RUSSIA
VOLUME
$4J.oo
PRINCIPIA
MATHEMATICA
BY
A.N.
WHITEHEAD
AND
BERTRAND RUSSELL
Principia Mathematica
in 19 1 013
this
is
the
was
first
fifth
published
impression of
The
as
one of the
century.
clearly
It
the
intellectual
was the
close
landmarks of the
book
first
relationship
from
been recognized
logic.
to
show
between
Starting
No
other book
on the subse
the set
of 3 volumes
PRINCIPIA MATHEMATICA
BY
BERTRAND RUSSELL,
VOLUME
F.R.S.
SECOND EDITION
CAMBRIDGE
AT THE UNIVERSITY PRESS
1963
PUBLISHED BY
THE SYNDICS OF THE CAMBRIDGE UNIVERSITY PRESS
Bentley House, 200 Euston Road, London, N.W. 1
American Branch 32 East 57th Street, New York 22, N.Y.
:
West African
Office:
1910
Second Edition 1927
1950
Reprinted
First Edition
1957
1960
1963
PRESTON
v>ri>
^4i 3<^ y
Jteik
'tf(^
84
Reprinted by offsetMho
by Messrs Lowe
&
Brydone (Printers)
Ltd.,
London,
N.W.
10
PREFACE
mathematical treatment
THE
the
work, has
necessary period of growth, has now, thanks to Peano and his followers,
acquired the technical adaptability and the logical comprehensiveness that are
essential to a mathematical instrument for dealing with what have hitherto
been the beginnings of mathematics. From the combination of these two
studies two results emerge, namely (1) that what were formerly taken, tacitly
or explicitly, as axioms, are either unnecessary or demonstrable; (2) that the
same methods by which supposed axioms are demonstrated will give valuable
results in regions, such as infinite number, which had formerly been regarded
Hence the scope of mathematics is
as inaccessible to human knowledge.
enlarged both by the addition of new subjects and by a backward extension
into provinces hitherto
abandoned to philosophy.
PREFACE
VI
An
familiar.
We
that the axioms stated by us are sufficient to prove them.) At the same time,
full proofs are necessary for the avoidance of errors, and for convincing
those who may feel doubtful as to our correctness, yet the" proofs of propositions may usually be omitted by a reader who is not specially interested in
that part of the subject concerned, and who feels no doubt of our substantial
accuracy on the matter in hand. The reader who is specially interested in
some particular portion of the book will probably find it sufficient, as regards
earlier portions, to read the summaries of previous parts, sections, and
numbers, since these give explanations of the ideas involved and statements of
the principal propositions proved. The proofs in Part I, Section A, however,
are necessary, since in the course of them the maimer of stating proofs is
explained. The proofs of the earliest propositions are given without the
omission of any step, but as the work proceeds the proofs are gradually
compressed, retaining however sufficient detail to enable the reader by the
help of the references to reconstruct proofs in which no step is omitted.
though
The order adopted is to some extent optional. For example, we have treated
and relationarithmetic before series, but we might have
To a great extent, however, the order is determined by
cardinal arithmetic
treated series first.
logical necessities.
VU
PREFACE
A very large part of the labour involved in writing the present work has
been expended on the contradictions and paradoxes which have infected logic
and the theory of aggregates. We have examined a great number of hypotheses for dealing with these contradictions many such hypotheses have been
advanced by others, and about as many have been invented by ourselves.
Sometimes it has cost us several months' work to convince ourselves that
a hypothesis was untenable. In the course of such a prolonged study, we
have been led, as was to be expected, to modify our views from time to time
but it gradually became evident to us that some form of the doctrine of types
must be adopted if the contradictions were to be avoided. The particular
form of the doctrine of types advocated in the present work is not logically
indispensable, and there are various other forms equally compatible with the
truth of our deductions. We have particularized, both because the form of
the doctrine which we advocate appears to us the most probable, and because
it was necessary to give at least one perfectly definite theory which avoids
the contradictions. But hardly anything in our book would be changed by the
adoption of a different form of the doctrine of types. In fact, we may go
farther, and say that, supposing some other way of avoiding the contradictions
to exist, not very much of our book, except what explicitly deals with types,
is dependent upon the adoption of the doctrine of types in any form, so soon
as it has been shown (as we claim that we have shown) that it is possible
to construct a mathematical logic which does not lead to contradictions. It
should be observed that the whole effect of the doctrine of types is negative
it forbids certain inferences which would otherwise be valid, but does not
permit any which would otherwise be invalid. Hence we may reasonably
expect that the inferences which the doctrine of types permits would remain
valid even if the doctrine should be found to be invalid.
;
in the
The symbolic form of the work has been forced upon us by necessity
its help we should have been unable to perform the requisite
without
reasoning.
great merit consists not so much in his definite logical discoveries nor in the
details of his notations (excellent as both are), as in the fact that he first
showed how symbolic logic was to be freed from its undue obsession with the
forms of ordinary algebra, and thereby made it a suitable instrument for
research. Guided by our study of his methods, we have used great freedom
in constructing, or reconstructing, a symbolism which shall be adequate to
deal with all parts of the subject. No symbol has been introduced except
on the ground of its practical utility for the immediate purposes of our
reasoning.
PREFACE
Vlll
We
notably
We have to thank the Council of the Royal Society for a grant towards the
expenses of printing of 200 from the Government Publication Fund, and also
the Syndics of the University Press who have liberally undertaken the greater
portion of the expense incurred in the production of the work. The technical
excellence, in all departments, of the University Press, and the zeal and courtesy
of its officials, have materially lightened the task of proofcorrection.
The second volume is already in the press, and both
appear as soon as the printing can be completed.
it
Cambridge,
November, 1910.
W.
will
CONTENTS OF YOLUME
I
PAGE
PREFACE
ALPHABETICAL LIST OF PROPOSITIONS REFERRED TO BY
NAMES
xii
xiii
'
Chapter
I.
Chapter
II.
Chapter
III.
of Part
..
37
....
66
87
90
#1.
#2.
#4..
98
109
115
#&
Miscellaneous Propositions
123
#3.
....
......
Section B.
#9.
91
.....
.
.....
#13.
Identity
#14.
Descriptions
#22.
Calculus of Classes
Calculus of Relations
#24.
The Universal
#25;
The Universal
Classes
213
.....
of Relations
Glass* the
161
187
200
205
#23.
151
187
#20.
127
138
168
173
127
216
CONTENTS
PAGE
232
238
#30.
Descriptive Functions
*31.
Converses of Relations
#32.
Term with
.
respect to a given
242
247
256
#33.
34.
35.
265
#36.
277
#37.
279
#38.
Function
Note to Section
Section E.
#40.
#41.
<#42.
#43.
PART
231
Logic op Relations
Section D.
II.
of
two Relations
296
299
302
The Relations
304
315
320
324
Summary
Section A.
of
Part II
329
331
#50.
333
#51.
Unit Classes
The Cardinal Number
347
#52.
340
1
....
#53.
352
#54.
Cardinal Couples
#55.
Ordinal Couples
#56.
359
366
377
#62.
#63.
386
Section B.
#60.
#61.
388
393
#64.
#65.
On
395
400
410
415
418
Section C.
#70.
given Classes
...
to
420
#72.
#73.
Similarity of Classes
#74.
#71.
426
441
455
468
CONTENTS
Section D.
Selections
XI
.........
#80.
#81.
Selections from
#84.
Miscellaneous Propositions
#91.
#92.
#93.
#94.
#96.
#97.
#95.
#85.
Inductive Relations
#88.
#90.
......
.......
....
.......
#82.
#83.
Section E.
...
ManyOne Relations
......
PAGE
478
483
496
501
508
517
525
536
543
549
558
573
579
588
596
607
623
APPENDIX A
#8.
APPENDIX B
#89.
635
Mathematical Induction
.......
650
APPENDIX C
TruthFunctions and others
LIST OF DEFINITIONS
659
667
Number
Abs
*201.
\~.py~p D
Add
*i*3.
bzp.pDq.^.q
.~p
zq.D .j>vq
Ass
*3
Assoc
#15.
Coram
*204.
Comp
*343.
Exp
*33.
Fact
*3 45.
zpv(qvr).3.qv(pvr)
b z.p .D .gOr:3: q .3 ,p"Dr
b z.pDq.p^r. D :p D .q.r
b .i.p.q.O.r D :p. D qDr
f :. p "5q . D p r D q .r
Id
*208.
b.pDp
Imp
Perm
*3'31.
Simp
35.
3 gO r D zp~q 3 . r
#14.
i : p v q
D qvp
b
zq.D
.pDq
*202.
b t.p
*326.
*3*27.
bzp.q.D.p
b :p .q.^.q
Sum
*l6.
Syll
*205.
qDr.Dzpvq.O.pvr
b z.qDr.D zpDq.y.pDr
*206.
*333.
bz pDq.qOr.^.pDr
bz.
z.pOq.^'.q^r.D.pDr
zqDr .pDq.'S.pDr
:pvp. D .p
*3 34.
Taut
*l2.
Transp
*203
bzpD^q.D.q D~p
*215.
5>
*216.
*217.
*337.
*41.
D ,<^q"Dp
b p D q D ~ qD~p
b ~ q D ~p D p D q
bz.p.q.D.rzDzp.^r."D.e^q
b p D q = ~ q "D ~ p
3J
*411.
bzp =
<^>pDq
q. = .<^p = <**>q
have entailed alteration of the references, which would have meant a very
great labour. It seemed preferable, therefore, to state in an introduction the
main improvements which appear desirable. Some of these are scarcely open
to question others are, as yet, a matter of opinion.
;
The most
definite
is
I,
logic
Section A, of the
one indefinable "p and q are incompatible" (or, alternatively, "p and q are
both false") for the two indefinables "notp" and "p or q." This is due to
Dr H. M. Sheffer]:. Consequentially, M. Jean Nicod showed that one
primitive proposition could replace the five primitive propositions *1'2*3"4'5*6.
From
up of
by a new chapter, #8,
given in Appendix
A to
this
is
replaced
Volume.
that there
is
is
no need
of the distinction between real and apparent variables, nor of the primitive
On
Principia Mathematica,
we have an
all
occasions where, in
or "h .fp" this is to be taken as meaning "rsequently the primitive proposition *1*11 is
"V .fx"
convention that,
when the
"
(x)
<fix
is
it
but in
" I
( gar)
<f>x
to
Mr F.
P.
Ramsey
well
That
is
the
be explicitly indicated
"h . <f>x " is to mean
is to say,
more
all."
is
I,
clearly than
Section
we
Boscovitch, of Gdttingen.
are indebted to
many
MX.
INTRODUCTION
XIV
They
are three in number, two being essential logical novelties, and the third
merely notational.
"
We
introduce in Section
"
(<&p) .fp
"
the
new
Section
B general propositions
to
mean
" (x)
" (x)
<]>x
p " is
<f>xvp."
from Section A.
it
recommended by Wittgenstein f
This
is
to
assume
the
way
is
shown
in Tractatus LogicoPhilosophicus
(loc. cit.
We are
not prepared to assert that this theory is certainly right, but it has seemed
worth while to work out its consequences in the following pages. It appears
required)
that everything in Vol. I remains true (though often new proofs are
it seems that the
but
survives
ordinals
and
cardinals
inductive
the theory of
;
* In his
j
X See Appendix C.
XV
INTRODUCTION
logic has
We now proceed
I.
to the detailed
mathematical treatment.
(x),
R*(x>y)
(x,y, z),
(x, y, z, w),
and
sorts r
triadic relation
(in intension)";
tetradic relation
(in intension)";
them
series
would be impossible.
But
logic is
it is
is
in these cases, the fact that the hypothesis can be framed proves that
true.
itself
belong to
it is
logic.
Given all true atomic propositions, together with the fact that they are all,
every other true proposition can theoretically be deduced by logical methods.
That
is to say, the apparatus of crude fact required in proofs can all be condensed into the true atomic propositions together with the fact that every
true atomic proposition is one of the following: (here the list should follow).
If used, this
infinite,
generality
this
though
is
R&W
possess.
h
INTRODUCTION
Xvi
We
to molecular propositions.
We
Let
p, q, r,
s, t
denote,
which may be read "p is incompatible with q"* and is to be true whenever
either or both are false. Thus it may also be read "p is false or q is false";
or again, "p implies notq." But as we are going to define disjunction, implication, and negation in terms of p q, these ways of reading p q are better
avoided to begin with. The symbol "p\ q" is pronounced: "p stroke q." We
\
now put
~P = P\P
pD q = p~<7
.
Df,
Df,
pv q = .^pl^q
p.q. = .~(p\q)
Df.
Df,
p3q. =
We
.p\(q\q)
of the stroke.
Df.
find that
p.D.q.r. =
Thus p D q
We
is
.p\(q\r).
can construct
for example, (p q)
new
propositions indefinitely
(r\s),
r),
(p q)
p (q
the permutative law (p q) = (q \p) but not the associative law (p\q)\r =p\(q\r).
(These of course are results to be proved later.) Note also that, when we
construct a new proposition by means of the stroke, we cannot know its truth
or falsehood unless either (a) we know the truth or falsehood of some of its
\
r,
(q
q, r, s.
are substantially
propositions of
For what follows, see Nicod, " A reduction in the number of the primitive
3241.
xix.
Vol.
pp.
logic," Proc. Camb. Phil. Soc.
*
INTRODUCTION
We
different propositions.
XV11
have in fact
<l)
by
method
this
"P q"
Thus molecular
propositions are
all
of the form
but
q,
of times.
hold
is
still; i.e.
If p,
q,
This
is
a primitive proposition.
We
now
can
p and p
(q
r),
we can
When
infer
r.
a molecular
manner,
it
which
of
is
p with
always true.
itself,"
It
which
is
The simplest
instance
P\(P\P)>
means "p is incompatible with the incompatibility
obvious.
Again, take
Again, take
"~jp.D.~pv~ q."
Again, "p
D .p v
This
(P
q"
"p
p."
This
is
P)
is
is
q)}>
is
p\i{(p\p)\(q\q)}\i(p\p)\(q\q)}l
p and q may be chosen. It is the
fact that
we can
important to
logic.
Logic
is helpless
62
INTRODUCTION
XV111
known
empirically.
known
evidence.
The laws
q, r, ...
may
F(p,q,r,...),
q, r, ...
may be "
is
is
(p,q,r, ...).F(p,q,r,
When
such a proposition
beginning.
true,
denoted by
is asserted,
we
shall
...).
...)" at the
Thus
"\.F{p,q,r,...V
q, r, ...
may
...) is
true
be.
The
logic,
is
logic,
is
is different.
rule
is
infer " h
H{p,
q,
may
be,
both
..)},"
q, r, ...)."
will
meet us
later.
the above
Nicod has shown that the logic of propositions (*1 *5) can be deduced,
by the help of the rule of inference, from two primitive propositions
and
\
.pDq.D.s\qDp\s.
m
pDq.DzqDf^s.D.pD^s,
XIX
INTRODUCTION
which
is
{p\(9\q)}\[{(s\q)\((p\s)\(p\s))}\{(s\q)\((p\s)\(p\s))}].
Nicod has shown further that these two principles may be replaced by
one. Written wholly in terms of the stroke, this one principle is
bl(gr)}[{*(**)}Kl)l((pl*)l(pl))}]be seen that, written in this form, the principle is less complex than
the second of the above principles written wholly in terms of the stroke.
When interpreted into the language of implication, Nicod's one principle
It will
becomes
p.0.q.r:^.tDt.s\qDp\s.
In this form,
it
pDj.D .s\qDp\s,
but in
complex.
itself it is less
From
we add one
proved, provided
proposition
(p, q,
r, ...)
F (p,
q> r, ...),
we may
viz. that,
substitute
given a
for p, q, r,
...
/,0>,
and
q, r, ...)
assert
(p,q,r,...).FUi(p,q,r,
where
f / f
1}
3 , ...
...),
Since
the former assertion applied to all elementary propositions, while the latter
applies only to some,
it is
II.
1.
We
R R R R
Here
2,
lt
...
Rm (x
xlt x2
a's
1}
#2
...
>
xm
Rn (x x
m is not equal
occur in
even
if
1}
...
xm )
xn ) can also occur as one of the x's in R m (x x2
n.
Terms which can occur in any form of atomic
x
. .
to
proposition are called " individuals" or " particulars"; terms which occur as the
is
XX
INTRODUCTION
Given an atomic proposition n (x1} x2 ... xn ), we shall call any of the x's
a "constituent" of the proposition, and R n a " component " of the proposition*.
We
shall
Rn (x
1} x2
parts of
q,
Definition of
2.
individual a
We
We may
call
namely
all
those that
We
When
<?> 9 P> 9
9> where q is the argument. Similar considerations apply when an
argument occurs more than twice. Thus p\(p\p) is a value of q\(r\s), or
P
9
(r
" ^
or 9
<l)>
(<?
r),
Fp," the
or q\(r\
r),
or q\(q\
is
q).
When we
to be varied whenever
(P)
p
assert a proposition of the form " (x)
assemblage indicated by
<f>x
it
meaning
<f>x,"
assert a proposition
occurs.
" all
We may
similarly
propositions of the
is
to be
varied.
3.
may happen
it
may happen
its
that
values
all its
is true.
we wish
to assert
it,
in
"h.(f)(x,y,z,
* This terminology is taken
is
...)."
from Wittgenstein.
written
it
all
XXI
INTRODUCTION
We
have already had assertions of this kind where the variables were elementary propositions. We want now to consider the case where the variables
are individuals and the function is elementary, i.e. all its values are elementary
propositions. We no longer wish to confine ourselves to the case in which it
is
to
make the
we
are true;
...)
desire to be able
proposition
(x y,z,...).<\>{x,y,z,
...)
(i.e.
"(a*)
Similarly the proposition that
il
(x, y, z,
<f>
<K'
. .
"sometimes true"
.) is
We
denoted by
(^x,y,z, ...).4>(x,y,z,...)."
)
. .
<f>
(x, y,z,...)
and (3a;, y, z,
Consider
two
is
...).<(#, y,
first
z, ...
),
a function of
can form
<t>
<i>
>
<t>
<f>
(y
We
a,
<f>
( a,
y)
and
(33/)
. </>
(a
y).
a a>)
(y)
<f>
(33/)
<\>
0*.
2/)
thence to
(x, b),
<f>
. <j>
(y)
{as)
. <j>
(x, y), (
(a*)
<f>
(x, y),
(ay)
(a)
(y)
(ay)
<f>
<f>
(x,
y)
te 2/)
(y)
(ay)
()
=
4> (a?,
(a*0
(3*0
0*.
2/)
We
have
y),
te y)
4>
are no other equivalences that always hold. For example, the disbetween " (x) (gy) <j> (x, y) " and " (gy) (x) <f> (x, y) " is the same
as the distinction in analysis between " For every e, however small, there is a
"
8 such that..." and " There is a 8 such that, for every e, however small
But there
tinction
XX11
INTRODUCTION
first
vary
x,
constant.
to the left of " (x) " or " fax)," the latter in the converse case.
The grounds
with the method of
matrices, which brings order into the successive generation of types of pro
from the
us,
start, to deal
not elementary.
will
it interferes
Take, for
types,
and that
y),
.
it
requires
which are
v q." This
be
\:.(p):.(q):q.D .pvq,
h:.(q):.(p):q.D.pvq,
or
and
or
(p)
This makes
we wish
it
q D p v q considered as a
:q.D .pvq considered as a
function of
q.
function of p,
seems necessary to
start with
(pi&l, x%,
we
Hence
an elementary function
x3
...
xn ),
<f>
(a^)
"
y) "
"
(y)
4>
0> y)
" is
(x,
is
fax)
(y)
case. (Some of these, however, are equivalent.) The process of obtaining such
propositions from a matrix will be called " generalization," whether we take
" all values " or "
"
some
value,"
general propositions."
We shall
later
we may
therefore say
A " matrix " is a function of any number of variables (which may or may
not be individuals), which has elementary propositions as its values, and is
used for the purpose of generalization.
INTRODUCTION
A
We
shall
XX111
is
4.
\.F(p,q,r,...y
where
place
F is built up by the stroke, and p, q,r, ... are elementary, we may rethem by elementary functions of individuals in any way we like, putting
P == Ji\ph)
q =j2\xii x2>
z
and so
on,
all
xlt x2
...
xn
What we
thus assert
take
x n)>
is less
(Any two
or
...
may
be identical.)
I .
#811.
I .
(g#, y)
fax)
(<f>a;
<f>a
<f>x
(<pa
and
<f>y)
<f>b)
to
<pa
and
(x)
<j>x
fax)
<j>a
<px
.
<f>b.
These two primitive propositions are to be assumed, not only for one variable,
but
for
(x1} x2
The
...
(<*!,
xn ).
proposition (x)
a2
an ) D
...
#2
<(#i,
.
<f>x
proving existencetheorems.
...
<j>a
But
v
x2
(g#i,
xn ). D.
<f>a
(g#)
or
<f>b,
<f>x
<f)b
xn )
^>(ai, Oa,
<f>
...
{x1>
x2
...
xn ),
&)
may be
it
...
v
.
written
<f>a
fax)
<f>b
<f>x,
For purposes of
inference,
we
<f>
<j>x
and
infer q,
even
(x)
INTRODUCTION
XXIV
\.f(x,x).
Since
<f>x
fay)
<f>y,
we can
infer
May) /toy).
i.e.
H:(#):(a2/)./(a;,y).
Similarly
Again, since
<j>
(x, y)
(y)
faz, w)
and
'
We may
sitions
w\ we
y).
can infer
y) "/(* y)
(a^
y).
<f>x
(z,
<>
H. (ay, )/(*,
We
by a simple example.
Hence, substituting
fax) .f(x,
for p,
{x)
<f>x
<f)X.
b
I"
rfa)
(y)
fax,
fa D
fax) <f>x D
(ay)
y)fa^
<f>y,
cf>y,
4>y
Apart from
special
III.
etc.,
for
Again, fax) x and (x, y) . $x O yfry enable us to infer (y) tyy; here again,
we wish to be able to state the corresponding implication. So far, we have only
.
But
practically, it is
This
of strokefunctions.
and there
is
propositions of #1
*5.
Hence
q, r, ...
introducing
The fundamental
shown
By
#1#5,
are elementary.
it will
XXV
INTRODUCTION
When
limited scope.
Scope
indicated by dots.
is
The occurrence
1(3*0
fa)
{(ay)
9.
is
definitions:
{(x).<j>x}\q.
These
replace *9 in Principia
tyy)
= .fax).<f>x\q
= (*) fa ?
Df,
Df
Df
>
(y)
what
fy
is
it
(0)
{fax)
The
<f>x}
fa]
<f>x}
is
.^ry}.=
[(y)
\{y)
yjry]
Thus
=
=
Kay) fy} =
=

is
fax)
<f>x\
(3*0
(32/)
(3*0
fa
(3*0
(y)
(x)
^y}
fa t2A
I
{(ay)
fay)
is
{(y)
fa
<f>x
M
tyy>
fy.
the last
obtained
And
is
true in the
(x)
is
then
true.
same circumstances.
is
way
i.e.
to
the fact that x only occurs on one side of the stroke and y only on the other.
The order of the variables in the prefix is indifferent whenever the occurrences
of one variable are all on one side of a certain stroke, while those of the other
are
all
(y)
it.
We
x (x
>
y)
iy)' (3*0
0> y);
XXVI
INTRODUCTION
is
side.
But we
do have
(ft)
The
iy) <l>a!\yjry:
'
(y)
(a*)
$x\^y.
when they
is
it is
convenient
proposition
applicable.
is
[{(a*)
<H
{(y)
= p
"fy}]
(a*)
(y)
All that
is
from
When
Take
y.
e.g.
necessary
is
written
yjry
(<j>x
or
{fx (fy

<f>x
yjry)
tyy)}
(y)
and
'
(a 57 )
(y)
(a)
[tyy
all
(a)
(<*
{<t>
iry)
1ry)>
ifry.
<f>x)
(Qx
<f>x)}.
all
those
v tyy
be changed.
~ v ^ ^y
.~^pv<v yfry
x v "tyy
4>x
V <^
{$y
By means
x ^y]
the matrix
may be
(y)
4>
<f>x
This
(ay)
[0*0
e.g.
we
<f>
and
ir.
p by
<f>x
or
<f>
(x,
y) or
etc.,
p
we must
(O)
treat it as a case of p
{()
(ay)
{(x)
* (> y)
(ay)
.
</>#},
and
]
>
first
eliminate
x.
Thus
f ( y)} =
<f>x)}
function.
INTRODUCTION
The
Thus
definitions of
~ {(x)
~ p, pv q, p .q, pOq
<f>x\
{(x)
(a0
(x)
p D
.
<f>x
\(x)
{(x):

(33/)
<f>x}
<f>x\
(<#
4>y\
fox <y),
p'l [{<*)**} {(*)**}]:
P {(3*0 (33/) (4& #)}
(x) (y) .p (<f>x <y),
(x) .</>#:
fa]
(rx)
xxvu
(y)
(x)
D p =
<f>x
{(#)
<f>x]
(rx)
(x)
<f>x
<f>x
\(p\p):
(p p) =
:
p v
.
(x)
(gar) .<f>xDp,
p = [~{().^}] ~p:
<f>x
(x):(y).(p\p)\(<f>x\<f>y).
be seen that in the above two variables appear where only one might
have been expected. We shall find, before long, that the two variables can be'
reduced to one i.e. we shall have
It will
(3*0
(32/)
(y)
(a;)
These lead to
<#
<f>x
~ {(x)
~ {(a^)
<f>y
<j)x}
4*}
=
=
'
<f>V
(3)
(a;)
4>
x *
I
<#
<j>x.
= (a) ~ fa,
= (#) ~ <j>x.
.
But we cannot prove these propositions at our present stage nor, if we could,
would they be of much use to us, since we do not yet know that, when two
;
may be
we have a
therefore, suppose
and we wish
" (z)
Thus the
<f>z."
strokefunction in which
to replace
(p
shall have to write the second occurrence of

p),
"
(y)
p by
<f>y,"
(x)
<f>x,
we
many
separate
The
They
we can
(4)
I
I
(a, y)
(g#)
$a
<f>x
tyx,
(<f>x
(<fxi
The extended
infer (x)
are as follows:
<f>b), i.e.
(x)
rule of inference,
even when
y by a certain
\:<f>a.D
<j>y), i.e.
<
and
<f>x
<f>a
<f>x.
.
<f>b.
i.e.
yfr
(ga?)
<j>x
i/r#
INTRODUCTION
XXViii
For (g#) (y) <f>x fy we can substitute (y) : (ga;) <f>% yjry, and
versa, even when this is only a part of the whole asserted proposition.
but
for
By means
vice
The above
variable,
any number.
the propositions of #1
it
all
p,q,r ... involved are not elementary. For this purpose, we make use of the
work of Nicod, who proved that the primitive propositions of *I can all be
t
deduced from
h
b
and
Op
.pDq.D .s\qDp\s
.p
Thus
when
p,
all
r),
we can
infer r."
we have
q, s,
Appendix
p and p\(q\
or
IV.
FUNCTIONS AS VARIABLES
is less
such
It is necessary
now
to consider
cases.
propositions.
just as
The values
is
still
elementary propositions,
As we
shall
We
we
replace
a?
by a constant
a,
(<f>).cf>l a,
(a<)
<f>
a.
propositions as
shall
INTRODUCTION
XXIX
These are not elementary propositions, and are therefore not of the form
! a.
The assertion of such propositions is derived from matrices by the method of
#8. The primitive propositions of #8 are to apply when the variables, or some
of them, are elementary functions as well as when they are all individuals.
</>
<f>
x,
i/r !
y,
its
z,
up by means of the
built
To obtain a
values*.
in place of p,
..
We
stroke.
q, r, ...
can then
<f>,
is
all
the characteristics of a
great general."
Taking the
all
"
a certain collection of propositions, generalized in " (x) . x loves Socrates " and
"(qx) . x loves Socrates." But " <f> ! Socrates" picks out those, among elementary
Socrates"
mention Socrates. The generalizations "(<)
propositions
which
and (a0)
Socrates " involve a class of elementary
"
"<j>
! Socrates
cannot be obtained from an individual variable. But any value of
is an ordinary elementary proposition
the novelty introduced by the variable
propositions, which
"
<f>
</> !
^> is
(x)
x loves Socrates,
(<)
(f>
new
On
among elementary
propositions.
#5.
A variable function which has values that are not elementary propositions
starts a new set. But variables of this sort seem unnecessary. Every elementary
proposition
is
a value of
(p) .fp. =
(<,
</> !
&
therefore
*)./(*
x)
(gp) fp
.
(a</>,
x) ./(<
x).
assumption is fundamental in the following theory. It has its difficulties, but for the
moment we ignore them. It takes the place (not quite adequately) of the axiom of reducibility.
It is discussed in Appendix C.
the matrix must be generalized. In other
f In a proposition of logic, all the variables in
general propositions, such as "all men are mortal," some of the variables in the matrix are re* This
placed by constants.
XXX
INTRODUCTION
We
have the
shall
all classes
now
Consider
x,
keeping
of propositions
one individual
<
y) or
<f>
on, until
vary
<f>
(x,
the matrix
first,
but
<f>(a,y),
"(y):(^as)"
involves
<f>
fixed (which
I
x,
<f>
y,
<f>
. . .
we have done
we make another,
But now suppose we
class,
all
varies.
for another.
and so
<f>
<f>
adopted.
vary
(a,
<f>
(x, y),
class
y) as
< (a,
giving classes
<f>
<f>
a,
i.e.
all
I b
and so on. The set of propositions
which are values of < ! a is a set not obtainable by variation of individuals,
i.e. not of the form fx [for constant / and variable x\
This is what makes
a new sort of variable, different from x. This also is why generalization of the
form (<f>) F (<f> 2, x) gives a function not of the form /! x [for constant /].
Observe also that whereas a is a constituent of/! a, /is not thus the matrix
x has the peculiarity that, when a value is assigned to x, this value is a
a constituent
<f>
<f>
<f>
when one
results
for another.
We
of individuals.
fl(<f>lz, ^r
\z,xlz,
...
x,y,z,
its values,
<f>
x,
...).
<f>
<j> !
(e.g.)
y,
<f>
z,
..
or of
<f>
a,
! z, yjr 1
, x %> > V> z and no others, if it is of the sort
<f>
that can occur explicitly in logic, will result from substituting <f>\x,<f>\y,$\z,
yfrlx, yfrly, yfrlz, %lx, % 1 y, % I z, or some of them, for elementary propositions
in
some
strokefunction.
INTRODUCTION
It
is
matrix
logical
or,
is
as
XXXI
"
we may
matrix."
call
is
" logical
it,
Thus p q
matrix
a logical
is
so is
<f>
x,
where
<f>
variable.
purposes
for different
but
all
merely an illustration of
The test of a logical
expressed
without
introducing
any symbols other
can be
above denned.
This
is
matrix
that
is
than those of
it
we must not
logic, e.g.
Consider
the expression
/!
(<f>
z,
yfr I
z,
z, #, y, z).
new
sort,
because
>
y>
>
containing the
takes functions
it
among
When
arguments.
its
. .
a value
assigned,
is
we obtain an elementary
alone, we obtain a matrix
a value to
<
There
which
<f>
alone,
is
! 1z
a, b, c, ...
and
<f>
is
no
#,
we
we
we
<
If
x.
but
if
This
we
we
is
give
give
logical
(<f>
2).
<j> I
<f>
b,
<f>
c,
. .
in
where
are constants; but these are not logical matrices, being derived
from the logical matrix <f> \x. Since <f> can only appear through its values, it
must appear, in a logical matrix, with one or more variable arguments. The
simplest logical functions of
<f>
fl(<f)lz, a?i,#2
is
<f>
x and (a)
<f>
x,
but these
matrix
logical
...
xn )
F(pi,Pz,Ps> >..p n )
by substituting
<p
xlt
(f>
x2
. . .
of r and
. . .
some values
s.)
R&W
The
first
of these is that,
when a value
is
assigned to
c
XXXU
INTRODUCTION
/, the result
may be a
second
first
three variables,/,
number)
point first:/!
and
The
(<j>
is
logic.
z, x), for
example,
a matrix containing
<j>lxD<f>lx, etc.
is
(among an infinite
by assigning a value to/:
x, (<j> x) (<f> x),
Similarly <f>lx2<f>ly, which is a logical matrix, results from
<,
x.
<f>
/ is
y),
(#, y>z),
Rn (a,b
b2
...
6 n_j),
a value of "
proposition
and
still
<f>
Rn (a, b
1}
<f>.
intolerably complicated if
of variable functions;
it
it
By assigning < and x in/! (< ! , x), while leaving /variable, we obtain
an assemblage of elementary propositions not to be obtained by means of
variables representing individuals and firstorder functions. This is why the
new variable /is useful.
!
INTRODUCTION
We
XXX111
...}
ele
V.
When we
(,
'
V)
(<)
F {<
z, x, y, ...},
<f>
can take are matrices, so that functions containing apparent variables are not
We
included.
can, if
functions such as
we
I ot,<f>
new
introduce a
like,
(y).<j>l($,y), (y,z).<f>l(x,y,z),
(ay) <!(,
...
y), ...;
by generalization
Let us denote any such
function by fax, or ty^sc, or Xl x, or etc. Here the suffix 1 is intended to indicate that the values of the functions may be firstorder propositions, resulting
as values of single
variables.
.fax. =
(<f>).
<f>lx: (fa y)
.xf)
(x,
y)
(0)
(ay) .<f>l(x,y):
etc.,
(fa) ./!
where/!
(fa%x)
{(y)
a value of
<j>,
<f>
say
=.
(<f>)
./
<f>
a, occurs,
3
(f> ,
fa,
substitute (y)
(</>)
./!
(</>)
/
similarly.
{(ay)
{(y,
{(y)
(w)
w)
'
(<f>)
./! {(ay)
. </> !
<
.
(a, y),
(z, y), x]
(% y, w), x)
<f>
{<
z, x},
is
>
<f>
(a w )
(33/>
<f>
w)
*(% y, w),
>
(w)
x]
(32/)
etc.,
We
define
Then
is
Similarly put
(fa) ./! (fa lz,x).
where
structed.
<j> !
(z, y), x) is
I
(< )
z, x),
<}>
and
x,
c2
INTRODUCTION
XXxiv
this,
is
not applicable.
such variables as
to
in
manipulate
fa.
No
therefore adopt
We
yjr 1}
...
ft,
may
/, (faz, ),
in
argument to
which
is (ay)
xRy, may be treated
the function. For example, x e D'JR,
x.
In virtue of #8, fax may be
without danger as if it were of the form
substituted for < ! x without interfering with the truth of any logical prox is a part. Similarly whatever logical proposition holds
position which
variables involved in a function are not of higher order than the
.
<f>
<f>
concerning fx (faz,
will hold
concerning/! (faz, x)
(3*)
x).
/! (*!*,*).
These are functions of x, but are obviously not included among the values
is the argument). If we adopt a new variable fa which is
x (where
for"
z can be an apparent variable, we shall obtain
functions
in which
to include
<f>
<f>
(f>
other
new
functions
ifa).f\{fa%x),
among values
for
(afc) ./!(#*,*)>
fax (where fa
now
involved,
is
is
which was formerly involved. However much we may enoccurs as apparent variable
large the meaning of <f>, a function of x in which
has a correspondingly enlarged meaning, so that, however <f> may be defined,
of values of
<f>
<f>
embraces among
We
there
is
its
values
all
variable,
and so
on.
XXXV
INTRODUCTION
The essence
of the matter
that
is
variable
a
may
travel through
any
well
defined totality of values, provided these values are all such that any one can
replace any other significantly in any context. In constructing fax, the only
totality involved is that of individuals, which is already presupposed. But
when we allow <j> to be an apparent variable in a function of x, we enlarge the
a;,
<f>
<j>
<f>
definitions of *8, together with the principle that a function can only occur
through its values. In virtue of the principle, a function of a function is a
And
complete asserted proposition should be derived from a matrix by generalization, and that, in the matrix, the substitution of constant values for the
variables should always result, ultimately, in a strokefunction of atomic
say " ultimately," because, when such variables as fa% are
propositions.
admitted, the substitution of a value for fa may yield a proposition still
We
We may
It seems, however,
functions fa, fi
Let us take
them.
a very simple
first
illustration.
>
<f>
made by
(g<)
But
in fact f{x, a)
<f>
. <j> !
./ (x, a).
a,
(ay)
(y)
Suppose
first
4>
0*> y).
(a*)
that fax
fax
<M
faa
to *)
(y)
=
D
<f>
tf>
y, z), ...,
fo y.*).>
(ay)
0 y. *)>
.
0,
(*)
(x,
y)
(y)
(y)
<f, I
(a</>)
<j> I
<
foy z^
Then
(x, y).
<
a.
tf>
(a,
y)
XXXVI
INTRODUCTION
=
3
D
fax .faa.
because
<j> I
(%,
y) v
Then
.<f>\{x, y).
O,
faz) .<f>l(a, z)
(gy)
'
(g;0)
(f>
<f>
a,
. <j> !
of the form
(x, z) is
. <j> I
?/)
I
<f>
when y and
x,
z are fixed.
It is
obvious that this method of proof applies to the other cases mentioned above.
Hence
fafa)
We
fax
faa
(>&<j>)
<f>
a.
<f>
can satisfy ourselves that the same result holds in the general form
by a similar argument.
strokefunction
F(p,q,r,...)
by substituting
a, </> !
(where a, b, ... are constants) for some of
<f>
<f>
the propositions p,q,r,... and g x l x, g2 lx, g3 lx, ... (where ^, 2
# gs ... are
constants) for others of p, q, r, ..., while replacing any remaining propositions
x,
b,
. . .
p, q, r,
...
suppose
= .(<f>la)\{(<f>lx)\(cl>lb)}.
fl(<l>lz,x).
We
that fax
first
(y)
<$> !
<f>
(x, y) is
(&
fab)
(y)
(a>r)
.=
putting \jrlx
with.
Hence the
of the form
(33/)
<j> !
yjr !
a (^ ! x
i/r
<f>
{<f>
!
x.
Then
(x, y).
w)}
{<f> !
faa
Then
(x, y).
(a;,
*)
<f>
(6,
w)}
b),
(ftl(x,z)v<j>l(x, w).
result follows.
Here
it is
(fa% x)
(<)
./!
(<
X x).
i.e.
(fa).f\(<t>l%x).1.(fa).f\(fa%x).
This follows from the previous case by transposition.
dependently as follows. Suppose, as before, that
fl(fa$,x).
and put
Then
first
fax
: (
H y)
It can also
be seen
.(faa)\(fax\fab),
(y) .<f>\(x, y).
(z,
(6, /)}.
in
XXXV11
INTRODUCTION
Thus we require
that, given
(ylr).(ylrla)\(yfrlx\^lb),
w)
(z,
.
<f>
(6,
<j> !
w)}.
Now
(yft)
yfr
a
\
(yfr
yfr !
b)
: .
!
<f>
(a, z)
<f>
:.
!
<f>
(a, *)
<
(#, z)
(a,
<f>
<f>
w) D
(a,
w) D
.
<
(b, z)
w)
(x,
<
<f>
(6, /) :.
w)
:.
>
~^>!(a,?).D:<^!(a,w;).D.</>!(,5).^!(6 w) (2)
Also
(l).(2).D:.(^).^!o(^!ar^!6):D:.(ay):^!(a,y).D.^!(a?,).^!(6,w)
which was
be proved.
to
Put next
Then
fax
The method
will
(33/)
<
(x, y).
z= w = y
and the
(a, y)
{</> !
(x, z)
<f>
(6,
w)}.
result follows.
Hence generally
(fa)
<j> !
Since the proof can only be conducted in each separate case, it is necessary
to introduce a primitive proposition stating that the result holds always. This
primitive proposition
h
is
:(*)./! (01
% x).D.fl(fa%x)
Pp.
<f>
<f>
<f>
containing no apparent variables except individuals, the above primitive proposition enables us to treat it as if it were denned by a matrix.
We
fax
We
want
./!
(<)
(<f>
% x)
to discover whether, or
(fa)
or fax
.g\(4>\^x) .1
gl(<f>lz,x).
.
(<f>
z,
x)
= a,
(gtf) ./!
(<f>
% x).
Then (fa g
g\(faz,x).
case.
Put
.<f>laD<f>lx.
according to #131.
(A)
INTRODUCTION
XXXV111
We
want
.
(<j>)
i.e.
to prove
aD
<j> I
<f>
<f> 2
aD
x,
(tf>
/!
(<*>
(a<*>)
Now/!
<f>2
(<f>).<t>laD<f>lx.3: () ./!
(0"! 2,
D (<f>) ./!(</>! % x)
l%a).D. ( a <) ./!(<! *, *).
a)
z,
a?)
F(p,q,r,...)
by substituting
some of
for
p, q,r,
As soon as
are constants.
b, c, ...
(<f>).(j>laD<f>lx.D
the values
...
is assigned, this
<f>
:/!
:(<!>)
(<
<j> I
is
% a) D
x,
./!
tf> I
b,
(<f>
c,
of the form
yfr
% x)
where
. . .
#.
Hence
D:(*)./!(*!*,a). 3. (*)./!(*!*,*):
Thus generally
axiom
It
is
(<)
</> !
<
(a<*>)
/'
x D
(<*> I
(<2 )
<f>
% a) ^
2a D
</> 2 a?
(3*) /!(* ! *, *)
without the need of any
of.reducibility.
must
not, however,
is
some strokefunction
F(p,q,r,...)
by substituting
for
being constants).
4> 2
Thus
<f, 2
What we want
x.
x.
x,
a,
<j> !
<f>
b,
...
(a, b,
...
= .{<j>).f\{4>\z,x).
= .(<}>). F((j>
to discover is
<
(B)
whether
{<\>).g\{^\%x).^.g\{^%x).
Now
(<f>
z,
G(p,q,r,...)
by substituting
x,
<f>
If
G
<f> 2
is
((f>)
then g
<f> 2
b', ...
x,
a,
<f> 2
some of p, q, r,
... in G(p, q, r,
for
<f> 2
b',
x)
is
known
x)
is
by substituting
to
for
some of
To obtain
...),
instead of
a new matrix.
z,
(p, q, r, ...)
a',
...
<f>lb',
<f> 2
z,
(<f> 2
((f)
!
put
We
x, <f>la', <f>lb',
true,
<j>la',
to
r, ...
p, q,
q, r, ...) is
it
is
always
obtained from
the propositions
<f> 2
x,
warranted.
We
Whenever
(</>)
gl(<j>lz,x)
known
is
to
be true because g
(<
z,x)
is
G(p,
which
is
true for
(of course) is
(< 2 )
values of p,
all
.
(<f> 2
z. x).
q, r, ...),
q, r, ...,
then g
(<f> 2
lz,
x)
is
also true,
and so
XXXIX
INTRODUCTION
where
may be
When
We
g (<f> ! 2, x) is not a
some values of x and
!
(< 2 2,
x)
is
some
(<j>)
true for
CLASSES
VI.
in another
all
i.e.
<t>x= x
This
is
of
for p,
<
for
q, r, ...
+x.l.f(p)=f(1rt).
obvious, since
<f>
as the substitution of
yfrx.
same truthvalue
Consequently there
We
= x tyx D
.
<f>%
we
is
to the truthfunction
= yjrx.
(<$>x),
On
we now have
aC/3./3C7.D.aC 7
will hold
whatever
may be
the orders of
a,
Take, as a
#,
first
x ep f K = a e k . D x e a.
.
Thus p'tc is a class of higher order than any of the members of k. Hence the
hypothesis (a) .fa may not imply f{p'ic), if a is of the order of the members
There is a kind of proof invented by Zermelo, of which the simplest
k.
example is his second proof of the SchroderBernstein theorem (given in #73).
This kind of proof consists in defining a certain class of classes tc, and then
of
showing that
p'tceic.
On
the face of
it,
"p'/ce/c"
is
is
INTRODUCTION
Xl
A class
said.
of classes k
Ox, x2
where
is
true.
It
may
stituted for
...):
(gyx y2
. .
.)
x2 e a,
Ffa e a,
"oe" means
a strokefunction, and
a,
k.
to be
is
is
. . .
yx e a, y2 e a,
. .
.),
is
happen that the above function is true when p'ic is suband the result is interpreted by #8. Does this justify us in
well
is
K = a (R"a Ca.aea).
Then
R"p'/cCp'K
so that, in a sense,
defining function of
k,
That
e k.
aep'/c
(see *40'81)
we
to say, if
is
substitute
definition of #90,
pK
for a in the
proposition.
By
the
4
R%. t a=p t K.
<
Thus
R%a
(a) .fa,
is
a secondorder
where a
is
Consequently,
class.
a firstorder
if
we have a hypothesis
we cannot assume
class,
(A)
(a)./a.D./CR*'a).
By
if (a)
./a
is
deduced by
logic
f(R%a)
(a) ./a is
may
be true. Thus we
will also
substitute
R#a
for a in
any asserted
proposition
" h
For example,
if
= a (R"a C
e a),
we have
Hence
many
In
a,
R"{ar\ 0)
C /3 a e
.
.p'ic
.ae/3.
(1)
/3
we
whence we obtain
R"(/3np<,e)C/3.ae/3.D.p t feC/3
(2)
i.e.
aR%z D w w e y8 a e
aR%oc D x e /8
aR^x D :. z e /3 aR%z Dz w .M/e/3:ae/8:D.#e/3.
z
or
This
is
aR%x.
Z)
for a in passing
valid,
p'tc
from (1) to (2). Therefore the proofs which use this form of
induction have to be reconstructed.
INTRODUCTION
It will
xli
be found that the form to which we can reduce most of the fallacious
seem plausible is the following:
inferences that
f(x, %)" we can infer " h : (x) : (gy) . f(x, y)." Thus given
(a) ./(a, a)" we can infer " V : (a) : (g) ./(a, #)." But this depends upon
Given
"
I .
" h
(x)
and we wish
The
to infer g$,
Da
/8
of another,
we do
got.
where
# is
D a #a D
/3 e k.
proposition
:.
(/3)
/e .
/3 e
/c .
gryS
This
is
is
only valid
::
if
(ga) :.ae
=$
.3 .ga.D
re
* = 3(a
p'tc e
plained above.
We
.D
.g/3.
inference
is fallacious if /3
a.
and we prove
fie k
Hence the
is possible.
C D'R
have a
class of classes
which
(#73"81),
to Zermelo's
We
ff.
is
x~e(0Q.'R)vIt"p t ic
l
we make a use
proof
i'x e k
to p'x Cp'tc
which
of it
only valid
For,
same
i'xe k
because
x,
if
when a e k
(a)
This
*7382).
D a p'k C a.
inference from
a
is
But
is
ae k
The
is
is
D a .p
:::
i'x
p
Da p
/c
(g/S)
Ca
(x)
::.
deduced from
a
p'/c t'xe k
to
.0
p'/cCp'ic
i'x
is
is
::
k.
:.
k D
:.
ft e
D .xe /S D x e
D xea
a.
principle that /(a, a) implies (g/3) ./(a, /3). But here the fi must be
same order as the a, while in our case a and /8 are not of the same
order, if a = p l K i lx and /3 is an ordinary member of k. At this point, therefore, where we infer p l K Qp'ic fc'#,*the proof breaks down.
by the
of the
It
is
easy, however, to
remedy
x~e(@<l'R)u
or,
conversely,
R"p'/c. D.x~ep'/c
All
we need
is
INTRODUCTION
Xlii
Now
x ep'/c
:.
Da
3a
i'x^e k
^(p _ a f jR C a 
X :xe0 Q.'R
D
:.
 CF#
xe
i'a>)
~ [R"(ol  i'x) C a 
v xe R"{<t
t'?}
l'x)
D a x e i"a.
.
Hence, by *72341,
a;
e><*
result.
indefinitely.
= /(/3Ca)
Cl'a
Now "0Ca"
members
survives.
We
put
Df.
significant
is
are of the
its
0Ca.Dp.ffr
the
/3
must be
of
some
definite type.
As a
we
rule,
be able to show
shall
that a proposition of this sort holds whatever the type of /3, if we can show
Consequently no difficulty
is of the same type as a.
that it holds when
arises until
W
to Cantor's proposition 2
we come
>
n,
proposition
~{(Cl'a)sm}
which
is
The proof
proved in #102.
WR
=a
R e 1  1
^
ye a. ye R'y .Oy.y^eg.yeu.
.
<J'i2
y~e
is
as follows:
C Cl'a
w
R'y .D y
= x [x e a
 R'x) D
.
.ye^:D:yea.D y .^ R'y
D :<<!<#.
As
Let a
= (A
x),
and
we
somewhat minutely.
let
xR${4>\z)). = .f\{4>\%x).
Then by our
data,
Alx.D.
(?[<!>).
fl(<j>l%x),
fl(<f>l^,x).D.Alx.cf>lyD y Aly
fl(<f>lz,x).fl(<l>lz,y).O.x
f
With these
(<f>
% x).f\{^r\%x).0.<i>\y =y.^
y.
data,
xea R'x =
.
Thus
>
= y,
'.
= {(<) A
'
A
'
x :/!(<!
:
(4>
'
2.
z,
x)
D^
<j> I
x ) D ~4> ' x )
x.
INTRODUCTION
xliii
<f>,
that
is
sition
With regard
classes.
variables,
is finite.
two
and we have
0(,)
= ^(,#) =
.
:<f>{x,y)
=,, .f(x,y).
The difficulties as regards^'X and Rl'Pare less important than those concerning
l
p'/c and Cl'a, because p \ and HYP are less used. But a very serious difficulty
occurs as regards similarity.
a sm
We
= (rR)
.
have
i2 e 1
*
D'i
/3
= (Pi*.
Here R must be confined within some type; but whatever type we choose,
can be correlated.
there may be a correlator of higher type by which o and
Thus we can never prove <~(asmy8), except in such special cases as when
either a or
2n
> n, which
is finite.
cerned in proving that two classes are similar, and these can
all be interpreted
remain valid. But the few propositions which are concerned with
proving that two classes are not similar collapse, except where one at least of
the two is finite.
so as to
MATHEMATICAL INDUCTION
VII.
E and
But the
proofs
when
of many of them become fallacious when the axiom of reducibility is not
assumed, and in some cases new proofs can only be obtained with considerable
labour. The difficulty becomes at once apparent on observing the definition
of "xR%y" in #90. Omitting the factor "xeC'R," which is irrelevant for
our purposes, the definition of " xR%y " may be written
(A)
zRw.'D z>w .<l>lz'2 4>l w.D^.^lxD^ly,
i.e.
"
remain
suitably interpreted.
valid,
x."
We
may,
as
we
is
onemany or many one, and fifthorder properties in other cases. But for
preliminary purposes it makes no difference what order of properties we take,
and therefore for the sake of definiteness we take elementary properties to
begin with. The difficulty is that, if 2 is any secondorder property, we
<f>
zRw Dz>w
.
. <j> 2
D <p 2 w D
:
<f> 2
xD
<f> 2
y.
(B)
INTRODUCTION
xliv
<j> 2
z.= .(<f>).fl(<frlz,z);
deduce
zRw D
.
Zj
w ./!
(<
% z) D4 /! ( % iv) D
!
:/!
:
(<
fax
% x)
D* ./!
(<
y)
(C)
. <j> 2 y.
<f>
of.(B).
Hence
it is
<f>
2 z,
where
z
(f>2
is
./! (*
(<f>)
it is
is
either
% z)
*,*),
w,
each elementary 0,
zRw.D ZtW ./l t4> ! % z)
for
is
D/I (0 ! f ;).
that, primd facie, an
,
inductive property
of the form
xR%. z
<f>lz
SeFotid'R.tfilS
or
or
ae
This
is
alone
is not, i.e.
NC induct
. <}>
I a.
when
<f>
we may have
. <ft !
and similarly
it
necessary to reexamine
still valid,
all
inductive proofs.
but
Appendix
it is
always possible.
still
The method
of
to this volume.
is,
made adequate
to
is Dedekindian. It is upon this that the theory of real numbers rests, real
numbers being defined as segments of the series of rationals. This subject is
dealt with in #310. If we were to regard as doubtful the proposition that the
series of real numbers is Dedekindian, analysis would collapse.
The
axiom of
XCD'Pe.D.s'XeD'Pe.
INTRODUCTION
For reasons explained above,
if
is
is
xlv
of the order of
members
of X, (a) .fa
may
yet
we cannot
reason, of the
is finite,
infer s'X e
DP
f
same order
= P"s'Pe"K
except when s*\ or s'Pe "\
as the
require reconstruction.
Exactly similar
difficulties
The
Bord^PCClex'C'PCCFminp)
PeBord = aCC'P.g
whence
Df,
D a .g aP"a.
!
P r\p'P"(a
substitution
is
r\
C'P), which
illegitimate,
is
we cannot prove
a.
If this
C'P
and having successors must have an immediate successor, without which the
theory of wellordered series becomes impossible. This particular difficulty
might be overcome, but it is obvious that many important propositions must
collapse.
It
might be possible
numbers is an integral part of ordinary matheand can hardly be the object of a reasonable doubt. We are therefore
justified in supposing that some logical axiom which is true will justify it.
The axiom required may be more restricted than the axiom of reducibility,
rigour, but the theory of real
matics,
but, if so,
The
it
remains to be discovered.
following are
publication of the
D.
P.
among
first
Jahresbeiicht
86.
H. Weyl. Das Kontinuum, Veit, 1918. Ueber die neue Grundlagenkrise der Mathematik,
Mat/iematische Zeitschrift, Vol. 10. Bandbemerkungen zu Hauptproblemen der
Mathematik, Mathematische Zeitschrift, Vol. 20.
INTRODUCTION
Xlvi
L. E. J.
tiker
Mathematical,
A. TajtelbaumTarski. Sur le terme primitif de la logistique, Fundamenta
Tom. IV. Sur les "truthfunctions" au sens de MM. Russell et Whitehead, ib.
Tom. V. Sur quelques theoremes qui equivalent a l'axiome du choix, ib.
F. Bernstein.
C.
Konig.
I.
Lewis.
Veit, 1914.
Nigod.
Camb. Phil.
L.
reduction in the
number
Proc.
Vol. 92.
INTRODUCTION
The mathematical logic which occupies Part I of the present work has
been constructed under the guidance of three different purposes. In the first
place, it aims at effecting the greatest possible analysis of the ideas with
which it deals and of the processes by which it conducts demonstrations,
and at diminishing to the utmost the number of the undefined ideas and
undemonstrated propositions (called respectively primitive ideas and primitive
it starts.
In the second place, it is framed with a
view to the perfectly precise expression, in its symbols, of mathematical
propositions) from which
propositions
which, in recent years, have troubled students of symbolic logic and the
theory of aggregates
what
it is
follows,
Of the above three purposes, the first and third often compel us to adopt
methods, definitions, and notations which are more complicated or more
difficult than they would be if we had the second object alone in view.
This
applies especially to the theory of descriptive expressions (#14 and #30) and
and
to a lesser degree
only temporary
real
meaning
is
The
sacrifice
is,
thought
for
though
its
i.e.
where
it is
we
it
believe to be
is
When
the notation
more complicated
In the body of the work,
more
correct.
in
it
Chapter
;
the
is
full
The use
r&w
strictly accurate
INTRODUCTION
and
employed are more abstract than those familiarly conAccordingly there are no words which are used mainly
in the exact consistent senses which are required here. Any use of words
would require unnatural limitations to their ordinary meanings, which would
(1)
The
ideas here
sidered in language.
be in fact more
entirely
new
difficult to
remember
symbols.
of
ideas of this
more
easily.
"
giving terse expression to a complicated fact while the true analysis of one
Accordingly
is a number" leads, in language, to an intolerable prolixity.
terseness is gained by using a symbolism especially designed to represent the
;
ideas
(3)
The adaptation
this work.
deduction aids the intuition in regions too abstract for the imagination
readily to present to the mind the true relation between the ideas employed.
For various collocations of symbols become familiar as representing important collocations of ideas
according
And
relations
ideas.
imagination.
(4)
The
(5)
all
first mentioned
namely
INTRODUCTION
in mathematics, necessitates both terseness
maximum
as possible.
(a)
is
is
thrown by a
employment
is sufficient to
against error.
In this connection
tions of Weierstrass
without
its help.
work
to
show
that, with
is
these subjects.
12
CHAPTER
The notation adopted in the present work is based upon that of Peano,
and the following explanations are to some extent modelled on those which
he prefixes to his Formulario Mathematico. His use of dots as brackets is
adopted, and so are many of his symbols.
Variables. The idea of a variable, as it occurs in the present work, is
more general than that which is explicitly used in ordinary mathematics.
meaning
is
be any set of
to circumstances.
and
"
Mr B
"
If a statement is
are variables
made about
"
Mr A and Mr
to
men.
according
Mr A "
variable may
may (in
B," "
the absence
of any indication of the range of values) have as the range of its values
it
all
occurs significant.
Thus when a textbook of logic asserts that "A is A" without any indication
what A may be, what is meant is that any statement of the form
as to
"A
is
"
is
true.
We may
call
a variable restricted
unrestricted.
it is
capable
falsely)
made
its
it
values are
we
shall call
represents any
significantly
{i.e.
either
unrestricted variable
shall
when
otherwise,
variable occurs,
always employ
is
it.
We
by the manner of
its
occurrence,
i.e.
things
the variable occurs, and are therefore intrinsically determined by this statement. This will be more fully explained later*.
To sum
up, the three salient facts connected with the use of the variable
is
throughout the same context, so that many variables can occur together in the
* Cf.
CHAP.
"
THE VARIABLE
I]
its
possible determinations of
is
may be
determina
is
is
all
variables in
it
suitable
determinations.
letters.
<f>,
letters,
and
will
be used
will
letters.
After the early part of the work, variable propositions and variable functions
We shall then have three main kinds of variables
In addition to this usage of small Greek letters for variable classes, capital
Latin letters for variables of type wholly
INTRODUCTION
[CHAP.
The general
as arguments.
is
employed in
But
this work.
there are four special cases which are of fundamental importance, since
the
all
They
Sum,
or Dis
junctive Function, (3) the Logical Product, or Conjunctive Function, (4) the
Implicative Function. These functions in the sense in which they are required
in this
all
independent and
;
if
though
not entirely
arbitrary
primitive.
first
asserting that
is
proposition p.
notp,
It will also
is
least
p, where
is
any proposition,
is,
the proposition
denoted by ^p.
Thus
^p
is
the
as
The Logical
and
is
to
means
This
some
ideas.
not true.
is
to
is
Sum
is
p and q
or q disjunctively, that
is true.
This
is
denoted by
is,
p and
q,
asserting that at
p v q. Thus p v q
is
the logical
p and q.
and
is
is
p and
both
or
by p
::
p and
q.
Thus p q
.
is
by p q, or by p :. q,
p and q as arguments. It
Accordingly p q means that both
:
~(~pv ~ q).
If any further idea attaches to the proposition " both
p and
q are true,"
it is
FUNCTIONS OF PROPOSITIONS
I]
p and
q,
and
proposition
is
~pv
alternative left
is
Thus
if
is true,
by the proposition
if
true.
implies
important that
it
q.
The
idea contained in
disjunction "not^ or
"
~ pyq"
is
"pD q."
This symbol
may
for
"p
q"
implies
i.e.
for
The
"
implication
"
from
The
is
it is
necessary
When
is
simply "implication"
common usage
is
often
explained immediately.
No
Equivalence.
of the formation of a
more complex
Newton was not a man " and " the sun is cold " are equivalent as being both
But here we have anticipated deductions which follow later from our
false.
"
formal reasoning.
Equivalence in
its
origin
is
stated above.
Truthvalues.
and falsehood
The
if it is
"
truthvalue
false *.
*
"
of a proposition
is
truth
if it is true,
This phrase
is
due to Frege.
;;
INTRODUCTION
[CHAP,
pvq,p .q,pOq, ~p, p = q depend only upon those of p and q, namely the
truthvalue of'pvq" is truth if the truthvalue of either p or q is truth,
and is falsehood otherwise that of " p q " is truth if that of both p and q is
truth, and is falsehood otherwise that of "pDq" is truth if either that of
p
is falsehood or that of q is truth
that of " ~ p " is the opposite of that of p
and that of " p = q" is truth if p and q have the same truthvalue, and is
.
falsehood otherwise.
in the present
repetitions.
if a proposition p occurs in any propof(p) which we shall ever have occasion to deal with, the truthvalue
f f(p) w iH depend, not upon the particular proposition p, but only upon
its truthvalue
i.e. if p = q, we shall have
f(p) =f(q). Thus whenever two
propositions are known to be equivalent, either may be substituted for the
other in any formula with which we shall have occasion to deal.
sition
We' may call a function f(p) a " truthfunction " when its argument p is
a proposition, and the truthvalue of
f(p) depends only upon the truthvalue of p. Such functions are by no means the only common functions of
For example, "A believes p " is a function of p which will
truthvalue for different arguments having the same truthvalue
propositions.
vary
its
A may
believe
one
and may
proposition
any general propositions we may make about functions but the particular
functions of propositions which we shall have occasion to construct or to con;
This fact
is
mathematics
become more
so
Assertionsign.
what
The
follows is asserted.
is
if not
always concerned
now obvious,
of classes
and
will
relations.
means that
The
bolism.
For example,
if "
I
(p D p) "
occurs,
it is
same purpose
in our
sym
to be taken as a complete
"pOp"
The process
of inference
is
as follows: a proposition
is
asserted,
and then
"p"
is
as a sequel
I]
is
The
asserted.
trust in inference
the
error,
is
final assertion is
p and
if
the
not in error.
determinations,
put
occurrence of " h
to write
" h
q"
It is of course convenient,
q."
it
on record. The
is
the
When
this is to
is
instead
"hpDbq,"
which
is
Thus
"
to be considered as a
" h
bpDbq" may
p " and
state,
what
is
part of
its
The use of
"
be read
p,
it is
and
" h q."
meaning, that
;
"
q)
is;
implies
An
q.
same
explicitly
inference
is
the
Dots on the
dots.
" h
(p
line of the
bracket off propositions, the other to indicate the logical product of two
"
proposition
general principle
is
that a larger
number of
by dots
dots into three groups which we
is
logical product.
The
mark a
The exact
will
name
I, II,
and
III.
Group
I consists of
by
definition
(= Df).
Group
(x,
y) or i^x) or
Group I is of greater force than Group II, and Group II than Group III.
The scope of the bracket indicated by any collection of dots extends backwards
or forwards beyond any smaller number of dots, or any equal number from a
group of less force, until we reach either the end of the asserted proposition
or a greater number of dots or an equal number belonging to a group of
equal or superior force. Dots indicating a logical product have a scope which
works both backwards and forwards; other dots only work away from the
*
The meaning
INTRODUCTION
10
[CHAP.
II.
When
we write
it,
"h
where the two dots after the assertionsign show that what is asserted is the
whole of what follows the assertionsign, since there are not as many as two
dots anywhere else. If we had written "p v q D q v p" that would mean
:
is true,
If
we wished
to assert
we should have to put three dots after the assertion sign. If we had
written "pvq D .q:v :p," that would mean the proposition " either 'p or q'
implies q, orp is true." The forms "p v .q D .qv p" and "pv q D .q v .p"
this,
have no meaning.
implies
r,
implies
implies
r,'
then
This
implies r."
in general untrue.
is
implies 'q
(Observe that
"P ^<l"
since
else,
the
scope of these two dots extends backwards to the beginning of the proposition,
and forwards
if
to the end.
"pvq .0 '..p
either p or 'q
this
is
v . q Dr :
implies r
to be asserted,
we
"\
(This proposition
to assert
is
true,
(what
D .pvr"
is
::
'
is
mean
"if either
or q
or r
will
is
is
true, then
pvq
'3
:.
v q
.
If
true."
D r D .pvr."
:
is
and either p or
then either
or r
is
true,"
we
write
" h :.
Here the
does not.
first
pvq p v
:
qD r D p v r."
:
11
DEFINITIONS
I]
first,
as they
show
its structure.
it is
of dots before
number
to everything that
is
comes
stated
after
it.
We
defined as meaning.
to the left
and the
"Df "
We express
what
is
names
the
It is to
to be explained shortly.
An example
of a definition
is
p"Dq.
<^>p
vq
Df.
no part of the
For a definition is concerned wholly with the
symbols, not with what they symbolise. Moreover it is not true or false,
It is to be observed that a definition
subject in which
is,
strictly speaking,
occurs.
it
volition, not of
a proposition.
definitions
are
instead,
appear among our primitive ideas, because the definitions are no part of our
subject,
but
become
tically, of course, if
we introduced no
so lengthy as to
superfluous.
In spite of the
nevertheless true that they often convey more important information than
First,
is worthy of
embodies our choice
careful consideration.
*
is
Hence the
collection of definitions
concern us at present.
It
INTRODUCTION
12
judgment
what
is
defined
is (as
as to
what
is
[CHAP.
most important.
may therefore
Secondly,
when
common
idea,
and
amounts to the statement that what he is dewhich has the properties commonly associated with the
word " continuum," though what precisely constitutes these properties had
not before been known. In such cases, a definition is a " making definite ": it
gives definiteness to an idea which had previously been more or less vague.
are
For these reasons, it will be found, in what follows, that the definitions
what is most important, and what most deserves the reader's prolonged
attention.
considered as a whole.
Summary
tive ideas
of preceding statements. There are, in the above, three primi" defined " but only descriptively explained.
Their
primitiveness
is
is
not absolute; though of course such an exposition gains in importance according to the simplicity of its primitive ideas. These ideas are symbolised
q,"
= ,oj(~pv~g')
p"Dq = .~pv
p = q = ,p"D q.qDp
p
Primitive propositions.
g.
Df,
Df,
Df.
proof,
These, as
PRIMITIVE PROPOSITIONS
I]
13
some slight limitation in the form of an increased stringency of enunand (2) the system must lead to do contradictions, namely in pursuing,
our inferences we must never be led to assert both p and notp, i.e. both " .p"
and "h ,~p" cannot legitimately appear.
require
ciation;
I
The
The
propositions.
letters
"Pp" stand
(2)
i.e.
if
(3)
i.e.
i.e.
if
bzq.D.pvq
is
(4)
if
(5)
i.e.
Y'.pvp.S.p Pp,
p is true, then p
or
if
then
true,
or q is true,
h
:p vq. D q
is
or q
vp
is
true.
Pp,
then q or
:pv(qv?~).
either
is true.
Pp,
p is true.
D.qv(pvr) Pp,
true or
,:
q or r"
is true,
then either q
is
is
true.
(0)
i.e.
if
</
(7)
I
qDr D
:.
implies
r,
\p~vq. D .^ vr Pp,
"_p or q" implies "p or
then
r."
we
is
to say, if
may
is
be), or,
assertion,
<f>
. .
of excluded middle:
h
This
is
*211
below.
We
.pv<^p.
The law
numbers given
to the
[CHAP.
INTRODUCTION
14
The law
The
principle of transposition,
i.e.
namely
(#41)
I
(#411)
t
:pDq. = .~qD~p,
:p = q. = ~p = ~q,
f
:p.^r.D .^q,
The law
(#4'24)
p = p .p,
.
"p
i.e.
or
and
is
true"
is
equivalent to "p
true and
is
true."
is
its
ordinary algebra.
The law
of absorption:
(#471)
i.e.
is
equivalent to "p
do
it
so.
An
(*473) h:.q.D:p.
the following:
is
.p.q.
and commutative
The second
laws,
p qv r = p q v p r,
.v .q .r: = :pv q .pvr.
(#44)
\
(#441)
b :.p
:.
and multiplication.
Propositioned functions.
it
possible determinations to
false.
Also
if
"x
is
x in "x
is
is
PROPOSITIONAL FUNCTIONS
I]
15
value of that function. Accordingly though "x is hurt" and "y is hurt" occurring
in the same context can be distinguished, "ft is hurt" and "y is hurt" convey
no distinction of meaning at
the propositional function
for x,
<f>a is
More generally,
and when a definite
all.
<}>$,
an unambiguous value of
<f>x is
an ambiguous value of
a is substituted
signification
<f)x.
Propositional functions are the fundamental kind from which the more usual
kinds of function, such as "sin a;" or "logo?" or "the father of x," are derived.
The
functions."
<a>,
there
total variation.
is
the propositions (true or false) which can be obtained by giving every possible
<px.
value of x for which <f>x is true will be said to
determination to x in
"satisfy"
Now
given by three propositions of which one at least must be true. Either (1) all
propositions of the range are true, or (2) some propositions of the range are
true, or (3) no proposition of the range is true. The statement (1) is symbolised
by "(a?) . <f>x," and (2) is symbolised by "(g#) . <f>x." No definition is given of
The symbol
system.
or
"<f>x is
true for
"(x)
all
<j>x"
may be read
<j>x is
always," or
The symbol
"<f>x
"<f>x is
"(ga?)
always true,"
.
<f>x"
may be
an x satisfying
hand.
Accordingly
4>cb
is
contradicts a value of
(1)
and
an obvious
untrue. This is
<f)x\.
<f>8l
number (3)
as stated above.
at once, for
oj {(x) .
<f>% is
Hence
~ (x)
x =
.
~ [(x)
(f>x\
Df.
[CHAP.
INTRODUCTION
16
is to
will
portant idea.
The
proposition (x)
<f>x is
be explained in Chapter
will
II,
<.
as a primitive idea
when
negation of fax)
4>x as
being (x)
~ {(x)
~ {(3^)
<t>x}
$*'}
Thus we put
~<f>x.
.
<px
p =
:
(x)
<fix
v p Df,
"either <f>x is always true, orp is true" is to mean "'fyxoxp' is always true,"
with similar definitions in other cases. This subject is resumed in Chapter II,
and in #9 in the body of the work.
i.e.
Apparent variables. The symbol "(x) <f>x" denotes one definite proposition,
and there is no distinction in meaning between "(x) <px" and "(y) fyy" when
they occur in the same context. Thus the "a?" in "(x) <j>x" is not an ambiguous
constituent of any expression in which "(.:). <f>x" occurs; and such an ex.
ambiguity of the
x in the
"<f>x."
The symbol
"(x)
the symbol
<j>
(x)
dx
neither case
is
x.
The range
is
determinate.
The x which
(following Peano) an
"
.x = x"
mean "everything
is
x.
Thus
e.g. "(.r)
will
A
x.
is
the field
x,
meaning
APPARENT VARIABLES
I]
all
all
17
"<f>x" is
If
the scope of #;
if
we
we
are
are
"(g#)"; that
is
mean
"<j>x
will
mean
"if
always implies
tyx,"
but
"(x).(f>x.0.ylrx"
Note that
<f>x is
(x)
the two
x'a
yjrx is
argument
x."
in the proposition
.
<f>x
ypx
first
is
rather than
(x)
(x)
<f>x
<f>x
D
D
yjry
yfrx,
since the use of different letters emphasises the absence of connection between
and
Ambiguous
$x can be
of $ob
is
is
it is
letters,
letter.
assertion
asserted.
symbolised by
"h.^x."
Here a and
is
expressed, though
them
propositions are
all
of
are true.
above, pp. 9
R&W
and
I
of the type of
Group
II as explained
10.
INTRODUCTION
18
be legitimately asserted
i.e. if
all
middle
if,
[CHAP;
is
equivalent to
" (a, b)
is
b."
it
necessary to keep
distinguished.
When we
assert
"K# =
we
e.g.
#,"
When we
assert
some
(x)
x = x"
or
we
are asserting, in the first case all values, in the second case
some value
determined as to give a
false proposition.
be determined,
Thus
it
might be so
we have
\
we may
And
infer
,x = x
h.(x).x x.
an assertion containing a real variable x, we may transvariable into an apparent one by placing the x in brackets at
generally, given
When we
many
something containing a
assert
real variable,
we only obtain a
we cannot
strictly
definite proposition
by assigning a value
to the variable,
has not at
example
"sin 2 #
all
+ cos # = 1"
2
does not assert this or that particular case of the formula, nor does it assert
that the formula holds for all possible values of x, though it is equivalent to
this latter assertion; it simply asserts that the formula holds, leaving
undetermined; and
it is
x wholly
x may
results.
SEAL VARIABLES
I]
Definition
and
When
real variables.
19
its
on both
the definition of
"~ {{%)
<f>x}"
<f>z,
In
is
on the
left.
omitting an apparent variable where (as in the case before us) this
argument
plainly
say
"ob
if,
if
is
~ {(x)
But
which
to the function
we take some
<&,
particular function,
in which the
x,
"& = a."
practice.
In
fact
either as constants
to
[e.g.
we
namely
but would be clumsy in
itself,
we have found
explanatory portions
the
definition of
x = a]
is
=a]
work.
"
When
More
that
in
This translates
all."
an assertion, we
it
may turn
itself into
it
what
the rule
into an apparent
the assertionsign.
(2)
"
What
holds of
all,
holds of any,"
h
"If
<f>y
<f>x is
(x)
<f>x
is true,
then
<f>x is
i.e.
$y.
true."
<\>y is
sometimes
\:<f>y.D. (rx)
true,"
i.e.
<t>x.
22
20
INTRODUCTION
An
theorem," namely
what
position gives
we always have
infer " (g#)
an x
"
</>&'
" existence
expresses an
which
for
is
[CHAP.
to find
If
<f>x.".
its
In virtue of
"
" h
and
(x) ,<f)x.D.<f>y"
what
" h
$y D (g#)
always true
we have
<#,"
is
is
"any."
" If
(4)
true,"
<j>x is
'
yjrw
<f>x
'
always
is
i.e.
We
type.
also holds
i.e.
(x)
: .
and
same
yfrx is
(f>x
(x)
yfrx
(x)
<f>x
\^x.
yfr
(x)
: .
(f>x
yfrx
."D:(x).
<f>x
(x)
yjrx.
propositions assumed, on this subject, in the body of the work (*9), are the
following
(1)
I"
if
<f>x
either
</>.*.
is true,
or 4>y
is
If
(3)
.
<f>x
<px
<f)x
a *)
<\>z.
in the
necessity
we can
i.e.
yfrx, is
formally implies
be said
(On the
is true.
cj>z
remarks on #911
(x)
h:4>xv<f>y.D.(^z).cf>z,
(2)
i.e.
to state
formal
such as "'Socrates
sition of the
yfrx
form
In such a case, we
is
i.e.
when
(x)
feel
<f>x
implications.
\]rx "
"
an implication, say
\jrx,
(x)
we
z^.D.^x "
will
" 4>x
When
equivalence.
assert
in a case in
is
mortal,' "'we
which
"
(x)
:</>*.
have a propo
yfrx
"
is true.
Thus it has come about that implications which are not particular
cases of formal implications have not been regarded as implications at all.
cation.
There
is
also a practical
ground
for
for,
speaking
FORMAL IMPLICATION
I]
hypothesis
known when
already
it is
is false
21
true
is
known
and
in neither of these
for
known
already.
Thus
chiefly
it is
be known
directly.
These reasons, though they do not warrant the complete neglect of implications that are not instances of formal implications, are reasons which
make
all
Since
of
"
what
is
effectively stated is that, for all these values, tyx is true. Thus propositions
of the form " all a is /3," " no a is /9 " state formal implications, since the first
(as appears
said) states
D x
is
is
an a D x
an a
is
/3,
And any
values of
"
x which
satisfy*
may be
</>#
not a
is
"
interpreted as
" All
satisfy tyx,"
<f>x
interpreted as
yS.
may be
yjrx "
No
values of
x which
satisfy
<f>x
satisfy tyx."
We
have similarly
for "
some a
(g#)
and
for "
some a
is
not
/3
"
Two
functions
<f>x,
i.e.
yfrx
is
the formula
an a
is
is
not a
/3,
the formula
(g#)
is /3 "
an a
is
/9.
when
(x)
<f>x
yfrx,
value of x
is
said to satisfy
<f>x
or
tj>x
when
<f>x
is
INTRODUCTION
22
[CHAP.
it
</>
has just been said explains the connection (noted above) between
the fact that the functions of propositions with which mathematics is specially
concerned are all truthfunctions and the fact that mathematics is concerned
What
Convenient abbreviation.
often
<f>x
<px
Dx
yjrw
=x
ifrx
=
=
(x)
<f)X
:(x):<f>x.
D
=
tyx
Df,
tyx
Df.
" is
This notation
"
<f>x
Ox
yfrx
<px
D x ijrx
.
= x yfrx =
cf>x
(x)
.(#).
D ijrx
=
yjrx
<f>x
Df,
<f>%
Df.
In practice however, when </> and yjr^c are special functions, it is not possible
to employ fewer dots than in the first form, and often more are required.
The
or
more variables
(x, y)
and
so
on
</>
and so on
<f>
for
any number
(x,
y)
for
Identity.
(x, y)
D x>y .^(x,y): =
any number of
The
(x)
(x, y)
we
(cf.
< (x,
y)
Df,
4>(x,y)
.1 .^
(x,
y)
Df,
variables.
x=
in the definition,
(y)
of variables
is
identical with
y"
is
expressed by
y.
shall here
omit
it (cf.
Chapter
II).
We
have, of course,
.x
The
first
23
IDENTITY
I]
between y and a;. The third proposition expresses that identity is a transitive
relation a relation is called transitive if, whenever it holds between x and y
and between y and z, it holds also between x and z.
:
We
mathematics
all
If x and y are identical, either can replace the other in any proposition
without altering the truth value of the proposition; thus we have
r :
This
x= y D
.
<f>x
<j>y.
is
mostly follow.
It
since
can only hold between x and y if x and y are different symbols for the same
object. This view, however, does not apply to what we shall call " descriptive
it
phrases,"
"
i.e.
important, as
the soandso."
we
A proposition
and
A. class
relations.
(which
is
propositional function
one which
is
is
always
false,
be the
its roots
class of
fractions,
and so
men
if
<f>x is
if <f>x
class,
When
the class
Thus Greek
be
will
be
< <
on.
functions.
"
is "
will
same
it is
may
letters,
Any
other letter
may be
used instead of
z.
is
'
24
INTRODUCTION
[CHAP.
There are two kinds of difficulties which arise in formal logic one kind
with classes and relations and the other in connection
with descriptive functions. The point of the difficulty for classes and relations,
so far as it concerns classes, is that a class cannot be an object suitable as an
argument to any of its determining functions. If a represents a class and <j>x
one of its determining functions [so that a = z {<f>z)\ it is not sufficient that
<]>a be a false proposition, it must be nonsense.
Thus a certain classification
of what appear to be objects into things of essentially different types seems
;
arises in connection
to
be rendered necessary.
is
the theory of types, and the formal treatment in the systematic exposition,
is
is
The
is
it is
It
is
though perhaps, if we had regarded some solution which held classes and resome real sense objects as both true and likely to be universally
received, we might have simplified one or two definitions and a few preliminary
propositions. Our symbols, such as " ct (<f>%) " and a and others, which represent
classes and relations, are merely defined in their use, just as V 2 standing for
lations to be in
dx*
df
+
dz*
The
x, y, z on which to operate.
which we use classes corretheir use in ordinary thought and speech and whatever
way
in
sponds in general to
may be the ultimate interpretation of the one
is
CLASSES
I]
The
is
25
by the notation
following Peano,
area.
Here e is chosen as the initial of the word earL " x e a " may be read " x is
an a." Thus "x e man" will mean "x is a man," and so on. For typographical
convenience
we
shall
put
x~ea. = .~(xea.)
x, y eOL. =
For
We
have
we
"'x
i.e.
is
satisfies
.xea. .y ea Df.
Df,
x e z (<f>z) =
.
or to
class."
<f>x,
<j)z,'
is
is
<f>z'
equivalent to 'x
"
'
<f>x is
true.'
is
known, that
is,
there
cannot be two different classes having the same membership. Thus if <f>x, yfrx
are formally equivalent functions, they determine the same class ; for in that
case, if
a?
is
Thus we have
h
The
:.
i.e.
is identical
" is
(<f)z)
= z (tyz) =
:.
with the
ft
class
determined by
yfrx.
<fiz
<f>x,
is
(f>x;
= ft ,= :x60t.= x .xe/3,
& (x a)
= a,
is "
formally equivalent to
i.e.
is
.(f)X.= x
= 2 (<j>z) = x e a = x
b :.a
i.e.
an a
members
is
an o "
is a,
same
in other words,
of a
r.2(*)eCls,
i.e.
is
class.
It Will be seen that, according to the above, any function of one variable
can be replaced by an equivalent function of the form "xea." Hence any
extensional function of functions which holds when its argument is a function
of the form "zea," whatever possible value a may have, will hold also when
its
argument
is
any function
<f>z.
Thus
we
are concerned.
all
INTRODUCTION
26
[CHAP.
. . .
. . .
Such
relations.
relations will
not concern us until we come to Geometry. For the present, the only relations
we
relations
Any
function
(f>
determines a relation
(x, y)
between x and
by the function
(x,
< (x,
y) for which
y) will
</>
(x,
y)
The
is true.
y.
If
< (x,
we
y) is
relation determined
be denoted by
We" shall use a capital letter for a relation when it is not necessary to specify
Thus whenever a capital letter occurs, it is to be
understood that
The
it
x has the
relation
to
y"
will
be expressed
by the notation
xRy.
This notation
which,
when
is
Owing
we
Rel
";
thus "
is
R e Rel "
common
it
:.
x<f> (x,
y)
= xyfr (x,y).=
:<f>
means "R
we
(x, y)
shall
.
=x
language,
between its
and so on. For
it
is
a relation."
have
y .ty
0>
y),
two functions of two variables determine the same relation when, and only
when, the two functions are formally equivalent.
i.e.
We
have
{xf/rf*
(x,
y)}w =
,<f>
(z,
w),
(x, y) is different
to distinguish
it
"z has to
i.e.
to
27
CALCULUS OF CLASSES
I]
<f>
w)
(z,
R = %< (x, y)
#%
= x ,y<f> (,
I
:.
<j>
y)"
(x,
is
equivalent
2/)>
\.$f/(xRy)
= R,
K^(a?,y)}eRel.
These propositions are analogous to those previously given for classes. It
them that any function of two variables is formally equivalent to
some function of the form xRy; hence, in extensional functions of two variables,
variation of relations can replace variation of functions of two variables.
results from
Both
classes
and
members
of both.
/3 is
This
common
their
represented by a
is
ar\fi
This gives us
i.e.
"x
is
logical
I :
member
product of
"
part,
i.e.
xea
r\
is
member
of a "
Thus we put
Df.
x e a x e {3,
.
logical product
r\ fi.
= <x}(xea.xe0)
most of those of
The
and
"
and 0"
is
is
equivalent to the
member
of fi"
members
are
and the
of either
V \.xeoL\J
fi
'.xea..v .xe
ft.
The
We
are not
members
of the negation of
a.
Thus the negation of a class a is the class of terms of suitable type which
are not members of it, i.e. the class tc(x~ea). We call this class ",a" (read
"hota"); thus the definition
is
a = x(x<^e a)
Df,
xe a. =
is
given by
x<^>ea.
if
x e a D x x e 0.
.
We
write " a
if all
C /3 "
for
class a is said
put
Df.
INTRODUCTION
28
[CHAP.
The
following are
classes
common
i.e.
the
i.e.
"
is
i.e.
"
is
not a
and
part of a
ft is
member
= (au/3),
an/3
the negation of
x e (a u
"
nota or not/3
"
a),
of a or nota "
h
member
#~e (a n a),
and nota "
of both a
h.a = (a),
h:aCft. = .ftCa,
The two
The law
\
Thus
for
example
" all
aCft = .a=
.
Cretans are
ar\
ft.
liars " is
Just as we have
so
hzjj'Dq.qDr.D.p'Dr,
we have
I
a Cft
ftCy D
.
aC^
interchanged)
are 7's."
with
If all a's are ft's, and all ft's are 7's, then all a's
observed
be
that syllogisms are traditionally expressed
" therefore," as if
'"
way
is
is,
really asserted is
The syllogism
subject
in
is
e ft
ft
C 7 D x e 7,
men are mortals, then
.
" if
mortal."
Socrates
is
This, as
This point
is
will
become
is
here mistaken.
The nature
CALCULUS OP CLASSES
I]
29
We
h x ( R n S) y = xRy xSy.
RvS = $y (xRy v xSy) Df,
which leads to
Similarly
^R = xy {{xRy)}
RGS. = xRy D^ xSy
:
Df.
Generally,
and
Df,
for classes,
when it has
Thus we put
"
denoted by
a."
3
The
class
Any
"A."
at least one
(g#) x
is
Df.
member
always
and
is
denoted by
determines the nullclass. One such function is known to us already, namely "x is not identical
with x," which we denote by " x 4= x." Thus we may use this function for defining A, and put
is
A = x(x$x)
The
class
Df.
universal class,
and
is
is
We
the negation of V.
h
"
i.e.
'
is
member
of
'
"
'x
is
member
of
A'
(x)
always
is
h
i.e.
a R " means
!
false."
.
xy (x ^ x y 4= y).
.
Also
a,
We
put
(g#, y) xRy,
.
that there
holds.
will
relation
and
x~e A,
= A = ~g
3 R =
!
the relation
We
Y.{x,y).~{xky),
and
called the
Df.
"
is
.xeY,
i.e.
always true
have
(x)
is
Ji.e.
is
represented by V; thus
V = &(# = #)
Thus
false
R = A. = ~g! R.
.
INTRODUCTION
30
There are no
classes
[CHAP.
is
By
mean a phrase
a "description" we
If
<f>&
is
is used in our
can always be read as " the x which satisfies
<&,"
x which satisfies
phrase as embodying a primitive idea. Every use
of " (ix)
on hand.
An
"
in
use can always be read as "the x's which satisfy <f>&." Both symbols are incomplete symbols defined only in use, and as such are discussed in Chapter III.
The symbol " cb (<f>x) " always has an application, namely to the class determined
by <f>x but " (ix) (<f>x) " only has an application when </> is only satisfied by
;
one value of x, neither more nor less. It should also be observed that the
meaning given to the symbol by the definition, given immediately below, of
E ! (ix) (<f>x) does not presuppose that we know the meaning of " one." This is
also characteristic of the definition of any other use of (ix) (<f>x).
We
now proceed
satisfying
<f>x
exists."
to define "
(It will
(ix)
we
(<f>x)
express by
.
"
g.")
is
can be read
Its definition is
(gc) :<f>x.= x .x
=c
" the
a different meaning of
Df,
DESCRIPTIONS
I]
i.e.
is
" the
The
The
satisfy
" there is
:.
E!
(ix)
(<f)x)
such that
cf>x
is
(gc)
: <f>c :
(gc)
(gc)
:.
E!
(ix) (<f>x)
\
:.
(7a?)
($#)
an object
D x x = c,
D XjV .x = y,
<f>c : <f>x
<f>y
Dx ~ <#
a? 4= c
<f>c
<f>x
c satisfying
<f>$!,
<f>
does not
</>&."
The kind
example
" the
is
(gc)
c is
cases.
Thus
for
most perfect = x . x = c,
.
although
many
mean
an object
"there
31
is
is
equivalent to
is really
of the
same
logical form,
Scott
Here
(as
we observed
= (ix) (x
the.
wrote Waverley).
The notation " (ix) (<f>x)" which is long and inconvenient, is seldom used,
being chiefly required to lead up to another notation, namely "R'y" meaning
" the object having the relation R to y." That is, we put
R'y
= (ix) (xRy)
Df.
inverted
would stand
for
y.
Instead of a variable
INTRODUCTION
32
[CHAP.
A descriptive function
while y belongs to a certain domain, but not outside that domain thus if we
are dealing with positive rationals, sjy will be significant if y is a perfect
;
mean
real
mean
the negative
square root), \'y will be significant provided y is positive, but not otherwise
and so on. Thus every descriptive function has what we may call a " domain
of definition
or a "
"
domain of
function in question
R'y,
is
existence," which
its
domain of
may
be thus defined
If the
the class of those arguments y for which we have E! R'y, i.e. for which
! (ix) (xRy), i.e. for which there
is one x, and no more, having the relation
to y.
If
any
is
function."
we
relation,
will
"
associated descriptive
In such
cases, it is easier
assigning the meaning of the descriptive function, and to deduce the meaning
of the relation from that of the descriptive function.
of
x and
y.
Thus greater
husband of
nition
wife, etc.
is
the converse of
The converse
of
less,
is
written*
Cnv'R
or R.
The
defi
is
R = $$(yRx)
Cnv'R =
Df,
Df.
is
called symmetrical if
with
its
converse,
r\
i.e.
less,
metrical, as are all other relations of the sort that lead to series.
many asymmetrical
*
The second
relations
of these notations
that of
DESCRIPTIVE FUNCTIONS
I]
33
wife's brother*.
for
Ca
both/be true
= /8,
will
if
The
called
of
class
all
with respect to R. These two classes are denoted respectively by R'y and
R x.
l
Thus
R'y = $(xRy)
Df,
R'x = {xRy)
Df.
relation
to y; it
members
of
R x.
l
It runs
The
R is
of
x.
notations R'y,
R'x
= the
We
have
If
parents of
y,
R'x = the
children
R'y = xRy
.
y e R'x . = xRy.
These equivalences are often embodied in common language. For example,
we say indiscriminately u x is an inhabitant of London" or "x inhabits London."
If we put "R"for "inhabits," "x inhabits London" is "xR London," while "x
and
I
is
"
sg'R
=R
"
Df,
gs'R = R Df.
These notations are sometimes more convenient than an arrow when the
relation concerned is represented by a combination of letters, instead of a
single letter such as R.
Thus
e.g.
we should
The
class of all
called the
r\
write sg'(.R
domain of R. Thus
if
is
sister's
R&W
husband.
I
S), rather
than put
to something or other is
the relation of parent and child, the
the
r\
8).
is strictly
when the
asymmetrical.
[CHAP.
INTRODUCTION
34
We
class of parents.
D<E = {( a y).*%}
by
Df.
all
= $ {(a) xRy) Df
Q'R
The sum
of the
is
and
is
since
any).
The
if it exists, is
it is
(if
any)
and (I'R
any),
(if
term,
first
is
be
all
member
member
of G.'R
is
The
relative
D'R n Q.'R,
Similarly
of
the only
is
will
the only
R and S is
follower
is
a pre
These conditions
between x and z when there is an intermediate term y such that x has the
relation R to y and y has the relation S to z. The relative product of R and
S is represented by R S thus we put
;
RlS^^K^.xRy.ySz}
whence
" is
is
Df,
is
it
The
"
maternal grand
relative product
is
i.e.
\.(P\Q)\R = P\(Q\R).
It also obeys the distributive
relations,
i.e.
logical addition of
we have
h.P\(QuR) = (P\Q)v(P\R),
\.(QvR)\P = (Q\P)v(R\P).
But with regard
we have only
b.P\(QnR)G(P\Q)r.(P\R)
h.(QnR)\PG(Q\P)n(Q\R).
>
The
have in general
R R = R. We
put
R*
= R\R
Df.
tautologjr,
i, e .
we do not
I]
Thus
= (father)
grandmother = (mother)
2
paternal grandfather
maternal
A relation
i.e.
2
.
when
called transitive
is
G R,
i.e.
when,
if
when
xRy yRz
.
"^
.
x ,y,z
xRz.
x>y .y> z
P
35
t.
Dx
thus
e.g.
,x>z.
is
precedes
z,
The
z."
class of relations
which generate
transitive and
If
but
i.e.
For
in this case
we have
if
i.e.
there
Many
possession of
R to some
Thus
then
for
member
example
let
and a the
integral values of
of
R be
R"a = inhabitants
rationals,
a, i.e. all
class of
of a class a
is
denoted by
definition
of towns.
Let
then
is
R"a
1.
If
P is
series,
and a is any class of members of the series, P"a will be predecessors of as, i.e. the
segment defined by a. If P is a relation such that P ly always exists when
yea, P"a will be the class of all terms of the form P'y for values of y which
are
members
of a
i.e.
be the father of
y,
where y is some great man. In other cases, this will not hold for instance,
then
let P be the relation of a number to any number of which it is a factor
;
32
INTRODUCTION
36
[CHAP.
P" (even
factor apiece.
identical with x,
i'x
i.e.
"i'x"
The
a;
means "the
x and y
class consisting of
aw
to a class a will be
be a
will
i'x.
(We
= ${y = x)
Df,
class of objects
be i'x
will
\j i
x."
l'x\
write a
ft
as
an abbreviation
for a
r\
ft.)
It will be observed that unit classes have been defined without reference
number 1
number is defined
to the
we use
in fact,
number
1.
This
i.e.
= a{{<&x).a = l'x}
Df.
This leads to
I
From
:.
e 1
r
"
= (&x) y e a = y y =x:
% (<f>z)
is
a unit
If a e 1, t'a
is
e 1
% (<t>z)
a stands for
" (ix)
(<f)x),"
and
the only
z(<f>z).
is
= E (ix) {x e a),
= E (ix) (<f>x),
!
member
whence
i.e.
i.
of
a,
for
Thus
In practice, "t'a"
the only
"iV
is
x satisfying
<f>fc
member
exists."
of a
is
the only
nitions belong, for the most part, rather to mathematics than to logic.
reader
who
The
has mastered the symbols explained above will find that any
CHAPTER
II
theory of logical types, to be explained in the present Chapter, reitself to us in the first instance by its ability to solve certain
commended
on
its
own
it
An
analysis of the paradoxes to be avoided shows that they all result from
a certain kind of vicious circle*. The vicious circles in question arise from
means of the
can only be
" all
propositions are either true or false." It would seem, however, that such a
by statements about
it
cannot do
We
if
new
shall, therefore,
More
generally, given
have a total, it will contain members which presuppose this total, then such a set cannot have a total.
By saying, that a set has "no total," we mean, primarily, that no significant
statement can be made about "all its members." Propositions, as the above
any
illustration shows,
must be a
we suppose the
set having
This
The
is
when
sets,
is
true, as
we
shall
set to
a.
In such
each of which
is
cases, it is
capable of a
at effecting.
may be
a collection must not be one of
provided a certain collection had a total,
stated as follows:
the collection";
"Whatever involves
or,
conversely: "If,
all of
would have members only definable in terms of that total, then the said
collection has no total." We shall call this the " viciouscircle principle," because it enables us to avoid the vicious circles involved in the assumption of
illegitimate totalities. Arguments which are condemned by the viciouscircle
it
Revue de Metaphysique
et
de Morale,
Cf. also
Mai
1906, p. 307.
et
38
INTRODUCTION
[CHAP.
circumstances,
may
clusions to which
it
fallacious.
" all
is
is
true or false,
in
some way limited before it becomes a legitimate totality, and any limitawhich makes it legitimate must make any statement about the totality
tion
fall
who
asserts that
is
if
The paradoxes
sitions., classes,
as
we
cardinal
etc.
fallacies.
explained in Chapter III) which reduces statements that are verbally con
cerned with classes and relations to statements that are concerned with
propositional functions, the paradoxes are reduced to such as are concerned
The paradoxes
that concern
We
functions.
all
sitional functions.
sitional function.
We
PROPbSITIONAL FUNCTIONS
II]
and
its
avoidance leads, as
we
to a propositional
39
is
very
where
When we
etc.,
say that
"<f>x"
ambiguously denotes
<f>a, <f)b,
<f>c,
<f>a, <$>b,
etc.,
one, but
an undetermined one.
meaning
be ambiguous)
a function
is
if
the objects
etc., Ave
mean
of
it is
are welldefined.
It follows
defined.
$>a,
</>c,
all its
its
essence to
That
is
to say,
its
values
anything which presupposes the function, for if it had, we could not regard
the objects ambiguously denoted by the function as definite until the function
was definite, while conversely, as we have just seen, the function cannot be
definite until its values are definite. This is a particular case, but perhaps the
It will
It is sufficiently
* "When the word " function " is used in the sequel, "propositional function "
Other functions will not be in question in the present Chapter.
is
always meant.
40
INTRODUCTION
[CHAP.
its being necessary to apprehend its values severally and individually. If this
were not the case, no function could be apprehended at all, since the number
of values (true and false) of a function is necessarily infinite and there are
necessarily possible arguments with which we are unacquainted. What is
necessary is not that the values should be given individually and extensionally,
but that the totality of the values should be given intensionally, so that, concerning any assigned object, it is at least theoretically determinate whether or
is
we
(Any other
letter
in place of x.)
is
a proposition," but
sitional
is
a proposition,"
may be used
"<]>& is
a prepo
we mean to state
though we do not decide
We
Now
given a function
<f>,
all
pro
must be no propositions, of
the form <px, in which x has a value which involves
(If this were the case,
the values of the function would not all be determinate until the function
was determinate, whereas we found that the function is not determinate unless
its values are previously determinate.) Hence there must be no such thing as
the value for <f>Zc with the argument <&, or with any argument which involves
That is to say, the symbol "<f> (<f>)" must not express a proposition, as
cf>x.
"<f>a" does if <f>a is a value for </>. In fact "<f> (<f>)" must be a symbol which
does not express anything: we may therefore say that it is not significant. Thus
given any function <j>x, there are arguments with which the function has no
value, as well as arguments with which it has a value. We will call the
arguments with which <px has a value "possible values of x." We will say
that cf)x is "significant with the argument x" when <f>x has a value with the
argument x.
positions of the form
<px.
<j><jc.
We
case the
shall speak in this Chapter of "values for (f>x" and of "values of 4>x" meaning in eacli
thing, namely (pa, <f>b, <pc, etc. The distinction of phraseology serves to avoid
same
ambiguity where several variables are concerned, especially when one of them
is
a function.
II]
When
"
41
said that
it is
false, it is
e.g.
(<2)" is meaningless,
<f>
argument a, we assert <pa, we are not meaning to assert "the value for <j>^c with
the argument a is true"; we are meaning to assert the actual proposition
which is the value for <f>x with the argument a. Thus for example if $ is
"x is a man,"
(Socrates) will be "Socrates is a man," not "the value for
the function x is a man/ with the argument Socrates, is true." Thus
<f>
"</>
nonsense, but
we can
(#2)"
man'
is
is
meaningless,
we cannot
is
is false,
is
is
the function.
We
will
<f>x"
the proposition
i.e.
<f>,
<f>&.
"<f>x
always*,"
This proposition
assertion of
than
what
"<j>x
judged.
is
When we
is
Definition
"all
men
is
are mortal,"
we judge
judge
is
but
truly,
it
need
be
mortal."
Since "(x)
principle,
<f>.
asserting
We
*
will
mean
all
in
some
cases."
all
is false."
Such a
proposition,
we
Similarly "sometimes"
INTRODUCTION
42
Hence we
true.
[CHAP.
because "p
false,
is
{(p)
is
^ se
is false,"
i.e. we should be led to a proposition in which "(p) .p is false" is the argument to the function "p is false," which we had declared to be impossible.
Now it will be seen that "(p).p is false," in the above, purports to be a
and that, by the general form of the viciousmust be no propositions about all propositions. Nevertheless, it seems plain that, given any function, there is a proposition (true or
Hence we are led to the conclusion that "p is
false) asserting all its values.
false" and "q is false" must not always be the values, with the arguments p
and q, for a single function " p is false." This, however, is only possible if the
word "false" really has many different meanings, appropriate to propositions
proposition about
all
propositions,
of different kinds.
not difficult to
is
Let us take any function <f>x, and let <f>a be one of its values. Let us call
the sort of truth which is applicable to </>a "first truth." (This is not to assume
that this would be first truth in another context: it is merely to indicate that
it is the first sort of truth in our context.) Consider now the proposition
see.
(x)
<f>x.
that (gav)
thus we
for
<f>x
(f)x
may
has
Thus
to falsehood.
for
"
has
<j>$
{(a#)
4>x)
falsehood,"
first
if
<a?}
"(g#)
falsehood,"
i.e.
"(qx)
"(x)
(cf)X
has
first
(</>#
mean
falsehood)."
is
has
first
value
falsehood),"
while
Thus the
different
mean "some
$$ have
first
to a particular proposition.
first
If,
for
we
see
example,
is meant, the function "p has first falsehood" is only signiwhen p is the sort of proposition which has first falsehood or first
Hence "{p).p is false" will be replaced by a statement which is
falsehood
ficant
truth.
have
is false,"
first
falsehood."
first
truth or
first
falsehood
is
not
II]
"p has
"<f>
43
first
{(#)
Thus the
must be meaningless
falsehood."
<j>x}"
disappears.
Similar considerations will enable us to deal with "notp" and with "p or q."
if these were functions in which any proposition might
appear as argument. But this is due to a systematic ambiguity in the meanings of "not" and "or," by which they adapt themselves to propositions of any
order. To explain fully how this occurs, it will be well to begin with a
It
might seem as
The universe
consists of objects
in various relations.
Some
When an
object
complex.
and falsehood.
complex,
is
it consists
of interrelated parts.
Let
is
namely a and b and R and the percipient. The perception, on the contrary, is
a relation of two terms, namely "aintherelationi2to6," and the percipient.
Since an object of perception cannot be nothing, we cannot perceive "aintherelationi?to6 " unless a is in the relation jR to b. Hence a judgment of
perception, according to the above definition, must be true. This does not
mean that, in a judgment which appears to us to be one of perception, we
are sure of not being in error, since we may err in thinking that our judgment
has really been derived merely by analysis of what was perceived. But if our
judgment has been so derived, it must be true. In fact, we may define truth,
where such judgments are concerned, as consisting in the fact that there is a
complex corresponding to the discursive thoughtwhich is the judgment. That is,
when we judge "a has the relation R to b" our judgment is said to be true
when there is a complex "<xintherelationJSto6," and is said to be false
when this is not the case. This is a definition of truth and falsehood in relation to judgments of this kind.
It will be seen that, according to the above account, a judgment does not
have a single object, namely the proposition, but has several interrelated
objects. That is to say, the relation which constitutes judgment is not a
relation of two terms, namely the judging mind and the proposition, but is a
of the proposition.
44
INTRODUCTION
[CHAP.
When
this."
a judgment
alone.
It follows from
of a judgment,
not one.
is
judgment has
in judgment
several objects,
(as
opposed to
they do not appear to have realized clearly what was meant by this epithet.
Owing
what we
call
judgment,
is
it
follows that
distinguished from
judgment
proposition 'Socrates
The judgments we have been dealing with hitherto are such as are of the
same form as judgments of perception, i.e. their subjects are always particular
and definite. But there are many judgments which are not of this form. Such
are "all men are mortal," "I met a man," "some men are Greeks." Before
dealing with such judgments, we will introduce some technical terms.
We will give the name of "a complex" to any such object as "a in the reR to b" or "a having the quality q" or "a and b and c standing in the
lation
universe and
We
not simple.
will call a
sponding complex.
But take now such a proposition as "all men are mortal." Here the
judgment does not correspond to one complex, but to many, namely "Socrates
* See Chapter III.
GENEBAL JUDGMENTS
II]
is
is
mortal," "Plato
is
45
it
men
judgment that
all
judgments. It
is not,
number
of elementary
{e.g.)
is
mortal
is
not necessary to
radically
judgments.
men
are mortal,
where x
is
we
mean
shall
that
all
We may define
if
we
is
mortal,"
this as "truth of
and
judgment
call this
is
mortal,
p, then
"p
is
where
a?
is
a man,"
where
a;
is
a man."
46
INTRODUCTION
may
take are
all
[CHAP.
is
significant.
We
asserts all
mortal"
equivalent to
is
"(x)
i.e.
man' implies
\x is a
'x is a mortal,'"
"(x)
As we have
is
not a
man
or
is
mortal."
meaning of truth which is applicable to this proposition is not the same as the meaning of truth which is applicable to "# is a
man" or to "x is mortal." And generally, in any judgment (x) <f>x, the sense
in which this judgment is or may be true is not the same as that in which <f>x
just seen, the
is
may be
or
to
true.
If
<f>x is
an elementary judgment,
But
a corresponding complex.
(x)
<f>x
it is
true
when
it
points
values of
x.
made by Epimenides
his
judgments
order.
may make n
assertions of the form "all the judgments of order m made by Epimenides are true," where m has all values up
to n. But no such judgment can include itself in its own scope, since such a
judgment is always of higher order than the judgments to which it refers.
the nth
is
the highest, we
"(a?)
We
$x."
is
of
or
"<f>x
sometimes,"
is
a judgment which
whole symbol may be read "there exists an x such that <f>x." We take the
two kinds of judgment expressed by "(x) <j>x" and "(g#) <f>x" as primitive
.
The
<f>x.
(f>x and (gp)
be denoted by the symbol "~p." Then the
.
negation of (x)
<f>x
will
will
be defined as meaning
"(
(gp?)
<f>x
will
a #).~<K'
be defined as meaning "(x)
<f>x."
mative
is
meaning of negation
for
such propositions
is
affir
Hence the
from the meaning of
particular affirmative
Thus,
different
SYSTEMATIC AMBIGUITY
II]
An
47
We
ment
We
suppose that
an elementary proposition, and that $x is always an elementary proposition. We take the disjunction of two elementary propositions as
a primitive idea, and we wish to define the disjunction
is
"p v
.
This
is
to
(x)
<f)X."
may be defined as "(x) .pv <j>x" i.e. "either p is true, or <f>x is always
mean " 'p or <f>x' is always true." Similarly we will define
"P v (3^) $x
as meaning "(g#) .p v
for
which
$x v
"either
<f>x
is
always true."
<f>x," i.e.
we
we can
"
define "either
is
Similarly
true."
"( x )
<f>x
an x
is
is
for
is
an x
which either p or
(f>x is
true or there
(y)
will
yfry"
By
this
method we
.
meaning
forms (x)
<f>x
vtyy,"
i.e.
or tyy' is
obtain definitions of disjunctions con
true"
<f>x.
(f>x
or (ftx)
(qx)
<j>x
in terms of disjunctions
of "disjunction"
<f>x,
as
it
'"<f)X
was
for
is
elementary pro
positions.
and
disjunction.
IV.
The
Why a
it
is it
yjr^ is
<f>z
to
have
itself
another function
such that there are arguments a with which both "<f>a" and "yfra" are significant, then yfr$ and anything derived from it cannot significantly be
way
arises
if it is to
is
essentially
it
must
unambiguous statement has resulted. A few illustrations will make this clear.
Thus "(x) <f)x," which we have already considered, is a function of <j>; as soon
.
as <& is assigned,
But
is
it is
obvious that
not a function: "(x).<f)x" means "<# in all cases," and depends for its
upon the fact that there are "cases" of <f>%, i.e. upon the
significance
48
[CHAP.
INTRODUCTION
ambiguity which
the fact that,
which
is
characteristic of a function.
significantly as argument,
something
is
versely,
is
it
up
in such a statement as
possible.
Hence
V.
We
"{(x)
we
"p
is
<#} is
But
a man," where
a
man"
is
is
a proposition, this
is
not
meaningless.
are thus led to the conclusion, both from the viciouscircle principle
and from direct inspection, that the functions to which a given object a can
be an argument are incapable of being arguments to each other, and that they
have no term in common with the functions to which they can be arguments.
We are thus led to construct a hierarchy. Beginning with a and the other
terms which can be arguments to the same functions to which a can be argument, we come next to functions to which a is a possible argument, and then
to functions to which such functions are possible arguments, and so on. But
the hierarchy which has to be constructed is not so simple as might at first
The functions which can take a as argument form an illegitimate
appear.
totality, and themselves require division into a hierarchy of functions.
This
is easily seen as follows.
Let f (<j>z, x) be a function of the two variables $z
and x. Theii if, keeping x fixed for the moment, we assert this with all possible
values of
<f>,
we obtain a
proposition
(<f>).f(^,x).
Note that statements concerning the significance of a phrase containing "02" concern the
symbol "<p2," and therefore do not fall under the rule that the elimination of the functional
ambiguity is necessary to significance. Significance is a property of signs. Cf. pp. 40, 41.
*
t Cf Chapter III.
.
n]
49
requires that
"whatever
since
we
We
of identity.
is
function of
we
shall
all
values of
"<j>x
implies
<f>y
"
regarded as a
<j>
<f>
function of x;
<f>y"
we
shall
hence, if
identical with a,
a legitimate value of
it is
be able to
infer,
and x
is
<f>
in
"<f>x
always implies
x is
y is identical with a.
sound, the reasoning embodies a viciouscircle
definition, that if
<f>,
This difficulty
to be explained later.
by the
If
we here
all
significant for a
afunctions,
and
replace a by a variable,
we obtain an
afunction; but
member of our
whole of the selection. Let the selection consist
those functions which satisfy/ (<f>z). Then our new function is
viciouscircle principle this afunction
cannot be a
of all
((f))
where x
is
afunctions
*
the argument.
we may make,
{f(<fiz) implies
It
<f>x},
lie
outside our
When we
speak of "values of <p$" it is <p, not z, that is to be assigned. This follows from
the explanation in the note on p. 40. When the function itself is the variable, it is possible and
simpler to write <j> rather than <f>z, except in positions where it is necessary to emphasize that an
R&W
to secure significance.
[CHAP.
INTRODUCTION
50
selection.
arise
"types," each of which contains no functions which refer to the whole of that
type.
When
something
is
some (undetermined)
apparent, after Peano.
The presence
really present
example "A
a variable time occurs as apparent variable.
for
The
Now
an apparent
variable, as will
be shown in Chapter
III.
Thus
in
what we have
in
awareness of oneself.
if
<f>^
complexity.
Thus we must
many
variables
51
MATBICES
II]
Thus
we
.
<f>
(x, y),
which
is
a function of
(x)
. <]>
(x,y),
which
is
a function of y,
y)
is
shall derive
(y)
(x,
This
for
from
x,
true with
all
possible values of
last is
it
i.e.
x and
y."
no variable except
apparent variables.
It is thus plain that all possible propositions
from matrices by the process of turning the arguments to the matrices into
apparent variables. In order to divide our propositions and functions into types,
we shall, therefore, start from matrices, and consider how they are to be divided
we
functions,
soandso."
The
first
matrices that occur are those whose values are of the forms
4>Xy^{x,y\x{oD,y,z...),
where the arguments, however many there may be, are all individuals.
tjr, %..., since (by definition) they contain
no apparent
functions
variables, and have no arguments except individuals, do not presuppose any
totality of 'functions.
From the functions ^r, ^ ... we may proceed to form
i.e.
The
<f),
other functions of
(y)
'
(3^)
X (x
>
x,
V> ?)>
We
of individuals.
such as (y)
on
an(^ so
yjr
(x, y),
(gy)
% {>, y, z),
x,
Such functions we
We may
We
will
function by
"<f>
x."
Thus
"<f>
x" stands
for
any value
for
a function of two variables, namely < Vz and x. Thus <f> I x involves a variable
which is not an individual, namely < ! z. Similarly "(x) . <f> ! x" is a function
of the variable
Again,
if
is
<j> 1
1,
"<f>lx implies
is
individual.
a given individual,
a function of
x,
but
it is
(f>
a with
all
possible values of
an (apparent) variable
which is not an
"predicate" to any firstorder function
I
<f>
<f>
^>
x,
<f>"
because
it
involves
individual.
tb.
42
INTRODUCTION
52
is
[CHAP,
Then the
ment
"<f>lx implies
<f)
a with
possible values of
all
<f>"
may be
state
apparent variables, but contains the two real variables cf> ! z and x. (It should
be observed that when <f> is assigned, we may obtain a function whose values do
involve individuals as apparent variables, for example
But
if
so long as
<
variable,
is
a definite individual,
is
<j> !
is
<f>
2,
yfr
t,
and so
<f>
on.
is
(y)
yfr
variables.)
(x, y).
Again,
two variables
if
x contains no apparent
<f)
We
yfr
b"
is
<f>
2.
a function of the
new
matrices,
by turning
all
its
new
Thus we
variables.
We will
.g(<$>\z,
">jrlz),
which
is
a function of
(x) ,F(<j>lz,
x\ which
is
a function of
<
(0) .F((f>lz,
x),
which
is
a function of
x.
yjr I
z.
z.
firstorder functions
firstorder functions
and
give the
individuals.
We
name
of second
order functions to such as either are secondorder matrices or are derived from
argument
no other variables
We
place,
firstorder function.
We
will
first
is a
denote a variable function of this kind by the
x,
(<
z),
fl(tf>l z) is a function of
<j> I
<f>
firstorder functions."
SECONDORDER FUNCTIONS
II]
In the second
of which
is
As soon
as x
place,
we have secondorder
53
we
assigned,
is
shall
a predicative function of x
simplest possible case, if /!
if
(<f>
us a predicative function of
f\
(<f>
2,
we
assign a value to
x)
z,
x, in
is
!
<f>
<
2.
we have
<f>
<f>
gives
But
x. "
if
We
x.
apparent variable.
We
If our
z.
shall obtain
of a value to
will all
<f>
<j> !
we
(</> !
z,
by turning
x)
for
<f>
into
an
them.
We may
which
now proceed
same way
in exactly the
to thirdorder matrices,
firstorder functions
We
its
argument,,
i.e.
it is
of the
having that argument. If a function has several arguments, and the highest
if it is
compatible with
of the
its
f
1th order,
i.e.
the nth,
again, if
it
has.
it is
we
call
the function
function of several
arguments is predicative if there is one of its arguments such that, when the
other arguments have values assigned to them, we obtain a predicative function
of the one undetermined argument.
It is important to observe that all possible functions in the above hierarchy
where
./!
(<f>
is
z,x) or (gtf>)
/!
(<f>
% x) or
(<f>,
f) ./!
(<f>
And
z,
f
% x) or
etc.,
speaking generally, a
INTRODUCTION
54
[CHAP.
all
we need not
may be
<j> 1
(<)
if
$ u
!
(</> !
is
, x).
Thus
F is a predicative
The nature
by a succession
of steps
we find
that,
predicative function of
where
speaking generally,
will
function of
<f>
u and
x.
may be
restated as follows.
meaning
what comes to the same thing, the totality of
its possible arguments. The arguments to a function may be functions or
propositions or individuals. (It will be remembered that individuals were
defined as whatever is neither a proposition nor a function.) For the present
we neglect the case in which the argument to a function is a proposition.
Consider a function whose argument is an individual. This function prefunction, as
the totality of
we saw
at
an
apparent variables,
however,
it
it
it
contains functions as
it
If,
cannot
be defined until some totality of functions has been defined. It follows that
first define the totality of those functions that have individuals
we must
as
is
variables.
These are
the possible values of the argument, and those that are presupposed by any
Thus a
argument
is
argument.
that a function
the function
is
55
n]
"
is
is
human
" is not
Elementary and
human.")
its
values
no
totality
except (at most) the totality of individuals. They are of one or other of the
three forms
<j>lx; (x).<f>lx;
where
<f>
x
{Qx).<p\x,
,
is
If follows that, if
sition,
and
The
VI.
meaning
some given
" all
object.
argument a"
In some
will
cases,
we can
properties of a,"
see that
harm results from regarding the statement as being about " all properties of
a" provided we remember that it is really a number of statements, and not
a single statement which could be regarded as assigning another property to
a, over and above all properties. Such cases will always involve some systematic ambiguity, such as that involved in the meaning of the word "truth,"
as explained above.
Owing
it will
be
possible,
sometimes, to combine into a single verbal statement what are really a number
of different statements, corresponding to different orders in the hierarchy.
This
is
statements
should be broken up
kind of falsehood.
56
INTRODUCTION
[CHAP.
For two
<j>
(&, p),
variables,
there
is
we
(^)
=x
(f>x
yjrl x.
(a^r)
<f>
(x,
y)
= x>y
yfr
i.e.
(x, y).
In order to explain the purposes of the axiom of reducibility, and the nature
it true, we shall first illustrate it by applying it
some particular
to
If
we
call
cases.
among
Take
for
all
make
a great general."
We may interpret
this as
is
true of
its properties.
all
is
z) implies
<j> !
(Napoleon).
common and peculiar to great generals. In fact, it is certain that there is such
a predicate. For the number of great generals is finite, and each of them
certainly possessed some predicate not possessed by any other human being
for example, the exact instant of his birth. The disjunction of such predicates
will constitute a predicate common and peculiar to great generals*. If we
call this predicate yfr z, the statement we made about Napoleon was equivalent to yfr (Napoleon). And this equivalence holds equally if we substitute
any other individual for Napoleon. Thus we have arrived at a predicate which
is always equivalent to the property we ascribed to Napoleon, i.e. it belongs
to those objects which have this property, and to no others. The axiom of
!
When
(finite) set of
predicate, because
i.e.
is
It]
57
some predicate.
We may next
our principle by
illustrate
its
application to identity.
In
It is plain that, if
x and y are
identical,
and
it
for
any
<j>x
implies
is true,
<f>cc
may be
then $y
is true.
Here
But we cannot say, conversely " If, with all values of <,
then
x and y are identical " because "all values of <f> " is
<j>y,
inadmissible. If we wish to speak of "all values of <f>," we must confine
function.
We may confine
We may
and
and
its
if all
of x belong to
y, for to
have
all
say
" all
the
so on.
example,
all
to predicates, or to
<f>
the predicates of
a;
is
y,
predecessors
then
all
for
predicates
a secondorder property,
and therefore
make
to
this property
Some
things indiscernible
is
limitation of the
common
therefore implied
properties necessary
common
Then the
axiom of
reducibility.
belong to
y,
y has
all
x, since
is
the predicates of x
<f>
hence y
as identical
.
<f>
y.
We
is
when
it
belongs
identical with x.
all
It
the predicates of x
definition of identity*:
x=y. = :(<f>):<f>lx.D.<f>ly
*
is
defined
is
Df.
[CHAP.
INTRODUCTION
58
But apart from the axiom of reducibility, or some axiom equivalent in this
we should be compelled to regard identity as indefinable, and to
connection,
The axiom
of reducibility is even
more
we assume
the existence of
function
$z
is
Hence "<#"
is
equivalent to
"x belongs
to
or."
that
We
all
is
fixed)
form.
The axiom of
reducibility
is
II]
we
is
59
satisfy a function
z),
there
assert that
is
substituted for the axiom of reducibility in symbolic deductions, since its use
would require the explicit introduction of the further assumption that by a
finite number of downward steps we can pass from any function to a predicative
function, and this assumption could not well be made without developments
that are scarcely possible at an early stage. But on the above grounds it seems
plain that in fact, if the above alternative axiom is true, so is the axiom of
reducibility. The converse, which completes the proof of equivalence, is of
course evident.
Axiom of Reducibility.
is selfevident is
is
is
new evidence
its
Infallibility is
never attainable, and therefore some element of doubt should always attach
to every axiom and to all its consequences. In formal logic, the element of
doubt
is less
although
it
But
false,
Here the combination or disjunction is supposed to be given intensionally. If given extenbut in this case the number of
(i.e. by enumeration), no assumption is required
predicates concerned must be finite.
sionally
60
INTRODUCTION
[CHAP.
it is
VIII.
We
are
now
The Contradictions.
in a position to
For
this
we
purpose,
shall
will
he
lying,
(2)
is
Let
be the
x may
be,
"
"w
is
is
"w
w"
is
is
equivalent to
w"
is
"oc is
equivalent to
not a w."
(3) Let T be the relation which subsists between two relations R and 8
whenever R does not have the relation R to S. Then, whatever relations
R and S may be, "R has the relation T to S" is equivalent to "R does not
have the relation R to S." Hence, giving the value T to both R and 8,
"T has the relation T to T" is equivalent to "T does not have the relation
T to
T."
BuraliForti's contradiction*
(4)
may be
stated as follows
It can be
shown that every wellordered series has an ordinal number, that the series of
ordinals up to and including any given ordinal exceeds the given ordinal by
one, and (on certain very natural assumptions) that the series of all ordinals
(in order of
*
"Una
xi. (1897).
magnitude)
is
wellordered.
questione sui numeri transfiniti," Rendiconti del circolo matematico di Palermo, Vol.
See *256.
ENUMERATION OF CONTRADICTIONS
n]
an ordinal number,
ordinals has
X2 say.
But
61
number O +
number of all
than
.XI.
nameable
Among
(6)
transfinite ordinals
some can be
which
ordinal,"
(7)
is
as follows
number
hence
is
a contradiction J.
Richard's paradox
:
Consider
of words
let
is
It
all
Let
be a number
defined as follows: If the nth figure in the nth decimal is p, let the nth
figure in
be p + 1 (or 0, if p = 9). Then
is different from all the members
of E, since, whatever finite value n may have, the nth figure in
is different
from the nth. figure in the nth of the decimals composing E, and therefore
is different from the nth decimal. Nevertheless we have defined
in a finite
number of words, and therefore
ought to be a member of E. Thus
both
is and is not a member of E.
its
....
In
all
indefinite
number) there
is
as selfreference or reflexiveness.
itself in its
selves, are
own scope. If all classes, provided they are not members of themmembers of w, this must also apply to w and similarly for the
;
* This contradiction
was suggested
to us
by
Mr
The
definition,"
and
is
Cf. Poincare,
Mai
(1906), p. 149
fit.
[OHAP.
INTRODUCTION
62
whose ordinal number causes the difficulty is the series of all ordinal
numbers. In each contradiction something is said about all cases of some kind,
and from what is said a new case seems to be generated, which both is and is not
of the same kind as the cases of which all were concerned in what was said.
series
But
this
is
Hence
them
in
constructed.
When
(1)
defined
all
we have
we
our contradictions are illustraIt only remains to show, therefore, that the
man
says "I
am
ment
is
fall
within
its
own
scope,
Hence the
state
and therefore no
contradiction emerges.
we
"I
am
te
As no
order
the
is
first
is
False.
This statement
am
is
statement "I am making a false statement Of the second order" is true. This
is a statement of the third order, and is the only statement of the third order
which is being made. Hence the statement "I am making a false statement
of the third order"
false
is false.
statement of order 2n
a false statement of
contradiction.
(2) In order to solve the contradiction about the class of classes which are
not members of themselves, we shall assume, what will be explained in the
next Chapter, that a proposition about a class is always to be reduced to a
class,
i.e.
VICIOUSCIRCLE FALLACIES
II]
63
class is
and
is
whenever
relation
defined by
is
function
<j>
whenever
"R
be the arguments to
significantly
</>
presupposes
4>.
its
Suppose the
to S."
But (assuming,
and
can
as will appear in
is
defined in
^>, and this no function can do, as we saw at the beginning of this
Chapter. Hence "R has the relation R to S" is meaningless, and the contra
terms of
diction ceases.
(4)
The
developments
a series
is
At
this stage, it
number
is
must
some further
suffice to
class of series.
observe that
(These state
ments are justified in the body of thework.) Hence a series of ordinal numbers
is a relation between classes of relations, and is of higher type than any of the
series which are members of the ordinal numbers in question. BuraliForti's
"ordinal
(5)
For the word " nameable" refers to the totality of names, and yet is allowed
to occur in what professes to be one among names.
Hence there can be no
such thing as a totality of names, in the sense in which the paradox speaks
*
*256.
The
means
is
given in detail ia
INTRODUCTION
64
of "names." It
is
[CHAP.
We may, in
Elementary names
will be such as are true "proper names," i.e. conventional appellations not
involving any description, (b) Firstorder names will be such as involve a
distinguish
description
names
(a)
is to say, if
<j> I
&
is
first
order function, "the term which satisfies <j>\" will be a firstorder name,
not always be an object named by this name, (c) Secondbe such as involve a description by means of a secondorder
function; among such names will be those involving a reference to the totality
of firstorder names. And so we can proceed through a whole hierarchy. But
though there
order
names
will
will
at no stage can
which specify the type to which the definition is to belong. And however
the type may be specified, "the least ordinal not definable by definitions of
this type" is a definition of a higher type; and in Richard's paradox, when
we confine ourselves, as we must, to decimals that have a definition of a given
type, the number N, which causes the paradox, is found to have a definition
which belongs to a higher type, and thus not to come within the scope of our
previous definitions.
An
indefinite
number
terms of
itself;
and
members defined
in
In most
to a contradiction,
type.
VICIOUSCIRCLE FALLACIES
Il]
65
valid.
Thus by employing typically ambiguous words and symbols, we are able to make
one chain of reasoning applicable to any one of an infinite number of different
cases, which would not be possible if we were to forego the use of typically
ambiguous words and symbols.
Among
may differ, in respect of truth or falsehood, according to the typical determination which they receive, are existencenotions practically the only ones which
theorems.
total
22
is
existencetheorems which, hold for higher types but not for lower types. Even
number of individuals is not asserted, but is
merely assumed hypothetical^, we may replace the type of individuals by any
other type, provided we make a corresponding change in all the other types
here, however, so long as the
members
That
is,
we may
give the
name
"relative in
R&W
t,
CHAPTER
III
INCOMPLETE SYMBOLS
Descriptions..
(1)
is
a symbol which
only defined in
is
'certain contexts.
In ordinary mathematics,
for
example,
^ and J^
are in
we have anything
complete symbols: something has to be supplied before
be called a "definition in use."
significant. Such symbols have what may
Thus
if
we put
22
3*
da?
df
_+V!
y =
&
;?+ dz*
Df,
we
but
V by
2
itself
For
we cannot
first
exist,
we have a
"~E l(ix)(<f>x)"
~{(ac):^c.=*.0 = c},
i.e.
or
some equivalent.
"~E!(?#)(<M," (*)(**)
symbol.
* Cf. pp. 30, 31.
is
(ix)(<f>x)
has
an ^complete
CHAP.
DESCRIPTIONS
Ill]
By
67
(ix)
(<f>x) is
"
Scott
is
a = (ix)(<f>x)
is
one which
= a;
may be
true or
may be
false,
but
is
never merely
trivial, like
be either
this
may be
anything,
has meaning,
it
it
(<f>x) is
nothing.
it
might be suggested that " Scott is the author of Waverley " asserts that
"Scott" and "the author of Waverley" are two names for the same object.
But a little reflection will show that this would be a mistake. For if that
were the meaning of " Scott is the author of Waverley," what would be required
for its truth would be that Scott should have been called the author of
Waverley: if he had been so called, the proposition would be true, even if
some one else had written Waverley; while if no one called him so, the propositionwould be false, even if he had written Waverley. But in fact he was
the author of Waverley at a time when no one called him so, and he would
not have been the author if every one had called him so but some one else
had written Waverley. Thus the proposition "Scott is the author of Waverley"
is not a proposition about names, like "Napoleon is Bonaparte"; and this
illustrates the sense in which "the author of Waverley " differs from a true
proper name.
It
Thus
all
phrases (other than propositions) containing the word the (in the
For
or " Scott
is
" the
which it plainly does not; nor can "the author of Waverley" mean
anything other than " Scott," or " Scott is the author of Waverley " would be
false.
Hence "the author of Waverley" means nothing.
Scott,"
It follows from the above that we must not attempt to define " {ix) (<#),"
but must define the uses of this symbol, i.e. the propositions in whose symbolic
expression it occurs. Now in seeking to define the uses of this symbol, it is
important to observe the import of propositions in which it occurs. Take as
52
[CHAP.
INTRODUCTION
68
an illustration: "The author of Waverley was a poet." This implies (1) that
Waverley was written, (2) that it was written by one man, and not in collaboration, (3) that the one man who wrote it was a poet. If any one of these fails,
the proposition is false. Thus " the author of Slawkenburgius on Noses was
'
because no such book was ever written; " the author of The
Maid's Tragedy' was a poet" is false, because this play was written by
Beaumont and Fletcher jointly. These two possibilities of falsehood do not
a poet "
is false,
'
arise if
"
x wrote Waverley,"
a poet."
Now
taking
of
to replace
it is
requires (1) (gar) . (<f>x) and (2) </>x . 4>y . D XtV . x = y; here (1) states that at
least one object satisfies <f>x, while (2) states that at most one object satisfies
<f>x.
are equivalent to
(ftc):<f>x.= x
which we defined as
(ix)(<f>x).
<f>x
=x
x = c. Thus we
f{(ix) (4>x)}.
(ix)
(<f>x).
of
what
is/
If our proposition
.x=c,
is
[(ix)(<f>x)},
what
further affirmed
is
is
have
= :(Rc):<j>x.= x .x = c:fc
satisfies /#"
Df,
i.e.
"the x satisfying
that
<fix is
true when,
<f>x
is
to
is c,
The
"<f>x
= x ,x=a."
For,
by the
is
i.e.
to " <px
= x x = a."
.
"
x wrote Waverley
Thus although
be equivalent to
to
.x=c:a = c
shown
definition, it is
(gc) z<j)x.= x
" there is a c for
easily
" is true
is
always equivalent to x
'
when x
"(ix)((f)x)" has
is
is
Scott,'"
when x
is
not Scott.
some other
proposition,
we
We
(<f)x)
than "/{('*)
(<P
X )}" in future.
is
69
Ill]
We may interpret
this as
~{El(ix)(<l>x)},
or as
<^
if " <f>x"
that a proposition
is
is
King
{(gc)
= (ix) (<j>x)},
In either
of France."
in which (ix)
(<f>x)
occurs
The same
and
asserted is
is
this proposition
King
what
case,
is false,
of France would be
House of Orleans."
is
ambiguous;
(ix)
(<fix)
it
may deny
it
f{(ix)(<f>x)}, in
which case
it
will
be true
if
may mean
(<f>x)
exists.
In ordinary language,
For example, the propo
being held to
"it
is
false
^(ix)(<f>x).D.p,
p.^.^(ix)(^>x),
(ix)
>/r
(<f>x)
D x 0#) (<H>
.
and
we must be able by our notation to distinguish whether the whole
or only part of the proposition concerned is to be treated as the "f(ix) (<f>x)"
of our definition. For this purpose, we will put " [(ix) (<f>x)]" followed by dots
at the beginning of the part (or whole) which is to be taken as f(ix) (<f>x), the
so on,
i.e.
f(ix)(<f>x)
to
Thus
"f 0*0 (0*0 D P
= x x = c tyc D p,
[(ix) (</>#)]
will
mean
but
will
It is
(gc)
<f>x
[(ix) (<f>x)]
mean
(gc)
<j>x
( ix)
=x x = c
.
first is
false.
(<f>x)
tyc
D p
.
Again
[(ix) (<f>x)]
yjr
p.
(ix)
((f>x)
not
exist,
the
INTRODUCTION
70
mean
will
(gc)
\jr
mean
Here
.= x .x = c
<frx
while
will
[CHAP.
when
again,
(7a?) (cfrx)
yjrc,
(<f>x)}
yjrc}.
first is false
amend our
By means
of this definition,
[(lx)((f>x)],f(ix)((j)x).
(ix)(<f>x),
we
Df.
true.
D p
.
the scope
is
f(i%) (##)
~ {[(ix)
in
the scope
is
the scope
is
(<f>x)]
(<)>x)
.D.p
P'>
./(ix)
(4>x)}
but in
(ix) (<f>x);
[(ix)(<t>x)].~f(ix)(4>x)
~y(?#) (<f>x).
be seen that when
It will
concerned for
(ix)(<jix);
its
but when'
may
proposition in which
as
we
(ix)
concerned
much
true unless
(ix) (<px)
(<f>x),
occurs.
it
If a proposition contains
we have
to distinguish
distinguish
(1)
[(ix)
(2)
[(ix) (yfrx)]
The
first
(<H]
(gc)
(3)
(<f>x),
(ix) (fx)},
{(ix)
(<j>x),
(ix) (yjrx)}.
[(ix) (<f>x)]
<f>x.= x .x
(yjrx),
(gc) :.<j>x.= x
[(**) (yjrx)].f{(ix)
(ix)(<f)x).
= c:
[(ix)
(fx)]
becomes
x = c:.(<zd):fx.= x
results
if,
.x=c :f(c,
we eliminate first
when (ix)(<f>x)< and
in (1),
d),
(ix)(yjrx)
and
(ix)(yjrx) are
eliminated,
fad)
(5)
(4)
and
:. yfrx
=x x = d
.
:.
(gc) :<}>x.= x .x
= c :/(c, d).
is
71
CLASSES
Ill]
It will
scope
is,
Thus
1 0*)
[(**) (**)]
will occur
much more
this reason it is
of (ix)
(<f>x) is
:.
(</>*)
^
( *).
[(wO (W]
X (w) (W
3 X O) (**)
of an occurrence
Thus
mean
p D a = (ix) (<#)
p D [(ix) (4>x)~\ a = (J x ) (
p.D. (ga) a = (?#) (0a;)
p.D. (g) [(?&) (<f>x)] a = (ix) (<f>x)
.
and
e.g.
.
will
example
frequently than
[(M>) (a?)]
For
for
'>
mean
will
mean
mean
p D
and
will
but
will
~ {[(?#)
(</>#)]
a=
(ix)
(<f>x)}.
This convention enables us, in the vast majority of cases that actually
occur, to dispense with the explicit indication of the scope of a descriptive
symbol; and it will be found that the convention agrees very closely with the
Thus
for
example,
if
"0)(W
is
"
which would generally be allowed to hold if " the soandso " Hoes
not exist. Ordinary language is, of course, rather loose and fluctuating in its
implications on this matter; but subject to the requirement of definiteness,
our convention seems to keep as near to ordinary language as possible.
soandso,"
mean
will
(gc)
(f>x
=x x = c
.
will
= (ix) (yfrx)
[(ix)
(yfrx)']
while
mean
(<\>x)
(gd)
yjrx
=x x = d
.
easily
[(ix)
= (ix) (tyx),
(<f>x)
(<f>x)~\
shown
to
(ix)
(<f>x)
= d.
be equivalent.
descriptions, are, in
(2) Classes. The symbols for classes, like those for
our system, incomplete symbols their uses are defined, but they themselves
are not assumed to mean anything at all. That is to say, the uses of such
:
INTRODUCTION
72
[CHAP.
symbols are so defined that, when the definienste substituted for the definimdum,
there no longer remains any symbol which could be supposed to represent
a class. Thus classes, so far as we introduce them, are merely symbolic or
if
they are
individuals.
an old dispute whether formal logic should concern itself mainly with
In general, logicians whose training was mainly
philosophical have decided for intensions, while those whose training was
mainly mathematical have decided for extensions. The facts seem to be that,
It is
complete symbols.
definite proof,
was possible to prove that they are inIn the case of classes, we" do not know of any equally
it
classes, it is
necessary
first
This
is
effected
by
The
false.
Two
value,
(This expression
due
when they
if it is
truth
to Frege.)
i.e.
Two
is
and falsehood
if it is true,
formally equivalent to
is
is
"ob is identical
function of a function
argument
is
when they
when any argument which
versa. Thus "ob is a man " is
i.e.
with
is
an even prime"
is
2."
called extensional
when
its
That
is
to
If there is
Ill]
say, f(<f>z) is
an extensional function of
provided
<f>z if,
73
yfrz is
formally equiva
lent to
<f&, f(<f>z) is
We
function of a function
is
called intensional
when
it is
not extensional.
function of "&
is
A may
a man," because
may
person
who
men are
may have
all
featherless
Again the
z is n " is
unchanged if
number
<f>
<f>
all
since, if
North
'a?
is
Pole."
a white
"
$,
men
two white
man who
we
The
truth or
a white
claims to have reached the North Pole " any other statement which
man who
<p
is
unaffected
if
is
Hence it is an extensional
a strange coincidence that two white
men should claim to have reached the North Pole," which states "it is a
strange coincidence that two arguments should satisfy the function '& is a
white man who claims to have reached the North Pole,'" is not equivalent to
"it is a strange coincidence that two arguments should satisfy the function
same arguments, and
But the proposition "it
holds of the
function.
'ab
is
<f*t tb
Dr Cook
or
Commander
of no others.
is
Peary.'"
Thus
"it
is
is
an intensional function of
<f>lx.
INTRODUCTION
.74
[CHAP.
The above instances illustrate the fact that the functions of functions with
which mathematics is specially concerned are extensional, and that intensional
functions of functions only occur where nonmathematical ideas are introduced,
Hence
fact.
it is
When two
we may
the
same
extension.
its
In such a
argument.
case, it is convenient to
many and
an object, called a
class,
important,
which
is
it is
Since exten
all
the
Thus
e.g. if
we say
ment
it is
men, namely those who were Apostles, rather than as attributing the property
of being satisfied by twelve arguments to the function "& was an Apostle."
This view is encouraged by the feeling that there is something which is
We
may be
regarded
a predicative function
yfr
it is
is
extensional,
as well as the property of being significant (by the help of the systematic
ambiguity of equivalence) with any argument d>z whose arguments are of the
same type as those of yjr z. The derived function, written "f {z(<f>z)} ," is defined as follows: Given a function /(^ z), our derived function is to be "there
is a predicative function which is formally equivalent to <f>z and satisfies/."
If <f)Z is a predicative function, our derived function will be true whenever
f{4>z) is true. If f(<j>z) is an extensional function, and <j>z is a predicative
I
75
DEFINITION OF CLASSES
Ill]
may sometimes be
derived function
if
If f(<f>2)
is
not an ex
original function
But
is false.
in
always extensional.
is
true; thus in
<f& is
when the
true
is
it
for
any function
predicative function.
an argument
<f>z,
cative function
yfrlz,
it is
it is
ambiguity as to type that belongs to truth and falsehood, and can therefore
hold between functions of any two different orders, provided the functions
take arguments of the same type. Thus by means of our derived function we
have not merely provided extensional functions everywhere in place of intensional functions, but we have practically removed the necessity for considering differences of type among functions whose arguments are of the same
type. This effects the same kind of simplification in our hierarchy as would
z)
it will
be found that, in
be substituted
for
yfr I
is
z in
type as those of
f(<j>z)
At
the function
<f>z,
yjr
z~
is
this point,
<f>z,
in
If,
with a given
equivalent to f(yfr ! z)
then/ (2 ($2)} is equivalent to
interpreted,
formally equivalent to
is
provided there
there always
f(<f>z), so
Thus
argument
whenever
any function
! can significantly
is
is
As was explained
above,
it is
<f>%.
Now we
which
satisfy
<f>z,"
(4>z)}" will
INTRODUCTION
76
[CHAP.
reducibility.
if
<j>2
is
is
predicative function.
many
argument
in iavour of the
The above
axiom of
is
to be regarded as an
reducibility.
by the function
is,
cf>z,"
or rather,
in symbols, as follows:
/{S^}.:^):^.^.^!*:/^*}.
Df.
We
require of classes,
if
CLASSES
Ill]
77
(3) Conversely,
must be formally equivalent; in other words, when the class is given, the
membership is determinate two different sets of objects cannot yield the same
class.
(4) In the same sense in which there are classes (whatever this sense
may be), or in some closely analogous sense, there must also be classes of
classes. Thus for example " the combinations of n things m at a time," where
the n things form a given class, is a class of classes; each combination of
:
things
is
class,
is
member
classes.
Again,
is
number
It
the number
class is
therefore a class
is
1,
(5)
2.
have
k . =a a
k" is supposed
a
Since,
But
but
it
K ~
.
a.
significant, the
above equivalence,
K ~>
k,
i.e.
6 K.
this is a contradiction f.
meaningless.
a,
e
In general, there
is
In the
first place,
might appear as
if
the class of
all
classes
false.
were a
In the second
class,
i.e.
as if
But
this
com
meaning of
to have a different
be supposed
requisites,
it is
more
As explained
shortly, "a; is
of reducibility, there
(p. 25), "area" means "x is a member of the class a*" or,
an a." The definition of this expression in terms of our theosy of classes
in Chapter I
f This
is
II.
INTRODUCTION
78
[CHAP.
2 (<f>z)
first
also satisfies/.
requisites together
classes z(<f>z)
and
z tyz) should be identical when, and only when, their defining functions are
formally equivalent, i.e. that we should have
of
^(<f>z)^(^z)"
to be derived,
is
by means of a two
x !a = 6?!t.
whichis
by the general
x lz = 6lX"
= :(/):/!%!^.D./!6'!^ Df
definition of identity.
by our
definition,
(ax) : x =* X ! x
which, by eliminating 2 (yjrz), becomes
'
<l>
(ax) = </>=*
which
%'
 (a#)
fx
'
(3%. 0)
which, again,
* =*%WZ )>
=x
<f>x
(ax
= x X x fx = x
l
Thus our
'
4>
x =x x! x
'
down
laid
h
:.
=x
i.e.
= 0l,
it
for classes,
2 ((f>z)
* x x
!
is
z,
>
equivalent to
i.e.
= z (yjrz) =
.
<#
= x fx.
.
it will
by <f>z."
by means of our general
must be derived,
xef z
!
This definition
it
x x z = 6\
tyx.
f\z (<f>z)},
^rx
'
(<j>z) is
6\
axiom of reducibility,
<f>x
meaning
x : x!
01
equivalent to
is
tions,
equivalent to
is
member
is
gives
We
therefore put
Df.
is,
of/ \z (<f>z)},
(>&<f):<t>y.=y.1rly:ylrlx.
It thus appears that "xez(<j>z)" implies
is
equivalent to
<j>x;
also, in
is
<f>x,
since
it
implies
fix, and
a predicative function
<f>x
fix
implies
f formally equivalent to
<f>,
79
CLASSES
Ill]
and x must satisfy i^>, since x {ex hypothesi) satisfies <. Thus in virtue of the
axiom of reducibility we have
m
x ez (<f>z) = <f)X,
i.e. # is a member of the class z (<f>z) when, and only when, x satisfies the
which defines the class.
function
I
<j>
We
class of classes.
As we have
we may
.(<&g):fp.^ .g\pF{g\a)
any expression of the form z (yfrl z).
F{&(fa)}.
where
fi
We
shall
Df,
we put
Thus we
find
H
Df,
Df.
.gl/3:yefflZ.
fi
xe^Mz = .ty\x
7 eg\ a = g\ 7
Just as we put
so
= :(zg):f/3.=
we
:.
deduce
easily
(30)
/{*(*'
*)}
{*(*
5".
Thus
I
classes as arguments,
members
offers
*)},
H:to)://3.=0.<7!/3.
7 e a (fa) = .fy.
i.e.
which
to functions of
satisfy the
no
i.e.
every function
difficulty.
We
(a^)
i.e.
$ x *
x
it
'
'
e ty
^>
would mean
(g>/r)
e yjrl z
(i/r!
<f>x
=x
yfr
yfrl
I
x Df,
:
yfr
(yjr
2 ).
The use
explained shortly.
3,
INTRODUCTION
80
As
[CHAP.
as to the scope of
larger proposition.
(<j>z) if it
But
of reducibility, namely
an ambiguity
is
we always have
part Of a
the axiom
which takes the place of El(ix)(<f>x), it follows that the truthvalue of any
proposition in which "z{^>z) occurs is the same whatever scope we may give to
z (<f>z), provided the proposition is an extensional function of whatever functions
it may contain. Hence we may adopt the convention that the scope is to be
always the smallest proposition enclosed in dots or brackets in which z (<f>z)
occurs. If at
followed
by
is
required,
"
for [(?#)($#)]
Similarly when two class symbols occur, e.g. in a proposition of the form
/ {z (cf)z), 1z (tyz)}, we need not remember rules for the scopes of the two symbols,
since all choices give equivalent results, as it is easy to prove. For the preliminary propositions a rule is desirable, so we can decide that the class symbol
which occurs first in the order of writing is to have the larger scope.
is
ambiguous, in so
far as it is
z (^z), z (xz )> etc i* i s to stand for, where $z, fz, xz,
functions of the class. According to the choice
determining
are the various
of the symbols z
etc.
now be understood.
undecided as to which
The
((f>z),
result.
But
all
"r
Hence
unless
notion of a class
of a
is
:$*=.**. D
we wish
./{
(<f>z)}
=/{3 (**)}."
is really
we
is
really "f{$(4>z)\,"
1
"(af
where
<j>
is
Thus
we
"/(a),"
are led to
where a
is
is, it is
<f>
is
variable function.
be led into errors unless we confine ourselves to predicative determining functions. These errors especially arise in the transition to total variation (cf.
%
Accordingly
Df
BELATIONS
81
for
yiT
for
if
'
(*):(a*).*l*s.^!./ftr!3}.
class, (2)
fla=.(^).<f,lx= x ^lx.fl{^l^}
Thus a predicative function of a
Df.
We
We
a t) x4>{x,y): = x>y
(x, )} is
an extensional function of
is
<.
a relation
On
is
determined by
yfrl
(x,y)
:f^l(x,p)}
Df.
^''%H{x,y) = x^(x,y). =
i.e.
and we use
We put
its
x{+l($,y)}y.
z" we put
.fl(x,y) Df*.
is
its
own
sake,
oo{xg<j>{x,y)}y.
Ot)
le
4>(*>y)
this, in virtue
is
y,
"(at)
is
i^)<i>{^y)=x,y.^\(^y):^\{x,y),
and
equivalent to
<f>
(*,
y)
=x, y
to y)>"
<f>(x,y).
:z {$4>(x,y)} y
<f>(x,y).
R&w
[CHAP.
INTRODUCTION
82
by a single
of a relation
is
we may
not relevant,
capital letter.
above,
r
R S = xRy .= x y xSy,
\.R = &4>(x,y).= xRy = x>y
:.
\.R
and
Classes of relations,
.4>(x,y),
= $$(xRy),
and
Just as a class must not be capable of being or not being a member of itself,
so a relation must neither be nor not be referent or relatum with respect to
cannot
itself. This turns out to be equivalent to the assertion that 4> ! (x, fi)
again,
This
principle,
(as,
in
or
x
arguments
<
!
the
of
either
y).
y
significantly be
results from the limitation to the possible
We may
sum up
this
arguments
to a function explained
II.
The use
it
follows.
directly represented
(ix) (4>x)
(ix) (4>x)
= (ix) (yfrx)
(ix) (4>x) = (ix) (tyx)
(ix) (4>x)
The use
= (ix) (<j>x),
= (ix) (<f>x),
(ix) (fx) = (ix) ( Xx). D
.
(ix) (ifrx)
(or of a single
(ix) (4>x)
letter,
= (ix) (%).
such as
a,
to represent
b:(a).fa.D.f{x(4>x)},
x(4>x)
adjusted,
argument
by the theorems
\:(R).fR.D.f{x4>(x,y)},
y)^x^ (x, y) D
\.Zp4>(x,y) = x4>(x,y),
b:tif/4> (x,
\:x4>
b:$<f>
./{$$
4>
0>
y))
=ffi
y)
t 0> V))>
y),
0..x4>(x,y)
= xpx(>y)'
incomplete symbols
hi]
83
is possible.
symbols are obedient to the same formal rules of identity as symbols which
we only
resulting variable (or constant) values of propositional functions and not their
identity.
summed up
as follows.
is
relevant
i.e.
that
is
is
is
always true
does not
Thus
exist.
for
may
is
(<f)x)
may not be
(x) .fx
when E
(ix) (4>x).
(<f>x).
(x)
all
true.
example we have
inference
only valid
incompleteness
(<j>x) is
The
its
a function which
This
(<f>x),
= x,
.x
.f{ix)
(<f)x)
As soon as we know E
(ix)
(<f>x),
In regard to
different.
It
may
z(tf>z)
without having
For,
by a
= ty !
yfr !
(<f>z)
shall
(<f>z) = yfr
=x
somewhat
is
may have
z.
Thus we
incompleteness
we
find that
(f>x = x \fr
x.
have
h :<f>x= x
for
for
but the function "2 = the author of Waverley" has the property that George IV
wished to know whether its value with the argument "Scott" was true, whereas
*
Cf. p. 8.
62
INTRODUCTION
84
the function " 2
= Scott "
[CHAP.
Ill
Hence there
is
x = y.x = z."D.y = z,
which holds without any exception, and yet does not hold when for x we
substitute a class, and for y and z we substitute functions. This is only
possible because a class is an incomplete symbol, and therefore "(<.?)= ^S 2"
is not a value of " x = y"
It will be observed that
i/r
z.
"Olz ^lz"
of 2 (<f>z)
2(<f>z)
If
is
is
= y}rlz.z(<f>z) = x
Z'
z(<f>z),
the product
is
equivalent to
(a#)
and
this does
<f>x
but
is
apt to
> ^
=%
x %
'
2.
we
become
functions,
yjr I
=x e x e
imply
(<f>z)
is
an incomplete symbol
PART
MATHEMATICAL LOGIC
SUMMARY OF PART
In
this Part,
symbolic
logic, or
that
shall,
is
we
shall deal
deserve to belong to
it
We
functions, classes
tions,
namely negation,
the last two can be defined in terms of the first two. Throughout this first
section, although, as willbe shown at the beginning of Section B, our propositions, symbolically
unchanged,
will
distinction
the Introduction. Its importance and purpose, however, are purely philosophical,
and so long as only mathematical purposes are considered, it is unnecessary to
remember
which
is
Section
apparent variables
and
{i.e.
variables are
We
show
that,
where
we can define
such a way that their
concerned,
and implication in
We
show
also that
formal im~
'plication,
of
i.e.
reducibility,
variables.
An
the next topic considered in Section B. Finally, this section deals with
It is
i.e. phrases of the form "the soandso" (in the singular).
shown that the appearance of a grammatical subject "the soandso "is deceptive,
is
descriptions,
88
MATHEMATICAL LOGIC
[PART
C deals with classes, and with relations in so far as they are analogous
Classes and relations, like descriptions, are shown to be "incomplete
symbols" (cf. Introduction, Chapter III), and it is shown that a proposition
Section
to classes.
Section
for classes.
D deals
which are constantly needed throughout the rest of the work. Most of
the
properties of relations which have analogues in the theory of classes
are comparatively unimportant, while those that
greatest utility.
Section E,
finally,
by
members
of
some
class.
summands
with an
or factors.
and (3), in so far as they are analogous, are dealt with in Section
C.
have, for each of the three, the four analogous ideas of negation,
addition,
multiplication, and implication or inclusion. Of these, negation is
analogous
We
(2)
positions
classes,
fact
two relations
two
classes is their
common
I]
89
one
class in
in the fact that every pair of terms which has the one relation also has the
other relation. It is then shown that the properties of negation, addition,
multiplication and inclusion are exactly analogous for classes and relations,
are, with certain exceptions, analogous to the properties of negation, ad
and
dition, multiplication
and implication
"p
implies
for propositions.
q"
(The exceptions
arise
is itself
classes, is
corresponding to the division of propositions into true and false, but a threefold division, namely into (1) the universal class, which contains the whole of
a certain type, (2) the nullclass, which has no members, (3) all other classes,
which neither contain nothing nor contain everything of the appropriate type.
The resulting properties of classes, which are not analogous to properties of
analogues
SECTION A
THE THEORY OF DEDUCTION
The
is
explicit,
and to
effect the
it is
itself, i.e.
assumptions
its logical
If
it
deduction of
first
all
those that are required to make deduction possible. Symbolic logic is often
regarded as consisting of two coordinate parts, the theory of classes and the
theory of propositions. But from our point of view these two parts are not
coordinate
for in
But the
subject to be treated in
Now
it is
what
how one
proposition
may be
all
all
necessary,
common forms
and
it is
of inference.
it will
be shown
shown
them might
It will not be
number
of
be diminished. All that is affirmed concerning the premisses is (1) that they
are true, (2) that they are sufficient for the theory of deduction, (3) that we
do not know how to diminish their number. But with regard to (2), there
must always be some element of doubt, since it is hard to be sure that one
never uses some principle unconsciously. The habit of being rigidly guided
is a safeguard against unconscious assumptions; but
not always adequate.
is
*1.
PRIMITIVE IDEAS
AND PROPOSITIONS
Since all definitions of terms are effected by means of other terms, every
system of definitions which is not circular must start from a certain apparatus
of undefined terms. It is to some extent optional what ideas we take as
undefined in mathematics; the motives guiding our choice will be (1) to
make
the
number
systems in which the number is equal, to choose the one which seems the
simpler and easier. We know no way of proving that such and such a system
of undefined ideas contains as few as will give such and such results*.
Hence
we can
only say that, such and such ideas are undefined in such and such
a system, not that they are indefinable. Following Peano, we shall call the
is
meant; but
the.
explanations do not constitute definitions, because they really involve the ideas
they explain.
shall first
we
being preceded by a star; thus "fcl/Ol " will mean the definition or proposition
so numbered, and " #1 " will mean the chapter in which propositions have
numbers whose integral part is 1, i.e. the present chapter. Chapters will
generally be called " numbers."
Primitive Ideas.
Elementary propositions. By an "elementary " proposition we mean
one which does not involve any variables, or, in other language, one which
does not involve such words as " all," " some," " the " or equivalents for such
words. A proposition such as " this is red," where " this " is something given
in sensation, will be elementary. Any combination of given elementary
(1)
propositions
*
fundamentals.
92
MATHEMATICAL LOGIC
[PAST
be elementary.
letters p, q, r, s will
(2)
By an
we
function"
mean an
shall
"
elementary propositional
tuent,
i.e. a variable, or several such constituents, and such that, when the
undetermined constituent or constituents are determined, i.e. when values are
elementary proposition,
We
shall
(3)
notp "
show in *9 how
#5)
numbers (#1
"
is
an elementary propositional
function.
to propositions
Any proposition may be either asserted or merely conIf I say " Caesar died," I assert the proposition " Caesar died,"
Assertion.
sidered.
if
"
'
is
be de
signated by
"\.p."
The
sign "h" is called the assertionsign*; it may be read "it is true that"
(although philosophically this is not exactly what it means). The dots after
the assertionsign indicate its range 'that is to say, everything following is
;
asserted until
we
number
Thus "
The
truef."
is
of p or of
q,
p D
.
means "p
"
means
true
" it is
therefore
q
of these does not necessarily involve the truth either
.
is
Assertion
(4)
first
"
of a propositional function.
definite propositions,
we need what we
Besides
the assertion of
argument
This
"A
is
is
is
w; then
we may assert
when the law
</>?
A ."
*
Here
We
is left
is
as.
A may
have adopted both the idea and the symbol of assertion from Frege.
of identity
be deter
SECTION a]
primitive ideas
and propositions
we
<f>x,
93
leaving x undetermined,
This
is
only legitimate
" V
"(
Thus
:pvp D .p,"
.
triangle;
it is
All the assertions in the present work, with a very few exceptions, assert
propositional functions, not definite propositions.
one.
As a matter
The
which
i.e.
with
all
some
possible value.
name "individual"
with itself" or the proposition " there are individuals " will be a proposition
belonging to logic. But these propositions are not elementary.
(5)
Negation.
false," will
If
is
any
"p
is
proposition.
(6)
Disjunction.
"either
i.e.
is
If p
true or q
"
This
will
is
is
"
mean
"~
p
p or q is true," which is
(~ p v ~ q)" will mean "it is false that
is false
or q
is
true";
either
"
PVq
equivalent to
"p and q
either
p is
and so
on.
q is false," which
For the present, p and
false or
94
MATHEMATICAL LOGIC
[PART
The above are all the primitive ideas required in the theory of deduction.
Other primitive ideas will be introduced in Section B.
When
p, sp that, if
what
artificial;
adopted
The
is true,
very
is
q must
also
but although there are other legitimate meanings, the one here
much more convenient for Our purposes than any of its rivals.
we
But
anything, and
determine
what,
if so
is that, if
is
is
implies
require of implication
true."
is
this
this property
then
it
What
is false
is
is true,
to be defined to
*1'01.
then
mean
pDq. = ~ pvq
.
Here the
letters "
"What
is
"
is
it
does
true and
The most
if
either
p"
Df " stand for " definition." They and the sign of equality
together are to be regarded as forming one symbol, standing for " is defined
to mean*."
mean
Whatever comes
it.
Definition is not among the
primitive ideas, because definitions are concerned solely with the symbolism,
not with what is symbolised ; they are introduced for practical convenience,
Primitive Propositions.
The
denned
is
"Df "
will
later.
38.
PRIMITIVE IDEAS
SECTION A]
95
AND PROPOSITIONS
The above
principle is used
from a proposition.
present work are assertions of propositional functions, i.e. they contain an
undetermined variable. Since the assertion of a propositional function is a
different primitive idea from the assertion of a proposition, we require a
primitive proposition different from *11, though allied to it, to enable us to
"
deduce, the assertion of a propositional function " yfrx from the assertions of
" (f>x"
position is as follows
*1'11.
When
<f>x
Pp.
real variable.
This principle
is
also to
be assumed
and " h . <j>x D yfrx " then the " x " in <f>x is
not absolutely anything, but anything for which as argument the function "<j>x"
"
similarly in " <f>x D yfrx " the x is anything for which " <f>x D yfrx
is significant
is significant. Apart from some axiom, we do not know that the x's for which
" <px D yfrx" is significant are the same as those for which " <f>x " is significant.
of propositional functions " r
<f>x
"
The primitive
proposition #111,
can also be asserted, secures partial symbolic recognition, in the form most
useful in actual deductions, of an important principle which follows from the
theory of types, namely that, if there is any one argument a for which both
"
" and "
" are significant, then the range of arguments for which "<f>x
"
(fta
yfra
"yfrx"
is significant is
It is
nificant.
the same as the range of arguments for which " yfrx " is sigobvious that, if the propositional function " <f)X D yfrx " can be
must be arguments a for which " <f>a D yfra " is significant, and
for which, therefore, "<j>a" and "yfra" must be significant. Hence, by our
principle, the values of x for which " <f>x " is significant are the same as those
asserted, there
for
p.
which
15)
is
i.e.
called the
"axiom
it
yfrx.
<f>x
The primitive
(cf.
pro
of identification of type."
what
96
MATHEMATICAL LOGIC
The above
proposition #1*11
is
[PART
it is first
"hz.pDq.D.qDr .D.pDr."
We
:.
D r D p D q D .p D r.
#205 by means of *204, which
.
For
we
obtain, as
we
\
.qDr:D:q.D .pDr.
p by q D>, q by p D
an instance of #204, the proposition
replace
and r by p D
q,
#111 enables us
is
asserted by #205.
r,
(1),
*1'2.
:.p.D
in this proposition,
if,
is
h.pvp.D.p
Pp.
"
If either
is
and
true or
will
is true,
then
is
true."
title of "
#13.
by
.
their
D pvq
.
This principle
"today
is
numbers.
Pp.
states
"If q
Wednesday" and p
is
is
true,
"today
" If today is
is
true,
will
*14.
h:pvq.D.qvp
it false.
The
principle
Pp.
This principle states that "p or q" implies "q or p." It states the
permutative law for logical addition of propositions, and will be called the
" principle of permutation."
It will be referred to as " Perm."
#15.
"
If either
is true,
or
'
'
is
as "Assoc."
The
proposition
power, and
less
deductive
SECTION A]
*1'6.
\: K
AND PROPOSITIONS
PRIMITIVE IDEAS
q'2r.'5:pvq.'2.pvr
" If
97
Pp.
q implies
r,
'
and conclusion without impairing the truth of the implication. The principle
be called the "principle of summation," and will be referred to as "Sum."
will
*1'7.
#1*71.
If p is
If
position.
an elementary proposition,
p and
~p is an elementary proposition.
pvq
is
Pp.
an elementary pro
Pp.
*1*72.
function. Pp.
This axiom
<j>
Introduction becomes relevant, and any view of logic which justifies these
axioms justifies such subsequent reasoning as employs the theory of types.
This completes the list of primitive propositions required for the theory
of deduction as applied to elementary propositions.
R&W
*2.
Summary
The
o/#2.
In such
any instance of themHence when a general rule
in noticing that they are instances of the general rules given in #1.
cases,.these rules are not premisses, since they assert
selves,
is
when
<>jp is
Thus
written in place of p.
"
"
Taut
is
duction from general rules, but cannot itself be erected into a general rule,
since the application required is particular,
explicitly
Again,
when two
different sets of
and one of
is
round brackets.
The propositions in this number are all, or nearly all, actually needed in
deducing mathematics from our primitive propositions.. Although certain
abbreviating processes will be gradually introduced, proofs will be given very
fully, because the importance of the present subject lies, not in the propositions themselves, but (1) in the fact that they follow from the primitive
propositions, (2) in the fact that the subject is the easiest, simplest,
and most
employ the
and functions
all
in the entities
considered.
* Later on we shall cease to mark the distinction between a premiss and a rule
according to
which an inference is conducted. It is only in early proofs that this distinction is important.
SECTION A]
IMMEDIATE CONSEQUENCES
99
number
are the
following
*202. h.q.l.plq
q implies that p implies q, i.e. a true proposition is implied by any
This proposition is called the " principle of simplification " (re
I.e.
proposition.
b'.pZ^q.Z.qD^p
*203.
~p"5q 3. ~qOp
zpDq.D. ~gO ~p
*215. h:
*216.
~gO ^p.D.p'Dq.
*217. h:
They
negative.
of an equation
are thus analogous to the algebraical rule that the two sides
may be
\:.p.0.q^r:D:q.0.pDr
*204.
This
is
provided q
is true.
H:.gOr.D:jpDg.D.jOr
\:.p3q.D:qDr.3.p3r
*205.
*2;06.
These two propositions are the source of the syllogism in Barbara (as will
be shown later) and are therefore called the "principle of the syllogism"
(referred to as " Syll "). The first states that, if r follows from q, then if q
follows from p, r follows from p. The second states the same thing with the
premisses interchanged.
b.pDp
*208.
I.e.
and
itself.
This
It is not the
I.e.
The
h:
is
it (cf.
is
#1315).
~p.O .pDq
propositions in
forms.
"
called the " principle of identity
same
*2'21.
is
number
We now proceed to
subsumed under
more compendious
are mostly
results in
formal deductions.
72
[PART
MATHEMATICAL LOGIC
100
*201. \:pD
~p.D ~p
.
This proposition states that, if p implies its own falsehood, then p is false* It
is called the "principle of the reductio ad absurdum," and will be referred to as
Abs."* The proof is as follows (where "Bern" is short for " demonstration"):
'
Dem.
Taut ^
[(l).(*r01)] h
~> v ~jp
D ~p
(1)
p D ~p D
.
~.p
*202. i:g.3.;Og
Dem.
^
Add
L
:/0~</.
~p v a
(1)
D.qO~p
Dem.
Perm
~g D ~ov ~
~r> v
(1)
2 J
2>
#204.
V.q.D.pDq
[(1).(*101)]
#203.
i>
\:pD~q.D.qD~p
[(l).(*r01)]
h:.p.D.^Dr:D:^.D.pDr
Dem.
Assoc 21Pl2^3
? J
P.
h:.~i)v(~ovr).D.~ov(~vr)
h:.p.D.^Dr:D:g.D.^Dr
[(1).(*101)]
#205.
(1)
Dem.
[SumP
#206.
:.
D r D ~p v q D ~p v r
L
p J
[(l).(*r01)] h
:.
D r D :pD # 3 .p D
(1)
hz.^O^.DigOr.D.pDr
Dem.
Comm<L
:>r
'
pDq
p,
q,
[#2*05]
[(1).(2).*111]
In the
last
line
'
p:>r~\
r
br.qDr.D.pDq.D.pDr:.
D :.pDq'.D:qDr .D .pDr
F:.gDr.D:pD5.D.)Dr
h:.pD^.D:^Dr.D.joDr
(2)
#111" means
that
(1)
(2)
we
are
to
*rn.
*
There
is
an interesting
historical article
on
this principle
by
Vailati,
"A
proposito d'
un
IMMEDIATE CONSEQUENCES
SECTION A]
101
"
gism in Barbara
is
bzp.O.pvp
*207.
*l3
proved
is
*208.
h.jOp
is
written in place of
to be
q.
Dem.
*205
py^M
H ::p
vp 3 p 3
[(1).(2).*111]
h :.p.
[*207]
\
[(3).(4).*111]
b.~pvp
*211.
b.pv^p
p O.pvp: D p'Sp
:pvp. D .p
"S.pvp D p^p
.p. D .pvp
[Taut]
*21.
:.
(1)
(2)
(3)
(4)
.p^p
[*208
(*101)]
Dem.
>p,p
Perm
[(1).*21.*111]
This
*212.
is
~p vp D ,p v*>p
(1)
\.pv~p
h.jO<^(~p)
Dem.
r*2n^
L
(i)
.
[(1).(*101)]
*213.
h.pD~(~p)
Kjov~{~(~p)}
This proposition
is
a lemma
Dem.
Sum
'(~P)}1
~p D ~ (~(~p)} D
r
:.
[*212^]
I"
~jp D ~{no(f^p)}
(2)
[(1).(2).*1'11]
p vr^p D p v~ (~(~jp)}
(3)
(1)
[(3).*2n.*rii]
MATHEMATICAL LOGIC
102
[PABTI
K~(~p)3j?
*214.
Bern.
h
L
q
[(i).*2i3.*rii]
D. ~{^(<^)} yp
K~{~(~/>)} Vp
\.~(~p)Dp
[(2).(*101)]
*215.
:p v~{j(~p)\
(l)
(2)
:~pDq .2 .cvq'Sp
Dem.
^205^~ ~ g) 1
(
P>
*212 5]
(2)
Xl).(2).*lll]
:p"Dq D
I
p 3~(~o/)
(3)
L2 03^^1
L
j"
P>
cK
(4)
9.
~g~<~P>>P
2*05
~
~(~p) Dp D
..
~ g D ~(~p)
D ~g Dp
.
[(5).*214.*lll]
(6)
~
3 "^^)
^^g'^P^^gX^g
P D ~P 3 ~(~g)> ~g 3~(~p)
'
"!
h: .
g>
i>>
~pD~(~g). D ~gO~(~p) D
.
ojpDg'.
[(4).(7).*ril]
[(3).(8).*111]
#2*05
D ~/>D~(i~g') D r^pOq D
:
<7 .
<*jqDe**>(~p)
(8)
(9)
::~a
D~(~) D ~qDp
.
J
<^gOoo(<>^p)
D :~pDq. D
D :.~pDq. D
[(6).(10).*1'11] h :. ~p D g D ~? 3 ~(~p) D
.
on
H ~_p D g
p yp
~gOp
(10)
(11)
D ~g Dp
.
^e proof of #2'15. In
it will
be seen that
(3),
p^p
2,
results
called "
time
~^0<7.D.~gOp
.
(4), (6)
iV^ote
"
g>
[(9).(11).*111]
(7)
*
2>,
~p Z> ? D ~p D ~(~?) D
D ~g D ~(~^>)
^jp D
~p D q D ~? D~(~p)
:.
I :
:.
^*
(5)
p>
*2'05
r*205
(1)
it is
used
it will
[Syll]h.(a).(6).(o).Dh.(d),
where (a) is of the form p1 Dp2 (b) of the form p 2 Op3 (c) of the form p 3 Dp 4
and (d) of the form p x "Dp 4 The same abbreviation will be applied to a sorites
of any length.
,
IMMEDIATE CONSEQUENCES
SECTION A]
Also where
be proved,
it is
103
is
the proposition to
K^,"
[etc.]
where
be a reference to the previous propositions in virtue of which
the implication "pi^pa" holds. This form embodies the use of #1\L1 or #1*1,
and makes many proofs at once shorter and easier to follow. It is used in the
first
two
bip^q. D.~gO~p
*216.
Dem.
f.?D~(~?).D
[*2\L2]
[*205]
*203
b:p5q.3.pD~(~q)
'
:j?D~(~<7). D
(1)
,~qD~p
(2)
b.(l).(2).Db:pDq.D.~qD~p
[Syll]
The
Note.
when
write " b
etc.
is
Prop". Thus
we
shall
less
corresponds to "Q.E.D."
*217. b i^q'Dr^p
D .p"5q
Dem.
b:~(~q)3q:3
b:pD~(~q).O.pDq
[Syll]
r.(l).(2).3h.Prop
[*214]
#2*15, *2'16
be
all
*218.
transposition,
(l)
(2)
and
will
h:~pDp.D.p
Dem.
[*212]
[*205]
This
is
h.jO~(~p).D
r.~jpDp.D.~jO~(~p)
:~pD~(~p). D
[*=*]
I
[Syll]
I .
[*214]
b.~(~p)Dp
[Syll]
K(3).(4).DKProp
(1)
(2)
I :
~(~j)
~p Dp D ~(~p)
.
(1)
(2)
(3)
(4)
MATHEMATICAL LOGIC
104
states that a proposition
hood
[PART
its
own
false
is true.
hzp.D.pvq
*2'2.
Dem.
Add D
r .
*221.
p D
r :
q vp
[Syll]
(1)
...
[Perm]
q vp
(1)
D pvq
(2) D h Prop
.
(2)
h:~p.3.pl q ~*22^1
p D
*2*24.
*2'25.
\~
z.pzv
~p D q [*2*21 Comm]
zpvq.D .q
.
Dem.
*21 D
[Assoc] D
f .
1 :
1
~(p v q) v
.
:^>
{~(p v q) v
*226.
H:.~p:v:^D ? .D.g
*225
*227.
H.p.D:pOg.D.?
[*2'26]
*2'3.
(p v q)
q)
Prop
v (o v r) D p v (r v o)
.
Dem.
Perml^
H:<7vr.D.rv<7:
Sum
'
*2'31.
p v (g v r) D . (p v
.
q)
I :
p v (q v r) D p v (r v q)
.
vr
This proposition and *2*32 together constitute the associative law for
aD6,
6 3c,
cOd,
[(1).(2).*111]
[(3).(4).*1'11]
[Syll]
[(5).(6).*111]
[(7).(8).*111]
is
often
is
as follows
:. a D b
D bDc D
h:a.D.&
h bDc. D .aDc
(
have a series of
and "aDd" is the
Dc
(1)
(2)
(3)
H&.D.c
(4)
bza.O.c
:.aDc.D:cDrf.D.a!>^
:cDd. D aDd
(6)
:c.O .d
:a .D .d
but
When we
all asserted,
(5)
(?)
(8)
It
is
105
IMMEDIATE CONSEQUENCES
SECTION A]
full.;
we
h:a.DJ.
where
"aDd"
[etc.]
[etc.]
D.dzOh. Prop,
is
We
by
implications hold.
not " a
D b,"
We
put one dot (not two) after " 6," to show that
that implies
c.
is
concerned.
If
it is b,
"a D d "
is
now
we put
za.D .b
[etc.] D .c
[etc.] D d
(1).
and then
means
"
D d." The
proof of *2*31
is
as follows
Dem.
[*2'3] h
Assoc
p v (q v r) D p v (r v q)
rJ
9>
i_
fperm^^"
L
*2'32.
I
P>
r v (p v q)
p v q) v r D
:
Prop
I .
:(pv^)vr.D.pv(q'vr)
Dem.
Perm
Assoc
I
p v <?) v r D
.
'
D .p v (g v r) D
:
I .
Prop
pvqvr. = (pvq)vr Df
.
#236.
r v (p v 9)
^' ^
p, q, r
[*23]
*2'33.
9.
\
for
q .D .rvp
z.qDr .D zpw
Dem.
Vzpvr.D.rvpz
[Perm]
\sy\\
pVq,pVr
[Sum]
'
r*P
~\Dh:.pvq.D.pvr:D:pvq.3.rvp
hz.qOr.
"2
h.(l).(2).Syll.Dh.Prop
#237.
*2'38.
Perm Sum]
.
hz.qOr.Dzqvp.O.rvp
[Syll
Perm Sum]
.
(1)
(2)
106
"
MATHEMATICAL LOGIC
[PART
The proofs of #2*37 38 are exactly analogous to that of #2*36. (We use
*2'37'38 " as an abbreviation for " *2*37 and *2"38." Such abbreviations will
be used throughout.)
The use
The
the inference. If
we
treated
it
which the
made, but
is
as a premiss,
is
not
itself
we should need
a premiss in
either
it
or
some
other general rule to enable us to infer the desired conclusion, and thus
we
inference
The
is
drawn.
premisses, that is to say, in the case of " Syll " for example, the proposition
.
we have
square brackets. It
is,
manner
tinguish in the
indicated
it
We
of the reference.
both
"h.(l).(2).Syll.Db.Prop"
" [Syll] h (1) (2) D h Prop "
rather than
*2*4.
.pyq D pvq
:.p. v
'
Dem.
h
.#231
h \.p. v\.p v q
D :pvp v
*2'41.
D:pvq:.Dh. Prop
[Taut.*238]
\:.q.v.pvq:'D.pvq
Dem.
Assoc "tliA
p,
q,
b:.q.v.pvq:0:p.v.qvq:
D p v.q
[Taut.Sum]
~p .w .pDq:D pOg
#24
*2'42.
f
*243.
b:.p.D.pDq:0.pDq
[*2"42]
*2"45.
h:~(pvq).D. ~p
[*2'2
*2'46.
:.
~ (p v q)
~q
[*1 3
Transp]
Transp]
:.
Prop
i.e.
we
IMMEDIATE CONSEQUENCES
SECTION A]
107
#247.
\:~(pvq).0.<s>pvq
*245.*2'2^.Sylf
*2*48.
f:~(pvg).D.pv~#
*2'46 . *l3
*2*49.
H:~(pv<7).0.~pv<>^
*2'45
*22
Syll]
<~
<**>p,
Syll]
p>
:^(pDgr). D.yp>gr
247
^'
*2'5.
I
*2'51.
r:%(jOg).D.0^g'
*248
P
P
*2'52.
>:~(^02).D.~pO~2
*2'49
^
*2 521. h:~(jOg)'.D.gOj>'
<1
PA
[*2'5217]
hzpv^.D.^^Og'
*2'53.
h .*21238
I
:pv^ D ~(~jp) vg D
.
*254.
H^jO^.D.jovg
*255.
h:.~j.
.g
[#2*53
*256.
h:.~g D zjjvg.D.jp
#2*55
*26.
Drjjvg.D
~p D g D p D q D
:
.Prop
[*21438]
:.
Comm]
^:
Perm
Bern.
h:.~pDg'.D:~pvg'.D.g'V(/
[*2'38]
(1)
3 qvg: D
D
v
h
D
D
D
:.~p
D
:~p
g
(1) (2)
g
q :. D H Prop
[*26 Comm]
.p"Dq.D: ~p D ^ D .q
[*2536 Syll]
pvq D:p^q.O .q
[*2'62 Comm]
.pOq.D :pvq. D .q
Syll
:~p v #
*2'61.
*262.
*2621.
1
*263.
*264.
.^}
*2*65.
\:.pDq.D:pD~q.D.r*>p [*264^]
*2 67.
h:.)Vf.D.jf':D.)D9
. <?
(2)
p v ^ D ~jp v ^ D ^
v q.
D :pv~^. D
[*262]
.^)
T*263 &?
P>2
L
Perm
Bern.

[*2 54.Syll]
[*2'24.Syll]
1 .
(1)
(2)
\:.~pDq.D.q:D.pDq
Syll
Prop
(1)
(2)
[PART
MATHEMATICAL LOGIC
108
\r:.p2q.2.q:D.pvq
*2'68.
Bern.
^2] hz.pOq.O.q.^.^pOq
1*267
h
(1)
*2*54
Prop
Perm *2'62
*269.
\:.pOq.^.q:0'.qDp.^.p
[*268
*2'73.
\:.p3q.0:pvqvr.3.qvr
[*2621'38]
*274.
H.^Dp.D^vgvr.D.^vr
*2*75.
h::pvg.D:._p.v.?Dr:D.)vr
*274^2
*2'74
*276.
h.p.v.prOspv^.D.pvr
*277.
hz.p.O.qlr.lipDq.O.p'Dr
*2'8.
r
[*2'75
Syll
*2'53'31
*276
J
*2 53 Perm . D

f
:.
qvr
[*2'38]
I
Comm]
2>
*2'81.
(1)
::
~r D g
D ~r vs.D.gvsr.DH. Prop
:
q.D.rOszDz.pvq.Dzpvr.'S.pvs
Bern.
> *276
.
(1)
(2)
K(l).(2).Dh.Prop
*2'82.
H.pvgvr .D:^v~rv$.D.jjvgvs
*2'8
*281
o,
*2'83.
H
:: jp
. <j
Dr D
:
:.
jp
rD*
D: ^> D
.
r,
gO
L2 82 jm^l
.
[Add.Syll]
H
*2'55
::
~jp
(2)
Comm D
.
I
:f:.j)Dg.>..)Dr:3
(2)
[#2'54]
*2'86.
(1)
D :.^> v r D r :.
Or.^vg. D.pvrz D :^vg .D. r:.
Oi.pyq.D.pvr'.OzqDr
:.
_pvg. D .pvr D :~p. D .gOr
[(1).*2'83]
.
? J
K:.pvg.D.r:D.gDr
3H
[Syll]
p>
D :p. v.gOr
jp
3 5Dr
.
*2'85
^
*3.
Summary
The
two propositions p and q is practically the proBut this as it stands would have to be a
logical product of
"p and q
position
new
q/"#3.
We therefore
primitive idea.
~(~^v~j),
i.e.
obviously true
*301. p.q.
is false
that either
is false
or q
is false,"
which
is
Thus we put
= .~(~^v~g) Df
is
*302.
"it
q.
pDqDr. = .pDq.qDr Df
When we
" H
fyx"
we
have
shall
yjrx "
<f>x
whenever
<f>
and
\fr
the same type. This will be proved for any functions in #9 for the present,
we are confined to elementary propositional functions of elementary pro;
positions.
In this
By *1'7,
fore,
by
by
~(j>p
case,
the result
is
proved as follows
and ~yjrp are elementary propositional functions, and thereyjrp is an elementary propositional function.
Hence
~<)pv~
#1*72,
#211,
h:~^pv~i^p.v.< v(~<^pv~^fp).
,
by
:.
<f>p
yjrp
D ~(~^,jo v o>yJfp),
.
*3'01,
I
Hence by
#1'11,
This proposition
to functions of
The above
is
:.
is
irjp
.D
,<j)p
typ.
when we have "h. <\>p" and "V .typ" we have "V .fyp.typ."
#3*03.
It is to
two or more
of real variables
<f>p
be understood,
like *1'72, as
applying "also
variables.
*1'72).
means
If <f)X contains, in any way, a constituent %(#, y, z, ...) and yjrx contains,
any way, a constituent %(#, u,v, ...), then both <f>x and yjrx take arguments
of the type of the argument x in ^ (x, y,z, .. .), and therefore both <}>x and yjrx
take arguments of the same type. Hence, in such a case, if both <j>x and yjrx
can be asserted, so can <f>x yfrx.
in
MATHEMATICAL LOGIC
110
As an example
[PART
We
there prove
and
and what we wish
(1)
(2)
to prove is
p'Dr.qOs.Ozp.q.O.r.s,
which
#3*47.
is
everywhere
(as
Now
in
and
in (1)
(2), p, q, r, s are
Section A);
hence
by
elementary propositions
applied
#l'7"7l,
repeatedly,
whence the
The
I.e.
so
is
ele
::pDr.q'Ds.D:p.q.D.q.r:.pDr.qDs.D:q.r.D.r.s,
m
\
*3'2.
result follows
by #3'43 and
*3*33.
bz.p.D.q.D.p.q
"p implies that q implies p
q,"
i.e.
number
if
is
true,
*326.
V.p.q.l.p
*3'27.
bip.q.D.q
the logical product of two propositions
I.e. if
propositions severally
\:.p.q.D
*3'3.
is true,
is true.
.r:D:p.D.q^r
"
"
Exp."
\:.p.D.qDr:D:p.q.'D.r
This is the correlative of the above, and will be called (following Peano)
importation " (referred to as " Imp ").
*335.
b.p.p^q.D.q
I.e. "ifp is true, and q follows from it, then q is true." This will be called
the "principle of assertion" (referred to as "Ass"). It differs from #11 by
the fact that it does not apply only when p really is true, but requires merely
I
is true.
z.pDq.pDr D
.
:p
D .q.r
will
*3*45.
I.e.
This
is
z.pDq.^zp .r .D .q.r
as " Fact."
by Peano the
may be
" principle of
multiplied by a
common
factor.
SECTION A]
347. Hr.pDr.gOff.Drp.g.D.r.f
I.e. if p implies q and r implies s, then p and
The law
jointly.
of contradiction, " h
its
~ (p
~p),"
111
is
.<f>p
.
<f>p
and
"
.
yfrp.
Bern.
V
*l'772
*211
~ (~
#2*32
I .
(2)
(*3'01)
.Dh:.^).D:^).D.#.^)
(3)
*1\L1
(*1'01)
312.
3*13.
314.
<>jp.
"*>q
:. <j>p
[Id
q)
<?
hz.p.l.q.O.p.q
321. h:.gr.D:^.D.p. ?
322. h'.p.q.D.q.p
is
yfrp
v ~ yjrp)
<f,p
(1)
(2)
(3)
(3 01)]
,
211
311
v .p . g
Transp]
[#3l Transp]
.
2.
This
Id . (*301)]
"
~*> v ~g
D ~p v ~
~p v~q.D.~(p.q)
<^ (^>
Prop
3*11.
(1)
hip.q.3.~(~pv~q)
h :~(~pv~^). D pq
*31.
I .
[*3'12]
[*3'2
Comm]
Ls13
L
~ (q
p)
Note
that, in the
(1)
~ q v ~p
Transp D
.
above proof,
w
"
Prop
(1)
~(3.p).3.(p.j),
3'24.
h.~(p.~p)
Dem.
211
r^p'
P
314 
The above
is
2.~(p.q)
[#314]
.
D.~pv~g.
[Perm]
?J
i>>
J
I
<
'
( JO
<^p)
the proposition
(1)
MATHEMATICAL LOGIC
112
*3*26.
[PART
bip.q.l.p
Bern.
*202
:p.D *q"2p
r:
[(1).(*101)J
~qvp;
pv~q
*231]
*253
q>p
(1)
v .p:
;<^>(~pv<^>q).
zp.q.O .p
D ,p
(2)
p>
[(2).(*301)]
*327.
bzp.q.l.q
Dem.
[*322]
zp.q.'D.q.p.
\
D.grDKProp
#326 2l21L
SJ
#33. h :.p .q ,D
.r:D :p .D .qOr
Dem.
\:.p.q.O
[Id.(*301)]
~ p v ~ q)
(
[Transp]
D ~r D .p D^g
D:j?.D.~rD~g':
D p
q D r :. D h Prop
[Id.(*l01)]
[Comm]
[Transp.Syll]
*331.
hz.p.D.qDr.'D.p.q.D,
Dem.
Id.(*101)]
r.jp.D.^Dr
:D:~^.v.~^vr
D :~j3V~^. v.r
#2*31]
pv ~q
*253
P ~(~j9 v oj<2) D
:
2>>
D:j).g.D.r:.DK
[Id.(*3'01)]
*3 33.
hzpDq.qDr.D .p Dr
[Syll
Imp]
*334.
\:qDr.pDq.D.pDr
[Syll
Imp]
Prop
hzp.pDq.l.q
#337.
h :.p. q.
[*2"27
:p,<*>>r
Imp]
.
D .~q
Dem.
h
Transp
.OhiqDr.D.^rD^q:
[Syll]
(2)
(1)
r
(1)
(2)
D h :.p. D\ qDr D p D ~r D ~^
D z.p.q. D .r D :^>. D # Dr
DF:.^).D.'>JrD~g :^:p.~r.D.~g
(3) Syll D K Prop
:
h.Exp.
h .Imp
(3)
SECTION A]
This
*3*4.
*3'41.
is
D pDq
p
bz.pDr.Dzp.q.O.r
H
113
[*326.Syll]
\z.q3r.Dzp.q.D.r [*327.Syll]
*343. tz.pDq.pDr.Dzp.O.q.r
*342.
Dem.
K*3 2.DH:.?.D:r.D.g.r
>
(1)
\.(l).Syll.D\zzpDq.Dz.p.Dzr.D.q.rz.
Oz.pDr.Dzp.D.q.r
[*277]
f .
*344.
(2)
is
z.qDp.rDp.Dzqvr .D .p
This principle
#3*44
(2)
analogous to #343.
is
Dem.
~q3r
Dp 3 ~ q 3 p
DzqDp.D.p
~
Exp
::
3
3
r
3
r 3p 3
(1)
3p 3 p :.
q
[Comm.Imp]
3 :. # Dp r 3 p 3 p
(2) Comm 3 h
q Dp r 3 p 3 ~ q 3 r 3 .p
h
Syll
:.
[*2*6]
I .
I .
f
:.
: <j
[*2'53.Syll]
#3*45.
I
:.
(2)
:.
Prop
may
by a common
We
(1)
factor;
hence
it is called
factor."
Dem.
I
.Syll
z.pDq.
[Transp]
[Id.(*l01.*301)]
*3"47.
3:gO~r.3.p3~r:
:
:.
Prop
its
analogue for
calls it
classes,
"praeclarum theorema*."
Dem.
K*326.3h:.p3r.3s.3:p3r:
[Fact]
Dzp.q.D.r.qz
[*322]
Dzp.q.D.q.r
(1)
R&w
:
:
[PABT
MATHEMATICAL LOGIC
114
D:g.r.O.*.r:
[*3'22]
D.q.r.O.r.s
K(1).(2).*303.*283.D
:.p D r ? D s D p q ..? r
*348.
I
(2)
:.
D H Prop
.
h:.^Dr.grDs.D:j>v^.D.rvs
This theorem
is
Bern.
V
#326
f
:.p
Dr
D s, D p 3 r
:
Dipvg.D.rv^:
Ozpvq.O.qvr
[Sum]
[Perm]
I .
*327
(1)
q D D q Ds
Dztjvr.D.svr:
D :.> Dr
I
[Sum]
Dr^vr.D.rvs
[Perm]
(2)
H.(1).(2).*283.D
I
:.
p D r # D s D :p v <f D r v s :. D
.
Prop
*4.
Summary
of #4.
is
as follows
The
entities considered
"
numbers which are all either or 1
p D q" is to have the value
~p is to be 1 if p
otherwise it is to have the value 1
if p is 1 and q is
are
both
and
and is to be in
if
be
1
to
1,
is
if
is 0, and
q
p
p is 1 p q
is to be 1 in any
both
and
are
and
if
0,
any other case p v q is to be
q
p
that
what
follows has the
mean
to
is
other case; and the assertionsign
are to be
1.
For
this reason,
we
with what
When
equivalent,
*401.
each of two propositions implies the other, we say that the two are
which we write
"
p=
q."
We put
p=q. = .pDq.q0p Df
It is obvious that
call
the truth
Thus two
false.
p = q,
case,
[PART
MATHEMATICAL LOGIC
116
The reason
The
*4"1.
#4*11.
I
zp'Dq. =
number
~gD~p
:p = q. = .<>jp =
~q
h.^ = ~(~#)
#4:13.
This
is
the falsehood of
i.e.
a proposition
is
equivalent to
negation.
its
\.p=p
*421. h :p = q.=.q=p
#422. H :p~q.q = r D .p = r
#42.
is
reflexive,
symmetrical and
transitive.
:p.= .p .p
#424.
#425.
h:p. = .pvp
I.e.
is
p" and
to " p or
p"
the law of tautology, and are the source of the principal differences between
the algebra of symbolic logic and ordinary algebra.
#4'3.
h:p.q.=.q.p
This
is
h:pvq. = .qvp
#431.
This
The
is
#4 32.
#433.
h :(p
The
#44.
sum
(p
q)
vq) vr
=
.
(q
of propositions.
r)
.p v(gvr)
two forms
\:.p.qvr. = :p.q.v.p.r
distributive law in the
#441.
in ordinary algebra.
I.e.
position
implies q when, and only when, p is equivalent to p . q. This proused constantly; it enables us to replace any implication by an
is
equivalence.
#4'73.
I.e.
:.
D :p = .p q
may be dropped from
.
a true factor
SECTION A]
117
p = q* = .pOq.q Op Df
*402. p = q = r.=z.p = q.q = r Df
#401.
41.
:^>Dg. =
ojqD~p
[21617]
411.
:p = q. = .f*p = ~q
412. h zp = ~q = ,q = r*>p
[2161 7. 34722]
\
[20315]
p = <v* (~jp)
D.r = :_p.~r< >. ~g
z.p.g.D.^r: = :g.r 3. <vj)
413.
\r .
414.
415. h
i.p.q.
[21214]
[*337. *413]
[#322
41314]
4*2.
H.^)=^
[Id.*32]
4*21.
h:p = ?. = .?=;>
:^ = gf.g = r.D.p = r
[*322]
4*22.
Dem.
h.*3*26.
~2t:p
= q.q = r.'D.p = q.
ipiq
[326]
D.gDr
#2*83 .3\:p = q.q = r.3.pDr
[326]
h
(1)
*3'27
(2)
Db:p=q.q = r.0.q = r.
K*326.
.q = r.0.p = q.
D.rOq
[*3'27]
3bip = q
3qOp
[327]
= q.q = r.0.rOp
(4)
K(3).(6).Comp.Dh.Prop
b
Note.
is
propositions
is reflexive
and
(1)
K*3*27.
(3)
(4)
(5)
(6)
transitive,
(2)
The
properties of being
symmetrical, transitive, and (at least within a certain field) reflexive are
essential to
*4'24.
any
\:p.
relation
which
is
to
= .p.p
Bern.
\.*B26.Db:p.p.D.p
h.*32. DH.jp. D zp.D.p.p:.
Db.p.O.p.p
[243]
I .
*4 25. h:p.z=.pvp
(1)
(2)
*32
(1)
(2)
K Prop
Taut Add ^
.
*4'24'25 are two forms of the law of tautology, which is what chiefly
distinguishes the algebra of symbolic logic from ordinary algebra.
Note.
MATHEMATICAL LOGIC
118
*43.
= .q.p
\:p.q.
[*3'22]
Whenever we
Note.
[PART
p and
may
have,
<f>(p,q).y.<f>(q,p),
we have
also
<f>(p,q).= .<f>(q,p).
For
{* (p,q).D.<f> (q,p)}
: <j>
(q,p)
(p,
tf>
<?).
*431. H
*4'32.
%* D
Dem.
K#4 15.
D\:.p.q.D.~r: = :q.r.'2.~p:
=:2>.D.~(gr.r)
[*412]
h
(1)
*4'11
[(*r01.*301)]
h
h
~ (p
~ r)
q.
tacit, as
The
above.
j)
(
(1)
~ (^
r)}
Prop
jVofe.
is
D
D
principle
is
~r
:jp
The use
(<?
which
often be
r),"
of #4*22 will
implication in #231.
*433.
\:(pvq)yr.
The above
= .pv(qvr)
[*231'32]
we introduce the
avoid brackets,
and addition.
To
following definition
p.q.r = .(p.q).r Df
[Faet.*347]
h:.p = q.D:p r.=.q.r
[Sum.*347]
*437. \:.p = q.D:pvr. = .qvr
=
[*3'47
*432 *322]
.r.s
\:.p=r.q=s.D:p.q.
*438.
*439. V'..p = r.q=s.^:p\q. = .rys [*3*4847 *432 *3'22]
*4*34.
*4'36.
*4*4.
This
z.p.qvr = :p.q v .p .r
.
is
the
first
Bern.
K*3'2.
0\::p.D:q.'2.p.qz.p.'2:r D,P
[Comp]
[*3'44]
0\::p.D:.q.y.p.q:r.D.p,
Dz.qvr. D.p.q.v.p.
0\:.p.qvr.D:p.q. v,p.r
D\:.p.q.D.p:p.r.D.p:.
Dhz.p.q. v .p.r.D.p
h.*3'27.
D\:.p.q.y.q:p.r.'D.r:.
[*3'48]
f.(l).Imp.
K#3"26.
[*348]
h
(3)
(4)
r.(2).(5).
yli.p.q.v .p.r'.y.qvr
Comp Df :.p.q. v .p.rzD.p. qvr
.
DKProp
r
r
::.
:,
(1)
(2)
(3)
(4)
(5)
is
119
.q.r:= .pvq.pvr
*4'41. \:.p.v
is
SECTION A]
This
By
law a
r)."
Dem.
.
*3'26
I .
*3'27
Sum
Sum
D h :.p. v. q.r D .p vq
D up v q r D .p v r
r
(1)
(2)
K(l).(2).Conip.Dh:.jp.v.g.r:D._pvgr.jpvr
DH:._pvg'.^vr.3:~)Dg'.~i)3r:
h . *2*53 *3*47
D:~p.D.q.r:
[Comp]
(3)
D:p.v.g.r
[*2'54]
(4)
DKProp
:.^>. = zjj.g. v.^.^^
K(3).(4).
*4*42.
f
Ztewi.
K*3'21
...
[*211]
h.*326.
H
(1 )
(2 )
Dh:.5V~g.D:^.3.j5.^v~g:.
Orrp.D.^.grv^
I : .
h:.p.
(2)
=:p.q.v .p.~q:.Dh
[*4*4]
*4'43.
(1)
Dh:j).grv~g. j.>
.O
p = p qv~q
.
Prop
= :pvq.pv~q
Dem.
Dhip.D .pvqzp.D.pv^q:
Dbip.'Z.pvq.pv^q
t.*2*2.
[Comp]
b
*265
I
~p D g D ~p 3 ~ o D .p
:.
:.
Dh :.~j)Dgf.~^D~gr .D.p:.
Dhz.pv^.pv^^.D.^
[Imp]
[*253.*3*47]
(2)
Oh. Prop
K(l).(2).
#444. \:.p.
(1)
= :p.v.p.q
Dem.
h
h
*445.
I
The
p = p
.
*2'2
Id . *3'26
^Vi.p.^.p.y.p.q
D V :.p "Dp :p'. q O .p
[*3'44]
Dbt.p.v.p.qiD.p
K(l).(2).
DKProp
J?
[*3'26
(1)
:.
(2)
*2*2]
De Morgan, or rather, are the propoDe Morgan for classes. The first
of them,
product.
it
will
logical
::
120
MATHEMATICAL LOGIC
*4'5.
= .~(^v<v ?)
= .<^>j)V~g'
*451.
*452. >
*453.
f
*454.
r.
*455
h:
*456.
I:
p
*^
(p
<^
<^J>
= p v v q
).
[*4'5412]
.<vj. =
.~(pvq)
= .pvq
*457. h :^>(<^>jp.no).
The
f*45^.*413]
(*301)]
[*4*512]
045212]
q)
~p. q. = ^ (p v m q)
o>(cvp
[*45612]
how
and vice
denials of products,
They
sums or into
be observed that the first of them
versa.
It will
*4'6.
p3q.= r^pwq
~(pDq).=.p.~q
1
*4'62.
*463.
r:
~(pD~q) = .p.q
*464. h:
r^pDq = .pv q
r*>
p
(<~J3 D q) =
*4/65.
o>p^e>jq.
*466. h:
*47.
. <>>>
.pvt**>q
(~p 3 ~ q) = ~p
.pDq.= :p .0 .p.q
<**>
H t]
[*2'53'54]
(*101)]
[462115]
(**>
[4'6'll52]
p D~ q =. ~)v<v q
h:
*467. h:
[*42
*461. h:
[*4641156]
[464
^2]
[*466ll'54]
Bern.
h.*321.Sy\l.D\:.p.D.p.q:0.p3q
KComp.
O^z.pOp.pOq.Dzp.'D.p.q:,.
[Exp]
D\::pDp.D:.pOq.D:p.D.p.q::
[Id]
Dh i.pDq.'Dzp.D .p.q
DKProp
h.(l).(2).
*471.
>
pDq.
(1)
(2)
:p.=.p.q
Dew.
K*321.
Dhi.p.q.D.p.D
Dh:.p.D.jp. ? :D p. = .p.q
3h:.p. = .p.q:3 p.Z.p.q
[*3'26]
K*3'26..
K(l).(2).
K (3)
*4722
r*4.5^.* 4 .i3"
= .~(~pvq)
<**>q
[*42
[PART
3r:.f.D..p.g: P = 2>3
D h Prop
.
(1)
(2)
(3)
::
SECTION A]
The above
proposition
is
It enables us to transform
constantly used.
121
is
an advantage
if
symbolic logic
and
it is
is
we wish to
But when
Dem.
\ .
#41 D b
.
:.
~ q D ~p
pDq =:
.
=:~o. = .~g.~p:
#4*71 "^'"^l
P>
<l J
[*4'12]
[*4'57]
[#4*31]
is
it
may be
omitted from a product without altering its truth or falsehood, just as a true
hypothesis may be omitted from an implication.
#474. \:.~p.Dzq.
= .pvq
[#221
#472]
#441 ^.(#101)1
#477.
#478.
*2'2]
Dem.
\
.#4*2 .(#101).
[#4*33]
[#4'31*37]
[#4*33]
[*4*25'37]
=:p.D.gvr:.DI.Prop
[*42.(*101)]
Dem.
[#2 15]
[*4'2.(#301)]
Note.
The
E:~j).0.~5v~r:
[#4*78]
= z(<*^qv r) .D .p:
=:q.r.D.p:.D\.?rov
#47879
are
false.
Take,
is
e.g.
#478,
contained in q
[PART 1
MATHEMATICAL LOGIC
122
[*2'01 Simp]
l:pD~p. = .~>
[*2'18.Simp]
#481. \:^pDp. = .p
*4 82. h r p Dg . p 3 ~ q = ~p [*2'65 Imp *2'21 Comp]
[#2'61 Imp Simp . Comp]
*483L \:pOq.r^p~yq. = .q
Note. *4'82*83 may also be obtained from *443, of Tyhich
*4r*8.
ally other
*485.
*487.
forma
*4'84 \'..p=q.^:pyr.
*486.
I"
:,
[Exp
Comm
Imp]
principle.
and im
MISCELLANEOUS PROPOSITIONS
*5.
Summary
of #5.
The present number consists chiefly of propositions of two sorts: (1) those
which will be required as lemmas in one or more subsequent proofs, (2) those
which are on their own account illustrative, or would be important in other
developments than those that we wish to make. A few of the propositions of
this number, however, will be used very frequently. These are
*5*1*
I.e.
I
D p=q
if
that two propositions are equivalent if .they are both false is #5 '21.)
*5*32.
I
:.p
= r := p.q. = .p.r
of
p and
#56.
I.e.
r.
This
is
hz.p.f^q.D
"p and
Among
.r
= :p. >.#vr
is
equivalent to
"p
implies q or r."
that,
given
any two propositions p, q, either p or >p must imply q, and p must imply
either q or not*?, and either p implies g or q implies p; and given any third
proposition r, either p implies q or q implies r*.
either to
or to
The proofs
respectively,
q.
and we
shall therefore
p. q, D p = q
#6'1.
#511.
zpOq.v .r^pDq
#512. \zpDq.v
#513.
#514.
.pD~q
h.p^q.v.qDp
H p Dq v qDr
:
* Cf. Schroder,
JW422]
[*25'54]
[*251*54]
[*2'521]
Band
270
MATHEMATICAL LOGIC
124
#5*15.
[PART
h:p = q.v.p = ~q
Bern.
V
*4'61
DH
~ (p D q)
D .p
~q
D.p = ~q:
[*51]
~5
Dh:jpDgr.v.p =
[*254]
~ (y D j>)
(1)
~p
[#51]
D
D
[#412]
0.p=~r>jq;
#461 D
= ~p
Dhs^Dp.v.jpi^g'
[*2 54]
(2)
h.(I).(2).#4'41.DKProp
#516.
~ (p = q
p = ~ q)
Bern.
.*326.
D.~p
[*4'82]
I .
*3'27
[Syll]
(1)
(2)
Comp D
.
\
(2)
p = q p D ~ q 3 ~p
.
Exp D
:.
[Id.(*101)]
[*4'51 .(#401)]
~<?
D.~(~ 2 :>p)
f*465^
K (3)
D.~?
[Abs]
I .
(1)
Op /> D ~ q
D. 3 D~0.
D h :p = q .p D ~ q D
p = q D p D ~? D ~(~2 Dp)
3:~(/0~g).v.~(~2:>2>):
D ~ {p = ~ q) D h Prop
.
: .
g)
(3)
= ,p = ~q
Dem.
Ohipvq = .~qOp
}\ ;~(p .q). = .pO~q
K*4'64'21.
h
#463 Transp
(1)
(2)
*43821
~ (p = ~ q)
h:p = q. =
#519.
K~(p = ~p)
*5'21.
#522.
[*51 #411]
\:~p.~q.0.p = q
h:.~(p = q). = :p.~q.v.q.~p [#4'6151'39]
#523.
\:.p
#5'24.
#525.
(2)
D H Prop
#518.
(1)
'
1,5 "' g
*518 ^.*4.
= q. = :p.q.v.~p.~q
m
[*262'68]
SECTION A]
MISCELLANEOUS PROPOSITIONS
From *5*25
disjunction,
"pDq.D
it
as
125
"pvq"
have defined
as
meaning
q."
= :p.q.'2.p.r
*53.
\:.p.q.0.r:
*5'31.
b:.r.p^q'.0:p.D.q.r
[Simp.Comp]
b:.p.3.q = r: = :p.q. = .p.r [*4'76 *3331 *53]
*532.
This proposition
p qD r = p p
*535.
b:.pDq.pDr.^:p.D.q = r
b:p.p=q. = .q.p = q
b:.p.D.pDq: = .pDq
*54.
;*
[*47384
[Comp
*5*1]
[*2778G]
[*53
*532]
[Simp.*243]
bz.pDq.D ,p'Dr:=ip.7J.q'2r
*542. b ::p D q"5r: = :,p D :q.O .p .r
*544 b :: p D q D :,p D r = p D q r
*5*5.
b :.p. D jOg. = .q
.
[Ass . *438]
*541.
is
*533.
*5 36.
[Simp*CompSy)l]
*487]
[*476
[Ass
*5 501 .h:.p\3:q. = .p = q
*5332]
Exp Simp]
[*5*1
Exp
Ass]
*5'53.
*5
54.
[*4 73
*5
55.
b :. p q = p v p q =
q
b: pvq. = .p;v:pvq. = .q
[*1* 3
*51
*474]
*56.
b:.p.~q.D.rz = :p.'5.qvr
1*487
^
*4648ol
*561.
*5 62.
b :.p
*563.
*57.
b :.pv7'
*571.
[*477]
pvq
<^q
= ,p
.q.v .~q =
:
\.pv q.
~q
[*4* 74.*532]
^~
.pv~q
*47
[*562~^1
ip .v .cop q
.
= .qvr = ir.v.p^q
:. qD~r.'D:pyrq.r. = .p.r
.
q>
p\
[*4 74.*l3.*51.*437]
"Hp" means
the hypothesis
Dem.
Db:.pvq.r.= :p.r.v.q.r
h.*44.
V *462ol .O b
.
::
Hp D :. ~ (q
.
r)
(2)
*422
Prop
(1)
:.
D:.p.r.v.q.r: = :p.r
[*4'74]
b .(1)
(2)
MATHEMATICAL LOGIC
126
*5*74.
:.
[PART
Dem.
b
#5*41
r .
(1)
*4'38
h ::p
Dg s
.
jO r
.= :..p
= :. jp
[#4"76]
*5'75.
b:.r'D<^>qip. =
.qvr:D:p.<^q.=
(1)
D .ql>r:p. D rO^ :.
D 3 = r :: D .Prop
.
I
.r
Dem.
h.*56.
I .
Df:.Hp.D:p.~^.D.r
v r D .p
h :. Hp D
*327 D
.
g
D:rDp
[*477]
(2)
h.*326.Dh:.Hp.D:rD~2
(2) (3) Comp D h :. Hp D
.
f .
[Comp]
H
(1)
(1)
(4)
Camp D b
.
:.
(3)
:
r
r
Dp
.
r
jp
D~q
.
~q
Hp .D:).~g. = .r:.Dh.
(4)
Prop
SECTION B
THEORY OF APPARENT VARIABLES
EXTENSION OF THE THEORY OF DEDUCTION FROM
LOWER TO HIGHER TYPES OF PROPOSITIONS
*9.
Swnmary
of*9.
In the present number, we introduce two new primitive ideas, which maybe expressed as "<f>x is always* true" and "<f>x is sometimes* true," or, more
correctly, as "<j>x always" and "<fix sometimes." When we assert "0# always,"
we
are asserting
all
values of
<&,
<f>x
is
itself,
as
is
We
shall denote
"$x
$x,
is
number of dots to
The form in which
i.e.
such
a proposition as
(x)
<f>Xm
y.tyx,
always implies
"<f>x
This is the form in which we express the
yfrx."
universal affirmative "all objects having the property have the propertyi/r."
i.e.
Here "g" stands for "there exists," and the whole symbol may be read
"there exists an x such that <j>x."
In a proposition of either of the two forms (x) .
fa, fax) fyx, the x is
an apparent variable. A proposition which contains no apparent
variables is called "elementary," and a function, all whose values are elementary propositions, is called an elementary function. For reasons explained
in
Chapter II of the Introduction, it would seem that negation and disjunction
and their derivatives must have a different meaning when applied to elemen.
called
tary propositions from that which they have when applied to such
propositions
<j>x or fax) . $x. If $x is an elementary function, we will
in this number
as (x) .
call (x)
*
<f>x
We use
and fax)
<f>x
"firstorder propositions."
to "sometimes."
all eases,"
not "at
all
times."
A similar remark
fact
applies
MATHEMATICAL LOGIC
128
[PABT
to
must
assertion of a
number
we must
different
Likewise in regard to
we must
to elementary propositions,
elementary propositions, it
and the
is
p, q, r of
*1
disjunction of firstorder propositions, and proofs of the analogues, for first6. (*11 and #1*11
order propositions, of the primitive propositions *1*2
which extends the theory of deduction from elementary to firstorder propositions. Thus by merely repeating the process set forth in the present
number, propositions of any order can be reached. Hence negation and
disjunction may be treated in practice as if there were no difference in these
ideas as applied to different types^Nthat
is
to say,
when
"
~ p"
or
"pvq"
positions of
any order
or,
in the case of p v
q,
of
limitation,
same in all types, would only arise if we ever wished to assume that there is
some one function of p whose value is always ~ p, whatever may be the order
of p, or that there is some one function of p and q whose value is always p vq,
whatever may be the orders of p and q. Such an assumption is not involved
so long as p (and q) remain real variables, since, in that case, there is no need
to give the same meaning to negation and disjunction for different values of
p (and q), when these different values are of different types. But if p (or q)
is going to be turned into an apparent variable, then since our two primitive
and restrict
ideas (x) 4>x and (g#) <f>x both demand some definite function
it follows that negation
the apparent variable to possible arguments for
and disjunction must, wherever they occur in the expression in which p (or q)
.
<f>,
</>,
is
SECTION B]
129
if
we
"\.pv~p"
there
is
of any order, and then give to the negation and disjunction involved those
argument
is
order.
But
if
we
assert
p, of a function
and
p and
We
we can
and prove analogous ideas and propositions as applied to propositions of the forms (x) . <f>x and (<^x).<f>x. By mere repetition of the analogous
process, it will then follow that analogous ideas and propositions can be defined
and proved for propositions of any order; whence, further, it follows that, in
all that concerns disjunction and negation, so long as propositions do not
define
and
disjunction.
be relevant in practice
number.
interest of the present
number
propositions, we can
deduce the theory of deduction for propositions containing apparent variables
from the theory of deduction for elementary propositions. From the purely
to
development
is
resumed.
number, we prove
they hold for propositions containing n apparent variables, also hold for such as contain n + 1,
yet
if
may be used
to infer
r& w
MATHEMATICAL LOGIC
130
number of apparent
containing any
Mathematical induction
variables.
is
[PAKT
is (as will
is
appear) incapable
What we
means of the
propositions in the present number, is to prove our desired result for any as
say
Thus we can
prove, concerning
The
by step
principle
We
<f>x.
built up.
We
Definition of Negation.
(go?)
have
first to
always true"
<f>x is
is to
<f>%
mean
~ {(x)
~> {fax)
To avoid
v
fax)
*9011.
*9 021.
<f>x}
(gar)
<f>x
we
brackets,
shall write
<px in place of
~ (x)
~ fax)
<\>x
<f>x
~ 4>x,
i.e.
<f>x is
<j>x
"it is
not$#
to be defined as
Df
= .(#). ~ <f>x Df
<f>x}
as fax)
(gar)
is
*902.
and
is essential,
Definition of Disjunction.
propositions concerned
To
in place of
<f>x
~ {(x)
<f>x} y
and
Df
Df
<f>x\
define disjunction
is
Thus
(f>x}
~ (x)
when one
we have
or both of the
as follows:
*903.
#904.
#9*05.
*9*06.
*9'07.
*908.
Df
:=.(#) <f>x vp
= (x) p v <f>x Df
p v (x) <#
=
(j>x
v
Df
p
fax)
fax) <f>x vp
=
fax) .pv<px Df
p v fax) ,<f)xz
(x) <f>x v fay) yfry: = (x) fay) .fyxytyy
Df
=
yjry
v
(x)
(a?)
v
<f>x
fay)
fay) yfry <f>x Df
(x).
.
<f>x
j>
.
to apply also
when
<f>
and
i/r
elementary functions.)
<f>x
or (x)
<f>x
it
does not really occur, since the scope of the apparent variable really extends
SECTION B]
theory of deduction
is
131
concerned, (gar)
$x and (a?)
<f)X
The
be transferred unchanged to
The above
(x)
definitions can
<f>x
and (g#) .
be repeated
<f>x.
The
Primitive Propositions.
may be
number, and
which
namely
propositions,
tions,
effect the
*91.
b.(f>x.D.(^z).<f>z
*911.
Of
which
<j>x
<f>y
We
have
first
two
Pp
(rz)
Pp
<f>z
true;
i.e.
if
we can
find
is "sometimes true." (When we speak of a function as "sometimes" true, we do not mean to assert that there is more than one argument
for which it is true, but only that there is at least one.) Practically, the above
the function
it is
Instances of such axioms are the multiplicative axiom (#88) and the axiom of
infinity (defined in #12003).
in
$z.
is
We
.
<f>x
But
if
we
try to infer
proposition
(a^)
<z
and
<f>y
<f>x
v <y
where
faz)
D
is
<f>z.
faz) . <f>z,
(gs) . <z.
Now
it
will
be
found, on referring to *4'77 and the propositions used in its proof, that this
proposition depends upon #12, i.e. pvp.D .p.
Hence it cannot be used by
us to prove (ga?)
to
v (ga?) ^x : D (ga;)
<f>x
assume the primitive proposition *9'11.
We have
<j>x,
next two propositions concerned with inference to or from propoopposed to implication. First, we have,
92
[PART
MATHEMATICAL LOGIC
132
new meaning
for the
What
is
Pp.
is true.
That
is
<f>
from a
an apparent variable, namely "when <f>y may be asserted, where y may
be any possible argument, then (x) . <f>x may be asserted." In other words, when
is true however y may be chosen among possible arguments, then (x) <j>x
<f>y
That is to say, if we can assert a wholly
is true, i.e. all values of <f> are true.
real to
ambiguous value $y, that must be because all values are true. We may express
this primitive proposition by the words " What is true in any case, however
the case may be selected, is true in all cases." We cannot symbolise this proposition, because if we put
:
"\:<f>y.3.(x).<l>x"
that means: "However y may be chosen, <f>y implies (x) <f>x," which is in
general false. What we mean is: "If <f>y is true however y may be chosen, then
(x) <f>x is true." But we have not supplied a symbol for the mere hypothesis
.
what
is
asserted in " r
primitive proposition
4>y" where y
is
is
[<f>y]
(x)
is
(f>x
Pp.
when we
actually have
*913. In any assertion containing a real variable, this real variable may be
turned into an apparent variable of which all possible values are asserted to
Pp.
We
We
(cf. p.
say that x
51).
is
"individual"
types.
if
a; is
These
neither
133
SECTION B]
The
type."
following
is
a stepbystep
We say that u and v "are of the same type" if (1) both are individuals, (2) both
same
and v
negation, (4)
is its
is <f>x
or ^Sb, and v
u is a function
type, (3)
$c v
is
where
tylb,
y) and v
and tyx
where
<pfc
4* (P* P)>
^ (^ 9) are f tne
a proposition and v
^& are
type.
(7)
is
<a\and
Our
*9'14.
same
is (z)
ty (x, z),
If "<f>x"
significant,
*915.
of the
If,
and vice
is significant,
and vice
be seen
(x)
is
if
(Cf.
is
of the
same type as
note on *10*121,
a proposition
p. 140.)
then there
<f>a,
a, "<j>a" is
a function $&,
is
Pp.
versa.
It will
Pp.
versa.
some a, there
for
then
"D
p means
i.e. (gyi?)
(ga?) *<f>x
~(a?) .<f>x.v.p,
.<^<f>xvp,
i.e.
i.e.
(gai)
(x).<*j<f)xvp,
i.e.
(x)
(gar) .~<f>x
<$>x
Op
i.e.
(x) .~<f>x
<f>x
v . p,
p,
Dp
In order to prove that (x) $x and (ga;) . <fyx obey the same rules of deduction
as (ftx, we have to prove that propositions of the forms (x) x and (ga;) . <f>x
may replace one or more of the propositions p, q, r in #1'2 *6. When this has
been proved, the previous proofs of subsequent propositions in #2 #5 become
applicable. These proofs are given below. Certain other propositions, required
.
(a?)
The above
^wzr 
to the particular,
i.e.
"what holds in
all cases,
Dem.
h.*21.0h.~<yv<f>y
(1)
^Dhrv>0yv^y.D.(aa?).~^v^
h.(l).(2).*lll.Dl.(aa;).~^v^
H.*9*l
1
[(3).(*905)]
r
[(4).(*901.*r01)]
\:(x).<f>x.0.<l>y
(a)
<f>x
(2)
(3)
(4)
<f>y
In the second line of the above proof, " <^> ^>y v <j>y" is taken as the value,
for the argument y, of the function "
<f>x v <f>y," where x is the argument.
A similar method of using #91 is employed in most of the following proofs.
applications of definitions.
Hence
all
it will
steps
not be
134
MATHEMATIcili I#GIC
[PART
important.
*9'21.
:.(x) .<f>x"Dyfrx
I
always implies
J.e. if <f>x
is
then
yjrx,
"<f>x
always" implies
"yfrx always."
The
Dem.
h.*208.
h
*91
Db:<f>zD^z,D.<f>z^^rz
(1)
Db:fay):<f>zDylrz.D.<f>yDyfrz
(2)
K(2).*9'l.
Dh:.fax):.(^y):4>xDyjrx.D.<f>yDiJrz
(3)
h.(3).*913.
Dh::(z)::fax):.fay).(l>x^yjrx.D.(f>yD^z
(4)
(1)
[(4).(*9'06)]
\::(z)::fax):.<f>xDylrx.D:fay).<f)yD^z
[(5).(*r01 .*908)]
: .
fax)
[(6).(*9'08)]
:.
fax)
[(7).(*1'01)]
Y:.(x).<\>x^^x.^'.(y).4>y.^.(z).y\rz
This
<f)X,"
is
I.e. if <f>x
always implies
#921,
is
tyx,
then
(6)
(7)
<f>y" is
yjrx."
D.fax).tyx
t:.(x).<f)xDylrx. D:fax).<f>x.
proposition, like
(5)
is
sition as "(as)
*9'22.
~
~
if <f>x is
sometimes
true, so is tyx.
This
Dem.
K*2'08.
0jO fy D
.
<y
(1)
K(l).*91.
Db:faz):<f>yDfy.D.<f>yDfz
h.(2).*91.
D\:.fax):.faz):<f>xDfx.D.<l}y^fz
(3)
D\::(y)::fax)'..faz):<j>x^yfrx.D.<f)yDylrz
(4)
I .
(3)
*913
fax) :.<f>xD+x.D
,(2)
[(4).(*9'06)]
[(5).(*r01.*908)]
\::fax).~(<f>xDylrx):v:(y):faz).<l>yDylnz
[(6).(*101.*9'07)]
I"
::
(y)
::
" (3)
~(^ D ^a?) v
:
faz) .cfyyO^z
(y)
(5)
(6)
~<y v (gs> ^*
.
(7)
[(7).(*101.^90V02)]b:.(x).<f>xDy{rx.D:fay).<f>y.O.faz).yJrz
the proposition to be proved, because fay) . <j>y is the same proposition as fax) . <f>x, and faz) . tyz is the same proposition as fax) . yfrx.
This
is
[Id
*91321]
[Id
*9\L322]
*923.
H:(ar).^B.D.(*).^c
*9 24.
*925.
\:.(x).pv<f>x.D:p.v.(x).<j>x
<f>x
[*9"23
(*9 04)]
<j>x
"6,
or fax).
replacing
<f>x.
The
SECTION B]
#9'3.
H :.(?).
<f>x.
3. (a;).
v.(x).<f>xi
135
0a*
Bern.
Dh.^v^.D.^
(1)
Dh:(gy):<a*v0y.3.<a;
(2)
h.*l'2.
K(l).#91.
K(2).*913.
D>:.(a):.(gy):<a*v</>y.D.</>aV
[(3).(*9050104)]
(4)
#9*21
This
(gar)
:.
is
: .
<f>x
v (y)
<y
(3)
<j>x
(4)
y\:.(x):<f>x.v.(y).<l>y:D.(x).<f)x
[(5).(*903)]
#9*31.
:. (a*)
. <*Sa* .
:. (a?)
(ga*)
<a?
<a*
v.(y).<f>y:
(gas)
D ..(a?)
<a :.
(5)
D H Prop
.
<f>x
Dem.
h. #9*1113.
Db:(y):<f>xv<l>y.D.(^).<f>z
h
[(l).(*90302)]
h.(2).*913.DH:(a;):( a2/)./>a
#9*32.
[(3).(*90302)]
:.
(ga*)
[(4).(*905*06)]
1
:.
(ga*)
<f>z
v^.D.(^).^
v <y : D . (gs) . *
(gy)
</>y : D . (g*) . </>*
<f>x . v . (gy)
(1)
(2)
(3)
(4)
h:.^.D:(a?).^>a?.v.^
Dem.
K*l*3.
3h:.q.D:<l>x.v.q
h (1) #913 D h :. (x) :. ? D fs
.
r
:. g
:. (a?)
/>
(2)
[Proof as above]
,<f>x.v.q
(ga*)
. <j&a .
h:.q.O:(x).<f)X.y.q
[(2).(*903)]
#9*33.
(1)
.
Dr:.g.D:(a*):<a*.v.2
[*925]
#9*34.
v . (x)
. </>a*
Dem.
h.*l3.
1 .
(1)
#913
#9*35.
#9*36.
\"..
: .
(3)
D p v
v (x)
<j>x
(3)
(#904) D
(2)
D\:(x).(f>x.D.(x).pv<f>x
.
<f>x
(ga;)
.
(1)
D\i(x):(f>x.D.pv<f>x
K(2).*9*21.
I
3\:4>x.D.pv<f>x
(ga:)
(x)
<f>x
Prop
[Proof as above]
<a*
.
Dem.
H.*l4.
H
I
#9*361.
I
*9'37.
#9*371.
r
D\
:pv
<f>x."D
p D p v (a;) cf>x
:.^>. v.(ga?). <a*: D :(ga*). <f>x v.p
(ga;) <px
:. (ga?) .<f>x.v ,p:D:p .v
:.
(x)
.<j>xvp
D\:(x).pv<f>x.D.(x).<l>xvp
(1) #913*21
(2) . (#903*04) D h Prop
.
<j>x
[Similar proof]
[Similar proof]
[Similar proof]
(1)
(2)
136
MATHEMATICAL LOGIC
*9'4.
::p
I
v q v
:
(x)
D :. q
<f>x :.
[PART
v :p . v . (x)
.<f>x
Dem.
r.*l'5.*9'21
Dh.(*):j).v.gv^s 3; (a;) :q
v .pv<f>x
(1)
l.(l).(*904).DH.Prop
*9'401.
I
*941.
r
*9'411
p v
p v
p
v
::
*942.
I
*9421
\
*95.
h ::p
::
q . v . fax)
::
(x)
q:v:p.v
>:.
<f>x :.
v r :. D
<f>x
:.
(x)
<$>x
fax)
[As above]
<a;
pv
[As above]
[As above]
::
::
fax)
x v q v r :. D :. g v (ga?) <j>x v r
D :. p v (x) . (fix : D : q v (x) <$>x
[As above]
Dq
[As above]
Dem.
r
#16
:.pDq.D:py<j>y.D.qv<f>y
:.pD q. D :fax)zpv<f>x. D .qv<f>y
Dh
H.(l).*9l.(*906).
K(2).*913.(*904).Dr
D9
Dg
r.pDq.
:: p 3
::p
[(3).(*908)]
r ::p
[(4).(*9'01)]
J
[(5).(*904)]
I
*9501. h ::p
>g D :.p
$9*51.
f
::p
(x).
(ga?)
<? .
v.r
(x).<f>x.
"Dt.pvr. 3
<j)x:
(1)
(2)
Dem.
K#r6.
h
(1)
(2)
*9'511. h
#952.
::
*91321
I
:.pD"<j>x.
D :pvr. D. ^rvr
(1)
Db:'.(x).pD<f>x.3:.(x):pvr.D.tf>xvr
(*90304) . D h . Prop
p . D fax)
.
h ::'()
<f>x
(2)
<#
"D :.
:.
pvr D
.
(x)
<}>x
(ga?)
.V . r
<f>x
v r
[As above]
2 qv r
Dem.
K#r6.
D H:.^a?D^.D
(1)
*91322
K (2)
(*90501)
(3)
(*903)
I .
*9 521.
\
*9"6.
(a;)
fax)
::
.
<j>x,
<f>x
~(a?)
D
.
:<a?vr
D.gvr
(1)
3b::fax).4>xDq.O:.fax)z<l>xvr.D.qvr
D H :: (a;) <j>x D g : D :. ().^vr.D q vr
DH.Prop
^>a;,
D :. (ga;)
^>ar
v .r:
<j>x
gvr
(2)
(3)
[As above]
are of the
same
tjpe.
[*9*131,(7)and(8)]
*9"61.
If
$e and
tyic
same
type, there
is
function ^vir^.
Dem.
By
*9'14*15, there
is
and therefore so
Hence the result by *915.
significant,
an a
for
is "<j>a
for functions of
any number of
variables.
SECTION B]
137
*9 62.
^> is
If <(, )
iy) .<}>($,y).v.
^a, (a^)
<t>
(^
y)
^>
Dem.
<f>
*9'63.
If
<f>
$ (&,
$),
i/r
<f>
(&,
($,
We
. <f>oc,
"(ga?)
<#,
*5
number
of
*10.
Summary of *10.
The
number is to extend
x D yjrx) as many
formal implications
(i.e.
Then we have
Socrates
a Greek
that
e.g.
as
we have proved
i.e.
in *3*33 that
pOq.qDr.D.p'Dr.
p = Socrates is a Greek,
q = Socrates is a man,
r = Socrates is a mortal.
Put
'
pDq. Thus
form
to
is
" if
man
implies
'
'
'
Socrates
implies
'
Socrates
a Greek
is
Socrates
'
'
is
men
all
implies
a mortal,'
a mortal.' "
if all
is
But
'
it
Socrates
is
follows that
a man,' and
'
Socrates
is
all
Greeks are
mortals.
= x is a Greek,
sjrx
x is a man,
XX = x is a mortal,
Putting
we have
<fcx
to prove
(x)
<$>x
D yfrx
(x)
yfrx
Many
tytb.
(x)
<f>x
D %x.
0& and
D %x D
We
assume in this number, what has been proved in #9, that the
*5 can be applied to such propositions as (x) <px and
of
Instead
the method adopted in *9, it is possible to take negation
shall
propositions of *1
(gaj)
it is
<f>x.
*1001.
= .'^(x).^>^x Df
( 3/c).<f)x.
l
In order to make
we
shall, in
it
clear
how
this alternative
139
SECTION B]
The two
is
often
*1002.
<f>xDx ^x.
*10 03.
<f>x
for
#2#5.
= x tyx
(x)
$x D tyx
or (x).<px
yfrx.
= .(x).<f>xDylrx Df
(x) <f>x = yfrx. Df
.
The
following propositions
The
*101.
I.e.
number
The
(a;)
what
<#
true in
is
are very
propositions
<y
cases is true in
all
any one
case.
:.
(x)
p D <f>x = p D
*10'22.
:.
(x)
<f>x
The
yfrx
:(x).
{as)
<f>x:
be,
<py
<f>x
(x).tyx
demand
that
<j>
and
yfr
*1023.
b:.(x).<f>x'2p.
I.e. if <f>x
*1024.
:('2Lx).<j)x.D.p
<j>y
$y
I.e. if
method
(roc)
is true,
if <f>x is
ever true,
is true.
<f>x
then there
an x
is
which
for
<f>x is
true.
This
is
the sole
of proving existencetheorems.
*10 27.
I.e. if
:.
<j>z
(z)
<f>z
D tyz D
.
always implies
(z)
<f>z
then "
tyz,
<f>z
(z)
tyz
yjrz
always."
The
three following propositions, which are equally useful, are analogous to *10"27.
*10'281. H
:.
#1035.
:.
*10271.
f
*10'28.
V :.(x). <f>x'2i{rx.
:.
(z)
4>z
*10'42.
:.
#10
: .
fax)
5.
<j>x
fa D
.
fx
fax)
= fax)
.
<f>x
.<f>xvyfrx
fax)
yfrx
MATHEMATICAL LOGIC
140
[PART
This
the source of
is
many subsequent
multiplication.
#10*51.
:..
~ {(gas)
This proposition
<f>x
tyx)
</>#
'3*
~ tyx
analogous to
is
~ (p
q)
= p 3 <>*q
.
Of the remaining
at a
much
#1001.
later stage.
(ga*).<j!>#.
This definition
= .~(a ).~<a* Df
i
is
only to be used
when we
discard the
method of *9
(ga;) .<f>x.
<j>x= x '\Jrx.
.(x).<f>x = $'X Df
#10*1.
.3 $y
#10*11.
(x)
. <f>x
$tx.
^D^rx Df
#10*03.
~(a*)
in
we have
[*9*2]
argument y may
true. ^*9l3]
This proposition
"
What
true of
is
is,
in
all is .true
#1012. h
of
\x)
is
true of
p v <f>x, 3.: p v
.
(a?)
all."
[#9*25]
tf>x
"qSq,"
ifxc" is significant,
and vice
then
if
is
of the
same type
as x, '*#" is
versa. [#D:1 4]
must be
#10122.
And
df the
If,
same type;
for
vice versa.
some
a,
for if a?
there
is
is
a function ^c,
[*9'I5J
arid
141
SECTION b]
Bern,.
By
there
(3),
a function
is
~^v~^.
<^<f>xv<^y^x. v
(1)
<f>x.tyx
h.(l).*232.(*l01).Dh:.^.D:'/ra;.3.^,^
Prop
h (2) *912 D
.
*1014
h :.(#).<#: (a?).
This proposition
when
significant
and
<f>
\fr
it is significant,
is significant.
this.
Hence,
be applied when
if it is to
same
Dem.
(1)
ty take
arguments
(I)
D\z(x.).<f>x.y.<l>y
(2)
Dhr^.^.D.ifry
*1013 D h (x) <jxc D fyy
[*3*47]
(2)
z
(x)'. far
\:.(x).pv
yfr
m
D\i.{x).<f)x:(x).'fx:'2>.<f}y.'^y
*102.
^ are given, or
type.
K*10l.
K*10\L.
H
and
and
<
or vice versa,
given as a function of
is
yfr
true whenever
should take arguments of the same type, while the hypothesis does
not demand
when
is
^arO .<f>y.yjry
hypothesis
its
(2)
..
DKProp
Dem.
h
*10'1
*l6
K(l),(2).
h :..(#) 2>
^ </>* =
This proposition
*10'22.
:.
(x)
<f>x
K*1012.
*1021.
D H :. p v (x)
D f :;.(y) :. p v
<f>x :
D p v <f>y :.
.
:0.pv<f>y
.
.
yh:.p.v.(x).<j>x:D.(y).pv<l>y
3>:.(y)._pv^y.3:p.v.(a?).^aj
Dr.. Prop
[*1011]
[*1012]
=2>
0a?
(a?)
z.
(1)
(2)
3 (*0<^ [*102^1
is
.
Dem.
[*1021]
D ^y yfry
D.^y:
DI:.(y):(ar).^.^.D.^:.
>.h:.(#)/^.^.D.(y).<y
h.(l).*3'27.
y\i.\x).<\>x.^x.^.^zi.
[*1011]
D H : (x) $x
*10'1
[*326]
[*1011]
(3)
(z)
(x)
^rx
$x
yjrx
(1)
fz
(2).
z .
D\:.(x).sf>x.ylrx.D.(z).fz
[*10'21]
t .
(2)
*101411
I : .
<f>y
(z)
[*1Q21]
K(4).(5).
Dh.Prop
(3)
(4)
(5)
MATHEMATICAL LO0JC
142
The above
proposition
true whenever
is
(#).<#: (x)
yjrx
it is
significant;
but, as
is
it
[PART
was
when
" is significant.
#10*221. If
<fxc
x u
stituent
>
>
v>
...),
where
ia
t>,
<f>
>
>
>
>
(wf)
~x ( a
>
~fa v x ( a
y) v 0a > (y)
>
y)>
b:.(x).<f>x^p. = :fax).<f>x.D.p
#10*23.
Dem.
h
#42
(#9'03)
:.
(x)
~<a; v
p =
.
(1)
(x)
~<f>x
s.(a*).^.Di>
[(*902)]
h
(#101)
(1)
DKProp
proceeds as follows.
#10*23.
\:.(x).(f>xDp.
= fax)
:
<f>x
D.
fax)
<f>x
Dem.
b
Transp (#1001) D
.
I
:.
D p = ~p D
.
(x)
~<f>x
[#1021]
=:(a;):~jj.D.~^r:
[#101]
D:~p.D.~<j)x:
[Transp]
[#1011]
I
:.
(x)
:.
fax)
.<f>x.
:<f>x
D p .O
.
.D .p
:
<f>x
:.
D p
.
:.
(1)
143
DH:.(g^).^c.D.j?0:(a;):f.D.p
[*10'21]
.
SECTION B]
Dh:.(jr):^.O.p:D:^ Dp
#10'1
(2)
D:~j3.D.^^a;i.
[Transp]
3b:.(x):<f>x.D.p:'2:(x)z~p. 2.~<f>x:
[*10'11'21]
D:(a^).^.D.p
[(!)]
(3)
DKProp
K(2).(3).
1024.
This
f :
is
<j>y
fax)
<f>x
*9'1.
it
as follows.
is
Dem.
K*101 .Oh:(a;).~^.D.~^:
D h <f>y D ~(#)
Prop
[(*1001)] D
[Transp]
~<f>x
I .
1025.
b:(x).<f>x.D.fax).<j>x
*10 251.
10252.
:~{fax)
10 253.
(x)
~ \{x)
~ {(x)
<#}
<f>x
<f>x]
[*10124]
[*1025 Transp]
<#}
(x).~<f>x
[*42
fax) .~<j>x
[*4"2
(*9'02)]
(*901)]
<f>x is
*10252.
10 253.
[fax)
<f>x}
~ {(x) .<f>x}. =
==
(x)
fax)
[*413 (*1001)]
.
<f>x
<f>x
Dem.
K*10'l.
Dh:(a?).^c,D,^y.
[*212]
[101121] D h
[Transp]
Dh
[(1001)] D
(x)
<f>x
~ {(y) ~
.
D.~(~0y):
D (3/) .~(~tf>*/)
.
( <o
^>y)}
~ {(#)
<#}
K*101. DH:(y).~(~03/).
D .~{(*) <M
D.~(~<#)
[214]
2.<j>x:
I :
[*101121]
[Transp]
3h
Db
(ay) .~*y
(y)
~(~#)
~ {(#)
[(1001)]
K(l).(2).Dh.Prop
<f>x}
D
D
(x)
(1)
<f>x
{(3/)
~ ~ $3/)}
(
D.(ay).~*y
(2)
MATHEMATICAL LOGIC
144
*1026.
"If
[*101
b:.(z).<f>zyyjrz:<f)xiy.fa;
[PART
is
all
is
men
and Socrates
are mortal,
is
Imp]
This
tyz
a man,
mortal."
is
h:.(<z).<f>zyylrz.D:(z).<f>z.D.(z).yfrz
This
is
*9*21.
as follows.
is
Dem.
b
#1014
(z)
:.
D yjrz
<f>z
(z)
D
D
* <f>z :
[Ass]
Dh
Dh
[#101]
[*10'21]
I
*10271.
:.
(y)
:.
{z)
:.
.
(z)
D yjrz
cf>z
0* D
<fz
(1)
tyy
<j>z :
(*)
0* : D
(z)
^z
(y)
<f>y
:.
.
yjry :.
C1 )
^2/
Exp y r Prop
D :.(*) . 0* =
:. 0)
0z =
(z)
D ijry
4>y
Dem.
K*10'22.
K*1022.
'
This
is
: .
(1)
(x)
D:(s).^s.>.(.!r).0*
(2)
Comp . y h Prop
D (ga;) 0a; D
(2)
(ga?)
^a;
#922.
(1)
DJ:.Hp.D:(*).^eO0s:
[*10'27]
#10*28.
>:(z).<f>z.D.(z).yfrz
[*1027]
as follows.
is
Dem.
h
*ioi
y\
(x)
[Transp]
[*10\L1~21]3>
:.
{x).<f>x
y tyx
[*1027]
[Transp]
D
y
: .
(a;)
*1029.
:.
.<j>x= tyx
(rx)
.
<f>y
.~fyy~<f>y :
y:(y).~ylry.y.(y).~<f>y:
D (ay) . 0y 3 (ay) ^y 3
:
(y)
*10 28L
~ tyy D ~
<f>x
0a;
(a)
D %x =
:
(x)
<f>x .
yfrx
xx
Prop
r .
[*1022"28
^ra?
Comp]
Dem.
h
*lQ22
D %a;
= (ar) 0a; D ^a; 0# D ^a;
0a; D tyx XXZm
D"r :. 0a; D^ar* 0a; D ;*?. =
D K :. (a;) :. 0a; 3'i^tt'. 0a; D %a? = 0a; D tyx ^a; :.
y b :. (a?) 0a;O ^ar. 0a; D %a? = (x) 0a; D ^ra? xx
D
:.
(x)
0a?
D tyx
(a;)
0a;
t
#4*76
[#1011]
[#10'27l]
r .
(1)
This
is
(1)
(2)
SECTION B]
#10*3.
This
(x)
:.
<f>x
D yfrx
(x)
D ^ D
^ra;
(#)
D xx
<a"
145
is
Dem.
\
#10*22*221
D r Hp D
(x).
(x) .<f>xO
[Syll .#1027]
#10*301. H
(x)
:.
<f>x
i/ra;
#1 0*22*221
I .
(a*)
:.
I
= %x : D
i/ra?
D yfrx
<f>x
(a;)
Hp D
(x)
<f>x
(a*)
0a;
[*4*22.*10*27]
yfrx
D xx
Xx;Oh.
Prop
= ^
<f>x
= yfrx yfrx = xx
= X a? :. D h Prop
.
In the second line of the proofs of #10*3 and #10*301, we abbreviate the
way which is often convenient. In #10*3, the full process
process of proof in a
would be as follows:
D xx D <f>x D xx
yfrx D x x ^ $x ^
[#10*11]
D
D
[#10*27]
Dh().^Dfx,fa;D^.D.
Syll
r :
<f>x"D yfrx
I :
(x)
<f>x
yfrx
D yfrx
(a?)
X^
<#
D %#
:.
(x)
<f>x
D yfrx D
.
(x)
<f>x
xx D
.
^ra;
xx
Dem.
K Fact. #10* 11
Dh :.(x):. <f>x^yfrx.
K(l). #10*27.
#10*311.
:.
I
(x)
<f>x
D:<f>x.xx. ^.yfrx.xx
(1)
Dr. Prop
yfrx
(x)
<f>x
xx =
tyx
Xx
Dem.
r
#4*36
#10*11
D :. (x) :. $x =
Dr. Prop
yfrx
I
K(l). #10*27.
The above two propositions are extensions
#10*32.
h <f>x = x yfrx = yfrx = x x
:
<a?
xx =
"f
xx
(1)
Dem.
V
#10*22
I :
= x yfrx . =
<f>x
=
=
[*43]
[#10*22]
<f>x
Dx yfrx
#10*321. H
yfrx
D* <f)x .
yfrx
Dx <f>X
yfrx
=, ^v
Prop
symmetrical.
is
tyx %XX
Dem.
V
#10*32
Fact
Hp D
yfrx
= x <f>x
yfrx
=x x D
[#10301]
#10*322.
I
i/r#
=x
<f>a: .
xx = x <f>x D
.
\frx
==
<f>x
= x xx
Prop
x xx
Dem.
V
#1032 D
[#10*301 ]
R&W
Hp D
yfrx
= x <j>x
yfrx
=x Xx D
<f>x
= x xx
.
Prop
10
[PART
MATHEMATICAL LOGIC
14$
*10 33.
I
:.
(x)
<f>x
.p =
(x) .<l>x:p
Dem.
Dh:.(x):<f>x.p:D.<f>y.p.
h.*10'l.
(1)
D.p
[*327]
(2)
(#):<#. p :D
<y
p D
(y)
[*101121]
D
D
K(2).(3).
D\:.(x):<f>x.piD:(y).<l>y:p
I .
(1) . *3'26
:.
I
:.(#):
<f>x
(3)
<f>y
K*101.
D\:.(y).<f>y.
[Fact]
Dh:.(y).^:p:D.fB.p:.
[101121]
I
(4)
(5)
D.<jk:.
I
:.
(5)
0\'..(y).<f>y:p:0:(x):<f>x.p
D h Prop
.
*1034.
(4)
(g)
is
In the alternative
as follows.
Dem.
K*42.(*1001).D
= :~{(x) .~(<f>xDp)}
H:.( a#).<aOjp.
[*461.*10'271]
=:~{(a?):^r.~p:
[*1033]
[*453]
=
ty] v .p
s :(x).<f>x.D.p
:
[*46]
*10 35.
: .
~ {()
~ {(*)
Dh:p.</>a:.D..p:
[*1011]
[*1023]
Dh:(#):p.<^.O.p:
D h (a*) p <a? D p
Db.p.Qx.D.fa:
*327
~/>}
'.
(1)
[*1011]
D\:(x):p.4>x.D.<f>x:
[*10'28]
Dh:(ga;).p.^c.D.(aa!).^
K*3'2.
"D\:.p.Di<f>x.D.p.<f>x.
[*101121]
t
:.p
(x)
<#.
D .p
<#
(2)
D:(ga;).<^.3.(aa;).p.^B
.(1).(2).(3).
(3)
= :(a).^.v.p
h:.(a)./>vp.
H.*326.
[*10'28]
proof
<j>#
(ga;) .p.<f>x.
r .
*1036.
as follows.
Dem.
h
*464
[*1011]
[*10'281]
[*1034]
[*4'6.(*1001)]
D
D
D
I :
<j>xvp
I :
(x)
1
:.
<j>x
,~<j>x'Dp:
vp =
.
=
(a) .^vj).
=
=
<f>x
Dp
.~<#
Dp
(ga?)
(x) ,f^><j)X.
(3^) .f.v.])!.3
D .p
K Prop
:
SECTION B]
The above
proposition
*10 37.
I
:.
#1039.
:. <f>x
(gar)
is
p 3 <f>x = p 3
.
D x xx
147
^* 8x
'
(gar)
*1036
. <j>a
tyx
<fc
^x
P
Xx x
Dem.
r .
*1022
3f
:.
Hp 3
(a?)
<f>x
3 % ^ 3
(a)
cf>x
[*347.*1027]
This proposition
fco
when the
only true
is
.0 .xx
0#
:.D b . Prop
Occ
conclusion
is
significant;
On
the
the
#104.
:. <f>x
= x xx
.tyx
= v 0x .D
<f>x
yfrx
= x xx @x
Dem.
h
*1022
1
: .
Hp 3
.
[*1039]
(1)
(2)
Comp 3
:.
:,
Hp 3 x x
Hp 3 <f>x
"f*
&x ^x
tyx
[*1022]
In #10*4 and
significant
mate
3* xx
<f>x
D:<f)X.yfrx.D x
Similarly
I
<\>x
Dx
yfrx
=x
3* Gx
xx Ox
'
(2)
<f>
.
(1)
x tyx
xx @x Xx x ^* fa
xx Bat :. 3 H Prop
s
"x
'
i.e.
to pass
yjr,
must be such as to
<f>,
x*
have overlapping ranges of significance. In virtue of *1'0*221, this is secured if
they are of the forms F{x, x (x,p,%...)}J{x,xM,%...)},0{x, (x,'PX)},
X
or x an(^ ^
9 x X (x > $> )} It is also secured if ^ and yfr or <f> and
or x ar) d
are of such forms, for $ and
x must have overlapping ranges of
\
>
>
*10"41.
be
:.{x).$x. v .(x).yjrx:
(#)
and
significant,
.
so
must
^
and
0.
favyfrx
Dem.
h
*101
I
:(x).<f>x.
[*22]
r .
(1)
<f>y
^.cj>yvfy
*101
(1)
Dh:(*).^.D.^.
D.^yv^y
[*l3]
(2)
[#3'44]
[*101121]
*10 13 3
.
3
3
I
:. (a?)
<#
<y
yfry
(a)
i/r#
:. (.*)
$a v
(x)
^x 3
<f>y
1
:. (a?)
<f>x
(a?)
yjrx
(y)
v
.
^v
(2)
yjry :.
^y
<jty
^y :. 3
r .
Prop
Observe that in the above proof the uses of #2*2 and *1*3 are only legitimate
and ^ry have overlapping ranges of significance, for otherwise, if y is such
that there is a proposition <f>y, it is such that there is no proposition yfry, and
if <f>y
conversely.
10.2
148
MATHEMATICAL LOGIC
*10"411. h
:. <f>x
= x j(x
yfros
= x 0x
"D
v yjrx
<f>x
[PART
= z x v Ox
.
Dem.
h
*10*14
Hp D
:.
= %x
<#
yfrx
= 0x
^:<t>xvtyx. = .xxvdx
h.(l).*10ir21.DKP.rop
[*4'39]
*10412.
(1)
H :
#10413. H
Dem.
r
*10411412
I ..
*10414.
D:<f>xDylrx.= x
[(*101)]
= x %x
:. <f>x
tyx = x 0x D
= i/r# =3
<f>x
.
xxD6x:.Dh.T?rov
= 6x
Dem.
H
*10413
The
(1)
$xD Xx
CO
DKProp
*104
replaced by
<
or
is
replaced by
is
in
yjr,
*1042.
:.
fax)
<f>x
fax)
yjrx
= fax) .Qxvfx
.
Dem.
h .*10"22
[*4'11]
[*4'51'56.*10271]
[*10 253]
:.
I .
This proposition
#10*5, in
*10"43.
<f)Z
= z <fyz ,(f>x. =
.<f>z= z
yjrz
yfrx
Dem.
K*10\L.
h
*105.
:.
fax)
(1)
<f>x
*532
yfrx
^h:<f)Z= z yjrz.D.<f,x=yfrx
H . Prop
(1)
D:fax).<f>x: fax).\jrx
Dem.
b
*326 #1011 D
.
[*10'28]
h
*3'27
[*10*28]
(x)
<j>x
fx
D ,<f>x
Db:fax).(f>x.ylrx.D.fax).(l>x
.
*10'11
:.
(x)
<f>x
yjrx
(1)
tyx:
Db:fax).<f>x.fx.D.fax).yJrx
h.(l).(2).Comp.Dl:.Prop
(2)
SECTION B]
The converse
source of
above proposition
of the
is
The
false.
an
proposition states
149
many subsequent
differences
is
the
logical
h :~{fax)
yfrx]
:.~ {fax)
</>#
<f>x
Dx .~yfrx
:<f>x.
Bern.
h
#10252
yfrx]
[#4*51*62.*10*271]
#10*52.
:.
fax)
(f>x
(x)
<f>x
Dp = .p
.
Dem.
h
#5*5
::
Hp D :.p =
fax) .<f>x.D.p:
(x)
[#10*23]
#10*53.
I
~fax)
:.
<j>x
<f>x
Dx
<f>x
Dp
: :
Prop
yfrx
Dem.
h. #2*21. #10*11.3
h
#10*541.
(x)
:.
:.
oxfrx
<f>x
yfrx :.
[#10*27]
Dh:.(x).^<f>x.D:(x):<f>x.D.yfrxi.
[#10*252]
\:.<f>y
h :.~(aa;)
<f>x
(x)
<f>x
yfr
:.
D h Prop
.
Dem.
h
#4*2
= (y).pv~<f>yvyfry:
= p.v.(y).~<f>yvyfry:
[Assoc.*10*27l]
[#10*2]
p. v.
[(#1*01)]
The above
proposition
is
This proposition
*10'55.
:.
is
<f>yDy
yfry:.Db. Fro?
= :p.D.<f>yDy yfry
a lemma
f*10*541^l
for #84*43.
Dem.
I .
#4*71
I
:. <&#
D yfrx D
.
<f>x
yfrx
<f>x
(1)
h.(l).*1011'27.D
I
:.
[#10*281]
b. (2). #5*32.
This proposition
is
D fax) <fix
Dr. Prop
lemma
for
yfrx
#117 12 121.
,
<f>x
.
fax)
<f>x
(2)
MATHEMATICAL LOGIC
150
#10'56.
I
:. <j)x
Dx
yjrx:
(g#).
$x Xx: ^ (3^)
.
[PART
ty x
'Xx
Dem.
h
Imp D
#10'31
:. <f>x
Dx
yfrx
[#10*28]
h
(1)
D
D
cf>x
yx D x
.
(ga;)
<#
tyx
%a;
yx
(gar)
fa; . ^a;
(1)
Prop
This proposition and #1057 are used in the theory of series (Part V).
#1057.
\
:.<f>x
.D x
Fact
v yx
yjrx
~(g#)
(g#)
fac
yx
Dem.
*1051
I
:. <f>x
Da,
fa; v ^a;
[*10*29]
[*5'61]
h.(l).*56.Dh.Prop
<f>x
yx
Dx
D :<f>x .D x
D <a; D*
"D
<f>x
yjrx
v yx
fa; v
fa;
yx
<f>x
~ yx
D x <^yx
.
'
(1)
#11.
Summary
of #11.
one variable in *10 are to be
In this number, the propositions proved for
a few propositions having no
extended to two variables, with the addition of
<f>
{x,
is
y)
y.
The
definition
#1101
need to be taken as a
shows that " the truth of all values of (x, y)" does not
"
all values of yfrx."
new primitive idea, but is definable in terms of the truth of
becomes a function of one
reason is that, when x is assigned, <f>(x,y)
<f>
The
possible value of x,
namely. y, whence it follows that, for every
introduced in *9. But
"(y).4>(x,y)" embodies merely the primitive idea
one variable, namely x, since y has
"(y) . <f> (x y)" is again only a function of
the definition #1101 below ilHence
here become an apparent variable.
variable,
legitimate.
We put:
Df
Df
#1102.
Df
*n03.
Df
*ll04. (^x,y,z).<f>(x,y,z).^:(^x):('3_y,z).<f>(x,y,z)
Df
*ll05. <j>(x y).D x>y .1r(x,y)i = i(x,y):<j>{x,y).3.+(x,y)
Df
aj
*ll06. ^(a?,y).s,, y .^(,y): = :(.y)!*( aj.y)=^ ( 'y)
= \(x):(y).<i>(x,y)
= (x) (y, z) (*, y, z)
Qc,
z).
(x, y,
y, z)
(aa?,y).^(a?,y). = :(a*):(ay)^(^y).
#1101.
(x,y).<j>(x,y).
<J>
. <f>
>
that
may
The
number
of variables
occur.
is
all
be extended to any
finite
number
of #10, but
frequently.
#111.
\:(x,y).(f>(x,y).D.<f>(z,w)
#1111.
If
(x,
y)
<j>
(x,
<f>
(z,
y)
w)
is
is
true
following:
w may
be,
then
152
MATHEMATICAL LOGIC
[PART
*ll2.
I.e.
y"
values of
true for
all possible
*11'3.
f
This
p D
.
\:.(x, y)
<f>
(x,
y>.
y) always.'"
This
^ (x, y) O
:
r :.(x,y)
I.e. if
<f>
vice versa.
*ll46.
<f>(x,
y)
D .p =
is
(x,
y)
true for
is
fax, y)
<f>
(x,
.
<f>
.
<f>
then
and are
(x,
t:.(&x,y):p.<t>(x,y):
This
p D
y{r(x,y),
*ll35.
possible
ail
y)
{x,y)
'<f>(x,
is
y)
x."
(x,
I.e.
(>
values of
(x,
<j>(x,
all
y) <f>
y) :=:(x, y)
the analogue of *1021.
: .
is
*1132.
'f
is
.,(*,
y)
^ (x, y)
y) always' implies
much used.
(x,y) .D.p
also
<f>
(x,
y)
ever true,
is
is true,
and
:p:(>&x,y).<f>(x,y)
*H54.
^:.faso,y).<f>x.ylry.^:fax).<f>x:fay).yfry
This proposition is useful because it analyses a proposition containing
two apparent variables into two propositions which each contain only one.
"<j>x.y]ry" is a function of two variables, but is compounded of two functions
of one variable each. Such a function is like a conic which is two straight
lines: it
*11'55.
I.e.
may be
called
is
function.
r : . fax,
y) . $x . f (x, y) . = : fax)
to say " there are values of x and
equivalent to saying
there
an "analysable"
"
there
is
<f>x
a value of x for
yfr
(x,
y)
(ay)
yjr
(x,
y)
for
is true."
*ll6.
*ll62.
::
:: <f>x
is
:.
$x D,
.
<f>(x,
y)
*1105.
*H06.
<f>(x,y).= x,y.ylr(x,y)t
(cc,y).<t>(x,y).
(x,y,z).<\>tx,y,z).
*ll03.
*1104.
Xx ^y
D y x (x,
.
y)
often useful.
= ;{x);{y).<l>{x,y)
= :{x)i{y z).<\>{x,y,z)
(^,y).<f,(x y). = :fax):fay).<f>(x,y)
(^,y,z).<f>{x,y,z). = :fax):fay,z).(f>{x,y,z)
<l>(x,y),D Xt y.yfr(x y): = :(x,y):<f>(x,y).^.^(x,y)
*1101.
*1102.
>
= :(x,y):<f>(x,y). = .,jr(x,y)
Df
Df
Df
Df
Df
Df
*1107.
possible
y interchanged except in
"<f>(x, y)".
Pp.
SECTION B]
*H1.
(x,
y)
$ (x, y) D
153
w)
(2,
Dem.
h.*101.DI:Hp.D.<y).*(*,y).
OlO'l]
*11'11.
If
(x,
(x,
y)
<f>
</>
w)
(z,
y)
w may be,
is
then
true.
is
Dem.
By #1011,
possible
#1112.
argument
h
y) .p v
:. (x,
<f>
(x,
y)
D p v
(x,
Dem.
H.*10'12.Dh:.(y) v pv$tey).
y)
<f>
(x,
y)
[*10'll27]Db:.(x,y).pv<f>(x,y).D:(x):p.v.(y).<l>(x,y):
D p
[*1012]
This proposition
#1113.
If
<f>
(&, y),
#1114.
: .
jr (x,
(x,
y)
we have
yy
. <j>
{x,
y)
<f>
(x,
y)
:.
Prop
(&,
is
yfr
(x,
y)
(x,
"f
y)
. <j>
. ir
first
(x,
(x,
y)
: <j>
w)
(z,
yjr (z,
w)
Dem.
h.*1014.DH.Hp.D:(y).^(^,y):(y).^(^y)
D
(z, w) yfr (z, w) :. D h
[#1014]
Prop
: </>
is
This proposition, like #1014, is not always significant when its hypothesis
true. *1113, on the contrary, is always significant when its hypothesis is
true. For this reason. #1113 may always be safely used in inference, whereas
#1114 can only be used in inference {i.e. for the actual assertion of the conclusion when the hypothesis is asserted) if it is known that the conclusion is
significant.
#112.
I
(as,
y)
<f>
(x,
y)
(y,
x)
<fj
(x, y)
Dem.
V
#111
(1)
D h {x, y)
D r :. (w, z)
#110711
<f>
(x,
(x,
y).D.<f>(z, w)
y)
. <j>
(x,
y)
(I)
<f>
(z,
w)
(2)
V
Similarly
I .
(3)
(4)
is
z)
<f>
:.
(x,
I
: .
(w, z)
(z,
y).(f>(x,y).D. (w,
.
<f>
(z,
w) D
.
(x,
z) ,<f>(z,
w)
(3)
(x,
y)
(4)
y)
<f>
Prop
w)"
is
"
(y,
x)
<f>
(x,
y)";
it.
MATHEMATICAL LOGIC
154
1121.
I :
(x, y, z) .(f>(x,y,z).
(y,
[PART
x).$ (x, y,
z)
Dem.
[(110102)]
I
::
(*,
[*ll2]
.(y):.(x):(z).<f>(x,y,z):.
=
=
[*112.*1 0*271]
[(H'0102)]
1122.
.(jj)i.(z)\(x).4>(x;y,z)u
.
x)
(y, z,
<f>
(x, y, z) ::
Prop
= .~{(x,y).~(f>(x,y)}
b:(<Rx,y).<f>(x,y).
Dem.
b
10252
Transp (*ir03) D
.
b:(^x,y).<f>(x,y).
=.~i(*):~(ay).^(a?,2/)}
[10252271]
[(1101)]
H'23. b:(^x
y). <f>(x,y).
<
*"
Prop
= .( 3_y,x).<j>(x,y)
l
Dem.
b
*11'22
.Db: K <c, y)
<j>{x,
y)
=
=
=
[*ll2.Transp]
[1122]
1124.
~ {(#, y) ~
~ {(y, #) ~
.
<*>
(x, y)}
<j>
(x, y)}
i:(a,y,^).^(^y,^). = .(a2/,^).^(^y./)
Dem.
[(110304)] b
:: (gar,
:.
(a*)
:.
y, *) :.
[ii23]
=  (ay) . (a* )
[*ii23.^io28i]
=:(ay):.(a^):(a*)^(,y,)
= :. (ay, *,).(*,&*):: 3 Prop
(a*)
4> (>
&>*)'
[(110304)]
1125.
b:~{(<&x,y).<l>(x,y)}.
1126.
r : .
(a) : (y)
I
(a?, y) 3
10128 . D
(1)
:.
101121
(a*)
.(x,y).~<l>(x,y)
:
(y)
(a>
(y)
< (a?,
4> (* 2/)
y)
>
(a*)
< 0*.
^)
C1 )
Prop
Note that the converse of this proposition is false.. E.g. let <j>(x,y) be the
propositional function " if y is a proper fraction, then x is a proper fraction
greater than y." Then for all values of y we have (a#) < (, yX s0 tnafc
the
In fact
(y) (ft)
( x $)" expresses
(y) : (3*) 4* x V) is sa ti8ned
'
>
'
<t>
>
is
false.
1127.
b :.(a*,y) { TKz)^(x,y;z)' s?
(a*) *
to *)
'{'&x y,z).<$>(x,y,z)
>
is
SECTION B]
155
Bern.
K #4*2. (#11*03). 3

i
(a*> y)
(a^)
4>ix y> z)
>
4> {*>
y> *>
(*)
K*4*2.(#ll*03).O
h
(ay* *)
(2)
#(^ u> z)
fr.,(2)
bz.p.O.
*ll3,
(x,y)
<fi
(x,
y)
r=
(x,y)
more
or
(3)
p *D4> (x, y)
Dem.
b
#1021.
>h
:.p
(ayy)
4>{x,y)
[#1 021271]
*11'31.
I
(x,
: .
y)
(a;
. <j>
y)
(a>,
y)
i(x)
(x,
f (x, y)
:p~D
.(y)
:p . D
y)
(x,
y)
^(, y)
< (*,.y>:..
$ (xx y)
>
I"
Prop
ty (#, y)
h.*l0'22.Dh::(x,y).<]>(x,yy.(x,y).ylr(x,y):
=
=
[*10*22*271]
The
proofs of
:..
:.
()
(*, y) :</>(*,
used
is
may
When
be.
we
shall
<f>
(x,
If
<f>
yfr
(x, y),"
as in #10*13.]
#11*32.
b:.(x,y):<f)(x
*11*33.
b :.(x, y)
<f>
>
{x,
y).D.^ (x, y) D
:
y).
.ty (x,y)
(x,
(x,
y)
.
<f>
(x,
y)
(x,
y)
f
(x,
y)
(x y)
t
[#10*271]
#11*34:
:.
(x,y)
:
<f>
(x,
y)
^ (, y)
:.
(x,
y)
<f>
{x,
y)
^ (x, y)
(rx, y).<f>(x,y).
#11*35.
b:.(x, y)
#11*36.
<f>
(z,
: <j>
(x,
y)
D p =
w) D (rx, y)
.
(x,
(rx, y)
(rx, y)
. cf>
(x, y)
^ (x y)
y
3 p
[*1027l*281]
[*10*23*271]
y)
Dem.
b #1 1*1 D h : (x, y) ~>
(x, y) D\
.(l).Transp.DI.Prop
.
<f>
<f>
(z,
w)
(1
1^6
MATHEMATICAL LOGIC
*11'37.
[PART
^.(x,y):(f>(x,y).D.yJr(x,y):.(x,y):yJr(x,y).D.
x (x
y):.
3:(x,y):<f>(x,y).3.
Dem.
x (x,y)
In the following demonstration, " Hp" means the hypothesis of the proposition to be proved. We shall employ this abbreviation, whenever convenient,
in all cases where the proposition to be proved is a hypothetical, i.e. is of the
form "p3q." Similarly "Hp (1)" will mean "the hypothesis of (1)," and
so on.
H.*ll'Sl.Dh::Hp.D:.(*,y):.^(,y).D.^(*,y):^(^y).D. X (*
y)
(1)
Syll.*llll.Dh:'.( y):.^(,y).D.^(*;y).:Vr(*,y).D.x(*.y):
>
[*ll32]
K /n
(1)
l
/ox
(2)
x (x,y):.
D\:.(x,y):<f>(x,y).D.\jr(x,y):^(x,y).D.
x (x,y):
X (x,y)
M
D
2:<f>(x,y).D.
Syll
^:(x,y):<f>(x,y).D.
(2)
Prop
I
y): ^(x,y).
:: (a?,
= .^ (x, y) :. (x, y)
3:.(x,y):<f>(x,y).
*1138.
= x (as, y) :.
= x (x,y) [*U311133]
f (x, y)
\::(x,y):<f>(x,y).D.ylr(x,y):.l:.
X (x
(x,y):<f)(x,y).
*11'39.
>
y).^.y(r(x,y).
*11391. h
::
y)
{x,
y)
(x,
(x,
X
<f>(x, y) D
y)
y)
f (x,
^ (x, y)
=
:.
:(x
>
Dem.
H.*476.
X (x y).^.0(x,y):.D:.
)
(x,
[*3'47
y)
y)
z <)>
(x,
y)
x (x,y):
:<t>(x,y).D.yfr(x,y).x
(x,
y)
:.
x (x,y):
= &> V) :<}>(x y).O.yfr (x, y) X {x,y)::
>
^^:(x y):<j>(x,y).D.f(x,y):.(x
)
=
D
*ll4.
*1111'32]
'
[*1131]
D x (x, y) :.
y):<f>(x,y).D.yjr(x,y). x (x,y)
{x,
3h:.<j>(x,y).0.+(x,y):<f>(x y).D.
=
[111133]
y)
X (x,y)
y):(f>{x>
y).D. x (x,y):.
i(x,y):<f>(x,y).D.'f(x y).
>
Prop
> :: (x, y)
<f>
(x,
y)
yfr
:. (x, y)
X (x, y). =
= .yjr(x y).0(x,y)
(x,y)
(x,y):<f>(x,y). x (x,y).
x (x,y)::
.0(x, y)
:.
D :.
>
Dem.
^.^llSl.Dh::Hi? .D:.(x
[*438.*llll32]
I .
Prop
>
y):.<f>(x y).
>
3:.(x,y):<f>(x,y).
^(x
x (x
y):
X (x
y).
:0(x,y):.
y). = .ylr(x,y).e(x,y)::
#11401. b::(x,y):<f>(x,y).
(x,
I
y)
4> (x,
y)
f (, y)
:.
:.
fax, y)
<f>
(x,
157
= .yjr(x,y):D:.
x (#, y) =
=
*11'42.
SECTION B]
*11'41.
X (x
^ (x, y)
r*ll4^.Id]
V)
>
[*1042'281]
:fax,y):<f>(x,y).v.yjr(x,y)
y)
(x,
yjr
y)
fax, y)
(x,
<f>
y)
fax, y)
yfr
(x,
y)
[*105]
#11421. h
:.
(x,
*ll42
*11'43.
: .
fax, y):<f>(x,y).D.p:
#1144.
: .
(x,
*11'45.
:.
fax, y):p.4>{x,y):
#1146.
:.
fax, y):p.3.<t>(x,y):
*11'47.
:.
(x,
*ll5.
:.
y)
<f>
y) :p
y) v p =
(x,
(x,
<f>
fax) :~{(y)
y)
<f>(x,y)}
y)
(x,
.
<f>
(x,
y)
(x,
<f>
= :~{(x,y)
(x,
<f>
y)
y) v
.
yjr
(x,
y)
[#10'34281]
[*102271]
[*1035281]
y)
. <j>
(x,y)
[*1037'281]
[*103327l]
y)
(f>(x,y)}
Transp #456]
:(x,y).<f>(x,y).D.p
(x,
= :p.D. fax, y)
:
y)
^J*'^
= :pifax,y).<\> (x,
= p
(x,
= fax,y)
:
.~<f>(x,y)
Bern.
h.*lO2o2.D):.fax):~{(y).<l>(x,y)}:=:~{(x):(y):<l>(x,y)}:
[(*noi)]
= :{(*, y)*(*,y)}
I .#10253 . D I= . fay).~<f>(x, y) :
:~{(y) . <f>(x, y)} .
(i)
[(#1103)]
(2)
='fax>y)~4>(%,y)
DKProp
K(l).(2).
#1151.
: .
fax) :(y).(f>(x,y):
= :~ {(x)
fay)
<f>
(x, y)}
Dem.
V
I .
#10252
Transp D
.
fax) (y)
:
<f>
(x,
y)
(2)
#1152.
(1)
= ~ [(x) ~ (y)
(x, y)]
= fay)
(x, y) :.
= (a;) fay) .~<f> (x, y) :.
= =~{() (3y).~0(*,y)}
:
<f>
(1
[Transp]
.
: .
D h :. (x) :~(y)
(#, y)
D h :.~[(*) :~{(y) ^(^y)0D h Prop
[#1011271]
#10253 .Dh:.~(y).<f>(x,y).
<f>
(2)
:.
fax, y)
<f>
(x,
y)
yfr
(x,
y)
Dem.
h
*45162
D
=
h:.~\<f>(x,y).\lr(x,y)}.
H.(l). #111133.
l:.(^,y)
I :
(2)
#11*521. H
:.
{<f>(x,y).yJr(x,y)}:
Transp *1122
.
~fax, y) $ (x,
.
:<f>(x,y)
.D
.~yjr(x,y)
(1)
D
.
y)
= :(x,y):<f>(x,y).D.~y}r(x,y)
(2)
Prop
~f (x, y)
(x,
y)
<f>
(x,
y)
.
(x, y)
>^r(x,y)
yjr(x,y)
[PART
MATHEM&TI'CAIi LOGIC
158
#11*53.
4:. t>,
0aO<fy.
y).
^ %) "^
=? :(!#. 0a? .
Dem.
1 .
*1021271
(a?)
0a?. D (y).. ^t/.:
=r(a).<^.D.(2/).^:.DH.Prop
[#1023]
*11'54.
I
:.
i/ry
(ga;)
0a?
(gy)
fy
Dem.
#1035. 3 h
1*1011 281 ] 3 h
t
:.
(ga>,
: .
0a;
(33/)
y)
ijry
. jfxc
^y =
f*1035]
This proposition
#1155.
:. (gas,
0a?
0a;
(gy)
..
(gy)
^y
:.3> .Prop
fy
is
y)
^y :
(gy)
(gar)
..
0a?
==::
^(a?, #)
(g?)
0a;
(gy)
ijr (a;,
y)
Dem.
f .
#1035
DV
D
D
#1011]
#10281]
(gy )
: .
56.
(gy)
f
:.<gaj)
y)
fix, y)
*jr (x, y).=
(ga?)i 0a? :
(a?,
(a?,
y)
y)
:.
(y)
= $x : (gy)
= (f>x (gy)
yjr (w,,
..0a;
(gy),. 0a?
is
(x)~<j>x
:.
:. (a?) 1.
This proposition
#11
0a;
s :
tyy.
.(*,
y)
*0a?
. ry
Dera.
>
.#1033
:: (a?)
0a;
..
K'*10'38.D.r:.
(y)
:(y)
(y)
<j>x
[*10'113
[#10271]
:. (a;)
:.0a;
^y
fy
#1157.
(1)
(2)
(x)
0a;
=
0y :. =
=
0a>
r :
>
*ll59.
(ga;)
0a;
:.
0a,
Dx
sfrx
(ga>,
:. (a?) :.
0a;
(y)
(y). <f>x.fy
(y) .<f>x.fy:.
(a?)
(y)
<f>x
fy
(1)
:.
yfry
:(x,y).<f>x.fy
^y *T1S6
(2)
Prop
(x, y)
ipy
.[(#1101)]'
y)
(f>x
#424]
0a;
.0y
0y
(a;)
0a;
and
(y)
0y
are
D*, y
Bern.
h .#11*57
D f:s.
0a;
3^ ty
'
[*347.*ll32]
1
[ .
#111
D> :. {x, y)
 *424 3
( 2)
.
f. (3) #101121
.
:.
(x,
:.
'
(a?,
y)
(f>x
r <a?,
y)
0a;
.
: <f>x .
(f>y
Hp (2) D
.
D
:
tyx
0a;
..
yjrx
(f>y
0y D .^x.tyy
$x 0y D
:D
tyy
:
yjry
(1)
y}rx
i/ra;
^y
(2)
(3)
D
y)
l.(l).(4).Dh.Prop
0a;
0y . D  frx
^ry
0a;
D^
^ra;
(4)
SECTION B]
*ll6.
::
I
(g)
:.
(ay)
This proposition
y)^y XXi s

<f>(x,
:*
(32/)
159
(3*) $(* V)
'
Xx t$
is
Dem.
h.*10*35.
[*1011281]
<f>
:
:.
[fcio352813
1161.
'..
[*n23]
I
i.
(ay) :<f>x.Dx
.4r (x,
y)
Dx
<f>x .
Dem.
Dr
Hp D
*ll26
*1037 D
r
:.
[*1011'27]D
I
::.<)
.
.
:.:
(ay)
:. (a?) :.
<*
:.<ay)
(ay)
^(a;,
ty.O + (x, y)
y) : D : <fce 3 (3y) lK* y)
(1)
(2)
h.(l).(2)OI.Prop
*11;62.
I
:: <f>x
^ (x, y)
x>y
% (a?, y) =
:. <f>x
D*
(x,
yjr
y)
3y
x(a?,y)
h.*487.*llll33.D
:: <# . yfr (x, y)
[*1021'11271]
I
DKProp
1163.
:.
(a, y)
(x,
y)
Dem.
r .
*221
*1111 D
I
I
[*1I'32]
[1125]
117.
J : .
:.
(x, y) i.~<f>(x,y)
<f>(x,
y)
>,^r(x, y)u
y).1.^{x,y) :.
3^:~(ay)*<*,y)3s<*,y)^<*,y).3.^(*,yMDh.Prqp
(ayy) v (y, ar) : = (a#, y) $ (x, y)
(a, y)
:
:.
(x, y) ~<*<f>(x,
<f>
y)
: (a?,
y)
< (x,
Dem.
V
*1 1 41
3h
:.
(a*, y)
^ (^, y) v
#<y, ar)
'
C*11 23 ]
[*4'25]
In the
last line of
(a y)
are the
The
same
first
is
made
i> (y,
proposition.
160
MATHEMATICAL LOGIC
#1171.
::
(qz)
fa : (aw)
%w D
:
fa Dz
.
#101 #347
:.
fa Dz
.
yfrz
%w D w 6w = fa xw Dz>w .yfrz.Bw
.
%w
Bw
D fa yw D
yjrz
&w
(1)
(1)
#101 .Dhr.fa. x w
"^z, w ^ z
[#1028]
[#1035]
(3)
h
(3)
Coram #3*26 D
I
(4)
#101121
Similarly
I
::
::
::
(gtt>)
(gw)
(qz)
xw '^
%w D
.
:.
'
<l>
Xw
.0w:
Dzfa .0 .yfrz
}*,
%w D Z)W
.fa.Di.fa. \w D2)W
.
(5)
(6)
Comp D
%w
D z>w
fa
#347
:'.
(4)
%?#
tyz
.D w .0w
(5)
(6)
:: Hp
D
h.(2).(7).DKProp
tyz
^ 0w
2 fa D
^ Ow
.
D
H
:.
\Jrz
Dem.
V
[PART
0w D
:
<z
Dz
\r^
^ty
,"2>
w .0w
(7)
*12.
The
true,"
i.e.
<j>x
<$>x"
But whatever
are true."
mean
function
<f>
"<f>x is
always
may be,
there
be arguments x with which <f>x is meaningless, i.e. with which as arguments $ does not have any value. The arguments with which <\>x has values
form what we will call the "range of significance" of cf>x. A "type" is defined
will
and
yjrx
are significant,
i.e.
some
function.
In virtue of *9*14,
is yfry.
From
if
<j>x,
<f>y,
this it follows
The
by the viciouscircle
These fallacies show that there must be
no totalities which, if legitimate, would contain members defined in terms of
themselves. Hence any expression containing an apparent variable must not
be in the range of that variable, i.e. must belong to a different type. Thus
fallacies
which otherwise
arise*.
As explained
from
by the
<f>a is
we
proposition containing a;
will give
the
name
of generalization to the
process which turns <f>a into (x) . <j>x or (ga?) . <f>x, and we will give the name
of generalized propositions to all such as contain apparent variables. It is
plain that propositions containing apparent variables presuppose others not
.
containing apparent variables, from which they can be derived by generalization. Propositions which contain no apparent variables we call
elementary
propositions^, and the terms of such propositions, other than functions, we call
individuals.
Then
type, or even
is
of variables are relevant; thus the lowest type occurring in a given context
may be called that of individuals, so far as that context is concerned. Accord
is
R&W
11
[PART
MATHEMATICAL LOGIC
162
way
in
so, is
to be secured
We
same
type, since, as
No
etc.
will call
such a function
a,
predi
we may without
by means
of the
taken as argument, provided the right meanings are given to the negations
account of what
clearness,
is
which asserts that the function in question is true with all possible values or
with some value of one of the arguments, the other argument or arguments
(x, y) we shall be able
remaining undetermined. Thus e.g. from the function
<f>
(x).<f>(x,y),
(a#).<0,
y)>
{y)>4>( x>y)>
(ay)<M^y)>
first
SECTION B]
163
Thus
variables.
if
their arguments,
number
<j>
name
(x,
y)
is
a matrix,
(x,
y)
(x,
<f>
We
a proposition.)
is
y)
and we
name
any
As we have already
"<j>
z" is used for any elementary
x" represents any value of any elementary
function of one variable. It will be seen that "<j)lx" is a function of two
variables, namely
! z and x.
Since it contains no apparent variable, it is
a matrix, but since it contains a variable (namely
I z) which is not an in(f>
dividual, it is not a firstorder matrix. The same applies to
a, where a is
some definite constant. We can build up a number of new matrices, such as
Thus
"(f)
<j>
~<jE>!a,
<f>lx
~<f>lx,
D .y\r\x,
(fylxv (f>ly,
Qlx.^rlx,
fylxvtylx,
<)!a;vf
(f>lxvyfrly,
!yv^!0, and
so on.
among their arguments. Such matrices we will call secondorder matrices. From these matrices,
by applying generalization to their arguments, whether to such as are functions
or to such
secondord&r matrix
among
and
its
individuals.
secondorder function
to a secondorder matrix.
A secondorder proposition
y aPPly
is
(2)
is
I
<f>
<f>lx.D x .yfrlx,
where a and
<f)lx.= x .ylrlx,
b are constants,
(x)
<
x,
x,
we may
(g#)
z and yjrlz:
(ga?) .<j>x.fx,
<f>
a D
.
f
x,
is
<f>
where g z
b,
and so on
112
MATHEMATICAL LOGIC
164
[PART
.<j>loc,
(3) Functions in which the argument is an individual x
(<f>)
(3$) <!#,<!# 3$ <M a, where a is constant, and so on.
I z and x:
$!#,<!#. D .<! a,
(4) Functions in which the arguments are
where a is constant, (g//r) < x = yjr x, and so on.
:
<f>
in
definitely,
Thus
a,
<f>
a and
<f>
aD
<j>
a,
!
is
a firstorder matrix,
while the rest are either individuals or firstorder matrices, will be called
predicative
It will
if it is
a matrix.
heading
(3).
Such functions we
functions of individuals have already been defined as being such as are of the
first order.
argument
may be
of
its
arguments.
<f>
generally, a matrix
at least
is
of a function
<f>
f\(4>\%^\%...SD,y,...).
Such a matrix is not of the first or second order, since it contains the new
variable /, whose values are secondorder matrices. We proceed to construct
ot
these constitute thirdorder
new matrices as we did with the matrix
matrices. These together with the functions derived from them by generalization are called thirdorder functions, and the propositions derived from thirdorder matrices by generalization are called thirdorder propositions.
<f>
In this way we can proceed indefinitely to matrices, functions and propoand higher orders. We introduce the following definition:
sitions of higher
function
is
said to be predicative
is
when
all
it
is
a matrix.
It will be
(cf.
pp.
127, 128).
The
is
a primitive idea.
by a note of
letter.
The variables occurring in the present work, from this point onwards, will
be either individuals or matrices of some order in the above hierarchy.
Propositions, which have occurred hitherto as variables, will no longer do so
all
SECTION B]
165
p.
is
made. In practice,
may be regarded
a function of the same order and takes
is
In practice, we never need to know the absolute types of our variables, but
That is to say, if we prove any proposition on the
assumption that one of our variables is an individual, and another is a function
of order n, the proof will still hold if, in place of an individual, we take a
function of order m, and in place of our function of order n we take a function
only their relative types.
of order n
+ m,
for
be involved. This results from the assumption that our primitive propositions
are to apply to variables of
We
any
order.
For
q, r, s) for
functions,
we
variables of the
We
which
is
and
relations,
is
more
convenient in practice.
<f>
<f>
this practice.
It should
individual
fl((f>lz, yfrlz,
...
x, y, z, ...),
assigned.
whose
argument,
(<f>
is
not given.
be made into an apparent variable, unless we suppose its order previously fixed.
As the only purpose of the notation is to avoid the necessity of fixing the order,
such a function will not be used as an apparent variable; the only functions
which will be so used will be predicative functions, because, as we have just
seen, this restriction involves
no
loss of generality.
166
MATHEMATICAL LOGIC
We
[PART
mathematics
in the Introduction,
is
statement of the form "x belongs to the class a." Now assuming that there
is such an entity as the class a, this statement is of the first order, since it
involves no allusion to a variable function.
that
Indeed
its
it is
of the
main reason which makes them linguistically convenient, is that they provide
a method of reducing the order of a propositional function. We shall, therefore,
not assume anything of what may seem to be involved in the commonsense
admission of
classes,
except
this,
is
equivalent,
ments.
Thus a predicative function of an individual is a firstorder function; and for higher types of arguments, predicative functions take
the place that firstorder functions take in respect of individuals. We assume,
predicative function.
.is
SECTION B]
167
we have any use for, and little enough to avoid the contradictions
which a less grudging admission of classes is apt to entail. We will call this
assumption the axiom of classes, or the axiom of reductibility.
We shall assume similarly that every function of two variables is equivalent,
of classes as
the axiom of relations or (like the previous axiom) the axiom of reducibility.
In dealing with relations between more than two terms, similar assumptions
variables. But these assumptions are not
would be needed for three, four,
indispensable for our purpose, and are therefore not made in this work.
. . .
Stated in symbols, the two forms of the axiom of reducibility are as follows:
*121.
h:
Pp
Pp
*1211.
We
call
<f>(x }
y).= x>y
may
yfrx,
and
.yjr(x) y).
of one or
two variables
is
be.
the above two axioms, the first is chiefly needed in the theory of classes
and the second in the theory of relations (#21). But the first is also
essential to the theory of identity, if identity is to be defined (as we have done,
in #13'01); its use in the theory of identity is embodied in the proof of #13101,
Of
(*20),
below.
We may sum up what has been said in the present number as follows:
(1) A function of the first order is one which involves no. variables except
whether as apparent variables or as arguments.
individuals,
(2)
A function of the (n + l)th order is one which has at least one argument
which
is
function or
(3)
i.e.
is
...
or a function of order n.
predicative function
a matrix. It
is possible,
is
without
loss of generality, to
use no variables
Any
is
formally equivalent to a
IDENTITY
*13.
Summary
The
We
q/"#13.
is
identical with
y"
will
be written "x
y."
common
uses
shall find that this use of the sign of equality covers all the
The
definition
as follows:
is
#=y. = :(<):0!#.D.<!> Df
1301.
This definition states that x and y are to be called identical when every
predicative function satisfied by x is also satisfied by y. We cannot state that
every function satisfied by # is to be satisfied by y, because x satisfies functions
of various orders, and these cannot
But in
tyx,
all
where
yjr
is
13101,
i]ry (cf.
would be
if it
x.
Note that the second sign of equality in the above definition is combined
is not really the same symbol as the sign of equality
which is defined. Thus the definition is not circular, although at first sight
with "Df," and thus
it
appears
of
The
them
so.
number
I.e. if
x and y are
13101
This includes
any property of y
is
13151617, which
13191.
I.e.
identical,
)r:x^y.D.ylrx =
1312.
to. Most
The most important
\\x = y."^.^rxZ>^y
13101.
is
number
any property of x
is
a property of
yfry
a property of
if
x and y are
.(f>y
identical
x.
h :.y = x.Dy
y.
and
transitive.
= .<f>x
13195.
I.e.
(gy) .y
= x.<f>y =
to state that
.<f>x
13*22.
This
h
is
(32,
.<f>(x,y)
is
SECTION b]
x = y.
*1301.
The
= :(<f>):<f>lx.'2.<f>ly Df
embody abbreviations which
following definitions
x^y. = .~(x = y)
*1302.
#1303.
x = y = z.
*131.
\:.a}
*13'101. h
169
identity
Df
= .x = y.y = z T)f
= y. = :<f>lc.3t.<l>ly [*4'2
x = y D yjrx D <^y
.
(*1301)
(*10'02)]
Bern.
V
J
#121 D
*131 D
.
:.
(a<)
::
Hp D :.
:.
yfrx
x D*
<f>
.</>!
.
#
.
<
fy =
.
I
<f>
(1)
:.
^y =
[*4'8485.*1027]
D :. tyx =
[*1023]
cf> !
<f>
y D* ^a; 3
:
i/ry :.
(2)
H.(l).(2).Dh.Pro.p
In virtue of this proposition,
= y, y
satisfies
is satisfied
We
by
x.
which x and y agreed. Strict identity would, in this case, have to be taken as
a primitive idea, and #13101 would have to be a primitive proposition, as would
also
*13151617.
#1311.
\.:x
= y. = i<l>lx.=4
>
.<l>ly
Bern.
h
#1022
<
>
D:x = y
[*131]
K #13101.
I .
*13'101
#17
(1)
3\:.x = y.3.<l>lx3<f>ly
D\:.x = y.D.~<j>lx1~<f>ly.
[Transp]
.<j>lyD<j>lx
(3)
h.(2).(3).Comp.Dh:a; = y. D .<f>lx=<j>ly:
[*10'ir21]
DH:.# = 2/*D:<!#.=*.0!y
K (1)
$1312.
(4)
x=y D
#13'101
<*}rx
(4)
Prop
ijry
Bern.
[Transp]
x = y .O
[#13101
yjrx
*1314.
yfrx. f^yfry
#1315.
)r
,x = x
[Id.
#1316.
\x y. =.y x
[#1311. #1032]
yfry
D x^y
.
Comm. Imp]
#1313.
[#1313. #414]
#1011. #131]
(2)
170
MATHEMATICAL LOGIC
[PART
#1317.
Dem.
V
#131 D
::
Hp D
:.
[*103]
<f>
:. <
x D*
# D^
.
<j> !
::
y D$
Prop
.
<f>
z.\.
functions of
<j> !
<f>
z.D.y = z
[*1316\l7]
= x.z = x.0.y = z
[*131617]
h:x = y.x^z.D.y^z
[#1317 .#414]
h:x = y.y^z.D.x^z
[#13171 #414]
\:.x = y.D:z = x. = .z = y [#1317172 Exp. Comp]
h :. x = y = z = x = z z ~ y
*13172. Y:y
#1318.
*13181.
#13182.
#13183.
Dem.
h.
h
3:x = y
[*13*15]
h
(1)
(1)
(2)
D K Prop
(2)
#1319.
#13191. h:.y
= x.D y .(f>y: =
.<j)x
Dem.
r
#10'1
y = x Dy
:.
tf>y
D x=x D
D <f>x
[#1315]
(1)
This proposition
#13192. h
:.
(gc)
(2)
is
x=
<f>x
(1)
^.<f>y:.
D.<f>y:.
[#101121] D\:.<j>x.D:y
= x. D y .<f>y
(2)
Prop
=x x =
.
sjrc
yjrb
Dem.
.D\::fb.D:.x=b.= x .x = b:^b:.
D:.(^c):x=b.= x .x = c:yfrc
.#101 .0\ :.x = b .= x .x c zyjrc :"D :b = b = ,b = c:\lrc:
[*5501.*1315]
D:6 = c.^c:
f .
#42
#32
[#1024]
"
f
D:^/r&
[#1313]
D h :.
K(l).(3).Dh.Prop
h
(1)
(2)
#101123
This proposition
is
(gc)
x= b = x
.
(2)
x = c ^c
:
yfrb
(3)
SECTION B]
#13193.
IDENTITY
:
<f>x
=y =
.
x=y
<f>y
171
Dem.
h
Simp
Dh:(f}x.x = y.D.x = y
K #13*13.
.
This proposition
h :(f>x.
(4)
D.<f>x.x = y
(4)
Prop
is
xy =
.<f>x. (f>y
This proposition
:
= x.
D.<f>x.y
^1.
(3)
(3)
*1316.Fact]
#13*195. b
(2)
(3)
#13194.
(1)
3\:<f>x.x = y.1.<f>y
(1) (2) Comp .D\ .<j>x.x = y.'D.<l>y.x = y
#1316 Fact
Dh (j>y .x = y .D <f>y .y = x
=y
.x
[#13*13
#4*71]
is
(ay) .y = x
<f>y
,<f>x
Dem.
h
#3*2
#13*15
D\:<f>x.D.x = x.<j)X.
[#10*24]
#1313 #1011 D
[#10*23]
D
:.
:.
\
The use
#13*196. h
(1)
(2)
:.
~(f>x
Dy.y^x
i(f>y.
= x w = y DZjW
.
< (z,
w)
[#13*191]
(2)
[#13*195
is
very frequent.
Transp #10*51]
.
.<f>(x,y)
= z = x Dz w = y O w
= .w = y.D w .<f>(x,tv):.
= <f>(x,y)::D\ .Prop
.
<\>
(z,
w)
:.
This proposition
h
(1)
Prop
.<j>(z,w):
[#13*191]
#13*22.
r/
h :.z = x .w = y .D ZiW
Dem.
h. #11*62.3
::
#13*21.
D.(ny).y = x.<f>y
(y) y = as <\>y D <f>x
= * cf>y D <\>x
(ay)
:
(a^, w)
is
=x w=y
.
< (z,
w) =
(f>
(x,
y)
Dem.
h
#11 *55
: .
(a^,
w) z = x
.
w=y
<f>
(z,
w)
[#13*195]
= x (aw) .w = y .<f>(z,
(a^) z
(aw) .iv = y.<f>(x,w):
[#13*195]
4>(x, y)
This proposition
is
:.
I
iv)
Prop
It
is fre
The
to
show
following proposition
that, if
we have
is
<f>avo*><f>a,
is
any argument
then
for
which
"<f>x" is significant
"
<f>a
Its purpose is
" is significant,
i.e.
for
which
is
either
172
MATHEMATICAL LOGIC
[PART
that, if "<j>a"
In the following proof, the chief point to observe is the use of #10*221.
There are two variables, a and x, to be identified. In the first use, we depend
upon the
fact that.
of
both
<f>a
been
in
$a and x = a both
justifies
the two
x = a in both
"x = a"
is
#13*3.
ambiguous
typically
the occurrence
(Unless the a's had been already identified, this would not be
x's.
legitimate, because
and
a's,
<**><f>a
.D
is
if
neither x nor a
by the
justified
fact that
of
is
both
<f>a
(6).
:.<f>xv <f>x
= :x a.v .x^a
Dem.
h.*2*ll.
Dh.^vv^B
(1)
h.(l).Simp.
(2)
f.*211.
Dhzx = a.v.x^a
H. (3). Simp.
r.
I .
#13*101
(4)
(5)
I
<av~<a. D .<j>xv~<f)x
(3)
I
:.
</av~^a, D :x = a.
Comm Dh
:.
<av~0a D :x = a D
*1013*221
v.x^a
(4)
(5)
<f>xv<^><f>x
K (7). Simp.}
I:: <j>av oj(jxi
t i
(6)
(8). #6*35
<f>x
v~<jkc
f
::
:.^v~^a. D
<f>av<^^>a
:.
# = a v a?4=a
.
<f)xv<fix
(8)
=:#
(7)
= . v. #={= a::
D
Prop
DESCRIPTIONS
*14.
Summary of #14.
description
explicitly, "
is
where
<f>&,"
<& is
more
etc.," or,
some function
satisfied
by one and only one argument. For reasons explained in the Introduction
III), we do not define " the x which satisfies <f>$," but we define any
" The term x
proposition in which this phrase occurs. Thus when, we say
which satisfies <f>x satisfies yjrx," we shall mean " There is a term b such that
That is, writing
<f>x is true when, and only when, x is b, and tyb is true."
" (ix) (cf>x) " for " the term x which satisfies
yjr
(ix)
<f>x,"
(<f>x) is to mean
(Chapter
(a&)
This, however,
<f>x
=x x = b
.
yfrb.
is
is
when
(ix)
(<f>x)
doubt
"^(ix)(<f>x).'
(a&)
or
<px
= x x == b
.
{'&b):.<f>x.= x
If " (a&)
<f>x
= x x = b " is
false. Thus
.
second must be
The
yjrb
D .p
.x = b:'fb.Z).p.
false,
it is
the scope of
(ix)(<f>x).
scope of (ix)
(<f>x) is yfr
Thus
(ix)
yjr
(<f>x),
be called
two propositions, the
p.
Thus we
[(ix) (<f>x)]
f (ix) (<j>x).D.p,
[(ix) (<f>x)]
i/r
is
(ix)
(<f>x)
.D.p.
e.g.
aJr(ix)(<f>x)
~ [a = (ix)
(<f)x)}
we
shall
have
=
( a &)
fa .= x
= ~ {(g6) $sc = x
:
x = b a^b,
x = b a = b).
!:
MATHEMATICAL LOGIC
174
Of
these the
first
We
does not.
<f>x
[PART
.= x .x =
b,
put
*1402. El(ix)(<f>x).^:fab):<f>x.= x .x =
Df
This defines " The x satisfying <f> exists," which holds when, and only
when, <f>ai is satisfied by one value of x and by no other value.
:
[(ix)
(<j>x),
[(ix)(<f>x)]
(<f>x),
[(ix)
(ix) (yjrx)}
(fx)] ./{(ix)
(<j>x),
(ix) (fx)}
Df
It will be shown (#141 13) that the truth value of a proposition containing
two descriptions is unaffected by the question which has the larger scope.
Hence we shall in general adopt the convention that the description occurring
first typographically is to have the larger scope, unless the contrary is expressly
Thus
indicated.
e.g.
= (lx)(yfrx)
fab) <f>x .= x .x = b:b = (ix) (yfrx),
= x x = b :. fac) tyx .= x .x = c:b = c.
:.
<\>x
fab)
(lx)(<f>x)
will
mean
i.e.
*1404.
[(ix)
(fx)] ./{(ix)
(cf>x),
(ix) (+x)}
[(ix)(^x) (ix)
}
Whenever we have E
argument
to
(ix)((f>x),
any function
in
(<f>x)]
Df
which
it
may
This fact
occur.
is
embodied in
*1418.
That
: .
is
everything.
the present
(ix)
to say,
(<f>x)
when
(ix)
(x)
(<f>x)
fx D f (ix) (<j>x)
.
exists, it
King
(<f>x)
for
example,
or not bald.
If (ix)
(<f>x)
it
must
exist.
This fact
is
stated
in the proposition
*14'21.
h:^(ix)(<f>x).D.El(ix)((f>x)
This proposition
of
"
yjr
(ix) (4>x)."
is
When,
is,
by the
definitions, part
said to "exist,"
SECTION b]
poems
"),
descriptions
which
is
(<f>x).
(<f>x)
175
"
be significantly applied to subjects immediately given i.e. not only does our
definition give no meaning to " E ! x," but there is no reason, in philosophy, to
suppose that a meaning of existence could be found which would be applicable
;
to
among
*14 202.
From
:. <f>oc
the
*14204. r
(ix)
(cf>x)
=b =
:
(ix) (<j>x).
exists
when
a 6)
there
(ix)
(<f>x)
it
(f>x
=x
=x =
:
(ix) (<f>x)
follows that
=b
something which
is
(ix)
(<f>x) is.
have
*14 205.
I.e.
first
We
=x x = b =
r
(ix)
yjr
(ix)
(<f)x)
(<j>x)
(36)
= (ix) (<f>x)
yfr when
tyb
there
is
something which
is (ix) (<f>x)
yjr.
We have to prove that such symbols as " (ix) (<j>x) " obey the same rules
with regard to identity as symbols which directly represent objects. To this,
however, there
is
we
= (ix) (<ja),
((f>x)
only have
*1428.
r
(ix) ($x)
(ix)
= (ix) ($x)
(<f>x)
if (ix) (<f>x)
exists.
#1413.
*14'131.
I" :
a = (ix)
(<j>x)
(ix) (<f>x)
(ix) (<px)
= (ix) (tyx) =
.
i.e.
(<f>x),
we have
=a
(ix) (yjrx)
= (ix) (<f>x)
*1401.
*1402.
*1403.
is
proved in #1414'142144.
= :(Rb):<l>x.= x .x = b:fb Df
=
E\(ix)((f>x).
:('g b):<j>x.= x .x = b
Df
[(ix) (4>x), (ix) tfx)] .f{(ix) (cf>x), (ix) (yjrx)} =
[(ix)(<f>x)].f(ix)(<l>x).
i
*14"04.
[(ix)
(W]
Df
Df
[(ix) (4>x)]
[(ix)
MATHEMATICAL LOGIC
176
#141.
:.
[{ix) {<f>x)]
(ix)
yjr
(<f>x)
(36)
<f>x
[PART
.
=x x = b
yjrb
[*42.(*1401)]
#14101.
I
#1411.
#14111.
= x x = b tyb
:. E
(ix) (<f>x) = (36) :<f>x.= x .x = b
=
:. i(w) (yfrx)] ./{(ix) (<f>x), {ix) (fx)}
=
(36, c) <f>x
x x = b
:.
f (ix)
(<f>x)
(36)
(f>x
[#141]
[*42
(*14'02)]
is
yfrx
= x x = c :/(&, c)
.
Dem.
#42
::
(*14'0403)
[(ix)
{{ix) (yfrx)]
[#141] =
=
[#1155] =
[#141]
#14112.
=* ^ = b:f{b, (ix)(^x)}
:.
[(M)ty*)]
:.
:.
(36, c)
H :./{(?>)
<j>x
(a 6 )
**
.= x .x = c
:.
yjrx
(<H> (w)(^)}
(36, c)
<#
(w) (^)j
:.
:.
= x .x = b:f{b, c) :.
:/(&, c) :: D h Prop
= x .x = b
<#
=x x = b
.
yfrx
= x x = c :/{b, c)
.
[Proof as in #14111]
p. 174,
#14113. h [{ix) (fx)] ./{{ix) (<f>x), (ix) (fx)} = ./{() (<f>x), (ix) (f *)}
[#14111112]
This proposition shows that when two descriptions occur in the same proposition, the truth value of the proposition is unaffected by the question which
.
#1412.
:.
{ix)
{<f>x)
<f>x .<f>y
Dx>y .x = y
Dem.
Dl:.Hp.D:(a6):^aJ.s.a? = 6
K*1411
K
V
:. (}>x
= x .x = b D
:
[#13172]
h
(1)
(2)
<$>x
$y = x>y .x = b .y=b.
.
Dx y .x = y
3 h :. (36) <f>x s. x = b D $x
Dh.Prop
x = b $0 = a oc c D b = c
:.
4>x
=x
K(l).(3).
.
#101 .Dhz.Hp.D:
Bern.
<&.
\
.& = &:#>.
Di<f>b:<f>b. = .b = o:
[#1315]
D 6 = c:.DH.Prop
[Ass]
= x x=b = <j>x .Ox x = b
:. (frx
= :<f>x .Dx .x = b: (30) <#
.
<f>y
D XtV x = y (3)
#14*122.
(2)
#101123
#14121. h
: <f>b :
.6
= c:
SECTION b]
descriptions
177
Bern.
V
#10'22
Dh
:. <j>x
=x x = b =
.
[*13'191]
K*47l.
DK
[*101127]
(2)
<f>x .
<j>x
"2
Z>*
x=b x=b
x=b
<f>x
(1)
:. <f>x
D x x = b fax) .<f>x: =
.
<f>x
Dx x =
.
(2)
<j>b
(3)
DKProp
K(l).(3).
The two
the analogy with #14122, but they are not used until
of couples (#55 and #56).
*14'123. H
<&
D,
\
.#532 3
:.<#.
[*1 0281]
[#13195]
h
w) = z>w z = x
:. < (z,
w=y
we come
to the theory
:<f>(z,w).
:<f>(z,w).
. <j>
(z,
w)
Bern.
Y .#1131
Y :.<f>(z,
=
=
w) .= ZyW z = x .w = y
.
[*1321]
K*471.
[#111132]
<f>
<f>
=.<f>(x,y)
#532
n\ /ox
r .(l).{6)
#14124.
:.
fax, y)
"
#111
w) = ZyW .z =
.
'{'Spc,
#347
y)
h
:.
(f>
.<f>(u, v)
(z,
w) = z>w .z = x.w
: <j>
(z,
<f>
<f>
(x,
y)
(1)
(u, v)
=x w=y u=x
D.z = u.w = v
D
=y
(2)
w) = ZtW .z = x.w = y:
.
= y:
D:<f>(z w).<f)(u,v).D ,z =
R&W
w)
K (2). #111135. D
fax, y)
w)
(z,
[#13172]
:.
x.w = y:
#327.D
r
(3)
Dem.
V. #14123.
\<\>{z,
'
(2)
(z,
:.
^ L Prop
x>
Jb.
I
<f>
[#11341]
(2)
01322]
I .
:<f>(z,w).
u.w = v
(3)
12
MATHEMATICAL LOGIC
178
[PART
h.(3).*ll'll3.D
\'..(^x,y)i^(z
#111
:.
<f>
(x,
y)
w)
(z,
<j>
<j>
. <j>
v)
(it,
<f>
.
0:<f>(x, y)
[*533]
: <j>
[*1 4123]
D.(j>(z >
:
(f>
(x,
y)
<f>
(u, v)
Oz w,u,v .z = u.w = v
(4)
.z = u.w = v:
z>Wl u, v
(z,
w)
(z,
w) D z> w z = x
< (x,
y)
Zt
=x w=y
w = y:
w.z
(5)
(6)
h.(5).*iril'3445.D
I"
= (a#> y) >4>(x,y):4> 0, w)
(4)
(6)
*1413.
a=
(1)
K Prop
(ix)
Dem.
h.*14l.
(<j>x)
<f>
(u, v)
=a
(ix) (<f>x)
b:a = b
=
=
:b = a
.=
x=b
b
:. <f>x
= x x=b :a
h *1316 *436 .1
<f>x
x
[*10'11281]
D :. (a&) :(f>x.= x .x = b:a = b:
i'Sfi) <f>x =z b
= :(ix)(<fix) = a
[*141]
.
D\:.a=(ix)(<f)x).
\
:('3b):<f)X.= x .x =
:
(1)
I
'
(1)
(2)
(ix)(<j>x)"
is
=a
(2)
Prop
This proposition
"a =
'
is
*14131.
(ix)
(<j>x)
= (ix) (yfrx) =
.
(ix) (yfrx)
= (ix) (<f>x)
Dem.
*141
:.
= (ix)(yjrxy';
The
eliminated.
but in
as
The above
#141 11 D
proposition
t
" (ix)
may
also
(ix)(*jrx) is to
be
first
be proved as follows:
[#43.*1316.*l 11 1341]
[*112.*14111]
tyx)
SECTION b]
#1414.
179
descriptions
a=b b=
.
(ix)
Bern.
V
#141 D
Hp D
: :
: .
(g&) :'<f>x.= x .x
= b:a = b:.
(gc)
[#1035]
Dem.
#14111
(ix)
(<j>x)
I
::
= (ix)(yfrx)
Hp D
:.
(ga, b)
(gc, d)
:.
(ga)
[#1154]
=x
:.
(ga, c)
<f)X
[#14111]
D:.(ix)(4>x)
= (?#) (<f>x)
:.(ga,e): <#
(ix)
(<f>x)
= (ix) ()(x)
x= c xx *' x ~ c
\
'"
= x x = a yfrx .= x ,x = a:
.= x .x=c: ^x = x .x = c :.
= x .x = a x# = x .x c a = c
.
:.
= (ix)(xx)::1\.~Proy
= (?#) (yfrx) D
yfrx
4>x
= x .x = b :a = b :.
xx x x = d c = d :.
x = a yjrx. = x .x = a :.
,= x .x = a:
^c = x x = c
[#14121. #11*42]
= (ix) (\Jrx) :.
<j)x
yjrx
= (ix) (xx) D
(ix) (yjrx)
(gc) ityx ,= x
[#13195]
#14145.
= x ~x = c
[*327.*13195]
\ .
[#14121]
#14144.
<px
[#13'195]
= (ix) (yjrx)
(ix)
(<f>x)
(g&)
<#
=3 x = 6 a = 6
Dem.
V
#141
I
:.
= (ix) (<f>x) =
(1)
#141
I
::
Hp =
.
=
D
[#1035]
[#14111]
#1415.
=:<f>x.= x .x = a
[#13195].
H
(ix)
:.
r
=x x = a
(gi)
(1)
:.
ipx
^a;
:.
(g&) :.<f>x.= x .x = a:
yjrx
:.
a,
x=
a=6
:.
.= x .x = b:a = b:.
I .
Prop
(<j>x)
Dem.
h.*141.0
h::Hp. D :. (gc) <j>x = x .x = c:c = b :.
[*13195]D:.^.= x .a; = 6
K (1) #141 D
:: Hp
D:.yjr {(ix) (<f>x)} = (gc) :x = b.=x x = c
=:^6::Dh.Prop
[#13192]
:
I
#1416.
(ix)
:.
(1)
(cf>x)
y]rc
Dem.
\ .
#141
#14
D
D
:.
Hp D
::
fyx
(g&)
(f>x
=x x = b D
.
=x x = b
.
= (ix) (yjrx)
(1)
X (O) ( Wl = (a c ) x = & =* * = c %c
= X
[#13192]
:.
:
l>
(2)
122
180
r
MATHEMATICAL LOGIC
[PART
*14\L315 D
.
K(2).(3).
r :.b
(3)
(4)
H.(l).(4).*10l23.DKProp
*1417.
I" : .
(ix) (0a?)
=b =
.
yfr I
.^.yjrlb
(ix) (0a?)
Bern.
K*1415.*10ir21.D
H:.(?a?)(0a?)&.D:^!(?a?)(0a?).=*.^!&
\ .
(1)
(ix)(<f>x)
=*
&
0 !
= 6. =
D:(?a?)(0a?) = 6
D:(?a?)(0a?)
[*13'15]
.&
= &:
(2)
h.(2).Exp.*1011'23.D
b::(^ X):x^=x^ = b:D:.yjrl(ix)(<l>x).= .ylrlb:^.(ix)( (fix) = b
(3)
ri/
K*12l.
D\:(<Kx ):
(3)
(4)
(1)
(5)
X lx.= x .x = b
3 H :. yfr (?a?)(0a?) =*
D h Prop
.
(ix) (0a?)
~E
(4)
&
(ix)(<f>x)
=b
(5)
for, if
0
(ix)
0
(<f>x),
(ix)
all
*14171.
I
values of
(ix)
:.
(<f>x)
(ix)
not have
0
(<f>x) is
0
holds for
we do
=6 =
D+.yjrlb
(ix) (<#),
always
D^
(<f>x)
false,
0
and therefore
But we do have
b.
=6 =
.
0>
D*
0
(?a?) (<f>x)
Bern.
K*14\L7.
3
#101 *121 D
I .
I
:.
(ix) (0a?)
:.
= 6. D
D,,,
0<
:0!&
(?a?) (0a?)
D^,. 0
:
[*1315]
h
(1)
(2)
*101
(?a?)(0a?)
=6 D
.
D:(?a?)(0a?)
*1418.
[Fact]
: .
(?a?)
(a?)
I
:.
D
D
0a?
0a;
D
D
(a?)
yfrb
=3 x = b
.
0a;
0a?
= b:
:D:<f>x.= x .x = b:yJrb:
D :(>&b) <f>x .= x .x =
[*10'35]
[*14'111]
The above
formally)
all
Hence when
(2)
(?a?)(0a?)
0 (ix) (0a?)
(a?).<0a?:
=&
(x)
[*10ll28]D\:.('3b):<j)X.= x .x
=6
Prop
(0a?)
(1)
(m?) (0a?)
(a?)
0>a?
yfrxz.D
0
yjrb:.
.= x = b
yfrb :.
(g;6)
(?a?) (0a?) :.
0a?
Prop
it
has (speaking
(<f>x)
exists,
it is
it
occurs.
SECTION B]
#14*2.
DESCRIPTIONS
(ix)
181
(xa) = a
Dem.
h
#14101
"5
= a) = a. =
(ix) (x
:.
[*13*195]
h
(1)
*14'201.
Id
(1)
Dh.Prop
H:E!(?a;)(^).D.(aar).^
Dem.
h
#14 202.
#1411 D
Hp D
:.
[*10\L]
[*1315]
:. <f>x
=x x = 6 =
.
#14"1
Dh
(ix)
:.
(<j>x)
=6 =
.
(gc)
<f>x
D h Prop
.
: .
=b x
(px
(ga?)
<f>x
b *=
ar
:<#.=,,..#=&:.
is
(ix) (<f>x)
=b =
(a&) . 6
(ix) (<f>x)
[*13195]
#14*203. h
(g6) :<j>x.= x .x = b:
<f>x
<f>y
a?
=c
= (?a?) (<#)
=b
Dh.Prop
first half.]
DXt y x = y
.
Dem.
h
#1412201
#10'1
3b:.El(ix)(<f>x).D:('zx).<f>x:<l>x.<f>y.3Xty .x = y
I
:. <f>b
<f>x
<f>y
x>y
(1)
.x = y:D:<l>b:<f>x.<f>b.Dx .x = b:
[#533]
0:<j>b:<f>x.Dx .x = b:
[#13191]
Di^&.D*.^:
[#1022]
.D x .x = b:
0:<f>x.=x .x = b
(2)
(f>%
I .
(2)
#10128
[#1035]
D x>y .x = y:0
(fib)
D.* = y:D
(Kb):<l>x.= x .x = b:
:.
[#1411]
K(l).(3).
#14204. h
= x x = b :.
(g&):<6 :<j>x.<f>y.
Dh:.( a &).<&:<^.<^.
<f>x
El(ix)(<f>x)
(3)
Dh.Prop
:.
(ix)
(<f>x)
(g&)
(w) (0a>) = 6
Z)em.
h. #14202. #1011
h
:.
(6)
:. <]>x
[*10281] h
=x x = b = (ix) (<f>x) = b :. D
:. (
3 6) <f>x = x x = b = (g&) (?#) (^) = b
.
h.(l).*14\Ll. Dh.Prop
#14205. h
#1421.
yjr
(ix) (<f>x)
(36) . b
= (ix) ($x) ^b
.
[*14202'1]
b:ylr(ix)(<f>x).O.El(ix)(<f>x)
Dem.
#141
:. yfr
[#105]
D
D
[#1411]
D E
{(ix) (<f>x)}
.==x
.x = b:yfrb:
(36)
(gi) :<l>x.=x .x = b:
<f>x
(?#) (<#)
:.
D h Prop
.
(1)
[PART
MATHEMATICAL LOGIC
182
This proposition shows that if any true statement can be made about
(ix) (fa), then (ix) (<f>x) must exist. Its use throughout the remainder of the
work
be very frequent.
will
When
(ix)
(<f>x)
but
"(ix)(fa)" occurs,
still
it
proposition concerned
is
which
(ix) (fa),
yjr
is
i.e.
the asserted
#1422.
I
(ix) (fa)
Dem.
#14122.
1 .
(1)
#471
[*1011281]
.<(>
(ix) (fa)
D\:.fa.= x .x = b:D.cf>b
D b :. fa = x x b = <j>x = x x = b <f>b :.
D :. (g&) fa = x x = b = (g&) <f>x .= x .x=
Prop
D h E l(ix) (<}>x) = .</> (ix) (fa)
:
I
[*1411\L01]
As an
'
OK
is
(1)
proposition
(f>b :.
"The
following:
equivalent to 'the
man who
"the man
*1423.
(fa tyx) . =
(ix)
a false proposition.
is
{(ix)
<
(fa
tyx))
Dem.
J .
#1422
:.
(ix)
=
[*10'5.*326]
h
^rx)
[(ix)
(<f)X
yjrx)]
(f>{(ix) (<j>x
yjrx)}
sjr
</>
{(ix) (<f>x
yfrx)]
(ix)
{(ix)
(fa
^rx))
(1)
3:<f>{(ix)(<f>x.yjrx)}
OK
O K Prop
#1421
K (1)
(<f>x
(fa ^x)
(2)
(2)
Note that in the second line of the above proof #105, not only #3*26, is
required. For the scope of the descriptive symbol (ix) (fa tyx) is the whole
.
product
<
[(ix) (<f>x
yjrx)}
i/r
{(ix)
(g&)
fa
(fa
fx)},
tyx
that,
becomes
first line
.
so
.= x .x = b:<f>b.
>Jrb
fa
yjrx
~x
x=b
<f>b,
<j>{(ix)(fa.yjrx)}.
i.e.
#1424.
(ix)
(fa).
[(ix) (fa)]
<t>V
=y V = O) (fa)
Dem.
h
#141 D
.
I
: .
[(ix) (fa)]
= (a 6 ) '.$y.= y .y = bi$y.= y .y = b:
:
183
descriptions
SECTION b]
s (36) :<f>y .= y .y = b:
= E (?a?) (<M : 3 K Prop
[*424.*10'281]
[*14'11]
I : .
(ix)
(<f>x)
= y y = O) (<M
<f>y
Dem.
K #14203. 3h::Hp.3
.<f>y.<f>x.D.y=x:.
[Exp]
[#101121]
(j>y
3H::Hp.3
(j>y
[#471]
[#13191]
<jty
[#1022]
:.
<j>x
<f>x
3 y = a; "
.
"D x
y=x
:.
= : <f>y:<]>x.D x .y = x:
= :y = x.D x .<}>x:<f>x.'2 x .y = x:
= :<f>x.= x .y = x:
= :y= (ix) (<f>x) :: 3
Prop
[#14202]
#1425.
3
D
(?a?) (<#)
:ir&.
Dem.
h
#484 . *1027'271
3H
:: <j>x
=x x = b 3
.
3* tyx
:. <f>x
=6
a?
[#13191]
=.:^r6:
[#14242]
=.
(1)
#101123 3
.
:.
(36)
<Jxb
=*
a?
=b
3
: .
D x fx =
<#
ir
(?#) (<#)
(2)
31. Prop
H. (2). #1411.
#1426.
>*
(ix)
(<f>x)
(ga?)
<#
.~.<f>xD x
f {(m?) (<#)}
fx
Dem.
#1411. 3
H:.Hp.D:(a6):^B.= x .af = 6
#10311 3 :: <j>x = x x = b 3 :.
3 :. (gar)
[#10281]
:
I
I .
CO
<f>x
4>x
~x x = b
= (ga;)
yjr x
yjrx
[#13195]
=.yjrb.
[#14242]
I .
I
(2)
:.
#101123
(36)
1 .
(1)
#1427.
<#
(3)
:.
tyx
a?
:.
= 6 fx
.
^ {O) (<HJ
(2>
= . x = 6
#1425 3
(ix) (4>x)
(g#)
Prop
:
<f>x
<#
= x ^x. =
\}rx
(ix)
yjr
(<j>x)
(3)
{(?#) (<#)}
= (ix) (fx)
Dem.
3b
#101127 . 3
J .
#48621
(1)
I
::
</>*;
[#10271]
[#14202]
[#14242]
I .
(2)
::<#.
#101123 . #1411
x=b
x=6
OK Prop
=:(?a;)(<^) = (?)(^)
(2)
:.
'
184
MATHEMATICAL LOGIC
#14271. b
:.
fa = x ,yfrx D E
:
(ix) (fa)
= E
.
[PART
(ix)
(fx)
Dem.
b. #4*86.
[#101127]
D
D
::
fa~
::
Hp
: .
[*101121]
::
Hp
..
fa = x
.
yjrx
[*10281]
b
[*10'271]
#14272.
^a?.
Dem.
h.*4'86.
D h ::<# = yfrx.D fa ,=.x=bi=: yjrx = x=b
[*10ir414]Dh::Hp.
D :. <a;. =^.a;=6 = ^ra;. = x .x = b :.
[Fact]
D :.fa.= x .x=b:x^> = ^x =* a; = 6 ^6 :.
[#101121] Dh::Hp.
D (b) 0a; =x a; = b ^6 = yjrx .= x .x=b:xb:.
[#10281]
D :. fab) fa = x x = b xb =
: .
'
: .
:.
[*14'101]
: .
Prop
The above two propositions show that E ! (?#) (<a;) and x( lx)(fa) are
extensional" properties of fa, i.e. their truthvalue is unchanged by the
substitution, for fa, of any formally equivalent function yfrft.
"
*14'28.
b:El(ix)(fa).~.(ix)(fa) = (ix)(fa)
Dem.
b #1315 #473 D b
K(i).*ioir28i.:>
.
:.
fa = x x = b = fa = x x = 6
.
=6
(1)
(2)
l.(2).*14lll.Dh.Prop
This proposition states that (ix) (fa) is identical with itself whenever it
but not otherwise. Thus for example the proposition "the present
exists,
King
of France
is
the present
is false.
The
of
argument
ever concerned.
[(ix)(fa)].f{ x (ix)(fa)}.
is
185
descriptions
SECTION b]
This
we have
elsewhere, and
proofs.
*W3.
\z.p
= q Dp q .f(p) =f(q) E
(70)
x (70) (00)} =
[(ix) (*)]
f{\(ix) (00)]
(00)
Dem.
h
*14'242
=x x = b
<f>x
h.(l).DH:.p S 5.D
r : .
*14242
:.
00 = x x = b D
.
(1)
/{[(w) (Ml
% (70) (00) = X 6
./(p)=/(j):^c.5..6:D:.
[(70) (00)]
xO*)(M
= /(%*)
(2)
[(10) (0*)]
(3)
r.(2).(3).Z>
I
:.p
I .
The
(4)
*10'23 . *14 11
=6
[(10) (00)]
(4)
Prop
They
are,
: :
(70) (00)
: .
[(?0) (00)]
p v x (*#) (<H
= p v
:
[(?0) (00)]
x () (<H
Dem.
.*14242
I .
I
(2)
The
(1)
'
(3)
hence we shall merely give references to the propositions used in the proofs.
*1432.
: .
(70) (00)
[(70) (00)]
~X
( lx )
=
[*1 4242
(<K>
~ {[(70) (00)]
X (M?) (<f>)}
The equivalence asserted here fails when ~ E (70) (00). Thus, for example,
let <f>y be " y is King of France." Then (70) (00) = the King of France. Let
XV be "y is bald." Then [(70) (00)] .~%(70)(00) = the King of France
!
exists
and
King
is
not bald
but
~ {[(70) (0#)]
is
bald.
Of
X ( lx ) (4*)} =
these the
it is false
first is false,
that the
MATHEMATICAL LOGIC
186
[PART
biguous
as the
but
it
is
is
am
equivalent
*1433.
: :
(ix) (fa)
:.
[(?*) (fa)]
*14331. V
::
(ix) (fa)
:.
[(ix) (fa)]
% (ix) (fa) Op
= [(ix) ((fix)] x X) (<\>x ) 3P
.
[*484
*14 332.
::
*14'242
(ix)
*1023 *1411]
(fa)
[*486
*14'34,
t : .
p =
.
[(ix) (fa)]
x (ix) (fa)
[(ix) (fa)]
X () (fa) =
[(**) (fa)]
(ix) (fa).
Devi.
V
*14
:.p
[(ix)(fa)]
[*10'35]
I
[*14*1]
q,
~p, p D q and p
it is
.
q.
or (b) the proposition in which (ix) (fa) has the smaller scope
#14'34, and is the reason
! (ix)
(fa). The second case occurs in
(ix) (fa).
The
implies
why we
proposition in
SECTION
CLASSES AND RELATIONS
*20.
Summary o/#20.
The following theory of classes, although it provides a notation to represent
them, avoids the assumption that there are such things as classes. This it does
by merely defining propositions in whose expression the symbols representing
we
The characteristics of a class are that it consists of all the terms satisfying
some propositional function, so that every propositional function determines a
class, and two functions which are formally equivalent {i.e. such that whenever
either is true, the other is true also) determine the same class, while conversely
two functions which determine the same class are formally equivalent. When
two functions are formally equivalent, we shall say that they have the same
extension.
objects.
occurs
<
may
value, upon the particular function ^, or they may depend only upon the
extension of <f>. In the former case, we will call the proposition concerned an
intensional function of </>; in the latter case, an extensional function of <.
for
example, (x)
<f>x
or (g<c)
is
<f>x
<f>
it
by no means
mark
<j> !
<f*x:
The
is
\x .= x .^lx: D
:/(< ! 3) . = ./(^ ! S).
<f>
"
when we wish to speak of the function itself as opposed to
(We write "<f> !
its argument.) The functions of functions with which mathematics is specially
concerned are
When
all
extensional.
a function of
<$>
is
extensional,
it
may be
regarded as being
about the class determined by <f> 1% since its truth value remains unchanged
so long as the class is unchanged. Hence we require, for the theory of classes,
a method of obtaining an extensional function from any given function of a
function.
This
is
effected
by the following
definition:
188
MATHEMATICAL LOGIC
f fi^z)}. = :(<&<!>):
#2001.
Here
{<
/ { ($z))
in reality a function of
is
/ { (yfrz)}
regard
as though
which
yfrz,
<f>
had an argument
defined whenever
is
But
t.
Df
<f>lx.= x .fx:f{<f>l$}
[PART
it is
convenient to
This
((f>z)
it
= z (yfrz) =
.
(x)
<$>x
(yfrz),
yfrx.
With regard to the scope of 2 (yfrz), and to the order of elimination of two
such expressions, we shall adopt the same conventions as were explained in
#14 for (ix) (<f>x). The condition corresponding to
E
which
is
<f>
x =x
.
yfrx,
xez
"x
to express
is
member
(^frz)
by
yfrz."
We
therefore
#2002. x e (0 2) =
x Df
In this form, the definition is never used
!
<f>
it is
proposition
H
xe2
:.
(yfrz)
(g<)
^y = y
<f>
<\>
by the help
We
classes,
xez
(yfrz)
yfrx
of #1 2*1.
Greek
letters (other
than
e, i, it,
(<f>z)
or z
yjr,
(f>,
(<f>
%, 6) to represent
When
z).
a small
cerned.
is
The use
<j>
(<f>z)
or z
(<f>
z)
intolerably cumbrous. Thus "x e a" will mean "x is a member of the class a,"
and may be used wherever no special defining function of the class a is in
question.
The
#2003.
Cls
= o {( a <). a = z(<f>
z)}
What
=z
(<f>
*)},"
is
meant by a
class.
Df
= 2(0
z)}" has
no meaning in
such expressions.
{(g<). a
symbol "a
what
;;
SECTION
C]
189
type.
thus
<f>
is
definitions, is
conventions,
ambiguous as to
is to
represent the
is
convenient
Thus the type of " Cls " is fixed relatively to the lowest type concerned
but if, in two different contexts, different types are the lowest concerned, the
meaning of "Cls" will be different in these two contexts. The meaning of " Cls"
only becomes definite when the lowest type concerned is specified.
of
a.
Equality between classes is defined by applying #13*01, symbolically unchanged, to their defining functions, and then using #20*01.
The propositions of the present number may be divided into three sets.
we have those that deal with the fundamental properties of classes
these end with #2043. Then we have a set of propositions dealing with both
First,
classes and descriptions; these extend from #20*5 to #20*59 (with the exception of #2053 54). Lastly, we have a set of propositions designed to prove
that classes of classes have all the same formal properties as classes of in
dividuals.
In the
#2015.
I.e.
first set,
:.
two
yjrx
= x %? =
.
: .
1.0. two
members.
#2043.
z (^z)
(\Jrz)
#2018.
I.e. if
:.
two
This
is
and z (xz).
(<f>z)
./{z (yjrz)}
also to the
is
= p. = :x6a.= x .x@
in place of z
other.
This
= z (xz )
is
z ($>z)
\:.a
This
is reflexive,
symmetrical
transitive.
#203.
(<j>x)
'59),
it satisfies
we show
may be
that,
the defining
under suitable
MATHEMATICAL LOGIC
190
first set,
"
i.e.
[PART
Here it is to be remembered that "a" stands for "2(<f>2)" and that "fa"
therefore stands for " f {2 (<f>e)\." This is, in reality, a function of <pz, namely
the extensiOnal function associated with f(yjrlz) by means of #2001. Thus
an expression containing a variable class is always an abbreviation for an
expression containing a variable function.
we prove
all
43),
= :(^)'.^\x.= x .^rx'.f{j>\z} Df
#2001. f{z(^z)}.
x e (<f>
#20*02.
Ob = 3.{<a0).
#2003.
The
z)
Df
Df
=*($!*)}
x,yea. = .xa.yea
Df
#2005. x, y,zea. = .x, yea.zea Df
Df
#2006. x~ea = ~(xea)
*20'04.
The
#20071.
= .(<j>).f{z(<t>lz)}
(go) ./. = (30) /(* (0
#20072.
[(?a) (<)]
#2007.
(a).fa.
#2008.
#20081. ae
The
ir
&
/< <0)
.
(30)
*)}
(a*/)
fa =.
.
. <f>
. r !
Df
Df
0 = = 7 fj Df
Df
a :/(0 &)
Df
I
:./{$(**)}
#201.
#2011.
I" : .
s <30)
:
:<\>\x
.= x .tyx:f{4>\$]
#486
'
(#2001)]
::
Hp D
.
:.
= x fx =+
as .
D:.<f>lx.= x
[#10281]
:.
(g0)
[#436]
<f>
x = x x '=x x
as
=*
X /{0
.
,.
.fx :/{0
lx.= x fx :/{<f>.l z)
z]
= (g0)
[#2011
[*42
fx = x xx D / (2 ( Wl = / I* (X*))
Dem.
\
:.f{z(irz)\
ee
.f{z( X
z)}
<f>
*}
'
:</>!
x. =..
=> ^op
Xx :/[0
lz}:.
SECTION C]
191
depend
determining the
class,
for
*20lll.hi.f($l2).=t.g(<f>l2):3:f{2(<j>lz)}.=t.g{2($lz)}
Dem.
KFact.
[*10281]
[#201]
r .
(1)
*101121
20112.
:.
Prop
*)}
{z
(<f>
*)}
Dem.
K (1)
I
#2013.
f : .
1
(g<)
fyx
of reducibility
still
(<f>
*)
The meaning of
#2001 to #1301, remembering the convention that
scope than 2 (%z) because
Dem.
#20 1 3
; :
[#201]
I .
it
occurs
= 2 (Xz) s
2(tyz)
to have
a,
= ! a; < ! 2 = 5 (^^)
= :.(^4>,0);^x.= x .<f>\x:xx^ x .nx:<f>l2 = 0l2
2 (yjrz)
(g<)
: .
larger
^a;
(1)
::
Hp 3 :. (g$>)
.
(1)
^a
3 :. (a0, 0)
(2)
#2014.
: .
:.
=x
l& yx .= x .$>\x\.
. <f>
2 (ifrs) = (^)
tyx
=x x x
.
[#10322]
.:.
(2)
Prop
Dem.
#201 3 h :: 2(^s)2(#er) = :. (g) : ^a?. =,..^
= :. (a<, 0):.^x.= x .<f>lxzxx.= x 01x:
[*20l]
[#13195] = :. (g</>) :, yjrx = x < x xx =* < * :
h
is
first.
#121 #10321 3
[#13195]
f .
(1)
x =x .
<
=a,
i^a;
This proposition
=x
is
^"
">*"
a?: <f>l2
= 2( Xz)
$12=0
:.
:.
Prop
#2015.
MATHEMATICAL LOGIC
192
#20151.
(g<)
= (0
2 (^z)
[PART
z)
Bern.
>.*2015.
3h:.ylrx.= x
Db
[*10ll28]
J.
(1).*1 21.
:.(%<!>):
.<f>lix::D.'z*(ylrz)
= 2(<f>lz):.
(1)
Dh. Prop
all classes
when
classes are
used as apparent
variables.
fore,
members,
its possible
its
is
i.e.
significant.
*2016.
(a</>)
*2017.
(<)
#2018.
h:.2(<K> =
*2019.
\
:.
.f{H<f>
(<f>
3 ./{2 (yfrz)}
*)}
WM
*0K>3:/
$ {&) = z { X z)
Dem.
.#2018 .*1011'21 D
.
[*20'12]
z))
[*20'16
"
#101]
./{*(*)} [*201115]
:.$We) = 2(x*) ?
(/):/!*(**). D./!*(x*)
h
(1)
/!*(*!*). D./!*<0!*)
(2)
D
h::^!a;.s..^:^!*.s,. X*:.(/):/Ii(^).D./!^(x*):3
I .
(2)
*10ll2733
D :.<!#.=*.<! a D :<!#.=*.
D :.<!#. =*.#!#:.
[*20112.*10l]
[#42]
[#1030132.Hp]
D :. 2 (>K> = *(%*)
[#2015]
l.(3).*10ll2335.D
(3)
h::(a^,^:^!*.s..^:^!*.s..X*:.(/):/^(^)3/^(X*)
D.*'W*)S( X*)
h.(4).*12l.Dh:.(/):/!^(^).D./!t(^):3.t(^) = a(^)
DH.Prop
h (1) (5)
h
:.
#20 191.
2(^) = * O) = :(/) :/! ty*) s ./! 2( X *)
.
[#201819
#202.
(4)
(5)
#2015
(<f>z)
K (1)
#1022]
= 2 (<*)
D H :. 2 (<f>z) = z((f>z) =
Prop
#42 #1011 D
.
r
<f>x
<f>x
(1)
SECTION
C]
#2021.
#2022.
[#2015 #1032]
 2 (^)
(<f>z)
2 (<fz)
= 2 ( %*) D
.
193
.
2 (*)
= a (%*)
[*2015 #10301]
.
The above
for a reason
f{z (<f>z)}
is
= y"
= ^( Xz).D.z(yfrz) = ^( X z)
=
=
z(<j>z).z'( X z)
b:2(ylrz)
%(<l>z).D. z(^z) = 2( Xz)
h :.a = z((j)z).= a
= 2tyg) = .z(cf>z) = z(fz)
#2023.
\:z(<f>z)^^(^z).^(<f>z)
[#202122]
#2024.
#2025.
ct
[*20'2122]
Dem.
h
#101
:.
= z(<f>z) .= a
.a
= 2(yjrz) D
z(<j>z)
= 'z((f>z). =
.^(<j>z)
= ^(fz):
D:^(<f>z) = ^(yjrz)
b #2022
D
a = (<*) (<^) = (^) D a = (yfrz)
[Exp.Comm]Dh:.a(^) = a(^).D:a = ^(^).D.a = t(^).
h. #2024.
3 h :.$ (<}>z) = z (yfrz) a = z (yjrz) .D a = 2 (cf>z)
[Exp]
D h %{<j>z) = 2{fz) D a = (^) D a = (?)
h.(2).(3). Dh:J(^) = ^(^).D:a = t(^). = .a = J(^)
[*101121]
Dl:.f(^) = t(^).D:a=f(^).=.a = t(^)
[*20'2]
\
(1)
(2)
:.
:.
(3 )
: .
Dh.Prop
\;xe%{tyz). =
(4)
l.(l).(4).
#203.
.tyx
Dem.
h.
#201.3
r : :
xez
{*\rz)
[(#2002)]
[#1043]
[#1035]
=
=
=
[#121]
(a</>)
iry .= y .<\>\y.^x
:.
yjrx::
#2032.
\.x~{x6z(<!>z)}=$(<l>z)
#2033.
b:.a =
(yjrz)
Prop
member
#2031.
:.
..
.= .(a<^>): sjry.^y.^ly^yjrx:.
by
yjr
[#20153]
[#20315]
Dem.
h. #2031.
^\:.a
h. (1). #203.
DK Prop
(1)
R&W
jo
[PART
MATHEMATICAL LOGIC
194
#20*34.
y.='.xa.Dayea
[~:.cc
Dem.
h
#4*2
(#2007)
I
x e a D a y e a : = x e Z (<f>
: .
z)
D* y e 2 (<f>
.
z)
D^.^ly:
=:x = y:.Db.Frop
=:</>!#.
[#20*3]
[#13*1]
The above
apparent variables.
= y. = :xea.= a .yea
a e Cls = ( 3 0) a = %(<]>
#2035.
\:.x
#20 4.
[*20"3
z)
[#20*3 .#1311]
*2041. h.f(^)eCls
#2042. H z (z e a) = a
(#2003)]
[*20'4151]
Greek
the form 2
letter,
(<f>z)
such as
a, is
Dem.
h
#203
#1011
D
D
[#201 5]
#2043.
:.
I
:x eztyz) ~ x
fx
= (^#) D
[x e z tyz)}
Prop
The
a
to the contrary, adopt the convention that the descriptions are to have
#1401
and
#2001.
definitions
the
applying
in
classes,
the
than
larger scope
#205.
(ix)
(<f>x) e
z (^z)
= .^{(ix)
z (tyz)
(<*)}
Dem.
h
#141
:: (iae) (<f>x)
[#203]
[#141]
#2051.
r
:. (7a?) (<jix)
= b. =
(ix)
=
=
=
(<f>x)
(a c )
.
'
(go)
4>x
<f>x
.= x .x = c:cez (yjrz)
= x x = c tyc :.
.
:.
::DK Prop
.yfr{(ix)(<f)x)}
ea.=..6ea
Dem.
#2053
:.
(ix)(<j>x) ez(yfr
[#1011] h
z)
:. (las) (<j>x)
.b ez(yfrl z)
= &
.
[#1417]
*2052.
: .
(i)
(</>#)
= (a&)
:
=
=
(') (*)
(far)
O) (<H e a
(<K>
=
.
=&: 3
= 6
($x)
=
f
r
<f
:.
Prop
:.
(g&)
(ix) (<\>x)
h.(l). #14204.
#2053.
This
l:./3
is
=b =
.
Dh. Prop
= a.D0.</>/3: =
.</>a
(a&)
(ix)
=a
&ea
(1)
SECTION
C]
195
Dem.
K*101.
[*202]
h
:.
D:0a
=
0a
a
D
D
/3
0/3 :.
.
[Comm]
DH:.0a.:>:/3 = a.D.0/3:.
D\:.(f>a.D:/3^a.D
(1)
{ft/3)
(2)
/3
This proposition
(1)
[*101121]
h
#2054.
*2()1821
fi
(2)
.(f>^
DKProp
= a.
0/3
= 0a
.
is
Dem.
h
#2018 #1011 D
.
K #20'2
#32
z (0*)
0/3
Dp. 0a
I
0a D a = a 0a
(1)
K(l).(2).
I .
=a
Dh:(g/3)./3 = a.0/3.D.0a
[*1024]
#20 55.
/3
I :
[#1023]
= (?a) (xea.= x
(2)
<f>x)
Dem.
*20'56.
*2057.
:.
(?a)
.Dem.
h
#141
::
Hp =
.
[*2054]
:. (
a /3) /a = a a = /3
= :./a.= a .a = (00)
.
(00)
:.
(1)
K*141.
0\:.g{(ia)(fa)\.^:(^):fa.= a .a = /3:g^
(2)
h.(l).(2).3H::Hp.D:. r{(ia)C/a)}.= (>&&):a = z( (j>z).= .a = {3:g/3:
= ( 3 /3).(00) = /3.#/3:
[*13183]
= g{z(<f>z)} ::Dh. Prop
[#2054]
ll
fi
K 2(00) = (7a)
#2058.
[a
= (00)}
Dem.
= f(00) = a .a = z(<j,z)
[#2054]
:. (
a /3) a = (0*) = a a =
h $ (0^) = ( ?a ) { = 2 (0*)} D h
[#141]
#2059.
h 2 (0^) = (id) (fa) = (la) (fa) = z(cf>z)
h
*42
#1011
D
D
D
I
: .
2 (0s)
Prop
= /3 :.
Dem.
f .
#201
[#1413]
[#201]
:.
z (00)
= (?a) (/) =
.
=
=
(ftyjr)
:^>x.= x ^lx:yjrlz=:
(g^r)
0a;
(7a)(/a)
=,
i/r
(7a) (fa)
(7a) (fa)
= i/r
= 2(00):.Dh.Prop
132
MATHEMATICAL LOGIC
196
[PART
the
all
classes as individuals
is
only necessary to
we
shall
proofs.
original propositions
h:(aa)./a. = .~{(a).~/a}
#206.
Dem.
3
=
h:(aa)./a. .(a0)./{3(0!*)}.
[(#1001)]
= ~[(*).~/{*(* **)}].
= ~ {(a) </}
Prop
[(#2007 )]
h
#42
(#20071)
This
is
#2061.
:()./. D.//3
Dem.
h
OK
#101
(#2007)
I
(a) ./a
D ./{z (<f>
z))
Prop
r:()./.D./{0K>}.
This
We
is
be,
This
is
(got)
(yfrz)
= a.
#2041.
When
#2062.
may
#2017.
need further
This
/3
is
then
//3
is
true,
Dem.
h
be,
#1011
then
#2063.
This
(cf>)
./{z((f>
z)} is true,
z)) is true,
i.e.
\:.(a).pvfa.D:p.v .(a).fa
is
is true.
<j>
may
SECTION
197
C]
Dem.
h
#4*2
(#20*07)
h:.(a).j>v/. = :($).pv/{*(*!*)}:
= p. v. ()./{$ (*!*)}:
= p v (a) ./a D H Prop
[#1012]
[(#20*07)]
#20*631.
and vice
significant,
This
is
: .
same type
of the
/3 is
as
a, " fft" is
versa.
Dem.
By #20151,
$ 2.
function of
/r
z.
is
Similarly
#20*632.
some
for
If,
and vice
a,
there
is
and
of the form z
/3 is
to
z and
</>
therefore,
(yjr
t/r
and
z),
by #20*01, fa
is
ffi is a function of
is
a function fa,
versa.
Dern.
By
f{z (yjr
^)] is
a function of
i/r !
Hence the
"Whatever
#20*633.
class /3
may be"
possible class a
#20*64.
whatever possible
/3
inter
fi)."
is
/3) is
may
/3.
\:.(a).fa:(a).ga:D.f/3.g/3
Dem.
#4*2
:.
(#2007) D
(a) ./a
(a)
ga
(<j>)
.f{z (<&
*))
[#10*14]
(<)
^{S (<
{2(yjr !*)}:.
*)}
.
Prop
Observe that "/3" is merely an abbreviation for any symbol of the form
(ir ! z). This is why nothing further is required in the above proof.
The above
it
proposition
is
it is
significant, it
same
type.
This
/
is
and g should be
not required for
significant.
#20*7.
This
H:(asr):/a.= B .$r!a
is
#20*701. V
[#20112]
a <7) :/ [z
(<f>
z),
x}.=^ x .g
[z
(</>
*),
x\
is
is
MATHEMATICAL LOGIC
198
#20 702.
:/{x, H (<j>
<jr)
z)}
~^ x g
{*,
2 (<
[PART
*)}
[Proofasin*20'701.]
#20 703.
*),
a (*
z)}
=*,* . g
{2 (<
z),
(yfr
z)}
Dent.
\.*lO3ll.D\:.f{ x
<j>lx= xX lx.
(1)
h.(l).*lln3341.D
h:.Hp(l).
[*20l.*1035]D:/{a(</)!^),^(t^)}=*^..9J{^!^^^}
(2)
h.(2).*10ll281.D
Yi.{^g):f{ X
(W) f{*
(<t>
*)
(*
*)}
=** 9
#2071.
This
{*
(0
*)
* (*
(3)
*)}
[#2019]
is
all
From
This
be
much
is
possible.
The two
The
"type" of any object x will be defined in #63 as the class of terms either
identical with x or not identical with x. We may define the "type of the
arguments to
i.e.
the class
that
<f>z"
&(</>&
as the class of
</>#).
if "(f)(i" is significant,
arguments x
Then the
first
for
which
"<f>x" is significant,
is
if
"<a"
and
"yjra" are
is
the type of a;
both significant,
because each
is
the type of
a.
is
&
SECTION
*20 8.

199
C]
<f>av<>><f>a
(<f>xv~<f>x)
ob{x = a.y/.x^a)
Bern.
K*133.*101121.D
:: Hp. D :.(f)x v~<#. = x :x = a.v .x^a:.
[*20'15] D :. (<f>xv~<f>x) = x~(x=a .v .x^a)::D\ Prop
I
H :</>av~^>a.T^av~i/ra.D.^(^>a?v~</>a?)
*2081.
= ^(ira;v~'^ra;)
l.*20'8.D!:Hp.D.^(^v~^)=^(aj = a.v.a;4=a)
*20*8 D h Hp D $ (yfrxv~y}rx) = cb (x = a v x a)
h (1) (2) *10\L2M3 Comp D
I .
I
Hp
[*2024]D
(1)
(2)
=)=
D.(<j>xv~<fix)
Prop
In the third line of the above proof, the use of #10121 depends upon the
fact that the "a" in both (1) and (2) must be such as to render the hypothesis
significant, i.e. such as to render
significant.
in (1)
*1013 we can
(2)
Since a type
is
if <f>x is
a function
which is always true, z (<f>z) must be a type. For if a function is always true,
the arguments for which it is true are the same as the arguments for which
hence z
it is significant;
(<j>z) is
if (x)
<f>x
holds.
Thus any
may
be,
<f>
*21.
Summary
The
q/"#2I.
definitions
and propositions of
stood in extension:
it
may
ty (x,
y)
is
its
true.
(x,
y) for which
determining function.
We
^ (&,
{/) is
put
f{H(x>y)} =
This
is
:(R<f>):<t>l(x,y).= x
Df
shown that
It will be
that two relations, as above defined, are identical when, and only when,
they are satisfied by the same pair of arguments.
i.e.
For substitution in
when a
for the
arguments a and
b is to
and a
for y.
where a
#2102.
{<f>
(&, ))
&.
= .</>
Df
(a, b)
a{4>l($,x)}b.
.<}>l(b,a)
= .<f>l (a, b)
This definition
is
{<j> !
{tf> !
(x, )}
(, tb)}a.
not used as
it
<
(b,
a)
Df
Df
Df
a 0)
(x,
y)
=x> y .^(x
y):
<f>l
(a, b)
SECTION
C]
201
which results from #21 01 02. We shall use capital Latin letters to represent
variable expressions of the form &y<\> (x, y), just as we used Greek letters for
!
z).
(22)./R.
The use
= .(*)./{^!(*
Df.
y)}
is
a practically
indispensable convenience.
The
following
#2103.
Rel
is
= 5{(a0).i2 = ^!(a?,y)}
it
Df
letters,
notation
the notation
t(
xRy"
The
will
and
practically convenient,
is
will, after
relation
R to y."
This
{scp<f>(x, y)} y.
number
they are exactly analogous to those of #20, merely substituting #1211 for
#121,
for propositions in
#10.
The propositions of this number, like those of #20, fall into three sections.
Those of the second section are seldom referred to. Those of the third section,
extending to relations the formal properties hitherto assumed or proved for
individuals and functions, are not explicitly referred to in the sequel, but are
#2115.
I.e.
first
i.
+ ix, y) =*,.*(*, y) = x$^(x, y) = $$ X (x, y)
two relations are identical when, and only when, their defining functions
#2131.
I.e.
:. xfiyjr
(x, y)
= xy X {oo,y). =
two relations are identical when, and only when, they hold between
The same fact is expressed by the following
#2143.
\:.R
relations is
reflexive,
symmetrical and
transitive.
#213.
I.e.
r
.+ (x, y)
two terms have a given relation when, and only when, they
defining function.
satisfy its
202
MATHEMATICAL LOGIC
*21151.
I.e.
[PART
= x$<j>\{x,y)
V .(>&$). &y\r(x,y)
we have a relation
as apparent variable,
no
*2ioi.
On
to &,
Hence when,
loss of generality.
Df
/{^^(^y)}.:(a0)":0K^y)^ir^(^y)=/{0K^)}
the convention as to order in #21 01 '02,
cf. p.
so that
fm^(x
#2102.
a{<f>l(x,
#2103.
B,e\
The
Df
y)}.
>
:('
= B{(>a<l>).R = $g<f>l(x,
Df
y)}
little
modifi
#2107.
(R) ./R
.f{x<j>l(x, y)}
.(<]>)
#2108.
Df
Df
<\>R
=R
R = 8 :/S
Df
(R,
#21081.
Df
{<j> I
The convention
#21082. f{R(fR)}
#21 083.
Re
#211.
<f>
as to typographic
.
Df
:./{xpf
(x,
Df
R = .<! R
y)}.
here retained.
is
a 0)
<f>
(x,
* (x, y) :/{</>
y) .=,,.
(fi,
v)}
[*42.(*2101)]
#2111.
:.
ir (x, y) .=*,.
X (x,
[#48636. #10281.
y)
D :f{%H (x,
y)}
= ./(% (*.
V))
#211]
#21111.
b :.f{<j>l(x,y)}
^.g{<f>l(x,y)} :
[Fact #11113 #10281 . #211]
.
#21112. h
:.
D :/{*#$ l(<c,
(^)
#1211,
(x,
y)}.=*.gl
[%H
(a, y)}
.g{%$<j>\(x,y)}
</>) :.
<
(x,
[*121 #21111]
.
=*
:./{</>
It is #121, not
#2112.
y)\
y)
.= x>y
^ (x, y) \f[M}+
namely
(<c,
y)}
(f>,
we
although that
./{</>
(x, y)}
[#2111. #1211]
This is the
*20701702703.
#2113.
first
\:.y}r(x,y).= Xt y.x(^y)^
$t 0>
y)
= *X 0> V)
SECTION
*2114.
203
C]
:.
=*,j/
% 0* 2/)
[Proof as in #2014]
#2115.
h:.^(*,y).s (M ,.x(*.y):s^^(*y)^X(
fl?
[*2H314]
y)
This proposition states that two double functions determine the same
relation when,
by the same
pairs of arguments.
$f (x, y) =
#2116.
l:(a*):/{^(*,y)}.s./{^!(*,y)} [#2112]
#2117.
h :(<}>). fm<l>l(x,y)}.1fl$H(x,y)}
#2118.
(a<^>)
:.
a$4> (x, y)
are satisfied
[#2115 #1211]
#21151.
i.e.
is
(x,y)
[#2116. #101]
[#211115]
#2119.
*2ii9i.
^%
..
#212.
#2121.
#2122.
h:
#2123.
1
(/) f *Q *

(*,
y)
(a?,
y)
y^ (a, y) = y
</>
h:x{xyyjr(x,y)}y.
(,
y)
M)x ( x
>
y)
2]
*10'32]
[#212122]
= $00 (*> y) D
= .^(x,y)
M)^{x,
:.
= %(*, y) =
y)
[*212122]
</>
#2131.
[#2115
y)
&9lr(w,y) = &9x(*'y)
#213.
%vx (*.
><
%f(^y) = ^ X (^,y)
#2124.
/'
[#211819]
[#2115 #4
:oc
is
i/r
{$pyjr(x,y)}
[#21153]
= &y> (#, y)
[#21315]
#2132.
#2133.
y [# {&y^ (#,
y)} y]
>
[#21313]
written for some expression of the form xfity (x, y). The use
of a single capital letter for a relation is convenient whenever the determining
Here
function
is
irrelevant.
#214.
#2141.
#2142.
#2143.
is
Rel . =
$$<t> (x,
y) e
(g<)
.R = $g<f>l
Rel
h.$P(xRy) = R
\:.R = S. = :xRy.= x
(x,
y) [#203
(#2103)]
[*214'15.1]
[#21315]
.xSy
[*21'153]
204
MATHEMATICAL LOGIC
*2153.
*21'54.
h:.(^S).S = R.<f>S. =
*21*55.
I
#21*56.
#21*57.
$y<j> (x,
E (iR) {xRy .= x
y)
$$<!> (x, y)
Coram .*10'ir21]
#141]
:.
[*101 .*21*218*21
.<l>R
[PART
[#21*55
= {iR)(fR) .D:g
= g
.
#14*21]
{(iR) (fR)}
Vixy<j>(x,y)
The
= (iR){R = $cy$(x,y)}
[#4*2 .#1011
ff.,
.*2154 #14*1]
.
purpose.
#21
6.
#21*61.
#21 '62.
may
a J?) fR. =
m
~ {(R)
~/R]
[Proof as in #20*6]
be, (R)
.fR
y)
is
\:.(R).p v/R
#2163.
#21631. If
significant,
#21 632.
"fR" is
and vice versa.
If,
for
possible relation
is
[Proof as in #20*631]
some R, there
is
is
a function
[Proof as in #20*632]
versa.
"Whatever
#21*633.
f{R, S)
(x,
[Proof as in #2062]
is true.
possible relation
S may be"
may
be,
f(R, S)
is
true whatever
may
S may
be,
be."
[Proof as in #20*633]
#21*64.
\:.(R).fR:(R).gR:D.fS.gS
h (g#) :/R = R g ! R
#21701. \:(Rg),:f(R,x).= B x .gl(R,x)
#21*702. h fag) :f(x, R) = BjX .gl(R, x)
#21*7.
[Proof as in #20*64]

[Proof as in #20*7]
[Proof as in #20*701]
#21*703. r
#21*704.
(Rg)':f(R,
(g#) :f(R, a) . = Rt
(R, S)
[Proof as in #20*702]
[Proof as in #20*703]
a)
[Proof as in #20*703]
[Proof as in #20703]
I
#21705. h
#21*71.
g (R,
!
[Proof as in #20*71]
From
all
a relation or not.
for classes
and
so on.
CALCULUS OF CLASSES
*22.
Summary o/*22.
In
this
symbolic
logic.
startingpoint of
0) are always to
stand for expressions of the form cb
x),
or,
where
the
Greek letters are
(<f>
not apparent variables, ik (<j>x). The small Latin letters may either be such as
have a meaning in isolation, or may represent classes or relations; this is
</>,
*2201.
This defines
#2202.
avj/3
/3,"
/S's."
common
/3.
= $i(a:ea.v.xel3) Df
sum
of two classes
#2204.
put
ar\P=*&(xea.xe&) Df
#2203.
We
it is
all
all
a = x(x~ a) Df
which
"
below
a,
" is
xea
meaningless.
Thus
it
#2205.
but
it
a/3 = an~j3 Df
is
often convenient.
The
avb
is
in the class
whenever a and
16.
anb
is
in the class
whenever a and
II
a.
There
is
an element
such that a u
A=a
for
II
6.
There
is
an element
such that a
V=a
for every
Ill
a.
Ill
b.
r\
a,b,
avb
and
a, b,
ar\b and
5,
u a are
n a are
July 1904,
every element
element
in the class.
in the class.
p. 292.
a.
a.
MATHEMATICAL LOGIC
206
IV a.
= (avb)r\(av c) whenever a, b, c, a u b, a u c, b n c,
a\j(br\c)
and (av
IV b. a c\
(b
b)
r\(au
= (a r\b) \j (a r\
c)
c)
whenever
a, b, c,
and
a = V and
There are at
VI.
r\ b,
r\ c,
u c, an (6 u c),
is
and 116
an element
exist
and are
such that
a = A.
ar\
least
in postulates II a
a u (b r\ c),
c)
If the elements
c)
and (anb)yj (a n
V.
[PART
y, in
the
class,
The form
pendent,
i.e.
nullclass
I a, in *2237,
namely
"h
/3 e
Cls
in *2236,
namely
"h
/3 e
Cls
"
"
b,
II
a,
in #24*24,
namely
"KowA = a"
II
b,
in *2426,
namely
"h
III
a,
in *2257,
namely
III
b,
in *2251,
namely
"Kowj3 = j8ua"
"h .ar\ fi = fi n a"
IV a,
IV b,
in #2269,
V,
in
in
VI,
an
V= a"
"
n (a w y) = a w (/3 n 7)
"
#2268, namely " h (o o /3) u (a n 7) = a n (^ u 7)
#242122, namely "h.ano = A" and "r. aw _a = V"
namely
"h
(a
yS)
namely
in #241,
"r
A+V"
it
(1)
Kar = y8na
h.au/8rs0ua
These embody the commutative
#2257.
#2252.
(a
/3)
#227.
(a
/3)
n 7 = a r\
u7=au
(/3
(/?
law.
n 7)
u 7)
A,
A, V.
The
#2251.
r\,
=a
a u a= a
a
r\
*2256. h
These embody the law of tautology.
.
number
for
SECTION
C]
#2268.
#22
69.
CALCULUS OF CLASSES
n ) u
(a
(a
n 7) = a n (# u
207
7)
These embody the distributive law. It will be seen that the second
by everywhere interchanging the signs of addition and
multiplication.
#22 8.
I
This
is
 (
a)
=a
h:aC/3. = ./3Ca
#22 81.
This
(2)
is
h:aC/3./3C 7 .D.aC7
*22441. h a C ft. xea.D.xeft
These embody the two forms of the
#2244.
#22
62.
#22621.
syllogism in Barbara.
\:aCft. = .auft = ft
h:aC/3. = .an/3 = a
to transform
any inclusion
C ft)
is
excluded
which
a."
*2201.
#2202.
#22 03.
#2204.
a
#2205.
aft
#221.
\:.aCft.
#222.
I
=x(x~ea)
=an~ft
Df
Df
= :x a.Dx .xeft
[*42
(#2201)]
[*20'2
(*22'02)]
#223.
= % (x a x e ft)
y.avft = x(xea.v.xeft)
[#20*2
(#22*03)]
#2231.
K = ^(^
[#20*2
(#2204)]
[#20*2
(#2205)
ft
e o)
#2233.
h.aft = $(xea.x~ft)
\:xar\ft. = .xea.xeft
#2234.
\:.xea\jft.
#2232.
#2235.
#22351.
into
h.au/3 = au(/3)
from
(a
[#203
= :xea.v.wep [#203
\:xea. = .x~ea
[#203
Ka=f=a
*22'2
*22'2]
*22'3]
#22*31]
Dem.
h.*2235.*519.Dt:~{a;ea. = .a;ea}:
[#1011]
Db:(x):~{xea. = .xea}:
[#10251]
0\:~{(x):xea. = .xea}:
[*2043.Transp]
*20"32]
MATHEMATICAL LOGIC
208
[PART
two
at least
relations.
#2238.
Kan/3eCls
KauygeCls
h  a e Cls
#2239.
#22*36.
#2237.
[#2041]
[#2041]
[#2041]
n % (yftz) =
(<f>z)
z(<f>z. yfrz)
Dem.
h
#2233
xez
(<f>z)
[#203]
h
(1)
#2033
*22'4.
(1)
[Similar proof]
= z(~<f)z)
[Similar
=
:.aC/3./3Ca. =:xea. x x eft
#22392. \.z
Prop
= z{<t>zv ^z)
.<f>x.^x
(<f>z)
proof]
Dem.
h.*22l
[#4 38]
D
D
::
aC ft ~
::
C ft
x e a D x x e ft
Ca =
ft
:.
:.
ft
.
#2241.
b:aCft.ftCa. = .a = ft
[#22'4
#2242.
h.aCa
H:a^/3Ca
[Id. #1011]
This
is
"D x
xea
:.
#2043]
[#326 .#1011]
r:aC/3./3C 7
#2244.
=:.xea.= x .Xft::D\.Frop
[#1022]
#2243.
C a = x e ft D x x e a :.
x e a D x x e ft x e ft
.D.aC 7
[*10%3]
is
the following
#22441.
#2245.
t :
Dem.
b
#22"1
=:xea.D x .xeft.X<y:
=:xea.D x .X&ny:.Dh. Prop
[#10*29]
[*2233.*10413]
#2246.
bzxea.aCft.D.xeft
[#22441
#2247.
[2243;44]
#2248.
#22481.
Perm]
[#1031]
Dem.
b
#2241
[#22*48]
[#2241]
:.
Hp D a Cft ft C a
DianyCftny.ftnyCany:
D:an7 = /3n 7 :.DKProp
.
C]
*2249.
\:aC0.yC8.D.anyC/3n8
h a n a=
$22*5.
209
CALCULUS OF CLASSES
SECTION
[*1039]
a.
Dem.
V
$22*33
:.
x e a c\ a
[*424]
h
The above
is
(1)
*10'11
Kan = /3na
*22 52.
Thus
/3)
[*22'33
n7=
(1)
Prop
*43 *1011
.
*20'43]
laws.
*22
(a
*2043 D
*22 51.
= xea xea
= :#ea
*2043
be omitted.
onj3n 7 = (ani3)ft7 Df
53.
h:.a
*2254.
= /3.D:aC7. = .y9C 7
[*20'18]
= /3.D:yCa. = .y.C/3
*22551.\;a = /3.D.auy = /3uy
*22 56. Kaua =
*2255.
\z.a
The above
[*2018]
[*10411]
[*425 .*10'11]
is
*22 57.
Kau=ua
[*4\31
*2258.
\.aCav/3./3Ca\jj3
[*l'3.*22]
*2259.
h:aCy.0Cy. =
*1011]
a yjj3Cy
Dem.
K*221 .D h
Hy =
:z
[*1022]
[*4*77.*10271]
We
is false.
=:.(x):x6ayj/3.D.xey::3\. Prop"
[*2234.*10413]
The analogue
:. x e a
D x x e y x e fi
x e y 1.
^z.(x}z.Xa.D.xeyzxe/3.D.xyz.
= :. (x) :. x e a v x e/3 D x ey :.
of *4*78,
i.e.
aCfi.v.aCy. = .aCl3yjy
have only
aC0.v.uQy:D.aCfi\jy.
similar
\:.xea\J@.
*22*6.
Cf. *22'64'65.
= :aCy.@Cy.D y .xy
Dem.
.*2259.DH:.aC7.y3C7.D:a;6auy8.D.a;e7:.
[Comm]
Dhz.xeav fi.DzaCy.ftCy. D .xey:.
[*101121] D h :. x e a u /3 D a C 7 /3 C 7 D v x e 7
h *101
D h :. 0C7
C7 D y x e 7 D aC a u /S C a u D # e
h
[*2258]
r
(1)
D:xea\j/3
(2)
R&W
(1)
v,
(2)
Prop
14
MATHEMATICAL LOGIC
210
*2261.
\:ctCj3.D.aCl3uy
#2262.
h:aCj8. = ,auj8 = j8
[PART
[#224458]
Bern.
Y
#472
xe
;:
D x e ft =
oc .
:.
x e a .v x e @ = x
.
=i.x<a\j/3.
[*2234]
C/3. = i.xea v
f.(l).*10271.Dr:.:
.an^ =
e ft
.a;e/3
:.
(1)
^.ae/S :.
= :.au = ::DI.Prop
[#2043]
.[#471]
The
as in #22*62.
proposition #22*621
is
one of the
h:au(an
#2263.
The
8)
=a
[#4*44]
is
of the
same kind
as the
process employed in the proofs that have been written out in this number.
is
referred
We
to.
*22'631.
Kfln(au/3) = o
#22*632.
[#22*42*621]
#22 633.
[#22*551*621]
#22*64.
h:.aC7.v./3C 7 :D.arv/3C7
[#22*58*621]
/
Dem.
h
*22*47*51
C7 D
/3
C 7: /3 C 7 D
.
/3
C7
(1)
#22 65.
h:.aCy3.v.aC 7 :D.aC/3u7
#2268.
is
[*22*61*57 .#477]
untrue.
h:aC 8.D.u 7 C 8u 7
h (a n 0) u (a n 7) = a n ( v 7)
[#238]
Dem.
f
.#22*34.
::
xe
[(
n ^) u (a
r\
7)}
:.
ocean @. v .xeOLr><y\.
[#22*33]
= :.xea.xe/3.v.xea.xey:.
[#4*4]
:. xea: xe (3 .v ,xe<y
= :.xea.xe/3 U7 :.
[#22*34]
=:.xean(/3vry)
[#2233]
r
(1)
#1011 #2043 D
.
r
Prop
:.
(1)
SECTION
K(uj3)ft(au7) = au(j3ft7)
#22*69.
211
CALCULUS OF CLASSES
C]
The above
Note that either results from the other by interchanging the signs of
addition and multiplication.
law.
*227.
K(auj3)u 7 = ou(i3u7)
#22
uj3u7=(ou^)u7 Df
71.
[#433]
#2272.
#2273.
#2274.
Dem.
b
#22*43
#473
Db:ar\/3Cy. = .an/3Ca.an/3Cy.
=.an9Can7
[#2245]
/3
h .(l)X'
h
(1)
(2)
P>
*438
[#2241]
6f
h.(a) = a
H:aC/3. = ./3Ca
[#4"1]
*22811.
h:aC^.E.^Ca
[#41.*228]
#2282.
[#414]
#22831. l:o
= /3.E./3 = a
K(an/3) = av/3
Kan = (av)
K(an) = au/8
[#412]
#2284.
[#451]
#2283.
#2285.
#2286.
h.aftj8 = (u8)
*2284858687 are De Morgan's
#2288.
h. (a?), are (aw a)
#2287.
This
is
#2289.
This
is
(2)
#2281.
#228.
(1)
[#413]
[#411]
[#2284831]
[#457]
[*22'86831]
formulae.
[#211]
(a?)
*~ e (a  a)
[*3'24]
#229.
K(av/3)/3 = a/3
#2291.
r.au/3
[#561]
= au(a)
Dem.
b
#5 63

[*22333435]
DH:.#ea.v.#e/S:E:#ea.v
Dh
:.
=:arOu(/9o)
[#2234]
h
(1)
#1011
.xe/3 . ar~ea
:.
#2043
(1)
.Db. Prop
142
212
MATHEMATICAL LOGIC
*2292.
H:aC.D./3 = av(/3a)
*2293.
r.a/9 = a(an/3)
[PART
[#22'91*62]
Dem.
I
#4*73
Transp
I
E.^e^n^):.
[*2233]
[#5'32]
I
[*22'3533]
[*1011.*2043]
*22'94.
r:(a)./a.= .(a)./(a)
Dem.
K*10\L.
Dh:Ya)./a.D./(a):
Dh:(a)./.D.(a)./(a)
h *101
D h (a) ./( a) D ./{_ ( )}
[*22'8.*20\L8]
3/:
[*1 01121]
.
[101121]
h
(1)
(2)
(1)
(2)
Prop
This proposition
*90102, which
*2295.
is
H:(aa)./a.s.( a a)./(a)
Dem.
K*2294.Dh:(a).~/a. = .(a).~/(a)
H (1) Transp *20'6 D
Prop
.
I .
(1)
CALCULUS OF RELATIONS
#23.
Summary
The
of #23.
definitions
of those of #22.
and propositions of
this
number
will not be dealt with till Section D. Proofs will be omitted in the present
number, as they are precisely analogous to those of analogous propositions in
#22. In this number, as always in future, capital Latin letters stand for
expressions of the form ob<f> (x, y), or, where they are not being used as
apparent variables, for &$<(#, y). The principal propositions of this number
!
#2301.
#2302.
Rf>S = $(xRy.xSy)
#2303.
#2304.
RvS = ob (xRy v
R = $p{~ (xRy)}
#2305.
R^S=Rn^S
xSy)
Df
Df
Df
Df
Df
#233.
#2331.
#2332.
I .
#231.
#232.
^
R = ty
co (xRy)\
R^S= x [xRy
~(xSy)}
#23 35.
\:x(RnS)y. = .xRy.x8y
= xRy .v.xSy
:. x (R \j
8)
h x^Ry = .~(xRy)
#23351.
h.R^R
#2336.
#2337.
KiZntfeRel
KiZotfeRel
#2338.
h.^ReUel
#2339.
#2333.
#2334.
f
#23391. r
$$4>
{x,
x$<f> (x,
y)
y)
c;
#234.
#2341.
\:RGS.SGR.= .R = S
#23392. V
#2342.
b.RGR
#2343.
#2344.
\.RnSdR
\'.RGS.SGT.D.RGT
#23441.
b.RQS.xRy.O.xSy
xSy
214
MATHEMATICAL LOGIC
*2345.
biRQS.RQT.D.RQSnT
*23'46.
*2347.
*2348.
\:RGT.D.R*SGT
\:RGS.D.RnTGSnT
*23481.
^:R = S.D.RnT=SnT
*2349.
\:PGQ.RGS.D.PnRGQr*S
\.RnR = R
*235.
xRy .RdS.D.xSy
*2351.
\.RnS = SnR
*23
K(EnS)nT=En(/SnT)
52.
*2353.
*2354.
*2355.
*23551.
*23 56.
*23
57.
RnSnT=(RnS)nT
Df
h:.P = .D:PGr. = .#GT
h:.# = .:>:rGP.==.TG
h:P = S.:>.PiyT=#vyr
\.RvR = R
\.RvS = SvR
*23'59.
V.RZRvS.SGRvS
\:R<iT.S(iT. = .RvSGT
*236.
*2358.
*23 61.
hiPGS.D.PGSc/T
*23
= ..Rvy,Sf=
h:PG. = .Pn =
h.Rv(RnS) = R
62.
*23621.
*23'63.
H:PG,Sf.
*23631. h.E/S(Ea>Sf)
*23632. \:R =
*23 633. h:
=E
S.D.R = RnS
RG*S.D.^c/T = ( RniSf)oT
*2369.
\:.RGT.v.SGT:D.RnS<iT
hr.EG^.v.EGTzD.i^G^vyr
\:RdS.D.RvT<iSvT
h.(R*S)v(RnT) = Rn(SvT)
h .(Rv S) * (Rv T) = R v (S n T)
*237.
b.(RvS)vT=Rv(SvT)
*2371.
RvSvT=(RvS)vT
*2372.
*2374.
h.PGE.QGS.D.PvyQGEc/S
h:P = P.Q = >Sf.D.PvyQ = Pvy^
\:PnQGR.PnR<lQ. = .PnQ = PnR
*2364.
*2365.
*23
66.
*23'68.
*23
73.
Df
*23&
h.(P) = P
*2381.
hziZGS.EE.SG^iZ
*23811.
h:PG#. = .G^J2
*2382.
h^nSGr.^.PTG^S
*2383.
h:P = . = .^E=
[PART
SECTION
CALCULUS OF RELATIONS
C]
*2387.
h:R = S. = .S = R
b.(RnS) = R^S
\.RnS = (^RvS)
\.^(R^^8) = RvfS
\.RnL.S = (RvS)
*2388.
\.(x,y).oc(Rv^R)y
*23 89.
\.{x,y).~\x(RR)y]
*23
*2392.
\.(RvS)S=RS
\.R\jS = Rv(S^R)
b:RGS.D.S = Rv(SR)
*23 93.
*2394.
\:(R).fR.==.{R).f(R)
*23 95.
Ma^./tf.EE.^)./^!?)
*23 831.
*2384.
*2385.
*2386.
9.
*23 91.
RS = R(R * #)
215
*24.
Summary
The
q/"#24.
Any
but this
The
V,
it is
by V,
is
the class of
all
Thus V,
ambiguous as
letters,
i.e.
of
to type.
Its
x,"
definition is as follows
*2401.
is
V = &(# = ) Df
of
A is
*2402.
A = V Df
When
to exist.
a class a
W e write " g
T
in *1402.)
*2403.
is
it
it is
said
"
a " for a exists."
The
definition is
a!a. = .(ga;).#ea Df
In the present number, we shall deal first with the properties of A and V,
then with those of existence. In comparing the algebra of symbolic logic with
ordinary algebra,
1
and of
oo
Among
this
I.e. "
0,
while
number
*241.
and
h.A+V
nothing
is
is
A and V,
whose only member is the one individual. Our
primitive propositions do not require the existence of more than one individual.
that only one individual exists, there would be only two classes,
V being
SECTION
217
C]
*24102103 show
universal class,
class.
*24 2122 give forms of the laws of contradiction and excluded middle, namely
" nothing is both a and nota " (a r a = A) and " everything is either a or
"(au = V).
nota
#242324 2627
multiplication,
h:aC/9. = .a/3 = A
#243.
I.e. "
is
contained in
/S
nothing
is
no a
is
/3
nothing
is
both a and
ft."
h:j8C.D.B = )8u(aj8)
h:a/3Cy. = .aC#v 7
*24'411.
#24 43.
As a
rule, propositions
concerning
are
much
less
The
we
is
the contradictory of a = A, as
is
Thus
proved in #2454.
h~(aCi8). = .a!ai3
#2455.
I.e. "
all a's are /9's " is equivalent to " there are 's which are not /3's."
the familiar proposition of formal logic, that the contradictory of the
universal affirmative is the particular negative.
This
not
is
We
#24 56.
have
h:.a!(ouj8). = :a!o.v.a!j8
I.e. if a sum exists, then one of the summands exists, and vice versa
a product exists, both the factors exist (but not vice versa).
The
#2401.
#2402.
*2403.
#241.
*24101.
V = (a> = a?)
A = V
a !a. = .( a*).
b A + V
h V= A
number
Df
Df
Df
[*22351
022831 (*24*02)]
(*24'02)]
offer
no
difficulty.
and
218
MATHEMATICAL LOGIC
*24102.
I
(x)
. (f>cc .
(<f>z)
[PART
=V
Dem.
t
#1315 #5501 D
D
.
[#1011271]
:. (fyx
h :.(x).
<j>x
= x = x :.
= (x) (j>x = x = x
= 2 (<*) = S (ar = a;)
= $ (</>*) = V :. D h Prop
(fix
[*20 15]
[(#2401)]
is
class,
vice versa.
#24103.
(x)
<j>x
= 2 (<f>z)
=
Devi.
V
#24102 D
.
:.
[*22392]
[#22831]
=:t(0s) = V:
[(#2402)]
=A
(</>*)
: .
I
Prop
#24104. h.(ar).ajeV
Dem.
\.^20S.D\:xeY = .x^x
(1)
.(1)
1
*24105.
*13'15
.(x).x~eA
Dem.
h. #2235.
[#412]
K (1)
#2411.
Dh:a;A. = .a5~eV:
DH:;~eA. = .tfeV
D
*1011271
*24104<
#24104 . #101
DKareV.
(1)
Prop
K(o).oCV
Dem.
h
#2412.
xea.D .xeY
[Simp]
"Dh
[*1011.*221]
Dh:aCV:
[#1011]
(a)
CV D
I
Prop
K(a).ACa
Dem.
V
#24105 #101
.
DK~eA.
Dr:a?eA.D.a?ea
[#221]
h
#2413.
(1)
#1011
#221
t
(1)
Prop
h:a=A. = .aCA
Dem.
V
#2412 #473
.
.Dh:aCA. = .aCA.ACa.
=
[#2241]
#2414.
r
(a?)
xea
h. #24102.
[#2032]
=A D
:
Prop
a=V
[:(*). are a.
= J(ajeo) = V.
= a=V D
Prop
.
r
#24141.
219
SECTION C
hVCa. = .V = a
Bern.
.Dh:VCa. = .aCV.VCa.
= a = V D h Prop
[#2241]
V ;(x).x~ea.= a = A
Y
#2411 #473
.
#2415.
Bern.
V
=.a = A:Dh.
[#2032]
#2417.
#2421.
[#24103 *22\89]
#2422.
h.ana = A
Kav,a = V
#2423.
KanA = A
[#2412 #22621]
#2424.
h.auA = a
[#2412 #2262]
Prop
(#2402)]
[*22'88
#24102]
to zero.
#2426.
KnV =
[#22621. #2411]
h.awV = V
to 1.
[#2262 .#2411]
V
V
to oo
h:aC/3. = .a/3 = A
Bern.
D
D x e ft =
#4536
: .
xea
~ (x e a
x<^> e ft)
= :~(xea.xeft):
= :~(xeaft)
[#2235]
[#2233]
(1)
h.(l).*101127l.D
h
[#2415]
The above
#2431.
~ (x e a ft)
=.a/3 = A.OKProp
C ft =
proposition
is
(x)
h:aC/3. = .av/3 = V
Bern.
r
.#4*6
~2>\
[#2235]
[#2234]
= zx^a..v .xe ft i.
x~ea v .xe ft
= {x) zxe a.v.xeft:
= (x) x e ( a u ft)
= au/3=V:.Dh.Prop
..xeoL.'S
[#1011271] Dh:. a
,xe ft
C/S.=
(x)
[#2414]
it is
we
220
MATHEMATICAL LOGIC
[PART
Bern.
h
.Dh:.xea.D.Xl3: = :xa.D.x^e0:
= :~(xea.xe0):
*2235
[*4"5162]
[*2233]
>
(1)
*10 11271
[*2415]
*24312. H
a C/3 =
*2235
:.
/3
=:~(* ar>/3)
(x) #~e a n
= .an/3 = A:Dh.Prop
C  /3 =
(1)
=V
 a C /S =
.
a?~e a . D,
a?
[*22:34]
=
=
[*2414]
=:aw/3 = V:.DH.Prop
[*4*64]
:(#):#ea.v.tf;e/3:
:(x).xeav/3:
*2432.
H:.oui9 = A. = .a = A.^ = A
Dem.
V
9CA:
=:aCA./3CA:
=:a=A./3 = A:.D!.Prop
[*2259]
[*24'13]
*2433.
h:a = V.D.aw/3 = V
I .
*22551
.Dh:Hp.D.av = Vu/3
= V:Dh.Prop
[*2427.*2257]
*24'34.
h:a = A.D.an/3 = A
*2435.
*2436.
*24 37.
[*22481
*2423]
Dem.
(x)
x~e(a r\ /3)
=:(*).~(a;e.*6/9):
[*2233]
= :(x,y):x = y.D.~(xa.ye/3):
[Transp]
= (x, y) x e a y e @ D x j= y :. "D
*2438. h:.an/3 = A.D:a4=yS.v.a = A.y8=A
[*13 1 91]
\
Prop
Dem.
h
[*225]
[*2023]
h
(1)
[*46]
Ex p D
.
: .
D.a = A.
Z>.a = A./3 = A
D a=A = A
a n /3 = A D a =
Z>:a + /3.v.a = A./3 = A:.DKProp
.
(1)
221
SECTION
C]
*2439.
*244.
Dem.
\
*2431 l.Dh:an0
[*22 621]
k(l)^.
=.(au 8)a = S
Dh:/3na = A. = .08u)/3 = :
[*225157]
Dh:an/3 = A. = .(au/3)/3 = o
[*229]
*24401.
(1)
#C a
= A. = .0Ca.
s./3a = /3.
(2)
(1>
(2)
Prop
D ( u 7)  a = 7  a
.
Dem.
(1)
(2)
*24402.
Dew.
h
*22'49
Hp D
*2441.
Dem.
=*
(a
77
C a n /3
[*2413]
h
D.^n^CA.
[*2255]
n 0) w
(a  /S)
=nV
[*2426]
= a.DKProp
h:/3Ca.D.a = 8u(fl
*24411.
8)
Dem.
H..*22633^^^.DI:y8Ca.D./3w(a[*2441]
'
8)
'
= (an/3}u(a/3)
= a: Dr. Prop
Dem.
[*24323]
= (a 
[*2268]
=={(a/3)u/3}7
 7) u (0  7)
=a7:DI.Prop
[*24411]
This proposition
is
h:an/8C7.a/3C7. = .aC7
*24'42.
Dem.
r
*2259
[*2441]
.3\:an0Cv.a0Cry.==.(cLr\0)v(a0)Cy.
=.oC7:Dh.Prop
MATHEMATICAL LOGIC
222
[PART
h:/3C7. = .aC/3w 7
#2443.
Dem.
h
#5'6
[#2235*33]
"D
::
xea
::
xea
"D
x e y = :. x e a. D x e ft v % e y
D x e y = :. x e a D x e /3 v x e y :.
=.:..D.e(/3u 7 )
x<^> e /3
ft
[*2234]
:.
(1)
h.(l).*10ll271.Dh.Prop
= (anj3)u(<xy)v(/3ny)
*24431. \.(cHJy)o(/3vy)
for #24"44.
Dem.
V
#2268
(a
v y) n
(/3
 7) =
{(a u 7 ) n } v, {(a u 7 ) r\  y]
=(an/3)u( 7 n/3)vj( a  7 )u( 7  7 )
[*2268]
[#24r21]
= (a
[*2424]
=(an/3)v,( 7
[#225157]
= (a n ) u
/3)
/3) u (a  7 ) u A
n)u(a 7 )
(y
(a
 7) u ( n 7) D
.
[#2251]
=(ar\fir\y) u
*242235
a n
/3
Prop
= (a n /3) n (7 u  7)
(an 8n 7 )u(an/3 7 )
8)u(a7)u( 8n 7 )
[#2268]
[#22551]
K (a n
/3)
(a
7) = (a
(a
70 /3)
r>
/?
n 7) v (an
7 n /3) u (a  7)
=(any3r>7)o(a7)
=(a7)u(an/8r7).
[#2263]
[#2257]
[#22551]
(a
r>
/3)
(a
 7) u (/3 n
7)
[#2263]
#2444.
I
(a
u 7) n
(/3
 7) = (a n  7)
#24*45.
I
(a
n 7) u
(/3
7)
=A =
.
/3
(/3
[#24431 432]
7)
C7 7C
r
Dem.
h
#2432
I
(a
n 7) v
(/3
[#243311]
#2446.
f
(a
n 7) u
#2445
(/3
 7) = A = a n 7 = A /3  7 = A
=. 7 Ca>./3C7:Dh.Prop
.
 7) = A D
.
a n
/3
=A
Dem.
.
#2244 D
.
[#22811]
[#24311]
The
following propositions,
for use in
much
down
/3 C  a
D.aC/3.
D.an/3 = A:DKProp
Hp D
.
lemmas inserted
SECTION
*2447.
223
C]
!:ar/3
Dem.
K*24311.DI:an/3=:A. = ./3Ca
h *2241
DF:au,8=7. = .au/3C7.7Cav;/3.
[*2259.*2443]
=.aC7.^C 7 7 aCi3
K(l).(2)Oh:a = A.au/8 = y.=s.Ca.aC7.
(1)
(2)
[#4'3]
[#2245]
[#2241]
8C7.yaC.
=.aCy.^Cy.y8Ca.yaC/9.
= .aCy./8Cya.yaC.
=.aCy. S = ya:Dh.Prop
Ca.fCa.77C/3.7/'C/S.an/3 = A.D:
wi7 = fwV = ? = !'. = V
/
h:.
#2448.
17
Dem.
K #2273.
F
#22481
[#2268]
h. #22621.
#2248
Dh:97C/3.an^=A.D.^naCA.
[#2255]
D.7;na = A
(3)
(4)
Dh:.Hp.D:(^a)u( na) =
}/
[#2424]
r
(3)
(4)
h:i/'C^.an/8 = A.D.Va = A
Similarly
.
(3)
Dh:,Cj8.D.j?naConj8:
[#2413]
I
(2)
[#347]
h
(1)
(5)
:.
Hp D
.
('
H.(2).(6).(7).Dh:.Hp.D:fu7; =
Similarly
:. Hp
D v 1; =
.
(6)
='
(7)
^ju^aj^'uA
[#2424]
f
(5)
^A
f uV.D.^^r
'
77'
(8)
D v^v'
(9)
h.(l).(8).(9).Dr.Prop
The above proposition, besides being used in the next two, is used in the
theory of couples (*54 6), in the theory of greater and less (#117632), and in
the chapter on the ordering of classes by the principle of first differences

(#17068).
#24481. r:.an/3
./3
=y
Dem.
#2448
:.aC
L~^rJ^l D
.
.aC
a.
ft
C
.yC~ a. a a = A D
.
a w/3
h
#2242
#2421
=aw7 =
.
= a ./3 = y
h:.aCa.aCa./3C K.^Ca.a a = A =
.
/3
C a^C a
(1)
224
MATHEMATICAL LOGIC
[PART
= .an = A.an 7 =A
[*24311]
K*20'2.*473.DH:a = a./?=7. = . = 7
K(l).(2).(3).Dh.Prop
The above
proposition
is
(2)
(3)
theory of greater and less (#ll7 582), and in the theory of transfinite induction

(*257).
*24482. h:.Ca.77C.a/3
#2448
t^4
$>y
The above
#2242 ]
J
proposition
is
h:.ar==A.D:aC/3u 7 = .aC 7
#2449.
Dem.
b
022/68]
#24*24
K
I
(1)
(2)
[*22621]
*24491.
.o=an(^u 7 )
= (an)u(an 7 )
Dh:an/3=A.D.(an^)u(an 7) = on 7
D h Hp D a C v 7 = a = a n 7
= .aC 7 :DKProp
*22621 .Dl:oCj9w7. =
:.
hj3n7 = A.Cj3u7.
D a fi = a.r\
.
(2)
7 .a y =
(1)
an j3 .a = (a /3) u (a 7 )
2)em.
K*22'621.
[*22481]
0244]
h Hp D a=ar 7
Dh:Hp.D.(a/3)v(a 7) = (an 7 )u(an)
= a n (7 v )
=a
Similarly
h.(l).(2).
02268]
022621]
I .
(1)
The above
(2)
proposition
Prop
is
(2)
(3)
series (#21184).
h.)3Ca.ai3 = 7.D.a7 = j8
Dem.
K #22481
O
O 22 8 9
22 8 86 ]
'
'
'
Hp D a 7 = a (a/3)
=an(au)
= '>
The above
proposition
It is first
=/3:Dh.Prop
2 2621]
series.
segments of a
in the theory of
#24492.
(3)
(1)
is
used
SECTION
225
C]
#24493. h:/Sn 7 =
A.D. = (/9)u(7)
Dem.
\
#2284 . #2417
[#24*26]
[*22'68]
Hp 3  /3 w  7 = V
3.a = an(u 7 )
= (a/3)u(a 7 ):3r
.
.Prop
#24494.
3h:Hp.3.a = A
K#243.
K#2431I.
(1)
3h:Hp.3./3Ca.
3
[#2244]
C
.3.170 =
[#22621]'
(2)
17
K #2268.
3K(ui7)a = (a)v(i7a)
h.(l).(2).(3).*2424.3l:Hp.3.(^ui )a =
Similarly
h
(4)
(5)
This proposition
#24495. h:ar 7
I .
(3)
i7
(4)
Hp 3 .(%v v )0=
(5)
Prop
is
Dem,
h
#228768
(a   7) u (7   7)
= a _ 3 7
K*24311.#22621.3H:Hp.3.a7=a
H
(a
u 7)  ( v 7) =
[#2421]
The above
K Prop
h.(l).(2).
proposition
used
is
(1)
(2)
the
in
theory
of
minimum
points
(*2058383284).
Many
of classes.
number we
shall
fact that to say a class exists is equivalent to saying that the class is not equal
to the nullclass.
This
proved in #2454.
is
#245.
#24 51.
~g
a=
[*42 (#2403)]
.
Dem.
[#10252]
~ {(g.r) x e a]
=.().aj~ea.
[#2415]
I .
#2452.
is
g V
!
#245
[#24511
I
~jj a
!
.a
= A:3b.Prop
Transp]
This proposition states that the class of all objects of the type in question
not null, but has at least one member. The assumption that there is someit
&w
15
MATHEMATICAL LOGIC
226
[PART
#10*1, that
is
#24*53.
K~g!A
#2454.
h:g!a.=
#24*55.
*:~(aC/3).
#24*56.
h.a!(auj3). = :a!a.v.a!/3
#24*561. h
#24*57,
.a==A
= .a!a
n ) D
#22*481 .DI:an/3
a a a /3
l:.ar> = A.D:g;!a.D.a=f
!
(a
[#24*51
#20*2]
[#24*51
Transp]
[#24*3
Transp #24*54]
#22*34]
[#10*42
[#10*5
#22*33]
Bern.
h
[#22*5]
= A.a=/S.D.ara = A.
D a=A
.
#24*571. H:
(1)
3..~3!a
[#24*51]
Exp Transp D
.
(1)
Prop
a !a.a = 8.D.a!(artyS)
/
Dem.
V
#24 5 7
*
C omm D
.
: .
[Transp]
D a n j3 = A D a + /3
D : = . D .ar\j3$ A
D.g!(an/3)
:
'
[#24*54]
K (1)
Imp D
.
(1)
Prop
#24*58.
h:.aC/3.0:a!a.D.a!/9
#246.
[#10*28]
Dem.
h
#22*41
Transp
O h Hp
:.
a+
Oh:a = /3.D./3a=A
h. #24*21.
.
.~(/3
C a)
D.a!/3a
[#24*55]
K (2)
K(l).(3)'.
a a.D
*
Dh.Prop
#24*61.
H:~ a !.D.av/8 = a
[*24*51*24]
#24*62.
h:~a!.D.an = A
[*24'5123]
+ /3
(1)
(2)
(3)
SECTION
#2463.
In
be a
q]
I
:.
227
The
condition "a
hypothesis in
hypothesis
tt
a"
one required as
is
Bern.
V
*13191 .Dh:.A~e/c.
[Transp]
[*2454]
:a =
A.D a .a~e:
=:ae/c.D .a4=A:
a e k D a 3 a :. D
Prop
152
#25.
Summary
q/"#25.
This number contains the analogues, for relations, of the definitions and
propositions of #24.
in #24.
"R
exists"
The
is
"g
written
R."
propositions of this
number
are
made
#2501.
V = $p(x = x.y = y)
Df
#2502.
Df
#251.
A = V
a R =
KA + V
#25101.
r.V=^A
#25102. r
(x, y)
(ga?, y)
. <j>
(x,
#25103. \:(x,y).~<f>
#25104. h
(x,
y)
#25105. h
(x,
y)
y)
(x,
Df
xRy
=V
y) > = &<j> (x, y) = k
.
= 5$
.
<f>
(x,
xYy
.~(xAy)
#2511.
\.(R).RQV
#2512.
\ .
#2513.
\:R = A. =
#2514.
h:(x,y).xRy. = .R = Y
#25141.
#2517.
\:YQR. = .Y = R
(x, y) ~ (xRy) = .R = A
\:R = Y. = .^R = A
*2521.
\.Rn^R = A
#2515.
I
(R)
AGR
.RCA
in
number.
#2503.
for
y)
SECTION
229
C]
*2522.
\.Rv^R = Y
#223.
h.RnA = A
*2524.
KPoA=P
*25 26.
I
*2527.
RAV=R
KPvyV = t
*25*3.
\:R<ZS. = .R^S = A
*2531.
h:PG#. =
*25311.
\:RQ^S.~.RnS = A
h:^PG. = .Pc/# = t
h:RnS = A. = .R8 = R
K:Pvy = A.2.P = A.# = A
h:P = V.D.Pc/#=V
H:P = A.:>.Pn,Sf=A
l:P = V.D.Pn =
\:R = A.D.RvS = S
H :: R r\ S = A = \.xRy zSw ^ x y z w :oc^z.v .y^w
h.EA^=A.D:E^.v.ii = A.Sf = A
b:.Rr\S = A.= xRy Dx>y ,~(a:Sy)
h:Pn(3 = A.= .(PoQ)P=Q. = .(Pc;Q)^Q = P
\:QGP.D.(QvR)P = RP
\:PnQ = A.RQP.SdQ.D.RnS = A
\.R = (RnS)v(R^S)
!:>8'G R.D.i2 = 5c/(iB^^)
h:QGP.GQ.D.(PQ)vy(Q^S) = P,S
h:PnQGR.P^Q(iR. = .PG.R
\:P^.QGR. = .P(lQvR
h.(PvyJS)n(Qvy^i2) = (PnQ)vy(P R)a(QnE)
\.(P^R)v(QnR)=(PnQ)v(P^.R)v(QnR)
(P vy 22) A (QvR) = (P fy^R) c/(QnP)
h (P n P) o (QP) = A = Q G P P G^P
h:(PniJ)a(Q^) = A.D.PnQ = A
h:PAQ = A.PiyQ = P. = .PGP.Q=P^P
K GP P' GP G Q ' G Q P n Q = A D
h
*25312.
*25 313.
*2532.
*2533.
*2534.
*25 35.
*2536.
*2537.
*2538.
*25 39.
*254.
*25401.
*25402.
*2541.
*25411.
*25412.
*2542.
*25'43.
*25'431.
*25432.
*25'44.
*2545.
*25 46.
*25 47.
*2548.
.uPiy,Sf =
I
::
Pc/=P
c;,Sf
.= .P = P'.,Sf=,Sf'
*25 482.
*2549.
*25'481.
230
MATHEMATICAL LOGIC
[PART
\:Q*R = A.PGQvR.3.
25491.
*25492.
*25493.
*25494.
t".RGP.SGQ.P*QA.^ (RvS)PS.(RvS)Q = R
H:PAP = AO.(Pc/P)^(^c/P) = Pi.Q
g R = a #, y)
H:~g!P. = .P = A
*25495.
*25
5.
*25'51.
*2552.
*2553.
*25'54.
*25
55.
*2556.
f :
Kg!V
K~g!A
h:g!P. = .P^A
h~(iJGS). = .a!iJ^
h:.g!(Bvy). = :a!E.v.a!
*25561. h:a!(JJnS.).D.a!iJ.a!^
*2557. f:.PnSf = A.D:g!P.D.P4=^
*25571..
I
R R=
.
D a (R n S)
.
*2561.
H.PGS.DrgliZ.D.alS
\:.RGS.D:R$S. = .rISR
h :~a D P vy = #
*2562.
h:~a!.:>.Pn,Sf = A
*2563.
l:.A~6. =
*2558.
*256.
:22eff. Dje.gl.fi.
SECTION D
LOGIC OF RELATIONS
Id the present section
we
shall
The
be of fundamental importance.
*30.
Summary
The
DESCRIPTIVE FUNCTIONS
o/*30.
functions such as a
/3,
their values.
i.e.
of mathematics, such as
for
x*, sirxx,
log
same
by means of
number
its
yet
as they would be
were substituted for sin w/2. This appears e.g. from the proposition
= 1," which conveys valuable information, whereas " 1 = 1 " is trivial.
Descriptive functions, like descriptions in general, have no meaning by themselves, but only as constituents of propositions*.
if 1
The general
#3001.
R'y = (ix)(xRy) Df
That
"
is,
R'y "
is to
mean
"
R y,
(
about
i.e.
apostrophe in
to son, "
all
"R'y" may be
"
read "of."
Thus
If
if
R is
is
to y."
to y, all propositions
false.
The
more.
All the functions that occur in ordinary mathematics are instances of the
above definition all are obtained in the above manner from some relation.
;
Thus
in our notation
"R'y"
commonly be
We
to
y when x
= sin y.
R'y = (ix)(xRy), where the meaning given to the
a description, must be understood to mean that the term
definition such as
term defined
is
defined (in this case R'y) and the description assigned as its meaning (in this
case (ix) (xRy)) are to be interchangeable in use the definition is, in a sense,
more purely symbolic than other definitions, since the description assigned as
:
f(R'y).
= .f{(ix)(xRy))
* Cf. *14, above.
It
Df.
233
DESCRIPTIVE FUNCTIONS
SECTION D]
But even
this definition
= .\(ix){xR y )\.f{{ix)(xRy))
[R'y].J(R'y).
Df.
unnecessary to adopt this form of definition, provided it is understood that the definition #3001 means that "R'y" may be written for
" (ix) (xRy) " everywhere, i.e. in indications of scope as well as elsewhere. The
But
it is
which
[R'y] .f(R'y)
=s
[(ix)
(xRy)] .f(ix)(xRy),
#301, below.
is
It is to
R'y = (ix)(xRy).
For
this,
by the
definition, is equivalent to
(ix)
(xRy)
= (ix) (xRy),
R to
there
is
one term,
y.
R'x,
i.e.,
in the absence of
any contrary
R'x
is
to
be
in question occurs.
We put
#3002.
R S'y=R'(S'y)
t
Df
It is to be in
terpreted as meaning
[R<S'y].f(R'S<v).
= .[R<(S<y)].f{R<(S<y)}
Df.
That
is,
when R'y
exists,
the
"'
#3018.
:,
E R'y
!
(z)
<f>z :
Z>
<f>
(R'y)
we
We
have
234
MATHEMATICAL LOGIC
[PART
so that
exists.
This
results immediately from #1418, and shows that, provided R'y exists, the fact
that " R'y " is an incomplete symbol does not prevent its being substituted
as a value of z
function
One
whenever we have
(z)
(f>z,
or
<f>z.
number
is
h:.x=R'y. = :zRy.=~z .z = x
#303.
#3031.
I.e.
:.
x = R'y
"x = R'y"
= xRy zRy .D z .z = x
R to y is identical with
#3037.
x.
is
E R'y
In the hypothesis,
of
them
The use
of
#3037
is
Suppose,
descriptive functions.
chiefly in cases
for
S'w
for z if
substitute
it for
In like manner,
if
#304.
exists.
if
By
Hence whether
exist,
and there
we may
z and obtain
h
we
:
E R'y
we
replace y by T'v,
!
obtain
R'T'v = R'S'w.
is
This proposition states that, provided R'y exists, to say that a is the term
which has the relation R to y is equivalent to saying that a has the relation
R to y. Thus for example " a is the occupier of the house y " is equivalent
to " a occupies the house y," " a is the writer of Waverley " is equivalent to
" a wrote Waverley," " a is the father of y " is equivalent to " a begot y." But
we cannot argue from " John Smith inhabits London " to " John Smith is the
inhabitant of London."
always true.
for
which
we
is
When
R is such
a R'y =
.
aRy
that
E R'y is
!
always true,
both
Thus
that
that
is
useful in cases
where
#3041.
.
235
DESCRIPTIVE FUNCTIONS
SECTION D]
if
we know
R and S are
#3001.
R'y = (ix)(xRy) Df
#3002.
R'S'y = R'(S'y)
identical,
Df
is
to be treated as
[R'y] ./(R'y)
#301.
#3011.
:.
The
s
[R'y] ./(R'y)
[(ix)
(xRy)]./(ix) {xRy)
(a&)
[*42
(#3001)]
ff.,
made
#3012.
#3013.
\
#3014.
::
[*14'31]
X (R'y)}[*Um
= P 3 {R'y] % (12'y)
.
[#1433]
#30141. h
::
E R'y D
!
[fi*y]
:.
x (2ty) D^ =
.
[ity]
X (R'y) D p
.
[#14331]
#30142.
::
X (R'y)
[R'y]
[#14332]
#3015.
\z.pz [R'y]
The
X (R'y) =
z
[R'y] .p.
X (R'y)
[*14'34]
#3016.
b:[R'y]./(R'y,S'z). = .[S'z]./(R%S'z)[*U113]
#3017.
\:.[R'y]./(R'y,S'z).
(g;6, c)
#3018.
Yz.E\R'y\(z).$z:1.4>(R'y)
#3019.
h:.R'y = b.D:yjr(R'y). =
#302.
fz.
E R'y =
.
a &)
c)
[#14112]
[#1418]
[#1415]
.y}rb
xRy .= x .x^b
[*42
#1411
(#3001)]
In proving #30*2, we have to use the definition #3001, not #30'1, because
if we attempt to apply
which
leads
to
an
expression containing
(<f>x),
the meaningless constituent E b. But by the definition #3001, every typographical occurrence of the symbol "R'y" means what results when this
symbol is replaced by " (ix) (xRy)," hence " E R'y " means " E (ix) (xRy)"
!
(ix)
(<f>x) is
(ix)
MATHEMATICAL LOGIC
236
#3021.
: :
E R'y =
!
:. (
[PART
[#14203 (#3001)]
.
#3022.<