisfiable in some generalized Henkin
model, so is S itself.
67
This Theorem is a version of Gentzen's famous Haputsatz, or cut elimination theorem, for higher-order logic. It is an important result about
classical first-order logic that closed tableaus for SU{ <P} and for SU{ --,cp}
can be constructively converted into one for S. There is no constructive
proof for the higher-order case, but the result can be obtained provided
we are willing to drop constructivity. Such a proof was given in [Pra68]
and in [Tak67], and their argument has appeared here, in disguise, as
a completeness proof. To finish things off I sketch the remaining ideas
involved in a proof of the Theorem.
Proof Suppose there are closed tableaus for S U {<P} and for S U {--,cp}.
Then neither set is satisfiable. It follows that S itself is not satisfiable,
for if there were a generalized Henkin model in which its members were
true, one of <P or --,cp would be true there. It remains to show that the
unsatisfiability of S implies there must be a closed tableau beginning
with S.
Suppose the contrary: there is no closed tableau beginning with S,
so that S is a consistent set. Proposition 4.28 says a consistent set of
L( C) sentences can be extended to a maximal consistent, E-complete
set-the same proof can easily be made to work even if the set contains
parameters, provided it omits infinitely many of them. Since S is finite,
it certainly omits infinitely many parameters, so we can extend it to
a maximal consistent, E-complete set, which must be a Hintikka set.
Corollary 4.24 says Hintikka sets are satisfiable. Since Sis a subset of a
satisfiable set, it too must be satisfiable, but it is not. This contradiction
concludes the proof.
This immediately gives us the following important result.
4.33 (CUT RULE) The addition of the following Cut Rule
to the basic tableau system does not change the class of provable formulas:
at any point split a branch, and add --,cp to one fork, and <P to the other,
where <P can be any grounded formula.
COROLLARY
The way this result is most often used is embodied in the following.
4.34 If <I> has a tableau proof, <I> can be added as a line to
any tableau, without expanding the class of provable formulas.
COROLLARY
68
That is, we have used an application of a cut. Now, on the left branch,
introduce the steps appropriate to close it, which exist because we are
assuming there is a closed tableau for~- This leaves the right branch.
The net effect has been to add ~ to the tableau.
Now, go back through the proof of completeness given earlier. Proposition 4.28 said we could extend a consistent set to a maximal consistent,
E-complete one. Using the work above, it follows that a maximal consistent set must contain either ~ or --,~ for every grounded formula ~.
Since this is the case, each grounded term can, in fact, have only one possible value associated with it. Thus the particular model constructed in
the completeness argument must have countable Henkin domains, since
the family of grounded terms for each type is countable. We thus have
the following.
4.35 (LOWENHEIM-SKOLEM) Let 8 be a set of closed formulas of L(C). If S is satisfiable in some generalized Henkin model, S
is satisfiable in a generalized Henkin model whose domain function 1t
meets the condition that 1t(t) is countable for every type t.
THEOREM
The results above have both good and bad points. It is obviously good
to be able to prove such powerful model-theoretic facts about a logic-it
provides tools for the construction of useful models. The bad side is
that Lindstrom's Theorem says, since the version of higher-order logic
based on generalized Henkin models satisfies the theorems above, it is
simply an equivalent of first-order logic. This does not mean nothing has
been gained. The higher-order formalism is natural for the expression
of things whose translation into first-order versions would be unnatural.
And finally, if a sentence is not provable, it must have a generalized
Henkin counter-model, but if it is provable, it must be true in all generalized Henkin models, and among these are the standard higher-order
models! Thus we have a means of getting at higher-order validities-we
just can't get at all of them this way.
Chapter 5
EQUALITY
The basic tableau rules of Chapter 3 do not give any special role to
equality. It is time to bring it into the picture. This is done by adding
axioms to the tableau system, which has the effect of narrowing things
to normal generalized Henkin models. In addition, some useful derived
tableau rules will be presented.
1.
Adding Equality
Leibniz's principle is that objects are equal just in case they have
the same properties. This principle is most easily embodied in axioms,
rather than in tableau-style rules.
DEFINITION 5.1 (EQUALITY AXIOMS) Each sentence of the following
form is an equality axiom:
I will show that a closed formula <I> of L( C) is valid in normal generalized Henkin models if and only if <I> has a tableau derivation from
EQ. But before that is done I give some handy derived tableau rules,
and examples of their use.
2.
There are two derived rules involving equality that are more "tableaulike" in flavor, and are what I primarily use in constructing tableau
proofs and derivations. I do not know if they can serve as full replacements for the official Equality Axioms, since I have been unable to prove
69
M. Fitting, Types, Tableaus, and Gdel's God
Kluwer Academic Publishers 2002
70
a completeness theorem using them. Nonetheless, the derived rules below are the ones I generally use in practice.
DEFINITION 5.2 (DERIVED REFLEXIVITY RULE) For a grounded term
T of L +(C), at any point in a proof (T = T) may be added to the end of
a tableau branch. Schematically,
(\fa)(\f/3)[(a = /3)
= (\f'Y)(r(a) :J 'Y(/3))]
(\f'Y)(/(T) :J 'Y(T)) 6. (T = T)
1.
2.
7.
DEFINITION
<J?( T1)
(T1=T2)
<J?( T2)
Justification of Derived Substitutivity Rule
Assume T1 and T2
are grounded terms of type t, and <I>(TI) and (T1 = T2) occur on a tableau
71
EQUALITY
branch. I show <P( 72) can be added to the end of the branch.
(71
/~
/
(Aa. <P( a)) (7I)
<1>(71) 11.
9.
10.
1.
(c=c) 4.
(c=c) 5.
In this, 2 is from 1 by an abstract rule; 3 is from 2 by a universal rule;
4 is from 3 by an abstract rule, and 5 is by the derived reflexivity rule.
The next example shows how, by using the derived rules, we can
reverse things and prove a version of the equality axioms.
72
EXAMPLE
of
(Va)(V,B)[(Vr)('y(a)
1(,8))
~(a=
,B)].
--,(AX.--,(X = Q))(P)
--,--,(P = Q) 9.
7.
(AX.--,(X = Q))(Q)
--,( Q = Q) 10.
(Q = Q) 11.
8.
Exercises
2.1 Prove the following characterization of equality-it says
it is the smallest reflexive relation.
EXERCISE
(Vx)(Vy){(x = y)
EXERCISE
= (VR)[(Vz)R(z,z)
R(x,y)]}
= ,8(1))]
EQUALITY
3.
73
THEOREM 5.6 Let <P be a closed formula and let S be a set of closed
formulas of L( C).
1 <P is valid in all normal generalized Henkin models if and only if <P
74
Proof Suppose 01 = 02, that is, 01 =I 02, and say 01 and 02 are of
type (t). In Exercise 2.2 you were asked to give a tableau derivation of
(Va)(Vj])[(a = (3) ::J (V'Y)(a('Y) = {3(/))] from EQ. Then by soundness,
this sentence is valid in (M, A). It follows that (V'Y)(a('Y) = (3(!)) is
also true with respect to any valuation assigning 01 to a and 02 to (3.
From this we immediately get that the sets (01) and (02) must be the
same. A similar argument applies if 01 and 02 are of type (tb ... , tn)
The Lemma above justifies the following. For 0 E 1{ ( (t1, ... , tn)), set
(0) = {(01, ... , On) I (01, ... , On) E (0)}.
We have now created a new generalized Henkin frame M = (?t,I,).
LEMMA
= (1t,I, )
is normal.
= v2
then A( v1, T)
=I
A(v2, r).
Proof For convenience say T just has one free variable, 1; the more general case is treated similarly. From now on I'll write T as r('Y). Assume
v1 = v2, hence in particular, v1 (!) =I v2 (!); I'll show A( v1, r('Y)) =I
A(v2, r('Y)).
Let a and (3 be variables of the same type as 1, that do not occur
in r('Y) (free or bound). Since all members of EQ are true in (M, A),
(Va)(Vf3)[(a = (3) ::J (r(a) = r((3))] is true in it, and hence r(a) = r((3)
is true in (M, A) with respect to any valuation v such that v(a) =I v(f3).
Set w to be a particular valuation such that w(a) = v1(!), w(f3) =
v2(!), and otherwise w is arbitrary. Since v1('Y) =I v2(!), we have
w(a) =I w((3), so by the paragraph above, r(a) = r((3) is true in (M, A)
with respect tow, in other words, A(w,r(a)) =I A(w,r(f3)). Also
75
EQUALITY
A(v~ah},T(a))
A(w, T(a))
=I A(w, T(,6))
=
{:::::::::>
Proof Part 1 follows for variables, constant symbols, and predicate abstracts by definition of v, I, and A respectively.
Part 2 is by an induction on the degree of <P, which I leave to you.
Finally, for part 3 it is necessary to verify the three parts of Definition 2.27. I check part 3 and leave the other parts to you. Let 0' be a
substitution that is free for (>.a1, ... , an.<P), a term which I abbreviate
as T. It must be shown that A(v, TO') = A(W, T). We have the following.
A(vu,T)
= A(vu,T)
76
v(o:o-)
= v(o:o-)
= VU(o:)
=
Exercises
EXERCISE 3.1 Give the details of the proof that
=x is an equivalence
relation.
EXERCISE 3.2 Supply the proof of part 2 of Lemma 5.11.
Chapter 6
EXTENSIONALITY
1.
Adding Extensionality
DEFINITION
2.
Suppose T1 and
T2 are two grounded terms, both of type (t1, ... , tn) At any point in a
tableau construction the end of a branch can be split, with one fork labeled
77
M. Fitting, Types, Tableaus, and Gdel's God
Kluwer Academic Publishers 2002
78
(71 = 72), and the other fork labeled (7I(PI, ... ,Pn)
=
=
4.
~~
P=Q
P(Q)
8.
9.
Exercises
EXERCISE 2.1 Give a proof of formula (3.1) from Example 3.12.
EXERCISE 2.2 Show that the rule contained in Definition 6.2 is, in fact,
a derived rule, using EXT.
79
EXTENSIONALITY
3.
and so this sentence is true in (M, A). Let v be a valuation such that
= 0 and v(f3) = 0'. Then
v(a)
II
MODAL LOGIC
Chapter 7
MODAL LOGIC
SYNTAX AND SEMANTICS
1.
Introduction
83
M. Fitting, Types, Tableaus, and Gdel's God
Kluwer Academic Publishers 2002
84
What is presented here is a modification of the Montague/Gallin approach, in which both extensions and intensions are first class objects.
What are the underlying intuitions? An extensional object will be
much as it was in Part I: a set or relation in the usual sense. The added
construct is that of intensional object, or concept, and this is treated
in the Carnap tradition. A phrase like, say, "the royal family of England," has a meaning, an intension. At any particular moment, that
meaning can be used to determine a particular set of people, constituting its extension. But that extension will vary with time. For other
phrases, there may be different mechanisms for determining extensions
as circumstances vary. The one thing common to all such intensional
phrases is that they, somehow, induce mappings from circumstances to
extensions. Abstracting to the minimum useful structure, in a possible
world model an intensional object will be a function from possible worlds
to extensional objects.
Here is an example using the terminology just introduced. Suppose we
take possible worlds as people, with an 85 accessibility relation~every
person is accessible to every other person. And suppose the ground-level
domain is a bunch of real-world objects. Any one person will classify
some of those objects as being red. Because of differences in vision,
and perhaps culture, this classification may vary from person to person.
Nonetheless, there is a common concept of red, or else communication
would not be possible. We can identify it with the function that maps
each person to the set of objects that person classifies as red. And
similarly for other colors. In addition, each person has a notion of color,
though this too may vary from person to person. One person may think
of ultra-violet as a color, another not. We can think of the color concept
as a mapping from persons to the set of colors for that person. If we
assert that red is a color for a particular person, we mean the red concept
is in the extension of the color concept for that person. The extension
of the red concept for that person plays no role for this purpose.
Sometimes extensions are needed too. Certainly if we ask someone
whether or not some object is red, the extension of the red concept, for
that person, is needed to answer the question. Here is another example
in this direction. Assume the word "tall" has a definite, non-fuzzy,
meaning. Say everybody gets together and votes on which people are
tall, or say there is a tallness czar who decides to whom the adjective
applies. The key point is that the meaning of "tall," even though precise,
drifts with time. Average height of the general population has increased
over the last several generations, so someone who once was considered
tall might not be considered so today.
85
('v'x)OT(x)
0(\lx)T(x)
(7.1)
(7.2)
Formula (7.1) refers to those alive now, and says at some point they
will all be tall. Formula (7.2) refers to those alive at some point in the
future and asserts, of them, that they will be tall. All this is standard;
the problem is with the adjective "tall." Do we mean that at some point
in the future everybody (read either way) will be tall as they use the
word in the future, or as we use the word now? If we interpret things
intensionally, T(x) at a possible world would be understood according
to that world's meaning of tall. There is no way, using the present
machinery, to formalize the assertion that, at some point in the future,
everybody will be tall as we understand the term. But this is what is
most likely meant if someone says, "Someday everybody will be tall."
Here is another example, one that goes the other way. Suppose a
member of the Republican Party, call him R, says, "necessarily the proposed tax cut is a good thing." Suppose we take as the possible worlds
of a model the collection of all Republicans, and assume a sentence is
true at a world if that Republican thinks the sentence true. (We assume Republicans are entirely rational, so we don't have to worry about
contradictory beliefs.) Let us now take OX to mean that every Republican thinks X is the case, which means X is necessary for Republicans.
(Technically, this gives us an 85 modality.) How do we formalize the
sentence above? Let c be a constant symbol whose intended meaning
is, "the proposed tax cut," and let G be a "goodness" predicate. Then
OG(c) seems reasonable as a formalization. What should it mean to say
it is true for R?
86
One possibility is that R means every Republican thinks the tax cut
is good, as R understands the word good. This may not be what was
meant. After all, R might consider something good only if it personally
benefitted him. Another Republican might think something good if it
eventually benefitted the poor. Such a Republican probably would not
think a tax cut good simply because it benefitted R but he might believe
it would eventually benefit the poor, and so would be good in his own
sense. Probably R is saying that every Republican thinks a tax cut is
good, for his own personal reasons. The notion of what is good can
vary from Republican to Republican, provided they all agree that the
proposed tax cut is a good thing. But the mere fact that we can consider
more than one reading tells us that a simple formalization like DG(c) is
not sufficient.
Here will be presented a logic of both intension and extension, of
both sense and reference. In one of the examples above, color is an intensional object. It is a function from persons to sets of concepts like
red, blue, and so on. As such, it is the same function for each person.
The extension of color for a particular person is the color function evaluated at that person, and thus it is a particular set of concepts, such
as red but not infra-red, and so on, quite possibly different from person
to person. We need a logic in which both intensions and extensions are
first-class objects. The machinery for doing this makes for complicated
looking formulas. But I point out, in everyday discourse all the machinery is present but hidden-we infer it from our knowledge of what we
think must have been meant. Formalization naturally requires complex
machinery-it is making explicit what our minds do automatically.
2.
Now begins the formal treatment, starting with the notion of type. I
want it to include the types of classical logic, as defined in Section 1.
I also want it to include the purely intensional types of the Montague
tradition, as given in [Gal75].
DEFINITION 7.1 (TYPE) The notion of a type, extensional and inten-
1 0 is an extensional type.
2 If t1, ... , tn are types, extensional or intensional, (t1, ... , tn) is an
extensional type.
3 If t is an extensional type, jt is an intensional type.
A type is an intensional or an extensional type.
87
7.2
(TERM OF
88
term of L( C) of the intensional type j (t1, ... , tn). Its free variable
occurrences are the free variable occurrences of <I>, except for occurrences of the variables a1, ... , an.
3 If 7 is a term of L(C) of type jt then 17 is a term of type t. It has
the same free variable occurrences that 7 has.
The predicate abstract (Aa1, ... , an. <I>) is of type j(t1, ... , tn) above,
and not of type (t1, ... , tn), essentially because <I> can vary its meaning
from world to world, and so (Aa1, ... , an.<I>) itself is world dependent.
Case 3 above makes use of what may be called an extension-of operator, converting a term of an intensional type to a term of the corresponding extensional one. Continuing with the color example, suppose
r is the intensional notion of red, of type j(O), mapping each person to
that person's set of red objects. Then for a particular person, 1r is that
person's set of red objects-the extension of r for that person, and an
extensional object of type (0).
Of course the symbols j and 1 were chosen to suggest their roles-in
a sense 1 'cancels' j. Nonetheless, 1 is a symbol of the language, while j
occurs in the metalanguage, as part of the typing mechanism.
DEFINITION 7.3 (MODAL FORMULA OF L(C))
mula of L( C) is as follows:
1 If 7 is a term of either type (t1, ... , tn) or type i (t1, ... , tn), and
71, ... , 7n is a sequence of terms of types t1, ... , tn respectively,
then 7(71, ... , 7n) is a formula {atomic) of L(C). The free variable
occurrences in it are the free variable occurrences of 7, 71, ... , 7n
2 If <I> is a formula of L( C) so is -,q>. The free variable occurrences of
-,q> are those of <I>.
3 If <I> and \lT are formulas of L( C) so is (<I> 1\ \lf). The free variable
occurrences of (<I> 1\ \lf) are those of <I> together with those of \lf.
89
the set of colors for a particular person-call the person George. Also
let C be the intensional version of color, mapping each person to that
person's extension of color. cis of type (i(O) ), and Cis of type j(j(O)) I
take both C(r) and c(r) to be atomic formulas. If we ask whether they
are true for George, no matter which formula we use, we are asking if r
is a color for George. But if we ask whether they are true for Natasha,
we are asking different questions. C(r) is true for Natasha if r is a color
for Natasha, while c(r) is true for Natasha if r is a color for George.
No matter which, both c(r) and C(r) make sense, and are considered
well-formed.
I use 0 to abbreviate ...,o..., in the usual way, or I tacitly treat it as
primitive, as is convenient at the time. And of course other propositional
connectives and the existential quantifier will be introduced as needed.
Likewise outer parentheses will often be dropped.
3.
Should quantifiers range over what does exist, or over what might
exist? That is, should they be actualist or possibilist? This is really a
first-order question. A flying horse may or may not exist. In the world
of mythology, such a being does exist. In the present world, it does
not. But the property of being a flying horse does not exist in some
worlds and lack existence in others. In the present world nothing has
the flying-horse property, but that does not mean the property itself is
non-existent. Thus actual/possible existence issues really concern type 0
objects, so the discussion that follows assumes a first-order setting.
As presented in [HC96] and [FM98], the distinction between actualist
and possibilist quantification can be seen to be that between varying
domain modal models and constant domain ones. In a varying domain
modal model, one can think of the domain associated with a world as
what actually exists at that world, and it is this domain that a quantifier ranges over when interpreted at that world. In a constant domain
model one can think of the common domain as representing what does
or could exist, and this is the same from world to world. Of course a
choice between constant and varying domain models makes a substantial difference: both the Barcan formula and its converse are valid in a
constant domain setting, but neither is in a varying domain one.
As it happens, while a choice between constant and varying domain
models makes a difference technically, at a deeper level such a choice is
essentially an arbitrary one. If we choose varying domains as basic, we
can restrict attention to constant domain models by requiring the Barcan
formula and its converse to hold. (Technically this requirement involves
an infinite set of formulas, but if equality is available a single formula will
90
4.
91
3 For sets A and B, A B is the function space, the set of all functions
from B to A.
DEFINITION 7.5 (OBJECTS, EXTENSIONAL AND INTENSIONAL)
Let
g be a non-empty set (of possible worlds) and let V be a non-empty
set (the ground-level domain). For each type t, I define the collection
[t, V, Q], of objects of type t with respect to V and Q, as follows.
1 [O,V,Q] =V.
2 [(t1, ... , tn), V, Q]
92
5.
Truth in a Model
means M, r lf-w <P where w is the a1, ... , an variant of v such that
w(ai) = 01, ... , w(an) =On.
93
94
4 M, r
w.
lf-v (Va)<P if M, r lf-v <P[a/0] for all objects 0 of the same type
as a.
5 M, r lf-v D<P if M, D. lf-v <P for all D. E g such that
rnD..
6.
Validity in a modal setting is now straightforward to define, but consequence has a few surprises, so I've devoted a separate section to the
matter. Let us begin with what is simple.
DEFINITION 7.11 (VALIDITY) A formula <P of L(C) is valid in a model
n, V, I) provided M, r If-v <P for all valuations v and all worlds
M = (g,
rEg.
<I> is K-valid if it is valid in all models, and is S5-valid if it is valid in
all models for which the accessibility relation is an equivalence relation.
95
Factual items would, most naturally, be local assumptions, while logical principles would be global. Tableau rules differ for the two kinds of
assumptions. [Fit83] has a detailed discussion of the notions, including
appropriate versions of the deduction theorem. [Fit93] has a somewhat
more abbreviated treatment.
7.
Examples
96
Say cp(x) is read as "x has no living parents." We are asking about
the truth of (3x)Dcp(x). The key question is, what type of variable
is x? If we think of the quantifier as ranging over objects-so x is of
type 0-then when we say the oldest person in the world instantiates
the existential quantifier we are saying a particular person does so. If
we designate the oldest person now as the value of x, instantiating the
existential quantifier, while cp(x) is certainly true now for this value of
x, there are earlier worlds in which the person who is the oldest now
had living parents. Thus we do not have Ocp(x), where x has as value
the oldest person in the present world. The proposed instantiation for
the existential quantifier does not work. More generally, it is easy to see
that (3x)Dcp(x) can never be true, now or at any other point of time,
provided we think of quantifiers as ranging over objects or individuals.
On the other hand if quantifiers range over individual concepts-so
that x is of type jO-we would certainly have the truth of (3x)Dcp(x)
since taking the value of x to be the oldest-person concept would serve
as a correct instantiation of Ocp(x).
The type theory of [Bre72] makes intensional objects basic. The
second-order logic of [Coc69] quantifies over extensional objects at the
first-order level, and over intensional objects at the second-order level.
The higher-order modal logic of [Fit98], which is a forerunner to this
book, had quantification only over extensional objects. The first-order
treatment of [FM98] involves a kind of mixed system, and more will be
said about it shortly. The system of [FitOOb] has quantifiers over types
0 and jO. Clearly many variations are possible.
Now for some further examples, which will be treated more formally.
ExAMPLE 7.13 Suppose x is a variable of type 0 and P is a constant
symbol of type j(O). The following formula is valid, where X is of type
j(O).
(7.3)
(7.4)
97
optimists and pessimists, separated by a volcano. Let us say the optimists, while generally positive in their outlook, are quite insecure and
don't accept something as possible (even if true) unless the pessimists
believe it. Further, the optimists think the volcano is beautiful, while
the pessimists think nothing is beautiful.
Take for P the concept of beauty-it maps each society to the set of
things that society accepts as beautiful. For the optimists the formula
(.XX.O(:Jx)X(x))(lP) is true for the following reasons. In the optimist
society the extension of P is the set consisting of the volcano, so the
formula asserts O(:Jx)X(x) is the case, when X is understood to be
that set. For optimists, (3x)X(x) is possible if the pessimists believe
it, and even the pessimists would agree that something is in the set
consisting of the volcano. On the other hand, O(.XX.(:lx)X(x))(lP) is
not true for the optimists, because (.XX.(:Jx)X(x))(lP) is not the case
for the pessimists, and this happens because the pessimists do not think
anything is beautiful.
This informal example can be turned into a formal argument. Here
is a model, M = (g, 'R, 'D,I), in which (7.4) is not valid. The collection
of worlds, Q, contains two members, r and ~' with rn~. Think of
r as the optimists and ~ as the pessimists. The domain, 'D is the set
{7} (think of the number 7 as the volcano). I show 7 available at both
worlds as a reminder that domains are constant. The constant symbol
P is interpreted to be a type j(O) object: the function that is {7} at r
and 0 at ~. Thus I(P) is true of 7 at r, and of nothing at ~. This
gives us the model presented schematically below.
[2]
1
[2]
I(P, r)
= {7}
I(P,~) = 0
M, r 11-v (.XX.O(:lx)X(x))(lP).
(7.5)
Since (v *I* r)(lP) = (v *I* r)(P, r) = I(P, r) = {7}, by Proposition 7.10 we will have (7.5) provided we have
(7.6)
98
M, ~ 11-v (3x)X(x)[X/{7}].
(7.7)
(7.8)
M, r IYv O(.XX.(3x)X(x))(!P)
(7.9)
M, r 11-v O(.XX.(3x)X(x))(lP).
(7.10)
M, ~ 11-v (.XX.(3x)X(x))(!P),
(7.11)
0,
M, ~ 11-v (3x)X(x)[X/0].
(7.12)
It is easy to see we can not have this, and thus we have (7.9).
EXAMPLE
consideration, although deeper thought says it should not be. The following formula is valid, with types of variables and constants as in ( 7.4).
(.XX.0(3x)X(x))(!P) :J (.XX.(3x)X(x))(!P)
(7.13)
M, r 11-v (.XX.0(3x)X(x))(tp).
(7.14)
M, r 11-v 0(3x)X(x)[X/O]
(7.15)
Then
~ E
g such that
rR~,
M, ~ 11-v (3x)X(x)[Xj0]
(7.16)
99
(7.17)
(7.18)
since o E 0. Consequently
M, r 11-v (3x)X(x)[X/O]
(7.19)
M, r 11-v (>.X.(3x)X(X))(!P)
(7.20)
and finally,
DP(c) :J (3x)DP(x)
(7.21)
M, r 11-v DP(c)
Let Ll be an arbitrary world such that fRLl. We must have
(7.22)
100
M, ~ 11-v P(c)
(7.23)
M, ~ 11-v P(x)[xji(c)].
(7.24)
Since
M, r 11-v DP(x)[x/I(c)]
(7.25)
M, r 11-v (3x)OP(x)
(7.26)
and hence
Since we went from (7.22) to (7.26), the validity of (7.21) has been
established.
7.16 This continues the previous example. Let c again be of
type jO, but now let x be of type 0 and P be of type j(O). The following
formula is not valid.
EXAMPLE
DP(lc)
:=)
(7.27)
(3x)DP(x)
~@J
I(c,~) =
I(P,~) =
{1}
n@J
I(c,n) = 2
I(P,O) = {2}
101
= y)(lZ))(lZ)
(7.28)
(V'z)(..\x.D(..\y.x = y)(z))(z)
(7.29)
(V'Z)(..\X.D(..\Y.X = Y)(Z))(Z)
(7.30)
(V'Z)(..\x.D(..\y.x
Of the formulas above, (7.28) is not valid, but (7.29) and (7.30) are both
valid. I leave the work to you. I note that in [FM98] it was shown
that, in a first-order setting, the constructions used above relate directly
to rigidity. Both extensional and intensional objects, as such, are the
same from world to world, but the extensional object designated by an
intensional object can vary. This is what the example illustrates.
Exercises
EXERCISE 7.1 Show that formula (7.3) is valid.
EXERCISE 7.2 This is a variation on formula (7.13); the formula looks
the same, but the types are different. Show the validity of
(..\X.0(3x)X(x))(jP) :::> (..\X.(3x)X(x))(jP)
where xis of type jO and Pis of type j(jO). The fact that ground level
quantification is possibilist---constant domain-will be needed.
EXERCISE 7.3 Show the validity of the following, which looks a little
like a version of the Barcan formula: 0(3x)P(x) :::> (3X)OP(lX). In
this x is of type 0, X is of type jO and P is of type j(O).
EXERCISE 7.4 Show the non-validity of the following, where x is of type
8.
Related Systems
102
9.
Henkin/Kripke Models
In the classical case there were good reasons for introducing nonstandard higher-order models, and those same reasons apply in the
103
0.
0.
(v*T*r*A)(r)(r).
104
DEFINITION
For
w.
E 1l(t).
7.22
(HENKIN/KRIPKE MODEL)
(M, A) is a Henkin/Kripke model provided that, for each predicate abstract (.Xa1, ... , an.<I>) of L(C) of type jt, A(v, (.Xa1, ... , an.<I>)) is the
function f given by the following:
The various theorems concerning uniqueness of an abstraction designation function, if one exists, and the good behavior of substitution
(Section 6) all carry over to the modal setting. I leave this to you.
The semantics just presented is extensional, in the sense of Part I. A
modal analog of generalized Henkin models can also be developed, along
the lines of Section 5. Objects in the Henkin domains are no longer sets,
and an explicit extension function must be added. The generalization is
straightforward but complex, and I also leave this to you.
Chapter 8
MODAL TABLEAUS
1.
The Rules
There are several varieties of tableaus for modal logic. This book uses
a version of prefixed tableaus. These incorporate a kind of naming device
for possible worlds into the tableau mechanism, and do so in such a way
that syntactic features of prefixes reflect semantic features of worlds.
Prefixed tableau systems exist for most standard modal logics. Here I
only give versions for K and 85 since these are the extreme cases. I
refer you to the literature for modifications appropriate for other modal
logics-see [FM98] for instance.
1.1
Prefixes
There are two versions of what are called prefixes. The version for K
is more complex, and variations on it also serve for many other modal
logics. The version for 85 is simplicity itself.
DEFINITION 8.1 (PREFIX) A K prefix is a finite sequence of positive
integers, written with periods as separators {1.2.1.1 is an example). An
85 prefix is a single positive integer.
Think of prefixes as naming worlds in some (unspecified) model. Prefix
structure is intended to embody information about accessibility between
worlds. For K, think of the prefixes 1.2.1.1, 1.2.1.2, 1.2.1.3, etc. as
naming worlds accessible from the world that 1.2.1 names. For 85 one
can take each world as being accessible from each world, so prefixes are
simpler. Prefixes have two uses in tableau proofs, qualifying formulas
and qualifying terms. I begin with terms.
105
M. Fitting, Types, Tableaus, and Gdel's God
Kluwer Academic Publishers 2002
106
L+(c).
Think of CJ <I> as saying that formula <I> is true at the world that CJ
names. Note that this use of prefixes does not compound, that is, CJ <I>
is a prefixed formula if <I> is a formula, and not something built up from
prefixed formulas.
DEFINITION 8.3 (GROUNDED) l call a term or a formula of L+(C)
grounded if it contains no free variables, though it may contain parameters.
As usual, tableau proofs are proofs of sentences--closed formulas-of
107
MODAL TABLEAUS
1.2
Propositional Rules
Since the modal tableau rules are rather complex, I've divided their
presentation into categories, beginning here with the propositional ones.
These are much as in the classical case, except that prefixes must be
"carried along." In these, and throughout, I use a, a', and the like to
stand for prefixes.
DEFINITION
aXI\Y
aX
aY
DEFINITION
a (X V Y)
a (X ::) Y)
aX
aY
aX
aY
aX =:Y
aX:=>Y
aY:=>X
aX
aX
DEFINITION
aXVY
aX laY
aX::)Y
a .x a Y
a (X 1\ Y)
X a ,y
(X
a (X ::) Y)
= Y)
I a (Y ::) X)
1.3
Modal Rules
Naturally the rules for modalities differ between the two logics we are
considering. It is here that the structure of prefixes plays a role. The
idea is, if OX is true at a world, X is true at some accessible world,
and we can introduce a name-prefix-for this world. The name should
be a new one, and the prefix structure should reflect the fact that it is
accessible from the world at which OX is true.
DEFINITION
a OX
a.nX
aDX
a.n X
108
aDX
nX
Notice that for both logics there is a newness condition. This implicitly treats 0 as a kind of existential quantifier. Correspondingly, the
following rules treat 0 as a version of the universal quantifier.
8.9 (NECESSITY RULES FORK) If the prefix a.n already
occurs on the branch,
DEFINITION
OX
a.nX
aOX
a.nX
8.10 (NECESSITY RULES FOR 85) For any positive integer n that already occurs on the branch,
DEFINITION
a OX
nX
aOX
nX
1.4
Quantifier Rules
DEFINITION
a (:l<i)cJ>( oJ)
a cf>(pt)
a (Vo:t)cf>( o:t)
a -,cf>(pt)
Terms of the form lT may vary their denotation from world to world
of a model, because the extension of the intensional term T can change
from world to world. Such terms should not be used when instantiating
a universally quantified formula.
8.12 (RELATIVIZED
is a relativized term.
DEFINITION
term,
lT
TERM)
If T is a grounded intensional
MODAL TABLEAUS
109
a ('v'at)<P(at)
a <I>(7t)
1.5
a (3at)<I>( at)
a
<l>( 7t)
Abstraction Rules
The rules for predicate abstracts essentially correspond to Proposition 7.10. Note the presence of a subscript (prefix) on the predicate
abstract. We must know at what world the abstract is to be evaluated
before doing so. The next subsection provides machinery for the introduction of these subscripts. Note that the subscript on the abstract, and
the prefix for the entire formula need not be the same.
DEFINITION 8.14 (ABSTRACT RULES) In the following, 71, ...
,7n
are
non-relativized terms.
a' (>.a1, ... , an.<P( a1, ... , an)) u( 71, ... , 7n)
a<fl(71, ... ,7n)
a' (.Aa1, ... , an.<P( a1, ... , an)) u( 71, ... , 7n)
a --,<fl(71, ... , 7n)
1.6
Atomic Rules
a7(71, ,7n)
a(l7)(71, ,7n)
110
= T.
O"T(Tl,. ,Tn)
0" 1 T( Tt, ... , Tn)
1. 7
0"-,T(Tl, ,Tn)
0"1 -,T(Tl, ... , Tn)
MODAL TABLEAUS
111
DEFINITION 8.23 (TABLEAU DERIVATION) A sentence <I> has a derivation from global assumptions S and local assumptions U if there is a
closed tableau beginning with 1 --,<]>, allowing the use of U and S as local
and global assumptions respectively.
This concludes the presentation of the basic tableau rules. It is a
rather complex system. In Section 2 I give a few examples of proofs using the rules. I omit soundness and completeness proofs. The arguments
are elaborations of those given earlier for classical logic. Complexity of
presentation goes up, but no fundamentally new ideas arise. Consequently they are left as a huge exercise.
There is one important consequence of the completeness proofs that
we will need, however, and that is the fact that the system has the cutelimination property-see Theorem 4.32. Just as in the classical case
(Corollary 4.34), it is a consequence of this that any previously proved
result can simply be introduced into a tableau.
2.
Tableau Examples
Tableaus for classical logic are well-known, and even for propositional
modal logics they are rather familiar. The abstraction and predication
rules of the previous section are new, and I give two examples illustrating their uses. The examples use the K rules; I do not give examples
specifically for S5 here.
EXAMPLE 8.24 This provides a proof for (7.3) which was verified valid
in Example 7.13. The formula is
(>.X.()(:Jx)X(x))(P) :J ()(>.X.(:lx)X(x))(P)
in which x is a variable of type 0 and X is a variable and P a constant
symbol, both of type j(O).
112
1 ..,[(.XX.O(::Jx)X(x))(P) :J 0(-XX.(::Jx)X(x))(P)]
1 (.XX.O(::Jx)X(x))(P) 2.
1 --,0(-XX.(::Ix)X(x))(P). 3.
1
l(.XX.O(::Jx)X(x))(P) 4.
1 (.XX.O(::Jx)X(x))I(P) 5.
1
O(::Jx)P(x) 6.
1.1 (::lx)P(x) 7.
1.1--,(.XX.(::Jx)X(x))(P) 8.
1.1--, l(.XX.(::Ix)X(x))(P) 9.
1.1--,(.XX.(::Jx)X(x))u(P) 10.
1.1--,(::lx)P(x) 11.
1.
In this, 2 and 3 are from 1 by a conjunctive rule; 4 is from 2 by intensional predication; 5 is from 4 by extensional predication; 6 is from 5 by
predicate abstraction; 7 is from 6 by a possibility rule; 8 is from 3 by a
necessitation rule; 9 is from 8 by intensional predication; 10 is from 9 by
extensional predication; and 11 is from 10 by predicate abstraction.
It should be obvious that useful derived rules could be introduced.
For instance, the passage from 2 to 4 to 5 to 6 could be collapsed. Such
rules are given in the next section.
EXAMPLE
earlier.
{.XX.O(::Jx)X(x))(jP)
:J
(.XX.(::Ix)X(x))(jP)
1
1
1
1
1
1
1
1
--,[(.XX.O(::Ix)X(x))(jP) :J (.XX.(::Jx)X(x))(jP)]
(.XX.O(::Jx)X(x))(jP) 2.
--,(.XX.(:3x)X(x))(1P) 3.
1(-XX.O(::Ix)X(x))(jP) 4.
--, 1(-XX.(::Jx)X(x))(lP) 5.
(.XX.O(::Jx)X(x))I(Pl) 6.
--,(.XX.(::Ix)X(x))I(PI) 7.
O(::Jx)P1(x) 8.
1 --,(::Jx)PI(x) 9.
1.1 (:3x)P1 (x) 10.
1.1 P1(p) 11.
1
P1(p) 12.
1
--,pl (p)
1.
13.
113
MODAL TABLEAUS
Exercises
EXERCISE 2.1 Give a tableau proof of the following
(>.X.O(:lx)X(x))(tP) :J (>.X.(:lx)X(x))(lP)
where xis of type jO, X is of type (jO) and Pis of type j(jO).
EXERCISE 2.2 Give a tableau proof of the following
O(:lx)P(x) :J (:lX)OP(lX)
where x is of type 0, X is of type jO and Pis of type j(O).
3.
The tableau examples in the previous section are short, but already
quite complicated to read. In the interests of keeping things relatively
simple, a few derived rules are introduced which serve to abbreviate
routine steps.
DEFINITION 8.26 (DERIVED CLOSURE RULE) Suppose X is a world
independent atomic formula. A branch closes if it contains a X and
a' -,x.
The justification for this is easy. Using the World Shift Rule, if a X
is on a branch, we can add a' X, and then the branch closes according
to the official closure rule.
The official rule concerning intensional predication has a slightly more
efficient version, in which we first apply intensional, then extensional
predication rules.
DEFINITION 8.27 (DERIVED INTENSIONAL PREDICATION RULE) Let
be a grounded intensional term, and T1 , ... , Tn be arbitrary grounded
terms.
Also here are two derived rules for predicate abstracts, one in which
the abstract has a prefix (subscript), one in which it does not.
114
DEFINITION
DEFINITION
Chapter 9
MISCELLANEOUS MATTERS
1.
Equality
1.1
Equality Axioms
(71 = 72)
IP(72)
115
M. Fitting, Types, Tableaus, and Gdel's God
Kluwer Academic Publishers 2002
116
(>.x.O(lp = x))
Informally, this predicate applies to a person at a particular time if
there is some possible future in which that person is the President of the
United States.
Next, call a property of persons statesmanlike if it will always apply to the President. Thus we are using statesmanlike as a property
of properties of persons-being diplomatic is hopefully a statesmanlike
property, for instance. As such, being statesmanlike is of type j((O)).
It is expressed by the following abstract, where X is of type (0), and
applies to those properties that will always belong to the President, no
matter who that will be.
(>.X.DX(lp))
Now, the extension of the property of being Presidential material is a
statesmanlike property since, no matter who turns out to be President,
that person must have been of Presidential material. The following gives
a tableau verification for this.
EXAMPLE
(>.X.DX(lp))(t(>.x.O(lp = x)))
117
MISCELLANEOUS MATTERS
1 (.AX.DX(lp))(l(.Ax.O(lp = x)))
1 D(.Ax.O(lp = x))I(lp) 2.
1.1(.Ax.O(lP = :t))I(lP) 3.
1.1(Ax.O(lp = x))I(pu) 4.
1 O(lP=pu) 5.
1.1(1P = Pu) 6.
1.1(Pu = Pu) 7.
1.1 (pu = Pu) 8.
1.
1.2
Extensionality
Extensionality can, of course, be imposed by assuming the Extensionality Axioms of Chapter 6, Definition 6.1, as global assumptions. The
trouble is, doing so for intensional terms yields undesirable results, as
the following shows.
9.2
Assume the Extensionality Axioms apply to intensional terms. If a and f3 are of intensional type j(t), then the following
is valid.
PROPOSITION
DEFINITION
(Va)(V,B)[D(la
=l/3)
~(a= ,B)]
I will assume this at some points, but I will be explicit when. For
extensional terms, the extensionality axioms pose no difficulty and will
always be assumed. Let me make this official.
118
Extensionality Assumptions From now on, the extensionality axioms will be assumed for extensional terms as global assumptions. For
intensional terms extensionality, Definition 9.3, will only be assumed if
explicitly stated.
I restate the extensionality axioms here for convenience.
9.4 (EXTENSIONALITY FOR EXTENSIONAL TERMS)
Each sentence of the following form is an extensionality axiom, where
a and {3 are of type (t1, ... , tn), 11 is of type t1, ... , In is of type tn.
DEFINITION
2.
De Re and De Dicto
119
MISCELLANEOUS MATTERS
(Vo:)[(.Xf3.Do:(f3))(1T)
= D(.Xf3.o:(f3))(1T)]
(Vo:)[D(.A,B.o:(,B))(lT)
:J
(.A,B.Do:(,B))(tr)]
120
we used schemas instead of the formulas given above, because that was
a first-order treatment and we did not have the higher-type quantifier
(Vo:) available. The interesting fact about the three notions above is:
they all say the same thing.
PROPOSITION
T,
121
MISCELLANEOUS MATTERS
/
1.1-.cf>(Tl)
19.
Exercises
EXERCISE 2.1 Give the tableau proof needed to complete the argument
for Proposition 9.6.
3.
Rigidity
122
DEFINITION
PROPOSITION
123
MISCELLANEOUS MATTERS
1 (A :J B) 1.
1
(,\,6.0(,6 =17))(17) 2.
1 (Va)[0(,\,6.a(,6))(17) :J (>.,6.0a(,6)}(17)]
1 [0(>.,6.<1>(,6))(17) :J (>.,6.0<1>({3))(17)] 4.
1
0(,\,6.<1>(,6))(17) 5.
1 (A,6.0<I>(J3)}(17) 6.
1 D<I>( 71) 7.
1.1<1>(71) 8.
1.1 (,\,6.<1>({3))(17) 9.
1.1 <l>( 71.1) 10.
1
0(71 =17) 11.
1.1 (71 =17) 12.
1.1 71 = 71.1 13.
1.1 <l> (71.1) 14.
3.
4.
1 0(>.,6.(Af'. 17 = /)(,6))(17)
1
(,\,6.0(Af'. 17 = 1)(,6)}(17)
8.
14.
D(>.,. 17 = ,)( 71) 15.
1.2 (A,6.(Af'. 17 = 1)(,6))(17) 9. 1
1.1 (Af'.l7=f'}(71) 16.
1.2(Af'.l7 = 1)(71.2) 10.
1.1 (17=71) 17.
1.2 (17 = 71.2) 11.
1.1 71.1 = 71 18.
1.2 (71.2 = 71.2) 12.
1.1(71 = 71) 19.
1.2 71.2 = 71.2 13.
1.1 71 = 71 20.
124
4.
Stability Conditions
PROPOSITION 9.10 In S5, (\la)[T(a) ::J DT(a)] and (\la)[()T(a) :::> T(a)]
are equivalent.
MISCELLANEOUS MATTERS
125
Exercises
EXERCISE 4.1 Complete the proof of Theorem 9.11 by giving appropriate closed tableaus. Recall that extensionality is assumed for extensional
terms, and we have the derived extensionality rule given in Definition 6.2.
5.
Definite Descriptions
As is well-known, Russell treated definite descriptions by translating them away, [Rus05]. His familiar example, "The King of France is
bald," is handled by eliminating the definite description, "the King of
France," in context, to produce the sentence "exactly one thing Kings
France, and that thing is bald." It is also possible to treat definite
descriptions as first-class terms, making them a primitive part of the
language. In [FM98] we showed how both of these approaches extend to
first-order modal logic. Further extending this dual treatment to higherorder modal logic adds greatly to the complexity, so I confine things to
a Russell-style version here.
Suppose we have a formula <P, and we form the expression m.<P, which
is read as the a such that <P, and is called a definite description. Syntactically it is treated like a term. Its free variables are those of <P, except
for a, and its type is the type of a. In a more formal presentation, all
this would have been built into the definition of term and formula given
earlier, but doing so adds much complexity at the start of the subject,
so I am taking the easier route of explaining now what could have been
done.
DEFINITION 9.12 (DESCRIPTION DESIGNATION) The definite description 1a.<P designates, or is defined at the possible world r of M =
126
(Q, R, 1-l,I) if
M, r 11-v (3,6)('v'8)[(>.a.<P)(8)
=(,6
= 8)]
(3,6){('v'8)[(>.a.<P)(8)
127
MISCELLANEOUS MATTERS
5 l( m. <I>) (T1, . . .
, Tn)
is an abbreviation for
(:Jz1){('v'z2)[(.\y.B(y))(z2)
(3z4){('v'zs)[(.\x.A(x))(zs)
=(z1 = z2)]/\
= (z4 =
zs)]/\ (.\z3.Z3(zl))(z4)}}
(9.1)
(:Jz2){('v'z3)[(.\x.A(x))(z3)
(.\zl-(3z4){('v'zs)[(.\y.B(y))(zs)
= (z2 =
= (z4 =
z3)]/\
zs)]/\ z1(z4)})(z2)}
(9.2)
Fortunately, (9.1) and (9.2) are equivalent. In general, the elimination procedure is confluent-different reduction sequences for the same
atomic formula always lead to equivalent results.
In a sense there are two kinds of definite descriptions, intensional and
extensional, depending on the type of the variable a in 10:. <I>. Extensional
definite descriptions are rather well-behaved, and I say little about them,
but for intensional ones, some interesting issues can be raised. In Definition 9. 7 I characterized a formal notion of rigidity. That definition
can be extended to definite descriptions: call m.<I> rigid at a world if the
following is true at that world.
PROPOSITION
= (,B =
8)]
128
2 (\1,6)[(-\a.<P)(,B) :J 0(-\a.<P)(,B)]
3 (\1,6)[0(-\a.<P)(,B) :J (-\a.<P)(,B)] ..
In other words, this Proposition says (1a. <P) is rigid if and only if
(m.<P) designates and (Aa.<P) satisfies the stability conditions.
Exercises
EXERCISE 5.1 Show the equivalence of (9.1) and (9.2). (For this classical tableaus can be used, since modal operators do not explicitly appear.)
EXERCISE 5.2 Use K tableaus to prove Proposition 9.14. (This is a
long exercise.)
6.
Choice Functions
129
MISCELLANEOUS MATTERS
DEFINITION
(Vo:)[D(:3{3)o:(f3)
::::>
(:3'Y)Do:(h)]
Informally, the axiom says that if, at each world the set of things such
that o: is non-empty-0(:3f3)o:(f3)-then there is a choice function 'Y that
picks out something such that o: at each world-(::l"f)Do:(h). I give one
example of a Choice Axiom application. Suppose o: is an extensional
variable, and m.<I> designates in every possible world. That is, in each
possible world, the <I> is meaningful. Then, plausibly, there should be an
intensional object that, in each world, designates the thing that is the <I>
of that world-that is, the term ?(.D(Ao:.<I>)(t() should also designate.
More loosely, the <I> concept should also designate. Recall, Definition 9.12
says what it means for a definite description to designate, and since
(A(.0(Ao:.<I>)(K))("7) = D(Ao:.<I>)(t'f]), things can be simplified a little.
PROPOSITION
D(:3{3)(V8)[(Ao:.<I>)(8)
= ({3 =
8)]
::::>
(:3'Y)(V"7)[D(Ao:.<I>)(t"7)
= ('Y =
17)]
= ({3
D(:3f3)(V8)[(Ao:.<I>)(8)
= ({3 =
(9.3)
8)]
which is equivalent to
0(:3{3){ (Ao:.<I>)(f3)
1\
(V8)[(Ao:.<I>)(8)
::::>
({3
= 8)]}.
1\
(V8)[(Ao:.<I>)(8)
::::>
("7
= 8)])
(9.4)
130
(9.5)
which is equivalent to
(9.6)
Since the Barcan and converse Barcan formulas are valid in the semantics, this is equivalent to
(9.7)
(9.8)
(9.9)
(9.10)
which is equivalent to
(:l!)(V1J)[O(>,a.<P)(11J)
= (t = 17)]
(9.11)
Exercises
EXERCISE 6.1 Give tableau proofs of the Barcan formula, and of the
converse Barcan formula.
EXERCISE 6.2 Give a tableau proof to show (9. 7) implies (9.8).
III
ONTOLOGICAL ARGUMENTS
Chapter 10
..
1.
Introduction
There are many directions from which people have tried to prove the
existence of God. There have been arguments based on design: a complex universe must have had a designer. There have been attempts to
show that the existence of an ethical sense implies the existence of God.
There have been arguments based on causality: trace the chain of effect
and cause backward and one must reach a first cause. Ontological arguments seek to establish the existence of God based on pure logic: the
principles of reasoning require that God be part of ones ontology.
For religion, as contrasted with philosophy or logic, it does not matter
if proofs for God's existence have holes. Religious belief, like much that
is fundamentally human, is not really the product of reason. We are
emotional animals, and one of the uses of proof, in the various senses
above, is to sway emotion. Proof is often just a rhetorical device, one
among many.
But this takes us too far afield. Here we are interested in ontological
arguments only. Independently of whether one believes their conclusion to be true, the logical machinery used in such arguments is often
ingenious, and merits serious study. It is generally accepted that such
arguments contain flaws, but saying exactly where the flaw lies is not
easy, and is subject to controversy. It happens that different analyses
of the same argument will locate an error at different points. Often
this happens because the notions involved in a particular ontological
argument are vague and subject to interpretation. Godel's ontological
argument is rather unique in that it is entirely precise-the premises are
clearly set forth, and the reasoning can be formalized. But we will see
133
M. Fitting, Types, Tableaus, and Gdel's God
Kluwer Academic Publishers 2002
134
that here too there is room for interpretation, and things are not as clear
as they first seem.
2.
Anselm
Put into more modern terms, Anselm speaks of a maximally conceivable being. This term-maximally conceivable being-must denote
something, since ''whatever is understood is in the mind." But a maximally conceivable being must have the property of existence, because if
it did not, we could conceive of a greater being, namely one that also
had the existence property.
My understanding of this is that, read with some charity, it shows the
phrase "maximally conceivable being," if it designates anything, must
designate something that exists. The flaw lies in the failure to properly
verify that the phrase designates at all-to show it is not in the same
category as "the round square." Indeed, Anselm's way of justifying this,
by claiming that it exists in the mind, is exactly what was attacked by
his contemporary Gaunilo, in his counter-argument, A Reply on Behalf
of the Fool. A modern translation of this can also be found in [Cha79].
Anselm's argument was the ancestor of various later versions, all of
which involve some notion of maximality. An easily accessible discussion
of the family of ontological arguments in general is in the on-line Stanford
Encyclopedia of Philosophy [Opp96b], and [Opp95, Pla65] are recommended as more detailed studies. A full examination of the Anselm argument can be found in [Har65]. In addition, a detailed book in progress
is available on the internet, [SobOl]-the Anselm argument is discussed
in Chapter 2, "Classical Ontological Arguments."
3.
Descartes
Descartes (1598 - 1650) gave different versions of an ontological argument. Here is one, in which he defines God to be a being whose
135
Here is a somewhat different argument, using existence rather than necessary existence. This version is from The Meditations, book V, [Des51].
. . . because I cannot conceive God unless as existing, it follows that existence is inseparable from him, and therefore that he really exists; not that
this is brought about by my thought, or that it imposes any necessity on
things, but, on the contrary, the necessity which lies in the thing itself, that
is, the necessity of the existence of God, determines me to think in this way,
for it is not in my power to conceive a God without existence, that is a being
supremely perfect, and yet devoid of an absolute perfection, as I am free to
imagine a horse with or without wings.
Taking some liberties with the first of the Descartes proofs above: God
is the most perfect being, the being having all perfections, and among
these is necessary existence. Put a little differently, necessary existence
is part of the essence of God. And here we have reached an ontological argument that can be easily formalized. Recall the discussion in
Chapter 7, Section 3. The type-0 objects are possibilist-they represent
what might exist, not what does. If we want to relativize things to what
actually exists, we need a type-(0) "existence" predicate, E, about which
nothing special need be postulated at this point. Now, suppose we define
God to be the necessarily existent being, that is, the being g such that
DE(g). If such a being exists, it must satisfy its defining property, and
hence we have
E(g)
::::>
DE(g).
(10.1)
D[E(g)
::::>
DE(g)]
(10.2)
136
From (10.2), using the K principle D(P :::) Q) :::) (()P :::) OQ) we have
the next implication.
OE(g) :::) ODE(g)
(10.3)
(10.4)
:::) DE(1a.DE(a)).
(10.5)
This is not a valid formula of K, but that logic is too weak anyway, given
the step from (10.3) to (10.4) above. But (10.5) is valid in 85, a fact
I leave to you as an exercise. In fact, using 85, the argument above is
entirely correct!
The real problem with the Descartes argument lies in the assumption that God's existence is possible. In 85 both OE(g) :::) E(g) and
E(g) :::) OE(g) are trivially valid. Since OE(g) :::) DE(g) has been shown
to be valid, we have the equivalence of E(g), OE(g), and DE(g)! Thus,
assuming God's existence is possible is simply equivalent to assuming
God exists. This is an interesting conclusion for its own sake, but as an
argument for the existence of God, it is unconvincing.
Exercises
EXERCISE 3.1 Give an 85 tableau proof of the following, where P and
137
E(m.DE(a))
DE(1a.OE(a)).
(.X,8.E(,8))(1a.OE(a))
D(.X,B.E(,B))(m.OE(a))
(.X,8.E(,8))(1a.DE(a))
(.X,B.OE(,B))(m.OE(a))
4.
Leibniz
Leibniz's remedy amounted to an attempt to prove that God's existence is possible, where God is defined to be the being having all
perfections-again a maximality notion. Intuitively, a perfection is an
atomic property that is, in some sense, good to have, positive. Leibniz based his proof on the compatibility of all perfections, from which
he took it to follow that all perfections could reside in a being-God's
138
5.
Godel
Godel (1906- 1978) was heir to the profound developments in mathematics of the late nineteenth and early twentieth centuries, which often
involved moves to greater degrees of abstraction. In particular, he was
influenced by David Hilbert and his school. In the tradition of Hilbert's
book, Foundations of Geometry, Godel avoided Leibniz's problems completely, by going around them. It is as if he said, "I don't know what
a perfection is, but based on my understanding of it intuitively, it must
have certain properties," and he proceeded to write out a list of axioms.
This neatly divides his ontological argument into two parts. First, based
on your understanding, do you accept the axioms. This is an issue of
personal intuitions and is not, itself, subject to proof. Second, does the
desired conclusion follow from the axioms. This is an issue of rigor and
the use of formal methods, and is what will primarily concern us here.
Godel's particular version of the argument is a direct descendent of
that of Leibniz, which in turn derives from one of Descartes. These arguments all have a two-part structure: prove God's existence is necessary,
if possible; and prove God's existence is possible.
Godel worked on his ontological argument over many years. According
to [Ada95], there is a partial version in his papers dated about 1941.
In 1970, believing he would die soon, Godel showed his proof to Dana
Scott. In fact Godel did not die until1978, but he never published on the
matter. Information about the proof spread via a seminar conducted by
Dana Scott, and his slightly different version became public knowledge.
Godel's proof appeared in print in [Sob87], based on a few pages of
Godel's handwritten notes. Scott also wrote some brief notes, based
on his conversation with Godel, and [Sob87] provides these as well. In
fact, [Sob87] has served as something of a Bible (pun intended) for the
139
6.
140
itive.
It follows that there must be positive properties. If we call a property
that is not positive negative, it also follows that there are negative properties. By Informal Axiom 1, a negative property can also be described
as one whose complement is positive.
Suppose we say property P entails property Q if, necessarily, everything having P also has Q.
INFORMAL AXIOM 2
itive.
INFORMAL PROPOSITION
P.
Proof Suppose P is positive. Let N be some negative property (the
complement of P will do). It cannot be that P entails N, or else N
would be positive. So it is not necessary that everything having P has
N, that is, it is possible that something has P without having N. So it
is possible that something hasP.
Leibniz attempted a proof that "all perfections are compatible with
each other or can be in the same subject," that is, having all perfections is a possibly instantiated property. Godel instead simply takes
141
the following as an axiom-it is an immediate consequence, using Informal Proposition 1, that having all positive properties is a possibly
instantiated property.
INFORMAL AXIOM 3 The conjunction of any collection of positive properties is positive.
This is a problematic axiom, in part because there are infinitely many
positive properties, and we cannot form an infinite conjunction (unless
we are willing to allow an infinitary language). There are ways around
this, but there is a deeper problem as well-we will see that this axiom
is equivalent to Godel's desired conclusion (given Godel's other assumptions). But further discussion of this point must wait till later on. For
now we adopt the axiom and work with it in an informal sense.
Now Godel defines God, or rather, defines the property of being Godlike, essentially the same way Leibniz did.
INFORMAL DEFINITION 2 A God is any being that has every positive
property.
This gives us part one of the argument rather easily.
INFORMAL PROPOSITION 3 It is possible that a God exists.
142
Proof Let's state what we must show a little more precisely. Suppose
G is the conjunction of all positive properties, so having property G is
what it means to be a God. It must be shown that if an object g has
property G, then G is the essence of g.
Suppose g has property G. Then automatically we have part 1 of
Informal Definition 4.
Suppose also that P is some property of g. By Informal Axiom 1, if
P were not positive its complement would be. Since g has all positive
properties, g then would have the property complementary toP. Since
we are assuming g has P itself, we would have a contradiction. It follows
that P must be positive. Since G is the conjunction of all positive
properties, clearly G entails P. Since P was arbitrary, G entails every
property of g, and we have part 2 of Informal Definition 4.
Here is the second of Godel's two new notions.
INFORMAL DEFINITION 6 An object g has the property of necessarily
existing if the essence of g is necessarily instantiated.
And here is the last of G6del's axioms.
INFORMAL AXIOM 5 Necessary existence, itself, is a positive property.
INFORMAL PROPOSITION 7 If a God exists, a God exists necessarily.
Proof Suppose a God exists, say object g is a God. Then g has all
positive properties, and these include necessary existence by Informal
Axiom 5. Then the essence of g is necessarily instantiated, by Informal
Definition 6. But the essence of g is being a God, by Informal Proposition 5. Thus the property of being a God is necessarily instantiated .
143
OP
:::::>
:::::>
DP is
Exercises
EXERCISE 6.1 Show that only God can have a positive essence. (This
exercise is due to Ioachim Teodora Adelaida of Bucharest.)
Chapter 11
..
1.
General Plan
The last Chapter ended with an informal presentation of Godel's argument. This one is devoted to a formalized version. I'll also consider some
objections and modifications. There are two kinds of objections. One
amounts to saying that Godel committed the same fallacy Descartes did:
assuming something equivalent to God's existence. Nonetheless, again
as in the Descartes case, much of the argument is of interest even if it
falls short of establishing the desired conclusion. The second kind of
objection is that Godel's axioms are too strong, and lead to a collapse
of the modal system involved. Various extensions and modifications of
Godel's axioms have been proposed, to avoid this modal collapse. I'll
discuss these, and propose a modification of my own. Now down to details, with the proof of God's possible existence coming first. I will not
try to match the numbering of the informal axioms in the last chapter,
but I will refer to them when appropriate.
2.
Positiveness
146
as
[P(X)
1\
147
11.4
(EXISTENTIAL RELATIVIZATION)
('v'Ex)<I> abbrevi-
AXIOM
Now we have a result from which the possible existence of God will
follow immediately, given one more key assumption about positiveness.
PROPOSITION
11.8
1)
Proof The idea has already been explained, in the proof of Informal
Proposition 1 in Section 6. This time I give a formal tableau, which is
displayed in Figure 11.1. In it use is made of one of the Propositions
above. Item 1 negates the proposition in unabbreviated form. Item 2 is
from 1 by an existential rule (with P as a new parameter); 3 and 4 are
from 2 by a conjunctive rule; 5 is Axiom 1; 6 is from 5 and 7 is from 6
by universal rules; 8 and 9 are from 7 by a disjunctive rule; 10 and 11
148
::::>
P(X)]
= (W)[Z(Y)
::::>
Y(x)]}
::::>
(\iX)[P(X)
::::>
DP(X)]
0-.P(X)]
149
150
P has been taken to be an intensional object, of type j(j(O)). Axiom 11.11 and Theorem 9.11 tells us that Pis rigid. In effect the intensionality of P is illusory-since it is rigid it could just as well have been
an extensional object of type (j(O)).
Exercises
EXERCISE 2.1 Give a tableau proof that (.Ax.(x = x)) = (.Ax.x
More generally, show that for a type (0) term T, (T) = T.
= x).
EXERCISE 2.2 Show that ('v'X)[P(X) :J DP(X)] follows from Axiom 11.11 together with Axioms 11.3A and 11.3B.
EXERCISE 2.3 Show Axiom 11.10 implies Axiom 11.9. Hint: use equality.
3.
(>.x.(W)[P(Y)
= Y(x)]).
151
=G*(x)].
Axiom 11.3B is a little problematic, but it is essential to the Proposition above. If, eventually, we show something having property G exists,
and G and G* are equivalent, we will know that something having property G* exists. But the converse is also the case: if something having
property G* exists, Axiom 11.3B is the case, even if the existence in
question is possibilist. Here is a formal statement of this. Once again I
leave the proof to you.
PROPOSITION 11.15 InK, (3x)G*(x) :J (VX)[P(X) :J 'P(X)].
Now we can show that God's existence is possible. Godel assumes the
conjunction of any family of positive properties is positive. Since G* is,
in effect, the conjunction of all positive properties, it must be positive,
and hence so must G be.
PROPOSITION 11.16 InK Axiom 11.10 implies P(G).
Once again I leave the formal verification to you. What must be
shown is the following.
('v'Z)('v'X){[pos(Z)
1\
('v'X){P(X) :J 0(3Ex)X(x)},
hence trivially,
('v'X){P(X) :J 0(3x)X(x)}.
By the Proposition above, P(G). The result is immediate.
Note that the full strength of Proposition 11.8 was not really needed
for the possibilist conclusion. In fact, if we modify Axiom 11.5 so that
quantification is possibilist,
152
(\fX){P(X) :J 0(:3x)X(x)}
and the Godel proof would still go through.
Exercises
EXERCISE 3.1 Give a tableau proof that G entails any positive property:
(\fX){P(X) :J D(\fy)[G(y) :J X(y)]}. You will need Axiom 11.11.
EXERCISE 3.2 Give a tableau proof for Proposition 11.14.
EXERCISE 3.3 Give a tableau proof for Proposition 11.15.
EXERCISE 3.4 Give a tableau proof for Proposition 11.16.
EXERCISE 3.5 Give a tableau proof of
(\fZ)(\fX){[pos(Z)
4.
1\
Objections
Godel replaced Leibniz's attempted proof of the compatibility of perfections by an outright assumption, given here as Axiom 11.10. Dana
Scott, apparently noting that the only use Godel makes of this Axiom
is to show being Godlike is positive, proposed taking P( G) itself as an
axiom. Indeed, Scott maintains that the Godel argument really amounts
to an elaborate begging of the question-God's existence is simply being
assumed in an indirect way. In fact, it is precisely at the present point
in the argument that Scott's claim can be localized. Godel's assumption
that the family of positive properties is closed under conjunction turns
out to be equivalent to the possibility of God's existence, a point also
made in [SobOl].
We will see, later on, Godel's proof that God's existence is necessary,
if possible, is correct. It is substantially different from that of Descartes,
and has many points of intrinsic interest. What is curious is that the
proof as a whole breaks down at precisely the same point as that of
Descartes: God's possible existence is simply assumed, though in a disguised form.
The rest of this section provides a formal proof of the claims just
made. Enough tableau proofs have been given in full, by now, so that
abbreviations can be introduced as an aid to presentation. Before giving
the main result of this section, I introduce some simple conventions for
shortening displayed tableau derivations.
153
a X 3.
2.
4.
aY
The left branch is closed, and the branch below 4 continues as if we had
used the derived rule.
Here are a few more derived rules, whose justification I leave to you.
aX
a (X 1\ Y) :J Z
aY :J Z
aX
aX=:Y
aY
aX
a X=:Y
a--,Y
11.18 Assume all the Axioms to this point, except for Axiom 11.10 and Axiom 11.9. The following are equivalent, using 85:
THEOREM
1 Axiom 11.10;
2 P(G);
3 O(:JEx)G(x);
154
0(3x)G(x).
1 P(G) 1.
1 -{v'X)(W){[P(X) 1\ P(Y)] :::> P(X 1\ Y)} 2.
1 {[P(A) 1\ P(B)] :::> P(A 1\ B)} 3.
1
P(A) 1\ P(B) 4.
1 P(A 1\ B) 5.
1
P(A) 6.
1
P(B) 7.
1
('v'X)('v'Y){[P(X) 1\ D('v'Ex)(X(x) :::> Y(x))] :::> P(Y)} 8.
1
[P(G) 1\ D('v'Ex)(G(x) :::>(A 1\ B)(x))] :::> P(A 1\ B) 9.
1
D('v'Ex)(G(x) :::>(A 1\ B)(x)) :::> P(A 1\ B) 10.
1 D('v'Ex)(G(x) :::>(A 1\ B)(x)) 11.
1.1(\:IEx)(G(x) :::>(A 1\ B))(x) 12.
1.1(\:lx)[E(x) :::> (G(x) :::> (A 1\ B)(x))] 13.
1.1-,[E(c) :::> (G(c) :::>(A 1\ B)(c))] 14.
1.1 E(c) 15.
1.1(G(c) :::> (A 1\ B)( c)) 16.
1.1 G(c) 17.
1.1-,(AAB)(c) 18.
1
('v'X)[P(X) :::> DP(X)] 19.
1
P(A) :::> DP(A) 20.
1 P(B) :::> DP(B) 21.
1
DP(A) 22.
1
DP(B) 23.
1.1 P(A) 24.
1.1 P(B) 25.
1.1 (>.x.('v'Y)[P(Y) :::> Y(x)])(c) 26.
1.1 ('v'Y)[P(Y) :::> Y(c)] 27.
1.1 P(A) :::> A(c) 28.
1.1 P(B) :::> B(c) 29.
1.1 A(c) 30.
1.1 B(c) 31.
1.1(.\x.A(x) 1\ B(x))(c) 32.
1.1[A(c) 1\ B(c)] 33.
/~
1.1A(c)
34. 1.1B(c)
Figure 11.2.
35.
155
156
Exercises
EXERCISE 4.1 Give a tableau proof showing that 0(3x)G(x) implies
Axiom 11.10.
5.
Essence
Even though we ran into the old Descartes problem with half of the
Godel argument, we should not abandon the enterprise. The other half
contains interesting concepts and arguments. This is the half in which
it is shown that God's existence is necessary, if possible. For starters,
Godel defines a notion of essence that plays a central role, and is of
interest in its own right. [Haz98] makes a case for calling Godel's notion
character, reserving the term essence for something else. I follow Godel's
terminology. The essence of something, x, is a property that entails every
property that x possesses. Godel says it as follows.
cp Ess x = (\17/!){7/J(x)
0(\fy)[cp(y)
7/J(y)]}
As just given, it does not follow that the essence of x must be a property
that x possesses. Dana Scott assumed this was simply a slip on the part
of Godel, and inserted a conjunct cp(x) into the definition. I will follow
him in this.
cp Ess x
= cp(x) 1\ (V7j!){7j!(x)
D('v'y)[cp(y)
7/J(y)]}
157
THEOREM
('v'x)[G(x) :) ( G, x)].
Rather than giving a direct proof, if we use Proposition 11.14 it follows
from a similar result concerning G*, provided Axiom 11.3B is assumed.
Since such a result has a somewhat simpler proof, this is what is actually
shown.
THEOREM
(\fx)[G*(x):) (G*,x)].
Proof Here is a closed K tableau to establish the theorem.
1 ('v'x )[G* (x) :) ( G*, x )] 1.
1 -.., [G* (g) :) (G*, g)] 2.
1 G*(g) 3.
1-.E(G*,g) 4.
1-.{G*(g) A ('v'Z){Z(g):) D('v'Ew)[G*(w):) Z(w)]}}
5.
/~
1-.G*(g)
7.
158
8.
able.
(VX)(VY)(Vz){[t'(X, z)
1\
= Y(w)]}
159
1 E(P, a) 1.
1 E(Q, a) 2.
1D('v'Ew)[P(w) Q(w)] 3.
1 P(a) 4.
1 ('v'Z)[Z(a) :J D('v'Ew)[P(w) :J Z(w)]]
1 Q(a) 6.
1 ('v'Z)[Z(a) :J D('v'Ew)[Q(w) :J Z(w)Jl
1 Q(a) :J D('v'Ew)[P(w) :J Q(w)] 8.
1 P(a) :J D('v'Ew)[Q(w) :J P(w)] 9.
1Q(a)
5.
7.
/~
1P(a)
11.
/~
13.
/~
1.1 P(b)
1.1Q(b)
18. 1.1P(b)
19. 1.1 Q(b)
20.
21.
160
22.
23.
/~
1.1--,E(b)
24.
25.
/~
1.1--,P(b)
26.
1.1 Q(b)
27.
Exercises
EXERCISE 5.1 Give a tableau proof for Theorem 11.23. Hint: for a
parameter c, one can consider the property of being, or not being, c,
that is, (.Xx.x =c) and (.Xx.x #c). Either property can be used in the
proof.
EXERCISE 5.2 Give a tableau proof to establish Theorem 11.20 directly,
without using G*.
6.
(.Xx.(VY)[(Y, x)
:J
0(3Ez)Y(z)]).
161
The idea is, something has the property N of necessary existence provided any essence of it is necessarily instantiated. Godel now makes a
crucial assumption: necessary existence is positive.
AXIOM
'P(N).
Given this final axiom, Godel shows that if (some) God exists, that
existence cannot be contingent. An informal sketch of the proof was
given in Section 6 of Chapter 10, and it can be turned into a formal
proof-see Informal Propositions 7 and 8. I will leave the details as
exercises, since you have seen lots of worked out tableaus now. Here is
a proper statement of Godel's result, with all the assumptions explicitly
stated. Nate that the necessary actualist existence of God follows from
His possibilist existence.
11.26 (FORMALIZING INFORMAL PROPOSITION 7)
Assume Axioms 11.3B, 11.11, and 11.25. In the logic K,
THEOREM
THEOREM
D(3Ex)G(x).
Proof By Theorems 11.27 and 11.17.
162
Exercises
EXERCISE 6.1 Give a tableau proof to show Theorem 11.26. Use various
earlier results as assumptions in the tableau.
7.
Going Further
Godel's axioms admit more consequences than just those of the ontological argument. In this section a few of them are presented.
7.1
Monotheism
Does there exist exactly one God? The following says "yes." You are
asked to prove it, as Exercise 7.1.
PROPOSITION 11.29 (3x)('v'y)[G(y) =: (y = x)j.
This Proposition has a curious Corollary. Since type-0 quantification
is possibilist, it makes sense to ask if there are gods that happen to be
non-existent. But Corollary 11.28 tells us there is an existent God, and
the Proposition above tells us it is the only one God, existent or not.
Consequently we have the following.
COROLLARY 11.30 ('v'x)[G(x) :J E(x)j.
Proposition 11.29 tells us we can apply the machinery of definite descriptions. By Definition 9.12, 1x.(W)[P(Y) :J Y(x)] always designates,
and consequently so does 1x.G(x). Proposition 9.14 tells us this will be
a rigid designator provided G(x) is stable. It follows from Sobel's argument in Section 8 that it, and everything else, is. But alternative versions
of Godel's axioms have been proposed-! will discuss some below-and
using them the stability of G(x) does not seem to be the case. That
is, it seems to be compatible with the axioms of Godel (as modified by
others) that, while the existence of God is necessary, a particular being
that is God need not be God necessarily. If this is not the case, and a
proof has been missed, I invite the reader to correct the situation.
7.2
Proposition 11.8 says that positive properties are possibly instantiated. In [Sob87], it is observed that a consequence of Corollary 11.28 is
that every positive property is necessarily instantiated.
PROPOSITION 11.31 ('v'X){'P(X) :J 0(3Ex)X(x)}.
I leave the easy proof of this to you.
163
Exercises
EXERCISE 7.1 Give a tableau proof for Proposition 11.29. Hint: you
will need Corollary 11.28, Theorem 11.20, and Theorem 11.23.
EXERCISE 7.2 Provide a tableau proof for Proposition 11.31. Hint: by
8.
More Objections
(11.1)
(11.2)
(11.3)
Now to undo the technicality just introduced, note that since y does not
occur free in Q, (>..y.Q)(x) = (>..y.Q)(w) = Q, and so we have
(11.4)
164
Since x does not occur free in the consequent, (11.4) is equivalent to the
following:
(11.5)
(3x)G(x)
(11.6)
(11.7)
(11.8)
(11.9)
Finally (11.9), and Corollary 11.28 again, give the intended result,
Q :::> DQ.
(11.10)
9.
A Solution
Sobel's demonstration that the Godel axioms imply no free will rather
takes the fun out of things. In this section I propose one solution to the
problem. I don't profess to understand its implications fully. I am
presenting it to the reader, hoping for comments and insights in return.
Throughout, it has been assumed that Godel had in mind intensional properties when talking about positiveness and essence. But,
suppose not-suppose extensional properties were intended. I reformulate Godel's argument under this alternative interpretation. It is one
way of solving the problem Sobel raised.
165
(VX)(VY){[P(X) A D(V'Ex)(X(x)
:J
Y(x))]
:J
P(Y)}
Note that this has the same form as Axiom 11.5, but the types of
variables X and Y are now extensional rather than intensional. This
will be the general pattern for changes. The definition of negative, for
instance, is modified as follows. For a term T of type (0), take T as
short for l(.Xx.T(x)). Then Axioms 11.3A and 11.3B, 11.10, and 11.11,
all have their original form, but with variables changed from intensional
to extensional type.
The analog of Proposition 11.8 still holds, but with extensional variables involved.
(VX){P(X)
:J
0(3Ex)X(x)}
Analogs of G and G* are defined in the expected way. G is the following type j(O) term, where Y is type (0) and, as noted before, Pis of
type j((O)).
(.Xx.(VY)[P(Y)
:J
Y(x)])
(.Xx.(W)[P(Y)
= Y(x)]).
:J
D(V'Ew)[Y(w)
:J
Z(w)]})
Theorem 11.21 plays an essential role in the Godel proof, and it too
continues to hold, in a slightly modified form:
(>.x.('VY)[E(Y, x) :J D(3Ez)Y(z)).
166
{b}.
3 The designation of N is also rigid, with extension {a} at each world.
4 All the Axioms are valid, as modified in this section.
Now take Q to be the closed formula (::JEx)(::JEy)-,(x = y). Since it
asserts two objects actually exist, it is true at r, but not at D., and hence
Q :J OQ is not true at r.
We now know that Sobel's argument must break down in the present
system, but it is instructive to try to reproduce the earlier proof, and see
just where things go wrong. The attempted argument takes on a rather
formidable appearance--you might want to skip to the last paragraph
and read the conclusion, before going through the details.
167
(11.11)
which, unabbreviated, is
(Vx)[G(x) ::)
(.XY,x.Y(x) 1\ (VZ){Z(x)::) D(VEw)[Y(w)::) Z(w)]})(1G,x)]
(11. 12 )
where Y and Z are of type (0), unlike in (11.2) where they were of type
j(O).
The variable xis of type 0, and it is easy to show the following simpler
formula is a consequence of (11.12).
(\ix)[G(x) ::)
(11.13)
(Vx)[G(x) ::)
(.XY.(VZ){Z(x)::) D(VEw)[Y(w)::) Z(w)]})(1G)]
(11.14)
(Vx)[G(x) ::)
(.XY, Z.Z(x) ::) D(VEw)[Y(w) ::) Z(w)])(1G, 1(-Xy.Q) )]
(11.15)
(11.16)
168
The variable x does not occur free in (>.y.Q) and Y does not occur
in Z(x), so (>.Y,Z.Z(x))(lG,l(>.y.Q)) is simply equivalent to Q, and
(11.16) reduces to the following.
('v'x){G(x) :J
(11.17)
(3x)G(x) :J
(11.18)
(11.19)
(11.20)
(11.21)
(11.22)
(>.Z.DZ(a))(l(>.y.Q) )].
(11.23)
169
Q:) (>.Z.DZ(a))(l(>.y.Q)).
(11.24)
Q :) (.>.Z.DZ)(lQ)
(11.25)
and this is quite different from Q :) DQ. In the abstract, the variable
Z is assigned the current version of Q-its truth value in the present
world. Perhaps an example will make clear what is happening. Suppose
it is the case, in the real world, that it is raining-take this as Q. If
we had validity of Q :) DQ, it would necessarily be raining-DQ-and
so in every alternative world, it would be raining. But what we have is
Q :) (.>.Z.DZ)(lQ), and since Q is assumed to hold in the real world, we
conclude (>.Z.DZ)(lQ). This conclusion asserts something more like: if
it is raining in the real world, then in every alternative world it is true
that it is raining in the real world. As it happens, this is trivially correct,
and says nothing about whether or not it is raining in any alternative
world.
10.
Anderson's Alternative
(.Xx.(W)[P(Y) := DY(x)]).
There is a corresponding change in the definitions of essence and necessary existence. Definition 11.19 gets replaced by the following
170
[A abbreviates
(Vx)[GA(x) :J eA(GA,x)].
I leave it to you to verify the theorem, using tableaus say.
Next, Anderson replaces Axiom 11.25 with a corresponding version
asserting that his modification of necessary existence is positive.
(3x)GA(x) :J 0(3Ex)GA(x).
Once again, I leave the proof to you. These are the main items. The
rest of the ontological argument goes through as before. At the end, we
have the following.
THEOREM 11.39 Assume all the Axioms 11.3A, 11.5, 11.10, 11.11, and
11.37. In the logic S5,
171
Thus the desired necessary existence follows, and with one fewer axiom
(though with more complex definitions). And a model, closely related to
the one given in the previous section, can be constructed to show that
these axioms do not yield Sobel's undesirable conclusion-see [And90]
for details.
Exercises
10.1 Supply a tableau argument for Theorem 11.36. Do the
same for Theorem 11.38.
EXERCISE
11.
Conclusion
=<I>(x)].
172
puzzles, in [Smu83], Chapter 10. You should find this fun, and a bit of
a relief after the rather heavy going of the book you just finished.
References
Note to reader: At the end of each bibliography item is a list of the pages on which there
is a reference to the item.
(Ada95]
(AG96]
(And71]
[And72]
[And86]
[And90]
(Bre72]
Aldo Bressan. A General Interpreted Modal Calculus. Yale University Press, 1972. pages 96, 102
(Car 56]
(Cha79]
[Chu40]
173
174
[Coc69]
[Des51]
[Fit83]
Melvin C. Fitting. Proof Methods for Modal and Intuitionistic Logics. D. Reidel Publishing Co., Dordrecht, 1983. pages 95, 102,
110
[Fit93]
[Fit96]
Melvin C. Fitting. First-Order Logic and Automated Theorem Proving. Springer-Verlag, 1996. First edition, 1990. pages 8, 10, 33, 37,
41, 47
[Fit98]
[FitOOa]
[FitOOb]
[FJWDG+95]
[FM98]
[G70]
[Gal75]
[GHR94]
Dov M. Gabbay, C. J. Hogger, and J. A. Robinson, editors. Handbook of Logic in Artificial Intelligence and Logic Programming. Oxford University Press, 1993-94. pages 174, 176
North-
REFERENCES
175
[Gil99]
Paul C. Gilmore. Partial functions in an impredicative simple theory of types. In Neil V. Murray, editor,. Automated Reasoning
with Analytic Tableaux and Related Methods, number 1617 in Lecture Notes in Artificial Intelligence, pages 186--201, Berlin, 1999.
Springer. pages 4
[Gil01]
Paul C. Gilmore. An intensional type theory: motivation and cutelimination. To appear, Journal of Symbolic Logic, 2001. pages 4,
33
[Gol81]
Warren D. Goldfarb. The undecidability of the second-order unification problem. Theoretical Computer Science, 13:225-230, 1981.
pages 42
[Haj95]
Petr Hajek. Der Mathematiker und die Frage der Existenz Gottes.
To appear in Wahrheit und Beweisbarkeit- Leben und Werk von K.
Godel, Buld, Kohler, and Schimanowicz., 1995. pages 171
[Haj96a]
Petr Hajek, editor. Godel '96, Berlin, 1996. Springer. pages 173
[Haj96b]
[Har65]
[Haz98]
[HC96]
[Hen50]
[Hue72]
[Hue73]
[Hue75]
Gerard P. Huet. A unification algorithm for typed >.-calculus. Theoretical Computer Science, 1:27-57, 1975. pages 42
[Koh95]
Journal of
176
[Kri80]
(Lei 56]
[Lei94]
(Mag88]
171
[Man96]
Maria Manzano. Extensions of First Order Logic. Cambridge University Press, Cambridge, UK, 1996. pages xii
(Mon60]
Richard Montague. On the nature of certain philosophical entities. The Monist, 53:159-194, 1960. Reprinted in (Tho74], 148-187.
pages xv, 102
(Mon68]
Richard Montague. Pragmatics. pages 102-122. 1968. In Contemporary Philosophy: A Survey, R. Klibansky editor, Florence, La Nuova
Italia Editrice, 1968. Reprinted in [Tho74], 95-118. pages xv, 102
(Mon70]
(Opp95]
Graham Oppy. Ontological Arguments and Belief in God. Cambridge University Press, Cambridge, UK, 1995. pages 134
[Opp96a]
Graham Oppy.
Gooelian ontological arguments.
56(4):226-230, October 1996. pages 176
[Opp96b]
(Opp01]
(Pla65]
Alvin Plantinga. The ontological argument, from St. Anselm to contemporary philosophers. Anchor-Doubleday, Garden City, NY, 1965.
pages 134
(Pra68]
[Rob77]
The Gospel of Philip. In James M. Robinson, editor, The Nag Hammadi Library in English. Harper and Row, New York, 1977. pages
ix
Analysis,
REFERENCES
177
[Rus05]
[Sch60]
[Sha91]
[SmaOl]
[Smi93]
[Smu68]
[Smu83]
Raymond M. Smullyan. 5000 B. C. and Other Philosophical Fantasies. St. Martin's Press, New York, 1983. pages 172
(Sob87]
(SobOl]
[Tak67]
[Tar55]
[Tho74]
Richmond H. Thomason, editor. Formal Philosophy, Selected Papers of Richard Montague. Yale University Press, New Haven and
London, 1974. pages 176, 176, 176
(Tho87]
(Tol75]
Sue Toledo. Tableau Systems for First Order Number Theory and
Certain Higher Order Theories. Number 447 in Lecture Notes in
Mathematics. Springer-Verlag, Berlin, Heidelberg, NY, 1975. pages
xii, 33, 33, 35, 46
Index
E-complete, 63
entity, 48, 49
equality, 115
equality axioms, 69, 115
essence, 141, 142, 156, 170
evaluation at a prefix, 109
existential relativization, 147
extensional object, 84, 91
extensionality, 117
assumptions, 118
axioms, 77
for extensional terms, 118
for intensional terms, 117
finite support, 8
formula, 6
modal, 88
prefixed, 106
truth, 13, 22, 26, 93, 104
frame
augmented, 91
extensional, 30
Henkin, 20
generalized, 25
relative generalized, 50
Henkin/Kripke, 103
Kripke, 91
free, 9
free variable, 6
de dicto, 118-121
de re, 118-121
Dedekind, R., 15
defined at, 125
definite description, 125-128
degree, 8
Descartes, R., 134-136, 152, 156
description designation, 125
designates, 125
domain, 91
domain function
Henkin, 20, 103
Gaunilo, 134
global assumption, 95, 111
Gi:idel, K., 138-143, 145, 147, 148, 150,
152, 156, 158, 162-164, 166,
171
grounded, 34, 106
Hajek, P., 171
Henkin domain
179
180
relative, 50
Hintikka set, 47
impredicativity, 4
inconsistent, 63
intensional object, 84, 91
interpretation, 11, 30, 51, 73, 92, 103
allowed, 50
K, 105
L(C), 5
.X abstraction, 3
Leibniz, G., 137-140, 145, 148, 152
Lindstrom, P., 68
local assumption, 95, 111
Liiwenheim-Skolem, 66, 68
Magari, R., 171
model
classical, 12
extensional, 30
general, 19
generalized Henkin, 28, 104
Henkin, 19, 22, 23
Henkin/Kripke, 104
modal, 91
standard, 24
monotheism, 162
necessary existence, 142, 160, 170
negative, 141, 146
non-rigid, 102
normal, 25
order, 5
parameter, 34, 108
perfection, 137-139
positive, 138-142, 145, 146, 162
possible value, 49
possible world, 91
predicate abstract, 5
predicate abstraction, 3
prefix, 105
pseudo-model, 47, 48, 51
quantification
actualist, 89, 91
possibilist, 89, 91
rigid, 121-124
rule
abstract, 37, 109
branch extension, 35
conjunctive, 35, 107
derived
85, 105
satisfiability, 14, 28
Scott, D., 138, 152, 156, 158
sentence, 6
Sobel, J. H., 163, 164, 166, 171
sound, 43, 46, 73
stability, 124-125
substitution, 8
free, 9
tableau, 33
basic, 35
derivation, 37, 111
prefixed, 105
proof, 37, 110
satisfiable, 43
term, 6, 87
denotation, 12, 21, 26
designation, 93, 103
relativized, 108
type, 4, 86
extensional, 86
Gallin/Montague, 102
intensional, 86
relation, 11
validity, 14, 28, 94
valuation, 12, 20, 26, 92
variable, 5
variant, 12
varying domain, 89
Wittgenstein, L., 139
world independent, 110
Zermelo-Fraenkel set theory, 17
TRENDS IN LOGIC
1.
2.
E. Ejerhed and S. Lindstrom (eds.): Logic, Action and Cognition. Essays in Philosophical Logic. 1997
ISBN 0-7923-4560-6
3.
ISBN 0-7923-5205-X
4.
ISBN 0-7923-5238-6
5.
H.J. Ohlbach and U. Reyle (eds.): Logic, Language and Reasoning. Essays in Honour
ofDov Gabbay. 1999
ISBN 0-7923-5687-X
6.
7.
ISBN 0-7923-5720-5
R.L.O. Cignoli, I.M.L. D' Ottaviano and D. Mundici: Algebraic Foundations ofmanyISBN 0-7923-6009-5
9.
V.F. Hendricks: The Convergence of Scientific Knowledge. A View From the Limit.
2001
ISBN 0-7923-6929-7
10.
11.
12.
ISBN 0-7923-6940-8
ISBN 1-4020-0604-7