A Survey of Symbolic Logic

SEMICENTENNIAL PUBLICATIONS
OF THE
UNIVERSITY OF CALIFORNIA
1868-1918
A SURVEY OF
SYMBOLIC LOGIC
BY
C.
I.
LEWIS
UNIVERSITY OF CALIFORNIA PRESS
BERKELEY
1918
PRESS OF THE NEW ERA PRINTING COMPANY LANCASTER, PA.
TABLE OF CONTENTS
PREFACE
v
I.
CHAPTER
SECTION
THE DEVELOPMENT OF SYMBOLIC LOGIC

I.
The Scope
and
Symbolic Logic. Symbolic Logic Summary Account of their Logistic.

I
5-
of
Development
SECTION
SECTION SECTION
II.
Leibniz
III.
From Leibniz to De Morgan and Boole
18
IV.
De Morgan
Boole
37
51
SECTION
SECTION
V.
VI.
Jevons
Peirce
72
79
107
SECTION VII.
SECTION VIII.
Developments since Peirce
CHAPTER
SECTION SECTION SECTION
SECTION
II.
THE CLASSIC, OR BOOLE-SCHRODER AL GEBRA OF LOGIC

I.
118
General Character of the Algebra.
The Postulates
118
and
II.
their Interpretation
Elementary Theorems
General Properties of Functions
122
III.
132
of Equations.
. .
SECTION IV.
V.
Fundamental Laws Fundamental Laws
of the
Theory
144
SECTION VI.
of the Theory of Inequations. 166 Note on the Inverse Operations, "Subtraction" 173 and "Division"
CHAPTER
III.
APPLICATIONS OF THE BOOLE-SCHRODER
ALGEBRA
SECTION
SECTION SECTION
I.
175
175
Diagrams
for the Logical Relations of Classes ....
II.
III.
SECTION IV.
The Application The Application The Application
to Classes
184
to Propositions
213 219
to Relations
CHAPTER
SECTION
IV.
SYSTEMS BASED ON MATERIAL IMPLI CATION

The Two-Valued Algebra
iii
222
222
I.
IV
Table of Contents
SECTION
SECTION
II.
III.
The Calculus of Prepositional Functions. Func 232 tions of One Variable Functions of Two or More Variables. 246 Prepositional
Derivation of the Logic of Classes from the Calcu
lus of Propositional
SECTION IV.
Functions
260 269
SECTION
V.
SECTION VI.
The Logic The Logic
of Relations of Principia
Mathematica
279
291
CHAPTER
SECTION
SECTION
V.
I.
THE SYSTEM OF STRICT IMPLICATION...

Primitive Ideas, Primitive Propositions, and
Im
292
299
II.
mediate Consequences Strict Relations and Material Relations
SECTION
III.
The Transformation
of Consistencies
{-/-)
306
SECTION IV.
Extensions of Strict Implication.
The Calculus
of
and the Calculus
Ordinary
316
Inference
SECTION
V.
The Meaning
of
"Implies"
324
CHAPTER
SECTION
VI.
SYMBOLIC LOGIC, LOGISTIC, AND MATHE MATICAL METHOD

General Character of the Logistic Method.
"Orthodox"
340
I.
The
340
View
SECTION
II.
Two
Varieties of Logistic
Method
Peano
Formu343
laire
and Principia Mathematica.
The Nature
of
of Logistic Proof
SECTION
III.
"Heterodox"
View
of
of the
Nature
Mathe
354
362
matics and of Logistic
SECTION IV. SECTION

V.
The
Logistic
Method
Kempe and Royce

.
Summarv and
Conclusion
367
APPENDIX.
TWO FRAGMENTS FROM
LEIBNIZ
373 389 407
BIBLIOGRAPHY
INDEX.
PREFACE
The student who has completed some elementary study
of symbolic
logic and wishes to pursue the subject further finds himself in a discouraging He has, perhaps, mastered the contents of Venn s Symbolic situation.
The Algebra of Logic, or the chapters concerning this subject in Whitehead s Universal Algebra. If he read German with sufficient ease, he may have made some excursions into
Logic or Couturat
s
admirable
little
book,
Schroder
Vorlesungen uber die Algebra der Logik.
These
all
concern the
is
classic, or Boole-Schroder algebra,
and
his
knowledge
of symbolic logic
probably confined to that system. His further interest leads him almost inevitably to Peano s Formulaire de Mathematiques, Principia Mathematica
of of the
Whitehead and Russell, and the increasingly numerous shorter studies same sort. And with only elementary knowledge of a single kind of
difficult
development of a small branch of the subject, he must attack these most and technical of treatises, in a new notation, developed by methods
which are entirely novel to him, and bristling with logico-metaphysical If he is bewildered and searches for some means of further difficulties.
preparation, he finds nothing to bridge the gap.
Schroder
work would
be of most assistance here, but this was written some twenty-five years
ago; the most valuable studies are of later date, and radically
new methods
have been introduced.
What
ject
such a student most needs

will familiarize
is
a comprehensive survey of the sub

single
one which
him with more than the
system
which he knows, and
not only the content of other branches and the alternative methods of procedure, but also the relation of these to
will indicate
the Boole-Schroder algebra and to one another.
The
present book
is
an
attempt to meet this need, by bringing within the compass of a single

volume, and reducing to a
of the requirements of a
common
notation (so far as possible), the most

logic.
If,
important developments of symbolic

"handbook"
in addition to this,
some
are here fulfilled, so
much
the better.
But
in
this
survey does not pretend to be encyclopedic.
A
is
gossipy recital
of results achieved, or a superficial account of
methods,
of
no more use
symbolic logic than in any other mathematical discipline.
What
is
presented must be treated in sufficient detail to afford the possibility of real

insight
and grasp.
This aim has required careful selection of material.
vi
Preface
historical
The
of
summary
in
Chapter
attempts to follow the main thread
development, and no reference, or only passing mention, is given to those studies which seem not to have affected materially the methods of
later researches.
In the remainder of the book, the selection has been
governed by the same purpose. Those topics comprehension of which seems most essential, have been treated at some length, while matters less
fundamental have been
set forth in outline only, or
omitted altogether.
My
own contribution to symbolic logic, presented in Chapter V, has not earned the right to inclusion here; in this, I plead guilty to partiality. The discussion of controversial topics has been avoided whenever possible
and, for the
rest, limited to
the simpler issues involved.
Consequently,
the reader must not suppose that any sufficient consideration of these
questions
accurate.
is
here given, though such statements as are
made
will be, I
hope,
Particularly in the last chapter, on

",
"Symbolic
Logic, Logistic,
and Mathematical Method

adequate account of the
size of this one.
it
is
not possible to give anything like an
facts.
I
Rather,
That would require a volume at least the have tried to set forth the most important and
arbitrarily
critical considerations
is
somewhat
and
and dogmatically,
since there
not space for argument
to provide such a
map
of this difficult terri
tory as will aid the student in his further explorations.

Proofs and solutions in Chapters II, III, and IV have been given very
fully.
Proof
is
of the essence of logistic,

fair
and
of
it is
my
observation that stu
dents
even those with a
knowledge
mathematics
seldom
command
the technique of rigorous demonstration.
In any case, this explicitness can do no harm, since no one need read a proof which he already understands.
I
am
indebted to
many
friends
and colleagues
for valuable assistance in
preparing this book for publication: to Professor

tions of
W.
A. Merrill for
emenda
my
translation of Leibniz, to Professor J.
H. McDonald and
Dr. B. A. Bernstein for important suggestions and the correction of certain errors in Chapter II, to Mr. J. C. Rowell, University Librarian, for assistance
in securing a
number
of rare volumes,
and to the
officers of the
University
Press for their patient helpfulness in meeting the technical difficulties of

printing such a book.
manuscript, eliminated
Mr. Shirley Quimby has read the whole book in many mistakes, and verified most of the proofs.
indebted to
But most
of all, I
am
my
friend
and teacher, Josiah Royce,

failed to give
is
who
first
aroused
my
interest in this subject,
and who never

that
me encouragement and
due to him.
BERKELEY, July
10,
wise counsel.
Much
best in this
C>
book
is
L LEWIS.
1917.
CHAPTER
I.
THE DEVELOPMENT OF SYMBOLIC LOGIC

THE SCOPE OF SYMBOLIC LOGIC. SYMBOLIC LOGIC AND SUMMARY ACCOUNT OF THEIR DEVELOPMENT
The
to as
LOGISTIC.
subject with which

"
we
are concerned has been variously referred

"algebra
logic",
"symbolic
logic",
logistic",
of
logic",
"calculus
of
logic",
"mathematical
logic",
"algorithmic
And none
because
of these
is
satisfactory.
We
in
and probably by other names. have chosen "symbolic logic"

in this country,
it is
the most
commonly used
is
England and
understood.
and
is
because
its
signification
pretty
well
Its
inaccuracy
obvious: logic of whatever sort uses symbols. We are concerned only with that logic which uses symbols in certain specific ways those ways which are exhibited generally in mathematical procedures. In particular,
logic to
relations,
be called
symbols for the logical and must so connect various relations that they admit of "trans
"symbolic"
must make use
of
formations"
and
"operations",
according to principles which are capable
of exact statement.
If
we must
is
give
some
definition,
we
shall
hazard the following: Symbolic
Logic
the development of the most general principles of rational pro
cedure, in ideographic
of these principles
symbols, and in a form which exhibits the connection

Principles which belong exclusively
e.
one with another.

of rational procedure
to
some one type
g.
to dealing with
is
number and
quantity
are hereby excluded,
and generality
designated as one of the
marks
form.
of symbolic logic.
Such general principles are likewise the subject matter
of logic in
any
To
be sure, traditional logic has never taken possession of more

field
than a small portion of the
are unnecessarily restricted.
As we
which belongs to it. The modes of Aristotle shall have occasion to point out, the
reasons for the syllogistic form are psychological, not logical: the syllogism,
made up of the smallest number of propositions (three), each with the small est number of terms (two), by which any generality of reasoning can be attained, represents the limitations of human attention, not logical necessity. To regard the syllogism as indispensable, or as reasoning par excellence, is
2
1
Survey of Symbolic Logic
the apotheosis of stupidity.
And
the procedures of symbolic logic, not

to
being thus arbitrarily restricted,
matter between
it
and the traditional
accidental, not essential,

logic
all
is
mark a difference of subject But any such difference is logic. and the really distinguishing mark of symbolic
may seem
the approximation to a certain form, regarded as ideal.
There are
degrees of such approximation; hence the difficulty of drawing any hard

fast line
between symbolic and other logic. But more important than the making of any such sharp distinction is the comprehension of that ideal of form upon which it is supposed to
and
depend.
The most convenient method which the human mind has

is
so far
devised for exhibiting principles of exact procedure

call,
the one which
we
in
general terms, mathematical.

(1)
The important
which
characteristics of
this
form are:
the use of ideograms instead of the phonograms of

(2)
ordinary language;
to
the deductive method
may
here be taken
is
mean simply
that the greater portion of the subject matter
derived
from a relatively few principles by operations which are

(3)
"exact";
and
the use of variables having a definite range of significance.
Ideograms have two important advantages over phonograms. In the first place, they are more compact, + than 3 than "three", etc.
"plus",
This
is
no inconsiderable gain, since

of
it
makes
possible the presentation of a
formula in small enough compass so that the eye

glance and the image
reference with a
it
(in visual or other terms)
may apprehend it at a may be retained for
minimum
of effort.
None but
a very thoughtless person,
or one without experience of the sciences, can fail to understand the enor
mous advantage
notation
is
of such brevity.
In the second place, an ideographic
superior to
any other in precision.
Many
ideas which are
quite simply expressible in mathematical symbols can only with the greatest
be rendered in ordinary language. Without ideograms, even arithmetic would be difficult, and higher branches impossible.
difficulty
The deductive method, by which a considerable array
of facts
is
sum
marized in a few principles from which they can be derived, is much more than the mere application of deductive logic to the subject matter in It both requires and facilitates such an analysis of the whole question.
body
of facts as will
most
precisely exhibit their relations to one another.

is
In fact, any other value of the deductive form
largely or wholly fictitious.
presentation of the subject matter of logic in this mathematical form constitutes what we mean by symbolic logic. Hence the essential
characteristics of our subject are the following:
(1) Its
The
subject matter
is
The Development of Symbolic Logic

the subject matter of logic in any form
that
is,
the principles of rational
or reflective procedure in general, as contrasted with principles which

(2) Its belong exclusively to some particular branch of such procedure. medium is an ideographic symbolism, in which each separate character
represents a relatively simple and entirely explicit concept.

all
And,
ideally,
non-ideographic symbolism or language
is
excluded.
"
(3)
Amongst the
system)
ideograms, some will represent variables (the
terms"
of the
having a definite range of significance.
Although
"
it
is
non-essential, in
"individuals",
any system
they
so far developed the variables will represent
or classes, or relations, or propositions, or

will represent
prepositional
functions",
or
ambiguously some two or more of these. (4) Any system of symbolic logic will be developed deductively that is, the whole body of its theorems will be derived from a relatively few principles, stated
in symbols,
We
by operations which are, or at least can be, precisely formulated. have been at some pains to make as clear as possible the nature of
logic,
symbolic
because
its
distinction
from
"ordinary"
logic,
on the one
hand, and, on the other, from any mathematical discipline in a sufficiently abstract form, is none too definite. It will be further valuable to comment
briefly
on some of the alternative designations
for the subject
which have
been mentioned.
would not have served our purpose, because "logistic" is used to denote symbolic logic together with the application of commonly its methods to other symbolic procedures. Logistic may be defined as
"Logistic"
the science
which deals with types of order as such. It is not so much a subject as a method. Although most logistic is either founded upon or makes large use of the principles of symbolic logic, still a science of order
in general does not necessarily presuppose, or begin with,
symbolic
logic.
Since the relations of symbolic logic, logistic, and mathematics are to be
the topic of the last chapter,
that matter here.
We
is
have mentioned have

in the
we may postpone any further discussion of it only to make clear the meaning
pages which follow.
It
which
"logistic"
to
comprehends
symbolic logic and the application

fies
of such
methods
as symbolic logic exempli

is
to other exact procedures.

"Algebra
Its subject
matter
not confined to
logic.
of
logic"
is
hardly appropriate as the general
name
for our
subject, because there are several quite distinct algebras of logic,
and
because symbolic logic includes systems which are not true algebras at all. "The algebra of logic" usually means that system the foundations of
which were
laid
by Leibniz, and
after
him independently by
Boole,
and
which was completed by Schroder.

"
We
shall refer to this
system as the
a
Boole-Schroder Algebra
"Calculus"
".
is
a more general term than
"algebra".
By
"calculus"
will
be meant, not the whole subject, but any single system of assumptions
their consequences.
for symbolic logic
and
The program both

clear form,
and
for logistic, in anything like a
sketched by Leibniz, though the ideal of logistic seems 1 Leibniz left frag to have been present as far back as Plato s Republic.
was
first
mentary developments of symbolic logic, and some attempts at logistic which are prophetic but otherwise without value. After Leibniz, the two
interests
somewhat
diverge.
Contributions to symbolic logic were
made by
Ploucquet, Lambert, Castillon
and others on the continent.
This type of
contribution
research interested Sir William Hamilton and, though his
own
was
slight
and not
essentially novel, his papers were, to
some extent at
least, responsible for
the renewal of investigations in this field which took
place in
Boole.
England about 1845 and produced the work of De Morgan and Boole seems to have been ignorant of the work of his continental
predecessors, which
proved so
subject,
much
probably fortunate, since his own beginning has more fruitful. Boole is, in fact, the second founder of the
is
and
all later
work goes back
to his.
The main
line of this
develop
ment runs through Jevons, C.

Vorlesungen
iiber die
S. Peirce,
and MacColl to Schroder whose

I,
Algebra der Logik (Vol.
1890) marks the perfection
of Boole s algebra
and the logical completion of that mode of procedure. In the meantime, interest in logistic persisted on the continent and
fostered
by the growing tendency to abstractness and rigor in mathe matics and by the hope for more general methods. Hamilton s quaternions
was
and the Ausdehnungslehre of Grassmann, which was recognized as a con tinuation of the work begun by Leibniz, contributed to this end, as did also
the precise logical analyses of the nature of
Also, the elimination
number by Cantor and Dedekind.

of all
from
"modern
geometry"
methods
.of
proof
dependent upon
"intuitions
of
space"
or
"construction"
brought that
subject within the scope of logistic treatment, and in 1889
such a treatment in I Principii di Geometria.

Begri/sschrift of
Frege 1879 to the Grundgesetze der Arithmetik (Vol.
Peano provided works, from the

I,
1893;
Vol. II, 1903) provide a comprehensive

logistic
1
development
of arithmetic
by the
method.
or philosophic development of mathematics in Bk. vi, Step. 510-11
See the criticisms of contemporary mathematics and the program for the dialectic and Philebus, Step. 56-57.
In 1894, Peano and his collaborators began the publication of the
Formulaire de Mathematiques, in which
all
branches of mathematics were to

logistic.
be presented in the universal language of

logic
In this work, symbolic
and
logistic are once
more brought
together, since the logic presented
in the early sections provides, in a
way, the method by which the other
branches of mathematics are developed.

production.
logistic,
The Formulaire
is
a monumental
But
its
mathematical interests are as much encyclopedic as

the possibilities of the method are utilized or
and not
It
all
made
in
clear.
remained for Whitehead and Russell, in Principia Mathematica,
to exhibit the perfect union of symbolic logic and the logistic
method
mathematics.
The
publication of this
in the history of the subject. of the algebra of logic
work undoubtedly marks an epoch The tendencies marked in the development

in the
development of the and in the foundations algebra of relatives from De Morgan to Schroder, for number theory of Cantor and Dedekind and Frege, are all brought 2 Further researches will most likely be based upon the together here.
formulations of Principia Mathematica.
from Boole to Schroder,
We
history
must now turn back and

3
trace in
more
detail the
development
if
of
symbolic
is
logic.
history of the subject will not be attempted,
by
are
meant the report
of facts for their
own
sake.
Rather, we
interested in the cumulative process

interest us
by which those
results
which most
today have come to be. Many researches of intrinsic value, but lying outside the main line of that development, will of necessity be neglected. Reference to these, so far as we are acquainted with them, will
be found in the bibliography. 4
II.
LEIBNIZ
logistic properly begins
The
In the
say:
6
history of symbolic logic
and
with Leibniz. 5
New
"I
Essays on the Human Understanding, Philalethes is made to begin to form for myself a wholly different idea of logic from
I
that which I formerly had.
regarded
it
as a scholar
it,
diversion, but I
now
2
see that, in the
way you understand
it is like
a universal mathe-
Hilbert, Fieri,
Perhaps we should add "and the modern development of abstract geometry, as by and others", but the volume of Principia which is to treat of geometry has
not yet appeared. 3 The remainder of this chapter is not essential to an understanding of the rest of the book. But after Chapter i, historical notes and references are generally omitted. 4 Pp. 389-406. 5 Leibniz regards Raymond Lully, Athanasius Kircher, John Wilkins, and George Dalgarno (see Bibliography) as his predecessors in this field. But their writings contain
little
6
which is directly to the point. Bk. iv, Chap, xvn, 9.
A
As
matics."
this passage suggests, Leibniz correctly foresaw the general

it
character which logistic was to have and the problems

to solve.
would
set itself
But though he caught the
large outlines of the subject
and
actually delimited the field of work, he failed of

of the difficulties to be met,
any and he contributed comparatively

Perhaps
partial explanation,
is
clear understanding
little
to
the successful working out of details.
this is characteristic of the
man.
But another explanation, or
possible.
Leibniz
expected that the whole of science would shortly be reformed by the appli cation of this method. This was a task clearly beyond the powers of any
one man, who could, at most,

plan.
offer
life,
only the
initial
stimulus and general
And
and
so,
throughout his
he besought the assistance of learned
societies
titled patrons, to
the end that this epoch-making reform might
be instituted, and never addressed himself very seriously to the more

limited tasks which he
7 might have accomplished unaided.
Hence
of
his
studies in this field are scattered through the manuscripts,

still
many
them
unedited, and out
of five
hundred or more pages, the systematic
results
attained might be presented in one-tenth the space. 8
Leibniz
conception of the task to be accomplished altered somewhat
during his
life,
but two features characterize
all
the projects which he

language"
entertained:
language"
(1) a universal
"universal
medium
("universal
or
of
"rational
or
characteristic")
for the expression
science;
and
(2) a calculus of reasoning (or
"universal calculus")
designed to display
the most universal relations of scientific concepts and to afford some sys
tematic abridgment of the labor of rational investigation in all fields, much as mathematical formulae abridge the labor of dealing with quantity and
number.
that
is
"The
true
method should furnish us with an Ariadne
thread,
and palpable medium, which will guide the mind as do the lines drawn in geometry and the formulae for 9 operations which are laid down for the learner in arithmetic."
to say, with a certain sensible
This universal medium is to be an ideographic language, each single character of which will represent a simple concept. It will differ from
existing ideographic languages, such as Chinese, through using a
7
combina-
editor s introduction to "Scientia Generalis. in Gephardt s Characteristic Philosophischen Schriften von Leibniz (Berlin, 1890), vn, gives an excellent account of Leibniz s correspondence upon this topic, together with other material of historic interest. (Work hereafter cited as G. Phil.}
a"
The
See Gerhardt, op.
tit.
especially iv
will
(1901), gives a survey study of the sources.

9
which
and vn. But Couturat, La logique de Leibniz prove more profitable to the general reader than any
vii, 21.
Letter to Galois, 1677, G. Phil,

tion of symbols, or
some
similar device, for a
compound
idea, instead of
having a multiplicity of characters corresponding to the variety of things.
So that while Chinese can hardly be learned in a lifetime, the universal 10 The fundamental char characteristic may be mastered in a few weeks.
acters of the universal language will be few in
number, and
will represent
the
"alphabet
of
human
thought":
The
fruit of
many
analyses will be the
catalogue of ideas which are simple or not far from

catalogue of primitive ideas
of science
is
n
simple."
With
this
this alphabet of
human thought
the whole
to be reconstructed in such wise that its real logical organiza

its
tion will be reflected in
symbolism.
In spite of fantastic expression and some hyperbole,

the program of logistic.
If
we
recognize here
is
the reconstruction of
all
science
a project too
desirability
ambitious,
still
we should maintain the

its
ideal possibility
and the
of such a reconstruction of exact science in general.
And
the ideographic
language finds
matica, and
realization in
Peano
Formulaire, in Principia Mathe-
in all successful applications of the logistic
method.
Leibniz stresses the importance of such a language for the more rapid
and orderly progress

least effect of it
".
of science
. .
and
of
human thought
in general.
The
will
be the universality and communication of
different nations.
Its true use will
be to paint not the word
but the
.
. .
thought, and to speak to the understanding rather than to the eyes.
Lacking such guides, the mind can make no long journey without losing with such a medium, we could reason in metaphysics its way
.
and
in ethics
very
much
fix
as
we do
characters would
our ideas,
12
geometry and in analytics, because the which are otherwise too vague and fleeting
in
it
in such matters in
which the imagination cannot help us unless
be by
the aid of
characters."
"The
The
lack of such a universal
medium
prevents
cooperation.
human
race, considered in its relation to the sciences
which serve our welfare, seems to

in confusion in the darkness,
comparable to a troop which marches without a leader, without order, without any
me
word or other
one another.
of the road,
signs for the regulation of their
Instead of joining hands to
march and the recognition of guide ourselves and make sure

another."
13
we run
hither
and yon and

complex
(?),
interfere with one
The
10
11
"
"alphabet
of
human
of
thought" is
more
visionary.
The
possibility
of constructing the whole of a

Letter to the
science from a few primitive conG. Phil., vn, 24-25.
Duke
Hanover, 1679
G. Phil., vii, 84. G. Phil., vii, 21.
13
G. Phil, vii, 157.
A
is,
cepts
indeed, real
vide the
few primitives of Principia Mathematica.
But we should today recognize a certain arbitrariness in the selection of The these, though an arbitrariness limited by the nature of the subject.
secret of Leibniz s faith that these primitive concepts are fixed in the nature
of things will
believes that
all
be found in his conception of knowledge and of proof. He predicates are contained in the (intension of the) subject
and may be discovered by analysis. Similarly, all truths which are not absolutely primitive and self-evident admit of reduction by analysis into
such absolutely
thing
"real"
first
truths.
And
finally,
only one real definition of a

is
14
as opposed to
"nominal"
possible;
that
is,
the result
in
of the correct analysis of
any concept
is
unambiguously predetermined
the concept
itself.
The
and he
construction, from such primitives, of the complex concepts of

"synthesis"
the various sciences, Leibniz speaks of as

is
or
"invention",
concerned about the

is
"art
of
invention".
But while the
result of
analysis
synthesis,
character.
always determined, and only one analysis is finally correct, like inverse processes generally, has no such predetermined In spite of the frequent mention of the subject, the only im
of a calculus of reasoning.
portant suggestions for this art have to do with the provision of a suitable
medium and
essay,
To be
sure there are such obvious
counsels as to proceed from the simple to the complex,
and
in the early
De
Arte Combinatoria, there are studies of the possible permutations

"syntheses"
and combinations or
later regarded this
of
fundamental concepts, but the author

value.
Specimina he says that the utmost which we can hope to accomplish at present, toward the general art of invention, is a perfectly orderly and finished reconstruction of existing science in terms of the
Scientice novce Generalis,
study as of
little
And
in Initia et
absolute primitives which analysis reveals. 15
After
two hundred
years,
we
are
fields
without any general method by which logistic may be used in as yet unexplored, and we have no confidence in any absolute primi
still
tives for such investigation.
The
for the
calculus of reasoning, or universal calculus,
is
to be the instrument
development and manipulation

to get its complete generality
of systems in the universal language,
and
it is
from the fact that

of
all
science will be
expressed in the ideographic
symbols
that universal medium.
The
calculus will consist of the general principles of operating with such ideo14
See G. Phil, vn, 194, footnote.

G. Phil, vii, 84.
15

graphic symbols:
16
"All
our reasoning
stituting of characters,
images."
is nothing but the relating and sub whether these characters be words or marks or
Thus while the
characteristica universalis
is
the project of the
logistic
treatment of science
in general, the universal calculus is the pre
cursor of symbolic logic.
The plan
for this universal calculus
changed considerably with the
development of Leibniz s thought, but he speaks of it always as a mathe matical procedure, and always as more general than existing mathematical methods. 17 The earliest form suggested for it is one in which the simple
concepts are to be represented by numbers, and the operations are to be
merely those of arithmetical multiplication, division, and factoring. When, later, he abandons this plan of procedure, he speaks of a general calculus which will be concerned with what we should nowadays describe as "types
with combinations which are absolute or relative, symmetrical or unsymmetrical, and so on. 1 * His latest studies toward such a calculus
of
order"
form the
earliest presentation of
what we now
call
the
"algebra
of
logic".
But
it
is
doubtful
if
Leibniz ever thought of the universal calculus as
restricted to our algebra of logic:
we can only say
that
it
was intended
to
be the science of mathematical and deductive form in general (it is doubtful whether induction was included), and such as to make possible the appli
cation of the analytic
scientific
method
of
mathematics to
all
subjects of
which
knowledge is possible. Of the various studies to this end our
chief interest will be in the early
essay,
in the fragments which attempt to an algebra of logic. 20 develop Leibniz wrote De Arte Combinatoria when he was, in his own words,
De Arte Combinatorial and
vix egressns ex Ephebis,
and before he had any considerable knowledge

tells
of
mathematics.
consent.
It
was published, he
us,
without his knowledge or
The
it
intention of the work, as indicated
by
its title, is
it.
to serve the
general art of rational invention, as the author conceived
As has been
mentioned,
seems that this end
is
to be accomplished
by a complete
is
analysis of concepts of the topic under investigation and a general survey

of the possibilities of their combination.
large portion of the essay
concerned with the calculation of the possible forms of this and that type
16
G. Phil, vii, 31.
17
18 19
See New Essays on the Human Understanding, Bk. iv, Chap, xvn, 9-13. See G. Phil, vn, 31, 198 jf., and 204. G. Phil, iv, 35-104. Also Gerhardt, Leibnizens mathematische Schriften (1859),
Scientia Generalis.
Characteristica,
V,
1-79.
20
xv-xx, G.
Phil.,
vn.
10

etc.,
of logical construct:
the various dyadic, triadic,

of elements;
complexes which
can be formed with a given number

of the syllogism;
of the
moods and
figures
of the possible predicates of a given subject (the
com
the key to the predicates plexity of the subject as a concept being itself which can be analyzed out of it); of the number of propositions from a
given
number
of subjects, given
21
\
number
of predicate relations,
and given
number of quaestiones terms, and so on. In

of permutations to discover
of the variations of order with a given
number
of
fact so
much
space
is
occupied with the computation

of his contemporaries failed
and combinations that some
any more important meaning
of the essay,
and
it is
most
fre
22
quently referred to simply as a contribution to combinatorial analysis. Beyond this the significance of the essay lies in the attempt to devise a
symbolism which
will preserve the relation of
analyzable concepts to their
primitive constituents.
The
particular device selected for this purpose

is
representation of concepts
itself is of interest.
by numbers
unfortunate, but the attempt
Leibniz makes application of this method to geometry
and suggests
it
for other sciences. 23
In the geometrical illustration, the

Class
1
concepts are divided into classes.
consists of concepts or terms
regarded as elementary and not further analyzable, each of which is given a number. Thereafter, the number is the symbol of that concept. Class 2
consists of concepts analyzable into (definable in terms of) those of Class 1.
By
the use of a fractional notation, both the class to which a concept

its
belongs and
place in that class can be indicated at once.
nator indicates the
number
of the class
and the numerator
is
The denomi the number of

is
the concept in that class.

represented by 7/2.
those in Class
analysis of
1
Thus the concept numbered

2,
7 in Class 2
Class 3 consists of concepts definable in terms of
and Class
is
and
so on.
By
this
method, the complete

its
any concept
supposed to be indicated by
numerical symbol. 24
21 Leibniz tells us that he takes this problem from the Ars Magna of Raymond Lully. See G. Phil., v, 62. 22 See letter to Tschirnhaus, 1678, Gerhardt, Math., iv, 451-63. Cf. Cantor, Geschichte d. Math., m, 39 ff. 23 See the Synopsis, G. Phil, iv, 30-31.
24
See Couturat, op.
tit.,
appended Note
vi, p.
554 ff.
The concepts
are arranged as follows (G. Phil., iv, 70-72):
"Classis I; 1. Punctum, 2. Spatium, 3. intervallum, 4. adsitum seu contiguum, 5. dissitum seu distans, 6. Terminus seu quae distant, 7. Insitum, 8. inclusum (v.g. centrum est insitum circulo, inclusum peripheriae), 9. Pars, 10. Totum, 11. idem, 12. diversum, 13. unum, 14. Numerus, etc. etc. [There are twenty-seven numbered concepts in this class.] "Classis II; 1. Quantitas est 14 T&V 9 (15). [Numbers enclosed in parentheses have their usual arithmetical significance, except that (15) signifies an indefinite number .] 2. Indudens est 6.10. III. 1. Intervallum est 2.3.10. 2. Aequale A rijs ll.. 3. Continuum est A ad B, si TOV A j 9 est 4 et 7 B.; etc.
TU>
etc."
11
In point of fact, the analysis (apart from any merely geometrical defects) Leibniz uses not only the inflected Greek falls far short of being complete.
article to indicate various relations of
concepts but also modal inflections

etc.
indicated
by
et, si,
quod,
quam faciunt,
its
In later years Leibniz never mentions this work without apologizing for
it,
yet he always insists that
main intention
is
sound.
This method
of assuming primitive ideas which are arbitrarily symbolized, of introducing
other concepts by definition in terms of these primitives and, at the same

time, substituting a single symbol for the complex of defining symbols
this
is,
in fact, the
method
of logistic in general.
Modern
logistic differs
from
this
logistic
two respects: (1) modern attempt of Leibniz most would insist that the relations whereby two or more concepts are
notably in
while Leibniz regards his set of primitive concepts as the
united in a definition should be analyzed precisely as the substantives are
analyzed;
(2)
necessary result of any proper analysis, modern logistic would look upon
them
as arbitrarily chosen.
Leibniz
s later
work looks toward the elimina
tion of this first difference, but the second represents a conviction from
which he never departed.
At a much
later date
come various
studies (not in Gerhardt),
w hich
r
attempt a more systematic use of number and of mathematical operations in logic. 25 Simple and primitive concepts, Leibniz now proposes, should be
symbolized by prime numbers, and the combination of two concepts (the qualification of one term by another) is to be represented by their product.
Thus
if
3 represent
"rational"
and 7
"animal",
"man"
will be 21.
No
a
prime number
animal.
cation:
"man",
will enter
more than once
into a given combination

is
rational rational animal, or a rational animal animal,
simply a rational
Thus
logical synthesis
is
represented by arithmetical multipli
logical analysis
by
resolution into prime factors.

its
The
analysis of
"rational",
21,
would be accomplished by finding

7.
prime factors,
s
3,
and
"animal",
In accordance with Leibniz
conviction that
all
knowledge
is
analytic
"All
and
all
P"
valid predicates are contained in the subject,

will
the proposition
the concept
25
is
be true
if
the
number which
represents
is
divisible
by that which represents P.
Accordingly the
Dated
April, 1679.
"Elementa
as follows:
Couturat (op. cit., p. 326, footnote) gives the titles of these Characteristicae Universalis (Collected manuscripts of Leibniz in
the Hanover Library, PHIL., v, 8 b); Calculi universalis Elementa (PHIL., v, 8 c); Calculi universalis investigations (PHIL., v, 8 d); Modus examinandi consequentias per numeros (PHIL., v, 8 e); Regulae ex quibus de bonitate consequentiarum formisque et modis syllogis-
morum
many
categoricum judicari potest per numeros (PHIL., v, 8f)." These fragments, with others, are contained in Couturat s Opuscules et fragments inedits de Leibniz.
12
universal affirmative proposition
(where y
is
a whole number)
= Py may be symbolized by S/P = y or S By the plan of this notation, Py will represent

within the genus P,
is y.
some
species whose
"difference",
Similarly
Sx
Hence the particular affirmative, "Some will represent a species of S. S is P," may be symbolized by Sx = Py, or S/P = y x. Thus the uni versal is a special case of the particular, and the particular will always be
true
when the
universal
is
true.
this
There are several objections to
scheme.
In the
first
place,
it
the class as genus. presumes that any part of a class is a species within defensible on the ground This is far-fetched, but perhaps theoretically that any part which can be specified by the use of language may be treated
as a logical species.
worse defect
lies in
the fact that Sx
= Py
will
always be true.
will satisfy
For a given S and P, we can always find x and y which = P, the equation Sx = Py. If no other choice avails, let x
or
some multiple
"man-angel"
and y = S, or some multiple of S. although no men are angels. "Spineless

of P,
"Angel-man"
man"
"ra
tional
invertebrate",
but
it
is
false that
some men are invertebrates.
third difficulty arises because of the existential import of the particular a difficulty which later drew Leibniz s attention. If the particular affirma
tive
is
true,
then for some x and

4=
y,
Sx = Py.
The
But
universal negative should,
then, be Sx
Py.
And
since the universal affirmative
particular negative should be

practically
for all
will
4=
Py>
this
S = Py, the symbolism would be

is
unworkable because the inequations would have to be verified values of x and y. Also, as we have noted, the equality Sx = Py
4=
always hold and Sx
Py, where x and y are arbitrary, will never be

further,
true.
Such
difficulties led
Leibniz to complicate his symbolism
still
introducing negative numbers and finally using a pair of numbers, one But this scheme also breaks positive and one negative, for each concept. down, and the attempt to represent concepts by numbers is thereafter
abandoned.
in
Of more importance to symbolic logic are the later fragments included the plans for an encyclopedia which should collect and arrange all known
Leibniz cherished the
26 science as the proper foundation for future work.

26
G. Phil., vii, xvi-xx. Of these, xvi, without title, states rules for inference in terms of inclusion and exclusion; Difficultates quaedam logicae treats of subalternation and conversion and of the symbolic expression for various types of propositions; xvm, Specimen Calculi universalis with its addenda and marginal notes, gives the general prin ciples of procedure for the universal calculus; xix, with the title Non inelegans specimen
13
notion that this should be developed in terms of the universal characteristic.

In these fragments, the relations of equivalence, inclusion, and qualification by another, or combination, are defined and used. These
of one concept
relations are always considered in intension
when
it is
is
ing the calculus to formal logic.

concepts, not simply of
"Equivalence"
a question of apply the equivalence of

"for
two
classes
which have the same members;
to include
or
to be included in
27
is
to affirm the predicate
universally
However, Leibniz evidently considers the calculus to have many applications, and he thinks out the relations and illustrates them frequently in terms of extensional diagrams, in which A, B, etc., are
of the subject
A".
represented by segments of a right
line.
Although he preferred to treat

If
logical relations in intension, he frequently states that relations of intension
are easily transformed into relations of extension.

in intension,
is
included in
is
included in
in extension;
and a calculus
may
be inter
preted indifferently as representing relations of concepts in intension or

relations of individuals
and
classes in extension.
Also, the inclusion rela
tion
may
be interpreted as the relation of an antecedent proposition to a
consequent proposition.
just as the subject
The hypothesis A
all
includes
includes the predicate B.~ B
its consequence B, This accords with his

is
frequently expressed conviction that

these studies are by no
demonstration
analysis.
means
to be confined to the logic of intension.
Thus As
one
title
suggests, they are studies demonstrandi in abstracts.
demonsirandi in abstractis struck out, and xx, without title, are deductive developments of theorems of symbolic logic, entirely comparable with later treatises.
place of symbolic logic in Leibniz s plans for the Encyclopedia is sufficiently by the various outlines which he has left. In one of these (G. Phil., vu, 49), divisions 1-6 are of an introductory nature, after which come
The
indicated
scientiarum instauratione, ubi de Systematibus et Repertoriis, et de Encyclo paedia demonstrativa codenda. Elementa veritatis aeternae, et de arte demonstrandi in omnibus disciplinis ut in Mathesi.
"7.
De
"8.
De novo quodam Calculo generali, cujus ope tollantur eos qui in ipsum consenserit; est Cabala sapientum. "10. De Arte Inveniendi.
"9.
omnes disputationes
inter
"11.
"12.
"13.
De Synthesi seu Arte combinatoria. De Analysi. De Combinatoria speciali, seu scientia formarum,
sive qualitatum in genere (de
Characterismis) sive de simili et dissimili. "14. De Analysi speciali seu scientia quantitatum in genere seu de "15. De Mathesi generali ex duabus praecedentibus composita."
magno
et parvo.
Then
cine, psychology, political science, economics, military science, jurisprudence, theology, in the order named.
27 28
various branches of mathematics, astronomy, physics, biological science, medi and natural
G. Phil, vii, 208.

"
"Generates
Inquisitiones
(1686): see Couturat, Opuscules
etc.,
pp. 356-99.
14
It is a
failed to
frequent remark upon Leibniz s contributions to logic that he accomplish this or that, or erred in some respect, because he
chose the point of view of intension instead of that of extension.

facts are these:
The
Leibniz too hastily presumed a complete, or very close,
analogy between the various logical relations. It is a part of his sig nificance for us that he sought such high generalizations and believed in their validity. He preferred the point of view of intension, or connotation,
partly from habit and partly from rationalistic inclination.
As a conse
quence, wherever there
is
a discrepancy between the intensional and exis
tensional points of view, he
likely to overlook
it,
and to follow the former.
This led him into some

opposite inclination
difficulties
and choice
of example, but it also led
which he might have avoided by an him to make which has since been overlooked and
his
some
distinctions the importance of
to avoid certain difficulties into
which
commentators have
fallen. 29
In Difficultates quaedam logicae, Leibniz shows that at last he recognizes

the difficulty in connecting the universal and the corresponding particular.
He
sees also that this difficulty
is
connected with the disparity between the
intensional point of view

tions.
existential import of particular proposi In the course of this essay he formulates the symbolism for the four
and the
propositions in
two
All
different ways.
The
first
formulation
is:
30
Univ.
aff.;
is
B:
is
AB =
not B;
A, or
non-5 does not
exist.
Part, neg.;
Some
A
is
AB
4=
A, or
exist.
non-5
exists.
Univ. neg.;
Part,
aff.;
No A
Some A
B;
is
AB does not 5; AB exists.

men
AB = A
man"
and
AB
=}=
A may
be interpreted as relations of intension or of

are mortal, the intension of
"man",
extension indifferently.
is
If all
"mortal
the same as the intension of

is
and likewise the
class of
mortal
men
identical in extent
with the class of men.
The statements
concerning existence are obviously to be understood in extension only. The interpretation here put upon the propositions is identically that of
contemporary symbolic logic. With these expressions, Leibniz infers the subaltern and the converse of the subaltern, from a given universal, by
impossible,
For example, it led him to distinguish the merely non-existent from the absurd, or and the necessarily true from the contingent. See G. Phil, vn, 231, footn6te; and "Specimen certitudinis seu de conditionibus," Dutens, Leibnitii Opera, iv,
Part in, pp. 92
footnote.
30
.
29
,C
ff.,
also Couturat,
La Logique
de Leibniz, p. 348, footnote,
and
p. 353,
G. Phil, vii, 212.
15
means
of the hypothesis that the subject,
A,
exists.
Later in the essay, he

31
gives another set of expressions for the four propositions:

All
is
B:
is
AB =
not B:
A.
Some A
AB
4=
A.
Ens.
No A
is
B:
is
Some A
B:
AB does not exist, or AB 4= AB AB exists, or AB = AB Ens.

"in
AB before the sign of equality represents the AB s or the AB the region of possible represents existing AB s, or actual members of the class AB. (Read AB Ens, which AB = AB Ens thus represents the fact that the class AB has members; AB AB Ens, that the class AB has no members. A
In the last two of these,
ideas";
"AB
Ens"
"AB
exists".)
=f=
logical species of the genus
A,
"some
A",
may
be represented by
"some
YA;
exist
YA Ens
ing
--
will represent existing
members
of that species, or
if
A".
Leibniz correctly reasons that
AB = A
(All
is
B),
YAB
A is B); but if AB 4= A, it does not follow that YAB 4= YA, YA. Again, if AB 4= AB Ens (No A is B), YAB B, YAB 4= YAB Ens (It is false that some A is B); but if AB = AB Ens (Some A is B), YAB = YAB Ens does not follow, because Y could assume values incompatible with A and B. For example, some men are wise, but it does
YA
if
(Some
=
for
=--
not follow that foolish
men
32
are foolish wise persons, because
"foolish"
is
incompatible with
division of
"wise".
The
distinction here between
AB,
a logical
A
s
or of B,
and
AB
Ens, existing
AB s,
is
ingenious.
This
is
our author
intension,
most
successful treatment of the relations of extension
and
are
and
of the particular to the universal.

"principles
In Specimen calculi universal, the announced as follows: 33

1)
"Whatever is
of the
calculus"
concluded in terms of certain variable letters
may
be
concluded in terms of any other letters which satisfy the same conditions; for example, since it is true that [all] ab is a, it will also be true that [all]
be
is
and that
[all]
bed
is be.
2)
"Transposing
coincides with ba,

3)
"Repetition
"One
letters in terms changes nothing; animal rational with rational animal
for
.
example ab
of a letter in the
same term
is
useless.
4) proposition can be made from any number by joining all the subjects in one subject and all the predicates in one predicate: Thus, a is b and c is d and e is /, become ace is bdf.
.
31 32
G. Phil,, vii, 213-14. G. Phil., vii, 215: the illustration

G. Phil, vii, 224-25.
is
mine.
33
16
A
5)
"From
any proposition whose predicate is composed of more than one term, more than one proposition can be made; each derived proposition having the subject the same as the given proposition but in place of the
given predicate some part of the given predicate. u a is b and [all] a is c and [all] a is
d."
If [all]
is
bed,
then
[all]
If
we add
ab
is
to the
of
number
of these,
two
principles
(1)
which are announced

a
is
under the head
"self-evident
propositions"
included in a;
and
(2)
included in a
we have
logic.
here the most important of the funda

Principle 1 or of
is
mental principles of symbolic
usually qualified
of
by
some doctrine
of the
"universe
of
discourse"
"range
significance",
but some form of
it is is
indispensable to algorithms in general.

call
The law
the
numbered 2 above
3, the
"principle
what we now
the
"principle
of permutation";
5,
of
tautology";
4, the
"principle
of composition";
"principle
of
division".
And
the two
"self-evident propositions"
are often
included in sets of postulates for the algebra of logic.
There remain
for consideration the
two fragments which are given
in
translation in our Appendix,

acteristica.
sir andi
XIX
and
XX
of Scientia Generalis:
Char-
The
first
of these, with the title
Non
inelegans specimen
is
demoninter
in abstractis, stricken out in the manuscript,
rather the
or ab
is
more
esting.
Here the relation previously symbolized by
AB
represented
by A+B. And A+B = L signifies that A is contained or included in A scholium attached to the definition of this inclusion relation (est in} B.
distinguishes
from the part-whole relation. Comparison of this and other passages shows that Leibniz uses the inclusion relation to cover
it
(1)
the relation of a
member
of the class to the class itself;
(2)
the relation
(3)
of a species, or subclass, to its genus
a relation in extension;
a relation of intension.
the rela
first
tion of a genus to one of its species
The
of
these
is
our e-relation;
is
(2) is
the inclusion relation of the algebra of logic;
and
(3)
the analogous relation of intension.

it is
Throughout both these

theorems
in
fragments,
clear that Leibniz thinks out his
terms of
extensional diagrams, in which classes or concepts are represented by segments of a line, and only incidently in terms of the intension of concepts.
The
different interpretations of the

If
tinguished.
is
"rational"
and
is
symbols must be carefully dis "animal", and A and B are taken
m intension,
then
A+B
will represent
"rational animal".
But
if
and
are classes taken in
extension^ then
A B
+
is
the class
made up
of those
things which are either

34
or
(or both).
Thus the
inclusion relation,
4.
and
5.
ties of
universal affirmative propositions.
are stated without qualification because this study is confined to the proper 4. is true also for universal negatives.
17
A +B =
in
L,
may
is
".
This
is be interpreted either in intension or in extension as little confusing to us, because we should nowadays invert
"A
the inclusion relation
when we pass from

meaning
+
of
intension to extension;
instead
B"
of this, Leibniz changes the
A +B
from
If
"both
and
(in
intension) to
mal",
"either
or
B"
(in extension).
A
If
is
"rational",
"ani
and
"man",
then
B = L
is
true in intension,
"man".
"rational animal"
"man"
or
"rational"
is
contained in
A, B, and
L are
classes
of points, or segments of a line, then

class of points
comprising the points
A + B = L will mean that L is the in A and the points in B (any points
common
segments
to
and
counted only once), or the segment made up of
and B.
The
relation
If
A +B
is
exclusive.
does not require that A and B should be mutually a line, A and B may be overlapping segments; and, in
intension,
and
B may
be overlapping concepts, such as
"triangle"
and
"equilateral",
Leibniz also introduces the relation
each of which contains the component "figure". L A, which he calls

that
detractio.
A =
+
N signifies
is
contains
and that
[ ]
if
be taken from
the
if
remainder
N.
The
relations [+] and
are not true inverses:
B =
L,
it
does not follow that

Leibniz
s
==
B, because
.
overlapping
(in
terms, communicantia)
If
A and B may be A = N, A and N

if
must be mutually exclusive (incommunicantia} A and B have a common part, M, L A = B

.
Hence
A+B
= L and
M.
(If
the reader will
take a
clear.)
line,
L, in which
and
are overlapping segments, this will be
This makes the relation of detractio somewhat confusing.
In
extension,
it
is
A may
and
calls
be interpreted
"L
which
is
not
".
In intension,
"rational"
more
difficult.
Leibniz offers the example:
"man"
=
is
"brute",
our attention to the fact that

or
"man"
"man"
"rational"
not
"non-rational man"
"non-rational".
35
In intension, the
relation seems to indicate an abstraction; not a negative qualification.
But there
"
are difficulties, due to the overlapping of concepts.

"woodworking"
Say that
pp. 377-78)
man"
35
r
"carpenter"
w^~
and
"man"
"white-skinned"
-^^
&
G. Phil., vn, 231, footnote.
Couturat in commenting on
this (op.
cit.,
says:
"Ailleurs
Leibniz essaie de preciser cette opposition en disant:
"Mais il oublie que le ne"ant Nihilum. Sed A non-A est Absurdwn. (non-Ens) n est pas autre chose que ce qu il appelle 1 absurde ou 1 impossible, c est-a-dire
est
le contradictoire."
Non-existence may It may be that Couturat, not Leibniz, is confused on this point. be contingent, as opposed to the necessary non-existence of the absurd. And the result of abstracting A from the concept A seems to leave merely non-Ens, not absurdity.
-
18
A
"Caucasian".
Then
"
Caucasian
"
"
"
carpenter
"man"
"white-
skinned"
penter"
"woodworking".
"white-skinned",
Hence
("Caucasian"
"carpenter")
"car
because the
common
constituent
is,
"man"
has
been abstracted in abstracting

"carpenter"
"carpenter".
That
which
the abstraction of
"Caucasian"
from
"Caucasian
carpenter"
leaves, not
is
but
only that part of the concept

"carpenter".
"Caucasian"
wholly absent in
because
is
"man"
We
cannot here say
"white-skinned
man"
is
abstracted, nor
"man":
"white-skinned animal"
because
"animal"
contained
in
we can only say

is difficult
"white-skinned"
as a pure abstraction.
Such
abstraction
to carry out
and
of very little use as
an instrument
of logical analysis.
manuscript, and
his
theorems
in
Leibniz s illustration is scribbled in the margin of the seems clear that at this point he was not thinking out terms of intensions.
it
Fragment by
[
XX differs from XIX in that it lacks the relation

is
].
This
a gain rather than a

]
loss,
[ ]
and because [+ and XX is more carefully developed more of the simple theorems are proved, and more illustrations are given. Otherwise the definitions, relations, and
interpretation
:
symbolized both because of the difficulty of are not true inverses. Also
methods
of
proof are the same.
In both fragments the fundamental

is
operation by which theorems are proved

expressions.
If
the substitution of equivalent
the successors of Leibniz had retained the breadth of view which

7
characterizes his studies
and aimed to symbolize
relations of a like generality,
these fragments might well have proved sufficient foundation for a satis
factory calculus of logic.
III.
FROM LEIBNIZ TO DE MORGAN AND BOOLE

w ere made to develop a
r
After Leibniz, various attempts

logic.
calculus of
Segner, Jacques Bernoulli, Ploucquet, Tonnies, Lambert, Holland,
Castillon,
and
others, all
made
studies toward this end.
Of these, the
most important are those

of Holland s
in extension.
is
of Ploucquet,
Lambert and
it
Castillon, while one
of particular interest because

this
intends to be a calculus
success,
But
is
attempt was not quite a

either
and the net
result
of the others
to illustrate the fact that a consistent calculus of logical

is
relations in intension
most
difficult or quite impossible.

s
Of Segner
36
w ork and Ploucquet

r
36 copies of these writings are available.
we can give no account, since no Venn makes it clear that Ploucin this country,
secure
There seem to be no copies of Ploucquet s books them from the continent have so far failed.
and attempts to
19
quet
calculus
was a calculus
of intension
and that
it
involved the quanti
fication of the predicate.
Lambert 37 wrote voluminously on the subject

important contribution to symbolic procedure
is
of logic, but his
most
contained in the Seeks
Versuche einer Zeichenkunst in der Vernunftlehre.* 8

separate studies,
These essays are not

later essays presuppose
made from
different beginnings;
those which precede and refer to their theorems; and yet the development
is
not entirely continuous.
Material given briefly in one
will
be found
set forth
more
at length in another.
And
discussion of
lems of the theory of knowledge and of scientific introduced. But the important results can be presented as a continuous
more general prob method are sometimes
development which follows
in general the order of the essays.

list
Lambert
gives the following
of his symbols:
The symbol
of equality (Gleichgultigkeit)
=
+
addition (Zmeizung)
abstraction (Absonderung)
opposition (des Gegentheils)
universality
particularity
copula
given concepts (Begriffe)
a, b, c, d, etc.
undetermined concepts
n,
m,
I,
etc.
unknowns
the genus
the difference
x, y, z.
7
5
developed entirely from the point of view of intension: the letters represent concepts, not classes, + indicates the union of two
calculus
is
[
The
concepts to form a third,
represents the withdrawal or abstraction of
some part
term
of the connotation of a concept, while the product of a
and
represents the
common
part of the two concepts.
"multiplied"
into them.
difference of a.
Much
use
is
qualify any Thus ay represents the genus of a, ad the made of the well-known law of formal logic
8
7 and
that the concept (of a given species) equals the genus plus the difference.
(1)
37
ay + ad = a(y +
6)
Johann Heinrich Lambert (1728-77), German physicist, mathematician, and astrono He is remembered chiefly for his development of the equation x n +px = q in an infinite series, and his proof, in 1761, of the irrationality of TT. 38 In Logische und philosophische Abhandlungen; ed. Joh. Bernoulli (Berlin, 1782),
mer.
vol.
i.
20
A
is

a.
ay + a8
the definition or explanation (Erkldrung) of
As immediate
consequences of (1),
(2)
we have
also
(3)
]
ay = a
it
ad
ad
]
ay
Lambert takes
tions.
If
for granted that [+
and
are strictly inverse opera
We
+
b.
two concepts,
(a
b)
then
have already noted the difficulties of Leibniz on this point. a and b, have any part of their connotation in common, b will not be a but only that part of a which does not belong
"European"
"+
also to
If
and
"carpenter"
have the
common
is
part
"man",
then
("
European
"carpenter")
"man".
minus
"carpenter"
]
not
"European"
but
"European"
minus
And
[+
and
will
not here be true
inverses.
But this difficulty may be supposed to disappear where the terms of the sum are the genus and difference of some concept, since genus anddifference may be supposed to be mutually exclusive. We shall return
to this topic later.
More complex laws

fact that the genus of
of
genus and difference

is
may
be elicited from the

"explained,"
any given a
also a concept
and can be
as can also the difference of a.

(4)
a
Proof:
= a(y+
2
<5)
= ay
+ ayd + ady + ad 2
ad
ay = ayy + ayd
ad.
and
= ady +
add
But a = ay +
That
is
Hence Q.E.D.
to say:
giving
its
one wish to define or explain a, one need not stop at genus and difference, but may define the genus in terms of its
if
genus and
difference,
and define the difference
similarly.
Thus
is
equiva
lent to the genus of the genus of a plus the difference of the genus of a plus
the genus of the difference of a plus the difference of the difference of
a.
This
may
be called a
"higher"
definition or
"explanation"
of a.
Obviously, this process of higher and higher "explanation" may be carried to any length; the result is what Lambert calls his "Newtonian
formula".
We
shall best
understand this
if
we take one more preliminary
step. Suppose the explanation carried one degree further and the resulting terms arranged as follows:
3 a(7 + 776 + ydd +
6 )
+ 767 + dyd
+ dyy + ddy
The
three possible arrangements of
two y
s,
and one
might be summarized
21
the three arrangements of two 6 s and one 7 by 8752. With this convention, the formula for an explanation carried to any degree, n, is:
by 3y
d;
This
"Newtonian formula" is
Two
(6)
a rather pleasant mathematical conceit. further interesting laws are given

:
a
Proof:
ad + ayd + ay 2 d + ay*d +
etc.
= ay +
ad
But
and
2 ay = ay + ayd 2 2 ay = ay* + ay d
ay*
(7)
= ay 4 + ay*d,
2
etc. etc.
. .
a
Proof:
= ay + ady + ad y + ad*y + a = ay + ad
ad
etc.
But
and
ad 2
ad*
= ady + ad = ad y + ad*
2 2
=
is
ad*y + ad*, etc. etc.
Just as the genus of a
represented by ay, the genus of the genus of
a by ay 2
n
etc., so a species of
which a
is
genus
may
be represented by ay~
2
,
l ,
and a species
ay
genus of the genus by ay~ etc. In general, as a. genus above a, so a species below a may be represented by represents
of
is
which a
ay~
a
or
yn
Similarly afny concept of which a
is
difference of the difference of the differ
ence
etc.,
may
be represented by
ad~ n
or
--
Also, just as a
tion",
a(y +
n d)
,
where a
is
a concept and a(y +
d)
its "explana
so-
a,
where
^n ls ^ ne concept and a
the
"explanation"
of
it.
Certain cautions in the transformation of expressions, both with respect

to
"multiplication"
39
and with respect to

5.
"division,"
need to be observed. 40
40
Seeks Versuche, p. Ibid.j pp. 9-10.
22
A
is
The concept ay 2 + ady

(8)
very different from the concept (ay + ad)y, because
(ay + ad)y
2 7 + ady
= a(y+d)y = ay(y+d) = ay
while
of a.
is
the genus of the genus of a plus the genus of the difference

distinguished from
i.
Also
- y must be
7
which a
is
77
.
- y
is
the genus of any
species x of
the genus,
e.,
(9)
-7 =
But ay/y
is any species of which the genus of a is the genus, x such that a and x belong to the same genus. species
i.
e.,
any
We
turn
now
is
to consideration of the relation of concepts
which have a
and
common
in
part.
Similarity
so far as,
identity of properties.
Two
concepts are similar
if,
they comprehend
identical properties.
In respect to the
41 remaining properties, they are different.
ab represents the
common
properties of a and
a.
b.
a
a + b
ab represents the peculiar properties of

ab
ab represents the peculiar properties of a together with

b.
the peculiar properties of

It is evident
from
this last that
Lambert does not wish
to recognize in
b
his
system the law a + a a; else he need only have written a + If x and a are of the same genus, then
ab.
xy = ay
If
and
ax
= ay = xy
is
now we symbolize by a
b that part of a
which
different
from
42
6,
then
(10)
a\b + b\a + ab + ab
=
x
a+ b
Also
x
ax
a
x a
ay,
or
= ay + x\a
ad
ax
ad
ax + ad
a
ad
ax
41 42
= ay = xy
sometimes a
:
Ibid., p. 10.
Lambert sometimes uses a
b for this,
b.
23
And
since
x
ax + a x ax
= ay
ax + x\a
a\x
a ax
x\a
x\a
a\x
=
is
ax
The
y
fact that y
a property comprehended in x
= xy
or
by y + x y
x.
The manner
first is
in
may be expressed by which Lambert deduces the

43
second of these expressions from the of x, then y x is null. But by (10),
interesting.
If
is
a property
2xy + x\y + y\x
x+y
Hence
in this case,
2xy + x y
\
=
=
x+y x +y
And
since y
xy,
2y + x\y
Hence
y+x\y =
This
is
He
2y
has subtracted y from both

y
sides, in the last step,
and we observe that

it
=
y>
rather characteristic of his procedure;
follows,
throughout, arithmetical analogies which are quite invalid for logic.
With the complications

little
of this calculus, the reader will
probably be
concerned.
There
is
no general type of procedure
for elimination or
solution.
Formulae
of solution for different types of equation are given.
They
are highly ingenious, often complicated,
and
of dubious application.
It is difficult to
of the development, so far as outlined, there
judge of possible applications because in the whole course is not a single illustration of a
solution which represents logical reasoning,
and very few
illustrations of
any kind.
The shortcomings
of this calculus are fairly obvious.
There
is
too
much
and
reliance
upon
the analogy between the logical relations symbolized
their arithmetical analogues.
Some
of the operations are logically
uninterpretable, as for
example the use of numerical coefficients other than

in the "Newtonian
formula",
and
1.
These have a meaning
but 2y either
has no meaning or requires a conventional treatment which is not given. And in any case, to subtract y from both sides of 2y = x + y and get y = x
represents no valid logical operation.
of the relations
b,
43
employed
is
lacking,
Any adequate study of the properties x = a + b is transformed into a = x

b
regardless of the fact that a

Seeks Versuche, p. 12.
and
may have
common
part and that
24
x
A
man =
b represents the abstraction of the whole of b
from
x.
s
Suppose, for
procedure,
example, should have also rational

out this
difficulty,
rational + animal.
Then, by Lambert
animal.
we
= man
Since Leibniz had pointed
that addition and subtraction

it is
meanings) are not true inverses, should err in this.
(with exactly these the more inexcusable that Lambert
There
is
still
deeper difficulty here.

"
As Lambert himself remarks, 44
no two concepts are so completely dissimilar that they do not have a

part.
"
common
"
One might say that the concept

is
thing"
ing
common to every pair of concepts.

a<5;
word) or be This being the case, + and
(Lambert
are never really inverse operations.
Hence the
and a
difficulty will
not really
a8
disappear even in the case of ay and

will
ay =
ad, a
= ay
not be strictly valid.

"subtraction"
use of
of a
in a calculus
In fact this consideration vitiates altogether the based on intension. For the meaning
[ ]
ventional inverse of
becomes wholly doubtful unless be treated as a wholly con + ], and in that case it becomes wholly useless. The method by which Lambert treats the traditional syllogism is only
b
[
remotely connected with what precedes, and
its
value does not entirely
depend upon the general validity of his calculus.
He
reconstructs the
whole of Aristotelian logic by the quantification of the predicate. 45 The proposition "All A is has two cases:
B"
(1)
A =
B, the case in which
it
has a universal converse, the concept

particular, the concept
is
identical with the concept B.
(2)
>
B, the case in which the converse

in the
is
comprehended concept A. The particular affirmative similarly has two cases (1) A B, the case in which the converse is a universal, the subject comprehended within the predicate B.
:
<
(2)
The
case in which the converse
is
particular.
In this case the
comprehended within a subsumed species of the predicate and the predicate within a subsumed Lambert says species of the subject.
subject
is
this
may
be expressed by the pair:
mA
not
>
and
<
nB
must
<
Those who are more accustomed to
logical relations in extension
make the mistake

is
44
here of supposing that
A
is
>
mA, and
mA
A.
mA
a species of A, and in intension the genus

pp. 93
jf.
contained in the species,
Ibid., p. 12.
45
7m,

not vice versa.
at first glance.
species
25
Hence
mA
in
>
does not give
>
B, as one might expect

"Some
We
see that
comprehended
by mA, a A, making the same assumption which occurs in

or portion of a class
is
Lambert here
translates
"
Leibniz, that
any subdivision
capable of being treated
as
some
species
comprehended under that
class as its genus.
Lambert says the subject and predicate each have peculiar properties by virtue of whose comprehension neither is contained in the other. But if the peculiar properties of the subject be taken away, then what remains is contained in the predicate; and if the peculiar properties of the predicate be taken away, then what
remains
is
In a universal negative proposition
contained in the subject.
Thus the universal negative
is
repre
sented by the pair
m
The
is
and
A>
B n
particular negative has
two
cases:
i.
(1) When it has not B but all B is
a universal affirmative converse,
e.,
when some
A.
This
is
expressed by
.4
<B
(2)
When
it
has not a universal affirmative converse.

is
In this case a
contained in the predicate, and a sub sumed species of the predicate in the subject.
subsumed
species of the subject
mA
Either of the signs,
terms.
<
>
B
>,
and
<
nB
by transposing the
"
and
>
may
be reversed
/,
And
if
<
Q,
P, then for some
P =
IQ.
Also,
multi
plication"
and
"division"
are strict inverses.
Hence we can transform
these expressions as follows:
>
is
equivalent to
A = mB nA = B
kB]
]
mA =
IA A_
= nB
or
pA = qB
_B
k
or
m
A =
"
B_
It is
evident from these transformations and from the prepositional equiva-
26
lents of the
"
A
inequalities"

that the following
is
the
full
expression of these
equations:
(1)
A = mB:
All
is
and some
is
not A.
(2)
(3)
nA = B: Some A is not B and all B is A. mA = nB: Some, but not all, A is B, and some, but
not
all,
is
A.
(4)
-=-
No A
is
B.
that
The
first
noticeable defect here

(4)
is
A/m = B/n
is
transformable into
nA = mB and
fact,
can mean nothing different from (3). Lambert has, in four different propositions, if he sticks to the laws of his calculus: only
(1)
(2) (3) (4)
A = B: All A is all B. A = mB: All A is some B. nA == B: Some A is all B. mA = nB: Some A is some
B.
These are the four forms which become, in Hamilton s and De Morgan s A little scrutiny will show that treatises, the four forms of the affirmative.
Lambert
all, it is
treatment of negatives
necessary that
"
is
a failure.
For
it
to be consistent at
should not be transformed. But Lambert constantly makes such transformations, though he carefully re frains from doing so in the case of expressions like A/m = B/n which are supposed to represent universal negatives. His method further requires that m and n should behave like positive coefficients which are always
fractions"
greater than
and such that
4= n.
This
is
unfortunate.
It
makes
it
impossible to represent a simple proposition without

If
"entangling alliances".
he had taken a leaf from Leibniz
as affirmatives
book and treated negative propositions with negative predicates, he might have anticipated the
s
calculus of
De Morgan.
In symbolizing syllogisms, Lambert always uses A for the major term, for the middle term, and C for the minor. The perfectly general form of
is:
proposition
mA
P
Hence the
nB
(
1
46
:
perfectly general syllogism will be
AT
46
mA
p
nB
q
Major
Ibid., pp. 102-103. Ibid., p. 107.
47
27
Minor
nC
vB
Conclusion
fj.n
C =
mv
pp
irq
The indeterminates
letters.
in the
minor are always represented thus by Greek

as follows:
The conclusion
is
delved from the premises
The major premise
gives
B =
np
A.
The minor
gives
B =
irv
C.
Hence
np
A-C.
irv
mv and therefore -- C = - A. irq pp

The above being the general form of the moods in the first figure is the following:
classification only so far as indicated
syllogism,
it
Lambert
scheme of
coincides with the traditional
by the use of the traditional names:
28
nB = A
IV.
Fideleo
C
7T
X.
Pilosos
nB = A V.C = B
nC
= A
XL
Romano
nB = mA C = B nC = mA
nB = mA C = B
7ijjiC
XII.
Somnio
= mA
The
since
difficulty
about
"division"
it is
only required that
if
does not particularly affect this scheme, one of the premises involve fractions",
"
the conclusion must also.

cal in
lies in
It will
be noted that the

Lilii
mood
Hilario
is
identi
form with Romano, and

the fact that
with Somnio.
The reason
for this
nB =
mA
has two partial meanings, one affirmative

Hilario and
Lilii
and one negative
(see
above).
take the affirmative
interpretation, as their
the negative. indicate; Into the discussion of the other three figures, the reader will probably not care to go, since the manner of treatment is substantially the same as
in the above.
names
Romano and Somnio,
method
There are various other attempts to devise a convenient symbolism and for formal logic; 48 but these are of the same general type, and
they meet with about the same degree and kind of success. Two brief passages in which there is an anticipation of the logic of
relatives possess
attributes",
some
interest. 49
Relations,
Lambert
says, are
"
external
by which he means that they do not belong

(i.
to the object
an
"
sich.
Metaphysical"
e.,
non-logical) relations are represented
by
Greek
letters.
For example
if
=
/
fire,
h
: :
=
h
heat,
and a = cause,
The symbol
48
represents a relation which behaves like multiplication:

vi.
See in Seeks Versuche, v and
Also fragments
i,
"Uber
die Vernunftlehre", in
Logische und Philosophische Abhandlungen,

p.
xix and xx; and Anlage zur Architektonik,
190 jf.
49
ec/is
Versuche, pp. 19, 27 ff.

h is in fact what Peirce and Schroder later a Lambert transforms the above equation into:
: :
29
"relative product".
called a
Fire
h
is
to heat as cause to effect.
Fire
is
to cause as heat to effect.
-*
<*
Heat
is
to
fire
as effect to cause.
The dot
here represents Wirkuny (it might be, Wirklichkeit, in consonance with the metaphysical interpretation, suggestive of Aristotle, which he It has the properties of 1, as is illustrated elsewhere 50 gives to Ursache).
by the fact that 7 may be replaced by this symbol. Lambert also uses powers of a relation.
If
=
<p
b,
and
=
<p
c,
=
<p
(p
(p-
And
if
=
<p~
c,
o
<p-
a c
j and
V7
a \|-
And more to the same effect. No use is made of this symbolism;

Lambert could have used
it.
indeed
it
is
difficult to see
how
felt
Yet
it is
interesting that he should have
that the powers of a relation ought to be logically important, and that he here hit upon exactly the concept by which the riddles of "mathematical
induction"
were later to be solved.

s
Holland
bert.
51
attempt at a
it
logical calculus
is
contained in a letter to
Lam
He
himself calls
an
"unripe thought",
and
in a letter
some three
52 years later he expresses a doubt
if
logic
is
really a purely formal discipline
capable of mathematical treatment.
But
this
study
is
of particular interest
because
it
treats the logical classes in extension
the only attempt at a
symbolic logic from the point of view of extension from the time of Leibniz
to the treatise of Solly in 1839.
Holland objects to Lambert

concepts by the relation of
50
61
method
of representing the relation of
lines,
one under the other, and argues that the
Ibid., p. 21.
52
Joha n. Lamberts deutscher Gelehrten Brief wechsel, Brief m, pp. 16 ff. See Ibid., Brief xxvn, pp. 259 ff.
30
"
A
"men"
relation of
to
mortals" is
not sub but
inter.
He
is
apparently not
aware that
this
means exchanging the point
of view of intension for that
of extension, yet all his relations are consistently represented in extension,
as
we
(1)
shall see.
If
S represent the
subject,
the predicate; and p,
TT
signify
is
unde
a part a P.
is
termined variable
numbers, S/p =
P/TT will
come
to:
part of S
of P, or certain of the
are certain of the
P s,
all
or (at least) an
is
This expression
is
the general formula of
possible judgments, as
evident by the following:

(2)
A member
If
is
either positive or negative,

shall see in
is
and
in
IT
both
cases,
is
either
finite or infinite.
(3)
We
what fashion p and
can be understood.
p =
less
attains
it
its logical
can
still
all S, and in this way S/p cannot become less than 1, Since, then, p disappear and consequently cannot become negative.
1
in S/p, then
S/p as
many
as
maximum.
TT.
The same
(4)
is
true of
Therefore p and
TT
TT
cannot but be positive and cannot be
less
than
If
or
(5)
infinite, the concept becomes negative. If /expresses a finite number 1, then the possible forms of judg
>
becomes
ment
are as follows:
All
(1)
--=
^
--=
^
j
is all
P.
(2)
All
is
some P.
Now
an
expresses negatively
what
l/oo expresses positively.

is
To
say that
infinitely small part of a curved line

is
straight,
means
exactly:
No
part
of a curved line
straight.
(3)
~ =
1
All
CO
is
not P.
(4)
j-
-.
Some S
p
is all
P.
o
(5)
- =
Some S
is
some P.
(6)
S - = -P
Some S
is not-
P.
(7)
53
oo
=^ 1
All not-S
is all
P.
See
Ibid., Brief iv.
31
S
(8)
P j
P
All not-S
is
some P.
S
(9)
All not-S
is all
not-P.
(1),
(2),
and
and
(9)
(8),
(3), (7), (6),
says are universal affirmative propositions; universal negatives; (4) and (5), particular affirmatives;
Holland
a particular negative.
As Venn has
Boole.
If
said, this notation anticipates, in a
way, the method of

of the
instead of the fraction

it,
we take the value
numerator
indicated by
the three values are
where
Boole
is
<
<
1,
and
/co
-.=
Q-S.
But the
differences between this
and
form
procedure are greater than the resemblances.

unfortunate in that
it
The
fractional
little
suggests that the equations

results
may
be cleared
of fractions,
and
s
this
would give
which are
logically uninterpretable.
But Holland
calculus.
notation can be
That he did
made the basis of a completely successful not make it such, is apparently due to the fact that
he did not give the matter sufficient attention to elaborate the extensional
point of view.
He
gives the following examples:
Example
1.
All All
men
H are mortal M
E are men H
Europeans
P
Ergo,
[All
Europeans are mortal]
Example
2.
All plants are organisms
P =
P
All plants are
no animals
oo
Ergo,
- =
A
-
[Some organisms are not animals]
32
A
Example
3.
All
men
are rational
//
P
T>
All plants are not rational
P =
CO
Ergo, All plants are no
men
vH P=00
In this last example, Holland has evidently transformed //
= R/p
into
pH =
"Some
R, which
is
not legitimate, as
the rational
It
men
are
all
we have noted. pH = R would be And the conclusion P = pH/ao beings".

"All
is
also misinterpreted.
should be,
plants are not
some
men".
correct reading
would have revealed the invalid operation.
Lambert
replied vigorously to this letter, maintaining the superiority
method, pointing out, correctly, that Holland s calculus would not distinguish the merely non-existent from the impossible or
contradictory (no calculus in extension can), and objecting to the use of c in this connection. It is characteristic of their correspondence that each
of the intensional
pointed out the logical defects in the logical procedure of the other, and neither profited by the criticism.
Castillon s essay toward a calculus of logic
is
contained in a paper
presented to the Berlin
Academy in
1S03. 54
The
letters S,
concepts taken in intension,

synthesis"
A,
etc.,
represent
is
an indeterminate, S +
the withdrawal or
of
S and M, S
from
S.
M thus represents a genus concept in which S

"difference"
- M,
M represents the abstraction of M

is
subsumed,
being the logical
of
in
- M.
Consonantly S + M,
to S, represents a
symbolizing the addition of some

species concept
"further specification"
which contains
(in intension)
the concept S.
is
The
predicate of a universal affirmative proposition
contained in the
subject (in intension).
Thus
"All
is
A"
is
represented by
S =
A
is
M
+
is
The
universal negative
"No
S
-
is
A"
symbolized by
= (- A)
M
withdrawn
classes,
is
The concept S
lusory".
is
something,
M, from which A
no A.
Particular propositions are divided into
two
"real"
and
"il
real particular is the converse of a universal affirmative;

sur
the
"Memoire
un nouvel algorithme
logique",
in
Memoires de V Academic des Sciences

See also his paper,
"Reflexions
de Berlin, 1803, Classe de philosophie speculative, pp. 1-14. sur la Logique", loc. cit., 1802.

illusory particular, one
lar affirmative is
33
whose converse
also
is
particular.
The
real particu
A = S since this
is
is
M
The
illusory particular affirmative
the converse of S
= A
M.
represented by
S =
A=?
M
A
which
Castillon s explanation of this is that the illusory particular judgment gives us to understand that some S alone is A, or that S is got from A by ab
straction (S
= A - M), when
==
abstraction (S
M + A).
in reality
this
it
it is
is
Thus
put +
M]
one can, then, indicate

is,
judgment puts by S = A ^ M.
M where
drawn from S by
it
should
The
fact
of course, that
"Some
the relations of the concepts S and
This means, in intension, that
if
indicates nothing about that they are not incompatible. except one or both be further specified in proper
is
"
might well be symbolized by S + N is really suspect that Castillon s choice of S = A =F governed by the consideration that S = A + be supposed to give may
fashion, the results will coincide.
It
= A
M.
^F
We
S =
the universal to give its subaltern, and that A will also give S = A =F M, that is to say, the real particular
M,
= S which
M
is
"All
is
S"
will also give

"All
S =
A"
=F
M.
Thus
"All
"Some
is
A"
may
is
be
derived both from
is
and from
is
S",
which
a de
sideratum.
The
illusory negative particular
is,
correspondingly,
=F
M
symbolism.
Immediate inference works out
fairly well in this
The S =
is
universal affirmative and the real particular are converses.
A
-is
M gives A
= S
M, and
S+
=
vice versa.
The
universal negative
directly convertible.
S =
ticular
A+M
J/
gives
A =
-
J/,
and
vice versa.
The
S
illusory
par
also convertible.
S =
A T
gives
=F
M.
Hence
A =
=F
comes back to
S = A
M, which
M.
subaltern
A
S
universal gives
its
S =
=
A
-
gives
S =
A
=
=F J/,
-
and
=T=
A+M
gives S
A
M,
M.
And a real A = S 4
particular gives also the converse illusory particular, for

gives
S =
34
A
its
which gives which gives
subaltern,
S =
A ^ M,
figures of the syllogism
A =
M.
All the traditional
moods and
may
be symbolized
in this calculus, those
both for the real

All
All
which involve particular propositions being valid For example: particular and for the illusory particular.
Mis A
M
.
SisM
Sis
,4
.
All
= A+N S = M+P S = A + (N + P) = - A+N S = M+P

S
=
No
All
M
S
is is
M
:.
M
A
No Sis A
All
(N + P)
is
M
.
.
Some S is Some S is A
is
= A+N S = M =F P S = (A + N)
or
=F
S =
or
M-
P
+ N)
S = (A
- P
This
the most successful attempt at a calculus of logic in intension.
The
and
not
in
difficulty
about
"subtraction"
in the
Lambert
"
calculus, arises because

is
"
or
M which
If,
XIX Fragment of P does not mean

then
[
Leibniz,
"M
but
not
P
is
".
If it
mean
this,
and
are not
true inverses.
on the other hand,
M-
P indicates the
abstraction from
the concept
difficult or
of all that
involved in the concept P, then
-P
is
impossible to interpret, and, in addition, the idea of negation
cannot be represented by [-]. How does it happen, then, that Castillon s notation works out so well when he uses [-] both for abstraction and as
the sign of negation?
It
would seem that
his calculus
ought to involve
him
in
both kinds of
is
difficulties.
method
by good luck, hit upon a which nothing is ever added to or subtracted from a determined concept, S or A, except an indeterminate, or N or P, and this indeter
in
The answer
that Castillon has, apparently
M
]
minate, just because

are not true inverses.
it is
indeterminate, conceals the fact that [+
and [-]
And when
the sign
appears before a determinate,
A,
serve as the sign of negation, because no difficulty arises from supposing the whole of what is negated to be absent, or abstracted. Castillon s calculus is theoretically as unsound as Lambert s, or more
it
may
so
if unsoundness admits of It is quite possible that it was worked degree. out empirically and procedures which give invalid results avoided.
35
Whoever studies Leibniz, Lambert and Castillon cannot fail to be con vinced that a consistent calculus of concepts in intension is either immensely difficult or, as Couturat has said, Its main difficulty is not impossible.
the one which troubled Leibniz
s system the failure This can be avoided by treating negative propositions as affirmatives with negative predicates, as Leibniz did. The more serious difficulty is that a
[ ]
Lambert
and which constitutes the main defect in of + and [-] to behave like true inverses.
calculus of
"
"concepts
is
not a calculus of things in actu but only in possibile,
and
in a rather loose sense of the latter at that.
Holland pointed this out
admirably
in a letter to
Lambert. 55
He
gives the example according to
Lambert
method,
All triangles are figures.
All quadrangles are figures.
T =
tF
Q = qF
tQ
Whence,
T =56
or
qT =
and he then proceeds:

"In
the conclusion nA = be drawn, the calculus cannot determine whether the ideas nA and
general,
if
from
==
mC
and
B = nC
mB mB
consist of contradictory partial-ideas, as in the foregoing example, or not.
The thing must be judged according

This example also
tributed middle term.
of the concept
calls
to the
matter."
attention to the fact that Lambert
calculus,
by operations which he continually

If
"some
uses, leads to the fallacy of the undis

is
A"
simply some further specification
A, then
this
mode
is
not fallacious.
And
this observation
brings
down
the whole treatment of logic as a calculus of concepts in in
tension like a house of cards.
The
more
relations of existent things cannot be
determined from the relations of concepts alone.
The
calculus of Leibniz
is
successful than
s is
continental successors
unless Ploucquet
any invented by his an exception. That the long

successful
period between
him and De Morgan and Boole did not produce a

is
system of symbolic logic

sional point of view.
probably due to the predilection for this inten-
It is
successful after the initial interest

logical relations in extension,
no accident that the English were so quickly was aroused; they habitually think of
and when they speak
of
"intension"
it
is
usually clear that they

"intension"
55 56
do not mean those
relations of concepts
which the
of traditional logic signifies.

I,
Ibid., pp.
Deutscher Gelehrter Briefwechsel, 262-63.
Brief xxvu.
36
A
The beginning
of

this subject in
England is marked by the proposing some modification of the traditional logic by quantifying the predicate. As Sir William Hamilton 57 notes, the period from Locke to 1833 is singularly barren of any real con
thought upon
publication of
numerous
treatises, all
tributions to logic.
quantification of the predicate.

least as Leibniz.
About that time, Hamilton himself proposed the As we now know, this idea was as old at
Both Hamilton and
his student
Ploucquet, Lambert, Holland, and Castillon also had
quantified the predicate.

tion Ploucquet;
Thomson men
but this new burst of logical study in England impresses one as greatly concerned about its own innovations and sublimely indifferent
to its predecessors.
establish his
Hamilton quarrelled at length with De Morgan to 58 This is the more surprising, since priority in the matter.
George Bentham, in his Outline of a
New
System of Logic, published in 1827,

of propositions:
had quantified the predicate and given the following table

1
.
2. 3.
4.
5.
6.
7.
8.
is
|
X in toto = Y ex parte X in toto Y ex parte; X in toto = Y in toto; X in toto Y in toto; X ex parte = Y ex parte; X ex parte Y ex parte; X ex parte = Y in toto; X ex parte 7 in toto.
;
!
| | | | |
(|
here the sign of
"diversity").
in
But Hamilton was certainly the center and inspirer of a new movement logic, the tendency of which was toward more precise analysis of logical
significances.
Thought are
there
Bayne s Essay on the New Analytic and Thomson s Laws of the most considerable permanent record of the results, but
fervid discussion of logical topics in various peri
was a continual
was
odicals; logistic
in the air.
This movement produced nothing directly which belongs to the history of symbolic logic. Hamilton s rather cumbersome notation is not made the
basis of operations, but
is essentially only an abbreviation of language. scheme of representing syllogisms was superior as a calculus. But
Solly
57
See Discussions on Philosophy, pp. 119 ff. This controversy, begun in 1846, was continued for many years (see various articles in the London Athenceum, from 1860 to It was concluded in the pages of the Con 1867).
temporary Review, 1873.

this
37
in fact
it
movement accomplished two things for symbolic logic: though not always in name the point of view of
it
emphasized
extension,
and
aroused interest in the problem of a newer and more precise logic. These may seem small, but whoever studies the history of logic in this period
will easily
never have been revived.

Boole.
in
convince himself that without these things, symbolic logic might Without Hamilton, we might not have had
record of symbolic logic on the continent is a record of failure, England, a record of success. The continental students habitually
The
emphasized intension; the English, extension.

IV.
DE MOP CAN
of symbolic logic only
De Morgan 59
permanent value
of certain
tions.
is
known
to
most students
through
of
the theorem which bears his name.
But he made other contributions

of
60
discourse",
the idea of the
"universe
the discovery
new types
of propositions,
and a beginning
of the logic of rela
Also, his originality in the invention of

illustrations,"
new
logical forms, his
ready
wit, his pat
and the
clarity
and
liveliness of his writing did

\
yeoman
of
service in breaking
down
in
the prejudice against the introduction
His important writings on logic logic. are comprised in the Formal Logic, the Syllabus of a Proposed System of
"mathematical"
methods
Logic,
and a
series of articles in the

1
Transactions of the Cambridge Philo
sophical Society?
59
Augustus De Morgan (1806-78), A.B. (Cambridge, 1827), Professor of Mathematics London 1828-31, reappointed 1835; writer of numerous mathematical treatises which are characterized by exceptional accuracy, originality and clearness. Per haps the most valuable of these is "Foundations of Algebra" (Camb. Phil. Trans., vu, vm); the best known, the Budget of Paradoxes. For a list of his papers, see the Royal For many years an active member of the Cambridge Philosophical Society Catalogue. Society and the Royal Astronomical Society. Father of William F. De Morgan, the novelist and poet. For a brief biography, see Monthly Notices of the Royal Astronomical Society,
in the University of
xii, 112.
60
The
idea
is
introduced with these words:

It
is
"Let
us take a pair of contrary names,
as
man and
not-man.
or real, in the universe.

universe, but
plain that between them they represent everything, imaginable But the contraries of common language embrace, not the whole
idea. Thus, of men, Briton and alien are contraries: The same may be said of be one of the two, no man can be both. integer and fraction among numbers, peer and commoner among subjects of a realm, male and female among animals, and so on. In order to express this, let us say that the whole idea under consideration is the universe (meaning merely the whole of which we are considering parts) and let names which have nothing in common, _but which between them contain the whole of the idea under ^nsideratiqn, be called contraries in, or with respect to,
some one general
every
man must
that
universe."
61
after
(Formal Logic, p. 37; see also Camb. Phil. Trans., vm, 380.) Formal Logic: or, The Calculus of Inference, Necessary and Probable, 1847. to be cited as F. L.
Here
38
Although the work of De IN [organ is strictly contemporary with that of Boole, his methods and symbolism ally him rather more with his prede
cessors than with Boole
and those who
follow.
Like Hamilton, he
is
bent
upon improving the traditional Aristotelian logic. His first step in this direction is to enlarge the number of typical propositions by considering
all
the combinations and distributions of two terms,

It is a feature of
and F, and
their
negatives.
De Morgan
notation that the distribution of
affirmative or negative of the proposition are indicated, these being sufficient to determine completely the type of
each term, 62 and the quality

the proposition.
That a term
^before or after
A"
is
distributed
is
indicated
by writing
half a parenthesis
A"),
it,
with the horns turned toward the

is
7
)A".
letter, thus:
or (X.
An
undistributed term
A"(,
marked by turning the half-parenthesis the other

A"))}
,
way, thus: that
or
for example, indicates the proposition in
which the subject, X,

is,
"All
A"
is
distributed
and the predicate, F,

63
is
undistributed,
is
F".
X()Y
indicates a proposition with both terms un

is
F".
distributed, that
is,
"Some
The negative
of a term,
is
X,
is
indi
negative proposition placed between the parenthetical curves; thus "Some

.r;
cated by
of
F by
y, etc.
indicated
by
a dot
will
is
not
Y"
an affirmative proposition. All the different forms of proposition which De Morgan uses can be from two types, the universal, "All is generated ./ and the
A"(-(F.
be
64
Two
dots, or none, indicates
particular,
tive,
.r,
"Some
F and
y.
is by using the four terms, For the universals we have:

.
.,"
X and
its
nega
Syllabus of a Proposed System of Logic, 1860. Hereafter to be cited as Syll. Five papers (the first not numbered; various titles) in Camb. Phil. Trans., vm, ix, x. The articles contain the most valuable material, but they are ill-arranged and inter
spersed with inapposite discussion. Accordingly, the best get these articles and the Formal Logic, note in a general
way way
to study
De Morgan
is
to
the contents of each, and
then use the Syllabus as a point of departure for each item in which one is interested. 62 He does not speak of "distribution" but of terms which are "universally spoken
cr
"particularly
of"
spoken
of
",
or of the
"quantity"
of a term.
63 This is the notation of Syll. and of the articles, after the first, in Camb. Phil. Trans. For a table comparing the different symbolisms which he used, see Camb. Phil. Trans.,
ix, 91.
64
It is
sometimes hard to determine by the conventional
criteria
whether
De Morgan
propositions should be classed as affirmative or negative.

rule for distinguishing of and false of
them
X and
(Syll., p. 13):
"Let
gives the following ingenious a proposition be affirmative which is true
He
Thus Every A is Y is Some things are neither Some things are neither
things are neither
or x; negative, which is true of and x, false of and X. affirmative: is is true; is x is false. But Every Every
not-A"
A s nor F s A s nor A s
is
is is
also affirmative,
true,
A s nor z s
though in the form of a denial: though superfluous in expression; Some
false."

(1) (2) (3) (4)
39
A))
x))y
MIX
All
is
Y.
is
All not-
A"
not- 7.
X))y
is
not-7.
7.
x))Y
:
Allnot-Zis
and
for particulars
we have
(5)
X()
Some
X X
is
7.
is
(6) (7) (8)
x()y
Some not-X
not- 7.
X()y
x()Y
Some
Some
is
not- 7.
is
not-JT
7.
The
its
rule for transforming a proposition into other equivalent forms may be stated as follows: Change the distribution of either term that is, turn
parenthetic curve the other way, change that term into its negative, and change the quality of the proposition. That this rule is valid will appear if we remember that "two negatives make an affirmative", and note
that we introduce one negative by changing the term, another by changing the quality of the proposition. That the distribution of the altered term
should be changed follows from the fact that whatever proposition distrib
utes a term leaves the negative of that term undistributed,
and whatever
proposition leaves a term undistributed distributes the negative of that
term.
Using this rule of transformation, we get the following table of equivalents for our eight propositions:
(a)
(1) (2)
(3)
(b)
(c)
(d)
XY
x))y
= X)-(y
=
-
--=x((y
=
x(-)Y
x)-(Y
X((Y = X(-)y
*()?
-
X))y
X).(Y = x((Y
a-)-Q/
=
(4) .r))7
(5)
X((y
X(-)Y
T()
7 = X(-(y
*x(.(Y
-.=
-=x)(y
=
=x).)Y
(G) x()i/
(7) (8)
XQy
a-()7
Z)(7-Z)-)y = x)(Y = x)-)y X(.(Y

-
=x(.(y
X}(y -=X)-)Y
is
It will
be observed that in each line there
one proposition with both

eight different types
terms positive,
A and
Y.
Selecting these,
we have the
of propositions:
40
A
(la)
(2c)

is
A))F
All
A"
Y.
is all
A ((F
X)-(Y
X(-)Y
r
Some
F;
or, All
is
X.
(36)
NoZis
Some
Some
Some
All
Y.
is
(4d)
Everything
either
X or
Y.
(See below.)
(5o)
(6c)
A ()F
X)(Y
X(-(Y X)-)Y
A
A"
is
7.
things are neither

is
X nor
F.
(See below.)
(76)
not Y.
or,
(Sd)
is
not some Y;
is
Some 7
any one
is
not X.
Since the quantity of each term
indicated,
of these propositions
may
be read or written backwards
that
is
is,
w ith Y
r
subject
(4d)
and
X predicate X
is
provided the distribution of terms

cult to understand.
F",
preserved.
and
(6c) are diffi
We
might attempt to read X(-)Y
"Some
not
but we hardly get from that the difference between X(-)Y and some is not (any) A (-(F, "Some Also, A"(-)F is equivalent to uniis not some would make it par and the reading, "Some versals,
F".
F",
ticular.
"Xo
A"(-)F
is
equivalent to
cc))
F,
"All
not-Z
is
F",
and
to x)-(y,
not-X
is,
is
not-F".
The only equivalent

(in
of these
with the terms

is
A"
and F
"Everything
the universe of discourse)

likely to read
either
or
F
or
(or
"
both)".
(6c),
X)(Y, we should be
";
"All
is all
F",
X and
F are equivalent
but this would be an error, 65 since

(6a),
its
equivalents
is
not-F".
are
particular
propositions.
of this in
The equivalent
neither
terms of
xQy, X and
is
"Some
not-X
is
plainly,
"Some
things are
X nor
F".
Contradictories 66 of propositions in line

of those in line (2), in line (8);
(1) will
be found in line (7);
of line (3), in line (5); of line (4), in line (6).

:
We
give those with both terms positive

(la)
(2c)
X))Y
X((Y
A>(F
contradicts
"
(76)
X(-(Y
(Sd)
"
AF
X
",
(36)
(5a)
(6c)
XQY
X)(Y
X(-)Y
65
which it might seem that De Morgan himself has fallen. See e. g., is all F SylL, p. 25, and Camb. Phil. Trans., ix, 98, where he translates X)(Y by "All or "Any one X is any one Y But this belongs to another interpretation, the "cumular", which requires X and Y to be singular, and not-X and not- Y will then have common members. However, as we shall note later, there is a real difficulty.
error into
".
An
66
De Morgan
he
calls contradictory
propositions
"contraries"
(See F. L., p. 60; Sytt.,

"contraries".
p. 11), just as
calls
terms which are negatives of one another
41
Thus the
of terms
rule
is
that two propositions having the same terms contradict

is
one another when one

is
affirmative, the other negative,
and the distribution
exactly opposite in the

for transforming
two
cases.
The
rule
propositions which has been stated and
exemplified, together with the observation that
may
any symbolized proposition be read or written backwards, provided the distribution of the terms
from universals, particulars from particulars.
"Each
be preserved, gives us the principles for the immediate inference of universals
For the
rest,
we have
67
the rule,
universal affirms the particulars of the
same
quality".
For
syllogistic reasoning, the test of validity

6i
and
rule of inference are
as follows:
"There is inference: 1. When both the premises are universal; 2. When, one premise only being particular, the middle term has different quantities
in the
two premises.
conclusion
is
"The
[parenthetic
curves]."
found by erasing the middle term and its quantities This rule of inference is stated for the special
arrangement of the syllogism in which the minor premise is put first, and the minor term first in the premise, the major term being the last in the
second premise.
Since
any proposition may be written backward,

According to the rule,
"No
this
arrangement can always be made.

is
F",
X))Y,
"All
X
is
is
F",
and F) (Z, "No Y is and F(-(Z, "Some F

things are neither
A"
Z",
give X)-(Z,
X
-
is
Z".
A>(F,
"No
is
not
Z",
give
A")
(Z, or
A ) (Z, which
"Some
nor
Z."
The
reader may, by inventing other

is
examples, satisfy himself that the rule given

reasoning, with
sufficient for all syllogistic
any
of
De Morgan
eight forms of propositions.
De Morgan
also invents certain

in a fashion
pound syllogisms
"1.
compound propositions which give com somewhat analogous to the preceding: 69
X)0)Forboth A))Fand
AF
All
Xs
and some things be
sides are
2.
s.
For
both
A ))Fand A ((F
X((Y and X(-(Y
All
3.
X(0 (F
)
or both
X s are F and X Among X s are all the

s,
all
are
s.
and
some things
4.
besides.
A O (For
both X)-(Y and
X)(Y
Nothing both some things
and
F and
neither.
67
Sytt., p. 16.
**Sytt.,p. 19. 69 SyU., p. 22.
42
A
5.
Z|- For both

|
A>(Fand
X(-)Y
and F and Nothing both everything one or the other.

Everything either
6.
X(O)Y
or both
X(-)Y and
XQY
or
Y and
some things
both."
Each
sign
of these propositions
may, with due regard
for the
meaning
of the
O, be read or written backward,
just as the simple propositions.

is
The
rule of transformation into other equivalent forms
slightly different:
Change the quantity,

by
its
or distribution, of
any term and replace that term
negative.
We
are not required, as with the simple propositions, to
change at the same time the quality of the proposition. This difference is due to the manner in which the propositions are compounded.
The
rules for mediate, or

70
"
syllogistic",
inference for these
compound
or 6,
propositions are as follows:

"If
any two be joined, each of which is [of the form of] with the middle term of different quantities, these premises
clusion of the
1, 3, 4,
yield a con
same kind, obtained by erasing the symbols of the middle term and one of the symbols [O]. Thus X)O(Y(O)Z gives X)O)Z: or
if
nothing be both
X and
7 and some
things both,
things neither, and

it
if
everything be
tivo lots
either
other
"
Z and some things are Z s.

Y
or
in
follows that
all
and
of
In any one of these syllogisms,

or
it
follows that
may
be written for
one place, without any alteration of the conclusion, except )O( )O) reducing the two lots to one. But if this be done in both places, the con clusion is reduced to or and both lots disappear. Let the reader
|
examine
for himself the cases in
which one
of the premises
is
cut
down
to a
simple universal.
"The Letters written following exercises will exemplify what precedes. under one another are names of the same object. Here is a universe of 12
instances of which 3 are
and the remainder

the remainder
P s;
5 are
and the
remainder Q
s;
7 are
Z s and
R s.
XXX PP PP PPPPP
YYY YY QQ QQQQQ Z Z Z Z Z Z Z RRRRR
We
can thus verify the eight complex syllogisms
X)0)Y)0)Z P(0)Y)Q(R
70
P(0)Y)0)Z
P(0(Q(O)Z
X)0)Y)0(R
X)Q(Q(0(R
P(O(Q(O(R X)O(Q(O)Z
Sytt., p. 23.

In every case
it
45f
be seen that the two lots in the middle form the quantity of the particular proposition of the conclusion."
will
In so
too
much much tied
of his
work
as
we have thus
far reviewed,
to his starting point in Aristotelian
De Morgan is still He somewhat logic.
simplifies traditional
methods and makes new generalizations which include

distinctly the old logic.
old rules, but
it is still
He
does not question the
inference from universals to particulars nor observe the problems there involved. 71 He does not seek a method by which any number of terms
may
be dealt with but accepts the limitation to the traditional two. And his symbolism has several defects. The dot introduced between the
parenthetic curves
is
read ()
as,
"It
is
not the sign of negation, so as to false that The negative of ()

()".
make
is )
it
possible to
(,
so that this
simplest of all relations of propositions
is
represented by a complex trans

in the
formation applicable only when no more than two terms are involved
prepositional
relation.
Also, there are two distinct senses in which a
term
in a proposition
may
be distributed or
"mentioned universally",
and
De Morgan,
following the scholastic tradition,
fails
to distinguish them and
symbolizes both the same way.
This
is
is
the secret of the difficulty in reading
X)(Y, which looks

neither
A"
like
"All
A"
all
7",
and
really
is
"Some
things are
Mathematical symbols are introduced but without any mathematical operations. The sign of equality is used both corresponding for the symmetrical relation of equivalent propositions and for the un".
nor
72
73 symmetrical relation of premises to their conclusion. His investigation of the logic of relations, however,
is
more
successful,
and he
is
This topic suggested to him by consideration of the formal and material elements
laid the
field.
foundation for later researches in that
in logic.
71
He
says:
7l
But he does make the assumption upon which

is
particular from a universal
simple term has members. ideas, is tacitly claimed for the terms of every syllogism". 72 A universal affirmative distributes its subject in the sense that it indicates the class to which every member of the subject belongs, i. e., the class denoted by the predicate.
Similarly, the universal negative,
all inference (in extension) of a necessarily based: the assumption that a class denoted by a He says (F. L., pp. 110), "Existence as objects, or existence as
not-A
is F, indicates that every is not- Y, every Y is No The particular nega particular proposition distributes a term in that sense. tive tells us only that the predicate is excluded from some unspecified portion of the class and Y in this sense only. Comparison with denoted by the subject. A)(F distributes
.
No
its
equivalents shows us that
it
can
tell us,
of
A",
only that
it is
specified portion of not-F; and of Y, only that it is excluded from of not-A. is wholly included in Y, or cannot infer that
We
excluded from some un some unspecified portion Y in X, or get any other
relation of inclusion out of

73
it.
In one passage (Camb. Phil. Trans., x, 183) he suggests that the relation of two C. premises to their conclusion should be symbolized by A B
<
74
Camb.
Phil. Trans., x, 177, footnote.
44
there
A
"Is
any consequence without form?

Is
is
of the
machinery?
Is not consequence an action not logic the science of the action of the machinery?
act of the
Consequence
to ask,
on,
always an
mind
the
What
kind of act?
What
is
act,
on every consequence logic ought as distinguished from the acted

?
and from any
inessential concomitants of the action
For these are
of
the form, as distinguished from the matter.

".
The copula performs

. . .
certain functions;
it is
competent to those
functions
validate
because
. . .
it
has certain properties, which are sufficient to

is,
its use.
it
The word
which
identifies,
does not do
is
its
work
because
identifies,
except insofar as identification

is
a transitive and
;
convertible
motion:
is
that which is
B means A
is
,
is
and
A
is
is
B
fit
means
Hence every
transitive
and convertible relation
as
and by the same proof in each case. Some forms are valid when the relation is only transitive and not Thus if Y represent convertible; as in give and 7 connected
.
to validate the syllogism as the copula
by
a transitive copula, Camestres in the second figure
is
valid, as in
EveryZ
... In the
7,
No
X
:
Y,
therefore
No
is
X~Z.
following chain of propositions, there
exclusion of matter,
form being preserved at every step
Hypothesis
(Positively true)
Every man Every man Every

is
is is
animal
Y
a
/?
has existence.
X 7 Every X 7 a of X 7
(Probability
(3)
X has existence.
is
a transitive relation.
<
is is
a fraction
or or
1.
a of .Y
a fraction
<
1.
The
last is nearly the

it,
purely formal judgment, with not a single material
point about
"...
except the transit! veness of the copula. 75 I hold the supreme form of the syllogism of one middle term to
is
be as follows: There
there
is
the probability a that
is
in the relation
the probability
the probability a /3
"...
M to Z; that X in the relation L of M to Z.

that
is
to 7;
is
is
in the relation
whence there
76
The copula
of the
post hoc
is
and effect, of motive and action, of all which form and propter hoc (perhaps) of the matter, will one day
of cause
77
logic."
be carefully considered in a more complete system of

75 76
Ibid., pp. 177-78.

Ibid., p. 339.
77
Ibid., pp. 179-80.
45
De Morgan
general,
is
and to consideration
thus led to a study of the categories of exact thinking in of the types and properties of relations.
His division of categories into logico-mathematical, logico-physical, logico78 metaphysical, and logico-contraphysical,
is
inauspicious,
and nothing
"The
much comes
of
it.
But
it,
in
connection with
this,
and an attempt to rebuild

con
is
logic in the light of
he propounds the well-known theorem:

[logical
trary [negative] of an aggregate
sum]
the
compound
[logical
product] of the contraries of the aggregants: the contrary of a is the aggregate of the contraries of the components." 79
compound
For the
L, of
logic of relations,
X, Y, and
will
represent the class names;

is
M, N,
.
relations.
LY will
is
signify that
A"
some one
of the objects
thought which stand to Y
in the relation L, or is
one of the
. .
of 7. 80
X LY will signify that X not any one of the L s of Y. X (LM) Y or X LM Y will express the fact that AT is one of the L s of one of the M s of y, or that X has the relation L to some Z which has the relation M to Y. X LM Y will mean that X is not an L of any M of Y.
.
.
It should be
noted that the union of the two relations

"
L and
.
M
. .
is
.
what
.
we should
If
"
call
today
.
.
their
relative product
.
that
. .
is,
A"
together give A^
LM Z, but A
"brother
of"
LY and A
is
the relation
"
and
MY do not give
"aunt
is
the relation
Y".
will
mean
is
a brother of an aunt of
of
(Do not
MZ LM Y. A LM Y of say hastily, X
LY and Y
A"
",
"
is
uncle of
7".
"Brother
an
aunt"
is
not equivalent to
written by
"uncle"
since
some uncles have no

that
sisters.)
L, or
M,
itself, will
represent
which has the relation L, or
M,
that
is,
a brother, or an aunt, and

to F, that
is,
LY
stands for any
X which
has the relation
a brother of T. 81
In order to reduce ordinary syllogisms to the form in which the copula
has that abstractness which he seeks, that
is,
to the form in
it is
which the
necessary
copula
may
be any relation, or any relation of a certain type,
to introduce symbols of quantity.
L of Accordingly every M, that is, something which has the relation L to every member of L*M is to indicate an L of none the class (say, a lover of every man). but The mark of quantity, * or *, always s (a lover of none but men).
is
LM*
to signify an
M
78 79
See
ibid., p. 190.
Ibid., p. 208.
"On
See also
Sytt., p. 41.
3,
Pp. 39-60 of Syll. present in

in
General."
summary
the ideas
of the paper,
80
81
the Syllogism, No.
and on Logic
Camb.
Phil. Trans., x, 341.
We
1 tried at first to
make De Morgan
follow the order of the paper from this point on. s symbolism more readily intelligible by intro
ducing the current equivalents of his characters. But his systematic ambiguities, such as the use of the same letter for the relation and for that which has the relation, made this impossible. For typographical reasons, I use the asterisk where he has a small accent.
46
goes with the letter which precedes

the letter which follows.
that
it is
it,
but
L*M
L
of
is
read as
if
*]
modified
suggests
To
obviate this difficulty,
De Morgan
"An
L*M
an Z,
be read,
"
"An
every-!, of If;
an
M in every way in which

L
of
but we
shall stick to the simpler reading,
none but
s".
LM*X means an L of every M of X: L*MX, an L of none but M s of X:

L*M*, an L
of every
M and
1
of
none but 3f
s:
LMX*,
of
an
of
an
M of
.
every X, and so on.
If
Two more L is
".
symbols are needed.

of",
The converse
of";
L is
l symbolized by L~
"lover
L-
is "beloved
if
is
"aunt",
L-
is
"niece
or
nephew
ized
The contrary
(or as
we should
say, the negative) of
L is symbol
:
by m. by 1; the contrary of In terms of these relations, the following theorems can be stated
Contraries of converses are themselves contraries.
(1) (2) (3)
Converses of contraries are contraries.
The contrary
If
of the converse
is
the converse of the contrary.

(a)
(6)
(4)
the relation
L
l
,
be contained
is
converse of L,
L~
in, or imply, the relation M, then contained in the converse of M, M~ l and

\
the
the
contrary of
M, m,
is
contained in the contrary of L,

of"
I.
For example,
if "parent
is
contained in
"ancestor
of",
(a)
"child
of"
is contained in "descendent
"not
of",
and
(b)
"not
ancestor
of"
is
contained in
parent
of".
compound relation is accomplished by converting both components and inverting their order; thus, (LM) 1 = M~ l L~ l
(5)
.
The conversion
of a
If
X be teacher of the
child of Y,
is
When
a sign of quantity
parent of the pupil of X. involved in the conversion of a compound

is
relation, the sign of quantity changes
its
place on the letter; thus,
If
X be
teacher of every child of F,

in a
is
parent of none but pupils of X.

is
(6)
When,
compound
relation, there
its
a sign of quantity,
if
each
component be changed into

from one component to
the resulting relation
is
contrary, and the sign of quantity be shifted the other and its position on the letter changed,
equivalent to the original;
thus
LM* =
l*m and
L*M =
A
lm*.
lover of every
man
is
a non-lover of none but non-men; and a lover
of none but
men
is
a non-lover of everv non-man.

(7)
is
47
When
compound
relation involves
no sign of quantity,
its
contrary
found by taking the contrary of either component and giving quantity

"
to the other.
"Not
The contrary of LM is IM* or L*m. is "non-lover of every man" (lover of a man) but non-men"; and there are two equivalents, by
((>).
or
"lover
of
none
But
if
there be a sign of quantity in one component, the contrary
is
taken by dropping that sign and taking the contrary of the other component.
The contrary
"Not
of
LM*
"
is
IM;
of
L*M
is
Lm.
of a
man";
(lover of every
man)" is "non-lover
".
and
"not
(lover
of
none but men) is "lover of a non-man So far as they do not involve quantifications, these theorems are familiar though
seems not generally known that they are due to The following table contains all of them:
it
to us today,
De Morgan.
Combination
Converse
Contrary
l
Converse of Contrary Contrary of Converse
LM
M- L~
l
/J/* or
1
L*M
J/*-
/- 1
or
l
m~ Ll
LM*mhm
L*Morlm*
The
M*~ L~
l
or
or
M^L-
m~Hm*- l~
}
*
l
IM
M~
l
l~ l
l
Lm
is
m~ L~
"contained
in"
sense in which one relation
said to be
or to
"imply"
another should be noted:
is
contained in
which has the relation L to any Y has
also the relation
M in case every X M to that Y. This

between two
rela
is
is
must not be confused with the

tive terms.
relation of class inclusion

is
Every grandfather
also a father, the class of grandfathers

"grandfather
of"
contained in the class of fathers, but

"father
of",
not contained in
because the grandfather of

of"
is
not also the father of Y.

of",
The
uses
"All
relation "grandfather
is
contained in
"ancestor
since the grand
father of
is
also the ancestor of Y.

"
But De Morgan appropriately

M"
the same symbol for the relation
contained in
is,
that he uses for
is
M",
where
L and
M are class terms, that

of
L))M.
In terms of this relation of relations, the following theorems can be

stated
(8)
:
If
L))M, then the contrary
is
contained in the contrary of L,
that
is,
L))M
gives m))l.
Applying
(8
-)
this
theorem to compound
relations,
we have:
LM))N
If
gives w))/J/* and n))L*m.

l 1
(8")
LM))N, then L~ n))m and wJ/-
))/.
l
l
.
Proof: If
LM))N, then n))lM*.
Whence nM- ))lM*M~
But an
of
48

be an
l
every
M of an
M~
of
Z must
of Z.
l
Hence
nM~ ))l.
l
.
Again;
if
LM))N,
l
then ri))L*m.
relation converse-of-Z to
Whence L~ n))L- L*m. But whatever has the an L of none but m s must be itself an m. Hence
K"
L~ n))m. De Morgan
(9)
If
calls this "theorem
from
1
its
use in Baroko and Bokardo.
LM
= N,
then
Proof: If
for
any X,
LM MM~ X
1 1
L))NM~ and M))L~ N. = N, then LMM~ = NM~ and

l l
l
L^LM = Z~W. Now

X; hence
the
and L~
LX
l
are classes which contain
theorem.
We
do not have
L =
==
NM~
and
l
M
--=
==
L~ N, because
1
:
it is
not generally
of the of the
MM~ X X and Lr LX X. of X may not be X but brother parent

true that
A"
For example, the child but the class children

"
parent of
then
X"
will contain
l
X.
1
The
relation
MM~
or
cancel out.
MM~
1 1
and
M~ M are always symmetrical relations

1
M~ M will not always

1
;
if
XMM~
YMM~ X. If X is child of a parent of Y, then Y is child of a parent of X. But MM- and M~ M are nctf exclusively reflexive. XMM^X does
If we know that a child of the parent of X is a celebrated we may not hastily assume that X is the linguist in question. linguist With reference to transitive relations, we may quote 82
not always hold.
"A
relation
is
transitive
when a
relative of a relative
is
a relative of
so on.
the same kind; as symbolized in ZZ))Z, whence

"A
ZZZ))ZZ))Z; and
transitive relation has a transitive converse, but not necessarily a transitive contrary: for L~ l L~ l is the converse of LL, so that ZZ))Z gives
L L
1
its
From these, by contraposition, and also by theorem contrapositions, we obtain the following results:
*))L
l
.
and
L
I
is
contained in
LL-
*,
Z*H, HZ*, L*~
ZZ"
1
,
Z-!Z ...... Z- 1
omit demonstration, but to prevent any doubt about correctness of
printing, I subjoin instances in words:

82
Z signifies ancestor and L~
descendent.
Camb.
Phil. Trans., x, 346.
For
treats all reciprocal relations, such as
XLL^Y,
this discussion of transitive relations, De Morgan as also reflexive, though not necessarily
exclusively reflexive.

"An
49
ancestor
is
always an ancestor of
all
descendents, a non-ancestor
none but non-descendents, a non-descendent of all non-ancestors, and a descendent of none but ancestors. A descendent is always an ancestor of
of
none but descendents, a non-ancestor of

ancestor
all non-descendents, a non-descend ent of none but non-ancestors, and a descendent of all ancestors. A nonis always a non-ancestor of all ancestors, and an ancestor of none but non-ancestors. A non-descendent is a descendent of none but non-
descendents,
and a non-descendent
all
of
all
descendents.
Among
all
nonnonall
ancestors are contained

ancestors
of
descendents of non-ancestors, and
descendents.
Among
all
non-descendents
are
contained
ancestors of non-descendents, and
non-descendents of
ancestors."
In terms of the general relation, L, or

syllogisms of traditional logic
M,
representing any relation, the

83
may
be tabulated as follows:
The Roman numerals here

clusion, there are
indicate the traditional figures.
All the con
clusions are given in the affirmative form;
but for each affirmative con

relation
.
two negative conclusions, got by negating the

its
1
and
l
replacing
.
it
by one or the other of

.
contraries.
A IH*Z and A L*mZ;

and so on
63
1M~ Z
gives
LMZ gives Thus LM~ *Z and A hm- Z,

A"
for
each of the others.
Ibid., p. 350.
50
A
When
the copula of
all

three propositions
is
limited to the same transitive

84
relation, L, or its converse, the table of syllogisms will be:
X..LY
I
X.LY
Y..L-1Z
X..LY
Y.L~ Z X L~ Z
1
Y..LZ
X ..LZ
X .LZ
Y ..LX
IV
Y .LX
Y ..LX
1
Z.L- Y
1
X.LZ
Here, again, in the logic of relations,
Z..L~ Y X.L~ Z
1
Z..LY X..L~
1
Z"
De Morgan would
very likely have

itself.
done better
if
he had
left
the traditional syllogism to shift for
The
introduction of quantifications and the systematic ambiguity of L, M, etc., which are used to indicate both the relation and that which has the
hurry him into complications before the simple analysis of rela tions, and types of relations, is ready for them. This logic of relations was
relation,
destined to find
its
importance
in the logistic of
of,
mathematics, not
in
any
applications to, or modifications

cations of
Aristotelian logic.
And
these compli
De Morgan
s,
due largely to
his following the clues of formal logic,
had to be discarded
Boole
s
later, after Peirce
discovered the connection between

logic of relative
algebra and
relation theory.
The
terms has been
reintroduced by the work of Frege and Peano, and more especially of Whitehead and Russell, in the logistic development of mathematics. But
it is
there separated
and has to be separated

Nevertheless,
it
of the relations themselves.
from the simpler analysis should always be remembered
was De Morgan who laid the foundation; and if some part of his work had to be discarded, still his contribution was indispensable and of
that
it
permanent value.
84
In concluding his paper on relations, he justly remarks
85
:
Ibid., p. 354. Ibid., p. 358.
85

"And
51
here the general idea of relation emerges, and for the first time in the history of knowledge, the notions of relation and relation of relation
are symbolized.
And
here again
is
is
seen the scale of graduation of forms,

is
the
difference of form at one step of the ascent difference of matter at the next. But the relation of algebra to the
manner
in
which what
higher
developments of logic
It will hereafter
it
is
a subject of far too great extent to be treated here.
be acknowledged that, though the geometer did not think

et
necessary to throw his ever-recurring principium
tion of
Omnis homo
est
animal, Sortes
est
homo,
etc.,
exemplum into imita yet the algebraist was
living in the higher

relation, before it
atmosphere of syllogism, the unceasing composition of was admitted that such an atmosphere existed." M
V.
BOOLE
The beginning from which symbolic logic has had a continuous develop 87 is that made by His significant and vital contribution George Boole. was the introduction, in a fashion more general and systematic than before,
ment
of
Indeed Boole allows operations which have no direct logical interpretation, and is obviously more at home in mathe matics than in logic. It is probably the great advantage of Boole s work
that he either neglected or was ignorant of those refinements of logical theory which hampered his predecessors. The precise mathematical
mathematical operations.
development of logic needed to make its own conventions and interpreta tions; and this could not be done without sweeping aside the accumulated
traditions of the non-symbolic Aristotelian logic. As we shall see, all the nice problems of intension and extension, of the existential import of universals
and
particulars, of
It
is
empty
classes,
and
so on, return later
and demand
consideration.
enough
well that, with Boole, they are given a vacation long to get the subject started in terms of a simple and general procedure.
s first
book, The Mathematical Analysis of Logic, being an Essay toward a Calculm of Deductive Reasoning, was published in 1847, on the
86
Boole
omit, with
some misgivings, any account
of
De Morgan
contributions to prob
and judgment. (See SylL, pp. 67-72; 384-87, and 393-405.) His work on this topic is less closely connected with symbolic logic than was Boole s. The allied subject of the "numerically definite syllogism" (see Syll., pp. 27-30; F. L., Chap, vm; and Camb.
Phil. Trans., x, *355-*358)
87
ability theory as applied to questions of authority F. L., Chap, ix, x; and Camb. Phil. Trans., vm,
is
also omitted.
George Boole (1815-1864) appointed Professor of Mathematics in Queen s College, Cork, 1849; LL.D. (Dublin, 1852), F.R.S. (1857), D.C.L. (Oxford, 1859). For a biographi cal sketch, by Harley, see Brit. Quart. See also Proc. Roy. Rev., XLIV (1866), 141-81.
Soc.,
xv
(1867), vi-xi.
52
o
~*same day as
Calculus of
article
De Morgan s Formal Logic. 88

and
The next
year, his article,
"The
Logic,"
appeared in the Cambridge Mathematical Journal.
This
summarizes very
posed by Boole.
in
briefly clearly the important innovations pro But the authoritative statement of his system is found
An Investigation of the Laws of Thought, on which are founded the matical Theories of Logic and Probability, published in 1854. 89
Boole
s
Mathe
algebra, unlike the systems of his predecessors,
is
based squarely
upon the relations of extension.

his
The
three fundamental ideas

"
upon which
(1) symbols"; (2) the Jaws of thought expressed as rules for operations upon these symbols; (3) the observation that these rules of operation are the same which would hold for an algebra of the numbers and I. 90
method depends
are:
the conception of
elective
For reasons which

objects" is
will
appear shortly, the

1.
"
universe of conceivable
represented by
All other classes or aggregates are supposed
to be formed from this

in
1, all
by
selection or limitation.
l-.r
This operation of
electing,
all
the
A"s,
is
represented by
or x;
the operation of electing
the
s is
similarly represented
by
1-y or y,
and
so on.
Since Boole does
not distinguish between this operation of election represented by x, and the result of performing that operation an ambiguity common in mathe matics
.r
becomes, in practice, the symbol for the class of

representing ambiguously
all
the A^
s.
Thus
x, y, z, etc.,
operations of election or classes,

"elective symbols"
are the variables of the algebra.

to distinguish
Boole speaks of them as
them from
coefficients.
This operation of election suggests arithmetical multiplication: the suggestion becomes stronger when we note that it is not confined to 1.
1
"
-x -y or
xy
will represent the
universe",
operation of electing, first, and from this class by a second operation,
all all
the
in the
the Fs.
The
result of these
A"s
and F
s.
two operations Thus xy is the
will
be the class whose members are both
class of the
common members
of x
and
to
y;
z,
xyz, the class of those things which belong at once to and so on. And for any x, l-x = x.
x, to y,
and
The operation of "aggregating parts into a whole" is represented by + + y symbolizes the class formed by x combining the two distinct classes,
.
x and
y.
It is a distinctive feature of
Boole
algebra that x and y in x + y
must have no common members.

88
The
"On
relation
may
be read,
"that
which
Camb.
See
De Morgan s note to
the article
Propositions Numerically
Definite",
Phil Trans., xi (1871), 396. 89 London, Walton and Maberly.

90
Work
This principle appears for the hereafter cited as L. of T.
first
time in the Laws of Thought.
See pp. 37-3g.

is
53
either x or y but not both". Although Boole does not remark it, x + y cannot be as completely analogous to the corresponding operation of ordinary algebra as xy is to the ordinary algebraic product. In numerical
number may be added to itself: but since Boole conceives the terms of any logical sum to be quite distinct 91 mutually exclusive classes, x + x cannot have a meaning in his system. As we shall see, this is very awkward, because such expressions still occur in his algebra and have to be
algebras a
"
",
dealt with by troublesome devices.
But making the relation x + y completely disjunctive has one advantage it makes possible the inverse relation of "subtraction". The "separa
tion of a part, x, from a whole,
?/",
is
represented by y
x.
If
x+
then since x and
z
]
have nothing
in
common, y
and y
= =
y,
x.
Hence
x +
of x or
and
are strict inverses.
y,
then, symbolizes the class of those things which are either

of y,
members
members
but not of both,
x-y or xy symbolizes the class of
those things which are both

sents the class of the
members of x and members of y. x y repre members of x which are not members of y the x s
two
classes
except the y
s.
i.
=
]
represents the relation of
which have the
same members,
e.,
have the same extension.
These are the fundamental
relations of the algebra.
The
entity
(1
.r)
.r
is
of especial
all
universe except the
s,
or
importance. This represents the It is, then, the things which are not x s.
supplement or negative of x.
With the use

classes, x and
y,
of this
symbolism for the negative of a class, the sum of two which have members in common, can be represented by
xy + x(l
-#)
(!
x)y.
s
The
first
term of
this
s
sum
is
the class which are both x

s;
and y
s;
the second,
those which are x
but not y
the third, those which are y
but not x
s.
Thus the three terms represent classes which are all mutually exclusive, and the sum satisfies the meaning of + In a similar fashion, x + y may
.
be expanded to
x(l
y)
(l
x)y.
Consideration of the laws of thought and of the meaning of these bols will show us that the following principles hold
:
sym
x.
(1)
xy
yx
(2)
91
x +y
y+x
What What
is is
both x and y is both y and either x or y is either y or .r.
See L. of
T., pp. 32-33.
54
A
(3) z(x
+ y)
zx + zy
That which
is
is
both
z
and and
(either
z
.r
or y)
y.
either both
is
and x or both
z
(a;
and
(4) z(.r
y)
zx
zy
That which
is
both
but not
2
y)
y.
both
and x but not both
and
(5)
If
.T
y,
then
= z + x = x-z =
Z.T
.r
zy
z
+y
(6)
.r
= -
y +
This
last is
an arbitrary convention: the
first
half of the expression gives
the meaning of the last half.

It is a peculiarity of
"logical
symbols"
that
if
the operation
.r,
upon
1,
be repeated, the result

l-x
is
not altered by the repetition:

.
.
=
(7)
l-x-x
2
=
X
x x x
Hence we have
.T
Boole
calls this
the
"index law".
92
All these
laws, except
logic,
"If
(7),
hold for numerical algebra.
It
may
At
be
noted that, in
glance, this
.r
y,
then zx
=
zy"
is
not reversible.
first
may
seem to be another difference between numerical algebra
and the system
in question.
But
"If
zx
zy,
then x
=
y"
does not hold
Law (7) is, then, the distinguishing in numerical algebra when z = 0. The only finite numbers for which it holds are principle of this algebra. and 1. With 1. All the above laws hold for an algebra of the numbers and
this observation, Boole
2 modified by the law x
adopts the entire procedure of ordinary algebra, x, introduces numerical coefficients other than
and
1,
and makes
use,
properties of functions,
to serve his purpose. 93
and
on occasion, of the operation of division, of the of any algebraic transformations which happen
This borrowing of algebraic operations which often have no logical interpretation is at first confusing to the student of logic; and commen
tators have seemed to smile indulgently
upon
it.
An example
it,
will help:
the derivation of the

of
duality",
"law
of contradiction" or, as Boole calls

94
the
"law
from the
"index law".
92 In Mathematical Analysis of Logic he gives he avoids this, probably because the factors of x n
it
also in the
form x n
x) are
(e. g.,
z3
= x, but in L. of T. not always logically
iriterpretable.
93 and 1, and the fractions which can This procedure characterizes L. of T. Only be formed from them appear in Math. An. of Logic, and the use of division and of fractional coefficients is not successfully explained in that book.
94
L. of T., p. 49.

Since x 2
55
.r,
a-
0.
Hence, factoring,
.r(l
x)
0.
Whoever says
is
This transformation hardly represents any process of logical deduction. 2 "What is both x and a-, a* is equivalent to x; therefore what
,
both x and not-z
is nothing"
may well
if
.r 2
be asked for the steps of his reason
ing.
Nor should we be
satisfied
.r
he reply by interpreting in logical terms
the intermediate expression,
0.
Nevertheless, this apparently arbitrary

algebraic processes
is
way
of using uninterpretable
thoroughly sound. Boole s algebra may be viewed as an abstract mathematical system, generated by the laws we have noted, which has two interpretations. On the one hand, the "logical" or "elec
tive"
cal
symbols may be interpreted as variables whose value is either numeri or numerical 1, although numerical coefficients other than and 1 are
be remembered that such coefficients do not obey which holds for "elective" symbols. All the usual alge
it
admissible, provided
the
"index law"
braic transformations will have an interpretation in these terms.
On
the
other hand, the
symbols may be interpreted as classes. For this interpretation, some of the algebraical processes logical of the system and some resultant expressions will not be expressible in terms
"logical"
or
"elective"
of logic.
But whenever they are interpretable, they will be valid conse quences of the premises, and even when they are not interpretable, any
further results, derived from them, which are interpretable, will also be
valid consequences of the premises.
It
must be admitted that Boole himself does not observe the
proprieties
His consistent course would have been to develop this al without reference to logical meanings, and then to discuss in a thorough gebra
of his procedure.
fashion the interpretation, and the limits of that interpretation, for logical
such a method, he would have avoided, for example, the + .r. We should have + x = 2.r, the interpretation of difficulty about
classes.
By
.1-
.1-
which
classes
for the
numbers
and
1 is
obvious, and
its
interpretation for logical
which
will
would depend upon certain conventions which Boole made and be explained shortly. The point is that the two interpretations
should be kept separate, although the processes of the system need not be limited by the narrower interpretation that for logical classes. Instead
of
making this separation of the abstract algebra and its two interpretations, Boole takes himself to be developing a calculus of logic; he observes that
"axioms"
95
its
are identical with those of an algebra of the
numbers
and
95
1;
L. of T., pp. 37-38.
56
hence he applies the whole machinery of that algebra, yet arbitrarily rejects from it any expressions which are not finally interpretable in terms of logical The retaining of non-interpretable expressions which can be relations.
transformed into interpretable expressions he compares to
ment
of the uninterpretable
"
of trigonometry.
96
It
"the employ symbol V 1 in the intermediate processes would be a pretty piece of research to take Boole s
algebra, find independent postulates for
it (his
laws are entirely insufficient
as a basis for the operations he uses), complete
it,
and systematically
investi
gate
its
interpretations.
of
But neglecting these problems

logical relations in
will etc.
method, the expression of the simple

will
Boole
symbolism
will
now be
entirely clear.
Classes
y),
be represented by
x, y, z, etc.;
their negatives,
by
(1
is
x), (1
That which
.r(l
is
both x and y
be
;r(l
i.
y), etc.
That which
is
x but not y will x or y but not both, will be x + y, or x or y or both will be x +

or
y)
(1
be xy\ that which
y) + (1
x}y.
is
That which
is
x)y
e.,
that which
x or not x but y
xy + x(l
that which
"universe"
(1
x)y
x.
1
is
both x and y or x but not y or y but not

"everything".
represents the
is
or
The
logical significance of
determined
Oy = 0: the only class which remains unaltered of electing from it whatever is the class "nothing". by any operation Since Boole s algebra is the basis of the classic algebra of which logic
by the
fact that, for
any
y,
the topic of the next chapter it will be unnecessary to comment upon those parts of Boole s procedure which were taken over into the classic
is
algebra.
These
will
be clear to any
who understand
the algebra of logic
in its current
II.
form or who acquaint themselves with the content of Chapter
turn our attention chiefly to those parts of his method which are peculiar to him.
shall, then,
We
Boole does not symbolize the relation
"x
is
included in
?/".
Conse
quently the only copula by which the relation of terms in a proposition can be represented is the relation =. And since all relations are taken in
extension, x
membership.
y symbolizes the fact that x and y are classes with identical Propositions must be represented by equations in which
or == 1, or else the predicate must be quantified. something is put = Boole uses both methods, but mainly relies upon quantification of the This involves an awkward procedure, though one which still predicate.
survives
96
the introduction of a symbol v or w, to represent an indefinite
L. of T., p. 69.

class
57
and symbolize
vy:
"Some
class"
"Some".
Thus
#",
"All
is
(some)
If
v,
i/"
is
represented by
"the
is
(some)
by wx =
this
vy.
or w. were here
indefinite
or
v,
"any
class",
method would be
that
less objectionable.
it
But
class
in
such cases
or w,
must be very
but
this,
definitely specified:
it
97
must be a
of
"indefinite
in all respects
contains some
members
the class to whose expression
it is prefixed".
The
universal affirmative
can also be expressed, without this symbol for the indeterminate, as .r(l y) = 0; "All xisy" means "That which is x but not y is nothing". Negative
cate.
propositions are treated as affirmative propositions with a negative predi So the four typical propositions of traditional logic are expressed as
98
follows:
All x
is is
y:
y:
is
x
.r
= =
vy,
v(l
or,
x(l
y)
0. 0.
Xo
y), y),
y),
or
or, or,
Some x
y:
v.r
= =
w(\
xy = v =
v
xy.
Some
Each
.r
is
not y:
vx
w(l
x(l
y).
of these has various other equivalents
which
may
be readily deter
mined by the laws
of the algebra.
To
reason
by the
aid of this symbolism, one has only to express his
proper manner and then operate upon the resultant equation according to the laws of the algebra. Or, as Boole more explicitly
premises explicitly in the
puts
"
it,
valid reasoning requires:
"
1st,
That a
fixed interpretation be assigned to the
in the expression of the data;
and that the laws
of the
symbols employed combination of these
symbols be correctly determined from that interpretation. "2nd, That the formal processes of solution or demonstration be con
ducted throughout in obedience to all the laws determined as above, with out regard to the question of the interpretation of the particular results
obtained.
"3rd,
That the
final result
be interpretable in form, and that
it
be
actually interpreted in accordance with that system of interpretation which
has been employed in the expression of the
data."
As we
97
shall
see,
Boole
methods
of solution
is
sometimes involve an
uninterpretable stage, sometimes not, but there

L. of T., p. 63.
is is
provided a machinery by
the above statement
This translation of the arbitrary v by "Some" is unwarranted, and inconsistent with Boole s later treatment of the arbitrary coefficient.
null.
There
98 99
no reason why such an arbitrary coefficient may not be See Math. An. of Logic, pp. 21-22; L. of T., Chap. iv.
L. of T., p. 68.
58
A
may

be reduced to a form which
is
which any equation

this
interpretable.
To
we must first understand the process known as the develop comprehend ment of a function. With regard to this, we can be brief, because Boole s method of development belongs also to the classic algebra and is essentially
the process explained in the next chapter. 100
Any
the form
expression in the algebra which involves x or
(1
x)
may
it
be
called a function of x.
A
.r).
function of x
is
said to be developed
when
has
Ax + B(l
It is here required that
x be a
"logical symbol",
susceptible only of the values
and
be
1.
But the
"logical
coefficients,
symbol"
not so limited: A, or B,
"law
may
such a
and B, are which obeys the

or
1,
of
duality",
or
If
it
may
be some number other than
or involve
such a number.
the function, as given, does not have the form
Ax
+ B(1
.r), it
may
be put into that form by observing certain interesting
laws which govern coefficients.
Let
/(.r)
Then
/(I) /(O)
And
Hence
= Ax + B(l - x) = A-1 + B(1 ~1) = A - A-Q + B(l - 0) = B =

/(I) -x +/(0)
(1
f(x)
x)
Thus
if
f(x)
~,
2x +
;
Hence
f(x)
(1
x)
developed function of two variables, x and y
will
have the form:

x)(l
y)
And
for
any function,
/(.r, y),
the coefficients are determined by the law:
f(*,y)
/(I, !).?# +/(!,()) x(l
-y)
+/((),
!)(! -x)y
See Math. An. of Logic, pp. 60-69; L. of T., pp. 71-79; "The Calculus of Logic," Cambridge and Dublin Math. Jour., m, 188-89. That this same method of development should belong both to Boole s algebra and to the remodeled algebra of logic, in which + is not completely But a completely developed function, disjunctive, is at first surprising. either algebra, is always a sum of terms any two of which have nothing in common. This accounts for the identity of form where there is a real and important difference in the meaning of the symbols.
[
59
Thus
if
f( x y)
,
= = =
ax + 2by,
a-1 + 26-1
a1
/(I, 1) /(I, 0)
/((), 1)
=
=
a + 26
+ 26-0
26-1
a
26
/((),
0)
= a-0 + = tt-0 +
= =
26-0
Hence
/(.r,
y)
a + 21) xy + ay (I
y)
+ 26(1
.r)y
An
exactly similar law governs the expansion and the determination of
coefficients, for functions of
any number
of variables.
In the words of
Boole:
101
"The
general rule of development will

will relate to the
consist of
two
parts, the
formation of the constituents of the expansion, the second to the determination of their respective coefficients. It is as
follows
:
first of
which
"1st.
To expand any function of
the
symbols x,
y, z
Form
a series
of constituents in the following
manner: Let the
first
constituent be the
product of the symbols: change in this product any symbol z into 1 z, for the second constituent. Then in both these change any other symbol Then in the four constituents y into 1 y, for two more constituents.
thus obtained change any other symbol .r into 1 .r, for four new constit uents, and so on until the number of possible changes has been exhausted.
"2ndly.
To find
the coefficient of
any
constituent
If
.r
that constituent
into
1
;
involves x as a factor, change in the original function involves

1
but
if it
x as a factor, change in the original function .r into 0. Apply the same rule with reference to the symbols y, z, etc.: the final calculated
value of the function thus transformed will be the coefficient
sought."
Two
solutions
is 0.
further properties of developed functions, which are useful in
and interpretations,
are:
(1)
The product
of
any two constituents
If one constituent be, for example, xyz, any other constituent will have as a factor one or more of the negatives, 1 1 z. .r, y, 1 Thus the product of the two will have a factor of the form x(l And .r).
where
is
.r
is
a
0.
"logical
symbol
",
susceptible only of the values

of
1.
and
1,
x(l
x)
always
ficient 1, the
And (2) if each constituent sum of all the constituents is

which
it
any expansion have the coef
All information
may
be desired to obtain from a given set of
premises, represented by equations, will be got either (1)
by a
solution, to
.r,
determine the equivalent, in other terms, of some

101
"logical
symbol"
or
L.
o/T
.,
pp. 75-76.
60
A
by an
(2)
elimination, to discover
what statements
(equations), which are

in
independent of volve
x, or (3)
some term
x, are warranted
by given equations which
by a combination
/,
of these two, to determine the equivalent
of x in terms of
other
result.
from equations which involve x, t, u, v, and some logical" symbol or symbols which must be eliminated in the desired Formal reasoning is accomplished by the elimination of "middle"
u,
v,
" "
terms.
The student
of equations
of symbolic logic in its current
form knows that any
set
may
involving a term x
be combined into a single equation, that any equation = 0, and that may be given the form Ax + B(l x)
the result of eliminating x from such an equation is solution of any such equation, provided the condition
will
AB = AB =
0.
Also, the
be satisfied,
s
be x
= B
1,
+ v(l
A), where
is
undetermined.
Boole
methods
achieve these same results, but the presence of numerical coefficients other
than
and
as well as the inverse operations of subtraction
and
division,
necessarily complicates his procedure.

of solutions in the
And
he does not present the matter,
in which we should expect to find it but in a more complicated fashion which nevertheless gives equivalent results. We have
form
now
Boole obviates the
by which which have been mentioned. algebra The simplest form of equation is that in which a developed function,
difficulties of his
to trace the procedures of interpretation, reduction, etc.
of
any number
of variables,
is
equated to
0, as:
Ax + B(l -x) =
Axy +
It
is
0,
or,
.r)(l
B.r(l
y)
+ C(l
x)y + D(l
y)
0,
etc.
an important property of such equations that, since the product of in a developed function is 0, any such equation gives any one of its constituents, whose coefficient does not vanish in the develop ment, = 0. For example, if we multiply the second of the equations given
any two constituents
by
we
xy, all constituents after the first will vanish, giving shall have xy = 0.
Axy =
0.
Whence
Any equation in which a developed function is equated to reduced to the form in which one member is by the law;
1
may be If V = 1,
1
"
--
V =
0.
The more
symbol",
general form of equation
is
that in which some
logical
w,
is
suppose x
x =
equated to some function of such symbols. For example, yz, and it be desired to interpret z as a function of x and y.
yz gives z
x/y-
but this form
is
not in terpre table.
We
shall, then,

develop x/y by the law
).(!
61
x)y
+/(0,0)-(1 -x)(l
-y)
By
this law:
If z
3*
-, then
= x+-x(l
These fractional
7)
+ 0(1
coefficients represent the sole necessary difference of Boole s
methods from those at present familiar

of division
in
a difference due to the presence
system. Because any function can always be de and the difference between any two developed functions, of the veloped,
his
same
variables,
is
always confined
to
the
coefficients.
If,
then,
we can
interpret
and successfully deal with such

Boole
s
fractional coefficients, one of the
difficulties of
system
is
removed.
The fraction 0/0 is indeterminate, and this suggests that a proper inter pretation of the coefficient 0/0 would be to regard it as indicating an unde termined portion of the class whose coefficient it is. This interpretation may be corroborated by considering the symbolic interpretation of "All
x
is
?/",
which
y)
is
x(l
0,
y)
0.
If
x(l
then x
xy
and x =
xy.
Whence y
x/x.
Developing x/x, we have y

If
"All
x + -
(1
x).
is
?/",
the class y
is
portion of the class
not-.r.
made up Whence
of the class x plus
an undetermined
0/0
is
equivalent to an arbitrary
parameter
or as
"All,
v,
which should be interpreted as some, or none

of".
"an
undetermined portion
of"
The
coefficient 1/0 belongs to the general class of

"index
law",
symbols which do not

"law
obey the
.r(l
x~ =
.r,
or
its
equivalent, the
of
duality",
x)
0.
At
least Boole says it belongs to this class,

fact,
numerical properties of 1/0 would, in

belong to Boole
s
though the depend upon laws which do not

with the class of
system.
But
in
any
case, 1/0 belongs
such coefficients so far as
its logical
interpretation goes.
Any constituent of a
developed function which does not satisfy the index law must be separately
equated
to 0.
Suppose that
in
any equation
w = At + P
62
A
be a
"
logical
symbol",
whose
be the
law.
coefficient
be a constituent of a developed function does not satisfy the index law, A 2 = A. And let P
t
and
sum Then
of the remaining constituents
whose
coefficients
do
satisfy this
w =
2
w,
t*
t,
and
P = P
2
Since the product of any two constituents of a development
is 0,
Hence
w = (At+P) = A w = A t+P
2 2 2
2 2
t
Subtracting this from the original equation,
(A
Hence
since
- A )t = = A(l = A(l - A) 4= 0,
2
A)t
Hence any equation
of the
form
w = P+OQ+
is
IR+IS
S =
equivalent to the two equations
w = P + vR
which together represent
It will
its
and
complete solution.
be noted that a fraction, in Boole s algebra, is always an am biguous function. Hence the division operation must never be performed: the value of a fraction is to be determined by the law of development only, except for the numerical coefficients, which are elsewhere discussed.
We
have already remarked that ax = bx does not give a = b, because .r may have the value 0. But we may transform ax = bx into a = bx/x and determine this fraction by the law
f(b,x)
/(!, l).fcr+/(l,0).&(l
-*)+/<)
!!
- bx
We
shall
then have
a= -==
and
this
is
bx
bx *
(1
~^
6.
not, in general, equivalent to
Replacing 0/0 by indeterminate
coefficients, v
If
and w,
then
a
this gives us,
ax
bx,
x)
63
And
this result
is
always
valid.
Suppose, for example, the logical equation
rational
men = married men

Our equation
and suppose we wish to discover who are the rational beings.

will
not give us
rational
married
but instead
married men + v married non-men + w non-married non-men our hypothesis is satisfied if the class "rational is, beings" consist of the married men together with any portion (which may be null) of the
rational
-
That
class
"married women"
and a similarly undetermined portion of the

s
class
"unmarried women".
If
we
consider Boole
system as an algebra of
and
1,
and the
fact that
for
any
fraction, x/y,
--a*+ 6 *(l-)+
we
shall find,
(1)
(l
-*)(!-)
-
by investigating the cases

and y =
1;
(2)
=
x
1,
y
0,
0;
(3)
0,
0,
1;
and
that
it
(4)
requires these three possible cases:

.
Or, to speak
susceptible of the values
more accurately, it requires that 0/0 be an ambiguous function and 1.
Since there are, in general, only four possible coefficients, 1, 0, 0/0, and such as do not obey the index law, of which 1/0 is a special case, this means
that any equation can be interpreted, and the difficulty due to the presence of an uninterpretable division operation in the has disappeared. system
And any equation can
be solved for any
"logical
symbol"
.r,
by trans
ferring all other terms to the opposite side of the equation,
by subtraction
is
or division or both, and developing that side of the equation. Any equation may be put in the form in which one member
by
64
bringing all the terms to one side.

fully
When
this is done,
and the equation
expanded, all the coefficients which do not vanish may be changed to Boole calls this a "rule of unity, except such as already have that value.
102
interpretation".
Its validity follows
from two considerations:

0,
(1)
Any
not
constituent of an equation with one
member
whose
coefficient does
vanish in development,
may
the constituents thus separately equated to
be separately equated to 0; (2) the sum of will be an equation with one
member
in
which each
coefficients
coefficient will be unity.
Negative
may
be eliminated by squaring both sides of any
equation in which they appear.

are not altered
The
"
logical symbols" in
any function
x),
by squaring, and any expression

symbol",
is
of the
form
(1
where
is
"logical
not altered, since

is
it
and
1.
Hence no constituent
altered, except that its coefficient
can have only the values may be
altered.
And any
negative coefficient will be
made
positive.
Xo new
effected
terms
will
be introduced by squaring, since the product of any two terms

is
of a developed function
always
null.
is
Hence the only change

is
by squaring any developed function

ficients into positive.
the alteration of any negative coef

of
Their actual numerical value

1
no consequence,
because coefficients other than

above.
can be dealt with by the method described

single equation, Boole
103
For reducing any two or more equations to a

first
proposed the
"method
of indeterminate
each equation, after the first, is equations then added. But these indeterminate multipliers complicate the process of elimination, and the method is, as he afterward recognized, an
inferior one.
by which multiplied by an arbitrary constant and the

multipliers",
be reduced, by the methods just described, to the form in which one member is 0, and each coefficient is 1. They may then be simply added; the resulting equation will combine
simply, such equations
More
may
the logical significance of the equations added.
Any
logic.
"logical
symbol"
which
is is
not wanted in an equation

familiar to
is
may
be
eliminated by the
method which
all
students of symbolic
To
eliminate x, the equation
reduced to the form

a-)
Ax + B(l The
result of elimination will be 104
AB =
102
L. of T., p. 90.
See Math. An. of Logic, pp. 78-81; L. of T., pp. 115-120. See L. of T., p. 101. We do not pause upon this or other matters which will be entirely clear to those who understand current theory.
104
103
65
By
these methods, the difference between Boole

it is
algebra and the classic
algebra of logic which grew out of

results obtainable
reduced to a minimum.
classic algebra
Any
logical
Boole
by the use of the procedures. The difference is
may
also be got
by
solely
one of ease and mathematical

classic algebra
neatness in the method.
Two
s
important laws of the
which
do not appear among Boole

(1)
principles are:
x,
2xy y). with coefficients other than unity, 2x may be replaced in the equation by x, and 2xy + x(l - y) by xy + x(l - y), which
algebra, the methods of reduction which have been discussed will always give x in place of x + x or of x + xy, in any equation in which these appear. The expansion of x + x gives 2x-, the expansion of x + xy gives + x(l
x + xy These seem to be inconsistent with the Boolean meaning of + the first them does not hold for x =-- 1; the second does not hold for x = 1, y = But although they do not belong to Boole s system as an abstract
(2)
x +x
and
of
1.
By
the
method
for dealing
is
equal to x.
should
The methods of applying the algebra to the relations of logical classes now be sufficiently clear. The application to propositions is made
familiar device of correlating the
"
by the
times
logical
symbol",
x,
with the
when some
proposition,
is
true,
xy
X and
times
will represent the
times
1"
when
are both true; x(l
y), the times
when
of
A"
is
true
and
is false,
and so on.
Congruent with the meaning

A"
x+y
will represent the
the times
when either when A
or
(but not both)
is
true.
or
or both are true,

1,
we
xy + x(l
- y)+
and
(1
In ord-r to symbolize must write x + (1 x)y, or

represent
"all
x)y.
the
"
universe",
will
times"
or
"always";
"A
will
be
"no
time"
or
==
"never",
will represent
is
is
always
true";
or
(i
.r)
1,
"A"
is
never true,
always
false".
Just as there
is,
classes, so there is
with Boole, no symbol for the inclusion relation of no symbol for the implication relation of propositions.
is
T"
For
classes,
"All
A"
or
"A"
is
contained in
Y"
becomes x
vy.
Cor
"If
respondingly,
"All
times
when
then
A"
"
or
"A
implies
F"
when
is
true and the times
when Y is or = vy. x = y will mean, "The times when Y is true are the same" or implies
A"
is
true are times
true"
is
"A
Y and Y
The
follows:
105
implies
\
"secondary
propositions"
is
entire procedure for

105
summarized as
L. of T., p. 178.
66
A
"Rule.
"Eliminate
Express symbolically the given propositions. separately from each equation in which
.
it
is
found the
indefinite
symbol
"Eliminate
the remaining symbols which
it is
desired to banish from
the final solution: always before elimination, reducing to a single equation

those equations in which the symbol or symbols to be eliminated are found.
Collect the resulting equations into a single equation [one
isO],
member
of
which
V =
0.
"Then
proceed according to the particular form in which
it is
desired
to express the final relation, as

1st.
If in
the form of a denial, or system of denials, develop the

all
function V, and equate to
those constituents whose coefficients do
not vanish.
2ndly.
If in
the form of a disjunctive proposition, equate to
the
sum
of those constituents
If in
whose
coefficients vanish.
3rdly.
the form of a conditional proposition having a simple

1
element, as x or
x,
for its antecedent, determine the algebraic
expression of that element, and develop that expression.
the form of a conditional proposition having a com .r)(l pound expression, as xy, xy + (1 y), etc., for its antecedent,
4thly.
If in
equate that expression to a new symbol
t,
and determine
as a developed
. .
.
function of the symbols which are to appear in the consequent.
5thly. only be desired to ascertain whether a particular elementary proposition x is true or false, we must eliminate all the
If it
...
symbols but
tion
is
x\
then the equation x
true, x
that
it is false,
= 1 =
will indicate that the proposi
that the premises are insuf

false."
ficient to
determine whether
it is
true or
It is
a curious fact that the one obvious law of an algebra of

is
and
which Boole does not assume
exactly the law which would have limited
the logical interpretation of his algebra to propositions.

If
is is
The law
1
4=
1,
and
if
4= 0,
exactly the principle which his successors added to his system

to be considered as a calculus of propositions.
his
when
it
This principle would have
made
and
system completely inapplicable to logical classes. For propositions, this principle means, If x is not true, then x
"
is false,
if
is
not
false, it is
true".
But
careful attention to Boole

in his
interpre
tation for
"propositions"
shows that
system x
should be inter-

preted
"x
67
is
"x
false at all times (or in all

is
terpreted
"
true at
all times".
1 should be in This reveals that fact that what Boole

cases)",
"
and x =
calls
propositions"
is,
are
what we should now
that
statements which
may
prepositional functions be true under some circumstances and false
call
",
upon what we now call "propositions" namely that they must be absolutely determinate, and hence simply true
limitation put
or false
sitional
under others.
The
does not belong to Boole

symbols"
system.
And
his
treatment of
"prepo
gives
them
in the application of the algebra to probability theory the character of prepositional functions" rather than of our
"
absolutely determinate propositions.
one hundred and seventy-five pages of the Laws of Thought are devoted to an application of the to the solution of problems in algebra 106 This application amounts to the probabilities. of a new
last
The
method
is
inyention involved in the problem logical analysis performed as automatically as the purely mathematical operations. a
method whereby any
We
can make this provisionally clear by a single illustration All the objects belonging to a certain collection are classified in three
:
ways
as
^4 s
or not, as
or not, and as
s
C
s
or not.
(2)
It is
then found
that (1) a fraction
m/n
of the .4
are also
B
s
and
the
consist of the
s.
A
it
which are not
together with the

if
which are not
Required: the probability that will also be a C.
one of the
be taken at random,
By
premise
(2)
C = A(l - B)
Since A, B, and
Hence,
C are = A* (I - B) AC
"logical
+ .1(1
s
- A)B = A and A (I - A) = symbols", A - A)B = A(l - B).

+
(l
2
0.
The
that
which are also C
are identical with the
which are not
s.
Thus the probability that

it is
a given
is
also a
is
exactly the probability
m/n, which is the required solution. In any problem concerning probabilities, there are usually two sorts of difficulties, the purely mathematical ones, and those involved in the logical
;
not a
or
by premise
(1 ), 1
analysis of the situation
The methods
106
of Boole
upon which the probability in question depends. algebra provide a means for expressing the relations
and then transforming these
logical
of classes, or events, given in the data,
See also the Keith Prize essay "On the Application of the Theory of Probabilities to the Question of the Combination of Testimonies or Judgments", Trans. Roy. Soc. Edinburgh, xxi, 597 ff. Also a series of articles in Phil. Mag., 1851-54 (see An article on the related topic "Of Propositions Numerically Definite" appeared Bibl).
Chap. 16 ff.
posthumously;
Carafe. Phil. Trans., xi,
396-411.
68
equations so as to express the class which the quaesitum concerns as a func It thus affords a method for untangling tion of the other classes involved.
the problem
often the most difficult part of the solution.
The
Boole
is
parallelism between the logical relations of classes as expressed in
algebra and the corresponding probabilities, numerically expressed,
striking.
the event
107
A".
Suppose x represent the class of cases (in a given total) in which the occurrence of occurs or those which "are favorable
to"
The
then
p be the probability, numerically expressed, of the event X. total class of cases will constitute the logical "universe", or 1; the
And
let
null class will be 0.
Thus,
if
1
is
if all
the cases are favorable to

If
X
0.
p =
the probability of
X
=
"certainty".
0,
then p
Further, the class of cases in which

1
X does
1.
not occur,
is
will
be expressed by
1
x] the probability that

(1
X will not occur

p)
the numerical
p.
Also,
x+
x)
and p +
(1
This parallelism extends likewise to the combinations of two or more If x represent the class of cases in which events. occurs, and y the class
of cases in
which
occurs, then xy will be the class of cases in
which
and 1 both occur;

(1
x(l
y), the cases in which
occurs without Y;
(1
x)y, the cases in
which
occurs without
y~)
X;
x)(l
y),
the
cases in
X or
and
which neither occurs; x(l occurs but not both, and so on.
events,
+ y(l
x), the cases in which
Suppose that
X and
of
Y are
X,
"
"
simple
"independent"
and
let
p be the probability
q the
prob
ability of Y.
Then we have:
of events
s
Combination
expressed in Boole
algebra
Corresponding probabilities numerically expressed
xy
pq
p(l
(1 (1
- y) x(l - x)y (1
(1
- q) - q)p
+
(1
-*)(! -y)
-p)(l -q)
x(l
y) +
(1
x)y
p(l
q)
p)q
Etc. etc.
In fact, this parallelism
is
complete, and the following rule can be
formulated:
107
108
asserting
thing.
108
Boole prefers to consider x as representing the times when a certain proposition, an occurrence, will be true. But this interpretation comes to exactly the same
L. ofT., p. 258.

"If
69
p, q,
r,
events, x, y,
z,
are the respective probabilities of unconditioned simple the probability of any event V will be [V], compound
,
this function [V] being

x, y, z,
. . .
formed by changing,
.
.
in the function F, the
symbols
into p,
q, r,
"According
the event
V occur,
to the well-known law of Pascal, the probability that if the event will occur with it, is expressed by a fraction
whose numerator
the probability of the joint occurrence of V and V, and whose denominator is the probability of the occurrence of V. We can then extend the rule just given to such cases:
is
"The
probability that
if
the event
occur,
any other event
will
also occur, will be
[V
-
V]
-
where [V
V]
denotes the result obtained by
multiplying together the logical functions

result x, y,
z,
. . .
and V, and changing
in the
into p,
q, r,
."
The inverse problem of finding the absolute probability of an event when its probability upon a given condition is known can also be solved. Given The probabilities of simple events x, y, z, are respectively when a certain condition V is satisfied. r, p, q,
:
To determine:
the absolute probabilities
/,
m,
n,
of x, y,
z,
By
the rule just given,
U f]
>
[yV]
[zV]
W\
>
And
I,
the
7i,
m,
number of such equations will be equal to the number of unknowns, ... to be determined. 109 The determination of any logical expres
form
sion of the
xV
is is
peculiarly simple since the product of x into
any
developed function
as a factor.
if
the
sum
of those constituents of
V which
*)(!
contain x
For example:
xyz + x(l
V =
xV = xyz + x(l y)z = xyz+(l -x)y(l - z) yV zV = xyz + x(l - y)z + (1 Thus any equation
of the
y)z + (1
x)y(l
z)
(1
y)z,
.r)(l
y)z
form
109
On
Question in the Theory of
xxm
ibid.,
certain difficulties in this connection, and their solution, see Cayley, "On a Probability" (with discussion by Boole), Phil. Mag., Ser. iv,
(1862), 352-65,
and Boole,
"On
a General Method in the Theory of
Probabilities",
xxv
(1863), 313-17.
70
is
example
"
readily determined as a numerical equation. in illustration: no
Boole gives the following
Suppose that in the drawings of balls from an urn, attention had only been paid to those cases in which the balls drawn were either of a particular
white/ or of a particular composition, marble/ or were marked by both of these characters, no record having been kept of those cases in which
color,
which was neither white nor of marble had been drawn. Let it then have been found, that whenever the supposed condition was satisfied, there
ball
was a probability p that a white ball would be drawn, and a probability q that a marble ball would be drawn: and from these data alone let it be
required to find the probability
ence at
all
that in the next drawing, without refer to the condition above mentioned, a white ball will be drawn;
also a probability
"Here
n that a marble
ball will
be drawn.
if
ball,
the condition
x represent the drawing of a white ball, y that of a marble F will be represented by the logical function
xy + x(l
y) + (1
x)y
Hence we have
xV = yV =
Whence
[xV]
xy + x(l y) + (1 - x)y xy
= =
or
m,
[yV]
= n
and the
final
equations of the problem are
m
mn + m(l
n) + (1
m)n
m)n
+ =p
= P
=
q
mn + m(l
from which we find
n) + (1
m =
p+
q
q
... To meet a possible objection, I here remark that the above reasoning does not require that the drawings of a white and a marble ball should be independent, in virtue of the physical constitution of the balls.
"In
general, the probabilities of
any system
being given, the probability of

logical equation of the
any event
X may
+
independent events be determined by finding a

of
form
x
= A+OB
^C
ID
by a change
of letters.
110
L. of T., p. 262.
have
slightly altered the illustration
71
where A, B, C, and D are functions of the symbols of the other events. As has already been shown, this is the general type of the logical equation, and its interpretation is given by
= A D =
x
+ vC,
where
v is arbitrary
and
By
the properties of constituents,
we have
also the equation,
A+B+C+D
and, since
=
1
D =
0,
A+B + C
A+B+ C
abilities
thus gives the
universe
of the events in question,
and the prob
given in the data are to be interpreted as conditioned by A + B + C = 1, since = is the condition of the solution x = A + vC. If the given probability of some event S is p, of T is q, etc., then the supposed absolute
probabilities of S, T, etc.,
may
be determined by the method which has
been described.
Let
V = A+B+C,
[sV]
then
W]
where
[sV],
[tV],
etc.
-=P,
M
[IV]
-?,
probabilities"
are the
"absolute
sought.
These,
being determined,
may
be substituted in the equation

Prob.
w =
[A + vC]
\yr~
which
will furnish the required solution.
"The
term vC
will
appear only
in cases
where the data are

it
insufficient
to determine the probability sought.

this probability
Where
v
does appear, the limits of
may
be determined by giving
the limiting values,
and
1.
Thus
Lower
limit of Prob.
w =
\A] -
[A +
Upper
limit
cr
method, and with the theoretical difficulties of its application and interpretation, we need not here concern ourselves. Suffice it to say that, with certain modifications, it is an entirely workable
detail of this
With the
marked advantages over those more matter of surprise that this immediately useful application of symbolic logic has been so generally overlooked.
method and seems
to possess certain
generally in use.
It is a
72

VI.
JEVONS
s
"calculus
It
has been shown that Boole
of
logic"
is
not so
of
much
system of logic as an algebra of

or propositions,
reasoning.
If
the numbers
and
1,
some
whose ex
pressions are capable of simple interpretation as relations of logical classes,
and some
of
whose transformations represent processes

Boole himself failed to recognize this
of
the entire algebra can, with sufficient ingenuity, be inter

still
preted as a system of logic,
fact,
and
this indicates the difficulty
and unnaturalness
of
some parts
s
of this
interpretation.
Jevons 111 pointed a
way
to the simplification of Boole
algebra, dis
carding those expressions which have no obvious interpretation in logic,
and laying down a procedure which is just as general and is, in important In his first book on this subject, Jevons says: l12 respects, superior.
"So
long as Professor Boole
system of mathematical logic was capable

it
of giving results
beyond the power of any other system,
had
in this fact
an impregnable stronghold. Those who were not prepared to draw the same inferences in some other manner could not quarrel with the manner
of Professor Boole.
But
if
it
chapters
is
of equal
power with Professor Boole
be true that the system of the foregoing s system, the case is altered.
of
There are now two systems of notation, giving the same formal results, one which gives them with self-evident force and meaning, the other by dark
processes.
and symbolic
The burden
of proof
is
shifted,
and
it
must be
in
for the author or supporters of the
dark system to show that
it is
some
way
superior to the evident
system."
He sums up
follows:
"1.
the advantages of his system, compared with Boole
s,
as
113
Even- process is of self-evident nature and force, and governed by laws as simple and primary as those of Euclid s axioms.
"2.
The The
process
is infallible,
and gives us no uninterpretable or anom

less labor
alous results.
"3.
inferences
may
be drawn with far
than in Professor
Boole
system, which generally requires a separate computation and

for
development
111
each
inference."
William Stanley Jevons (1835-1882), B.A., M.A. (London), logician and economist; professor of logic and mental and moral philosophy and Cobden professor of political economy in Owens College, Manchester, 1866-76; professor of political Uni
economy,
versity College, London, 1876-80.

112 113
Pure Logic, or
Ibid., p. 74.
the Logic of Quality apart
from Quantity,
p. 75.
73
The
unduly
his
third
of these
observations
is
not entirely warranted.
Jevons
restricts the operations

is
and methods
of Boole in such wise that
own procedure
be expeditious.
often cumbersome and tedious where Boole s would Yet the honor of first pointing out the simplifications
in the algebra of logic belongs to
which have since been generally adopted

Jevons.
He
the
discards Boole
of a
inverse operations, a
"
and
a/6,
and he interprets
sum
and
b as
either a or
shall
6,
where a and
b are not necessarily
exclusive
classes".
(We
114
A+B
is
or
-J^.)
symbolize this relation by a + b: Jevons has Thus, for Jevons, a + a = a, whereas for Boole a + a
if it
not interpretable as any relation of logical classes, and

in the algebra of
be taken as
an expression
so that a + a
and
1, it
obeys the usual arithmetical laws,

is
2a.
As has been
and
indicated, this
a source of
much awk
ward procedure
in Boole s system.
The law a + a = a
eliminates numerical
1, and this is a most important simplification. Jevons supposes that the fundamental difference between himself and Boole is that Boole s system, being mathematical, is a calculus of things taken in their logical extension, while his own system, being "pure
coefficients, other
than
logic",
is
a calculus of terms in intension.
It
is
true that mathematics requires
that classes be taken in extension, but
it is
not true that the calculus of
logic either requires or derives important advantage

of intension.
from the point of view
Since Jevons
system can be interpreted in extension without

supposed difference.
:
the slightest difficulty,
we
shall ignore this
The fundamental
(1)
ideas of the system are as follows
a b denotes that
which
is
both a and
b,
or (in intension) the
sum
of the
meanings of the two terms combined. (2) a + b denotes that which is either a or
b or both, or (in intension) a
term with one of two meanings. 115 = b a is identical with 6, or (3) a

(4)
1
(in intension) a
means the same

Boole
s
as
b.
-b
116
Not-6, the negative of
b,
symbolized
in
system by
6.
(o)
"excluded
114 115
indicates that which is contradictory or According to Jevons, from thought". He prefers it to appear as a factor rather than
\
A +B
in Pure Logic; A Jevons would add "but
it is
in the other papers. not known which".
(See Bibl.) (See Pure Logic, p. 25.)
But
hardly correct; it makes no difference if it is known which, since the meaning of a + 6 does not depend on the state of our knowledge. Perhaps a better qualification would be "but it is not specified which". 113 Jevons uses capital roman letters for positive terms and the corresponding small
this is
italics for their negatives.
Following
De Morgan,
he
calls
and a
"contrary"
terms.
74
117
A
itself.

is
by
intension.
a proper interpretation of the symbol in Its meaning in extension is the null-class or "nothing", as with
The meaning given
Boole.
Jevons does not use any symbol for the "universe", but writes out the This "logical alphabet", for any number n of ele "logical alphabet".
ments,
a, 6, c,
. . .
n consists of the 2 terms which, in Boole 1. b,
system, form
the constituents of the expansion of

the
"logical
alphabet"
consists of a
z,
terms, x - Zf -x
y, z, it
consists of x y
x y
-z,
two elements, a and 6, Thus, a-b, -ab, and -a -b. For three x -y z, x -y -z, x y -z, -x -y z, and
for
Jevons usually writes these in a column instead of adding them -y and putting the sum == 1. Thus the absence of 1 from his system is simply
a
whim and represents no The fundamental laws

(1) (2) (3) (4) (5) (6)
(7)
real difference
from Boole
procedure.
of the
}
system
of Jevons are as follows:
(8) (9)
If a = b and b = c then a = c. a b = b a. a a = a. a -a = 0. a + b = b + a. a + a = a. = a. This law is made use of but is not stated. a+ = a b + a c and (a + b)(c + d) =ac + ad + bc + bd. a(b + c) = a. This law, since called the of absorption", a+ a b
"law
allows
a direct simplification which

multiplication
is
not possible in Boole.
Its
analogue for
a(a+
follows from
(8), (3),
b)
a
of absorption extends to
and
(9).
The law
also
.
any
number
of terms, so that
we have
a+
(10)
ab + ac + ab d+
This
is
=
a,
a(b + -b)(c + -c) ....
the rule for the expansion of

b, c,
any term,
it
with reference to any other terms,
etc.
For three terms
gives us
a(b + -b)
is
(c
+ -c)
= abc
ab-c + a-bc + a-b-c

in
This expansion
identical with that

.
which appears
Boole
system, except
But the product any two terms of such an expansion will always have a factor of the form a -a, and hence, by (4), Thus the terms of any expansion will always represent classes will be null.
for the different meaning of +
of
117
See Pure Logic, pp. 31-33.
75
which are mutually exclusive.

the different meaning of +
,
This accounts for the fact that, in spite of developed functions in Boole s system and in
Jevons
(11)
always have the same form.
The
"logical
alphabet" is
made up
this
.
of
any term plus

will
its
negative,
a + -a.
It follows
immediately from
and law
.
.
(10) that the logical
alphabet for any number of terms,
a, b, c,
be
and
will
have the character which we have described.

1
the expansion of
its
in Boole s
system because
it is
corresponds to a developed function and
It
terms are mutually exclusive.
procedure by which Jevons sets great store is the "substitution of similars of a for b or b for a when a = b. Not only is this procedure valid
",
when the
expressions in which a and b occur belong to the system, but it holds good whatever the rational complex in which a and b stand. He
first
considers this the

Aristotle
s
principle of reasoning,
et nullo. 11 *
more fundamental than
and yet there

is
is undoubtedly correct, another principle, which underlies Aristotle s dictum, which equally fundamental the substitution for variables of values of these is
dictum de omni
In this he
variables.
And
this procedure
is
not reducible to any substitution of
equivalents.
The only copulative

All a
relation in the system

is
of simple logical propositions

is b:
hence the expression the same as with Boole: substantially

is
[
= ];
No
is
is
a
c a
Some
"U"
is b:
"
= = =
a b a -b
c
ab
or
Ua = Va
used to suggest
of
Unknown".
The methods
than Boole
s,
working with
this calculus are in
some respects simpler
in
some respects more cumbersome.

s
But, as Jevons claims,
they are obvious while Boole

"intrinsic"
are not.
Eliminations are of two sorts,
and
"
extrinsic".
Intrinsic eliminations
substituting for any part of one

other.
member
get
of
may be performed by an equation the whole of the

=
a,
Thus from a =
a
b c d,
we
a c d
= ab
a c
a d
a b
This rule follows from the principles a a
b a,
and
if
b,
ac =
If
be.
For example
a-a
118
= bcd = bed-bed = bb-cc-dd = bcd-d =
ad.
See Substitution of Similars, passim.
76

-6), or of the
Also, in cases
where a factor or a term of the form a(b +
form
a -a,
is
involved, eliminations
0.
is
may
be performed by the rules a(b + -b)
and a -a =
which
Extrinsic elimination
that simplification or
or
"solution"
may
occur
when two
more are united.
of equations Jevons does not add or
multiply such equations but uses them as a basis for striking out terms in the same "logical alphabet".
This method
is
equivalent, in terms of current procedures, to
first
forming the expansion of 1 (which contains the terms of the logical alphabet) and then putting any equations given in the form in which one member is and "subtracting" them from the expansion of 1. But Jevons did not
hit
upon the current procedures.

"1.
His own
is
described thus:
119
Any
premises
being given, form a combination containing every
term
into
involved therein.
its
Change successively each simple term

all
of this
contrary [negative], so as to form
the possible combinations of

if a,
the simple terms and their contraries.
[E. g.,
b,
and
are involved,
form the
"logical
alphabet"
of all the
terms in the expansion of
(a
"2.
+ -a)(6 + -&)(c + -c).]

[or
Combine
successively each such combination
term, as a be,]
with both members of a premise.
When
the combination forms a con
tradiction [an expression having a factor of the form (a -a)] with neither side of a premise, call it an included subject of the premise; when it forms a
contradiction with both sides, call
it
an excluded subject
call it
of the premise;
when
it
forms a contradiction with one side only,
a contradictory
com
bination or subject,
"We
and
strike it out.
an included or excluded subject a possible subject as distinguished from a contradictory combination or impossible subject. Perform the same process with each premise. Then a combination
may
call
"3.
is
an included subject of a
series of premises,
of of
any one;
any one;
it is it
a contradictory subject
is
an excluded subject
when it is an included subject when it is a contradictory subject when it is an excluded subject of
every premise.
consists of
any term [as a or b] involved in the premises the included and excluded subjects containing the term, treated as alternatives [in the relation + ].
"4.
The
expression of
all
"5.
Such expressions
may
be simplified by reducing
all
dual terms
[of
119
Pure Logic, pp. 44-46.

the form a(b + -b)], and by intrinsic elimination of in the expression.
"
77
all
terms not required
6. When it is observed that the expression of a term contains a com bination which would not occur in the expression of any contrary of that term, we may eliminate the part of the combination common to the term
and
in
its
"7.
expression.
Unless each term of the premises and the contrary of each appear one or other of the possible subjects, the premises must be deemed in
consistent or contradictory.
possible subjects.
Hence there must always remain at
least
two
Required by the above process the inferences of the premise a = b c. possible combinations of the terms a, b, c, and their contraries are as given [in the column at the left, which is, for this case, the logical
"
"The
alphabet
].
Each
of these being
combined with both
sides of the premise,
we have the
following results:
abc
ab-c
a-b
c
ale
ab-c
a-b
c
= ab = ab
abc
included subject
contradiction contradiction
c -c
ab
-c
a -b -c
a-b
-c
= ab-bc = = a b -b c -c = =
-a
b c
a -b c a -b -c
contradiction
-abc
-ab-c
-a -be
-a -b -c
"It
= = = =
a -a b c
a -a b -c
-abc
=
contradiction
a -a -b c
= -ab c -c = -ab-b c
= -ab
-b
c -c
-a
b -c
c
excluded subject
excluded subject
=0
=
-a -b
a -a -b -c
-a -b -c excluded subject
appears, then, that the four combinations ab-c to -abc are to be struck out, and only the rest retained as possible subjects. Suppose we now require an expression for the term -b as inferred from the premise
b c.
Select from the included

c
and excluded subjects such as contain

c
-6,
namely -a -b
"Then
and -a -b
-c.
-b
= -a
c
-b
+ -a -b -c, but as -a
occurs only with -b, and

eliminate -b from
not with
6,
its
contrary,
we may, by Rule
0,
-a-b
hence -b
-a
+ -a -b
-c."
This method resembles nothing so much as solution by means of the Venn diagrams (to be explained in Chapter III). The "logical alphabet"
is
compartments in such a diagram those marked "contradiction" are the ones which would be struck out in the diagram by transforming the equations given into the form in which one member is 0.
a
list
of the different
78
The advantage which Jevons claims

ness,
for his method, apart from its obvious namely, that the solutions for different terms do not require to be
is
separately performed,
all
also
an advantage
of the diagram,
which exhibits
the possibilities at once.

If
any problem be worked out by

it will
of Boole,
alphabet"
this method of Jevons and also that be found that the comparison is as follows: The logical consists of the terms which when added give 1, or the universe.
"
Any term marked

ficient
"
"
contradiction
will,
by Boole
will
method, have the coef

will
or 1/0;
any term marked
"included subject"
have the coef
ficient 1;
any marked
is
"excluded subject"
have the
coefficient 0/0, or v
where
arbitrary.
If,
then,
we remember
that, according to Boole,
and thus eliminated, we two methods give substantially the same results. The single important difference is in Boole s favor: the method of Jevons does not distinguish decisively between the coefficients 1 and v. If, for example,
see that the
terms with the coefficient 1/0 are equated to
the procedure of Jevons gives x or x = v -y z.
x -y
z,
Boole
will give either
-y
One
further, rather obvious, principle
may
be mentioned
12
:
Any
subject of a proposition remains an included, excluded, or con
tradictory subject, after combination with any unrelated terms. This means simply that, in any problem, the value of a term remains its value as a factor when the term is multiplied by any new terms which may be
introduced into the problem. In a problem involving a, 6, and c, let a -be be a "contradictory" term. Then if x be introduced, a-bcx and a -b c -x will be "contradictory".
On
the whole Jevons
methods are
likely to be tedious
and have
little
of mathematical nicety
about them.
Suppose, for example, we have three
equations involving altogether six terms.

consist of sixty-four
The
will
"logical
alphabet"
will
members, each of which
have to be investigated
separately for each equation,

operations.
of rejecting
making one hundred and ninety-two separate

It
Jevons has emphasized his difference from Boole to the extent
much
that would better have been retained.
remained for
others,
notably Mrs. Ladd-Franklin and Schroder, to accept Jevons s amended meaning of addition and its attendant advantages, yet retain
s
Boole
methods
solution.
development and similar methods of elimination and But Jevons should have credit for first noting the main clue to
of
this simplification
120
the laws a + a
a and a + a
a.
Pure Logic,
p. 48.

VII.
79
PEIRCE
contributions of C. S. Peirce 121 to symbolic logic are more numerous and varied than those of any other writer at least in the nineteenth
century.
The
Boole
He understood how to profit by the work of his predecessors, and De Morgan, and built upon their foundations, and he anticipated
when he did not
himself.
the most important procedures of his successors even
work them out

be
recent developments in
Again and again, one finds the clue to the most the writings of Peirce. These contributions may
(1)
summed up under
three heads:
He improved
more
the algebra of Boole
by distinguishing the relations which are

are
characteristic of logical
classes (such as multiplication in Boole s algebra)
from the relations which
more
closely related to arithmetical operations (such as subtraction
and
division in Boole).
The
resulting algebra has certain advantages over the

it
methods of develop ment, transformation, elimination, and solution, and certain advantages over the algebra of Boole because it distinguishes those operations and
retains the mathematical
relations
system of Jevons because
which are always interpretable

"illative"
for logical classes.

in",
Also Peirce
into
introduced the
relation,
"is
contained
or
"implies",
Following the researches of De Morgan, he marked advance in the treatment of relations and relative terms.
symbolic
logic.
(2)
made The
method
of dealing with these
is
made more
precise
and
"mathematical",
and the laws which govern them are related to those classes. Also the method of treating "some" and
of Boole s algebra of
"all"
propositions as
sums
(2) and products (II) respectively of "propositions" containing This is the historic origin of "formal variables was here first introduced.
implication"
and
all
that has been built upon

(3)
it
in the
more recent develop
ment
of the logic of mathematics.
logic to be the science of mathematical form in general,
Like Leibniz, he conceived symbolic and did much to
revive the sense of logistic proper, as
we have used that term.
He worked
out in detail the derivation of various multiple algebras from the calculus of relatives, and he improved Boole s method of applying symbolic logic to
problems
121
in probability.
Charles Saylnders Peirce (1839-1914), son of Benjamin Peirce, the celebrated mathematician, A.B. (Harvard, 1859), B.S. (Harvard, 1863), lecturer in logic at Johns Hopkins, ^596- ?. For a number of years, Peirce was engaged in statistical researches for the U. S. Coast Survey, and was at one time head of the Department of Weights and Measures. His writings cover a wide variety of topics in the history of science, meta According to William James, his physics, mathematics, astronomy, and chemistry.
articles
on
"Some
Illustrations of the Science of
Logic",
Pop. Sri. Mo., 1877-78, are the
source of pragmatism.
80

in the order
is
We
shall take
up these contributions
named.
mainly in the 122 and in Logic",
The improvement
brief article,
"On
of the Boolian algebra

s
set forth
an Improvement in Boole
the Algebra
.of Logic".
123
Calculus of
two papers,
It will
"On
be remembered that Boole
calculus has four operations, or rela

of the
tions
a + b indicates the class

6;
[
made up
[
two mutually
if
exclusive classes,
a and
is
the strict inverse of
] t
so that
x+b
a,
then x
b;
denotes the class of those things which are common to a and 6; and division is the strict inverse of multiplication, so that if x b = a, then
b or a b
ax
afb.
[ ]
and
These relations are not homogeneous in type. Boole s [+] have properties which approximate closely those of arithmetical
If [n]x indicate
addition and subtraction.

class x,
the
number
of
members
of the
[n]a+ [n]b
[n](a + b)
is
because a and b are mutually exclusive classes, and every a member of (a + b) and every member of b is a member of
member
of a
relation, then, differs
a and b
larly,
This (a + b). from arithmetical addition only by the fact that are not necessarily to be regarded as numbers or Simi quantities.
[n]a
[n]b
s
[n](a
b or
b,
b)
But
in contrast to this, for
Boole
a x
[n]ax
will
[n]b
[n] (a b)
not hold except for
and
this relation
is
is
not of the type of

its
its
arith
metical counterpart.
And
the same
true of
inverse, a/6.
Thus, in
Boole
calculus, addition
as arithmetical addition
are different in
and subtraction are relations of the same type and subtraction; but multiplication and division from arithmetical multiplication and division. type
Peirce rounds out the calculus of Boole

relations,
by completing both sets of these adding multiplication and division of the arithmetical type, and
124 addition and subtraction of the non-arithmetical type. character of these relations is as follows
:
The
general
122
Proc. Amer. Acad., vn, 250-61.

".
This paper will be referred to hereafter as
"Boole
Calculus
123
will
Amer. Jour. Math., in (1880), 15-57, and vn These two papers (1885), 180-202. be referred to hereafter as Alg. Log. 1880, and Alg. Log. 1885, respectively.
124
"Boole
s Calculus,"
pp. 250-54.

A.
(1)
81
The
"
Non-Arithmetical" or Logical Relations

s
a+
denotes the class of those things which are either a

inverse of the above, a
or 6
or
both.
125
(2)
The
\-b.
\-b,
is
such that
if
x+b
a,
then
Since x and
6,
in
x+
b,
an ambiguous function.
a
\-
need not be mutually exclusive Suppose x + b = a and all b is x.
classes, a h b is
Then
x,
and
a h b
a
b
i.
Thus a
isx.
h b has
\-
an upper
b coincides
limit, a.
But suppose that x +

b (a
=
e.,
a and no b
Then a
with a
x,
which
a
\-
is
not a
6)
f-
and
6,
b
it)
Thus a
a -b.
f-
has a lower limit, a

in
or (as
we elsewhere symbolize
all
And
any
case, a
(-
b is
not interpretable unless
b is a, the
class b contained in the class a.
We may
summarize
[0]
all
these facts
by
a h b
a -6 + v a b +
[0]
b
it is
where
v is
undetermined, and
null.
indicates that the term to which
prefixed
(3)
must be
Boole
a
a b denotes the class of those things which are both a

s
b.
if
and
s.
This
is
(4)
The
is
126
inverse of the preceding, a/b such that

s a/b.
b x
o,
then x
a/6.
This
is
Boole
a/b
an ambiguous function.
Its
upper limit
is
a + -b;
i.
its
lower
limit, a.
It is uninterpretable unless b is contained in a
e.,
a/b
a b + v -a -b +
[0]
a -b
B.
(5)
The
"Arithmetical"
Relations
b denotes the class of those things 6 are
which are either a

This
is
or 6
s,
where a and
mutually exclusive classes.

a
Boole
a +
6.
a -6 + -a 6 +
[0]
a 6
"a
(6)
is
The
6".
inverse of the preceding,
b signifies the class
which
not
(7)
125
a
i
X
rce
pe
As has been mentioned, it coincides with the lower limit of a [-6. 6 and a -f- 6 are strictly analogous to the corresponding relations indicates the logical relations by putting a comma underneath the sign of
is
the relation: that which

126
"
both a and
b is a, b.
:
Peirce indicates the upper limit by a only in the paper Boole s Calculus".
6,
the lower limit by a
-5-
6.-
These occur
82
of arithmetic.
"
logical"
They have no such connection with the corresponding b. relations as do a + b and a Peirce does not use them
relations, the following familiar laws are stated:
except in applying this system to probability theory.
For the
"logical"
a+ a
= =
a
b
c
a a
a
b a
a+ b
(a +
+ a
ab =
a+
(b
b)
c)
(a b)c
= =
a(b
(a
c)
(a + b)c
= ac
+ bc
a b+ c
c) (b
c)
The
last
Peirce
two are derived from those which precede. s discussion of transformations and solutions
in this
system
is
inadequate.
Any sufficient account would carry us quite beyond what he has given or suggested, and require our report to be longer than the We shall be content to suggest ways in which the methods original paper.
As has been pointed
out,
if
of Boole s calculus can be extended to functions involving those relations
which do not appear in Boole. be developed by Boole s laws,

/(*)
<f>(*>
any function
=/(!)- x +/(0).-z,
y}
=
<p(l,
l)-xy +
<p(l,
Q)-x-y +
^(0, !)
-xy +
<p(Q,
0)-
-x-y,
Etc., etc.,
the terms on the right-hand side of these equations will always represent mutually exclusive classes. That is to say, the difference between the
"logical"
relation,
+, and the
"arithmetical"
relation,
+, here
vanishes.
Thus any
it
can be interpreted by developing to the above laws, provided that we can interpret these rela according
s
relation in this system of Peirce
tions
when they appear
in the coefficients.
And
the correct interpretation
of these coefficients can always be discovered.
Developing the
x+y
"logical"
sum, x +
y,
we have,
0)-
(l
+ l)-xy+(l +
Q)-x-y+(0+l)- -zy+(0 +
and (0+0) =
\-b,
-x-y
(1
Comparing
=
this
==
with the meaning of x + y given above, we find that

1,
1)
1,
(1
+ 0)
(0+1)
==
1,
0.
Developing the
x
\-y
"logical"
difference, a
we have
(0 f-0)
.
(1
\-l)-xy+(l hO)-z-2/+(0 hi)- -x y +

with the discussion of x
\-y
-x -
y
is
Comparing
127
this
above,
0;
we
see that (1 hi)

(1
equivalent to the
"Boole
undetermined
pp. 250-53.
coefficient
that
hO)
1;
that
s Calculus,"

(0 hi) is
83
it is
equivalent to
null,
[0],
which indicates that the term to which

(0 |-0)
prefixed
must be
and that
0.
The
interpretation of the
"arithmetical"
relations,
and ^,
in coef
not to be attempted. These are of service only in probability theory, where the related symbols are numerical in their
significance.
ficients of class-symbols is
The reader does not

Logic of
require to be told that this system
is
too complicated
to be entirely satisfactory.
Relatives", all
In the
"
Description of a Notation for the

-f-
these relations except

"
are retained, but in later
papers we find only the
logical"
in"
relations, a + b
and a
b.
The
relation of
"inclusion
or
"being
as small
as"
(which we shall
c) appears for the first time in the "Description of a Notation for the Logic of Relatives". 129 Aside from its treatment of relative terms and the use of the "arithmetical" relations, this monograph
symbolize by
gives the laws of the logic of classes almost identically as they stand in the
128
algebra of logic today.

(1)
The
130 following principles are stated.
If
cy and
b,
y cz, then x cz.

is
(2) If (3)
acb, then there

a c
such a term x that a + x such a term y that b y
=
a.
b.
If
then there
a,
(c
is
b.
(4) If b
(5)
=
6,
then a c
+ a) c
(c
If
a c
b)
(6)
(7) (8)
(9)
If
If
a cb,
a
c
a cb, a
cc b. cb c.
abca.
xc(x
y).
(10) x + y = y + x. (11) (x + y) + z = x + (12) x(y + z) (13) x y = y
(y
z).
= xy +
x.
xz.
(14)
(xy)z
x(yz).
(15) x x
(16) (17)
128
= x. x -x = O. 131 x + -x = 1.
is
<
Peirce s symbol
which he explains as meaning the same as
<
but being sim
pler to write.
129 130
Memoirs of
the
"Description
Amer. Acad., n. s., ix (1867), 317-78. of a Notation for the Logic of Relatives," loc.
is
cit.,
pp. 334-35, 338-39,

"
342.
131
In this paper, not-x
1 symbolized by n
"different
from every
x,
or
by
z
<r~
84
(18) z +
(19)
A
= = = x.
1.
x+
(20)
(21)
<p(x)
?(1)
[*(!)
*+*>(())
-a.
<K.r)
(22)
If
<p(x)
(23) If
= =
0,
1,
<p(x)
equation of condition and the elimination re Boole had stated (22), which sultant for equations with one member 1. is the corresponding law for equations with one member 0, but not (23).
The
last of these gives the
Most
by
of the
above laws, beyond
(9),
had been stated
either
Jevons.
(1) to (9) are, of course, novel, since the relation
by Boole or c appears
here for the same time since Lambert.
Later papers state further properties of the relation c

If
notably,
x c y, then -y c -x.
And
the methods of elimination and solution are given in terms of this

Also, these papers extend the relation to propositions.
"If
relation. 132
In this
is
interpretation, Peirce reads x cy,
a;
is
true,
is
true,"
but he
well
aware
of
"a;
of the difference
between the meaning of x

says:
133
cy and
is
usual significance
implies
is
y".
He
"It
stated above that this
means
if a: is
true, y
true
But
this
meaning
is
greatly modified by the circumstance that only the actual state

is
of things
referred
to.
... Now
the peculiarity of the hypothetical

if
proposition [ordinarily expressed by
is
true, y
is
true
is
that
it
goes
out beyond the actual state of things and declares what would happen were The utility of this is that it puts things other than they are or may be.
us in possession of a rule, say that if A we afterward learn something of which
is
true,
is
true
such that should
we
are
now
ignorant,
namely that
is
true, then,
by virtue
of this rule,
we
shall find that

.
we know something
.
.
else,
namely, that
true
if
is
true.
if
[In contrast to this]
the proposition,
is false.
ac
.
.
b, is
.
is
false or
b is true,
but
is false if
is
true while b
For example, we shall see that from -(x cy) [the negation of x cy] we can infer z ex. This does not mean that because in the actual state of
things x
is
true and y false, therefore in every state of things either z

;
is
false or x true
but
it
does
mean
that in whatever state of things

is false
we
find x
true and y
false, in
is
that state of things either z

anyway]."
or x
is
true [since,
ex hypothesi, x
132
true
133
2. Alg. Log. 1880, see esp. Alg. Log. 1885, pp. 186-87.
85
We now call this relation, x

theorems which are true of
of
it
"
cy,
material
implication,"
and the peculiar
are pretty well

if
them.
They
will
be intelligible
is
the reader
"The
actual state of things
not one in
Peirce gives a number remember that x c y means, which x is true and y false".
"A
known.
(1)
arc (yea:).
This
is
the familiar theorem:
true proposition is
implied by any
(2)
proposition".
[(xcy) ex] ex.
If
"x
implies
y"
implies that x
is
true,
then x
is
true.
(3) [(a: cy) ca] ex, where a is used in such a sense that means that from x cy every proposition follows.
(ar
cy) c a
The
meaning
time. 134
difference between
of
"implies"
"material
implication"
and the more usual

this
is
a difficult topic into which
But
it
is
interesting to
its
we need not go at note that Peirce, who introduced

some
of his successors
the
relation, understood
limitations as
have not.
Other theorems in terms of this relation are:

(4) (5) (6)
(7)
ar
c x.
[xc(ycz)] c [yc(xcz)].
xc[(xcy)cy]. (ar cy)c[(ycz)c

of the
(xcz)]..
This
is
a fundamental law, since called
the
"Principle
Syllogism".
Peirce
worked most extensively with the
logic of relatives.
His interest
here reflects a sense of the importance of relative terms in the analyses of
mathematics, and he anticipates to some extent the methods of such later

researches as those of Peano and of Principia Mathematica.
his successive papers
To
follow
on
this topic
would probably
shall
result in
complete con
fusion for the reader.
Instead,
Peirce:
we
(1)
make
three divisions of this entire
subject as
treated
by
the modification and extension of
De
Morgan
calculus of relatives
by the introduction of a more
"mathe
matical"
symbolism for the most part contained in the early paper, "Description of a Notation for the Logic of Relatives"; (2) the calculus
of relations, expressed without the use of exponents
and
in a
form which
makes
it
an extension of the Boolean algebra

its
a later development which
may
be seen at
best in
"The
Logic of
Relatives",
Note B
in the Studies
in Logic by members of Johns Hopkins University; consideration of the theory of relatives, which
papers, but has almost complete continuity.
134
and
(3) the systematic
is
scattered throughout the
But
see below, Chap, iv, Sect,
i,
and Chap,
v, Sect. v.
86
A
The terms
of the algebra of relatives

"ancestor",
may
usually be regarded as simple

etc.
relative terms, such as

will
"lover,"
Since they are also class
names, they obey may be taken for granted without further discussion. But relative terms have additional properties which do not belong to non-relatives; and it is to these that our attention must be given. If w signifies "woman" and s,
all
"servant,"
the laws of the logic of classes, which
logic
is
"logical
product"
concerned not only with such relations as "servant woman", s + w, the sum" "logical
both)
",
iv,
the
"either
servant or
in the
woman (or class women
and
relations
first
cw, "the class servants which belong to class-terms

s
is
contained
in general of a
but also with the relations

woman," "servant
symbolized
"
by De Morgan,
of
U5
"servant
of every
woman,
and
"servant
none but
This
is
women".
We may
represent
"servant
of a woman"
by s\w.
a kind of
"multiplication"
relation.
It is associative,
s\(l\w)
"Servant
=
is
(s\l)\w
"
of a lover-of-a-woman
"
servant-of-a-lover of a
woman".
Also,
it is
distributive with respect to the non-relative

,
"addition"
symbol
ized
by +
s\(m + w)
"Servant
= s\m + s w
a
"
of either a
".
man
or a
woman" is "servant of
:
man
or servant of
"
a
is
woman
But
it is
not commutative
"lover
s\lis not
l\s,
servant of a lover
not equivalent to
of a
servant".
To
distinguish s
from
s w,
or s x
call s
|
w w
the class of those
who
s
are both servants
and women
w
,
we
shall
the relative product of

"servant
and w.
proposed
s
For
of every woman" Peirce

s
and
for
"servant
of none but
women"
w.
As we
shall see, this notation is suggested
by
certain mathematical analogies.

of the class
as
W W W
ly
We may
and the
represent individual
class of all the
,
members
2,
3,
etc.,
Ws
as
W +W
l
+ ....
Remembering
the interpretation of +
we may
write
w =
and
that
this
is,
W W
1
W, +
means,
"The
class-term, w, denotes
or
or
s.
or
...,"
denotes an unspecified
member
of the class of PF
The servant
of a (some, any)
woman
is,
then, s w.
sw
"A
woman" is either
135
or
W
is s
or
3,
etc.;
s,
"servant
of a woman"
is
either
Peirce s notation for this
w, he uses
for the simple logical product.

servant of
87
or servant of
or servant of PF3 etc.

,
of every woman" is servant of

etc.
;
Similarly,
z
"servant
and servant of
,
and servant of
z,
or remembering the interpretation of x
n represents the relative product, and x s\ n of the non-relative logical product translated by "and". The represents above can be more briefly symbolized, following the obvious mathematical
"s
,"
where, of course,
analogies,
w = 2 W s\w = 2 w (s W) w = H s w (s\W)
|
Unless
represent a null class,
we
shall
have
or
jc*w
"servants
The
class
"servants
of every woman"
is
contained in the class
of a woman".
This law has numerous consequences, some of which are:
lover of a servant of
all
women
is
a lover of a servant of a
(/]*)
woman.
Kc
A
lover of every servant-of-all-women stands to every
woman
in the rela
tion of lover-of-a-servant of hers (unless the class s w be null). s w c Is w I

|
lover of every servant-of-a-woman stands to a (some)
woman
in the
relation of lover-of-a -servant of hers.
From
the general principle, 136
m\[U x f(x)]cU f (m\f(x)]

136
The
proof of this
theorem
...
. . .
is
as follows
a
or a
= abc = abc
+ab
-c ...
+a
-b c ... + ...,
+ P, where
is
the
sum
of the
remaining terms.
+
,
Whence,
if
represent any
relation distributive with respect to

.
Similarly,
mOa = mQabc mOb = mQa be mOc = mQa b c

Etc., etc.
let a, 6, c, etc.,
+mQP + mOQ
+
rnQR
Now
be respectively
the
left side,
/(.Ci), f(x*),
fM,
etc.,
and multiply together
all
the above equations.
On
we have
88
A
also,
we have
^c*,
If
/
or
8cs
woman
in the
lover of a (some) servant-of-every-woman stands to every
relation of lover-of-a-servant of hers.
We
have also the general formulae of inclusion, c s,

then
l
w w
c sw c
I
s
and,
If s
c w, then
The
first of
these means:
If all lovers are servants,
then a lover of every

:
woman is also a servant of every woman. The second means are women, then a lover of every woman is also a lover of
These laws
are, of course, general.
If all
servants
We
=
I
have
also:
every servant.
(l\s)\w
(s\w)
The
or a
last of these is read:
lover of every person
who
is
either a servant
woman
is
a lover of every servant and a lover of every
woman.
An
interesting law
which remainds us of Lambert

(I
"Newtonian formula" is,
s) w
+ 2 q (l w
-<*
x s9) + s w
One who
is
either-lover-or-servant of every
woman,
women,
is
is
either lover of every
woman
or, for
some portion
q of the class
lover of every
q, or, finally, is
woman
servant
except members of q and servant of every

of every
member
of
woman. Peirce also gives this law in a form which approximates even more closely the binomial theorem. The corresponding law for the
product
is
simpler,
(I
xs)
xs w
which
is
On
the right side,
we have
(raOP) + (mQQ) +
(mOabc
where
...) +
(mOR)
..,
or
(mQabc...)+K
is
is
sum
of other terms. ...}

is
But
(mOabc
?nQ[f(x
x f(x2 ) x /(z3 )
.],
which
mOH x f(x)
Hence [mOHf(x)]+K = U x [mOf(x)]. Hence m O n /Or) c n x [m Of(x)
]
.
Peirce does not prove this theorem, but illustrates

"Description
137
it
briefly for logical multiplication (see
of a
Notation",
p. 346).
"Description of
a Notation, p. 334.
89
One who
woman
both-lover-and-servant of every and a servant of every woman.

is
woman,
is
both a lover of every
Peirce introduces a fourth term, and summarizes in a diagram the inclu sion relations obtained by extending the formulae already given. 138 The
number
of such inclusions, for four relatives,
is
somewhat more than one

precise
hundred eighty.
He
challenges
the
reader to accomplish the
formulation of these by means of ordinary language and formal logic. An s of none but members of w, Peirce symbolizes by 8 w. He calls this
operation
"
"backward involution",
relatives",
and
relatives of the type
w he
refers to
as
infinitesimal
on account of an extended and
difficult
mathe
matical analogy which he presents. 139
The laws
l
of this relation are analo
gous to those of
sw
If s
If all
c w, then
c w
l
servants are
women, then a
If
/
lover of none but servants
is
lover of none
but women.
c s, then
wc w
l
If all
lovers are servants, then a servant of none but
women
is
a lover of
none but women.

i(*
w) =
("OH;
The
lovers
of
none but servants-of-none-but-women are the
lovers-of-
servants of none but
women.
i+*w
= w
l
Those who are either-lovers-or-servants of none but women are those who
are lovers of none but
women and
servants of none but
women.
(w xv)
= w
s
The servants
those
of
none but those who are both women and
violinists are
who
are servants of none but
women and
)&
servants of none but vio
linists.
<">MJC
Whoever
is
lover-of-a-servant of
none but
women
is
a lover-of-every-
servant of none but
women.
l\
wc
<*>w
lover of one
who
is
servant to none but
women
is
a lover-of-none-but-
servants to none but
women.
l
wc
(s
w)
138
Ibid., p. 347. Ibid., pp.
139
348 jf.
90
A
to a

in the relation of lover-of-nothing-but-servants
Whoever stands
of hers
is
woman
a lover of nothing but servants of
women.
by the laws:
The two kinds
of involution are connected
lover of none but those
who
are servants of every
woman
is
the same as
one who stands to every

servants of hers.
woman
i
in the relation of a lover of
none but
-/
- HO
It
Lover
from
s
of
none but servants

x
is
non-lover of every non-servant.

:
appears
this last that x

)
and y are connected through negation

a lover of every servant a lover of
is
= -O) =
-(I
-l\s,
l\-s,
Not Not
non-lover of a servant.
is
none but servants
lover of a non-
servant.
l
-s
~(l\s}
-I s ,
A
,
lover of none
but non-servants
is
one who
is
not lover-of-a-servant, a non-lover of every servant. -i s = = l~ s A non-lover of none but servants -(-/| -s)
is
one
who
is
not a non-lover-of-a-non-servant, a lover of every non-servant. We have the further laws governing negatives: 141
In the early paper,

of
Relatives",
the Description of a Notation for the Logic negatives are treated in a curious fashion. A symbol is
"On
used for
"different
s".
from
"
and the negative
of s
is
represented by
s
,
"
differ
ent from every
Converses are barely mentioned in this study. In the paper of 1880, converses and negatives appear in their usual notation, "relative addition" is brought in to balance "relative multiplication", and
the two kinds of involution are retained.
in the
But
in
"
The Logic
of Relatives
"
Johns Hopkins Studies in Logic, published in 1882, involution has

"relative addition"
disappeared, converses and negatives and
are retained.
This last represents the
final
form of Peirce
calculus of relatives.
We
have here,
(1)
Relative terms,
a, b,
... x, y, z.
(2)
140
141
The negative
ibid., p. 353.
of x, -x.
See
Not-z
is
here symbolized by
(1
z)
Alg. Log. 1880, p. 55.

(3)
91
The converse
a;
of x, ^x.
If
is
"lover",
^.r is
"beloved";
if
KC
is
"lover",
is
"beloved".
(4)
Non-relative addition, a +
b,
"either
a or
6".
(5)
(6) (7)
Non-relative multiplication, a x6, or a

"a
b, "both
a and
6".
Relative multiplication, a\b, of a Relative addition, a t b, of everything but 6

6".
"a
s,
a of every non-6
".
(8)
(9)
The The
relations
= and c
as before.
1,
universal relation,
"consistent
with,"
which pairs every
term with
(10)
itself
and with every
other.
1.
(11)
itself.
The The
null-relation, 0, the negative of
relation
"identical
with",
7,
which pairs every term with
(12)
The
relation
"different
from",
142
N, which pairs any term with
every other term which
is
distinct.
In terms of these, the fundamental laws of the calculus, in addition to

those which hold for class-terms in general, are as follows:
(1)
(2)
(3) (4) (5)
- a = w(- ) -(wo) = (-6 c (a c b)

.(.a)
If If
acb, then (a\x)c(b\x) and (x\a)c(x\b). acb, then (a t x) c (6 t x) and (x t a) c (x t
6).
(6)
(7)
(x a)|6 x\(a\b) x t (a t b) = (x t a) t b
(8)
(9)
x\(a-tb)c(x (a t b) x c a t
a)
t6
(6 x)
(10) (a\x) + (b\x)c(a+b)\x

(11) (12)
z|(at&)c(3ta)(a:t&)
(13)
(14)
(a+b)xc(a\x) + (b\x) (at*)(6tz)c(a|6) to: -(at 6) - -a -6

|
(15) -(a 6)
1
(16) (17)
(18)
(19)
= -at -6 ^(a+ 6) = â+ ^b = a ^6 (a b) t 6) = â t ^6 w(a w( |6) = v a |w6

1, 0,
/,
For the relations

given
142
:
and N, the following additional formulae are

may
see
have altered Peirce
notation, as the reader
by comparison.
92
A
(20)

(21)
Ocx
= x + 1 = 1 ztl = 1 1 t.T = 1 sttf = x Art* = x + -x = 1
x+
.r
xcl
= = z-0 z|0 = 0\x =
x
(22) (24) (26)

(28) (30) (32)
(23) x-l
(25)
(27)
(29)
(31)
z|/-cg
(33) /|z
(34)
(36)
or
(35) (37)
= x x -x =
fcr|
7c[a;1X-aO]
is,
(-*)]
ctf
This calculus
ciples of solution
as Peirce says, highly multiform,

laid
and no general prin
and elimination can be
down. 143
Not only
the variety
of relations, but the lack of

relative addition,
e. g.,
symmetry between
relative multiplication
and
in (10)-(13) above, contributes to this multiformity.
But, as
we now know, the

from
less
its
chief value of
any calculus
s
of relatives
is
not in
any elimination
or solution of the algebraic type, but in deductions to be
made
directly
formulae.
Peirce
is
devices for solution are, there
fore, of
much
importance than
his calculus of relatives is built.
the theoretic foundation upon which It is this which has proved useful in later
research and has been
made
the basis of valuable additions to logistic
development. This theory
is
practically unmodified throughout the papers dealing

"Description
Relatives"
with relatives, as a comparison of of Relatives" with "The Logic of
of a
in the
Notation for the Logic Johns Hopkins studies
and with the paper

"Individual"
of 1884 will indicate.

"elementary"
or
relatives are the pairs (or triads, etc.)
of individual things.
etc.,
If
the objects in the universe of discourse be A, B, C:

will constitute the
then the individual relatives
two-dimensional array,
A:A,A\B,A:e,A:D,
B
:
... ... ...
A,
B,
C,
D,
C:A, C:B, C:C, C:D,

. . .
Etc., etc.
itself is
be noted that any individual thing coupled with relative but that in general A B differs from B A
It will
: :
an individual
individual relatives
are ordered couples.
A
143
general relative
"Logic
is
conceived as an aggregate or logical

in Studies in Logic by
sum
of such
of
Relatives"
members of Johns Hopkins University,
p. 193.

individual relatives.
If b
93
represent
b
"benefactor",
then
= Z,S
y (6) t.,.(/
,/),
where (&)/
factor of J,
is
a numerical coefficient whose value

0,
is
in case / is a
bene
and otherwise
and where the sums are to be taken
for all the
individuals in the universe.
That
is
to say, b
is
the logical
is
sum
of
all
the
benefactor-benefitted pairs in the universe.

of
"definition
This
the
first
formulation
in
in
extension",
now widely used
in logistic,
though seldom
exactly this form.
By
It is
b
this definition, b is the aggregate of all the individual
relatives in our two-dimensional array
which do not drop out through having

of the form,
r
:
the coefficient 0.
some expression
= (X: y)!+(Z:
logical
Y) 2
+(X
,
) 3
+...
If,
"b
now, we consider the

is
meaning
of +
we
see that this

".
may
be read,
either
(X
F), or
(X
F) 2 or
(X
F) 8 or ...
is,
To say
that 6 repre
b repre
it
sents the class of benefactor-benefitted couples
then, inexact:
sents an unspecified individual relative, any one of this class.
should represent
"some"
in a sense
(That which denotes more than one at once

admits
is
which the meaning of +

fact that
exclusive.)
"real
in the general case
precluded by the
any two
distinct individual relatives are ipso facto

relative, so defined, is
mutually
calls
genera)
what Mr. Russell
variable".
Peirce discusses the idea of such a variable in a most
144 illuminating fashion.
"Demonstration of the sort called
mathematical
is
founded on suppo
the required condi
case, his
sition of particular cases.
The geometrician draws

quantity
a figure; the algebraist
assumes a
tions.
esis
is
letter to signify a certain
fulfilling
But while the mathematician supposes a particular
hypoth
yet perfectly general, because he considers no characters of the
must belong to every such case. The ad vantage of his procedure lies in the fact that the logical laws of individual terms are simpler than those which relate to general terms, because indi
individual case but those which
viduals are either identical or mutually exclusive, and cannot intersect or be subordinated to one another as classes can.
. .
.
"The
old logics distinguish between
indimduum signatum and
indi
mduum vagum. Julius Caesar is an example of the former; man of the latter. The indimduum vagum, in the days when
,
a certain
such con
ceptions were exactly investigated, occasioned great
difficulty
from
its
having a certain generality, being capable, apparently, of logical division.

144
"Description
of a Notation, pp. 342-44.
94
If
A
we
,

as
include under
indimduum vagum such a term

all
any individual
is
man
these difficulties appear in a strong light, for
what
true of
any
individual
man
is
true of
it
men.
Such a term
is in
it
one sense not an
individual term; for
represents every
man.
is
But
represents each
man
as capable of being denoted by a term which

it is
not
.
itself
. .
terms.
so, though an individual term, it stands for any one of a class of such The letters which the mathematician uses (whether in algebra
individual; and
or in geometry) are such individuals
by second
intention.
All the
formal logical laws relating to individuals will hold good of such individuals
by second
intention,
and at the same time a universal proposition
may
be
substituted for a proposition about such an individual, for nothing can be
predicated of such an individual which cannot be predicated of the whole

class."
The
It
is
relative
b,
fitted pairs in the universe,
denoting ambiguously any one of the benefactor-beneis such an individual by second intention.
of the
:
"
defined
by means
prepositional
function",
"I
"I
benefits
J"
J",
as the logical
sum
of the (7
J) couples for
which
benefits
is
true.
The compound
If
relations of the calculus can be similarly defined.
S,-Sy(a)iy(Z
J),
and
SiSy(&)<y(/
J),
then a +
S,-Sy[(a),-y +
(&),-,](/
J)
:
That
"7
is,
if
"agent"
is
the logical
sum
of all the (7
is
J) couples for which

of all the (7
:
is
agent of
J"
is
true,
and
"benefactor"
the
sum
"
J)
couples for
is
which
"7
benefits
J"
is
true, then
"either
agent or
benefactor"
is
the logical
sum
J"
of all the (7
is
J) couples for which
Either 7
facts
agent of
or 7 benefits
true.
We
might indicate the

+
same
6) t-,-.
more simply
by
defining only the
"prepositional function", (a
145
The
definition of a + b given above, follows
immediately from this simpler

relations are similar:
equation.
The
definitions of the other
compound
(ax6)
or
"Both
(a) t-yx(6) iy
:
a xb
SiSy[(a) tV x (6) iy ](7
J)
:
agent and
"7
benefactor" is
the logical
sum
J"
of the (7
is
J) couples for
which
is
agent of
J and
(a
I
is
benefactor of
true.
&).-/
S*{(a)*x(6) Ay
<A
or
145
a\b
",
S,-Sy[S*{(a)
188.
(b) hi ]](I
J)
See
"Logic
of Relatives
loc. cit., p.

"Agent
95
J) couples such
is
of a benefactor"
is
the logical
sum
//
is
of all the (7
that, for
some
//,
"I
is
agent of 77
and
benefactor of
J"
true.
There are two
difficulties in
"agent
the comprehension of this

of a benefactor".
last.
The
first
concerns the meaning of

treats
Peirce, like
De Morgan,
itself
his relatives as denoting
ambiguously either the relation

or the class
or
the things which have the relation
either relations or relative terms.
is
either the relation
"agent
of"
name
"agent".
Now
note
that the class

relation.
clear.
name denotes the first term in With this in mind, the compound
of a benefactor"
"
the pairs which have the

relation,
s
a\b, will
:
become
which
"Agent
names the 7
in the /
".
pairs
make up the field of the relation, agent of a benefactor of Any reference J s at the other end of the relation is gone, just as omits The second any reference to the J s in the field of the relation "agent concerns the operator, SA, which we have read, "For some difficulty
to the
"agent"
of".
H".
is
Consider any statement involving a "prepositional the variable representing the individual of which
function",
<p
<pz,
where
is
asserted.
That
is,
2z
3
<pz
symbolizes
",
"Either
<p
is
true of Zi or
<p
is
true of
"For
or
z
(p
is
true of
or ...
and
this
is
most simply expressed by

<pz
some
(some
"/
2 or other),
<pz".
In the particular case in hand,

is
is
(o)t*x(6) A y,
is
agent of // and //
the relation
"7
benefactor of
J".
The
terms, I and J, which stand in

are those for which there is
is
is
agent of a benefactor of
is
J",
some
H or other such that 7
agent of
"
H and H
benefactor of J.
<pz
Suppose we
ator n.
consider any
propositional
function",
with the oper
Tl 2 (pz
=
<f>Zi
(pZ/z
<pZs
x
<p
...
is
That
of
3
is,
"
<pz
symbolizes
",
is
<p
true of Z\ and
z".
true of
and
<p
is
true
Z and
or
b.
"
is
<p
true for every
This operator
is
needed
in the
definition of a t
(ot&),/- n A {(a)
"
+(6) A
is
,-}
agent of everyone but benefactors of either 7 is agent of 77 or is benefactor of

is
"
J"
equivalent to
For every
77,
H ot6-
J".
2&[Il h {(a) ik +(b) h j}](I:J)

is is
"Agent
of all non-benefactors"
the logical
sum
of all the (7
is
J) couples
such that, for every H, either 7
agent of
H or 77
benefactor of J.
The
same considerations about the ambiguity
of relatives
denoting either the
96
relation itself or those things

in this case also.
which are
first
terms of the relation

still
applies
We
need not, for the relations
to be discussed, con
"prepositional
sider the step

tions",
from the definition of the compound

is
func
(a t b)a in the above, to the definition of the corresponding relation,
a t
b.
This step
always taken in exactly the same way.
The
converse, converse of the negative, and negative of the converse,
are very simply defined.
That the negative
of the converse
-(6).-/
is
the converse of the negative follows

.
from the obvious fact that
(-6)/ f
All the formulae of the calculus of relatives,

also to the calculus of non-relative terms, 146
definitions.
beyond those which belong may be proved from such
For example:
To
prove, v(a + b)
(a
â + ^b
(a
&),-/
&),-,-
(a) /t-+
(6) yf
t
(&)/<
But
(a)ji
= (a)
t -/,
and
w o) (
(w6) tV
Hence
(a + &)/
:
-/+
(W>)v
of
Hence S<S,Ka + 6),-,-} (7 J) S;Z y {(-a) t7 + (&)/}(/ J) Q.E.D. For the complete development of this theory, there must be a discussion the laws which govern such expressions as (a) or in general, expressions
:
,-,-,
of the
is
form
(px,
where
<px
is
a statement which involves a variable, x, and
(px
either true or false
whenever any individual value

"
of the variable is
functions".
147
specified.
(a)*,
or in
Such expressions are now called the more convenient notation,

or
<p(x,
propositional
is
<px,
a propositional function of
one variable;
(a),-/, y), may be regarded as a propositional function of two variables, or as a function of the single variable, the individual rela
tive (7
J), or
(X
Y).
is
This theory of propositional functions
stated in the paper of 1885,
On
the Algebra of
Logic".
It
is
assumed, as also in earlier papers, that
4 he laws
of the algebra of classes hold for propositions as well. 148

is
The
time.
Additional law which propositions obey
stated here for the
first
46 The formulae of the calculus of classes can also be derived from these, considered as themselves laws of the calculus of propositions (see below, Chap, vi, Sec. iv). 147 Peirce has no name for such expressions, though he discusses their properties acutely
(see Alg. Log. 1880,

148
2).
first
This assumption
appears in Alg. Log. 1880.
97
The current form

"If
of
this
law
is
"If
4= 0,
then x
is
1",
which gives
is
& * 1, then x = immediately and if x is not true, then x is false".

"false",
0"
"If
not
false,
then x
true,
Peirce uses v
is
instead of
and
0,
and the law

(x
stated in
and / for the form
"true"
and
-f)(v-x)
=0
though derived from the
"x
But the calculus

true nor false.
in
of
propositional
is
functions,
algebra for propositions,
not identical with
it.
is
man" is
neither
A
"x
propositional function
a:
may
a
be true in some cases, false

true in
all cases,
some
cases.
"If
is
a man, then x
is
mortal" is
or
true of any x;
of x.
is
man" is
true in some cases, or true for
some values
For reasons already suggested,
2x
"
<px
represents
<px
<?x
is
2
true for
is
some value
<px
of the variable, x
. .
."
that
is,
either
is
true or
<px
true or
3 is
true or
Similarly,
TL x TL t
tf
<px
=
<px
(pxi
<px
<px
...
represents
true and
(a) xy ,
,
<px
tpx
2
is
true for
all
is
values of the variable, x

true and
..."
that
is,
<pxi
is
is
true and #r 3
If
or
more conveniently,
\f/(x,
<p(x,
y), represent
"x
is
agent of
y",
y",
and
(b) xy or
more conveniently,
IL x
y),
mean
"x
is
benefactor of
then
2 y [<p(x,y) *t(x,
y)]
y,
"x
will
mean
is
that for
all
values of x and some values of

y"
is
agent of y
and x
benefactor of
is
true
that
is,
it
represents the proposition
"Everyone is
if
both agent and benefactor of

[<f>(x,
someone".
This
will
appear
we expand n x 2 y
y) x^(.r, y)]:
2/0
2/j)
xiKzi,
,
2/i)]
+
+
[<?Oi,
2/2)
x x
{[<p(x
2,
x^(.r 2 yi)]
[<f>(x
z,
y z ) x^(z z y 2 )] +
,
..)
.
U</>(>3,
2/i)
x^(.r 3
2/i)]
[^(.r 3 , 2/2)
x^(.r 3 2/2)]+
,
... Etc., etc.

"
This expression reads directly

factor of yi] or [xi
[x z is
is
{Either [xi
is
agent of y 2 and Xi
a: 2
agent of y\ and x\ is bene benefactor of y z or .[ and {either

is
]
. .
agent of
2/1
and
. .
is
benefactor of
{either [x s is
.r 3
y\] or [x 2
is
agent of y 2
is
.
and x 2
is
bene
factor of y 2 ] or
2/i]
.}
and
agent of y\ and x 3
]
.
.
benefactor of
.
or [# 3 is agent of y 2
8
and
is
benefactor of y z or
and
Etc.,
etc".
98
A
Z,

"Boolian"
The operator
it,
which
is
nearer the argument, or
as Peirce calls
II,
indicates the operation,
within the
lines.
i.
The
e.,
outside operator,
in the
indicates the operation,
the subscript of
columns; and the operator nearer the Boolian indicates the letter which
x, between the lines
line to line.
little
varies within the lines, the subscript of the outside operator, the letter
which varies from

sional array.
Three operators would give a three-dimen
With a
patience, the reader
may
learn to interpret
any such expression directly from the meaning of simple logical sums and For example, with the same meanings of logical products. y) and
<p(x,
t(x, y),
lUZj^Cr,
will
y) xifr(y, x)]
(y)
mean
"Everyone
(x)
is
agent of some
benefactor of
himself".
(Note the order of the variables in the Boolian.)
And
"
will
symbolize
There
is
is
is
agent of z or z
benefactor of
is
some x and some y such that, for every z, either or, more simply, "There is some
y";
pair,
x and
y,
such that x
agent of
all
non-benefactors of
y".
The laws
for the
:
manipulation of such Boolians with

149
n and 2
operators
are given as follows

"1st.
The
different premises
having been written with distinct indices
(the same index not being used in two propositions) are written together, and all the n s and Z s are to be brought to the left. This can evidently be
done, for
[Or in the
more convenient, and probably more

Il x
<f>x
familiar, notation,
Xll y
<py
= U xU y
(<px
x
<py)
x
<py)]
"2d.
Without deranging the order

and Z
s
of the indices of
any one premise,
the
n
149
belonging to different premises
may
be
moved
relatively
left
to one another,
and as
pp.-
far as possible the
should be carried to the
Alg. Log. 1885,
196-98.

of the
II s.
99
We
have
n,-ny.T t-y
S.-SyZiy
= n itey = SyZtf.-y
;
[Or, [Or,
mix*,
S.S^fo
y)
y)
= njiôr, y)] = S y S z ^(a-, y)]

x^y)J
and
also
this
Silly&.-yy
= n/Z^y
[Or, SJI y (^r x

i
^) = n y S x (^r
But
formula does not hold when
and
j are not separated.
We do
have, however,
SillyZiy
It will, therefore,
-<
IlyZtf.v
[Or,
ZJIâr,
y)
c n,2 x ^(.r,
y)]
be well to begin by putting the 2 s to the left as far as because at a later stage of the work they can be carried to the possible,
right but not [always] to the left.
premises are
orders
II
For example, if the operators of two and 2 JI 2 z we can unite them in either of the two ÎI*
tf
and
shall
often be
. . .
room
"oth.
usually obtain different conclusions accordingly. There will for skill in choosing the most suitable arrangement.
The next
step consists in multiplying the whole Boolian

itself
part, by the modification of

of
produced by substituting for the index

to the left of
it
any n any other index standing
in the Quantifier.
Thus,
for
Silly/,-,-
[Or, for
2JI 2JI
tf *>(.r,
y),
we can write
"6th.
2 JI//^
tf (
<p(x,
y) x ^(.r, x)}]
The next
step consists in the re-manipulation of the Boolian

in
adding to any part any term we like; 2d, from any part any factor we like, and 3d, in observing that
part, consisting, 1st, in
dropping
xx
so that
"7th.
= =
/,
x+x
(x
v,
xxy +
+ x + y)z
and S
in the Quantifier
whose indices no longer appear
in
the Boolian are dropped.

"The
fifth
step will, in practice, be combined with part of the sixth
and seventh.
Thus, from 2
!! ,-/,-/
we
shall at
once proceed to 2
/ lt if
we
like."
We may
<pxi
say, in general, that the procedures
which are valid

x
in this
calculus are those which can be performed
by treating 2 x v?.r
tpx\
<^.r
as a sum,
<^.r
<.r
<px
and
Il x
<px
as
product,
x,
2 JI,,^(,r,
y) as a
sum, for the various values of
of products, each for
100
the various values of y, and so on.
Thus
this calculus
may
be derived from
the calculus of propositions.
of the principles of the system,
But Peirce does not carry out any proofs and he notes that this method of proof
"It
is to be remarked that SjX,- and would be theoretically unsound. 150 IliXi are only similar to a sum and a product; they are not strictly of that
nature, because the individuals of the universe
may
be
innumerable."
Another way
of saying the
calculus of propositions
same thing would be this: The laws of the extend to 2iX and Ito, because the extension cannot
-
of these laws to aggregates in general,

cal analogies of
sum and product

is
by the method which the mathemati suggest, would require the principle of
mathematical induction, which

gate
is
not sufficient for proof in case the aggre
infinite.
The whole
that
of the calculus of relatives
may
be derived from this calculus
of propositional functions
is,
by the methods which have been exemplified

relation, 6, as 2*2/(&){/(7
of",
:
by representing any
II
J),
relations, such as
"converse
"relative-product,"
and defining the etc., which dis
tinguish the calculus, as
We
of
of the elementary relatives. need not enter into the detail of this matter, since Sections II and III
will
and 2 functions
Chapter IV
modification of Peirce
develop the calculus of propositional functions by a s method, while Section IV of that chapter will show
will indicate the
how
the calculus of classes can be derived from this calculus of propositional
functions, Section
tions
manner
s
in
which the calculus

will suggest
of rela
may
be similarly derived, and Section
VI
how, by a
further important modification of Peirce

logic of
It
method, a theoretically adequate
mathematics
may
be obtained.
remains to consider briefly Peirce s studies toward the derivation of other mathematical relations, operations, and systems from symbolic logic.
The most important paper,

Mathematics".
151
in this connection,
is
"Upon
the Logic of
Certain portions of the paper,

Logic",
Boole
Calculus of
and
of the
an Improvement in monograph, "Description of a Nota

"On
tion for the Logic of
Relatives",
are also of interest.

is
The
+
,
first-mentioned of these
etc., of arithmetic
concerned to show
how
the relations
cal
can be defined in terms of the corresponding logi =, relations, and the properties of arithmetical relations deduced from
logical analogues.
152
theorems concerning their

"Imagine
150
151
...
a particular case under Boole
calculus, in
which the
Alg. Log. 1885, p. 195. Proc. Amer. Acad., vn, 402-12. Loc.
tit.,
152
pp. 410-11.

letters are
101
no longer terms of
first
first
intention, but terms of second intention,
and
that of a special kind.
... Let
the letters
relate exclusively to
the extension of
things
intensions.
Let the differences of the characters of

let
the letters signify only the differ In other words, the only logical comprehension which the letters considered as terms will have is the greater
ences of classes as wider or narrower.
or less divisibility of the class.
and events be disregarded, and
Thus, n in another case of Boole

States
;
calculus
might, for
example, denote
all
it
New England
make
what
but in the case now
supposed,
neglected,
the characters which
these states
what they
are, being
would
signify only
essentially belongs to a class
which
has the same relation to higher and lower classes which the class of
New
England States has,

11
that
is,
a collection of
six.
if
In this case, the sign of identity will receive a special meaning. For, denotes what essentially belongs to a class of the rank of sides of a
,
cube
State
then [the
is
logical]
m=
will
imply, not that every
New England
the side of a cube, and conversely, but that whatever essentially belongs to a class of the numerical rank of New England States essentially
belongs to a class of the rank of sides of a cube

of this particular sort
If a, b,
c,
and conversely.
Identity
may
be termed equality.
..."
etc. represent
thus the number of the classes,
a,
b,
c,
etc.,
then the arithmetical relations can be defined as logical relations. The logical relation a + 6, already defined, will represent arithmetical addition:
And from
will follow
is
the fact that the logical that the arithmetical
+
is
is
commutative and
associative,
it
so also.
Arithmetical multiplication
153
more
a
difficult to deal
with but
may
be defined as follows:
b are events
b represents
an event when a and
only a
is
if
these events
is
are independent of each other, in

logical product].
which case a
X
3,
[where a b
the
is
By
two
the events being independent
meant that
B\,
it
possible to take
series of terms, A\,
A A
2,
,
etc.,
and
B B
2,
3,
etc.,
such that the following conditions are

individual or class, not nothing;
,
satisfied.
, ,
(Here x denotes any
A m A n Bm B n
(.4
two
series of terms,
and 2 A, 2 B, 2
B)
logical
any members sums of some
of the
of the
A n s,
the
B n s,
and the (A n
BJ
respectively.)
Condition
1.
2. 3.
4.
153
No A m is A n No m is n x = 2 (.4 B) a = 2 A
Loc.
cit.,
p. 403.
102

5.
Condition
= 2B
is
6.
Some A m
Bn
of the
This definition
a b
is
somewhat involved: the crux

with a
matter
is
that
will, in the case described,
have as many members as there are combina
tions of a
member
will
of a
member
of
b.
Where the members
of a are
distinct (condition 1)
and the members
of 6 are distinct (condition 2), these
combinations
It is
be of the same multitude as the arithmetical a
b.
worthy
of
remark
that, in respect
both to addition and to multi
plication, Peirce has here hit
of
upon the same fundamental ideas by means which arithmetical relations are defined in Principia Mathematical
"second
The
intention"
of a class
term
is,
in Principia,
Nc
b,
in
Peirce
sum"
discussion, corresponds to
what
is
there called the

is
"arithmetical
of
two
logical classes,
s
and a
b to
what
called the "arithmetical

all
product".
But Peirce
discussion does not
meet
the difficulties
it
that
could hardly be expected in a short paper. In particular, define the arithmetical sum in case the classes summed have
does not
in
members
common, and
class,
it
it
though
does not indicate the manner of defining the number of a does suggest exactly the mode of attack adopted in Prin
cipia,
namely, that number be considered as a property of cardinally similar

for the derivation of the laws of various numerical
classes taken in extension.
The method suggested
algebras from those of the logic of relatives is more comprehensive, though here it is only the order of the systems which is derived from the order of
the logic of relatives; there
is
no attempt to define the number or multitude
of a class in terms of logical relations. 155
We
individual in which
are here to take a closed system of elementary relatives, every is either a T or a P and none is both.
Let
c
s
= (T T) = (P:P)
:
p = (P:T)
t
= (T :P)
:
Suppose T here represent an Individual teacher, and P an individual pupil the system will then be comparable to a school in which every person is
either teacher or pupil,
pupil.
354 155
and none
is
The
relative
term, c, will
both and every teacher teaches every then be defined as the relation of one
See Vol. n, Section A. "Description of a Notation, pp. 359 jf.

teacher to another, that
"
103
P), the rela
is,
colleague".
is,
Similarly, s
is
(P
"
tion of one pupil to another, that

is
"schoolmate".
The
relative
term, p,
any pupil to any teacher, that is, pupil". And the relative term, t, is (T P), the relation of any teacher to any pupil, that is, "teacher". Thus from the two non-relative terms, T and P, are generated the four elementary relatives, c, s, t, and p.
(P
:
:
T), the relation of
The
properties of this system will be clearer
if
we venture upon
certain
explanations of the properties of elementary relatives which Peirce does not give and to the form of which he might object. For any such relative
(7
:
J),
(1)
where the /
and the J
are distinct,
we
shall
have three laws:
(I:J)\J =
1
(7
:
Whatever has the
J) relation to a
J must
be an 7: whoever has the
teacher-pupil relation to a pupil

(2)
must be a
teacher.
(7:./)l/
=
relation to a teacher (where teachers
Whatever has the teacher-pupil
and
pupils are distinct) does not exist.

(3)
(I
:J)\(II:K) =
[(I
J)\H]
K
(7
:
The
relation of those
:
which have the
J) relation to those which have

:
the (77
K)
relation
is
the relation of those-which-have-the-(7
J)-relation-
to-an-77 to a
K.
law which
is
It is this third
the source of the important properties of
the system.
t\p
For example:
= (T :P)|(P
of
fit
T)
s
[(T
P) |P]
T = (T
s
1
:
The teachers
or his
any person
pupils are that person
colleagues.
(Our
illustration, to
the system, requires that one
may
be his
own
colleague
own
schoolmate.)
c\c
= (T
= (T
T)\(T
s
T)
[(T
T)\ T]
s
T = (T
T)
The
colleagues of one
t
t
:
colleagues are one

:
|
colleagues.
T]
:
P) (T
P)
[(T
P)
P = P =
(0
P)
There are no teachers of teachers in the system.
= (P
T)
(P
P)
[(P
T) \P]
(0
P)
There are no pupils of anyone s schoolmates in the system. The results may be summarized in the following multiplication table,
in
which the multipliers are
in the
column at the
right
and the multiplicands
104

156
:
at the top (relative multiplication not being commutative)

t
The symmetry
of the table should be noted.

it
The reader may
easily in
terpret the sixteen propositions which
gives.
To
the algebra thus constituted

letters.
may
be added modifiers of the terms,

"French",
symbolized by small roman

of the system in case
If f is
will
be a modifier
French teachers have only French pupils, and vice versa. Such modifiers are "scalars" of the system, and any expression of the form
a c +
bt+
p+d
s
a, b, c,
where
c,
t,
p,
and
s are
the relatives, as above, and
d are scalars,
Peirce calls a
is
"logical
quaternion".
The product
of a scalar with a
term
commutative,
bt = tb
since this relation
is
that of the non-relative logical product.
Inasmuch as
any (dyadic,
triads, etc.)
triadic, etc.) relative is resolvable into a logical

it is
sum
of (pairs,
elementary relatives,
plain that any general relative what
ever
is
If
sum of logical quaternions. we consider a system of relatives, each of which

resolvable into a
is
of the
form
az +
where
i,
bj + ck + dl+
...
j, k,
I,
etc. are
each of the form
mu + nv + oiv+
we
...
where m, n, o, etc. are scalars, and u, v, w, etc. are elementary relatives, shall have a more complex algebra. By such processes of complication, multiple algebras of various types can be generated. In fact, Peirce says: 157 can assert, upon reasonable inductive evidence, that all such
"I
[linear
on the principles of the present notation in the same way as those given above. In other words, all such algebras are complications and modifications of the algebra of (156) [for which the multiplication table has been It is very likely that this given].
155
associative] algebras can be interpreted
Ibid., p. 361. Ibid., pp.
157
363-64.

is
105
true of
all
algebras whatever.
The algebra
of (156),
which
is
of such a
pure algebra and our logical nota tion, has been shown by Professor [Benjamin] Peirce to be the algebra of
in reference to
fundamental character
Hamilton
s quaternions."
Peirce gives the form of the four fundamental factors of quaternions
and
of scalars, tensors, vectors, etc., with their logical interpretations as relative
terms with modifiers such as were described above.
One more item

of probabilities.
of importance
is
is
Peirce
modification of Boole
calculus
This
in
set forth
s
an Improvement
identity of
Boole
with extreme brevity in the paper, "On Calculus of Logic". 158 For the expression of
the relations involved,
we
shall
two
classes in extension
need to distinguish the logical relation of from the relation of numerical equality.
We
of
may, then, express the

b,
fact that the class a has the

all
fe
as the class
or
all
are
s,
by a
b,
and the
of
fact that the

of
b,
same membership number

by a =
b.
members of a is the same Also we must remember the

pressed
as the
number
members
distinction between the logical relations ex

relations
by a +
b,
ab, a
b,
\-b,
and the corresponding arithmetical

and a
expressed by a
"Let
b,
b.
Peirce says:
159
every expression for a class have a second meaning, which is its meaning in a [numerical] equation. Namely, let it denote the proportion
of individuals of that class to be
in the long run.
"Then
found among
all
the individuals examined
we have
If
"
as a class,
Let b a denote the frequency of the 6 s among the a s. Then considered if a and b are events b a denotes the fact that if a happens b happens.
X
"
ba
ab
It will
ties of
be convenient to set
.
down some obvious and fundamental proper

a
the function b a
Xb =
a,
X
1
ab
C) a
<p(b
Ca ) b) a
(1
= =
<p(b,
ba
158 159
Proc.
Amer. Acad., vn, 255 ff.

255-56.
Ibid., pp.
106

I a ---7- X
&
0(l_a)
The
abilities
(1)
chief points of difference
between
this modified calculus of
prob
and the
original calculus of Boole are as follows:

p, q, etc. for the "probability of a, of b,
etc.",
Where Boole puts

from the
in passing
logical to the arithmetical interpretation of his
equa
simply changes the relations involved from logical relations to the corresponding arithmetical relations, in accordance with the foregoing,
tions, Peirce
and
lets
the terms
a, b, etc.
stand for the frequency of the a
s,
s,
etc.
system under discussion. Boole has no symbol for the frequency of the a s amongst the 6 s, which Peirce represents by a b As a result, Boole is led to treat the
in the
(2)
.
probabilities of all unconditioned simple events as independent
a pro
cedure which involved him in

(3)
many
difficulties
and some
errors.
Peirce has a complete set of
four
logical
operations,
and four
analogous operations of arithmetic.
This greatly facilitates the passage
from the purely
logical
expression of relations of classes or events to the
arithmetical expression of their relative frequencies or probabilities.
Probably there is no one piece of work which would so immediately reward an investigator in symbolic logic as would the development of this calculus of probabilities in such shape as to make it simple and practicable.
Except
for a
monograph by Poretsky and the
studies of H. MacColl, 160
the subject has lain almost untouched since Peirce wrote the above in 1867.
Peirce s contribution to our subject is the most considerable of any up to his time, with the doubtful exception of Boole s. His papers, however, are brief to the point of obscurity: results are given summarily with little
or no explanation
the most valuable of
and only infrequent demonstrations. As a consequence, them make tremendously tough reading, and they
serves. 161
have never received one-tenth the attention which their importance de If Peirce had been given to the pleasantly discursive style of
clearly accurate
De Morgan, or the detailed and work on symbolic logic would fill

160
manner
of Schroder, his
several volumes.
an unusually
logic.
161
Since the above was written, a paper by Couturat, posthumously published, gives clear presentation of the fundamental laws of probability in terms of symbolic
See Bibl.
find our report of Peirce requested to consult the original papers.
Any who
work unduly
difficult or
obscure are earnestly
107
VIII.
DEVELOPMENTS SINCE PEIRCE
of Peirce
Contributions to symbolic logic which have been made since the time need be mentioned only briefly. These are all accessible and in a
sufficiently close to current notation to be readily intelligible.
form
Also,
they have not been superseded, as have most of the papers so far discussed; consequently they are worth studying quite apart from any relation to later work. And finally, much of the content and method of the most *
important of them
later chapters, or
will
is
is
substantially the
same with what
will
is
be set forth in
there set forth
such that
its
connection with what
be pointed out.
But
for the sake of continuity
and perspective, a
summary account may be given of these recent developments. We should first mention three important pieces of work contemporary
with Peirce
s later treatises. 162
Robert Grassmann had included
in his encyclopedic Wissenschaftslehre

3
a book entitled Die Be griffslehre oder Logik,

Begriffen,
(2)
containing
(3)
(1)
Lehre von den
Lehre von den Urtheilen, and

is
Lehre von den Schlussen.
The
Begriffslehre
this
the second book of Die Formerdehre oder Mathematik,
and as
would
indicate, the
development of
s
logic
is
entirely mathematical.
An important
character of Grassmann
procedure
is
the derivation of the
laws of classes, or Begriffe, as he insists upon calling them, from the laws governing individuals. For example, the laws a + a = a and a- a = a,
where a
where
it,
e,
is
a class, are derived from the laws e + e
e,
e-e
e,
e\-e<i
0,
e\, e 2
represent individuals.
This method has
much
to
commend
but
it
has one serious defect
the supposition that a class can be treated
as
an aggregate of individuals and the laws of such aggregates proved As Peirce has observed, this method generally by mathematical induction.
breaks
down when
the
number
of individuals
is
may
be
infinite.
Another
difference
between Grassmann and others
the use throughout of the

of extension,
language of intension.
But the method and the laws are those

diagrammatic
areas.
and
in the later treatise, there are
illustrations in
which
in
"concepts"
162
are represented
by
Although somewhat incomplete,
Alexander MacFarlane, Principles of the Algebra of Logic, 1879, gives a masterly presentation of the Boolean algebra. There are some notable extensions of Boole s methods and one or two emendations, but in general it is the calculus of Boole unchanged. MacFarlane s paper "On a Calculus of Relationship" (Proc. Roy. Soc. Edin., x, 224-32) re
sembles somewhat, in
its method, Peirce s treatment of "elementary relatives". But the development of it seems never to have been continued. 163 There are two editions, 1872 and 1890. The later is much expanded, but the plan and general character is the same.
108

s
other respects Grassmann
calculus
is
not notably different from others
which follow the Boolean tradition.
Hugh MacColl
164
first
two papers on
paper
"On
"The
Calculus
of
Equivalent
165
in
printed 1878-80, present a calculus of propositions which has essentially the In others words, it is properties of Peirce s, without II and 2 operators.
Statements",
and
his first
Symbolical
Reasoning",
it
a calculus of propositions, like the Two-Valued Algebra of Logic as we know today. And the date of these papers indicates that their content was
s
arrived at independently of Peirce
studies
which deal with
this touic.
In fact, MacColl writes, in 1878, that he has not seen Boole. 166
The
calculus
167
set
forth
in
MacColl
book, Symbolic
Logic
and
its
Application*,
is
of
an entirely different character.

traditional truth values,
"certain",
Here the funda
mental symbols represent propositional functions rather than propositions;
and instead
have
tain
77,
of the
two
"true"
and
"false",
we
"true",
"false",
"impossible"
and
"variable"
(not cer
and not impossible). and procedures

of
8 respectively.
These are indicated by the exponents r, i, e, The result is a highly complex system, the fundamental
ideas
which suggest somewhat the system
of Strict
Implication to be set forth in Chapter V.
The
calculus of Mrs. Ladd-Franklin, set forth in the paper

Logic"
"On
the
Algebra of
in the
Johns Hopkins studies,
168
differs
.
from the other

b are
systems based on Boole by the use of the copula v or "Some a is classes, a v b represents is-partly
"a
Where a and
6",
6",
av
"
b,
represents
a is-wholly-not-6
b to
"
",
or
0.
No
is
".
and its negative, Thus a v 6 is equiva

a
lent to a b =h 0,
and a v
a b
These two relations can, between
them, express any assertable relation in the algebra,
cb
will
be a v-6,
and a = b is represented by the pair, (a v-b)(-a v b). For propositions, a v b denotes that a and b are consistent a does not imply that b is false and b does not imply that a is false. And a v6 symbolizes and b are
"a
inconsistent"
if
is
true, b
is false;
if
6 is true, a is false.
The
is
use of the
terms
"consistent"
and
"inconsistent"
in this connection
possibly mis
leading:
164
any two true propositions or any two

ix,
false propositions are con177-86.
166
(1) Proc. London Math. Soc., Mind, v (1880), 45-60. Proc. London Math. Soc., ix,
9-20;
(2) ibid., ix,
178.
Longmans, 1906. 38 The same volume contains an interesting and somewhat complicated system by O. H. Mitchell. Peirce acknowledged this paper as having shown us how to express uni versal and particular propositions as n and S functions. B. I. Oilman s study of relative number, also in that volume, belongs to the number of those papers which are important
in connecting symbolic logic with the theory of probabilities.
167

sistent in this sense,
109
and any two propositions one

This
is
"it is
of
which
is
true
and
the other false are inconsistent.

"consistent"
not quite the usual meaning of
and
"
inconsistent
"
related to
what
is
usually
these terms exactly as the

is
material implication a c b
a".
is
related to
meant by what
x
usually
meant by
"b
can be inferred from
That a given
class, x, is
empty, or a given proposition, x,
is false,
0,
may
be expressed by x v
1.
oo,
where
co
is
"everything"
represented by
That a
class, y,
has members,
is
most systems symbolized by y v oo.

in
where y is a proposition, since Mrs. system does not contain the assumption which is true for propositions but not for classes, usually x =(= 0, then expressed, x = 1, and if x =|= 1, then x = x v oo may be abbreviated to xv,
is
This last
of doubtful interpretation
s
Ladd-Franklin
"If
0".
a b v
v and T/VGO to yv c dv co to c d v understood that if one term of a relation v or v

co
to a b
etc., since it is
always
is
missing, the missing
term
tion:
is
This convention leads to a very pretty and convenient opera v or v may be moved past its terms in either direction. Thus,
oo
.
(a
v 6) =
(a
by) =
va
vx
6)
and
(xvy) = (xyv) =
y)
But the forms (va6) and (vxy) are never used, being redundant both logically and psychologically.
Mrs. Ladd-Franklin
logic particularly well
:
system symbolizes the relations of the traditional
All a
is b.
-b,
or or
or or
a -6 a b a b
No
is b.
avb, avb,
6.
Some Some
Thus v
a a
is b.
is
not
a v -6,
a -6
characterizes a universal,
v a particular proposition.
"
And any
then
all
pair of contradictories will differ from one another simply
by the difference
between v and v
a
"
The
syllogism,
If all
is
and
all b is c,
is c,
will be represented
by v -b) (b v -c) v
(a
(a
v c)
where v, or v, within the parentheses

pression
is interpreted for classes, and v between the parentheses takes the propositional interpretation. This ex
may
also be read,
"
All a
is
and
is
all b is c
is
inconsistent with
the negative (contradictory) of

(a
Some
not
".
It is
equivalent to
-b) (b
v -c) (a v -c) v
110
A
three propositions,

All a
is
"The
All b
is c,
and
Some
is
not
are inconsistent
three syllogisms:
they cannot
all
three be
true".
This expresses at once
(1)
"If
(a
v-6)(6 v-c) v
all
(a
v-c)
all
is
and
all 6 is c,
then
(a
is
c";
(2)
"If
v -b) (a v -c) v
c,
(b
v -c)
not
c";
all
is
and some a
(3)
is
not
(b
then some
b is
v -c) (a v -c) v
c,
(a
v -6)
is
"
If all b is c
and some a
is
not
then some a
not b
".
Also, this
method
gives a perfectly general formula for the syllogism

(a
v -b) (b v c) (a v c) v
where the order of the parentheses, and their position relative to the sign v which stands outside the parentheses, may be altered at will. This single
rule covers all the
modes and
169
figures of the syllogism, except the
illicit
particular conclusion drawn from universal premises.

this
We
shall revert to
Chapter III. The copulas v and v- have several advantages over their equivalents, = and =j= 0, or c and its negative: (1) v and v are symmetrical rela tions whose terms can always be interchanged; (2) the operation, mentioned above, of moving v and v with respect to their terms, accomplishes trans
formations which are
the copula;
(3) for
matter
in
less
simply performed w ith other modes of expressing

r
various reasons,
it is
psychologically simpler and
more
terms
natural to think of logical relations in terms of

of
v and v than
in
=
,
and
=|=
0.
But v and v have one disadvantage

For better or
.
as against
4=
and c
they do not so readily suggest their mathematical analogues in

for
other algebras.
worse,
symbolic logicians have not

that of Ernst Schroder.
generally adopted v and v
Of the major contributions since Peirce, the
first is
In his Operationskreis des Logikkalkuls (1877), Schroder pointed out that the logical relations expressed in Boole s calculus by subtraction and divi
sion were all otherwise expressible,
of
meaning
has, first
as Peirce had already noted. The + given by Boole is abandoned in favor of that which it now introduced by Jevons. And the "law of duality", which con
,
nects theorems which involve the relation +
or +
and
1,
with corresponding
0, is
theorems
169
in
terms of the logical product

ff.
x, or
x and
emphasized.
See below, pp. 188

(This parallelism of formulae had been noted by Peirce, in his but not emphasized or made use of.)
first
111
paper,
we know it today. This perfected and elaborated in Vorlesungen uber die Algebra der Logik (1890-95). Volume I of this work covers the algebra of classes;
resulting system
is
The
the algebra of logic as
system
is
Volume
II the algebra of propositions;
and Volume III
is
devoted to the
calculus of relations.
The algebra
is
of classes, or as
we
shall call
it,
the Boole-Schroder algebra,
170 the system developed in the next chapter. rated the theory of functions, but in all essential
We
have somewhat elabo
respects,
we
give the algebra
There are two differences of some importance between Schroder s procedure and the one we have adopted. Schroder s assumptions are in terms of the relation of subsumption, c instead of the
it
as
appears
in Schroder.
relations of logical product
and =; which appear in our postulates. And, second, Schroder gives and discusses the various methods of his predecessors,
as well as those characteristically his own.
The calculus of propositions (Aussagenkalkid) is the extension of the Boole-Schroder algebra to propositions by a method which differs little from that adopted in Chapter IV, Section I, of this book. The
Peirce
s
discussion of relations
is
based upon the work of Peirce.
But
methods are much more

is
the scope of the calculus
precisely formulated by Schroder, and much extended. We summarize the funda
mental propositions which Schroder gives for the sake of comparison both with Peirce and with the procedure we shall adopt in Sections II and III
of
Chapter IV.
1)
A, B, C, D,
symbolize
"elements"
or individuals. 171
These
are distinct from one another
and from
0.
symbolizes the universe of individuals or the universe of discourse of

first
the
order.
j, k,
/,
3)
2,
m,
n, p, q
represent any one of the elements A, B, C, D, ...
of
4)
70
Sri
For an excellent summary by Schroder, see Abriss der Algebra der Logik ; ed. Dr. Eugen Miillor, 1909-10. Parts i and n, covering Vols. i and n of Schroder s Vorlesungen, have so far appeared.
171
The
in, 3-42.
Many
propositions here noted will be found in Vorksungen uber die Algebra der Logik, others, and much discussion of theory, have been omitted.
112
A
5) i 6)
(i
:

i
j represents
any two elements,
and j,
of I 1 in a determined order.
j)
(i
i),
(i 4= j)
*j
for every
7)
and
j.
i:j*0
I
1
Pairs of elements of
may
be arranged in a
"block":
A:A, A:B, A:C, A:D,

B:A, B:B, B:C, B:D,
C:A, C:B, C
:
...
...
...
C,
D,
:A,
D:B, D:C, D:D,
These are the
"individual
binary
:
relatives".
(A
+ (B + (C
A) + (A A) + (B A) + (C
:
B) + (A
B) + (B
:
C) + C) + C) +
.
.
B) + (C
represents the universe of binary relatives.

10)
I
2
S,-S,- (i
j)
= S -S,t
(i
:j)
= S
iy
(i
j)
9)
and
10)
may
1
be summarized in a simpler notation:
Ziji :j =
A :A+A :B + A :C+
+
...
B :A
C :A
+
+
B :B + B C+
:
11)
C :B
C+
...
+
12)
13)
i
........
"individual
:j
h will symbolize an
S^SyS,(i
:
ternary
relative".
A)
= 2 A t:j:A
Various types of ternary relatives are

14)
A A
:
A,
B A
:
A,
A B
:
A,
A A
:
B,
A B C
:
:
It is
obvious that we
may
similarly define individual relatives of the

order.
fourth, fifth,
...
or
any thinkable
113
The
general form of a binary relative,
a, is
= Sy an
is 1
(i
j) j) pairs in
where a i;
is
a coefficient whose value

j,
for those
(i
which
has
the relation a to
1
and
is
otherwise
0.
==
Ziii-.j
the null class of individual binary relatives.
/=
JV
(a b)ij
S -/(i =j)(i:j) = 2i(i:i)

t
172
S,- y
(i=t=
j)(i:j)
(a
aij bij
-a.-/
6) i?
= a, =
&,-/
(-a),-,-
-(a,y)
(a &);y
1
= 2A
a,-*
6 A/
(a t 6),v
n* (a* +
fe*,-)
general laws which govern propositional functions, or Aussagenschemata, such as (ab) iit 2 h a ih b hi U h (a ih + b hj ), II a a i} S a a,-y, etc., are as
, ,
The
follows
Au
case,
symbolizes any statement about u\

in case,
and only
is
Au =
for every
there
"A
at least one
u"
for every
u such that A u == and 2 U A U means A u for some

"
U U A U will have the value 1 in u; 2 U A U will have the value 1 if 1. That is to say, U U A U means
u".
a)
j8)
U u A u cA c2 u A u K U A U = A U V A U1
v V
-[2uA u ]c-A v c-[U u A u
2UA U
0,
= A
+ 2UA U
(The subscript
7)
d)
u, in
a and
-/i u
,
represents
any value
]
of the variable u.)
-[n u ,4 u
= 2U
If ^4 M is
independent of
)
u,
then
-[s^ = n u -A* U U A U = A, and S u .4 =

tt
A.
e)
r)
U u (AcB u = (Acn u B u ), n, or n M n,(^ M c5 =

r
i;
U u (A u cB) =
(2 u ^l M
(2 u A u
cB)
cn 5
y
v)
77)
2 u (A u cB) = (n u A u cB), S
I
f
2 u (AcB u ) =
(H u A u
i),
(Ac2 u B u
= =
0)
or 2 u 2 v (A u
(.-i
cB
v)
= =
c2,B
9)
nw
=
=
i)
(n u .4
n u (,i u =
S B (.4
==
0)
1)
;
= =
(a
|
(z M ,i M
(S
tt
o)
1)
172
\2u(A u 0) We write I_ where

-a for a;
<-a
= (n u /l u =
Schroder has
0),
1
;
.4
where he has
6) for (a; 6); (a
6)
for (a j 6);
for a.
114
A
H u (A u cB u )c]
\
(U U A U c
UUBU
K)
(S.4
tt
c ZU
(The reader should note that

Principia Mathematica,
X)
(x).<px
U U (A U ^B U
o\f/x.)
is
"formal
implication",
in
2 UBU = SM
AB
= S
U)
A + U U B U = U U (A + B u
Au B
V)
//)
(2 U A U )(2 V B V )
yl
ttf
II U
A U + U B = n,
V
V
(A U +
)
v)
nu
= nu A Bu
A
,
+ 2UBU
,
= ? U (A
Bu
(U U A U )(U V B V )
= UU
A U B = UU A U BU
V
o)
2 u n^4 M
Ii v
Su
Au
From
these fundamental propositions, the whole theory of relations
is
developed.
Though Schroder carries this

s
much
further than Peirce, the
general outlines are those of Peirce

esting of the
in the
new items
of
Schroder
calculus. Perhaps the most inter treatment are the use of "matrices"
relatives,
form of the two-dimensional array of individual binary

of the calculus of relatives to
and the application

"chains
",
Dedekind
theory of
as contained in
Was
sind und was sollen die Zahlen.
Notable contributions to the Boole-Schroder algebra were made by
Anton Poretsky
logiques
in his three papers, Sept lois fondamentales de la theorie
des egalites logiques (1899), Quelques lois ulterieures de la theorie des egalites
(1901),
and Theorie des
non-egalites logiques
(1904).
(With
will
his
earlier works, published in Russian, 1881-87,
we
are not familiar.)

of
Poret-
sky
Law
of
Forms,
Law
of Consequences,
and Law
Causes
be
II. As Couturat notes, Schroder had been influenced overmuch by the analogies of the algebra of logic to other algebras, and these papers by Poretsky outline an entirely different procedure which, though based on the same fundamental principles, is somewhat more
given in Chapter
"natural" to logic. Poretsky s method is the perfection of that type of procedure adopted by Jevons and characteristic of the use of the Venn
diagrams.
The work
of Frege,
though
its
intrinsically important, has its historical
interest largely
through
influence
upon Mr. Bertrand
Russell.
Although
the Begrifsschrift (1879) and the Grundlagen der Arithmetik (1884) both

precede Schroder
there; and
laire is
s
115
hardly more than mentioned his influence upon Peano and other contributors to the Formusurprisingly small when one considers how closely their task is re
Vorlesungen, Frege
is
lated to his.
Frege
is
but, in thorough
German
concerned explicitly with the logic of mathematics fashion, he pursues his analyses more and more
a development of arithmetic of unprecedented but a more or less complete treatise of the logico-metaphysical problems rigor
deeply until
we have not only
concerning the nature of number, the objectivity of concepts, the relations

of concepts, symbols,
his
and
objects,
and many other
subtleties.
In a sense,
fundamental problem is the Kantian one of the nature of the judgments involved in mathematical demonstration. Judgments are analytic, de pending solely upon logical principles and definitions, or they are synthetic.
His
thesis, that
mathematics can be developed wholly by analytic judg

logical,
is
ments from premises which are purely

Russell
s
likewise the thesis of
Principles of Mathematics.
metik, like
Grundgesetee der ArithFrege Principia Mathematica, undertakes to establish this thesis for
s
And
arithmetic
by producing the required development.

s
Besides the precision of notation and analysis, Frege
work
is
important
is
as being the first in which the nature of rigorous demonstration

ficiently understood.
suf
His proofs proceed almost exclusively by substitu
tion for variables of values of those variables,
and the substitution

is
of defined
it is notation, Frege almost diagrammatic, occupying unnecessary space and carrying the eye here and there in a way which militates against easy understanding. It is probably this forbidding character of his medium, combined with the
equivalents.
it
must be admitted
against him:
unprecedented demands upon the reader s logical subtlety, which accounts for the neglect which his writings so long suffered. But for this, the revival
of logistic proper
Frege
might have taken place ten years earlier, and dated from Grundlagen rather than Peano s Formulaire.
publication, beginning in 1894, of
The
Peano
Formulaire de Mathelogic.
matiques marks a
new epoch
in the history of
symbolic
Heretofore,
the investigation had generally been carried on from an interest in exact

logic
and
its possibilities, until,

it
as Schroder remarks,
we had an elaborated
his collaborators, the
instrument and nothing for

situation
of
:
to do.
With Peano and
is reversed symbolic logic is investigated only as the instrument mathematical proof. As Peano puts it: 173 The laws of logic contained in what follows have generally been found
"
173
Formulaire,
(1901), 9.
116
in the

rules, the
by formulating,
in
form of
deductions which one comes upon
mathematical
demonstrations."
The immediate
rate
result of this altered point of
view
is
new
logic,
no
less elaborate than the old
destined, in fact, to
become much more elabo
but with
its
elaboration determined not from abstract logical con
siderations or
of
in
by any mathematical prettiness, but solely by the criterion De Morgan had said that algebraists and geometers live application.
"a
seems to have required the mathe matical intent to complete the rescue of logic from its traditional inanities. The outstanding differences of the logic of Peano from that of Peirce
higher realm of
syllogism":
it
and Schroder are somewhat as

(1)
follows:
Careful enunciation of definitions and postulates, and of possible

postulates,
alternative
marking an increased emphasis upon rigorous

of a
deductive procedure in the development of the system.

(2)
The prominence
new
relation,
e,
the relation of a
member
of a
class to the class.

(3)
The prominence
implication"
of the idea of a prepositional function

"
and
of
"formal
and
formal
equivalence",
as against
"material
implication"
and
"material equivalence".
"existence"
(4)
Recognition of the importance of
and
of the properties
of classes,
(5)
members of classes, and so on, with reference to their "existence". The properties of relations in general are not studied, and "relative
addition"
in
does not appear at all, but various special relations, prominent mathematics, are treated of.
The disappearance
gain.
(6)
its
of the idea of relation in general is a real loss, not a
The
increasing use of substitution (for a variable of
some value
in
range) as the operation which gives proof.
We
The
here recognize those characteristics of symbolic logic which have
since been increasingly emphasized.
publication of Principia Mathematica would seem to have deter

of further investigation to follow that general direction
mined the direction

indicated by the
The Principia is con cerned with the same topics and from the same point of view. But we see here a recognition of difficulties not suggested in the Formulaire, a deeper
of Frege
work
and the Formulaire.
and more lengthy analysis of concepts and a corresponding complexity of procedure. There is also more attention to the details of a rigorous
method
174
of proof.
All these belong also to the Logica
Mathematica of C. Burali Forti (Milan, 1894).
117
is
The method by which the mathematical logic of Principia Mathematica developed will be discussed, so far as we can discuss it, in the concluding
We shall be especially concerned to point out the sometimes lost sight of, between it and the older logic of Peirce connection, and Schroder. And the use of this logic as an instrument of mathematical
section of Chapter IV.
analysis will be a topic in the concluding chapter.
CHAPTER
THE
I.
II
CLASSIC,
OR BOOLE-SCHRODER, ALGEBRA OF LOGIC

THE POSTULATES AND
GENERAL CHARACTER OF THE ALGEBRA.

THEIR INTERPRETATION
The algebra of logic, in its generally accepted form, is hardly old enough warrant the epithet "classic". It was founded by Boole and given its to present form by Schroder, who incorporated into it certain emendations
which Jevons had proposed and certain additions particularly the relation or implies" which Peirce had made to Boole s system. contained
"
"is
in"
It is
sound judgment that the result is still an algebra, simpler yet more powerful than Boole s calculus. Jevons, in simplifying Boole s system, destroyed its mathematical form; Peirce, retaining the
due to Schroder
mathematical form, complicated instead of simplifying the original calculus. Since the publication of Schroder s Vorlesungen uber die Algebra der Logik
certain additions
of
and improved methods have been offered, the most notable which are contained in the studies of Poretsky and in Whitehead s Uni
versal Algebra. 1
But
this
if
the term
it
"classic"
is
inappropriate at present,
still
we may
out,
all
venture to use
by way
of prophecy.
As Whitehead has pointed

"algebra",
system
is
a distinct species of the genus

its
differing
from
other algebras so far discovered by
non-numerical character.
It is
certainly the simplest mathematical system with any wide range of useful
applications,
and there are indications that

of
it
will serve as the
parent stem
from which other calculuses

eral
such have appeared.
an important type will grow. Already sev The term "classic" will also serve to distinguish
the Boole-Schroder Algebra from various other calculuses of logic.
Some
of these, like the system of Mrs. Ladd-Franklin, differ through the use of other relations than +
1
,
and =
and are otherwise equivalent
versal Algebra,
For Poretsky s Bk. n.
See Whitehead s Uni studies, see Bibliography; also p. 114 above. Whitehead introduced a theory of "discriminants" and a treatment
of existential propositions by means of umbral letters. This last, though most ingenious interesting, seems to me rather too complicated for use; and I have not made use of "discriminants preferring to accomplish similar results by a somewhat extended study of
and
",
the coefficients in functions.
118
The Classic, or Boole-Schroder, Algebra of Logic

that
is
119
to say, with a
"
of these systems
may
dictionary" of equivalent expressions, any theorem be translated into a theorem of the Boole-Schroder
Algebra, and vice versa.
they go, but partial.

Mathematica,
are
finally, there are
Others are mathematically equivalent as far as some, like the calculus of classes in Principia logically but not mathematically equivalent. And,
And
systems such as that of Mr. MacColl s Symbolic Logic which are neither mathematically nor logically equivalent. Postulates for the classic algebra have been given by Huntington,
the Abriss}, by Del Re, by Sheffer and by Bernstein. 2 set here adopted represents a modification of Huntington s third
(in
by Schroder
The
set.*
It
has been chosen not so
much
for
economy of assumption
as for
"
natural
ness"
and obviousness.
Postulated:
A
1-1
class
If
K of elements
b are
b.
a, b, c, etc.,
and a
relation
b is
x such that:
in
a and
elements in K, then a x
an element
K, uniquely
determined by a and
1
For any element
a,
a xa
=
6,
a.
1 1
1
For any elements a and
a x
c,
x
x
a.
4
5
For any elements
a, b,
and
a x
(b
c)
(a x 6)
x c.
There
a.
is
a unique element,
0, in
such that a x
for every ele
ment
1
For every element

1-61
If
a,
there
is
an element,
-a, such that

x,
x x-a y xa
0,
then x xa
and 1-62
If
y and y x -a
y,
then y
0.
The element
These
1-7 1-8
and the
relations +
:
and c do not appear
in the above.
may 1 = -0
be defined as follows
Def.
Def.
o+ 6 a cb
It
= -(-ax-6)
is
1-9
equivalent to a x6
a
is
Def.
remains to be proved that -a

it
which
will follow that

b.
is
uniquely determined by a, from unique and that a + b is uniquely determined
by a and
2 3
See Bibl. See "Sets of Independent Postulates for the Algebra of Logic", Trans. Amer. Math. Our set is got by replacing + in Huntington s set by x and Soc., v (1904), 288-309. Thus 1 can be replacing the second half of G, which involves 1, by its analogue with 0. omitted. Postulate J is not strictly necessary. defined, and postulates E and
,
120
A
The
sign of equality in the
above has
if
its
usual mathematical meaning;

<p(x)
j.
e .,
=
j j
is
a relation such that

<p(x)
y and
function in the system, then

interchangeable.
in the system,
It follows
and
<p(y)
an unambiguous are equivalent expressions and

is
from this that
if
$(x)
is
an ambiguous function
terms of
that -a,
x,
is
every determined value of ifr(x), expressible in Suppose, for example, similarly expressible in terms of y.
y,
and x
"negative
of
a",
is
an ambiguous function
"negative
of a.
Still
we may
write
-a to mean, not the function
of
a"
itself,
but to mean some
(any) determined value of that function
any one
of the negatives of a
and
if
-a
b,
then
is
<p(-a)
and
<p(b)
will
be equivalent and interchangeable.
This principle
We
terms,
shall
important in the early theorems which involve negatives. develop the algebra as an abstract mathematical system: the
a, b, c, etc.,
,
and x
and c
may be any entities which have the postulated properties, may be any relations consistent with the postulates.
we
give
But
for the reader s convenience,

all,
two
possible applications:
(1) to
the system of
continuous and discontinuous, regions in a plane, the
null-region included,
and
(2) to
the logic of classes. 4
(1)
For the
first
interpretation, a x b will denote the region
common
to a
and
a
b (their
is
overlapping portion or portions),
and a +
b will
denote that
"Region
which
is
either a or 6 or both,
a c b will represent the proposition,

b
completely contained in region

+
(with or without
remainder)".
1
will
itself,
represent the null-region, contained in every region, and

or the
will
"sum"
{
the plane
of all the regions in the plane.

a, all
For any region
a,
-a
be the plane except
that
is
not-a.
The
postulates will then hold
as follows:
1
1
b,
If
is
a and b are regions in the plane, the region

in the plane.
If
common
to a
is
and
b,
a x
a and 6 do not overlap, then a x b
the null-
region, 0.
1
2
3
For any region
a,
the region
to a
common
to a
and
a,
a x
a, is
itself.
1 1
The
The
c
is
region region
common
and
b is the region
is
common
to b
and
a.
and
1-5
1
the
common region common

common
a,
to a and b x c
the region
common
to a x b
to all three.
The
region
to
any region a and the

is its
null-region, 0,
is 0.
6
4
For every region

Both
there
negative, -a, the region outside or
of these interpretations are
more
fully discussed in the next chapter.

not contained in
1-61
a,
121
and
If
this region
is
such that
common, then the region

and 1-62
in a,
If
-a and any region x have only the null-region in common to x and a is x itself, or x is contained in a;
the region
common
to y
and a
is y,
or y
is
contained
-a, then
and the region common to y and -a is y, or y is contained in must be the null-region which is contained in every region.
That the
definitions, 1-7, 1-8,
and
1-9, hold, will be evident.
(2)
For the second interpretation,

in extension
a, b, c, etc., will
be logical classes, taken
that
is,
b will
mean
and
that a and b are classes composed of
identically the
same members,
a x b will represent the class of those

of b both;
things which are

are either
tion
members
of a
a+
b,
those things which
members of a or members of b or both, that all members of a are also members of b,

is is
a c b will be the proposi or that a is contained in b
(with or without remainder).
the null-class or class of no members;

is
r
and the convention

1 is
required that this class

discourse"
contained in every
class.
the
"universe
of
or the class
w hich contains every
entity
in the system.
of all
For any class a, -a represents the negative of a, or the class The postulates will hold as fol things which are not members of a.
lows
1-1
If
a and b are logical classes, taken in extension, the
members com
mon
in
to a
and
b constitute a logical class.

is
In case a and b have no members
common,
this class
the null-class,
to a
0.
1-2
1
The members common The members common

a.
and a constitute the and

b are the
class a itself.
to a
same as those common to
b
1
and
-4
The members common to a, b, and c, all three, are the same, whether we first find the members common to b and c and then those common to a and this class, or w hether we first find the common members of a and b
r
and then those common to

1
this class
and
c.
The members common

For every
"
to
any
class a
and the
null-class are none, or
the null-class.
1
class a, there
is its
negative, -a, constituted by

a,
all
members
of the
universe of
discourse"
not contained in
and such that:

in
1-61
If
-a and any class x have no members
common,
122
then
A
all
members
and 1-62
of
x are
common
to x
of
and
a,
or x
is
contained in a;
a,
If all
members
and common
1-7
also to
y and -a,
of
any class y are common to y and then y must be null.

"everything",
The
"universe
discourse",
is
the negative of the
null-class,
1
"nothing".
That which
is
"a
is
either a or 6 or both
is
identical with the negative of
that which
1-9
both not-a and not-6.

"
That
is
contained in b
of a
is
equivalent to
b
".
"The
class a is identical
with the
common members
and
proved by these interpretations. In the form given, they are not independent, but they may easily be made so by certain alternations in the form of statement. 5
is
That the postulates are consistent
The
following abbreviations and conventions will be used in the state

of theorems:
ment and proof

1.
ax6
will generally
be abbreviated to a b or a -b,
ax(bxc)
to a (be),
ax-(6x-c)
2.
to a-(fc-c) or a--(&-c), etc.
In proofs,
we
shall
sometimes mark a lemma which has been established

in that proof refer to the
"Q.E.D."
as (1), or (2), etc.,
and thereafter
shall
lemma by
this
number.
Also,
we
sometimes write
instead of repeating
the theorem to be proved.

3.
The
principles (postulates, definitions, or previous theorems)
by which
any
step in proof is
If
taken
x
will usually
brackets, thus:
0,
be noted by a reference in square then [1-5] a x = 0. Reference to principles
whose use
is
more
Theorems
insertion
will
proceed. be numbered decimally, for greater convenience in the
or less obvious will gradually be omitted as
we
of
theorems without alteration of other numbers.
The non-
decimal part of the number will indicate some major division of theorems, as 1- indicates a postulate or definition. Theorems which have this digit
and the one immediately following the decimal point in common different forms of the same principle or otherwise closely related.
II.
will
be
ELEMENTARY THEOREMS
b c
2-1
If a
6,
then a
and
c b.
This follows immediately from the meaning of

v
= and
1.
2 2
b is equivalent to the pair, a

If
c b and
a.
b.
6,
then [1-2]
ab =
a and b a
s set.
On
this point,
compare with Huntington
The
Classic, or Boole-Schroder, Algebra of Logic
123
And
But
if
a b
=
a b
a and b a
b,
then
[1 -3]
a b
[1 -9]
is,
is
equivalent to a c b and b
= b a = b. a = b to 6 c a.
Equality
2-3
then, a reciprocal inclusion relation.
aca.
a
a,
hence [2-2] Q.E.D.

is "contained
in"
Every element
2-4
a -a
itself.
=
[1-2]
= -a a. a a = a. = (a -a) a. = (a -a) -a. (-a -a) (a -a) a = (a -a) -a = a -a, then a -a = 0. a = a -a. Hence also, -a a = 0.
-a
(a a)
Hence
[2-1, 1-4, 1 -3] a
-a
a (a -a)
Also [1-2] -a -a
-a.
Hence a -a = a
But
[1-62]
[1-3]
if
And
-a
Thus the product of any element modulus of the operation x

.
into its negative
is
0,
and
is
the
2-5
a -b
=
If
is
equivalent to a b
a and to a c
b.
=
if
a,
then
[1-4-5,. 2-1-4]
a -b
(a b) -6
(b -b)
And
[1-61]
a -6
=
is
0,
then
By (1) and And [1 9] a
(2),
a -b a
=
a c
ab = a and ab = a
c6
(2)
are equivalent.
6.
We
shall derive other equivalents of a
later.
The above
"1-4-5"
is
required
"2-1-4"
immediately.
instead of
In this proof,
1-5"
we have
"2-1,
written
and
"1-4,
and
2-4".
This kind of abbreviation in
references will be continued.
2-6
If
acO, then a =
If
0.
a cO, then [1-9] o-O

b,
=
c
a.
But
c b.
[1-5]
a-0
0.
2-7
If
a c
then a
c,
and
a c
If acb, then But [1-2-3-4] (a
[1-9] a b
b) c
a and [2-1]
(ac)
b) c
=
Hence, by
(ac)(cb)
= (b a) c = b = (ac)(bc)
a c
c b
6,
= (ac)
c)
= ac = [a (c c)
(1)
b]
= [(ac)
a
c
c]
(2)
(1)
and
(2), if
then (a c)(b
b c.
= ac and
also c a
[1 -9]
b.
b c.
And
2-8
-(-a)
[1-3] c a
= ac and
-
Hence
cc
a.
[2-4] -(-a) --a
0.
Hence
Hence
[2-5] -(-a)
ca
ca-a.
(1)
By
(1),
-[-(-)] c-a.
[2-7] a- -[-(-a)]
124
A
But
[2-4] a
-a
0.
Hence a --[-(-a)] cO.
Hence
[2-G] a --[-(-a)]
and
[2-5] a
c-(-a)
(2)
[2-2] (1)
and
(2) are
equivalent to -(-a)
a.
3-1
a cb
is
equivalent to -b c -a.
[2-5]
ac6
is
equivalent to a -6
0.
And And
If
[2-8]
a-b = -ba = -b-(-a).
-6 -(-a)
of
is
equivalent to -b c-a.
be transposed by negating both. then the portion of the plane not in b is contained in the portion of the plane not in a: if all a s are 6 s, all non-6 s are non-a s. This theorem gives immediately, by 2-8, the two corollaries:
The terms
is
any
relation
may
region a
contained in region
b,
3-12 3-13
3-2
a
a c -b
is
equivalent to b c-a; and equivalent to -6 c

a.
-a c
is
b is
equivalent to -a
-6.
[2-2] a [3-1]
=
is
b is
equivalent to (a
cb and bed).
cb
equivalent to -b c-a, and b
ca
to -a c-6.
is
Hence a =
lent to
b is
equivalent to (-a
-6.
c-b and -6 c-a), which
equiva
-a
The negatives
3-22
a
of equals are equals.
By
2-8,
we have
also
-b
is
equivalent to -a
b.
Postulate 1-6 does not require that the function
"negative
of"
be
There might be more than one element in the system having the properties postulated of -a. Hence in the preceding theorems, -a
unambiguous.
must be read
proved of -a,
in
"any
negative of a
",
-(-6)
6,
the negatives of any given negative of

etc.,
and
must be regarded as any one of so on. Thus what has been

But we can now demonstrate
has been proved to hold for every element related to a

postulate.
is
the
manner required by the
that for every element a there

properties postulated of -a.
one and only one element having the
3-3
-a
is
uniquely determined by
a.
By
such.
1-6, there
is
is
at least one element -a for every element a.

let
Suppose there
more than one:
-a x and -a 2 represent any two

[3-2] -a,
Then
Since
all
[2-8] -(- Gl )
-(-a 2 ).
Hence
-a 2
the relation x
functions in the algebra are expressible in terms of the negative, and 0, while is unique and a x b
,
a, b, c, etc.,
is
uniquely
125
determined by a and b, it follows from 3-3 that all functions in the algebra are unambiguously determined when the elements involved are specified.
(This would not be true
"division"
if
the inverse operations of
"subtraction"
and
were admitted.)
1 is
3-33
The element
[1-5]
is
unique.
is
unique, hence [3-3] -0
unique, and [1-7]
-0.
3-34
-1
0.
[1-7] 1
-0.
Hence
[3-2]
Q.E.D.
is
3-35
If
a and b are elements in K, a + 6

b.
an element
in
uniquely deter
mined by a and
The theorem
3-37
If
follows from 3.3, 1-1,

c
and
c
8.
6,
then a +
and
+a
b.
The theorem
,
follows from 3-35
and the meaning
of
=.
3-4
-(a +
6)
-a
-b. b
[1-8] a +
Hence
.3-41
[3-3, 2-8] -(a
-(-a -b). + b)
-[-(-a-b)}
-a -b.
-(a
6)
-a +
-6.
[1-8, 2-8]
-a + -6
-[-(-a) --(-6)]
-(a
6).
3-4 and 3-41 together state
De Morgan
sum
is
the product of the negatives of the
Theorem: The negative of a summands; and the negative of a

s
product is the sum of the negatives of its factors. The definition 1-8 is a form of this theorem. Still other forms follow at once from 3-4 and 3-41,
by 2-8:
3-42
3-43
-(-a + -b)
-(a + -6)
3-44
3-45
-(-a +
b)
= ab. = -ab. = a-b.

-a +
b.
-(a-b)
-(-ab)
= =
3-46
a + -b.
s
From De Morgan
3-34, -1
Theorem, together with the

equals",
principle, 3-2,
1
"The
negatives of equals are
the definition 1-7,
-0,
and theorem
x there
is
0, it
follows that for every theorem in terms of

in
corresponding theorem
replaced by
its
terms of +
If in
any theorem, each element be

result
is
negative,
and x and + be interchanged, the

can, of course, be replaced
valid theorem.
The negative terms
by
positive,
126
since
A
we can suppose x =
system there
is

-a, y
-b, etc.
Thus
for every valid
theorem
in the
another got by interchanging the negatives and 1 and the symbols x and + This principle is called the Law of Duality. This law is to be illustrated immediately by deriving from the postulates
. .
their correlates in terms of +
The
correlate of
1 is
3 35, already proved.
4-2
a+a
a.
[1-2]
-a-a =
-a.
Hence
[1-8,
3-2,
2-8] a + a
-(-a -a)
-(-a)
a.
4-3
a + b
a.
[1-3]
-a -b
-b -a.
Hence
[3-2]
-(-a -6)
-(-6 -a).
Hence
4-4
a+
(b
\1
[1-8]
c)
4.1
Q.E.D.
(a
(
=
n
+
A
b)
^\
c.
h\
Hence
[3-2] -[-a (-b -c)}
-[(-a -6)
-c].
But
[3-46, 1-8] -[-a(-fc-c)] [3-45, 1-8] -[(-a-6)-c]

1.
And
4-5
= a + -(-6-c) = a+(6 + c). = -(-a-b)+c = (a + b)+c.
a+1 =
Hence
[1-5]
-a-0
0.
Hence
[3-2] -(-a-0)
[3-46] a +
-0
-0,
and
x.
[1-7]
= a+1 = = =
-0.
1.
4-61
If
-x + a
If
1,
then x a
1,
-x + a
[2-5]
then [3-2-34-44] x -a
is
-(-z + a)
x.
=-1=0.
And
4-612
x -a
equivalent to x a
If -a;
+a
1,
then x + a
= =
y,
a.
[4-61] If
-a + x
if
1,
then a x
1,
a,
and
a.
1.
[3-2]
-a + -x
-a
(1)
By
4-62
If
(1)
and 2-8,
-x + a
x+a
= =
a)
y+a
If
y and y + -a
y, [3-2]
= =
then y
And
But
4-8
a + -a
y+a = if y + -a
if
-y -a
y,
-y a
-(y + -(y + -a)
=
-y. -y,
-y.
[1-62]
-y a
-y and -y -a
= =
-y
and y = -0 =
1.
l=-a + a.
-a a
(Correlate of 2-4)
[2-4]
0.
Hence
[3-2] a
+ -a
is 1.
-(-a a)
=-0 =
all
1.
Thus the modulus

4-9 -a +
6
of the operation +
1,
a+ b
=
0,
b,
a -6
=
a,
0,
a,
and a c b are
equivalent.
[2 5]
a -b
6
a 6
and a cb are equivalent.
[3-2]
-a +
1 is
equivalent to a -b
-(-a +
b)
=-1=0.
-M,V)

[4-612] If -a + b
127
=
b,
1,
a+ b
b.
And
if
a+b
= =
[3
37] -a + b
-a +
(a
b)
(-a + a) + 6
= 1+6
1.
Hence a +
b is equivalent to
-a + b
1.
We
5-1
If
turn next to further principles which concern the relation c

a c 6 and b cc, then
[1-9] a
If
ace.
cb
a
is
is
equivalent to a b
c
a,
and
a
cc
to b c
=
a.
b.
a b
=
c
a and b
b,
(a b) c
c.
(b c)
a b
But a
equivalent to a c
the Syllogism.
This law of the transitivity of the relation c is called the Principle of It is usually included in any set of postulates for the algebra which are expressed in terms of the relation c
.
5-2
a b
ca and ab cb. = a (a b) = (a a) b = a b. (a b) a But (a b) a = a b is equivalent to ab ca. = a (b b) = a b, and ab Similarly, (a b) b

a ca + b
[5 2]
cb.
5-21
and
ca +
b.
-a -b c -a and -a -b c -b.
Hence
[3-12]
ac-(-a-b) and 6c-(-a-6).
But -(-a
-b)
(a
b).
Note that 5-2 and 5-21 are

general,
correlates
by the Law
of Duality.
In
,
having now deduced
the fundamental properties of both x and +

in
we
shall give further
theorems
such pairs.
A
5-22
corollary of 5-21 is:

a b
ca + b.
[5-
1-2- 21]
c
5-3
If
a c b and
[1-9] If
d,
then a
c
cb
d.
a c b and
cd, then a
b
c,
a and
c
d
d.
c.
Hence
5-31
If a
(a c) (b d)
b
=
d,
(a b) (c d)
=
c
b
and a
cb
and
then a +
d.
If
a c 6 and
[5-3] [1-8]
cd, [3-1] -b c -a and -d c -c.

[3-1]
Hence Hence
-b-dc-a-c, and
Q.E.D.
a and a + a
-(-a -c) c-(-fc-d).
By
5-32
the laws,
If
aa =
a,
5-3 and 5-31 give the corollaries:
a c
and
b cc,
then
ab
cc.
128
A
If

c.
5-33
a c c and b c
c,
then a + 6 c
c.
5-34 5-35 5-37
If a If If
and ace, then a cb
a c 6 and ace, then a c 6 + a c6, then a +

c
c.
c6
c.
(Correlate of 2-7)
[2-3]cce.
5-4
a + ab =
a.
Hence
[5-31]
Q.E.D.
[5-21]
[2-3]
aca + ab
a,
(1)
a c
and
[5-2]
(2),
ab ca.
Hence
[5
33] a + a
ca
(2)
[2-2] If (1)
and
a.
then Q.E.D.
5-41
a (a +
b)
[5-4]
-a + -a-b
-a.
But
[3-4] -(-a +
--&) =
-(-a + -a-b) a --(-a -6) =0(0 + 6).

[3-2]
Hence
-(-a)
a.
Law of Absorption. We have next to prove the Distributive Law, which requires several lemmas.
5-4 and 5-41 are the two forms of the
5-5
a
(b
c)
= ab +
ac.
Lemma
1:
[5-2] [5-2] a b
ab + acca(b + c). a6ca and acca.

c
b
Hence
[5-33] a b + a c
ca
cb+
c.
(1)
and a
c
b
c
c
c.
But
6
[5-21] b
c.
and
Hence Hence
[5-1] a b
[5
and ac cb +
c
+
c
33] a b + a c
(2)
[5-34] If (1)
and
q
(2),
then a
+a
ca
is
(b
c).
Lemma
If
pc
-q.
is false,
then there
an element
x, 4= 0,
such that
x c p and x c
p-q
[4-9]
if
is
such an element, for [5-2]

if
p-qcp
q
and p-qc-q- and

then p -q
4= 0.
(This
p-q = 0, then pcq, hence lemma is introduced in order to

3
:
pc
is false,
simplify the proof of
Lemma
3.)
Lemma
4=
(6
c)
+a
c.
Suppose this 0, such that
false.
Then, by lemma
2,
there
is
an element
x,
xca(b + c)
.TC-(6 + ac) But [3-12] if xc-(b + ac), then b + acc-x [5-1] If (1), then since [5-2] o (6 + oc) co, z
(1)
and
(2)
co
(3) (4)
(5)
and
Also
also, since a(6
+ a
c)
c, .r
[5-1] If (2),
[5
1] if
then since [5-21] b cb + a c, b c -x and [3-12] xc-b (2), then since [5 21] a c c b + a c, a c c -x and [3-12]
zc-(oc)
(6)
The
129
From
(6)
and
x c
(3), it
a,
follows that
[5
xcc must
But
if
be
c
false;
for
if
.TCC
c),
and
(3)
then
34]
x ca
c.
x ca
and
(6)
x c -(a
then [1-62] x
0,
But
if
x c
be
false,
which contradicts the hypothesis x 4= 0. then by lemma 2, there is an element y, 4=
0,
such that
ycx
and
[5-1] If (7)
If (8)
If (7)
(7)
?/c-c,
(5),
or [3-12]
and
(9),
and
and
(4),
then y c-b and [3-12] then [5-33] b + cc-y and [3-12] yc-(b + then [5-1] ycb + c
cc-y bc-y
(8) (9)
c)
(10)
(11)
[1-9] If (11),
But
y
if
(12),
then y(b + c) then [1-62] y
=y, and if = 0, which
(10),
y~(b +
c)
(12)
contradicts the condition,
*0.
false
is
Hence the supposition that a(b + c)cb + ac be supposition, and the lemma is established.
a false
Lemma
a(b + c)cab + ac. By lemma 3, a (b + c) c Hence [2-7] a [a (b + c)] c a

4
:
+ a
c.
(b
+ a
c)
But a [a (b + c)] = (a a)(b + c) = a (b + c). And a (b + a c) = a (a c + b). Hence a (b + c) c a
(a c
b).
But by lemma
3,
a (a c + b) c a c + a
c.
b.
And
Proof of
a c + a b
a b + a
Hence a
1
(b
c)
a b + a
c.
the theorem:
[2-2]
Lemma
c.
and lemma 4 are together equiva
lent to a (b + c)
a b + a
This method of proving the Distributive

"Sets
Law
is
taken from Huntington,

".
the long and
The proof of Algebra of Logic due to Peirce, who worked it out for his paper of 1880 but mislaid the sheets, and it was printed for the first time in
of
Independent Postulates
difficult
for the
lemma
is
Huntington
5-51
(a
paper.
+ b)(c + d) = (a c + [5 5] (a + b)(c + d)
b c)
(a
d+
b d).
(a +
b) c
(a
b)
(a c
+b
c)
(a
d + b d).
5-52
a + b c
(a + b)(a
c).
(Correlate of 5 5)
[5-51] (a +
b) (a +
c)
(a a + b a) + (a c + b c)
=
But
[5-4] (a + a
b)
[(a
+ a
b)
+ a
c]
c.
+ a
a + a c
a.
Hence Q.E.D.
follows:
300, footnote.
Further theorems which are often useful in working the algebra and
which follow readily from the preceding are as

6
See
"Sets
of
Independent Postulates,
etc.",
loc. rit., p.
10
130
A
a-1
5-6
1 -a.
[1-5]
a-0
if
0.
Hence a--l =
0.
But
5-61
[1-01]
a--l
= =
0,
then a-1
a.
acl.
[1-9] Since a-1
a,
acl.
-a--0
5-63
-a-1
-a.
Hence
[3 2]
a+
-(-a--0)
-(-a)
a.
Oca.
0-a
a-0
0.
Hence
[1-9]
Q.E.D.
5-64
ca
is
equivalent to a
1.
[2-2] a
1 is 1
equivalent to the pair,
acl and
ca.
But
5-65
a c
[5-61] a c
is
holds always.
Hence Q.E.D.
equivalent to a
0.
[2 2]
is
equivalent to the pair, a c
and
a.
But
5-7
If
[5-63]
a;
ca
holds always.
Hence Q.E.D.
x.
a + 6
If
= and a = 0, then 6 = a = 0, a + 6 = + 6 = 6. =
x and a
5-71
If a 6
= =
1,
then 6
x.
If a
= =
1,
a 6
1-6
6.
5-72
a + 6
If
=
a
is
equivalent to the two equations, a
and
0.
and
=
=
0,
then a + 6
And
But
if
if
a+ 6
-a -6
if
= =
0,
1,
-a -6
a
-(a +
6)
= =
= =
1.
0.
-0
a-1
a(-a -6)
(a -a)
-6
0--6
0.
And
5-73
a 6
[5-7]
1 is
a+ 6
and a
0,
then 6
0.
=
If
equivalent to the two equations, a

1
-
and
1.
and 6 = 1, then a 6 = 1 1 = 1. = 1, -a + -6 = -(a 6) =-1=0. Hence and -6 = 0. But [3-2] if -a = 0, a == 1, and if -6 = 0, 6 = 1.

a
==
And
if
a 6
[5-72]
-a
5-7 and 5-72 are important theorems of the 5-7, "Any null algebra. term of a sum may be dropped", would hold in almost any system; but a sum is null, each of its summands is 5-72, is a special law null",
"If
characteristic of this algebra.
It is
tains no inverses with respect to
due to the fact that the system con and 0. a and -a are inverses with

respect to
131
x and
and with respect to + and

less useful.
a.
1.
5-71 and 5-73, the
correlates of 5-7
and 5-72, are

a b + a -b
(b
5-8
(b
+ -b)
[5-5]
=
a
+ -6)
=
1.
a b + a -b.
And
5-85
a+
b
[4-8] b
+ -b
b.
Hence a
(b
+ -b)
a-l
a.
a + -a
= ab + -ab. Hence a + b = a + (a b + -a b) = (a + a But [5-4] a + a b = a. Hence Q.E.D.

[5-8] b
It will
b)
+a
flb.
be convenient to have certain principles, already proved for two terms or three, in the more general form which they can be given by the
use of mathematical induction.
Where the method
of such extension is
obvious, proof will be omitted or indicated only.

are associative,
Since both
x and
we can
dispense with parentheses by the definitions:

Def.
5-901
a+ b+
c=(a + b)+c
Def.
5-902
5-91
abc = (ab) c a = a (b + -b) (c

[5-8]
1
+ -c) (d + -d)
5-92
=
a
(a
+ -a)(6 + -6)(c + -c)...

[4-8]
5-93
a+
ab + ac + ad+...
(a + d)
[5-4]
5 93 1
=
+
a (a + b) (a +
[5-41]
c)
5-94
(b
+ d + ...)
[5-5]
= ab + ac
ad+
.
...
5-941
a + bcd.,..
[5-52]
--=
(a
+ b)(a + c)(a + d)
5-95
-(a + b +
If
= -a
-b -c ...
the theorem hold for n terms, so that

-(ai + a 2 +
.
+ a n)
-a-i
-a z
-a n
then
it will
hold for n
terms, for by 3-4,
And
[3-4] the
theorem holds
for
two terms.
Hence
it
holds for
any
number
of terms.
132
A
-(a

-c +
5-951
bed...) = -a + -b +
a+ b + c +
-d+
Similar proof, using 3-41.
5-96
+-a-b
-c
[4-8, 5-951]
5-97
a + b+
is
equivalent to the
set,
0, 6
0, c
0,
[5-72]
5-971
abed...
[5-73]
--=
1 is
equivalent to the
set,
==
1,
1, c
I,
5-98
a-b
[1-2] a a a a
=ab-ac-ad... = a.
.
5-981
a +
(b
+d+
)
.
=
. .
(a
b)
(a
c)
+ (a + d) +
[4-2] a + a + a +
a.
s
The
extension of
De Morgan
Theorem by 5-95 and 5-951
is
especially
important.
5-91, 5-92, and 5-93 are
different forms of the principle
by
which any function

whatever
may
be expanded into a
originally involved in the function
sum and any elements not introduced into it. Thus any expression
be regarded as a function of any given elements, even though in the expression, a peculiarity of the algebra. 5-92, the expression of the universe of discourse in any desired terms, or expansion
may
they do not appear

the basis of
of
1, is
many important
procedures.
if
The theorems 5-91-5-981

involved be
induction.
III.
finite, since
are valid only
the
number
of elements
proof depends upon the principle of mathematical
GENERAL PROPERTIES OF FUNCTIONS

etc., to
We may
only
use/Or), $(x, y),

of the class
members
denote any expression which involves and the relations x and + The further
.
requirement that the expression represented by /(.r) should involve x or its negative, -x, that $0, y) should involve x or -.r and y or -y, is unnecessary, for if .r and -x do not appear in a given expression, there is an equivalent
expression in which they do appear.
a
By
5-91,
-?/)
a (x +
-.r)
a x + a
-x
(a
x + a -x) (y +
= axy + ax-y + a-xy +

a x + a -x
to x.
a -x -y, etc.
reference
reference
may be called the expansion, or development, of a with And any or all terms of a function may be expanded with
and
so
of elements.
to x, the result expanded with reference to y,
on for any elements

ele-
and any number
Hence any expression involving only
133
ments
in
and the
relations
x and +
may
be treated as a function of
any elements whatever. If we speak of any a such that x = a

of x being given, the value of
as the
"value
of x
",
then a value
as in
any
other.
any function of x is determined, in this algebra But functions of .r in this system are of two types: (1)
those whose value remains constant, however the value of x
may vary, and those such that any value of the function being assigned, the value of x (2)
thereby determined, within limits or completely. Any function which symmetrical with respect to x and -x will belong to the first of these
in general, a function
is
is
classes;
which
is
not completely symmetrical with
respect to x
and -x
will
belong to the second.
But
it
must be remembered,
in this connection, that a
unless
it
symmetrical function may not look symmetrical be completely expanded with reference to each of the elements
involved.
For example,
a + -a b + -b
is
symmetrical with respect to a and -a and with respect to panding the first and last terms, we have
a
(b
and
-b.
Ex
+ -b) + -
(a
+ -a) -b
b.
it
a b + a -b + -a b + -a -b
whatever the value of a or of
Any
is
function in which an element,
.r,
does not appear, but into which
introduced by expanding, will be
symmetrical with respect to x and -x. The decision what elements a given expression shall be considered a
function of
is,
in this algebra, quite arbitrary except so far as
it
is
deter
mined by the form

and
"variables"
of result desired.
"unknowns"
The
distinction between coefficients
or
is
not fundamental.
In fact, we shall
frequently find
it
say
of x
and
y,
convenient to treat a given expression first as a functionthen as a function of z, or of x alone. In general, coef
ficients will
be designated by capital letters.

of
The Normal Form

can be given the form
a Function.
Any
function of one variable,
/(.r),
A
where
x+
-x
is
and
are independent of x.
This
the normal form of functions
of one variable.
Any function
of
one variable, /(.r),

.r,
is
such that, for some
and some
which are independent of
/(.r)
= A
x +
-x
134
A
Any
the relations
and expression which involves only elements in the class x and + will consist either of a single term a single
or of a
element, or elements related by x
sum
of such terms.
Only
x,
four kinds of such terms are possible:

(2)
(1)
those which involve
those which involve -x, (3) those which involve both, and (4)
those which involve neither. 7 Since the Distributive Law, 5

of x, of -x
5,
allows us to collect the coefficients

of such
and
of (x -x), the
most general form

(x -x)
an expres
sion
is
px + q -x + r
where
p,
q, r,
and
s are
independent of x and -x.

0.
But
[2-4]
And
[5-9]
r(x-x) = r-0 s = s x + s -x.

+ q -x +
r (x
Hence
px
-x) + s
q +
=
s,
(p + s) x + (q + s) -x.
Therefore,
A = p + s, B =
gives the required reduction.
The normal form
of a function of
C2,
+
Xn
,
variables,
+!
may
be defined as the expansion by the Distributive

f(Xi,
Law
of
X2
...
X n ) -X n +i +/
(.Tl, .T 2 ,
X n ) -X n+ i
x*>,
where/ and/
in the
are each
some function
is
of the n variables, Xi,

step"
...
x n and
,
normal form.
This
"step
by
definition; the normal form
of a function of
two variables
is
defined in terms of the normal form of
functions of one variable; the normal form of a function of three variables

in
terms of the normal form for two, and so on. 8
Thus the normal form
of a function of
two
variables, $(x, y), will be
found by expanding
-y
(A x +
It will be,
-x)
y+(Cx + D -x)
Axy + B -x y + C x -y + D -x -y
of a function of three variables, ty(x, y, z), will be
The normal form
Axyz + B-xyz + Cx-yz + D-x-yz + Exy-z + F-xy-z

+
G x -y
-z + II -x -y -2
And
7
so on.
Any
function in the normal form will be fully developed with

"involves"
By
a term which
is
meant a term which

in
either is x or has
is
factor".
But
"factor"
seems inappropriate
an algebra
in
which h x
x a always contained
"as
in x, h
8
x ex.
This definition alters somewhat the usual order of terms in the normal form of func tions. But it enables us to apply mathematical induction and thus prove theorems of a generality not otherwise to be attained.

reference to each of the variables involved
135
that
is,
each variable, or
its
negative, will
appear
in
every term.
be given the normal form.
of
6-11
Any
function
may
1
,
(a)
By
If
(>
any function
one variable
may
be given the
normal form.
(6)
functions of n variables can be given the normal form,
then functions of n
1
,
variables can be given the normal form, for,
Let $(1,
Xz, ...
x n Xn+i) be any function of n

will
variables.
By
definition, its
normal form
be equivalent to
(Xi,
f(Xi, Xz, ...
X n ) Xn+l +f
XZ
,
...
X n ) -Xn+l
wLere / and /
form.
are functions of x\, x 2
...
x n and in the normal
By
the definition of a function, $(#1, x 2)
x n x n +i)
,
may
be re
garded as a function of x n +i.
Hence, by
i
6-1, for
some
and some
which are independent

$(Xi, Xz,
. . .
of x n ^
,
X n Xn+l)
= A
X n+ i +
-X n +l
7
Also,
by the
definition of a function, for
some/ and some/

x n)
.
A = f(xi,
and
Hence, for some / and /
&(Xi, X 2
,
. . .
x2
...
.
.
B =
(xi, Xz,
xn)
of x n +\
which are independent

f(Xi,
-To,
.
. .
X n Xn+l)
,
X n ) X n +l +/
(.Ti,
X Zt
X n )Xn+l
Therefore,
if
the functions of n variables, / and /

. . .
>
can be given the
x n x n +\) can be given the normal form. normal form, then $(0*1, Xz, be given the normal form, (c) Since functions of one variable can
and
since
if
functions of n variables can be given the normal form,
functions of n
variables can be given the normal form, therefore

of variables
functions of any
number
can be given the normal form.
The second
is
step, (6), in the
above proof
may seem
,
.
arbitrary.
That
it
valid, is
due to the nature of functions

of
in this algebra.
. -
x n ), the normal form will n variables, &(xi, x 2 all the combinations of x it positive or be a sum of 2 terms, representing with .r n positive or negative, with x z positive or negative, with negative, each term having its coefficient.
6-12
For a function
n
136
A
(a)
by
will
definition of the
normal form function of one variable has two terms, and normal form of functions of n + 1 variables, if
functions of k variables have 2* terms, a function of k
variables
have 2 k
(b)
2 k or 2 A+1
,
terms.
normal form function of one variable has the further character described in the theorem; and if normal form functions
of k variables
will
have
this character, then functions of k
-f-
variables
have
1
it,
since, by definition, the normal form of a function of
variables will -consist of the combinations of the (k
l)st
variable, positive or negative,
with each of the combinations repre
sented in functions of k variables.

Since any coefficient
tain terms
coefficient
may
1.
be
0,
the normal form of a function

coefficient for a
may
con
which are
is,
null.
of course,
Where no The order
of terms in
term appears, the the normal form of a
function will vary as the order of the variables in the argument of the function is varied. For example, the normal form of y) is, by defini
3>(x,
tion,
A
and the normal form
xy+
B -x y + C x -y + D
is
-x -y
of V(y, x)
P yx + Q-yx + Ry-x + S-y -x

Except for the
tive
coefficients, these differ
order of the elements in the terms.
only in the order of the terms and And since + and x are both associa
is
and commutative, such a difference
not material.
differ materially
6-15
Any two
functions of the
same variables can
only
in the coefficients of the terms.
The theorem
that, for
follows immediately from 6-12.
In consequence of 6-15,
we
can, without loss of generality,

of the
assume
in the
any two normal form functions

of
same variables with which

in
we may be concerned, the order

arguments of the functions
n
is
.
terms and the order of variables
the same.
,
And
is
also,
any function
of
variables,
3>(
Xl
x2
. .
xn
.r
n+1 ),
which
(X,,
1
.To,
equated to
.
/Ol,
x n+l
it
.1-2,
Xn)
X n+1 +f
X n ) -X n+l
may
be any chosen one of the n

last is consistent
variables.
The convention that

normal form
is
always the
with complete generality of the proofs.

of a function in the
is
6-17
null.
The product
of
any two terms
137
By
6-12, for any two terms of a function in the normal form,
there will be
some
variable, x n
such that x n
is
positive in one of
them and negative in the other; since otherwise the two terms would represent the same combination of x\ positive or negative,
t
with
0*2,
positive or negative, etc.

will
any two terms
Consequently, the product of involve a factor of the form x n -or n and will
,
therefore be null.
Unless otherwise specified,

tion
it
will
be presumed hereafter that any func
mentioned
is
in the
normal form.
The
Coefficients in a Function.
The
itself.
coefficients in
any function can be
expressed in terms of the function
6-21
If f(x)
= A = A
y.
11 -.r,
then /(I)
= A.
+ B-0
For /(I)
6-22
If /(.r)
= A-1
x+
+ U--1
= A
A.
For/(0)
6-23
/(.r)
B -.r, then /(O) = B. = A-O + B-Q = + 7M

+/(0)~z.
==
B.
=/(!)
a;
The theorem
These laws,
first
follows immediately from 6-1, 0-21
and 0-22.
stated
by Boole,
are very useful in reducing compli

if
cated expressions to normal form.

^(.r)
For example,
x + -d -x) +
(c
c (d
.r)
reduction by any other
method would be
tedious.
But we have
d
*(1)
and
*(0)
= ac (VM+-d-0) = a c (d-Q + -rf-1)

of ty(x)
is
+ (c+l) d +
(c
= acd+cd+d =
+ 0) d
a c -d + c d
Hence the normal form
given by
c
\F(.r)
d x + (a
-d +
c d)
-x
for functions
Laws analogous to 6-23, also stated by Boole, may be given of more than one variable. For example,
/(.r,
y)
/(I, l)-x
y +/(0,
?/
1)
-a y +/(!,
1, 1)
-.)*
0)
-.r
-y
+/((), 0)
0,
-.*
-//
and
$(.r, y, z)
$(1,
1, l)-.r
+ $(0,
+
<I>(1,
l)-x-ijz
--.r
+ $(0,0, l)--.r-z/2+ $(1, l,0)-.r z/-s + $(0, 1,0)

+ $(1,
0,
y-z
0) x -y -z + $(0,
0, 0)
-.r -//
-z
We
can prove that this method of determining the coefficients extends to
functions of anv
number
of variables.
138
A
If
Surrey of Symbolic Logic
6-24
I -.TI -.i- 2 -.r 3
-x n
f J
be any term of
^(.TI, xz,
x3
...
#)>
then
-j
[-
will
be the coefficient, C.
I 0, 0, 0,
...
(a)
(6)
By
If
6-23, the theorem holds for functions of one variable.

A*
the theorem hold for functions of
variables,
it
will
hold
for functions of k
variables, for,
By
is
6-11,
such that,
any function of k + 1 for some / and some /
variables, $(#1, Xz, ...

,
.T*,
0, 0,
...
Therefore,
if
every term of / be of the form
{1,1, I 0, 0,
...
...
ll
j
f
I
.n
xz...
. . .
xk
\
}
-x, -x 2
is
-x k
then every term of

f
in
which x k+
r
positive will be of the form
fi,i,...ii
I
0, 0, ...
J
x,
x^.. xk
. . .
\
J
-.n -x 2
-xk
\
(^
and the
coefficient of
any such term
will
be /
r U, U,
... U J
which,
And
similarly,
if
every term of /
...
...
be of the form
.,f 1,1,
I 0, 0,
11
j
-.rj -,r,
-x t
then every term of

fl,
*
1,
in
...
which
x/,+i is
f
negative will be of the form
l|
Tl
x,...
.rt l
Mo.o,. .ol
i-.,,-.,,.
.-,,|
139
and the
coefficient of
any such term
will
be /
which,
Hence every term

r
I
of
<
will be of the
form
x2
...
.
1,1,.
..MI
j
0, 0, ... 0,
I -.?! -.r 2
-x k -xk+l
(c)
Since the theorem holds for functions of one variable, and
since
of k
if it
hold for functions of k variables,

variables, therefore
it
it
will hold for functions
holds for functions of any
number
of
variables.
For functions
but
less useful If f(x)
of
one variable, further laws of the same type as 6-23
have been given by Peirce and Schroder.

+ B)
Ax + B-x:
= f(-A
+ -B).
6-25
/(I)
/(())
= f(A
6-26
=f(A-B) =f(-A-B). 6-27 j(A) = A + B =/(-J8) = f(A-B) =f(A+-B)

6-28
f(B)
= A-B =
/(-4)
= f(-A-B) = f(-A
+ B)
The
proofs of these involve no difficulties

it
and may be omitted.

be convenient to denote the coef
,
. .
In theorems to be given later,

ficients in functions of the
will
form
$(.?!,
x2
x n ) by A\,
z,
3,
A^,
or
by
Ci,
C2 C3
,
.,
etc.
This notation
is
perfectly definite, since the

is
order of terms in the normal form of a function

of
,
. .
fixed.
If
the argument
,
x n ), then any one of the variables, x k any function be (a*i, .r 2 be positive in the term of which Cm is the coefficient in case
.
will
p-2
~l
<
m^
(p
l)-2
~l
.r^
where p
= any
in the term.
even integer (including 0). Thus it may be determined,
Otherwise
for
will
be negative
in the
is
each of the variables
function, whether
coefficient,
it is positive or negative in the is
term
of
which Cm
the
and the term

it
thus completely specified.
We make
no use
of this law, except that

it will
validates the proposed notation.
be convenient to distinguish the coefficients of those Occasionally in which some one of the variables, say x k is positive terms in a function from the coefficients of terms in which x k is negative. We shall do this
,
140

PS,
for coefficients of
by using different letters, as PI, P 2 which Xk is positive, and Q b Q 2 Q 3

,
terms in
for coefficients of
terms
in
which Xk
of terms,
is
negative.
This notation
is
perfectly definite, since the

is
,
number
it is
is
for a function of
is
n variables,
n always 2 the number of those in which Xk
always equal to the number of those in which and the distribution of the terms in which x k is positive, or
positive
is
negative,
is
negative,
determined by the law given above. The sum of the coefficients, Ai + Ao +

cated by
Since the
+
2
.,
3
will frequently
...
be indi
h
Â
is
or
Â
h
the product,
Ai-A A
by JJ/l or YL^h-
number
of coefficients involved will

r
always be fixed by the func

to indicate numerically the
tion which
in question, it
range of the operators
w ill be unnecessary and IJ

.
The Limits
the coefficients.
of a Function.
The lower
limit of
any function
is
is
the prod
uct of the coefficients in the function, and the upper limit
the
sum
of
6-3
A B
cAx + B-xcA + B.
(A B)(A x +
-x)
= AB
x +
AB -x
= A B
Hence
[1-9]
ABcAx + B-x.
B -x)(A
+ B)
And (A
But
x +
Ax + AB -x + ABx + B -x
=
(A
B + A)
and
x +
(AB + B)
A
x+
-x.
A B + A == A, and AB + B = B. Hence (.1 x + B -x) (A + B) = A x + B -x,

[5-4]
[1-9]
-x
cA + B.
6-31
f(B)cf(x)cf(A). [(3-3 and 6-26, 6-27]

If
6-32
then
the coefficients in any function, F(x lf x 2
;i-),
be
C C2 C
lt
,
3,
(a)
(b)
By
3,
the theorem holds for functions of one variable.

,
.
. .
Let $(xi, x z
for
.
. .
x k x k +i) be any function of k

,
variables.
By (Ml,
$(.ri, *2,
some / and some /

x k Xk+i)
,
f(x lt x z
xk )
-iCfc+i
(.TI,
+/
Since this last expression
x2
x k ) -xk+ i
(1)
may
x2
be regarded as a function of Xk+i in
which the
i,
coefficients are the functions
/ and /
$(.ri,
[6-3]
,
. .
x2
x k ) x/
(#,,
xh ) c
x2
x k x k +i)
,
The Classic, or Boole-Schroder, Algebra of Logic Let Ai{$},

Ai[3>\,
141
{3>},
etc.,
be here the coefficients in

<J>;
Ai{f\,
^2 {/I, A*\f},
etc.,
etc.,
the coefficients in/; and A,{f

.
\,
{/
[,
{f
},
the coefficients in /
If
rU{/} c/and IU{/
c/
then [0-3]
and, by
(1),
EU
{/J
JL
,
!/
<.
But
in
since (1) holds,
any
/
coefficient in
will
be either a coefficient
/ or a
coefficient in
and hence
!
lUf/! xIL-i{/
Hence
if
==IU!*i
x2 x k ),
the theorem hold for functions of k variables, so that
II A
then
{/}
c/Cd, * 2
xk )
...
and
,
EU
l,
{/
c/
(x lt
II^l^!
if
c$(a-i, x z
x k x k+} ).
Similarly, since (1) holds, [0-23]
Hence
/c
.!{/}
and / c
$c/+/ then Z^f/

+
[5-31]
*<=Z^{/1
Z^{/
But
since
any
,
coefficient in
is
either a coefficient in
==
ficient in
/ z^f/i Hence $ c Â{$}. Thus if the theorem hold
z^m
1
/ or a coef
z^j.
it
for functions of k variables,
will
hold for functions of k

(c)
variables.
Since the theorem holds for functions of one variable, and
since
if it
hold for functions of k variables,

variables, therefore
it
it
will hold for functions
of k -f
holds generally.
As we
shall see, these
the basis of the
method by which eliminations

Since
all
theorems concerning the limits of functions are are made.

functions of the same variables
Functions of Functions.
be given the same normal form, the operations of the algebra be performed simply by operating upon the coefficients.
may may frequently
6-4
If f(x)
= A
x+
-x, then
4/0)] -
-.-1
.r
-B
-x.
[3-4]
-(A x +
-x)
= -(A
= =
x)(B -.r)
.r)
(-A + -x) (-B +
-.1
-B
+ -.1
.T
-B -x
(-A -B
+ -A) x + (-A
+
-B
+ -B) -x
But
[5-4] -.1
-B
+ -.1
-x)
Hence -(,1 x +
== -A and -.1 -B = -A x + -B -.r.
-B = -B.
142
A
The negative
of

in the
6-41
any function,
normal form,
its
is
found by
re
placing each of the coefficients in the function by

(a)
(b)
negative.
By
If
6-4, the theorem
is
true for functions of one variable.

it
the theorem hold for functions of k variables, then
will
hold for functions of k
variables.
Let F(XI, xz, Then by 6-11 and 3

.
Xk, Xk+i)
2, for
be any function of k
,
variables.
some / and some /
+f
But
f(xi, Xz,
.
. .
(xi,
x2
x k )--x k+l
x k )-x k+ i
+f
Oi>
^,
Zk)
-Xk+i
may
be regarded
as a function of a^+i.
Hence, by 6-4,
,
a-o,
x k ) -x k+l +/
(xi,
xz
x k ) -xk+ i]
Hence
if
the theorem be true for functions of k variables, so that
the negative of /
its
is found by replacing each of the coefficients in / by and the negative of / is found by replacing each of the negative
coefficients in
by
its
negative, then the negative of

in
will
be
found by replacing each of the coefficients as has just been shown, any term of
F by
its
negative, for,
-x
in
l}
xz,
x kj
its coefficient is
which x k +i
is
positive
is
such that
X2
,
a coefficient in
-[/(>!,
...
X k )]
and any term
of
-[F(x lt
in
xz,
xk
which x k +\
is
negative
is
such that
its coefficient is
a coefficient in
-[/ (*
(c)
X*
...
.T,; )]
Since (a) and
(b)
hold, therefore the theorem holds generally.

is
Since a difference in the order of terms
not material, 6-41 holds not
only for functions in the

pletely
normal form but
for
any function which
is
com
expanded
so that every
or negative, in each of the terms.
element involved appears, either positive It should be remembered that if any

in the
term
an expanded function is missing, its coefficient is 0, and negative of the function that term will appear with the coefficient 1.
of
143
6-42
The sum
#2,
.
.
of
.
any two functions

is
of the
same
variables, $(0:1, T 2
xn)
and V(xi,
x n ),
another function of these same variables,

F(Xi, To, ... X n ),
such that the coefficient of any term in the corresponding terms in $ and ^.
is
the
sum
of the coefficients of
By
6-15,
$(0:1,0:2,
...
x n ) and V(xi, x 2
...
xn)
cannot
differ
except in the coefficients of the terms.
Let AI, At,
3,
etc.,
be the coefficients in $;
BI,
B
TI
2,
B
x*
3,
etc.,
the
coefficients of the corresponding
terms in ^.
. .
For any two such corf

.
r responding terms,
f
Ak J 1
.!
TI
TO
Tn
L -0*1 -.1 2
r ~.l n J
and B k
Tn
1
f-
xn
.
.
]
}
l
I -.Ti -.r 2
.
-X n
*1
*2
...
Ah
f
nk
f
-j
T!
0: 2
...
.
. .
-Ti -1 2
-Xn
-Xi -X 2
Xn
Ti
T2
Xn
-x n
of the
-TI -T 2
And
since addition
is
two functions
and commutative, the sum equivalent to the sum of the sums of such
is
associative
corre
sponding terms, pair by

6-43
pair.
The product
^f(xi,
0*2,
of
two functions
is
of the
same
variables,
$(1,
o: 2 ,
...
xn)
and
...
x n ),
another function of these same variables,

F(Xi,
.T 2 ,
-
Xn),
such that the coefficient of any term in

cients of the corresponding terms in
T Let Ak J
"j
F
^.
f
is
the product of the
coeffi
$ and
Xi
X2
Xn
-TI -T 2
-x n
and Bk
^.
Tl
0*2
Ti
T2
...
.
. .
Xn
}
r }
-T! -T 2
-x n
be any two
corresponding terms in
$ and
J
I
A k \[
,
TI
T2
n 1
...
.
. .
Xn
-Ti -T^
& H -X n Jfx5
-Ti -T 2
-X n
-(4*Jfc>{ -Ti -To t
-T n
By
6-15,
$ and ^ do not
differ
except in the coefficients, and by
6-17, whatever the coefficients in the normal form of a function, the
product of any two terms is null. Hence all the cross-products of terms in $ and ^ will be null, and the product of the functions will
144
A
be equivalent to the
pair
sum
of the products of their corresponding terms,
by
pair.
Since in this algebra two functions in which the variables are not the same may be so expanded as to become functions of the same variables,
these theorems concerning functions of functions are very useful.
TV.
FUNDAMENTAL LAWS OF THE THEORY OF EQUATIONS
We
may
of
have now to consider the methods by which any given element be eliminated from an equation, and the methods by which the value
"
an
unknown
"
may
be derived from a given equation or equations.

of equation for eliminations
0.
The
is
most convenient form

equation with one
and solutions
the
member
Equivalent Equations of Different Forms. If an equation be not in the form in which one member is 0, it may be given that form by multiplying
each side into the negative of the other and adding these two products.
7-1 a
b is
equivalent to a -b + -a b
0.
[2-2] a
[4-9] a
b is
equivalent to the pair, a c b and b c a. = equivalent to a -b = 0, and b c a to -a b

6 is
0.
And
[5
72] a -b
and -a
are together equivalent to a -b
+ -a b
0.
The transformation
7-12
a
of
an equation with one member
1 is
obvious:
1 is
equivalent to -a
0.
[3-2]
By
in
6-41,
in
form
xn) = 1 any equation of the form f(xi, ,r 2 which one member is simply by replacing each
,
. . .
is
reduced to the
of the coefficients
/ by
its
negative.
is
Of especial interest
the transformation of equations in which both
members
7-13
If
are functions of the

<J>(a,-i,
same
variables.
,
.T 2 ,
Xn)
and
ty(x 1} x 2
...
x n ) be any two functions of the
same
variables, then
is
if
equivalent to F(XI, x 2
xn)
0,
where
<,
A A A
i,
2,
3,
etc.,
be the coefficients in
and
F is a function such B B B etc., be the

1} 2, s,
that
coef
ficients of the
corresponding terms in ty, then the coefficients of the corre sponding terms in F will be (Ai -Bi + -A l #0, (A 2 -B 2 + -A 2 B 2 ), (A 3 -J5 3 + -.4 3 # 3 ), etc.
145
By
By
7-
1,
-<
$ = ^
is
equivalent to ($ x-\F) +
of the
(-<!>
x ^)
0.
<
6-41,
and -^ are functions
same variables as
and ^.
be a
Hence, by 6-43, $
x-^ and -$
(i-42,
will
each be functions of these

ty)
same
variables,
and by
-^,
(3>
x-^) + (-$ x
will
also
function of these same variables.
Hence
3>,
^,
-<,
<J>
x-^, -$ x ^, and ($ x-^) +
(-<
x ^) are
all
functions of the same variables and, by 6-15, will not differ except in the coefficients of the terms.
If Ate
be any coefficient in
<,
and
Bk
the corresponding coefficient

coefficient in
will
then by 6-41, the corresponding and the corresponding coefficient in -^

in ^,
-$
will
be
-A k
be
.
be
-Bk
Hence, by 6-43, the corresponding coefficient x Ak -B k and the corresponding coefficient in

,
-3>
in
$x-^
be
will
fy will
(<
-A k Bk
(-<
will
Hence, by 0-42, the corresponding be A k -B k + -A k B k

.
>^)
coefficient in
x-^) +
x ^)
Thus ($ x-^) + (-$ x

the theorem holds.
is
the function F, as described above, and
By
in the
7-
1,
for every equation in the algebra there
is
an equivalent equation
be
form in which one member is 0, and by 7 13 the reduction can usually made by inspection. One of the most important additions to the general methods of the
s
algebra which has become current since the publication of Schroder

is
work
Poretsky
Law
of
Forms.
By
t
this law, given
any equation, an equiva
lent equation of
which one member
may
be chosen at will can be derived.

t.
7-15
is
equivalent to
a -t + -a
If a
And
Since
if t
- 0, a-t + -at = Qt+l-t = = a -t + -a then [7-1]

t,
t.
(a -t
t
+ -a
t)
-t
+ (a
+ -a
-t)
a -t + a
may
:
here be any function in the algebra, this proves that every
equation has an unlimited number of equivalents. of the law is
The more
general form
7-16
b is
1]
equivalent to
a
(a b
+ -a -6)
+ (a -b + -a
b) -t.
[7
b is equivalent to a -b
+ -a b
0.
And
[6-4] -(a -b
+ -ab) = ab
+ -a -b.
Hence
[7-15]
Q.E.D.
The number
9
of equations equivalent to a given equation

fondamentales de la theorie des egaliies logiques, Chap.
and expressible
i.
See Sept
11
lois
146
in
A
will
Survey of Symbolic Logic be half the number of distinct functions which

n
terms of n elements
~ can be formed from n elements and their negatives, that is, 2 /2. The sixteen distinct functions expressible in terms of two elements,
a and
b,
are:
b,
f
a, -a,
-b,
(i.
e.,
a -a, b -b, a + -6, -a +
etc.), 1
6,
(i.
e.,
a + -a, b + -b, etc.), a
b, b.
-^
_a i
_ rt
-^
a +
6,
-a +
-b,
ab
+ -a -b, and a -b + -a
In terms of these, the eight equivalent forms of the equation a

a
=
b are:
-=-&;
=
a -6 + -a &
a 6 + -a -6
b.
ab =
a + b;
a -b
-ab; -a -b
-a + -b: and a + -6
-a +
Each
of the sixteen functions here appears
on one or the other
side of
an
equation, and none appears twice. For any equation, there is such a
set of equivalents in
terms of the
elements which appear in the given equation. And every such set has = a -b + -a b) what may be called its "zero member" (in the above,
and and and
its
"whole
member"
(in
the above,
a &+
--&).
If
we observe
member"
the form of 7-10,

"whole
t
we
shall note that the functions in the
"zero
member"
are the functions in terms of which the arbitrarily
chosen
is
is
determined.
Any
the
which contains the function

.
=
{
0}
contained in the function
=
{
1
t,
The
validity of the
i.
law depends
It is
simply upon the fact that, for any
Octfcl,
e.,
1-/ + 0--/.
rather surprising that a principle so simple can yield a law so powerful.
Solution of Equations in
sible
One Unknown.
Every equation which
is
pos
according to the laws of the system has a solution for each of the
un
to
knowns involved.
equations
in
This
is
a peculiarity of the algebra.
We
turn
first
one unknown.
Every equation
in x,
.
if it
be possible in the
algebra, has a solution in terms of the relation c
7-2
x +
-x
is
equivalent to
c x c -A.
[5-72]
x +
-x
is
Ax
and
B -x = 0. [4-9] B -x = And A x =
7-21
is
equivalent to
ex.
is
equivalent to x -(-A)
A"
0,
hence to x c-A.
solution in the form // c x c

is
is
indeterminate whenever the equa

-.T.
tion which gives the solution

First,
if
symmetrical with respect to x and
the equation be of the form

is,
x+
-x
0.
The But
solution then
if
A
1.
x+
-x
A ex c-A. = 0, then A A
--=
(x
+ -x)
= A
x+
-x
0,
and
-A =
147
Hence the
satisfied
solution
is
equivalent to
x.
Oc.rcl, which [5-61-63]
is
by every value of
In general, any equation symmetrical with respect to x and -x which gives the solution, // c x c K, will give also // c -x c K. c x and // c -x, then [4 9] // x = But if and // -x = H.
Hence
[1-62] II
0.
And =
if
xcK
and -xcK, then
[5-33] x +
-xcK, and
1.
[4-8, 5-63]
I.
Hence
H ex c K will
be equivalent to
if
c a: c
It follows directly
from 7-21 that
neither x nor -x appear in an equa
tion, then although they
may
be introduced by expansion of the functions
involved, the equation remains indeterminate with respect to x.
7-22
ever
An A =
equation of the form
x +
-x
determines x uniquely when
-B,
B = -A. [3-22] A = -B
is
and
-.1
= B
are equivalent;
hence either of
these conditions
[7
21
equivalent to both.
is
it is
x+
if
Hence
B -x = B = -A,
[2-2] to
equivalent to
equivalent
B c x c -A. to B ex cB and
x+
to -.1 c.r c-.4,
and hence
= B = -A.
In general, an equation of the form
-x
=
is
determines x be
tween the
if,
limits
B and
-A.
Obviously, the solution

is
unique
if,
and only
these limits coincide;
and the solution
wholly indeterminate only
when they are respectively and 1, the limiting values of variables generally. 7-221 The condition that an equation of the form A x + B -x = be pos
sible in the algebra,
and hence that
its
solution be possible,
[5-65]
if
is
AB =
0,
0.
By 6-3, A B A B = 0. Hence if A B
=(=
cAx + B-x.
0,
Hence
x +
B -x =
then
then
x+
B -x =
must be
false for all values
of x.
And A x + B -x = A B = is called the

+
is
and the solution B ex

"
equation of
c-A are equivalent. condition" of A x + B -x = 0:

To
call it
it is
a necessary, not a sufficient condition.
the condition that
B -x =
have a solution seems inappropriate:
the solution
Bc.xc.-A
false, or
equivalent to
x +
-x
=
will
0,
whether
x +
-x
be true,
impossible.
solution of
The
sense in which
AB =
made
conditions other forms of the
x+
B -x =
be
clear in
what
follows.
The equation
of condition
is
frequently useful in simplifying the solution.
148
(In this connection,
A
it
should be borne in mind that

if
AB =
follows
from
x +
-x
0.)
For example,
a b x + (a +
6)
-x
=
of condition
is
then (a +
b) c.r
c-(a
b).
But the equation
a b (a + b)
= ab =
0,
is
or,
-(a
b)
Hence the second

solution
half of the solution
indeterminate, and the complete
may
be written a+ b
ex
is
However,
this simplified
form of the solution
equivalent to the original

is
equation only on the assumption that the equation of condition
satisfied
and a
0.
Again suppose
ax
+ b -x +
=
we have
Expanding
with reference to
(a
x,
and
(b
collecting coefficients,
c)
x +
c)
-x
and the equation

(a
of condition is
+ c)(b +
c)
a b + a
b
+b
= ab + =
The
solution
is
c x c -a -c
c
0,
But, by 5-72, the equation of condition gives
and hence -c =
1.
Hence the complete solution may be written

b
c x c -a
is
But here
again, the solution b
ex c-a
equivalent to the original equation
only on the assumption, contained in the equation of condition, that c = 0. This example may also serve to illustrate the fact that in any equation
is 0, any terms which do not involve x or -x may be dropped without affecting the solution for x. If a x + b-x + c = 0, then by 5-72, a x + b -x = 0, and any addition to the solution by retaining c will
one member of which
be indeterminate.
All
terms which involve neither the unknown nor

constituent"
its
negative belong to the "symmetrical
of the equation
to be
explained shortly.
Poretsky
is
Law
-x
of
Forms
gives immediately a determination of x which
equivalent to the given equation,
whether that equation involve x or not.
7 23
x+
B
=
=
x +
is
equivalent to x
= -A
x +
-x.
[7-15]
-x
is
equivalent to
(A x +
-x) -x + (-A x +
-B
-r)
r- B
-x +
-Ax
149
of Jevons. 10
This form of solution
is
also the one given
by the method
is
Although
it
is
mathematically objectionable that the expression which

-.r,
gives the value of x should involve x and

logically simple
this
in reality a useful
and
form
is
of the solution.
It follows
from 7-2 and 7-23 that
= -A
x +
B -x
equivalent to
B
is
cx c
-.1.
Many
writers on the subject have preferred the form of solution in
which the value of the unknown
given in terms of the coefficients and an
undetermined (arbitrary) parameter.

form.
This
is
the most
"mathematical"
7-24
+
If
AE =
is
0,
as
the equation
x+
B -x =
= B
requires,
then
Ax
is
-x
=
-x
satisfied
by x
= B -u + -A
is
u, or x
+ u -A, where u
arbitrary.
x +
B
.
And this = there
solution
is
some value
complete because, for any x such that of u such that x = B -u + -A u = B
+u
-A
(a)
By
if
6-4,
if
-u + -A
?/,
u,
then -x
= -B -u + A
u.
Hence
x +
B -x
if
= B -u + -A then = A (B -u + -B u) + B
= B -u
(-B -u +
A u) = AB
-?/
AB
= A
Hence
of u,
x+
A B = and B -x = 0.
.r
-A
u,
then whatever the value
Suppose x known and such that A x + Then if x = B -u + -A u, we have, by 7-1,

(b)
-x
0.
(B -u +
-A
if)
-x + (-B -u +
n) x
= (A
The condition that
x +
-A
-x) u + (B -x +
-B
is,
x)
-u =
this equation hold for
some value
x + -.1
of
by 7 221,
(A x + -A -x) (B -x + -B x)
This condition
is
= A -B
-x
satisfied
if
A
(c)
(B + -B) x + (A + -A)
A x + B -x = 0, for then B -.r = AB + A -B x + -A B -x =

-x
and by 5-72,
If If
A -B x + -A B
0,
0.
AB =
then
A B = 0, then A B u Hence B -u + -A u = B -u + -A (B + -B) u + A B u = B -u + (A + -A) B u + -.1 -B u = B (-u + u)

But
10
B -u + -A = 0.
= B
u-A,
for:
-A -B u = B + -A -B
u.
[5-85]
B + -A -Bu =
B+n -A.
See above, p. 77.
150
Only the simpler form
of this solution, x
= B+
u -A,
will
be used hereafter.
The above
solution can also be verified
by substituting the value given
for x in the original equation.
We
then have
A
And
if
(B -u + -A
0,
?/)
(-B -u +
is
Au) = A B-u + ABu = A B

the same as
AB =
the solution
verified for every value of u.
That the
solution, x
= B-u + -Au = B + u-A, means

if
B ex c-A,
parameter,
If If
will
u, is
be clear
we
reflect that the significance of the arbitrary
to determine the limits of the expression.
u =
B -u + -A u = B + u -A = B. = 1, B-u + -Au = -A and B + u-A = B + -A. But when

0,
AB
Hence x =
stated
0,
B + -A = -A B + -A =
-A.
fact,
B -u + -A u = B + u-A
of condition of the
by
B ex
1,
simply expresses the c-A, that the limits of x are B and -A.
otherwise
The equation
and the solution
for equations of the
form
C x + D -x =
7 25
and
form
x+
B -x = C x + D -x,
1 is
follow readily
from the above.
The equation
of condition that
C x + D -x =
C.
C+D =
1,
and the
solution of
Cx
D -x
6-3,
1 is
-D c x c
(a)
By
if
Cx + D-xcC + D.
any value
1.
==
Hence
there be
of x for
which C x +
D -x =
0,
1,
then
necessarily
If
C+
D =
Cx + D-x (6) -DcxcC.

7-26
x
If
1,
then [6-4]
-Cx + -D-x =
1
and
[7-2]
C+
D =
1,
then the equation

is
Cx + D-x =
equivalent to
is
satisfied
by
= -D
uC, where u
Since [6-4]
arbitrary.
1
is
Cx + D-x =
is
-Cx + -D-x =
0,
and C +
D =
equivalent to
-C -D =
0,
the theorem follows
from 7-24.
7-27
If
x+
B -x = C x
B -D
-x, the equation of condition
is
(A -C +
-A C)(B -D
-BD) =0
-C, or
and the
solution
is
-B
cx c
A C + -A
+
+
= B -D +-B D By
7 13,
+ u (A
C + -A
-C), where
is
arbitrary.
x +
-x
= Cx
-x
is
equivalent to
(A-C + -AC)x
(B-D
-Bd-x =
0.
151
Hence, by 7-221, the equation of condition And by 7-2 and 7-24, the solution is
is
as given above.
B -D + -B D c x c - (A -C + -A C) x = B -D +-B D + u--(A -C + -A C), where u And [6-4] -(A -C + -AC) = A C + -A -C.

,
or
is
arbitrary.
The
subject of simultaneous equations
is
very simple, although the

is
clearest notation
we have been
able to devise
in
somewhat cumbersome.
7-3
The condition that n equations
one unknown,
x+
-x
0,
is
A*x +
-x
0,
x+
-x
h
0,
may
)
be regarded as simultaneous,
the condition that
(A
fi,
Bk =
is
And
the solution which they give, on that condition,
^B
k
cxc
is
H-A
k
or x
^B
k
u>
II -^* where u
k
arbitrary.
By
6-42
and
5-72,
x+
5
1
-x
0,
A*x +
-x
0,
x+
-x
0,
are together equivalent to
1 (A +
A n )x +
,
(B + B* +
B")
-x
or
k
z +
k
-*
=
is
By
7 23, the equation of condition here
A* x
A:
B =
k
A
But
ZM* x Z
A-
fc
=
1
(.I
+
. .
,I
+...+.4
= A B
1
+
l
B- +
+
.
.
A Bn + A B
1
+ A*
+
A"
A*B
+ A*
+ A* B- +
B B
+ +
A B + An B
+
2
Z(^*).
h,k
And by 7-2 and
7-24, the solution here
is
or
*=
A-
A-
And by
It
5-95,
-{^M
k
==
II ~A
k
k
.
may
be noted that from the solution in this equation,
?i
partial solu
152
tions of the form
A
B
h
c.r
c-A>
can be derived, for
Wc^B*
k
2 Similarly, 2
"
and
H -A*
k
c-A>
partial solutions
can be derived by taking selections of
members
of
k ^Z H
an d II
-A k
Symmetrical and Unsymmetrical Constituents of Equations. Some of the most important properties of equations of the form A x + B -x = are
by dividing the equation into two constituents the most comprehensive constituent which is symmetrical with respect to .r and -x, and a completely unsymmetrical constituent. For brevity, these may
clear
made
be called simply the

constituent
is
".
"symmetrical constituent"
and the
unsymmetrical
In order to get the symmetrical constituent complete, it necessary to expand each term with reference to every element in the
function, coefficients included.
Thus
in .1
x+
-x
it is
necessary to
expand the
first
term with respect to B, and the second with respect to A.
A
By
(B + -B) x + (A + -A)
5 72, this
.1
is
-x
= A Bx
AB -x + A -B x + -A B -x
.1
equivalent to the two equations,

-.r)
is
(x
.1
B =
and
-B
x + -.4
-x
is
=
the
The
first of
these
the symmetrical constituent; the second
unsym
metrical constituent.
The symmetrical
constituent will always be the equa
tion of condition, while the unsymmetrical constituent will give the solution. But the form of the solution will most frequently be simplified con
by
sidering the symmetrical constituent also.
The unsymmetrical
is
constituent
will always be such that its equation of condition the equation of condition of
satisfied a priori.
Thus
A -B
is (.1
x +
-A B -x =
-B)(-A B) = 0, which is an identity. By this method of considering symmetrical and unsymmetrical con stituents, equations which are indeterminate reveal that fact by having no unsymmetrical constituent for the solution. Also, the method enables
any term
us to treat even complicated equations by inspection. Remembering that in which neither x nor -x appears belongs to the symmetrical
constituent, as does also the product of the coefficients of x
and
-.1-,
the
separation can be
made
(c
directly.
For example,
+ x) d + -c -d + (-a + -x) b

will
153
have as
its
equation of condition
c
d + -c -d + -a
and the solution
will
be
b
c .r c -d
symmetrical constituent
is
Also, as
we
shall see shortly, the
always the
complete resultant of the elimination of x. one
The method does not readily apply to equations which do not have member 0. But these can always be reduced to that form. Plow it
extends to equations in more than one
unknown
will
be clear from the
treatment of such equations.

Eliminations.
The problem
of elimination
is
the problem, what equa
tions not involving x or -x can be derived from a given equation, or equa

tions,
which do involve x and
-.r.
In most algebras, one term can, under
favorable circumstances, be eliminated from two equations, two terms from three, n terms from n 1 equations. But in this algebra any number
terms (and their negatives) can be eliminated from a single equation; and the terms to be eliminated may be chosen at will. The principles whereby such eliminations are performed have already been provided in
of
theorems concerning the equation of condition.

7-4
A B =
By
contains
all
the equations not involving
.r
or
-.r
which can
be derived from
x+
-x
0.
7-24, the complete solution of
x+
B -.r =
is
= B
-?/
-A u
we have
Substituting this value of x in the equation,

.1
(B -u +
-A
11)
B
is
(-B -u +
u)
= A B -u
AB
= A B =
.r.
Hence
It is at
AB =
the complete resultant of the elimination of
once clear that the resultant of the elimination of x coincides
with the equation of condition for solution and with the symmetrical con
stituent of the equation.
7-41
If
,
n elements,
,T 3 ,
.
. .
x\,
0,
.r 2 ,
.r 3
xn
be eliminated from any equation,

is
F(XI, # 2
xn )
the complete resultant

in
the equation to
of the
product of the coefficients

(a)
F(x it x 2
.r 3
x n ).
is
By
If
6-1 and 7-4, the theorem
true for the elimination of
one element,
(b)
= 0. .r, from any equation, /(.r) the theorem hold for the elimination
of k elements, x\,
X2
154 xk
A
. . .
from any equation,
$(.TI,
x Zt
...
.TI,
xk )
.T L
>,
=
.
.
0,
.
then
,
it will
hold
for the elimination of k
elements,
x k Xk+i, from any
equation, V(xi, x z
x ki xk +i)
...
,
0, for:
/(.ri,
By (Ml,
V(xi,
.1-2,
xk x k+ i) =
xz
x k )-x k+i xz
,
.
+f
(xi,
x k )--Xk+i.
And
the coefficients in
will
be the coefficients in / and /
By
7-4, the complete resultant of eliminating x k +i

/(.Ti, .To,
is
.
.
from
i
X k ) Xk+i + /
. .
.
(Xi,
XZ
Xk)~X k+ =
a *)
f(xi,
.T 2 ,
x k ) x/ Oi,
.
.r 2 ,
=
=
is
And by
to
<(.TI,
6-43,
2*2.,
/(.TI,
a;*;)
.T 2 ,
x k ) X/
(.TI, or 2 ,
a**)
equivalent
if
...
0,
2,
where $
3,
is
a function such that
the
coefficients
in/ be PI,
P P
etc.,
and the corresponding
coefficients
<
in/
will then the corresponding coefficients in be PiQi, P 2 2, PsQa, etc. Hence if the theorem hold for the elimina x k ) = 0, this tion of k elements, .TI, x Z) ... x k from $(.TI, x z
be Qi,
Qz Q
,
s,
etc.,
elimination will give

(P.Qi)(ft<2.)(PiC.).
.
(f iftftis
.ddQ..
0,
where PiP^Pzor in /
-QiQzQsi.
the product of the coefficients in

SF.
.TI,
3>,
and /
if
e.,
the product of the coefficients in
Hence
.
.
the theorem hold for the elimination of k elements,

<(.TI,
xz
x k from
,
xz
xk )
,
0, it will
hold for the elimination of

^(.TI,
elements,
i
Xi,
xz
...
x k x k+ i, from
xz
x k x k+ i)
,
0,
provided x k+ be the first eliminated. But since the order of terms in a function
is
immaterial, and for
any order of elements in the argument of a function, there is a normal form of the function, x k+ in the above may be any of the k + 1 elements in \F, and the order of elimination is immaterial.
i
(c)
Since (a) and
(b)
hold, therefore the
theorem holds
for the of
elimination of any
number
of elements from the equation to
any function
of these elements.
it is
By this theorem,
possible to eliminate simultaneously
elements from any equation, by the following procedure:

equation to the form in which one
(2) Develop the other
any number of (1) Reduce the
form;
member is 0, unless it member of the equation

(3)
already have that

as a normal-form
function of the elements to be eliminated;
Equate to
the product of
the coefficients in this function.

resultant.
This
will
be the complete elimination

Occasionally
it is
155
convenient to have the elimination resultant in the

1,
form of an equation with one member gives the resultant have that form.
7-42
especially
if
the equation which
The complete
resultant of eliminating
,
n elements,
x\
x2
...
xn
from any equation, F(XI, x 2 of the coefficients in F(XI, x 2

Let AI,
...
...
etc.,
xn) x n ).
==
1,
is
the equation to
of the
sum
x n ).
2,
A
=
3,
be the coefficients in F(XI, x 2

Xz, ...
F(XI, xz, ... x n )
1 is
,
equivalent to -[F(xi,
x n )]
,
0.
And
$
by 6-41, -[F(xi, x 2
that
will
if
...
x n )]
is
a function, $(xi, x 2
...
x n ), such
any
-A/c.
coefficient in
be AU, the corresponding coefficient in
be
Hence, by 7-41, the complete resultant of eliminating x\ x 2 from F(XI, Xz, ... x n ) = 1 is
9
...
xn
U-A
But
[5-95] -{
=0,
=--
or
-{ 11-^1
-=
1
11-4}
Z A.
Hence Q.E.D.
For purposes
tion
is
of application of the algebra to ordinary reasoning, elimina
a process more important than solution, since most processes of

"middle"
reasoning take place through the elimination of
terms.
For
example
If all b is x,
c x,
-x
ax
and no a
then
is x,
= =
0,
ax
+ b -x
0.
Whence, by elimination, a
0,
or no a
is b.
The complete solu unknown may be accomplished by tion of any equation in more than one eliminating all the unknowns except one and solving for that one, repeating
Solution of Equations in
more than one Unknown.
the process for each of the unknowns.
Such solution
will
be complete
because the elimination, in each case, will give the complete resultant which is independent of the unknowns eliminated, and each solution will be a
unknown, and complete, by previous theorems. How ever, general formulae of the solution of any equation in n unknowns, for each of the unknowns, can be proved.
solution for one
any equation in n unknowns is identical with the resultant of the elimination of all the unknowns; and this resultant is the condition of the solution with respect to each of the unknowns sepa
7 5
The equation
of condition of
rately.
(a)
If the
equation in n unknowns be of the form

z,
.
xn )
0:
156

...
Let the coefficients in F(XI, x 2
x n ) be AI,
2,
3,
etc.
Then,
by 6 -32,
II A c F(xi, x 2
and
[5-65]
...
xn)
JJ
A =
=
0*2,
is
a condition of the possibility of
F(x lt
0-2,
z) Xi,
And
.
[7-41]
,
UA
.
is
.
the resultant of the elimination of

.
x2
x n from F(x\,
(6)
xn)
0.
If
.
the equation in n
.
unknowns have some other form than

it
F(XI, x 2)
of that
Xn)
0,
then by 7-1,
has an equivalent which

its
is
form, and
its
equation of condition and
elimination
resultant are the equivalents of the
equation
of
condition and
elimination resultant of
its
equivalent which has the form

,
F(x lf X 2
(c)
...
X n)
=0
the
The
result of the elimination of all
unknowns
is
the
,
equation of condition
because
(1)
:
with respect to any one of them, say x k
ing
all
(2)
The equation to be solved for Xk will be the result of eliminat the unknowns but Xk from the original equation; and The condition that this equation, in which Xk is the only
is,
unknown, have a solution for Xk result of eliminating Xk from it.
by
(a)
and
(6),
the same as the
Hence the equation

other
of condition with respect to Xk is the

first
same as
all
the result of eliminating, from the original equation,
the
unknowns and then x^

(6),
And by 7-41 and

independent
the result of eliminating the unknowns
is
of the order in
it
which they are eliminated.

be unnecessary to investigate separately
Since this theorem holds,
will
the equation of condition for the various forms of equations;
they are
already given in the
theorems concerning elimination.

,
. . .
7-51
Any
its
equation in n unknowns, of the form F(XI, x 2

as follows: Let x k be
x n)
0,
provided
of the
equation of condition be satisfied, gives a solution for
each
unknowns
and
any one
of the
unknowns;
,
. . .
let P],
2,
Pa, etc., be the coefficients of those terms in F(XI, x 2

is
x n ) in which Xk
positive,
Qi,
Q2 Q
,
etc.,
the coefficients of those terms in which Xk

is
is
negative.
The
solution then
II Q ex* c
2 -P,
or
Xk
= II Q
+ u-
2 -P,
where u
is
arbitrary.
157
(a)
By (Ml,
to
for
some / and some /

.r 2
,
. . .
F(x\,
.
.
.TO,
.
x, n )
=
--=
is
equivalent
/(.ri,
.r n _i)-.r n
+/
3,
(.ri, .T 2 ,
.r n _i)--.r n
0.
Let the coefficients
in
/ be PI,
P P
2,
etc., in
be î,
z,
Qs, etc.
Then
in
PI,
P
Xk
is
2,
3,
etc., will
be the coefficients of those terms in
F
in
which
Xk
is
positive, Qi, Q*,
Q3
the coefficients of terms in
which
negative.
.T n _])-.r n
If /(.TI,
.TI, .TO,
.
.T 2 ,
be regarded as a function of the variables,

,
n _i, its
.r 2
,
. .
coefficients will be Pi.r n

.
2 .r n ,
3 .T n
etc.
.TI,
o* 2 ,
And
. .
.
r n -i) -x n be regarded as a function of / Oi, x n -i, its coefficients will be Q -x n Q 2 -x n (? 3 -x n etc.

if
1
,
Hence, by 6-42,
/(.?!,
X2
X n -l) X n + / (1,
^(.TI,
.TO,
. .
.
.T 2 ,
Xn-i) -X n
=
a function in
is
equivalent to
x n -i)
0,
where
is
which the
+
coefficients
etc.
are
(PiX n + Qi-x n ),
is
(P 2 x n +
Q2
.
-.r n ),
(P 3 ^n
3-^ n ),
^(iCi,
And
T2
...
.r n _i)
equivalent to F(XI,
.r 2 ,
xn)
,
. .
0.
.r n _i 7-41, the complete resultant of the elimination of 0*1, .r 2 = will be the equation to of the product x n -i) from ty(xi, x 2
By
of its coefficients,
But any expression

tion of x n
.
of the
form
P nr n + Q
-x n
is
a normal form func
Hence, by 6 43,
By
7-2 and 7-24, the solution of
H P^n+
Qr -^
is
And
for
[5-951]
-{II P]
----
T.-Pis
(b)
Since the order of terms in a function
immaterial, and
is a any order of the variables in the argument normal form of the function, x n in the above may be any one of the .r n _ ) and / (.n, .r 2 .T n ), and /(.TI, .i 2 variables in F(x^ .T 2
of a function there
.T n
_0 each some function
of the
Therefore, the theorem
holds for
1 variables. remaining n one of the variables, x k any

.
That a
is
single equation gives a solution for
any number
of
unknowns
another peculiarity of the algebra, due to the fact that from a single equation any number of unknowns may be eliminated.
158
As an example of the last theorem, we give the solution of the exemplar equation in two unknowns, first directly from the theorem, then by elimina
tion
and solution
for
(1)
x y+
-x y +
each unknown separately. C x -y + D -x -y = has the equation of condition,
ABCD
Provided this be
satisfied,
the solutions for x and y are

or
BDcxc-A+-C,
CDcyc-A+-B,
(2)
x
y
= BD
+ u (-A + -C)
or
= CD
is
+u
(-A+-B)
x y+
-x y +
(a)
C x -y + D -x -y =
equivalent to
-x) -y
-.r
(Ax
B -x) y+(Cx + D
=
=
and
to (b)
(a),
(A y + C -y) x + (B y +
D -y)
Eliminating y from
we have
+
(Ax
The equation
B-x)(Cx
D-x) =
ACx + BD-x
is,
of condition
with respect to x
then,
(AC)(BD) =
And
the solution for x
C),
is
ABCD
C).
BDcxc-(A
or
(6),
BD + u--(A
And -(A
C)
= -A
-C
Eliminating x from
we have
y+
is,
(Ay+C-y)(By + D-y) = A B
The equation
of condition with respect to y
is
C D -y =
then,
BCD
0.
And
the solution for y
CDcyc-(AB),
Another method
or
CD + v-(AB).
-y
in
And -(A B) = -A+-B

two unknowns, x and y, coefficients, with x and u
this value
of solution for equations in
would be to solve
for y
and
for
terms of the
as undetermined parameters, then eliminate y

of
it
by substituting
in the original equation,
and solve
for x.
By
a similar substitution,
then be eliminated and the resulting equation solved for This y. method may inspire more confidence on the part of those unfamiliar with
may
this algebra, since it is a general algebraic
method, except that
in other
algebras
more than one equation

solution of
is
required.
The
y
x y+
-x y +
(C x +
D -x)
+ u--(A x +
C x -y + D -x -y = for y is B -x) = (C + u -A) x +(D + u -B) -x
159
The
solution for -y
is
-y
= (A
x +
B -x)
v-(C x
D -x) =
(A +
-C) x + (B +
-D) -x
Substituting these values for y and -y in the original equation,
(A x +
B
=
-x)[(C + u -A) x + (D + n -B) -x] + (Cx + D -x)[(A +
(C + u -A) x +
7?
=
Hence
AC x + D -x
B (D = 0.
+ u -B) -x +
C (A +
-C) x + (B + v -D) -x\ -C) x + 1) (B + v -D) -x

v
B D ex c-A
it is
+ -C.
to equations in
Theoretically, this
method can be extended
any number
of
unknowns:
7 52
practically,
too cumbersome and tedious to be used at
all.
Any
equation in n unknowns, of the form

F(XI, x 2
,
...
xn)
f(xi,
x2
...
xn)
gives a solution for each of the

of the
unknowns
Qi, Qi,
3,
as follows:
Q.a,
Let x k be any one
unknowns;
If 2
,
let PI,
.
. .
P P
2,
3,
...
be the coefficients in F,
and
J/i,
3/3,
AI,
A2 A
,
... the coefficients of the
terms
in /, so that
and
1/V are coefficients of
corresponding terms in which x k is positive,

is
and Q r and
are coefficients of terms in which Xk

is
negative.
The
solu
tion for Xk then
II
or
Xk
(Qr
N
(Qr
+ -Q,
r)
Xk
C
+ U
r
(P r
M
^
+ ~P, - J/ r )
= II
r
-N
i,
+ -Qr
...
r)
(Pr
By
i,
7-13,
-I
7 X.i
n)
;i 2 ,
.r n )
=
is
#2,
0,
where $
/Oi, 2 ... .r n ) is equivalent to a function such that if A r and l? r

,
be coefficients of any two corresponding terms in coefficient of the corresponding term in $ will be
F and /, then the A -B + -A B,.

r r
r
Hence, by 7-51, the solution
will
be
II
Or
(Qr
"N r
+ -Qr
r)
C Xk C
-(P r
-M
-P
3/ r )
= II
[0-4]
(Qr
-N
-Q r
r)
r)
+ H
-(P r - J/ + -P,
r
J/r)
And
7-53
-(P r -J/ r + -P J/
r
= (P
J/ r +
-P -J/
r
r ).
equations in n unknowns, each of the form be regarded as simultaneous, is as follows: xn) 0, may F(XI, x, x n ) = 0, Let the coefficients of the terms in F 1 in the equation F (x lf .r 2 1 1 let the coefficients of the correbe Pi 1 , Pa 1 P3 1 QJ, Qs Qi
The condition that

. . .
160
A
in
spending terms
F
Qs
2
,
in the
.
equation
(.ri, ,r 2 ,
.r n )
0,
be Pi 2
2 2
,
Pa
2
,
2
,
.
Q\~, Qz~,
Fm
z
in the
,
equation
F m (x
lt
the coefficients of the corresponding terms in m xz x n ) = 0, be PI W P 2 m P 3 W Qi Qzm

,
. .
....
The condition then

;
is
Or
if
*
r
be any coefficient, whether
or Q, in
P\
the condition
is
And
the solution which n such equations give, on this condition, for any
Xk, is
one of the unknowns,
as follows:
Let PI^, P/,
P
h
7i
...
0, in
be the coef
ficients of those terms, in
any one
,
of the equations
F =
is
which
x-k
is
positive,
and
let
Qi
h
,
Qi
h
3
,
...
be the coefficients of those terms, in

solution then
F =
h
0, in
which Xk
is
negative.
The
nizvic^czn-w]
r
ft
or
**
= ![ tZ
r
Qr
w-Z
r
III
h
-Pr"]
By
.
.
6-42,
equations in n unknowns, each of the form F(x\, x z

&(x-i,
x n)
xn)
0, 0,
are together equivalent to the single equation
xz
where each of the

1
,
coefficients in
2
<
is
the
sum
of the
l
F That is, if corresponding coefficients in F F F P r m be the coefficients of corresponding terms in P1 F 2

3
, ,
. .
P
.
P, Fm
2
,
then the coefficient of the corresponding term in
<
will
h
be
P, +
1
P
Q
r
2 r
+ P, .-,
or
2 Pr
h
and
1
,
if
2
,
Qr
.
Qr
2
,
.
.
m be the coefficients of corresponding terms in

<
F F
will
Fm
then the coefficient of the corresponding term in
be
h
The equation
of condition for
0,
<>
0,
and hence the condition that

is
F =
1
0,
F =
2
...
Fm =
may
be regarded as simultaneous,
that
is,
the equation to
of the product of the coefficients in $;
or
nizjvixiiizcr i-o h
r
And by
7-51, the solution of $(#1, T 2,

-
#n)
for Xk
is
II
S Qr
h
]
c TA c

or
161
%
5-95, -[
h
And by
7 54
P\ =
h
-P/.
The condition that
m
,
equations in n unknowns, each of the form

...
F(XI, x 2
x n)
f(xi,
x2
...
xn)
l
,
may
be regarded as simultaneous, is as follows: Let the coefficients in F l 1 in the equation F 1 =f be Px 1 P, Pj 1 and let the QJ, Q Q. in the equation F = f l be coefficients of the corresponding terms in /
1 1
,
, ,
,
. .
.
Mi MJ, Ms
1
,
1
,
.
. .
Ni As AY,
,
let
,
the coefficients of the corresponding
terms in
2
,
in the
equation
F =f
2
be P^,
2 3
,
.
.
and
.
.
let
2
,
the coefficients of the corresponding terms in /
QS, be
M M
2
i
, ,
.
Q 22 Q 32
,
the coefficients of the corresponding terms 2 , 3 Ni m = m P m P m m be m m W in the equation F Pi 2 3 Qi Qz $3 f m let the coefficients of the corresponding terms in / m be Mi m z T m m AT m .... The condition then is A
.
A A
2
let
Ms 2 in F m
,
.,
,
and
Ms m
A
h
II
r
H (Pr
-MS + ~Prh
MS)] X JJ
r
E
h
(Qr*
-N, +
-(? r
A/)]
=
represent
is
or
if
h
r
represent any coefficient in

h
,
Fh
+
whether
P
=
or Q,
and
h
r
the corresponding coefficient in /
whether
M or A, the condition
B
h
r
II
IZ
m
(Ar
-BS
-A
h
r
)}
And
the solution which
one of the unknowns,
.TA-,
such equations give, on this condition, for any h h be the coefficients is as follows: Let P r and r
of those terms, in any one of the equations
Fh = f h
in
which a^
is
positive,
and
is
let
h
r
and
h
r
be the coefficients of the terms, in

is
F = fh
h
in
which x k
negative.
The
solution then
or
a*
- II
By
Z
.
(Qr
-A7/ +
-9,"
AV)] +
.
E [II
a? 2 ,
7-13, ^(.TI,
. .
#2,
0,
ar n )
= / A (-TI
is
...
n)
is
equivalent to
if
h Q and
r
i, .r 2 ,
xn)
where
a function such that
h
r
be coefficients of corresponding terms in of the corresponding term in * will be
Ph
r
and Jf r A be
h
coefficients of
and / h the coefficient * h Q r -Nr + -Q r AV, and if h h the corresponding terms in F and /
,
Fh
coefficient of the corresponding
term
h
r
in
h
r
^ will
+
be
P h -M h + -P h
r r
r r
h
r
.
And -(P r -MS

12
-P
h
r
h
r )
= P
-P/
-M \
Hence the theorem
follows from 7 53.
162 x n)
A
F(XI,
.1*2,
. .
f(xi,
xz
a- n )
is
a perfectly general equation, since
F
of
and / may be any expressions any number

and
this
in the algebra,
developed as functions of
the variables in question.
7-54 gives, then, the condition and the solution
of simultaneous equations, in
any number
is
of
unknowns,
for
each of the unknowns.

tion,
is its
This algebra particularly lends

It
itself to
generaliza
most general theorem.
the most general theorem
concerning solutions in the whole of mathematics.
Boole
General Problem.
Boole proposed the following as the general

11
problem of the algebra of logic. Given any equation connecting the symbols x, y, ... w, z, .... Re quired to determine the logical expression of any class expressed in any
way by
.
. .
the symbols x, express this:

t
y,
... in
t
terms of the remaining symbols u\
z,
....
z,
We may
)
;
Given
f(x, y,
z,
.)
and
is
$>(.r,
y,
.)
ty(w,
to determine
. .
.
in
.
terms of w,
.
....
This
perfectly general, since
if
is,
x, y,
and w, z, 7-1 and 5-72, a by
for solution
=
V(w,
2,
by any number of equations, there The rule single equation equivalent to them all. = f(x, y, .) and .) may be stated: Reduce both t y, to the form of equations with one member 0, combine them .)
.
are connected
3>(.r,
by addition into a single equation, eliminate x, 7-1, the form of equation with one member
form.
y,
is
and solve
for
t.
By
equivalent to the other
And by
5-72, the
sum
of
equivalent to the equations added.
two equations with one member is Hence the single equation resulting
will contain all the data.
is
from the process prescribed by our rule

result of eliminating will be the
The
complete resultant which
independent
of these,
of
t
and the solution

z,
. . .
for
will
thus be the most complete determination
in
terms of w,
afforded by the data.
Consequences
of
Equations in General.
word
of caution with refer
ence to the manipulation of equations in this algebra may not be out of As compared with other algebras, the algebra of logic gives more place.
room
for choice in this matter.
Further, in the most useful applications
of the algebra, there are frequently
problems of procedure which are not
resolved simply by eliminating this and solving for that.
The
choice of
method must,
then, be determined with reference to the end in view.
But
the following general rules are of service:

(1)
Get the completest possible expression

==
0,
or the least inclusive
possible expression
1.
a+b+
c+
gives
0, 6
0, c
0,
a+ 6
0,
a+
0,
of Thought, p. 140.

etc.
163
But a
will
. .
a+ b
1,
a+
==
not generally give a + b = 0, etc. Also, a == == 1 will not 1, but a + b generally give a == 1.
of equations, with
gives
(2)
Reduce any number

is
which
it is
necessary to deal,
by first reducing each to the form in which and then adding. The various constituent equations can always be recovered if that be desirable, and the single equation gives
to a single equivalent equation,
one
member
other derivatives also, besides being easier to manipulate.
Do
not forget
that
it is
possible so to
If
than the data.

or a + b
we have a =
combine equations that the result is less general and 6 = 0, we have also a = b, or a b = 0,
0,
according to the
mode
of combination.
But a +
is
equivalent to the data, while the other
two are
less
comprehensive.
A
fact,
general
desired
method by which consequences of a given equation, in any 12 and is, in terms, may be derived, was formulated by Poretsky,
a corollary of his
Law
of
Forms, given above.
We
have seen that
this
law
may
a-b + -ab =
be formulated as the principle that if a = b, and therefore and a b + -a -b = 1 then any t is such that a -b + -a b c t
,
and tcab + -a-b, or any

member" of this set.
"zero
the
which contains the
"zero
member"
"
of the set of equations equivalent to a
b,
t,
and
is
contained in the
whole
Now
Law
if
x c
t,
ux c
for
any u whatever, and thus the

arbitrarily
if
member" of
the
of
Forms may be multiplied by any

Similarly,
chosen u which we choose to introduce.
tcy, then
tcy + v,
by the
and the
"whole
member" in
the
Law
v.
of
Forms may be
increased
addition of any arbitrarily chosen
This gives the

t
Law
of Consequences.
7-6
If
b,
then
= (ab + -a
-b + u)
v (a
-b + -a
b) -/,
where u and
are arbitrary.
[7
1
12]
If a
b,
then a -b + -ab
+
v (a
= =
and ab + -a-b =
(1
1.
--
Hence
(a b +
-a -b +
all
it) t
-b + -a
b) -t
?/) t
+ v-0--t
t.
This law includes
First, let us see that it is
the possible consequences of the given equation. more general than the previous formulae of elimina
tion
for
t.
and
solution.
Given the equation
.r
B -x =
0,
and choosing
.1
we should
get the elimination resultant.

0,
ItAx + B -x =
then
AB
(-A x + -B -x + u)
AB
Since u and
fore
12
are both arbitrary
and
may assume
the value
0,
there
AB =
0.
Sept
lois, etc.,
Chap. xn.
164
A
this
is
But
only one of the unlimited expressions for
AB
which the law
gives.
Letting
u =
0,
and
1,
we have
x+
A B = A -B
Letting u
-A B -x
= A and
B,
we have
AB
And
so on.
AB + -AB-x
AB
be found that every one of the equivalents of which the law gives will be null.
it will
But
Choosing x for our

If
t,
we should
get the solution.
x+
-x
0,
then x
= =
(-A x + -B -x + u) x + (-A + u) x + v B -x.

0,
(A x +
B -x)
-x
Since u and v
may
both assume the value

x
= -A
x,
or
x c
-A
(1)
And
since
u and
may both assume the value 1, x = x + B -x, or B -x ex
But
if
B -x c x,
(1)
then
(2),
1,
Hence,
and
v
B -x = (B B cxc-A.
Law
x
of
-.T)
0,
or
cx
(2)
When u =
of
and
the
Consequences becomes simply the
Law
Forms.
For these values
in the above,
= -A
x+
B -x
in the
which
is
the form which Poretsky gives the solution for x.

of the arbitraries,
The introduction
all
u and
of
v,
Law
of
Consequences
it
extends the principle stated by the
Law
Forms
so that
all
covers not
only
equivalents of the given equation but also
the non-equivalent
As the explanation which precedes the proof suggests, this is accomplished by allowing the limits of the function equated to t to be expressed in all possible ways. If a = b, and therefore, by the Law of
inferences.
Forms,
t
(a b
+ -a -b)
(a
-6 + -a
6,
b) -t
the lower limit of

1,
t,
0, is
expressed as a -b + -a
-b.
and the upper
limit of
/,
is
expressed as
ab + -a
its
In the
Law
of Consequences, the lower

is,
limit, 0, is expressed as v
(a-b + -ab), that

in
can be derived from
expression as
is,
expressed as a b + -a -6 + u, that
ways which a-b + -ab; and the upper limit, 1, is all possible ways which can be derived
in all possible
from
its
expression as a b + -a -b.
t
Since an expression of the form

t
(a b
+ -a -b)
(a
-b + -a
b) -t

or of the form
t
165
(a b
+ -a -b + u)
v (a
-b + -a
b) -t
determines
only in the sense of thus expressing

all
its limits,
and the Law

it
of
Consequences covers
all
possible
ways
of expressing these limits,
covers
possible inferences
is,
from the given equation.
The number
of such
inferences
of course, unlimited.
The number
expressible in
terms of n
elements will be the number of derivatives from an equation with one
member and the other member expanded with reference to n elements. The number of constituent terms of this expanded member will be 2 n and the number of combinations formed from them will be 2 2 Therefore,
,
".
since pi + pz + p$ +
=0
number
of
of consequences of
p z = 0, p s = 0, etc., this is the a given equation which are expressible in terms

gives pi
0,
n elements.
Poretsky gives the sixteen determinations of a in terms of the three elements, a, b, and c, which can be derived from
As one
illustration of this law,
the premises of the syllogism in Barbara:

If all
13
is b,
a-b =
b -c
0, 0,
and
all 6 is c,
then a-b +
a
b -c
0,
and hence,
=
=
a
5 b
= a (-b + c) = a + 6 -c = a b = a (b c + -b -c) _ c + a (b c + _5 _ c = a c = b -c + a c = a (-b + c) + -a b -c = a b c -c + ab c = a (b c + -b -c) + -a b -c = a c + -a b -c = a 6 c + -a b -c

(b
+ -c)
(b
c)
)
The Inverse Problem

that
of
Consequences.
Just as the
Law
of
Conse
quences expresses any if a-b + -ab =
inference from a
0,
1
;
then (a-b + -ab)v
by taking advantage = 0, and if ab + -a-b =

any equation which
of the fact
1,
then a b + -a -b + u
the inference
fl( a 5
so the formula for
will give
+ _a a -b + - a
_fr)
= =
b
i}
can be expressed by taking advantage of the fact that if then ab + -a-b = 1, and if a-b + -ab + u = 0, then
thus get Poretsky
s
l>
o.
We
Law
of v
of Causes, or as
14
it
would
be better translated, the
Law
of Sufficient Conditions.
7-7
If for
some value
t
of
u and some value

t
(ab + -a -b)
+ (a -b + -a b + u)
-t,
then a
b.
[v
= v(ab + -a-b)t+(a-b + -ab + u)-t, = (a b + -a -b) t + (a -b + -a b + u) -t] -t

If
t (
then
[7-1, 5-72]
a -^ + - a
1)
+ u ) -t
= (a-b + -a
b) -t
+ u -i
13
14
Ibid., pp. Ibid.,
98 /. Chap. xxm.
166
A
Hence
(a
-b + -a
b) -t
t
=
+ -a -b)
t
t,
(1)
Hence
also [5-7]
v (a b
and
[4-9]
b)
t--[v (a b
+ -a -b)]
(a
=
b)
t
(-v
+ a -b + -a
t-v + (a -b + -ab)
Hence
[5
72]
-b + -a
(
=
(t
(2)
By
(1)
and
(2),
a -b + -a
b.
b)
-t)
a -6 + -a
6.
Hence
[7-1] a
Consequences and the Law of Sufficient Conditions are more general than the Law of Forms, which may be derived from either.
of
Both the Law
Important as are these contributions of Poretsky, the student must not

be misled into supposing that by their use any desired consequence or
equation can be found automatically. The only sense in which these laws give results automatically is the sense in
sufficient condition of a given
which they make it possible to exhaust the list of consequences or conditions And since this process is expressible in terms of a given set of elements.
ordinarily too lengthy for practical purposes, these laws are of assistance
principally for testing results suggested
"intuition
".
One has
to discover for himself
by some subsidiary method or by what values of the arbitrages
u and
v will
give the desired result.
V.
FUNDAMENTAL LAWS OF THE THEORY OF INEQUATIONS

.
The
In this algebra, the assertory or copulative relations are = and c denial of a = b may conveniently be symbolized in the customary way:
8-01
4= b is
equivalent to
"a
"a
b is false
".
Def.
We
or
"a
lent to a b
might use a symbol also for = a and to a -b = 0,

4= 0.
",
b is
false.".
But
since a
b is
its
negative
may
b"
be represented by a b
equiva 4= a
by a -b
c
It is less
b is false
since
"a
is
necessary to have a separate symbolism for not contained in is seldom met with in logic
in
true,
is
except where a and b are mutually exclusive,
which case a
then Q
is
=
",
0.
For every proposition of the form
"If
P
it.
is
true
there
is
Q is false, then P is false ad absurdum,oT the simplest form

"
another,
If
".
This
the principle of the rednctio
of
In terms of the relations
and
=|=,
the more important forms of this principle are:

"If
(1)
(2)
"If
= = =
b,
then then
c
6,
= gives = d&ndh =
d",
".
also,
k",
"If
c 4= d,
then a
=(=
<Z,
4= b
".
gives also,
"Ifc
then a
4=
b",
and
then
"If
4= k,
then a
b
4= b
(3)
"If
and
"If
c c
c 4=
d",
and
d,
then h
=
k",
gives also,
4= b
".
"If
and h
4= k,
d and h
4= k,
then a

(4)
"a
167
= =
b is equivalent to c
d"
gives also,
"a
4=
b is equivalent
toe
4=
d".
(5)
"a
"a
b is equivalent to the set, c
=
<7,
k,
".
..,"
gives also,
4= 6 is
equivalent to
Either
c 4=
</
or h
4= A
or ...
16
The
But the
from
general forms of these principles are themselves theorems of the

of propositions"
"calculus
the application of this algebra to propositions.
calculus of propositions, as an applied logic, cannot be derived
this algebra
without a
circle in
the proof, for the reasoning in
demon
stration of the theorems presupposes the logical laws of propositions at
every step.
We
must, then, regard these laws of the reductio ad absurdum,
like the principles of proof previously used, as given us
by ordinary logic, which mathematics generally presupposes. In later chapters, 16 we shall discuss another mode of developing mathematical logic the logistic
method
which avoids the paradox of assuming the principles of logic in order to prove them. For the present, our procedure may be viewed simply
as an application of the reductio ad
absurdum
in
ways
in
which any mathe
matician
feels free to
make
use of that principle.
Since the propositions concerning inequations follow immediately, for the most part, from those concerning equations, proof will ordinarily be
unnecessary.
Elementary Theorems.
sitions are as follows:
The more important
of the elementary propo
8-1
If
ac
4=
be, then a
4= b.
[2-1]
8-12
If
a+c
=t=
c,
then a
4= b.
[3-37]
8-13
4= b is
equivalent to -a
4"
-b.
[3-2]
8-14
a + b
4= b,
ab
4= a,
-a + b
4= 1,
and a -b
4=
are
all
equivalent.
[4-9]
8-15
If
a+b
[5-7]
x and b
4= x,
then a
4=
8-151
If
=
=
and
b 4= x,
then a + b
4= x.
[5-7]
8-16
If
ab
x and b
4=
.r,
then a
4=
1.
[5-71]
15
"Either
...
or
..." is
here to be interpreted as not excluding the possibility that

v.
both should be true.

1(5
Chap,
iv, Sect, vi,
and Chap.
168
A
If
8-161
and
b 4= x,
then a b
4= x.
[5-71]
8-17
If a
+ b
4=
and a =
and a
0,
then b
4= 0.
[5-72]
8-18
If
a b
4= 1
1,
then
b 4= 1.
[5-73]
8-17 allows us to drop null terms from any sum

a rule by which an equation and an inequation for example, a + b 4= and x = 0.
4= 0-
In
this, it gives
may
be combined.
Suppose,
a+ b
=
if
(a + b)(x
Hence ax + bx
But
x
[8
+ -x) =ax+bx+a-x+b -x. + a-x + b-x 4= 0.
0,
then a x
and
0.
Hence
8-2
If
17]
a-x +
-x
4= 0.
4= 0,
then a + b
4= 0.
[5-72]
8-21
If
4= 1,
then
064=!.
4=
[5-73]
8-22
If
a 6
4= 0,
then a
and
b 4= 0.
[1-5]
8-23
If
+ 6
[4-5]
4=
1,
then a
4= 1
and
b 4= 1.
8-24
If a b 4-
x and a
x,
then
b 4= x.
[1-2]
8-25
If a 4=
and a c6, then 64-0.
[1-9] If
ac6, then a
4=
a.
4= 0.
Hence Hence
8-26
a+ 6
4-
if
and a c
6,
then a 6
[8-22] 6
is
0.
"Either
equivalent to
4=
or 6
4=
".
[5-72]
8-261
as
4=
ai
+ a 2 + a3 +
".
4=
is
equivalent to
"Either
ax
4=
or a 2
4=
or
0, or ...
8-27
a 6
4= 1 is
equivalent to
"Either
4=
or 6
4= 1
".
[5-73]
8-271
3
aia 2
1
+1
=
is
equivalent to
"Either
ai 4= 1
or
a2
4= 1
or
or ...
".
The
tions
difference
b
between 8-26 and 8-27 and their analogues for equa

is
5-72 a +
equivalent to the pair, a
and
0,
and
The
5-73 a
b
169
1 is
equivalent to the pair, a
==
and
points to a neces
sary difference
inequations.
between the treatment of equations and the treatment of Two or more equations may always be combined into an
equivalent equation;
two
or
more inequations cannot be combined
into
an equivalent inequation. and b =|= 0. pair, a 4=
But, by 8-2,
a+
b 4=
is
a consequence of the
Equivalent Inequations of Different Forms.
The laws
of the equiva
lence of inequations follow immediately from their analogues for equations.
8-3
4= b is
equivalent to a -b + -a b 4
0.
[7-1]
8-31
4=
1 is
equivalent to -a
=h 0-
[7-12]
8-32
If
&(xi,
2,
...
x n ) and V(xi,
.r 2 ,
x n ) be any two functions of
the same variables, then

,
X2
.
Xn)
4=
is
equivalent to F(XI,
.T 2 ,
xn)
Ai,
4= 0?
where
3,
is
a function of these same
variables
f
and such
,
that,
if
z,
etc.,
be the coefficients in
$ and
Bi 2 3 etc., be the coefficients of the corresponding terms in ^, then the coefficients of the corresponding terms in F will be Ai-Bi + -AiBi,
,
A, -B 2 + -A z
B A -B
2,
3
+ -At B,, etc.
[7-13]
Poretsky
8-33
a
4=
Law
of
Forms
for inequations will be

t
is
equivalent to
3=
a -t+ -at.
[7-15]
Or
in
more general form

a
4= b is
8-34
equivalent to
=h
(ab + -a -b)
(a
-b + -a
b) -t.
[7-16]
Elimination.
The laws governing the
elimination of elements from an
inequation are not related to the corresponding laws governing equations
by the
reductio ad absurdum.
But these laws
follow from the
same theorems
concerning the limits of functions.
8-4
It
x +
B -x
4= 0,
then
[6-3\Ax + B -x
8-41
If
A + B + 0. cA+B. Hence
[8
25]
Q.E.D.
,
.
. .
the coefficients in anv function of n variables, F(XI, x z
x n ),
170
be Ci,
A
C
2,
To, etc.,
and
if
F(x 1}
.r 2 ,
.r n )
=f=
0,
then
ZC=NO
[0-32]
F(x lf
.r,
a- n )
cC.
Hence
[8-25]
Q.E.D.
Thus, to eliminate any number of elements from an inequation with one member 0, reduce the other member to the form of a normal function
of the elements to be eliminated.
The
elimination
is
then secured by
putting
=f=
the
sum
of the coefficients.
The form
of elimination resultants
It is
for inequations of other types follows
immediately from the above.
obvious that they will be analogous to the elimination resultants of equa tions as follows: To get the elimination resultant of any inequation, take
the elimination resultant of the corresponding equation and replace
= by
=f=
and x by +
A
a
is
is
universal proposition in logic

"
is
represented by an equation:
"All
by a -b
"
0,
No
"
is b
by
a b
0.
Since a particular proposition
always the contradictory of some universal, any particular proposition may be represented by an inequation: "Some a is by a b =f= 0, "Some
b"
by a -b =j= 0. The elimination of the "middle" term from an which represents the combination of two universal premises equation But elimina gives the equation which represents the universal conclusion.
is
not
b"
tion of terms from inequations does not represent

process.
an analogous
logical
Two
particulars
give
no conclusion:
a particular conclusion
is
requires one universal premise.
The drawing
of a particular conclusion
represented by a process which combines an equation with an inequation,
by 8-17, and then

All
simplifies the result,
by 8-22.
. .
For example,
c
aisb,
a
is c,
a-b =
a
c
=J=
0. 0.
a -b
a b
c
0.
Some
+ a -b
4= 0.
c =h 0.
abc
[8-17] [8-22]
Some
"
fr
is c.
6c4=0.
Solution
"
of
an Inequation.
of
An
inequation
may
be said to have a
solution in the sense that for
inequation one
member
4=
is
any inequation involving x an equivalent which is x can always be found.

4=
8-5
x +
-x
equivalent to x
-A
x +
-x.
[7-23]
8-51
i.
A
to
x+
-x
4=
is
equivalent to
or
"Either
-x
or
",
e.,
"Either
B ex
is false
xc-A
is false
".
[7-21

Neither of these
"
171
"
"solutions"
"
determines x even within limits.

x";
B ex
is
is
false
does not
x".
mean
"
is
excluded from
is
it
means only
is
false"
"
not
wholly within
Either
Bex
false
or
xc-A
does not
determine either an upper or a lower limit of x; and limits x only by ex cluding B + u-A from the range of its possible values. Thus "solutions"
of inequations are of small significance.
Consequences and
Sufficient Conditions of
an Inequation.
By
Poret-
sky s method, the formula for any consequence of a given inequation follows from the Law of Sufficient Conditions for equations. 17 If for some value
of
u and some value

t
of
v,
x (a b
+ -a -6)
+ (a -b + -a b + u) -t
then a 8-52
=
If
b.
Consequently, we have by the reductio ad absurdum:

4= b,
then
^= v (a b
+ -a -b)
+ (a -b + -a
+ u)
-t,
where
and
v are arbitrary.
[7-7]
The formula
lows from the
for the sufficient conditions of
an inequation similarly
If
fol
Law
t
of
Consequences for equations.

(a b
b,
then
+ -a -b + u)
v (a
-b + -a
b) -t
where u and
8-53
If for
are arbitrary.
of
Consequently, by the reductio ad absurdum:

of
v,
some value
t
u and some value

t
3=
(a b
+ -a -b + u)
v (a
-b + -a
b) -t
then a
=|=
b.
[7-6]
an Equation and an Inequation. If we have an equation in one unknown, x, and an inequation which involves x, these may be combined in either of two ways: (1) each may be reduced to the form in which one
System
of
member
either.
and expanded with reference to all the elements involved in Then all the terms which are common to the two may, by 8 17,
is
be dropped from the inequation; (2) the equation and this value substituted for x in the inequation.
may
C
be solved for
x,
8-6
If
B -x = and Cx + D -x * 0, then [5-8] If Cx + D-x

x+
=(=
0,
then -.1
x +
-BD -.r
4= 0.
ACx + -ACx + BD-x + -B D

17
-.r
See Poretsky, Theorie des non-egaliUs logiques, Chaps. 71, 76.
172
A
[5-72] If
Ax + B-x =
and
0,
then
0.
and
B -x =
0,
and hence
A Cx =
Hence
[8
BD
-x
+
=
-B
17] -.4
Cx
-x
4= 0.
The
result here is not equivalent to the data, since
for
one reason
x +
the equation
A Cx+BD
mode
-x
is
not equivalent to
is
B -x =
0.
Nevertheless this
of combination
the one most frequently useful.
8-61
The condition that the equation

4=
x+
B -x =
is,
C x + D -x + -B D 4=
may
x
4=
be regarded as simultaneous
of x
AB =
is
and the inequation and -A C
0,
and the determination

(-A -C +
which they give

-x
A-D)x + (BC + -B D)
is
[7-23]
A x + B -x =
equivalent to x
= -A
x+
-x.
Substi
tuting this value of x in the inequation,
C (-A x + B
or (-A
[8-4]
-x) +
(A x +
4= 0.
is
-B
-x)
4=
C + A D) x+(B C + -BD) -x
condition of this inequation
4=0,
or
(-A+B) C+(A
But the equation

that
+ -B) D 4= 0. x + B -x = +A =
requires that
^45 =
0,
and hence
-A + B = -A and -B
if
-B.
Hence
the equation be possible and
AB =
0,
the condition of the
inequation reduces to
-A C + -B D
4= 0.
[8-4] If the original inequation
this condition is
be possible, then already present in -A C + -B D 4=

if
C+
4= 0.
But
and
0, since
-A C c C
and hence
hence
to
"
[8-25]
if
-B
4= 0,
-.4(74=0, then C 4= 0, and then D 4= 0, while [8-26] C + D 4=
-BDcD
is
equivalent
Either
"Either
C 4= -A C 4=
or
4=
",
and -A C + -B
".
4=
is
equivalent to
or
-B D
4=
Hence the
entire condition of the
system
is
expressed by
AB =
And
[8
5]
and
-A C + -B
4=
the solution of the inequation,

4= 0,
is
(-AC + AD)x + (BC + -BD)-x

x
4=
(-A-C + A-D)x
(BC + -BD)-x
x, in
This method gives the most complete determination of
the form of
an inequation, afforded by the data.

VI.
173
NOTE ON THE INVERSE OPERATIONS,

"DIVISION"
"SUBTRACTION"
AND
It is possible to define
"subtraction"
and
a.
"division"
in the
algebra.
Let a
a.
b be
x such that
+x
And
let
be y such
that
by =
However, these inverse operations are
more trouble than
they are worth, and should not be admitted to the system. In the first place, it is not possible to give these relations a general meaning. We cannot have in the algebra: (1) If a and 6 are elements
in
A",
then a
b is
b is
an element
in
in
K; nor
If
(2) If
a and 6 are elements in K,

y,
then a
it
an element
b
if
K.
a.
must be true that

5-2, a cb.
by
Thus
y a and
But
b
then for some y a, then, by 2-2, a cb y and, y be so chosen that a cb is false, then a b
a
:
b is b
an element,
if
cannot be any element in K. To give a b a general meaning, be required that every element be contained in every element
:
it
would
is,
that
that
x, in
all
elements in
for
be identical.
Similarly,
6
if
b
a.
be an element,
A, then
some
x, it
must be true that
+x
=
if
But
if
+x
a,
then,
by
2-2, b
xca
and, by 5-21,
b
:
be a.
Thus
It
a and b be so chosen
in
A".
that b c a
is false,
then a
cannot be any element
Again, a since a + -a
1
and a
are ambiguous.
1
might be expected that,
=
=
1,
the value of
is
satisfied
1,
by any x such that -a c

is
a would be unambiguously -a. But x. a = x is equiva For 1
lent to x + a
which
equivalent to
-(x + a)
-1
=
x.
-a -x
it
And -a -x =
that, since a -a
:
is
equivalent to -a c
:
Similarly,
might be expected
y,
or a
= 0, the value of a would be unambiguously -a. = y = 0, is satisfied by any y such that y c -a. a y
But
and
y c -a are equivalent.
Finally, these relations can always be otherwise expressed.
of a
:
The value
b is
the value of y in the equation, b y
a.
is
equivalent to
-a b y + a -b + a -y
=
0.
The equation
condition,
is
of condition here is a -b
And
the solution, on this
+u
(a
+ -b)
= ab + u-a-b, where u
is
undetermined.
b
The value
is
of a
b is the value of
x in the equation,
+x
a.
+x
equivalent to
-a
+ -a x + a -b -x
174
The equation
condition,
is
of condition here
is,
-a
0.
And
the solution, on this
a -b + v a
a -b + v a
6,
where
v is
undetermined.
of the expression,
In each case, the equation of condition gives the limitation of the meaning and the solution expresses the range of its possible values.
CHAPTER
There are four applications
III
APPLICATIONS OF THE BOOLE-SCHRODER ALGEBRA

of the classic algebra of logic
which are
commonly
relations.
considered:
(1) to spatial entities, (2) to the logical relations
of classes, (3) to the logical relations of propositions, (4) to the logic of
The
application to spatial entities

line,
may
be
made
to continuous
and
discontinuous segments of a
or to continuous
and discontinuous regions
in a plane, or to continuous
dimensions.
Segments
of a line
and discontinuous regions in space of any and regions in a plane have both been
used as diagrams for the relations of classes and of propositions, but the
application to regions in a plane gives the
more workable diagrams,
for
obvious reasons.
And
since
it
is
only for diagrammatic purposes that
the application of the algebra to spatial entities has any importance,

shall confine
we
our attention to regions in a plane.
I.
DIAGRAMS FOR THE LOGICAL RELATIONS OF CLASSES

a, b, c, etc.,
For diagrammatic purposes, the elements of the algebra,

will
denote continuous or discontinuous regions in a given plane, or in a 1 represents the plane (or circumscribed circumscribed portion of a plane.
portion)
itself.
is
the null-region which
every region.
a,
i.
e.,
not-a.
b.
For any given region, a, The "product", a xb or a

If
b,
b,
is supposed to be contained in -a denotes the plane exclusive of

b, is
",
that region which

is
is
com
0.
mon
The
over.
to a and
"sum",
a and 6 do not
"overlap
then a b
the null-region,
a+
determining a +
denotes the region which is either a or b (or both). In the common region, a b, is not, of course, counted twice
a + b
a -b + a b + -a
b.
This
is
a difference between + in the Boole-Schroder Algebra and the
= b, signifies that a and b denote the same equation, a ac6 signifies that a lies wholly within b, that a is included or region. contained in b. It should be noted that whenever a = b, a c b and b c a.
of arithmetic.
The
Also, a
c a holds always.
.
Thus the
relation
is
analogous not to
<
in
arithmetic but to
175
176
A
While the laws of
this algebra hold for regions, thus denoted,
however
those regions
may
be distributed in the plane, not every supposition about

is
their distribution
classes.
equally convenient as a diagram for the relations of
All will be familiar
with Euler
"All
"
diagrams, invented a century

is
earlier
than Boole
algebra.
"
is
6"
represented by a circle a a and

b,
wholly within a
intersect;
circle b;
No
is b
by two
a
is
circles,
which nowhere
"Some
is
6"
and
"Some
not
6"
by
intersecting circles,
sometimes with an asterisk to indicate that division of the diagram which The defects of this style of diagram are obvious represents the proposition.
:
All a
is
No
is
cuto;
Some
a
is &
x-*>c^\
^-^^\
Some
a
is
not 6
FIG. 1
the representation goes beyond the relation of classes indicated by the propo
sition.
In the case of
"All
is
&",
the circle a
b is
falls
a",
within b in such wise
as to suggest that
we may
infer
"Some
"No
not
b"
but this inference
is
not valid.
The representation
b",
of
a
is
is
similarly suggests
"Some
things are neither a nor
which
also
unwarranted.
With these
is
dia
grams, there
is
no way
of indicating
whether a given region

is
null/
But
Yet
the general assumption that no region of the diagram

misinterpretations
null leads to the

similar.
mentioned,
and
to
others
which are
Euler
still
diagrams were in general use until the invention of Venn, and are
specifically to represent the relations
doing service in some quarters.
The Venn diagrams were invented

of these
of logical classes as treated in the Boole-Schroder Algebra. 1
The
principle
diagrams is that classes be represented by regions in such relation to one another that all the possible logical relations of these classes can be
indicated in the same diagram.
for
any
possible relation of the classes,
then be specified
null.
the diagram initially leaves room and the actual or given relation, can by indicating that some particular region is null or is notis,
That
Initially the
or
1.
diagram represents simply the "universe of discourse", For one element, a, 1 = a + -a. 2 For two elements, a and b,
1
(a
+ -a) (b + -b)
a b + a -b + -a b + -a -b
See Venn, Symbolic Logic, Chap. v. The first edition of this book appeared before Schroder s Algebra der Logik, but Venn adopts the most important alteration of Boole s the non-exclusive interpretation of a + b. original algebra 2 See above, Chap, n, propositions 4-8 and 5-92.
Applications of the Boole-Schroder Algebra
177
For three elements,

1
a, b,
and
c,
(a
+ -a) (b + -b) (c + -c)
a b c + a b -c + a -b
+ -a b
b -c
+ a -b -c
c
+ -a
+ -a -b
+ -a -6 -c
Thus the
"
universe of
discourse"
r(
for
any number
of elements, n,
must
correspond to a diagram of 2
expansion of
1.
If
divisions, each representing a term in the the area within the square in the diagram represent
-a
FIG. 2
-a-b
the universe, and the area within the circle represent the element a, then the remainder of the square will represent its negative, -a. If another element,
6, is
to be introduced into the
same universe, then
may
be repre
sented by another circle whose periphery cuts the first. The divisions, (1) into a and, -a, (2) into b and -b, will thus be cross-divisions in the uni
verse.
If
a and b be classes, this arrangement represents a

b,
all
the possible
subclasses in the universe;
those things which are both a and b;
a-b, those things which are a but not b; -ab, those things which are 6 but not a; -a -6, those things which are neither a nor b. The area which
represents the product, a
b,
will readily
be located.
We
have enclosed
by a broken line, in figure 2, the area which represents a + b. The negative of any entity is always the plane exclusive of that
For example, -(a b + -a -6), in the above, two divisions of the diagram, a -b + -a 6.
If it
entity.
will
be the
sum
of the other
be desired to introduce a third element,
c,
into the universe,

into
it is
necessary to cut each one of the previous subdivisions
two
one
part which shall be in c and one part which shall be outside c. This can be be accomplished by introducing a third circle, thus It is not really necessary to draw the square, 1, since the area given to the
figure, or the
whole page,
is
may
as well be taken to represent the universe.

it
But when the square

13
omitted,
must be remembered that the unenclosed
178
area outside
A
all

is
the lines of the figure

-ft,
a subdivision of the universe
the entity -a, or -a

involved.
or -a -b -c, etc., according to the
number
of elements
-a-b-c
FIG. 3
If
a fourth element,
d,
be introduced,
it is
no longer possible to repre
sent each element
by a
circle, since
a fourth circle could not be introduced
each previous subdivision into two parts one part in d and one part outside d. But this can be done with ellipses. 3 Each
in figure 3 so as to cut
-a-b-c-d
FIG. 4
3
We have deformed the ellipses slightly
and have indicated the two points
of junction.
This helps somewhat in drawing the diagram, which is most easily done as follows: First, draw the upright ellipse, a. Mark a point at the base of it and one on the left. Next,
179
one of the subdivisions in figure 4 can be "named" by noting whether it is in or outside of each of the ellipses in turn. Thus the area indicated by
6
is
a b c -d, and the area indicated by 12

it
is
-a -b
c d.
With a diagram
of
c,
four elements,
requires care, at
first,
to specify such regions as a +
These can always be determined with certainty by developing each term of the expression with reference to the missing ele
a
c
d,
+b
+ -d.
ments. 4
Thus
ac +
= ac
a b
c
(b
+ -b) (d + -d) +
b c
b
c
d (a + -a) (c + -c)
d + a -b
c
d+ a
-d + a -b
-d + a
b -c
d + -a
b c
d + -a b -c d 4 by the
The terms
divisions
of this
sum,
in the order given, are represented in figure

6,
numbered
10,
9, 5,
14,
11,
15.
Hence ac + bd
one
is
the region
which combines these.
With a
it
little
practice,
may
identify such
b
regions without this tedious procedure.

easily identified
1, 4, 5, 8.
Such an area as
2, 3, 6, 7,
+ -d
is
more
and
by inspection:
comprises
10, 11, 14, 15,
Into this diagram for four elements,

e, if
it is
possible to introduce a fifth,

5.
it
we
let e
be the region between the broken lines in figure

diagram"
The
principle
all
of the
"square
(figure 6) is the
same
as
Venn
s:
represents
FIG. 5
draw the horizontal ellipse, d, from one of these points to the other, so that the line con necting the two points is common to a and d. Then, draw ellipse 6 from and returning to the base point, and ellipse c from and returning to the point on the left. If not done in
this
way, the
first
attempts are likely to give twelve or fourteen subdivisions instead of
the required sixteen. 4 See Chap, n, 5-91.
180
the subclasses in a universe of the specified
number
of elements.
No
diagram
is
really convenient for
more than four elements, but such are

-a
-a
-a
a-b
-a -b
FIG. 6
frequently needed.
slightly the square
The most convenient

-a
are those
made by modifying
Figure 7
diagram of four terms, at the right
in figure 6. 5
b<
-b
h
-e
-e
-h
FIG. 7
gives,
by
this
method, the diagrams

for seven
for five
and
for six elements.

is
We
give
also the
diagram
(figure 8)
since this
frequently useful and
not easy to
make in any other way. The manner in which any function
in the algebra
may
be specified in a
diagram of the proper number of divisions, has already been explained. We must now consider how any asserted relation of elements any incluwe
See Lewis Carroll, Symbolic Logic, for the particular form of the square diagram which adopt. Mr. Dodgson is able, by this method, to give diagrams for as many as 10 terms, 1024 subdivisions (p. 176).
6

sion, a
181
b,
or
any equation, a =
relation, or
b,
or inequation, a
=)=
may
be repre
sented.
such relations, can be completely in these diagrams by taking advantage of the fact that they specified
Any such
any
set of
or to the can always be reduced either to the form of an expression = form of an expression 4= 0- Any inclusion, a c b, is equivalent to an equa
tion,
a -6
one of the
And every equation of the form a = b form a -b + -a b = O. 7 Thus any inclusion or

O.
6
is
equivalent to
equation can be
represented by some expression form a --r b is equivalent to one

asserted relation
any inequation of the Thus any of the form a-b + -ab =H O. 8 whatever can be specified by indicating that some region
Similarly,
is null,
= 0.
(continuous or discontinuous) either
=
{
Oj, or
is
not-null,
0}.
in
We
can illustrate
this,
and at the same time indicate the manner
which such diagrams are

Given: All a
.
useful,
by applying the method to a few syllogisms.

a -b
b -c
is 6,
is c,
a cb,
b cc,
and
All b
= =
0.
0.
6 7 8
See Chap, n, 4-9. See Chap, n, 6-4. See Chap, n, 7-1.
182
A
have here indicated
striking
it

the a which
T
We
by
that
(figure 9) that a -b
lines).
is
not
is
null
out (with horizontal
Similarly,
w e have
indicated
by striking out b -c (with vertical lines). Together, the two have eliminated the whole of a -c, thus indicating that a -c = 0, operations
all b is c
or
"All
is c
".
FIG. 9
For purposes
of comparison,
9
we may
derive this
same conclusion by
algebraic processes.
Since a -b
and and
b -c
= =
= =
therefore, a -b c
+ -c) = a -b c + a -b -c, b -c (a + -a) = a b -c + -a b -c, + a -b -c + a b -c + -a b -c = 0,

a -b
(c
[5-72] a b -c + a -b -c
a -c
(b
+ -b)
a -c.
The equation in the third line, which combines the two premises, states The last exactly the same facts which are represented in the diagram.
equation gives the conclusion, which results from eliminating the middle term, b. Since a diagram will not perform an elimination, we must there
"look for"
the conclusion.
illustration of this kind
is b,
:
One more
Given: All a
a -b
b c
and No
b is c,
is
= =
0.
0.
The
first
premise
indicated (figure 10) by striking out the area a -b (with
horizontal lines), the second
by
striking out
be (with
vertical lines).
c,
To
gether, these operations have struck out the whole of a
giving the con
clusion
9
ac =
0,
or
"No
is
c".
orem
in
Throughout this chapter, references in square brackets give the number Chap, u by which any unobvious step in proof is taken.
of the the

In a given diagram where
all
183
the possible classes or regions in the uni
verse are initially represented, as they are by this
method
null or
of
diagramming,
we
cannot presume that a given subdivision
is
is
not-null.
The
actual state of affairs
may
require that
null
some regions be
not.
null, or
that some
be not-null, or that
some be
and others
Consequently, even
when we have struck out the

the premises.
All
regions which are null,
we cannot presume
null-regions,
that all the regions not struck out are no/-null.
This would be going beyond
we can
say,
when we have struck out the

then,
is that, so far as the premises represented are concerned,
any region not
struck out
may
is
be not-null.
If,
we wish
to represent the fact that a
given region
there
is
definitely not-null
that a given class has members, that

indicate this
it is
mark
a
in
b
some expression 4= the diagram. For

=}=
we must
this purpose,
by some
b.
distinctive
convenient to use asterisks.
That a
0,
may
be indicated by an asterisk in the region a

If
But here
further difficulty arises.
say, a,
a
it
b -c.
the diagram involve more than two elements, a b c and b, and c, the region a b will be divided into two parts, a b 4= 0, does not tell us that a b c =}= 0, and Now the inequation,
tell
does not
then,
us that a 6 -c
4= 04=
it
It tells us
only that a b
+a
b -c
=|=
0.
If,
we
wish to indicate a b
by an asterisk
that
is all
in the region a b,
we
c.
shall not be
warranted
in putting
either inside the circle c or outside
It belongs in one or the other or both
we know.
Hence
it is
convenient to indicate a b
of a b
4=
by placing an
by a broken
asterisk in each of the divisions
and connecting them

is
line,
to signify that at least one of
these regions
not-null (figure 11).
ab =0
FIG. 11
We
shall
show
later that a particular proposition
is
best interpreted
by
0.
an inequation;
"Some
is
6",
the class a b has
members, by ab
184
A
we have:
is b, is c,
Suppose, then,
Given:
All a
a-b =
a c
b is
c",
0.
and Some a
H= 0.
is
The
conclusion,
"Some
indicated (figure 12) by the fact that
one of the two connected asterisks must remain the whole region a b c + a -be cannot be null. But one of them, in a -b c, is struck out in indi
cating the other premise, a -6
0.
Thus a
b c
=}=
0,
and hence a
c 4= 0.
FIG. 12
The
by
entire state of affairs in a universe of discourse
may
be represented
by and remembering that any region which is neither struck out nor occupied by an asterisk is in doubt. Also, the separate subdivisions of a region occupied connected asterisks are in doubt unless all but one by
are not-null,
of these connected asterisks
striking out certain regions, indicating
asterisks that certain regions
any regions which are

course, be
left in
made
specifically
occupy regions which are struck out. And doubt by a given set of premises might, of null or not-null by an additional premise.
In complicated problems, the use of the diagram is often simpler and more illuminating than the use of transformations, eliminations, and solu
tions in the algebra.
tions, the
All the information to be derived

(for
diagram gives
one who can
"see"
it)
at a glance.
from such opera Further

diagrams in con
illustrations will be unnecessary here, since
we
shall give
nection with the problems of the next section.

II.
THE APPLICATION TO CLASSES
interpretation of the algebra for logical classes has already been 10 explained. a, b, c, etc., are to denote classes taken in extension; that is 10 Chap, n, pp. 121-22.
The

to say, c signifies, not a class-concept, but the aggregate of
all
185
the objects
denoted by some class-concept. Thus if a b, the concept of the class a not be a synonym for the concept of the class b, but the classes a and b may
must
nifies
consist of the
same members, have the same extension,

of the class a
is
acb
sig
b.
that every
product",
member
b,
also a
member
of the class
The
"
bers of a
and members
denotes the class of those things which are both mem of b. The "sum a + 6, denotes the class of those
",
things which are either
members
of a or
members
of b (or
members
of both).
denotes the null-class, or class without members.
Various concepts
may
an empty class "immortal men", "feathered invertebrates", "Julius Caesar s twin," etc. but all such terms have the same extension; they denote nothing existent. Thus, since classes are taken in extension,
denote
there
is
but one
null-class, 0.
Since
it
is
a law of the algebra that, for
every
x,
the null-class
bers of
ex, we must accept, in this connection, the convention that is contained in every class. All the immortal men are mem
class, since there are
"universe
any
the
no such.
",
represents the class

"universe
".
"every
thing
is
",
of discourse
or simply the
This term
if
pretty well understood.

of the class a,
"universe
X"
But
A"
it
may
be defined as follows:
a n be any
member
then the
"a
and
represent the class-concept of the class x,
of
discourse" is
the class of
If
all
the classes,
x,
such that
blind"
is
an
is
either true or false.

"fixed
"The
fixed stars are

class
is
neither true nor false, then
stars"
and the
"blind"
do not
belong to the
same universe
of a, -a,
is
of discourse.
The negative
of discourse:
a class such that a and -a have no
members
+ -a
==
in
common, and a and -a between them comprise everything

a -a
in the universe
0,
"Nothing is
-a".
both a and
not-a",
and
1,
"Everything is
either a or
6,
Since inclusions, a c
equations, a
b,
and inequations, a
4= b,
repre
sent relations which are asserted to hold between classes, they are capable
of being interpreted as logical propositions.
And
would
the operations of the

are
algebra
transformations,
eliminations,
and
It
solutions
capable
of
interpretation as processes of reasoning.
hardly be correct to
say that the operations of the algebra represent the processes of reasoning
from given premises to conclusions: they do indeed represent processes of reasoning, but they seldom attain the result by just those operations which are supposed to characterize the customary processes of thinking.
In fact,
it is
the greater generality of the symbolic operations which

*
makes
their application to reasoning valuable.
186
A
The representation
of propositions
by
inclusions, equations,
and
in
equations, and the interpretation of inclusions, equations, and inequations

in the algebra as propositions, offers certain difficulties,
due to the fact
that the algebra represents relations of extension only, while ordinary logical
In discussing the representation of the four typical propositions, we shall be obliged to

propositions quite frequently concern relations of intension.
consider some of these problems of interpretation.
The
(1) (2)
"
universal affirmative,
a
All a
is
6",
has been variously represented as,
= =
b,
fl-c6,
(3)
v b,
where
v is
undetermined,
(4) a -b
0.
All of these are equivalent. 11
The only
its
possible
doubt concerns
(3)
v b,
where
v is
undetermined.
But
equivalence to the others
may
be
demon
strated as follows:
[7- 1]
v b is
equivalent to a--(v
b)
+ -a
v b
v b
0.
v b.
But a
Hence
-(v b) + -a v b
[5-72]
if
a (-v + -b) + -a
a -v + a -b + -a
vb, then a -b
0. v
(i.
And
if
b}
then for some value of

of
"All
e., v
a),
v b.
These equivalents
(1)
is
6"
would most naturally be read:

s
The a
is
are identical with those things which are a

b:
and
6 s both.
b.
(2) a
(3)
contained in
class a
is
every
member
of a
is
also a
member
of
The The
identical with
some (undetermined) portion

of a but not
of the
class
6.
(4)
class of those things
which are members
members
of 6
is null.
If
we examine any one
of these symbolic expressions of
"All
is
6",
we
not only may it hold when a 0, but it always = 0-6, c 6, and 0--6 = 0, will be true for every when a = 0. = b for some value of element b. And is always true for v = 0. = means that a has no members, it is thus clear that the algebra Since a
shall discover that
holds
"0
v"
requires that
"All
is
6"
be true whenever no members of a
exist.
The
actual use of language
that
"All
ambiguous on this point. We should hardly say sea serpents have red wings, because there aren t any sea ser
is
pents";
11
yet
we understand the hero

%
of the novel
who
asserts "Whoever
See Chap. n. 4-9.

enters here
assert that
is
187
must pass over

any one
my
dead
body".
This hero does not mean to
will enter the
defended portal over his body: his desire
that the class of those

is
who
enter shall be null.

serpent"
The
difference of the
two cases
concept
portal"
this: the concept
"sea
does not necessarily involve the

"those
"having
red
wings",
while the concept of
who
enter the
as conceived
by the hero
does involve the concept of passing

of
over his body.
We
readily accept and understand the inclusion

of the
an
empty
class in
some other when the concept
of the other
when
the relation
is
one of intension.
one involves the concept But in this sense, an

but in some only.
.r,
ex",
empty
class is
not contained in any and every
class,
In
order to understand this law of the algebra,
"For
every
we must
bear in mind two things:

only,
(1)
that the algebra treats of relations in extension
that ordinary language frequently concerns relations of intension, and is usually confined to relations of intension where a null The law does not accord with the ordinary use of language. class is involved.
and
(2)
This
is,
however, no observation upon
its
truth, for
is
it is
a necessary law
of the relation of classes in extension.
It
an immediate consequence of
"All
the principle,
are also
"For
every
y,
cl",
that
is,
members
of
any
class, y,
members
of the class of all
things".
One cannot accept
this last
without accepting, by implication, the principle that,

class is contained in
in extension, the null-
every
class.
is
The
interpretation of propositions in which no null-class

difficulty,
involved
is
not subject to any corresponding
both because
in
such cases the
relations predicated are frequently thought of in extension and because the relation of classes in extension is entirely analogous to their relation in
intension except where the class
or the class
all cases,
is
involved.
But the
interpretation of the algebra must, in
be confined to extension.
in the algebra as stating
In brief:
"All
is
6"
must always be interpreted

class,
b",
a relation of classes in extension, not of class-concepts, and this requires

that,
whenever a
is
an empty
"No
"All
is
6"
should be true.
The
do not
proposition,
6",
is
is
represented by
ab =
of a
"Nothing is
both a and
or
"Those
"No
things which are
members
and
of b both,
exist".
Since
is
6"
is
equivalent to
"All
is not-6", it
may
v is
also be represented
by
a -b
-b, a
c-6, 6c-a,
or a
=
is
v -b,
where
undetermined.
In the case of this proposition, there

the ordinary use of language.
no discrepancy
between the algebra and
has been a problem to representation of particular propositions not clearly conceived the symbolic logicians, partly because they have
The
188

tried to stretch the algebra to cover traditional
If
"Some
relations of classes relations
and have
which hold
in -intension only.
is
6"
that
it is false
when the
"All
class a
6",
has no members, then

"All
"
"Some
be so interpreted a is will
6"
not follow from
a
"
is
for
is
is
6"
is
true whenever a
0,
0.
But
diffi
6"
on the other hand,

culties:
(1) this
if
Some a
be true
when a =
we have two
does not accord with ordinary usage, and (2) "Some a is will not, in that case, contradict "No a is For whenever there are no
6".
members
"Some
of a
is
6"
(when a = 0), "No a can be true when a =
is
6"
(a b
0) will be true.
Hence
a
is
if
0,
then
"Some
is
b"
and
"No
6"
can both be true at once.

that
"Some
is
b"
solution of the difficulty lies in observing as a relation of extension requires that there be some a
of the class a exist.
"Some
The
that at least one
member
are interpreted in extension,

is
6",
is
6"
does not follow from
Hence, when propositions "All a

is
6"
precisely because
whenever a =
"All
0,
"All
will be true.
But
"Some
is
b"
does follow from
is 6,
and members
a
of a
exist".
To
is
interpret properly
"Some
a
6".
is
6",
we need only remember that

"No
it
the contradictory of
"No
is
Since
0,
is
6"
is
interpreted
by
a b
are
0, "Some
is
6"
will
be a
=(=
that
is,
"The
class of things
which
members
of a
and
of b
both
is not-null".
It is surprising
what blunders have been committed

"Some
of particular propositions.
is
y"
in the representation has been symbolized by x y = v,
where
v is
undetermined, and by u x
v y,
where u and
v are
undetermined.
Both
of these are incorrect,
and
for the
same reason: An
1
"undetermined"
element
may have
the value
or the value
or
any other
value.
all.
Conse
are
quently, both these equations assert precisely nothing at both of them true a priori, true of every x and y and in
They
But
all cases.
For
them
to be significant,
u and
must not admit the value
0.
in that
case they are equivalent to x y =J= 0, which is much simpler defined laws which are consonant with its meaning.
and obeys
well-
Since
we
are to symbolize
"Some
"All
is
b"
by a -b =
=|=
0, it is clear
that
its
contradictory,
is
not
6",
will
be a -b
0.
To sum
follows:
up, then:
the four typical propositions will be symbolized as
A. All a
is b,
a -b
E.
I.
No
is b, is b, is
= 0. = 0. ab
a b
b,
=|=
O.
Some a Some a
0.
=}=
not
a -b
0.
12
Each
12
of these four has various equivalents: See Chap. 11, 4 9 and 8 14.

A. a -b
all
189
0,
b,
-a +
= =
1
1,
-a + -b
-a, a
b,
and -b c-a are
equivalent.
E. a b
all
=
4
1
0,
=
4=
a -b, -a + -b
1,
-a +
-a,
ac-6, and be -a
s
are
equivalent.
I.
a b
0, 0,
a -&, -a + -6 4
1, 1,
and -a +
4 -a are
4=
all all
equivalent. equivalent.
O. o -6
=}=
4=
6,
-a +
and -a + -b
-a are
The reader will easily translate these equivalent forms for himself. With these symbolic representations of A, E, I and O, let us investigate
the relation of propositions traditionally referred to under the topics,
"The
Square of
Opposition",
and
"Immediate Inference".
That the
is
traditional relation of the

If
two
pairs of contradictories holds,

=f=
at once obvious.
a -b
is
=
is
is
true, then a -6
is
false;
if
a -b
is false,
then a -b
=)=
true.
Similarly for the pair, a b

defined:
and a
b 4= 0.
The
relation of contraries
Two
propositions such that both
may
be false but both cannot be true are
"contraries".
This relation
is
traditionally asserted to hold between
A
a
and E.
It does not hold in
ex
tension:
it fails
to hold in the algebra precisely

is
the two propositions
a null-class.
If
0,
whenever the subject of and a b = O. 13 then a -b =
That
is
to say,
"All
if
no members of a
6"
extension,
is
and
"No
is
6"
exist, then from the point of view of But if it be assumed are both true.
or stated that the class a has
members
(a
=|=
0),
then the relation holds.
(b
if
+ -b) a
a b + a -b.
Hence
4= 0,
then a
+ a -b
4= 0.
[8-17] If a b + a -b 4=
and a -b =
0,
then a
b 4= 0.
4= 0.
(1) (2)
And
if
a b + a -b
=h
and a
b
:
0,
then a -b
We may
(1)
is false.
read the last two lines

there are
If
members
of the class a
and
all
is b,
then
"No
is
6"
(2)
is false.
If
there are
members
of the class a
and no a
is b,
then
"All
is
6"
By
tradition, the particular affirmative should follow
from the universal
affirmative, the particular negative
from the universal negative.

0.
As has
been pointed out, this relation fails to hold when a = T ever a 4= 0. e can read a b 4= 0, in (1) above, as
But
a
it
holds
6"
when
"Some
is
instead
"
of of
"
No
13
is
is
b b
is
is
false",
and
a -b 4= 0, in (2), as
"Some
is
not
6"
instead
All a
false
".
We
then have
See Chap, n, 1-5.
190
If If
A
(1) (2)
"

of a, of a,
there are there are
members members
"
and
all
is b,
is b,
then some a
is b. is
and no a
then some a
not
b.
Subcontraries
are propositions such that both cannot be false but
both
may
be true.
Traditionally
"Some
is
6"
and
"Some
=j=
is
not
6"
are subcontraries.
false,
"
But whenever a =
and a -b
0,
a b
a
"
4=
and a -6
0, it
are both
and the
is
relation fails to hold.

b
is false
",
is
Some a
(1)
When = is
holds.
is
Since a b
1
=
we
Some
a
not b
is
false",
can read
(1)
is
and
(2)
above:
of a,
If there are
b.
members
and
"Some
is
6"
is false,
then some a
not
(2)
If there are
members
of a
and
"Some
is
not
6"
is false,
then some
is 6.
To sum
class
up, then: the traditional relations of the
"square
of
opposition"
hold in the algebra whenever the subject of the four propositions denotes a
which has members.
When
the subject denotes a null-class, only
the relation of the contradictories holds.

are, in that case,
The two
universal propositions
both true, and the two particular propositions both false. The subject of immediate inference is not so well crystallized by tradi
tion,
and
class
for the good reason that it runs against this very difficulty of the without members. For instance, the following principles would all
:
be accepted by some logicians

"No
is
6"
gives gives
"Xo
b is b
a".
"No
b is
b is
a"
"All
is not-a".
"All
not-a"
not-a"
gives gives
gives
(6)
"Some
"Some
b is
not-a".
"Some
b is
"Xo
not-a
is
6".
Hence
is
6"
"Some
"
not-a
is
6".
Xo cows
implies
(a) are inflexed
gasteropods
implies
(a)
"Some
non-cows are
"
inflexed gasteropods":
"Some
"Xo
mathematician
has squared the circle

circle".
(b)
non-mathematicians have squared the
These
infer
ences are invalid precisely because the class b

cessful circle-squarers
is
inflexed gasteropods, suc

it was presumed Those who consider the
an empty
"Some
class;
b is
and because
that
"All
b is
not-a"
gives
not-a".
algebraic treatment of null-classes to be arbitrary will do well to consider
the logical situation just outlined with some care.

particular
The
inference of any
proposition
from
the
assumption that either the
class
corresponding universal requires the denoted by the subject of the particular
proposition or the class denoted by its predicate ("not-6" regarded as the is a class which has members, predicate of "Some a is not
6")
191
The
tive
is
"
conversion"
of the universal negative
and
is
of the particular affirma

",
validated by the law

a".
ab =
a
is
"
b a.
6",
Xo
4= 0,
a b
0,
gives b a
0,
"No
b is
And
a
is
"Some
a 6
gives b a
4= 0,
is
4= 0, "Some b is
"Some
a".
Also,
"Some
not
of
6",
a -6 4
0,
gives -6 a
not-6
is
a".
The
"converse"
the universal affirmative
simply the
"converse"
of the corresponding particular, the inference of
which from the universal
has already been discussed.
What
are called
"obverses"
i.
e.,
two equivalent propositions with

is
the same subject and such that the predicate of one

predicate of the other
or
the negative of the
are merely alternative readings of the
same equation,
is
a depend upon the law, -(-a) a is is also "Xo a which is

"All
6",
14
.
Since x y
=
And
"
is
Xo x
a
",
a -b
is
0,
is not-6".
since a b
equiva
lent to a -(-b)
0,
"Xo
is
6"
is
equivalent to
"All
is not-6".
A
S
convenient diagram for immediate inferences can be
made by
putting
(subject)
and
(predicate) in the center of the circles assigned to them,

divisions of
-*S,
-S between the two

divisions.
and -P between
its
two constituent
which the dia-
The
eight arrows indicate the various
ways
in
(Mverl Prop,
Converse
-S-P
FIG. 13
and thus suggest all the immediate inferences which are valid. For example, the arrow marked "converse" indicates the two terms which will appear in the converse of the given proposition and the
gram may be
read,
order in which they occur.

14
In
this diagram,
we must
specify the null
and
See Chap, n, 2-8.
192

if
not-null regions indicated
wish
If
add the
"No
by the given proposition. And we may qualification that the classes, S and P, have members. is and S and P have members:
P",
we
SP
=
5.
4=
-4.
-S-P
7.
6.
FIG. 14
Reading the diagram of figure 14 in the various possible ways, we have: 1. No S is P, and 1. Some S is not P. (According as we read what
is
indicated by the fact that
is null,
or
what
is
indicated by the fact
that S
2.
3.
-P
All
is
not-null.)
is
not-P,
and
2.
Some S
is
not-P.
All
4. 5.
6. 7.
P is not-S, and 3. Some P is not-S. No P is S, and 4. Some P is not-S.

P.
S.
Wanting.
Some not-S is Some not-P is

Wanting.
8.

if
193
Similarly,
"All
is
P",
and S and
0,
f.
P
0,
have members:
S -P =
P
5.
4=
FIG. 15
Reading from the diagram

1.
(figure 15),
we have:
All
is
P,
and
1.
Some S
is
P.
2.
No S
is
not-P.
3.
Wanting.
4.
5. 6. 7.
Some P is S. Some not-$ is

Wanting. No not-P
All
is is
not-P.
S.
8.
not-P
not-S,
and
8.
Some not-P
is
not-S.
The whole
trivial.
so simple as to be almost subject of immediate inference is Yet in the clearing of certain difficulties concerning null-classes
the algebra has done a real service here.

results of syllogistic reasoning algebraic processes which give the in those examples we carried out the have already been illustrated. But
The
14
194
operations at unnecessary lengths in order to illustrate their connection
with the diagrams.
The premises
of
concerns, altogether, three classes.
any syllogism give information which The object is to draw a conclusion
which gives as much of

the
"middle"
this information as
is
term.
This
can be stated independently of the kind of result which elimination exactly

is
gives in the algebra.
And
it
elimination
is
very simple.
O.
15
eliminating x from
of a syllogism
is
A x + B -x
may
of
AB =
is 0,
The result of Whenever the conclusion
universal,
be obtained by combining the premises
in a single equation
one
member
which
and eliminating the
"middle"
term.
For example:
No
is y,
x y
z
0. 0.
All z
is x,
-x
Combining
these, x y + z -x
x,
is
"No
=
=
or
0.
0.
Eliminating
y y
is
Hence the
valid conclusion
z",
"No
z is
y".
Any
syllogism with a universal conclusion

"If
may
a
b,
also be symbolized so
b cc,
that the conclusion follows from the law,
c6 and
then
ace".
By
this
method, the laws, -(-a)
a and
"If
a c
then -b
c-a",
are
some
times required also. 16
For example:
No
is y, is x,
x c -y.
z
All z
ex.
"No
Hence
There
is
c-y, or
"No
z is
y",
and y c-z, or
is
z".
no need to treat further examples of syllogisms with universal conclusions they are all alike, as far as the algebra is concerned. Of course,
:
there are other

clusion,
and of getting the con but the above are the simplest. When a syllogism has a particular premise, and therefore a particular conclusion, the process is somewhat different. Here we have given one
ways
of representing the premises
equation
(1)
1=0) and
one inequation
{+0j.
We
proceed as follows:
expand the inequation by introducing the third element; (2) multiply the equation by the element not appearing in it; (3) make use of the prin If a + b =f= and a = 0, then b 4= ciple, to obtain an inequation with
"
0",
only one term in the

the
"middle
15 16
literal
member;
(4) eliminate the element representing
term"
from
this inequation.
Take, for example,
1 1 in
See Chap, n, 7-4. See Chap, n, 2-8 and 3-1.

the third figure:
All x
is z,
195
x -z x y
-z.
0.
=
=}=
0.
Some x
xy
is y,
0.
[1-5]
[8
+ -z) = x y z + x y Since x -z = 0. x y -z =
x y
(z
Hence, x y
+ x y -z
=f=
0.
17] Since
xy z + xy -z
yz
4= 0,
4=
and
y
is
-.r
0,
therefore x y z
=f=
0.
Hence
[8-22]
or
"Some
a".
An
exactly similar process gives the conclusion for every syllogism with a
particular premise.
have omitted, so far, any consideration of syllogisms with both premises universal and a particular conclusion those with "weakened" in the third and fourth figures. These conclusions, and A A I and E A
are
all
We
invalid as general forms of reasoning.

is
They
If
involve the difficulty
which
now
familiar:
a universal does not give a particular without an
added assumption that some class has members.

of such syllogisms the
we add
to the premises
assumption that the
class
denoted by the middle

Take, for
term
is
a class with members, this makes the conclusion valid.
example,
AA
I in the third figure:

is y,
All x
x-y =
x -z
0, 0.
and x has members,
=|=
0.
All x
is z,
Since x
=k Q,
xy + x
-y
4= 0,
and
since x -y
0,
x y 4
0.
Hence x y
Since x -z
+ x y -z 4= 0, x y -z
(2),
0.
(1)
0.
(2)
z
=f=
By
(1)
and
xy
0,
and hence y
0,
or
"Some
is
z".
Syllogisms of this form are generally considered valid because of a tacit assumption that we are dealing with things which exist. In symbolic
reasoning, or
any other which
is
rigorous,
any such assumption must be

is
made
If
explicit.
An alternative treatment
we take
of the syllogism
due to Mrs. Ladd-Franklin. 17
the two premises of any syllogism and the contradictory of its three conclusion, we have what may be called an "inconsistent triad" that if any two of them be true, the third must be false. propositions such For if the two premises be true, the conclusion must be true and its con17
See
"On
the Algebra of
Logic",
in Studies in Logic by
University, ed.
by
Peirce;
also articles listed in Bibl.

in a modified form,
We
members of Johns Hopkins do not follow Mrs. Franklin s
symbolism but give her theory
due to Josiah Royce.
196
A
And
false.
if

the contradictory of the conclusion be true,
i.
tradictory false.
if
e.,
the conclusion be
false,
and
either of the premises true, then the other
premise must be
sponds to three valid syllogisms.
As a consequence, every inconsistent triad corre Any two members of the triad give the
For example:
contradictory of the third as a conclusion.
Inconsistent Triad
1.
All
a:
is is
y
z
is
2. 3.
All y
Some x
not
z.
Valid Syllogisms
1.
Allan s?/
All
1.
M\
is
y
is is
2.
All y
is 2 is is
2.
.
.
is
3.
. .
All x
is 2.
Some x Some y
not
z
z.
3.
.
not
Some x Some x
not 2 not
y.
Omitting the cases
in
which two universal premises are supposed to
give a particular conclusion, since these really have three premises and
are not syllogisms, the inconsistent triad formed from any valid syllogism
will consist of
two universals and one
particular.
For two universals
will
give a universal conclusion, whose contradictory will be a particular; while
one premise be particular, the conclusion will be particular, and its contradictory will be the second universal. Representing universals and
if
particulars as
sistent triad,
we have done, this means that if we symbolize any incon we shall have two equations = Oj and one inequation 0}.
{
={=
And
the two universals
as a conclusion.
must give the contradictory of the This means that the contradictory of the
{
Oj
particular particular
must be
expressible as the elimination resultant of
a x + b -x
0,
because we have found
all
an equation of the form conclusions from two universals
to be thus obtainable.
will
Hence the two universals
be of the form a x
=
is
and
a b
-x
0,
any inconsistent triad The elimination respectively.

of
will
resultant of a x + b -x
= =
whose contradictory
:
be a
0.
Hence every inconsistent

ax
triad will
0,
have the form
-x
0,
a b
=f
where a and
b are
any terms whatever
positive or negative,
and x
is
any
positive term.
The
from
validity of
any syllogism m-ay be tested by expressing
its
proposi
it
tions in the form suggested, contradicting its conclusion
by changing
{=0}
to
{=t=.0} or the reverse,
and comparing the resulting triad
197
with the above form.

considering
And
the conclusion of any syllogism
may
be got by
how
the triad must be completed to have the required form.
Thus,
if
the two premises are
No
and
is y,
is y,
x y
-z -y
= = =
All not-z
the conclusion must be universal.
The
particular required to complete

is
the triad
is
x -z
it
4= 0.
Hence the conclusion
x -z
0,
or
is
"All
is
z".
(Incidentally
may
be remarked that this valid syllogism

Again,
if
in
no one of
the Aristotelian moods.)
the premises should be x y
and
yz =
0,
no conclusion
is
to the possible, because these two cannot belong
same inconsistent
triad.
We can, then, frame a single canon for all strictly valid syllogistic reason
ing:
The premises and the contradictory of the conclusion, expressed = 0} or 4= 0}, must form a triad such that symbolic form,
{ {
in
(1)
There are two universals
(2)
and one particular The two universals have a term in common, which
{
0}
{4= 0).
is
once positive
and once negative.

(3)
The
particular puts
4=
the product of the coefficients of the
com
mon term
in the
two
universals.
few experiments with traditional syllogisms will make this matter clear The validity of this canon depends solely upon the nature to the reader. three terms, three propositions and upon the law of
of the syllogism
elimination resultants,
"If
.r
+ b -x
0,
then a
0".
Reasoning which involves conditional propositions hypothetical argu same process, if we first ments, dilemmas, etc. may be treated by the A reduce them to syllogistic form. For example, we may translate cases in which where x is the class of is J5, then C is by "All x is All cases in which .1 in which C is Di. e., A is B, and y the class of cases
"If
D"
y",
"
is
B
"
are cases in which

All 2
is
is
D".
And we may
translate
"But
is
B"
by
",
where
z is
the hypothetical argument: is represented by the syllogism: C is

"If Z)",
the case or class of cases under discussion. Thus But ,1 is B. Therefore, .1 is B, C is D.
"All
"
cases in which
all
is
are cases in which
is
D.
is
But
the cases in question are cases in which
B.
is
Z)."
"Hence all
the cases in question are cases in which
And
in some other arguments of this type are reducible to syllogisms to Thus the symbolic treatment of the syllogism extends similar fashion.
all
198
A
also.

is
them
But conditional reasoning

of the algebra
more
easily
and simply treated

lies in
by another interpretation
the interpretation for propositions.
The
its
chief value of the algebra, as
an instrument of reasoning,
liberating us
from the limitation to syllogisms, hypothetical arguments,

logic.
dilemmas, and the other modes of traditional

the narrowness of formal logic
still
Many who
object to
do not
realize
logical point of view) its limitations are.

etc.,
how arbitrary (from the The reasons for the syllogism,

be worth while to exemplify
are not logical but psychological.
It
may
this fact.
We
shall offer
two
illustrations designed to show,

logical possibilities
each in a
different
logic.
way, a wide range of

first
undreamt
of in formal
The
of these turns
upon the properties of a triadic relation

to think in terms of dyadic relations:
whose
significance
was
first
18 pointed out by Mr. A. B. Kempe.
It is characteristically
human
we
habitually break up a triadic relation into a pair of dyads.

is
In fact, so
ingrained
this disposition that

It
some
will
be sure to object that a triadic
relation is a pair of dyads.

all
dyadic relations are triads with a null
would be exactly as logical to maintain that member. Either statement is

simply one of point of view
psychological
logical sense
correct
enough
If
the difference
is
preference.
there should be inhabitants of
Mars whose
coincided with our own, so that any conclusion which seemed valid to us would seem valid to them, and vice versa, but whose psychology otherwise
differed
prejudice in favor of triadic relations.
from ours, these Martians might have an equally fundamental We can point out one such which
the
they might regard as

or inclusion.
elementary relation of logic

all
as
we regard
equality
In terms of this triadic relation,
their reasoning
might
be carried out with complete success. Let us symbolize by (ac/b), a-bc +
diagrammed
+ -ab -c
is
as in figure
16,
since
-ab-c = 0. This relation may be a-bc + -ab-c = is equivalent to

(ca/b) are equivalent, since
c.)
accbc(a + c).
This relation
(Note that
(ac/b)
and
a -be
symmetrical with respect to a and

(ac/b)
precisely the information which we discard in drawing a syllogistic conclusion from two universal habitually If all a is b and all b is c, we have premises.
represents
a -b
=
+ -c) +
and
(a
6 -c b -c
=
0,
Hence a -b
18
(c
+ -a)
the Relation of the Logical Theory of Classes and the Geometrical Theory of Points/ , Proc. London Math. Soc., xxi, 147-82. But the use we here make of this relation is due to Josiah Royce. For a further discussion of Kempe s triadic relation,
"On
See his paper
see below, Chap, vi, Sect. iv.

Or, a -6 c + a -b -c + a b -c + -a b -c
[5-72] This equation
(1)
is
199
0.
a -b -c + a b -c
and
(1) is
(2)
a -b
+ -a
b -c
"All
= =
a
a -c
0.
is
(b
+ -b)
a -c
0,
the syllogistic conclusion,

feel
c";
(2) is (ac/b).
all
Perhaps most
the information
that precisely
of
us would
that a syllogistic conclusion states
given by the premises: the Martians might equally well
feel
FIG. 16
what we overlook
curious
"illogical"
is
the only thing worth mentioning.
And
yet with this
be capable of understanding and of getting for themselves any conclusion which a syllogism or a hypo
prejudice, they
still
would
thetical
argument can
stateable in terms of
and many others which are only very awkwardly a is our formal logic. Our relation, a c6, or
give,
"All
6",
would
be, in their terms, (Ob/ a).

1
(Qb/a)
is
equivalent to
a -6 +
-a 6
=
"
a -b
hence (Oc/a) This would, in fact, be only a special case of a more general principle which is one of those we may suppose the Martians would ordinarily rely upon
Hence the syllogism
in
Barbara would be
(06/a)
and
(Oc/6),
".
for inference:
principle holds,
"If
(xb/a)
and
(xc/b),
then
(arc/a)".
That
this
general
is
proved as follows:
(xb/a)
(xc/b)
is
-x a -6 + x -a -x
b -c
b
c
is
+ x -b
= =
=
These two together give

-x a-b - x a _5
(e
or?
+ -c) + x -a b (c + -c) + -x b -c (a + -a) + x -b c (a + -a) + - x a -b -c + x -a b c + x -a b -c + -x a b -c + -x -a b -c + x a -b

c
0,
+ x -a -6 c
0.
200
[5-72] This equation
(1)
-.r
A
is

equivalent to the pair, -ab c + x -a -b
a b -c + -x a -b -c + x
c c (b
= =
(2)
-x a-c
(b
+ -b) + x -a
+ -b)
-x a -c +
c
a*
-a
0.
x -a
b -c
+ -x -a b -c + x a -b
+ -x a -b c
b -c (x
= =
(1) is (xc/a), of
-a
+ -x) + a -b
c
c (x
+ -x)
-a
b -c
+ a -b
0.
which our
syllogistic conclusion is a special case;
(2) is a
similar valid conclusion,
though one which we never draw and have no
language to express.
Thus these Martians could deal with and understand our formal
by treating our dyads
as triads with one
logic
member
null.
In somewhat
similar fashion, hypothetical propositions, the relation of equality, syllo
gisms with a particular premise, dilemmas, etc., are all capable of state ment in terms of the relation (ac/b). As a fact, this relation is much more
powerful than any dyad for purposes of reasoning. Anyone who will trouble to study its properties will be convinced that the only sound reason for not using it, instead of our dyads, is the psychological difficulty of
keeping in mind at once two triads with two members in
common
but
Our differently placed, and a third member which is different in the two. But the operations of the algebra are inde attention-span is too small.
pendent of such purely psychological limitations that is to say, a process too complicated for us in any other form becomes sufficiently simple to be
clear in the algebra.
logic
The algebra has a
generality and scope which
"
formal
"
cannot attain.
traditional modes of reasoning.
This illustration has indicated the possibility of entirely valid nonWe shall now exemplify the fact that by
are not so remote from familiar processes of reasoning,
modes which
any
number of non-traditional conclusions can be drawn. we make use of Poretsky s Law of Forms: 19
x
This law
is
For
this purpose,
is
equivalent to
-x +
-t
-tx
"All
=
a
-t-0
is
6"
evident enough: if x = 0, then for any t, t-x = t-1 = t, and = t. Let us now take the syllogistic premises, + 0, while
and
"All
6 is
c",
and
see
what
sort of results can be derived
from them by
this law.
All a
All b
19
is b,
is c,
a-b =
b-c =
0. 0.
See Chap, n, 7-15 and 7-16.
201
Combining
these,
a-b +
b -c
And
[3-4-41]
-(a-b + b-c)
= 0. = -(a
-b) --(b -c)
(-a + b)(-b +
c)
=
Let us make substitutions, in terms of
a+ b
a, 6,
-a -b + -a
t
+b
c.
and
c,
for the
of this formula.
= (a + b) (-a -b + -a c + b c) = abc+-abc + bc = be
is
+ -a -b (a-b + b -c)
What
is
either a or 6
identical with that

"All
which
6
is
both
and
c".
c.
This
is
non-syllogistic conclusion from
is
and
all
b is
Other such
conclusions
a
may be got by similar substitutions in the formula. + c = (a + c) (-a -b + -a c + b c) + -a -c (a -b + b -c) = a b c + -a -b c + -a c + b c + -a b -c = a b c + -a (b + c)

is
What
is
is
either a or c
identical with that
which
is a, b,
and
c, all
three, or
not a and either

-5
c
b or c.
= -b c (-a -b + -a c + b c) + (b + -c) (a = -a -b c + b -c + a -b -c = -a-b c +

but not
c is identical
c.
-b + b -c)
(a
b)
-c
That which
or
is
is b
with what
is c
but neither a nor b
either a or b but not

"All
The number
"All
of such conclusions to be got

is
from
of
the premises,
is
6"
and
b is
c",
limited only
c,
by the
this
number
functions which can be formed with

stitutions in terms of these
is,
a, b,
and
and the limitation to sub
of course, arbitrary.
By
method, the
is
number
of conclusions
which can be drawn from given premises
entirely
unlimited.
In concluding this discussion of the application of the algebra to the more involved logic of classes, we may give a few examples in which problems
than those usually dealt with by formal logic are solved. The examples chosen are mostly taken from other sources, and some of them, like the
first,
are fairly historic.
Example
I.
20
certain club has the following rules:
(a)
The
financial
committee
shall
shall be
chosen from
among
(c)
the general committee;
(b)
No
one
be a
member both
on the
of the general
and
library committees unless he be also

of the library
on
be
the financial committee;
Xo member
committee
shall
financial committee.
Simplify the rules.

10
See Venn, Symbolic Logic, ed.
2, p.
331.
202 Let /
g
/
A
= member = =
" "
of financial committee.
"
general
"
library
The premises then become:

(a)
(6)
(c)
fcg,
(gl)cf,
or or
/ -g
0.
0.
-fgl =
// =
0.
We
can discover by diagramming whether there
is
redundancy here.
(c)
In
figure 17, (a) is indicated
by
vertical lines, (6)
by
horizontal,
by
oblique.
(a)
and
(c)
both predicate the non-existence of / -g
I.
To
simplify the
rules, unite (a), (6),
and
(c)
in a single equation:
-g)
[5-91]
= f -g +
-f g
+ fg
+ f- g
=f-g+(-f+f)gl=f-g
is
+ gl
0.
And
Thus the
(a
}
[5-72] this
equivalent to the pair, f -g

:
and g
0.
simplified rules will be

financial
The
committee
shall be
chosen from
among
the general
committee.
(6
)
No member
of the general
committee
shall be
on the library com
mittee.
203
Example
a
is
2.
21
The members
s
of a certain collection are classified
in three
ways
as
or not, as 6
or not,
and as
c s
or not.
It is
c s
then found that the class 6

c s
made up
is
precisely of the a s
which are not
and the
which are not a
s.
How
the class c constituted?

b b
Given:
a -c + -a
b (c
c.
To
6 c
solve for
b -c.
c.
c.
22
+ -c) = Hence, b c + b -c
a -c + -a
c
Hence [7-27] a -6 + -a b c Or [2-2] c = a-b + -ab.
c a -b + -a
b.
The
c s
comprise the a
which are not
6 s
and the
6 s
which are not a

b
s.
Another solution of
+ -ac to the form
[7
1]
{
this
problem would be given by reducing
a -c
0}
and using the diagram.

c is
a -c + -a
equivalent to
6 --(a -c +
-a
c)
+ -b
c
(a -c
+ -a
c)
And
[6 4]
-(a -c + -a
c)
+ -a
-c.
c
Hence, a
b c
+ -a
6 -c
+ a -6 -c + -a -6
c
0.
We
observe here (figure 18) not only that
a -b + -a
b,
but that the
FIG. 18
relation of a,
6,
and
c,
stated
c.
by the premise
is
totally symmetrical, so that
we have
21
also a
b -c + -6
of
Adapted from one

i
Venn
s, first
printed in an article on
"Boole
System
of
Logic",
Mind,
22
(1876), p. 487.
intelligible
if
This proof will be
the reader understands the solution formula referred
to.
204
A
3.
23
Example
If
x that
is
is
not a
is
the same as
6,
and a that
is
not x
is
the same as
c,
what
x in terms of
b
a, 6, c
and
c?
-.r.
Given:
= -ax and
To
solve for
.r.
[7-1] 6
=
6
-a
.i
is
equivalent to
-(-a
.T)
+ -a -b x
(a
-.r) 6
+ -a -6
a:
=
And
c
a b + b -x + -a -6
.r
(1)
=
c
a -x
is
equivalent to
-(a -x)
+ a -c -x
(-a +
.T)
+ a-c -x
= -a c
Combining
(1)
x + a -c -x
(2)
and
(2),
a b + -a c + (-a -b + c) x + (b + a -c) -x
(3)
Hence
[5
72]
(-a -b + c) x + (b + a -c) -#
(4)
[7-221] This gives the equation of condition,
(-a -b + c)(6 + a -c)

[7-2]
6 c
(5)
The
solution of (4)
(b
is
+ a -c) c # c -(-a -6 +
c)
And by
(5),
c)
-(-a -6 +
=
x
-(-a -6 +
c)
+ 6
(a
b)
-c + b c
=
Hence
[2-2]
a -c + b
(c
+ -c)
+ a -c
b + a -c.
FIG. 19
2S
See Lambert, Logische Abhandlungen,
i,
14.
205
This solution
is
verified
combines
all
the data.
by the diagram (figure 19) of equation Lambert gives the solution as

x
(a
(3),
which
b) -c
This also
is
verified
by the diagram.
Example
4.
24
What
is
the precise point at issue between two disputants, one of
whom,
A, asserts that space should be defined as three-way spread having points

as elements, while the other, E, insists that space should be defined as
three-way spread, and admits that space has points as elements.
Let
asserts: s
s
=
p
p.
= space, t = three-way spread, p = having points as elements. B states: s = and s cp.

s
t
is
equivalent to
s--(tp)+-stp =
s
s -t
-p + -s
(1)
(2)
cp
is
equivalent to
-p =
t
And
(2)
is
equivalent to s -t + -s
=
=
(3)
and
(3)
together are equivalent to

s -t
-p + -s
(4)
(1)
represents
.4 s
assertion,
and
(4)
represents
s.
The
difference between
FIG. 20
the two
is
that between -s tp
=
t
and -*
-s
t
=
t
0.
(See figure 20.)
-s
24
p + -s
-p
Quoted from Jevons by Mrs. Ladd-Franklin,
loc, tit., p. 52.
206
A
difference
is,
The
then, that
would be easy to misinterpret
way
spread not having points nificantly assert this, for he has denied the existence of any space not having
".
-st-p = 0, while A does not. It is t-pcs, "Threethis issue, -st-p = But B cannot sig as elements, is space
asserts
s
points as elements.
Both assert
p.
The
real difference is this:
definitely asserts that all
space, while
has
left
three-way spread has points as elements and is open the possibility that there should be three-way
spread not having points as elements which should not be space.
Example
5.
Amongst the
enough to sink a
unless
it is
objects in a small
useful.
boy
pocket are some bits of metal
which he regards as
either
But
all
the bits of metal which are not heavy
fishline are bent.
And
he considers no bent object useful

is
heavy enough
to sink a fishline or
fishline,
not metal.
And
the
only objects heavy enough to sink a

are bits of metal that are bent.
which he regards as
useful,
Specifically
what has he
in his pocket
which
he regards as useful?
Let x
= y = z = w =
bits of metal,
objects he regards as useful,
things heavy enough to sink a fishline,
bent objects.
Symbolizing the propositions in the order stated, we have
xy *
x -z c w,
or
x -z -w
or
y w c (z + -.i ), or zy ex ID,
x y -z w = -x y z + y z -w =
z
Expanding the inequation with reference to

x y
z
and w,
w+x
-w + x y
-z
w+
x y -z
-w
=H
Combining the equations,

x -z -w (y + -y) + x y -z w + -x y
or
z
(w + -w) + y
z
-w
(x
+ -x)
z
=
-w =
x y -z
-w + x -y
-z
-w + x y
-z
w + -x y
w + -x y z-w + x y
terms of the inequation appear also in this equation, with the exception of x y z w. Hence, by 8- 17, x y z w 4= 0. The small boy has
All the
207
some bent
useful.
bits of
metal heavy enough to sink a
fishline,
which he considers
This appears in the diagram (figure 21) by the fact that while
FIG. 21
some subdivision
It
of x y
must be
not-null, all of these but x y z
is null.
appears also that anything else he may have which he considers useful may or may not be bent but is not metal.
Example
6.
25
The annelida
and
either
consist of all invertebrate animals having red blood in a
double system of circulating vessels.
And
all
annelida are soft-bodied,
naked or enclosed
in a tube.
Suppose we wish to obtain the

in tubes are placed (by virtue
relation in
which soft-bodied animals enclosed
with respect to the possession of red blood, of an external covering, and of a vertebral column.
of the premises)
Let a
= = s n = = i = r =
/
annelida,
soft-bodied animals,
naked,
enclosed in a tube,
invertebrate,
having red blood,
etc.
t
and a cs (n + t), with the implied condition, n eliminate a and find an expression for s t.
Given: a
i
0.
To
25
See Boole, Laws of Thought, pp. 144-46.
208
a
A
=
i r is
equivalent to
-(i r) a + -a
i r
a -i + a -r + -a
i r
(1)
cs (n+t)
n+
s
t)
is
-0
= 0. equivalent to a --(s n + s t) = (-s + -n) (-s + -0 -(s ii) -(s t)

-
-s + -n
-t.
Hence,
a-s + a-n-t =
(1)
(2)
Combining
and
(2)
and n
0,
a -i + a -r + -a
i r
+ a -s + a -n -t + n
(3)
Eliminating
(-i
a,
by 7
4,
+ -r + -s +
-7i -2
+ n
t) (i
+n
t)
= n
26
i r
-s +
i r
-n
-t
The
solution of this equation for s

its
is
i r
cs.
And
solution for
[5-3]
tisir-nctc-n.
s
t
Hence
ir-nc.stc.-n, or
= ir-n+u--n, where u
is
un
determined.
The
soft-bodied animals enclosed in a tube consist of the invertebrates
FIG. 22
26
See Chap, n, Sect,
iv, "Symmetrical
and Unsymmetrical Constituents
of
an Equa
tion".
209
which have red blood
in a double system of circulating vessels and a body with an undetermined additional class (which may be covering, together This solution may be null) of other animals which have a body covering.
by the diagram of equation (3) the square formed by the two crossed
verified
(figure 22).
In this diagram,
is
rectangles.
The lower
half of this
inner square exhibits the solution.

i r
Note that the

is
qualification, -n, in
-n c s
t,
is
necessary.
i
In the top row
a single undeleted area repre

s
t.
senting a portion of
r (n)
which
is
not contained in
Example
7.
27
"All
Demonstrate that from the premises a is either b or c is no conclusion can be drawn which involves only two
"All
a",
c",
and
of the
classes, a, 6,
and
(b
c.
Given: a c
c)
and
a.
To prove
is
that the elimination of any one element gives a result which
either indeterminate or contained in one or other of the premises.
a c
(b c
c) is
is
equivalent to a -6 -c
0.
And
= 0. Combining these, a -b -c + -a c = 0. = 0. Eliminating a [7-4], (-b -c) c = 0, which is the identity, = 0. Eliminating c, (a -b) -a = 0, or c + a -c) -a c = -a c = 0, which is the second Eliminating b, (-a
ca
equivalent to -a c
premise.
Example
8.
A
green,
set of balls are all of
them spotted with one
or
more
of the colors, red,
and
blue,
and are numbered.

All the
And
all
the balls spotted with red are

balls,
also spotted with blue.
odd-numbered blue
and
all
the even
numbered balls which are not both red and green, are on the table.
scribe the balls not
De
on the table.
Let
e r
g
t
Given:
27
(1)
-r -b -g
= = = = = =
even-numbered, -e
spotted with red,
= odd-numbered,
spotted with blue, spotted with green,

balls
0.
Logic, p. 123.
on the
table.
See
De Morgan, Formal
15
210
r
A
(2)
(3)
-6
0.
[-eb + e-(rg)]ct,
x,
or
(-e b + e -r + e -g) -t
0.
To
find
an expression,
such that -t c x, or -t x
-t.
Such an expression
-t
should be as brief as possible.

respect to
e, r,
Consequently we must develop

all
with
b,
and
g,
and eliminate
-t,
null terms.
(An alternative
is
method would be
-t
to solve for
(r
but the procedure suggested
briefer.)
-t(e + -e)
+ -r) (b + -6) (g + -g)
= -t(erbg+erb-g+er-bg + e-rbg + -erbg + er-b-g

+ +
e -r b
e
-g + -e
r b
r
-g +
-r -b g + -e -r b g + -e
-b g
(4)
-r -b -g + -e
-b -g + -e -r b -g + -e -r -6 g + -e -r -b -g)
From
(1), (2),
and
(3),
-t (-e b
-r +
-g +
-b + -r -b -g)
(5),
(5)
Eliminating from
-t
(4)
terms involved in
g),
-t (e r b g
+ -e -r -b
or
-t
(e r b
g + -e -r -b g)
all
All the balls not
on the table are even-numbered and spotted with colors or odd-numbered and spotted with green only.
three
g...
211
In the diagram (figure 23), equation (1)

(2)
is
indicated by vertical lines,
by oblique,
9. 28
(3)
by
horizontal.
Example
Suppose that an analysis of the properties of a particular class of sub stances has led to the following general conclusions:
1st.
That wherever the properties a and

c,
b are
combined, either the
property
2d.
or the property d,
is
That wherever the properties
present also; but they are not jointly present. b and c are combined, the properties
a and d are either both present with them, or both absent.

3d.
ties c
c
and d are both absent
That wherever the properties a and b are both absent, the proper also; and vice versa, where the properties
and d are both absent, a and b are both absent also. Let it then be required from the above to determine what may be con cluded in any particular instance from the presence of the property a with
respect to the presence or absence of the properties b and regard to the property d.
c,
paying no
Given:
(1) (2) (3)
a b c
(c
-d + -c
d).
bcc(ad + -a-d). -a -b = -c -d.

a.
To
eliminate d and solve for

(1) is (2) is
equivalent to a b--(c -d + -c d) equivalent to b c--(a d + -a -d)
= =
0.
0.
But [6-4] -(c-d + -cd) =cd + -c-d, and -(a d + -a -d) = -a d + a -d. Hence we have, a b (c d + -c -d) =abcd+ab-c-d = Q and b c (-a d+ a -d) =-abcd+abc-d = ()
(3) is
(4)
(5)
equivalent to
(c
-a -b
+ d) +
(a
b)
-c -d
=
Combining
function of
(4), d, (5),
-a -b
(6),
+ -a -b d + a -c -d + b -c -d
result the
(0)
and
and giving the
form of a
(-a -b c + -a -6 + a b c + -a b
c)
d
c
+ (-a -b
28
+ a -c +
b -c
+ a b -c + a b
c)
-d
See Boole, Laws of Thought, pp. 118-20.
Franklin, loc. cit., pp. 51-61, Venn, Symbolic Logic, Chap, Logik: Vol. I, Dreizehnte Vorlesung.
For furfher problems, see Mrs. Larldxm, and Schroder, Algebra dcr
212
A
Or, simplifying,
(_ a
-fc
by 5-4 and 5-91,

+ b
c)
d + (-a -b
d,
+ a -c +
b -c
+a
b c)
-d
Hence
[7
4]
eliminating
I)
(_ a
-I)
c ) (-a
-b
+ a-c +
b -c
ab
c
c)
= -a-b
ab
Solving this equation for a [7-2], -b
cac
(-b + -c).
always present when c is present and b absent, and when ever a is present, either b is absent or c is absent. The diagram (figure 24) combines equations (4), (5), and (6).
The property a
is
FIG. 24
is
As Boole correctly claimed, the most powerful application of this algebra to problems of probability. But for this, additional laws which do not
belong to the system are, of course, required.
Hence we omit
it.
Some
thing of what the algebra will do toward the solution of such problems will be evident if the reader imagine our Example 8 as giving numerically the
proportion of balls spotted with red, with blue, and with green, and the
quaesitum to be
"If
a ball not on the table be chosen at random,
what
r
is
the probability that
it will be spotted with all three colors? that it w ill be with green?" The algebra alone, without any additional laws, spotted answers the last question. As the reader will observe from the solution,
all
the balls not on the
ta*ble
are spotted with green.

III.
If,
uct",
213
THE APPLICATION TO PROPOSITIONS

a, b, c, etc.,
in
our postulates,
represent propositions, and the
"prod
b,
represent the proposition which asserts a and b both, then
we
have another interpretation of the algebra. Since a+b is the negative of is false that a and b are both false", or -a -6, a + b will represent
"It
"At
least
"
one of the two, a and

6",
6, is
true".
It
has been customary to read

is
a+
b,
Either a or
or
"
Either a
"Either
is
true or b
or
true
".
But
this
is
some
in
what misleading,
algebra
is
since
...
..."
frequently
denotes,
ordinary use, a relation which
is
to be understood in intension, while this
incapable of representing relations of intension.

affirm
"Either
For instance,
"
we should hardly
parallels
meet at
finite intervals or all

.
men
are
mortal".
We
might well say that the

"
"Either
or
relation here predicated fails to hold because the

irrelevant.
But
at least one of the two,

mortal",
Parallels
two propositions are meet at finite intervals"
and
"All
men
are
is
a true proposition.
by +
and
in the algebra holds
between them.
in
The relation denoted Hence, if we render a + b by

necessary connection of a
"Either
a or
6",
we must bear
mind that no
6,
no relation of
"relevance"
or
is "logical import",
intended.
"a
The negative
false".
of a, -a, will be its contradictory, or the proposition
is
It
might be thought that -a should symbolize the

that
if
"contrary"
of a as well,
a be
"All
men
are mortal
",
then
"
Xo men
are mortal
"
But if the contrary as well as the contradictory be denoted then -a will be an ambiguous function of a, whereas the algebra by -a, 29 requires that -a be unique.
should be -a.
The
and
its
interpretation of
and
1 is
most
easily
made
clear
by considering
a,
the connection between the interpretation of the algebra for propositions

interpretation for classes.
The
prepositional sign,
may
equally
well be taken to represent the class of cases in which the proposition a is a b will then represent the class of cases in which a and b are both true,
true; -a, the class of cases in
which a
is false,
and
so on.
The
"
universe", 1,
will be the class of all cases, or all
"actual"
cases, or the universe of facts.

is
Thus a =
"a
represents
"The
cases in which a
"a
true are
all
cases",
or
is
true in point of
cases,
fact",
or simply
"a
is
true".
Similarly
or
"a
is is
the class
of
no
It
and a
will
mean
is
true in no
case",
false".
might well be asked:

die
May
not
a, b, c, etc.,
represent statements which

"Today
are sometimes true or

"The
and sometimes
ace"?
false,
such as
is
Monday"
is
shows an
May
not a symbolize the cases in which a
29
See Chap, n, 3-3.
214
A
and these be not
1
true,
all
but only some of the cases?

true",
And
should not
be read
"a
is
always
"a
as distinguished from the less
com
prehensive statement, thus suggested can be
is true"?
The answer
But symbolic
is
that the interpretation
made and
that Boole actually
made
it
in his chapters
on
30
"Secondary
Propositions".
logicians
have come to
distinguish between assertions which are sometimes true and sometimes In the sense in which "Today is Monday" is false and propositions.
sometimes true and sometimes
false, it is called
a propositional function
and not a proposition.

first place,
There are two principal objections to interpreting

In the
the Boole-Schroder Algebra as a logic of propositional functions. the logic of propositional functions
is
this algebra,
and
in the
second place,
it is
algebra to propositions by the additional

if
much more complex than much more useful to restrict the a H= 0, then a = 1, and law
"If
=^ 1,
then a
0",
and avoid any confusion
of propositions with asser
tions
which are sometimes true and sometimes
false.
In the next chapter,
we
shall investigate the
tions but not for classes or for propositional functions.
consequences of this law, which holds for proposi We need not pre
:
sume
this
law at present
the Boole-Schroder Algebra, exactly as presented
in the last chapter, is applicable
remember that a proposition

tion
is
is
throughout to propositions. But we shall either always true or never true if a proposi
:
true at
all,
it is
always
==
true.
Hence
"a
in the interpretation of the

true"
algebra for propositions, a

indifferently
false"
means
is
or
"a
is
always
either
true"
the two are synonymous.

is
And
means
"a
is
or
"a
always
b,
false".
The
is false
true".
relation a c
since
it
is
equivalent to a -b
",
=
"If
0,
may
is
be read
"
It
is
that
is
true and b
like
is false is
or loosely,
true, then b
But acb,
and
==
a+
b,
here a relation which does not signify

"logical
"relevance"
=
or a connection of
"Christmas
is
import".
Suppose a
that
"2
+2
"If
4"
b =
holiday".
We
Yet
should hardly say

"2
4,
then Christmas
is
is
holiday".
it is false
+2
and Christmas hence a c b will

it is
not a
holiday":
in this
example a -b
called
is
true,
and
";
hold.
This relation, a c
b, is
"material
implication
a relation of extension, whereas
we most frequently
"a
6"
interpret
"implies"
as a relation of intension.
But
acb
common
true,
30
case in which a
with our usual meaning of is true but b is false does not occur.
b will
has one most important property in when a c b is true, the implies

If
a c b holds, and a
is
then
Laws
not be
false,
though
it
may
be irrelevant.
Thus
"material
of Thought, Chaps, xi-xiv.

implication" is
215
of ordinary
a relation which covers
more than the

"a
"implies"
logic: a
c6
holds whenever the usual

in
implies
6"
holds;
it
also holds in
some cases
which
"a
implies
6"
does not hold. 31

is
The
application of the algebra to propositions

its
so simple,
and so
resembles
application to classes, that a comparatively few illustrations
will suffice.
sitions,
We
give
some from the elementary
logic of conditional
propo
and conclude with one taken from Boole.
Example
If
1.
is
B,
is
C
B.
is
D.
(1) (2)
And A
Let x
= A
is B-,
= C
is
D.
are
:
The two premises then

(1)
(2)
xcy,
x
or [4-9] -x + y
or -x
==
1.
1,
=
1
is
0.
[5-7] Since
-x + y
"
=
C
and -x =
0,
1.
1 is
the conclusion
".
Example
(1) (2)
2.
If
is
B,
is
is
D.
But C
not D.
Let
= A
(1)
is
B- y
= C
is
D.
.re?/, or
-x + y
=
1
1.
(2)
0.
[5-7] Since
-x + y
=
"A
and y =
0,
-r
or
1.
-x
1 is
the conclusion
is 1? is
false",
"A
is
not
Example
(1)
If
3.
(3)
A isB,C isD; But either A is B

x
and
or
(2)
if
is
F,
is
H.
is Z).
Let
w = A isB;
(1)
(2) (3)
31
= ClsD; y = E = 10. icc.x, or [4-9] w x = # ycz, or 2/3 w + y - 1.

discussed
isF;
= G
is //.
"Material implication" is
more at length
in
Chap,
iv, Sect,
i,
and Chap.
v, Sect. v.
216
Since w + y
A
=
1,

icx +
= w and y z = Hence [4 5] -w x + -w x + y z + -y z = 1 + -w Hence # (ir + -w) + z (y + -y) = x + z = 1.

and w x
?y,
yz =
1.
x + -y
1.
#+3 =
is
the conclusion
if
Either
is
Z)
or
6?
is
//".
This dilemma
may = (1) w-x
be diagrammed
0,
we put our equations

0,
in
the equivalent forms
(2)
y-z =
(3)
-10
-y
0.
In figure 25,
-z
is
struck
FIG. 25
out with horizontal
lines,
y-z with
is
vertical,
is
-H>
-# with oblique.
That
everything which remains
either x or z
evident.
Example
(1)
(2)
4.
.1 is 7?
Either Either
Either
or
is
not
F.
7).
is 7)
is Z?
(3)
or E or E
is is
not
2
/?.
Let x
= A
(1)
is /*;
= C
is 7);
= E
0.
is
^.
(2) (3)
* + -2/= 1. y + s = 1, or -y-z + -3 = 1, or -xz

(1),
a;
= =
0.
By
+ -#
(2),
(3
Hence by
+ -3) = x + x + -yz = 1
z
And by
(3),
-x -y
0.
+ -y -3 = 1. = x + -y z (x + -x) = x + x -y Hence x + x -y z = x = 1.
-?/ z
+ -x -y
z.
give the categorical conclusion "A is 5", indi cating the fact that the traditional modes of conditional syllogism are by
Thus these three premises
no means exhaustive.
Applications of the Boole- Schroder Algebra
217
Example
5.
32
Assume the premises:

a necessary being, either the property of gravitation necessarily present, or it is necessarily absent.
1.
If
matter
is
is
2.
If
gravitation
is
necessarily absent,
and the world

exist.
is
not subject to
any presiding
3.
intelligence,
motion does not
If gravitation is necessarily present, If

If
vacuum
is
necessary.
4.
5.
vacuum
is
matter
not a necessary being. a necessary being, the world is not subject to a presiding
is is
necessary, matter
intelligence.
Let x
= Matter is a necessary being. = Gravitation is necessarily present. y z = The world is not subject to a presiding w = Motion exists. = Gravitation is necessarily absent.
t
intelligence.
A vacuum
are
f),
:
is
necessary.
The premises then

1
(1)
(2)
xc(y +
t
or
x-y-t =
0.
0.
(3)
(4)
(5)
z c -w, or tzw = y cv, or y -v = 0. v c -x, or v x = 0. x c z, or x -z = 0.
And
since gravitation cannot be both present

(6)
and absent,
0.
:
Combining these equations
x-y
From
between
-t
+ tz
w+
it
ij
-v + v x + x -z + y
first,
=
any
(7)
these premises, let
be required,
being",
to discover
collection
x, "Matter is
a necessary
it
and
y, "Gravitation is necessarily
present".
For
this purpose,
is
sufficient to discover
whether any one

since these are
of the four, x y
0,
x -y
0,
-x y
0,
or
-.T
-y
0,
the relations which state any implication which holds between x, or -.r, and ?/, or -y. This can always be done by collecting the coefficients of
y,
x -y, -x
y,
and -x
(7),
such as equation
32
comprehensive expression of the data, and finding which of them, if any, reduce to 1. But
-y, in the
See Boole, Laws of Thought, Chap. xiv.

s
The premises assumed
are supposed to be
borrowed from Clarke
metaphysics.
218

is
sometimes, as in the present case, this lengthy procedure
not necessary,
because the inspection of the equation representing the data readily reveals
such a relation.
From
Hence
If
if
(7), [5-72]
vx + -vy =
0.
(v
[1-5] v
x y + -v x y
+ -v) x y
x y
is
0,
or x c -y, y c -x.
matter
is
a necessary being, then gravitation

is
not necessarily present;
gravitation necessarily present, matter is not a necessary being. Next, let any connection between x and w be required. Here no such
is
relation
easily to be discovered
by
inspection.
Remembering that
if
0,
then a b
and a -b
(-y -t +
t
=
;
From
(7),
+ y -v +
z +
+ -z + y
t)
wx
-w x
(8)
y -v + y t) w -x + (-y -t + y -v + v + -z + y + (y-v + yt) -iv -x = +

(t
t)
Here the
coefficient of
wx
y+
reduces to
v,
1,
for [5-85],
t
-i)
and
-y
-t
+ -z
+ -z
and hence the
coefficient
)
is
But
[5-90]
(-y -t + y +
0, or
+ y + t + v + -z + y t. + -z + y t = l+v + -z + yt
1.
Hence w x =
c -x, x c -iv.
7-
-IV
-y-
--{
FIG. 26
219
None
being;
of the other coefficients in (8) reduces to
1.
Hence the conclusion

is
which connects x and

if
is:
"If
motion
exists,
matter
not a necessary
matter
is
a necessary being, motion does not
exist".
Further conclusions, relating other terms, might be derived from the same premises. All such conclusions are readily discoverable in the dia
gram
of equation (7).
In fact, the diagram
is
more convenient
for such
problems than the transformation of equations in the algebra. Another method for discovering the implications involved in given data is to state the data entirely in terms of the relation c and, remembering
,
that
-b
"If
ac6 and bcc,

to seek directly
then
ace",
as well as
"acb
is
equivalent to
c-a",
sitions which are in question.
any connection thus revealed between the propo Although by this method it is possible to
exists, the
overlook a connection which
danger
is
relatively small.
IV.
THE APPLICATION TO RELATIONS

is
The
application of the algebra to relations

is
relatively unimportant,
because the logic of relations
immensely more complex than the Boole-
Schroder Algebra, and requires more extensive treatment in order to be of We shall, consequently, confine our discussion simply to the service.
explanation of this interpretation of the algebra.
A relation,
etc.,
taken in extension,
is
the class of
all
couples, triads, or tetrads,
which
of"
have the property of being so related.

is
That
is,
the relation
is
"father
the class of
all
those couples,
(x\y)>
such that x
father
of y: the dyadic relation
the relation
R to
which have the

(x; y)
R is (.r; R y. The extension of a relation is the class of things y, We must distinguish between the class of couples relation.
the class of all couples
y) such that x has
and the
class of couples (y- x), since not all relations are

differs
symmetrical
and
.r
R y commonly
from y
R x.
Since the properties of relations,
so far as the laws of this algebra apply to them, are the
same whether they
are dyadic, triadic, or tetradic, etc.,

sufficient.
the discussion of dyadic relations will be
The
(x; y)
"
product
",
x S, or
S, will represent the class of all those couples

",
such that
all
xRy and
couples
;
the class of
R + S, will be x S y are both true. The "sum .r R y and y) such that at least one of the two,
of R, -R, will be the class of couples
x S
y, holds.
The negative
is false.
0;
y) for
which x
Ry
(t\
The
couples
null-relation, 0, will be the null-class of couples.
If
the class of
u) for
which
Ru
is
true,
is
a class with no members, and the
220

v
class of couples (u; w) for
which
Sw
is
true
is
also a class with
then
R
t
and S have the same extension.
It is this
no members, extension which repre

t
sents.
Thus
R =
signifies that there are no two things,
and
u,
such
that
Ru
is
true
that nothing has the relation

1, is
R to
anything.
Similarly,
the universal-relation,
course).
the class of
all
couples (in the universe of dis
The
for
inclusion,
RcS,
represents the assertion that every couple (x; y)
which x R y is true is also such that x S y is true; or, to put it otherwise, that the class of couples (x; y) for which x R y is true is included in the
class of couples (u; v) for
which u S
v is true.
Perhaps the most satisfactory
reading of
RcS
$".
"
is
The presence
of the relation
R implies
the presence of
the relation
signifies
R =
S, being equivalent to the pair,
RcS
and S c R,
that
and S have the same extension
that the class of couples
(x; y) for
which x
v is true.
Ry
is
true
is
identically the class of couples (u; v) for
which u S
It is
obvious that
all
the postulates, and hence
all
the propositions, of
the Boole-Schroder Algebra hold for relations, so interpreted.

1-1 If
and S are
true,
relations (that
is, if
there
is
a class of couples (x; y)
such that x
Ry is
and a
class of couples (u; v)
such that u S
-K
is
true),
then
R xS is a relation (that is, z) w R z and w S z are both true). If R and S be such that there is no couple (w; z) for which w R z and w S z both hold, then R x $ is the null-relation,
there
is
a class of couples (w;
such that
i.
e.,
the null-class of couples.
1-2
The
class of couples (x; y) for
which x
Ry
and x
Ry
both hold
is
simply the class of couples for which x Ry holds. 1 3 The class of couples denoted by R x S is the same as that denoted
by
SxR
1-4
namely, the class of couples

m
(x
y)
such that
xRy
y,
and
xSy
all
are both true.
The
class of couples (x
y) for
which
xRy,
x S
and x T y
hold
i.
is
identically the same in whatever order the relations be combined
e.,
Rx(SxT) = (RxS) xT. 1-5 R xO = e., the product

i.
of the class of couples for
which x
Ry
holds and the null-class of couples

1
is
the null-class of couples.

class of couples
for
For every relation, R, there is a relation -R, the which x Ry is false, and -R is such that:
1-61
If
the relation
is
Rx-S
Sy
is
is
null (that
is,
if
there
is
no couple such
is,
that x
Ry
true and x
is false), is
then
R xS = R
(that
the class of
couples for which x
Ry
true
;
identically the class of couples for which
Ry
and x S y are both true)
and
221
1-62
If
couples for couples for
R xS - R and Rx-S = R, then R = e., if the class of which x R y and x S y are both true is identically the class of which x R y is true, and if also the class of couples for which
i.
x x
R y is true and x S y is false is identically the class of couples for which R y is true, then the class of couples for which x R y is true is null. 1-71= -0 e., the universal class of couples is the negative of the
i.
null-class of couples, within the universe of discourse of couples.
1-8
R+S =
-(-Rx-S)-i.
e.,
the class of couples

y, is
(x\y)
such that
at least one of the two, x

of couples for
Ry
and x S
true
is
the negative of the class
which x
Ry
and x S y are both

i.
false.
1-9
for for
RxS =
R Ry
which x
Ris equivalent to R c S y and x S y are both true is

is
e., if
the class of couples (x y)

;
identical with the class of couples
which x
if
true, then the presence of
implies the presence of
and
x
the presence of
R
is
implies the presence of S, then the class of couples

true
is
33
(x; y) for
which x
Ry
identical with the class of couples for
which
Ry
33
and x S y are both

For a further discussion
true.
of the logic of relations, see
Chap,
iv, Sect. v.
CHAPTER
We
IV
SYSTEMS BASED ON MATERIAL IMPLICATION

are concerned, in the present chapter, with the
"
calculus of propo
its
sitions"
or calculus of
"material
implication",
and with
extension to
pro positional functions.

procedure, and
it is
We
shall
discover here two distinct
modes
by
of
part of our purpose to set these
two methods
side
side.
procedure takes the Boole-Schroder Algebra as its foundation, the elements of this system as propositions, and adds to it a interprets
first
The
postulate which holds for propositions but not for logical classes.
result
is
The
what has been

then x
called the "Two-Valued
Algebra",
because the
additional postulate results in the law:
and
if
=t=
0,
1.
For any x, if x =|= 1, then x = 0, This Two- Valued Algebra is one form of the
where x n
is
calculus of propositions.
The extension
<px
propositions of the
{
form
,
of the Two-Valued Algebra to an individual member of a class
a* 3 etc., gives the calculus of prepositional functions. composed of x II and 2 functions have a special significance in this system, and the relation
,
a>,
of of
"formal implication", it,
Hx
(<px
c\j/x), is
particularly important.
In terms
the logical properties of relations
including the properties treated
in the last
chapter but going beyond them
can be established.
This
is
the type of procedure used by Peirce and Schroder.
The second method

is
that of Principia Mathematica
begins with the
calculus of propositions, or calculus of material implication, in a form
which
simpler and otherwise superior to the Two-Valued Algebra, then pro ceeds from this to the calculus of propositional functions and formal impli
cation,
and upon
of
this last bases not only the

classes".
treatment of relations but also
the
"calculus
It is especially
important for the comprehension of the whole subject
of symbolic logic that the agreement in results of these
and the difference of method, two procedures, should be understood. Too often they appear to
the student simply unrelated.

I.
THE TWO-VALUED ALGEBRA

.
. .
a, b, p, q, etc., represent propositions, and a x b or a b represent the joint assertion of a and b, then the assumptions of the
If
the elements
summary
See Schroder, Algebra der Logik: n, especially Fiinfzehnte Vorlesung. is contained in Schroder s Abriss (ed. tiller), Teil n.
An
excellent
222
Systems Based on Material Implication
223
and p = 1, is true". Since and 1 are unique, it follows that any two propositions, p and q, such that and q = 0, p = or such that p = 1 and q = 1, are also such that = q. p = q, in the p
sent
"p
Boole-Schroder Algebra will all be found to hold for propositions, as was 2 As was there made clear, p = explained in the last chapter. will repre
is
false",
"p
algebra, represents a relation of extension or
"truth
value",
not an equiva
lence of content or meaning.
-p symbolizes the contradictory or denial of p. The meaning of p + q is readily determined from
its definition,
P+
p+q
"At
-(-P -q)
q
is
false",
is
the denial of
"p
is
false
and
q,
or
it
is
the proposition
least
one of the two, p and

is
is
true",
p+q
may
be read loosely,
"Either
p
is
true or q
is
true".
The
possibility that both
p and
q should
be true
is
not excluded.
equivalent to p q
pc
= p and
to
p -q =
0.
;;
material implication.
We
shall consider its properties
c q is the relation of with care later in

q
the section.
"It
is
false
For the present, we may note simply that p c that p is true and q false". It may be read
(materially) implies
q".
means exactly
p
is
"If
true, q
is
true",
or
"p
With the
interpretations here given,
all
the postulates of the Boole-
Schroder Algebra are true for propositions. Hence all the theorems will also be true for propositions. But there is an additional law which holds
for propositions:
p =
"The
(P
=
p
i)
is
proposition, p,
this that
is
equivalent to
true
".
It follows
immediately
from
-P = (-P
"-p
=
".
1)
(P
0)
is
equivalent to
is
false
It also follows that
-p = -(p =
1),
and hence
-(P
1)
is
(P
0),
and
-(p
0)
=
is
(p
1)
is
=
"p
1 is
false
equivalent to p
=
==
0",
and
"
p =
is
false
equivalent
to
1".
Thus the
is
calculus of propositions
a two-valued algebra:
every proposition
either
or
1,
We
may, then,
proceed as follows:
2
All the propositions of the Boole-Schroder Algebra
However, many
and
solutions, are of little or
of the theorems, especially those concerning functions, eliminations, no importance in the calculus of propositions.
224
which were given in Chapter II may be regarded as already established in the Two- Valued Algebra. We may, then, simply add another division of propositions the additional postulate of the Two-Valued Algebra and the
additional theorems which result
from
it.
Since the last division of the

shall
orems
in
Chapter
.
II
was numbered
8-,
we
number the theorems
of
this section 9
The
9-01
additional postulate
is:
For every proposition

for
p,
(p
1).
:
And
9-02
convenience we add the convention of notation
-(p
q) is
equivalent to p
of 9-01,
=|=
q>
As a consequence
-(p
we
shall
have such expressions as -(p
=
=|=
1)
1
and
and
=
0.
0).
9-02 enables us to use the more familiar notation, p
p*
It follows immediately from 9-01 that the Two-Valued Algebra cannot be viewed as a wholly abstract mathematical system. For whatever p and 1 may be, p = 1 is a proposition. Hence the postulate asserts that
any element,
interpretation
p, in the
system,
is
a proposition.
But even a
necessary
be abstracted from in one important sense no step in need be allowed to depend upon this interpretation. This is the proof
may
procedure we shall follow, though it is not the usual one. It will appear shortly that the validity of the interpretations can be demonstrated within
the system
itself.
In presenting the consequences of 9-01 and 9-02,
we
shall indicate
previous propositions by which any step in proof is taken, by giving the number of the proposition in square brackets. Theorems of Chapter II may, of course, be used exactly as if they were repeated in this chapter.
9-1
-p = (p =
[9-01]
0).
-p = (-p =
1).
And
[3-2]
-p =
1 is
equivalent to p
0.
9-12
-p = (p
[9-01]
+
=
1). 1).
p = (p =
(p
Hence
[3-2]
-p = -(p =
1)
(p
1).
9-13
(p*
1)
0).
9-14
(p
[9-M2] + 0) = (P =
[9-13, 3-2]
1).
9-13 and 9-14 together express the fact that the algebra
is
two-valued.
Every proposition
is
225
written the
.",
Up
logical
.
. .
to this point
relations
. .
.",
that
.
is,
,
"If
throughout Chapter II "Either then

.
we have
. .
.",
or
"Both
and
etc.,
not in the symbols of the system but just as they
in arithmetic or geometry or any other mathematical have had no right to do otherwise. That "... c ..." system. + is by inter and then is by interpretation or does not warrant us in identifying the "Either pretation
would be written
We
"If
".
."
",
.",
theorem
"If
a c
6,
then
-be
-a"
with
. . .
"
(a
,
c b) c
.
(-/;
.
c -a)
".
We
have
."in theorems with then had no more reason to identify c ..." than a geometrician would have to identify the period at "... end of a theorem with a geometrical point. The framework of logical the
"If
relations in terms of
which theorems are stated must be distinguished from
the content of the system, even
to interchange the joint But we can now prove with pcq, etc. We can then of p and q wath p xq, assertion p, demonstrate that if p and q are members of the class A then p c q is a And we can is equivalent to p cq. member of K, and that p, then
r
when that content that w e have a right

"If
<y",
is logic.
"If
</",
demonstrate that this
is
true not merely as a matter of interpretation but
by the necessary laws

"Both
of the
system
in
.
.
itself.
We
. .
can thus prove that writing

"Either
.
the logical relations involved
the theorems
,
or
.,"
...
and ..
.,"
"If
then
."in
terms of
+,
x,
c,
etc., is
a valid procedure.
in
which these things are proved are never needed here and their after, except in the sense of validating this interchange of symbols we need not give them any section number. Consequently interpretation.
The theorems
(1)
If
is
an element
in
A, p =
is
and
[9-01] If
If
is
is
an element in
= A, p = 1
/;
are elements in
is
an element
in
[9
A.
1]
/;
[1-0]
p is an element in A, -p an element in K.
q,
an element
in
A, and hence
(2)
The two, p and
are together equivalent to p x

[5-73]
q,
or p
q.
= (pq = 1). [9-01] pq = 1 and q = 1, and pair, p

(3)
If
pq =
is
is
equivalent
pair,
to
the
hence [9-01] to the
p and
in
q.
p and
[4-9]
if
q are
elements in A, then p c q
an element
A.
But
(4)
p cq is equivalent to p -q and q are elements in A, [1 6, p

"
0,
1
and hence
1]
[9-1] to -(p-q).
-(p -q)
is
an element
in
A.
-p
is
is
equivalent to
p
4=
is
false".
[9-12]
false",
-p =
(p
1),
and
[8-01]
"p
4=
is
equivalent to
=
"p
and hence
[9-01] to
is
false".
16
226
A
(5)
pc
is
equivalent to
"If
p,
then
1,
g".
[5-64]
"If
p eg
g".
gives
"If
p =
then q
=
it
1",
and hence
[9-01]
p,
then
p,
And = g
[2-2]
"If
then
g"
gives
peg,
for [9-01]
gives
"If
p =
g
1,
then
1",
and
(a)
Suppose as a fact p
1.
Then, by hypothesis,
1,
and
peg.
Suppose that p
={=
(6)
1.
Then
[9-14]
p =
0,
and
[5-63]
in
peg.
(6)
If
p and
[7-1]
g are elements in
K, then
p =
+
is
an element
0,
K.
[9-1]
p =
is
equivalent to
p-g -pg =
and hence
to-(p-g + -pg).
(7)
Hence
[1-6, 1-1, 3-35]

is
Q.E.D.
g".
p =
g is equivalent to
"p
equivalent to
"peg
[2-2]
p =
g is equivalent to
"p
and
cyj".
By
if g,
(5)
above,
p".
eg and
this
is
cp"
is
equivalent to
"p
"If
p,
then
g,
and
then
And
equivalent to
is
equivalent to
in
g".
(8)
If
p and
g are elements in
K, then p
g).
4= g is
an element
K.
[9-02] (p
g)
=-(P =
Hence, by
(9)
(6)
above and 1-6, Q.E.D.

"p
=|=
is
equivalent to
is
not equivalent to
g".
By
(10) p + g
is
(4)
and
(2)
above, Q.E.D.
"At
equivalent to
least

equivalent to
is
g, is
true.
[1-8]
p+q =
and
(2)
-(-p-g).
above, -(-p-g)
is false)
g, is
".
By
(p
(4)
is
is
"It
is
false
that
false
and
And
this
equivalent to
"At
least
one
of the two,
p and
true".
In consideration of the above theorems, we can henceforth write

"
".
for
"If
then
...","...
=
. .
..."
for
.",
".
is
equivalent
to
...","...
..."
for
"Either,
or
etc.,
for
we have
and the proved that not only all expressions formed from elements in relations x and + are elements in K, but also that expressions which in
volve
c, and =, and
=j=
are elements in the system of the

of
"If
.
Two-Valued
. .
Algebra.
of
"Both
The equivalence
. .
then
.
..."
with
".
.",
no longer a matter = (p = 1). Also, we can of interpretation but a consequence of 9-01, p go back over the theorems of Chapter II and, considering them as propositions then of the Two-Valued Algebra, we can replace etc.,
.
and
..."
with
".
.",
etc., is
"If
.",
227
by the symbolic equivalents.

corresponding theorem which
is
Each theorem not wholly in symbols gives a wholly in symbols. But when we consider
the Boole-Schroder Algebra, without the additional postulate, 9-01, this procedure is not valid. It is valid only where 9 -01 is one of the postulates
i.
e.,
only in the system of the Two-Valued Algebra. Henceforth we shall write all our theorems with pcq for
q for
"If
.
"
If p,
then
7",
p =
use
"p
is
equivalent to
r/",
etc.
But
".
in the proofs
.
we
shall frequently
then
..."
instead of
.
",
etc.,
because the
symbolism sometimes renders the proof obscure and makes hard reading. (That this is the case is due to the fact that the Two-Valued Algebra does
not have what we shall hereafter explain as the true
"logistic"
form.)
9-15
+
(p4=
1.
=
.9-16
q)
0.
Hence [9-13]0
(-p
4= 1.
=
If
=
q
q)
(p
=
1,
-?).
(1)
And
if
p = = 1, p
If
and p
[3-2]
= = -p
then q
4= 1
and
q.
[9-13] q
0.
0.
Hence -p =
then [9-13] p
= 0, and [3-2] -p = =h 1, (2) p 4= q = 1 = -p. Hence if p q, then q 4= 0, and [9- 14] q If -p = q and q = 1, then -p = 1, and [3-2] p = 0. (3)
and p
=j=
1.
Hence
[9-15]
If
4= q.
= 0. (4) -p = q and q 4= 1, then -p 4= 1, and [9-13] -p Hence [3-2] p = 1, and p 4= qBy (1) and (2), if p 4= q, then -p = q. And by (3) and (4), = q are equivalent. Hence p 4= q and q-p = q, then p And [3 -2] (-p = q) = (p = -q).
=)=
->
if
This theorem illustrates the meaning of the relation, =, in the calculus or 7; = If p 4= (/. then either p = 1 and q = of material implication. = 0, and if p = 0, then -p = 1. Hence and q = 1. But if p = 1, then -p
the theorem.
Let p represent
home".
"Caesar died",
died"
and
q represent
"There
"There
is is
no place no place
is
{
like like
If
"Caesar
is
not equivalent to
die"
home",
then
"Caesar
did not
is
is
equivalent to
{
"There
no place
like
home".
The equivalence
one of truth values
0} or
1}
not of content or logical significance.
9-17
p =
(p
1)
(p
4= 0)
(-p
0)
(-p
4= 1).
[9-OM3-14-16]
9-18
-p
(p
0)
(p
4=
1)
(-P
1)
(-P
0).
[9-M3-14-16]
228
9-2
A
(p
l)(p
[2-4]
0)
0. 0.
p-p =
is
And
[9-01]
false.
p =
(p
1);
[9-1]
-p = (p =
0).
No
proposition
(p
both true and
9-21
l)(p
0)
0.
9-22
(7;
= 0. And [2-4] -pp = l) + (p = 0) == 1.

[4-8]
[9-18J
-p =
(p
1);
[9-17]
p = (p
0).
p + -p =
is
1.
Hence [9-01-1] Q. E. D.
Every proposition
9-23

0)
== 1.
(p*
l)
(p*
[4-8, 9-01-1]
Theorems
of the
same
sort as the above, the proofs of
which are obvious,
are the following:
9-24
(p q)
9-25
= (p = l)(g = 1) = (p 4 0) (g = == - (p = l)(g 4= 0) = -(-p + -g) (p * 0)(g 1) = (-P + -g = 0) = [(p = 0) + (g = 0) = 0] = [(P * 1) + (g * 1) = 0], etc., etc. = (p + g = 1) = (p + g 4= 0) = (p = 1) + (g = 1) (p + g) = = -(-p -g) = [(p = 0)(g = 0) = 0] (P * 0) + (g 0) = [(P * l)(g * 1) 4= 1], etc., etc.
=
(p
1)
(p g
0)
=t=
0)
=(=
These theorems
illustrate the variety of
relation can be expressed in
ways in which the same logical the Two-Valued Algebra. This is one of the
redundancy
of forms.
defects of the
system
its
In this respect, the

neater calculus of
alternative method, to be discussed later, gives a

propositions.
much
.
We
turn
now
to the properties of the relation c
We
shall include here
some theorems which do not require the additional postulate, 9-01, for the sake of bringing together the propositions which illustrate the meaning of
"material implication".
9-3
(per/)
(-p +
q)
(p -q
0).
1).
[4-9] (peg) = (p-g = 0) == (-p+g = [9-01] (-p + g = 1) = (-p + g).

"p
materially implies
"It
g"
is
is
equivalent to
true
"Either
is
false or g is
true",
and to
is false
that p
and
false".
"
Since
pc
g has
been proved to be an element in the system,

g"
It is false
that p materially implies
may
be symbolized by -(peg).
229
9-31
-(p
cry)
[3 4]
(-p + q
q)
0)
=
-q.
(p-q).
-(-p +
</)
= p
And
[9
3]
-(p c
q)
-(-p +
q)
[9
"p
.02] -(-? +
(-/;+</
is
0).
does not materially imply

true",
q"
equivalent to
"It
is
false
that either p
is
false or q is
and to
"p
is
true and q
false".
9-32
(p
0)c(pc<7).
[5-03] Ocr/.
If
Hence Q.E.D.
ry,
is false,
the famous
proposition".
then for any proposition or notorious theorem:
p materially implies
false
q.
This
is
"A
proposition
implies any
9-33
(q=-
})c(pcq).
[5-01]
pel.
Hence Q.E.I).
"A
This
is
the companion theorem:
true proposition
is
implied by any
proposition".
9-34
-(pcq)c(p =
since
if
1).
The theorem
-(p
c<y),
follows from 9-32
then [9-32] p
q,
4= 0,
by the reductio ad absurdum, and [9-14] p = 1.

is
If
there
is
any proposition,
is
which p does not materially imply, then p
true.
This
simply the inverse of 9 32.

0).
A similar consequence of 9
=
0.
33
is:
9-35
-(pcq)c(q =
If
If
-(pcq), then [9-33] q =(= 1, and [9-13] p does not materially imply q, then q is false.
-(/;
9-36
q)
(7;
-q)
-(p c
cry),
0,
q)
c (-p c
q)
1
-(/;
q)
c (-p c
-q).
[9-34-35]
[3-2] If
If
-0;
then p
p =
1, -y;
and
if
= =
and
0.
=--
0,
then -q
1.
[9-32] If
[9-331 If
If
-p = -q =
0,
1,
then -p cq and then y;c-ry.

r/,
-;; c-ry.
p does not materially imply or denial, of and the negative

ry,
then
y>
materially implies the negative,

r/,
of
p implies
and the negative
of
p implies
the negative of
"The
ry.
If
"Today
is
Monday"
does not materially imply
made of green cheese", then "Today is Monday" implies "The is not made of green cheese", and "Today is not Monday" not Monday" implies "The moon is made of green cheese", and "Today is implies "The moon -is not made of green cheese". Some of the peculiar properties of material implication are due to the
moon moon
is
230
fact that the relations of the algebra
were originally devised to represent

exhibits properties of material
the system
of logical classes.
But 9-36
implication which have no analogy amongst the relations of classes.

is
9-36
a consequence of the additional postulate,

"is
p =
(p
1).
For
b,
classes,
represents
contained
in":
but
if
is
not contained in
it
does not
follow that a
of
b.
is
contained in not-6
may
be partly in and partly outside
9-37 -(pcq) c(qcp).

[9
36] If
"(p
q),
then -p c
q, if
-q,
and hence
[3
1]
p.
q,
Of any two propositions, p and

materially implies p.
p does not materially imply
then q
9-38
(pq) c[(pcq)(qcp)].
[9-24]
pq =
(p
l)(q
==
1).
Hence
[9-33]
Q.E.D.
If
p and
q are both true, then each materially implies the other.
9-39
(-p-q)c[(pcq)(qcp)].
[9-24]
-p-q =
(-p
l)(-q
1)
(p
0)(g
0).
Hence
If
[9-32]
Q.E.D.
p and q are both false, then each materially implies the other. For any pair of propositions, p and q, there are four possibilities:
1)
2) 3)
4)
= p = p = p =
p
1,
q q
q q
0, 0,
1,
= = = =
1:
true, q true.
false, q false. false, q true.
0:
1
:
p
p
0:
true, q false.
1
Now
in
the algebra,
cO,
1,
and
1;
but
cO
is false.
Hence
in
the four cases, above, the material implications and equivalences are as
follows
:
1)
2) 3)
pc pc
q,
q,
qcp, p = qcp, p =
q.
q.
4)
pcq, -(qcp), p 4= ~(pcq), qcp, p ^
q.
q.
This summarizes theorems 9-31-9-39.

the content or meaning of p and
"implication"
These relations hold regardless
of
q.
Thus p cq and p =
q are not the
and
"equivalence"
of ordinary logic, because, strictly

"propositions"
speak
"truth
ing,
p and
q in the algebra are not
but simply the
values"
of the propositions represented.
In other words, material impli-
231
cation and material equivalence are relations of the extension of proposi whereas the "implication" and "equivalence" of ordinary logic are relations of intension or meaning. But, as has been mentioned, the material implication, pcq, has one most important property in common with
tions,
"q

does not hold.
p"
in ordinary logic;
if
is
true and q false,
the relation of material equivalence, p connects a true proposition with a false one.
And
pcq
never
q,
These theorems should make as clear as it can be made the exact meaning and character of material implication. This is
important, since
many theorems whose
significance
would otherwise be very puzzling follow
from the unusual character of
this relation.
Two more propositions,

9-4
(p q
of
some importance, may be given:

[p
r)
(q
p c
r)
(q
r)]
[q
c (p c
q)
==
r)}.
[1-3]
pq =
qp.
Hence
[3-2]
-(p
-(q p), and
[-(/; q)
r]
= (p q c r), and [-(q p) + r] = And [3 41] [-(p q) + r = + -?) + r] = [-p + (-ry + + r] = Iqc(pcr)]. Similarly, [-(q p)
But
[9
3]
[-(p q) +
r]
(q
r)]
[(-/>
per). = (p c
(q
r)}
This theorem contains Peano

[(p
"If
Principle of Exportation,
[p
q)cr]c
c(qc r)]
r";
pq
implies
r,
then p implies that q implies
and
his Principle of
Importation,
[y;
"
c
r,
(q
r)]
if
[(p q)
r]
If
p implies that
[(pq)c
r]
q implies
[(p -r)
then
p and
c
q are
both true,
"
r is true.
9-5
-q]
[(q -r)
-y,].
[9 -3]
[(p q)
r]
= =
[-(p q)
[(-</
r]
=
-y;]
[(-p + -q) +
r]
[(-p +
r)
-ry]
[-Q; -r) + -q]
= =
[-(q -r) + -p].
[9-3]
[-(p- r ) +
-ry]
-[(p-r)c-g], and
[-(q-r)+-p]
l(q-r) c-p].
p and q together imply r, then if p is true but r is false, q must be false, and if q is true but r is false, p must be false. This is a principle first stated by Aristotle, but especially important in Mrs. Ladd-Franklin s theory of
If
the syllogism.
We
have now given a

to
sufficient
number
of
theorems to characterize the
Two-Valued Algebra
illustrate
the consequences of the additional
232
A
p =
(p

of
p c q. Any further theorems of the system will be found to follow readily from the foregoing. A convention of notation which we shall make use of hereafter is the
postulate
1),
and the properties
following:
sign
;
=, unless enclosed
,
in parentheses, takes precedence
over
any other sign a sign c unless enclosed in parentheses, takes precedence and the sign + unless enclosed in parentheses, takes over any + or x
;
precedence over a relation x
This saves
many
parentheses and brackets.
II.
THE CALCULUS OF PROPOSITIONAL FUNCTIONS.

VARIABLE
FUNCTIONS OF ONE
The
calculus of prepositional functions
is
an extension of the Two-
Valued Algebra to propositions which involve the values of variables. Fol 3 lowing 3Ir. jlussell, we may distinguish propositions from prepositional
functions as follows:
or false;
proposition
is
is
any expression which
is
either true
an expression, containing one or more variables, which becomes a proposition when each of the variables is re
a propositional function
placed by some one of
its
values.
"
There
is
one meaning of
1,
Today
2,
is Monday"
.
ambiguously Jan.
or Jan.
or
etc.
is
for which today denotes For example, when we say

",
Today
Jan.
1 is
is
Monday
implies
Tomorrow
Tuesday
we mean that
if
is
Monday, then Jan. 2 is Tuesday; if Jan. 2 is Monday, then Jan. Tuesday; if July 4 is Monday, then July 5 is Tuesday, etc. Today
variables,
and tomorrow are here

etc.,
whose values are Jan.
1,
Jan.
2,
Jan.
3,
that
is,
all
the different actual days.
variable sense,
"Today is Monday" is
today is used in this sometimes true and sometimes false, today

,
When
or more accurately,
it is
true for some values of the variable
and
false for other values.
"Today is Monday" is
here a propositional function.

Monday" for
There
today
is
is
a quite different meaning of
"Today is
which
In
not a variable but denotes just one thing

if
Jan. 22, 1916.
this sense,
"Today
is
Monday" is
its
true
it
is
always true.
its
"
It is either
simply true or simply

change.
false:
meaning and
"
truth or falsity cannot
today Today is Monday is a proposition. When Today, meaning Jan. 16, 1916, is one value of the variable today this value is substituted for the variable, then the propositional function is
this
For
meaning
of
turned into a proposition.

3 See Principles of Mathematics, Chap, vii, and Principia Mathematica, i, p. 15. Mr. Russell carries out this distinction in ways which we do not follow. But so far as is here
in question, his
view
is
the one
we
adopt.
Principia Mathematica
is
cited hereafter as
Principia.
233
We may
functions, in
use
^r,
t(x, y),
f (*, y z )
t
e tc.,
x, or
to represent prepositional
y, or x, y, and z, etc. These propositional functions must be carefully distinguished from the functions discussed in Chapter II. We there used /, F, and the Greek
which the variable terms are
x and
capitals, $, *, etc., to indicate functions;

letters.
here
Also, for
any function
<px,
of one variable,
x.
we use only Greek small we here omit any parenthesis
around the variable
\l/y,
f(x), V(x, y), etc., Chapter II are confined to representing such expressions as can be formed from elements in the class A and the relations x and + If x and y in *(*, y) are logical classes, then *(.r, y) is some
.
in
logical class,
such as x + y or a x + b -y.
Or
if
x in
/(.r)
b.
is
a proposition,
then
/(.r)
is
some proposition such as

$(x
t
functions,
<px
<px,
y},
(x, y, z),
etc.,
propositional are subject to no such restriction.

it
a x or -.r +
The
becomes a proposition when a- is replaced by one of its values, but does not necessarily become any such x proposition as a x or -.r + b. x is a citizen of y, Monday, these are y is between x and z
propositional functions.
is
typical
They
are neither true nor false, but they
become
either true or false as soon as terms denoting individual things are sub stituted for the variables x, y, etc. All the functions in this chapter are
such propositional functions, or expressions derived from them.
A
is
fundamental conception
"range
of the theory of propositional functions

".
is
that of the
of significance
The range
class,
of significance of a function
determined by the extent of the

its
or classes, of terms which are

x, in
,
values of
<px,
variables.
,
All the
terms which can be substituted for

^.r.
and make sense

mortal
constitute the range of

is
If
<px
be
is
mortal
the range of this function
the aggregate of
all
the individual terms for

"range
which x
is
is
is
Thus the
"universe
of
significance"
is
to propositional functions
what the
of
discourse"
to class
terms.
Two
propositional functions,
<px
and
<px,
r/^,
is
may
be such that the
class of values of
x in
<px,
or the range of
identical with the class of
values of y in
\J/y,
or the range of $y.

.r
different ranges of significance,
is
Or the two functions may have a man and .r is a poet will have the
true will differ.
.r
same
for
is
range,
though the values of

is
.r
for
which they are

is
Any
.r
which x
man
is
either true or false,

.r
also such that

.r
is
a poet
But some
s
.r
for
which
is
is
a poet
.r
is
either true
or false are such that

.r
x precedes
nonsense,
is
a poet
and
precedes x
have different ranges. 4
It is
important to note that the
4 According to Mr. Russell s "theory of types" (see Principia, i, pp. 41-42), the one fundamental restriction of the range of a propositional function is the principle that nothing
234
range of
function.
is
A
<px

x,
determined, not by
but by
<p.
<px
and vy are the same
we have a prepositional function of two variables, say x is a citizen of y we must make two substitutions in order to turn it into a proposition which is either true or false. And we conceive of two aggregates or classes
If
,
the class of values of the

variable, y.
first
variable, x,
and the
class of values of the
second
These two
classes
It
may,
for a given function, be identical, or

"John
they
may
be different.
Turkey" is
depends upon the function.

"Turkey is
Jones
is
citizen of
Jones" is
either true or false;

"3
a citizen of John
is
nonsense.
3".
"5
precedes
is either true or false, as But precedes The range of x and of y in $(x, y) depends upon
5"
also
^,
not
upon x and
y.
<px
xz
is
is by x\, x 2 convenient method of representing the values of x in This is not to presume that the number of such values of x in etc.
<px
finite, or even denumerable.
Any
If
sort of tag
all
,
these values as individual would serve

of
.TI,
,T 2 ,
which would distinguish the uses which we shall make

,
3,
etc., equally well.
<px-i,
<f>x*,
<p%3,
etc., will
x, x 2 x z be propositions; and
is
etc.,
<px
are individuals, 5 then

will
n is
be a proposition.
the values of x
,
<pxz
is
a proposition about a specified individual;
<px
a proposition about
if
a certain individual
in
\I/(x,
which
.r 3 ,
not specified. 6
Similarly,
y) be
,
Xi, x*,
etc.,
,
and the values

etc., are
of y be yi, y z ys, etc.,
then
$(%*>
2/s),
t(x z y n ), t(x m y n ),
propositions.
II
We
shall
now make
new
use of the operators
and
S, giving
them
a meaning similar to, but not identical with, the meaning which they had in Chapter II. To emphasize this difference in use, the operators are here
set in a different style of type.
We
x
.
shall let
2x
<px
represent
<pXi
<px
<f>Xz
to as
many
<pxi
terms as there are distinct values of x in

x
<px
<px.
And U x
<px
will represent
z
T
x
e
<px
to as
many
terms as there are distinct
values of x in
<px.
(W
have heretofore abbreviated
ax6
to a b or a -b.
is
But where prepositional functions are involved, the form
of expressions
that presupposes the function, or a function of the same range, can be a value of the func It seems to us that there are other restrictions, not derived from this, upon the tion. range of a function. But, fortunately, it is not necessary to decide this point here. 5 which is the only "Individuals" in the sense of being distinct values of x in
<px
conception of
6
"individual"
which we require.
<px
It
may be
is
urged that
n is
not a proposition but a prepositional function.
The
most difficult, and we cannot enter upon it. But this much may be said: Whenever, and in whatever sense, statements about an unspecified individual can be certain gentle If any object to this, we shall reply n is a proposition. asserted, man is confused". Peirce has discussed this question most acutely. (See above, pp.
question
"A
<px
93-94.)

likely to
235
chapter, always
be complex.
write
"products"
Consequently we with the sign x .)
shall,
in
this
might be an infinite set of values of x in does not affect the theoretical adequacy of our definitions. For nothing here depends upon the order of m ^x n p and it is only required that the values of x which are distinct should be identifiable or The ob "tagable".
<px <px
The
fact that there
<px
jection that the values of x might not be even denumerable
is
more
serious,
but the difficulty

Since
<pxi,
<px
may
z,
<f>x
be met by a device to be mentioned shortly. 3 etc., are propositions, #TI + 2 + 3 +

,
<px <px
is
proposition
the proposition,
"For
"Either
<pxi
or
<px
or #r 3 or ...
<px
etc.".
Thus
2x 2x
of
all
<px
represents
is
some value
Similarly,
3,
of x (at least one),
is
true".
And
"For
<px
a proposition.
^.TJ
<px
^.r 3
x ...
is
the joint assertion
<ATi
and
<px
and
<px
<px
etc.
",
Thus H t
<px
represents the proposition
values of x,
true",
is
true
We may
"
translate
<px
Z x ^.r loosely by
true".
"
<px
is
sometimes
and
lation fails of literal
by inasmuch accuracy
<px
nx
loosely
is
always
This trans
<px
as the variations of x in
may
not be confined to differences of time.
The conception
3,
of a prepositional function,
<px,
and
of the class of values
of the variable in this function, thus give us the

<px
(px n ,
2x
<px,
and
U x(px.
of proposition, Since the laws of the Two-Valued Algebra
new types
hold for propositions generally, all the theorems of that system will be true when propositions such as the above are substituted for ...
a, b,
p, q,
etc.
(We must,
of course,
remember that while
a, b,
... p, q, etc., in
the
Two-yalued Algebra represent propositions, x in yx, tion but a variable whose values are individual
to follow,
etc., is not
things.
a proposi In the theorems

in
we
shall
sometimes need a symbol for propositions
which no
variables are specified.
To avoid any
possible confusion,
we
shall represent
such propositions by a capital
letter, P.)
We
may, then, assume as already

cr,
proved any theorem which can be got by replacing any proposition of the Two-Valued Algebra, by
6,
<px
...
n,
p, q, etc., in
<px,
<px
3,
2x
or
Ti x
<px.
Additional theorems, which can be proved for propositions involving values of variables, will be given below. These are to be proved by reference to
Chapter II and in Section I of this chapter. As before, theorem by which any step in proof is taken will be given in square brackets. Since the previous theorems are numbered up to 9-, the additional theorems of this section will be numbered beginning with 10earlier theorems, in
the
number
of the
One
will
additional assumption, beyond those of the
Two- Valued
Algebra,
be needed.
The
propositions which have been proved
in sufficiently
236
general form to be used where

are in question
all
sums and products
of
more than three terms
require for their demonstration the principle of

If,
mathe
matical induction.
of this section,
then, AVC wish to use those theorems
in the proofs
we
values of x in
<f>x,
by the difficulty that the number of may and hence the number of terms in 2 x #r and H x
are confronted
<px
not be
finite.
And any
use of mathematical induction, or of theorems

for proof, will
dependent upon that principle

nection.
then be invalid
in tttis
con
Short of abandoning the proposed procedure, two alternatives are open to us: we can assume that the number of values of any variable in a propositional function is always finite; or we can assume that any
law of the algebra which holds whatever finite number of elements be involved holds for any number of elements whatever. The first of these assumptions
would obviously be
false.
But the second
is
true,
and we
shall
make
it.
This also resolves our difficulty concerning the possibility that the number of values of x in might not be even denumerable, and hence
<px
that the notation

inadequate.
in
<pxi
<px
<px
and
<pxi
x
if
tpx 2
<px
might be
of values of
"2
We
(f>x
can make the convention that

<px,
the
<px
number
+
.
.
any function,
x
<px
be not
or
finite,
<fXi+
<px
or
<px,
and
<?Xi
Hx
<px,
shall
be so dealt with that any theorem to
be proved will be demonstrated to hold for any finite number of values and this being proved, our assumption allows us to extend the of x in
<px\
theorem to any case in which the values of the variable in the function are This principle will be satisfactorily covered by the infinite in number. x shall always + and convention that 3 x 2 x 3 + 2 +
. . . . . .
<pxi
<px
<px
<pxi
<px
<px
be supposed to have a finite but undetermined
number
of terms,
and any
theorem thus proved

distinct values of
shall
be presumed independent of the number of

operative, will be
7 any variable, x, which is involved. This postulate, and the convention which makes
it
supposed to extend also to functions of any number of variables, and to

sums, products, and negatives of functions. Xo further postulates are required, but the following definitions are
needed
10-01
2p.r
10-02 10-03
7
H(px
-<p.r
= 2x = Hx
=
=
<px
<pxi+
<px
+
x
tftf 3
+ ....
x
.
Def.
=
<px
<pxi
(f>x
<pxz
..
Def.
-{<f>x}.
Def.
is
This procedure, though not invalid,
far
from
ideal, as are
many
other details of
this general
criticisms together in the last section of this chapter. But it is a fact that in spite of the many defects of the method, the results which it gives are without exception valid.
method.
We
shall gather the
main

10-031
- xn 9
237
=
-{<px
n }.
Def.
Def.
Def.
10-04
10-05
-U x
-2 x
<px
<px
= -{n x = -{S,0*r}
<px}.
The
one
merely serve to abbreviate the notation. Elementary theorems concerning propositions which involve values of
variable are as follows:
2<px
last four
10-1
-n-^.r.
<f>xi+
[5-951]
<f>x+
<px
...
-{-^.n x-^.T 2 x-î-3 x
J.
10-12
U<px
-2-<px.
[5-95]
îx^r
"For
10-1 states that

denial of
"For
some values
<px
of x,
is
<px
is
true"
is
equivalent to the
"For
all
is
values of x,
true"
false".
10-12 states that
all
"For some values These two represent the extension of De Morgan s Theorem to propositions which involve values of variables. They might
<px
values of x,
<px
is
equivalent to the denial of
of x,
is
false".
be otherwise stated:
false that
"It
is
true that
all
is
"
is
is
equivalent to
"
"It
is
some x
is
is
not
";
and
is
"It
true that some x
is
is
equiva
lent to
"It
false that all

n.
not
10-2
Il(px
<px
[5-99]
<pxi
x x
x
<px-~
x x x
<**3
x x
c c c
<px
and and
10-21
<px
<pxi
(px%
<^r
<px
<px%
<px!
<px
^.r 3 ,
etc., etc.
c2<f>x.
[5-991]
^c^+^roH.^*...
<pXz
and
c c
(pXi
<pxi
(px^
<px$
+ +
. . .
and
<pXs
+ pxz +
<px
etc., etc.
it is
By
of
10-2,
or
if
<px
is
true for
all
values of
true of
.r,
then
true for
any given value

10-21, If
.r,
<px
.r,
"What is
true of
all is
.r,
any given
It
one".
By
is
true for one given value of

is
then
it is
true for some value of
or
"What
true of a certain one
is
true of
is
some".
implication stated by 10-21
reversible.
<px
might be thought that the But we do not have 2^.r c ^.r n

,
because
<px
example,
let
may =
<^.r
be
<p.r
2,
and 2
<px
would not hold generally.
For
"Today (x) is Monday".
Then
2^.r will
mean
"Some
day
is Monday",
"Today
but
<px
n will
mean
"Today
(Jan. 1)
"Some
is Monday",
or will
mean
(Feb. 23)
is Monday", etc.
day
is Monday"
does
238
not imply
"
.1
Jan.
1 is Monday",
and does not imply

is
"Feb.
23
is
Monday
n
"-
does not imply that any one given day

certain value of
.r"
Monday.
x n in
<px
means
"
in a sense
which
<px
is
not simply equivalent to
"some
value of
respect.
.r".
Xo
translation of
will give its exact significance in this
10-22
II<>.r
cS^.r.
[5-1, 10-2-21]
Whatever
10-23
is
true of
is
all is
true of some.
"Whatever
3
H<px
equivalent to
value of x, in
(<pxi
<px,
xn x
.
may
.
be,
<px
".
Ii<px
=
<pxi
<px
x
<px
<px
x
<^r
=
. .
<px
<px
=1)
to
[9-01]
set
And
[5-971]
<pxi
x
1,
==1
is
equivalent
the
=
<pXi
1,
<pX
I,
pXz
....
<px
And
[9-01]
tt<px
<px
I is
equivalent to
Hence
is
equivalent to the set
<px
lf
<px
z,
<p.T
3,
....
This proposition is not tautological. It states the equivalence of the x tpxz x with the system of separate propositions 3 x product
<pxi
. .
<f>x
<pxi,
<f>x
z,
<f>x
s,
etc.
It is
tt<px
by virtue
as
"For
of the possibility of this proposition
that the translation of
all
values of x, ^x
is
true"
is
legitimate.
In this proof
we make
use of the principle,
p =
(p
1)
the only case in
which
it is
directly required in the calculus of propositional functions.
we can pass directly from any theorem of the TwoBy Valued Algebra to a corresponding theorem of the calculus of propositional functions. If we have, for example, pc.p+q, we have also "Whatever
virtue of 10 23
value of x, in
10-23,
<px,
xn
may
be,
<px
p.r n
P".
And hence we
importance of
have, by
this:
it
Ux
[<px
<px
+ P].
We
shall
later
see the
gives us,
for every
theorem concerning
"material
implication",
a cor
responding theorem concerning
"formal implication".
Next, we give various forms of the principle by which any proposition may be imported into, or exported out of, the scope of a II or S operator.
10-3
Z#c
2
<px
P = S^r + P). + P = +
(
<.ri
<px<2
<px
+...) +
+ P) +
P
...
P) +
(^
(^3 + P)+
[5-981]
10-31
P+Zô- - ? X (P+
Similar proof.
p.r ).
10-3
may
be read;
"
Either for some
x,
<px
is
true, or
is
true
is
equiva-

lent to
239
10-31
For some
"Either
x, either
is
<?x
is
true or
is
true
".
And
is
may
be
read:
P
is
true or, for some x,

<px
<px
is
true
equivalent to
For
some
10-32
x, either
true or
(<^.r
is
true
".
U<f>x
P = nx II (pX + P =
+
+ P).
<pXz
(<pXiX
<pX
P
P) x... [5-941]
10-33
P+tt<px
= U X (P +
tpx).
Similar proof.
<(
Either
P is true
is
or, for
<px
every
true."
x,
<px
is
true"
is
equivalent to
"For
every
x,
either
true or
is
10-34
S X (^ + P) = S X (P
[4-3]
2<px
+*>*).
+P = P+
2<px.
Hence [10-3-31] Q.E.D.
10-35
ILOi
+ P)
= n x (P+
<px).
[10-32-33]
is
Exactly similar theorems hold where the relation of the two propositions x instead of + The proofs are so simple that only the first need be
.
given.
10-36
2<px
xP = S*(#E xP). 2 px x P = ^.TI +

(
<f>x
<px
P
xP) +
. . .
=
(<^.ri
xP) +
(^-2
xP) +
(cp.i-3
[5-94]
"
<px
is
true for some x, and

are both
true".
is
true",
is
equivalent to
"For
some
.r,
<px
and
10-361
10-37
10-371
- 2 X (P x x P). n^ xP = n Pxn^r = n x (Px^r).
xS^.i-
<r>.r).
j:
<px
10-38
10-381
S z (^ xP) = S,(P x
^.r).
Hx
(<px
xP) = n x (P
<^.r).
should perhaps expect that a proposition, P, might be imported into and exported out of the scope of an operator when the relation of P But here the matter is not to the other member of the expression is c
.
We
quite so simple.
240
10-4
A
Pc?<px
= S,(Pc
<px).
[9-3]
PC
=
2(f>x
-P+2<px
= -P
[5-981]
= (P c ^rO = S X (P c
The
But
it is
+ (P c
^.r 2 )
+ (P c
^.r s )
[9
3]
<px)
relation
c, in the above,
is,
of course, a material implication.
tedious to read continually
"p
materially implies
9",
g".
We
g".
shall,
then, translate
pcq
"P
10-4 reads:
"For
implies simply by that for some implies

"p
or
<^.r
by
is
"If
p,
then
is
.r,
true"
equivalent to
some
.r,
implies that
:
is true".
^>.r
This seems clear and obvious,
but consider the next

10-41
Stf>.rcP
IlzO.rcP).
U-<px
[9-3]
S^ccP = -S#r + P =
2
P
. .
[10-12]
= (-î x-^.r X-^.TS x .) + P = (-^d + P) x + P) x (-^3 + P) = (^.TiCP) X(^.T CP) X(^.T CP)... = U x (<pxcP)
(-<pxz
[5-941]
[9-3]
"
<px
is
true for
P".
some x
implies
P"
is
equivalent to
"For
every
x,
<p
implies
It is
If
the
first also:
<px
easy to see that the second of these two expressions gives is sometimes true, P must always implies P, then if
<px
be true.
It is
it
not so easy to see that

"If
<px
2<px
can put
thus:
P".
is
ever true,
P gives then P is
c
tt x (<px
c P).
But we
"
true"
must mean
<px
always implies
10-42
PcU<px
=
c
Tl x (Pc<px).
[9-3]
U<px
= -P + Il^.r = (-P +
[5-941]
==
-P +
(<pxi
<pxi)
x (-P + x (P c
<px
x ^r 2 x 3 x x (-P + #c 8 ) x z)
.
.)
...
- (P c = Ut (P c
"P
<pxj
<px
2)
x (P c ^c s ) ...
[9-3]
<px)
implies that
<px".
<px
is
true for every
x"
is
equivalent to
"For
every x
implies
10-43
n^TcP = 2 = -n^ + P [9-3] n^rcp

x (<pxcP).
S-^.T +
+<X3
P
. .
[io-i]
.
<X
+-
.1*
241
= (-^1
+ P) +
(-^
(<^
+ P) +
(-<pOr
+ P) +
+
[5-981]
C P) +
C P) +
(^3 C P)
[9-3]
"
<px
is
true for every x

implies
P".
implies that
is
true"
is
equivalent to
"For
some
x, (px
At
first
sight this
theorem seems to commit the

"fallacy
"fallacy
of
division"
going one way, and the
of composition"
going the other.
It suggests the ancient
and baldness.
out".
Suppose
<px
be
"If
x
is
is
example about the separate hairs a hair of Mr. Blank s, x has fallen
And
of
let
be
"Mr.
Blank
bald".
Then
H<px
cP
will represent
is
bald".
"If
all
Mr. Blank
hairs have fallen out, then

"There
Mr. Blank
And 2 x (#rcP)
that
Ti(px
if
will represent
is
some
is
hair of
bald".
Mr. Blank
such
this hair has fallen out,
Mr. Blank
(<px
In this example,
their equivalence
is
cP is
obviously true, but 2 x
cP)
is
dubious, and
seems likewise doubtful.
The explanation
".
of the equivalence
<px
this:
"
we
is
here deal with material implication, and false that n is true but P is false)
(<px
"
P means
means,
simply
It
U<px
cP
in this
example,
is
It is false
bald";
that
all
Mr. Blank
hairs have fallen out but

"There
Mr. Blank
s
not
and 2 I (#rcP) means

is
is
some one
of
is
Mr. Blank
is
hairs
such that
This hair has fallen out but
Mr. Blank
not bald
false".
No
necessary connection
predicated between the falling out of any single

is
hair
and baldness
material implication
last four
not that type of relation.
If
we compare the
theorems,
we observe
that an operator in
the consequent of an implication is not changed by being extended in scope to include the whole relation, but an operator in the antecedent is changed
from
to
-<p
to 2,
f
from 2 to
II.
This
is
due to the fact that
pcqis
equivalent
where the sign of the antecedent changes but the consequent remains the same; and to the law -n() = 2-(), -2() = n-().
+ q
The above
tional function
principles, connecting
any proposition, P, with a preposi
used in later proofs. In fact, all the proofs can be carried out simply by the various forms of this principle and theorems 10-1-10-23. Since P, in the above, may be any propo
and
its
operator, are
much
sition,
#r n 2#r, II#c, etc., can be substituted for P in these theorems. In order that + \l/x) and x^j-) are, of course, functions of x. must be significant and $x must be significant, + \I/x) be significant, ( par Such or $x" have meaning. further requisite that "Either and it is
,
(<px
(<px
<?x <px
considerations determine the range of significance of complex functions A value of x in such a function must be at + \I/x) and like x^.r).
(<px
(<px
17
242
once a value of x
(<px
A
in
<px

of x in fa:
and a value
x n in
<px
and
in
fa n
in
+ fan), denotes identically the same individual.
10-5
<px
+ 2 fa
= Sx
^ + #r)
is
Since addition
associative
and commutative,
"
Either for some

<px
.r,
^,
or for
some
x,
fa"
is
equivalent to
"For
some
x,
either
or
fa".
If it
be supposed that the functions,

i.
<px
and fa, may have

is
different
ranges
e.,
that the use of the same letter for the variable
not indicative
(<p.r
of the range
then S^.r + 2 fa might have meaning when 2 x
+ fa) did
not.
will
But
in
in
not have meaning.
such a case the proposition which states their equivalence We shall make the convention that x n in n
<px
and x n
wherever
<px
n +fa n ) and n xfa n ), but fa n are identical, not only in and fa are connected, as in S^.r + S#r. Where there is no
(<px
(<f>x
such presumption, it is always possible to use different letters for the But even without this convention, the above variable, as S#e + 2^y.
theorem
will
always be true when
it is
significant
i.
e., it is
never
false
and a similar remark applies to the other theorems

10-51
of this section.
<px
Ufa = n z (
is
<px
x fa)
Since x
associative
and commutative, similar
proof.
it
We
"For "For
might expect 2<pxx2fa some .r, x is ugly, and for some
= 2 x (<pxxfa)
x, x
to hold, but
does not.
to,
is beautiful", is
not equivalent
some
x,
is
ugly and x
is beautiful".
Instead of an equivalence, we
have an implication:
10-52
Sx (
[5-21
<px
x fa)
n
( (p.i-1
x fa
,
(<px
xfa n ) c
<p.r
and
(<px
xfa n )cfa n
2,(<px
Hence Hence
Similarly,
x,
[5-31]
[5-34]
U<px
2 x (^.r xfa) cS^.r, and 2x xfa) c2<px xZfa

(<px
xfa) c^fa
Hfa = U x
is
(<f>x
+ fa)
",
fails
is
.T
to hold.
"Either
for every
is
ugly, or for every x, x

is
beautiful
".
either x
ful.
ugly or
.r
is
beautiful
Some
not equivalent to, "For every x, may be ugly and others beauti
But we have:

10-53
II
243
<px
Ufa ctt x
<px
(<px
+ #r)
[5-21]
c(<px
n +\I/x n ),
(<px
Hence
Hence
[5-3]
[5
H<px
cH x
+
and $x n c \f/x), and Il^.r

<px
cU x ((px +
lemma
in
if/x)
33] II
<px
n^x c n x (
+ $x)
In the proof of the last two theorems,

of writing
(<px
we
write a
for
<px
instead
it
for
<px
<pxi,
for
<pxz,
for
<px
s,
etc.
For example,
10-52 we write
x^.r n ) c
n,
instead of writing
(<pxi
<px3,
etc., etc.
The
proofs are
somewhat more obvious with
this explanation.
This method
of writing such
lemmas
will
be continued.
<px
With two
prepositional functions,
and
\f/x,
we can form two
impli
S x (#rc^) and IL^.r c i^.r). But 2 x (#rc#r) states is false or \j/x is true: is a value of x for which either only that there and this relation conveys so little information that it is hardly worth while
cation relations,
<px
to study
its
properties.
is
Ux
The
(<px
c^.c)
the relation of
is
<f>x
"jormal
implication"
is
"For
every
.r,
at least one of the two,

negative of
"It
false
and $x
true
is
a true statement"
(
<px
Ux
(<f>x
c^x)
is
S x (^.r
is
x-^.r), so that
.r
Ux
is
i//.r)
may
l
also
be read
is
false that there
any
such that
"At
<px
true and $x
false".
The material
false
implication,
is
peg,
states only
statement";
least
"It
one of the two,

false that
is
and
q
true
is
a true
or,
n
is
is
true
least
and
false".
The material
is
<p
implication,
<px
c\l/x n
states only
,
"At
one of the two,

or
"It
false of
<px
x n and
is
true of x n
false".
is
a true
statement";
is false
that
is
true and
^.r n is
But the formal impli

it
cation, n s (^.rc^.r), states that however x n be chosen,

is
is
false that
is
<px
true and $x n
is
false
in the
whole range of
false.
<px
and $x, there

II X
not a
case in which
<px
is
true and $x
To put
it
another way,
(^
^.r)
means
"Whatever
has the predicate ^ has also the predicate
^"
plies"
of "im This relation has more resemblance to the ordinary meaning it should than material implication has. But formal implication,
be remembered,
nx
(>.r
i//.r)
is
material implications; simply a class or aggregate of c ^.ri, ^.r 2 c ^.r, ^.r 3 c tx~, simply the joint assertion of ^,r,
is
is
etc.,
s
where each separate assertion

The whole question
"implies",
a material implication.
ing of
of material implication, formal implication, is discussed in Section v of Chap. v.
and the usual mean
244
A
The
properties of formal implication are especially important, because

in the calculus of classes
upon this relation are based certain derivatives and in the calculus of relations.
10-6
nârcâ-) = U x
[9
(-<px
+ fa)
<pX
= Hx
-(<px
x-fa).
n
3]
(pX n
Cfan = Q.E.D.
n
+ fan
-(
<pX
X -fa n )
Hence
10-61
[10-23]
tt x (<pxcfa)
c(<px
cfa n ).
materially implies fa n
[10-2]
If (px
formally implies fa, then

[II X (
<px
<px
10-611
c $x) x
10-61]
<px
n]
c fa^
[9-4,
If
<px
formally implies fa and $
is
true of
ar n ,
then ^
is
true of x n
This
is
is
for every x, a; one form of the syllogism in Barbara: for example, man implies x is a mortal and Socrates is a man, then Socrates
"If
,
is
mortal".
10 -62
nx nx
( ??ar
c #r) c S x (
<^x
c fa)
[10-22]
10 63
<px
c fa)
c(Il<pxc
Ufa).
then
<px
[10-61] If
U x (<pxcfa),
cfa n
Hence
10-631
[5-3]
Q.E.D.
[U x (<pxcfa)
[9-4,
xn^cnô*.
<px
10-62]
is
If (px
always implies fa and
always true, then fa

.
is
always true.
10 -64
nx
<^.r
c fa)
c(2<px
c.
2 fa)
[10-61, 5-31]
10-641
[n x (
<f>x
cfa)*2
10-64]
<px]
c 2 fa.
[9-4,
If (px
always implies fa and

[n x (
<px
<px
is
sometimes true, then fa c
is
sometimes
true.
10 65
c fa) x
If
n x (^.T
fa;)]
nx
^.T
far).
[10-61]
iA.r n
n x (^.rc^.iO and U x (fa

.r,
far),
then
<px
cfa n and
far n .
Hence
[5-1]
whatever value of
xn
may
be,
<^.r
far B .
Hence
[10-23]
n x (ârcfar)
245
It is
This theorem states that formal implication

(
is
<px
= x is a another form of the syllogism in Barbara. For example let 10-65 will then read: Greek \f/x = x is a man and x = x is a mortal
,
l "If
for every x,
is
a Greek implies x
,
is
man
#
and
for every x,
x x
is
is
man
implies
".
.r
is
a mortal
then for every
x,
is
a Greek implies
mortal
10-65
10-651
may
also be given the form:
Ux
(<px
c #c) c [n x (#r c
10-65]
x)
Ux nx
(<px
fr)].
[9-4,
10 652
U r (ifrx
[9-4,
fa;)
c [U x (
<px
c $x) c
<px
far)].
10-65]
10 66
nx
#r c
[3-1]
= n x (-#c c - #x) (^ci/O - Htfc -?.!)

i/a*)
.
Hence
[2-2, 5-3]
Q.E.D.
Any
further theorems concerning formal implication can be derived
from the foregoing.

"Formal equivalence" is
reciprocal formal implication, just as material
equivalence
is
reciprocal material implication.
The
properties of formal
equivalence follow immediately from those of formal implication.
10-67
Ux
to
=
(<px
jx)
[tt x (vx
c $x) xll x (tx c
<px)].
Whatever value of x, x n may the pair, n c \//x n and \f/x n c

<px
be, [2-2]
n
.
<px
is
equivalent
<px
Hence
10-68
[U x
[10-23]
Q.E.D.
=
(<px
jx)
xn x (^.r =
f-r)]
cn
x (^.i
T-r).
<px
Whatever value of .r, .r n may = rr. Hence [10-23] Q.E.D.
be,
if
<px
= ^x and
i:
==
far,
then
10-681
Ux
=
(<px
^.r) c[II,(^.r
far)
[10-68, 9-4]
10-682
n z (^.r -
f.c)
c[H x (^r =
^.r)
[10-68, 9-4]
Formal equivalence, as indicated by the

relation.
last three
theorems,
is
a transitive
10-69
Ux
and
(<p
n x (^c -
[2-2,
10-61-62-63]
246
10-691
A
n(^c =
[3-2,
If
$x)
= Ux
=
(-<px
-#r).
10-23]
to investigate the propositions
(<px
we wish
which can be formed from

sig
functions of the type of

nificance of
<px
may
differ
xfy) and from that of

tf
(<px
+ ty), where the range of
\l/y,
we
find that these will involve
2 Jl^r; ty), H S x (^r; ^), etc. And these are special + \l/y) are special cases of a function of two variables. x\f/y) and
two operators
(<px
(<px
cases of f (.T, y).

in general.
Hence we must
first
investigate functions of two variables
III.
PROPOSITIONAL FUNCTIONS OF
Two
OR
MORE VARIABLES
y), gives
prepositional function of
<p(x
two
y n ),
variables,
<p(x,
the derivative
y)-,
propositions
m y n ),
,
U*<P(X,
S*2
tf
p(a:, y),
2 y n x p(a:,
etc.
The
range of significance of
<p(x,
<p(x,
y) will
comprise
all
the pairs
(x,
y) such that
y)
is
.T 3 ,

etc.,
s
We
i/
here conceive of a class of individuals,

,
#1, xz,
and a
class of individuals, y lf y,, y s
etc.,
such that for any
one of the x
and any one
of the
s,
<p(x,
y)
is
As has already been pointed out, the function may be such that the class of values of x is the same as the class of values of y, or the values of x
may
be distinct from the values of

#",
y.
If,
for
example,
<p(x,
y)
be
"x
is
brother of
the class of
.r
for
which
<p(x,
y) is significant consists of
identically the
nificant.
all
9
same members as the
class of y s for
which
<p(x,
<p(x,
y) is sig
In such a case, the range of significance of
y) is the class of
the ordered couples which can be formed by combining any

itself
member
of
of
the class with

class be
i,
or with
any
other.
Thus
if
the
members
such a
G2
>
10 as, etc., the class of couples in question will be
(ai, ffi),
(fli,
(
2,
a 2 ),
a,),
(i,
a-s),
(a z
ai),
(2, a 3 ),
,
...
(a 3 ,
.
.
aO,
3,
a 2 ), (a 3 a 3 ), ...
Etc., etc.
is
But
y",
if
<p(x,
y) represent
"x
a citizen of
y",
?/",
or
"a:
is
a proposition about
or
"x
is
member
is
of the class
the class of x
and the
class of y s
for
which
9
<p(x,
y) is significant will
"A
be mutually exclusive.
is
Presuming that
Schroder treats
first
brother of
A"
significant
i.
e., false.
10
all relatives
as derived
Alg. Log., in,

relatives.
three chapters.)
But
this
from such a class of ordered couples. (See is an unnecessary restriction of the logic of
247
does not neces
Although
<p(x,
y) represents
some
relation of x
and
y, it
sarily represent
any
relation of the algebra, such as x
cy
or x
y;
and
it
cannot represent relations

<P(XI,
which are not assertable.

prepositional functions of one variable, y.
<p(x
y),
<p(x*,
y), etc., are
Hence H v v(xi,
and properties
y),
Uv
<p(x,
y),
^v
lt
y), etc., are propositions,
the meaning
of
And U y
<p(x,y),
2 v v(x,y),
which follow from preceding definitions and theorems. and 2 t Il x <p(x,y), y) are propositional
<t>(x,
functions of one variable.
We
can, then, define propositions involving

follows:
two variables and two operators, as

11-01
11-02
11
UxUy 2xUy
n*Z Sx S
<f>(x,
y)
==
ll t
{n y
<t>(x,
y)}.
Def
Def.
Def. Def.
==
<p(x,y)
2 x lU y <p(x,y)}.
03
tf <p(x,
y)
==
nx
sy
<p(x,
y)
} .
11-04
tf
?(.r, y)
--=
2,{S^(.r, y)}.
from these definitions that our explanation of the range of ^significance of functions of two variables was not strictly required; it The same con follows from the explanation for functions of one variable.
It will be seen
vention regarding the

the propositions
is
number
of values of variables
and interpretation
of
also
extended from the theory of functions of one
variable to the theory of functions of two.
(Where the
will
first
variable has a subscript, the

is
comma between
.r,
the two
be omitted:
Since
<p(x_y)
p(.r 2
y), etc.)
Uy
<f>(x,
y)
is
a propositional function of one variable,
the defini
tion. 10-02, gives us
IUIX*,
And
y)
==
UÎl^x,
y)}
--=
Tl y
<p(x,y)
xll y
<f>(x
z y)
the expansion of this last expression, again by 10-02,

{
is
<p(xiyi)
x
x
{
<x
x Ôi*/ 2 ) x x 2) x
<>x?
<p(.ri?/s)
x
x
^(.r 2 z/ 3 )
<p(x*yi)
<p(x
3 y) x
x ... Etc., etc.
And
similarly,
y)
by 10-01,
SJlXz,
And
= S x {n^(;r,
y)}
--=
U
is
the expansion of the last expression, by 10-02,
x
x
x +
. .
Etc., etc.
248
Or, in general,

is
any prepositional function with two operators
expanded
into a two-dimensional array of propositions as follows:

(1)
The operator The
nearest the function indicates the relation (+

line.
or
x)
between the constituents in each

(2)
subscript of the operator nearest the function indicates the

the lines.
(
letter
(3)
which varies within
The operator
The
from
to the left indicates the relation
+ or
between
each two lines.

(4)
subscript of the operator to the left indicates the letter which

line to line.
varies
Some
2xUv
<p(x,
caution must be exercised in interpreting such propositions as

y}, etc.
It is usually sufficient to
it
read ZJI y
x,
"For
some
such
x.
ami.
every
y",
but
<p(x,
strictly
should be
"For
"For
some
every y
is
is
that".
Thus 2 x H y
true".
y} should be
<p(x,
some
x,
every y every
such that
<p(x,
y) is
And H v 2 x
is
true".
y) should be
"For
y,
some x
is
such that
v(x, y)
which
ma
The two here chosen be made clear as follows:

2 x ll y
<p(x
illustrate the necessity of caution,
y)
=
"Either
That
is,
2x
IIy<p(x,
y)
means
<p(x,
for x\
or for # 2
...
or for
and every y, y) is true, or for some other particular x and every
and every y, .T 3 and every y,

y,
<p(x,
<p(x,
y) is true, y)
is
<p(x,
true,
y)
is
true".
On
the
other hand,
Tl y
2x
<p(x,
y)
= 2x
<p(x,
i/i)
xS x ^(.r,
*/ 2 )
xS x
yi,
<^(.T,
y
y) is true,
<p(x,
That
Uy^ x (p(x, some x and y z

is,
,
y)
means
"For
some x and
for
<p(x,
and
for
<p{x,
y) is true,
and
is
some x and y 3
y) is true, y)
and
.";
or
"Given
any y, there
one x (at
least) such that
<p(x,
is true".
The
following illustration of the difference of these two is given in Principia Mathematical n Let y) be the propositipnal function y is a proper
<p(x,
"If
fraction, then x of y,
is
a proper fraction greater than

?/), so
y".
Then
for all values
we have 2 x
<p(x,
that
"If
Ii y
2x
<p(x,
y)
is
satisfied.
In fact,
U^
is
<p(x,y)
expresses the proposition: y a proper fraction greater than

sition:
tion",
is
a proper fraction, then there
always
?/".
But 2 x li y
is
<p(x,
y) expresses the
propo
frac
"There is
a proper fraction which
greater than
any proper
which
is false.
if
In this example,
11
we should read S x n y
"For
some x and every
#";
See
i,
p. 161.
Systems Eased on Material Implication
249
these two
IlySz
"For
every y and some
x",
we should make equivalent

this caution
is
very different propositions.

infrequent, as
But cases where
required are
we
r
shall see.
Where both operators
are
II
or both 2, the two-dimensional array of
since every rela propositions can be turned into a one-dimensional array, tion throughout will be in the one case x in the other + and both of
,
,
these are associative
and commutative.
It follows
from our discussion of
the range of significance of a function of two variables that any such func the ordered tion, y), may be treated as a function of the single variable,
<p(x,
couple, (x, y).
Hence we can make the further conventions:
11-05
SsSy^Gc, y)
tt x ll y
<p(x,
= S (x
V )<p(x,
y)
y)
==
S Xf
<f>(x,
y).
y).
11-06
y)
= n (x
V )<p(x,
==
nx
<?(x,
The second
1 1
half of each of these serves merely to simplify notation.

<p(x,
is
x r and y a be any values of x and y, respectively, in == r y s ). such that a value of (x, y) y) n say, (.T, y) n
07
If
y)
there
<f>(x,
<p(x
11-05 and 11-06 could be derived from 11-07, but the process is tedious, and since our interest in such a derivation would be purely incidental,
we
prefer to set
If
down
all
three as assumptions.
y
<p(x,
we wish
2x 2y
to identify a given constituent of S*,

<p(x,
y) in
with a con
stituent of
is
y),
if
some convention
of the order of
terms
<p(x,
Sz
<p(x,
y)
required, because
the order of constituents in 2 x 2 v
y) be unaltered,
this identification will
determined
be impossible unless the number of values of y is we which, by our convention, need not be the case. Hence
of
make, concerning the order

tion:
if
<p(x
terms
in
S Xf
<
<p(x,
y),
the following conven

in
m yn) precedes
<p(x
y s)
if
m+
r
y
s,
and where
n =
+ *,
<
s.
Thus the order

+
<p(x
of
terms
in S*,
<p(x,
y) will be
<p(xiyj
2 yi)
This arrangement determines an order independent of the number of values r y e ) can always y) n in terms of of x, or of y, so that the equivalent of
<p(x,
<p(x
to govern the be specified. exactly similar convention is supposed with the terms y) and their identification arrangement of terms in n*, y
12
An
<p(x,
of
U x ll y
12
<p(x,
y).
These conventions
of order are
never required in the
proof of theorems:
we note them
here only to obviate
any
theoretical
familiar device for denumerating the rationals i. corner. agonals, beginning with the upper left-hand
the This arrangement turns the two-dimensional array into a one-dimensional by successive e., by proceeding along
250
A
The
y)

of
is it
objection.
identification
2x2y
will
<p(x,
y)
with
Sx
<p(x,
y),
and
of
HJIy^Gr,
with
nx
is
<p(x,
y),
of little
consequence for the theory of

of functions of
propositional functions
of relations
itself,
but
be of some importance in the theory
which
to be derived
from the theory
two or
more
to
variables.
Having now somewhat tediously cleared the ground, we may proceed the proof of theorems. Since 2 X y y) may be re y) and H x y
,
<p(x,
<p(x,
garded as involving only one variable,

at once from
(x, y),
many theorems
here follow
those of the preceding section.

y)
11-1
sx
<p(x,
= ^X^ V
<P(X,
y}
= -H z
y -<p(x,
y)
= -[U x U v
-<t>(x,
y)}.
[11-05-06, 10-05]
11-12
Xf y
<p(x,
y)
= U x ll y
<p(x,
y)
= -S Xf
v -<p(x,
y)
= -{S^-pfo
y)}.
[11-05-06, 10-04]
11-2
Hx
<p(x,
y)
<p(x,
y) n
[10-2]
11-21
<p(x,
y) n
c2 Xry
y)
<p(x,
y).
[10-21]
11-22
nx
nx
<p(x,
c s*
<p(x,
y)
[10-22]
11-23
(x, y) n
<f>(x,
y) is
equivalent to
".
"
Whatever value
of (x, y), in
<p(x,
y),
may
be,
<p(x,
y) n
[10-23]
11-24
U x Tl y
<p(x,
y)
is
equivalent to
be,
<p(x
"Whatever
values of x and
y,
in
v(x, y), x,
and
y,
may
r
r y^".
[10-23]
IL y
<p(x,y),
HJIÔ, y) is x may be, H y

value of
y, in
<p(x
equivalent to
<p(x
"Whatever
<f>(x
value of
.r,
in
yY\
y
s
And H y
y)
is
equivalent to
"Whatever
(f>(x,y),
may
be, v(x
r y^)".
But
y).
[11-01]
the values of x in
Hy
y) are the values of
x in
<p(x,
Hence
Q.E.D.
"
11-25
is
"Whatever
value of (x,y), in
values of
.r
<p(x,y),
(x,y) n
may
be,
<p(x,
y) n
equivalent to
"
"Whatever
and
y, in (p(x, y),
x r and y,
may
be,
<p(x
ys]
[11-06-23-24]
11-26
n x n,
<p(x,
y)
tf
<f>(x
n y)
[11-01, 10-2]
251
11-27
n x n^(.r,
S x S^Gr,
y)
is
= U v TL x
<p(x,
?/).
Since x
associative
and commutative, Q.E.D.

y).
11-28
y)
==
S S x ^(a;,
tf
Since +
11
is
associative
.
and commutative, Q.E.D.
29
IIJI,
<p(x,
y)
c n x ^(jr, y n )
[11-2(3-27]
11-291
lUI^C*,
[2-2,
y)
c<p(av#,).
11-24]
y)
11-3
IIJIÔ,
c SJI^Gc,
y).
[11-01, 10-21]
11-31
SJI^(a;,
[11-03]
y)
cU
2x
<p(x,
y).
ny s x
<p(;e,
y)
vfayi) + vfayd + + ^22/2) + x

I {
<p(x
yi)
+ + +
<K#i2/2)
^(^3^/2)
^(.TSZ/S)
<p(xiyz)
^(.r 2 z/ 3 )
x ... Etc., etc.

Since x
to the
is
distributive with reference to +

of the products of each
is,
this expression is equal
sum
column separately, plus the sum
of all the cross-products, that
to
z)
+ +
p(.ri2/i)
<f>(x
x x
<f>(xiy
<p(.ri#
3)
x x
z yi) x
<p(x
yz)
<f>(x
y3)
Etc., etc.
where
/I is
the
sum
is
of all cross-products.
But
[11-02] this
SJIÔ,
==
y)
+ A.
/).
Hence SJIÔr, i/) +^4 Hence [5-21] S x n y ^(.r,
n^S^Gr,
y)
c I^Z^Gr,
y).
We
is
have already called attention to the fact that the implication of 11-31 not reversible that S x n^(.r, y) and I^S^Cr, y) are not equivalent.
11-32
nx
2y<t>(x,
y)
c Sx S
tf
^(.r, ?/).
[11-03] n,S,
[11-04]
S x S^(a-,
S
c2
And
[5-992]
We
have
also the propositions concerning formal implication
where
252
functions
of

of
two variables are concerned.

<f>(x,y)
$(x,y)
by
f
The formal implication y) c \//(x y)] may be written either U Xt v

[<p(x t
or
UsUy
11-4
[<f>(x
y) Ci//(x, y)].
By
11-06, these two are equivalent.

first of
We
shall
give the theorems only in the

IT X
,
these forms.
y) y)
y [<p(x,
y)
c t(x,
y)]
= H = Ux
x>
v [-<p(x, v -[<p(x,
+ j(x,
y)]
y)].
x-^(z,
[10-6]
11-41
nx
y [<p(x,
y)
c t(x,
y)]
c[<p(x,
y) n
c f(x,
y) n ].
[10-61]
11-411
( (
nx
y [<p(x,
y)
c t( x
y)]
<p(x,
y) n }
c j(x, y).
[10-611]
11-42
nx
y [<p(x,
y)
c$(x,
y)]
c Sx
y [<p(x,
y)
cf(x,
y)].
[10-62]
11-43
n*,
[<f>(x,
y)
^(.r, y)]
c [n,
^>(a;,
?/)
Xf y
$(x, y)].
[10-63]
11-431
{n x
,[^(.r,
y)cf(x,
y)]
xH x
^(x,
?/)}
nx
y }(x,
y).
[10-631]
11-44
nr
y [<p(x,
y)
^(.T, ?/)]
c [2 Xt
y<p
(x, y)
c S,
y ^(.r,
?/)].
[10-64]
11-441
{n x
tf
[^(a;, y)
^(ic, y)]
x Sx
,^(.T, y)}
2,,
y $(x,
y).
[10-641]
11-45
{n x
y [<p(x,
y)
cj(x,
y)]
xH
If v [^(a;,
y)
cf(a
[10-65]
11-451
n Xftf [^(ic,2/)c^(a:,y)]
c {n x
[10-651]
,
[^fe
2/)
cr(.T, y)]
cn
x>
J^GT,
y)
c{(x, y)}}.
11-452
nx
[*(x,y)ct(x,y)] c {n^,
y [<^(^
y) c^(.r,
?/)]
cn
tf
[^(a;, y)
c rfe
y)]}.
[10-652]
11-46
n,,
y [<p(x,
y) c^(;r,
?/)]
= nx
,[-^(.1%
i/)
c-^(.r,
y)].
[10-66]
Similarly,
we have the theorems concerning the formal equivalence
of
functions of two variables.
253
11-47
n*,
v [v(x, ?/)
}(x, y)]
=--
(n x
v [<p(x,
y)cf(x,
y)]
y)c<p(x,
xn,, [*(*,
[10-67]
y)]}.
11-48
{n*.
y [*>(ar,
y)
}(x, y)} xll x
v [t(x,
y)
(x, y)}}
cU
[10-68]
y [<p(x,y)
(x, y)}.
11-481
II X
[*>(*,
?/)
iKz,
?/)]
{II X
,[^0, y)
f(z, y)]
f
c n,
[10-681]
[*>(*,
y)
f(x,y)]|.
11-482
n,
y [^(o;,
y)
- rfe
2/)]
c fn x
tf
[^(ar, y)
t(x, y)]
v [<p(x,y)
cU,,
[10-682]
= rfe
11-49
Hx
J^GT, y)
t(x, y)}
[^(a;, y) n
t(x, y) n ]
c[U c[2 x
[10-69]
x>l
,<p(x,y)
<p(x,
y)
= Ht = S Xf
y)
t(x,y)]
t
*(x
y)].
11-491
n,
tf
[^(.r,
i/)
- ^fe
?/)]
= nx
y [-^(ar,
-j(x,
y)].
[10-691]
Further propositions concerning functions of two variables are simple

consequences of the above.
The method by which such functions are treated readily extends to those of three or more variables. y, z) may be treated as a function
<p(x,
of three variables, or. as a function of one variable, the ordered triad
(a:,
y, z)
of x and y, or of the ordered just as \l/(x, y) can be treated as a function with each extension of Strictly, new definitions are required pair (x, y). number of variables, but the method of such extension our theory to a larger
will
be entirely obvious.
For three
y, 2)
variables,
f
we should have
t
njijiÔ,
SJIyll
,>(&,
y, z)
= u x {u v u = S x {lI n
1/
tf>(x
y, z)}
^(.T, y, z)\
Etc., etc.
It is interesting to note that the
most general form
for the analogues of
11-05 and 11-06
will
be
n
and
Since
*)*>(.r,
y, z) y, z)
==
S (I
y, z)
==
,)<t>(x,
njl^, S x S (y
,
g )<p(x t
y, z)
f)
^(ar, y, z)
z y,
njl (v
,)<p(x,
= n
(j/
,)<p(xiy,
z)
xn
(y ,
g}
<p(x
z)
xn
(l/
,)^(x 8 y, z)
254
x
A
.
.,
and
n (y
y
,
Z )<p(x n y,
2)
= Uy
llz<p(x
n y,
2), etc.,
we
shall be able to deduce
n
And
( x,
Z)<P(X,
y, 2)
= nji = n n
tf
(I ,,
2)
^(.r,
?/,
2)
(I>
Z)<P(X,
y, 2)
= n îwo, #, 2) = njijiÔ, y, 2)
(X|
<p(x,
similarly for
S (I
2)
This
calls
our attention to the fact that
as a function of three variables or as a function y, 2) can be treated not only of one, but also as a function of two, x and (y, 2) or (x, y) and 2 or (x, 2)
and
?/.
In general, the conventions of notation being extended to functions of any number of variables, in the obvious way, the analogues of preceding
theorems for functions of two
will follow.
We
failed to treat of
such expressions as
U<px
xUfy,
2<px
U\f/y, etc.,
under the head of functions of one variable.
The reason
for this omission
was that such expressions

of the type H. x l\. y of functions of
(<f>x
find their significant equivalents in propositions
x^z/),
2 x Hy
(<px
+ ty),
etc.,
and these are
special cases
of the
two
variables.
We may
<p
also remind the reader

<px
+ U\l/x and n tpx + U\{/y. between two such expressions as II need not be identical; there and The ranges of the two functions,
difference
\f/,
may
be values of
.r
in
<px
which are not values

,
of y in $y.
But
in
any
expression of the form
<px
x fa n x n as a value of x in #x must be identical
with x n as a value of
.r
in ^.r.
For
this reason,
is
we have adopted the con

Hence the case where
vention that where the same letter
used for the variable in two related
functions, these functions have the

\[/y is
same range.
the more general case, in which the functions are not we have px and Theorems involving functions of this type restricted to the same range.
not always be significant for every choice of v and \J/. There may even be cases in which an implication is not significant though its hypothesis is
will
significant.
But
true;
for
whatever functions such theorems are
significant,
they
they will never be false for any functions, however chosen. + \f/y) follows from the The meaning of an expression such as 2 x U y (
will
be
<px
definition of
2x
Tly<f>(x,
y).
(
i
S x IIj,(
<px
\j/y)
= II, =
+ +
<px
i/^)
z
Il y (
<px
.r
<xi
+ $y) + U v ( + \/ z x
<px
+
\/
^)
3
+
.
.
<xi
{(VP.TS
. .
.
Î) x( ^3 +
^2/2)
x( ^3 +
^2/3)
Etc., etc. of
And
for
any such expression with two operators we have the same type

two-dimensional array as for a function of two variables
only difference x \{/y, etc. or
<px
255
in general.
<^.r
The
+
\fry
is
that here the function
itself
has a special form,
12
(1)
[1-3]
II
<>.r
n^.r
(2)
II\l/y
= =
<pX]_
</?.r
x
(<p.r
(<f>X!
x nî/) x
(>.r 2
x n^z/) x
x Ityy) x
[5-98]
xn
(3)
tf
(^r 8 x^y)x...
[10-371]
[ii-oi]
By
(2)
and
1-3,
tpxi)
x (n^y x
<^.r
2)
(H^
3)
(p.r 3 )
x ...
n y (^
xll
.
<^.r
2)
tf
(^x^.r
^.r)
x ...
[10- 37]
[11-01]
(4)
"
Similarly, ll^y
.r
xH^r = U y U x (\f^y
is
=
is
11^(^.1- x^y).
"For
<px
is
a;
true for every
and
\f/y
true for every

true",
z/"
equivalent to
every
12-2
and every
y,
<px
and
^z/
are both
etc.
S^.r + S^z/
= 2^z/+2^.r = ^^^(^.r
Sî- +
(1) (2)
[4-3]
2<p.i-
S^
- S^+S^r.
i
S^
= (^
^2 + ^3 +...)
->/
[5-981]
[10-31]
SxZ^ +
ifc/).
[11-04]
(3)
By
(2)
and 4 -3,
[10-3]
S x S,(^+^r).
tf
[11-04]
(4)
Similarly,
S^+2^.r = S
256
for
A
some
x,
<px,

or for
\f/y",
"Either
some
etc.
y,
tyy"
is
equivalent to
"For
some
x and some
12-3
y, either
<px
or
S^r
=
[1-3]
S5
(1) (2)
S^c S#e S^y =

x X S^y) +
O?2
(
X S^y) + (^T 3 X
[5-94]
+ S
tf
x $y) + S y ( +
x ^y)
[10-361]
...
S^C^cx^).
S
x
i
[11-04]
(3)
By
(2)
and
x
1-3,
Si =
S^
y
x ^cj +
<^x
+ Sy
=
(^x^ )+
8
tf
..-
[10-36]
SA(^x^).
#,
^?/"
[H-04]
#r)
(4)
"For
Similarly,
<^.r,
S^y xSô; = S y S,(^ x

some
is
- S S x (^c x^).
"For
some
y,
<px
x,
and
for
equivalent to
some x and
some
12-4
and
+
^y",
etc.
v?.r
n^
H<px
n^ + n^
IL<f>x
= nji^z
=
^) = n*n y (^ +
(<px
<?x)
UyUx
3
== Il v
U x (frj+
<px).
(1) (2)
[4-3]
+ TIty
= U^y +
x
<^x
U^px.
+ n^?/
=
=
(âji
(<pxi
<px
+ ttty) x
(^2
+ n^z/ ) + nî/) x (^T 3 + n^y) x

. .
...
[5-941]
+
=
M
x
x...
[10-371]
n s n (^ + ^).
tf
[ii-oi]
(3)
By
(2)
and 4 -3,
+ n^?/
= (n^y +
ô x (n^ +
^2)
n
.T 3 )
x...
[10-37]
(4)
"Either
Similarly,
x,
<px,
Uty
H<f>x
= U v U x (^y +
\f/y"
<px)
= U y tt x
(<px
for
every
or for every y,
is
equivalent to
"For
every x
257
and every
invalid.
y, either
<px
or
\f/y",
etc.
At
<px
first
glance this theorem

If
a-
may seem
it is
One may
If
say:
"Suppose
it is
be
.
is
a number,
odd
and $y be
is
a number,
even
Then
even
,
U<px
U^y
will
be
\f/y)
Either
will
every number is odd or every number Every number is either odd or even
illustration lies in misreading TL x
is
".
but
U x ll v
(<px
be
The mistake
\^y).
of this supposed
Uv
(<px
It
is
is
legitimate to choose,
as in this case,
<px
and
\f/y
such that their range

+ $y) as
if
identical:
but
it is
not
legitimate to read
UxUy Ux
(<px
with
a corresponding value of y.
lly(<px
To put
each given value of x were connected nzllx (#c + #c), it another way:
as a special case of
either
of x
<px
or
\l/x",
but would be
<px
+ ty), would not be "For every value of x, "For any two values of x, or for any value
$x"
.
and
itself,
either
or
Thus HJIj/^T +
is
1/^)
in the
supposed
illustration would not be as above, but
in fact
is
or for
any number and

+ TLfy and
itself,
either one
any odd or the other

"For
pair of
numbers,
so
is even"
that
lent.
U<px
Il x lly(<px
+ $y) would here both be
false,
and are equiva
A
12-5
somewhat
similar caution applies to the interpretation of the next
two theorems.
2<f>x
The analogues
of these, in
<p(x,
y),
do not hold.
^x
Wy
By
= n^y + 2
<px
= 2 X II V
<f>x
+ ty)
==
(1)
(2)
[4-3]
proof similar to
(2) in 12-2,
2<px
And by (3) By proof similar to And by proof similar to

"Either
proof similar to (3) in 12 2,

(2) in 12-4,
<px
Ufy = S x n v (^r + + Ufy = S,n (^ +

tf
ty)
<px)
(3) in 12-4, Uifry

is
H^y + S + S^r
==
Uy ^x
"For
for
some
or,
^r, or for every y,

<?x
^"
equivalent to
some x
and every
12-6
y, either
or
^?/",
etc.
s<^.r
S^r By proof similar to And by proof similar to

(1)
[1-3]
(2)
(2) in 12-3,
(3) in 12-3,
S^c xltyy
2<px
==
==
S*n (^c x
tf
xH^z/
S,H y (^ x
By proof similar to And by proof similar to

(3)
"For
(2) in 12-1,
(3) in 12-1,
is
Ityy
Ityy
xS^c xS^r
==
n H
tf
2 2
some
y,
<px
x,
<px,
and
for every y,
etc.
fy"
equivalent to
"For
some x and
every
18
and
^?/",
258
A
We may
generalize theorems 12-1-12-6
by saying that
for functions
of the type
(<px
\f/y)
and
(px
\j/y)
the order of operators and of
members
in the function is indifferent;
and
for propositions of the type
n
l 2J
the operators
may
be combined, and the functions combined in the relation
between the propositions. It will be unnecessary to give here the numerous theorems which follow from 10-5-12-6 by the principles pqcp, pcp + q, and Il^xc Spx, etc.
For example, 10-51,

IlipX
gives at once
(i)
nx
(<px
(2) (3)
(4) (5) (6)
Etc., etc.
And
12-2,
2<px
2\j/y
= 2x 2v
(1) (2) (3) (4) (5) (6)
+
(<f>x
\fry),
etc.,
gives
S,
2<px
2^z
from
Etc., etc.
Another large group

the combination of
of theorems, only a little less obvious, follow

cH.<px,
H<px
or
S#ccS#r,
with
Ii\f/y
c S^y, giving
by 5 -3,
(3) (4)
<px
+
x
Tl\{/y
TL\f/y
<px
c 2 (px + ^^y x 2^?/ c2

<px
Etc., etc.
Each
of these has a
whole
set of derivatives in
which
n (px + U\I/y is replaced
259
etc.
by
H. x lly(<px
1/
?/),
etc.,
H<px
xZi/ Z/ by
(2),
n x S (^x^),
J/
We
give, in
summary form,
the derivatives of
by way
any one
of illustration:
Any one
of
of
xU\l/y
x
x
<px)
<px)
S n x (^ x
tf
U y Ux
9x x ^)
S vnz (
<px
x
<?x
n ipx
zty
xn
etc., etc.
etc., etc.
^),
This table summarizes one hundred
fifty-six
etc., etc.
theorems, and these are only a
portion of those to be got by such procedures. and ((px + Functions of the type of
(<f>xx$y)
g ive four different

(2)
kinds
(3)
of
implication
,
relation:
(4)
(1)
tf
njl^rc^);
$y).
"implications",
n x S v (^c^);
of the
of
all
2 x n y (<pxcty)
and
S x S (^c c
field
With the exception

though
first,
these relations are unfamiliar as

of
them
could be illustrated from the

ticularly useful:
mathematics.
Xor
are they par
the results to be obtained by their use can always be got

material
implications
or
by means
of
formal
implications.
Perhaps
UJl y
12-7
(<px
c $y)
tf
is
of sufficient interest for us to give its elementary properties.
n x n (#cc
[11-01]
y^
i
c Ufy) x
(^
c
...
[10-42]
And
this last expression is equivalent to the set

<pxi
Uifry,
<pxz
c H^y,
<px
c nty,
etc.
12-71
{IIJI
J/
Ozci/
?/)
x^r B
[9-4, 12-7]
260
If for
A
every x and every
is

implies
y,
<px
\l/y,
and
for
some given
x,
<px
is
true
then $y
12-72
true for every y.

<px
Il t Tl v (
c }y) = S ^.r c Ilty =
II
- #c
<p
(1)
If
IIJIÔo; c^y), then [12-7]
Hence
[5-991]
And
if
S^cltyi/. S^ccltyy, then [10-42]
tt y (2
<px
c ty)
and hence [10-41]
(2)
[9-3]
is
S^c
"If
HxHy
(<px
c $y)
is
equivalent to
y".
there
is
some x for which px
is
true,
then $y
12-73
true for every
{nji w (#rc^)
[12-72]
If
xn y n,(^cf2)} cn x n,(^c^).
IlJIyC^rc^) and n v n(î/cfz), then

Hence
[5-1]
and S^y c Hfz. But [10-21] Il^yc^y.
S*ccII3, and
[12 72]
:
This implication relation is here demonstrated to be transitive. In fact, it is, so to speak, more than transitive, as the next theorem shows.
12-74
{(S^ccS^) xn n,(^cfz)} cIIJI.C^c cfz). = s^z/cn^. [12-72] n v n (^crz) And [5-1] if S^c cZfy and S^y c Hfz, then S^.T
tf
c Hfz, and [12-72]
12-75
(n x
[12-72]
ns
if 2<pxcttty
And
[5-1]
and H^y c Hfz, then S^c cHfz, and [12-72]
IV.
DERIVATION OF THE LOGIC OF CLASSES FROM THE CALCULUS OF

PROPOSITIONAL FUNCTIONS
The
logic of classes
and the
logic of relations
can both be derived from
the logic of propositional functions.
In the present chapter, we have
begun with a calculus of propositions, the Two-Valued Algebra, which includes all the theorems of the Boole-Schroder Algebra, giving these
theorems the propositional interpretation.
sidered as belonging to the calculus of
have proved that, con propositions, these theorems can

"If
.
We
validly be given the completely symbolic form:
then
."

being replaced by
"...
261
...","...
is
equivalent to
..."
by
".
.",
etc.
The
it
Two- Valued Algebra does not presuppose the Boole-
Schroder Algebra;
simply includes it. the calculus of propositions the Two-Valued Suppose, then, Algebra our fundamental branch of symbolic logic. We derive from it
we make
the calculus of prepositional functions by the methods of the last

sections.
We may
then further derive the calculus of
logical classes,
two and a
calculus of relations,
by methods which are to be outlined
in this section-
and the next.
The present section will not develop the logic of classes, but will present the method of this development, and prove the possibility and adequacy At the same time, certain differences will be pointed out between of it. the calculus of classes as derived from that of propositional functions and
the Boole-Schroder Algebra considered as a logic of classes.
distinguish class-symbols from the variables, x,
tions,
y, z, in
In order to
propositional func
we
shall here represent classes
by
a, 0, y, etc.
For the derivation
of the logic of classes

is
from that of propositional
conceived as the aggregate of individuals for functions, a given class is a man", n represent which some propositional function is true. If
"a:
<px
then the aggregate of x
s for
which
<px
is
true will constitute the class of
men.
If,
then,
propositional
represent the aggregate of individuals for which the will be "the class determined by is true, function
z(<?z)
<pz
z(<pz)
the function
acter
etc.
13
<>".
<pz",
or
"the
class
determined by the possession of the char
We
can use
a, 0, 7, as
"a
an abbreviation
for
z(<pz),
2(^2), z(&),
<pz",
z(tpz) will
(In this connection,
is the class determined by mean and we should remember that

<px
the function
<pz
are the
same
function.)
The
relation of
an individual member
"
of a class to the class itself will

is
be symbolized by e. x n e a represents x n This relation can be defined. n is an

"x
member
of
a "-or
briefly
a".
here borrow the notation of Principia. The corresponding notation of Peirce and Schroder involves the use of S, which is most confusing, because this S has a meaning But in of a propositional function. entirely different from the S which is an operator such does not represent an aggregate of individuals; it represents Principia, And z(tpz) is not a primitive idea but a notation supported by an elaborate that Our procedure above is inelegant and theoretically objectionable: we adopt it theory. because our purpose here is expository only, and the working out of an elaborate technique would impede the exposition and very likely confuse the reader. As a fact, a more satis
13
We
"
z(<pz)
<pz".
factory theory on this point
makes no important
difference.
262
13-01
"x
A
xn
a
z(<pz)

Def.
=
<px
n is
member
of the class
determined by
<pz"
is
equivalent to
"
<px
n is
true".
(For convenience of reference,
we continue
is
to give each definition
and
theorem a number.)
The
class
"a
relation
"a
is
contained in
of
ft"
the relation of the class a to the
ft
is
when every contained in

and
member
ft"
a
ft.
is
by a c
"is
a member of (3 also. We shall symbolize The sign c between a and ft, or between
in";
z(vz)
2(^2), will be
",
contained
e
c between propositions
e
will
be
"implies
as before.
xn
is,
of course, a proposition; x
a, a prepo
sitional function.
13-02
acft =
is
U x (xeacxeft)
"For
Def.
c
a c
ft
equivalent to
e o:
every
x,
is
an a implies x
is
ft
".
nx
(>
cx
e (3)
is
a formal implication.
is
that the logic of classes
It will appear, as we proceed the logic of the formal implications and formal
equivalences which obtain between the propositional functions which deter
mine the
13-03
(
classes.
ft)
= U x (x e a =
"For
e (3)
{
Def.
a
to
=
x
ft
is
equivalent to
every
is
member
a c
of
/3
".
a =
e.,
ft
ft
x is a member of a is equivalent x, thus represents the fact that a and ft
have the same extension

xn
e
i.
consist of identical
members.
propositions.
a,
ft,
and a
are assertable relations
But
the logical product of two classes, and the logical sum, are not assertable
relations.
They
are, consequently, defined
not by means of propositions
but by means of functions.

13-04
axft
= x[(xea) x(xe
of
ft)}
ft,
Def.
is
The product
which
this
two
classes,
"x
a and
the class of x
is
propositional function
is
is
an a and x
ft".
determined by the The class of the x s for

s
true constitute
a*
ft,
the class of those things which are
both as and
/3 s.
The relation x between a and ft is, of course, a different relation from x between propositions or between propositional functions. A similar remark applies to the use of + which will represent the logical sum of two classes,
,
two propositions or propositional functions. This double use of symbols will cause no confusion if it be remembered that a and ft, and 2(^2), etc., are classes, while x e a is a propositional function,
as well as of
(<pz)
and x n
a,
a c
ft,
and a =
ft
are propositions.

13-05
263
a+
of
/3
= x{(xea)
classes,
+ (x6(3)}
Def.
is
The sum
two
x
is
a and
/?,
the class of z
,
such that at least one
of the two,
an
is
and x
an a or x
is is
a
/3.
is
true, or loosely, the class of z s
such that either Z
The negative
13-06
of a class
can be similarly defined:

Def.
is false.
-a = x-(xta)
The negative of a is the class of or s for which x is an a The "universe of discourse", 1, may be defined by the
device of selecting
some prepositional function which is true for all values of the variable. Such a function is (f.r cfz), whatever prepositional function x may be.
13-07
1 is
= x({xcx)
xs
x
s.
Def.
the class of
all
for
which x implies
"null-class",
x.
Since this
is
always true,
1.
1 is
the class of
The
Def.
0, will
be the negative of
13-08
=
is,
-1
That
by
13-06,,
= x-(txcx), and
s will
/3
since
-(fac^)
is
false for all
values of x, the class of such x
be a class with no members.

2(^2).
This will be tt x #r). c$x), and a = assertable relations of establishes at once the connection between the
Suppose that a Hence a c ft will be
2(^2)
and
Then, by 13-01, x n
==
(<px
==
<px
n.
Ux
(<f>x
classes
and formal implication and equivalence.
To
illustrate the
way
in
which
this connection enables us to derive the logic of classes
from that
of
propositional functions,
It will
prove a number of typical theorems. be convenient to assume for the whole set of theorems:
we
shall
a
13-1
= 2O),
]8
2(*z),
&(&)
= x-({xc{x). = -1. Hence [13-06] = x -(x * 1). = f.r c c. Hence = x -(s*.r [13 01 07] x e 1
f .r)
13-2
n x (xel).
[13-01-06] x n el
Hence
U x (x
-
1)
== (x n cxn). = U x (xc{x).
But
[2
2]
fr n c
r-r n .
Hence
a
[10 23] of the
n,(^ c
x)
Every individual thing
is
member
.
"universe
of
discourse".
the range of discourse", 14 This defines, not the universe of discourse, but "universe With 1 so defined, propositions which invol of significance of the chosen function, f whenever t, etc., a the classes (**), (*), etc., and 1, will be significant
<p,
the same range, and true
if
significant.
264
13-3
II x
A
-(.eO).
[13 -01 -06 -07]
a- n
eO =
-(f.r n cf.r n ).
Hence
[3-2] ~(x
f.r.
n e
0)
(r n
c far)
But $x n c
For every
class.
Hence
that x
[10-23]
e
n x (fx
fr),
and
is
H x -(a;
0).
x, it is false
no individual
member
of the null-
13-4
cl.
[13-01-06] X n tl
Since a
frr n Cf.r n ).
e
%(<pz),
[13-01] x n
ipx n -
c [*c n c (fz n c &)]. [9-33] r n ). Hence since f.r n c f.r n n c ($x n c c ({x c{x)], and [13-2] a cl. Hence [10-23] tt x
cf.r n )
,
<p.T
On
[<px
Any
class,
a, is
contained in the universe of discourse.
It will
be noted
(13-2 and 13-4) that individuals are members
of 1, classes are contained in 1.

"A
In the proof of 13-4, we
make
c $x n
use of 9-33,
is
true proposition
it is
is
implied
n.
by any
proposition".
$x n
true.
x,
Hence
implied by
<px
And
since this holds,
whatever value of
xn
may
be, therefore,
But
(px is
the function which determines the class

1.
fa:
cfa:, the function
which determines
Hence
<px
n is
xn
a,
and
xn
cx
n is
xn
e 1.
Conse
quently we have 11^ (a:caca:cl). contained in", this is a c 1.

"is
And by
the definition of the relation
13-5
Oca.
[9-1]
-O n e 0)
is
e
equivalent to (x n
0)
0.
Hence
[13-3] (x n
0)
0,
and
a,
[9-32]
(x n e 0)
<px
n.
Hence
Hence
[13-01] x n
[13-02]
is
eOcx n e
a.
and [10-23]
U x (xeQcxe
a).
The
"A
null-class
contained in every
class, a.
In this proof,
-(fa: n
we
use 9-32,
false proposition implies

.
<px
hence implies
and
<px,
and any proposition". But -(fxcfx) is the function which determines 0; n the function which determines a. Hence c a.
cfa: n )
is false,
The
proofs of the five theorems just given are fairly typical of those
which involve
and
1.
But the great body

This connection
of propositions
of classes
make more
and formal
direct use of the connection
between the relations
implications or equivalences.
following:
may
be illustrated by the

13-6
z(
265
= n x ( (px c \j/x) [13-02] z((pz) cz(^z) = U x [xe z((pz) c.ce = (px n and x n e z(\f/z) = $ [13-01] x n e z((pz} ex e (^z)] = U x Hence [2-1] U t [x e
(pz)
z(^z)
z(<pz)
(<px
c
\f/z"
"The
class
determined by
"For
(pz is
contained in the class determined by

\frx".
is
equivalent to
every
x, (px implies
[13-03]
[13-01] x n
e
=
[z(<pz)
z(^z)]
<px
= Hx
[xez(<pz)
a-ez(^z)].
.
z((pz)
e
==
n,
e
and x n
z(^z)]
z(^z)
(<px
Hence
"The
H x [x
=
z(<pz)
= Hx
= \l/x n = fa)^z"
class
determined by
"For
<pz
is
equivalent to the class determined by

is
is
equivalent to
every
x,
<px
equivalent to
$x".
[10-66]
Il x [x e
z(<pz)
Hence
Hence
13-9
II x (a: e
ex
e
e /3)
ex e z(tz)] = n x [-(.i- e
==
E x (-[.r
c-(x
e
e
z(^z)] c-[.r
z(^z)]).
/3)
a)].
.T e
[13-01-06]
-(a;
a)
a; e
-a, and -(x
0)
-/3.
[13-02]
(c/3) - (-/3c-).
[(aC|8) x(/3c T )]
[13-6] (a c
j8)
c(aC7).
= Ux
c
(<px
c ^ r),
t
(|8
7)
= H x (tx c
^.i-),
and (a c
7)
And
The
[10-65] [U x
"is
(<f>x
\f/x)
x
is
n x (^.r
.r)]
n x (â;
is
c ^).
the
first
relation
contained
in"
transitive.
is:
13-9
form
of the
syllogism in Barbara.
The second form

c j8).
13-91
[(aC]8) x(.Tnea)] c(ar B

[13-6] (aC]8)
= n x (^c^.r).
<px
[13-01] (x n
a)
==
n,
and
(x n
0)
^
.
n.
And
If
is
[10-611] [n x (^.r c ^r) x

is
^.r n ] c\j/x n
the class a
a
contained in the class

/3.
&
and
.r n
is
member
of a, then
.r n
member
[(a
of
13-92
]8)
x(/3
\<2
==
7)]
c (a
=:
7)~
[lo-/J
Py
-iix^cp.c
/>
\fj
And
The
[10-68] [H x (^,r
^.r)
xH x (^.r =
^.r)]
cH x (^ =
^r)-
last three
theorems
illustrate particularly well the direct connection
266
between formal implications and the relations of classes. 13-6 and 13-7 = 0. Similar alternative defini are alternative definitions of a c /3 and a
tions of the other relations
would be
l5
:
-[z(<pz*)]
z(-<pz)
z(<pz)
We may
give one theorem especially to exemplify the
way
by
in
which
every proposition of the
Two-Valued Algebra,
since
it
gives,
10-23, a
formal implication or equivalence, gives a corresponding proposition con cerning classes. We choose for this example the Law of Absorption.
13-92
[a+(aX]8)]
a.
[13-04-05] [a+(aX|8)]
= x{(xea)
x/3)]}
+ [(xea) x(.re/3)]}.
Hence
[13-01]
{x n e
[a + (a
=
But
But
[13-03]
{(x n
ea) +
[(x n
ea)x(x n
0)]}
==
e(3)}}.
(1)
{[+(
=
x/3)]
=
e
a}
a) + [(x
e
n,[{(.T
a) X (x
,
(x
)J.
(2)
[13-01]
(x n e
a)
=
<px
n,
(x n e 0)
= $x n and by
+
(<px
(2),
{[a+(a*{3)]
=
+
a}
But
[5-4]
[<pX
+(<pX n X\l/X n )]
= Ux =
xifrx)]
{[<px
xjx)]
=
<px}
V%n>
Hence
[10-23]
TL x
=
<px}.
{[<px
(<?x
All but the last
two
lines of this proof are

/3)]
the connection between [a + (a x
= a and
concerned with establishing the formal equivalence

<px]
n a {[#c+(#ex#r)] =
Once
this connection
is
made, we take that theorem

[a + (a x
if/x n
of the
Two- Valued
Algebra which corresponds to substitute in it n for p and

<px
/3)]
namely
5-4, (p + .p q)
p,
for
g,
and then
generalize,
by 10-23,
to
the formal equivalence which gives the proof.
An
exactly similar pro
cedure will give, for most theorems of the Two-Valued Algebra, a corre sponding theorem of the calculus of classes. The exceptions are such
propositions as
p =
its
(p
or an equivalence.
which unite an element p with an implication In other words, every theorem concerning classes can
1),
be derived from
analogue in the Two- Valued Algebra.
We may
15
conclude our discussion of the derivation of the logic of classes
As a fact, these definitions would be much more convenient for us, but we have chosen to give them in a form exactly analogous to the corresponding definitions of Principia (see
i,
p. 217).
267
set of postulates
will
from the
logic of prepositional functions
by deriving the
II.
for the Boole-Schroder Algebra given in
prove that, beginning with the Two- Valued Algebra, as a calculus of propositions, the This procedure may have the appear calculus of classes may be derived. ance of circularity, since in Section I of this chapter we presumed the
propositions of the Boole-Schroder Algebra without repeating them.
Chapter
This
But
the circularity
is
apparent only, since the Two- Valued Algebra
is
a distinct
system.
The
postulates of Chapter II, in a form consonant with our present
notation, can be proved so far as these postulates express symbolic laws.
The postulates of the existence, in the system, of -a when a exists, of a x when a and exist, and of the class 0, must be supposed satisfied by the fact that w e have exhibited, in their definitions, the logical functions which
r
determine a x
14-2
(a Xa)
(3,
-a, and
a.
e
O.
16
[13-01] x n
=
<fX
n.
Hence
Hence
[13-04] x n
[13-03]
e(a*a) =
a]
[(x n
a) x (x n
a)]
[(aXa) =
x
= H x {[.re
(a x a)]
= = x = Ht
(<px
^.r n ).
a]
[(<px
==
<px)
<&].
But
[1-2]
(<f>x
=
<pxn)
<px
n.
Hence
14-3
[10-23]
(/3
Ut
a).
[(<px
==
<px)
<px].
(ax0) =
[13-03]
[(ax0)
= U x {[x e (a x 0)] = = II x {[(.r6a)x(.r60)] = [(.r e /3) x(.r =

(0 x a)]
[x e (ft
x a)]}
a)]}.
[13-01-04]
Hence
= (0 x a)] = nj(#rx#r) = [13-01] [(ax/3) = (#r x ^r n ). But [1-3] n xtx n = (^ x ^.r)]. Hence [10-23] U x x^.r)
(<px
(*x x
<px)}.
[(<f>x
14-4
(aX0) X T
[13-03]
==
ax(0x 7 ). [(X0)X 7 =
==
X(0X T )]
x
(.r
= n x [{.T6[(aX/3)x 7 ]l = n x [{[(.T e a) x (x e (3)]

Hence
[13-01]
{a?e[ax(0x 7 )]}J
e
7)
==
J {
(.r e
a) x
[(.r
0)
(.r e
7 )] }J.
[13-01-04]
[(x0)
x 7
- ax(0x 7 )]
is possible when the satisfactory derivation of these existence postulates functions is treated in greater detail. See Principia, I, pp. 217-18. theory of propositional
16
more
268
A
But
[1-04]
(<px
xtx n )
{[(<px
xx
=
<px
n x(\f/x n
Hence
14-5
[10-23]
0.
Hx
x^x) x &]
=
[<px
x^x n ). x(+x x &)]}.
axO =
[13-1-01] .TncO
= (x n eO)} = -(fzcMOj. n.u^x-o-zc^)] But [2-2, 9-01] (r.rcfar B ) = 1, and [3-2] -({x n c{x n ) = 0. = = -(fr cc). Hence [1-5] n x-(x n c x n )] Hence [10-23] n s {[^r x-frr c^)] = -(fa; cfr)}.
[13-03-04]
eQ)]
=
[<px
= -(frcr B ). [axO = 0] = U x l[(x n ea)x(x n
[13-01]
0, in
the fourth and fifth lines of the above proof,

of the calculus of classes.
is
the
of the
Two-Valued
Algebra, not the

these proofs will
Since the general
method
of
now
be clear, the remaining demonstrations can be some
what abbreviated.
14-61
[(x-|8)
0]
c[( a x/3)
a].
is
[13 -01 -02 -04 -06]
The theorem
equivalent to
But
[13-3]
Il x
-(xeO), and hence [9-1]

is
IL x [(x e 0)
0].
Hence the theorem

IT*
{
equivalent to
(xeQ)]c[(<pxxifrx)
[Or x -i/a*) =
is
But
[13-3]
H x -(xeQ), and
Ux
{[(<px
hence [9-1]
TL x [(x e 0)
= =
<px]}
0].
Hence the theorem
equivalent to
x-fr)
=
0]
0]
But
[1-61]
[(<f>X
n X-\j/Xn)
C[(<px
c[(^ x^) = = n X$X n )

=
0).
<px]}
<pX
n }.
Hence
14-62
[10-23]
Q.E.D.
{[(Xj8)
a]
x[(x-/3) =
is
a]}
c (a
The theorem
But
equivalent to
==
n,[{[(#rxr)
[13-3, 9-1]
^]x[(^x-^) =
0)
<px]}
c[<px
(x
0)]]
n x [(.r
is ==
0].
Hence the theorem
equivalent
to,
n,[{[(#cx^.T)
î ]x[(^x-^.r)
==
<px
^r]}
==
c(<px
0)]
But
[1-62]
{[(?.rxr n )
Q.E.D.
n]
x[(^x n x-^x n }
<px
n }}
c(^.r n
0).
Hence
[10-23]
The
definition, 1
-0, follows readily
from the definition given
of
in
this section.
The other two
definitions of
Chapter II are derived as follows:
14-8
O+/3) - -(- ax -j8).
The theorem
is
equivalent to
Il x [(<px
\f/x)
= -(-$x
x-\f/x)].
269
But
[1-8]
(<px
+tXn) =
Q.E.D.
-(-<?x
x-\l/x n ).
Hence
14-9
(aCjS)
[10-23]
[(ax/3)
is
a].
V/.r)
The theorem
But
[1-9]
(<f>x
equivalent to IIjOz c
is
= Uz
tt )
[(<px
x$x)
==
<px].
c^x n )
equivalent to
[(^cxîc
==
v*c n ].
Hence [10-23-69] Q.E.D.

Since the postulates and definitions of the calculus of classes can be deduced from the theorems of the calculus of prepositional functions, it
follows that the whole system of the logic of classes can be so deduced. differences between the calculus of classes so derived and The
important
the Boole-Schroder Algebra, as a logic of classes, are two: (1) The BooleSchroder Algebra lacks the e-relation, and is thus defective in application,
since
it it
which
cannot distinguish the relation of an individual to the class of of which is con is a member from the relation of two classes one
(2)
tained in the other;
The theorems
of the
Boole-Schroder Algebra
of the cannot validly be given the completely symbolic form, while those functions can calculus of classes derived from the calculus of prepositional
17 be given this form.
V.
THE
is
LOGIC OF RELATIONS
The
tions of
logic of relations
two or more
variables, just
derived from the theory of propositional func as the logic of classes may be based
of one variable. upon the theory of propositional functions in extension when we logically exhibit the A relation, R, is determined
If to y. y) class of all the couples (x, y) such that x has the relation then x $[<p(x, y)} is the relation "parent of of is
<p(x,
"
represent
"x
parent
y",
is This defines the relation in extension: just as the extension of so the the class of all those things which have the property of being red,
"red"
extension of the relation
"parent
of"
is
the class of
all
couples in the universe.
a property that is the extension of the relation is, couples (or triads, etc.) of a certain class; The calculus of relations, like the calculus thus, the class of couples itself.
relation
is
the parent-child common to all the
of propositions,
"Oftentimes,
and
of classes,
is
a calculus of extensions.
i,
as in Schroder, Alg. Log.,
the propositional of classes have been represented in the symbols of the Book-Schroder Algebra. calculus has been treated otherwise than as an interpretation of to the system, tl in such a case, if these symbols are regarded as belonging But
is
the relations of propositions in the algebra calculus befo
invalid.
270
A We

the relation
assume, then, the idea of relation
R meaning
etc.
the class
of couples (x, y) such that x has the relation
R to
y.
R =
This notation
is
x y(x Ry),
S = w z(w S
z),
simpler
S = w
z[\[/(iv, z)],
but
it
and more suggestive than means exactly the same thing.
R = x A triadic
y[<p(x,
y)],
relation,
T, will be
such that
T = x
or
y z[T(x y
}
z)}
is
the class of triads

is
(x, y, z)
for
which the prepositional function
T(x, y, z)
true.
But
all
relations can be defined as dyadic relations.
triadic relation can be interpreted as a relation of a
dyad to an individual
y, z),
that
is
to say,
any function of three variables, T(x,
can be treated as a
.
two variables, the couple (x, y) and z, or x and the couple (y, z) This follows from the considerations presented in concluding discussion of the theorems numbered 11-, in section III. 18 Similarly, a tetradic relation
function of
can be treated as a dyadic relation of dyads, and so on.

of dyadic relations
is
Hence the theory
a perfectly general theory.
Definitions exactly analogous to those for classes can be given.
15 -01
(x, y) n e z
w[R(z,
ic)]
R(x, y) n
Def.
It is exactly at this point that
our theoretical considerations of the equiva
lence of
<p(x,
y) n
and
<p(x
y a ) becomes important.
For
this allows us to
treat R(x, y), or (x
Ry), and by 11-07, we can give our

15-01
"The
as a function of one or of
two
variables, at will;
definition the alternative form:
(xm
y n )ezw(zRw)
xm
R yn R yn
Def.
or extension, of the relation deter
true.
couple (x m y n ) belongs to the

(z
field,
mined by
15-02
R w)
is
"
means that x m
is
RcS = U
x, v
[(xRy) c(xSy)].
Def.
This definition
strictly parallel to 13-02,
(ac/3)
because, by 15-01, (x
= Hx
y)
e
(xe
acxe
(x
/3)
R y)
,
is (x,
and
y) is (x, y)
S.
similar
remark applies to the remaining

15-03
definitions.
y)].
(R
S)
= nx
y [(x
Ry) = (xS
Def.
and
and S are equivalent in extension when, (x S y) are equivalent assertions.

18
for every
x and every
y, (x
R y)
See above, pp. 253
ff.
271
15-04
R xS =
x y
[(x
Ry)* (x
y)].
Def.
product of two relations, R and S, is the class of couples (x, y. such that x has the relation R to y and x has the relation S to y. If R is
The
logical
"friend
of",
and S
is "colleague
of",
R x S will be
Def.
S,
is
"friend
and colleague
of")
15-05
R + S = xy[(xRy)
sum
of
y
+ (xSy)].
The
will
logical
tw o relations,
R and
the class of couples
(x, y)
such
that either x has the relation
to y or x has the relation
S to
y.
R+S
be
"
Either
fl of or
of".
15-06
-R = xy-(xRy).
the relation of x to y
Def.
-R
is
important to + S)y, and # -E y are significant assertions. z(#

It is
}
when x does not have the relation R to y. note that R x S, R + S and -# are relations: x(R x *$)#,
and the
"null-relation"
The
"universal-relation"
are also definable
after the analogy to classes.
15-07
x y
[?(x,
y)c{(x,
y)].
Def.
is
x has the universal-relation to y in case there

$(x, y)
a function,
f,
such that
c(x,
y),
i.
e.,
in case x
and y have any
relation.
15-08
-1.
Def.
for relations different
Of course, 0,1, + and x have different meanings
from their
meanings for classes or for propositions. 0, + etc., are strictly analogous.

,
But these
meanings ot
As was pointed out
in Section III of this chapter, for
every theorem
involving functions of one variable, there is functions of two variables, due to the fact that a function
a similar theorem involving

<p(x
y)
may
be
for
regarded as a function of the single variable (x, y). Consequently, each theorem of the calculus of classes, there is an exactly corresponding
theorem
in the calculus of relations.
We
may, then,
cite as illustrations of
this calculus the analogues of

classes;
the theorems demonstrated
to hold
for
and no proofs
will here
be necessary.
These proofs follow from the
theorems of Section
are given
III,
numbered
11-, exactly as the proofs for classes

in Section II,
by the corresponding theorems

y)].
numbered 10
15-1
= x$-[t(x,y)c(x,
null-relation
i.
The
is
the relation of x to y
when
it is
false that $(x, y) implies

is
(x, y), (x, y)

19
e.,
when
x has no relation to #. 19
Of course, there
no such
couple which can significantly be called a couple. As in the case of the 1 and of the class calculus, the 1 and
of relations, defined as
272
15-2
A
nx
,
Surrey of Symbolic Logic
y [(x,
y)
e 1].
Every couple
is
member
of the universe of couples, or has the universal
(dyadic) relation.
15-3
nx
-[(x, 2/)eO].
No
couple has the null-relation.
15-4
15-5
jRcl. Ocfl.
relation, R, is implied
Every
by the
null-relation
and implies the universal
relation; or, whatever couple (.r, y) has the null-relation has also the relation 72, and whatever couple has any relation, R, has also the universalrelation.
15-6
(RcS) =
S".
tt x
For
relations,
RcS
By
is
[(xRy)c(xSy)}. more naturally read

"
"
implies
S"
than
"
is
con
tained in
if
15-6,
x has the relation
R to
y,
means "For every x and every y, implies then x has the relation S to Or R implies
S"
"
y".
S"
means
"Every (x,
y) couple related
y [(x
by
are also related
by
S".
15-7
(R =
S)
= Ux
Ry) =
(x
y)].
Two
relations,
R
S,
and
S, are equivalent
when the
couples related by
JR.
are
also related
by
and
vice versa (remembering that
is
always a reciprocal
relation
c).
15-8
(RcS) = (-Sc-R).
If the relation
implies the relation S, then
when S
is
absent
also will
be absent.
15-9
[(RcS) *(ScT)]c(RcT).
implication of one relation
The
by another
is
15-91
If
[(RcS)x(x m
Ry
)]c(x m
Sy n ).
this couple are
implies
related also
S and a given couple are related by R, then by S.

S)
15-92
[(R
x(S
==
T)]c(R =
is
T).
The equivalence
If it
of relations
transitive.
be supposed that the postulates concerning the existence of rela tions are satisfied by exhibiting the functions which determine them, then,
we have
nificant,
defined them, are such that propositions involving them are true whenever sig and significant whenever the prepositional functions determining the functions in question have the same range.

as in the case of classes,
273
we can
derive the postulates (or remaining postu
lates) for a calculus of relations
sitional functions.
from the theorems of the calculus of prepo The demonstrations would be simply the analogues of
those already given for classes, and
may
be omitted.
16-2
16-3
(RxR) = (RxS) =
R.
(S xR).
16-4
16-5
(RxS) xT = Rx(SxT). RxO = 0.

[(Rx-S) = 0}c[(RxS) =
R] H(RxS) = -(-Rx-S). (R + S) (RcS) = [(RxS) = R].
----
16-61
R].
16-62 16-8 16-9
x[(Rx-S) =
R}}
c(R =
0).
These theorems
may
also be taken as confirmation of the fact that
the Boole-Schroder Algebra holds for relations.

relations"
In
fact,
"calculus
of
most frequently means

a, b, c,
just that
the Boole-Schroder Algebra
with the elements,
etc., interpreted as relations taken in extension.

is
So
far,
the logic of relations

are
a simple analogue of the logic of classes.
But there
analogies,
many
properties of relations for
which
classes present no
h. Fact, and these peculiar properties are mosl important, of mathematics, worked out by Peirce, Schroder, the logistic development Peano and his collaborators, and Whitehead and Russell, has de
Frege,
While pended very largely upon a further study of the logic of relations. manner we can do no more, within reasonable limits, than to suggest the of this development, it seems best that the most important of these proper
ties of relations
should be given in outline.
But even
this outline
cannot
be complete, because the theoretical basis provided by our previous dis

cussion
is
not sufficient for completeness.

relation, R, has a converse,
Every
17-01
If
*R, which can be defined as follows:
*R =
y x (x Rij).
Def.
x has the relation
to y, then y has the converse relation,

e
R, to
x.
It follows at
once from the definition of (x m y n )
that
xm
because
(x m
R yn = R
yn
*R x m
e
"R
Ry
n)
(x m y n )
R =
(y n x m )
is
Vn
"R
xm
>
The converse
of the converse of
R.
w(wfl)
= R
19
274
since
A
=
(*>R)
x y (x R y) proof would require that we demonstrate

x y (y
R x) =
R.
(This
is
not
proof:
n,
[(*, y) e
~(~R)
(x, y) e
R]
But
it
is
obvious that such a demonstration
may
be given.
In general,
we shall not pause for proofs here, but merely indicate the method of proof.) The properties of symmetrical relations follow from the theorems con
cerning converses.
versal relation,
(x 1 y)
1,
The uni For any symmetrical relation T, T = and the null-relation, 0, are both symmetrical:
*->T.
[rO, y) cf(s,
y)]
==
==
[f(y, x)
c{(y,
.r)]
(y I x)
(The
"1"
in the
middle of this
proof
is
obviously that of the calculus of
propositions.
(
Similarly for
in the next.)
y)
-[r(.r, y)
if
$(*, y)]
-[f (y, x)
{(y, x)]
(y
a-)
It is
obvious that
two
relations are equivalent, their converses will be
equivalent:
(#
(R
$)
in
Not
quite so obvious
is
the equivalent of
(RcS),
terms of
Instead
*R and
^S.
We
might expect that (R c S) would give (/S
c^).
we have
(RcS) = (*R c
for
(fl
S)
y [(y
"R
,S)
= H Xf
y [(.i-
E y)
(x
y)]
= Ux
is
x)
(y
^S
x)]
= (vRc^S)
"
Parent of implies
of".
ancestor
of"
equivalent to
"Child
of
implies
descendant
The converses
of
compound
relations
is
as follows:
(/JxS)
for
= vRx^S
(x
(R x S)y
= y(R
x S)x
(y
R x) x(yS x) =
y ) x (x ^S y)
=
If
x(*Rx"S)y
is
employer and
by".
exploiter of y, the relation of y to
.r
is
"employee
of
and exploited
Similarly
If
is
either
employer or benefactor of
y,
the relation of y to x
is "either
by". employee Other important properties
of or benefitted
of relations
concern
"relative
sums"
and

"relative
275
These must be distinguished from the non-relative sum and product of relations, symbolized by + and x The non-relative of "friend and "colleague is "friend and product colleague
products".
.
of"
of"
of":
their relative product
is
"friend
of a colleague
of":
of".
Their non-relative
sum
by
is "either
friend of or colleague
of".
their relative
sum
is "friend
of
every non-colleague
We
shall
denote the relative product of
and S
S, their relative
sum by
R t S.
Def.
relation
z
17-02
S = xz{2 y [(xRy) x(ySz)]}.
R\S
to y
for
(x, z) when for some y, x has the and y has the relation S to z. x is friend of a colleague of some y, x is friend of y and y is colleague of z.
is
the relation of the couple
when,
17-03
R-tS = xz{n y [(xRy)

relation of x to z
+ (ySz)}}.
Def.
R t S is the R to y or y
It is
has the relation S to

y, either
when, for every y, either x has the relation z. x is friend of all non-colleagues of z
is
when, for every
is
friend of y or y
colleague of
z.
noteworthy that neither relative products nor relative sums are commutative. "Friend of a colleague is not "colleague of a friend
of"
of".
Nor
is "friend
of".
of all non-colleagues
of"
the same as
"colleague
of all non-
friends
But both
relations are associative.
R\(S\T) = (R\S)\T
for
2 t {(wR
x)
x X (S T)z]
[
2 X (w R x) x z v [( x Sy)x(yTz)]} = 2 y 2 x {(wRx)x[(xSy)x(yTz)]}
--=
{
= 2 v 2,{[(wRx)x(xSy)]x(yTz)} = 2 v (2 x [(wRx)x(xSz)]x(yTz)} = 2 v {[w(RS)y]x(yTz)}

"Friend
of a (colleague of a neighbor
of".
of)"
is "(friend
of a colleague) of a
neighbor
Similarly,
"Friend
R
all
t (S t T)
= (R
t S) t
T
of)"
of all (non-colleagues of all non-neighbors
is
"(friend
of all
non-colleagues) of
non-neighbors
of".
De Morgan
ucts.
Theorem holds
for the negation of relative
sums and prod
-(R\S)
for
= -flt-S
--=
}
-{
2 t [(x
Ry)*(yS z)]
tf
-[(*
R y)
x (y S
z)}
276
A
of
"friend

is "non-friend
The negative
Similarly,
of a colleague
of"
of all colleagues
(non-non-colleagues)
of".
-(R t S) = -R -S
of
"friend
The negative
colleague
of".
of all non-colleagues
of"
is "non-friend
of a
non-
Converses of relative sums and products are as follows:
for
x *(R S)z
|
= z(R
S)x
= ? y [(z Ry)*(yS x)] = 2 v ((y S x) x(zR y)]
If
is
fitted
employer of a benefactor of by an employee

of".
z,
then the relation of
z to
"
is
bene-
Similarly,
If
^(R t
all
/S)
= ^S
^R
a;
is
hater of
non-helpers of
by".
z,
the relation of z to
is "helped
by
all
who are not hated The relation of

relative addition.
relative product
is
distributive with reference to non-
for
x[R (S + T)]z
\
R\(S+T) = (RS) + (R T) = ? y {(x Ry)x [y(S + T)z]

}
X ^^2/
f 1
(T v**
7? tl/
ii\ c//
(11 L\c/
F
Si
?^
/
+ +
T?y
\"
y 2^1 /J
1
)
= 2 y {[(x Ry)*(yS
Similarly,
"Either
z)]
[(x
Ry)x(y T z)}}
(R + S) T = (R T) + (S T)
\ \
\
friend or colleague of a teacher
of"
is
the same as
"either
friend
of a teacher of or colleague of a teacher
of".
somewhat curious formula
is
the following:
It holds since
x[R (S x T)]z
\
R\(S*T)c(R\S)x(R\T) = ? y {(x Ry)x [y(S x

=
T)z]
Z y {(xRy)x[(ySz)x(yTz)}}
and
since a x (b xc)
(a
x6) x (a xc),
=
z y {[(x
2,[(.T
Ry)*(yS z)] x [(x Ry)x(yT z)] Ry)*(yS z)] x ? y [(x Ry)*(yT z)]
And
this last expression
is
[x(R\S)z] x[x(R\ T)z].

If
277
is
student of a friend and colleague of

z.
z,
then x
is
student of a friend and
student of a colleague of
cause
"student
The converse
colleague"
of a friend
and
implication does not hold, be requires that the friend and the
friend and student of a col colleague be identical, while "student of a where S y is repeated, last step in the proof does not. (Note the league"
,
and observe that

Similarly,
this step carries exactly that significance.)
(flxS) T c (R\ T) x (S T)
\
The corresponding formulae with t

and seldom
useful; they are omitted.
instead of
are
more complicated
is
no particular importance, but the relative product In terms of this idea, "powers" of a relation are a very useful concept.
The
relative
sum
is
of
definable
# = R
2
JR,
# = R
3
2
|
R,
etc.
transitive relation, S,
S n cS.
The
predecessors
S 2 c 8, and hence distinguished by the fact that of predecessors ... of x are predecessors of predecessors of the powers of a relation plays a This of x.
is
conception
prominent
ties of
part in the analysis of serial order,
and
of the
fundamental proper
the
of
number
20
series.
By
use of this and certain other concepts, the
method
"mathematical
induction"
can be demonstrated to be com
pletely deductive.
were not given were letters by which relations were symbolized separate treatment. The Any also interpreted as relative terms by a sort of systematic ambiguity. class of entities which have that relation also stood for the
In the work of
De Morgan and
Peirce,
"relative terms"
relation
symbol
to something.
But
in the logistic
development
of
mathematics, since that
time, notably in Principia
Mathematical
relative
terms are given the
The "domain" of a given separate treatment which they -really require. relation R to some that is, the class of entities which have the relation, R be defined as follows: fl, which can thing or other-may be symbolized by
17-04
DR =
of
&[2y(xRy)].
Def.
"For
The domain
x has the
R is the class of x s determined by the function D R will If R be "employer of relation R to

",
some
y,
be the class
y".
of employers.
The
20
21
"converse
i,
domain"
of fl
that
is,
the class of things
See Principia,
See
i,
Bk. n, Sect. E.
*33.
The notation we
use for domains and converse domains
cipia.
278
which something or other has the relation and similarly defined:

17-05
(I
R may
be symbolized by
CL
R =
x,
y[2,(xRy)].
Def.
is
The
will
converse domain of
the class of y
"For
some
a-
has the relation
to
y".
determined by the function dR If R be "employer

of",
be the class of employees.
The domain and converse domain

the
"field "of
of a relation, R, together constitute
R,
R. + (yRx)]}.
all
17-06
C R = xi2 y [(xRy)
The
field of
will
If
be the class of be
"employer
terms which stand in either place in
the relation.
of",
CR
is
the class of
all
those
who
are
either employers or employees.
The elementary
properties of such
"relative terms"
are
all
obvious:
xn
eC R = 2 y [(x n Ry) C R = D R+a R
(yRx n )}
However,
are of
"
development of mathematics, these properties 22 the highest importance. We quote from Prirwipia Mathematical
for the logistic
Let us ... suppose that
is
the sort of relation that generates a series,
say the relation of less to greater among integers. Then that are less than some other integer = all integers, Q
that are greater than some other integer
case,
DR =
R =
all
integers integers
all
all
integers except
0.
In this
C R = all integers that are either greater or less than some other = all integers .... Thus when R generates a series, C R becomes integer
important.
..."
have now surveyed the most fundamental and important characters of the logic of relations, and we could not well proceed further without
elaboration of a kind which
is
We
here inadmissible.
But the reader
is
warned
that
we have no more than
scratched the surface of this important topic.

"What
About
1890, Schroder could write
a pity!
it".
developed instrument and nothing to do with
To have a highly And he proceeded to
make
beginning
in the bettering of this situation
of relatives to the logistic
development
in the
of certain
by applying the logic portions of Dedekind s

of
theory
of
number.
Since that time, the significance of symbolic logic has
been completely demonstrated

22
1,
development
Peano
Formulaire
p. 261.
279
and
of Principia Mathematica.
is
And the very head and front of this develop
ment
a theory of relations far
previously given.
We
more extended and complete than any can here adapt the prophetic words which Leibniz
"I
puts into the
begin to get a very different opinion I had regarded it as a scholar s of logic from that which I formerly had. in the way you understand it, it is a kind of diversion, but I now see that,
of Philalethes:
mouth
universal mathematics."
VI.
THE LOGIC OF
Principia Mathematica
of the Boole-Schroder Algebra
We
have now presented the extensions
the Two-Valued Algebra, prepositional functions and the propositions derived from them, and the application to these of the laws of the Two-
Valued Algebra, giving the calculus of propositional functions. Beyond with the Twothis, we have shown in outline how it is possible, beginning Valued Algebra as a calculus of propositions, to derive the logic of classes in a form somewhat more satisfactory than the Boole-Schroder Algebra, In so doing, we have presented of relations and relative terms. and the
logic
as
much
work
of that
of
the
being followed the exact forms which that development took

considerably modified
it
But, our purpose
development which begins with Boole and passes through Peirce to Schroder as is likely to be permanently significant. we have not here expository rather than historical,
Instead,
we have
in the light of
since the publication of the
work
of
what symbolic logicians have learned Peirce and Schroder.

in detail our divergence
Those who are interested to note

historical
from the
development
of
will
and VIII
Chapter
II
I.
be able to do so by reference to Sections VII But it seems best here to point out briefly what
these alterations are that

interpreted S^.r,
^r,
we have made.
In the
first
place,
we have
S^(.r, y), etc.,

,
propositions of the form ^r n iK-r m y n ), this, in consideration of the serious

<px
sums or products of Peirce and Schroder avoided etc. But while theoretical difficulties.
explicitly as
still as an actual product, S#c as an actual sum, they did not treat II which result the laws which they give for propositions of this type are those method by from such a treatment. There is no slightest doubt that the and formulated these laws is substantially the one
which Peirce discovered
which we have exhibited.
symbol for a a sum, makes demonstration possible where product, S#r as the symbol of made and, for further otherwise a large number of assumptions must be resorted to. a much more difficult and less obvious style of proof
And
this explicit use of n^.r as the
principles,
280
In this part of their work, Peirce and Schroder can hardly be said to have formulated the assumptions or given the proofs.
In the second place, the Boole-Schroder Algebra
of
the general outline
which
is
already present in Peirce s
work
probably seemed to Peirce

(though there are indications
and Schroder an adequate calculus
of classes
With this system before in the paper of 1880 that Peirce felt its defects). them, they neglected the possibility of a better procedure, by beginning
with the calculus of propositions and deriving the logic of classes from the
laws which govern propositional functions. And although the principles which they formulate for propositional functions are as applicable to func
tions of one as of
their interest
relatives
two
variables,
and are given
for
one as well as for two,

of
was almost
entirely in functions of
two and the calculus
which
may
be derived from such functions.

is,
The
logic of classes
which we have outlined

for,
then, something which they
laid the
foundation
but did not develop.

exposition thus far in the chapter have been
relation of this earlier treatment of symbolic
The main purposes of our two: first, to make clear the
logic with the later and better treatment to be discussed in this section;
and second, to present the logic of propositional functions and their deriva tives in a form somewhat simpler and more easily intelligible than it might
otherwise be.
The
theoretically sounder
and more adequate
logic of Prin-
cipia Mathematica
is
given a form which

is
so far as propositional functions

its
and
their derivatives
concerned
seems to us to obscure, by
notation,
the obvious and helpful mathematical analogies, and requires a style of

proof which
is
much
disclaim any idea
With regard to this second purpose, we that the development we have given is theoretically
less
obvious.
adequate
its
chief value should be that of

difficult
tory to the
more complex and
an introductory study, prepara treatment which obviates the the
oretical shortcomings.
Incidentally, the exposition
which has been given
will serve to indicate
how much we
are indebted, for the recent development of our subject, to
the earlier work of Peirce and Schroder.
The Peirce-Schroder symbolic

Peano
s
logic
is
closely related to the logic of
connection
Formulaire de Mathematiques and of Principia Mathematica. This is easily overlooked by the student, with the result that the sub
the Boole-Schroder Algebra and
its
ject of his first studies

likely to
logistic
applications
is
seem quite unrelated to the topic which later interests him the development of mathematics. Both the connections of these two
281
and
their differences are important.
We
shall
attempt to point out both.
And
because, for one reason, clearness requires that
we
stick to a single
illustration,
our comparison
will
be between the content of preceding

logic of
sections of this chapter
and the mathematical
Book
I,
Principia
Mathematical
The Two- Valued Algebra
is
a calculus produced by adding to and re
interpreting an algebra intended primarily to deal with the relations of And it has several defects which reflect this origin. In the first classes.
place, the
same
logical relation is expressed, in this system, in

"If
two
q
different
ways.
per",
We
have, for example, the proposition

p, q,
peg
. .
and
.
,
r,
then
..."
where
and
"
are propositions.
relation
"
But
"if
then
q,
is
supposed to be the
Also,
same
in
which
is
r
expressed by c in p c
is
r,
and per.
system
"either
"and
pc
q and q c
"
the relation which
is
other
wise expressed
by
and
use
so on, for the other logical

of
"if
.
. .
relations.
The
. .
.",
involves
.
the
. .
.",
then
to
.
",
"...
and
or
.
.
"...
is
equivalent
.
.",
and
"...
is
not
equivalent to
.
",
just as
,
exactly the relations c
any mathematical system may; yet these are + = and =H whose properties are supposed to
,
,
be investigated in the system.
Thus
the
system takes the
laics of the logical
relations of propositions for granted in order to prove them.
Xor
is
this
paradox removed by the fact then ..." and c, ability of

"if
.
.
that we can demonstrate the

of
".
interchange-
and
..."
and x,
etc.
For the very demonstration of this interchangeability takes for granted the logic of propositions; and furthermore, in the system as developed,
it is
form impossible in most cases to give a law the completely symbolic until it has first been proved in the form which involves the non-symbolic
So that there
demonstration of the laws of
expression of relations.
in the
which the circularity propositions can be removed in this

is
no way
in
system.
Another defect
of the
"p
Two- Valued Algebra

is
true"
is
the redundance of forms.

p,
s
The
23
proposition p or
is
symbolized by
method
of
by p =
1,
by p
is
4=
Logically, as well as historically, the
Peano
Formulaire
a sort of
The general intermediary between the Peirce-Schroder mode of procedure and Principia. method of analysis and much of the notation follows that of the Formulaire. But the Formulaire is somewhat less concerned with the extreme of logical rigor, and somewhat more concerned with the detail of the various branches of mathematics. Perhaps for this fundamentals which is the dis reason, it lacks that detailed examination and analysis of
tinguishing characteristic of Principia. of the relation D (in our notation, c
):
For example, the Formulaire retains the ambiguity p?q may be either the class p is contained in
q".
the class
laire
or "the proposition p implies the proposition contains no specific theory of propositions.

q",
In consequence, the Formu
282
the negation of
A
p
or

is
false"
etc.,
"p
by
-p,
p =
0,
-p =
1,
=}=
1,
etc.
These various forms may, it is true, be reduced in number; p and -p may be made to do service for all their various equivalents. But these equivalents
cannot be banished, for in the proofs
that
it is
necessary to
0)
make
is
use of the fact
p =
(p
1)
(p 4= 0),
-p =
(p
(-p
1), etc., in
order to
demonstrate the theorems.

avoidable.
Hence
this
redundance
not
altogether
calculus of propositions in Principia

etc.,
Both these defects are removed by the procedure adopted for the Mathematical Here p = 1, p = 0,
are not used;
instead
we have simply p and

it
its
negative, symbolized
by ~p. may every mathematical system has always taken

And, impossible as
ideas
are:
seem, the logic of propositions which

for granted
is
not presumed.
(2)
The
primitive
(1)
elementary propositions,
elementary
the
propositional functions, (3) assertion, (4) assertion of a prepositional func

tion,
(5)
negation, (6) disjunction, or the logical sum;
and
finally,
idea of
but
is
which does not belong in the system a notation to indicate that one symbol or complex of symbols merely
"equivalent
by
definition",
be replaced by another. An elementary proposition is one which does not involve any variables, and an elementary propositional function is such as "not-p" where p is an undetermined elementary proposition.
may
The
what would be supposed a proposition may be asserted or merely considered. The sign h prefaces all propositions
idea of assertion
is
just
which are asserted.

where
An
q.
asserted propositional function
is
such as
"A
is
"
is
undetermined.
p v q, instead of p + p and q, is true".
pv
The disjunction of p and q is symbolized by At least one of the two propositions, q means
"
The
#1-01
"p
postulates and definitions are as follows:
poq. =
(
~pvq.
q"
Df.
is
is is
(materially) implies
the defined equivalent of

true
,
"At
least
one of
the two,
is
false
and
is
a true
is
proposition".
(The explana
tion of propositions here
ours.)
its
p^q
the same relation which
we
have symbolized by p c
q,
not
converse.
Principia.
(The propositions quoted will be given the number which they have in The asterisk which precedes the number will distinguish
in earlier chapters or earlier sections of this
them from our propositions

chapter.)
The
24
logical
i,
product of p and q
is
symbolized by p
q,
or p
q>
See Bk.
Sect. A.
283
*3-01
"p
p.q. =
true
~(~p
v~ry).
is
I)f.
"It
is
and
is
true"

</,
is
false that at
least
This is false". one of the two, p and Morgan s Theorem in our notation, (p q) =
is,
of course, a
form of
De
.
-(-/>
-(/).
The
#4-01
"p
(material) equivalence of
p and
q
.
is
symbolized by p
q or
r/.
= ,p?q q^p.
.
q"
Df
is
is
(materially) equivalent to

p".
"p
(ma
terially)
implies q
q)
and
g (materially) implies
In our notation, this

==
would be (p =
= (pcq)(qcp).
Note that
...
...
and
Df
are different relations in Principia.
The dots
in these definitions serve as punctuation in place of parentheses
and brackets.
takes precedence over one, as a bracket over a In #4-01 we have only one dot after =, parenthesis, three over two, etc. because the dot between p D q and q D p indicates a product a dot, or two
dots,
:,
:
Two
dots, indicating a product
is
always inferior to a stop indicated
by the
same number
of dots
but not indicating a product.
The
follows: postulates of the system in question are as
*!!
*1-11
Anything implied by a true elementary proposition

stands for
<px
is
true.
Pp.
("Pp."
"Primitive proposition".)
can be asserted, where x is a real variable, and <pxo\f/x can be asserted, where x is a real variable, then #r can be asserted, where x
When
is
a real variable.
Pp.
A
#1-2
"real
variable" is
such as p in -p.
Pp.
p.
\-:pvp.?.p.
In our notation, (p + p} c
*l-3
\-lq*-D.pvq.
Pp.
q).
In our notation, q c (p +
*l-4
\-mpvqmOmqvp.
q)
Pp.
In our notation, (p +
c(q +
p}.
*l-5
\-:p
v(q vr) .3.
v (p vr).
r)]
Pp.
r)].
In our notation, [p +
(q
+
.
[q
+ (p +
r.
#1-6
h!
^
<7
pv
c
pv
+
Pp.
r)].
In our notation,
(</
r)
[Q;
(/)
c Q; +
Note that the
the sign of assertion in each of
above
is
followed by a
284
sufficient
A
number
of dots to indicate that the
whole of what follows
is
asserted.
$K-l-7
If
is
an elementary proposition, ~p
is
an elementary proposition.
PP.
*1-71
If
p and
Pp.
<pp
q are
elementary propositions,
pvq
is
an elementary
proposition.
#1-72
If
and
\I/p
are elementary prepositional functions
which take
elementary propositions as arguments,

tional function.
v
<pp
\f/p
is
an elementary proposi-
Pp.
list of assumptions. The last three have to do with the method by which the system is developed. By *l-7, directly any proposition which is assumed or proved for p may also be asserted to
This completes the
hold for ~p, that
is
any proposition
or q or
r,
of
~p may be substituted for p or q or r, etc., in the system. By #1-71, p v q may be substituted for p
to say,
72,
if
etc.
And by *1
"
any two complexes
of the foregoing
symbols
in the
which make sense as
statements"
can be treated in a certain
way
all
system, their disjunction can be similarly treated.

of these,
.
By
q,
the use of
three
q
any combination such as p v
q,
q,
p3
q,
i>
p,
~p v . p v q, ~p v ~ry, etc., etc., may be substituted for p or q or r in any assumed proposition or any theorem. Such substitution, for which no
postulates would ordinarily be stated,
is
one of the fundamental operations
by which the system

Another kind
for
is
developed.
of substitution of
which
is
fundamental
is
the substitution
exists.
".
any complex
is
symbols
of its defined equivalent,
where such
This operation
covered by the meaning assigned to "... = ... Df Only one other operation is used in the development of this calculus
elementary propositions the operation for which *!! and ^1-11 are assumed. If by such substitutions as have just been explained there
of
results a
complex
of
and
if
that part of
which the main, or asserted, relation is o the expression which precedes this sign is identical with a
symbols
in
,
postulate or previous theorem, then that part of the expression which

follows this sign
may
. . .
be asserted as a
lemma
or
new theorem.
In other
words, a main, or asserted, sign D has, by *!! and *1-11, the significant then This property is explicitly assumed property of
"If
,
.".
in the postulates.
The main thing

not so
to be noted
about
this operation of
inference
strictly
much a piece of reasoning as a mechanical, or mathematical, operation for which a rule has been given. No
is
that
it is

"mental"
285
involved except that required to recognize a previous proposition followed by the main implication sign, and to set off what
operation
is
follows that sign as a
then,
assertion. The use of this operation does not, that the processes and principles of ordinary logic are tacitly presupposed as warrant for the operations which give proof.
new
mean
What is the significance of this assumption of the obvious in # 1

*l-7,
1,
#1
1 1,
#1-71, and *l-72?

so
Precisely
this:
these
is
postulates explicitly
assume
much
of the logical operations as
necessary to develop the

is
system, and beyond
this the logic of propositions simply

it
not assumed.
To
illustrate this fact,
will
be well to consider carefully an exemplary
proof or two.
#2-01
\-ipD~p.o.~p
Dem.
Taut M-: ~p v ~p
01)]
.
~p
(1)
Vip?~p.?.~p
#1-2 above.
}
"Taut"
is
the abbreviation for the Principle of Tautology,
~plp indicates that
~p
is
substituted in this postulate for p
giving (1).
This operation is valid by #1-7. Then by the definition #1-01, above, p D ~/j is substituted for its defined equivalent, ~p v ~p, and the proof is
complete.
*2-05
\-mmqDrm3ipoqm3mp3r
Dem.
Sum [(1)
.
M-:
=>
~p v
~p v
.
(1)
O1-01)]
\-: .
qor .D: pDq .? p?r

And
(1) is
Here ~p
is
"Sum"
refers to
*l-6, above.
what
*!(>
becomes when
Then, by #1-01, p^q and pir are substituted for their defined equivalents, ~pvq and ~pvr, in (1), and the resulting
substituted for p.
expression
is
the theorem to be proved.

illustrates the use of
m"3l
The next proof
*1
and #1-11.
*2 06
\-l ."3
-zr
."3
Dr
Dem.
p,
Dr.D:pDg.D.pDr:.3:p3g
[#2 -05] hs
[(1)
.
.3Sg
DrD.p3r
por
(1) (2)
3r
D:p3
-
s. par
(2)
*1-11] h:
/>3r/.D:
q?r .0.
286
"
A
is
"Comm
When,
it
in
*2-04, previously proved, which is p . D . q D r : D : q D . p D this theorem, q D r is substituted for p, p D q for q, and p D r for
r.
r,
becomes the long expression (1). Such substitutions are valid by *l-7, #1-71, and the definition ^1-01: if p is a proposition, ~p is a proposition; if ~p and q are propositions, ~p v q is a proposition; and p D is the defined If we replace equivalent of ~p v q. Thus p D q can be substituted for p.
ry
the dots by parentheses,

h
{
etc., (1)
becomes
r)]
}
(q
r)
[(p
D
</)
D
(/>
(p
ry)
[(ry
r)
D (p D
r)]
But, as (2) states, what here precedes the main implication sign is identical with a previous theorem, *2-05. Hence, by #1-11, what follows this
the theorem to be proved can be asserted. sign Further proofs would, naturally, be more complicated, but they involve no principle not exemplified in the above. These three operations sub
stitutions according to *l-7,
main implication
*1-71, and *l-72;
substitution of defined
are the only
equivalents;
and
"inference"
according to #1-1 and #1-11
processes which ever enter into any demonstration
in the logic of Principia.
The
result
is
of propositions for
that this development avoids the paradox of taking the logic granted in order to prove it. Nothing of the sort is
explicitly stated postulates
assumed except these

served.
whose use we have ob

is
And
it
results
from
this
mode
of
development that the system
completely symbolic, except for a few postulates, #1 1, #1-7, etc., involving no further use of then "either ... or ...","... and
"if
. .
.",
"
etc M
\^
We have now seen that the calculus of propositions in Principia Mathematica avoids both the defects of the Two-Valued Algebra. The further comparison of the two systems can be made in a sentence Except for the
:
absence, in the logic of Principia, of the
=|=
0, etc., etc.,
redundance of forms, p, p = 1, and the absence of the entities and 1, the two systems
of this part of Principia
are identical.
into a valid
Any theorem
of the
can be translated
Two-Valued Algebra, and any theorem of the Two-Valued Algebra not involving and 1 otherwise than as {=0} or
can be translated into a valid theorem of Principia. In fact, the qualification is not particularly significant, because any use of and 1 in the Two-Valued Algebra reduces to their use as = and = 1 For
{
theorem
as a term of a sum,
and
as a factor, immediately disappear, while the

1 in
presence of
as a factor
wise expressed.
and the presence of But p = is -p, and p =
sum can always be
other
1 is p.
Hence the two systems
287
are simply identical so far as the logical significance of the propositions
they contain
is
concerned. 25
of our treatment of propositional functions with the
The comparison
same topic
In the
in Principia is
first place,
not quite so simple. 28

is,
there
in Principia, the
"theory
of
types,"
which
concerns the range of significance of functions.

sideration of this.
But we
of
shall
omit con
Then, there are the differences
notation.
Where
we
or
write
U<px,
or
H x (px,
by
Principia has
.
<px.
2x
<px,
Principia has (Kx)
and where we write S#r, (or) further and more important difference
.
<px\
may
be
made
-{(.?)
clear
citing the
.
assumptions of Principia.
~#r.
~<px.
*9-01
.?&}. =
(3.r)
.
Df. Df.
Df.
*9-02
*9-03
25
~{(3a:)
(x)
.
<px
.?&}. =
fa)
(x)
.vp
<px
vp.
This
may be proved by
are contained
amongst the propositions
noting that, properly translated, the postulates of each system of the other. Of the postulates in Principia,
rendered in our notation:
^1-01
is
(peg)
(-p +
q),
is is
which
is
contained in our theorem 9-3.
^1 -2
sfcl
is
is,
(p + p) cp, which
a consequence of our theorems 2-2 and 5-33.
-3
pc
p+
(p + q),
which
our theorem 5-21.
^1 -4
^1-5
by
êl-6
is
is,
(p + q) c(q + p), which follows from our theorem 4-3,
by
2-2.
+
(</
r)
cq
+ (p +
?),
which
is
a consequence of our theorems 4-3 and 4-4,
2-2.
is (q cr) c [(p + q) c(p + r)], which is a consequence of our theorem The remaining (non-symbolic) postulates are tacitly assumed in our system.
5-31,
by
2-2.
Of our postulates, 1-1-1-9

1 is
in
Chap,
n and
9-01 in Chap, iv:
a consequence of
^1-7 and ^1-71
in Principia.
2
3
is
is
-^4 -24 in Principia.
^-4-3 in Principia.
4
5
of
is
^2-3
in Principia.
"If
is
equivalent to
0,
then a x
=
0",
hence to -x c -(a
x),
which
is
a consequence
*3-27
in Principia,
by *2-16.
= x), is a consequence of -#-4-71 and ^-4-61 in Principia, 1-61, in the form -(x -a) c (x a by ^4-01 and *3-26. = y)]c-y, is a consequence of #4-71, *5-16, and 1-62, in the form [(y a = y)(y -a ^2-21 in Principia. to [(a; = 1)Q/ = 0)] c (x = -?/), hence to (x -y) c (x = -y), which is an 1-7 is
equivalent
immediate consequence
1-8
1-9
is is
of ^-5-1 in Principia.
^K-4-57 in Principia.
^4-71
is
in Principia.
(q
9-01
equivalent to
1)
[p
(p
q)},
hence to q c
[p
(p
q)],
which
is
an
immediate consequence
26
of -^5-501 in Principia.
See Principia,
i,
15-21.
288
In this
A
last,
two
sides.
note the difference in the scope of the "quantifier" (x) on the If the dots be replaced by parentheses, *9 03 will be
{[(x)
.
^r]
vp] =
{(x)
[tpxyp]}
in the scope of each of the further definitions. 27
A similar difference
*9-04
p
.
(x) or (3.r)
on the two sides characterizes
(x)
<px
<px
=
:
(x)
.
pv
.
<px.
Df.
p.
*9 05
#9 -OG
*9-07
(3.r)
v
.
= =
.
(3.r)
<px
Df
p ***
(.r)
.
("3.x)
<px
(3z)
:
pv
(#)
(.T)
:
<px.
Df.
.
<px
mv
.
(3#)
.
(.I-)
^
^r
= =
(3#)
^ v ^.
$y v
^.r
.
Df.
*9 OS
(3y)
^
.
(3y)
Df
Besides these definitions, there are four postulates (in addition to those
which underlie the calculus of elementary propositions).
*9 *9
1 1 1
h: hs
<?x
(3
.
z)
<pz.
Pp.
.
<p
<f>x
^?/
(3
2)
Pp.
is
*9- 12
What
is
implied by a true premiss
true.
Pp.
#9-13 In any assertion containing a real variable, this real variable may be turned into an apparent variable for which all possible values are asserted
to satisfy the function in question.
Pp.
is
By
#9-01
for
<px.
proved proposition.
is
-H<px
our method, every one of these assumptions, except ^9-12, In our notation,
= S
=
-<px,
which
is
our theorem 10-1, with
-<px
substituted
*9-02
is
-2^.i<px.
n-^.r, which
is
our theorem 10-12, with
-<?x
substi
tuted for
*9-03
is
tt<px
P = n x (^c + P),
II X
which which
is
is
our theorem 10-32.

10-33.
10-3.
*9-04
*9-05 *9-06
is
is
is
is is
P + n^r = S^ + P = P+ S^r U<px
(P+
<px),
2 x (^-c + P), which
is is
2 X (P+
<px),
which
$y),
^.r),
10-31.
is is
*9-07
*9-08
+ Z^y = U x 2 y = U x 2 y (ty + 2ty+tt<px +

(<px
which which
contained in 12-5,
also contained in 12-5.
The
<py,
<pz,
etc.,
The postulates require explanation. to represent values of the function

<px
<px.
authors of Principia use In other words, where
we have written
"Ibid.,
i,
they simply change the
letter.
This
is
a valid con-
135-38.

vention (though
is
289
it
often renders proofs confusing) because the range of

<p,
<px
determined by not by x, and x is conventions aside indifferent. z in where we should write n is called a "real variable", x in (x) . and (3or) . an "apparent variable". With this it is clear
<pz,
<px
<px
<px,
explanation,
that:
*9-l
#9-11
is
<px
c2<px,
which
c2<px,
is
10-21.
is
is
<px
m+
<px
which
an immediate consequence
of 10-21,
by 5-33.
*9-13
is
"If
whatever value of
is
x, in
<px,
xn
may
be,
<px
n,
then
n^.r,"
and
this implication
contained in the equivalence stated by 10-23.

in Principia
These principles which are assumed

ficient to give all further propositions
Mathematica are suf
concerning functions of one variable,

<px^
<px
without assuming
etc.), (3a-)
.
<px
(x)
.
<*r
to be the product of
of
<px
2,
etc. (or
<py,
<pz,
to be the
sum
.
lf
<f>x
2,
"
etc.
<px
These are simply assumed

x",
as
new
primitive ideas, (x)

"
<?x
meaning
a:".
for all values of
(3or)
.
<px
meaning
<px
for
some values
of
This procedure obviates

<px
all
questions
about the number of values of x
in which troubled us and secures the universality of theorems involving prepositional functions without any discussion or convention covering the cases in which the values of the vari
able are infinite in number.

of
The
proofs in Principia reflect this difference
method.
all
They
+
are, in general,
what ours might have been

etc.,
if
we had
based
further proofs directly upon 10-23 and the propositions con
necting
2<f>x
with 2 x
(<px
+ P),
not making any use, after 10-23,

2<^.r
of the properties of TLtpx as a product, or of
as a sum.
The theory
*11 *11
of functions of
two
variables, in Principia Mathematica,
requires two further assumptions:

-01
(x, y)
.
<p(x,
y)
=
.
(x)
:
(y)
:
.
<p(
y).
<p(
Df.
,
-03
(3z, y)
<p(x,
y)
(3.x)
(30)
y).
Df.
These are identically our assumptions:

11-06
11-05
nx
<p(x,
y)
2 XlV
<p(x,y)
H x ll v y), and = S x S^fey).

==
<f>(x,
The
difference
between the treatment
which
we have given and
the treatment in Principia is not necessarily correlated with the difference between our treatment of propositions and theirs. The
method by which we have developed the theory

20
290
might exactly as well have been based upon the calculus of elementary A few minor propositions in Principia as upon the Two-Valued Algebra.
alterations
would be
sufficient for this change.

is
The
different procedure
for propositional functions, in the twr o cases,
a difference to be adjudged
independently, without necessary reference to the defects of the
Twothere
to the
Valued Algebra which have been pointed out.
Beyond the important

different use of notation.
differences
w hich have been mentioned,

r
are minor and trivial divergences between the
two systems, due
Neglecting these, we may say that the two methods give the same results, with the following exceptions: 1. There are certain complexities in Principia due to the theory of
types.
2.
3.
In Principia the conditions of significance are explicitly investigated. Principia contains a theory of "descriptions", account of which is
here omitted.
But none
of these exceptions is a necessary difference.
They
are due to
the more elementary character of our presentation of the subject. may, then, say loosely that the two methods give identical results.
We
The
calculus of classes
and
of relations
which we have outlined
in the
preceding sections bear a similar relation to the logic of classes and of relations in Principia; that is to say, there is much more detail and com
plexity of theory in Principia, but so far as our exposition goes, the
two are
roughly the same.
And
here there
is
no important difference of method.
It should now be clear how the logic of Principia is related to the logic we have presented, following in the main the methods of Peirce and Schroder. There is much difference of method, and, especially in the case of the cal
culus of propositions, this difference
is
in favor of Principia.
And
in
Principia there
is
much more
of theoretical rigor
and consequent complexity:

"descrip
also there are important extensions, especially in the theory of

tions"
and the
logic of relatives.
But
so far as the logic
which we have
expounded goes, the two methods give roughly identical results- When we remember the date of the work of Peirce and Schroder, it becomes clear
what
is
our debt to them for the better developments which have since
been made.
CHAPTER
The systems
implication,
is
V
1
THE SYSTEM OF STRICT IMPLICATION

discussed in the last chapter were
"The
all
*
based upon material
a false
is
statement, p is true and q false/ p cq meaning exactly We have already called attention to the fact that statement".
"implies".
this
not the usual meaning of

of ordinary inference
is
Its
divergence from the

"A
"implies"
exhibited in such theorems as
false
proposition implies any

2
proposition",
and
"A
true proposition
is
implied
by any proposition". The present chapter intends

sitions
to present, in outline, a calculus of propo

"
based upon an entirely different meaning of implies one more in accord with the customary uses of that relation in inference
is
"-
which
and
proof.
We
shall call it the
system of Strict Implication.
And we
shall
refer to Material Implication,
meaning
it
either the
Two-Valued Algebra or
the calculus of propositions as
appears in Principia Mathematica, since
It will appear that the logical import of these two systems is identical. Strict Implication is neither a calculus of extensions, like Material Impli
and the Boole-Schroder Algebra, nor a calculus of intensions, like It includes relations the unsuccessful systems of Lambert and Castillon. Strict of both types, but distinguishes them and shows their connections.
cation
Implication contains Material Implication, as it appears in Principia Mathematica, as a partial-system, and it contains also a supplementary
partial-system the relations of which are those of intension. The numerous questions concerning the exact significance of implication,
and the ordinary or

Section V.
It will
"proper"
meaning
of
"implies",
will
be discussed in
be indicated
how
Strict Implication,
by an extension to proposi-
tional functions, gives a calculus of classes and class-concepts which exhibits In this, it provides the their relations both in extension and in intension.
Various studies toward this system have appeared in Mind and the Journal of Phi losophy (see Bibliography). But the complete system has not previously been printed. We here correct, also, certain errors of these earlier papers, most notably with reference to
1
triadic
2
"strict"
relations.
illustrations, see
For further
Logic,"
Chap, n, Sect,
etc.
i,
and Lewis,
"Interesting
Theorems in
Symbolic
Jour. Philos., Psych.,
(1913), p. 239.
291
292
calculus of intensions, so often attempted before, so far as such a calculus

is
possible at
all.
I.
PRIMITIVE IDEAS, PRIMITIVE PROPOSITIONS, AND IMMEDIATE CONSE
QUENCES
The fundamental
Symbolic Logic and
1.
ideas of the system are similar to those of
MacColl
its
Applications.
q, r,
They
are as follows:
Propositions: p,
etc.
"p
2.
3.
Negation: -p, meaning

Impossibility: ~p,
3
true".
is
false".
meaning
"p
is
impossible",
or
"It
is
impossible
that p be
4.
The
q
logical product:
is
true".
p xg or p
q,
meaning
"p
and
both",
or
"p
is
true
and
5.
Equivalence:
p =
q,
the defining relation.
Systems previously developed, except MacColl s, have only two truthThe addition of the idea of impossibility and values,
"true"
"false".
gives us five truth- values,

(1)
(2)
all of
which are familiar
logical ideas:
p,
"pis
true".
-p,
"pis
false".
(3)
(4)
~p,
"p
is impossible". is is
p,
"It
false that
is impossible"
i.
e.,
i.
"p
is possible".
"p
(5)
true".
p,
"It
impossible that p be
false"
e.,
is
necessarily
Strictly, the last
two should be written -(~p) and -(-p): the parentheses

at
are regularly omitted for typographical reasons.
The reader need be

it is
no pains to grasp -~p and
as simple ideas:
understand -p and ~p, and to remember that each such It should prefix affects the letter as already modified by those nearer it. is equivalent be noted that there are also more complex truth-values. p as will be shown, but - ~ -p, to p, etc., are irreducible. p,
sufficient to
p,
We
shall
have occasion to make use of only one of these,
"
p,
It is
4
false".
,
false that it is impossible that
p be
true"
i.
e.,
"p
is
possibly
Each one
3
of these
complex truth-values
is
a distinct and recognizable idea,
though they are seldom needed
in logic or in
mathematics.
here use a symbol, ~, which appears in Prindpia Mathematica with a different meaning. The excuse for this is its typographical convenience. 4 is possibly is possibly true" and MacColl uses a single symbol for -~p, p,
"p "p
We
false".
The System of
Strict Implication
293
The dyadic
truth- values
relations of propositions can be defined in terms of these

logical product,
and the
p q?
Def.
true"
1-01
~(pq),
are
Consistency.
"It
poq = -~(pq}.
"It
is
impossible that p and q both be

is Hence -~(pq), and q are consistent".
would be
"p
and
inconsistent".
possible that
p and
q both be
true",
represents
"p
1-02
Strict Implication.
p-*q
~(p-q)-(p-q).
Def.
1-03 1-04
Material Implication.
Strict Logical
pcq =
p+
q
Def.
Def. Def.
p)
.
Sum.
p *q = ~(-p-r/).
1-05
Material Logical Sum.

Strict Equivalence,
= =
-(-p -q)(p
*
1-06
(p
q)
q) (q
-*
Def.
We
here define the defining relation
itself,
because by this procedure we
establish the connection between strict equivalence

Also, this definition
and
strict implication.
makes
it
possible to deduce expressions of the type,
p =
5 something which could not otherwise be done. q mains a primitive idea as the idea that one set of symbols
But p =
q re
may
Def.
be replaced
by another.
1 07
Material Equivalence,
(p
q)
(p c
q) (q
c p).
These eight relations the seven defined above and the primitive rela = q are the relations divide into two sets, p q, p c q, p + q, and p tion, p q
which
figure in
any calculus
of Material Implication.
We
shall refer to
them
as the
"material
relations",
p o
q,
-i
q,
p*q, and p =
".
q involve
the idea of impossibility, and do not belong to systems of Material Impli We may anticipate a These may be called the strict relations cation. of these two sets, which results from the exhibit the
"
little
and
analogy
theorem
~(pq)
shortly to be proved.
Strict relations
:
= -(pti
Material relations
piq =-(pO-q) pAq = _(_ p

(p
pcq =
P+<1
-(p -q)
o-<7)
q)
= -(po-q) x-(qo-p)
(p
=
"
q)
= -(-P-0) = ~(p -tf) *-( c
l~P">
hesitates to The "circularity" here belongs inevitably to logic. No mathematician If theorem A, then theorem B, of two propositions by showing that prove the equivalence know that a reciprc and if theorem B, then theorem A". But to do this he must already And the italicized relation is equivalent to an equivalence. then . .
"if
. .
.
."
to"
be assumed. represents a relation which must
294
The reader will, very likely, have some difficulty in distinguishing in meaning p q from p c q, p A q from p + q. The above comparison may be of assist
-J
ance in this connection, since
it
translates these relations in terms of

q,
"p
p oq
and p
g,"
q.
We
q,
shall be in
no danger of confusing p o
true".
is
consistent with
with p
"p
and
q are both q
Both p A
and p +
would be read
"Either
p or
q".
But p A
q denotes
"
a necessary connection;
p + q a merely factual one.

q,
"2
day
is
Monday",
and
+
is
.
4:".
Let p represent To Then p+q is true but p A q is
false.
In point of
fact, at least one of the
two
is
propositions,
"Today is
Monday"
and
"2
4",
true;
.
.
but there
."
no necessary connection
in this respect.
between them.
the
"Either
or
is
ambiguous
Ask
members
or 2
of
Monday
"Either
+
or
any company whether the proposition "Either today is = is true, and they will disagree. Some will confine
4"
...
..."
to the
p*q
meaning, others
will
make
p =
it
include
the p + q meaning; few, or none, will

Similarly, the difference
make
the necessary distinction.
between p
and p
q is that
q denotes
an equivalence of logical import or meaning, while p = q denotes simply an equivalence of truth-value. As was shown in Chapter II, p = q may be and q are both true or both false". Here again, accurately rendered
"p
the strict relation, p

relation,
q,
symbolizes a necessary connection; the material
g,
a merely factual one.
The
1-1
If
postulates of the
q-i q
system are as follows
p
q are both true, then q
p and
q
and p are both
true.
1-2
-I
p
true, then
If q
and p are both

-*p
is
is
true.
1- 3
If
p
then p
is
true,
true and
is
true.
1-4
If
p(qr)4q(pr) p
is
true and q and
are both true, then q
is
true and
p and
are both
true.
1-5
If
p-i-(-p)
is
true, then
it is
false that
is false.
1-6
If
(p-tq)(q-lr)-l(p-lr)
strictly implies q
and
q strictly implies
r,
then p strictly implies
r.
The System of
1 7
Strict
Implication
295
~p-i-p
If it is
impossible that p be true, then p
is false.
1-8
p
"p
-J
~q
-J
~p
"
strictly implies
is
q"
is
equivalent to
is
impossible
strictly
im
plies
impossible
".
The
as
first six of
these present no novelty except the relation
-J
They
But,
do not, so
far, distinguish this
system from Material Implication.
we
shall see shortly, the postulates 1 7
and 1-8 are
principles of trans
formation;
and thus introduce the distinguishing characteristics of the system.

late 1 -7
is
they operate upon the other postulates, and on themselves, Postu

Postulate
If
1 -8 is
q,
.
(p-iq)
obvious enough. -i -J
(-
~p
~q)
p implies
is
then
is
possible
implies
possible
is
(
(~p
-J
~q)
-J
(-p
-J
-q)
If
impossible
is
implies
q
is
is
impossible
.
then
false
implies
false
These two
propositions are
more
"self-evident"
than the postulate, but
they express exactly the same relations.
(To eliminate parentheses, as far as possible, we make the convention

that the sign =, unless in parentheses, takes precedence over any other o and x that A that -1 and c take precedence over A + relation
;
and + take precedence over o and x c Thus

.
and that
-J
takes precedence over
and
is pq + -p -q -ipcq = (p c q)(p c r) pcqr

is
[(p q)
is
+ (-p
[pc.(q
-*
-</)]
(p
q)
r)]
[(p
c q)(p c
shall
r)]
However, where there

parentheses.)
a possibility of confusion,
we
put in the
The
1.
operations
by which theorems are to be derived from the postulates
are three:
Substitution.
If
Any
proposition
may
be substituted for p or q or
If
r,
etc.
is
a proposition, -p and ~p are propositions.
p and q are
propositions,
is
a proposition.
Also, of
any
is
pair of expressions related
by =
2.
either
may
be substituted for the other.

If
Inference.
is
asserted
and p
is
-J
asserted, then q
may
be
asserted.
cation,
3.
(Note that this operation

q.)
not assumed for material impli
pc
Production
It
p and
q are separately asserted,
may
be asserted.
These are the only operations made use
of in proof.
296
In order to
A
make
clearer the nature of the strict relations,
and particu
larly strict implication,
we
shall
correlates in terms of strict relations.
wish to derive from the postulates their This can be done by the use of
postulate
8 and
its
consequences, for by 1-8 a relation of two material
relations can be transformed into a relation of the corresponding strict

relations.
But
as a preliminary to exhibiting this analogy,
we must prove
number
of simple but
fundamental theorems.
These working principles
will constitute the
remainder of this section.

will
The
first
theorem
be proved in
full
and the proof explained.
The
conventions exemplified in this proof are used throughout.
2-1
pq-tp
1-6 [pq/p; qp/q; p/r}:
1-1 xl-2-i (p
1-6,
q*p)
q
is
This proof
may
be read:
for
r,
"Proposition
when p
substituted for p,
qp
(P q
is
for
*
q,
"
and p
states that propositions 1-1

of the proposition
and 1-2 together imply

line of proof
is
P)
The number
which states any

is
given at the beginning of the line.

"p
Next, in braces,
indicates that
indication of
any
substitutions to be made,
q/p"
in the proposition cited, for p;
"p+ q/r"
p q would indicate that

proposition
1
to be substituted,
p+
was to
be substituted for
r,
etc.
Suppose we take
6,
which
is
and make the substitutions indicated by {p

get
q/p;
qp/p; p/r}.
We
then
(pq-lqp)(qp-lp)
This
since
is
(pqlp).
But
the expression which follows the brace in the above proof.

q
-l
is
1-1,
and
-J
is
1-2,
we
write 1-1 x
2 instead of
q-iqp)(qplp). This calls attention to the fact that what precedes the main implication sign is the product of two previous propositions.
(p
Since 1-1 and 1-2 are separately asserted, their product
may
be asserted;
and
since this product
may
be asserted, what
it
implies
the theorem to be
is its
proved may extreme brevity.
be asserted.
The advantage of this way of writing the proofs Yet anyone who wishes to reconstruct the demon
stration finds here everything essential.
2-11
(p
= q)4(p4q)
2-1 {p-lq/p; q-ip/q}: (p *
1-06: (p
q) (q *
p) 4 (p
* q)
q)
(p
The System of
Strict Implication
297
2-12
(p
= (/HfeHp)
Similar proof, 1-2 instead of 2-1.
2-2
(p-i q)
-J
(~g
-J
~p)
1-06: 1-8
[(p4q)4(~q4~p)}[(~q*~P}*(P*<l)}
l)
2-1: (1) H Q.E.D.
In this last proof,

(1),
we
introduce further abbreviations of proof as follows:
and placed after a lemma which has been established, thereafter in the same proof we write (1), or (2), etc., instead of that lemma. line of proof instead Also, we shall frequently write "Q.E.D." in the last In the first line of this proof, the of repeating the theorem to be proved. substitutions which it is necessary to make in order to get
or (2), etc.,
is
1 -8
[(p
q)
* (~q
-*
~p)][(~<7
H ~p)
-J
(P
<
</)]
are not indicated because they are obvious.
And
in the
second
line, state
ment
of the required substitutions
is
omitted for the same reason.
Such
abbreviations will be used frequently in later proofs. Theorem 2-2 is one of the implications contained in postulate 1-S. of a By the definition, 1-6, any strict equivalence may be replaced by pair
strict implications.
By
either of these postulate 1-2 and theorem 2-1,
implications
may
2
be taken separately.
2-21
(~q*~p)4(p4q)
1
:
[(1) in
proof of 2
2]
-J
Q.E.D.
This
is
the other implication contained in postulate 1-8.

q)
2-3
(-P *
* (~q * P)
1-1
2 .2
{-q/p
-p/q}:
-q-p-*-p-q
-q/q}
:
{-g -p/p;
-p
(1)
-J
-*
[~(-/>
-<?)
~H
"rfl
1-02: (2)
= Q.E.D.
pp-lp
2-4
pip
1-2 {p/q}:
1-G: l-3x(l)-{ Q.E.D.
2-5
-(-p)*p
2--i
{-p/p}:
-p*-p
(1)
-I
2-3 {-p/q}:
Q.E.D.
2-51
-(-p}
=p
2-5x1-5 = Q.E.D.
1-06:
298 2-6
A
(-p
-i
-g)
-J
(g
-J
p)
-J
2-3 {-q/q}: (-p

2-51:
(1)
( fy
-j
-g)
-i
[-(-q) H p]
(1)
= Q.E.D.
-p)
2-61
(p
-j
-g)
-i
2-6 {-pip}:
2-51:
(1)
K-p) -gM = Q.E.D.

-*
(g
<
-P)
(1)
2-62
(p
q)
-I
(-g
-p)
-i
2-61 {-q/q}: [p
2-51:
(1)
-(-g)]
-J
(-g H -p)
(1)
= Q.E.D.
2-6J
(png) = H-t-p)
1-06:
2-62x2-6 = Q.E.D.
2-61x2-3 = Q.E.D.
and 2-62 are the four forms
is
2-64
(p*-q)
= (q*-p)
2-3, 2-6, 2-61,
of the familiar
1-06:
Theorems
terms.
principle that an implication
converted by changing the sign of both
2-7
(~p4~q}4(-p4-q)
2-21 [plq; q/p}: (~p -J ~g) -j (q * p ) 2-62 {p/q; q/p}: (q -J p) H (-p -t -q) 1-6: (1) x (2) H Q.E.D.
(1)
(2)
2-71
(-p
-J
-q)
-i
(~p
-i
-J
q)
2.6: (-p
-?) H (g
p)
-i
(1)
2-2 {^/p; p/gj:

1-6:
(g S p)
(~jM
~g)
(2)
(l)x(2)^Q.E.D.
(-p-t-q)
1-06: 2-71
= (~p*~q) x2-7 = Q.E.D.

:
2-7 {-p/p; -g/g}

2-51:
(1)
(~
-p
-J
-g)
-J
[-(-p)
-J
-(-g)]
(1)
= Q.E.D.
:
2-73
(p4q)-i(~-p-i~- q )
2-71 {-p/p -q/q} [-(-p) 2-51: (1) = Q.E.D.
,
-i
-(-g)] 4 (-
-p s -
-g)
(1)
2-7J1
(p-ig)
= (~-p^~-g) 1-06: 2-73x2-72 = Q.E.D.

:
2-74
(p^g)H(-^p^-^)
2-62 {-g/p; ~p/g}
1-6:
(~g ^ -p) H (-
^p
-i
^g)
(1)
2-2x(l)^ Q.E.D.
The System of
Strict Implication
299
2-75
(-
~p
-i
~g)
-i
(p H
ry)
2-6 {~p/p; ~q/q]
(-
~p
-J
~q)
-J
(~g
-I
~p)
(1)
1-6: (1) x2-21-jQ.E.D.
2-76
(p-Jg)
= (-T
=
(-
!--</)
1-06:
2-74x2-75 - Q.E.D.
-i
2-77
(p
-i
~p -i - ~g) - (~ -p ~ -g) = 2-76 x2-731 x2-G3 x2-712 = Q.E.D.

g)
"
(-g
-{
-p)
(~g
-I
~p)
"
"p
implies
g"
is
"
equivalent to
p
"
is
possible
q
is
implies
q
"
is is
possible
is
equivalent to
"q
is
necessary
implies
is
necessary
"
equivalent to
implies
is
false
implies
".
is
false
equivalent to
is
impossible
is
impossible
2-6-2-77 are various principles for transforming a strict implication. These are all summed up in 2 77. The importance of this theorem will be
illustrated shortly.
2-8
pq = qp
1-1
{g/p; pfq}: q
-I
(1)
1-06: 1-1 x(l)
- Q.E.D.
(1)
2-81
p =
pp
1-2 {plq}: pplp 1-06: 1-3 x(l) = Q.E.D.
2-9
p(qr) = q(p r) 1-4 {g/p; plq}: q(p

1-06: 1-4 x(l)
r)
-i
p(q
r)
(1)
= Q.E.D.
q)
q)
2-91
p(qr)
(p q)r
2-8: p(qr)
2-9: p(rq)
2-8: r(pq)
= = =
p(r
r(p
(p q)r
The above theorems

briefly
constitute a preliminary set, sufficient to give
most further
proofs.
II.
STRICT RELATIONS AND MATERIAL RELATIONS
We
proceed
now
to exhibit a certain analogy between strict relations
and material
necessities,
relations;
between truths and
falsities
on the one hand and

This analogy
possibilities,
and
impossibilities
it is
on the other.
runs
all
through the system:
exemplified by 2-77.
300
l-l
If
A
pq-iqp
p and
3-llpOq-iqOp
If
q are both true, then q

true.
p and
q are consistent,
then q
and p are both

1-2
q
If q
is
and p are
3-12
true,
consistent.
pip
and p are both
then p
qop*-~p
and p are
consistent, then
It q
it is
true.
possible that p be true.
1-3
If
p-tpp
p
is
3-13
then p and p are
-~p
is
pop
that p be true,
itself.
true,
If it is possible
both true.
1-4
then p
consistent with
(q r) 4 q
p(qr)4q(pr)
3-14
po
o (p
r)
The correspondence exhibited in we should note with care that while

p(q
r)
the last line seems incomplete.
But
q(p
"
r)
(p q)r
and any one

is not
"p,
of these
may
all
is
be read
p, q,
and
(q
".
are
all
true",
po(qor)
consistent
q,
and
are
q
is
with the proposition

day";
p o consistent with r
consistent",
or) means
"p
is
Let
p =
is is
"Today is
Friday".
Tues
"Today
Thursday";
"Tomorrow
Then
is
qor
q
is
true.
And
r
it
happens to be Tuesday, so p
true.
Since p and
(q
or
are both true in this case, they

"p,
must be
consistent:
p o
or)
true.
But
q,
and
are
all consistent" is false.

is
"Today is Tuesday" is
incon
sistent
with
"Today
Thursday"
and with
"Tomorrow
is
Friday".
o q o r. Then Suppose we represent q, and r are all consistent" by p as a fact, p o q o r will not be equivalent to p o (q or). Instead, we shall
"p,
have
p o
"p,
or = p o
(qr)
o (p
r)
=
"
(p q)
or
consistent with the
q,
and
are
q
all consistent" is
equivalent to
",
is
proposition
definitions:
and
are both true
etc.
We
may, then, add two new
3-01
3 02
-
pqr =
is
p(qr).
Def.
Def.
poqor =
po(qr).
3-02
typical of triadic, or polyadic, strict relations:
when parentheses
are introduced into them, the relation inside the parentheses degenerates
into the corresponding material relation.
In terms of the new notation of
3-01 and 3-02, the last
line of
-i
the above table would be

r
q r
poqorlqopor
clearly.
which exhibits the analogy more
The System of
Strict Implication
301
We
3-11
must now prove the theorems
in the right-hand
column
of the table
poq-iqop
2-74 Iqp/p
1-01: (1)
,
qp/q}:
-I
[-
~(P
q)
-I
~(q p}}
(1)
= Q.E.D.
(1)
3-12
qop-l-~p
2-74 (qp/p; p/q}: l-2*[- ~(qp) *- ~p]
1-01: (1)
= Q.E.D.
(1)
3-13
op
= Q.E.D.
1
2-74 {pp/p}: l-3-J[-~iM-~(Pp)]

1.01: (1)
3-14
p o(qr) -lqo(pr)
2-74 [p(qr)/p-, q(pr)/q}:
1.01: (1)
-4
-I
~[p(r/ r)]
-J
~[?(p
r)]
(1)
= Q.E.D.
But
what 1 4 is stated to imply should such cases, since no confusion will be oc
omit the brace.)
r)
(In the above proof, the whole of
be enclosed in a brace.
casioned thereby,
in
we
shall hereafter
3-15
po
(q r)
(pq) or
= qo
(p
2-76:
2-9H-~[p(<7r)]
~[q(P
r)]
(1)
2-76: 2-91-J-~[p(r/r)]
1-01: (2)x(l)
~[(p q)r]
(2)
= Q.E.D.
sum, p A
An
p+
q,
exactly similar analogy holds between the material logical sum,

strict logical
q.
and the
3-21
p + q-iq + p
least one of the two,
3-31
p*(j-lqAp
either
"At
p and
one
"Necessarily
p or
im</"
q,
is
true"
implies
"At
least
plies "Necessarily
either q or
p".
of the two, q
and
p,
is
true".
3-22
If
p*p
p
is
+ q
3-32 one of
If
- -p
-J
pA
q
true,
true, then at least

q, is
is
necessarily
then
the two, p and
true.
necessarily either
p or
is
true.
3-23
p+p+p
is
3-33
p*p4~-p
p
is
If at least
If necessarily either
true or
p, is true,
then p
r)
true.
r)
is
true, then
p
r)
is
-J
necessarily true.
?
3-24
p+(q +
is
4q+(p +
3-34
p A
(0
A (p +
r)
As
it
before, the analogy in the last line
really
complete.
And
the explanation
seems incomplete, and as before, is similar, p + (7 + r) and
302
+ (p +
(q
A
r) r)

least
"
both mean
"At
one of the three,
p, q,
and
r, is true".
But
pA
would not mean

it
One
of the three, p,
"One
true".
Instead,
would mean
of
q, and r, is of necessity the two propositions, p and
To distinguish p*(q + r) necessarily either q or r\ is necessarily true". from p A (q A r) is rather difficult, and an illustration just now, before we have discussed the case of implication, would probably confuse the reader.
be content to appeal to his intuition to confirm the fact that "Necessarily one of the three, p, q, and r, is true" is equivalent to "Neces and this last is sarily either p is true or one of the two, q and r, is true"
shall
We
pA
(q
r).
If
we chose
to
make
p+
+
definitions here, similar to 3-01
and 3-02,
they would be
and
= p+ = pA p^qAr
q
r
(q (q
+ +
r) r)
Proof of the theorems in the above table
is
as follows
3-21
p + q-iq + p
1-1
{-q/p; -p/q}:
-q-pl-p-q
(1)
(2)
2-62:
(l)*-(-p- q )i-(- q -p)
1-05: (2)
= Q.E.D.
1 -2 in
3 22
-
p-tp +
Similar proof, using
place of 1-1.
3 23
-
p + p-ip
1
3.
3 24
-
p+
(q
r)
-i
q + (p + r)
(1)
(2)
1-4 {-q/p; -p/q; -r/r}: -q(-p -r) -J -p(-q -r) 2-62: (1) i-[-p(-q- r )] -j -[- g (- p - r)
]
2-51:
1
-
(2) (3)
05
1-05: (4)
= -{-p-[-(- q -r)]} * -{-q -[-(-p - r )]} = p + -(-q -r) q + -(-p -r) = Q.E.D.
-J
(3)
(4)
3 25
-
p+ pA
(q
r)
(p + q) + r
q + (p +
r)
Similar proof, using 2-9 and 2-91, and 1-06.
3-31
-J
Ap
{-q/p; -p/q}:
1-1
-q-pi-p-q
(1) (2)
2-2: (l)4~(- p
1-04: (2)
-q)4~(- q -p)
= Q.E.D.
3 32
-
-p H p A
Similar proof, using 1-2 in place of 1-1.
The System of
3 33
Strict Implication
303
pAp
p A
-j
p
A (p +
Similar proof, using 1-3. 3 34

-
(q
r)
-J
?)
-J
1-4 {-q/p; -p/q; -r/r}: -q(-p -r)

2 2
-
-p(-q -r)
(1) (2)
(3) (4)
(1)
-I
~[-p(-q -r)}
-J
~[-q(-p -r)}
2-51: (2)
1
-
04
(3)
1-05: (4)
- H-p-hH-r)]} H-H+(-p-r)ll = p A -(- - r ) q A -(-p _ r) = Q.E.D.

fy
.,
3 35
-
pA
A (p + /) Similar proof, using 2-9 and 2-91, and 1-06.

(q
r)
(p + q)
ry
Again, an exactly similar analogy holds between material implication,
peg, and
3-41
If
is
strict implication,
q.
(pcq)-t(-qc-p)
2-62
g,
(p
2?
-I
q)
-J
(-q
-I
-p)
g,
then q
(
If
strictly implies
then
q
.
is
false
.
materially implies
is
false
strictly implies
is
false
false
3-42
If
-p-*(pcq)
p
is false,
3-52
If
~p-l(p-l0)
then p materially
q.
is
impossible (not self-con
implies
any
proposition,
sistent, absurd),
plies
then p
strictly
q.
im
any proposition,
3-43
If
(pc-p)-i-p
is false.
3-53
its
(p-*-p)4~p
p
strictly implies its
is
own
If
own nega
self-
negation, then p
tion, then p
impossible (not
consistent, absurd).
3-44
ccr]^[c(cr)]
3-54
[-t(cr)]4[*(pcr)}
The comparison
of the last line presents peculiarities similar to those
noted in previous tables. The significance of 3-54 is a matter which can be better discussed when we have derived other equivalents of p -J (q cr).
The matter will be taken up in detail further on. The theorems of this last table, like those in previous
by transforming the postulates
1-1, 1-2, 1-3,
tables, are got
and
1-4.
In consideration
of the importance of this comparison of the two kinds of implication, we add certain further theorems which are consequences of the above.
may
3-45
If
p-i(qcp)
p
is
3-55
If
~-p-l(</-p)
true, then every proposi-
is
necessarily true, then
is
tion, q, materially implies p.
strictly
implied
by any proposi
tion, q.
304
A
(-pep)
p
is
3-46
If
-ip
its
3-56
If
(-pipH~-p
p
is
materially implied by
strictly implied
by
its
own
own
3-47
denial, then
is
true.
denial, then
is
necessarily true.
-(pcq)-ip
q,
3-57
If
-(p-ig)-i-~p
If
p does not
q,
.
strictly
imply any
is
any proposition,
3-48
If
then p
is
true.
proposition,
then p
possible
(self-consistent)
-(pcq)4-q
is false. q,
3-58
If
q -(p * q) -* p does not strictly imply

is
g,
then q
then p
possibly false (not neces

.
sarily true)
Note that the main or asserted implication, which we have translated

is always a strict implication, in both columns. then 3-42 and 3-45-3-48 are among the best known of the "peculiar" the orems in the system of Material Implication. For this reason, their ana
"
If
",
logues in which the implication

first
is strict
deserve special attention.
Let us
note that ~ -p
-J
(-p
-J
p)
is
a special case of 3 55.
This and 3 56 give
us at once
-P
(-P
-J
p)
necessarily true
is
This defines the idea of

e.
g.,
"I
"necessity".
proposition
strictly
this,
am",
as conceived
by Descartes
is
one whose denial
implies
it.
Similarly,
p-l(-pcp)
a special case of 3-45.
And
with 3-46, gives
p = (-p c p)
true proposition
is
one which
is
materially implied
This point of comparison throws some light
by the two upon

is
its
own
or
denial.
relations.
The negative
~p
-i
of
a necessary proposition
impossible
absurd.
(p
-i
-p)
is
a special case of 3-52.
This, with 3-53, gives

-*
-P =
(P
-P)
And p
with
-i
-p
is
equivalent to -(pop).
sition is
one which strictly implies
Thus an impossible or absurd propo its own denial and is not consistent
itself.
Correspondingly, we get from 3-42 and 3-43
-P = (P c -P)
false proposition is
one which materially implies
its
own
negation.
It is
obvious that material implication, as exhibited in these theorems,
The System of
is
Strict Implication
305
not the relation usually intended by "implies", but it may be debated whether the corresponding properties of strict implication are altogether We shall revert to this question later. At least, these propo acceptable.
sitions serve to define
more sharply the nature

is
of the
two
relations.
Proof of the above theorems
as follows:
3-41
(pcq)
1-1
-i
(-r/c-p)
[p/q; -q/p}: -q
(1) 4 -(p -q) (2)
-J
p -q
(1)
2- 62:
-(-q p)
-(-p)}
(2)
(3)
2-51:
=-(p-q)*-(- q
1-03: (3)
= Q.E.D.
(1)
(2)
3 42
-p
4 (p
q)
1-2 {p/q; -q/p}: p -q 4 p

2-62:
(1) 4
-p
4 -(p -q)
1-03: (2)
- Q.E.D.
1
3 43
(p
c.
-p) 4 -p
-3.
3 44
p c
(q
r)
4 q
,
c (p c
r)
1-4 {q/p p/q-, -r/r}: q(p -r) 4 p(q -r) 2-62: (l)^-[p(g-r)]H-[g(p-r)]
2-51:
1
(1)
(2)
(3) (4)
(2)
(3) (4)
03
1-03:
= -{p-[-(g-r)]} 4 -{ g .[-(p-f)]} - 2J c -(ry -r) c -(/; -r) - Q.E.D.

-I
ry
:
3 45
p4(qc.p)
3-42 {-p/p; -q/q]
3.41: (_ 2;c
_
ry )
-(-/>)
-I
(-/>
c-f/)
(1)
H [-(- r/ )c -(-;;)]
(2)
2-51:
(2)
- (-pc-r/)-i(7Cp) = Q.ED.
(3)
(4)
1-6: (l)x(3)H-(-y;)-i(r/c^
2-51:
3 46
(4)
(-p c p) -j p 3-43 {-p/ 2;}: [-pc-(-p)H-C-p)
(1^
2-51:
(1)
= Q.E.D.
q) 4
3-47
~(pcq)4p
2-62 {-p/p; pcq/q}: 3-42 4 -(p c
2-51:
(1)
-(-p)
= Q.E.D.
q4(pcq)
(D
3-48
-(pcq)4-q
3-45 {^/p; p/^}:
2-62:
(1)
-i
Q.E.D.
21
306
A
~p
-J
3-52
(p
-J
g)
2-1
1. 8:
i-q/q}:
(1)
p-q*p
Q.E.D.
(1) (2)
=~p-i~(p-g)
1-02: (2)
3-53
(p-J-p)-J~p
3-54
[p-*(gcr)H[g-J(pcr)]
1-4 [q/p; p/q; -r/r}: q(p -r) * p(q -r)
1-8: (1) = ~[p(g-r)H~[g(p-r)] 2-51: (2) - ~{p-[-(g- r )]H~{<H-(p-r)]) 1 -02: (3) = [p S -r)] * [g H -(p -r)] 1-03: (4) = Q.E.D.
-(<?
(1)
(2) (3)
(4)
3 55
~ -p * (g -J p) 3-52 !-p/p; -g/g)

2-6: (-p
-l
~
p)
-p
-J
(-p
-^
-g)
(1)
-g)
-l
(g
-l
(2)
1-6: (1) x (2)
-i
Q.E.D.
3-56
(-p-lp)4~-p
3-53 {-p/p}: [-p-j-(-p)H~-:p
2-51:
(1)
(1)
=
-^
Q.E.D.
3-57
-(p-ig)-*-~p
2-62: 3-52
Q.E.D.
J-58
-(p-J g)
-J
g
-^
2-62: 3-55
Q.E.D.
between material relations and
The presence
strict relations in
of this extended analogy
the system enables us to present the total character of

1
the system with reference to the principles of transformation,

in brief
and
8,
and systematic form.

III.
This
will
be the topic of the next section.
THE TRANSFORMATION
{-/~<
We
-p
-}
have not, so
"If
far,
considered any consequences of postulate

is
false".
7,
-p,
is
impossible, then p
They
are rather obvious.
4-1
~-p*p
1-7 {-pip}
2-51:
(1)
.
--P-I-(-P)
(1)
= Q.E.D.
p
is
If
is
necessary, then
true.
4-12
p-i-~p
is
If
2-61 {~p/p; p/q}: 1- 7 * Q.E.D. true, then p is possible.
The System of
Strict
Implication
307
4-13
~-p-J-~p
1-6: 4-1
x4-12-iQ.E.D.
is
If
is
necessary, then p
possible.
4-14
pq-ipoq
4-12 lpq/p}:
1-01: (1)
pq*-~(pq) = Q.E.D.
(1)
4-15
(p*q)4(pcq)
1-7 (p-q/p}: ~(p-q)*~(p-q)
1-02: (1) 1-03: (2)
0)
(2)
= (pi q )-i-(p- q) = Q.E.D.

-(-p-gH-Hp-tf)
-I
</
4-16
p*q-ip +
1.Q4:
q
(1)
1-7 {-p-q/p}:
(1)
1-05: (2)
= pA -(-p -g) - Q.E.D.
(2)
4-17
-(poq)4-(pq)
2-62: 4-14
= Q.E.D.
any
strict implication
By
i.
virtue of theorem 4-15,
which
is
asserted
e., is
the main relation in the proposition
may
implication.
asserted,
And by
4 -1C, any
"
strict logical
be replaced by a material sum, A, which should be

.
The be reduced to the corresponding material relation, + It consistent with", o is a little different. relation case of the strict
may
follows from 4-17 that for every theorem in the
main
x
relation
in
.).
is
denied,
that
is,
-(.
is
.),
there
is
an exactly similar theorem

is,
which the main which
relation
It
is
that of the logical product, that
-(.
our immediate object to
show that
for every strict relation
is
assertable in the system, the corresponding material relation is also assertable. It is, then, important to know how these various relations are present
in the system.
-J
The only
== is
relations so far asserted in

-i
,
any proposition are

-J
and =
Since
expressible in terms of
we may take
it.
as the
fundamental relation and compare the others with

4-21
-J
= -p A q
1-02:
p*q =
~(p-g)
~[-(-p)~q]
C1)
2-51: (1)
1-04: (2)
=p*q =
= Q.E.D.
q
2)
4-22
pA
= -p
-J
4-21 {-p/p}: -p-i?

2-51:
(1)
=*
= -Hp) A?
Q.E.D.
308
For every postulate and theorem in which the asserted relation is -J there is a corresponding theorem in which the asserted relation is A and
,
,
vice versa.
Consider the analogous relations, c and + 4 23
pc
= -p +
1-03:
2-51:
pcq = -(p-q) = pcq = -[-(-p)-q] (1)

= Q.E.D.
:
0)
(2)
1-01: (2)
4-24
p+q
= -pcq
4-23 {-p/p}
2-51: (1)
-pcq =
-(-p) + q
(1)
= Q.E.D.
is
For every theorem in which the asserted relation sponding theorem in which the asserted relation is +
there
is
a corre
and
vice versa.
The exact parallelism between 4-21 and 4-23, 4-22 and orates what 4-16 tells us: that wherever the relation A is
corresponding material relation, +
,
4-24, corrob
asserted, the
may
be asserted.
4-25
p-iq
= -(po-q) 1-02: p*q =

1-01:
~(p-q)
-[--(?-?)]
(1)
2-51:
(l)=p*q = - Q.E.D. (2)
(2)
4 26
po
-(p -i -q) 4-25 {-q/q}:

q
2-51:- (1)
p*-q = -[po -(-?)] = p*-q = -(poq)
(1)
(2)
(3) (4)
2-11: 2-12:
2-3:
(2)-l(2M-0H-(pog) (2)4-(poq)*(p*-q)
(3)-ipo0-i-(2>-*-0)
-*
(5)
r/
2-61: (4)
-(p
-l
-ry)
po
(6)
1-06: (5) x(6)
= Q.E.D.
-*
there
For every postulate and theorem in which the relation is a corresponding theorem in which the main relation
is
is
asserted,
is
o but this
relation
is
is
denied:
and
for every possible
theorem
in
which the relation o
asserted, there will be a corresponding

-J
theorem
-i
in
which the main relation
but that relation
is
denied.
o and
are connected
q.
by negation.
An
1-03
exactly similar relation holds between p q and p c
pcq = -(p-q) 4-27 pq = -(pc-q)

Proof similar to that of 4-26, using 1-03 in place of
1 -02.
The System of
Strict Implication
309
The
parallelism here corroborates 4-17:
for every possible
theorem
in
which the relation
is
denied, there
is
a theorem in which the corre
sponding logical product is denied. But the implications of 4-1-4-17 are not reversible, and a theorem in which the relation o is asserted does not
give a theorem in which
way:
of
p, p,
any material relation and - ~p, the weakest is

truth- value of
is
asserted.
p and
is
[
it
To put it another cannot be further

po
q
reduced.
But the
any consistency
=
is
(p q).
The reduction
4-28
of
to the corresponding material relation,
=,
obvious.
q. Hypothesis: 2.11: Hyp. H (p
p =
To
-J
prove: p
q.
q)
(1)
(2) (3)
2-12: Hyp.
<
(<M
p)
4-15:
(IHCpcg)
==
4-15: (2)-i(?cp)
1-07: (3)x(4)
(p
(4)
q)
For every theorem in which the relation responding theorem in which the relation =
=
is
is
asserted, there
is
a cor
asserted.
have now shown at length that, confining attention to the main relations in theorems, there are two sets of strict relations which appear
in the system:
(1) relations (2)
We
=,
-i ,
and A which are
asserted,
and
relations
o which are denied;
relations
o which are asserted, and relations

relation of the first described
-i ,
and A which are denied.

it
Wherever a
set appears,
Any
is
relation of the second set

its
asserted
be replaced by the corresponding material relation. will be equivalent to some relation o which Such relations cannot be further truth- value will be [- ~].
may
reduced; they do not give a corresponding material relation. But under what circumstances will relations of this second sort appear? Examination
show that they can occur as the main relation in the orems only through some use of 1-7 and its consequences, for example, In other words, they can occur only ~p. pq-ipoq, p-l-~p, and - -p
of the postulates will
-J
where the corresponding material relation is already present in the system. Hence for every theorem in the system in which a relation of the type
p o
q is asserted, there is a
theorem
in
which the corresponding material
relation,
q, is
asserted.
is
in which the main relation Consequently, for every theorem in the system is the strict there is an exactly similar theorem in which the main relation
corresponding material relation.

relations appear as subordinate, or unasserted, These are reducible to the in theorems, the situation is quite similar.
strict relations
W herever
T
310

its
corresponding material relations through some use of 1-8 and quences. Note particularly theorem 2 77,
(P * q)
conse
(~
~P
-*
~?)
(~
-P
-*
~ -9)
(-?
-P)
(~?
-*
~P)
or [~ -]
rela
The
or
[
truth-value of any strict relation will,

].
by
its definition,
-J
be
[~]
And where two such
are connected
by
or
any equivalent
tion, they may be replaced by the corresponding relation whose truth-value and this is always a material relation. is simply positive or is [-]
We may now
illustrate this reduction of subordinate strict relations:

-J
43
[(p
q)
-J
(r
*)]
-*
[(p
g)
:
-I
(res)]
-i
2-7 {p -q/p;
1-02: (1)
r -s/q]
-J
[~(p -q) H -(r -s)}

-I
[-(p -g)
-I
-(r -$)]
(1) (2)
1-03: (2)
4- 31
[(p
= [(p q) = Q.E.D.
*
(r
-I
*)] *
[-(p -q)
-(r -s)]
A g)s
(r
A s)]
[(p
q)
-J
(r
*)]
Similar proof, (-p -q) in place of (p -q), etc.
4-32
[(poq)-l(ros)]*[(pq)-i(rs)] 2-75 {pq/p; rs/q}: [- ~(p

1-01: (1)
?)
-1
- ~(r
s)H
[(p ?)
* (r *)]
(1)
= Q.E.D.
*)]
4- 33
[(p
-i
q) H (r
-J
S [(p
g)
(r
c 5)]
(1)
4-15:
[(pcg)-l(rc5)H[(pcg)c(rc*)]
A*)]
-^
1-6: 4-3 x(l)-i Q.E.D.
4-34
[(p
q)
(r
[(p
g)
(r
s)]
Similar proof, using 4-31 in place of 4-3.
4-35
[(pog)H(ro*)H[(pg)c(r*)]
relation,
Note that as a subordinate
p o
1
q reduces directly.
In theorems 4 3-4 32, postulate
8 only has been used, and the reduc

is
tion of strict relations to material relations
incomplete.
In theorems
4 33-4 35, postulates 1 8 and 1 7 have both been used, and the reduction In these theorems, dyads of dyads are dealt with. The is complete.
reduction extends to dyads of dyads of dyads, and so on.
this
We may illustrate
by a
single
example which
q)
-J
is
typical.
-I
Hypothesis: [(p 4
(-p A
q)]
4 [(p o -q)
[(p -q)
-(p
-J
q)]
To prove:
[(p
q)
c (-p +
is
q)]
c -(p c
it
q)}
(The hypothesis
true,
though
:
has not been proved.)

-I
2-71 lp -q/p; p -q/q}
[-(p -q) H -(p -q)]
[~(p -g)
-I
-(p -q)]
(1)
1-02, 1-03, 1-04, 1-05, and 2-51:

(1)
[(p
q) H
(-p +
q)]
-I
[(p
-i
g)
4 (-p
g)]
(2)
The System of
Strict Implication
311
1-6: (2) xllyp. -I [(per/) -i (-p + r/)] -i [(p o-g) -t -(/; -J g)l 2-72 {p -g/p; p -g/g} [- -(? -g) -i - ~(p -r/)] -I [(p -g) (p
:
<
(3)
-r/)]
(4)
1-01, 1-02, 1-03,
and 2-51:
(4)
=
c
((p
1-6: (3) x (5)
-I
[(p
g)
o -g) -i -(p -i g)] S (-p + g)] -J [(p -r/)
[(p -r/)
-*
-t
-(p c
g)]
(5)
(6)
-(p c
g)]
4-33:
(6HQ.E.D.
In any theorem in which ~p is related to ~g, or p to ~g, or ~ -p to ~ -g, ~ may be replaced by -. This follows immediately from 2-77. We
illustrate briefly the reduction in those cases in
is
which
~r, or
r,
or
r,
related to
po
g,
or
-1
pA
g,
or
r)
:
-J
g.
4-36
(pog-J-~r)
1-01: (1)
(p g H
r/g}
2-75 {p g/p;
[-
-(p
g) H -r]
-J
(p q H
r)
(1) (2) (3)
= (poq*-~r)-l(pq4r)
4-15: (pg-ir) -* (pgcr) 1-6: (2) x (3) -i Q.E.D.
4-37
(pAg-i~r) ^(p + gc-r)

2-7 {-p -g/p; r/g} [-(-p -q) H -r] -J [-(-p -g) 4 15 [-(-p -q) -* -r] ^ [-(-p -g) c -r]
: . :
-J
-r]
(1) (2) (3)
1-6:
1-04 and 1-05:
(l)x(2HK-p-gH~rH[-(-p-g)c-r] = Q.E.D. (3)

e.
g.,
dyad
of triadic strict relations,
p o
(g r)
-J
o (p
r),
reduces
just like a
relation
of dyads, because a triadic strict relation is a dyadic strict with a dyadic material relation for one member. But a triad
dyad
of dyadic strict relations behaves quite differently.
Such
is
postulate
6,
This does not look
like a strict triad,
but
r)
it is,
being equivalent to
r)]
(p
-l
g)
-l
[(g
-J
c (p
-{
which obviously has the character
of strict triads generally.
The sub
direct use of ordinate relations in such a triad cannot be reduced by any be reduced. 1-8 and its consequences. However, all such relations can The method will be illustrated shortly by deriving
(pcg)(gcr) c(pcr)
from the above.
What
strict relations, then,
material relations?
cannot be reduced to the corresponding The case of asserted relations has already been dis-
312
cussed.
A
For subordinate

relations, the question
admits of a surprisingly simple answer. All the relations of the system can be expressed in terms of some product and the various truth values the truth values of ~ -p, p, Let us remind ourselves ~p, -p, and ~p.
:
po
--=
pi
pA
q q
= =
~(p ~(
q)
g)
~(-p -q)
of
pq = pcq = p+q =
-(p
q)
-(p -q)
-(-p -q)
of -p,
The
and
difference
between the truth-value
p and that
between ~p
p, p and -p, does not affect reduction, because p can be regarded as - -(p) or as ~(-p); as or as - ~(-p); p (p)
between
and p is also -(-p). Hence we may group the various types of expression which can appear in the system under three heads, according to truth-value
:
Hor[~-]
p-lq
or
[-]
[--] or [---]
p =
ptq p = q
p*q
-(poq)
p+q pq
-(pq)
poq
~(p*q)
-(p
-(p+q)
-(p
q)
=
p
q)
-(pcq)
~(p-lq)
-p
-p
p
~-p
-~p
In this table, the letters are quite indifferent: replacing either letter by other letter, or by a negative, or by any relation, throughout the table, any
gives a valid result.
The blank spaces
in the table could also be filled;
But, (p -q). as the example indicates, the missing expressions are more complex than any which are given, and possess little interest. The significance of the
table
is
for example, the first line in the third
column would be
this:
//,
in any theorem, two expressions which belong in the same
column
of this table are connected, then these expressions
may
be reduced by
its consequences. For, by 2-77, a relation of any two in the same column gives the corresponding relation of the corresponding two in either of the other columns. But any theorem which relates expressions
postulate 1-8 and
which belong
in different
columns of
is
this table
is
not thus reducible, since
any such difference of truth- value
ineradicable.
This table also sum-
The System of
Strict
Implication
313
marizes the consequences of postulate 1-7:

gives the expression on the
right.
same
line
with
it
any expression in the table and in the next column to the

left also
-J
It follows that expressions in the
column to the
give the
expressions on the same line in the column to the right, since
is
transitive.
Just as postulate
is
the only source of asserted strict relations

relations, so also
which are not replaceable by the corresponding material

sequences of 1-7.
the only theorems containing irreducible subordinate relations are con
For
this postulate
is
is
the only one in which different
truth-values are related, and

tute - for ~ in
the only assumed principle by which an
asserted (or denied) truth- value can be altered.

1
But
if
we simply
substi
7, it
every proposition
values,
[~],
[
in
becomes the truism, -p -* -p. As a consequence, for the system which contains strict relations or the truth],
],
or [-
],
in
any form,
in
such wise that these truth[-],
values cannot be reduced to the simple negative,
or the simple positive
(the truth- value of p), by the use 1-8, the theorem which results if we - for ~ in that Or, to proposition is a valid theorem. simply substitute
put
it
more
clearly,
if
less accurately;
it
if
any theorem involve
[~],
explicitly
or implicitly, in such wise that
cannot be reduced to [-] by the use of 1-8, ~ for - is valid. For example, 4 13, ~ -p ^ still the result of substituting ;;, ~ truthcannot be reduced by 1-8; ~p are irreducibly different -p and - for ~, we have -(-p) --(-;;), and hence -(-p) But values.
substituting
c-(-p), or
p^p- Propositions such as the c-(-p) may be called "pseudo-analogues".
pair
If
~-p4-~p
and -(-p)
so
we reduce completely,
far as possible, all the propositions which involve [~] or strict relations, by the use of 1-7 and 1-8 and their consequences, and then take the pseudo-
analogues of the remaining propositions, we shall find such pseudo-analogues redundant. They will all of them already be present as true analogues of This transformation by propositions which are completely reducible.
means
scheme
of postulates 1-7
and
1-8,
by which
strict relations give the cor
responding material relations,
may
q,
be represented by the substitution
pc
q,
-(p
q),
p+
q,
-p,
~l>*
p,
~(p)
P*q, p =
</,
-(p</)
PM
~-P* -(-~P)
o as a main rela put -(p o q) and -(p q), -(- ~p) and -(p), because p q when it is denied, and - ~p is reducible tion in theorems is reducible only
We
for triads of only through its negative. As we have now shown (except dyads, the reduction of which is still to be illustrated), propositions involving
314

still
expressions below the line are
valid
when the corresponding
expressions
above the
line are substituted.
The transformation by {-/~} of all the assumptions and theorems of system of Strict Implication which can be thus completely reduced, and
the the
rejection of remaining propositions which involve expressions below the line

(or the substitution for
them of
their pseudo-analogues), gives precisely the
system of Material Implication.

All the postulates
and theorems
of Material Implication
can be derived
the system of
from the postulates and

further illustrate this
definitions of Strict Implication:
Strict Implication contains the
fact
system of Material Implication. We may by deriving from previous propositions the
postulates and definitions of the calculus of elementary propositions as

it
appears in Principia Mathematical
pcq =
is
-p +
(Principia, ^1-01)
theorem 4-23.
(Principia, ^3-01)
(i) (2) (3)
4-41
pq = --p + -q
1-05 {-p/p -q/q}: -p + -q = -[-(-p)-(2-51: (1) =-p + -q = -(p q) 2-63: (2) = -(-p + -q) = -[-(p q)]
,
2-51:
(p
(3)
= Q.E.D.
(Principia, *4-01)
=
is
q)
= (pcq)(qcp)
the definition, 1-07.

(Principia,
(1)
4 42
-
p+p cp
3-23: p + p-lp
4-15:
4 43
-
(1)
-i
Q.E.D.
(Principia,
cp+
1-2 {-q/p; -p/q}: -p-q-i-q

2-61:
(1)
-J
CD
(2)
q ^ -(-p -q)
1-05: (2)
=
-i
q-lp + q
(3)
4-15:
(3)
Q.E.D.
(Principia,
4-44
p+
cq + p
3-21: p + q-iq + p 4-15: (!)H Q.E.D.
CD
(Principia,
4-45
p+
+ (p + 3-24: p + (q + r)
(q
r)
cq
r)
-J
q + (p +
r)
CD
4-15: (1)-^ Q.E.D.

6
Pp. 98-101, 114, 120.
The System of
Strict Implication
315
For the proof

certain
of the last postulate in the set in Principia Mathematica
lemmas
p
q
are needed which are of interest on their

(q
own
account.
4-51
= qc(pcr) 1-03 {pq/p-, r/q}: pqcr = -[(p q) -r] 2-91 and 2-9: (1) = pry cr = -[p(q -r)] = -[q(p -r)] 2-51: (2) = p0cr = -{p-[-( 9 -r)]} = -|<H-(p-r)]} 1-03: (3) = Q.E.D.
c
r
= pc
r)
(1) (2)
(3)
4-52
p q-lr = p-l (qcr) = ql (per)

1-02 {p q/pi r/q}
:
pq
-J
~[(p q) -r]
Remainder
4-53
of proof, similar to the above.
[(pcq)p]-lq 2-4 {pcq/p}: (p c 0) -1 (p c g) 4-52 {pcq/p; p/q-, q/r}: (I) = Q.E.D.

"
(1)
is
an immediate consequence of 4-53 that If p is asserted and p c q asserted, then q may be asserted", for, by our assumptions, if p is asserted
It
is
and p c
q is asserted,
then
[(p
q)p]
may be asserted.
"
And
is
if
this is asserted,
then by 4-53 and our operation of inference", note that the relation which validates the assertion of q
q can be asserted.
But
the relation s in
in the
the theorem. This principle, deduced from 4-53, is required of Material Implication (see Principia, #1-1 and #1-11).
system
4-54
(~pc~q) *(qcp)
4-3: 2-21-^Q.E.D.
4-55
(p-*q)l(prcqr) 1-6 {-r/r}: (p-lg)(^

4-52:
(1)
-r)
-i
(p l -r)
(2)
r)]
1-02 and 2-51:
(p-Jg)HKg-l-r)c(p-l-r)] = (p -I g) -* r) c ~(p (2)
Kg
(3)
(4)
4-54: -(gr)c-(pr)-l(2;rcgr) 1-6: (3) x (4) Q.E.D.

<
4-56
(pcq) c(prcgr)
4-55 {(pcq)p/p}: 4-53 H
2-91:
(1)
(2)
[(p
c?)p]r
r
eg
r
(2)
4-51:
= [(pc 9 )(pr)]c 9 = Q.E.D.
4-57
p cq = -qc-p
4-3:
2-62H(pCtfH(-?c-p)
(1) (2)
4-3: 2-6-l(-gc-p)-j(pC(/)
1-06: (l)x(2)
= Q.E.D.
316
4-58
(pcq)(qcr)c(pcr)
4-56 {-r/r}: (p c
4-57: (1)
1-03: (2)
q)
c (p -r c
q -r)
(1)
(2)
4-51: 4 58
is
(3)
- (pc</)c[-(g-r)c-(p-r)] = (pcq)c[(qcr)c(pcr)] = Q.E.D.
(3)
the analogue, in terms of material relations, of 1-6.
The method
the proof
by which we pass from 1-6 to 4-58

strict
illustrates the reduction of triads of

first line of
of
dyads 4-55. Here
in general.
1
This reduction begins in the

in the
-6
is
put
form
(p
*?)*[(?*
12
The
istic
345
r)
(p
* r)]
relation
numbered 4
is
already a material relation.
This
is
character
of strict triads.
Relations 3 and 5 are reduced together by some
consequence of 1-8, in a form in which the asserted relation is material. Then, as in 4 56, relations 1 and 2 are reduced together by the use of 4 53
as a premise.
clearer
if
This use of 4-53
is
quite puzzling at
"
first,
but
will q
is
become
we remember its consequence,

be
asserted".
If
is
asserted and
pc
asserted,
it,
then q
may
can
(q
This method, or some obvious modification of
applies to the reduction of
any
triad of strict
dyads which the system
gives.
We
4 59
-
now prove
r)
the last postulate for Material Implication.

(Principia, #1-6)
(1) (2)
c[(p + q) c(p + r)] 4-58 {-pip} (-pcq)(qcr) c (-per) 4-51: (1) = (gcr)c[(-pcg)c(-pcr)]
:
4-24:
(2)
- Q.E.D.
symbolic postulates for Material Impli cation, as the development of that system from them, in Principia Mathcsufficient set of
These are a
system of Strict Implication, those theorems which belong also to Material Implication are not necessarily derived from the above set of postulates. They can be so derived, but the transformation {-/~} produces them, more simply and directly, from their
matica, demonstrates.
However,
in the
analogues in terms of strict relations.
IV.
EXTENSIONS OF STRICT IMPLICATION.
THE CALCULUS OF CONSIST AND THE CALCULUS OF ORDINARY INFERENCE ENCIES

converted by negating both members,
From
the symmetrical character of postulate 1-8, and from the fact

is
i.
that postulate 1-7
e.,
-i- ~p,
The System of
it
Strict
Implication
is
317
follows that, since the transformation

{~/-},
is
j-/~)
possible,
an opposite
transformation,
possible.
And
since implications are reversed
by negating both members, those expressions which are transformed directly by j-/~) will be transformed through their negatives by |~/-), while those
expressions which are transformed through their negatives
by
{-/~j will
be transformed directly by }~/-J.
Hence we have
-(peg),
-(/;
g),
pq,
-(p +
This substitution scheme
page 312.
in the
The
be verified by reference to the table on transformation }-/~} represents the fact that expressions
may
column to the
left, in this table,
give expressions in the middle column:
middle column give ex {~/-J represents the fact that expressions in the {-/-} eliminated strict relations: pressions in the column to the right. As in the previous case, so here, a j~/-j eliminates material relations.
dyad
of dyadic relations, or a relation connected with p, -p, ~p, etc., can

its
be transformed by 1-8 and
consequences when and only when the
connected expressions appear in the same column of that table.

transformation J~/-j
is
Thus the
is
subject to the same
sort of limitation as
(-/-}.
The
transformation {~/-J, eliminating material relations, has already

in those tables in Section II, in
been illustrated
strict relations
which theorems
in
terms of
terms of
were compared with analogous propositions
in
material relations.
result
Theorems
in the right-hand
column
of those tables
from
those in the left-hand
column by the transformation {-/-).
3-52, 3 proofs of 3-11, 3-12, 3-13, 3-14, 3-31, 3-32, 3-33, 3-34, and 3-55 indicate the method of this transformation. Theorem 3-54 indi As we have noted, triadic strict relations are not cates a limitation of it. Consequently, in the case of expressible in terms of strict dyads alone. carried triadic relations, the transformation {-/-} cannot be completely
The
out.
for
This
is
an important limitation, since postulate

of proof,
is
(>,
which
It
is
necessary
any generality
of logic in
a triadic strict relation.
means that any
system
its
which there are no material relations cannot symbolize

Since strict relations are the relations of intension,
own
is
operations.
this
transformation
an important observation about calculuses of intension in general. The vertical line in the substitution scheme is to indicate that the considered to be complete when no is
{-/-}
arbitrarily
p and -p will be transformed when connected with a material relation which is transformed; when not
material relations remain in the expression,
318
so connected,
p and -p remain.
is
They
could be transformed in
instructive.
all cases,
but the result
needlessly complex
and not
The system, or partial-system, which

{~/-}
results
from the transformation

It
may
be called the Calculus of Consistencies.

:
can be generated
independently by the following assumptions
(4)
Let the primitive ideas be: p o q, (5) p = q.

Let the other
(1) propositions, p, q,
r,
etc., (2) -p, (3)
~p,
strict relations
be defined
II.
-J
-(p O -q)
:
For postulates assume

III.
poq-iqop
q
IV.
op
-J
p pop
V.
VI.
~p
-J
"
Assume the operations

but
is
of
Substitution
"
and
"Inference"
as before,
-J
in place of
"Production"
put the following: If p
-J
is
asserted and q
is
asserted, then p -J r may be asserted. By this principle, proof without the introduction into the postulates of triadic relations.
possible
intensions.
The system generated by these assumptions is purely a calculus of It is the same which would result from performing the trans
{~/-j
of
it,
formation
upon all the propositions of Strict Implication which and rejecting any which still contain expressions, other than p and -p, below the line. It contains, amongst others, all those theorems concerning strict relations (except the triadic ones) which were exhibited
admit
in Section II in
comparison with analogous propositions concerning material
relations.
More
inference,
interest attaches to another partial-system contained in Strict

If
Implication.
our aim be to create a workable calculus of deductive

q, q,
need to retain the relation of the logical product, p but material implication, p eg, and probably also the material sum, p +
shall
we
may
be rejected as not sufficiently useful to be worth complicating the
system with.
complications.
Inference.
The
ideas of possibility
and impossibility
also are unnecessary
Such a system
may
be called the Calculus of Ordinary

it.
The
following assumptions are sufficient for
The System of
Primitive Ideas:
(5)
(1)
Strict Implication
319
Propositions, p,
q, r, etc., (2)
-p, (3)
pH
q, (4)
g,
p -
Definitions:
A.
B.
C.
[D.
= -p g pOq = -(p*-q) = q) = (p-ig)(g-ip) (p

pA
g
-J
p+
-(-p-g)]
Optional.
Postulates:
E.
F.
(-p
-J
g)
<
(-g
-l
p)
-J
G.
p-pp
p(g
r)
-J
H.
I.
g(p
r)
H -(-p)
J.
(p-!g)(g-ir)-l(p-Jr)
K.
L.
-J
po
*
(p g
r s)
(p o g H
o 5)
system
is
is
All of these assumptions are propositions of the
of Strict Impli
cation.
A.
is
is
4-22, B.
is
4-26, C.
I. is
is
1-06,
and D.
1-05;
E.
is
2-3, F.
is
is
2-1, G.
1-3,
H.
is
1-4,
1-5, J. is
1-06, K.
4-26, and L.
an im
mediate consequence of 4 32 and 4 35. The Calculus of Ordinary Inference It consists of all is, then, contained in the system of Strict Implication.
those propositions of Strict Implication, which do not involve the relation But of material implication, peg [or the material logical sum, p + g].
where, in Strict Implication,
we have
-p,
we
shall have, in the Calculus of
-p will be replaced Ordinary Inference, -(pop) or p -J -p. Similarly by -(-p o -p) or -p -J p, and ~p by p o p or -(p -J -p). In other words, for p is impossible we shall have p is not self-consistent or p implies
{
its
own negation
shall
for
is
necessary
(
we
shall
have
;
the negation of p
is
not self-consistent
or
we
p
have p
is
and for p is possible the negation of p implies p self -consistent or p does not imply its own negation
The Calculus
g,
of
-J
pA
q,
and p
of Ordinary Inference contains the analogues, in terms which of all those theorems of Material Implication r/,
It does not contain the useless
are applicable to deductive inference.
and
doubtful theorems such as
"A
any
proposition",
and
"A
true proposition
is
implied by any
proposition".
As a working
320
A
logic,
it

it
system of symbolic
respect,
is
superior to Material Implication in this

q.
and
also in that
it
contains the useful relation of consistency, p o
On
the other hand,
avoids that complexity which
may
be considered an
objectionable feature of Strict Implication.
The system of Strict Implication admits of extension to prepositional functions by methods such as those exhibited in the last chapter. For the working out of this extension, several modifications of this method are desirable, but, for the sake of brevity, we shall adhere to the procedure
which
is
already familiar so far as possible.
In view of
this,
the outline
to be given here should be taken as indicative of the general

results
method and
and not as a theoretically adequate account.

is
Since, as
we have
and S#E,
demonstrated, the system of Material Implication

Implication,
It follows that,
contained in Strict
U<px
with suitable definitions of
the whole theory of prepositional functions, as previously developed,
may
be derived from Strict Implication. px is true in explicitly than before,

"
H<px
will here
be interpreted more
or
"
all
(actual) cases/
<px
is
true of
(actual)
every x which
case",
.exists
".
And
Znpx will
mean
"
tpx is
<px
true in
true".
some
or
"There
exists
at least one x for which
is
The novelty
~ -p, and the
of the calculus of prepositional functions, as derived

tion, will
from
Strict Implica
p,
come from the
presence, in that system, of ~p,

<px
We might expect that if is a prepositional function, would be a prepositional function. But such is not the case: ~(px is a ~$x For example, x is a man but not is impossible that proposition.
strict relations.
"It
mortal
"
is
a proposition although
it
contains a variable.
So
is "Nothing
1
can be both
and x
is
A not- A
and
".
not-.4",
It
is A which predicates the impossibility of would be an error to suppose that all the propositions
".i
which contain variables are such because they involve the idea of impossi bility, or necessity, but the most notable examples, the laws of mathe
matics, are propositions,
reason.
and not prepositional functions,
for precisely this

i.
When
stated in the accurate hypothetical form
e.,
as the
implications of certain assumptions Since ~ is a proposition,

<px
they are necessary truths.

<px,
<px,
and
If ipx
all
the strict relations of

\f/x
propositional functions will be propositions.

functions, then
and
are prepositional
<px
\f/x
is
the proposition - ~(
<px
\f/x)
<px
\[/x
is is
the proposition
~(-<px
x-^r);
<px
* \f/x
the proposition ~($x x-\I/x)
The System of
Strict
Implication
is
321
impossible, then
We
then
shall
have the law,

cases".
~<px4U-<px,
"If
<px
it
is
false in all
<px
Hence
also,
<px
-J
- - #c,
"If
<px
is
sometimes true,
is possible".
The
first of
these gives us one
most important con
sequence,
(
-i
<px
fa)
-1
II X (
<px
c fa)
false,
"If
it is
impossible that
<px
be true and fa
"If
<px
then in no (actual) case

\l/x,
is
<px
true but fa
false",
or
strictly
implies
then
<px
formally
This connects the novel theorems of this theory of prepo implies fa". sitional functions with the better known propositions which result from
the extension of Material Implication.
(
<px
Similarly
(
<px
we
shall
have
A fa)
<px
-j
nx
-j
+ fa)
and
If
S, (
x fa)
<px
o fa)
we
x
s
use
z(<pz)
to denote the class determined

<px
by
<pz,
that
is,
the class
of all
such that
is
true, then
we
derive the logic of classes from this
calculus of propositional functions, by the same general type of procedure If we let a = as that exhibited in Section III of Chapter IV.
(3
will be as follows: &(fa), the definitions of this calculus
On
"x
a)
(px n
is
member
of the class a,
determined by the function
<pz
means
"
<px
n is true".
(ttC/3)
U*(<pxcfa)
(a (a
/3)
=
(<px
=
II X (#C
fa)
j8)
= = -
s
or
^.T)
-a =
(a x (a +
1
==
-a = ^
or or
x(-<f>x),
-0
a)
]8)
|3)
x(<px
^.r),
(a x
(a +
= =
x[(x
a) x (x a) +
e /3)]
j8)
.T(^T +
-J
^),
j3)
^[(a; e
(a; 6 /3)]
(Sx
fa:)
=
a c
/3
-1
"All
is
the relation
It
is
members
by
of
a are also members of
0"
a relation
<px
of extension.
defined
or fa
false
is
true";
or
"There is
"-n x
(^c^r)
is
true",
==
U,(-<px
is false every (actual) case, either no (actual) case in which ?x is true and fa is the cor + fa) - II, x-**). a -I
"In
-(<p*
responding relation of intension:

false or
it is
defined
by
"Necessarily
either
<px
is
fa
or
"It
is
impossible that
<f>x
be true and fa
false",
22
322
that
A
is,
(vx-*$x)
"
(-<pxA\f/x)
=
~(<px
x-\f/x).
is,
<p,
interpreted
The
ft,
class-concept of a, that
be correctly contains or implies the

-J
ft
may
class-concept of
that
is,
^".
That
will
this should be true
may
not be at
once clear to the reader, but
it
become
so
if
he study the properties of
-J ft,
and
of
<?x
\f/x t
in this system.
(<px
Since
we have
have also
-*
\f/x)
-i
-J
Ux
#x c
\[/x)
we
shall
(a
ft) -J (
0)
If the class-concept of
a implies the class-concept of

of
(3.
ft,
then every
-i ,
member
of
of
will
be also a
member
c.
The
intensional relation,
implies the
"law"
extensional relation,
But
the reverse does not hold.

ft
The
ft
old
formal
logic,
that
if
is
contained in
in extension,
then
is
contained in a
extension
in intension,
and
is
vice versa,
is false.
The connection between
and intension
by no means
so simple as that.
This discrepancy between relations in extension and relations in inten

sion
is
particularly evident in cases where one of the classes in question

0, or the universe of discourse,
1.
is
the null-class,
As was pointed out

x,
in
Chapter IV, we
shall
have
x
for
e 1,
every
"individual",
and
have
-(x
0)
Also, for
any
class, a,
we
shall
ac
These
last
1,
and
ca
always true when significant,
<px,
two
will
hold because, since
fa: * fa: is
-(fa: H fa:)
always
false,
we
shall have, for
any function,
and
ILKfa:
"A
-J
fa:)
<px]
We
shall
have these because
false proposition materially implies

is
proposition",
sition."
and
since
any
true proposition
does not hold that
"A
materially implied
by any propo
But
it
"A
false proposition strictly implies

is
any
proposition",
or that
true proposition
strictly
implied by any
proposition",
we
shall not have

<px
-J
(fa:
-J
-{
far)
-i
<px
or
-(fa:
fa:)
And
consequently we shall not have
a
If
-*
1,
"
or
All
-J
a
of
is
a null class,
we
shall
have
members
7 are also members of
ft,
The System of
Strict
Implication
323
class-concept of y
whatever
class
may
be".
But we
0,
shall not have class
"The
implies the class concept of
whatever
may
be".
The
implications
of a class-concept are not affected
by the
fact that the class has no
members.
"The
The
classes
relation,
0,
is
material or extensional equivalence,
a and
consist of identical members";

"The
a =
ft
is
strict or intensional
0".
equivalence,
It is
class-concept of
is
equivalent to the class-concept of
obvious that
(a
0)
-I
(a
0)
but that
2/z0
reverse does not hold.
The
relation
between intensions and
extensions
is
unsymmetrical, not symmetrical as the medieval logicians

it.
would have
And, from the point of view of deduction, intension are more powerful than relations of extension.
relations of
a+0
and
none.
and a x
(3
are relations of extension

of
the familiar
"logical
sum"
"logical
product"
two
classes.
What about
the corresponding
there are
relations of intension?
This most important thing about them
Consider the equivalences,

(a A 0)
and
(px
<px
(a o
0)
= =
x(<px
A #r) =
o #r)
x[(x
e e
a)
(x (x
0)]
(3)}
x(<px
x[(x
a) o
i/a*
is
a proposition
the proposition
(<px
x^.r),
"It
is
possible that
it
and
\l/x
both be
it
true".
And
is
being a proposition, either

0,
is
true of
every x or
1
is
true of none.
<^.r
So that a o
so defined,
would be either
.r
or 0.
Similarly
or;
A
(3
\f/x
a proposition, either true of every

either at
all.
or false
of every
and a A
would be
or 0.
Consequently, a A
and
a o
(3
are not relations of a
and
The product and sum
of classes
are relations of extension, for which no analogous relations of intension

exist.
This
is
the clue to the failure of the continental successors of Leibniz.

is
They sought
it
a calculus of classes in intension: there

-J
no such calculus, unless

in
be confined to the relations a
and a =
to
7
(3.
Holland really came
sight of this fact

"multiplication"
when he pointed out

and
logical
Lambert the
difficulties of logical
"division".
The presentation
involve
of the calculus of propositional functions
and calculus
would
it
of classes here outlined,
and
of the similar calculus of relations,
many
subtle
and vexatious problems.

results
But we have thought
worth while to indicate the general
which are possible, without dis But there is one important problem which involves cussing the problems. the whole question of strict implication, material implication, and formal
7
See above,
p. 35.
324
implication, which
is

the meaning of
must be discussed
"implies".
This
the topic of the next section.
V.
THE MEANING OF
"IMPLIES"
It is impossible to escape the
assumption that there
is
some
definite
and
"proper"
meaning
of
"implies".
The word denotes that

any reference
it
relation
which
is
present
when we
"validly"
pass from one assertion, or set of

to additional
"evi
assertions, to another assertion, without

dence".
If
a system of symbolic logic

"implies"
is
to be applied to such valid infer
ence, the meaning of
which
figures in
must be such a
is
"proper"
meaning.
We
should not hastily assume that there
only one such
meaning, but we necessarily assert that there is at least one. This is no more than to say: there are certain ways of reasoning that are correct or
valid, as
opposed to certain other ways which are incorrect or invalid. Current pragmaticism in science, and the passing of "self-evident
in
axioms"
mathematics tend to confuse us about
this necessity.
Pure
mathematics is no longer concerned about the truth either of postulates or of theorems, and definitions are always arbitrary. Why, then, may not
symbolic logic have this same abstractness? What does it matter whether the meaning of "implies" which figures in such a system be "proper" or
not, so long as
logic
it is
entirely clear?
The answer
is
that a system of symbolic
may have
But
this kind of abstractness, as will be

it
demonstrated in the next
chapter.
cannot be a criterion of valid inference unless the meaning,

"implies"
or meanings, of
which
it
involves are
"proper".
There are
two methods by which a system of symbolic logic may be developed: the non-logistic method exemplified by the Boole-Schroder Algebra in Chapter
II,
or the logistic
method exhibited
in Principia
Mathematica and in the
development
is
of Strict Implication in this chapter.
The
non-logistic
method
takes ordinary logic for granted in order to state its proofs.
This logic which
taken for granted

it
is
either
"proper"
or the proofs are invalid.

it
And
if
the logic
takes for granted
is
not the logic
develops, then we have a
most curious situation. A symbolic logic, logistically developed i. e., without assuming ordinary logic to validate its proofs is peculiar among mathematical systems in that its postulates and theorems have a double
use.
They
are used not only as premises
deduced, but also as rules of
from which further theorems are inference by which the deductions are made.
A
it
system of geometry, for example, uses its postulates as premises only; gets its rules of inference from logic. Suppose a postulate of geometry
The System of
Strict
Implication
325
to be perfectly acceptable as an abstract mathematical assumption, but

false of
"our
space".
Then the theorems which

u
spring from this assump

still
tion
truly
may
be likewise false of
our
space".
But
the postulate
will
imply these theorems. However, if a postulate of symbolic logic, used as a rule of inference, be false, then not only will some of the theorems
be
false,
but some of the theorems
will
be invalidly inferred.
"
The
its
"
use of
the false postulate as a premise will introduce false theorems;

rule of inference will
use as a
produce invalid proofs.
Abstractness
in
mathe
matics has always meant neglecting any question of truth or falsity in postulates or theorems; the peculiar case of symbolic logic has thus far
been overlooked.
But we
are hardly ready to speak of a
"good"
abstract
mathematical system whose

it is
proofs are arbitrarily invalid.

"implies"
Until
we
are,
requisite that the meaning of

"proper"
in
any system
of symbolic
logic shall be a
one,
and that the theorems
used as rules of
inference
shall be true of this

"implies"
meaning.
"proper"
meaning, there is no criterion of is one or validity, no possibility even of arguing the question whether there
Unless
has some
not.
And
yet the question
What
is
the
"proper"
meaning
of
"implies"?
remains peculiarly difficult. It is difficult, first, because there is no common agreement which is sufficiently self-conscious to decide, for example, about
"material implication"
or
"strict
implication".
Even those who
feel quite
decided in the matter are easily confused by the subtleties of the problem. And, second, it is difficult because argument on the topic is necessarily
petitio
principii.
One must make the
Socratic presumption that one

of
"implies",
interlocutor already
knows the meaning
one
and needs only to that the meaning in denotation is clear to all, as the meaning of pose is clear, though the definition remains to be determined. or
s self,
"life"
be made aware
of that fact.
and agrees with One must sup

"cat"
If
two
should have different persons should really disagree about "implies" be nothing to hope for from their argument. there would "logical sense" involution of logical questions, the In consideration of this
peculiar
best procedure
is
to exhibit the alternatives in

"implies",
some
detail.
When
the
to
nature of each meaning of

be the
"proper"
and the consequences

rests.
of taking
it
one have been exhibited, the case
We have already drawn attention, both in this chapter and in Chapter II,
on material to the peculiar theorems which belong to all systems based r e may repeat here a few of them: implication.
(1)
any proposition; -p c
(p
q)
326
A
(2) (3)

is
A
If If If
true proposition
implied by any proposition; q c (p c
q)
p does not imply

p does not imply
q, q, q,
(4)
(5)
p does not imply

-(per/)
-(p cq) c-g then p implies that q is false;

is false;
then p then (7
is
true; -(p c
q)
cp
c(pc-q)
g,
(6)
(7)
If
If
p does not imply p and q are both p and

g are
then
is
false
implies g; -(p c
q)
(-29
g)
true,
then p implies q and g implies p;
pqc(pcq)(qcp)
(8)
If
both
false,
then p implies q and q implies p\
-p-qc(pcq)(qcp)
These
sufficiently characterize the relation of material implication. It is
obviously a relation between the truth-values of propositions, not between
any supposed content or logical import of propositions, materially means is false that p is true and q false". All these the implies orems, and an infinite number of others just as "peculiar" follow necessarily
"p
q"
"It
The one thing which this relation has in common with other meanings of implies "a most important thing of course is that if p is true and q is false, then p does not materially imply q. As has been said, there are any number of such "peculiar" theorems
from
this definition.
"
in
any calculus
of propositions based
on material implication. 8
to valid inference.
is
These the
orems do not admit of any application
In a system of
material implication, logistically developed, there

their being used as rules of inference, but
nothing to prohibit
when
so used they give theorems
which are even more peculiar and quite as useless. If we apply these theorems to non-symbolic propositions, we get startling results. "The moon is made of green cheese" implies because q c (p c q). + 2 ==
"2
4",
Let q be
the
"2
==
4"
and p be
is
"The
moon
is
made
of green
cheese".
Then, since
"If
"2
+
s
==
4"
true, its
filled
puppy
teeth are
s
consequence above is demonstrated. with zinc, tomorrow will be Sunday".

filled
Because the puppy
teeth are not
with
zinc, and,
anyway,
to be Saturday as I write.
A
9
false proposition implies any,
happens and a true

it
proposition
is
implied by any.
There
sitions,
8
are, then, in the system of Material Implication, a class of propo which do not admit of any application to valid inference. And
Every theorem gives others by substitution, as well as by being used as a rule of And there are ways whereby, for any such theorem, one other which is sure to be "peculiar" also can be derived from it. And also, it can be devised so that no result of one shall be the chosen result of any other. Hence the number is infinite. 9 Lewis Carroll wrote a Symbolic Logic. I shall never cease to regret that he had not
inference.
heard of material implication.
The System of
all
Strict Implication
327
the other, non-" peculiar theorems of Material Implication find their analogues in other systems. Hence the presence of these peculiar and use less theorems is a distinguishing mark of systems based upon material
",
implication.
There can be no doubt that the reason why the relation

implication
is
of material
s
the basis of every calculus of propositions except MacColl

is
and
Strict Implication
a historical one.
it
classes;
he then discovered that
Boole developed his algebra for could also be interpreted so as to cover

s
certain relations of propositions.
Peirce modified Boole
algebra by intro
.
a cb ducing the relation of inclusion, which we have symbolized by c all the properties of the relation between a and b when every member of has
is
also a
member
of b.
It
has one notable peculiarity:

class
if
is
a class which
then for any class x, a ex. Now the idea of "zero" in any branch of mathematics seems a little more of an arbitrary convention than the other numbers. The arithmetical fact that
has no members
"zero"
<-8
seems
"queer"
to children,
and
it
would, most likely, have seemed

defined, its
"queer"
"queer"
to an ancient Roman.
Once
is
properties,
as well as the obvious one, 8
+
=
is
8,
are inevitable.
is
"
It is similar
with
the
is is
"null class",
=
(a b
0,
"That
which
necessary.
And
a)
(a c6),
&
",
both a and nothing is nothing", That which is both a and 6,

"
is
equivalent to
If
All a
leads to the necessary consequence

"
a.
there are no sea serpents, then
All sea serpents are arthropods
necessarily follows.
because
it is
"sea
concept
This consequence seems more "queer" and arbitrary a relation of extension with no analogue in intension. The does not imply the concept "arthropod" as has
serpent"
been pointed out, -i a does not hold. And in our ordinary logical thinking we pass from intension to extension and vice versa without noting the
because the relations of the two are so generally analogous. But once we make the necessary distinction of relations in extension from relations in intension, it is clear that Oca in extension is a necessary conse
difference,
remarks apply to quence of the concept of the null-class. Entirely similar in everything" does not is contained the proposition a c 1, except that
"a
seem
so
"queer".
Boole suggested that the algebra of classes be reinterpreted as a calculus

of propositions
by
letting a,
6, c, etc.,
tions A, B, C, etc., are true.
represent the times when the proposi Then Peirce added the postulate which
holds for propositions but not for classes (or for propositional functions), \ proposition is either true in all cases, or true in none. a = ( a = i).
328
A
class of cases in

is
The
which any proposition
true
is
either
or
1.
This
If we add gives the characteristic property of the Two-Valued Algebra. to this the interpretation of a cb, "All cases in which A is true are cases in
which
is
true",
or loosely
of
"If
A, then
B",
we have the
source of the
peculiar
propositions
1
Material Implication.
For Ocfr. acl, OcO,

6.
1,
and
follow from the laws which are thus extended from classes
to propositions.
false proposition
1].
[=
0]
implies
And
a implies any
true proposition [=
the other.
Of any two
false
Of any two true propositions [= 1] each implies false propositions [= 0] each implies the other.
"A
false proposition any true one. means precisely there are no materially implies any proposition" cases in which A is true (if a = 0) then all cases in which A is true are
And any
proposition implies
"If
also cases in
which
is
true".
It does not
mean
is
"
any
false proposition".
"A
true proposition
materially implied by any
proposition"
means
only,
"If
is
true
[=
1],
then the cases in which

cases) in
is
true are contained

It does not
among
the cases
(i. e.,
all
which
is
true".
mean
"Any
true proposition can be inferred from

logical import, intension,
A".
Inference
depends upon meaning,

of extension.
Is
a c
b,
b is a relation
purely
this material
implication, a c
a relation which can
validly represent the logical nexus of proof
and demonstration?
Choose any true. Cer
It
Formal implication IL x means "For every value
(<pxc
\f/x) is
defined in terms of material implication.
of x,
z,
<px
materially implies
\f/x".
value of
x, say for convenience
and unless
(pz is false,
fa
is
tainly this relation approximates

"implies".
more
closely to the usual
meaning
of
But the
is
precisely
accurate
interpretation
of
x".
of
Hx
(<pxc\f/x)
depends upon what
meant by the
H. x
"values
them
the
as
"cases"
or
"individuals".
It
(<px
have spoken of makes a distinct difference whether

are
all
We
"cases"
comprehended by
c\f/x)
the possible cases,
all
conceivable individuals, or only all actual cases, all individuals

(in
which
exist
the universe of discourse).
Either interpretation
may
consistently be
chosen, but the consequences of the choice are important. briefly the more significant considerations on this point.
Let us survey
In the first place, supposing that the second choice is made and n x be taken to signify "for all x s which exist", what shall we mean by "exist"? This is entirely a matter of convenience, and logicians are by no means at
one in their use of the term.

it
to temporal
and physical
reality or to
But any meaning what
of
is
"exist"
which confines
"the
sometimes called
factual" is
inconvenient because, for example,
we may wish
to distinguish
The System of
Strict Implication
329
the status of curves without tangents in mathematics from the status of
the square of the

the former
circle.
This distinction
is
usually
made by
saying that
"curve"
"exist",
since the general mathematical idea of a
admits such cases, and their equations

the circle
is
demonstrably impossible.
be given; Again, it is inconvenient to say that
may
while the square of
Apollo exists in Greek mythology, whereas the god Agni does not. Now the god Agni is not inconceivable in Greek mythology; we find no record
of him, that
is all.
Similarly, while the usual illustrations of mathematical

is still
"non-existence"
are impossibilities, there
a difference between what
"does not exist" in a mathematical system and what is impossible. Sup we have an "existence postulate" in a set which are consistent and pose
independent each of the others the 0-postulate in the Boole-Schroder Without this postulate, the remainder of the set Algebra, for example.
generate a system in which
does not
exist.
But
it is
possible, as the
con
The frequent sistency of this postulate with the others demonstrates. existence" is the same as "possibility" is a statement that "mathematical
very thoughtless one. The most convenient use of
"exist"
in logic
is,
then, one which
makes
the meaning depend upon the universe of discourse, but one which does and the "existent" not, as is sometimes supposed, thereby identify the
possible.
("Possible"
similarly varies its

it is
meaning with the universe

"exist"
of
discourse.)
On
the other hand,

is
inconvenient to use
".
so widely
that
"existence"
synonym
for "conceivability
This
is
so obvious
it
in the
most
frequent universe of discourse,

in this sense, in
"phenomena",
that
hardly
needs to be pointed out.
Using
"exists"
which
"existence"
is
narrower than
but may, in some universes of discourse, be wider than "the "possibility" whether 11* in U t (^xcfa) denotes only factual", it makes a difference
existent x
s
or
all
possible z
<px
s.
All
American
silver coins silver coin
dated 1915 have

1915",
milled edges.
let
Let
be
"x
is
an American
There
is
dated
and
fa be
"
x has a milled
edge".
no necessity about milled edges

universe of discourse,
II X
for silver coins, unless one speak in the

this illustration,
false
if
if it
"legal"
U x (<pxcfa)
all
will
s.
be true
if
denote only actual
.r
s;
denote
possible x
One
s,
illustration
Il x (<px
<px
c fa)
refer to all possible x

false".
(<px
is as good as a hundred; is impossible c fa) means

"It
that
be true and fa
If
U^xcfa)
"The
be confined to actual x
s,
then
is
it
of extension, signifies a relation

is
class of things of
is
true".
which
<px
true
contained in the class of things of which fa
330
It
might be thought that the meaning of H x ((px c\f/x) is sufficiently determined by saying that the "values of in a function, px, are all the entities for which is either true or false. But this is not the case, for
a;"
<px
there
is
question whether, of an x which does not exist,
<px
is
always true,
or always false, or sometimes true true or false.
and sometimes
is,
false, or never either
Here again, the question

it is
in part,
one of convention.
if
<px
From
the point of view of extension,

all
obvious that
can be predi
cated at
of
<p,
an x which does not

of
exist, it will
x,
something,
an
"individual",
always be false. (Predicating which does not exist, should be dis

a,
tinguished from asserting that an empty class,
which
"The
exists
has no members,
bald"
is
included in some other, a c
is
of the latter.)
an example of the former; "All And the point of view of extension

In this sense,
"x
King sea serpents have green

is
6.
of
though France
it
is
wings",
frequently that of
common
brother,
sense.
is
man" is
false of
my non-existent
twin brother
twin
and even
identical propositions such as
"My
is
my
of
twin
brother"
are false of the non-existent.

is
intension,
it
an identical proposition
But from the point of view always true, and may be true
<px
or
may
be false of a non-existent
x.
If
the point of view of extension be

<px
taken with reference to prepositional functions, then

significant or false of the non-existent,
is
either not-
and
\j/x
is
similarly not-significant
or false.
c\f/x)
If
<px
and
\f/x
are not significant of the non-existent, then

x,
<px
Ux
If
is
(<px
means
"For
every existent
materially implies
$x".
<px
and
\l/x
are significant
and
false of the non-existent,
then
<px
\f/x
true
of every non-existent x, since of
two
false propositions,
each materially
implies the other.

for all possible x s
Hence on
and
will still
this interpretation,
Ux
(<px
c\l/x) is significant
<px
Hence
x,
<px
c \j/x. every existent x is such that be accurately rendered by "For every existent If the point of view of intension be taken materially implies $x".
true in case
its
meaning
with reference to propositional functions, or
if it
be
left
open, then
$x",
Hx
or
(<px
cfa) may mean

(px
"For
every possible
confine
10
x,
<px
materially implies
"
we
may, by convention,
materially implies
still
ê".
it
to the
meaning
For every
existent x,
10 We would gladly have spared the reader these details, but we dared not. If logicians do not consider one another s views, who will? In this connection we are reminded of a passage in Lewis Carroll s Symbolic Logic (pp. 163-64) anent the controversy concerning
the existential import of propositions:

"The
writers,
and
editors, of the Logical text-books
which run in the ordinary
hope inoffensive) title The Logi cians take, on this subject, what seems to me to be a more humble position than is at all necessary. They speak of the Copula of a Proposition with bated breath"
grooves
I shall hereafter refer
to
whom
by the
(I
The System of
Strict Implication
331
to the simpler
The ground being now somewhat

quences of taking
cleared,
we return
considerations which are really more important.
What
are the conse
H x ((px
is
c\f/x)
in
is
and most important
this.
It
one or the other meaning? The first a desideratum that we should be able to a c
derive the calculus of classes from the calculus of prepositional functions.
And
(<pz)
in this calculus of classes, the inclusion relation of classes
or
z(\j/z)
can be defined by
z(<pz)
c z(tz) =
(<px
I/M-)
by some equivalent definition. If then a c 0, or which exist,

or
<pxc\}/x",
ll z ((px c\J/x)
mean
"For
all
&(<pz)
cz(\f/z), so defined,
is
the use
of
ful relation of extension,
"All
the existing things which are
members
are also
"All
members
of
/3".
American
silver
Such a relation can represent such propositions as coins dated 1915 have milled edges". If, on the other
x
(<px
c $x) to mean "For all possible z s, <pxc\l/x, then two courses are open: (1) we can maintain that whatever is true of thus abrogating a useful and all existent things is true of all possible
hand, we interpret
probably indispensable logical distinction; or

is
(2)
its
we can allow that what

nature as conceived or
is
true or false of the possible depends upon

If
defined.
we make
c
z(\j/z),
the second choice here, the consequence

defined
z(<pz)
that
a c
|S,
or
z(<pz)
by
c z(tz) =
as
nx
c
(<t>x
t/a-)
or in
any equivalent fashion, such

o:C/3
= Hi(xe acxe
)8)
is
the relation
|S".
of
concept of
a implies the classThis relation does not symbolize such propositions as

intension
"The
class-concept of
what it almost as if it were a living, conscious Entity, capable of declaring for itself that we, poor human creatures, had nothing to do but to ascertain chose to mean and what was its sovereign will and pleasure, and submit to it. In opposition to this view, I maintain that any writer of a book is fully authorised If I find to use. in attaching any meaning he likes to any word or phrase he intends the at the beginning of his book, Let it be understood that by an author saying, word "black" I shall always mean "white", and that by the word "white" however injudicious I may think always mean "black", I meekly accept his ruling, whether a Proposition is or is not to be "And so, with regard to the question writer may understood as asserting the existence of its Subject, I maintain that every it is consistent with itself and with tl adopt his own rule, provided of course that
"
accepted facts of Logic. various views that may logically b "Let us consider, one by one, the I shall hold myse thus settle which of them may conveniently be held; after which to hold." free to declare which of them / intend
332
A
American
"All
silver coins
March",
dated 1915 have milled
edges"
or
"It
rained
every week in
or in general, the frequent universal propositions
which predicate
this relation of extension.

Hi(<f>x
And
whichever interpretation of
c\f/x)
be chosen,
we can now
"
point out one interesting peculiarity of it. We quote from Principia Mathematica: n "In the usual instances of implication, such as Socrates
is
man
"
implies
Socrates
is
a mortal
"IL
",
we have
"
a proposition of the
true.
form
<px
\j/x"
in a case in
which
x (<px
\f/x)
is
In such a case,
".
we
feel
the implication as a particular case of a formal implication

"
It
might be added that

is not
Socrates
it is
is
man
implies
Socrates
is
a mortal
a formal implication:
a material implication and a strict impli

object:
"But
cation, but not formal.
One may
as a fact, in such cases

,
there
is
a tacit premise of the type
All
".
men
are mortal
and
this
is
precisely
this
implication,^^ c \j/x) premise, and still the implication is men are mortal and Socrates is a
the formal
"All
Granted, of course.
strict
man"
But add
and material, but not formal.

does not formally imply
of
"implies"
"Soc
rates
is mortal".
12
If
the
"proper"
meaning
is
one in which
or one
"Socrates is
man"
in
which
"All
and truly implies men are mortal and Socrates

really
is mortal",
"Socrates is mortal",
is
man"
really
and truly
implies
"Socrates
meaning. However much support such an inference, it cannot
then formal implication is not that proper any formal implication may lie behind and
state
it.
One
to
oj
further consideration
exist,
is
worthy
of note: If IL x
(<px
c
"
\f/x)
"
be restricted
is is
which
then
1915
it will
can
silver coin of
denote such propositions as implies x has a milled edge

";
x x
an Ameri
a
Monday
of last
March
implies #
".
is
a rainy
day
";
"x
has horns and divided hoofs
implies x chews a cud

"contingent",
In other words
it
will
denote relations which are
and due to
"coincidence".
It
may
be doubted whether
this question the
such relations are
"properly"
implications.
But upon
reader will very likely find himself in doubt.
What we
regard as the
reason for this doubt will be pointed out later.
The
11
strict implication,
-i
q,
means
"
It is
impossible that p be true
1, 12
p. 21.
It
may
We render the symbolism of this passage in our own notation. be objected that the calculus gives the formal implication
which
not
<?
is the formal implication of \J/z c \f/x) x which is required. But this is by H x what is required. The variables for all values of which this proposition is asserted are and not x and z. The reader will grasp the point if he specify and ^ here, and then allow them to vary in his illustration.
(<px
<pz,
<A>
<p
The System of
Strict Implication
333
Similarly $x
or
"the
and
of
false",
"It
or
"p
is
inconsistent with the denial of

<?x
-t
q".
fa
means
<px
is
impossible that
be true and fa
fa".
false",
assertion
"im
is
inconsistent with the denial of

or
"inconsistent"
Some
explanation of
possible"
may seem
called for here.
These terms can
either of
them be explained by the other, but one or the other must be taken Yet the following observations may be of assistance: An for granted. assemblage or set of propositions may be such that all of them can be true
They are mutually compatible, compossible, consistent. There may be more than one such set. Whoever denies this on metaphysical grounds must assume the burden of proof. And whether, in fact, the
at once.
possible
and the
actual, the consistent
and the concurrently
true-in-fact
are identical, at least one

differs
must admit that our
concept of the possible
from our concept of the actual: that we mean by "consistent" some thing different from "concurrently true-in-fact". Any set of mutually
consistent propositions
"case"
may
be said to define a
"possible
situation"
or
or
"state
of
affairs".
And
a proposition
may
be
"true"
of
more
than one such possible situation Whoever understands "possible

sistent
propositions",
situation"
belong to more than one such set. situation" thereby understands "con
may
and
vice versa.
also
q
And whoever understands

"inconsistent
"im
possible
understands
translate
propositions".
In
these terms,
true
we can
-J
by
"Any
situation in
which p should be
s
and
q false
is impossible".
But
"times
"situation"
as here used
true".
should not be confused with Boole

is
when
is
proposition, once true,

"situations"
always
true.
A
is
proposition
may
be true of some possible
and
false of others,
but
it
must
be in point of fact either simply true or simply false.
This
what
nor
constitutes the distinction between a proposition

"x
and a prepositional
function such as
false.
13
is
man".
This
last
is,
in point of fact, neither true
Of special interest are the cases than two propositions are involved.
triadic relations take the
itself
of strict implication in
which more
We
form
of strict
have already seen that strict is dyads, one member of which

expect,
(q
a non-strict or material dyad.

instead
or (p q) o
Where we might
r.
-i
(q-*r),
we have
p o
(q r)
13
p
r.
4 (q
r)
or p q 4
Instead of p o
r)
we have
the reason
We may now
it is
discover the reason for this
be true of some impossible to deny that a proposition may one is prepared to maintain that whatever assertion And whoever can be referred to different possible "circumstances", is not a proposition. one true proposition, the asserts this must, to be consistent, recognize that there is only whole of the truth, the assertion of all-fact, the Hegelian Idee.
On
the other hand,

,
"situations
and
false of others unless
334
not in mathematical-wise but in terms of
common
sense about inference.

set.
Suppose that p,
cannot
all
q,
and the negation
of
r,
form an inconsistent
They
this
be true of any possible situation.
We
have symbolized
by
-(poqo-r).
- (p oqo -r)
If p, q,
-lp o (q -r)]
= p
So
-J
-(q -r)
= p-i(qcr) = p
-*
and -r form an inconsistent

r
set and, in point of fact
true,
then
q)
must be true
is strict.
also.
much
is
quite clear.
p and q are both The inference

all
from (p
to r
But suppose
p, q,
and -r cannot
be true in
is
any
possible situation
to be true.
We
(in the actual situation) p only can then conclude that q is true, r is true".
"If
and suppose
known
-(poqo-r) = p-l(qcr)
This inference
"If
. . .
is
.
also strict, but our symbolic equivalents tell us that this

.
then
."is
not
is
itself
strict
implication;
is
it
it
is
qcr, a
the
material implication.
other?
That
is
the puzzle;
If p, q,
why
not
all
strict like
The answer
simple.
if
and -r cannot
be true in any
it
possible situation
q
and
is
true of the actual situation,

is,
follows that
it
and -r are not both true
of the actual situation, that

in
-(q-r), but
does not follow that q
and -r cannot both be true

false)
it
some
other possible
situation (in which
p should be
does not follow that q

it
and -r are
does not follow that qlr, Consequently that q strictly implies r. we begin with an a priori truth (holding If, then, for all possible situations), that p, q, and -r form an inconsistent set, and
to this
inconsistent, that -(qo-r).
add the (empirical) premise

"If
"p
is
true",
we
get, as a strict
con
sequence, the proposition
is
true, r
is
true".
But the truth
of this
If,
consequence
this case,
is
confined to the actual situation, like the premise p.

infer r
in
we go on and The
it
from
q,
our inference
may
be said to be valid
because the additional premise, p, required to

granted.
describe
as
inference depends on pq-ir. an inference based on material implication, which
make it strict, is taken for Or we may, if we prefer,

is
valid
because
it is
confined to the actual situation.
Much
of our reasoning is
of this type.
We
state, or
have
explicitly in
which are required to give the conclusion
mind, only some of the premises We have omitted or strictly.
forgotten the others, because they are true and are taken for granted. In this sense, much of our reasoning may be said to make use of material or formal implications. This is probably the source of our doubts whether
such propositions as x has a milled edge

l
"x
is
an American
"x
silver coin
dated 1915
implies
",
and
has horns and divided hoofs
implies
The System of
Strict
Implication
called implications.
335
In
chews a cud
"
represent
what
are
"properly"
such cases, the reasoning is valid only if the missing premises, which would render the implication strict, are capable of being supplied.
The
case
w here two premises

r
are strictly required for inference
is
typi
r,
cal of all those
are required for
which require more than one. a conclusion, s, we have
Where
three, p,
q,
and
p(qr)*s = p*(qrc s) = p
[q
(r
c *)]
And
similar equations hold
where
four, five, etc., premises are required

is strict.
for a conclusion.
strict implication
Only the main implication
In other words, a
may
be complex but
is
always dyadic.
Another
x
significant property of
strict implication, as
is
implication or formal implication,

implies implies
is
that
if
we have
"
"x
opposed to material is a man strictly

is
a mortal
is
",
we have
",
likewise
Socrates
man
strictly
Socrates
a mortal
and
vice versa.
Propositions strictly imply
each other when and only when any corresponding prepositional functions According to this view of implication, similarly imply one another. is not simply felt to Socrates is a man implies Socrates is a mortal
"
"
be the kind of relation upon which most inference depends: it is the rela Strict implication is the tion upon which all inference does depend.
symbolic representative of an inference which holds equally well whether its terms are propositions or prepositional functions. One further item concerning the properties of strict implication has to
do with the analogues of the "peculiar" propositions of Material Impli These analogues are themselves somewhat peculiar: cation.
3-52
3 55
.
~p
-*
(p 4 q)
(q
.
If
p
If
is
impossible, then
is
p implies any p
is
proposition,
q.
~-pi
-j
p)
necessarily true, then
implied by any
proposition, q
These two are the
critical
members
in this class of propositions:
the re
mainder follow from them and are of similar import. In the "proper" sense of "implies", does an absurd, not-self-consistent proposition imply answer is contained in the anything and everything? A part of the
observation that
"necessary"
and
"impossible"
in
every-day use are

is
"impossible"
commonly
proposition
hyperbolical and no index.
Xo
proposition
in the sense of
is
~p except such
as imply their
own
contradiction;
and no
"necessary"
in the sense of -
contradictory.
indication of
Again, the implications of
what would be true
if
-p unless its negation is selfan absurd proposition are no that absurd proposition were true.
336
It

it is
is
the nature of an absurd proposition that

it
not logically conceivable
that
should be true under any possible circumstances.

that, in the ordinary sense of
"implies",
we can demonstrate
the
And, finally, an impossible

is
proposition implies anything and everything.

"proper"
It will
be granted that in
true"
sense of
it
."implies",
(1)
"p
and
if
q are both
implies
q
"q
true".
And
r,
will be granted that

r,
(2)
is g,
two premises p and

false.
imply a
conclusion,
and that conclusion,
false,
while one of the premises,
say p,
is
true,
liars"
then the other premise,
must be
while
That
Blank
is,
if
"All
men
is
are
and
"John
Blank
is
is
man"
together imply
"John
"John
Blank
man"
liar",
but
"John
Blank
liar"
is false,
is
is
true,
then the other premise,
"All
men
are
liars",
must be
false.
And
it
be granted that (3) If the two propositions, p and q, together imply r, and r implies s, then p and q together imply s. These three principles being granted, it follows that if q implies r, the impossible proposition q is true
will
"
but
false"
implies anything
"p
and everything.
true"
For by
r.
(1)
and
(3), if q
"p
implies
r,
then
true"
and
q are both
r,
"q
implies
r is
But by
(2), if
"p
and
are both
implies
r,
is
true but
false"
implies
is
false".
Hence
if
q implies
then
"
q is true
but
r is
false"
implies the negation of
any proposition,
is
p.
And
since
p
is
itself
may
is
be negative, this impossible

implies "Tomorrow
is
proposition implies anything.

Tuesday".
"
"Today
Monday"
Hence
"
"Today
Monday and
is
the
moon
"
not
is
made
of
green cheese
implies
is
Tomorrow
or
Tuesday
"It
".
Hence
Today
but tomorrow
not
Tuesday"
"The
implies
is
false that the
Monday moon is not
made
of green
cheese",
moon
is
made
of green
cheese".
be taken as an example of the fact that an absurd proposition It should be noted that the principles of the implies any proposition.
This
may
demonstration are quite independent of anything we have assumed about strict implication, though they accord with our assumptions.
We
and a
shall
now demonstrate:
first,
that there are a considerable class of
propositions which imply their

class of propositions
own
contradiction and are thus impossible,
thus necessary;
proposition,
will
which are implied by their own denial and are and second, that an impossible proposition implies any
and a necessary proposition is implied by any. These proofs be similarly free from any necessary appeal to symbolism, making use
proposition which should witness to the falsity of a law of logic,
of
only of indubitable principles of ordinary logic.
Any
or of
any branch
"
absurd.
p implies
p"
mathematics, implies its own contradiction and is is a law of logic; and may be used as an example.
The System of
In general, any implication,
"p
Strict Implication
337
"p
implies
is
</,"
shown
false
p"
is
true
and
q is
false".
Thus the law
"p
implies
"p
by the fact that would be disproved
by the discovery
This
its
is,
of
any proposition p such that

"p
is
true
and p
is
is
false".
then, an impossible proposition,
is
true
and p
false"
implies
true".
own
"p
negation, which
is
is
"At
least one of the two, not-p and p,

"p
is
For
"At
true
and p
"
is
false"
implies
q, is
is
true".
And
"p
is
true"
implies
least
q.
".
true".
"
And
not-?;,
"p
is
false,"
may
be this
is
Hence
"
p
"
is
true
true
implies
At
"
least
true
Hence
is
is
and p
is
false
one of the two, p and not-p, implies At least one of the two,
"
p and not-p,
least
true".
one of the two,
The negation of and not-ry, is true".

"p
"r/
is
true
and
r is
false"
is
"At
If
we have
as the negation of
is
true".
is
true
it is
two, p and not-p,
And
and p is this which
p here replace both q and r, false", "At least one of the

"p
is
true and p
is
false"
has been shown to imply.
Merely
for purposes of comparison,
we resume
this proof in the
symbols
of Strict Implication:
pH
p,
and p
~(p -p).
-J
Hence
(p -p)
-J
is
an impossible proposition.
By the principle p q p, we have p -p p. And by the principle q4p + q, we have p (-p + p). r), this gives (1) r) H (p By the principle (p
-1 -J -J -J
</)(</
(2)
p -P
-*
("P
+ P)
-I
But (-p +
This
is
p)
-(p -p).
Hence p -p
-(p -p).
any
which might be carried out in only one illustration of a process and number of cases. Take any one of the laws of Strict Implication
it
transform
into a
form which has the
prefix, ~.
For example,
The
in the example [(p q) p)], impossible proposition thus discovered, The reader will easily see can always be shown to imply its own negation. be done. Such illustrations are quite generally too complex how this
-(</
may
of symbolic abbreviation, but only to be followed through without the aid for the proofs. the principles of ordinary logic are necessary a Wherever we find an impossible proposition, we find necessary propo
sition, its negation.
is
true"
For example,
"At
least
one of the two, p and not-p,
is
a necessary proposition.
its
We
have just demonstrated that

of late to
implied by
own
denial.
(Some
23
logicians
have been inclined
deny the ex
338
necessary propositions and of impossible, or self-contradictory, propositions.
We
beg their attention to the above, and request their criticisms.)
We
shall
now prove
that every impossible proposition

its
i.
e.,
every
negation implies anything and every WTe have If p implies not-p, then p implies any proposition, q. thing. is true and r is false" implies any already shown that if q implies r, then Hence if p implies not-p, is true but not-p is false", proposition.
proposition which implies
"q
own
"p
that
is,
"p
is
true
"p
equivalent to
is q.
and p is true", implies any proposition, q. But p is true and p is true". Hence if p implies not-p, p implies
i.
any proposition,
Any
implies
necessary proposition,
its
e.,
any proposition,
q,
whose
r.
denial, not-g,
own
negation,
is
implied by any proposition,

q,
This follows
"
from the above by the principle that if p implies is false". In the theorem just proved,
"p
then
"q
is
false
implies
"If
implies
any proposition, then have not-g implies

"If
g",
let
g,
p be
"not-r
"not-g",
p implies not-p, then p and q be We

"not-r".
then not-g implies any proposition,

is
false"
not-r".
And
i.
if
not-g implies not-r, then

g.
implies
"not-g is
false",
e., r
implies
"a
Hence
if
not-g implies
g,
then any proposition,
r,
implies
g.
But
man
convinced against his
will is of the
same opinion
is:
still".
In what honest-to-goodness sense are the
"necessary"
principles of logic
and mathematics implied by any proposition?

sense of presuppositions.
The answer
that?
In the
And what,
precisely,
is
Any
principle,
A,
may be said to be presupposed by a proposition, B, if in case A were false, B must be false. If a necessary principle were false, anything to which
it is
at
all
relevant would be false, because the denial of such a principle,

itself.
being an impossible proposition, implies the principle

principle and
is
And where
its
negative are both operative in a system, anything which
proved
is
liable to disproof.
Imagine a system
in
which there are con
tradictory principles of proof.
That the chaotic
results
which would ensue
are not, in fact, valid, requires as presuppositions the truth of the necessary laws of the system. These laws those strictly "necessary" are always
logical in their significance.
very pretty
affair
The logic of "presupposition" is, in fact, a we have no more than suggested its character here.
of
The time-honored
ists
principles
rationalism
are
thoroughly sound and

the historic rational
capable of the most rigid demonstration, however
much
have stretched them to cover what they did not cover, and otherwise misused them.
In this respect, then, in which the laws of Strict Implication seemed
The System of
possibly not in accord with the
"
Strict
Implication
"
339
"
proper
sense of
implies
",
we have demon
strated that they are, in fact, required by obviously sound logical principles,
though
It
in
ways which
it is
easy to overlook.
may
be urged that every demonstration we have given shows not
only that impossible propositions imply anything and necessary proposi tions are implied by anything, but also that a false proposition implies
anything, and a true proposition
that an impossible proposition
is
implied by anything.
of course,
The answer
is
is false,
and a necessary proposition
is true. anyone think that this validates the doubtful theorems of Material Implication, it is incumbent upon him to show that some proposi
But
if
tion that
is
false but not impossible implies anything

is
and everything, and

implied by
all
that some proposition which

sitions.
true but not necessary
is
propo
And
shall
"
this
cannot be done.
We
for the
not further prolong a tedious discussion by any special plea as against material implication propriety" of strict implication
and formal implication. Anyone who has read through so much technical and uninteresting matter has demonstrated his right and his ability to
draw
his
own
conclusions.
CHAPTER
METHOD
I.
VI
SYMBOLIC LOGIC, LOGISTIC, AND MATHEMATICAL
GENERAL CHARACTER OF THE LOGISTIC METHOD. VIEW
THE
"
ORTHODOX"
The method of any science depends primarily upon two factors, the medium in which it is expressed and the type of operations by which it
is
developed.
"Logistic"
may
be taken to denote any development of
scientific
matter which
is
expressed exclusively in ideographic language and
uses predominantly (in the ideal case, exclusively) the operations of sym bolic logic. Though this definition would not explicitly include certain
cases of
later to
what would undoubtedly be called logistic", and we shall wish present an alternative view, it seems best to take this as our point
"
of departure.
geometry" differs from Euclid most fundamentally by the modern geometry no step of proof requires any principle except It was the extra-logical principles of proof in the principles of logic. Euclidean geometry and other branches of mathematics which Kant
"Modern
fact that in
noted and attributed to the

source of
"synthetic
"pure
intuition"
of space (and time) as the
judgments a
priori"
in science.
The character
cannot take
of
space (or of time), as apprehended
a priori, carries the proof over places

logic
"analysis"
where the more general principles of

Certain operations of
it.
thought are, thus, accepted as valid in geometry
because geometry
is
valid for spatial entities,
thought about space, and these transformations are though they might not be valid for other things.
The
principal impetus to the
modern method
in
geometry came from the
non-Euclidean systems which must necessarily proceed, to some extent, without the aid of such space intuitions, a priori or otherwise.
discovery of
And
1
the perfection of the
modern method
is
attained
when geometry
is
entirely freed
from dependence upon figures or constructions or any appeal
In the opinion of most students, Euclid himself sought to give his proofs the rigorous character which those of modern geometry have, and the difference of the two systems is in degree of attainment of this ideal. But Euclid s successors introduced methods which
still
further depended
upon
intuition.
340
Symbolic Logic, Logistic, and Mathematical Method

to the perceptual character of space.
341
geometry is thus freed from this appeal to intuition or perception, the methods of proof are simply those which are independent of the nature of the subject matter of the science
that
is,
2 the methods of logic, which are valid for any subject matter.
When
Coincidently with this alteration of method comes another changegeometry is now abstract. If nothing in the proofs depends upon the fact
that the terms denote certain spatial entities, then, whatever may be meant by "point", "plane", "triangle", "parallel", etc., if the assumptions be
true, then the
theorems
will be.
Or
in
any sense
can be asserted, in that same sense

asserted.
all
which the assumptions the consequences of them can be

in
The student may carry
in his
mind any image
of
"triangle"
More or "parallels" which is consistent with the propositions about them. than this, even the geometrical relations asserted to hold between "points",
"lines",
etc.,
may
be given any denotation which
is
consistent with the
In general, this means for relations, that any meaning may be assigned which is consistent with the type of the relation transitive or intransitive, symmetrical or unsymmetrical, one-one or e.
properties assigned to them.
g.,
one-many,
etc.
and with the distributions
of such relations in the system.

in arithmetic, or
"alge
Essentially the
bra".
same evolution has taken place
Any
reference to the empirical character of tally
marks or
collections
-of
c,
of pebbles has
become unnecessary and naive.

"A
The
K,
"indefinables"
arithmetic are specified, very likely, as

etc.,
class,
of elements, a,
6,
and a
relation (or
operation
+".
Definitions have
come
to have
is
the character of what

to say, they
Kant
called "transcendental definitions "that
to comprehend those properties which differentiate the entity which distinguish it for sense be defined by its logical relations, not those The real numbers, for instance, no longer denote the possible perception.
lengths of a
line,
but are the class of
all
the
"cuts"
that can be
made
(logi-
the widespread notion 2 1 cannot pass over this topic without a word of protest against of Kant s Transcendental that the development of modern geometry demonstrates the falsity notion that such "synthetic" Aesthdik. It does indeed demonstrate the falsity of Kant s that But, in general, it is accurate to say are indispensable to mathematics. principles with the source of our certainty about the world of nature, is concerned
Kant
s account in his day. Nothing is mo not with the methods of abstract science which did not exist the abstractness of modern geometry comes about through definitely obvious than that the certainty of its applicability renouncing the thing which Kant valued in geometry When geometry becomes abstract, the content of the science of space spin to our space. of space, which is con into two distinct subjects: (1) geometry, and (2) the metaphysics This second subject has been much discui of geometry. cerned with the application in the skeptical or "pragmatic since the development of modern geometry, usually and to me it seems a fact that Kant s basic argu But it is
(vide Poincare).
possible
with qualifications, capable of being rehabilitated as arguments concerning i. e. as a metaphysics of space. certainty of our knowledge of the phenomenal world,
t)
ments
are,
342
cally specified) in a dense,

rationals.
denumerable
series, of
the type of the series of
Thus abstractness and the

matics.
rigorously deductive
in the
method
of
development
have more and more prevailed
most
careful presentations of
mathe
When
these are completely achieved,

less
a mathematical system becomes
nothing more nor
than a complex logical structure. 3
Consider any two
mathematical systems which have been given this ideal mathematical form. They will not be distinguished by the entities which form their "subject
matter",
for the
terms of neither system have any fixed denotation.
And
not be distinguished by the operations by means of which they they are developed, for the operations will, in both cases, be simply those of
will
logical demonstration.
word
of caution
upon the meaning

all
of
"operation"
is
here necessary.
peculiarly mathematical operations that a system comes to have the rigorously deductive form. For the
It is exactly
by the elimination of
grocer
who
represents his putting of one sack of sugar with another sack
by 25
25
50,
[+]
is
a symbol of operation.
For the child who learns
the multiplication table as a means to the manipulation of figures, [X]

represents an operation, but in
arithmetic, in
any rigorously deductive development

and Multiplication
in
of
Dedekind
Was
sind und was sollen die Zahlen, or Hunting-
ton
"Fundamental
Laws
of Addition
Elementary
Algebra",
[+] and [X] are simply
done, performed.
The only things performed
are the logical operations
operation something an abstract deductive system variables are not added or multiplied. But,
in
still
relations.
An
is
unfortunately, such relations as [+] and [X] are likely to be

of as
"operations".
spoken
Hence the caution.
Since abstract mathematical systems do not differ

of their terms,
by any
fixed
meaning
and
since they are not distinguished through their operations,
they will be different from one another only with respect to the relations
of their terms,
and probably
also in certain relations of a higher
order-
relations
differ,
of
relations.
And
the relations, being likewise abstract, will
from system to system, only in type and in distribution in the systems; that is, any two systems will differ only as types of logical order. 3 M. Fieri, writing of Geometrie envisagee comme systeme purement logique", tiens pour assure que cette science, dans ces parties les plus elevees comme says:
"La
"Je
dans les plus modestes, va en s affirment et en consolidant de plus en plus comme I etude d un certain ordre de relations logiques; en s affranchissant peu a peu des liens qui 1 attachent a 1 intuition, et en revetant par suite la forme et les qualites d un science ideale purement
deductive
et abstraite,
comme V Arithmetique"
(Bibliotheque du congres Internationale de
Philosophic, in, 368.)
343
The connection between

is,
abstract or
"pure"
mathematics and
logistic
thus, a close one.
But the two cannot be simply identified.

is
For the
its
logical operations
by which the mathematical system

not themselves be expressed
there are symbols for
in
generated from
assumptions
may
ideographic symbols.
Ordinarily they are not:

"triangle"
"four"
and
"congruent",
and
. .
"plus",
but
.
the operations of proof are expressed

. .
by
"If
Only when the logical ideographic symbols do we have logistic. operations also are expressed In other words, all rigorously deductive mathematics gets its principles of from symbolic operation from logic; logistic gets its principles of operation
. . .
then
.",
"Either
or
.,"
etc.
in
logic.
Thus
logistic, or
the logistic development of mathematics,
is
name
for abstract mathematics the logical operations of whose development
are represented in the ideographic symbols of symbolic logic. Certain extensions of symbolic logic, as we have reviewed
for the satisfactory expression of these larly certain further
it,
are needed
mathematical operations
particu
developments
"descriptions"
of the logic of relations,

in
of
what
are called
necessary additions in
no wise
affect
logistic
and the theory But these Principia Mathematica. what has been said of the relation
development
of
between symbolic
II.
logic
and the
mathematics.
S
Two
The
VARIETIES OF LOGISTIC METHOD:
PEANO
Fornmlaire AND
Principia Mathematica.
logistic
THE
NATURE OF LOGISTIC PROOF And

gives, in
method
is,
then, a universal method, applicable to

it
body of exact knowledge. the a most precise and compact development, displaying clearly matics, are order which characterizes the system. However, there of
sufficiently coordinated
any mathe
type
logical
certain variations of the logistic method,
and systems so developed may which have nothing directly to do differ widely from one another in ways difference with the type and distribution of relations. One most important is carried out. has to do with the degree to which the analysis of terms
"Number,"
may
may, an important item
be taken simply as a primitive idea, or it notions be defined in terms of more fundamental notions. And these is carried, is The length to which such analysis in turn, be defined.
for example,
may
Correldetermining the character of the system. as primitive, or they and [X] may be taken atively, relations such as [+] which generate may be defined. And, finally, the fundamental propositions
in
the system
sis
be simply assumed as postulates, or they may, by the analy derived from those of a more elementary discipline. just mentioned, be
may
344

"terms"
In general, the analysis of
and
of relations
and the derivation

use
of
this
of
fundamental
propositions
go together.
And
the
analytic
method
requires, to
some extent at
least,
a hierarchy of subjects, with
symbolic logic as the foundation of the whole.
To illustrate these possible differences between logistic systems, it will be well to compare two notable developments of mathematics: Formulaire
de Mathematiques 4 of
Peano and
his collaborators,
and Principia Mathe-
matica of Whitehead and Russell.

in
the respects
just mentioned.
These two are by no means opposites Principia Mathematica represents the
farthest reach of the analytic
method, having no postulates and no primitive

mathematical branches. 5
ideas save those of the logic, while the Formulaire exhibits a partially hier
archic, partially independent, relation of various
for arithmetic,
For example, in the Formulaire, the following primitive ideas are assumed which immediately succeeds "mathematical logic".
signifies
number
zero
.
and
a
is
the
common name
of 0,
1, 2, etc.
signifies
+
The
signifies
6
.
plus
If
is
a number, a
indicates
the
number suc
ceeding a
primitive propositions, or postulates, are as follows:
1-0
1-1
No
Cls
OeN
acNo.D.a
e
1-2
1
+ eNo
3 s
Cls
e
1-4
a, 6
NO
Oes:aes.D,,.a + $SD.NoS a -f- =6 + .D.a = 6
l-5aeN .3.a + -=0

The symbol D here
in"
contained represents ambiguously "implies" or the relation c of the Boole-Schroder Algebra. This and the idea of
"is
a class,
"Cls",
and the
e-relation, are defined

logic".
and
their properties
demon
strated in the "mathematical

sitions
4
In terms of these, the above propo
may
be read:
be to the
All our references will
national language, Interlingua,
and
fifth edition, which is written in the proposed inter entitled Formulario Mathemalico, Editio v (Tomo v de
Formulario complete). 5 The independence of various branches in the Formulaire is somewhat greater than a superficial examination reveals. Not only are there primitive propositions for arithmetic and geometry, but many propositions are assumed as "definitions" which define in that discursive fashion in which postulates define, and which might as well be called postulates. and X, to be quoted shortly. Observe, for example, the definitions of
6
7
Section n,
Ibid.
1,
p. 27.

1-0
1-1
345
No
is
is
a class, or
number
is
common name.
is
a number.
is
1-2 1-3
If
a number, then the successor of a

a class, and
implies
if
a number.
if,
If s is s
is
contained in
s,
and
for every a,
,
is is
contained in contained in
(1-3
is
the successor of a is contained in s

is
then No
(every
number
member
if
of the class s).

induction".)
=*
the principle of
mathematical
1-4
of
b,
If
a and b are numbers, and
the successor of a
the successor
then a
If
=
is
b.
1-5
a number, then the successor of a

are then defined in the obvious
4= 0.
The numbers
3
is
= 2+,
etc.
The
relation
+, which
9
differs
= 1 +, -f, 2 way: 1 = from the primitive idea, a +
then defined by the assumptions: = a 3 1 a e No . 3 . a +

a
is
(If
a number, then a -f
a,
a.)
3-2
(If
fceNo.D.o
(6
+) =
(a
b)
+
the successor of
a and b are numbers, then a

6
.)
the successor of 6
The
X is defined by: 10 = 1-0 a, 6, ceN .3.a X 1 -01 a, b, c No s a X (b + 1) =

relation
e
-
(a
6)
It will
be clear that, except for the expression of logical relations, such

,
as
and D
in ideographic symbols, these postulates
and
definitions are of
the same general type as any set of postulates for abstract arithmetic. of members a, b, c, etc., is assumed, and the idea of a +, "suc A class,
cessor of
a".
The substantive
"successor
11
notions,
"number"
and
"zero",
the de
scriptive function,
of,"
the relations
but are taken as simple notions. have by virtue of being members of a
and X, are not analysed However, the properties which numbers

class,
are not taken for granted, as
would necessarily be done

set
in a non-logistic treatise
"mathematical
is
they are specifically
forth in propositions of the
logic"
which precedes.
And
the other principles by which proof
accomplished are similarly

to which this explicit-
demonstrated.
Of the
specific differences of
method
ness of the logic leads,

8
9
we
shall
speak shortly.
See
ibid., p. 29.
Ibid.
10
"
See
ibid.,
2, p. 32.
Peano does not suppose them
si
nos pote defini significa ~ ~ composite per signos noto

,
si 1
.
He says (p. 27): to be unanalyzdble. nos pote scribe aequalitate de forma,
-,
quod non
est
facile".
= expressione (This was written after the
but before Principia publication of Russell s Principles of Mathematics,
Mathem
346
In Principia Mathematica, there are no separate assumptions of arith

metic, except definitions which express equivalences of notation
and make
possible the substitution of a single symbol for a complex of symbols. There are no postulates, except those of the logic, in the whole work. In
other words,
all
the properties of numbers, of sums, products, powers,
etc.,
are here proved to be
what the
relations
what they are, solely on account of what number is, and X are, etc. Postulates of arithmetic can be
dispensed with because the ideas of arithmetic are thoroughly analysed.
The
lengths to which such analysis
must go
is
in order to derive all the proper
ties of
number
solely
from definitions
naturally considerable.
We should
be quite unable, within reasonable space, to give a satisfactory account of the entities of arithmetic in this manner. In fact, the latter half of Volume
I
and the
first
half of
Volume
II of Principia Mathematica
may
be said to
do nothing but just

with patience,
it is
this.
However, we may, as an
"cardinal number".
illustration, follow
will
out
the analysis of the idea of
This
be tedious but,
highly instructive.
We
shall first collect the definitions
the definition of cardinal
which are involved, beginning with number and proceeding backward to the definition
of the entities in terms of
which cardinal number
is
defined,
and then
to
the entities in terms of which these are defined, and so on. 12
*100-02
NC - D
Xc.
Df
"the
"Cardinal number" is
domain
of (the rela
tion)
Xc".
*33-01
"D"
D = aR[a =
.r{(32/l
R y].
Df
is
the relation of (a class) a to (a relation) R,

is
when a and R
are such
that a
(the class) x
is
which has the relation

.T
That
is,
"D"
the relation of a class of
S,
R to (something or other) y. each of which has the relation
R
"
to something or other, to that relation
itself.
*30-01
Ry =
"the
(ix)(xRy).
Df
y"
means
x which has the relation
to
?/".
Putting together this definition of the use of the

definition of
"D",
we
is
see that
"D
Xc"
is
"the
and the symbol x which has the relation

of
to
12
Nc",
and
this x
a class a such that every
member
a has the
rela-
The place of any

translations"
definition quoted, in Principia,
The
"
of these definitions are necessarily

"translation"
and, of course, any

precedes.
is indicated by the reference number. ambiguous and sometimes inaccurate, must anticipate what here follows but in Principia

tion
If
347
Nc
to something or other.
"D
/J"
is
"the
domain
of the relation
R".
"R"
be
"precedes",
then
"D7T will
"Cardinal
be
"the
class of all those things

"NC,"
which precede
"D
anything".
number",
is
defined
as
Nc",
"the
domain
of the relation
Xc".
We now
*100-01
"
turn to the meaning of
"Nc".
Nc = sm.
Df
"
Nc
"
is
the relation of the class of referents of
sm
"
"
to
sm
"
itself.
First,
let
us see the meaning of the arrow over
"sm".
*32-01
"
R = ay[a =
relation of
$(xRy)}.
y,
Df
is
R"
is
s,
"the
a to
where a and y are such that a
the class
of x
"
each of which has the relation

be the relation of the class
"sm".
"
to
y"
If
R"
7
be
"precedes",
R"
will
"predecessors of y
to y
itself.
Now
for
We
.
shall best not
study
its
definition but a
somewhat
simpler proposition.
*73-l
"a
asm/3. =
is
(3/0 .flcl
"For
->1
I) /?
= G7?
sm
0"
equivalent to
some
is
relation R,
is
a one-to-one relation,
/?".
while a
is
the domain of
and
the converse-domain of
of
"G
/f"
We
have here anticipated the meaning
and
"
of
>
1".
*33-02
"G"
Q = M[0 =
is
y{(Xx).xRy}].
for
Df
is
"the
relation of (a class)
to (a relation) R,
when
and
are
such that
the relation
"D
R"
the class of #
to
y".
s,
each of which (something or other) x has

this
Comparing
"Q
/?",
with the definition of
"D"
and
"
of
above, we
see that
the converse-domain of
/?, is
the class
R"
of those things to
be
"precedes",
"G*/?"
which something or other has the relation R. If will be the class of those things which are preceded
by something or
other.
R(R"d
*71
-03
!->!==
R"D
1).
"
Df
1
This involves the meaning of
"#",
of
",
and
"
of
*32-02
"5"
R =
to
0.r|0
= $(xRy)}.
to x,
Df
signifies
i/
"the
relation of
when
and
.r
are such that
is
the
class of
which x has the relation R.
^37-01
"#"0"
fl"0
if-((3y)
.r
.yep.xRy}.
?/,
Df
?/
is
"the
class of
such that, for some
is
member
of the
of 0,
s)
and z
is
has the relation
to
?/.
In other words,
"#"0"
(the
Rs
the
348
class of things
which have the relation

be
"precedes",
/?.
to
some member
or other of the
class 0.
all
If
"R"
"fl"0"
will
be the class of predecessors of
(any)
members
of
#71-03.
With the help "1
of this last
is
"the
and
of preceding definitions,
we can now read
1"
class (of relations) R,
such that whatever has
the relation
to
is
any member
contained in
of the class of things-to-which-anything-has1;
the-relation-72,
and whatever Or more

if
is
such that any
member
"
of the class of those-things-which-have-the-relation-#-to-anything has the
relation
is
R to it, is contained in
(3
"
freely
13
and
intelligibly
"
the class of relations, R, such that
is
"
true, then
"
is
a class of
all
one
member and
is
a class of one
member:
"asmjS"
>
is
the class of
is
one-to-one correspondences.
Hence
means
"There
a one-to(3.
"sm"
one correspondence of the members of

is
a.
with the members of
the relation of classes which are (cardinally) similar.
The
analysis of the idea of cardinal
number has now been

are
all
carried out
until the undefined symbols, except
"1",
of
them
logical
symbols;
of propoy)]\
of relations,
R; of
classes, a,
/3,
etc.;
of individuals, x, y, etc.;
is
sitional functions
"
such as x
(3.x)
.
Ry
is
[which
a special case of
e,
<p(x,
of
(px
for
some
"the
x",
<px;
the relations
true".
c, and =; and the idea
(1 x)(<px),
x for which px
This last notion occurs in various
special cases, such as

"
D^R,
R"0,
etc.
"
is
also defined in terms
involved are incapable of
which reduce to these, but the definitions more accurately, ordinary precise translation
language
is
incapable of translating them.

I
1
*52-01
*ol-01 *50-01
"7"
==
a{&x) .a =
7.
x}.
Df
Df
(x
= xy
y).
Df 13
"i"
is
the relation of identity;
is
the class of those things which have

"
the relation of identity to something or other; and

classes,
i.
"
is
the class of such
e.,
the class of
"
all
classes
having only a single
member.
Thus the
t
definition of
"
is
given in terms of the idea of individuals, x and y
of
the relation
of classes,
and the idea involved

translation
in the use of the

1 is
arrow
in
over
7,
which has already been analyzed.
This definition of
no
is
wise circular, however
much
its
may
1
".
suggest that
it is;
nor
there any circularity involved in the fact that the definition of cardinal
number
13
requires the previous definition of

=, which
"
Strictly, analysis of
required.
But the lack
of this does
differs from the defining relation, [. not obscure the analysis, so we omit
... Df], is
it
here.
349
We have now completely accomplished the analysis of the idea of cardinal

number
into constituents
all
of
which belong to mathematical
logic.
The
important significance of this analysis for the method involved we must

postpone for a moment to discuss the definition itself. If we go back over these definitions, we find that the notion of cardinal
number can now be

all
defined as follows:
"Cardinal number" is
the class of
those classes the
members
of
which have a one-to-one correspondence

"Cardinal
(with
all
members
of
some other
class).
number" is
is
the class of
all
the cardinal numbers;
and a cardinal number
the class of
those
of a
classes
whose members have a one-to-one correlation with the members

This
is
given class.
definition
"in
extension".
We
most frequently think

Definition
of the cardinal
number
of a class, a, as a property of the class.
in extension determines
of all those things
any such property by logically exhibiting the class which have that property. Thus if a be the class com
John, the cardinal number of a
will
posed of Henry,
Mary and
be deter
mined by
will,
logically exhibiting all those classes
which have a one-to-one
correlation with the

"3"
members
of
i.
e., all
the classes with three members.

"4",
then, be the class of
all
classes
having three members;
the
class of all classes of four, etc.
And
"cardinal number"
in general will be
the class of
It
all
such classes of classes.
be well to observe here also that, by means of ideographic sym could not bols, we can represent exactly, and in brief space, ideas which in mind in any other terms. be grasped or expressed or carried
may
possibly
him it Perhaps the reader has not grasped those presented: we can assure And if the symbolism appals clear. is not difficult once the symbolism is of by its unfamiliarity, we would call attention to the fact that the number nor is their meaning more obscure than different symbols is not greater, It is the persistent accuracy of the those of the ordinary algebraic signs. far be it from us to suggest that we do not that has troubled him;
analysis
like to
think accurately.
appeal to us as unnecessary and burdensome. But observe the consequences of it for the method. When "cardinal
So
much
analysis
may
"D
number" is
defined as
Xc,"
all
the properties of cardinal

"Xc"
number
follow
from the properties

represented by
.
of
"D"
and
and the relation between these

"sm"
And when
these in turn are defined in terms of
and the idea expressed by the arrow, and the properties of the entities which define them.
constituents of
"cardinal
so on, their properties follow from
And
finally,
when
all
the
number",
and the
other ideas of
arithmetic
350
sitions about cardinal
have been analyzed into ideas which belong to symbolic logic, all the propo number follow from these definitions. When analysis
of the ideas of arithmetic
is
complete,
all
the propositions of arithmetic
follow from the definitions of arithmetic together with the propositions of

logic.
Now
its
in Principia
Mathematica
it is
the ideas of mathematics.
Hence the whole
found possible to so analyze all of mathematics is proved
from
definitions together with the propositions of logic.
And, except
the logic, no branch of mathematics needs any primitive ideas or postulates of

its
own. It is thus demonstrated by this analysis that the only postulates and primitive ideas necessary for the whole of mathematics are the postu lates and primitive ideas of logic.
In the light of
this,
we can understand Mr.

the class of
all
Russell
definition
of
mathematics:
"Pure
14
Mathematics
is
propositions of the form
p im
plies q
where p and
q are propositions containing
one or more variables,
the same in the two propositions, and neither p nor q contains any constants
except logical constants.

in
And
logical constants are all notions definable
terms of the following: Implication, the relation of a term to the class of which it is a member, the notion of such that, the notion of relation, and
such further notions as
of the
may
be involved in the general notion of propositions
above
form."
The content
in
of
mathematics, on this view,
is
the assertion that certain

all
propositions imply certain others,
and these propositions are

that
is,
expressible
terms of
"logical
constants",
the primitive ideas of symbolic
logic.
These undefined notions, as the reader is already aware, need not be numerous: ten or a dozen are sufficient. And from definitions in terms,
finally, of these
and from the postulates
of symbolic logic, the
whole of
mathematics
is
deducible.
The
logistic
metic of
development of a mathematical system may, like the arith the Formulaire, assume certain undefined mathematical ideas and
in
mathematical postulates
logical ideas
terms of these ideas, and thus
differ
from an
ordinary deductive system of abstract mathematics only by expressing the
by ideographic symbols and by using principles of proof supplied by symbolic logic. Or it may, like arithmetic in Principia Mathematica, assume no undefined ideas beyond
which occur
in its postulates
those of logic, define
all its
mathematical ideas
in
terms of these, and thus
require no postulates except, again, those

14
of logic.
Or
it
may
pursue an
Principles of Mathematics, p. 3.

intermediate course, assuming some of
its
351
ideas as primitive but defining
others in terms of a previously developed logic, and thus require

postulates of its
some
own but
still
dispense with others which would have been
necessary in a non-logistic treatment.
adopted, the general method of proof in logistic will be the same, and will differ from any nonof these of procedure
is
But whichever
treatment.
modes
logistic
non-logistic
development
will
proceed from postulates
to theorems
by immediate
inference or the use of syllogism, or

logical methods.
or the reductio
ad absurdum, and such general

of
enthymeme, Or it may,
which
upon occasion, make use
methods
It
of reasoning the validity of
depends upon the subject matter.

induction",
if
may make
use of "mathematical
which requires the order of a discrete series with a first term. Or proofs of consistency and independence of the postulates are offered, these will make use of logical principles which are most complex and difficult of comprehension principles of which no thoroughly satisfactory account
has ever been given. The principles of all this reasoning will not be men tioned; it will be supposed that they are understood, though sometimes
they are clear neither to the reader nor to the mathematician who uses them, and they may even be such that nobody really understands them.
(This
is
not to say that such proofs are unsound.
Proofs by
"
mathematical
induction" were valid before Frege and Peano showed that they are strictly deductive in all respects. But in mathematics as in other matters, the
assurance or recognition of validity rests upon familiarity and upon prag matic sanctions more often than upon consciously formulated principles.)
As contrasted with
each step of proof.
this,
the logistic
method
requires that every principle
state of proof be explicitly given, because these principles are required to
by any extended in arithmetic or geometry do not differ in proof of Chapter V. Proofs method from proofs in the logic, and the procedures there illustrated are An examination of these proofs will show that postu universal in logistic.
The method
of proof in logistic
is
sufficiently illustrated
lates
and previously established theorems are used as

q,
r,
principles of proof
by
substituting for the variables p,
etc., in
these propositions, other
as values of their variables. expressions which can be regarded

principle
(p
-i
</)
The
general
-*
(-?
-*
-P)
can thus be made to state
352
A
for
it

for
by substituting p q and n e D Xc for q,

[(/z e
p and p
q.
Or
if
/*
XC
-*
be substituted for p
states
-i
NC)
(M
Nc)]
<
[-(M
Xc)
-(M
XC)]
is is
Thus any
stated
special case
which comes under a general
logical principle
by that principle, exactly the manner in which the

development
when
the proper substitutions are made.

principles of proof
This
which belong to sym
bolic logic state the various steps of any particular proof in the logistic
of arithmetic or geometry.
first
Returning to our
example, we discover that in
the
first
half,
p,
it
is is
itself
a true proposition.
proved
as, in fact,
is
in the last chapter.

is,
We
Suppose this already can then assert what
-i
stated
by the above to imply, that

-P
-*
"(p
q)
We
thus prove this new theorem by using p q H p as a premise. To use a previous proposition as a premise means, in the logistic method, exactly
to
this:
make such
substitutions in a general principle of inference, like

(p
-J
q)
-I
(-q
-*
-p)
first
that the theorem to be used as a premise appears in the

expression
half of the
sign.
the part which precedes the
main implication
That
then
part of the expression
which follows the main implication sign
may
be asserted as a consequence of this premise.
There are two other operations which may be used in the proofs of the operation of substituting one of a pair of equivalent expressions logistic
for the
other,
and the operation
of
15
combining two previously asserted
propositions into a single assertion.
The
first
of these
is
exemplified
whenever we make use

from which our
of a definition.
is
For example, we have,
in the
system
illustration
borrowed, the definition
p+q =
and the theorem
-(-p-q)
15
The operation
of
combining two propositions, p and
q,
into the single assertion,
q, is
not required in systems based on material implication, because
we have
pqcr = pc(qcr)

If in
353
the definition,
we
substitute -p for
p and -q
r/ )]
for
q, it
states
-p + -q = -[-(-p)._(-
And
then,
making the substitutions which the theorem p = -(-p)

-p +
allows,
we have
=
-<i
-(P
</)
which
may
be asserted as a theorem.
Again,
if
we return
to the
theorem
proved above,
-*
"P
-(p
</)
we
are allowed,
by
this last equivalence, to
make
the substitution in
it
of
-p + -q for -(p
q).
Thus we prove
-p
-*
-p + -q
in proof
This sufficiently illustrates the part played

equivalent expressions.
by the substitution of
We may now
(1)
see exactly
what the mechanics
of the logistic
method
is.
The only operations
required, or allowable, in proof are the following:

logic, other,
In some postulate or theorem of symbolic
and usually
q,
r,
more complex, propositions

substitutions are
are substituted for the variables p,
etc.,
which represent propositions.
The
postulate or theorem in which these

states,
made
is
thereby used as a principle of proof which

results
in this particular case,
the proposition which
when
these substi
tutions are made.

(2)
The
in
of proof postulate or theorem of logic to be used as a principle
may, and
in
most cases does,
state that something implies something else.

will
In that event, we
may make
such substitutions as
produce an expression
which that part which precedes the main implication sign becomes of logic, of identical with some postulate or previously proved theorem That part of the expression which arithmetic, of geometry, or whatever.
follows the
main implication sign may then be separately asserted as a new theorem, or lemma, which is thus established. The postulate or previously the main implication proved theorem which is identical with what precedes
sign, in
such a case,
is
thus used as a premise.
of any It should here be noted that propositions of logic, of geometry, of symbolic as premises; but only propositions logistic system, may be used
logic,
which state implications, are used as general principles of inference. one of a pair of equivalent expres (3) At any stage of a demonstration,
sions
may
be substituted for the other.
24
354
for example,
A
(4)
If,

for a certain desired
two premises are required
consequence, and each of these premises has been separately proved, then the two may be combined in a single assertion.
These are
tions
all
the operations which are strictly allowable in demonstra
by the
logistic
method.
To
their
simplicity
and
definiteness
is
attributable a large part of the precision

is
and
rigor of the
method.
Proof
here not a process in which certain premises retire into
ing faculty, there to be transformed

in the
form of the conclusion.
by the alchemy of The whole operation takes

is
somebody s reason thought and emerge

place visibly in
the successive lines of work, according to definite rules of the simplest

possible description.
The
process
as infallible
and as mechanical as the
adding machine
as the reader
is
except in the choice of substitutions to be made, for which,

discover
may
by experiment, a
certain
amount
of intelligence
required,
if
the results are to be of interest.
III.
"HETERODOX"
VIEW OF THE NATURE OF MATHEMATICS AND OF

LOGISTIC
have now surveyed the general character of forth what may be called the orthodox" view of it.
"
We
logistic
and have
set
As was stated
earlier
in the chapter, the account
which has now been given is such as would exclude certain systems which would almost certainly be classified as And these excluded systems are most naturally logistic in their character.
allied
with another view of
logistic,
forth.
The
differences
between the
which we must now attempt to set "orthodox" and this "heterodox"
view have to do principally with two questions: (1) What is the nature of the fundamental operations in mathematics; are they essentially of the
nature of logical inference and the
like,
or are they fundamentally arbitrary
and
extra-logical?
(2)
Is logistic ideally to
is its
be stated so that
all its
assertions
are metaphysically true, or
principal business the exhibition of logical
types of order without reference to any interpretation or application?
The two questions
are related.
It will
appear that the systems which the
previous account of logistic did not cover are such as have been devised from a somewhat different point of departure. One might characterize the logistic of Principia Mathematica roughly by saying that the order of
logic
is
assumed, and the order of the other branches then follows from the
meaning of their terms.

attempt to
set
On
the other hand, the systems which remain to be
discussed might, equally roughly, be characterized
up a type
of logical order,
which
shall be general
by saying that they and as

inclusive as possible,
355
and to
to cover
let
the meaning of terms depend upon their
properties of order
is
and
it
relation.
all
Thus
this
"heterodox"
view of
logistic
investigations and developments of types of logical order which involve none but ideographic symbols and proceed by operations which may be stated with precision and generality.
one which takes
In any case,
it
must be granted that the
operations of the logistic
method
are themselves pre-logical, in the sense that they underlie the proofs of
logic as well as of other branches.
The assumption
of
in a
all
of these operations
substitution, etc.
logistic.
is
the most fundamental
the assumptions of
It is possible to
view the subject
way which makes such

and does not
pre-logical
principles the fundamentally important thing,
regard as essential the use of symbolic logic as a foundation.

priety of the term logistic for such studies
may
be questioned.
The pro But if
such a different view
is
consistent
and
useful, it is of little
consequence
what the method ought
to be called.
if
We
any
will
see at once that,
such a view can be maintained, Mr. Russell

for
definition of mathematics,
"logistic"
quoted above, is arbitrary, development which is not based upon

all.
by that definition logic as # foundation
not be mathematics at
As a by
If to
fact, it will
be simplest to present this
"heterodox"
view of
logistic
first
presenting and explaining the cor
relative view of mathematics.
the reader
later
we seem
here to wander
from the subject, we promise to return
and draw the moral.

marks in which
remainder derived from these
mathematical system
is
any
set of strings of recognizable
some of
the strings are taken initially
and
the
by operations performed according
to rules
which are independent of any mean

consist of
ing assigned
of
to the
marks.
is
That a system should

it
is
marks instead
sounds or odors
immaterial, but
string-like
is
convenient to discuss mathe

is
matics as written.
of notation.
in
The
there
arrangement
due simply to our habits

single
And
no theoretical reason why a

"string".
mark may
it
not,
some
cases, be recognized as a
The
distinctive feature of this definition lies in the fact that
regards
tri
mathematics as dealing, not with certain denoted things

angles,
numbers,
nor with certain symbolized "concepts" or etc. but solely with recognizable marks, and dealing with them
it is
"meanings",
in
such wise
that
sent.
wholly independent of any This might be called the external view of

"
question as to
what the marks repre mathematics or mathe

"
"
matics without
of
meaning".
It distinguishes
mathematics from other
sets
marks by
precisely those criteria
which the external observer can always
356
Whatever the mathematician has in his mind when he develops a system, what he does is to set down certain marks and proceed to manipulate them in ways which are capable of the above description. This view is, in many ways, suggested by growing tendencies in mathe
apply.
matics.
Systems become
"abstract",
entities
with which they deal
"have
no properties save those predicated by postulates and definitions", and It becomes recognized propositions lose their phenomenal reference. for which lies in the properties of the that any procedure the only ground
things denoted
as
"constructions"
in
geometry
is
defective
and un-
mathematical. Demonstrations must take no advantage of the names by which the entities are called. But if Mr. Russell is right, the mathe
matician has given over the metaphysics of space and of the infinite only
to be plunged into the metaphysics of classes
of empirical possibility
"logical"
and
of functions.
Questions
of
and factual existence are replaced by questions

"existence"
possibility
questions about the
of
classes,
about
the
empty
"
or null-class, about the class of

",
all classes,
about
"individuals",
about
descriptions
"values"
about the relation of a class of one to
its
only member,
of func
"system
about the
tions,
of variables
and the
"range
of
significance"
"types"
about material and formal implication, about
and
atic
ambiguities"
and
"hierarchies
of
propositions".
And we may
be
pardoned for wondering if the last state of that mathematician is not worse than the first. It is possible to think that these logico-metaphysical
questions are essentially as non-mathematical as the earlier ones about
empirical possibility and phenomenal existence.
One may maintain that

order.
distri
nothing
is
essential in a
And
the type of order
mathematical system except the type of may be viewed as a question solely of the
like the
bution of certain marks and certain complexes of marks in the system.
The question
of logical
meaning,
question of empirical denotation,
may
be regarded as one of possible applications and not of anything internal

itself.
to the system
tration.
it may prove best to give an illus Let us choose a single mathematical system and see what we shall make of it by regarding it simply as a set of strings of marks.
Before discussing the matter further,
We
take initially the following eight strings:

(P D
tf)
(~JP
0)
(pxq) = ~(~pv~q) = ((P*q) (P = q)
*(<7=>p))
357
((p v(ry
vr))D(r/v(p vr)))
((r/Dr)D((^ vry)D(p vr)))
We
must now
state rules according to
from the above.
In stating these rules,
which other strings can be derived we shall refer to quids and quods:
these words are to have no connotation; they serve merely for abbreviation
in referring to certain
(1)
marks.
x, D,
r,
(2)
The marks +, The marks p, q,

above
and =, are quods. are quids; and any recognizable mark not appear
==,
ing in the
(3)
)
may
be taken arbitrarily as a quid.
Any
(,
expression consisting of two quids, one quod, and the
marks
and
(4)
in the order (quid
quod
quid),
may
be treated as a quid.
The combination
of
~ any quid preceded immediately by the mark

be repeated.
may
be treated as a quid.
(5)
(0)
Any Any
any
string in the set
may
quid which
is
separated only by the
mark
==
from some other
quid, in
string in the set,
may
be substituted for that other quid any
where.
(7)
In any string in the
initial set, or in
any
string
added to the
list
according to rule,
any quid whatever may be substituted for p or q or r, When a quid is substituted or for any quid consisting of only one mark.
for
any mark in a string, the same quid must also be substituted for that same mark wherever it appears in the string. of a quid consisting of (8) The string resulting from the substitution more than one mark for a quid of one mark, according to (7), may be added
to the
(9)
list
of strings.
list,
In any string added to the
according to
is
(8),
if
that portion of
string
the string which precedes any

in the set, preceded
mark D
identical with
some other
by
(,
then the portion of that string which follows the

repeated, with the omission of the
mark D referred to may be separately final mark ), and added to the set.
important not to refer to
theses,"
These rules are unnecessarily awkward.

"propositions",
In the illustration,
it
was
"relations", "variables",
"paren
etc., lest it
should not be clear that the rules are independent of
the meanings of the marks.
The
original eight strings
But though cumbersome, they are still precise. of marks are, with minor changes of notation,
358
definitions
A
and postulates
all

of divisions
#1
to
*5
in Principia
all
Mathematica.
By
following the rules given, anyone
may
derive
the theorems of these
divisions
and
other consequences of these assumptions, without knowing
anything about symbolic logic either before or after. In fact, these rules formulate exactly what the authors have done in proving the theorems
from the postulates. 16 For this reason, it is unnecessary to carry our illus tration further and actually derive other strings of marks from the initial The process may be observed in detail in Principia Mathematica: set.
it is,
in all
important respects, the same with the process of proof exhibited

in Principia
in our
Chapter V.
of
The method
r
development
Mathematica
differs
from the
one w e have suggested, not in the actual manipulation of the strings of marks, but most fundamentally in that the reasons why the principles
of their operations are to be found, not in explicitly stated rules, but in
discussions
and assumptions concerning the conceptual content

in discussions of the nature
of
the
system.
In fact, the rules of operation are contained in explanations of
the meaning of the notation
of "elementary propositions", "elementary
and properties propositional functions", and
so forth.
For example, instead

for p,
q,
r,
of stating that certain substitutions
may
be
made
etc.,
they assume as primitive ideas the notions of

p,
q,
r,
"elementary
propositions"
functions"
etc.
the
notion
of
of
"elementary
propositional
px,
\J/z,
etc.
and the idea
"negation",
indi
cated by writing ~ immediately before the proposition.
And
in part, the
rules of operation are contained in certain postulates, distinguished
by
an
their non-symbolic form:
"If
is
an elementary proposition, ~p
is
p and q are elementary propositions, pvq elementary proposition", is an elementary proposition", and (pp and ij/p are elementary propo
"If
"If
sitional functions
is
v $p which take elementary propositions as arguments, an elementary propositional function. The warrant for the substitution
<pp
of various complexes for p, q }
r,
etc., is
contained in these.
whiich states
which requires our complicated rule

be done,
is
(9),
The operation precisely what may
covered by their assumptions:

is true",
<px
"Anything
<px
elementary proposition
is
and
"When
implied by a true can be asserted, where x

is
a real variable, and

ipx
\f/x
can be asserted, where x

is
a real variable,
then
can be asserted, where x
a real
nection between these and our rule,

16
variable". To make the con we must remember that 3 is the
With the single and unimportant except ion that they do not add every new string which they arrive at, to the list of strings. Many such are simply asserted as lemmas, used immediately for one further proof and not listed as theorems.
symbol for implies", that if what precedes this sign is identical with some other string in the set, that means that what precedes is true or is asserted, and that the number of open and close parenthetic marks will
"
indicate whether the implication in question

cation.
is
the main, or asserted, impli
We
"
have chosen
this particular
system to
illustrate the
requirements
will
of
mathematics without
meaning" for a special reason,
which
appear
shortly.
But the same
sort of modifications
would be and
sufficient to bring
any good mathematical system into modifications would be necessary.

If,
this form;
in
most cases such
system in question were one of the better-known class, K, of elements, a, b, c, etc.", algebras, w e should probably have a and b are elements in K, a + b is an element such assumptions as and
for example, the
r
"a
"If
in
These would do duty as the principles according to which, for or example, x + y would be substituted for a in any symbolic postulate less radical, hardly theorem. The changes in such a system would be
K".
more than
alterations in phraseology, but
still
necessary.
upon meanings for the validity of the method has obvious is measuring two advantages. It is simple and natural and clear. (So
Ileliance
line-segments with a foot rule to prove equality.)
It also
has disadvantages.
Besides the logico-metaphysical questions into which this reliance upon it works a certain meanings plunges us, there is the disadvantage that confusion of the form of the system with
of these
is
its
content.
The
clear separation
the ideal set by "mathematics

free
without
all
meaning".
Not only
must mathematical procedure be

empirical data;
it
from
appeal to intuition or to
should also be independent of the meaning of any special
matter of the system. No alteration concepts which constitute the subject of mathematical procedure anywhere should be covered by or
abridgment relations or other the names which are given to the terms. Only those a system as a particular type of order should properties which determine be allowed to make a difference in its manner of development. to To secure complete separation of form from special content, and to use the system as purely formal and abstract, means precisely
present
principles of operation
which are capable

"
of statement as rules for the
manipulation of marks
"
any reference
consideration
to
is
avoidance of though, in general, the meticulous The important meanings would be a piece of pedantry.
the fact that the operations of any abstract and, really rigorous to truth mathematical system are capable of formulation without any reference
360
or meanings. 11
A We
are less interested in
any superiority
If
of this
"external
view of
mathematics",
or in the conjectured advantages of such procedure
as has been suggested, than in its bare possibility.
the considerations here
presented are not wholly mistaken, then the ideal of form which requires
17 It is possible to regard such manipulation of marks, the discovery of sufficiently precise rules and of initial strings which will, together, determine certain results, and the exhibition of the results which such systems give, as the sole business of the mathematician.
Mathematics, so developed, achieves the utmost economy of assertion. Nothing is There are no primitive ideas. Since no meanings are given to the characters, the strings are neither true nor false. Nothing is assumed to be true, and nothing is asserted It is not even necessary to assert that certain operations upon certain marks as "proved". give certain other marks. The initial strings are set down: the requirements of pure mathematics are satisfied if the others are got and recorded. Yet these initial strings and the rules of operation determine a definite set of strings of marks determine unambiguously and absolutely a certain mathematical system. To many, such a view will seem to exclude from mathematics everything worthy of the name. These will urge that the modern developments of mathematics have aimed at exact analysis into fundamental concepts; that this analysis does, as a fact, bring about such simplification of the essential operations as to make possible mechanical manipulation of the system without reference to meanings; but that it is absurd to take this shell of refined symbolism for the meat of mathematics. To any such, it might be replied that the development of kinematics as an abstract mathematical system does not remove the physics of matter in motion from the field of experimental investigation; that abstract
asserted.
geometry still leaves room for all sorts of interesting inquiry about the nature of our space that for every system which is freed from empirical denotations there is created the separate investigation of the possible applications of this system. Correspondingly, for every
:
system which
is made independent of classes, individuals, relations, and so on, there is created the separate investigation of the metaphysical status of the classes, individuals, and relations in question of the application of the system of marks to systems of more special "concepts", i. e. to systems of logical and metaphysical entities. That we are
more
interested .in
interested in the applications of a system than in its rigorous development, more its "meaning" than in its structure, should not lead to a confusion of meaning
with structure, of applications with method of development. It may be further objected that this view seems to remove mathematics from the field of science altogether and make it simply an art; that the computer would, by this
But there is one feature of mathematics, even as a definition, be the ideal mathematician. system of marks, which is not, and cannot be made, mechanical. Valid results may be obtained by mechanical operations, and each single step may be essentially mechanical,
"required" or "interesting" or "valuable" results will need an in and ingenious manipulator. Gulliver found the people of Brobdingnag (?) feed ing letters into a machine and waiting for it to turn out a masterpiece. Well, master pieces are combinations achieved by placing letters in a certain order! However mechanical the single operation, it will take a mathematician to produce masterpieces of mathematics. A machine, or machine-like process, will start from something given, take steps of a deter mined nature, and render the result, whatever it is; but it will not choose its point of departure and select, out of various possibilities, the steps to be taken in order to achieve a
yet the derivation of

telligent
desired result.
Is not just this ingenuity in controlling the destination of simple operations
which mathematics requires? The mathematician, like any other sci engaged upon what are, from the point of view of the finished science, inverse processes: he gets, by trial and error, or intuition, or analogy, what he
skill
the peculiar
entific investigator, is largely
presents finally as rigidly necessary. To produce or reveal necessities previously noticed this is the peculiar artistry of his work.
un
361
that mathematics abstract not only from possible empirical meanings but
also
from
if
logical or
is
And
is
this
metaphysical meanings is a wholly attainable ideal. possible, then Mr. Russell s view that "Pure Mathematics
propositions of the form
ideal of
the class of
all
p implies
it
q
is
etc.",
is
an arbitrary
definition,
and the
its
form which
imposes
not a necessary one,
but must take
chances with other such ideals.
The
decision
among
these will, then, be a matter of choice, dependent upon the advantages to
be gained by one or the other form.

"
There
is
no a
priori reason
why
systems which are generated by mathematical" may be peculiar to the system and meaningless
"sound"
operations, some
of
which
in logic, are not just as
and
"good"
and even
of
"ideal"
as systems developed
by the com
reduces
all
pletely
analytical
method
Principia Mathematica which
development operations to those of logic. And "extra-logical" since symbolic logic itself may be be just as universal as the "logical may
of
",
modes
developed by the "extra-logical" method. It was to make this clear that we chose the particular system which we did for our illustration.
In fact, symbolic
logic, or
that branch of
it
which
is
developed
first
as a
by operations the validity of which is It may, indeed, be the case presumed apart from that logic is developed by methods which it validates by its own theorems,
basis for others, must be developed
the logic so developed.
when
these are proved;
it
may
thus be
"self-critical",
or
"circular"
in a
sense which
to the point:
means consistency rather than

if
fallacy.
is
But
this
is
not really
the validity of certain operations

it is
presupposed, then that
validity is presupposed, whether

is,
symbolic logic
arbitrary
There afterward proved valid or not. that a system of then, a certain advantage in the explicit recognition a set of strings of marks, manipulated by certain is
merely
and
"extra-logical"
principles.
It
is,
in fact,
only on this view
that symbolic logic can be abstract.
lates
been pointed out, is and theorems are used to

logically valid, these postulates
For symbolic logic, as has already in that its postu peculiar among mathematical systems
state proofs.
If,
then, the proofs are to be

true,
and theorems must be

"proofs"
and the system

"valid"
cannot be abstract.
But
if
the
are required to be
only
in the sense that certain arbitrary and extra-logical rules for manipulation have been observed, then it matters no more in logic than in any other There branch whether the propositions be true, or even what they mean. is in the case of other mathe of choice here that there same
is
the
possibility
alterna phrased most sharply as the "external view of mathematics tive between the Russellian view and the ^ It may be noted that if mathematics consists of "propositions of the form p im-
matical systems
the choice which
is
362
If we take this view of mathematics, or any view which regards arbitrary mathematical operations as equally fundamental with the operations of logic, we shall then give a different account of logistic and of its relation
to logic.
We
shall, in
that case, regard symbolic logic as one mathematical
system, or type of order,
among
others.
We
shall recognize the possibility
of generating all other types of order

find
from the order
of logic,
but we
shall
no
necessity in this proceeding.
We
may,
possibly, find
some other
be derived.
very general type of order from which the order of logic
may
And
the question of any hierarchic arrangement of systems will then depend
upon convenience or simplicity or some other pragmatic consideration.

Logistic will, then, be defined not
by any
relation to symbolic logic but as
the study of types of order as such, or as any development of mathematics
which seeks a high degree of generality and complete independence of
any particular subject matter.

IV.
THE
LOGISTIC
METHOD
OF
KEMPE AND ROYCE

view of
mathematics"
We
should not care to insist upon the

of logistic
"external
and the consequent view

siderations aside,
of
it
which has been outlined.
Other con
mathematical form when there
seems especially dubious to dogmatize about the ideal is no common agreement on the topic
among mathematicians.
prefaced this discussion:
But we can now answer the questions which

Are the fundamental operations of mathematics
those of logic or are they extra-logical?

stated that
all its
And
is
logistic ideally to
is it is
be so
assertions are metaphysically true or
to exhibit certain general types of order?
The answer
concerned simply that it is entirely a
matter of choice, since either view can consistently be maintained and mathematics be developed in the light of it. 19 This is especially important
for us, since, as has
been mentioned, there are certain studies which would
most naturally be
<?
called logistic
which would not be covered by the
"ortho-
where p and q are propositions containing one or more variables, etc.", and the plies theorems about "implies" are required to be true if proofs are to be valid, all mathematics must be true in order to be valid. On this view, abstractness can reside only in the range of the variables contained in p and q. 19 One case in which the "external view of mathematics" is highly convenient, is of There are various symbolic which differ from one another especial interest to us. both in method and in content. Discussion of the correctness and relative values of these is almost impossible unless we recognize that the order of logic can be viewed quite apart from its content that a symbolic logic may be abstract, just like any other branch of mathematics and thus separate the question of mathematical consistency (of mere ob servance of arbitrary and precise principles of operation) from questions of applicability
,
" "
"logics"
of a system to valid reasoning. The difficulty of discussion of "implies" in the last chapter.
making
this separation
hampered our

view, since they are based, not
363
upon logic, but upon an order prior These studies exemplify a method which differs in notable respects both from that of Peano and that of Principia MatheAnd it seems highly desirable that we should discuss this alterna matica.
dox"
to or inclusive of logic.
tive
method without
a
initial prejudice.
It is characteristic of this alternative

initially
field,
method that
more
it
seeks to define
or class of entities,
and an order
in this field,
which
shall
be
mathematically as inclusive as possible, so that

specified
special orders
may
be
by principles of selection amongst the entities. It is distinguished from the method followed by Peano in the Formidaire by the fact that it
seeks to get special orders, such as that of geometry, without further
"existence postulates",
and from the method
of Principia Mathematica
by
tion
the attempt to substitute selection ivithin an initial order for analysis (defini by previous ideas) of newly introduced terms. The result is that this
method
fields
is
different special particularly adapted to exhibit the analogies of
the partial identities of various types of order. The application of this method has not been carried out extensively
enough so that we
tions.
may
is,
feel certain either of its
The method
or, say, all
in
advantages or of its limita a certain sense, exemplified wherever we have

all
various mathematical systems
of
which
satisfy a given set of postulates,
but each
but one
satisfying
some one
field
or
more
of the postulates
"vacuously".
For here we have an ordered

specified
within which other and

("Selection"
more limited systems are
by
a sort of selection.
is
not the proper word, but no better one has occurred to us.) It is particu in two studies of the relation of geometry to logic that the method has
larly
been consciously followed: 20 in a paper by A. B. Kempe, "On the Relation between the Logical Theory of Classes and the Geometrical Theory of
21
Points,"
and
in
Josiah lloyce
study,
"The
Relation of the Princi
ples of Logic to the
Foundations of
detail,
22
Geometry".
We
shall
hardly wish to
go into these studies in
but something of the mode of procedure and be indicated briefly. general character of the results achieved may of exact Kempe enunciates the principle that "... so far as processes matter depend solelj thought are concerned, the properties of any subject of i. e., that it consists of a number on the fact that it possesses form
2
Peirce s system of
of this
"logical
quaternions"
(see above,
pp.
102-04)
also exhibits
method. something 21 Proc. London Math. 22 Trans. Amer. Math.
Soc.,
xxi (1890), 147-82.
Some
article, "Types
Soc., vi (1905), 353-415. s are here reprinted and portions of the discussion of this paper in Phil. Rev. for May, 1916. S," of Order and the
Kempe
System
364
entities, certain individuals, pairs, triads, &c., certain of
each other
in all their relations,
and certain not; these
which are exactly like like and unlike indi

system of
viduals, pairs, triads, &c., being distributed through the whole

entities in a definite
23
way".
In illustration of this theory, he seeks to
derive the order both of logical classes
and
of geometrical sets of points
from assumptions in terms of a triadic relation, a c b, which may be read The type of this relation may be illustrated as between a and is follows: Let a, b, c represent areas; then ac-b symbolizes the fact that b
"b
c".
includes whatever area
is
common
is
to a
and
c,
and
is itself
included in the
to
area which comprises
what
either a or c or
common
-ab-c =
both.
Or
it
may
be expressed in the Boole-Schroder Algebra as

(ac)
cb c
(a + c),
or
a-bc
The
a
essential properties of serial order

If
may
be formulated in terms of this

is
relation.
and
and
b
c is
ac-b and ad-c, then also ad-b and bd-c. If b between a and d, then also b is between a and
rf.
between
d,
and
c is
between
and
24
abed
most fundamental property of linear sets. with reference to which precedence is deterloc. cit., p.
Thus the
If a
23
24
relation gives the
be regarded as the origin

"On
the Relation between,
etc.",
147.
Assuming
ac b to be expressed in the Boole-Schroder Algebra (as above)
by
(ac) c
this
(a +
c)
deduction
ac-b
is
is
as follows:
c)
equivalent to (a
c6 c
(a + c).
d).
ad-c
is
equivalent to (ad) cc c (a +
the laws of the algebra,
By
(a c)
b is equivalent to a -b c
c) is
0.
equivalnt to -(a + c) b = -a b -c (ad) c cis equivalent to a -c d = 0.
(a
0.
(a
+ d)
is
equivalent to -(a + d)
these premises,
c
i.
c
e.
-a
-d
0.
Combining
adding the equations,

0.
we have
a -b c + -a b -c + a -c d + -a
-d
Expanding each term

elements,
a -b
c
a, b, c, d,
c
d + a -b
of the left-hand member with reference to that one of the not already involved in it, -d + -a b -c d + -a b -c -d + a b -c d + a -b -c d + -a b c -d + -a -b c -d =
"If
By
(1)
the law
a +b
0,
=
0",
we
get from this,
and and
+ -c) + -a b -d (c + -c) = Q = a-bd + + (2) b -c d (a + -a) + -b c -d (a + -a) = (1) is equivalent to (ad) cb c(a+d), or ad-b,
a -b d
(c
-ab
-bc
-d,
0=b-cd
-d.
(2)
is
equivalent to
(b d)
(b
d),
or bd-c.
365
mined, ac-b
will represent
"6
precedes
Since ac-b and a d-c together give a d-b,

"c
precedes
d",
then
"6
precedes
rf".
and a d-c that precedes and we have: if precedes Hence this relation has the essential
c",
"c
d".
"b
c"
transitivity of serial
order, with the added precision that

"
it
retains reference
to the origin from which
precedes" is
determined.
The
last-mentioned property of this relation

it
makes
possible an inter
pretation of
for logical classes in
which
it
becomes more general than

If
the inclusion relation of ordinary syllogistic reasoning.
there should
Mars whose logical sense coincided with our own so that any conclusion which we regarded as valid would seem valid to them, and vice versa but whose psychology was somewhat different from ours,
be inhabitants of
these Martians might prefer to remark that
"b
is
between
a and
c",
rather than to note that
"All
is
and
all b is
c".
These Martians might
then carry on successfully all their reasoning For ac-b, meaning relation.
-a
is
in
terms of this triadic between
b -c
+ a -b
a general relation which, in the special case where a is the "null" class or is contained in contained in every class, becomes the familiar
"b c"
"
All b
is c
".
By
virtue of the transitivity pointed out above,
c b
and
d-c
all
together give
c is d,
d-b,
which
d".
is
the syllogism in Barbara,
"If
all b is c
and
Hence these Martians would possess a mode of own and including our own as a reasoning more comprehensive than our
then
all b is
special case.
The
all
triadic relation of
Kempe
is,
then, a very powerful one, and capable
of representing the
most fundamental relations not only in logic but in those departments of our systematic thinking where unsymmetrical
(serial)
transitive
relations
are
25
important.
"base
In terms of these triads,
Kempe
states the properties of his
system",
from whose order the
and geometry both are derived. The "base system" each having an consists of an infinite number of homogeneous elements, It is assumed that triads are disposed in infinite number of equivalents.
relations of logic
this
laws: system according to the following

2*
26
relation
this triadic should be pointed out that while capable of expressing such relations, be true not necessarily unsymmetrical: ac-b and ab-c may both for these two triads that case, b = c, as may be verified by adding the equations contained in itself, for any a and b, ab-a and ab-b always hold-& is always that any term serial order with the qualification triadic relation
It
is itself
represents
itself or is
"between"
itself
and any other-an

pp. 148-49.
entirely intelligible
and
vention.
26
See
Kempe
paper,
loc. cit.,
366
1.
A
If
2.
3.
If
4.
we have a b p and c b-q, r exists such that we have a q r and c p r. we have a b p and c p r, q exists such that we have a q r and c b r. 27 If we have a b-c and a = b, then c = a = b. If a = b, then we have a c-b and b c-a, whatever entity of the system
-
may
To
be.
these,
"No
Kempe adds
entity
is
a fifth postulate which he calls the
Law
of
Con
tinuity:
present".
absent from the system which can consistently be
From
these assumptions
is
and various
definitions in terms of
the triadic relation, he

classes
able to derive the laws of the symbolic logic of

of geometrical
sets
and fundamental properties
of points.
But
further
and most important properties

"base
of geometrical sets
system"
depend upon the

28
selection of such sets within the

If
by the law:
we have
and
then
a-p-q,
b-p-q,
a-b-p
does not hold;
p =
q.
a-p-q
least of
here represents a relation of

will hold.
a, p,
and
q,
such that some one at
If we call ab-c a "linear triad", ap-q, aq-p, pq-a then the set or locus selected by the above law will be such that no two
linear triads of the
points
comprised in
it
can have two non-equivalent

"It
points
in
common.
precisely the set of
Of such a geometric set, Kempe says: 29 is entities which is under consideration by the geometrician
when he
unlimited
is
considering the system of points which
make up
s
flat
space of
dimensions".
But there are

fessor
fluous.
certain dubious features of
Kempe
also.
procedure.
1
As Pro
Royce notes, the
Law
of Continuity
makes postulates
it
and 2 super
in spite of
"base
set",
And
there are other objections to
Moreover,
the fact that
Kempe
has assumed an infinity of elements in the
there are certain ambiguities
and
difficulties
about the application of his

as
principles to infinite collections.
In Professor Royce
the
paper,
we have no such blanket assumption

relations defined
Law
of Continuity,
and the
may
be extended without
"base
difficulty to infinite sets.
We
have here,
2",
in place of the
"
system"
"0-rela-
and
triadic relations, the

27
"system
the
^-relation
"
"
and the
If the reader will draw the triangle, a betweens" as indicated, be, and put in the the geometrical significance of these postulates will be evident. I have changed a little the order of Kempe s terms so that both 1 and 2 will be illustrated by the same triangle.
28 29
See
Kempe
paper,
loc. dt.,
pp. 176-77.
Ibid., p. 177.
367
tion".
The
^-relation
is
a polyadic relation such that F(ab.
./.rz/.
.)
is
expressible in the Boole-Schroder
Algebra as
+ -a -6
s
a b ... -x -y
...
x y
=0
F(ac/b).
This
tion
is
the generalization of
Kempe
a c-b, which
is
The
0-rela-
is a polyadic symmetrical relation which expresses simultaneously a whole set of equivalent F-relations. 0(abc. .) is expressible as
.
a b c
+ -a -b -c
=
but
in
We
have used the algebra
of classes to express these relations,

is,
Professor
Royce
paper, this order
of course, reversed.
In terms of
0-
relations, the ideas of the logic of classes are defined,
and from the postula-
tion of certain
rived.
-relations, the laws of the
symbolic logic of classes are de
And,
order of
ties of
most interesting ways which we cannot here discuss, the the system 2 is also shown to possess all the fundamental proper
in
geometric sets of points.

"the
The system 2 has
a structure such that
it
might be called
logical continuum",
and there are good grounds
for
presuming that types of order in the greatest variety may be specified within In the words of Professor Royce: 30 the system simply by selection.
"
Wherever a
is
linear series
is
in question,
wherever an origin of coordi

,
nates
employed, wherever
cause and effect
orientation in space or direction of tendency in
ground and consequence, time are in question, the
diadic asymmetrical relations involved are essentially the

tion here symbolized
same as the
;
rela
by p
</
-<
,,
q,
q is
"between"
y and p
or,
with y
as origin,
in q
is
;
or, in
or, p precedes terms of propositions,

;
where y
is
the null-class,
1
is
contained
p implies
].
This expression, then,
due to certain of our best established practical instincts and to some of our best fixed intellectual habits. Yet it is not the only expression for the
It
is
relations involved.
in several respects inferior to the

.
.
more
direct
When, in fact, we expression in terms of 0-relations. relations of the system S merely in terms of the antecedentscribe the choice of relation, we not only limit ourselves to an arbitrary
.
attempt to de
consequent
origin [y in
of
p more than a
<
q],
but miss the power to survey at a glance relations

character."
diadic, or triadic
V.
SUMMARY AND CONCLUSION

There
is,
There
first,
are, then, in general, three types of logistic procedure.

"simple
the
logistic method", as
we may
etc.,"
call it
the most obvious
30 "The
Relations of the Principles of Logic,
loc. tit.,
pp. 381-82.
368
one, in which the various branches of pure mathematics, taken in the nonlogistic
but abstract form, are simply translated into the logistic terms which symbolize ideographically the relations involved in proof. When this
is
translation
made, the proofs
in arithmetic, or
geometry,
etc., will
be
simply special cases of the propositions of symbolic logic. But other branches than logic will have their own primitive or undefined ideas and
their
own
postulates in terms of these.
We
have used Peano
Formulaire
as
an
illustration of this
the characters of
method, although the Formulaire has, to an extent, the procedure to be mentioned next. Second, there is
the hierarchic method, or the
method
of complete analysis, exemplified
by
is first
Here the calculus
of propositions (or implications)
developed, because
by
its
postulates
and theorems
all
all
the proofs of
other branches are to be stated.
And, further,
i.
the terms and relations
of other branches are to be so analyzed,
e.,
defined, that
from their
defini
tion and the propositions of the logic alone, without additional primitive
ideas or postulates,
third, there
is
the
generate initially
And, and Royce. This method aims to Kempe an order which is not only general, as is the order of logic,
all
the properties of these terms will follow.

of
method
but inclusive, so that the type of order of various special fields (in as large number and variety as possible) may be derived simply by selection i. e.,
by postulates which determine the class which exhibits this special order as a selection of members of the initially ordered field. 31 For this third method,
other types of order will not necessarily be based
in the only
upon the order
of logic:
derived from a more inclusive order.
good examples which we have of the method, logic is itself The sense in which such a procedure
may
still
Which
be regarded as logistic has been made clear in what precedes. of these methods will, in the end, prove most powerful, no one
can say at present.

developed.
The w hole
T
subject of logistic
is
too
new and un
hierarchic or
But
it,
certain characters of each, indicating their adaptability,
or the lack of
to certain ends, can be pointed out.
The
completely analytic method has a certain imposing quality which right One feels that here, for once, we have got to fully commands attention. the bottom of things. work in which this method is extensively carried Any
out, as
it
is
in Principia Mathematica,
is
certainly
monumental.
Further,
the
method has the advantage

31
of setting forth various branches of the
subject investigated in the order of their logical simplicity.

Professor
And
the step
Royce used to say facetiously that the system S had some of the properties of a junk heap or a New England attic. Almost everything might, be found in it: the ques tion was, how to get these things out.
3G9
from one such division to another based upon it is always such as to make clear the connection between the two. The initial analyses definitions
which make such steps possible

once the
are, indeed, likely to tax
our powers, but
initial analysis is correctly
performed, the theorems concerning the
derived order will be demonstrable by processes which have already become
and even stereotyped. The great disadvantage of this completely analytic method is its great complexity and the consequent tediousness of
familiar
its
application.
It is fairly discouraging to realize that the properties of
cardinal
number
require
some four hundred pages

or
of
prolegomena
in a
symbolism
it is
of great
compactness
for their demonstration.

"scientific"
To
those whose
interests are simply
"mathematical"
in the ordinary sense,
forbidding.
logistic
The simple
method
offers
an obvious short-cut.
It preserves
the notable advantages of logistic in general

of ideographic symbols,
the brevity and precision

of correctness.
and the consequent assurance

it
And
since
it differs
from the non-logistic treatment
in little save the introduction
of the logical symbols,
makes
possible the presentation of the subject in
hand
in the briefest possible form.
When
successful,
lies in
it
achieves the
acme
of succinctness
and
clearness.
Its
shortcoming
the fact that, having
which cannot be accomplished without logistic, it achieves attempted little more than is attained by the ordinary abstract and deductive presenta
little
tion.
For what
it is, it
cannot be improved upon; but those who are inter
ested in the comparison of types of order, or the precise analysis of mathe
matical concepts, will ask for something further. No one knows how far the third method that of
Kempe and Royce
can be carried, or whether the system 2, or some other very inclusive type of order, will be found to contain any large number, or all, of the various
special orders in
it
which we are
interested.
But we can
see that, so far as
works, this method gives a

It
maximum
of useful result with a
minimum
of
complication.
yet
it is
avoids the complexities of the completely analytic method, certain to disclose whatever analogies exist between various systems,
its
by the fact that
has the relations in question, or relations of precisely that type.
terms are allowed to denote ambiguously anything which In another

also,
important respect,
advantage seems to
lie
with this method.
One
;
would hardly care to invent a new geometry by the analytical procedure it is difficult enough to present one whose properties are already familiar. Nor would one be likely to discover the possibility of a new system by
the simple logistic procedure.
25
With
either of these
two methods, we need
370
to
know where we
method
greater.
is
are going, or
we
shall
go nowhere.
By
contrast, the
is
third
that of the pathfinder.
The prospect
of the novel
here
much
of order
The system S may, probably does, contain new continents whose existence we do not even suspect. And some chance trans
formation
may
put
us,
suddenly and unexpectedly, in possession of such

fields.
previously unexplored
The outstanding
its possibilities,
all
difficulty of the
method,
it
apart from our real ignorance of

rely
seems to be that
It
must
set
upon
devices which are not at

it is
obvious.
may
not tax severely
the analytical powers, but
certain to tax the ingenuity.
Having
up, for example, the general order of geometrical points,

loss
one
may
be at a
how
to specify
"lines"
having the properties of Euclidean

is
parallels.
In this respect, the analytic method
superior.
generality without complexity, which the third
But the prospect method seems to offer,
of
is
most
enticing.
have spoken of symbolic logic, logistic and mathematics. It may well be questioned whether the method of logistic does not admit of useful
application beyond the field of mathematics.
We
ment
as
much more
science
is
flexible
Symbolic logic is an instru and more powerful than Aristotelian logic as
more complex than its medieval counterpart. Some of the advantages which might have accrued to alchemy, had the alchemists reduced their speculations to syllogisms, might well accrue to modern sci
modern
ence through the use of symbolic logic.

is
The
use of ideographic symbolism
capable of
making quite the same difference in the case of propositions and

it
reasoning that
It
is
has already
made
in the case of
reported that the early Australian settlers could
numbers and reckoning. buy sheep from the
Bushmen only by holding up

It
against one sheep the coins or trinkets repre
senting the price, then driving off that sheep and repeating the process.
might be reported
of the generality of our thinking that
it is
possible to
get desired conclusions only

off
by holding up one or two propositions, driving the immediate consequences, and then repeating the process. Symbolic
is
logic
capable of working the same transformation in the latter case that
arithmetic does in the former.

credit this
it is,
Those unfamiliar with
logistic
may
not
but upon this point we hesitate to press the analogy.

full
Certain
that for the
benefit that symbolic logic

it,
is
capable of giving,
we
should need to be brought up in

arithmetic.
as
we
are in the simpler processes of
What
the future
may
bring in the widespread use of this
new instrument, one hardly ventures to prophesy. Some of the advantages which would be derived from
the wider use of

logistic in science,
371
one can make out.
The
logistic
method
is
applicable
wherever a body of fact or of theory approaches that completeness and systematic character which belongs to mathematical systems. And by the
use of
it,
the same assurance of correctness which belongs to the mathe
matical portions of scientific subjects
may
be secured for those portions
which are not stateable
in
terms of ordinary mathematics.

of the possible use of logistic in
shall
Dare we make one further suggestion

science?
Since
it
seems to us important, we
in scientific investigation
venture
it,
with
all
due
apologies for our ignorance
and our presumption.
considerable part
is
seems to be played
less certainly
extraneous to the real body of
by imagery which The scientific law.

for the
more or
scientist is
satisfied to accept a certain body of facts

"laws,"
directly or indirectly observed
and hypotheses which, phenomena, need not be questioned. But beyond this, he
directly nor indirectly observed, but serves
ation.
time being at
least,
finds a use for
what
is
neither
somehow
to represent the situ
physicist, for example, will indulge in mechanical
models of the
ether, or mechanical
models of the atom which, however much he

fact.
may hope
of
to verify them, he knows to run beyond established
The value
such imagery is, in part at least, its concreteness. The established relations, simply in terms of mathematics and logic, do not come to possess their full
significance unless they are vested in
deal of
what passes
for
"hypothesis"
something more palpable. A great and "theory" seems to have, in part

if it
at least, this character and this value;

gestiveness" of the concrete,
were not for the greater
"sug-
much
of this
would have no reason
for being.
whoever has worked with the precise and terse formulations of logistic realizes that it is capable of performing some of the offices of concrete
Now
imagery.
Its brevity enables
more
facts to be
"seen"
at once, thought of
And a logistic formulation can be free together, treated as a single thing. from the unwarranted suggestions to which other imagery is liable. Perhaps
a wider use of logistic would help to free science from a considerable body of
not in their logical implications but in their the reader will take this conjecture psychological "suggestiveness". But that for the presentation only for what it is worth. What seems certain is
"hypotheses"
whose value
lies
of a systematic
and
theories,
body of theory, for the comparison of alternative hypotheses and for testing the applicability of theory to observed facts,
its
logistic is
an instrument of such power as to make

of view of
eventual use almost
certain.
Merely from the point
method, the application of
logistic to
372
subjects outside the field of mathematics needs no separate discussion.
For when mathematics

quantity, but as
it is
is
no longer viewed as the science of number and
viewed by Mr. Russell or by anyone who accepts the alternative definition offered in this chapter, then the logistic treatment
any subject becomes mathematics. Mathematics itself ceases to have any peculiar subject matter, and becomes simply a method. Logistic is
of
the universal method for presenting exact science in ideographic symbols.

It is
the
"universal
mathematics" of Leibniz.
Finis
BIBLIOGRAPHY
The
logic
and
it is
following bibliography contains titles of logistic in the strict sense with which
of
all
am
the positive contributions to symbolic acquainted, as well as some, taken
from the bibliographies
of works which I have not been able to get. few of the numerous general discussions of symbolic logic and logistic have also been listed, and certain mathematical books and papers which, though not strictly logistic, are of special interest to the student of that subject. What are considered the most important contributions to symbolic logic, are indicated by an asterisk (*) preceding the title, while certain studies which should be especially helpful
Venn and Peano,
But
not complete.
to students of this book are indicated

are given in bold-faced type.
by a
"dagger"
(f).
Volume numbers
of periodicals
ADAMSON, E.
1897.
The
logical copula
and quantification
of the predicate.
London, Nutt,
ALSTED,
J.
H.
AMATO, V.
La
Logicae systema harmonicum. Herbornae Xassoviorum, 1614. BoU. Matem. teoria dei numeri reali fondata sui concetti del Russell.
(Rome), 9 (1910), pp. 198-205. ANDREOLI, C. Su un nuovo simbolo nell algebra della logica. Atti Soc. progr. Sci. (Rome), 4. Riunione Napoli, dicembre 1910, pp. 741 ff. BAIN, A. Logic: Part I, Deduction. London, Longmans, 1870. BARDILI, C. G. Grundriss der ersten Logik. Stuttgart, Loffland, 1800. BAYNES, T. S. An essay on the new analytic of logical forms. (Hamilton Prize Essay, 1846.) Edinburgh, Sutherland, 1850.
Who
discovered the principle of the quantification of the predicate?
Contemp.
Rev., 22 (1873), pp. 318-23.
BENEKE, E. C.
pp. 12-29.
On
the logical meaning of proper names.
Proc. Aristot. Soc., 3 (1905),
Outline of a new system of logic. London, Hunt, 1827. BERNOUILLI, J. Parallelismus ratlocinii logici et algebraici (1685): Opera (Geneve, Cramer, 1744), 1, p. 214. terms BERNSTEIN, B. A complete set of postulates for the logic of classes expressed in of the operation "exception", and a proof of the independence of a set of postulates
BENTHAM, G.
due to del Re.

*
Univ. Calif. Publ. Math., 1 (1914), pp. 87-96. Trans., Amcr. four independent postulates for Boolian algebras. Math. Soc., 17 (1916), pp. 50-52. A simplification of the Whitehead-Huntington set of postulates for Boolian
set of
algebras.
Amer. Math. Soc. Bull, 22 (1916), pp. 458-59.
BOCHER, M.
BOLZANO, B.
4 vols.
The fundamental conceptions and methods
of
mathematics.
Amcr.
Math. Soc. Bull, 11 (1904), pp. 115-35.

Wissenschaftslehre: Versuch einer ausfiihrlichen Darstellung der Logik. Alois Hafler, Sulzbach, Seidelschen Buchh., 1837. Neu herausgegeben von
*BOOLE, G.
Leipzig, Felix Meiner, 1914. The mathematical analysis of logic.
*_
The
calculus of logic.
On
the theory of
Cambridge, Macmillan, 1847. Camb. and Dublin Math. Jour., 3 (1848), pp. 183-98. on Mitchell s problem of the probabilities, and in particular
distribution of the fixed stars.
Phil. Mag., ser. 4, 1 (1851), pp. 521-30. Further observations on the theory of probabilities. Ibid., ser. 4, 2 (1851),
pp. 96-101.
389
390
*
An investigation of the laws of thought. London, Walton, 1854. Also re printed as vol. 2 of Boole s Collected logical works; ed. by Jourdain, Chicago, Open Court Publ. Co., 1916.
Solution of a question in the theory of probabilities.
(1854), pp. 29-32. Reply to some observations published
Phil. Mag., ser. 4, 7
by Mr. Wilbraham
Ibid., ser. 4,
8 (1854), pp. 87-91. On the conditions by which the solutions of questions in the theory of prob abilities are limited. Ibid., pp. 91-98.
Further observations relating to the theories of probabilities

175-76.
Ibid., pp.
On On On
a general method in the theory of probabilities. Ibid., pp. 431-44. certain propositions in algebra connected with the theory of probabilities.
9 (1855), pp. 165-79. the application of the theory of probabilities to the question of the combina tion of testimonies or judgments. Trans. Roy. Soc. Edin (Keith Prize Essay.) burgh, 21 (1857), pp. 597-653.
Ibid,, ser. 4,
*
On On
the theory of probabilities.
Phil. Trans. Roy. Soc. (London), 152 (1862),

Phil.
ser. 4, 25 (1863), pp. 313-17. Trans. Camb. Phil. Soc., 11 (1868), pp.
pp. 225-52.
the theory of probabilities.
Mag.,
Of propositions numerically
396-411.
definite.
BOOLE, MARY. Symbolical methods of study. London, Kegan, 1884. BOURDILLAT, F. La reforme logique de Hamilton. Paris, Hachette, 1891. BROWN, H. C. The logic of Mr. Russell. Jour, of Phil, Psych, and Sci. Methods, 8
(1911), pp. 85-91.
BRUNSCHWIEG, L. Les etapes de la philosophic mathematique. Paris, Alcan, 1912. BRYANT, MRS. S. On the nature and functions of a complete symbolic language. Mind,
13 (1888), pp. 188-207. The relation of mathematics to general formal logic. 2 (1901), pp. 105-34. Proc. Aristot. Soc.,
BUFFA, P.
Alcune formule
di logica.
Principi di logica.
Period,
Rev. de Mathem. (Turin), 7 (1900), pp. 56-58. Matem. (Livorno), 16 (1909), pp. 295-303, and
17 (1901), pp. 292-300.
BURALI-FORTI, C.
Teoria delle grandezze: Parte
IV
del Formulario.
Riv. di
Matem.
(Turin), 3 (1893). Sulle classi derivate a destra e a sinistra.

I
Atli Accad. Torino, 29 (1894).
numeri negativi. Riv. di Matem., 3 (1893), pp. 138 ff. Logica matematica. Milano, Hoepli, 1894. Exercise de traduction en symboles de logique mathematique.
elementaire (Turin), 1897. Sur les differentes methodes logique pour la definition du Cong. Int. de Phil. (Paris), 1900, 3, pp. 289-307.
Bull, de
Mathem.
Bibl.
du nombre
reel.
Sui simboli di logica matematica. II Pitagora, 1900, pp. 65 ff., 129 ff. Sur 1 egalite et sur 1 introduction des elements derives dans la science. Enseign. mathem., 1901, pp. 246-61.
Gli enti astratti definiti
come
enti relativi ad
un campo
di nozioni.
Rend.
Accad. Lincei (Rome),

(1903), pp. 256-82.
ser. 5, 21:
2 (1912), pp. 677-82.

Atti Accad.
Sulla teoria generale delle grandezze e dei numeri.
Torino, 39
BUSCH, M. Anfangsgriinde der logikalischen Algebra. Tubingen, Cotta, 1768. CANTOR, G. Beitrage zur Begrundung der transfiniten Mengenlehre. Math. Annalen, 46 (1895), pp. 481-512, and 49 (1897), pp. 207-46.
Bibliography
Theory
of
391
translation
transfinite
numbers:
by P. E. B. Jourdain.
Open Court CARROLL, LEWIS [pseud.]

Chicago,
Publ. Co., 1915.
see S. H. HODGSON. Alcune identita. Rev. de Mathem. (Turin), 7 (1900), pp. 58 ff. CASTILLON, G. F. Reflexions sur la logique. Berlin Accul. Mem., 1802. Sur un nouvel algorithme logique. Ibid., 1803. CAYLEY, A. On a question in the theory of probabilities (with discussion by Boole). Phil. Mag., ser. 4, 23 (1862), pp. 352-65. Note on the calculus of logic. Quar. J. of Pure and Appl. Math., 11 (1871),
CASTELLANO, F.
pp. 282-83.
On compound
pp. 99-100.
combinations.
Mem.
Manchester
Lit.
and
Phil. Soc.,
26 (1877),
CIPOLLA, M. Theoria de congruentias intra numeros integros. Rev. de Mathem. (Turin), 8 (1905). Specimen de calculo arithmctico integrale. Ibid. *CLIFFORD, W. K. Types of compound statement involving four classes. Mem. Man chester Lit. and Phil. Soc., 26 (1879), pp. 81-96. COHEN, M. R. The present situation in the philosophy of mathematics. J. of Phil., Psych. &c., 8 (1911), pp. 533-46. COUTURAT, L. La logique mathdmatique de M. Peano. Rev. de Metaphysique et de
Morale, 7 (1899), pp. 616-46.
Sur une definition logique du nombre. Ibid., 8 (1900), pp. 23-36. Sur la definition du continuum. Ibid., pp. 157-68. L algebre universelle de M. Whitchead. Ibid., pp. 323-62. La logique de Leibniz, d apres des documents ine"dits, Paris, Alcan, 1901. Histoire de la langue universelle (en collaboration avec M. Leau.)
Hachette, 1903. Sur I utilit6 de la logique algorithmique.
1904.
Paris,
C.-R., II Cong. int. de Phil., Geneve
Les principes des math<5matiques. Rev. de Met. et de Mor., 12 (1904), pp. 19-50, 211-40, 664-98, 810-44, 13 (1905), pp. 224-56. Definitions et demonstrations mathe"matiques. Enseign. Mathem., 1905.
*
f f
Les definitions mathematiques. Ibid., pp. 27 ff. Les principes des mathematiques, avec un appendice sur la philosophic des mathematiques de Kant. Paris, Alcan, 1905. L algebre de la logique. Paris, Gauthier-Villars, 1905. The algebra of logic; translated by L. G. Robinson. Chicago, Open
Court Publ. Co., 1914. Pour la logistique: r6ponse a M.

pp. 208-50.
Poincare".
Rtv. de Met.
et
de Mor., 14 (1906),
La
f
Die Prinzipien der Logik.

pp. 137-201.
logique et la philosophic contemporaine. Ibid., pp. 318-41. Encyc. d. phil. Wiss., hrsg. v. A. Ruge,
1,
Logik,
The
Sci., 1, Logic, pp.
principles of logic;
translated
by B. E. Meyer.
Encyc. of Phil.
136-239.
existentielle.
For
Mind, 22 (1912), pp. 481-523. logistics. Des propositions partlculieres et de la porte*e
Rev. de Met.
et
de
Mor., 21 (1913), pp. 256-59. Sur les rapports logique des concepts et des propositions. pp. 15-58.
Ibid.,
24 (1917),
La logique algorithmique et le calcul de probability. Ibid., pp. 291-313. Ed. DALGARNO, G. Ars signorum, vulgo character universalis et lingua philosophica.
Glasgow, Maitland Club, 1834.
4.
392
DARJES, Ed.
J.
A
G.
2.
Introductio in artem inveniendi seu logicam theoretico-practicam.

1776.
Jenae, 1747. Weg zur Wahrheit.
DAVIS, E.
W.
Some groups
DEDEKIND, R.
Was
in logic. Amer. Math. Soc. Bull, 9 (1903), pp. 346-48. sind und was sollen die Zahlen. Braunschweig, Vierwig. 1888.
und irrationalen Zahlen. Ed. 3. Braunschweig, Vierwig, 1905. Essays on the theory of numbers. (The two preceding translated by W. W. Beman.) Chicago, Open Court Publ. Co., 1909. DE LAGUNA, T. On certain logical paradoxes. Phil. Rev., 25 (1916), pp. 16-27. Opposition and the syllogism. J. of Phil, Psych., &c., 9 (1912), pp. 393-400. DELBOEUF, J. Essai de logique scientifique. Liege, Desoer, 1865. Logique algorithmique. Rev. Phil, 2 (1876), pp. 225-52, 335-55, 545-95.
Stetigkeit
t*DEL RE, A.
Also published separately, Liege et Bruxelles 1877. Lezioni di algebra della logica. Naples, 1907.
;
Sulla indipendenza dei postulati della logica.
Rend. Accad. Sci. (Naples),
ser. 3,
*DE
17 (1911), pp. 450-58. MORGAN, A. Formal logic;
or,
The
calculus of inference, necessary
and probable.
London, Taylor, 1847.

Series of five papers,
1)
"On
the syllogism,
etc."
Trans. Camb. Phil. Soc.
8 (1846), pp. 379-408. 2) 9 (1850), pp. 79-127. 3) 10 (1858), pp. 173-230.

4) Ibid. (1860), 331-*358.
*
5) Ibid. (1863), pp. 428-87. Syllabus of a proposed system of logic. London, Walton, 1860. On indirect demonstration. Phil. Mag., ser. 4, 4 (1852), pp. 435-38.
s
*DINES, L. L. Complete existential theory of Sheffer Amer. Math. Soc. Bull, 21 (1915), pp. 183-88.
postulates for Boolean algebras.
DROBISCH, M. W.
Neue Darstellung der Logik nach
ihren einfachsten Verhaltnissen.
Leipzig, Voss, 1836. DUFUMIER, H. Les theories logico-metaphysiques de Rev. de Met. et de Mor., 17 (1909), pp. 620-53.
MM.
B. Russell et G. E. Moore.
Les tendances de
la
La generalization mathematique. Rev. La philosophie des mathematiques de

(1912), pp. 538-66.
logique contemporaine. Rev. Phil, 74 (1912), pp. 359-78. de Met. et de Mor., 19 (1911), pp. 723-58. Russell et Whitehead. Ibid., 20
MM.
La logique des classes et la theorie des ensembles. Ibid., 23 (1916), pp. 623-31. ELLIS, A. J. On the algebraical analogues of logical relations. Proc. Roy. Soc. (London), 21 (1873), pp. 497-98.
ELLIS, R. L.
Boole
laws of thought.
Report
of British Assoc. for
Advancement
of
Science, 40 (1870), (Sect.) pp. 12-14. EULER, L. Letters on different subjects in natural philosophy addressed to a New York, Harper, 1872. princess.
German
FAGGI, A.
Sulla natura delle propozioni logiche.
FRANKL, W. M.
Quadrat.
Palermo, Reber, 1898. Gegenstandstheoretische Beitrage zur Lehre vom sogenannten logische Arch. f. sys. Phil, 13 (1907), pp. 346-66.
Ein Kalkul fur kategorische (Gewissheits-) Schliisse. Ibid., 19 (1913), pp. 1-8. FRANKLIN, F. A point of logical notation (proposing a numerical notation). Johns Hopkins Univ. Circ., 1 (1879-82), p. 131, Apr. 1881. FRANKLIN, MRS. LADD-. See LADD-FRANKLIN, C. FRECHET, M. Sur quelque points du calcul fonctionnel. Rend. Circ. Matem. di Palermo,
22 (1906), pp. 1-74. Les ensembles abstraits et
le
calcul fonctionnel.
Ibid.,
30 (1910), pp. 1-26.
Bibliography
*FREGE, G.
Begriffsschrift, eine der arithmetischen nachgebildete
393
Formelsprache des
f.
reinen Denkens.
Naturwiss., 1879.
Halle, Nebert, 1879.

Begriffsschrift:
Anwendung der
Vortrag.
Silz.
d.
Jena. Gesell.
Med.
u.
Ueber den Zweck der Begriffsschrift. Ibid., Jan. 1882. Die Grundlagen der Arithmetik, eine logisch-mathematische Untersuchung iiber den Begriff der Zahl. Breslau, Koebner, 1884. Ueber formale Theorien der Arithmetik. Sitz. d. Jena. Gesell. f. Med. u.
Naturwiss., 1885, pp. 94-104.
der Jena. Gesell. f.
Vortrag gehalten in der Sitzung vom 9. Januar. 1891, Med. u. Naturwiss., Jena, 1891. Ueber Sinn und Bedeutung. Zeitsch. f. Phil., 100 (1892), pp. 25-50. Ueber Begriff und Gegenstand. Viertelj.f. wiss. Phil., 16 (1892), pp. 192-205.
Function und Begriff.
Grundgesetze der Arithmetik, begriffsschriftlich abgeleitet.

Pohle, 1893-1903.
vols.
Jena,
Translated portions. Monist, 25 (1915), pp. 481-94, 26 (1916), pp. 182-99, 27 (1917), pp. 114-27. Kritische Beleuchtung einiger Punkte in E. Schroders Vorlesungen iiber die Algebra der Logik. Arch. f. sys. Phil., 1 (1895), pp. 433-56.
Ueber
Klasse
die Begriffsschrift des
Herrn Peano und mein eigene.

6. Juli,
Ber. d. ?nath.
d. Gesell. d.
Wiss. zu Leipzig,
1896.
GEISSLER, K.
Ueber Xotwendigkeit, Wirklichkeit, Moglichkeit und die Grundlagen

Phil, 11 (1905), pp. 1-26. Beitriigen zur Lehre von den Mannigfiiltigkeiten. Phil., 126 (1905), pp. 168-88. D. Essai de dialectique rationelle. Ann. de Mathem. pure et appl., 7,
f. sys.
der Mathematik.
Identitat
Zeitsch. f.
Arch.
und Gleichheit mit
GERGONNE,
pp. 189
J.
ff.
GEYSER,
J.
Logistik
und
Relationslogik.
Phil. Jahrb., 22, 2. Heft.
Grundlagen der Logik und Erkenntnisslehre. Mtinster, Schoningh, 1909. GILMAN, B. I. On propositions and the syllogism, with remarks by C. S. Peirce; and on propositions called spurious. Johns Hopkins Univ. Circ., 1 (1879-82), pp. 240Aug. 1882. Operations in relative number, with applications to the theory of probabilities. In Studies in Logic by members of Johns Hopkins Univ., pp. 107-25. GRASSMANN, H. Die lineale Ausdehnungslehre (1844). Leipzig, Wigand, 1878. *GRASSMANN, R. Die Wissenschaftslehre oder Philosophic. Stettin, Grassmann, 187241, 90.
GREGORY, D.
Issued in parts. F. On the real nature of symbolic algebra. 14 (1838), pp. 208-16.
J.
Trans. Roy. Soc. Edinburgh,

Rev. gen. des
HADAMARD,
La
logistique et la notion de
nombre
entier.
sci.,
16 (1906),
pp. 906-14.
HAGEMANN, G. Logik und Noetik. Freiburg, Herd sche HAHN, O. Zur Axiomatik des logischen Gebietkalkuls.
pp. 345-47.
*
Verlagsbuchh., 1873. Arch. f. sys. Phil., 15 (1909),

ihre
Ueber die Koeffizienten einer logischen Gleichung und

Lehre von den Schlussen.
Ibid.,
Beziehungen zur
f.
*HAHN,
O., und NEURATH, O. (1909), pp. 149-62.
Zum
16 (1910), pp. 149-76. Dualismus in der Logik.

s logical
Arch.
sys. Phil.,
15
HALSTED, G. B.
pp. 134-37.
Professor Jevons s criticism of Boole
system.
Mind, 3 (1878),
Boole s logical method. J. of Specul. Phil, 12 (1878), pp. 81-91. Statement and reduction of the syllogism. Ibid., pp. 418-26.
Algorithmic division in
logic.
Ibid., 13 (1879), pp. 107-12.
27
394
The modern
A
logic.

Ibid.,
Algebras, spaces, logics. De Morgan as logician.
HAMILTON, W.
Rational Geometry. Discussions on philosophy. London, Longmans, 1852. Lectures on logic. Blackwood, Edinburgh, 1860. HAMILTON, W. R. Elements of quaternions. London, Longmans, 1866. HARLEY, R. Memorial of George Boole. Brit. Quart. Rev., 44 (1866), pp. 141-81. Boole s analysis of logic. Rep. Brit. Assoc. Adv. Sci., 36 (1866), (Sect.) pp. 3-6.
17 (1883), pp. 210-13. Pop. Sci. Mo., 17 (1880), pp. 516-22. J. of Specul. Phil, 18 (1884), pp. 1-9. New York, Wiley, 1907.
Boole
laws of thought,
Ibid.,
s
40 (1876),
(Sect.) pp. 14-15.

Lit.
Remarks on Mr. Murphy

(1884), pp. 36-40.
paper.
Mem. Manchester
and
Phil. Soc., 23
HARRIS,
J.
R.
On
the syllogism.
Johns Hopkins Univ.
Circ., 3 (1884), pp. 130-31.
HAUBER, K. F. Scholae logico-mathematicae. Stuttgardiae, 1829. HEATH, A. E. The geometrical analysis of Grassmann and its connection with Leibniz
characteristic.
Monist, 27 (1917), pp. 36-56.
HERIGONE, P.
Cursus mathematicus, nova, brevi, et clara methodo demonstratus per notas reales et universales, citra usum cuiuscunque idiomatis, intellectu faciles. (In both Latin and French.) Paris, 1644. HERMANT, P., et VAN DE VAELE, A. Les principales theories de la logique contemporaine.
Paris, Alcan, 1909.
HICKS, L. E.
and adjacent space. Mind, n. s. 21 (1912), pp. 410-15. a valid inference? J. Phil, Psych., &c., 9 (1912), pp. 65-70. A rejoinder. Mind, n. s. 23 (1914), pp. 96-98. Is inversion a valid inference?
Euler
s circles
Is inversion
HILBERT, D. Grundlagen der Geometric. Leipzig, Teubner, 1899. HILDEBRANDT, T. H. A contribution to the foundations of Frechet s calcul functionnel. Amer. Jour. Math., 34 (1912), pp. 237-390. HILLEBRAND, F. Die neuen Theorien der kategorischen Schliisse. Wien, Holder, 1891.
HOCKING, W.
Two
G.
extensions of the use of graphs in elementary logic.

of logic by Lewis Carroll. London, Macmillan, 1896.
Univ. Calif.
Publ. Philos., 2 (1909), pp. 31-44.
HODGSON,
S.
The game
logic.
London, Macmillan, 1887.
Symbolic
Exacte Darstellung aller Urteile und Schltisse. Arch. f. sys. Phil., See Nachtrag, Ibid., 12 (1906), pp. 55-58. Sprachliche Logik und Mathematik. Arch. f. sys. Phil, 19 (1913), pp. 43-49. *HOLLAND, G. J. VON. Abhandlung liber die Mathematik, die allgemeine Zeichenkunst,
HOFFMANN,
J. J.
11 (1905), pp. 317-22.
und
die Verschiedenheit der Rechnungsarten. Tubingen, 1764. In Deutscher Gelehrter Briefwechsel, 1 (1781). Briefe an J. H. Lambert.
HOMANS, J. HONTHEIM,
Rev. Neo-Schol., 9, no. 2. logische Algorithmus in seinem Wesen, in seiner in seiner philosophischer Bedeutung. Berlin, Dames, 1895.
La
logique algorithmique.
J.
Der
Anwendung und
Amer. Math. Soc.
HOSKINS, L. M.
General algebraic solutions in the logic of classes.
Bull, 15 (1908), pp. 84-85.
HUGHLINGS, I. P. The logic of names: An introduction to Boole s laws of thought. London, Walton, 1869. HUNTINGTON, E. V. Simplified definition of a group. Amer. Math. Soc. Bull., 8 (1902),
pp. 296-300. t*-
second definition of a group. Ibid., pp. 388-91. Sets of independent postulates for the algebra of logic. Soc., 5 (1904), pp. 288-309.
Trans. Amer. Math.
set of postulates for real algebra.
Ibid.,
The continuum
by the Harvard
as a type of order. Univ. Press.
6 (1905), pp. 17-41. Reprinted from Annals of Math. (1905)
Bibliography
395
Proc. Nat. Acad. Sci.,
set of
independent postulates for cyclic order.
(1916), no. 11.
Complete existential theory of the postulates for serial order. Amer. Math. Soc. Bull., 23 (1917), pp. 276-80. Complete existential theory of the postulates for well ordered sets. Ibid., pp.
280-82.
HUNTINGTON, E. V., AND KLINE, J. R. Sets of independent postulates for betweenness. Trans. Amer. Math. Soc., 18 (1917), pp. 301-25. HUSSERL, E. G. Recension, Schroders Vorlesungen iiber die Algebra der Logik. Gottingische Gelehrte Anzeige, 1891, pp. 243-78.
Philosophic
der
Arithmetik,
psychologische
und
logische
Untersuchungen.
15 (1891),
Halle, Pfeffer, 1891.
Der Folgerungscalcul und
die Inhaltslogik.
.
Viertelj. f. iviss. Phil.,

.,
pp. 167-89, 351-56. See also, Erwiderung Ibid., 17 (1893), pp. 504-11.
von A. H. Voigt,
also Antwort,
A. Voigts
"elementare Logik"
und meine Darstellung zur Logik des
logischen
Calculs, Ibid., 17 (1893), pp. 111-120.
JAGER, J. N. *JEVONS, W.
Handbuch der
S.
Pure
logic,
Logik. Wien, Heubner, 1839. or The logic of quality apart from quantity.
London,
Stanford, 1864. The substitution of similars, the true principle of reasoning. millan, 1869.
London, Mac-
On the mechanical performance of logical inference. Phil. Trans. Roy. Soc. (London), 160 (1870), pp. 497-518. On a general system of numerically definite reasoning. Mem. Manchester Lit. and Phil Soc., 24 (1870), pp. 330-52. On the inverse, or inductive, logical problem. Ibid., 25 (1871), pp. 119-30. Who discovered the quantification of the predicate? Contemp. Rev., 21 (1873),
pp. 821-24.
The
*
principles of science, a treatise on logic
and
scientific
method.
Ed.
3.
London, Macmillan, 1879.

Studies in deductive logic. London, Macmillan, 1880. Elementary lessons in logic. New ed. London, Macmillan, 1904. * JOHNSON, W. E. The logical calculus. Mind, n. s. 1 (1892), pp. 3-30, 235-50, 340-57. Sur la th6orie des Equations logique. In Bibl. Cong. int. de Phil., Paris, 1900,
3,
pp. 185-99.
JONES, E. E. C.
n.
s.
Mr. Russell
u
objections to Frege s analysis of propositions.
Mind,
19 (1910), pp. 379-87.
A new law of thought" and its implications. Ibid., 20 (1911), pp. 41-53. JOURDAIN, P. E. B. De infinite in mathematica. Rev. de Mathem. (Turin), 8 (1905),
pp. 121 ff. Transfinite
pp. 93-118.
f
numbers and the
principles of mathematics.
Monist, 20 (1910),
of
Development
matics.
of theories of mathematical logic
and the principles
mathe
Quar. J. of pure and appl. Math., 41 (1910), pp. 324-52, 43 (1912), pp. 219-314, 44 (1912), pp. 113-25. Some modern advances in logic. Monist, 21 (1911), pp. 564-66. The philosophy of Mr. B*rtr*nd R*ss*ll. Ibid., 21 (1911), pp. 481-508, 26
(1916), pp. 24-62.
Mr. Bertrand Russell

(1912), pp. 149-58.
s first
work on the
Ibid.,
principles of mathematics.
Ibid.,
22
The
KAULICH, W.
logical
work
of Leibniz.
26 (1916), pp. 504-23.
Handbuch der
Logik.
Prag,
Lehmann,
1869.
396
KEMPE, A. B.
A Sumy
of Symbolic Logic
Memoir on the theory of mathematical form. Phil. Trans. Roy. Soc, 177 (1886), pp. 1-90. See also, Note, Proc. Royal Soc. (London), 42 (London), (1887), pp. 193-96. On the relation between the logical theory of classes and the geometrical theory Proc. London Math. Soc., 21 (1890), pp. 147-82. of points. The subject matter of exact thought. Nature, 43 (1890), pp. 156-62.
KEYSER, C.
Some outstanding problems

thesis of
for philosophy.
J. of Phil, Psych., &c., 2
(1905), pp. 207-13.
The
KIRGHER, A.
KONIG,
J.
modern logistic. Science, 30 (1909), pp. 949-63. Ars magna sciendi, in XII libros digesta. Amsterdam, 1669. Neue Grundlagen der Logik, Arithmetik und Mengenlehre. Hrsg.
v.
Konig.
Leipzig, Veit, 1914.
44 (1894), pp. 156-57. KORSELT, A. Bemerkung zur Algebra der Logik. Math. Annalen, Die Logik der Geometrie. Jahresber. d. Math. Ver. (Leipzig), 17 (1908), pp.
98-112.
Was
KOYRE, A.
722-24.
ist
Sur
Mathematik? Arch. f. Math. les nombres de M. Russell.
(Leipzig), 21 (1913), pp. 371-73. Rev. de Met. et de Mor., 20 (1912), pp.
KRAUSE, K. C. F.
Abriss des Systemes des Logik als philosophische Wissenschaft.

Viertelj. f. iciss.
Gottingen, Dieteiich, 1828. Bolzanos. KREIBIG, J. K. Ueber die Paradoxen in der Logik
Phil,
28 (1904), pp. 375-92.

1905. KRIENELKE, K. J. H. Lambert s Philosophic der Mathematik. Berlin, Mayer, Leibnizens Logik, nach den Quellen dargestellt. Prag, Tempsky, 1857. F. B. KVET,
LADD, CHRISTINE. See Ladd-Franklin. f*LADD-FRANKLiN, C. On the algebra of

University, pp. 17-71.
logic.
In Studies in Logic by Johns Hopkins
On some
543-67.
characteristics of symbolic logic.
Amer.
J. of Psych., 2 (1889), pp.
Some proposed reforms

Dr. Hillebrand
in
common
logic.
Mind, 15
(1890), pp. 75-85.
The
(1892), pp. 527-30. reduction to absurdity of the ordinary treatment of the syllogism.
s syllogistic.
Mind,
n.
s. 1
Science,
13 (1901), pp. 574-76. Implication and existence in logic.
Phil. Rev., 21 (1912), pp. 641-65.
The antilogism an emendation. J. of Phil, Psych., &c., 10 (1913), pp. 49-50. Articles on Syllogism and Proposition in Baldwin s Diet, of Phil, and Psych. 256-88. LALANDE, L. Le mouvement logique. Rev. Phil., 63 (1907), pp. 1764. LAMBERT, J. H. Neues Organon. 2 vols. Leipzig, Wendler, * D e universaliori calculi idea, disquisitio. Nova Ada Eruditorum, 1765, pp.
441
ff.
Riga, Hartknoch, 1771. v. Logische und philosophische Abhandlungen; hrsg. 1782-87. Berlin, Bernouilli,
Anlage zur Architectonic.
J.
Bernoulli.
2 vols.
*J.
H. Lamberts Deutscher Gelehrter Briefwechsel;
hrsg. v. Bernouilli.
4 vols.
Berlin,
Bernouilli, 1781-84. 1877. LANGE, F. A. Logische Studien. Iserlohn, Baedeker, Inventum novi quadrati logici universalis in trianguli quoque I. C. LANGE,
formam
commode
*LEIBNIZ, G.
*
redact!.
Giessen,
tiller,
1714.
Neucleus logicae Weisianae.
W.
Giessen, Miiller, 1712. Philosophische Schriften, hrsg. v. C. I. Gerhardt.

1887.
. . .
7 vols.
Berlin,
Weidmannsche Buchh.,
Opuscules et fragments inedits royale de Hanovre, par L. Couturat.
extraits de manuscrits de la Bibliotheque
Paris, Alcan, 1903.
Bibliography
Leibnitiana, Elementa philosophiae arcanae, Kazan, Imp. Univ., 1913.
397
rerum.
De summa
Ed. Ivan Jagodinsky.
LEWIS, C.
Implication and the algebra of logic. Mind, n. s. 21 (1912), pp. 522-31. Interesting theorems in symbolic logic. J. of Phil., Psych., &c., 10 (1913), pp. 239-42. A new algebra of implications and some consequences. Ibid., pp. 428-38.
I.
The calculus of strict implication. Mind, n. The matrix algebra for implications. J. of
589-600.
s.
23 (1914), pp. 240-47.
Phil., Psych., &c., 11 (1914), pp.
A too-brief set of postulates for the algebra of logic.

Types
his Sixtieth
Ibid., 12 (1915), pp.
523-25.
of order
and the system.. .(Among Papers in honor Phil Rev., pp. 407-19, May 1916. Birthday.}
of Josiah Royce
on
The issues concerning material implication. J. of Phil, Psych., &c., 14 (1917), 350-56. pp. Rev. Phil., LIARD, L. Un nouveau systeme de logique formelle, M. Stanley Jevons. 3 (1877), pp. 277-93. La logique algebrique de Boole. Ibid., 4 (1877), pp. 285-317.
f
Les logiciens anglais contemporains. Ed. 5. Paris, Alcan, 1907. LICHTENFELS, J. Lehrbuch der Logik. Wien, Heubner, 1842. Mathematik. Phil. Studien, 14 LIPPS, G. F. Untersuchung iiber die Grundlagen der
(1898), pp. 157-241.
[Anonymous.] Glasgow, 1764. Logicae Compendium. Bull des sci. Mathem., 18 (1894), LORIA, G. La logique mathe matique avant Leibniz. 107-12. pp. eine logische Unter LOURIE, S. Die Prinzipien der Wahrscheinlichkeitsrechnung,
LOWENHEIM,
*
1910. suchung des disjunktiven Urteils. Tubingen, Mohr, Ueber das Auflosungsproblem in logischen Klassenkalkul. L.
Silzber.
math. Gescll. (Berlin), 7 (1908), pp. 89-94. Auflosung von Gleichungen im logischen Gebietekalkul.
(1910), pp. 169-207.
Math. Annalen, 68
Ueber Transformationen im Gebietkalkul. /tad., 73 (1913), pp. 245-72. Transfor Potenzen im Relativkalkul und Potenzen allgemeiner endlicher
mationen.
Sitzber. math. Gestll. (Berlin), 12 (1913), pp. 65-71.
Ueber eine Erweiterung des Gebietekalkuls, welche auch die gewohnliche

21 (1915), pp. 137-48. Algebra umfasst. Arch. f. sys. Phil., Aus der Algebra der Relative, nach dem 3. Bd. von E. Schroders Vorlesungen J. Jahresber. d. Deutsch. Math.-Ver., 13 (1904), pp. 73-1 iiber die Algebra der Logik. 1796. MAASS, J. G. E. Grundriss der reinen Mathematik. Halle, Ruff, der Logik. Halle, Ruff, 1807. Grundriss with an application to mathen MACCOLL, H. Symbolical or abbreviated language Math. Questions, 28 (1877), pp. 20-23, 100. probability. * The calculus of equivalent statements. (Seven papers and a note.) 16-2 London Math. Soc., 9 (1877), pp. 9-20; ibid., pp. 177-86; 10 (1878), pp. 28 (1897), pp. 555-79; 29 ( 11 (1880), pp. 113-21; 28 (1896), pp. 156-83;
LiiROTH,
330-32. pp. 98-109; Note, 30 (1898), pp.

(1897),
(1903),
On
the growth and use of symbolical language.
Mem.
Manchester
Lit.
and
Phil. Soc., 27 (1881), pp. 225^48.
Implicational and equational logic.
Question
for logicians.
Mind,
n.
s.
Phil. Mag., ser. 5, 11 ( 9 (1900), p. 144, 288, 432.)
>D,
PP- 4
398
A
La
Logique tabulaire. Rev. de Met. et de MOT., 10 (1902), pp. 213-17. logique symbolique et ses applications. Bibl. du Cong. Int. de
Phil., Paris
1900, 3, pp. 135-83.
p.
La logique symbolique. Enseign. Mathem. (Paris), 6 (1904), pp. 372-76. Symbolic Logic and its applications. London, Longmans, 1906. Symbolic logic: A reply. Mind, n. s. 16 (1907), pp. 470-73. Mind, n. s. 19 (1910), pp. 186-99, 337-55. Linguistic misunderstanding. Problems, in Math. Questions 29 (1878), pp. 66-68; 33 (1880), p. 113; 35 (1881), 29; 36 (1881), pp. 50, 55, 72.
*MACFARLANE, A.
Principles of the algebra of logic. Edinburgh, Douglas, 1879. a calculus of relationship (algebra of relationship). Proc. Roy. Soc. Edin burgh, 10 (1879), pp. 224-32; 11 (1880), pp. 5-13; 11 (1881), pp. 162-73. An analysis of relationship. Phil. Mag., ser. 5, 11 (1881), pp. 436-46.
On
18 pp. London, Analysis of relationships, of consanguinity and affinity. Harrison, 1882. Analysis of relationships applied to various problems. /. of Anthrop. Inst.
(London), 1882.
The
logical spectrum.
Phil.
Mag.,
ser. 5,
Principles of the algebra of physics.
19 (1885), pp. 286-90. Proc. Amer. Assoc. Adv. Sci., 40 (1891),

Trans. Texas Acad. of Sci., Feb.
pp. 65-117. On exact analysis as the basis of language. 1892, pp. 5-10.
The fundamental principles of algebra. Proc. Amer. Assoc. Adv. Sci., 48 (1899), pp. 21-51. Les idees et principes du calcul geometrique. Bibl. du Cong. Int. de Phil., Paris 1900, 3, pp. 405-23.
Problems, in Math. Questions 32 (1879), pp. 17-19, 74-77, 90; 35 (1881), pp. 10304; 36 (1881), pp. 27-28, 78-81. MAIMON, S. Versuch einer neuen Logik. Berlin, Felisch, 1798.
MALLY, E.
Gegenstandstheorie
Zur Gegenstandstheorie des Messens. In Meinongs Untersuchungen zur und Psychologic. Leipzig, Earth, 1904.
Erganzungsheft
Gegenstandstheoretische Grundlagen der Logik und Logistik. zu 148. Bd. Zeitsch.f. Phil. (1912).
MANNOURY, G.
Haarlem,
Methodologisches und Philosophisches zur Elementar-Mathematik.
Visser, 1910.
MARQUAND, A.
term
A machine for producing syllogistic variations, and A note on an eightmachine. In Studies in Logic by Johns Hopkins Univ., pp. 12-16. A new logical machine. Proc. Amer. Acad. Arts and Sci., 21 (1885), pp. 303-07. MEDOLAGHI, P. La logica matematica ed il calcolo delle probabilita. Boll. Ass. Hal.
logical
incr. scienza attuari (Milan), 18 (1907), pp. 20-40.
Intorno
200.
al calcolo
formale delle probibilita.
Gior.
Matem., 45 (1907), pp. 192-
MICH, J. Grundriss der Logik. Wien, Buchholz, 1871. MINEO, C. Logica e matematica. Riv. Ital. di Filos., 3 (1911), pp. 49-70. MITCHELL, O. H. On the algebra of logic. Johns Hopkins Univ. Circ., 1 (1879-82),
p. 208,
May
a
1882.
of logic.
On
22-106.
new algebra
In Studies in Logic by Johns Hopkins Univ., pp.

. . .
MONTAGUE, W.
MOORE, E. H.
Note on Venn s review of Studies in Logic. Mind, 9 (1884), pp. 321-22. P. The meaning of identity, similarity and non-entity: a criticism of
s logical
Mr. Russell
pp. 485-92.
J. of Phil., Psych., &c., 3 (1906), pp. 127-31. puzzles. definition of abstract groups. Trans. Amer. Math. Soc., 3 (1902),
Bibliography
On
402-24.
the foundations of mathematics.
Amer. Math. Soc.
Bull.,
9 (1903), pp.
Introduction to a form of general analysis. (1910}] Yale Univ. Press, 1910; pp. 1-150.
New Haven Math.
Colloquium
MULLER, E.
See also Schroder, E.
Ueber die Algebra der Logik: I. Die Grundlagen des Gebietekalkuls; II. Da3 Eliminationsproblem und die Syllogistik. Programm des Gymnasiums Tauberbischofsheim, 1900. Ueber die Algebra der Logik und iiber die hinterlassene algebraische-logische Schriften von Ernst Schroder. Ber. int. Kong. Phil, Heidelberg 1908, 3, pp. 686-91.
MURPHY,
1875.
J. J.
Relation of logic to language.
Belfast Nat. Hist,
and
Phil. Soc., Feb.
Fundamental logic. Mind, 2 (1877), pp. 47-55. On an extension of ordinary logic, connecting it with the logic of relatives. Mem. Manchester Lit. and Phil. Soc., 27 (1879), pp. 90-101. On the addition and multiplication of logical relatives. Ibid., 27 (1881), pp.
201-24.
On the
transformations of a logical proposition containing a single relative term.
NAGY,
Ibid., 28 (1882), pp. 132-38. A new system of logical notation. Ibid., 32 (1888), pp. 22-32. A new symbolic treatment of the old logic. Ibid., 34 (1891), pp. 201-15. A. Sulla rappresentazione grafica delle quantita logiche. Rend. Accad. Lincei,
6 (1890), pp. 373-78.
Fondamenti
I
Gior. Matem., 28 (1890), pp. 1-35. del calcolo logico. teoremi funzionali nel calcolo logico. Riv. di Matem. (Turin), 2 (1892), pp.
177-79.
"
Principi di logica, esposti secondo 1892.
le
dottrine moderne.
Torino, Loescher,
Ueber Bcziehungen zwischen logischen Grossen.
Monatshefte f. Math. u. Phys.,
4 (1893), pp. 147-53. Ueber das Jevons-Clifford sche Problem. Ibid., 5 (1894), pp. 331-45. de Mathem. (Turin), 7 (190 NASSO, M. Alcuni teoremi di arithmetica. Rev.
42-55.
NATORP,
P.
Zu den
Mathematil logischen Grundlagen der neueren
Phil, 7 (1901), pp. 177-209, 372-438. NEURATH, O. See also Hahn and Neurath.
Produkte a b. Eindeutigkeit und Kommutativitat des logischen 342-44. Phil., 15 (1909), pp. Ernst Schroders Beweis des 12. Theorems. Fur die identischen Oper Ibid., pp. 104-06. gilt das "Kommutationsgesetz"? Gleichheit. Ibid., 16 (1 Definitionsgleichheit und symbolische elementare della logica. II metodo delle matematiche e 1 insegnamento NICOLI, P. F. Riv. Ital. di Filos., 11 (1908), pp. 355-63.
NEWLIN, W.
PADOA, A.
1898
J.
A new
logical diagram.
la logique
Conferences sur
mathematique.
J. of Phil., Psych., &c., 3 (1 Univ. nouvelli
Algebra elementare logicamento esposta. di Pavia, 1899, pp. 35 ff

.
Conferenze tenuto nella R. Univ.

i
d une Essai d une theorie algebrique des nombres entiers, precede Int a une theorie deductive quelconque. Bibl. du Cong. logique
^Theorie des nombres

1-41.
entiers absolus.
Rev. de
Mathem. (Turin), 7
(1900), pp.
400
A
di

of
"Essai
Numeri interi relativi. (Translation Mathem. (Turin), 7 (1900), pp. 73-84.
d une theorie
".)
Rev.
di
Logica matematica e matematica elementare. Atti del II. Congr. dei Professori Matem., Livorno 1902, pp. 186-200. Le probleme nr. 2 de M. David Hilbert. Enseign. Mathem., 5 (1903), pp. 85-91.
Che
cos e una relazione? Atti Accad. Sci. (Turin), 41 (1905-06), pp. 1084-92. Ideografia logica. Ateneo Veneto, 1906. Sul teorema Cantor-Bernstein-Peano. Period. Matem. (Livorno), 22 (1906),
pp. 23-28.
f
et
La logique deductive dans sa derniere phase de developpement. Rev. de Met. de Mor., 19 (1911), pp. 828-83; 20 (1912), pp. 48-67, 207-31. La logique deductive dans sa derniere phase de developpement. Paris, Gau,
thier-Villars, 1912.
Des consequences d un changement d
idees primitives dans
une theorie deductive
["*-
quelconque. Rev. de Met. et de Mor., 24 (1917), pp. 315-25. PASCH, M. Vorlesungen iiber neuere Geometrie. Leipzig, Teubner, 1882. PEACOCK, G. Report on recent progress and present state of certain branches of analysis. Rep. Brit. Assoc. Adv. Sci., 3 (1883), pp. 185-352. On symbolical algebra and its applications to the geometry of position. (Vol. 2 of A treatise on algebra.) Cambridge, Deighton, 1845. PEANO, G. Calcolo geometrico secondo 1 Ausdehnungslehre di H. Grassmann, preceduto dalle operazioni della logica deduttiva Torino, Bocca, 1888. Arithmetices principia, nova methodo exposita. Torino, Bocca, 1889. Sur la definition de la limite d une function; exercise de logique mathematique. Amer. J. of Math., 17 (1894), pp. 37-68. Sui fondamenti della geometria. Riv. di Matem. (Turin), 4 (1894), pp. 51-90. Notations de logique mathematique; introduction au Formulaire de Mathe
Torino, Bocca, 1894. Saggio di calcolo geometrico. Atti Accad. Torino, 31 (1896), pp. 3-26. Studii in logica matematica. Ibid., 32 (1897), pp. 3-21. Les definitions mathematique. Bibl. du Cong. Int. de Phil., Paris 1900, 3, pp. 279-88. Arithmetica generale e algebra elementare. Torino, Paravia, 1902. La geometria basata sulle idee di punto e di distanza. Atti Acad. Torino, 38
matique.
(1902), pp. 6-10.
Super theorema de Cantor-Bernstein. Rev. de Mathem. (Turin), 8 (1906). *PEANO, G., redacteur. Formulaire de Mathematiques. Tome I, 1895; II, 1897;
IV, 1902; V, 1908. Torino, Bocca, 1895-1908. In addition to Peano, the chief contributor, the other contributors were MM. Arbicone, Bellazzi, Boggio, Burali-Forti, Cantoni, Castellano, Ciamberlini, Giudice,
III, 1901;
Fano, Padoa, Vacca, Vivianti, et new, revised and enlarged editions. PEIRCE, B. PEIRCE, C.
al.
The
Tome V is written in
successive volumes are comparable to the proposed international
language, Interlingua. Linear associative algebras. New edition, with notes and addenda by C. S. Peirce. Amer. J. of Math., 4 (1881), pp. 97-229.
S.
Complete bibliography, by M. R. Cohen, see J.
of Phil, Psych., &c.,
13 (1916), pp. 733-37.
*PEIRCE, C. S. On an improvement in Boole Arts and Sci., 7 (1867), pp. 250-61.
s calculus of logic.
Proc. Amer. Acad.
the logic of mathematics. Ibid., pp. 402-12. Description of a notation for the logic of relatives. and Sci., 9 (1870), pp. 317-78.
Upon
Mem. Amer. Acad.

Proc.
Arts
On the application of logical analysis to multiple algebra. Arts and Sci., 10 (1875), pp. 392-94.
Amer. Acad.
Bibliography
Note on Grassmann
s "Calculus of Extension".
401
Ibid., 13 (1877), pp. 115-16.
the algebra of relatives. cation of the publisher.
On On On
the algebra of logic. the logic of number.
Amer. J. of Math., 3 (1880), pp. 15-57.

Ibid.,
4 (1881), pp. 85-95.

Separately published without indi
Circ.,
6 pp., 1882.
On
p. 179.
the relative forms of quaternions.
Johns Hopkins Univ.
Feb. 1882,
A
t*
*
theory of probable inference.
In Studies in Logic by Johns Hopkins Univ.,
pp. 126-82.
Ibid., pp. 187-203. logic of relatives. the algebra of logic; a contribution to the philosophy of notation. Amer. J. of Math., 7 (1885), pp. 180-202. The critic of arguments: Exact thinking, Open Court, 6 (1892), pp. 3391-94;
The
On
introduced to relatives, Ibid., pp. 3415-19. s Vorlesungen, Bd. Ill: The regenerated logic, Monist, 7 19-40; The logic of relatives, Ibid., (1897), pp. 161-217. (1896), pp. PESLOUAN, C. L. DE. Sur les fondements de I arithmetique et autres questions connexes. Rev. Phil, 10 (1907), pp. 372-97, 489-509, 568-92, 11 (1907), pp. 182-206.
The reader
is
Recension, Schroder
PIERI,
M.
Les systemes logique et la logistique. Paris, Riviere, 1909. Sur la geometric envisaged comme un systeme purement logique.
Rev. de Met.
et
Bibl.
du
Cong. Int. de Phil., Paris 1900, 3, pp. 367-404. Sur la compatibility des axiomes de 1 arithmetique. 14 (1906), pp. 196-207.
de Mor.,
Sopra una definizione aritmetica degli

1906.
irrazionali.
Accad. Gioenia di Catania,
Uno sguardo
al
nuovo
Discorso letto inaugurandosi

18
indirizzo logico-matematico delle scienza deduttive. anno accademico 1906-07 nella R. Univ.
PLOUCQUET, G. Fundamenta philosophiae speculativae. Tubingen, Cotta, des Herrn Untersuchung und Abanderung der logicalischen Constructionen
Professor Lambert.
di^Catania. 1759.
Tubingen, Cotta, 1765.

Calcul Herrn Professor Ploucquet s
<
Sammlung der Schriften welche den logischen

betreffen.
Frankfurt, 1766.
philosophiae
"
Institutiones philosophiae theoreticae sive de arte cogitandi. Elementa contemplativae, sive de scientia ratiocinandi.
gart, 1778.
POINCARE, H.
Les mathe*matiques et la logique. Rev. de Mil. pp. 816-35, 14 (1906), pp. 17-34, 294-317. A propos de la logistique. Ibid., 14 (1906), pp. 866-68. La logistique de 1 infini. Ibid., 17 (1909), pp. 461-82. The new logics. Monist, 22 (1912), pp. 243-56. The latest effort of the logistlcians. Ibid., pp. 524-3
1878.
et
de Mor.,
15),
POKORNY, I. Neuer Grundriss der Logik. Wien, Graeser, Russian.) PORETSKY, P. On the subject of mathematical-logic. (In
Math. Soc. Kazan, 1881.
Ibid., li (In Russian.) First principles of mathematical-logic. and inverse method of Logical equations, methods of solution,
logic.
Bull.
mathemat
*_
(In Russian.) i Solution of the general problem of the theory of probabilities by Ibid., 1887. (In Russian.) mathematical-logic. U On the monograph of M. Volcov, Le calcul logique. (In Russian.) Ibid., 5-8. AJO, iwi vie racines ^n i^gjivjv.^ La loi de t.a\ji.^.a en logique. Rev. de Mathtm. (Turin), 6 (1896), pp. Bu s lois fondamentales de la thSorie des elite logiques.
.
Ibid., 1884.
Soc.
Sept Kazan, 1898.
402
Expose
A
e"le"mentaire
Rev. de Met.
*
de la thSorie des Sgalites logique & deux termes a and 6. de Mor., 8 (1900), pp. 169-88. Quelques lois ulterieures de la theorie des egalite"s logiques. Butt. Phys.-Math.
et
Kazan, 1900-01. Theorie des egalites logiques a trois termes, de Phil, Paris 1900, 3, pp. 201-33.
Soc.
a, b,
and
c.
Bibl.
du Cong.
int.
* are reprinted and issued separately by the Imperial Univ. of Kazan.) PRANTL, K. v. Geschichte der Logik im Abendlande. 4 vols. Leipzig, Hirzel, 1855-70. PROCHAZKA, I. J. Gesetzbuch fur das Denken. Wien, Gerold, 1842. REIMARUS, H. S. Vernunftlehre. Hamburg, Bohn, 1790.
Theorie des non-6galites logiques. Bull. Phys.-Math. Soc. Kazan, 1903. Theorie conjointe des egalites et des non-4galites logiques. Ibid., 1908. (The four most important papers by Poretsky those indicated by
REYES Y PR6SPER, V.
en
la logica simbolica.
Matematica americana y su influencia Matem. (Zaragoza), 1 (1891), no. 12. Ernesto Schroeder: Sus merecimientos ante la 16gica, su propaganda logico:
Christina Ladd-Franklin
Progr.
REYMOND, A.
infini.
mathematica, sus obras. Ibid., 2 (1892), no. 14. Logique et mathematiques: Essai historique critique sur
Paris, Alcan, 1908.
le
nombre
Note sur le theoreme d existence des nombres entiers, et sur la definition logistique du zero. Rev. de Met. et de Mor., 17 (1909), pp. 237-39. RIEBER, C. H. Is inversion a valid inference? Mind. n. s. 22 (1913), pp. 258-59. RIEHL, A. Die englische Logik der Gegenwart, mit besonderer Beriicksichtigung von W. S. Jevons. Viertelj. f. wiss. Phil, 1 (1877), pp. 150-80. RIQUIER, C. Des axiomes mathematiques. Rev. de Met. et de Mor., 3 (1895), pp. 269-84.
Ross, G. R. F.
Mind, *ROYCE, J.
Inversion and the diagrammatic representation of negative terms. 22 (1913), pp. 254-57. The relations of the principles of logic to the foundations of geometry. Trans. Amer. Math. Soc., 6 (1905), pp. 353-415. Prinzipien der Logik. In Ruges Encyc. d. philosoph. Wiss., Bd. I, Loqik, pp. 61-136.
n.
s.
An
617-33.
Principles of Logic; translated by B. E. Meyer, pp. 67-135. j_j. .LVA^y ci ^ .LOU. MJ I- JK extension of the algebra of logic. /. of Phil, Psych., &c., 10 (1913), pp.
i
.!_-/.
RUSSELL, B. A. W.
and Russell. Mind, n. s. 5 (1896), pp. 1-23. On the relations of number and quantity. Mind, n. s. 6 (1897), pp. 326-41. An essay on the foundations of geometry. Cambridge Univ. Press, 1897. Sur la logique des relations. Rev. de Mathem. (Turin), 7 (1901), pp. 115-
(See correction, p. 672.) See also Whitehead
The
logic of
geometry.
148.
On
the notion of order. Mind, n. s. 10 (1901), pp. 30-51. Theorie generate des series bien ordonnees. Rev. de Mathem. (Turin), 8 (1902),
pp. 12-43. t*Principles of mathematics:
I.
Camb. Univ.
Press, 1903.
No more
volumes
published.
On
Russell,
Mind, n. s. 14 (1905), pp. 479-93. See also Notes by MacColl, and Shearman, Ibid., pp. 295, 398, 401, 578. Sur la relation des mathematiques a la logistique (avec une note de M. Whitedenoting.
Rev. de Met.
et
head).
de Mor., 13 (1905), pp. 906-17. Amer. J. of Math., 28 (1906), pp. 159-202. Les paradoxes de la logique. Rev. de Met. et de Mor., 14 (1906), pp. 627-50.
The theory
of implication.
On some
Proc.
difficulties
in the theory of transfinite
numbers and order
types.
London Math.
Soc., ser. 2,
4 (1906), pp. 29-53.
Bibliography
Mathematical
(1908), pp. 222-62.
logic as
in:;
based on the theory of types.
Amer. J.
of Math.,
30
La
the*orie des types logiques. Rev. de Met. et de A/or., 18 (1910), pp. 263-301. importance philosophique de logistique. Ibid., 19 (1911), pp. 281-91. RSponse a M. Koyre". Ibid., 20 (1912), pp. 725-26. The philosophical importance of mathematical logic. Monist, 23 (1913), pp.
481-93.
RUSSELL, F. C.
Hints for the elucidation of Peirce
s logical
work.
Monist, 18 (1908),
pp. 406-15. SCHADEN, E. A. VON.
SCHEFFLER, H.
System der positiven Logik. Erlangen, Palm, 1841. Die Naturgesetze und ihr Zusammenhang mit den Prinzipien der
abstrakten Wissenschaften. Leipzig, Foerster, 1876-80. See especially Part 3. SCHLOTEL, W. Kleine Bausteine zu einem Denkmale. Freiburg i. B., 1876. SCHMIDT, K. Studies in the structure of logical systems: The separation of problems. The deductive system form. Ibid., J. of Phil., Psych., &c., 9 (1912), pp. 197-204.
Postulates. pp. 317-21. 10 (1913), pp. 64-75.
f
Ibid.,
pp.
431-40.
The generating problem.
Ibid.,
Inversion.
Ibid.,
J.
9 (1912), pp. 232-34.
SCHNIPPENKOTTER,
fur die Logik.
Bedeutung der mathematischen Untersuchungen Coutourats
Phil. Jahrb. (Fulda), 23 (1910), pp. 447-68.
SCHRODER, E.
J-*
viele Algorithmen zur Auflosung der Gleichungen. Math. Annalen, 2 (1870), pp. 317-65. Lehrbuch der Arithmetik und Algebra. Leipzig, Teubner, 1873. Ein auf die Einheitswurzeln beziigliches Theorem der Funktionenlehre. Zeitsch. 183-90. f. Math. u. Phys., 22 (1877), pp. Der Operationskreis des Logikkalkuls. Leipzig, Teubner, 1877. Ueber den Operationskreis (Note). Math. Annalen, 12 (1877), pp. 481-84. 25 (1880), HistorRecension, Freges Begriffsschrift. Zeitsch. f. Math. u. Phys.,
Ueber unendliche
isch-literarisch Abt., pp. 81-94.
the most commodious and comprehensive calculus: Exposition of a logical of logic but overlooked by the ancient logi principle, as disclosed by the algebra 411-12. cians. Rep. Brit. Assoc. Adv. Sci., 53 (1883), (Sect.) pp. Ueber das Eliminationsproblem im identischen Kalkul. Tageblatt der Natur-
On
f*
forscherversammlung zu Strassburg, 1885, pp. 353-54. Ueber Algorithmen und Kalkuln. Arch. f. Math. u. Phys., 1887, pp. 225-78. Vorlesungen iiber die Algebra der Logik (exakte Logik), Leipzig, Teubner; Bd. 2. Abth., hrsg. v. E. Miiller, 1. Bd., 1890; 2. Bd. 1. Abth., 1891; 3. Bd., 1895; 2.
1905.
Eine Berichtigung
36 (1890), p. 602.
zum ersten Band meiner Algebra der
Logik.
Ueber das Zeichen. Festrede bei dern Direktoratswechsel an der Technu Hochschule zu Karlsruhe am 22. November 1890. Karlsruhe, 1890. 46 (: Ueber die Algebra der binaren Relative (Note). Math. Annalen,
>95),
pp. 144-58. Ueber Pasigraphie, ihren gegenwartigen Stand und die pasigraphische Beweg Verh. der 1. int. Math. Kong., Zurich 1898. in Italien. On Pasigraphy; its present state and the pasigraphic movemen
Italy;
(See corrigenda, p. 320.) translation. Monist, 9 (1898), pp. 44-62. de Phil., Pans 1900, ^Sur une extension de 1 idde d ordre. Bibl. du Cong. Int.
1.
pp. 235-10. Abriss der Algebra der Logik; hrsg. v. E. iMiiller. Leipzig, Teubner; (The third part not yet published.) 1909; 2. Teil, 1910. Amer. J. of Math., 31 relations. SCHWEITZER, A. R. A theory of geometrical 35 (1913), pp. 37-56. pp. 365-410,
3,
404
Concerning linear protective order. Ibid., 34 (1912), pp. 169-76. Les iddes directrice de la logique gene"tique des mathe matiques. Rev. de Met. et de Mor., 22 (1914), pp. 175-97; also published by Univ. of Chicago Press, 1915. SEGNER, J. A. Specimen logicae universaliter demonstratae. Jena, Crocker, 1740. SEMLER, C. A. Versuch iiber die combinatorische Methode, ein Beitrag zur angewandten Logik und allgemeinen Methodik. Dresden, Walther, 1811.
SETON, J. Dialectica. 1611. fSHAW, J. B. What is Mathematics? Review of Principia Mathematica, vol. I. Amer. Math. Soc. Bull, 18 (1912), pp. 386-411. Logistic and reduction of mathematics to logic. Monist, 26 (1916), pp. 397414.
SHEARMAN, A. T.
n.
s.
Some
controverted points in symbolic logic.
Proc. Aristot. Soc.,
5 (1904-05), pp. 74-105.
of symbolic logic. London, Williams, 1906. views on logical existence. Mind, n. s. 15 (1906), pp. 143-44. Certainties, impossibilities and variables (reply to Mr. MacColl). Mind, 16 (1907), pp. 315-16. Definition in symbolic logic. Mind, 19 (1910), pp. 387-89. The scope of formal logic. London, Univ. Press, 1911.
The development
Mr. MacColl
s
*SHEFFER, H. M.
A set of five independent postulates for Boolean algebras, with appli cation to logical constants. Trans. Amer. Math. Soc., 14 (1913), pp. 481-88. SIGWART, H. C. W. Handbuch zu Vorlesungen iiber die Logik. Ed. 3. Tubingen,
Osiander, 1835.
A syllabus of logic. Cambridge, Deighton, 1839. SPOTTISWOODE, W. Remarks on some recent generalizations of algebra. Proc. London Math. Soc., 4 (1871), pp. 147-64. STAMM, E. On the theory of relations and operations. (In Polish.) Prace mat.-fiz. Warszawa, 21 (1919), pp. 42-54. What is and what should be Mathematics. (In Polish.) Wiad. mat. Warszawa,
SOLLY, T.
14 (1910), pp. 189-96. Principles of the algebra of logic. (In Polish.) Ibid., 15 (1911), pp. 1-87. STERZINGER, O. Logik und Naturphilosophie der Wahrscheinlichkeitslehre; ein umfassender Losungsversuch. Leipzig, Xenien Verlag., 1911.
STOKL, A. Lehrbuch der Logik. 8 Aufl., neuarbeitet v. Dr. G. Wohlmuth. Mainz, Kirchheim, 1905. STOHR, A. Logik im psychologisierender Darstellung. Wien, Deuticke, 1910. STOLZ, O. Vorlesungen iiber allgemeine Arithmetik, nach den neueren Ansichten
bearbeitet.
2 vols.
STRACHEY, O.
Mr. Russell and some recent
Leipzig, Teubner, 1885-86. criticisms of his views.
Mind,
n.
s.
24
(1915), pp. 16-28.
THOMSON, W.
Outline of the necessary laws of thought.
(From the 4th London
ed.)
New
TONNIES,
1752.
York, Sheldon, 1860.

I.
H.
De
logicae scientiae ad exemplar arithmeticae instituenda ratione.
TRENDELENBURG, F. A. Logische Untersuchungen. Ed. 3. Leipzig, Hirzel, 1862. TRUAN-BORSCHE, M. Die ersten Schritte zur Entwicklung der logischen und matheTWESTEN, A. D. C.
1825.
matischen Begriffe. Langensalza, Beyer, 1912. Die Logik, insbesondere die Analytik.
Schleswig, Taubst-Inst.,
1
VAILATI, G. 103 ff.
Un
teorema di logica matematica.
Riv. di
Matem. (Turin),
(1891), pp.
Le proprieta fondamentali delle operazioni della logica deduttiva, studiate dal punto di vista d una teoria generale delle operazioni. Ibid., pp. 127 ff.
Sui principi fondamentali della geometria della retta.
Ibid.,
2 (1892), pp. 71
ff.
Bibliography
Dipendenza
405
fra le proprieta delle relazioni. Ibid., pp. 161 ff. dans lea logique math6matique et sa nouvelle phase de developpement Rev. de Met. et de Mor., 7 (1899), pp. 86-102. Merits de M. J. Peano. einer VATER, J. S. Pasigraphie und Antipasigraphie; oder Ueber die neueste Erfindung fur alle Volker, und von Wolkens, Leibnitzens, Wilkins allgemeinen Schriftsprache Leipzig, 1799. und Kalmars pasigraphischen Ideen. Trans. Amer. Math. Soc., 5 (1904), of axioms for geometry. O. A
La
VEBLEN,
system
pp. 343-84. The foundations of geometry. Math., pp. 1-51.
In Young
Monographs on Topics
of
Modern
VENN,
Consistency and real inference. Mind, 1 (1876), pp. 43-52. Boole s logical system. Ibid., pp. 479-91. On the various notations adopted for expressing the common propositions of Proc. Camb. Phil. Soc., 4 (1880), pp. 35-46. logic. On the diagrammatic and mechanical representations of propositions and
J.
Phil. Mag., ser. 5, 10 (1880), pp. 1-15. reasoning. On the forms of logical proposition. Mind, 5 (1880), pp. 336-49. of geometrical diagrams for the sensible representation of On the
employment
logic.
logical propositions.
j-*
Proc.
Symbolic
Ed.
2.
Camb. Phil. Soc., 4 (1880), pp. 35-46. London, Macmillan, 1894.

Mind, 6 (1881), pp.
.
Princeton Rev., Sept. 1883, pp. 247-67. Notice of Castillon s Sur un nouvel algorithm logique.
447-48.
to the library by John Venn. Catalogue of a collection of books on logic presented Univ. Library Bull. Extra series. Cambridge, 1889. Wien. Schaumburg, 15 VICTORIN, A. Neue naturliche Darstellung der Logik. der Logik. Leip A. Die Auflosung von Urtheilssystemen ... in der Algebra VOIGT,
. .
zig,
1890.
289-332. Logik? Viertelj. /. mss. Phil., 16 (1892), pp. Wesen der Mathematik. Leipzig, Teubner, 1908.^ Voss, A. Uebcr das Grau, 1( WAGNER, L. H. Grundriss der reinen und allgemeinen Logik. Hof., 8 des mathematiques de M. Couturat. Rev. de Phil., WARRAIN, G. Les principes 517-29, 658-73. pp. WERNER, H. Ein Beitrag zur Lehre logischer Substitutionen.
Was
ist
Camb. Univ. Press, 189 t*WniTEHEAD, A. N. A treatise of universal algebra. Amer. J. of Math., 23 Memoir on the algebra of symbolic logic. t*
* *
t*WHiTEHEAD, A. N., and RUSSELL, B.
_ _ -
139-65, 297-316. On cardinal numbers.
Ibid.,
24 (1902), pp. 367-S

substitution groups,
The
logic
of
relations,
logical
and cardi
Ibid.,
31Ibid-., 26 (1904), pp. Phil. Trc mathematical conceptions of the material world. ser. A, 205 (1906), pp. 465-525. (London),
25 (1903), pp. 157-78. Theorems on cardinal numbers.
On
The axioms The axioms
of protective geometry.
of descriptive geometry.
Camb. Umy, Camb. Univ.

Rev. de Met.
Press, 1907.
et
Introduction logique a la geometric.
de
Qd._ QQ
La
theorie relationniste de
espace.
Ibid.,
23 (1916), pp. 423-54
iplification of the logic of relatives.
Ibid., 17 (1914), pt. 5, pp,
387-90.
406
*
Certain formal invariances in Boolian algebras. Trans. Amer. Math. Soc., 18 (1917), pp. 65-72. WILBRAHAM, H. On the theory of chances developed in Professor Boole s Laws f Thought. Phil. Mag., ser. 4, 7 (1854), pp. 465-76.
J. Essay toward a real character and philosophical language. London, 1668. Mercury, or the secret and swift messenger. Ed. 2. London, 1694. WILSON, E. B. Logic and the continuum. Amer. Math. Soc. Bull, 14 (1908), pp. 432-43. Symbolic logic. Ibid., pp. 341-44. WINTER, M. Metaphysique et logique mathematique. Rev. de Met. et de Mor., 13
WILKINS,
(1905), pp. 589-619.
Application de Talgebre de logique a une controverse juridique.

(1906), pp. 617-25.
Ibid.,
14
Sur Introduction logique a

186-216.
la theorie
des foncotins.
Ibid.,
15 (1907), pp.
Les principes du calcul fonctionnel. Ibid., 21 (1913), pp. 462-510. WOLF, A. Existential import of categorical predication. Camb. Univ. Press, 1905. WUNDT, W. Logik. Ed. 3, 3 vols. Stuttgart, Enke, 1906-08. Refer especially
vol.
1.
to
P. Remarks on Professor Boole s mathematical theory of the Laws of Thought. Canad. Jour., 10 (1865), pp. 161-82. YOUNG, J. W. Lectures on the fundamental concepts of algebra and geometry. New
YOUNG, G.
YULE, G. U.
91-133.
York, Macmillan, 1911. On the theory of the consistence of logical class-frequencies and its geo metrical representation. Phil. Trans. Roy. Soc. (London), ser. A, 197 (1901), pp.
An
introduction to the theory of statistics.
Ed.
2.
London,
Griffin, 1912.
INDEX
(References here given are in addition to, not inclusive the Table of Contents).
of,
those readily suggested
by
Absorption, law of, 74. Absurd, see Impossible. Addition, arithmetical, 343; definition of arithmetical by Peirce, 81 ff., by Peano, 345; relative addition, 91, 95, 275 ff. See also Sum. Alphabet, logical, 74.
Aristotle, 231. Bayne, A., 36. Bentham, G., 36. Bernoulli, J., 18. Bernstein, B., 119 Boole, G., 3, 4, 31, 78-79, 118, 137, 207,
Either or, meaning of, 213. Elective symbols, 52. Elimination, in Boole s system, 59-60, 64; in Jevons s, 75-76; in the BooloSchroder Algebra, 153 ff., 194. e-relation, 16, 261-62, 270, 345. Equivalence, in the Boole-Schroder Al gebra, 120; of classes, 262; of relations, 270; of propositions, strict, and mater ial, 292-93.
Euclid, 342.
Euler
diagrams, 176.
ff.
211, 212, 217; Jevons on system of, 72; system of, compared with Peirce s, 80 162. ff.; his general problem,
Existence, 12, 14, 186-88, 328-29. Exponents, in the logic of relatives, 87
Calculus, 4, 6, 8. See also Classes, Proposi Rela tions, Propositional functions,

tions.
Cantor, G., 4, 5. Carroll, Lewis, see Hodgson, S. G. Castillon, G. F., 4, 18, 32-35. of Classes, conception of, 261 application Boole-Schroder Algebra to, 121-22; calculus of, derived from Strict Impli cation, 321 ff. in Coefficients, in Boole s system, 59-63; the Boole-Schroder Algebra, 137 ff. Consistency, relation of, in Mrs. LaddFranklin s svstem, 109; in the system of Strict Implication, 293; meaning of,
;
Extension, Chap. I, Sects. II and III, passim; in Leibniz s system, 13-14; Boole s logic based on, 52; classes in, 184-86; relations in, 219; propositions in, 230-31, 328; definition in, 349. See also Intension. Formal implication, see Implication. Formulaire de Mathtmatiques, v, 5, 7, 115-16, 278, 28078!, 368.
Fractions, see Division. Frege, G., 4, 5, 50, 114-15, 273, 351. F-relation, 366. Functions, as developed by Boole, 58, 82; in the Boole-Schroder Algebra, 125, 133 ff. See also Pro positional Functions.
333.
Contained
65,
in (the relation
C),
43, 16, 47,
118-19, 262, 270; Peirce on the meaning of, 83-84, 96. Contradictory propositions, 189; as treated by De Morgan, 40, Jevons, 76-77. Contrary propositions, 189; as treated by Jevons, 76-77. Converse, propositions, 191; Converse relations, 46, 74, 91, 273-74.
Geometry, 10, 340-42; 363-67. Grassmann, H., 4. Grassmann, R., 107-08. Hamilton, Sir W., 4, 36-37. Hamilton, W. R., 4.
Hilbert, D., 5. Hodgson, S. G., 326; his diagrams, 180 Holland, G. J. von, 18, 29-32, 323.
ff.
Dalgarno, G., 5. Dedekind, R., 4,5, 114,342. del Re, A., 119.
DeMorgan,
A., 4, 5, 79, 209;
DoMorgan
Theorem, 125, 237, 283.

Descriptions, 290, 343. Distribution of terms, 43. 21 Division, in Lambert s system, Holland s, 30 ff.; in Boole s, 61
Peirce
s,
Huntington, E. V., 119, 342. Ideographic language, 2, 6-8, 340. Boole s sys Implication, no symbol for, in tem, 65; material, 84-85, 214-15, 231, Chap. V, passim, see esp. 303-04 and Sect. V; formal, 243, 328 ff.; four types See also Contained in. of, 259. See also Con Impossible, 32, 292. 336-39.
sistency. Inclusion, see Contained in. Inconsistent triad, 19.5-97.
ff.;
ff.;
in in
Indeterminate, see Undetermined.
81
ff.
of Boole, Duality, law of, in terminology 58/in Boole-Schroder Algebra, 126.
Index law, 54. Individual, 324. See also e-relation. Induction, see Mathematical induction.
Index
Inference, 57-58, Chap. Ill, Sects. II
III,
and
passim; in the logistic method, 352-53; immediate, 14, 33, 39, 41, 77, 190 ff.
Intension, 8, 13 ff., Chap. I, Sects. II and III, passim, 73; postulates for calculus of propositions in, 318; calculus of classes in, 323. Jevons, W. S., 4, 118, 149, 205. Kant, I., 341.
s, 81; in the Boole-Schroder Algebra, 119; of classes, 120, 185, 262; of regions in space, 175; of relations, 86 ff., 219, 271, 275; of two functions, 143; strict logical, 293. See also Multi plication and IT and 2 operators. Prepositional functions, 94, 113; meaning
Peirce
Kempe, A.
Kircher, A.,
B., 198.
5.
Ladd-Franklin, Mrs. C., 78, 108-10; 118,

195, 205, 211, 231.
232-33; range of significance of, 233, 254; in Principia Mathematica, 287; calculus of, derived from Strict Implication, 320 ff. Quantification of the predicate, 19, 24, 36, 38 ff., 56. See also Undetermined co
of,
242,
efficient.
Lambert, J. H., Leibniz, G. W.,
4,
18-29, 32, 204, 323.

372.
3, 4, 79,
Linear Sets, 365, 367. Logical product, see Product. Logical sum, see Sum.
Logistic, vi, 5-7, 11. Lully, R., 5. MacColl, H., 4, 108, 119, 327. Mathematical induction, 29, 131, 236, 351.
Quaternions, 4; Peirce s logical, 103 ff. Reductio ad absurdum, 166-67. Regions in a plane, application of the Boole-Schroder Algebra to, 120-21,
175.
Relations, as treated
by DeMorgan,
85
ing
of,
ff.,
Peirce
Multiplication, Peirce, 101-02,
arithmetical,
ff.,
denned by
ff.
by Lambert, 28-29, 45 ff., by Peirce, 102-05, by Schroder, 111 ff.; s paper on logic of, 100; mean
37,
of, in
extension, 219, 269; calculus

of classes,
by Peano, 345;
275
multiplication 86
relative See also
compared with calculus
Product. Necessary, 17, Chap. V, passim. Negative, in the Boole-Schroder Algebra, 119, 124; terms, 38, 53, 73; classes, 121, 185, 263: relations, 46, 220, 271; propo sitions, 14-15, 25-26, 30-31, 32-33, 40, 57, 108, 188-89, 213. 292. Null-class, 185-90, 327. See also 0.
Null-proposition, see 0.
Null-relation, see 0. Number, 80, 101; in Peano s Formulaire, 344; denned in Principia Mathematica,
271; converse, 91, 273-74, 276; powers of, 29, 277; domain, converse domain, and field of, 277-78; calculus of, de rived from Strict Implication, 323. Relative terms, 277-78; in DeMorgan s system, 45 ff. Peirce s treatment of, 85 ff.; Schroder s treatment of, 111 ff. See also Relations.
;
Royce,
J., vi,
195.
346 ff. Universe of discourse; see also under Operation, meaning of, 342; nature of, in logistic, 358 ff.
1,
see
4, 5, 78, 110 ff., 211, 246, 261, 279. Science, exact, 7, 370-71. Segner, J. A., 18. Self-contradictory, see Impossible.
Schroder, E., v,
Sheffer,
operator, see
H. M., 119. II and 2 operators.
Order, logical,
3, 342, 364. Peano, G., 50, 115. See also Formulaire de
Mathematiques
Peirce, C. S., 4, 261, 279. Plato, 4.
Solly, 7, 29, 36. Solutions, in Boole s system, 60-63; in Jevons s, 76-77; in Peirce s, 98-100; by means of diagrams, 77, 181 ff. of some
;
Ploucquet, C.,
4, 18.
operators, 79, 97 ff., 140, 234. Fieri, M., 5, 342. Poretsky, P., 114, 145-46, 163-66, 200. Possible, 15, 329. See aho Impossible.
n and 2
See logical problems, 201-12, 215-19. in also Equations and Inequations Table of Contents. Square of opposition, 190. Subcontrary propositions, 190. Substitution of similars, in Jevons, 75.
treated arithmetical, Subtraction, by Peirce, 80-81; in Leibniz s system, 1718; in Lambert s, 19; in Castillon s, 32; in Boole s, 53; in Peirce s, 81 ff. Sum, in Leibniz s system, 16 ff. in Lam bert s, 19; in Castillon s, 32 ff.; in
;
Premise, in logistic, 352-53. Primitive concepts, in Leibniz, 7. Primitive ideas and propositions, of Prin cipia Mathematica, 282, 287-88; of arithmetic in Peano s Formula-ire, 34445.
Principia Mathematica; v, 5, 7, 8, 102, 116, 222, 261, 277, 279, 281, 314-16, 324,
361. Probability, Boole s treatment of, 67 Peirces treatment of, 105-06. Product, in Leibniz s system, 12 ff.
ff.;
Lambert
s,
19;
in
Boole
s,
52;
in in
s, 52-53; in Jevons s, 73; in Peirce s, 81-82; in Schroder s, 111; in the Boole-Schroder Algebra, 119; of classes, 121, 185, 263; of propositions, 213; of relations, 271; of two functions, 143; of prepositional functions, 94; See also strict logical, 291, 301-02. Addition.
Boole
Index
s system, 26 ff.; 31-32; in Castillon s, 34; DeMorgan s, 41, 49 ff.; in Boole s, 57-58; in Jevons s, 75; in Mrs. LaddFranklin s, 109-10, 195-97; Peirce s principle of, 85; application of the Boole-Schroder Algebra to, 181-82, 193-95; in Barbara, 245, interpreted
Syllogisms, in
in in
Lambert
Holland
s,
in Jevons s, 75; in 82; in the Boole-Schroder Algebra, 186. Universe of discourse, 37; diagrams of, 177
Boole
s,
50-51;
Peirce
s,
ff.
See aho
refs.
under
0.
Variables, 3, 232
Peirce on, 93; in Principia Mathematica, 289. Venn, J., v, 18, 201, 203, 211; diagrams,
ff.;
by
Kempe,
365;"
conditional,
197;
77, 176
ff.
limitation of, System-S, 366.
1,
198-201.
Whitchead, A. N., v, 118. cipia Mathematica.

Wilkins,
0: in
J., 5.
See also Prin
Thomson, W.,
36.
Tonnies, I. H., 18. Truth value of propositions, 227, 230, 294. Two-valued Algebra, defects of, as a calcu
Undetermined
lus of propositions, 281. class, in Holland s system, 30; in Castillon s, 32. See also Undeter
mined coefficient. Undetermined coefficient, in Leibniz s in system, 15; in Lambert s, 24 ff.;
Boole s system, 52 ff.; in Jevons s, 73-74; in Peirce s, 82; in Schroder s, 111; in the Boole-Schroder Algebra, 119; in the Calculus of classes, 121, 185 ff., 263; in the system of regions in a plane, 181; in the calculus of proposi tions, 213-14, 223 ff.; in the calculus of relations, 218-19, 271; Boole s algebra and 1, 52. is an algebra of
<-
(X
b,<.
J3IND1NG
UCU
19/U
University of
Ton vc
:
>
CheckCut Racs
020V/0!
12:33pni
:
Item: A survey
o symfji
li:
eg
Due
Date:
3/3/;iOi. ):./::
r>-.
ltem:Handbookofh(i
.-tote-
is.1or
::>
"i!
^n^ OD
iiS
BC 135
Lewis, Clarence Irving A survey of symbolic 1

A Survey of Symbolic Logic

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

A Survey of Symbolic Logic

Uploaded by

Copyright:

Available Formats

SEMICENTENNIAL PUBLICATIONS

UNIVERSITY OF CALIFORNIA PRESS

PRESS OF THE NEW ERA PRINTING COMPANY LANCASTER, PA.

THE DEVELOPMENT OF SYMBOLIC LOGIC

Symbolic Logic. Symbolic Logic Summary Account of their Logistic.

From Leibniz to De Morgan and Boole

Developments since Peirce

THE CLASSIC, OR BOOLE-SCHRODER AL GEBRA OF LOGIC

General Character of the Algebra.

Fundamental Laws Fundamental Laws

APPLICATIONS OF THE BOOLE-SCHRODER

for the Logical Relations of Classes ....

The Application The Application The Application

SYSTEMS BASED ON MATERIAL IMPLI CATION

The Logic The Logic

THE SYSTEM OF STRICT IMPLICATION...

mediate Consequences Strict Relations and Material Relations

Extensions of Strict Implication.

and the Calculus

SYMBOLIC LOGIC, LOGISTIC, AND MATHE MATICAL METHOD

and Principia Mathematica.

matics and of Logistic

SECTION IV. SECTION

Kempe and Royce

TWO FRAGMENTS FROM

373 389 407

Vorlesungen uber die Algebra der Logik.

classic, or Boole-Schroder algebra,

have been introduced.

such a student most needs

a comprehensive survey of the sub

him with more than the

which he knows, and

the Boole-Schroder algebra and to one another.

attempt to meet this need, by bringing within the compass of a single

notation (so far as possible), the most

important developments of symbolic

are here fulfilled, so

survey does not pretend to be encyclopedic.

of results achieved, or a superficial account of

symbolic logic than in any other mathematical discipline.

presented must be treated in sufficient detail to afford the possibility of real

This aim has required careful selection of material.

attempts to follow the main thread

In the remainder of the book, the selection has been

fundamental have been

set forth in outline only, or

the simpler issues involved.

here given, though such statements as are

Particularly in the last chapter, on

and Mathematical Method

not possible to give anything like an

not space for argument

of this difficult terri

tory as will aid the student in his further explorations.

of the essence of logistic,

observation that stu

even those with a

the technique of rigorous demonstration.

for valuable assistance in

preparing this book for publication: to Professor

translation of Leibniz, to Professor J.

Press for their patient helpfulness in meeting the technical difficulties of

and teacher, Josiah Royce,

and probably by other names. have chosen "symbolic logic"