University of Southern California Dissertations and Theses

A Reexamination Of F. H. Bradley'S Critique Of Relations

(USC Thesis Other)

A Reexamination Of F. H. Bradley'S Critique Of Relations

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content INFORMATION TO USERS
This material was produced from a microfilm copy of the original document. While
the most advanced technological means to photograph and reproduce this document
have been used, the quality is heavily dependent upon the quality of the original
submitted.
The following explanation of techniques is provided to help you understand
markings or patterns which may appear on this reproduction.
1. The sign or "target" for pages apparently lacking from the document
photographed is "Missing Page(s)". If it was possible to obtain the missing
page(s) or section, they are spliced into the film along with adjacent pages.
This may have necessitated cutting thru an image and duplicating adjacent
pages to insure you complete continuity.
2. When an image on the film is obliterated with a large round black mark, it
is an indication that the photographer suspected that the copy may have
moved during exposure and thus cause a blurred image. You will find a
good image of the page in the adjacent frame.
3. When a map, drawing or chart, etc., was part of the material being
photographed the photographer followed a definite method in
"sectioning" the material. It is customary to begin photoing at the upper
left hand corner of a large sheet and to continue photoing from left to
right in equal sections with a small overlap. If necessary, sectioning is
continued again — beginning below the first row and continuing on until
complete.
4. The majority of users indicate that the textual content is of greatest value,
however, a somewhat higher quality reproduction could be made from
"photographs" if essential to the understanding of the dissertation. Silver
prints of "photographs" may be ordered at additional charge by writing
the Order Department, giving the catalog number, title, author and
specific pages you wish reproduced.
5. PLEASE NOTE: Some pages may have indistinct print. Filmed as
received.
Xerox University Microfilms
300 North Zeeb Road
Ann Arbor, Michigan 48106
I
: l
76-5238
CURTIS, Carl Harold, 1929-
A REEXAMINATION OF F. H. BRADLEY'S CRITIQUE
OF RELATIONS,
U n iv ersity o f Southern C a lifo r n ia , P h.D ., 1975
Philosophy
Xerox University Microfilms r Ann Arbor, M ichigan 48106
(D Copyright by
Carl Harold Curtis
1975
THIS DISSERTATION HAS BEEN MICROFILMED EXACTLY AS RECEIVED.
A REEXAMINATION OF F. H. BRADLEY'S
CRITIQUE OF RELATIONS
by
Carl Harold Curtis
A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(Philosophy)
June 1975
UNIVERSITY O F SO U T H E R N CALIFORNIA
THE GRADUATE SCHOOL
UNIVERSITY PARK
LOS ANGELES. CALI FORNIA S 0 0 0 7
This dissertation, written by
da.*" ( Q. vair
under the direction of hJS.... Dissertation Com
mittee, and approved by all its members, has
been presented to and accepted by The Graduate
School, in partial fulfillment of requirements of
the degree of
D O C T O R OF P H I L O S O P H Y
Dean
Date.02fa^.&&4 .d±2£.
[SSE^TATION ,CQMMITTB$:
...........
Chairman
CONTENTS
Page
INTRODUCTION................ 1
Chapter
I. BRADLEY'S STATEMENT OF THE CONTRA
DICTIONS IN RELATIONAL THOUGHT .......... 7
A. Statement of the Thesis.............. 7
Our purpose is to determine whether
Bradley's arguments that the re
lational view contradicts itself
was ever refuted.
B. The Importance of the Problem of
Relations . . . . ................. 7
"Thing" is ambiguous, which
problem refers us to reality, not
words. A real thing includes its
relations, which we ignore. This
creates problems.
C. Preliminary Clarifications .... 12
What is the relational view?
Important concepts are dis
cussed.
D. The Relational View Stated
Abstractly . ....................... 17
It presupposes a metaphysical
plurality.
E. The Contradictions of the Abstract
Relational View Shown in Detail . . 18
1. Every Relation Is and Is Not the
Whole Situation.................18
The relation is (a) the whole
complex, and (b) an element in
that complex.
ii
Chapter Page
2. Terms and Relations Both Do and
Do Not Qualify One Another . . . .
They are both adjectival and
substantial
3. Every Relation Is Multiple
Indefinitely ..................... 30
There is a vicious proliferation
of elements.
4. Every Relation Is (a) Both
External and Internal, and
(b) Both Extrinsic and
Intrinsic....................... 39
Difficulties cannot be avoided
by appeal to the internal-
external distinction, nor to
intrinsic and extrinsic
qualities.
5. Every Relation Is and Is Not
Capable of Degree.............. 43
Degree of relation is relative
to scope of instances allowed.
F. The Relational View Is Not Another
Ultimate Form of Unity.............. 54
Relational thought has no field of
its own in which it is consistent.
G. Relational Thought Stands on Immediate
Experience While Attempting to
Supersede I t ......................... 58
It must be grounded, but that
grounding is ignored.
H. Relational Experience Cannot Be Taken
as Either Intelligible or Con
sistent in Itself................... 64
It is conditional, for it cannot
have denotation only, but with
connotation it is beyond itself and
relative.
I. Summary of Chapter I ................... 73
II. BRADLEY1S 'REALITY' AND LINGUISTIC
PHILOSOPHY............................... 86
iii
Chapter Page
A. Bradley's Meaning for "Reality"........... 86
It is the transcendent ground and
subject of every judgment.
B. Moore's Objections to Bradley's
Use of "Reality"....................... 94
1. Bradley Used "Real" in Two Senses,
Contradicting Himself .......... 95
2. Bradley Was Misled by a False
Theory about Words .............. 96
3. "Reality" Is Indefinable.............98
4. Reality Cannot Have Degrees .... 102
C. Reply to Moore's Criticisms.............. 104
1. Did Bradley Contradict Himself by
Using "Real" in Two Senses? . . . 106
Confusion about Bradley's view
led Moore to accuse him of
ambiguity.
2. Some Difficulties with Moore's
Conception of Definition ........ 113
"Simple indefinable" contradicts
itself, is meaningless if taken
seriously, and makes inference
impossible.
3. A Defense of Bradley's Criterion
for Reality: Reference to
Conditions ............ ..... 121
Moore's objection is ambiguous,
contradictory, and fails to
convict the criterion of circu
larity.
4. Bradley'3 Conception of Reality Is
Acceptable to Common Sense . . . . 127
D. Austin's Critique of "Reality" ......... 131
All his examples are compatible with
Bradley's "reality."
E. Did Austin 'Really' Contradict
Bradley?.............................. 140
Chapter Page
To characterize reality as Austin
does is another way of agreeing
with Bradley.
F. Summary of Chapter I I ..................147
III. COMMON SENSE, CATEGORICAL JUDGMENTS, AND
THE PCS PRINCIPLE........................159
A. The Objection to Bradley's 'Reality'
from Common S e n s e................. 159
It contradicts Bradley if it means
to assert a categorical proposition.
B. Common Sense and the New Philosophy:
the PCS Principle................... 162
Moore meant his propositions to
be categorical, so he asserted
the PCS Principle.
C. Need the PCS Principle Be Justir-
fied? . ........................... 166
It does not justify itself
against other views.
D. Moore's Justification of Common
Sense............................... 169
He appeals to the simple in-
dubitability of his Common Sense
propositions.
E. Commentary on Moore's Justification
of Common S e nse..................... 173
His defense presupposes what
itself needs proof.
F. Moore's Conception of the Given, and
Its Alternative..................... 180
Moore makes an assumption about
the given, but there is an
alternative.
G. A Difficulty with Moore's 'Given' . . . 185
Philosophy is trivial if merely
descriptive.
H. Summary of Chapter I I I ..................188
v
Chapter Page
IV. MOORE'S REPUTATION OF INTERNAL
RELATIONS.............. ................
A. The Relevance of the Question of
External Versus Internal
Relations ...........................
This question is crucial to Moore's
conception of philosophy.
B. Moore's Procedure Against Internal
Relations ............ ............
Moore set out to (1) state the
proposition implied by internal
relations, and (2) show it to be
false.
C. Two Preliminary Comments on
Moore's Procedure ...................
1. He inserted a questionable entity.
2. His conception of part-to-whole
relation is false.
D. Moore's Demonstration of the Fallacy of
Internal Relations .................
It turns upon the crucial dif
ference between "must" and "would
necessarily be."
E. Did Moore Confuse Abstract "Things"
with Concrete Existents? ..........
"Would necessarily be" states no
actual necessity.
F. Is Every Matter of Fact a Necessary
Truth? .................. ........
It is crucial to internal relations
that every matter of fact be a
necessary truth. Is there an
alternative?
1. Necessity Is Psychological in
Origin .............................
2. Necessity Is A Linguistic
Convention .........................
Linguistic formalism presumes three
questionable theses.
.197
197
198
201
205
210
214
215
220
Chapter Page
G. The Relevance of Alternate Views of
Necessity to Moore’s "Common
Sense View of the World"............225
Common Sense propositions. What
was his alternative to
philosophical necessity?
H. In Defense of Universal Necessity • • • 229
Three arguments are offered.
I. A Logical Difficulty with the Mere-
Matter-of-Fact ..................... 232
A criticism of Moore's mere matter-
of-fact is offered. The logical
force of "mere;"
J. Summary of Chapter IV ................237
V. THREE LATER CRITICS OF BRADLEY .......... 248
A. Russell's Attack from Asymmetrical
Relations.......................... 248
Bradley's "logic of mysticism"
reduced relations to properties,
but the asymmetrical relation was
a fatal exception.
B. Reply to the Attack from Asymmetrical
Relation............................ 254
1. Russell's attach was an
Ignatio Elenche..................254
2. Russell Mistook Bradley's
View .......................... 257
3. Bradley Had Noticed Asymmetrical
Relations...................... 259
4. What Did Russell Mean By
"Patent Fact*’? ..................260
C. Did Bradley Confuse a Relational
Complex with a Mere Collection? . . 266
McTaggart argued that the regress
of secondary relations proves
primary relations.
vii
Chapter Page
1. What Distinguishes Secondary
Relation from a Property of
Primary Relation? ................. 271
2. Mere Relation Generates No
Secondary Relatednes ............ 272
3. McTaggart Misunderstood New
Regress Arises ................... 273
D. Bergmann's Reply to Bradley: the
Nexus or Fundamental T ie............273
Bergmann argued that the regress
only proves another entity: the
nexus or fundamental tie. But,
1. The Nexus is also related.......... 275
2. Mbnism Is Not Avoided.............. 276
3. The Nexus Does Not E x i s t.......... 277
E. Gram's Answer to Bradley..............279
Gram argued that distinguishability
does not entail separability. But,
1. The Relation Is Not
the Fact ................ 281
2. Gram Mistook Bradley's
Fundamental Assumption..........282
Bradley did not assume that
distinguishability entails
separability in metaphysics,
but in logic. But the fact
remains unaccounted for.
F. Summary.................................283
VI. UNIVERSALS, IDENTITY, AND THE CHALLENGE
FROM RESEMBLANCE........................ 291
A. The Relation of Resemblance is Required
by Universal Ideas .......... ...292
Blanshard would replace the identity
theory of universals with the
resemblance theory. This is fatal
to Bradley's view.
Chapter Page j
I
B. Did Blanshard Contradict Himself? .... 298
His theory of universals seems j
to conflict with his claim about
relations.
C. Blanshard, James, and the Resemblance
Theory........................... 302
Blanshard presents and defends
James's criticisms of Bradley.
1. Identity Cannot Always be Fonnd . . . 304
2. The Identity Theory Leads to
Vicious Regress ..................... 305
3. The "Chinese Boxes" Objection .... 306
D. Reply to the Arguments for Simple
Resemblance.....................307
1. James's Logic Is Faulty.............. 310
2. James's Regress Objection Takes
Identity in Abstraction ............ 313
3. James Makes an Irrational Demand . . 315
To demand a non-abstract
identity found through
analysis is to demand the
impossible.
E. May No Identity Be Found?.......... 317
Identities are found in qualitative
and generic universals.
F. Is A Qualitative or Generic Universal
Never Experienced?.................... 321
1. Meaningful Experience Requires
Universals.......................... 322
2. Universals Endure through Change
of Resemblance among Specifics . . . 324
G. Is There No Problem with Specific
Universals?.......................... 327
ix
Chapter
Page
1. It Is False that Immediate
Experience Is of Nothing but
Specific Universals ............... 328
2. Merely Possible Repetition Makes
No Universal....................... 329
H. Summary of Chapter V I ....................330
VII. CONCLUSION..................................337
A. A Tu Quoque Justification of Our
Conclusion cannot be criticized on
the grounds that reality is unknow
able, for our opposition must regard
ultimate truth as "unexpressible" and
"ineffable*" ....................... 338
B. The Superior Satisfactoriness of Our
Conclusion ........................... 341
APPENDIX
A. What Is an Idea? .........................354
B. What Is A Fact? .........................361
C. Sameness and Difference.................. 365
D. Was Bradley's View Incompatible
with Science? .........................370
E. Is Philosophy's Task Descriptive? .... 375
F. Subjectivity or Objectivity?.............. 385
G. On the Supposed Difference Between
Logical and Causal Sequence ............ 388
WORKS CITED................................ 392
INTRODUCTION
I propose to re-examine F. H. Bradley's critique
of "the relational view of things." That critique carried
philosophical skepticism further, perhaps, than it was
ever carried before? yet it had a curious result: when
pushed to such an extreme it led to a positive conclusion
about reality that could be reached in no other way.
So, at any rate, he thought. Bradley believed
that he had proved reality, contrary to appearances, to be
some kind of harmonious unity that transcended the
ordinary world. The more conventional belief, of course,
is that reality consists of a large number of things with
relations between them. Some of those relations are, on
that view, merely casual and accidental. The relations
between some things are not necessary; things and some
relations are separable, at least in thought, with no loss
of intelligibility. This conventional view is what I shall
call "the relational view."
Bradley's skepticism was more extreme than that of
his famous predecessor, David Hume.'*' Hume had attacked
necessary connections among matters of fact. We relate
them, said Hume, only out of habit arising from constant
. . 9
conjunction in experience. But what is it for things to
be related? By pressing this question to new depths,
1
Bradley carried skepticism into the very meaning of the
particular matter of fact or existent itself— into the
3
very terms by which we express xt. If we are to consider
a certain matter of fact, it must have some significance;
but if so, how? There is here a problem about relations.
Is not the very meaning or intelligibility of a matter of
fact relational? Is it some isolated event or phenomenon
"shining alone in the wilderness?" Does its significance
not relate, instead, to our world? Does its significance
not reach beyond merely the fact itself in order to be
meaningful to us? It does, and so it is through and
through relational. But further, the fact has parts. Are
not the parts of the fact themselves related, so that
together they make the fact? We have, then, the problem
of getting them together, for what becomes of the fact if
we take its parts piecemeal? The pencil in my hand has
parts: lead, wood, eraser, and paint. There is no
necessary connection among them. Then why are they to
gether? There is more to the "together" than just the
parts, and it is the "more" that makes the pencil. The
wholeness of the fact is a problem. This much is obvious.
What is not so obvious is that the significance of
all these parts is no less relational. Lead, wood,
eraser, and paint all have meaning to us, but that meaning
is not limited to them as the internal elements of a
pencil. Indeed, if they were so limited, they would not
2
have the significance they do have. Neither is the
pencil, taken as one whole fact, a significant fact if not
related to our whole life in some way. We would not know
what to do with a pencil, would not even know what to call
it, nor would it make sense to say it was yellow or thin
if these qualities had no wider significance to us.
Relations seemed very important to Bradley,
entering into the very essence of every existent. Hume
might have realized this had he carried his skepticism
further. To the degree that anything means anything, as
everything does, relations are involved. Yet relations
pose a problem. For what necessary connections have the
meanings of the mere elements in the relation, so that the
significance of any whole fact can come from them? The
pencil i£ a pencil, and its significance is relational
throughout. But is it not also more than relational?
Could it have the significance that it does have if merely
relational. If by "relational" we meant merely that each
element of the fact were only itself, limited to itself,
and if the relation among them merely were added to them,
externally; would there then be any whole fact deserving of
the name? Would there not be, instead, only a collection
of independent parts and external relations completely
devoid of meaning? It seemed so to Bradley. What, then,
is to explain the meaning that in fact we do find? We do,
after all, claim to find it, not make it.
Such questions about relations led Bradley to
j conclude that the skeptical road ends in a situation that
is worse, even, than Hume had thought. At least he has
given that impression.^ It is not only that we have no
necessary connection among events. Even the event itself
is devoid of significance or intelligibility. Our darkness
is total.
Is there no escape from so dark a prospect?
Perhaps so. I have said that Bradley's skepticism led to
a positive result that could be reached in no other way.
I shall return to that at the end of this essay, but a few
words are in order here. The dark prospect comes from
taking things as terms in relation: from our intellectual
habit or taking "things" as merely casually or externally
juxtaposed. As intellectuals we take "the relational
view." But there may be a message for us here. If we do
find significance in "things," and if Bradley's dark
prospect simply cannot be accepted, then reality cannot
be identified with what is intellectually apprehended.
What we take of life through the relational view is not
reality, for there is more to reality than we think. That
is an indication of the positive result to which this
skepticism leads.
This essay is about the negative aspect of
Bradley's philosophy: his critique of relations. In this
critique one may find his special contribution to
4
|philosophy. One hears little about that contribution
| today because his critique of relations is thought to have
i been refuted. Perhaps it was not, however. That is what
this essay is about. Was Bradley refuted?
Throughout this essay I shall argue as the defender
of Bradley's position. Of course, I shall try to state
the case for the opposition fairly, at or near the begin
ning of each chapter; but I shall also present what I
think Bradley would have insisted upon as its weakness. I
take Bradley's position because it now stands most in need
of a spokesman, and also because my sympathies lie with it.
In the chapters to follow, the first one examines
the arguments that comprise Bradley's critique of
relations. In the next one I shall examine linguistic
arguments against Bradley's use of "reality." Following
that, I consider an argument for a Common Sense view which
disagrees with Bradley, and I then examine what is believed
to be a formal refutation of Bradley's view of relations.
I next examine arguments exploiting a purported failure by
Bradley to distinguish between a "complex" and a
"collection." My last chapter considers a recent defense
of an old attack upon Bradley's view of identity. The
sequence of these chapters is historical, but an attempt
will be made to show the relevance of one to another.
5
FOOTNOTES— INTRODUCTION
^This should not be taken to imply that the two
men had more than their skepticism in common.
2David Hume/ An Inquiry Concerning Human Under
standing/ ed. Charles W. Hendel (New York: The Liberal
Arts Press, 1955), pp. 187-89.
Gilbert Ryle stresses Bradley's contribution to
this problem (of meaning) in The Revolution in Philosophy
(London: Macmillan and Co., 1956), pp. 6-8.
^Bradley's dark view of the powers of reason is
remarked upon by M. R. Cohen in A Preface to Logic
(Cleveland, Ohio: World Publishing Co., Meridian Books,
1956), p. 206. Cohen speaks of "the essential tragedy of
human thought itself— its unavoidable task and its in
escapable frustration." Similarly, T. W. Silkstone says,
"The proper conclusion is Bradley'si that discursive
thinking cannot grasp reality." (See his article, "Bradley
on Relations," Idealistic Studies IV, No. 2 [May 1974]:
169.)
It is fair to add, for Bradley, that "Metaphysical
speculation has led me, if I may speak for myself, neither
to scepticism nor pessimism." (F. H. Bradley, Essays on
Truth and Reality [Oxford: Clarendon Press, 1914],
p. 444.) There Is not space here to fully discuss the
sense in which Bradley was skeptical, although certainly
he did employ it and had much to say about it. He did
press skepticism very hard against relations, which is our
concern here.
CHAPTER I
BRADLEY'S STATEMENT OF THE CONTRADICTIONS
IN RELATIONAL THOUGHT
A. Statement of the Thesis
As a central tenet of his metaphysical position,
F. H. Bradley maintained that the "relational view"
gives to those who adopt it an object which is only
appearance, not reality itself. For several decades it
has been commonly assumed in philosophical circles that
Bradley has been shown to be wrong on this matter. This
essay intends to demonstrate that Bradley has not been
proven wrong in his view that the relational conception
of the world is beset by pervasive and overwhelming
difficulties. In developing this modest thesis an attempt
will be made to provide evidence that can later be used in
support of the stronger position that Bradley was right
about the relational view.
B. The Importance of the Problem
of Relations
We would all agree that the world contains
"things," and might even go on to say that reality
consists of them. When we have named and described them
all, we suppose, we shall have given an account of
reality. Bradley thought there was a problem lurking
here, however, that damages our early and easy assurance
if we allow ourselves to ask just what a "thing" is.
A leather briefcase seems to be a "thing;" so is
the handle. The leather pieces that compose it may be
"things;'/ and even the rivets and thread that hold it
together. They seem to exist separately, and do not in
themselves, imply the larger "thing" that is the brief
case. They could exist without it— it seems. On the
other hand, however, they seem more problematic than the
briefcase. Probably we would not call the dye that
colors the leather a "thing," nor the initials engraved
upon it, and surely the use of the briefcase is not a
"thing," nor is the history of its manufacture.
Let us consider the least problematic "thing":
the leather briefcase itself. Its presence here and now
is plain fact. We can verify it. We can lift it and
turn it, open and shut it, note that it has pockets in
side, a color outside, a shape and a texture. Slap it,
and it makes a hollow sound. Here, we may say, is the
very sort of thing the real world is made of— no doubt
about it.
But what is there no doubt about? The real
thing? The present fact? Let us look into that. The
fact, we suppose, is a significant fact: something
intelligible to us. That is, the presence of the brief
case means something to us. If not, it is no briefcase
and we would not call the presence of a briefcase a fact.
It must be a "what" as well as a mere "that". The fact so
far as significant involves the function that briefcases
serve in our complex civilization where books and papers
must be carried about. Our civilization requires brief
cases, and so enters into the fact. Thus a great many
other things are involved with this briefcase, if it is
what we claim or what we understand it to be. Our claim
requires a relation of the briefcase to all those other
things, ultimately. That relation is very complex. Yet
it is only one total structure, for this is only one
briefcase. We see that the relation is becoming a prob
lem, a many in one, and the significance of our briefcase
is threatened.
Then let us disregard that. We still have the
"thing" itself, we presume. It still is what it is. It
still has its own meaning. As such it is distinguishable
from, say, the telephone. It is of pliable leather while
the telephone is of rigid metal and plastic. And yet, it
seems that here again relation is involved. The plia
bility of the leather stands out against the rigidity of
the metal and the plastic. Take them away and what be
comes of the pliability? Do we not need to compare them,
and so relate them, if we are to assert or withhold this
judgment of quality? Perhaps it is what it is, but if we
are to judge that it is pliable we need to compare it.
And if we go on to say that it is made of leather, then
we have right there— so far as leather is intelligible to
us— a reference to the leather industry. This thinqhood
involves the slaughtering and tanning of hides with all
the history of its development. But is it also a leather
article? Then we make implicit reference to factories,
to wholesale and retail sales and distribution, to Wall
Street financing, and all the rest. They are necessary,
are they not, to the actual article before us? Here
again a very complex relation is required.
It is still before us, presumably. But has it
color, size, and shape? These too, as intelligible
aspects, are all relational. Is it a mere object
irrespective of its distinguishing qualities? But
"object" is a relational term. It needs to be related to
us, for it is our object. That relation i|3 not it, for
it is presumably independent of the relation it has to
us— and so, likewise, are we of it. And yet if we dis
regard that relation, in order to get away from a
problem, then what right have we to speak of it?
Hence Bradley found a difficulty with the
relation of the "thing1 ,1' both within and without. He
complained that we overlook the infinite complexity of
our "things'.'" He tried to bring to our attention that
what we call sijnply a "relation" comes nowhere near to the
richness of reality. We may judge, "The leather brief-
10
case is before us," and so, assert a relational "fact",."
But in the actual fact, as it is in reality, the relation
that makes the fact is infinitely more complex than any
that may occupy our attention. If we mean to assert the
fact really, of the thing as it truly is (how can the fact
be anything less?), then we mean to assert all of that
infinite complexity. We do not succeed in explicitly
uttering it, and we hardly ever even try.
So, what do we manage to get into our explicit
judgment? Certainly not reality, but rather an appearance
of reality.
Why do we not try to tell the whole truth of the
fact? Because we rely upon implicit reference to the
context to provide what our explicit utterance per
impossible could not express. When we refer to the leather
briefcase we do not mention the state of our civilization
with its books and papers, nor do we mention the story of
leathercraft which provides us with the manufactured
article. We pyesume all that in presuming it meaningful
to speak of briefcases. Most of the story of the actual
"thing" we ignore. The "things" so far as we make explicit
what we mean, is but a conventional selection from
reality 'pulled out' of it for utilitarian reasons. So
long as that is all we mean to do when we say "the leather
briefcase;*" then Bradley would have no complaint. But that
does not assert the actual fact. It is not an assertion
of, but rather a presumption upon, reality.^
! C. Preliminary Clarifications
1 What is the relational view? What positive
i character does it have? What alternative is there to
adopting it?
The relational view would be held by anyone who
believed that there truly are many real things with
genuine relations between them that are what they seem,
individually, apart from those relations and apart from
the other things. It is 'realism' where that ambiguous
word is taken to mean that things are as they seem: that
there is a one-to-one correspondence between ideas^ and
reality.^ In a loose sense, of course, there is no one
who does not believe this, for as a practical matter the
relational view seems impossible to avoid. But not every
one uses it as a first principle in philosophy. Some
philosophers have wondered whether it was warranted to
presume that reality falls into place according to the
distinctions people commonly make. An even smaller
number, such as Bradley, have offered a proof that reality
could not be so construed. They have found it necessary
to distinguish between appearance and reality.
The relational view is marked by the activity of
thought. Thought names, characterizes, classifies,
compares, and interprets. All perceiving falls within the
relational view, for it involves coupling some immediate
12
j "this” to get a judgment of "what.'' The judging of
i things in this manner becomes so habitual that everything
seems to come already sorted out into tables, chairs,
briefcases, and so on— especially to adults. That is why
it seems so absurd to deny (as Bradley did) that ideas
copy reality. Yet philosophers, of the most widely
differing views on other matters, agree that a
distinction is to be drawn between what is immediately
presented in life and what comes to us only through
reflective thought. Russell and Moore, two of Bradley's
ablest critics, recognized the distinction. Others have
made a comparable one. Even Ludwig Wittgenstein (who
seemed to many to have changed all the rules) recognized
an inside and an outside to his fly-bottle. They may
have disagreed on which side was to receive the emphasis,
but they did allow the distinction.
But what alternative is there to taking the
relational view? If this question is taken to mean,
"What is to displace the relational view?" the answer
is that nothing can. Bradley did not urge that reason
be thrown out, nor that perception be forever distrusted.
What he urged was that careful attention to how things
are shows that the relational view is secondary and
artificial. It needs to be supplemented or grounded.
That does not make it worthless, nor does it make it even
untrustworthy, if it be remembered that in truth it i£
13
grounded? and that grounding (and the limits implied) be
kept in mind. Yet in philosophy it is most interesting
to notice that no "thing" can be what it is without
involving one's whole world, and nowhere short of that
whole can one stop without impoverishing the "thing;1"
Anyone can verify this to his own satisfaction, Bradley
thought. One assumes the context, one feels the unity,
and only within it have the divisible parts any place and
significance. Thus we might say that feeling is the
alternative to the relational view— not in the sense that
it is exclusive of it, or ought to replace it? but in the
sense that feeling is needed by, presumed in, and wider
5
than, the relational view.
Bradley's "feeling" alternative to the relational
view may be more clear if one understands the sense in
which he criticized the latter. He did not deny the
proposition "this table is in this room" in the sense of
holding that, after all and really, there is no table
here. He was not saying that we are dupes of an
illusion. His point was that the actual table was
designed by someone, built by someone, transported by
someone, and bought by someone. It is made of some kind
of wood and finished with some kind of varnish. If we
really know what we mean when we speak of the table, then
it is all this and more that we mean. These are not
extraneous details to be whisked away by a distinction
14
I drawn between essential and accidental attributes, for
; they are the actual conditions of this table being in this
|
room. They actually enter into the fact. If we wish to
assert that fact then we must get these aspects of the
"what" of our table into that assertion. "This table is
in this room" does not do the job. An important part of
the story has been left to supposition.
In the presence of the actual fact, when one says
with all possible assurance, "this table is in this room,"
the support of the meaning and conditions implied are
felt to be present somehow; whatever may be the paucity of
one's detailed knowledge. But then, the conditions making
the actual fact, although felt and implicit, must be
regarded as unknown. All that the proposition directly
asserts is that under unknown conditions a table is in a
room. But, on the other hand, under unknown conditions a
table is not in a room. The proposition is compatible
with falsity. It cannot then be asserted as true. If in
our logic there are no degrees of truth— if the only
alternative to absolute truth is utter falsity— then the
proposition is false as it stands. It is in this sense
that Bradley denied such propositions. In this sense he
denied the relational view, and suggested that we rely
instead upon feeling— for in truth We do. The relational
assertion may say something but what it says falls short
of reality.^
What quality has the relational view that
provides the difference by which we distinguish it? It
is the quality of being a many-in-one, an identity-in-
difference. We do become aware of a plurality of
separate "things;" but our awareness was initially
unitary (in immediate experience, below the level of
reflection), and from it we begin to mark out differences
and to relate them. By so doing we get an intellectual
grasp of things. The original feeling of wholeness gives
way to a secondary activity of distinguishing "this" from
"that," each acquiring its own "what" from the whole, and
thus we get ".ideas" to use in what we call "reasoning,«
It is as though an original fluid whole had crystallized
into an interlocking mosaic, from which bits and pieces
may be chosen. Choosing bits and pieces, we reason. It
is no doubt a valuable and even humanly indispensable
activity, and no doubt reality is such as to permit it,
but it is secondary and artificial. One may forget that
what thought manipulates is dependent on the original
whole. One may call the individual bits and pieces "the
given»"
Bradley's critique of relations is found in an
n
essay on which he was at work when he died. He had
completed only the first part but his outline and notes
were also found. The remaining sections of this chapter
are arranged according to his plan of development. In my
discussion of it his other works will be referred to,
for work on the problem of relations pervades them all.
He was trying to show in detail the contradictions in
the relational view, and was also trying to show that it
is grounded in some pre-relational mode of awareness—
perhaps misleadingly called "feeling." In the next
section the relational view will first be stated
abstractly so as to have before our attention a state of
mind that is usually adopted ■uncritically. Following that
its contradictions will be presented.
D. The Relational View Stated
Abstractly
If the relational view is taken strictly, as an
adequate rendering of things as they are, then the real
world consists of an irreducible diversity of elements.
It is a metaphysical plurality of things in themselves,
each of which is a complete entity with a being and a
character of its own. Each element has sufficient
character to maintain itself, though the others should
change or even cease to be. Some of these elements are
terms, others are relations, and the relation holds
between the terms while maintaining its own character
against any intrusion from outside.® The coming together
of the terms with the relation is not due to any character
of their own. They are indifferent to each other. The
character of each term and relation is limited to what in
it is essential, and what is outside of it, the fact of
g
relatedness, is accidental. While the degree of strict
ness with which the relational view is held may vary with
the preference of the person who holds it, still we may
say that he does hold it if some relations, at least, that
something has, could be changed without changing the
essential character of the thing itself. The plum tree at
the corner of the house, for example, would still be the
plum tree that it is were it not at the corner of the
house. Even though, contrary to fact, it had been planted
next door or in an orchard in Afghanistan, it would
maintain its same identity,'*'® according to the relational
view.
All of this is what thought, relative and dis
cursive, takes for granted. It takes ideal contents for
diverse unities and couples them, by means of relations,
into a whole. It presumes to create a unity out of
12
differences apart from any felt and prior totality. It
views reality as a collection of indifferent "whats>," as
Bradley said, "loosened from the immediacy of any 'this'
or ' that'."'*'^ The relational view takes this collection
of elements as ultimate, and regards as non-existent the
14
conditions of their occurrence.
E. The Contradictions of the Relational
View Shown in Detail
1. Every Relation Is and Is Not the Whole Situation
18
a) Presumably a relation is experienced and
! actual, and if it is not, then the fact of relatedness
is missing. The relation is a whole which at the same
time has parts, and the fact of relatedness cannot be less
than the whole and the parts. A relation is never merely
between terms. It "implies an underlying unity and an
inclusive whole," said Bradley, and this whole must be
experienced or there is no relation. That whole may be
wider than we supposed, for, in addition to other factors,
there is also an experiencing subject to whom the related
terms are given. The relation, with terms, is thus more
than they alone, but in the absence of this "more" the
15
fact of relatedness is lost. It is the conditio sine
qua non of the relation. This was denied, of course, by
Russell, who argued that in "Edinburgh is north of London"
the relation "north of" does not "depend on" anything
mental to be a fact. The relation is a universal: an
"object of an act of thought*" But his argument does not
meet the difficulty. If something is an "object*" then
its correlate, "subject*" is internal to the meaning of it.
Can it be denied that, as mere words in the language,
"object" and "subject" are correlates? Can it be denied
that if something in the world is correctly called an
object, that it must be an object for some subject? Or
did Russell, who italicized the word "object*" not mean to
be taken seriously when he said that every relation i£ an
19
object? It would seem, then, that there must be more to
the fact of relatedness than the merely abstract
'relation' or it could not be a relation. Without the
"more" the 'relation' is thinkable, perhaps, but not
thinkable as something actual. This "more" is not
included in the bare proposition "Edinburgh is north of
London .1 1 The relationally whole fact is more than the
terms taken singly with a bare relation between them.
But on the other hand,
b) This must be denied. There is some merit to
Russell's argument. The relation is thinkable somehow as
coming between the terras and not including them. The
terms in their turn are not mere modifications of the
relation, but rather are themselves thinkable as indi-
17
vidual entities. Although there never was a "north of"
that was not an object for some subject, yet the "north
of" somehow is. "North of" is a universal and it is
neither limited to, nor altered by, its coming between
Edinburgh and London. Similarly, London is, and
Edinburgh is, independently of their relation.
M. S. Gram, who argued that Bradley's critique of
relations never was refuted, nevertheless proposes to
reject this first argument of Bradley's against relations
as a sophism. Said Gram,
The only reason why somebody might believe
that a relation must be equated with the whole
complex of which it is a part is that the
distinction between a relation and its terms
20
might be thought to imply the independence of the
relation from its terms [emphasis mine].18
Gram argues that Bradley was merely playing upon the
ambiguity of the word "relation" in saying that it "is and
is not the whole situation." It ia true, said Gram, that
the very meaning of the word "relation" implies terms, and
possibly a great deal more; yet we do distinguish it from
its terms, as is done with a large number of things even
though they couldn't exist without fulfilling certain
conditions. Something can exist and yet be distinguished
from the conditions of its existence. A substantive term
can be distinguished from its adjectives. The same word
can be used for the thing or the substantive term whether
or not there is any intent to include reference to
conditions, attributes, or adjectives. The mere
distinguishability of some term, or of some thing, does
not need to imply at all that it is independent of what it
is distinguished from. Calling our attention to its
conditioned and dependent nature does nothing to weaken
the justifiability of that distinction.
But in reply for Bradley it must be said that,
although one can and does abstract when drawing a
distinction, and although one does use the same word for
the subject distinguished, whether abstract or not: yet
to the extent that it is abstracted it is to that extent
unreal. In speaking of the 1 relation1, was there an
intention to say something truly? Is some real "thing"
21
the subject? If so, then the conditions without which
there is no "thing" (and which the "thing" therefore
implies) must (and do) remain implicitly present so that
they continue to apply. What is spoken of as "the
relation" is but the grammatical subject of an actual
relational fact. We might or might not call a "relation"
a "thing." Gram may be right that no one ever meant to
imply that the relation is independent of the whole and
actual situation— but if not independent then to what
extent is there any "relation" unto itself? If the cat
is on the mat, then the relation "on" referred to is
something very special between this cat and this mat here
now. But if one goes on to say it is the same relation
that exists between the cup and the saucer, the hat and
the head, and the book and the shelf, then why not call
it independent? Gram's defense fails to be convincing.
Russell did mean "independent" and so Bradley's argument
is no sophism.
2. Terms and Relations Both Do and Do Not Qualify
One Another
A relation that is actual is a unique fact: it
19
is a relational situation. If divided into terms and
relation, that division is made, presumably, so that the
parts are relevant to the whole. Nothing is assigned to
either terms or relation except upon taking the whole
situation into account. The abstracted 'relation', so
22
i called, is dependent for its meaning upon that situ-
I ation. Terms and relation are parts of the situation,
j
and so they qualify the whole and each other. They
cannot at the same time be both actual parts and also
individuals.^
They cannot be individuals and yet they must be
individuals. At differing times we affirm one or the
other side of this contradiction. We commonly say that
things are what they are irrespective of "accidents ." Yet
on the other hand we are familiar with the story of the
kingdom, the rider, and the horseshoe nail; and know
that, at least sometimes, small matters affect great
events. But now Bradley is saying that it simply is not
true that there are ever any accidents. That is
astonishing to our common sense. We are apt to call
science to our support, and to argue that scientific law
(composed of universals of relation) provides a prime
example of the very thing that refutes Bradley: a
relation unaffected by changes of content. It is
"accidental" to Boyle's Law which volume of gas it governs
or describes.
Poets may affirm truly that not a leaf in the
forest falls but alters the stars in their courses. But
it is not true, we may object, that this warrants going on
to say that everything modifies everything. What about
the Inverse Squares Law itself, which governs (or at least
23
describes) the trajectory of the leaf’s fall and
j
the star in its course? Is it not plain that the
relation that their masses have to each other,
through the Inverse Squares Law, is indifferent
to both? How can it make any difference, to
that sublimely indifferent Law, which two items
in the vast universe it comes between and
joins? Science takes it that, although there is
no mass in the tiniverse which does not provide
matter for its Law, yet it matters not to the Law
which mass is provided. The relation expressed
by the Law is not qualified by the otherwise
differing characters of the terms.
Russell took over the notion of merely
external relations and modelled his logic after this
presumption of science. Science is thought to
be vindicated by laboratory observation. Some
relations are observed to be independent of the
terms related. These relations are often called
"laws.” Yet in a strict sense no such law
is ever verified. No prediction based on scientific
law is ever precisely borne out by observation,
even under controlled (artificial and therefore
abstract) conditions.
There is always some deviation between the measurements
taken of the particular event and the predicted resuit.^
The deviation may be so small as to be called "negli
gible '," but its very presence needs explaining. Even if
this were not a persistent problem, however, and even if
prediction and observation coincided precisely, it still
would not prove that the relation was unqualified by its
terms. It would not even prove that it was the relation
(the only and "essential" relation) between, say, A and
C, at all. The "relation" as taken is highly abstract.
Out of all the competing candidates for the title "the
relation" between A and C, something is singled out.
Laboratory procedures are then set up to measure that
relation. Even if this were unobjectionable, and even
if measurements between A and C coincided with those
between other instances of "the relation," say between B
and D, it still would not follow that the relation was
unmodified by its terms. What one must get, to have a
proved instance of some relation unmodified by the special
character of its terms, is a relation without any terms.
That is never to be had. If measurements between A— C
and B— D agree (more or less), then it is plausible and
useful to presume some constant relation between them.
But, strictly speaking, it does not prove an external
relation. All that is proved is that science finds it
helpful to presume there is such. There is some reason
25
to believe in it.
But there is also some reason to doubt that there
are any relations unmodified by the terms. Suppose that
one distinguishes, with Russell, the relational universal
"north of" from the particular terms that it joins, and
argues that the solution of the puzzle over relations lies
with a right understanding of the nature of universals.
"North of" is a universal of relation, and so long as one
thinks of universals as the sort of thing that they are,
and does not press against them objections pertinent only
to sorts of things that they are not, the puzzle, one
might argue, will disappear. Now universals are not like
particular things; particulars are the very sort of thing
that they are not. Universals are ubiquitous and in
finitely elastic, or, as Russell expressed it, they need
exist "nowhere and nowhen ." Grant all of that. There is
no need here to raise the question of the existence of
universals, for, if among all else that appears they
appear also, why then, of course, they do. But does it
solve the puzzle? Does it now make it clear how the
original reality was no more than terms-in-relation?
Does it show how there could be any relation if there were
no "more"? And if this "more" is so important to the
universal of relation that without it there wouldn't be
any, does it not then seem very likely that "the nature of
universals" is involved? In the preceding section it was
argued that certain conditions must be satisfied before
one is warranted in saying that there actually is any
(universal of) relation. Without conditions, no relation.
Is it warranted procedure, then, to appeal to 'the nature
of universals1 to explain why those very conditions can
be disregarded? It seems doubtful.
According to Russell the universal of relation is
non-mental. But what is one being asked to do when one is
asked to understand the nature of universals so that the
puzzle of relations will lose its fascination? Is one not
being asked to accept something utterly non-mental as the
explanation of the undeniable progress of thought? Some
thing here begs the question, for the universal of
relation, as Russell himself said, is an object of
thought. Then, contrary to Russell, it is relative to
thought. Is it however not wholly relative, in the sense
that while it may be partly mental, yet as an object it is
not? Then it is relative in part and non-relative in
another part, and so is divided within itself in a way
that needs to be explained. It is not, then, well under
stood how these two discrepant parts join to convey one
universal meaning.
Perhaps one might try to bypass ultimate problems
raised by the metaphysical thing-in-itself and restrict
concern to only the meanings of words. The universal-word
"relation" has a certain use in the language system, and
27
whatever meaning may not be covered by that (in the
problem of expressing the whole relational fact) may be
made up for by adding yet other words to serve that
purpose. Perhaps one ought to more painstakingly
analyze the elements and use the appropriate word for
each. Each word will be chosen to stand for a distinction
drawn among the elements so that various qualities and
functions are divided up among them. That is what we do
when we use words. It may be true, we might reply to
Bradley, that words have discrete meanings, and it may
also be true that no whole can be composed of absolutely
autonomous units. But to use that as an objection to any
verbal formulation is to badly misunderstand how words
work. Shall this mere peculiarity of words, that they
are a "many" conveying "one" meaning, be allowed to be
fuddle philosophers? Before we became philosophers we
learned to use words. Recognize the plain fact that
words do successfully describe and do not be misled by a
merely verbal difficulty. If the whole fact is not
described except in part by some word, yet the description
can be completed by adding yet other words. For example,
one might add to terms in relation falling short of the
22
fact something like "relational properties" to fill out
the fact. The "more" is accounted for by a relational
property or something of the sort.
This method of escape from the problem over
28
relations seems to have been what Bradley meant by "an
appeal to the diversity of respects." His reasons for
rejecting it seem to be as follows: The words chosen to
describe the relational fact are chosen, presumably, to
account for elements discerned in the fact. If not, then
why just these words and not others? The problem is thus
more than merely verbal. Furthermore, to concede that
words can no more than describe the fact is to confess
that the fact is more than a mere description, and it was
23
of this "more" that one had sought an account. Adding
something like a "relational property" only compounds the
puzzle. One must then ask, "Is there anything in the
term related that requires just this relational property
and not another?" If the answer is yes, then term and
relation have a necessary property in common and the
relation is not merely external. If on the other hand
the answer is no, then why is the relational property
together with them? Still another word is called
for to describe this relation (failing which the
fact is not wholly described), and with it the
same old problem breaks out again.
The relation and its terms both must and must not
qualify one another. On the one hand, as meanings,
supposedly they endure unchanged despite their conjunction
in some fact. They are not dependent for their being upon
their specific use to express that fact. Yet on the other
hand they somehow do express the unique and whole fact.
| That is puzzling. The meaning units are universals. In
i '
1 :
| themselves they are discrete, subsisting, semantic
i
; individuals; but their togetherness in the expression of
the fact is at once both more and less than they alone.
They are less, for they must stretch themselves to cover
the whole and unique fact. They are more, for, although
they themselves are "nowhere and nowhen," yet they accept
with catholic indifference every time and place. Although
these universals undeniably somehow are, and somehow
relate and somehow qualify the expression of the fact, yet
that "somehow" is the most interesting part of the story—
a part that remains untold. As expressive of the fact,
the terms and relation qualify one another; but as
meanings in themselves they do not.
3. Every Relation is Multiple Indefinitely
Order and direction seem obviously important
within some relational situations.2^ Temporal and sphtial
relations involve such situations. The common pre
positions "in," "out," "above," "belowi" "to," and "from,"
all involve spatial relations. "South of" and "north of"
are also spatial relations. "Before,'1 "after," "during,"
and any word implying change (all the verbs) involve
temporal relations. It is hard to see how normal discourse
could be carried on without being involved with temporal
30
and spatial relations. They are necessary to scientific
discourse, moreover, and science is reaching forth to en
compass the whole range of experience. If the spatial or j
temporal aspect of any concept were to be rendered !
questionable, then they would be questionable themselves.
But space and time are contradictory, Bradley argued, for
they are endlessly multiple.
The difficulties he found with space and time are
as follows: The question may first be raised, "Are space
and time relations?" It seems that space cannot be a re
lation, for it must be something on its own. There must be
some substance, or some positive quality, meriting the name
"space." Space is divisible so there must be some "what"
to be divided. Thus space is more than a relation. Yet on
the other hand it cannot be more than a relation. Space
consists of spaces (every space is divisible) that are no
less to be called "space" than the whole of space itself.
If space were some substance or positive quality, it would
be the terms related by the divisions falling between them.
It would be progressively less deserving of the name
"space" as it was progressively more divided. Yet, it is
not so. Space is infinitely divisible, and contains end
lessly yet more possible divisions and relations. It can
not be the terms of the relation so it must be the relation
itself. It is an endless relation among relations. And,
indeed, so it would seem because "space," taken in itself,
is but an unreal abstraction that only falls between terms.
Time, apprehended in the same way we apprehend
space (as we commonly do, under a spatial form),
encounters similar difficulties. It is a stream of which
past and future are parts taken as though they coexisted
together with the present. None of these parts, taken
each in itself, has real duration for they all occur
together. It is their togetherness that we call "time>"
yet how can the positive quality called "time" come from
adding these timeless parts together? Time is thus a mere
relation and no positive quality, and a multiple relation
at that (an internally discrepant relation) for it
contains the contradictory relations "before" and "after*"
Time is, and yet it cannot be, a relation. Take it,
then, as a substance or positive quality, apprehended as
itself uniquely and not (under a spatial form) as the
instantaneous juxtaposition of past, present, and future.
But if time is a real substance or quality it must
be a unity, not a plurality. It therefore repels "before"
and "after” from its "now." These are all incompatible,
one with another. Yet can any of these be dropped from
the meaning of "time"? Does real time not contain, some
how, these three all together? Or, shall we say that
these three are not incompatible? Perhaps we could say
that these three are not parts but rather are aspects of
time. Time includes them, perhaps, as, say, a cup in
cludes its volume together with its inside and its outside.
32
But is this move permissible since "before." "after;"
and "now" are themselves real times? They are all actual
durations. "Now" is a time, and so are "before" and
"after.'' Time presents itself as "now"— but it also
presents its "before" and "after.'' Its "now" is meaning- j
less as time or anything else, except in relation to
"before"and "after." They are not within the "now" lest
they destroy it. The "now" must fall between them.
"Now" is a relation between "before" and "after." Our old
quandary persists. "Now,'" by itself, does not deserve
the name of "time;" for it contains no "before" and
"after." It is only a relation between them. It takes
all three together to make time. Time is not a mere
relation, for it contains a relation together with terms.
On the other hand it is no positive substance or quality,
for it is not one thing but several: two terms and a
relation. No one of those terms, or the relation, can be
taken in itself as real duration (although, inconsistently,
they are so taken), for the relational multiplicity breaks
out again within them as we have seen.
What does this mean? Does it mean that on the
basis of Bradley's criticism, say, of space, a simple
assertion of spatial relation such as "Chicago is north of
New Orleans" must be denied?
Yes, strictly speaking, it does. An affirmation
of spatial relation involves endless multiplicity: a many
33
I I
I is taken as one and so the affirmation as it stands gives I
j appearance, not reality. Space (and time) contradicts
! itself in concept. But reality cannot contradict itself
so space (and time), as people think of it, cannot pro
vide reality's ultimate forms. Bradley's meaning may be
more clear if it is seen as a critique of meaning; as
laying stress upon apprehending that meaning as a
condition for judgment of truth or falsity.
What, then, was Bradley getting at? One might
begin answering by asking a question. If you affirm
"Chicago is north of New Orleans" as fact, do you mean to
affirm less than the actual fact or the full fact? Only
with the latter alternative have you affirmed, or meant to
affirm, reality— for surely the fact can be nothing less.
In the former case (with less than the whole and actual
fact) you have affirmed something somewhat superficial,
p C
which Bradley called "appearance^ The fullness of
detail which makes the actual fact, and which you actually
do mean, you have left to supposition. You may feel that
it is present, you may take it for granted that the other
person knows it, but you have not explicitly affirmed it.
You haven't said it and so that part of the whole meaning
that you have, objectively, still leaves the largest part
! to be filled in somehow by some other means. "Chicago is
north of New Orleans" still leaves on your hands the job
of filling in all these other relations with other terms
| (all no less part of the fact than what you have
| affirmed) making the togetherness of Chicago, New Orleans,
I and "north of;" Two large cities are involved, and with
; them many other relational facts. A spatial relation
"north of" is involved, and with it an orientation to the
planet and the universe. You may wish to make this a
problem about words instead of things, yet surely the
meanings are involved, with whatever relations they may
have that make them what they are. Ultimately the whole
universe is involved if what you really mean is full and
clear— without which you have no right to judge the
assertion true.
The meaning of the whole universe sub specie
"Chicago is north of New Orleans" is what you haven't got.
With what you've got, you've skimmed the surface. What
you've said may be 'true' so far as it goes, but it
doesn't go far enough to qualify as knowledge of reality.
Isn't that what everyone means by "truth"? Doesn't
everyone know, as well, that the bare relation of two
otherwise empty terms falls far short of what more is to
be known about even "Chicago is north of New Orleans"?
Where is that "more" to fall if not in reality? But if
what you say falls short of it, where is it to fall
except in non-reality? If the radical "either-or" is
forced upon us as it is when you insist, "Either affirm or
deny that Chicago is north of New Orleans," then the
truth of the proposition is to be denied. To call it
"appearance”is to be quite generous.
Someone may object that this example ignores the
important difference between terms and properties of
terms. You cannot treat the relation between a city and
its properties, someone may say, as though the city were
related to them in the same way that it is related to
another city. There is a difference demanding a
distinction here, and it destroys your defense of
Bradley. Properties differ from terms. Terms have
properties but terms are related only to other terms.
The above example does not prove that every relation is
endlessly multiple, but proves only that a compound term
(one having properties) carries those properties into
the relation. Even that is required only if the property
is essential to the meaning of the term. The question of
properties, whether essential or not, is irrelevant to the
relatedness of terms.
In reply, we can only repeat that if two terms
are known to be related, then one must suppose that the
actual existence which the terms and relation represent—
or stand for, or point to, or correspond with, or copy,
or carry into the objective understanding by whatever
means— must actually be there if the fact is known. What
ever falls short of that, falls short. Falling short, it
is less than the whole truth for it has not managed to co-
36
| incide with reality. It is relatively meaningless and
j thus not wholly true.
If the drawing of distinctions determines the
outcome of philosophical arguments (as it seems to), then,
with respect to the above objection, three issues have
been raised: In the first place a question may be raised
about what is to be called a "relation." Properties are
not terms to be related, it is objected. They are to be
distinguished from terms, we suppose, by an actual dif
ference. Yet in the absence universally acceptable
criteria as to when there are actually terms related to
other terms (versus when these terms merely have
properties), it seems safest to follow wide practice. It
is a practice wide enough to include Bradley's critics
which calls "different from" a relation.^® From this it
follows that whenever there is discriminable difference,
of whatever kind, then there is relation. A term can be
distinguished from its properties, so it is related to
its properties.
In the second place, it remains to be shown that
there ever are, in actual fact, terms that are not
compound— that is, infinitely compound. Every actual
subject for a term has an infinitude of other relations
(other than the relation chosen for emphasis), and the
other related things are divided from the subject by an
actual difference. All those differences that the subject
37
term has (with whatever differs from it) actually belong
to it because of something inherent in the term itself.
I Thus, although "different from" may be a relation, yet at
the same time the "different from" indicates also a
property of the related term. Taken as properties, they
are internal to the term and make it infinitely compound.
In the third place, it is difficult to see how
there could be any atomically simple relation. Relations
are universals, and so are outside time and space. They
are "nowhere and nowhen"; they do not 'exist' but 'sub
sist' . But what does a relation need to be, to be a
relation? It needs to be a "different from. " It is_ a
"different from." But if it is a "different from" then it
also differs from everything else (whether universal or
not), for it is distinguishable by a difference. That
difference, however, is not merely a property of the
relation, nor of that with which it differs, for it
signifies a relation and comes between them. The
difference, coming between them, differs from them; and
that new difference (the difference, which we had not
noticed, that differs from the difference they have from
each other), is another genuine relation. As such, how
ever, it is not a mere property of either (not a property
either of the terms, nor of the relation of difference
that they have from each other, nor of any other relation
which they have with each other), and is another genuine
relation. Being distinguishable, it is divided by a
difference from the difference, and from the difference
between the terms and the relation. Yet another dif
ference has come to our attention. But neither can that
difference belong to all those others— taken as terms and
relations— for it, itself, being different, is a relation
on its own that requires (and has) those others as terms
to relate. This process of finding new relations of
2 7
difference may (and must) be carried on endlessly. '
It is not being asserted here that this endlessly
regressive set of wretchedly abstract differences
qualifies the actual existence. What is being argued here
is that a formal difficulty has been revealed in the
relational view of things, if taken as real and true.
4. Every Relation is (a) Both External and Internal,
and (b) Both Extrinsic and Intrinsic
(a) The distinction between relations as external
and internal leads to contradiction, for there is no
relatedness that is not both. The meaning of "relation"
both does and does not involve its context. Because of
this contradiction, Bradley called the relational view of
2 8
experience I f a useful makeshift" yielding "relative truth
n ..29
only;
39
An external relation would be independent of the
particular circumstances of its appearing here and now.
Its union with its terms would be a . mere isolated oc
currence, inexplicable and irrational. No "how" or "why"
would help understanding pass from one diverse aspect to
another.^® The relation "north of>,1 1 between Chicago and
New Orleans, for example, could just as well come between
any other two terms. It is nothing to "north of" that it
links these cities. Pushing it to its extreme, "north of"
might as well come between two toots of a whistle, or two
thoughts in a head. It is nothing to the mere relation
what it comes between.
A difficulty remains when the relation is taken
as internal. Relatedness repels internality. The
relation, called "internal" to escape the above dif
ficulty over "external" relatedness, is nevertheless
taken as independent as soon as it becomes the subject of
exclusive attention. "North of" taken internally differs
from "north of" taken externally only because in the
former case one goes on to attribute it to, say, Chicago,
as its (Chicago's) character. "Internal" here means that
there is something essentially "north" about Chicago. As
Russell correctly pointed out, however, no single
character of one term will get it together with another.
Or, as Bradley had put it years before, no character of a
32
term singly gives it a "together" or a "between;" An
40
internal relation uniting them would be possessed by
them both? but because they essentially differ, then that
relation (as a common property) differs from the nature of
each of the terms. The relation, though it began as a
common property, is repelled by that difference and is
forced to become external. Only then can it come between
them; only then can it unite them, for their individual
differences must be respected. They cannot pass one into
the other bodily, for then their individuality would be
lost. They would have achieved the "together" at the
sacrifice of themselves, but by remaining themselves they
could not be got together by a mere quality of one or the
other. "North of" cannot be a mere quality of Chicago,
for as such it says nothing about New Orleans. To the
extent that it is internal to New Orleans, it says nothing
about Chicago.
The difficulty with either external or internal
relations may then be stated as follows: The relation
must be internal because the relation must belong, and be
suited to belong, to the terms making the actual fact.
But the relation must be external so that it can remain
itself, thus maintaining its function of coming between
the terms. Both views of relation are inconsistent and
rest upon abstractions which satisfy only by covert appeal
to a more fundamental and non-relational unity: a unity
presumed by thought when it acts but ignored by thought
| when it formulates its principles. The mere relation is
I ;
no more the actual existence than are the terms, and from ;
these elements no real whole could be composed. Intro
ducing any third element only complicates the relation
and compounds the failure. A further distinction implies
yet another problematic difference and so would not save
the situation, nor can it be saved by an appeal to a
33
"plurality of respects," as was said before.
(b) The problem of intrinsic versus extrinsic
qualities requires that the terms themselves be con
sidered. Although it may be supposed that a term is
either simple or compound, yet, to the contrary, it must
be both. As a mere term, it is simple. But if it is
related, then there must be a. reason for it. Some
quality of the term requires or permits the quality to
apply to this term.
We may in practice sometimes divide the term from
its quality, thinking of it as merely accidentally added
on. Thus it becomes an "extrinsic" quality. Or, for
another purpose, we may take the quality to be essential
to the term, and call it an "intrinsic" quality. Subject
to that qualification, we may justify dividing qualities
into extrinsic and intrinsic. We must remember, however,
O A
that the division involves the view we take and is not
35 . . .
just something in the term. The term thus divided is
an abstraction. What is extrinsic in one situation might
42
not be so under other conditions for other purposes. The
intrinsic-extrinsic distinction would, in any case, be
dependent upon suprarelational considerations.
Chicago is north of New Orleans. Chicago has some
quality, call it "location;?" permitting (or requiring)
this relation. Is it an intrinsic quality? No, not to a
Chicagoan. Chicago is Chicago, though he never even heard
of New Orleans. If one took away from Chicago only that
one quality, the quality that permits "north of New
OrleansV" and left everything else unchanged— the Loop,
the stockyards, the names of the streets and buildings— it
would still be the Chicagoan's home town. "North of
New Orleans" would never be missed. "Location" is only an
extrinsic quality. But what if you lived in New Orleans
and wanted to go to Chicago? Then the situation would
be quite different. "North of New Orleans" would be the
most important quality had by Chicago. "North of New
Orleans" would be an intrinsic quality.
5. Every Relation Both Is and Is Not Capable of Degree
Suppose it were possible, as relational thought
takes it to be, that abstract relations could remain
meaningful and could continue to relate, even though their
separation from the whole were taken seriously. Suppose
that every relation retained each its specific character
in isolation, and that some were and some were not capable
of degree. Then one could set out, like Russell, to
43
I catalog and classify the kinds of relation? and one of
i
! the principles of classification would be degree.
I
"Similar," for example, would be a relation capable of
degree, but "equal" would not be. Only the relations as
they are in themselves need be considered, of course.
Actual occurrence in this or that instance would be ir
relevant to the capacity of a relation to have degree.
Capacity for degree would be a quality of each relation.
This raises a difficulty. If reference to
conditional eontext were irrelevant to a relation having
degree, or even to its latent capacity for degree, then
that determination is impossible. How could any relation
have degree? Each relation would be an individual and
absolute; for what was other than it could have no power
over it, nor elicit any response from it, differing in
degree on one occasion or another. Yet is it not solely
in relation to others that anything may be said to
possess degree? Degree of relation ij3 relevant to other
instances.
Still, and on the other hand, suppose that a
comparison of instances were allowed. By what might
they be known to be instances of the required relation?
How might one know, for example, that one had an instance
of "similar" and not one of "equal"? Must not the
relation be_ there, absolutely? The mere comparison of
instances would not give that information. No, one must
44
| know that instances of the relation are the relation by
| the presence of the relation itself, relative to nothing.
| If the relation were merely relative, admitting a more or
less, then how could any relation fail to have degree?
Things merely relatively equal, for example, would have
equality in degree— if there were any way to know that
"equal" applied at all.
Consider examples. Consider "similarity" and
"difference." Now when one considers examples one must
mean all examples, for by what right may they be limited
to a few? Certainly in such a wide context one can see
that "similarity" and "difference" are relations capable
of degree. But now consider "equality;" One must con
sider all examples of it, must one not? And "equality;"
remember, was a. relation not capable of degree. Do all
instances of "equality" show it not having degree? No.
In so wide a scope of instances, "equality" also has
degree. Consider Bradley's example, that of two
collections of coins. The first collection has more
pennies in it. One could then say that the first is more
equal than the second. It has "equality" in greater de-
37
gree.
The crucial question here is, "How wide a
comparison is allowed?" If all instances are allowed,
; then perhaps all have degree. But if none, or one, or a
carefully chosen few are allowed, then perhaps none have
45
degree. The relation# to be what it is# must absolutely
be itself and have no degree. But to have degree it
implies another, and is relative. If both of these
conditions are satisfied then it is correct to say that
relations both have and do not have degree.
This is an important point. Every relation of
degree must rest on an underlying identity discerned
among related terms# although it is not always explicit.
Degree implies degree of something. Every quality may
be taken as such an identity among instances. Take the
case of degree of redness. It is red that is present as
an identity uniting the instances. An identical "what"
38
underlies the relation. The "what" either is there or
it is not# and so it itself is not capable of any degree.
But just red by itself we never do see, in utter and
absolute isolation. As a meaningful "what" it occurs in
a context wider than itself, here and now. There is
comparison and contrast. There is reference (whether or
not explicit use is made of it) to the past. It is a
meaningful "what #" so there is reference to color and the
other colors# and to all other qualities which conspire
to give red its place and significance. Thus, if one
speaks of more or less redness# it is because red both
is an absolute "what" and yet is not something unto
itself. By finding it red in degree, attention is
called (perhaps not as simultaneous perceptions) to the
other colors and qualities 'round about it' which give
it place in a wider series, and which contribute meaning
to the judgment, even though unseen.
Giving instances of relation presents another
difficulty relevant to whether a relation has or has not
got degree. It has to do with the ambiguity of the word i
used to name the relation. It has been brought to our
39
attention by linguistic philosophy. The problem is
raised by the question, "When you compare instance A
with instance B of some relation, are you sure that just
this relation is meant because the same word is used?"
When you compare "equality;" for example, between two
collections of coins and say that one has a greater
degree of equality than the other, because the one has
more pennies, have you got "equality" in the same sense
in the two cases? Is the equality between two
collections the same in sense as is the equality between
two pennies? If not, then you have not got a contra
diction in "equality" but two different meanings, both of
which you have chosen to call "equality." Bradley's
conundrum is thus merely a verbal one. Because of his
love for paradoxes, it might be suggested, he chose to
call different things by the same name, and then recoiled
in amazement before his 'discovery' that the meaning stood
in contradiction to itself. Bradley's puzzlement, say
linguistic philosophers, was but the predictable result of
his failure to notice ambiguity.
Linguistic philosophers will support their
position by offering examples, so that we may 'see* that
the problem is merely a verbal one. Take as an example
the notoriously ambiguous relational preposition "in*"
Consider,
(1) The hole is in the doughnut
Here "in" is used in an absolute sense. In order for a
doughnut to be what it is, it must have a hole in it.
Here "in" is not capable of degree. But in other usages,
"in" stands for a very relative relation. Consider,
(2) The dog is in his doghouse
The dog may be wholly or only partly in his doghouse. The
relation is one of degree. Shall one say that "in" is
the same in both (1) and (2) and so say, with Bradley,
that "in" both is and is not capable of degree? Or shall
one say, with linguistic philosophers, that the use of the
word "in" differs, showing that the whole problem is only
verbal? Perhaps we could use the linguistic method on yet
more examples, to help you decide. Consider,
(3) The coffee is in the pot
(4) The pot is in the coffee (think of a pot
immersed in coffee)
(5) The money is in the bank
(6) The man is in debt
(7) The man is in love
(8) Salt is in the sea
(9) The fish is in the sea
(10) Pitcairn's Island is in the sea
The puzzlement over how one word can mean so many dif
ferent things can be avoided simply by rephrasing these
statements, using words different from "in" to get an
equivalent meaning. Thus:
(1) The hole is through the doughnut
(2) The dog occupies his doghouse
(3) The pot contains the coffee
(4) The pot is immersed in coffee (the coffee
surrounds the pot)
(5) The bank credits to him the money
(6) The man owes money (The man has a debt)
(7) The man is preoccupied by erotic affection
(8) The sea is salty
(9) The fish is indigenous to the sea
(10) Pitcairn's Island is encompassed by sea
Depending upon what you meant, these more or less re
construct the ten instances of "in." By rendering the
meanings using other words, we can 'see' that the problem
of contradictory meanings in one word has been avoided.
Is it necessary to put up with contradictions and para
doxes in philosophy when they can so easily be avoided by
attending merely to verbal structures?
How could Bradley have defended his position? I
49
will try to speak for him, but I will try to make it
brief. Linguistic philosophy will receive more attention
in the next chapter.
Philosophy, like every other authentically human
activity (they all use words), is the attempt to say what
you mean and mean what you say. It differs from the
other activities by being an unusually determined effort
to do this. It is possible to get by, more or less,
without doing this, but to that degree it lapses from a
purely human activity. It is then not intelligent
activity, and this must be said of the mere use of words
to facilitate practical affairs, even if it were possible
to so use them in neglect of their ideal aspect (as pure
meaning). In philosophy one tries to get the meaning not
only down but out. One wants to get the whole story of
the nature of things down into words, to make it
explicit. There may be good reasons why this cannot
wholly be done, but it is hard to deny that this is the
goal. By "meaning" one refers, in the end, to reality—
that is what was meant— and the words are meant, somehow,
to be it. To some extent and somehow, reality is conveyed
by words as meaning, although perhaps no one has the
correct analysis of how it happens. Meaning, then, is of
more concern than words. If ambiguity can be avoided,
then by all means avoid it. Linguistic philosophy may
help— it may alert us to what we had overlooked— but it
! cannot lift from our shoulders the burden of getting the
i
j two sides together: the reality meant and the words that
somehow mean it. If contradictions persist even though
words are carefully chosen and seriously intended, then
’ why is not that a genuine problem? Is the failure to
mean reality not a genuine failure? Does it show the
problem to be merely verbal? No one is arguing here, on
Bradley's behalf, that reality is truly characterized by
the contradictions marking meaning's failure; but rather
one is saying that the contradictions afflict things as
they appear to us— as we understand them. In this sense,
perhaps, the problem is 'in' the words, but is it any less
a genuine problem? For human beings, the world of which
they speak is the world they mean and the problems there
are genuine problems.
The allegation that all paradoxes (e.g.,the
contradiction in the meaning of "equality") arise out of
the misuse of words, needs to be proved. Is a word
"misused" if used in more senses than one? Is the contra
diction of two meanings in one word always to be avoided
by substituting another word for the second meaning? Does
our ability to avoid paradox— or, as linguistic
philosophers say, "mental cramp"— by this device prove
that every meaning is specific to some native and natural
use? Suppose it did, and suppose, moreover, that the
device of 'showing' the specificity of use were sufficient
to prove that philosophical problems, such as paradoxes, I
were verbal in origin. Then a worse mental cramp would
j
be engendered because the device proves too much. Taken
as a principle, it proves that meaning is never the same
when contexts differ. Notice the difference, one is I
urged, between "The hole is in the doughnut" and "The dog
is in his doghouse." The situations are different so the
meaning of "in" is different. Being different, and not
the same meaning at all, the relation "in" does not
contradict itself. The mental cramp is banished? the
philosophical problem is 'seen' to be merely verbal.
Well, yes, but are not these different instances
of "in" at the same time similar? If similar, are they
not similar in some shared respect? And is not "in" that
very respect? But if one takes as a principle the notion
that any discernible difference makes them totally
distinct, then how is comparison possible? Are they
instances of anything, or not? If the slightest dif
ference cannot be ignored, then is not a basic presumption
struck out that is made by everyone, not by philosophers
only? That presumption is that the same thing can recur
although different in detail. The presumption may be in
defensible, as the critique of relations so far has al
ready shown. Yet how much of the world as people know it
would survive if the presumption were banished? Bradley's
"sameness in difference" persists in every concept by which
52
the world has meaning, and without it there i£ no J
!
meaning. Every difference is absolute. But if the
world is intelligible, then meaning is universal.
j
Universals, as meanings, are the same amid differing
contexts. They cannot be, and yet must be, one and the
same. Here, Bradley thought, is a genuine paradox? a j
glaring contradiction with which linguistic philosophy,
too, must deal. In reply to a criticism, Bradley long ago
asked the question: "Does just any difference justify a
distinction, or not?" The question is relevant here. If
one answers "Yes" to the question, then has one not
affirmed a principle of limitless fission making it im
possible even to speak at all?
Linguistic philosophers offer us no refuge from
the paradox of meaning: the paradox of one in many. When
on the attack they insist that meanings are never quite
the same? yet when stating their own positive doctrine,
they presume that the meanings of the words they use are
shared, communicable, the same for the speaker as for the
hearer. They will say something like, "Don't ask for the
meaning, look at the use';"40 But if the conscientious
philosopher, worried over meanings and beset by apparent
paradoxes, tries to take the advice seriously, he may find
only more trouble. He may find that "look" as used here
| (excised from its natural home) is but a crippled metaphor?
that "use" as meant here is actually "meaning" by the wrong
name: mutilated by having its ideal aspect ignored.
i
What then? On the linguistic philosophers' own principles,
j
|the use of a word is not the sort of thing that can be
looked at, and "use" (taken from its home in ordinary
usage and made to function as a substantive noun) means
"meaning." The advice can't be followed.
If the contradictions found by Bradley in such
things as relations could easily have been avoided by
using other words, then the burden falls upon his opponents
to show precisely where he erred, and how his puzzlement
could have been avoided. It is not enough to appeal to a
general a priori principle that all paradoxes merely
exhibit the misuse of words. Bradley was alert to the
problem of ambiguity? he understood that there are some
times only verbal difficulties. He used words carefully
and dealt in meaning, not verbalism. Linguistic
philosophy orders philosophers to 'look at' meanings.
That is precisely what Bradley did. He found them wanting.
He examined relations, as "meanings"— separate, self
subsisting semantic entities— and found them contradicting
themselves. He found that they both were and were not
capable of degree.
F. The Relational View Is Not Another
Ultimate Form of Unity
It might be thought that, while relational unity
presupposes an infrarelational unity (in feeling— a point
54
| to be discussed in the next section), the way is still j
open to argue that relational unity is itself one
ultimate form of unity. One might argue that relational
unity need not be contradictory so long as one does not
try to equate it to some other sort of unity: one must
only understand what sort of unity relational unity is.
Although grounded in feeling and not replacing it, and
although feeling might still be.prior in experience, yet
all of this need not bar relational unity from being an
unique development on its own. The relational unity
expressed by "The cat is on the mat" would be unproblematic
within its own field.^
In reply one must ask, "What does it mean to call
the unity of feeling 'prior'?" What does it mean to admit
that relational unity is "grounded"? Does it mean any
thing more than that for certain; purposes the actual
placing in existence, through feeling, of terms-in-
relation, can be ignored because it is presumed to be
present? One ought not to forget that the attempt in
philosophy is to state the conditions of truth in a way so
that they are included in one's writing of philosophy.
Granted that for the most part one can, in everyday affairs,
ignore fundamental philosophical questions, yet one cannot
do so here. By ignoring the grounding of terms-in-relation,
did one mean to imply that at no time does the actual
grounding make any difference to the fact of relatedness?
55
This might mean only that the grounding can be ignored
if terms-in-relation are grounded in ways only negligibly
different. If so, it would not follow that the actual
relation were indifferent to its ground. To defend the
autonomy of the relational view, first define the view
strictly, and, keeping to that, inquire whether there
still is no problem there. That is what was just done,
at the start of this essay. It was pointed out that terms
and relations do not relate apart from some prior whole.
The relational situation cannot truly be taken as con
sisting ultimately of an atomistic conjunction of terms,
united (somehow) by relations indifferent to the unique
situation. The relation relates only if the whole
situation is implicitly referred to. If not, and if it
were taken seriously that the relational view "has its
own field." then the relation expressed would fall short
of the actual situation. Falling short, it would cease to
relate. This was the first point to be made in response
to the question, "What does it mean to say that a relation
is grounded and unproblematic so long as one keeps to its
own field?"
A second point of rebuttal may also be put as a
42
question: "Can the relational view have its own field?"
In a trivial sense, of course, it can: for presumably one
means something actual when one refers to "the relational
view*" Somehow it qualifier reality, for there is nothing
56
else to qualify. But the difficulty is with "view." The j
i
"view" of reality is not reality, for ex hypothesi the re- i
I
lational view has its own field. But the purpose in philo-j
sophy is to understand, and that means, sub specie reality.;
Nothing less will do. Many and various are the "views" it
I
is possible to take, but views are beset by a difficulty: j
they are incomplete. A "view" of something has some object
without which it would be no view. The relational view, to;
return to that, is peculiar in taking everything for its
objects, and what it cannot so take it rejects as unreal.
There is contradiction here. When on the offensive, the
relational view asserts its sovereignty over all things,
including "views." But, when put on the defensive and
asked to justify itself, it retreats to the position that
"within its own field" it is autonomous. It takes all
fields and makes its own 1 science' of them, and then calls
that its "own field." Its "own field" is not an exclusive
preserve. What is "own field" to mean? It is co-extensive
with the universe, and yet not. It is autonomous, and yet
admits to being grounded. If by "own field" it means
merely itself, then it means hardly more than that to take
things as terms-in-relation is to take them as terms-in-
relation."
Of the autonomy of the relational view, Bradley
said,
We have a right everywhere to make and to imply
whatever assumption seems to aid us most to a better
understanding in the world . . . [but] . . . in the
57
end it is not true that a sameness and difference
in our terms and relations can be one-sided,
separable, and independent.43
Terms and relations involve both a sameness and difference.
As individual entities, eabh term and relation must be, or
have, some character of its own. Yet, to permit in
ference, they must pass beyond themselves to enter re
ality. This entering of reality is something from which
even the "essence" of terms and relation cannot stand
apart. Yet neither can they include it.
G. Relational Thought Stands on
Immediate Experience While Attempting
to Supersede It.
Relational thought presupposes or rests upon what
Bradley called "feeling.Although unstable and giving
rise to the relational view,^ yet feeling is always
implicitly present. By "feeling" Bradley did not mean
anything merely emotional, such as mere pleasure or pain
(themselves abstractions), nor did he mean something merely
subjective. In the beginning there is no awareness of
self separate from not-self— a distinction that must be
46
prior if feeling is called merely subjective. Indeed,
at the level of primary feeling there are no distinctions
at all. That level is unstable and full of inconsisten-
47
cies when one seeks to grasp it in thought. It sets the
problem for relational thought, which is,
To complete itself and pass into the other. It
implies a substantial totality beyond relations
58
and above them; a whole endeavoring without
success to realize itself in their d e t a i l .^8
Further, the level of feeling in experience "points to" a
transcendent unity harmonizing the inconsistencies of
particularity in relational experience with the totality
of the felt.^ The relational mode (experience understood
through terms-in-relation) rests upon the felt and "forms
50
a new and incompatible feature" with it. Feeling, then,
is not the whole of reality. "Feeling*" said Bradley,
"supplies us with a low and imperfect example of an
C " I
immediate whole*" It supplies an immediate "this" of
which relational thought tries to explicate the "whatv"
and in so doing the lower and the higher (feeling and
thought) fall asunder.
One must not suppose that relational thought is
to be regarded as a mistake, nor that discursive reason
is to be avoided so that, in the name of truth, one may
reimmerse oneself in the unity of the felt whole.
Relational experience is somehow a necessary construction
upon feeling, without which "we should be confined to the
feeling of a single moment." But, nevertheless, primary
experience is felt and felt in its unity.
Why did Bradley use such a word as "feeling" to
name the level of initial experience? Possibly, because he
did not think it was misleading if used in its correct
sense— a sense corrected by eliminating from it such
59
accretions of meaning as ought more rightly to belong to
such words as "prejudice," "superstition," "irrationality,"'
"passion;'!1 or words naming some particular emotion. Let
us at least try to get the 'feel' of Bradley's usage. To
him, "feeling" meant,
The general condition before distinctions and
relations have been developed, and where as yet
neither any subject nor object exists. . . . It
means . . . anything which is present at any stage
of mental life, insofar as that is only present
and simply is . . . Everything actual, no matter
what, must be felt; but we do not call it feeling
except so far as we take it as failing to be
more.53
Note that everything is initially felt, that feeling as
such simply i£, but that feeling may 'crystallize' or
'polarize' into something more definite in character.
Its own character, so far as it may be said to have one,
is fleeting, forever changing,
And both from within and from without, feeling
is compelled to pass off into the relational
consciousness. It is the ground and foundation
of further developments. But it is a foundation
that bears them only by a ceaseless lapse from
itself.54
What is important here is that we not ignore the
contribution of feeling to cognition, and its priority to
the relational view. Feeling, "being more than merely
simple, holds a many in one, and contains a diversity
within a unity which itself is not relational.
Is feeling ever found pure? Does feeling ever
actually come prior in time, either in the individual or
60
in the human race? So far as one can tell# it may never
be found pure, but rather feeling contains what Bradley
called "clots" of relational experience within the fluid
whole of feeling. Such relational structures within the
felt would not destroy its unity. The question for
Bradley, moreover, is not what is prior in time, but what
is initial in experience. Perhaps nothing is ever felt
except that it is felt serially, and so one may go on to
say that it is in time. Then one may ask what is first
in that order of time. But to do all that does not render
questionable the claim that whatever is, is first felt,
and felt as a unity that somehow makes for further
distinctions. The priority of this felt unity is not at
the mercy of those distinctions, temporal or otherwise,
made upon reflection— however effectively one may argue
that they are found rather than made. It has already
been seen that such a unity must be, both given immedi
ately and as required as a "middle term" for rational
inference. That unity is unaffected by any sense of
temporality in experience. Neither is it obvious that
the sense of spatiality or temporality is even basically
relational, for those senses involve a feeling of effort
or passage. Whether or not space or time are relational
calls for a review of the contradictions in them, as
56
relationally taken.
Another problem for "feeling" appears. It seems
61
to contradict itself. If whatever is must first be
felt and felt as it is, and if distinctions ever; may be
warranted, then the parts of feeling disrupt the unity of
the felt as much as do terms and relations disrupt the
unity of relational thought. Furthermore, if some
distinguishable quality justifies calling the felt,
"feelingthen the not-felt must stand over against it;
that is, in relation to it, so that the "feeling" quality
stands out. A distinguishable quality requires a dif
ference in reality to support it, and "different from" is
a relation. Thus, "feeling" is a term in an actual
relation. How, then, can it be prior to all relations?
So, "feeling" has parts in relation, which explode it
from within, and "feeling" is a term in relation, from
without. Bradley's critique of relations destroys his
prior "feeling/" and he has no right to call it, or any
distinguishable character, (any "what") prior. He is
brought down at last to the mere something-or-other as
prior, from which all positive quality has vanished and
of which further talk must cease.
This objection is an example of the pervasive
(if not pernicious) influence of relational thought.
57
Bradley's reply ran as follows: feeling does go on to
clash with the other parts of experience, and so, cannot
ultimately be real either. It is primitive or original
for us. It does not clash within itself, for the 'parts'
as such do not exist within it. They arise in
relational thought as we reflect on our feeling. But
feeling does clash with and relational thought and one is
brought at last to the Absolute. It is unfair, however,
to call the Absolute "the mere Something-or^other» " for
it, too, has its distinguishing quality. If "quality" is
the word, it is a quality peculiar in that it reconciles
all contradictions and thus deserves the naitfe "pure
harmony^" It satisfies completely both feeling and
thought, and so is pleasant above all things. That is
another story. To return to feeling, it is only being
maintained here that it is the prior ground for thought.
It is not maintained here that feeling is ultimately real.
To say that it clashes, moreover, is inaccurate, for it
goes on to clash with experience taken as bare terms-in-
relation. How this can be is "inexplicable'; " but it is
not contradictory to say it. Feeling does, not clash,
particular feelings clash. When we are immersed in just
feeling, there is no conscious conflict. Only when we
withdraw from feelings do they become 'particularized*
and clash. They are not able to remain mere feeling when
taken in reflection. Here feeling has begun to break up
into the abstractions of reflective life. I have used
the metaphor, "crystallization#" here. A "this" comes
over against a "that," for change has come between them.
The "is" separates from the "was," and "whats" come in to
qualify "this" against "that,V Feeling vanishes and
terms-in-relation take its place.
58
In pure feeling there is no contradiction for
no 'other* disrupts it. That is why Bradley said that
we have an immediate, if low level, apprehension of the
Absolute. This is no petitio principii. The claim rests
on direct experience. The claim is reinforced by reason
which is to be trusted, but which needs a ground to
guarantee it. That is why we said, to begin with, that
Bradley's skepticism gives us a positive result that
could be reached in no other way.
H. Relational Experience Cannot Be Taken
Either as Intelligible or Consistent
in Itself
We often think that the popular device of
"denotative" definition solves the problem of getting the
immediate and the reflective sides of experience together
in some intelligible way. The belief is mistaken, Bradley
thought. According to him, "this" cannot be used to ground
a term and then be dropped leaving terms-in-relation to
carry on, on their own. "This" is merely a symbol
indicating that the immediate is in hand and that all
conditions required for attribution of truth to ideas are
taken to be satisfied. But the conditions are not thereby
got into the judgment, and if not, the judgment is
compatible with falsity.
64
If relations and terms are ultimate individuals
then they are unintelTigibie. What does this mean? Is
"the cat is on the mat" unintelligible? Bradley would say
that it is intelligible only because the context ["the
unity of the felt whole"] is implicitly referred to.
"Intelligible" implies that the context is somehow present
and is present as a unified system in which the term or
relation has its place. He spoke of "the ideal
transcendance of the given finite." This means that a
necessary condition to calling something "intelligible" is
the feeling of satisfaction that one has grasped the
system placing the meaning. As part of a system, a term
or relation is intelligible, and if not, not. It would
seem that for Bradley an intelligible term has its
meaning continuous with its implication; along that
continuum one 'finds' the essential term, and limits
meaning (from its implication) according to some prac
tical (or, in the case of philosophy, theoretical)
purpose [see section E, 4, above).
What is this "implication"? It is what Bradley
called the "self-development" of the term or relation,
verifiable in the process of "ideal experiment"59 which
calls into mind yet other ideas from which the specific
meaning cannot abide in utter isolation.By
"meaningful" one understands that other ideas accompany
the initial idea in a way that is not merely contingent.
The term is accompanied by other content of meaning in a
way such that a system or series is presented, of which
both initial and accompanying ideas are understood as
members. "Red" is meaningful, for example, as a member
of the color series; "color" in turn is meaningful as
aligned with the series of modes of sensation; "sensation"
in the series of other modes of experience, and this
goes on indefinitely to the degree that "red" is meaning
ful. "Red" as wholly meaningful encompasses, to the
uttermost detail, the whole of whatever may be meant by
"reality." A term is intelligible to the extent that all
this is involved. Intelligibility springs up entire
out of the unity of the whole. It is not built up
piecemeal.
Is "feeling" intelligible?*** No, not as mere
62
feeling. But the question of intelligibility arises
only with the attempt to understand. There is a shift
from feeling to terms-in-relation. To understand, as
Bradley put it,
Is to have the real as something before us as an
object [my emphasis] so that our mind can pass,
as to its diversity, from one of the many
differences to another, and from each and all
to the One, and from the One to each and all—
and do this without thereby altering the real
. . . or . . . to show a jar or break in its
continuity.
But in feeling
We may have . . . so far no object and in any case
no passage from this feature or aspect to that.
66
We have simple qualification of each by the other,
and all by the whole which they qualify? and
there is hence no question of any intellectual
contradiction nor any inquiry as to 'how' . . .
[but] . . . it is 'unintelligible', so far, only
in the sense that to try to understand it is to
transform it into something beyond itself and
something not itself.®-*
It is not possible to remain in the merely felt for one
demands understanding. Abstraction and analysis are then
involved (as an inborn tendency of thought) in the
attempt to understand, and so, are justified. Yet,
neither is it possible to remain merely among the
products of analysis and abstraction. A unity is sought
for, and a passage; and so far as anything yet remains
outside, as an impediment to passage, there, one may say,
is a region of the unintelligible.
Bradley wished to deny that any instantaneous
apprehension ever is intelligible if taken merely as
such. Now in contrast to Bradley's view there is, as we
know, the view that implication need not be a function
of ideal self-development, nor need it be a function of
an idea in a system. It has been held, in contrast to
Bradley, that system among ideas is not prior but
posterior; that a logic may be built up from individual
parts, having each its own nature, separate initially and
only "materially" related— that is, externally. It has
been held that the requirements of meaning are satisfied
by the- device of "denotation": by the coupling of ideas,
vacant in themselves, with existents. Once this
one-to-one and piecemeal correspondence has been set
up, then and only then can system be brought in.
The use of such words as "thisii " "that,',' "there,"!
"here* V "now>" "mine)’! ' with perhaps some others,
indicates "denotatively" that something actual and
immediate is in hand. To be sure, everyone understands
such a phrase as "this is my hand," but the question is:
"If universality, wholeness, and system were not already
present on both sides of the copula, then could it ever
be got at all?" And conversely, "If immediacy were not
already present with the judgment, could the mere sign
get it inside of the judgment?" Bradley's answer to both
questions was, "No."
But feeling is not easily obtained in its pure
form, and the inveterate habit of the mind is to apprehend
feeling, as it apprehends all things, through terms-in-
relation. Our 'thinking' is mostly habitual, and our
'feeling' is usually a mummified replica of contact with
original being. It becomes easy to believe that "this"
is exactly equivalent to the concept by which we take it,
and to believe in "denotative" definition. However
difficult it may be, we must re-learn to distinguish the
"this" of actual immediacy from its mere conceptual
symbol; the feeling as felt from the feeling as thought.
If we wish adequately to depict that act of centering
attention, from which the "denotative" definition of terms
68
arises (meaning's ultimate and last appeal, many
believe), we must remember that at that level nothing is
given separately, but rather it is taken separately.
Hence, denotation affords no springboard for the leaping
into being of discrete and individual pellets of
intelligibility.
Feeling is unstable, giving rise to thought,
and out of the plurality of moments which spring up in
reflection a selection is made. In selecting, one may
vaguely remember that it was out of felt immediacy that
the selected abstraction sprang; and, because the
abstraction is singular, it may be supposed that it is
grounded by an equally singular or discrete "this';"
Thus, by reversing the wheels of the whole process, one
arrives at the 'ultimacy1 of "denotative" definition,
makes "connotative" definition dependent upon it, and,
because the latter is remembered to be dependent, it is
presumed that the former is not. This process, whereby
in "denotative" definition one has made a show of
grounding the ideal in the actual, emboldens one to
declare that "This is a cat" is a categorical truth able
to stand alone. One thus convinces oneself that there are
ultimate, simple, and singular meanings that spring out of
"denotative" definition and survive, in " connotation ' I, "
64
with a life of their own, true to reality. To reason,
to reflect, demands this: The subject is cut off and
69
lifted out of context at the margin beyond which further
contextual consideration seems to make no difference to
practical purpose. But difficulty results when one for
gets that the subject (actually only a grammatical
subject), occurring in and conditioned by its context,
65
is conditional and dependent.
In any relational situation the relation must
couple actual existents, and judgment proceeds as though
what were taken from the situation were truly given there
in a form and content independent of it. One presumes
that form and content continue as they were, adequately
expressed by terms-in-relation. But how can the
relational coupling be had without reference back to
an underlying unity that makes the reality? Without it
one has no subject, for the terms are only conditionally
related and so, are not the reality of which our state
ment of terms-in-relation is true. Although what we
call "denotation" is not useless, it will not serve to
give us the real subject for assertion— a subject not
conditional and abstract. Consequently, "this" does not
ground the term. The ground of the term, correctly so
called, is all those conditions coming together making
the actual existent. "This )," even though it were un
problematic otherwise, would not give them to us. It is
the prior unity of the felt totality that we want to
indicate with our "this," to ground the term and to insert
some "what" into living actuality. But "this" does not
even suggest the conditions of actuality, and so, remains
abstract.
If we are to believe in the sufficiency of
denotative definition, we must go so far as even to deny
the relevance of the prior unity of felt totality to the
actual grounding of our term. Let us try to take that
denial seriously, fearlessly concluding that the felt
unity no longer has any business being even implicitly
present. But now we have on our hands an interesting
situation. On the one side of an unbridgable gulf lies
infrarelational experience (feeling), and on the other
side lies the denotable world,, rationally taken (as terms-
in-relation) . The latter has no need of the former. The
former is declared to be not only infrarelational but
irrational, and for that reason of no concern. But now a
problem arises that is all-pervasive for relational
thought, breaking apart not only terms from actual
relation with other terms but even dividing each term
from its character. It can be nothing at all. As
grounded by designation, it is still presumably "this."
But "this" as merely designative is no definition unless
by its means some quality is added to "this;." A "this-
what" or a "this-and" is needed by any subject permitting
inference. Now the "— what" or the "— and" is the very
quality that a proper name has not got.
A separation has been thrust between
substantive and adjective that destroys meaning and takes
inference with it.
An illustration may help here. Let the separation
between felt unity and relational thought be taken
seriously. Consider a case where one would say "this
red." Here is a redness that, in its positive reference,
entails no position in any color series,in its negative
reference entails no exclusion from them (for as an
allegedly atomic and simple meaning it does not
involve them at all)— yet it is supposedly 'grounded'
through a "thisness" which cannot refer thought back to
any further intelligibility in larger experience. No idea
is "connoted" by a logically proper name. Then how can
the bare concept (or whatever) be called intelligible?
Even a question-begging label like "material implication"
is indefensible. "Red" is shut up all in itself, and
thought can do nothing with it. No process, no passage,
no series, no implication among ideas is possible: no
reality can be got to. Because the unity of the felt
whole has been forbidden to remain even tacitly on stage,
red must cease to be a color. It neither affirms nor
excludes, and so cannot relate to, yellow, green, and
blue— and "this" grounds nothing. "This" becomes the
vaguest of all universals, applying indifferently to all
alike; "a mere unspecifiable X.
72
tt. Summary of Chapter X
The conclusion may be dravm from Bradley's
arguments* presented in the seven preceding sections,
that reality cannot be a casual collection of independent,
self-existing, individual entities.
The arguments run as follows: Suppose that
reality were made up of individual entities. One assigns
to them terms,in order to speak of them? but now the
terms are independent, one of another, and thus are
independent of relationship. The relation, which cannot
fall within the term as an essential quality, must like
wise be independent of the terms. Section D gives
Bradley's statement of this conception of things, taken
by the intellect to be ultimate and true, which he
called, "The relational view stated abstractly."
But the relational view contradicts itself.
Section E gives five ways by which it does this:
1. Every relation is and is not the whole
situation.
2. Terms and relation both do and do not
qualify one another.
3. Every relation is multiple indefinitely.
4. Every relation is both external and internal?
both extrinsic and intrinsic.
5. Every relation both is and is not capable of
degree.
73
Because of these contradictions, the proposition |
summarizing Section F follows: "The relational view is
not another ultimate form of unity."
If the relational view posits a plurality which
contradicts itself and so cannot be real, then what are
terms and relations? They are provisional entities:
stopping places for thought in its attempt to understand.
They 'are' to the extent that they qualify and express
their ground, which is immediate experience, or feeling.
They are meaningful so long and so far as that ground
continues to be called up by them. As stated in
Section G, relational thought stands on immediate
experience while attempting to supersede it. Thought
separates "this" from "what" to concentrate upon and use
the "what." Thought uses ideas, but it uses their
reference, not their existence. It claims to detach the
"whats" from reality while implicitly assuming that they
refer to it still. It takes its "whats" separately, but
it cannot put them together again, to deal with reality,
except by impliaifc reference to the unity of the felt
whole.
Section H gives Bradley's rebuttal to a recent
attempt to solve the riddle of "this" separated from
"what" by the device of denotative definition. But that
device runs into difficulties making it even more plain
that relational experience cannot be taken either as
74
; consistent or intelligible in itself. The intelligibility;
of words used merely as proper names is questionable, for
proper names supposedly connote nothing. But neither
can proper names ground ideas in reality, for they are
abstract and discrete.
The next chapter will be devoted to an expli
cation of the meaning of the word "reality." That word
has been much employed in stating and defending Bradley's
position, here in this first chapter, on the assumption
that it has a meaning such as Bradley believed. But that
assumption has been challenged in recent philosophy, and
if the challenge is well founded then the result could be
fatal to Bradley's whole view. In the next chapter I will
try to show that "reality" has the singular meaning that
Bradley intended.
75
FOOTNOTES— CHAPTER I
1The word "fact1 1 is used here in its broadest
sense. The "thing" as a "fact" is the thing as it is in
reality. Whatever really is true about it makes the fact
and is the fact. Thus the fact includes the thing's
history, its context, and whatever may be included under
"feelingi." Try to deny any of these to any actual fact.
See Appendix B, "What Is A Fact?"
2W. A. Sinclair argues for a similar conclusion
with respect to "things". Says he, "'The tree', 'the
pipe', . . . are not independently existing separate
things . . . [but only] . . . names for certain
selections and groupings within the situation I experi
ence because I follow these ways of selecting and
grouping and not others." Our judgment of some "thing"
as a real fact, he argues, is a function of our
epistemological attitude of highest generality.
It is arbitrary in the sense that it serves our special
purpose. An epistemological attitude, even of high
generality, does not exist to give us reality, but rather
to select from reality in a most useful way. Only from
within such an atittude can we order "knowledge" and say
that this or that is or is not a "thingi" See his
The Conditions of Knowing (New York: Harcourt, Brace and
Co., 1951), pp. 78-79.
2See Appendix A, "What Is An Idea?"
4The problem of "reality" is taken up in the
chapter to follow. The passage above, in failing to
place reality wholly either in the immediate or the
reflective regions of awareness, should not be taken to
mean that reality is not wholly present to us somehow.
The intent above is to establish a distinction between
the relational view (reflective) and immediate
experience.
5"Feeling" was used by Hamilton to name that
third great division of mind that was neither conation nor
cognition. Bradley differed from that usage by making
feeling prior to cognition or conation in referring to
that whole unity of experience within which distinctions
are later made by thought. It seemed to Bradley, as it
seems to us, that this prior unity was felt. Bradley
used "feeling" to refer to our sense of the coherent
76
i structure or experience within which we find meaningful
; terms to relate, and upon which their significance
: continues to depend.
^See Appendix F, "Subjectivity or Objectivity?"
h . Bradley, Collected Essays, "Relations ,,"
2 vols., ‘ ed. H. H. Joachim (Oxford: Clarendon Press,
1935; rpt. Freeport, N. Y.: Books For Libraries Press,
1968), 2: 628-76. Hereafter cited as "RelationsCE.
^Bradley, "Relations," CE, pp. 633-34.
^Ibid., p. 645.
^-®See Appendix C, "Sameness and Difference. "
■^F. H. Bradley, Appearance and Reality, 2nd ed.
(Oxford: Clarendon Press, 1897), p. 150. Hereafter
cited as AR.
■^Ibid., p. 159.
■*-^Ibid., p. 171.
■^F. H. Bradley, The Principles of Logic, 2nd ed.,
2 vols. (London: Oxford Univ. Pr., 1928) 2:605.
Hereafter cited as Logic.
■^F. H. Bradley, Essays on Truth and Reality
(Oxford: Clarendon Press, 1914), pp. 193-94. Hereafter
cited as ETR.
16
Bertrand Russell, The Problems of Philosophy
(New York: Oxford Univ. Pr., 1912; Galaxy Books, 1959),
pp. 97-99.
■^Bradley, "RelationsCE, p. 637.
18
M. S. Gram, "The Reality of Relations .1 1 New
Scholasticism 44 (Winter 1970): 49-68.
19
Why not meet this assertion with a flat denial.
Suppose we say that the relation is a universal and not
any unique fact or situation. The relation has existence
as universals have, and not as any sentient fact; and
Bradley's whole conundrum over relations rests upon a
misunderstanding of what a relation is, and what a
universal is.
A reply for Bradley would go as follows:
77
Universals do exist/ and are used from the first by
cats and dogs as well as humans. In themselves uni
versals are abstract, but they do not exist if merely
abstract, and that is the point at issue here. Thus, to
refer to the existence of a relational universal is to
refer to the relational situation. Further, the relation
(dependent for its affirmation upon the actual situation)
is entered into by the situation in the sense that it is
modified, or changed, or somehow genuinely conditioned to
it. Even though there may be cases where no difference
is discriminable, between situation A and situation B,
that does not entail its independence.
^Bradley, "Relations,” CE, pp. 638-41.
21
See Appendix D, "Was Bradley's View Incompatible
with Science?"
22
The term "relational property" is Moore's, and
it will be discussed in a later chapter. See G. E. Moore,
Philosophical Studies, "External and Internal Relations,"
ed. C. K. Ogden (London: Routledge and Kegan Paul, Ltd.,
1922; rpt. Littlefield, Adams and Co., 1968), p. 281.
2
JThe difficulty cannot be escaped by retreating
to the now-popular position that philosophy is linguistic
and "merely descriptive-" No one would claim that the
description is the fact, yet on the other hand a
description HTat does not seize the actual bodily is not
worthy of the name. It cannot be a true description. It
may sound strange to say that the claim in describing, the
aim in knowing, is to identify with the object; but this
is because we have ignored the absurd implications of
the alternative. Is the truth of the description not
some identity between it and the thing described? Is the
description independent in any sense, to the degree that
it is true? Is it worth calling a "description" if
identity with the object is missing altogether? These
questions seem to answer themselves.
^Bradley, "Relations," CE, 2:638-41.
^One might dispute over whether "appearance" is
the right word here, and I am open to suggestions.
"Illusion" will hardly do.
26See Bertrand Russell, Our Knowledge of the
External World, revised ed. (London: Allen & Unwin Ltd.,
1926), pp. 58-61. Note that it is the difference
retained by an asymmetrical relation that forbids reducing
78
it to a mere property of its term. Difference is the
crucial aspect determining relatedness.
2 7
'Here is Bradley's famous Endless Regress
Argument. It is sometimes taken to be the whole soul and
substance of Bradley's objection to relations, but I will
not discuss it here beyond what is presented above. It
will be found in the first two chapters of Bradley's AR,
and I presuppose the reader's familiarity with it. I
take the Endless Regress Argument as the nominal summary
of, rather than the whole of, Bradley's objection to
relations.
2®Bradley, "Relations," CE, 2:641.
29Ibis., 2:659.
30Bradley, ETR, pp. 229-30.
31
Of course, no one would think of linking two
toots, or two thoughts, with the relation "north of."
The illustration is absurd— which is precisely the
point. If the relation is merely external, what dif
ference does it make what it comes between?
33Bradley, Logic, 2:697.
33
Let us distinguish between (a) internal
relations as they were criticized above, and (b) internal
relations in the sense that Bradley preferred them over
external relations. (a) As used above, a relation is
internal if it is a property of the related term.
(b) As preferred by Bradley, a relation is internal if it
is a property, not of the finite term, but of Reality.
The relation enters or modifies the term only in the
sense that part of the term's meaning includes this
relation that it has: a relation, within Reality.
(b) is the sense in which Bradley upheld internal
relatedness. He never defended it in sense (a), as can
be seen from his earliest writings. Critics have taken
him to have advocated sense (a).
3^I take "view" as inclusive of "purpose." Our
view of something is, after all, a portion of the whole
concrete existence, and so, if included, is nearer to the
'fact'.
35
See A. E. Talyor, Elements of Metaphysics
(London: Methuen & Co., Ltd., 1903? rpt. New York:
Barnes & Noble, University Paperbacks, 1961), p. 55.
79
36gee Bradley, "Relations , " CE, 2:647, 667-69.
! The whole relational fact includes the purpose in
selecting just these terms and relation for emphasis,
I because surely the principle of limitation ought to be in
evidence if the selection is not arbitrary. Shall the
"fact’ ;," however, be limited to ohly the terms-in-
relation? Such an interpretation of "fact" has been
rejected in this essay because, by "fact;)' one had thought
that one had meant to refer to reality. But the matter is
confused. How much of reality shall be called "fact"?
In this essay it is assumed that one meant all of it,
for it is all relevant. But shall one distinguish between
"fact" and "conditions" for the fact? Where does one draw
the line? By what principle may it be known that just
this much and no more is to be called "fact"? The rest
is to be called "conditionst" or something else. For the
principle of distinction, I have appealed to "purpose"
(inherent in "view"), for I regard "fact," exclusive of
the principle of limitation, as arbitrary and ambiguous.
But no fact can be arbitrary or ambiguous. It, of all
things, one had thought, is precise. Then some clear
definition of "fact" seems in order. What is the
difference between a "fact" and a "state of affairs"?
"Fact" seems singular, but "state of affairs" seems
plural. "Fact" sounds like something one and whole,
while "state of affairs" sounds vague, with many things
entering in. Is a "state of affairs" the objective part
of "fact"? Or, is it the objective part of a "propo
sition"? Does the "fact" come between the "proposition"
and the "state of affairs"? How can it, if the propo
sition is the objective part of a "judgment"? Or, is
"judgment" now out of fashion? At any rate "fact" seems
ambiguousi
Fact?
; ^
purposive subject— judgment— proposition— state of
affairs— 'reality'? (things-in-themselves)
Does "factV" or does it not, include all this? Does
"fact" instead, fall somewhere between two of the above-
mentioned elements? If so, where? Some clear definition
of "fact" is needed. See Appendix B, "What Is a Fact?"
■^Bradley, "Relations," CE, 2:675.
38Ibid., p. 631.
39
The term "linguistic philosophy" leaves the
impression that there is a monolithic school or doctrine
80
called Linguistic Philosophy. It is not clear that this
is so, which makes "linguistic philosophy" difficult to
discuss. It is sometimes called a "movement" or a
"method" rather than a "school" or "doctrinei'.' In this
broad sense it was sometimes shared by Moore and Russell,
by Ayer, and especially by Wittgenstein, whose work was
taken up in the Cambridge and Oxford schools.
What is shared among them, that justifies
speaking of "linguistic philosophy"? It is hard to say,
but I suggests A linguistic philosopher is one who
believes that at least some philosophical problems arise
out of the misuse Of Words. "Linguistic analysis" is his
method, which consists (more or less) of comparing
philosophical usages to ordinary usages, to 'see' (in
some unspecified sense) whether the philosophical usages
have any meaning.
For a discussion of the development of "linguistic
philosophyi " from its inception in Moore and Russell to
its culmination in Wittgenstein, see M. J. Charlesworth,
Philosophy and Linguistic Analysis (Pittsburgh, Pa.: Du-
quesne U. Pr., 1961).
40
See Ludwig Wittgenstein, The Blue and Brown
Books (New York: Harper and Row, 1958) , pp. 4, 7, 23,
43. See also his Philosophical investigations, tr. G.
Anscombe (New York: Macmillan Co., 1953), pp. 31-32.
41
Bradley attributed this opinion to De Witt
H. Parker. Today, Gustav Bergmann defends the view in
the name of Russell.
42The personification of "relational view" (and
perhaps some other terms) is done for the sake of
simplifying discourse. We do not mean to imply that
"the relative view," etc., actually is the agent, any more
than linguistic philosophers mean to imply that a word
can mislead us.
^ B r a d l e y , "Relations/" CE, 2:641.
44Ibid., 2:631.
4^In Bradley's opinion, the "giving rise to"
process of the relational view out of feeling must remain
"inexplicable" or "not intelligible " [ETR, pp. 189,
239]. "It loads us with nothing worse than the in
explicable," he said, adding, "Immanence of the
Absolute . . . [is] . . . inexplicable," and "The unity
of immediate experience . . . is a one in many and a
81
many in one which for thought is not intelligible or
unconditional," "Immanence of Reality . . . remains
inexplicable [ETR, pp. 272, 278]."
He did not believe that this inexplicability in
any way weakened his criticism of the relational view.
^Bradley, AR, pp. 90-91.
47Ibid., p. 141.
48Ibid., p. 205.
^Note the word "further*" It is important to
remember that feeling, for Bradley, although the
beginning, is not the ending of what is real and whole.
Feeling and relational thought (discrepant if isolated
by analyiis’ ) must be and are reconciled in the higher
reality. Thus Wolheim's conclusion (that Bradley gave
emotion its revenge) is misleading.
^Bradley, AR, p. 205.
51
Ibid., p. 215. What critics overlook is that
for Bradley there is, in felt immediacy (the beginning
for all thought), the presence of indefinite and unitary
wholeness. Thus, it is mistaken to criticize Bradley's
Absolute as "a mere intellectual absolute": a creature
which follows as the conclusion drawn from certain
premises. Randall cites such a criticism with approval
in his attempt to show that in the end Bradley adopted
Dewey's "20th Century Naturalism," together with Dewey's
"denotative method." [John Herman Randall, "F. H.
Bradley and Working-Out of Absolute Idealism/'' J. Hist.
Phil. 5 (July 1967): 245-67, See esp. pp. 257-59]."
Unity is both felt and rationally required.
There is no discrepancy between feeling and thought on
this point. Dewey's argument, approved by Randall, was
that Bradley's Absolute could only be had as the
conclusion from dialectical argument. The dialectic
first assumes a relational multiplicity which it then
criticizes and denies— which then requires an "absolute"
to unite the discrepant pieces. That makes Bradley's
Absolute "a mere intellectual absolute" whose claim to
actual existence can be no stronger than is the truth
of the initial premises— which in the end are denied.
But Bradley had replied to this criticism years before,
as Randall even mentions.
52
Bradley, AR, p. 307.
82
53Ibid., pp. 406-07.
54
Ibid. Note here Bradley's reason for
criticizing designative definition.
^Bradley, "Relations> " CE, 2:633.
3®See Bradley, AR, Ch. IV and XVIII. Bergson
has distinguished between true and "spatialized"
(objectified) time. But without objectifying it, it is
no "what’ ." Why call it "time"? Why not, with Bradley,
call it "feeling"?
^Bradley, "Relations," CE, 2:658-63. See
note 8.
5 p
But, we object, there is contradiction, in
every matter of fact. The fact, Bradley said, is one,
but terms in relation are many, and it is in feeling that
the many are brought together. Feeling is one. Very
well. But isn't it flatly contradictory to say that the
many are one? Feeling is one; terms-in-relation (with
or without the hyphens) are many. "Many" does not mean
"one," nor does "one" mean "many." It is flatly
contradictory to say it. How, then, can Bradley say that
the togetherness of feeling and thought (one and many)
"loads us with nothing worse than the inexplicable"?
But now, enter the Absolute upon the scene. It solves
the problem. All contradictions dissolve into the pink
mist. But is this appeal to a deus ex machina to be
allowed? Is it that there is no contradiction between
one and many, by grace of the Absolute? With such a
handy helper as the Absolute around, why need any
philosopher bother with consistency?
The answer, at the moment, is not quite clear.
It seems to have been Bradley's view, however, that
(a) the 'fact' (a most ambiguous word) is not one as
taken, but is internally discrepant, and so, it is merely
an appearance; (b) the demand for consistency is an innate
demand of most minds, but whoever in philosophy rejects
consistency, cannot have it forced upon him. No ultimate
"how" or "why” troubles him; (c) both the unity of
feeling and the multiplicity of the entities of thought
are actually and undeniably experienced. They are not
the hypothetical creatures of Bradley^s philosophy. He
only aimed to point them out. The contradiction, once
you see it, stares you in the face. But, in philospphy,
a contradiction cannot be allowed. Something must be
appealed to, to satisfy the demand of thought for
83
I consistency (and, in primitive feeling, there is an
indication of its nature), even if that something be only
the naming of an Inexplicable. "In the end," Bradley
hands us over to Religion.
59Bradley, Logic, 2:599-601.
60Ibid., 1:85-86.
^ Mere feeling. But again, feeling as viewed
through idea is an abstraction.
62Bradley, "Relations,!' CE, 2:659.
6^"Denotative . . . connotation." Bradley
would not suffer this "slovenly terminology," saying
that "nothing but ambiguity has arisen out of these
perversions;!' He preferred the established usages,
"read in extension," and "read in intension.”
He reserved the best of what Blanshard calls his
"Bluebeard style" of rhetoric for "denotative definition,"
What was it that he found so offensive?
It comes to this: Nothing is ever 'denoted'
without having it 'connote' as well, so what justifies
making connotation dependent on denotation? We have a
world, and we can point to various things and qualities
in it, saying, "this is what I mean by 'red'" and point
to something red; "this is what I mean by 'round'," and
point to something round"; this is what I mean by an
'apple'," and point to an apple, and so on. But in each
instance the pointing would be the same! 'Round',
'red', and 'apple', as "whats," have got to be on both
sides of the copula, or, in other words, we have got to
already know these words before we can begin to 'find'
them denotatively. That is to say, the meanings have
somehow got to be "latent," or what you will, or we
couldn't follow our denoted instructions. This old
insight, which Bradley did not claim to have originated,
has recently been rediscovered by Wittgenstein and his
followers. But see Bradley, Logic 1: 59, 169, 185, and
foil.
65Ibid., pp. 50, 100. "Our conditioned truth is
only conditional. . . . it hangs from a supposition;
I I
• • •
Shall existence be one thing and meaning
84
another, so that "this red" (the red that is) is aloof
from the problem of separable meanings? No, for it
raises the problem of getting the meaning and the
existence together again. Suppose this problem solved,
so that they can stand apart. But now red is aloof from
yellow, green, and blue. Then red is not a color, for
color involves all the colors. Red cannot imply a color
either, for what congress have the other elements with it,
that it is related to them? There is no middle term.
6^Bradley, Logic, 2:497. For "this" is also
taken abstractly. How else can it be taken? Nothing
immediate connotes an idea, hut the object of the whole
denotative process is surely to get an idea, in concrete
actuality, which the symbol "this" represents. The unity
of the immediately felt, let us remember, has been
expressly denied.
85
CHAPTER II
BRADLEY'S 'REALITY' AND LINGUISTIC PHILOSOPHY
A. Bradley's Meaning for "Reality"
Critics have objected to the arguments of the
preceding chapter that Bradley misused the term "reality"
in his arguments against "the relational view'.-" It is
true that Bradley attached a special importance to the
word "reality," and even a quite special meaning to it.
This section will clarify his meaning, and also explain
why he used the word as he did. Bradley made two
assumptions: first, that in judgment ideas are taken to
qualify reality, in the sense that "in judgment we have
passed beyond the state of mere perception or feeling?"'1 '
and second, that to have truth or reality is to have or
be something that is not true and real merely because of
something else.2 The meaning of these two assumptions
will become clear— and, hopefully, justified— in the
discussion to follow.
We begin by explicating two propositions which
he accepted. The first of these is,
(1) Reality is transcendent to every judgment.
Reality is supra-cognitive in the sense that the "what"
of any "this" is never fully disclosed. There is always
86
more of anything than what is known of it, and that
"more” is essential to the truth about it: essential to
the true significance of the terms used to speak of the
known. Beyond that, reality is supra-cognitive in being
also felt, and this aspect of experience is further
complicated by not being confined to any definite feeling.
Feeling is not something with clearly definable limits
and, in its way, includes everything. To get reality
into our "what" we must somehow get it as felt and also
as thought— and get these two divergent aspects together
in some intelligible way. Bradley did not deny the
relational side of experience. Said he, "Relations do in
3
fact exist and immediacy is in fact transcended." Thus
there is in reality a duality of feeling versus cognitive
terms-in-relation, and neither side can be denied. Yet
neither side is self-sufficient: each wants explanation,
and neither side is fully explained by the other. This is
so for terms-in-relation (apropos the preceding chapter),
but so as well for feeling. It is hard to understand.
Feeling exists in finite centers, which cannot be finally
explained. Neither can it be explained how feeling is
transcended, in the sense that it "contains/" so to
speak, "a world which goes beyond itself."^ So, unless
one is prepared to say that there really are "surd,"—
that reality even when thoroughly known yields no
intelligible answer to legitimate questions— then it must
87
be that neither in thought nor in feeling have we got
reality. Two outstanding questions of "How in the end?"
remain. How does one unified feeling embrace many
existing things in a finite center? How do terms and
relations maintain each their separate identity and yet
combine to make one individual fact?
The second proposition to be discussed is,
(2) Reality is the subject of every judgment.
In judging, we mean to predicate an ideal content to a
c
real subject. But no judgment is being made if the ideal
content is taken merely as some existing mental event,
and not as a meaning that refers beyond itself. A
"merely floating" idea, taken all as content but with no
reference beyond, cannot with confidence be attached to
any subject.® Every idea has two sides: existence, and
reference. It is not the existence of the psychic content
which is used as an idea, but rather, the reference.
The idea, taken logically, is not an event but a symbol.
7
It refers beyond itself to reality.
An idea is some psychic content detached and used
as a meaning. By predicating this ideal content of some
subject, an appearance is being meaningfully related to
reality. Thus "idea," as the word is used here, is not
merely an image or a mental picture. The idea, as a
meaning, is a fragment of a whole, initially felt, brought
to attention as an object of thought; and by predicating
88
it in an act of judgment one tries to reconstruct the
original felt whole in a way that may be called
objective, explicit, or intelligible. To do this, one
uses ideas as adjectives to qualify what is other than
0
their own being. An idea is a content separated from
reality, and in judgment one tries to restore (on the
level of intellectual apprehension) the broken unity. By
doing so successfully, one gets understanding. An idea
is always a predicate and always a universal. An idea
is never the subject of a judgment, for the aim in judg
ment is to say the truth about a subject: to qualify
reality with some meaning.
But is not "glass is breakable" a judgment?
Here, plainly, it is two ideas that are being joined
together, and not one idea being predicated of reality.
How could Bradley believe that judgment is always one
idea predicated of reality? How could every idea be a
predicate and an adjective, when "glass" here is a
subject and a noun?
In reply, we must point out that "glass,."" used
as the real subject, is not being used as an idea, i.e.
as a universal. But if "glass" were being used as an
idea, it would be a predicate and an adjective. If we
actually 'judge, "glass is breakable,!' the idea of glass
is not the subject. How can an idea be breakable? We
mean to say that the real glass is breakable. So far as
89
"glass” is an idea, it qualifies real. It indicates
that a portion of reality is qualified by whatever is
significant of glass. "Glass" is thus a predicate and an
adjective.
There is still more reason to say that reality is
the true subject of every judgment. Reality is what
supplies the conditions for coupling ideas. Conse
quently, any finite subject such as "glass" is not the
true subject. Another subject is implied, transcending
this merely grammatical subject: reality itself. The
finite subject cannot itself alone be presumed to supply
conditions sufficient for attaching its predicate; and,
if not, then it cannot of itself be said to own that
predicate. Mere "glass" does not imply "breakable ," nor
is "breakable" limited to "glass." Glass can exist
under conditions where it is not breakable, and
breakability is not limited to glass. There are con
ditions to getting "glass" and "breakable" together,
which conditions their mere meanings do not supply.
Presumably the judgment is true, but neither nominal
subject nor predicate justify us in saying so. But
apart from that justification the judgment may as well
be called false. As Bradley said
With every finite subject, the content of
that subject is and passes beyond itself. Hence
every assertion made of the subject implies that
which is not contained in it. The judgment in
other words is made under a condition which is not
90
specified and is not known. The judgment, as it
stands, can therefore . . . be both affirmed and
denied.*
This state of affairs must be taken into account
in any statement of the logical form of judgment. For
this reason Bradley argued that the form of judgment is
not simply "S is P," but rather it is, "Reality is such
that S is P." The "Reality is such that" allows for the
inclusion within the judgment of the necessary conditions
of getting S and P together.
Why could not "This apple is red" serve as a
true, explicit, or "categorical" judgment? How is it
compatible with falsity? What does it add to say,
"Reality is such that this apple is red?" Here is an
apple. We can see that it is red. Why demand extra
embellishment?
We repeat that "Reality is such that" adds the
conditions necessary for joining "this apple" and "red"
together. "This apple" asserts no real apple unless we
mean to imply that words like "this," "here," "now," and
so on complete the story of how, indeed, there is an
apple. The story is all there, in reality, but is it all
present in "this"? Does the meaning of "this" bring
those important conditions into the judgment? No. It
does not assert them at all. It merely presumes that
they are present and satisfied. It does not satisfy them.
91
The meaning of "this" can apply to anything and every
thing depending ori conditions. As a meaning, "this" is
viciously circular. And again, any ideal apple, if by
that we mean the content of some mental event— a content
present whenever we think of an apple— is "this apple*"
Are the conditions there that make it red? Think of them
as you please. Mere "this" does not make it a real
apple. We desire that "this apple” be coupled with
"red:" But because, apart from the assertion of
reality present and "such that)" "this apple" means mere
"apple," then it is only possibly red. The apple is
possibly red, but it is also possibly yellow or green.
The judgment "This apple is red" is compatible with
falsity.
Because reality must be the ultimate subject of
every judgment, there is but one real subject for all
judgments. There is no plurality of subjects. All is
one. Bradley's reasoning was that if there were a
plurality of real subjects, then judgments about them
must take the form, "Reality is A*" "Reality is B,"
"Reality is C»" and so on. But further, if there were a
plurality of real subjects, then by "Reality is A" we
must mean "Reality is mere A." In that case, "Reality is
A" is contradicted by "Reality is B," for B, likewise,
must mean mere B. "Reality" is a singular term, but
mere A, B, and C state a real plurality.^
92
Is this argument vulnerable to linguistic
criticism? The linguistic philosopher would object, to
Bradley's argument, that it is only the word "reality"
that is singular. To presume that every meaning must be
singular is an exploded superstition. Was Bradley right,
or was he basing metaphysics upon a merely verbal require
ment? That question will be pursued in this chapter.
First, however, it will be helpful to summarize Bradley's
argument that reality is the transcendent subject of
every judgment. The argument may be phrased as follows:
(1) A meaningful term of any judgment (an idea)
has two aspects: identity (psychic
existence) and reference (that to which the
idea refers)
(2) The coupling of contents in judgment is
justified by reference; that is, by that to
which the idea refers
(3) The truth of any judgment is dependent upon
what lies beyond it, implied but not stated,
as its conditions
(4) If those conditions are absent, the judgment
is false
(5) The grammatical subject of any judgment does
not assert those conditions
(6) The grammatical subject does not, of itself,
own the predicate of the judgment
(7) That which supplies the conditions of
predication owns the predicate
(8) Reality in toto is that which supplies the
conditions of predication
/, (9) Reality is the unconditional subject of every
judgment
93
but (10) the total conditions of predication are
never experienced
(11) Reality is the transcendent subject of
every judgment
Bradley himself determined the meaning of
"reality" through a critique of the conditions for
meaningful assertion. Later philosophers also centered
their attention upon the problem of meaning, as a problem
of prior importance. They became known as "linguistic"
philosophers. Their investigations led them to deny
meaningfulness to the word "reality" in Bradley's sense.
In this chapter the criticisms of two linguistic
philosophers will be examined. G. E. Moore will be taken
1 2
as the originator of the linguistic method, and
J. L. Austin will be taken as the spokesman for it at it
height. Both these excellent thinkers had much to say
about the philosophical use of "reality." We must
determine whether or not their criticisms seriously
undermined Bradley's position.
B. Moore's Objections to Bradley's
Use of "Reality"
The issues Moore raised will be discussed here at
length. His criticisms changed the whole course of
philosophy. His critical discussion of "reality" is
directed mainly against Bradley, to whom Moore referred
by name. Moore could not accept "reality" as Bradley used
it for several reasons, which I shall treat separately.
94
They are as follows:
1. Bradley Used "Real” in Two Senses, Contradicting
Himself
Moore found Bradley's view strange. Bradley
seemed to be saying that nothing ever did or ever could
exist in the way we ordinarily suppose, but rather must
exist in some timeless sense utterly removed from our
world. Said Moore,
We do find it very difficult to conceive how
anything could be truly said to exist or be real
at all [my emphasis], if it neither exists or
existed in the past nor in the future? the notion
of a timeless existence is certainly a very
difficult one to grasp.13
What is 'real' for Bradley must be timeless, Moore
reminded us, for on Bradley's view existence in time is
relational and hence contradictory. Note, however, in
the quotation above Moore ran together "exist" and "be
real ," which he did throughout his discussion. To
Bradley, on the other hand, these terms meant different
things and must be kept separate. The thing that appeared
in judgment as the grammatical subject could exist but
it was not real. This differing opinion of the meanings
of "real" and "exist" was characteristic of Bradley
14
versus Moore, and must be kept in mind. Later on we
must examine Moore's attempt to dispose of the distinction
between "real" and "exist.”
In his essay on time, Moore concerned himself
95
with problems of practical import. Inasmuch as
practical problems are temporal problems, the question
of whether or not time is real, and in what sense, is
dealt with. We may agree with Moore that no philosopher
who denied the reality of time would be "able to divest
himself of particular beliefs which contradict that
15
opinion." He would still believe that he ate his
breakfast before he ate his lunch. If the philosopher
then went on to say that, nevertheless, time is unreal,
then either he would be denying with his mouth what he
affirmed with his actions, or else he would be using
"real" in two senses. One sense of "real" would describe
our world, the other would apply to some region we know
nothing of. Bradley, Moore believed, was stuck with two
senses of "real. "
2. Bradley Was Misled by a False Theory about Words
Moore pointed out that "reality" is a word whose
meaning people do not ordinarily question. They use the
word confidently without even trying to analyze or define
it, and seem to have no trouble. But Bradley took great
pains to define "real"; and, because his definition failed
to apply to ordinary things, he declared them to be un
real.
Bradley was wrong to do that, Moore believed,
because not all words are definable. To be definable a
word must be analyzable, and so, compound. Many words are
96
not analyzable, such as "life/" "good," and "yellow,"
Might that not be Bradley's difficulty? Perhaps "real,"
"existslr" "is a fact," and "is true" express meanings
which are not compound/ and sof not definable. But need
they be defined? No, Moore implied, for,
I say, that the question as to the definition of
these words is not logically relevant to the
much more important question as to which of the
sentences in which they occur are true and which
false.
And lower down,
A man's views on more important questions may,
therefore, be very much influenced by his views
as to the meaning of a word. 6
Clearly, he was suggesting that Bradley was misled "on
more important questions" by a mistaken belief about the
meanings of words.
According to Moore, definitions are not possible
for all words, but only for those with complex
meanings.^ Complex meanings are to be got by putting
together simple ones. But the simple notions themselves,
such as "life," "good," "yellow,?' "real," "exists," "is
a fact," and so on, are indefinable. They have no parts
and any attempt to define them must end in confusion,
for one inevitably has tried to say what something is by
restating it in terms that it is not. It seems that
Bradley's 'proof' of a suprasensible "Reality" had been
arrived at on account of his mistaken notion about words.
For all the brilliance of his argument and for all the
97
historical respectability of his view, he was nevertheless
confused by a verbal difficulty: the indefinability
of "real" or "reality."
3. "Reality" is Indefinable
It has already been noticed that Bradley had a
different meaning for "real" than for "exist."
Philosophical consistency forced him to distinguish them,
he believed. This implies that he believed both words
were definable, and that a philosopher has a right to
define words according to his own requirements. It also
implies a unitary meaning, at least within what has be
come known as a "universe of discourse!" But in meta
physics, furthermore, where that universe is unrestricted,
our words must have a unitary meaning everywhere.
Bradley did not think that words .must be so strictly used
in practical matters. Moore, by contrast, said explicitly
- » ■ >
that he was concerned with practical matters.“ Were they
not, then, addressing different points? It seems so.
But Moore went on to attack the notion of unitary meaning
in Bradley's position.
Let us examine Moore's criticism. He argued that
Bradley used "time" with contradictory meanings. Bradley
called time (a) a "contradictory appearance" and so, not
18 19
real? but he also said that time "belongs to reality."
This seems to make nonsense of any distinction between
"real" and "exist." Even if Bradley were allowed to say
98
that time "exists" but is not "real>'f how could he go
on to say that time "belongs to reality"? Is the time
that belongs to reality the same time that exists? Then
why deny reality to it? But if it is a different 'time'
than the one that exists, then Bradley used "time" in
contradictory senses. Consistency required the
distinction between "real" and "exist/" but it makes for
a contradiction in "time.1 ! Thus Bradley gained nothing
in consistency, so he may as well have left "real" where
he found it in ordinary common usage. Such is Moore's
argument.
Moore did not claim that there was no difficulty
with the meanings of "real" and "existJ"' He merely
20
pointed out that Bradley's distinction gained nothing.
What i£ the correct meaning for "real"? Moore did not
say. He said, instead, that if we try to say its
central and invariable meaning, we fall into the trap
which certain words, make for philosophers. We become
puzzled. Upon analysis, "real" has several meanings.
We may even be mistaken to suppose that we use the word
in the same way, although we may have intended to.
Taking the trouble to define the word does not assure us
21
of using the word in a way always the same. No doubt
Bradley thought he was using "real" as he had defined
it; but nevertheless he vacillated, as may be seen from
the example of time which is both real and unreal.
The problem is not an easy one. The question is:
"Is the meaning of some word ["real"] called up in the
mind on occasion A the same as is called up on occasions
B or C?” 22 Moore seems to answer that there is no way
to say. We can't tell, for when we reach atomically
simple meaning, as we must eventually do with words like
"yellow," "is a fact," "realv" and "exist," then there is
no further standard of meaning and truth. The questions,
"What does this mean?" or "Is this meaning the same as
before?" when applied to words with atomically simple
meanings, are nonsense questions. No answer can be given.
They are often only rhetorical questions by means of
which some philosophers, misled by a false belief in the
equivalence of definability and meaningfulness, give an
air of plausibility to strange notions.
A definition is a criterion for using a word.
But words, like "realf" standing for simple notions,
don't need and can't be used according to a further
criterion. To prove it, Moore considered two suggested
criteria for "reality," namely: "mind-independence" and
"presence of connections . "
Suppose we try "mind-independent" as the
criterion for reality. Now we all agree that a horse is
real and a griffin is unreal. The horse is mind-
independent, but the griffin is not for the griffin
exists only as imagined. In this example mind-inde-
100
; pendence works as a criterion of reality. But it is not
the universal criterion, for it gets into trouble when we
try it on other examples such as acts of the mind. Are
they real, or not? We all judge them real, but certainly
they are not mind-independent. Said Moore,
My mental acts are certainly both in my mind and
dependent on it in a sense; they are not at all
except in my mind; and therefore it would seem
that if everything, which is only in the mind or
dependent on it is imaginary, my thought of a
griffin is just as imaginary as the griffin himself.23
Mind-independence fails as a criterion for reality.
The other criterion Moore mentioned involves the
presence of effects or connections, in the case judged
"real," which would not have been present otherwise. For
example, a lion really present in this room has
connections or effects that are absent when dreaming or
imagining a lion. But this criterion, too, has a fatal
flaw. Its flaw is that the word "real" must be used to
state the criterion. The criterion, in other words, needs
the meaning of "real" as its criterion. The "presence of
connections" criterion says, "To say a thing is real, is
to say that it has some connection or other which it
would not have had, if it had not— (been real),".2^
That won't do. If the criterion actually defined
reality, the definition could be used wherever the word
"real" was used. That is to say, we could substitute
the definiens, "having some connections which it would
101
not have had if it had not— (been real)," for the defini-
I
endum, "real." We can't, for "real" appears again in the |
definiens. If we try to substitute the definiens for j
"real," we get an infinite regress, for we have: "A thing ;
is real if it has some connection which it would not have
had if it had not— had some connection which..........ad
infinitum." The definiens in a good definition ought to
state the essential property of the thing defined. But
here it does not. No positive property is named, so there
is no definition here at all. The purported definition
only says that reality must have some property or other,
present with reality, absent otherwise. There is no way to
know what that property might be.
4. Reality Cannot Have Degrees
With the above result established, Moore argued
that it is false to believe, with Bradley, that reality
may be present in degree. It also is false to believe
that reality itself is the highest degree, the perfect
degree. Existence in degree must involve something which
admits of degree, but there is no way to know what that
something is or whether there is any one such thing at
all. Look into your own mind, Moore urged, when you are
sure of your judgment that something is real, and see if
there is any one property present but also present to a
lesser extent when something is judged to be unreal. Is
102
there any one thing? No. What one thing, for example,
is present to a higher degree when you judge a horse real
and a griffin unreal? There is not anything. Whatever
makes the difference is something "merely positive
25
Moore said, not relative. It is a difference m quality
or kind, not degree.
1, 2, 3, and 4 above are, then, statements of
Moore's criticisms of Bradley's conception of reality, or
use of the word "reality;1 . 1 It may be helpful to review
them before trying to reply to them.
1. Bradley contradicted himself by using "real"
in two senses. According to him, time is real ("belongs
to reality") but again it is unreal (a "contradictory
appearance"). But if time is both real and unreal, then
"real" must be occurring in two senses. Either that, or
"time" is being used in two senses. In that case Bradley
did not prove that time, or any other "mere appearance,"
is unreal at all. For the 'time' which he found 'unreal'
is not the time of the world of 'appearance', but a
special kind of time which never appears to us at all.
2. Bradley was misled by a false theory about
words. He assumed that all words were definable, and
that the meaning must be unitary wherever the word
occurred. But only words with compound meanings are
definable, i.e.,statable as a relation among simples.
103
; Terms with simple meanings, stating simple notions, need
! not and cannot be defined. To define them would be to try I
to state one simple notion by means of others which it is
not. It would create puzzlement where none was present.
3. One may not state the nature of reality.
Attempts to do so fail, as may be seen from examples:
(a) "Independent versus mind-dependent" will not do
for a criterion. It may sometimes serve, but it fails in
the case of the mental act which is mind-dependent but
also real.
(b) "Having connections versus absence of con
nections" will not do either, for (1) it specifies no posi
tive connection, and (2) as a definition it is circular.
4. Reality cannot have degrees. Immediate ex
perience discloses no one quality occurring in differing
degree when we distinguish real from unreal, yet our
ability to make that judgment is unimpaired. Hence,
Bradley's opinion that 'reality' versus 'appearance'
depends on degree, is indefensible.
C. Reply to Moore's Criticism
Before attempting a reply to the foregoing
criticism of Bradley, the reader may be wondering what,
precisely, Bradley's definition of "reality" would be.
Even after having read carefully the first pages of this
chapter, the reader might find himself agreeing with
104
Moore that "reality" is not clearly definable.
The "reality" of which Bradley spoke is meta
physical reality. It is, in the end, the only "reality"
worthy of the name— and this is no mere philosopher's
opinion. The great mathematical physicist Max Planck
26
argues for metaphysical reality very strongly. Putting
it into the simplest of terms, Bradley's reality is that
which is self-existent. Note that this simple definition
does agree with what we all ordinarily mean by saying that
something is "real . " But Bradley the metaphysician would
point out that most of what we call "self-existent” truly
is not, for its existence is conditioned. Things which
exist only on conditions which they themselves do not
supply cannot, strictly speaking, be called "self-
existent,"
If something is real, it is itself real and is not
real on account of something else. Otherwise, it is a
predicate of that which owns the conditions of predi
cation. Reality is that which owns all of the conditions
of predication. To call something "real" is to judge it
real, and judgment passes beyond immediate awareness as
mere feeling or sensation. The purpose here, in saying
these things, is not to argue these points (they will be
argued in pages to come) but rather to keep before us what
Bradley's position was. Perhaps Bradley's "reality"
still seems unclear. One could perhaps digress here to
10S
discuss at length the philosophical meaning of "clarity,"
but the task at hand is to examine Moore's criticisms of
Bradley's notion of reality. Are his criticisms of
themselves without difficulty? They seem not to be.
1. Did Bradley Contradict Himself By Using "Real"
in Two Senses?
Bradley may have been guilty of ambiguous use of
"reality," "exists," and "is a fact" in his earliest
work, his L o g i c He very carefully avoided such a
blunder, however, in Appearance and Reality, the work
which Moore specifically cited. Bradley did say "time is
unreal," That was a conclusion he reached only after
careful argument. Moore did not examine that argument in
his own article, "Is Time Real?" Bradley had found the
0 f t
concept of time contradictory. ° But Bradley did not
ever say "time is real," and Moore has not supplied an
instance where he did say that. In the passage quoted
by Moore, Bradley said, "Whatever exists must belong to
reality." He also said, "Anything that in any sense is,
29
qualifies the absolute reality and so is real." But to
belong to something, or to qualify something, is not to
be that something— is not to be qualified by it. In the
quotation just cited, time is real in the sense that it
qualifies reality. But Moore has turned the notion
around and has reality qualifying time. Surely these two
106
30
conceptions differ.
Moore presumed that reality inheres in some
31
things as a property. He did not get that from Bradley,
for Bradley maintained that "things" inhere in reality as
its properties. This is quite different. Thus, Bradley
has not contradicted himself by using "real" in two
senses. Time is not both real and unreal. "Real" is
used in one sense only, and "time" is used in one sense
only. To the degree that the conditions of actuality are
presumed present with the judgment wherein one says "time
is,1 1 time to that extent and subject to those conditions
qualifies reality, and so is real. But to the extent that
time is regarded in itself, in abstraction, as a separate
32
thing on its own, then it is contradictory. Replaced
in immediate experience, from which it was severed by an
act of abstraction, "time" is a useful idea and is, as
Moore held, central to practical concern. But regarded
as something in itself, which it never even appeared to
be, "time" is a contradictory abstraction which breeds
all kinds of bewilderment. Bradley's view of reality may
be difficult to understand but it is not impossible, and
it is not contradictory.
The important point of disagreement between
Bradley and Moore here is with what we ought to mean by
"X1exists." To Bradley it meant "X is a property of
reality.” To Moore it meant "Reality is a property of
107
I X .." For Moore it meant that existence (which, for him,
was the same as reality) could be predicated as a
property of some things. Now it had been thought long
ago to be settled that existence was not a predicate, and
so Bradley believed. But Moore disagreed. He expressed
33
his reasons in his article "Is Existence a Predicate , "
arguing that it is significant to say of something that it
34
exists. And it does seem so. We must look carefully at
his reasoning before proceeding.
Existence is plainly a predicate in the
grammatical sense, but is it so in the logical sense as
an attribute of some things? When we mention something
and add, "and it also exists, ’ ’ does the "also" here add
anything? Moore thought so, but others thought not.
Moore allowed that to predicate existence does, indeed,
sometimes lead to difficulty, because one can't always do
the same things with existence assertions that one can
do with other assertions. The existence assertion "tame
tigers e x i s t " can't be used in the same way as the plain
assertion "tame tigers growl. " One can go on to say,
"Some tame tigers do not growl," but one cannot say,
"Some tame tigers do not exist."
Russell had attempted to explain this difficulty
by his theory of the "propositional function." The
correct interpretation of "some tame tigers growl," said
Russell, is "'X is a tame tiger and X growls' is a
108
propositional function of which at least one value is
true." "Propositional function" refers to the logic of
statements of this sort, or way in which this sort of
statement is put together. The propositional function, as
Russell states it above, works for "tame tigers growl,"
but won't work for "tame tigers exist"; because it is bad
logic to say, '"X is a tame tiger and X exists' is a
propositional function of which at least one value is
true." 'X is a tame tiger and X exists' is redundant and
misplaced. According to Russell's theory of propositional
functions, 'X exists' is not a part of a propositional
function, but rather it is, or is expressed by, the
"existential quantifier," which may attach to a
propositional function. "Existence is essentially a
property of a propositional function [my emphasis],"
said Russell.^
Moore disagreed with Russell here in part. Said
he, "[This is] . . . part of what is meant by saying that
'exist', in this usage, does not stand for an attri
bute."^ All that Russell meant, to Moore, was that there
are some cases where the attribution of existence adds
nothing. There are other cases, however, where Russell's
view of "exists" (that it is only a property of a
propositional function) fails to account for our meaning;
as, for example, when we say "some tame tigers do not
exist" and mean "some tame tigers are imaginary."
109
Further, Moore argued, "this exists" says something
meaningful when we mean to refer to a sense datum of a
physical object, rather than to a sense datum of an
imaginary object. But further still, we might say "this
exists" significantly if it is significant to say, "this
might not have existed." Surely this last expression is
significant, but it would not be significant if it were
37
not also significant to say "this does in fact exist."
Moore concluded that there was no good reason why
"exists" should not be significant, and stand for a
genuinely predictable attribute of some things. In the
case of every sense datum that he had, Moore said, it was
logically possible that it might not have existed.
Logically, the contradictory of the significant is
significant; and, because logically possible non-existence
is contradicted by actual existence, the attribution of
existence is significant. In summary, here are Moore's
arguments that "existence" is a predicate: (a) Some
people such as Russell argue that to point to something
and say "this exists" asserts no proposition; it says
nothing. But this is only partly true. (b) It i£
significant (i.e., non-redundant) to point and say "this
exists" when it asserts "this is a sense-datum of a
physical object." (c) "This exists" is also significant
in the sense that it says "this might not have existed."
But Moore1s arguments seem vulnerable to the
110
following objections: (a) Russell's opinion that
"existence" was a property only of propositional functions
was integral with his conception of the correct logical
form for all empirical judgments. Thus, if it admits of
an empirical exception it is false and false utterly.
Moore generously called it "part of the truth," but it is
not clear what a logical half-truth could be. It seems
like a contradiction in terms. It is difficult to believe
that it could express truly what is sometimes meant by
saying that existence is not an attribute.
(b) Moore argued that "this exists" is signi
ficant in the sense of "this is a sense-datum of a
physical object." But the mere sense-datum of a physical
object does not differ from one of an imaginary or dreamt
object. There is nothing in "this exists" to indicate
which of these it is. The gesture of pointing-to will not
help; for one might imagine or dream, along with the
object, an act of pointing and saying "this exists" to an
audience of bystanders who would confirm it. There is
nothing in the mere pointing and saying to indicate
38
whether it is 'real' or imaginary pointing and saying.
(c) Is "this exists" significant in the sense
that it says "this might not have existed?" Surely "this"
is intended to indicate something actual, and so, it is
not something merely logically possible. And on the other
hand, logical possibility is removed from the actual
111
"this With mere possibility we are in the realm of
supposal and semi-ignorance. Moore has argued that
significant.logically possible non-existence shows that
existence is a predicate, but is this a significant
possibility or only a possible (not an actual) signi
ficance?
Is it a significant possibility that this X, which
does exist, might not have existed? Here we must inquire
about the meaning of "possible." In the first place,
significant possibility for most of us refers to the
future, not to the past. But Moore's "might not have
been" refers to the past. In the second place, it is
significant to speak of possible predicates only when
actual predicates are unknown. We use "possible" only
where our information is too scanty to permit a precise
determination. Thus, for example, it is possible that
there are men on Jupiter because we do not know it to be
otherwise. "Not known to be otherwise" is the meaning of
"possible" so we may, in that weak sense, predicate human
inhabitants of Jupiter. But if it actually is known that
there are men on Jupiter, then the predication of mere
possibility is no longer significant. To say "is so, but
possibly otherwise" is mere confusion. It is not
possibly otherwise for the best of reasons: it is known
39
to be so.
112
2. Some Difficulties with Moore's Conception of
Definition
Moore has presented (in rudimentary form) the
now-familiar objection that metaphysics grows out of
mistaken belief in unitariness of meaning and universal
definability of words. Moore argued that some meanings
are simply known, without the need for any criterion of
use. It cannot be that every word is definable, he
argued, for some words convey simple notions which are
neither analyzable nor translatable into other ideas.
40
"Reality" is such a word. Moore meant by "simple
notion" (or simple idea, simple property, simple in
definable) , "the character which we rightly attribute to
it when we say that it is . . . [good, real, a fact,
yellow, etc.] . . . ." He did not mean, by "simple
notion" (simple indefinable), "the complex of characters
41
which justify one's use of the word."
Moore distinguished between "rightly attributed
character" and "complex of characters justifying the
word" to save his conception of definition from an attack
by H. W. B. Joseph. The details need not concern us
here, but it is worth pointing out that if a distinction
can be drawn between what something is and what justifies
our use of a word, then Bradley's "real" could as easily
escape Moore's criticism as could Moore's "simple
indefinable" escape Joseph's attack. Bradley certainly
I did not intend to say that reality is his definition, but
; rather he meant to say that one's use of the word
"reality" is justified where conditions are assumed to be
irrelevant. If the words "where cpnditions are assumed
to be irrelevant" asserts a complex of characters which
justify one's use of a word, then reality is not that,
for Bradley's reality was not any complex of self- .
existing characters. By "character" Moore meant some
thing non-relative, an essential quality on its own. By
"complex" of characters Moore meant an external relation
among these individualities. But for Bradley nothing does
really possess such individuality, save Reality itself.
Finite things are what they are (what they appear to be)
in virtue of their internal relation in Reality. Some
thinkers have found this a difficult notion to grasp, and
it presents a problem which needs to be settled. What
concerns us here is that if a philosopher may distinguish
between reality and the means by which a sign refers to it,
in order to save his theory, then Bradley's theory of
Reality need not be troubled by an attack upon the
formulation of its verbal symbol.
Is it true, as Moore believed, that there must
be at rock-bottom, some simple indefinable qualities?
Some atoms of meaning? One might ask for the force of the
"must" here. One might ask for it because, to some
thinkers, it has seemed that the "must" results only from
114
a prejudice on behalf of analysis. For, if oneinsists
that the correct occupation for a philosopher is to anal
yze and if one presumes that there is some way of knowing
when he has done his job, then the end product must be a
set of simples. Then the simples (whether called "inde-
finables," "notions," "qualities," "properties," "sense-
data»" or whatever) have not exactly been discovered in,
but rather have been imputed to, the subject from a prior
theory about the nature of the philosopher's task. It is
likely that such a theory was held by Moore, for he did
not think that philosophy could discover new truth.
Philosophers could only 'describe' in their special way.
The truth of things was to be found in Common Sense, and
the philosopher's job was to analyze it. Where philo
sophy and Common Sense threatened to collide, Common
Sense had the right-of-way. The impending collision was
a sure sign that the philosopher was wrong. That is one
way to look at the philosophers' task. But, looking at
it that way, it appears that the conception of meaning as
composed ultimately of little mosaic blocks of simple
indefinables) has not been given as true, nor discovered
true, but rather has been taken as true. Whether or not
it is^ true depends on further considerations.
One of those considerations is that atomic
meaning must at least be possible. Moore said that
definitions are composed of "simple notions" which are
indefinable.^ A "simple notion" is not an idea or a
concept, apparently, but nevertheless it is divisible into
kinds. "Notion" is a vague word. It differs from a con
cept or an idea in being indefinable, but it is known
that there are kinds of notion of which "simple" is one
kind. Then the meaning of "notion" is not simple, what
ever may be simple about it, for it itself allows for a
diversity— an internal diversity— of kinds. But then,
by a similar argument, neither is the meaning of "simple”
itself simple; for, if it means what it seems to mean,
then its meaning is a correlate of "complex'." Part of
the meaning of "simple" is "not-complex." Then the
meaning of "complex" is involved in the meaning— the
intrinsic meaning— of "simpleThis result is puzzling,
for the meaning of "simple notion" is somehow a con
junction (and therefore compound) of two other meanings,
neither of which is simple. The problem of "simple
notion" seems formidable. Its meaning involves
"definable," for it is indefinable; it refers to other
non-simple notions through "kinds," of which it is only
one; and it refers to "complex" if the meaning of "simple"
is known. Moreover, whatever the conjunction may be
which welds "simple" and "notion" together into one
meaning, it is surely not to be left out, for "simple"
and "notion" are distinguishable meanings. But this is
not all, for, as Moore insisted, we must also add the
character which we rightly attribute to the simple
notion when we say that it is^ a simple notion. We have
only been talking about the complex of characters which
justify our use of the words "simple notion . " We must
also add to this complex of characters the yellowness,
for example, of the thing that is yellow.
There must be some way to get all three of these
things together: the words, the meanings, and the
yellowness of the thing that is yellow. For "yellow1 ;"
said Moore, is a simple notion. But it seems not to be.
If it is one, yet it is not one simply: it is one-in-
many, and the many-ness is not irrelevant to it as one.
Surely part of the meaning of "yellow" is "simple" plus
"notion," for otherwise how could Moore be sure (as he
was) that yellow is a simple notion? There is no
intention here to be facetious, or to make sport of
Moore's serious effort. This is an old problem, and
Moore, at least, has helped us to face it. But he could
not escape the problem by drawing a distinction between
what a thing intrinsically is and the complex of charac
ters justifying our use of the words, for then Bradley's
Reality would leap through the gap.
And yet, it may still seem that Moore has saved
his "simple notion" by his distinction. He argued that
yellow is a simple notion "in the character which we
rightly attribute to it when we say that it is [yellow],"
and not in "the complex of characters which justify
calling it [yellow]." Have we forgotten this distinction
in our criticism above? Would it save "simple notion"?
Consider carefully. Now "yellow i£ a simple notion,"
Moore said. May we ask what function the copula has
here? Surely analysis is justified. What of this "is"?
It cannot be the copula of equivalence, for, if on one
side of the copula is "the character which we rightly
attribute to it [in calling it yellow], and on the other
side is "the complex of characters which justify calling
it [yellow]," then the "is" is no "is" of equivalence.
Moore insisted upon it. Perhaps, then, the "is" affixes
a predicate to a subject. "Simple notion" is a predicate
of "yellow." But either it is a necessary predicate
or only a contingent one, and which is it? It cannot be
a contingent one, for then an exception could be admitted
and "yellow is not a simple notion" could be sometimes
true. "Is" must affix a necessary predicate. But in
that case "yellow" would not be simple for it would have
a necessary predicate. Obviously, the "is" is not the
"is" of predication. What else could it be? Perhaps we
might call it the "is" of denotation. "Yellow" denotes,
and so, it is a proper name. But then, we must ask, "Is
not denotation a species of definition?” If not, then
the consequences are serious for Bradley's opposition and
Moore has lost a valuable tool. But if so, on the other
hand, then what is to be said for the claim that
"yellow" is indefinable? But this is not the worst.
If "yellow" is a proper name, then it connotes nothing;
it means nothing— and "yellow" is, ex hypothesi, an atom
of meaning. If not, then how can the simple notion serve
as the foundation stone for all definition?
But are we still forgetting Moore's distinction,
and confusing the mere word "yellow" with its meaning?
Then forget, if possible, the mere word and focus upon
the meaning. With the meaning well in mind can it even
then be said that it connotes nothing more than itself,
to be itself? Does "yellow" not also connote "not red-
43
green-blue , " or something of that sort? It seems so.
And if it did not, then how could one know, by inference,
that those meanings were related to it?
There is another difficulty. A "unit of meaning"
seems like a contradiction in terms. A meaning is surely
a passageway for thought, not just a stopping-place for
thought, How, otherwise, is inference possible? Then
the notion of the atom, however useful, must be a
fiction. The meaning, even if thinkable as unitary, must
also refer away to another. It is that quality
(presumably a quality of something) which enables infer
ence to be accomplished. But yellow is a simple quality.
Then presumably (in fact emphatically) we must attend to
the actual quality, occurring (inhering) as it actually
119
does. We must not attend to the mere word, lest we fall
into a trap of verbal sophistries. But this admonition
leads us to an interesting conclusion: actual yellow is
something's predicate, its adjective. Presumably that
something is real. But in real life, even in Moore's
sense, is there ever mere yellow? Is it not always a
yellow something? It seems so. And if it is always a
yellow something, then the yellow is always conjoined and
compound. For is it nothing to the yellow that it is
thus dependent in its actual occurrence, arid has con
ditions imposed upon it? Yellow is always a predicate;
it is never a subject. How, then, can it be an atom?
In actual fact yellow is no disembodied wraith. But if it
is always a predicate and never a subject and a predicate
of something real, then it is precisely what Bradley said
it was: a property of reality. An atomic quality is
a contradiction, not only in terms but in fact.
It may be helpful to summarize the criticisms of
Moore's conception of definition in this section as
follows:
. (1) Moore believed that to define is to analyze
into atomically simple qualities, notions, or indefin-
ables. He meant by such simples, found through analysis,
the "character itself,',1 and not that "complex of
characters" by which a word is attached to it. But this
distinction, required by Moore to save his theory,
120
defeats his attack upon Bradley's conception of "reality ."
For Moore emphasized such things as the purported
circularity of Bradley's definition, and Bradley's
supposedly mistaken notion about the universal defin
ability and unity of words.
(2) Moore's belief that the correct activity for
philosophers (as guided by the dicta of Common Sense)
assumes but does not prove "simple indefinables/" "simple
qualities, properties, notions," et al. That assumption
must be rejected,^ for "simple notion"
(a) is compound and thus contradictory,
(b) would support or permit no inference,
(c) does not actually occur apart from, and is
meaningless if thought to be independent of, the "complex
of characters which justify use of the word."
(d) Either denotative definition does or does not
identify the "complex of verbal characters" with the
"rightly attributed character." But Moore both affirmed
and denied this identification.
3. A Defense of Bradley's Criterion for Reality:
Reference to Conditions
Moore criticized Bradley for having thought that
the sense in which he used "reality" was unitary when
actually (Moore argued) it was at least dual. No
criterion could be found. Whether or not Moore himself
believed that no meaning is unitary is hard to say.
“ ‘ 121
Certainly he believed in atomically simple meanings.
Certainly he believed that the meaning of "exists" and
"is real" are the same. But on the other hand he
believed that such meaning is 'incorrigible*. Thus, to
the question, "Is the meaning of words like 'real',
'good1, 'is a fact', and 'yellow' always one and the
same?" we should expect the reply, "There is no way to
tell." Moore applauded the investigations of Wittgen
stein, who certainly did not believe that meaning was
unitary. But a clear and unambiguous challenge to the
unitariness of meaning must await the discussion of
J. L. Austin, to follow.
What must be emphasized here is Moore's belief
that all meaning is ultimately indefinable. Common Sense
uses words like "real" and it uses them always correctly,
Moore thought, for there.is no other standard of correct
usage. Common Sense uses words at the 'atomic' level,
without any conscious criterion at all, and any criterion
revealed to philosophical hindsight will fail in some
respect. There will be some case or other, where Common
Sense judges something "real," to which the supposed
standard does not apply. Older philosophies erred by
dogmatizing on the basis of their definitions, but the
new trend in philosophy is to search out exceptions to
those definitions, enshrined in common speech patterns,
to show that the rules of definition were based upon
122
convention like the rules of a game. That activity of
searching out is an important part of what is called
"linguistic analysis."
Moore discussed the proposed criterion for
reality, "A thing is real if it is objectively independent
and unreal if it is mind-dependent." The criterion fails
to account for the mental act, Moore objected, so the
criterion fails. Now this criticism, if valid, would
be damaging to Bradley's position; for he had maintained
that any kind of relative dependence versus independence
was a criterion for the application of the word "reality."
That thing was the more real which was independent to the
greater degree. But note here a confusion that seems to
have crept into Moore's thought: the mental act is not
mind-dependent in the same sense as is the object of that
act. An act of imagining is a 'real' event in someone's
•
history in a sense obviously less conditioned and depend
ent than is whatever is imagined. If one picks up some
one's history at the point of his performing an act of
imagining, the act (even though conditioned by previous
events), if taken merely at that point in time, is a
'brute' fact and so, independent. The object of the
imagining, however (at that point in time), is
inextricable from its mind-dependence. It is in that
sense unreal. Reality among appearances is a matter of
degree— a degree which Moore seems to have overlooked in
123
calling both the act of imagining and the object of that
act "mind-dependent .1 1
Note something interesting about Moore's state
ment of his reason for rejecting "independent versus mind-
dependent" as the criterion for reality: Said Moore,
"If everything, which is only in the mind or dependent on
it is imaginary (unreal), my thought of a griffin is just
45
as imaginary as the griffin himself." Note the "^ust
as'." It implies that Moore recognized here a degree of
mind-dependence greater for the griffin than for the act
of imagining. Here, plainly, the difference between
reality and unreality (between the act and the griffin)
is one of degree, and is not a property of either one.
But Moore had called them both "mind-dependent" in no
qualified sense. How would this example be an exception
to the criterion if the griffin and the act itself con
formed to it in a differing degree? Not only does Moore's
conclusion fail to follow but his own example proves him
wrong.
A second criterion which Moore examined and
rejected was "A thing is real if it has connections or
effects which it would not have had were it not real."
This statement comes close to Bradley's meaning for
relative reality if we change "connections or effects"
for '.'conditions . " This criterion does refer to the
inclusion of the larger context in the case of the thing
judged more real.^® For Bradley, a thing "has or belongs
to reality" more than something else if in thinking of it
we must include within its meaning more of the whole
condition for existence than we must for something else.
Thus, for example, a human being is more real than a
table or chair, for the meaning of "human being" includes
not only the bodily appearance, but also the subjective
determinations which condition bodily appearance.
Moore rejected the criterion above because he
thought it circular. The definlendum recurred in the
definiens and the definiens could not be substituted for
it without generating an endless regress. We get,
"A thing is real if it has conditions which it would not
have had if it had not . . . had conditions which it
would not have had if it had not . . . had conditions
which . . . ad inf." This of course is not a definition
of Bradley's reality, for at best it is a statement of
the degree to which a thing may be said to belong to
reality. It defines relative reality. Moore did find a
place where Bradley said, "appearances have or belong to
reality," but this "have" is meant in the sense that a
tool "has" its function, a citizen "has" his duty, or a
slave "has" his master. Without this "having" in virtue
of which something is what it is, the tool qua tool, the
citizen qua citizen, or the slave qua slave, are not what
they seem.
125
The definition which Moore found circular is a
definition of relative reality, as of one thing against
another such as of a horse to a griffin, or a human being
to a table or chair. But it needs to be rephrased in
order to be accurate. It should read, "One existence is
more real than another (definiendum) if it includes within
its meaning conditions of occurrence not included in the
meaning of the other (definiens)." That eliminates the
circularity. It is not necessary to state which con
ditions the more real thing includes, so long as it
includes the conditions of the other plus at least one
more condition.
Summarizing this section, it is fair to say that
Bradley's criterion for "reality" is: "Something is more
real than something else if the former is less con
ditioned in its occurrence than the latter." Moore
criticized this definition for the reason that both a
mental act and an imagined object are conditioned by being
mind-dependent, yet one is real, the other, not. But (a)
to call the mental act of imagining and the object ima
gined "mind-dependent" is misleading, for it uses "mind-
dependent" ambiguously. Moreover, (b) Moore himself
admitted that the mental act and the object of that act
(the thing imagined) differ in degree of mind-dependence,
thus implying, in spite of himself, that there are degrees
of reality. Moore argued that reference to conditions
126
(or "connections or effects") as a criterion for reality
leads to circular argument and endless regress. But if
taken as a definition of relative reality, as Bradley
intended, reference to conditions is usable as a
criterion and may be stated so as to eliminate circu
larity.
4. Bradley's Reality is Acceptable to Common Sense
Moore argued that reality cannot have degrees,
because the judgment "real" is made without anything
present which is susceptible of degree. For example.
Common Snese would judge a seen polar bear to be real and
a dreamt or imagined one unreal. What is present to all
cases but possessed by the seen bear to greater degree?
The conditions of its occurrence? Then do we first
count the conditions and compare the tally before de
claring the seen bear to be the real one? No. In fact,
the conditions for actually seeing a polar bear are very
complicated and difficult to achieve, while one may
dream or imagine one with ease. Yet there is no doubt
that the seen bear is the real bear. The conditions of
occurrence are different in each case; there seems to be
no one thing to be more or less present; we do not (at
least consciously) concern ourselves with analyzing
conditions; the judgment "real" is immediate, un-
ref lective, and indubitable.
12 7
So why, if Bradley's were the criterion, is the
seen bear real and the dreamt one not even though it may
be very rare and difficult that we actually see one? And
there is still more to be said for Moore's view. As a
matter of fact, the seen bear, the real bear, exists far
more conditionally than does the imagined one. The real
polar bear needs food and a favorable climate; he is
constantly subject to the depredations of big-game
hunters and disease. The imaginary bear is free from all
that. But if the example of the polar bear isn't con
vincing, then consider the Abominable Snowman. He is
judged unreal although no one knows what conditions need
be met in order for there actually to be one. Conditions
are irrelevant.
In reply, the above may be met with a flat denial.
The judgment "real or unreal?" is the subject under
discussion. There is no such thing as a simple, immedi
ate, unreflective judgment, not even for Common Sense.
The judgment involves comparison— in this case between
seeing and imagining. The imagined object is more
conditional in its existence even to Common Sense than the
seen one, for it involves an effort on our own part.
To see it all we have to do is look. As for dreaming it,
judgment of "real or unreal?" about that, too, is
comparative and reflective and is not made at the moment
of dreaming. The judgment is not simple and immediate,
128
and the conditions (or connections, or effects, or
whatever) are quite relevant to the object’s existence.
So far as the Abominable Snowman "is" at all,
he is imaginary, and beyond that neither science nor
Common Sense can say. But Common Sense being now the
issue let us keep to that, keeping to the 'real', seen,
sensible, actual polar bear uncomplicated by irrelevant
problems about sense-data or irrelevant news about animal
husbandry or geography. Now if, to Common Sense, a polar
bear in fact is present (ignore him if you can), and
Common Sense judges "Here is a polar bear," then no
condition applies so far as concerns Common Sense.
Common Sense takes this to be a categorical judgment.
Here science and Common Sense part company. Science
stresses the conditions present but satisfied. Common
Sense disregards them altogether.
Science surely is right. The judgment "Here is a
polar bear" is not really categorical, but rather it is a
conditional judgment with the conditions taken to be
satisfied. Common Sense does, nbt state the conditions,
and even if it presumes them, it ignores them. It must.
It has more practical work at hand. A full statement
of the conditions would describe the whole universe,
emptied of time. But time, here, means a lot to Common
Sense. To science, 'time' is but a variable called "t"
in a formula to which all is simultaneous. Not so for
12 9
Common Sense.^ Reality for it is all strung out upon
the stage of time. So, as regards the real seen polar
bear, whatever the conditions are, they are taken by
Common Sense to be satisfied. This is not the case for
the imaginary bear. Without constant effort, he is not
present. Conditions for his existence cannot be ignored.
He is present only on condition that we keep on imagining
him— by an effort on our part. It cannot be taken for
granted. It is not satisfied by the object. Compara
tively, the actual bear is simply present, but the
imaginary bear's presence is compounded by a condition
outside himself. That condition makes a difference to
Common Sense— that is, to us. It makes the difference
between a real bear and an imaginary bear. Of the
imaginary bear, Common Sense says, with Bradley,
"conditional, therefore unreal."
So, Common Sense offers no good reason to con
clude that the judgment "real or unreal?" does not turn
upon a knowledge of conditions. Science affirms that
flesh and blood existence in space and time is highly
conditional, and thus confirms Bradley's opinion about
the ultimate reality of it. Yet scientific knowledge
must be set aside if we wish to return to the view of
Common Sense. But even Common Sense supports Bradley's
criterion. Let us look at the problem of "real or un
real?" through our Common Sense. To Common Sense there
130
are no conditions for the existence of the polar bear.
He satisfies them all. But the condition for imagining
or dreaming him we have to satisfy. That condition
makes a difference within Common Sense. There is thus no
conflict between Bradley's criterion and the actual
procedure of Common Sense.
D. Austin's Critique of "Reality"
We have now examined Moore's criticisms of
Bradley's "reality" and, moving forward in time by a
A Q
full generation, let us look at the linguistic method*
as it was sharpened and practiced by others. John
49
Langshaw Austin was one of the best exponents of it.
He agreed with Moore that philosophers can be wrong in
their use of words like "reality ,1 ^® and before we can do
philosophy we must examine ordinary meaning to find out
the correct usage. He differed from Moore, however, over
the notion that there was at the root of every word a
51
simple quality, as the meaning. Some words are
exceptions and "real" is one of them. It was a normal
word in the sense that it was at home in ordinary usage,
Austin said, but it was an abnormal word in having no
52
"single, specifiable, always-the-same meaning."
Austin launched a vigorous, witty, and unam
biguous attack upon the belief that "reality" has any
unitary meaning. He was more direct than Moore. He did
13'1
appeal directly to our good sense. We must maintain that
"reality1 1 always means "that which occurs unconditionally."!
i
That is its unitary meaning. We must argue, in support of j
Bradley, that this unitary meaning is always compatible
with the result of the judgment "Real or unreal?" wherever
it occurs in common speech. Mere compatibility with
ordinary usage will be taken as a sufficient defense of
Bradley, inasmuch as our purpose here is to show that
Austin's criticism did not refute Bradley's view. It is
probably not possible to do more than that. It is probably
not possible to prove, whenever the "real or unreal?"
distinction is made by anyone, that the criterion "exists
unconditionally if real" is consciously and deliberately
held in mind.
53
Austin's attack, or his method of demonstration,
was to present an assortment of unquestionably common
usages of the "real or unreal?" distinction, in an attempt
to show that there was no always-the-same meaning (to
"real") among them. "Show" is the important word when
speaking of Austin's method. We are to 'examine', 'look
at', or 'see' for ourselves what meaning the word "real"
has in this, that, or the other expression. These are
ordinary, common expressions which we all use and under
stand when (as linguistic philosophers like to say) "the
word is at work in its natural home." Thus, Austin's
'attack' is not an essay in formal proof, arguing that
such and such premises lead to contradiction. His
132
method of ‘demonstration1, as I have called it, is not
logical demonstration, but rather it consists of an
appeal to our own sincere discernment, our good sense.
He wants to 'show' that "reality" has no one, always-the-
5 4
same meaning; he wants us to 'see' his point.
Let us consider now, with Austin, his various
examples showing where "reality" and "real" occur in
common speech— ex hypothesi correctly used— to 'see'
whether or not there is something common to all those
usages. "Reality," if definable, must have an always-
55
the-same usage.
(1) "That isn't real cream." It isn't but, as
Austin pointed out, neither is it illusory cream nor
non-existent cream. It will still color the coffee. It
is fake, phony, artificial cream. This example poses no
problem for Bradley's standard definition of "real." We
are functioning within the parameters of Common Sense.
In that region, genuine, real, standard cream is dairy
cream, cow cream. That is what Common Sense takes for
the unconditioned object. So, fake cream is "object
plus condition .." That condition, which makes a dif
ference to Common Sense, is that someone made the cream
in a machine. Unlike cow cream, which can be obtained by
primitive means, it takes a high degree of engineering
skill to get artificial cream.
(2) "That isn't the real color of her hair."
M3
Again this one is easy prey for Bradley's definition.
"Naturally occurring" is the standard or unconditional
'real1, for Common Sense. Her natural hair has a certain
color. She has dyed it, say, with henna rinse. A
condition enters the picture; a conditional color is
applied. Again the situation if one of "object plus
condition." "Conditioned and therefore unreal," says
Common Sense.
(3) A deep-sea fish lies on the deck of the
ocean trawler. Its color is pale grey. "That isn't the
real color of this fish, for in the ocean depths it glows
with the iridescent and multi-hued color of the rain
bow!" says the deep-sea diver. He has seen the fish in
its natural habitat. That is (again) its standard, un
conditioned state. But a condition has been forced upon
it now that it is netted and hauled up, lying dead on the
deck. Again, conditional and therefore unreal.
(4) "What is the real taste of saccharine?"
Saccharine dissolved in tea makes it sweet, but "taken
neat” (placed directly in the mouth) it tastes bitter.
How shall we proceed with this example, for it is a
Common Sense example but Common Sense seems not to know
how to proceed? First it must choose a standard taste,
or else the question is senseless. Usually such a
standard is implicit, or 'understood'. But here, in
philosophical discussion, we must be explicit. We must
134
ask, "What is to be called the standard taste? " One may
opt for "taken neat." In that case the conditioned
taste, the unreal taste, will be the taste that
saccharine has in tea. Again, conditional and therefore
unreal, but in this case even Common Sense has to choose
a standard. Otherwise the question "Real or unreal?"
does not arise.
(5) "What are the real colors of the sky, or the
sun, or the moon, or of a chameleon?" Again the same
problem recurs as in the case of the saccharine. One
must first choose a standard to which the conditioned
(by which known to be conditioned) will be unreal. If,
as Austin suggested, we cannot decide for, say,
"conditions of standard illumination" on the sun, why,
then, the question is pointless. And why shouldn't it
be? In ordinary speaking people don't always ask of
everything, "Is this real, or unreal?" And, in ordinary
conversation, sometimes people do rebuke one another for
asking foolish questions. It is no refutation of a
definition that sometimes one is confused about whether
or not to apply it. Rather, what refutes the definition
is the case where there is no doubt about its applica
bility, and yet it fails to fit. But in this case there
is no insuperable problem about fitting it to the fact.
Suppose that the chameleon, normally encountered, is at
rest in a tree and looks green; or suppose that the sun
HB5
on a clear day high overhead is yellowish white. Common
Sense does take these for standard conditions. Then any
deviation, such as the sun through smoke, or the chameleon
frightened and plucked from his tree, will present a
conditioned and therefore unreal color.
(6) "What is the real color of a pointilliste
painting?" Presumably its standard color will be what it
displays from a normal viewing distance, say, halfway
across the room. There the little colored patches blend
to give the appearance of a unified whole. Very obviously
the success of this sort of painting all depends on
viewing conditions, and we must choose a standard in
order to make the ordinary-language judgment about the
'real* color. The selection of a standard is in all
cases prior to the applicability of ordinary language—
which was not devised, moreover, to describe pointilliste
paintings. If the selection of a standard were denied
then it would be correct to say, "I just don't know how
to answer."
(7) "What is the real color or an after-image?"
Plainly, this example was chosen to illustrate a diffi-
56
culty for the Sense-Datum School. Austin's point was
that an after-image possesses no 'incorrigible' color.
It seems to be many colors but none of them fixed. The
same general comment may be made about the 'real' shape
of a cat. That fey creature has a repertoire of shapes
136
ranging from proud to ridiculous. As he pointed out, in
such cases we do not know how to proceed. But is this not
57
because such cases (if indeed there are any such cases)
are so obviously conditioned, and those conditions so
obviously must be specified that, without it, the "Real
or unreal?" judgment of color or shape is of no interest?
There is no reason here to think that Bradley's position
has been weakened. To refute him one must produce a case
in clear violation of his criterion.
(8) "Is this a real diamond?" What else could it
be? If there is no alternative then surely it is a
diamond. If there is an alternative then presumably that
alternative is known. We have our standard; we can pro
ceed. Might it be a fake diamond? This sort of case is
now familiar. Might it be, instead, a zircon or a piece
of quartz or glass? Then it is no diamond at all, and we
might with reason declare the question wrongly asked.
Again, we must first identify a standard object in order
to entertain the question "Real, or not?" Presuming that
the diamond in question is a diamond, then the question
is about the conditions under which it was produced— for
there are now chemically genuine but artificial diamonds.
The same sort of situation applies to Austin's question,
5 8
"Is this a real duck?" What are the alternatives?
Presumably they are that it might be a toy or decoy duck.
Presumably we have already excluded the possibility of it
137
being a grebe# a goose, or an aquatic Plymouth Rock. The 1
possibility of it being a hallucinatory, illusory, or
imaginary duck has already been touched upon in the
preceding section on Moore.
Let us add to Austin's list of examples by
supplying one of our own, asking,
(9) "Is that a real Santa Claus?" We grownups
know that there isn't any real Santa Claus. There isn't
really any jolly old elf in a red suit who lives with his
wife at the North Pole, supervising the other elves
making toys, to distribute, at Christmas, to good girls
and boys. But how do we know that if, even from the
Common Sense standpoint, there isn't any real, naturally
existing, unconditioned Santa Claus? Haven't we, here,
what Austin believed he had found, namely a usage of
"real" that is sui generis?
Oh, the answer is very simple, we might reply,
because the term "Santa Claus" means just this uncon
ditioned, naturally existing jolly old elf. We have
there our standard of "real ." But the department store
Santa Claus, the unreal one, occurs very conditionally.
He comes to work, like any employee, and puts on his fake
beard and tummy.
But perhaps we're doing too much 'cooking' on this
example, to make it supple enough to bend into the para
meters of Bradley's definition. Have we, indeed, got the
1 3 8
meaning of "Santa Claus" in "jolly old elf from the North
Pole who doesn't exist?" What about "Santa Claus" in the
story the man wrote, for his little girl, entitled, "Yes,
Virginia, There Really Is A Santa Claus?" The meaning
with which we're truly dealing, here, is that of myth,
legend, or allegory. In this tale, the object isn't any
jolly old elf at all— in the literal sense. We're not
dealing, here, with a literal meaning. That makes this a
different, and difficult, sort of example. For myself,
I don't think that there ever is any strictly literal
meaning, but this raises a problem I'd hoped to avoid. It
involves the problem of relating the ideal to the actual
and arguing, as I think we can, that without ideality no
meaning at all can be found in the actual world.
But even that would not be enough: "ideal" does
not mean "real ” ," and in this example we have more than a
mere ideal. This is a very dynamic 'ideal'. The meaning
of the Christmas Spirit, and its relation to our human
drama, and whether or not it is real, threatens to lead
us in over our heads. The question of the reality of that
spirit, by which temporarily we put off selfishness and
greed, is not a question like the others. If we glimpse,
for a moment, all humankind as suffering children who
need, for their realization, some touch of Love, some
contact with Original Feeling, then can we call all this
unreal? If we feel impelled, by noble impulse, to give it
139
flesh to dwell among us, is it unreal? If, at midwinter
solstice, when our great dying Light is reborn to new
glory, we can, through it, refreshed, know the enshrined
significance of our ancient tradition, shall we deny it
its right? It conditions us. It makes u£ real.
The discussion above exhausts Austin's list of
examples. We have even added one of our own. By running
through them the reader will by now have begun to get the
hang of applying Bradley's criterion for reality. The
question we must first ask is, "What is being taken for
the normal or standard object? That is the "real-) "
Plainly, as Austin insisted, the "unreal" need not be non
existent nor some kind of trick of the mind. It may be
fake cream, a decoy duck, or synthetic hair color. It
was not Bradley but Moore who claimed that the unreal was
non-existent. Austin is wrong, however, to have insisted
that there is nothing that all cases of "real or unreal?"
have in common. The distinction is always based upon a
condition known that makes a difference from the standard
taken by Common Sense (if the context is ordinary lan
guage) to be self-existing.
E. Did Austin 'Really* Contradict Bradley?
The word "real, " or "really;"' can be used as
carelessly as any other, which is one difficulty with the
method of trying to prove anything by examining ordinary
140
usages. But when those ordinary usages are seriously
intended and thoughtfully chosen because no other usage
could do the job so well, then an examination of them
"shows' that they accept Bradley's criterion without
distortion. The criterion is, "The real is the self-
existent." If, to us, no conditions need be met before
something can be, then that something is real. "No
conditions" means, ordinarily, "no conditions which matter
to us♦1 1 That is why the standard meaning of "real" in
ordinary usage is "naturally existingV and that is why
59
Moore ran together the meanings of "real" and "exist."
Then it should not surprise us that Austin's
investigations reinforce Bradley's view. Notice what
Austin went on to say: "'Real' is a word that we may
call substantive hungry." And the reason for that is,
We must have an answer to the question, "A real
what?" if the question "Real or not?" is to have
a definite sense, to get any foothold.
And lower down,
"Real or not?" does not always come up, can't
always be raised. We do raise the question only
when, to speak rather roughly, suspicion assails
us.6®
This is Austin's way of saying what was alluded to above,
that unless there is present in mind some alternative,
then there is no question and no judgment of "Real or un
real?". Some standard must be known, and some alternative
possible, to raise a question about the reality of any
given content. If Common Sense knows, or cares, what it
takes to be the unconditioned existence, then it is
possible to make a distinction where conditions enter.
Recall the example of the 'real' color of the sun. What
is substantive here? When is the real sun present, color
and all? Without an answer to that, present or presumed,
the question "Real, or not?" is without point. Without
61
that the question isn't even asked. The substance is
what is aimed at with the word "real." Austin called it
"substantive hungry." It aims at what Bradley called
"the unconditioned individual".
Austin did not intend to discuss "real" in a
metaphysical sense, but rather he concerned himself solely
with the use of "real" as a modifier in common speech.
He said,
The attempt to find a characteristic common to all
things that are or could be called "real" is doomed
to failure; the function of "real" is not to
contribute positively to the characterization of
anything, but to exclude possible ways of being
not real.62
Bradley would have challenged both points made here. He
would have denied both (a) that no characteristic is
common to all things called "real;1' ; and (b) that a
contribution can be made to understanding by mere
exclusion (that any judgment or exclusion is possible if
not based on knowledge of some positive property that
excludes the property denied). The latter point need not
142
be argued here if the former point may stand. Has Austin
proved that no characteristic is common to all judgments
of "real”? Not yet. It has already been emphasized, and
bears repeating, that none of the objects called "real"
by Common Sense are real. They are conditioned, both in
the occasion of their appearing and in the ideal meaning
of their terms. But if, on the other hand, Common Sense
takes them for unconditioned existences, then such ex
istences are the standard or normal case. They are the
answer to Austin's always-prior question, "A real what?"
When the word "real" is used in ordinary discourse it is
used to refer to the characteristic of being unconditioned
even if further analysis, with which Common Sense cannot
concern itself, discloses that the imputed unconditioned
existence is not the case. This characteristic of being
unconditioned is the "simple, non-natural property" which
Moore was looking for— but it is not a property of any
existent 'thing', unless by a presumption of Common
Sense. Austin has not produced any examples incompatible
with this defining characteristic, of "reality' •"
The quotation cited above was made by Austin to
support his claim that "real" is a "trouser word*" He
said, "It is the negative use that wears the trousers."
"Real" has sense only if it be known how it might not
have been real. Real cream is distinguishable on account
of fake cream, a real duck from a decoy duck, and so on.
143
We must, on the contrary, insist that mere negation never
stands aloof from prior affirmation of the conditions of
denial. Austin might even be charged with contradicting
himself here for he has just told us, "We must have an
answer to the question, 'A real what?1 if the question
'Real or not?'1 is to have a definite sense, to get any
foothold." But even so, there is a point to what he
said, for as we said before, the alternatives must be
known if, in ordinary speaking, we bother to ask the
question "Real, or not?". If, in what is real to Common
Sense it is the state of being taken as unconditioned
that makes it so, and if anything at all may be taken so,
then without some contrast imposed on the object by some
thing similar but differing from the object in a way that
makes a difference, why, the judgment of "real" is point
less. It is in this sense that Austin said, "It is the
negative use that wears the trousers;" One may grant him
the point. But is it not perfectly compatible with
Bradley's criterion? It is the conditional use that
wears the trousers. The conditions must be known if there
is any point to calling something "unreal," for, after
all, whatever we experience i£ experienced. In that
sense, everything is "real’.,"
The conditional use is known to Common Sense; the
condition matters to Common Sense. Upon that basis it
makes its judgment of real or unreal. The known con-
144
dition modifies the meaning in such a way as to bring to
mind a distinction that makes a difference in the use of
the word. It is curious that Austin maintained that there
was no criterion for "real" when he went on to say what
that criterion was. It was knowledge, relevant to the
given situation, in answer to the always-prior question,
"A real what?"— which brought with it the awareness of "a
6 3
specific way in which it might be . . . not real."
The meaningful use of "real" depends upon knowledge of
conditions required by an alternative.
This point was made even more emphatically as
Austin went on to speak of "real" as a "dimension word."
"Real" is
The most general and comprehensive term in a
whole group of terms of the same kind [notice
"same kind"], terms that fulfill the same function
[again, "same function"], . . . [such as] . . .
"proper;" "genuine;" "live;" "true," "authentic;"
"natural"; and on the negative side, "artificial,'!
”fake,V "false;!' "bogus,!' "makeshift;" "dummy;"
"synthetic;" "toy",— and such nouns as "dream",
"illusion," "mirage;! ' "hallucination" belong here
as well.64
All of these are compatible with Bradley's criterion and
may be understood as referring to the conditioned versus
the unconditioned occurrence of something. Our point may
be made more clear if explained as follows: In any line
of endeavor, or in any "form of life" (a preferred phrase
among linguistic philosophers), there is some standard or
norm— a systematic set of ideals— by which that life-form
145
is known and defined. This standard or norm has an
internal structure which is referred to in linguistic
philosophy as the "grammar" or "logic" of the "game"
(this terminology is used to discourage flights of meta
physical speculation). Immersion in that structure
(acceptance and use of it) as a frame of reference (as a
"universe of discourse") furnishes one with the uncon
ditioned . The unconditioned existences are those
presumed to be self-existing and individual within the
frame of reference and upon which its "grammar" depends.
65
They are what is referred to by "real, "natural,"
"true,," "authentic , " "genuine1 , " "proper , " and so on.
That is to say, elements which have their "natural home"
in such a structure are, from that point of view, taken
to be of the highest "dimension" and so, unconditioned
and "real." Where the element exists within the
structure only as qualified by a condition that matters
within the life-form, then that element is "unreal" in one
of several senses; such as "artificial;" "bogus;," "fake,"
"toy," decoy," "dummy," "imitation," and so on. A down-
to-earth example may illustrate what we mean. Perhaps
a member of the house-painting profession says, "That
isn't real paint I" He certainly doesn't mean to say that
the liquid in the can is illusory or non-existent. He
means that it doesn't measure up to the standard Which
defines the profession, or "form of life;" It's cheap
146
dimestore stuff; it's chalkwater and glue. It isn't
full bodied, with high-hiding, long-lasting, non-fading
pigmentation, compounded in an enduring resinous vehicle.
"Real," Austin went on to say, "belongs to a
large and important family of words that we may call
66
adjuster-words." Language needs such words to give it
flexibility, to meet the unforeseen demands upon it.
Something new may come along which resembles the familiar,
but not exactly. We say, "It's like an X, but not a real
X." "Like" and "real" are here adjuster-words. So Austin
said, and it's probably true, for he had an impressive
grasp of his subject. Let him be granted the point.
But is it not, again, quite compatible with the criterion
we have been defending? Ordinary language reveals, in its
very structure, the need always to be grounded in the self-
existent, even though further analysis indicates it to be
only presumed self-existent. Ordinary language cannot
operate without the presumption of an unconditioned
standard. Ordinary language has certain words which
represent that standard. "Real" is, as Austin said, "the
most general and comprehensive term in a whole group of
words” which do this.
F. Summary of Austin's Criticism
Austin's criticism does not refute, but rather in
the main supports, Bradley's criterion for the use of
"reality”.
(a) A careful consideration of Austin's examples
purporting to show that "reality" has no unitary
linguistic function (such as could serve as a criterion
of use) proves the contrary. All of his examples of
"real" are compatible with Bradley's criterion, "un
conditional if real."
(b) Austin went on to point out the nature of
"reality". It is "substantive hungry", or demands a
standard or criterion instance. It is a "trouser word";
the negative use "wears the 'trousers', i.e. alternative
usages must be known if "real, or not?" is significantly
asked. It is a "dimension word," relating any judgment
of "real" to an unconditioned standard. These results
are just what would be expected if ordinary usages were
compatible with Bradley's conception of "real."
148
FOOTNOTES— CHAPTER II
F. H. Bradley, Essays on Truth and Reality
(Oxford: Clarendon Pr., 1914), p. H I . Hereafter cited
as ETR, and taken as his definitive statement of position
on "realityy"
2Ibid., pp. 289-90.
3Ibid., p. 190.
4
Ibid.
5In judgment, the mere affirmation or denial of
an idea or content to another is not the whole intent.
The attachment is meant always in a larger whole which
supplies the conditions. The judgment is thus always
partial and relative to a larger whole which is presumed
to be present. The conditions which that presentness
supplies are never wholly specified, so the judgment,
as explicit, is never wholly true.
®To judge, one couples one idea to another.
But one does more than that, for one means to call the
coupling true. This it cannot be, if the ideas coupled
are merely "floating" ideas. In judging, one treats the
ideas as initially indifferent, yet if this assumption
is taken seriously then one must wrestle with the
question, "Then in virtue of what do you maintain that
they may be coupled truly?"
There must be some peripheral or marginal
adhesions to every idea (call them what you will) which
enhble it to enter judgment so that the relation thus
formed may be called true. The idea does have these
adhesions, although in judgment we may provisionally
assume the idea to "float"" See ETR, pp. 32-33. Those
adhesions are so less important in the negative judgment.
See ETR, pp. 41-42.
^See Appendix A, "What is an Idea?"
8Bradley, ETR, p. 251.
^Ibid., p. 254.
10See ETR, p. 333-34.
^Ibid., p. 226.
1 2
It is no more fair to call Moore a "linguistic"
philosopher than to call Bradley an "idealist’ ;." Each mem
denied the label. Indeed, the mature Moore was a
platonist, according to Klemke (E. D. Klemke, ed.,
Studies in the Philosophy of G. E. Moore [Chicago:
Quadrangle Books, 1969]).
On the other hand, Blanshard says, "[Wittgenstein]
. . . has been regarded as the founder of the newer
linguistic analysis. But the honour, if honour it is,
should go to G. E. Moore. . . . He suggested that common
sense and its language supplied to philosophy both its
main problems and a touchstone by which its speculative
claims might be checked.[Brand Blanshard, Reason and
Analysis (La Salle, 111.: Open Court Pub. Co., 1962),
p. 31b]." Norman Malcolm says that Moore's main
accomplishment was that of being the first to see the
philosophical primacy of ordinary language (Norman
Malcolm, Knowledge and Certainty [Englewood Cliffs, N. J.,
Prentice-Hall Inc., 1963t, pp. 163-183). Bergmann credits
Moore with having shown the necessity for taking what he
calls "the linguistic turn" (Gustav Bergmann, The
Metaphysics of Logical Positivism [New York: Longmans,
Green and Co., 1954]. p. 35). These scholars would agree
that to call Moore a "linguistic philosopher" is at least
defensible.
"Linguistic philosopher," as I use the term,
applies to those who criticize others not because their
premises are wrong, nor their reasoning faulty, but
because they have used words whose ordinary meaning
differs from the ordinary. I think that Moore did use
this method of criticism against Bradley.
I do not call "linguistic analysis" worthless.
We should "speak with the vulgar" if we can. Bradley
himself said as much. But it does not follow that
ordinary language is the philosophers' touchstone, nor
that they have no right to a "language game" of their own.
■^G. E. Moore, Some Main Problems of Philosophy
(New York: The Macmillan Co., 1953; Collier Books, 1962),
p. 221. Hereafter cited as Main Problems.
14Moore acknowledged the distinction between (a)
the flat denial of something's existence in time, and
(b) its qualification as mere appearance in time and
only ultimately unreal. Only (a) contradicts Bradley,
while (b) was the view he held. Moore claimed that
(b) was contradictory. See Ibid., pp. 220-21.
150
15Ibid., p. 222.
16Ibid., p. 226.
•^G. e . Moore, Principia Ethica (Cambridge, Mass.:
Cambridge U. Pr^, 1903), pp. 1-17.
3-8P. H. Bradley, Appearance and Reality, 2nd ed.
(Oxford: Clarendon Pr., 1930), p. 36. Hereafter cited
as AR.
^■9Ibid., p. 30.
20
I am not sure that Moore did understand that
Bradley was making this distinction. I am trying, above,
to give Moore the benefit of doubt, and make his rebuttal
as strong as possible. See Main Problems, pp. 245-51.
21Ibid., p. 241.
22Ibid., p. 244.
22Ibid., "The Meaning of Real/"p. 247.
24Ibid., p. 252.
25Ibid., p. 255.
26
Max Planck, Scientific Autobiography,
tr. F„ Gaynor (New Yorlci Philosophical Library, 1949),
p. 102.
27
I find no instance where Bradley did this
except in his Logic, published in 1883, where Bradley
was consciously trying to avoid metaphysical problems.
Bradley himself called attention to whatever ambiguities
beset his use of "real" and "is a fact" (in footnotes
to later editions) to a degree that was almost unneces
sary. See for example footnote #4, p. 108 (F. H. Bradley,
The Principles of Logic [London: Oxford. U. Pr., 1928],
vol. I. Hereafter referred to as Logic).
28See AR, Ch. IV and XVIII, esp. pp. 33-36.
2^As quoted by Moore, Op. cit., p. 230.
30
In an early article (Philosophical Studie s,
"The Conception of Reality" [London: Routledge & Kegan
Paul, 1922; rpt. Totowa, N. J., Littlefield, Adams &
151
Co., 1968], pp. 197-219), Moore (at that time) seemed
aware that Bradley's notion of "real" things was that
they were properties of reality, but Moore there rejected
that idea as simply nonsense. He said, "It is quite
bbvious that what we mean to assert [when we say 'lions
are real'] is not any such nonsense as this [p. 213]."
He made his point by appeal to ordinary language, asking
repeatedly, "What ought Bradley to mean [pp. 200, 209,
211]." Ordinary language was to be the supreme arbiter
of meaning, to which philosophers are bound by an
"ought . "
The notion that Bradley regarded "real" things as
properties of reality— and not that "reality" was the
property of some things— seems to have disappeared from
Moore's consideration by the time of writing the article
now under discussion ("The Meaning of Real," Main Problems,
Op. cit.).
31
"... the property which we express by the
word 'real'." Moore, Main Problems, "The Meaning of
Real," p. 252, line 4.
32Reality, remember, is the ultimate subject to
which, in Bradley's words,,all "appearance" is "adject
ival", in the sense of "a property of’."
33G. E. Moore, Philosophical Papers, Op. cit.,
pp. 114-25.
■^Before Moore could argue that existence is
predicable of some things as a property he must show
that "real" and "exist" may be used interchangeably.
He has not shown that, but he did argue for it in his
"Defence of Common Sense" (Philosophical Papers, Op.
cit.), pp. 32-59. Later that esaay will be examined,
and we will maintain that Moore failed to show the
equivalence of "real" and "exist.But for the sake of
discussion here, assume that he did.
33As quoted by Moore, ibid., p. 122. This seems
strange, It does seem that, even for Russell, "exists"
is a property, but not of the existing thing so named;
for "exists" is a property of a "propositional
function"."
36Ibid., p. 123.
37Ibid., pp. 123-25.
3®See Moore, Main Problems, pp. 248-49. The
152
Waverly consideration would prove the opposite of
Moore's position. That which distinguishes real from
imaginary, dreamt, etc., i.e. the act of attending,
stands outside of the content, as a context, and is not
got into the act by saying "this''. "
3^Bradley, Logic 2:707.
4®This conclusion may be drawn from Moore's
"Defence of Common Sense;.," Philosophical Papers, Op. cit.
jcn "The Meaning of Real';." Main Problems, Op. cit., he
spoke of the "property" called up before the mind by
"real)" "exist ,," etc. Our inability to point out any
such property did not make him speak of it in any terms
other than singular. Presumably he meant simple property.
41
"Is Goodness A Quality?" Philosophical Papers,
Op. cit., pp. 99-100.
42 , .
Pnncipia Ethica, Op. cit., Loc. cit.
43
Bradley emphatically denied that there could be
any such thing as "denotation" exclusive of "con
notation. " He was aware of the doctrine of Proper Names.
"Denotation vs. connotation" was for Bradley an ambiguous
jargon concocted to cover a false theory of definition—
a doctrine claiming meaningful separability of logical
extension and intension. See Logic 1:168-69.
44More precisely, we must reject the notion that
philosophers can do no more with meaning than to find,
by analysis, the simple notions (enshrined in common
speech) of which definition is the conjunction. That
there are simple qualities in some sense Bradley would
admit, for it is useful to presume them. That presumption
may be indefensible, yet no serious presumption is so
nonsensical that it is not founded on some truth. But
Bradley would insist that the contradictions in such a
presumption relegate it to the realm of appearance only,
not reality.
45
Main Problems, "The Meaning of Real," p. 247.
4®For Bradley, the determination of truth, as
reality, turns upon the question, "How much of my world
as I know it would be involved?"
47The contrast between the views of science and
Common Sense has been drawn too sharply here. For
expository purposes this was unavoidable, but it needs to
be reemphasized that even for Common Sense judgment has
153
its beginning there. There is, even at the Common Sense
level, reflective activity and timeless cogitative aware
ness. Common Sense vacillates between the Given and the
Taken, and may not be definitely located at either
extreme. Even the sciences, on the other side, are not
free from such indecision.
48
I must repeat that I am only following the
conclusions of other scholars (Blanshard, Bergmann, and
Malcolm) by referring to Moore as the originator of the
linguistic method. I have already documented this, and
I have already explained what I mean by "linguistic
method/" "linguistic analysis)'" and "linguistic
philosophy."
It may help us to understand the "linguistic"
method better if, instead of trying to say what it is, we
try instead, or in addition, to understand the motivation
which led to its origin.
Something called "idealism" dominated the skies
around the turn of the century, and Moore was educated
in that philosophical tradition. But he disliked it.
Blanshard tells us of that period when idealism dominated
the philosophical skies, and refers to them as skies
which were marred only by, in his words, "the cloud no
bigger than a man's hand which hovered over the spot
where Moore sat brooding." (See Brand Blanshard,
"The Philosophy of Analysis," Proceedings of the British
Academy, vol. XXXVII [London: Geoffrey Camberlege Amen
House, E. C. 48, 1952], pp. 39-69.) Blanshard refers
to Moore, then, as the "enfant terrible" of philosophy.
See Brand Blanshard, Reason and Analysis (La Salle,
111.: Open Court Publ Co., 1964), p. 314.
At that time Moore did not have his arguments
against idealism worked out. he knew he didn't like it,
but he couldn't say precisely why. He only had a feeling
against it. We may sympathize with.that feeling or we
may not, but I think we ought at least to notice that it
was feeling in Moore that later 'was resolved into terms-
in-relation', or 'worked its way out', or 'crystallized1,
giving rise to the method of argumentation used by him and
others later. So the little cloud spread and covered the
skies, for many others found themselves sympathizing with
Moore's feeling.
Feeling is beyond the readh of argument. I have
tried to deal only with Moore's arguments.
154
Moore's own philosophical development shows an
interesting progression from the enfant terrible of the
early days of anti-idealism to an almost platonic view
in his last years. Along the way he retracted his own
'refutation* of idealism, contradicted himself about the
'reality* of sense-data, and repudiated the results of
the linguistic method which he had originated.
As with every matter of supreme importance,
feelings run high over this one. That is what makes us
move. But I say again, this is my opinion. Whoever
finds it an illuminating way to look at that watershed
figure, G. E. Moore, is welcome to share it with me.
I think I am correct; I think it can be documented. But
that would be a large undertaking and I do not offer
to do so here.
49 .
The reader may have noticed that no mention is
being made here of Ludwig Wittgenstein. This is not from
prejudice, or ignorance, but is deliberate for several
reasons: (a) It is difficult to know just where to
place him. He seems to have resented any attempt to
summarize his 'doctrine' or even to impute one to him.
Those who claim to have understood him suggest that what
he advocated was, rather, a method. Just how there is
to be a method which yields no cumulative results worthy
of being called a doctrine is not easy to understand.
(b) Experts who do take from his writings a doctrine
seem to disagree among themselves. If Wittgenstein left
no mutually agreed-upon doctrine, even as a "description"
of his method, yet he must have exerted some kind of
influence on someone somehow. What kind of influence was
it? If it cannot be put into words then perhaps it is
permissible to pass by in silence. (c) There are passages
in Wittgenstein's writings which could be selectively
quoted to uphold most if not all of Bradley's teachings.
However and on the other hand there are passages which
seem to repudiate everything which Bradley ever stood
for. The difficulty (for an outsider) in arriving at
Wittgenstein's 'true view' lies with the fact that his
writings are a collection of aphorisms with little
discernible system among them or connection between
them. His 'method' may have produced in some men
certain insights but those insights seem all to have
been Bradley's lifelong possessions. The 'method',
however, was also productive of confusion inasmuch as
Wittgenstein made no attempt to be systematic— and system
was to Bradley all-important for intelligibility. On
the other hand Wittgenstein was not consistent even in
this (anti-system) for he seemed at times to believe
155
| that meaning and truth are understandable only within
: a system (the little sub-system comprised by the "rules"
; of the "game" employed) outside of which they were
nothing. See Appendix E.
^°J. L. Austin, Sense and Sensibilia, ed.
G. J. Warnock (New York: Oxford U. Pr., 1962? rpt.
Galaxy Books, 1964), p. 63,
51 .
Ibid., p. 64. Austin believed that Moore had
held to a unitary conception of meaning, a belief
difficult to reconcile with Moore's own discussion of
"reality;" But from what Austin says, then we were
correct in this ess^y to say that Moore believed "reality"
to designate a simple quality similarly to "yellow"
and "good .” (I have not discussed Moore's further
distinction between "natural" and "non-natural" qualities.)
52ibid. Austin was very critical of "non-natural
quality1 / " a term referring to such "simple qualities"
as are possessed by "good*)," for Moore. I have questioned
whether my assumption that Moore did hold such a belief
about "real" was correct, but according to Austin it
was. Moore did not explicitly Say so in his article on
"Reality.” He seems to have dropped the issue after
arguing that Bradley had been indefinite. Whether or
not "real" and "reality" designate a simple non-natural
quality must remain a question, if we keep to the
"Reality" article.
C O
Perhaps "attack" is the wrong word here. I use
it only in that sense which has become philosophically
common usage, in the 'adversary system' of justice which
has sprung up among us. I do not use "attack" in the
sense of "with malice aforethought"." Austin makes us
laugh. His witty discussion reveals that good humor
characteristic of first-rate thinkers, which character
istic Moore also shared. While I do suggest that Moore's
'attack' on idealism derived from a feeling he had about
it, I would refrain from suggesting that he was an
ignorant sorehead, or a 'bad man' in any sense.
^This is a difficult demand to carry out, for
"seei," here, has a special, philosophical sense. Even
as practitioners of the linguistic method, we had better
not contradict ourselves lest logic return to take its
vengeance. "See/" here, has no ordinary sense at all,
for we are to 'look at' meanings. But if meaning is
nothing always-the-same, but rather is, like quick
silver, always changing and running away, then how can
156
i we examine it? If we are to 'look at' it, must 'it* not
| be first recognizable and always-the-same? If not, then
j what are we to 'look at'. Looking at meanings is not
like looking at marbles or inspecting fuel-pumps ®n an
assembly line.
CC , 9
See ibid., pp. 64-67 for these examples.
^6I.e., A. J. Ayer's The Foundations of Empirical
Knowledge.
~*^I have never heard anyone ask, "What is the
real shape of a cat?" And, of course, all the talk about
after-images is highly contrived. That may be Austin's
point, but we ought to be sure that our illustrations of
common usage are common usage. Otherwise cadit
guaestio (putting it in uncommon usage).
CQ
I am stressing the similarity between the two
cases, above. I ddmit, of course, that they are also
different. People do make fake diamonds and palm them
off for 'real'. People do not do this with ducks. Thus,
one might object, the question "Is it real, or not?"
could have quite different meanings in the two cases.
But I must reply (a) Surely our language is not
so impoverished that we cannot be more precise in asking
our questions, and (b) "Real, or not?" does inquire
into the conditions of occurrence in both cases.
59I do not mean to imply that something we make,
like a chair (and which therefore exists only on a
condition) cannot, in ordinary speech, be called "real."
But that is because there is no even less conditioned
chair. It is our standard, compared to which a chair
painted on canvas as a stage prop is unreal.
60Austin, op. cit., p. 68.
^Etymologically, sub (under) plus stare (to be
under, or present, to stand firm).
62Ibid., p. 70.
63Ibid., p. 70.
64Ibid., p. 71.
On this cardinal point, regarding what we take
to be the meaning of "nature ," see R. G. Collingwood,
157
The Idea of Nature (London; Oxford U. Pr., 1945;
; rpt. New York; Galaxy Books, ,1960). See also
Bradley, AR, Ch. XXII.
^Austin, op. cit., p. 73.
158
CHAPTER XII
COMMON SENSE, CATEGORICAL JUDGMENTS,
AND THE PCS PRINCIPLE
A. The Objection to Bradley's 'Reality*
from Common Sense
In the preceding chapter a further discussion was
promised, dealing with Bradley's reasons for believing
that 'reality' must go beyond momentary sense
presentation. To call something "real" is to judge it
real, he thought, and yet no explicit judgment was ever
complete. What is judged of, and the reasons for it,
belong to reality; so reality is (1) transcendent to,
and (2) the subject of, every judgment. He was saying,
in other words, that all empirical judgments are
conditional, and so not capable of being known as true
unless implied by a reality beyond them. But to say that
is to affront Common Sense, for are there not some
judgments about the world which we know to be true,
whatever else may be unknown?
No one, not even Bradley, could possibly deny
that some Common Sense judgments are true.1 What must
occupy our attention here is (pace Moore) the sense of
"true" that is meant, i.e. categorically true. Under
159
some conditions Bradley could not have said "A piece of
paper is now before me” was possibly false. What he
would have added, however, was that the conditions making
the proposition true were not irrelevant: the mere
proposition was not itself true. But no reference is made
to those conditions as the proposition is set forth. We
may be certain that the proposition is true, and so
'understand' that the conditions are satisfied, and so
we can admit that some Common Sense assertions are "true"
in the sense of "psychologically certain." What cannot
be said of Common Sense propositions like "A piece of
paper is now before me" is that they are categorical.
But then, did any advocate of Common Sense ever suggest
that they were? Is there, then, any disagreement between
what Moore was claiming for Common Sense propositions and
what Bradley was denying to them?
Moore's defense of Common Sense is relevant to
Bradley's view only if Moore meant to say that judgments
like "A piece of paper is now before me" is a categorical
judgment. What is a categorical judgment? It is a
direct, affirmative judgment admitting of no conditions
or exceptions. Bradley denied that there were any such,
as empirical judgments. Common Sense disagrees with
Bradley only if it means to say that there are. In that
case a premise is denied in the argument proving Bradley's
notion of reality. That argument was given at the
160
beginning of the previous chapter. In simplified form,
it went:
(1) Whatever is real is self-existent, which
the subject of a categorical judgment must
be
(2) Any empirical judgment, if true, is known
true only if relevant conditions are
satisfied
(3) Any empirical judgment is a conditional
judgment
(4) All predication of meaning to the subject of
the judgment is conditional
(5) The subject is only a conditional subject
(6) The conditions of the subject are in part
unknown
(7) The subject of any empirical judgment is
unreal (or ambiguous, or problematic in
itself)
2
It is crucial to this argument to understand that the
meaning of the subject term of any empirical judgment is
always a conditional meaning, altered by increased
knowledge of those conditions. The meaning is not
irrelevant to the truth of the judgment. Thus, in the
more complete version of Bradley's argument presented in
Chapter II, Section A, proposition (3) read, "The truth
of any judgment is dependent upon what lies beyond it,
implied but not stated, as its conditions." That is to
say that the subject of any empirical judgment is some
thing abstracted from the whole reality in which it
originally occurred, and it loses its meaning unless the
161
whole be understood as implicitly present still. That is ;
what Common Sense denies. As its defender, Moore believedi
that parts can be known irrespective of the whole; that
accurate piecemeal knowledge is possible; that meaning is
a stable value occurring in atomic form, unalterable, like;
the truth or falsity of Russell's propositions. The
general view, held by champions of Common Sense and the
'realists' who followed them, was of a particulate
conception of meaning and reality: that among the basic
particles, relations were merely added on from outside.
These philosophers held to the "relational view';," already
discussed. Here, then, is the clear opposition between
Common Sense and Bradley; here is the relevance of our
present discussion.
B. Common Sense and the New
Philosophy: The PCS Principle
At the turn of the Century, Moore led a revolt
against the philosophical tradition Bradley represented.
Moore's arguments for the primacy of Common Sense judg
ments in philosophy won over such important thinkers as
Broad3 and Russell.4 According to Broad, the argument
that turned the trick was Moore's "Refutation of
Idealism,! ' which later was repudiated (at least in some
measure) by Moore himself.5 Nevertheless the revolt was
underway, and we must examine its basic principle
carefully.
162
The new movement began by calling into question
the whole nature of philosophical explanation and
justification.^ It assumed that there were certain funda
mental propositions which make the subject matter for all
philosophizing, the truth of which is initially presumed,
and so are not themselves subject to philosophical
explanation and justification. These fundamental
propositions simply are not questionable. These are what
Moore defended when he defended Common Sense. They are
such propositions as "The earth has existed for many
years past." Philosophy had nothing to add to these
privileged propositions: its role was descriptive, not
prescriptive. It was a mistake to suppose that rational
dialectic could force out new truths about 'reality'.
Philosophy could at most "describe," in the sense of
"correctly analyze," the Common Sense propositions.
Correct analysis was quite a different matter from
discovery of new truth. It was logically independent of
7
the latter.
What was the principle of this new movement? To
find it, suppose that certain advocates of the traditional
philosophy, such as Bradley, had thought it possible to
discover, by merely rational criticism, such interesting
truths about the world as that ordinary physical objects
Q
do not really exist. That is shocking to Common Sense,
not only to Moore's but to ours. It shocked Russell's.
163
! We will give Moore's defense of Common Sense much
attention later. Russell espoused the truth of Common
Sense propositions in principle although he preferred to
speak of "patent facts • " He said,
It has become natural to suspect a fallacy in any
deduction of which the conclusion appears to
contradict patent facts.
And lower down
The empirical outlook has become part of most
educated people's habit of mind; and it is this,
rather than any definite argument, that has diminished
the hold of the classical tradition [my emphasis].
Russell added,
[Logic is] . . . analytic rather than con
structive . . . [for it]_. . . refuses to legislate
as to what the world is.
Logic refuses to legislate about the world. Why? Be
cause something else is in the privileged position, namely,
"the empirical outlook »" This opinion about the dimin
ished role of logic, paradoxically, changed the
philosophers' world, at least. Henceforth, where there
was a clash between reasoned conclusions and sensuous
observation the deliverances of the latter were to be
believed. Any philosophy which challenged the truth of
Common Sense propositions was, in Moore's words, "as
profoundly mistaken as any view can be."10 Philosophers
were to tailor their logic to fit the form of (to "give
an account of"), in Russell's words, "the world of science
11
and daily life," not to "convict it of unreality."
164
What was the guiding principle of the new
philosophy? What characterizes the movement away from
"the hold of the classical tradition"? It was unswerving
fidelity to "the empirical outlook"; a conviction that
philosophical methods and conclusions were to be fitted
to the measure of Common Sense. That conviction is very
much alive today. So important a philosopher as Gustav
Bergmann writes,
[Philosophy has] . . . the task of elucidating
common sense, not of either proving or disproving
it. In this form the common sense doctrine also
represents . . . that philosophy is descriptive.
The defining characteristic of the new movement may now
be stated as a principle. Let us call it the Principle
of the Philosophical Primacy of Common Sense, or, the
PCS Principle. It says,
In all matters of philosophical dispute involving
disagreement between conclusions got from
philosophical principles versus conclusions got
from Common Sense, tne Common Sense Judgment xs
to be taken as true and philosophical principles
are to be rearranged accordingly.1^
By this principle categorical judgments may be
got from empirical ones. Any merely philosophical argu
ment purporting to prove that all empirical judgments are
beset by the difficulty of being conditional may be done
away with. "Categorical" means just that property which
some Common Sense judgments have. There is nothing more
it could mean. Such judgments are not just true, but
categorically true. There is, indeed, no other meaning
for "true-" If some other philosopher has a meaning for
"categorical" or "true" which is not satisfied by the
sort of truth that Common Sense judgments have, then so
much the worse for his philosophy.
C. Nedd the PCS Principle Be Justified?
Before criticizing the PCS Principle, we must
pause to consider the possibility that it needs no
justification and so is immune to criticism. It would
also be helpful to clarify the sense in which Bradley's
view is incompatible with it.
If the PCS Principle means to say no more than
that the aim in philosophy, in "describing"1^ reality,
is that the fit between them should be perfect, then
Bradley would have insisted upon it. But he would have
gone on to say that "fit" is a metaphor here, and that
the aim in philosophy is not merely to describe but to
understand— in the sense that the description and the
15
reality described should coxncide, should become one.
Until that occurred there would remain the problem of the
relation between them, a relation very important for
philosophical understanding and which must be included
somehow within reality as well. This oneness cannot be
accomplished short of transcending the subject-object
dichotomy, which means the transcending of all relational
thought. Philosophy, for him, required its own
16
transcendence — as Wittgenstein later discovered.
Perhaps such a notion was too extravagant for
those who adopted the PCS Principle. For them, then,
the problem of relations remains to be'solved, e.g. the
relation between the description and the reality de
scribed. The PCS Principle does not offer a solution to
this problem. Indeed, it seems like a proposal to
ignore it. Yet ’reality1 so described has not been shown
to be unaffected by the conditions of apprehension and
interpretation. Until that is shown, it might very well
be that the ’real* object, as it appears to us, is as
dependent upon whatever conditions the description as the
description is dependent upon the object. It just might
be so, and the PCS Principle would forever prevent us
from finding that out. What reason has been given for
supposing that philosophical procedures depend upon
Common Sense judgments? What reason has been given for
binding truth to them? What reason is there for giving
them priority? Why not suppose, instead, that those
judgments are themselves dependent upon whatever we take
for philosophical procedures? Some philosophers have
argued that the meanings we find, and the judgments we
make which depend upon those meanings, reflect our way
17
of selecting and grouping the materials at hand. Does
this view not merit consideration? To be sure, the PCS
Principle rules out inquiry into the question posed by
167
such a view, but can it simply legislate its problems
away? By its own rule, it can only describe them.
But the PCS Principle legislates important
problems away. It claims that the Common Sense judgment
is fundamental and needs no justification, for it has a
stronger kind of truth than any philosophical principle
that might be used to call it into question. It claims
that philosophical justification has got to stop some
where so criticism stops with the Common Sense judgment.
That is an important claim. It is to claim, among other
things, that the object of knowledge is logically
independent of its being known. It is to claim that the
knowledge relation is external and so does not affect
the terms. That is where the PCS Principle is in
compatible with Bradley's view.
The PCS Principle seems to have settled some very
important questions with a fiat. It would not be unfair,
then, to ask for the justification of the PCS Principle.
It does need justification. Perhaps Moore was right that
Common Sense judgments need no justification for common
and practical purposes. But Moore was speaking as a
philosopher. He himself went on to give a justification
for Common Sense. It would not be unfair to Moore to
examine that justification, inasmuch as he seems to have
18
contravened his own rule. More than that, even if it
were admitted that some Common Sense judgments need no
168
justification, still, it might be helpful to know how
Common Sense judgments can be distinguished from others.
They comprise a class which is basic to all philosophi
zing.
D. Moore's Justification of Common Sense
Moore admitted that the judgments of Common Sense
may change.^ In that case, we may object, why make such
judgments primary to philosophizing? No doubt he intended
to do just that, so a justification is called for.
This is how it went: Moore began by making a
list of propositions that he knew for certain. There
were two classes of them. The first had to do with outer
events such as that there is a body that is my body which
changes from birth to death. There are other bodies
similar to mine, and other things at various distances
from my body. The second class had to do with inner
events: states of awareness such as that I know I am a
human being having a succession of experiences, perceiving
the body and other things, and knowing facts about them
and relations between them. Also in the second class
were such facts as that I know other facts although I do
not observe them, such as that my body existed yesterday,
that it had relations with other things, that there were
expectations, dreams, and fancies, and that other human
beings have had such inner states of experience also.
169
Moore was then able to express yet another proposition
about these two classes of proposition, also known for
certain but which could not have been stated without
them, viz. that other people have known propositions
20
corresponding to all of these.
Moore knew all of these propositions were true,
using "true" in the ordinary acceptation of that word.
He was not using "true" in any sense in which a
proposition that was partially false could still be true.
He meant the ordinary sense of "true" which everyone
knows to belong to "The Earth has existed for many years
past-" He wished to make it clear that no philosopher
could question the truth of such a proposition by saying
that it all depends on knowing whether you mean by
"true," so-and-so, or so-and-so-and-so. The truth of his
propositions was beyond argument and was not to be
obscured by talk about the nature and meaning of "truth.r
That could come later. Philosophers who did question his
propositions, said Moore, were confusing understanding
the meaning of the proposition with the entirely dif
ferent question of giving a correct analysis of what it
21
meant. The latter question could not even be raised
unless the meaning were first understood.
There were two ways to disagree with this, said
Moore. One could raise (a) the bold metaphysical
objection that none of these propositions are true, or
170
one could offer (b) the more timid epistemological
objection that even if they were true, yet there was no
way to know it. Now philosophers of type (a), such as
Bradley, have denied material existence in space and
time, have denied selves, plurality, change, relations,
and all of the sorts of proposition Moore said he knew
for certain. His reply to them was simple: if what they
alleged were true then no philosopher has ever existed—
and surely nothing is more certain than that they have
existed. Moreover, no philosopher has ever held position
(a) consistently, for he says "we;1" he criticizes his
opposition, and otherwise acts in ways implying tacit
acceptance of what his official philosophical utterances
will not allow.
There was a conclusive refutation to arguments of
type (a), said Moore. All the propositions he stated were
known to be true. Their truth was beyond the reach of
philosophical argument. There was no question about it,
so the question became, "How did philosophers of type
(a) get confused? Moore had the answer ready. They
confused analysis of meaning with knowledge of truth.
But might not position (b) hold? It merely says
that Moore's propositions can11 be known to be true.
Bradley took position (a), but he also took position
(b). But (b) won't do either, said Moore; for unlike
(a), the more modest position (b) entails incompatible
171
consequences. If (b) were taken seriously, then the
proposition "There is? more than one human being"
expresses a belief which is not certain. Philosophers
holding (b) would be saying that, although there might be
other human beings, yet "we" couldn't know it; Common
Sense might be right, but there was no proof of it.
That entails incompatible consequences. Even to
recognize a meaning to "Common Sense" entails the
meaning of a plurality of selves. To speak of a "common"
error leads to the same difficulty. Opponents of Common
Sense persist in saying "we;" "you,!1 and "they;!1 and yet
argue that "we" cannot know of a plurality of selves.
That entails incompatible consequences.
In this way Moore justified his "Common Sense view
of the world." That phrase he used to characterize the
general proposition, meant to be implied when the term
"Common Sense" was used, viz. that he and others have
known and do know with certainty that some propositions
such as those he mentioned are true. And, moreover, the
sense of "true" that he meant was known prior to the
raising of any philosophical questions. This he said in
order to ward off equivocation over the word "true','y for
he recognized that all philosophers without exception
have agreed with him that, within the Common Sense view,
such propositions are true. Where certain of them
differed with him was in holding that, from some other
172
! point of view (.developed from an anlaysis of meaning),
these propositions must he false. And that was, said
22
Moore, "the height of absurdity."
Moore's position may be summarized by.saying that
he believed some propositions to be known for certain.
Philosophy begins with these, and it cannot go on to deny
them without destroying its own basis. Some philosophers
have denied them, therefore they are wrong. They con
fused a question about the analysis of meaning with the
logically independent question of what was meant, and
whether it was true. But the correct analysis of some
proposition's meaning is logically independent of the
question of its truth.
E. Commentary on Moore's Justification
of Common Sense
Moore's defense consisted of appealing to the
self-evident meaning and truth of certain propositions.
He gave no philosophical justification, for he believed
that philosophizing was a secondary activity and
dependent upon certain primary truths that could only be
pointed out. Whether or not the taking of this position
also needs justification (whether one is ever entitled
to import bodily into philosophy its unalterable
foundation stones) will be discussed in the next section.
At present we must ask whether, in defending Common
173
I Sense, Moore justified a position that contradicted
Bradley.
Did Moore justify the PCS Principle? Unless he
meant to say that Common Sense propositions are
"Categorical truths;'? then he did not deny Bradley's
position. This is not just a quibble over fashionable
terminology. Did Moore mean to say that the truth of a
proposition taken in abstraction could still be taken
seriously? Did he mean that the truth of "The earth has
existed for many years past" adheres to it as a property,
after the context within which it is meaningful and true
has been disregarded? He did not say so but he laid no
stress on the importance of the conditions under which his
truisms were true, so it must be assumed that they were
irrelevant.
But on the other hand Moore said that Common
2 3
Sense truisms may change so xt xs hard to see how they
could be categorical, or even undeniable. Somehow the
Common Sense proposition has got to allow for change,
either in its meaning or its truth. A difference must
fall somewhere that affects either the absoluteness of
the meaning or the sense in which it is taken to be true.
Moore did not allow relevance to any distinction
in the sense of "true," so far as concerned his truisms.
They were paradigmatic.^ He appealed back to his
Common Sense propositions for the meaning of words like
174
"true"earthV’ "ejcistsV and "past",." '"The earth has
existed for many years past* is the very type of an un
ambiguous expression [emphasis mine]," he said. He
emphasized "the ordinary and popular meaning" of such
25
expressions. Now he did not emphasize these things in
the same article in which he said the Common Sense
propositions could change, but unless he meant something
different by "Common Sense proposition" in the two
cases, then might one not question that ordinary and
popular meaning? With change, does the proposition not
alter in meaning, and so, alter in truth? And is that
kind of truth, after all, what we popularly and
ordinarily mean by "truth"? Are we to take that sort of
truth as paradigmatic? For, without wishing to weary
the reader with the obvious, the proposition "The earth
has existed for many years past" once meant that it was
the flat earth that so existed. Nowadays that propo
sition, taken in that sense, would be called false.
Surely an analysis of the meaning of the proposition is
called for if truth is predicated of it absolutely. But
if that is denied, then a distinction in "truth" is
called for. The difference must fall somewhere.
Another objection to Moore's procedure deserves
notice: If Common Sense propositions are to guide
philosophy, and if their truth cannot be denied without
stultifying the whole enterprise, then there must be
175
some way to know such propositions when we find them.
26
They form a class of propositions, but Moore did not
mean that just any proposition, uttered popularly and
ordinarily, fell into that class. He meant by Common
Sense propositions those which belong to "a set of
propositions'," which "resemble" in some respect, which
belong to a "type" of unambiguous expression, which
27
"correspond to" those which he mentioned. Surely he
did not mean to limit us to just these propositions.
Surely there are more. He didn't claim to have exhausted
the extension of the class, so presumably there are more.
How might we find more? They are a "type" and "corre
spond to" the "class" of those which he mentioned, but
what type? Which class? How do they correspond? What
principle defines the class? Is it that no principle
is needed, nor can be given? But then, would just any
content of immediate experience qualify as'real and true?
Suppose that any word meant whatever it seemed to, but
the meaning of "true" was not to be discussed, nor any
distinctions of its sense allowed. But are these
suppositions not, after all (and having been laid down)
first principles for doing philosophy? Would they not
continue to be such even if their utterance were
forbidden? But on such generous principles as these, how
could any disagreement ever be settled? Why would not
just any proposition be meaningful and true if it seemed
176
to be? Why would it not pass for a Common Sense
proposition? Or shall we remain with just these
propositions, taken upon authority, with philosophy
doomed never to pass beyond endless dissection of the
merely commonplace? That is something which will be
taken up later.
Might we restrict ourselves to analysing only
the truisms given by Moore without bothering to find the
boundaries of the class? If so, then how shall we reject
from consideration any other proposition claiming
membership in the privileged class? But if, ignoring
that difficulty, we do restrict ourselves to Moore's
preferred instances, then would that not be stultifying
to the task of philosophy, as stated by Moore, "to give
a general description of the whole of the universe?"
Does not Moore's "whole range of philosophy" not
28
suffer here from his restriction? There is no intent
here to be facetious. The problem seems genuine.
Surely we could not restrict philosophy to the proffered
instances (in order to ignore the problem of determining
the extension and intension of the class) without doing
violence to what is ordinarily meant by "universe "
The difficulty with using Common Sense as a court
of final appeal to settle our disputes lies with its
ambiguity. It is a court so generous in its verdicts
that philosophers holding contradictory theses may each
177
think himself upheld by Common Sense. For example, there
is a view of Common Sense which holds that recourse to it
is recourse to the sense of immediacy^9— a mistrust of
intellectual abstractions. This was what Bradley every
where insisted upon as the ground for all assertion— a
ground that does not disappear and drop from relevance
with the intellectual act of taking terms-in-relation.
It is a ground which remains present and continues to
uphold after judgment is made; after terms and relations
have been selected, throughout the process of inference
and into the validity of the conclusion. It is that by
which the judgment is true, if it is true. Abstract
from that and we abstract from truth. But then we are
involved with a condition demanding what Moore called
"an analysis of meaning" to make good the crucial
omission and predicate truth.
To contradict Bradley, Moore must have sought
from Common Sense an unconditionally true empirical
30
proposition. But to Bradley such a proposition xs an
abstraction from experience and as such (the mere
proposition) it is true or false by conditions lying
beyond itself. Moore surely knew all of this. We must
hesitate, therefore, to say he meant his Common Sense
propositions for categorical truths. But it is truly
hard to tell whether he did or not. How much weight is
the Common Sense proposition to carry, in philosophy?
178
If nothing takes precedence over its truth, then it is
fairly called "categorical! ' But perhaps that was not
what Moore meant. If not, then cadit quaestio. We
need a clear statement on precisely this point. Moore
did not discuss conditions, he discussed propositions
that were obviously true. At this juncture Bradley
would have insisted that the conditions, although dis
regarded, must be understood to still apply. If other
wise (if abstraction is to be taken seriously), then the
proposition cannot be known to be true, and cannot
safely be taken to be so. If the only alternative to
true is false, then the proposition is false.
We cannot deny the truth of "The earth has
existed for many years past". But to say that is a
different thing from saying that it is true whatever
else is false. Moore no doubt would have agreed, but he
should have discussed a matter which his opposition used
as the basis for their claim. "The earth has existed
for many years past" is true, but in it the conditions
are understood to be relevant and satisfied. If the
truth is known, then they are known. Now one may seek
to save Moore's defense by pointing out that he did not
say that the proposition was unconditionally true, so
perhaps he meant only that it was undeniably true. This
change of terms involves us in psychology. One can
179
know for certain without letting conditions enter one's
head. In that case Moore did not contradict Bradley.
But surely he did mean to contradict Bradley. That was
the whole point of his "Defence." Surely he did mean to
assert the PCS Principle. He did not succeed in
justifying it by appealing to Common Sense.
F. Moore's Conception of the Given and
Its Alternative
The most important thing Moore asserted was the
PCS Principle. It involves the problem of the given in
experience. If indeed it is true that something is
simply given— and if that something (whatever it is),
remaining the same throughout, continues to enjoy
privileged status as the base from which all philo
sophical results must be derived— then it is hard to see
how the PCS Principle can be avoided. Consequently,
Moore's defense of Common Sense might have addressed
itself to the question: Does the empirically given enjoy
privileged epistemological status? He seems to have
assumed an affirmative answer to this question, on the
basis of which he judged philosophy's task to be
descriptive. But the question itself is at least
arguably within the bounds of philosophy.
There is another conception of the given which
differs from Moore's. It was held by Bradley but also
by, among others, John Dewey. These men believed that
180
although something or other may be given, yet propo
sitions are not to be called "true" just because they
mirror the given "something;" Rather, they are true if
they fit into the system of our ideas. Here "true" and
"valid" come together. Truth is not a copy of "actual
existence1 , ' " for 'actual existence', so far as intelligible,
so far as understandable, so far as ideally apprehended
as terms-in-relation, so far as it is a taking in one of
diverse elements which permit and do not repel relations—
so far, that is, as philosophy has not been abandoned—
is a construction. And whatever it was that Moore may
have thought was given (so far as it was any "what" at
all), it receives its definition and place in that
constructed or ideal system. If it cannot be placed in
such a system, it is rejected as error. Now whatever
the merits or demerits of this view of the given may be,
yet it is obvious, at least, that it differs from
Moore's. For him, truth was a property of some propo-
sitions. For them, it was place m a system.
Both Bradley and Moore began philosophizing from
immediate experience. But for Bradley the point of
beginning was not important. Reason needed no such
external propping up. To him, Moore's position was the
result of confusion, That confused view, deriving from
Locke, begins with the general perceptual situation as
its given. Then, through logical analysis, it hopes to
reach the simple and ultimate elements. But no
perceptual situation is the outcome of such elements.
The actual and immediate situation is the beginning.
Yet, if it is the situation which is the beginning and
the court of last appeal, by what warrant may it be
dissected within or limited without? The perceptually
'given' is not so much a given as a taken; for through
analysis, limitation and separation enter. With that,
a step is taken away from the actual and immediate
situation which was (allegedly) the home of reality and
truth.
But even with so little as the perceptually
given (itself the product of discrimination and
limitation), it is unthinkable that it in truth consists
of logically simple elements. Because it i£ a whole,
it is analyzable. What is no more than a diverse
collection is not analyzable. It continues to be a
whole, and only so do the products of analysis continue
to be meaningful with reference to it. Try for a
moment to take seriously the notion that these products
are themselves ultimates, and we will find that they
may not even be called "elements " "Element" and
"ultimate individual" are not reconcilable meanings.
Without the meaningful whole they are meaningless, and
without the unity they are disunited. How may they be
got together? How is it even intelligible to say that
182
they are together? One must dispute the initial
assumptions of Moore’s theory of the given.^2 we may
try to put the objection in other words. Seeing the
task of philosophy to be descriptive rather than
constructive, and viewing the given as a collection of
individual simples that are epistemologically privileged,
those who defend Moore's position suppose that
philosophy's task is first to copy reality in terms of
its structure: thus to derive a descriptive logic.
They then extend that logic to cover the whole of
reality. As one writer put it,
Moore assimilates, within his foundation, the
idea that one fundamental fnnction of philosophy is,
to speak, to spread the validity which is the
birthright of a certain sphere to, in this sense,
other less fortunate spheres.33
The privileged sphere is, of course, the class of the
Common Sense proposition.
Now while Dewey was a "doer" while Bradley was a
"knower," yet both denied that cognition arises out of
any privileged sphere. The birth of cognitive endeavor,
according to them, comes with the experience of
disturbance or indeterminacy. Something jolts our
rational system. Something external, of course, may
deliver the jolt, but the moment we try to say just
what (to understand or know or be clear on the problem),
we find that it is our whole relational context that is
involved. Jolts may be called 'external', but that
183
only shows our system to be incomplete. We weren't
'clear' to begin with. Is it true that the earth has
existed for many years past? We never ask that question/
ordinarily, because it doesn't disturb our system. If
this is a genuine question we must ask, "What prompts
you to ask"? That is to say, what are the circumstances
under which this is a genuine problem? It must be a
genuine problem if it deserves a Yes or a No. The
context cannot be disregarded if the problem for
cognition is meaningful. Then and only then have we
any data. Otherwise, we will find ourselves agreeing
with Moore that this is an unquestionably true
proposition. Or, as the writer quoted above put it,
Each problematic situation defines its own
ultimate data within the context in which it comes
to birth, the context of primary experience. And
so we see that Dewey's whole position runs counter
to the tradition which assumes that there is
something which is at once ultimate and primary,
an absolute foundation for knowledge. For Dewey
there cannot be any description of such a
foundation. There can be, at most, a description
of how such foundations, in the plural, came to
be with the birth of thought and inquiry due to
specific problems which arise in the context of
"doing, suffering, and enjoyment [my emphasis]."34
The Common Sense proposition will not serve as a
foundation stone for philosophizing. By itself it
neither raises nor answers any question. Of course it is
not given by itself, and there are a few propositions, of
a least-common-denominator sort, of which the context
may be more or less presumed. But it may not be
184
presumed irrelevant. There is no privileged
proposition.
Summarizing this alternate view of the given we
may say that, contrary to Moore, it is possible to argue
that no propositions are epistemologically privileged.
It is not the function of philosophy to describe them.
The whole presumption is fundamentally mistaken. Rather,
it is the function of cognition to systematically re
arrange and alter such problematic judgments as arise,
so that conflict may be avoided. Cognition cannot be
content with just a description, for description is
dependent and ancillary: it is always conditioned by the
conflict-ridden situation. The end an aim of cognition,
for a "doer" like Dewey, is practical problem solving.
35
For a "knower" like Bradley the end is understanding,
only in the light of which are the elements known and
clear.
G. A Difficulty with Moore's 'Given*
We need not take Moore's view with respect to
the given. Indeed, the alternative position suggested
above is represented by such contemporary thinkers as
W.V.O. Quine and Wilfred Sellars. They have almost
swept the field before them, and rendered the Bradley/
Dewey view something of a commonplace. We need not
accept the notion of the given as an epistemologically
185
privileged ultimate. Then how shall we decide which
view is correct? I have tried to point out some of the
difficulties in Moore's position. Even more might be
done. One might argue that if philosophy can no more
than describe the structure of certain privileged
propositions, then one is committed to the "copy” theory
of truth. One might review the difficulties besetting
that theory. One might then go on to justify in its
place the theory that truth is "comprehensive doherence.”
But that would greatly prolong this discussion, and it
is doubtful whether we would say anything new.
There may be another way to incline decision in
Bradley's favor. Now Moore thought philosophy's task
was descriptive, as we have already noticed. In answer
to the question "What is philosophy?" he used the word
36
"describe" twelve times in his first three paragraphs.
Why did he give it such emphasis? Perhaps because his
conception of the given requires that philosophy be
merely descriptive. Philosophy, whose conclusions
cannot supersede Common Sense judgments, can do nothing
37
but 'clarify' those judgments. Clarity and precision
in our use of words, especially in the asking of our
questions, is very important to philosophy, but does it
deserve to be called "philosophy"? Not in the ordinary
usage of "philosophy ." But then, perhaps ordinary usage
is wrong. If so, we must confess to a wistful feeling of
regret. We had hoped for more, from philosophy.
But above all let us be fair with Moore. Did he
seriously mean to reduce philosophy to an exercise in
descriptive analysis of the merely commonplace? No less
an authority than Normal Malcolm seems to have thought
so. He writes.
The actual efficacy of Moore's . . . misnamed
"defence of common sense" consists in reminding
us that there is a proper use for sentences like
"I see the broom under the bed." ..............
He "appealed to common sense, ' 1 in the common
meaning of the words, when he reminded other
philosophers of the plain facts of knowledge. . .
Moore's misnamed "defence of common sense" was a
philosophical step of first importance. Its
effect is to alter one's conception of the
nature of philosophy...........................
In order to grasp . . . that philosophical work
cannot interfere with the actual use of language
but must "leave everything as it is" . . , one
must understand what is right in Moore's defense
of ordinary language. The latter was an advance
in philosophy because it brought us nearer to a
true understanding of philosophy itself [my
emphasis].38
We do stand on authority when we say that Moore, in saying
that philosophy is "descriptive," meant merely
descriptive.
But if so, and if, as Malcolm says, Moore's
"misnamed 'defence of common sense'" was actually a
"defense of ordinary language" which instructs us to
"leave everything as it is," then let us have a look at
this word "description:1 * It is an ordinary word. But
187
if "description" meant mere description, for Moore, then
what did he mean to deny to philosophy? He meant to deny
to philosophy the sense of discovery. He meant to deny
that philosophy's task is to discover more truth. That
is the sense in which Moore was using the ordinary word
"description”" But Moore did not take that limited
meaning of "description" from ordinary usage, for it has
the 'discovery1 sense, the 'more truth1 sense, in its
natural home, and we ought to leave it there. Why do we
"describe','!1 in ordinary language? Hopefully, in order to
find out something. We want to get as much as possible
of the whole situation before us, in reflection, so that
we can understand it. We describe in order to under
stand. Surely the word "understand" brings us closer to
39
the true and and aim of philosophy.
H. Summary of Chapter III
Moore wished to conceive philosophy as
descriptive. He wished to lay it down as a principle
that certain Common Sense propositions are to be taken
in philosophy as incontrovertibly true and so,
logically independent of merely philosophical con
sequences, This principle we have called the PCS
Principle. According to it, philosophy can only reflect
the meaning and structure of the privileged Common Sense
propositions. If so, two problems need to be solved.
188
1. May meaning and truth be imported into
philosophy from elsewhere? For Moore this was so, but
there are two difficulties with such a view: (a) Moore
is operating with two senses of "true." He has the
'atomic property' kind which belongs to the Common Sense
proposition, and he has the 'philosophical' kind pos
sessed by the analytical parts of a philosophical
description. This latter kind of truth is possessed not
by the products of analysis themselves, but rather they
have it because of their place in the whole paradigm
Common Sense proposition. (b) If contradiction in
"true" is to be avoided, by restricting "true" to only
that property which Common Sense propositions have, then
philosophy cannot yield more truth. Philosophy, re
stricted to tending to its own analytical and descriptive
business, must be totally empty.
2. Truth is to be imported bodily into
philosophy from elsewhere, in a plurality of inde
pendently true propositions. But what of the relations
among them? Is this not a philosophical problem? If
not, then what is to become of philosophy whose task,
according to Moore, was to describe the whole universe?
Shall yet other compound propositions be imported into
philosophy which state the relation between the initial
propositions? With that, the initial propositions have
traded off their Common Sense property of truth for
189
philosophical truth. For, as parts of a new proposition
(the compound proposition that states their relation),
they may be analyzed? and analytical truth (philosophical
truth) is dependent upon structure and place in the
compound proposition. The parts do not possess truth as
a property of their own.
Now there is no limit upon how compound a
proposition may be? nor does it, by being compound, cease
to be one proposition. And surely only one vastly com
pound proposition is to be allowed if all the relations
in the universe are to be imported descriptively into
philosophy. And so it would seem that, in the end, if
the whole universe is to be described, then Common Sense
piecemeal atomic property truth has got to give way to
philosophical truth.
Presumably some details of the universe may be
left out of a merely philosophical description, but the
fact of its wholeness (for the pieces are related, and
so the propositions describing them must describe the
relation) surely cannot be ignored. And so we see the
problem of relations continues to be a vexing one, and i
has returned to claim our attention. In the coming
chapter we will examine Moore's attempt to prove that
some (and even, all) relations are merely external.
190
FOOTNOTES “ CHAPTER III
For Bradley, empirical judgments are true in
the sense that the objects exist. But they are not
real. Their existence and truth are conditional, and
those conditions are never got fully into the empirical
judgment, in its explicit form. In this sense "The
earth has existed for many years past" is both true and
false. The nominal subject "earth" may be taken in a
sense not intended; a sense that must enter the explicit
judgment (proposition) if it is to have the property
of truth. The precise sense does not enter, so the
proposition is false. Something may be thought of,
called "the earth"," to which is predicated "existed for
many years past)" but the subject is not the real
subject. The real subject is the earth as seen and
known really— sub specie aetemitatis. Does that seem
startling? But the notion is simpleand plain. We do
intend it (and so are familiar with it) although it is
doubtful whether we ever achieve it. But if not, then
our predication is about a partly unknown subject over
which we have made a mere show of words. Shall we try
to justify ourselves and say that whatever the earth
may really be, yet it is indubitable that it has
existed for many years past? But now, what has existed
for many years past? The "what" is important here. It
makes the difference between false and true predication.
2
The crucial point in the shortened argument
above is between (3) and (4). Presumably the subject
may be analyzed. If other than its analytical properties
we have not got the meaning of it. If it is_ those
properties, yet why are they together? They are together
but the "why" is so important that we have not got here
a categorical judgment.
3J. H. Muirhead, ed., Contemporary British
Philosophy, 2 vols. (London: 1925; rpt. London: Allen
and Unwin, 1956), 1:25.
4
"It was towards the end of 1898 that Moore and
I rebelled .... Moore led the way, but I followed
closely in his footsteps. I think that the first
published account of the new philosophy was Moore's
article in Mind on "The Nature of Judgment'.''.' Although
neither he nor I would now adhere to all the doctrines
in this article. I, and I think he, would still agree
191
with its negative part— i.e. with the doctrine that
fact is in general independent of experience" [my
emphasis]. See Bertrand Russell, My Philosophical
Development (New York: Simon and Schuster, 1959), p. 54.
5
"This paper now appears to me to be very
confused, as- well as to embody a good many downright
mistakes? . . ." See G. E. Moore, Philosophical Studies
(London: Routledge and Kegan Paul, 1922; rpt.
Totowa, N. J., Littlefield, Adams & Co., 1968), p. vii.
6M. J. Charlesworth, Philosophy and Linguistic
Analysis (Pittsburgh, Pa.: Duquesne U. Pr., 1961),
p. 14.
?G. E. Moore, Philosophical Papers, "A Defence of
Common Sense" (New York: Macmillan Co., 1959; rpt.
Collier Books, 1962), p. 37. Hereafter referred to as
"Defense".
O
No doubt fault can be found with the way
philosophers like Bradley expressed themselves, for they
did leave the impression with many people that they
thought there wasn’t any ordinary world. But Bradley
nowhere denied that such objects appear as they do.
They exist. But it is not possible to get that into a
discrete proposition. On the basis of his rational
critique, Bradley distinguished between "real" and
"exist.1 1 Moore did not. What Moore was denying was
that a merely rational critique was sufficient to
establish that distinction. Hence his commitment to
ordinary language.
Q
Bertrand Russell, Our Knowledge of the External
World, rev. ed. (London: Open Court Pub. Co., 1914;
rpt. Allen and Unwin, 1969), pp. 18-19. Hereafter cited
as OKEW.
10Moore, "Defense ,"p. 37.
■^Russell, OKEW, p. 55.
1 2
Gustav Bergmann, The Metaphysics of Logical
Positivism (New York: Longmans, Green and Co., 1954),
p. 35.
l3No one could fairly accuse Russell of having
held to this principle consistently. Reference is here
being made to the beginnings of "the revolt into
pluralism'1 -" Russell and Moore were together long enough
192
to lead this revolt, but they soon drew apart in their
views.
•^The word "describe" is set apart here for
special emphasis because it is popular today to agree
that philosophy's task is descriptive. This is a part
of the legacy which Moore has left us. He said, "The
most important and interesting thing which philosophers
have tried to do is no less than this; namely: To give
a general description of the Whole of the universe"
(Main Problems, "What is Philosophy?", p. 13). He used
the word "describe" twelve times in the first three
paragraphs of this article, while affirming his intention
"to try to give a description of the whole range of
philosophy." But see Appendix E, "Is Philosophy's
Task Descriptive?"
Briefly, the trouble with the view that philosophy
is "descriptive" (taking that word in its normal sense)
is.that we are led to believe we have, in our description,
the way things are: reality. The word is tendentious;
it is philosophically 'loaded*. No doubt we do intend
to 'describe' reality, in a special philosophical sense,
in the end, but if we make description also our
beginning, in philosophy, then we are, from the start,
bound to the Common Sense judgment for our paradigms
of logical structure and truth. That is why I say that
Moore's conclusions were implicit, from the start, in
the beginning which he made; in his very conception of
the nature of philosophy's task.
15
There is a difficulty here, we must admxt. In
philosophy we are trying to say the truth; to actually
get reality down and out, to get it into words. We try
to remain within the relational view while also trying
to transcend it. Perhaps we cannot, wholly. And yet
we can achieve a feeling of satisfaction that now, by
our words, we have ourselves got our ideas systematically
arranged, and therefore, 'clear'. We have become
something better. We understand. And even if we do fall
short of that, we ought not to use words, such as
"describeV which forever put reality beyond our grasp.
"Describe" extrudes us from the very universe with which
Moore wished to deal in its "whole range",/' in a way that
"understand" does not. For in part, at least, we can
and do understand. "Describe" raises pseudo-problems,
such as "other minds', the 'thing-in-itself', and
rules out inquiry into that famous old standby, the 'mind-
body problem'. Sundering the describer from the reality
described, it needs a miraculous faculty for getting them
together again.
193
Not its own denial.
17
See W. A. Sinclair, The Conditions of Knowing
(New York; Harcourt, Brace and Co., 1951), pp. 172-73.
lfl
°Why did Moore give a defense of Common Sense
when, to him, it needed none? Was he speaking ex
cathedra (not as a philosopher)? It seems so. He was
reminding philosophers of their own presmmption of
truth in certain propositions, which they must have to
go on and philosophize. They seemed, to him, to have
forgotten what they took for granted. Some of them
seemed to have developed a peculiar form of amnesia which
allowed them to go on and say strange things. I think
Moore did hold that opinion about certain of his predeces
sors, and I shall try to show that it was an indefensible
presumption.
19Main Problems, "What Is Philosophy?", pp. 20-21.
29 "Defense'i" pp. 32-45.
21Ibid., pp. 36-37.
22Ibid., pp. 44-45.
23"What is Philosophy?", op. cit., pp. 20-21.
24
Moore thus deserves at least part of the
credit for discovering the "protocolsatze"
25"Defense'; " p p . 3 6 - 3 7 .
26Ibid., p. 32.
27Ibid., pp. 32-25, 37.
23See Moore, "What Is Philosophy?", pp. 13, 27.
2 9
In this sense Bradley was also a champion of
Common Sense. See F. H. Bradley, Collected Essays,
ed. H. H. Joachim, 2 vols. (Oxford, Clarendon Pr.,
1935? rpt. Freeport, N. Y.: Bookd For Libraries, Pr.,
1968), 2:68.
39I use "empirical" to mean "knowledge derived
from experience through the senses," and so, to include
the Common Sense judgment. The manner of derivation
may not have been important, to Moore, but its truth did
take precedence over all.
194
31
But why xs not "place in a system" a property
of some propositions? If so, Moore was right after all.
Can we regard 'place* in some system of meaning-
conferring coordinates as a property of the thing which
'has* the place? Certainly we. can, for we do. But the
question is, "Can we do so consistently?” We cannot.
No doubt we can regard "north of New Orleans" as a
property of Chicago, but, as we have seen in Chapter I,
this so-called 'property' is relative, and the relation
cannot be reduced to a property of the term"! By similar
argument, we cannot regard truth, conceived of as place
in a system, as a property of a proposition.
32
Defenders of Moore would be quick to poxnt
out that the products of philosophical analysis do
derive their meaning and truth from their place in the
paradigm Common Sense proposition. Thus, within these
limits, they are "elements" so far as parts within a
whole.
But now we seem to have two senses of "truth"
going here. There is the truth which Common Sense
propositions have, which is ultimate and unquestionable.
It is a simple property of some propositions. But then
there ..is another kind of truth: philosophical 'truth',
which deals with the standard of correctness for
descriptive analysis. It is place in a system: the
little sub-system which the Common Sense proposition
gives us. We must be faithful to the structure of that
little system if we wish to claim 'truth' in philosophy.
It is the paradigm, and our business is to 'describe'
it.
I have assumed that Moore believed in ultimate
simples, but I may be wrong. According to an admirer of
Moore (Norman Malcolm, Knowledge and Certainty, "Direct
Perception" [Englewood Cliffs, N. J., Prentice-Hall
Inc., 1965], p. 73ff), it is no worse than "a curious
inclination" that Moore held "two apparently contra
dictory" views on perception: (a) We directly see parts
of the surfaces of physical objects, and (b) we cannot
directly see a physical surface, for we see sense-data.
Moore justified this apparent contradiction by arguing
that it depends on how you mean "see"'.' If you mean so-
and-so, then (a). But if you mean so-and-so-and-so,
then (b). It is curious that what was, for Moore's
admirers, no worse than a curious inclination would have
been, if done by Bradley et al., a fatal error and "the
height of absurdity". But the more direct issue here
195
is whether, for Moore, knowledge is built up piecemeal
out of ultimate simples. Apparently so. In "A Defence
of Common Sense" Moore said, "But there is no doubt at
all that there are sense-data, in the sense in which
I am now using that term [op. cit., pp. 53-54]." That
statement seems definite enough, apart from some
equivocation about what we mean by "sense-data'V In
that case he did not really see a hand, for he saw sense-
data. He directly saw ultimate simples. Either that or
Moore was in fiatcontradiction of himself, and it is
hard to say which it was.
An admirer of Moore must explain three apparent
contradictions in his master's doctrine: (1) Moore
forbade equivocation over so difficult and profound a
word as "truth,"" but did not hesitate to equivocate over
so ordinary and simple a word as "see"." (2) he (a)
directly saw a hand, but (b) he only inferred a hand from
sense-data. (3) Although he forbade equivocation over
"true,'" he had two senses of it going himself. There was
(a) the simple-property kind, had by Common Sense
propositions, and (b) the kind which guides philosophical
analysis, i.e. coherence among parts so that they render
the C. S. Proposition faithfully.
33C. Miller, "Moore, Dewey, and the Problem of
the Given/"" Modern Schoolman, 1961-62 (39), p. 382.
^5
Bradley vigorously disputed the notion that
the end of cognition is mere practice. For his remarks
on Dewey see ETR, pp. 127-142.
36
Main Problems, "What Is Philosophy?" op. cit.,
pp. 13-14.
37
While passages in Moore's later writings seem
to belie this assertion, others support it. It is
difficult to give any consistent summation of what
Moore's true position was, but the opinion cited above
is supported by Brand Blanshard (Reason and Analysis
(La Salle, 111., Open Court Publ. Co., 1962), p. 310),
Norman Malcolm (Knowledge and Certainty [Englewood
Cliffs, N. J.: Prentice-Hall, Inc., 1963], pp. 163-83),
and Gustav Bergmann (The Metaphysics of Logical
Positivism [New York: Longmans, Green and Co., 1954),
p. 35).
33Malcolm, op. cit., pp. 181-83.
39See Appendix E, "Is Philosophy's Task
Descriptive?"
196
CHAPTER IV
MOORE'S REFUTATION OF INTERNAL RELATIONS
A. The Relevance of the Question of External
Versus Internal Relations
At the conclusion of the preceding chapter the
problem of relations had again assumed importance.
That problem entered through the emphasis placed by
Moore on the wholeness of the universe which, according
to him, philosophers try to describe in their special
way. Is the fact of this wholeness to be disregarded?^"
Did Moore, whole italicised the word "whole;" mean
to say that he was taking that wholeness as fact? If he
did, then is that consistent with a view taking reality
to be a property of independently existing things which
together make the universe? What is the logical force
of the "together" here? The "together" would seem all-
important to the universe. But is the "together" nothing
to the isolated fact? Is it sometimes something to it,
but not always? Or is the universe just a bare con
junction of externally independent facts, even though,
within some facts, there may be an occasional internal
relation?
The important issue here is that of external
versus internal relations. The issue is this: Granted
197
that to think of the universe is usually to think of a
plurality of facts in relation— presumably Moore's view—
2
yet is the togetherness of all those facts a necessary
relation? Obviously, three views are possible on this
issue: (a) No relations among facts are necessary,
(b) Some such relations are necessary, and (c) All
relations among facts are necessary, if we mean to speak
seriously of a universe. This last is the internal
relations view of the world. Ahy combination of (a) and
(b) would be an external relations view of the world.
It is a view which may be held in varying degree. The
demonstrated reality of any one external relation is
sufficient to refute the internal relations view. Moore
not only held view (b), he believed that the same
argument that refuted (c) would bring one at last to
B. Moore's Procedure Against Internal
Relations
We will examine Moore's argument aimed at
4
refuting the internal relations view, (c) above. In
that argument, Moore first set out to define clearly one
proposition which is always implied by the internal
relations view, and to show that proposition to be
certainly false. He believed it to be not easy to find
such a proposition, but there are two propositions which
definitely do not capture the meaning of the internal
relations view. The first of these unsatisfactory
propositions says that external relations assert only a
"togetherness" of terms and relation, whereas internal
relations assert something more than that. This mere
"more" could not be the crucial difference, Moore said,
for the "more" must say how the relations relate. Mere
"more" tells us nothing about "relate .," It is true, as
Bradley had insisted, that the relational fact is more
than the terms plus relation. No fact is identical with
its constituents, in the sense that it is reducible to
them, but this was obviously true of all relations,
whether internal or external. The proposition asserting
the internal relations view could not mean "merely this
5
obvious proposition."
The second unsatisfactory proposition was "All
relations modify their terms."® The proper sense of
"modify" was "change';" but it was obviously false that
all relations change the terms, for some terms enter
relation without changing at all. But, Moore went on,
if proponents of internal relations do not mean "change"
by "modify;!' what do they mean? Moore had the answer:
"They must be using 'modify' in some metaphorical
sense other than its natural one."
Having thus proved that proponents of internal
relations had been misled by a metaphor, Moore set out to
answer why they had made such a blunder over the meaning
199
of "modify.'!' It might have come about from an instance
of such a relation as "father of" , ! ' where A is father to
B, C, and D. A's children, B, C, and D, are each related
to him in a slightly different way. It may seem plausible
in some cases that a man would be different were he not
related to his child as he is. Edward VII would seem to
have been a different man were he not father of George V.
But this can be explained simply by making a distinction?
not in the man, nor in the fatherhood relation, but
rather by imputing to the man a relational property that
is particular to each case. It modifies the man, but
"it is quite clear that this property is not itself a
relation." The confusion between relation and relational
property, said Moore, enabled proponents of internal
relations to overlook the metaphorical sense of
"modify’ . ' V8
Having thus disposed of propositions that did
not state the doctrine of internal relations, Moore found
the one he was looking for. It was: "Of any term A
which has relational property P, any term which has not
9
P would necessarily be different from A1 ."
Next, Moore set about his customary task of
clarifying this proposition. What does "would necessarily
be" mean here? It means that if any term A has property
P, it follows that it also has Q, in the same sense of
200
"follows" we use when we say that if A is a right angle,
it follows that it is also an angle.'*'® What does
"different from" mean here? It means that if any term
X has not property P, it follows that it is not the same
as A: it is not only qualitatively different, but
numerically different. Qualitative difference does not
entail numerical difference, Moore said, for examples can
be found where the former occurs but not the latter. A
colored patch, half red and half yellow, has relational
properties got from each component; and in that case the
loss of either would entail numerical difference. But
this is a mere matter of fact, as we see if we consider
either half alone. "The red patch might perfectly well
have existed without being part of that particular whole,"
said Moore. Even though the red half, say, had got a
relational property of its own from its place in the
whole, it would not have been numerically altered by loss
of that property. Moore thought that he had now found
the fatal flaw in the dogma of internal relations, and
proceeded to a logical demonstration which would baring
11
the absurdity out openly.
C. Two Preliminary Comments on Moore's
Procedure
Before turning to a detailed consideration of
Moore's demonstration, two remarks may be made here.
First, the introduction of a "relational property" alters
201
the question which proponents of the dogma were trying
to ask. Why could not the question be precisely the one
which Moore said it couldn't be, viz., If relation
changes, do the terms change? Why not mean "modify"
in its "proper" sense, which is "to change"? It is not
obviously absurd to ask it. Ex hyp, we see the red and
yellow halves together in the whole colored patch, and
we only imagine them separate and unchanged. Proponents
of the dogma had an argument to prove that change of
relation must change the terms. It is no reply to say
that, inasmuch as they obviously don't change, then
proponents of the dogma couldn't have been asking what
they were asking. It is not obviously absurd at all.
Is it a meaningful question to ask, "If relation changes,
do the terms change?" Apparently so, for Moore answered
the question in the negative. So, there is no need to
reword it.
But Moore did reword it, because it was so
obviously not so that proponents of the dogma must have
been misled. How? By a metaphor. So the question
became, "How can we account for alteration in the term
which does occur in some cases, such as in the case of
George'V's father?" Well, obviously Edward VII wasn't
changed by begetting a son. The term was not changed,
and the relation was independent, so the difference must
202
fall elsewhere. Moore then brought in a "relational
property" to account for the difference we seem to find
in a king with, rather than without, an heir to the
throne. But in doing so Moore made the original question
meaningless. If, wherever a change of relation seems
to change the term, the truth is that (instead) the term
has only acquired a relational property, then can we ever
ask, "If relation changes, do the terms change?" No, for
our critic could always reply: "Ah, but you see the
relation hasn't changed, it's just a matter of adding or
subtracting relational properties." And what if one
persisted by asking: "Well, then, does adding or sub
tracting relational properties change the terms?" One
could expect the reply: "Ah, but you see the terms
don't change, they just lose or acquire a property." By
the use of Moore's technique the question, "Does change
of relation change the term?” never could be asked. It
would always be buffered by relational properties. Moore
set the problem up so that he wouldn't have to answer
that question. His question was, "If term A has property
P, would A still be A without P?" And the answer to that
question is, "Obviously so, if you keep to the bare and
abstract A— if, that is, you can substitute without
penalty an unconditional for a conditional judgment."
Moore thus set about refuting the dogma by changing the
question.
203
The second remark we must make deals with Moore's
red-and-yellow patch. His conclusion was that, although
the patch would be altered were the red part removed,
yet "the red patch might perfectly well have existed with
out being part of that particular whole." Thus some
relations, such as that of part to whole, are merely
external. But on what does this last assertion of
Moore's depend, other than his prefatory phrase, "It
seems quite clear that . . . "?^2 Whether or not a part
remains unchanged in abstraction from the whole in which
it actually occurs is a question which has demanded
the prolonged attention of the very subtlest minds.
Merleau-Ponty inquired into this question at length,
and reached a conclusion opposite to Moore's. It seemed
to him quite clear that a specific quality, such as red,
gets that quality from the whole context in which it
occurs. Contrast, for example— such as is furnished by a
yellow part of the whole patch— affects our perception
of red. In Merleau-Ponty's words,
We now discover the core of reality: a thing is a
a thing because, whatever it imparts to us, is
imparted through the very organization of its
sensible aspects.^3
The original patch was the whole one, as parts of which
the red and yellow had their place. They were "real"
only as parts, and such they should remain. But in
Moore's discussion the original relationship was dis-
204
i regarded. Then how could their relation be thought of
Other than as external?
D. Moore's Demonstration of the
Fallacy of Internal Relations
Moore next turned to his demonstration of the
fallacy of the dogma of internal relations. To do that,
he first drew a distinction between two forms of the
proposition always implied by the dogma, which proponents
of it confuse together. He said,
These two propositions, the one which I admit to
be true (1) that if A has P, and X has not, it
does follow that X is other than A, and the one
which I maintain is false (2) that if A has P,
then from that proposition with regard to any
term X that it has not got P, it follows that X
is other than A, are, I think, easily confused
with one another. And it is in fact the case
that if they are not different, or if (2) follows
from (1), then no relational property is external.
For (1) certainly is true, and (2) is certainly
equivalent to asserting that none are.
To make his point more clear Moore added,
(1) asserts that if A has P, then any term which
has not, must be other than A. (2) asserts that
if A has P, then any term which had not, would
necessarily be other than A. And when they are
put in this form, it is, I think, easy to see why
they should be confused: you have only to confuse
"must" or "is necessarily" with "would necessarily
be."14
It is important to see that the "would necessarily be"
of (2) does not follow from the "must" of (1).
Moore said, of the above distinction, that he
"cannot help thinking that it is not clear to every
body," so he set about showing the difference between (2)
205
and (1) in perfectly strict form. Here is how it went:
Let A be the term to which relational property P belongs,
and let that proposition (2), which is always implied by
the dogma of internal relations, be: "Of any term A
which has P, any term X which has not got P would
necessarily be other than A." That is, it follows that
the other term X is not A, in the same sense that it
follows that a right angle is also an angle. Let
"entails" be the converse of "follows Thus, if q
follows from p, then p entails q. So to affirm, of any
X, that from its having property P it is deducible that
it has also property Q, we can say simply that XP entails
XQ. To deny it, we say XP entails XQ. And to say that
any two things, A and B, are not the same we say,
"B = A". Let "There is an A having property P" be "AP";
use for "and," and now we can write the dogma of
internal relations in symbolic form, thus: "(XP . AP)
entails (X = A) ." We have written, in symbolic form, what
is logically equivalent to saying "(X = A . AP) entails
XP", moving from modus tollens to modus ponens. Or, the
modus ponens form reads, "Of any term X which is identi
cal with A, where A has property P, it follows that X has
property P."
Next, a way was needed to symbolize relational
property P as internal to any proposition possessing it.
To say that, we say, "There is nothing X which has Q and
206
not R," or, "It is not the case that XQ is true and XR
is false." This may be written "XQ * XRi."1^ Moore said
that this does not mean "XQ implies XR" in the sense,
called "material implication^" which Russell used in his
Principles of Mathematics; "For that leads to saying
that if AQ is false, then it implies every other
preposition, both true and false, and this is not what
16
'implies' ordinarily means."
Now proposition (2) can be written in symbolic
form: (2) XP * [YP entails (Y = X)] or XP * [(Y = X)
entails YP]. This expresses the dogma of internal
relations as it pertains to every case of relation.
Next, Moore set about expressing proposition (1),
which proponents confuse with (2). Moore wished to make
very clear the difference between the "must" of (1) and
the "would necessarily be" of (2). He regarded (1) as
true, but (2) was false. (1) said that if A had
relational property P, then nothing not having P could
be identical with A. That could be written,
"(XP . AP) entails XP * (x = A) •" This is true of some
A. To assert the same of anything whatever, write,
"XP entails [YP * (Y = X)] . ” Now compare (2), which
Moore said was false, with what we have written for (1),
which is true:
(1) XP entails [YP * (Y = X)]
(2) XP * [YP entails (Y = X)]
207
These are obviously different sets of symbols, so doesn't
that prove incontrovertibly that (1) and (2) are
different propositions?
Perhaps the difference of content is not
immediately clear from comparison of these two formm-
lations. But it can be seen that (2) does not follow
from (1), Moore said, if we substitute propositions of
ordinary language for the symbols. Let X stand for
"every book on this shelf"; let P stand for the property,
"blue"; let Y be "my book"." Now (1) says, "From the
fact that every book on this shelf is blue, it follows
that, if my book is not blue, it is not a book on this
shelf." But (2) says something very much more remarkable.
(2) says, "It is neeessarily true that every book on this
shelf is blue, if from the fact that my book is not blue
it follows that it is not a book on this shelf." But "as
a matter of fact," said Moore, this deduction is "quite
17
certainly false."
"Every book on tihas shelf is blue" does entail
that my book, if on this shelf, is blue, as (1) truly
asserts. But (2) is obviously false, for it could be
that "Every book on this shelf is blue" were true, yet
false that "My book is on this shelf entails that my
book is blue." Here we can see the force of Moore's
"would necessarily be" in (2), as contrasted with the
mere "must" of (1). Proposition (1) states the
208
uninteresting truth that, from the given fact that all
books on this shelf are blue, we can deduce that no non
blue book is on this shelf. This is the force of the
"must*".' But (2), with its "would necessarily be," says
something astonishing. It says that it cannot be false
that every book on this shelf is blue, if from the fact
that it is blue we may deduce that it is a book on this
shelfI That is obviously false, said Moore. We cannot
deduce that a book is on a shelf just from its being
blue. There is no necessary connection between the color
of a book and its spatial location!
The dogma of internal relations seems to be
saying that if X has relational property P, then its
having of that property is a necessary truth. Proponents
of external relations, on the contrary, are saying that
the having of relational property P is a mere matter of
fact. They are rejecting a view about "follows.1 " ! Moore
was saying that, from "XP * XQ"'/ it does not follow that
"XP entails X Q ' ." ! If it did follow, said Moore, "then
of course it would be impossible to distinguish between
(1) and (2)'r. and so the dogma would be true. But that
it is not true can be seen by noticing the difference
between "Q can be deduced from PP" and "not both P true
and Q false." To believe that these mean the same, said
Moore, "is simply an enormous 'howler', and I do not
think that I need spend any time trying to show that it
209
is so."^^
Moore added one last example to clinch his
argument: If in fact all persons in this room are more
than five years old, then it follows that anyone not more
than five years old is not in this room. But it is not
true that anyone in this room would necessarily be more
than five years oldl Being in this room and more than
five years old are not related as is the property of being
I Q
a right angle to that of being an angle 1
E. Did Moore Confuse Abstract "Things" with
Conc'rete' Existents ?
What would Bradley have said about Moore's refusal
to grant that "my book" and "blue" entails location on
some shelf? He would have insisted that Moore was quite
right. The bare conjunfction of two abstract terms was
never, for Bradley, a necessary truth. But neither could
one say, from such meagre information, that the relation
was not necessary. On what grounds did Moore mean to
deny it? Was the affair over the blue book and the shelf
a matter of actual judgment, intended to qualify reality
truly, or not? By what right did Moore appeal to his
Common Sense to determine absolutely that the bare con
junction of "blue" and "on the shelf" was a mere matter-
of-fact? Bradley, too, appealed to common sense, for he
declared that by immediate experience we know that the
actual affair is not limited to the bare conjunction of
210
! two abstract terms. In order to live we must act, to
i
act we must judge, and to judge we must abstract terms-
in-relation. But we cannot say that reality consists of
; them.
Moore appealed to immediacy with phrases like
"my blue book" and "on this shelfbut the designative
words such as "this"," "my"," and so on, never get the
conditions of actuality into an explicit judgment. It
remains conditional and abstract. The relational affair
referred to is no bare conjunction of "shelf" and "blue
book" and the mere use of designative words does not make
the affair complete. If we want reality in our judgment,
then there is nowhere to stop, with our naming and re
lating, until we have got the whole reality into the one
judgment.^® When the importance of doing this is
recognized, then the proposition "'My book, if on this
shelf, is blue' is a necessary truth" seems far less
absurd. It was Moore's burden to prove that it was
absurd, not that it merely seems absurd at first glance.
In his article "Internal and External Relations,"
Moore sat out to do two things: (a) to find that
proposition always implied by the dogma of internal
relations, and (b) to show it to be false. We have
already argued that his formulation of the proposition is
questionable. We must also argue that his demonstration
of falsity in the proposition which he did formulate is
211
questionable. The crucial point is, as he said, with
the meaning of "follows*" If it is true that no non
blue book can be on a shelf, does it follow that it would
necessarily be blue to be on the shelf? We must argue
that, if only we knew the whole situation, in which this
affair of the blue book and the shelf has its place, then
it not only would, but in reality it does, actually
follow that this book being blue, and on the shelf, is
necessarily true. Whether or not we can deduce it
depends on how much of the real situation we know. But
our mere ignorance cannot be used to deny a necessary
truth. And yet to affirm it seems too bold, for it means
affirming, as Leibniz did, that everything we take for a
mere matter-of-fact is, in reality, a necessary truth.
And yet, Moore did not prove the contrary. His
appeal to Common Sense to justify mere matter-of-fact
amounts, in the end, to an ad ignorantiam fallacy. We
call the relation of the book to the shelf a mere
matter-of-fact because we do not know it to be other
wise. But that makes it only a possible mere matter-of-
fact. In that case it might be impossible, in reality,
that the relation is not a necessary truth. Moreover,
the appeal to a "relational property" only transfers the
difficulty. What, we must ask, of this "having" by A of
relational property P? Is that not a relation? Does
calling it a "property" solve the difficulty? Can we
212
avoid a philosophical problem by a change of words?
Would this not be a circumlocution, like the famous
'explanation' of how clocks keep time, by referring
to their "time-keeping function"? Surely this problem
is genuine. What is the relation between a thing and its
property? Is that relation external, or internal, or
sometimes one and sometimes the other? That is the new
question arising with the move to "relational property."
It is the same old question about external relations.
Bradley denied them for they begat vicious regress.
Chapter I (section E) of this essay presents five
ways in which the relational view contradicts itself.
While those contradictions apply to any view which takes
reality as terms-in-relation (whether taken’ as externally
or internally related— see section E, 4), yet, as
Bradley elsewhere said, the internal view is a "truer by
21
far" than the taking of relations as merely external.
The internal relations view does not overlook the
relevance of the Whole situation, to the relation between
the book's place and its color. It is that whole
situation which we must take into account, if we are
either to affirm or deny that the relation is necessary.
On what grounds did Moore deny it? Merely
because propositions (1) and (2) were formulated
differently? No, for he pointed out that unless we
first see that "not p true and q false" does not mean
213
"p entails q^' then (1) and (2) are equivalent. We must
see the difference between "must” and "would necessarily
be". What is that difference? Let us notice that dif
ference. The "would necessarily be" of (2) is the
conditional tense of the "is necessarily" of (1). But
Moore nowhere mentioned the conditions that would enable
us to deduce the relation between the book's place and
its color. He just said that it was "an enormous howler"
to think that "must be : ; , " or "is necessarily " means the
same as "would necessarily be". Tt was merely absurd
simpliciter to think that a book's place and color are
necessarily connected. He appealed in the end to our
22
Common Sense; that is, to our ignorance. Then what did
he accomplish by putting his argument into strict
logical form? He didn't clinch his argument by logic;
in the end he invoked the PCS Principle.
F. Is Every Matter-Of-Fact
A Necessary Truth?
If we wish to defend the doctrine of internal
relations we must say that every matter of fact is,
in reality, a necessary truth. Moore has proved that
much, at least. We must try to prove that the universe
is throughout a necessary system— which is a large under
taking. We may need help. Let us call on Brand Blan-
shard.^
According to Blanshard, it is not possible to
214
prove that the universe is a necessary system— that there
is no mere matter-of-fact— yet we may postulate it.^
Then we can proceed with an empirical study to see to
what degree necessity applies in the world. Blanshard
believes that the postulation of necessity is justified
by such a study. But let us first review the alterna
tives to universal necessity to see why Blanshard
believes that they all have failed.
The alternatives to universal necessity fall
into two groups: 1. Necessity is psychological in
origin, and 2. Necessity applies only to syntax, as a
linguistic convention. These alternatives will now be
examined.
1. Necessity Is Psychological in Origin
The most famous argument that the idea of
necessity is psychological in origin was advanced by
Hume. If the world were a necessary system, so that
every matter-of-fact were a necessary truth, then,!
according to Hume, we must presume that the logic
followed by human reason coincides with the order of
causation among events in the world. Hume attacked
this presumption. He argued that there is no observed
compulsion among events, and all that can be observed
is regularity of sequence. After repeatedly observing
event A followed by event B, a habit of thought is
215
formed so that we expect B, given A. "If A, then B" is
the structure of the habit thus formed. We then project
this 'logic' upon the world, and seem to find 'causation'
among events. But we cannot identify logic with the
order of the world, Hume argued, for in logic, given an
antecedent, the consequent must be affirmed. But in the
world it does not have to be so. There is no contra
diction in saying that some 'cause' or 'effect' occurred
in the world without its consequent or antecedent. If
the order of events in the world were a matter of logical
necessity, then causally related events couldn't even be
imagined separately. But plainly they can be.
Hume's argument depends upon the truth of three
propositions: (a) all ideas come from sense-experience,
(b) cause and effect may be conceived separately without
contradiction, and (c) necessity is never apprehended in
2 c ,
the natural order of events. None of these propositions
is convincing.
(a) It is not obvious that all ideas must
26
originate from sense experience. Those who believe in
universal necessity do not claim to have got the idea
from experience. So, it would be no refutation to them
even if it were true that there is no sense-impression
of necessity. But is necessity ever found in experi
ence? Yes. It is necessarily true that pink is more
like red than blue; that hardness and softness are
216
different sensations; that the angle inscribed in a
semi-circle is a right angle; that a camel cannot pass
through a needle's eye.
(b) Is causal event A conceivable apart from
effect B? Certainly, if we are content to manipulate
superficial appearances in abstraction from the reality
where they actually occur, and yet call the result
'truth'. But if reality is meant, the case is other
wise. Hume stated the relation between cause and effect
in terms of imagination. We can imagine one billiard
ball striking another without rebound, but that proves
nothing. That we can juxtapose in our imagination two
abstract images contrary to the order in which they
occur does not prove that the natural order is not a
27
necessary one.
(c) Necessity is apprehended in the causal
order of nature. While it is true that a conclusion
arrived at in thought is the result of a complex of
factors, yet, as Blanshard says, "the logical relation
connecting premise to conclusion . . . has a part in the
28
emergence of the conclusion as a psychical event."
We see this through several examples: Given a series
of numbers in uniform sequence, we are able to 'go on in
the same way'. Given any three points we can determine
a plane, upon that plane determine an arc of regular
curvature, and from it complete only one circle. Such
217
processes as these apply to objects in the universe.
Now we may grant that life in the full is richer and more
complex'- than an exercise in geometry? so we must not
say that logical necessity, as we apprehend it, is always
a sufficient condition for arriving at a correct
conclusion. A logical conclusion, if thought, is also
a psychical event for some thinker, to which other causal
factors may apply. For most of us the scope of our
knowledge and the powers of our concentration soon reach
a limit; and yet, if some logical necessity can be seen
in nature, and if beyond that it is only human weakness
that raises a barrier, then we may draw from that no
argument against the universal rationality of the world. 1
He must be a super-rationalist indeed who could infer
each and all details of the world from the rest, but the
mere observation that somewhere there is a region of
ignorance does nothing even to suggest that there really
are surds in nature.
There is another reply to Hume, not mentioned
by Blanshard, which I think Bradley would have insisted
upon. Let us put it as a question: What would Hume
have replied if one simply and flatly denied his own
first premise, viz. that epistemologically privileged
experience consists of a succession of discrete sense-
impressions? Suppose we said that this was not rock-
bottom at all. Suppose we said, contrary to Hume, that
218
to us life is a felt unity within which distinctions are
made, but that the necessary connection among the dif
ferences was prior to the setting of them out in
30
thought? Does this view of ours not agree at least as
well, if not better, with experience? If we disregard
involved argument and betake ourselves to life, as a man
like Hume would wish to do, what do we find? Do we find
a hodge-podge of discrete sense-impressions, each setting
on its own bottom, aloof from the rest? Are they hard
little self-contained autonomies which are intelligible
solely upon their own warrant?
Do we not find, to the contrary, an initially
vague unity within which the differences, when made
thinkable, are related and in order? Does it not take an
additional effort to disjoin them in thought? This last
view is, at least, as plausible as Hume's. On what
grounds would he have forbade us to adopt it? There is
none. He set the problem up too hastily; he did not
exhaust the possibilities.
In summary, the psychological objection to
necessity is unsatisfactory. It denies necessary
connection in the world, as cause and effect, because the
31
latter, it says, is never found. And because xt takes
32
all ideas to be of sensory origin, it concludes that
there can be no idea of necessary connection. But there
is a difficulty here. Is it known that all ideas are of
219
sensory origin? If not, then cadit quaestio. But if so,
then the connection between idea and sensory object must
be necessary. How else could it be known that all ideas
are of sensory origin? But this result ruins Hume's
conclusion. He said,
All belief of matters of fact or real existence is
derived merely from some object present to the
memory or senses and a customary conjunction
between that and some other object.33
His conclusion, that "customary conjunction" makes all
knowledge of matters of fact doubtful, is inconsistent
with his knowledge that all ideas are of sensory origin.
There is known to be a constant conjunction between the
sensory object and any idea, but if the sensory object
is unknowable, how can that claim be made?
2. Necessity Is A Linguistic Convention
Next, we must examine the objection to universal
necessity called "linguistic formalism." This view of
necessity regards it as applying only among parts or
acts of speech, as the result of an agreement among
speakers. The 'necessity' which prevails in philosophy
is that of tautology. For example, it is necessarily
true, but quite uninformative, to say that a bald-headed
man has no hair. The only necessity is the syntactical
one which words have within some established convention.
Three closely connected propositions summarize this view
of necessity:
220
(a) Necessary connection holds only among parts
or acts of speech.
(b) Such connection is conventional, hence,
arbitrary.
(c) Necessary connections are analytic.
But there are difficulties. (a) If necessary
connection holds only among parts or acts of speech, then
it does not hold among items in the world. Then it is
denied that any red thing is necessarily colored, or that
the diagonals of a square are necessarily equal. This
result would be astonishing to painters and carpenters.
It is true that we do have a convention by which the words
"red/" "colored;" "square1 ;" "equal," and "diagonal" are
used in a certain way, but is this a mere convention? Is
just our use of words involved? Then how are we to
report our use of words so as to construct that syntac
tical system to which, it is claimed, necessity is
limited? If we attend only to words, and not to the
meaning of those words, how are we to know the different
uses, or conventions, which those words serve? Words
qua words are all alike, and as sounds or marks on paper
they are all different. They cannot be sorted out as
parts or acts of speech unless we deal with them as
meanings. But with necessity among meanings, thesis
(a) collapses, for then the word must refer beyond the
mere parts or acts of speech.
221
(b) Can the classification of the parts of speech
be conventional, in the sense of merely arbitrary, so as
to ignore the difficulty of meaning? What would it mean
to say that such conventions are arbitrary? It cannot
mean that for one word another might be substituted, for
34
who .would disagree? Surely the linguistic formalist
means to assert more than that, and upon the basis of
that assertion to deny something of opposing philosophical
views of necessity. But "conventional" cannot mean that
for one judgment another might be substituted, for here
again the linguistic conventionalist wishes to make a
truth claim for his view which he could not do if, for
his judgment, another might be substituted. But it can
not be that for the reality judged of, another reality
would do equally well, for it is reality which makes the
judgment true, if it is true. And so, it is hard to tell
just what "conventional" in the sense of "arbitrary" is
to mean, for linguistic formalism.
(c) Are necessary connections merely analytic?
The question is of paramount importance because, if a
necessary statement is but an "unpacking" in the pre
dicate of the meaning "contained in" the subject, then
indeed philosophy is but a "descriptive" exercise in that
trivial sense mentioned in Chapter III. If "synthetic
apriori knowledge" is a delusion, then there is no
genuine entailment at all. The predicate merely repeats
222
in whole or in part what the subject term has already
asserted. To take such a view is to regard necessity
as "analytic", but it has difficulties.
Although "red is a color" is true by definition,
yet to call their connection analytic begs the question.
The question is, "In thinking from 'red1 to 'color', have
we added anything?" We have placed "red" in the system
of reality, among the order of visual sensations, in an
identity which includes yellow and blue, related also to,
but distinguished from, sounds, tastes, and textures.
Mere "red" does not "contain" that true and useful
information, and yet it is added by reason alone.
It is said that a proposition is analytic when
its denial is contradictory. But to answer the question
whether or not denial of the predicate contradicts the
subject, involves the meaning of the subject term. Thus,
to call some proposition "analytic" requires that first
we settle the prior question of which are the essential
versus accidental characteristics. Is it essential to
gold that it be a yellow metal? Then it is necessary,
in the sense of tautologous, that gold is a yellow metal.
To deny it is contradictory. But it would not be so to a
chemist. If, by various tests, he finds something with a
certain atomic weight and number, placing it in his
Periodic Table, then it is gold, whether yellow or not.
Is a sergeant a soldier with three stripes on his coat-
223
sleeve? Would it be contradictory to call him "sergeant"
with his coat off? It seems absurd, but it illustrates
the need to fix a non-arbitrary definition: a difficult
accomplishment if necessity does not go beyond linguistic
convention.
What does it mean to say that an analytic, or
necessary, predicate is contained in the subject? It is
essential to "red" that it exclude "blue ," so "not-blue"
is contained in the subject. So are "not-green", "not-
loud1 ' , . " "not-sharp",;" and "not-prickly v" How about any
thing else whatsoever that we can think of, such as "not-
a-pumpkin"? Must "red" not deny that too? Could red be a
pumpkin? If not, if the meaning of "red1," whether fixed
by convention or not, excludes anything else in the
universe, then the whole universe is contained in, or
analytic to, "red". It is a tautology. But here words
seem a little strained.
Is the linguistic formalist's assertion, "All
necessary truth is analytic," analytic, or synthetic?
If philosophers are concerned mainly with forms of words,
then this is a legitimate question. If the assertion is
synthetic, then it is empirical.. If it is empirical,
then it could admit of exceptions. But it is not
intended to admit of exceptions. Then it must be
analytic. But if so, it says nothing true or false, for
it says only that linguistic formalists have agreed to
224
mean "analytic" when they say "necessary truth.'1 " They
thus have no grounds for denying alternate views of
necessity.
The logic favored by linguistic formalists, which
reads implication in extension only, limits the meaning
of "implies" to the merely contingent truth of certain
combinations of propositions. P implies Q in all cases
except where P is true and Q is false. Thus, if P says
"the sea is salty" and Q says "flowers bloom in Mayi" , 1 P
will imply Q. If P said that the sea were not salty, or
were sweet, or hot, or pink lemonade, it would still imply
that flowers bloom in May. It would also imply that they
do not bloom in May, or that they grew with their roots
in the air. But surely this is a grotesque result for a
doctrine which began by founding philosophy on actual
experience, as an antidote to the excesses of meta-
. . 35
physicians.
G. The Relevance of Alternate Views of
Necessity to Moore's "Common Sense View
of the World"
What has our discussion of psychological and
linguistic objections to necessity got to do with Moore?
That cautious and sensible thinker only had said that there
is such a thing as a mere-matter-of-fact. He did not try
to reduce logic to psychology. He did not say that
necessity holds only within some verbal convention. He
225
did not deny that some features of the world are
necessarily connected, but only held that some of them
are not.
But the preceding section is relevant to Moore
because he wrote a passage that was to change the course
of philosophy. He wrote,
"The earth has existed for many years past" is
the very type of an unambiguous expression, the
meaning of which we all understand. Anyone who
takes a contrary view must, I suppose, be confusing
the question whether we understand its meaning
(which we all certainly do) with the entirely
different question whether we know what it means,
in the sense that we are able to give a correct
analysis of its meaning.36
This passage was most interesting, for philosophers took
Moore to be saying that (a) there are discrete propo
sitions incorrigibly known? (b) the meaning of some
propositions is beyond the reach of philosophical
discussion, (c) philosophical discussion consists of
analyzing meaning "correctly," i.e.,so as not to alter
the Common Sense meaning. Meaning and truth, in other
words, lie outside philosophy's range, and all that it
can do is 'describe'.
Now if (a) is considered along with Moore's
statement that there is "no good reason to suppose"
37
physical facts not to be independent of mental events,
then one has the problem of relating mental events and
physical facts. And further, Moore did believe in sensory
atomism? for while he believed that "I see a hand" is true
beyond the reach of philosophical argument, yet what I
see are sense-data. Thus, if Moore held that there were
sensory atoms, and if there is a problem with connecting
physical facts and mental events, then a view of logic is
called for that does not ignore these results.®® This
Russell tried to give, using the notion of merely
"material" implication.
Further, if one takes [from (b) and (c) above]
the Common Sense meaning to be absolute and supposes
that philosophical inquiry merely analyses the meaning
without distortion, then philosophy does indeed seem
launched upon a merely linguistic course. For this
reason Moore has been credited with paternity of the
"linguistic method" by such diverse thinkers as Blanshard,
Malcolm, and Bergmann. Moore may have rejected such
responsibility, but how else are we to take his dictum
that we ought not to confuse analyzing the meaning of a
proposition "with the entirely different question" of its
39
meaning and truth? How are we to grapple with reality
except through meaningful terms? And if that meaning
fulness is as taken in the first instance— a sense which
establishes it for all time— then what is left to
philosophy but to 'unpack' the meaning 'contained in' the
Common Sense proposition? If the matter for philoso
phical analysis is Common.Sense truth, with which
philosophers may not tamper with respect to that meaning
227
and truth, and if, in taking Moore's "Common Sense view
of the world," such propositions are paradigmatic of
meaning and truth, then what is left for philosophy?
We have left a job of analysis which is irrelevant to
truth.
Because philosophical necessity must fall else
where than in the truth of things, it must fall within
language: within the conventions of our expressions of
the world. That was the reasoning followed by linguistic
formalists. Was it wrong? Is that not what Moore
intended? Were Moore's followers, men like Malcolm and
Bergmann, unable, after all, to see what Moore was
driving at? Then perhaps there wasn't much to it. But
how else are we to take seriously the passage, quoted
above, than as an admonition to devise a theory of
'necessity' which does not perpetuate the ancient
and traditional confusion?
If Moore was right, that the most influential
minds of the Western world from Plato to MacTaggart went
wrong simply by confusing the knowledge of truth with the
correct analysis of meaning, then surely some new theory
is wanted that will explain to us what philosophy ought
to be. If we are asked to believe that venerated
authorities could have deflected the whole course of
intellectual history by making a mistake about meaning
with respect to truth— a mistake which would be simply
228
ludicrous were it not so productive of tragic con
sequences— then surely a new philosophy is wanted, to
rescue human thought from its delusion. Sensory atomism,
linguistic formalism, and their ’logic1 of merely
material implication, sought to fill that want witln .a
view of necessary connection which was true to Moore's
prescription. But psychological atomism and linguistic
formalism have not withstood criticism. We must conclude
that Moore's words, in the passage quoted above, were not
meant to suggest any philosophical principle.
H. in Defense of universal Necessity
Moore1s words were not to be used as a
prescription for a new view of philosophical necessity.
He himself refused to reduce entailment to merely material
implication, except to the extent needed to make room
for the mere-matter-of-fact. With that restriction the
dogma of internal relations could be refuted. With even
one mere-matter-of-fact the dogma could be refuted. For
the defender of the dogma, the task remaining is to close
that last gap.
So far the situation is this: Three views of
philosophical necessity were under consideration:
(1) Necessity is psychological in origin, (2) necessity
is a linguistic convention, and (3) necessity actually
40
holds among matters-of-fact. (1) and (2) have both
failed. The necessity among our ideas, or propositions,
229
also holds among matters-of-fact. There is some
"isomorphism How much? Moore would have agreed that
there is some. Red necessarily is a color and the
diagonals of a square are necessarily equal. But not
every matter-of-fact is necessarily connected with
each and all others, Moore thought. His book need not be
blue to be on his shelf. But the dogma disagrees. That
is the issue.
Blanshard has helped us carry the argument for
universal necessity quite a distance by refuting (1) and
(2), above. He increased the probability of universal
(3) by pointing out that the progress of science tends to
confirm it. There is increasing empirical confirmation
of universal necessity. That does not prove our case,
but, on the other hand, what has been done to prove the
mere matter-of-fact? How could one prove a mere matter-
of-fact? One need only point out some matter-of-fact,
about which the whole actual situation is fully known,
of which there is no reason, or cause, or explanation, of
why it should be. It is a mere happenstance. Such a
state of affairs would be repugnant to our common sense,
but so much the worse for it. We would be inclined to
say that a book was blue because the printer had bound
it so, and that it was on this shelf because someone had
put it there, but if such information were not forth
coming although the full situation were fully known, then
230
we would have to admit that here, indeed, was a mere
matter-of-fact.
But the burden of proof has been shifted to us.
And to bear that burden what must we do? We must give,
apparently, an empirical proof that every matter-of-fact
is necessarily connected to every other; and until that
proof is given, the dogma is— what?— refuted? Why? It
seems unreasonable. It seems unreasonable because it is
empirically impossible. We cannot explain every matter-
of-fact. But surely an empirical proof must be
empirically possible. All that we can possibly offer is
empirical confirmation. And that, as we have said, is
all on the side of necessity.
The physical evidence is on the side of
necessity. So is the psychological evidence. As Hume
himself pointed out, we cannot help believing that the
relation between being in the fire and being burned are
necessarily connected. There are some cases where
"chance" seems to reign, such as when we roll dice. But
even here it is only the number of spots turned up that
is "chancy," That the dice when thrown should not roll,
or should change shape, or lose their spots— these
'possibilities' are never considered. And even with the
number of spots turned up, we do not suppose that there
is no reason for it. It is just that we don't know it,
or can't control it.
231
But must we prove that necessity is universal,
if we can prove, instead, that it is universally pre
sumed? Let us try to prove this latter by what we may
call "the argument from reflexive implication." Now we
know that good philosophy rests upon solid argument.
Our thought passes smoothly over it, there are no gaps
in it, and every idea, or proposition, follows from the
rest. Is this not so? It seems so, when we review the
criticisms which philosophers aim at other philosophers.
What are they? First of all, non sequiturs of any kind:
Bradley's view, said Moore, didn't follow. But he was
also ambiguous, contradicted himself, was inconsistent,
he misclassified, he failed to include all the 'facts'
within his system. These are typical "errors," and they
are sufficient to refute the poor philosopher in whose
work they occur. Why are they errors? Why this predi
lection for consistency, and an inclusion of all the
facts? Because, to philosophers as to men of common
sense, reality is truth and what is not the one is not
the other. And reality i£ a consistent system which
includes all the facts. Then know them by their works:
by "reflexive implication","
I. A Logical Difficulty with the
Mere-Matter-of-Fact
Someone may object that we have not proved
232
universal necessity, in the passages above. The argument
is ad hominem— in the sense that while we may have proved
something about what all philosophers believe some of the
time, yet we have not proved that their belief is true.
Blanshard himself did not think it possible to prove
universal necessity. He said,
Though one cannot without some absurdity say that
anyone has ever proved the world to be a necessary
system, one can still postulate it and then
examine how far the actual exercise of reason goes
toward justifying the postulate.^1
If this be the case, then Moore has not refuted the
dogma, for we are still free to postulate the universally
necessary truth which requires that his book necessarily
be blue to be on the shelf. But on the other hand,
neither have the heirs to Moore's viewpoint been silenced.
They are still free to postulate their mere-matter-of-
fact. They would still reply, to our arguments above
that the advance of science suggests universal necessity,
that our case is far from proved. While they would admit
that our ignorance does not warrant a denial of universal
necessity, yet they would point out that our ignorance is
vast beyond comprehension. However much we may flatter
ourselves in our scientific achievements, yet, compared to
what is not found out, it is but a desultory raid upon the
darkness. Thus, Blanshard has not silenced advocates
of the mere-matter-of-fact. We need a logical argument
to settle the issue.
But surely no one would say that the advance of
science does nothing to increase the plausibility of
universal necessity, and surely is a weakness on the
other side if no verifiably mere matter-of-fact has been
produced. It, empirically, the evidence favors necessity;
and if, psychologically, it is not possible to keep from
believing in it at least sometimes; and if, further, the
argument for necessity has advanced to the point where
we can now postulate universal necessity— then a logical
argument for it seems to offer itself.
The possibilities have been narrowed to this
alternative: either universal necessity is true, or we
may postulate its truth. Let us now ask, "What is it to
postulate some truth?" What do we mean to say when we
say that a truth may be postulated? "Postulate" as a
verb means to stipulate, to lay down as an assumption,
to assume without proof. Very well. But now what are we
to postulate? Universal necessity, i.e. that everything
follows from everything else; that everything that in
truth is, could not possibly be otherwise. And we are
to postulate this. We are to assume without proof that
nothing could possibly be otherwise than it is. But to
assume it, short of proof— is that not to assume that it
could possibly be otherwise? But how can we assume that
nothing could possibly be otherwise while qualifying that
234
assumption by saying that it could possibly be other
wise? In a merely postulated universal necessity we are
trying to force together, into one meaning, two mutually
repugnant meanings. Now a logical proof must come from
the meanings of the terms alone. Here are two meanings
that contradict. So here, we submit, is a logical proof
that, whatever else we may postulate, we cannot
postulate, without some absurdity, universal necessity.
And yet, while we agree that what is needed
against the mere-matter-of-fact is a logical proof,
someone may still be dissatisfied with any mere mani
pulation of abstractions. What of Moore's blue book
which, he had plausibly argued, need not be blue to be
on his shelf? Let us make one last effort to get down
to the 'fact'.
Let us try to dispose of this mere-matter-of-
fact in a way which will rid us of it, and all others.
Moore has said,
It seems to me so plainly true that, in the caee
of every sense-datum I have, it is logically
possible that the sense-datum in question should
not have existed--that there should simply have
been no such t h i n g .42
For Moore it was significant to say that this book here
now, although blue, could possibly have been not-blue.
This we must deny. It could not possibly be not-blue.
It is not even significant to use "possibly" in this
sense. Now we are not speaking of some imaginary book
235
which we imagine in the place of this blue one here,
actually in place on the shelf. We mean this book. We
are saying that it makes no sense to say that this book
could possibly be not-blue. "Possibly” means "and
perhaps also otherwise". But it is known not to be
also perhaps otherwise for the best of reasons: it is
known to be blue.
"Aha, but now we have youl" says our critic.
"You have made the very^blunder that Moore warned
against, and which destroys your so-called 'proof'.
Moore, in speaking of possibly otherwise, used it in
conjunction with 'should have been', or 'could be' or
'would necessarily be'— but you are treating his pro
positions as though he were saying 'is', or 'is
necessarily', or 'must be'. Surely you do not mean to
say that this book, to be on this shelf, would
necessarily be blue!"
We do indeed. "Could be1 '," "should have been, ' !
or "would necessarily be" are the conditional tense of
"is;" or "is necessarilyit helps his argument not
at all to phrase his propositions in the conditional
rather than in the indicative form. "Would necessarily
be" invites us to consider the conditions under which we
have any right to say that this blue book is on this
shelf here now. We are invited to explore its history,
from its coming into being at the printer's, through the
236
story of how we came to purchase and use it, up to our
placing of it here upon our shelf. Nowhere, in that whole
narrative, is there anything 'mere' or 'happenstancy1.
It is all quite well in order and satisfactorily
explained. No meaningful question arises as to why it is
this rather than that, or here rather than there. To be
this book, satisfying these conditions, as it does, it
would necessarily be this blue one, here on the shelf.
J. Summary of Chapter IV
The doctrine of internal relations states that
"A term is internally related to another term if the
absence of that relation would entail difference in the
term". Moore rejected this "dogma/" for it means that
relation changes the term. His reason for that rejection
is doubtful, and his rephrasing of the proposition en
tailed by the dogma, to include such a dubious entity as
a "relational property'/" serves only to obscure the
question while contributing nothing to its solution.
Moore rejected the universal internality of
relation because, although there may be relatedness, yet
it was a "howler" to suppose that in some cases the terms
would necessarily be related as they are. Such
relatedness was "a mere-matter-of-factV"
But the dogma is not thus refuted unless we
suppose that necessity is not universal. This Moore did
237
suppose by calling the matter-of-fact "mere." There are
several arguments which make Moore's position on
necessity doubtful, however.
(1) The alternatives to genuine (philosophical,
logical) necessity are indefensible.
(2) There is physical and psychological evidence
that necessity is universal.
(3) Universal necessity, in philosophy, is
presupposed. It is presupposed in the conventions by
which philosophers define their task. They cannot deny
necessity without absurdity.
Further, the "mere" matter-of-fact leads to
contradiction, for a proposition incorporating necessity
in the sense of "would necessarily be" is a conditional
proposition.^ It is compatible with falsity. It says
no more than "necessary, presuming other conditions are
satisfied," or "necessary, so far as we can tell." If a
statement of actual necessity is intended, it must be
phrased in the indicative form. Merely possible
properties may be predicated only when actual ones are
unknown— but in that case no 'fact' is being asserted.
238
FOOTNOTES--CHAPTER IV
Moore seemed to have accepted this aim of
traditional philosophy with the emphasis he placed on
"whole". See G. E. Moore, Some Main Problems of
Philosophy (New York: The Macmillan Co., 1953;rpt.
Collier Books, 1962), p. 27. Hereafter cited as Main
Problems.
I am talking about "facts" here while Moore
speke about "true propositions". I presume that by
asserting a plurality of true propositions Moore meant to
assert a plurality of facts, and so I am not guilty of
confusing, here, the truth of a proposition with the
logically independent question (according to Moore) of
its correct analysis.
What is a "fact"? See Appendix B. Bradley's
view of "fact" is shown in the following quotations:
"The view which I advocate takes them all as in principle
fallible. On the other hand, that view denies that there
is any necessity for absolute facts of sense. Facts for |
it are true, we may say, just so far as they work, just
so far as they contribute to the order of experience . . .
there is no 'fact' which possesses an absolute right
.../... facts are valid so far as, when taken
otherwise than as 'real', they bring disorder into my
world. . . . 'then no judgment of perception will be more
than probable?' Certainly that is my contention [my
emphasis]." See F. H. Bradley, Essays on Truth and
Reality (London: Oxford U. Pr., 1914), pp. 210-11.
Hereafter cited as ETR.
On Bradley's view it is a mistake to suppose that
reality consists of unique 'facts'. These are
erroneously supposed to be upheld by an appeal to
'denotation' and thus rest upon the "this," "now," and
"here" of personal feeling— from which we then transfer
those contents bodily to a privileged region in thought
where they provide a "test of absolute truth and sheer
error." Ibid., p. 261.
"Matter of fact" is an ambiguous phrase in which
Bradley distinguished three senses:
(1) The merely felt, exclusive of judgment, as in
imagination, memory, and observation on the level below
239
j udgment.
(2) The above, taken in analysis, abstractly,
with connections regarded as external or "devoid of
intrinsic reason" (that is, as unconditional because
conditions are taken to be irrelevant).
(3) The above, taken as "belonging to and as
dependent on a point in our 'real' series," i.e. as
conditional with conditions known and satisfied. Ibid.,
pp. 378-79.
3
G. E. Moore, Philosophical Studies, "External
and Internal Relations" (London: Routledge and Kegan
Paul, 1922? rpt. Totowa, N. J., Littlefield, Adams & Co.,
1968), p. 286. Hereafter cited as PS, "Relations."
4Ibid., pp. 276-309.
^Moore has already given his case away. He has
admitted (a) that the relation must relate, and (b) that
no relational fact is reducible to its constituents.
Had he begun, as we did in this essay, by defining the
relational view strictly, he would have seen that to give
away (a) and (b) leaves nothing more for "external
relations" to mean.
®Moore did not bother to tell us where Bradley
said explicitly that relations "modify" their terms, nor
have have we found the place. Bradley did say that
relations "make a difference to" the terms, but this is
hardly metaphorical language.
7Ibid., p. 279.
®Ibid., p. 281.
Q
Note that in addition to introducing a
"relational property" Moore has also injected
conditionality.
■^To Bradley this mere tautology asserts nothing.
See his The Principles of Logic, 2nd ed., 2 vols.
(London, Oxford. U. Pr., 1928), 2:141-42. Hereafter
cited as Logic.
Moore uses "follows" in a sense which is in
significant. No "also" is implied. A significant
assertion, in geometry, would be, "If A is a right angle,
it follows that A could be inscribed in a semicircle."
It means, in the end, to assert the geometrical system
of reality. Significant assertion is not the unpacking of
a tautology, nor even the predicating of one idea of
240
another. See Appendix A, "What Is An Idea?" See the
opening discussion of Chapter II. See Bradley, Ibid.,
p. 10.
^Moore, PS, "Relations", pp. 188-89.
12Ibid., p. 288.
1 1
Maurice Merleau-Ponty, Phenomenology of
Perception, trans. Colin Smith {London: Routledge and
Paul, Ltd., 1962), p. 323.
■^Moore, PS, "Relations", pp. 288-89.
15
* « was Moore's way of writing what most
logicians would write " (P . @)", or "(P Q)"
" " means "materially implies."
■^Moore, PS, "Relations," p. 295.
Moore, ibid., pp. 295-301.
18Ibid., p. 304.
I Q
Thus, the "would necessarily be" of (2), the
conditional tense of "is necessarily;" requires that
anyone believing in the dogma should have been able to
deduce the spatial location of certain persons from their
age, with the surety of a geometrical demonstration.
What would Bradley have said to this?
Bradley would have replied as follows: "My true
views had been available to Moore for thirty years, and
he must have known that I made no such extravagant claim.
The issue, I suppose, is this: If two abstract
propositions, p and q, are not false together, then can
we go on to say that p implies q? Certainly not. No
bare proposition implies, or entails, or follows from,
another. In that impoverished sense (p . q) does not
mean psq. But a proposition uses signs; a sign is a sign
of something: of reality. That is why I have always
Insisted that reality isthe true subject of our judg
ments. An implicit "reality is such that" must preface
our propositions. And if reality is such that p and q
are not false together, then might"there not be, in
reality, a reason for it? Doesn't it agree better with
experience, and make sense out of what is otherwise
senseless, to suppose some connection? And on the other
hand, is there any reason not to suppose such necessary
connection?
Now, there are mediating conditions between bare
Kegan
later
where
241
p and q, as Moore implied by using "would necessarily
be”) ' instead of "is necessarily," in his grotesque
formulation of my view. He did not state those
conditions— he did not get them into his formulation—
he merely acknowledged that they applied. But with
them, to get the whole meaning of our p and q, it is not
absurd to go on and say, p implies q."
20
See ETR, p. 226. But then, did Bradley mean
that no empirical judgment can be known for certain?
This would seem to lead to absurdity (See Norman Malcolm,
Knowledge and Certainty [Englewood Cliffs, N. J.:
Prentice-Hall Inc., 1965], pp. 58-72). Does this mean
that no one could ever be certain that on page 211 of
Essays on Truth and Reality Bradley wrote, "1 Then no
judgment of perception will.be more than probable?'
Certainly that is my contention?"
Bradley was certain that no judgment of
perception can be known for certain? No, he might be
certain of it and yet not know it to be more than
probably. Bradley distinguished between the psychic
content of a judgment (of which one could be psycholo
gically certain) and the logical connections one means
to assert in calling the judgment true. Those con
nections, the conditions of truth, are required to be
satisfied but are not presented wholly satisfied, and so
involve interpretation, the work of the mind, universality
of meaning which transcends presented content, "the ideal
transcendence of the given finite."
That which verifies the judgment, whether
perceptual or not, is never at one moment wholly before
our attention. It depends for its satisfaction upon the
veracity of memory, which is beset by the contradiction
of taking the past for the present. If somehow all the
evidence for the truth of any judgment were to be veri
fied, thfet plurality of evidences must be verified
serially and singly, with memory employed to conserve
the progressive result. This is a heavy burden to place
upon memory.
Even though, as Malcolm insists, it is stupid
practically to repetitively and endlessly verify the
same bit of evidence, yet the variety of evidence is
endless, as evidence for the evidence; and by what
theoretical principle may we cut the effott short? For
every 'verification' of every bit of evidence involves
making another judgment, which is in principle as
questionable as the first.
Psychological certainty, although attainable, is
momentary and transitory. It is possible to forget what
242
is the subject of a sentence while attending to the
i predicate. It is possible to 'lose' your glasses be
cause you are not attending to what is resting on your
nose. And, really now, is it true that Bradley wrote
such-and-such on page so-and-so of some book or other?
Better find the page and look again. But, now, is this
the book? Better look on the title page, and see. But,
now, did he write "judgment;!' or was it "assertion"?
Did he write "likely" where, we remember "probable"? Turn
to page so-and-so, and read it again. But, now, did we
suffer psychic blindness when checking that title?
Better recheck. And so on.
If someone wishes to call such momentary results
as this "knowledge,7 which now is certain and now is
not— which Bradley would have called "mere appearance"—
then that is his affair. But if, on the other hand, one
wishes to assert as veridical some existent as it
qualifies and is qualified by reality, of which the truth
is nothing less, then one is involved with the problem
which is the subject of this entire essay.
2 "^Bradley, ETR, p. 312.
22
". . . to our ignorance." in the sense that we
justify denying that a relation is in truth necessary
merely because our ignorance of the full situation does
not permit us to affirm its necessity. In the logic
textbooks this is called the ad ignorantiam fallacy.
(See W. H. Werkmeister, An introduction to Critical
Thinking, rev. ed. [Lincoln, Neb.: Johnsen Pub. Co.,
1957] , p. 56).
22The discussion of objections to universal
necessity, in this section, owes much to Blanshard's
Reason and Analysis (La Salle, 111.: Open Court Publ Co.,
1961), Chapters VII-XI. Hereafter cited as R&A.
24Ibid., p. 383.
25Ibid., p. 449.
26
Even though it were true that some sensory
occasion were the necessary condition without which no
idea could come to consciousness, yet we still have a
problem about the Given. In Appearance and Reality
(2nd ed. [Oxford: Clarendon Pr., 1897, rpt. 1966)) Brad
ley wrote: "Nothing can be taken as real because, for
psychology, it is original; or, again, as unreal
because it is secondary [p. 30]." Also, "There can be no
origin, nor want of origin, that warrants our predicating
nonsense [p. 52]," and "Any deliverance of consciousness—
243
whether original or acquired— is but a deliverance of
consciousness. It is in no case an oracle and a
revelation which we have to accept. . . . The 'given' of
course is given? it must be recognized, and it cannot
be ignored. But between recognizing a datum and
receiving blindly its content as reality is a very wide
interval [p. 182]."
With respect to Hume and the privileged sense-
impression, the re-levance of Bradley's remarks is
this: While the universal meaning of a term is vital
to the truth of the judgment in which some idea appears
as a term, yet the mere occasion of its appearing has
nothing to do with its meaning and truth. Subsequence
is not consequence. If one awakes to a belief subse
quent to a blow on the head, is the blow the logical
antecedent of the belief? Blows on the head are all
alike, but why is an idea believed? There is no cog
nition determined by blows on the head, or sensory
occasions of any other sort, unless the blows or
occasions are of assorted kinds. But with "kinds" we
are involved with universal meanings, among which only
logical necessity is the movxng force.
Even if some form of the Sensory Impact Theory
(or any psychological theory purporting to derive
cognition from non-cognitive afferent stimuli) were
true (in the sense that cognition may arise only with
N' neural discharges), is it defensible to think of
cognition as only their decaying echo reverberating in
the brain? No, for in order to get cognition— in order
for a sign to be detected— neural discharge must make a
pattern. And a pattern is a many-in-one; an identity
with internal structure knowable only in reflection.
It is not a naked particular but a universal, and so,
within the domain of logic.
27
] C cannot imagine one ball striking another
without rebound, unless, by an effort, I suppress the
imaginary rebound. But the question here is whether we
have the right to identify what we may imagine with the
natural order.
28Ibid., p. 454.
29See Ibid., p. 444-61. So Blanshard thought,
but can we say that Bradley would agree? To be logical
throughout, must not the world consist of terms-in-
relation throughout? What, then, of infra-relational
Feeling, and suprarelational Reality?
Bradley would reply that so far as Feeling and
244
Reality are merely thought, they do not conflict with
thought's own nature^ forthey are correctly thought of
as the ground and consumation demanded by thought, to be
as it is. It is no contradiction to say that the nature
of thought's demand requires grounding, and also
transcendence for completion. Nor is such transcendence
unthinkable. We experience consummation in trans
cendence with every realized desire: the object, pre
viously ideal, is secured, and with that the desire is
obliterated. So it is for thought, in its pursuit of
truth and reality. There is thus no contradiction, but
"only an inexplicable'.1"
To confront Bradley with a contradiction in his
conception of Feeling and Reality, one must first render
Feeling and Reality into terms-in-relation, and then
show that the meanings conflict with some other part of
his system. But Feeling and Reality, thought as
Bradley thought them, do not conflict in their meanings
with anything else in thought; their own meanings are
required if the contradiction of bare terms-in-relation
is to be overcome. Feeling and Reality ground and
complete, something that is required by thought. In
themselves unthinkable, being beyond thought, they are
not as such items of conceptual furniture. They are not
internal distinctions within thought and thus do not
conflict as ideas.
3®See this essay, Ch. I. Sec. 6.
31David Hume, Hume Selections, ed. C. W. Hempel
(New York: Charles Scribner's Sons, 1927; rpt. 1955), p.
120. "The mind can never possibly find the effect in the
supposed cause, by the most accurate scrutiny and
observation."
32David Hume, An Inquiry Concerning Human
Understanding (New York: Library of Liberal Arts, 1955),
p. 26. "These faculties (memory and imagination] may
mimic or copy the perceptions of the senses, but they
never can entirely reach the force and vivacity of the
original sentiment."
^Ibid., p. 60.
34Wittgenstein's Philosophical Investigations
was puhlished in English and German, on facing pages,
so that his meaning would be clear.
■^See Blanshard, R&A, pp. 255-307.
3®Moore, "Defencep. 57.
245
37Ibid., p. 45.
38
There seems to have been a conflict in Moore's
thought about implication. We have seen that "entails"
did not mean "not P true and Q false:" He rejected
(P'» q) as the meaning of po q because "this is not what
'entails" ordinarily means."
Yet, based upon the same Common Sense, he
defined "reality," or tried to define it, while
restricting its meaning to the range of particular
meanings it has in common speech. The universal,
"reality;" was thus dependent for its meaning upon the
particular. The principle thus adopted would lead him
to logical atomism. And from logical atomism it is
impossible to argue for any necessary connection, except
through tautology, if we are to believe that avowed
logical atomists did their job responsibly.
On the ambiguity of Moore's Common Sense view of
the world, see Blanshard, R&A, pp. 315-16.
39
We may assume that Bradley was Moore's bette
noire. Then why did Moore think that Bradley had denied
Common Sense propositions because analysis of them
failed? Bradley did not confuse what C.S. propositions
mean with an analysis of them. He took them at face
value; he denied what they claim. They claim to be
categorical truth (Common Sense sometimes takes them
so), and Bradley denied it. He did not first reduce
those C.S. propositions to their constituents.. In what
sense is a correct theory of meaning crucial to Bradley's
denial? In what sense had he confused meaning with
analysis of meaning? It was because the C.S. proposition
had meaning that Bradley denied it unconditional truth.
40
The discussion to follow does not distinguish
between causal and logical necessity, for it is a
distinction without a difference. See Appendix G, "On
the Supposed Difference Between Logical and Causal
Sequence." See also Appendix A.
^^Blanshard, R&A, p. 383.
4 2
G. E. Moore, Philosophical Papers (New York:
Macmillan Co., 1959; rpt. Collier Books, 1962), p. 125.
^The possibility that Moore meant "true now and
always irrespective of conditions" by his "would
necessarily be" has not been considered. It has not,
because it was presumed that he meant to refute- Bradley,
whom he mentioned by name in his opening sentence. But
246
Bradley never suggested that any finite subject,
signified by any term, has any predicate (relational or
not) irrespective of conditions. And to impute to him
the notion thfat two abstract ideas, or propositions, are
necessarily connected bare and in themselves, would be
monstrous.
247
CHAPTER V
THREE LATER CRITICS OF BRADLEY
The preceding chapter was devoted to Moore's
logical refutation of Bradley's view of relations.
Others have voiced logical objections; they have dharged
that Bradley failed to apprehend correctly the role of
logic in philosophy, or failed to understand the role of
the relation in that logic— failed even, to understand
what a relation actually was. And so, the inquiry into
logical difficulties besetting Bradley's view will be
1
continued here. Russell, McTaggart, and Bergmann have
each expressed similar objections. We begin with Rus
sell, who followed Moore in time as well as doctrine, and
was second only to him in the effect of his criticism
2
upon the popularity of Bradley's philosophy.
A. The Attack from Asymmetrical Relations
"Mrw Bradley," said Russell, "has worked out a
theory according to which, in all judgment, we are
ascribing a predicate to Reality as a whole; and this
3
theory is derived from Hegel." Russell was referring
to Bradley's belief that the form for all judgment was
"Reality is such that S is P." That belief, Russell
held, was a conclusion which was unwittingly thrust upon
248
Bradley from an unquestioned (but questionable) assumption
made in his logic. The assumption was, said Russell,
That every proposition ascribes a predicate to a
subject, and from this it easily follows that there
can be only one subject, the Absolute, for if there
were two, the proposition that there were two would
not ascribe a predicate to either.^
Russell was saying that Bradley, along with his predeces
sors and admirers, had been misled by their faulty logic
into believing as they did. The suggestion struck the
philosophical world as ingenious. Philosophers who
followed Russell could now see the reason why Bradley
believed that unitary Reality was the transcendent subject
of every judgment: Bradley was the dupe of an outmoded
subject-predicate logic.
The influence of the old logic used by Bradley is
pernicious, Russell held. Arguments which seem to
establish it are only the distorted result of having pre
supposed it. It was not true that, in logic, a predicate
is always being attached to some subject, but
This belief, being traditional, scarcely self-
conscious, and not supposed to be important,
operates underground, and is assumed in arguments
which, like the refutation of relations, appears to
establish its truth. This is the most important
respect in which Hegel uncritically assumes the
traditional logic [my emphasis].
There was now an alternative, Russell said, in
"logistical or mathematical logic," which was "the ful
fillment of a hope which Leibniz cherished," but which he
did not realize because "respect for Aristotle prevented
249
Leibniz from realizing that this was possible."^
But now the hope for an universal mathematics of
propositions was to be realized. Mathematical logic turns
the philosopher to the use of "the principle of
abstraction," said Russell, which is the exercise of
discerning among some group of things a kind of
similarity which may be called "form’,," Logic is con
cerned with this form. The discerning of such forms
enables philosophers to see that the subject-predicate
proposition is not the only kind of proposition. In
fact, it is not the simplest nor is it even very common.
Russell pointed out the kind of proposition which was
simplest and most common: the proposition stating a
relation among things. It has a very different form from
the subject-predicate proposition.
Why did philosophers like Hegel and Bradley not
notice the pernicious influence of their outmoded logic?
Russell had the answer:
If they had been honestly anxious . . . [to give
an account of the world and daily life] . . .
they would probably have discovered their error
very quickly; but most of them were less anxious
to understand the world . . . than to convict it
of unreality in the interests of a super-sensible
"real" world. Belief in the unreality of the
world of sense arises with an irresistible force
in certain minds . . . The conviction born of
those moods is the source of most mysticism and
of most metaphysics. When the emotional intensity
of such a mood subsides, a man who is in the habit
of reasoning will search for logical reasons in
favour of the belief which he finds in himself.
250
And lower down:
They remained— to borrow a useful word from
Mr. Santayana— "malicious" in regard to the world
of science and common sense.
Russell then proceeded to show that Bradley's logic, "The
logic of mysticism, . . . shows, as is natural, the
defects which are inherent in anything malicious."7
Bradley's "malicious" logic had led him to assume
that relations must be reduced to properties of the re
lated terms. It was "unable to admit the reality of
relations." While there were many ways to refute the
belief that relations are properties of the related terms,
said Russell, he chose the "easiest way": that of proving
the "reality" of "asymmetrical" relations. His proof went
as follows: Relations may be symmetrical, holding
between B and A just as they hold between A and B, as in
"similar"; or they may be non-symmetrical, where they do
not necessarily hold between B and A as they do between A
and B, as in "brother of." But relations may also be
asymmetrical, where the relation never holds between B and
A the same as between A and B. "Father of" is an example
O
of such a relation.
Russell agreed that it was possible to regard
symmetrical relations as predicates of the terms related,
for symmetrical relations are transitive: they hold
between B and A just as they hold between A and B. The
order is unimportant. Thus, if A is different from B, B
251
is no less different from A: the relation "different
from" is symmetrical and transitive. Thus "different I
from" can be thought of as a property of A and B. But
the case is not so for a symmetrica1 relations. They are
intransitive. If A is greater than Bf B cannot be greater j
than A. "Greater than" is not a property of A and B. It
is a different 'property' for A than for B. The order is
important. "Greater than" is that which asserts the order
between A and B. Suppose that "greater than" refers to
the sizes of A and B. They are of course different sizes,
so the "different from," referring to sizes, being a
transitive relation, can be a property of A and B. But
the relation between these sizes, which "greater than"
asserts, cannot be such a property. "Different from" can
be possessed by, or assimilated into the meaning of, A,
or B. Each can have it. But "greater than" cannot be so
assimilated. Its meaning compels it to stand apart from
them in order to express what it does express. It relates
them; it is not a property.
The asymmetrical relation is a very important
one. In every series, spatial, temporal, or expressive
of magnitude or degree of subordination, it is involved.
Thus without it we cannot refer to the actual world. But
if our logic cannot deal with it, without destroying what
it means, then we are prevented by our logic from even
252
dealing with the actual world at all. Our logic compels
us to condemn the actual world. Russell said,
To those whose logic is not malicious, such a
wholesale condemnation appears impossible. And
in fact there is no reason except prejudice, so
far as I can discover, for denying the reality of
relations. When once their reality is admitted,
all logical grounds for supposing the world of
sense to be illusory disappear.
The logic used by Bradley distorted reality; it could not
even express one of its most fundamental forms: the
asymmetrical relation. It could not even deal with
relations, except as properties. But there is no reason
for not recognizing logical relations except prejudice.
The old logic produced a false metaphysics.
According to it every proposition predicates a property of
some subject. Every relation is treated as a property, to
be assigned to a subject. But if at first we try to refer
to a plurality of many subjects, we cannot claim that there
truly are many, for we must reduce their relations to
properties. Every property belongs to a subject, and it
cannot belong to more than one subject. For if there are
two subjects, we have two related things whose relation
has not been reduced to a property. Our (misconceived)
logic compels us to assign those actual things, in the
external world taken as real subjects, to a supra-sensible
Something-or-other as properties. Our logic forbids real
relations. So long as we continue to maintain that there
is more than one thing in the universe, our logic compels
253
us to assign that plurality to some subject as a
property. It logically follows, therefore, that there is
really only one thing in the universe, as the ultimate and
transcendent subject to which all these properties belong.
This indefensible metaphysical view is called "monism:"
Russell's proof of the reality of asymmetrical relations
shows it to be a delusion.
B. Reply to the Attack from
Asymmetrical Relations
However desirous we may be of realizing Leibniz's
hope for a mathesis universalis, still we must here in
sist that Russell has been hasty in his judgment. He has
not only said that Bradley and his sort were mistaken, but
that they were not even honestly mistaken. They had not
been "honestly anxious" to find out the truth; they were
"malicious" with regard to the truths of science and
ordinary life; they let the "emotional intensity of such
a mood" dictate their logical procedures. These are
grave charges. Could no reply be made? We shall attempt
such a reply in the following sections 1 through 4.
1. Russell's attack was an Ignoratio Elenchi
Russell charged that the reason why Bradley et al.
believed that all judgment predicates of reality was
because their outmoded logic required it. But we have
already discussed the reason why Bradley held that view,
254
in Chapter II, A above. The reason we cited there, for 1
saying that all judgment must take the form "Reality is j
such that S is P," was that whatever truly owns a predi- j
cate must also own the conditions under which subject and ;
predicate may be truly joined. The mere proposition does
not mention those conditions even when that proposition
states a relation, asymmetrical or otherwise. If it
does not own the conditions or predication or relation,
then the truth of the proposition lies outside of it., too.
Contrary to what Russell thought, it does not own its
truth as a property.
In and of itself the proposition, whatever it is,
is compatible with falsity. But we say it is true. Then
the conditions under which it is true are not irrelevant—
either to the proposition's meaning, for they complete
it; or to its truth, for it depends upon them. Yet those
conditions have not been got inside the proposition.
Then it, of itself, is neither meaningful nor true. But
we do not mean to say that the truth and meaning of the
proposition depends upon our whimsy. A proposition, after
all, is the objective part of a judgment. Then something
does own the conditions of meaning and truth, and guarantee
them to our proposition. It supplies the conditions, and
so, owns the proposition as a qualification of itself; or
as predicated of itself, or as a character in its self— or
255
some acceptable manner which conveys the notion that it,
itself, is what this whole affair of judging meaning and
truth of propositions is all about.
If that which guarantees the meaning and truth of
propositions is left unaccounted for but is merely under
stood to be present, then the explicit proposition does
not present the whole truth and is, by itself, not true at
all. The titular subject of the proposition is but a
grammatical subject. By its means we wish to say some
thing which is true. Now we may, with Russell, wish to
say that some relation is true. We do not wish to predi
cate that relation as a property, for we may wish to say
"The cat is on the mat" and not mean that "on the mat" is
a property of "the cat!" We do not mean "S is P ‘ ," for we
mean "a R b>." "On," here, is an asymmetrical relation
and not a property belonging to the cat, or the mat, or
both of them together, for it is this very "together"
which we mean to assert in a synthetic proposition. And
yet the relational proposition itself is not true.
Unless we mean it under certain conditions, then "The cat
is not on the mat" is equally true. Let "X" express
those conditions. Thus, if we wish to say that our
proposition is true, we must say "aRb on condition of X."
Reality supplies the unspecified X. Thus, if we wish to
call our proposition true, we must say, "Reality is such
that aRb." If we wish to call our proposition true, then
256
reality is the genuine subject of which our relational
proposition states a property. This is what Bradley was
driving at.
Such was Bradley's argument purporting to estab
lish his belief that all judgment of truth predicates of
reality. Even if the argument were faulty, Russell's
criticism ought to have been directed to it. It is
difficult to see how Bradley's belief was the result of a
mistaken logic, when, like Russell himself, Bradley was
trying to determine what basic form the logical propo
sition ought to take.
2. Russell Mistook Bradley's View
It is difficult for mere outsiders like ourselves
to be convinced that Russell failed to understand the
view which he criticized when he, like Moore, had been
brought up in it. And yet we must accept that possibi
lity. Russell had charged that the demonstrable reality
of asymmetrical relations provided an exception fatal to
Bradley's view— sometimes called "monism." But, as
Bradley pointed out, the proved reality of any relations
at all would refute his doctrine. He said quite clearly
how his doctrine could be refuted. Before trying to re
fute it, however, a critic ought to determine for himself
just what that objectionable view asserted. A lengthy
quotation of Bradley's own words may help to clarify what
257
his position was, and why he rejected Russell's attack
from asymmetrical relations:
[Russell] assumes that every Monism must take the
universe as throughout a simple unity of qualities
or adjectives in and of a whole which completely
determines them, and is bound (I understand also)
in this way to account for and explain all its
contents— which it obviously cannot do in the case
of 'asymmetrical relations'.
Now nothing, I agree, can be more obvious [than]
that not only some, but any form of relational
experience is, on such a view of Monism, impossible,
while (as a fact) it is undeniable. But for myself
(if I may speak as a Monist) this from the first has
been obvious, and I do not take the ultimate reality
as above, and I had hoped that so much was clear;
• • •
Hence, when it is objected against me as a
Monist that all that I as such have a right to is
the terms and the whole, while the order and di
rection is in neither— my answer is that no whole
is really a simple whole, and in every whole are
always conditions unexpressed, and that in those
conditions falls the difference required here, and
here is the reason why ARB and BRA are incompatible
(that is, when and where they are so).
Bradley had contended (we may remember) that all
relational experience (including that expressed in the
so-called subject-predicate proposition) gives appearance
only, not reality, because the relational view is
contradictory. But reality as a whole need not be a
simple whole— indeed, as we have seen, nothing real is
sipiple. It would be most difficult to find any example
where order and direction were wholly lacking. Reality
appears richly variegated. Parts appear, which do not
seem to be properties of the whole— because the whole is
not before our narrow gaze. Iri reality, where there are
258
finite centers of consciousness, or selves, those selves
can and obviously do take up the relational view and go on
to 'describe1. Some of those selves may take up philo
sophy. They then err by presuming that reality is
throughout as taken in the relational view.
To a finite center of consciousness, with such a
view, order and direction appear among the parts, along
with the parts. The order and direction are then
abstracted, called "relations," some of the relations
called "asymmetrical," and declared "real" because they
are not reducible to the parts as properties of them.
Bradley did not deny that things do appear to us this
way, but his question was, "Have we the right to assume
tllat this is reality?" He thought not because he found
that the assumption led to contradictions. His critique
was of the whole relational viewI It itself is con
ditioned, those conditions are unexpressed and even
ignored, but they are vital to the truth of the judgments
we make. Ignoring those conditions raises contradictions
within the relational view itself, when it takes itself
for reality.
3. Bradley Had Noticed Asymmetrical Relations
Russell charged that Bradley tried to reduce all
relations to properties of terms, having failed to notice
what the case for asymmetrical relations proves: that
259
some relations cannot be so reduced. But Russell was
mistaken. Bradley had noticed it. As early as 1883,
when his Logic first appeared, Bradley was saying such
things as
The ultimate connection of quality and relation is
a most difficult problem. But it is clear that in
their phenomenal appearance the one can not be
reduced to the other [my emphasis].12
In a footnote to later edictions Bradley emphasized the
point which he wished to bring out here: that terms and
relations are both abstractions from one whole and
original situation. That point was sufficiently stressed
and elaborated upon in Chapter I of this essay.
4. What Did Russell Mean by "Patent Fact"?
Russell criticized Bradley for having "convicted
the world and daily life of unreality" instead of "giving
an account of" it. Russell added,
It has become natural to suspect a fallacy in any
deduction of which the conclusion appears to
contradict patent facts [my emphasis].13
Bradley may have had his reasons for rejecting such
"facts," but nevertheless such reasons were to be
rejected i£ they contradicted the patent facts. Never
mind that to Bradley the "facts" contradict themselves.
Russell's point of view, expressed here, reminds us of
Moore's. Philosophy, to Moore, must restrict itself to
giving a "description" of Common Sense propositions? to
Russell, it must restrict itself to "giving an acconnt
260
of" the world and daily life. Can we say, then, that
Russell took common sense (or Common Sense, in Moore's
sense) for his guide?
No, we cannot say that, and so we cannot apply
the same criticism to Russell which we applied to Moore's
position. Russell did seem to believe is Moore's Common
Sense propositions, but Russell also seems to have
reserved the right to criticize and rephrase any and all
propositions. Consider these words of Russell's:
Relations of terms to themselves are unavoid
able . 14
Does this sound like a common sense utterance? Would we
ever say, in ordinary usage, that something can be
related to itself? Would we ever say that this was the
very sort of thing of whose meaning and truth we were
most sure? No. Our common sense would incline us to
agree with Bradley, who said,
To myself it [self-relation] is either meaningless
or self-destructive. In making an ideal
experiment I either have no diversity, or else
the terms are different? and, when I suppress the
difference, the relation is destroyed.15
Russell said that a relation could have two, three, four,
or any number of terms, adding,
Jealousy, for example, is a relation between
three people.1®
Would our common sense incline us to say this? No, for
after thinking about it we would again have to agree with
Bradley who said,
261
The fact, I presume, is the jealousy that I feel.
And is it a relation, or is it, on the contrary,
a unity of another sort? To myself, as I under
stand the question, the answer is obvious. It
seems to me monstrous to describe the felt whole
as a. relation.
Jealousy does not seem to be a multiple relation among
three people, for it does not seem, in itself, to be a
relation at all. It is a feeling in one person about a
a relation between the other two, and that relation need
not even be actual so long as the jealous person suspects
it to be so. Even that third party need not be actual
so long as the feeling is there. Then why did Russell
call it a relation when, to our common sense, jealousy
is a feeling? Because, we suggest, his logic could deal
with feeling in no other way. And so what he said of
Bradley is true of Russell himself: his logic compelled
him to distort reality. He did not discern the form in
the fact.
But Russell cared little for common sense. He
cared much more for the inquiries which occupied him in
his attempt to establish a universal mathematics of
propositions. He wrote,
The correct statement of formal truth requires
the notion of any and every term, but not the
collective notion of all terms.1°
This astonishes our common sense, for when we say "every"
do we not mean "all"? We are not questioning, of course,
Russell's reasons for making the distinction between
262
"every" and "all/" for we suppose that they were very
good ones. What we are saying is that our common sense
is perplexed. We do not condemn Russell for insisting
upon such niceties of expression, for, indeed, we admire
such an elaborate and careful a train of thinking as must
be involved in finding the contradiction,, justifying the
distinction, between "every" and "all." The contradiction
was there, ordinary usage notwithstanding. Said Russell,
The above contradiction, which springs directly
from common sense, . . . can only be solved by
abandoning some common sense assertions [my
emphasis].
Then can we say that Russell used common sense for his
guide? It does not seem so. He may have believed in
Moore's Common Sense propositions, but that would have
involved him with making a distinction between Common
Sense and common sense propositions. He seems to have
criticized the latter without explicitly espousing the
former, and so we cannot call him a believer in the PCS
Principle.
But if Russell did not follow common sense what
did he follow? Reason? No, not always, for Bradley1s
use of reason was
An example of a method . . . calculated to produce
bewilderment rather than conviction; because there
is more likelihood of error . . . in a very subtle,
abstract, and difficult argument than in so patent
a fact as the interrelatedness of .things in the’
world [my emphasis].
He then added,
263
It has become natural to suspect a fallacy in any
deduction of which the conclusion appears to
contradict patent facts.
And lower down.
What we may call the empirical outlook has become
part of most educated people's habit of mind; and
it is this, rather than any definite argument, that
has diminished the hold of the classical tradition
upon students of philosophy and the instructed public
generally.20
What guiding principles can we draw from these statements
by Russell? They are that (a) reason can lead us into
error as well as truth, (b) patent fact takes precedence
if the reasoning is difficult, (c) appearances to the
contrary indicate that reason has gone wrong, (d) edu
cated people take the empirical outlook, and (e) Brad
ley's tradition was not refuted but was just abandoned.
There is nothing here to indicate a commitment to reason,
even though Russell said these things in a chapter
entitled, "Logic as the Essence of Philosophy.”
Russell committed himself to the patent facts,
not to reason. But what we are trying to determine here
is the principle by which to call something a "patent
fact." Why did he not just follow along with Moore?
What role did reason play, if any, in determining the
patent fact as distinguished from the object of just any
common sense assertion we might make?
Sometimes reason is to be followed, sometimes
not. When is it not to be followed? When the patent
264
facts take precedence over it. The patent facts
referred to are the important facts which are to settle
philosophical disputes and take precedence over reason,
at least sometimes. But in virtue of what are these
facts patent facts? It cannot be by acquiescence of
common sense, for Russell explicitly dismissed it, at
least sometimes. He found a contradiction in it. But the
facts are not patent by an appeal to mere rational con
sistency, for Russell, with equal emphasis, has dismissed
difficult reasoning when it issues in a conclusion
appearing to contradict the "patent facts ." To what,
then, did Russell appeal? It must have been to some
alternative not yet considered; some third or middle way
that was neither common sense nor difficult rational
argument. He appealed to "patent facts" which, he said,
had diminished the hold of the old rational philosophy
upon the instructed public. This public, as distinguished
from the ordinary ruck and run, had qualified itself by
adopting "the empirical outlook." Such persons would
"naturally suspect a fallacy" in whatever contradicted
their instruction and was at odds with their outlook,
Russell appealed to "most educated people’s habit of
mind."
This seems to have been Russell's foundation for
"patent facts;'.' Will it serve? It seems on the face of
it patently indefensible. It seems patently circular.
265
The facts are patent to whomever has adopted the empi
rical outlook. The empirical outlook is a determination
to be guided only by the patent facts. Can circularity
be more vicious? Russell may be right that the
empirical outlook with its patent facts has become the
foundation of public instruction: that public to whom
the facts are patent. To challenge Russell is to ask
whether there even is any such thing as "most educated
people's habit of mind," in spite of the instruction.
It involves asking whether, even if there were any such
thing, it ought to have any status in philosophy. These
questions will be left unanswered. Still, Russell's
position on "patent facts" is hardly tenable. It
appeals, in the end, to "most educated people's habit of
mind"— in virtue of which, we must presume, they were to
be called "educated." But such an appeal fails to pro-
22
duce total conviction.
C. Did Bradley Confuse A Relational
Complex with a Mere Collection?
Did our reply to Russell miss his point? His
point may have been that relational complexes are
irreducible in the sense that they are uniquely what they
are, and must be apprehended as such. He meant to
illustrate that point by the example of the asymmetrical
relation. He was saying, perhaps, that every relation is
what it is, qua relation, of which there are many sorts;
266
such as, asymmetrical relations, multiple relations of
varying degree of complexity, and so on. But each, as a
case of relation, is not to be understood in a way that
applies only to something else. That may have been his
message. And so, no relation is a mere collection of
things, nor can it be so conceived without destroying its
character— a character which we all recognize when we
agree to discuss it. A relation is a unity of a special
kind that analysis destroys. For example, of the
relation of "difference" in the proposition "A differs
from B," Russell said,
The constituents of this proposition, if we
analyze it, appear to be only A, difference, B.
Yet these constituents, thus placed side by side,
do not reconstitute the proposition. The dif
ference which occurs in the proposition actually
relates A and B, whereas the difference after
analysis is a notion which has no connection
with A and B...................................
. . . A proposition, in fact, is essentially a
unity, and when analysis has destroyed the unity,
no enumeration of constitutents will restore the
proposition.23
And yet Bradley did analyze relations, complaining
afterwards that the original whole had been destroyed,
All he had left on his hands were disparate parts, a
mere collection. Yet was he misled, or did he even seek
to mislead, by a word? The word was "collection,v which
suggests no formal sense of "together^'1 Had he thought
of an analyzed relation as a "complex" instead, perhaps
this confusion need not have arisen.
267
One might defend Russell in this way and one
would not stand alone. McTaggart took a similar line of
r e b u t t a l .24 The issue is Bradley's charge that a fact,
as merely relational, entails contradiction and vicious
regress; it is not true as it stands. To regard a fact
as a relation is to regard it as a grouping of separable
parts, yet from separate parts a whole fact can never
be had. Remember the sad case of Humpty-Dumpty. Such
was Bradley's charge.
Both Russell and Mctaggart sought to answer that
charge by drawing a distinction. Russell seems to have
drawn his distinction between "collection" and "complex,"
but his distinction does not meet the objection. Whether
we call it a "collection" or a "complex*;" yet do we not
distinguish between the relation and the terms? Then we
have asserted a difference which justifies the dis
tinction and, as we have just seen in Russell's own
words, "different from" is another relation.
It is no adequate reply to say that in the end a
relational complex cannot be exhaustively analyzed, for
that is what raised the problem insisted upon by Bradley
all along. l£ there an unanalysable remainder? l£ the
whole more than the mere sum of its parts? Then this
introduces into the fact an unknown X of such importance
that it can hardly be ignored by logic. As Bradley
himself pointed out,
268
We are brought back once more to the fundamental
question of 'unities'. Is there, in the end and
really, for Mr. Russell such a thing as a whole
which is non-relational or supra-relational? Un
less we know the answer to this question, the
entire position seems doubtful throughout. Now
on the one hand Mr. Russell appears to be fully
committed to such unities, but, if so, how this
doctrine stands to his other views, I am unable to
conceive. On the other hand, in his interesting
little book the Problems of Philosophy the idea
seems to have disappeared. There is even a
tendency to imply that a complex unity consists of
a relation (p. 202). But this problem surely (if
in philosophy there are problems) is second to none
in importance.25
Did Russell regard aRb as a unity, in the serious sense
that a, and R, and b, could not indifferently and un
affected enter into other relations with other terms?
His answer seems unclear.
McTaggart had the answer. The fact of relation
was to be saved, true enough, by the drawing of a
distinction; and it was also true that the possibility of
further analysis did not destroy that fact. But the
distinction is not to be drawn between "complex" and
"collection for both words suggest parts, and so, allow
for the generation of a regress. Let us then begin
differently by recognizing that, whatever discussion may
follow the fact of some relation, yet in fact relations
are given. There are merely relational facts, such as
that X is north of Y. To go on from there and to recog
nize further that there is an X, and a relation, and a
Y, is to recognize the possibility of a regress; but the
269
regress is not vi'c'ions. It is only upon further analysis
that the regress is involved— but the mere possibility
of further analysis— even endless analysis— is simply
irrelevant to the initial and primary fact of the
relation. The regress is secondary to the primary fact.
What further relations may come between X and its
primary relation to Y are found by reflection upon the
primary fact of X's and Y's relatedness.
McTaggart thus drew his saving distinction
between two kinds of relatedness; primary and secondary.
Primary relations are the fact with which analysis begins
They allow for an infinite series of secondary relations
but that does not alter the fact of X being north of Y.
Thus, (1) "X has the relation of a term in the relation
R(X,Y)" is a first-order generated relation. It follows
that (2) "X has the relation of a term in the relation of
being a term in the relation R(X,Y)" is a second-order
generated r e l a t i o n .^6 This process can be infinitely
extended. Every relation generates a series of orders of
secondary relation, but the regress is not vicious: it
does not call into question the fact of the primary
relation. Bradley had overlooked a consequence of his
premises that relation generates an infinite series.
Grant the premise. Yet it is precisely that premise
which unites the series, for it asserts it as a fact.
The infinite series shows that the relation must be real,
not i;hat it cannot be real.27
1. What Distinguishes Secondary Relation from a
Property of Primary Relation?
If the distinguishing of a relation from a
property is ever warranted, then we should call the
"secondary relation" a mere property of a primary relation
which, if McTaggart's presentation of the case were
correct, the primary relation would necessarily have. The
difference justifying the distinction between primary and
secondary was found by observing that, given some
relation R(X,Y ) t X is disclosed as a term in that
relation. That would make "X is a term in R(X,Y)" a
statement of a property the relation has. But the more
serious question is, "Is such a relation given as a
fact? Now the fact of relatedness Bradley did not deny,
nor did he deny that "facts" (however we may use the
term) could be analyzed. What he denied was that the
mere relation was real, or a "fact'." Let us speak, now,
of "facts": whatever the word means, it means at least
that portion of the real which we experience. We
experience some relation. Can we say, "R(X,Y) is a
fact?" No. We must say, "The fact is such that R(X,Y)."
There is a difference between these two formulations, and
if a distinction is to be drawn, it should be drawn here.
We must abstract from the fact to speak of the mere
271
relation in the fact.
But with respect to the question of secondary
relations, if even further abstraction discloses mere
X as a term in that already abstract relation, have we
theright to say that X has the relation of a term in
iR(X,Y)? No. We observe no still further relation that
0 f t
is secondary. R(X,Y) has X as a property. °
2. Mere Relation Generates No Secondary Relatedness
Hft is not true that the primary relation
R(X,Y) generates an infinite series of secondary
relations, for it generates no relations at all. This
formulation begins with the questionable recognition of
a barely relational fact. Then it calls the parts of
that relation facts of another order; from which it
moves to the affirmation that those parts have the
relation of parts to the fact. But to say that of
R(X,Y) is not to add anything significant. It generates
no further relation to point out the parts. If, however,
we were to move from the initial fact of relatedness to
an abstraction written R(X,Y), but keeping the original
whole fact in mind, we might then justify an infinite
series of further assertions. If there were to be made
a series of further assertions, then this demands that
the initial fact should be infinitely richer than
R(.X,Y). That is precisely what Bradley had insisted
272
upon.
3. McTaggart Misunderstood How Regress Arises
McTaggart's orders of generated relation,
resulting from taking some relation in successively more
abstract acts of reflection, are not the way by which
relations involve infinite regress. The regress arises
with noticing that the real fact could not be a fact if
not more than relational. The mere relation is a many-
in-one, but the manyness of the parts does not account
for the oneness of the fact. Bradley's question was a
bold one: Can a relation express a fact? The most
primary of relations must fail, he argued, for no
distinguishable parts can account for their connection
in the whole. That connection is another relation, but a
relation no more successful in making the fact than the
first. Thus the regress arises. This difficulty is not
met by saying that the existence of a secondary series
generated out of a primary relation implies that primary
relation.
D. Bergmann's Reply to Bradley; the
Nexus of Fundamental Tie
Bergmann allied himself with Russell and the
notion that the task of philosophy was to construct a
"meta-language" which would explicate formal structure.
He also allied himself with Moore, for the formal struc
273
ture to be brought out was that of the common sense
utterance, which "sets the style,"
assigning to philosophy the task of elucidating
common sense, not of either proving or disproving
it. In this form the common-sense doctrine also
represents . . . that philosophy is descriptive.29
He admitted that there was some point to Bradley's
criticism of the relational complex— that it, simply in
itself, seemed to generate an infinite regress— but his
solution to the problem was to say that it instead
implied another sort of non-relational unity: the
fundamental tie or nexus. Agreeing with Bradley, he
said,
A relation is a thing, a kind of universal. If
this individual is to the left of that, there are
three things, two individuals, and a relation,
but added,
. . . held together by relational exemplification;
which is not a thing and hence not a relation,
but a fundamental tie.30
There was a problem with mere complexes, for
A complex is an entity among whose constituents
are at least two things. A thing which is not a
complex is a simple.31
And the problem was that
There must be ties, having ontological status, which
tie the simples into complexes. What then, one may
ask, ties the ties to the simples? There are only
two possibilities. One is, paradoxically, an
infinite regress, which is the way Bradley took to
monism. The other is my solution. There are
fundamental ties, I also call them nexus, which _
tie without themselves being tied to what they tie.
274
And so, here is Bergmann's solution to the problem of
relations. Either accept a paradox, with Bradley, or a
fundamental tie or nexus.
Bergmann's solution recommends itself to us.
Who but a desperate man would accept a paradox?
Bradley's regress afflicting relational complexes proves
the unreality of them only if we may assume that our
analysis of the complex is complete, when we have found
the terms and the relation. But perhaps we have over
looked another kind of entity, also present but dif
ferent from either the terms or the relation: the nexus
or fundamental tie.
1. The Nexus is Also Related
But Bergmann's solution has difficulties of its
own: (1) The fundamental tie i£ not the relational
complex for it is required by^ the relational complex.
Then does it not need another tie to tie it to the
complex? Or, if that tie is already had by the complex,
as one of its parts, what ties these parts together?
He denied that the fundamental tie needs a tie, but the
reason he gave for taking that position, and not
Bradley's,was that Bradley's led to regress. Why was the
fundamental tie needed to begin with? Because terms and
relations are distinguishable from each other, and so,
an actual difference comes between them. But the
275
fundamental tie is also distinguishable. It is_ not the
terms and relation. Then how is regress avoided.^
2. Monism Is Not Avoided
Bergmann rejected Bradley's "monism?'.' preferring
recourse to the fundamental tie or nexus. The alterna
tive to tegress is the nexus. Then suppose that the
nexus delivers the relational complex from threat of
regress. Yet how is "monism" avoided? There must be more
than one relation if "monism" is avoided. Indeed, there
seem to be many of them. Here the curtain is over the
window. This is one relation. There the chair is next
to the table. That is another relation. But these
related things are also related; the curtain over the
window is to the left of the chair next to the table.
This makes another relational complex, and it, like the
others (the curtain over the window and the chair next to
the table) has its nexus. The nexus of its relation
unites the other relations, with their nexus. But the
relation of these related things is still more compound,
for the rug is on the floor. That relational complex is
related to our other related things, with their nexus.
"Below" is the relation that the rug on the floor has to
the chair next to the table which is to the left of the
curtain over the window. And so, that makes for an even
more compound relation uniting all these related things.
276
It makes one relation, for, as Russell said, a relation
can have any number of terms. And to be one relation it
needs, and has, a nexus. The process of compounding our
relation must be continued until, at last, we have
related all things— and "things" include relations. This
relation uniting all things has its nexus. At last we
have one mighty Supernexus over all things! Then how is
34
"monism" avoided?
3. The Nexus Does Not Exist
Bergmann's meaning for "nexus" or "fundamental
tie," and "relation,is unclear. Is or is not a
relation a nexus or fundamental tie? Plainly not, it
seems, where in one place he says,
There are three things, two individuals, and a
relation, held together by (relational)
exemplification, which is not a thing, and
hence not a relation, but a fundamental tie
[my emphasis].
And yet the difference is not so plain in another place,
where he says,
Exemplification is not the only fundamental tie.
Conjunction, disjunction, and so on, represented
by the connectives, are others [my emphasis].36
And again,
I merely want to call attention to the connectives,
conjunction, and so on, since they are a kind of
nexus.3^
He says that a fundamental tie is not a relation, but
also says that conjunction, disjunction, and the
277
connectives, are fundamental ties. But isn't con
junction a relation? Bradley had thought so, and
Russell and Moore agreed. We had thought that con
junction, and other connectives, were relations.
But Bergmann escapes our criticism by again
drawing a distinction. He distinguishes between entities
of form-i and entities of form... A relation is a
1 2 ‘-
"thing," so it is an entity of form^. A conjunction,
which is a connective, is an entity of form2. > Thus a
connective can be a kind of nexus and yet not be a
■3 0
relation. °
Here again there appear to be difficulties.
It would seem that to distinguish between entities of
form^ and entities of form2 requires that there be a
difference between them. What is the form of that
difference? (a) If it is of a different form than
either form-^ or form2, then there is again a difference
between it and the forms with which it differs. That
difference again has its form, which is different from
form^, or forn^, or the form of the difference between
them— and thus an infinite regress is generated in the
old familiar way. But (b) if the form of the difference
between form^ and form2 is of the same kind as form^, or
form2/ then we have a formal difference that is not a
formal difference, which is absurd.
But Bergmann's answer is ready. He would here
278
invoke the "ineffability thesis."39 That thesis says that
any statement of form has itself a form different from the
form which it states. Thus the difference of which we
spoke, which leads Bergmann's saving distinction into
either regress or absurdity, is a difference which cannot
be stated. Therefore our objection cannot be stated.
Perhaps Bergmann has escaped. But the region to
which he has escaped offers no base from which he can
attack us. We find that these ever more abstract levels
of form, up which Bergmann has climbed to escape our
pursuit, do not "exist," they "subsist."^® The different
forms, and the difference between them, lies in a realm
which does not exist, but only subsists. Then we may as
well leave it to him. Bergmann has eluded us. Our prob
lem was to find out whether relations were real or not,
not to find out whether there was something else that
was even less real than they are.
E. Gram's Answer to Bradley
In this chapter so far we have examined proposed
solutions, to the problem of relations, given by the three
men of whom Gram has spoken. Russell's, McTaggart's,
and Bergmann's replies to Bradley have failed, as Gram
himself concluded. We have amplified his arguments to
that conclusion and added arguments of our own. Now
we must consider Gram's own proposed solution to
279
the conundrum of relations.
Russell, McTaggart, and Bergmann failed to answer
Bradley, says Gram, because they accepted Bradley's
fundamental assumption that whatever' is distinguishable
can exist separately. If separate existence follows from
distinguishability, then the relation is a separate
individual, and so, cannot account for the whole rela
tional complex. But Gram challenges this assumption. It
does not follow from the fact that a relation is
distinguishable in some fact, that it can exist as
another separate fact. Bradley's argument "trades upon
the ambiguity" of his initial assumption: upon the
ambiguity of the Axiom of Indiscernibles. It may mean
(a) a relation can exist outside its complex, as a bare
particular— in which case Bradley's conclusion follows;
or it may mean (b) we can find it outside this particular
complex in which we happen to find it— in which case
Bradley's conclusion does not follow.
If a relation is distinguishable in some complex
which "instantiates" it, it does not follow that it can
exist separately, and so, it does not follow that the
relation fails to actually relate the elements of that
complex. It does not need a further tie to tie it to
those elements. The mere fact that relation R relates C
to D, as well as A to B, in two complexes where the terms
are different, does not imply that R is a separate
280
individual.
Gram admitted that mere R does not explain how
C and D, which differ from A and B, are particularly
related in the particular instance. To explain that
would require that he
account for one contingent member of the universe
rather than another. But my failure to do this
permits no inference to a corresponding failure
to give an account of the possibility of an
entity of one kind of combine with entities of
another kind.
Thus Bradley's argument is "not powerful enough to
eliminate relations from an ontological inventory of
the world.
Merely because R comes between C and D as well as
between A and B does not imply that R exists without any
terms, and so, needs another relation. R has a character
of its own such that it can combine with others. How
absurd to say that because it has a character of its own,
it can't combine with others! It has an R character!
That seems to be the summary of Gram's argument.
1. The Relation Is Not the Fact
Gram, like the others, persists in speaking of
the relational complex as though it were a fact^2 which
the relation "accounts fori" The relation, he says,
"genuinely relates" the elements, even though A and B are
different elements from C and D. But we have been
281
through all this before. The "fact," if by that we mean
a real fact, is more than merely two terms and a
relation and so that fact is not accounted for by
pointing out bare R in R(A,B), or R(C,D). No doubt the
fact is such that abstract R comes between the abstrac
tions C and D, as well as A and B, but we would like to
know how, if the fact is accounted for. It is a
conditional fact, and the conditions are included if the
fact is to be accounted for. Gram's admission that he
cannot explain how R comes between the particular terms
leaves us wondering how to justify the claim that R
accounts for any fact.
2. Gram has challenged Bradley's "fundamental
assumption," the Axiom of Indiscemibles: that the
apparent identities present to our experience are to be
taken as independent of their surroundings, so long as
not discernibly changed by them.43 Gram says,
"Bradley's argument can get started only if it can be
shown that distinguishing one entity from another
renders both of these entities capable of independent
existence. "44
Gram's implication is that Bradley took
"distinguishabllity implies separate existence" as a
true premise, from which he then argued to his con
clusion. Bradley would have been appalled. He nowhere
282
took this for a true premise. He argued that if we take
this for a true premise, then disaster follows. He argued
j
that we do take this for a true premise, within the re
lational view of things, so our world, so conceived, must
be no more than an appearance. He said that in logic we
take this for a true premise:
The Principle of Identity . . . the Axiom of In-
discernibles . . . [is] . . . that inference rests
on the principle that what seems the same i£ the
same, and can not be made different by any diversity,
and that so long as an ideal content is identical no
change of context can destroy its unity. The
assumption in this principle may be decried as
monstrous, and I do not deny that perhaps it is false.
In a metaphysical work this question would press us,
but in logic we are not obliged to discuss it.4^
Later, in Appearance and Reality, he discussed the
pressing question of this principle, and rejected it.
But we must continue to presume the Axiom of Indis-
cernibles in logic. That is what poses the conundrum.
Was Gram talking about a presumption indispensible
to logic when he placed the relation in "an ontological
inventory of the world"? The word "ontology" suggests
metaphysics. And the trouble is, how are we to prove
that a bare relation is real? An instantiated relation,
of course, exists, but it is hard to call something real if
it exists only on condition of something else— as is the
case with relations, Gram agreed.
F. Summary of Chapter V
Russell claimed that the asymmetrical re1ation
283
offered a fatal exception to Bradley's view of relations/
for it could not be reduced to a property of the related
terms. Bradley's logic, which took "S is P" as the form
of all judgments, could not express such relations with
out distortion. His faulty logic, which took judgment
to predicate of reality, erroneously led him to believe
that relations are unreal.
But Bradley did not hold that judgment predicates
of reality merely because his logic^required it. Rather,
he believed that logic required it because otherwise
logic could not allude to the conditions which make a
proposition true. Russell was mistaken to think that
Bradley had overlooked asymmetrical relations, or that
he had tried to reduce all relations to properties of
finite terms. Russell misunderstood Bradley's "monism
And finally, Russell's patent fact, which dictates the
form of logical schemata, is itself troublesome. The
criterion for "patent fact" was not the common sense
proposition, for sometimes common sense could be wrong.
Yet Russell did not allow "patent fact" to be sub
servient to merely rational requirements.
Perhaps Russell's point was that Bradley had
confused a relational complex with a mere collection.
McTaggart made the point more strongly, arguing that
Bradley's criticism of relation proved, rather than
disproved, relations. It proved that primary relation
284
generates secondary relation. Bradley had to assume the
reality of primary relation in his initial premise.
But McTaggart had confused a "secondary" relation
with what could be called, with equal justice, a property
of primary relation. Moreover, he was mistaken to think
that a bare relational "fact" generates any secondary
relatedness. And he was mistaken about how relation
generates regress: the regress results if relation is
taken as itself a fact.
Bergmann argued that Bradley's criticism of
relational complexes proves only the existence of
another sort of entity: the nexus or fundamental tie.
But the nexus, being different from the complex, surely
must be related to it; the nexus of relation between
other nexus threatens to resassert "monism"? and the
difference between "nexus" and "relation" does not exist,
it subsists. It does not therefore help solve the real
problem of relations.
Gram argued that the others had failed to answer
Bradley because they accepted his false fundamental
assumption: The Axiom of Indiscernibles. Distinguish-
ability does not prove separate existence, so Bradley's
argument cannot get started. But Gram has confused a
relational complex with a fact, and he was mistaken to say
that Bradley had assumed the truth of "distinguishability
entails separate existence." He assumed it as a working
285
iple in logic, but found it indefensible in
prxnc
metaphysics
FOOTNOTES— CHAPTER V
I
I must acknowledge my debt to M. S. Gram for the
grouping together of these three men. See his "The
Reality of Relations," New Scholasticism 44 (Winter 1970);
49-68.
Gram begins his article with the words, "It
remains a scandalous fact of philosophical criticism that
there has been no successful answer to the argument
[Bradley's, against relations]." Gram ends with a pro
posed answer to Bradley free from the "patent defects"
of others, and Gram's proposal, too, will be examined in
this chapter. But at least, until Gram, Bradley had not
been answered. This essay was complete, in an early
draft, before Gram. While his arguments have influenced
the content of this chapter to some degree, yet we have
tried here to present the material differently and in
more detail.
^Russell will be remembered for his contribution
to mathematical logic, but as he himself said,
"Mathematical logic, even in its most modern form, is
not directly of philosophical importance except in its
beginnings [Bertrand Russell, Our Knowledge of the
External World, rev. ed. (London: Open Court Pub. Co.,
1926? rpt. George Allen & Unwin Ltd., 1969), p. 50.
Hereafter cited as OKEW]." Here we are concerned with
mathematical logic only as it is of direct philosophical
importance.
•^Ibid., p. 48.
4Ibid.
5Ibid.
®Ibid., p. 49.
^Ibid., p. 55-56.
®Ibid., pp. 56-59.
^Ibid., p. 59.
■^There is some truth to the charge that Bradley
reduced relations to properties of Reality— but none to
the charge that he made them properties of the mere
287
related terms. He did make relations qualify Reality
lest, being merely isolated, they cease to relate. But
he did not make them vanish into the terms.
Indeed, far from reducing all relations to
properties of terms, Bradley made all properties
relational. Every subject-predicate assertion becomes,
in his hands, a relational assertion. We have already
spoken of the problem of relating a "thing" to its
property. The inherence of qualities, characters, or
properties was, for Bradley, a perplexing problem in
relations. The mere "thing;," as an identifiable subject,
and its property, which is no less identifiably something
else, stand apart. But our judgment says they are
together. How do we account for this "together"? What
supplies the conditions for being together? The relation
is the outstanding problem.
H f. H. Bradley, Collected Essays, "Relations ,
ed. H. H. Joachim, 2 volsT (1935? rpt. Freeport, N. Y.:
Books for Libraries Pr., 1968), 2:672. Hereafter cited
as "Relations CE.
F. H. Bradley, The Principles of Logic, 2nd ed.,
2 vols (1922: rpt. London: Oxford U. Pr., 1967),
1:289. Hereafter cited as Logic.
13Russell, OKEW, p. 18.
14Bertrand Russell, The Principles of
Mathematics (1903; rpt. New York: W. W. Norton and Co.,
1969), p. 44. Hereafter cited as Principles of Math.
15p. H. Bradley, Essays on Truth and Reality
(1914; rpt. Oxford: Clarendon Pr., 1968), p. 290.
Hereafter cited as ETR.
16Russell, OKEW, p. 60.
17Bradley, ETR, p. 304.
18Russell, Principles of Math, ibid.
19
Russell, xbid.
20Russell, OKEW, pp. 17-18.
21I do not deny that there is, less or more or
even for the most part, something we call "the climate
of prevailing opinion- ." By that we mean (when we take
it for our guide), "the opinion of qualified persons."
And yet, in philosophy, must we not ask for these quali-
288
fications before using such a thing to overthrow an
otherwise sound argument?
Perhaps Russell did not intend "most educated
people's habit of mind” for a criterion of truth, but
if not, then why did he speak of it as the reason for
abandoning the classical tradition? And if he did not,
then in virtue of what are the facts "patent"? I mean by
that, in virtue of what is some empirical judgment taken
for categorical truth? Why is the significance of this
premise beyond question? How much weight shall I attach
to it?
It is true that we do not always use a criterion
for truth. When I see the cat on the mat I do not
consult a criterion before daring to say, "The cat is on
the mat.1 1 But I do use one when matters cease to be
trivial— when a genuine question assails me. Then,to
the question, "What does this appearance signify?" I ask
myself, "How does its seeming significance fit in with
all else which I think that I know? If it disrupts or
contradicts, then I must reinterpret the significance—
perhaps in toto. First appearances are not patently
incorrigible. They do not bear their own patent. They
are subject to correction, and the criterion seems to be,
comprehensive coherence, as I understand the term.
Reason is my guide.
2 2 See Appendix B, "What Is A Fact?"
^Russell, Principles of Math, pp. 49-50. It is
curious that here Russell uses an argument, to prove the
irreducibility and uniqueness of relations, which is
similar to what Bradley used to prove that relational
thought must give appearance only, not reality.
2^The similarity between Russell's and
McTaggart's views on relations is pointed out by Gram,
"The Reality of Relations op. cit., pp. 58-60.
^Bradley, ETR, p. 303.
2®This is Gram's formulation of McTaggart's view.
2^See Gram, "The Reality of Relations," op. cit.,
pp. 55-60.
2®Ibid. But let me emphasize that the above
formulation is my own.
29Gustav Bergmann, The Metaphysics of Logical
Positivism, "Reconstruction of Metaphysics" (New Yorks
Longman's, Green and Co., 1954), p. 35.
289
3®Gustav Bergmann, Logic and Reality (Madison,
Wis.: U. of Wis. Pr., 1964), p. 94.
31Ibid., p. 245.
32Ibid., p. 229.
33gee Gram, op. cit., pp. 61-63.
34Surely this Supernexus is real, for what but a
real nexus would be powerful enough to guard real
relations which, but for that nexus, would collapse in
vicious regress?
05
Bergmann, Logic and Reality, op. cit., p. 94.
36Ibid., p. 88.
37Ibid., p. 199.
38Ibid., pp. 52-53.
38Ibid., p. 55.
40Ibid., p. 56 .
4^Gram,
•
"The Reality of Relations, " op. cit.,
42Again, on the ambiguity of "fact" see
Appendix B.
43My own statement of the Axiom, but not un
acceptable to Gram, I think.
^Gram, ibid., p. 66.
43Bradley, Logic, 1:288-9.
290
CHAPTER VI
UNIVERSALS, IDENTITY, AND THE CHALLENGE
FROM RESEMBLANCE
We have examined the arguments which started the
"revolt into pluralism;"^ We have considered the
"refutations" of Bradley's view of relations, from the
inception in Moore and Russell, to Bergmann who tried to
combine and perfect what these two had started. We have
found the arguments unconvincing. We were helped to that
conclusion by the work of Blanshard and Gram, but Gram was
another critic whose criticism failed. Now we find that
2
Blanshard, too, rejected Bradley's view of relations.
We turn to his criticism.
Blanshard's is perhaps the most serious challenge
of all. He was, for years, an advocate of Bradley's view.
He had nothing to gain by changing his mind. He was no
enfant terrible seeking to score a succes de scandale.
His criticism is sober, well-considered, and informed.
It may well be fatal to our attempt, in this essay, to
defend Bradley.
Blanshard appears to have given up the doctrine
of internal relations over the problem of universals.
In order to give an account of them he felt obliged to
recognize one external relation— a very pervasive one:
291
the relation of resemblance. Thus Bradley's view of
relations was abandoned in principle. According to
Blanshard, without the relation of resemblance to account
for universals, thinking cannot even begin.
A. The Relation of Resemblance is
Required by Universal Ideas
Why was it necessary to recognize the external
relation of. resemblance? It was needed to solve the
problem of universal ideas. We cannot deny them, and
Bradley recognized that they are used even by such
3
creatures as the dog. But Bradley seems to have failed
to give an adequate account of universals. He seems to
have believed in an "identity" theory of universals, i.e.
that what was universal was always a common element among
instances.4 But this cannot be true of all instances of
universal ideas, Blanshard said. There are difficulties
to so simplistic and indiscriminate a view of
universals.
Even at the lower levels of life, Blanshard
agrees, "action follows lines of connection between
characters."5 Yet we must be discriminate about what we
affirm of their existence. Some distinctions are needed.
There are three types of universals: specific,
qualitative, and generic. A generic universal is one
whose instances are such things as men, horses, and
stones. A qualitative universal would be a color, a
292
sound, or a shape. Both of these sorts of universal
are subdivisible. The generic universal "man," for
example, is subdivisible into "animal," "rational," "two-
legged and so on. The qualitative universal "color”
is subdivisible into "yellow," "red*," "blue," and so on.
But now we reach the specific universal. It is not
capable of subdivision. It is specific. It is this
shade of red, this degree of loudness, this precise
shape, and so on. It is universal because it in principle
could be repeated. It is very common. Blanshard says,
There is no fully specific quality we ever
experience that is not a universal of this kind.
Every sound or odour or taste that anyone ever
experiences, every sensation of hot or cold,
every ache or pain, has a quality that in prin
ciple could be repeated and hence is a universal.
These are sensible qualities . . . [or] . . . non-
sensible characters such as numbers . . . [or] . . .
spatial qualities as immediately apprehended.
7
Only the specific universal exists, says Blanshard.
Generic and qualitative universals exist only through
resemblance among specific ones. Generic and qualitative
universals are derivative and dependent. "Color;" for
example, is but a resemblance among specific yellow, red,
and blue. It does not exist in nature.
But how do we know that generic and qualitative
universals do not exist? "The prime difficulty is simply
that it is hard to verify the existence of these
universals."8 It is hard to isolate and think clearly
the identical bedness of all beds, or to think color when
293
it is no specific color. Such invariant characters
cannot be isolated, even in concept. Further, if we are
to abstract some one identical character from any
generic or qualitative universal in virtue of which it is
an instance of such, the range of instances must clearly
be known. If we are to abstract the common element in
"blue there must be some limit clearly demarcating the
blue things from these which are violet or green. This
is the problem with qualitative universals. The diffi
culty with generic universals is similar. The dog, for
example, is a carnivore. Yet some dogs like to eat
fruit and vegetables. Are these latter dogs to be in
cluded among the carnivores? Again, what is the
quintessential cat? The housecat hates water, but the
Siberian tiger loves it. The cheetah can't retract his
claws, but he is a true cat, nonetheless. The civet cat
isn't a cat at all. He only resembles one. Some cats
are tailless, or hop like rabbits, or will retrieve like
a dog. The Manx cat possesses all three of these
unfeline qualities. He is still a cat. It seems that
neither the cat nor any other natural species is con
sistently one within its own class, nor clearly marked
off from the other species.
What is to be done about this difficulty? It
appears that the identity theory of universals will have
to be abandoned. It cannot be true of generic and
294
qualitative universals, for it is hard to isolate and
think clearly any one thing that is the same throughout
the class of instances for which the universal stands.
Yet on the other hand we cannot just simply dispense with
universals altogether, for, as Blanshard says, "action
follows lines of connection between characters."
Inference is central to intelligent activity, and in
ference needs universals. The solution of the problem
must lie somewhere between the two extremes: between the
identity theory applied to all universals, and the utter
denial of any universals.
Now it has already been said that specific
universals are free from the problems that afflict
generic and qualitative universals. The solution of the
problem must lie in a distinction between specific
universals, on the one hand, and generic and qualitative
universals, on the other. It must be that this latter
pair of universals, for whose universality the identity
theory does not account, borrows, so to speak, its
universality from the others. How does it borrow from
specifics? By resemblance. Qualitative universals such
as "blue" are derived from a resemblance which some
specific shades share among them. Similarly, "color" is
derived from a resemblance which this red, this blue, and
this yellow have among them. That yields qualitative
universals from specific ones. Again, the generic
295
universal "cat" is derived from the resemblance that
certain animals share among them. There is a resemblance
among groups of specific universals, such as this hair
texture, this shape, this sort of tooth, claw, cry, and
style of locomotion. That yields a generic universal
from specific ones. Blanshard says we must "combine an
identity theory for specific universals with a
resemblance theory for qualitative and generic
g
universals."
What does "resemblance" mean here? It means a
relation among specific universals in a way that is
"ultimate, and is not resoluble in all cases into partial
identity." Some universals, such as "color" and "cat,"
are such by resemblance only, and not by common identity.
Only the specific universal satisfies the identity theory
of universals.
How does all this affect Bradley's view of
relations? It pulverizes his one reality into a vast
swarm of specific universals, each one a somewhat which
could, ex hypothesi, exist unchanged elsewhere.
"Our experience is of nothing but specific universals,"
Blanshard says.^ The emphasis laid upon "nothing but"
is his own. Then we directly experience nothing but
these independent specifics, which are related,
externally, through resemblance, to create the universals
of greater extension and intension. But not only does
296
this theory admit the external relation of resemblance,
for it also admits a great many others. He has said,
"Every sound or odour or taste that anyone experiences,
every sensation of hot or cold, every ache or pain, has
a quality that in principle could be repeated."^2 He
went on to say that "non-sensible characters such as
numbers" are also specific universals. "Threeness" is
such a character. Then why not admit, to status as a
specific universal, the "betweenness" of a specific
between, or the "differentness" of a specific difference?
Could they not also in principle be repeated? But they
are specific universals of relation. Then they are
external relations.
Change need not affect the specific universals,
even those which are relational. In principle they
could be repeated, so they need not relate just these
terms. The terms could be different, and indeed must
be different, if the specific relation is repeated. The
two occurrences, of which one repeats the other with
respect to the specific universal, must be distinguish
able if we are to call them "two." That entails
qualitative difference. Mere numerical difference does
not support a distinction, as Blanshard himself insists,13
so it cannot support "repetition,"
All ordinary things such as men, trees, and
stones are but composites woven up out of specific
297
universals, through the relation of resemblance. Only
specific universals are encountered in immediate ex-
|
perience. These are the ones that exist. If one seems to j
i
encounter a tree, it is by derivation from a whole swarm of;
specific universals, through resemblance. Thus we can
hardly say that resemblance comprises an isolated and unim-:
i
portant exception to Bradley's view of relations. The re
lation of resemblance is everywhere, for Blanshard, and it :
does not alter the terms it relates. Thus, it is difficult
to conceive any philosophy more opposed to Bradley's view
than this view. It is a radically atomistic view in which
the atoms are specific universals. Moreover, all re
lations must be external, for the specific universal (the
only -universal that is, and includes relations) , could
appear elsewhere unchanged. Bradley's view of relations
has been thoroughly demolished.
B. Did Blanshard Contradict Himself?
To so respected a scholar as Blanshard we cannot
bring ourselves to say, "et tu, Brute?" but we are here to
defend Bradley, not to bury him. We will continue to
defend him if we can.
In The Nature of Thought, published in 1939,
Blanshard said, "It is hard to see how the conclusion is
to be resisted that the universe of existing things is a
system in which all things are related internally."1^
298
Twenty-three years later, in Reason and Analysis, we find
him saying, "The argument of the rationalists to the
effect that all things are connected through the relation
of difference, that this is in truth a necessary
relation, and hence that all things are connected
15
necessarily, is a valid argument." Blanshard seems
steadfastly to have claimed a belief in universal
internal relatedness. Yet he has told us that all things
are composites of specific universals, which alone exist
to immediate experience. "Things," such as stones and
men, exist only dependently through the ubiquitous
external relation of resemblance. How are we to under
stand this experience, which consists cf a vast swarm
of specific universals able to exist elsewhere in other
contexts,unless the relatedness is external?
Blanshard seems to have espoused a view of universals
that contradicts his view, elsewhere expressed, about
relations.
But we hesitate to make against Blanshard so
serious a charge. Are we laboring under an unwitting
assumption of the identity theory of universals? We
have called the ubiquitous relation of resemblance an
external relation. Are we unwittingly assuming that
there is some one thing common to all instances of
resemblance,.by which they may be said to be instances
299
of resemblance, and not, say, difference? If so, then
we are guilty of precisely the blunder which Blanshard
warns against. Perhaps if we apply the resemblance
theory of universals, not the identity theory, to the
relational universal of resemblance itself, objections to
his theory may collapse. Instances of resemblance might
just resemble. Two cabbages could resemble, two
concertos could resemble, and yet these two instances of
resemblance might be called "resemblances" because the
17
relation only resembles itself in each of the cases.
But resemblance must still be an external
relation. It cannot be internal. Even though the
relation of resemblance merely resembles itself, and so,
differs with the instances related, still it must be
external to those instances. What is there really, in
rerum natura, to relate? Only specific universals. They
are simple qualities, such as this red, this loudness,
or this roughness. They do not in themselves refer to
the others. They have no other quality, other than their
own simple one, to account for the relatedness. They
resemble, but it is nothing to them that they should
resemble. The relation must be added on from outside.
Speaking of the generic or .qualitative universal,
Blanshard said, "neither of these exists outside our
thought."^® But specific universals do exist outside of
19
thought, for they are what immediate experience is of.
300
It is the resemblance among the specifics that gets them
into our thought, as generic and qualitative universals.
If the resemblance were internal to the specifics, then
so would be the generic and qualitative universals.
They would exist in rerum natura along with the specific
universals. But they do not. Then resemblance must be
an external relation.
To defend Blanshard against the charge that his
specific universals must be externally related, one might
point out that he didn't say so, and perhaps we have mis
conceived his intent. What he claimed for the inde
pendence of specific universals was that in principle they
20
could be repeated. They are not in principle unique.
He did not say that they actually were indifferent to
their context. All he said was that the same specific
universal could appear again, that there was nothing
about it to prevent it from appearing again. Surely
this capacity of appearing again^ is the minimum
necessary for any universal. We must find some way to
explain universals. Even Bradley insisted upon uni
versals. Indeed, he argued that every phenomenon is
» oo
universal, and not particular. *
But a false explanation is worse than no
explanation. Bradley did not claim that a fully
intelligible account could be given of universals, even
though nothing is more sure than that they exist. The
301
word "exist" is fairly loose: loose enough to allow
for ambiguity; loose enough to allow relational thought
to make an appeal to what lies beyond itself in order to
support its claims. But if we try to make relational
thought our object, as we must do when we try to give an
account of universals, then we find ourselves perplexed.
(See Chapter I, sections F, G, and H.) Bradley did not
call universals real. They too, if taken in themselves,
2 3
are mere appearance. Relational thought and the
universals it claims belong to reality somehow, but the
question is: "How?" The question remains acute, for the
standing contradiction of self-contained terms and
external relations continues to afflict them. But
Blanshard did not distinguish between appearance and
reality, to place universals among the former. His
specific universals are real. His avowed belief that all
relations are internal does seem to conflict with his
theory of universals.
C. Blanshard, James, and the
Resemblance Theory
The resemblance theory of universals is correct,
and Blanshard did not contradict himself, for resemblance
is an internal relation differing with the specific
differences, or the related terms. It only resembles
itself. That, at least, must be the rationale which a
defender of Blanshard's position would have to use to
302
escape what seems like a contradiction. His view of
relations (that they are all internal) seems clearly in
contradiction to his theory of universals (that generic
and qualitative ones are generated by resemblance). It
must be that resemblances can differ, only resembling
each other. That is because the relation is simple—
which means that two resembling things need have no
common identity by which they resemble.
Now we have argued elsewhere that there can be no
such thing as a "simple," even as a relation. It must be
internally compound. It must be what it is, independent
of the context, yet it must relate just what it does
relate, which refers it again to the context.
"Resemblance" may be "simple," but we must know what that
means. Surely it does not mean that the relation of
resemblance is a property of the terms, for the relation
must have positive quality enough to merit the name
"resemblance" which we give it. And yet there must be
something about it which allows, facilitates, and even
requires it to join with just the terms it does join.
Thus there seems to be a problem infecting all cases of
relation, and one which, we had thought, Blanshard
recognized. But his relation of resemblance seems to be
a special case. Perhaps we do not understand what
Blanshard was driving at. Whatever the "simplicity" of
the relation may mean, it is at least a quality that lets
it escape Bradley's criticism. The latter affects only
relations interpreted under the "identity" theory. Then
let us examine the reasons why the identity theory is
unacceptable. It will perhaps make the resemblance theory
appear more plausible.
William James advocated the resemblance theory
against Bradley's identity theory, and Blanshard approves
of what James said. Blanshard had sided with Bradley, who
was the better logician, but later came to believe that
O A
James was right. We must examine their arguments.
Bradley and James had debated the issue of identity
versus simple resemblance in the pages of a learned
journal, where the issue in its day became a cause
celebre.
1. Identity Cannot Always Be Found
James's first argument was siiriptly that the
identity theory of universals fails because sometimes the
purported identity cannot be found, and so it is not a
necessary condition for universals. Sometimes, of course,
it can be found, as is the case with two similar lines.
They are similar in having length, and length is the
identity which explains their similarity. But it is not
necessary that an identity be detectable for us to say
that one thing resembles another. Consider any two
colors. Unlike the two lines, there is nothing in two
colors analogous to "length" in which they either
304
resemble or differ— and which justifies saying that they
both are instances of color. What could it be? Blue and
yellow are both colors, and they differ. But according
to the identity theory they must also share some common
element so as to be called "color." "Is blue yellow plus
2 6
something?" asked James. "If so, plus what?" There is
between blue and yellow a simple difference of kind.
That "kind" is got from their resemblance: a simple
resemblance between them in which they differ as specific
determinations. "We are forced to admit," said James,
"that differences of simple kind form an irreducible sort
of relation. James himself emphasized these words,
and doubly emphasized the word "kind ." What is found in
all series such as color, sound, texture, etc. is
qualities which simply resemble, and the resemblance
gives the "kind . "
2. The Identity Theory Leads to Vicious Regress
James1s second argument was directed against the
identity theory again on the grounds that a series
cannot be explained by saying that there is some common
shared identity among the members. According to the
identity theory, a series of tones a, b, c, .... must
have some X among them common to all. Thus, to ade
quately state the series we ought to write it Xa, Xb, Xc,
..... But even the addition of this common X to the
305
members of the series is not enough to state it ade
quately, objected James, for the differences of the
members a, b, c, .... forms another series which remains
unstated, so what is gained? If you must state the
identity, beside the members, then must you not state the
difference also? The series of these differences may be
written a, 3, y, ..... But because they form a series,
they must have a common element X', by the identity
theory. To state it we must write, X'a, X'3, X'y ......
But the series of these differences within the original
series have differences among themselves, as members,
that form yet another series which we must include in our
statement; and these differences, forming a series, must
have another common element X'1. To write that series
we must write X''a, X''3/ X''y . ..., and the regress
is obvious. The identity theory cannot be true. What
are we to do? James had the answer. The oneness of the
series cannot be got from some identity shared among the
members which is, at the same time, other than them.
Rather, what makes a series of the members is simply a
high degree of likeness.among them. Identity thus is
2 f i
"nothing but a high degree of likeness," James said.
3. The "Chinese Boxes" Objection
James' third argument was like the second in
306
claiming that the identity theory of universals led to
vicious regress. The theory cannot be true, for it leads
to a set of encapsulating 'identities' to whose pro
liferation there is no end. Suppose that two balls
resemble each other. The identity theory explains that
resemblance by saying that they have a common quality,
call it M. Two apples also resemble, and so, have a
common quality N. But the balls resemble the apples so,
internal to both the M and the N, which explains the
shapes that resemble, there is another common quality 0,
to exolain the resemblance of the shapes. But two goose
berries also resemble, by a common P, which resembles the
0 of the resembling shapes of the apples and balls by a
common quality Q. But there are also two pumpki.is which
resemble by an R, which resembles the Q by an S, and we
have not yet accounted for the further resemblances of
planets, boulders, and all the rest which resemble, and
whose resemblance resembles the resemblances mentioned.
Each of these likenesses of shape is like the likenesses
of the others, by an identity shared between them, and
so, each ex hypothesi somehow encapsulates each and all
of the others and is in turn encapsulated by them.
29
Surely it is simpler to accept resemblance per se. ?
D. Reply to the Arguments for Simple
Resemblance
The identity theory is an attempt to account for
307
universals, as is the resemblance theory. But we must
admit that universals can be used without an adequate
explanation. The cat can use the universal "rat" al
though he does not bother to explain whether it is the
simple resemblance among specific rats that enables him
to know them, or whether he has subsisting in his sub
liminal process an identity under which particular rats
are subsumed. At any rate, it matters not to him whether
the rat is brown or grey or inhabits the attic or the
cellar. He knows a rat when he finds one. Or he may be
even more catholic in his action, acting upon the more
inclusive universal, "prey." It covers, without
discrimination, rats, bats, birds, and mice. Indeed, as
a kitten the range of his -universal was wider still,
including such things as balls, spools, and strings
dragged along the floor. But his discrimination of
specific differences increased with his experience: a
truth which is embarrassing for the notion that the
generic universal is dependent upon a prior acquaintance
with specifics. This notion is presupposed by the
resemblance theory. You may seek to avoid the embarrass
ment by declaring that the specifics are universal too,
but you have not accounted for the order by which they
come to be known. The baby calls all men "Father."
Bradley found none of James's arguments con
vincing and offered replies to them. He did not change
308
his opinion, nor did James. At last, in an attempt to
settle the issue, Bradley said,
I will make Professor James this offer. If he
will state the principle on which he objects
to identity (a thing which, let me remind him, he
has not yet even attempted to do), I will also take
that same principle, whatever it is. And I will
show that, judged by it, Professor James's thesis
as to Resemblance . . . is indefensible.30
The offer seems fair, but there is no evidence that James
accepted it.
James's arguments do seem weak. In the first
argument he said, "differences of simple kind form an
irreducible sort of relation." He used the words "sort"
and "kind" with emphasis and double emphasis. Then are
we to take sorts and kinds seriously? James asked,
rhetorically, "Is blue yellow plus something?" But
no one suggested that the "kind" or "sort" we call
"color" consisted of something added on, and so
externally related to, its specific determinations.
But James spoke of "sorts" and "kinds !I1 He spoke of
differences that were "of" them. Then might we not have
an identity of them? He said that identity was "nothing
but the extreme degree of likeness." The emphasis again
was his. But "degree"? Degree of what? Degree of
likeness? And yet, on the resemblance theory, there can
be no one thing of which there is any degree.
Jame's arguments do not seem convincing, but
Blanschard found them so.^ He found fault with Brad-
309
ley's replies. Then let us turn from James to Blanshard j
: I
I and see if we can find in Blanshard's criticism of j
Bradley's replies what we have failed to find in James's !
arguments.
1. James's Logic is Faulty
James's first argument was that identity between
resembling things was not always discernible. Blanshard,
supporting James, argued that Bradley had assumed the
Axiom of Indiscernibles which states that distinction
32
must be based upon observed difference. Bradley had
replied to James's first argument that it was not
necessary to observe a common quality in order to posit
identity.-*3 Identity need not be distinguished in order
to be asserted, Bradley implied. Bradley thus contra
dicted himself in his own procedure. Indeed, in this
very chapter ve have used the presumption that distinction
must be based on observed difference. "Bradley doesn't
show that resemblance must be analyzed through identity?
he does not show that ultimate resemblance is nonsense,"
says Blanshard.3^
So far Blanshard is right. Bradley both affirmed
and denied the Axiom of Indiscernibles. We have followed
his example, but we have also tried to follow a rule. We
have assumed the Axiom in our logical procedure, for it
seems indefensible to make distinctions without an actual
310
difference. But we have denied that those differences,
even when observed, are ultimately real as they appear:
because reason requires it. Now Blanshard denies our
right to affirm oneness when we sometimes don't observe
it, because the Axiom forbids it. But as we understand
the Axiom, it only requires that we justify our
distinctions if we make them, or adhere consistently to
meanings as we take them. It does not say observation
must always confirm what it sometimes confirms.
But now, who assumes the burden of proof? Was it
not Blanshard, following James, who set out to refute the
identity theory? James's first argument doesn't do the
job. The argument is that identity sometimes isn't found.
Let us ask, on Bradley's behalf, "What principle is being
used here?" It is that what isn't found can't be. Or,
that whatever is, appears to us. Or, reality is as it
appears. These are worthless principles. Human discern
ment isn't total. The argument, if we can call it that,
is ad ignorantiam. But what makes this bad argument
worse is that even if the identity theory were proved
worthless, still it does not prove that the resemblance
theory is the only and correct alternative. This logical
procedure is defective. We agree with Bradley who said,
35
"This argument appears to be thoroughly unsound."
In his first argument James emphasized "sorts" and
"kindsBut if there is a sort or kind, we must identify
it as the sort or kind that it is, and resemblance cannot
help us. According to the resemblance theory the sort or
kind is not presented in the instance itself, which
consists of nothing but specific universals. The sort or
kind must be brought in from elsewhere. Blanshard tells
us that it is made by the mind, out of resembling
specifics. Then which resemblance are we to select? We
are aware of more than one single specific universal, at
any time. There are an indefinite number of resemblances.
Indeed, how are we to fasten upon resemblance as the
important relation? There are an indefinite number of
relations present. This whole process wants explanation.
We are not even conscious of it, for what we seem to be
conscious of are things like men, trees, and stones.
Suppose we fastened upon resemblance. Yet our diffi
culties would not end, for mere resemblance does not
dependably unite specific universals of one genus. To get
the genus "color," resemblance should unite two specific
universals of color, should it not? Yet it doesn't,
necessarily. Some people have said that a certain tint of
red resembles the sound of a trumpet. Here resemblance
is across qualities. A color resembles a sound. If we
asked such a person, "Which does this tint of red more
resemble, this tint of green or the sound of a trumpet?"
we could expect the reply, "Why, the sound of a trum-
pet." For such a person, the resemblance between this
I red and something else is not another specific universal
color/ but of something like arousal to action.
Mere resemblance is not sufficient to give us this
universal rather: than that.
2. James's Regress Objection Takes Identity in
Abstraction
James argued that the identity theory led to
vicious regress. Bradley replied that he was not
supposing that the identity of the specific tints called
"color" was a separate individual 'somewhat', in the
series, and James seemed to assume that he did. But
James, says Blanshard was not supposing that the identity
was separable, but only that it was logically distinguish
able . Even logical distinguishability is too much for
the identity theory, for the identity sometimes is not
distinguishable.
But having recalled this much of what Bradley
said, let us review the rest. Let us try to get his
view. Now we are trying to analyze a relational complex,
such as the color series united through resemblance. We
get identity in the universal, and difference in the
members. Taken merely as such, as abstract identity and
difference, there is a regress just as James said. Mere
identity has not accounted for the unity of the series;
the differences remain differences. But from that can we
313
conclude that there Is no identity among them? No. Mere j
i
identity is not enough; there is more to the truth of j
i
things. But certainly identity in a relation is as
|
justifiable an abstraction as are mere differences.
Bradley's view on this matter was that the unity
of the differences cannot be rationally explained. But
it does not follow that the series consists of mere
differences, united through a resemblance that is "simple"
and so neither external nor internal to those differences.
We have a dilemma, but we have not solved it by taking
sides with the differences against the identity.
Relational thought cannot avoid treating the
series as a plurality and yet somehow also a unity. We
cannot explain why this is so, we can only point it out.
Identity-in-difference is somehow inherent in the prin
ciple of our mental process. This is what Bradley meant
when he replied to James that the identity and the
difference in a series are not separate parts but in
separable aspects.^ They are internally related.
Regress results whenever the abstract products of
analysis are taken as ultimate constituents.
Blanshard accused Bradley of "imperfect analy*
sis."88 When is analysis ever perfect? When are the
parts enumerated so well that nothing is left out and the
whole is accounted for? Will not the original unity of
the analyzed whole always be left out? Enumerated parts
314
are a plurality but the whole is a unity, and no exercise
in juxtaposition will ever restore it. And it seems j
strange to be criticizing Blanshard on this account. He
expressed quite well the limitations of analysis.^ What
did he think Bradley failed to analyze correctly?
Bradley found that every series, every relational complex,
expresses the divergent aspects of identity and dif
ference. He emphasized that these aspects are in
separable. Is this what he failed to analyze correctly?
Should he have come up with differences plus the relation
of resemblance, which on Blanshard's own principles
cannot be either internal or external because it is
"simple"?
Blanshard applauds James's argument that
discriminable identity among the differences leads to
regress. How could it help it? Bradley knew it, and
Blanshard also knew it. But if analysis is such that it
can produce only abstractions, then why criticize Bradley
because the identity of a series is abstract? His critics
have demanded what analysis cannot produce: a non
abstract abstraction, an indistinguishable distinction.
That seems like an irrational demand.
3. James Makes An Irrational Demand
We may be tempted to agree with Blanshard and
James that it is simpler to just accept resemblance,
315
| rather than to try to understand the mind-boggling notion
of identity-in-difference.^® Even Bradley agreed that it
i was simpler. He added, wryly, that it would be even j
simpler to ignore fundamental problems altogether.^1
"Simple resemblance" seems to suggest that we do just
that. By "simple resemblance" Blanshard meant "resem
blance may be ultimate, and is not resolved in all cases
into partial identity."^2 It seems strange to make
"ultimate" the meaning of "simple’ ." By "simple" we
normally mean either (a) that it has no parts, and is un-
analyzable, or (b) that we understand it easily and with
out effort. But (b) won't do, for "simple resemblance’ ."
The problem that afflicts relations has now been dis
cussed ad nauseam. And (a) won't do either. We have
given our reasons in the discussion of Moore's "simple
indefinable"" By "simple resemblance" our critics must
mean just "ultimate,;" This is where discussion stops.
Here is where reasonable men must declare themselves
to be satisfied.
The unity of differences has been "explained" by
invoking "simple resemblance." All is simple. And if
our opponent persists in raising objections, we can
impose him the burden of producing the unproducible, of
explaining the inexplicable. We can demand an identity
that is not an identity of differences, but is identity
simpliciter. Then, when he fails, we can declare that
316
the resemblance theory has been proved true. But our
procedure is unwarranted. It is unwarranted procedure to
demand that intellectual analysis produce an idea of
identity that is other than abstract. Identity-in-
difference is inexplicable to relational thought. But is
it not more honest, and truer to our experience, to say
so and give our reasons? That is what Bradley did.
E. May No Identity Be Found?
It is unwarranted procedure to demand that
relational thought express itself in terms other than
terms-in-relation: to demand that identity of universals
be made to stand aloof from differences and yet not be
troubled by regress. That can’t be done. The best that
can be done is to point out that wherever there is
identity there is also difference, that these are cor
relative terms, that both of these are abstractions from
what is actual. To speak of identity-in-difference is to
be reminded of this. But perhaps this caution can be
borne in mind. Perhaps it can be remembered that re
lational thought does not stand alone, that it is not
ungrounded in the felt unity of the whole, and so ijs
able to exercise its legitimate function. With that in
mind, then, perhaps we can extend to Blanshard, a
courteous philosopher, the courtesy of answering his
question in the manner in which he asked it. His
317
question was: What common element have the instances of
any generic ore qualitative universal that justifies the
identity theory of universals?
(1) One might answer the question right off by pointing
out that Blanshard himself has already named one thing
all universals have in common: repeatability. Is
repeatability not one and the same quality? Does it, too,
fall victim to the resemblance theory, so that it merely
resembles itself, rather than actually repeats itself?
Does it differ among horse and man and purple and loud?
It hardly seems to, and we would not say that it did
unless we had some theory to defend. Shall we adopt the
principle that just any difference justifies a
distinction?
Blanshard defends his resemblance theory by
distinguishing between the repeatability of specific
universals, and that of qualities and genera. Specific
repeatability ie repeatable, but qualitative and generic
repeatability only resembles itself. This is strange,
for it makes "stone1 ;," "man',;' "purple';," and "loud"— what
people thought were universals— not strictly universals?
but on the other hand "this pink,!' "that loudness ;." "the
prickliness that I feel now"— what people had thought
were particular— are the only genuine universals.
(2) Long ago Plato wished to know what was the identical j
I
"bedness" common to all beds. Blanshard answers that i
i
there is none. 3 There is nothing identical among all I
beds that could qualify as a generic universal. These
do not even exist in rerUm natura, he says, but are only ;
the work of the mind. Only specific universals exist
in rerum natura. This is shocking. No horses, stones,
or men? No colors, sounds, nor textures? Is it truly so
difficult to find such identities?
The question is, "What, if anything, is common
to all beds?" But first we must understand the
question, for it is no refutation of our theory that we
cannot answer a meaningless question. What is meant by
"bed"? Perhaps it does not mean a flower bed nor a road
bed but a man's bed. In that case some common element
is not so difficult to find. We can find at least three.
All beds must have (a) a size near to that of a man,
(b) a shape that will accommodate a man, and (c) a
surface comfortable enough to lie upon for rest. That
excludes most things in the universe. And yet such
otherwise differing things as an army cot, a sultan's
divan, a nest of leaves, and a sandy hollow under a
bridge all qualify as beds under the identity theory.
Shall it be objected that these three things must be made
one? No, what was wanted was one thing common to all,
and any of the three will do. It would be even easier
319
to find something common to all if use could be appealed
to. And if "repeatability" will serve an identity
theory among specific universals, then it is hard to see
why "usability" will not serve for a large number of
other universals, such as beds.
(3) The principle on which Blanshard objects to the
identity theory for generic and qualitative universals
is that we cannot "verify" or "isolate even conceptually
an invariant character" common to all. "One cannot even
be sure of the species from which one is supposed to
extract it," he says.^ The problem is that qualities
and genera overlap, so that one can't be sure when one
is dealing with universal A or universal B. But why is
this a difficulty? The practical problem of distinguish
ing A from B is irrelevant. The problem as stated is,
"Given instances of A, or B, can anything be found
common among those instances? Can it? Why, yes.
"Locomotion" is common to animals, "Breathes carbon
dioxide" is common to plants. "Has a velocity of 1100
feet per second at sea level" is characteristic of
sound. "Within the wave frequency of visible light" is
characteristic of color. "Cat" can be identified by
chromosome structure.
In the case of "color" we have referred to a
range of specific values, and so, it might be objected,
320
one range has been identified by referring to another.
But the range falls within clear limits, the values are
measurable, and within that range a definite mathematical
formula applies: by its means these appearances can be
predicted and controlled. That gives an identity among
them.
It is true that certain small organisms exhibit
traits of both A, animal and B, plant. The universals A
and B overlap within the range of their instances. What
of it? How does the identity theory forbid that two
universals should overlap so that one instance might
exhibit both? Any instance exhibits more than one
universal. It might be useful to refer to a hybrid
uinversal AB, and then go on to ascribe to it a uni
versality of its own, but what of it? The only thing to
forbid universals to overlap would be that one instance
AK
should not exhibit contradictory characters. J But in
that case, where are the instances? Such an example has
not been produced by any opponent of the identity theory.
F. Is A Qualitative or Generic
Universal Never Experienced?
Blanshard has denied qualitative and generic
universals to immediate experience. Perhaps he is right.
But what he meant was that there aren't any such
universals, except through the relation of resemblance.
"These cloudy and elusive entities" (he says, referring
321
to qualitative and generic universals), "so indefinite
as to arouse grave doubts as to their conceivability,
46
have no place in nature." He is talking about qualitie
like color, heat, and tone; about things like men, trees,
and stones. Is there no such universal in experience
except as mediated through resemblance?
1. Meaningful Experience Requires Universals
No such identity can be found, he said. But in
the preceding section it was claimed, to the contrary,
that such could be found. Sometimes, as in the case of
color, it was necessary to appeal to the identity of a
range of instances taken under a scientific formula.
That may have aroused suspicion. Even though some range
may be used, precisely defined, and its instances sub
sumed under formulas for wave propagation and so forth,
still the specific colors do differ among themselves.
It is they that exist. If so, then invariant identity
is not among them and the identity theory fails. Color
is typical of those universals in which the identity is
indirect and only inferential. But is it not experi
enced at all?
We must argue that all generic and qualitative
universals are actually experienced— to the degree that
anything interpreted and meaningful can be called
"experienced." Is that which is passed through the
322
reflective thinking process experienced? One may raise
the question whether anything taken in thought as terms- j
I
in-relation, subsumed under universals as identity-in- i
difference, can be called "experienced.?1 But can we say
that the product of the reflective process is not
experienced? We need not grapp^ with these questions
here. If nothing taken as universal— taken in
reflection— is to be called "experiencedf” then of
course specific universals aren't experienced either.
A difference has been thrust between "specific" and
"universal i I t may be true that the whole experience
is not to be reduced to relations among universals, but
it is something else to deny those universals. The
relation aRb is nothing actual, but on the other hand the
actual is such that aRb:. The relation, the universal,
is somehow there. Surely Blanshard, who elsewhere
argued for universal necessity, will not now reverse
himself and tell us that specific tints of color are
only arbitrarily connected through "color." Are they
necessarily connected through some universal other than
"color"?
Aren't men, stones, trees, sounds, and colors
experienced as truly as are this loudness, that tint of
pink, or such-and-such a roughness? "Appearances" they
may be, as Bradley called them, but they do appear. The
difficulty seems to be in finding some invariable some-
323
thing among them. Perhaps we looked in the wrong
place. Why should the universal be among the specifics— i
I
or even in space at all? The specifics may appear in
i
space, but where does space appear? In space? The
universal is not specific, and so it is not in space.
The universal does not need space. But space needs
universals. It ie a universal. Then why should not the
universals be experienced in a way of their own?
2. Universals Endure Change of Resemblance Among
Specifics
Consider the difficulties that would assail us
if we seriously denied true universals. Blanshard
denied them. With only specific universals truly
repeatable, then the generic and qualitative universals
have got to be a down-graded and derivative sort of make
shift universal. The defining trait of universality,
that of identical repeatability, is not among them. But
suppose that we went looking for universals among the
specifics, in space. For example, where, in this stadium
filled with 50,000 people, are the true universals?
We would have thought that there were true universals
like humanity, rationality, two-leggedness and so on,
repeated 50,000 times. But no. Then what truly does
possess repeatability so that it can qualify un
reservedly for universality? Only specific universals.
324
They are there, and innumerable, but none of them need
actually be repeated even once. It is enough that they
be repeat-able. The specific redness of this man's hair,
though it is not repeated once in the whole stadium,
is the sort of universal that is genuine. The 50,000
instances of humanity do not quite qualify, but this tint
of red does qualify, although it is not repeated even
once— under the repeatability criterion. This seems
strange. We can hardly resist saying that it is more
strange, even, than the mind-boggling notion of
identity-in-difference.
Let us keep to normal experience and pursue the
question. We are looking for "color" in common examples,
having been told that it was not there. Only the
specific tints are there, with their simple resemblance.
Now suppose that we have here two cans of pink
paint. Beside them are two more, one red and one green.
The colors in all cans are specific. "Color" can be got
only from resemblance among them, and if there is more
resemblance, then there is more color. But shall we say,
then, that the two cans of pink paint, which could hardly
resemble each other more, are more colored than the red
or the green? After all, resemblance is resemblance
pure and simple, and it is not to be qualified as to
kind. Yet isn't the result absurd? Who would say that
the pink is more colored than the red or the green?
Surely color is there along with the pink, the red, and
the green, and resemblance has nothing to do with the
matter.
Suppose that a painter has contracted to paint a
room pink, but his customer has changed her mind. Now
she wants grey. Her husband dislikes pink, but he can
endure grey, and besides, grey doesn't show dirt. Her
mind is made up: she must have grey. What will the poor
painter do? If he goes back to the store with the pink,
to trade it for grey/ he will lose time. If he charges
the lady extra for the time, he may lose his customer.
But he is prepared. He takes from his truck a tube of
chrome green and stirs it into the pink. Green, he knows,
will 'kill' pink. The paint turns grey. The lady is
happy; the painter is happy. But look at the unhappy
consequence for the resemblance theory: one specific
universal has annihilated the other; the green has
killed the pink. In doing so it committed suicide.
Now all is grey. First they were there but now they are
gone; a sinister stranger has taken their place. Yet
"color" was present throughout, was it not? The change
of resemblance could hardly have been more dramatic, yet
the qualitative universal, color, was serenely present
all along. It did not leave, with the specifics. It
only changed partners. It changed only its appearance,
not its presence. Ignore, if you like, the scientific
law that "explains" all this; but notice, as you must,
that what endures through change must be more real than
what changes. Color was there all along.
G. Is There No Problem Over Specific
Universals?
The resemblance theory for generic and quali
tative universals rests upon an identity theory for
specific universals. They are genuinely, identically
repeatable. They need not actually be repeated, it is
sufficient that they "could be" repeated.^ They com
prise everything we actually experience objectively.
"Our immediate experience is of nothing but specific
universals," says Blanshard, with emphasis upon the
"nothing butV^® They exist in rerum natura. Other
universals are only the work of our mind. Specific
universals are precise, he adds, unlike generic and
qualitative universals.^ They are as precise as this
blueness, that threeness, the roughness that I now feel.
Yet that does not make them particular, he says, for
"There is no reason in principle," why they should not
have their "duplicate" elsewhere.5 *® There is no reason
in principle why this blue should be unique. The only
condition under which "this blue" should be taken in a
sense that is unique, according to Blanshard, would be
thht it were taken together with its relations to every
thing else in the universe. But in that case all that
327
has been asserted is that the universe is unique— which i
1 i
is true by definition. But there are problems here. i
t
1. It is False that Experience Is of Nothing but
Specific Universals
The universe is not only unique in itself and on
the whole, by definition, but every One of its internal
constituents is unique if taken as it actually is, with
all its internal relations included. "This blue if
taken as it is and not as an abstraction, is no
exception. This blue, since it i£ actual, brings its
conditions of actuality along with it? and when it is
asserted as actual we imply that those conditions are all
known to be satisfied. Thus we do assert the universe
whenever we assert "this blue That makes the blue
unique. That is why we spoke of this blue, rather than
that blue or some other blue or just blue in general.
By "this blue" we meant to assert the particular, even
though we did not succeed. Of course we don't know all
the conditions in the universe, but we do mean to
assert them— and this is the reason in principle why
"this blue" should be unique.
It just is not true that "our immediate
experience is of nothing but specific universals." The
proposition is false. What of feeling? Is that not
experienced? Is it not an objective experience? What of
328
the feeling of unclarity that Blanshard confessed over J
i
Bradley's identity-in-difference? Feeling, it seems, !
enters even into a philosopher's determination of what isj
and what is not a valid point of argument. We do not
challenge his right to have and use his feelings. What
we insist upon is that he include his feeling in his
experienced universe. It has its right to be there.
Feeling may be subsumed under universals, as diverse
feelings (in the plural), but feeling itself is not a set
of any kind of universals. Feeling presents the unique,
and-we sometimes do mean to assert it in our universals.
(See Chapter I, Sections F, and G.)
2. Merely Possible Repetition Makes No Universal
Consider again the specific universals. "There
is no reason in principle" why any of them should not be
repeated elsewhere. That makes them universal. They
need not actually be repeated, it is enough that they
could be repeated— but the "could be" presents a
difficulty. Repetition ex hypothesi makes for uni
versality; and if actual experience is of universals, it
must be that they actually are repeated. "Could be" is
not enough. It means only that something possibly is
repeated, and "possible" means only "not known to be
impossible.But it does not exclude "impossible,"
"possible" and "impossible" are not contradictories. J
329
Merely possible repetition makes no actual
repetition, for it does not exclude the contradictory. . j
It does not exclude "no repetition"; it does not esclude j
particularity. Nevertheless, someone might reply, the
repetition is a real possibility. Then certain con
ditions would actually produce it. But its repetition is I
still hypothetical, not actual. The addition of "real"
to "possibility" only asserts that there are known to
be regularities in the universe which would result in
repetition if they were all satisfied, and with "real"
we mean to say that they are known to be in part
satisfied.^ Yet it could be otherwise, and so the
repetition of the specific universal is thus no more
than ideal or mental. Blanshard says of the qualitative
and generic universal, "neither of these exists outside
C C
our thought." But specific universals are no
exception.
H. Summary of Chapter VI
Blanshard's belief that immediate experience is
of nothing but specific universals, united artificially
to yield other universals through the relation of
resemblance, would, if true, beafatal to Bradley's view.
It would pulverize the unity of Bradley's reality into a
vast swarm of logical atoms, united by an external
relation that is all-pervasive. What justification does
330
Blanshard give for his belief? He relies on the
correctness of James's arguments against the identity
theory for generic and qualitative universals. But
James's argument that there is no identity among such
universals is logically faulty. His attempt to convict
the identity theory of vicious regress makes the
irrational demand that analysis should be carried out
without disrupting the unity of the original whole with
which it started. Furthermore, Blanshard's stipulation
that repeatability is the quality that accounts for
universality is itself a universal of the sort whose
existence he had denied. Beyond even that, it is false
that no identity among qualitative and generic
universals can be found. It is false that experience
consists of nothing but specific universals. It is false
that the specific universal occupies a privileged
position, for possible repetition makes nothing universal.
331
FOOTNOTES— CHAPTER VI
Russell's phrase. See My Philosophical
Development (New York: Simon and Schuster, 1959),
p. 54.
o
Brand Blanshard, Reason and Analysis
(La Salle, 111.: Open Court Pub. Co., 1964), p. 411.
Hereafter cited as R&A.
F. H. Bradley, The Principles of Logic,
2nd ed. (1928; rpt. London: Oxford, U. Pr., 1967),
1:36-7. Hereafter cited as Logic.
^See Appendix C, "Sameness and Differencei"
5Blanshard, R&A, p. 390.
6Ibid., p. 399.
7
Ibid., see also p. 402.
8Ibid., p. 405.
9Ibid., p. 409.
^0Blanshard does not say "Any specific universal
could exist elsewhere in other contexts," but he does
say, "Every sensation . . . [or "non-sensible
characters such as numbers"] . . . has a quality that in
principle could be repeated and hence is a unlversaT
[my emphasis]. See R&A, p. 399, once again.
"Repeatable" entails "universal." What does
"repeatable" mean? It does not mean "occurs again in
the same context;" because sameness means qualitative
sameness [See R&A, 395-8]. If all qualities of the so-
called "repeated" occurrence are the same, then we do
not have two occurrences; we have only one. Hence,
"repeatable" must mean that the specific universal can
recur in a different context, itself unchanged.
Ibid., p. 399.
^■2Note that Blanshard carefully limits himself
with a "could be." The specific universal "could be"
repeated. He doesn't say, "is actually repeated." But
could be repeated says only that it possibly is
332
repeated. Does possible repetition guarantee an actual
universal? Don't we need an actual is repeated?
■^Ibid., pp. 395-8. The point here is that if
we are justified in saying that any specific universal
is repeated, the two instances must differ qualitatively,
it is important that we see this, and elsewhere, for
another purpose, Blanshard points it out. That means,
if a quality is repeated, its relations must change; and
if a relation is repeated, the quality of its terms must
change. Something must support the distinction between
instance A and instance B.
Brand Blanshard, The Nature of Thought,
2 vols. (1939; rpt. New York: Humanities Pr. Inc.,
1969), 2:516.
•^Blanshard, R&A, p. 490.
iD". . . able to exist elsewhere in other
contexts ..." These are my words, not Blanshard's,
but they are justifiable. See above, where I have
discussed "repeatable . "
Let us also discuss this suffix, "— able. . ' ! 1 it
refers us to possible conditions of actuality.
Blanshard makes "repeatable" the condition of actuality
for specific universals, but it means only "possibly
repeated", or "repeated, under certain conditions." But
those conditions are unspecified, and there is no way to
know whether they ever will be satisfied, or not. But
if they are not satisfied, then the specific universal
will not be repeated. Thus "repeatable" does not exclude
"but intruth not repeatable."
Blanshard, like Moore, has resorted to the
conditional tense of a verb, and to the appeal to mere
possibility. But if we have forgotten the lesson of
Moore's error, let us learn it here. Possibility does
not exclude impossibility. We use "possible" only when
we are ignorant of the truth. We say, "possibly is
repeated," or "could be repeated" meaning, "so far as we
know the necessary conditions might be met" in the sense
that "I know of no reason not to think so." Still, it
might not be possible, and in truth it either is or
isn *t.
In reality there may be no way that this or that
"specific universal" could be repeated. Therefore we
cannot say of it that it could be repeated, with any
assurance.
■^Blanshard, R&A, p. 415.
333
18Ibid., p. 402.
18Ibid., p. 399.
28Ibid., p. 401.
21
And surely I need not repeat again that
"appearing again” means that some difference of context
must accompany that second appearance, if the "again"
is significant. As Blanshard himself said, numerical
difference is no difference without qualitative
difference. Again, see R&A, pp. 395-8.
22Bradley, Logic, 1:45.
23F. h. Bradley, Appearance and Reality, 2nd ed.
(1893; rpt. Oxford: Clarendon Pr., 1966), pp. 353-4,
149. See ETR, pp. 296-9.
^Blanshard, R&A, pp. 409-11.
23Ibid., pp. 409-14. We will compare Blanshard'
appraisal with the originals as reprinted in James's
Principles of Psychology and Bradley's Collected Essays.
28William James, The Principles of Psychology,
2 vols. (1890; rpt. New York; Dover Pub. Inc., 1950)
1:532. Hereafter cited as Principles.
27Ibid., pp. 493-4.
2 8
James, Principles, 1:532. See also p. 533n
and Bradley, CE, 1:290. James's regress argument was
originated by Stumpf..
2 9
This third argument was omitted from the 1890
edition of James's Principles, but it is mentioned by
both Blanshard (R&A, pp. 412-3) and Bradley (CE
1:296-7).
28Bradley, ibid., p. 299.
31Blanshard, R&A, p. 409.
32Ibid., p. 395.
33This was not quite Bradley's reply. He spoke
of identity as a "general character" of which he was
"vaguely aware ." See Bradley, CE, 1:289.
•^Blanshard, R&A, p. 411. j
j
35Bradley, CE, 1:288. I
j
•*®Why should we care about the opinion of someone j
who thought that red was like the sound of a trumpet?
Because, according to James himself, such a person is a
genius. He said, "Some people are far more sensitive to
resemblances . . . than others are . . . A native
talent for perceiving analogies . . . [isl I ~the
leading fact in genius of every order." See James,
Principles, pp. 299-300. The emphasis upon these words
is not mine, but. Jama's own.
"^Bradley, CE 1:292.
•^Blanshard, R&A, p. 411.
■^See ibid., Chapter IV.
4®see Blanshard, R&A, p. 412.
41Biradley, CE 1:299.
43Blanshard, R&A, p. 409.
43Blanshard, R&A, pp. 404-05.
44Ibid., p. 406.
430n Bradley's view universals do contradict in
the sense that they are both A, an identity, and B,
exhibited in different instances. There is identity in
difference. This is not strictly a contradiction but
rather an "ultimate inexplicable> But it requires that
we refer to degree of reality, and that the universal be
recognized as more real than the instances. That is
because something is more real than something else if
in its meaning it includes the other.
4®Blanshard, R&A, p. 416.
4^Ibid., p. 339.
49Ibid., p. 419.
^9Ibid., p. 401.
51Ibid.
335
^Bradley, Logic 2:203.
53Ibid., 2:668.
54Ibid.r 1:208-09.
55Blanshard, R&A, p. 402.
CHAPTER VII
CONCLUSION
Our purpose in writing this essay was to
determine whether or not Bradley's criticism of "the
relational view of things" ever was refuted. We must
conclude that it never Was refuted. It was believed to
have been refuted, but those who supposedly refuted it
have largely failed even to understand what Bradley was
saying.
We had wished to find out if Bradley's view had
been refuted: and our further aim, if we succeeded in
showing that it had not been, was to gather evidence to
support the stronger conclusion that he was right. This
we have done, through our discussion of the defenses
Bradley himself would have used in reply to his critics.
We will therefore conclude not only that Bradley's
critique of relations was never refuted, but also that
he was right. Reality transcends our world; our
propositions about it fall somewhere short of the truth.
We offer this carefully considered conclusion in the
same courteous sense in which Bradley himself would have
offered it: with the invitation, to our critics, to
Hhow us precisely where and how we have erred. Until
337
that is done, we do not think it is too bold to presume
that we have not erred.
Our effort throughout this essay has been
largely confined to answering criticism. But now we will
try to summarize our conclusion in one positive propo
sition. What has our inquiry taught us? We have learned
that Bradley's argument must be allowed to stand: the
relational view gives appearance only, not reality. In
finding that out we have also seen that the terms and
relations used by the intellectual process are the
products of abstraction from an original and prior whole
of being. They are dependent upon that original and
prior whole for their meaning, and for the truth of their
conjunction if we relate them, in our propositions. Then
let this be the proposition that summarizes our con
clusion: The products of analysis derive their meaning
from the whole.
A. A Tu Quoque Justification of
Our Conclusion
Let us try to ward off criticism of our
conclusion. It may be objected that this concluding
proposition is curious; for by "the whole" we mean the
whole reality as we take it, so the "meaning" of our
proposition, if it is consistent with itself, depends
upon our taking of reality as we do. In other words, it
is meaningful and true only to someone who is already
convinced of it. How, then, can we recommend it to
someone taking an opposing view? And, he will object,
even we cannot understand this proposition in its full
meaning and truth, for by our own principle reality is
never fully known, Our own conclusion is partly
meaningless and therefore false. Its full meaning is not
before us, for it is conditioned by the unknown. If it
were known, that meaning would be changed, and so what
we believe would be changed from what it now is. So
what we now believe is false. Our conclusion, therefore,
is false.
We must reply. The meaning and truth of our
concluding proposition presumes familiarity with all that
has gone before in this essay. We have shown the un-
tenability of our critics' position. The critic must
have some positive and consistent basis from which to
deny our claim, and he has none. Further, although the
meaning and truth of our proposition may not be self-
evident to everyone— the meaning and truth of the
proposition is not that mere proposition— yet we claim
to have shown its meaning and truth to anyone who has
followed our discussion.
We found that our opposition came mainly from two
camps: Moore with his followers, and Russell with his
followers. The first group is represented by linguistic
philosophy, of whom Austin and Malcolm are representative
339
in this essay.1 Wittgenstein, too, who owed much to
Moore and Russell for the influence he later exercised,
was a foremost spokesman for linguistic philosophy.
But Wittgenstein, having reached Das Unausprechliche
at the end of his Tractatus, threw his logical ladder
2
down. He later widened that devastating conclusion,
declaring all logical arguments to be nonsense. The
truth, it seems, became "unexpressible."
The second opposing camp is represented in this
essay by Bergmann. Although he too took both Moore and
Russell for his masters, his efforts, a.la Russell, were
bent toward displaying the logical forms of true state
ments. And yet, as we saw, those forms themselves fell
into a plurality of orders, the relations among which
were again new forms of their own that turned out, in the
end, to be "ineffable." And yet the relations among those
orders of forms among themselves, and to reality, were
precisely what was needed to call them true. Otherwise
they did not even exist, but subsisted.
And so, the truth of the propositions maintained
by our critics against us was either "unexpressible" or
"ineffable" Then upon what positive and meaningful
basis might they be used to deny our claim? If they
claim to have "shown" the truth of their propositions,
at least they have not managed to say it— and so cannot
claim to have demonstrated it. Their truth claim rests,
340
it seems, on some kind of intuitive insight. After
reading a while one simply "sees" that Moore, or
Russell, or Wittgenstein, or Bergmann, was right. But
then, we 'see' that Bradley was right. Is our insight
worse off then theirs? We, at least, do not see it so.
The matter has now gone beyond the reach of argument—
or, if there is any longer an argument, it is the tu
quoque argument. Because the opposition cannot
demonstrate their truth, we are free to claim it.
And we do claim it. We have critically discussed
relational thought, with its limitations. We have seen
that it begins with subrational feeling, and is abandoned
in the end for supra-rational reality. Relational
thought is a mean between the two. In feeling it is
grounded; toward reality it reaches for its directing
purpose. Our mere ideas are not reality for they are
signs of reality."*
B. The Superior Satisfactoriness Of
Our Conclusion
The tu quoque justification leaves us dis
satisfied. A gentlemanly agreement to disagree on
fundamental principles leaves us with a feeling of
futility. Are we free to go on uttering our falsehoods
because our opponents, too, are liars? Is there no way
to strengthen our own position at the expense of the
opposition? Let us try to find a way.
341
If we ourselves are dissatisfied with our tu
quoque justification then we cannot rest content with it,
for we have insisted upon satisfaction as the touchstone
of truth from the first. We must try for satisfaction.
Then let us first ask, "What is the purpose in thinking
at all?" In Bradley’s words it is "to get rid of mere
conjunctions in the soul."^ That is to say, we seek to
understand, which is, to become aware of rthe plurality of
appearances in a way that brings them into a unity; to
'see' beneath the apparently accidental conjunctions
among things a connection that is necessary. Experi
enced Content is richly varied. In it we find terms to
relate. These are "mere conjunctions" to us at first
because the "how" and "why" of their coming together
C
escapes us. The missing "how" and "why" leaves us
feeling dissatisfied. By thinking we seek to change
that troublesome state: to find out the "how" and the
"why ' . . " Thus science has arisen, and with such success
as to leave little doubt that here thought finds its
rightful function.
And yet, with all our self-satisfaction, can we
say finally that we have reached the truth? In reaching
the whole truth do we not mean to have reached reality—
indeed, to have become one with reality? For, if our
'truth' fell short of reality, then would it not fall
short of what was truer still than merely what we have
342
sometimes reach intellectual satisfaction. To that
degree we feel justified in calling our newly discovered
proposition "true "!'
The truth is unreachable in the sense that human
finitude and thought's own nature limit its fullest
development. This much we do concede to our opposition's
use of the words "unexpressible" and "ineffable" ' . 1 And
yet we do claim that our own concluding proposition is
truer by far than any which deny it. By saying that
"The products of analysis derive their meaning from the
whole" we mean to say that the truth of this and every
other proposition is dependent upon the degree to which
7
reality is taken as included. We do not claim that thxs
or any other proposition is absolutely true in and of
itself. But if our critic still finds our conclusion
unsatisfactory to some degree, let us ask him to notice
our opposition's claims. Let our critic ask himself
whether they are not much worse than ours.
The opposition claims that its basic propositions
are true. It must do so in order to deny our claim.
Thus, for the opposition, some propositions are true,
and known to be true. Moreover, the opposition claims to
have "described" in a better way than we have and so,
we presume, reality is what has been superiorily
described. But let us notice that the opposition regards
the truth or falsity of a proposition (a genuine,
344
attained? the relation between our 'truth' and reality
would remain unexplained. Thought seeks to dissolve all
merely external relations, into a superior identity. Yet
it cannot, utterly. Thought, by its nature, is an
effort to say of some "this" that it is also a "what."
The "this" is its subject, the "what" is the predicate.
And their coming together, so long as merely thought,
remains ideal only. The actual coming together is
forbidden so long as mere thought is the mean between
them. Thought, by its nature, asserts a separation
between "this" and "what." Thought seeks, to be
sure,
an arrangement of ideas, self-consistent and
complete; and by this predicate it has to
qualify and make good the Reality.
Still,
Truth should mean what it stands for, and stand
for what it means; but these two aspects in the
end prove incompatible. There is still a
difference, unremoved, between the subject
and the predicate, a difference which, if
removed, would wholly destroy the special essence
of thinking.
We seek the union of "this" and "what," but thought must
preserve its distinctions while supposing they are over
come. Thus, in thought the union must remain merely
ideal. And yet even this merely ideal union, when
achieved by thought, is marked by a special sense of
satisfaction. Like Archimedes shouting "Eureka!" we do
344
synthetic, informative proposition) as something
external to that proposition, even though possessed
(somehow) by it as a property. The proposition need not
be true in order to be the proposition that it is.
Neither need the proposition be known to be true in order
to be true. And reality, of which the proposition is
true (if it happens to be true), is what it is
irrespective of whether any propositions about it are
known or not. From this we conclude that (a) truth is
never a necessary property of any informative propo
sition, because (b) it merely copies, or corresponds to,
or describes something else.
Truth is something a proposition may or may not
have, indifferently to itself. Its truth is somehow
dependent upon an external relation between the propo-*
sition and something else. That relation is sometimes
called "correspondence,/1 "copying,," or "description1 ’. "
These notions are fundamental to our opposition's use of
the word "true"
But the difficulties besetting the opposition's
view of truth are formidable. Although the opposition
treats the truth of some proposition as though it were
a property of the proposition, yet how can it be? For
truth is imputed to the bare proposition only on
condition of some external relation between the propo
sition and something else. The proposition itself is
345
neither true nor false except through the agency of
some tertiam quid whose presence is neither necessary nor
accounted for. That presence is largely ignored, although
belatedly noticed and suggested through such words as
"unexpressible" and "ineffable."
But even ignoring that, there are difficulties
about the "something else" to which the proposition is
related for its truth. The "something else';," we must
presume, is also present in the proposition in order for
words like "correspondence1 ," "description," or "copy" to
be warranted. But (1) what "something else" is essenti
ally future, or past, and yet present in some proposition
about them? A proposition copies, or corresponds to,
or describes in a way warranting saying "is true of",
something else called a "state of affairs." But the
proposition surely exists and the past or future state of
affairs surely does not. Or how, and in what sense, can
anything in a bare proposition correspond to a feeling?
Ex hypothesi the bare proposition is something objective,
but a feeling is something subjective. (2) How much of
the copied state of affairs is the work of the mind?
(3) Does the proposition copy the whole state of affairs?
How can it do less? (4) What state of affairs corres
ponds to a disjunctive, hypothetical, or negative
proposition? They are not all false. (5) The most
desirable kind of truth is an explanation. But an
346
explanation does not copy sensuous perception, nor even
past and future events. What corresponds to, or is
described by, Boyle's Law, or Quantum Theory, or the
Coefficient of Friction?
The main error or the opposition's use of "true"
is to treat knowledge as though it were piecemeal, to
treat truth as though it were something divisible from
knowledge— and even from propositions, and to treat know
ledge and truth as though they were other than reality.
These discrepant parts cannot then be reunited except
through an external relation that must remain "unexpres
sible" or "ineffable." But "truth" surely implies that
they are united, and known to be united.
Our opposition, then, has difficulties of its
own. Those who follow the linguistic road recognize
this and have already given up the game. The truth of
things lies beyond philosophy altogether, and the
"grammar" of linguistic expression alone concerns it.
Solipsism, it turns out, is the truth of our intellectual
predicament— but even that cannot truly be said. The
truth of things, for philosophers, is radically
"unausprechliche I ' 1 . ' But for those who follow the logical
road the case is no better. Again the forms of
propositions concern them, but in order to escape
paradox they must distinguish between the forms of
"things" and the forms of propositions which express
347
them. But the: relation between those kinds of forms, ]
i
of propositions versus states of affairs or "things"— a
!
relation of copying or correspondence or description—
which must be known if the proposition is true, is
"ineffable.".'
And so, if someone started out along the
opposition road, reassured to be told that common sense
propositions, and "patent facts," were utterly to be
trusted, he was in for a shock. For, although the
propositions of common sense and patent fact were taken
as absolutely true, yet the relation of them to the
knower, by which they were known to be true, turned out
to be absolutely "unexpressible" and "ineffable." And
with this result, we submit, absolute satisfaction
gives way to utter dismay.
Our dismay is not lessened when we hear that
knowing that is logically independent of knowing how.
If we do not know how the ineffable and unexpressible
relation of correspondence, or copying, or description,
maintains its function of coming between the knower, with
his proposition, and the object known, then we have no
right to presume that this relation leaves unaffected
the "what" of the "that" which is purportedly known.
Our opposition (at least sometimes) seems to presume that
this relation is external. But the whole point of this
essay was to show that presumption to be indefensible.
348
What is wanted is a better theory of truth, and
this Bradley sought to provide with his doctrine of
degree of truth and reality. It is implicit in our
concluding proposition, "The products of analysis derive
p
their meaning from the whole." To the degree that any
proposition is satisfying, and so far true, the whole of
our reality is present to lend meaning and not to clash.
So far as we call the proposition "true,'.' reality is
coherent and in order. Thus there is, so far as the
proposition is true, no separation between it and re
ality. The merely external relation, whatever we may
have called it, has been transcended.
It follows from our view that no proposition is
wholly true, for reality remains in part unknown; the
unknown would and does modify the meaning. Moreover, the
"this" and the "what" are only ideally connected. And
conversely, no proposition is wholly false. Let our
opposition take comfort. For surely any proposition has
some meaning, to him who utters it, intended to qualify
reality in a way that is true. The proposition presents
some aspect of reality, however deficient it may be in
those further connections our satisfaction craves. But
the difference between them, "true" and "falseis
precisely that "presence of connections" our opposition
rejected but which we have defended. To the degree that
349
those connections are felt or understood to be present
the proposition is, so far, true.
Our meaning for "true" does agree with common
sense and ordinary usage. Those of us with any respect
for common sense, and for that term as preserved in
ordinary usage, appeal to it in support of judgments
deemed important, not trivial. As is the case with any
judgment, one which deserves the employment of our common
sense must answer some genuine question. Thus we never
say, "It's just common sense that stones are hard."
Why bother to invoke the authority of common sense? Who
cares what some other people think about such matters?
No, when we invoke common sense we do so in a statement
like, "It's just common sense to vote for Jones rather
than Smith." We mean to add, by the invocation of
common sense, "other sensible people think so too." It
is "common" in being widely shared among people who have
"sense/" that is to say, sound judgment. In judgments
of trivial matters we trust our own. Whoever may
g
believe that stones are soft is not wrong, he's crazy.
We appeal to common sense, as to other authority, when
our judgment needs support. And often it does need
support. Ordinarily, sensible people readily admit that
their knowledge isn't total, nor absolute, on any subject;
at all. When struggling to decide some important
question they often say that one answer seems truer than ;
350
another— although they would admit some truth to the
alternative. They say, for example, that Jones speaks
more of the truth than Smith. Thus, even though he
falls short of their ideal, he deserves their vote. He
is more in harmony with the world of things as they
really are. By "world” people mean that coherent
structure which they have learned, out of years of
agonizing reappraisal, to regard as real. Although
common language is somewhat inconsistent, implying at
times that all judgment divides between the absolutely
true and the utterly false, this is because people must
decide either to act or not to act. But in their
reflective moments their speech is filled with allusions
to the relativity of truth and error. Their speech
thus implies degrees of truth, especially in those
judgments that matter most.
Bradley's doctrine of degrees of truth and
reality was formulated to meet the requirements of
theoretical consistency. But it is also consistent with
common sense— unless by this term we mean some
philosophical doctrine of false absolutism neither
widely believed nor preserved in general usage. The
doctrine of degrees of truth and reality was referred
by Edward Caird as "the greatest thing since Kant."
It was offered to meet the difficulty over relations
discussed throughout this essay. It was offered to
351
escape the difficulty of equating truth to mere
practice. It was offered to escape the difficulties
of atomistic absolutism that regards truth as a con
tingent property of some propositions but leaves no
place in philosophy for anything deserving the name
"implication V"
FOOTNOTES— CHAPTER VII
Our reasons for calling the heritage from
Moore "linguistic philosophy" is by now, we hope, well
enough known.
2
Wittgenstein, Ludwig, Tratatus Logico-
Philosiphicus (New York: Harcourt, Brace & Co., 1922),
p. 189 (654) .
O
JSee Appendix A.
^F. H. Bradley, Appearance and Reality, 2nd ed.
(Oxford: Clarendon Pr., 1897), p. 315. Hereafter cited
as AR.
5
See Bradley, Essays on Truth and Reality
(Oxford: Clarendon Pr~ 1914), p. 273. Hereafter cited
as ETR.
^Bradley, AR, p. 319.
7 See Bradley, ETR, p. 430, AR, p. 321.
O
At this point our opponents will demand of us
as we asked of Blanshard and James: "Degree? Degree of
what? What one thing have all propositions in common
that may be had in degree?" And we answer in a word,
"Reality It is not ourselves but our opponents who
made propositions into discrete entities, independent of
being known and of the reality known. It is our
opponents, not ourselves, who denied that reality is one
and whole. But the positive basis supporting their
denial does not seem to have been made plain, and to
accept the claim merely as a working presumption leads
to insuperable difficulties.
^In the sense that he lives in a world of his own.
353
APPENDIX A
WHAT IS AN IDEA?
Bradley gave his answer to the question by first
pointing out that philosophy is concerned with the
activity of judging. We do not have a judgment unless we
claim that something is either true or false. To do
that, something has got to be related to something else,
but because actual reality is not true or false, the
"something" that we relate in judgment has got to be an
aspect, abstraction, or appearance of reality: we call
such things "ideas;'" So, the first thing we can say,
about ideas, is
(1) Ideas are abstract aspects of reality, used
in judgment.
But furthermore, we cannot judge until"’we use
ideas as ideas. This means that we must become aware
that they are mere ideas: that the reality from which
they borrow their meaning lies beyond them. They are
mere signs: signs of an existence other than their
own. So,
(2) Ideas are signs of existence other than
their own.
Now ideas do exist, although just what they are
354
is difficult to verify, for they are not their objects;
but when we try to verify and examine them they must
become objects. But so far as we can tell, ideas
exist as states of mind, or phenomena, or psychical
events. Let us speak of their existence as a Vpsychic
event". This term, of course, begs the next question,
but we can ignore saying just what a "psychic event” is,
because we don't need to know. We do not use the idea so
far as it i£ a psychic event: we use the idea for what it
means. We do not use the idea because of its own
existence, we use the signification of the idea as it
refers to another existence. So
(3) Ideas, as signs, are what they mean.
But what is a sign?1 An idea, when used, refers
away from itself. It is a sign. It is an idea of some
existence. A sign has three aspects: (a) its own
existence, (b) its own character, and (cj what it means.
It (a) exists, and so (b) it has its character; a content
distinguishable from other contents. But (a) and (b) do
not distinguish the sign from anything else. Everything
has existence and content. A sign is a class of
existents with a third and distinguishing side: it has
a meaning. A sign is an existent with a meaning. Brad
ley said,
For logic, at least, all ideas are signs. Each we
know exists as a psychical fact, and with particular
qualities and relations. It has its specialty as an
355
event in ray mind. It is a hard individual, so
unique that it not only differs from all the
others, but even from itself at subsequent
moments. And this character it must bear when
confined to the two aspects of existence and
content.
But such an idea becomes of interest for philosophy only
when
It begins to exist for the sake of its meaning. And
its meaning, we may repeat, is a part of the content,
used without regard to the rest, or the
existence. . . . [it] . . . has become an
universal.
The "idea," if that is the psychical state, is in
logic a symbol. But it is better to say, the idea
is the meaning, for existence and unessential
content are wholly discarded. The idea, in the
sense of mental image, is a sign of the idea in
the sense of meaning.
This is not to say that (a) there is an idea apart from
its being used, for its defining attribute is that it be
used as a sign. Neither are we saying (b) that we must
become aware of it as a mental "thing" in order to use
it, nor (c) that there even is an image present when
there is an idea. That is to say, it is questionable
whether we can, in every instance, verify the idea as a
psychic event.2
In Essays on Truth and Reality, published thirty
one years after the above opinions were expressed in
his Logic, we find Bradley still holding the same
i
opinions about ideas, but with the addition that no idea
merely "floats"— that is to say, there is no such thing
as an idea which is not entirely dependent upon reality
356
for its meaning. He credited this addition to the
insight of Bernard Bosanquet. The importance of this is
that ideas in themselves are not classifiable. In them
selves they are not general of specific, real or
imaginary, true or false, meaningful or meaningless. How
are they being used a^ ideas? is the crucial question.
This is interesting, for it is the essential message
supplied us by currently popular linguistic philosophy.
Ideas are not floating, disembodied characters.
This erroneous notion, Bradley says, is "based upon a
false assumption as to the limits of the real world."
What we call our "own world" or "real world" is a
construction made round the body during waking hours.
The body, unified through feeling and conceived in space
and time, is taken for actual 'fact', and all else is in
relation 'unreal'. From this basis dreams, imaginings,
and whatever else is contert to a psychic event, are
real only to the extent that they are connected with the
bodily basis. That connection does not always appear;
we are unaware how some of our ideas do qualify reality,
and so, conclude that they must be something other than
qualities of reality: they are floating, disembodied,
on their own.
But 'reality' so conceived fails to account for
past and future events, abstract and general ideas, and
other elements which we need to complete the meaning of
357
what we call "fact;" even from the bodily basis. Another
i
I
view of reality must be sought, so Bradley suggested the j
following: Every man's world is one world unified |
1
through feeling, although he may think of it as plural and!
may exclude from it the 'worlds' of other men. The unity
of one's world is experienced as actual. Within, or
superimposed upon, this unified felt whole are several
'worlds' thought of as independent, although incon
sistently. Such "worlds" are: the "outer facts of
sense" world; the "inner realm" of intellect; the realm
of intimate feelings and passing moods; the "worlds" of
the different professions, trades, and arts; and (lest
we forget) the realm of religious or moral duty. To this
collection of 'worlds' Bradley adds the regions of "hope,
desire, dream, madness and drunkenness and error, all
'•unreal' if you please"— but which nevertheless "pene
trate and on the other side transcend" the world of
sensation and normal bodily expression.
Now an idea that lives with a life of its own,
and so seems to float, stands in violation of Bradley's
understanding of an idea attached to some 'world' as a
meaning. There are no such floating ideas. They only
seem so in one 'world' because they are attached to
another 'world': they qualify another; they are
centered in another for their meaning. The meaning
continues to inhere although the attachment is neglected
or forgotten. Thus we find ourselves with 'false' ideas, |
'meaningless' ideas, merely imaginary ideas: ideas which
are self-contained because they are not taken as |
qualifying anything real. Because we force upon "real"
a limited meaning of our own, we reject, as unattached
and therefore self-contained, all ideas that stand
outside the circle of our 'real' world.^ Thus, we must
add to our statement of what an idea is, a fourth
proposition:
(4) The meaning (of the idea) is a function of
the 'world' (the universe of discourse) to which it
belongs.
These four propositions answer the question
"What is an idea?"
359
FOOTNOTES— APPENDIX A
■ * • 1 shall follow Bradley's example and not
emphasise here any difference between a sign and a
symbol. The difference is not crucial to ideas.
Apropos "idea," however, Bradley did add these words:
"A ‘sign1 or 'symbol' implies the recognition of its
individual existence, and this recognition is not implied
in an 'idea' [F. H. Bradley, The Principles Of Logic,
2nd ed., 2 vols (London: Oxford U. Pr., 1928!) 1:38,
note 8]."
^The material discussed in this paragraph,
together with the quotations supplied, will be found in
ibid., pp. 2-7, 38-9.
^See F. H. Bradley, Essays on Truth and Reality
(Oxford: Clarendon Pr., 1914), pp. 28-33.
360
APPENDIX B
WHAT IS A PACT?
We would not deny that the word "fact" has its
use, and perhaps linguistic philosophers can tell us
what it is. We might not go so far as to say the mere
word "bewitches us''1 . ' but we do sometimes notice our use
of "fact" has become too careless. Its established
meaning restricts it to matters capable of direct
verification, but unless we qualify its meaning in
philosophy we find it is tendentious. We are led to
suppose there are matters capable of direct verification,
in the sense that some highly privileged knowledge is
not based upon 'judgment. To speak of "factsi " as we
carelessly do, is to suggest a plurality of categorical
truths acceptable to all and beyond argument. But
philosophers like Bradley would not agree.
The word "fact" is overworked in writing
philosophy. Nothing is added to our serious assertions
by calling them "facts;"! And to assume that the objects
of our propositions are discrete self-existing "facts"
is to take a metaphysical position. The phrase "the
fact that”' . ! moreover, is stylistically objectionable.
It is a superfluous expression that we hide behind as a
361
screen; an evocative embellishment that casts an aura of
scientific indubitability over our discourse. "Fact" is
a convenient word and we all tend to retreat before
whomever invokes it. It will take time to break the habit
of using it, but we ought to try. We will find that it
is almost never needed. The opposition has compelled me
to use it, but I have tried to do so with caution. I have
carefully avoided "the fact that.’ " !
Russell strengthened our habitual appeal to
"fact" by the emphasis he put upon it, and upon the even
stronger "patent fact.;" I will discuss his position in
order to point out the difficulties into which it leads
him. The word "fact" seems almost ruinously ambiguous.
Russell gave no criterion for his "fact" or
"patent fact" because he thought that these are the very
sort.of things that carry their own "patent" or warrant
of originality and self-sufficiency. What could give
greater certitude than the proposition, "A paper is now
before me," to you as you read this paper? It seems that
nothing could. If nothing could, then what sense does it
make to ask for a criterion for even greater certitude?
The paper before you is a "patent fact” , , " and the use of
this term certifies that rock-bottom has been reached.
But in reply, for Bradley, we must insist that
the mere proposition, "a paper is now before me," is
certainly true only if special conditions are satisfied,
! which the proposition alone does not guarantee. In itself
j
it is not the "fact" because, in abstraction from the
conditions, it is compatible with falsity. Sometimes it
is true, sometimes not. It needs some outside support,
and that is what we mean by "criterion;" Adding words
like "this,"" "here,"" and "now" do not get those
conditions into the assertion of "fact;",! the proposition,
for a proposition is the objective aspect of judgment.
Looked at with the detachment of objectivity, any thing,
time, or place may be "this," "now,"" and "here;'.' The
proposition asserts of reality only if taken as some
particular "this’1 , ' "now/V and "here." This uniqueness
cannot be achieved except by the addition of feeling—
something that has no business in a proposition. Feeling
loses its uniqueness when taken objectively; that is,
when translated into a proposition. The "this," "now;1 , 1
and "here" become general (See Bradley, ETR, pp. 261-62).
Further, if "A paper is now before me" is to be our
paradigm of "fact," then "fact" is limited to the
sensuously present. Then what is to become of (a) his
torical "fact;" and (b) The conclusion of any inference?
(See Bradley, Logic 2:588). Both (a) and (b) are lost to
reality unless "fact" is prolonged into the realm of the
ideal. If (a) and (b) are lost, how much of the
original sense of "patent fact" are left to us? But if
(a) and (b) are included, then the sensuously present is
dependent and fragmentary. It alone is not the fact.
A criterion for "patent fact" is needed because
three questions need answers:
1. Is the "patent fact" independent of the
context that makes it a fact?
2. Is it exclusive of historical or inferential
"fact"?
3. Is it exclusive of feeling?
364
APPENDIX C
SAMENESS AND DIFFERENCE
We will discuss identity here, but we will not
exhaust the subject. To do the latter would leave
nothing out. To mention just one aspect of identity;
a cataloguing of the kinds of sameness that there are,
which differences fall within or outside of this or that
sameness, and where they do fall, would be to arrange in
a hierarchy all the categories of being. Such an under
taking is beyond the scope of this essay.
But we will try to say what identity is. One may
have wondered whether, in this essay, we meant to deny
such a thing as identity. If every change of relation
changes every related term, what becomes of identity?
Qeesn't identity mean "sameness through change of
relation?"
No, not if we mean "identity" in any sense that
may be possessed by a finite thing. Finite things change
relation. We search for identity, through the process of
change, and our search carries us beyond thS finite
thing.1 It carries us to universals, which endure, and
so are more real— but then, because they are plural,
and are replaced one by another, we must seek something
more enduring still. We seek the highest universal of
365
all, which includes the changing "thing" in its
meaning; but, because it is unchanging, includes the
"thing" also in its reality. This is why the search for
identity carries us to transcendent Reality, or "the
Absolute . "
But we mean nothing of that sort by "identity".
We impute identity to finite things, such as this red
pen. But then, in doing so, we wish to refer to its
uniqueness. This red pen does not have uniqueness,
except in relation to everything else in the universe.
The universe is unique, and so is the finite thing, the
red pen, taken related within it. Taken in itself the
pen is not unique. If we do not look to the whole
universe, which includes this red pen— but merely look at
the red pen— then it isn't unique. Possibly, in reality,
this isn't the same red pen with which we started. We
don't know the difference, and so the actual identity of
the red pen makes no difference to us. Only in the whole
reality does the identity make a difference— that is,
stand out as unique. To us "red" and "pen" are uni-
versals, and don't make the unique. "This," as a uni
versal, is the vaguest of all, so it doesn't make the
pen unique. But we feel the red pen is unique; we take
it here, now, in its actual setting. But in feeling we
have abandoned the finite and limited. The lived-in
whole, taken through feeling, stretches out to the
universe. Again we have referred to the whole reality,
for the uniqueness of our red pen.
But the identity had by this red marks it off
from the whole reality. It is an identity, to which the
rest is difference. j
But still, the universe is referred to in the
identity of our red pen. To say that it has identity
!
amid differences is not to say that its identity is just j
!
another difference among the differences. No. The I
identity of the red pen is the subject to which all those |
o i
differences are predicated. What are all those dif
ferences, as differences, if not differences to it? If |
the differences which the red pen has within the whole |
i
marks off the red pen from the whole, then the dif- !
ferences are differences to it. Then the identity of the j
red pen is taken for the center of the universe, to j
whose whole meaning the universe is referred by relation, j
and for whose unique existence the whole universe I
supplies the conditions. When we say, "This red pen is j
i
what it is, and so, has its identity," we must exclude j
from the universe anything else having the same predi--
i
cates. "Sameness" takes in the universe. j
We have denied identity to external relations.
Its elements are isolated universals whose identity is
not in them, for they are not unique, apart or 'together'.
Their "together" has not been accounted for; the uni-
367
verse has not been referred to.. The "together" is, as
Moore said, a "mere matter of fact"'.' The terms and
relation differ, but the "of" of which they are dif
ferences is missing. They are wandering adjectives in
search of a substantive.^ The "this" of actual
presentation is separated from the "what" of the
adjectives? the external relation is nothing real. The
4
separation between "this" and "what" makes "appearance '
368
FOOTNOTES— APPENDIX C
*F. H. Bradley, Appearance and Reality, 2nd
(1893; rpt. Oxford, Clarendon Pr., 1966), p. 526.
3Ibid., pp. 62-63.
3Ibid., pp. 525-33.
4Ibid., p. 526.
APPENDIX D
WAS BRADLEY'S VIEW INCOMPATIBLE
I
WITH SCIENCE? |
Morris R. Cohen speaks favorably of Bradley, but
qualifies his praise by adding, j
l
i
Too lightly does he dismiss science as valid and j
useful but incompetent to put us in contact I
with reality— thus helping to ppread the j
superstition that science is but a practical ;
device or a useful fiction.^
We need not challenge this statement. Perhaps Bradley
did have the effect Cohen mentions, and if so, perhaps
that is regrettable. Science certainly has told us many
things we otherwise would not have known. But there j
seems little danger that Bradley's view, even if regret
table, will become so popular that it will pose any
threat.
We will maintain, here, that Bradley's view,
whatever it may have led many to believe, is not in
compatible with science. Bradley himself did not think j
j
that his view was incompatible with science. The j
|
contrary is true. Russell had brought that charge
against Bradley, but he replied, "The doctrine which I
hold I hold largely because it seems to me to remain,
i
more than others, in harmony with life as a whole." j
| He added that Russell's view took "facts" to be invi
olable but argued that such an opinion "would make
indefensible the constructions in and by which the entire
2
body of history and natural science consists."
Bradley spoke of the "constructions" of science.
This view, both of scientific "facts" and scientific
"laws," is more in harmony with informed scientific
i
opinion than is the view that both "facts" and "laws" are
inviolable, or, , , as we now say, "incorrigible*"Fact/ "
as observed, is some "this" which is also a "what*" If
we say truly that we observe an object fall with an j
increasing velocity, universals are involved in our
statement; past experience lends meaning to it. These
transcend the sensuous presentation and justify calling j
our statement a "construction." The finding of a j
regularity among such observed "facts" leads to an even !
more general formulation which we call a scientific "law."j
But further scientific inquiry reveals that the "law" j
itself is conditioned,3 and an even more general and i
comprehensive formulation is called for. As has been j
well stated, j
Every law— and regardless of the form in which it
is stated— is but a hypothesis because it tells
us, not what actually happens, but only what will
happen if certain conditions are fulfilled [my
empha s is].4
Further inquiry reveals the effect of further conditions,
prompting the search for ever more sophisticated formulae.j
371
This advancing progress shows no sign of coming to an
end, and tends to confirm Bradley's view of reality.
Indeed, the relativity cf "facts" and "laws" to
the conditions under which observations are made, is
widely recognized— so much so that to say "Everything is
relative" sounds trite. The old immovable concepts which j
i
I
we associate with Newtonian physics, such as "time",
"energy", and "mass", have been shown all to be relative.
Equations expressing their relation and interdependence
I
have been formulated and cleverly 'verified'. "Facts" |
I
must be observed, of course, in the verification of any j
i
experiment— but the "facts" prove that the relativity is
true. The "facts" are constructions; highly conditioned
manifestations of a more comprehensive and unified whole. !
With the words "comprehensive and unified whole"
i
we are brought up to Bradley's conception of reality: j
!
the Reality which "includes and on every side transcends"
our bodies, with their "facts.". Is that reality—
metaphysical reality— incompatible with science? No.
Max Planck tells us, j
Metaphysical reality does not stand spatially
behind what is given in experience, but lies fully
within it. The essential point is that the world
of sensation is not the only world which may
conceivably exist, but that there is still another
world. To be sure, this other world is not directly
accessible to us, hut its existence is indicated,
. . . by the labor of science . . . , the j
scientist's assumption of the actual existence of |
a "real world", in the absolute sense of the word, j
eventually grows into a firm conviction which
nothing can shake any more.5
Then Bradley's view of reality is consistent with
informed scientific opinion.
FOOTNOTES— APPENDIX D
•^Morris R. Cohen, A Preface to Logic (Cleveland,
Ohio: World Pub. Co., 1965), p. 207.
2F. h. Bradley, Essays on Truth and Reality
(1914; rpt. London: Oxford U. Prd., 1968), p. 291.
•^Boyle's Law, for example, which we learned in
Basic Physics, turns out to be true only for a "perfect
gas:" It is not absolutely true for an actual gas.
Actual gases change state, or perform unpredictably (by
Boyle's Law) at exotic pressures and temperatures:
Boyle's Law must be augmented. That Law, so comforting
to novices in their early search for scientific 'truth',
turns out to be true only under "normal conditions' . " . 1
It is, as Bradley would say, a construction made round
the body; that is, a 'truth' true only within the narrow
range of temperatures and pressures normally encountered.
4W. H. Werkmeister, A Philosophy of Science
(Lincoln, Nebr.: U. of Nebr. Pr., 1940), p. 462.
^Max Planck, Scientific Autobiography and Other
Papers, tr. F. Gaynor (New York: Philosophical Library,
1999), p. 102.
374
APP.ENDIX E
IS PHILOSOPHY'S TASK DESCRIPTIVE
The conception of philosophy's task as merely
descriptive, and the de-emphasis of its enabling role for
understanding, stems from the current philosophical
interest in language. Moore's contribution to it has
been discussed. Wittgenstein carried it further.
According to him, thinking and understanding (along with
remembering, intending, and feeling) were private and
obscure processes hidden behind linguistic behavior.
Bradley disagreed. He thought philosophy's task was "to
say what we mean and mean what we say"— implying that
there was more in the "meaning" than in the mere "saying."
He meant to include his notion of "intelligibility;" the
feeling of satisfaction as thought passes "from each and
all to the One, and from the One to each and all . . .
without . . . a jar or break in its continuity."'*'
Wittgenstein argued that all this involves secret
processes, hidden from verification and public view. It
is not statable. So far as we can talk about under
standing, it is linguistic behavior, and nothing more.
So far as it is open to examination, communicable, and
shared, "understanding" is a way of behaving under cer-
375
tain circumstances. The "secret process" theory of
understanding is beset by this difficulty: How do we know
what to look for, when we look for it? How do we know
that there is one thing in two heads called "under
standing"?
Linguistic behavior is descriptive of the public
world. Philosophy is to describe that behavior. How can
it do anything else? We have argued, on Bradley's be
half, that description is only justifiable if it serves
understanding. But if the description needs justification
in understanding, what justifies the understanding? It
is secret, hidden. As Wittgenstein put it,
We are trying to get hold of the mental process
of understanding which seems to be hidden behind
those coarser and therefore more readily visible
accompaniments. But we do not succeed; . . . For
even supposing I had found something that happened
in all those cases of understanding, why should it
be the understanding? And how can the process of
understanding have been hidden, when I said, "Now I
understand" because I understood?! And if I say it
is hidden— then how do I know what I have to look
for? I am in a muddle.2
And lower down.
In the sense in which there are processes (in
cluding mental processes) which are characteristic
of understanding, understanding is not a mental
process.^
In short, understanding consists in the behavior, which
is open and publicly describable. In the case of
philosophy, understanding (typically, "knowing how to go
on") consists of linguistic behavior. Philosophy cannot
376
be an attempt to seize and hold a secret process called
"understanding;"!
Then what sort of thing ought philosophers to do?
They ought to describe linguistic behavior, Wittgenstein
said,
Philosophy may in no way interfere with the actual
use of language; it can in the end only describe
it. For it cannot give it any foundation either.
It leaves everything as it is.4
He previously had said,
We must do away with all explanation, and
description alone must take its place. And this
description gets its light, that is to say its
purpose— from the philosophical problems. These
are, of course, not empirical problems; they are
solved, rather, by looking into the workings of
our language and that in such a way as to make us
recognize those workings: in despite of an urge to
misunderstand them. The problems are solved, not
in giving new information, but in arranging what
we have always known. Philosophy is a battle
against the bewitchment of our intelligence by
means of language.5
The purpose of philosophical description is got from the
philosophical problems. These are not empirical problems.
Then what are they like? "A philosophical problem has
g
the form: 'I don't know my way about',", he said.
Lower down, he described more exactly how a purpose is got
from a philosophical problem:
It throws light on our concept of meaning something.
For in those cases things turn out otherwise than we
had meant, foreseen. That is just what we say when,
for example, a contradiction appears: "I didn't mean
it like that."
The civil status of a contradiction or its status
in civil life: there is a philosophical problem.7
377
In summary, Wittgenstein was saying that (a) one
does not, ink philosophy, seek explanation or understanding!
as a secret or superior process. To "understand" can
mean nothing more than linguistic behavior. Rather, in
philosophy one seeks to (b) describe problems, not solve
them. Philosophical problems are but instances of
linguistic confusion, and to 'solve' such problems is to
describe more clearly what one meant to say. An apparent
contradiction, for example, can be 'solved' descriptively
when the speaker accused of uttering a contradiction goes
back and describes what he meant? how he was using the
words found objectionable. Thus philosophy is
descriptive, not explanatory. It does not lead to a
higher process called "understanding.! ' It simply makes
problems disappear by describing linguistic behavior.
We must attempt a reply to Wittgenstein. Bradley
did not deny that description, in the sense of explaining
what was meant by certain words, was useful. But just
why that excludes "explanation" is not clear. Why such
description cannot convey something worth calling "under
standing," from the speaker to the hearer, is not clear
either. If it involves a "secret process," then so do
other processes which Wittgenstein seemed to think were
describable, such as that other people had failed to
'see' what we meant.
We must be careful about accusing Wittgenstein of
378
uttering a contradiction, because he wrote in German, not
in English. Whether or not he thought it was possible
to translate his meaning into English is a question we
cannot answer. But according to our translation, he
said "... in despite of an urge to misunderstand them."
In the German it reads, "... entgegen einem Trifeb, es
misszuverstehen." "Misunderstand" seems correct here.
Then apparently, according to him, in the absence of an
adequate description there ie something called "mis
understanding", yet even with adequate description we, in
philosophy, cannot get "understandingi'" That seems
puzzling. Have we misunderstood him?
Suppose someone utters a contradictory statement,
and seeks to extricate himself "descriptively','',' i.e. by
the rather unremarkable process of explaining what he
meant. Was it not because the speaker understood his own
position, but 'saw' that the hearer had not, that the
speaker sought to arouse, through other words (by
"description") a similar understanding in the hearer?
People ordinarily think so, and the word "understand"
has that function in the "grammar" of their "language
game.“" The goal is mutual understanding. The description
only leads to it (in the hearer), having been based upon
it (in the speaker). The understanding iis not the
description. There, is the plain difference here between
means and ends. The means for understanding is not the
379
understanding. In pursuit of understanding there was
verbal behavior, but we cannot conclude that the under
standing is nothing but verbal behavior. We would
destroy the 'grammar' of the word "describe" if we
thought that this were so.
In saying these things we are not "giving new
information", we are just "arranging what we have always
known."
In order to "describe" what is actually going on
in moving from description to understanding, it is a
misuse of words to say that we "see" that our hearer has
misunderstood. We know that he did misunderstand, but we
don't 'see' it, we find it out in another way. Perhaps
what we actually observe of the hearer's misunderstanding
is his behavior— but not his linguistic behavior. Then
what word best describes our 'seeing' of another's
misunderstanding? C. S. Peirce used the word "abduction"
to describe our ability to grasp an explanation. In
addition to induction and deduction, the mind also has a
power of "abduction". What is abduction like? It is like
saying to ourselves, somehow, "If A were true, C would
be a matter of course." A stands for an explanatory
g
hypothesis; it accounts for the behavior C. Abduction
is not a form of seeing, it is a form of inference.
Bradley did not say that understanding is a
hidden process, nor that it was superior in the sense of
380
being exclusive of other behavior. It is not "exclusive
of" or "other than/" it is "inclusive of" or "more than,'''!
The difference here is that of a whole which is more than
the parts taken abstractly. Moreover, it would be
foolish to use words suggesting that, to Bradley, under
standing takes place somewhere that is spatially
locatable, such as in a special chamber of the brain.
If we must use spatially suggestive words, it is better
to say that understanding is the "reaching" or "grasping"
of reality; not retreating from it, hiding behind it, or
withdrawing to some chamber within it.
The difficulty with philosophy conceived as
merely descriptive lies not with what it affirms, but
with what it denies. It denies understanding. If
philosophy's task were merely descriptive, then the denial
of the "more, " implicit in the "mere,"", would destroy
description itself* Perhaps this has already become
clear. But if not, recall what Wittgenstein himself
said: a philosophical problem takes the form, "I don't
know my way about." Description seeks to provide the
missing knowledge. It seeks to fill in the detail; to
give the circumstances. That is to say, by describing
the 'grammar' of the word in question, we try to 'show'
how it fits into the language. It is fair to ask
linguistic philosophers, however, who recommend this
sort of thing, not to contravene their own rule. We must
381
stick to the 'grammar' of the 'game'. Very well.
"Describe" is one word in the language game; "under
stand" is another. The word "describe" has been much
used by linguistic philosophers, but "understand" has not
found favor. But both of these words have their "natural
home" in the grammar of the language.game, and that
grammar, by their own rule, ought to be left intact. In
the actual situation where "describe" finds its normal
use (its grammar) there is also to be found a describer,
an object described, a hearer of the description, and a
purpose in the describing. "Purposep y we have seen,
found favor with Wittgenstein. If philosophical
description is to give us a picture, then these are also
parts of the picture: i.e. describer, object, hearer,
and purpose. Now somewhere in that picture the word
"understand" finds its grammar. It finds that grammar
in "purpose of the description^" All these elements:
describer, object, hearer, purpose, and understanding, fit
together to give us a picture of the grammar of the word
"describe•"
If "describe" is abstracted from any of these,
then its "grammar" (the use which it had in its natural
home in ordinary language) has been mutilated. The
philosopher who ignores these other elements has been
misled (or has deliberately misled) by a mere word,
viciously abstracted. But there is no problem of vicious
382
abstraction, in ordinary language, with "describe , " for
among its other circumstances it is at home with its
purpose. Its purpose is to secure understanding. In the
presence of "understand*' ! "describe" knows its way about.
Wittgenstein told us,
— but for us it is the circumstances under which
he had such an experience that justify him in
saying that in such a case that he understands,
that he knows how to go on.^
Why should it be different with "describe"? Is it not,
after all, the circumstances which justify (or render
questionable) the use of "describe"? Surely so. And on
the other hand, is it not knowledge of the circumstances
which justifies the use of "understand" the way
philosophers have used it? Does that use occur in a
vacuum? No. Philosophy is itself a "form of life
arising out of a genuine and common need. That form of
life "describes," if you will, the grammar of "under
stand'^ which is its goal.
383
FOOTNOTES— CHAPTER E
■^As explained in Chapter I, Section H.
2Ludwig Wittgenstein, Philosophical Investi-
gations, tr. G. Anscoxnbe (New York: MacMillan Co.,
1962), p. 60.
3Ibid., p. 61.
4Ibid., p. 49.
3Ibid., p. 47.
, 6Ibid., p. 49.
^Ibid., p. 50.
8 See Philosophical Writings of Peirce, ed.
Justus Buchler (1940; rpt. New York: Dover Pub. Inc.
1955), p. 151.
9Wittgenstein, op. cit., p. 61.
APPENDIX F
SUBJECTIVITY OR OBJECTIVITY?
It seems that Bradley denied that we can know the
"thing" as object, just as it is in itself. The object
is always conditioned through subjective form. Then how
can one be "objective"? And, that being impossible, it
is but one short step further to the conclusion that no
"knowledge" worthy of the name is possible. Or, in other
words, Bradley was insisting upon subjective idealism.
Such an appraisal of Bradley's view is too
simple, and it is wrong. Bradley was an "idealist" only
in the sense that he refused to accept any concept that
was empty of experienced content. That is no more than
to say he was a very cautious thinker. No one who is
acquainted with his critique of solipsism [Appearance and
Reality (Oxford: Clarendon Press, 1897), pp. 218-30]
would call him a "subjective idealist' It ij3 possible,
on his view, for experience to transcend the 'self1.
He did not assume that the 'self' is as wide as
experience [See ibid., p. 464]. He did not apply the
term "idealist" to his own view. On the question of
whether his philosophy was to be called realist or
idealist, he said, "I do not know and I have not cared to
385
inquire [ibid., p. 484]." His own view, he said,
"neither puts thoughts or ideas first, nor again does it
permit us to assert that anything else by itself is more
real Iibid]."
But surely "objectivity" demands that we should
insist upon putting the object first, ahead of our
thoughts and ideas formed about it. We are to detach
ourselves from subjectivity for the sake of truth.
Objectivity, we have been taught, is a paramount
intellectual virtue, fundamental to science and every
other serious study. How is it to be practiced if we
reject.in principle the very possibility of putting the
object first?
The answer to our question is that there is no
such objectivity. There never was nor can be. As a
modern writer puts it, "It is a very 'old hat' notion of
science that thinks of it as a detached and disinterested
inquiry into predetermined and self-subsistent
structures [see T. F. Torrance, God and Rationality
(London: Oxford Univ. Pr., 1971), pp. 6-7]." In modern
science, the criterion of perceptibility (and even
predictability) has been found to be very limited in
applicability— i.e. applicable only within the narrow
range where we are solely concerned with what we presume
to be nature's determinate aspects. Even here, it is to
abstractions from nature that we apply this kind of
386
"objective" inquiry and explanation. In other words, we j
i
can be "objective" only within that narrow range of life
where we are determined in advance to ignore subjective
forms. To borrow again from Torrance (words which he in
turn has borrowed from M. Polanyi), the notion of
"objectivity" as "detachment", " . . » is not the
sign of rationality but of open-mouthed imbecilityt
[ibid., p. 8]I" Or, to put it more politely, to approach
a scientific problem with an attitude of "detachment"
is to act upon the presumption that one's basic
epistemological presumptions are true beyond question.
This kind of "objectivity" is, paradoxically, actually
a kind of closed-minded dogmatism which forbids challenge
to its preconceptions.
True "objectivity" is an attitude of willingness
to subject one's basic conceptions to critical review.
387
APPENDIX G
ON THE SUPPOSED DIFFERENCE BETWEEN
LOGICAL AND CAUSAL SEQUENCE
We in philosophy have all leanred from the first,
almost as though it were an axiom, that the cause-and-
effect relation is to be carefully distinguished from the
logical antecedent-consequent relation. "Necessity"
was restricted to the latter. It held, we were told,
only among our ideas, or our propositions, because their
relationship was through a unity which we had given them
(put of habit, perhaps), and so their relationship did
not hold of the 'real world'. The logical antecedent-
consequent relationship was tautological, in the sense
that it held only among our meanings. But in the real
world, it was not so: the flaming of a match, (b), could
not be inferred from the striking of its head, (a), in the
way that we can infer that A is an angle, given that it is
a right angle.
But the above distinction is indefensible. The
radical difference between a causal sequence and a
logical sequence rests upon the supposal that the real
world consists of real 'events', while the unreal world
(our mere ideas, or our verbal propositions), is a mere
abstraction from the real world. But an 'event' is also
388
an abstraction from the real world. The whole reality is
not cut up into discrete events, as we take them, and
when we inquire into the principle which we employ when
we regard events separately, we shall solve the mystery
of how it is possible to reunite them.
Mere (a) does not produce (b), even if we allow
ourselves to suppose that (a)— (b) is two events and not
one. (a)-— (b), related as cause to effect, are truly so
only within an unspecified background. As Hume truly
said,
The mind can never possibly find the effect, in
the supposed cause, by the most accurate serutiny
and observation.!
Certainly not, for "the mind” has already broken the
original unity of these abstract "events,''", and sheared
off, to discard, the relevant details! But if we have
the background disregarded, the motivation ignored, the
whole situation broken into "events'," with the details
sheared off, need we wonder that (a) is one "thing" and
(b) is another? It is ourselves who wielded the hammer,
the saw, and the scissors. But now (a) and (b) have
become conditional and abstract. They are not real.
They are but ideas, with meanings, and if we wish to ask
whether they are together ar apart, we must make that
meaning full and complete. We must try to retrieve as
much as we can of the original situation, with its
detail. But to deal with reality as ideas, with
meanings, connected together through mutuality of
meaning shared, is inference.^
The simplified relation (a)— (b) is ideal, their
separation is ideal, and so is their reunion which
justifies our speaking of a "causal" sequence. But
reality is what reality is. It is ourselves who
abstracted a "causal" series, and then, from that,
abstracted an (a) and a (b)— to express our astonishment
that (a) was (a) while (b) was (b). To see that the
relation between (a) and (b) is necessary, we need but
reverse our wheels.
390
FOOTNOTES— APPENDIX G
*David Hume, Hume Selections, ed. C. W. Hempel
(New York: Charles Scribner's Sons, 1955), p. 120.
2See Bradley, AR, pp. 192-93, Logic, 2:536-40.
391
WORKS CITED
Austin, John Langshaw. Sense and Sensibilia. London:
Oxford University Press, 1964.
Bergmann, Gustav. The Metaphysics of Logical Positivism.
New York: Longmans, Green & Co., 1954.
...... Logic and Reality. Madison: University of
Wisconsin Press, 1964.
Bergson, Henri. Matter and Memory. New York: The
Macmillan Co., 1913.
Blanshard, Brand. The Nature of Thought. 2 vols.
London: George Allen & Unwin, 1939.
Reason and Analysis. La Salle, 111.: Open
Court Publishing Co., 1964.
Bradley, Francis H. The Principles of Logic. 2nd ed.,
2 vols. London: Oxford University Press, 1928.
' • _____. Appearance and Reality. 2nd ed. Oxford:
Clarendon Press, 1930.
________ . Essays on Truth and Reality. Oxford:
Clarendon Press, 1914.
________ . Collected Essays. 2 vols. Edited by
H. H. Joachim. Oxford: Clarendon Press, 1935;
reprint ed., Freeport, N. Y.: Books for Libraries
Press, 1968.
Charlesworth, Maxwell J. Philosophy and Linguistic
Analysis. Duquesne Studies. Philosophical Series
vol. 9. Pittsburgh: Duquesne University Press,
1961.
Cohen, Morris R. A Preface to Logic. Cleveland:
World Publishing Co., 1944; Meridian Books, 1956.
Collingwood, R. 3. An Essay on Metaphysics. Oxford:
Clarendon Press, 1940.
Gram, M. S. "The Reality of Relations." New
Scholasticism 44 (Winter 1970): 49-68.
392
Hume, David. An Enquiry Concerning Human Understanding, j
Chicago: Henry Regnery Co., Gateway ed., 1956.
________ Hume Selections. Edited by C. W. Hendel, Jr., !
New York: Charles Scribner's Sons, 1927? reprint
ed., 1955.
James, William. The Principles of Psychology. 2 vols.
New York: Dover Publications, 1890.
Klemke, E. D., Ed. Studies in the Philosophy of
G. E. Moore. Chicago: Quadrangle Books, 1969.
Malcolm, Norman. Knowledge and Certainty. Englewood
Cliffs, N. J.: Prentice-Hall, Inc., 1963.
Merleau-Ponty, Maurice. Phenomenology of Perception.
Translated by Colin Smith. London: Routledge and
Kegan Paul, Ltd., 1962.
________. The Structure of Behavior. Translated by
Alden L. Fisher. Boston: Beacon Press, 1963;
Beacon Paperback, 1967.
Miller, Clarence. "Moore, Dewey, and the Problem of the
Given." Modem Schoolman 39 (1961-62): 379-82.
Moore, George Edward. Principia Ethica. Cambridge:
Cambridge University Press, 1903; reprint ed. 1922.
________. Philosophical Studies. Edited by C. K. Ogden.
London: Routledge and Kegan Paul, Ltd., 1922;
reprint ed. Totowa, N. J.: Littlefield, Adams &
Co., 1968.
________. Some Main Problems of Philosophy. New York:
The Macmillan Co., 1959? Collier Books, 1962.
________. Philosophical Papers. New York: The Macmillan
Co., 1959; Collier Books, 1962.
Muirhead, J. H. Ed. Contemporary British Philosophy.
2 vols. London: 1925; rpt. Allen & Unwin, 1956.
Peirce, Charles Sanders. Philosophical Writings of
Peirce. Edited by Justus BuOnler. London:
Routledge and Kegan Paul Ltd., 1940? reprint ed.,
New York: Dover Publications, Inc., 1955.
Planck, Max. Scientific Autobiography and Other Papers.
Translated by F. Gaynor. New York: Philosophical
Library, 1949.
393
]
Randall, John Herman, Jr. The Making of the Modern Mind. |
Cambridge: Houghton Mifflin Co., 1954. j
"P. H. Bradley and the Working-Out of Absolute !
Idealism." Journal' of the History of Philosophy,
5 (July 1967): 245-67.
Ryle, Gilbert, ed. The Revolution in Philosophy.
London: Macmillan & Co., 1956; reprint 1960.
Russell, Bertrand. The Principles of Mathematics.
New York: W. W. Norton & Co., n.d.
The Problems of Philosophy. New York: Oxford
University Press, 1912; reprint ed. Galaxy Books,
1959.
• Our Knowledge of the External World : As A
Field for Scientific Method in Philosophy. Revised
ed. London: Allen & Unwin Ltd., 1926; reprint
ed. 1969.
_________ My Philosophical Development. New York:
Simon & Schuster, 1959.
The Autobiqgraphy_ of Bertrand Russell.
New York: Little, Brown & Co. reprint ed., New York:
Bantam Books, 1967.
Saxena, Sushil Kumar. Studies in the Metaphysics of
F. H. Bradley. London: Allen & Unwin, 1967.
Sinclair, W. Angus. The Conditions of Knowing. New York:
Harcourt, Brace & Co., 1951.
Taylor, A. E. ”F. H. Bradley.” Mind XXXIV (1925): 1-12.
Elements of Metaphysics. London: Methuen &
Co., 1903; University Paperbacks, 1961.
Torrance, Thomas F. God and Rationality. London:
Oxford University Press, 1971.
Werkmeister, W. H. An Introduction to Critical Thinking.
Rev. ed. Lincoln, Nebr. Johnsen Publishing Co.,
1957.
Wittgenstein, Ludwig. Philosophical Investigations.
Translated by G. Anscombe. New York: Macmillan
Co., 1962.
394
Wittgenstein, Ludwig. Tractatus Logic-philosophicus
New York: Harcourt, Brace & Co., 1922.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Le "Journal" De Maine De Biran. (French Text)

PDF

The Concept Of Sincerity In John Oman'S Thought

PDF

The influence of Kant's moral argument on three British personal idealists: A. S. Pringle-Pattison, W. R. Sorley, C. C. J. Webb

PDF

The Role Of Relativity In Berkeley'S Philosophy

PDF

John And The Synoptics--A Discussion Of Some Of The Differences Between Them

PDF

Friends In California--A Study Of The Effect Of Nineteenth Century Revivalism Upon Western Quakerism

PDF

A William Temple Word-Book: A Comprehensive Philosophical And Theologicalindex To His Major Published Writings

PDF

A Critical Examination Of Heidegger'S And Jasper'S Interpretations Of Nietzsche

PDF

The meaning of judicium and its relation to illumination in the philosophical dialogues of augustine

PDF

Rhetoricians On Language And Meaning: An Ordinary Language Philosophy Critique

PDF

Early Analytic Philosophy: The history of an illusion

PDF

Benoit-Constant Coquelin: The Art Of A Rhetorical Actor

PDF

Prescriptive Deontic Logic: A Study Of Inferences From Linguistic Forms Expressing Choice And Conditional Permission And Obligation

PDF

The Concept Of "Presence" In Selected Theories Of Rhetoric

PDF

California Land Grant Disputes, 1852-1872: A Rhetorical Analysis

PDF

The Word Within The Word: A Literary Examination Of Lancelot Andrewes' Presentation Of The Life Of Christ

PDF

A Critical Study Of Contemporary Aesthetic Theories And Precepts Contributing To An Aesthetic Of Oral Interpretation

PDF

Illusion And Illocution: J. L. Austin'S "Sense And Sensibilia."

PDF

The Effects Of Variant Forms Of Competition On The Cognitive Learning Outcome Of An Educational Game Played By College Students In English Classes

PDF

Kant'S Doctrine Of Existence As A Predicate

Asset Metadata

Creator Curtis, Carl Harold (author)

Core Title A Reexamination Of F. H. Bradley'S Critique Of Relations

Degree Doctor of Philosophy

Degree Program Philosophy

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag OAI-PMH Harvest,Philosophy

Language English

Contributor Digitized by ProQuest (provenance)

Advisor Willard, Dallas (committee chair), Fisher, Walter R. (committee member), MacGregor, Geddes (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c20-561970

Unique identifier UC11226226

Identifier 7605238.pdf (filename),usctheses-c20-561970 (legacy record id)

Legacy Identifier 7605238.pdf

Dmrecord 561970

Document Type Dissertation

Rights Curtis, Carl Harold

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA